ADVANCES IN
FOOD AND NUTRITION RESEARCH VOLUME 39
ADVISORY BOARD DOUGLAS ARCHER Gainesville. Florida
JESSE F. GREGO...
30 downloads
1041 Views
20MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
ADVANCES IN
FOOD AND NUTRITION RESEARCH VOLUME 39
ADVISORY BOARD DOUGLAS ARCHER Gainesville. Florida
JESSE F. GREGORY I11 Gainesville, Florida
SUSAN K. HARLANDER Minneapolis, Minnesota
DARYL B. LUND New Brunswick, New Jersey
BARBARA 0. SCHNEEMAN Davis, California
SERIES EDITORS GEORGE F. STEWART
(1948-1982)
EMIL M. MRAK
(1948-1987)
C. 0. CHICHESTER
(1959-1988)
BERNARD S. SCHWEIGERT (1984-1988) JOHN E. KINSELLA
(1989- 1993)
STEVE L. TAYLOR
(1995-
)
ADVANCES IN
FOOD AND NUTRITION RESEARCH VOLUME 39
Edited by
STEVE L. TAYLOR Department of Food Science and Technology University of Nebraska Lincoln, Nebraska
ACADEMIC PRESS San Diego
New York
Boston
London
Sydney Tokyo
Toronto
This book is printed on acid-free paper.
@
Copyright 0 1996 by ACADEMIC PRESS All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher.
Academic Press, Inc. 525 B Street, Suite 1900, San Diego, California 92101-4495, USA http://w.apnet.com
Academic Press Limited 24-28 Oval Road, London NW I 7DX, UK h t t p : / / w . hbuk.co.uk/ap/ International Standard Serial Number: 1043-4526 International Standard Book Number: 0-12-0 16439-6 PRINTED IN THE UNITED STATES OF AMERICA 96 97 9 8 9 9 00 O l Q W 9 8 7 6 5
4
3
2
1
CONTENTS
CONTRIBUTORS TO VOLUME 39
. . . .. .. .. .. . . . . . . . .. . .. .. . .. . .
ix
The Rheology of Semlliquid Foods
Gustavo V. Barbosa-Chnovas, Li Ma, Jozef L. Kokini, and Albert Ibarz
I. 11. 111. IV. V. VI. VII.
. . . . . . . . . . . . . . . . . . . . .. . . . .. . . . . . . . . . . . Introduction Basic Rheology Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . Experimental Methods . . . . . . . . . . . . . . . . . .. . . . . . . . . . . Instrumentation in Fundamental Rheology .. . . . . . . .. . Constitutive Models . . . . . . . .. . . . .. .. . . .. . . . . . .. . .. . Applications of Rheology in Characterizing Engineering Properties of Foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary and Research Needs . . . . . . . . . . . . . . . . . . . . . . References .......................................
2 3 12 23 33 44 61 63
Control of the Dehydration Process in Production of IntermedlateMoisture Meat Products: A Review S. F. Chang, A. M. Pearson, and T. C. Huang
I. 11. 111. 1v. V. VI. VII. VIII. IX.
Introduction ...................................... Traditional Production of IM Meat Products . . . . . . . . . . Technology of Producing IM Pet Foods . . . . . . . . . . . . . . Preservation Principles and Their Application to IM Meats ............................................ Problems in Production of Different IM Meats , . . . . . . . Effects of Slaughtering, Handling, Chilling, Freezing, Storage, and Thawing on Muscle Properties . . . . . . . . . . . Effects of Predrying Treatment and Handling of Muscle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Factors Influencing Absorption Phenomena in Meats/Meat Mixtures . . . . . .. . . .. . . . . . .. . . . .. . . .. . .. Mechanisms Involved in Meat Dehydration Systems . . .
73 74 83 85 88 90 94
97 100 V
vi
CONTENTS
X. Quality Attributes as Affected by Dehydration and Its Associated Processes ............................... XI. Process Optimization for IM Meats . . . . . . . . . . . . . . . . . . XI1. Energy Costs for Production of IM Meat Products ..... XI11. Research Needs for IM Meats ...................... XIV . Summary ......................................... References .......................................
114 138 143 147 151 152
Cheese: Physical. Biochemical. and Nutritional Aspects
P. F . Fox.T. P. OConnor. P . L . H. McSweeney. T . P . Guinee. and N . M. O’Brien I. I1. I11. IV . V. VI . VII . VIII . IX . X.
Introduction ...................................... Composition and Constituents of Milk ............... Conversion of Milk to Cheese Curd ................. Biochemistry of Cheese Ripening .................... Cheese Flavor ..................................... Cheese Texture ................................... Accelerated Cheese Ripening ....................... Processed Cheese Products ......................... Nutritional and Safety Aspects of Cheese . . . . . . . . . . . . . Perspective ....................................... References .......................................
164 168 169 195 235 254 255 259 277 292 296
Biogenic Amines in Fish and Shellfish
Dafne D . Rawles. George J . Flick. and Roy E. Martin I. I1. 111. IV . V. VI . VII . VIII . IX.
Introduction ...................................... Nonvolatile Amine Formation ...................... Amine Detoxification .............................. Bacterial Species with Decarboxylase Activity ......... Amines Presence in the Marine Ecosystem . . . . . . . . . . . Amines Occurrence in Seafood ...................... Scombrotoxicosis .................................. Amine Formation as an Indicator of Freshness in Seafoods ......................................... Recommended Limits of Amine Content .............
330 330 330 331 336 336 346 348 351
CONTENTS
X . Determination of Biogenic Amines in Fish ........... References .......................................
INDEX
...................................................
vii 353 358 367
This Page Intentionally Left Blank
CONTRIBUTORS
Numbers in parentheses indicate the pages on which the authors’ contributions begin.
Gustavo V. Barbosa-CBnovas, Biological Systems and Engineering, Washington State University, Pullman, Washington 99164 (1)
S. F. Chang, Department of Food Science and Technology, Pingtung Agricultural Institute, Pingtung, Taiwan (71) George J. Flick, Department of Food Science and Technology, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 (329)
P. F. Fox, Department of Food Chemistry, University College, Cork, Ireland (163) T. P. Guinee, National Dairy Products Research Center, Teagasc, Moorepark, Fermoy, Cork, Ireland (163) T. C. Huang, Department of Food Science and Technology, Pingtung Agricultural Institute, Pingtung, Taiwan (71) Albert Ibarz, Food Technology, University of Lleida, 25006 Lleida, Spain (1) Jozef L. Kokini, Center for Advanced Foods, The State University of New Jersey, New Brunswick, New Jersey 08903 (1) Li Ma, Biological Systems and Engineering, Washington State University, Pullman, Washington 99164 (1)
Roy E. Martin, National Fisheries Institute, Arlington, Virginia 22209 (329) P. L. H. McSweeney, Department of Food Chemistry, University College, Cork, Ireland (163)
N. M. O’Brien, Department of Food Chemistry, University College, Cork, Ireland (163) ix
X
CONTRIBUTORS
T. P. O’Connor, Department of Food Chemistry, University College, Cork, Ireland (163) A. M. Pearson, Department of Animal Sciences, Oregon State University, Corvallis, Oregon 97331 (71) Dafne D. Rawles, Department of Food Science and Technology, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 (329)
ADVANCES IN FOOD AND NUTRITION RESEARCH, VOL. 39
THE RHEOLOGY OF SEMILIQUID FOODS GUSTAVO V. BARBOSA-CANOVAS Biological Systems Engineering Washington State University Pullman, Washington 99164-6120
JOZEF L. KOKINI Center for Advanced Foods The State University of New Jersey New Brunswick, New Jersey 08903
LI MA Biological Systenzs Engineering Washington Slate University Pullman, Washington 99164-6120
ALBERT IBARZ Food Technology University of Lleida 25006 Lleida, Spain
I. Introduction 11. Basic Rheology Concepts A. Viscous Flow B. Ideal Elastic Behavior (Hookean Body) C. Time Effects D. Viscoelasticity 111. Experimental Methods A. Steady Shear Measurement B. Extensional Flow
1 Copyright 0 1996 by Academic Press. Inc. Ail rights of reproduction in any form reserved.
2
GUSTAVO
IV.
V.
VI.
VII.
v. BARBOSA-CANOVAS
ef
C. Transient Flow D. Linear Viscoelasticity Instrumentation in Fundamental Rheology A, Capillary Tube Geometry B. Cone-Plate Geometry C. Plate-Plate Geometry D. Concentric Cylinder Geometry Constitutive Models A. Rheological Model for Steady Shear Flow B. Dilute Solution Molecular Theories C. Concentrated Solution/Melt Theories D. Solid Foods Applications of Rheology in Characterizing Engineering Properties of Foods A. Steady Shear Viscosity of Fluid Foods and Dilute Food Polymer Solutions B. Characterization of Entanglement in Concentrated Solution C. Application of Constitutive Equations in Dough Rheology D. Rheology of Emulsions E. Extensional Flow F. Formulations Development Summary and Research Needs References
I.
INTRODUCTION
Food rheology is the study of the manner in which food materials respond to an applied stress or strain. The science of rheology has many applications in the fields of food acceptability, food processing, and food handling. Determination of rheological properties of foodstuffs provides an instrumental quality control of raw material prior to processing, of intermediate products during manufacturing, and of finished goods after production. Determination of rheological properties of foodstuffs is also useful in elucidation of structure and composition of food and analysis of structural changes during processing (Escher, 1983). The rheological properties of foodstuffs are important in food process engineering since they are the essential parameters in plant design (pumping requirements, pipe and valve dimensions, and mixing operations, etc.) and in the calculationof basic heat, mass, and momentum transfer (Szczesniak, 1977). Instrumental quality control before, during, and after manufacture is one area to which food rheology makes important contributions. For example, the measurement of apparent viscosity and yield stress of ketchup helps to predict how well tomato ketchup drains from a bottle. A number of tests have been developed using either basic rheological instruments (rotational viscometer, capillary viscometers, etc.) or instruments simulating the situation in which the rheological properties are of importance (Bostwick
THE RHEOLOGY OF SEMILIQUID FOODS
3
Consistometer) (Gould, 1983; Harper, 1960; Rao et al., 1981; Rao and Cooley, 1983; Rani and Bains, 1987). Rheological quality control measurements are also used to follow changes in foods during processing and storage. For instance, in manufacturing an apple nectar beverage by enzymatic breakdown of apple pulp, Struebi et al. (1978) measured flow curves after various lengths of treatment. An optimum suspension stability and consumer-determined quality of the final beverage can be achieved by monitoring viscosity in the pulp liquefying process. The chemical analysis of changes in pectic substances and other cell wall material of the apple tissue was much more difficult and did not yield a conclusive picture as to how far these substances had to be enzymatically altered for optimum quality. In sensory evaluation consumers estimate fruit firmness on the basis of the deformation resulting from physical pressure applied by the hand and fingers. The toughness or tenderness of meat is subjectiveIy evaluated in terms of the effort required for the teeth to penetrate and masticate the flesh tissues. Therefore, determination of rheological properties of foodstuffs is important in evaluation of consumer-determined quality by correlating rheological measurements with sensory tests. Foods, however, are complex materials structurally and rheologically. In many instances they consist of mixtures of solid as well as fluid structural components, e.g., solid cellwall material, water and colloidal liquids, and intercellular gases. Many foodstuffs are neither homogeneous nor isotropic and have properties which vary from one point to another within their mass. All these complicating factors make the study of food rheology more difficult than the study of rheology in other fields such as polymer rheology. Nevertheless, investigators report that many foods do behave in a predictable manner and that concepts from elasticity, viscosity, and viscoelasticity theories can be used to interpret their response to applied deformation or applied force. In this review, the basic food rheology concepts such as shear rate, shear stress, viscosity, elasticity, and viscoelasticity are introduced; the experimental methods for steady shear; extensional flow, and dynamic measurements are described; the principles of instrumentation are discussed;constitutive models based on steady shear flow, diluted solution molecular theories, and concentrate solutions are presented; and applications of rheology in characterizing engineering properties of foods are reviewed. II.
BASIC RHEOLOGY CONCEPTS
Food rheology is concerned with the description of the mechanical properties of food materials under various deformation conditions. To under-
4
GUSTAVO V. BARBOSA-CANOVAS et al.
stand and use rheological information it is essential to have a mechanistic basis for the interpretation and correlation of the experimental data. Such interpretation of behavior in terms of theoretical mechanisms provides the guidelines needed to make sense of observations, to relate behavior to composition and structure, to predict and to modify properties, and to compare one experimental method with another. Under external force, the food materials exhibit the ability to flow, or accumulate recoverable deformations, or both. According to the extent of recoverable deformation, the basic rheology concepts can be classified into viscous flow, elastic deformation, and viscoelasticity. A. VISCOUS FLOW
Viscosity is a measure of a fluid’s ability to resist motion when a shearing stress is applied. Considering a simple geometry (Fig. l), the upper plate is caused to move with a velocity (w) relative to the lower plate. This velocity is due to the application of a shearing force ( F ) per unit area. The layers of fluid contacting the plates are considered to move at the same velocities as the surface they contact; i.e., the assumption is made that no slipping occurs at the walls. The fluid then behaves as a series of parallel layers, or lamina, whose velocities are proportional to their distance from the lower plate. The differentiation of velocity with respect to the distance (dv/dy) is defined as shear rate
.
dv dY’
Y=--
For a class of fluid known as Newtonian fluid, there is a linear relationship between shear stress and shear rate. The dynamic viscosity (or coefficient of viscosity) is defined as the ratio of the shear stress to the shear rate
FIG. 1. Schematic diagram of flow induced by displacement of one of a pair of parallel planes.
THE RHEOLOGY OF SEMILIQUID FOODS
77''
7
Y
or 7
=
w,
(2b)
where q is called Newtonian viscosity, .jl is the shear strain rate or shear rate, and T is the shear stress. Curve a in Fig. 2 illustrates the relationship between the shear stress and shear rate of a Newtonian fluid. The plot is a straight line passing through the origin. The slope of the line is, by definition, the viscosity. However, most food materials are very complex (i.e., emulsions, suspension of solids, biopolymers such as carbohydrates, proteins, food gums, etc.), and their rheological behaviors do not obey Eq. (2). Instead, food materials exhibit either pseudoplastic (shear thinning) or dilatant (shear thickening) behavior. In these situations the viscosity is no longer constant, but dependent on the shear rate. The shear rate-dependent viscosity is called apparent viscosity (ya). Curves b and c in Fig. 2 show the relationships between shear stress and shear rate for the shear thickening and shear thinning fluid behaviors, respectively. Compared to Newtonian fluid behavior (curve a), they either exhibit an upward or downward curvature. The typical equation to characterize the shear thickening and shear thinning fluid is the power law Shear stress
t
Shear rate FIG. 2. Rheograms for time-independent fluids.
6
GUSTAVO
v. BARBOSA-CANOVAS el ai. T =
Kj"
or
where K is the consistence index (Pa * sec"), n is flow index, and qa is apparent viscosity (Pa sec). When n > 1, the plot for shear stress-shear rate will be an upward curvature, which represents the shear thickening fluid; when 0 < n < 1, the plot for the shear stress-shear rate will be a downward curvature, which represents the shear thinning fluid. When n = 1, Eq. (3) is reduced to Eq. (2) and represents the behavior of a Newtonian fluid. Some other food materials exhibit a yield stress which may be defined as a minimum shear stress required to initiate flow. Although the existence of yield stress in fluids is still a controversial topic (Barnes and Walters, 1985; Cheng, 1986; Hartnett and Hu, 1989; Steffe, 1992a; Evans, 1992; Schurz, 1992), there is little doubt that yield stress is an engineering reality (Hartnett and Hu, 1989)which may strongly influence process calculations. Yield stress imparts stability to food emulsions in low-stress situations (e.g., during storage and transportation, where the stress involved is usually lower than the yield stress). The most common model for characterizing nonNewtonian fluids with yield stress is the Herschel-Bulkley equation
or
where T is the shear stress, i, is the shear rate, n is the flow index, 70 is the yield stress, K is the consistency index, and qa is apparent viscosity. , non-Newtonian behavior can be Based on the magnitude of n and T ~ the classified as shear thinning, shear thickening, Bingham plastic, pseudoplastic with yield stress, or dilatant with yield stress (see Fig. 2 and Table I). The Herschel-Bulkley model is able to describe the general flow properties of fluid foods within a certain shear range. The discussion on this classification and examples of food materials has been reviewed by Sherman (1970), DeMan (1976), Barbosa-Cinovas and Peleg (1983), and Barbosa-Cgnovas et al. (1993).
7
THE RHEOLOGY OF SEMILIQUID FOODS TABLE I CLASSIFICATION OF NEWTONIAN AND NON-NEWTONIAN FLUIDS
Newtonian Non-Newtonian Shear thinning (pseudoplastic) Shear thickening (dilatant) Bingham Pseudoplastic with yield stress Dilatant with yield stress
70
n
K
0
1
K = q
0 0 >O >O
O l 1 O l
>O
>O >O
>O >O >O
B. IDEAL ELASTIC BEHAVIOR (HOOKEAN BODY) An ideal elastic body (also called Hooke’s body) is defined as a material that deforms reversibly and for which the strain is proportional to the stress, with recovery to the original volume and shape occurring immediately upon release of the stress. In a Hooke body, stress is directly proportional to strain, as illustrated in Fig. 3. The relationship is known as Hooke’s law, and the behavior is referred to as Hookean behavior. Based on Hooke’s law, the following relationships have been established for elastic, homogeneous, and isotropic materials under tensile or compressive stresses (see Fig. 4a)
where E is the elastic modulus (or Young’s modulus, Pa), D is the tensile or compressive stress (Pa), and E is the tensile or compressive strain [ E = (L’ - L)/L, dimensionless].
Suaill
FIG. 3. Ideal elastic behavior.
8
GUSTAVO V. BARBOSA-CANOVAS et al.
FIG. 4. Hooke’s elasticity in (a) tension and (b) shear.
When the Hooke solid is subjected to distortion by shear stresses (Fig. 4b), the shear modulus or modulus of rigidity is given by
where G is shear modulus (Pa), strain (= AWL, dimensionless).
T
is the shear stress (Pa), and y is shear
C. TIME EFFECTS Time-dependent rheological properties reflect the nature of a system’s structure and can be due to viscoelasticity, structural changes, or both (Cheng and Evans, 1965; Harris, 1972). Structure breakdown can result in a decrease in the viscosity of a substance. It occurs in emulsions, suspensions, and sols. The characterization of the time-dependent flow properties of food systems is important for process design and control, for product devel-
THE RHEOLOGY OF SEMILIQUID FOODS
9
opment, to establish relationships between structure and flow, and to correlate physical parameters with sensory evaluation. Most foods exhibit time-dependent rheological properties (Rha, 1978; Shama and Sherman, 1966; Tiu and Boger, 1974; Mitchell, 1979; Bloksma, 1972; Peleg, 1977; Dickie and Kokini, 1982). The classical approach to characterize structural breakdown is the measurement of the hysteresis loop, first reported by Green and Weltmann (1943). A sample is sheared at a continuously increasing, then continuously decreasing shear rate, and a shear stress-shear rate flow curve is plotted. If structural breakdown occurs, the two curves do not coincide, creating a hysteresis loop. The area enclosed by the loop indicates that degree of breakdown. The major disadvantage of this method is that the investigator arbitrarily sets up the time of shearing and the maximum shear rate (Mewis, 1979). Thus, comparison of data from different investigations becomes difficult. D. VISCOELASTICITY The term “viscoelastic” means the simultaneous existence of viscous and elastic properties in a material (Barnes et al., 1989). Some food materials (i.e., food gels) have a unique configuration that will return to the original structure after releasing the deforming stresses. However, the viscous resistance to deformation makes itself felt by delaying the response of the material to a change in stress. The mechanical assemblies called Maxwell and Kelvin models illustrate this point (Fig. 5). These one-dimensional mechanical models consist of springs and dashpots arranged, in parallel
Maxwell model
Kelvin-Voigt model
FIG. 5. Schematic diagram of Maxwell and Kelvin-Voigt models.
10
GUSTAVO V. BARBOSA-CANOVAS et 01.
or in series so that the overall system behaves analogously to a real material, although elements themselves may have no direct analogues in the actual material. The correspondence between the behavior of a model and a real material can be achieved if the differential equation relating force, extension, and time for the model is the same as that relating stress, strain, and time for the material. However, due to the complex composition (i.e., foods contain numerous compounds) and structure of food materials, the use of models such as the Maxwell and the Kelvin-Voigt model and their combination is valid only when the experimental data are obtained within the linear viscoelastic range. Linear viscoelasticity is the simplest viscoelastic behavior in which the ratio of stress to strain is a function of time alone and not of the strain or stress magnitude. Under a sufficiently small strain, the molecular structure will be practically unaffected, and linear viscoelastic behavior will be observed. At this sufficiently small strain (within the linear range), a general equation that describes all types of linear viscoelastic behavior can be developed by using the Boltzmann superposition principle (Dealy and Wissbrun, 1990). For a sufficiently small strain (-yo) in the experiment, the relaxation modulus is given by
Consider a sequence of small shear strains as shown in Fig. 6. The shear stress resulting from the strain that occurs at time f1 will be
Assuming the incremental response of the material to this second step strain is independent of the strain introduced at time tl, the stress resulting from the strain at time t2 can be simply added on as
For any combination of N small strains, the contributions to the stress can continue to be added on, and, in general N
7(t) =
X G(t - ti)6y(ti) i=l
(t > t ~ ) .
For a smooth strain history not consisting of finite steps, the following expression can be obtained by use of the definition of the integral
THE RHEOLOGY OF SEMILIQUID FOODS Sb
11
L
FIG. 6. The diagram of a sequence of step strains. The resulted stress can be superposed
if all strain is within the linear viscoelastic range [see Eqs. (8-10); after Dealy and Wissbrun, 19901.
s’, G(t - t’)dy(t’).
7(t) =
(11)
Noting that the strain that occurs during the time interval dt’ is simply r(t’)dt’,this can also be written as
7(t) =
s’, G(t - t’)r(t’)dt’.
(12)
The use of the lower limit of minus infinity is a mathematical convenience; it implies that to calculate the stress at time t, in the most general case, one must know the strain history infinitely far into the past, i.e., at all times t’ prior to time t. In practice this is not necessary. In general, an experiment is started at some time ( t = 0) when the material is in a stress-free state. In this case, ~ ( 0 )= 0, and 7(t) =
si G(t
-
t’)dy(t’).
(13)
The above equations only apply to shearing deformations, but they can be generalized for any type of deformation by using another feature of linear viscoelasticity. The relaxation process is independent not only of the
12
GUSTAVO V. BARBOSA-CANOVAS et al.
magnitude of the strain, but also of the type (kinematics) of the deformation. Thus, the shear strain can be replaced by the strain tensor for infinitesimal strain, and the shear stress can be replaced by the stress tensor to obtain the following alternative form of the Boltzmann superposition principle ~ i , ( t )=
$-, G(t - t’)dyij(t’).
(14)
Using the Boltzmann superposition principle, it is possible to calculate the stress components resulting from any types of deformation, as long as the deformations are sufficiently small or slow so that linear behavior is exhibited.
Ill. EXPERIMENTAL METHODS
The rheological concepts such as viscosity, elasticity, and viscoelasticity have just been reviewed. To determine the material functions by carefully defined experimental methods is important in food rheology. In general, two types of tests have been used to characterize rheological properties of food materials: fundamental and empirical tests. Empirical tests have been developed from practical experience as a rapid way to measure something that is related to textural quality. These tests are usually arbitrary, poorly defined, have no absolute standard, and are effective for only a limited number of foods (Bourne, 1994). In contrast, fundamental tests are conducted on a given material by imposing a well-defined stress and measuring the resulting strain or strain rate or by imposing a well-defined strain or strain rate and by measuring the stress developed. The simple shear flow and extensional flow are most common approaches in the fundamental rheological tests (Bird et al., 1987). In this section various tests such as steady shear measurements, extensional flows, and linear viscoelasticity used to characterize the rheological properties of food materials are introduced. The mathematical constraints and simplifications associated with the experimental methods for rheological measurements are discussed. A. STEADY SHEAR MEASUREMENT Consider the flow shown in Fig. 1, where a fluid between two plates is sheared as the top plate moves with velocity V, in the x direction. The
THE RHEOLOGY OF SEMILIQUID FOODS
13
velocity gradient or shear rate is given by i.X, = -dVx/dy = y, and macroscopically is given by VJS, where S is the plate separation. The stresses generated by the flow act both parallel to the direction of shear (i.e., shear stress) and perpendicular to the direction of shear (normal stress Pyyin Fig. 7). The experimentally observable stress perpendicular to the direction of flow includes the stresses arising from fluid motion and the isotropic hydrostatic pressure. It is customary to eliminate the isotropic pressure by taking the difference between normal stresses. The material functions for steady shear flow are defined as (1) viscosity 17 =
(16)
-.yx/i;x
(2) primary normal stress coefficient *l
= -
Pxx
-
PYY
Y&
(3) secondary normal stress coefficient
where
T~~
(= Pyx)is shear stress (Pa), Pxx- PyY (= N 1 ) is primary normal
FIG. 7. Force components acting on a cube shaped volume element.
14
GUSTAVO V. BARBOSA-CANOVAS et al.
stress difference (Pa), and Pyy - PZz (= N2) is secondary normal stress difference (Pa). For a Newtonian fluid, the apparent viscosity 7 is a constant, and the normal stress differences Nl and N2 are zero for all. For non-Newtonian fluids, these material functions vary with shear rate.
B. EXTENSIONAL FLOW The extensional flow is a deformation that involves stretching along streamlines. According to the resulting deformation, it can be classified as uniaxial, biaxial, or planar extension. 1.
Uniaxial Extension
The uniaxial extension is illustrated in Fig. 8. If the x1 axis is aligned with the principal stretching direction, it can be noted that two fluid particles located on this axis will get further apart as the deformation proceeds. It is convenient to use a coordinate system in which the axes are oriented in the directions of the principal strain axes. For this choice, the strain rate components always have the general form
where ul, a2, a3 have units of reciprocal time and al + a2 + u3 = 0 for an incompressible material. The uniaxial extensional flow, with regard to describing both the deformation and the resulting stresses, is uniform shear free flow, in which the strain rate is the same for every material element, and there is no relative
ffl + T p ,......... ................
..... .......'
before extension
............. ....c' .............. ...... ........__ ..... .... after extension
FIG. 8. Effect of simple extension on an initially cubical fluid element.
THE RHEOLOGY OF SEMILIQUID FOODS
15
rotation of perpendicular axes fixed in a fluid element. It is convenient to use the principal strain directions as the coordinate axes, and, in this case, the rate of strain tensor has only diagonal components, as shown in Eq. (19). The shear free flow that has been most used in experimentalrheology is uniaxial extension, also called simple extension, which is an axisymmetric flow with stretching in the direction of the axis of symmetry. The effect of uniaxial extension on the shape of an initially cubic fluid element is sketched in Fig. 8. In the simple extension, the velocity distribution is given by
since the flow is axisymmetric, it can also be conveniently described using cylindrical coordinates
v, =
EZ
1. v, = ---Er 2 ve = 0, where 6, the Hencky strain rate, is defined as
.
de dt
&=-=--=-
1 dL L dt
d(ln L ) dt *
The components of the rate of deformation tensor are given as
It is also noted that the principal stresses for extensional flow are in the same direction as those for the strain rate. Thus, the stress tensor also has only diagonal components
16
GUSTAVO V. BARBOSA-CANOVAS et al.
For incompressible fluids, only the differences between normal stress components have a rheological significance, and the “net tensile stress,” in uniaxial extension, is defined as CrE
3
U]1
- u22 =
u11
- u33.
125)
If h is independent of time, the flow is steady simple extension, and since this is a motion with constant stretch history, it should be associated with a material function in which times does not appear as an independent variable. For an axially symmetric flow, the stress is also symmetric with a22
= a 3 3 = urr.
(26)
Thus, there is one independent normal stress difference, and a material function having the units of viscosity can be defined as
This property is called the “extensional viscosity.”
2. Biaxial Extension Another axisymmetric extensional flow is biaxial extension. In this deformation, there is a compression along the axis of symmetry that stretches in the radial direction, as shown in Fig. 9. The principal strain rate is defined as
before extension
after extension
FIG. 9. Biaxial extension.
THE RHEOLOGY OF SEMILIQUID FOODS
17
For a circular disk having a radius R, this “biaxial strain rate” would be
.
1 dR R dt
Eb=--=-
d(ln R ) dt *
Furthermore, when this flow is looked upon as biaxial stretching, it is thought of as being generated by a radial, tensile stress rather than an axial, compressive stress. Thus, the “biaxial extensional viscosity” is defined as
C. TRANSIENT FLOW One of the characteristics of the viscoelastic foods is that when a shear rate is suddenly imposed on them, the shear stress displays an overshoot and eventually reaches a steady state value. Campanella and Peleg (1987c), Kokini and Dickie (1981), and Dickie and Kokini (1982) presented stress overshoot data for different foods. Figure 10 illustrates stress overshoot as a function of shear rate. The stress overshoot data can be modeled by means of an equation which contains rheological parameters related to the stress (normal and shear) and shear rate. One such equation is that of Leider and Bird (1974)
where ulzis the shear stress, K and IZ are power law parameters relating the shear stress and the shear rate, $ is the imposed sudden shear rate, t is the time, h is a time constant, and a and b are adjustable parameters. A number of models have been developed to describe transient viscoelastic behavior and one must have at hand carefully obtained rheological data in order to test the applicability of the models. Another example of the applicability of models to viscoelastic data is the study of Leppard and Christiansen (1975) in which the models proposed by Bogue and Chen, Carreau, and Spriggs were evaluated. In the case of foods, the empirical models have been developed to describe the transient data on stick butter, tub margarine (Mason er d.,1982), canned frosting (Kokini and Dickie, 1981; Dickie and Kokini, 1982), and mayonnaise (Campanella and Peleg, 1987~).
18
GUSTAVO V. BARBOSA-CANOVAS et al. 4.5
0
10
20
30
Time (sec) FIG. 10. Shear stress development for peanut butter at 25°C and comparison of the prediction of the Bird-Leider equation with experimental data (Kokini, 1992).
D. LINEAR VISCOELASTICITY For a number of viscoelastic materials, a linear viscoelastic response can be achieved experimentally if the deforming stresses are kept sufficiently small such that recovery occurs upon unloading. There are a number of tests which may be used to study viscoelastic materials to determine the relationships between stress, strain, and time for a given type of deformation and a given type of loading pattern. The most important tests include creep compliance and recovery test, stress relaxation, and the dynamic oscillatory test. As shown in the thin slab shown in Fig. 1, different deformation patterns can be applied on the upper plate. Assume that the slab is so thin that inertial effects can be ignored and the slab can be regarded as homogeneously deformed with the amount of shear y(t) variable in time. Let 7(t) be the shearing stress, the force per unit area on the slab. The
19
THE RHEOLOGY OF SEMILIQUID FOODS
following three simple shearing motions illustrate the three experimental methods of stress relaxation, creep compliance, and sinusoidal shear deformation. 1. Stress Relaxation Test
When the slab is subject to a single-step shear history y(t) = yoH(t), where H(t) is the Heaviside unit step function, zero for negative t and unit for t zero or positive, the stress response can be used to characterize the rheological properties. When the materials are subjected to a step strain as shown in Fig. l l a , the different stress responses are obtained as shown in Fig. l l b . If the material were perfectly elastic, the corresponding stress history would be of the form 7(t) = T ~ H ( ~constant ), for t positive (curve a in Fig. llb). If the material were an ideal viscous fluid, the stress would be instantaneously infinite during the step and then zero for all positive t, like a Dirac delta, 8(t) = H’(t) (curve b in Fig. llb). For most real materials, like semisolid foods, the stress response shows that neither of these idealizations is quite accurate. The stress usually decreases from its initial value
a
I
P
t
0
b
‘t
0
FIG. 11. Stress relaxation of different types of materials in response to the shear strain.
20
GUSTAVO
v. BARBOSA-CANOVAS et UL
quite rapidly at first and later more gradually, as it approaches some limiting value <m). The limiting value is a subjective matter since no one can wait that long, but it is a convenient idea. If the limiting value is not zero, it is likely to call the material a solid (curve c in Fig. llb), and if the limiting value approaches zero rapidly, the material is then called a fluid (curve d in Fig. llb).
2. Creep Test Supposing the upper slab subjected to a one-step stress history is ~ ( t=) ~ , $ f ( t )(Fig. 12a), the response of strain can be used to characterize the creep/recovery behavior of food materials. The creep/recovery response may be classified into several different categories as shown in Figs. 12b to 12e. For a perfectly elastic solid, compliance rises instantaneously to the equilibrium value. When the stress is released, there is an instantaneous recovery (Fig. 12b). All the energy is stored in the solid and there is no energy dissipation. For a viscous fluid, flow occurs in response to the applied stress. As a result, the compliance increases linearly with time with a slope of 1/77,where 77 is the viscosity. The input energy is totally dissipated due to the motion of the liquid and there is no energy stored. When the stress is released, the compliance does not decrease (since there is no energy release), but stays constant at the final value (Fig. 12c). The response of viscoelastic materials lies between these two extremes. When a constant stress is applied, there is an instantaneous rise in compliance. The compliance then increases with time to the equilibrium value. When the stress is released, there is an instantaneous drop in the compliance followed by a time-dependent decrease. For a viscoelastic “solid,” all the energy is stored, and hence there is a total energy release upon removal of the shear stress. As a result, the final equilibrium compliance is zero (Fig. 12d). For a viscoelastic “liquid,” however, viscoelastic flow takes place and there is only a partial recovery when the stress is removed (Fig. 12e). 3. Sinusoidal Oscillatory Shearing
The properties of viscoelastic materials can also be described in terms of the responses to sinusoidal inputs. In a sinusoidal shear test, the applied strain or stress to the sample is sinusoidal and, in general, the response of the stress or strain is dependent on both shear frequency and the rate of shear strain (Fig. 13). A convenient means of manipulating oscillatory quantities is in terms of their complex equivalents, which is based upon the Euler identity
21
THE RHEOLOGY OF SEMILIQUID FOODS
't
'k- . ............
(c)
t
g
........................
n
Y
*
t
f
Je
FIG. 12. Creep recovery response of different types of material to the shear stress.
&e =
cos(8)
+ isin(@),
where i = fl(imaginary). Hence, oscillatory shear strain and stress become y(wt) = yOeiot
(33) (34)
22
GUSTAVO V. BARBOSA-CANOVAS et al.
nmr FIG. 13. Small amplitude oscillatory test.
where y(wt) is the oscillatory strain input in terms of complex notation, is the amplitude of strain, o is the shear frequency, dot) is the stress output in terms of complex notation, T~ is the amplitude of stress, and 6 is the phase angle between strain and stress. During oscillation, the complex shear modulus, G*,is defined as the ratio of oscillatory stress to oscillatory strain
Substituting Eqs. (33) and (34) into Eq. (35), one obtains
It is customary to write the complex modulus G* in terms of real and imaginary parts G* = G’
+ iG”.
(37)
Combining Eqs. (36) and (37), one obtains G’ = 7o/yo cos(6)
(38)
G = TO/Y~ sin(8).
(39)
The “in-phase” component, G’, represents the elastic character of the material and hence is called the storage modulus (since elastic energy is stored
THE RHEOLOGY OF SEMILIQUID FOODS
23
and can be recovered). The “out-of-phase’’ component, C“, represents the viscous character and is called the loss modulus (since the viscous energy is dissipated or lost). The tangent of the phase angle 6 is sometimes called the loss tangent, since tan(@ = G’YG’.
(40)
In the linear viscoelastic range, various other material functions relate to one another (Ferry, 1980)
where 7’ is the dynamic viscosity, q” is the out-of-phase component of the complex viscosity, and q* is the complex viscosity. Normal stress is an “extra” stress which is developed in viscoelastic material under shear, in directions normal to the plane of shear. The normal stress is expressed by the first and second normal stress coefficients (Bird et al., 1987) 711
*dY> =
-
k(Y)=
-
-
722
Y2 722
i.’- 733
(44)
(45)
is the first normal stress coefficient, and &(?) is the second where normal stress coefficient.
IV. INSTRUMENTATION IN FUNDAMENTAL RHEOLOGY
The above-discussed material functions (v,G’, G , E’, E”, q f , q”,tan(6), etc.) are important concepts to characterize the flow and viscoelastic properties of food materials. Furthermore, a rheometer or relevant instrument is an essential tool to determine and evaluate these material functions. One of the most important components of the rheometer is the measuring geometry, because it provides a particular way in deforming the tested material.
24
GUSTAVO V. BARBOSA-CANOVAS et al.
The common configurations of measuring geometries are capillary, coneplate, plate-plate, and concentric cylinder. The following is a review of working equations associated with each geometry and its limitations.
A. CAPILLARY TUBE GEOMETRY When a fluid is pumped through a certain length of tube, a pressure drop is observed due to the viscous drag effect of the fluid. This pressure drop is a function of the geometric size of the tube (inner diameter and length of tube) and the flow rate. Thus, this relationship is used to determine the viscosity of a fluid. Figure 14 shows a schematic diagram of a capillary tube. For Newtonian fluids, the viscosity, shear rate, and shear stress at the wall can be determined by the following relationships (Bird et al., 1960; Sherman, 1970)
nAPR4
v=xGjTR
=
APR 2L
constant force or constant rate
I /
piston
removable capilary
/
L
+i." FIG. 14. Schematic diagram of a capillary geometry (Bird er al., 1987).
(47)
THE RHEOLOGY OF SEMILIQUID FOODS
25
where R is the radius of the capillary, Q is the flow rate, L is the length of the capillary, and AP is the pressure difference between the two ends of the capillary. For non-Newtonian fluids, the working equations are modified as follows (Barnes et al., 1989)
(49) The apparent viscosity for non-Newtonian fluids can then be obtained by combining Eqs. (47) and (49)
where qo is the apparent viscosity. When a non-Newtonian fluid in a capillary tube is subjected to a small amplitude oscillatory pressure gradient, the viscoelastic properties of the fluid can be determined (Darby, 1976).The explicit equations for determining such material functions as 77’ and 7’’ are fairly complicated. Equations (51) and (52) are the implicit expressions
v’’= -81 nR4X(w), where R is the radius of the capillary, R(w) and X ( w ) are the solutions of the following governing differential equations [Eqs. (53-55)J and boundary conditions [Eq. (56)]
ap az
- --
P a t -
1a r ar
-I- - -(rTr2)
(54)
26
GUSTAVO V. BARBOSA-CANOVASet a/.
v, = 0 at dV,/dt = 0 where @ ( dis)the oscillatory pressure input with complex argument, $ is the strain rate, and 77" is the complex viscosity. However, the above equations [Eqs. (46-56)] are only valid under certain assumptions. These assumptions include: (1) the fluid is incompressible; (2) the fluid velocity is zero at the wall (no slippage at the wall); (3) the normal stress is isotropic; and (4) a unique function of y = f ( ~ relates ) the rate of shear (q) to shear stress ( T ) . Not all of these assumptions are met by a capillary rheometer. In the capillary geometry, the source of errors can be: (1) entrance (end) effects; (2) kinetic energy imparted to a sample; (3) wall effect; (4) turbulent flow; and ( 5 ) plug flow (Sherman, 1970). The entrance effect in capillary flow is due to an abrupt change in the velocity profile and shear distribution when the material is forced from a large diameter reservoir into the capillary tube. However, this effect can be eliminated by using a long entrance region and determining the pressure drop as the difference of two pressure values measured in the fully developed laminar flow region (Kokini, 1992). It is assumed that there is no slip at the boundaries in derivation of equations to calculate shear rates from capillary and rotational viscometer data. However, many foods form a thin layer of liquid at solid boundaries and this in turn results in deviation from the no-slip boundary condition. The errors due to slippage at walls were studied by Kokini and Plutchok (1987a) in capillary viscometers and by Higgs (1974) in capillary and concentric cylinder viscometers in accordance with the pioneering theoretical study of Mooney (1931). In the capillary viscometer, the slip phenomena can also be observed due to inhomogenous flow (Cohen and Metzner, 1986; de Vargas et al., 1993).Therefore, the influence of the slip phenomenon must be considered when analyzing the data. In the case of 0.2% xanthan gum solution, the slip velocity is an increasing function of the wall shear stress and also of the length to diameter ratio WD. However, the slip velocity becomes independent of W D at large W D (de Vargas et al., 1993).
THE RHEOLOGY OF SEMILIQUID FOODS
27
B. CONE-PLATE GEOMETRY Cone-plate geometry consists of a flat plate and a rotating cone forming a very small angle with the plate (Fig. 15a). The angle (Oo) between the cone and plate is usually between 0.5" and 3". The food material to be tested is sheared between the rotating cone and the fixed plate. The following expressions are used to determine the steady shear rheological parameters (Bird et al., 1960, 1987)
y=-
n sin( Oo)
7 =
3T 2TR3 sinij5 - Oo)
2F
A(Y) = nR2j2
FIG. 15. Schematic diagram of cone-plate geometry
(59)
28
GUSTAVO
v. BARBOSA-CANOVAS el ~ i .
where R is the radius of the plate, 0, is the angle between the cone and plate, Q is the angular velocity of the cone, T is the measured torque, F is the force required to keep the tip of the cone in contact with the circular plate, roo is the measured pressure by a flush-mounted pressure transducer located on the plate, ,)I is the first normal stress coefficient, and & is the second normal stress coefficient. Since 0, is very small (ca. 0.5-3", or 0.0087-0.0523 radians), sin2 ( d 2 0,) will close to one, and will be nearly independent of position [see Eq. (59)]; that is, the tested material between the gap will experience uniform shear stress. This is the advantage of cone-plate compared to other geometries (i.e., capillary tube and parallel disk). For example, at an angle of lo, the percentage difference in shear stress between cone and plate is 0.1218% (Fredrickson, 1964).This is within the precision of measurements that must be made; therefore, one can assume that shear stress, and, hence, shear rate and apparent viscosity, are uniform throughout the fluid. When the upper cone oscillates within a small deformation, the viscoelastic properties are calculated by the following governing equations (Bird et al., 1987) =
30,T0 sin(@ 2T~3wy,
where yo is the deformation, To is the amplitude of the oscillating torque, S is the phase angle, w is the oscillating shear frequency, 0, is the angle between cone and plate (usually less than 4'), and R is the radius of the cone. Using the relationships in Eqs. (41-43), the other linear viscoelastic material functions, such as G', G', and tan(s), can be obtained. In practice, the cone is often truncated by a small amount which avoids contact with the cone tip (which might become worn) and the plate (which might become indented). Figure 15b shows the schematic diagram of a truncated cone-plate geometry. A truncated cone also facilitates tests on suspensions (Barnes et al., 1989). If R1 < 0.2R, the torque is reduced by less than 1%.The total torque is reduced by much less than 1%because the parallel plate section near the axis will contribute to the torque (Whorlow, 1980). For cone-plate geometry, the major errors are the edge and end effects which arise from the fact that the geometry has finite dimensions and a fracturing effect (Walters, 1975).
THE RHEOLOGY OF SEMILIQUID FOODS
29
C. PLATE-PLATE GEOMETRY The plate-plate geometry consists of two disks (Fig. 16). The food material to be tested is placed between two parallel plates. During the measurement, one plate is stationary while the other is rotating or oscillating, depending on the measuring mode. Compared to cone-plate geometry, plate-plate geometry has the advantage of a flexible gap which allows for a wider measuring range and more applications. The disadvantage is that the distribution of shear stress in the test material is not uniform. In rotational mode, the following expressions can be used to determine the apparent viscosity and the normal stress from the measurement of the torque on the fixed bottom plate (Bird et al., 1987)
.
YR =
RR
H
where R is the radius of the plate, H is the separation of the plates, R is
FIG. 16. Schematic diagram of plate-plate geometry.
30
GUSTAVO V. BARBOSA-ChOVAS et al.
the angular velocity of the upper plate, t,hl and & are the first and second normal stress coefficients, respectively, T is the torque required to rotate the upper plate, yR is the shear rate at the edge of the plate, F is the force required to keep the separation of the two plates constant, Po is the atmospheric pressure, and &(R) is the normal pressure measured on the disk at the rim. When the upper plate oscillates sinusoidallywith frequency o and angular amplitude yo (yo < 1), the time-dependent torque (T) is measured. The following expression is used to determine 7’ and V” from the measurements with yo 4 1 (Bird et al., 1987)
where T and yo are the measured torque and angular amplitude, respectively. The phase angle, 6, is the measured phase shift (0 5 S Id2). The limitation of plate-plate geometry is that the shear rate has to be below 500 sec-’ (Connelly and Greener, 1985). The major sources of error associated with the plate-plate viscometer may be surface fracture, radial migration, and wall slippage. The advantage of plate-plate geometry is that it provides flexibility for the material such as coarse dispersions, which are intolerant of the narrow gaps associated with either cone-plate or concentric cylinder rheometers. The plate separation provides a simple means to extend the shear rate range and to test for sample slip (Shoemaker et al., 1987; Yoshimura and Prud’homme, 1988). Specimen slippage has been previously reported. Navickas and Bagley (1983) detected slip between gels formed by wheat starch granules and the parallel plates of a Rheometrics Model KMS-71 mechanical spectrometer. Slippage has also been detected in the measurement of flowability of molten cheese by capillary rheology (Konstance and Holsinger, 1992), in the measurement of apple sauce by concentric cylinder rheometer (Qiu and Rao, 1989,1990), and in the measurement of mayonnaise with parallel plates of a Physica rheometer (Ma and Barbosa-Citnovas, 1995a). Several methods have been employed to overcome the slippage problem during rheological measurement. One is to use a cyanoacrylate ester adhesive to attach the gel to the plates (Konstance and Holsinger, 1992).
THE RHEOLOGY OF SEMILIQUID FOODS
31
Another approach to prevent slippage is to use an upper plate with 80-pm teeth (Rosenberg et al., 1995).
D. CONCENTRIC CYLINDER GEOMETRY The basic features of concentric cylinder geometry are shown in Fig. 17. Depending on whether the inner cylinder or outer cylinder rotates, the concentric cylinder geometry is referred to as Couette or Searle type. In a Searle type (Fig. 17a), the inner cylinder rotates, while in a Couette type (Fig. 17b), the outer cylinder rotates. For both types, the fluid to be tested is sheared in the gap between the cylinders. The torque necessary to maintain the position of the fixed cylinder is a measure of the shear stress after certain corrections have been made, and the rotational speed of the moving cylinder is the measure of the rate of shear. Equations (70) through (72) represent the working equations for both types of concentric cylinder geometry (Couette and Searle type) with the assumptions: (1) steady laminar, isothermal flow; and, (2) negligible gravity and end effect.
-
-
FIG. 17. Schematic diagram of concentric cylinder geometry.
32
GUSTAVO V. BARBOSA-CANOVAS et al. 72-
T 2rrLRf
The viscosity of Newtonian fluid is given by Margules’ equation:
where L is the length of the bob, Ri is the radius of the bob, R, is the radius of the cup, T is the torque, and K is defined as the ratio of the radius of the cup to the radius of the bob (RJR,). If the gap is very small compared to the radius of the cylinder, there is little variation in the shear rate across the gap. Therefore, the shear rate can be approximated by the simplified equation
When the outer cylinder undergoes a very small sinusoidal oscillation in a tangential direction, this motion causes the inner cylinder, suspended by a torsion wire, to oscillate with the same frequency but with different amplitude and phase. The viscoelastic properties of the fluid can be determined by the expressions (Bird et al., 1987) -pB WAsin(8)
+ A2 - 2Acos(8)
(74)
pBWA[A - COS(~)] 1 + A2 - 2Acos(S)’
(75)
“I’ = 1
”’
=
where p is the density, W, A and S are all functions of the dimensionless frequency 0,and B is the instrument parameter defined by B
=
(a - 1 ) 2 R 2 ( V Z @ ,
where K is the torsion constant of the wire and Z is the moment of inertia for the bob.
THE RHEOLOGY OF SEMILIQUID FOODS
33
The major sources of errors associated with concentric cylinders are the end effects and turbulent flow in the end region. Princen (1986) proposed a modification to eliminate the end effects in a concentric cylinder rheometer. He proposed adding a pool of mercury at the bottom of the cup that essentially eliminates the torque exerted on the bottom of the inner cylinder and on the sample in the gap. However, the limitation of this modification is that the fluid to be tested must be considerably more viscous than mercury, and the angular velocity of the cup must be kept below the levels where centrifugal force or normal force effects start to significantly alter the shape of the samplelair and sample/mercury interfaces. Another feature of some commercial rheometers is a well in the bottom of the inner cylinder in order to minimize the end effect (Fig. 17c). Its purpose is to trap an air bubble and thus eliminate the torque on the bottom. However, it has been demonstrated that the bubble immediately balls up and assumes the shape of a sessile bubble, so that a major fraction of the bottom area is again in contact with the field (Fig. 17d). Alternatively, the shape of the end of the cylinder can be chosen as a cone (see Fig. 17e) (angle equal to tan-'[(R, - Ri)/ R,]) such that the shear rate in the liquid trapped between the cone and the bottom is the same as that in the liquid between the cylinders. V.
CONSTITUTIVE MODELS
Rheological studies become particularly useful when predictive relationships between rheological properties and the responsible structural unit of food materials are being developed. However, most food materials have a complex structure and composition (i.e., foods contain numerous compounds) which makes modeling very difficult. In order to understand the relationship between rheological properties and the structures of food materials, idealization of their conformations is necessary. A typical example is the freely jointed chain consisting of springs and beads (Bird et al., 1987). This idealization leads to models which describe stresses developed in materials as a result of an applied deformation. The models which are able to describe the relationship between components of stress and strain as well as strain rate are referred to as constitutive equations (Bird et al., 1987; Bagley, 1992a; Darby, 1976). A. RHEOLOGICAL MODEL FOR STEADY SHEAR FLOW The simplest rheological equation of state is for the Newtonian fluid where viscosity is the only material property needed to characterize the
34
GUSTAVO
v. BARBOSA-CANOVAS et ai.
fluid flow (Prud’homme, 1991). For a non-Newtonian fluid, the Newtonian model may be generalized by allowing the viscosity to be a function of shear rate, leading to the generalized Newtonian fluid model (Prud’homme, 1991). The power law (Eq. 3b) and the Herschel-Bulkley equation (Eq. 4) are examples of a generalized Newtonian fluid model. Another popular model is the Casson model which has the expression
where T is the shear stress, T~ is the yield stress, K is the consistency index, and is the shear rate. The power law, Herschel-Bulkley equation, and Casson model are simple and easy to use. However, these equations only work for modeling steady shear flows rather than transient or elongational flows. Thus, many other models have been proposed to fit experimental data more closely for food materials. Among these, it is worth mentioning the Ree-Eyring equation which has three constants
r
where fl is a molecular relaxation time, and 170 and vmare limiting viscosity at zero shear rate (++ 0) and at a very large shear rate (++ m), respectively. This equation has been used by Doublier and Launay (1976,1981) for guar gum and locust bean gum solutions with satisfactory results. However, when the shear rate range was extended up to or near the first Newtonian zone (a very low shear rate region (+ -+ 0) where the apparent viscosity is independent of shear rate), the model fit was not as close (Launay and Pasquet, 1982). One equation, initially developed by Cross (1965) for dispersed systems, has given good fits to the experimental flow curve for several polysaccharide solutions and dispersions over a very large shear range, extending, in a few cases, from the first to second Newtonian zones [the zones where the apparent viscosity is independent of shear rate at a very low shear rate region + 0) and a very large shear rate region (+ + CQ), respectively]
(r
where t is a relaxation time and m is a nondimensional exponent. When both 17 4 7)o and 17 S- qm,Eq. (79) predicts a power behavior where (1 m) approaches the flow index (n). Doublier and Launay (1981) have calculated the four parameters of the Cross equation for guar gum by a computerized nonlinear regression
THE RHEOLOGY OF SEMILIQUID FOODS
35
method, but no physical meaning was assigned to vmbecause it was extrapolated outside the measurement range. The mean relative deviations (MRD) between calculated and measured viscosities were typically 1-3%, but the Newtonian viscosity, vo,was not reached experimentally.
B. DILUTE SOLUTION MOLECULAR THEORIES Dilute solution molecular theories are beginning to find applications in food polymer rheology. Chou et al. (1991) tested the Rouse (1953) and Zimm (1956) theories for random coils and the Marvin and McKinney (1965) theory for rod-like molecules in reference to citrus pectin solutions. In the Rouse concept, the molecules are totally free draining without considering hydrodynamic interactions. In the Zimm approximation, on the other hand, it is assumed that hydrodynamic interactions have a significant effect and are taken into account when calculating the spectrum of relaxation times. At the other extreme, it is possible to develop molecular models which approximate the flow behavior of elongated molecules as rigid rods. Examples include Yamakawa’s cylinder, the Shishkebab model, and U1mann’s cylinder (Ferry, 1980). Marvin and McKinney (1965) also developed a model approximating rod-like behavior, but when Kokini and Chou (1993) compared the dilute solution behavior of apple pectin with the Zimm, Rouse, and rod-like theories, the Zimm model gave the best approximation. The equations to predict the reduced moduli and relaxation time of flexible random coil molecules of Rouse and Zimm types are
and
where w is the frequency, T~ is the spectrum of relaxation time, [G’IRis the reduced storage modulus, and [ G ] Ris the reduced loss modulus. The reduced storage modulus, reduced loss modulus, and relaxation time spectrum are given by
36
GUSTAVO V. BARBOSA-CANOVAS
[G’]R =
et
al.
((3’’ - UTJM
c-ro
(83)
cRT
where G’ and G” are the storage modulus and loss modulus of the dilute solution, respectively, M is the molecular weight, c is the concentration, R is the ideal gas constant, T is the absolute temperature, and qs is the solvent viscosity. To predict rigid rod behavior several theories are available. The general form of these predictions are
and
and m is given by
m
=
(ml +
m2)-l,
where ml and m2 are empirical constants. Table I1 shows the values of these constants for five different rigid rod models. The predicted reduced moduli from the theory of Marvin and TABLE I1 EMPIRICAL CONSTANTS 1111 AND mz FOR THE ELONGATED RIGID ROD MODEL‘
Model
ml
m2
m
Cylinder (Yamakawa, 1975) Cylinder (Ullman, 1969) Rigid dumbbell (Marvin and McKinney, 1965) Prolate ellipsoid (Cerf, 1952) Shishkebab (Kirkwood and Auer, 1951)
0.60 0.46 0.60 0.60 0.60
0.29 0.16 0.40 0.24 0.20
1.15 1.61 1.00 1.19 1.05
‘Source: Kokini, 1993.
THE RHEOLOGY OF SEMILIQUID FOODS
37
McKinney (1965) for rigid dumbbells as a function of W T are given in Fig. 18, and the predicted moduli for random coil theories of Rouse and Zimm are shown in Fig. 19. At high frequencies, the reduced moduli of the Rouse theory become equal and increase together with a slope of 1/2, while those in the Zimm theory remain unequal and increase in a parallel manner with a slope of 213. C. CONCENTRATED SOLUTION/MELT THEORIES
The Bird-Carreau Model
1.
The Bird-Carreau model is an integral model which involves taking an integral over the entire deformation history of the material (Bistany and Kokini, 1983). This model can describe non-Newtonian viscosity, shear rate-dependent normal stresses, frequency-dependent complex viscosity, stress relaxation after large deformation shear flow, recoil, and hysteresis loops (Bird and Carreau, 1968). The model parameters are determined by a2, a nonlinear least squares method in fitting four material functions (a,, A l , and A*).
I
3
2.
n 1
9 -m 0
2
g
0. -1 -2
m
O
-3-4
.
-5 -3
-2 -1
log
0
1
2
W T
FIG. 18. Predictions of reduced moduli for the rigid rod theory of Marvin and McKinney (1965).
38
GUSTAVO
v. BARBOSA-CANOVAS e t a / .
Rouse
log w r
Zimm
log w r
FIG. 19. Prediction of reduced moduli for flexible random coils as proposed by Rouse (1953) and Zimm (1956).
The Bird-Carreau model (Bird and Carreau, 1968 Carreau et al., 1968) prediction for 7 is m
'
=
5
1+
7P (Alp
y)*
and at larger shear rates, the above equation is approximated by
THE RHEOLOGY OF SEMILIQUID FOODS
where Alp
=
Al[2/(p
39
+ 1)]"1
Tp = Tofbp~T1p
Z(a1) = 2
It-"].
The Bird-Carreau prediction for 7' is
and, at high frequencies, 7' is approximated by
Finally, the prediction for f l w is
which converges to the following equation at high frequencies
where h2, = h2 (2/p + l)"? The Bird-Carreau model employs the use of four empirical constants (a1, a2, h l , and h2) and a zero shear limiting viscosity (qo)of the solutions. The constants alraz,hl, and h2,can be obtained by two different methods: one method is using a computer program which can combine least square method and the method of steepest descent analysis for determining parameters for the nonlinear mathematical models (Carreau et aL, 1968).Another way is to estimate by a graphic method as illustrated in Fig. 20: two constants, a1 and A ] , are obtained from a logarithmic plot of 77 vs and the other two constants, a2 and h2, are obtained from a logarithmic plot of 7'vs w.
r,
2. Doi and Edwards Theory
A molecular theory of viscoelasticity of molten, high molecular weight polymers that makes use of the reptation concept has been developed by
40
GUSTAVO V. BARBOSA-CANOVAS e t a / .
4 F
-m 0
slope=
1-a1 "1
log Y
slope=
t n
1'"i a2
log w
FIG. 20. Determination of the Bird-Carreau constants Al,
log w A2,
a,,a2 (from Bird et al., 1977).
Doi and Edwards (1978, 1979, 1986). They started with the Rousesegmented chain model for a polymer molecule. Because of the presence of neighboring molecules, there are many places along the chain where lateral motion is restricted, as shown in Fig. 21. To simplify the representation of these restrictions, Doi and Edwards assume that they are equivalent to placing the molecule of interest in the "tube" as shown in Fig. 22. This tube has a diameter d and length L. The mean field is represented by a three-dimensional cage. The primitive chain can move randomly forward or backward only along itself. For a monodisperse polymer, the linear viscoelasticity is characterized by
FIG. 21. Sketch showing one entire molecule together with the segments of other molecules that are located near to it and restrict its motion (from Dealy and Wissbrun, 1990).
THE RHEOLOGY OF SEMILIQUID FOODS
41
FIG. 22. Sketch showing the hypothetical tube assumed by Doi and Edwards to be equivalent in its effect to the segments shown in Fig. 21 (from Dealy and Wissbrun, 1990).
TI is given by
where GNois the terminal region plateau modulus, w is the frequency of the applied shear field, and p is the normal mode of motion. The lower limit of summation forp is 1. The upper limit is estimated from the viscosity versus concentration plot as p(upper limit)
=
4C -, 4e
where +e is the volume fraction of polymer in a solution and 4c is the volume fraction of polymer in solution at the critical concentration. For a polydisperse system the theory has been modified by Rahalkar et al. (1985)with the results
where p is dimensionless molecular weight, and f ( p ) is normalized molecular weight distribution.
D. SOLID FOODS Some foods exist in the form of a gel, so it is useful to determine the mechanism of gelation in new product development. Oakenfull and Scott
42
GUSTAVO
v. BARBOSA-CANOVAS e t a / .
(1988) demonstrated that for weak gels, information about the junction zone-the regions of polymer where the molecules interact to form a gel network-can be derived from the kinetics of gelation and from the relationship between shear modulus and concentration. The equations to predict the junction zone (n) are
G =
(RTdM)* (M[J]MJ - c
C)
and
where M is the number average molecular weight of the polymer, MJ is the number average molecular weight of the junction zone, [J]is effectively the molar concentration of the junction zone, n is the number of crosslinking loci that form a junction zone, c is the concentration of the gum solution, Kj is an association constant, R is the gas constant (1.987 cal * g/ (mole OK), and T is the temperature (OK). Combining Eqs. (101) and (102) to eliminate the quantity [J],a relationship between shear modulus ( G )and concentrations (c) in terms of M, MJ, Kj, and n can be obtained using numerical methods. The best fit experimental data of G versus c, the number of polysaccharide chains involved in a junction zone (n),as well as information about the. size and the thermodynamic stability of junction, can also be obtained. Table I11 shows the results obtained by Oakenfull and Scott (1988) after applying Eqs. (101) and (102). The data in Tabel I11 indicate that kappa-carrageenan has a larger junction than iota-carrageenan by a factor of more than two; kappazone (MJ) carrageenan's junction zones are also thermodynamically more stable with TABLE I11 SIZE AND THERMOSTABILITY OF THE JUNCTION ZONE IN GELS OF KAPPA-CARRAGEENAN
CALCULATED FROM SHEAR MODULUS DATA AT
M Mf n KJ
AGJ" No. of monomers units per junction zone AG," per monomer
298"~
Kappa-carrageenan
Iota-carrageenan
10,500 8,900 6.4 1.65 X lW3 -133 37 -3.58
17,800 4,100 2.1 554 -15.6 14 -1.12
THE RHEOLOGY OF SEMILIQUID FOODS
43
a free energy of formation (AGJ)of -133 kJ/mol for kappa-carrageenan compared to -15.6 kJ/mol for iota-carrageenan. Thus, iota-carrageenan forms softer, weaker gels than kappa-carrageenan at equivalent concentrations. Ainsworth and Blanshard (1980) demonstrated that the creep curves of gels could be described by a Maxwell element in series with one or more Voigt elements as described by the equation
where J ( t ) is the measured compliance, J, and Ji are the compliance of the Maxwell and Voigt springs, respectively, qNis the viscosity of the Maxwell dashpot, and q is the retardation time associated with the Voigt element. A modified Maxwell model and a nonexponential model were used by Nussinovitch et al. (1989) for characterization of the stress relaxation of agar and alginate gels as described by the equation
or in its linear form
where Fo is the initial force, F(t) is the decaying force, and kl and k2 are the constants. The force-time relationship in relaxation was fit to Eq. (105) using linear regression to yield the constants kl and kZ.The same relationship normalized and fit using nonlinear regression is described by the modified Maxwell model
Equations (105) and (106) are two different mathematical models for characterizing the relaxing behavior of agar and alginate gels. The main advantage of Eq. (105) is that it is a simple mathematical form which has the possibility of calculating its constant by simple linear regression. The model expressed by Eq. (106) is mathematically more elaborate, and it provides a more detailed account of the shape of the relaxation curves of
44
GUSTAVO
v. BARBOSA-CANOVAS et ai.
gels. This detail, however, comes at the expense of having to use a nonlinear regression procedure to determine its coefficients (Nussinovitch et al., 1989).
VI. APPLICATIONS OF RHEOLOGY IN CHARACTERIZING ENGINEERING PROPERTIES OF FOODS A. STEADY SHEAR VISCOSITY OF FLUID FOODS AND DILUTE FOOD POLYMER SOLUTIONS
The steady shear flow properties of fluid foods, including most beverages such as tea, coffee, beer, wines, and soda pop, and biopolymer materials, such as carbohydrates and protein, can be used in engineering design. Simple liquids, true solutions, low molecular weight solvents, dilute macromolecule dispersions in low molecular weight solvents, and noninteracting polymer solutions exhibit Newtonian behavior. Sugar solutions also exhibit these flow characteristics. Several researchers have studied the viscosity of sucrose solutions because they are often used to calibrate viscometers (Muller, 1973; see Table IV). Milk, which is an aqueous emulsion of butter fat globules of 1.5-10 pm in diameter and contains about 87% water, 4% fat, 5% sugar (mainly lactose), and 3% protein (mainly casein), is a Newtonian liquid. FernindezMartin (1972) pointed out that the viscosity of milk depends on temperature, concentration, and the physical state of fat and proteins which in turn is affected by thermal and mechanical treatments. Fernfindez-Martin (1972) TABLE IV COEFFICIENT OF VISCOSITY OF SUCROSE SOLUTIONS AT
20"c (MULLER, 1973)
Sucrose (%)
gll00 g water
Viscosity (mPa . sec)
20 25 30 35 40 45 50 55 60 65 70 75
25.0 33.2 42.9 53.8 66.7 81.8 100.0 122.2 150.0 185.7 233.3 300.0
2.0 2.5 3.2 4.4 6.2 9.5 15.5 28.3 58.9 148.9 485.0 2344.0
45
THE RHEOLOGY OF SEMILIQUID FOODS
found that unconcentrated milks were Newtonian liquids, but concentrated milk showed a weak dependence on shear. All oils have a fairly high viscosity because of their long chain molecular structure. The longer the chain of the fatty acids, the higher the viscosity. Polymerized oils have a much higher viscosity than nonpolymerized oils. The viscosity of an oil also increases with saturation of the carbon double bonds. At a temperature higher than melting point, the artificial fat demonstrates a Newtonian behavior in the experimental shear range, but the viscosity is greatly dependent on its composition (Drake et al., 1994; see Table V). Generally, it appears that the greater the molecular interaction, the greater the viscosity. Table V demonstrates the dependency of viscosity of oil on the fatty acid composition, temperature, and its substitutes. Some fruit juices also exhibit Newtonian flow, such as filtered apple juice up to 30" Brix, Concord grape juice up to 50" Brix, and filtered orange juice of 10 and 18" Brix (Saravacos, 1970). This behavior is found in the 20-70°C temperature range. Depectinated and clarified fruit juices (apple, peach, pear, etc.) also exhibit Newtonian behavior (Rao et al., 1984;
TABLE V VISCOSITY OF SELECTED OIL AND ARTIFICIAL OIL
Name Corn oil Olive oil Rapeseed oil Soybean oil
Temperature ("C)
Viscosity (Pa . sec)
References
25 38 10 40 70 0 20 30 30 50
0.0565 0.0317 0.138 0.0363 0.0124 2.53 0.16 0.096 0.046 0.0206 0.0078 0.016 0.272 0.215 0.200 0.104
Steffe et al. (1986)
90
Mfat" M SPEb MM SPEC MC SPEd MT SPEP
50 50 50 50 50
%fat, anhydrous milkfat. bM SPE, milkfat sources polyester. 'MM SPE, milkfat :myristate sources polyester. dMC SPE, milkfat :coconut sources polyester. 'MT SPE, milkfat: tallow sources polyester.
Steffe et al. (1986) Steffe et al. (1986) Steffe ef al. (1986)
Drake et al. Drake et al. Drake et al. Drake et a!. Drake et al.
(1994) (1994) (1994) (1994) (1994)
46
GUSTAVO V. BARBOSA-CANOVAS et al.
Ibarz et al., 1987,1989,1992). Other important foods that show Newtonian behavior are table syrups such as honey, corn syrup, blends of sucrose, and molasses. Examples of the rheological properties of several foods were listed by Barbosa-Cfinovasand Peleg (1983), Barbosa-Cfinovaset af. (1993), Steffe (1992b), and Kokini (1992). However, most fluid foods exhibit a non-Newtonian behavior instead of the Newtonian behavior discussed above. Thus, the power law equation (Eq. 3), Herschel-Bulkley equation (Eq. 4),and the Casson equation (Eq. 77) are often used in characterizing flow properties of food. The parameters of these models for some foods are shown in Table VI. One has to keep in mind that the power law equation, Herschel-Bulkley equation, and Casson equation are often valid for about two or three logarithmic cycles of shear rate, and the range of shear rate tested should always be recorded to prevent erroneous results from extrapolation outside this range. For a wider shear rate range, the Ree-Eyring equation or Cross equation [Eqs. (78) and (79)] needs to be used. Table VII shows the application of the Ree-Eyring equation for orange juice at different temperatures (Vitali and Rao, 1984). It is noted that food products such as apple sauce, mustard, and tomato ketchup are time dependent, since the parameters calculating from the first shear curve differed from the second shear curve (see Table VI) (Barbosa-Cfinovas and Peleg, 1983). Many semisolid food materials portray yield stress. The yield stress in polysaccharide dispersions results from intermolecular hydrogen bonding and molecular entanglements (Gencer, 1985). A precise quantitative knowledge of yield stress is necessary since it is very important in pumping operations, stability of suspensions, appearance of coated materials, and consumer acceptability (Kee and Durning, 1990). Yield stresses can be measured with a variety of techniques. These include measuring the shear stress at vanishing shear rates, extrapolation of data using rheological models that include yield stresses and stress relaxation experiments, and others (Barbosa-Cfinovas and Peleg, 1983; Lang and Rha, 1981). Kee and Durning (1990) reviewed two principal methods of measuring yield stresses: dynamic and static methods. One example of the dynamic method is the extrapolation from the flow curve. Equation (4) is often used to determine the yield stress of gum solutions. Table VIII lists the examples of yield stress of several selected food commodities measured from different methods. It is noted that for the same food product, different methods have different yield stress value. In addition to the measuring method, the embedded factor-the composition of food products-also needs to be emphasized. For instance, in mayonnaise, the concentration of oil and xanthan gum significantly affected the yield stress since it increased from
TABLE VI RHEOLOGICAL PARAMETERS OF SELECTED FOODS
Power law Product Apple sauce (Matt's) Apple sauce (Stop & Shop) Mustard (Gulden's) Mustard (Stop & Shop) Tomato ketchup (Heinz) Tomato ketchup (Stop & Shop) Orange juice (low PdP) -18.5"C -14.1"C -9.3"C -5.O"C -0.7"C 10.1"C 19.9"C 29.5"C Concentrated orange juice (Hamlin, early) (Hamlin,late)
TSS ("Brix) 18.2 18.1 -
-
23.2 34.4
Herschel-Bulkley model
Casson model
Flow curve
K
n
1st 2nd 1st 2nd 1st 2nd 1st 2nd 1st 2nd 1st 2nd
TQ
K
n
SS
TQ
K
SS
33 34 26 30 20 35 17 20 21 24 15 15
31 6.8 33 6.4 37 5.5 16 3.4 9.4 2.2 6.4 2.0
0.16 0.42 0.15 0.43 0.21 0.52 0.29 0.56 0.38 0.61 0.4 0.6
49 42 216 116 462 251 1244 634 46 36 43 74
8.0 6.5 7.8 6.1 7.9 6.4 6.1 4.8 5.8 4.9 4.8 4.0
0.16 0.24 0.16 0.25 0.26 0.34 0.25 0.32 0.26 0.49 0.24 0.27
461 82 705 153 1196 280 1406 641 133 53 88 75
References Barbosa-Cbnovas and Peleg (1983)
-
65.0
29.16 14.58 10.80 7.88 5.93 2.72 1.64 0.91
Rao er al. (1984)
0.712 0.757 0.743 0.721 0.711 0.725 0.721 0.739
65 1.32 0.94
0.083 0.031
0.859 1.055
Crandall ef af. (1982)
48
GUSTAVO V. BARBOSA-CANOVAS et al.
TABLE VII REE-EYRING EQUATION PARAMETERS FOR PERANP SAMPLE"
Temp ("C)
')o
P
11-
SSb
-18.8 -14.5 -9.9 -5.4 -0.8 9.5 19.4 29.2
76.61 42.04 27.03 16.09 9.98 4.25 2.62 1.69
0.2338 0.1211 0.1096 0.0747 0.0500 0.0259 0.0393 0.0421
65.02 35.77 23.54 13.71 8.37 3.63 2.15 1.41
600 1.531 291 433 522 571 154 29
"Source: Vitali and Rao, 1984. PERANP was a 65"Brix, 5.7% pulp sample made from Pera oranges. *SS is the sum of the squares of deviation between the data and the model.
23 to 235 Pa as the oil concentration increased from 75 to 85% or from 55 to 195 Pa as the xanthan gum concentration increased from 0.5 to 1.5% (Ma and Barbosa-Cinovas, 1995b). Yoshimura et al. (1987) compared the yield stress of a series of model oilin-water emulsions determined from three techniques: concentric cylinder, parallel disk, and a vane. The techniques give comparable results, with parallel disk having the largest uncertainty when measuring the yield stress. No one technique is superior to the others but each has advantages and limitations. Experimentally it is easier to change gap spacing on parallel disks with the sample in place than to remove the sample, change cylinders, and reload the concentric cylinder geometry. Experiments using the vane method with a stress rheometer are easy to perform and are of high precision. Viscosity information cannot be obtained since the flow field around the vanes is quite complex after flow begins. The technique requires large volumes of sample relative to the other two methods.
B. CHARACTERIZATION OF ENTANGLEMENT IN CONCENTRATED SOLUTION Viscoelastic properties of biopolymeric materials such as carbohydrates and protein can be used to characterize their three-dimensional configuration in solutions. This configuration affects their functionality in many food products. An understanding of how the molecular structure of polymers affects their rheological properties can make it possible to predict and improve the flow behavior of newly developed food products that have such
THE RHEOLOGY OF SEMILIQUID FOODS
49
TABLE VIII YIELD STRESS OF SELECTED FOODS
Products
Yield stress (Pa)
Apple sauce Apple sauce Apple sauce Ketchup Ketchup Ketchup Ketchup Mayonnaise Mayonnaise Mustard Mustard Tomato puree Tomato puree Emulsion EMlB
58.6 45-87 46-82 22.8 15.4-16.0 18-30 26-30 24.8-26.9 81-91 34.0 52-78 23.0 25-34
Extrapolation Stress to initiate flow Squeezing flow Vane method Stress to initiate flow Squeezing flow Extrapolation Squeezing flow Stress decay Vane method
120 t- 5 100 t 50 104 ? 2 240 2 5 270 2 50 241 t 2 460 2 10 450 t 50 471 2 4 570 ? 10 460 t 50 536 ? 4 42 2 1 80 2 20 48 Ifi 2
Concentric cylinder Parallel disks Vane Concentric cylinder Parallel disks Vane Concentric cylinder Parallel disks Vane Concentric cylinder Parallel disks Vane Concentric cylinder Parallel disks Vane
EM2B
EM3B
EM2A
EM2C
Method Stress decay Squeezing flow
References Charm (1963) Campanella and Peleg (1987b) Qiu and Rao (1988) Ofoli et al. (1987) De Kee et al. (1980) Campanella and Peleg (1987b) Missaire et a/. (1990) D e Kee et al. (1980) Campanella and Peleg (1987b) Ofoli et a/. (1987) Campanella and Peleg (1987b) Charm (1963) Missaire et a/. (1990) Yoshimura e t a / . (1987)
polymers (Liguori, 1985). Examples can be found at the consistency and stability improvement of emulsions by using polymers with enhanced surface activity and greater viscosity and elasticity. Constitutive equations were applied to simulate viscoelasticity of concentrated food polymer dispersions. Some fundamental and empirical models have been discussed in Section V. Among them, the Bird-Carreau constitutive model [Eqs. (89-94)] have been used for food polymer dispersions (Kokini et al., 1984; Kokini and Plutchok, 1987b; Plutchok and Kokini, 1986). In Fig. 23, the Bird-Carreau model is compared to the experimental data of a 1.0% guar solution in the frequencylshear rate range of 0.1 to 100 sec-*.
50
GUSTAVO
v. BARBOSA-CANOVAS
et UI.
-model lo-'
100
10'
102
Frequency, Shear Rate Sec-l FIG. 23. Comparison of predictions of the Bird-Carreau constitutive model and experimental data for 1% guar solution (Kokini el al., 1984).
For ~/'/w, the model very accurately predicted the high-frequency region. The low-frequency region was predicted somewhat less accurately. Experimental values of log 7'versus log w and log 77 versus log i, were fit quite accurately by the model. Such constitutive models can also be used to predict the rheological properties of concentrated gum blend systems as well. Plutchok and Kokini (1986) developed empirical equations capable of predicting TO, h l , and h2, as well as the slope of the non-Newtonian region of and q', using concentration and molecular weight data. The equations to predict rheological constants of CMC, guar gum, and CMCguar blend ratios of 3: 1, 2: 1, 1:1, 1:2, and 1:3 in the concentration range of 0.5 to 1.5% by weight are
-
3.63 rlo = 1.06 .1019 c blend
.
1.75 blend
.
1.72 blend
.
hl = 1.81 . l o 5 c
-
,
h2 = 1.63 109 c
-2.94 w,blend -0.92
w,blend -1.72 w,blend
THE RHEOLOGY OF SEMILIQUID FOODS
s, = 1.63 .107
0.50 blend
.
0.32 s,. = 2.17 . l o 7 . cblend
51
-1.26 w,blend
-1.26 Mw,blend'
where M w = X1Mwl + X2 Mwz;Xiis mass fraction, M , is weight-average molecular weight (g/g * mol), Mwl and MW2 are molecular weight of each component of the blend
where vi is the volume fraction and Cblend is the concentration (g/lOO ml). Using these empirical equations in conjunction with the predictions of the Bird-Carreau model, it is possible to predict 77' and q''/o.An example of such a plot is shown in Fig. 24 for a 1.0% CMC:guar blend (3: 1). Experimental data are superimposed on these plots to judge the aptness of the model. The steady shear viscosity 77 and the dynamic viscosity 77' are well predicted in the shear rate range of 0.1 to 100 sec-*.The experimental data, as well as the theoretical prediction, portray commonly observed behaviors by polymeric dispersions. In this instance, 77 and v' for this blend ratio tend to some value, a property suggested by the Bird-Carreau model at low shear rate (Kokini et al., 1984).
-I
v) .-
g
10'
v
3 > 100 h
.-%
g
y
10'
F
i
-Bird-Carreau model oA3 Experimental
.-
-~
102
10'
lo*
70'
10-2
Frequency, Shear Rate (sec-1) FIG. 24. Comparison of prediction of the Bird-Carreau constitutive model and experimental data for a 3 :1 CMC: guar blend at a total concentration of 1% (Kokini, 1993).
52 104-
*'
102-
3 C
+ A
10'A
b
10-2.
0
1,
+
+
0"
A
0
A
v
4+'
0
+A'+
v)
+*+f:
0 0 0
0 0
+
ExperimentalG' Monodisperse G'
0
A
Polydisperse G'
10-4
FIG. 25. Predictions of G' values for a 5% pectin solution using the Doi-Edwards model (Kokini, 1993).
To predict G' and G' values for 5% pectin dispersion the equations assuming a monodisperse polymer as well as the equations assuming a poiydisperse polymer were used. For the polydisperse case, a computer program was developed to account for a small molecular fraction measured using low-angle light scattering coupled with HPLC. Figures 25 and 26 show the plot of predicted values along with the experimental values for the simulations of G' and G", respectively. It is clear that the polydisperse
''I
101
1 00
t
ExperimentalG"
0
Monodisperse G"
A
Polydisperse G"
10'
1 02
10
Freq (rad/sec) FIG. 26. Predictions of G" values for a 5% pectin solution using the Doi-Edwards model (Kokini, 1993).
THE RHEOLOGY OF SEMILIQUID FOODS
53
model explains the experimental data much better than the monodisperse model, which is in agreement with the polydispersity ratios (Mw/Mn)ranging from 15 to 45 (Chou and Kokini, 1987; Shrimanker, 1989). C . APPLICATION OF CONSTITUTIVE EQUATIONS IN DOUGHRHEOLOGY
The ability to accurately predict the nonlinear viscoelastic behavior of a wheat dough is of practical interest to scientists developing new products or technologies in the food industry. Because of the many different processing schemes in use it is necessary to accurately predict the rheological behavior, especially the steady viscosity and the primary normal stress coefficient, through a shear rate range which is relevant to processing. The BirdCarreau model, although semiempirical, provides the accuracy and the versatility which should make it of particular interest to those working with wheat dough. Using small amplitude oscillatory properties, Smith et al. (1970) showed that as protein content increased in a protein (gluten)starch-water system the magnitude of both the storage modulus ( G ’ )and loss modulus ( G ) increased. Hibberd and Wallace (1966) reported a critical strain of approximately 0.5% beyond which wheat dough rheological properties became nonlinear. Dus and Kokini (1990) showed that the steady shear rheological properties (7, N I )and small amplitude rheological properites (7’and f / o ) could be successfully simulated using the Bird-Carreau constitutive model. The experimental and predicted values of the steady viscosity function versus shear rate are given in Figure 27. The experimental data are well
v)
4
$’ 8 v)
5 9
2
105104d Predicted
lo3 . A Stress Rheom. 102. 0 RMS 0
10”
.
D
Capillary
10-6
104
10.2
100
102
104
Shear Rate (s-1) FIG. 27. Experimentalvalues and those predicted using the Bird-Carreau model of apparent viscosity (7)as a function of shear rate for hard flour dough sample (Dus and Kokini, 1990).
54
GUSTAVO V. BARBOSA-CANOVAS et aL
I
A
A Predicted 0 Experimental
<
1021 10-5
10-3
A
10’
10-1
103
Frequency (s-’) FIG. 28. Experimental and predicted values of dynamic viscosity (7’) versus frequency for the 40% moisture hard flour dough sample (Dus and Kokini, 1990).
simulated by the Bird-Carreau model. Comparison of experimental data versus the predicted data shows a high degree of superposition throughout the range of viscosity. The prediction of 7’and the experimental data versus frequency in Fig. 28 show that the prediction works well. This is especially so at the intermediate frequencies where the prediction is very close. The experimental data are seen to deviate slightly at the extremes of the frequency range. Figure 29 shows the predicted and experimental values of rJ’lw versus frequency. In the range of frequencies tested there is a very high degree 10’1
-A
109 -
A
A 0
a u,
107-
e 9
105 -
0
AA
F
lo3
-
101 ..
0 0 0 0 0
0
A Predicted o Experimental
0 0 0 0
0
0
FIG. 29. Experimental and predicted values of q”/o versus frequency for the 40%moisture hard flour dough sample (Dusand Kokini,1990).
55
THE RHEOLOGY OF SEMILIQUID FOODS
of correlation between experimental and predicted data. The performance of this prediction is better than those of the other functions. The rheological properties of a number of commercial glutens have been examined and compared with their baking performance when reconstituted with starch and flour water soluble material. Figure 30 shows the relationship between loaf volume and the dynamic rheological properties G' and G' for a range of glutens hydrated to 65% moisture. There is an obvious relationship in that dough with higher values of the storage and loss moduli have consistently lower loaf volumes (LeGrys et al., 1981). Thus, for a gluten to perform well in a baking test it should be extensible without an excessive amount of elastic recoil.
D. RHEOLOGY OF EMULSIONS Emulsions are dispersions of one liquid phase in the form of fine droplets in another immiscible liquid phase. The immisciable phases are usually oil and water, so emulsions can be broadly classified as oil-in-water or waterin-oil emulsions, depending on the dispersed phase. Some typical food emulsions are mild cream, ice cream, butter, margarine, salad dressing, and meat emulsions. The results from rheological measurements can allow for a better understanding of how various emulsifiers/stabilizers interact to stabilize emulsions. Understanding the effect of additives such as food
\\
0
G' 0
Modulus x 103,dyne cm-2
FIG. 30. Relationship between loaf volume and storage and loss moduli for commercial glutens rehydrated, after freeze drying, to 65% moisture (frequency = 10 radlsec, strain = 0.25; LeGrys et nl., 1981).
56
GUSTAVO
v. BARBOSA-CANOVASel al.
gums, salts, and sugar on the stabilizing action emulsifiershtabilizers allows the development of the best mixture of stabilizers to function effectively in food systems. Steady shear measurements were used to determine flow properties and to estimate the degree of structure breakdown with shear (Elliott and Ganz, 1977). The power law equation (Eq. 3) has been used to describe the shear stress-shear rate behavior of salad dressings (Figoni and Shoemaker, 1983; Paredes et al., 1988, 1989). The flow behavior index of five commercial salad dressings at different temperatures and storage times of up to 29 days were all less than one, indicating that they were pseudoplastic fluids. The consistency index ( K ) decreased with the increase in product temperature. Shear modulus is a measure of the elasticity of an emulsion. The higher the shear modulus, the greater the capacity of an emulsion to store energy. In practice, a high value of shear modulus is indicative of enhanced emulsion stability in low-stress situations. On the other hand, high shear modulus usually implies high viscosity which also increases pumping and handling difficulties (Gladwell et al., 1985a,b). It is also possible to link the shear modulus with colloidal interactions between droplets (Buscall et al., 1982). In soy oil-water emulsions, most of the viscoelastic behavior can be attributed to the interactions between xanthan gum microgel and emulsion droplets, which indicates that the droplets behave as soft deformable spheres rather than as rigid spheres (Gladwell et al., 1985a,b). The compliance response of viscoelastic materials in a creep/recovery test is due to three mechanisms: instantaneous elastic, retarded elastic, and viscous flow. As oil concentration increases, the emulsions become more elastic, with the instantaneous elastic mechanism becoming more dominant. In the nonlinear region the viscous flow is more dominant, with the emulsions becoming less elastic with increasing shear stress (Gladwell et al., 1985a,b). A four parameter model was employed to describe the creep compliance data (Paredes et al., 1988). Major changes in the rheological parameters took place during the initial 7 days of storage. Egg yolk is used in food emulsions, such as salad dressing and mayonnaise, to lower the oil-water interfacial tension. Hydrocolloid gum such as propylene glycol alginate, carrageenan, guar, and xanthan are the main emulsion stabilizers in commercial salad dressings. Because of their complex composition (e.g., egg yolk, food gums, salt, etc.), salad dressings exhibit complex and thixotropic rheological properties (Paredes et al., 1988). Vernon-Carter and Sherman (1980) reported that the creep was influenced by pH, electrolytes (NaC1, CaC12),and storage time. The change in rheological properties during storage was dependent on the rate of droplet coalescence and the ability of the mesquite layers in adjacent oil droplets to interpenetrate and form
57
THE RHEOLOGY O F SEMILIQUID FOODS
new linkages. Table IX shows the magnitudes of the four creep parameters for the six salad dressings (Paredes et al., 1988). Structural breakdown is a time-dependent process resulting in a decrease in the viscosity of a product. The classical approach to characterizing structural breakdown is the measurement of the hysteresis loop, first reported by Green and Weltmann (1943). A sample is sheared at a continuously increasing, then continuously decreasing, shear rate, and a shear stressshear rate flow curve is plotted. If structural breakdown occurs, the two curves do not coincide, creating a hysteresis loop. The area enclosed by the loop indicates the degree of breakdown. Dynamic measurements, at strain amplitudes within the linear viscoelastic limit, were made to establish the properties of the essentially undisturbed samples (Elliott and Ganz, 1977). Oscillatory experiments are a powerful tool to study the effects of aging, the amount and type of ingredients, and additives such as food gums on the rheological properties and quality of salad dressings (Munoz and Sherman, 1990). Small amplitude dynamic viscoelastic properties of apple butter, mustard, table margarine, and mayonnaise were compared to their respective properties in steady shear flow in the range of shear rates and frequencies of 0.1 to 100 sec-’ (Bistany and Kokini, 1983). Comparisons of dynamic and steady viscosities showed that dynamic viscosities (q*) are much greater Consequently, the Cox-Merz rule is not obeyed than steady viscosities (7). (Bistany and Kokini, 1983). This phenomenon can be explained by a signifi-
TABLE IX FOUR CREEP PARAMETERS OF SIX COMMERCIAL SALAD DRESSINGSa
Ah ARCC
Bd BRC‘ Cf
Dg
55.2 55.2 55.2 55.2 22.8 22.8
1.42 0.55 2.04 1.15 0.62 0.61
1.o 0.52 2.08 0.27 1.02
0.48
“Source: Paredes et al., 1989. bA is bottled creamy style A. ‘ARC is bottled creamy style A-reduced calorie. dB is bottled creamy style B. ‘BRC is bottled creamy style B-reduced calorie. fC is dry creamy mix style C. g D is dry creamy mix style D.
16.2 22.5 38.5 7.05 9.10 1.16
79.8 148.8 587.4 67.2 19.8 3.6
58
GUSTAVO V. BARBOSA-ChOVAS et al.
cant destructive effect during the steady shear measurements. However, the structure breakdown due to shear strains between 0.001 and 0.1 was found to be fully recoverable in a small amplitude oscillatory test of butter (Rohm and Weidinger, 1993). No qualitative differences in the rheological behaviors were observed between the samples. Temperature-induced variations were mainly ascribed to changes in the solid fat content. Correlations between selected rheological measures in the linear viscoelasticregion (i.e., complex modulus, G*, relaxation modulus, G(t),and modulus of deformability, Md) were affected by the solid fat content (Rohm and Weidinger, 1993). Margarine and tablespread are oil-in-water emulsions. Melting characteristics of these products are important for flavor release and consumer acceptance. Oscillatory measurements as a function of temperature and drop points were used to quantify rheological changes accompanying melting, The rheology of low-fat spread is governed by emulsion characteristics such as the proportion of the aqueous phase and the size of the water droplets. E. EXTENSIONAL FLOW
While shear rheological properties of food materials have received a great deal of attention (Kokini and Dickie, 1981; Plutchok and Kokini, 1986), extensional properties of food materials are only recently being investigated (Bagley et al., 1988; Bagley, 1992a; Bhattacharya and Padmanabhan, 1992). Among the wide variety of food materials, wheat dough is one of the most complex and also the most interesting. Schofield and Scott-Blair (1932) studied the relaxation and elastic recovery of dough during extension and suggested that the elastic behavior of wheat dough resulted from the protein fraction of flour. Hlynka and Anderson (1952) examined the relaxation behavior of dough in tension using a spectrum of relaxation times. They found that gluten relaxograms were of the same type as those of dough, but starch relaxograms were different from those of dough. Glucklich and Shelef (1962) showed wheat flour dough to be nonlinear viscoelastic in uniaxial compression past a critical stress value. They also noted that dough might be regarded as almost incompressible due to a very large elastic bulk modulus. Tschoegl et al. (1970a,b) confirmed nonlinear viscoelasticbehavior in large deformation. They found that extension rate, temperature, water content, and flour type had a drastic effect on stress-strain behavior. Bagley (1992a) measured the apparent biaxial elongational viscosity of wheat flour dough. The upper convected Maxwell model was considered to be adequate in explaining both the effect of crosshead speed and sample
THE RHEOLOGY OF SEMILIQUID FOODS
59
dimensions. The Leonov constitutive model was also used by the same author in analyzing both extensional and shear flows. Senouci and Smith (1988) used a simplified analysis for converging flow in a piston-driving capillary rheometer at 120-130°C and obtained ratios of uniaxial extensional viscosities to shear viscosities of maize grits and potato powder in the range of 60-3900. Gas cell expansion for loaf volume development during baking is largely biaxial stretching flow (Bloksma and Nieman, 1975; de Bruijne et al., 1990). Extensional viscosity data of wheat dough are necessary to relate functional properties of bread such as loaf volume to biaxial extensional rheological properties. Bloksma (1988) estimated that extensional rate ranged from sec-' during fermentation of bread dough. To predict the to performance of these processes it is necessary to obtain accurate biaxial extensional viscosity data, preferably in this range of extension rate. Doughs with different protein contents (13.2, 16.0, and 18.8% based on 14%MB) showed different biaxial extensional viscosities (Huang and Kokini, 1993). Figure 31 shows that the biaxial extensional viscosity approached 677 at an extensional rate of 7.3 X sec-'. Strain thinning behavior was observed in dough during biaxial extension. F. FORMULATIONS DEVELOPMENT Most food producers want to process at the highest rate possible for maximum output. However, the shear history during processing can I
16.0% protein
h
fn
m
k .-2 fn
8
.-fn
> 105
104
10-5
10-4
10-3
10-2
10-1
Shear or Extension Rate l/s FIG. 31. Biaxial extensional viscosity and shear viscosity vs extension and shear for wheat flour dough with 16.0%protein content (Huang and Kokini, 1993). m is biaxial extensional ~ that the biaxial extensional viscosity. 17 is shear viscosity. 6*17 superimposed with 1 ) indicate viscosity are about five times larger than shear viscosity at the equivalent shear rate.
60
GUSTAVO v . BARBOSA-CANOVAS et a[.
change the fluid microstructure, which, in turn, may affect the characteristics of the final product. As consumers are becoming more concerned about health and nutrition, food researchers seek to adjust product formulations to reduce calories and still provide good quality. But it is important to fully compare ingredients before switching formulations, as some may appear the same but behave differently in processing. Figure 32a compares the results of three different starch solutions at 70°C,while Fig. 32b compares the same solutions at 130°C. Comparing the order of these three solutions based on the viscosity at the two temperatures shows that the order is reversed. The molecular weight and conformation of the starch molecules in solution have a great effect on the enhancement of the viscosity. The gel point in materials represents the point where behavior changes from viscous (liquid-like) to elastic (solid-like). The conditions under which this occurs are critical to such food constituents as wheat-soya solutions used as setting agents within reconstituted meat products. In this case, oscillatory-controlled stress experiments, in which a small sinusoidal stress is applied to the material, provide a convenient method for evaluating elastic and viscous properties without destroying the delicate structure of soft semisolids.
a
b
100 m
m
2 la .b @I
8m
5
1
0.1
0.1
1
10
Shear rate ljS
100
0.5 .5 1 2
5 1020 5 0 1 0 0 2 m
Shear rate
FIG. 32. Viscosity as a function of shear rate for three starch solutions (Race, 1993). A, starch solution A; B, starch solution B; C, starch solution C. (a) Measured at 70°C in the high pressure cell. (b) Measured at 130°C in the high pressure cell. Note that in (a) the viscosity was measured at 70"C, the starch solution A behaves similar to B but significantly smaller than starch solution C; in (b) the viscosity was measured at 130"C, the viscosity starch solution A was significantly greater than that of solution B and C, while solution B and C behaves in a similar way.
61
THE RHEOLOGY OF SEMILIQUID FOODS
Figure 33 illustrates the ability to detect the temperature of gelation for two aqueous wheat solutions with different gelation enhancing enzymes. The temperature of gelation point is where the G’ and G” curves cross (G’ and G” being the storage and loss modulus, respectively, and the crossover point at which there is a balance between solid- and liquidlike structure). Even though the different enzymes yield the same gelling temperature, the final gel strength G ‘ is much higher for enzyme A. This information can be used to gain an appreciation of the food product’s behavior during production and when consumed. For example, at the gel point, the character of food changes significantly. Therefore, it may be crucial to pump products into packaging before this point is reached. The gel strength, on the other hand, has a great bearing on a texture of food, and, hence, the appeal when consumed. In the case of reconstituted meat products, gel strength also dictates whether or not they hold together during heating and cooking. VII.
SUMMARY AND RESEARCH NEEDS
An attempt has been made to review the essential food rheology concepts and the current state of knowledge regarding rheological properties of food materials and food products. The progress made in the measurement and simulation of the viscoelasticity of semisolid food and their biopolymeric I
1031
103
102 h
m
iLY
10
1
lo-’
60
70
80 Temperature (“C)
90
A: Wheat solution treated with enzyme A 8: Wheat solution treated with enzyme B
FIG. 33. Gelation of wheat solutions (Anonymous, 1993). A, wheat solution treated with enzyme A. B, wheat solution treated with enzyme B.
62
GUSTAVO
v. BARBOSA-CANOVAS et a[.
components has also been discussed. Various constitutive equations have been employed for the interpretation of experimental data and prediction of the rheological properties of food material and food products. In particular, constitutive models were shown to be useful in many ways. First, dilute solution theories have been shown to be useful to characterize long range conformation and flexibility of carbohydrate- and protein-based polymers. Rheological measurements and constitutive theories predicting rheological properties compliment short range studies such as neutron scattering. One important advantage of combining short range vs long range studies was to understand the effect of short range flexibility on long range conformation. This would be a fertile area for new research. Second, semiempirical constitutive models such as the Bird-Carreau model were shown to be useful to generate a database for predicting steady and small amplitude rheological properties of viscoelastic materials. This database should enable estimation of important parameters such as recoverable strain, which affects die swell and recovery processes in general in the processing of viscoelastic materials. While these models do not originate from specific molecular information, they do incorporate key assumptions pertaining to network formation and dissolution which clearly occurs during deformation processes. The semiempiricism facilitates the estimation of parameters and makes the model easily applicable to complex food material such as dough. Third, a constitutive model with accurate molecular and conformational origins such as the Doi-Edwards model was able to simulate concentrated pectin dispersion rheology quite accurately. Such studies enabled correlation between chemical structure and functionality. While state of the art modeling does not permit the use of such detailed understanding with complex food systems, it nevertheless provides major clues about how to design molecules for functionality. Such knowledge can provide design guidelines to genetically engineer materials of biological origin with optimal functionality. Clearly there is need for much more work in these areas of food rheology. It is first necessary to develop better methods for extensional viscosity of fluid and semigels. In addition, it is necessary to bring constitutive models into the field that are able to predict extensional properties. Constitutive models which can predict rheological properties of dispersed system (suspensions, emulsions) need to be developed. Advances in techniques and numerical methods are necessary so that nonlinear constitutive models can be developed that are capable of incorporating the diverse and complex structural properties of foods. The predictive capability developed through this effort will result in product design and process improvement rules leading to improved food products for the consumer.
THE RHEOLOGY OF SEMILIQUID FOODS
63
REFERENCES Ainsworth, P. A., and Blanshard, J. M. V. (1980). Effect of thermal processing on structure and rheological properties of carrageenanlcarob gum gels. J. Text. Stud. 11, 149-162. Anonymous (1993). TA Instruments Application Notes. Bagley, E. B. (1992a). Constitutive models for doughs. In “Food Extrusion Science and Technology” (J. L. Kokini, C. Ho, and M. V. Karwe, eds.), pp. 203-212. Dekker, New York. Bagley, E. G., Christianson, D. D., and Martindale, J. A. (1988). Uniaxial compression of a hard wheat flour dough: Data analysis using the upper convected Maxwell model. J. Text. Stud. 19, 289-305. Barbosa-Canovas, G . V., and Peleg, M. (1983). Flow parameters of selected commercial semiliquid food products. J. Text. Stud. 14, 213-234. Barbosa-Canovas, G . V., Ibarz, A., and Peleg, M. (1993). Propiedades reol6gicas de alimentos fluidos. Rev. Aliment. 241, 43-93. Barnes, H. A., and Walters, K. (1985). The yield stress myth? Rheol. Actu 24,323-326. Barnes, H. A., Hutton, J. F., and Walters, K. (1989). “An Introduction to Rheology.” Elsevier, New York. Bhattacharya, M., and Padmanabhan, M. (1992). On line rheological measurement of food dough during extrusion cooking. In “Food Extrusion Science and Technology” (J. L. Kokini, C. Ho, and M. V. Karwe, eds.), pp. 213-232. Dekker, New York. Bird, R. B., and Carreau, P. J. (1968). A nonlinear viscoelastic model for polymer solutions and melts. 1. Chem. Eng. Sci. 23, 427-434. Bird, R. B., Stewart, W. E., and Lightfoot, E. N. (1969). “Transport Phenomena.” Wiley, New York. Bird, R. B., Armstrong, R. C., and Hassager, 0. (1977). “Dynamics of Polymeric Liquids.” Wiley, New York. Bird, R. B., Armstrong, R. C., and Hassager, 0.(1987). Fluid dynamics of rheometry. “Dynamics of Polymer Liquids,” 2nd ed., Vol. 1, Chapter 10. Wiley, New York. Bistany, K. L., and Kokini, J. L. (1983). Comparison of steady shear rheological properties and small amplitude dynamic viscoelastic properties of fluid food materials. J. Text. Stud. 14,113-124. Bloksma, A. H. (1972). Rheology of wheat flour doughs. J. Text. Stud. 3, 3-17. Bloksma, A. H. (1988). Rheological of the breadmaking process. Pap., Inr. Cereal Bread Congr., 8th, Lausanne, Switzerland, 1988. Bloksma, A. H., and Nieman, W. (1975). The effect of temperature on some rheological properties of wheat flour doughs. J. Text. Stud. 6, 343. Bourne, M. C. (1994). Covering from empirical to rheological tests on foods-it’s a matter of time. Cereal Foods World 39(1), 37-39. Buscall, R., Goodwin, J. W., Hawkins, M. W., and Ottewill, R. H. (1982). Viscoelasticproperties of concentrated lattices. Part 2. Theoretical analysis. J . Chem. Soc., Furaduy Trans. I 78,2889. Campanella, 0. H., and Peleg, M. (1987b). Determination of the yield stress of semi-liquid foods from squeezing Bow data. J. Food Sci. 52,214-215. Campanella, 0.H., and Peleg, M. (1987~).Analysis of the transient flow of mayonnaise in a coaxial cylinder viscometer. J . Rheol. 31, 439-452. Carreau, P. J., MacKonald, I. F., and Bird, R. B. (1968). A non linear viscoelastic model for polymer solutions and melts. Part 2. Chem. Eng. Sci. 23, 901. Charm, S. E. (1963). The direct determination of shear stress-shear rate behavior of foods in the presence of yield stress. J. Food Sci. 28, 107-113.
64
GUSTAVO
v. BARBOSA-CANOVAS
et
Cheng, D. C. H. (1986). Yield stress: A time dependent properties and how to measure it. Rheol. Acta 25, 25-554. Cheng, D. C. H., and Evans, F. (1965). Phenomenological characterization of the rheological behavior of inelastic reversible thixotropic and antithixotropic fluid. Br. J. Appl. Phys. 16,1599-1617. Chou, T. C., and Kokini, J. L. (1987). Rheological properties and conformation of tomato paste pectins, citrus and apple juice pectins. J. Food Sci. 52, 1658. Chou, T. C., Pintauro, N., and Kokini, J. L. (1991). Conformation of citrus pectin using small amplitude oscillatory rheometry. J. Food Sci. 56, 1365-1371. Cohen, Y., and Metzner, A. B. (1986). An analysis of apparent slip Row of polymer solutions. Rheol. Actu 25,28-35. Connelly, R. W., and Greener, J. (1985). High-shear viscometry with a rotational paralleldisk device. J. Rheol. 29(2), 209-226. Crandall, P. G., Chen, C. S., and Carter, R. D. (1982). Models for predicting viscosity of orange juice concentrates. Food Techno/. 36(5). 245. Cross, M. M. (1965). Rheology of non-Newtonian fluid: A new flow equation for pseudoplastic systems. J . Colloid Sci. 20, 417-437. Darby, R. (1976). “Viscoelastic Fluids: An Introduction to their Properties and Behavior.” Dekker, New York. Dealy, J. M., and Wissbrun, K. F. (1990). “Melt Rheology and Its Role in Plastics Processing.” Van Nostrand-Reinhold, New York. de Bruijne, D. W., de Looff, J., and van Eulem, A. (1990). The rheological properties of bread dough and their relation to baking. In “Rheology of Food, Pharmaceutical and Biological Materials with General Rheology” (R. E. Carter, ed.), pp. 269-283. Elsevier, New York. De Kee, D., Turcotte, G., and Fildey. K. (1980). New method for the determination of yield stress. J. Text. Stud. 10, 281. DeMan, J. M., ed. (1976). “Rheology and Texture in Food Quality.” Avi Publ. Co., Westport, CT. de Vargas, L., Perez-Gonzalez, J., and Romero-Barenque, J. D. (1993). Experimental evidence of slip development in capillaries and a method to correct for end effects in the flow of xanthan solution. J. Rheol. 37(5), 867-878. Dickie, A., and Kokini, J. L. (1982). Use of the Bire-Leider equation in food rheology. J . Food Process. Eng. 5, 157-174. Doi, E., and Edwards, R. H. (1978). J. Chem. Soc., Faruday Trans. 2 74,1789. Doi, E., and Edwards, R. H. (1979). J. Chem. Soc., Furuduy Trans. 2 75,38. Doi, E., and Edwards, R. H. (1986). “The Theory of Polymer Dynamics.” Oxford Univ. Press, Oxford. Doublier, J. L., and Launay, B. (1976). Rheological properties of galactomannan in aqueous solutions: Effect of concentration and molecular weight. Proc. Int. Congr. Rheol., 6th, pp. 532-533. Doublier, J. L., and Launay, B. (1981). Rheological galactomannan solutions: Comparative study of guar gum and locust bean gum. J. Text. Stud. U,151-172. Drake, M. A,, Ma, L., Swanson, B. G., and Barbosa-CBnovas, G. V. (1994). Rheological characteristics of milkfat and milkfat blend sucrose polyesters. Food Res. Znt. 27,477. Dus, S. J., and Kokini, J. L. (1990). Prediction of the non-linear viscoelastic properties of a hard wheat flour dough using the Bird-Carreau constitutive model. J. Rheol. 34(7), 1069-1084. Elliott, J. H., and Ganz, A. J. (1977). Research note: Salad dressing-Preliminary rheological characterization. J. Text. Stud. 8, 359-371.
THE RHEOLOGY O F SEMILIQUID FOODS
65
Escher, F. (1983). Relevance of rheological data in food processing. In “Physical Properties of Foods” (R. Jowitt, F. Escher, B. Hallstrom, H. T. Th. Meffert and G. Vos, eds.), pp. 103-110. Applied Science, New York. Evans, I. D. (1992). Letter to editor: On the nature of the yie1dstress.J. Rheol. 36(7), 1313-1316. Fernandez-Martin, F. (1972). Influence of temperature and composition on some physical properties of milk concentrates. I. Viscosity. J. Dairy Rex 39, 75. Ferry, J. D. (1980). “Viscoelastic Properties of Polymers.” Wiley, New York. Figoni, P. I., and Shoemaker, C. F. (1983). Characterization of time dependent flow properties of mayonnaise under steady shear. J. Text. Stud. 14, 431-442. Frederickson, A. G. (1964). “Principles and Applications of Rheology.” Prentice-Hall, New York. Gencer, G. K. (1985). Interactions of selected gums in solution and the effect of common food ingredients on these interactions. Ph.D. Dissertation, University of Massachusetts, Amherst. Gladwell, N., Rahalkar, R. R., and Richmond, P. (1985a). Rheological behavior of soya oil water emulsion: Dependence upon oil concentration. J. Food Sci. 50, 440-443. Gladwell, N., Rahalkar, R. R., and Richmond, P. (1985b). Creephecovery behavior of oilwater emulsions: Influence of disperse phase concentration. J. Food Sci. 50, 1477-1481. Glucklich, J.. and Shelef, L. (1962). An investigation into the rheological properties of flour dough: Studies in shear and compression. Cereal Chem. 39,243-255. Gould, W. A. (1983). “Tomato Production, Processing and Quality Evaluation,” 2nd ed. Avi Publ. Co., Westport, CT. Green, H., and Weltmann, R. N. (1943). Analysis of the thixotropy of pigment vehicle suspensions: Basic principles of the hysteresis loop. Ind. Eng. Chem., Anal. Ed. 15, 201-206. Harper, J. C. (1960). Viscometric behavior in relation to evaporation of fruit purees. Food Technol. 14, 557-561. Harris, J. (1972). A continuum theory of structural change. Rheol. Acta 11, 145-151. Hartnett, J . P., and Hu, R. Y. Z. (1989). Technical note: The yield stress-an engineering reality. J. Rheol. 33(4), 671-679. Hibberd, G. E., and Wallace, W. J. (1966). Dynamic viscoelastic behavior of wheat flour doughs. Part I. Linear aspects. Rheol. Actu 5, 193-198. Higgs, S. D. (1974). An investigation into the flow behavior of complex non-Newtonian foodstuffs. J. P h y ~ D . 7, 1184-1191. Hlynka, I., and Anderson, J. A. (1952). Relaxation of tension in stretched dough. Can. J. Tech. 30, 198. Huang, H., and Kokini, J. L. (1993). Measurement of biaxial extensional viscosity of wheat flour doughs. J. Rheol. 37,879-891. Ibarz, A,, Vicente, M., and Graell, J. (1987). Rheological behavior of apple juice and pear juice and their concentrates. J. Food Eng. 6,257-267. Ibarz, A., Pagan, J., Guitierrez, J., and Vicente, M. (1989). Rheological properties of clarified pear juices concentrates. J. Food Eng. 10,57-63. Ibarz, A., Gonzalez, C., Esplugas, S., and Vicente, M. (1992). Rheology of clarified fruit juices. 1. Peach juices. J. Food Eng. 15,49-61. Kee, De D., and Durning, C. J. (1990). Rheology of materials with a yield stress. In “Polymer Rheology and Processing” (A. A. Collyer and L. A. Utracki, eds.). Elsevier, New York. Kirkwood, J. G.. and Auer, P. L. (1951). The visco-elastic properties of solutions of rod-like macromolecules. J. Chem. Phys. 19,281-283. Kokini, J. L. (1992). Rheological properties of foods. In “Handbook of Food Engineering” (D. R. Heldman and D. B. Lund, eds.), 1st ed. Dekker, New York.
66
GUSTAVO V. BARBOSA-CmOVAS et af.
Kokini, J. L. (1993). Constitutive models for dilute and concentrated food biopolymer systems. In “Plant Polymeric Carbohydrates” (F. Meuser, D. J. Manners, and W. Seibel, eds.), pp. 43-75. Royal Society of Chemistry, Cambridge, UK. Kokini, J. L., and Chou, T. C. (1993). Comparison of the conformation of tomato pectins with apple and citrus pectins. J. Text. Stud. 24,117-137. Kokini, J. L., and Dickie, A. (1981). An attempt to identify and model transient viscoelastic Bow in foods. J. Text. Stud. 12,539-557. Kokini, J. L., and Plutchok, G. J. (1987a). Predicting steady and oscillatory shear rheological properties of cmc/guar blends using the Bird-Carreau constitutive model. J. Text. Stud. 18931-42. Kokini, J. L., and Plutchok, G. J. (1987b). Viscoelastic properties of semisolid foods and their biopolymeric components. Food Technof. 41(3), 89-95. Kokini, J. L., Bistany, K. L., and Mills, P. L. (1984). Predicting steady shear and dynamic viscoelastic properties of guar and carrageenan using the Bird-Carreau constitutive model. J. Food Sci. 49, 1569-1576. Konstance, R. P., and Holsinger, V. H. (1992). Development of rheological test methods for cheese. Food Technol. 46(1), 105-109. Lang, E. R., and Rha, C. (1981). Determination of the yield stress of hydrocolloid dispersions. J. Text. Stud. 12,47. Launay, B., and Pasquet, E. (1982). Interaction of hydrocolloids. In “Gums and Stabilizers for Food Industry 1” (G. 0. Phillips, D. J. Wedlock, and P. A. Williams, eds.). Elsevier, New York. LeGrys, G. A., Booth, M. R., and Al-Baghdadi, S. M. (1981). The physical properties of wheat proteins. In “Cereals: A Renewable Resource” (Y. Pomeranz and L. Munck, eds.). AACC, St. Paul, MN. Leider, P. J., and Bird, R. B. (1974). Squeezing flow between parallel disks. I. Theoretical analysis. lnd. Eng. Chern. Fundarn. 113,336-341. Leppard, W. R. (1975). Viscoelasticity: Stress measurements and constitutive theory. Ph.D. dissertation, University of Utah, Salt Lake City. Liguori, C. A. (1985). The relationship between the viscoelastic properties and the structure of sodium alginate and propylene glycol alginate. M. S. Thesis, Rutgers University, New Brunswick, NJ. Ma, L., and Barbosa-Canovas, G. V. (1995a). Rheological characterization of mayonnaise. Part I. Slippage at different oil and xanthan gum concentrations. J. Food Eng. 25,397-408. Ma, L., and Barbosa-Canovas, G. V. (1995b). Rheological characterization of mayonnaise. Part 11. Flow and viscoelastic properties at different oil and xanthan gum concentrations. J. Food Eng. 25,409-425. Marvin, R. S., and McKinney, J. E. (1965). In “Physical Acoustice” (W. P. Mason, ed.), Vol. B. Academic Press, New York. Mason, P. L., Bistany, K. L., Puoti, M. G., and Kokini, J. L. (1982). A new empirical model to simulate transient shear stress growth in semi-solid foods. J. Food Process. Eng. 6,219-33. Mewis, J. (1979). Thixotropy-A general review. J. Non-Newtonian Fluid Mech. 6, 1-20. Missaire, F., Qiu, C.-G., and Rao, M. A. (1990). Yield stress of structured and unstructured food suspensions. J. Text. Stud. 21, 479-490. Mitchell, J. R. (1979). “Rheology of Polysaccharides Solutions and Gels. Polysaccharides in Food.” Butterworth, London. Mooney, M. (1931). Explicit formulas for slip and fluidity. J. Rheof. 2, 210-222. Muller, H. G. (1973). “An Introduction to Food Rheology.” Crane, Russak & Co., New York.
THE RHEOLOGY O F SEMILIQUID FOODS
67
Munoz, J., and Sherman, P. (1990). Dynamic viscoelastic properties of some commercial salad dressing. J. Text. Stud. 21, 411-426. Navickis, L. L., and Bagley, E. B. (1983). Yield stresses in concentrated dispersions of closely packed, deformable gel particles. J. Rheol. 27(6). 519-536. Nussinovitch, A., Peleg, M., and Normand, M. D. (1989). A modified Maxwell and a nonexponential model for characterization of the stress relaxation of agar and alginate gel. J. Food Sci. 54,1013-1015. Oakenfull, D., and Scott, A. (1988). Size and stability of the junction zones in gels of iota and kappa carrageenan. In “Gums and Stabilizers for Food Industry 4” (G. 0. Phillips, D. J. Wedlock, and P. A. Williams, eds.). Elsevier, New York. Ofoli, R. Y., Morgan, R. G., and Steffe, J. F. (1987). A generalized rheological model for inelastic fluid foods. J. Text. Stud. 18, 213-230. Paredes, M. D. C., Rao, M. A., and Bourne, M. C. (1988). Rheological characterization of salad dressings. 1. Steady shear, thixotropy and effect of temperature. J. Text. Stud. 19,247-258. Paredes, M. D. C., Rao, M. A., and Bourne, M. C. (1989). Rheological characterization of salad dressing. 2. Effect of storage. J. Text. Stud. 20,235-250. Peleg, M. (1977). Operational conditions and the stress strain relationship of solid foodstheoretical evaluation. J. Tent. Srud. 8, 283-295. Plutchok, G. J., and Kokini, J. L. (1986). Predicting steady and oscillatory shear rheological properties of CMC and guar gum blends from concentration and molecular weight data. J. Food Sci. 51, 1284-1288. Princen, H. M. (1986). A novel design to eliminate end effects in concentric viscometer. J . Rheol. 30,271-283. Prud’homme, R. K. (1991). Rheological measurement. In “Polymers as Rheology Modifiers” (D. N. Schulz and J. E. Glass, eds.). Maples Press, New York. Qiu, C. G., and Rao, M. A. (1988). Role of pulp content and particle size in yield stress of apple sauce. J. Food Sci. 53, 1165-1170. Qui, C. G., and Rao, M. A. (1989). Effect of dispersed phase on the slip coefficient of apple sauce in a concentric cylinder viscometer. J. Texture Stud. 20, 57-70. Qui, C. G., and Rao, M. A. (1990). Quantitative estimates of slip of food suspensions in a concentric cylinder viscometer. In “Engineering and Food,” (W. E. L. Spiess and H. Schubert, eds.), Vol. 1. Elsevier, New York. Race, S. W. (1993). Bohlin Instruments Application Notes No. 7. Rahalkar, R. R., Javanaud, C.. Richmond, P., Melville, I., and Pethrick, R. A. (1985). Oscillatory shear measurements on concentrated dextran solutions: Comparison with Doi and Edwards’ theory reptation. J. Rheol. 29(6), 955-970. Rani, U., and Bains, G. S. (1987). Flow behavior of tomato ketchups. J. Text. Stud. 18,125-135. Rao, M. A., and Cooley, H. J. (1983). Applicability of flow models with yield of tomato concentrates. J. Food Process. Eng. 6, 159-173. Rao, M. A., Bourne, M. C., and Cooley, H. J. (1981). Flow properties of tomato concentrates. J. Text. Stud. 12, 521-538. Rao, M. A., Colley, H. J., and Vitali, A. A. (1984). Flow properties of concentrated juices at low temperatures. Food Technol. 38(3), 113-119. Rha, C. K. (1978). Rheology of fluid foods. Food Technol. 7, 32-35. Rohm, H., and Weidinger, K. H. (1993). Rheological behavior of butter at small deformations. J. Text. Stud. 24, 157-172. Rosenberg, M., Wang, Z., Chuang, S. L., and Shoemaker, C. F. (1995). Viscoelastic property changes in cheddar cheese during ripening. J. Food Sci. 60(3), 640-644.
68
GUSTAVO
v. BARBOSA-CANOVAS
et al.
Rouse, P. E. (1953). A theory of the linear viscoelastic properties of dilute solutions of coiling polymer. J. Chem. Phys. 21, 1272. Saravacos, G. D. (1970). Effect of temperature on viscosity of fruit juices and purees. J. Food Sci. 35, 122. Schofield, R. K., and Scott-Blair, G. W. (1932). The relationship between viscosity, elasticity and plastic strength of soft materials as illustrated by some mechanical properties of flour doughs. Proc. R. SOC. London, Ser. A 138,707-717. Schurz, J. (1992). Letter to the editor: A yield value in a true solution. J. Rheol. 36(7), 13191321. Senouci, A., and Smith, A. C. (1988). An experimental study of food melt rheology. 11. End pressure effects. Rheol. Acta 27, 649-655. Shama, F., and Sherman, P. (1966). The texture of ice cream. 11. Rheological properties of frozen ice cream. J. Food Sci. 31, 699-716. Sherman, P. (1970). “Industrial Rheology with Particular Reference to Foods, Pharmaceuticals and Cosmetics.” Academic Press, New York. Shoemaker, C. F., Lewis, J. I., and Tamura, M. S . (1987). Instrumentation for rheological measurements of food. Food Technol. 41(3), 80. Shrimanker, S. H. (1989). M. S. Thesis, Rutgers University, New Brunswick, NJ. Smith, J. R., Smith, T. L., and Tschoegl, N. W. (1970). Rheological properties of wheat flour doughs. 111. Dynamic shear modulus and its dependence on amplitude, frequency and dough composition. Rheol. Acta 9,239-252. Steffe, J. F. (1992a). Yield stress: Phenomena and measurement. In “Advances in Food Engineering” (R. P. Singh and M. A. Wirakartakusumah, eds.), pp. 363-376. CRC Press, London. Steffe, J. F. (1992b). Viscoelasticity. I n “Rheological Methods in Food Process Engineering” (J. F. Steffe, ed.), 1st ed., pp. 168-194. Freeman Press, East Lansing, MI. Steffe, J. F., Mohamed, I. O., and Ford, E. Q. (1986). Rheological properties of Ruid foods: Data compilation. In “Physical and Chemical Properties of Food” (M. R. Okos, ed.), ASAE Publ., p. 1. Am. SOC.Agric. Eng., St. Joseph, MI. Struebi, P., Escher, F., and Neukom, H. (1978). Use of macerating pectic enzyme in apple nectar processing. J. Food Sci. 43, 260. Szczesniak, A. S. (1977). Rheological problems in the food industry. J. Text. Stud. 8, 119. Tiu, C., and Boger, D. V. (1974). Complete rheological characterization of time dependent food products. J. Text. Stud. 5,329-338. Tschoegl, N. W., Rinde, J. A., and Smith, T. L. (1970a). Rheological properties of wheat flour doughs. I. Method for determining the large deformation and rupture properties in simple tension. J . Sci. Agric. 21, 65-70. Tschoegl, N. W., Rinde, J. A., and Smith, T. L. (1970b). Rheological properties of wheat flour doughs. 11. Dependence of large deformation and rupture properties in simple tension on time, temperature and water absorption. Rheol. Acta 9,223-238. Ullman, R. (1969). The viscoelastic properties of solutions of rod-like macromolecules of finite diameter. Macromolecules 2(1), 27-30. Vernon-Carter, E. J., and Sherman, P. (1980). Rheological properties and applications of mesquite tree (Prosopis julifora) gum. 2. Rheological properties and stability of o/w emulsions containing mesquite gum. J. Text. Stud. 11,351-365. Vitali, A. A., and Rao, M. A. (1984). Flow properties of low pulp concentrated orange juice: Effect of temperature and concentration. J. Food Sci. 49,882-888. Waiters, K. (1975). “Rheometry.” Chapman & Hall, London. Whorlow, R. W. (1980). “Rheological Techniques.” Wiley, New York. Yamakawa, H. (1975). Viscoelastic properties of straight cylindrical macromolecules in dilute solutions. Macromolecules 8,339-342.
THE RHEOLOGY OF SEMILIQUID FOODS
69
Yoshimura, A. S., Prud’homrne, R. K., Princen, H. M., and Kiss, A. D. (1987). A comparison of techniques for measuring yield stresses. J . Rheol. 31(8), 699-710. Yoshimura, A. S., and Prud’homrne, R. K. (1988). Wall slip corrections for couette and Parallel disk viscometers. J . Rheol. 32( l), 53-67. Zirnm, B. H. (1956). Dynamics of polymer molecules in dilute solutions: Viscoelasticity, Bow birefrigence and dielectric loss. pp. 241-269.
This Page Intentionally Left Blank
ADVANCES IN FOOD AND NUTRITION RESEARCH, VOL. 39
CONTROL OF THE DEHYDRATION PROCESS IN PRODUCTION OF INTERMEDIATE-MOISTUREMEAT PRODUCTS: A REVIEW S. F. CHANGt AND T. C. HUANG Department of Food Science and Technology Pingtung Agricultural Institute Pingtung, Taiwan
A. M. PEARSON Depariment of Animal Sciences Oregon State University Corvallis, Oregon 97331
I. Introduction 11. Traditional Production of IM Meat Products
111.
IV.
V.
VI.
A. IM Meat Products of Turkey B. Chinese and Malaysian IM Meats C . IM Meat Products of European Origin D. Some IM Meat Products of Africa E. North and Latin American IM Meats Technology of Producing IM Pet Foods A. Principles Involved in Production of IM Pet Foods B. Humectants Used in IM Pet Foods C. Mycostats Preservation Principles and Their Application to IM Meats A. Stabilization during Processing B . Comparison of Heated and Raw Products Problems in Production of Different IM Meats A. Processing Equipment and Raw Materials B. Improvement of Processes and Products Effects of Slaughtering, Handling, Chilling, Freezing, Storage, and Thawing on Muscle Properties A. Changes in Muscle Structure and the Moisture Component
t Deceased. 71 Copyright 0 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
72
S. F. CHANG ei al.
VI1.
VIII.
IX.
X.
XI.
XII.
XIII.
B. Effects of Rigor Mortis C. Freezing and Storage D. Raw Prerigor Muscle Effects of Predfying Treatment and Handling of Muscle A. Chopping or Grinding B. Role of Myofibrillar Proteins in Water Binding C. Fermentation D. Effects of Heating E. Influence of Osmotic Treatment Factors Influencing Absorption Phenomena in Meats/Meat Mixtures A. Bound and Unbound Water in Meat Products B. Relationship of Water Vapor Pressure to Water Activity C. Sorption Phenomena in Multicomponent Food Systems D. Sorption Phenomena in Meat Systems E. Influence of Muscle Protein Denaturation on Sorption Isotherms Mechanisms Involved in Meat Dehydration Systems A. Moisture Removal during IM Meat Processing B. Drying Rate Curves for Heated Muscle Bundles C. Physical and Structural Changes in IM Meats D. Dehydration Parameters and Profiles in Relation to Physical Changes during Dehydration of IM Meats E. Dehydration in Production of IM Meats F. Fermented IM Meats G . Dehydration by Heated Osmosisihfusion of IM Meats H. Morphological Changes in Muscle Bundles during Heating-Drying Quality Attributes as Affected by Dehydration and Its Associated Processes A. Q, as an Important Parameter B. Influence of Nonthermodynamic Factors on Quality of IM Foods C. Quality Attributes of IM Meats D. Control of Microbial Growth in IM Meats E. Other Considerations for Controlling Microbial Growth in Some IM Meats F. Texture of IM Meats G . Effect of Precooking H. Use of Humectants I. Color J. AromaFlavor Development and Retention K. Nutritive Value Process Optimization for IM Meats A. Background B. Concept for Optimization for Heat Processing of Meat C. Mathematical Modeling for Heat and Mass Transfer D. Development of an Appropriate Kinetic Model Energy Costs for Production of IM Meat Products A. Energy Consumption in Food Processing B. Energy Costs Associated with IM Meats C. Factors Affecting Dryer Selection Research Needs in IM Meats A. Effects of Ante- and Postmortem Treatments on Properties of IM Meats
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
73
B. Influence of Freezing and Thawing C. Fermentation D. Synergistic Stabilization E. Effects of Heat F. Effects of Osmotic Treatments G. Gel Formation H. Role of Glass Transition I. Modeling J. Design of Drying Equipment K. Thermal Data L. HACCP Systems XIV. Summary References
I.
INTRODUCTION
Intermediate-moisture (IM) meats are temperature stable products with a moisture content around the equilibrium moisture content of the meat mixture at ambient temperature and humidity. The products can be consumed as they are without rehydration, having a desirable texture without brittleness or overdryness. Most of the traditional IM meat products have evolved from natural drying of the meat mixture. In the drying process, the ultimate water activity (a,,,) of IM meat approaches 0.60 to 0.90, which is equivalent to a relative humidity (RH) of 60-90% at ambient temperature (Leistner, 1987). The products do not require either strict moisture impermeable packaging or refrigeration during storage. Indeed the products are processed and dried to a stable state and can be kept at ambient temperatures without spoiling. There are many traditional IM meats in all parts of the world. Their production varies with the climatic conditions and the economic and technological status in each country. Man observed naturally sun-dried grains and fruits before learning to dry fish and thin slices of meat by hanging them in the air and sun. Drying of these animal products usually took a long time, so that bacterial spoilage during the time-consuming operation often occurred (Shin and Leistner, 1983). Thus, the use of salt, sugar, and smoke as further preservative agents gradually evolved in combination with drying. Sun-drying is still in use in many parts of the world including the United States. While sun-drying in some parts of the world and for certain products is the most economical method of drying, it has several obvious disadvantages according to Leistner (1987):
1. Sun-drying is dependent upon the elements;
74
S. F. CHANG et al.
2. It is slow and not suitable for many high quality products; 3. It generally will not lower the moisture content below about 15%, which is too high for storage of numerous food products; 4. It requires considerable space; and, 5 . The food is subject to contamination and loss from dust, mold, insects, rodents, and the like.
There has been a renewal of interest in IM meat products as costs of refrigeration have increased and with the development of new technologies. It is the purpose of this review to update and consolidate the latest information on drying procedures that have been used or show promise for producing IM meat products.
II. TRADITIONAL PRODUCTION OF IM MEAT PRODUCTS
Processing equipment used in traditional meat drying is simplistic, and in some cases primitive, making process control difficult, if not impossible. The processor is at the mercy of the weather for sun or smoke-fire drying. Product quality is uncertain and variable, so standards are met only by luck and sheer artistry. Yet, nearly every country in the world produces its own characteristic IM meat product. A few products from different countries are used here as examples of successful IM meat products. However, limitations in manufacturing are frequently encountered making their production difficult or self-limiting. Generally these products are safe, although they may have an abundant microflora (Gibbs, 1986). A. IM MEAT PRODUCTS OF TURKEY Production of pastirma or basturma, which is an IM meat product of Turkey, has been described by Leistner (1987). It is preferably produced from September to November, since flies are not prevalent during this season. The air temperature is not as high as in the summer, and the relative humidity is moderate due to scanty rainfall. According to Leistner (1987) pastirma is produced by dry curing thin strips (50-60 cm long by 5 cm in diameter) of meat at ambient temperature. The strips of meat are dry-cured followed by pressure treatment and air drying. After salting, drying, and application of pressure, the meat is covered with 3-5 mm of a paste containing garlic and other spices and dried for 1 day in a pile and hung for 5-12 days in a room with good ventilation. The product is then ready for distribution and consumption. The final product has an a, of 0.85-0.90.
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
75
Berkman (1960) investigated the survival of salmonellae, anthrax bacilli, pathogenic clostridia, rinderpest virus, and tapeworm larvae in pastirma and found this product was virtually absent of these organisms. Krause et af. (1972) recovered micrococci and lactobacilli, but rarely any Enterobacteriaceae. The lactic acid producing organisms appear to play an important role in assuring the safety of this product. Nevertheless, the problems encountered in producing pastirma and the labor intensive operation limit its production. Genigeorgis and Lindroth (1984) have investigated the safety of basturma from salmonella and found it is generally safe. B. CHINESE AND MALAYSIAN IM MEATS 1. Slice-Cured Chinese and Indonesian (Dending) Products
Slice-cured meat (pork or beef) is produced for the Chinese and Malaysian markets (Leistner, 1987; Chuah et af., 1988; Wang and Chen, 1989) and utilizes essentially the same process as Indonesian dending (Buckle et al., 1988). The sliced-cured Chinese product has been described by Leistner (1987) and can be produced by one of two methods. The first procedure utilizes paper-thin (0.2 cm) slices of lean meat (pork or beef) cut parallel to the grain. The meat is cured with a mixture of sugar, salt, soy sauce, monosodium glutamate, and spices. The traditional cure does not use nitrite or nitrate. The meat is cured for 24-36 hr at 4°C and then the thin strips are placed side by side (slightly overlapping) on an oiled bamboo basket or wire rack and dried at 5O-6O0C, until they lose approximately 50% of their original weight. The meat is removed from the containers and further dried by cooking over charcoal or deep fat fried to give a final a , of about 0.69. The product is then ready for distribution. Dending of Indonesia is produced by essentially the same process, although procedures and formulation may vary slightly as described by Buckle et al. (1988).
2. Cooked and Dried Meat Cubes In another process described by Leistner (1987) and Buckle et af. (1988), the whole muscle is boiled for 40-45 min, after which it is cut into cubes or pieces (5 x 5 x 10 cm). Although beef is preferred because of its fibrous nature, pork and chicken are sometimes used (Lo, 1980). The cubes are added to the cure in a steam kettle and cooked until nearly all of the cure has evaporated. The meat is removed from the steam kettle and dried in a hot air dehydrator. The final a , is about 0.69. Leistner (1987) concluded that an a , < 0.69 is critical for Chinese dried meats, although Ho and Koh (1984) suggested that an a , < 0.61 is needed to prevent mold growth.
76
S. F. CHANG et al.
3. Zousoon and Dried Pork Floss
Two other Chinese products are produced. One is a semidry product called Zousoon and the other a dry product known as dried pork floss. The composition and water activity of these products are shown in Table I. Production of these two products has been described briefly by Chang et al. (1991). Both products traditionally are prepared from prerigor pork, which is boned, cut into pieces parallel to the muscle fibers, added to a gas-fired scraping frypan along with a small amount of water, sugar, and salt, and cooked to the desired degree of dryness. Zousoon has an a, of about 0.65 and dried pork floss of about 0.45. Zousoon is more desirable because of its fibrous texture, lighter color, better flavor, and higher yield (Chang et al., 1991). 4. Lup Cheong (La Zang)
Lup Cheong (Cantonese) or La Zang (Mandarin) is the most typical Chinese sausage according to Leistner (1987), with some variation being produced in different areas (Lo, 1980; Ho and Koh, 1984). Leistner (1987) described production of Lup Cheong, which is a raw nonfermented sausage made from coarsely ground pork (preferably ham) and pork fat mixed with sugar, salt, soy sauce, Chinese wine, potassium nitrate, five-spice powder (anise, clove, fennel, and watchau), and monosodium glutamate. Up to 25% water is sometimes added to give a wrinkled appearance to the sausage after drying. The mixture is stuffed into small hog casing tied at 15-cm intervals. The casings are punctured regularly to allow the escape of entrapped air and water during drying. It is dried by heating over charcoal at TABLE I PROXIMATE COMPOSITION AND WATER ACTIVITY OF ZOUSOON (SEMI DRY CHINESE LONG-FIBER PRODUCT) AND DRIED PORK FLOSS (DRY ROASTED SHORT-FIBERED CHINESE
Protein content (above) (%) Carbohydrate content (below) (%) Fat content (below) (%) Ash content (below) (%) Moisture content (below) (%) a, (range) a
PRODUCT)"^
Zousoon (semi dry)
Dried pork Floss (dry)
47
32 15
16 12 9 13 0.60-0.65
43b 7 4 0.40 or below
Source: Chang et al. (1991). Reproduced with kind permission from Elsevier Science Ltd. After adding lard during finishing process.
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
77
45-50°C for 1-2 days and then held at room temperature for 2-3 days to allow for moisture equilibration. Lup Cheong can be stored for 1-3 months without refrigeration if mold growth is controlled. It has a reddish-brown color and has a fat speckled appearance. It is always heated before consumption and is consumed sliced and steamed with rice. It has an intense aroma and a few slices are sufficient to season an entire dish. Leistner (1987) concluded that the microbiological stability of Lup Cheong mainly is due to rapid reduction in a,,,, which is aided by its salt (2.8-3.5%) and sugar (1-10%) contents, the thin casings (26-28 mm), the high ripening temperature (45-5OoC), and a low relative humidity (65-75%). The pH is not important since it is relatively high (5.7-5.9) and the lactic acid bacteria count is low (<106/g). Leistner (1987) reported an average a,,,of 0.75 and a pH of 5.9 for 24 samples imported from China and analyzed. Ho and Koh (1984) reported an a,,, of 0.6-0.7 for Lup Cheong samples produced in Singapore, which was probably due to the higher sugar content (1520%). Leistner (1987) stated that the number of Gram-positive bacteria in the raw meat should be moderate. The a,,,must be decreased to C0.92 within 12 hr and to <0.90 by 36 hr. This can be achieved by drying for 36 hr at 48°C and 65% relative humidity. If drying is not carried out over charcoal, it should be lightly smoked at 48°C and 65% relative humidity. The product should be held at 20-25°C and 75% relative humidity for 3 days to allow for moisture equilibration or until an a,,,of <0.80 is achieved. The sausage should then be vacuum-packaged, since this improves the flavor and inhibits mold growth. C. IM MEAT PRODUCTS OF EUROPEAN ORIGIN Fermented meat products have been produced in Europe for centuries, with the bacteria being indigenous to each processing plant and producing a characteristic product. Natural fermentation was later controlled to some extent by using small amounts of product from a previous batch to give the same bacterial flora, which is called backslopping (Romans el at., 1985). Salami is an example of a semidry (IM) sausage. Today most semidry sausages are produced adding pure or mixed bacterial cultures, which gives better control of the fermentation process as explained by Bacus (1986). These fermented products are preserved by a combination of low pH and drying. 1. Salami and Other Fermented Products
Pearson and Gillett (1995) have described the processing of salami. The meat is first ground or chopped and the cure is added during the process.
78
S. F. CHANG er al.
The cure is composed of salt, sugar, and nitrite or nitrate. The culture is also added during the chopping process. Seasonings are added as desired and the product is stuffed into synthetic or natural casings. Fermentation takes place in a fermentation or “greening room” held at 24°C and 75% RH for 36 hr. The sausage is then dried for 9-10 weeks in a drying room at 10°Cand 75% RH. Salamis and other fermented European-type sausages have an a, of 0.62-0.80 and are now widely produced in other countries. The use of nitrite and salt helps to inhibit the growth of the food-poisoning bacteria during ripening (Leistner, 1978,1987,1990a).However, the production of lactic acid during fermentation lowers the pH and provides further safety as the a , is lowered during drying (Bacus, 1986; Leistner, 1987). 2. Other Fermented European Sausages
Although many fermented semidry and dry sausages originated in Southern Europe (Romans et af., 1985; Bacus, 1986), they are now widely produced in other developed countries. Cervelat or summer sausage originated in Germany, but it is now the most popular semidry sausage according to Romans et al. (1985). Cervelat is made from a mixture of beef and pork and usually utilizes a culture. It contains salt, sugar, and nitrite and characteristically contains whole black peppers. After fermentation it is smoked and dried to give an a,,,of about < O M . Pepperoni is also a common IM sausage that is now widely produced, although it originated in Italy as explained by Romans et al. (1985). It typically contains pork, beef, salt, sugar, nitrite, and spices and is fermented while holding in a greening room, followed by smoking and drying. Pepperoni may reach an a, of about 0.65 and shrinks by about 35%(Pearson and Gillett, 1995). Typical Italian pepperoni is dried but not smoked (Romans et al., 1985). Farmer sausage originally was produced by farmers in northern Europe and commonly is made of 65%beef and 35%pork, which is chopped medium fine, seasoned, stuffed in beef middles, and heavily smoked (Romans et af.,1985).The smoking and drying of quite small pieces produces a relatively stable product with an a, of 0.85. Holsteiner is a similar product, but the ends are fastened together (Romans et al., 1985). Mortadella originated in Bologna, Italy, according to Romans et al. (1985). It is typically made of 75% pork and 25% beef with garlic in the seasonings. It can also be manufactured using turkey, chicken, goat, or mutton and often contains green pistachio nuts and/or olives. Traditionally it is fermented and dried or cooked as described by Pearson and Gillett (1995). It normally has an a, of >0.85 and is popular with Italian and Spanish populations. It is a large volume product in the world market.
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
79
Rosselld et at. (1995) have described the production process of Sobrasada, which is typically produced in the Balearic Island off the coast of Spain. It usually contains approximately 50% fresh raw lean pork, 40% fresh pork fat, 5% paprika, 2% salt, and 5% white pepper. The lean and fat are finely ground and kneaded together until the mixture has a paste-like texture, after which the paprika, salt, and white pepper are mixed with the meat and stuffed into natural casings. The mixture is held at about 4°C for 24 hr and ripened at ambient winter temperature of about 8-15%"C and a relative humidity of 60-85% for about 4 months. The pH falls rapidly to about 5.3 during the first few days. The initial a , falls from about 0.93 to 0.88 to 0.83, which coincides with the loss of about 15% of the initial water content. The lactic acid bacteria become the predominant microflora and by the end of about 30 hr enterobacteriaceae organisms are virtually absent. 3. Low Acid IM Meats
Some IM meat products from Europe are dried without fermentation and have a final pH of about 6.0 and a, values in the range of 0.87-0.90 as outlined by Incze (1991, 1992). Drying is initiated at low temperatures to prevent spoilage and only later during processing is the temperature increased to accelerate drying. Long-cured hams are cured with a mixture of salt, sugar, and nitrate at temperatures below 5°C until they are placed in the smokehouse after being partially preserved by salt and then smoked until the a, reaches >0.90. Low acid sausages are first refrigerated 6°C and then when cured are smoked to further reduce the a, to about 0.88. The heat is added gradually as explained by Incze (1991,1992). Although sugar may be added to both long-cured hams and low-acid sausages, it is not necessary. The final pH is unimportant, but usually is about 6.0. Thus, low-acid IM meat products are produced by salting and low temperatures during the early stages of preservation and by salt and drying in the latter stage of processing. Nitrite may also provide an additional hurdle as explained by Leistner (1978). Prosciutti and capacola are examples of low-acid IM meats originating in Europe.
D. SOME IM MEAT PRODUCTS FROM AFRICA 1. Biltong An IM meat product produced in South Africa is known as biltong. Its production is described by Ledward (1981, 1985), Leistner (1987), Obanu
80
S.
F. CHANG et al.
(1988), and Van der Riet (1982). It is usually made from long strips of beef muscle, which is commonly cured by dry salt, although sugar and spices are often added. Nitrite or nitrate may be used to stabilize the color of biltong. South African regulations allow addition of 0.1% potassium sorbate to prevent mold growth. The original procedure as described by Obanu (1988) points out that sun-drying is common and is carried out by hanging the salted strips on barbed wire or galvanized wire fences. Van den Heever (1970) analyzed 70 samples of biltong and found an average of 25% moisture, 6.6% salt, an a, of 0.74, and a pH of 5.9, which can be compared to 23% moisture, 5.6% salt, an a, of 0.70, and a pH of 5.7 for 20 samples analyzed by Van der Riet (1976a,b). Shin and Leistner (1983) and Shin (1984) found 25 biltong samples had a, values ranging from 0.36 to 0.93, with most samples falling within a range of 0.65-0.85. The pH varied from 4.8 to 5.8, with most being around 5.5. Salt levels ranged from 5 to 15%,with an average of 7%. Some 32% of the samples were spoiled by yeasts, molds, and Micrococcaceae during transport and storage. Several salmonella species have been reported to be present in biltong by a number of investigators (Bokkenheuser, 1963; Van den Heever, 1965, 1970;Prior and Badenhorst, 1974). Salmonellosisin humans has been traced to biltong by Bokkenheuser (1963) and Botes (1966), which emphasizes the importance of good hygiene and sound manufacturing practices. Although Aspergillus flavus is frequently isolated from biltong, aflatoxins are not normally found in biltong (Prior and Badenhorst, 1974;Leistner et al., 1981).
2. Kilishi and Bande Okonkwo (1984) and Obanu (1988) have discussed production of kilishi and bande, which are made by sun-drying and hot smoking-cooking, respectively, in Africa. Igene et al. (1992) have described the traditional production of kilishi as produced in Nigeria. Typically these products are made from lean beef or goat meat, which is cut into thin strips and dried in the sun or smoked-cooked over a low fire until they are preserved. Salt may be added to hasten drying, but often is not added. Smoking may also aid in preservation. Flies are normally a problem, so smoking-cooking is preferred. E. NORTH AND LATIN AMERICAN IM MEATS Four IM meat products that are indigenous to North America include pemmican, jerky, country hams, and Lebanon bologna.
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
81
1. Pemmican Pemmican is a product originally made by American Indians and its production is described by Ashbrook (1955) and Binkerd et al. (1976). It was made from lean buffalo meat or venison. Processing is carried out by either sun-drying or smoking at low temperatures followed by pounding the dried meat into a shredded mass. It then had dried fruit pounded into the dried meat and was embedded in melted fat. It was sewed in rawhide bags and used by Indians on the warpath or in times of scarcity and later by mountain men and Arctic and Antarctic explorers (Stefansson, 1956). Although interest in pemmican was revived during World War 11, it is no longer produced. 2. Jerky and Dried Beef Ashbrook (1955) also has described production of jerky (beef or venison), which was cured in a hot brine and smoked over a low burning fire. Production of dried beef was also described by Ashbrook (1955), which is cured in a brine with a high-sugar content, followed by draining and smoking. A similar product was produced from dried salted mutton by Zapata et al. (1990). Today jerky is a popular product in the United States with a number of companies specializing in its production. Modern processing in temperature and humidity controlled smokehouses produces jerky in 10-24 hr. During the early phases of drying-smoking, humidity must not be lowered too fast or case hardening will occur. The final a, is between 0.70 and 0.75. 3. Country-Cured Hams
Country-cured hams are made by dry curing hams so as to be stable at ambient temperatures and are produced in a similar form in many countries (Kemp et al., 1983; Pearson and Tauber, 1984; Romans et al., 1985). These products have a high-salt content, which lowers their a, and makes them stable. During storage, the indigenous enzymes, particularly the cathepsins, play a role in flavor development (Toldra and Etherington, 1988; Lopez et al., 1992). Country-cured hams have an a, of about 0.80-0.85 and are in demand in the southern United States. 4.
Lebanon Bologna
Production of Lebanon bologna originated in Lebanon Pennsylvania. Typically, it is made from lean whole carcass cow beef, to which is added
82
S. F. CHANG et at.
2% salt and it then is held for 8-10 days at 2-4°C to permit fermentation by the natural microflora as described by Pearson and Gillett (1995). Today most Lebanon bologna is produced by using starter cultures, which permits production while smoking. The cow meat is ground and mixed with salt, sugar, sodium nitrite seasonings, and the starter culture and smoked until the pH reaches less than 5.0. The finished product has a pH of 4.7 to 5.0, a salt content of 4.5 to 5.0%, a moisture content of 52-56%, and an a , of about 0.85. The final product has a tangy acid and salty taste. It is often called Lebanon style bologna when made in areas other than Lebanon, Pennsylvania. Recently, 275 tons of Lebanon bologna were recalled by a Pennsylvania processor due to possible salmonella contamination (Anonymous, 1995b). This emphasizes the importance of good manufacturing practices and careful control of microbial contamination in processing of IM meat products. 5.
Charqui
Torres etal. (1989) have described the production of the popular Brazilian IM meat product, charqui, which is cured in dry salt, washed, and sun dried. Heating-smoking can be utilized instead of sun drying. Pardi (1961) concluded that charqui is stable for periods exceeding 6 months at ambient Brazilian temperatures. The same principles apply to the production of charqui as to other IM meats that were discussed earlier herein.
Reyes-Can0 et al. (1995) have described the process utilized in production of cecina-an IM meat product that is widely produced and consumed in Mexico. Fresh lean beef from the hindquarter is sliced parallel to the direction of the muscle fibers in long strips at a thickness of about 5 mm and immersed in a 15% salt solution at a ratio of 1:2 meat to solution for 4 hr. The meat then can be dried by oven drying at 50°C for 1 hr or sun dried for several days. Still another procedure adds salt to the thinly sliced meat and folds the meat over the salt until curing is complete. The final product has an a, of about 0.85 and is similar in composition to beef jerky (Reyes-Can0 et al., 1994). The final pH is about 5.4 to 5.5 (Reyes-Cano et al., 1995). Production and processing of cecina is highly variable in different states of Mexico, with some products having added oil or vinegar (Reyes-Cano et al., 1994), whereas other use salt only. The only constants in production are the use of salt and drying of the thin strips of lean beef, which may be
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
83
aided by application of heating in an oven or sun drying. The addition of vinegar or oil also aids in the drying process. 7. Salchichon Salchichon is a popular Spanish style sausage with the Spanish name meaning large sausage. Serrano-Moreno (1979) reviewed the literature on production of salchichon and stated it can be made from lean beef, lean pork, or a combination of the two. Pork backfat is added along with seasoning (1-4% salt, nitrate and nitrite, and white pepper) and sugar (1% of a 1:1 mixture of sucrose and dextrose). It is normally held at a temperature of 25-30°C and a relative humidity of 80-90% for 5 days, after which it is held for an additional 60 days at 30-37°C and 70-80% relative humidity for maturation. This reduces the moisture content of about 50-60 to 2635%. The final a, is 0.80-0.87, and the final pH is 4.6-4.8. Serrano-Moreno (1979) has described the production and physicochemical characteristics of salchichon, which was made from lean ground cow beef (40%), lean pork (30%), and pork fat (30%). It contains about 3% NaC1, sucrose (1%),dextrose (l%), white wine (0.5%), ground pepper (0.25%), whole pepper, potassium nitrate, and sodium nitrite, which were ground and mixed together. Incubated at 20°C and a relative humidity of 80% for 5 days after stuffing in large casings (natural or artificial), salchichon then undergoes maturation at about 12°C at 70% relative humidity for an additional 60 days. Salchichon made in this way has a final chemical composition of about 25% protein, 27% moisture, 43% fat, and 5% ash. It has a final pH 5.90 and an a, of 0.85 (Serrano-Moreno, 1979). Micrococci and lactobacilli are the predominant microorganisms, with the former being the most important by the end of maturation. The salt content rose from 2.69 initially to 4.03% in the final product. Ill. TECHNOLOGY OF PRODUCING IM PET FOODS
Development of IM pet foods has proceeded that of IM foods for humans and many problems can be prevented by applying principles and techniques learned in the commercial manufacture of IM pet foods. Some of the principles and techniques used in their production will be described. A. PRINCIPLES INVOLVED IN PRODUCTION OF IM PET FOODS Development of IM semimoist pet foods was discussed by Karel(1976), since it was considered to be sufficiently important to have application in
84
S. F. CHANG et al.
producing IM foods for humans (Davies et al., 1976). Burrows and Barker (1976) pointed out that two techniques have been utilized in production of IM pet foods, namely, (1) making a slurry of the ingredients, heating them to a pasteurizing temperature and extruding them to produce a semimoist product; and (2) infusing whole chunks of meat or meat analogs with humectants to lower their water activity in order to produce a stable semimoist pet food. Burrows and Barker (1976) stated that raw meat products utilized in producing pet foods have an a, of about 0.97 and a moisture content of about 60-70%. They concluded that the a, must be lowered to about 0.80 by adding of humectants and drying. At this a, the product is palatable and shelf stable, except for mold growth. However, mold growth can be prevented by addition of a mycostat, such as potassium sorbate. Products produced in this way were both palatable and shelf stable. The humectants that have been used to lower the a, of pet foods include sugar, salt, and polyhydric alcohols (Burrows and Barker, 1976). Addition of these humectants results in products containing about 35-45% moisture. The salt content is limited due to its effect upon acceptability so the polyhydric alcohols and sugars are normally utilized at higher levels (Davies et al., 1976) in order to lower the a,. Propylene glycol is the polyhydric alcohol of choice, although others, such as glycerol, have also been utilized (Burrows and Barker, 1976). Table 11, which was taken from Karel (1976), gives the range of the most common ingredients used in pet food formulations. Salt may also be utilized but is limited to about 2-4%. The wide range of other components in the products means that they must also be balanced, i.e., adjusted up or down according to the amount of other constituents in the formulations. TABLE I1 RANGE OF TYPICAL INGREDIENTS UTILIZED IN IM PET FOODSa
%
Meat or meat by-products Sodium caseinate Sugar Propylene glycol Starch Nutritional supplements Flavor and color additives a
Source: Karel (1976).
30-70 7.5-25 15-30 2-10 0.5-10 1-5 As desired
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
85
B. HUMECTANTS USED IN IM PET FOODS
Table I1 shows that sugar and propylene glycol are the main humectants utilized in producing IM pet foods. However, since sodium caseinate and starch also are added in the dry form, they also assist in lowering the water activity of the final products and acts as humectants. Although glycerol also acts as a humectant, propylene glycol is commonly preferred because of a desirable effect upon the flavor (Burroughs and Barker, 1976). The high-sugar content in some formulations may lead to the Maillard reaction and produce melanoidins, thus having a deleterious effect upon both color and flavor (Davies et af., 1976). C. MYCOSTATS Federal meat inspection regulations permit soaking of sausage casings in 2.5% potassium sorbate or 3.5% polyparaben solution to inhibit mold growth (Rust, 1988). The same fungicides also are used in semimoist pet foods to prevent mold growth during storage at ambient conditions. IV. PRESERVATION PRINCIPLES AND THEIR APPLICATION TO IM MEATS
A. STABILIZATION DURING PROCESSING Stabilization of meat and meat products is achieved during processing as well as in storage. The stability achieved is due to the cumulative effects of the processing operations, treatment, and final storage condition for each product (Gould, 1989a,b), although mechanical failure may occur in biological systems (Atkins, 1987). B. COMPARISON O F HEATED AND RAW PRODUCTS
Certain meat products are preserved by heating. Heating is used to destroy some of the pathogenic bacteria and to make the product more palatable (Pearson and Tauber, 1984). However, some raw IM meats are consumed without any prior heat treatment, such as is the case with salami (Romans et af., 1985). For some processed (heated) products, heating is applied during processing, where its major function is to achieve stability rather than the effects on palatability. Most heated meat products still require preconsumption reheating, except for some IM meats (Leistner, 1987). Currently, there is a trend to heat more of the formerly uncooked
86
S . F. CHANG etal.
raw meats to improve their safety and stability (Leistner and Rodel, 1976; Leistner et al., 1981). 1. The Simple Stabilization Operation
High temperature transient heating of meat produces in closed containers makes sterilized canned meat items stable when stored at ambient temperature at a, = 1 (Potter, 1986). The product is heat processed and does not require further heating before consumption. Chilling and subsequent storage at -20°C also renders frozen meat items stable at a, = 1 (Potter, 1986). The high degree of moisture removal and subsequent holding of the products at an a, < 0.6 renders dehydrated meat items stable at ambient temperature (Leistner, 1987). However, the products normally do require rehydration before consumption.
2. Preservation by Complex Stabilization Treatments Many of the same principles apply to all IM foods as explained by Gee et al. (1977). Refrigerated meat items are subjected to mild chilling in combination with other stabilization treatments and then stored at <4"C at a, = 1 (Romans et al., 1985). Shelf stable meat items are subjected to a moderate degree of transient heating (less than canned meat items) to achieve an a, of 0.90-0.97 along with other stabilization treatments (salt and pH) and stored at ambient temperatures (Leistner, 1987). For IM meats, several optional stabilization treatments usually are involved. The only required operation is dehydration, with some IM meats being produced without involving heating (Okonkwo, 1984; Leistner, 1987; Obanu, 1988). However, dehydration is applied only to remove the moisture to a certain level, which is designed to be about in equilibrium with ambient relative humidity at ambient temperature. The products are finally stored at the achieved a,, which is equivalent to the ambient relative humidity (Ledward, 1981,1985; Leistner, 1987; Obanu, 1988; Okonkwo et al., 1992a,b). Indeed, the products are dehydrated in combination with other optional stabilization operations and/or treatments to produce a stable moisture state. Stable storage conditions for each product are defined in relation to their storage surroundings in terms of temperature and relative humidity. The sequence of the applied stabilization operations andlor treatments is variable among available IM meats as explained by Obanu (1988). For example, heating can be applied either in a high-moisture or in a lowmoisture atmosphere. In the latter case, the heating may be a complete process or may not be complete until heating before consumption. The
A REVIEW O F DEHYDRATION CONTROL IN IM MEAT PRODUCTS
87
water activity and storage temperature of IM meats also varies with the prevailing climatic conditions to which each product is exposed (Ledward, 1985; Obanu et al., 1975a,b, 1976). In tropical areas, water activity and storage temperatures of products will be higher than those of the same products in a temperate area (Leistner, 1987). The stability of various meat products depends on the whole series of stabilization factors, which must be achieved in the preservation procedure as shown in Tables I11 and IV. The final water activity, in itself, is simply a physical index of product stability in terms of chemical and microbial deterioration (Troller, 1980; Gould, 1989b). IM meats can be preserved at ambient temperature without any stabilization treatment other than dehydration according to Leistner (1987). The commonly associated soluble solids (sugar and salt) in most IM meats are added to achieve stability during the early stages of dehydration. However, when a meat mixture has an a, between 0.90 and <0.97, it is subjected to a certain degree of in package heating, so that the processed mixture is classified as a shelf stable product (Table IV). When a meat mixture is TABLE I11 EFFECTS OF TREATMENTS ON VARIOUS PARAMETERS INFLUENCING SPOILAGE AND/OR GROWTH OF FOOD POISONING MICROORGANISMS
Treatment Drying Salting Sugar Fermentation Acidification Removal of O2 Smoking Addition of C02 Irradiation Freezing Cooling Preservatives Bacteriocins High pressurization Electrical current Addition of spices
Effects on stability characteristics Lowering a, to 0.90 or less. Prevents bacterial growth and spoilage. Preserves by lowering a, by accelerating drying. Lowers a, and inhibits bacterial growth. Counteracts water loss due to salt. Lowers pH and inhibits growth of spoilage and food poisoning organisms. Assists in removal of water. Lowers pH and aids in removal of unbound water. Aids in drying. Prevents growth of aerobic microorganisms. Adds organic acids and other preservative substances. Inhibits growth of aerobic bacteria. Lowers surface pH. Low level radiation reduces number of viable bacteria. High level inactivates. Stops bacterial growth and delays deterioration of meat. Slows down rate of bacterial growth and spoilage onset. Inhibits growth of microorganisms. Bactericidal and bacteriostatic agents (nisin). Destroys or inhibits microorganisms. 3000-4000 atm. Destroys vegetative cells and inactivates enzymes. Application of high levels of electrical current destroys bacteria without heat. Some spices have bactericidal and bacteriostatic effects.
88
S . F. CHANG et al.
TABLE IV TARGET ORGANISMS CONTROLLED IN VARIOUS PRODUCTS BY DIFFERENT STABILIZATION TREATMENTS
Type of product
Target organisms"
Fermented meats Canned meats (sterilized) Modified atmosphere packaged meats Frozen meats Cured meats (salt and nitrite) Irradiation-pasteurization dosage Irradiation-sterilization dosage Chilled meats (raw) (0 to -4T) Dry and semidry meats ( a , = 0.90 to 0.40) Cooked uncured meats
~~
Pseudomonads S. aureus and C. botulinum, Enterococceae E. coli 0157-H7, Salmonella, Listeria, Pseudomonads C. perfringens C. botulinum
Pseudomonads, Yersinia (pork), Aeromonas, Salmonella, Listeria, E. coli, Campylobactor, Pseudomonads, C. perfringens S. aureus, E coli 0157:H7 Salmonella, Listeria, C. botulinum (Modified atmosphere packaging) C. botulinum, Salmonella, Listeria and S. aureus
Freeze-dried meats ~~
Staphylococcus aureus and Escherichia coli 0157:H7 Clostridium botulinum Pseudomonads, C. botulinum (vacuum packaging)
~~
~
" Microorganisms which are generally targeted
~
~
~
for control of their growth. This list may
not be complete.
subject to a high degree of in package heating at a, = 1, the processed mixture becomes a sterilized product, which in concert with other treatments, such as nitrite can greatly increase the stability of the product as explained by Sebranek (1988). Preservation of IM meat is complex in nature. The identified stabilization operations/treatments in processing, water activity, and storage temperature of IM meat are never greater than those for other preserved meats (Leistner, 1987).The specific preservation procedure or step for each empirical product cannot be taken for granted, but must be determined by experimentation. The possible involved consequence and/or synergistic stabilization factors developed in the preservation procedure should be subjected to more scientific scrutiny. V.
PROBLEMS IN PRODUCTION OF DIFFERENT IM MEATS
Production of IM meats can be divided into different steps. The manufacture of dry and semidry sausages is steeped in art according to Pearson and Tauber (1984) and Romans et al. (1985). However, the art is slowly
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS 89
being replaced by scientific principles. In modern practice, dry and semidry sausages are the most difficult and time-consuming to produce. Ironically, these types of sausages are thought to have been the first produced in early history, dating back to the Babylonian culture around 1500 BC, approximately 3500 years ago. A. PROCESSING EQUIPMENT AND RAW MATERIALS Processing equipment used in traditional meat drying is simplistic and in some cases primitive, making process control difficult, if not impossible as outlined by Leistner (1987) and Obanu (1988). The processor is often at the mercy of the weather for sun-drying or smoke-drying. Product quality is uncertain and variable so stability is achieved only by luck or sheer artistry. Quality of both the raw materials and the finished products are variable and often inferior. Dried meat production in tropical Africa has remained shrouded in secrecy as explained by Obanu (1988). Salmonellae are often troublesome, the recovery of salmonella spp. from biltong has been reported by Bokkenheuser (1963) and Van den Heever (1965). Biltong with such indigenous infection has also caused salmonellosis in humans (Bokkenheuser, 1963; Botes, 1966). The introduction of biltong for general use in developing countries is obviously hampered by health-related requirements. Production of dending in Indonesia generally is limited in scale to that which can be made in the home or in butcher shops, with there being relatively few large scale processors according to Buckle et al. (1988). Production statistics for dending in Indonesia are not available, but substantial quantities are produced. High storage temperatures and humidities and high initial levels of microbial contamination can lead to significant problems by spoilage from chemical, physical, and microbiological changes unless production and storage conditions are controlled. Studies on Malaysian IM meats by Chuah et al. (1988) led them to conclude that the quality of different batches differs widely due to poor quality control and the lack of proper equipment. However, these production problems are slowly being overcome with the introduction of more modern equipment, which not only speeds up the rate of production but also results in more consistent quality products due to more precise control (Buckle et al., 1988). For example, product particle size, temperature, air speed, and other processing parameters can be standardized and improve the end products. B. IMPROVEMENT OF PROCESSES AND PRODUCTS Since production of Zousoon is rather an empirical process and suffers from lack of control of time and temperature, a study was undertaken by
90
S.
F. CHANG
ef
al.
Chang et al. (1991) to investigate the design and operating parameters for better controlled heating-drying equipment in producing this Chinese product. The process is described in greater detail later in the present review. Similarly preliminary results in producing pork slices by Lin (1981), and I. M. Lin et al. (1981), and S. L. Lin et al. (1983a,b) showed that an improved process was superior to the conventional process in respect to labor cost, quality of product, and rate of production. The new process can be summarized as follows: Pork (sliced)-seasoning-molding-freezing-slicingdrying-roasting-final product. The product produced by the conventional process requires a great deal of labor and time, in addition to being unsanitary and depending upon atmospheric air drying, according to Lin et al. (1981). For industrialized countries, traditional IM foods in developing countries are of considerable interest. In cooperation with scientists from developing countries, the principles behind these IM meats have been studied (Chang et al., 1991;Linetal., 1981;Chuahetal., 1988;Buckle etal., 1988).Resultssuggest that processing and shelf life can be improved without impairment of their sensory and nutritive properties. The improved formulas should be made widely available, because they could be of great benefit in many parts of the world, if the products are acceptable to local tastes. In addition to traditional meat-based intermediate-moisture foods, new and promising ideas for product development in industrialized countries could emerge, although these meat items are based upon centuries-old trial-and-error processes. There is a trend toward food preservation methods that simultaneously provide extended shelf life and minimum changes in food quality. Nowhere in the world is this need more urgent than in less developed countries where the lack of refrigeration makes perishable foods unavailable for widespread consumption. With these facts at hand, the remainder of this review will focus upon: (1)The effects of predrying upon muscle structure and the moisture compartment as affected by slaughter and handling operations and by manufacturing, including heating and osmotic treatments; (2) Mechanisms involved in the dehydration process; (3) The influence of storage stability on the quality attributes as affected by dehydration and the associated processes, and finally (4)Process optimization and appropriate technology. VI. EFFECTS OF SLAUGHTERING, HANDLING, CHILLING, FREEZING, STORAGE, AND THAWING ON MUSCLE PROPERTIES
A. CHANGES IN MUSCLE STRUCTURE AND THE MOISTURE COMPONENT
Schmidt et al. (1981) have demonstrated that the protein matrix in muscle has a marked effect upon its functionality and properties. Every process
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
91
involved in the conversion of muscle to meat alters the characteristics of the structural elements (Stanley, 1983). The most obvious physical postmortem change in muscle is the stiffening and the loss of extensibility as a result of rigor mortis as explained by Lawrie (1979). During rigor, multiple attachments are formed between actin and myosin filaments, which lock the interdigitating mechanism into a rigid, inextensible structure (Pearson and Young, 1989). After a period of time muscle begins to undergo a tenderizing process. In terms of the physical properties, a decrease is noted in the modulus of elasticity as rigor mortis becomes complete, but extensibility does not return (Bendall, 1969). B. EFFECTS OF RIGOR MORTIS The tensile and adhesive properties and the structure of selected beef muscle strips undergoing rigor mortis have been followed at various times postmortem by Currie and Wolfe (1980). They found that changes in the mechanical properties of the muscles correlate well with final pH and the rate of pH fall. Additionally, and perhaps most importantly, the shapes of the curves generated over the postmortem aging times were correlated with the changes in the extracellular spaces. Thus, the authors concluded that intrafiber water must be considered as an important factor in meat tenderness, in addition to the effects of myofibrillar contraction and connective tissue orientation. One factor that may influence the extracellular spaces may be cooking which was studied by Locker and Carse (1976). Another factor that also may alter extracellular space is cold shortening, which has been described by Voyle (1969) and Marsh et al. (1974). 1. Structural Alteration Greaser (1986) has discussed the physical and biochemical changes that take place during the conversion of muscle to meat. According to Offer and Knight (1988a) the fibers in beef sternomandibularis muscle almost touch and the fiber bundles fill the perimysial network after 2 hr postmortem at 10°C. Between about 4 and 6 hr postmortem, however, a significant change can be seen. The fiber bundles shrink away from each other and gaps of about 5-50 pm develop between the fiber bundles and the perimysial network. Nevertheless, at this stage the fibers still seem to fill the endomysial network. The results imply that up to this stage both the cell membrane and the endomysium shrink with the fibers. At about 24 hr postmortem, roughly at the time the fibers enter rigor, another structural change is apparent. In addition to the gaps between fiber bundles, gaps develop between the fibers and the endomysial network. It, thus, appears that the
92
S . F. CHANG et al.
gaps at the endomysial-perimysial junction do not act as a sink for all the fluid expelled from the myofibrils as explained by Offer and Knight (1988b), which is responsible for drip losses. A possible reason for this is that the endomysial network may resist shrinkage beyond a certain point. Alternatively, at the point of rigor onset in a fiber, the shrinking force may rise so fast that ruptures occur before much of the fluid can flow through the endomysial network. Thus, the histological results demonstrate that in muscle at rigor there are two extracellular spaces: first, the gaps between fibers and the enclosing endomysial sheaths, and second the gaps between the fiber bundles and the perimysial network. Stanley (1983) concluded that postmortem events influence the physical properties of meat, not only through rigor mortis, but also as a result of the action of numerous endogenous enzymes on myofibrillar structure, and perhaps, connective tissue as well. A major structural alteration that has been observed in postmortem muscle is Z-disc degradation. The unreactive chemical nature of collagen may preclude any major attack by endogenous muscle enzymes on this fibrous protein (Offer et al., 1988). 2. Fragmentation of Muscle Fibers According to Cia and Marsh (1976) electrical stimulation results in contracture bands that cause tearing and fragmentation of myofibers/myofibrils. McLoughlin (1971) and Van der Wal (1971) suggested that death caused by electrical stunning and C 0 2 gas brings about marked changes in the turnover rate of ATP in the muscle and is responsible for fragmentation of the myofibers. Brief electrical stimulation directly or via the motor nerve depletes creatine phosphate, accelerates the breakdown of ATP, and increases the formation of lactate. McLoughlin (1971) suggested that it is necessary to maintain the intracellular environment as close as possible to the in vivo state after death in order to reduce postmortem glycolysis and concluded that commercial methods of stunning and slaughter presently used do not meet this requirement.
3. Other Structural Alterations Other structural alterations observed during electrical stimulation to the contractive bands include intracellular edema, swollen membranous organelles, myofibrillar fragmentation, and other changes consistent with accelerated autolysis of muscle resulting from tissue disruption (McLoughlin, 1971). Stanley (1983) has shown that several other abnormalities in texture (structure) also may result as a consequence of failure to properly cool
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
93
prerigor meat immediately following slaughter. Jones (1977) has followed the ultrastructural changes occurring in meat postslaughter. 4. Densely Stained Cross-Bands
Bendall and Wismer-Pedersen (1962) have observed irregular wavy, densely staining cross-bands in the muscle fibers of pale, soft, and exudative (PSE) pork. They observed that muscle filaments ran through the dense bands and concluded that the change was not of myofibrillar origin. They theorized that myofibrillar material from a number of sarcomeres accumulated to form the dense bands. On either side of the dense bands, they observed recognizable but short sarcomeres. Bendall and Wismer-Pedersen (1962) further pointed out that the dense bands were similar to the contracture bands or zone of supercontraction that can be induced in muscle by other treatments, such as electrical stimulation. Zones of supercontraction alternating with zones of stretched sarcomeres also have been observed in muscle exposed to an elevated temperature prior to the onset of rigor, after thaw rigor, after cold-shortening, or in electrically stimulated muscle according to Bendall nd Wismer-Pedersen (1962). All this suggests that some of the muscle fibers undergo excessive shortening prior to rigor, leading to a heterogeneity of sarcomere length and irregular zones of supercontraction. However, supercontracted fibers are not always present at rigor, and it is possible that zones of supercontraction persist only in the more extreme forms of PSE muscle (Bendall et al., 1963). The reason, although not clear, is most probably a result of contraction still being in a reversible state. C. FREEZING AND STORAGE Freezing of meat is a common preservation technique which has been discussed by Stanley (1983). It has been demonstrated that when whole muscle tissue is frozen and stored under accepted processing and packaging conditions, little if any alteration in structure or physical properties is produced during a normal storage period (Pearson and Miller, 1950). D. RAW PRERIGOR MUSCLE Raw muscle in a specific postmortem physiological stage and/or stored under proper temperature conditions has been strictly selected and used as the raw material for some specific IM meat products due to the eventual physical properties of the finished items (Chang et al., 1991). Examples include Zousoon and cooked-dried pork pieces or cubes (Leistner, 1987;
94
S. F. CHANG ef al.
Chang et al., 1991). The idea of boning the carcass prior to chilling has been examined because the procedure has many economic advantages, not the least of which is a large energy savings in terms of refrigeration (Henrickson and Asghar, 1985). Some IM meat products require hot-boned prerigor or warm meat as a source of raw material, not because of the concern for energy savings, but rather because of their specific structural and/or physical properties (Chang et af.,1991;Chang and Pearson, 1992). For developing countries, hot boning is a general practice for postmortem handling of meat (Obanu, 1988). For the domestic pork market in Taiwan, people still prefer pork from hotboned meat (S. F. Chang, unpublished observations). VII.
EFFECTS OF PREDRYING TREATMENT AND HANDLING OF MUSCLE
A. CHOPPING OR GRINDING Romans et al. (1985) stated that to achieve the best particle definition, the raw meat should be ground or chopped at very low temperatures (-4 to -6°C). Mixing and unnecessary handling of the meat should be avoided as much as possible. The cold meat temperature reduces protein extraction and helps to eliminate unnecessary particle deformation. Mixing should be only enough to allow uniform distribution of the curing ingredients andl or other additives (Pearson and Tauber, 1984). B. ROLE OF MYOFIBRILLAR PROTEINS IN WATER BINDING The myofibrillar proteins, especially the myosin molecule, are extracted at low-ionic strength (0.1-0.3 M KCl at pH 6.0-8.0) and spontaneously form aggregates as explained by Katsaras and Budras (1992). Some of the aggregates are long and spindle-shaped and are called synthetic filaments. The synthetic filaments thus become building components in a threedimensional network. Some of the reactive aggregates are bound to each other by hydrogen bonds and electrostatic charges to form a continuous but reversible coagulated structure (Offer and Trinick, 1983). Within the three-dimensional protein network a large amount of free water is absorbed, as described by Wismer-Pedersen (1971) and by Offer et al. (1988). The water has two functions: (1) as a hydration layer it separates the protein aggregates; and (2) via the hydrogen bonds it constitutes a link between the protein network. If there is a sufficient amount of water available the protein remains in a colloidal “sol-state”. The coagulation bonds are still
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
95
rather unstable and, in addition, are weakened by the intermediate layer of water molecules. C.
FERMENTATION
All phases of meat fermentation were reviewed in a recent book edited by Campbell-Platt and Cook (1995). The discussion covers the safety aspects of fermentation in meat products as well as its use world wide and the techniques that are utilized. During the process of sausage fermentation and after a certain period of adaptation, bacterial populations increase in number as outlined by Katsaras and Budras (1992). The growth of lactic acid bacteria is dominant because these bacteria enjoy the advantage of selection in the milieu of fermented sausages. The lactic acid bacteria break down sugar and produce lactic acid. As soon as the sausage has reached the isoelectric point of about pH 5.3, the negative and positive charges compensate for each other and the bound water is released, i.e., the water binding capacity in the fermented sausage has reached its minimum. Consequently, the amount of immobilized water between the protein threads is reduced and its function as a “spacer” becomes increasinglyineffective so that the protein aggregates and the meat particles gradually approach each other. The spindle-shaped protein aggregates are subjected to a series of changes, and the intermolecular interactions successively create new bonds, until an extended, dense, three-dimensional network of protein threads are built as explained by Katsaras and Budras (1992). The process, which is accompanied by water loss and shrinkage of the sausage in a drying atmosphere, is called syneresis. Fermentation results in proteolysis and an increase in the amount of insoluble protein as reported by Astiasaran et al. (1990), which may assist in dehydration. Ibanez et al. (1995) compared dry-fermented sausages formulated so that part of the NaCl was replaced with KCI from the standpoint of their effects on carbohydrate fermentation and nitrosation. Both sausages attained similar final pH and a, values, at pH 5.06 and 4.99 and a, 0.85 and 0.88 for the control NaCl and the KC1-containing sausage. Final a, values were achieved in less than 20 days and were well below the a, value of <0.90, which is required for stability of IM meats. D. EFFECTS OF HEATING
According to Sebranek (1988) heating is believed to cause the denaturation of the muscle proteins even below 60°C, but not enough to greatly affect shear resistance. The decrease in shear observed at 60°C was attrib-
96
S. F. CHANG et af.
*
uted to collagen shrinkage. Hardening at 70-75°C was believed to be due to increased crosslinking and water loss by the myofibrillar proteins, while decreasing shear at higher temperatures may indicate solubilization of collagen. Heating produces major changes in muscle structure. Voyle (1981) has reviewed modifications in cooked tissue observable with the scanning electron microscope. Several authors (Hsieh et al., 1980b; Voyle, 1981) have reported alteration in muscle structure due to heating, which include coagulation of the perimysial and endomysial connective tissue, sarcomere shortening, myofibrillar fragmentation and coagulation of the sarcoplasmic proteins. Heating and/or drying intensifies the detachment of the myofibrils from the muscle fiber bundles, which is caused mainly by electrical stunning or stimulation and improper conditioning following slaughter (Chang and Pearson, 1992). The bovine sarcolemma with its associated endomysium has been shown to maintain its postrigor integrity except for the occurrence of some small perforations in the sarcolemma (Rowe, 1989a,b).The close structural association of the three component parts, i.e., the plasmalemma, basement membrane, and endomysium, persists even during physical disruption of the muscle, such that when a tear occurs all three parts are involved according to Rowe (1989a,b). After 1hr at W C , the collagen fibrils of the endomysium appear beaded, which is brought about by their close association with the heat-denatured noncollagenous proteins in the extracellular spaces. Heat denaturation of the lipoprotein plasmalemma results at a temperature of 60°C for 1 hr. The breakdown products of the plasmalemma are large granules and are often associated with the basement lamina, which appears to survive intact even after heating at 100°C for 1 hr (Rowe, 1989a,b). When an animal or plant is killed, its cells become more permeable to moisture as pointed out by Potter (1986). When the tissue is blanched or cooked, the cells may become still more permeable to moisture. Generally, cooked vegetable, meat, or fish will dry more easily than their fresh counterparts, provided cooking does not cause excessive shrinkage or toughening (Potter, 1986). Cooking also results in a decrease in WHC (Stanley, 1983). Liquid components, such as water and lipids, are removed. Shrinkage in fiber diameter and length occurs and the apparent density increases (Stanley, 1983).Other changes, such as denaturation of the sarcoplasmic proteins, are also observed (Bendall and Wismer-Pedersen, 1962).
E. INFLUENCE ON OSMOTIC TREATMENT Osmotic concentration can be considered a simultaneous water and solute diffusion process. As explained by Lerici et al. (1988) the membrane is only
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
97
partially selective, thus, there is always some leakage of solute from the solution into the food and from the food into the solution. By means of a suitable choice of osmotic solutions and of the processing conditions, the solute diffusion can be minimized or enhanced, depending on the desired characteristics of the food. It is this duplicity of action in respect to the food, (partial dehydration and controlled chemical composition) which makes direct osmosis so interesting in food processing (Ponting et al. 1966). Guilbert (1988) demonstrated that an edible layer of material could be utilized to protect from moisture during storage of tropical fruits dried by osmosis. Muguruma et al. (1987) demonstrated that low-temperature osmotic dehydration improved the quality of IM meat products. They utilized a socalled “dehydrating sheet” to develop osmotic pressure and improve the rate of water removal at relatively low temperatures, which resulted in better quality IM meat products. On immersing either a whole muscle or single isolated fibers in a hypertonic medium, the single fibers behave as osmometers and shrink substantially as explained by Offer and Trinick (1983). In contrast, the whole muscle shrinks only slightly, because the fibers within it shrink substantially, which is almost matched by an increase in the extracellular space. This would be expected, since connective tissue sheaths are presumably not permselective (selective semipermeable membranes), and therefore, neither the perimysial nor the endomysial networks should alter in volume in the presence of an osmotic agent. Shrinkage of myofibrils (or fibers) leads to a greater proportion of loosely held water, which has the potential of being lost from the meat. The greater the amount of shrinkage, the greater is the potential for water loss (Offer and Knight, 1988a).The loss, however, will only occur under appropriate conditions, such as pressure. VIII. FACTORS INFLUENCING ABSORPTION PHENOMENA IN MEATS/MEAT MIXTURES
A. BOUND AND UNBOUND WATER IN MEAT PRODUCTS Moisture may be bound to the solids so that full vapor pressure is no longer exerted according to Watson and Harper (1987). They pointed out that the terms unbound (free) and bound are commonly used to distinguish moisture in the relatively large spaces from that held more tightly by other forces. They further stated that there is a continuous transition from unbound to bound moisture, and it is not possible to make a precise dividing line. Unbound moisture is frequently defined as that which exerts the
98
S. F. CHANG ef al.
normal vapor pressure, while bound moisture has a lower vapor pressure (Watson and Harper, 1987). Offer and Knight (1988b) concluded that muscle protein molecules in an aqueous solution interact with water, and when it moves through the solvent it carries some water with it. Part of this bound water is believed to be hydrogen-bonded to the surface of the protein molecule, while part may be present in clefts or pockets. Both are in dynamic exchange with “free” water. These authors stated that the amount of water associated with proteins in this way can be measured by a variety of techniques. However, it amounts to only about 0.5 g of water per gram protein. The total concentration of protein in muscle is about 200 mg/ml, so that as emphasized by Hamm (1960,1986), only about a 10th of the water in muscle can be considered to be closely bound with the proteins. Measurement of water vapor pressure of a food as a function of moisture content provides information on the vaporization potential at a given moisture content with respect to the specific solids in a food (Watson and Harper, 1987). Thus, water vapor pressure is a useful measure of moisture content in food products. B. RELATIONSHIP OF WATER VAPOR PRESSURE TO WATER ACTIVITY Water activity was defined by Scott (1957) using the equation a, = P/po, where P is the water vapor pressure of food and PO is the water vapor pressure of pure water at the same temperature. Water activity is, thus, an index of degree of freeness (the potential to be vaporized) of the solid bound moisture relative to pure water. Relative humidity has been defined by Aguilera and Stanley (1990) using the equation RH (%) = P/po X loo%, where P is the partial water vapor pressure in an atmosphere and PO equals the water vapor pressure of pure water at the same temperature. Thus, the water activity of a food can be determined by the prevalent relative humidity in the atmosphere surrounding the food when in an equilibrium state. It can be expressed by the equation a, = P/po = RH% X 1/100 (in equilibrium state) = ERH% X 1/100, where ERH is an expression of equilibrium relative humidity (Watson and Harper, 1987). Water activity is, thus, an index of water vapor pressure of food or its moisture state (in our case meat) and is a function of: (1) the moisture contentholid content of a meat or meat mixture, (2) the components in and composition of a meat or meat mixture, (3) the microstructure of a meat or meat mixture, (4) the temperature, and ( 5 ) the state of some component solids (i.e., sugar).
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
99
C. SORPTION PHENOMENA IN MULTICOMPONENT FOOD SYSTEMS When food components differing in a, are put into the same system, the components of higher a, give up moisture to those with a lower a, until the mixture reaches equilibrium as described by Potter (1986). A practical consequence of this is that each component of a mixture can be prepared separately under specific conditions of formulation and/or infusion. When these components are subsequently blended and reach the equilibrium a, of the mixture, they will retain different amounts of water in keeping with their individual water sorption isotherms and texture. This principal is employed in producing complex mixtures. Rockland and Nishi (1980) stated that isotherms appear to be related to different modes of water binding. Their statistical study suggests that local isotherms do not always give a precise and unequivocal definition of the state of water in heterogeneous mixed-component systems. Linko et al. (1981) have discussed the additive model of water absorption by calculating the adsorption isotherm of skim milk powder and its components. They obtained good agreement between adsorption isotherms of skim milk powder and those of various binary and ternary mixtures. Casein appeared to be the main adsorber of water at a, < 0.2. In the range of 0.2 to 0.6, the sorption behavior of dried milk products was dominated by the physical state of lactose. At a, > 0.6, the salts present in the milk powder had the major influence on water adsorption. Van der Riet (1976b) utilized sorption isotherms to predict the critical moisture content for storage of biltong. Aguilera and Stanley (1990) concluded that at high-moisture contents an aqueous environment prevails and many components, in particular salts and low molecular weight sugars, exist in solution. The undissolved components are high molecular weight materials, such as proteins, and form the structural matrix of the food, which contains the aqueous phase. Liquid water itself has a stable structure due to hydrogen bonding that is perturbed every time a solute is introduced. The same authors stated that water activity for actual solutes in a high-moisture range can be estimated by Raoult’s law. They further concluded that high molecular weight polymers, such as proteins, may immobilize large quantities of water in threedimensional structures called gels. Sorption isotherms for gels extend over the whole range of a d . The contribution of gels to food microstructure and their depression of water activity deserve further research aimed at fabricating shelf stable products. Finally it was stated by these researchers that a, depression by capillary effects in a food is complicated by difficulties in determining pore size, swelling of the matrix during sorption, and lack of information about the actual radius of curvature of the meniscus. Additional
100
S. F. CHANG et al.
work is needed to demonstrate and quantify the capillarity effects of a, in foods. D. SORPTION PHENOMENA IN MEAT SYSTEMS IM meats are multicomponent systems as explained earlier by Chang et al. (1991). The isotherms of the meat or meat mixture are a relationship of averaged moisture content and the common water activity. In such systems, each component has a common water activity. However, each component may have a different moisture content. The added soluble solids will increase the average moisture content of the IM meat mixture within the specific water activity range of 0.60-0.90 as shown by Chang et al. (1991). Indeed, the soluble solids in the amorphous state adsorb more water than that of the muscle solids within a water activity range of IM meats. In production of IM meats, in addition to selecting raw muscle with good sorption properties, one should try to modify the shape of the isotherm in the meat by formulation'(adding sugar andlor salt). Added sugar and salt change the sorption phenomena of the meat solids by addition of these nonmeat ingredients. Thus, in the range of a, 0.60-0.90, which is equivalent to an open storage relative humidity of 60-90%, the meat mixture will have a higher moisture content than that of the muscle alone. The purpose can be achieved by manipulation of the involved solids through selection of the proper components and composition and by altering the structure of the solid matrix, such as developing a gel or capillary structure. E. INFLUENCE OF MUSCLE PROTEIN DENATURATION ON SORPTION ISOTHERMS Water holding capacity refers to the potential of the meat solids to bind the water in a high-moisture range as explained by Hamm (1960). Muscle with a high water holding capacity will have good water adsorption properties at a moderate to low-moisture range. Consequently, raw muscle with less protein denaturation is preferred for IM meat formulations. Raw muscle with good water adsorption properties will improve the textural properties of IM meats within the specific a, range (Chang et al., 1991). Water sorption in relation to texture will be discussed further later herein. IX. MECHANISMS INVOLVED IN MEAT DEHYDRATION SYSTEMS
A. MOISTURE REMOVAL DURING IM MEAT PROCESSING
Van Arsdel (1963) and Keey (1972) have explained that dehydration is a process of moisture removal from a solid by thermal means. This definition
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
101
distinguishes drying from mechanical methods of moisture removal from solids, but does not fully differentiate between drying and evaporation in which heat is used to evaporate large amounts of water from solutions or slurries. Predrying osmotic treatment of meat causes shrinkage of muscle fibers as outlined by Offer and Trinick (1983) and Offer and Knight (1988a,b). They stated that shrinkage of the fibers leads to a greater proportion of loosely held water in the extracellular space. Mechanical pressing of the muscle separates the loosely held liquid from the solids. The pressing operation removes most of the moisture from the muscle. However, some solids are lost in the expressed liquid. The predrying heating of muscle decreases its water holding capacity. The cooked muscle exudate in combination with the muscle per se can further be heated and boiled to “dryness,” with the leached solids being reabsorbed. The “cooked-dry’’ boiling process also removes most of the moisture from the meat but without solid losses. The predried meats or meat mixtures are generally further dried by thermal dehydration (Chang et af., 1991). Weight losses during the heating process are caused through exudative and/or evaporative moisture losses. Prevention of the cooking loss is a major concern in heat processing of meat. In meat dehydration, however, the weight lost by evaporation must be effectively achieved, but solid losses should be avoided (Sebranek, 1988). Chang et al. (1991) have studied the dehydration process in relation to IM meats. They stated that if the unbound moisture in meats is defined as that which exerts water vapor pressure like that of pure water, all unbound moisture must evaporate before equilibrium can be achieved with air that is less than saturated. In other words, the water will evaporate until the water vapor pressure of the meat is equal to the partial water vapor pressure in the air. Data on the equilibrium moisture content-relative humidity (i.e., isotherms) of meat or meat mixtures are needed over a wide range of temperatures for dehydration applications. For example, isotherms for meat/meat mixtures below ambient temperature are needed for salami and raw ham, and isotherms above steam temperatures are needed for hightemperature finished dried IM meats (Chang et af., 1991). Dehydration is basically a simultaneous heat and mass transfer operation as explained by Van Arsdel (1963). In applications where rates of drying are low, such as in IM meats, consideration of heat transfer alone is a quite satisfactory approach, and often a preferred one. However, in order to obtain a true understanding of drying and to develop a sound fundamental theory, one must better understand the mass transfer process, both internally and externally (Lubuza, 1976). In addition, one must develop a better
102
S. F. CHANG et al.
understanding of the forces that bind liquids to solids and control their movement (Aguilera and Stanley, 1990). Most thermal dryers embody convective heating as the drying condition and can be readily controlled by the temperature and humidity of the (warm) air that evaporates and removes the moisture (Keey, 1972; Chang et al., 1991). Air dryers for handling solid materials are by far the most common type of equipment. Thus, the discussion that follows is devoted to air drying. Potter (1986) stated that in a strict sense dehydration refers to the nearly complete removal of water from foods under controlled conditions that causes minimum or ideally no other changes in their properties. In the dehydration process, the technological challenge is especially difficult, since very low moisture levels are required to obtain maximum product quality, but are not easily obtained with a minimum change in the foods per se. Further, optimization of quality frequently can be attained only at the expense of increased drying costs as stated by Potter (1986). Since IM meats are produced by removing only part of the water, the increased costs of drying may not apply to these products (Chang et al., 1991). Controlled studies on dehydration by Watson and Harper (1987) have demonstrated that both the temperature and humidity change as the air moves past the moist material. Thus, the moisture content of the solids varies from one point to another in the dryer. This means that the operation cannot be analyzed easily to provide the drying characteristics of the material itself. Constant control in batch drying experiments is better suited for this purpose and more generally provides a better understanding of the mechanisms involved in the drying process (Chang et al., 1991). The concept of a constant rate of drying period followed by one or more falling rate periods, with sharp discontinuities in a rate at critical moisture contents, is firmly entrenched in the literature and may be regarded as constituting the classical theory of drying (Keey, 1972; Watson and Harper, 1987). B. DRYING RATE CURVES FOR HEATED MUSCLE BUNDLES Chang et al. (1991) designed a series of experiments to investigate the structural modifications in cooked muscle bundles during predrying tumbling with regard to possible shifts in moisture removal mechanisms. The drying rate curves in the study were shown to vary in a stepwise fashion, since the muscle bundles had more than one constant rate period (Fig. 1). Muscle bundles tumbled for longer times developed more constant rate periods. This was true even though the greater amount of disintegration could not be seen without magnification of the sample.
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
0.5
1.o
103
1
Moisture Content (g H20/g solid) FIG. 1 . Drying rate curve for heated pork muscle bundle. Drying rate curves were determined at a dry bulb temperature ( T d b ) of 40"C, a wet bulb temperature (Twb)of 35°C. and an air velocity ( V ) of 1.0 d s e c . The specified time in minutes for each curve (0, 10,20, and 30 min) represents the tumbling time for the heated muscle bundles before the determination. Source: Chang et al. (1991). Reprinted with kind permission from Elsevier Science Ltd.
The reappearance of each constant rate period followed each transient falling rate period. Among the treatments, the drying rate varied with length of tumbling, with those bundles tumbled longer having a faster drying rate than those tumbled for a shorter period of time at the same moisture content (Fig. 1). The samples dried without tumbling (0 min) became hard and dry, whereas samples tumbled for 10-30 min were soft and porous in texture. Predrying tumbling was shown by Chang et al. (1991) to play an important role by modifying the structure of the muscle bundles. The tumbling-created shear force overcomes the adhesion of the heated connective tissue. The loosened muscle structure modified by the previous heating and tumbling is intensified by the stress caused by connective tissue shrinkage, which separate it from the elements of the bundle and further aids in fiber dehydration. During development of separation, there is an instant increase in the internal free surface vaporization, which is considered to contribute to the reappearance of the constant drying rate period that follows each transient falling rate period as shown in Fig. 1. This contributes to an increase in
104
S. F. CHANG et al.
the drying rate of the tumbled muscle bundles. Otherwise, the drying rate will fall continuously, giving the bundles a hard and compact texture similar to the heated untumbled muscle bundles. Norback (1980) has summarized some techniques that may be utilized in optimizing the drying process. The stepwise drying rate curves obtained by Chang et al. (1991) are quite different from those for so-called “structureless” material. Conventional drying curves generally consist of one constant rate period with one or two consecutive falling rate periods. The stepwise curves obtained by Chang et ai. (1991) can be explained in term of specific changes due to modification of muscle structure that occur during predrying heating and tumbling and in the subsequent effects of those developed during the drying process per se. The key explanation is that heating and tumbling intensifies formation of a “capillary-like” structures in the supporting connective tissues, which are partially solubilized during heating. Some of the capillary-like structures formed in connective tissues may then gradually contribute to a more open structure as the drying process progresses. The changes during the constant rate of drying period mean that more vaporization of unbound moisture is occurring at the free and open internal surfaces, which correspond to openings in the bundledfibers surface during the complex effect of heating, tumbling, and drying. Muscle bundles tumbled for longer times developed more constant rate periods because their internal structure disintegrated more than those subjected to shorter tumbling times (Chang el ai., 1991). Drying rate curves for cooked muscle bundles implies that the classical fundamental studies on drying properties and mechanisms under constant external drying conditions are empirical in nature. This is shown in the specific study of controlled muscle bundle dehydration, which indicated that for intact tissue the embedded capillaries may become more open and thus create internal free vaporization surfaces as demonstrated in studies by Chang et at. (1991). Internal free surface vaporization operating during unbound moisture removal resulted in identification of the mechanism and is reflected in the reappearance of the constant drying rate periods (Fig. 1) and the multiple critical points and by the porous texture of the product. These studies by Chang et al. (1991) showed that during muscle dehydration, the shifting of the drying mechanisms and drying properties are always possible if the pretreatment and/or drying conditions are varied, which can cause structural modifications in the muscle. If the embedded connective tissue network is intact, capillary movement and external surface vaporization become the major mechanisms for moisture removal. On the other hand, if the connective tissue network is open, below surface and free vaporization become the major mechanisms involved. With meat mixtures of a complex structure, the relative importance of each mechanism depends on the specific structure of the meat and the manipulative changes that
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
105
occur in the process. If below surface free vaporization of unbound water is involved, the apparent moisture diffusion path decreases since the drying rate increases sharply (Chang et al., 1991). C. PHYSICAL AND STRUCTURAL CHANGES IN IM MEATS Loss of water from both raw and precooked meat is accompanied by a decrease in the space between groups of muscle fibers and between the individual fibers and by a progressive reduction in muscle fiber diameter as explained by Wang et al. (1953). The rate of moisture removal and of muscle fiber shrinkage is more rapid with heated than with raw meat and proceeds further. Jason (1958) studied the falling-rate period during drying of cod muscle. Although he did not observe the capillary or porous structure, a “continuous gel structure” was seen, which behaved almost as an isotropic medium. He concluded that the main mechanism of water removal was molecular diffusion in a solid medium. The capillary channels were closed after the capillary movement of the moisture was complete. Whether other mechanisms functioned during the altered stages of drying depended on the state of the modified connective tissue, i.e., if it had been torn or solubilized during rigor development or precooking, respectively. In summary, drying rate data do not ordinarily provide sufficient information to distinguish the mechanisms involved in moisture removal. Further, the mechanisms will usually change as drying proceeds. Drying rate data have been analyzed mathematically on the basis of particular mechanisms for moisture movement through the solid. The complex nature of the moisture movement process, however, makes it impossible to attach any fundamental significance to rate constants obtained from such analysis. The fact that experimental data do not conflict with a theory does not prove its validity. Essentially all moisture movement mechanisms may lead to the same general form of drying rate curves. Drying rate constants can be looked upon only as empirical expressions of the rate of moisture loss during dehydration. Because of the empirical nature of the constants, data must be obtained over a much wider range of variables than would be necessary if there were a dependable theory. The moisture content at the dividing point between the constant and falling rate periods on the drying rate curve is commonly called the critical moisture content. However, critical moisture content is not entirely a property of the material, but depends on the manner of loading on the bed and on drying conditions (Chang et al., 1991). The stepwise drying rate curve of the tumbled cooked muscle bundles as discussed by Chang et al. (1991) further proves the above statement (Fig. 1). Indeed, reappearance of the constant rate period makes it impossible to approximate the critical moisture content. Limited knowledge on the
106
S. F. CHANG et al.
critical moisture content restricts its application in prediction of drying rates during the falling rate period. However, the real value of the more fundamental approach is that once it is related to practice through actual large scale experiments, it provides a basis for determining the effects of variation in operating conditions and to establish optimum drying schedules without the necessity of extended experimental programs for each individual situation.
D. DEHYDRATION PARAMETERS AND PROFILES IN RELATION TO PHYSICAL CHANGES DURING DEHYDRATION OF IM MEATS One main reason for the difficulty in developing a broad fundamental theory for drying is due to the fact that many foods and other solid materials are dried. Differences in the structure and the chemical composition of various foods and other solid materials make it difficult to develop a sound theory of drying based on internal moisture flow that would be applicable to all materials. The drying theories that have been developed over the years have originated from two points of view as explained by Keey (1972): (1) the effects of the internal mechanisms of liquid flow within a product include (a) air temperature, (b) humidity, (c) air velocity, (d) agitation of the solid material, (e) method of supporting the solid material, and (f) contact between the hot surface and wet solid. (2) The external theory of drying considers the effects of external variables to be more useful in correlation of operational data and equipment design than the internal theory. Internal mechanisms of liquid flow are a function of its solid structure. The internal mechanism of drying may be more fundamental and provides a better insight on how a solid dries than the external variables, which may be more useful for design purposes (Keey, 1972). Factors affecting heat and mass transfer, such as temperature, humidity, and air velocity, are relatively easy to control and largely determine dryer design (Keey, 1972). Far more subtle are the properties of food materials that affect heat and mass transfer and that may change during dehydration. Dehydration profile (time-product temperature-moisture contentla,) of a material can be discussed mainly in terms of air conditions (dry bulb temperature, air velocity, and wet bulb temperature/relative humidity) and relative efficiency of heat transfer vs mass transfer of the material to the surrounding air as explained by Sebranek (1988) and Chang et al. (1991). The difference between dry bulb and wet bulb temperatures or product temperature is an index for the driving force of heat on the material and the potential for dehumidification or vaporization of water from the raw material. If heat transfer efficiency is greater than mass transfer efficiency,
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
107
less moisture is evaporated. On the contrary, if mass transfer efficiency is greater than heat transfer efficiency, the product cools with greater evaporation of moisture. If heat and mass transfer efficiency are in balance, the product is at the medium temperature and suffers a moderate moisture loss. Indeed, product temperature, which fluctuates between the dry and wet bulb temperature of the air, depends on the instantaneous relative efficiency of heat vs mass transfer in the dehydration system (Sebranek, 1988; Chang: er aL, 1991).
E. DEHYDRATION IN PRODUCTION O F IM MEATS Discussion will focus on meat dehydration based on information that is derived from heat processing of meat in which dehydration is not the major concern, but is a necessary step in its preparation, i.e., cooking of raw meat. However, the partial surface drying that occurs in cooking of meat is similar to that involved in dehydration. Sebranek (1988) has discussed heat processing of meat as it applies to both fresh and cured pork. In addition to heat transfer and diffusion considerations in processing of meat cuts, heating rates are influenced by mass diffusion and mass transfer characteristics. These are a result of water release by meat proteins, diffusion of the water through the product and evaporation (or drip) loss from the surface. The primary effect of mass transfer on the heat process is the evaporative Cooling that occurs at the product surface during the phase change of liquid to water vapor. Since the phase change alone absorbs 1000 BTU per pound of water evaporated, when surface evaporation occurs considerable energy is used without any increase in product temperature (Sebranek, 1988). Consequently, methods of heating that minimize evaporative cooling result in more efficient transfer of energy to the product. One implication of this is that the humidity of the air during heat processing (or in dehydration) is an important factor that affects heating (or drying) rate efficiency to a greater extent than air temperature. However, relative comparisons between heating methods in terms of permitting or limiting evaporative cooling are dramatic. For products heated in various situations involving air, relative humidity becomes a major consideration since evaporation of water from the product will continue during heating until the partial water vapor pressure in the air equals that of the product. Thus, heat transferred, in the case of air heating, is dependent not only on temperature differential (air velocity and product thermal conductivity) but also upon the moisture content of the air (Sebranek, 1988). Complete convective heating without moisture loss by evaporation is rarely encountered in processed meat products. Accordingly, heat processing of meat in a convective air stream can be considered as a specific case
108
S . F. CHANG et al.
of dehydration. For example, the drying phenomena in comminuted meat products heated in air and the concomitant heating and drying effects best can be observed in meat gels. Mittal and Blaisdell (1983) evaluated the moisture mobility in frankfurter emulsions during cooking. Since they did not observe any water removal rates, they concluded that internal moisture movement was the controlling factor in moisture losses. They suggested that the bulk of unbound moisture is fixed in the gel matrix, thus, bulk flow of moisture could not contribute to a constant rate period. Evaluation of the effects of relative humidity at 69°C by Mittal and Blaisdell(l983) showed variable results. As relative humidity was increased from 41 to 60%, moisture losses increased. As relative humidity increased further to 87%, the moisture loss was reduced. The low relative humidity probably permitted excessive surface drying and skin formation or case hardening in the frankfurters, thus limiting further removal of moisture. Higher relative humidities permitted a balance between heat transfer and evaporative cooling, thus reducing moisture losses. Steep moisture gradients were present within the frankfurters during heating in the air, with virtually no moisture removal from the center. Thus, the authors concluded that even though muscle proteins may release water at around 50"C,low moisture diffusivity in the frankfurters will not allow the water to migrate to the product surface. Among physical stabilization factors in meat mixtures, the heat-set protein gel matrix is generally recognized as the predominant factor controlling water retention. Gelation of meat proteins during heating takes place to some degree in all products. It first involves unfolding, then the interlinking of muscle proteins (or even possibly nonmeat proteins if included) to form a three-dimensional continuous network (Offer and Knight, 1988a). This network is effective in trapping and stabilizing water in the mixture. During extended dehydration, the heated gel will shrink severely, resulting in a very tough texture. F. FERMENTED IM MEATS Fermentation lowers the pH toward the isoelectric point of the meat proteins (Bacus, 1986),which assists in the removal of water from fermented sausages. Carbohydrates are added to provide food for the bacteria, and thus, accelerate fermentation. Acton et al. (1977) have discussed utilization of various carbohydrates and their role in the fermentation process. Fermentation and simultaneous dehydration of meat products is one of several basic processing steps (Campbell-Platt and Cook, 1995). However, few meat products are dehydrated as a separate process. In many processed meat products having a reduced moisture content, drying occurs simultaneously with ripening. Ripening occurs under controlled dehydration condi-
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
109
tions as explained by Romans et al. (1985). This process involves keeping the manufactured products for varying time periods under controlled temperature and humidity conditions. The length of the ripening period for different products varies from a few days to several months. The actual fermentation process requires specific environmental conditions for optimal product quality and has been described in reviews by Incze (1991, 1992). Larger dry sausage processors typically use a greening room to hold the product during fermentation. This room can be carefully controlled for temperature, humidity, and air velocity. Dry sausage is typically fermented at 20-25°C and 75-80% relative humidity with slow steady movement of air until the desired pH is attained (usually 1 to 3 days). The final pH of fermented sausages typically ranges from about 4.8 to 5.4, depending on the tanginess desired and the individual product. Carefully controlled environmental conditions are also critical during the drying process as explained by Bacus (1986). If the product dries too fast, a problem known as “case hardening” occurs. Case hardening refers to a condition in which the outside of the sausage becomes hard and dry, inhibiting further moisture migration from the interior of the sausage that still has a high moisture content. Sausages that have undergone case hardening are prone to internal spoilage by anaerobic bacteria. Case hardening is caused by the humidity being too low during drying. On the other hand, excessive humidity causes the product to dry too slowly and will often result in excessive mold or yeast growth and bacterial surface slime on the product surface. Theoretically, the drying rate at the surface of the sausage should be slightly greater than that required to remove the moisture which migrates from the inside of the sausage. Good drying conditions are achieved by various combinations of temperature, humidity, and air velocity. There is little agreement as the the most effective combination of these three variables. As a general recommendation, the temperature of the drying room should be maintained between 15 and 18°Cat a relative humidity of 70-72%. Air velocity in the room should be between 15 and 25 air changes per hour. According to Demeyer et al. (1986) a large amount of water is retained within the three-dimensional protein network of manufactured sausage batters. The water has two functions: (1) as a hydration layer it separates the coherent protein aggregates and (2) it constitutes a link between the protein threads via the hydrogen bonds. The firmness of the coagulated batter is still rather unstable and is weakened by the intermediate layer of water molecules. The coagulated structure of the batter is still capable of flowing and is flexible. Final firmness of the coagulated batter, however, can only be achieved by release of the immobilized water molecules that occupy the spaces between the protein aggregates. Fermentation and dehydration are thus essential prerequisites for firmness in the sausage batter
110
S. F. CHANG ef al.
as explained by Incze (1991,1992). For continuous dehydration, it is necessary that at the onset of dehydration conditions in the ripening chamber be properly adjusted and controlled, so that the degree of moisture differs only slightly (3 to 5%) between the sausage and air in the chamber (Incze, 1992). At this moisture differential, the sausage continuously loses considerable water. The loss of water is accelerated by an increase in nonprotein nitrogen and free amino acids as fermentation and drying occurs (DeMasi et al., 1990). Water losses from the interior of the sausage take place in part by diffusion due to the moisture difference between the sausage and the ripening chamber and partially by water lost by capillary moisture movement via slits and gaps as explained by Incze (1992). The formation of slits and gaps results from syneresis, halolytic, and enzymatic processes of decomposition between myofibrils and the larger meat particles, i.e., between the perimysium and endomysium as well as the muscle cells. Therefore, it can be concluded that the initially unstable coagulation bonds of the manufactured batter are gradually converted into firmer condensation bonds by acid denaturation and gradual drying of the sausage, i.e., the viscous protein system is transformed from the sol state into the colloidal viscous gel state (Demeyer et al., 1986; Incze, 1992). Indeed, water losses are achieved both in manufacturing and in the ripening-dehydration of the meat batter of fermented sausages. Many different chemical compounds have been identified as contributors to meat flavor by MacLeod (1986). Lipolysis and protein breakdown are responsible for development of flavor in fermented sausages and hams according to Demeyer et al. (1979a,b, 1986) and Verplaetse (1994). This is supported by results obtained by Dominguez-Fernandez and Zumalacarreui-Rodriguez (1991) during ripening of chorizos and by Lopez et al. (1992) for the hams from Iberian pigs, in which the diet of the pigs was shown to have a marked effect. However, the particular compounds responsible for the specific flavors have not been identified. Drying, no doubt, is responsible for the concentration of the compounds. G. DEHYDRATION BY HEATED OSMOSIS/INFUSION OF IM MEATS
Many drying techniques or treatments given to a food before drying are aimed at making the structure more porous so as to facilitate potential mass transfer and thereby speed up the drying rate (Demeyer et al., 1986). However, porous structures are excellent insulating bodies and will slow down the rate of heat transfer into the food. The net result depends on whether the change in porosity has a marked effect on the rate of mass
Water evaporated
NaCl (15)
t Lean pork
4
Water vapor
Water vapor
t
t
Semi-dry
+
disintegration
(1000)
Water
storage
finishing
t
t
t
Sucrose
Starch cell
Lard
(100)
or Dry-roasted product
(20-120)
(0-40)
FIG.2. Flow diagram showing the successive steps (processes) involved in production of Zousoon. The numbers in parentheses are given as weight ratio relative to raw meat = 1OOO. The starch cell preparation may be varied from 0 to 40 (weight ratio) and has a proximate composition of 22.4%protein, 68.0% carbohydrate, 1.9%fat, 0.8%ash, and 7.5%moisture. Source: Chang et aL (1991). Reprinted with kind permission from Elsevier Science Ltd.
112
S. F. CHANG etal.
transfer or heat transfer in the particular food material and drying system. A new direct dehydration method for producing Zousoon improves both heat and mass transfer efficiency throughout the process and uses the benefits of the readily controllable direct drying system as outlined by Chang et al. Figure 2 shows a flow diagram for production of Zousoon. In the traditional method, indirect heating is applied in the drying process, which results in specific structure problems as a result of an inadequate porous muscle system and poor time-temperature control of the product. The muscle is subjected to heating-osmosis, i.e., infusion, and boiled dry prior to dehydration as described by Chang et al. (1991). The compositional and structural properties of the muscle have been modified and most of the moisture in the muscle is removed during the predrying treatment. Chang et al. (1991) have discussed in some detail the possible moisture removal mechanisms of the heated meat mixture and described the design and operation profile developed to achieve balanced disintegration and drying of cooked muscle fiber bundles in a rotary dryer. The dryer was equipped with concave baffles to achieve greater shear force on the muscle bundle during tumbling in the dryer. The profile makes use of intermittent drying and tempering operations. In the process described by Chang et al. (1991) the heating (drying) cycles operated at 0-40,80-120, and 160-200 min at a temperature of 70°C with complete exhaustion of the air. During each period of heating, the moisture was removed along with the exhausted air, which decreased the moisture content of the muscle bundles. At the same time, the dry bulb temperature was increasing with a corresponding decrease in relative humidity. The tempering cycles were operated from 40 to 80 and from 120 to 160 min at a temperature setting of 35"C, with complete recirculation of the air but without exhausting it. During the tempering cycles, the moisture content of the muscle bundles was relatively constant since no moisture was removed by the air. At the same time, the dry bulb temperature of the air was decreasing and was accompanied by a corresponding increase in relative humidity. The tempering cycles were designed to minimize the moisture gradient within the muscle in order to prevent case hardening, which can result from excessive local surface drying while the interior moisture content is still high. The intermittent operation of the heating (drying) and tempering cycles helped to achieve a balance between drying and disintegration of the muscle bundles. Superheated steam drying-finishing of the predried muscle bundles makes use of the potential drying properties of steam at temperatures above 100°C as explained by Chang et al. (1991). However, the relative humidity at the steam temperature range is rather higher than that of the ambient air heated to the same temperature. Thus, superheated steam
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
113
drying-finishing achieved a greater drying effect than dry air could have done at the same temperature. The application of superheated steam in the finishing process was designed to achieve a controllable browning effect, which is characteristic and needed for the flavor and color of the product (Huang et al., 1989), but which cannot be fully controlled in the traditional indirect drying-finishing method (Chang et al., 1991). Overall results from the studies by Chang et al. (1991) indicate that finishing Zousoon with steam heat is a viable procedure, with the best color and consistency being achieved by using steam at a temperature of 150°C for 7 min. It may be possible to further refine the time and temperature for finishing Zousoon by controlling steam injection, temperature, steam velocity, and relative humidity during processing. Figure 3, which is taken from the research of Chang et al. (1991), shows a schematic diagram of a proposed mechanism by which muscle bundles disintegrate and form the fibrous characteristics that is typical of Zousoon at the end of the drying process. As indicated in the diagram tumbling creates a shear force, which causes disintegration of the muscle bundles as the heat is absorbed and the water evaporates. The combined forces cause the bundles to be separated into their component fibers, which on further drying and disintegration and final drying-finishing form the long fibrous Zousoon. The finished product is a light yellowish-brown in color and has a fibrous appearance, although some bundles persist due to the collagenous fibers holding small units together. The drying and disintegration processes should take place simultaneously and be balanced in order to achieve optimum drying-disintegration. H. MORPHOLOGICAL CHANGES IN MUSCLE BUNDLES DURING HEATING-DRYING Figure 4, which also comes from studies by Chang et al. (1991), illustrates the morphological changes that occur in the appearance of the muscle fiber bundles during balanced disintegration-drying in a convective heated rotary dryer. Figure 4A shows a bundle of fibers after removal from the boiling water and demonstrates that the fibers are bound together in a compact bundle. Figure 4B and 4C illustrates how the bundle size is gradually reduced by the effects of heating and tumbling during the early stage of predrying in the modified clothes dryer. Figures 4D, 4E, and 4F show how the apparent bundle size is expanded with the endomysial capillary moisture being removed. Figure 4F shows the texture of the muscle bundle as it becomes more loosened. The residual osmotic-held sarcoplasmic moisture may diffuse to the fiber surface and evaporate in the final stages of predrying. Each successive stage of convective heating in the predryer and
114
S. F. CHANG etal.
Heat
H20
r creating
Disintegrated structure formed
Component bundle held together
Component bundle separated
$.
$.
Further disintegrating & drying
FIG. 3. Proposed mechanism for muscle bundle disintegration during tumbling and drying. Simultaneous application of tumbling and heat creates shear forces and drives off moisture causing the muscles to separate into smaller bundles. Source: Chang el al. (1991). Reprinted with kind permission from Elsevier Science Ltd.
dryer-finisher results in greater separation of the large bundles into smaller bundles to achieve a proper (optimum) size bundle, which has a coarse hair-like exterior. The fibers still retain their identity as bundles due to the presence of some residual connective tissue (Chang et a/., 1991).
X. QUALITY ATTRIBUTES AS AFFECTED BY DEHYDRATION AND ITS ASSOCIATED PROCESSES Quality attributes on IM meats are mainly affected by a,, temperature, and the state of other crystalline components (i.e., sugars and other carbohy-
FIG. 4. Photographs showing normal pork muscle fibers during different predrying steps in production of Zousoon. (A) Muscle fibers following cooking before predrying. Note how fibers are held together in compact bundles. (B) and (C) show muscle fibers as they separate into smaller bundles during early predrying process. (D)and (E) show fibers as they progressively expand as drying continues. (F) shows Zousoon following predrying operation. Note the loosened fibers being held together by connective tissue. Source: Chang er al. (1991). Reproduced with kind permission from Elsevier Science Ltd.
This Page Intentionally Left Blank
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
115
drates). Troller and Christian (1978) and Troller (1972,1980) have pointed out that development of microbiology during the 18th and 19th centuries led to recognition that addition of salt, sugar, and dehydration was merely a way of preserving food through delaying or preventing microbial growth. The influence of such processes on other characteristics of foods, such as nonenzymatic browning, oxidative rancidity, flavor, texture, and nutritional quality, was only recognized and understood much later (Troller, 1987; Chang et al., 1991). Specific changes in color, aroma, flavor, texture, stability, and acceptability of raw and processed food products have been associated with relatively narrow a, ranges (Huang et al., 1989; Chang et al., 1991). The a, may have direct uncomplicated effects upon various chemical reactions (Labuza, 1980), enzymatic reactions (Schwirnmer 1980), and the proliferation of microorganisms (Troller, 1972, 1980; Troller and Christian, 1978). A. a, AS AN IMPORTANT PARAMETER Rockland and Nishi (1980) have explained that the a, theory involves the basic premise that independent and interdependent reactivities of individual chemical moieties are related to the aqueous molecular environment surrounding the reactive materials which affect the properties of natural products. Within heterogeneous systems, i.e., food products, the reactivity of each constituent is influenced by its affinity for water molecules and the competing influences of neighboring hydrophilic or hydrophobic chemical groups. Therefore, the structural-chemical architecture of the system is influenced by these forces. Changes in the environment-heat treatment, pH, modification of partical size, light, and pressure may alter the molecular state of the water and thereby influence the constituent reactivities and their functional properties. In more general terms, the properties of a system are influenced by the water binding energies of specific molecules and interactions among hydrophilic chemical constituents. The total binding energies of constituent chemical groups are reflected in the equilibrium water vapor pressure. At constant temperature, the vapor pressure may be expressed as equilibrium relative humidity, i.e., a,. It is clear that for some heterogeneous food systems, at least two a, optima exist, as indicated in the uppermost curve shown in Fig. 5 , which is taken from Rockland and Nishi (1980). This curve represents the relationship between a, and the integrated resultant relative stability based on the summation of a series of independent and/or interdependent chemical reactions, which are characterized diagrammatically in the figure. This figure presents an updated diagrammatic summary of a, and its major effects upon some chemical, enzymatic, and microbiological properties of foods.
116
S . F. CHANG et al.
MOISTURE
RELATIVE
FREE
FA1
WATER
ACTIVITY, % R.
H.
FIG. 5 . Diagrammatic representations of the influence of water activity upon some chemical, enzymatic, and microbiological changes and on overall stability and moisture sorption properties of food products. Source: Rockland and Nishi (1980).
Aguilera and Stanley (1990) pointed out that the term water activity as defined is valid only at equilibrium. Since equilibrium may not prevail in many processes or even during storage of foods, they pointed out that the concept of water activity has to be used cautiously.
B. INFLUENCE OF NONTHERMODYNAMIC FACTORS ON QUALITY OF IM FOODS Recently Gould (1989b) has suggested that the mobility or the diffusivity of molecules in foods, or the intrinsic viscosity, may be more useful than a, or equilibrium relative humidity as a meaningful determinant of biological and chemical activity. Extreme cases, such as glass transition formation, may raise viscosity by many orders of magnitude (Slade and Levine, 1987) so that for practical purposes biochemical and chemical changes are arrested. More accurately, such changes are greatly slowed down so as to be irrelevant over a normal time scale. However, such systems are metastable and therefore, are not at equilibrium, so they are not defined by equilibrium
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
117
concepts like thermodynamic water activity. Slade and Levine (1991) proposed that the potential for microbiological activity in partially dried or intermediate moisture foods could also be estimated through a knowledge of the distance (in terms of moisture content or temperature of a particular food or component) from the glass transition point. However, more work needs to be carried out before the potential value of such new correlations can be judged. Removal of water by evaporation results in formation of an amorphous state according to Vuataz (1988) and Roos and Karel (1991a,b, 1992). Amorphous foods are produced from carbohydrates by rapid cooling. The most important change that is characteristic of the amorphous state is noticed at the glass transition temperature (Tg), which involves transition from a solid “glassy” to a liquid-like “rubbery” state. Some diffusionlimited deteriorative reactions are controlled by the physical state in the vicinity of Tg. The “state diagram” shows the physical state is a function of temperature and concentration and illustrates the roles of relevant water content, temperature, and the time-dependent phenomena on the amorphous food components according to Roos and Karel(1991a,b, 1992). The viscosity at the glass transition temperature is about 10-11 Pa-sec or about 10-14 CP and it decreases above Tg. The most dramatic changes occur within a fairly narrow temperature range above Tg. At the temperature range of Tg to Tg + lOO”C, various time-dependent physical phenomena become evident, for example, stickiness, collapse, and crystallization. Roos and Karel (1991a) found that the widely used “sticky point” was governed by Tg. The critical viscosity for stickiness of about 10.7 Passec correlates with the viscosity at the end point of glass transition. Materials, such as fruit juices with high amount of monosaccharides (e.g., fructose), exhibit low Tg values and a low sticky point. Isoviscosity lines can be used to show critical viscosities for stickiness and the time needed to a given degree of crystallization or collapse, which may be used as a quantitative prediction of stability. Crystallization in the glass state below Tg is kinetically inhibited. It is initiated concurrently with structural changes above Tg, showing an increasing rate with increasing T-Tg. Crystallization can be accounted for by the moisture dependence of Tg according to Roos and Karel (1991b). Crystallization releases water, which in closed containers is absorbed by the amorphous portion of the food. As a result when Tg drops, T-Tg increases and rapid crystallization follows. Products, which have high moisture transfer rates in the environment, will lose water, but the moisture content in the amorphous part remains fairly constant. Crystallization proceeds at a rate defined by a constant T-Tg. Crystallization leads to a complete change in physical structure. It may considerably decrease stability. Lactose crystallization in milk powders
118
S. F. CHANG et al.
leads to an increase in free fat and concomitant flavor deterioration as explained by Roos and Karel(l992). They found that the increase in water activity caused by crystallization also increased the browning rate and loss of lysine in whey powder. Crystallization also resulted in complete release of any encapsulated compounds. Thus, volatiles are lost and lipids become exposed to oxygen. State diagrams showing the relationships between product composition and its physical state provide means for formulation or reformulation of food products to meet processing requirements and product stability during storage. State diagrams may also be used to design processes, equipment, packaging, and storage conditions that meet product requirements and achieve maximum stability (Roos and Karel, 1991a,b, 1992). It should also be noted that the Tg values of amorphous food components decrease linearly with increasing LZ, within the typical range of low and intermediate-moisture foods. Roos (1995) has pointed out that at temperatures about Tg various physical properties of foods, such as molecular mobility and viscosity and collapse (loss of structure) and crispness, are significantly affected.
C. QUALITY ATTRIBUTES OF IM MEATS Leistner (1987) stated that the most successful applications in development of intermediatemoisture meats have been utilized in production of pet foods. He gave several reasons why novel IM meats for humans have not become more acceptable: (1) newly developed intermediate-moisture meats are often not sufficiently palatable, (2) they are too expensive, (3) they contain too much additives (chemical overloading of the food), and (4)they may pose legal problems with respect to approval of greater amounts of new additives. However, there are many traditional IM meat products, which are highly acceptable in different parts of the world. In Europe meat products in the a, range 0.60-0.90 are not very common. Yet, traditional meat products, such as raw ham, fermented sausage, and dried beef, are dried sufficiently to reach an a, < 0.90. Traditional meat products in the intermediate-moisture range are frequently found in countries where the climate is hot and refrigeration is expensive or unavailable. For industrialized countries, traditional IM meats found in developing countries are of interest from the standpoint of shelf stability and adding variety to the diet. By studying the principles involved in production of IM meats, processing and shelf life may be improved without impairment to their sensory and nutritive properties. Leistner (1990a) pointed out that for most raw sausages, e.g., Italian or German salami, fermentation by microorganisms, which grow in the interior and also on the surface of some products, is essential for preservation and acceptance. An exception is Chinese raw
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
119
sausage, which is stabilized by a, only, since a low pH caused by fermentation would not be acceptable to Chinese consumers (Lin et al., 1983a,b; Kuo and Ockerman, 1985). Similarly, consumers of the Western world object to the sweet taste of Chinese dried meat. Buckle er af. (1988) observed that Indonesian dending has a sweet taste due to its high-sugar content together with the strong flavor of the spices. The sweetness and spices in the dried meat give dending a characteristic flavor, which is highly acceptable to most Asian consumers. Hodge (1953, 1976) stated that the appetizing aroma of fresh bread, coffee, roasting nuts, pralines, and barbecue depend largely on sugar-derived browning products. Although training in food science is not a prerequisite to appreciation of the art of the chef and his masterful control of induced browning reactions, it does require one with some scientific background to determine exactly which compounds contribute to the aroma and flavor of foods and how these sensory perceptions can be improved in less palatable products (Hodge, 1967).
D. CONTROL OF MICROBIAL GROWTH IN IM MEATS Leistner (1987) and Gould and Christian (1988) concluded that a, is the most important factor contributing to the storage stability of IM foods. Product a, in itself does not guarantee the safety of a product after processing, however, since deterioration may have already occurred before treatment. The moisture removal process as well as the associated stabilization operations and/or treatments also may play an important role in controlling microbial quality of meat products. 1. Effects of Temperature
Regardless of the specific mechanism, the osmoregulatory capacity of a particular cell determines to a large extent its osmotolerance in most practical situations. This is because the environmental solutes that are most often present in foods fall into classes that do not readily penetrate the cell membrane and are, therefore, able to effectively plasmolyze the cell (Gould and Christian, 1988).Muscle fibers are more or less subject to the same effect during osmosis treatment or dehydration. Consequently, any measurement that correlates closely with osmolality in a particular substrate will normally give a good indication of the potential for growth and the metabolic activity of particular groups of food spoilage and food poisoning microorganisms. Similarly, such data can be applied to favorable microorganisms utilized in the desired fermentation process. It is for this reason that direct or indirect measurements that allow rapid estimation of water activity or
120
S. F. CHANG
el
al.
equilibrium relative humidity have found increasing use. Scott (1953,1957) was the first to propose that as far as effects on the physiology of food spoilage and food poisoning microorganisms were concerned, the thermodynamic water activity (a,) was a key determinant, independent of the means by which the particular value was obtained. The growth limits for different types of microorganisms have since been commonly described in this way, with the “a, limits for growth” being quoted for each species or strain below which growth supposedly does not occur (Gould and Christian, 1988). Therefore, like Eh (available oxygen supply), pH values, and temperature, a , has been increasingly promoted as a practical useful determinant of microbial activity (Scott, 1957; Gould and Christian, 1988). Troller (1980) concluded that a , interacts with other stabilization factors, including numerous factors in the environment, to produce additive microbial inhibition. In fact, if one considers the many foods preserved by a , limitation, combination effects are the rule rather than the exception. While achieving the a, for IM meat products during the moisture removal process, the specific interaction among the other factors is more important. On the other hand, the interaction of temperature, with such factors as a,, pH, and chemicals, has been extensively discussed as it applies to thermally processed meats by Leistner (1978, 1987, 1990a) and Sebranek (1988). However, control and inhibition of microbial growth are often manipulated with all possible involved factors independent of the category of the meat product (Leistner, 1990a). 2. Interaction of Temperature with a , Troller (1980) stated that temperature interacts directly with a,, which suggests that a higher a, level will limit growth as the minimal or maximal temperature for growth is approached. In terms of heating lethality, however, a different picture emerges because very low a, levels are protective as opposed to heating at more elevated a, levels. In this context, the terms “wet heat” and “dry heat” are used to describe the water vapor concentration existing at the prevailing heating environment (Troller, 1980; Sebranek, 1988).Dry heat is irrelevant at lower a, values, unless the product is heated-dried to an equilibrium state. Salmonella species are especially heat-resistant in the absence of unbound water (low a,), particularly if sucrose is added to reduce the a , (Troller, 1980,1987). From the viewpoint of preservation, some IM meats are heated before the moisture is removed. That is, heating is applied at a high a , stage. On the other hand, some products are dehydrated before achieving a high temperature in the final stages, i.e., the heat is applied at low a,. In addition to the possible different lethality effects, there may be a difference in microbial recontamination
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
121
by the two processing procedures. In-package pasteurization of some IM meats has been shown to eliminate possible post-process recontamination before packaging (Chuah et al., 1988). Sanitary/aseptic packaging following heat pasteurization, such as “superheat steam flashing,” would be expected to be very efficient in microbial control. However, some IM meats are still preserved in the raw state, including Chinese sausage and Indonesian dending, with preconsumption heating being applied to achieve cooking effects before eating (Chow et al., 1989; Chang et al., 1991). 3. Effects of p H
Troller (1980, 1987) concluded that as the a, of a food is lowered, the pH limits within which growth will occur are narrowed. These effects have been described by Ohye and Christian (1967) for Clostridium perfringens and by Troller (1972) for Staphylococcus aureus. Similar effects occur with yeast and molds. 4.
Influence of Oxygen
The minimal a, at which growth will occur is lower under aerobic than anaerobic conditions for facultative organisms. For example, Scott (1953) demonstrated that the minimal a , at which growth of S. aureus occurred was 0.86 under aerobic conditions and 0.92 under anaerobic conditions. 5. Effect of Chemicals
The effects of nitrite, nitrate, and potassium sorbate have been determined by Gould and Measures (1977) and by Troller (1980). The role of combined nitrite and low a, (added NaCl) in preserving meat by inhibiting the growth of C. botulinum has been discussed by Gould and Measures (1977), Troller (1980), and Gould and Jones (1989). As a general rule, the minimal a , for growth of a microorganism is raised as the level of any additional inhibiting factor or substance is increased. Another way of stating this is that if bacterial growth factors are less than ideal, growth inhibition through a, reduction is increased (Troller and Christian, 1978). E. OTHER CONSIDERATIONS FOR CONTROLLING MICROBIAL GROWTH IN SOME IM MEATS 1. Nonheated IM Meats
Potter (1986) has discussed some nonsterilization aspects of the moisture removal process. Unless heated specifically for sterilization, virtually no
122
S.
F. CHANG et al.
quality dehydrated food emerges from a dryer in a sterile condition. While a large proportion of the microbial load may be killed during most drying operations, many bacterial spores are not affected. This becomes still more significant if the dehydration method is designed to be gentle to protect highly delicate foods. The nonsterilizing aspects of food dehydration also apply to certain natural food enzymes that may survive usual drying methods. In some foods, the necessity of retaining latent enzymes for later activation during hydration requires special drying regimens in which microorganisms cease to grow but latent enzymes may nevertheless survive as explained by Schwimmer (1980). In some cases microbial growth can, potentially at least, resume when these products are rehydrated and allowed to remain at room temperatures for prolonged periods. Hence, the microbial status of these products must be vigilantly monitored and reflected in specifications for such items. Microbial and enzyme management should be involved in the empirically estabished pattern which is emerging in the drying of food products. This suggests that in many instances superior product quality attainment is associated with slow removal of water at relatively low temperatures. Low-temperature-long-time drying has been used for the manufacture of such food products as smoked sausages according to Leistner (1987). Deeper insight and better understanding of the process afforded by fundamental studies with model systems and foods can be utilized to dispel much of the empirical approach currently pervading the processing of intermediate moisture and other low moisture foods. For some products, controlled microbial activity, i.e., fermentation, is achieved during dehydration and associated operations and/or treatments and are referred to as hurdles, which are discussed by Leistner (1987). The sequence of hurdles is intricate in fermented sausages, such as salami. In salami the hurdles occur in a sequence explained by Leistner (1987) and are particularly important at certain ripening stages in order to effectively inhibit food poisoning organisms (salmonella spp., C. botulinum and S. aureus) as well as other bacteria, yeasts, and molds which cause spoilage. On the other hand, the sequence of hurdles also favors selection for the desired competitive flora (lactic acid bacteria and nonpathogenic staphylococci), which contribute to the flavor and stability of fermented sausages (Leistner, 1987, 1990a). An important hurdle in the early stage of the ripening process for salami is nitrite, which is added with the curing salts, since addition of 125 mgfkg of sodium nitrite inhibits the growth of salmonella and clostridia as reviewed by Leistner (1987). The nitrite hurdle diminishes during the ripening process, since the nitrite is depleted. Due to the multiplication of bacteria in salami, the redox potential of the product decreases, and this in turn enhances the Eh hurdle, which inhibits the growth of aerobic organisms and
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
123
favors selection of competitive flora. The growth and metabolic activity of lactic acid bacteria, which then flourish, cause acidification of the product and, thus, an increase in the pH hurdle. This is of particular importance in the microbial stability of rapidly fermented sausages, which are not completely dried. Nitrite, Eh, competitive flora, and pH diminish with time, because in long ripening salami the nitrite level and the count of lactic acid bacteria decrease, while the Eh and pH somewhat increase. Therefore, only the a , hurdle is strengthened with time, and this hurdle, therefore, is mainly responsible for the stability of long-ripened fermented sausages (Leistner, 1987). Leistner (1987) also discussed the stability of raw ham, in which it is essential that the initial count of organisms in the interior of the product be low. The pH should be 6.0 or less and the temperature should below 5°C at the beginning of the curing process. The low temperature should be maintained until sufficient salt (i.e., 4.5% NaC1, which corresponds to an a, below 0.96) has penetrated into all parts of the ham. After the a , in the interior of the ham has decreased to 0.96 or below, the product can be further ripened and smoked at room temperature in order to achieve the desired flavor associated with enzymatic action. The microbial stability of traditional Chinese sausage is due mainly to the rapid reduction of a, according to Leistner (1987). This is aided by the addition of salt and sugar, the thinness of the casing, and a high drying temperature at a low relative humidity. On the other hand, the pH hurdle is not important for stability, because the pH of the raw sausage is relatively high and the number of lactic acid bacteria low. The principle used in the preservation of Chinese sausage, i.e., the quick decrease of a,, is also of interest, since the product demonstrates that raw sausage may also be successfully processed at 48°C and 65-70% relative humidity. Leistner (1987) also pointed out that the microbiologicalstability of Turkish pastirma is superior to biltong. The stability of biltong is apparently based on a low a , with little contribution from the pH hurdle, whereas, in traditional pastirma several hurdles are inherent according to Leistner (1987). These include the competitive flora (lactic acid bacteria), which probably contribute to inhibition of Enterobucferiuceae, and salmonella spp, and the cover paste, which is added to pastirma. Thus, besides a,, pH, and competitive flora, the preservatives present in garlic are also an effective hurdle. This results in inhibition of undesirable microorganisms, including toxigenic molds (Leistner, 1987).
2. Heated IM Meats The preparation of Chinese dried meats has been described in considerable detail by Leistner (1987). Dried pork slices are dried in the raw state
124
S. F. CHANG ef al.
and heat processed before packaging. Dried pieces of meat (cubes or bundles) are preheated and then dehydrated. On the other hand, dried pork floss and Zousoon are preheated and undergoes a disintegrationdehydration and finishing process. The microbial stability of Chinese dried meats depends primarily on the heat treatment. Chinese dried meats are indeed safe products, because the heat treatment eliminates most microorganisms present in the raw material as well as the survivors and other organisms, which recontaminate the product. These microorganisms are inhibited and inactivated by a, control. Molds and yeasts associated with IM meats have been discussed by Hocking (1988). He concluded that a, is the dominant factor governing the stability of IM meat products and has a strong influence upon the types of microorganismsgrowing on and spoiling these foods. Many manufactured meats, such as salamis, hams, and other cured meats, have a relatively high a, of 0.80-0.95 as stated by Hocking (1988). These products rely on a combination of factors for their microbial stability, e.g., nitrite, reduced pH, addition of salt, reduction Eh by vacuum packaging, and sometimes a heat treatment during manufacturing. Dried meats, such as biltong and Chinese dried meat products, rely primarily on reduction of a, for their stability. Consequently, the microflora of these two classes of meat products are somewhat different, although both groups contain large numbers of organisms (Leistner, 1987).
F. TEXTURE OF IM MEATS Bourne (1975) has described texture in foods and the methods used in its measurement. The texture of meat products has been described by Purslow (1987), who stated that meat texture is affected by the structure of the solid matrix. He concluded that it is important to have a fundamental understanding of the fracture behavior of meat and how it relates to the structure of the material. The long-term aim of his studies was to explain and predict variations in the perceived texture of meat on the basis of variation in composition and structure and hopefully to able to control and optimize texture by manipulation of these factors. Stanley (1983) stated that many researchers now believe the major structural factors affecting meat texture are associated with connective tissues and myofibrillar proteins. Therefore, he suggested that these structures merit particular attention. He concluded that two other components, muscle membranes and water, also deserve consideration, not because of their inherent physical properties, but rather as a result of the indirect influence they have on the physical properties. It should be noted that sarcoplasmic
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
125
proteins may be important for the same reason, although little information on their role is available. Chang and Pearson (1992) obtained results that indicated electrical stunning of hogs caused fragmentation and breakage of the muscle fibers so that the meat was not suitable for production of Zousoon-a semidry shredded Chinese pork product-and other similar items prepared from prerigor pig muscle. Heating and drying intensified the detachment of the myofibrils from the muscle fiber bundles, which was caused mainly by electrical stunning and improper conditioning following slaughter. The combined effects of electrical stunning and heating-drying appear to be responsible for the fragmentation of the muscle fiberdmyofibrils and contribute to the unsuitability of prerigor muscle from electrically stunned pigs for use in production of Zousoon. Katsaras and Budras (1992) reported that the protein matrix is important in production of the desired texture in fermented sausages that are suitable for slicing. They stated that formation of the protein network is predominantly induced by gelation and syneresis of myosin and actin during fermentation and drying. During chopping, the salt brings about a change in the original structure by causing swelling and partial solution of the myofibrils. The dissolved proteins are transformed into a fluid colloidal transition state or the so-called sol state with its unstable coagulation bonds. During sausage ripening, as a result of denaturation by lactic acid and due to gradual loss of water (drying), the unstable bonds are replaced by condensation bonds, and thus the sol state is converted into the gel state. Both gel formation (condensation) and water evaporation (syneresis) result in the development of the protein matrix in fermented sausage and, consequently, produce the texture in the sliceable product. Kuprianoff (1958) referred to the possible adverse effects of removing bound water from foods, which he enumerated as: (1) denaturation of protein by concentration of the solutes, (2) irreversible structural changes leading to textural modification upon rehydration, and (3) storage stability problems. Stanley (1983) stated that the water holding capacity of muscle is related to its sorption properties. H e suggested that although water accounts for approximately 75% by weight of the fresh tissue, more important than the total amount of water present is the water holding capacity (WHC) of the tissue. On the other hand, of greater importance of IM meats are the sorption properties of the muscle components. Thus, at an a, of 0.60-0.90, muscle could bind more moisture than at a,s outside of this range. The bound water in the muscle is primarily a result of its association with the myofibrillar proteins as indicated by Wismer-Pedersen (1971). Protein-water interactions significantly affect the physical properties of the meat (Hamm, 1960). Changes in WHC are closely related to
126
S . F. CHANG et aI.
pH and to the nature of the muscle proteins. Measuring the water vapor pressure (i.e., isotherms), however, might be a better index of water binding by muscle (Chang et al., 1991). The isoelectric point of a protein is defined as that pH at which the net charge is zero (Wismer-Pedersen, 1971). Since protein-protein ionic interactions are promoted at this point, it would be expected that the protein matrix would shrink and WHC would be at a minimum (Kapsalis, 1975). It follows that increasing the pH away from isoelectric point would also result in a higher WHC, since protein-water interactions are favored (Hamm, 1960). Bouton et al. (1971) were able to increase the ultimate pH and WHC of meat by preslaughter injection of epinephrine and showed that tenderness increased directly with pH values. Further work by Bouton et al. (1972) and Bouton and Harris (1972) showed that as pH increased from normal values of 5.5 to 7.0, tenderness of the tissue increased and became independent of the contracture state. G. EFFECT OF PRECOOKING
The use of humectants has been especially succcessful in production of semimoist pet foods, which commonly include high levels of sugars, propylene glycol, or sorbitol (Corbin, 1992). Direct dehydration without preheating maintains a structure, which favors water binding and enhanced tenderness in dried meat products. Accordingly, most IM meats are dehydrated from the raw state, except for some Chinese IM meat products which utilize prerigor or warm muscle as a raw material (Chang et al., 1991). PSE pig muscle is associated with a rapid pH drop immediately after death (WismerPedersen, 1960). PSE muscle is not suitable for production of IM meats, since the proteins are denatured (Bendall and Wismer-Pedersen, 1962; Bendall et al., 1963), which results in poor sorption properties. In comparison with fresh or frozen meat, freeze-dried meat is somewhat lower in tenderness and juiciness. Some of the difference is attributable to “woodiness”, although this characteristic is sometimes observed in fresh (frozen) meat (Lawrie, 1979).The benefits of high ultimate pH induced by preslaughter injection of adrenaline in protecting muscle proteins and in enhancing their water-holding capacity are reflected by greatly enhanced tenderness and diminished woodiness, with these benefits being retained after freeze drying according to Lawrie (1979). H. USE OF HUMECTANTS Humectants are edible substances that lower the a, to 0.60-0.90 and offer advantages in production. Webster et al. (1982) investigated utilization
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
127
of glycerol in production of IM meats. Later they also examined sorbate and glycerol and studied their effects on meat lipids (Webster et al., 1986). A specific humectant, such as sucrose, may be in an amorphous rubber state in the a , range after the dehydration-cooling process (Chang et al., 1991). The contribution of amorphous sucrose may help in explaining the plasticized texture in Chinese IM meats (Leistner, 1987), being distributed on the fiber surface in the meat mixture as a result of specific osmosis and the disintegration treatment (Chang et al., 1991). A cellular food, such as intact animal tissue, has a definitive structure and some rigidity at drying temperatures (Kapsalis, 1975). A concentrated sugar solution, on the other hand, lacks structure and softens and melts at some drying temperature (Chang et al., 1991). Thus, if a sugar solution is dried the solids will be in a thermoplastic tacky condition, giving the impression that they still contain moisture (Chang et al., 1991). On cooling, however, the thermoplastic solids may harden into an amorphous glass or a crystalline structure depending on the prevailing cooling rate and the residual moisture level (Potter, 1986). When components differing in a , are put into the same food system, those with a higher a, give up moisture to components of lower a , until the mixture reaches a single equilibrium a, (Potter, 1986; Chang et al., 1991). When these components are blended and reach the equilibrium a, of the mixture, each component retains different amounts of water in keeping with its individual water sorption isotherm and texture as outlined by Potter (1986). This principle is employed in producing complex mixtures such as American pemmican. In some foods the attainment of an a, low enough to inhibit microbial growth by dehydration alone will yield a product which is too dry for consumer consumption without adding water. The use of sugar, salt, glycerol, and other additives may impart undersirable flavors to the product and limit their use, i.e., sweet and sour pork, heavily salted meats, etc. (Potter, 1986; Leistner, 1987). Measurement of textural changes in dending demonstrates that these changes parallel sensory scores for toughness or hardness during extended storage according to Buckle et al. (1988). They suggested that this is partially due to moisture losses on storing dending in polyethylene bags, but also is related to changes in proteins and possibly nonenzymatic browning reactions. I. COLOR Color changes in cured meats are induced by curing and/or drying and during storage (Sebranek, 1988). Depending on the process of preservation, a meat mixture may be subject to temperatures between ambient and
128
S . F. CHANG et al.
freezing or up to above the boiling point. Forrest et al. (1975) have reviewed some factors that influence color changes and transformations occurring in the heme pigments of fresh and processed meats. In fermented sausages, Bacus (1986) has pointed out that cure development is necessary during fermentation and drying. When the curing mixture contains only nitrate, time must be allowed for growth of the nitrate-reducing bacteria in order to convert nitrate to nitrite. Unless nitrate is added, nitrite must be included in the cure to develop color. Okonkwo et al. (1992a,b) have discussed methods used to produce intermediate-moisture smoked meats. They prepared smoked beef by cook-soaWequilibrium in a solution containing sodium chloride, sodium nitrate, and potassium sorbate. Half of the samples were smoked for 18 hr (heavy smoking) and the other for 4 hr (light smoking) at 50°C. All samples developed the pink-red color of nitrite-cured meat. Sebranek (1988) has reviewed the effects of heat on denaturation of the proteins. Dehydration by heat denatures the muscle proteins, particularly the sarcoplasmic proteins. This induces a rather dramatic change in meat color. The heme pigments, which provide most of the color of fresh meat, serve as a general indication of doneness or temperature history. In the case of cured products, heme pigments react to form nitric oxide hemochromogen, which contributes the characteristic pink cured meat color (Pearson and Tauber, 1984). Maillard-type nonenzymatic browning reactions in processed meat products also contribute to their external surface color (Sebranek, 1988). Pearson et al. (1962, 1966) demonstrated that the main browning reaction involves the reaction of carbonyl compounds with amino groups, although lesser amounts of carbonyl browning also occur. Muscle usually contains small amounts of carbohydrates in the form of glycogen, reducing sugars and nucleotides, while the amino groups are readily available from the muscle proteins. Browning occurs at temperatures of 80-90°C and increases with time and temperature. A loss of both amino acids and sugars from the tissue occurs as a result of the browning reaction. Lysine, histidine, threonine, methionine, and cysteine are some of the amino acids that may become involved in browning (Hsieh et aL, 1980a). Maillard browning reactions are essentiaI in production of Chinese shredded dried pork (Zousoon) during processing as shown in studies by Yen et al. (1981). Although sucrose may be degraded during processing, its effect on color is slight since it is not a reducing sugar. Soy sauce as an ingredient of Zousoon plays an important role in browning of Zousoon, since it contains Maillard browning products and reducing sugars. The effects of various sugars on the browning of Zousoon was also studied by Yen et al. (1981). They found the effects to be in the following order:
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
129
glucose > fructose > sucrose. The desirable color of the finished products can be reached by using only a small amount of reducing sugars, such as glucose or fructose, and drying at a low temperature. Buckle et af. (1988) discussed the stability of dending and the role of nonenzymatic browning in color and flavor development. Significant amounts of browning can occur during storage of dending, even though the product is initially dark in color. They found that erroneously high results were obtained in measuring browning if absorbance due to the meat pigments and spices, etc. was not taken into account. Browning initially increased, but then decreased during extended storage, perhaps as the pigments become involved in interactions with oxidizing lipids. The addition of nonmeat ingredients, such as coconut sugar, to dending significantly increased the extent of nonenzymatic browning. Although coconut sugar increased browning, salt may cause a decrease if spices are present. During storage of dending, hydrolysis of sucrose occurs, with glucose being made available for the browning reaction (Buckle et af., 1988). Control of the Maillard reaction in Zousoon has been studied by Chang et af. (1991), who found that finishing Zousoon with steam heat was a viable procedure. The best color and consistency was achieved in a rotary finisher by using steam at a temperature of 150°C for 7 min. The steam dryingfinishing process aided in achieving a final a, of 0.60-0.65 and completion of browning. This process prevented drying from being too fast and causing inadequate mobility as explained by Karel et af. (1975). It also resulted in developing the proper product temperature-awlmoisture relationship to produce browning and slow moisture removal, which prevented product inhibition as described by Karel et af. (1967, 1975). Potter (1986) stated that Maillard browning proceeds most rapidly during drying if the moisture content is decreased to a range of 15-20%. As the moisture content drops further, the reaction rate slows, so that in products dried below 2% moisture further color change is not perceptible, even during subsequent storage. Drying systems or heating schedules generally are designed to dehydrate rapidly through the 15-20% moisture range so as to minimize the time for Maillard browning. However, some products, such as Zousoon may require longer periods of time in the 15-20% moisture range to develop the desirable color (Yen et af., 1981; Chang et af., 1991). Concomitant with browning, caramelization of sugars may occur and add to the brown color (Pearson et af., 1966; Nursten, 1986a,b). In carbohydrate foods browning can be controlled by removing or avoiding amines and, conversely, in protein foods by eliminating the reducing sugars (Pearson et al., 1962, 1966; Nursten, 1986a,b). Maillard browning can readily be distinguished from biologically derived alterations in food quality by inspection of a, vs response profiles or iso-
130
S.
F. CHANG
ef
al.
therms, such as the well known Labuza concept (Labuza, 1976,1980).These profiles reveal that nonbiologically derived changes exhibit both a maximum and a minimum. Nonenzymatic browning changes can be further distinguished from biological ones by their marked enhancement at elevated temperatures. An example of the practical consequences of browning is the problem of intermediate-moisture meat products stored in the tropics, which may develop other types of deteriorative alterations such as oxidative breakdown (Obanu et al., 1975a,b, 1977). J. AROMA/FLAVOR DEVELOPMENT AND RETENTION 1. Preheating
Nonsterilization during food dehydration can result in deterioration of quality due to survival of the indigenous food enzymes that survive drying according to Schwimmer (1980). Although precooking is widely used in IM and other meat products, it does have disadvantages. This is largely due to the effects of excessive heating. It should be possible, theoretically at least, to prevent enzymatic problems by alternative means. One would be by simply removing water without excessive heating, such as by freeze drying. Fortunately, one can frequently achieve the same results by removal of only a part of the water to lower the a, below a certain critical value as demonstrated by Schwimmer (1980). It is not always advantageous to inactivate or prevent the action of enzymes. Enzyme management may be involved in the empirically established pattern, which is emerging in the drying of food products, and suggests superior product quality can be attained by a slow removal of water at relatively low temperatures. Under such conditions, most enzymes are preserved and to some extent active. Low-temperature-long-time drying has been used for the manufacture of such diverse food products as smoked sausages, fruits, nuts, and nutritional supplements. In this range of temperatures cellular membranes are destroyed and enzymes are potentiated (Schwimmer, 1980).
2. Ripening The process of ripening involves keeping the processed product for varying periods of time under controlled temperature and humidity conditions as explained by Pearson and Tauber (1984). Development of a distinctive flavor results from microbial fermentation during ripening. Quite often products that are subjected to ripening are not fully heat processed, but are only subjected to a cold smoke. For example, semidry fermented sausages are heat processed at a minimum temperature of 57°C whereas, dry
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
131
sausages, such as summer sausage and salami, are never held above a temperature of 32°C (Bacus, 1986). 3. Aroma and Flavor Formation
Aromas and flavors may be produced by proteolysis as outlined by Garcia de Fernando and Fox (1991). Proteolysis in fermented sausages is caused by enzymes from the starter culture as well as by the indigenous meat enzymes. Tolda and Etherington (1988) found several cathepetic enzymes played an important role in production of dry-cured hams. Rico et al. (1991) later confirmed that cathepsin D was present in dry-cured hams and may be involved in flavor development. Understanding of the type and extent of proteolysis can help to achieve better quality sausages, since the nature and concentration of protein degradation products may contribute to the flavor and texture. Meat flavor development has been described as being similar to that of coffee or bread in that it is highly temperature dependent and contributes desirable aromas and flavors as explained by Sebranek (1988). Dry heating has been described as initially producing a flavor similar to moist heat, but as surface moisture evaporates, migration of soluble components to the meat surface occurs. This leads to continued concentration of compounds at the surface, which results in further chemical reactions that help to develop flavor and aroma. The carbonyl compounds that contribute to meat flavor are derived from fat as a result of heating and include several different aldehydes, ketones, and related compounds as shown by Sanderson et al. (1966) and are produced by condensation of carbohydrates with amino acids during cooking (nonenzymatic browning reactions) as explained by Sebranek (1988). They are believed to be significant flavor contributors. Carbonyls may not only contribute flavor themselves but also may be important in secondary reactions, such as formation of pyrazines according to Huang et al. (1989). These compounds are believed to be formed by reaction of carbonyls with amino-containing compounds or free ammonia. Pyrazines have been the basis of several patents involving synthetic meat flavors. Flavor development upon heating cured meat containing sodium nitrite has been reported to be different from that of uncured meat (Cross and Ziegler, 1965; Cho and Bratzler, 1970; Gray and Pearson, 1984). These studies show that the volatile compounds produced by cooked-cured meats differ from that in uncured products. 4. Control of AromdFlavor Development
Meat flavor can also be considered a transient characteristic since it continuously changes as heat is applied according to Sebranek (1988).
132
S.
F. CHANG
et
al,
Flavor desirability continues to increase as heat is applied. It reaches a point of maximum desirability and then deteriorates into harsh unpleasant flavors if heating continues. However, there is less information concerning specific temperature effects on the volatile compounds produced. Led1 (1987) indicated that in model systems, heating can produce flavor volatiles between 70 and 140°C. The specific compounds produced during heating are temperature dependent (MacLeod, 1986). The amount of volatiles produced can be correlated with heating time at a given temperature. Drying temperatures was shown by Huang et al. (1989) to be a critical parameter that determined the flavor quality of Zousoon. Cooked meat aroma increased directly as the heating temperature was increased from 134 to 172°C. Below 130°C neither cooked meat aroma nor brown color developed. Amino acid analysis confirmed that Maillard browning reactions occurred during the drying process, since basic amino acids such as lysine and arginine decreased significantly, whereas, other amino acids decreased only slightly. A selective purge-and-trap method was developed by Huang et al. (1989) to investigate pyrazine formation in Zousoon samples. The basic fractions of the collected volatiles identified in Zousoon contained 16 alkylpyrazines.A combination of these alkylpyrazines seemed to contribute to the characteristic cooked meat aroma of Zousoon. Although some oxidative changes may aid in development of the flavor of IM meats (Gray and Pearson, 1984), some toxic components, including cholesterol oxidation products, may be produced by oxidative reactions (Pearson et al., 1983). 5. Retention of AromalFlavor
The diffusion coefficient of water in concentrated solutions behaves differently from that of other substances according to King (1988). Diffusion coefficients of water and of other solutes decrease substantially as the water concentration falls in aqueous solutions of carbohydrates and other food components. However, the diffusion coefficient of water decreases by less than those of other substances. The result of this general phenomenon is that above some dissolved solids content, the diffusion coefficients of other substances become much less than that of water (King, 1988). Therefore, it is possible to reach a high concentration of dissolved solids at the surface of the material being dried before there is any massive losses of volatile flavor and aroma constituents. The remaining volatiles may become imprisoned because the surface becomes effectively impermeable to them. The retention of volatiles during spray drying can be improved if a high concentration of dissolved solids is built up on the surfaces of the droplets early enough in the drying process according to King (1988). This goal can be accomplished by rapid initial drying, since high rates of water evaporation and transport create substantial concentration gradients within the droplets.
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
133
Build-up of high surface concentrations of dissolved solids can be assisted by supplying a more concentrated liquid feed (e.g., by adding sucrose) to a spray dryer. This will result in lower coefficients and a lower gradient, which is needed to produce the critical surface concentration. The amorphous sucrose layer on the muscle fibers of Zousoon after boiling indicate that sucrose is deposited on the fibers by osmosis as reported by Chang et al. (1991), which is demonstrated by Fig. 6. Steam heating instantly increases the solid concentration on the fiber surface. However, when the internal fiber temperature reaches over 100°C, the internally generated steam passes through the concentrated melted sucrose layer. During fast cooling following finishing, the amorphous sucrose forms a layer on the fibers. Since the added sucrose has increased the initial dissolved solid concentration of the meat mixture and early fast finishing-drying achieved an increase in the surface solid concentration, most of the developed aroma/ flavor volatiles are retained by the amorphous sucrose layer. Thus, muscle fibers are encapsulated in the amorphous surface sucrose layer and contain concentrated aromas and flavors (Wientjes, 1968). 6. Control of Oxidation
The low awlmoisturecontent of IM meats may prevent lipid oxidation according to Simatos and Karel (1988). They presented data showing that water retards lipid oxidation in the intermediate moisture range. One factor that may be important is the production of browning products, which are known to have antioxidant activity. The effectiveness of nonenzymatic browning products in preventing lipid oxidation was demonstrated by Griffith and Johnson (1957) and is one of the mechanisms hypothesized by Karel(l986) to prevent lipid oxidation. Several investigators (Griffith and Johnson, 1957; Karel, 1986; Simatos and Karel, 1988) have confirmed that intermediates in the complex set of reactions involved in nonenzymatic browning are effective as antioxidants. Furthermore, intermediate-moisture contents would maximize the concentration of these intermediates according to Eichner and Ciner-Doruk (1981). Since high water activity promotes browning, this may be one of the explanations for the observed effects. Studies on purified model systems containing no components capable of forming antioxidants through browning have shown, however, that oxidation also can be retarded by increasing water content. The explanation for these effects is based on the fact that water is produced both at the initiation and termination steps of the chain reaction. In purified systems, water interferes with the normal bimolecular decomposition of hydroperoxides by hydrogen bonding with the amphipolar hydroperoxides formed at the lipid-water interface as outlined by Simatos and Karel(l988). In the pres-
134
S. F. CHANG et af.
FIG. 6. Scanning electron micrograph of muscle fiber covered with a layer of crystalline sucrose (arrows). Starch cell fragments labeled C are also shown. Source: Chang et nl. (1991). Reproduced with kind permission from Elsevier Science Ltd.
ence of trace metals added to systems of lyophilized emulsions, humidification at moderate levels retards oxidation because of hydration of metal ions. The reduction in rate depends on the type and hydration state of the added metal salt as well as on the water content. The monomolecular rate period is primarily affected by this mechanism, although bimolecular rates are also decreased (Karel et al., 1967). The effects of water on the destruction of the protective food structure in some specific dehydrated foods is probably involved in prevention of lipid oxidation in heated meat systems (Karel, 1986). In systems in which there are both surface lipids and lipids encapsulated within a carbohydrate, polysaccharide, or protein matrix, the surface lipids oxidize readily when exposed to air. The encapsulated lipids, however, do not oxidize until the structure of the encapsulated matrix is modified and/or destroyed by adsorption of water as shown by Simatos and Karel (1988). In some IM meats, muscle may be considered as being encased in or surrounded by a humectant matrix, However, free lipid may be left on the surfaces. The unwarranted overuse of lipids, which often happens in the indirect drying process to improve heat transfer and to prevent burning, is detrimental to the products. Prevention of this structural change is of considerable
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
135
importance in protecting unsaturated lipids from oxidation in dry preparations. The most dramatic demonstration of the increased capability of various reactants to diffuse as water activity increases is provided by the glassy transition stage formed by fast drying of sugar-containing foods as described by Roos and Karel(l992). When the “glassy” products are humidified at high water activities, they ‘‘collapse’’or crystallize, at which point some mobility is possible. The mobility allows the glassy products to be converted from an amorphous state to a mixture of crystals and amorphous components and permits the diffusion of various gases or vapors, which then may be encapsulated within the glassy material (Roos and Karel, 1992). The resistance of diffusionin foods is important, since the fat encased within the glassyproducts are protected from oxidation until the water activity of the matrix causes it to become sufficiently permeable to allow oxygen to penetrate. The diffusion coefficient of oxygen in sucrose solution decreases much more rapidly with increasing sucrose concentrations than does the diffusion coefficient of water (King, 1988). According to King (1988) when diffusional limitations exist, the local oxygen concentrations may be different from those that would be expected if the head space or surrounding air were instantly and continuously maintained in equilibrium. Studies related to oxidation of IM meats have been discussed by Okonkwo (1984) and Okonkwo et aZ. (1992a,b). With smoked beef prepared by cook-soak equilibrium in a solution containing sodium chloride, sodium nitrite, and potassium sorbate and smoked for 4 or 18 hr at 50”C, oxidation was not a serious problem. TBA values were low and all samples possessed no detectable rancidity. In dending prepared without nitrite, the chemical changes observed in the lipids were quite complex as shown in studies by Buckle et al. (1988). TBA numbers invariably decreased during the early stages of storage, perhaps as intermediates of lipid oxidation, as reactants in the Maillard reaction, or as browning products or their precursors that developed antioxidative activity. Subsequent increases in TBA numbers during storage (>6 months at 37°C) were not always related to the sensory assessment of rancid odors or flavors. Lipids are derived not only from the meat, but also from spices, such as coriander. Metal impurities occurring in the refining agent used to clarify sugar before crystallization or in unrefined salt may also increase the tendency of fat to undergo oxidation is discussed by Buckle et d. (1988). Production of dried sliced pork and control of various parameters utilized in production were described by Kuo and Ockerman (1985). They considered the effects of nitrate, packaging methods, and storage time on residual nitrite, m A values, and sensory properties. Residual nitrite decreased with increasing storage times at 1°C. The addition of nitrate plus vacuum
136
S.
F. CHANG
ef
al.
packaging caused a greater residual nitrite level and a lower TBA value during storage. Nitrite and/or nitrate acted as an antioxidant and retarded oxidative rancidity (TBA values). Dried pork manufactured by the technique described had no major rancidity problem and had an acceptable shelf life. Zousoon production methods were discussed by Lin et al. (1983a,b). They reported that the product packed in gas-impermeable bags under nitrogen or at atmospheric pressure could be kept at room temperature for at least 6 months without any major changes in quality. Vacuum packaging was unsuitable because of the poor appearance of the product and development of off-flavors during storage. Dried beef bundles are a popular product in Indonesia and their production has been described by Chuah et af. (1988). The finished products are packed in aluminum laminate pouches and vacuum sealed. They are then submerged in boiling water until the temperature at the center of the product reaches 95°C. The pouches are then cooled immediately in running water until the internal temperature reaches 37-40°C. This process is similar to that of pasteurized products having an a, < 0.85. After 6 months storage at room temperature, the products were still found to be acceptable with no rancid taste being detectable. The moisture content, free fatty acid, and peroxide values remained relatively unchanged during the entire storage period.
K. NUTRITIVE VALUE Erbersdobler (1986) stated that dehydration of food is one of the most important achievements in man’s history, making him less dependent upon a daily food supply even under adverse environmental conditions. Nutritional damage to food is of practical significance only if the complete daily diet does not provide an adequate intake of the nutrients in question as summarized by Schweigert (1987). In general, losses of B vitamins are usually less than 10% in dried foods. Dried foods do not greatly contribute to dietary requirements for thiamin, folk acid, and vitamin B-6. Although vitamin C is largely destroyed by heating-drying, meat per se is not a good source. Even though most amino acids are fairly resistant to heating-drying, lysine is quite heat labile and likely to be borderline or low in the diet of humans and especially so in developing countries where high quality animal proteins are scarce and expensive (Erbersdobler, 1986). In the case of protein, heat treatment of pork does not greatly affect retention as long as the critical temperature or time is not greatly exceeded according to Sebranek (1988). This has been suggested to be about 100°C and less than 1hr, respectively. Heating methods (microwave, steam, infra-
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
137
red, or hot air convection) do not appear to have any major effects on the amino acids. Losses of protein are not important as long as drippings are included for edible purposes. Some specific amino acids, however, may be lost during relatively severe heat treatments, i.e., lysine, methionine, and tryptophan as summarized by Schweigert (1987). Those losses most likely occur as a result of oxidative reactions, which are due to the direct effects of heating or from heat-induced crosslinking between amino acids. Crosslinks are not hydrolyzed during the digestive process. The nutrients most susceptible to loss through heating appear to be the B vitamins, particularly thiamin. With moist heat, losses of thiamin have been reported to average 60%; niacin and B-6, 50%; pantothenic acid, 40%; riboflavin, 30%; and B12,20% (Schweigert, 1987). With dry heating methods, a 20% loss is likely for thiamin, niacin, B-6, riboflavin, and B-12. Two trials were conducted by Lin et al. (1981) to determine the protein quality of Zousoon. The adjusted PER (protein efficiency ratio) values in the first trial for control samples (cooked freeze-dried pork) compared to Zousoon were 2.90 and 2.21, respectively, whereas, in the second trial the respective values were 3.10 and 2.28. Approximately a 25 to 30% decrease in PER and net protein utilization values for Zousoon were found in comparison to control samples. Compositional data show that the sulfurcontaining amino acids were the limiting ones in both the Zousoon and control samples. The Zousoon samples had lower methionine and lysine chemical scores than control samples. The reason for the decrease in protein quality of Zousoon was thought to be partly due to the addition of a lowquality food extender (starch cells or wheat flour) and partly because of the long-time-high-temperature drying process, which may cause further destruction of the amino acids. The available lysine content of dending is lower than that of control dehydrated meat samples according to Buckle et af. (1988). During storage at 50°C, available lysine levels decreased to about 60% of the original levels in the finished product. Muchtadi (1986) found the available lysine levels decreased substantially when dending was fried before serving. Protease inhibitors were found in both stored and fried dending but not in the boiled meat. The KCL-soluble N (nonprotein N) decreased by 22% during storage of dending at 50°C, while it increased slightly in dehydrated meat. Protein solubilized by sodium dodecyl sulfate and P-mercaptoethanol (i.e., denatured protein) decreased by more than 50% in dending after 3 months storage at 50"C, presumably due to development of stable cross-links and other reactions associated with browning. Dry pork sausage was formulated, fermented, and dried for 41 days and chemically analyzed by Garcia de Fernando and Fox (1991). The amount of water soluble nitrogen, water soluble nitrogen permeate, phosphotungs-
138
S . F. CHANG et al.
tic acid soluble nitrogen, and free amino acids all increased during processing. On the other hand, the salt soluble nitrogen and phosphotungstic acid insoluble nitrogen decreased. Electrophoretic studies demonstrated that proteolysis of the heavy myosin chain, a-actinin, and actin occurred during processing. XI.
PROCESS OPTIMIZATION FOR IM MEATS
A. BACKGROUND According to Schwimmer (1980) the two major considerations upon which successful adoption of a particular step in a food processing procedure hinges have traditionally been economic feasibility and consumer acceptance. He stated that these can be translated into innovation in processing efficiency and product quality and stability, with heavy emphasis on sensory quality and appearance. Such hitherto subsidiary parameters as nutritional, environmental, and energy conservation considerations are rapidly attaining equivalent status. The preparation and distribution of dry and dried foods, especially, intermediate-moisture foods, have proven to constitute no exception to this trend and have required contributions from applied and basic research to provide relevant scientific information upon which improvement in processing can be based. Saguy and Karel(l980) stated that improvements have been made possible by the increase in knowledge on the kinetics of food deterioration using advanced analytical methods and by the availability of computer modeling. The latter can simulate behavior of complex systems and save the time and expense of actual experiments. When no correction of the model is anticipated, the formulated model may be used for optimization, prediction and analysis. B. CONCEPT FOR OPTIMIZATION FOR HEAT PROCESSING OF MEAT The heating concept can be considered in many ways, but all require an understanding of the process involved, the conditions existing during the process, the range of possible results, and limitation of the equipment being used with each of these areas being reviewed by Sebranek (1988). In the case of heat processing of meat, Sebranek (1988) pointed out that the desired end results include a desirable color, flavor, texture development, gelation stability, and inactivation of the microorganisms and enzymes. Negative effects include loss of nutrients, oxidation of fat, moisture loss, and potentially undesirable flavor and texture changes. Consequently, processes first should be examined for basic principles, after which optimization can
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
139
be considered. The use of mathematical models to represent processes lends a great deal of potential to finding optimum processes. The objective of these models is to optimize sensory, chemical, and nutritional characteristics while achieving the necessary effects of heat processing. The concept of model thermal processes has potential when product variables are considered, for example, predicting the effects of ingredients or handling changes. Color, texture, and shrink can be predicted on the basis of ingredient functions or modified processing steps. The advantage of predictive modeling is that it is less expensive and time-consuming than is direct experimentation. However, any predictions based on models must be substantiated with experimental products to validate the effectiveness of the model. Improved basic research would increase the successful predictability of mathematical modeling. Additional and more precise information is needed on the thermal properties and heat transfer characteristics of meat products according to Roos and Karel (1991b). This is especially true during application of heat energy because thermal characteristics of meat are modified as the product changes during heating. Predictability of results from mathematical modeling has not always been highly successful, but is useful and should be pursued. A successful model, once developed, would be eminently useful in demonstrating the limits and flexibility of the heating process. Instead of a constant temperature or even constant step-up heat treatments (At), perhaps a process should be designed specifically for what is happening on a molecular basis. Such a process might be initiated to investigate the effects of high environmental temperatures. This could be followed by a step-down in temperature while passing through a critical heat rate zone for gelation and completed with a final finishing treatment to achieve microbial control. This approach might also be useful for new equipment applications as part of the process, for example, microwave heating by stages in a process where temperature rise can be achieved quickly. Optimization of the dehydration process for IM meats can be better understood by examining the basics involved in dehydration. The complexity of the phenomena involved, e.g., destruction of heat sensitive material, during concentration and drying is obvious. Optimization of such a process is, therefore, difficult. The approach used in traditional food technology is based on employing well-experienced craftsmen, who can make decisions on a case by case basis using a combination of objective and subjective criteria “integrated” into a decision in their minds (Thijssen and Kerkhof, 1977; Karel, 1988). Optimization fundamentals of the dehydration process for IM meats and for dried foods are likely the same. However, IM meats are dehydrated only to a water activity of 0.60-0.90 and are accompanied with a number
140
S. F. CHANG etal.
of side reactions in the process and by rather high-soluble solids in the formulation. The solids are derived from diffusion treatments. Indeed the products are the result of a stable state and can be stored at open atmosphere and need not to be stored in isolated or closed systems for stability. However, some physical-chemical and structural changes associated with dehydrated products may also occur in some low a, IM products. C . MATHEMATICAL MODELING FOR HEAT AND MASS TRANSFER
Aguilera and Stanley (1990) concluded that modeling is important in order to quantitate the effect of changes in variables and parameters on the drying rate, moisture content, and product temperature of foods. The ultimate objective is to be able to predict the final conditions in the dried product and the couse and extent of the reactions taking place during drying, such as browning and microbial growth andlor inhibition. Since the falling rate periods account for a major proportion of the drying time and is usually controlled by internal mass transfer, the model commonly has applied simple Fickian diffusion to the last stage of drying. Simultaneous with mass transfer, heat transfer also occurs. Lewis numbers greater than 60 or a characteristic dimension smaller than 3 cm thermal gradient can be neglected and the temperature considered uniform throughout the sample, but varying with time. When heat transfer is the limiting factor, then the thermal gradient inside the food particles increases. Fourier’s law can then be applied and the principles become similar to those governing mass transfer according to Aguilera and Stanley (1990). The important thermal properties of foods involved in the heat transfer process include thermal conductivity, specific heat, and thermal diffusivity. The transport of water in structured food materials is difficult to describe mathematically. Several correction criteria have been introduced to compensate for the complexity in mass transfer and microstructural effects. Improved results are obtained when variable diffusion coefficients are used. Moisture distribution profiles can be obtained by modeling Fick’s second law in which the diffusion coefficient varies with the moisture content. Mathematical models for drying in which particular physical or geometrical characteristics are relevant are also available. For example, Loncin (1988) developed a model for heterogeneous material having surface resistance. Another example can be found in the research of Crapiste et al. (1995) in drying of fruit in which the cellular structure of fruit was modeled as series/parallel arrangement with water flux between cells and along the walls being of a similar order of magnitude. This model is thought to be similar to the drying of meats and could be useful as a possible model.
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS 141
Engineering calculations based on model systems can go a long way toward selecting favorable drying condition, but seldom are sufficient per se to accurately predict drying behavior according to Potter (1986). This is because food materials are highly variable in initial composition, in amount of free and bound water, in shrinkage and solute migration patterns, and, most importantly, in the way the properties change throughout the drying operation. For these reasons, in selecting and optimizing a drying process, experimental tests with the food to be dried must always supplement engineering calculations based on less variable model systems (Lois et al., 1987).
D. DEVELOPMENT OF AN APPROPRIATE KINETIC MODEL Karel (1988) has investigated concentration effects in a kinetic model, in which concentration was chosen as the index of quality. He 'stated that quality loss may then be represented as dQ/dt = -dCldt
- k Cn - C n2 . . .,
where Q is the quality index, C is concentration of the heat sensitive components, n is reaction order, d is reaction rate, t is time, and k is a constant. Various reaction orders may be involved. Nonenzymatic browning very often follows zero-order kinetics. 1.
Temperature Dependence
Karel(l988) concluded that temperature dependence behaves according to the Arrhenius equation
K
=
KOexp (-EJRT),
where K is the reaction rate constant, KO is a constant, E, is activation energy, R is the ideal gas constant, and T is absolute temperature. This equation constitutes the soundest approach for modeling temperature dependence. 2.
Water Activity and Moisture Content
The equilibrium relationship between water activity and moisture content is expressed by sorption isotherms and may be used to specify local moisture contents or to calculate average moisture content as outlined by Karel (1988)
142
S . F. CHANG et al.
where m is moisture content, f is a function of moisture content, and x,y,z are the coordinates specifying positions. In many food and biological systems, physical and/or chemical changes occur concurrently with, and are often caused by, sorption of water. The kinetic model requires knowledge of the dependence of the reaction rate constant on either m or a,. These relationships can be extremely complex, and simplified relations are needed. For example, in many oxidative reactions at a very low moisture content, a useful mathematical model is K = b/m n,
where K is a reaction rate constant, b is a constant, m is moisture content, and n is an exponent with a value usually close to 1.0. Karel(l988) has discussed some problems arising from complex kinetics or physical phenomena. Optimization for retention of heat-sensitive material becomes very difficult when the reaction kinetics cannot be described simply or when the properties of the materials being heated and/or dried change dramatically during the process. For example, Karel (1988) stated that many of the reactions causing deterioration of food quality involve free radicals, and the kinetics are those of chain reactions. These kinetics may be simplified by focusing on only one of the several potential indicators of oxidation and by simplifying the assumption. He further suggested that changes in structure, and particularly in glass transitions, have a profound effect on rates of chemical reactions for heat-sensitive materials and that water content affects these transitions substantially. Diffusion of sugars is also important in nonenzymatic browning. Considerable amounts of work have been conducted in these areas, with research being carried out in connection with diffusion of water and of flavors during drying. The diffusion coefficient for organic compounds drops even more rapidly than that for water. This phenomena is the basis of the selective diffusion theory, which was formulated by Thijessen and Kerkhof (1977). This theory is generally accepted with some modifications as the basis for flavor retention during drying of foods. 3, An Example of Optimization
An applicable example of optimization, which was derived from dehydration of a model dehydrated food, has been described by Mishkin et al. (1982). They suggested that food engineering has lagged behind other engineering disciplines in implementing process optimization techniques
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
143
because the complex characteristics of foods makes mathematical modeling of their behavior difficult. For illustrative purposes, they chose a dehydration system of slabs composed of water, cellulose, and ascorbic acid, which was dried in a tray dryer. In this process, their aim was to vary the air temperature during drying in a manner that resulted in the highest ascorbic acid retention while achieving the desired final moisture content at any specific drying time. XII.
ENERGY COSTS FOR PRODUCTION OF IM MEAT PRODUCTS
A. ENERGY CONSUMPTION IN FOOD PROCESSING In food process operations, energy requirements must be examined in relation to their contribution to total operating costs according to Flink (1977a,b). Whenever efforts are made to improve the economic efficiency of a process by reducing operating costs, energy utilization must be among the factors considered. In the past, however, when energy costs were relatively cheap, there was little incentive to make significant capital investment to improve the design or operation merely to reduce consumption of energy.
B. ENERGY COSTS ASSOCIATED WITH IM MEATS Flink (1977a,b) has pointed out that energy costs have always been a factor of importance and concern in the food processing industry. Most aspects of production of food from the farm through the factory to the home require significant inputs of energy. Hirst (1974) has estimated that the U.S. food cycle uses about 12% of the total U.S. energy. Of this amount of energy, food processing accounts for almost 50%, with distribution and trade (wholesale and retail marketing) adding an additional 20%. Much of the remainder is energy use associated with purchasing, storage, and cooking by the ultimate consumer. For IM meats, the product could be naturally dried, even without using solar energy, to a characteristic a,,,, which is in equilibrium with the relative humidity or the atmosphere, to which the meat is exposed. However, dehydrated foods need extra energy in order to remove the moisture and reach an a, below the relative humidity of the open atmosphere. Often times, the low moisture levels for maximum product stability of a food are not easily obtained and frequently can be approached only at the expense of increased dehydration costs. For example, the high costs of freeze drying salted prerigor meat in order to maintain a high water holding capacity after rehydration make the process too expensive to be practical (Judge et al., 1981). The dehydrated meat is excellent
144
S. F. CHANG et at.
as raw material for various sausage types but costs for freeze drying are prohibitive. 1. Energy Analysis of the Dehydration Process
Flink (1977a) concluded that air drying is the cheapest in terms of energy costs and is followed by drum or pan drying while freeze drying is the most expensive process. He showed that the basic energy cost to remove 1 kg of water is much lower for air drying and drum drying than for freeze drying. Shin (1984) reported on the energy requirements for producing IM meats, while Stiebing et al. (1982) investigated energy savings during manufacturing of raw sausages.
2. Factors Affecting Energy Efficiency Flink (1977a,b) pointed out that energy efficiency can affect the economic viability of the dehydration process and will depend on the weight of product per unit weight of water removed. The higher the initial concentration of solids in the food, the more economical the drying process, provided that the concentration process used prior to drying has a lower cost for removing water than the drying process per se. One way to accomplish efficiency in the drying operation is to increase the inlet air temperature to the drying chamber (Chang et al., 1991). This, in part, is related to the reduction in drying rates observed at lower air temperatures. When these kinetic factors are negligible, however, the opposite behavior occurs. As most drying operations have some heat and mass transfer kinetic limitations, the air discharge temperature can be reduced by lowering the air flow rate in the dryer as described by Chang et al. (1991). This will allow a longer contact time between the sample and the heating medium. The effective air discharge temperature can be reduced by using air recirculation, i.e., by operating the dryer as a somewhat closed system in which the humid air at the dryer outlet is, for the most part, reheated and used again. In this case, the amount of energy being discharged in the exit air per unit of water removed is reduced, and the energy input required to heat the air being introduced to the drying chamber is lowered by the difference in temperature between the “discharge” air and the ambient air. There will, however, be some effects associated with the increase of average absolute humidity of the air in the process. Since IM meats do not require complete drying, the process of recirculating moist air improves the overall efficiency of drying. Air recirculation will also increase the specific heat of the circulating air, thus reducing the required volumetric flow.
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
145
Most current industrial approaches involve energy reclamation in the air dryer by either heat exchange or recycling. The potential advantages of the use of osmosis as a method of increasing the solid contents of the feed material could be useful in future development of dehydration, such as for IM meats (Chang et al., 1991).
3. Equipment for Producing IM Meats Although the wind and sun have long been used by man to produce dried foods, this can only be accomplished under favorable climatic conditions as outlined by Okonkwo et al. (1992a,b). However, some novel drying methods had been proposed by the end of the 19th century. At the beginning of this century, Hausbrand (1901) wrote a short monograph on the heating and ventilation for a given drying load, but little further work was done until recently. Outstanding examples of such later work are Lwikov’s (1966) analysis of heat and moisture transfer and their inseparable association and Krischers’ (1963) treatise on the scientific basis for drying technology. An infusion of new knowledge of this kind into old arts is likely to bring about significant saving in costs and radical changes in practice. Processing equipment used in traditional meat-drying is simplistic and, in some cases, primitive, making process control difficult, if not impossible. The quality of different batches of product may differ due to poor quality control, resulting from lack of proper equipment. However, these production problems are slowly being overcome with the introduction of more modern equipment, which not only speeds up the rate of production but also results in more consistent quality products due to more precise control of the processing parameters (Chang et al., 1991). Ockerman and Kuo (1982) described production of dried pork by tumbling and forced air drying techniques, which they compared to the more conventional method of dry curing with the sun and nonheated air-drying. The latter method requires more time and labor in addition to being unsanitary due to the exposure during drying. The predried pork was cooked (180 + 2°C) on a grill for 1 min on each side. They used many methods for final finishing, such as deep fat frying, radiant heat roasting, and open fire roasting. The production of Zousoon was described by Chang et al. (1991) as traditionally produced in Taiwan. The production process is empirical and is more of an art than a science. These workers described a modified clothes dryer developed to tumble and dry the product, which was shown to result in improved heat transfer and greater shear force that gives better control of evaporation of water while causing the muscle bundles to disintegrate into smaller subunits. The predried product was finished in a steam-heated
146
S . F. CHANG 'et d.
dryer-finisher. The final product had a yellowish-brown color and fibrous appearance, being more uniform in color and texture than that produced in the traditional gas-fired, scraping frypan. Leistner (1990a) has described the development of dry and semidry sausages which originated in Southern Europe. He pointed out that these products first were produced by fermentation induced by naturally occurring microorganisms, but today commonly use starter cultures that preserve the sausages by fermentation and drying. Stiebing et al. (1982) introduced an improved method for control of relative humidity in the ripening room during sausage fermentation and achieved an energy saving of up to 70%. By using the surface a, of the sausages as a measure of control, Stiebing and R6del (1989) suggested that optimization and automation of the ripening process could be controlled with microprocessors. Since freeze drying requires high capital investment and is in itself an expensive process of drying (Judge et al., 19Sl), it has not been widely used by the meat industry. Users of the freeze-dried products have encountered no major problems except for the high costs. Some studies with IM meats have involved infusion of a low a, solution into the sponge-like structure of freeze-dried meat to improve its absorptive properties. 4. Control System
Most thermal dryers embody convective/direct heating since drying can be readily controlled by the temperature and humidity of the air that evaporates and conveys away the moisture as explained by Chang et al. (1991). There is some insurance against overheating the drying materials, since its temperature can never exceed that of the incoming air. On the other hand, drying rates for the various methods of indirect drying are not as readily estimated as those for direct drying because the heat and mass transfer coefficientsat the point of contact between solid phases are not well established. Frequently, radiation and conduction cause the temperature of evaporation to exceed the wet-bulb temperature of the air in the early stages of drying. It is necessary to estimate the true surface temperature in order to calculate the constant rate. The true surface temperature, however, can be estimated from heat transfer data using either graphical or trial and error methods. C. FACTORS AFFECTING DRYER SELECTION The type of dryer selected should be determined by: (1) physical characteristics of the material being handled when wet and dry, and (2) the drying properties of the material, which include (a) type of moisture being
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
147
removed, i.e., free or bound moisture, (b) initial and final moisture content, and (c) the final product quality, such as shrinkage, state of subdivision, bulk density, contamination, overdrying, and chemical changes (Chang et al., 1991). The interior surface of a rotary dryer used in production of Zousoon is equipped with baffles, which are longitudinal fins extending inward toward the center, and was described by Chang el al. (1991). The solid material is continuously carried up by rotation of the fins until it spills off and falls through the air. The high degree of turbulence and excellent contact between the air and solids provide very high rates of drying for materials that have this capability. Material that would tend to mat or pack on a tray or belt may easily be handled in a rotary dryer. The abrasion caused by the tumbling action rules out the rotary dryer for many food materials. Furthermore, particle sizes may frequently be too large to take full advantage of rapid drying by use of the rotary dryer. XIII.
RESEARCH NEEDS FOR IM MEATS
A. EFFECTS O F ANTE- AND POSTMORTEM TREATMENTS O N PROPERTIES OF IM MEATS It was pointed out by Chang et al. (1991) that prerigor meat was preferred for production of some Chinese IM meat products, as it yielded superior products. Chang and Pearson (1992) showed that meat from electrically stunned hogs was not suitable for producing Zousoon, since the muscle fibers did not retain the fibrous structure which is characteristic of this product. Yet there is little, if any, information on the effects of various prerigor and postrigor treatments on the structural and physical properties of other IM meats. Experimentation is needed to see if prerigor treatments improve the quality of other IM meats, especially other Chinese IM products. What are the effects of using prerigor meat on the efficiency of the drying process? It has been shown by several researchers (Bacus, 1986; Demeyer et af., 1986; Katsaras and Budras, 1992) that lowering the pH toward the isoelectric point of the muscle proteins aids in the dehydration process. This would suggest that the efficiency of water removal would be improved in postrigor meat, particularly at low pH values. Since PSE pig muscle loses water readily (Briskey, 1964; Borosova and Oreshkin, 1992), it would be interesting to see if its use would improve the efficiency of drying. If not, would there be an advantage for using dark, firm, and dry muscle for production of IM meats? What whould be the effect of altering postmortem
148
S.
F. CHANG et al.
pH by feeding sugar to pigs, or, conversely, to deplete the muscle glycogen by exhaustive exercise as reported by Briskey et al. (1959) on the efficiency of drying and the quality of IM meats? If either a low or high pH should have an advantage for production of IM meats, this could be achieved by enzymatic manipulation as explained by Bouton and Harris (1972) and by Bouton et al. (1971, 1972). B. INFLUENCE OF FREEZING AND THAWING Pearson and Miller (1950) studied the effects of freezing rate and length of freezer storage on the quality of beef and found that freezing rate was relatively unimportant but drip losses increased during 90 days of freezer storage. The increased losses of water associated with freezing and freezer storage suggest that this phenomenon could be utilized to assist in removal of water in order to lower the a, during production of IM meats. It has been shown by Judge et al. (1981) that freeze drying produces a high quality product that is useful as a sausage ingredient. However, the process is too expensive to be economically feasible. It is not known whether there would be an advantage in using meat that has been frozen and thawed with the resultant water losses improving the efficiency of water removal. Studies need to be carried out to see if this technique could be used to improve transfer of moisture in both conventional drying and freeze drying of meat. C. FERMENTATION It has been known for many years that fermentation plays an important role in production of IM meats (Lawrie, 1995; Campbell-Platt, 1995; Zeuthen, 1995; Krockel, 1995;Leistner, 1995). The fermentation process assists in preservation in two ways as pointed out by the above-listed authors. First, the production of acid during fermentation lowers the pH and inhibits growth of spoilage microorganisms.Second the drying process is accelerated by fermentation since the lowered pH assists in removal of moisture as it moves toward the isoelectric point of the meat proteins. Thus, the acidic pH helps in lowering the a,. Although bacterial fermentation is commonly used in producing sausages (Bacus, 1986), it has not been commonly used in other products. Jessen (1995) has reviewed the procedures involved in production of meat products by bacterial cultures. Thus, research on the use of bacterial fermentation in production of other meat products is needed. Lowering the pH by adding acids should also be explored as a means of lowering a, and imparting an acidic environment, i.e., especially the use of organic acids.
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS 149
Fungal growth is generally considered to be responsible for toxins in meat (Pestka, 1995), yet some traditional fermented meat products are produced by selected fungal strains (Cook, 1995). Leistner (1990a,b) and Leistner et al. (1989) have demonstrated that a commercial culture of Penicillium nalgiovense can be utilized in production of mold-ripened sausages. The number of molds and yeasts that have been investigated as cultures for meat fermentation still need further investigation (Pestka, 1995). A dense coating of P. nalgiovense protects against growth of undesirable molds (Leistner, 1995).However, the mechanism by which they protect against undesirable molds needs further study. Yeasts also have been utilized as starter cultures in sausages according to Leistner (1999, usually Debaryornyces hansenii, which reduces the Eh and causes the meat to turn red. The cultures produce catalase, which delays development of rancidity and improves the aroma of the sausages as was demonstrated by Miteva et al. (1989). Nevertheless, the use of yeasts for fermentation of meat needs additional research particularly in regard to the utilization of different strains and their mechanism(s) of action. D. SYNERGISTIC STABILIZATION
Leistner (1978,1987,1995) has explained the synergisticaffects of various treatments, such as salt, nitrite, smoke, and a,,,, on production of IM meats. Some or all of these factors may be involved in providing hurdles that assist in producing safety in IM meats. Further research is still needed to ascertain the role of each hurdle in preventing both spoilage and growth of food pathogens in various IM meat products as pointed out by Leistner (1987).
E. EFFECTS OF HEAT Although application of heat or cooking has been widely utilized in production of IM meats (Keey, 1972; Ledward, 1981; Okonkwo 1984; Buckle et al., 1988; Leistner, 1987; Chang et al., 1991), only a few of these investigations have concentrated on the role of heat in the transfer of moisture from the product to the environment. Large scale studies are needed on factors influencing the efficiency of water removal. F. EFFECTS OF OSMOTIC TREATMENTS Although Lerici et al. (1988) demonstrated that osmotic treatment greatly affected drying rate, little research has been done on a combination of heat and osmotic treatment. It may be that altering the solid matrix by adding salt and sugar assists in removal of water, yet such information is not available.
150
S.
F. CHANG
et
al.
Offer and Trinick (1983) and Offer and Knight (1988a) have demonstrated that osmotic treatment causes shrinkage of the muscle fibers. This may assist in removal of water during heating and not only prevent growth of food spoilage and pathogenic microorganisms, but also aid in lowering the a,. Investigations are needed to ascertain the role of altering osmotic pressure by use of additives, such as salt and sugar, on the efficiency and rate of water removal in production of IM meats. G. GEL FORMATION In a number of sausage products, gelation is involved (Aguilera and Stanley, 1990). This phenomenon probably plays a key role during fermentation, although virtually nothing is known about the effects of gelation on the rate of water removal. Since gels bind water, the question needs answering if gel formation slows down water losses or if by some unknown mechanism may assist in water removal. Since fermentation lowers the pH toward the isoelectric point of the muscle proteins perhaps gel formation is not involved in fermented meat products. These questions need to be answered in order to help in understanding whether gelation is involved in water transfer of IM meats. H. ROLE OF GLASS TRANSITION Slade and Levine (1987,1991) have shown that heated foods may form a glassy transition phase, yet little is known about the effects of this transition. Does it accelerate or delay heat transfer? Does the glass transition have an advantage or disadvantage in production of IM meats? These questions need to be answered before recommendations can be made on the influence of the glass transition on drying. I. MODELING Mathematical modeling of the effects of drying on IM meats are needed similar to those carried out by Karel(1986,1988). However, results of the modeling studies then need to be applied to the IM meat system to show that the results are applicable in the IM meat system. The model is only good if the results are applicable to IM meat products. J. DESIGN OF DRYING EQUIPMENT
Chang et al. (1991) demonstrated that the process of producing the Chinese IM meat product, Zousoon, could be improved by developing new
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS 151
equipment. They also demonstrated that recirculation of the air improved the efficiency of the process. The need for new equipment designed for more efficient operation and more effective removal of water during the drying process is an urgent one. Newly designed equipment for producing IM meat products could lead to their wider availability and greater popularity. K. THERMAL DATA
Chang et al. (1991) pointed out that data is needed on the properties of IM meats that affect drying rates and efficiency of drying. What compositional and structural constraints result in changes in drying efficiency? Furthermore, the factors controlling mass transfer need to be elucidated. Mass transfer efficiency vs heat transfer efficiency needs to be understood in order to improve the efficiency of drying IM meats. L. HACCP Systems Mortimore and Wallace (1994) have published a practical approach to HACCP, which needs to be developed for IM meats. HACCP is important since salami has been found to have been contaminated with Escherichia coli 0157:H7 (Anonymous, 1995a) and salmonella organisms have survived processing in Lebanon bologna (Anonymous, 1995b). With emphasis on food safety, critical control points (CPs) in production of different IM products must be developed. The CPs will, no doubt, be quite different from product to product. Thus, systems must be developed for each different process and put in place before production parameters are set. One cannot assume that IM meats are safe, but recognize possible problems before they cause difficulty. HACCP must be an important part of each IM meat product. Safety will be crucial to the success of all products and must be developed during the experimental stage of production. XIV. SUMMARY
IM meat products are produced by lowering the a, to 0.90 to 0.60. Such products are stable at ambient temperature and humidity and are produced in nearly every country in the world, especially in developing areas where refrigeration is limited or unavailable. Traditionally IM meats use low cost sources of energy for drying, such as sun drying, addition of salt, or fermentation. Products produced by different processes are of interest since they do not require refrigeration during distribution and storage.
152
S. F. CHANG et al.
Many different IM meat products can be produced by utilizing modern processing equipment and methods. Production can be achieved in a relatively short period of time and their advantages during marketing and distribution can be utilized. Nevertheless, a better understanding of the principles involved in heat transfer and efficiency of production are still needed to increase efficiency of processing. A basic understanding of the influence of water vapor pressure and sorption phenomena on water activity can materially improve the efficiency of drying of IM meats. Predrying treatments, such as fermentation and humidity control, can also be taken advantage of during the dehydration process. Such information can lead to process optimization and reduction of energy costs during production of IM meats. The development of sound science-based methods to assure the production of high-quality and nutritious IM meats is needed. Finally, such products also must be free of pathogenic microorganisms to assure their success in production and marketing.
REFERENCES Acton, J. C., Dick, R. L., and Norris, E. L. (1977). Utilization of various carbohydrates in fermented sausages. J. Food Sci. 42,174. Aguilera, J. M., and Stanley, D. W. (1990). “Microstructural Principles of Food Processing and Engineering.” Elsevier, London. Anonymous (1995a). “Lean Trimmings,” Sept. 24, p.1. National Meat Association, Oakland, CA. Anonymous (1995b). “Lean Trimmings,” Oct. 23, p.4. National Meat Association, Oakland, CA. Ashbrook, F. G. (1955). “Butchering, Processing and Preservation of Meat.” Van Nostrand, New York. Astiasaran, I., Villanueva, R., and Bello, J. (1990). Analysis of proteolysis and protein insolubility during the manufacture of some varieties of dry sausage. Meat Sci. 28,111. Atkins, A. G. (1987). The basic principles of mechanical failure in biological systems. In “Food Structure and Behavior” (J. M. V. Blanshard and P. Lillford, eds.), p. 149. Academic Press, San Diego, CA. Bacus, J. N. (1986). Fermented meat and poultry products. Adv. Meat Res. 2, 123. Bendall, J. R. (1969). “Muscle, Molecules and Movement.” Heinemann, London. Bendall, J. R., and Wismer-Pedersen, J. (1962). Some properties of the fibrillar proteins of normal and watery pork muscle. J. Food Sci. 27, 144. Bendall, J. R., Hallund, O., and Wismer-Pedersen, J. (1963). Postmortem changes in the muscles of Landrace pigs. J. Food Sci. 28, 156. Berkman, L. (1960). fiber die Haltbarkeit von Krankheitserregern in einem spezifisch tUrkischen Fleischerzeugnis. Fleischwirtschaft 40,926. Binkerd, E. F., Kolari, 0. F.. and Tracy, Y. (1976). Pemmican. Proc. Annu. Reciprocal. Meat Con$ 29,37. Bokkenheuser, V. (1963). Hygienic evaluation of biltong. S.Afr. Med. J. 37, 619.
A REVIEW O F DEHYDRATION CONTROL IN IM MEAT PRODUCTS
153
Borosova, M. A., and Oreshkin, E. F. (1992). On the water condition in pork meat. Meat Sci. 31, 257. Botes, H. J. W. (1966). Biltong-induced Salmonella enteritidis var. typhimurium food poisoning-A case report. J. S. Afr. Vet. Med. Assoc. 37, 173. Bourne, M. C. (1975). Is rheology enough for food texture measurement? J. Text. Stud. 6,259. Bouton, P. E., and Harris, P. V. (1972). The effects of some post-slaughter treatments on the mechanical properties of bovine and ovine muscle. J. Food Sci. 37, 539. Bouton, P. E., Harris, P. V., and Shorthose, W. R. (1971). Effect of ultimate pH upon the water-holding capacity and tenderness of mutton. J. Food Sci. 36, 435. Bouton, P. E., Harris, P. V., and Shorthose, W. R. (1972). The effect of ultimate pH on bovine muscle: Mechanical properties. J . Food Sci. 37, 351. Briskey, E. J. (1964). The etiological status and associated studies of pale, soft and exudative porcine musculature. Adv. Food Res. 13, 89. Briskey, E. J., Bray, R. W., Hoekstra, W. G., Phillips, P. H., and Grummer, R. H. (1959). The effect of exhaustive exercise and high sucrose regimen on certain chemical and physical pork ham muscle characteristics. J. Anim. Sci. 18, 173. Buckle, K. A,, Purnomo, H., and Sastrodiantoro, S. (1988). Stability of dending. In “Food Preservation by Moisture Control” (C. C. Seow, ed.), pp. 137-148. Elsevier, London. Burrows, I. E., and Barker, D. (1976). Intermediate moisture petfoods. I n “Intermediate Moisture Foods” (R. Davies, G. G. Birch, and K. J. Parker, eds.), pp. 43-53. Applied Science, London. Campbell-Platt, G. (1995). Fermented meats-A world perspective. In “Fermented Meats” (G. Campbell-Platt and P. E. Cook, eds.), pp. 39-52. Blackie, London. Campbell-Platt, G., and Cook, P. E., eds. (1995). “Fermented Meats.” Blackie, London. Chang, S. F., and Pearson, A. M. (1992). Effect of electrical stunning or sticking without stunning on the microstructure of Zousoon, a Chinese semi-dry pork product. Meat Sci. 31, 309. Chang, S. F., Huang, T. C., and Pearson, A. M. (1991). Some parameters involved in production of Zousoon, a semi-dry, long fibered pork product. Meat Sci. 30,303. Cho. I. C., and Bratzler, L. J. (1970). Effect of sodium nitrite on flavor of cured pork. J. Food Sci. 35, 668. Chow, H. M., Liu, M. B., Hsu, P. C., Hong, W. S., and Lin, C. I. (1989). Effect of hurdle technology on the preservation of restructured dried pork. Anim. Znd. Res. Znst. TSC, Annu. Res. Rep., p. 227. Chuah, E. C., Yeoh, Q. L., and Hussin, A. B. H. (1988). Traditional Malaysian low and immediate moisture meat products. I n “Food Preservation by Moisture Control” (C. C. Seow, ed.), pp. 149-159. Elsevier, London. Cia, G., and Marsh, B. B. (1976). Properties of beef muscle cooked before rigor onset. J. Food Sci. 41, 1259. Cook, P. E. (1995). Fungal ripened meats and meat products. I n “Fermented Meats” (G. Campbell-Platt and P. E. Cook, eds.), pp. 110-129. Blackie, London. Corbin, J. E. (1992). Inedible meat, poultry and fish by-products in pet foods. Adv. Meat Res. 8, 329. Crapiste, G. H., Rotstein, E., and Urbicain, M. J. (1985). Drying of foods while the cellular structure prevails. Lat. Am. J. Chem. Eng. Appl. Chem. 15, 77. Cross, C. K., and Ziegler, P. (1965). A comparison of the volatile fractions from cured and uncured meat. J. Food Sci. 30, 610. Currie, R. W., and Wolfe, F. H. (1980). Rigor related change in mechanical properties and extracellular space in beef muscle. Meat Sci. 4,123.
154
S . F. CHANG et al.
Davies, R., Birch, G. G., and Parker, K. G., eds. (1976). “Intermediate Moisture Foods.” Applied Science, London. DeMasi, T. W., Wardlaw, F. B., Dick, R. L., and Acton, J. C. (1990). Nonprotein nitrogen and free amino acid contents of dry, fermented and nonfermented sausages. Meat Sci. 27, 1. Demeyer, D., Hoozee, J., and Moermans, R. (1979a). Specificity of lipolysis during dry sausage ripening. J. Food Sci. 39, 293. Demeyer, D., Vanderckove, P., and Moermans, R. (197913). Compounds determining pH in dry sausages. Meat Sci. 3, 161. Demeyer, D., Verplaetse, A., and Gistelinck, M. (1986). Fermentation of meat: An integrated approach. Proc. Eur. Meet. Meat Res. Workers 32,241. Dominguez-Fernandez, M. C., and Zumalacarreui-Rodriguez, J. M. (1991). Lipolytic and oxidative changes in “chorizo” during ripening. Meat Sci. 29, 99. Eichner, K., and Ciner-Doruk, M. (1981). Formation and decomposition of browning intermediates and visible sugar-amine browning reactions. In “Water Activity: Influences on Food Quality” (L. B. Rockland and G. F. Stewart, eds.), pp. 567-603. Academic Press, New York. Erbersdobler, H. F. (1986). Loss of nutritive value on drying. I n “Concentration and Drying of Foods” (D. MacCarthy, eds.), pp. 69-87. Elsevier, London. Flink, J. M. (1977a). Energy analysis in dehydration process. Food Technol. 31(3), 77. Flink, J. M. (1977b). A simplified cost comparison of a freeze-dried food with its canned and frozen counterparts. Food Technol. 31(4), 50. Forrest, J. C., Aberle, E. D., Hedrick, H. B., Judge, M. D., and Merkel, R. A. (1975). “Principles of Meat Science.” Freeman, San Francisco. Garcia de Fernando, G. D., and Fox, P. F. (1991). Study of proteolysis during the processing of a dry fermented pork sausage. Meat Sci. 30,367. Gee, M., Farkas, D., and Rahman, A. R. (1977). Some concepts for the development of intermediate moisture foods. Food Technol. 31(4), 58. Genigeorgis, C., and Lindroth, S. (1984). The safety of Basturma, and armenian-type dried beef with respect to salmonella. Proc. Eur. Meet. Meat Res. Workers, 30, 217. Gibbs, P. A. (1986). Microbiological quality of dried foods. In “Concentration and Drying of Foods” (D. MacCarthy, ed.), pp. 89-111. Elsevier, London. Could, G. W. (1989a). Heat induced injury and inactivation. I n “Mechanisms of Action of Food Preservation Procedures” (G. W. Could, ed.), pp. 11-42. Elsevier, London. Gould, G. W. (1989b). Drying, raised osmotic pressure and low water activity. In “Mechanisms of Action of Food Preservation Procedures” (G. W. Could, ed.), pp. 97-117. Elsevier, London . Could, G. W., and Christian, J. H. B. (1988). Characterization of the state of water in foods-Biological aspects. In “Food Preservation by Moisture Control” (C. C. Seow, ed.), pp. 43-56. Elsevier, London. Gould, G. W., and Jones, M. V. (1989). Combination and synergistic effects. In “Mechanisms of Action of Food Preservation Procedures” (G. W. Could, ed.), pp. 401-421. Elsevier, London. Could, G. W., and Measures, J. C. (1977). Water relations in single cells. Philos. Trans. R. SOC. London, Ser. B 278,151. Gray, J. I., and Pearson, A. M. (1984). Cured meat flavor. Adv. Food Res. 29, 1. Greaser, M. L. (1986). Conversion of muscle to meat. In “Muscle as Food” (P. J. Bechtel, ed.), pp. 37-102. Academic Press, Orlando, FL. Griffith, T., and Johnson, J. A. (1957). Relation of the browning reaction to storage stability of sugar cookies. Cereal Chern. 34, 159.
A REVIEW O F DEHYDRATION CONTROL IN IM MEAT PRODUCTS
155
Guilbert, S. (1988). Use of superficial edible layer to protect intermediate moisture foods: Application to the protection of tropical fruits dehydrated by osmosis. In “Food Preservation by Moisture Control” (C. C. Seow, ed.), pp. 199-219. Elsevier, London. Hamm, R. (1960). Biochemistry of meat hydration. Adv. Food Res. 10,355. Hamm, R. (1986). Functional properties of the myofibrillar system and their measurements. In “Muscle as Food” (P. J. Bechtel, ed.), pp. 135-199. Academic Press, Orlando, FL. Hausbrand, E. (1901). “Drying by Means of Air and Steam.” Greenwood & Sons, London. Henrickson, R. L., and Asghar, A. (1985). Cold storage energy aspects of electrically stimulated hot-boned meat. Adv. Meat Res. 1, 237. Hirst, E. (1974). Energy for food: From farm to home. Trans. ASAE 17,323. Ho, H. F., and Koh, B. L. (1984). Processing some Chinese meat products in Singapore. Proc. SIFST Symp. Adv. Food Process., 4th, Singapore, 1984, p. 94. Hocking, A. D. (1988). Moulds and yeasts associated with foods of reduced water activity: Ecological interactions. In “Food Preservation by Moisture Control” (C. C. Seow, ed.), pp. 57-72. Elsevier, London. Hodge, J. E. (1953). Chemistry of browning reactions in model systems. J. Agric. Food Chern. 1, 928. Hodge, J. E. (1967). Origin of flavor in foods: Nonenzymatic browning reactions. In “The Chemistry and Physiology of Flavors” (H. W. Schultz, ed.), pp. 465-491. Avi Publ. Co., Westport, CT. Hsieh, Y. P. C., Pearson, A. M., Morton, I. D., and Magee, W. T. (1980a). Some changes in the constituents upon heating a model meat flavour system. J. Sci. Food Agric. 31, 943. Hsieh. Y. P. C., Cornforth, D. P., and Pearson, A. M. (1980b). Ultrastructural changes in preand post-rigor beef muscle caused by conventional and microwave cookery. Meat Sci. 4, 299. Huang,T. C., Chang, S. F., Lin, C. S., Shieh, D. E. C., and Ho, C. T. (1989). Aroma development in Chinese fried pork bundle. ACS Symp. Ser. 409,487. Ibanez, C., Quintanilla, L., Irigoyen, A., Garcia-Julon, I., Cid, C., Astiasarian, I., and Belo, J. (1995). Partial replacement of sodium chloride with potassium chloride in dry fermented sausages: Influence on carbohydrate fermentation and the nitrosation process. Meat Sci. 40,45. Igene, J. 0..Farouk, M. M., and Akanbi, C. T. (1992). Preliminary studies on the traditional processing of Kilishi. J. Sci. Food Agric. 50, 89. Incze, K. (1991). Raw fermented and dried meat products. Inr. Congr. Meat Sci. Techno/., Kulmbach, Germany, 1992, Vol. 2, p. 829. Incze, K. (1992). Raw fermented and dried meat products. Fleischwirtschaft 72,62. Jason, A. C . (1958). A study of evaporation and diffusion processes in the drying of fish muscle. In “Fundamental Aspects of Dehydration of Foodstuffs,” pp. 103-135. Society of Chemical Industry, London. Jessen, B. (1995). Starter cultures for meat fermentations. In “Fermented Meats” (G. Campbell-Platt and P. E. Cook, eds.), pp. 130-159. Blackie, London. Jones, S. B. (1977). Ultrastructure characteristics of beef muscle. Food Technol. 31(4), 82. Judge, M. D., Okos, M. R., Baker, T. G., Potthast, K., and Hamm, R. (1981). Energy requirements and processing costs for freeze-dehydration of prerigor meat. Food Technol. 35(4), 61. Kapsalis, J. G. (1975). The influence of water on textural parameters in foods at intermediate moisture levels. In “Water Relations of Foods” (R. B. Duckworth, ed.), pp. 627-653. Academic Press, New York. Karel, M. (1976). Technology and application of new intermediate moisture foods. In “Intermediate Moisture Foods” (R. Davies, G. G. Birch, and K. J. Parker, eds.), pp. 4-31. Applied Science, London.
156
S . F. CHANG et al.
Karel, M. (1986). Control of lipid oxidation in dried foods. In “Concentration and Drying of Foods” (D. MacCarthy, ed.), pp. 37-68. Elsevier, London. Karel, M. (1988). Optimising the retention of heat sensitive materials in concentration and drying. In “Preconcentration and Drying Food Materials” (S. Bruin, ed.), p. 217. Elsevier, Amsterdam. Karel. M., Labuza, T. P., and Maloney, J. F. (1967). Changes in freeze-dried foods and model systems. Cryobiology 3, 1288. Karel, M., Schaich, K., and Roy, R. B. (1975). Interaction of peroxidizing methyl linoleate with some proteins and amino acids. J. Agric. Food Chem. 23, 159. Katsaras, K., and Budras, K. D. (1992). Microstructure of fermented sausage. Meat Sci. 31,121. Keey, R. B. (1972). “Drying: Principles and Practice.” Pergamon, Oxford. Kemp, J. D., Langlois, B. E.. and Fox, J. D. (1983). Effect of potassium sorbate and vacuum packaging on the quality and microflora of dry-cured, intact boneless hams. J. Food Sci. 48,1709. King, C. J. (1988). Spray drying of food liquid and volatiles retention. In “Preconcentration and Drying of Food Materials” (S. Bruin, ed.), p. 147. Elsevier, Amsterdam. Krause, P., Schmoldt, R., Tolgay, Z., and Yurtyeri, A. (1972). Mikrobiologische und serologische Unterschungen an Lebensmitteln in der Tiirkei. Fleischwirtschaft 52, 83. Krischers, 0. (1963). “Die Wissenschaftlichen Grundlagen der Trocknungstechnik,” 2nd ed. Springer, Berlin. Krockel, L. (1995). Bacterial fermentation of meat. In “Fermented Meats” (G. CampbellPlatt and P. E. Cook, ed.), pp. 69-109. Blackie, London. Kuo, J. C., and Ockerman, H. W. (1985). Effect of salt, sugar and storage time on microbiological, chemical and sensory properties of Chinese style dried pork. J. Food Sci. 50,1384. Kuprianoff, J. (1958). Bound water in foods. In “Fundamental Aspects of Dehydration of Foodstuffs,” pp. 14-23. Society of Chemical Industry, London. Labuza, T. P. (1976). Drying food: Technology improves on the sun. Food Technol. 30(6), 37. Labuza, T. P. (1980). The effect of water activity on reaction kinetics of food deterioration. Food Technol. 34(4), 36. Lawrie, R. A. (1979). “Meat Science.” Pergamon, Oxford. Lawrie, R. A. (1995). The structure, composition and preservation of meat. In “Fermented Meats” (G. Campbell-Platt and P. E. Cook, eds.), pp. 1-38. Blackie, London. Ledl, F. (1987). Analysis of volatile sugar degradation products. Lebensmittelchem. Gerichtl. Chem. 41(4), 83. Ledward, D. A. (1981). Intermediate moisture meat products. In “Developments in Meat Science” (R. Lawrie, ed.), Vol. 2, p. 157. Applied Science, London. Ledward, D. A. (1985). Novel intermediate moisture meat products. In “Properties of Water in Foods” (D. Simatos and J. L. Multon, eds.), pp. 447-463. Martinus Nijhoff, Dordrecht, The Hague, The Netherlands. Leistner, L. (1978). Hurdle effect and energy saving. In “Food Quality and Nutrition” (W. K. Downey, ed.), pp. 553-562. Applied Science, London. Leistner, L. (1987). Shelf stable products and intermediate moisture foods based on meat. In “Water Activity: Theory and Applications to Foods” (L. B. Rockland and L. R. Beuchat, eds.), pp. 295-327. Dekker, New York. Leistner, L. (1990a). The variety of methods in raw ripened sausage manufacture. Fleischerei 41, 570. Leistner, L. (1990b). Mould-fermented foods: Recent developments. Food Biotechnol. 4,433. Leistner, L. (1995). Stable and safe fermented sausages worldwide. In “Fermented Meats” (G. Campbell-Platt and P. E. Cook, eds.), pp. 160-175. Blackie, London.
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
157
Leistner, L., and Rodel. W. (1976). The stability of intermediate moisture foods. In “Intermediate Moisture Foods” (R. Davies, G. G. Birch, and K. J. Parker, eds.), pp. 120-145. Applied Science, London. Leistner, L., Rodel, W., and Krispien, K. (1981). Microbiology of meat and meat products in high and intermediate moisture ranges. I n “Water Activity: Influences on Food Quality” (R. B. Rockland and G. F. Stewart, eds.), pp. 855-916. Academic Press, New York. Leistner, L., Geisen, R., and Fink-Gremmels, J. (1989). Mould-fermented foods of Europe: Hazards and developments. Bioact. Mol. 10, 145-164. Lerici, C. R., Mastrocola. D., Sensidoni, A,, and Dalla Rosa, M. (1988). Osmotic concentration in food processing. In “Preconcentration and Drying of Food Materials” (S. Bruin, ed.), p. 123. Elsevier, Amsterdam. Lin. I. M., Chung, Y. J., Tsai, R. T., and Lee, T. C. (1981). Studies on nutritional value of Chinese dried seasoned fried pork. Recent Adv. Food Sci. Technol. 2,295. Lin. S . L. (1981). New process for manufacturing dried pork slices (I). Food Ind. Res. Dev. Inst., Res. Rep. (Taiwan),No. 212. Lin, S. L.. Tsai, S. F.. Chen. C. R., and Li, C. F. (1983a). New process for manufacturing dried pork slices (11). Food Ind. Res. Dev. Insf.. Res. Rep. (Taiwan), No. 280. Lin, S. L., Huang, C. S.. Chang, P. Y., and Li, C. F. (1983b). New process for manufacturing dried pork slices (111). Food Ind. Rex Dev. Inst., Res. Rep. (Taiwan) No. 287. Linko, P., Pollari. T., Harju. M., and Heikonen, M. (1981). Water sorption properties and the effect of moisture on the structure of dried milk products. Lebensm. Wiss. Technol. 15(1), 26. Lo. C. X. (1980). “Processing Foods of Animal Origin. Textbook for Agricultural Universities of China.” pp. 157-165. Eastern-Northern Agriculture University, Peking (in Chinese). Locker, R. H., and Carse, W. A. (1976). Extensibility, strength and tenderness of beef cooked to various degrees. J. Sci. Food Agric. 27, 891. Lois, A. L., Guitierrez, L. M., Zumalacarregui, J. M., and Lopez, A. (1987). Changes in several constituents during ripening of “chorizo.” A Spanish dry sausage. Meat Sci. 19, 169. Loncin, M. (1988). Activity of water and its importance in preconcentration and drying of food. In “Preconcentration and Drying in Food Materials” (S. Bruin, ed.), p. 15. Elsevier, Amsterdam. Lopez, M. 0..de la Hoz, L., Cambero, M. I., Gallardo, E., Reglero, G., and Ordonez, J. A. (1992). Volatile compounds of dry hams from Iberian pigs. Meat Sci. 31, 267. Lwikov, A. V. (1966). “Heat and Mass Transfer in Capillary-Porous Bodies.” Pergamon, Oxford. MacLeod, G. (1986). The scientific and technological basis of meat flavors. In “Developments in Food Flavors” (G. G. Birch and M. G. Lindley, eds.), p. 191. Elsevier, London. Marsh, B. B., Leet, N. G.. and Dickson, M. R. (1974). The ultrastructure and tenderness of highly cold-shortened muscle. J. Food Technol. 9, 141. McLoughlin, J. V. (1971). Stunning, the death reaction and metabolism postmortem of porcine skeletal muscle. Proc. Int. Symp. Condition Meat Qua1 Pigs, 2nd, 197I, pp. 123-137. Mishkin, M., Karel, M., and Saguy, I. (1982). Application of optimization in food dehydration. Food Technol. 36(7), 101. Miteva, E., Kirova, E., Gadjeva, D., and Radeva, M. (1986). Sensory aroma and taste profiles of raw-dried sausages manufactured with a lipolytically active yeast culture. Nahrung 30,829. Mittal, G. S., and Blaisdell, J. L. (1983). Weight loss in frankfurters during thermal processing. Meat Sci. 9, 79. Mortimore, S., and Wallace, C. (1994). “HACCP-A Practical Approach.” Chapman & Hall, London.
158
S . F. CHANG et al.
Muchtadi, D. (1986). “Studies on the Nutritional Value of Dending-An Indonesian Traditional Preserved Meat Product.” Food Technology Development Centre, Bogor, Indonesia. Muguruma, M., Katayama, K., Nakamura, M., and Yamaguchi, M. (1987). Low-temperature osmotic dehydration improves the quality of intermediate moisture meats. Meat Sci. 21,99. Norback, J. P. (1980). Techniques for optimization of food process. Food Techno[. 34(2), 86. Nursten, H. E. (1986a). Aroma compounds from the Maillard reaction. In “Developments in Food Flavors” (G. G. Birch and M. G. Lindley, eds.), p. 173. Elsevier, London. Nursten, H. E. (1986b). Maillard browning reactions in dried foods. In “Concentration and Drying of Foods” (D. MacCarthy, ed.), pp. 53-68. Elsevier, London. Obanu, Z . A. (1988). Preservation of meat in Africa by control of the internal aqueous environment in relation to product quality and stability. In “Food Preservation by Moisture Control” (C. C. Seow, ed.), pp. 161-173. Elsevier, London. Obanu, Z . A., Ledward, D. A., and Lawrie, R. A. (1975a). Storage of intermediate moisture meat at tropical temperatures. I. Changes in protein solubility and electrophoretic pattern. J. Food Technol. 10, 651. Obanu, Z . A., Ledward, D. A., and Lawrie, R. A. (1975b). Storage of intermediate moisture meat at tropical temperatures. 11. Effect on eating quality. J. Food Technol. 10, 667. Obanu, Z . A., Ledward, D. A., and Lawrie, R. A. (1976). Storage of intermediate moisture meat products at tropical temperatures. 111. Differences between muscles. J. Food Technol. 11, 187. Obanu, Z . A., Ledward, D. A., and Lawrie, R. A. (1977). Reactivity of glycerol in intermediate moisture meats. Meat Sci. 1, 177. Ockerman, H. W., and Kuo, J. C. (1982). Dried pork as influenced by nitrate, packaging method and storage. J. Food Sci. 47, 1631. Offer, G . , and Knight, P. (1988a). The structural basis of water-holding in meat. Part 1: General principles and water uptake in meat processing. In “Developments in Meat Science-4’’ (R. Lawrie, ed.), pp. 63-171. Elsevier, London. Offer, G., and Knight, P. (1988b). The structural basis of water-holding in meat. Part 2. Drip losses. In “Developments in Meat Science-4” (R. Lawrie, ed.), pp. 173-243. Elsevier, London. Offer, G., and Trinick, J. (1983). On the mechanism of water holding in meat: The swelling and shrinking of the myobrils. Meat Sci. 8, 245. Offer, G., Elsey, J., Parsons, N., Cousins, T., and Knight, P. (1988). In “Basic Interactions of Water and Foods” (M. Le Maguer and W. Powrie, eds.), pp. 112-130. Dekker, New York. Ohye, D. F., and Christian, J. H. B. (1967). Combined effects of temperature, pH and water activity on growth and toxin production by Clostridium types A, B, and E. Botulism 1966, Proc. In?. Symp. Food Microbiol. 5th, 1966, p. 217. Okonkwo, T. M. (1984). Studies on the commercial production, consumption and quality of hot-smoked meat in Nigeria. MSc. Thesis, University of Nigeria, Nsykka. Okonkwo, T. M., Obanu, Z . A., and Ledward, D. A. (1992a). Characteristics of some intermediate moisture smoked meats. Meat Sci. 31, 135, Okonkwo, T. M., Obanu, Z. A., and Ledward, D. A. (1992b). The stability of some intermediate moisture smoked meats during storage at 30°C and 38°C. Meat Sci. 31,246. Pardi, M. C. (1961). Aelaboracao do charqui do Brasil: Da conveniencia de novos rumos, p. 44. Tese de mestrado, Faculdade Fluminense de Medicina Veterinaria, Rio de Janeiro. Pearson, A. M., and Gillett, T. A. (1995). “Processed Meats,” 3rd ed. Chapman & Hall, New York. Pearson, A. M., and Miller, J. I. (1950). The influence of rate of freezing and length of freezerstorage upon the quality of beef of known origin. J. Anim. Sci. 9, 13.
A REVIEW OF DEHYDRATION CONTROL IN IM MEAT PRODUCTS
159
Pearson, A. M., and Tauber, F. W. (1984). “Processed Meats,” 2nd ed. Avi Publ. Co., Westport, CT. Pearson, A. M., and Young, R. B. (1989). “Muscle and Meat Biochemistry.” Academic Press, San Diego. CA. Pearson, A. M., Harrington, G., West, R. G., and Spooner, M. E. (1962). The browning produced by heating fresh pork. I. The relation of browning intensity to chemical constituents and pH. J. Food Sci. 27, 177. Pearson, A. M., Tarladgis, B. G., Spooner, M. E., and Quinn, J. R. (1966). The browning produced on heating fresh pork. 11. The nature of the reaction. J . Food Sci. 31, 184. Pearson, A. M.. Gray, J. I., Wolzak, A. M., and Horenstein, N. A. (1983). Safety implications of oxidized lipids in muscle foods. Food Technol. 37(7), 121. Pestka. J. (1995). Fungal toxins in raw and fermented meats. In “Fermented Meats” (G. Campbell-Platt and P. E. Cook, eds.), pp. 194-216. Blackie, London. Ponting, J. D., Watters, G. G . , Forrey, R. R., Jackson, R., and Stanley, W. L. (1966). Osmotic dehydration of fruits. Food Technol. 20(10), 125. Potter, N. N. (1986). “Food Science.” Avi Publ. Co., Westport, CT. Prior, B. A,, and Badenhorst, L. (1974). Incidence of salmonellae in some meat products. S. Afr. Med. J. 48, 2532. Purslow, P. P. (1987). The fracture behavior of meat-A case study. In “Food Structure and Behavior” (J. M. V. Blanshard and P. Lillford, eds.), p. 178. Academic Press, New York. Reyes-Cano, R.. Dorantes-Alvarez, L., Hernandez-Sanchez, H., and Guitierrez-Lopez, G. F. (1994). A traditional intermediate moisture meat: Beef cecina. Meat Sci. 36, 365. Reyes-Cano, R., Dorantes-Alvarez, L.. Hernandez-Sanchez, H., and Guitierrez-Lopez, G. F. (1995). Biochemical changes in an intermediate moisture cecina-like meat during storage. Meat Sci. 40,387. Rico, E., Toldra, F., and Flores, J. (1991). Assay of Cathepsin D activity in fresh pork muscle and dry-cured ham. Meat Sci. 29,287. Rockland, L. B., and Nishi, S. K. (1980). Influence of water activity on food product quality and stability. Food Technol. 34(4), 42. Romans, J. R., Costello. W. J., Jones, K. W., and Carlson, C. W. (1985). “The Meat We Eat,” 12th ed. Interstate Printers and Publishers, Danville, IL. Roos, Y. R. (1995). Glass transition-related physicochemical changes in foods. Food Techno/. 49(10), 97. Roos, Y., and Karel, M. (1991a). Plasticizing effect of water on thermal behavior and crystallization of amorphous food models. J. Food Sci. 56,38. Roos, Y., and Karel, M. (1991b). Applying state diagram to food processing and development. Food Technol. 45(12), 66. Roos, Y . . and Karel, M. (1992). Crystallization of amorphous lactose. J. Food Sci. 57,775. Rossello, C., Barbas, J. I., Berna, A., and Lopez, N. (1995). Microbial and chemical changes in sobrasada during ripening. Meat Sci. 40,379. Rowe, R. W. D. (1989a). Electron microscopy of bovine muscle: I. The native state of postrigor sarcolemma and endomysium. Meat Sci. 26,271. Rowe, R. W. D. (1989b). Electron microscopy of bovine muscle: 11. The effects of heat denaturation on post rigor sarcolemma and endomysium. Meat Sci. 26,281. Rust, R. E. (1988). Production of edible casings. A d v . Meat Res. 5,261. Saguy, I.. and Karel, M. (1980). Modeling of quality deterioration during food processing and storage. Food Technol. 34(2), 79. Sanderson, A., Pearson, A. M., and Schweigert, B. S. (1966). Effect of cooking procedure on flavor components of beef, carbonyl compounds. J. Agric. Food Chem. 14,245.
160
S . F. CHANG et at
Schmidt, G. R., Mawson, R. F., and Siegel, D. G. (1981). Functionality of a protein matrix in comminuted meat products. Food Technol. 35(5), 235. Schweigert, B. S. (1987). The nutritional content and value of meat and meat products. In “The Science of Meat and Meat Products” (J. F. Price and B. S. Schweigert, eds.), 3rd ed., pp. 275-305. Food & Nutrition Press, Westport, CT. Schwimmer, S. (1980). Influence of water activity on enzyme reactivity and stability. Food Technol. 34(5), 64. Scott, W. J. (1953). Water relations of staphylococcus aureus at 30°C. Austr. J. Biol. Sci. 6,549. Scott, W. J. (1957). Water relations of food spoilage organisms. Adv. Food Res. 7,83. Sebranek, J. G. (1988). “Meat Science and Processing.” Paladin House, Lake Geneva, WI. Serrano-Moreno, D. A. (1979). Evolucionde varias microfloras y su interdependencia con las condiciones fisico-quimicas durante la maduacion del salchichon. Alimentaria 100, 39. Shin, H. K. (1984). Energiesparende Konservierungsmethoden fur Fleischerzeugnisse, abgeleitet von traditionellen Intermediate Moisture Meats. Ph.D. Thesis, Universittit Hohenheim, Stuttgart-Hohenheim, West Germany. Shin, H. K., and Leistner, L. (1983). “Mikrobiologische Stabilitst traditioneller IM-Meats, importiert aus Afrika und Asien.” Jahresber. Bundesanst. Fleischforsch., Kulmbach, C21, Germany. Simatos, D., and Karel, M. (1988). Characterization of the condition of water in foods. Physicochemical aspects. In “Food Preservation by Moisture Control” (C. C. Seow, ed.)., pp. 1-41. Elsevier, London. Slade, L., and Levine, H. (1987). Polymer-chemical properties of gelatin in foods. Adv. Meat Res. 4, 251. Slade, L., and Levine, H. (1991). Beyond water activity: Recent advances based on an alternative approach to the assessment of food quality andsafety. Crit. Rev. FoodSci Nuir. 30,115. Stanley, D. W. (1983). Relation of structure to physical properties of animal materials. In “Physical Properties of Foods” (M. Peleg and E. B. Bagly, eds.), p. 157. Avi Publ. Co., Westport, CT. Stefansson, V. (1956). The Fat of the Land. MacMillan Co., New York. Stiebing, A,, and Rodel, W. (1989). Continuous measurement of the surface water activity of raw ripened sausage. Mitt. Bundesanst. Fleischforsch. 104, 221. Stiebing, A., Rodel, W., and Kletterner, P. G. (1982). Energy savings during raw sausage manufacturing. Fleischwirtschaft 62, 1383. Thijssen, H. A. C., and Kerkhof, P. J. A. M. (1977). Effect of temperature-moisture content history during processing on food quality. In “Physical, Chemical and Biological Changes in Food Caused by Thermal Processing” (T. Hoyem and 0. Kvale, eds.), pp. 10-30. Applied Science, London. Toldra, F., and Etherington, D. J. (1988). Examination of cathepsins B, D, H and L activities in dry-cured hams. Meat Sci. 23, 1. Torres, E., Pearson, A. M., Gray, J. I., Ku, P. K., and Shimokomaki, M. (1989). Lipid oxidation in charqui (salted and dried beef). Food Chem. 32,257. Troller, J. A. (1972). Effect of water activity on enterotoxin A production and growth of Staphylococcus aureus. Appl. Microbiot. 24,440. Troller, J. A. (1980). Influence of water activity on microorganisms in foods. Food Technol. 34(5), 76. Troller, J. A. (1987). Adaptation and growth of microorganisms in environments with reduced water activity. In “Water Activity: Theory and Applications to Foods” (L. B. Rockland and L. R. Beuchat, eds.), pp. 101-117. Dekker, New York. Troller, J. A., and Christian, J. H. B. (1978). “Water Activity and Food.” Academic Press, New York.
A REVIEW O F DEHYDRATION CONTROL IN IM MEAT PRODUCTS
161
Van Arsdel, W. 8. (1963). “Food Dehydration,“ Vol. 1. Chapter 5. Avi Publ. Co., Westport, CT. Van den Heever, L. W. (1965). The viability of salmonellae and bovine cysticerci in biltong. S. Afr. Med. J. 39, 14. Van den Heever, L. W. (1970). Some public health aspects of biltong. J. S. Afr. Vet. Med. Assoc. 41, 263. Van der Riet, W. B. (1976a). Studies on the microflora of biltong. S. Afr. Food Rev. 3(1), 105. Van der Riet, W. B. (1976b). Water sorption isotherms of beef biltong and their use in predicting critical moisture contents for biltong storage. S. Afr. Food Rev. 3(6), 93. Van der Riet, W. B. (1982). Biltong ein siidafrikanisches Trockenfleischprodukt. Fleischwirtschaft 62, 970. Van der Wal, P. G . (1971). Stunning procedures for pigs and their physiological consequences. Proc. Int. Symp. Condition Meat Qual. Pigs, 2nd, 1971. Verplaetse, A. (1994). Influence of raw meat properties and processing technology on aroma quality of raw fermented meat products. Int. Congr. Meat Sci. Technol. 40, 45 (Main papers). Voyle, C. A. (1969). Some observations of cold-shortened muscle. J. Food Technol. 4,275. Voyle, C. A. (1981). Scanning electron microscopy in meat science. Scanning Electron Microsc. 3, 405. Vuataz, G . (1988). Preservation of skim milk powders: Role of water activity and temperature in lactose crystallization and lysine loss. In “Food Preservation by Moisture Control” (C. C. Seow, ed.), pp. 73-101. Elsevier, London. Wang, C. T., and Chen, Y. S. (1989). Studies on using pre- and post rigor ground pork to manufacture dried shredded, fried shredded and dried sliced pork. Taiwan Livestock Res. 22, 1. Wang, H., Andrews, F., Rasch, E., Doty, D. M.. and Kraybill, H. R. (1953). A histological and histochemical study of beef dehydration. Food Res. 18, 351. Watson, E. L.. and Harper, J. C. (1987). “Elements of Food Engineering.” Van NostrandReinhold, New York. Webster, C. E. M., Nuilez-Gonzalez, F. A., and Ledward, D. A. (1982). The role of lipids and glycerol in determining the shelf-life of glycerol desorbed intermediate moisture meat products. Meat Sci. 6, 181. Webster, C. E. M., Allison, S. E., Adelakum, I. O., Obanu, Z. A., and Ledward, D. A. (1986). Reactivity of sorbate and glycerol in intermediate moisture meat products. Food Chem. 21, 133. Wientjes, A. G. (1968). The influences of sugar concentration on the vapor pressure of food odor volatiles in aqueous solution. J. Food Sci. 33, 1. Wismer-Pedersen, J. (1960). Quality of pork in relation to rate of pH change post-mortem. Food Res. 25, 789. Wismer-Pedersen, J. (1971). Water I n “The Science of Meat and Meat Products” (J. F. Price and B. S. Schweigert, eds.), pp. 177-191. Freeman, San Francisco. Yen, G . C., Tsai, R. Y. T., and Lee, T. C. (1981). Studies on Maillard browning in Chinese shredded fried pork. Natl. Sci. Counc. Mont. Republic of China 9(3), 232. Zapata, J. F. F., Ledward, D. A., and Lawrie, R. A. (1990). Preparation and storage stability of dried salted mutton. Meat Sci. 27, 109. Zeuthen, P. (1995). Historical aspects of meat fermentations. In “Fermented Meats” ( G . Campbell-Platt and P. E. Cook, eds.), pp. 53-68. Blackie, London.
This Page Intentionally Left Blank
ADVANCES IN FOOD AND NUTRITION RESEARCH, VOL. 39
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS P. F. FOX, T. P. O’CONNOR, AND P. L. H. MCSWEENEY Department of Food Chemistry University College Cork, Ireland
T. P. GUINEE National Dairy Products Research Centre Teagasc, Moorepark Ferrnoy, Co. Cork, Ireland
N. M. O’BRIEN Department of Nutrition University College Cork. Ireland
I. Introduction A. Historical B. Cheese Production and Consumption C. Cheese Science and Technology 11. Composition and Constituents of Milk 111. Conversion of Milk to Cheese Curd A. Rennet Coagulation of Milk B. Fresh Acid-curd Cheese Varieties C. Ultrafiltration Technology in Cheesemaking IV. Biochemistry of Cheese Ripening A. Cheese Ripening Agents: Assessment of Contribution to Ripening B. Metabolism of Lactose and Lactate during Ripening C. Citrate Metabolism D. Lipolysis E. Proteolysis 163 Copyright 0 19% by Academic Press, Inc. All rights of reproduction in any form reserved.
164
P. F. FOX ei al. V. Cheese Flavor A. Introduction B. Analytical Methods C. Inter- and Intravarietal Comparison of Cheese Ripening D. Factors that Affect Cheese Quality VI. Cheese Texture VII. Accelerated Cheese Ripening A. Elevated Temperatures B. Exogenous Enzymes C. Modified Starters D. Cheese Slurries VIII. Processed Cheese Products A. Introduction B. Classification of Processed Cheese Products C. Manufacturing Protocol D. Principles of Manufacture of Processed Cheese E. Structure Formation on Cooling F. Properties of Emulsifying Salts G. Influence of Various Parameters on the Textural Properties of Processed Cheese Products IX. Nutritional and Safety Aspects of Cheese A. Introduction B. Protein C. Carbohydrate D. Fat and Cholesterol E. Vitamins F. Minerals G. Nisin and Other Additives in Cheese H. Cheese and Dental Caries I. Mycotoxins J. Biogenic Amines in Cheese X. Perspective References
I. INTRODUCTION A. HISTORICAL
Cheese is the generic name for a diverse group of fermented milk-based foods produced in at least 500 varieties throughout the world. Cheese evolved in the “Fertile Crescent” between the Tigris and Euphratres rivers, in Iraq, some 8000 years ago. From a very early stage in the Agricultural Revolution, Man consumed milk from domesticated animals but since milk is a rich source of nutrients for contaminating bacteria, it has a short shelf life, especially in warm climates. Certain bacteria (lactic acid bacteria, LAB) ferment milk sugar, lactose, as a source of energy, producing lactic acid as a by-product. When sufficient
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS 165
acid is produced, the principal proteins of milk, the caseins, coagulate, i.e., at their isoelectric points (-pH 4.6), to form a gel, entrapping the fat. Acidcoagulated (fermented) milks are produced throughout the world and are increasing in popularity. The ability of LAB to produce just enough acid to coagulate the caseins is quite fortuitous: neither the LAB nor the caseins were designed for this function. The caseins were “designed” to be enzymatically coagulated in the stomachs of neonatal mammals at ca. pH 6. The ability of LAB to ferment lactose, a sugar specific to milk, is encoded on plasmids, suggesting that this characteristic was acquired relatively recently. Their natural habitats are vegetation, from which they colonized the intestinal tract and the teats of mammals, contaminated with milk; through evolutionary pressure, these bacteria probably acquired the ability to ferment lactose. When an acid-coagulated milk gel is broken, the pieces synerese, expressing whey. Acid-coagulated casein is the starting material for a family of cheeses that are usually consumed fresh, i.e., are not matured (ripened); major examples are Cottage and cream cheeses, Quarg, and fromage frais. An alternative mechanism for milk coagulation exploits the ability of many proteinases, referred to as rennets, to modify the milk protein system, leading to coagulation. Rennets from bacteria, moulds, and plant and animal tissues may be used, but the principal traditional source was stomachs of neonatal mammals, the principal proteinase in which is chymosin. The properties of rennet-coagulated curds are very different from those of acidcoagulated curds, e.g., they exhibit better syneretic properties which make it possible to produce low-moisture cheese curd without hardening, thus permitting the production of more stable products than is possible from acid-coagulated curds. Therefore, rennet coagulation is exploited in the manufacture of most cheese varieties, which are normally ripened. Cheese manufacture accompanied the spread of civilization throughout the Middle East, Egypt, and Greece and was well established in the Roman Empire. Cheese making remained a localized, essentially farmstead, enterprise until the mid-19th century. Due to particular local circumstances, certain varieties of cheese evolved in specific regions and remained localized due to limited communications. Hence, several hundred varieties of cheese evolved, most of which are still produced locally although the principal varieties, Cheddar, Dutch, Swiss, Camembert, and fromage frais types, are now produced internationally. Information on the history of cheese can be found in Davis (1965), R. Scott (1986), and Fox (1993b). B. CHEESE PRODUCTION AND CONSUMPTION World production of cheese in 1990 was -14 X lo6 tonnes [Food and Agriculture Organization (FAO), 19941 and is increasing at an average
166
P. F. FOX et al.
annual rate of -4%. Europe, with a production of -6 X lo6 tonnes per annum, is the largest producing region (Table I). Cheese consumption varies widely, as shown in Table I1 for countries for which up-to-date information is available. Some form of cheese is produced throughout the world and some interesting minor varieties are produced in “nondairying” countries. C. CHEESE SCIENCE AND TECHNOLOGY Cheese is the most diverse, most scientifically interesting, and most challenging group of dairy products. While most dairy products, if properly manufactured and stored, are biologically, biochemically, and chemically very stable, cheeses are biologically and biochemically dynamic and, consequently, inherently unstable. Cheese manufacture and ripening involves a complex series of consecutive and concomitant microbiological, biochemical, and chemical events which, if synchronized and balanced, lead to products with highly desirable flavors but when unbalanced, result in off-flavors. Considering that a basically similar raw material (milks from a very limited number of species) is subjected to a generally common manufacturing protocol, it is fascinating that such a diverse range of products can be produced. Acid-coagulated cheeses are consumed fresh but the production of rennet-coagulated cheese can be subdivided into two phases, manufacture and ripening: Manufacture Milk
Ripening Fresh cheese curd
Mature cheese
The manufacturing phase comprises those operations performed during the first 24 hr and although the protocols for the various varieties differ in detail, the basic steps are common to most varieties, i.e., (1) acidification, (2) coagulation, (3) dehydration (cutting the coagulum, cooking, stirring, pressing, salting, and other operations that promote gel syneresis), (4) shaping (kneading, moulding, pressing) and ( 5 ) salting. Cheese manufacture is essentially a dehydration process in which the fat and casein of milk are concentrated 6- to 12-fold,depending on the variety. The degree of hydration is regulated by the extent and combination of the above operations and, in addition, by the chemical composition of the milk. In turn, the levels of moisture and salt, the pH, and the cheese microflora regulate and control the biochemical changes that occur during ripening and hence determine the taste, aroma, and texture of the finished product. Thus, the nature and quality of the finished cheese are determined largely
TABLE I CHEESE PRODUCTION IN THE LEADING COUNTRIES IN
Regions
1993" Cheese production (tomes xi03)
Europe France Germany Italy Netherlands United Kingdom Denmark Poland Russia Greece Spain Switzerland Czech Republic Sweden Hungary Irish Republic Austria Finland Norwayh BelgiudLuxembourg Portugal Slovakia Iceland
1,528.7 1,282.9 885.1 632.3 324.4 323.3 293.3 280.0 211.6 147.0 134.6 115.5 115.0 96.8 92.3 90.6 89.0 77.7 69.0 57.5 34.5 2.1
Americas U.S.A. Argentina Canada Mexico Brad
3,268.5 315.0 291.1 116.4 60.2
Africa Egypt Sudan South Africa Uruguay Zimbabwe
325.0 70.0 38.0 19.8 4.2
Others Australia China New Zealand Japan Israel Total
210.7 156.1 145.5 100.4 84.9 12,089
Food and Agriculture Organization (FAO) (1994). The International Dairy Federation (IDF) (1988).
* From
168
P. F. FOX er al.
TABLE I1 CONSUMPTION OF CHEESE (KG PER CAPUT) ~~
Country France Greece Italy Belgium Federal Republic of Germany Iceland Luxembourg Switzerland Sweden Denmark Netherlands Finland Canada Norway USA Austria Czech and Slovak Republics Australia United Kingdom New Zealand Russia Hungary Spain Ireland South Africa Japan India @
1991a
~
Hard and processed cheeses 15.3 21.8 14.5 13.7 9.7 10.3 11.3 13.5 14.8 14.0 13.6 11.4 12.7 12.9 11.4 6.2 6.6 7.4 4.9 3.4
Fresh cheese
Total
7.5 0.8 5.5 4.3 7.8 6.0 5.0 2.8 0.9 1.3 1.6 2.4 1.1 0.2 1.5 5.3 4.0
22.8 22.6 20.0 18.0 17.5 16.3 16.3 16.3 15.7 15.3 15.2 13.8 13.8 13.1 12.9 11.5 10.6 8.8 8.2 8.0 7.7 7.6 7.0 5.6 1.8 1.24 0.2
-
-
0.8 2.8 4.2 -
1.7 1.2 -
0.1 0.04 0.2
-
From International Dairy Federation (IDF) (1993).
by the manufacturing steps. However, it is during ripening that the characteristic flavor and texture of the individual cheese varieties develop. The principal scientific aspects of cheese manufacture and ripening will be discussed in this review; more comprehensive coverage can be found in textbooks edited by Fox (1987, 1993a). II. COMPOSITION AND CONSTITUENTS OF MILK
The coagulability of milk by acid or proteinases is due to certain characteristics of the milk proteins and is influenced by other milk constituents,
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
169
especially milk salts. The development of cheese flavor and texture is due to certain changes in the constituents of the cheese curd during ripening. Therefore, knowledge of chemistry of milk constituents is essential for a thorough understanding of cheese manufacture and ripening. However, such topics are outside the scope of this review and the interested reader is referred to textbooks on dairy chemistry by Walstra and Jenness (1984) and Fox (1982, 1983,1985, 1989b, 1992,1995). 111.
CONVERSION OF MILK TO CHEESE CURD
The conversion of milk to cheese curd essentially involves coagulating the casein, either isoelectrically or enzymatically; if present, the milk fat is occluded in the curd. The mechanisms of rennet and acid coagulation of casein and the subsequent manipulation of the coagula to produce cheese curd are described in the following sections. A. RENNET COAGULATION OF MILK The rennet coagulation of milk is a two-stage process: the primary phase involves the enzymatic production of para-casein and TCA-soluble peptides [(glyco)macropeptides] while the secondary phase involves the aggregation or gelation of para-casein Ca2+ at temperatures >20°C; the two stages overlap somewhat. The subject has been reviewed by Fox (1984) and Dalgleish (1992, 1993). 1. Primary Phase of Rennet Action
K-Casein was first isolated by Waugh and von Hippel (1956), who showed that this protein is responsible for the stability of the casein micelles and that its micelle-stabilizing properties are lost on renneting. Only K-casein is hydrolyzed to a significant extent during the primary phase of rennet action. The primary cleavage site is Phelo5-MetlM(Delfour et al., 1965), which is many times more susceptible to hydrolysis by acid proteinases (all commercial rennets are acid proteinases) than any other peptide bond in the milk protein system. The unique sensitivity of the Phe-Met bond has aroused interest. The dipeptide, H-Phe-Met-OH, is not hydrolyzed nor are tri- or tetrapeptides containing a Phe-Met bond. However, this bond is hydrolyzed in the pentapeptide, H-Ser-Leu-Phe-Met-Ala-OMe (Hill, 1968,1969) and reversing the positions of serine and leucine in this pentapeptide, to give the correct sequence for K-casein, increases the susceptibility of the Phe-Met bond to
170
P. F. FOX et al.
hydrolysis by chymosin (Schattenkerk and Kerling, 1973). Both the length of the peptide and the sequence around the sectile bond are important determinants of enzyme-substrate interaction. Serlm appears to be particularly important (Hill, 1968, 1969) and its replacement by Ala in the above pentapeptide renders the Phe-Met bond very resistant to hydrolysis by chymosin (Raymond et al., 1972) but not by pepsins (Raymond and Bricas, 1979); even substituting D-Ser for L-Ser markedly reduces its sensitivity (Raymond and Bricas, 1979). Extension of the pentapeptide H-Ser-PheMet-Ala-Ile-OH (i.e., K-CNf104-108) from the N- and/or C-terminal to reproduce the sequence of K-casein around the chymosin-susceptible bond increases the efficiency with which the Phe-Met bond is hydrolyzed by chymosin (Visser et aL, 1976,1977).Taking the pentapeptide K-CNf104-108 as a standard, studies at pH 4.7 showed that extending the peptide toward the C-terminal by three residues, i.e., K-CN f104-111, caused a 6-fold increase in the catalytic ratio, kca,lK,, while addition of Leu103 to a pentapeptide, i.e., to give K-CNf103-108, increased the ratio 600-fold. Addition of Hislozand Prolol (i.e., K-CNf101-108) increased kc,,lKm a further 5-fold. The sequence K-CNf98-111 includes all the residues necessary to render the Phe-Met bond as susceptible to hydrolysis by chymosin as it is in intact K-casein;it is hydrolyzed -66,000 times faster than the parent pentapeptide (K-CNf104-108), with a kc,,lKm of -2 rn-lsec -l, which is similar to that for intact K-casein (Visser et al., 1980). K-Casein and the peptide K-CN f98-111 are also readily hydrolyzed at pH 6.6 but smaller peptides are not. The residues Phe and Met are not intrinsically essential for chymosin action. Replacement of Phe by Phe (NOz) or cyclohexylamine reduces kc,/ K , 3- and 50-fold, respectively (Visser et al., 1977). Oxidation of Metlo6 reduces kc,,lKm 10-fold but substitution of norleucine for Met increases this ratio 3-fold. Neither porcine nor human K-casein possesses a Phe-Met bond [both have a Phe-Ile bond at this position (Brignon et al., 1985; Chobert et al., 1976; Fiat et al., 1977)], yet both are readily hydrolyzed by calf chymosin, although more slowly than bovine K-casein; in contrast, porcine milk is coagulated more effectively than bovine milk by porcine chymosin (Foltmann, 1987). Thus, the sequence around the Phe-Met bond, rather than the bond itself, contains the important determinants for hydrolysis. The particularly important residues are Ser lm, the hydrophobic residues Leulo3and Ileloe,at least one of the three histidines (residues 98, 100, or 102), some or all of the four prolines (residues 99, 101, 109, and llO), and Lyslll. Visser et al. (1987), using chemical and enzymatic modifications of the peptide K-CN f98-112, attempted to identify the relative importance of residues in the regions of 98-102 and 111-112. They suggest that the sequence Leulo3-Ilelo8,which probably exists as an extended 0-structure
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
171
(Loucheux-Lefebvre et af., 1978; Raap et al., 1983), fits into the active site cleft of acid proteinases. The hydrophobic residues, Phelos,Metlos,Leulo3, and Ilelos, are directed toward hydrophobic pockets along the active site cleft while the hydroxyl group of Serlm forms a hydrogen bridge with some counterpart on the enzyme. The sequences 98-102 and 109-111 form pturns (Loucheux-Lefebvre et al., 1978; Raap et af., 1983) around the edges of the active site cleft and are stabilized by Pro residues at positions 99, 101, 109, and 110. The three His residues, at positions 98, 100, 102, and Lyslll are probably involved in electrostatic bonding between enzyme and substrate. Small peptides mimicking or identical to the sequence of K-casein around the Phe-Met bond, especially chromogenic peptides, are very useful for determining the activity of rennets, independent of variations in the nonenzymatic phase (e.g., Hill, 1969; Raymond er af., 1973; de Koning et al., 1978; Martin et al., 1981; Salesse and Garnier, 1976; Visser and Rollema, 1986). Since the specific activity of different rennets on these peptides varies, methods for quantifying the proportions of different acid proteinase in commercial rennets have been proposed (e.g., de Koning et al., 1978; Martin et al., 1981; Salesse and Garnier, 1976). a. Rennets. The rennets used in cheesemaking are crude preparations of selected proteinases. Many proteinases will coagulate milk under suitable conditions but most are too proteolytic relative to their milk clotting activity and cause excessive proteolysis during coagulation and subsequent ripening, leading to reduced cheese yields and/or inferior cheese quality. Traditionally, the most widely used rennets were extracts of the stomachs of young calves, kids, or lambs in which the principal proteinase is chymosin. Extracts of certain plants have been used since ancient times and extracts of certain species of thistle are still used for certain cheeses, e.g., Serra de Estrela in Portugal. Chymosins are aspartyl (acid) proteinases produced by neonatal mammals for the specific purpose of coagulating milk in the stomach, presumably to delay its discharge into the intestine, and thereby improve the efficiency of digestion. The general proteolytic activity of chymosinsis low in comparison with their specific action on K-casein and they probably play little or no role in the general digestion of proteins. The chymosins are well characterized at the enzymatic and molecular levels (see reviews by Foltmann, 1987,1993). As the young animal ages, chymosin is gradually replaced by pepsin as the principal gastric proteinase at a rate which is affected by the animal’s feed, being slow on a milk diet. Owing to increasing world cheese production (-4% p.a. over the past 30 years), concomitant with a reduced supply of calf rennet (due to a
172
P. F. FOX
el al.
reduced number of calves and a tendency to slaughter more mature calves), the supply of calf rennet has been inadequate for many years and has led to a search for rennet substitutes. Although many proteinases can coagulate milk, only six rennet substitutes have been found to be more or less acceptable: bovine, porcine, and chicken pepsins and the acid proteinases from Rhizomucor miehei, R. pusillus, and Cryphonectria parasitica. Chicken pepsin is the least suitable of these and is used widely only in Israel. Bovine pepsin is probably the most satisfactory and many commercial “calf rennets” contain up to 50% bovine pepsin; its proteolytic specificity is generally similar to that of calf chymosin. The proteolytic specificities of the three principal fungal rennets are considerably different from that of calf chymosin but the acceptability of most cheese varieties made using fungal rennets is fairly good. Microbial rennets are widely used in the United States but to only a limited extent in Europe. The extensive literature on rennet substitutes has been reviewed by Sardinas (1972), Ernstrom and Wong (1974), Nelson (1975), Green (1977), and Phelan (1985). Although rennets are relatively cheap, they represent the largest single industrial application of enzymes with a world market of ca. 30 X lo6 liters of standard rennet per annum (worth US$250-350 X lo6). The gene for calf chymosin has been cloned in Kluyveromyces marxiaus var. lactis (Gist Brocades), Escherichia coli (Pfizer), and A. niduians (Hansen’s). Genetically engineered chymosins, which have been cleared by Food and Drug Administrations in many but not all countries, have given very satisfactory results in large-scale cheesemaking trials on several varieties and are now widely used commercially in several countries (see Teuber, 1990). b. Principal Factors Affecting the Hydrolysis of u-Casein. The optimum pH for chymosin on K-CNf98-112 is 5.3-5.5 (Visser et al., 1987) and for the first stage of rennet action in milk is -6.0 (Van Hooydonk et al., 1986b). The pH optimum for the general proteolytic activity of chymosin is -4. The optimum temperature for the coagulation of milk by calf rennet at for the hydrolysis of pH 6.6 is -45°C. The temperature coefficient K-casein in Na caseinate is -1.8, activation energy, E,, is -10,000 cal/mol, and activation entropy, AS, is --39 cal/deg/mol (see Fox, 1988). Severe heat treatments, especially >8O”C, adversely affect the rennet coagulation of milk. Although changes in calcium phosphate equilibria are contributory factors, complexation of P-lactoglobulin and/or a-lactalbumin with K-casein via intermolecular disulfide bond formation is the principal factor responsible (see Fox, 1988). Most authors agree that the primary and, especially, the seconday (nonenzymatic) phase of rennet coagulation are adversely affected by severe heat treatments, as is the strength of the
(el,,)
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS 173
resulting gel. The adverse affects of heating on the rennetability of milk can be reversed by acidification before or after heating or by addition of CaC12 (see Fox, 1988). 2.
Secondary (Nonenzymatic) Phase of Coagulation
Hydrolysis of K-casein during the primary phase of rennet action releases highly charged, hydrophilic macropeptides, representing the C-terminal '/3 of K-casein, which protrudes from the micelle surfaces, thereby reducing the zeta potential of the casein micelles from ca. -20 to ca. -10 mV and destroying their steric stabilizing layer. When ca. 85% of the total K-casein has been hydrolyzed, the micelles begin to aggregate; reducing the pH or increasing the temperature from the normal values (-6.6 and -31"C, respectively) permits coagulation at a lower degree of K-casein hydrolysis (see Fox, 1984, 1988 Dalgleish, 1992, 1993). Coagulation of rennet-altered micelles depends on a critical concentration of Ca2+,which may act by crosslinking micelles via serine phosphate residues or by charge neutralization. Colloidal calcium phosphate is also essential for coagulation, but its role can be partially fulfilled by increased [Ca'.]. Partial enzymatic dephosphorylation of casein impairs rennet coagulability. Cationic species predispose casein micelles to coagulate and may even coagulate unrenneted micelles. Chemical modification of histidine, lysine, or arginine residues inhibits coagulation, presumably by reducing the positive charge on the micelles. It has been suggested that coagulation occurs via electrostatic interaction between a positively charged cluster toward the C-terminal of para-K-casein, which is exposed on removal of the macropeptide, and an unidentified, negatively charged cluster on neighboring micelles. The coagulation of renneted micelles is very temperature-dependent (Qlo-16) and normal bovine milk does not coagulate <18"C unless [Ca2+] is increased. The marked difference between the temperature dependence of the enzymatic and nonenzymatic phases of rennet coagulation has been exploited in the study of the effect of various factors on rennet coagulation, in attempts to develop a system for the continuous production of cheese or casein curd and in the application of immobilized rennets. The very high dependence on temperature of rennet coagulation suggests that hydrophobic interactions play a major role. 3. Gel Assembly and Strength
The assembly of renneted micelles into a gel has been studied using various forms of viscometry, electron microscopy, and light scattering. The
174
P. F. FOX et al.
micelles remain discrete until -60% of the visual coagulation time has elapsed, after which they begin to aggregate steadily, without sudden changes, into chain-like structures which eventually link up to form a network (for reviews, see Fox, 1984; Dalgleish, 1992,1993; Green and Grandison, 1993). Aggregation of rennet-altered micelles can be described by the Smoluchowski theory for diffusion-controlled aggregation of hydrophobic colloids when allowance is made for the need to produce, enzymatically, a sufficient concentration of particles capable of aggregating. Gel assembly is poorly understood and will not be discussed further. The strength of renneted milk gels (curd tension, CT) is very important, especially from the viewpoint of cheese yield. This subject has been reviewed (Fox, 1984; Green and Grandison, 1993). Suffice it to say here that curd tension is positively affected by protein concentration, [Ca’+], and reduced pH to -5.9 and adversely by high heat treatments. Thus, CT is affected by the same variables that affect rennet coagulability. 4. Curd Syneresis
When the renneted milk gel has reached the desired degree of firmness, usually assessed subjectively by the cheesemaker but for which objective instrumental methods are sought, the gel is cut or broken. If left undisturbed, the gel is stable over a long time period but when broken, the curd particles syneresis, expressing whey. The tendency of renneted casein gels to synerese enables the cheesemaker to control the extent of dehydration and thereby the composition of the resulting cheese curd, which in turn strongly affects cheese texture and the various biochemical reactions that occur during ripening. The individuality of cheese varieties could be said to commence with the syneresis process. The mechanism of syneresis at the molecular level is poorly understood but the physical and operational aspects have been well described and reviewed (van den Bijgaart, 1988;Pearse and MacKinlay, 1989;Akkerman, 1992; Walstra, 1993); the subject will not be reviewed here. Syneresisis promoted by increasing temperature, decreasing pH, addition of CaC12,fine cutting of the gel, vigorous stirring, and high casein concentrations and retarded by high heat treatment of the milk and increasing fat content. Curd for many cheese varieties is pressed to shape and then consolidated to remove some whey. Additional whey is removed by salting (for hard and semihard cheeses, -2 kg H 2 0 are lost per kilogram NaCl absorbed).
5. Acidification during Cheese Manufacture One of the primary events in the manufacture of most, if not all, cheese varieties is the fermentation of lactose to lactic acid by selected lactic acid
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
175
bacteria added as a culture (starter) or, in traditional cheesemaking, by the indigenous microflora. The rate and point in the process at which lactic acid is principally produced are characteristic of the variety; e.g., in Cheddar-type cheeses, most of the acid is produced prior to moulding while in most other varieties it occurs mainly after moulding. For rennet-coagulated cheeses, the pH reaches ca. 5 within 5 to 12 hr. Acid production affects almost all facets of cheese manufacture and is probably the most important factor affecting cheese quality. Among the most important consequences of acid development are:
1. Activity of the coagulant during manufacture. 2. Retention of coagulant in cheese curd. 3. Activity and perhaps specificity of the coagulant during ripening. 4. Activity of plasmin (indigenous milk alkaline proteinase). 5. Curd tension. 6. Curd syneresis. 7. Solubilization of colloidal calcium phosphate, which, among other factors, affects curd (cheese) texture, stretchability, and meltability. 8. Shelf-life.The spectrum of cheese varieties exhibits a range of storage stability from low (e.g., Cottage, Camembert) to high (e.g., Parmesean); pH and a , play complementary roles in determining cheese stability and rate of ripening. 9. Growth and/or survival of pathogens. 10. The taste of acid-coagulated and young rennet-coagulated cheeses is strongly affected by the concentration of acid in the cheese. 11. Lactate serves as substrate for the production of propionic acid, acetic acid, and C 0 2 during the ripening of Swiss-type cheeses. The significance of some of these functions is discussed further below. The metabolism of lactose by lactic acid bacteria is well understood but will not be discussed here; the interested reader is referred to reviews by Cogan and Daly (1987), Fox et af. (1990), and Cogan and Hill (1993). a. Buffering Capacity. The pH of cheese curd is determined directly by the amount of lactic acid produced and indirectly by the buffering capacity of the curd which is determined primarily by the casein with contributions from phosphate and citrate. Cheese curd has a relatively low buffering capacity in the pH range 6.8 to 5.5, but buffers strongly at pH 5.5 to 4.5. Due to the high buffering capacity below pH 5.5, even extensive acid production in Cheddar curd after salting has little effect on the final pH of the cheese and explains why the pH of highly acidic cheese, e.g., Cheshire (pH 5.0), increases little during ripening. The higher the moisture content of the curd, the higher its lactose content and hence the higher ratio of lactate to buffering substances and the lower the pH. Lawrence
176
P. F. FOX et al.
and Gilles (1982) suggested that the buffering capacity of curd is determined largely at whey drainage since 85 to 90% of the total calcium, phosphate, and citrate removed during cheesemaking are removed at this stage. b. Structure and Texture. The pH at whey draining determines the mineral content of cheese curd. The loss of calcium and phosphate from the casein micelles determines the extent to which they are disrupted and largely determines the basic structure and texture of a cheese (see Lawrence et ai., 1983; Fox et al., 1990, for references). Curds with a low pH tend to have a crumbly texture, e.g., Cheshire, while high pH curds tend to be elastic, e.g., Emmental. Scanningelectron microscopy showed that the aggregates in Swiss or Gouda (high pH, high calcium) are globular, like the original submicelles in milk, whereas in Cheshire (low pH, low calcium), the protein aggregates are smaller and less well organized and occur as strands or chains. The aggregates in Cheddar (intermediate pH) are immediate between those in Gouda and Cheshire. Other cheesemaking variables, e.g., the direct effect of pH on protein charge, curd composition (levels of water, fat, and protein), also influence the nature of the protein aggregates. The texture of Cheddar cheese is considered to be more dependent on pH than on any other factor (see Lawrence and Gilles, 1982; Fox et al., 1990, for references); for the same calcium content, the texture of Cheddar varies from “curdy” (pH 2 5.3), to “waxy” (pH 5.3 >5.1), to “mealy” (pH 4 . 1 ) . Suggested explanations for this pH dependence include micelle hydration, especially in the presence of NaCl, and the extent to which colloidal calcium phosphate (CCP) is solubilized. Proteolysis during ripening modifies cheese texture. The casein in lowpH Cheddar is hydrolyzed more rapidly than in normal pH cheese partly because solubilization of CCP causes micellar dissociation and renders the caseins more susceptible to proteolysis (O’Keeffe et al., 1975) and partly because more chymosin is retained, and is more active, at low pH (Holmes et al., 1977; Creamer et at., 1985). Cheddaring involves the development of a fibrous curd structure, but this occurs only at a curd pH d 5.8 and is a consequence of the loss of calcium from the protein matrix (Lawrence and Gilles, 1982). Optimum stretchability of Mozzarella occurs at pH 5.2-5.4.
c. Enzyme Retention and Activity. The retention and activity of proteinases in cheese are influenced by the pH of the curd and cheese. Thus, other conditions being equal, cheese pH can be expected to influence the rate and extent of proteolysis during ripening and therefore cheese texture and flavor.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
177
As discussed in Section IV E, the proteolytic activity of residual rennet is responsible for primary proteolysis in cheese. Most (>90%) of the rennet is lost in the whey but the level retained depends on pH and cooking temperature. The chymosin content of Gouda is inversely related to the pH at renneting and to cooking temperature (Stadhouders and Hup, 1975). Progressively more chymosin and pepsin is retained in Cheddar curd as the pH at coagulation is reduced from 6.6 to 5.2 but the retention of microbial rennets is not affected (Holmes et al., 1977; Creamer et al., 1985). On renneting at pH 6.6, -6% of the original chymosin activity is retained in Cheddar cheese after pressing, while only 2-3% of microbial rennets and no detectable pepsin activity remain. Pepsin is denatured rapidly at pH 6.6 but becomes more stable at lower pH. Why microbial rennets behave differently from chymosin is not obvious. The activity of residual chymosin and other rennets, which are acid proteinases, is likely to be increased at low pH. Chymosin is partly or totally inactivated in high-cooked cheeses, e.g., Swiss (Creamer, 1976b; Matheson, 1981; Garnot and Molle, 1987), but plasmin is stable at 85°C for 5 min (Creamer, 1976b). Plasmin is associated with the casein micelles in milk, but is released 5pH 4.6 (see Grufferty and Fox, 1988). Its activity in Swiss and Dutch-type cheeses is considerably higher than that in Cheddar (Richardson and Pearce, 1981); the high plasmin activity in high-cook cheeses appears to be due to inactivation of plasmin inhibitors or the inhibitors of plasminogen activators (Farkye and Fox, 1990). Cheese contains high levels of plasminogen, which may be activated to plasmin (Ollikainen and Nyberg, 1988), especially if the inhibitors of the piasminogen activator are inactivated. In the case of washedcurd cheeses, e.g., Gouda, high plasmin activity may be due to the removal of inhibitors on washing.
6. Salting of Cheese All cheese are salted but the concentration varies markedly, as does the method of salt application (for review, see Guinee and Fox, 1987, 1993). Typical concentrations are: Emmental, 0.5%; Cheddar, 1.7%;Gouda, 2.0%; Blue, 4.0%; Feta, 7%. Converted to salt-in-moisture, these values range from 1.5 to 12%. Salting is usually performed by immersion in brine for a period ranging from 30 min to several weeks, depending on the size of the cheese, its moisture content, and the desired salt content. Initially, there is a steep salt gradient from the surface to the center, which affects biological and biochemical events in the different regions of the cheese. Salt diffuses inward and, eventually, equilibrium is established throughout the cheese if time permits. Some cheeses, e.g., Blue, are salted by the application of
178
P. F. FOX et af.
dry salt on the surface of the cheese. The salt dissolves in the surface moisture, causing the outward movement of moisture. Salt diffuses into the cheese, in effect as in brine-salted cheeses. Cheddar and other British cheeses are salted by mixing dry salt with milled or broken curd. Each curd piece then behaves as a minicheese, with salt diffusing from the surface to the center. There is no overall gradient throughout the cheese and salt distribution should be uniform from the start. However, salt distribution is frequently uneven and in such cases equilibrium is not established, probably due to the lack of an overall gradient. Since the salt concentration in Cheddar cheese is sufficient to inhibit or retard acid production, the pH of Cheddar curd is close (-5.3) to its final value (-5.1) at salting and prior to moulding, whereas acidification of surface-salted (brined or dry) occurs mainly after moulding. Salt plays several important roles in cheese ripening:
1. Controls the growth and survival of microorganisms, including food poisoning and pathogenic bacteria. 2. Influences the activity of enzymes. 3. Regulates water activity (u,,,), especially of young cheese. 4. Affects cheese composition due to loss of water (2 kg H20 lost per 1 kg NaCl absorbed). 5. Affects cheese flavor directly and indirectly via its effect on bacteria and enzymes. 6. Affects cheese texture via dehydrating effects on the caseins, cheese composition, and its effect on proteolysis. 7. Nutritional: high dietary intakes of NaCl are considered undersirable; however, in most cases, cheese is a relatively small contributor to dietary NaC1. B. FRESH ACID-CURD CHEESE VARIETIES Fresh acid-curd cheeses refer to those varieties produced by the coagulation of milk, cream, or whey via acidification or a combination of acid and heat and which are ready for consumption directly after manufacture (Fig. 1). They differ from rennet-curd cheeses, where coagulation is induced by the action of rennet at a pH value of 6.4-6.6, in that coagulation occurs close to the isoelectric point of casein, i.e., pH 4.6, at 30°C or at higher values when a higher temperature is used, e.g., Ricotta (pH 6.0, 8OOC). A very small amount of rennet may be used in the production of Quarg, Cottage, and Fromage frais to provide firmer coagulum and minimize casein losses on whey separation, but it is not essential. Annual world production of fresh, acid-curd cheeses amounts to about 3.2 million tonnes, ca. 25% of
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
Acid coagulated Quarg-type - skim milk Quarg - Full fat Quarg - Tvorog
179
Acid-heat'coagulated Queso Blanco Ricotta Mascarpone
Fromage h s i Labneh Labawh Fresh cheese preparations Cream cheese-type - doublelsingle Cream cheese - Petit Suisse - Neufchatel Cottage cheese-type - Low/fat Cottage cheese - Bakers Cheese
Whe'y based Ricottone
Brown 'cheese' - Mysost - Gudbrandsalost - Eke Geisost - Floteost
FIG. 1. Fresh acid-curd cheese varieties (from Guinee et al., 1993).
total cheese [International Dairy Federation (IDF), 1986; Milk Marketing Board (MMB), 19911; Quarg, Cottage, Cream, Fromage frais, and Ricotta are the most important types. Consumption has grown by ca. 4% per annum during the past decade. Factors contributing to this growth include: (i) The large variety available in terms of consistency and flavor as effected by variations in processing parameters, blending of different cheese types to create new products and the addition of sugars, fruit purees, spices, or condiments. (ii) Their soft, ingestable consistency, which makes them safe and attractive to very young children. (iii) Their healthy perception by diet-conscious consumers. In general, their fat content is lower than that of rennet-curd cheeses; double cream cheese, an exception in the group, has a fat content similar to that of
180
P. F. FOX et al.
Cheddar. However, acid-curd cheeses are relatively low in calcium compared to rennet-curd cheeses such as Cheddar (ca. 0.75% Ca) and Swiss (ca. 0.95% Ca) (Table 111). This review will concentrate on those varieties produced by coagulation at pH values close to the isoelectric point of casein at 20-40°C (e.g., Quarg, Fromage frais, cream cheese); acid-heat coagulated fresh cheese varieties (e.g., Ricotta) were reviewed by Torres and Chandan (1981a,b). 1. Production
Production generally involves pretreatment of milk (standardization, pasteurization, and/or homogenization), slow quiescent acidification, gelation, TABLE 111 APPROXIMATE COMPOSITION OF VARIOUS FRESH CHEESES'
% (w/w)
Variety Cream cheese Double Single Neufchatel Labneh Quarg Skim milk Full fat Cottage cheese Low fat Creamed Fromage frais Skim milk Queso blanco Ricotta Whole milk Part skim Ricottone Mysost Gudbrandsalost Floteost
Dry matter
Fat
Protein
Lactose (lactate)
Salt
40 30 35 25
30.0 14.0 20.0 11.6
8-10 12 10-12 8.4
2-3 3.5 2-3 4.3
0.75 0.75 0.75 -
80 100 75 -
4.6 4.6 4.6 4.2
18 27
0.5 12.0
13 10
3-4 2-3
-
120 100
4.5 4.6
21 21
2.0 5.0
14 13
-
-
90 60
4.8 4.8
14 49
1.0 15.0
8 23
3.5 1.8
3.9
28 25 18
13.0 8.0 0.5
11.5 12 11
3.0 3.6 5.2
-
82 80
30.0 19.0
11 11
-
38
-.
46
-
Ca (mg/100 g)
0.15 200 280 400
pH
4.4 5.4 5.8 5.8 5.3
Compiled from data by USDA (Posati and Orr,1976), Kosikowski (1982), Winwood (1983), Oterholm (1984), Patel er al. (1986), Sohal er al. (1988), Tamime et af. (1989), and Jelen and Renz-Schauen (1989).
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
181
whey separation, and curd treatments (pasteurization, shearing, addition of salt, condiments, and stabilizers, and/or homogenization) (Fig. 2). Acidification is generally slow, 12-16 hr at 21-23°C (long set) or 4-6 hr at 30°C (short set), via the in situ production of lactic acid by a Lactococcus starter culture; less frequently, an acidogen, e.g., glucono-&lactone, is used. Many processing factors (e.g., pasteurization, rate and temperature of acidification,level of gel-forming protein, gel pH) influence coagulum structure and hence the textural/organoleptic attributes and the physicochemical Standardized M&
I
Pretreatment - Pasteurization - Homogenization and/or - Partial acidification
I
Cooling 22-30°C (-Starter Incubation (Quiescent)
( - 1%)
I k-- - - - -Rennet (0.5-1 rnl/lOO)
I
Gelled Acidified Milk (PH 4.6)
I
Separation (Dehydration)
I
W h e y l P e r m e a t e i Curd -Cold Pack-wProduct: Quarg Fromage frais I Cottage cheese I Pasteurization, Hydrocolloid and Condiment addition, and/or Homogenization
I
Hot, Treated Curd - Hot Pack-+Product: Cream cheese
Other Fresh cheeses Cream, andlor Yogurt andlor Condiments
I
Other
-7 Heat, lend Homogenize
I
Hot Blend
Hot Pack
Fresh cheese preparations
FIG. 2. Generalized scheme for the production of fresh cheese products (from Guinee et aL, 1993).
182
P. F. FOX et al.
stability of the end product. This is especially true for cold-pack products where the curd, following whey separation and concentration, is not treated further. In hot-pack products, curd treatments (i.e., pasteurization, homogenization, hydrocolloid addition) have a major impact on the quality of the final product (Guinee et al., 1993).
2. Principles of Acid Gel Formation Slow acidification of milk under quiescent conditions is accompanied by two opposing sets of physicochemical changes: 1. A tendency toward disaggregation of the casein micelles to a more disordered system as a result of (a) solubilization of CCP, which, at 20-30°C, is fully soluble at pH 5.2-5.3 (Creamer, 1985;van Hooydonk et al., 1986a); (b) a pH- and temperature-dependent dissociation of individual caseins, especially p , from the micelles with a concomitant increase in the level of serum casein [(Creamer, 1985; Roefs et al., 1985;Dalgleish and Law, 1989); casein dissociation decreases with decreasing pH to -6.2, then increases to a maximum at pH 5.3-5.6 (depending on temperature), and thereafter decreases to a minimum at the isoelectric pH (Snoeren et al., 1984)l; (c) An increase in micelle solvation and porosity as a consequence of a and b, over the pH range 6.7 to 5.3-5.4 (Vreeman et al., 1989). 2. A tendency for the casein micelles to aggregate to a more ordered system due to: (a) The reduction of the negative surface charge on the micelles and hence intermicellar repulsive forces (Darling and Dickson, 1979; Schmidt and Poll, 1986); (b) a decrease in casein hydration in the pH range 5.4 to 4.6 (Creamer, 1985);and (c) the increase in the ionic strength of the milk serum (due to the increased concentrations of calcium and phosphate ions), which has a shrinking effect on the matrix of casein micelles (Vreeman et al., 1989).
-
At pH values greater than that at the onset of gelation, i.e., -5.1-5.3 at 20-30°C, disaggregating forces predominate and hence a gel is not formed. At lower pH values, forces that promote aggregation predominate and gelation occurs. a. Structural Changes. Slow quiescent acidification of milk from pH 6.6 to 4.6 is accompanied by a number of concerted structural changes which are summarized below: 1. At pH values > 5.6, no major changes are observed: the individual micelles retain their shape, dimensions and integrity; a wide spectrum
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
2. 3.
4.
5.
183
of particles with an average diameter of -120 nm are present (Heertje et al., 1985). At the pH of maximum casein dissociation (-pH 5.5), the micelles become more porous. On further acidification to pH 5.2, where practically all CCP is solubilized, smaller particles, in addition to the “original” micelles, are formed; these new particles are probably formed by aggregation of dissociated caseins (J. Visser et af.,1986 Roefs, 1986; Rollema and Brinkhuis, 1989). At pH -5.2, i.e., the pH at the onset of gelation, a heterogeneous distribution of casein aggregates (composed of aggregated casein micelles and new casein particles) with a range of sizes is observed. Further reduction in pH is paralleled by a touching of aggregates which initiates the formation of loose, porous strands. Eventually, on close approach to the isoelectric pH, dangling strands touch and crosslink to form a three-dimensional particulate gel network which extends, more or less continuously, throughout the serum phase.
b. Prerequisites for Gel Formation. Acid casein gels may be considered as particulate network gels, i.e., they consist of overlapping, cross-linked strands which are composed of particles (i.e., casein aggregates) linked together by various types of bonding (Harwalkar and Kalab, 1980; Schellhaass and Morris, 1985; Heertje et al., 1985). Aggregation and structural rearrangement of casein during quiescent acidification of milk may result in the formation of a gel, as described, or in a precipitate, depending on the extent of aggregation. Gelation occurs when aggregation forces slowly overcome repulsive forces, resulting in the formation of relatively loose, porous, hydrated aggregates with a small density gradient between them and the serum phase in which they are dispersed. Owing to the relatively low density gradient, the aggregates have sufficient time to knit together, via strand formation, to form a continuous network before sedimenting. When conditions promoting aggregation are more extreme (i.e., rapid acidification under nonquiescent conditions at high temperature), casein particles aggregate more rapidly to form smaller, less porous, and less hydrated aggregates which, owing to their relatively high density, sediment as a precipitate which lacks the matrix continuity and water-holding characteristics of a gel (Kinsella, 1984). To obtain a gel rather than a precipitate, the number of attractive forces, and hence the surface area of contact, between the dispersed particles must be limited (Walstra et at., 1985; Walstra and van Vliet, 1986). Such limited interparticle attractions are promoted by an optimum ratio of attractive to repulsive forces between the conformationally rearranged casein particles (Kinsella, 1984), which in turn is achieved by the desired rate of concerted
184
P. F. FOX
er al.
physicochemical changes. As the number of interparticle points of attraction increases (e.g., when the rate of acidification is increased), the resulting gel becomes less structurally organized, coarser, less voluminous, and closer to a precipitate. Alternatively, if the number of interparticle attractive sites is lower than optimum, slowly forming aggregates may have sufficient time to precipitate before fusing to form into strands of a network. An example of the latter is the defect in Cottage cheese production known as “major sludge formation,” whereby phage infection of the starter, after acid development has progressed to an advanced stage (pH 52-53), leads to casein precipitation rather than gelation (Grandison et al., 1986) or clumping of starter bacteria by agglutinins causes localized concentration of bacterial cells and consequently excessively fast acid production in some regions of the vat and vice versa in other regions (Salih and Sandine, 1980).
-
3. Structure-Quality Relationships Gel structure is a major determinant of quality attributes, such as mouthfeel (smoothnesskhalkiness), appearance (coarseness/smoothness), and physicochemical stability (absence of wheying-off and graininess during storage) of fresh acid-curd cheese products, especially cold-pack varieties. The relationship between product quality and gel structure may be explained by reference to Fig. 3, which depicts the structural differences between fine and coarse gels in which the concentrations of gel-forming protein are equal. In the fine gel, Fig. 3A, the micelles have formed into thin strands (chains) giving a highly branched, continuous gel network. Such a structure has an even distribution of gel-forming protein, a relatively
FIG. 3. Schematic diagram of a fine-structured yogurt gel (A) made from heated (90°C) milk and a coarse-structuredyogurt gel (B) made from unheated milk (modified from Harwalkar and Kalab, 1980).
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
185
low porosity, a high water-holding capacity, and a low tendency to syneresis spontaneously (i.e., in the absence of external pressure) (Harwalkar and Kalab, 1980; Green, 1980). Conversely, in the coarse gel, the micelles have fused to a much greater degree, giving thicker strands and a gel which is more discontinuous and porous and more susceptible to syneresis on storage. a. Syneresis. Syneresis requires rearrangement of the gel matrix into a more compact structure as effected by the breaking of bonds within strands and the consequent formation of new bonds (Walstra er al., 1985). Rearrangement necessitates stress (for strand breakage), which can be effected by the application of external pressure, e.g., centrifugation, gravity, stirring, and/or cutting, or internally by the spontaneous breaking of gel strands due, possibly, to micelle flow and thermal motion of gel strands (Walstra et al., 1985). Shrinkage of the casein particles in the network, as induced by a reduction in pH and/or increase in temperature following gel formation, may enhance both types of syneresis. Simultaneously, the outward flow of aqueous phase is impeded by the sieve effect of the pores which becomes increasingly greater as progressive structural rearrangement leads to matrix contraction and reduced porosity. Because of the reduced syneresis for a given pressure, due to the sieve effect, and the decrease in the net pressure acting on the entrapped liquid (due to the counteraction of syneretic pressure by elastic reaction forces building up in the matrix), the rate of outward migration of aqueous phase decreases with time. For unidimensional flow through a porous medium, such as a milk acid gel, the rate of syneresis (v) may be expressed by Darcy’s law (Walstra er al., 1985; Roefs, 1986). u=
BAP/hl,
where u is liquid flux (i.e., volume flow rate in the direction 1, divided by the cross-sectional area, perpendicular to 1, through which the liquid flows) (msec-I), B is the permeability coefficient of the matrix, which corresponds to the average cross-sectional area of the pores, h is viscosity of the serum flowing through the matrix, AP is the pressure gradient arising from syneretic pressure exerted on the entrapped serum by the matrix, and 1 is the distance over which the serum flows. The permeability coefficient, B, depends on the volume fraction of the protein matrix and the spatial distribution of the matrix strands (i.e., gel fineness/coarseness). The greater the permeability coefficient, the less is the resistance to the flow of moisture through the gel for a given syneretic pressure. Hence, fine gel struc-
186
P. F. FOX et al.
tures, which have narrow pores, have lower permeability than their coarsestructured counterparts and are less susceptible to syneresis. b. Rheology. Structure also has a major influence on the rheological properties of gels. The response of the structure to stress depends on many factors, including: 1. The number of strands per unit area. Considering a gel to which a relatively small stress (i.e., much less than the yield value) is applied in the direction x, the elastic modulus ( G ' ,i.e., ratio of stress to strain, a/ y, in the linear viscoelastic region) can be related to the number of strands per unit area according to the equation (van Vliet and Walstra, 1985) G' = CN dzA/dx2,where N is number of strands per unit area in a cross section, perpendicular to x, bearing the stress, Cis the characteristic length determining the geometry of the network, and d A is the change in Helmholtz energy when the particles in the strands are moved apart by a distance dx. 2. Gel homogeneity. The homogeneity of the gel determines the number of stress-bearing strands. For a coarse gel network, there are fewer stressbearing strands than in a fine gel; however, the thickness, and hence strength, of the stress-bearing strands are, on average, greater in the coarser gel because of the greater number of attractions between the aggregates. 3. The number and type of bonds between the basic building units, i.e., aggregates, within a strand; the smaller the number and the weaker these bonds, the more susceptible the strand is to deformation. In fine gel structures, there are probably fewer intra- and interaggregate attractions/ bonds than in coarser gels. Hence, products with finer gel structures may be considered to exhibit less elastic, and more viscous, behavior than those with coarser networks. d. Sensory Attributes. Structure may also influence the sensory characteristics of gels, especially in products such as set yogurt where the gel, formed in its package, is not subjected to concentration or other treatments. In such cases, the formation of large, dense protein conglomerates (fused aggregates) causes a chalky or gritty mouthfeel (Modler et at., 1989). These defects are more prevalent in products where the gel, following fermentation, is concentrated andlor heated; these conditions promote protein dehydration and hence the formation of large protein conglomerates. 4. Factors That Influence Gel Formation and Structure
a. Level of Gel-Forming Protein. For a given protein type and degree of gel fineness, higher concentrations of gel-forming protein result in a
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS 187
denser [higher number of strands (of equal thickness) per unit volume], more highly branched, less porous network (Harwalkar and Kalab, 1980; Modler and Kalab, 1983; Modler et al., 1983). However, for a given protein concentration, the structure is strongly influenced by the ratio of casein to whey protein (Modler and Kalab, 1983; Modler ef al., 1983; Tamime et al., 1984). Reducing the ratio from 4.6 : 1 to 3.2 : 1 results in set yogurt with a finer, more highly- branched and less-porous structure, with a relatively low propensity to syneresis. Factors which increase the effective protein concentration include: (i) addition of milk protein supplements (e.g., skim milk powder), as practiced in yogurt production, (ii) homogenization of milk (as practiced in yogurt and Cream cheese production) which converts fat globules to pseudoprotein particles (van Vliet and Dentener-Kikkert, 1982), (iii) high-heat treatments which cause denaturation and binding of whey proteins to casein; undenatured whey proteins are soluble and do not participate in gel formation.
b. Milk Heat Treatment. It is well known that high-heat treatment of milk prior to culturing to fermented products, such as yogurt, gives a smoother, more viscous consistency although the level of the effect varies considerably with the type of preheating, i.e., whether in-vat, hightemperature-short time, or UHT treatments (Labropoulus et al., 1981a,b; Parnell-Clunies et al., 1986a,b). It also results in the onset of gelation at a higher pH (Kalab et al., 1976). These effects may be attributed to extensive (>50%) whey protein denaturation which results in a higher effective concentration of gel-forming protein and a more finely structured gel network with reduced propensity to spontaneous wheying-off (Kalab et al., 1976 Harwalkar and Kalab, 1980, 1981; Parnell-Clunies et al., 1986a,b). The above effects are attributed to the formation, via disulfide interaction, of a complex between K-casein and denatured P-lactoglobulin which results in the formation of filamentous appendages which protrude from the micelle surfaces (Harwalkar and Kalab, 1980; Heerjte et al., 1985; Modler and Kalab, 1983) and prevent the close approach, and hence large-scale fusion, of micelles on subsequent acidification (Roefs, 1986). c. Zncubation Temperature. Higher temperatures (in the range 2043°C) during acidification result in: (i) the onset of gelation at higher pH values, i.e., pH 5.5 at 43°C compared to pH 5.1 at 30°C (Heertje et al., 1985) and (ii) coarser gel structures which are more prone to wheying-off on storage (Green, 1980; Schellhaass and Morris, 1985). These effects may be attributed to a higher ratio of aggregating to disaggregating forces during the early stages of acidification due to reduced casein dissociation from the
188
P. F. FOX et al.
micelles, increased protein hydrophobicity (Grigorov, 1966), and a faster rate of acidification. d. Rennet Addition. It is common practice during the manufacture of some fresh cheese products, such as Quarg and Cottage cheese, to add a small quantity of rennet (0.5-1.0 ml single-strength per 100 liters) to the milk shortly after (i.e., -1-2 hr) culture addition when the pH is -6.1-6.3. The rennet hydrolyzes some K-casein with concomitant decreases in zetapotential, casein dissociation and micelle solvation over the pH range 6.6-4.6 (Green and Crutchfield, 1971; Pearse, 1976; Darling and Dickson, 1979;Creamer, 1985; van Hooydonk et af.,1986a,b). These changes contribute to an enhanced aggregation of micelles and gelation begins at a higher pH. Hence, a firmness suitable for cutting is obtained at a higher pH (i.e., 4.8-4.9); in the absence of added rennet, cutting is performed at pH 4.6-4.7 so as to prevent excessive loss of fines on whey separation. e. Rate of Gelation. Gelation rate is faster with more rapid acid development and with increased incubation temperature in the range 20-45°C (provided starter growth is not inhibited) (Heertje et al., 1985; Kim and Kinsella, 1989). Higher rates of gelation result in the onset of gelation at higher pH values and a coarser network with a greater propensity to spontaneous syneresis (Emmons et af., 1959; Harwalkar and Kalab, 1980, 1981; Schellhaass and Morris, 1985). In an extreme situation, rapid acidification to pH 4.6 promotes rapid aggregation of casein with the formation of large dense aggregates which precipitate rather than gel. Gel formation by rapid acidification is, however, possible when the tendency of micelles pH 4.6 at low temperatures to coagulate is reduced by acidifying to (0-4°C) and subsequently heating slowly (-OS"C/min) under quiescent conditions to -30°C (Harwalkar and Kalab, 1981; Roefs, 1986)
-
5. Influence of Processing Parameters on Quality
The structure and textural characteristics of fresh, acid-curd products are influenced by many factors, e.g., milk composition (levels of fat and protein), processing parameters (heat and homogenization treatments), conditions of gel formation (incubation temperature, rate of acidification, addition of rennet, final pH), and further curd treatments. a. Texture. Increasing protein concentration, for a given rate of gelation, results in firmer and more continuous gels (Modler et af., 1983; Roefs et aL, 1985; Roefs, 1986; Mottar et af., 1989). However, for a given level of gel-forming protein, increasing the casein to whey protein ratio results in
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
189
coarser, firmer gel networks (Modler et al., 1983; Tamime et al., 1984). For a given level of solids, increasing the fat content gives a weaker gel because of the decrease in the level of gel-forming protein and the physical interference of fat globules with the protein network. However, if the milk (or cream) is homogenized, the fat globules become partially covered with casein and whey proteins and are converted to pseudoprotein particles, which then participate in network formation. Increasing the effective protein concentration in this way results in firmer gels (van Vliet and DentenerKikkert, 1982; Beyer and Kessler, 1988). The viscosity of soft fresh cheeses, the firmness of natural, fortified yogurt, and the elastic modulus, G', of acidified caseinate and reconstituted skim milk powder gels increase with increasing protein concentration (Modler et al., 1983;Tamime et al., 1984; Roefs, 1986; Korolczuk and Mahaut, 1989). High-heat treatment of milk causes an increase in the viscosity and firmness of low-solids acid-curd products, such as natural yogurts (Harwalkar and Kalab, 1981; Labropoulus et al., 1981a,b; Parnell-Clunies et al., 1986a,b; Korolczuk and Mahaut, 1989). [However, Kim and Kinsella (1989) found that preheating (90°C X 15 min) of milk caused a decrease in the firmness of gluconic acid Glactone (GDL)-acidified milk gels]. The former effect is attributed to extensive whey protein denaturation and binding of denatured whey proteins to the micelles which effects an increase in the level of gelforming protein and a finer, more highly branched continuous network. The number of stress-bearing strands (though possibly weaker) per unit volume of such a gel would be greater than in the coarser, lower-density matrix gel formed from unheated milk. For similar levels of whey protein denaturation, the type of heat treatment appears to have a significant influence on the textural parameters of fermented milks; for similar levels of whey protein denaturation, ultrahigh temperature treatment (130-150°C x 2-15 sec) gives natural yogurts of lower viscosity and firmness than high-temperature-short time (-80-90°C X 0.5-5 min) treatments, which in turn give lower values than those obtained for yogurt made using batch (63-80°C X 10-40 min) heat treatments (Labropoulus et al., 1981a,b; Parnell-Clunies et al., 1986a,b). While there is generally a strong positive correlation between the viscosity and firmness of yogurt as functions of whey protein denaturation, variations for similar levels of denatured whey protein may be due to the different rates of whey protein denaturation which alter their binding to casein micelles and hence gel structure (Morr, 1985). High incubation temperatures promote coarser gel networks which are more elastic and firmer than those formed at a lower temperature (Harwalkar and Kalab, 1981; Schellhaass and Morris, 1985; Bringe and Kinsella, 1990). It is possible that in the coarser gel, the higher stress-bearing capacity
190
P. F. FOX
el
al.
of the relatively thick strands overrides the effects of a greater gel discontinuity which would, otherwise, be expected to give a weaker more easily deformed network. The firmness of acid milk gels increases with decreasing pH toward the isoelectric point of casein (Harwalkar et al., 1977); the optimum pH for the elastic modulus, G’, is -4.5 (Walstra and Jenness, 1984; Roefs, 1986). The increase in firmness is attributed to a greater number of bonds with greater strength between the strand-forming aggregates. b. Syneresis. In the manufacture of fresh acid-curd cheeses such as Quarg, Labneh, Cream cheese, Cottage cheese, the gel, following incubation, is concentrated (i.e., removal of whey) by cutting, stirring, cooking, whey drainage, and/or mechanical centrifugation. Further syneresis of the product may occur during storage, depending on processing conditions and product composition, which affect structure and porosity, and on the absencelpresence of hydrocolloids which bind the aqueous phase. However, for other fresh cheese-type products, such as Laban, Fromage frais, set yogurt, and fresh cheeses made by recombination technology (i.e., reconstituting various dairy ingredients, e.g., skim milk powder and cream, in milk andlor water and standardizing the blend to the desired product composition prior to culturing), the gel, which is packaged, is the final product and hence whey removal is not practiced. In both classes of product, spontaneous syneresis, following packaging, is undesirable but occurs frequently because of the relatively high moisture to protein ratio compared to rennetcurd cheeses (e.g., -17.6 g HzO/g protein in yogurt versus -1.44 g H20/g protein in Cheddar cheese). Acid milk gels formed in the package, such as set natural yogurt, show little tendency to syneresis if left undisturbed. However, even in this situation, spontaneous syneresis may occur with time to a greater or lesser extent depending on the level of fortification and processing conditions (such as preheating of milk which affects differences in porosity and gel structure). This may be due partly to slow proteolysis of the casein, as affected by enzymes of starter bacteria, pH decrease, and temperature fluctuations. Hydrolysis of casein may be responsible for the widely different practical experience (day-to-day in-factory and interfactory inconsistencies) regarding syneresis in fermented set milk products (Walstra et al., 1985). Disturbance of these products, e.g., by movement during cartoning and transport, which provides stress for bond breakage and matrix rearrangement, may initiate or accentuate syneresis. For a given level of syneretic pressure, syneresis increases with increasing gel surface area to volume ratio. The shape of the package containing the gel may also influence syneresis; for example, in a package with sloping
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS 191
walls, the gel may have a tendency to detach from the walls, leading to a stress in the gel which induces syneresis (Walstra et al., 1985). Increasing the level of total solids and gel-forming protein results in gels which are less susceptible to syneresis (Emmons et al., 1959; Modler et al., 1983). Harwalkar and Kalab (1983) found an inverse relationship between the level of total solids and susceptibility to syneresis for natural yogurt (10-15% DM); however, the relationship differed for yogurts from heated (90°C X 10 min) and unheated milks. As inverse relationship between gel firmness and susceptibility to syneresis has been found for natural yogurt produced from reconstituted skim milk (Modler et al., 1983;Harwalkar and Kalab, 1983), Cottage cheese gels containing 8 to 15% solids (Emmons et al., 1959), and chemically acidified skim milk gels (Harwalkar and Kalab, 1981).The decrease in syneresis with increasing level of gel-forming protein may be attributed to the formation of denser (greater number of strands of a given thickness per unit volume), less porous gels. However, for a given level of gel-forming protein, the susceptibility to syneresis strongly depends on the proportions of casein and whey protein (Modler et al., 1983). Spontaneous syneresis of acid skim milk-based gels decreases with increasing severity of heat treatment, homogenization, and lower gelation temperatures (Harwalkar and Kalab, 1981, 1983; Schellhaass and Morris, 1985; Parnell-Clunies et al., 1986a,b). Reducing the pH of Cottage cheese gels (9-11% DM) at cutting from 4.92 to 4.59 reduced the level of syneresis on holding the cut gel for 1 hr at 32°C (Emmons et al., 1959). A decrease in pH during syneresis results in greater syneresis than if the gel is brought to the same pH before cutting (i.e., before initiating external syneresis) (Walstra et al., 1985). A wide variety of hydrocolloids (including gelatin, pregelatinized starch, cellulose derivatives, alginates, and carageenans) are used in practice to immobilize water and reduce syneresis in fresh acidcurd products, especially yogurts. Their effects on yogurt quality have been studied extensively (Kalab et al., 1975; Modler et al., 1983). The use of slime-producing cultures in yogurt has also been found to reduce syneresis considerably (Schellhaass and Morris, 1985). c. Further Treatments of the Gel. In the production of many fresh cheese products, the gel produced following acidification is subjected to a number of further processing steps such as stirring, whey separation/ concentration, heating, homogenization, agitation, and cooling (Fig. 2). Various materials, such as cream, sugar, salt, fruit purees, and/or hydrocolloids, may be added to the curd. Such treatments influence the structural, rheological, and syneretic properties of the final product, as discussed below. Cutting of the gel into cubes, as in Cottage cheese, initiates syneresis which is enhanced by cooking and stirring, as in the manufacture of rennet-
192
P. F. FOX er al.
curd cheese. Stirring of the gel (as in Quarg, Cream cheese, Fromage frais) breaks the matrix strands to an extent depending on the severity of agitation. Cooling of the gel to temperatures <20"C, to retard a further decrease in pH (before heating and whey separation), may result in greater destruction of the gel for a given degree of agitation as the strength of hydrophobic bonds, which play an important role in the structure of acid-curd cheeses, decreases with decreasing temperature (Kinsella, 1984; Hayakawa and Nakai, 1985). Increasing the temperature, in the range 25-85"C, enhances whey separation. Any factor which increases the firmness of the gel at separation (such as proximity to the isoelectric pH, rennet addition, milk homogenization, higher level of gel-formingprotein, increased temperature of gelation) makes it less susceptible to breakage for a given shear. Whey separation, which may be performed by pouring the hot fluid onto cheesecloths, ultrafiltration, or centrifugation, results in concentration and aggregation of the broken pieces of gel to a greater or lesser degree. Collision during concentration may be expected to result in the formation of large irregularly shaped conglomerates (of varying thickness and length) which are forced into close proximity. The moisture content of the curd is closely related to the degree of aggregation; factors which enhance aggregation (conditions that promote a coarser gel structure, increased separating temperature, and increased syneretic force during separation) reduce the water content and increase the coarseness and firmness of the resulting curd. Homogenization and/or shearing results in destruction, to an extent depending on the magnitude of the shear, of conglomerates and thereby yield a more homogeneous size and spatial distribution of the matrix-forming material (Kalab and Modler, 1985; Modler et al., 1989). Holding hot cream cheese, while shearing, at 75 to 85°C may result in a thickening of the consistency,which is rather similar to the "creaming" process in pasteurized processed cheese products. However, while imparting a greater elastic character to the product, it may also lead to grittiness (Modler et al., 1989), an effect which may be attributed to protein dehydration and the consequent formation of compact protein conglomerates. Slow cooling probably accentuates this defect. The degree of matrix reformation during cooling of the homogenized hot-packed product is uncertain; the matrix of the cooled cream cheese is more or less continuous, with the degree of continuity being governed by the size and spatial distribution of the matrix-forming material and the rate of cooling (Kalab and Modler, 1985; Modler et al., 1989). A finer matrix manifests itself in a product which has a smoother appearance and mouthfeel and which is less susceptible to spontaneous wheying off on storage. The addition of hydrocolloids to the curd also minimizes syneresis; stabilizers which interact with casein, particularly K-
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS 193
carrageenan, may interrupt matrix formation and yield a smoother, softer product. C. ULTRAFILTRATION TECHNOLOGY IN CHEESEMAKING Ultrafiltration (UF), as a technology for cheese manufacture, was introduced in the early 1970s and has been investigated extensively and reviewed (Zall, 1985; Ernstrom, 1985; Ottosen, 1988; Lelievre and Lawrence, 1988; Lawrence, 1989; Spangler et al., 1991). It has attracted the attention of cheese and equipment manufacturers, primarily because of the potential to increase yield, through the recovery of whey proteins in the cheese. Other advantages include its potential to reduce production costs and to produce new cheese varieties with different textural and functional characteristics. In this section, some of the more important aspects of UF in cheesemaking are highlighted. The most successful commercial applications of UF in cheese manufacture to date have been in the production of cast Feta in Denmark (Tamime and Kirkegaard, 1991), fresh acid-curd varieties (Quarg, Ricotta and Cream cheeses) in Germany and other European countries (Pedersen and Ottosen, 1992), and the standardization of milk protein, to 4-5%, for the production of Camembert and other varieties (Coton, 1986; Korolczuk et al., 1987; Puhan, 1992). Based on the degree of concentration and whether whey expulsion following concentration is neccessary, UF in cheesemaking may be classified into three general areas: (i) Low concentration factor (LCF), followed by cheesemaking and whey removal using conventional equipment. The main application of LCFUF is the standardization of milk to a fixed protein level to obtain a more consistent end-product; variations in gel strength at cutting, buffering capacity, and rennet-to-casein ratio are minimized. However, when using conventional cheesemaking vats, concentration appears limited to a maximum CF of -1.5, or 4-5% protein, because of difficulties in handling the curd and yield losses (Green et al., 1981a; Guinee et al., 1994). (ii) Medium concentration factor (2-6X) to the final solids content of the cheese without whey expulsion. The main attraction of this type of UF technology is the increased cheese yield associated with retention of whey proteins and increased moisture when whey proteins are denatured prior to UF. The main commercial application is in the production of highmoisture cheeses which are consumed fresh (e.g., Quarg, Cream cheese) or are not very dependent on proteolysis during ripening for flavor development (e.g., Feta). Feta produced by this method (by addition of rennet to
194
P. F. FOX et al.
a concentrate, i.e., precheese, without curd cutting) has a smoother more homogeneous texture than the more “curdy-textured” traditional product, hence the name “cast” Feta. There are numerous reports on the use of UF concentration to the final cheese dry matter level for the production of soft or semihard rennet-curd cheeses, including Camembert, Blue, Havarti, and Mozzarella (Glover, 1985; Qvist et al., 1987; Mohr et al., 1989; Lawrence, 1989). Manufacture essentially involves preacidification, ultrafiltration/diafiltration, starter addition, rennet addition, coagulation and automated curd cutting using specialized equipment (e.g., Al-Curd or Ost Retentate coagulators), moulding, pressing, and brining. To date, uptake of UF technology by the industry for the production of the latter cheeses has been limited; apart from the uncertainties concerning the regulatory status of such cheeses and the relatively low reported increases in yield, the main drawbacks include changes in cheese texture, flavor, and functionality (Le., meltability and stretchability). (iii) High concentration factor, followed by whey expulsion in novel equipment. Since the upper limit of concentration by UF is -7 :1 for whole milk, it is not possible to achieve the dry matter levels required for hard cheeses such as Cheddar and Gouda; hence, further whey must be expelled following coagulation of the retentate and cutting the coagulum. Owing to the high curd-to-whey ratio, efficient curd handling (i.e., stirring and heat transfer) is not feasible in conventional systems.The only continuous system capable of handling such concentrates is the Siro-Curd which has been used for the production of Cheddar cheese in Australia since 1980 (Anonymous, 1986). The cheese produced by this process, which gives a yield increase of 4 to 6%, is claimed to be indistinguishable from Cheddar manufactured using standard equipment. On renneting at a fixed dosage level, increasing milk protein level results in a reduced rennet coagulation time, an increase in the level of soluble (nonaggregated) casein at the point of gelation, increased rate of curd firming, reduced set-to-cut time when cutting at a given curd strength, a decrease in the degree of aggregation at cutting, and a coarser gel network (Dalgleish, 1981; Guinee et al., 1992,1994; McMahon et al., 1993). Micelles which are not modified, or aggregated, at the onset of gelation are presumably modified later and incorporated into the gel to a greater or lesser degree. Owing to the rapid rate of curd firming, it becomes increasingly difficult, as the milk protein level is increased, to cut the curd cleanly, without tearing, before the end of the cutting cycle (Guinee et al., 1994). Reflecting the tearing of curd, and consequent shattering of curd particles, fat losses in the whey are greater than those predicted on the basis of volume reduc-
-
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS 195
tion (due to UF) for milks with protein concentrations >5%. Similar findings by Green et al. (1981a,b) were attributed partly to the poorer fat-retaining ability of higher protein curds which had coarser, more porous protein networks. Reduction of the setting temperature, in the range 31 to 27T, and the level of rennet added give set-to-cut times and curd firming rates for concentrated milks closer to those of the control milk (Guinee et al., 1994). Increasing the concentration of protein in the cheesemilk also results in slower proteolysis during ripening when equal quantities of rennet on a milk volume basis are used (Green er al., 1981a; Green, 1985; Creamer et al., 1987; Spangler et al., 1991). The slower rate of proteolysis in cheeses made from ultrafiltered milks may be attributed to a number of factors, including: (i) the lower effective rennet concentration, i.e., rennet-to-casein ratio, and hence activity in the cheese (Green et al., 1981a; Creamer et al., 1987); (ii) the inhibition of the indigenous milk proteinase, plasmin, by retained P-lactoglobulin in cheeses containing a significant quantity of whey proteins (Qvist et al., 1987);(iii) the concentration, during UF, of proteinase/ peptidase inhibitors (Hickey et al., 1983b), and/or (iv) the resistance of undenatured whey proteins to degradation in cheese where they represent a substantial portion of the protein, i.e., -18.5% (de Koning et al., 1981). However, Creamer et al. (1987) found that at equal rennet-to-casein ratios, the level of aSl-caseinhydrolysis was higher in control Cheddar cheese than in that made from milk concentrated fivefold by UF. The reduced surface area-to-volume (SA/V) ratio of the protein network in cheeses made from concentrated milks, resulting from their coarser networks (Green et al., 1981b), may also contribute to the observed reduction in proteolysis (Guinee er al., 1992). It is conceivable that for a given level of enzyme activity in the cheese curd, casein degradation decreases as the S A N ratio of the matrix decreases. Cheese becomes progressively firmer (i.e., requires a higher compression force to induce fracture), more cohesive, mealier, and drier, and the structure of the protein matrix becomes coarser and more compact (fused), with increasing concentration factor. The reduced rates of protein degradation result in slower softening and flavor development during maturation. IV. BIOCHEMISTRY OF CHEESE RIPENING
The conversion of milk to cheese curd is only the first stage in the production of most cheese varieties. Essentially all hard and most soft cheeses are ripened for periods ranging from a few weeks to 2 years or longer. During this period, cheeses undergo numerous biochemical changes which lead to the development of the appropriate flavor and aroma. The
196
P. F. FOX
et
al.
biochemistry of cheese ripening, which is very complex, involves three primary processes, glycolysis, lipolysis, and proteolysis, the relative importance of which depends on the variety, and numerous secondary reactions which are not well understood and may be mainly responsible for the finer points of cheese flavor. A. CHEESE RIPENING AGENTS: ASSESSMENT OF CONTRIBUTION TO RIPENING Ripening agents in cheese generally originate from five sources: the coagulant, the milk, starter bacteria, secondary or adjunct starter bacteria, and nonstarter bacteria. The residual coagulant and enzymes from the starter, and probably the nonstarter microflora, are common to nearly all ripened cheeses. The secondary starter (i.e., microorganisms added to cheesemilk for purposes other than acidification) can exert considerable influence on maturation in cheese varieties in which they are used (e.g., Penicillium roquefurti, P. camemberti in mould-ripened varieties or Brevibacterium linens in smear-ripened cheeses). Exogenous enzymes used to accelerate ripening could be added to the above list and, when present, can be very influential. The role of the individual ripening agents in cheese maturation has been studied using model cheese systems in which the action of one or more of the above agents is eliminated. The most comprehensive studies of this nature involved the manufacture of cheese under controlled microbiological conditions, with inactivation of the rennet andlor chemical acidification, allowing the contribution of individual ripening agents to be studied. A milk supply of exceptionally high bacteriological quality is essential to eliminate nonstarter bacteria. Sterile teat cups, clusters, and bucket milking plant have been used by a number of authors (Perry and McGillivray, 1964; O'Keeffe et at., 1976a) to obtain milk with an initial bacterial count of ca. lo2cfu ml-'. Kleter and de Vries (1974), who included a cooling coil between the cluster and bucket, obtained counts averaging 46 CFU ml-'. Reiter et al. (1969) withdrew milk aseptically by using a teat cannula but the quantities obtained were sufficient to produce only very small cheeses (ca. 100 g). Heat treatment of aseptically drawn milk is necessary to further reduce bacterial counts. Perry and McGillivray (1964) used batch pasteurization (68°C X 5 min) in a steam-jacketed cheese vat. LTLT pasteurization (63°C X 30 min) was also used by Reiter et al. (1969) and O'Keeffe et al. (1976a). HTST pasteurization was used by Chapman et al. (1966), Visser (1976, 1977a), Kleter (1976), and Paulsen et al. (1980). UHT-treated milk was used by Le Bars et al. (1975); CaC12,more rennet, and a higher setting
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
197
temperature were used to offset the ill-effects of the high-heat treatment on the rennet coagulability of milk. Turner et al. (1986) concluded that a heat treatment of 83°C X 15 sec or 72°C X 58 sec is necessary to ensure a reduction of lo8, which was deemed necessary to produce cheese with nonstarter counts 4 0 CFU kg-' cheese from milk if the initial count is lo3 CFU ml-'; HTST pasteurization (72°C X 15 sec) is sufficient for milk with an initial count of 10 CFU mi-'. The manufacture of cheese under aseptic conditions can be achieved using enclosed cheese vats (Mabbitt et al., 1959;Kleter, 1976; Visser, 1977a; Paulsen et al., 1980), a sterile room with a filtered air supply (Le Bars et al., 1975), or a laminar air-flow unit (O'Keeffe et al., 1976a,b; McSweeney et al., 1994a); the latter is probably the simplest of these techniques. The acidifying role of starter can be simulated closely using an acidogen, usually GDL (O'Keeffe et al., 1975), although the rate is faster than that which occurs in biologically acidified cheese. To determine the role of the coagulant in cheese ripening, it is necessary to inactivate the rennet after coagulation, for which three techniques have been developed. One approach (Visser, 1976) involved separation of the first and second stages of rennet action: milk, depleted of Ca2+and Mg2+by treatment with an ion-exchange resin, was renneted (but did not coagulate), heated (72°C X 20 sec) to inactivate the rennet, and cooled to <5"C and CaC12 was added; the milk was then heated dielectrically and allowed to coagulated. O'Keeffe et al. (1977) used porcine pepsin as coagulant. This was inactivated after coagulation by adjusting the pH of the curd-whey mixture to 7.0. Mulvihill et al. (1979) used piglet gastric proteinase to prepare rennet-free curd in small-scale cheesemaking trials; the enzyme hydrolyzes bovine K-casein but has little activity on as'-or &casein. The noncompetitive plasmin inhibitor 6-aminohexanoic acid (AHA) was used by Farkye and Fox (1991) to study the significance of plasmin in cheese ripening. It was necessary to use a high concentration of AHA which affected curd syneresis and the moisture content of the cheese; also, since AHA contains N, the background level of soluble N was increased greatly. Plasmin is inhibited by several proteins, including soybean trypsin inhibitor, which may be suitable as a plasmin inhibitor in cheese. Plasmin is also specifically inhibited by dichloroisocoumarin (Harper et al., 1985), but neither it nor the inhibitory proteins have been investigated in cheesemaking. The high heat stability of plasmin [its activity is increased by high cooking temperatures (Farkye and Fox, 1990)] suggests that it may be possible to develop a model system based on aseptic curd in which the rennet is denatured by a suitable cooking temperature and acidified by GDL, in which to assess plasmin activity in isolation.
198
P. F. FOX
el al.
B. METABOLISM OF LACTOSE AND LACTATE DURING RIPENING The primary glycolytic event, the conversion of lactose to lactate, is normally mediated by the starter culture during curd preparation or the early stages of cheese ripening. In cases where glycolysis has not been completed by the starter, nonstarter lactic bacteria may contribute. The metabolism of lactose was discussed in Section IIIAS. Although 98%of the lactose in milk is removed in the whey as lactose or lactate (Huffman and Kristoffersen 1984), Cheddar curd at milling typically contains 0.8 to 1.0% lactose (Turner and Thomas, 1980; Huffman and Kristoffersen, 1984). The residual lactose is fermented relatively rapidly to an extent dependent on the percentage salt-in-moisture (S/M) in the curd (Turner and Thomas, 1980; Thomas and Pearce, 1981). At low S/M concentrations and low populations of NSLAB, residual lactose is converted mainly to L-lactate by the starter. At high populations of NSLAB, e.g., at high storage temperatures, considerable amounts of D-lactate are formed, partly by fermentation of residual lactose and partly by isomerization of L-lactate (Turner and Thomas, 1980). At high S/M levels (e.g., 6%), and low NSLAB populations, the concentration of lactose decreases slowly and changes in lactate are slight. The quality of cheese is strongly influenced by the fermentation of residual lactose (O’Connor, 1974): the pH decreases after salting at S/M levels <5%, presumably due primarily to the continued action of the starter, but at higher levels of S/M, starter activity decreases abruptly, as indicated by high levels of residual lactose and high pH, accompanied by a sharp decrease in cheese quality. Dutch-type cheese contains -1.4% lactose at pressing but this decreases to <0.1% after pressing and to undetectable levels after brining (Raadsveld, 1957). Typical levels of lactate in Camembert, Swiss, and Cheddar are 1.0, 1.4, and 0.5%, respectively (Karahadian and Lindsay, 1987; Turner et al., 1983; Turner and Thomas, 1980). The fate of lactic acid in cheese depends on the variety. Initially, Cheddar contains only L(+) lactic acid but as the cheese matures, the concentration of D-lactate increases. The latter could be formed from residual lactose by lactobacilli (Turner and Thomas, 1980 Thomas and Pearce, 1981; Tinson et al., 1982) or by racemization of Llactate by NSLAB. Except in cases where the post-milling activity of the starter is suppressed (e.g., by S/M > 6%), racemization is likely to be the principal mechanism (Thomas and Crow, 1983). Racemization of L-lactate, which appears to occur in several cheese varieties (Thomas and Crow, 1983), is probably not significant from the flavor viewpoint. However, calcium Dlactate may crystallize on the surface of cheese, causing undesirable white specks (Pearce et al., 1973; Severn et aL, 1986; Dybing et al., 1988).
-
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
199
Lactate in cheese may be oxidized to acetate. Pediococci produce 1 mol of acetate and 1 mol of C 0 2 and consume 1 mol of O2per mole of lactate utilized (Thomas et al., 1985). The concentration of lactate in cheese far exceeds that required for optimal oxidation, and lactate is not oxidized until all sugars have been exhausted. The oxidation of lactate to acetate in cheese depends on the NSLAB population and on the availability of 02, which is determined by the size of the block and the oxygen permeability of the packaging material (Thomas, 1987). Acetate, which may also be produced by starter bacteria from lactose (Thomas et at., 1979) or citrate or from amino acids by starter bacteria and lactobacilli (Nakae and Elliott, 1965), is usually present at fairly high concentrations in Cheddar cheese and is considered to contribute to cheese flavor, although high concentrations may cause off-flavors (see Aston and Dulley, 1982). Thus, the oxidation of lactate to acetate probably contributes to Cheddar cheese flavor. Presumably, the oxidation of lactate to acetate also occurs in other hard and semihard cheeses but studies are lacking. Production of lactate in Romano cheese was monitored by Mora et al. (1984). As with other varieties, L-lactate predominated initially, reaching a maximum of 1.9%at 1 day (Deiana et al., 1984). The concentration began to decrease at 10 days and was 0.2 to 0.6% at 150 to 240 days. Some of the decrease was accounted for by racemization to D-lactate, which reached a maximum at -90 day (up to 0.6%in some cheeses) and then declined somewhat. In some cheeses, acetate reached very high levels (1.2%) at -30 days, but decreased to 10.2% at 90 days; the agents responsible for the metabolism of acetate were not identified, but yeasts (Debaryomyces hansenii) may have been involved. The metabolism of lactate is very extensive in surface mould-ripened varieties, e.g., Camembert and Brie. The concentration of lactic acid in these cheeses at 1 day is -l.O%, produced mainly or exclusively by the mesophilic starter, and hence is L-lactate. Secondary organisms quickly colonize and dominate the surface of these cheeses-first Geotrichum candidum and yeasts, followed by Penicillium caseicolum, and, in traditional manufacture, by Brevibacterium linens. G. candidum and P. caseicolum rapidly metabolize lactate to C 0 2 and H20, causing an increase in pH. Deacidification occurs initially at the surface, resulting in a pH gradient from the surface to the center and causing lactate to diffuse outward. When the lactate has been exhausted, P. caseicolum metabolizes proteins, producing NH3, which diffuses inward, further increasing the pH. The concentration of calcium phosphate at the surface exceeds its solubiIity at the increased pH and it precipitates as a layer of Ca3(P04)2at the surface, thereby causing a calcium phosphate gradient within the cheese. The elevated pH stimulates the action of plasmin, which, together with residual coagulant, is largely responsible for proteolysis. Although proteinases se-
-
200
P. F. FOX et al.
creted by the surface microorganisms are very potent, they diffuse into the cheese to only a very limited extent; however, peptides produced from surface proteins may diffuse into the cheese. The combined action of increased pH, loss of calcium (necessary for the integrity of the protein network), and proteolysis are necessary for the very extensive softening of the body of Brie and Camembert. B. linens does not grow at pH < 5.8 and does not colonize the cheese surface until the pH has increased (Lenoir, 1984; Karahadian and Lindsay, 1987). The metabolism of lactose and lactate in Swiss-typecheeses was described comprehensively by Turner ef al. (1983). Typically, Emmental cheese contains -1.7% lactose 30 min after moulding, which is rapidly metabolized by S. thermophilus with the production of L-lactate. Only the glucose moiety of lactose is metabolized by S. fhermophilus and consequently galactose accumulates to a maximum of -0.7% at -10 hr, when the lactobacilli begin to multiply. These metabolize galactose to a mixture of D- and L-lactate, which reach -0.35 and 1.2%, respectively, at 14 days, when the galactose is metabolized completely. On transfer to the warm room, Propionibacterium metabolize lactate, preferentially the L-isomer (Crow, 1986), to propionate, acetate, and C 0 2
H
I
3 CH3-C-COOH + 2 CH3CH2COOH + CH3COOH + CO2 + H20
I
OH The C 0 2 generated is responsible for eye development, a characteristic feature of these varieties. The significance of the primary fermentation of lactose to L-lactate is well recognized in cheese manufacture (see Section IIIA5). However, less significance has been attached to the subsequent changes in lactose, and lactate metabolism received relatively little attention in most varieties until recently. However, these latter changes are of major proportions; they are critical in some varieties, e.g., Swiss and Camembert, and are probably of some importance in all varieties. C. CITRATE METABOLISM The relatively low concentration of citrate in milk (-8 mM) belies the importance of its metabolism in some cheeses made using mesophilic cultures (for reviews, see Cogan, 1985; Cogan and Hill, 1993). Citrate is not metabolized by L. lactis or L. cremoris but is metabolized by L. lactis subsp.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
201
diacetyfactis and Leuconostoc spp. with the production of diacetyl and C02. It is not metabolized by S. thermophifus or by thermophilic lactobacilli (Hickey et af.,1983a),but several species of mesophilic lactobacilli metabolize citrate with the production of diacetyl and formate (Fryer, 1970). Citrate is not used as an energy source by L. factis subsp. diacetyfactis or Leuconostoc spp., but is metabolized very rapidly in the presence of fermentable carbohydrate. C 0 2 produced from citrate is responsible for the characteristic eyes of Dutch-type cheese and for the undesirable openness and floating curd in Cheddar and Cottage cheeses, respectively. Diacetyl is very significant in the aromalflavor of Cottage cheese, Quarg, and many fermented milks. It also contributes to the flavor of Dutch-type cheeses and possibly of Cheddar (Manning, 1979a,b;McGugan, 1975;Aston and Dulley, 1982). Acetate may also contribute to cheese flavor. Approximately 90% of the citrate in milk is soluble and is lost in the whey: however, the concentration of citrate in the aqueous phase of cheese is approximately three times that in whey (Fryer er af.,1970), reflecting the concentration of colloidal citrate. Cheddar cheese contains 0.2 to 0.5% (w/ w) citrate which decreases to 0.1% at 6 months (Fryer et af., 1970; Thomas, 1987). Inoculation of cheesemilk with Lb. pfantarum accelerated the depletion of citrate; pediococci did not appear to utilize citrate (Thomas, 1987).
D. LIPOLYSIS Lipases in cheese originate from milk, rennet preparation (paste), starter, adjunct starter, or nonstarter bacteria. The degree of lipolysis in cheese varies widely between varieties, from -6 meq free fatty acids in Gouda to -45 meq/100 g fat in Danish Blue (Gripon, 1987,1993).Lipolysis in internal bacterially ripened varieties, such as Gouda, Cheddar, and Swiss, is generally low but is extensive in mould-ripened and certain Italian varieties. In general, in those varieties in which extensive lipolysis occurs, lipases originate from the coagulant (rennet paste, which contains pregastric esterase, as used for some Italian varieties) or from the adjunct starter [Penicilfium spp., which produce a number of lipases (Gripon, 1987, 1993), in mouldripened varieties]. 1. Indigenous Lipases
Milk contains substantial amounts of an indigenous lipoprotein lipase (LPL) which is well characterized (Olivecrona and Bengtsson-Olivecrona, 1991; Olivercrona et af., 1992). The physiological role of LPL is in the metabolism of plasma triglycerides and, although it is generally believed that LPL occurs in milk as a result of leakage, it may have a function
202
P. F. FOX et al.
in milk (see Olivecrona and Bengtsson-Olivecrona, 1991). LPL is rather nonspecific for the fatty acid but is specific for the Snl and Sn3 positions of mono-, di-, and triglycerides and the 1 position of glycerophospholipids. Therefore, lipolysis in milk leads to preferential release of short and medium chain acids, which in milk triglycerides are esterified predominantly at the Sn-3 position. In bovine milk, more than 80% of the LPL is associated with the casein micelles (Olivecrona et al., 1992) and is incorporated into cheese curd. LPL probably causes significant lipolysis in raw milk cheese and may also contribute to lipolysis in pasteurized milk cheese as heating 278°C X 10 sec is required for complete inactivation of LPL (Driessen, 1989). 2. Lipases from Rennet Rennet extract should contain no lipolytic activity. Rennet pastes used in the manufacture of hard Italian varieties (e.g., Romano, Provolone) contain a potent lipase, pregastric esterase (PGE), which is responsible for the extensive lipolysis and the characteristic “piccante” flavor in such varieties. The literature on PGE was comprehensively reviewed by Nelson et al. (1977). PGE, also called lingual or oral lipase, is secreted by glands at the base of the tongue. Its secretion is stimulated by suckling and it is subsequently washed into the abomasum by milk and siliva. Rennet paste is prepared from the abomasa of calves, kids, or lambs slaughtered immediately after suckling. The abomasa are partially dried and ground into a paste which is slurried with milk or water before being added to cheesemilk. Rennet pastes are considered unhygienic and their use is not permitted in several countries; instead, partially purified PGEs are used. Calf, kid, and lamb PGEs were partially purified from commercial preparations (Lee et al., 1980) and calf PGE from oral tissue (Sweet et al., 1984; Bernback et al., 1985). The enzyme appears to be a glycoprotein with a p l of 7.0 and a mol wt of about 49 kDa. The gene for rat lingual lipase has been cloned and sequenced and the primary structure of the enzyme deduced (Docherty et af., 1985). PGE is highly specific for short chain acids esterified at the Sn-3 position (Nelson et al., 1977) and therefore releases high concentrations of short and medium chain acids from milk fat. Slight specificity differences between calf, lamb, and kid PGEs allow the manufacture of Italian cheeses with slightly different flavor characteristics. Most other lipases are unsuitable for the manufacture of Italian cheeses because of incorrect specificity but it has been claimed that certain fungal lipases may be acceptable alternatives (see Fox, 1988/1989; Fox and Stepaniak, 1993). The use of PGE to accelerate the ripening of other cheese varieties was discussed by Fox (1988/1989).
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
203
3. Microbial Lipases Lactococcus spp. and Lactobacillus spp. have low lipolytic activities compared to other genera of bacteria (e.g., Pseudomonas). Fryer et al. (1967) considered that, although weakly lipolytic, lactococci will hydrolyze milk fat to a significant extent if present at high numbers for long periods (e.g., during cheese ripening). Small aseptic cheeses acidified with GDL instead of starter had lower free fatty acid (FFA) concentrations which did not increase during ripening (Reiter et al., 1967). Umemoto et al. (1968) found that the cell-free extracts of various dairy lactic acid bacteria were most active on tributyrin at pH between 6 to 8 and at 37°C. Singh et al. (1973) found intracellular, but no extracellular, lipases in L. lactis ssp. luctis and L. luctis ssp. cremoris which were more active on tributyrin than on tripalmitate or triolein. Harper el al. (1980) reported that L. luctis ssp. lactis had a more complex esterase system than L. lactis ssp. cremoris; the former showed two bands on a zymogram stained with a-naphtyl acetate, while the latter showed only one. In a comparative study of the lipase activity of a number of strains of lactic acid bacteria, including L. luctis ssp. lactis biovar. diacetylactis, Singh et al. (1981) found that mutants prepared by UV treatments produced more long chain acids than the parent strains. Piatkiewicz (1987) found interstrain differences in the lipolytic and esterolytic activity of Lactococcus spp. which were influenced by the composition of the growth medium and the physiological age of the culture; he also reported that lactococci were more lipolytic than lactobacilli. Kamaly et al. (1990) studied the lipolytic activity of a number of strains of Lactococcus. Lipases, in cell-free extracts, were most active at 37°C and at pH 7 to 8.5 (on tributyrin) or 7.0 (milk fat emulsion) and were more active on triglycerides containing short chain acids (Cdz0 to Clo:o)than long chain acids (Clz:oto C18:1).Starter bacteria can liberate FFA from mono- and diglycerides in milk produced by other lipases, e.g., LPL or lipases from Gram-negative bacteria (Stadhouders and Veringa, 1973). Kamaly et al. (1988) found quantitative interstrain differences in the esterolytic activity of lactococci assayed on p - and o-nitrophenyl esters of fatty acids. The esterase/lipase system of Luctococcus has received relatively little attention in comparison with its proteolytic system. Unlike the situation with lactococcal proteinases and peptidases, little is known about the genetics of lactoccal lipase/esterase. Isolation of lipase/esterase-negative variants of Lactococcus would permit the significance of these enzymes in cheese ripening to be assessed. The lipolytic/esterolytic activity of Lactobacillus has also received limited attention. An esterase of Lb. plantarum was purified by Oterholm et al. (1972) who found that the enzyme was maximally active at pH 6.7 and
204
P. F. FOX et al.
40°C (on triacetin). The enzyme was not affected by heavy metals or low concentrations of cyanide or azide but was inhibited by higher concentrations; it preferentially hydrolyzed acetylesters and exhibited a strong preference for soluble over emulsified substrates. The intracellular esterases of Lb helveticus, Lb. delbrueckii ssp. bulgaricus, and Lb. lactis were studied by El Soda et al. (1986) on 0- and p nitrophenyl derivatives of fatty acids. The optimum temperature for esterase production was 40 to 45°C and the cells had little esterolytic activity if grown at 35 or 55°C. The esterases of these strains were generally specific for short chain acids. Piatkiewicz (1987), who investigated the lipase and esterase activities of Lb. casei, found that cells had higher activities in the logarithmic than in the stationary phase of growth. Khalid et al. (1990) reported that Lb. helveticus CNRZ 32 (a strain thought to have potential in accelerated cheese ripening), Lb. helveticus ATCC10797, and especially Lb. delbrueckii ssp. bulgaricus, possessed esterolytic activity. Thus, several Lactobacillus spp, the principal nonstarter bacteria in Cheddar and Dutchtype cheeses, possess both lipolytic and esterolytic activity but none of these enzymes has been isolated and fully characterized. Other nonstarter bacteria (e.g., Micrococcus and Pediococcus) also produce lipases. It is generally believed that lipases from Micrococcus spp., when present in cheese, can contribute to lipolysis during ripening (Bhowmik and Marth, 1990b).The lipase of M . freudenreichii was strongly inhibited by organophosphates and divalent metal ions, but less so by EDTA or pCMB (Lawrence et al., 1967). Bhowmik and Marth (1990a) studied the esterases in cell-free extracts of five strains of Micrococcus, all of which showed esterolytic activity; most strains hydrolyzed p-nitrophenyl derivatives of fatty acids faster than onitrophenyl derivatives. The esterase of Micrococcus spp. ATCC 8459 was studied in detail; it was optimally active at pH 8.0 and 40°C and was inhibited by organophosphates, divalent metal ions, NaCI, and redox reagents. The lipase/esterase of Pediococcus spp. has received little attention. Tzanetakis and Litopoulou-Tzanetaki (1989) found only weak esterase and lipase activities in a number of strains of P. pentosaceus of dairy origin by means of the API-ZYM system. Bhowmik and Marth (1989) found esterase activity in six strains of P. pentosaceus but none in two strains of P. acidilactici. The lipases of Propionibacteriurn sherrnanii studied by Oterholm et al. (1970) were optimally active at pH 7.2 and 47°C on tributyrin; the enzymes showed a high preference for tripropionate and tributyrin and were inhibited by Hg2+and Na2HAs04but not by pCMB or EDTA. Some esterase activity was observed but the enzyme was more active on emulsified than on soluble substrates.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
205
Lipases and esterases of Brevibacterium linens were described by Sgrhaug and Ordai (1974) and Foissy (1974). Sgrhaug and Ordal(l974) found only intracellular lipase and esterase activities in Br. linens. Most of the strains studied were more active on tributyrin than on triacetin or methyl butyrate. According to Foissy (1974), 15 Br. linens isolates possessed intracellular esterase(s); they also possessed weak extracellular activity, which, based on zymogram patterns, may have been of intracellular origin. Extensive lipolysis occurs in mould-ripened cheese, particularly blue varieties. In some cases, up to 25% of the total FFA may be liberated (see Gripon, 1987, 1993). However, the impact of FFA on the flavor of blue mould-ripened cheeses is less than in hard Italian varieties, possibly due to neutralization as the pH increases during ripening and to the dominant influence of methyl ketones on the flavor of blue cheese. Lipolysis in mould-ripened varieties is due primarily to the lipases of Penicillium roqueforti or P. camemberti, which secrete potent extracellular lipases. Penicillium lipases are well characterized (see Kinsella and Hwang, 1976; Gripon, 1987, 1993).P. camemberti appears to excrete only one lipase which is optimally active at ca. pH 9.0 and at ca. 35°C. P. roqueforti excretes two lipases, one with a pH optimum at 7.5 to 8.0 (or perhaps 9.0 to 9.5), the other at pH 6.0 to 6.5 The acid and alkaline lipases exhibit different specificities. Lipases of Geotrichum candidum have been studied by Sidebottom et al. (1991) and Charlton et al. (1992). This organism produces two lipases of different substrate specificities. Psychrotrophs, which can dominant the microflora of refrigerated milk, are a potentially important source of potent lipases in cheese. Cousins et al. (1977) considered that active lipase would be present in cheese if psychrotroph numbers exceed lo7 CFU ml-I. Many psychrotroph lipases are heat stable and thus may cause rancidity in cheese over the course of a long ripening period. The subject of psychrotroph enzymes in cheese was discussed by Mottar (1989). Unlike psychrotroph proteinases, which are largely water-soluble and are lost in the whey, psychrotroph lipases adsorb onto the fat globules and are therefore concentrated in the cheese. 4.
Pattern and Levels of Lipolysis in Selected Cheeses
Lipolysis is considered to be undesirable in most cheese varieties. Cheddar, Gouda, and Swiss-type cheeses containing even a moderate level of free fatty acids would be considered rancid; however, certain cheese varieties are characterized by extensive lipolysis (e.g., Romano, Parmesan, and Blue cheeses). Bills and Day (1964) quantified FFA (G:o to C18:3)in 14 Cheddar cheeses with wide variations in flavor but found only small differences, qualitatively or quantitatively, between cheeses of different flavor. The
206
P. F. FOX et al.
proportions of FFAs (G:oto ClSx3) in cheese were similar to those in milk fat, indicating that these FFAs were released in a nonspecific manner. However, free butyric acid was found at a higher concentration than could be explained by its proportion in milk fat, suggesting that it was selectively liberated or synthesized by the cheese microflora. Lipolysis in hard Italian varieties is extensive and due primarily to the action of PGE in the rennet paste used in the manufacture of these cheeses. Lipolysis in Blue cheese varieties is extensive due to the action of lipases from Penicillium spp. Free fatty acid levels in a number of cheese varieties are listed in Table IV.
5. Catabolism of Fatty Acids The taste and aroma of Blue cheese are dominated by saturated n-methyl ketones, a homologous series in which odd-numbered carbon chains from C3 to C15,as well as a number of even-numbered carbon chains (including C, to Clo), is present (Patton, 1950). Concentrations of methyl ketones in Blue
TABLE IV TYPICAL CONCENTRATIONS OF FREE FAlTY ACIDS
(FFA) I N SOME CHEESE
VARIETIES~
Variety
FFA (mg kg-’)
Sapsago Edam Mozzarella Colby Camembert Port Salut Monterey Jack Cheddar Gruyere Gjetost Provolone Brick Limburger Goat’s milk Parmesan Romano Blue (US.) Roquefort
211 356 363 550 681 700 736 1,028 1,481 1,658 2,118 2,150 4,187 4,558 4,993 6,754 32,230 32,453
“Adapted from Woo er al. (1984) and Woo and Lindsay (1984).
CHEESE; PHYSICAL. BIOCHEMICAL, AND NUTRITIONAL ASPECTS 207
cheese fluctuate, presumably due to reduction to secondary alcohols; however heptan-Zone, nonan-Zone, and undecan-Zone are dominant (Dartley and KinseIla, 1971). Jolly and Kosikowski (1975a) showed that C5,C,, C9, and Cll were dominant methyl ketones in a Blue cheese flavor concentrate. The metabolism of fatty acids in cheese by Penicillium spp. involves four main steps (Fig. 4): (1) release of fatty acids by the lipolytic systems discussed above (Sections IVD2-IVD4), (2) oxidation to P-ketoacids, (3) decarboxylation to methyl ketone with one less carbon atom, and (4) reduction of methyl ketones to the corresponding secondary alcohol (Hawke, 1966); step 4 is reversible under aerobic conditions (Adda et al., 1982). The concentration of methyl ketones is related to lipolysis. Methyl ketones can also be formed by the action of the mould on the ketoacids naturally present at low concentrations in milk fat (ca. 1% of total fatty acids). They could also be formed by the oxidation of monounsaturated acids, but Adda et Saturated Fatty Acids (C,)
CoA-SH
f3-Oxidation, -2H,+ HzO
Keto Acyl-CoA
Acetyl-CoA + Acyl-CoA (Cz,,-d
CoA-SH + P K e b Acid
t Methyl Ketone (C&
I
+ CO,
Reductase
secondsry Alcohol (C,.i)
FIG. 4. Catabolism of fatty acids by Penicillium roqueforri (modified from Kinsella and Hwang, 1976).
208
P. F. FOX
er
al.
al. (1982), who discussed the implications of such a pathway for methyl ketone formation in cheese, considered the evidence for such a pathway is equivocal. A number of factors affect the rate of methyl ketone production, including temperature, pH, physiological state of the mould, and the ratio of the concentration of fatty acid to the dry weight of spores (Adda et al., 1982). Fan et al. (1976) found that both resting spores and fungal mycelium are capable of producing methyl ketones. The rate of production of methyl ketones does not depend directly on the concentrations of FFA precursors; indeed high concentrations of FFAs are toxic to P. roqueforti. Lactones are cyclic esters resulting from the intramolecular esterification of a hydroxyacid through the loss of water to form a ring structure. a- and P-Lactones are highly reactive and are used or occur as intermediates in organic synthesis; y- and S-lactones are stable and have been found in cheese. Lactones possess a strong aroma, which although not specifically cheese-like, may be important in the overall cheese flavor impact. Eriksen (1975) concluded that y- and 6-lactones in freshly secreted milk originated from the corresponding y- and 6-hydroxyacids esterified in triglycerides. Dimick er al. (1969) reported a 6-oxidation system for fatty acid catabolism in the mammary gland of ruminants and thus oxidation within the mammary gland is the primary source of lactone precursors. The potential for lactone production depends on such factors as feed, season, stage of lactation, and breed (Dimick et al., 1969). The formation of y- or Slactones is spontaneous following release of the corresponding hydroxy acid. In a study on Cheddar cheese, Wong et al. (1975) found that longer chain lactones (CI4to c16) increased disproportionately to other lactones in rancid cheese. Two alternative methods of formation were proposed. The first involved the reduction of the corresponding keto acid, but investigations tended to disprove this hypothesis. The other possible method involved the microbial metabolism of homoricinoleic acid to shorter chain hydroxy acids and lactones. S-Lactones have very low flavor thresholds (Kinsella et al., 1965). Jolly and Kosikowski (1975b) found that the concentration of lactones in Blue cheese was higher than that in Cheddar and concluded that the extensive lipolysis in Blue cheese influences the formation of lactones; 6-CI4and Sc 1 6 were the principal lactones in Blue cheese (as found also for Cheddar) (Wong et al., 1973). A stronger typical Blue cheese flavor was found in cheeses containing added lipase, perhaps because lactones blend or modify harsher flavors. O’Keefe et al. (1969) identified y-C12,y-C14, y-c16 6-Clo, S-Cl2,S-c14, SCIS,8C16,and &Cls lactones in Cheddar cheese. The presence of most of these (7-CI2, S-Clo, S-CI2,S-C14, y-c16) in Cheddar cheese was confirmed
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
209
by Wong et al. (1973) who showed a correlation between the number and concentration of lactones with age and flavor, suggesting that certain lactones are significant in Cheddar cheese flavor. In a further quantitative study of lactones in Cheddar cheese, Wong et af. (1975) failed to find a close correlation between flavor and lactone concentration. In general, the above &lactones were produced more quickly and to higher concentrations than y-CIz. Lactone levels increased more rapidly early in the ripening period and the levels found were well above the flavor threshold; it was considered likely that they influence flavor.
E. PROTEOLYSIS
1. Introduction Of the three primary biochemical events that occur during cheese ripening, proteolysis is the most complex and, according to many investigators, the most important. It is primarily responsible for textural changes, including changes in stretchability, meltability, adhesiveness, and emulsifying properties, and makes a major contribution to cheese flavor and the perception of flavor (through release of sapid compounds); unfortunately, some peptides produced are bitter and if present at sufficient concentrations will cause bitterness, which is probably the principal flavor defect in cheese. Proteolysis during maturation is essential in most cheese varieties. The extent of proteolysis varies from very limited (e.g., Mozzarella) to very extensive (e.g., Blue mould varieties) and the products range in size from large polypeptides, comparable in size to intact caseins, through a range of medium and small peptides to free amino acids. Clearly, no one proteolytic agent is responsible for such a wide range of products. 2. Assessment of Proteolysis Techniques for the assessment of proteolysis in cheese fall into two general classes: specific and nonspecific methods. The latter include determination of nitrogen soluble in, or extractable by, one of number of solvents or precipitants (e.g., water, pH 4.6 buffers, CaC12,NaCl, ethanol, trichloroacetic acid, phosphotungstic, or sulfosalicilic acids), or permeable through ultrafiltration membranes or the formation of reactive groups (e.g., -NH2). Such methods are valuable for assessment of the overall extent of proteolysis and the general contribution of each proteolytic agent. Nonspecific techniques are generally simple and are valuable for the routine assessment of cheese maturity since soluble nitrogen correlates well with cheese age and to a lesser extent with quality.
210
P. F. FOX et al.
Specific techniques, such as chromatography and electrophoresis, resolve individual peptides. Various forms of chromatography have been used to study cheese peptides, including paper, thin-layer, ion-exchange, gel permeation, metal chelate, and, more recently, a variety of highperformance techniques, particularly reverse-phase high-performance liquid chromatography (RP-HPLC). Electrophoresis is a very effective and popular technique for assessing primary proteolysis in cheese, especially alkaline urea-PAGE, but also SDS-PAGE and isoelectric focusing. Capillary electrophoresis has not yet been used widely but this technique will probably find widespread application for the analysis of peptides from cheese. Techniques for the assessment of proteolysis in cheese during ripening have been the subject of a number of recent reviews, including Grappin et al. (1985), Rank et al. (1985), Fox (1989a), IDF (1991a), McSweeney and Fox (1993), and Fox et al. (1995). 3. Relative Importance of Proteolytic Agents in Cheese
Much of the information on the relative importance of individual proteinases has been obtained from studies on model cheese systems. The most comprehensive of these studies is that of Visser (l976,1977a,b,c) and Visser and de Groot-Mostert (1977) in which the relative importance of enzymes from rennet, starter bacteria, and milk to proteolysis in Gouda cheese was assessed. The results indicated that rennet is mainly responsible for initial proteolysis and the production of most of the water- or pH 4.6-soluble nitrogen. However, the production of small peptides and free amino acids is due primarily to the action of enzymes from starter bacteria. The results of other studies on controlled-microflora cheese (Reiter et al., 1969; Gripon et al., 1975; Desmazeaud et al., 1976; R. B. O’Keeffe et al., 1976a,b; A. M. O’Keeffe et al., 1978) were generally similar. Visser (1977~)found that only -5% of the total nitrogen in a 6-month-old aseptic starter-free, rennet-free cheese was soluble at pH 4.6, with very low levels of free amino acids, indicating only a minor role for plasmin. Farkye and Fox (1991), who inhibited plasmin in Cheddar cheese by AHA, found differences between electrophoretograms of experimental and control cheeses, especially ycaseins (produced from &casein by plasmin) and in the level of watersoluble nitrogen, suggesting a role for plasmin in the initial hydrolysis of caseins. Bovine milk also contains an indigenous acid proteinase which appears to be cathepsin D (see Kaminogawa et al., 1980); procathepsin D has been identified in bovine milk (Larsen et al., 1993). The action of cathepsin D
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
211
on the caseins is very similar to that of chymosin (Kaminogawa et al., 1980; McSweeney et al., 1995). Although lactic acid bacteria, including the genera Lactococcus and Lactobacillus, are weakly proteolytic, they possess a great variety of proteolytic enzymes, especially peptidases, which contribute significantly to the later stages of proteolysis during cheese ripening, i.e., the formation of small peptides and amino acids. The proteolytic system of lactic acid bacteria is discussed in Section IV E6. Although nonstarter lactic acid bacteria (NSLAB) can dominate the microflora of Cheddar-type cheese during much of its ripening (see Peterson and Marshall, 1990;Khalid and Marth, 1990a),their influence on proteolysis in cheese has been neglected by most authors. Visser (1977a,b,c) used aseptic cheesemaking techniques to eliminate NSLAB, as did Desmazeaud etal. (1976), R. B. O’Keeffe et al. (1976a), and A. M. O’Keeffe et al. (1978). A wide range of proteolytic enzymes have been identified in NSLAB (see Peterson and Marshall, 1990; Kahlid and Marth, 1990a), and therefore it is likely that they contribute to proteolysis in cheese. The influence of adjunct starters, e.g., Penicillium roqueforti, on proteolysis can be great in varieties in which they are used. P. roqueforti degrades both asl-and P-caseins rapidly and releases large amounts of small peptides and free amino acids (Desmazeaud et al., 1976). The progress of proteolysis in most ripened cheeses can be summarized as follows: initial hydrolysis of caseins is caused primarily by residual coagulant, and to a lesser extent by plasmin and perhaps cathepsin D, resulting in the formation of large and intermediate-sized peptides which are subsequently degraded by the coagulant and enzymes from the starter and nonstarter flora of the cheese. The production of small peptides and free amino acids results from the action of bacterial proteinases and peptidases. This general outline of proteolysis can vary substantially between varieties due to differences in manufacturing practices. In Mozzarella, Swiss, and other high-cook varieties, coagulant is extensively or completely denatured and plasmin is therefore a more important contributor to primary proteolysis than in Cheddar. 4.
Proteinases from the Coagulant
Chymosin (EC 3.4.23.4), the principal proteinase in traditional rennets used for cheesemaking (Rothe et al., 1977), is an aspartyl proteinase of gastric origin, secreted by young mammals. The principal role of chymosin in cheesemaking is to coagulate the milk. However, about 6% of the chymosin added to cheese milk is retained in the curd for Cheddar and plays a
212
P. F. FOX et al.
major role in the initial proteolysis of caseins in many cheese varieties (Fox, 1989a). The action of chymosin on the B-chain of insulin indicates that it is specific for hydrophobic and aromatic amino acid residues (Fish, 1957). Chymosin is relatively weakly proteolytic; indeed, limited proteolysis is one of the characteristics to be considered when selecting proteinases for use as rennet substitutes (Fox, 1989a). The primary chymosin cleavage site in the milk protein system is the Phelo~-Metlo6bond in rc-casein. This bond is many times more susceptible to chymosin than any other bond in milk proteins (Vreeman et al., 1986) and its hydrolysis leads to coagulation of the milk (see Section IIIA). Cleavage of K-casein Phel05-Metl06 yields para-K-casein (K-CNfl-105) and glycomacropeptides (GMP; K-CN f106-169). Most of the GMP is lost in the whey but the para-rc-casein remains attached to the casein micelles and is incorporated into the cheese. asl-,as2-,and @-caseinsare not hydrolyzed during milk coagulation but may be hydrolyzed in cheese during ripening. A number of authors (Pelissier et af., 1974; Creamer, 1976a; Visser and Slangen, 1977; Carles and Ribadeau-Dumas, 1984) have investigated the action of chymosin on P-casein. In solution in 0.05 M Na acetate buffer, pH 5.4, chymosin cleaves P-casein at seven sites: Leu192-Tyr193> Ala189-Phe190> LeUl65-Serl66 2 Ghl67-Serl68 2 LeUl63-Serl64 > LeU139Leul4o2 L e ~ ~ ~ ~ - T h(Visser r 1 2 ~ and Slangen, 1977). The Michaelis paramer,~~ ters, K , and kcat,for the action of chymosin on the bond L e ~ , ~ - T y are 0.075 mM and 1.54 sec-', respectively, for micellar @-caseinand 0.007 mM and 0.56 sec-' for the monomeric protein (Carles and Ribadeau-Dumas, 1984). NaCl inhibits the hydrolysis of @-caseinby chymosin to an extent dependent on pH; hydrolysis is strongly inhibited by 5% NaCl and is completely inhibited by 10% (Mulvihill and Fox, 1978). The primary site of chymosin action on asl-casein is Phe23-Phe24(Hill et al., 1974; Carles and Ribadeau-Dumas, 1985). Cleavage of this bond is believed to be responsible for softening of cheese texture and the small peptide (asl-CN fl-23) is further hydrolyzed by starter proteinases. The -casein in solution was studied by Pellissier specificity of chymosin on aS1 et af.(1974), Mulvihill and Fox (1979), Pahkala etaf.(1989), and McSweeney et af. (1993b). The hydrolysis of asl-casein by chymosin is influenced by pH and ionic strength (Mulvihill and Fox, 1977,1979). In 0.1 M phosphate buffer, pH 6.5, chymosin cleaves aSl-casein at Phe23-Phe24,Phe28-Pr029, Leu.~-Ser41, Leu149-Phe150, Phe1~3-Tyr154,Le~156-Asp1~7,Try1~9-Pw60, and Trp1a-Tyr165 (McSweeney el af., 1993b). These bonds are also hydrolyzed at pH 5.2 in the presence of 5% NaCl (i.e., conditions in cheese) and, in addition, Leull-Prol2, Phe32-Gly33, Leulol-Lys102, Leu142-Alal43, and Phe179-Serl80. The rate at which many of these bonds are hydrolyzed de-
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
213
pends on the ionic strength and pH, particularly Leulol-Lyslo2,which is cleaved far faster at the lower pH. The k,,, and K,,, for the hydrolysis of Phe23-Phe24bond of a,,-casein by chymosin are 0.7 sec-I and 0.37 mM, respectively (Carles and Ribadeau-Dumas, 1985). aS2-Caseinappears to be relatively resistant to proteolysis by chymosin; cleavage sites are restricted to the hydrophobic regions of the molecule, i.e., residues 90-120 and 160-207: Phess-Tyrs9, T ~ r ~ ~ - L e u ~ ~ , T ~ r ~ ~ - L Phe163-Leu164, eu~~, Phe174-AlaI,s, Tyr179-Leu180(McSweeney er al., 1994b). Although parmecasein has several potential chymosin cleavage sites, it does not appear to be hydrolyzed either in solution or in cheese (Green and Foster, 1974). Calf rennet contains about 10% bovine pepsin (EC 3.4.23.1, Rothe er al., 1977). The proteolytic products produced from Na-caseinate by bovine pepsin are similar to those produced by chymosin (Fox, 1969), although as far as we are aware the specificity of bovine or porcine pepsins on bovine caseins has not been rigorously determined. For several years, the supply of calf rennet has been insufficient to meet demand and much effort has been expended on searching for suitable rennet substitutes for cheesemaking (see Green, 1977; Phelan, 1985). A number of enzymes have been studied, including bovine, porcine, ovine, and chicken pepsins and proteinases from Cryphonectria parasitica, Rhizomucor pusillus, R. miehei, Penicillium janthinellum, Rhizopus chinensis, and Aspergillus usameii. The specificity of many of these enzymes on the oxidized B-chain of insulin were summarized by Green (1977). The specificity of the fungal proteinases on caseins differ substantially from chymosin but have not been determined rigorously. Recombinant calf chymosins,expressed in Aspergillus niger var. awamori, Kluveromyces marxianus var. lactis or E. coli. were introduced recently and have been accepted by the regulatory authorities in many countries for use in foods; they are now used widely for cheesemaking. Cheesemaking trials, involving a number of cheese varieties, have shown only small differences between cheese made using calf rennet or recombinant chymosins (Green et al., 1985; Hicks et al., 1988; Bines et al., 1989; van den Berg and de Koning, 1990; O’Sullivan and Fox, 1991; Nuiiez et al., 1992). Recombinant chymosins contain only one genetic variant of this enzyme (Harboe, 1992), while calf rennet can contain three chymosin variants (A, B, and C Teuber, 1990) as well as bovine pepsin.
5. Indigenous Milk Proteinases a. Plasmin. Plasmin (fibrinolysin, EC 3.4.21.7) has been the subject of much study (for review, see Grufferty and Fox, 1988). The physiological
214
P. F. FOX et al.
role of plasmin is solubilization of fibrin clots; it is a component of a complex system consisting of the active enzyme, its zymogen, activators and inhibitors of the enzyme, and its activators, all of which are present in milk. Plasmin, plasminogen, and plasminogen activators are associated with the casein micelles in milk, while the inhibitors are in the serum phase (see Grufferty and Fox, 1988; Fox and Law, 1991). Plasmin is a trypsin-like serine proteinase with a pH optimum at about 7.5 and a high specificity for peptide bonds involving lysyl residues. It is active on all caseins, but especially on as2-and @caseins (Grufferty and Fox, 1988). Plasmin cleaves p-casein at three primary sites: Lys28-Lys29, Ly~1~5-His106, and L ~ s ~ ~ with ~ - the G formation ~ u ~ ~ of~the ~ polypeptides, @-CNf29-209 (yl-CN), f106-209 (y2-CN), and f108-209 (y3-CN), &CN fl-105 and fl-107 (proteose peptone 5), P-CN f29-105 and f29-107 (proteose peptone 8-slow) and 0-CN fl-28 (proteose peptone %fast) (see Eigel et al, 1984). Additional cleavage sites are at Lysl13-Tyrl14and Arg183Asplu (Fox et al., 1994). Plasmin cleaves as2-caseinin solution at eight sites: Lys21-Gln22, LysZ4Asnzs, &3114-ASnli~,LySi49-LyS1~0,Lysl~o-Thrl~~, LYSi8i-Thrim LySi88Alals9, and L y s 1 9 ~ -T h r~ (Visser ~ ~ et al., 1989; Le Bars and Gripon, 1989), producing about 14 peptides, 3 of which are potentially bitter (Le Bars and Gripon, 1989). Although plasmin is less active on aSl-caseinthan on as2-and @caseins, the formation of A-casein, a minor casein component, has been attributed to its action on asl-casein (Aimutis and Eigel, 1982). The specificity of plasmin on aSl-caseinin solution has been reported by McSweeney et al. (1993~)who found the principal plasmin cleavage sites to be Arg2~-Phe~~, Arg90-Tyr91, Lys102-Lys103, Ly~103-Tyr10.1,L Y ~ I O S - VLYSWG~UIZ, ~~O~, and ArglSl-Glnl52. Plasmin has very low activity on K-casein, although it contains several potential sites; Eigel(l977) found no hydrolysis of K-casein under conditions adequate for the hydrolysis of asl-casein.However, Andrews and Alichanidis (1983) reported that 4% of the peptides produced by indigenous plasmin in pasteurized milk stored at 37°C for 7 days and detectable by PAGE originated from K-casein. The specificity of plasmin on K-casein has not been determined. Cathepsin L). The indigenous acid proteinase in milk has received little attention. This activity was first recognized by Kaminogawa and Yamauchi (1972), who isolated and characterized the enzyme and considered it to be similar to the lysozomal acid proteinase, cathepsin D (EC 3.4.23.5). The presence of procathepsin D in milk has been reported (Larsen et al., 1993). Cathepsin D is relatively heat labile (completely inactivated by 70°C X 10
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
215
min) and has a pH optimum of 4.0 (Kaminogawa and Yamauchi, 1972). The specificity of cathepsin D on the caseins has not been determined, although electrophoretograms of caseins incubated with milk acid proteinase (Kaminogawa and Yamauchi, 1972) or cathepsin D (McSweeney et al., 1995) indicate a specificity very similar to that of chymosin; surprisingly, it coagulates milk very slowly (McSweeney et al., 1995). Other Indigenous Milk Proteinases. The presence of other minor proteolytic enzymes in milk has been reported, including thrombin and a lysine aminopeptidase (Reimerdes, 1983) and proteinases from leucocytes (Grieve and Kitchen, 1985; Verdi and Barbano, 1991), but they are considered not to be very significant (Grieve and Kitchen, 1985; Grufferty and Fox, 1988). 6. Proteolytic Enzymes from Starter
Although lactic acid bacteria (LAB) are weakly proteolytic they do possess a proteinase and a wide range of peptidases which are principally responsible for the formation of small peptides and amino acids in cheese. The genus most widely used as a cheese starter is Lactococcus, the proteolytic system of which has been studied thoroughly at the molecular, biochemical, and genetic levels. The proteolytic system of Lactobacillus spp. is less well characterized than that of Lactococcus, but the systems of both genera appear to be generally similar. The extensive literature has been comprehensively reviewed by Monnet et al. (1993), Tan ef al. (1993a), Visser (1993), and Law and Haandrikman (1996), to which the reader is referred. The principal properties of the peptidases isolated to date are summarized in Table V. The proteolytic system is capable of hydrolyzing casein completely to free amino acids; the sequential action of the peptidase system is shown schematically in Fig. 5. This complex proteolytic system is required by LAB for growth to high numbers in milk which contains a low concentration of small peptides and free amino acids. The proteinase in LAB is anchored to the cell membrane and protrudes through the cell wall, giving it ready access to extracellular proteins. All the peptidases are intracellular although some, e.g., Pep X, appear to be oriented toward the outer surface of the cell membrane (Tan et al., 1992). The oligopeptides produced by the proteinase are actively transported into the cell where they are hydrolyzed further by the battery of peptidases. Cell wall-associated proteinases of Lactococcus can be classified into three groups, PI-, PIII-,and mixed-type; P,-type proteinases degrade p- but not aSl-casein at a significant rate, while PIII-typeproteinases rapidly degrade both a-and p-caseins ( S . Visser et al., 1986).The nucleotide sequences
TABLE V YEYlIDASES OF LACTIC ACID BACTERIA
Organism
Principal assay substrate
MW/kDa
Oligoendopeptidases (LEP, MEP, NOP, PepO) Lc. bv. diacetylacrir CNRZ 267 Peptides 49 Lc. lactis ssp. cremoris H61 Peptides 98 Lc. lactis ssp. cremoris H61 a,l-CN fl-23 80 Lc. lactis ssp. cremoris Wg2 Metenkephalin 70 Lc. lactis ssp. cremoris HP ~I-CN fl-23 180 Lc. la& ssp. cremoris C13 (Y~I-CN fl-23 70 Lc. lactis ssp. lactis MG 1363 q,-CN fl-23 70 Lc. lacris ssp. cremoris SKll Bradykinin 70 Aminopeptidases Aminopeptidase N (Generalaminopeptidme, AMP, PepN) Lc. bv. diacetylacris CNRZ 267 Lys-p-NA 85 Lc. lactis ssp. cremoris ACI Lys-p-NA 36 Lc. lactis ssp. cremoris Wg2 Lys-p-NA 95 Lb. delbrueckii ssp. lactis 1183 Lys-p-NA 78-91 Lb. acidophilus R-26 Lys-p-NA 38 Lb. delbrueckii ssp. bulgaricus Lys-p-NA 95 CNRZ 397 Lb. helveticus CNRZ 32 Lys-p-NA 97 Lb. delbrueckii ssp. bulgaricus Lys-p-NA 95 B14 Lb. helveticus LME-511 Leu-p-NA 92 Lb. casei ssp. casei LLG Leu-p-NA 87 Lb. delbrueckii ssp. bulgaricus Lys-p-NA 98
Optimum Activity pH (“C) Subunits
Class
Reference
7-7.5 6 6-6.5 8-9 6-7 7.5 6.0
40 37 30-38 42 35 40 -
-
-
1 2 1 >2 1 1 1
M‘ M M M N M M
6.5 7 7 6.2-7.2 -
35 40 40 47.5
-
Desmazeaud and Zevaco (1979) M Geis er al. (1985) M, -SH Tan and Konings (1990) M Eggimann and Bachmann (1980) Machuga and Ives (1984) M M Atlan et al. (1989)
-
1 1 1 -
-
-
-
-
Desmazeaud and Zevaco (1976) Yan er al. (1987b) Y a n et al. (1987a) Tan et al. (1991) Baankreis (1992) Baankreis (1992) Stepaniak and Fox (1995) Pritchard et al. (1994)
M
7
50
1
M M
Khalid & Marth (1990~) Bockelmann et al. (1992)
7 7 6
37 39 40
1 1 1
M M M
Miyakawa et al. (1992) Arora and Lee (1992) Tsakalidou er aL (1993)
6.5
50
1
M
Blanc et al. (1993)
ACA-DC233
Lb. helveticus ITGLl
Lys-p-NA
97
Str. thermophilus CNRZ 1199 Lys-p-NA 89 6.5 Str. thermophilus CNRZ 302 Lys-p-NA 97 7.0 Aminopeptidase A (Glutamyl aminopeptidase, GAP, PepA) Lc. lactis ssp. cremoris HP GlulAsp-p-NA 130 245 8 Lc. lactis ssp. lactis NCDO 712 Glu-p-NA Glu-p-NA 520 8 Lc. lactis ssp. cremoris HP Aminopeptidase C (Thiol amiopeptidase, PepC) Lc. lactis ssp. cremoris AM2 His-P-NA 300 7 Lb. delbrueckii ssp. bulgaricus Leu-Gly-Gly 220 6.5-7 B 14 Pyrrolidonyl Carboxylyl Peptidase (pyroglutamyl aminopeptidase, PCP) Lc. lactis ssp. cremoris HP Pyr-p-NA Lc. lactis ssp. cremoris HP Pyr-p-NA 80 8-8.5 X-Prolyldipeptidyl aminopeptidase (XPDA, PPDA, XAP, PepX) Lc. lactis ssp. cremoris P8-2-47 X-Pro-p-NA 180 7 Lc. Iacris ssp. lactis NCDO 763 Ala-Pro-p-NA 190 8.5 Lc. lactis ssp. cremoris AM2 Gly-Pro-NH-Mec 117 6-9 Lc. lactis ssp. lactis H1 X-Pro-p-NA 150 6-9 Lb. delbrueckii ssp. lactis X-Pro-p-NA 165 7 72 7 Lb. helveticus CNRZ 32 X-Pro-p-NA 82 7 Lb. delbrueckii ssp. bulgaricus X-Pro-p-NA CNRZ 397 170-200 6.5 Lb. delbrueckii ssp. bulgaricus Ala-Pro-p-NA B14 Lb. acidophilus 357 Ala-Pro-p-NA 170-200 6.5 Lb. delbrueckii ssp. bulgaricus Gly-Pro-p-NA 270 6.5
35 36 50-55 65 50 40 50
-
1 1
M M
Tsakalidou et al. (1993) Rul et al. (1994)
3 6 -10
M M M
Exterkate and de Veer (1987) Niven (1991) Baankreis (1992)
-SH
Neviani et al. (1989) Wohlrab and Bockelmann (1993)
6 4
-
-SH -
37
2
S
Exterkate (1977) Baankreis (1992)
45-50 40-45 -
2
S S S S S S S
Kiefer-Partch et al. (1989) Zevaco et 01. (1990) Booth et al. (1990a) Lloyd and Pritchard (1991) Meyer and Jordi (1987) Khalid and Marth (1990d) Atlan et al. (1990)
-
50-55 40 50
2 -
2 1 -
45
2
S
Bockelmann et al. (1991)
45 50
2 3
S S
Bockelmann et al. (1991) Miyakawa el al. (1991)
50
1
S
Miyakawa et al. (1994)
LBU-147 Lb. helveticus LHE-511
Gly-Pro-p-N A
90
6.5
(continues)
z
-4
TABLE V (Continued) Organism
Proline Iminopeptidase (PIP) Pr. freud ssp. shermnii 13673 Lc. lactis ssp. cremoris HP Lb. delbrueckii ssp. bulgaricus CNRZ 397 Lb. casei ssp. casei LLG Dipeptidnses (DIP) Lc. bv. diacetylach CNRZ 261 Lactococcus spp. Lc. lactis ssp. cremoris H61 Lc. lactis ssp. cremoris Wg2 Lb. delbrueckii ssp. bulgaricus B14 Prolidase (PRD) Lc. lach ssp. cremoris H61 Lc. lactis ssp. cremoris A M 2 Tripeptidase (TRP) Lc. bv. diacetylactis CNRZ 267 Lc. lactis ssp. cremoris Wg2 Lc. lactis ssp. cremoris AM2 Lc ktis ssp. crenwris IMN-Cl2 Lb. delbrueckii ssp. bulgaticus B14
Principal assay substrate Pro-Gly-Gly Pro-p-N A
MW/kDa
-
Optimum Activity pH (“C) Subunits
Class
Reference
100 100
8.5 6-7
37 40
2 3
M S
Panon (1990) Baankreis (1992) Gilbert et al. (1994)
Pro-AMC
46
7.5
40
1
SH
Habibi-Najafi and Lee (1995)
Leu-Leu Dipeptides Leu-Gly Dipeptides Dipeptides
51 25 & 34 100 49 51
7.5 7 8 8 7
30 50 50
1 1 1
M M M M M
Desmazeaud and Zevaco (1977) Law (1979) Hwang et al. (1981) van Boven et al. (1988) Wohlrab and Bockelmann (1992)
43 42
6.5-1.5 7.35-9
-
-
M M
Kaminogawa et al. (1984) Booth et al. (1990b)
1
35 35 33 40
-
M M M -SH M
Desmazeaud and Zevaco (1979) Bosman et al. (1990) Bacon ef al. (1993) Sahlstrom et al. (1993) Bockelmann ef al. (1995)
Leu-Pro Leu-Pro Tripeptides Leu-Leu-Leu Tripeptides Leu-Leu-Leu Leu-Gly-Gly
75 103-105 105 72 85
7.5 8.6 5.8 6.0
M, metallo; S, serine; SH, thiol; N , neutral; AMC, aminomethyl coumarin.
-
-
-
-
2 2 3 >1
-
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
219
+
pep0 pyro-Glu-Lys-Ala-Glx-Gly-Ro-Leu-Leu-Lu-Ro-His-~e
+
PCP
PIP
pyro-Glu-Lys-Ala-Glx-GIy-Ro-Lu-Lu-Leu
\cH
Ro- is-Phe
prN
Lys-Ala-Glx-GIy-Ro-Lu-Leu-Leu
'i"
Ala-Glx-GIy-Ro-Lu-Lu-Leu
PepX
DIP
1
Leu-Leu
FIG. 5. Schematic representation of the hydrolysis of a hypothetical dodecapeptide by the combined action of endo- and exopeptidases of Lactococcusspp. Modified fromFox etal. (1995).
of the genes for both PI- and PIrI-typeproteinases are known (Kok et af., 1988; Vos et af., 1989a,b);few differences are apparent and alteration of a few residues by site-directed mutagenesis can alter the specificity of the proteinase (Kok, 1990). However, the specificity of the lactococcal proteinases appears to be more diverse than proposed by S . Visser et af. (1986), e.g., Exterkate et al. (1993) classified the proteinases of 16 Lactococcus strains into seven groups based on their specificity on a,l-CN fl-23. The specificity of the CEP from several Lactococcus strains on as1-aSz-, 0- and K-caseins and short peptide substrates has been established; these studies have been reviewed by Fox et af. (1994) and are summarized in Figs. 6-10. The lactococcal CEP appears to be of great importance in proteolysis in cheese (see Section IV E9). 7. Proteolytic Enzymes of the Nonstarter Microjlora Despite the findings by a number of authors that the NSLAB can dominate the microflora of Cheddar-type cheese during much of its ripening,
220
P. F. FOX et al. I
1
I1
f1I
I
PI R ~ P X E P I X K Q G L P Q E V L N E N L ~ O L R F F V A P F P E V ? G K E K V N E 4 ~ S X D I G S ~ S T E ~ ~
I11
I
1
["
D
Q A
I4 E D I X Q
"60 E
A
121
¶
L E Q L L RlOO
I
I I
11 1
[11
1
1
E P I P P P I E I V P N S V I 9 K Hgo I Q K E 0 V P 6 E R Y L G
I
A
R Q ~ Y Q L D ~ Y P ~ ~ ~ S G A X Y Y V P L G T Q Y T D A P S F ~ ~ ~ ~ D I P N P T G S ~ I ~ S E K T ~ ~ P
FIG. 6. Specificity of lactococcal cell envelope-associated proteinase on ar,l-casein. [l]Lactococcus lactis spp. cremoris SK112 (Reid et al., 1991). [2] L. lactis ssp. lactis NCDO 763 (Monnet e? al., 1992). Modified from Fox et al. (1995).
the proteolytic system of NSLAB has received little attention compared with that of Lactococcus. The proteolytic specificity of proteinases from NSLAB on the caseins has not been determined. The predominant NSLAB in Cheddar and Dutch-type cheeses are mesophilic Lactobacillus, which possess a cell wall-associated and intracellular proteinases. A range of intracellular peptidases, including dipeptidases, aminopeptidases, and endopeptidases, have been identified in Lactobacillus (see reviews by Khalid and Marth 1990a; Peterson and Marshall, 1990). Interestingly, carboxypeptidase activity, which has not been found in lactococci, has been reported in Lactobacillus casei (El Soda et al., 1978). p-Casein is preferentially degraded by a number of strains Lactobacillus plantarum and L. casei, but some strains degraded aS1-caseinalso (Khalid and Marth, 1990b). NSLAB also include Micrococcus and Pediococcus. All Micrococcus spp. appear to produce intracellular proteinases and some also produce K
N
T
I
I
~
~
V
~
~
~
~
E
~
I
I
~
Q
C
T
1
Y
~
~
K
Q
C
~
N
K
~
I
1
C Y S I G 8 S 8 C C ~ ~ S A C V A T ~ C V K l T V D D X K Y Q X ~ ~ A L N ~ I N C F Y Q K ? P Q Y L Q Y L Y ~ ~
A
1
A
Q Q P I V L A P Y D Q V X R H A V P I T ~ ~ ~ P T L N R C Q L B T 8 t I I W S K X T V D ~ q ~ n E 6 T t V F T X X ~ ~
1
A
1
i
14
1
T ~ L T ~ ~ ~ ~ R R ~ ~ ~ L I I L ~ ~ I ~ Q R Y Q K ~ A L P ~ Y L ~ ~ ~ K T V Y Q W Q K
A I P 1 V R Y L207
FIG. 7. Specificity of the cell envelope-associatedproteinase from Lactococcus lactis ssp. lactis NCDO 763 (Monnet et al., 1992) on as*-casein. Modified from Fox et al. (1995).
N
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
221
FIG. 8. Specificity of lactococcal cell envelope-associated proteinase on p-casein. [l] L. lactis ssp. cremoris H2 (Reid et al., 1991b). [2] L. luctis ssp. cremoris SK112 (Reid el al., 1991b). 131 L. lactis ssp. cremoris AM, (Visser et al., 1991). [4] L. lacris ssp. crernoris JP (Visser et ab, 1988). [5] L. lactis ssp. cremoris AC1 (Monnet er al., 1989). [6] L. lactis ssp. lactis NCDO 763 (Monnet er al., 1989). [7] L. lactis ssp. lactis NCDO 763 (Monnet et al., 1986). Modified from Fox et al. (1995).
FIG. 9. Specificity of lactococcal cell envelope-associated proteinase on K-casein. [l] Lactococcus lactis ssp. lactis NCDO 763 (Monnet et al., 1992). [2] L. lactis ssp. cremoris H2 (Reid et al., 1994). [3] L. l a d s ssp. cremoris S K l l (Reid et al., 1994). [4] L. lactis ssp. cremoris AM1 (Visser et al., 1994). Modified from Fox et at. (1995).
m N d W
m I4 m
m W 4 9.l
W m
d
m 0
N A
7
0 N
W N
O
0 rn
W
ri
w
a
A m d W
? m!
OL
.!i
2
4
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
223
extracellular proteinases. All strains studied by Nath and Ledford (1972) and Bhowmik and Marth (1988) possessed intracellular proteoloytic activity. Baribo and Foster (1952) studied some of the characteristics of the intracellular proteolytic activity of M . freudenreichii 325. Micrococcus spp. also produce extracellular proteinases with alkaline pH optima. Desmazeaud and Hermier (1968a,b) studied an extracellular neutral metalloproteinase from M. caseolyticus and Desmazeaud and Hermier (1971) determined its specificity on glucagon; this enzyme, which did not possess exopeptidase activity, rapidly cleaved at bonds containing Phe, Leu, or Ala. Nath and Ledford (1972) found that 3 of the 18 strains of Micrococcus studied produced extracellular proteinases which preferentially hydrolyzed aSl-caseinwhile intracellular proteinases preferetially hydrolyzed 0-casein. The extracellular metalloproteinase (28.9 kDa) of Micrococcus MCC-315 was isolated by Prasad et al. (1986) who found this enzyme to be optimally active at pH 10.6 and at temperatures between 50 and 60°C (whole casein) or 37 to 40°C (@-casein).Garcia de Fernando and Fox (1991) purified two extracellular metalloproteinases (23.5 and 42.5 kDa) from Micrococcus GF, a strain which had been isolated from farmhouse Blue cheese. The enzymes were optimally active at -45°C and pH 8.5 to 11. One proteinase preferentially hydrolyzed 6-casein while the other hydrolyzed both as]and 6-caseins at approximately the same rate. Membrane-associated and intracellular proteinases have also been found (see review by Bhowmik and Marth, 1990b). There are few reports on the proteolytic activity of Pediococcus; Tzanetakis and Litopoulou-Tzanetaki (1989) found Leu and Val aminopeptidase activities in P. pentosaceus and El-Soda et al. (1991) reported aminopeptidase and dipeptidase activity in Pediococcus sp. LR.
8. Proteinases from Secondary Starter Proteinases and peptidases from the secondary (adjunct) starter can play an important role in proteolysis in cheese varieties where such adjuncts
FIG. 10. Specificity of lactococcal cell envelope-associated proteinase or endopeptidase (Pep 0)on various peptides. [ l ] L. luctis ssp. cremoris H61 (Kaminogawa et ul., 1986). [2] L. luctis ssp. cremoris HP (Exterkate et ul., 1991). [3] L. luctis ssp. cremoris AM, (Exterkate et ul., 1991). [4] L. lucris ssp. cremoris (Baankreis, 1992). [ 5 ] L. luctis ssp. cremoris C13 (Baankreis, 1992). [6] L. lucris ssp. lactis MG 1363 (Stepaniak and Fox, 1995). [7] L. luctis ssp. cremoris H61 (Yan et ul., 1987b). [8] L. luctis ssp. luctis NCDO 763 (Monnet et ul., 1992). [9] L. luctis ssp. cremoris H61 (Yan etal., 1987a). [lo] L. lactis ssp. cremoris AMI,pH 6.5 (Exterkate and Alting, 1993). [ l l ] L. luctis ssp. cremoris AM,, pH 5.2 + NaCl (Exterkate and Alting, 1993). Modified from Fox et ul. (1995).
224
P. F. FOX
et
al.
are used, A number of authors (e.g., Broome et al., 1990, 1991) have used lactobacilli as adjuncts in the manufacture of Cheddar; their proteolytic system is presumably the same as that of starter and nonstarter lactobacilli, as was discussed in Section IV E6 and IV E7. In this section, enzymes from traditional adjuncts, i.e., Penicillium spp. (mould-ripened varieties), Brevibacterium linens (smear-ripened varieties), and Propionibacterium spp. (Swiss varieties) will be discussed. Blue-veined cheeses are characterized by the growth of Penicillium roqueforti throughout the cheese and Camembert and Brie by the growth of P.cumemberti on the surface. These moulds produce aspartyl and metaland p-caseins loproteinases which have generally similar specificities on as](Trieu-Cuot and Gripon, 1981). The aspartyl proteinases from both species cleave p-casein at positions L y ~ ~ ~ - ILysg7-Va198, le~~, and Lysw-Glulw, releasing five peptides: p-CN f98-209, f100-209, f1-97/99, f30-209, and fl-29 (Trieu-Cuot and Gripon, 1982). The aspartyl proteinase of Penicillium spp. also cleaves aSl-casein at a number of sites; according to Trieu-Cuot and Gripon (1982), the primary site is at or near the primary site of chymosin action, i.e., PheZ3-Phez4. The metalloproteinases of P. roqueforti and P. camemberti are generally similar; the primary cleavage sites for the P. camemberti enzyme on pcasein are LYS28-LYS29, P r ~ ~ ~ - G land u ~ ,Gluloo-Alalol , (Trieu-Cuot et al., 1982). Intracellular acid proteinase(s) and exopeptidases (amino and carboxy) have been found in P. roqueforti and P. camemberti but have not been well studied (see Gripon et aL, 1991; Gripon, 1993). P. roqueforti excretes a carboxypeptidase with an acid pH optimum and an alkaline metallo aminopeptidase (see Gripon et al., 1991; Gripon, 1993). Brevibacterium linens, a characteristic component of the surface microflora of smear-ripened varieties, has a strong proteolytic activity (Gripon et al., 1991). The extracellular proteinases secreted by a number of Br. linens strains have been partially purified (see Rattray et al., 1995). The proteinase secreted by Br. linens ATCC 9174 was purified to homogeneity by Rattray et al. (1995) and found to be a serine proteinase with optimum activity at pH 8.5 and 50°C. It appears to be a trimeric enzyme with a monomeric mass of 56 kDa. The sequence of the 20 N-terminal amino acids showed no homology with the published sequences of other bacterial proteinases (Rattray 1995). Foissy (1974) reported on both intra- and extracellular proteinases but they were not studied further; an extracellular aminopeptidase was purified by Foissy (1978). SBrhaug (1981) described a number of peptidase activities in cell-free extracts of Br. linens. There are few studies on the proteinase and peptidase activities of Propionibacterium shermanii, the characteristic microorganism in Swiss cheese
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
225
varieties. Langsrud et al. (1977,1978) studied the release of Pro by Propionibacterium in growth media. The location of peptidases from P. shermanii a and ATCC 9614 was studied by Sahlstrom et al. (1989) who found a number of peptidases (assayed on a range of di- and tripeptides) associated with the wall, membrane, and intracellular fractions. Perez Chaia ef al. (1990) studied the aminopeptidase and proline iminopeptidase activities of cell-free extracts of P. freudenreichii ssp. shermanii ATCC 1367 and E22, which were optimal at 37-45°C and pH 6.4-7.2. Panon (1990) isolated a proline iminopeptidase from P. shermanii 13673; this serine enzyme (61 kDa) was optimally active at pH 8 and 40°C. P. shermanii (NZ) was one of the bacteria studied by El-Soda et al. (1991) who found that its partially purified aminopeptidase was optimally active at pH 7.5 and 50°C. El-Soda et al. (1992) studied the intracellular peptide hydrolase system of P. freudenreichii, P. acidipropionicci, and P. jensenii; aminopeptidase and dipeptidase activities were found in all species studied but no carboxypeptidase or endopeptidase activities were detected. The majority of the strains studied were able to degrade casein. The cell wall-associated peptide hydrolase activities of a number of cheese-related microorganisms, including P. acidipropionici, were studied by Ezzat et al. (1993).
FIG. 11. Urea-polyacrylamidegel electrophoretograms of casein (CN) and water-insoluble fractions from a number of Cheddar (C), Swiss (S), Edam (E), Gouda (G), or Blarney (D) cheeses (E. Olthoff and R. Schmidt, unpublished).
226
P. F. FOX et al.
9. Characterization of Proteolysis in Cheese
As mentioned in Section IV El, the extent of proteolysis varies from very limited, e.g., Mozzarella, to very extensive, e.g., blue-mould varieties. The use of PAGE showed that the proteolytic pattern, as well as its extent, exhibit marked intervarietal differences (Ledford et al., 1966; Marcos et al., 1979). The PAGE patterns of both the water-insoluble and water-soluble fractions are, in fact, quite characteristic of the variety, as shown in Figs. 11 and 12 for a number of Cheddar, Dutch, and Swiss-type cheeses. RPHPLC of the water-soluble fraction or subfractions thereof also shows varietal characteristics (Fig. 13). Both the PAGE and HPLC patterns vary and become more complex as the cheese matures and are in fact very useful indices of cheese maturity and to a lesser extent of its quality (O’Shea, 1993). Therefore, they have potential in the objective assessment of cheese quality. Complete characterization of proteolysis in cheese requires isolation and identification of the individual peptides. Various fractionation techniques were compared to Kuchroo and Fox (1983), some of which have been developed further by O’Sullivan and Fox (1990), Singh et al. (1994), and Fox et al. (1994). Using these techniques, many of the water-insoluble and water-soluble peptides in Cheddar cheese have been isolated and identified
FIG. 12. Urea-polyacrylamide gel electrophoretograms of casein (CN) and water-soluble fractions prepared from a number of Cheddar (C), Swiss (S), Edam (E), Gouda ( G ) , or Blarney (D) cheeses. (E. Olthoff, R. Schmidt, and P. F. Fox, unpublished).
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
227
0.16 0.14.
n
.? 0.10-
-1
::[ ~ , j $ j ~ ~ l L J J L d
0 02
q c;;'"I
000, 0 00 10 00 20 00 30 00 40 00 50 00 60 00 70 00
Minutes
0 1 2 I'
0 12
.? 0 1 0 0 08
Minutes
~ 0 1 0
I
0 0 0 1000 2000 3000 4000 5000 6000 7000
0 08
0 0 0 1000 2000 3000 4000 5000 6000 7000
Minutes
Minutes 0 20 0.18 0 16
0 14
0.14
0 12
0 12
F
I
0 02 0 00 I 0 00 10 00 20 00 30 00 40 00 50 00 60 00 70 00
Minutes
0 00 0 0 0 1000 2000 3000 4000 5000 6000 7000
Minutes
FIG. 13. Typical RP-HPLC (CHcolumn, acetonitrile/watergradient, TFA as ion pair reagent, detection at 214 nm) chromatogramsof the water-soluble fraction of Cheddar (A), Emmental (B), Edam (C), Gouda (D), Leerdamer (E), and Jarlsberg (F).
by amino acid sequencing and mass spectrometry. The results of these studies have been reported by Singh et af. (1994, 1995), Fox et af. (1994), and McSweeney et af. (1994~)and are summarized in Fig. 14. All the principal water-insoluble peptides are produced either from aSl-caseinby chymosin or from p-casein by plasmin (McSweeney et af., 1994~).In mature Cheddar (>6 months old), all the a,,-casein is hydrolyzed by chymosin at Phe23-Phe24.The peptide LyS1-CNfl-23 does not accumulate but is rapidly hydrolyzed at the bonds Gln9-Glylo and Gln13-Glu14by the lactococcal
91
A
-
93 -106
85 -92
93
24-. 85 55
25 25 25
34
75
35 39
75 75
25
75 4
7570 -7s
30
34
24 24
~1~~~
ntt t
1
* 4041 . I
14
115
-
88/89
90199
98/99
28/29 40141
1-9 13 -
110110
121/22
I
1JOBl 142/43 149/Jo 161’62 169/70 1 11 I1 I1 1 I
tt
2 3 ~ 4 ~3283 s 25-
180 -188
115 -121
=
74fl3 84/85
I
I-. I d 1 . - *
-. -. -m
70-76 70 -78
29
ai911d’~~,~~38/14
1
91)
91
8s26
-
~-indnva~cd.Ilcs
in11102 105-•
44-.
*
wtt
14919
t 1w65
‘%6”
191i92
t
1
199
119n#1
1W59
m .
41-*
FIG. 14. Identity of peptides isolated from the water-soluble fraction of Cheddar cheese. Peptides derived from aSl-casein are shown in A, those from p-casein in B. The principal chymosin cleavage sites in %,-casein, the principal plasmin cleavage sites in @casein, and the principal cleavage sites of Iactococcal cell envelope proteinase on a,,-and p-casein are shown by arrows (from Fox ef al., 1994,1995, unpublished).
.-
t6 56 9L 26 ""
65 65 65 65
CL
8L-EL EL
86
16 E6
POI
LS LS
-
O8
Z8 t8-69
801-L6
8s
69
E6
69 26 16 68
58-69 E8-69
*E -S
*-
it
6Z
g*-LS
69
*-
ES
LLI
230
P. F. FOX et al.
cell wall proteinase. A significant amount of the larger peptide asl-CN f24-199 is hydrolyzed at Leulol-Lyslo2.In mature Cheddar, -50% of the p-casein is hydrolyzed, mainly by plasmin, to y-caseins (P-CN f29-209, f106-209, and f108-209) and proteose peptones (0-CN fl-28,'fl-105, f l 107, f29-105, f29-107). These polypeptides do not appear to be hydrolyzed by chymosin or lactococcal proteinases. Although aS2-caseingradually disappears from PAGE patterns (McSweeney et al., 1993a), the polypeptides produced from it, if any, have not been identified. para-#-Casein appears to be resistant to proteolysis and no peptides produced from it have been identified. Most of the peptides in the UF retentate of the water-soluble fraction are derived from p-casein, especially from the region residues 53 to 91 (Fig. 14). In contrast, most of the peptides in the UF permeate are produced from asl-CN (Fig. 14). The N terminus of some of these peptides corresponds to a chymosin (aSl-CN)or plasmin (p-CN) cleavage site but many appear to be produced by the lactococcal cell wall proteinase. However, the N terminus and especially the C terminus of many peptides does not correspond precisely to the known cleavage sites of chymosin, plasmin, or lactococcal proteinase. This strongly suggests the action of bacterial aminopeptidases. Carboxypeptidase activity would explain why the C terminus of many peptides does not correspond to known proteinase cleavage sites but Lactococcus spp. have not been reported to possess a carboxypeptidase and there is only one report (El-Soda et al., 1978) of carboxypeptidase activity in Lactobacillus. It must be presumed that other proteinase, e.g., from NSLAB, or starter or NSLAB endopeptidases (Pep 0),are involved or perhaps other cleavage sites for lactococcal cell wall proteinase remain to be identified. The N-terminal sequence of asu,,-CN fl-9 and fl-13 is RPKHPIK; therefore, it should be susceptible to Pep X. The accumulation of these peptides in Cheddar and the apparent absence of peptides with a sequence commencing at Lys3 of crsl-CN suggest that Pep X is not active in cheese; isolated Pep X is in fact inactive on asl-CN fl-23 (W. Bockelmann and P. F. Fox, unpublished). The very small peptides in the UF permeate have not yet been identified. A number of authors (Aston and Creamer, 1986; Cliffe et al., 1993; Engels and Visser, 1994) have shown that the very small peptides ( 4 0 0 Da) make a significant contribution to Cheddar flavor; therefore, identification of the small peptides should prove interesting. Fractionation and identification of the small water-soluble peptides in cheese varieties is much less advanced than for Cheddar. Although not fractionated systematically, a large number of 12% TCA-soluble and -insoluble peptides were identified in water extracts of Parmesan by Addeo
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
231
et al. (1992,1994) using fast atom bombardment mass spectrometry. Parmesan is quite an exceptional cheese; while it undergoes extensive proteolysis and has a very high concentration of free amino acids, it contains low concentrations of medium-sized peptides (Resmini et al., 1988). Although very extensive proteolysis occurs in Blue cheese and some of the larger peptides detectable by PAGE have been partially identified (see Gripon, 1993),very little work has been done on the small (pH 4.6-soluble) peptides. The only study we are aware of is the partial identification of four PTA-soluble peptides from Gamonedo Blue cheese by Gonzalez de Llano et af. (1991). Tsuda et al. (1993) identified the dipeptide Leu-Leu in Camembert using capillary isotacophoresis. Some of the peptides resulting from the cleavage of (rSl-CNfl-23 (produced by chymosin) by lactococcal cell envelope-associated proteinase have been identified in Gouda (Kaminogawa et af., 1986; Exterkate and Alting, 1995). Proteolysis in Swiss-type cheeses has been studied using PAGE and RP-HPLC (Steffen et al., 1993; Bican and Spahni, 1993) but as far as we are aware, small peptides have not been isolated and characterized. Significant concentrations of amino acids, the final products of proteolysis, occur in all cheeses that have been investigated (see McSweeney and Fox, 1993). Relative to the level of water-soluble nitrogen, Cheddar contains a low concentration of amino acids (see Fig. 19, Section VC2). The principal amino acids in Cheddar are Glu, Leu, Arg, Lys, Phe, and Ser (Wilkinson, 1992) (Fig. 15). Parmesan contains a very high concentration of amino 3000 AM2 ClllQ5 HP
2000 0
.-
J .?
2 -5
M
5
1000
n ASP THR SER GLU PRO GLY ALA CYS VAL MET ILE LEU TYR PHE HIS LYS ARG
AMINO ACID
FIG. 15. Free amino acids in Cheddar cheese made using different starters and ripened at 10°C for 42 days (Wilkinson, 1992).
232
P. F. FOX et al,
acids which appear to make a major contribution to the characteristic flavor of this cheese (Resmini et al., 1988). The presence of amino acids in cheeses clearly indicates aminopeptidase activity;since these enzymes are intracellular, their action indicates lactococcal cell lysis. On the presumption that amino acids contribute to cheese flavor, interest is now being focussed on a search for fast-lysing lactococcal strains, e.g., heat-induced (Feirtag and McKay, 1987), phage-induced (Crow et al., 1996), or bacteriosin-induced (Morgan et al., 1995). Amino acids have characteristic flavors (Table VI); although none has a cheese-like flavor, it is believed that they contribute to the savory taste of mature cheese. 10. Catabolism of Amino Acids and Related Events
Catabolism of free amino acids probably plays some role in all cheese varieties but is particularly significant in mould- and smear-ripened varieties. The first stage in amino acid catabolism involves decarboxylation, deamination, transamination, desulfuration, or perhaps hydrolysis of the amino acid side chains. The second stage involves conversion of the resulting compounds (amines and a-ketoacids), as well as amino acids themselves, to aldehydes, primarily by the action of deaminases on amines. The final level of amino acid catabolism is the reduction of the aldehydes to alcohols or their oxidation to acids (Hemme et al., 1982). Sulfur-containing amino acids can undergo extensive conversion, leading to the formation of a number of compounds, including methanethiol and other sulfur derivatives. General pathways of amino acid catabolism are summarized in Fig. 16. The catabolism of amino acids has been reviewed by Hemme et al. (1982) and Law (1984, 1987). Decarboxylation involves the conversion of an amino acid to the corresponding amine, with the loss of COz.The presence of primary amines (e.g., tyramine) in cheese can be explained in terms of simple decarboxylation, although the presence of secondary and tertiary amines is more difficult to explain. The principal amine in cheese is tyramine (Law, 1987). The decarboxylase activities of microorganisms of significancein cheese ripening were discussed by Hemme et al. (1982). Deamination results in the formation of NH3 and a-ketoacids. Ammonia is an important constituent of a number of cheeses, such as Camembert, Gruyere, and Comte (Hemme et al., 1982). Ammonia can also be formed by the oxidative deamination of amines (yielding aldehydes), and products of complex reactions involving amino acid side chains can also be deaminated. Transamination results in the formation of other amino acids by the action of transaminases. Aldehydes formed by the above processes can then be oxidized to acids or reduced to the corresponding alcohols.
TABLE VI TASTE DESCRIPTOR AND THRESHOLD VALUES OF AMINO ACIDS (ADAPTED FROM O'CALLAGHAN,
Concentration" in Cheddar
Q Amino acid
(cal mol-I)
GlY Ser
0 300 400 500 0
Thr His ASP
Glu 4% ' Ala Met LYS Val
Leu Pro Phe TYr Ile TrP
0 750 500 1300 1500 1500 1800 2600 2500 2300 2950 3400
Taste threshold (mg 100 d - I )
(pg g-I)
(mg 100 d extract)b
Taste" l
Perception' -
90
440 610 5080 1740 340 870 2330 2020 4610 390 2400 610 470
15.6 51.1 27.4 18.6 25.7 214.4 73.4 14.3 36.7 98.3 85.2 194.5 16.5 101.3 25.7 19.8
90
-
-
-
130 150 260 20 3 5 50 60 30 50 40 190 300 90
370 1210 650
1994)
-
Sweet
Sour
*** ***
***
5
+ + +
Salt
***
*
**
***
-
***
+ + +
**
***
-
***
***
*
**
*** ** ***
+
-
Umami
*
*** *** ***
2
Bitter
*** ***
Wilkinson (1992); total concentration 24.1 mg g-' cheese.
'50 g cheese containing 37% moisture to 100 ml water.
Amino acids in Cheddar cheese are deemed to be perceived if their concentration in the water extract is greater than their threshold concentration. Asterisks indicate degree of taste sensation.
234
P. F. FOX et al. CASEINS
1 Decarboxylalion
Amines
Proleolysis
Transsmination
A d n o Acids
Aldehydes I
Alcohols
t Acids Sulphur Compounds
FIG. 16. Catabolism of amino acids in cheese (adapted from Fox et al., 1995).
Amino acid side chains can also be modified in cheese. Hydrolases can release ammonia from Asn or Gln or by the partial hydrolysis of the guanidino group of Arg, forming citrulline or its degradation to ornithine (Hemme et al., 1982). Phenol and indole, in addition to pyruvate and ammonia, can be produced by the action of C-C lyases on Tyr and Trp. Volatile sulfur compounds are found in most cheeses and can be important flavor constituents. The origin of sulfur-containing compounds is generally thought to be the sulfur-containing amino acids methionine and cysteine (Law, 1987). As Cys is rare in the caseins (occurring at low levels only in aq-and K-caseins, which are not extensively hydrolyzed in cheese), the origin of sulfur compounds must be primarily Met. Sulfur compounds formed from Met include H2S, dimethylsulfide, and methanethiol. The importance of methanethiol and related compounds in cheese aroma is discussed by Law (1987). A number of amines produced in cheese are biologically active, including tyramine, histamine, tryptamine, putrescine, cadaverine, and phenylethylamine. These biogenic amines can have imporant physiological effects for susceptible individuals,including migraine headaches and hypertension (see Section IXJ).
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
235
V. CHEESE FLAVOR
A. INTRODUCTION The flavor and texture of cheeses are their most important attributes. Generally, the former is the more important but there are exceptions, e.g., Mozzarella, which has very little flavor and is judged mainly by its textural properties, especially meltability and stretchability. Color and overall physical appearance are of some importance in all varieties; in fact, poor appearance and discoloration may be the most important attributes since they are the attributes by which the consumer initially assesses cheese quality or acceptability. Cheese flavor has been the subject of scientific investigation since the beginning of this century. Initially, it was believed that a single compound might be responsible for cheese flavor but according to the “Component Balance Theory” (Mulder, 1952; Kosikowski and Mocquot, 1958), cheese flavor results from the correct balance and concentration of numerous sapid and aromatic compounds. During the intervening 40 years, there has been extensive research on the flavor of several cheese varieties, but complete information is not yet available on the flavor chemistry of any specific variety. The extensive literature on cheese flavor has been reviewed by Reiter et al. (1966), Kristoffersen (1973, 1985), McGugan (1975), Aston and Dulley (1982), Adda et al. (1982), Lawrence et al. (1983), Olson (1990), Urbach (1993), Fox (1994), and Fox et al. (1995). Although it is not possible to describe the flavor of cheese in precise chemical terms, very considerable progress has been made on the identification of flavor compounds in cheese and elucidation of the biochemical pathways by which these compounds are produced. It is generally recognized that the aroma of cheese is primarily in the volatile fraction while taste is largely in the aqueous phase; until recently, most researchers focussed on the volatile fraction. Intervarietal comparisons should be a valuable approach toward identifying key flavor compounds. Although several such studies on the volatile compounds have been reported, there have been relatively few comparative studies on the aqueous phase. One of the major problems encountered in research on cheese flavor is defining what the typical flavor should be. Within any variety, a fairly wide range of flavor and textural characteristics is acceptable; this is particularly so for Cheddar which makes it especially difficult to chemically define its flavor. In cheese factories, wholesale or retail outlets and research laboratories, somebody decides what constitutes desirable and undesirable flavor, which may not be typical. Only recently have systematic attempts been
236
P. F. FOX et al.
made to objectively describe the sensoric attributes of cheese, e.g., McEwan et al. (1989), Muir and Hunter (1992), and Hirst et al. (1994). An international study in the EU FLAIR-SENS programme (FLAIR Concerted Action No. 2, Cost 902, Relating Instrumental, Sensory and Consumer Data) had a similar objective, especially for cheese varieties with Appelation d’Origine Contr6lke status. An agreed vocabulary is essential if the results of instrumental studies are to be related to the sensoric attributes and quality of cheese. Perhaps not surprisingly, chemical definition of off-flavors has been more successful than the definition of desirable flavors because off-flavors usually have a fairly well-defined cause, e.g., bitterness (peptides), rancidity (fatty acids), fruitiness (esters), and the specific compound(s) responsible can, usually, be identified. Since cheese texture has a major impact on flavor perception, these attributes should, ideally, be considered together. For example, it has been suggested (McGugan et al., 1979) that the main contribution of proteolysis to cheese flavor is due to its effect on cheese texture which affects the release of sapid compounds on mastication of the cheese. However, these two aspects of cheese quality are rarely part of the same investigation and cheese texture is even less well understood at the molecular level than cheese flavor. This section will deal with three aspects of cheese flavor: (1) analytical techniques, (2) intervarietal comparisons, and (3) effect of composition on cheese quality. The biochemical pathways through which the principal flavor compounds are generated were described in Section VI.
B. ANALYTICAL METHODS 1. Nonvolatile Compounds
Although studies on cheese flavor date from the beginning of this century, the techniques available prior to the development of gas chromatography (GC) in the 1950s were inadequate to permit significant progress. Early investigators recognized the important contribution of proteolysis and lipolysis to cheese ripening. Studies on proteolysis relied on changes in protein/ peptide solubility, e.g., in water, pH 4.6 buffers, TCA, ethanol, etc. Such techniques, which have been reviewed by Fox (1989a), IDF (1991a), and McSweeney and Fox (1993), are still widely used as useful indices of cheese maturity but are less effective indices of cheese quality (Aston et al., 1983). Since the products of proteolysis are nonvolatile, they contribute to cheese taste but not to aroma.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
237
More specific studies on proteolysis commenced with the development of paper and ion-exchange chromatography in the 1950s. Large watersoluble and insoluble peptides are best characterized by electrophoresis, especially PAGE, which was first applied to cheese by Ledford et al. (1966) and is now used widely for this purpose (for reviews, see Shalabi and Fox, 1987; Creamer, 1991; Strange et af., 1992). Isoelectric focusing has had limited application in studies on cheese ripening and is not particularly effective. Recently developed capillary electrophoresis is a very powerful technique but has had very little application to date in cheese research. The small peptides in cheese can be fractionated by various forms of chromatography, e.g., gel permeation, ion-exchange, and especially RPHPLC. Using these techniques, more than 200 peptides have been demonstrated in Cheddar cheese, many of which have been isolated and identified (see Section IVE). Free amino acids are usually quantified by ion-exchange HPLC with post-column derivitization using ninhydrin or by separation of fluorescent amino acid derivatives by RP-HPLC. The total concentration of free fatty acids is usually determined by extractionhitration methods or spectrophotometrically as Cu soaps. Early attempts to quantify the concentration of individual short-chain fatty acids involved steam distillation and adsorption chromatography. Complete separation and quantitation of free fatty acids can be achieved by GC, usually as their methyl esters, for which several preparative techniques have been published. Free fatty acids are major contributors to the flavor of some varieties, e.g., Romano, Feta, and Blue; in the latter, up to 25% of the total fatty acids may be in the free form. Short chain fatty acids are important contributors to cheese aroma, while longer chain acids contribute to taste. Excessive concentrations of either cause off-flavors (rancidity) and the critical concentration is quite low in many varieties, e.g., Cheddar and Gouda. Several other organic acids, especially lactic, are present in cheese, and are routinely analyzed by HPLC. Enzymatic methods are available for several acids, e.g., D and L lactic (the easiest method available to distinguish between the isomers), acetic, pyruvic, and succinic (Boehringer-Mannheim, 1986). 2. Volatile Compounds Compounds responsible for cheese aroma are volatile. While some preliminary work on the volatile constituents of cheese was done before 1960, e.g., short chain fatty acids and amines, significant progress was not possible until the development of GC in the 1950s. GC was first applied to the study of Cheddar cheese volatiles by Scarpellino and Kosikowski (1962) and McGugan and Howsam (1962), who used vacuum distillation and cold
238
P. F. FOX et al.
trapping to recover volatiles and GC with packed columns and thermal conductivity detectors to resolve and identify them. The introduction of flame ionization detectors, capillary columns, and interfacing GC with mass spectrometry (MS) markedly increased sensitivity and greatly extended the number of compounds detected, e.g., 200 for Cheddar (Aishima and Nakai, 1987). Vacuum distillation at ca. 70°C may generate artefacts and was replaced by head-space analysis. To increase sensitivity, head-space volatiles may be trapped, e.g., Tenax traps which can be inserted directly into the port of the GC (see Bossett and Gauch, 1993). Based on a survey of the published literature, Maarse and Vischer (1989) listed 213 volatile compounds that had been identified in 50 studies on Cheddar; these included 33 hydrocarbons, 24 alcohols, 13 aldehydes, 17 ketones, 42 acids, 30 esters, 12 lactones, 18 amines, 7 sulfur compounds, 5 halogens, 6 nitriles and amides, 4 phenols, 1 ether, and 1 pyran. The concentrations of many of these compounds were reported. The principal volatile compounds identified in Cheddar are listed in Table VII. Thus, a great diversity of potentially sapid and/or aromatic compounds have been identified in one or more cheese varieties-these include small TABLE VII SIXTY-ONE VOLATILE COMPOUNDS WHICH HAVE BEEN IDENTIFIED IN CHEDDAR CHEESE' ~
~
Acetaldehyde Acetoin Acetone Acetophenone P-Angelicalatone 1, 2-Butanediol n-Butanol 2-Butanol Butanone n-Butyl acetate 2-Butyl acetate n-Butyl butyrate n-Butyric acid Carbon dioxide pCresol y-Decalactone SDecalactone n-Decanoic acid Diacetyl Diethyl ether
Dimethyl sulfide Dimethyl disulfide Dimethyl trisulfide SDodecalactone Ethanol Ethyl acetate 2-Ethyl butanol Ethyl butyrate Ethyl hexanoate 2-Heptanone n-Hexanal n-Hexanoic acid n-Hexanol 2-Hexanone Hexanethiol 2-Hexenal Isobutanol Isohexanal Methanethiol Methional
'Adapted from Urbach (1993).
Methyl acetate 2-Methylbutanol 3-Methylbutanol 3-Methyl-2-butanone 3-Methylbutyricacid 2-Nonanone SOctalactone n-Octanoic acid 2-Octanol 2,CPentanediol n-Pentanoic acid 2-pent an
Pentan-2-one n-Propanol Propanal Propenal n-Propyl butyrate Tetrahydrofuran Thiophen-2-aldehyde 2-Tridecanone 2-Undecanone
CHEESE PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
239
peptides (200 or more), amino acids and more than 200 volatile compounds (fatty acids, other acids, carbonyls, amines, sulfur compounds, and hydrocarbons). All these classes of compounds occur in most or all cheeses-what appears to matter most is the absolute and relative concentrations of these compounds, not any major compound, or even class of compounds. 3.
Uff--flavorsin Cheese
Specific flavor defects are frequently encountered in cheese. While the desirable flavor of cheese has been difficult to define precisely in chemical terms, the specific cause(s) of many defects has been established more or less definitively. The principal flavor defects in cheese are described below. a. Bitterness. Bitterness is probably the principal taste defect in cheese. Although amino acids, amines, amides, substituted amides, long-chain ketones, some monoglycerides, N-acyl amino acids, and diketopiperazines may contribute to bitterness (Ney, 1979a; Roudot-Algaron et al., 1993), this defect in cheese usually results from the accumulation of hydrophobic peptides (Fox et al., 1995). Ney (1979b) suggested that the mean hydrophobicity (Q = XAfi/n, where AJ is side chain hydrophobicity and n is the number of residues) of a peptide, rather than any particular amino acid sequence, is of importance in bitterness. Although further work suggested that the nature of the terminal amino acids and certain steric parameters influence the perception of bitterness (see Lemieux and Simard, 1991, 1992), the mean hydrophobicity of a peptide remains the single most important factor determining its bitterness. Peptides of mol wt 0.1 to 10 kDa and Q < 1300 cal residue-’ are not bitter while those with Q > 1400 cal residue-’ and mol wt from 0.1 to 6 kDa are bitter (Ney, 1979b); above 6 kDa, even peptides with a Q value greater than 1400 cal residue-’ are not bitter. Hydrolysates of proteins with a high mean hydrophobicity are likely to contain bitter peptides although the distribution of hydrophobic residues along the polypeptide and the specificity of the proteinase used to prepare the hydrolysate also influence the development of bitterness (Adler-Nissen, 1986). Since the caseins, especially P-casein, are quite hydrophobic and the hydrophobic residues are clustered, casein hydrolysates have a high propensity to bitterness. As discussed in Section IVE, cheese contains a great diversity of proteinases and peptidases with different and complementary specificities. Although detailed kinetic studies are lacking, at least some of the peptides in cheese are transient and hence bitterness may be transient as bitter peptides are formed and hydrolyzed, or masked by other sapid compounds. It is very likely that all cheeses contain bitter peptides, which probably
240
P. F. FOX et al.
contribute positively to the overall desirable flavor; bitterness becomes a problem only when bitter peptides accumulate to excessive, unbalanced or &casein, levels. Although bitter peptides can originate from either as]the primary action of chymosin and/or lactococcal cell envelope-associated proteinase (CEP) on the very hydrophobic C-terminal region of P-casein may result in the early production of bitter peptides, while the peptides produced initially from asl-casein are generally not hydrophobic. The production of bitter peptides also depends on the specificity of the lactococcal CEP; e.g., PIII-typeCEP (Lc. lactis ssp. cremoris AM1) produced less bitter casein hydrolysates than P,-type CEP (Lc. luctis ssp. cremoris HP) (Visser et al., 1983), perhaps due to the initial release of more peptides from the hydrophobic C-terminal region of P-casein by the latter. The concentration of bitter peptides depends on the rate at which they are degraded by lactococcal peptidases ( S . Visser, 1993) and perhaps, in the case of larger bitter peptides, by the CEP. The debittering effect of aminopeptidase N on a tryptic digest of P-casein was demonstrated by Tan et al. (1993b). Certain starters have a propensity to cause bitterness (Lowrie et al., 1972; Lawrence et al., 1972; Stadhouders, 1974). Nonbitter cheese can be made using these strains provided they are used in combination with “nonbitter” strains. Since chymosin may cause the release of bitter peptides, factors that affect its retention in cheese curd (type and quantity used in cheesemaking, drain pH, and cook temperature) will influence the development of bitterness. Certain rennet substitutes produce bitter cheese, owing to excessively high activity andlor incorrect specificity. The pH of cheese also influences the activity of residual coagulant and other enzymes. Cheese with a low salt concentration is very prone to bitterness (Stadhouders et al., 1983; Visser er al., 1983),perhaps because the susceptibilityof p-casein to hydrolysis by chymosin, with the production of the bitter peptide, P-CN f193-209, is strongly affected by the NaCl concentration in cheese (Kelly, 1993). Salt also inhibits lactococcal CEP (Exterkate, 1990) and may promote the aggregation of large, nonbitter hydrophobic peptides which would otherwise be degraded to bitter peptides (Visser, 1993).Bitterness can be particularly problematic in low-fat cheeses, presumably as a result of reduced partitioning of hydrophobic peptides into the fat phase. Bitter peptides that have been isolated from cheese are summarized in Table VIII. As expected, bitter peptides originate principally from hydrophobic regions of the caseins, e.g., sequences 14 to 34, 91 to 101, and 143 to 151 of Lysl-casein, and 46 to 90 or 190 to 209 of 0-casein. As discussed by McSweeney et al. (1996) the majority of these peptides show evidence of some degradation by lactococcal proteinases and/or peptidases. b. Astringency. Astringency is a taste-related phenomenon perceived as a dry feeling in the mouth and a puckering of the oral tissue (Lindsay,
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
241
TABLE VIII BITTER PEPTIDES ISOLATED FROM CHEESE“
Hydrophobicity Cheese Cheddar
Origin a,,-CN f14-17 a,l-CN f17-21 asl-CN f26-32 a,,-CN f26-33 P-CN f46-67
B-CN f46-84
Gouda
Alpkase Butterklse
P-CN f193-209 P-CN f84-89 P-CN f193-207 0-CN f193-208 P-CN f193-209 a,,-CN f198-199 P-CN f61-69
Amino acid sequence
(Q,cal residue-’)
E.V.L.N. N.E.N.L.L A.P.F.P.E.V.F A.P.F.P.E.V.F.G Q.D.K.I.H.P.F.A.Q.T.Q.S.L.V.Y.P. F.P.G.P.1.P Q.D.K.I.H.P.F.A.Q.T.Q.S.L.V.Y.P. F.P.G.P.I.P.N.S.L.P.Q.N.I.P.P.L. T.Q.T.P.V.V.V
1162.5 1074.0 1930.0 1688.8 1580.5
Y.Q.Q.P.V.L.G.P.V.R.G.P.F.P.I.1.V V.P.P.F.L.Q Y.Q.Q.P.V.L.G.P.V.R.G.P.F.P.1 Y.Q.Q.P.V.L.G.P.V.R.G.P.F.P.I.1
Y.Q.Q.P.V.L.G.P.V.R.G.P.F.P.I.1.V L.W P.F.P.G.P.1.P.N.S
1508.5
1762.4 1983.3 1686.7 1766.9 1762.4 2710.0 1792.2
Adapted from Lemieux and Simard (1992).
1985). Astringency usually involves the formation of aggregates or precipitates between tannins (polyphenols) and proteins in the siliva (Lindsay, 1985). Harwalkar and Elliott (1971) isolated an astringent fraction from Cheddar cheese by extraction with chloroform-methanol but did not characterize it. Since these authors monitored the chromatographic separation of astringent and bitter compounds by absorbance at 280 nm, it appears that they may be peptides. The aqueous fraction of Comt6 contains N propionyl methionine, which is slightly bitter, astringent, and pungent, and also fractions which had umami or pungent tastes (Roudot-Algaron et al., 1993). c. Fruitiness. The principal compounds responsible for fruitiness in Cheddar are ethyl butyrate and ethyl hexanoate, formed by esterification of free fatty acids (FFAs) with ethanol: production of ethanol appears to be the limiting factor as FFAs are present in cheese at relatively high concentrations. Ethyl esters are present in low concentrations in nonfruity cheeses and thus the fruity defect occurs as a result of excessive production of ethanol or its precursors.
d. “Unclean” Off-Flavors. The origin of “unclean” and related flavors in Cheddar was investigated by Dunn and Lindsay (1985) who quantified
242
P. F. FOX
et
al.
a number of Strecker-type compounds, including phenylacetaldehyde, phenethanol, 3-methyl butanol, 2-methyl propanol, phenol, and p-cresol. Phenylacetaldehyde concentrations were elevated in cheeses with an “uncleanrosy” off -flavor and the addition of this compound to clean-flavored mild Cheddar reproduced this defect. At higher concentrations (> 500 pdkg), phenylacetaldehyde imparted astringent, bitter, and stinging flavors to cheese. Concentrations of phenethanol were similar in most of the cheeses studied (ca. 100 pglkg). p-Cresol imparted a “utensil”-type flavor when present at high concentrations. Dunn and Lindsay (1985) also discussed the potential of short-chain fatty acids to potentiate the flavor impact of p-cresol. Branched chain Strecker-type aldehydes (3-methyl butanal, 2-methyl butanal, and 2-methyl propanal) did not cause flavor defects when added at concentrations below 200 pg/kg to clean-flavored cheese. Phenol contributed to the unclean flavor of cheese and indeed enhanced the sharpness of Cheddar flavor. C. INTER- AND INTRAVARIETAL COMPARISON O F CHEESE RIPENING As discussed in Section VB, most ripened cheeses contain essentially the same sapid and aromatic compounds but at different concentrations and proportions. Therefore, it appears reasonable to presume that inter- and intravarietal comparison, especially of closely related varieties, might help to identify compounds most likely to contribute to characteristic cheese flavors. However, although both the water-soluble and volatile fractions of several cheese varieties have been analyzed, there are relatively few intervarietal comparisons, especially of the water-soluble fraction. In this section, the results of some such studies will be discussed. 1. Gel Electrophoresis
Gel electrophoresis, especially alkaline (pH 9.0) urea-PAGE, is the best method for characterizing the large water-insoluble (WISF) peptides and is also useful for characterizing the larger water-soluble (WSF) peptides. Ledford et al. (1966) and Marcos et al. (1979) showed differences between the electrophoretograms of 7 and 34 cheese varieties, respectively, by ureaPAGE; both studies involved only one sample, of unknown age, history, or quality, for each variety and hence may not have been typical. We have made several comparative studies on Cheddar cheese in relation to starter type, manufacturer, ripening temperature, age, and quality. In one such study, urea-PAGE of the WSF and WISF from 22 6-monthold cheeses made using single strain starters showed very little difference
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS 243
between the cheeses, which differed substantially in quality; perhaps, this is not surprising since starter proteinases contribute little to primary proteolysis in Cheddar, as detected by PAGE. In another study (McSweeney et al., 1993a), PAGE failed to show substantial differences between cheeses made from pasteurized, ultrafiltered, or raw milk, although the flavor of the raw milk cheese was very much more intense than that of the others. O’Shea (1993) characterized proteolysis in 60 commercial Cheddar cheeses ranging in age from 3 to 33 months and varying in quality. Urea-PAGE of the WISF generally reflected the age of the cheese and was a useful index of its textural quality (if the age was known), but not of its flavor. Since an elevated ripening temperature accelerates proteolysis (Folkertsma et al., 1996), the history of the cheese must be known if PAGE is to be a useful index of the age of cheese. Electrophoretograms are fairly characteristic of the variety, as shown in Figs. 11 and 12 for a selection of commercial samples of Cheddar, Swiss, and Dutch varieties. Studies of more samples and a wider range of varieties are warranted. The peptides detected by PAGE are probably too large to have a direct impact on cheese flavor but probably do reflect the overall ripening process and hence might be useful as an indirect index of quality, including maturity. 2. High-Performance Liquid Chromatography HPLC is more effective than PAGE, which does not detect peptides less than -3000 Da, for analysis of the WSF. HPLC profiles of the UF permeate and retentate of a mild, mature, and extramature Cheddar are shown in Fig. 17; clearly, the complexity of the chromatograms, especially of the retentate (medium-sized peptides), increased as the cheese matured. Unfortunately, it is not possible at present to relate cheese flavor or texture to HPLC chromatograms. HPLC of water- or TCA-soluble peptides shows clear varietal characteristics (Fig. 13). Engels and Visser (1994) concluded that very small peptides and free amino acids contributed significantly to the flavor of Cheddar, Edam, Gouda, Gruykre, Maasdam, Parmesan, and Proosdij. Glu, Leu, and Phe were the principal free amino acids, in agreement with other studies, e.g., Wilkinson (1992). The concentration of free amino acids (measured by reaction with Cd-ninhydrin) in Cheddar is highly correlated with age (Fig. 18) and hence with the intensity of cheese flavor. The ratio of free amino acids to water-soluble nitrogen (WSN) appears to be characteristic of the variety (Fig. 19); Cheddar appears to contain a very low level of free amino acids relative to small peptides. Parmesan contains a particularly high concentration of amino acids (Engels and Visser, 1994) which have a major effect of its flavor (Resmini et ai., 1988).
244
P. F. FOX el al.
2
0
5
10
15
20
25
30
35
40
0
5
10
15
20
25
30
35
40
0
5
10
15
30
35
40
20
25
b
10
0
O
20
20
10
10
20
30
30
30
40
40
40
50
50
50
60
60
60
76
70
70
FIG. 17. Reverse-phase HPLC (C, column, acetonitrile/water gradient, TFA as ion pair reagent, detection at 214 nm) profiles of (1) permeates and (2) renetates from (a) mild, (b) mature, and (c) extra mature Cheddar cheese (from O’Shea, 1993).
The WSF also contains short chain fatty acids (< Cq0)which impart a “cheesy” aroma; there are substantial intervarietal differences with respect to short chain fatty acids (Engels and Visser, 1994). This is the only study we are aware of on the free fatty acids in the WSF of cheese, on which further work appears to be warranted.
3. Cheese Volatiles A number of intra- and intervarietal comparisons of cheese volatiles have been published. An early example is that of Manning and Moore (1979) who analyzed head-space volatiles of nine fairly closely related varieties; considerable intervarietal differences were evident but the four samples of “Cheddar” also differed markedly. The intensity of cheese flavor was reported to be related to the concentration of sulfur compounds (peaks 1 and 2); 2-pentanone was also considered to be important for Cheddar cheese flavor.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
245
0 0
10
20
30
40
Age (months) FIG. 18. Concentration of total free amino acids (Cd-ninhydrin assay) in Cheddar cheese as a function of age (from O’Shea, 1993).
A more comprehensive study on Cheddar, Gouda, Edam, Swiss, and Parmesan (total of 82 samples) was reported by Aishima and Nakai (1987). The volatiles were extracted by CH2CI2and analyzed by GC. More than 200 peaks were resolved in every chromatogram, 118of which were selected as variables for discriminative analysis. Expression of the area of each of the 118 peaks as a percentage of total chromatogram area clearly permitted classification of the five varieties. The compounds likely to be responsible for the characteristic flavor of each variety were not discussed. Bosset and Gauch (1993) concentrated the head-space volatiles from six cheese varieties by a “purge and trap” method for analysis by GC-MS; a total of 81 compounds were isolated and identified (Fig. 20), 20 of which were found in all six varieties and a further nine in five of the six varieties. The authors concluded that “practically all types of cheese analyzed contain more or less the same constituents, but at varying concentrations” and that the flavor of these cheeses seems to depend not on any particular key compound, but rather on a “critical balance” or a “weighted concentration ratio of all components present.” Thus, the component balance theory still applies and we appear to have advanced relatively little during the last 40 years. While analytical techniques have improved greatly during that period and data are available on
246
P. F. FOX el al. 3
1
‘
0
0 Gouda
0
0 Edam
0
0
0 Swiss
0
2-
0
0
0 0
0 0 O 0
.
0
0
1-
4 0 u.
a D
I
0
I
o ! . 0.0
1
0.1
.
1
0.2
. ‘ . ‘ . ‘ . I 0.3
0.4
0.5
0.6
% WSN
FIG. 19. Relation between the concentration of total free amino acids by the Cd-ninhydrin assay, A507, and water-soluble nitrogen for some cheese varieties (E. Olthoff, R. Schmidt, and P. F. Fox, unpublished).
the concentrations of many flavor compounds present in cheese, we still do not know what the critical compounds are, if any. Further intervarietal comparisons may be useful if quantitative data are provided, although usually they are not. Perhaps it would be fruitful to reinvestigate cheeses with controlled microflora. There have been no studies on such models since the 1970s and these were concerned mainly or totally with proteolysis. It would seem to be particularly useful to combine studies on controlled microflora cheese with intervarietal comparisons but perhaps such an undertaking is beyond the capabilities of a single laboratory. D. FACTORS THAT AFFECT CHEESE QUALITY
As discussed in Section IV, the ripening of cheese, and hence its quality, is due to the activity of microorganisms and enzymes from four or five sources. Therefore, it might reasonably be expected that it should be possible to produce premium quality cheese consistently by controlling these agents; however, in spite of considerable research and quality control efforts, it is not yet possible to do so.
-
-
j
“5
Ed> %LA,
248
P. F. FOX er al.
Certain factordagents can be manipulated easily and precisely, while others are more difficult to control. In this section, possible factors responsible for the variability of cheese quality and how they might be controlled will be considered. Interactions (Fig. 21) between these factors were reviewed by Lawrence and Gilles (1980).
1. Milk Supply It is well recognized that the quality of the milk supply has a major impact on the quality of the resultant cheese. Three aspects of quality must be considered: microbiological, enzymatic, and chemical.
a. Microbiological. In developed dairying countries, the quality of the milk supply has improved markedly during the past 30 years-total bacterial count (TBC) is now usually <20,000 CFU/ml exfarm. The TBC probably increases during transport and storage at the factory, but growth can be minimized by thermization (65°C X 15 sec) of the milk supply, as is standard practice in some countries.
Breed of Cow
Stage of lactation
Climate
I
Milk comwsition
PetP nC --:A
Residual rennet
Starter activity in cheese
00
Growth of adventitious bacteria
1
0 0
/==
Storage temperature
Texture
Flavour development
FIG. 21. Principal factors that affect the quality of Cheddar cheese.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
249
Although many cheeses are made from raw milk, in quantitative terms, most cheese is made from milk pasteurized at or close to 72°C X 15 sec. If produced from good quality raw milk and if subsequently handled under hygienic conditions, pasteurized milk should have very low TBCs and therefore represents a very uniform product from a microbiological viewpoint. b. Indigenous Enzymes. Milk contains as many as 60 indigenous enzymes, the significance of which to cheese quality has not yet been researched adequately. Several of these enzymes have the potential to affect cheese quality, especially lipase, proteinase, acid phosphatase, and perhaps xanthine oxidase, sulfydryl oxidase, lactoperoxidase, and y-glutamyl transpeptidase. Most of these survive HTST pasteurization to a greater or lesser extent and at least some, e.g., proteinase (plasmin), are active during cheese ripening. Although precise information is lacking, it is our opinion that indigenous milk enzymes are not major causes of variability in cheese quality, although some contribute to cheese ripening and may contribute to the superior quality of raw milk cheese, a possibility that warrants investigation.
c. Chemical Composition. The chemical composition of milk, especially the concentrations of casein, fat, calcium, and pH, has a major influence on several aspects of cheese manufacture, especially rennet coagulation, gel strength, curd syneresis, and hence cheese composition. When seasonal milk production is practiced, as in New Zealand, Ireland, and parts of Australia, milk composition varies widely but there is some variability even with random calving patterns. It is possible to reduce, but not eliminate, the variability in the principal milk constituents by standardizing with respect to fat and casein content, not just the ratio (protein content can be standardized by adding UF retentate), the pH (using gluconic acid Glactone), and the calcium content (by adding CaC12). 2.
Coagulant (Rennet)
It is generally accepted that calf chymosin produces the best quality cheese. An adequate supply of chymosin from genetically engineered microorganisms is now available and therefore rennet quality should not be a factor in cheese quality. As discussed in Section IVE4, the proportion of added rennet retained in cheese curd varies with rennet type, cook temperature, and drain pH; these variables should be standardized if cheese of consistent quality is to be produced. Johnston et al. (1994) found that increased retention of the coagulant in the curd resulted in greater initial hydrolysis of aSl-casein
250
P. F. FOX et
af.
although this did not appear to be reflected in sensory assessment of cheese texture. It has been suggested (Exterkate and Alting, 1995) that the activity of chymosin in cheese curd is the limiting factor in cheese ripening; however, excessive rennet activity leads to bitterness. There have been relatively few studies on the significance of chymosin activity to cheese quality, an aspect which appears to warrant further research. 3. Starter
Since the starter plays a key role in cheese manufacture and ripening, it might reasonably be expected that differences between the enzyme profile of starter strains affect cheese quality. Modern single-strain starters produce acid very reproducibly and if properly managed, show good phage resistance. Lactococcus strains have been selected mainly on the basis of acidproducing ability, phage resistance, and compatibility. Empirical studies have indicated strains that produce unsatisfactory, especially bitter, cheese but systematic studies on strains with positive cheesemaking attributes are lacking. This probably reflects the lack of information on precisely what attributes of a starter are desirable from a flavor-generating viewpoint. Studies on genetically engineered strains that superproduce proteinase and/ or the principal aminopeptidase (Pep N) showed that cheese quality was not improved although proteolysis was accelerated (McGarry d al., 1994). Since all lactococcal enzymes, except the cell wall-associated proteinase, are intracellular or membrane associated, the cells must lyse before these enzymes can participate in ripening; therefore, the lytic rate of Lactococcus strains is being studied, with the objective of selecting strains with improved cheesemaking properties (see Section VIIC). Sulfur compounds have long been considered as contributors to Cheddar cheese flavor. Some strains of Lc. lactis ssp. cremoris, but not Lc. lactis ssp. la&, can absorb glutathione (y-Glu-Cys-Gly; GSH) from the growth medium (Fernandes and Steele, 1993). Release of GSH into the cheese on cell lysis may affect the redox potential (Eh) of cheese, and hence the concentration of thiol compounds. Cheesemaking studies using starter strains that accumulate glutathione or those that do not are warranted. It is very likely that the desirable cheesemaking properties of starters are due to a balance between certain, perhaps secondary, enzymatic activities, which have not yet been identified. 4. Nonstarter Lactic Acid Bacteria
The significance of lactobacilli for Cheddar cheese quality is controversial (see Sections IVD3 and IVE7). Many researchers consider their contribu-
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
251
tion to be negative (in the Netherlands, a maximum of 2 X lo6 NSLAB/g is specified for Gouda). Although there are several studies on controlled microflora cheeses, we are not aware of studies in which cheese free of NSLAB was compared with “control” cheeses containing “wild” NSLAB. McSweeney et af. (1994a) reported comparative studies on cheese made under aseptic conditions using Lactococcus starter alone or with selected Lactobacillus adjuncts; preliminary results suggested that inoculation of cheese milk with selected strains of Lactobacillus improves cheese flavor and possibly accelerates ripening. These findings confirm those of other researchers; Lactobacillus adjuncts are used commercially in Canada and are being investigated by starter supply companies. Since the numbers and strains of lactobacilli in cheese are uncontrolled, it is likely that they contribute to variability in cheese quality. Since it is impossible to eliminate NSLAB completely, even under experimental conditions, it appears worthwhile to determine what factors affect their growth. The number of NSLAB in Cheddar is strongly influenced by the rate at which the curd in cooled and subsequently ripened (Folkertsma et al., 1996). The growth of NSLAB does not appear to be influenced by the concentration of NaCl in the cheese (Turner and Thomas, 1980). It is likely that the moisture content of cheese affects the growth of NSLAB but we are not aware of studies in which this effect has been studied. NSLAB grow mainly after the lactose has been metabolized by residual starter activity. Although the growth substrates in cheese for Lactobacillus are not known, it is likely that they are limited (NSLAB normally plateau at -lo7 CFU/g) and hence it might be possible to outcompete wild NSLAB by adding selected strains of Lactobacillus, thereby offering better control.
5. Cheese Composition The quality of cheese is influenced by its composition, especially moisture content, NaCl concentration (preferably expressed as S/M), pH, moisture in nonfat substances (MNFS; essentially ratio of protein :moisture) and % fat in dry matter (FDM). At least five studies (O’Connor, 1971; Gilles and Lawrence, 1973; Fox, 1975; Pearce and Gilles, 1979; Lelievre and Gilles, 1982) have attempted to relate the quality of Cheddar cheese to its composition. While these authors agree that moisture content, %S/M and pH are the key determinants of cheese quality, they disagree as to the relative importance of these parameters. O’Connor (1971) found that flavor and aroma, texture, and total score were not correlated with moisture content but were significantly correlated with %NaC1 and particularly with pH. Salt content and pH were themselves strongly correlated, as were salt and moisture.
252
P. F. FOX et al.
Based on the results of a study on experimental and commercial cheeses in New Zealand, Gilles and Lawrence (1973) proposed a grading scheme which has since been applied commercially in New Zealand for young (14 day) Cheddar cheese. The standards prescribed for Premium grade were: pH, 4.95-5.10; %S/M, 4.0-6.02%; MNFS, 52-56%; FDM, 52-55%. The corresponding values for First grade cheeses were: 4.85-5.20%, 2.5-6%, 5057%, and 50-56%; young cheeses with a composition outside these ranges were considered unlikely to yield good quality mature cheese. Quite wide ranges of FDM are permitted; Lawrence and Gilles (1980) suggested that since relatively little Iipolysis occurs in Cheddar cheese, fat content plays a minor role in determining cheese quality but if FDM fell below about 48%, the cheese was noticeably more firm and less attractive in flavor. Pearce and Gilles (1979) found that the grade of young (14-day-old) cheeses produced at the New Zealand Dairy Research Institute was most highly correlated with moisture content; the optimum compositional ranges were: MNFS, 52-54%; %S/M, 4.2-5.2%; pH, 4.95-5.15%. Fox (1975) reported a poor correlation between grade and moisture, salt, and pH for Irish Cheddar cheeses but a high percentage of cheeses with compositional extremes were downgraded, especially those with low salt (<1.4%), high moisture (>38%), or high pH (>pH 5.4). Salt concentration seemed to exercise the strongest influence on cheese quality and the lowest percentage of downgraded cheeses can be expected in the salt range 1.61.8%(S/M, 4.0-4.9%); apart from the upper extremes, pH and moisture appeared to exercise little influence on quality. High salt levels tended to cause curdy textures, probably due to insufficient proteolysis; pasty body, often accompanied by off-flavors, was associated with low salt and high moisture levels. In the same study, the composition of extramature cheeses was found to vary less and the mean moisture content was 1% lower than that of regular cheeses. A very extensive study of the relationship between the composition and quality of nearly 10,000cheeses produced in five commercial New Zealand factories was reported by Lelievre and Gilles (1982). As in previous studies, considerable compositional variation was evident but was less for some factories than others. While the precise relationship between quality and composition varied between plants, certain generalizations emerged: (1) within the compositional range suggested by Gilles and Lawrence (1973) for premium quality cheese, composition does not have a decisive influence on grade, which decreases outside this range; (2) composition alone does not provide a basis for grading as currently acceptable in New Zealand (3) MNFS was again found to be the principal factor affecting quality; (4) within the recommended compositional bands, grades declined marginally as MNFS increased from 51 to 55%, grades increased slightly as %S/M
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECI'S
253
decreased from 6 to 4%, pH had no consistent effect within the range 4.9-5.2, and FDM had no influence in the range 50-57%. The authors stress that since specific interplant relationships exist between grade and composition, each plant should determine the optimum compositional parameters pertinent to that plant. One could summarize the results of the foregoing investigations as indicating that high or low values for moisture, salt, and pH lead to flavor and textural defects. The desired ranges suggested by Gilles and Lawrence (1973) appear to be reasonable, at least for New Zealand conditions, but within the prescribed zones, composition is not a good predictor of cheese quality. Presumably, several other factors, e.g., microflora, activity of indigenous milk enzymes, relatively small variations in milk composition, and probably other unknown factors, influence cheese quality but become dominant only under conditions where the principal determinants, moisture, salt, and pH, are within appropriate limits. Although the role of calcium concentration in cheese quality has received occasional mention, its significance was largely overlooked until the work of Lawrence and Gilles (1980) who pointed out that the calcium level in cheese curd determines the cheese matrix and, together with pH, indicates whether proper procedures were used to manufacture a specific cheese variety. As the pH decreases during cheese manufacture, colloidal calcium phosphate dissolves and is removed in the whey. The whey removed at running constitutes 90-95% of the total whey lost during cheesemaking and this whey contains, under normal conditions, -85% of the calcium and -90% of the phosphorus lost. Thus, the calcium content of cheese reflects the pH of the curd at whey drainage. Lawrence and Gilles (1982) showed strong correlations between the calcium content of cheese and the pH at 1 day, pH at 14 days, and the amount of starter used. Since the pH of cheese increases during ripening, the pH of mature cheese may be a poor index of the pH of the young cheese. Therefore, calcium concentration is probably a better record of the history of a cheese with respect to the rate of acidification than the final pH. Reduction in calcium phosphate concentration by excessively rapid acid development also reduces the buffering capacity of cheese and hence the pH of the cheese will fall to a lower value for any particular level of acid development. Unfortunately, no recent work on the levels and significance of calcium in Cheddar cheese appears to be available. 6. Ripening Temperature
The final factor known to influence the rate of ripening and cheese quality is ripening temperature. Ripening at elevated temperatures is normally
254
P. F. FOX et al.
considered with the objective of accelerating ripening but it also affects cheese quality. The literature on the accelerated ripening of cheese is discussed briefly in Section VII. VI. CHEESE TEXTURE
The rheological properties of cheese, commonly referred to as body and texture or consistency, are important aspects of cheese quality as perceived by the consumer. In many cases, texture is as important as a parameter of quality as taste and aroma. In addition to its significance to the eating quality of cheese, the texture of cheese is also very important with respect to the domestic and catering use of cheese as an ingredient, e.g., ease of cutting, grating, and melting and handling properties, e.g., whether the cheese holds its shape, whether grated cheese particles remain discrete or stick together (adhesiveness), and eye formation. The texture is characteristic for the variety and Ianges from crumbly, e.g., Parmesan and Cheshire, to elastic, e.g., Emmental or Gouda, from close, e.g., Cheddar, to open, e.g., Blue, from very firm, e.g., Gruyere, to fluid, e.g., Camembert. Texture is determined by composition, manufacturing conditions, the duration and extent of ripening, and assay temperature. For example, Young’s modulus, E, decreases with increasing temperature, increasing moisture content, increasing fat content (with a concomitant decrease in protein content), and increasing pH but increases with increasing salt content and age (partly due to loss of moisture and partly to proteolysis which reduces the amount of free moisture). The concentration of calcium in cheese (which is influenced mainly by the pH of the curd/whey at draining) and its association with casein (which is influenced by the final pH of the cheese) have major effects on the elasticity of cheese, e.g., cheeses with a high calcium/high drainage pH, e.g., Emmental, are elastic while those with a low calcium content (low draining pH), e.g., Cheshire, are crumbly. Proteolysis softens the texture of cheese, putatively owing to breakdown of the aSl-caseinmatrix which is commonly regarded as forming the continuous solid network in young cheese (although this is debatable). However, extensive proteolysis in low-moisture cheeses causes an increase in crumbliness due to increased water binding by the liberated carboxyl and amino groups. Cheese is an inhomogeneous system due to fissures/cracks of various types, eyes or mechanical openings, or differences in composition, e.g., between the rind and center of a cheese. Therefore, precise studies on cheese rheology are difficult. However, as the amount and depth of knowledge on the biochemistry, chemistry, and physical chemistry of cheese
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
255
has increased, researchers are showing increased interest in the study of cheese rheology. The subject is largely outside the scope of the present review and the interested reader is referred to a number of recent reviews/monographs on various aspects of cheese rheology and texture (Chen et al., 1979; Creamer and Olson, 1982; Lawrence et al., 1983; Masi and Addeo, 1986; Prentice, 1987; Walstra et al., 1987; Luyten, 1988; Tunick et al., 1990; IDF, 1991b; Pagliarini et al., 1991; Luyten et a)., 1991a,b; Jack and Peterson, 1992; Konstance and Holsinger, 1992; Prentice et al., 1993). Traditionally, cheese texture has been assessed subjectively by trained or untrained graders. Such procedures are expensive and not very reproducible and there is a major problem in describing and defining textural characteristics as determined by a grader. A useful attempt to standardize the evaluation and description of cheese texture is presented by Lavanchy etal. (1994). To overcome at least some of these limitations, objective instrumental methods for textural analysis of foods in general, and cheese in particular, have been developed. However, it is very difficult to evaluate and assess the sensory qualities of food by instrumental methods, largely because it is very difficult to replicate the process of food mastication. While various rheological properties of foods can be quantified readily by instrumental methods, the problem lies in relating these rheological parameters to sensory characteristics. These problems are discussed in some of the references cited above. VII. ACCELERATED CHEESE RIPENING
Cheese ripening is a slow, and hence an expensive, process, e.g., Parmesan and extramature Cheddar are ripened for at least 18 months. Ripening is still not controllable precisely, i.e., the quality and intensity of flavor cannot be predicted precisely. Therefore, there is an economic incentive for the development of methods for the acceleration of cheese ripening, provided that the flavor and texture can be maintained and characteristic of the variety. Of the three primary events in cheese ripening, i.e., glycolysis, lipolysis, and proteolysis, proteolysis is usually the rate-limiting one. Glycolysis is normally very rapid and is complete in most varieties within 24 hr; therefore, acceleration of glycolysis is not of interest. The modification and catabolism of lactate is either of little or no consequence (e.g., Cheddar or Dutch varieties) or is quite rapid-2-3 weeks (e.g., Swiss types, Camembert)-and consequently its acceleration is not important. Lipolysis is limited in most cheese varieties, exceptions being some Italian varieties, e.g., Romano and
256
P. F. FOX et al.
Provolone, and blue varieties. In the Italian and blue varieties, lipolysis does not appear to be rate-limiting and there appears to be little interest in accelerating it. As discussed in Section IVD, it is claimed that the intensity of the flavor of several varieties, including Cheddar, can be increased by adding exogenous Iipases, but as far as we are aware, such practices are not used commercially. Therefore, studies on the acceleration of cheese ripening have focussed on proteolysis, especially in hard, low-moisture varieties, in particular Cheddar. Low-fat cheeses have attracted much attention recently; such cheeses have poor texture and flavor and the techniques being considered to accelerate the ripening of normal cheeses are being applied to low-fat cheeses also. The third area of interest is the production of cheese-like products, e.g., enzyme modified cheeses, for use in the preparation of food products, e.g., processed cheeses, cheese sauces, cheese dips, etc. Literature on the acceleration of cheese ripening and related topics has been reviewed extensively (e.g., Law, 1984, 1987; Fox, 1988-1989; El-Soda and Pandian, 1991; El-Soda, 1993; Wilkinson, 1993). Therefore, it is not intended to exhaustively review the literature again but rather to provide a summary and suggest possible developments. Ripening can be accelerated by: 1. Increasing the ripening temperature. 2. Using exogenous enzymes. 3. Using modified starters. 4. Using cheese slurries.
Each of these methods has advantages and limitations (see Fox, 1988-1989). A. ELEVATED TEMPERATURES Traditionally, different cheese varieties are ripened at a characteristic temperature which is frequently chosen to suit the secondary microflora, e.g., Propionibacterium or moulds. Temperatures of 12-15°C are common, an exception being Swiss-type cheeses which are exposed at -20°C for a period of 3-4 weeks to induce the growth of Propionibacterium responsible for eye formation and typical flavor. Traditionally, Cheddar was ripened at -15°C but during the last 40 years or so it has become widespread commercial practice to ripen at 6-8°C. Ripening at lower temperatures reduces the risk of off-flavor development, but obviously retards ripening and probably the intensity of the final flavor attained. The use of low ripening temperatures is often augmented by rapidly cooling the curd, perhaps to as low as 10°C immediately after manufacture. The principal objective of rapid cooling is to retard the growth of NSLAB, which are
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
257
considered to have a negative effect on cheese quality. However, this conclusion depends on ones definition of Cheddar cheese flavor; almost certainly, high numbers of NSLAB lead to more strongly flavored cheese. The microbiological quality of cheesemilk today is vastly superior to that available even 20 years ago, total counts <20,000 CFU/ml are now common. Thermization of milk on receipt at the factory is common in some countries and it is normal commercial practice in most countries to pasteurize cheesemilk. When good hygienic practices are followed at farm and factory, milk is almost sterile as it enters the cheese vats; mechanized curd handling systems and enclosed cheddaring and salting systems (in the case of Cheddar) minimize contamination and consequently modern Cheddar, and probably Dutch varieties, are essentially free of NSLAB at the end of manufacture-in our experience, NSLAB counts of 4 0 0 CFU/g are normal for 1day-old Cheddar (although these grow to >lo7 during a 6-month ripening period). Starters have been greatly refined; it is now common practice to use a starter containing only one or two carefully selected strains of Lc. crernoris. While this practice leads to better control over cheesemaking, it further limits the diversity of ripening agents in cheese. The overall effect of these various practices, i.e., improved milk quality, pasteurization, defined starters, enclosed cheesemaking equipment, rapid cooling, and low ripening temperature, is the production of very mild cheese, free of off-flavors. While the latter is, obviously, a desirable development, not all consumers are happy with the very mild flavor of modern Cheddar. Several studies, especially in Australia, have shown that provided cheese of good composition and with a low count of NSLAB is used, the ripening of Cheddar cheese can be accelerated and its flavor intensified by using higher than normal ripening temperatures. Optimum results have been reported at 13-15°C at which the ripening time required for the production of mature cheese can be reduced by 50% (Folkertsma el al., 1996). Such practices are in fact reverting to traditional methods. B. EXOGENOUS ENZYMES On the assumption that proteolysis is the rate-limiting event in cheese ripening, there has been interest for several years in adding exogenous proteinases to cheese curd. The first problem encountered is the method of enzyme addition. Direct addition of the proteinase to the cheesemilk ensures its uniform distribution throughout the curd but since most proteinases are water-soluble, most of the added enzyme is lost in the whey, which is economically undesirable, and significant proteolysis may occur prior to coagulation with consequent loss of peptides in the whey and a reduction in cheese yield.
258
P. F. FOX et al.
The enzyme may be added to the curd but this method can be practiced only with Cheddar-type cheeses, the curd for which are milled prior to moulding; to facilitate uniform dispersion, the enzyme is usually diluted with salt but nonuniform distribution of enzyme may still occur, resulting in uneven ripening. Microencapsulation appears very attractive as a technique for preparing enzymes for addition to cheese and has attracted considerable attention. The microcapsules are added to the cheesemilk and are occluded in the curd, thus ensuring uniform distribution in the curd and minimizing losses of enzyme in the whey. The microcapsules disintegrate during cooking or ripening, releasing the entrapped enzyme into the cheese matrix. Liposomes are the preferred form of microcapsule and appear to give satisfactory results with respect to the efficiency of encapsulation and retention in the coagulum. However, microencapsulated enzymes have not been commercialized, probably owing to cost. The use of microencapsulation in cheese technology was reviewed by Skeie (1994). Early studies on the use of exogenous enzymes concentrated on individual proteinases, of which Neutrase (a neutral proteinase from B. subtilis) gave best results. However, gross proteolysis is not rate-limiting and current trends are to combine proteinases and peptidases. Such preparations appear to accelerate ripening (Wilkinson et al., 1992) but perhaps equally good results can be obtained using elevated ripening temperatures, with less inconvenience and risk and at no cost. Exogenous enzymes are not yet used commercially to accelerate the ripening of natural cheeses. C. MODIFIED STARTERS
Since current evidence indicates that the starter cells and their enzymes are responsible for the final stages of proteolysis, i.e., production of small peptides and free amino acids, probably the modification of amino acids and probably other important changes, it would appear that increasing starter cell numbers should accelerate ripening. Much of the recent work on acceleration of cheese ripening has been based on the above assumption and several approaches have been adopted. Increasing starter cell numbers is not satisfactory owing to the consequent increase in rate of acid production which has undesirable consequences (see Section IIIAS). High starter numbers may also cause bitterness. To overcome these problems, either lactase-negative (Lac-) or attenuated starters have been investigated and reported to accelerate ripening. The inclusion of a proportion of Lac- cells in cheese starters has also been recommended to reduce the incidence of bitterness and apparently is widely used for this reason in the Netherlands. Attenuated starters are prepared by heat-shocking or freeze-shocking Lactococcus or Lactobacilluscells such
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
259
that their acid-producing ability is destroyed but much of their proteolytic and especially their peptidolytic activity is retained. Attenuated starters can be regarded as microencapsulated enzymes. Several reports indicate that their use accelerates ripening but they are probably too expensive in most circumstances and to our knowledge they are not used commercially. The alternative approach is to develop starters that superproduce key enzymes. An engineered starter producing three times as much cell wallassociated proteinase as the parent did not accelerate ripening (Law et al., 1993), which supports the results of earlier studies using Prt- starters which indicated that the cell wall-associated proteinase is not rate-limiting in cheese ripening. McGarry et al. (1994) used an engineered strain that produced high levels of aminopeptidase (Pep N) in Cheddar cheese but found no beneficial effect. Thus, it is not clear at present which starter enzymes might be limiting. A further approach is to select or develop starter strains that lyse and release their intracellular enzymes quickly. This is based on the premise that if intracellular lactococcal enzymes are important, then the more rapidly they are released from the cells the better. Lactococcus strains die and lyse at considerably different rates and fast-lysingstrains have been reported to give faster rates of ripening in Cheddar (Wilkinson et al., 1994) and Saint Paulin (Chapot-Chartier et aL, 1994). Further work in this area appears warranted. D. CHEESE SLURRIES Flavor has been reported to develop very rapidly (1 week) in slurries containing -40% solids. Such systems have been used to screen exogenous enzymes. Fast-ripening slurries could be useful in the preparation of cheese sauces, cheese flavoring, processed cheeses, etc. Enzyme-modified cheeses, which can be regarded as being based on the slurry principle, are used commercially as ingredients in processed cheese and cheese products. Thus, there is undoubtedly a commercial economic incentive to develop techniques to accelerate the ripening of cheese. However, in spite of a considerable amount of published research, and presumably unpublished work, the number of viable options appears to be rather limited-at present the best method appears to be a higher ripening temperature. VIII.
PROCESSED CHEESE PRODUCTS
A. INTRODUCTION Pasteurized processed cheese products are produced by comminuting, melting, and emulsifying, into a smooth, homogeneous molten blend, one
260
P. F. FOX et al.
or more natural cheeses and optional ingredients using heat, mechanical shear, and (usually) emulsifying salts. Optional ingredients permitted are determined by product type, i.e., whether processed cheese, processed cheese food, or processed cheese spread and include dairy ingredients, vegetables, meats, stabilizers, emulsifying salts, flavors, colors, preservatives, and water (Tables IX and X). Although a product of recent origin compared to natural cheese, processed cheese products (PCPs) show a parallel increase in growth rate of -3% p.a. Documented world production amounts to 8-10% of total cheese manufactured (MMB, 1991). Factors contributing to the continued growth of these products are: (i) Their versatility as foods which offer almost unlimited variety in flavor, consistency, functionality, and consumer appeal as a result of differ-
TABLE IX PERMITTED INGREDIENTS IN PASTEURIZED PROCESS CHEESE PRODUCTS
Product Pasteurized blended cheese
Ingredients
Cheese; cream, anhydrous milk fat, dehydrated cream [in quantities such that the fat derived from them is less than 5% (wlw) in finished product]; water; salt; food-grade colors, spices, and flavors; mould inhibitors (sorbic acid, potassiudsodium sorbate, and/or sodiudcalcium propionates), at levels > 0.2% (wlw) of finished product. Pasteurized process cheese As for pasteurized blended cheese, but with the following extra optional ingredients: emulsifying salts [sodium phosphates, sodium citrates; 3% (wlw) of finished product], food-grade organic acids (e.g., lactic, acetic, or citric) at levels such that pH of finished product is 45.3. Pasteurized process cheese foods As for pasteurized process cheese, but with the following extra optional ingredients: dairy ingredients (milk, skim milk, buttermilk, cheese whey, whey proteins-in wet or dehydrated forms). Pasteurized process cheese spread As for pasteurized process cheese food but with the following extra optional ingredients: food-grade hydrocolloids (e.g., carob bean gum, guar gum, xanthan gums, gelatin, carboxymethylcellulose, and/ or carageenan) at levels < 0.8% (w/w) of finished products; food-grade sweetening agents (e.g., sugar, dextrose, corn syrup, glucose syrup, hydrolyzed lactose).
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
261
TABLE X COMPOSITIONAL SPECIFICATIONS FOR PASTEURIZED PROCESS CHEESE PRODUCTS"~
Product category Pasteurized blended cheese Pasteurized process cheese Pasteurized process cheese food Pasteurized process cheese spread
Moisture (%, w/w) 3 43 3 43 3 44 40-60
Fat (%,
WIW)
Fat in dry matter (96, WIW)
-
4 47
-
4 47
4 23 4 20
-
' Minimum temperatures and times specified for processing are 655°C for 30 sec. The compositional specifications for pasteurized process cheese may differ from those given, depending on the type of product; for more detail, see the Code of Federal Regulations (CFR) (1988).
(ii) (iii) (iv) (v)
ences in formulation, condiment addition, processing conditions, and packaging in various shapes and sizes. Lower cost relative to natural cheese due to incorporation of lowgrade natural cheese and cheaper noncheese dairy ingredients. Adaptability to the fast food trade, the most notable examples being the use of cheese slices in burgers and dried processed cheeses as snack and popcorn coatings. Relatively long shelf-life and no waste. Development of companies specializing in the manufacture of equipment, emulsifying salts, and other ingredients tailor-made to the industry's needs in fulfilling consumer needs.
B. CLASSIFICATION OF PROCESSED CHEESE PRODUCTS Four main categories of PCPs are identified [Code of Federal Regulations (CFR), 19881, namely, processed cheese, processed cheese foods, processed cheese spreads, and texturized blended cheeses (Table IX). The criteria for classification include permitted ingredients and compositional parameters. Processed cheese is usually sold in the form of sliceable blocks (e.g., processed Cheddar) or slices; cheese spreads and foods may be in the form of blocks, slices, spreads, dips, sauces, or pastes (e.g., in tubes). Texturized blended cheese, which is the least common category, is usually sold in forms giving a natural cheese image. Processed cheese analogueslsubstitutes, although similar to PCPs in make procedure and product characteristics, are not PCPs per se. Analogue production is based on vegetable fat-caseinate/ rennet casein blends and usually does not incorporate natural cheese, except where it is added in small quantities in the form of enzyme-modifiedcheeses,
262
P. F. FOX et al,
very mature cheese, or cheese powders to impart a more natural cheese flavor. Other processed cheese-derived products include cheese powders (obtained by drying pasteurized PCPs) and cheese sauce preparations (prepared by cooking blends of cheese powders, starches, skimmed milk solids, flavors, and water). C. MANUFACTURING PROTOCOL Manufacture involves the following steps: (i) formulation of blend, which involves selection of the correct type and quantity of natural cheeses, emulsifying salts, water, and optional ingredients, (ii) shreddinglcomminuting of cheese and blending with optional ingredients, (iii) processing of the blend, (iv) homogenization of the hot molten blend; this step is optional and implementation depends on the fat content of the blend, type of cooker used, smoothness, and body characteristics required in the end product, and (v) packaging and cooling. Processing refers to the heat treatment of the blend, with direct or indirect steam, with constant agitation; a partial vacuum may be used to regulate moisture, when using indirect steam injection, and to remove air from the product. In batch processing, the temperature-time combination varies (i.e., 70-95°C for 4-15 min), depending on the formulation, extent of agitation, the desired product texture, body, and shelf-life characteristics. At a given temperature, the processing time generally decreases with agitation rate which may vary, depending on the kettle (cooker) type, from 50 to 3000 rpm. In continuous cookers, mainly used for dips and sauces, the blend is mixed and heated to 80-90°C in a vacuum mixer from where it is pumped through a battery of tubular heat exchangers and heated to 130145°C for a few seconds and then flash-cooled to 90°C. Cooked product is then pumped to a surge tank which feeds the packaging machine. In the case of slice production, the hot molten cheese is pumped through a manifold with 8-12 nozzles which extrude ribbons of cheese onto the first two or three counter-rotating chill rolls over which the cheese ribbons pass and are thereby cooled from 70-80 to 30°C.The ribbons are automatically cut into slices, which are stacked and packed. D. PRINCIPLES OF MANUFACTURE OF PROCESSED CHEESE Natural cheese may be viewed as a three-dimensional particulate network, resembling a loose semirigid sponge, in the pores of which naturally emulsified fat globules and moisture are entrapped (Brooker, 1979; Kalab, 1979). The integrity of the network, which consists of overlapped and
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
263
crosslinked chains of partially fused aggregates (consisting of fused paracasein micelles) is maintained by various intra- and interaggregate bonds, including hydrophobic and electrostatic (e.g., calcium cross links via casein phosphoserine and ionized carboxyl residues) attractions (Knoop, 1977; Walstra and van Vliet, 1986). Application of heat (70 to 90°C) and mechanical shear to natural cheese, as in processing, in the absence of stabilizers, generally results in a heterogeneous, gummy pudding-like mass with extensive oiling-off and moisture exudation, particularly on cooling. These defects arise from shearing of fat globule membranes and aggregation and partial dehydration of the protein phase (especially in low-pH cheeses); consequently, free moisture and deemulsified liquified fat seep through the more porous, modified structure. The addition of emulsifying salts (1-3%) during processing promotes emulsification of free fat and rehydration of protein and thus contributes greatly to the formation of a smooth, homogeneous, stable product. However, it is also possible to manufacture stable pasteurized cheese products (i.e., pasteurized blended cheese) without the addition of emulsifying salts or hydrocolloids when certain changes in formulation and processing conditions are implemented (McAuliffe and O'Mullane, 1991; Guinee, 1991); such products are now commercially available in Ireland and elsewhere in the form of slices and spreads. In practice, the development of body, creaminess, and sheen during processing is commonly termed creaming of the product, especially when making spreadable, high-moisture PCPs. It is generally observed as an increase in viscosity on holding at elevated temperatures (75-90"C), especially when shearing continously. The emulsifying salts most commonly used for the manufacture of pasteurized processed cheese are sodium citrates, sodium orthophosphates, sodium pyrophosphates, sodium tripolyphosphates, sodium polyphosphates (e.g., Calgon), basic sodium aluminium phosphates (e.g., Kasal), and phosphate blends (e.g., Joha, Solva blends). These salts generally contain a monovalent cation (i.e., sodium) and a polyvalent anion (e.g., phosphate). While they are not emulsifiers per se, they cause, with the aid of heat and shear, a series of concerted physicochemical changes within the cheese blend which convert the aggregated, inactive calcium para-casein gel (network) into an active emulsifying and water-binding agent. These changes include calcium sequestration, pH displacement and stabilization, dispersal and hydration of para-casein, emulsification, and structure formation (Fig. 22) and are discussed briefly below. 1. Calcium Sequestration
This involves the exchange of the divalent Ca2+ (attached to casein via carboxyl and phosphoseryl residues) of the calcium para-caseinate net-
264 A.
P. F. FOX et at. Overall Reaction Raw cheese (calcium paracasein network with entrapped moisture and globular fat) water melting salts and energy
+
B.
+
Process ResDonsible for Transition PrOCeSS
Main Causative Agent
Ion exchange (Ca sequestration)
Emulsifying salt
pH buffering
Emulsifying salt
Protein dispersion
Emulsifying salt Thermal and mechanical energy
Protein hydration
Processing salt Thermal energy
Emulsification
Dispersed hydrated protein Mechanical energy
Structure formation
Protein-protein interactions Emulsified fat Globules (pseudo-protein particles)
FIG. 22. Summary of principles operative in process cheese manufacture (from Guinee, 1990).
work for the monovalent Na+ of the emulsifying salt (Nakajima et al., 1975; Lee and Alais, 1980). The removal (sequestration) of Ca2+results in: (i) disintegration of the inter- and intrastrand linkages keeping the paracaseinate network intact and hence a reduction in matrix continuity and gel-like structure, (ii) conversion of the para-caseinate gel network into sodium- and/or sodium phosphate para-caseinate dispersion (sol) (Tatsumi et al., 1975; Ito et al., 1976) to a greater or lesser degree, depending on the processing conditions and type of salt (calcium chelating strength, pH, and buffering capacity). The formation of sodium phosphate para-caseinate, as affected by the probable surface adsorption of the polyvalent emulsifying salt anions to the para-casein (van Wazer, 1971; Melnychyn and Wolcott, 1971; Shimp, 1985), is supported by the higher concentration of proteinbound phosphorus in processed cheese compared to natural cheese (Nakajima et al., 1975).
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
265
2. p H Displacement and Stabilization The use of the correct blend of emulsifying salts usually shifts the cheese pH from -5.0-5.5 in the natural cheese to 5.5-5.9 in PCPs and stabilizes it by virtue of their high-buffering capacity (Meyer, 1973; Tatsumi et al., 1975; Gupta et al., 1984; Caric and Kalab, 1987). This change contributes to successful processing by increasing the calcium-sequestering ability of the emulsifying salts per se (Irani and Callis, 1962; Lee et al., 1986) and the negative charge on the para-caseinate, which in turn promotes further disintegration of the calcium para-caseinate network and a more open, reactive caseinate conformation with superior water-binding and emulsification properties. Thus, the extent of pH displacement is a critical factor controlling the textural attributes of PCPs (Swiatek, 1964; Rayan et al., 1980; Gupta et al., 1984). 3. Para-Casein Dispersioflater Binding
Dispersion of para-casein, also called peptization, refers to the disintegration of the cheese network and conversion of the calcium paracaseinate into a charged, hydrated sodium (phosphate) para-caseinate as effected by the above-mentioned emulsifying salt-induced changes in combination with the mechanical and thermal energy inputs of processing (Lee et al., 1979, 1986; Kirchmeier et al., 1978; Buchheim and Thomasow, 1984). The conversion of calcium para-caseinate to sodium (phosphate) paru-caseinate during processing is the major factor affecting protein hydration. This is supported by the inverse relationship found between caseinbound calcium and casein solvation (Sood et al., 1979). Protein hydration is reflected in the large increases in nonsedimentable protein and bound water (Templeton and Sommer, 1936; Nakajima et al., 1975; Ito et al., 1976; M. A. Thomas et a!., 1980; Csok, 1982; Lamure et al., 1988); the level of hydration varies with formulation and processing conditions. However, prolonged holding of the molten processed cheese at a high temperature results in some reaggregation of the protein, a decrease in protein-bound water, and an increase in the level of sedimentable protein (Csok, 1982; Tatsumi et al., 1991). 4, Emulsification
Under the conditions of cheese processing, the dispersed, hydrated puracaseinate contributes to: (i) emulsification, via coating of dispersed free
266
P. F. FOX et al.
fat droplets, resulting in the formation of fat globule membranes, and (ii) emulsion stability by immobilization of large amounts of free water.
E. STRUCTURE FORMATION ON COOLING During the cooling of PCPs, the homogeneous, molten, viscous mass sets to form a characteristic body, which, depending on blend formulation, processing conditions, and cooling rate, may vary from a firm sliceable product to a semisoft spreadable consistency. While little or no information is available on the physicochemical mechanisms responsible for structure formation (setting) of PCPs on cooling, factors which probably contribute include: fat crystallization, protein-protein interactions, and incorporation of recombined fat globules, which may be considered as increasing the effective protein concentration, into the new structural matrix. Electron microscopical studies on processed cheese (Kimura et al., 1979; Taneya et al., 1979, 1980; Rayan et al., 1980; Heertje et al., 1981; Lee et al., 1981; Kalab et al., 1987; Savello et al., 1989; Tamime et al., 1990) indicate that the structure consists of emulsified fat globules dispersed in a protein network. The fat globules are evenly distributed (unlike natural cheese) and generally range from 0.3 to 5 wm in diameter. Fat globule size, as affected by the degree of emulsification, varies with formulation (i.e., type and quantity of emulsifying salt and other ingredients, age of cheese) and processing conditions (shear rate, temperature, and time). Rayan et al. (1980) found that the diameter of fat globules in processed Cheddar decreased progressively with processing time and at any given time the mean fat globule diameter was smallest on using tetrasodium pyrophosphate, largest with basic sodium aluminium phosphate (SALP), and intermediate with trisodium citrate or disodium phosphate. Indeed, after processing for 10 min with SALP, many of the fat globules had diameters greater than 10 pm (Rayan et al., 1980); hence, in practice, SALP is generally claimed to give processed cheeses with good melt and stretch properties. Lee et al. (1981) found that increasing the concentration of emulsifying salt (1-4%) and processing temperature (80-140°C) resulted in a progressive decrease in mean fat globule diameter and a parallel increase in firmness. The protein fraction of processed cheeses exists as relatively short strands which are connected to varying degrees, resulting in a matrix with different degrees of continuity, depending on product type. The matrix strands are much finer than those of natural cheese and appear to be composed of para-caseinate particles (20-30 mm diameter) which undergo limited aggregation or touching; it is suggested that these particles may correspond to casein submicelles (Kimura et al., 1979; Taneya et al., 1980; Heertje et al., 1981) released from the cheese para-caseinate network as a result of calcium
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS 267
chelation by the emulsifying salts. The para-caseinate membranes of emulsified fat globules, distributed uniformly throughout the protein phase, appear to attach to the matrix strands; the ensuing anchoring of the relatively short strands by the recombined fat globules probably contributes to the continuity and elasticity of the matrix in the cooled product. The positive correlation between the degree of emulsification (surface area of the fat phase) and firmness and the inverse relationship between the degree of emulsification and melt (Rayan el al., 1980; Savello ef al., 1989) lend support to this suggestion. The following differences exist between the protein matrices in hard and soft processed cheeses (Kimura et al., 1979; Taneya ef al., 1980): (i) The network in hard processed cheeses consists of interconnecting strands (up to 100 pm in length) made up of para-casein aggregates (20 pm diameter) strung together in a necklace-like structure. (ii) In soft products, the matrix-building para-casein aggregates are more dispersed and there are fewer interaggregate connections. The increase in the length of matrix strands with processing time and temperature (Heertje er al., 1981) probably reflects the decrease and increase in protein hydration and aggregation, respectively. The occurrence of more numerous and longer strands in hard processed cheeses ensure more interstrand connections and hence a more continuous and elastic protein matrix. Hence, Rayan et al. (1980) found that increasing the processing time, while scarcely changing the dimensions of the emulsified fat particles (i.e., when using sodium aluminum phosphate as emulsifying salt), resulted in processed Cheddar which was firmer, more elastic, and less meltable.
F. PROPERTIES OF EMULSIFYING SALTS The emulsifying salts most commonly used are citrates, phosphates, polyphosphates, and sodium aluminium phosphates (Caric and Kalab, 1987) (Table XI). Other potential emulsifying agents include gluconates, lactates, malates, ammonium salts, glucono lactones, and tartarates (Price and Bush, 1974). Today, salts are generally supplied as blends of phosphates (e.g., Joha C special) or phosphates and citrates (e.g., Solva NZ lo), tailor-made to the processor’s requirements. Citrates are usually used as sodium salts, although potassium salts have also been used (Gupta et al., 1984). The trisodium citrate is used most commonly; the mono- and disodium salts (NaH2C6H507and Na2HC6H507), when used alone, generally give overacid PCPs which are mealy, acid, and crumbly and show a tendency toward oiling-off due to poor emulsification
TABLE XI PROPERTIES OF EMULSIFYING SALTS FOR PROCESSED CHEESE PRODUCTSa
Group Citrates Orthophosphates Pyrophosphates
Polyphosphates
Aluminium Phosphates
Emulslfylng salt Trisodium citrate Monosodium phosphate Disodium phosphate Disodium pyrophosphate Trisodium pyrophosphate Tetrasodium pyrophosphate Pentasodium tripolyphosphate Sodium tetrapolyphosphate Sodium hexametaphosphate (Graham’s salt) Sodium aluminium phosphate
From van Wazer (1971) and Caric and Kalab (1987).
Formula 2Na3C&1507.lHzO NaH2P04.2H20 Na3HP04.12H20 Na~H2P207 NaD207.9H20 Na4P207.1OH20 NasP3010
Na6Pdk Na.+2Pn03n+l(n = 10-25) NaH14A13(P04)8.4H20
Solubility at 20°C (%)
pH value (1% solution)
High
6.23-6.26 4.0-4.2 8.9-9.1 4.0-4.5 6.7-7.5 10.2-10.4 9.3-9.5 9.0-9.5 6.0-7.5 8.0
40 18 10.7 32.0 10-12
14-15 14-15 Infinite
-
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
269
(Gupta et al., 1984). The dissociation constants (pKa’s) of citric acid at the ionic strength of milk are 3.0, 4.5, and 4.9 (Walstra and Jenness, 1984). Owing to their acidic properties, mono- and disodium citrates may be used to correct the pH of a cheese blend, e.g., when a high proportion of very mature cheese is used. The phosphates used in cheese processing include sodium monophosphates (orthophosphates; n = 1) and linear condensed phosphates such as pyrophosphates (n = 2) and polyphosphates (n = 3-25) (van Wazer, 1971; Wissmeier, 1972). Of the orthophosphates, disodium orthophosphate (Na2HP04)is the form normally used; the mono- and trisodium forms when used alone tend to give overacid and underacid products, respectively (Templeton and Sommer, 1936; Scharf, 1971; Gupta et al., 1984). Comparative studies (Gupta et al., 1984; Scharf, 1971) have shown that the potassium orthophosphates, pyrophosphates, and citrates give processed cheeses with textual properties similar to those made with the equivalent sodium salts at similar concentrations. Therefore, Gupta et af. (1984) suggested that potassium emulsifying salts may have potential in the preparation of reduced-sodium formulations. Sodium aluminium phosphate (NaHI4Al3 (PO4)** 4H20; van Wazer, 1971) is used only on a limited scale. The effectiveness of the different salts in promoting the various physicochemical changes during processing, summarized in Table XII, are discussed below. TABLE XI1 GENERAL PROPERTIES OF EMULSIFYING SALTS IN RELATION TO CHEESE PROCESSING~
Property ~
Citrates
Orthophosphates
Pyrophosphates
Polyphosphates
Low
Low
Moderate
High-very high
High
High
Moderate
Low-very low
Low
Low
High
Very high
Low
Low
Very high
Nil
Low
High
Very high (n = 3-10) -Low High-very high
Aluminium phosphate
~~
Ion exchange (calcium sequestration) Buffering action in the pH range 5.3-6.0 para-Caseinate dispersion (peptization) Emulsification
Bacteriostatic effects
Low
-
Very low
a Templeton and Sommer (1936), Glandorf (1964), Roesler (1966), Scharf (1971),van Wazer (1971), Tanaka et al. (1979, 1986), Rayan et al. (1980), Kosikowski (1982), Caric and Kalab (1987), and Marcy et nl. (1988).
270
P. F. FOX et al.
1. Calcium Sequestration Ion exchange is best accomplished by salts containing a monovalent cation and a polyvalent anion and effectiveness generally increases with the valency of the anion. The general ranking of the calcium sequestration ability of the common emulsifying salts in cheese is polyphosphates > pyrophosphates > orthophosphates > sodium aluminium phosphate = citrates (Wagner and Wagner-Hering, 1981; Nakajima et al., 1975; Lee et al., 1986). However, the sequestering ability, especially of the shorter chain phosphates, is strongly influenced by pH. This may be attributed to more complete dissociation, which gives a higher valency anion, at the higher pH values (van Wazer, 1971). Thus, for the shorter chain phosphates, calcium binding increases in the following order: NaH2P04, Na2HP04, Na2H2P207,Na3HPzO7,Na4P207,NasP3OIo(Caric and Kalab, 1987).
2. pH Changes and Buffering The buffering capacity of sodium phosphates, in the pH range normally encountered in PCPs (i.e., 5.3-6.0), decreases with increasing chain length and is effectively zero for the longer chain phosphates (n > 10). This reduction in buffering capacity with chain length is due to the corresponding reduction in the number of acid groups per molecule which occur singly at each end of the polyphosphate chain (van Wazer, 1971). The ortho- and pyrophosphates possess high buffering capacities in the pH ranges 2-3, 5.5-7.5, and 10-12 (van Wazer, 1971); thus, in cheese processing they are very suitable not only as buffering agents but also for pH correction. Within the citrate group, only the trisodium salt has buffering capacity in the pH region 5.3-6.0; the more acidic mono- and disodium citrates cause overacid, crumbly cheese with a propensity to oiling-off (Gupta et al., 1984). 3. Casein Hydration and Dispersion
The ability of the different groups of emulsifying salts to promote protein hydration and dispersion during cheese processing is in the following general order: polyphosphates > pyrophosphates = monophosphates = citrates (Lee et al., 1986; Tatsumi et al., 1975; Wagner and Wagner-Hering, 1981). The greater swelling effect of polyphosphates, which increases with chain length, over mono- and diphosphates can be explained in terms of the greater calcium sequestrating ability of the former. 4. Ability to Promote Emulsification
The effectiveness of different emulsifying salts to promote emulsification, as determined from electron microscopical and oiling-off studies, in pro-
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
271
cessed cheese is in the following general order: pyrophosphates > polyphosphates ( p = 3-15) > citrates orthophosphates (slightly) > basic sodium aluminium phosphates (Templeton and Sommer, 1936; Roesler, 1966; Scharf, 1971; Thomas et al., 1980; Shimp, 1985). This ability generally parallels their effectiveness in promoting hydration and dispersion.
-
5. Hydrolysis (Stability)
During processing and storage of PCPs, linear condensed phosphates undergo hydrolytic degradation which finally leads to complete conversion to orthophosphates (Glandorf, 1964;Roesler, 1966; Scharf, 1971;van Wazer, 1971; Meyer, 1973). The extent of degradation increases with increased duration of processing time and temperature, product storage time and temperature, and phosphate chain length (Glandorf, 1964). Other influencing factors include the type of cheese, quantity of emulsifying salt, and type of product being produced. In experiments with processed Emmentaler (Roesler, 1966), the level of polyphosphate (n > 4) breakdown during melting at 85°C varied from 7% for block cheese (processed for 4 min) to 45% for spreadable cheese (processed for 10 min). While the breakdown of condensed phosphates, through the tri- and di-forms, to monophosphates was complete in the spreadable cheese after 7 weeks, low levels were detectable in block processed cheese even after 12 weeks. The greater extent of polyphosphate degradation in the spreadable processed cheeses is also expected due to their higher pH and moisture content (Scharf, 1971; van Wazer, 1971). The practical consequences of hydrolysis include variations in the functionality of the emulsifying salt blend with processing conditions, an increased propensity to precipitation of dodecahydrate disodium orthophosphate (Na2HP04 12H20) on product storage (Scharf and Kichline, 1969) and labeling difficulties in relation to declaration of emulsifying salts used.
-
6. Bacteriocidal Effects
Cheese processing normally involves temperatures (70-95°C) which are lower than those used for sterilization. Thus, PCPs may contain viable spores, especially Clostridium spp., which originate in the raw materials (Briozzo et al., 1983; Sinha and Sinha, 1986, 1988; Caric and Kalab, 1987). Germination of spores during storage often leads to blowing of cans, putrefaction, and off-flavors (Meyer, 1973; Price and Bush, 1974). While bacterial spoilage is minimized through the addition of preservatives, some of the emulsifying salts also possess bacteriocidal properties. Polyphosphates inhibit many microorganisms, including Staphylococcus aureus, Bacillus subtilis, Clostridium sporogenes, and various Salmonella species (van Wazer,
272
P. F. FOX et al.
1971). Citrates, on the other hand, possess no bacteriostatic effects and may even themselves be subject to microbial degradation, thus reducing product keeping quality (Caric and Kalab, 1987). Orthophosphates have been found to inhibit the growth of CI. botulinum in processed cheese (van Wazer, 1971; Tanaka er al., 1979; Marcy et al., 1988). The inhibitory effect of sodium orthophosphates on Cl. botulinum, which has been found to be superior to that of sodium citrates in pasteurized process cheese spreads with moisture levels in the range of 52-58% (Tanaka et al., 1979), depends on the moisture and sodium chloride levels and pH of the processed cheese product (Tanaka et al., 1986). The general bacteriocidal effect of phosphates, which increases with chain length (Meyer, 1973;Kosikowski, 1982), may be attributed to their interactions with bacterial proteins (van Wazer, 1971) and chelation of calcium, which generally serves as an important cellular cation and cofactor for some microbial enzymes (Stanier et al., 1981). 7. Flavor Effects
While the effect of emulsifying salts per se on the flavor of processed cheese is difficult to quantify because of the influence of the many processing conditions thereon, it is generally recognized that sodium citrates impart a “clean” flavor while phosphates promote off flavors such as soapiness (in the case of orthophosphates) and bitterness (Templeton and Sommer, 1936; Scharf, 1971; Meyer, 1973; Price and Bush, 1974). Potassium citrates also tend to cause bitterness (Templeton and Sommer, 1936). G. INFLUENCE OF VARIOUS PARAMETERS ON THE TEXTURAL PROPERTIES OF PROCESSED CHEESE PRODUCTS
Numerous investigations have been carried out during the last decade on the development of cheese extenders and the functionality of ingredients for PCPs with the view to reducing formulation costs while retaining or improving textural attributes (Mann, 1986,1987,1990). Comparative studies on the effects on textural characteristics have frequently led to different conclusions; discrepancies may be attributed, at least partly, to differences in processing conditions and formulation and hence in the types of interactions between the component ingredients. The influence of various blend ingredients and processing conditions on product quality are discussed below. 1. Blend Ingredients
Cheese is the major blend constituent in PCPs, ranging from a minimum of 51%in spreads to 95% in processed cheeses (CFR, 1988). Hence, both
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
273
the type and maturity of the cheese used have a major influence on the rheological properties of the product (Meyer, 1973; Harvey et al., 1982; Caric and Kalab, 1987). This is reflected by formulation practices at the commercial level; block processed cheese with good sliceability and elasticity requires predominantly young cheese (7590% intact casein) whereas predominantly medium-ripe cheese (60-75% intact casein) is required for spreads. In model experiments with processed Gouda, Ito et al. (1976) found an inverse relationship between the age (and hence the degree of proteolysis) of the natural cheese and its emulsifying capacity. Hence, it is not surprising that the melt and firmness of PCPs increase and decrease, respectively, with maturity of the cheese blend (Thomas, 1970; Lazardis et al., 1981; Harvey et al., 1982; Mahoney et al., 1982; Tamime et al., 1990) since there is a positive relationship between the degree of emulsification and firmness and an inverse relationship between the degree of emulsification and firmness/hardness (Rayan et al., 1980). Working with chemically acidified curd systems, Lazaridis et al. (1981) treated curd with Aspergillus oryzae proteinase to induce variations in the level of proteolysis prior to processing. A strong positive relationship (r = 0.96) was found between meltability and the extent of proteolysis. Excessive proteolysis, however, was associated with textural defects, including excessive shortness. The same group (Mahoney et al., 1982) found that the optimum meltability of the processed acidified curd was obtained when the molecular weight of the proteolysis products was in the range 10-25 kDa; smaller peptides (<10,000) resulted in an excessively meltable soft cheese. Because of intercheese variation in structure and level of proteolysis, different types of cheese give processed products of different textural characteristics. Hence, it is generally recognized that mature semihard cheese varieties, such as hard Italian-types, Cheddar, and Emmentaler, give firmer, longer-bodied processed products that mould-ripened cheeses of the same age. “Rework” refers to creamed (having attained the desired increase in viscosity) processed cheese which is not packaged for sale; it is obtained from ‘‘leftovers’’ in the cookinglfilling machines, damaged packs, and batches which have overthickened (overcreamed) and are too viscous to pump. Added at a maximum level of -20% (w/w), rework promotes the rate of firming (i.e., enhances the creaming effect) during processing, especially in blends with a high moisture content (e.g., cheese spreads) and/or a high proportion of overripe cheese. Such cheese (overripe) tends to give poor emulsification due to very low levels of intact casein or to a high hydrophilicity; emulsification requires protein (emulsifier) with the correct hydrophile-lipophile balance. Addition of rework generally results in a
274
P. F. FOX et al.
higher apparent viscosity in the processed, hot, molten blend and in cooled products with higher resistance to deformation and reduced meltability, especially if the rework has overcreamed and is cooled slowly after processing (Kalab et al., 1987). The thickening effect of rework during processing becomes more pronounced when it is cooled slowly and stored for some time prior to reuse. In the above context, it is interesting to note that Tamime et af. (1990) found that the firmness of processed cheese increased during storage over 3 months with the effect becoming more pronounced as the storage temperature was increased from 10 to 30°C. Cheese base (CB) is being used increasingly as a cheese substitute in processed cheese manufacture, the main advantages being its lower cost and more consistent quality (i.e., intact casein content). Production generally involves ultrafiltration and diafiltration of milk, inoculation of the retentate with lactic culture, incubation to a set pH (5.2-5.8), pasteurization, and scraped-surface evaporation to 60% dry matter (Ernstrom, 1985; Mortensen, 1985); similar products are fermented, high dry matter, UF retentates. Increasing the level of substitution with CB normally results in “longer-bodied,” firmer, and less meltable products (Tamime et af., 1990; Younis et af., 1991; Collinge and Ernstrom, 1988). However, the effects vary depending on the method of CB preparation and the subsequent heat treatment during processing: (i) decreasing pH, in the range 6.6 to 5.2, of milk prior to UF resulted in CB with lower calcium levels and processed products with improved meltability (Anis and Ernstrom, 1984), (ii) rennet treatment of the UF retentate results in poorer meltability (Anis and Ernstrom, 1984), an effect which may be attributed to the higher degree of interaction between P-lactoglobulin and para-K-casein (than with native casein) during subsequent processing (Doi et al., 1983), (iii) treatment of retentate with other proteinases (i.e., Savorase-A, Aspergillus oryzae, and Candida cyfindracea), which increase the level of proteolysis in the CB, yields products which are softer and more meltable than those containing untreated CB (Sood and Kosikowski, 1979; Tamime et al., 1990), (iv) increasing the processing temperature in the range 66 to 82°C results in processed products with reduced meltability, an effect attributed to the gelation of whey proteins at the higher temperatures, especially when rennet-treated CB is used (Collinge and Ernstrom, 1988). In this context, it is noteworthy that processed cheese, resistant to melting on cooking, may be prepared by adding a heat-coaguable protein (3-7%, w/w, lactalbumin, egg albumen, etc.), at a temperature <70”C, to the cheese blend on completion of processing (Schulz, 1976). Noncheese dairy ingredients may account for a maximum of -15% of the blend in pasteurized processed cheese spreads. Addition of skim-milk powder at levels of 3-5% of the blend results in softer, more spreadable
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS 275
PCPs but increases the propensity to nonenzymatic browning on storage; higher levels (>7%) promote textural defects such as crumbliness (Thomas, 1969, 1970; Kairyukshtene and Zakharova, 1982). However, high levels (7-10%) may be added if the skim-milk powder is first reconstituted and then precipitated by proteinases or citric acid and the curd then added to the blend (Thomas, 1977). Added lactose, in the range of 0 to 5%, is reported to result in lower spreadability, lower water activity, and increased propensity to nonenzymatic browning during processing (especially at high temperatures) and storage (Kombila-Moundounga and Lacroix, 1991; Piergiovanni et aL, 1989). Excess lactose may also increase the propensity to form mixed crystals containing various species, e.g., Ca, P, Mg, Na, tyrosine, and/or citrate, in PCPs during storage. Owing to the relatively high level of bound water in processed cheese products (a maximum of 1.6 g/g solids nonfat; Csok, 1982), the effective lactose concentration, in the free moisture phase, may easily exceed its solubility limit (-15 g/lOO g H 2 0 at 21°C) (Uhlmann et al., 1983); the lactose crystals may then act as nuclei for crystallization of mineral species which are supersaturated. It is generally recognized that added whey proteins gel on heating and thereby result in firmer process cheese products with lower meltability (Schulz, 1976; Savello et al., 1989); the magnitude of the effect, however, appears to be related to the method of preparation and form in which the whey protein is added (Hill and Smith, 1992). The addition of milk protein coprecipitates (produced by high heat treatment of milk followed by acidification and calcium addition), at levels up to 5% of the blend, to processed Cheddar yielded products with increased firmness and sliceability and lower meltability and a reduced propensity to nonenzymatic browning (Thomas, 1970). However, the level at which meltability became noticeably impaired varied from 0.25 to 3.0% and varied with the source of the coprecipitate. Caseinates and caseins (acid and rennet) are used widely in processed cheese and cheese analogues, the main attractions being lower cost (relative to cheese protein), a consistent level of intact casein, good emulsifying capacity (of caseinates), and stretching properties of rennet casein, making it ideal for analogue pizza cheese. Caseinates (especially sodium) find most application in processed cheese spreads where their high water-binding capacity and good emulsification properties promote creaming. Gouda et al. (1985) found that partial replacement of cheese solids nonfat by calcium caseinate (5-7%) improved the meltability of cheese spreads. While caseinates may be used in spreadable cheese analogues (Marschall, 1990; Hokes et al., 1989), rennet casein is generally preferred in the manufacture of analogue pizza cheese, which is the major imitation cheese product (McCar-
276
P. F. FOX et al.
thy, 1990). With rennet casein, the colloidal calcium to casein ratio, and hence degree of intercasein aggregate crosslinking, may be controlled by the calcium-chelating strength of the emulsifying salts during processing, to give the desired degree of casein hydratiodaggregation and fat emulsification which in turn give the desired degree of meltability and stretchability on cooking the finished product. In this application, caseinates appear to overhydrate, resulting in a degree of casein aggregation which yields good meltability but which is too low to achieve satisfactory stretchability on cooking. Hydrocolloids, including carob bean gum, guar gum, carageenan, sodium alginate, gum karaya, pectins, and carboxymethylcellulose, are allowed in pasteurized processed cheese spreads at a maximum level of 0.8%.Owing to their water-binding andlor gelation properties (Phillips et aL, 1986), they impart viscosity, especially in instances of high water content or low creaming action (thin consistency) as affected by, for example, the use of overripe cheese. These materials, along with polysaccharides/polysaccharide derivatives (e.g., inulin) are finding increasing application, as fillers and texturizers, in the manufacture of lowfat products, including PCPs (Brummel and Lee, 1990; Smith et aL, 1992; Anonymous, 1993).
2. Composition Although the rheological attributes of PCPs with the same moisture level can differ significantlydue to variations in blend composition and processing conditions, increasing moisture content yields products which are softer, less elastic and viscous, sticky, and spreadable (Kairyukshtene and Zakharova, 1982). Product pH has a major effect on the texture (Scharf, 1971; Gupta et al., 1984; Shimp, 1985). Low pH (4.8-5.2), e.g., due to the use of monosodium citrate, monosodium phosphate, or sodium hexametaphosphate alone, give short, dry, crumbly cheeses which show a high propensity to oiling-off (Gupta et al., 1984). High pH values (>6.0) lead to very soft, overmeltable products (Gupta et al., 1984). 3. Processing Conditions
Increases in agitation speed, temperature in the range 7O-1OO0C, and duration of processing are generally accompanied by an increase in the creaming reaction with a concomitant increase in firmness and loss of spreadability and meltability (Ryan et al., 1980; Lee et at., 1981; Harvey et al,, 1982; Kalab et al., 1987; Tatsumi et al., 1991). The changes may be attributed to the increased degree of protein emulsification. Hence, high moisture formulations, as in processed cheese spreads, are generally sub-
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS 277
jected to conditions (higher temperature and more vigorous agitation) which promote a stronger creaming action (firming effect) than blockprocessed cheese. IX. NUTRITIONAL AND SAFETY ASPECTS OF CHEESE A. INTRODUCTION
Cheese is a nutritious, versatile food. A wide variety of cheese types are available to meet specific consumer requirements and allow convenience of use. The increasing popularity of cheese is apparent from consumption studies worldwide, Table I1 (IDF, 1984,1992;National Dairy Council, 1989). Cheese contains a high concentration of essential nutrients relative to its energy content. Its precise nutrient content is influenced by the type of milk used (species, stage of lactation, whole fat, lowfat, skim), method of manufacture, and to a lesser extent the degree of ripening. As outlined in detail elsewhere in this review, water-insoluble nutrients of milk (casein, colloidal minerals, fat, and fat-soluble vitamins) are retained in the cheese curd whereas the water-soluble constituents (whey proteins, lactose, watersoluble vitamins, and minerals) partition into the whey. However, loss of water-soluble B vitamins in whey may be compensated to a certain extent by microbial synthesis during ripening (Renner, 1987).
B. PROTEIN The protein content of cheese ranges from ca. 3 to 40%, Table XI11 (Holland et al., 1989). Protein content tends to vary inversely with fat content for any one type of cheese. During traditional cheese manufacture, most of the whey proteins are lost in the whey and represent only 2-3% of the total protein, the remainder being casein which is slightly deficient in sulfur-containing amino acids. Thus, the biological value of cheese protein is slightly less than that of total milk protein. If the essential amino acid index of total milk protein is given a value of 100, the corresponding value for the proteins in most cheese varieties ranges from 91 to 97 (Renner, 1987). When ultrafiltration is exploited in cheese manufacture (see Section IIIC), whey proteins are incorporated into the cheese, resulting in a biological value similar to milk protein. Cheese protein is almost 100%digestible as the ripening phase of cheese manufacture involves a progressive breakdown of casein to water-soluble peptides and free amino acids (see Section IVE). Hence, a significantdegree of digestion of cheese protein has occurred before it is consumed.
278
P. F. FOX et al.
TABLE XI11 COMPOSITION OF SELECTED CHEESES, PER I 0 0 G"
Energy Cheese type Brie Caerphilly Camembert Cheddar (normal) Cheddar (reduced fat) Cheshire Cottage cheese Cream cheese Danish blue Edam Emmental Feta Fromage frais Gouda Gruyere Mozzarella Parmesan Processed cheeseC Ricotta Roquefort Stilton
Water (g)
Protein (g)
Fat (g)
Carbohydrate (9)
Cholesterol (mg)
kcal
kJ
48.6 41.8 50.7 36.0
19.3 23.2 20.9 25.5
26.9 31.3 23.7 34.4
Trb 0.1 Trb 0.1
100 90 75 100
319 375 297 412
1323 1554 1232 1708
47.1
31.5
15.0
Trb
43
261
1091
40.6 79.1 45.5 45.3 43.8 35.7 56.5 77.9 40.1 35.0 49.8 18.4 45.7
24.0 13.8 3.1 20.1 26.0 28.7 15.6 6.8 24.0 27.2 25.1 39.4 20.8
31.4 3.9 47.4 29.6 25.4 29.7 20.2 7.1 31.0 33.3 21.0 32.7 27.0
0.1 2.1 Trb TP Trb Trb 1.5 5.7 Trb
Trb Trb Trb
90 13 95 75 80 90 70 25 100 100 65 100
0.9
85
379 98 439 347 333 382 250 113 375 409 289 452 330
1571 413 1807 1437 1382 1587 1037 469 1555 1695 1204 1880 1367
72.1 41.3 38.6
9.4 19.7 22.7
11.0 32.9 35.5
2.0 Trb 0.1
50 90 105
144 375 411
599 1552 1701
Adapted from Holland er al. (1989). Tr, trace. Variety not specified.
C. CARBOHYDRATE Most of the lactose in milk is lost in the whey during cheese manufacture and hence most cheese contain only trace amounts of carbohydrate (Table XIII).Furthermore, the residual lactose in cheese curd is usually fermented to lactic acid by starter bacteria. Thus, cheeses are suitable dairy foods for lactose-malabsorbing individuals who are deficient in the intestinal enzyme, lactase. D. FAT AND CHOLESTEROL The fat content of cheese varies considerably, depending on the milk used and the method of manufacture (Table XIII).Fat plays several roles
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
279
in cheese. It affects its firmness, adhesiveness, mouthfeel, and flavor. Its influence on cheese flavor depends on the variety; in some varieties, free fatty acids and their metabolites are important flavor constituents (see Section IVE). From a nutritional point of view, the digestibility of the fat of different varieties of cheese is in the range 88-94% (Renner, 1987). Most cheeses are potentially significant sources of dietary fat. For example, a 50-g serving of Cheddar provides 17 g fat (Table XIII), which is a significant amount when compared with typical intake of fat in affluent Western societies, e.g., a typical Western diet providing 2000 kcal (8400 J) with 40% energy from fat contains approximately 88 g fat. Cheese fat generally contains -66% saturated fatty acids, 30% monosaturated and 4% polyunsaturated. Thus, cheese represents a significant dietary source of both total fat and saturated fatty acids. Recent work on primates (Hayes et al., 1991) shows that of the many saturated fatty acids in milk, only C12:0, CI4:",and C I ~have : ~ the property of raising blood cholesterol and palmitic acid (CI6:") is relatively ineffective. Many sets of dietary guidelines issued by expert panels worldwide have recommended reductions in both total and saturated fat intakes in Western societies. Recently, the National Research Council (1989) recommended that total fat intake be restricted to not more than 30% and saturated fatty acids to not more than 10%of energy, respectively. In large measure, these recommendations are based on evidence that increased intakes of some saturated fatty acids elevate both total and low density lipoprotein cholesterol in blood which is associated with an increased risk of coronary heart disease. While some nutritionists dispute this hypothesis, the overwhelming body of medical opinion worldwide supports the concept of dietary guidelines. Market forces and consumers have responded to these guidelines and the market for food products low in fat, cholesterol, and sodium has expanded dramatically during the past decade. The cheese industry has responded by developing "light" cheese products with a reduced fat content (see Olson and Johnson, 1990). The cholesterol content of cheese is a function of its fat content (Table XIII) and ranges from ca. 10 to 100 mg/100 g, depending on the variety. Despite considerable consumer confusion and the widespread prevalence of misinformation, dietary cholesterol has much less influence on blood cholesterol levels than dietary saturated fat (Keys, 1984). Thus, the cholesterol content of cheese is of lesser importance than its saturated fat content. The majority of individuals show little or no response in blood cholesterol levels to increased dietary cholesterol intake in the range 250-800 mg/day. However, a minority (ca. 20%)of adults do exhibit increased levels of blood cholesterol in response to increased dietary intake (McNamara, 1987). Some dietary guidelines recommend restricting dietary cholesterol intake to not
280
P. F. FOX et al.
greater than 300 mg/day (National Research Council, 1989) while others make no specific recommendation regarding cholesterol [Committee on Medical Aspects of Food Policy (COMA), 19841. In recent years, a considerable body of research has been carried out on the effects of oxidation products of cholesterol on the etiology of atherosclerosis (Hubbard et af., 1989). However, cholesterol oxides are formed to a negligible extent in cheese under normal conditions of manufacture, ripening, and storage (Sunder et af., 1988).
E. VITAMINS The concentration of fat-soluble vitamins in cheese is influenced by the same factors that affect its fat content. Most of the vitamin A (80-85%) in milk fat is retained in the cheese fat. The concentration of water-soluble vitamins in cheese is generally lower than that in milk due to losses in the whey. The loss of some of the B vitamins is offset, to a certain extent, by microbial synthesis during cheese ripening. In particular, propionic acid bacteria synthesise significant levels of vitamin BI2in Swiss-type cheeses (Renner, 1987). In general, most cheeses are good sources of vitamin A, riboflavin, vitamin BI2,and to a lesser extent, folate. The vitamin content of a range of cheeses is shown in Table XIV (Holland et af., 1989). Cheese contains negligible amounts of vitamin C. F. MINERALS Cheese is an important dietary source of several minerals, in particular calcium, phosphorus, and magnesium (Table XV). A 100-g serving of hard cheese provides ca. 800 mg Ca, which represents the Recommended Daily Allowance for most adults (Food and Nutrition Board, 1980). However, acid-coagulated cheeses, e.g., Cottage, contain considerably lower levels of calcium than rennet-coagulated varieties (Renner, 1987). Calcium bioavailability from cheese is equivalent to that from milk. Recker et al. (1988) reported that 22.9, 26.7, and 25.4% of total calcium was absorbed from cream cheese, whole milk, and yogurt, respectively. While the etiology of osteoporosis is very complex, there is widespread consensus that adequate calcium intake during childhood and in the teenage years is important in assuring the development of optimum peak bone mass. Maximizing bone mass early in life is considered to be a major preventative factor against development of osteoporosis in later years (Heaney, 1991). The National Research Council (1989) recommended an increase in calcium intake by teenage and adult females. Cheese has a potential role in supplying extra, highly bioavailable, calcium.
TABLE XIV VITAMIN CONTENT OF SELECTED CHEESES, PER IOO G" (HOLLAND ET AL., Retinol
Vitamin D
Vilamm E
Thiamlne
Riboflavin
Niacin
Vitamin Bh
Vitamin BI2
Folate
Pantothenate
Biotin
Cheese type
(rg)
(a)
(tg)
(mg)
(mg)
(mg)
(mg)
(mg)
(PP)
(PP)
(mt9
(PP)
Brie Caerphilly Camemben Cheddar (normal) Cheddar (reduced fat) Cheshire Cottage cheese Cream cheese Danish blue Edam Emmental
285 315 230
210 210 315
1.2 1.1 1.1 1.1
58 50 102 33
5.6 3.5
0.07
0.15 0.11 0.22 0.10
0.35
0.05 0.03
0.43 0.47 0.52 0.40
0.43 0.11 O.%
225
0.84 0.78 0.65 0.53
0.04 0.03
325
0.20 0.24 0.18 0.26
0.36
3.0
165
100
0.11
0.39
0 03
0.53
0.09
0.13
1.3
56
0.51
3.8
350 44
220
0.24 0.03
0.70 0.08
0.03 0.03
0.48 0.26
0.11
0.09
0.13
0.08
0.9 0.7
40
10
27
0.31 0.40
3.0
385
220
0.27
1.0
0.03
0.13
0.06
0.04
0.3
11
0.27
I .6
280 175 320
250 150
0.23
50 40 20
0.53 0.38
220
33
2.7 18 30 2.4
100
Tf
245 325 240 345 270
145
225
0.25
0.48 0.07 0.10 0.19 0.13 0.05 0.04 0.08
1.0 2.1
Nb 0.50 0.05 0.24
0.41 0.35 0.35
Feta
0.03 0.03 0.05 0.04 0.04 0.03 0.03 0.03 0.03 0.03
0.12
140
0.76 0.48 0.44
0.19 0.65 0.43
Fromage frais Gouda Gruyere Mozzarella Parmesan Processed cheesed Ricotta Roquefort Stilton
185 295 355
Carotene
1989)
0.19
0.37 0.02 0.53 0.58 0.33 0.70 0.55
170
0.16
210 95
0.25
92
Nb Nb
0.03 0.55
0.02
10
185
0.27
0.61
0.03
0.21
0.04
0.21 0.40 0.30
0.39 0.31 0.44 0.28
0.09 0.09
0.07
0.12 0.10
0.10 0.08 0.11 0.09 0.13 0.08
0.09 0.57 0.49
0.03 0.09 0.16
'Adapted from Holland et al. (1989). The nutrient is present in significant quantities but there is not reliable information on the amount. Tr, trace. Variety not specified.
2.0 1.1 1.4
23
2.1 1.9 0.9
15 43 12 19 12 18
0.3 0.4 1.0
Nb 45 77
1.7 1.6
0.29 0.36
0.40 0.36
7.6
4.0
Nb 0.32 0.35
Nb
0.25
2.2
0.43 0.31
3.3 2.3
Nb 0.50 0.71
Nb 2.3 3.6
1.4 1.5
282
P. F. FOX el al.
TABLE XV MINERAL CONTENT OF SELECTED CHEESES, MG PER I 0 0 G"
Cheese type
Na
K
Ca
Mg
P
Fe
Zn
Brie Caerphilly Camembert Cheddar (normal) Cheddar (reduced fat) Cheshire Cottage cheese Cream cheese Danish blue Edam Emmental Feta Fromage frais Gouda Gruyere Mozzarella Parmesan Processed cheeseb Ricotta Roquefort Stilton
700 480 650 670 670 550 380 300 1260 1020 450 1440 31 910 670 610 1090 1320 100 1670 930
100 91 100 77 110 87 89 160 89 97 89 95 110 91 99 75 110 130 110 91 130
540 550 350 720 840 560 73 98 500 770 970 360 89 740 950 590 1200 600 240 530 320
27 20 21 25 39 19 9 10 27 39 35 20 8 38 37 27 45 22 13 33 20
390 400 310 490 620 400 160 100 370 530 590 280 110 490 610 420 810 800 170 400 310
0.8 0.7 0.2 0.3 0.2 0.3 0.1 0.1 0.2 0.4 0.3 0.2 0.1 0.1 0.3 0.3 1.1 0.5 0.4 0.4 0.3
2.2 3.3 2.7 2.3 2.8 3.3 0.6 0.5 2.0 2.2 4.4 0.9 0.3 1.8 2.3 1.4 5.3 3.2 1.3 1.6 2.5
Adapted from Holland ef al. (1989). Variety not specified.
Dairy products, including cheese, contribute little dietary iron (Table XV). Iron deficiency is commonly observed in both developing and developed countries, e.g., iron deficiency anemia in the United States has been estimated at 5.7% for infants, 5.9% for teenage girls, and 4.5% for young women (Dallman et al., 1984). Hence, there has been interest in fortifying dairy products with iron to enhance their nutritional value. Cheddar and processed cheese have been successfully fortified with iron (Zhang and Mahoney, 1989a,b, 1990, 1991). As discussed in Section IIIA6, NaCl serves several important functions in cheesemaking. A wide range of sodium levels are found in cheese due to different amounts of salt added during cheesemaking (Table XV). In general, the salt content of natural cheeses tends to be less than that of many processed cheeses. Adults in the United States consume, on average, 4-6 g sodium/day (Surgeon General, 1988). Similar intakes are reported for other affluent Western Societies and are substantially above recommended safe and ade-
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
283
quate intakes, i.e., 1.1-1.3 g/day (Food and Nutrition Board, 1980). Considerable evidence exists that high sodium intakes can contribute to hypertension in a susceptible minority (20%) of individuals who are genetically saltsensitive (National Research Council, 1989). Unfortunately, there is no simple diagnostic test to identify salt-sensitive individuals. Hence, dietary guidelines for the general public usually recommend that salt intake be restricted (National Research Council, 1989). However, even in countries with high consumption, cheese contributes only about 5-8% of total sodium intake (Renner, 1987).
G. NISIN AND OTHER ADDITIVES IN CHEESE Sorbic acid and sorbates are used in several foods, including hard and semihard cheeses, to inhibit yeast and mould growth. It is a harmless and effective additive (Renner, 1987). Nitrate is usually added to the milk for some varieties of cheese. It is reduced to nitrite which inhibits the growth of Clostridium spp. responsible for late gas blowing and flavor defects. Nitrite is rapidly destroyed in cheese so that the finished product contains only trace levels which pose no hazard to consumers. The contribution by cheese to the total intake of nitrite is negligible (Renner, 1987). Consumer resistance to the use of synthetic additives in foods has stimulated interest in natural additives and preservatives. The principal natural additive used in cheese is the bacteriocin, nisin. Bacteriocins are peptides which inhibit a limited range of bacteria, usually closely related to the producer organism. The potential of nisin, produced by Luctococcus luctis, as a food preservative was first demonstrated using nisin-producing cultures in the manufacturer of Swiss-type cheese to prevent spoilage by clostridia (Hirsch et al., 1951). To date, nisin is the only purified bacteriocin commercially exploited as a food preservative. It can be added to processed cheese products to prevent late blowing by clostridia, the spores of which, if present in the natural cheese, survive pasteurization (Barnby-Smith, 1992).
H. CHEESE AND DENTAL CARIES A simplified description of the etiology of dental caries involves metabolism of sugars by oral microorganisms to acids which gradually dissolve tooth enamel. However, it is now recognized that a number of dietary factors and nutrient interactions can modify the expression of dental caries (Herod, 1991).The cariogenic potential of food is influenced by its composition, texture, solubility, retentiveness, and ability to stimulate saliva flow (Morrissey et al., 1984).
284
P. F. FOX et al.
In recent years, a considerable body of research has been conducted on the cariostatic effects of cheese. Early work (Shaw et al., 1959; Dreizen et al., 1961) demonstrated that the incorporation of dairy products into the diet greatly reduced the development of dental caries in rats. Reynolds and Johnson (1981) confirmed these findings. Later work (Weiss and Bibby, 1966; Jenkins and Ferguson, 1966) indicated that if enamel is treated with milk in vitro and subsequently washed, the solubility of the enamel is greatly reduced. This effect was attributed to the high levels of calcium and phosphate in milk (Jenkins and Ferguson, 1966) or to the casein (Weiss and Bibby, 1966). Later work supports both viewpoints. Reynolds and del Rio (1984) found that both casein and whey proteins significantly reduced the extent of caries, with the former exerting the greater effect. Further evidence for the protective effect of casein was provided in a study on rats fed casein-enriched chocolate (Reynolds and Black, 1987). However, the palatibility of this innovative product was considered unacceptable for humans! Concentrates containing various levels of whey protein, calcium, and phosphorus, but negligible amounts of casein, significantly reduced the incidence of dental caries in rats (Harper et al., 1987). Thus, there is evidence that milk proteins, calcium, and phosphate all exert an anticariogenic effect. Rugg-Gunn et al. (1975) provided the first evidence that cheese consumption had an anticariogenic effect in humans. They showed that the consumption of Cheddar cheese after sweetened coffee or a sausage roll increased plaque pH, possibly due to increased salivary output. Similar effects were reported by Imfeld et al. (1978) who used a more sophisticated continuous wire telemetry procedure to monitor variations in plaque pH. The effect of eating patterns on dental caries in rats was investigated by Edgar et al. (1982). Rats fed additional meals of cheese while on a highsucrose diet developed fewer smooth surface carious lesions and exhibited increased salivary output (which buffers acid formed in plaque) and a reduction in numbers of Streptococcus mutans. Harper et al. (1983) suggested that the cariostatic effect of cheese in rats is due to its calcium phosphate and/or casein; the fat or lactose appeared to exert little influence. Further work by Rosen et al. (1984) on the effect of cheese, with or without sucrose, on dental caries and recovery of inoculated S. mutans in rats indicated beneficial cariostatic effects of cheese consumption but little effect on S. mutuns numbers. These data suggest that the cariostatic effects of cheese may not be related directly to effects on S. mutans. Work on the protective effects of four different cheese varieties in an in vitro demineralization system suggested that most, but not all, of the protective effects of cheese could be explained by mass action effects of soluble ions, particularly calcium and phosphate (Jenkins and Harper, 1983).
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
285
An elegant study on the effects of Cheddar cheese on experimental caries in humans was reported by Silva el al. (1986) using an “intraoral cariogenicity test” (ICT). Demineralization and consequent reduction in the hardness of enamel monitored in this test is assumed to be typical of the early stage of caries development. Enamel slabs were polished and their initial microhardness determined using a Knoop Indenter. The slabs were then wrapped in Dacron and fastened on a prosthetic appliance made specificallyfor each subject to replace a missing lower first permanent molar. The subjects chewed 5 g of cheese immediately after rinsing their mouths with 10%(w/v) sucrose. Chewing cheese immediately after sucrose rinses resulted in a 71% reduction in demineralization of the enamel slabs, raised plaque pH, but caused no significant changes in the microflora of plaque compared with controls. Silva et al. (1987) investigated the effects of the water-soluble components of cheese on human caries using the ICT procedure and an experimental protocol which avoided salivary stimulus caused by chewing cheese. An average reduction of 55.7% in the cariogenicity of sucrose was reported, indicating the presence of one or more watersoluble anticariogenic components in cheese. Further evidence that cheese may inhibit dental caries in the absence of saliva was provided by Krobicka et al. (1987); rats that had their saliva-secreting glands surgically removed developed fewer and less severe carious lesions when fed cheese in addition to a cariogenic diet when compared to appropriate controls. Trials on human subjects have confirmed that consumption of hard cheese leads to significant rehardening of softened enamel surfaces (Jenkins and Hargreaves, 1989; Gedalia et a/., 1991). While more research is necessary to define the precise mechanism(s) involved in the cariostatic effects of cheese, there is ample evidence to support the consumption of cheese at the end of a meal as an anticaries measure (Herod, 1991). Gedalia et al. (1992) consider the most plausible mechanism for the protective effect of cheese to be related to the mineralization potential of casein-calcium phosphate of cheese and to the stimulation of saliva flow induced by its texture and/or flavor. I. MYCOTOXINS
Mycotoxins are secondary metabolites of fungi which can cause acute toxic, mutagenic, teratogenic, and carcinogenic effects in animals. The fact that mycotoxins, such as aflatoxin, are among the most potent animal toxins and carcinogens known, warrants concern about the contamination of the human food supply, including dairy products, with mycotoxins. The presence of mycotoxins in milk and dairy products may result from contamination of the cows’ feedstuffs (indirect contamination) or contami-
286
P. F. FOX et al.
nation of dairy products by mycotoxin-producing fungi (direct contamination) (van Egmond, 1989).
1. Indirect Contamination Allcroft and Carnaghan (1962, 1963) first reported that intake of aflatoxin-contaminated feedstuff by dairy cows resulted in excretion of a toxic factor in their milk within a few hours. Subsequently, Allcroft et al. (1966) and Holzapfel et al. (1966) identified aflatoxin M1 (the 4-hydroxy derivative of aflatoxin B1) as the principal toxic metabolite in milk. Studies (Shreeve et al., 1979; Robinson et al., 1979; Prelusky et al., 1984) on the indirect contamination of milk with other mycotoxins such as ochratoxin A, zearalenone, T-2 toxin, sterigmatocystin, and deoxynivalenone have indicated that contamination of milk with these mycotoxins does not represent a significant public health issue. An average of 1-2% of ingested aflatoxin B1 is excreted in milk as aflatoxin M1 (Rodricks and Stoloff, 1977; Patterson et al., 1980; Lafont et al., 1980). The carryover level varies between animals, from day to day and between milkings for the same animal. The results of surveillance programs for aflatoxin M1 in milk and milk products in many countries have been summarized by Brown (1982) and van Egmond (1989); the incidence and levels of aflatoxin M1 in dairy products have decreased significantly in recent years, which has been attributed to the effect of legislation implemented in many countries on aflatoxin contamination of feedstuffs. However, a significant increase in aflatoxin Ml levels in dairy products was noted in the United States in 1988-1989 as a result of feeding maize products contaminated with aflatoxin B1 due to the severe drought in the U.S. Midwest in 1988 which created ideal conditions for aflatoxin production by Aspergillus jlavus. The surveillance programs discussed above contain few data on aflatoxin MI in cheese. In general, aflatoxin MI was not detectable or occurred at concentrations lower than the current legal limits (0.2-0.25 pglkg) in a few countries for aflatoxin MI in cheese (van Egmond, 1989). 2. Fate of Aflatoxin MI during Cheese Manufacture and Ripening
The fate of aflatoxin M1 in cheese milk is affected by the principal manufacturing steps. Early studies on the distribution of aflatoxin MI between the curd and whey gave contradictory results. Allcroft and Carnaghan (1962,1963) reported that toxic activity was associated only with the rennetcoagulated curd. In contrast, Purchase et al. (1972) reported that Cottage cheese made by acid coagulation of naturally contaminated milk contained
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
287
no aflatoxin MI, which was present in the whey. However, several other studies (below) have indicated aflatoxin MI present in milk partitions between the curd and whey in both acid-coagulated and rennet-coagulated cheeses. Several investigators (Stubblefield and Shannon, 1974; Kiermeier and Buchner, 1977a; van Egmond et al., 1977) have investigated the stability of aflatoxin MI during the conversion of milk to curd; recoveries of 88-111% of the aflatoxin M1 in the milk were found for combined curd and whey fractions, indicating that aflatoxin MI is stable during cheesemaking. About half (40, 57, and 47%, respectively) of total aflatoxin M1 was found in the curd fraction in these studies. The partition coefficient of aflatoxin MI in water would suggest that most of the toxin should partition into the whey. The anomaly can be explained by data which suggest that aflatoxin M1 tends to associate, possibly via hydrophobic interactions (Yousef and Marth, 1989), with casein micelles (Brackett and Marth, 1982c; Blanc et al., 1983), causing a greater than expected concentration of the toxin in cheese curd. The concentration of alfatoxin MI in Camembert and Tilsit (Kiermeier and Buchner, 1977b), Cheddar (Brackett and Marth, 1982a), and Brick cheese (Brackett et al., 1982) increased during the early stage of ripening, followed by a gradual decrease to the concentration observed in the initial stages of ripening. In contrast, the concentration of aflatoxin MI in Gouda (van Egmond et al., 1977) and Mozzarella did not vary significantly during ripening (Brackett and Marth, 1982b),while the concentration in Parmesan cheese decreased initially and then increased slowly (Brackett and Marth, 1982b). Despite these somewhat contradictory trends, it appears that aflatoxin M1 is stable in cheese during ripening.
3. Production of Toxic Metabolites in Mould-Ripened Cheese Cultures of Penicillium roqueforti and P. camemberti have been used for a long time in the manufacture of various types of blue-veined and white surface-mould cheeses. A report by Gibe1 et al. (1971), suggesting that these moulds could form toxic metabolites of potential public health significance, prompted a significant research effort on this topic. It was subsequently shown (Schoch, 1981) that strains of P. roqueforti and P. camemberti can produce a range of toxic metabolites. Some P. roqueforti strains can produce PR toxin, patulin, mycophenolic acid, penicillic acid, roquefortine, cyclopiazonic acid, isofumigaclavine A and B, and festuclavine. P. camemberti strains have been shown to produce only cyclopiazonic acid. Neither P. roqueforti nor P. cumemberti have been reported to produce aflatoxins, either in vitro or in cheese (Engel and von Milczewski, 1977).
288
P. F. FOX et al.
Cyclopiazonicacid, which is formed in vitro by all strains of P. camemberti examined to date, has been reported in samples of commercial Camembert and Brie (Le Bars, 1979; Schoch et al., 1983). It occurs primarily in the rind and values of <0.5 mg/kg whole cheese are normally found in cheese stored in the cold but up to 5 mg/kg may be encountered if the storage temperature is too high. Evaluation of toxicological data currently available, together with data on the consumption of Camembert and Brie, indicates that these levels cause no appreciable risk to human health (Engel and Teuber, 1989). Of the toxic metabolites produced by P. roqueforti, patulin, penicillic acid, and PR toxin have not been detected in commercial cheeses as the strains used for blue cheeses do not appear to produce these toxins in detectable amounts (Engel and Prokopek, 1979; Polonelli et af., 1978). Mycophenolicacid has been reported in commercial cheese samples (Lafont et al., 1979a; Engel et af., 1982) but at levels well below those which pose a potential human risk. Roquefortine and isofumigaclavine A and B have been reported in commercial blue cheeses (Scott and Kennedy, 1976;Ware et al., 1980) ripened at normal temperatures but the toxicity of these compounds is low and they are very unlikely to be hazards to human health (Engel and Teuber, 1989).The most compelling evidence that consumption of mould-ripened cheeses is not hazardous to human health is provided by studies in which P. roqueforti and P. carnemberti cultures and mould-ripened cheeses were fed to rats and rainbow trout (Frank et al., 1975,1977;Schoch et af., 1984); although the daily dose of mould was equivalent to human consumption of 100 kg cheese per day, no signs of toxicity were observed. 4. Direct Contamination of Cheese with Mycotoxins.
Unintentional mould growth on cheese during ripening and storage is a potential problem for manufacturers, retailers, and consumers; it results in financial loss, diminishes consumer appeal, and often necessitates trimming. However, mycotoxin production is a potentially more serious problem involving possible health risk. Such cheese has been reported to contain mycotoxins that are nephrotoxic (ochratoxin A, citrinin), teratogenic (ochratoxin A, aflatoxin B1),neurotoxic (penitrem A, cyclopiazonic acid), carcinogenic (aflatoxins B1and G1,ochratoxin A, patulin, penicillic acid, sterigmatocystin), or toxic antibiotics (patulin, penicillic acid, mycophenolic acid, citrinin) (Ueno, 1985). Of the mycotoxigenic fungi isolated from cheese. Penicilfiurn spp, are by far the most frequently reported, with Aspergillus spp. and others occasionally encountered. Moulds that develop on cheese during refrigerated storage are almost always Penicilfium spp., which, in contrast to Aspergillus spp., can grow at low temperatures (Bullerman, 1981). However, not all Penicil-
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
289
fium spp. isolated from cheese are toxigenic. Toxicity in chick embryoes was induced by 30% of Peniciflium isolates from Cheddar cheese (Bullerman and Olivigni, 1974) and 35% of Peniciflium isolates from Swiss cheese (Bullerman, 1976). Toxicity in ducklings was induced by 47% of Penicilfium isolates from South African Gouda and Cheddar (Luck et af., 1976). Cheese is generally a good substrate for fungal growth, given suitable conditions of temperature and humidity. Mycotoxin-producing moulds require oxygen and hence appropriate packaging of cheese is important; moulds are very unlikely to grow on vacuum-packed or wax-coated cheese. Good plant sanitation during manufacture and handling is also important in minimizing or preventing mould growth on cheese (Bullerman, 1977). Mycotoxins are unlikely to be produced during low temperature storage (Bullerman, 1981). The presence of mould growth does not imply that mycotoxins are present in cheese. Bullerman (1976) analyzed mouldy Swiss cheese for the presence of penicillic acid and reported that 4 of 33 samples were positive. Lafont et af. (1979b) reported a low incidence of penicillic acid in 1 of 110 mouldcontaminated Blue cheese samples, 3 of 48 samples of hard cheese, 5 of 39 semihard cheeses, and 2 of 18 goat-milk cheese samples. These investigators also reported a low incidence of patulin in hard cheese (1 of 48samples), semihard cheese (4 of 39 samples), and goat-milk cheese (1 of 18 samples), and mycophenolic acid in hard cheese (4 of 48 samples) and semihard cheese (7 of 39 samples). However, high incidence of ochratoxin A and citrinin were reported in mouldy cheese samples in the UK (Jarvis, 1983). Richard and Arp (1979) found penitrem A in refrigerated mouldy cream cheese. Several studies on mouldy cheese for Penicillium mycotoxins gave negative results (Nowotny etaf.,1983;Williams, 1985;Fritz and Engst, 1981). Work has also been conducted on the incidence of mycotoxins in cheese contaminated with Aspergiflus spp. A low incidence of sterigmatocystin was reported in a number of studies (Lafont ef al., 1979b; Northolt et af., 1980; Bartos and Matyas, 1982), but it was not detected in others (Steering Group on Food Surveillance, 1987; Nowotny et al., 1983; Luck et af., 1976; Bullerman, 1976). There is very little evidence that significant levels of aflatoxins are produced in cheese contaminated with Aspergiffusspp. and many surveys have reported negative results (see Scott, 1989). Some work has been undertaken on the ability of mycotoxins to migrate from the surface of cheese into the interior. Data on this topic are of significance in making objective decisions on whether or not to trim or discard mould-contaminated cheese. However, interpretation of much of the data is difficult since relatively high incubation temperatures were used. The Health Protection Branch of Health and Welfare, Canada (Anonymous 1981) has recommended that if a hard cheese is contaminated with a patch
290
P. F. FOX et al.
of mould growth, the cheese can be salvaged by removing the infected portion to a depth of 2.5 cm. In view of both economic and public health implications, more research on the migration of mycotoxins is required. J. BIOGENIC AMINES IN CHEESE Biogenic amines can induce significant physiological effects in Man and animals under certain conditions. Biogenic amines are found in a wide variety of foods, including cheese (Maga, 1978; Smith, 1981; Rice et al., 1976; Chang et al., 1985; McCabe, 1986; Joosten, 1988). In cheese, biogenic amines are produced via decarboxylation of amino acids during ripening. Levels vary depending on the ripening period, the intensity of flavor development, and the microflora (Renner, 1987). High levels of biogenic amines are most likely in cheese heavily contaminated with spoilage microorganisms (Joosten, 1987). The principal biogenic amines found in cheese are histamine, tyramine, tryptamine, putrescine, cadaverine, and phenylethylamine. Renner (1987) reported average values of histamine and tyramine in some cheeses (Table XVI). The ingestion of amine-containing foods may cause food poisoning (Rice et al., 1976; Smith, 1981). However, for most individuals, ingestion of even large concentrations of biogenic amines does not elicit toxicity symptoms since they are rapidly converted to aldehydes by mono- and diamine oxidases and then to carboxylic acids by oxidative deamination (Edwards and Sandine, 1981). However, if mono- and diamine oxidases are impaired either due to a genetic defect or administration of inhibitory drugs, adverse reactions may occur on ingestion of biogenic amines (Rice et aL, 1976; McCabe, 1986; Joosten, 1988; Voight et al., 1974; Diamond et al., 1987). TABLE XVI AVERAGE TYRAMINE AND HISTAMINE CONTENTS OF SOME CHEESE VARIETIES' ~
Cheese variety Cheddar Emmental Blue Edam, Gouda Camembert, Brie Cottage a
~~
Tyramine
Histamine
(PW
(CLg/g)
910 190 440 210 140
110 100
5
Adapted from Renner (1987).
400
35 30 5
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS 291
Histamine is a normal constituent of the body; it is formed from histidine by a pyridoxal phosphate-dependent decarboxylase and mediates several important body functions (Douglas, 1980). The concentration of histamine in blood is strictly regulated and orally administered histamine will cause poisoning only when the regulatory mechanism fails to counteract all the ingested histamine, e.g., ingestion of a very high dose or impairment of histamine metabolism by other toxic substances (Taylor, 1986). Toxic symptoms, sometimes referred to as scombroid poisoning because they often result from consumption of fish of the Scombroidae family, become apparent within several minutes to 3 hr after ingestion of the histamine-containing food. Initially, a flushing of the face and neck occurs, often followed by an intense throbbing headache. Other symptoms sometimes reported include cardiac palpitations, dizziness, faintness, rapid and weak pulse, gastrointestinal complaints, bronchospasms, and respiratory distress (Taylor, 1986). Oral ingestion of up to 1 mmol (ca. 100 mg) of histamine does not elicit toxic symptoms in normal individuals (Motil and Scrimshaw, 1979). However, vasodilation and increased heart rate result on intravenous administration of 0.07 pmol, demonstrating the importance of histaminemetabolizing enzymes in the digestive tract. More research is needed to define nontoxic levels of histamine in foods which may contain other substances that potentiate the action of histamine, e.g., putrescine and cadaverine (Bjeldanes et al., 1978). Construction of an overall biogenic amine index may be valuable for the establishment of regulatory limits (Joosten, 1988). Most cases of histamine poisoning are associated with fish (Taylor, 1986) and only a few cases due to the consumption of cheese have been reported. Gouda containing 85 mg histamine/100 g cheese was implicated in an outbreak in Holland (Doeglas et al., 1967). Salt-tolerant lactobacilli, which contaminated the rennet, were considered the most likely factor responsible for the high levels of histamine (Stadhouders and Veringa, 1967). Two outbreaks of histamine poisoning have been reported in the United States. In 1978, 38 people exhibited histamine toxicity symptoms following consumption of Swiss cheese containing more than 9 mmol/kg of histamine (Chambers and Staruszkiewich, 1978) and in 1980,6 people were poisoned by Swiss cheese containing 16.8 mmol/kg of histamine (Taylor et al., 1982). Unlike histamine, tyramine is normally present at very low concentrations in the body. If tyramine enters the bloodstream it causes a release of noradrenaline from the sympathetic nervous system ( Joosten, 1988). Noradrenaline is important in many physiological reactions, on which tyramine can exert an indirect effect. The most common tyramine-induced responses include increased blood pressure, increased cardiac output, peripheral vasoconstriction and headache (Voight et al., 1974; Rice et al., 1976; McCabe, 1986; Diamond et al., 1987).
292
P. F. FOX et al.
In humans, monoamine oxidase (MA0)-catalyzed oxidative deamination to p-hydroxyphenylactic acid is the main tyramine degradative pathway. However, if a genetic deficiency of M A 0 exists or if M A 0 inhibitor drugs are administered, the normal route of tyramine degradation may be impaired and the toxicity symptoms outlined above may be manifested. Individuals whose M A 0 function is impaired are susceptible to the toxic action of tyramine, sometimes referred to as the “cheese reaction.” This involves a hypertensive crisis, usually accompanied by severe headache and in certain cases could lead to intercranial hemorrhage, cardiac failure, and pulmonary edema. Fatal incidences have been reported (Asatoor et al., 1963). It is important to note that other foods besides cheese, such as marinated herring, dry sausage, and marmite, may contain high levels of tyramine (Crocco, 1979). The tyramine content of cheese varies with the degree of ripening, intensity of flavor development, and microflora (Renner, 1987) and may vary significantly within a block of cheese (Chang et al., 1985). Unripened cheeses, e.g., Cottage cheese, contain negligible levels of tyramine (Voight et al., 1974; Rice et al., 1976; McCabe, 1986), while long-ripened varieties, such as Cheddar, may contain high levels (McCabe, 1986). If mature cheeses with a high tyramine content are used in processed cheeses, these cheeses may also have a high tyramine content. Whenever M A 0 inhibitor drugs are prescribed, patients should be advised to avoid intake of tyramine-rich foods. Food poisoning unambiguously caused by the consumption of tyramine-rich foods in the absence of M A 0 inhibitor drugs has not been reported (Joosten, 1988). The toxicity threshold for tyramine alone has been estimated at 3 mmol (ca. 400 mg) and it has been concluded that healthy individuals can tolerate the consumption of large amounts of tyramine-rich cheese (Grind et al., 1986). X.
PERSPECTIVE
Cheese and wine are the outstanding examples of food biotechnology. Both originated at the dawn of civilization, perhaps 10,000 years ago. Both are surrounded by a certain mystique, have attained epicurean status, and are usually regarded as complementary. Cheese and wine are available in a great diversity of forms, both undergo long maturation periods, both have very complex flavor profiles, and, if properly produced, both improve with age. Although cheese and wine were discovered by accident, both would be regarded as outstanding biotechnological achievements if discovered today. Although cheese production is a very ancient and traditional craft, cheese still has a very young image. It enjoys a consistent and substantial growth
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
293
rate, with a healthy and positive image. It could be regarded as the original convenience food which may be consumed as the main component of a meal, as a dessert, as a snack, as a condiment, or as a food ingredient. This latter application has become a major growth area, e.g., it now represents -25% of total cheese consumption in the United States, largely through the increasing popularity of pizza, a product now common in many countries. Another significant feature of the cheese industry is that although there are at least 500 varieties, new varieties continue to be developed, usually by hybridizing varieties, e.g., Jarlsberg (Norway), Leerdamer (the Netherlands), Araglin (Ireland), and Proosdij (the Netherlands); interestingly, the first three of these are variants of Swiss-type cheese. Cheese is an exciting research subject; it is produced in a great diversity of varieties, is complex, dynamic, and vital, and offers a challenge to scientists from many disciplines, e.g., analytical and colloidal chemists, biochemists, microbiologists, rheologists, and nutritionists, as well as engineers and technologies. This broad scientific appeal, coupled with its economic and dietary importance, has led to an extensive scientific and technical literature on cheese. In this review, we have described the current state of knowledge on many aspects of cheese production. The review has concentrated on the chemical, biochemical, and nutritional aspects, with only cursory coverage of such important physicochemical aspects as rheology and texture. The microbiological and public health aspects of cheese were not considered (apart from mycotoxins and biogenic amines). The technological and engineering aspects of cheese production were omitted also. These important topics were excluded mainly in the interests of homogenity of presentation. It seemed pertinent to conclude this review by considering the directions research on the chemistry and biochemistry of cheese will probably take in the immediate future. The enzymatic aspects of milk coagulation are now rather well established at the molecular level and it is not readily apparent what new work of major significance might be done in this area. Undoubtedly, information on the enzymatic reaction will be refined but the principal reactions are already well established. While the mechanism of coagulation of the rennet-altered micelles is understood in general terms, the precise chemical or physicochemical interactions that lead to gelation remain obscure. Perhaps further advances must await further refinement of our knowledge of the structure of the casein micelle, especially detailed knowledge of its surface and the changes therein as a consequent of enzymatic hydrolysis of K-casein. Postaggregation phenomena, i.e., development of the gel network, its strength, and syneresis of the gel when cut or broken, are probably a continuation of the aggrega-
294
P. F. FOX et al.
tion phase, modified by changing conditions of temperature and pH. Further work on the physicochemical aspects of these phenomena is required. Considerable progress has been made during the past decade on the isoelectric coagulation of casein micelles in the formation of Quarg and fromage frais-type products. It is likely that work will continue on the coagulation mechanism, rheology, and physical stability of this important and increasingly popular group of dairy products. The principal features of the primary biochemical events in ripening, especially glycolysis and lipolysis, are well established. Proteolysis, especially in long-ripened cheeses, is much more complex than glycolysis and lipolysis; the initial proteolytic events are now well established and the great diversity of peptides and the enzymes that produce them are now being elucidated. It appears reasonable to predict that the proteolytic diversity, pathways, and kinetics in the principal cheese varieties will be established during the next 5-10 years. Undoubtedly, the products of these primary biochemical events, i.e., fatty and other acids, peptides, and amino acids, contribute to cheese flavor, perhaps very significantly in many varieties and proteolysis certainly has a major influence on the various rheological properties of cheese, e.g., texture, meltability, and stretchability. However, the finer points of cheese flavor are almost certainly due to further modification of the products of the primary reactions. The most clear-cut example of this is the oxidation of fatty acids to methyl ketones in blue cheeses. Catabolism of amino acids leads to the production of numerous sapid compounds, including amines, carbonyls, acids, thiols, and alcohols. Many of these compounds may interact chemically with each other and the compounds of other reactions via the Maillard and Strecker reactions. At present, relatively little is known concerning the enzymology of amino acid catabolism in most cheeses and even less is known about the chemical reactions. It is very likely that research attention will focus on these secondary and tertiary reactions in the short-term future. In spite of extensive and intensive research over the last 40 years, the flavor of cheese remains elusive. While very considerable qualitative and quantitative information is available on the aromatic and flavorful compounds in many cheese varieties is now available, it is not yet possible to fully define cheese flavor in chemical terms. Extensive comparative studies of the volatile and nonvolatile low molecular weight compounds, both of cheeses of different quality characteristics of the same variety and between varieties should be useful. Numerous studies on cheese flavor have been published but the number and diversity of cheese, both with respect to the range of quality attributes and cheese types analyzed in such studies, have been rather limited. Although expensive, such a large-scale study appears
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
295
warranted. To date, most studies on cheese flavor have concentrated on volatile compounds. The water-soluble, nonvolatile fraction, which contains small peptides, amino acids, various low molecular weight acids, and NaC1, has a definite savory, cheesy flavor. Any further large-scale study of cheese flavor should include analysis of both the volatile and nonvolatile fractions and the data should be subjected to multivariate analyses. It is almost certain that the flavor of cheese develops through the action of microorganisms andlor their enzymes. The starter cultures now used in the principal cheese varieties are very highly refined, perhaps only a single strain. Although these strains perform very reproducibly in terms of acid production and produce cheese free from off-flavors, many authorities feel that overrefinement has led to the production of cheese with a low flavor intensity. Apart from acid production, the mechanism of which is well understood, the precise contribution of the starter to flavor development is not yet known although many of their principal enzymes have been characterized at the biochemical and genetic level. However, the key compounds in cheese flavor, and hence the enzymes responsible for their production, are not yet known precisely. When these key reactions have been identified, the ability to modify the genetic make-up of starters will enable the engineering of strains that overproduce the key enzymes. Phage infection is still the principal problem faced by cheese makers. Mechanisms by which lactic acid bacteria resist phage infection are known and work is in progress to engineer strains with superior phage resistance. At present, many of the techniques used to genetically engineer starter bacteria are not food-grade but it is almost certain that food-grade techniques will be developed which will enable super tailor-made strains to be developed. Very exciting developments can be expected in this area. Improved hygiene on the farm and at the factory, pasteurization of cheesemilk, and the use of enclosed, automated cheesemaking equipment has reduced the numbers and diversity of nonstarter bacteria in cheese. While this has led to greater uniformity of cheese quality, it has probably also reduced flavor intensity. To offset this, NSLAB are sought which will permit the production of highly flavored cheese consistently and predictability. Although NSLAB are not yet well characterized at the genetic level, it will probably soon be possible to genetically engineer NSLAB to produce cheeses with a desired type and intensity of flavor. In this review we did not discuss the possibility of grading cheese objectively. As discussed in Section V, cheese is graded subjectively by trained graders who presumably reflect consumer preferences. Subjective grading is expensive and varies between factories and probably over time. Therefore, objective methods for quality assessment are being sought. While the maturity of cheese can be estimated fairly accurately by objective chemical and/
296
P. F. FOX et al.
or physical methods, it is not yet possible to reliably assess the quality of cheese by such methods. However, as the depth of our knowledge on the biochemistry and flavor chemistry of cheese increases, it is very likely that it will become possible to objectively assess the quality of cheese. Study of chemistry and biochemistry of cheese is at an exciting stage. It seems reasonable to predict that it will be possible in the not too distant future to describe completely the production and ripening of cheese at the molecular level.
REFERENCES Adda, J., Gripon, J-C., and Vassal, L. (1982). The chemistry of flavour and texture generation in cheese. Food Chem. 9, 115-129. Addeo, F., Chianese, L., Salzano, A., Sacchi, R., Cappuccio, U., Ferranti, P., and Molorni, A. (1992). Characterization of the 12% trichloroacetic acid insoluble oligopeptides of Parmigiano-Reggiano cheese. J. Dairy Res. 59, 401-411. Addeo, F., Chianese, L., Sacchi, R., Musso, S . P., Ferranti, P., and Molorni, A. (1994). Characterization of the oligopeptides of Parmigiano-Reggiano cheese soluble in 120 g trichloroacetic acid/l. J. Dairy Res. 61, 365-374. Adler-Nissen A. (1986). “Enzymic Hydrolysis of Food Proteins.” Elsevier, London. Aimutis, W. R., and Eigel, W. N. (1982). Identification of A-casein as plasmin-derived fragments of bovine a,1-casein. J . Dairy Sci. 65, 175-181. Aishima, T., and Nakai, S. (1987). Pattern recognition of GC profiles for classification of cheese variety. J . Food Sci. 52,939-942. Akkerman, J. C. (1992). Drainage of curd. Ph.D. Thesis, The Agricultural University, Wageningen, The Netherlands. Allcroft, R., and Carnaghan, R. B. A. (1962). Groundnut toxicity: Aspergillus flavus toxin (aflatoxin) in animal products. Vet. Rec. 74, 863-864. Allcroft, R., and Carnaghan, R. B. A. (1963). Groundnut toxicity: An examination for toxin in human products from animals fed toxic groundnut meal. Vet. Rec. 75, 259-263. Allcroft, R., Rogers, H., Lewis, G., Nabney, J., and Best, P. E. (1966). Metabolism of aflatoxin in sheep: Excretion of the milk toxin. Nature (London) 209, 154-155. Andrews, A. T., and Alichanidis, E. (1983). Proteolysis of caseins and the proteose peptone fraction of bovine milk. J. Dairy Res. 50, 275-290. Anis, S. M. K., and Emstrom, C. A. (1984). Calcium and the meltability of process cheese food made from ultrafiltrated retentate. J. Dairy Sci. 67, Suppl., 79 (abstr.) Anonymous (1981). “Mould-More than Meets the Eye.” Health Protection Branch, Health and Welfare Canada, Ottawa. Anonymous (1986). The world’s first continuous Cheddar cheese process. Caseus 3(2), llD-22D. Anonymous (1993). Healthy options. Food Process. 62(1), 26. Arora, G., and Lee, B. H. (1992). Purification and characterization of aminopeptidase from Lactobacillus casei ssp. casei LLG. J. Dairy Sci. 75, 700-710. Asatoor, A., Levi, A., and Milne, M. (1963). Tranylcypromine and cheese. Lancet 2,733-734. Aston, J. W., and Creamer, L. K. (1986). Contribution of the components of the water-soluble fraction to the flavour of Cheddar cheese. N. Z. J. Dairy Sci. Technol. 21,229-248.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
297
Aston, J. W., and Dulley, J. R. (1982). Cheddar cheese flavour. Aust. J. Dairy Technol. 37,59-64. Aston, J. W., Durward, I. G., and Dulley, J. R. (1983). Proteolysis and flavour development in Cheddar cheese. Aust. J. Dairy Technol. 38,55-59. Atlan, D., Laloi, P.,and Portalier, R. (1989). Isolation and characterization of aminopeptidasedeficient Lactobacillus bulgaricus mutants. Appl. Environ. Microbiol. 55, 1717-1723. Atlan, D., Laloi, P., and Portalier, R. (1990).X-Proline-dipeptidyl aminopeptidase of Lactobacillus delbrueckii subsp. bulgaricus: Characterization of the enzyme and isolation of deficient mutants. Appl. Environ. Microbiol. 56, 2174-2179. Baankreis, R. (1992). The role of lactococcal peptidases in cheese ripening. Ph.D. Thesis, University of Amsterdam, The Netherlands. Bacon, C. L., Wilkinson, M., Jennings, P. V., Ni Fhaolain, I., and O'Cuinn, G. (1993). Purification and characterization of an aminotripeptidase from cytoplasm of Lactococcus lactis subsp. cremoris AM2. lnt. Dairy J. 3, 163-177. Baribo, L. E., and Foster, E. M. (1952). The intracellular proteinases of certain organisms from cheese and their relationship to the proteinases in cheese. J. Dairy Sci. 35,149-160. Barnby-Smith, F. M. (1992). Bacteriocins: Applications in food preservation. Trends Food Sci. Technol. 3, 133-137. Bartos, J.. and Matyas, 2. (1982). Examination of cheeses for sterigmatocystin. Vet. Med. 27,747-752. Bernback, S., Hernell, O., and Blackberg, L. (1985). Purification and molecular characterization of bovine pregastric lipase. Eur. J. Biochem. 148, 233-238. Beyer, H. J., and Kessler, H. G. (1988). Optimierte Magermilch- und Vollmilchjoghurttechnologie. Dtsch. Milchwirrsch. (Gelsenkirchen, Ger.) 39, 992-995. Bhowmik, T., and Marth, E. H. (1988). Protease and peptidase activity of Micrococcus species. J . Dairy Sci. 71, 2358-2365. Bhowmik, T., and Marth, E. H. (1989). Esterolytic activities of Pediococcus species. J. Dairy Sci. 72,2869-2872. Bhowmik, T., and Marth, E. H. (1990a). Esterases of Micrococcus species. Identification and partial Characterization. J. Dairy Sci. 73, 33-40. Bhowmik, T., and Marth. E. H. (1990b). Role of Micrococcus and Pediococcus species in cheese ripening: A review. J. Dairy Sci. 73, 859-866. Bican, P., and Spahni, A. (1993). Proteolysis in Swiss-type cheeses: A high-performance liquid chromatography study. lnt. Dairy J. 3, 73-84. Bills, D. D., and Day, E. A. (1964). Determination of the major free fatty acids of Cheddar cheese. J. Dairy Sci. 47, 733-738. Bines, V. E., Young, P., and Law, B. A. (1989). Comparison of cheese made with a recombinant calf chymosin and with standard calf rennet. J. Dairy Res. 56,657-664. Bjeldanes, L. F., Schultz, D. E., and Morris, M. M. (1978). On the aetiology of scromboid poisoning: Cadaverine potentiation of histamine toxicity in the guinea-pig. Food Cosmer. Toxicol. 16, 157-159. Blanc, B., Lauber, E., and Sieber, R. (1983). Binding of aflatoxin to milk proteins. Microbiol. Aliments, Nut. 1, 163-177. Blanc, B., Laloi, P., Altan, D., Gilbert, C., and Portalier, R. (1993). Two cell-wall-associated aminopeptidases from Lactobacillus helveticus and the purification and characterization of APII from strain IT GL1. J. Cen. Microbiol. 139, 1441-1448. Bockelmann, W., Fobker, M., and Teuber, M. (1991). Purification and characterization of the X-prolyl-dipeptidyl-aminopeptidasefrom Lactobacillus delbrueckii subsp. bulgaricus and Lactobacillus acidophilus. lnt. Dairy J. 1, 51-66
298
P. F. FOX et al.
Bockelmann, W., Schulz, Y., and Teuber, M. (1992). Purification and characterization of an aminopeptidase from Lactobacillus delbrueckii subsp. bulgaricus. Int. Dairy J. 2,95107. Bochelmann, W., Beuck, H.-P., Lick, S., and Heller, K. (1995). Purification and characterization of a new tripeptidase from Lactobacillus delbrueckii ssp. bulgaricus B14. Int. Dairy J. 5,493-502. Boehringer-Mannheim, M. (1986). “Methods of Biochemical Analysis and Food Analysis.” Boehringer Mannheim GmbH, Biochemica, Mannheim, Germany. Booth, M., Ni Fhaolain, I., Jennings, P. V., and O’Cuinn, G. (1990a). Purification and characterization of a post-proline dipeptidyl aminopeptidase from Streptococcus cremoris AM2. J. Dairy Res. 57,89-99. Booth, M., Jennings, P. V., Ni Fhaolain, I., and O’Cuinn, G. (1990b). Prolidase activity of Lactococcus lactis subsp. cremoris AM2: Partial purification and characterization. J . Dairy Res. 57, 245-254. Bosman, B. W., Tan, P. S. T., and Konings, W. N. (1990). Purification and characterization of a tripeptidase from Lactococcus lactis subsp. cremoris Wg2. Appl. Environ. Microbiol. 56, 1839-1843. Bosset, J. O., and Gauch, R. (1993). Comparison of the volatile flavour compounds of six European ‘AOC’ cheeses by using a new dynamic headspace GC-MS method. Inr. Dairy J. 3, 359-377. Brackett, R. E., and Marth, E. H. (1982a). Fate of aflatoxin MI in Cheddar cheese and in process cheese spread. J. Food Prot. 45, 549-552. Brackett, R. E., and Marth, E. H. (1982b). Fate of aflatoxin MI in Parmesan and Mozzarella cheese. J. Food Prot. 45,597-600. Brackett, R. E., and Marth, E. H. (1982~).Association of aflatoxin MI with casein. Z. Lebensm.Unters. -Forsch. 174, 439-441. Brackett, R. E., Applebaum, R. S., Wiseman, D. W., and Marth, E. H. (1982). Fate of aflatoxin MI in Brick and Limburger-like cheese. J. Food Prot. 45,553-556. Brignon, G., Chtourou, A., and Ribadeau-Dumas, R. (1985). Preparation and amino acid sequence of human K-casein. FEBS Lett. 188,48-54. Bringe, N. A., and Kinsella, J. E. (1990). Acidic coagulation of casein micelles: Mechanisms inferred from spectrophotometric studies. J. Dairy Res. 57, 365-375. Briozzo, J., de Lagarde, E. A., Chirife, J., and Parada, J. L. (1983). Wacstum und ToxinProduktion von Clostridium botulinum Typ A in Medium and Schmelzkaeseaufstrich. Appl. Environ. Microbiol. 45, 1150-1 152. Brooker. B. E. (1979). Milk and its products. In “Food Microscopy” (J. G. Vaughn, ed.), pp. 273-311. Academic Press, London. Broome, M. C., Krause, D. A,, and Hickey, M. W. (1990). The use of non-starter lactobacilli in Cheddar cheese manufacture. Aust. J. Dairy Technol. 45, 67-73. Broome, M. C., Krause, D. A., and Hickey, M. W. (1991). The use of proteinase negative starter and lactobacilli in Cheddar cheese manufacture. Ausr. J. Dairy Technol. 46,6-11. Brown, C. A. (1982). Atlatoxin MI in milk. Food Technol. Aust. 34,228-231. Brummel, S. E., and Lee, K. (1990). Soluble hydrocolloids enable fat reduction in process cheese spreads. J. Food Sci. 55, 1290-1292. Buchheim, W., and Thomasow, J. (1984). Strukturveranderanderungen in hochfettem Frischkaese durch thermische Behandlungen und Schmelzsalze. Nordeuropaische Molkerei-2. 50(2), 38-44. Bullerman, L. B. (1976). Examination of Swiss cheese for incidence of mycotoxin producing molds. J. Food Sci. 41, 26-28. Bullerman, L. B. (1977). Mold control in the cheese plant. Ital. Cheese J. 6, 1-8.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
299
Bullerman, L. B. (1981). Public health significance of molds and mycotoxins in fermented dairy products. J. Dairy Sci. 64, 2439-2452. Bullerman, L. B., and Olivigni, F. J. (1974). Mycotoxin producing potential of molds isolated from Cheddar cheese. J. Food Sci. 39, 1166-1168. Caric, M., and Kalab, M. (1987). Processed cheese products. In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), Vol. 2, pp. 339-383. Elsevier, London. Carles, C., and Ribadeau-Dumas, B. (1984). Kinetics of action of chymosin (rennin) on some peptide bonds of bovine @-casein.Biochemistry 23,6839-6843. Carles, C., and Ribadeau-Dumas, B. (1985). Kinetics of the action of chymosin (rennin) on a peptide bond of bovine cy,,-casein: Comparison of the behaviour of this substrate with that of @-caseinsand waseins. FEBS Lett. 185,282-286. Chambers, T., and Staruszkiewich, W. (1978). Fluorometric determination of histamine in cheese. J. Assoc. Off: Anal. Chem. 61, 1092-1097. Chang, S. F., Ayres, J. F.,and Sandine, W. E. (1985). Analysis of cheese for histamine, tyramine, tryptamine, histidine, tyrosine and tryptophane. J. Dairy Sci. 68,2840-2846. Chapman, H. R., Mabbitt, L. A., and Sharpe, M. E. (1966). Apparatus and techniques for making cheese under controlled bacteriological conditions. Int. Dairy Congr. [Proc.], I7rh, Munich 1966, Vol. D l , pp. 55-60. Chapot-Chartier, M.-P., Deniel, C., Rousseau, M., Vassal, L., and Gripon, J. C. (1994). Autolysis of two strains of Lactococcus lactis during cheese ripening. Int. Dairy J. 4, 251-269. Charton, E., Davies, C., and McCrae, A. R. (1992). Use of specific polyclonal antibodies to detect heterogeneous lipases from Geotrichum candidum. Biochim. Biophys. Acta 1127,191-198. Chen, A. H., Larkin, J. W., Clark, C. J., and Irwin, W. E. (1979). Textural analysis of cheese. J. Dairy Sci. 62, 901-907. Chobert, J.-M.. Mercier, J. C., Bahy, C., and Haze, G. (1976). Structure primaire du caskinomacropeptide des casein porcine et humane. FEBS Lett. 72, 173-178. Cliffe, A. J., Marks, J. D., and Mulholland, F. (1993). Isolation and characterization of nonvolatile flavours from cheese. Peptide profile of flavour fractions from Cheddar cheese determined by reverse-phase high performance liquid chromatography. Inf. Dairy J. 3, 379-387. Code of Federal Regulations (CFR) (1988). Part 133: Cheese and related cheese products. In “Food and Drugs 21. Code of Federal Regulations. Parts 100 TO 169.” Office of the Federal Register, National Archives and Records Administration, U.S. Govnt. Printing Office, Washington, DC. Cogan, T. M. (1985). The leuconostocs: Milk products. I n “Bacterial Starter Cultures for Foods” (S. E. Gilliland, ed.), pp. 25-40. CRC Press, Boca Raton, FL. Cogan, T. M., and Daly, C. (1987). Cheese starter cultures. I n “Cheese: Chemistry Physics and Microbiology” (P. F. Fox, ed.), Vol. 1, pp. 179-249. Elsevier, London. Cogan, T. M., and Hill, C. (1993). Cheese starter cultures. In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), 2nd ed., pp. 193-255, Chapman & Hall, London. Collinge, S. K., and Emstrom, C. A. (1988). Relationship between soluble nitrogen at pH 4.6 and meltability of pasteurized process cheese food made from UF curd. J. Dairy Sci. 71Suppl. 1, 71 (abstr.). Committee on Medical Aspects of Food Policy (COMA) (1984). “Diet and Cardiovascular Disease,” H. M. Stationery Office, London. Coton, G . (1986). Ultrafiltration in cheesemaking. Dairy Ind. Int. 51(8), 29-32, 33. Cousins, C. M., Sharpe, M. E., and Law, B. A. (1977). The bacteriological quality of milk for Cheddar cheesemaking. Dairy Ind. Int. 42, 12-13, 15, 17.
300
P. F. FOX et al.
Creamer, L. K. (1976a). A further study of the action of rennin on &casein. N. 2. J. Dairy Sci. Technol. 11,30-39. Creamer, L. K. (1976b). Casein proteolysis in Mozzarella-type cheese. N. Z. J. Dairy Sci. Technol. 11, 130-131. Creamer, L. K. (1985). Water absorption by renneted casein micelles. Milchwissenschaft 40,589-59. Creamer, L. K. (1991). Electrophoresis of cheese. In “Chemical Methods for Evaluating Proteolysis in Cheese Maturation,” Bull. No. 261, pp. 14-28. Int. Dairy Fed., Brussels. Creamer, L. K., and Olson, N. F. (1982). Rheological evaluation of maturing Cheddar cheese. J. Food Sci. 41,631-636,646. Creamer, L. K., Lawrence, R. C., and Gilles, J. (1985). Effect of acidification of cheese milk on the resultant Cheddar Cheese. N. 2. J. Dairy Sci. Technol. 20, 185-203. Creamer, L. K., Iyer, M., and Lelievre, J. (1987). Effect of various levels of rennet addition on characteristics of Cheddar cheese made from ultrafiltered milk. N. 2. J. Dairy Sci. Technol. 22,205-214. Crocco, S. (1979). Monoamine oxidase inhibitors and dietary risks. JAMA 242, 87. Crow, V. L. (1986). Utilization of lactate isomers by Propionibacterium freudenreichii subsp. shermanii: Regulatory role for intracellular pyruvate. Appl. Environ. Microbiol. 52, 352-358. Crow, V. L., Martley, F. G., Coolbear, T., and Roundhill, S. (1996). Influence of phageassisted lysis of Lacrococcus lactis subsp. lactis ML8 on Cheddar cheese ripening. Int. Dairy J. 5 (in press). Csok, J. (1982). The effect of holding time on free- and bound-water contents of processed cheese. Proc. Int. Dairy Congr., Zlst, Moscow, Vol. I, Book 1, pp. 475-476. Dalgleish, D. A. (1981). Effect of milk concentration on the nature of curd formed during renneting-a theoretical discussion. J. Dairy Res. 48, 65-69. Dalgleish, D. G. (1992). The enzymatic coagulation of milk. In “Advanced Dairy Chemistry,” Vol. 1 (P. F. Fox, ed.), pp. 579-619. Elsevier Applied Science, London. Dalgleish. D. G. (1993). The enzymatic coagulation of milk. In “Cheese: Chemistry, Physics and Microbiology” (P.F. Fox, ed.), 2nd ed. Vol. 1,pp. 69-100. Chapman & Hall, London. Dalgleish, D. G . ,and Law, A. J. R. (1989). pH-Induced dissociation of bovine casein micelles. 11. Mineral solubilization and its relation to casein release. J. Dairy Res. 56,727-735. Dallman, P. R., Yip, R., and Johnson, C. (1984). Prevalence and causes of anaemia in the United States. Am. J. Clin. Nutr. 39,437-445. Darling, D. F., and Dickson, J. (1979). Electrophoretic mobility of casein micelles. J. Dairy Res. 46,441-451. Dartley, C. K., and Kinsella, J. E. (1971). Rate of formation of methyl ketones during Bluemould cheese ripening. J. Agric. Food Chem. 19,771-774. Davis, J. G. (1965). “Cheese.” Vols. 1 and 2. Churchill-Livingstone, London. Deiana, P., Fatichenti, F., Farris, G. A., Mocquot, G., Lodi, R., Todesco, R., and Cecchi, L. (1984). Metabolism of lactic and acetic acids in Pecorino Romano cheese made with a combined starter of lactic acid bacteria and yeasts. Lait 64, 380-394. de Koning, P. J., van Rooijen, P.J., and Visser, S. (1978). Application of a synthetic hexapeptide as a standard substrate for the determination of the activity of chymosin. Neth. Milk Dairy J. 32,232-244. de Koning, P. J., de Boer, R., Both, P., and Nooy, P. F. C. (1981). Comparison of proteolysis in a low-fat semi-hard type of cheese manufactured by standard and by ultrafiltration techniques. Neth. Milk Dairy 1. 35, 35-46. Delfour, A., Jollirs, J., Alais, C., and JollBs, P. (1965). Caseino-glycopeptides: Characterization of a methionine residue and of the N-terminal sequence. Biochem. Biophys. Res. Commun. 19.452-455.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
301
Desmazeaud, M. J., and Hermier, J. (1968a). Facteurs intervenant dans la production du systeme prottolytique chez Micrococcus caseolyticus. Ann. Biol. Anim., Biochim., BioPhyS. 8,419-429. Desmazeaud, M. J., and Hermier, J. H. (1968b). Isolement, purification et proprietts d’une protCase exocellulaire de Micrococcus caseolyticus. Ann. Biol. Anim., Biochim., Biophys. 8,565-577. Desmazeaud, M. J., and Hermier, J. H. (1971). SpCcificite de la protease neutre de Micrioccus caseolyticus. Eur. J. Biochem. 19,51-55. Desmazeaud, M. J., and Zevaco, C. (1976). General properties and substrate specificity of an intracellular neutral protease from Streptococcus diacetilactis. Ann. Biol. Anim., Biochim., Biophys. 16, 851-868. Desmazeaud, M. J., and Zevaco, C . (1977). General properties and substrate specificity of an intracellular soluble dipeptidase from Streptococcus diacetilactis. Ann. Biol. Anim., Biochim., Biophys. 17,723-736. Desmazeaud, M. J., and Zevaco, C. (1979). Isolation and general properties to two intracellular amino peptidases of Streptococcus diacetilactis. Milchwissenschaft 34, 606-610. Desmazeaud, M. J., Gripon, J.-C., Le Bars, D., and Bergere, J. L. (1976). Etude du role des micro-organisms et des enzymes au cours de la maturation des fromages. 111. Influence des micro-organisms. Lair 56, 379-396. Diamond, S., Freitag, F. G., Solomon, G. D., and Millstein, E. (1987). Migrane headache. Working for the best outcome. Postgrad. Med. 81, 174-176. Dimick, P. S., Walker, N. J., and Patton, S. (1969). Evidence of a &oxidation pathway for saturated fatty acids. Biochem. J . 111,395-399. Docherty, A. J. P., Bodiner, M. W., Angal, S., Verger, R., Riviere, C., Lowe, P. A,, Lyons, A., Emtage, J. S., and Harris, T. J. R. (1985). Molecular cloning and nucleotide sequence of rat lingual lipase cDNA. Nucleic Acids Res. 13, 1891-1903. Doeglas, M. H. G., Huisman, J., and Nater, J. P. (1967). Histamine intoxication after cheese. Lancet 2, 1361-1362. Doi, J., Ideno, S., Huang Kuo, F., Ibuki, F., and Kanamori, M. (1983). Gelation of the complex between K-casein and P-lactoglobulin. J . Nutr. Sci. Vitaminol. 29, 679-689. Douglas, W. (1980). Histamine and 5-hydroxytryptamine and their antagonists. In “The Pharmacalogical Basis of Therapeutics” (A. G. Gilman, L. S. Goodman, A. Gilman, S. E. Mayor, and K. L. Melman, eds.), 6th ed., pp. 609-646. Macmillan, New York. Dreizen, S., Dreizen, J. G., and Stone, R. E. (1961). The effect of cow’s milk on dental caries in the rat. J. Dent. Res. 40, 1025-1028. Driessen, F. M. (1989). Heat inactivation of lipases and proteinases (indigenous and bacterial). In “Heat-Induced Changes in Milk,” Bull. No. 238, pp. 71-93. Int. Dairy Fed., Brussels. Dunn; H. C., and Linday, R. C. (1985). Evaluation of the role of microbial Strecker-dervied aroma compounds in unclean-type flavors of Cheddar cheese. J. Dairy Sci. 68,2859-2874. Dybing, S . T., Wiegand, J. A., Brudvig, S. A., Huang, E. A,, and Chandan, R. C. (1988). Effect of processing variables on the formation of calcium lactate crystals on Cheddar cheese. J. Dairy Sci. 71, 1701-1710. Edgar, W. M., Bowen, W. H., Amsbaugh, S., Monell-Torrens, E., and Brunelle, J. (1982). Effects of different eating patterns on dental caries in the rat. Caries Res. 16, 384-389. Edwards, S. T., and Sandine. W. E. (1981). Public health significance of amines in cheese. J. Dairy Sci. 64, 2431-2438. Eggimann, B., and Bachmann, M. (1980). Purification and partial characterization of an aminopeptidase from Lactobacillus lactis. Appl. Environ. Microbiol. 40,876-882. Eigel, W. N. (1977). Effect of bovine plasmin on a,,-B and K-Acaseins. J . Dairy Sci. 60,13991403.
302
P. F. FOX et al.
Eigel, W. N., Butler, J. E., Ernstrom, C. A., Farrell, H. M., Jr., Harwalkar, V. R., Jenness, R., and Whitney, R. McL. (1984).Nomenclature of proteins of cow’s milk: Fifth revision. J. Dairy Sci. 67,1599-1631. El-Soda, M. (1993).Accelerated maturation of cheese. Int. Dairy J. 3,531-544. El-Soda, M., and Pandian, S. (1991). Recent developments in accelerated cheese ripening. J. Dairy Sci. 74, 2317-2335. El-Soda, M., Bergere, J. L., and Desmazeaud, M. J. (1978). Detection and localization of peptide hydrolases in Lactobacillus casei. J. Dairy Res. 45,519-524. El-Soda, M., Abd El-Wahab, H., Ezzat, N., Desmazeaud, M. J., and Ismail, A. (1986).The esterolytic and lipolytic activities of the lactobacilli. 11. Detection of esterase system of Lactobacillus helveticus, Lactobacillus bulgaricus, Lactobacillus lactis and Lactobacillus acidophilus. Lait 66,431-443. El-Soda, M., Macedo, A., and Olson, N. (1991).Aminopeptidase and dipeptidylaminopeptidase activities of several cheese related microorganisms. Milchwissenschafi 46,223-226. El-Soda, M., Ziada, N., and Ezzat, N. (1992).The intracellular peptide-hydrolase system of Propionibacterium. Microbios 72, 65-74. Emmons, D. B., Price, W. V., and Swanson, A. M. (1959).Tests to measure syneresis and firmness of Cottage cheese coagulum, and their application in the curd-making process. J . Dairy Sci. 42, 866-869. Engel, G.,and Prokopek, D. (1979). No detection of Penicillium roqueforfi-toxinin cheese. Milchwissenschafi 34,272-274. Engel, G., and Teuber, M. (1989).Toxic metabolites from fungal cheese starter cultures. In “Mycotoxins in Dairy Products” (H. P. van Egmond, ed.), pp. 163-192.Elsevier, London. Engel, G., and von Milczewski, K. E. (1977).Penicillium caseicolum, P. camemberti and P. roqueforti and their harmlessness to human health. I. Physico-chemical tests for the formation of known mycotoxins. Milchwissenschafi 32, 517-520. Engel, G.,von Milczewski, K. E., Prokopek, D., and Teuber, M. (1982).Strain specific synthesis of mycophenolic acid by Penicillium roqueforti in blue veined cheese. Appl. Environ. Microbiol. 43, 1034-1040. Engels, W. J. M., and Visser, S. (1994).Isolation and comparative characterization of compounds that contribute to the flavour of different cheese types. Neth. Milk Dairy J.
48,127-140. Eriksen, S . (1975).Flavors of milk and milk products. 1. The role of lactones. Milchwissenschafi
31,549-550. Ernstrom, C. A. (1985). Properties of products from ultrafiltered whole milk. In “New Dairy Products via new Technology,” Roc. IDF Semin., pp. 21-29. Dairy Fed., Brussels. Ernstrom, C. A., and Wong. N. P. (1974). Milk clotting enzymes and cheese chemistry. In “Fundamentals of Dairy Chemistry” (B. H. Webb, A. H. Johnson, and J. A. Alford, eds.), 2nd ed., pp. 662-771. Avi Publ. Co., Westport, CT. Exterkate, F. A. (1977). Pyrrolidone carboxylyl peptidase in Streptococcus cremoris: Dependence on the interaction with membrane components. J. Bacteriol. U9,1281-1288. Exterkate, F. A. (1990).Differences in short peptide-substrate cleavage by two cell-envelopelocated serine proteinases of Lactococcus lactis subsp. cremoris are related to secondary binding specificity. Appl. Microbiol. Biotechnol. 33, 401-406. Exterkate, F. A., and Alting, A. C. (1993).The conversion of the asl-casein-(1-23)-fragment by the free and bound form of the cell-envelope proteinase of Lactococcus lactis subsp. cremoris under conditions prevailing in cheese. Syst. Appl. Microbiol. 16, 1-8. Exterkate, F. A., and Alting, A. C. (1995).The role of starter peptidases in the initialproteolytic events leading to amino acids in Gouda cheese. Int. Dairy J. 5, 15-28.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
303
Exterkate, F. A., and de Veer, G. J. C. M. (1987). Purification and some properties of a membrane-bound aminopeptidase A from Streptococcus cremoris. Appl. Environ. Microbiol. 53, 577-583. Exterkate. F. A., Alting, A. C., and Slangen. C. J. (1991). Specificity of two genetically related cell-envelope proteinase of Lacfococcus lacris subsp. cremoris toward cr,,-casein-(1-23)fragment. Biochem J. 273, 135-139. Exterkate. F. A., Alting, A. C., and Bruinenberg, P. G. (1993). Diversity of cell envelop proteinase specificity among strains of Lacfococcus laciis and its relationship to charge characteristics of the substrate-binding region. Appl. Environ. Microbiol. 59,3640-3647. Ezzat, N., El Soda, M., El Shafei, H., and Olson. N. F. (1993). Cell-wall associated peptide hydrolase and esterase activities of several cheese-related bacteria. Food Chem. 48,19-23. Fan. T. Y., Hwang, D. H., and Kinsella, J. E. (1976). Methyl ketone formation during germination of Penicilliurn roqueforti. J . Agric. Food Chem. 24, 443-448. Farkye, N. Y., and Fox, P. F. 1990. Observations on plasmin activity in cheese. J. Dairy Res. 57,413-415. Farkye. N.Y., and Fox, P. F. (1991). A preliminary study of the contribution of plasmin to proteolysis in Cheddar cheese: Cheese containing plasmin inhibitor, 6-aminohexanoic acid. J . Agric. Food Chem. 39, 786-788. Feirtag. J. M.. and McKay, L. L. (1987). Thermoinducable lysis of temperature sensitive Streptococcus cremoris strains. J. Dairy Sci. 70, 1779-1784. Fernandes, L.. and Steele, J. L. (1993). Glutathione content of lactic acid bacteria. J. Dairy Sci. 76, 1233-1242. Fiat, A.-M., Jolles, J., and Jolles, P. (1977). Structural data concerning human caseinoglycopeptide and comparisons with ruminant caseinoglycopeptides. C. R. Hebd. Seances Ser. D 2 s q 393-395. Fish, J. S. (1957). Activity and specificity of rennin. Nature (London) 180, 345. Foissy, H. (1974). Examination of Brevibacterium linens by an electrophoretic zymogram technique. J . Gen. Microbioi. 80, 197-207. Foissy, H. (1978). Aminopeptidase from Brevibacterium linens: Production and purification. Milchwissenschafr 33, 221-223. Folkertsma, B.. Fox. P. F., and McSweeney, P. L. H. (1996). Acceleration of Cheddar cheese ripening at elevated temperatures. h i . Dairy J . (in press). Foltmann, B. (1987). General and molecular aspects of rennets. In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), Vol. 1, pp. 33-61. Elsevier, London. Foltmann, B. (1993). General and molecular aspects of rennets. In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), 2nd ed., Vol. 1, pp. 37-68. Chapman & Hall, London. Food and Agriculture Organization (FA0)(1994). “Production Yearbook,” Vol. 47, pp. 220221. FAO, Rome, Italy. Food and Nutrition Board (1980). “Recommended Dietary Allowances,” 9th rev. ed. National Academy of Sciences, National Research Council, Washington, DC. Fox, P. F. (1969). Milk-clotting and proteolytic activities of rennet, and bovine pepsin and porcine pepsin. J. Dairy Res. 36, 427-433. Fox. P. F. (1975). Influence of cheese composition on quality. Ir. J . Agric. Res. 14, 33-42. Fox, P. F. (1982). “Developments in Dairy Chemistry,” Vol. 1. Applied Science, London. Fox, P. F. (1983). “Developments in Dairy Chemistry.” Vol. 2. Applied Science, London. Fox, P. F. (1984). Proteolysis and protein-protein interactions in cheese manufacture. In “Developments in Food Proteins” (8.J. F. Hudson, ed.), Vol. 3, pp. 69-112. Elsevier, London. Fox, P. F. (1985). “Developments in Dairy Chemistry,” Vol. 3. Elsevier, London.
304
P. F. FOX et al.
Fox, P. F., ed. (1987). “Cheese: Chemistry, Physics and Microbiology,” Vols. 1 and 2. Elsevier, London. Fox, P. F. (1988). Rennets and their action in cheese manufacture and ripening. Biorechnol. Appl. Biochem. 10,522-535. Fox, P. F. (1988-89). Acceleration of cheese ripening. Food Biotechnol. 2, 133-185. Fox, P. F. (1989a). Proteolysis during cheese manufacture and ripening. J. Dairy Sci. 72,13791400. Fox, P. F. (1989b). “Developments in Dairy Chemistry,” Vol. 4. Elsevier, London. Fox, P. F. (1992). “Advanced Dairy Chemistry,” Vol. 1. Elsevier, London. Fox, P. F., ed. (1993a). “Cheese: Chemistry, Physics and Microbiology,” 2nd ed., Vols. 1 and 2. Chapman & Hall, London. Fox, P. F. (1993b). Cheese: An overview. In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), 2nd ed., Vol. 1, pp. 1-36. Chapman & Hall, London. Fox, P. F. (1994). Cheese flavour-an overview. Proc. Bienn. Cheese Ind. Conf: Ilth, Utah State University, Logan, 1994. Fox, P. F. (1995). “Advanced Dairy Chemistry,” Vol. 2. Chapman & Hall, London. Fox, P. F., and Law, J. (1991). Enzymology of cheese ripening. Food Biotechnol. 5,239-262. Fox, P. F., and Stepaniak. L. (1993). Enzymes in cheese technology. fnr. Dairy J. 3,509. Fox, P. F., Lucey, J. A,, and Cogan, T. M. (1990). Glycolysis and related reactions during cheese manufacture and ripening. CRC Crir. Rev. Food Sci. Nutr. 29, 237-253. Fox, P. F., Singh, T. K., and McSweeney, P. L. H. (1994). Proteolysis in cheese during ripening. In “Biochemistry of Milk Products” (A. T. Andrews and J. Varley, eds.), pp. 1-31. Royal Society of Chemistry, Cambridge, UK. Fox, P. F., Singh, T. K., and McSweeney, P. L. H. (1995). Biogenesis of flavour compounds in cheese. In “Chemistry of Structure/Function Relationships in Cheese” (E. L. Malin and M. H. Tunick, eds.), pp. 59-98. Plenum, New York. Frank, J. K., Orth, R., Reichle, G., and Wunder, W. (1975). Feeding experiments with rainbow trout using Camembert and Roquefort starters. Milchwissenschafr 30,594-597. Frank, J. K., Orth, R., Ivankowic, S., Kuhlmann, M., and Schmahl, D. (1977). Investigations on carcinogenic effects of Penicillium caseicolum and P. roqueforti in rats. Experientia 33,515-516. Fritz, W., and Engst, R. (1981). Survey of selected mycotoxins in food. J. Environ. Sci. Health, Part B B16, 193-210. Fryer, T. F. (1970). Utilization of citrate by lactobacilli isolated from dairy products. J. Dairy Res. 37, 9-15. Fryer, T. F., Reiter, B., and Lawrence, R. C. (1967). Lipolytic activity of lactic acid bacteria. J. Dairy Sci. 50, 388-389. Fryer, T. F., Sharpe, M. E., and Reiter, B. (1970). Utilization of milk citrate by lactic acid bacteria and ‘blowing’ of film-wrapped cheese. J. Dairy Res. 37, 17-28. Garcia de Fernando, G. D.. and Fox, P. F. (1991). Extracellular proteinases from Micrococcus G F 11. Isolation and characterization. Lair 71, 371-382. Garnot, P., and Molle, D. (1987). Heat stability of milk clotting enzymes in conditions encountered in Swiss cheese making. J. Food Sci. 52, 75-77, 87. Gedalia I., Ionat-Bendat, D., Ben-Mosheh, S . , and Shapira, L. (1991). Tooth enamel softening with a cola type drink and rehardening with hard cheese or stimulated saliva. J. Oral. Rehabil. 18, 501-506. Gedalia, I., Davidov, I., Lewinstein. I., and Shapira, L. (1992). Effect of hard cheese exposure with or without fluoride prerinse, on the rehardening of softened human enamel. Caries Res. 26.290-292.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
305
Geis, A., Bockelmann, W., and Teuber, M. (1985). Simultaneous extraction and purification of a cell wall-associated peptidase and p-casein specific protease from Streptococcus cremoris ACl. Appl. Microbiol. Biotechnol. 23.79-84. Gibel, W., Wagner, K., and Wildner, G. P. (1971). Experimental investigations on the cancerogenicity of Penicillium camemberti var. candidum. Arch. Geschwulstforsch. 38, 1-6. Gilbert, C., Atlan, D., Blanc, B., and Portalier, R. (1994). Proline iminopeptidase from Lactobacillus delbrueckii subsp. bulgaricus CNRZ 397: Purification and characterization. Microbiology 140,537-542. Gilles, J., and Lawrence, R. C. (1973). The assessment of Cheddar cheese quality by compositional analysis. N. Z . J. Dairy Sci. Technol. 8, 148-151. Glandorf, K. (1964). Verhalten und Abbau der Polyphosphate bei der Herstellung und Lagerung von Schmelzkase. Milchwissenschafi 19,332. Glover, F. A. (1985). Applications of ultrafiltration and reverse osmosis in the dairy industry. In “Ultrafiltration and Reverse Osmosis for the Dairy Industry” (F. A. Glover, ed.), Tech. Bull. No. 5, pp. 141-181. National Institute for Research and Dairying, Reading, England. Gonzalez de Llano, D., Polo, M. C., and Ramos, M. (1991). Production, isolation of low molecular mass peptides by high performance liquid chromatography. J. Dairy Rex 58,363-372. Gouda, A., El-Shabrawy, S. A., El-Zayat, A., and El-Bagoury, E. (1985). Use of calcium caseinate in processed cheese spread making. Egyp. J. Dairy Sci. 13, 115-119. Grandison, A. S., Brooker, B. E., Young, P., Ford, G. D., and Underwood, H. M. (1986). Sludge formation during the manufacture of Cottage cheese. J. SOC. Dairy Technol. 39, 119-123. Grappin, R., Rank, T. C., and Olson, N. F. (1985). Primary proteolysis of cheese proteins during ripening. J. Dairy Sci. 68, 531-540. Green, M. L. (1977). Review on the progress of dairy science. Milk coagulants. J . Dairy Res. 44,159-188. Green, M. L. (1980). The formation and structure of milk protein gels. Food Chem. 6,41-49. Green, M. L. (1985). Effect of milk pretreatment and making conditions on the properties of Cheddar cheese from milk concentrated by ultrafiltration. J. Dairy Res. 52, 555-564. Green, M. L., and Crutchfield, G. (1971). Density-gradient electrophoresis of native and of rennet-treated casein micelles. J. Dairy Res. 38, 151-164. Green, M. L., and Foster, P. D. M. (1974). Comparison of the rates of proteolysis during ripening of Cheddar cheeses made with calf rennet and swine pepsin as coagulants. J . Dairy Res. 41,269-282. Green, M. L., and Grandison, A. S. (1993). Secondary (non-enzymatic) phase of rennet coagulation and post-coagulation phenomena. In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox,ed.), 2nd ed., Vol. 1, pp. 101-140. Chapman & Hall, London. Green, M. L., Glover, F. A., Scurlock, E. M. W., Marschall, R. J., and Hatfield, D. S. (1981a). Effect of use of milk concentrated by ultrafiltration on the manufacture and ripening of Cheddar cheese. J. Dairy Res. 48, 333-341. Green, M. L., Turvey, A,, and Hobbs, D. G. (1981b). Development of structure and texture in Cheddar cheese. J. Dairy Res. 48,343-355. Green, M. L., Angal, S., Lowe, P. A., and Marston, F. A. 0. (1985). Cheddar cheesemaking with recombinant calf chymosin synthesized in Escherichia coli. J. Dairy Res. 52,281-286. Grieve, P. A., and Kitchen, B. J. (1985). Proteolysis in milk: The significance of proteinases originating from milk leucocytes and a comparison of the action of leucocyte, bacteria and natural milk proteinases on casein, J. Dairy Res. 52, 101-112. Grigorov, H. (1966). Effect of various types of heat processing of cow’s milk on the duration of the coagulation process and on the pH and acidometric titration values of Bulgarian sour milk (yoghurt). Int. Dairy Congr. [Proc.], 17th, Munich, 1966, Vol. E F , pp. 643-47.
306
P. F. FOX et al.
Grind, M., Siwers, C., Graffner, G., Alvan, L. L., Gustafsson, J., Helleday, J., Lindgren, J. E., Ogenstad, S., and Selander, H. (1986). Pressor response of oral tyramine in healthy men given amiflamine and placebo. Clin. fharmacol. Ther. 40,155-160. Gripon, J.-C. (1987). Mould-ripened cheeses. In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), Vol. 1, pp. 121-149. Elsevier, London. Gripon, J.-C. (1993). Mould-ripened cheeses. In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), 2nd ed., Vol. 2, pp. 111-136. Chapman & Hall, London. Gripon, J.-C., Desmazeaud, M. J., Le Bars, D., and Bergere, J.-L. (1975). Etude du role des micro-organismes et des enzymes au cours de la maturation des fromages. 11. Influence de la presure commerciale. Lait 55,502-516. Gripon, J.-C., Monnet, V., Lambert, G., and Desmazeaud, M. J. (1991). Microbial enzymes in cheese ripening. I n “Food Enzymology” (P. F. Fox, ed.), Vol. 1, pp. 131-168. Elsevier, London. Grufferty, M. B., and Fox, P. F. (1988). Milk alkaline proteinase: A review. J. Dairy Res. 55, 609-630. Guinee, T. P. (1990). Pasteurized processed cheese products. Co-op Ireland, February, pp. 25-28. Guinee, T. P. (1991). Natural stabilization of food. In “First Food Ingredients Symposium” (M. K. Keogh, ed.), pp. 74-87. National Dairy Products Research Centre, Moorepark, Fermoy, Co. Cork, Ireland. Guinee, T. P., and Fox, P. F. (1987). Salt in cheese: Physical, chemical and biological aspects. In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), Vol. 1, pp. 251-297. Elsevier. London. Guinee, T. P., and Fox, P. F. (1993). Salt in cheese: Physical, chemical and biological aspects. In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), 2nd ed., Vol. 1, pp. 257-302. Chapman & Hall, London. Guinee, T. P., Pudja, P. D., Mulholland, E. O., and Reville, W. J. (1992). Ultrafiltration in cheesemaking. In “Proceedings of the Third Cheese Symposium” (T. M. Cogan, ed.), pp. 21-29. National Dairy Products Research Centre, Moorepark, Fermoy, Co. Cork, Ireland. Guinee, T. P., Pudja, P. D., and Farkye, N. Y. (1993). Fresh acid-curd cheese varieties. In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), 2nd ed., Vol. 2, pp. 363-419. Chapman & Hall, London. Guinee, T. P., Pudja, P. D., and Mulholland, E. 0. (1994). Effect of milk protein standardization, by ultrafiltration, on the manufacture, composition and maturation of Cheddar cheese. J. Dairy Res. 61,117-131. Gupta, S. K., Karahidian, C., and Lindsay, R. C. (1984). Effect of emulsifier salts on textural and flavour properties of processed cheeses. J. Dairy Sci. 67,764-778. Habibi-Najafi, M. B., and Lee, B. H. (1995). Purification and characterization of proline iminopeptidase from Lactobacillus casei ssp. casei LLG. J. Dairy Sci. 78, 251-259. Harboe, M. K. (1992). “Chymogen, a Chymosin Rennet Manufactured by Fermentation of Aspergillus niger,” Bull. No. 269, pp. 3-7. Int. Dairy Fed., Brussels. Harper, D. S., Osborn, J. C., and Clayton, R. (1983). Cariostatic potential of four cheeses evaluated in a programmed-fed rat model. J. Dent. Res. 62,283 (abstr.). Harper, D. S., Osborn, J. C., Clayton, R., and Hefferren, J. J. (1987). Modification of food cariogenicity in rats by mineral-rich concentrates from milk. J. Dent. Res. 66, 42-45. Harper, J. W., Hemmi, K., and Powers, J. C. (1985). Reaction of serine proteases with substituted isocoumarins: Discovery of 3,4-dichloroisocoumarin,a new general mechanism based serine protease inhibitor. Biochemisny 24, 1831-1841. Harper, W. J., Carmona de Catril, A., and Chen, J. L. (1980). Esterase of lactic streptococci and their stability in cheese slurry systems, Milchwissenschafi 35, 129-132.
CHEESE PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
307
Harvey, C. D., Morris, H. A., and Jenness, R. (1982). Relation between melting and textural properties of process Cheddar cheese. J. Dairy Sci. 65,2291-2295. Harwalkar, V. R., and Elliott, J. A. (1971). Isolation of bitter and astringent fractions from Cheddar cheese. J. Dairy Sci. 54, 8-11. Harwalkar, V. R., and Kalab, M. (1980). Milk gel structure. XI. Electron microscopy of glucono-&lactone-induced skim milk gels. J. Text. Stud. 11, 35-49. Harwalkar, V. R., and Kalab, M. (1981). Effect of acidulants and temperature on microstructure, firmness and susceptibility to syneresis of skim milk gels. Scanning Electron Microsc. 3,503-513. Harwalkar, V. R., and Kalab, M. (1983). Susceptibility of yoghurt to syneresis. Comparison of centrifugation and drainage methods. Milchwissenschaji 38,517-522. Harwalkar, V. R., Kalab, M., and Emmons, D. B. (1977). Gels prepared by adding D-gluconoGlactone to milk at high temperatures. Milchwissenschaji 32,400-402. Hawke, J. C. (1966). Reviews of the progress of dairy science. Section D. Dairy chemistry. The formation and metabolism of methyl ketones and related compounds. J. Dairy Res. 33,225-243. Hayakawa, S., and Nakai, S. (1985). Relationships of hydrophobicity and net charge to the solubility of milk and soy proteins. J Food Sci. 50, 486-491. Hayes, K. C., Pronczuk, A., Lindsey, S., and Diersen-Schade, D. (1991). Dietary saturated fatty acids differ in their impact on plasma cholesterol and lipoproteins in non-human primates. Am. J. Clin. Nutr. 53,491-498. Heaney, R. P. (1991). “Evaluation of Publically Available Scientific Evidence Regarding Nutrient-Disease Relationships. 3. Calcium and Osteoporosis.” Life Sciences Research Office, Federation of American Societies for Experimental Biology, Rockville Pike, MD. Heertje, I., Boskamp, M. J., van Kleef, F., and Gortemaker, F. H. (1981). The microstructure of processed cheese. Neth. Milk Dairy J . 35, 177-179. Heertje, I., Visser, J., and Smits, P. (1985). Structure formation in acid milk gels. Food Microstruct. 4, 267-277. Hemme, D., Bouillanne, C., MBtro, F., and Desmazeaud, M. J. (1982). Microbial catabolism of amino acids during cheese ripening. Sci. Aliments 2, 113-123. Herod, E. L. (1991). The effect of cheese on dental caries. Aust. Dent. J. 36, 120-125. Hickey, M. W., Hillier, A. J., and Jago, G. R. (1983a). Metabolism of pyruvate and citrate in lactobacilli. Aust. J. Biol. Sci. 36, 487-496. Hickey, M. W., van Leeuwen, H., Hillier, A. J., and Jago, G. R. (1983b). Amino acid accumulation in Cheddar cheese manufactured from normal and ultrafiltered milk. Aust. J. Dairy Technol. 38, 110-113. Hicks, C. L., O’Leary, J., and Bucy, J. (1988). Use of recombinant chymosin in the manufacture of Cheddar and Colby cheese. J . Dairy Sci. 71, 1127-1131. Hill, A. R., and Smith, A. K. (1992). Texture and ultrastructure of process cheese spreads made from heat precipitated whey proteins. Milchwissenschaji 47,71-74. Hill, R. D. (1968). The nature of the rennin-sensitive bond in casein and its possible relation to sensitive bonds in other proteins. Biochem. Biophys. Res. Commun. 33,659-663. Hill, R. D. (1969). Synthetic peptide and ester substrates for rennin. J. Dairy Res. 36,409-415. Hill, R. D., Lahav, E., and Givol, D. (1974). A rennin-sensitive bond in aSlB-casein. J. Dairy Res. 41, 147-153. Hirsch, A., Grinsted, E., Chapman, H. R., and Mattick, A. T. R. (1951). A note on the inhibition of an anaerobic sporeformer in Swiss-type cheese by a nisin-producing Streptococcus. J . Dairy Res. 18, 205-207. Hirst, D., Muir, D. D., and Naes, T. (1994). Definition of the sensory properties of hard cheese: A collaborative study between Scottish and Norwegian panels. Int. Dairy J. 4 743-761.
308
P. F. FOX et nl.
Hokes, J. C., Hansen, P. M. T., and Mangino, M. E. (1989). Functional properties of commercial calcium caseinates for use in imitation cheese. Food Hydrocolloids 3,19-31. Holland, B., Unwin, I. D., and Buss, D. H. (1989). “Milk Products and Eggs: The Fourth Supplement to McCance and Widdowson’s: The Composition of Foods,” 4th ed. Royal Society of Chemistry/Ministry of Agriculture, Fisheries and Food, Cambridge, UK. Holmes, D. G . , Duersch, J. N., and Ernstrom, C. A. (1977). Distribution of milk clotting enzymes between curd and whey and their survival during Cheddar cheese making. J. Dairy Sci. 60,862-869. Holzapfel, C. W., Steyn, P. S., and Purchase, I. F. H. (1966). Isolation and structure of aflatoxins MI and M2. Tetrahedron Lett. 25,2799-2803. Hubbard, R. W., Ono, Y., and Sanchez, A. (1989). Atherogenic effects of oxidized products of cholesterol. Prog. Food Nutr. Sci. 13, 17-44. Huffman, L. M., and Kristoffersen, T. (1984). Role of lactose in Cheddar cheese manufacturing and ripening. N . 2. J. Dairy Sci. Technol. 19, 151-162. Hwang, I.-K., Kaminogawa, S., and Yamauchi, K. (1981). Purification and properties of a dipeptidase from Streptococcus cremoris. Agric. Biol. Chern. 45, 159-165. Imfeld, T. H., Hirsch, R. S., and Muhlmann, H. R. (1978). Telemetric recordings of interdental plaque pH during different meal patterns. Br. Dent. J. 139,351-356. International Dairy Federation (IDF) (1984). “Consumption Statistics for Milk and Milk Products,” Bull. No. 173, pp. 1-21. Int. Dairy Fed., Brussels. International Dairy Federation (IDF) (1986). “The World Market for Cheese,” Bull. No. 203, pp. 2-14. Int. Dairy Fed., Brussels. International Dairy Federation (IDF) (1988). “The World Market for Cheese,” Bull. No. 243. Int. Dairy Fed., Brussels. International Dairy Federation (IDF) (1991a).“Chemical Methods for Evaluation of Proteolysis in Cheese Maturation,” Bull. No. 261. Int. Dairy Fed., Brussels. International Dairy Federation (IDF) (1991b). “Rheological and Fracture Properties of Cheese,” Bull. No. 268. Int. Dairy Fed., Brussels. International Dairy Federation (IDF) (1992). “Consumption Statistics for Milk and Milk Products,” Bull. No. 270. pp. 1-21. Int. Dairy Fed., Brussels. International Dairy Federation (IDF) (1993). “Consumption Statistics for Milk and Milk Products,” Bull. No. 282. Int. Dairy Fed., Brussels. Irani, R. R., and Callis, C. F. (1962). Calcium and magnesium sequestration by sodium and potassium polyphosphates. J. Am. Oil Chem. SOC.39, 156-159. Ito, T., Okawachi, Y., and Muguruma, Y. (1976). Relationship between the emulsifying capacity of cheese and the size of casein during ripening. J. Fnc. Agric. Kyushu Vniv. 20, 79-85. Jack, F. R., and Peterson, A. (1992). Texture of hard cheeses. Trends Food Sci. Technol. 3,161-164. Jarvis, B. (1983). Mould and mycotoxins in mouldy cheese. Microbiol., Aliments, Nurr. 1, 187-191. Jelen, P., and Renz-Schauen, A. (1989). Quarg manufacturing innovations and their effects on quality, nutritive value and consumer accaptance. Food Technol. 43(3), 74-81. Jenkins, G. N., and Ferguson, D. B. (1966). Milk and dental caries. Br. Denr. J. l20,472-477. Jenkins, G. N., and Hargreaves, J. A. (1989). Effect of eating cheese on Ca and P concentrations of whole mouth saliva and plaque. Curies Res. 23, 159-164. Jenkins, G. N., and Harper, D. S. (1983). Protective effect of different cheeses in an in vitro demineralization system. J. Dent. Res. 62, 284 (abstr.). Johnston, K. A,, Dunlop, F. P., Coker, C. J., and Wards, S. M. (1994). Comparisons between the electrophoretic pattern and textural assessment of aged Cheddar made using various levels of calf rennet or microbial coagulant (Rennilase 46L). Inr. Dniry J. 4,303-327.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
309
Jolly, R. C., and Kosikowski, F. V. (1975a). A new blue cheese food material from ultrafiltered skimmilk and microbial enzyme-mold spore reacted fat. J. Dairy Sci. 58, 1272-1275. Jolly, R. C., and Kosikowski, F. V. (1975b). Quantification of lactones in ripening pasteurized milk blue cheese containing added microbial lipase. J. Agric. Food Chem. 23,1175-1176. Joosten, H. M. L. J. (1987). Conditions allowing the formation of biogenic amines in cheese. Factors influencing the amounts formed. Neih. Milk Dairy J. 41, 329-357. Joosten, H. M. L. J. (1988). The biogenic amine contents of Dutch cheese and their toxicological significance. Neth. Milk Dairy J. 42, 25-42. Kairyukshtene, I., and Zakharova, N. (1982). Investigation on the consistency of low-fat cheese. Proc. Ini. Dairy Congr., 21~1,Moscow, 1982, Vol. 1, Book 1, p. 495. Kalab, M. (1979). Scanning electron microscopy of dairy products. An overview. Scanning Electron Microsc. 111,261-271. Kalab, M., and Modler, H. W. (1985). Milk gel structure. XV. Electron microscopy of whey protein-based Cream cheese spread. Milchwissenschaft 40, 193-196. Kalab, M., Emmons, D. B., and Sargant, A. G . (1975). Milk-gel structure. IV. Microstructure of yoghurts in relation to the presence of thickening agents. J. Dairy Res. 42,453-458. Kalab, M., Emmons, D. B., and Sargant, A. G. (1976). Milk gel structure. V. Microstructure of yoghurt as related to the heating of milk. Milchwissenschafi 31, 402-408. Kalab, M., Yun, J., and Yiu, S. H. (1987). Textural properties and microstructure of process cheese food rework. Food Microsiruct. 6, 181-192. Kamaly, K. M., El-Soda, M., and Marth, E. H. (1988). Esterolytic activity of Streptococcus lactis, Sireptococcus cremoris and their mutants. Milchwissenschaff 43, 346-349. Kamaly, K. M., Takayama, K., and Marth, E . H. (1990). Acylglycerol acylhydrolase (lipase) activities of Streptococcus laciis, Sireptoiococcus cremoris and their mutants. J. Dairy Sci. 73, 280-290. Kaminogawa, S., and Yamauchi, K. (1972). Acid protease of bovine milk. Agric. Biol. Chem. 36, 2351-2356. Kaminogawa, S., Yamauchi, K., Miyazawa, S., and Koga, Y. (1980). Degradation of casein components by acid protease of bovine milk. J. Dairy Sci. 63,701-704. Kaminogawa. S., Azuma, N., Hwang, I.-K., Suzuki, Y., and Yamauchi, K. (1984). Isolation and characterization of prolidase from Streptococcus cremoris H61. Agric. Biol. Chem. 48,3035-3040. Kaminogawa, S., Yan T. R., Azuma, N., and Yamauchi, K. (1986). Identification of low molecular weight peptides in Gouda-type cheese and evidence for the formation of these peptides from 23 N-terminal residues of a,,-casein by proteinases of Strepiococcus cremoris H61. J. Food Sci. 51, 1253-1256. Karahadian, C., and Lindsay, R. C. (1987). Integrated roles of lactate, ammonia, and calcium in texture development of mold surface-ripening cheese. J . Dairy Sci. 70,909-918. Kelly, M. (1993). The effect of salt and moisture on proteolysis in Cheddar cheese. MSc. Thesis, National University of Ireland, Cork. Keys, A. (1984). Serum cholesterol response to dietary cholesterol. Am. J. Clin. Nutr. 40, 351-359. Khalid, N. M., and Marth, E. H. (1990a). Lactobacilli-their enzymes and role in ripening and spoilage of cheese: A review. J. Dairy Sci. 73, 2669-2684. Khalid, N. M., and Marth, E. H. (1990b). Proteolytic activity by strains of Lactobacillus plantarum and Lactobacillus casei. J. Dairy Sci. 73, 3068-3076. Khalid, N. M., and Marth, E. H. (1990~).Partial purification and characterization of an aminopeptidase from Lactobacillus helveiicus CNRZ 32. Appl. Environ. Microbiot. 55, 381-388.
310
P. F. FOX er al.
Khalid, N. M., and Marth, E. H. (1990d). Purification and partial characterization of a prolyldipeptidyl aminopeptidase from Lactobacillus helveticus CNRZ 32. Appl. Environ. Microbiol. 56, 381-388. Khalid, N. M., El-Soda, M., and Marth, E. H. (1990). Esterase of Lactobacillus helveticus and Lactobacillus delbrueckii ssp. bulgaricus. J . Dairy Sci. 73, 271 1-2719. Kiefer-Partch, B., Bockelman, W., Geis, A., and Teuber, M. (1989). Purification of a X-propyldipeptidyl aminopeptidase from the cell wall proteolytic system of Lactococcus lacris subsp. cremoris. Appl. Microbiol. Biotechnol. 31,75-78. Kiermeier, F., and Buchner, M. (1977a). Distribution of aflatoxin M1 in whey and curd during cheese processing. Z . Lebensm-Unters. -Forsch. 164,82-86. Kiermeier, F., and Buchner, M. (1977b). On the aflatoxin M1content of cheese duringripening and storage. Z. Lebensm.-Unters. -Forsch. 164, 87-91. Kim, B. Y., and Kinsella, J. E. (1989). Rheological changes during slow acid induced gelation of milk by D-glucono-Slactone. J. Food Sci. 54,894-898. Kimura, T., Taneya, S., and Furuichi, E. (1979). Electron microscopic observation of casein particles in processed cheese. Proc. In!. Dairy Congr.,20th, Paris, 1978,Vo1. E, pp. 239-240. Kinsella, J. E. (1984). Milk proteins: Physicochemical and functional properties. CRC Crit. Rev. Food Sci. Nutr. 21,197-262. Kinsella, J. E., and Hwang D. H. (1976). Enzymes of Penicillium roqueforti involved in the biosynthesis of cheese flavor. CRC Crit. Rev. Food Sci. Nutr. 8, 191-228. Kinsella, J. E., Patton, S., and Dimick, P. S. (1965). Chromatographic separation of lactone precursors and tentative identification of y-lactones of 4-hydroxynananoic acids in butterfat. J . Am. Oil Chem. SOC. 44,202-205. Kirchmeier, O., Weiss, G . , and Kiermeier, F. (1978). Fleissverhalten von Schmelzkaese unter der Wirkung verschiedener Phosphat. Z. Lebenbsm-Vnrer. -Forsch. 16,212-220. Kleter, G. (1976). The ripening of Gouda cheese made under aseptic conditions. 1. Cheese with no other bacterial enzymes than those from a starter Streptococcus. Nerh. Milk Dairy J. 30,254-270. Kleter, G., and de Vries, Tj. (1974). Aseptic milking of cows. Neth. Milk Dairy J. 28,212-219. Knoop, A. M. (1977). Die Ausbildung der Gallertenstruktur bei der Lab-und Sauregewinnung der Milch. Drsch. Milchwirsch. (Gelsenkirchen, Ger.) 35, 1154-1158. Kok, J. (1990). Genetics of the proteolytic system of lactic acid bacteria. FEMS Microbiol. Rev. 87,15-41. Kok, J., Leenhouts, C. J., Haandrikman, A. J., Ledeboer, A. M., and Venema, G. (1988). Nucleotide sequence of the gene for the cell wall bound proteinase of Streptococcus cremoris Wg2. Appl. Environ. Microbiol. 54, 231-238. Kombila-Moundounga, E., and Lacroix, C. (1991). The effect of combinations of sodium chloride, lactose and glycerol on rheological properties and colour of processed cheese spreads. Can. Inst. Food Sci. Technol. J. 24,239-251. Konstance, R. P., and Holsinger, V. H. (1992). Developments of rheological test methods for cheese. Food Technol. 46 (l), 105-109. Korolczuk, J., and Mahaut, M. (1989). Viscometric studies on acid type cheese texture. J. Text. Stud. 20, 169-178. Korolczuk, J., Maubois, J.-L., and Fauquant, J. (1987). Mechanisation en fromagerie de pates molles. Milk-Vital Force, Proc. lnt. Dairy Congr., 22nd, The Hague, 1986, pp. 123-128. Kosikowski, F. V. (1982). “Cheese and Fermented Milk Foods,” 2nd ed., pp. 382-412. F. V. Kosikowski and Associates, Brooktondale, NY. Kosikowski, F. V., and Mocquot, G. (1958). “Advances in Cheese Technology.” F A 0 Agric. Stud. No. 38. FAO, Rome. Kristofferson, T. (1973). Biogenesis of cheese flavor. J. Agric. Food Chem. 21, 573-575.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
311
Kristoffersen, T. (1985). Development of flavor in cheese. Milchwissenschaft 40, 197-199. Krobicka, A., Bowen, W. H., Pearson, S., and Young, D. A. (1987). The effects of cheese snacks on caries in desalivated rats. J. Dent. Res. 66, 1116-1119. Kuchroo, C. N., and Fox, P. F. (1983). A fractionation scheme for the water-soluble nitrogen in Cheddar cheese. Milchwissenschaft 38,389-391. Labropoulos, A. E., Lopez, A., and Palmer, J. K. (1981a). Apparent viscosity of milk and cultured milk yogurt thermally treated by UHT and vat systems.1.Food Prot. 44,874-876. Labropoulos, A. E., Palmer, J. K., and Lopez, A. (1981b). Whey protein denaturation of UHT processed milk and its effect on rheology of yogurt. J. Text. Stud. 12, 365-374. Lafont, P., Siriwardana, M. G., Combemale, J., and Lafont, J. (1979a). Mycophenolic acid in marketed cheeses. Food Cosmet. Toxicol. 17,147-149. Lafont, P., Siriwardana, M. G., and Lafont, J. (1979b). Contamination de fromages par des metabolites fongique. Med. Nutr. 15,257-262. Lafont, P., Lafont, J., Mousset, S., and Frayssinet, C. (1980). Etude de la contamination du lait de vache lors de l’ingestion de faibles quantities d‘aflatoxine. Ann. Nutr. Aliment. 34,699-708. Lamure, A., Pommert, J.-J., Klaebe, A., Lacabanne, C., and Perie, J.-J. (1988). Effect of polyphosphate binding on the chain dynamic of caseins: Investigation by differential scanning calorimetry and thermally stimulated currents. J. Dairy Res. 55,401-412. Langsrud, T., Reinbold, G. W., and Hammond, E. G. (1977). Proline production by Propionibacterium shermanii P59.J. Dairy Sci. 60,16-23. Langsrud, T., Reinbold, G. W., and Hammond, E. G. (1978). Free proline production by strains of propionibacteria. J. Dairy Sci. 61, 303-308. Larsen, L. B., Boisen, A., and Petersen, T. E. (1993). Procathepsin D cannot autoactivate to cathepsin D at acid pH. FEBS Lett. 319,54-58. Lavanchy, A., Berodier, F., Zonnoni, M., Noel, Y., Adamo, C., Sequella, J., and Herrero, L. (1994). “A Guide to the Sensory Evaluation of Texture of Hard and Semi-hard Cheeses,” pp. 1-40. Inst. Natl. Rech. Agron., Paris. Law, B. A. (1979). Extracellular peptidases in group N streptococci used as cheese starters. J. Appl. Bacteriol. 46,455-463. Law, B. A. (1984). Flavour development in cheeses. In “Advances in the Microbiology and Biochemistry of Cheese and Fermented Milk” (F. L. Davies and B. A. Law, eds.), pp. 187-228. Elsevier, London. Law, B. A. (1987). Proteolysis in relation to normal and accelerated cheese ripening. In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), Vol. 1, pp. 365-392. Elsevier, London. Law, J., and Haandrikman, A. (1996). Proteolytic enzymes of lactic acid bacteria. Inf. Dairy J. 6, (in press). Law, J., Fitzgerald, G. F., Uniacke-Lowe, T., Daly, C., and Fox, P. F. (1993). The contribution of lactococcal starter proteinases to proteolysis in Cheddar cheese. J. Dairy Sci. 76,24552467. Lawrence, R. C. (1989). “The Use of Ultrafiltration Technology in Cheesemaking,” Bull. No. 240, pp. 2-15. Int. Dairy Fed., Brussels. Lawrence, R. C., and Gilles, J. (1980). The assessment of the potential quality of young Cheddar cheese. N. Z. J. Dairy Sci. Technol. 15, 1-12. Lawrence, R. C., and Gilles, J. (1982). Factors that determine the pH of young Cheddar cheese. N. Z. J. Dairy Sci. Technol. 17, 1-14. Lawrence, R. C., Fryer, T. F., and Reiter, B. (1967). The production and characterization of lipases from a Micrococcus and a pseudomonad. J. Gen. Microbiol. 48,401-418.
312
P. F. FOX et al.
Lawrence, R. C., Creamer, L. K., Gilles, J., and Martley, F. G. (1972). Cheddar cheese flavour. I. The role of starters and rennets. N. 2. J . Dairy Sci. Technol. 7,32-37. Lawrence, R. C., Gilles, J., and Creamer, L. K. (1983). The relationship between cheese texture and flavour. N. Z. J. Dairy Sci. Technol. 18, 175-190. Lazaridis, H. N., Rosenau, J. R., and Mahoney, R. R. (1981). Enzymatic control of meltability in a direct acidified cheese product. J. Food Sci. 46,332-339. Le Bars, D., and Gripon, J.-C. (1989). Specificity of plasmin towards bovine as2-casein. J. Dairy Res. 56, 817-821. Le Bars, D., Desmazeaud, M. J., Gripon, J.-C., and Bergere, J. L. (1975). Etude du role des micro-organismes et de leurs enzymes dans la maturation des fromages. I. Fabrication aseptique d’un caille modelle. Lait 55, 377-389. Le Bars, J. (1979). Cyclopiazonic acid production by Peniciliurn camemberti Thom and natural occurrence of this mycotoxin in cheese. Appl. Environ. Microbiol. 38,1052-1055. Ledford, R. A., O’Sullivan, A. C., and Nath, K. R. (1966). Residua1 casein fractions in ripened cheese. J. Dairy Sci. 49, 1098-1101. Lee, B. O., and Alais, C. (1980). Etude biochimique de la fonte des fromages. Evolution des phosphates et des rnttaux. Lait 60, 130-139. Lee, B. O., Paquet, D., and Alais, C. (1979). Etude biochimique de la fonte des fromages. 1. Mesure de la peptisation. Lair 59, 589-596. Lee, B. O., Kilbertus, G., and Alais, C. (1981). Ultrastructural study on processed cheese. Effect of different parameters. Milchwissenschap 36, 343-348. Lee, B. O., Paquet, D., and Alais, C. (1986). Etude biochimique de la fonte des fromages. IV. Effect du type de sels de fonte et de la nature de la matiere prottique sue la peptisation. Utilisation d’un systerne modele. Lait 66, 257-267. Lee, H. J., Olson, N. F., and Ryan, D. S. (1980). Characterization of pregastric esterases. J. Dairy Sci. 63, 1834-1838. Lelievre, J., and Gilles, J. (1982). The relationship between the grade (product value) and composition of young commercial Cheddar cheese. N. Z. J. Dairy Sci. Technol. 17,69-75. Lelievre, J., and Lawrence, R. C. (1988). Manufacture of cheese from milk concentrated by ultrafiltration. J. Dairy Res. 55,465-478. Lemieux, L., and Simard, R. E. (1991). Bitter flavour in dairy products. I. A review of the factors likely to influence its development, mainly in cheese manufacture. Lair 71,599-636. Lemieux, L., and Simard, R. E. (1992). Bitter flavour in dairy products. 11. A review of bitter peptides from the caseins: Their formation, isolation and identification, structure masking and inhibition. Lait 72, 335-382. Lenoir, J. (1984). “The Surface Flora and its Role in the Ripening of Cheese,” Bull. No. 171, pp. 3-20. Int. Dairy Fed., Brussels. Lindsay, R. C. (1985). Flavours. In “Food Chemistry,” (0.R. Fennerna, ed.), 2nd ed., pp. 585-627. Dekker, New York. Lloyd. R. J., and Pritchard, G. G. (1991). Characterization of X-prolyl dipeptidyl aminopeptidase from Lactococcus lactis subsp. lactis. J. Gen. Microbiol. 137, 49-55. Loucheux-Lefebvre, M. H., Aubert, J.-P., and Joll6s, P. (1978). Predication of the conformation of the cow and sheep K-caseins. Biophys. J. 224,323-336. Lowrie, R. J., Lawrence, R. C., Pearce, L. E., and Richards, E. L. (1972). Cheddar cheese flavour. 111. The growth of lactic streptococci during cheesemaking and the effect on bitterness development. N. Z. J. Dairy Sci. Technol. 7,44-50. Luck, H., Wehner, F. C., Plomp, A., and Steyn, M. (1976). Mycotoxins in South African cheeses. S.Afr. J. Dairy Technol. 8, 107-110. Luyten, H. (1988). The rheological and fracture properties of Gouda cheese. Ph.D. Thesis, The Agricultural University, Wageningen, The Netherlands.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
313
Luyten, H., van Vliet, T., and Walstra, P. (1991a). Characterization of the consistency of Gouda cheese: Rheological properties. Neth. Milk Dairy J. 45, 33-53. Luyten, H., van W e t , T., and Walstra, P. (1991b). Characterization of the consistency of Gouda cheese: Fracture properties. Neth. Milk Dairy J. 45, 55-80. Maarse, H., and Visscher, C. A. (1989). “Volatile Compounds in Foods: Qualitative and Quantitative Data,” 6th ed., p. 49.1. TNO-CIVO, Food Analysis Institute, Zeist, The Netherlands. Mabbitt, L. A., Chapman, H. R., and Sharpe, M. E. (1959). Making Cheddar cheese on a small scale under controlled bacteriological conditions. J. Dairy Res. 26, 105-1 12. Machuga, E. J., and Ives, D. H. (1984). Isolation and characterization of an aminopeptidase from Lactobacillus acidophilus R-26. Biochim. Biophys. Acra 789, 26-36. Maga, J. A. (1978). Amines in food. CRC Crit. Rev. Food Sci. 10,373-403. Mahoney, R. R., Lazaridis, H. N., and Rosenau, J. R. (1982). Protein size and meltability in enzyme-treated, direct-acidified cheese products. J. Food Sci. 47, 670-671. Mann, E. J. (1986). Processeed cheese. Dairy Ind. Int. 51(2), 9-10. Mann, E. J. (1987). Processed cheese. Dairy Ind. Int. 52(2), 11-12. Mann, E. J. (1990). Processed cheese. Dairy Ind. Int. 55(11), 12-13. Manning, D. J. (1979a). Cheddar cheese flavour studies. 11. Relative flavour contributions of individual volatile components. J. Dairy Rex 46, 523-529. Manning, D. J. (1Y79b). Chemical production of essential Cheddar cheese flavour compounds. J. Dairy Res. 46, 531-537. Manning, D. J., and Moore, C. (1979). Headspace analysis of hard cheeses. J. Dairy Res. 46,539-545. Marcos, A., Esteban, M. A., Leon, F., and Fernandez-Salguero, J. (1979). Electrophoretic patterns of European cheeses: Comparison and quantitation. J. Dairy Sci. 62,892-900. Marcy, J. A., Kraft, A. A., Hotchkiss, D. K., Mollins, R. A., Olson, D. G., Walker, H. W., and White, P. J. (1988). Effect of acid and neutral pyrophosphates on the natural bacterial flora of a cooked meat system. J . Food Sci. 53,28-30. Marschall, R. J. (1990). Composition, structure, rheological properties and sensory texture of processed cheese analogues. J. Sci. Food Agric. 50,237-252. Martin, P., Collin, J.-C., Garnot, P., Ribadeau Dumas, B., and Mocquot, G. (1981). Evaluation of bovine rennets in terms of absolute concentrations of chymosin and pepsin A. J. Dairy Res. 40,447-456. Masi, P., and Addeo, F. (1986). An examination of some mechanical properties of a group of Italian cheeses and their relation to structure and conditions of manufacture. J. Food Eng. 5,217-219. Matheson, A. R. (1981). The immunochemical determination of chymosin activity in cheese. N. Z. J. Dairy Sci. Technol. 16, 33-41. McAuliffe, J. P., and O’Mullane, T. A. (1991). Preparation of cheese from natural ingredients. U. K. Pat. Appl. GB 2,237,178A. McCabe, B. J. (1986). Dietary tyramine and other pressor amines in MA01 regimens: A review. J. Am. Diet. Assoc. 86, 1059-1064. McCarthy, J. (1990). “Imitation Cheese Products,” Bull. No. 249, pp. 45-52. Int. Dairy Fed., Brussels. McEwan, J. A,. Moore, J. D., and Colwill, J. S. (1989). The sensory characteristics of Cheddar cheese and their relationship with acceptability. J. SOC. Dairy Technol. 4, 112-117. McGarry, A., Law, J., Coffey, A., Daly, C., Fox, P. F., and Fitzgerald, G. F. (1994). Effect of genetically modifying the lactococcal proteolytic system on ripening and flavour development in Cheddar cheese. Appl. Environ. Microbiol. 60,4226-4233.
314
P. F. FOX et al.
McGugan, W. A. (1975). Cheddar cheese flavor. A review of current progress. J. Agric. Food Chem. 23,1047-1050. McGugan, W. A., and Howsam, S. G . (1962). Analysis of neutral volatiles in Cheddar cheese. J. Dairy Sci. 45, 495-500. McGugan, W. A., Emmons, D. B., and Larmond, E. (1979). Influence of volatile and nonvolatile fractions on intensity of Cheddar cheese flavour. J. Dairy Sci. 62,398-403. McMahon, D. J., Yousif, B. H., and Kalab, M. (1993). Effect of whey protein denaturation on structure of casein micelles and their rennetability after ultra-high temperature processing of milk with or without ultrafiltration. Int. Dairy J. 3,239-256. McNamara, D. J. (1987). Effects of fat-modified diets on cholesterol and lipoprotein metabolism. Annu. Rev. Nutr. I , 273-290. McSweeney, P. L. H., and Fox, P. F. (1993). Cheese: Methods of chemical analysis.In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), 2nd ed., Vol. 1, pp. 341-388. Chapman & Hall, London. McSweeney, P. L. H., Nursten, H., and Urbach, G. (1995). Flavours and off-flavours in milk and dairy products. In “Advanced Dairy Chemistry” (P. F. Fox, ed.), Vol. 3. Chapman & Hall, London (in press). McSweeney, P. L. H., Fox, P. F., Lucey, J. A,, Jordan, K. N., and Cogan, T. M. (1993a). Contribution of the indigenous microflora to the maturation of Cheddar cheese. Int. Dairy J. 3,613-634. McSweeney, P. L. H., Olson, N. F., Fox, P. F., Healy, A. and Hdjrup, P. (1993b). Proteolytic specificity of chymosin on bovine aS1-casein.J. Dairy Res. 60,401-412. McSweeney, P. L. H., Olson, N. F., Fox, P. F., Healy, A., and Hdjrup, P. (1993~).Proteolytic specificity of plasmin on bovine a,l-casein. Food Biofechnol. 7,143-158. McSweeney, P. L. H., Walsh, E. M., Fox, P. F., Cogan, T. M., Drinan, F. D., and CasteloGonzalez, M. (1994a). A procedure for the manufacture of Cheddar cheese under controlled bacteriological conditions and the effect of adjunct lactobacilli on cheese quality. Ir. J. Agric. Food Res. 33, 183-192. McSweeney, P. L. H., Olson, N. F., Fox, P. F., and Healy, A. (1994b). Proteolysis of bovine aS2-caseinby chymosin. Z. Lebensm.-Unters. -Forsch. 119,429-432. McSweeney, P. L. H., Pochet, S., Fox, P. F., and Healy, A. (1994~).Partial identification of peptides from the water-insoluble fraction of Cheddar cheese. J. Dairy Res. 61,587-590. McSweeney, P. L. H., Fox, P. F., and Olson, N. F. (1995). Proteolysis of bovine caseins by cathepsin D Preliminary observations and comparison with chymosin. Int Dairy J. 5, 321-336. Melnychyn, P., and Wolcott, J. M. (1971). Interactions between phosphates and individual food components. In “Phosphates in Food Processing” (J. M. Deman and P. Melnychyn, eds.), pp. 49-64. Avi Publ. Co. Westport, CT. Meyer, A. (1973). “Processed Cheese Manufacture.” Food Trade Press Ltd., London. Meyer, J., and Jordi, R. (1987). Purification and characterization of X-prolyl-dipeptidylaminopeptidase from Lactobacillus lactis and from Streptococcus thermophilus. J . Dairy Sci. 70, 738-745. Milk Marketing Board (MMB) (1991). “EEC Dairy Facts and Figures.” Milk Marketing Board, Thames Ditton, UK. Miyakawa, H., Kobayashi, S., Shimamura, S., and Tomita, M. (1991). Purification and characterization of an X-prolyl-dipeptidyl-aminopeptidase from Lactobacillus delbrueckii ssp. bulgaricus LBU-147. 1. Dairy Sci. 74, 2375-2381. Miyakawa, H., Kobayashi, S., Shimamura, S., and Tomita, M. (1992). Purification and characterization of an aminopeptidase from Lactobacillus helveticus LHE-511. J. Dairy Sci. 75,27-35.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
315
Miyakawa, H., Hashimoto, I., Nakamura, T., Ishibashi, N., Shimamura, S., and Igoshi, K. (1994). Purification and characterization of an X-prolyl dipeptidyl aminopeptidase from Lactobaciilus helveticus LHE-511. Milchwissenschaft 49, 670-673. Modler, H. W., and Kalab, M. (1983). Microstructure of yogurt stabilized with milk proteins. J. Dairy Sci. 66, 430-437. Modler, H. W., Larmond, M. E., Lin, C. S., Froehlich, D., and Emmons, D. (1983). Physical and sensory properties of yogurt stabilized with milk proteins. J. Dairy Sci. 66,422-429. Modler, H. W., Yiu, S. H., Bollinger, U. K., and Kalab, M. (1989). Grittiness in a pasteurized cheese spread: A microscopic study. Food Microstrttct. 8,201-210. Mohr, C. M., Leeper, S. A., Engelgau, D. E., and Charboneau, B. L. (1989). Dairy applications. In “Membrane Applications and Research in Food Processing” (C. M. Mohr, S. A. Leeper, D. E. Engelgau, and B. L. Charboneau, eds.), pp. 56-66. Noyes Data Corp., Park Ridge, NJ. Monnet, V., Le Bars, D., and Gripon, J.-C. (1986). Specificity of a cell wall proteinase from Streptococcus laccis NCDO 763 towards bovine p-casein. FEMS Microbiol. Lett. 36, 127-131. Monnet, V., Bockelmann, W., Gripon, J.-C., and Teuber, M. (1989). Comparison of cell wall proteinases from Lactococcus l a d s subsp. cremoris AC1 and Lactococcus lactis subsp. lactis NCDO 763. 11. Specificity towards bovine p-casein. Appl. Microbiol. Biotechnol. 31, 112-118. Monnet, V., Ley, J. P., and Gonzalez, S. (1992). Substrate specificity of the cell envelope-located proteinase of Lactococcus lacris subsp. iactis NCDO 763. Int. J. Biochem. 24,707-718. Monnet, V., Chapot-Chartier, M. P., and Gripon, J.-C. (1993). Les peptidases des lactocoques. Lair 73, 97-108. Mora, R., Nanni, M., and Panari, G . (1984). Physical, microbiological and chemical changes in Parmigiano Reggiano cheese during the first 48 hours. Sci. Tec. Latt.-Casearia 35,20-32. Morgan, S., O’Donovan, C., Ross, R. P., Hill, C., and Fox, P. F. (1995). Significance of autolysis and bacteriocin-induced lysis of starter cultures in Cheddar cheese ripening. In “Proceedings of the Fourth Cheese Symposium” (T. M. Cogan, P. F. Fox, and R. P. Ross, eds.), pp. 51-60. Teagasc, Moorepark, Fermoy, Co. Cork, Ireland. Morr, C. V. (1985). Functionality of heated milk proteins in dairy and related foods. J. Dairy Sci. 68,2773-2781. Morrissey, R. B., Burkholder, B. D., and Tarka, S. M. (1984). The cariogenic potential of several snack foods. J. Am. Dent. Assoc. 109, 589-591. Mortensen, B. K. (1985). Recent deveIopments in the utilization of milk proteins in dairy products. Proc. Int. Congr. Milk Proteins (Galesloot, T. E., and Tinbergen, B.-J., eds.), Luxemburg, pp. 109-119. Motil, K. J., and Scrimshaw, N. S. (1979). The role of exogenous histamine in scombroid poisoning. Toxicol. Lett. 3,219-224. Mottar, J. F. (1989). Effect on the quality of dairy products. In “Enzymes of Psychrotrophs of Raw Foods” (R. C. McKellar, ed.), pp. 227-243. CRC Press, Boca Raton, FL. Mottar, J. F., Bassier, A., Joniau, M., and Baert, J. (1989). Effect of heat-induced association of whey proteins and casein micelles on yogurt texture. J. Dairy Sci. 72,2247-2256. Muir, D. D., and Hunter, E. A. (1992). Sensory evaluation of Cheddar cheese: The relation of sensory properties to perception of maturity. J. SOC. Dairy Technol. 45, 23-30. Mulder, H. (1952). Taste and flavour forming substances in cheese. Neth. Milk Dairy J. 6, 157-167. MulvihiIl, D. M., and Fox, P. F. (1977). Proteolysis of cusl-caseinby chymosin: Influence of pH and urea. J. Dairy Res. 44,533-540.
316
P. F. FOX et al.
Mulvihill, D. M., and Fox, P. F. (1978). Proteolysis of bovine &casein by chymosin: Influence of pH, urea and sodium chloride. Ir. J. Food Sci. Technol. 2, 135-139. Mulvihill, D. M., and Fox, P. F. (1979). Proteolytic specificity of chymosin on bovine a S I casein. J. Dairy Res. 46,641-651. Mulvihill, D. M., Collier, T. M., and Fox, P. F. (1979). Manufacture of coagulant-free cheese with piglet gastric proteinase. J. Dairy Sci. 62, 1567-1569. Nakae, T., and Elliott, J. A. (1965). Volatile fatty acids produced by some lactic acid bacteria. I. Factors influencing production of volatile fatty acids from casein hydrolyzate. J. Dairy Sci. 48,287-292. Nakajima, I., Kawaniski, G., and Furuichi, E. (1975). Reaction of melting salts upon micelles and their effects on calcium,phosphorus and bound water. Agric. Biol. Chem. 39,979-987. Nath, K. R., and Ledford, R. A. (1972). Caseinolytic activity of micrococci isolated from Cheddar cheese. J. Dairy Sci. 55, 1424-1427. National Dairy Council (1989). Cheese and its role in health. Dairy Counc. Dig. 60(1), 1-6. National Research Council (1989) “Diet and Health.” National Academy Press, Washington, DC. Nelson, J. H. (1975). Application of enzyme technology to dairy manufacturing.J. Dairy Sci. 58,1739-1750. Nelson, J. H., Jensen, R. G., and Pitas, R. E. (1977). Pregastric esterase and other oral lipases-a review. J. Dairy Sci. 60, 327-362. Neviani, E., Boquien, C. Y.,Monet, V., Phan Thanh, L., and Gripon, J.-C. (1989). Purification and characterization of an aminopeptidase from Lactococcus lactis subsp. cremoris AM2. Appl. Environ. Microbiol. 55,2308-2314. Ney, K. H. (1979a). Bitterness of Lipids. Ferte, Seifen, Anstrichm. 81, 467-469. Ney, K. H. (1979b). Bitterness of peptides: Amino acid composition and chain length. ACS Symp. Ser. 115,149-173. Niven, G. W. (1991). Purification and characterization of aminopeptidase A from Lactococcus lactis subsp. lactis NCDO 712. J. Gen Microbiol. 137, 1207-1212. Northolt, M. D., van Egmond, H. P., Soentoro, P., and Deijll, E. (1980). Fungal growth and the presence of strigmatocystin in hard cheese. J. Assoc. Off: Anal. Chem. 63,115-119. Nowotny, P., Bakes W., Kronert, W., and Weber, P. (1983). Thin layer chromatographic method for the determination of 22 mycotoxins in mouldy food. Chem., Mikrobiol., Technol. Lebensm. 8,24-28. Nufiez, M., Medina, M., Gaye, P., Guillen, A. M., and Rodriguez-Martin,M. A. (1992). Effect of recombinant chymosin on ewes’ milk coagulation and Manchego cheese characteristics. J. Dairy Res. 59,81-87. OCallaghan, D. M. (1994). Physicochemical, functional and sensory properties of milk protein hydrolyzates. Ph.D. Thesis, National University of Ireland, Cork. O’Connor, C. B. (1971). Composition and quality of some commercial Cheddar cheese. Ir. Agric. Creamery Rev. 24(6), 5-6. O’Connor, C. B. (1974). The quality and composition of Cheddar cheese: Effect of various rates of salt addition. Ir. Agric. Creamery Rev. 27(1), 11-13. O’Keefe, P. W., Libbey, L. M., and Lindsay, R. C. (1969). Lactones in Cheddar cheese. J. Dairy Sci. 52, 888. O’Keeffe, A. M., Fox, P. F., and Daly, C. (1977). Denaturation of procine pepsin during Cheddar cheese manufacture. J. Dairy Res. 44,335-343. O’Keeffe, A. M., Fox, P. F., and Daly, C. (1978). Proteolysis in Cheddar cheese: Role of coagulant and starter bacteria. J. Dairy Res. 45, 465-477. O’Keeffe, R. B., Fox, P. F., and Daly, C. (1975). Proteolysis in Chedddar cheese: InAuence of the rate of acid production during manufacture. J. Dairy Res. 42,111-122.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
317
O’Keeffe, R. B., Fox, P. F., and Daly, C. (1976a). Manufacture of Cheddar cheese under controlled bacteriological conditions. Ir. J . Agric. Res. 15, 151 -155. O’Keeffe, R. B., Fox, P. F., and Daly, C. (1976b). Contribution of rennet and starter proteases to proteolysis in Cheddar cheese. J. Dairy Res. 43, 97-107. Olivecrona, T., and Bengtsson-Olivecrona, G. (1991). Indigenous enzymes in milk: lipase. In “Food Enzymology” (P. F. Fox, ed.), Vol. 1, pp. 62-78. Elsevier, London. Olivecrona, T., Vilar6, S., and Bengtsson-Olivecrona, G. (1992). Indigenous enzymes in milk. 11. Lipases in milk. In “Advanced Dairy Chemistry” (P. F. Fox, ed.), Vol. 1, pp. 292-310. Elsevier, London. Ollikainen, P., and Nyberg, K. (1988). A study of plasmin activity during ripening of Swisstype cheese. Milchwissenschaft 43,497-499. Olson, N. F. (1990). The impact of lactic acid bacteria on cheese flavour. FEMS Microbiol. Rev. 87, 131-147. Olson, N. F., and Johnson, M. E. (1990). Light cheese products: Characteristics and economics. Food Technol. 44(10), 93-96. O’Shea. B. A. (1993). Objective assessment of Cheddar cheese quality. M.Sc. Thesis, National University of Ireland, Cork. O’Sullivan, M., and Fox, P. F. (1990). A scheme for the partial fractionation of cheese peptides. J. Dairy Res. 57, 135-139. O’Sullivan, M., and Fox, P. F. (1991). Evaluation of microbial chymosin from genetically engineered Kluyveromyces lactis. Food Biotechnol. 5, 19-32. Oterholm, A. (1984). “Cheesemaking in Norway,” Bull. No. 171, pp. 21-29. Int. Dairy Fed., Brussels. Oterholm, A., Ordal, Z. J., and Witter, L. D. (1970). Purification and properties of glycerol ester hydrolase (lipase) from Propionibacterium shermanii. Appl. Microbiol. 20, 16-22. Oterholm, A., Witter, L. D., and Ordal, Z. J. (1972). Purification and properties of an acetyl ester hydrolase (acetylesterase) from Lactobacillus plantarum. J. Dairy Sci. 55, 8-13. Ottosen, N. (1988). Protein Standardization,” Tech. Inf. pp. 1-33. APV Pasilac, Silkeborg Divison, Denmark. Pagliarini, E., Lembo, P., and Bertuccioli, M. (1991). Recent advancements in sensory analysis of cheese. Ital. J . Food Sci. 3, 85-99. Pahkala, E., Pihlanto-Leppala, A., Leukkanen, M., and Antila, V. (1989). Decomposition of milk proteins during the ripening of cheese. 1. Enzymatic hydrolysis of a,-casein. Meijeritiet. Aikak. 47, 39-47. Panon, G. (1990). Purification and characterization of a proline iminopeptidase from Propionibacterium shermanii 13673. Lait 70,439-452. Parnell-Clunies, E. M., Kakuda, Y. and de Man, J. M. (1986a). Influence of heat treatment of milk on the flow properties of yogurt. J. Food Sci. 51, 1459-1462. Parnell-Clunies, E. M., Kakuda, Y.,Mullen. K., Arnott, D. R. and de Mann, J. M. (1986b). Physical properties of yogurt: A comparison of vat versus continuous heating systems of milk. J . Dairy Sci. 69, 2593-2603. Patel, R. S.. Reuter, H., and Prokopek, D. (1986). Production of Quarg by ultrafiltration. J. SOC. Dairy Technol. 39, 27-31. Patterson, D. S. P., Clancy, E. M., and Roberts, B. A. (1980). The carry-over of aflatoxin MI into milk of cows fed rations containing a low concentration of aflatoxin B1. Food Cosmet. Toxicol. 18, 35-37. Patton, S. (1950). The methyl ketones of Blue cheese and their relation to its flavor. J . Dairy Sci. 33, 680-684. Paulsen, P. V., Kowalewska, J., Hammond, E. G., and Glatz, B. A. (1980). Role of microflora in production of free fatty acids and flavor in Swiss cheese. J. Dairy Sci. 63, 912-918.
318
P. F. FOX er al.
Pearse, K. N. (1976).Moving boundary electrophoresisof native and renneted-treated micelles. J. Dairy Res. 43, 27-36. Pearce, K. N., and Gilles, J. (1979). Composition and grade of Cheddar cheese manufactured over three seasons. N.Z. J Dairy Sci. Technol. 14,63-71. Pearce, K. N., Creamer, L. K., and Gilles, J. (1973). Calcium lactate deposits on rindless Cheddar cheese. N.Z. J. Dairy Sci. Technol. 8, 3-7. Pearse, M. J., and MacKinlay, A. G. (1989). Biochemical aspects of syneresis: A review. J. Dairy Sci. 72, 1401-1407. Pedersen, P. J., and Ottosen, N. (1992). Manufacture of fresh cheese by ultrafiltration. I n “New Applications of Membrane Processes,” Spec. Issue No. 9201, pp. 67-76. Int. Dairy Fed. Brussels. Pelissier, J.-P., Mercier, J.-C., and Ribadeau-Dumas, B. (1974). Etude de la proteolyse des caseines a,,-et @-bovinespar la presure. Ann. Biol. Anim., Biochim., Biophys. 14,343-362. Perez Chaia, A., Pesce de Ruiz Holgado, A., and Oliver, G. (1990). Peptide hydrolases of propionibacteria: Effect of pH and temperature. J. Food Pror. 53,237-240. Perry, K. D., and McGillivray,W. A. (1964). The manufacture of “normal” and “starter-free” Cheddar cheese under controlled bacteriological conditions. J . Dairy Res. 31, 155-165. Peterson, S. D., and Marshall, R. T. (1990). Nonstarter lactobacilli in Cheddar cheese: A review. J. Dairy Sci. 73,1395-1410. Phelan, J. A. (1985). Milk coagulants-An evaluation of alternatives to standard calf renent. Ph.D. Thesis, National University of Ireland, Cork. Phillips, G. O., Wedlock, D. J., and Williams, P. A., eds. (1986). “Gums and Stabilizers for the Food Industry 3.” Elsevier, London. Piatkiewicz, A. (1987). Lipase and esterase formation by mutants of lactic acid streptococci and lactobacilli. Milchwissenschaft 42,561-564. Piergiovanni, L., de Noni, I., Fava, P., and Schiraldi, A. (1989). Nonenzymatic browning in processed cheeses. Kinetics of the Maillard reaction during processing and storage. Ital. J. Food Sci. 1, 11-20. Polonelli, L., Morace, G., Monarche, F. D., and Samson, R. A. (1978). Studies on the PR toxin on Penicillium roqueforti. Mycopathologia 66, 99-104. Posati, L. P., and Orr, M. L., eds., (1976). “Composition of Foods, Dairy and Egg Products; Raw, Processed, Prepared.” Consumer and Food Economics Institute, Agricultural Research Service, US. Department of Agriculture, Washington, DC. Prasad, R., Malik, R. K., and Mathur, D. K. (1986). Purification and characterization of ewtracellular caseinolytic enzyme of Micrococcus sp. MCC-315 isolated from Cheddar cheese. J. Dairy Sci. 69,633-642. Prelusky, D. B., Trenholm, H. L., Lawrence, G. A., and Scott, P. M. (1984). Non-transmission of deoxynivalenol (vomitoxin) to milk following oral administration to dairy cows. J. Environ. Sci. Health, Part B B19,593-609. Prentice, J. H. (1987). Cheese rheology. In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), Vol. 1, pp. 299-344. Elsevier. London. Prentice, J. H., Langley, K. R., and Marshall, R. J. (1993). Cheese rheology. In “Cheese: Chemistry, Physics and Microbiology”, (P. F. Fox, ed.), 2nd ed., Vol. 2, pp. 303-340. Chapman & Hall, London. Price, W. V., and Bush, M. G . (1974). The process cheese industry: A review. 11. Research and development. J. Milk Food Technol. 37,179-198. Pritchard, G. G., Freebaim, A. D., and Coolbear, T. (1994). Purification and characterization of an endopeptidase from Lactococcus lactis subsp. cremoris SKll. Microbiology 140, 923-930.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
319
Puhan, Z. (1992). Standardization of milk protein content by membrane processes for product manufacture. In “New Applications of Membrane Processes,” Spec. Issue No. 9201, pp. 23-32. Int. Dairy Fed., Brussels. Purchase, I. F. H., Steyn, M., Rinsma, R., and Tustin, R. C. (1972). Reduction of aflatoxin M content of milk by processing. Food Cosmet. Toxicol. 10,383-387. Qvist, B., Thomsen, D., and Hoier, E. (1987). Effect of ultrafiltered milk and use of different starters on the manufacture, formation and ripening of Havarti cheese. J. Dairy Res. 54,437-446. Raadsveld, C. W. (1957). The course of lactose break-down in Dutch cheese. Meth. Milk Dairy J. 11,313-328. Raap, J., Kerling, K. E. T., Vreeman, H. E., and Visser, S. (1983). Peptide-substrates for chymosin (rennin)-conformational studies of K-casein and some K-casein-related oligopeptides by circular-dichroism and secondary structure prediction. Arch. Biochem. Biophys. 221,117-124. Rank, T. C., Grappin, R., and Olson, N. F. (1985). Secondary proteolysis of cheese during ripening: A review. J. Dairy Sci. 68,801-805. Rattray, F. P., Fox, P. F., and Bockelmann, W. (1995). Purification and characterization of an extracellular proteinase from Brevibacterium linens ATCC 9174. Appl. Environ. Microbiol. 61,3454-3456. Rayan, A. A., Kalab, M., and Ernstrom, C. A. (1980). Microstructure and rheology of process cheese. Scanning Electron Micros. 3, 635-643. Raymond, M. N., and Bricas, E. (1979). New chromophoric peptide substrates. J. Dairy Sci. 62,1719-1725. Raymond, M. N., Gamier, J., Bricas. E.. Cilianu, S., Blasnic, M., Chaix, A., and Lefrancier, P. (1972). Studies on the specificity of chymosin (rennin). I. Kinetic parameters of the hydrolysis of synthetic oligopeptide substrates. Biochimie 54, 145-154. Raymond, M. N., Bricas, E., Salesse, R. Gamier, J., Garnot, P., and Ribadeau Dumas, B. (1973). A proteolytic unit for chymosin (rennin) activity based on a reference synthetic peptide. J . Dairy Sci. 56,419-422. Recker, R. R., Bammi, A., Barger-Lux, M. J., and Heaney, R. P. (1988). Calcium absorbability from milk products, an imitation milk and calcium carbonate. Am. J. Clin. Nurr. 47,93-95. Reid, J. R., Moore, G. G., Midwinter, G. G., and Pritchard, G. G. (1991a). Action of a ceU wall proteinase from Lactococcus lactis subsp. cremoris S K l l on bovine crSl-casein.Appl. Microbiol. Biotechnol. 35, 222-221. Reid, J. R., Ng, K. H., Moore, C. H., Coolbear, T., and Pritchard, G. G. (1991b). Comparison of bovine p-casein hydrolysis by PI and PI1,-typeproteinase from Lactobacillus [sic] lactis subsp. cremoris. Appl. Microbiol. Biotechnol. 36, 344-351. Reid, J. R., Coolbear, T., Pillidge, C. J., and Pritchard, G. G. (1994). Specificity of hydrolysis of bovine K-casein by cell envelope-associated proteinases from Lactococcus lactis strains. Appl. Environ. Microbiol. 60, 801-806. Reimerdes, E. H. (1983). New aspects of naturally occurring proteases in bovine milk. J. Dairy Sci. 66, 1591-1600. Reiter, B., Fryer, T. F., and Sharpe, M. E. (1966). Studies on cheese flavour. J. Appl. Bacteriol. 29,231-243. Reiter, B., Fryer, T. F., Pickering, A., Chapman, H. R., Lawrence, R. C., and Sharpe, M. E. (1967). The effect of the microbial flora on the flavour and free fatty acid composition of Cheddar cheese. J. Dairy Res. 34,257-272. Reiter, B., Sorokin, Y.,Pickering, A., and Hall, A. J. (1969). Hydrolysis of fat and protein in small cheeses made under aseptic conditions. J. Dairy Res. 36, 65-76.
320
P. F. FOX et al.
Renner, E. (1987).Nutritional aspects of cheese. In “Cheese: Chemistry, Physics and Microbiology” (P.F. Fox, ed.), Vol. 1, pp. 345-363. Elsevier, London. Resmini, P., Pellegrino, L., Hogenboom, J., and Bertuccioli, M (1988). “GIi ammioacidi liberi nel formaggio Parmigiano-Reggiano stagionato Ricerca triennale scilla composizione e su alcune peculiari caratteristiche de formaggio Parmigiano-Reggiano,” pp. 41-57. Consorzio de Fromaggio Parmigiano-Reggiano, Reggio Emilia. Reynolds, E. C., and Black, C. L. (1987). Reduction of chocolate’s cariogenicity by supplementation with sodium caseinate. Caries Res. 21, 445-451. Reynolds, E. C., and del Rio, A. (1984). Effect of casein and whey protein solutions on caries experience and feeding patterns of the rat. Arch. Oral. Biol. 29, 927-933. Reynolds, E. C., and Johnson, I. H. (1981). Effect of milk on caries incidence and bacterial composition of dental plaque in the rat. Arch. Oral Biol. 26,445-451. Rice, S . L., Eitenmiller, R. R., and Koehler, P. E. (1976). Biologically active amines in food A review. J. Milk Food Technol. 39, 353-358. Richard, J. L., and Arp, L. H. (1979). Natural occurrence of the mycotoxin penitrem A in moldy cream cheese. Mycopathologia 67, 108-109. Richardson, B. C., and Pearce, K. N. (1981). The determination of plasmin in dairy products. N.Z. J. Dairy Sci. Technol. 16,209-220. Robinson, T. S., Mirocha, C. J., Kurtz, H. J., Behrens, J. C., Chi, M. S., Weaver, G., and Nystrom, S. D. (1979). Transmission of T-2 toxin into bovine and porcine milk. J. Dairy Sci. 62, 637-641. Rodricks, J. V., and Stoloff, L. (1977). Aflatoxin residues from contaminated feed in edible tissues of food producing animals. I n “Mycotoxins in Human and Animal Health” (J. V. Rodricks, C. W. Hesseltine, and M. A. Mehlman, eds.), pp. 67-79. Pathotox Publ. Park Forest South, IL. Roefs, S. P. F. M. (1986). Structure of acid casein gels, a study of gels formed after acidification in the cold. Ph.D. Dissertation, The Agricultural University, Wageningen, The Netherlands. Roefs, S. P. F. M., Walstra, P., Dalgleish, D. G., and Horne, D. S. (1985). Preliminary note on the change in casein micelles by acidification. Neth. Milk Dairy J. 39, 119-122. Roesler, H. (1966). Verhalten der Polyphosphate in Schmelzkaese. Milchwissenschafi 21, 104-107. Rollema, H. S., and Brinkhuis, J. A. (1989). A H-NMR study of bovine micelles; influence of pH, temperature and calcium ions on micellar structure. J . Dairy Res. 56, 417-425. Rosen, S., Min, D. B., Harper, D. S., Harper, W. J., Beck, E. X.,and Beck, F. M. (1984). Effect of cheese, with or without sucrose, on dental caries and recovery of Streptococcus mutans in rats. J. Dent. Res. 63, 894-896. Rothe, G. A. L., Harboe, M. K., and Martiny, S. C. (1977). Quantification of milk-clotting enzymes in 40 commercial bovine rennents, comparing rocket immunoelectrophoresis with an activity ratio assay. J. Dairy Res. 4, 73-77. Roudot-Algaron, F., Le Bars, D., Einhorn, J., Adda, J., and Gripon, J.-C. (1993). Flavor constituents of aqueous fraction extracted from Comte cheese by liquid carbon dioxide. J. Food Sci. 58, 1005-1009. Rugg-Gum, A. J., Edgar, W. M., Geddes, D. A. M., and Jenkins, G. N. (1975). The effect of different meal patterns upon plaque pH in human subjects. Br. DenL J. 139,351-356. Rul, F., Monnet, V., and Gripon, J. C. (1994). Purification and characterization of a general aminopeptidase (St-Pep N) from Streptococcus salivarius ssp. thermophilus CNRZ 301. J. Dairy Sci. 77,2880-2889. Sahlstrom, S., Espinoza, C., Langsrud, T., and S~rhaug,T. (1989). Cell wall, membrane, and intracellular peptidase activities of Propionibacterium shermanii. J. Dairy Sci. 72,342-350.
CHEESE PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
321
Sahlstrom, S., Chizanowska, J. and SBrhaug, T. (1993). Purification and characterization of a cell wall peptidase from Lactococcus lactis subsp. cremoris IMN-C12. Appl. Environ. Microbiol. 59, 3076-3082. Salesse, R., and Gamier, J. (1976). Synthetic peptides for chymosin and pepsin assays: pH effect and pepsin independent determination in mixtures. J. Dairy Sci. 59, 1215-1221. Salih, M. A,, and Sandine, W. E. (1980). Lactic streptococcal agglutinins: A review. J. Food Prof. 43, 856-858. Sander, B. D., Smith, D. E., and Addis, P. B. (1988). Effects of processing stage and storage conditions on cholesterol oxidation products in butter and Cheddar cheese. J . Dairy Sci. 71,3173-3178. Sardinas, J. L. (1972). Microbial rennets. Adv. Appl. Microbiol. 15,39-73. Savello, P. A., Ernstrom, C. A., and Kalab, M. (1989). Microstructure and meltability of model process cheese made with rennet and acid casein. J. Dairy Sci. 72,l-11. Scarpellino, R., and Kosikowski, F. V. (1962). Evolution of volatile compounds in ripening raw and pasteurized milk Cheddar cheese observed by gas chromatography. J. Dairy Sci. 45,343-348. Scharf, L. G., Jr. (1971). The use of phosphates in cheese processing. In “Phosphates in Food Processing” (J. M. Deman and P. Melnychyn, eds.), pp. 120-157. Avi Publ. Co., Westport, CT. Scharf, L. G., Jr., and Kichline, T. P. (1969). Properties and chemical characterization of a ‘bloom’ on process cheese slices. Food Technol. 23,835-837. Schattenkerk, C., and Kerling, K. E. T. (1973). Relation between structure and capacity to function as rennin substrate. Neth. Milk Dairy J. 27, 286-287. Schellhaass, S. M., and Morris, H. A. (1985). Rheological and scanning electron microscopic examination of skim milk gels obtained by fermenting with ropy and non-ropy strains of lactic acid bacteria. Food Microstruct. 4, 279-287. Schmidt, D. G., and Poll, J. K. (1986). Electrophoretic measurements of unheated and heated casein micelle systems. Neth. Milk Dairy J. 40,269-280. Schoch, U. (1981). Mycotoxins in mould ripened cheese-a review. Mitt. Geb. Lebensm. Hyg. 72,80-395. Schoch, U., Luthy, J., and Schlatter, C. (1983). Mycotoxins in mould ripened cheese. Mitt. Geb. Lebensmittelunters. Hyg. 74, 50-59. Schoch, U., Luthy, J., and Schlatter, C. (1984). Mutagenicity testing of commercial P. camernberti and P. roqueforti strains. Z. Lebensm.-Unrers. Forsch. 178, 351-355. Schulz, M. E. (1976). Preparation of melt-resistant processed cheese. U.S. Pat. 3,962,483. Scott, P. M. (1989). Mycotoxigenic fungal contaminants of dairy products. In “Mycotoxins in Dairy Products” (H. P. van Egmond, ed.), pp. 193-259. Elsevier, London. Scott, P. M., and Kennedy, B. P. C. (1976). Analysis of blue cheese for roquefortine and other alkaloids from Penicillium roqueforti. J . Agric. Food Chem. 24, 365-368. Scott, R. (1986). “Cheesemaking Practice,” 2nd ed. Elsevier, London. Severn, D. J., Johnson, M. E., and Olson, N. F. (1986). Determination of lactic acid in Cheddar cheese and calcium lactate crystals. J. Dairy Sci. 69, 2027-2300. Shalabi, S. I., and Fox, P. F. (1987).Electrophoretic analysis of cheese: Comparison of methods. Zr. J. Food Sci. Technol. 11, 135-151. Shaw, J. H., Ensfield, B. J., and Wollman, D. H. (1959). Studies on the relation of dairy products to dental caries in caries-susceptible rats. J. Nutr. 67, 253-273. Shimp, L. A. (1985). Process cheese principles. Food Technol. 39(5), 63-69. Shreeve, B. J., Patterson, D. S. P., and Roberts, B. A. (1979). The carry-over of alflatoxin, ochratoxin and zearalenone from naturally contaminated feed to tissues, urine and milk of dairy cows. Food Comet. Toxicol. 17, 151-152.
322
P. F. FOX ef al.
Sidebottom, C. M., Charton,E., Dunn, P. P. J., Mycock, G., Davies, C., Sutton, J. L., MacCrae, A. R., and Slabas, A. R. (1991). Geotrichurn candidurn produces several lipases with markedly different substrate specificities. Eur. J . Eiochern. 202, 485-491. Silva, M. F. de A., Jenkins, G . N., Burgess, R. C., and Sandham, H. J. (1986). Effect of cheese on experimental caries in human subjects. Caries Res. 20, 263-269. Silva, M. F. de A., Burgess, R. C., Sandham, H. J., and Jenkins, G. N. (1987). Effects of water soluble components of cheese on experimental caries in humans. J . Dent. Res. 66,38-41. Singh, A., Srinivasan, R. A,, and Dudani, A. T. (1973). Studies on exocellular and endocellular lipases of some of the lipolytic bacteria. Milchwissenschafr 28, 164-166. Singh, .I. B., Ranganathan, B., and Chander, H. (1981). Fatty acids liberated by mutants of lactic acid bacteria from milk fat. Milchwissenschafr 36, 742-743. Singh, T. K., Fox, P. F., Hojrup, P., and Healy, A. (1994). A scheme for the fractionation of cheese nitrogen and identification of principal peptides. Int. Dairy J. 4, 111-122. Singh, T. K., Fox, P. F., and Healy, A. (1995). Water soluble peptides in Cheddar cheese: Isolation and identification of peptides in the UF retentate of water-soluble fraction. J. Dairy Res. 62, 629-640. Sinha, R. N., and Sinha, P. R. (1986). Incidence of clostridia with reference to dairy processing. 1. Food Sci. Technol. India 23(2), 103-105. Sinha, R. N., and Sinha, P. R. (1988). Volatile and non-volatile acids produced by Clostridiurn sporogenes isolated from processed cheese. J. Food Sci. Technol. 25, 101-102. Skeie, S. (1994). Developments in microencapsulation science applicable to cheese research and development, a review. Int. Dairy J. 4, 573-595. Smith, D. A., Clark, K. M., and Firary, M. F. (1992). Method of making a non-fat Cream cheese product. U.S. Pat. 5,108,773. Smith, T. A. (1981). Amines in food. Food Chern. 6, 169-200. Snoeren, T. H. M., Klok, H. J., van Hooydonk, A. C. M., and Damman, A. J. (1984). The voluminosity of casein micelles. Milchwissenschafr 39, 461 -463. Sohal, T. S., Roehl, D., and Jelen, P. (1988). Rennet as a cause of bitterness development in Quarg. J . Dairy Sci. 71,3188-3196. Sood, M. S., Gaind, D. K., and Dewan, R. K. (1979). Correlation between micelle solvation and calcium content. N.Z. J. Dairy Sci. Technol. 14, 32-34. Sood, V. K., and Kosikowski, F. V. (1979). Process Cheddar cheese from plain and enzyme treated retentates. J. Dairy Sci. 62, 1713-1718. Serhaug, T. (1981). Comparison of peptide hydrolases from six strains of Brevibacteriurn linens. Milchwissenschafi 36, 137-139. Serhaug, T., and Ordal, Z. J. (1974). Cell-bound lipase and esterase of Brevibacteriurn linens. Appl. Microbiol. 27, 607-608. Spangler, P. L., Jensen, L. A., Amundson, C. H., Olson, N. F.. and Hill, G. G., Jr. (1991). Ultrafiltered Gouda cheese: effects of preacidification, diafiltration, rennet and starter concentration and time to cut. J. Dairy Sci. 74,2809-2819. Stadhouders, J. (1974). Dairy starter cultures. Milchwissenschafr 29, 329-337. Stadhouders, J., and Hup, G. (1975). Factors affecting bitter flavour in Gouda cheese. Neth. Milk Dairy J. 29, 335-353. Stadhouders, J., and Veringa, H. A. (1967). Texture and flavour defects in cheese caused by bacteria from contaminated rennet. Neth. Milk Dairy J. 21, 192-207. Stadhouders, J., and Veringa, H. A. (1973). Fat hydrolysis by lactic acid bacteria in cheese. Neth. Milk Dairy J. 27, 77-91. Stadhouders, J., Hup, G., Exterkate, F. A., and Visser, S. (1983). Bitter flavour in cheese. 1. Mechanism of the formation of the bitter flavour defect in cheese. Neth. Milk Dairy J. 3’7,157-167.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
323
Stanier, R. Y., Adelberg, E. A., and Ingraham, J. L. (1981). “General Microbiology,” 4th ed., pp. 27-35. Macmillan, London. Steering Group on Food Surveillance (1987). “Mycotoxins.” H. M. Stationery Office, London. Steffen, C., Eberhard, P., Bosset, J. O., and Ruegg, M. (1993). Swiss-type varieties. I n “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), 2nd ed., Vol. 2, pp. 83-110. Chapman & Hall, London. Stepaniak, L., and Fox, P. F. (1995). Characterization of the principal intracellular endopeptidase from Lactococcus lactis ssp. lactis MG 1363. Int. Dairy J. 5, 699-713. Strange, E. D., Malin, E. L., van Hekken, D. L., and Basch, J. J. (1992). Chromatographic and electrophoretic methods used for analysis of milk proteins. J. Chrornatogr. 624,81-102. Stubblefield, R. D., and Shannon, G . M. (1974). Aflatoxin MI: Analysis in dairy products and distribution in dairy foods made from artifically contaminated milk. J. Assoc. Off Anal. Chern. 57,847-851. Surgeon General (1988). “The Surgeon General’s Report on Nutrition and Health,” U.S. Dept. of Health and Human Services Publ. No. 88-50210. U.S. Govt. Printing Office, Washington, DC. Sweet, B. J., Matthews, L. C., and Richardson, T. (1984). Purification and characterization of pregastric esterase from calf. Arch. Biochem. Bzophys. 234, 144-150. Swiatek, A. (1964). Einfluss der Art und Menge des Schmelzsalzes auf die Konsistenz von Schmelzkaese. Milchwissenschaft 19,409-413. Tamime, A. Y., and Kirkegaard, J. (1991). Manufacture of Feta cheese-industrial. In “Feta and Related Cheeses” (P. K. Robinson and A. Y. Tamime, eds.), pp. 70-143. Ellis Honvood, New York. Tamime, A. Y., Kalab, M., and Davies, G. (1984). Microstructure of set-style yoghurt manufactured from cow’s milk fortified by various methods. Food Microstruct. 3, 83-92. Tamime, A. Y., Davies, G., Chehade, A. S., and Mahdi, H. A. (1989). The production of “Labneh” by ultrafiltration: a new technology. J. SOC.Dairy Technol. 42, 35-39. Tamime, A. Y., Kalab, M., Davies, G., and Younis, M. F. (1990). Microstructure and firmness of processed cheese manufactured from Cheddar cheese and skim milk powder cheese base. Food Microstruct. 9, 23-37. Tan, P. S. T., and Konings, W. N. (1 990). Purification and characterization of an aminopeptidase from Lactococcus lactis subsp. crernoris Wg 2. Appl. Environ. Microbiol. 56, 526-532. Tan, P. S. T., Pos, K. M., and Konings, W. N. (1991). Purification and characterization of an endopeptidase from Lactococcus lactis subsp. cremoris Wg 2. Appl. Environ. Microbiol. 57,3593-3599. Tan, P. S. T.. Chapot-Chartier, M. P., Pos, K. M., Rosseau, M., Boquien, C. Y., Gripon. J.-C., and Konings, W. N. (1992). Localization of peptidases in lactococci. Appl. Environ. Microbiol. 58, 285-290. Tan, P. S. T., Poolman, B., and Konings, W. N. (1993a). Proteolytic enzymes of Lactococcus lactis. J. Dairy Res. 60,269-286. Tan, P. S. T., van Kessel, T. A. J. M., van de Veerdonk, F. L. M., Zuurendonk, P. F., Bruins, A. P., and Konings, W. N. (1993b). Degradation and debittering of a tryptic digest from p-casein by aminopeptidase N from Lactococcus lactis subsp. cremoris Wg2. Appl. Environ. Microbiol. 59, 1430-1436. Tanaka, N., Goepfert, J. M., Traisman, E., and Hoffbeck, W. M. (1979). A challenge of pasteurized process cheese with Clostridiurn botulinurn spores. J. Food Prot. 42,787-789. Tanaka, N., Traisman, E., Plantinga, P., Finn, L., Flom, W., Meske, L., and Guggisberg, J. (1986). Evaluation of factors involved in antibotulinal properties of pasteurized process cheese spreads. J. Food Prot. 49,526-53 I.
324
P. F. FOX et al.
Taneya, S., Izutsu, T., and Sone, T. (1979). Dynamic viscoelasticity of natural cheese and processed cheese. In “Food Texture and Rheology” (P. Sherman, ed.), pp. 369-383. Academic Press, London. Taneya, S., Kimura, T., Izutsu, T., and Buchheim, W. (1980). The submicroscopic structure of processed cheese with different melting properties. Milchwissenschaft 35,479-481. Tatsumi, K.. Ohba, S., Nakajima, I., Shinohara, K., and Kawanishi, G. (1975). The effects of melting salts on the state of dispersion of casein. 3. The effect of melting salts on the texture of process cheese. J. Agric. Chem. SOC.Jpn. 49,481-489. Tatsumi, K., Nishiya, T., Ido, K., and Kawanishi, G. (1991). Effects of heat treatment on the meltability of processed cheese. J. Jpn. SOC.Food Sci. Technol. 38,102-106. Taylor, S . L. (1986). Histamine food poisoning: Toxicology and clinical aspects. CRC Crit. Rev. Toxicol. 17, 91-128. Taylor, S. L., Kiefe, T. J., Windham, E. S., and Howell, J. F. (1982). Outbreak of histamine poisoning associated with consumption of Swiss cheese. J. Food Prot. 45, 455-457. Templeton, H. L., and Sommer, H. H. (1936). Studies on the emulsifyingsaltsused in processed cheese. J. Dairy Sci. 19, 561-572. Teuber, M. (1990). “Production of Chymosin (EC 3.4.23.4) by Microorganisms and its Use for Cheesemaking,” Bull.No. 251, pp. 3-15. Int. Dairy Fed., Brussels. Thomas, M. A. (1969). Browning reaction in Cheddar cheese. Aust. J. Dairy Technol. 22, 185-189. Thomas, M. A. (1970). Use of calcium co-precipitates in processed cheese. Aust. J. Dairy Technol. 23,23-26. Thomas, M. A. (1977). “The Processed Cheese Industry,” p. 93. Department of Agriculture, Sydney, New South Wales, Australia. Thomas, M. A., Newell, G., Abad, G. A., and Turner, A. D. (1980). Effect of emulsifying salts on objective and subjective properties of processed cheese. J. Food Sci. 45,458-466. Thomas, T. D. (1987). Acetate production from lactate and citrate by non-starter bacteria in Cheddar cheese. N.2.J. Dairy Sci. Technol. 22,25-38. Thomas, T. D., and Crow, V. L. (1983). Mechanism of D(-)-lactic acid formation in Cheddar cheese. N.Z. J. Dairy Sci. Technol. 18, 131-141. Thomas T. D., and Pearce, K. (1981). Influence of salt on lactose fermentation and proteolysis in Cheddar cheese. N.Z. J. Dairy Sci. Technol. 16,253-259. Thomas, T. D., Ellwood, D. C., and Longyear,M. C. (1979). Change from homo- to heterolactic fermentation by Streptococcus lactis resulting from glucoselimitation in anaerobicchemostat cultures. J. Bacteriol. 138, 109-117. Thomas, T. D., McKay, L. L., and Morris, H. A. (1985). Lactate metabolism by pediococci isolated from cheese. Appl. Environ. Microbiol. 49, 908-913. Tinson, W., Radcliff, M. F., Hillier, A. J., and Jago, G. R. (1982). Metabolism of Streprococcus thermophilus. 3. Influence on the level of bacterial matabolites in Cheddar cheese. Aust. J. Dairy Technol. 37, 17-21. Torres, N., and Chandan, R. C. (1981a). Latin American white cheese-A review. J. Dairy Sci. 64,552-557. Torres, N., and Chandan, R. C. (1981b). Flavor and texture development in Latin American white cheese. J. Dairy Sci. 64, 2161-2169. Trieu-Cuot, P., and Gripon, J.-C. (1981). Casein hydrolysis by Penicillium caseicolum and P. roqueforri: a study with isoelectric focusing and two dimensional electrophoresis.Neth. Milk Dairy J. 35, 353-357. Trieu-Cuot, P., and Gripon, J.-C. (1982). A study of proteolysis during Camembert cheese ripening using isoelectric focusing and two-dimensional electrophoresis. J. Dairy Res. 49.501-510.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
325
Trieu-Cuot. P.. Archieri-Haze, M-J., and Gripon. J. C. (1982). Etude comparative de l’action des mCtalloprotCases de Penicillium caseicolum et Penicillium roqueforti sur les casCines alphasl et beta. Lait 62, 234-249. Tsakalidou, E., Dalezios, I., Georgalaki, M., and Kalantzopoulos, G. (1993). A comparative study: Aminopeptidase activities from Lactobacillus delbrueckii ssp. bulgaricus and Streptococcus thermophilus. J. Dairy Sci. 76, 2145-2151. Tsuda, T., Yamada, M.. and Nakazawa, Y. (1993). Measurement of lower molecular weight peptides in Camembert cheese using a computer simulation system of capillary isotachophoresis. Milchwissenschaft 48,74-78. Tunick, M. H., Nolan, E. J., Shick, J. J., Basch, J. J., Thompson, M. P., Maleeff, B. E., and Holsinger, V. H. (1990). Cheddar and Cheshire cheese rheology. J . Dairy Sci. 73, 16711675. Turner, K. W., and Thomas, T. D. (1980). Lactose fermentation in Chedder cheese and the effect of salt. N.Z. J. Dairy Sci. Technol. 15, 265-276. Turner, K. W., Morris, H. A., and Martley, F. G. (1983). Swiss-type cheese. 11. The role of thermophilic lactobacilli in sugar fermentation. N.Z. J. Dairy Sci. Technol. 18, 117-124. Turner, K. W., Lawrence, R. C., and Lelievre, J. (1986). A microbiological specification for milk for aseptic cheesemaking. N.Z. J . Dairy Sci. Technol. 21, 249-254. Tzanetakis, N., and Litopoulou-Tzanetaki, E. (1989). Biochemical activities of Pediococcus penfosaceus isolates of dairy origin. J. Dairy Sci. 72, 859-863. Ueno, Y . (1985). The toxicology of mycotoxins. CRC Crit. Rev. Tuxicol. 14,99-132. Uhlmann, G., Klostermeyer, H., and Merkenich, K. (1983). Kristallisationerschienungen in Schmelzkaeseprodukten. 1. Phaenomen und Ursachen. Milchwissenschafr 38,582-585. Umemoto, Y.. Umeda, H., and Sato, Y. (1968). Studies on lipolysis of dairy lactic acid bacteria. 11. On the lipolytic activities of cell-free extracts of lactic acid bacteria. Agric. Biol.Chem. 32,131 1-1317. Urbach, G. (1993). Relations between cheese flavour and chemical composition, Inf. Dairy J. 3, 389-422. van Boven, A,, Tan, P. S. T., and Konings. W. N. (1988). Purification and characterization of a dipeptidase from Streptococcus cremoris Wg2. Appl. Environ. Microbiol. 54,43-49. van den Berg, G . ,and de Koning, P. J. (1990). Gouda cheesemaking with purified calf chymosin and microbiologically produced chymosin. Neth. Milk Dairy J. 44, 189-205. van den Bijgaart, H. J. C. M. (1988). Syneresis of rennet-induced milk gels as influenced by cheesemaking parameters. Ph.D. Thesis, The Agricultural University, Wageningen, The Netherlands. van Egmond. H. P. (1989). Aflatoxin MI: Occurrence, toxicity, regulation. In “Mycotoxins in Dairy Products” (H. P. van Egmond, ed.), pp. 11-55. Elsevier, London. van Egmond. H. P., Paulsch. W. E., Veringa, H. P., and Schuller. P. L. (1977). The effect of processing on the aflatoxin M, content of milk and milk products. Arch. Inst. Pasteur 3-4,381-390. van Hooydonk, A. C. M., Hagedoorn, H. G., and Boerrigter, I. J. (1986a). pH-induced physicochemical changes of casein micelles in milk and their effect on renneting. 1. Effect of acidification on physico-chemical properties. Neth. Milk Dairy J. 40, 281-296. van Hooydonk. A. C . M., Boerrigter. I. J., and Hagendoorn, H. G. (1986b). pH-induced physico-chemical changes of casein micelles in milk and their effect on renneting. 2. Effect of pH on renneting of milk Neth. Milk Dairy J . 40,297-313. van Vliet, T.. and Dentener-Kikkert. A. (1982). Influence of the composition of the milk fat globule membrane on the properties of acid milk gels. Neth. Milk Dairy J. 36, 261-265. van Vliet, T., and Walstra, P. (1985). Note on the shear modulus of rennet-induced gels. Neth. Milk Dairy J. 39, 115-118.
326
P. F. FOX et al.
van Wazer, J. R. (1971).Chemistry of the phosphates and condensed phosphates. In “Phosphates in Food Processing” (J. M. Deman and P. Melnychyn, eds.), pp. 1-23. Avi Publ. Co., Westport, CT. Verdi, R. J., and Barbano, D. M. (1991).Properties of proteases from milk somatic cells and blood leukocytes. J. Dairy Sci. 74,2077-2081. Visser, F. M. W. (1976). Method for the manufacture of rennet-free cheese, Neth. Milk Dairy J. 30,41-54. Visser, F. M.W.(1977a). Contribution of enzymes from rennet, starter bacteria and milk to proteolysis and flavour development in Gouda cheese. 1. Description of cheese and aseptic cheesemaking techniques. Neth. Milk Dairy J. 31, 120-133. Visser, F. M. W. (1977b).Contribution of enzymes from rennet, starter bacteria and milk to proteolysis and flavour development in Gouda cheese. 2. Development of bitterness and cheese flavour. Neth. Milk Dairy J. 31, 188-209. Visser, F.M. W. (1977~). Contribution of enzymes from rennet, starter bacteria and milk to proteolysis and flavour development in Gouda cheese. 3. Protein breakdown: analysis of the soluble nitrogen and amino acid fractions, Neth. Milk Dairy J. 31,210-239. Visser. F. M. W., and de Groot-Mostert, A. E. A. (1977). Contribution of enzymes from rennet, starter bacteria and milk to proteolysis and flavour development in Gouda cheese. 4. Protein breakdown: A gel electrophoretical study. Neth. Milk Dairy J. 31, 247-264. Visser, J., Minihan, A., Smits, P., Tjan, S. B., and Heertje, I. (1986). Effects of pH and temperature on the milk salt system. Neth. Milk Dairy J. 40,351-368. Visser, S. (1993).Proteolytic enzymes and their relation to cheese ripening and flavor: An overview. J. Dairy Sci. 76, 329-350. Visser, S., and Rollema, H.S. (1986). Quantification of chymosin action of nonlabeled Kcasein-related peptide substrates by ultraviolet spectrophotometry: Description of kinetics by the analysis of progress curves. Anal. Biochem. 153,235-241. Visser, S., and Slangen, K. J. (1977). On the specificity of chymosin (rennin) in its action on bovine P-casein. Neth. Milk Dairy J. 31, 16-30. Visser, S.,van Rooijen, P. J. Schattenkerk, C., and Kerling, K. E. T. (1976).Peptide substrates for chymosin (rennin). Kinetic studies with peptides of different chain length including parts of the sequence 101-112 of bovine K-casein. Biochim. Biophys. Acta 438,265-272. Visser, S., van Rooijen, P. J. Schattenkerk, C., and Kerling, K. E. T.(1977).Peptide substrates for chymosin (rennin). Kinetic studies with bovine ~-casein-(103-108)-hexapeptideanalogues. Biochim. Biophys. Acta 481,171-176. Visser, S., van Rooijen, P. J., and Slangen, C. J. (1980). Peptide substrates for chymosin (rennin). Isolation and substrate behaviour of two tryptic fragments of bovine K-casein. Eur. J. Biochem. 108,415-421. Visser, S . , Hup, G., Exterkate, F. A., and Stadhouders, J. (1983).Bitter flavour in cheese. 2. Model studies on the formation and degradation of bitter peptides by proteolytic enzymes from calf rennet, starter cells and starter cell fractions. Neth. Milk Dairy J. 37, 169-180. Visser, S.,Exterkate, F. A., Slangen, C. J., and de Veer, G. J. C. M. (1986). Comparative study of action of cell wall proteinases for various strains of Streptococcus cremoris on bovine cxSl-, p- and K-casein. Appl. Environ. Microbiol. 52, 1162-1166. Visser, S., Slangen, C . J., and van Rooijen, P. J. (1987). Peptide substrates for chymosin (rennin). Interaction sites in K-CaSein-related sequences located outside the (103-108) hexapeptide region that fits into the enzyme’s active-site cleft. Biochem. J. 244,553-558. Visser, S., Slangen, K. J., Exterkate, F. A., and de Veer, G. J. C. M. (1988).Action of a cell wall proteinase (PI) from Srreptococcus cremoris HP on bovine p-casein. Appl. Microbiol. Biotechnol. 29, 61-66.
CHEESE: PHYSICAL, BIOCHEMICAL, AND NUTRITIONAL ASPECTS
327
Visser, S., Slangen, K. J., Alting, A. C., and Vreeman, H. J. (1989). Specificity of bovine plasmin in its action on bovine aS2-casein. Milchwissenschafr 44,335-339. Visser, S., Robben, A. J. P. M., and Slangen, C. J. (1991). Specificity of a cell-envelope-located proteinase (PIII-type)from Lactococcus lactis subsp. cremoris AM1 in its action on bovine p-casein. Appl. Microbiol. Biotechnol. 35,477-483. Visser, S., Slangen, C. J., Robben, A. J. P. M., van Dongen, W. D., Heerma, W., and Haverkamp, J. (1994). Action of a cell-envelope proteinase (CEPIII-type) from Lactococcus lactis subsp. cremoris AM1 on bovine K-casein. Appl. Microbiol. Biotechnol. 41, 644-651. Voight, M. N., Eitenmiller, R. R., Koehler, P. E., and Hamdy. M. K. (1974). Tyramine, histamine and tryptamine content of cheese. J. Milk Food Technol. 37, 377-381. Vos, P., Simons, G . , Siezen, R. J., and de Vos, W. M. (1989a). Primary structure and organization of the gene for a procaryotyc cell envelope-located serine proteinase. J. Biol. Chem. 264, 13579-13585. Vos, P., van Asseldonka, M., van Jeveren, F., Siezen, R., Simons, G., and de Vos, W. M. (1989b). A maturation protein is essential for production of active forms of Lactococcus lactis S K l l serine proteinase located in or secreted from the cell envelope. J. Bacteriol. 171,2795-2802. Vreeman, H. J., van Markwijk, B. W., and Booth, P. (1989). The structure of casein micelles between 5.5 and 6.7 as determined by light scattering, electron microscopy and voluminosity measurements. J. Dairy Res. 56, 463-470. Vreeman, W. J., Visser, S., Slangen, C. J., and van Reel, J. A. M. (1986). Characterization of bovine K-casein fractions and the kinetics of chymosin-induced macropeptide release from carbohydrate-free and carbohydrate-containing fractions determined by high performance gel premeation chromatography. Biochem. J. 240,87-97. Wagner, K.-H., and Wagner-Hering, E. (1981). Qualitaetsmerkmale des Schmelzkaesespraktische Erfahrungen und Wissenschaftliche Erkenntnisse. Milchwissenschaft 36, 744-747. Walstra, P. (1993). The syneresis of curd. In “Cheese: Chemistry, Physics and Microbiology” (P. F. Fox, ed.), 2nd ed., Vol. 1, pp. 141-191. Chapman & Hall, London. Walstra, P., and Jenness, R. (19M). “Dairy Chemistry and Physics,” pp. 294-297. Wiley, New York. Walstra, P., and van Vliet, T. (1986). The physical chemistry of curd-making. Neth. Milk Dairy J . 40,241-259. Walstra, P., van Dijk, H. J. M., and Geurts, T. J. (1985). The syneresis of curd 1. General considerations and literature review. Neth. Milk Dairy J. 39,209-246. Walstra, P., Luyten, H., and van Vliet, T. (1987). Consistency of cheese. Milk-Vital Force, Proc. Int. Dairy Congr., 22nd, The Hague, 1986, pp. 159-168. Ware, G. M., Thorpe, C . W., and Pohland, A. E. (1980). Determination of roquefortine in blue cheese and blue cheese dressing by high pressure liquid chromatography with ultraviolet and electrochemical detectors. J. Assoc. Off: Anal. Chem. 63, 637-641. Waugh, D. F., and von Hippel, P. H. (1956). K-Casein and the stabilization of casein micelles. J. Am. Chem. SOC.78,4576-4582. Weiss, M. E., and Bibby, B. G. (1966). Effects of milk on enamel solubility. Arch. Oral Biol. 11,49-57. Wilkinson, M. G. (1992). Studies on the acceleration of Cheddar cheese ripening. Ph.D. Thesis, National University of Ireland, Cork. Wilkinson, M. G. (1993). Acceleration of cheese ripening. In “Cheese: Chemistry, Physics and Microbiology”, (P. F. Fox, ed.), 2nd ed., Vol. 1,pp. 523-555. Chapman & Hall, London. Wilkinson, M. G., Guinee,T. P., O’Callaghan, D. M., and Fox, P. F. (1992).Effect of commercial enzymes on proteolysis and ripening in Cheddar cheese. Lait 72,449-459.
328
P. F. FOX et al.
Wilkinson, M. G., Guinee, T. P., O’Callaghan, D. M., and Vox, P. F. (1994). Autolysis and proteolysis in different strains of starter bacteria. J. Dairy Res. 61,249-262. Williams, B. C. (1985). Mycotoxins in foods and foodstuffs. In “Mycotoxins: A Canadian Perspective” (P. M. Scott, H. L. Trenholm, and M. D. Sutton, eds.), pp. 49-53. National Research Council of Canada, Ottawa. Winwood, J. (1983). Quarg production methods-past, present and future. J. SOC. Dairy Technol. 36, 107-109. Wissmeier, H. E. (1972). Grosstechnische Herstellung von Lebensmittelphosphaten in Ladenburg. Milchwissenschafr 27,276-280. Wohlrab, Y., and Bockelmann, W. (1992). Purification and characterization of dipeptidase from Lactobacillus delbrueckii subsp. bulgaricus. Int. Dairy J. 2, 345-361. Wohlrab, Y., and Bockelmann, W. (1993). Purification and characterization of a second aminopeptidase (Pep C-like) from Lactobacillus delbrueckii subsp. bulgaricus B14. Int. Dairy J. 3, 685-701. Wong, N. P., Ellis, R., La Croix, D. E., and Alford, J. A. (1973). Lactones in Cheddar cheese. J. Dairy Sci. 56, 636 (abstr.). Wong, N. P., Ellis, R., and La Croix, D. E. (1975). Quantitative determination of lactones in Cheddar cheese. J. Dairy Sci. 58, 1437-1441. Woo, A. H., and Lindsay, R. C. (1984). Concentration of major free fatty acids and flavor development in Italian cheese varieties. J. Dairy Sci. 67, 960-968. Woo, A. H., Kollodge, S., and Lindsay, R. C. (1984). Quantitation of major free fatty acids in several cheese varieties. J . Dairy Sci 67,874-878. Yan, T.-R., Azuma, N., Kaminogawa, S., and Yamauchi, K. (1987a). Purification and characterization of novel metalloendopeptidase from Streptococcus cremoris H61. Eur. J. Biochem. 163,259-265. Yan,T.-R., Azuma,N., Kaminogawa, S., and Yamauchi, K. (1987b). Purification and characterization of a substrate-size-recognizing metalloendopeptidase from Streptococcus cremoris H61. Appl. Environ. Microbiol. 53, 2296-2302. Younis, M. F., Tamime, A. Y., Davies, G., Hunter, E. A., Dawood, A. H., and Abdou. S. M. (1991). Production of processed cheese using Cheddar cheese and cheese base. 4. Microbiological and organoleptic qualities. Milchwissenschafr 46, 645-648. Yousef, A. E., and Marth, E. H. (1989). Stability and degradation of aflatoxin MI. In “Mycotoxins in Dairy Products” (H. P. van Egmond, ed.), p. 127-161. Elsevier, London. Zall, R. R. (1985). On-farm ultrafiltration. I n “New Dairy Products via New Technology,” Proc. IDF Semin., pp. 9-19. Int. Dairy Fed. Brussels. Zevaco, C., Monnet, V., and Gripon, J.-C. (1990). Intracellular X-prolyldipeptidyl peptidase from Lactococcus lacris subsp. lactis: Purification and properties. J. Appl. Bacteriol. 68, 357-366. Zhang, D., and Mahoney, A. W. (1989a). Effect of iron fortification on quality of Cheddar cheese. J. Dairy Sci. 72, 322-332. Zhang, D., and Mahoney, A. W. (1989b). Bioavailability of iron-milk protein complexes and fortified Cheddar cheese. J. Dairy Sci. 72, 2845-2855. Zhang, D., and Mahoney, A. W. (1990). Effect of iron fortification on Cheddar cheese. J. Dairy Sci. 73, 2252-2258. Zhang, D., and Mahoney, A. W. (1991). Iron fortification of process Cheddar cheese. J. Dairy Sci. 74, 353-358.
ADVANCES IN FOOD AND NUTRITION RESEARCH, VOL. 39
BlOGENlC AMINES IN FISH AND SHELLFISH DAFNE D. RAWLES' AND GEORGE J. FLICK Department of Food Science and Technology Virginia Polytechnic Institute and State University Blacksburg, Virginia 24061
ROY E. MARTIN National Fisheries Institute, Inc. Arlington, Virginia 22209
I. 11. 111. IV. V. VI.
VII. VIII. IX.
X.
Introduction Nonvolatile Amine Formation Amine Detoxification Bacterial Species with Decarboxylase Activity Amines Presence in the Marine Ecosystem Amines Occurrence in Seafood A. Finfish B. Shellfish C. Seafood Products Scombrotoxicosis Amine Formation as an Indicator of Freshness in Seafoods Recommended Limits of Amine Content A. Quality Control Purposes B. Toxic Level Determination of Biogenic Amines in Fish A. Liquid Chromatography B. Thin-Layer Chromatography C. High-pressure Liquid Chromatography D. Gas-Liquid Chromatography E. Enzymic Test F. Enzyme Electrode G. Test Strips
Present address: Perdue Farms, Inc., Bridgewater, VA 22812. 329 Copyright 0 I996 by Academic Press, Inc. All rights of reproduction in any form reserved.
330
DAFNE D. RAWLES etal.
H. Capillary Zone Electrophoresis References
I. INTRODUCTION
The term biogenic amines refers to the nonvolatile amines such as cadaverine, putrescine, spermidine, spermine, tyramine, tryptamine, and histamine produced post mortem in fish and shellfish products. The compounds are reported to originate from the decarboxylation of specific free amino acids in fish or shellfish tissue. Chemically, the biogenic amines are defined as low-molecular-weight aliphatic, alicyclic, or heterocyclic organic bases (Davidek and Davidek, 1995). The use of chemical compounds as objective product standards or indices of quality has long been suggested since these tests are rapid when compared to traditional microbiological analyses and less subject to individual interpretation than sensory analyses. 11.
NONVOLATILE AMINE FORMATION
The decarboxylation process can proceed through two biochemical pathways: decarboxylation through endogenous (naturally occurring) decarboxylase enzymes or by exogenous decomposition through enzymes released by the microflora associated with a seafood product. There is general agreement among researchers that the endogenous production of diamines is insignificant when compared to the exogenous pathway (Wendakoon and Sakaguchi, 1992). Polyamines, present in all biological materials, have been implicated in a wide variety of biological reactions (i.e., DNA, RNA, and protein synthesis) (Tabor and Tabor, 1976; Walters, 1987). They have also been shown to maintain the cell envelope integrity of bacteria (Tabor and Tabor, 1985). Besides, the covalent linking of cadaverine (Kamio et al., 1986) and putrescine (Kamio and Nakamura, 1987) to the peptidoglycan is necessary for normal growth of certain bacteria. Using thin-layer chromatography, Rolle et al. (1971) were able to detect ethanolamine, methylamine, dimethylamine, putrescine, and spermidine in the unicellular green algae Chlorella fusca and Scenedesrnus acutus. I 11.
AM1NE DETOXIFICATION
Histamine is formed by the decarboxylation of the amino acid histidine, which is found in high levels in tissues of scombroid fishes (Frank, 1985).
BIOGENIC AMINES IN FISH AND SHELLFISH
331
Upon ingestion, histamine is detoxified primarily by the enzymes diamine oxidase and histamine N-methyltransferase (Hui and Taylor, 1985). Relatively high doses of pure histamine have been administered to humans with no apparent ill effect (Arnold and Brown, 1978; Douglas, 1980; Ijomah et al., 1992), leading to the suggestion that scombroid poisoning by spoiled fish is caused by histamine acting synergistically with other diamines present in the fish, primarily putrescine and cadaverine (Bjeldanes et al., 1978). Putrescine (1,4-diaminobutane) is the decarboxylation product of the amino acid lysine and cadaverine arises from the decarboxylation of ornithine (Fig. 1). Both putrescine and cadaverine may interfere with the normal histamine detoxification system of the intestine by competing under certain conditions with histamine as substrates for diamine oxidase. Cadaverine and putrescine potentiate peroral toxicity of histamine in the guinea pig (Bjeldanes er al., 1978). Spermine and cadaverine increase histamine transport across the gastrointestinal wall (Jung and Bjeldanes, 1979) with little change in the relative amounts of histamine metabolites produced. Diamine oxidases belong to a widespread class of enzymes related to polyamine metabolism (Hill, 1971; Smith, 1985). Being a key enzyme in the metabolism of polyamines, essential for the growth and replication of all living cells (Bachrach, 1973; Walters, 1987), they catalize the oxidative deamination of a range of primary diamines (Fig. 2) (Equi et al., 1991). Inhibitors of diamine oxidases are known to possess antimalarial, antitrypanosomal, antibacterial, and antifungal activities together with possible roles in cancer chemotherapy (Equi et al., 1991). Clinical studies have considered the activity of diamine oxidases to be a useful parameter (Baylin, 1977; Luk et al., 1980) thus, many different assays to determine the activity of diamine oxidases have been proposed. Procedures using radioactive substrate (Okuyama and Kobayashi, 1961; Kusche ef al., 1973) achieve high sensitivity and most recently the use of highperformance liquid chromatography has been described (Biondi er al., 1984). IV. BACTERIAL SPECIES WITH DECARBOXYLASE ACTIVITY
It is known that different bacteria vary significantly in either the quantity of decarboxylase they produce and/or the specific activity (turnover number) of those decarboxylases (Wendakoon and Sakaguchi, 1992). It is also known that the composition of the fishery product significantly affects the amount of decarboxylase a bacterium may release. Primarily responsible for the decomposition of the scombroid fish, with the ability to decarboxylate histidine to form histamine (Table I), are
332
DAFNE D. RAWLES etal. H N
-
H N
H2N-(CH2)2
Histidine
CO2H
0
CO2H
I H2N -CH-(CH
Histamine
2)2-
IIC-NH
I
-H2N-
CH-(CH 2)4-C-NH2 Lysine
Glutamine
C02H
I HtN--CH+CH2)+-NH2
CO2H
NH2
I
I
H~N-CCH-(CHZ)~-NH-C=NH Arginine
Ornithine
i
r '
H2N-(CH2)4-NH2
NH,
I
H2N-(CH2)4-C=
t
Putrescine
.... .
a)
Agmatine
?.*\
H~N-(CH~)S-NH~
NH Cndaverine
H2N--(CH2)j-N-(CH2)4-NHZ Spermidine
i H2N-(Cli2)3-
ti
I
N - ( C H Z ) ~ - N - ( C H 2 ) r NH2 Spermine
FIG. 1. Biological pathways for the formation of histamine, putrescine, cadaverine, spermidine. and spermine.
the Enterobacteriaceae (Frank et al., 1985; Taylor and Sumner, 1986). Histamine-producing bacteria isolated from fish causing scombroid poisoning are Morganella (Proteus) morganii, Klebsiella pneumoniae, and Hafnia alvei (Taylor and Speckhard, 1983). Pleisomonas shigelloides, a bacterium frequently isolated from fish and aquatic environments, was identified as
BIOGENIC AMINES IN FISH AND SHELLFISH
n=4 (putrescines):
(I): R = H
333
(3): R = C2H5
(2): R = CH3 (4): R = n - C3H7
R1
I I
H~N-CHZ-CH~-C-CH~-CH~-
n=5 (cadaverines):
(5): R 1 = R2 = H (cadaverine)
NH2
R2
(6): R1 = H, R2 = Me (3-methylcadaverine)
(7): R 1 = R2 = M e (3,3 dimethylcadaverine) (8): R1 =OH, R2 = Me (3-hydroxy-3-methylcadaverine)
(9): RI = R2 = Me, R3 = R4 = H (2.2-dimethylcadaverine) (10): R 1 = R 3 = Me, R2 = R4 = H (2,4-dimethylcadaverine)
(1 1): R 1 = Me, R2 = R3 = R4 = H (2-methylcadaverine)
FIG. 2. Deanimation of primary diamines by diamine oxidases (DAO) (based on Equi al., 1991).
et
a new histamine former in fish by Lopez-Sabater et al. (1994b). In addition, bacteria isolated from skipjack tuna (Euthynnus pelamis) and found to produce histamine are Proteus vulgaris, Proteus mirabilis, Clostridium perfringens, Enterobacter aerogenes, and Vibrio alginolyticus (Arnold et al., 1980;Yoshinaga and Frank, 1982;Frank et ai.,1985).Three bacterial species not previously reported to have the potential to produce histamine; Acinetobacter Iwofi, Pseudomonas putrefaciens, and Aeromonas hydrophila, were isolated from decomposing Spanish mackerel (Scomberomorus maculatus) together with 11 other bacterial species (Clostridiumperfringes, Enterobacter aerogenes, Enterobacter sp., Hafnia alvei, Morganella morganii, Proteus
334
DAFNE D. RAWLES etal.
TABLE I MICROBIAL ISOLATES SHOWING HISTIDINE DECARBOXYLASE ACTIVITY"
Identification of isolate
Decomposition temperature(s) at which found ("C)
Acinetobacter lwoffi Aeromonas hydrophila Citrobacter freundii Clostridium perfringens Edwardsiella sp. Enterobacter aerogenes
0, 15 0, 15 15 15,30 15 15,30
Enterobacter sp. Escherichia coli Hafnia alvei Klebsiella pneumoniae Klebsiella sp. Morganella morganii
30 15.30 15,30 15 15 0, 15, 30
Proteus mirabilis Proteus vulgaris Proteus sp.
15 30 15
Pseudomonas fluorescens/putida Pseudomonas putrefaciens Pseudomonas sp. Vibrio sp. Vibrio alginolyticus
15 0, 15, 30 0, 15 15 0, 15, 30
Source Beef Skipjack tuna Skipjack tuna Mahi-mahi Skipjack tuna, tuna, mahi-mahi, pork, beef Food, tuna Tuna Tuna, skipjack tuna, mackerel, food Tuna, mahi-mahi, mackerel Skipjack tuna, mackerel Scombroid fish, pork, turkey, mahi-mahi Pork, turkey, tuna, skipjack tuna Beef, pork, turkey, tuna Fish, feces, tuna, mackerel, food, milk Food Mackerel Skipjack tuna
Activities determined using the carboylase base broth supplemental with histamine.
mirabilis, Proteus vulgaris, Proteus sp., Pseudomonas puorescens/putida, Pseudomonas sp., and Vibrio alginolyticus) with decarboxylase activity (Middlebrooks et al., 1988). Histamine forming bacteria in tuna fish (Thunnus thynnus) destined for canning are primarily members of the Enterobacteriaceae family (Lopez-Sabater et al., 1994a). Among these, the most active and frequent histamine former was Morganella morganii, followed in activity by Klebsiella oxytoca, Klebsiella pneumoniae, and some strains of Enterobacter cloacae and Enterobacter aerogenes. All these bacteria produced more than 500 ppm of histamine under experimental conditions. Using gas-liquid chromatography, Suzuki et al. (1988) studied the production of polyamines by putrefactive Pseudomonas type III/IV. They found that halophilic types produced spermidine in high ratios and that nonhalophilic types produced putrescine but did not produce spermidine. Photobacterium was also shown to be involved in the production of histamine, agmatine, and
BIOGENIC AMINES IN FISH AND SHELLFISH
335
cadaverine (Okuzumi et al., 1990).Histamine production by Photobacterium phosphoreum was greater under anaerobic than aerobic conditions, while the organism grew better under aerobic conditions (Morii el af., 1994). Behling and Taylor (1982) reported that the minimum temperature for two strains of Morganella morganii to produce toxicologically significant levels of histamine in tuna fish infusion broth (30 mg/lOO g) was 15°C. This was confirmed by Klausen and Huss (1987) who reported that M . morganii grows well at temperatures of 15°C (generation time is 2.6 hr at 15"C, 1.08hr at 25°C) or above and that the growth is greatly reduced at temperatures below 10°C (generation time is 10.2 hr at 1O"C, 14.9 hr at 5°C). They were able to confirm that large amounts of histamine (600-1400 ppm) can be produced by M. morganii in mackerel stored at low temperatures (0-5°C) following storage at higher temperatures (10-25°C). Frank et al. (1985) studied the composition and decarboxylase activity of bacteria isolated from decomposed mahi-mahi. They reported that the mesophilic bacteria isolated from fish incubated at 32°C for 24 hr were mainly Gram-negative rods; 89% of these were Vibrio alginolyticus. Strong histamine-producer (>lo0 mg/100 ml) mesophiles were Morganella morganii and Proteus mirabilis. Weak histamine-producer ( 4 0 mg/100 ml) mesophiles were all Vibrio alginolyticus. Psychrotrophic isolates obtained from fish incubated at 0°C for 14 days were essentially Gram-negative bacteria, 9% of which were Afteromonas putrefaciens, a weak histamine producer (<1 mg/100 ml) at 5 and 20°C. To evaluate a microorganism's ability to produce histamine in tuna fish, Omura et af. (1978) developed a culture medium from an extract of fresh skipjack tuna (Euthynnus pelamis). The medium was later modified by Arnold etaf.(1980), Yoshinaga and Frank (1982), and Ramesh et al. (1989). A further modification in the composition of the broth was further obtained by using sardine instead of skipjack. Variations in composition, due to species and quality of the raw material, were reported by Taylor and Woychik (1982) and Yoshinaga and Frank (1982). Taylor and co-workers (1978b) were able to eliminate some variation by using a medium consisting of trypticase soy broth supplemented with 0.1% histidine monohydrochloride with a pH adjusted to 6.8. To assure maximum histamine formation, Taylor and Woychik (1982) increased the histidine content to 2% and adjusted the pH to 6.3. The medium was further modified (Frank et al., 1985) by adding 0.01% pyridoxal hydrochloride as a cofactor. Early studies determined the bacterial histidine decarboxylase activity by using a Warburg manometer to measure the C 0 2volume released in the culture broth. Nowadays, the histamine produced in the culture broth is measured mainly by high-pressure liquid chromatography (HPLC), spectrofluorometry and thin-layer chromatography (TLC). Sumner and Taylor (1989) developed a
336
DAFNE D. RAWLES etal.
method for detecting histamine-producing lactic acid bacteria, using leucocrystal violet and diamine oxidase. With this technique it was possible to detect bacteria capable of producing more than 1.2 pmol of histamine per milliliter. However, the technique had only qualitative value due to interference from the culture broth (Sumner and Taylor, 1989). LopezSabater et al. (1994b) adapted an enzymatic method to measure histidine decarboxylase activity in bacteria isolated from fish. Quantification of histamine at levels as low as 3 ppm were achieved, with correlation of 0.99 between histamine content and the increase of absorbance in the concentration range between 3 and 30 ppm. V. AMINES PRESENCE IN THE MARINE ECOSYSTEM
The diamines putrescine (1,4-diaminobutane) and cadaverine (1 ,5-diaminopentane) belong to a group of natural polyamines which serve as stabilizing cations of the macromolecular structure of DNA and RNA and can be expected to be abundant in all living matter (Cohen, 1971, 1978). The abundance of these diamines is documented in all major groups of marine organisms such as algae (Rolle et al., 1971), invertebrates (Manen and Russell, 1973), vertebrates (Cohen, 1971), and microorganisms (Cohen, 1971). Therefore, the presence of diamines in the marine environments can be expected. Hofle (1984) investigated the potential in the marine microbial community to degrade diamines. They reported removal of putrescine and cadaverine from coastal waters supplemented only with these compounds within 48 hr. There was no increase in the bacterial biomass, growth rate, or viability when compared to the control (unsupplemented) cultures. Labeled putrescine experiments indicated that most putrescine carbon is mineralized to C 0 2rather than assimilated by the bacteria. At the concentrations added (500 Fg per liter), the diamines were not toxic to the marine bacteria. VI. AMINES OCCURRENCE IN SEAFOOD
A. FINFISH Fish muscle can support formation of a wide variety of amine compounds resulting from the direct enzymatic decarboxylation of amino acids. The substrate for the decarboxylase enzymes are free amino acids, therefore amine build-up normally occurs during a decomposition or spoilage process involving formation of free amino acids through proteolysis together with bacterial production and action of an amino acid decarboxylase (Eitenmiller
BIOGENIC AMINES IN FISH AND SHELLFISH
337
and De Souza, 1984). Amines frequently found in fish muscle include cadaverine from lysine, putrescine from ornithine, and histamine from histidine. Factors involved in the formation of histamine, related to scombroid food poisoning (histamine intoxication), have been widely studied in several different fish species. Tuna and other fish from the families Scomberesocidae and Scombridae and a nonscombroid fish, mahi-mahi or dolphin (Coryphaena hippurus), are known to contain high levels of histamine when spoiled. Muscle in scombroid fish as well as in mahi-mahi contains high levels of histidine, being readily transformable in toxic levels of histamine. Determination of histamine in such species is difficult due to the lack of rapid detection methods, because histamine can reach toxic levels in the absence of organoleptic indication of spoilage and because there may be an unequal distribution of the toxic agent in the fish. Kim and Bjeldanes (1979) determined concentrations of cadaverine, putrescine, histamine, spermidine, and spermine in canned wholesome tuna and in canned tuna that had been implicated in an outbreak of scombroid poisoning in humans (Table 11).Fresh tuna recently caught contains negligible quantities of histamine, usually less than 1 ppm (Frank et al., 1981). Fernandez-Salguero and Mackie (1987b) determined the levels of higher amines in canned tuna, mackerel, and sardine. They reported very low levels of histamine (below 0.5 mg/100 g) in all the samples analyzed, putrescine cadaverine ranged between 0 and 1.5 mg/100 g and spermidine and spermine between 0 and 2.5 mg/100 g. A good correlation has been found in Spanish mackerel (Scomberomorus maculatus) between the levels of histamine, cadaverine, and putrescine and the time and temperature of decomposition, between the ratios of cadaverine/histamine and putrescine/histamine levels and the temperature of decomposition, and between increasing total microbial counts and rising amine levels (Middlebrooks et al., 1988). No formation of amines during ice storage of mackerel (Scomberjaponicus) was reported by Wendakoon et af. (1990). However, at 20"C, histamine, putrescine, cadaverine, and tyramine were formed in large amounts. The low initial levels of spermidine in both dark and white muscle decreased during storage regardless of temperature. Amine production rate in dark muscle was higher and reached higher levels than in white muscle. Histamine poisoning (scombrotoxic fish poisoning) has been frequently associated with the consumption of spoiled scombroid fish, which usually have high levels of histidine in their muscle tissue, such as tuna and mackerel (Gilbert et al., 1980). However, pelagic fish, such as sardines, herring, and pilchards, and some types of cheese have been involved in outbreaks of this illness (Taylor, 1986). Thin-layer chromatography analysis of nine samples of smoked herring indicated that putrescine, cadaverine, and spermine
TABLE I1 TLC ANALYSIS OF AMINES IN COMMERCIAL (NONDECOMPOSED) AND DECOMPOSED CANNED TUNA FISH?
Range of amines (mg%) Fish sample
Cadaverine
nb
Commercial canned tuna fish
34
Canned decomposed' tuna fish
15
-
Avg 5 SE Median Avg % SE Median
Data from Kim and Bjeldanes (1979). Total number of cans analyzed. Lots D417 and D419 implicated in human poisoning.
0 3.70 1.05 -t 0.60 0.80 2.40 21.0 12.8 ? 1.66 10.8
-
Putrescine
-
0 2.50 0.35 2 0.10 0.12 0 5.60 1.53 ? 0.44 1.25
-
Histamine
-
1.0 17.0 2.14 2 0.71 1.97 9.70 200 116 ? 6.24 118
-
Spermidine
-
0.47 7.90 3.26 ? 0.33 2.5 0 4.90 2.37 ? 0.50 2.63
-
Spermine
-
0.25 2.90 1.23 ? 0.09 1.15 0 2.2 1.16 IT 0.15 1.30
-
339
BIOGENIC AMINES IN FISH AND SHELLFISH
were present at levels between 1 and 16 pg g-' (nine samples), while tyramine, spermidine, and histamine were present at low levels (1-8 p g g-') in five, five and seven samples, respectively (Shalaby, 1995).Tryptamine and phenylethylamine were not detected in any samples analyzed. In some cases, fish with low contents of histamine have been implicated (Murray et al., 1982), indicating that other substances might be involved, possibly as histamine toxicity potentiators (Bjeldanes et al., 1978;Klausen and Lund, 1986). Biogenic amines such as cadaverine and putrescine have been shown to potentiate the uptake of histamine in vitro (Lyons et al., 1983) and inhibit intestinal histamine-metabolizing enzymes in vivo (Hui and Taylor, 1985). Fletcher et al. (1995) in a survey of the histidine levels in retail fish in New Zealand, found levels above 1000 mg/100 g of muscle in the white muscle of Albacore (Thunnus alalunga) (4280 mg/100 g), Kingfish (Seriola grandis) (1580 mg/100 g), and Kahawai (Arripis trutta) (1242 mg/100 8). Changes in bacteria, amino acids, and biogenic amines in sardines (Sardina pilchardus) stored at ambient temperature were reported by Ababouch et al. (1991). Table 111 shows the bacterial counts for fresh sardines. Bacteria located initially on the skin and gills of the freshly caught sardines invaded and rapidly grew in the sardine muscle, reaching 5 X lo8 CFU/g after 24 hr at ambient temperature and 6 X lo8 CFUlg after 8 days in ice. The fresh sardines consisted of 64% water, 20% protein (N X 6.25), 2.4% fat, and 2.5% ash. Sardine amino acid content (Table IV) is rich in histidine, mainly free, arginine, phenylalanine, and lysine. During storage, a decrease in the levels of histidine, arginine, lysine, tyrosine, and methionine and an accumulation of the other amino acids, except proline and taurine, was observed in the fish muscle. After 24 hr of storage at ambient temperature, histamine, cadaverine, and putrescine accumulated to levels of 2350, 1050, and 300 ppm, respectively. In ice, histamine and cadaverine reached similar levels after 8 days, while putrescine formation was insignificant. The large amount of cadaverine accumulated in the sardines indicated that proteolysis had taken place during storage as the initial amount of its precursor, free lysine, was low (200 ppm). Arginine can be metabolized into ornithine and/ TABLE 111 (Sardina pitchardus)
BACTERIAL COUNTS OF FRESH SARDINES
PER GRAM OF GILLS, VISCERA, MUSCLE, OR MUSCLE
Skin Total bacterial count 2.5 x 106 Histamine producing bacteria 5 X 10
" Data from Ababouch et al. (1991).
Gills
Viscera
(CFU~CM' OF SKIN OR
+ SKIN)U
Muscle
1.2 x 105 3.1 x 104 3 x 104 3.3 X los 1.1 X lo4 6 X lo2
Muscle
+ Skin
1.3 x 106 4.4 X l@
340
DAFNE D. RAWLES eral.
TABLE IV AMINO ACID COMPOSITION OF FRESH S A R D I N E S (Sardina ~ pilchardus)
Type of amino acid
Free amino acid content (mg 100 g-l)
Thr Val Met CYS Ileu Leu Phe TYr LYS
TrP ASP Ser Glu Pro GlY Ala His Arg Taurine Ornithine (I
13.9 14.6 4.6 0 6.9 13.7 9.7 3.9 20.0 0 4.1 9.0 29.9 19.2 13.9 56.0 288.8 54.0 128.6 1.5
Total amino acid content (mg g-'1 10.3 15.9 9.2 1.6 13.4 22.1 12.3 9.4 29.0 0 17.9 9.6 40.9 10.0 15.7 17.4 10.6 17.7 1.3 0.085
Data from Ababouch et al. (1991).
or putrescine. Ornithine can also be derived from glutamic acid (Lehninger, 1982). Amine content in red perch and anchovy, determined by extraction with trichloroacetic acid and high-pressure liquid chromatography, are shown in Table V. The free amino acid content in fishery products is high when compared to terrestrial animals since the primary function of the compounds in aquatic organisms is to serve as osmoregulators. It is also known that the amino acids in fish can be rapidly decarboxylated. Haaland et uf. (1990) reported that the formation of free amino acids post mortem was temperature dependent. Most amino acids showed higher values at 2°C than at 20"C, and amine formation was higher at 20°C than at 2°C. They also reported that the formation of cadaverine and putrescine was higher in whole (ungutted) fish than in filets taken from stored whole fish. However, the reverse results were observed when gutted fish were studied. Fernandez-Salguero and Mackie (1987a) reported that histamine, cadaverine, and putrescine were produced more rapidly in haddock (Melunogramrnus aeglefinus) fillets than
BIOGENIC AMINES IN FISH AND SHELLFISH
341
TABLE V AMINE CONTENT (HPLC DETERMINATION) IN
RED PERCH AND ANCHOVY SAMPLES*
Average content (mglkg) Amine
Anchovy
Red perch
Putrescine Cadaverine Histamine Tyramine
18 107 650 111
13 109 ND’ ND
’Data from Feier and Goetsch (1993).
’ND, not determined.
in the whole gutted fish and that ungutted fish spoiled more rapidly than fillets. Reported concentrations (Fernandez-Salguero and Mackie, 1987a) of nonvolatile amines formed in haddock and herring during storage in ice and at 5°C are presented in Tables VI and VII. The rates of biogenic amine formation in fish can be summarized as whole ungutted fish > filets from whole ungutted fish; filets > whole gutted fish. In storage tests (Wei et al., 1995), fresh swordfish was shown to contain spermine, and the longer the storage, the higher in the amount of putrescine, cadaverine, histamine, and spermidine. The order of producing rate of biogenic amine was histamine > cadaverine > spermidine > putrescine. Amino acid formation is also dependent on the harvesting season and feeding activity prior to capture (Aksnes and Brekken, 1988).Fish harvested in summer during feeding quickly liberated large quantities of lysine and arginine, the microbial precursors of cadaverine and putrescine, respectively, due to the presence of large quantities of intestinal tract enzymes. Haaland et ul. (1990) concluded that the storage conditions of whole mackerel have little effect upon the nutritional composition of whole fish meal, unless the fish is stored for at least 7 days at 20°C. Dawood et al. (1988) studied the effect of holding (between 0 and 30°C) freshly caught rainbow trout (Sulmo irideus) for a period of 6 hr prior to chilled storage. Samples of whole and eviscerated fish were analyzed for putrescine, cadaverine, histamine, spermidine, and spermine by highperformance liquid chromatography of their benzoyl derivatives at intervals between 2 and 14 days of storage at 0°C. They observed an increase in the concentration of putrescine (C3.5 to 6-8 pg/g), cadaverine (<1 to 1-4 pglg), and histamine (4to 4.5-13 pglg) during storage, while the levels of spermidine (initially <5 pg/g) and spermine (initially <6 pg/g) decreased after an initial increase during the first 4 days. Tyramine was
TABLE VI PRODUCTION OF HIGHER AMINES (MG/IM) G TISSUE) IN HERRING AND HADDOCK HELD IN ICE AS FILLETS AND WHOLE
Putrescine
Time of storage (days)
Whole fish
Histamine
Fillet
Whole fish
Fillet
Whole fish
-
0 tr 0 0 1.65 9.67 18.99
0 0 0 0 6.81 15.46
0 2.75
Herring 0 2 4 6 8
tP tr tr tr
tr tr tr
tr tr
1.24 1.10 1.49
0.55
10
tr tr tr tr tr 0.96
tr tr tr tr tr
12 14 16
2.85 2.44 2.28
1.21 1.07 2.55
10 12
Haddock 0 2 4 6 8
Cadaverine
0 0 0 0 0
tr tr tr 0.78
Data from Fernandez-Salguero and Mackie (1987a). tr, <0.5 mg/lOO g tissue.
0 0 0 0 tr 1.56 2.82 5.11
tr 0.57 3.80 7.27 14.77
0
0 0 tr 1.10 1.06 7.38 7.40 7.23
Spermidine
FISH^
Spermine
Fillet
Whole fish
Fillet
Whole fish
-
tr
tr tr tr tr tr tr
tr tr tr tr tr tr
0.51
0
0 tr tr 18.77 12.68
-
0 0 0.70 0.79 4.02 9.82 9.15 28.11
tr tr tr tr tr tr tr tr tr
tr
tr tr tr tr tr tr tr
Fillet
-
tr tr tr tr tr
0.71 0.57 0.62 tr
0.89
0.92
1.35 1.39 1.08
1.22
0.89 0.68 1.10 1.70 1.44 1.70
1.oo
-
1.05 1.12 1.14 1.10 1.41 1.53 1.78
TABLE VII PRODUCTION OF HIGHER AMINES (MG/ICO G TISSUE) IN HERRING AND HADDOCK HELD AT
Putrescine
Time of storage (days) Herring 1 3 5 7
Histamine
Cadaverine
Whole fish
Fillet
Whole fish
Fillet
Whole fish
trb 0.89 4.78 6.24
tr tr 0.58 2.28
0 1.16 9.88 53.10
0 0.82 30.66 52.38
tr 5.20 47.54 36.66
tr
tr 0.52 0.54 1.79 2.09
tr tr 0.64 1.71 1.99
0 tr tr 3.57 4.96
0 0 2.31 9.24 7.67
tr 0.83 1.62 8.60 10.18
0 3.57 5.66 6.41 15.68
Fillet
3.38 29.64 38.66
5°C AS
FILLETS AND WHOLE FISH
Spermidine
Spermine
Whole fish
Fillet
Whole fish
Fillet
tr tr tr tr
tr
0.58 0.63 1.12 1.16
0.84 0.66 0.86 1.16
tr tr
tr tr tr
1.29 1.18 1.38 1.56 1.78
1.06 1.19 1.68 2.02 2.5 1
tr tr tr
Haddock 1 3 5 7 9
Data from Fernandez-Salguero and Mackie (1987a). tr, <0.5 mg/100 g tissue.
W W P
tr tr tr
tr
tr
344
DAFNE D. RAWLES etal.
not detected in any of the samples analyzed. Lower concentration of amines were found in the samples of eviscerated fish compared to the samples of whole fish. Capelin (Mallotus villosus) is an important raw material used for the production of fish meal and oil in Norway. Aksnes and Brekken (1988) studied the biochemical and microbial changes that occur during autolysis of bulk stored capelin with high contents of feed in the gut. They reported a fast release of arginine, serine, histidine, leucine, lysine, and tyrosine and a slower rate of release for glycine, aspartic acid, and glutamic acid. The amounts of tyrosine, lysine, serine, arginine, and histidine decreased rapidly due to bacterial activity, and the main products obtained from the bacterial decomposition of lysine, histidine, and arginine were cadaverine, histamine, and putrescine, respectively. Changes in bacterial flora and polyamine content during the storage of minced horse mackerel (Trachurus japonicus) meat were studied by Okuzumi et al. (1990). At spoilage stages (total aerobic count of 1.1 to 1.3 X 1O1O /g) in the samples stored at 5"C, Pseudomonas 1/11and Pseudomonas IIIIIV-NH, typical spoilage bacteria, were dominant in the bacterial flora and high amine contents of putrescine (2.3-54 mg/100 g), cadaverine (1115 mg/100 g), and histamine (7.2-12 mg/100 g) were detected. Samples stored at 30°C showed Vibrio and Photobacterium as dominant bacteria and high contents of histamine (210-1336 mg/lOO g) and cadaverine (74612 mg/100 g) were observed by the time the samples were at spoilage level (1.8-2 X lo9/g). Production of polyamines from arginine by Pseudomonas I/ 11,ornithine by Pseudomonas III/IV-NH, arginine and lysine by Photobacterium, and arginine by certain Vibrio was shown. It was suggested that Pseudomonas III/IV-NH produced putrescine in samples stored at 5°C and that Photobacterium produced agmatine and cadaverine in samples stored at 30°C. B. SHELLFISH Formation of nonvolatile amines during spoilage in the muscle of squid (Todarodes pacificus) and in a species of octopus (Paroctopus doJEeini dofleini) was studied by Takagi et al. (1971). During 4 days of storage they found cadaverine (10.1-7.6 mg% at 15°C and 27.7-6.0 mg% at 25°C) and very small amounts of putrescine (1.2 mg% at 15"C, 96 hr storage; 0.1 mg% at 25", 72 hr storage) in the muscle of squid and did not detect either tyramine or histamine. Cadaverine (2.4-6.0 mg% at 15°C and 3.9-4.5 mg% at 25"C), putrescine (3.4-7.1 mg% at 15°C and 1.5-1.0 mg% at 25°C) and small amounts of tyramine (trace at 15°C and 2.6 mg% at 25°C) were found in decomposing octopus muscle, and histamine was not detected.
34s
BIOGENIC AMINES IN FISH AND SHELLFISH
TABLE VIII (Patynopecten 5°C
CHANGES IN POLYAMINE CONTENTS (MG/IOO G ) I N JAPANESE SCALLOP
yessoensis)
ADDUCTOR MUSCLE DURING STORAGE AT
Storage time Sensory (h) Tyramine Putrescine Cadaverine Agmatine Tryptamine Spermidine rating' 0 24 48 72 96 a
0 0 0 0.03 0.15
0 0 0.12 0.49 2.71
0 0 0.14 0.37 0.88
0 0.38 0.40 0.69 0.79
0 0 0 0 0.08
0.10 0.09 0.08 0.08 0.08
1 1 1 2 3
Data from Yamanaka (1989). 1, acceptable: 2, initial decomposition; 3, advanced decomposition.
Yamanaka (1989) reported that storage at 5 and 15°Cof Japanese scallop (Putinopecren yessoensis) adductor muscle results in an increase in agmatine, putrescine, and cadaverine, while a significant decrease in arginine, with partial conversion into ornithine and agmatine, is observed. The changes in polyamine contents reported by Yamanaka (1989) during storage of Japanese scallop (Purinupecten yessoensis) adductor muscle at S and 15°C and the sensory ratings observed are shown in Tables VIII and IX. In crustaceans, the biogenic amines have a primary function as neurotransmitters and neuromodulators. In addition, some biogenic amines serve also as neurohormones in the hemolymph (Fingerman et ul., 1994). TABLE IX (Patynopecten 15°C"
CHANGES IN POLYAMINE CONTENTS ( M d I o o G) I N JAPANESE SCALLOP
yessoensis)
ADDUCTOR MUSCLE DURING STORAGE AT
Storage Sensory time (h) Tyramine Putrescine Cadaverine Agmatine Tryptamine Spermidine rating' 0 12 18 24 36 48 60 a
0 0 0 0 0.15 0.20 1.43
0 0.05 0.72 2.16 11.84 12.30 19.28
Data from Yamanaka (1989).
0 0.07 0.43 0.84 2.17 7.23 20.29
0 2.19 5.27 8.46 8.59 10.50 14.21
0 0 0 0.98 1.34 2.02 2.10
* 1, acceptable; 2, initial decomposition; 3, advanced decomposition.
0.10 0.08 0.08 0.08 0.29 0.60 0.70
1 1 2 3 3 3 3
346
DAFNE D. RAWLES etal.
C. SEAFOOD PRODUCTS Fermented fish products available mainly in Asian countries are particularly rich in histamine (Azudin and Saari, 1988; Wootton et al., 1986). Fermented fish paste, prepared from small fish or shrimp, is used frequently as a condiment for rice dishes in Southeast Asia. Fardiaz and Markakis (1979) tentatively identified the following amines in fish paste: ethanolamine, 2-methylbutylamine, 2-mercaptoethylamine, 2-phenylethylamine, cadaverine, and histamine in concentrations ranging from 0.5 to 64 mg/ 100 g. Histamine and 2-phenylethylamine were the major amines found with maximal amounts of 64.0 and 60.0 mg/100 g, respectively. Fermented sardine with rice bran is a traditional Japanese foods produced in a barrel for a period of 6 months to 1 year. In a study by Yatsunami et al. (1994), an increase in the numbers of halotolerant and halophilic histamineforming bacteria from 101-102Ig to 104-10s Ig after 6 months was observed. Putrescine, histamine, and tyramine content also had a considerable increase in the same time period. The isolates of halotolerant and halophilic histamine-forming bacteria from the raw sardines were identified as Staphylococcus, Micrococcus, Vibrio, Pseudomonas III/IV NH, and Pseudomonas IIVIV-H. The isolates from fermented sardine with rice-bran after 6 months were identified as Staphylococcus, Micrococcus, and Vibrio. VII.
SCOMBROTOXICOSIS
Histamine poisoning has been reported to be one of the major illnesses among foodborne diseases (Taylor et al., 1989; Morrow et al., 1991). In Britain, between 1976 and 1986, 258 incidents of suspected scombrotoxic fish poisoning were reported (Bartholomew et af., 1987). Originally associated with the consumption of fish belonging to the Scombridae and Scomberosocidae, scombrotoxicosis is a human intoxication which in severe cases is characterized by rapid onset (10 to 30 min). Symptoms of histamine poisoning include headache, nausea, vomiting, diarrhea, itching, oral burning sensation, red rash, and hypotension (Taylor er al., 1989; Hughes and Potter, 1991; Ijomah et al., 1992). Recently, this intoxication has also been associated with the consumption of certain nonscombroid species (Bartholemew et al., 1987; Hughes and Potter, 1991; Morrow et al., 1991) and in rare ocassions with the consumption of cheese, mainly Swiss cheese (Taylor et al., 1989). The level of histamine which constitutes a toxic level is uncertain as potentiators of toxicity which lower the effective dosage may be present in the fish (Arnold and Brown, 1978; Taylor, 1985). It is generally accepted that the risk of scombrotoxic poisoning in well-iced fish is very
BIOGENIC AMINES IN FISH AND SHELLFISH
347
low. Under normal conditions of storage, histamine concentration rarely reaches levels higher than 5 mg/100 g flesh (Arnold and Brown, 1978; Murray et al., 1982). The risk for histamine poisoning is associated not only with the consumption of fresh fish. Due to the heat resistance of histamine (Ijomah et al., 1992), it can be present in cans of tuna fish and other related species (Ienestea, 1971). The presence of histamine in toxic amounts, in sterilized cans, can also be due to the use of fish of poor hygienic quality as raw material or to defective handling of high quality tuna during processing. When incidents of poisoning have been reported for scombroid species the concentrations of histamine have often been found to be high, in excess of 100 mg/100 g flesh (Arnold and Brown, 1978). Scombrotoxic poisoning was attributed to the consumption of canned tuna containing 68-280 mg histamine/100 g fish (Merson et al., 1974). Possible vasoactive or psychoactive effects due to an excessive oral intake of biogenic amines have been reported (Malone and Metcalfe, 1986;Taylor, 1986;Joosten, 1988). Since destruction of these amines by physical methods such as freezing and heating is very difficult, it is important to prevent their formation. Prevention of amine formation in fish muscle extracts by some spices has been demonstrated (Wendakoon and Sakaguchi, 1992). A possible synergistic effect of clove essential oils and sodium chloride for the inactivation of the growth and amine production of Enterobacter aerogenes in mackerel muscle broth at 30°C was suggested by Wendakoon and Sakaguchi (1993). Contrary to widespread belief, new medical evidence has demonstrated that histamine present in the fish has but a minor role in the aetiology of scombrotoxicosis (Douglas, 1980; Clifford et al., 1989; 1991; Ijomah et aL, 1991,1992). Consequently, two hypotheses have been formulated: (1) that the action of dietary histamine is potentiated by some other component(s) of the toxic fish (Bjeldanes et al., 1978; Taylor and Lieber, 1979; Chu and Bjeldanes, 1981;Lyons eta1.,1983; Ijomah et al., 1992) or (2) that the toxin(s) is a mast cell degranulator, and the antihistamine therapy is effective because it eliminates the effect of endogenous histamine rather than the effect of dietary histamine (Clifford et al., 1991; Ijomah et al., 1991; Ijomah et al., 1991). Several reports of in vivo and in vitro studies suggest that the absorption, metabolism, and/or potency of one biogenic amine might be modified in the presence of a second amine (Bjeldanes et al., 1978; Taylor and Lieber, 1979; Lyons et al., 1983). Histamine taken in combination with wholesome tuna is reported to yield toxic effects in people even at moderate doses (100-180 mg/100 g tuna) (Motil and Scrimshaw, 1979). Diamines and polyamines inhibit the binding of histamine to mucin, therefore histamine could be released from intestinal mucin and increase the amount of histamine
348
DAFNE D. RAWLES etal.
absorbed in the intestinal lumen (Jung and Bjeldanes, 1979; Chu and Bjeldanes, 198l), resulting in an increase in histamine toxicity. Histamine taken with a meal (bread, milk and butter) has been reported to be absorbed to a greater extent than histamine consumed by itself (Mitchell and Code, 1954). Storage temperature has an important role in the production of histamine (Behling and Taylor, 1982). It is generally agreed that temperatures of 0°C or below inhibit histamine formation. However, the effect of storage at temperatures between 2 and 10°C is not that clear. Several studies have reported that at these temperatures there is little or no formation of histamine (Hardy and Smith, 1976; Smith et al., 1980; Klausen and Lund, 1986) while others have reported production of low levels of histamin at temperatures below 10°C (Baldrati et al., 1980; Morii et at., 1986). The differences in these reports could be due to the type and level of microbial flora of the fish used in the studies (Behling and Taylor, 1982). VIII.
AMINE FORMATION AS AN INDICATOR OF FRESHNESS IN SEAFOODS
The potential use of amine concentration as a criteria to assess freshness in finfish and shellfish has been discussed by many researchers. The volatile amines trimethylamine (TMA) and dimethylamine (DMA) have been widely used as indicator of freshness of marine fish (Castell et al., 1971). Tables X and XI give the concentration of TMA and DMA in herring and TABLE X CONCENTRATION OF TMA AND DMA (MG N/IOO G) IN HERRING AND HADDOCK HELD IN ICE AS FILLETS AND WHOLE
FISH^
Herring Time of storage (days) 0 2 4 6 8 10 12 14 16
Whole fish
Haddock Fiilets
TMA
DMA
TMA
DMA
1.46 3.79 2.28 5.40 19.64 20.79 22.88
0.50 0.45 0.30 0.25 0.20 0.20 0.20 -
0.95 1.28 2.71 7.49 35.84 33.52
0.20 0.40 0.40 0.40 0.54 0.56
-
Whole fish
Fillets
TMA
DMA
"MA
DMA
0.45 0.45 1.37 1.42 3.68 6.87 7.56 9.99 17.10
0.20 0.25 0.50 1.13 2.05 4.43 4.39 3.99 3.27
-
0.47 0.72 1.61 6.58 23.12 37.45 43.47 41.47
0.50 0.20 0.35 1.oo 0.80 0.50 0.55 0.60
Adapted from Fernandez-Salguero and Mackie (1987a).
349
BIOGENIC AMINES IN FISH AND SHELLFISH
TABLE XI CONCENTRATION OF TMA AND DMA (MG N/IOO G) IN HERRING AND HADDOCK HELD AT
5°C AS
FISH^
FILLETS AND WHOLE ~~~
~
~
~~
Herring
Time of storage (days) 1 3 5
7 9 a
Whole fish
Haddock Whole fish
Fillets
Fillets
TMA
DMA
TMA
DMA
TMA
DMA
TMA
DMA
1.86 20.34 50.25 57.68 -
0.70 0.45 0.50 0.45 -
1.49 9.56 39.85 53.62 -
0.20 0.30 0.40 0.44 -
0.70 5.45 10.31 37.71 40.86
0.18 1.13 1.62 1.90 4.88
0.80 8.69 39.31 50.45 50.01
0.50 0.50 0.20 0.90 1.30
Adapted from Fernandez-Salguero and Mackie (1987a).
haddock during refrigerated storage (Fernandez-Salguero and Mackie, 1987a). Luong and Male (1992) developed a biosensor system to measure the hypoxanthine concentration ratio as an indicator of fish freshness. Hypoxanthine is the autodegradation product formed from adenosine 5’triphosphate in fish tissue and is responsible for the bitter “off-taste” characteristic of fish which has lost its freshness. Dawood et af. (1988) reported that the concentration of putrescine and cadaverine in the flesh of chilled-stored rainbow trout (Salmo irideus) exceeded 1.10ppm by the second day of whole fish storage at 0°C. Therefore, they indicated that determination of these two diamines could be used to assess fish freshness. Putrescine and ornithine have been reported as potential indicators of freshness of scallop adductor muscle (Yamanaka, 1989). Spermidine and spermine are minor components in fish, changing only slightly during the period of storage over which fish is acceptable for human consumption (Ritchie and Mackie, 1980). Putrescine and cadaverine have a steady increase once bacterial spoilage begins, thus, they are considered potential indicators of fish quality (Fernandez-Salguero and Mackie, 1987a). Yamanaka et af. (1987) examined tyramine, putrescine, cadaverine, agmatine, and tryptamine variations in the muscle of common squid during storage at 0,3.5, and 15°C.Initially, even in the fresh muscle, agmatine was detected in small amounts. The agmatine concentration increased with storage time, exceeding 30 mg/100 g at the stage of initial decomposition and reaching 40 mg/100 g at the stage of advanced decomposition. They concluded that agmatine, formed from arginine, can be potentially a useful indicator for freshness of common squid. At 22°C a good correlation between the spoilage
350
DAFNE D. RAWLES etal.
of imitation crab meat and the amounts of total volatile acids, total volatile bases, cadaverine, putrescine, histamine, aerobic plate count, and protelytic count has been shown (Hollingworth et al., 1991). However, at 4 and 10°C, neither the chemical nor microbial indicators were adequate to assess quality of the product. Ethanol has proven to be a useful chemical indicator, especially in canned salmon (Hollingworth et al., 1987). An increase in the levels of putrescine, cadaverine, tyramine, and tryptamine was observed as the decomposition of sardine (Surdinops melanosticla) and saury pike (Cololubis saira) progressed during storage at 5 or 20°C (Yamanaka et al., 1986). Maximum formation was observed for cadaverine, reported to be the most useful index for decomposition of fish: below 15 mg/100 g of meat at the passable stage, between 15 and 20 mg/100 g of meat at the stage of initial decomposition, and over 20 mg/100 g of meat at the stage of advanced decomposition. While the presence of histamine in fish muscle is a good indication that decomposition has taken place, its occurrence is extremely variable. Its production is a function of time, temperature, and the type and level of microbial flora present. Variations within individual fish, depending upon the section of fish from which the meat sample originated (Lerke et ul., 1978; Frank et al., 1981), as well as variation between and within species (Edmunds and Eitenmiller, 1975) have been reported. Frank et al. (1984) reported that decomposed skipjack tuna frequently had histamine levels of <5 mg/100 g. In a survey of histamine levels in commercially processed scombroid fish products by Taylor et al. (1977), over 90% of these products had histamine levels below 5 mgllO0 g. Therefore, although histamine may confirm the presence of decomposed tissue, it is not a good indicator of the degree of decomposition in fish products. A chemical index (Fig. 3) that correlates well with sensory evaluation for canned tuna of varying qualities was proposed after the work of Hui and Taylor (1983) and Mietz and Karmas (1977). The index comprises levels of histamine, putrescine, cadaverine, spermidine, and spermine. Production of putrescine, cadaverine, and histamine and loss of spermidine and spermine were observed during the decomposition of raw and commercially canned tuna fish. Using the HPLC method described by Hui and Taylor (1983), histamine levels less than 5 mg/100 g can be detected. To increase sensitivity, fluorescence detection instead of uv detection can be used (Seiler, 1970).
(mglkg histamine + mglkg putrescine + mglkg cadaverine) quality index = (1
+ mglkg spermidine + mglkg spermine)
FIG. 3. Chemical quality index (based on Mietz and Karmas, 1977).
BIOGENIC AMINES IN FISH A N D SHELLFISH
351
A study of the decomposition of raw surimi and a surimi-derived flaked artificial crab stored at 4,10, and 22°C (Hollingworth et al., 1990) indicated that total volatiles acids and total volatile bases content has the most potential as chemical indicators of decomposition compared to ethanol, cadaverine, putrescine, and histamine content. IX.
RECOMMENDED LIMITS OF AMINE CONTENT
The US. Food and Drug Administration (FDA) in 1982 established a defect action level for histamine in scombroid fish products of 20 mg/ 100g, which indicates prior mishandling, and a hazard action level of 50 mg/ 100 g, which is considered a potential health hazard. Although the relationship between the level of histamine and the toxicity of a fish sample is not clear, based on the available present knowledge, it is considered that a histamine content >200 ppm in fish can be considered to be toxicologically significant and that a content of >50 ppm indicates that the fish has been exposed to higher temperature (Arnold and Brown, 1978; Murray et al., 1982).In France a maximum permitted concentration of histamine in marine products has been set at 10 mg/100 g (as applied to fresh tuna) (Gouygou et al., 1992). The FDA proposed in 1995 that its compliance policy guide be revised on decomposition and histamine to: (1) lower the histamine action level for decomposition to 50 ppm from 100 ppm; (2) extend application of the new action level to raw and frozen tuna and mahi-mahi; (3) eliminate the provision that findings of less than 200 ppm need to be confirmed by organoleptic testing; (4) allow other species to be considered as decomposed when histamine levels reach, 50 to 500 ppm; and (5) consider fish, other than tuna, implicated in histamine poisoning outbreaks as health hazards when histamine levels reach 500 ppm. It is anticipated that final action on this proposal will be announced after 1996 (Craven et al., 1995; National Fisheries Institute, Inc., 1995). A. QUALITY CONTROL PURPOSES Pelagic fish Histamine < 5 mg/100g (Murray et al., 1982) Canned skipjack tuna Putrefactive amines 0.4 to 0.5 pglg (Sims et af., 1992)
352
DAFNE D. RAWLES etal.
Canned tuna fish Histamine < 50 ppm (Lonberg et al., 1980) B. TOXIC LEVEL Histamine > 100 mgllOO g flesh (Arnold and Brown, 1978) The European Union has established regulations for species of fish belonging to the Scombridae and Clupeidae families and fixed 100 ppm of histamine as the limit of acceptance. A three-class plan for maximum allowable levels of histamine in fresh fish ( a = 9; c = 2, rn = 100 ppm; M = 200 ppm) and enzymatically ripened fish products (n = 9; c = 2, rn = 200 ppm; M = 400 ppm) from Scombridae and Clupeidae) families (where n = number of units to be analyzed from each lot, m and M = histamine tolerances, and c = number of units allowed to contain a histamine level higher than m but lower than M ) has been initiated (Anonymous, 1991). Histamine levels higher than 2000 ppm have usually been reported in sardine, mackerel, and tuna fish cans (Kim and Bjeldanes, 1979; Lonberg et al., 1980; Schulze and Zimmermann, 1980;Ababouch et al., 1986). Taylor et al. (1978a) reported that only 4% of mackerel and tuna fish cans showed a histamine level lower than 10 ppm. Ababouch et af. (1986) observed that 7% of tuna fish samples contained levels higher than 500 ppm of histamine as compared to 3.7 and 3.2% for sardine and mackerel, respectively. Sailfish filets (Zstiophorus platypterus) were involved in a 1994 food poisoning outbreak in Taiwan. Samples were collected from the victims' residues and analyzed for amines by a HPLC gradient elution technique. Analysis (see Table XII) indicated that tryptamine, histamine, spermine, TMA, spermidine, and cadaverine were prominent in all samples; other amines were not detectable (<2.5 mg/100 g) (Hwang et al., 1995). TABLE XI1 CONCENTRATIONS OF BIOGENIC AMINES IN SAILFISH TISSUE CAUSING SCOMBROID POISONING"
Amine level (mg/lOO g) Source
Cadaverine
Tryptamine
Spermidine
Histamine ~
Victim Victim
14.5
2
0.3
11.0 k 0.4
208 2 8 185 ? 11
Adapted from Hwang et a!. (1995).
50.0 t 3.9 20.0 2 1.6
168 2 8 180 ? 9
Trirnethylamine ~
24.5 ? 3.3 19.0 ? 0.5
BIOGENIC AMINES IN FISH AND SHELLFISH
X.
353
DETERMINATION OF BIOGENIC AMINES IN FISH
Various analytical techniques, including thin-layer chromatography, amino acid analyzers, liquid chromatography, gas chromatography, and enzymic tests have been developed for the determination of amines. Histamine has been usually determined by a specific fluorometric reaction with orthophthalaldehyde (Shore, 1971) and its concentration in fish with high levels of free histidine have been widely used as an index of spoilage and of potential scombrotoxic poisoning (Taylor, 1983). Histamine and the diamines have been determined by ion-exchange procedures in decomposed mackerel (Hatano et al., 1970). The fluorometric procedure is the official AOAC method most commonly used for the determination of histamine (AOAC Methods, 1980). The method involves extraction of the fish with methanol, separation of histamine from amino acids by passing the extract through an ion-exchange column, and reaction with o-phthalaldehyde under controlled conditions, followed by fluorometric measurement. This method has proven to be accurate and sensitive, with a detection limit of approximately 1 mg histamine/100 g fish sample (Gouygou et af.,1992). Drawbacks are that it is rather slow (maximum four or five samples per hour) and requires strict handling.
A. LIQUID CHROMATOGRAPHY Walters (1984) described a liquid chromatography (LC) system using a bonded cation-exchange column to resolve and detect histamine from histidine, cadaverine, and putrescine from fish. Detection of histamine by LC is difficult since histamine does not have enough ultraviolet absorbance or fluorescence nor is the molecule suitable for electrochemical detection. The introduction of a postcolumn reaction with o-phthalaldehyde to form a fluorescent derivative increased the detection sensitivity and specificity. The limit of detection was 1.5 ng histamine with a linear response in the range between 7 and 750 ng. Prior work by Hatano et al. (1970) used cation-exchange chromatography with a commercially available amino acid analyzer for amine analysis of samples containing between 0.1 and 1.0 pmol of the amines with high accuracy (within 3.5%) and short time (8 hr).
B. THIN-LAYER CHROMATOGRAPHY Low levels of spermidine, spermine, putrescine, cadaverine, and histamine can be determined with the TLC method developed by Spinelli et al. (1974). The method uses the fluorescent derivatives of the amines formed with dansyl chloride (5-(dimethylamino)-l-naphtalenesulfony chloride)
354
DAFNE D. RAWLES et al.
(Abdel-Monem and Ohno, 1975). Shalaby (1995) described the use of TLC with a multiple development technique to resolve the dansyl derivatives of histamine, cadaverine, putrescine, phenylethylamine, tyramine, tryptamine, spermine, and spermidine from fish, cheese, and meat samples. The procedure allowed for the detection in 14 samples of as little as 5 or 10 ng of the dansyl derivatives of the amines within 2 hr.
C. HIGH-PRESSURE LIQUID CHROMATOGRAPHY Polyamines are usually quantified by HPLC with precolumn (MorierTeissier et al., 1988; Yen and Hsieh, 1991) or postcolumn (Redmond and Tseng, 1979) derivatization by benzoyl chloride for measuring ultraviolet absorbance, dabsyl chloride for measuring visible absorbance, or dansyl chloride or o-phthalaldehyde for fluorescent detection (Abdel-Monem and Ohno, 1975;Desiderio et al., 1987).The dansyl derivatives of the polyamines have been detected with a 280-nm ultraviolet detector (Abdel-Monem and Ohno, 1975) and Mietz and Karmas (1977) and Hui and Taylor (1983) used a 254-nm detector and a gradient elution high-pressure liquid chromatographic system to separate the dansyl derivatives. The method was used to determine the quality of canned tuna and cheese. Later, Mietz and Karmas (1978) developed a chemical index to determine decomposition in rockfish, salmon, lobster, and shrimp. The dansylated polyamines, putrescine, cadaverine, spermidine, and spermine, and the amine histamine were quantified and an index formula was developed. Results from the chemical analysis of 21 samples (rockfish, salmon, and lobster) correlated well with the organoleptic analysis performed by 23 sensorists. The chemical index classified samples correctly 90.5% of the time versus an 83.9% correct classification by organoleptic determination. A good correlation between organoleptic and chemical index was also found in the analysis of shrimp composites. The dansyl derivatives, which have a napthtalene structure, are excellent derivatives for primary amines (Mietz and Karmas, 1978). They are easily formed and detected by a uv (HPLC) detector in amounts as little as 10ng, therefore the method does not require high sensitivity or fluorometric detectors. Gradient elution improves the separation, allowing for a broad range of derivatives to be separated in a relatively short time (40 min). Use of reverse-phase microparticular (5-10 pm) columns can further improve the HPLC separation (Gouygou ef al., 1992). Desiderio et al. (1987) described the use of a reversed-phase HPLC method for the quantification of putrescine, cadaverine, spermidine, and spermine from brain extracts as the dansyl derivatives.
BIOGENIC AMINES IN FISH AND SHELLFISH
355
Maruta ef al. (1989) described a rapid liquid chromatographic assay of urinary polyamines (putrescine, spermidine, spermine, and cadaverine) involving electrochemical detection with postcolumn-immobilized enzyme, polyamine oxidase, from soybean seedlings. The polyamines were separated by isocratic ion-pairing reversed-phase chromatography and then enzymatically converted, with release of hydrogen peroxide, via the postcolumn reactor with immobilized polyamine oxidase. The hydrogen peroixide was detected by electrochemical oxidation on a platinum electrode. Detection limits for the injected polyamines were 0.3,0.5,0.6 and 4 pmol for putrescine, spermidine, spermine, and cadaverine, respectively, with linear ranges of two to three orders of magnitude. Rosier and Van Peteghem (1988) described a rapid method for the extraction, derivatization, and determination by HPLC of the 5dimethylaminonaphthalene-1-sulphonyl (dansyl chloride) derivatives of putrescine, cadaverine, histamine, spermidine, and spermine from fish. Comparison of this procedure to earlier methods reflected considerable reduction in the time needed for sample preparation (from approx 8 to 0.5 hr) and cost (use of water and methanol instead of acetonitrile). Morier-Teissier et af. (1988) used precolumn derivatization with aphtalaldehyde and thiol combined with HPLC to measure spermine, spermidine, putrescine, and cadaverine. The derivatives were quantified by electrochemical detection instead of fluorescence, with an optimum potential from the working electrodes of + 0.65 V. A good linear relationship (Y = 0.99) existed between polyamines concentration and the peak height over the range 1 pmol to 10 nmol when the reaction time was carefully controlled. Spermidine gave the best sensitivity, with a signalhoke of 2 for 200 fmol. It was slightly lower for putrescine and cadaverine and seven to eight times lower for spermine.
D. GAS-LIQUID CHROMATOGRAPHY Yamamoto et af. (1982) developed a quantitative method for the determination of putrescine, cadaverine, spermidine, and spermine in foods. Separation of the amines from foods was achieved by eluting through a cationexchange resin column and then converted to their (ethyloxy) carbonyl derivatives by the reaction with ethyl chloroformate in aqueous medium before application to the gas chromatograph with a flame ionization detector. They used l,&diaminooctane as an internal standard and performed the separation and determination of the resulting derivatives on a 1.5% SE-3010.3% SP-1000 on Uniport H P column (0.5 m) under the temperatureprogrammed condition. Staruszkiewicz and Bond (1981) developed a GLC procedure for the quantitative determination of the diamines putrescine
356
DAFNE D. RAWLES eral.
and cadaverine using their perfluoropropionyl derivatives. Extraction of the amines from foods was performed with methanol and hexanediamine used as internal standard. A dry residue of the hydrochloride salts of the amines was prepared and derivatized with perfluoropropionic anhydride by heating for 30 min at 50°C. An alumina column was used to separate the reaction mixture and the derivatives were eluted with a 30% solution of ethyl acetate in toluene. GLC separations were performed on a 3% OV-225 column held at 180°C. Using this procedure, less than 1 p g diamine/g tissue could be quantitated using either an electron capture detector or a nitrogen-specific detector. Later, Farn and Sims (1987) applied the procedure on controlled spoilage studies of skipjack and yellowfin tuna and found, under the conditions of the test, that putrescine, cadaverine, and histamine developed in the raw fish only after severe temperature abuse (36 hr at 21°C). They concluded that whenever putrescine and cadaverine levels were found in amounts above background levels in canned tuna, it indicated that the original raw material had reached a stage of advanced decomposition prior to heat processing. A good correlation was observed between a panel of well trained sensory judges and the chemical analysis for putrefactive amines as an indicator of protein decomposition in canned skipjack tuna (Sims et al., 1992). The resolution by capillary gas chromatography of various substituted analogues of putrescine as their N,N’-perfluoroacyl derivatives was facilitated by the use of chiral stationary phases (Gaget et al., 1987). Bonilla et al. (1995) evaluated the use of a “cold on-column” GC injection technique in conjunction with a new base-deactivated fused silica capillary column for the direct analyses of putrescine and cadaverine by gas chromatography. They reported excellent resolution of the amines as well as symmetrical peaks. Good linear response was observed between concentrations of 20 and 400 ppm.
E. ENZYMIC TEST Lerke ef al. (1983) developed a rapid screening method to detect histamine in fish. The qualitative procedure uses a two-step sequential enzyme system. In the first step, the enzyme diamine oxidase catalyzes the breakdown of histamine with production of hydrogen peroxide. Detection of hydrogen peroxide is then performed by the formation of crystal violet from the leuco base in the presence of peroxidase at 596 nm. The method could be used to detect histamine in raw or heat-processed fish. Later, Lopez-Sabater et al. (1993) modified the procedure to achieve histamine quantification. Further modification of the Lerke method by Rodriguez-
BIOGENIC AMINES IN FISH AND SHELLFISH
357
Jerez et al. (1994) resulted in a rapid and reliable technique for histamine determination. Determination of the optimum wavelength depending on incubation time at a constant temperature of 37°C resulted in a recommendation of a wavelength of 580 nm, with an incubation time of 15 min. At these conditions the linearity was observed among 1 and 25 ppm (pg/g) of histamine.
F. ENZYME ELECTRODE Enzyme electrodes are formed by the joining of an electrochemical sensor and an enzymatic layer. Enzyme electrodes for oxidases are obtained by coupling an enzymatic film of immobilized oxidase with an oxygen electrode (Clark electrode) (Gouygou et al., 1992). Oxidase catalyzed reation: XH2
+0
2
+X
+ H202,
where XH2 is the reduced form of substrate and X is the oxidized form. Decrease in oxygen concentration, due to the enzymatic reaction in the film, is monitored amperometrically by the Clark electrode (Romette et al., 1982).
G. TEST STRIPS The test strips were obtained from an immobilized enzyme film containing both lentil seeds diamino oxidase and horse radish peroxidase (Gougouy et al., 1992). The films were made by coentraping the molecules of enzyme and an inactive matrix protein (gelatin or bovine serum albumin) with a bifunctional reagent (glutaraldehyde). A solid-phase assay (dipstick test) for histamine in tuna based on the coupling of diamine oxidase to a peroxidase/dye system was reported by Hall et al. (1995). The assay was linear to 1.O mM histamine, and the minimum detectable concentration was 0.07 mM, which corresponds to 2.3 mg% in tuna extracts. Putrescine did react slightly with the dipstick; however, determinations in fresh and spoiled tuna gave good agreement with the modified AOAC fluorometric method. H. CAPILLARY ZONE ELECTROPHORESIS Wang et al. (1994) reported on the use of capillary zone electrophoresis for the routine determination of biogenic amines in fresh fish samples. They obtained electrophoregrams of fluorescamine-derivatized histamine,
358
DAFNE D. RAWLES etal.
cadaverine, and putrescine in methanol and trichloroacetic acid under hydrostatic injection conditions. Although efficient peak separation of the standards was achieved within 7 min, the reproducibility of the injection varied among analyses due to difficultiesin regulating the amount of amine entering the capillary. They were able to improve the reproducibility maintaining an efficient peak separation by using an electrokinetic injection technique. ACKNOWLEDGMENTS This work was sponsored by the National Fisheries Institute, Inc., and by the National Oceanic and Atmospheric Administration (NOAA), Office of Sea Grant, U.S. Department of Commerce, under federal Grant NA90AAD-SG045 to the Virginia Graduate Marine Science Consortium and the Virginia Sea Grant College Program. The U.S. Government’s right to retain a nonexclusive royalty-free license in and to the copyright covering this paper, for governmental purposes, is acknowledged.
REFERENCES Ababouch, L., Alaoui, M. M., and Busta, F. F. (1986) Histamine levels in commercially processed fish in Morocco. J. Food Prot. 49,904-908. Ababouch, L., Afilal, M. E., Benabdeljelil, H., and Busta, F. F. (1991). Quantitative changes in bacteria. amino acids and biogenic arnines in sardine (Surdina pilchardus) stored at ambient temperature (25-28T) and in ice. Int. J. Food Sci. Technol. 26,297-306. Abdel-Monem,M. M., and Ohno, K. (1975). Separation of the dansyl derivatives of polyamines and related compounds by thin-layer and high-pressure liquid chromatography.J. Chromarogr. 107,416-419. Aksnes, A., and Brekken, B. (1988). Tissue degradation, amino acid liberation and bacterial decomposition of bulk stored capelin. J. Sci. Food Agric. 45, 53-60. Anonymous (1991). Normas sanitarias aplicables a la produccion y a la puestaen el mercado de 10s productos pesqueros. Directive 91\493\CEE, No. L 268, 22.07.1991. 08 J. Eur. Commun. 24.09.1991. AOAC Methods (1980). “Fish and Other Marine Products-Histamine,” 13th ed., pp. 294296, sections 18.012, 18.064-18.071. Arlington, VA. Arnold, S. H., and Brown, W. (1978). Histamine (?) toxicity from fish products. I n “Advances in Food Research” (C. 0. Chishester, E. M. Mrak, and G. F. Stewart, Eds.), Vol. 24, pp. 113-154. Academic Press, New York. Arnold, S. H., Price, R. J., and Brown, W. D. (1980). Histamine formation by bacteria isolated from skipjack tuna Katsuwonus pelamis. Bull. Jn. SOC.Sci. Fish. 46,991-995. Azudin, M. N., and Saari, N. (1988). Histamine content in fermented and cured fish products in Malaysia. “FA0 Fisheries Report No. 401 Supplement to the Seventh Session of the Indo-Pacific Fishery Commission Working Party on Fish Technology and Marketing,” pp. 105-111. Bangkok, Thailand, 19-22 April. Bachrach, U. (1973). “Function of Naturally Occurring Polyamines”. Academic Press, New York.
BIOGENIC AMINES IN FISH AND SHELLFISH
359
Baldrati, G., Fornari, M. B., Spotti, E., and Incerti, I. (1980). Effect oftemperature on histamine formation in fish rich in histidine. Industr. Con. 55, 114-122. Bartholomew, B. A., Berry, P. R., Rodhouse, J. C., Gilbert, R. J., and Murray, C. K. (1987). Scombrotoxic fish poisoning in Britain: features of over 250 suspected incidents from 1976 to 1986. Epidemiol. Infect. 99,775-782. Baylin, S . B. (1977). Histaminase (diamine oxidase) activity in human tumors: An expression of a mature genome. Proc. Nat. Acad. Sci. USA 74(3), 883-887. Behling, A. R., and S. L. Taylor. (1982). Bacterial histamine production as a function of temperature and time of incubation. J. Food Sci. 47, 1311-1317. Biondi, P. A,, Simonic,T., Secchi, C., and Ronchi, S . (1984). Sensitive assay for diamine oxidase activity using high-performance liquid chromatography. J. Chromatogr. 309, 151-155. Bjeldanes, L. F., Schutz, D. E., and Morris, M. M. (1978). On the aetiology of scombroid poisoning: Cadaverine potentiation of histamine toxicity in the guinea-pig. Food Cosmetol. T O X ~ C16, O ~ .157-159. Bonilla, M., Enriquez, L. G . ,and McNair, H. M. (1995). “Use of Cold On-Column Injection Technique for the Analysis of Putrescine and Cadaverine by Gas Chromatography.” Pittcon ’95. The Pittsburgh Conference. New Orleans, LA. Castell, C. H., Smith, B., and Neal, W. E. (1971). Production of dimethylamine in muscle of several species of gadoid fish during frozen storage, especially in relation to dark muscle. J. Fish. Res. Board Can. 28, 1-5. Chu, C.-H., and Bjeldanes, L. F. (1981). Effect of diamines, polyamines and tuna fish extracts on the binding of histamine to mucin in vitro. J. Food Sci. 47(1), 79-80, 88. Clifford, M. N., Walker, R., Wright, J., Hardy, R., and Murray, C. K. (1989). Studies with volunteers on the role of histamine in suspected scombrotoxicosis. J. Sci. Food Agric. 47,365-375. Clifford, M. N., Walker, R., Ijomah, P., Wright, J., Murray, C. K., and Hardy, R. (1991). Is there a role for amines other than histamines in the aetiology of sombrotoxicosis? Food Addit. Contamin. 8(5), 641-652. Cohen, S. S. (1971). “Introduction to the Polyamines.” Prentice Hall, Engelwood Cliffs, NJ. Cohen, S. S . (1978). What do the polyamines do? Nature 274,209-210. Craven C., Hilderbrand, K., Kolbe, E., Sylvia, G., Daeschel, M., Gloria, B., and An, H. (1995). “Understanding and Controlling Histamine Formation in Troll-Caught Albacore-Tuna: A Review of Preliminary Findings from the 1994 Season.” Corvallis, Oregon: Extension Sea Grant. Davidek, T., and Davidek, J. (1995). Biogenic amines. In “Natural Toxic Compounds of Foods” (J. Davidek, Ed.), pp. 108-123. CRC Press, Boca Raton, FI. Dawood, A. A., Karkalas, J., Roy, R. N., and Williams, C. S. (1988). The occurrence of nonvolatile amines in chilled-stored rainbow trout (salmo irideus). Food Chem. 27, 33-45. Desiderio, M. A., Davalli, P., and Perin, A. (1987). Simultaneous determination of y-aminobutyric acid and polyamines by high-performance liquid chromatography. J. Chromatogr. 419,285-290. Douglas, W. W. (1980). “The Pharmacological Basis of Therapeutics,” p. 590,5th ed. MacMilIan, New York. Edmunds, W. J., and Eitenmiller, R. R. (1975). Effects of storage time and temperature on histamine content and histidine decarboxylase activity of aquatic species. J. Food. Sci. SO, 516-519. Eitenmiller, R. R., and De Souza, S. C. (1984). Enzymatic mechanisms for amine formation in fish. Chap. 36 in “Seafood Toxins” (Edward P. Ragelis, Ed.). American Chemical Society, Washington, DC.
360
DAFNE D. RAWLES eial.
Equi, A. M., Brown, A. M., Cooper, A., Ner, S. K., Watson, A. B., and Robins, D. J. (1991). Oxidation of putrescine and cadaverine derivatives by diamine oxidases. Tetrahedron 47(3), 507-518. Fardiaz, D., and Markakis, P. (1979). Amines in fermented fish paste. J. Food Sci. 44, 15621563. Farn, G., and Sims, G .G. (1987).Chemical indices of decomposition in tuna. In “Developments in Food Science,” Chap. 15, “Seafood Quality Determination” (D. K. Kramer, and J. Liston, Eds.), pp. 175-183. Elsevier, New York. Feier, U., and Goetsch, P. H. (1993). Inter-laboratory studies on precision characteristics of analytical methods. Determination of biogenic amines in fish and fish products-HPLC method. Arch. Lebensm. 44, 129-152. Fernandez-Salguero, J., and Mackie, I. M. (1987a) Comparative rates of spoilage of fillets and whole fish during storage of haddock (Melanogrammus aeglefinus) and herring (Clupea harengus) as determined by the formation of non-volatile and volatile amines. h i . J. Food Sci. Technol. 22,385-390. Fernandez-Salguero, J., and Mackie, I. M. (1987b). Technical note: Preliminary survey of the content of histamine and other higher amines in some samples of Spanish canned fish. Ini. J. Food Sci. Technol. 22,409-412. Fingerman, M., Nagabhushanam, R., Sarojini, R., and Reddy, P. S. (1994). Biogenic amines in crustaceans: Identification, localization, and roles. J. Crustacean Biol. 14(3), 413-437. Fletcher, G. C., Summers, G., Winchester, R. V., and Wong, R. J. (1995). Histamine and histidine in New Zealand marine fish and shellfish species, particularly Kahawai (Arripis iruita). J. Aquai. Food Prod. Technol. 4(2), 53-74. Food and Drug Administration. (1982). Defect action levels for histamine in tuna: Availability of guide. Fed. Regis. 47(173), 40487-40488. Frank, H. A. (1985). Histamine-forming bacteria in tuna and other marine fish. In “Histamine in Marine Products: Production of Bacteria, Measurement and Prediction of Formation” (B. S. Pan, and D. James, Eds.), pp. 2-3. F A 0 Fisheries Technical Paper 252. Frank, H. A., Yoshinaga, D. H., and Nip, W. K. (1981). Histamine formation and honeycombing during decomposition of skipjack tuna, Kaisuwonus pelamis, at elevated temperatures. Mar. Fish. Rev. 43(10), 9-14. Frank, H. A., Yoshinaga, D. H., and Wu, I. P. (1983). Nomograph for estimating histamine formation in skipjack tuna at elevated temperatures. Mar. Fish. Rev. 45(4-6), 40-44, Frank, H. A., Rosenfeld, M. E., Yoshinaga, D. H., and Nip, W.-K. (1984). Relationship between honeycombing and collagen breakdown in skipjack tuna, Kaisuwonus pelamis. Mr. Fish. Rev. 46, 40-42. Frank, H. A. Baranowski, J. D., Chongiriwatana, M., Brust, P. A., and Premaratne, R. J. (1985). Identification and decarboxylase activities of bacteria isolated from decomposed mahimahi (Coyphaena hippurus) after incubation at 0 and 32°C. Ini. J. Food Microbiol. 2,331-340. Gaget, C., Wolf, E., Heintzelmann, B., and Wagner, J. (1987). Separation of the enantiomers of substituted putrescine and cadaverine analogues by gas chromatography on chiral and achiral stationary phases. J. Chromaiogr. 395,597-608. Gilbert, R. J., Hobbs, G., Murray, C. K., Cruickshank, J. G., and Young, S. E. J. (1980). Scombrotoxic fish poisoning: Features of the first 50 incidents to be reported in Britain (1976-1979). Br. Med. J. 281,71-76. Gouygou, J. P., Sinquin, C., Etienne, M., Landrein, A., and Durand, P. (1992). Quantitative and qualitative determination of biogenic amines in fish. Chap. 15 in “Pelagic Fish-The Resource and Its Exploitation” (J. R. Burt, R. Hardy, and K. J. Whittle, Eds.), pp. 178-186. Fishing News, Oxford.
BIOGENIC AMINES IN FISH AND SHELLFISH
361
Haaland, H., Arnesen, E., and Njaa. L. R. (1990). Amino acid composition of whole mackerel (Scomber scombrus) stored anaerobically at 20°C and at 2°C. Int. J. Food Sci. Technol. 25,82-87. Hall, M., Eldridge, D. B., Saunders, R. D., Fairclough, D. L., and Bateman, Jr., R. C. (1995). A rapid dipstick test for histamine in tuna. Food Biorech. 9(1/2), 39-57. Hardy, R., and Smith, J. G. M. (1976). The storage of mackerel (Scomberscombrus).Development of histamine and rancidity. J. Sci. Food Agric. 27, 595-599. Hatano, H., Sumizu, K., Rokushika, S., and Murakami, F. (1970). Automatic liquid chromatography of primary mono- and diamines on a cation-exchange resin. Anal. Biochem. 35, 377-383. Hill, J. M. (1971). Diamine oxidase (Pea seedling). Merhods Enzymol. B 17, 730-735. Hofle, M. G. (1984). Degradation of putrescine and cadaverine in seawater cultures by marine bacteria. Appl. Environ. Microbiol. 17(4), 843-849. Hollingworth, T. A., Jr., Throm, H. R., and Wekell, M. M. (1987). Determination of ethanol in canned salmon. In “Seafood Quality Determination” (D. E. Kramer and J. Liston, Eds.), Vol. 15, pp. 153-161. Elsevier, New York. Hollingworth, T. A., Jr., Wekell, M. M., Sullivan, J. J., Torkelson, J. D., Jr., and Throm, H. R. (1990). Chemical indicators of decomposition for raw surimi and flaked artificial crab. J. Food Sci. 55(2), 349-352. Hollingworth, T. A., Jr., Kaysner, C. A., Colburn, K. G., Sullivan, J. J., Abeyta, C., Jr., Walker, K. D., Torkelson, J. D., Jr., Throm, H. R., and Wekell, M. M. (1991). Chemical and microbiological analysis of vacuum-paced, pasteurized flaked imitation crab meat. J. Food Sci. 56(1), 164-167. Hughes, M. J., and Potter, M. E. (1991). Scombroid fish poisoning: From pathogenesis to prevention. New Engl. Med. 324(11), 766-768. Hui, J. Y., and Taylor, S. L. (1983). High pressure liquid chromatography determination of putrefactive amines in foods. J . Assoc. Off. Anal. Chem. 66(4), 853-857. Hui, J. Y., and Taylor, S. L. (1985). Inhibition of in vivo histamine metabolism in rats by foodborne and pharmacologic inhibitors of diamine oxidase, histamine N-methyltransferase and monoamine oxidase. Toxicol. Appl. Pharmacol. 81,241-249. Hwang, D.-F.. Chang, S.-H., Shiau, C.-Y., and Cheng, C.-C. (1995). Biogenic amines in the flesh of sailfish (Istiophorus plarypterus) responsible for scombroid poisoning. J. Food Sci. 60(5), 926-928. Ienistea, C, (1971). Bacterial production and destruction of histamine in foods, and food poisoning caused by histamine. Die Nahrung 15,109-113. Ijomah, P., Clifford, M. N., Walker, R., Wright, J., Hardy, R., and Murray, C. K. (1991). The importance of endogenous histamine relative to dietary histamine in the etiology of scombrotoxicosis. Food Addit. Contam. 8(4), 531-542. Ijomah, P., Clifford, M. N., Walker, R., Wright, J., Hardy, R., and Murray, C. K. (1992). Further volunteer studies on scombrotoxicosis. Chap. 17 in “Pelagic Fish: The Resource and Its Exploitation” (J. R. Burt, R. Hardy, and K. J. Whittle, Eds.), pp. 194-199. Fishing News Books, Oxford. Joosten, H. M. L. J. (1988). The biogenic amine contents of Dutchcheese and their toxicological significance. Netherlands Milk Dairy J. 42, 25-42. Jung, H. P. K., and Bjeldanes, L. F. (1979). Effects of cadaverine on histamine transport and metabolism in isolated gut sections of the guinea pig. Food Cosrn. Toxicol. 17,629-632. Kamio, Y., and Nakamura, K. (1987). Putrescine and cadaverine are constituents of peptidoglycane in Veillonella alcalescens and Veillonela parvula. J. Bacteriol. 169(6), 2881-2884. Kamio, Y., Poso, H., Terawaki, Y., and Paulin, L. (1986). Cadaverine covalently linked to an essential constituent of the peptidoglycan necessary for the normal growth in Selenornonas ruminantium. J . Biol. Chem. 261,6585-6589.
362
DAFNE D. RAWLES etal,
Karmas, E., and Mietz, J. L. (1978). Polyamine and histamine content of tuna fish and the relationship to decomposition. Leb. Wissensch. Technol. 11,333-337. Kim, 1.-S., and Bjeldanes, F. I. (1979). Amine content of toxic and wholesome canned tuna fish. J. Food Sci. 44,922-923. Klausen, N. K., and Huss, H. H. (1987). Growth and histamine production by Morganella morganii under various temperature conditions. Int. J. Food Microbiol. 5,147-156. Klausen, N. K., and Lund, E. (1986). Formation of biogenic amines in herring and mackerel. Zeitsch. Leb. Unter. Forsch. 182, 459-463. Kusche, J., Richter, H., Hesterberg, R., Schmidt, J., and Lorenz, W. (1973). Comparison of the 14C-putrescine assay with the NADH test for the determination of diamine oxidase: Description of a standard procedure with a high precision and an improved accuracy. Agents Actions 3(3), 148-156. Lehninger, A. L. (1982). “Principles of Biochemistry,” pp. 172-190. Worth, New York. Lerke, P. A., Werner, S. B., Taylor, S. L., and Guthertz, L. S. (1978). Scombroid poisoning. Report of an outbreak. West. J. Med. 129,381-386. Lerke, P. A., Porcuna, M. N., and Chin, H. B. (1983). Screening test for histamine in fish. J. Food Sci. 48(1), 155-157. Lonberg, E., Movitz, J., and Scorach, S. (1980). Histamine in tuna fish. Var FOda 32,114-123. Lopez-Sabater, E. I., Rodriguez-Jerez, J. J., Roig-Sagues, A. X., and Mora-Ventura, M. T. (1993).Determination of histamine in fish using an enzymic method. Food Addit. Contam. 10(5), 593-602. Lopez-Sabater, E. I., Rodriguez-Jerez, J. J., Roig-Sagues, A. X., and Mora-Ventura, M. A. T. (1994a). Bacteriological quality of tuna fish (Thunnus thymus) destined for canning: Effect of tuna handling on prescence of histidine decarboxylase bacteria and histamine level. J. Food Protec. 57(4), 318-323. Lopez-Sabater, E. I., Rodriguez-Jerez, J. J., Hernandez-Herrero, M., and Mora-Ventura, M. T. (1994b). Evaluation of histidine decarboxylase activity of bacteria isolated from sardine (Sardina pilchardus) by an enzymic method. Lett. Appl. Microbiol. 19,70-75. Luk, G. D., Bayless, T. M., and Baylin, S. B. (1980). Diamine oxidase (Histaminase)-A circulating marker for rat intestinal mucosal maturation and integrity. J. Clin. Invest. 66,66-70. Luong, J. H. T., and Male, K. B. (1992). Development of a new biosensor system for the determination of the hypoxanthine ratio, an indicator of fish freshness. Enzyme Microbiol. TechnoZ. 14, 125-130. Lyons, D. E., Beery, J. T., Lyons, S. A., and Taylor, S. L. (1983). Cadaverine and aminoguanidine potentiate the uptake of histamine in vitro in perfused intestinal segments of rats. Toxicol. Appl. Pharmacol. 70,445-458. Malone, M. H., and Metcalfe, D. D. (1986). Histamine in foods: Its possible role in nonallergic adverse reaction to ingestants. NER AZIergy Proc. 7, 241-245. Manen, C. A., and Russell, D. H. (1973). Polyamines in marine invertebrates. In “Polyamines in Normal and Neoplastic Growth” (D. H. Russell, Ed.), pp. 277-288. Raven Press, New York. Maruta, K., Teradaira, R., Watanabe, N., Nagatsu, T., Asano, M., Yamamoto, K., Matsumoto, T., Shionoya, Y., and Fujita, K. (1989). Simple sensitive assay of polyamines by highperformance liquid chromatography with electrochemical detection after post-column reaction with immobilized polyamine oxidase. Clin. Chem. 35(8), 1694-1696. Merson, M. H., Baine, W. B., Gangarosa, E. J., and Swanson, R. C. (1974). Scombroid Am. Med. Assoc. fish poisoning-Outbreak traced to commercially canned tuna fish. .I. 228(10), 1268-1269.
BIOGENIC AMINES IN FISH AND SHELLFISH
363
Middlebrooks, B. L., Toom, P. M., Douglas, W. L., Harrison, R. E., and McDowell, S. (1988). Effects of storage time and temperature on the microflora and amine development in Spanish mackerel (Scomberomorus maculatus). J. Food. Sci. 53(4), 1024-1029. Mietz, J. L., and Karmas, E. (1977). Chemical quality index of canned tuna as determined by high-pressure liquid chromatography. J. Food Sci. 42, 155-158. Mietz, J. L., and Karmas, E. (1978). Polyamine and histamine content in rockfish, salmon, lobster, and shrimp as an indicator of decomposition. J. Assoc. OffAnal. Chem. 61(1), 139-145. Mitchell, R. G., and Code, C. F. (1954). Effect of diet on urinary excretion of histamine. J. Appl. Physiol. 6, 387-391. Morier-Teissier. E., Drieu, K., and Rips, R. (1988). Determination of polyamines by precolumn derivatization and electrochemical detection. J. Liquid Chromatogr. 11(8), 16271650. Morii, H., C a m , D. C., Taylor, L. Y., and Murray, C. K. (1986). Formation of histamine by luminous bacteria isolated from scombroid fish. Bull. Jpn. SOC.Sci. Fish. 52(12), 2135-2141. Morii, H., Izumi, Y., Kasama, K., and Ishimoto, R. (1994). Factors influencing histamine formation by psychrotrophic luminous bacteria Photobacterium phosphoreum. Nipp. Suis. Gakk. 60(6), 773-777. Morrow, J. D., Margolis, G. R., Rowland, J., and Roberts, L. J. (1991). Evidence that histamine is the causative toxin of scombroid fish poisoning. New Engl. J. Med. 324,716-720. Motil, K. J., and Scrimshaw, N. S. (1979). The role of exogenous histamine in scombroid poisoning. Toxicol. Lett. 3,219-223. Murray, C. K., Hobbs, G., and Gilbert, R. J. (1982). Scombrotoxin and scombrotoxin-like poisoning from canned fish. J. Hygiene 88,215-220. National Fisheries Institute, Inc. (1995). “The NFInsider,” Vol. 3, Issue 32, August 11. Okuyama, T., and Kobayashi, Y. (1961). Determination of diamine oxidase activity by liquid scintillation counting. Arch. Biochem. Biophys. 95, 242-250. Okuzumi, M., Fukumoto, I., and Fujii, T. (1990). Changes in bacterial flora and polyamines contents during storage of horse mackerel meat. Nipp. Suis. Gakk. 56(8), 1307-1312. Omura, Y., Price, R. J.. and Olcott, H. S. (1978). Histamine-forming bacteria isolated from spoiled tuna and jack mackerel. J. Food Sci. 43, 1779-1781. Ramesh, A., Ananthalakshmy, V. K.,and Venugopalan, V. K. (1989). Histamine production in indian oil sardine and indian mackerel by luminous bacteria. MZRCEN J. Appl. Microbiol. Biotechnol. 5,105-108. Redmond, J. W., and Tseng, A. (1979). High-pressure liquid chromatographic determination of putrescine, cadaverine, spermidine and spermine. J. Chromatog. 170,479-481. Ritchie, A. H., and Mackie, I. M. (1980). The formation of diamines and polyamines during the storage of mackerel (Scomber scombrus). In “Advances in Fish Science and Technology” (J. J. Connell, Ed.), pp. 489-494. Fishing News (Books) Ltd., Farnham, U.K. RodriguezJerez, J. J., Grassi, M. A., and Civera, T. (1994). A modification of Lerke enzymic test for histamine quantification. J. Food Prot. 57(11), 1019-1021. Rolle, I., Payer, R., and Soeder, C. J. (1971).Amines in unicellular green algae. Arch. Mikrobiol. 77,185-195. Romette, J. L., Tran, N. D., Durand, P., Boitieux, J. L.,and Navarro, J. L. (1982). “Enzyme Engineering” (I. Chibata, Ed.), Vol. 6. Plenum, New York. Rosier, J., and Van Peteghem, C. (1988). A screening method for the simultaneous determination of putrescine, cadaverine, histamine, spermidine and spermine in fish by means of high pressure liquid chromatography of their 5-dimethylaminonaphthalene-1-sulphonyl derivatives. Zeitsch. Leb. Unters. Forsch. 186, 25-28.
364
DAFNE D. RAWLES etal.
Schulze, K., and Zimmermann, T. (1980). Detection of biogenic amines in canned tuna fish and sardines in oil. Fleischwirtschaft 60,2236-2240. Seiler, N. (1970). Use of the dansyl reaction in biochemical analysis. Methods Biochem. Anal. 18,259-337. Shalaby, A. R. (1995). Multidetection, semiquantitative method for determining biogenic amines in foods. Food Chem. 52,367-372. Shore, P. A. (1971). Analysis of biogenic amines and their related enzymes. In ” Advances in Fish Science and Technology” (J. J. Connell, Ed.), pp. 489-494. Fishing News (Books) Ltd., Farnham, U.K. Sims, G. G . , Farn, G., and York, R. K. (1992). Quality indices for canned skipjack tuna: Correlation of sensory attributes with chemical indices. J. Food Sci. 57(5), 1112-1115. Smith, T. A. (1995). The di- and poly-amine oxidases of higher plants. Eiochem. SOC.Trans. 13,319-321. Smith, J. G. M., Hardy, R., and Young, K. W. (1980). A seasonal study of the storage characteristics of mackerel stored at chill and ambient temperatures. I n “Advances in Fish Science and Technology” (J. J. Connell, Ed.), pp. 372-378. Fishing News (Books) Ltd., Farnham, U.K. Spinelli, A. M., Lakritz, L., and Wasserman, A. E. (1974). Effects of processing on the amine content of pork bellies. J. Agric. Food Chem. 22,1026-1029. Staruszkiewicz, W. F., Jr., and Bond, J. F. (1981). Gas chromatographic determination of Anal. Chem. 64(3), 584-591. cadaverine, putrescine and histamine in foods. J. Assoc. Off: Sumner, S. S., and Taylor, S. L. (1989). Detection method for histamine-producing, dairyrelated bacteria using diamine oxidase and leucocrystal violet. J. Food Prot. 52,105-108. Suzuki. S., Matsui, Y., and Takama, K. (1988).Profilesof polyamine composition in putrefactive Pseudomonas type III/IV. Microbios. Lett. 38,105-109. Tabor, C. W., and Tabor, H. (1976). 1,4-Diaminobutane (putrescine), spermidine, and spermine. Ann. Rev. Biochem. 45,285-306. Tabor, C. W., and Tabor, H. (1985). Polyamines in microorganisms. Microbiol. Rev. 49(1), 81-99. Takagi, M., Iida, A., and Oka, S. (1971). On the formation of unvolatile amines during putrefaction in the muscle of squid and octopus. Bull. Jpn. SOC.Sci. Fish. 37(11), 1079-1083. Taylor, S. L. (1983). “Monograph on Histamine Poisoning.” Paper presented at the Codex Committee on Food Hygiene, 19th session, 26-30 Sept., Washington, DC. Taylor, S. L. (1985). “Histamine Poisoning Associated with Fish, Cheese, and Other Foods,” pp. 1-47. World Health Organization, VPH/FOS/85.1. Taylor, S. L. (1986). Histamine food poisoining: Toxicology and clinical aspects. CRC Crit. Rev. Toxicol. 17,91-117. Taylor, S . L.. and Lieber, E. R. (1979). In vitro inhibition of rat intestinal histaminemetabolizing enzymes. Food Cosmet. Toxicol. 17,237-240. Taylor, S . L., and Speckhard, M. W. (1983). Isolation of histamine-producing bacteria from frozen tuna. Mar. Fish. Rev. 45(4-6), 35-39. Taylor, S. L., and Woychik, N. A. (1982). Simple medium for assessing quantitative production of histamine by Enterobacteriaceae. J. Food Prot. 45, 747-751. Taylor, S. L. and Sumner, S. S. (1986). Determination of histamine, putrescine, and cadaverine. In “Seafood Quality Determination” (D. E. Kramer, and J. Liston, Eds.), Vol. 30, pp. 235-245. Elsevier, Amsterdam. Taylor, S . L., Lieber, E. R., and Leatherwood, M. (1977). A survey of histamine levels in commercially processed scombroid fish products. J. Food Qual. 1, 393-397. Taylor, S. L., Lieber, E. R., and Leatherwood, M. (1978a). A simplified method for histamine analysis in foods. J. Food Sci. 43,247-250.
BIOGENIC AMINES IN FISH AND SHELLFISH
365
Taylor, S . L., Guthertz, L. S., Leatherwood, M., Tillman, F., and Lieber, E. R. (1978b). Histamine production of food-borne bacterial species. J. Food Saf: 1, 173-187. Taylor, S. L., Stratton, J. E., and J. A. Nordlee. (1989). Histamine poisoning (Scombroid fish poisoning): An Allergy-like intoxication. Cfin. Toxicul. 27, 225-240. Walters, M. J. (1984). Decomposition and filth in foods-Determination of histamine in fish by liquid chromatography with post-column reaction and fluorometric detection. J. Assoc. Anal. Chem. 67(6), 1040-1043. Walters, D. (1987). Polyamines: The Cinderellas of cell biology. Biologist 34(2), 73-76. Wang, Y., Enriquez-Ibarra, L. G., Bonilla, M., McNair, H. H., and Flick, G . J. (1994). “Analysis of Biogenic Amines in Finfish.” A Progress Report. Virginia Tech, Blacksburg, VA. Wei, Q.-K., Chen, T.-R., and Fang, C.-W. (1995). Analysis of biogenic amines in fish muscles by high performance liquid chromatography. J. Food Drug Anal. 3(4), 313. Wendakoon, C. N., and Sakaguchi, M. (1992). Effects of spices on growth and biogenic amine formation by bacteria in fish muscle. In “Quality Assurance in the Fish Industry” (H. H. Huss, M. Jakobsen, and J. Liston, Eds.), Vol. 30, pp. 305-313. Elsevier, Amsterdam. Wendakoon, C. N., and Sakaguchi, M. (1993). Combined effect of sodium chloride and clove on growth and biogenic amine formation of Enterobacter aerogenes in Mackerel muscle extract. J. Food Prof. 56(5), 410-413. Wendakoon, C. N., Murata, M., and Sakaguchi, M. (1990). Comparison of non-volatile amine formation between the dark and white muscle of mackerel during storage. Nippon Suis. Gakk. 56(5), 809-818. Wootton, M., Silahali, J., and Wills, R. B. H. (1986). Amine levels in some Asian seafood products. J. Sci. Food Agric. 49, 503-506. Yamamoto, S., Itano, H., Kataoka, H., and Makita, M. (1982). Gas-liquid chromatographic method for analysis of di- and polyamines in foods. J. Agric. Food Chem. 30,435-439. Yamanaka, H. (1989). Changes in polyamines and amino acids in scallop aductor muscle during storage. J. Food Sci. 54(5), 1133-1135. Yamanaka, H., Shimakura, K., Shiomi, K., and Kikuchi, T. (1986). Changes in non-volatile amine contents of the meats of sardine and saury pike during storage. Bull. Jpn. SOC. Sci. Fish. 52(1), 127-130. Yamanaka, H., Shiomi, K., and Kikuchi, T. (1987). Agmatine as a potential index for freshness of common squid (Todurodes pacifcus). J. Food Sci. 52(4), 936-938. Yatsunami. K., Takenaka, T., and Echigo, T. (1994). Changes in the number of halotolerant histamine-forming bacteria and contents of non-volatile amines in meat during processing of fermented sardine with rice-bran. Nipp. Shok. Kogyo Gakk. 41(11), 840-843. Yen, G-C., and Hsieh, C-L. (1991). Simultaneous analysis of biogenic amines in canned fish by HPLC. J. Food Sci. 56(1), 158-160. Yoshinaga, D. H., and Frank, H. A. (1982). Histamine-producing bacteria in decomposing skipjack tuna. (Katsuwonis pelamis). Appl. Environ. Microbiol. 44, 447-452.
This Page Intentionally Left Blank
INDEX
A Absorption, 97-100 Acid coagulation, 165, 166 Acid-curd cheese, 178-193 Acid gel formation, 182-188 Acidification, 166, 174-177,258 Adsorption, 99, 100 Aflatoxin, 286-287 Africa cheese, 167 IM meat, 79-80,89 Air drying, IM meat, 102, 144 Albacore, 339 Americas cheese, 167 IM meat, 80-83 Amines, biogenic, see Biogenic amines Amino acids, see also Biogenic amines cheese, 227,231-234 fish, 336-346 IM meat, 136-137 Amorphous food, 117 Anchovy, 340 Apparent viscosity, 5-6, 25, 274 Appelation d’Origine ContrBIBe, 236 Aroma, see also Flavor cheese, 206, 237 IM meat, 131, 132 Asia fish, 345-346 IM meat, see China; Indonesia; Malaysia Astringency, cheese, 240-241 a,, see Water activity
B Bacteria, see also Contamination; Fermentation; Microbial control; Ripening histamine-forming bacteria, 331-336
lactic acid bacteria, 164-165, 174-175 peptidases, 215,216-218 proteolysis, 211 nonstarter lactic acid bacteria, 198, 199, 211, 219-220,230 future research, 295 quality, 250-251 temperature, 256-257 starter bacteria, 196, 201, 203, 215-225, 250 modified, 258-259 Bande, 80 Basturma, 74-75 Beef, 75,80, 81-82, 83 Biaxial extension, 16-17, 58, 59 Biltong, 79-80, 89, 124 Bingham plastic behavior, 6 Biogenic amines, cheese, 290-292 Biogenic amines, fish and shellfish, 329-365; see also names of amines; names of f i h amine detoxification, 330-331 amine formation, nonvolatile, 330 amine occurrence, 336-346 finfish, 336-344 seafood products, 346 shellfish, 344-345 analytical techniques, 353-358 chromatography, 336,339,340, 341, 353-356 electrophoresis, 357-358 enzymic tests, 356-357 background information, 330 decarboxylase-forming bacteria, 331-336 definition, 330 ecosystem, 336 flavor, 349 freshness indicator, 348-351 367
368
INDEX
Biogenic amines, fish and shellfish (continued) limits, 351-352 quality control, 351-352 quality index, 350 scombrotoxicosis, 337, 346-348 Biopolymeric material, 48-53 Bird-Carreau model, 37-39,49-52, 54 Bitterness cheese, 239-240 fish, 349 Blue cheese contamination, 288 flavor, 206,208 lipOlySiS, 201, 205, 206-207, 208 peptides, 231 proteinases, 223, 224 proteolysis, 226 salting, 177 texture, 254 ultrafiltration, 194 Bologna, 81-82 Boltzmann superposition principle, 10-12 Bound moisture, 98 Brazil, IM meat, 82 Brick cheese, 287 Brie cheese, 199, 200, 224, 288 Browning, 128-129, 132,134, 135, 141, 275 Buffering capacity, 175-176
C Cadaverine cheese, 234 fish, 337-345, 349-356, 358 Calcium sequestration, 263-265, 270 Camembert cheese contamination, 287, 288 lactate, 198, 199, 200 proteinases, 224 ripening, 255 texture, 254 ultrafiltration, 194 Capacola, 79 Capillary, IM meat drying, 104 Capillary tube geometry, 24-26 Carbohydrate nutrition, 278 Caries, 283-285 Case hardening, 109
Casein, 165, 169-173, 175-176 fresh acid-curd cheese, 178, 182-184 hydration, 270 para-, 264, 265-267 proteolysis, 212-214,219-230 Casson model, 34,46 Cathepsin D, 214-215 Cecina, 82-83 Cervelat, 78 Charqui, 82 Cheddar cheese amino acids, 231 citrate, 201 contamination, 289 dental caries, 285 flavor, 209,235, 241-245 lactate, 198, 199 lipolysis, 201, 205, 208, 209 NSLAB, 211,219-220 peptides, 226,227, 237 quality, 250-251, 252 ripening, 255, 256 salting, 177, 178 texture, 176, 254 ultrafiltration, 194 volatile compounds, 238 Cheese, 163-328; see also names of cheeses acid gel formation, 182-188 acidification, 166, 174-177, 258 analytical techniques chromatography, 210, 237-238, 243-247,354-355 electrophoresis, 210, 226, 230-231, 242-243 mass spectrometry, 227, 231, 238 background material, 164-168 consumption, 166, 179 history, 164-165 manufacturing, 166, 168 production, 165-166,260 composition, 180 flavor, 232,233,235-254 analytical methods, 236-254 astringency, 240-241 bitterness, 239-240 defects, 236, 239-242 development, 166, 168, 169 emulsifying salts, 272 fatty acid catabolism, 206, 208-209 fruitiness, 241
INDEX quality factors, 246-254 varietal comparisons, 242-246 future research, 293-296 grading scheme, 252, 295-296 image, 292-293 milk constituents, 168-169 milk conversion to curd, 164-165, 166, 169- 195 fresh acid-curd varieties, 178-193 rennet addition, 188 rennet coagulation, 169-178, 249-250 ultrafiltration, 193-195 nonvolatile compounds, 236-237 nutrition, 277-283 carbohydrate, 278 fat and cholesterol, 278-280 minerals, 280, 282-283 protein, 277 vitamins, 280-281 PH milk conversion, 165, 172, 175-176, 178, 182-183 processed cheese, 265, 270,276 quality, 251-252 texture, 254 processed products, 259-277 blend ingredients, 272-276 classification, 261 cooling, 266-267 emulsifying salts, 263, 267-272 manufacturing, 262-266 structure formation, 266-267 texture, 272-277 quality attributes, 184-193; see also Cheese, flavor; Cheese, texture gelation, 186-188, 191-1 93 rheology. 186 sensory characteristics, 186 syneresis, 174, 185-186, 190-191 quality factors, 246-254 ripening, 166, 178, 195-234 accelerated, 255-259 agents, 196-197, 210-211 amino acids, 227, 231-234 assessment, 209-210 characterization, 226-232 citrate metabolism, 200-201 contamination, 286-288 glycolysis, 196, 198-201, 255 lactose metabolism, 174-177, 198-200
369
lipolysis, 201-209, 255 proteinases, 211-225, 257-258 proteolysis, 176-177, 209-234, 236,255 temperature, 253-254, 256-257 varietal comparisons, 242-246 safety, 283-292 additives, 283 biogenic amines, 290-292 dental caries, 283-285 mycotoxins. 285-290 salting, 166, 177-178 slurries, 259 technology, 166, 168 ultrafiltration, 193-195 texture, 168, 169, 254-255 flavor perception, 236 milk-curd conversion, 176 processed products, 272-277 processing factors, 188-190 volatile compounds, 237-239,244-246 Cheese base, 274 Cheshire cheese, 176, 254 Chicken, 75 China, IM meat production problems, 90 traditional methods, 75-77, 119, 123-124 Cholesterol, 278-280 Chopped meat, 94 Chorizos, 110 Chromatography cheese, 210,237-238,243-247,354-355 fish, 336,339,340,341, 353-356 meat, 354 Chymosin, 170-172, 177, 211-213, 249-250 Citrate, 200-201, 267-269 Coagulation, 165, 166 future research, 293-294 rennet action, 169-178, 197,249-250,293 Coefficient, stress, 13, 23 Coefficient of viscosity, 4-5 Color, IM meat, 127-130 Complex viscosity, 23 Compliance, creep test, 20 Compressive stress, 7-8 Concentrated solutionimelt theory, 37-41 Concentration, kinetic modeling, 141 Concentric cylinder geometry, 31-33 Cone-plate geometry, 27-28 Consistency index, 6, 56 Constitutive models, rheology, 33-44, 62
370
INDEX
Contamination biogenic amines, 290-292,330-358 casein precipitation, 184 control, 119-121 mycotoxins, 285-290 pet food production, 85 process improvement, 89-90 quality factor, 248-249 sources, 164 synergistic stabilization, 149 traditional production, 75, 77, 78, 80, 82 cost drying, 102 energy, 143-147 refrigeration, 74 Cottage cheese, 178, 201 contamination, 292 sludge formation, 184 syneresis, 190, 191 Couette cylinder geometry, 31 Crab, 350,351 Cream cheese, 187, 190, 193 Creaming action, 273, 276-277 Creep test, 20, 43, 56-57 Critical control point, 151 Critical moisture content, 105-106 Cross equation, 34, 46 Crustaceans, 345 Crystallization, 117-118, 135 Cubed meat, 75 Curd, 169-195 buffering capacity, 175-176 fresh acid-curd varieties, 178-193 rennet coagulation, 169-178 rennet inactivation, 197 syneresis, 174 ultrafiltration, 193-195 Curing, meat, see Dehydration control, IM meat Cylinder geometry, rheology, 31-33
D Danish blue cheese, 201 Darcy’s law, 185 Decarboxylase, 331-336 Decarboxylation cheese, 232 fish, 330,331, 336 Defect, flavor, 236, 239-242 Deformation, 3-4; see also Rheology, semiliquid foods
Dehydration, cheese manufacturing, 166 Dehydration control, IM meat, 71-161; see also names of meats absorption, 97-100 bound and unbound water, 97-98 protein denaturation, 100 sorption phenomena, 99-100 vapor pressure, 98 background information, 73-74 dryer selection, 146-147 drying equipment, 145-146, 150-151 energy costs, 143-147 hazard analysis and critical control point, 151 mechanisms, 100-114 cooking, raw meat, 107-108 drying rate curves, 102-105 fermentation, 95, 108-110, 118-119, 148-149 moisture removal, 100-102 osmosis, 110-113 physical changes, 105-107 muscle-meat conversion, 90-97 chopping or grinding, 94 drying rate curves, 102-105 freezing, 93, 148 heating, 95-96 morphological changes, 113-114 myofibers, 92,94-95,96 osmotic treatment, 96-97, 101, 149-150 prerigor processing, 93-94, 147-148 rigor mortis, 91-93 storage, 93 tumbling, 103 pet food production, 83-85, 118, 126 predrying treatment, 94-97, 102-105 preservation, 85-88, 93, 118-119 process optimization, 138-143 production problems, 88-90 quality attributes, 114-138 aroma and flavor, 110,118, 119, 122, 130-136 color, 127-130 microbial control, 119-124 nonheated versus heated meats, 121-124 nonthermodynamic factors, 116-118 nutritive value, 136-138 oxidation control, 133-136 oxygen effects, 121 pH effects, 121, 123, 124, 126, 148
371
INDEX precooking, 126, 130 product acceptability, 118-119 temperature effects, 119-121 texture, 124-126 water activity, 73, 87, 100, 119-127, 139-142 definition, 98 energy costs, 143 equilibrium, 115-116 microbial control, 119, 120-124 process optimization, 139-142 research needs, 147-151 traditional production, 74-83, 145-146 Africa, 79-80, 89 China, 75-77.90, 119, 123-124 contamination, 75, 77, 78, 80, 82, 89 Europe, 77-79, 118 Indonesia, 75, 89, 119, 136 Latin America, 82-83 Malaysia, 75, 89 North America, 80-82 Turkey, 74-75 Dehydration profile, 106 Denaturation, protein, 100 Dending, 75, 89, 119 Dental caries, 283-285 Detoxification, histamine, 330-331 Diffusion, Fickian, 140 Diffusion coefficient, 132, 135, 142 Dilatant behavior, 5, 6 Dilute solution molecular theory, 35-37 Doi-Edwards model, 39-41, 52-53 Dolphin, 337 Dough, 53-55, 58-59 Dried pork floss, 76 Dry heat, 120 Drying, 101-102; see also Dehydration control, IM meat convective heating, 113-1 14 energy costs, 146-147 process optimization, 140-143 rate data, 102-105 research needs, 150-151 theories, 106 Dynamic viscosity, 4-5
E Edam cheese, 243, 245 Eh hurdle, 122-123 Elasticity, 7-8, 9-12 Elastic modulus, 7
Electrical stunning, 125 Electrophoresis cheese, 210,226, 230-231, 242-243 fish, 357-358 Emmental cheese, 177,200, 254 Emulsification, 265-266 Emulsifying salts, 263, 267-272 Emulsion, rheology, 55-58 Endomysial network, 91-92 Energy cost, IM meat production, 143-147 Engineering property, rheology, 44-61 Entanglement, rheology, 48-53 Enzyme, see also Proteolysis cheese curd, 176-177 cheese ripening, 211-225,257-258 future research, 293 milk quality, 249 Enzymic test, 356-357 Equilibrium, water activity, 115-116 Error sources, rheological instrumentation, 26, 28, 33 Esterase, 203-205 Europe cheese production, 166, 167, 193 traditional IM meat, 77-79, 118 Evaporation, 101, 107, 117 Experimental methods, rheology, 12-23 Extensional flow, 14-17,58-59 Extensional viscosity, 16, 17
F Fat, nutrition, 278-280 Fatty acid, 203, 205-209 Fermentation, 95; see also Cheese; Salami; Sausage dehydration mechanisms, 108-110 dehydration quality attributes, 118-119 milk-based foods, 164-165, 174-175 research needs, 148-149 sardines, 346 Feta cheese, 177, 193-194 FFA, see Free fatty acid Fickian diffusion, 140 Finfish, amine occurrence, 336-344 Fish, biogenic amines, see Biogenic amines, fish and shellfish Fish paste, 346 Flavor cheese, see Cheese, flavor fish, 349
372
INDEX
Flavor (continued) IM meat, 110,122, 130-136 milk, 118 Floss, dried pork, 76 Flow index, 6, 56 Food poisoning, 290-292, 337, 346-348 see also Contamination Formulations development, rheology, 59-61 Frankfurter, 108 Free fatty acid, 203, 205-206, 208 Freeze drying, 144, 146 Freezing, IM meat, 93 Fresh acid-curd cheese, 178-193 Freshness indicator, seafood, 348-351 Fromage frais, 178, 179 Fruitiness, cheese, 241 Fruit juice, viscosity, 45-46
G GC, see Chromatography Gelation cheese, 173-174, 182-188,191-193 IM meat, 108, 150 milk, 165 versus precipitation, 183-184 rheology, 41-43, 60-61 sorption, 99 Gel electrophoresis, 242-243 Geometry, rheological instrumentation, 23-33 Glass transition, 117-118, 134-135, 150 Glycolysis, 196, 198-201, 255 Gouda cheese contamination, 289 flavor, 243,245 lipolysis, 201, 205 proteinases, 231 salting, 177 texture, 176, 254 ultrafiltration, 194 Grading scheme, cheese, 252,295-296 Ground meat, 94 Gruyere cheese, 243,254 Guar gum, 34-35 Guar solution. 49-51
H HACCP, see Hazard analysis and critical control point
Haddock, 341-342,349 Ham, 79, 81, 110, 123 Havarti cheese, 194 Hazard analysis and critical control point, 151 Heating, 101-102, 106-107, 187, 262; see nlso Temperature convective, 113-1 14 effects, 95-96 flavor, 131-132 microbial control, 120-121 osmosis, 110-113 preservation, 85-88 process optimization, 138-143 research needs, 149 ripening, 196-197 Hencky strain rate, 15 Herring, 339, 341-343, 348-349 Herschel-Bulkley model, 6, 34, 46 Histamine cheese, 234. 346 defect action level, 351-352 fish, 337,341-344,346, 350-357 Histamine detoxification, 330-331 Histamine-forming bacteria, 331-336 Histamine poisoning, 290, 291, 337, 346-348 Homogenization, cheese, 192 Hookean body, 7-8 Hooke’s law, 7 Horse mackerel, 344 HPLC, see Chromatography Humectant, 84,85,126-127 Hurdle, microbial control, 122-123 Hydrolysis, 271 Hysteresis loop, 9, 57
I IM meat, see Dehydration control, IM meat Indonesia, IM meat, 75, 89, 119, 136 Infusion, 112 Instrumentation, rheology, 23-33 Intermediate-moisture meat, see Dehydration control, IM meat Intrinsic viscosity, 116 Isotropic pressure, 13
J Jerky, 81 Junction zone, polymer, 42
INDEX
K Kahawai, 339 Kelvin model, 9-10 Kilishi, 80 Kinetic modeling, process optimization, 141-143 Kingfish, 339
L LAB, see Lactic acid bacteria Labneh cheese, 190 Lactate metabolism, 198-200 Lactic acid bacteria, 164-165, 174-175; see also Nonstarter lactic acid bacteria peptidases, 215, 216-218 proteolysis, 21 1 Lactose metabolism, 174-177, 198-200 Latin America, IM meat, 82-83 La Zang, 76-77 Lebanon bologna, 81-82 Leonov model, 59 Linear viscoelasticity, 10. 18-23 Lipase, 201 -205 Lipid oxidation, 134 Lipolysis, 201-209, 255 Lipoprotein lipase, 201-202 Loading pattern, 18 Lobster, 355 Loss modulus, 23 Low acid IM meat, 79 LPL, see Lipoprotein lipase Lup Cheong, 76-77
M Maasdam cheese, 243 Mackerel, 337, 341,344, 352 Mahi-mahi, 337 Maillard browning, 128-129, 132, 135 Malaysia, IM meat, 75, 89 Manufacturing technology, cheese, 166, 168, 193-195 Margules equations, 32 Marine ecosystem, 336; see also Biogenic amines, fish and shellfish Mass spectrometry, 227, 231, 238 Mass transfer, 140-141. 151 dehydration mechanisms, 101-102, 107, 110.112
373
Material function, rheology, 23 Mathematical modeling, 139-143, 150 Maxwell model, 9-10, 43, 58 Meat, dehydration, see Dehydration control, IM meat Melt theory, rheology, 37-41 Methyl ketone, 206-208 Mexico, IM meat, 82-83 Microbial control, 119-124.271-272; see also Contamination Milk, see also Cheese cheese quality, 248-249 citrate, 200-201 conversion to cheese, 164-165, 166, 169- 195 crystallization, 117-118 enzymes, 213-215,249. 257-258 fermentation, 164-165, 174-175 heat treatment, 187 lactate, 198-200 lactose, 174-177, 198-200 skim, sorption, 99 viscosity, 44-45 Minerals, nutrition, 280, 282-283 Modulus elastic, 7 loss, 23 molecular models, 35-37 rigidity, 8 shear, 8, 56 storage, 22-23 Young's, 7, 254 Moisture, see also Dehydration control, IM meat cheese, 178,198,251-253 IM meat, 73, 100-102.105-106, 141-142 Molecular model, 35-37 Mortadella, 78 Mozzarella cheese contamination, 287 flavor, 235 proteolysis, 211, 226 ultrafiltration, 194 Muscle-meat conversion, 90-97; see also Dehydration control, IM meat Mycostat, 85 Mycotoxin, 285-290 Myofiber, 92, 94-96, 124, 125
374
INDEX
N
Pastirma, 74-75, 123 PCPs, see Cheese, processed products Newtonian fluid, 4-6, 34 Pelagic fish, 351 Newtonian viscosity, 5 Pemmican, 81 Newtonian zone, 34 Pepperoni, 78 Nisin, 283 Peptidase, 215,216-218,220,223-225 Nitrite, microbial control, 121, 122-123, 124 Peptization, 265 Nonstarter lactic acid bacteria, 198, 199, Perch, 340 211, 219-220,230 Permeability coefficient, 185-186 future research, 295 Pet food production, 83-85, 118, 126 quality, 250-251 PGE, see Pregastric esterase temperature, 256-257 PH Normal stress, 13-14,23 cheese, see Cheese, pH North America, IM meat, 80-82 IM meat, 121, 123, 124, 126, 148 NSLAB, see Nonstarter lactic acid bacteria Phosphates, 269 Nutritive value Plasmin, 213-214 cheese, 277-283 Plate-plate geometry, 29-31 IM meat, 136-138 Polymer junction zone, 42 0 rheological properties, 48-53 octopus, 344 Polyphosphates, 267-269 Off-flavor Polysaccharide dispersion, 46 cheese, 236, 239-242 Pork, 75, 76, 79, 83 fish, 349 energy costs, 145 Oil, viscosity, 45 heat processing, 107 Optimization, IM meat processing, 138-143 microbial control, 123-124 Orthophosphates, 268-269 process techniques, 90 Oscillatory shearing, 20-23 structural alteration, 93 Osmoregulatory capacity, 119 texture, 125 Osmotic treatment, 96-97, 101, 110-113, Porosity, IM meat, 110 149-150 Power law, rheology, 5-6, 34, 46, 56 Oxidation Precipitation, versus gelation, 183-184 IM meat, 133-136 Pregastric esterase, 202 lactate, 199 Prerigor processing, IM meat, 93 Oxygen, microbial control, 121 Preservation, IM meat, 85-88, 93, 118-119 Processed cheese products, 259-277 P Proosdij cheese, 243 Prosciutti, 79 PAGE, see Electrophoresis Protein, see also Casein para-casein, 264, 265-267 gel-forming, 186-1 87 Parmesan cheese nutrition, 277 contamination, 287 surface area-to-volume ratio, 195 flavor, 243,245 Proteinase, 21 1-225, 257-258 lipolysis, 205 Protein denaturation, 100 peptides, 230-231 ripening, 255 Protein efficiency ratio, 137 texture, 254 Protein matrix, 124, 125, 266-267 Pasteurization Protein quality, IM meat, 136-137 cheese products, 259-277 Proteolysis IM meat, 121 cheese, 176-177,209-234,236,255
INDEX fish, 337 IM meat, 131 Pseudoplastic behavior, 5, 6 Pseudoplastic fluid, 56 Putrescine cheese, 234 fish, 337-339,341-346, 349-351, 353-356.358
Q Quality cheese, 184-193, 246-254 fish, 350-352 IM meat, 114-138 Quarg cheese, 178 citrate, 201 syneresis, 190 ultrafiltration, 193
R Rainbow trout, 349 Raoult’s law, 99 Recovery response, creep test, 20 Ree-Eyring equation, 34, 46 Relative humidity, 98, 108, 109; see also Water activity equilibrium, 115-116 Relaxation test, rheology, 19-20, 58 Rennet addition, 188 coagulation, 169-178,249-250,293 inactivation, 197 lipases, 202 substitutes, 172 Reptation, 39-40 Rework, processed cheese, 273 Rheology, cheese, 186, 254-255, 273; see also Cheese, texture Rheology, semiliquid foods, 1-69 background information, 2-3 basic concepts, 3-12 ideal elastic behavior, 7-8 shear rate, 4-6 shear stress, 4-6 time effects, 8-9 viscoelasticity, 9-12 viscous flow, 4-6 constitutive models, 33-44 Bird-Carreau, 37-39.49-52,54
375
Casson, 34, 46 concentrated solutionlmelt theories, 37-41 Cross equation, 34, 46 dilute solution molecular theories, 35-37 Doi-Edwards, 39-41, 52-53 Herschel-Bulkley, 6, 34, 46 Leonov, 59 Maxwell, 9-10, 43, 58 power law, 5-6,34, 46, 56 Ree-Eyring equation, 34, 46 rigid rod, 36-37 solid foods, 41-44 steady shear flow, 33-35 usefulness, 62 definition, 2. 3 engineering properties, 44-61 dough, 53-55,58-59 emulsions, 55-58 entanglement, 48-53 extensional flow, 58-59 formulations development, 59-61 polymers, 48-53 steady shear viscosity, 44-46 yield stress, 6, 46, 48 experimental methods, 12-23 biaxial extension, 16-17, 58, 59 creep test, 20, 43, 56-57 extensional flow, 14-17 linear viscoelasticity, 10, 18-23 sinusoidal oscillatory shearing, 20-23 steady shear measurement, 12-14 stress relaxation test, 19-20, 58 transient flow, 17 uniaxial extension, 14-16 gelation, 41-43, 60-61 instrumentation, 23-33 capillary tube geometry, 24-26 concentric cylinder geometry, 31-33 cone-plate geometry, 27-28 error sources, 26, 28, 33 plate-plate geometry, 29-31 research needs, 62 Rheometer, 23-24,48 capillary, 26, 59 cylinder, 33 Ricotta cheese, 178, 193 Rigidity, modulus, 8 Rigid rod model, 36-37
376
INDEX
Rigor mortis, 91-93 Ripening cheese, see Cheese, ripening IM meat, 108-110, 122, 130-131 Rockfish, 355 Romano cheese, 199,205 Rouse-segmented chain model, 40 RP-HPLC, see Chromatography
S Safety, 283-292; see also Contamination; Food poisoning Sailfish, 352 Salad dressing, 56-58 Salami, 77-78, 85, 118, 122 Salchichon, 83 Salmon, 355 Salt, emulsifying, 263, 267-272 Salting, cheese, 166, 177-178 Salt-in-moisture, cheese, 198, 251-253 Sardine, 339, 346, 352 Sausage, 76-79, 83 energy costs, 146 fermentation, 95, 109-110, 122, 125 preservation, 118-119 production problems, 88-89 Scallop, 345, 349 Scombrotoxicosis, 337, 346-348 Seafood, see Biogenic amines, fish and shellfish Searle cylinder geometry, 31 Semidry meat production, see Dehydration control, IM meat Shaping, cheese, 166 Shear oscillatory, 20-23 steady, 12-14,33-35,44-48 Shear force, 103 Shear history, 59-60 Shear modulus, 8, 56 Shear rate, 4-6, 12-14 rheological instrumentation, 24-33 Shear resistance, 95-96 Shear strain, 10-12 Shear stress, 4-6, 24-33, 28 Shear thickening, 5-6 Shear thinning, 5-6 Shelf life, 90, 118, 164; see also Preservation
Shellfish, biogenic amines, see Biogenic amines, fish and shellfish Shrimp, 346,355 Simple extension, 15 Skipjack tuna, 350, 351, 354 Slice-cured meat, 75 Slippage, rheological instrumentation, 26, 30-31 Slurries, cheese, 259 S/M, see Salt-in-moisture Sobrasada, 79 Solid foods, rheology model, 41-44 Sorption, 99-100,125 Spanish mackerel, 337 Spermidine, 337, 341-343,350,352 analytical techniques, 354-356 Spermine, 337-339,341-343,350,352 analytical techniques, 354-356 Spoilage, see Contamination Spreadability, 275 Squid, 344 Stabilization, IM meat preservation, 85-88 Starch, 60 Starter bacteria, 196, 201, 203, 215-225, 250 modified, 258-259 Steady shear flow, 33-35,44-48 Steady shear measurement, 12-14 Sterilization, IM meat, 86 Sticky point, 117 Storage, IM meat, 93 Storage modulus, 22-23 Strain Hencky rate, 15 oscillatory, 21-22 shear, 10-12 Stress compressive, 7-8 oscillatory, 21-23 shear, 4-6, 24-33, 28 tensile, 7-8, 16 yield, 6, 46, 48 Stress coefficient, 13, 23 Stress difference, 13-14 Stress overshoot, 17 Stress relaxation test, 19-20, 58 Stunning, electrical, 125 Sun-drying, 73-74 Surface area-to-volume ratio, 195 Surimi, 351
377
INDEX
Swiss cheese Ravor, 245 histamine poisoning, 346 lactate, 198, 200 lipolysis, 201. 205 peptidases, 224 peptides, 231 proteolytic agents, 21 1 ripening, 255 texture, 176 Syneresis, 174, 185-186, 190-191
T Taste, see Flavor Temperature, see also Heating creaming action, 276-277 energy costs, 144 Ravor, 131-132 freshness, 349-350 gelation, 192 glass transition, 117-118, 134-135, 150 histamine poisoning, 348 microbial control, 119-121 process optimization, 139, 140 quality factor, 187-188, 253-254 ripening, 253-254,256-257 Tenderness, IM meat. 91 Tensile stress. 7-8, 16 Texture cheese, see Cheese, texture IM meat, 124-126 Texturized cheese, 261 Thickening, shear, 5-6 Thinning, shear, 5-6 Tilsit cheese, 287 TLC, see Chromatography Training, food science, 119 Transient Row. 17 Trout, 349 Tumbling, IM meat, 103 Tuna fish, 337, 338, 350 chromatography, 353-354,355 quality control, 351-352 Turkey (country), IM meat, 74-75 Tyramine poisoning, 290, 291-292
U Ultrafiltration. cheese, 193-195 Unbound moisture, definition, 97-98 Uniaxial extension. 14-16
V Vapor pressure, 97-98 Viscoelasticity, 9-12 biopolymetric materials, 48-53 definition, 9 emulsions, 56-57 linear, 10, 18-23 Viscometer, 26, 30 Viscosity apparent, 5-6, 25, 274 cheese, 189 coefficient, 4-5 complex, 23 definition, 4 dynamic, 4-5 extensional, 16, 17 glass transition, 117 intrinsic, 116 Newtonian, 5 rework, 274 rheological instrumentation, 24-33 selected foods, 44-46 transient, 17 Viscous drag, 24 Viscous flow, 4-6 Vitamins. 280-281
W Water bound and unbound, 97-98,265 diffusion coefficient, 132, 135, 142 holding capacity, 100, 125-126 sausage fermentation, 109-110 vapor pressure, 97-98 Water activity, a, cheese ripening, 178 definition, 98 IM meat, see Dehydration control, IM meat Water adsorption, 100 Water holding capacity, 100, 125-126 Weight loss, IM meat, 101 Wet heat, 120 WHC, see Water holding capacity Wheat dough, 53-55, 58-59
Y Yield stress, 6, 46, 48 Yogurt, 187. 189 Young’s modulus, 7, 254
378
INDEX
2 Zousoon, 76, 137 color, 128-129 energy costs, 145-146, 147 flavor, 132,133,136
microbial control, 124 osmosis, 111-113 prerigor treatment, 93, 147 process improvement, 89-90 texture, 125
This Page Intentionally Left Blank
ISBN 0-12-016437-6