MICROWAVETECHNIQUESAND PROTOCOLS
MICROWAVETECHNIQUES AND PROTOCOLS Edited by
RICHARDT. GIBERSON, MS Ted Pella Inc., Redding, CA
and
RICHARD S. DEMAREEJR., PhD California State University, Chico, CA
HUMANA PRESS TOTOWA,NEWJERSEY
© 2001 Humana Press Inc. 999 Riverview Drive, Suite 208 Totowa, New Jersey 07512 All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording, or otherwise without written permission from the Publisher. All authored papers, comments, opinions, conclusions, or recommendations are those of the author(s), and do not necessarily reflect the views of the publisher. This publication is printed on acid-free paper. ANSI Z39.48-1984 (American Standards Institute) Permanence of Paper for Printed Library Materials. Cover design by: Patricia Cleary. Cover illustration: From Fig. 3A in Chapter 12 "In Vivo Microwave-Assisted Labeling of Allium and Drosophila Nuclei" by Mark A. Sanders and David M. Gartner. Production Editor: Kim Hoather-Potter. For additional copies, pricing for bulk purchases, and/or information about other Humana titles, contact Humana at the above address or at any of the following numbers: Tel.: 973-256-1699; Fax: 973-256-8341; E-mail:
[email protected]; Website: http://humanapress.com
Photocopy Authorization Policy: Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by Humana Press Inc., provided that the base fee of US $10.00 per copy, plus US $00.25 per page, is paid directly to the Copyright Clearance Center at 222 Rosewood Drive, Danvers, MA 01923. For those organizations that have been granted a photocopy license from the CCC, a separate system of payment has been arranged and is acceptable to Humana Press Inc. The fee code for users of the Transactional Reporting Service is [0-89603-903-X/01 $10.00 + $00.25]. Printed in the United States of America. 10 9 8 7 6 5 4 3 2 1 Library of Congress Cataloging-in-Publication Data Microwave techniques and protocols / edited by Richard T. Giberson and Richard S. Demaree, Jr. p. cm. Includes bibliographical references and index. ISBN 0-89603-903-X (alk. paper) 1. Microscopy--Technique. 2. Microwaves. 3. Electron microscopy. 4. Microwave drying. 5. Microwave devices. I. Giberson, Richard T. II. Demaree, Richard S. QH207.M59 2001 570'. 28'25--dc21
2001016992
PREFACE Microwave Techniques and Protocols brings to the clinical and research community a how-to manual based on contributions from laboratories that are using microwaves as a means to facilitate biological sample processing. This project actually began in 1992 when Ted Pella Inc. decided to develop microwave technology for clinical and research applications. A number of questions arose almost immediately: What can microwave technology do to improve sample processing for microscopy applications? How does it work? And, most importantly, why use it? In 1992 there was adequate literature to engender confidence that microwaves could be used to facilitate the following: (1) microwaveassisted stabilization or chemical fixation for light and electron microscopy, (2) enhanced special stain protocols for histology, (3) accelerated decalcification, (4) lower incubation times during immunocytochemistry, and (5) reduced processing times for small tissue biopsies into paraffin. The problem was where to begin when you didn't know anyone in the field. Microwave-assisted chemical fixation presented itself as the place to start to the literature interest at that time and owing to the fact that success or failure could be determined relatively quickly. It took almost three years before publication of our first paper on microwave fixation for electron microscopy (Giberson and Demaree, 1995). By 1995 we still could not locate anyone using the microwave to routinely fix tissue, and based on our experience, we could understand why. Success had come with a lot of effort, but few fundamental answers as to how or why the process worked. Much has changed since that original paper. Protocols now exist for microwave-assisted chemical fixation for both light and electron microscopy that can be done rapidly, reproducibly, and routinely. However, fixation is only a fraction of the time required in overall sample processing. The original 1995 paper was the seed for moving forward and demonstrating that the microwave could be used for each step in processing for electron microscopy (Giberson et al., 1997). That 1997 paper described a four-hour protocol that has since been shortened to two hours (see Chapter 2). A microwave workshop series that began in the summer of 1995, and continues to this day, is the basis for the contributions to this book.
vi
Preface
The contributors to this manual are uniformly from those laboratories routinely using microwave technology to facilitate their processing methods in the various fields of microscopy. The methods and results these authors describe are the tangible evidence that microwaves can be used routinely as the basis for improved sample processing for microscopy applications. These applications include complete sample processing protocols for light and electron microscopy, decalcification, and immunocytochemistry. The overall time savings, ease of use, and quality of results serve as justification for using microwaves in the laboratory. The question as to whether there is a "microwave effect" is alluded to, but not discussed in any great detail. When the term microwave technology is used, it is generic and intended to mean equipment designed for laboratory versus household use. Microwave Techniques and Protocols is designed for anyone with a background and experience in sample processing for immunocytochemistry, decalcification, light microscopy, or electron microscopy, and clearly demonstrates that microwave technology has a place in today's laboratory. Richard T. Giberson, Ms Richard S. Demaree Jr., VhD
REFERENCES Giberson, R.T. and Demaree, R.S., Jr. (1995) Microwave fixation: understanding the variables to achieve rapid reproducible results. Microsc Res Tech 32:246-254. Giberson, R.T., Demaree, R.S., Jr., and Nordhausen, R.W. (1997) Four-hour processing of clinical/diagnostic specimens for electron microscopy using microwave technique. J Vet Diagn Invest 9:61-67.
CONTENTS Preface ..................................................................................................
v
Contributors ........................................................................................
ix
Overview of Microwave-Assisted Tissue Processing for Transmission Electron Microscopy R i c h a r d S. Demaree, Jr. a n d Richard T. Giberson ....................................... 1
Vacuum-Assisted Microwave Processing of Animal Tissues for Electron Microscopy Richard T. Giberson ................................................. 13
Vacuum-Microwave Combination for Processing Plant Tissues for Electron Microscopy William A. Russin a n d Christina L. Trivett ............ 25
Basic Procedure for Electron Microscopy Processing and Staining in Clinical Laboratory Using Microwave Oven R o n a l d L. Austin ....................................................... 37
Specimen Preparation for Thin-Section Electron Microscopy Utilizing Microwave-Assisted Rapid Processing in a Veterinary Diagnostic Laboratory Robert W. N o r d h a u s e n a n d Bradd C. Barr ............ 49
6
Microwave Processing of Archived Pathology Specimens for Ultrastructural Examination Robert J. M u n n a n d Phillip J. Vogt ........................ 67
7
Microwave Fixation of Rat Hippocampal Slices Marcia D. Feinberg, Karen M. Szumowski, and Kristen M. Harris ......................................... 75
Microwave Processing Techniques for Biological Samples in a Service Laboratory Lou A n n Miller ......................................................... 89
°° VII
viii
Contents Microwave-Accelerated Decalcification: Useful Methods for Research and Clinical Laboratories Victoria J. Madden ................................................. 101 10
Microwave Processing of Sediment Samples Dawn Lavoie, Janet Watkins, and Yoko Furukawa .......................................... 123
11
Microwave Polymerization in Thin Layers of London Resin White Allows Selection of Specimens for Immunogold Labeling Jennifer E. Lonsdale, Kent L. McDonald, and Russell L. Jones ......................................... 139
12
In Vivo Microwave-Assisted Labeling of Allium and Drosophila Nuclei Mark A. Sanders and David M. Gartner ............... 155
13
Microwave-Assisted Cytochemistry: Accelerated Visualization of Acetylcholinesterase at Motor Endplates John P. Petrali and Kenneth R. Mills ................... 165
14
Microwave-Assisted Immunoelectron Microscopy of Skin: Localization of Laminin, Type IV Collagen, and Bullous Pemphigoid Antigen John P. Petrali and Kenneth R. Mills ................... 173
15
Microwave Paraffin Techniques for Botanical Tissues Denise Schichnes, Jeffrey A. Nemson, and Steven E. Ruzin .......................................... 181
16
Microwave-Assisted Formalin Fixation of Fresh Tissue: A Comparative Study Richard T. Giberson and Douglas E. Elliott ........ 191
17
Microwave-Assisted Processing of Biological Samples for Scanning Electron Microscopy Richard S. Demaree, Jr. ......................................... 209
Index ................................................................................................
217
CONTRIBUTORS RONALD L. AUSTIN " L S U Medical Center, Shreveport, BRADD C. BARR ° California Animal Health and Food
LA Safety Laboratory, School of Veterinary Medicine, University of California, Davis, CA RICHARD S. DEMAREE, JR. " D e p a r t m e n t of Biological Sciences, California State University, Chico, CA DOUGLAS E. ELLIOTT ° Ted Pella Inc., Redding, CA MARCIA D. FEINBERG " Department of Biology, Boston University, Boston, MA YOKO FURUKAWA • Naval Research Laboratory, Stennis Space Center, MS DAVID M. GARTNER " University of Minnesota, St. Paul, MN RICHARD T. GIBERSON • Ted Pella Inc., Redding, CA KRISTEN M. HARRIS ° Department of Biology, Boston University, Boston, MA RUSSELL L. JONES ° University of California, Berkeley, CA DAWN LAVOIE ° Naval Research Laboratory, Stennis Space Center, MS J E ~ - ~ R E. LONSDALE ° Department of Biology, University of California, San Diego, CA VICTORIA J. MADDEN • Department of Pathology and Laboratory Medicine, University of North Carolina, Chapel Hill, NC KENT L. McDONALD ° Department of Biology, Electron Microscopy Laboratory, University of California, Berkeley, CA Lou ANy MILLER ° Centerfor Microscopic Imaging, College of Veterinary Medicine, University of Illinois, Urbana-Champaign, IL KENNETH R. MILLS • US Army Medical Research Institute of Chemical Defense, Aberdeen Proving Ground, MD ROBERT J. MUNN ° Department of Medical Pathology, University of California, Davis, CA JEFFREY A. NEMSON ° Department of Plant and Microbial Biology, University of California, Berkeley, CA ROBERT W. NORDHAUSEN ° California Animal Health and Food Safety Laboratory, School of Veterinary Medicine, University of California, Davis, CA ix
x
Contributors
Comparative Medicine Division, Aberdeen Proving
JOHN P. PETRALI *
Ground, MD Department of Plant Pathology, University of Wisconsin, Madison, WI STEVEN E. RUZIN Biological Imaging Facility, College of Natural Resources, University of California, Berkeley, CA MARK A. SANDERS • Imaging Center, College of Biological Sciences, University of Minnesota, St. Paul, MN DENISE SCHICHNES • Biological Imaging Facility, College of Natural Resources, University of California, Berkeley, CA KAREN M. SZUMOWSKI • Department of Biology, Boston University, Boston, MA CHRISTINA L. TRIVETT Department of Botany, Weber State University, Ogden, UT PHILLIP J. VOGT * Department of Medical Pathology, University of California, Davis, CA JANET WATKINS GB Tech Inc., Naval Research Laboratory, Stennis Space Center, MS
WILLIAM A. RUSSIN ° •
•
"
1
Overview of Microwave-Assisted Tissue Processing for Transmission Electron Microscopy Richard S. Demaree, Jr. and Richard T. Giberson
INTRODUCTION The purpose of this book is to provide new, reliable, and recently updated protocols for processing many different kinds of samples, using microwave (MW) technology. Chapters included deal with a wide variety of methods and samples, everything from paraffin processing for light microscopy (LM) to scanning and transmission electron microscopy (TEM), to immunocytochemistry. MW-assisted processing of microscopy samples has a long, convoluted history beginning with the pioneering efforts of Mayers (1970) and Zimmerman and Raney (1972). The first LM and TEM report, utilizing MW-enhanced aldehyde fixation of tissues, was by Login and Dvorak (1982). Of the scattered reports that then began appearing in the literature, most reported using MW-assisted processing for the fixation step. Login et al. have chronicled these events in a series of reviews (Login and Dvorak, 1993; 1994; Login et al., 1996). In 1995, Giberson and Demaree reported ice-encased fixation in the MW as a means to control the effects of heating for TEM sample processing. Shortly thereafter, the authors described protocols for rapid, reliable MW polymerization of resin (Demaree et al., 1995) and complete MW tissue processing from fixation to polymerized resin blocks (Giberson and Demaree, 1995). In 1997, an updated, complete From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ
2
Demaree and Giberson
protocol was reported, utilizing microcentrifuge tubes (MT), which no longer required ice for cooling samples (Giberson et al:, 1997). Since 1997, the variety of MW protocols and samples processed has expanded rapidly, e.g., scanning electron microscopy (Fox and Demaree, 1999). Because much of this research is either unpublished or appears in a wide variety of journals, the authors felt that it was time to present a compendium of state-of-the-art MW sample processing. For tissues that require routine processing, Giberson et al. (1997) and Giberson and Demaree (1999) publications are a good starting point. Special adaptations for more difficult tissues are covered in this and subsequent chapters. Protocols for LM (Chapters 12, 15, and 16), scanning electron microscopy (Chapter 17), immunocytochemistry (Chapters 12-15), and decalcification (Chapter 9) are presented. Specific applications, such as clinical pathology (Chapters 4-6) are also covered. The remainder of this chapter presents updated suggestions and typical results from MW-processed tissues for TEM. MATERIALS AND M E T H O D S
Protocol Outline for TEM The steps in the protocol and the times required for each are shown below. The designation "MW" preceding a step, indicates that it is performed in the MW oven. This is an updated protocol based on earlier work (Giberson and Demaree, 1995; Giberson et al., 1997; Giberson and Demaree, 1999). Micrographs presented in the Results came from original work by the authors, or from contributed work of other MW users. A Model 3450 Microwave from Ted Pella was used for all MW processing steps. All the MW accessories used during the various processing steps are available from Ted Pella. During the fixation steps samples are processed in 1.7 mL MTs containing 600 ktL liquid. A Teflon ® holder for the MTs is placed in the established cold spot for MW processing. In step 7, specimens are transferred from the MTs to solvent resistent flowthrough baskets that are placed in 60 × 15 mm polypropylene Petri dishes containing approx 15 mL reagent (Giberson et al., 1997). MW-assisted resin polymerization is done underwater, in BEEM T M capsules sealed with Parafilm ®M, and held in place using a specially designed Teflon rack.
Determination of Sample Location Within MW The creation of a large cold spot, a region of homogenous MW heating, evolved from earlier work (Giberson and Demaree, 1995; Giberson
Overview of MW Processing for TEM et al., 1997). Two water loads (800-mL plastic beakers) are positioned in the MW cavity, so that the neon bulb array, which is 3 in. wide, will fit between the two beakers. The water load on the right is recirculated and cooled by the load cooler. Water-load placement and volume are adjusted until all, or nearly all, of the lights in the array do not light during 10 s MW exposure. These locations are marked on a grid map taped to the oven floor. The uncirculated water load is changed when it becomes hot to the touch (--50°C).
M W Processing Steps for TEM 1. MW aldehyde fixation (tissue in fix > 1 h: may be done under vacuum [vac]: see Chapters 2 and 3). Time: --2 rain. a. Fixative cooled to between 10 and 15°C. b. MW for 10-20-10 (10 s at 100% power; 20 s at 0% power; 10 s at 100% power) (temperature [temp] control is not used during fixation; place the temp probe in the water-load beaker). Note: In small fluid volumes, the temp probe can act as a MW antenna, which will cause excessive heating, compared to the other tubes. This will cause poor and/or unequal fixation between samples. 2. MW aldehyde fixation (tissue in fix <1 h). Time ~7 min. a. Fixative cooled to between 10 and 15°C. b. MW for 40 s at 100% power (temp control is not used during fixation, place the temp probe in the water-load beaker). Note the starting and ending temp of fixative: AT should be 10-15°C. Repeat steps a and b, before going to step c, if the temp change is less than 5°C. c. Repeat steps a and b if tissue has been in fixative <20 min. Note: For difficult-to-fix tissues, steps a and b may be repeated 2-6x. MW exposure time may be extended when using ovens with true variable-wattage control. When using fixatives containing paraformaldehyde (e.g., Karnovsky' s) it is best to use low wattage (350 W or less) for the initial fixation of fresh tissue, followed by a second MW step (b above) at 650 W or more. Otherwise, it is best to let fresh tissue sit in fixative for at least 60 min, before doing steps a and b. Note: MW ovens, from different manufacturers, are available with true variable wattage. 3. Buffer rinse. Time--6 min. a. Remove fixative, add buffer, and immediately replace with fresh buffer. b. Remove second rinse after 5 min.
4
Demaree and Giberson
4. MW Osmium (Os) Fixation. Time:--5 min. a. Cool fixative to 10-15°C. b. MW for 40 s at 100% power (temp control is not used during fixation, place the temp probe in the water-load beaker). Note the starting and ending temp of the fixative: AT should be 10-15°C. Repeat steps a and b before going to step c, if the temp change is less than 5°C. c. Repeat steps a and b for difficult-to-fix tissues. Note: For difficult-to-fix tissues, steps a and b can be repeated 2-6x. 5. Water rinse. Time:--4 min. a. Remove Os. b. Rinse tissue with distilled water. c. Transfer tissue to flowthrough basket, and place in polyproplyene Petri dish containing -- 15 ktL solvent. 6. MW Dehydration. Time:--7 min. a. 1 x 50%, 1 × 70%, 1 x 90%, 2 × 100% (ethanol or acetone). Set temp restriction for 37°C • Temp probe must be in solvent. b. MW each change for 40 s at 100% power. 7. MW Resin infiltration (may be done in vac [see Chapters 2 and 3]). Time:--45 min. a. Set temp restriction for 45°C: temp probe must be in resin, but not in a basket. b. MW 1:1 acetone:resin for 15 min at 100% power. c. MW 100% resin for 15 min at 100% power. d. MW 100% resin for 15 min. at 100% power. Note: When infiltration is done under vac, times are reduced. 8. MW Resin polymerization (Option: use standard oven protocols). Time: --1.25 h. a. Place tissue in fresh 100% resin in B EEM capsules. b. Slightly overfill the capsules with resin. A piece of Parafilm is placed in the capsule cap, before the capsule is sealed, which ensures that no water gets in the capsule during polymerization. c. Place BEEM capsules in a Teflon holder, which is placed in a polypropylene container and covered with --1000 mL water. d. Resins: 1. Epoxies: MW 10 min at 60°C, 10 min at 70°C, 10 min at 80°C, and 45 min at 100°C (all at 100% power). 2. LR White: MW 10 min at 60°C, 10 min at 70°C, 25 min at 80°C (all at 100% power). 3. UnicryVM: MW 10 min at 60°C, 10 min at 70°C, 25 min at 80°C, and --30 min at 90°C (all at 100% power). Total time, all processing steps: --3 h. i
Overview of MW Processing for TEM RESULTS Tissues ready to section, using the above protocol, require approx 3 h total processing time, for up to 12 individual samples. When vac is used during processing (see Chapters 2 and 3), the time can be shortened to approx 2 h. Results from using this protocol, and adaptations to it, are presented below and in subsequent chapters. Figure 1A illustrates rat kidney that has been processed completely in the MW, following the above protocol. In Figure 1B, the only difference was that the Os fixation was done on the bench (outside the MW) for 1 h, then processed the rest of the way in the MW. The preservation of both tissues is good, but there are subtle differences between the two. The most obvious difference is the appearance of the blood plasma in the capillary loops. Figure 1B has a grainy appearance to the plasma, indicating possible extraction and/or protein clumping. The appearance of the plasma in Fig. 1A is significantly more homogenous. That tendency is seen throughout the tissue in each figure. Figure 2 illustrates cytomegalovirus infection diagnosed from an AIDS patient at University of California/Davis Medical Center. The tissue was placed in aldehyde fixative (modified Karnovsky' s) for 24 h prior to MW processing, beginning with step 2. Excellent viral preservation can be seen in Fig. 2 (see insert). Figure 3 is a veterinary diagnostic sample of a Hexamita sp. infection in a turkey jejunium, initially fixed in 10% neutral buffered formalin, and sent to the California Animal Health and Food Safety Laboratory, UC Davis. The tissue was transferred to a modified Karnovsky's fixative and MW processed, using the protocol for tissue in fix > 1 h (step 1 above was the starting point). Parasite preservation was excellent, but overall tissue preservation was poor, because of extraction in the primary fixative (10% neutral buffered formalin). Figure 4 is rat hippocampus that was processed during a 3-d MW workshop held in the laboratory of Dr. Kristen Harris, Children's Hospital, Boston. The tissue was initially perfusion fixed, then processed in the MW. These results indicate excellent preservation of a difficult-to-process tissue. Figure 5 is rat lung, using the MW protocol outlined above for each step in processing. Note the outstanding lipid preservation (arrows) in the alveolar region. Lipid retention in other tissues, both animal and plant, has consistently been better than that achieved using bench protocols (unpublished results; and Giberson and Demaree, 1995).
6
Demaree and Giberson
Fig. 1. (A) MW-assisted processing was used from the glutaraldehyde fixation step through resin polymerization, for the specimen processing of this rat kidney sample. Note the homogeneous appearance of the blood plasma in the capillary loops of the glomerulus, when contrasted to that in (B). Bar - 1.0 ~tm. (B) MW-assisted processing was used for the glutaraldehyde fixation step; 1-h bench fixation was done for the Os fixation step, and the remaining processing, through resin polymerization, was MW-assisted, for this rat kidney sample. Note the flocculant appearance of the blood plasma in the capillary loops of the glomerulus, when contrasted in (A). B a r - 1 . 0 ~tm.
Overview of MW Processing for TEM
Fig. 2. MW-assisted processing was used for all steps after primary fixation. The primary aldehyde fixation was 24 h at room temp in modified Karnovsky' s. A cytomegalovirus-infected cell from the lung of an AIDS patient shows excellent viral morphology (see insert). Bar = 0.5 ~tm; insert: bar = 0.1 ~m. DISCUSSION These results demonstrate the flexibility of MW processing of a broad range of tissues/organisms, in a variety of circumstances. In each case, no two primary fixations were the same. Results were satisfactory in all cases, and processing times were significantly reduced, compared to bench processing protocols used by the different lab. MW processing can be utilized at any or all steps in tissue processing. The real benefit is derived from reduced time at each step. This permits flexibility and efficient time utilization in diagnostic and research settings. Sectioning is still the bottleneck in getting timely results. MW-assisted processing greatly enhances the ability to utilize time more efficiently, while achieving excellent results. Realize that the MW can be used for any step, depending on the needs of the investigator. It is a tool that is beginning to have a significant
8
Demaree and Giberson
Fig. 3. Formalin fixed turkey jejunum was MW-processed, beginning at the aldehyde step, for diagnostic evaluation by TEM. Jejunal cells show extraction characteristic of an initial formalin fixation. The parasites (arrows), Hexamita sp., were well preserved (see insert). Bar = 1.0 ktm; insert: bar = 0.5 ktm.
impact on the way microscopy is done. It is now possible to get diagnostic results from TEM in less than 3 h. This is quicker than any of the LM methods, except frozen sections. There are many recent developments in MW tissue processing, many still unpublished. The volume of fixative and other solvents required is often less than 600 ktL; sometimes, only 5-6 drops of solvent are used (see Chapter 8). This has the benefit of reducing any possible MW-induced heating. In addition, fixatives may be used for repeated microwaving, without changing to fresh fixative, thus reducing the amount of fixative used, and reducing hazardous waste volume. MW processing in vac has now been successfully used to prepare both animal (Chapter 2) and plant tissues (Chapter 3). Tissues that are hard to fix and infiltrate tend to benefit the most from vac-assisted MW processing.
Overview of MW Processing for TEM
Fig. 4. Rat hippocampus after perfusion fixation, was MW-processed from the aldehyde step to resin polymerization. This tissue was processed as part of a MW workshop held at Dr. Kristin Harris' lab, Children' s Hospital, Boston, MA. Tissue preservation was comparable to a 48-h bench protocol used in the lab. Bar = 1.0 gm; Insert - Bar = 0.5 gm.
Caution should be taken when using temp probes during the primary and secondary fixation steps. The probe can act as an antenna, causing rapid heating totally disproportionate to any of the surrounding tubes. If this happens, inadequate fixation will result. The authors' recommendation is to not use temp probes during fixation. Dehydration in the MW is remarkably forgiving. As long as the appropriate time (40 s) is used, adequate dehydration will occur, even if the temp probe is in the wrong location. The only possible problem in dehydration occurs if the temp restriction is too low and the MW does not come on. It is essential that resin infiltration be done very carefully. This step is not forgiving. The temp probe tip must be covered by resin. If it is not, the resin will probably polymerize, and samples will not be able to be transferred to embedding molds. Usually, both the sample and the processing container will be ruined.
10
Demaree and Giberson
Fig. 5. Rat lung was completely processed using the protocol outlined in this chapter. Note the excellent preservation, including lipid (arrows). Bar = 1.0~tm. The use of water immersion for MW resin polymerization (Demaree et al., 1995) removed all variability in heating of different blocks seen in earlier methods (Giammara, 1993). Different resins need to be processed with minor variations, for best results (see step 8 d. above). Personal communication with a number of labs where this technique is used, and several years of firsthand experience, have convinced the authors that MW-polymerized blocks are equal to or often better than, conventional oven-cured blocks. Epoxies are almost the same in trimming and cutting quality, using either MW or oven polymerization. In the authors' experience, LR White benefits most from MW polymerization. It is significantly less brittle when MW-polymerized. Unicryl benefits most in time saved, when using MW polymerization. Typical oven-curing requires 48 h; blocks may be polymerized in 1 h by MW. A number of labs routinely infiltrate epoxy resin mixtures, using the MW then place blocks in a conventional oven overnight for polymerization (see Chapters 5 and 8).
Overview of M W Processing for TEM
11
MW processing of biological tissues for TEM is a reality today. Tissues may be processed much more rapidly, compared to traditional methods (Hayat, 1989). Solvent volumes and hazardous wastes are likewise reduced, when compared to standard processing. MW processing offers flexibility to the user; a single step, multiple steps, or all processing may be done by MW, depending on the situation. MW-assisted processing has these advantages, while offering equal or better preservation, compared to standard processing. REFERENCES Demaree RS Jr, Giberson RT, Smith RL (1995) Routine microwave polymerization of resins for transmission electron microscopy. Scanning 17 (Suppl. 5):25-26. Fox NE, Demaree RS Jr. (1999) Quick bacterial microwave fixation technique for scanning electron microscopy. Microsc Res Tech 46:338-339. Giammara BL (1993) Microwave embedment for light and electron microscopy using epoxy resins, LR White, and other polymers. Scanning 15:82-87. Giberson RT, Demaree RS Jr (1995) Microwave fixation: understanding the variables to achieve rapid reproducible results. Microsc Res Tech 32:246-254. Giberson RT, Demaree RS Jr (1999) Microwave processing technique for electron microscopy: a four-hourprotocol. In: (Hajibagheri N, ed.) Electron Microscopy Methods and Protocols. Humana, Totowa, NJ, pp. 145-158. Giberson RT, Demaree RS Jr, Nordhausen RN (1997) Four hour processing of clinical/ diagnostic specimens for electron microscopy using microwave technique. J Vet Diagn Invest 9:61-67. Hay at MA (1989) Principles and Techniques of Electron Microscopy. Biological Applications, 3rd ed. CRC Press, Boca Raton, FL. Login GR, Dvorak AM (1982) Microwave energy fixation of tissue specimens for light and electron microscopy studies. Acad Gen Dentistry 30:92 (abstract). Login GR, Dvorak AM (1993) Review of rapid microwave fixation technology: its expanding niche in morphologic studies. Scanning 15:58-66. Login GR, Dvorak AM (1994) The Microwave Toolbook. A Practical Guide for Microscopists. Beth Israel Hospital, Boston, MA. Login GR, Tanda N, Dvorak AM (1996) Calibrating and standardizing microwave ovens for microwave-accelerated specimen preparation. Cell Vision 3:172-179. Mayers CP (1970) Histological fixation by microwave heating. J Clin Patho123:273-275. Zimmerman GR, Raney JA (1972) Fast fixation for surgical pathology specimens. Lab Med 3:29-30.
2
Vacuum-Assisted Microwave Processing of Animal Tissues for Electron Microscopy Richard T. Giberson
INTRODUCTION Microwave (MW) technology has found a permanent niche in the world of surgical pathology, biomedical research, and allied disciplines, for a few simple reasons: It can reduce the processing times for tissue samples destined for evaluation by light, scanning, or electron microscopy (EM); it is relatively inexpensive; and it is applicable to a wide range of applications that extend beyond tissue processing. MW-assisted processing of tissue samples, sections, or slides is a new tool in the way the art and science of tissue processing is approached. Understanding of the technology and how to apply it are probably the two most critical factors in determining overall individual success or failure. This report may provide a better understanding of the methods of MW-assisted processing, and introduce a new methodology for tissues being processed for evaluation by EM. The first reported use of a MW in tissue processing was by Mayer (1970), which described MW heating as a direct means of tissue stabilization. Since that first report, MW stabilization has become known as a process in which tissue samples, either fresh or taken from formalin, are placed in a solution (typically, normal saline), and heated to final solution temperatures (temps) between 45 and 70°C in the MW (Leong, 1991, 1994; Hopwood, 1993; Kok and Boon, 1992).
From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ
13
14
Giberson
MW-assisted chemical fixation was the next process to evolve, (first reported in the late 1970's [Login, 1978]), and it has become a method for rapid fixation of tissue being processed for EM (Login and Dvorak, 1988, 1993; Login et al., 1990; Benhamou et al., 1991; Giberson and Demaree, 1995). Rapid fixation was followed by the first report (to the author's knowledge) of MW energy being used for epoxy resin polymerization (Giammara, 1985). By the mid 1990' s, MW technology was gaining some converts in the EM field, as a method to facilitate the fixation process. Tissue processing for EM, however, is a time-consuming process that routinely requires 36-48 h to complete. The fixation steps (aldehyde followed by osmium tetroxide [OsO4]) usually account for <10% of the total time. Boon et al. (1986) had reported on MW-assisted dehydration, clearing, and paraffin infiltration for histology. The author et al. applied a similar approach to the remaining steps (dehydration; resin infiltration; resin polymerization) of tissue processing for MW (Giberson et al., 1995, 1997). A unified MW-assisted processing protocol evolved from this work. The time required to process fresh tissue to a polymerized resin block was effectively reduced from 36-48 h to <4 h. During this period, the combining of vacuum (vac) and microwave irradiation (MWI) was reported on as a histoprocessing technique (Kok and Boon, 1994, 1996; Kov~ics et al., 1996). Chapter 3 describes the original vac MW-assisted processing of botanical tissues for EM. This work describes vac MW-assisted processing of animal tissues for EM. Vac is applied during the fixation and resin infiltration steps in the MW. This protocol was designed around commercially available equipment (Ted Pella, Redding, CA), and has reduced tissue-processing times for EM to 2 h. MATERIALS A N D M E T H O D S
Equipment and Materials Required The MW used was a Model 3451 Microwave Processor (Ted Pella) which comes equipped with the following: a water-load recirculation device (load cooler), variable power controller (supplies continuous power output from 250 to 750 W), accessories for tissue processing and MW oven calibration, and a temp-restrictive (TR) temp probe to control sample temp maximums during MWI. A vac chamber (Ted Pella, no. 3435) was used for the vac MW steps. Standard fixatives (2% glutaraldehyde in 0.1 M sodium cacodylate, pH 7.2-7.3; 2% aqueous OSO4), dehydrating agents (50, 70, 90, and 100% acetone), and resins (Eponate 12, Ted Pella) were used. Rat liver, kidney, and lung were the tissues processed.
Vac-Assisted MW Processing for EM
15
Processing Protocol 1. MW oven calibration. The first step in processing is the determination of hot and cold spots within the MW cavity (MW calibration) (Giberson and Demaree, 1995; Login et al., 1996). This process is necessary to ensure uniform MW heating of the microcentrifuge tubes (MTs) during the fixation steps. A neon bulb array is normally used to identify hot and cold spots; however, it is too large to fit into the vac chamber. Two 100-mL water loads are positioned inside the vac chamber, as well as an additional load (1-L plastic beaker containing --800 gL water) positioned behind the vac chamber (Fig. 1). This load is recirculated and cooled to a constant temp during MWI. A trial fixation run (without samples) is done (see step 2 below) with the aldehyde fixative, and the same number of MTs that will be used during the actual fixation run. Temp maximums among all tubes should be within 3-4°C of each other after MWI. If not, increase water-load volumes, and/or reposition the vac chamber relative to the 1-L water load (see Fig. 1). 2. Aldehyde fixation. The vac chamber with water loads is positioned as shown in Fig. 1. The MTs, containing the samples, are placed in a Teflon ® holder (Fig. 2), and filled with 600 + 100 gL fixative. Fixative temp should be <20°C prior to MWI. The MW is programmed for the following time sequence: 1 rain. 0% power; 40 s 100% power at 650 W; 3 rain, 0% power. The holder is placed in the vac chamber, as shown in Fig. 1, and a vac of 20" Hg is drawn. The programmed time sequence is initiated. 3. Buffer rinse. Remove the holder with MTs from the vac chamber, and replace the fixative with buffer. There are two buffer changes during a 5-rain period outside the MW. 4. OsO 4 fixation. The same series of steps used during the aldehyde fixation are repeated for Os fixation. 5. Acetone dehydration. The tissue samples are transferred to baskets, for dehydration and resin infiltration. 1-7 baskets can be placed in the 55 mm diameter polypropylene Petri dish (Ted Pella). The Petri dish is filled with-~ 15 mL 50% acetone prior to tissue transfer from the MTs. A second water load is placed in the MW cavity, prior to starting MW-assisted dehydration. This is done to create a cold spot that can be checked using the neon bulb array. The samples are placed in the MW cavity, and the temp probe is placed in the acetone (Fig. 3). The continuous power output of the MW is set for 650 W, MWI at 40 s, and the temp restriction at 37°C, for each dehydration step. The following graded series is run: 1 × 50%; 1 x 70%; 1 × 90%; 2 × 100%. 6. Resin infiltration. The vac chamber is returned to the MW chamber without the two 100-mL water loads, and the water load added for the dehydration steps is removed. The Petri dish, containing 100% resin, is
16
Giberson
Microwave
Cavity
.8-1
L
x w
I \
v
k~ \
u
-
/
,
/
T II
I I_..7ZI:s2SL I
,
Iq Q P O
, II
/ 1
|
s
4
I
• '
~]m~N
f
/
I
/
H 0
a , ° .
~
°o°O°°Co f
J I
,
ooo9/
A
17 m M
: N G
I
\
~ \
o
\ IN
Vacuum
Chamber
.°
A ¢m.Oel.m
lul~l., m
A Iq~! IIn
Fig. 1. The location of the MTs (A) inside the vac chamber relative to the recirculated water load (.8-1 L) in the MW cavity and the two water loads (100 mL each) placed inside the vac chamber. placed in the vac chamber, and the temp probe is positioned in the resin. The lid to the vac chamber has a special port to allow the temp probe to monitor and control solution temps while under vac. The continuous power output of the MW is set for 650 W, MWI at 2 min, and the TR at 43°C. The following series is run: 3 × 100% resin. 8. Resin polymerization. This is done underwater, in a 1-L container, with BEEM TM capsules, according to Giberson et al., (1997). BEEM capsules, containing resin and sample, are capped with a piece ofParafilm TM in the lid. The capsules are placed in a Teflon holder, which is placed in a Rubbermaid TM ( no. 5) 940-mL container, which is filled with water (Fig. 4). The original recirculated water load (-800 mL) is left in the MW during polymerization. The continuous power output of the MW is set for 750 W, MWI for 75 min, and the TR is raised from 60, 70, and 80°C at 10-min intervals, for the first 30 min, then is raised to > 100°C for the remaining 45 min.
Vac-Assisted M W Processing for EM
17
i Sample
Rxative
Fig. 2. The MT holder is MW-transparent (i.e., does not heat), and can hold up to 14 sample tubes. The fixative vol (600 + 100 ktL) and tissue samples are depicted in two representative MTs placed in the holder.
Fig. 3. A 55-mm polyethylene Petri dish, containing two tissue-processing baskets is shown relative to the water load placements in the MW cavity. The TR temp probe tip is placed into the acetone in the Petri dish, to control temp maximums during MWI.
18
Giberson
4.5" I
! I
kv)
~
,
~
.
~
ObliqueView
f,,,¢ ater Level ~ 1/4-112"Above Capsule
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
LJ'. .
.
.
.
.
.
.
.
.
.
.
!
j
Fig. 4. BEEM capsules (or similar) containing resin and sample are capped with a piece of Parafilm in the lid, placed in a Teflon holder, which is centered in a Rubbermaid (no. 5) 940-mL container. The container is filled with tap water, as shown, and the TR temp probe placed in water to control temp maximums during the first 30 min of MW-assisted polymerization.
Comments on the Processing Protocol Each step in the MW is an independent event, which means that a step done on the bench can be followed by a step or steps in the MW, or visa versa. Although the buffer rinse is not routinely a MW-assisted step, because of its short duration, it can be done in the MW as a 40-s step identical to the parameters outlined for dehydration. For difficult-to-fix tissues, the best results come from using a low, continuous power setting (i.e., 250 W) for an extended time period (5-45 min) (unpublished results and personal communication), using the temp probe to control the temp maximum of the fixative. When using paraformaldehyde as the fixative, or in combination with glutaraldehyde, the 40-s MW step is done at 250 W. The 3-min 0% power period is then followed by 10 s at 650 W (see Chapter 16 for details regarding the influence of MW radiation on formalin). Polymerized resin blocks were sectioned, picked up on 200-mesh copper grids, poststained with uranyl acetate, and lead citrate, and viewed by a Philips T/EM 400 transmission EM. Photography was done by routine darkroom methods.
Vac-Assisted MW Processing for EM
19
RESULTS The three tissues processed by vac MW-assisted processing were rat lung, liver, and kidney. The samples were processed together, and polymerized resin blocks were ready for sectioning within 2 h, as measured from the aldehyde fixation step. Rat lung (Fig. 5A,B) demonstrates good preservation of the phospholipid in the lamellar bodies of the type II aveolar epithelial cells. Ribosomes are abundant and distinct, and mitochondrial density is good. Rat liver is shown in Fig. 6A,B. The cytoplasmic density is uniform, there is no extraction evident, rough endoplasmic reticulum is well-preserved and abundant; and the mitochondrial density and structures are well-preserved. Rat kidney is shown in Fig. 7A,B. Structural preservation of the glomerulus is good, and Bowman's capsule is evident in Fig. 7A, as are capillary loops. In Fig. 7B, good mitochondrial preservation is evident, as well as that of the basement membrane, capillary endothelium, and podocytes. Microtubules are evident in the podocytes (Fig. 7B). The quality of preservation, resin infiltration, and cutting properties were equal to, or better than, that achieved using routine bench processing protocols requiring 36-48 h to complete (results not shown). Electron beam stability of the MW-cured resin is excellent, as are the post-staining properties (uranyl acetate followed by lead citrate). DISCUSSION Vac MW-assisted processing of tissue for EM introduces tissue-processing flexibility that has not existed before (Giberson et al., 1995, Giberson, 1997). Each step (aldehyde fixation; OsO4 fixation, graded dehydration series, resin infiltration) can be completed in <10 min for the majority of tissues the author has tried to date, including: Chlamydomonas cells, yeast with cell walls, nematodes, and botanical tissues (unpublished results). MW-assisted resin polymerization is the time consuming step in the process, requiring 75 min for epoxy resins and 40 min for LR White (Giberson et al., 1995, 1997; Demaree et al., 1995). The polymerization step differs from other published methods (Giammara, 1985, 1993) in that polymerization is equal among all specimen blocks, does not require any MW oven calibration, and can be done as a routine procedure. The number of different samples that can be processed at one time is up to the individual. The author has done as many as 50 at one time, after the fixation steps, and prefers to fix no more than eight samples at one
20
Giberson
Fig. 5. (A) A low-magnification image of a type II aveolar epithelial cell. (B) A higher-magnification image of a type II aveolar epithelial cell, which demonstrates cytoplasmic detail (ribosomes, endoplasmic reticulum) and density. The phospholipid (arrow) in the lamellar bodies is well-preserved. time, however, after fixation, the number of samples being processed simultaneously is not a major issue, one way or the other. When a rapid answer is desired or required, it is most convenient to use MW-assisted
Vac-Assisted MW Processing for EM
21
Fig. 6. (A and B) are low and higher magnifications of the typical preservation seen with vac MW-assisted processing. In both figures, the cytoplasmic density of the hepatocytes is uniformly dense, and the mitochondrial structure is welldefined. Rough endoplasmic reticulum is easy to distinguish. resin polymerization for only those samples that one intends to section that day. The rest go to the convection oven for overnight polymerization. Even in workshop situations, in which a number of individuals are learning the technique for the first time, polymerized resin blocks are
22
Giberson ~ !iJ~~~:.~:i i~i ~:.
~i,!~
'
........':'~ii~........
~:
:::..: .~i:~ • :~.... :~: .,.............................~:,.
::::::i+'::'i :~::~i::::'':~:+ .............. :: ....... ;~ :: :::::+~:~::::,
,~,, ~ ~
.~!'~ ,~
~.......i ~:~
ii!i .... m i
!i,~,i~:i~,
Fig. 7 (A and B) are low and higher magnifications of the typical preservation seen in the glomerular region. A is typical of the outer regions of the glomerulus next to Bowman' s capsule (arrow). B is typical of the preservation seen in the capillary loops. Note the microtubules (arrow) in the podocyte. obtained by the participants routinely in 3.5-4 h (unpublished results from 18+ national workshops). Vac MW-assisted processing of tissue specimens for EM results in significant time savings, compared to any existing methodology. The author would argue that the quality of the
Vac-Assisted MW Processing for EM
23
results (Figs. 5A, B and 7A,B) are comparable to those obtained from routine bench, or other MW-assisted, protocols (Giberson et al., 1995, 1997). The implementation of vac MW-assisted processing for EM samples results in time saving of >30%, compared to previous MW-assisted protocols (Giberson et al., 1995, 1997), and to well over 90%, compared to bench protocols. REFERENCES Benhamou R, Noel S, Grenier J, Asselin A (1991) Microwave energy fixation of plant tissue; an alternative approach that provides excellent preservation of ultrastructure and antigenicity. J Electron Microsc Tech 17:81-94. Boon ME, Kok LP, Ouwerkerk-Noordam E (1986) Microwave-stimulated diffusion for fast processing of tissue: Reduced dehydrating, clearing, and impregnating times. Histopathology 10:303-309. Demaree RS, Jr, Giberson RT, Smith RL (1995) Routine microwave polymerization of resins for transmission electron microscopy. Scanning 17(Suppl.V):26. Giammara B (1985) Ultra rapid epoxy embedment using microwave energy. Electron Microsc EMSA Bull 43:706-707. Giammara B (1993) Microwave embedment for light and electron microscopy using epoxy resins, LR White, and other polymers. Scanning 15:82-87. Giberson RT, Demaree RS, Jr (1995) Microwave fixation: Understanding the variables to achieve rapid reproducible results. Microsc Res Tech 32:246-254. Giberson RT, Demaree RS, Jr, Nordhausen RW (1997) Four-hour processing of clinical/ diagnostic specimens for electron microscopy. J Vet Diagn Invest 9:61-67. Hopwood D (1993) Microwaves and tissue processing. USA Microsc Anal 1:23-25. Kok JP, Boon ME (1992) Microwave CookBook for Microscopists, 3rded. Coulomb, Leydon. Kok LP, Boon ME (1994) Nonchemical dehydration of fixed tissue combining microwaves and vacuum. Eur J Morpho132:86-94. Kok LP, Boon ME (1996) New developments of microwave technology in pathology: Combining vacuum with microwave irradiation. Cell Vision 3:224. Kovfics L, Szende B, Elek G, Lapis K, Horvfith O, Hiszek I, Tamfisi A, Schmidt O (1996) Working experience with a new vacuum-accelerated microwave histoprocessor. J Pathol 180:106-110. Leong AS-Y (1991) Microwave fixation and rapid processing in a large throughput histopathology laboratory. Pathology 23:271-273. Leong AS-Y (1994) Microwave technology for morphological analysis. Cell Vision 1:278-288. Login GR (1978) Microwave fixation versus formalin fixation of surgical and autopsy tissue. Am J Med Techno144:435-437. Login GR, Dvorak AM (1988) Microwave fixation provides excellent preservation of tissue, cells and antigens for light and electron microscopy. Histochem J20:373-387. Login GR, Dwyer BK, Dvorak AM (1990) Rapid primary microwave-osmium fixation. I. Preservation of structure for electron microscopy in seconds. J Histochem Cytochem 38:755-762. Login GR, Dvorak AM (1993) Review of rapid microwave fixation technology: Its expanding niche in morphologic studies. Scanning 15:58-66. Mayers CP (1970) Histological fixation by microwave heating. J Clin Patho123:273-275.
3
Vacuum-Microwave Combination for Processing Plant Tissues for Electron Microscopy William A. Russin and Christina L. Trivett
INTRODUCTION Preparation of botanical specimens for electron microscopy (EM) is hampered by several physical factors unique to plant cells. Impediments to penetration and action of various reagents in the tissue preparation process include the vacuole, plastids, cell wall (CW), and intercellular air space (Roland, 1978; O'Brien and McCully, 1981; Roland and Vian, 1991; Ruzin, 1999). The vacuole is a membrane-bound, fluid-filled sac that typically occupies more than 80% of the cell volume (Fig. 1). The plastids are double-membrane-bound organelles, which, like the vacuole, occupy a large percentage of cell volume (Fig. 1). Both of these cellular components contain an appreciable volume of fluid and may also contain large concentrations of solutes. Release of their contents during processing can affect fixation, by locally diluting the fixative solution, shifting the pH of processing solutions, or increasing the solute concentration in the vicinity of the tissue. CWs and their associated modifications are the first barriers that must be breached before fixation of cytoplasmic components can occur. Many plant cells retain thin, mostly cellulosic primary CWs (Fig. 1). However, plant organs often contain large areas of cells that have highly thickened primary (or secondary) CWs that may be modified chemically by the introduction of various constituents, such as cutin, suberin, and lignin. These CW materials are hydrophobic, and function in limiting the From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ
25
26
Russin and Trivett
Fig. 1. Cross-section of vac-MW processed soybean (Glycine max) petiole showing portion of vascular bundle in the region of xylem. CWs have different thicknesses in different cells; arrows labeled "CW" indicate thin cell walls; "CW" (without arrow) indicates the thick CW in tracheary elements. Note the large volume of the vacuole (V) and of plastids (P). Bar = 1 ~tm.
amount of evaporative water loss from the plant body. Both increased wall thickness and chemical modification can limit the penetration of fixatives and other processing solutions. Abundant intercellular spaces, which develop between plant cells, have a marked effect on tissue preparation protocols. Intercellular spaces are typically filled with air, which can be retained during processing, despite tissue dicing and the addition of surfactants (e.g., Tween-20) to
Vac-MW Plant-Tissue Processing for EM
27
processing solutions. Such spaces form another barrier to fixative penetration. A routine method for removing intercellular air is to apply a vacuum (vac) to diced tissues in fixative, then return the tissues to atmospheric pressure (O'Brien and McCully, 1981; Ruzin, 1999). This results in the replacement of intercellular air with fixative solution, thereby bringing fixative into contact with the cells. To completely remove intercellular air, it is often necessary to perform further processing steps under vac. Application of vac during processing has been shown to enhance fixative penetration, and thereby increase quality of the fixation image. Another method being employed more frequently to enhance and accelerate the fixation process is the application of microwave (MW) energy. Plant materials fixed in the presence of MW energy have shown excellent structural preservation at the light microscope and the transmission electron microscope (TEM) level (Walsh et al., 1989; Heumann, 1992; Giberson and Demaree, 1995). MW-processed material also has exhibited greater antigenicity in tissues used for immunolocalization (Benhamou et al., 1991), and greater probe specificity for in situ hybridization studies (Schichnes et al., 1999). In fact, MW processing has proved so useful that it has become a standard tissue-processing protocol in the authors' labs. For microautoradiographic studies of metabolite transport, tissue samples that are labeled are usually cryoimmobilized, then freeze-dried (FD) (lyophilized). FD is a dehydration technique that has been shown to limit the diffusion of water-soluble compounds (Eschrich and Fritz, 1972; Fritz, 1980; Fritz et al., 1983, 1989). Subsequently, the samples are either infiltrated with a mixture of diethyl ether and Spurr's resin (1:1), or are pressure-infiltrated directly with full Spurr's resin (Vogelmann and Dickson, 1982). Unfortunately, FD CWs are resistant to infiltration. Even when FD plant tissue is pressure-infiltrated, the epoxy resin often adheres poorly to the CWs so that sections disintegrate as they are cut (Vogelmann and Dickson, 1982). The most successful method of resin infiltration involves an initial vac treatment, followed by pressure infiltration (Vogelmann and Dickson, 1982). For a preliminary microautoradiographic study, the authors decided to couple the advantages of processing acceleration offered by MW treatment with the complete infiltration that vac treatment provides, to achieve well-fixed, well-infiltrated tissues with minimal diffusion artifacts. The precedent for combining vac treatment with MW processing was established for histological applications (Kok and Boon, 1995,
28
Russin and Trivett
1996). In those articles, no details about the chamber and design were given, and the extension of the technique to tissue processing for EM was not addressed. During the development of the method to process our FD material, the authors formulated a vac-MW fixation protocol for plant tissue preparation that has found general utility in these labs. This chapter describes the construction of the original vac chamber and a vac-MW fixation schedule that can be used with a wide variety of plant material. MATERIALS A N D M E T H O D S
Design and Construction of the Vac Chamber The vac-MW chambers that are commercially available have many useful features (e.g., modular design, improved viewing area, temperature [temp] probe port), which are lacking from the chamber described here. Description of the authors' chamber is included for those workers who either have a limited budget, or wish to modify a chamber for their own needs. The first prototype vac chamber was made by modifying a Scienceware Vacuum Desiccator (Bel-Art Products, Wayne, NJ) (Fig. 2A). The neoprene gasket included with the desiccator was not MW-proof, and was replaced with a silicon gasket (McMaster-Carr, Chicago, IL). The groove that held the gasket was enlarged slightly to better fit the gasket. A liberal application of Dow Coming High Vacuum Grease (Dow Coming, Midland, MI) to the top of the desiccator helped to seal the gasket. The exterior air-bubbler port of a Pelco TM 3450 Laboratory Microwave Processor (Ted Pella, Redding, CA) was connected by a heavy-duty vac hose to a standard rotary vac pump, and, on the inside, to the inlet on the desiccator assembly, by means of the neoprene hose supplied with the MW.
MW Calibration: Cold-Spot localization Before each tissue-processing procedure, the MW was calibrated and water loads placed to create a cold spot, as previously described (Giberson and Demaree, 1999). Two 1-L water loads in tri-cornered beakers were used during each fixation. One was temp controlled by a Pelco TM 3420 Microwave Load Cooler (Ted Pella) set to a temp limit of 40°C. The other was not controlled, but was changed frequently during the procedure, especially when it felt hot. The placement of water loads and vac chamber varied minimally among fixations (Fig. 2C).
Vac-MW Plant-Tissue Processing for EM
29
Basic MW Processing for Plant Tissues Preparation methods were modified from established protocols (Giberson and Demaree, 1995, 1999; Giberson et al., 1997; Bozzola and Russell, 1992). Tissue samples are diced in fixative (the authors typically use 2.5 % glutaraldehyde, 2.5 % paraformaldehyde, 0.05 % Tween20 in 50 mM sodium cacodylate buffer), then transferred into 600 ktL fixative in 1.5-mL microcentrifuge tubes (MTs), which are placed in a Teflon ® rack, then positioned in a vac chamber that is packed with ice (Fig. 2B). MTs must be left uncapped. If a temp probe will be used, it should be inserted into a dummy tube. For all MW steps, indicated below with an *, two water loads are used as described above (cold spot localization). The vac chamber is placed between the two water loads (Fig. 2C). All steps below are performed under vac (--0.25 mm Hg), with the MW operating at full power (unless otherwise noted). All solution changes should be performed in a fume hood. 1. Vacuum pretreatment: Place tissue in MTs with fixative, into ice-packed vac chamber (Fig. 2B), apply vac grease to chamber gasket, close lid, and put under vac (5 min). 2. Primary fixation*: Keeping the tissues under vac, MW-process the tissue (10 s, 20 s at 0% power, 10 s). 3. Rinse*: Release the vac. When chamber is at atmospheric pressure, open lid. Remove fixative, and replace with buffer at the appropriate molarity. Close chamber, put under vac; MW-process tissue (45 s). Repeat rinse steps for a total of three rinses. 4. Secondary fixation*: Release vac, open lid. Remove final buffer change, replace with 2% osmium tetroxide (aqueous). Close chamber, put under vac, and MW-process tissue (45 s). 5. Rinse*: Rinse as above. 6. Dehydrate*: With a release and reapplication of vac for each step, take the tissues through dehydration in 10, 30, 50, 70, 95, 100% acetone (45 s, each step). The tissues should be processed in each of the acetone concentrations twice. 7. Resin infiltration: For final embedding in Spurr's resin, the tissues should be passed through mixtures of 1:2 resin:acetone, 2:1 resin:acetone, and full resin (2.5 min each step). Vac can be applied at each step, if desired. Each change should be repeated for a total of two changes. 8. Embedding and polymerization: The author's standard method for embedding is to place all the tissue from a particular sample into an aluminum weighing boat (Fisher Scientific, Hampton, NH), then to polymerize the resin in an oven set to 70°C for at least 8 h, or overnight.
30
Russin and Trivett
Vac-MW Plant-Tissue Processing for EM
31
The authors have performed polymerization in the MW according to the published methods, using BEEM TM capsules (Giberson et al., 1997; Giberson and Demaree, 1999), with excellent results. RESULTS Results obtained by using the vac-MW method yielded well-infiltrated, easily sectioned material (Fig. 3A-D). The material also showed excellent ultrastructural preservation, judged on the basis ofultrastructural characteristics published in several standard references (O'Brien and McCully, 1981; Ruzin, 1999; Giberson et al., 1997; Bozzola and Russell, 1992; Bowers and Maser, 1988). These characteristics include the lack of structural artifacts (e.g., no swelling, shrinkage, or plasmolysis) (Fig. 3A,C,D), structure or organelle dissolution (e.g., mitochondria have a finely granular matrix and are not swollen) (Fig. 3C,D), and/or modification of staining characteristics of cellular components (e.g., evenness in cytoplasmic density, with no areas of clumping) (Ruzin, 1999; Giberson et al., 1997; Bozzola and Russell, 1992; Bowers and Maser, 1988). Tissues that are well fixed also show intact membranes, Cytoskeletal components are sometimes difficult to fix and image. The abovementioned method has yielded tissues showing intact, clearly defined microtubules (Fig. 3B). Phloem is a plant tissue that is often difficult to fix, at least in part because of the differences in solute concentrations among the various component cells (Russin and Evert, 1985; Beebe and Evert, 1992). It is especially difficult to avoid plasmolysis (shrinkage of the protoplast) in sieve-tube members, the defining cell type of higher plant phloem, because of their elongated nature and their high osmotic potential. The authors' vac-MW method has consistently given fixation of phloem tissues with all the above characteristics, in addition to a general lack of plasmolysis in the sieve-tube members (Fig. 3D). Persimmon endosperm, specifically, posed of cells that have thick primary CWs that resist infiltration (Fig. 3A). Despite the employment of many different techniques by a number of Fig. 2. (opposite page) (A) First prototype for vac chamber modified from Scienceware vac desiccator. Original neoprene gasket was not MW-proof, and was replaced by a silicone one. (B) Specimens arranged for processing in second prototype chamber manufactured from nylon stock by Ted Pella. MTs, containing tissue specimens and fixative, are placed in a Teflon holder, and packed with ice. MTs must remain open. (C) Standard setup for tissue processing. The vac chamber (VC), located in the cold spot generated by two water loads, one with temp control (C) by a load cooler, and one without temp control (U). Also visible are the temp probe and the vac line inserted into the VC.
32
Russin and Trivett +~!:++.:::::+++++i+
: . ......... +~..+
.....
A
Fig. 3. Diverse plant specimens that were processed using the vac-MW protocol described. (A) Portion of thick primary CW from persimmon (Diospyros) endo-sperm, a particularly difficult tissue to process. Ultrastructure is well-preserved; there is no plasmolysis. Darts indicate partial profiles of plasmodesmata, showing evidence of membranous substructure. Bar= 1 ~tm. (B) Glancing section of ph-loem parenchyma cell from squash (Cucurbita pepo) stem cell, just under the cell wall (CW), showing well-preserved microtubules (unlabeled darts). Cyto-plasm appears dense and well-preserved. Bar = 1 ~tm. (C) Section of fruit wall of red pepper (Capsicumfructescens). Note the smooth vacuolar membrane (membrane bounding vacuole, V) and dense cytoplasm. Plastid (P), mitochondria (M), and lipid droplets (unlabeled darts) are particularly well-preserved, and show little, if any, distortion or extraction. Little or no extraction of pigment (electrondense bodies within plastid) evident in plastid ultrastructure, and internal plastid membranes also well preserved. Bar = 1 ~tm. (I)) Portion of minor vein from Arabidopsis thaliana leaf in the region of phloem. No plasmolysis is evident, even in sieve-tube members (ST). Various cell types exhibit well=preserved vacuoles (V), plastids (P), and mitochondria (M). Bar = 1 ~tm.
Vac-MW Plant-Tissue Processing for EM
33
different workers, it had been impossible to produce adequately fixed and infiltrated material. Application of vac-MW processing has allowed preparation of well-fixed, well-infiltrated persimmon endosperm for examination with TEM (Fig. 3A). DISCUSSION MW processing methods already have been combined with other preparative techniques in specific applications, e.g., MW-freezing techniques for use in surgical pathology, histochemistry, and embryology (Boon and Kok, 1987; Login and Dvorak, 1994). A previous vac-MW combination for pathology (Kok and Boon, 1995, 1996) exploited the decreased boiling temp of solutions that occur in a vac, to enhance the infiltration steps in a paraffin technique. Processing of large tissue specimens in a vac was intended to expedite the evaporation of liquid molecules (ethanol or isopropanol) at lower temps/pressures. The vac-MW combination resulted in improved dehydration, clearing, and infiltration of the solvents used in a paraffin technique for light microscopy (Kok and Boon, 1995). As an additional advantage, the method allowed simultaneous processing of a large number of tissue samples (Kok and Boon, 1995; Leong and Leong, 1997). The authors' vac-MW method has yielded results that are at least as good as, and in many cases superior to, our standard chemical fixation protocols for TEM. In many cases, the quality of fixation obtained with the vac-MW technique was superior. This vac-MW process is not limited to standard fixations of typical plant materials, and can be modified to fit the purposes of each particular study. At any point, from second rinses to polymerization, the tissues can be processed at room temp by standard TEM tissue-preparation methods (e.g., Giberson and Demaree, 1999; Giberson et al., 1997; Bozzola and Russell, 1992), or by using other published MW embedding protocols. As an example, for plant materials with very thick CWs or copious intercellular air, the initial vac pretreatment can be repeated as often as necessary. In addition, both the primary and secondary fixation steps may be repeated as necessary, for materials that have high water content, or that may be difficult to fix for some other reason. Most standard methods of tissue fixation for postembedding immunochemical localizations recommend avoiding osmium tetroxide as a secondary fixative (Bozzola and Russell, 1992; Erickson et al., 1993). The authors have prepared tissues for immunochemical studies by eliminating the secondary fixation step and the second set of rinses, and proceeding directly to dehydration. LR White resin is often the
34
Russin and Trivett
embedding resin of choice for immunochemistry, and can be used in the vac-MW procedure. To dehydrate the tissue prior to infiltration with LR White resin, the authors have used the same schedule and concentrations as in the above dehydration, simply substituting ethanol for acetone. Since LR White has low viscosity, no need was found to do a graded infiltration, as with Spurr's resin. FD tissues to be used in microautoradiographic studies are notoriously difficult to infiltrate and section (Eschrich and Fritz, 1972; Fritz, 1980; Fritz et al., 1983, 1989; Vogelmann and Dickson; 1982). Application of the above method has yielded tissues that are well infiltrated and easily sectioned. Furthermore, FD tissues prepared by the authors' method are far less likely to separate from the embedding medium, as is common with other methods. To adapt the vac-MW method, all steps prior to the last infiltration step have been eliminated, which reduces the potential leaching of water-soluble and organic-soluble compounds, as much as possible (unpublished data). As described, this vac-MW technique is easily adaptable to a number of different studies with diverse types of plant material. The fixation obtained with the vac-MW technique was at least equal in quality to specimens processed by traditional chemical fixation methods. An additional benefit of applying the vac-MW technique is a reduction in processing times, especially during the fixation steps; traditional chemical fixation typically proceeds for 3 h (Russin and Evert, 1985; Beebe and Evert, 1992), compared to about 8 min for vac-MW processing. Preliminary evidence indicates that the vac-MW method may allow some reduction in tissue-processing times, even compared to MW processing alone. Considering the above evidence, the combination vac-MW processing technique is applicable to a wide variety of studies, and can provide highquality, well-fixed, and -embedded plant tissues, in much shorter times. REFERENCES Beebe DU, Evert RF (1992) Photoassimilate pathway(s) and phloem loading in the leaf of Moricandia arvensis (L.) DC. (Brassicaceae). Int J Plant Sci 153:61-77. Benhamou N, Noel S, Grenier J, Asselin A (1991) Microwave energy fixation of plant tissue: an alternative approach that provides excellent preservation of ultrastructure and antigenicity. J Electron Microsc Tech 17:81-94. Boon MA, Kok LP (1987) Microwave Cookbook of Pathology. The Art of Microscopic Visualization. Coulomb, Leiden, The Netherlands. Bowers B, Maser M (1988) Artifacts in fixation for transmission electron microscopy. In Crang RFE, Klomparens KL, eds. Artifacts in Biological Electron Microscopy, Plenum, New York, pp 13-42. Bozzola JJ, Russell LD (1992) Electron Microscopy: Principles and Techniques for Biologists. Jones and Bartlett, Boston, MA.
V a c - M W Plant-Tissue Processing for EM
35
Erickson PA, Lewis GP, Fisher SK (1993) Postembedding immunocytochemical techniques for light and electron microscopy. Methods Cell Bio137:283-310. Eschrich W, Fritz E (1972) Microautoradiography of water-soluble organic compounds. In Luttge U, ed. Microautoradiography and Electron Probe Analysis: Their Application to Plant Physiology Springer-Verlag, Berlin, pp. 99-122. Fritz E (1980) Microautoradiographic localization of assimilates in phloem: problems and new method. Ber Deutsch Bot Ges 93:109-121. Fritz E, Evert RF, Heyser W (1983) Microautoradiographic studies of phloem loading and transport in the leaf of Zea mays L. Planta 159:193-206. Fritz E, Evert RF, Nasse H (1989) Loading and transport of assimilates in different maize leaf bundles: digital image analysis of 14-C microautoradiographs. Planta 178:1-9. Giberson RT, Demaree RS (1995) Microwave fixation: understanding the variables to achieve rapid reproducible results. Microsc Res Tech 32:246-254. Giberson RT, Demaree RS (1999) Microwave processing techniques for electron microscopy: a four-hour protocol. In Hajibagheri N, ed. Methods in Molecular Biology, Humana, Totowa, NJ, pp. 145-158. Giberson RT, Demaree RS, Nordhausen RW (1997) Four-hour processing of clinical/ diagnostic specimens for electron microscopy using microwave technique. J Vet Diagn Invest 9:61-67. Heumann H-G (1992) Microwave-stimulated glutaraldehyde and osmium tetroxide fixation of plant tissue: ultrastructural preservation in seconds. Histochemistry 97:341-347. Kok, L. P., and Boon, M. E. (1995) Ultrarapid vacuum-microwave histoprocessing. Histochem. J. 27, 411-419. Kok LP, Boon ME (1996) New developments of microwave technology in pathology: combining vacuum with microwave irradiation. Cell Vision 3:224. Leong AS-Y, Leong FJ (1997) Principles, applications and protocols of microwave technology for morphological analysis. In Gu J, ed.AnalyticalMorphology: Theory, Applications and Protocols Eaton, Natick, MA, pp. 69-90. Login GR, Dvorak AM (1994) Methods of microwave fixation for microscopy. A review of research and clinical applications: 1970-1992. Progr Histochem Cytochem 27:1-127. O'Brien TP, McCully ME (1981) The Study of Plant Structure: Principles and Selected Methods. Termacarphi, Melbourne, Australia. Roland J-C (1978) General preparation and staining of thin sections. In Hall JL, ed. Electron Microscopy and Cytochemistry of Plant Cells. Elsevier/North Holland, Amsterdam, The Netherlands, pp. 1-62. Roland J-C, Vian B (1991) General preparation and staining of thin sections. In Hall JL, Hawes C, eds. Electron Microscopy of Plant Cells. Academic, San Diego, CA, pp. 1-66. Schichnes D, Nemson J, Sohlberg L, Ruzin SE (1999) Microwave protocols for paraffin microtechnique and in situ localization in plants. Microsc Microanal 4:491-496. Russin WA, Evert RF (1985) Studies on the leaf of Populus deltoides (Salicaceae): ultrastructure, plasmodesmatal frequency, and solute concentration. Am J Bot 72:1232-1247. Ruzin SE (1999) Plant Microtechnique and Microscopy Oxford University Press, New York, NY. Vogelmann TC, Dickson RE (1982) Microautoradiography of water-soluble compounds in plant tissue after freeze-drying and pressure infiltration with epoxy resin. Plant Physio170:606-609. Walsh GE, Bohannon PM, Wessinger-Duvall PB (1989) Microwave irradiation for rapid killing and fixing of plant tissue. Can J Bot 67:1272-1274.
4
Basic Procedure for Electron Microscopy Processing and Staining in Clinical Laboratory Using Microwave Oven RonaM L. Austin
INTRODUCTION The Ted Pella Microwave Oven Model 3450, has been used in this electron microscopy (EM) laboratory for the past 3 yr. In that time, over 750 specimens have been processed in the oven. It has reduced turnaround time from as long as 70 h to as little as 6 h (this time period includes sectioning and staining times). Processing has been the principal consumer of time, accounting for roughly 75% of the turnaround time. The reduction in processing time gives the microscopist an enormous advantage for diagnostic pathology, in terms of money saved, clients served, and allowing more freedom to attend to other important duties for both technician and doctor. The doctors of this lab are pathologists. Renal, muscle, nerve, and tumor biopsies are the primary tissues processed here, which means that the time it takes to return a diagnosis to the clinician can be essential to the patient care. The procedure seen in this chapter is not dramatically different from the protocol found in Giberson et al., 1997; however, modifications were made to accommodate a busy clinical EM facility. This procedure clearly demonstrates the practicality of microwave (MW) technology for an EM clinical service. From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ 37
38
Austin
The basic MW, load cooler, materials, methods, and staining procedures are all that are necessary to produce the results shown in this chapter. The method for EM processing and staining described is simple and reliable. The protocol offers dependable reproducible results. MATERIALS A N D M E T H O D S
Processing Tissue is routinely received at the EM lab in 3% glutaraldehyde with 0.1 M sodium cacodylate as a buffer, pH 7.3. The tissue fixation usually ranges from 1 h to 24 h, as a consequence of shipping time. After proper accessioning and grossing, the tissue is placed in a 2.0-mL microcentrifuge tube (MT), and filled to the 0.5 mL mark, then placed in a Teflon ®rack. The rack holding the MTs is placed in a predetermined sample area (Fig. 1). Before any MW procedure can be performed, a cool and safe area must be prepared, using water load, with the aid of an alphanumeric grid (Fig. 2) positioned on the bottom of the MW with a neon bulb array (Giberson et al., 1995). This neon array is used to locate the energy field inside the MW cavity, while it is running. Wherever a bulb lights up, a hot spot exists. The water loads are arranged on the grids, to eliminate these spots. The MW (in this lab) comes with water-recirculating load cooler, which maintains a constant water temperature (temp) of 39-40°C. The load coolers of the newer models have an adjustable temp setting. A 400-mL tri-pour beaker is used as the reservoir for the load cooler. Additional 200- and 400-mL water loads are needed to secure a suitable sample area (Fig. 2). Once this is done, the rack is positioned in the safe area, and the temp probe is placed in the dummy tube approx 2 mm below the surface of the fluid, for accurate temp measurement (Fig. 1). A 10-20-10:(10 s at 100% power, 20 s at 0% power, 10 s at 100% power) time sequence, without temperature restriction (TR) is used after the initial fixation (Giberson et al., 1997), to enhance tissue preservation. The tissue is then rinsed in 0.1 M Na cacodylate buffer 5x. Sural nerve is the only exception: It seems to need a longer time in the MW fixation cycle. Otherwise, infiltration problems occur. Three 40-s periods at a 37°C TR setting are necessary for sufficient fixation. The nerve is cooled (in an ice bath) to about 9 or 10°C, before and between each 40-s irradiation period. The ice-bath temp can be measured (in a dummy tube of glutaraldehyde) with the MW temp probe (Fig. 3). The secondary fixation (1% osmium tetroxide [OsO4] in 0.1 M Na cacodylate buffer, pH 7.3) is repeated as in the nerve fixation, but only two 40-s periods, at a 37°C setting, are necessary. After the fixation phase, the specimens are placed in flowthrough baskets in the bottom
39
EM Processing in Clinical Lab Using MW i~,,i 7 ¸ ............................... ~iiiii!iiiii~!
Fig. 2
Fig. 1
Fig.4
Fig.3 .
.
.
.
)...~.~),~
Fig.5
Fig.6
Fig. 1. MW-transparent Teflon rack with MTs and temperature probe placement. Fig. 2. Alphanumeric grid and neon array in relation to the water loads. Fig. 3. Ice bath showing positioning of the temp probe in the MT. Fig. 4.60-mm polypropylene plastic Petri dish, showing basket and temp probe positions. Fig. 5. Illustration of the damage that can happen if the temp probe is not correctly used. Fig. 6. Plastic water container, with the embedding capsule rack submerged in the water bath. half of a 60 mm polypropylene plastic Petri dish (Fig. 4). The specimens are then dehydrated, infiltrated, and polymerized. Table 1 (Giberson et al., 1997) summarizes this protocol. Three resins have been used in the MW: t-Epon, Maraglass 655, and Araldite 502. Araldite 502, which is
40
Austin Table 1 MW Processing Times and Temp Settings
Process
Time
3% Gluaraldehyde fixation a Fixation of sural nerve
Buffer rinse 1% OsO 4 in cacodylate buffer Rinse with 0.1 M cacodylate buffer Ethanol dehydration
Infiltration b cd
Polymerization ' •
Total time
•
Temperature
10-20-10 3 x 40 s each. Tissue must be cooled to 10°C or less between fixation period. 5× Brief rinses 2x 40 s each
No TR 37oc
2x Brief rinses
Outside the oven
50% 2× 40 s 70% 2x 40 s 95% 2x 40 s 100% 3x 40 s 1:1 resin and acetone 3x 10 min each. 100% resin, 3x 10 min 90 min without temp probe• Approx 4 h
45°C 45°C 45°C 45°C 50°C
Outside the oven 37°C
50°C No TR
a This author finds that a longer fixation time is necessary for sural nerve, to ensure complete fixation for proper infiltration and staining of the tissue. b Ice should be added to the non-circulation water loads in this phase, to ensure a cool working area for the length of time required for infiltration. c Use the caps from the standard truncated pyramid tip 00 BEEM capsule, they are deeper than the caps on the fiat-bottomed embedding capsules, and snap into place more securely.
d In the polymerization phase of this process, the technician can place a specimen in a conventional 60°C oven for overnight curing. This can save 2 h from processing time, freeing the individual for other tasks, if same day tumaround is unnecessary.
simple to prepare, has shown the best M W qualities, in terms of infiltration, polymerization, sectioning, and staining. It should be noted that care must be taken to make sure the temp probe is placed in the resin before the M W is started. Otherwise, serious damage will result (Fig. 5). It is the polymerization phase of the protocol that really contradicts all previous teaching about how EM processing must be done. The specimens are transferred to polyethylene embedding capsules (size 00), with a Parafilm ® liner pressed into the cap prior to closure. This creates a watertight seal (Giberson et al., 1997). The capsules are placed in a
EM Processing in Clinical Lab Using MW
41
Teflon holder that is then submerged under 900-1000 mL tap water in a plastic container (Fig. 6). Noncirculated water loads are removed, but the circulated load-cooler reservoir is maintained throughout the polymerization process. The technician should continue monitoring and maintaining proper water level. A 90-min polymerization period works best with Araldite 502. At the end of the 90-min polymerization stage, the capsules are removed, and are ready for thick and thin sectioning. The total time is roughly 6 h to the EM. In this lab, the pathologists make their diagnosis on the scope.
Staining This protocol was developed in this lab, out of necessity. The method used prior to obtaining the Ted Pella MW took as long as 20-25 min. The MW procedure takes only about 3 min for a given specimen, as exemplified in Tables 2 and 3. Staining with lead citrate (LC) and uranyl acetate (UA) can be done in the MW, on a routine basis. The combination that seems to work best is aqueous LC, made from a variation of Sato's (1967) technique, and 5% UA in 100% methanol (aq UA was tried without success). The variation is as follows: 1 g lead nitrate, 1 g LC, and 1 g lead acetate added, in order, to 91 mL millipored distilled water containing 9 mL standardized 0.2 N carbonate-free Na hydroxide (from Electron Microscopy Sciences). Stir well for a couple of hours, then allow to stand overnight. Decant the clear liquid into an airtight container (15-mL glass screwtop centrifuge tubes). Centrifuge at maximum setting in a clinical centrifuge for 5 min. This clear solution should stain very well, and last for--3-4 mo. Reynolds LC (Reynolds, 1963) stains as well, but only lasts --3-4 wk. The safest and most efficient way of staining in the MW is to cut previously unused silicone rubber flat embedding molds (Ted Pella, no. 10505) into 2-3 rows, and place in a 100 × 25 mm glass Petri dish (Fig. 7A,B). A glass cover is placed over the Petri dish to prevent splattering of the stains into the MW. Although temp does not seem to be a great concern in this procedure, experimentation was done to establish the average maximum temp of the two solutions. This temp was 30°C to 35°C. Generally, a TR setting of 35°C will work nicely, with the probe placed in the 400-mL noncirculating water load. RESULTS Figures 8 and 9 are a membranous glomerulonephritis. The difference between the two specimens is that Fig. 9 was processed in paraffin,
42
Austin Table 2 M W Staining Times for UA and LC
Stain
Time/Temp
5% UA in 100% methanol
1 min at 30-35°C
LC
1 min at 30°C-35°C (variation on Sato, 1967)
Total time
2 min to stain
Rinse times 20 dips 2x in 20 mL 100% methanol. 20 dips 1× in 20 mL millipored water. Blot dry with Whatman filter. 20 dips 1x in 20 mL 0.1 N NaOH. 20 dips 3x in 20 mL millipored water. Blot dry with Whatman filter. Approx 3 min to rinse.
Table 3 Conventional Staining Times for UA and LC
Stain
Time/Temp
5% UA in 100% methanol
10 min at room temp
LC
8 min at room temp (variation on Sato, 1967)
Total time
18 min to stain
Rinse time 20 dips 2x in 20 mL 100% methanol. 20 dips 1x in 20 mL millipored water. Blot dry with Whatman filter. 20 dips 1x in 20 mL 0.1 N NaOH. 20 dips 3x in 20 mL millipored water. Blot dry with Whatman filter. Approx 3 min to rinse.
before it was reprocessed in resin. Figure 8 is a well-developed membranous nephropathy with extensive subepithelial deposits (d), with welldefined basement membrane (bm). Figure 9 is also a well-developed membranous nephropathy with subepithelial deposits and clearly visible granular endoplasmic reticulium (ER) (arrowheads). Figure 9 was fixed in the same manner as resin-processed kidney. The paraffin temp averaged 65°C. The deparaffinization was simply a reverse of the paraffinization technique. Little difference can be seen in the subepithelial deposits, in terms of preservation. Figures 10 and 11 are, in essence, normal skeletal muscle and sural nerve. Figure 10 illustrates well preserved A, H, M, and Z bands, with the myofilaments appearing uniformly bound. Figure 11 is essentially a normal sural nerve, with slight depletion in the myelination (M), but no marked segmental nerve loss is identified. Both images show good EM processing by the MW. Figure 12 is a renal neoplasm with abundant mitochondria consistent with oncocytoma. One cell has an oval nuclei (Nu) with finely dispersed chromatin with occasional dilated rough ER (small arrows). Cell membranes are illustrated with arrowheads.
EM Processing in Clinical Lab Using MW
43
A
B
Fig. 7. (A) Two or three grids can be placed in each well. Place the Petri dish in the sample area, and maintain water loads as in the processing phase. (B) Petri dish in relation to the water loads.
Figure 13 is respiratory cilia (C) in origin. There is the usual array of 9 + 2 within the cilia. The inner and outer dynein arms are present (arrowheads). Basilar apparatus inside the cytoplasm of the cells is within normal limits.
Fig. 8. Membraneous glomerulonephritis; extensive subepithelial deposits (d), basement membrane (bm). Bar = 1 l,tm. Fig. 9. Membraneous glomerulonephritis, taken from paraffin, with subepithelial deposits (d), rough ER (arrowheads). Bar = 1 ~tm.
EM Processing in Clinical Lab Using MW
45
Fig. 10. Normal skeletal muscle with well preserved A, H, M, and Z bands. Bar = 0.5 ~tm. Fig. 11. Myelin sheath of myelinzed nerve (M), large granulated vesicles (V), microtubules (N). Bar - 0.5 ~tm.
46
Austin
Fig. 12. Mitochondria (M), nuclei (Nu), rough ER (small arrows), and cell membranes (arrowheads). Bar = 1 gm. Fig. 13. Cilia (C), inner and outer dynein arms (arrowheads). Bar = 0.1 gm.
DISCUSSION A technician can now start processing wet tissue at 8 AM, and have a completed stained section in an E M by 2 or 3 PM of the same day. As
EM Processing in Clinical Lab Using MW
47
many as six specimens can be processed in a single 60-mm Petri dish, and, with proper cold spots carefully selected, even more specimens can be processed with the use of a larger dish. The amount of materials used, in some cases, has been cut by more than one-half. One example is Os, one of the most expensive materials in the processing phase. Before implementation of the oven, the cost was about $300-400 every 8-9 mo, for a supply of 10 1-g ampules of Os. In 3 yr of using the MW oven, this lab has purchased only two supplies of Os. The quantity of Os used per specimen before was about 5 cc; now, it is 0.5 cc/specimen. In the past, processing time was a limiting factor that kept EM in the dark ages. Improvements were made to ultramicrotomes, with microglade knife stages and fiber optics for specimen and knife illumination, making ultrathin sectioning less stressful and less time-consuming. EMs were integrated with computers, modems, and digital imaging systems, to allow real-time examination of specimens and the rapid production of micrographic images. This virtually eliminated costly and chemically hazardous darkroom facilities. Until just a few years ago, the slow, laborious, and expensive processing phase of EM had been left out of this technological progress. A simple and innovative conversion of a commercially available MW oven into a processing machine has dramatically reduced all previous processing times, while maintaining consistently reproducible results. The significant contributions of this newly designed MW oven have made it possible for EM to compete on the same level as light microscopy. The procedure discussed here would be especially useful for serious diseases, in which a speedy diagnosis is essential, when ultrastructure evaluation is the only answer. It is also possible to do immunoelectron microscopy, with more advanced models of MW by direct or indirect methods using polyclonal and monoclonal antibodies producing as good, if not better, ultrastructural quality (Chicoine and Webster, 1998). The inventive MW oven, and the methodology found in this book, will enable the electron microscopist in both diagnostic and research facilities to spend more time in the analysis of high-quality results without lengthy processing times. The purpose of this chapter is to establish methodology, which is not to say results are insignificant or secondary, but results are only as reliable as the technique used. Without a clear, concise, consistent method the results would be unreliable. The constancy that the MW offers ensures the quality needed for the meticulous work demanded of EM, while dramatically reducing the time necessary to achieve dependable results.
48
Austin
The only real blunder that a technician can make in using this oven is to forget the proper use of the temp probe. The technician must remember to replace the probe after each refill of fluid, especially in the infiltration phase. Figure 6 illustrates the devastating effect such a mistake can make. The speed of processing, the reduction in volume of chemistry used, the variety of resins that can be used, the simple design allowing easy use, the capability of using water loads to eliminate hot spots, without compromising the fixation, dehydration, infiltration, and polymerization of the tissue, are exceptional, and negates a number of antiquated embedding rules. One of the unique advantages is the ability to interrupt the processing procedure at any point, without compromising the specimen integrity. The technician can go back and forth between MW and conventional processing, at will. This adds to a technician's flexibility in handling a variety of specimens. Now doctors and technicians alike have an additional device that can make their work faster and easier. Clearly, the advantages far outweigh any disadvantages. REFERENCES Chicoine L, Webster P (1998) Effect of microwave irradiation on antibody labeling efficiency when applied to ultrathin cryosections through fixed biological material. Microsc Res Tech 42:24-32. Giberson RT, Demaree RS (1995) Microwave fixation: understanding the variables to achieve rapid reproducible results. Microsc Res Tech 32:246-254. Giberson RT, Demaree RS, Nordhausen RW (1997) Four-hour processing of clinical/ diagnostic specimens for electron microscopy using microwave technique. J Vet Diagn Invest 9:61- 67. Reynolds ES (1963) The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J Cell Biol 17:208. Sato T (1967) Modified method for lead staining. J Electron Micros 16:133.
5
Specimen Preparationfor ThinSection Electron Microscopy Utilizing Microwave-AssistedRapidProcessing in aVeterinaryDiagnostic Laboratory Robert W. Nordhausen and Bradd C. Barr
INTRODUCTION Although transmission electron microscopy (TEM) has always been recognized as an useful tool in research laboratories for studying ultrastructure, it has had limited practical use in the diagnostic lab environment because of the technical expense and time necessary to prepare tissues for ultrastructural examination. Reduction of these two factors would enhance practical usefulness of TEM in the hospital/diagnostic environment. With the advent of microwave (MW) processing (Demaree, Giberson, and Smith 1995; Giberson and Demaree, 1995; Giberson, Smith, and Demaree, 1995; Giberson, Demaree, and Nordhausen, 1997), these obstacles of technician expense and time have, in fact, been significantly reduced. The MW rapid processing procedure has been tested and utilized for over 2 yr in the preparation of routine diagnostic specimens for thin section electron microscopic (EM) imaging at the California Animal Health and Food Safety Laboratory. This lab's mission includes the rapid diagnosis of spontaneous diseases in livestock herds and poultry flocks, including zoonotic diseases and surveillance for foreign animal diseases in the state of California, and incorporates multidisciplines, including full-service pathology, microbiology, virology, toxicology, immunology testing, and EM, to achieve maximum efficiency in diagFrom: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ
49
50
Nordhausen and Barr
nosis of disease for a wide range of domestic and exotic mammalian and avian species. TEM has proven its value in this lab environment, where new diseases ornew disease manifestations of recognized disease pathogens are frequently encountered. It has proven to be particularly useful in the recognition or diagnosis of diseases for which there are no other available commercial diagnostic reagents or tests (i.e., viral isolation, fluorescent antibody test, serology, antigen enzyme-linked immunosorbent assay, polymerase chain reaction, and so on) (Charlton and Bickford, 1995; Conrad et al., 1993; Cooper et al., 1995; Daft et al., 1996; Droual et al., 1995; Marsh et al., 1996; Patton et al., 1996; Walker et al., 1995; Woods et al., 1993, 1994, 1996, 1997). With the advent of the novel MW preparative methodology of Giberson, Smith, and Demaree (1995) and Giberson and Demaree (1995), thin-section EM has been transformed from an expensive, time-consuming procedure relegated to special projects, to a practical and reproducible procedure available to the diagnostic electron microscopist on a routine basis. This new methodology shortens turnaround time of specimen preparation for thin-section EM from 48-72 h to approx 4 h or less, with comparable ultrastructural quality. Thus, this new MW-assisted technique has far-reaching implications for the usefulness of EM imaging in thin-section cases, both in veterinary (Vet) and human medicine, in which diagnosis is dependant on ultrastructural examination. In human medicine, it can be of value for rapid diagnosis and monitoring of individual patients, including interpretation of renal biopsies, transplant status, and tumor cell identification, as well as for identification of certain infectious agents. In Vet medicine, rapid diagnosis is essential to prevent or control rapidily spreading diseases in livestock herds or flocks, in which enormous herd size and high animal density promote spread of highly contagious, infectious diseases, and where the presence of nutritional deficiencies or environmental toxins can also lead to high morbidity/mortality in a short period of time. It is critical to make rapid and accurate diagnoses of diseases under these circumstances. One example of such rapidly spreading disease is infectious laryngeal tracheitis (ILT), a serious herpes virus (HV) infection in chickens. This respiratory disease is rapidly contagious, and virulent strains cause high mortality in chicken flocks (Ritchie and Carter, 1995). ILT infection causes sudden onset of hemorrhagic or fibrinous inflammation in the tracheal mucosa of chickens, producing obstructing caseous plugs
MW Processing of Vet Specimens for EM
51
that can cause death caused by asphyxiation (Ritchie and Carter, 1995). Intranuclear inclusion bodies in the respiratory epithelium, as observed by light microscopy, are strongly suggestive of the HV, however, these inclusions may not always be found. Serum antibodies to the ILT HV can be useful for indicating exposure, but do not confirm the disease ILT. Confirmation of ILT has traditionally required time-consuming and problematical virus isolation, requiring up to 2 wk to complete. However, with the advent of rapid MW processing, HV from laryngeal lesions now can be easily demonstrated on a same-day basis, by thin section EM. Rapid diagnosis is even more important when highly contagious, exotic, or foreign animal diseases occur, such as recent horse morbillivirus outbreak in Australia (Nowak, 1995). In such instances, rapid diagnosis, followed by flock or herd quarantine or evaluation, may be necessary, in order to control disease spread. Another example serves to illustrate the usefulness of rapid MW processing, when conventional diagnostic test reagents are not available, and time is of the essence. In the fall of 1998, an international exhibition and competition of koi was held in the United States. During these competitions, expensive prize fish were commingled in a single tank for judging, then dispersed to the participants who returned them to their various home locations worldwide. Several weeks after this competition, koi became ill and died. The symptoms and etiology suggested a previously unrecognized viral disease. Obviously, many prized koi were at risk. To quickly identify the cause, a fish pathologist presented gill tissue and cell culture supernatant for EM, screening for any viruses present, since positive cytopathic effect in a novel cell line suggested a potential viral pathogen. TEM of the cell culture supernatant first yielded a HV by negative-stain EM, indicating that this might be the cause. Subsequently, within 5 working h, using rapid MW processing of gill tissue, HV was confirmed in nuclei of white blood cells in gill capillaries, by thin section EM thus proving the association of the cultured virus observed by negative-stain EM with in situ HV in the gills of the dead fish. With knowledge of this newly recognized herpes disease, the diagnostician began implementing immediate measures to control and prevent further spread. Other examples demonstrating the time saving usefulness of this MW-enhanced processing procedure, as well as the high quality of results obtained, follow.
52
Nordhausen and Barr MATERIALS A N D M E T H O D S
The first step in any MW processing protocol is the mapping of the MW cavity to identify hot and cold spots (Giberson and Demaree, 1995). The authors have found that this process is most easily accomplished through the use of a neon bulb array and water loads in the MW cavity. Water loads are positioned in the MW cavity, so that, when the neon bulb array (Ted Pella) is placed in a known location, the bulbs do not light when the MW is turned on. This region then becomes the sample site in the MW cavity. The MW used in this lab is a Model 3450 (800 W) Microwave Processor (Ted Pella). The instrument has a built-in temperature (temp) probe, which can be set to control the maximum temp of the processing solution. A water-load recirculation device also comes with the Model 3450, and is designed to circulate and cool the water load(s) used to create the cold spot in the MW cavity. All processing steps in the MW are done in a cold spot. The following protocol is used in this lab for MW processing. 1. Aldehyde fixation: Tissues submitted from diagnostic cases usually have been initially fixed in 10% neutral buffered formalin. These fixed tissues are trimmed and postfixed in 0.5 mL modified Karnovsky's (1965) in 1.5-mL microcentfifuge tubes (MTs), which are then placed in a holder (Ted Pella), and cooled in an ice bath (Fig. 1). The fixative in the MT is cooled to <15°C, then MW-irradiated for the following time sequence 10-20-10 (10 s 100% power; 20 s, 0% power; 10 s, 100% power). 2. Buffer rinse: The holder is removed from the oven, the fixative removed from the MT, and replaced with 0.2 M sodium cacodylate, pH 7.4. Two 2-min changes are done. 3. Reduced osmium (Os) fixation: The buffer is removed from the MT, and replaced with 0.5 mL 2% osmium tetroxide (OsO 4) in 2.5% potassium ferrocyanide. The holder with MT is again placed in an ice bath, which is placed in the MW cavity. The temp in the ice bath is controlled by the temp probe, which is set to maintain a 25°C temp maximum during the 2.5 min MW irradiation at 100% power. Temperature rise of ice water is typically <8°C, after MW irradiation. The tissues remain in fixative for an additional 15 min outside the MW. 4. Buffer rinse: Briefly rinse (<1 min) tissues in 0.2 MNa cacodylate buffer. 5. Dehydration: The tissues in the MT are transferred to baskets (Ted Pella), which are placed in 55-mm diameter polypropylene Petri dishes, for continued processing (Fig. 2). Approximately 15 mL dehydrating agent is placed in the Petri dish, for each step. MW-assisted dehydration is first carried out in a graded ethanol series (2 × 50%; 2 × 70%;
MW Processing of Vet Specimens for EM
53
Fig. 1. MTs contained in Teflon holder for the fixation steps. Note the location of the temp probe in a dummy tube (no sample). Reprinted with permission from Giberson et al., 1997.
The probe is sub~] mersed > 4mm II into fluid "~11
Fig. 2. Tissue processing baskets contained in a polypropylene Petri dish (Ted Pella) which is used for sample processing after the fixation steps. Reprinted with permission from Giberson et al., 1997.
54
Nordhausen and Barr
2 x 90%; 3 x 100%), which is followed by 3 x 100% in acetone. Each step in each series is done in the MW at 100% power for 2.5 min, with the temp probe set to maintain 56°C. 6. Infiltration: The last acetone step is removed from the Petri dish, and replaced with a 1:1 mixture of 100% acetone and Spurr' s resin. The dish with baskets is returned to the MW and irradiated at 100% power for 15 min, with the temp probe set to maintain 56°C (Fig. 2). This is followed by three changes (10 min each) in 100% resin, with the temp probe set to maintain 56°C. 7. Resin polymerization: The tissue is transferred to polyethylene embedding capsules, which are filled with fresh 100% resin. The embedding capsule cap is removed from the capsule, a piece of Parafilm ® placed in it, and it is then placed on the resin-filled capsule. The capsules are placed in a holder (Ted Pella), which is placed in a plastic container filled with water (1 L) (Fig. 3; Giberson et al., 1997). This is placed in the MW, and irradiated for 75 min at 100% power, with the temp probe set to maintain 96°C. Almost all tissues submitted to EM facilities, in Vet diagnostic labs, initially are fixed in 10% neutral buffered formalin, having initially been collected during routine necropsies. Following initial reading of routinely stained hematoxylin and eosin-stained histology sections on glass slides, the pathologist may order thin section EM examination of tissue. These tissues should be postfixed in a glutaraldehyde-based fixative, and processed for thin-section specimen preparation, using the rapid MW methods described by Giberson et al. (1995). The authors have modified these methods, because specimens are typically larger than in that reference, and consist of thin slices of 3-4 mm, or larger, in total surface area, by 1 mm or less in thinness. Specimens are postfixed in 2% OsO4 reduced with 2.5% potassium ferrocyanide (Russell and Burguet, 1977) in an ice bath for 2.5 min, in the MW at full power, with a temperature restriction (TR) on ice bath at 25°C. They are then fixed for an additional 15 min in Os fixative on ice, to ensure complete postfixation of specimens, which is another variation from the Giberson et al. (1995) protocol. Other modifications from the original protocol include the use of less-expensive and less-volatile ethanol for dehydration, rather than acetone, and extended time intervals for dehydration steps of 2.5-min duration. Dehydration is followed by transition through three changes of 100% acetone, 2.5 min each, at 56°C TR before infiltration for 15 min with 1:1 acetone:Spurr's media. This is followed by three infiltration
MW Processing of Vet Specimens for EM
~'0~
55
~0'
-
4.
....
! .
[~
.
.
.
.
.
,.
J rj
U/"
I Fig. 3. A 1-L (no. 5) Rubbermaid ®container is used to hold the embedding
capsule holder (Ted Pella), during the MW polymerization step. The holder with capsules is placed in the container, then filled with approx 1 L water to cover the capsules and holder. Water is added, as needed, to maintain coverage. Reprinted with permission from Giberson et al., 1997. steps, 10 min each, in the MW oven at 56°C TR, as originally described by Giberson, Smith, and Demaree (1995). In this lab, the more fluid Spurr' s resin is used, rather than the originally described Epon-Spurr' s mixture. Total processing time, from wet tissue through resin polymerization, is just 3 h, compared with previously reported processing times of 2.25 h (Giberson et al., 1995). Other modifications employed in this lab are more process-oriented, including incorporating various parts of MW processing into conventional processing technique. Depending on circumstances of workload and the time of day when a specimen arrives, the authors may fix conventionally in modified Karnovsky's (1965), followed by reduced OsO4 in the late afternoon, wash tissue overnight (ON) in 0.2 M Na cacodylate, then dehydrate, infiltrate with resin, and polymerize blocks the next morning in the MW. Another frequent variation used is to MW-fix, dehydrate, and start infiltration in the afternoon, continue infiltration in pure resin ON in a vacuum chamber at approx 5-10 mBar, followed by embedment and polymerization in the MW in the morning.
56
Nordhausen and Barr
RESULTS Figures 4-7 are images that demonstrate HV lesions in various animals. The specimens in these figures were processed either by the MW rapid-processing protocol entirely, from wet tissue to polymerized blocks, or by a combination of conventional and MW processing. The multinucleated epithelial syncytial cell (Fig. 4) is from a 1-2-moold chicken with necrohemmorrhagic to mucopurulent tracheitis, from a flock of 3000-4000 birds. This tissue was prepared by the authors' modified MW procedure in under 4 h, from wet tissue to polymerized block. The syncytial nuclei show marginated chromatin, typical of HV epithelial infections, and numerous naked viral nucleocapsids (inset). A HV suspect from a koi is illustrated in (Fig. 5A,B). These images show a portion of a nucleus from a plasma cell in a gill capillary of the koi infected with a previously unrecognized HV infection. This tissue was prepared by combining MW and conventional processing, including conventional fixation through osmification in the late afternoon, washing ON in 0.2 M Na cacodylate buffer, and performing dehydration, infiltration, and embedment the following morning, using the MW protocol. This image was produced in approx 5 working h. The duck esophagus epithelial cell (EC) expressing HV in (Fig. 6A,B) are from a case of a serious and often fatal disease called "duck viral enteritis" (DVE), or duck plague, which is very difficult to isolate in cell culture. This disease produces multiple hemorrhages in various organs throughout the body. In California, epithelial necrosis with intracytoplasmic inclusions are often seen in esophageal epithelium (Barr et a1.,1992). In Fig. 6A,B, a nucleus (N) with HV nucleocapsids is in an EC. The cytoplasm contains an enormous inclusion of enveloped virions, some with multiple nucleocapsids surrounded by a single envelope (arrowheads). This tissue was prepared entirely by MW rapid processing in approx 4 h. The next image (Fig. 7) shows HV cytoplasmic inclusions in esophageal epithelium from a parrot with Pacheco's disease, which most often is manifested by lesions in liver, spleen, kidney, and intestine (Ritchie and Carter, 1995), and may also, but less commonly, involve the esophagus (Kaleta and Brinkman, 1993). In this image, cytoplasmic inclusions containing membrane-bound virions (arrow) appear similar to, and suggestive of HV in duck plague. However, unlike DVE, in which cytoplasmic inclusions can be seen in esophageal epithelium by light microscopy, only intranuclear viral inclusions are usually seen with Pacheco's disease by light microscopy. This specimen was processed entirely by MW rapid processing.
MW Processing of Vet Specimens for EM
57
Fig. 4. Tracheal epithelial multinucleated syncytial cell from a chicken infected with HV ILT. Arrow indicates area of nucleus containing HV nucleocapsid that is enlarged by inset. Bar = 5.0 ~tm. Inset bar = 200 nm. The liver tissue in Fig. 8A,B are from a lovebird that was one of eight birds that had died from an aviary of 45 birds. The submitting Vet indicated that the birds had moderate hepatomegaly and moderate-tosevere splenomegaly. Histologically tissues were unremarkable, according to the pathology report, but an unusual granular appearance of the Kupffer cells in liver sinusoids, and in macrophages in the bile ducttriad region of the liver, was indicated. Liver tissue was subsequently processed with a modified MW procedure that included an ON infiltration of tissue in pure resin (Epon-Spurr' s 50:50) in a vacuum chamber at approx 5-10 mBar Polymerization was MW-assisted, and EM examination of the unusual sinusoidal Kupffer cell granularity revealed an intracellular microbe consistent with Chlamydia by morphology. Figure 9 shows an unusual presumptive attaching/effacing Escherichia coli discovered in chickens with severe gastroenteritis. This tissue was partially MW-processed with MW fixation, dehydration, and initial
58
Nordhausen and Barr
Fig. 5. (A) Low-magnification image of plasma cell in capillary from koi gill of previously unrecognized HV disease in koi fish. Arrow indicates intranuclear HV nucleocapsid. Arrowhead points to gill epithelium. Note the marginated chromatin causing the halo appearance of the nucleus. Bar= 1.0ktm. (B) Highermagnification electron micrograph of intranuclear herpes virions from koi fish (arrow). Bar = 0.5 ~tm.
59
M W Processing of Vet Specimens for EM :
!7"
: :i.¸ . ii
Fig. 6. (A) The esophageal epithelium of a duck infected with DVE, or duck plague. Esophagus EC infected with HV. Nucleus (N) with naked HP nucleocapsid (arrows). The cell contains a massive membrane-bound cytoplasmic inclusion with enveloped HP some of which have multiple nucleocapsids surrounded by a single envelope (arrowheads). Bar - 1.0 ~tm. (B) Higher-magnification image showing both intranuclear HV nucleocapsid (arrows) and intracytoplasmic enveloped herpes virions (arrowheads). Bar = 0.5 ~tm.
60
Nordhausen and Barr
Fig. 7. Parrot tracheal EC HV inclusion containing membrane enveloped HV (arrow). Bar = 0.5 gm. infiltration. However, the authors infiltrated the tissue ON in EponSpurr's in vacuo, and MW-polymerized tissue blocks the following morning. The tightly polar-bound rods have disrupted the microvillus absorptive surface of the intestinal mucosa, because of massive colonization. This disease is most often observed in young mammals, and is less frequently described in avian species (Fukui et al., 1995; Sueyoshi et al., 1997; Wada et al., 1995). Figure 10 is of a kidney tubular EC in a turkey. The nucleus of the infected cell shows an intranuclear paracrystalline adenovirus (Ad) inclusion. This tissue was processed by MW up to polymerization. Polymerization of the Epon-Spurr's resin was carried out by conventional means, in an oven at 60°C over a weekend. Figure 11 is of an Ad-infected hepatocyte in a fatal case of hepatitis in a bearded dragon lizard. Both intranuclear and cytoplasmic Ad are present in this image. This tissue was processed through infiltration in the Epon-Spurr' s mixture by conventional processing, but polymerized in the MW oven.
MW Processing of Vet Specimens for EM
61
Fig. 8 (A) Parrot liver with sinusoidal macrophage or Kupffer cell filled with Chlamydia (arrow). Bar marker 5.0 ktm. (B) Higher magnification of Kupffer cell Chlamydia bodies. Bar = 0.1 t.tm.
62
Nordhausen and Barr
¸
t
~ii~!iiiii~%ii'iiiiiii
Fig. 9. Chicken intestine infected with pathogenic attaching/effacing E. coli bacteria massively disrupting enterocyte absorptive surface. Bar marker = 1.0gm. DISCUSSION In this chapter, the authors have attempted to demonstrate the usefulness and flexibility of MW-assisted rapid processing, as described by Giberson, Smith, and Demaree (1995), in a practical Vet diagnostic setting. Examples were selected in which fast turnaround time in specimen processing has been especially helpful to the pathologist. In other cases, conventional processing was combined with the rapid MW procedure, depending on workload issues and time of day or week that the tissue arrived in the EM lab. It should be noted these selected tissues were in good postmortem condition. However, the postmortem condition can frequently be poor in cases in which necropsy or specimen collection is required from animals that expired several hours to days prior to submission. Various aspects of MW-assisted processing of tissues for thin-section EM have been employed for many years (for fixation, see Mayers, 1970; Login and Dvorak, 1985; Wild et al., 1989; Hopwood et al., 1984; Leong et al., 1985; and Chew et al., 1984. For embedment, (see Gia-
MW Processing of Vet Specimens for EM
63
Fig. 10. Turkey with Ad nephritis. Kidney tubule cell with abundant Ad forming a paracrystalline array in nucleus (N). Bar - 5.0 gm. Inset bar - 200 nm. marra, 1985, 1993), although it was not until the work of Demaree et al. (1995), Giberson et al., (1995), and Giberson and Demaree (1995) that a procedure was introduced that incorporates all aspects of tissue processing into a MW-assisted technique, which significantly reduced turnaround time, from 48-72 h to 3-4 h. This significant reduction in processing time allows for same-day imaging of thin-section cases. The authors have provided examples from the Vet diagnostic setting of serious and often fatal diseases that may be quickly and accurately diagnosed by ultrastructure, using thin section EM (Figs. 4, 6A,B, and 7). In the case of the novel virus in a koi fish, the authors were able to help a pathologist quickly identify a new contagious pathogen (Fig. 2). The diagnosis of Chlamydia in Kupffer cells in parrot liver represented an unusual presentation for a commonly encountered disease, which was resolved by TEM. Psittacosis, or parrot fever, is a zoonotic disease that can cause respiratory disease in humans, and is considered an important public health threat. The chicken with the presumptive
64
Nordhausen and Barr
Fig. 11. Bearded dragon lizard with Ad hepititis. Note the extensive involvement of the nucleus (N) with paracrystalline arrays of newly formed Ad. Bar = 5.0 ktm. Inset bar = 200 nm. pathogenic attaching/effacing E. coli was not a case that necessarily needed EM analysis. However, this pathogen has rarely been described in chickens in the United States, although it has been reported previously in Japan (Fukui et al., 1995; Sueyoshi et al., 1997; Wada, 1995), and was of special interest to the pathologist. Renal Ad nephritis in turkeys and hepatic adenovirus in the lizard were submitted for EM, in order to differentiate between HV and Ad as the source for intranuclear inclusions. In the authors' experience, these are among the more routine submissions or requests for thin-section EM. Nevertheless, a quick, positive virus identification is critical to the field practitioner for implementation of disease-management procedures. Although many of the specimens used to illustrate the usefulness of rapid MW-assisted tissue processing were embedded in the EponSpurr' s (50:50), as suggested by Demaree, Giberson, and Smith (1995), for MW-assisted polymerization, Spurr's media alone also yields fine results. The low viscosity of Spurr' s media expedites infiltration of the larger specimens used in the authors' casework. Batch processing of
M W Processing of Vet Specimens for EM
65
tissues through dehydration and infiltration steps is another aspect of MW processing that yields great savings in technician bench time. Six specimens can easily be batch-processed simultaneously. However, the underwater resin polymerization in BEEM TM capsules, as first described by Demaree, Giberson, and Smith (1995), is undoubtedly the most innovative and useful aspect of the MW rapid processing. This step alone reduces processing time by 15 h or more, and could be accomplished with almost any temp-probe controlled MW unit. In this lab, rapid MW processing has become a standard operating procedure, used daily to process tissue for thin section EM, and has elevated TEM, from an occasionally employed tool, to a standard diagnostic procedure that is now considered to be part of the authors' multidisciplinary approach to Vet diagnostics. REFERENCES Barr BC, Jessup DA, Docherty DE, Lowenstein LJ (1992) Epithelial intracytoplasmic herpes viral inclusions associated with an outbreak of Duck Viral Enteritis. Avian Dis 36:164-168. Charlton BR, Bickford AA (1995) Gross and histologic lesions of adenovirus group I in guinea fowl. J Vet Diagn Invest 7:552-554. Chew EC, Riches DJ, Lam TK, Hou Chan HJ (1984) Microwave fixation as a substitute for chemical fixation of tissue for light and electron microscopy. J Anat 138: 586. Conrad PA, Barr BC, Sverlow KW, Anderson M, Daft B, Kinde H, Dubey JP, Munson L, Ardans A (1993) In vitro isolation and characterization of a Neospora sp. from aborted bovine foetuses. Parasitology 106: 239-249. Cooper GL, Shivaprasad HL, Bickford AA, Nordhausen RW, Munn RJ, Jeffrey JS (1995) Enteritis in turkeys associated with an unusual flagellated protozoan (Cochlosoma anatis). Avian Dis 39:183-190. Daft B, Nordhausen RW, Latimer KS, Niagro FD (1996) Interstitial pneumonia and lymphadenopathy associated with circoviral infection in a 6-week-old pig. In: Proceedings 39th Annual Meeting of the American Association of Veterinary Laboratory Diagnosticians. Little Rock, AR, October 11-16. Demaree RS, Jr., Giberson RT, Smith RL (1995) Routine microwave polymerization of resins for transmission electron microscopy. Scanning 17(Suppl. V): 26-27. Droual R, Woolcock PR, Nordhausen RW, Fitzgerald SD (1995) Inclusion body hepatitis and hemorrhagic enteritis in two Africian Grey parrots (Psittacus erithacus) associated with adenovirus. J Vet Diagn Invest 7:150-154. Fukui H, Sueyoshi M, Haritani M, Nakazawa S, Naitoh S, Tani H, Uda Y (1995) Natural infection with Attaching/Effacing Escherichia coli (0 103 :H ) in chicks. Avian Dis 39: 912-918. Giammara B (1985) Ultra-rapid epoxy embedment using microwave energy. Proc Annu Meeting Electron Microsc Soc Am 43: 706-707. Giammara BL (1993) Microwave embedment for light and electron microscopy using epoxy resins, LR White, and other polymers. Scanning 15:82-87. Giberson RT, Demaree RS Jr. (1995) Microwave fixation: understanding the variables to achieve rapid reproducible results. Micros Res Tech 32:246-254.
66
Nordhausen and Barr
Giberson RT, Demaree RS, Jr., Nordhausen RW (1997) Four-hour processing of clinical/ diagnostic specimens for electron microscopy using microwave technique. J Vet Diagn Invest 9:61-67. Giberson RT, Smith RL, Demaree RS, Jr. (1995) Three hour microwave tissue processing for transmission electron microscopy: from unfixed tissue to sections. Scanning 17(Suppl. V):25-26. Hopwood D, Coghill G, Ramsay J, Milne G, Kerr M (1984) Microwave fixation. Its potential for routine techniques, histochemistry, immunocytochemistry and electron microscopy. Histochem J 16:1171-1191. Kaleta EF, Brinkman MB (1993) An outbreak of Pacheco's parrot disease in a psittacine bird collection and an attempt to control it by vaccination.Avian Patho122:785-789. Karnovsky MJ (1965) A formaldehyde - glutaraldehyde fixative of high osmolarity for use in electron microscopy. J Cell Bio127:137A. Leong SS-Y, Dayman ME, Milios J (1985) Microwave irradiation as a form of fixation for light and electron microscopy. J Pathol 146:313-321. Login GR, Dvorak AM (1985) Microwave energy fixation for electron microscopy. Am J Pathol 120:230-243. Marsh AE, Barr BC, Madigan J, Lakritz J, Nordhausen R, Conrad PA (1996) Neosporosis as a cause of equine protozoal myeloencephalitis. JAm Vet MedAssoc 209:1907-1913. Mayers CP (1970) Histological fixation by microwave heating. J Clin Patho123:273-275. Nowak R (1995) Emerging viruses: cause of fatal outbreak in horses and humans traced. Science 268:32. Patton JF, Nordhausen RW, Woods LW, MacLachlan NJ (1996) Isolation of a poxvirus from a Black-tailed deer (Odocoileus hemionus columbianus). J Wildlife Dis 32:531-533. Ritchie BW, Carter K (1995) Herpesviridae, In: Harrison L, ed, Avian Viruses, Function and Control. Wingers, Lake Worth, FL, pp. 171-222. Russell L, Burguet S (1977) Ultrastructure of Leydig cells as revealed by secondary tissue treatment with a ferrocyanide-osmium mixture. Tissue Cell 9:751-766. Sueyoshi M, Nakazawa M, Tanaka S (1997) A chick model for the study of "Attaching and Effacing Escherichia coli" infection. In Paul PS, ed, Mechanisms in the Pathogensis of Enteric Diseases. Plenum, New York, pp. 99-102. Wada Y, Kjondo H, Nakazawa M, Kubo M (1995) Natural infection with Attaching and Effacing Escherichia coli and adenovirus in the intestine of a pigeon with diarrhea. J Vet Med Sci 57:531-533. Walker RL, Read DH, Loretz KJ, Nordhausen RW (1995) Spirochete isolated from dairy cattle with papillomatous digital dermatitis and interdigital dermatitis. Vet Microbio147:343-355. Wild P, Krahenbuhl M, Schraner EM (1989) Potency of microwave irradiation during fixation for electron microscopy. Histochemistry 91:213-220. Woods LW, Hanley RS, Chiu PHW, Burd M, Nordhausen RW, Stillian MH, Swift PK (1997) Experimental adenovirus hemmorrhagic disease in yearling black-tailed deer. J Wildlife Dis 33:801-811. Woods LW, Latimer KS, Barr BC, Niagro FD, Campagnoli RP, Nordhausen RW, Castro AE (1993) Circovirus-like infection in a pigeon. J Vet Diag Invest 5:609-612. Woods LW, Latimer KS, Niagro FD, Riddell C, Crowley AM, Anderson ML, Daft BM, et al. (1994) A retrospective study of circovirus infection in pigeons: nine cases (1986-1993). J Vet Diagn Invest 6:156-164. Woods LW, Swift PK, Barr BC, Horzinek MC, Nordhausen RW, Stillian MH, et al. (1996) Systemic adenovirus infection associated with high mortality in mule deer (Odocoileus hemionus) in California. Vet Patho133:125-132.
6
Microwave Processing of Archived Pathology Specimens for Ultrastructural Examination Robert.}'. Munn and Phillip J. Vogt
INTRODUCTION Microwave (MW)-assisted processing has been shown to provide a considerable number of advantages for microscopic and ultrastructural examination of a wide variety of biological specimens, including increased antibody labeling on cryoultramicrotome sections (Chicoine and Webster, 1998), reduced processing times (Giberson et al., 1995), better retention of some ultrastructural features in animal tissues (Giberson et al., 1997), increased protein antigenicity (Lonsdale et al., 1999), rapid decalcification of bony tissues (Madden and Henson, 1997), increased immunocytochemical reactions (Madden, 1998), and increased enzyme activity (Rassner et al., 1997). Taken together, all of these benefits of MW-assisted processing provide additional evidence that many types of specimens can be rapidly processed without compromising microscopic or ultrastructural features. Universal acceptance and utilization of MW technology in pathology (PATH) continues to be slow, primarily because of hesitancy to use new or different protocols, lack of adequate instrumentation, and the inability to work out new specific techniques because of time and personnel constraints. In order for this rapid processing to become routine in the field of diagnostic electron microscopy (EM), it first must be determined that the procedure retains all of the pertinent ultrastructural features that the pathologist uses to render a specific From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ
67
68
Munn and Vogt
diagnosis. For this purpose, 14 previously fixed PATH samples, which already had been processed by conventional EM methods, were subsequently processed in a MW oven, in order to compare the ultrastructure utilizing the two procedures. MATERIALS AND M E T H O D S All specimens were regular diagnostic EM samples from the Department of Pathology, University of California Davis Medical Center, Sacramento, CA. They had all been fixed in modified Karnovsky's fixative, consisting of 2.0% paraformaldehyde and 2.5% glutaraldehyde in 0.06 M Sorensen's phosphate buffer, pH 7.2, and had been stored in the fixative, at 4-10°C, for periods of 2 h-11 mo. The diagnoses selected for investigation were one-each cases of adenocarcinoma of the ovary, adenosquamous carcinoma, amyloidosis of the lung, astrocytoma, cytomegalovirus infection of the lung, embryonal carcinoma, ependymoma, meningioma, mesothelioma, neuroendocrine carcinoma, pheochromocytoma, progressive multifocal leukoencephalopathy (papovavirus), rhabdoid tumor, and thymoma (Table 1). Four cases are illustrated in this chapter. For regular processing, tissues which had been fixed for 2 h-1 wk were cut into pieces about 1 x 2 x 2mm, in Karnovsky's fixative, and rinsed briefly (3 x 2 min) in 0.1 M Sorensen' s phosphate buffer, pH 7.2. The tissue pieces were postfixed in 1% osmium tetroxide (OsO4) in 0.1 M Sorensen's phosphate buffer for 1 h at 4-10°C, then rinsed briefly (3 x 2 min) in distilled water at 4-10°C. Dehydration was carried out in 50-70-90-100% acetone, 10 min each (3 × 10 min in 100%), bringing to room temperature (temp) in 70%. The epoxy resin formula for both processing protocols consisted of 50% (by wt) Eponate 12 resin (Ted Pella, Redding, CA) and 20% nadic methyl anhydride and 30% dodecenylsuccinic anhydride, with accelerator (2,4,6-tri(dimethylaminomethyl)phenol added at 8 drops/10 g of the thoroughly mixed resin components. Infiltration was in 1:1 acetone:epoxy resin 2-4 h at room temp with gentle rotation. Infiltration was continued with full-strength epoxy resin overnight at room temp with gentle rotation. Fresh full-strength epoxy resin was added to the tissue blocks in BEEM TM embedding capsules and polymerized overnight at 70°C. For MW processing, all indicated steps were carried out in a 900-W rated commercial MW oven, Pelco Model 3450 (Ted Pella). All processing steps were done in the MW at the 100% power setting, following the basic protocol of Giberson et al., 1997. Tissue samples
69
MW Processing of PATH Specimens Table 1 Selected Test Specimens Adenocarcinoma, ovary Adenosquamous carcinoma Amyloidosis, lunga Astrocytoma Cytomegalovirus, lung Embryonal carcinoma Ependymoma
Meningioma Mesothelioma a Neuroendocrine carcinoma Pheochromocytoma a Progressive multifocal leukoencephalopathy Rhabdoid tumor Thymoma a
a Illustrated.
were cut into pieces --1 x 2 x 2 mm, in Karnovsky's fixative, rinsed briefly in 0.1M Sorensen' s phosphate buffer (3 x 1 min) at 4-10°C, and placed into 1.7-mL microcentrifuge tubes containing approx 600 gL 1% OsO4 in 0.1 M Sorensen's phosphate buffer. Tissue was MW-irradiated for 3 min in a cold spot (low MW intensity), using a temp restriction of 37°C, followed by 3 min in the OsO4 outside the MW at 4-10°C. Tissue blocks were then rinsed briefly in Sorensen' s phosphate buffer for 3 x 10 s outside the MW, and transferred to plasticmesh baskets, which were placed in 60 X 15mm polypropylene Petri dishes (Ted Pella). All subsequent steps were done in the MW. With the temp maximum set at 45°C, dehydration was carried out in 50% acetone (two changes at 40 s each), 70% acetone (2 x 40 s), 90% acetone (2 x 40 s), and 100% acetone (3 x 40 s). With the temp maximum set at 50°C, infiltration consisted of 1"1 100% acetone:epoxy resin for 15 min, and fullstrength epoxy resin 3 x 10 min. Resin polymerization was done in the MW according to Demaree et al. (1995) and Giberson et al. (1997)., which entailed using Parafilm®-sealed and capped embedding capsules in a polytetrafluoreethylene rack, which were held under the surface of--900 mL water in a Rubbermaid TM no. 5 dish. Fresh epoxy resin was added to the embedding capsules, and the MW was turned on for 15 min, with the temp probe maximum set to 95°C. Fresh resin then was added again, and polymerization was completed for an additional 60 min, with no temp maximum setting. Under these conditions, the water in the dish boiled, and water was added as needed, to keep the embedding capsules under water during the entire polymerization. For sectioning, the polymerized blocks were trimmed with Teflon ®coated single-edge razor blades and 1.5-gm sections were cut on a RMC
70
Munn and Vogt
MT6000 ultramicrotome, using glass knives. The sections were dried on a hot plate, and stained with equal volumes of steaming 1% methylene blue in 1% Na borate and 1% azure II. Appropriate areas were selected, and ultrathin sections were cut at 60-90-nm thickness, using diamond knives. Sections were picked up onto 200 mesh copper grids without support. Grids were stained with 2% aqueous uranyl acetate and lead citrate, and viewed and photographed utilizing a Philips EM400 transmission electron microscope (FEI/Philips Electron Optics, Hillsboro, OR). RESULTS One experienced board-certified pathologist and one experienced electron microscopist examined the sets of micrographs. In all cases, no discernible differences were observed between the two processing conditions. Although the polymerized blocks from the MW processing were noticeably softer and slightly smaller, there were no differences in the sectioning qualities of the sets of blocks. The MW-processed sections stained just as well as the regularly processed sections, and they were stable in the electron beam. The diagnostic entities chosen for illustration here contain examples of comparisons of extracellular proteins (amyloidosis), cell-surface configurations (mesothelioma), cytoplasmic granules (pheochromocytoma), and cell-to-cell junctions (thymoma). Figures 1-4 illustrate the comparisons between the two processing protocols. In all cases, the blocks selected, and the areas photographed, show greater variation than the differences between the two processing procedures. Figure 1 shows deposition of amyloid fibrils in the lung of a patient with amyloidosis. Capillary lumina are at the far right and far left of each photograph, with the intervening space filled with amyloid fibrils. The fibrils measure about 10 nm in diameter, are nonbranching, and typically show a "pick-up-sticks" appearance. Figure 2 illustrates the typical, long, undulating microvilli on the surface of a mesothelioma, which originated in the thoracic wall. The microvilli themselves are relatively clean, with no glycocalyx. Figure 3 illustrates a pheochromocytoma from the adrenal medulla. There are numerous neurosecretory granules in the cytoplasm of the cells. Some granules contain eccentric electron-dense cores, indicative of norepinephrine secretion. Other granules have central cores, indicative of epinephrine secretion. Figure 4 is from a thymoma, which is derived from thymic epithelial cells. The cells are attached to each other with well-formed desmosomes,
71
MW Processing of PATH Specimens
~.~ . . . .
~.~
i Fig. 1. Amyloidosis of the lung, after regular (A) and MW (B) processing. The amyloid fibrils are located in the interstitium between capillaries. They are -~10 nm in diameter, nonbranching, and randomly arranged. Bar = 1.0 ~tm.
Fig. 2. Mesothelioma of the pleura, after regular (A) and MW (B) processing. The characteristic long, narrow, undulating microvilli are easily identified. Bar = 0.5 ktm.
which often are in clusters of 2-4. Cytoplasmic tonofilaments are present, and some course into the desmosomes. Notice the characteristic space with linear striations between the two adjacent cell membranes, and the dense, thickened subplasmalemmal plaque.
72
Munn and Vogt
.;
.
::7
Fig. 3. Pheochromocytoma of the adrenal, after regular (A) and MW (B) processing. This neoplasm arises from cells in the adrenal medulla, and is characterized by numerous cytoplasmic neurosecretory granules. Those granules with eccentric electron-dense cores are associated with norepinephrine secretion, and those with central cores with epinephrine secretion. Bar - 1.0 gm.
Fig. 4. Thymoma of the thymus, after regular(A) and MW (B) processing. This neoplasm arises from epithelial cells in the thymus, which exhibit well-formed desmosomes with cytoplasmic tonofilaments. Some tonofilaments insert into the desmosomes. Bar - 0.5 ~tm.
MW Processing of PATH Specimens
73
DISCUSSION Several methods for rapid processing of tissues for transmission EM have been published (Nesland et al., 1982; Erlandson, 1994). Very few of these techniques are used on a routine basis, however, probably because of variations in tissue density, tissue block size, and difficulties in consistently sectioning blocks that have been polymerized by high temp of 100°C, or more, in an oven. MW processing offers the advantage of short processing times, and, if done properly, will overcome the problems mentioned above. The advantages of MW processing have been reviewed (Leong, 1994), and commercial MW ovens are now available that allow rapid processing, with few, if any, disadvantages (Giberson et al., 1995, 1997). With the procedures described here, the processing times were reduced from --30 h, for regular processing, to less than 3 h for MW processing. Further refinements in technique may reduce the MW processing times to <2 h, from fresh tissue to polymerized blocks. MW processing for diagnostic transmission EM in PATH is one area in which the considerable time savings could improve patient care, through shorter turnaround times, decreased costs, and more accurate diagnoses. Acceptance of this technique has been slow, however, even though procedures for MW processing have been available for several years (Dardick, 1994). MW energy has been used, in fact, for many years to fix tissues for ultrastructural examination (Login and Dvorak, 1985), and to speed up the staining of ultrathin sections for ultrastructural PATH (Estrada et al., 1985). The present study determined that MW processing of PATH tissues, after prolonged fixation, had no detrimental effect on the ultrastructure of cells from a variety of conditions. In each of the 14 entities studied here, the ultrastructure of the cells closely matched the pertinent ultrastructural features as described in several leading textbooks on ultrastructural PATH (Dardick et al., 1996; Dvorak and MonahanEarley, 1992; Erlandson, 1994). One unexpected benefit of the present study was the validation of the use of prolonged storage of PATH specimens in appropriate fixative for periods approaching 1 yr. Even after more than 11 mo storage in Karnovsky's fixative at 4-10°C, no discernible differences could be detected between the more conventionally fixed and processed tissues and the MW-processed tissues. Another benefit of the prolonged storage is that it is not necessary to run the tissues through OsO4 before prolonged storage, which allows anatomical features to remain visible with a dissecting microscope, for accurate selection, trimming, and orientation.
74
Munn and Vogt
ACKNOWLEDGMENTS The authors gratefully acknowledge the excellent technical assistance of Grete Adamson and Paul Lee, whose untiring efforts made this endeavor possible. REFERENCES Chicoine L, Webster P (1998) Effect of microwave irradiation on antibody labeling efficiency when applied to ultrathin cryosections through fixed biological material. Micros Res Tech 42:24-32. Dardick I (1994) Outline for ultra-rush tissue processing for transmission electron microscopy using the microwave oven. Soc Ultrastructural Pathol Newslett 7:7. Dardick I, Eyden B, Federman M, Hammar S, Lajoie G, Malott R, et al. (1996) Handbook of Diagnostic Electron Microscopy for Pathologists-in-Training. Igaku-Shoin, New York. Demaree RS, Jr., Giberson RT, Smith RL (1995) Routine microwave polymerization of resins for transmission electron microscopy. Scanning 17(Suppl. V):V25-V26 (abstract). Dvorak AM, Monahan-Earley RA (1992) Diagnostic Ultrastructural Pathology I. A Text-Atlas of Case Studies Illustrating the Correlative Clinical-Ultrastructural Pathologic Approach to Diagnosis. CRC, Boca Raton, FL. Erlandson RA (1994) Diagnostic Transmission Electron Microscopy of Tumors. Raven, New York. Estrada JC, Brinn NT, Bossen EH (1985) Rapid method of staining ultrathin sections for surgical pathology TEM with the use of the microwave oven. Am J Clin Pathol 83:639-641. Giberson RT, Smith RL, Demaree RS (1995) Three hour microwave tissue processing for transmission electron microscopy: from unfixed tissues to sections. Scanning 17(Suppl. 5):26-27. Giberson RT, Demaree RS, Nordhausen RW (1997) Four-hour processing of clinical/ diagnostic specimens for electron microscopy. J Vet Diagn Invest 9:61-67. Leong AS-Y (1994) Microwave technology for morphological analysis. Cell Vision 1:278-288. Login GR, Dvorak AM (1985) Microwave energy fixation for electron microscopy. Am J Pathol 120:230-243. Lonsdale JE, McDonald KL, Jones RL (1999) High pressure freezing and freeze substitution reveal new aspects of fine structure and maintain protein antigenicity in barley aleurone cells. Plant J 17:221-229. Madden VJ, Henson MW (1997) Rapid decalcification of temporal bones with preservation of ultrastructure. Hearing Res 111:76-84. Madden VJ (1998) Microwave processing of cell monolayers in situ for post-embedding immunocytochemistry with retention of ultrastructure and antigenicity. Microsc Microanal 4(Suppl. 2):854-855. Nesland JM, Millonig G, Wilson A, Johannessen JV (1982) Rapid techniques in diagnostic electron microscopy. Ultrastruct Pathol 3:295-300. Rassner UA, Crumrine DA, Nau P, Elias PM (1997) Microwave incubation improves lipolytic enzyme preservation for ultrastructural cytochemistry. Histochem J 29:387-392.
7
Microwave Fixation of Rat
Hippocampal Slices Marcia D. Feinberg, Karen M. Szumowski, and Kristen M. Harris
INTRODUCTION The authors' research focuses on the structural basis of synaptic function. Knowing the dimensions and connectivity of presynaptic and postsynaptic elements is necessary for understanding the mechanisms of synaptic transmission in the central nervous system. To this end, electrophysiological (EP) methods are used in combination with confocal microscopy, serial electron microscopy (EM), three-dimensional reconstruction, and quantitative image analysis. Microwave (MW)enhanced methods for preparing hippocampal (HC) slices are described in this chapter. These procedures have been adapted from Jensen and Harris (1989) and Login and Dvorak (1985). MW-enhanced fixation has greatly improved the integrity of slice preparation for morphological study. Earlier procedures, using immersion fixation, required 1-2 h for the aldehydes to reach the center of the slice. Thus, the interior tissue became hypoxic during fixation, which led to severe deterioration of structure. Hence, it was desirable to have a procedure that produces rapid fixation throughout the slice, without additional manipulation of the slice prior to immersion fixation (Chang and Greenough, 1984; Harris et al., 1980; Petukhov and Popov, 1986; Reid et al., 1988). HC slices are prepared for EP by a variety of approaches, depending on the specific goals of the experiments (Harris and Teyler, 1984; Jackson et al., 1993; Sorra and Harris, 1998; Kirov et al., 1999). The authors From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ
75
76
Feinberg, Szumowski, and Harris
have based these methods on living slices that have been maintained at the interface of oxygenated air (95% O2, 5% CO2) and artificial cerebral spinal fluid components within a chamber. The best depth for recording EP responses in brain slices is between 70 and 150 ~tm below the air surface (Anderson et al., 1980; Jensen and Harris, 1989). Therefore, high-quality preservation of ultrastructure at this depth within the slice has been the main objective. The MW-enhanced fixation process produces high quality fixation to the center of the slice within seconds, so that one can detect those morphological changes that are likely to be associated with an experimental manipulation, as opposed to the effects of hypoxia during slow immersion fixation (Harris and Jensen, 1989). HC slices destined for EM undergo further processing, according to the schedule discussed below. Four d are needed for routine immersion processing, but only 3-4 h are required when MW-enhanced processing is used. Representative examples of tissue preserved from MW-enhanced fixation and perfusion-fixed HC in vivo are compared, and the processing details are presented. In addition, the authors demonstrate that resectioning the slices to 100 ~tm provides more uniform penetration ofreduced osmium (Os) fixative, and gives the most densely stained membranes, which is desirable for tracing dendrites and synapses through serial thin sections. These methods have been used primarily for HC slices and perfusion-fixed hippocampus in vivo (Sorra and Harris, 1998; Shepherd and Harris, 1998; Fiala et al., 1998; Jensen and Harris, 1989; Kirov et al., 1999), but the authors anticipate that they will prove to be applicable to all brain and spinal cord slice preparations, with slight modifications (White et al., 1994; Feirabend et al., 1994). MATERIALS A N D M E T H O D S
MWX-Enhanced Fixation of Slices Maintained In Vitro 1. Slices are prepared according to standard procedures for EP or pharmacological studies (Harris and Teyler, 1984; Jackson et al., 1993; Sorra and Harris, 1998; Kirov et al., 1999). 2. The HCs are removed from rats of various developmental ages, ranging from birth to young adult (65-75 d old). 3. 4-6 transverse slices (400 Bm) are cut from the middle-third, with a tissue chopper (Stoelting, Wood Dale, IL) or vibroslicer, VT1000S (Leica, Microsystems Inc., Bannockburn, IL), into ice-cold media containing (in raM) -117 NaC12, --5.4 KC1, --26.2 NaHCO 3, 1 NaH2PO 4, either 2.5 or 3.2 CaC12, --1.6 MgSO 4, and 10 glucose, equilibrated with 95% 0 25% CO 2, pH 7.4.
MW Fixation of Rat HC Slices
77
4. Slices are transferred to nets positioned over wells, with physiological saline at the interface of humidified 95% 02-5% CO 2, at 32°C, in a recording chamber (Stoelting), and maintained for varying times in vitro. 5. EP recordings are made according to experimental protocols (Harris and Teyler, 1984; Jackson et al., 1993; Kirov et al., 1999; Sorra and Harris, 1998). 6. At the end of each experiment, the electrodes are removed. 7. The slice, still attached to its net, is quickly transferred (less than 30 s) to 5.5 mL fixative (6% glutaldehyde, 2% paraformaldehyde, 2 mM CaC12, 4 mM MgC12 in 0.1 M cacodylic buffer at -23°C) in a 35-mm polystyrene Petri dish. 8. The dish is placed in a vented MW oven (Amana Radar Range, model RS4141,700 W), situated immediately near the recording rig. 9. Prior to experimentation, empirical procedures were used to calibrate the duration for MW-enhanced fixation. Slices are exposed to microwave irradiation (MWI), with fixative for increasing duration until an optimal temperature (temp) (35-58°C) is achieved. The temp is measured immediately after MW-enhanced fixation is complete. The final temp should not exceed 58°C, otherwise microtubules are lost. In the Amana MW oven, the optimal time is 8 s at full power in a MW cold spot, as determined with the neon bulb array (Giberson and Demaree, 1995; Jensen and Harris, 1989). Two water loads (800 mL each, in two beakers) are placed in the rear comers of the oven, to absorb reflected radiation. 10. Postfixation, up to an overnight period, after MW fixation in the same fixative, is not needed (Jensen and Harris, 1989); however, it is convenient, and serves to stabilize the tissue for subsequent handling. 11. Slices are transferred to 0.1 M cacodylic buffer for several rinses, and are stored in this buffer prior to further tissue processing. This procedure is optimized for 400-~t thick brain slices. However, if larger tissue samples are required, such as for uncut HC or another portion of the brain, the authors have found that steps 8 and 9 can be modified slightly. Namely, the 35-mm dish with 5.5 mL fixative is placed on a large bag of crushed ice, and microwaved for 1 min. The water loads are still present in the back of the MW oven. This procedure allows a longer exposure to the MW energy, to facilitate penetration of the aldehydes through the thicker tissue while maintaining a temp low enough so that the tissue at the surface is not overheated and destroyed.
MW.Enhanced Processing MW-enhanced processing is done in a programmable Pelco model 3450 (Ted Pclla) MW oven with the Pclco 3420 load cooler attachment and tcmp probe. This MW oven is kept in a fume hood.
78
Feinberg, Szumowski, and Harris
1. Embed the fixed 400-~tm slice (or larger piece of tissue) in agar, and trim manually, under a dissecting scope, to isolate the region of interest (e.g., HC area CA1) for processing. 2. Re-embed the region of interest in agar, and section at 100 ~tm, using the Leica VT1000S vibrating slicer, keeping the central two pieces of a 400-~tm slice. This procedure removes the outer 150-200 ~tm of cut surface, where the tissue is damaged. 3. Rinse, and store sections in 0.1 M cacodylic buffer. 4. Place sections and buffer in 1.7 mL polypropylene microcentrifuge tubes (MTs) positioned in a rack around a central water-filled MT, which serves as a temp monitor during MWI. 5. Position MTs in a designated cold spot. The rationale and calibration for optimal sample placement and strategic positioning of the two water loads within the MW oven has been previously described (Giberson and Demaree, 1995; Giberson et al., 1997). 6. Each water load contains 400 mL distilled water in a polypropylene beaker. One water load is always connected to the load cooler attachment (Ted Pella) via inlet and outlet tubes in the oven cavity. The water in the other beaker is changed frequently throughout the MW sessions, to avoid overheating and loss of MW-absorption capacity. 7. The volume amounts of all processing chemicals in steps 8-12 below are 600 ~tL in the MTs (Giberson et al., 1997). The tip of the temp probe is positioned 2 mm deep in the 600 ~tL water in the monitor tube. 8. Remove buffer from tissue and add 1% Os and 1.5% potassium ferrocyanide in 0.1M cacodylic buffer, and cool on an ice bath, until the temp in the monitor tube is less than 15°C. Then MWI for 2.5 min, up to 37°C. 9. Rinse tissue in 5-6 changes of 0.1 M cacodylic buffer, outside of MW oven, 1 min/change. 10. Add 1% osmium in 0.1M cacodylate buffer, cooled to less than 15°C on ice, and MWI for 2.5 min, up to 37°C. 11. Rinse tissue in 5-6 changes of 0.1M cacodylic buffer, outside of MW oven, 1 min/change. Rinse tissues again, but in two brief changes of distilled water. 12. Stain en bloc with 1% aqueous uranyl acetate cooled on ice to less than 15°C, and MWI for 2.5 min, up to 37°C, followed by two brief water rinses. 13. For steps 14-17 below, add 15-20 mL fresh acetones or resins in the 60-mm polypropylene petri dishes and position the temp probe 4 mm deep in acetone or resin in the 60-mm Petri dish (Giberson et al., 1997). Failure to maintain the probe 4 mm deep in resin results in premature polymerization, (R. Giberson, personal communication). 14. Transfer slices to the flowthrough baskets seated in 60-mm Petri dishes containing 50% acetone (aqueous).
MW Fixation of Rat HC Slices
79
15. Dehydrate through graded acetone series (50, 70, 90 and 100%, two changes each), and MWI for 40 s during each change, up to 37°C. 16. Infiltrate slices in the flowthrough baskets with 1 acetone: 1 Spurr'sEpon (1 Spurr' s: 1 Epon) resin (Ted Pella and Ladd Research, Burlington, VT), with the temp restriction level raised to 45°C and the temp probe tip submersed 4 mm deep in the resin. MWI for 15 min. 17. Infiltrate with two changes of 100% Epon-Spurr' s and MWI for 15 min each change. 18. Transfer into fresh 100% resin (Epon-Spurr' s or Epon alone) in either flat, coffin-shaped molds or BEEM TM capsules, then cure for 48 h in a 60°C oven. 19. Alternatively, embed in Epon-Spurr's in Pelco capsules (Ted Pella), and cure in the Pelco 3450 MW oven for 75 min according to procedures described in Giberson et al. (1997).
Microtomy 1. The cured blocks are cut with a diamond Histo knife (Diatome, US) on a Reichert Ultracut S (Leica) ultramicrotome, to obtain a full-face section of 1-~tm thickness. 2. The 1-~tm sections are dried onto glass slides, and stained with 1% toluidine blue stain at -60°C, until the edges of the stain become irridescent, then are rinsed with distilled water, followed by 70% ethyl alcohol, then water-rinsed again, dried, and examined under a light microscope. The 1% toluidine blue stain is made from equal amounts of 1% toluidine blue (aq) to 1% sodium borate (aq), diluted in half with 70% ethanol. 3. A raised rectangular or trapezoidal area, usually less than 100 x 100 ~tm, is made with a diamond square or wedge-shaped trim tool (Diatome), in the desired area of the block face. 4. Serial thin sections are cut with a diamond knife (Diatome). 5. The serial sections are mounted on Pioloform-coated (SPI Supplies, Westchester, PA) slot grids (Synaptek, Ted Pella). To prepare the grids, -- 1.2% Pioloform in chloroform is coated onto glass microscope slides, either by hand or under controlled-release, using the EFFA film caster (Ernest Fullam, Latham, NY). These coated slides are dried over desiccant for 1 min, then the Pioloform film is released from the edges with a razor blade, and floated onto water. Slot grids are gently pressed onto the surface of the Pioloform, picked up on a clean Parafilm ® coated glass slide, and stored in a dry place under glass. 6. Grids containing serial sections are stained in a small glass (-60 mm diameter)-covered Petri dish containing -3 mm dental wax, melted and cooled in the bottom of the dish. The dental wax has several rows of slits,
80
Feinberg, Szumowski, and Harris
prepared with a clean razor blade. Each grid is partially inserted on one edge into a slit, without damaging the Pioloform coating. Either filtered, saturated uranyl acetate (aq) alone or mixed 1:1 with 100% ethanol, is added to the grids for 5 min. The stain is then poured off, and grids are rinsed with a gentle stream of twice-distilled water, for several minutes. 7. The grids are wicked dry with filter paper and then Reynold's lead stain is applied for 5 min. To reduce excess CO 2 contamination, a 5.5-cmdiameter filter paper, moistened with 0.02 N NaOH, is adhered to the inner side of the glass Petri dish cover. The stain is then poured off, and grids are gently rinsed with 0.02 N NaOH, followed by twice distilled water and wicked dry. Grids are allowed to air-dry thoroughly. 8. Each grid is loaded into a gimbol set (Advanced Microscopy Techniques, Danvers, MA), or glued to a hexagonal ring (Gatan Inc., Warrendale, PA) and stored in a labeled gelatin capsule.
Serial Section Electron Microscopy Serial thin sections of uniform section thickness are desired for accurate measurement of synapses in three dimensions. Many factors will affect the uniformity of section thickness (Hyatt, 1989). The diamond trimming tools can be used to reliably prepare a small trapezoid for serial thin-sectioning. The square-shaped trim tool produces a rectangular mesa, resulting in serial sections that are exactly the same size. The 45 ° angled trim tool produces a pyramid, but the sections become sequentially larger, thereby making it more difficult to map the location of objects through serial sections during photography. Of the two, however, the angled trim tool produces sections with more uniform section thickness. Tissue, cured in either cylindrical BEEM capsules or coffin-shaped, fiat molds, is used, depending on the orientation of the tissue. Anecdotally, the cylindrical blocks seem to be held more securely in a microtome chuck, and are somewhat more reliable in producing uniform section thickness. However, BEEM capsules are not convenient for orienting tissue perpendicularly to the broad plane, which is needed in many of the authors' studies. Another consideration is the type of resin the tissue is in (Epon vs Epon-Spurr' s). These sectioning-related issues continue to be evaluated in relationship to MWI curing of the Epon. In the authors' experience, blocks cured with MWI can achieve uniform section thickness of a quality equal to those slowly cured for 48 h at 60°C (e.g., Fig. 1). To view a series of consecutive thin sections (--60 nm section thickness) in the same orientation, grids in their gimbols are placed in a rotating-stage specimen holder (EM-SRH, Jeol USA, Peabody, MA) in
M W Fixation of Rat HC Slices
81
100 Fig. 1. Portions of two ribbons from serial sections. The ribbon on the left is from a series prepared by conventional methods, and the series on the right was prepared by MW-enhanced fixation and processing. Although there are some irregularities in each of these ribbons, they demonstrate that the MW processing is comparable to conventional methods.
82
Feinberg, Szumowski, and Harris
the Jeol 1200EX electron microscope (JEOL USA, Inc.). In the same manner, a grid, attached to a Gatan hexagonal ring is placed in the Gatan model 650 rotational holder (Gatan, Inc.) and can be viewed in the JEOL 2010 electron microscope. The grid is rotated as needed, to maintain a reference fiduciary (such as a large cross-sectioned dendrite) in the center of the field. Alternatively, visual maps are drawn and edited as needed, to maintain several objects in the same positions relative to one another across serial sections. A calibration grid (Ernest Fullam) is photographed with each series. Section thickness (st) is determined by measuring the diameters (d) of longitudinally sectioned tubular objects (such as dendrites, mitochondria, or axons) in the micrographs (x-y plane), and counting the number of sections (n) that they span (z-axis), and calculating st = d/n, which is averaged across many measurements obtained throughout the series (e.g., see Shepherd and Harris, 1998). RESULTS Only about 50 gm surrounding the periphery of a tissue block is consistently penetrated by ferrocyanide-reduced Os (Hayat, 1989; Fig. 2A). Potassium ferricyanide~reduced osmium also exhibited a similar limit in stain penetration. The authors found, however, that, if the HC slice was resectioned to 100 gm or less, the neuropil was uniformly stained throughout the section (Fig. 2B). In slice preparations, the quality of the tissue improves in a graded manner, from the cut surfaces toward the middle slice (Fig. 2A; see also Jensen and Harris, 1989). Near the cut surface (Fig. 3A), there are dark, degenerating processes and vacuolated structures indicative of trauma caused by the cutting of the slices (Loberg and Torvik, 1993). Zones of excellent ultrastructure are found at --100-200 gm beneath the slice surface, in conjunction with the best EP (Fig. 2B). Excellent ultrastructural preservation of the neuropil is recognized by intact cytoplasmic and nuclear membranes, uniformly thick postsynaptic densities, a uniform distribution of presynaptic axonal vesicles, distinct microtubules, absence of vacuolation or shrinkage artifacts, and intact mitochondria. Similar results are published elsewhere (Harris et al., 1992; Kirov et al., 1999; Shepherd and Harris, 1998; Sorra and Harris, 1998). If a 400-gm-thick piece of tissue is processed, then, as one moves toward the center of the tissue, the density of staining falls off precipitously (Fig. 2A), so that the region of interest in the center of the block is not well-stained (Fig. 3C,D). This problem was completely overcome
MW Fixation of Rat HC Slices
A
Air Cut Surface
: :~:7 i ~ ~'~~ .~' ~ ~ ~ ° ...... i
)~ :~
m
83
~
.........~;,,.........
m
...............
Net Cut Surface
B
100 ~m
Fig. 2. Improvement in staining quality by reducing the thickness of the tissue. (A) 400-gm-thick HC slice. The "net cut surface" of the slice rested down on a net; the "air cut surface" was at the interface of the media and oxygenated, humidified air. The gray rectangle illustrates the region of optimal tissue quality that is sufficiently far from the cut edges, so that there is little or no damage to the processes (see Figs. 3 and 4). (B) 100-gm section from the middle of another 400-gm HC slice, which was resectioned to achieve complete penetration of the reduced Os to the region of interest. Both sections were processed by MW-enhanced methods, and photographed at x60 in the Jeol EM. They are not toluidine blue stained; hence, the rim around the outer edge of the thicker slice in (A) results from the incomplete penetration by reduced Os. by resectioning the slice to 70-100 gm thickness" Then, all portions of the region of interest have uniformly high-intensity staining of the plasma membranes (Fig. 4A-C). The quality of the neuropil in these sections is comparable to or better than, that obtained from perfusionfixed brain processed by standard methods. DISCUSSION MWI during tissue fixation and processing has greatly facilitated research (Kirov and Harris, 1999; Fiala et al., 1998; Shepherd and Harris, 1998; Sorra and Harris, 1998). The rapid fixation produces superior tissue preservation in HC slices; MW-enhanced processing has greatly
Feinberg, Szumowski, and Harris
84
:
i~
i¸
....
•
,~ ~
~
~
Fig. 3. Photographs from the 400-l.tm slice in Fig. 2A. (A) Damaged tissue near to the air cut surface of the HC slice. More cells die at the net surface, because it is less well oxygenated. All photographs in (B-D) are located 144 ~m from the air-cut surface of the slice, in the region of optimal tissue quality, as demonstrated both physiologically and morphologically. However, photographs in C and D have lower membrane staining, because of the lack of penetration by the reduced Os. The lower contrast and staining quality can result in broken membranes and less-optimal membrane staining for serial EM.
M W Fixation of Rat HC Slices
........ :~
t~.
.i~.74~i~~ ~
85
TM
;:
Fig. 4. All images photographed from the center of a resectioned HC slice are of high contrast and optimal staining quality of biological membranes through serial EM sections. Pictures in (A-C) are from the center of a 100-gm section through the middle of a HC slice, from a postnatal d 21 rat, fixed and processed by the enhanced MW procedures. The picture in (I)) is from a perfusion-fixed postnatal d 21 rat processed by enhanced MW procedures, although using a 400gm section. This picture is from the edge of the section near where some of the reduced Os would have penetrated the tissue section, and is similar to Fig. 2C.
86
Feinberg, Szumowski, and Harris
speeded the rate at which the authors achieve results. The mechanisms underlying MW-enhanced fixation are not fully understood, but they apparently result from enhanced diffusion and reaction rates, facilitated by internal heating (Kok and Boon, 1990; Evers et al., 1998; Horobin, 1998; Login and Dvorak, 1985; 1994b,c). MW-enhanced fixation has set new, higher standards in this laboratory for achieving quality tissue preservation in HC slices following electrophysiological and/or pharmacological experimentation. When using the household-type MW, it is important to follow the guidelines for careful calibration and standardization for fixation of biological tissues. Many factors can affect MWI, such as the unpredictability of MW power output and field distribution, nonuniformity of MW heating within the oven cavity, the position of water loads, sample size, fixative volume, and so on (Login and Dvorak 1994a, 1994b, 1994c; Login et al., 1998; Login, 1978; Kok and Boon, 1992). The authors overcame these potential problems by using the exact same spot (marked by an "X") in the Amana MW oven, after having calibrated it, and testing that it gave good tissue quality. Postfixation of the slice in aldehyde, following MWI, is another issue. Although the authors found that no postfixation of the slice is necessary for excellent tissue preservation following 8-11 s MWI (Jensen and Harris, 1989), preserving the slice for 1-12 h in the mixed aldehydes serves to stabilize the tissue during subsequent handling. Several investigators contend that postfixation of tissue, following MWI, is critical for complete chemical crosslinking of proteins with aldehydes (Mizuhira et al., 1990; Ohtani et al., 1990). The CaC12 and MgC12 also serve to maintain membrane integrity during MWI fixation (Ohtani, 1991). Until recently, the tissue-processing for EM involved a long procedure lasting several days. The programmable MW oven designed for laboratory use (Pelco 3450, 900 W, 2.45 GHz, with model 3420 load cooler attachment), has reduced the time required for this procedure to a few hours. ACKNOWLEDGMENTS This work was supported by National Institutes of Health grants, NS21184, NS33574, MH/DA57351 which is funded jointly by National Institute of Mental Health, National Institute of Drug Abuse, National Aeronautics and Space Administration (K.M.H.), and the Mental Retar-
M W Fixation of Rat H C Slices
87
dation Research Center grant P30-HD18655 (Dr. Joseph Volpe, PI). The authors thank Alex Goddard for preparing some of the HC slices and Betsy Velazquez for assistance in preparing the manuscript, and Dr. John Fiala for editorial assistance. REFERENCES Andersen P, Silfvenius H, Sundberg SH, Sveen O (1980) Comparison of distal and proximal dendritic synapses on CA1 pyramids in guinea-pig hippocampal slices in vitro. J Physiol (Lond.) 307;273-299. Chang FLF, Greenough WT (1984) Transient and enduring morphological correlates of synaptic activity and efficacy change in the rat hippocampal slice. Brain Res 309:35--46. Evers P, Uylings HB, Suurmeijer AJ (1998) Antigen retrieval in formaldehyde-fixed human brain tissue. Methods 15:133-140. Feirabend HK, Kok P, Choufoer H, Ploeger S (1994) Preservation of myelinated fibers for electron microscopy: a qualitative comparison of aldehyde fixation, microwave stabilisation and other procedures all completed by osmication. J Neurosci Methods 55:137-153. Fiala JC, Feinberg M, Popov V, Harris KM (1998) Synaptogenesis via dendritic filopodia in developing hippocampal area CA1. J Neurosci 18:8900-8911. Giberson RT, Demaree RS, Jr. (1995) Microwave fixation: understanding the variables to achieve rapid reproducible results. Microsc Res Tech 32:246-254. Giberson RT, Demaree RS Jr., Nordhausen RW (1997) Four-hour processing of clinical/ diagnostic specimens for electron microscopy using microwave technique. J Vet Diagn Invest 9:61-67. Harris KM, Cruce WLR, Greenough WT, Teyler TJ (1980) A Golgi impregnation technique for thin brain slices maintained in vitro. J Neurosci Methods 2:363-371. Harris KM, Teyler TJ (1984) Developmental onset of long-term potentiation in area CA1 of the rat hippocampus. J Physiol (London) 346:27-48. Harris KM, Jensen FE, Tsao B (1992) Three-dimensional structure of dendritic spines and synapses in rat hippocampus (CA1) at postnatal day 15 and adult ages: implications for the maturation of synaptic physiology and long-term potentiation. J Neurosci 12:2685-2705. Hayat MA (1989) Positive staining. In: Principles and Techniques of Electron Microscopy. CRC Press, Boca Raton, FL, pp. 272, 273. Horobin RW (1998) Problems and artifacts of microwave accelerated procedures in neurohistotechnology and resolutions. Methods 15:101-106. Jackson PS, Suppes T, Harris KM (1993) Stereotypical changes in the pattern and duration of long-term potentiation at postnatal days 11 and 15 in the rat hippocampus. J Neurophysio170:1412-1419. Jensen FE, Harris KM (1989) Preservation of neuronal ultrastructure in hippocampal slices using rapid microwave-enhanced fixation. J Neurosci Methods 29:217-230. Kirov SA, Sorra KE, Harris KM (1999) Slices have more synapses than perfusion-fixed hippocampus from both young and mature rats. J Neurosci 19:2876-2886. Kirov SA, Harris KM (1999) Dendrites are more spiny on mature hippocampal neurons when synapses are inactivated. Nat Neurosci 2(10):878-883.
88
Feinberg, Szumowski, and Harris
Kok LP, Boon ME (1992) Microwave Cookbook for Microscopists. Coulomb, Leiden, The Netherlands. Kok LP, Boon ME (1990) Microwaves for microscopy. J Microsc 158:291-322. Loberg EM, Torvik A (1993) Distinction between artefactually shrunken and truly degenerated 'dark' neurons by in situ fixation with microwave irradiation. Neuropathol Appl Neurobiol 19:359-363. Login GR (1978) Microwave fixation versus formalin fixation of surgical and autopsy tissue. Am J Med Techno144:435-437. Login GR, Dvorak AM (1985) Microwave energy fixation for electron microscopy. Am J Pathol 120:230-243. Login GR, Dvorak AM (1994a) Application of microwave fixation techniques in pathology to neuroscience studies: a review. J Neurosci Methods 55:173-182. Login GR, Dvorak AM (1994b) Methods of microwave fixation for microscopy. A review ofresearch and clinical applications: 1970-1992. Prog Histochem Cytochem 27:1-127. Login GR, Dvorak AM (1994c) The Microwave Tool Book. Beth Israel Hospital, Boston. Login GR, Leonard JB, Dvorak AM (1998) Calibration and standardization of microwave ovens for fixation of brain and peripheral nerve tissue. Methods 15:107-117. Mizuhira V, Notoya M, Hasegawa H (1990) New tissue fixation methods for cytochemistry using microwave irradiation. I. General remarks. Acta Histochem Cytochem 23:501-523. Ohtani H, Naganuma H, Nagura H (1990) Microwave-stimulated fixation for histocytochemistry: application to surgical pathology and preembedding immunoelectron microscopy. Acta Histochem Cytochem 23:585-597. Ohtani H (1991) Microwave-stimulated fixation for preembedding immunoelectron microscopy. Eur J Morpho129:64-67. Petukhov VV, Popov VI (1986) Quantitative analysis of ultrastructural changes in synapses of the rat hippocampal field CA3 in vitro in different functional states. Neuroscience 18:823-835. Reid KH, Edmonds HL, Jr., Schurr A, Tseng MT, West CA (1988) Pitfalls in the use of brain slices. Progr Neurobio131:1-18. Shepherd GMG, Harris KM (1998) Three-dimensional structure and composition of CA3-->CA 1 axons in rat hippocampal slices: implications for presynaptic connectivity and compartmentalization. J Neurosci 18:8300-8310. Sorra KE, Harris KM (1998) Stability in synapse number and size at 2 hr after long-term potentiation in hippocampal area CA1. J Neurosci 18:658-671. White EL, Amitai Y, Gutnick MJ (1994) Comparison of synapses onto the somata of intrinsically bursting and regular spiking neurons in layer V of rat SmI cortex. J Comp Neuro1342:1-14.
8
Microwave Processing Techniques for Biological Samples in a Service Laboratory Lou Ann Miller
INTRODUCTION The author's interest in microwave (MW) processing of biological samples for electron microscopy was initially stimulated by presentations and papers by Dr. Gary Login (Login and Dvorak, 1988, 1993, 1994.) The work of Giberson and Demaree demonstrated that routine MW processing was possible (Giberson and Demaree, 1995; Giberson et al., 1997; Giberson and Demaree, 1999). The following method is currently used in this university facility, which services both research and diagnostic clientele, and includes some adaptations for hard-to-process specimens, as well. Although some of those MW techniques are reserved only for the most experienced user, this tissueprocessing method is routinely used by staff members and students who train with us. This method is easily standardized for consistent results. The author' s method does require more hands-on work than that reported by Giberson et al. (1997), but these MW techniques are used to gain specimen quality, with some savings in overall turnaround time, as well. Some cellular components remain in the sample better with MW techniques than with standard procedures, and myelin preservation, fat retention, and filament structure are significantly better in MW preparations vs standard preparations. Hard-to-embed or -infiltrate samples fare better with MW embedding techniques than with room temperature (RT) methods. From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ
89
Miller
90
It has also been the author' s experience that fragile antigen preservation for electron microscopy immunolabeling is much improved with MW fixation, and that, in a few cases, was the only method found to actually work. For specimen quality and antigen preservation, this facility forsakes any automated methods or more routine batching, and spend hands-on time to use MW procedures, all the way to the final epoxyembedding steps, even with a workload of over 400 projects a year. MATERIALS AND METHODS
Materials 1. Microwave: Pelco 3450 Max Microwave Oven (Ted Pella). 2. Chemicals: a. Karnovsky's fixative in Sorensen' s phosphate buffer (2% glutaraldehyde, 2.5% paraformaldehyde). b. 0.1 M cacodylate buffer, no ions added. c. 2 % (aq) Osmium tetroxide (OsO4). d. 3% Potassium ferricyanide. e. 7% aq Uranyl acetate (UA). f. Ethanol Series: 25, 50, 75, 95, 100%. g. Acetonitrile (ACE). h. Lx112 epoxy (Ladd Industries). i. Dodecenyl succinic anhydride, anhydride, nadic methyl anhydride, tri(dimethylaminomethyl)phenol (Ted Pella). 3. Miscellaneous: a. Laser-printed block tags. b. Silicon embedding molds, type B (Electron Microscopy Sciences). c. Histology dryer oven (Lipshaw). d. Rotator that allows the specimen vials to remain upright.
Methods 1. Sample preparation: Tissues that need to be oriented should be cut to look like a slice of bread laying flat, in the proper orientation. Glass vials, from 2 to 4 mL are used for processing (e.g., 1-dram threaded vial, Fisher 03339-25B). Fluid volumes should be kept to a minimum: This reduces solution heating, which is an undesirable effect with MW-assisted processing. In some cases, volumes will be only 5-6 drops, or enough to barely cover the sample. The tissue must be cut small, and should have one dimension thicker than 0.75-1.5 mm. The sample, in fixative, is placed on an ice-water slurry to cool during the MW warm up. 2. Sample and water load placement: This is a variation of Giberson and Demaree (1995) and Login and Dvorak (1994), in determining hot and
MW Techniques in Service Lab
91
cold spots in the MW cavity. Liquid crystal sheets (Edmund Scientific) are used, in place of neon bulbs or other methods, to determine the cooler, less-active spots in the cavity. Areas that did not exceed 40°C, near the water loads, were chosen for sample placement. The water loads must be in place during the determination. For water load and sample placement, the same areas should always be used. 3-4 samples may be placed into the oven at one time for processing, however, they should be in the same area, but separated. There should always be two water loads in the MW. Both should be 350 mL. With this facility's MW, one of these loads is constantly recirculated and cooled, so that it does not need changing. 3. MW warm-up: With the sample cooling, the MW is run for 1 min prior to the beginning of processing. The uncirculated water load is changed after the warm-up.
Processing Procedure 1. Primary Fixation: Do the following steps 4x, changing the water load each time. a. Remove enough of the fixative, so that the sample is just covered. b. Place on ice for 30 s. c. With a Fresh water load in the proper place, place the uncapped sample in a designated area. d. MW (8 s on, 20 s off, 8 s on). The author programs the MW for this. e. Chill immediately for at least 20 s, and repeat this entire set of steps for a total of 4x without changing fixative solution. 2. Wash: Do the following steps 3x. a. Remove fixative/liquid from the tissue. b. Immediately add 0.1Mcacodylate buffer (containing no added ions). c. Mix well. 3. OsO 4 (Secondary Fixative): Repeat every 5 mins for 5x, changing water loads each time. a. Add just enough 2% aq Os to cover the sample. This will be all the OsO 4 needed. b. Briefly chill the sample on ice for 20-30 s. c. With a fresh water load in position, place uncapped sample in designated area. d. MW 8 s on, 20 s off, 8 s on; repeat, then chill for 20 s. e. Soak 5 min at RT, then repeat without changing the Os. 4. KCN-OsO4: Leave OsO 4 in vial and add an equal volume of 3% aq potassium ferricyanide. Rotate for 15 min at RT (no MW exposure). 5. Washwater: Rinse 3x briefly with distilled water.
92
Miller
6. UA en b l o c stain: a. Remove all water from the sample, and add just enough filtered, saturated UA to cover the sample. No chilling is needed. b. MW 8 s on, 20 s off, 8 s on. This improves the UA en b l o c staining. c. Cover, or dim lights and rotate for 30 min at RT. 7. Dehydration: a. Retain the recirculating water and remove the load beaker in the MW. b. Remove the UA, and rinse the tissue with 10% ethanol for 1 min. c. Rotate at RT 8 min each in a procession of ethyl alcohol concentrations: 25, 50, 75, 95, 2 x 100%. d. Rotate at RT 8 min in a 1:1 mixture of 100% ethyl alcohol and ACE, then at RT 2 x 8 min in pure ACE. Fast Alternative: The author has done this many times, and it works well. For each ethanol step, MW the sample for 25 s, then incubate for 2 min, and proceed to the next step. Note: Do not MW the ACE steps. Caution: This method should not be used with tissues sensitive to heat, such as liver and most plants. 8. Infiltration: a. Add a 1:1 mixture of ACE and epoxy resin to the sample. Vortex the sample, but make sure all the tissue ends up back at the bottom covered by the epoxy mixture. b. MW for 20 s. (Do not MW longer: the ACE will boil.) c. Vortex, and MW again, then rotate at RT for 10 min. d. Repeat this process 2x. 9. Infiltration 1:4 step: a. The epoxy is removed, and a mixture of 1:3 (ACE:epoxy embedding mixture) is added to the sample. b. Vortex, and make sure all the tissue ends up back at the bottom, covered by the epoxy mixture. c. MW for 20 s. (Do not MW longer: the ACE will boil.) d. Vortex, and MW again, then continue infiltrating, while rotating for 20 min. e. Repeat this process 2×. 10. Infiltration Pure Epoxy Embedding Mixture: a. Remove the epoxy-ACE mixture, and add pure epoxy embedding mixture to the sample. The volume is important: Too much, and the sample gets too hot, which can cause the resin to polymerize in the vial. It is best to keep the volume to about one-fourth of the vial and the MW times to less than 40 s. b. Vortex the sample, and make sure all the tissue is at the bottom, covered by the epoxy mixture.
MW Techniques in Service Lab
93
c. MW for 30 s. d. Vortex, and MW again then continue infiltrating, while rotating for 30 min. e. Replace with fresh epoxy embedding mixture, repeat MW steps and continue infiltrating at least 1 h more. 11. Polymerization: a. Fill the silicon embedding mold tips with one drop of epoxy mixture, then gently place the tissue on a Kimwipe ®, using a wooden tool. Allow a few moments for draining off excess epoxy, but do not allow the tissue to dry out. Gently lift the tissue with a wooden tool, and place into the epoxy drop in the bottom of the mold well. Check for air bubbles around the sample, and disperse with a wooden tool. b. Check orientation of the tissue, and nudge into position with a wooden tool, if necessary. c. Fill the mold with pure epoxy mixture, just short of the top, while avoiding specimen movement and air bubbles. d. Place the mold in the oven, set at 85°C (95°C, if using a Histodryer) overnight (ON) or 8-15 h. Do not place mold near walls or too close to any other mold. Allow the blocks to cool at RT for 20 min, before removing them from the molds. With silicon molds, removing the castings, while the mold is still hot, will damage the mold. Note: If more polymerization time is needed, set the oven timer to start several hours after work. On long weekends: Set the oven at 85°C, but set the timer to run only 6-8 h/day. (The author has observed that 2-3 d polymerization at 60°C will harden blocks, but cutting properties are better, using the 85°C method.) Rarely is there a procedure that will cover every tissue. Adjustments can and should be made for problematic tissue. The following is a list of ways this lab adjusts the procedure for different tissues: 1. Bone" Bone requires decalcification, which is usually done by fixing the tissue, washing in buffer, then using an ON incubation in ethylenediamine tetraacetic acid. Depending upon the tissue size and the amount of bone, the time for MW fixation should not increase, but the number of MW repetitions should. Primary fixation may receive, e.g., 8 MW exposures, with ice-cooling between each. If the tissue is deeper inside bone, the Os step might be prolonged by an additional 20-40 min. Epoxy infiltration will include sessions under vacuum (vac) and ON infiltration in pure epoxy. 2. Cell cultures" These samples often require shorter processing times, with significantly reduced times for very fragile cells. This lab has found that some cell cultures do not stain well; therefore, do not shorten the Os time. The shortcuts taken are in shorter times for
94
3.
4.
5.
6. 7.
8.
Miller dehydration (4-5 min), infiltration is reduced by 5 min in the 1:1 and 1:4 steps. For embedding in capsules, cell suspensions are placed in small BEEM capsules that are capped and placed in 1.5 mL microcentrifuge tubes (MTs), and spun down for 8-10 min. The entire capsule complex is placed into a 90°C oven ON. Robust cells are pelleted in MTs. A shaved wooden stick is used to mix the sample in each new solution. The suspension is spun down before removal of each solution. This is done to preserve quantity, to provide better mixing, and to give better concentration and processing than when using agar. Pellets should not be deeper than 1.5 mm from the bottom of a 1.5-mL MT. Extremely thin pellets are processed in 1.5-mL MTs. The pellets are not disturbed, and are embedded in the same tube. Only the tip of the MT should be filled and polymerized, allowing the tip to be glued to a blank block and sectioned as a normal block. Cell cultures on slides: Plastic slides, with cells growing on them, are processed using incubation wells. (Do not use propylene oxide as a transition solvent). Processing occurs in disposable Petri dishes, with the slide supported on broken wooden sticks to prevent the slide from touching the bottom of the Petri dish. The slides are covered with just enough solution to cover the cells. The slide can be placed on ice, and the back of the slide is wiped dry before microwaving, to prevent excessive heating. Slides should be touchtested on the bottom side, using gloves, after each MW step, to be sure that there is no overheating during the fixation steps. Gut: Treatment of gut tissue may require longer OsO 4 times and longer resin-infiltration times. As a rule, this lab lengthens the Os step from 25 min to 45-55 min for all gut tissue. Gut seems to need this extra Os time to prevent centers in the tissue from staining differently. This center artifact is more prominent by light microscopy than by electron microscopy. The use of vac and ON pure epoxy mixture infiltrations have been useful. Immunogold: Much care must be taken to prevent excess heating. Shorten the MW time by half to 4 s instead of 8 s at a time, with more repetitions and longer cooling times on ice. Typically, this lab only uses the MW for fixation; other steps are done on the bench. Liver: Do not run the UA en b l o c step: It will leach out the glycogen. The standard procedure is fine for all other steps. Nerve: Nerve is very sensitive to alcohols. Start dehydration in 50%, and do not lengthen incubation times. Test nerve, using a pilot study, before running the accelerated dehydration in the MW. Plant: Plant tissues require longer incubation and MW times and the use of vac. The author has used as many as 10-12 MW exposures during MW fixation. Os fixation can be 1-1.5 h, microwaving every 10 min.
MW Techniques in Service Lab
95
Accelerated dehydration is used with caution, because chlorophyll extraction may happen at this step. Epoxy infiltration usually is extended, in the 1:1 and 1:4 steps, to 1 h or more, microwaving every 10 min or so; vac is used for 50% of the infiltration time. Pure epoxy steps also involve microwaving about every 15 min. Vac may be used ON in a desiccator hooked to house vac. The epoxy is changed and infiltrated all day, or longer, for difficult samples. 9. Skin: Skin may require longer incubation times; usually, this is a slightly lengthened dehydration process, and more extended epoxy infiltration. RESULTS Because this lab receives so many different types of samples, it is imperative to have a procedure that will work, even for very difficult specimens. MW techniques work for delicate cell cultures and for hardy soybean seeds. The MW enables hand-processing and variation of that process for each specimen, yet achievement of a quick turnaround time. With years of prior experience to compare, this lab has actually experienced better results and better sectioning with MW techniques than with the more standard procedures. In diagnostic work, speed with accuracy is important. Adding a fast polymerization technique to the above method may allow a small batch of samples to be embedd, and to have thick sections done in the same shift. With more than one person on staff, two biopsies can be processed from tissue vial to prints, in less than two shifts, while working on other projects at the same time. Figure 1 shows the result of our MW technique, using extended Os and epoxy infiltration times to preserve soybean leaf cells. Hard-toembed samples, such as kidney stones, preserve well (Fig. 2). Nerve tissue containing myelin can separate, with poor processing. Figure 3 is an example of well-preserved MW-processed nerve tissue. Muscle fibers are distinct, and the vacuole contents near the mitochondria are well-preserved in Fig. 4. In diagnostic work, low magnifications are important. Figure 5 shows a low-magnification view of kidney tubules. Figure 6 shows that even very small structures can be well-preserved; these cilia cross-sections are from a canine nasal biopsy. DISCUSSION This quick method has been used for almost 4 yr. This lab has done hundreds of projects, and gotten consistent results with many types of samples. Even unique structures, such as kidney stones (Fig. 2) are well
96
Miller
~
0
Fig. 1. Soybean leaf cells MW-processed using extended times and vacuum. B a r = 1 ktm.
Fig. 2. Kidney stones MW-processed. Bar = 1 ktm.
M W Techniques in Service Lab
Fig. 3. MW-processed myelin sheath around nerve. Bar = 0.5 ktm.
Fig. 4. MW-processed skeletal muscle. Bar = 0.5 gm.
97
98
Miller
Fig. 5. Example of low magnification of MW-processed kidney tubules. Bar= 5 ~tm. preserved. Among the first things noticed, when this lab switched to MW processing, was the higher amount of lipids retained, better myelin preservation, and the increased amount of cellular fibers preserved. The MW oven used in this work has proven to be so stable over the years that a temperature (temp) probe is no longer used to restrict temp rise. During the first 2 yr of use, the author et al. checked temp at each step, and since the temp never exceeded 40°C, there was confidence that this procedure could be trusted, and the use of the temp probe was discontinued. Originally, the author et al. checked for hot and cold spots, using agar-saline-Giemsa blocks (Login and Dvorak, 1994), or agar-salineGiemsa plates. Results were also compared with liquid crystal temp sheets with 25-50°C range and the neon bulb array (Ted Pella). The cooler spots that did not exceed 40°C, near the water load were chosen for sample placement. For the fixative portion of the procedure, the sample is iced to keep it cool. At this point in the procedure, heat can cause the tissue to look
MW Techniques in Service Lab
99
Fig. 6. MW-processed canine nasal cilia. Bar = 0.1 gm.
soft, with poor resolution. The MW timing is an on-and-off type of procedure. We have found it much more beneficial for the tissue quality to do many shorter bursts of microwaving, getting the benefits of MW fixation without the harmful effects of building heat. So the MW is programmed to perform 8 s of microwaving, 20 s of waiting, and 8 s more of microwaving. Dehydration is different: By this time, the tissue is well fixed, and heat is more tolerable. This is why, starting at the dehydration point, the extra water load is removed, but the recirculating water bath remains. Solution volumes of dehydration solutions are larger than that allowed for fixation. MW times for infiltration are more fussy. If the 1"1 or 1:4 epoxy mixture is allowed to heat too much, the ACE will actually boil, disrupting structure, posing a danger for those who close the vial and try to mix the sample. At this point, the vial is only microwaved for 15-20 s. Pure epoxy resin also needs precautions. If the volume is too great, the entire sample can harden in 40 s of microwaving. Such hardening does not mean the sample is ruined. Many samples can be retrieved by
100
Miller
chopping up the epoxy and re-embedding into blocks. However, if using a glass vial, there is trouble. Hardened epoxy will not separate from the glass, and glass shards will damage knives. For this reason, only add a small amount of epoxy to the tissue, and only MW for a maximum of 30 s. This procedure, at least for our epoxy formulations, works well time after time. In fact, the author has not had the premature polymerizing problem at all, since lowering the volume and making the MW times 30 s or less. For routine embedding, this lab uses a histology drying oven. The entire oven can be put under the hood, and the airflow of the oven allows use of 85 or 90°C temps. As long as the molds do not touch each other, and do not touch the edge of the oven, the BEEM capsule molds will not melt. The author et al. have found that, with their epoxy formulations, better block quality is obtained with the 90°C method than with the 60°C method. For standard ovens, the 85°C method works well. This would be expected to vary for different epoxy formulations. ACKNOWLEDGMENTS The author would like to thank Dr. Birute Jakstys for her guidance, and for editing help with this chapter. REFERENCES Giberson RT, Demaree RS Jr. (1995) Microwave fixation: understanding the variables to achieve rapid, reproducible results. Microsc Res Tech 32:246-254. Giberson RT, Demaree RS Jr., Nordhausen RW (1997) Four-hour processing of clinical/ diagnostic specimens for electron microscopy. J Vet Diagn Invest 9:61-67. Giberson RT, Demaree, RS (1999) Microwave processing techniques for electron microscopy: a four-hour protocol. In: (Hajibagheri N, ed.) Methods in Molecular Biology, Humana, Totowa, NJ, pp. 145-158. Login GR, Dvorak, AM (1988) Microwave fixation provides excellent preservation of tissue, cells and antigens for light and electron microscopy. Histochem J20:373-387. Login GR, Dvorak, AM (1993) Review of rapid microwave fixation technology: its expanding niche in morphologic studies. Scanning 15:58-66. Login GR, Dvorak, AM (1994) The Microwave Toolbook. A Practical Guide for Microscopists. Beth Israel Hospital, Boston, MA.
9
Microwave-Accelerated Decalchqcation UsefulMethodsfor Researchand Clinical Laboratories
½"ctoriaJ. Madden
INTRODUCTION Morphological examination of calcified structures, by either light microscopy (LM) or electron microscopy (EM) often requires a decalcification (decal) step after fixation. Further processing steps must await the completion of the decal process. The time required for full decal with ethylenediamine tetraacetic acid (EDTA), a calcium (Ca) chelation agent, generally takes days to months, using routine methods. This time lag can be detrimental to both tissue morphology and antigenicity, and can hinder productivity and delay diagnostic results in clinical settings. Fortunately, decal times can be shortened dramatically with the use of a laboratory microwave (MW) oven (Keithley et al., 2000; Madden and Henson, 1997; Louw et al., 1994; Faria et al., 1992; Hellstrom and Nilsson, 1992; Ng and Ng, 1992; Roncaroli et al., 1991). Most of the tissues sent to this lab are research vs clinical in origin, and are processed for routine and immunoelectron microscopy. For the decal step, EDTA is used as the chelating agent, because it is less damaging to the ultrastructure of the specimens, compared to other commonly used acids, such as nitric or formic (Baird et al., 1967). The decal procedure presented here (Madden and Henson, 1997) has been updated to include the processing of multiple specimens. Also, a modified protocol is offered for more efficient sample cooling, to further help standardize the process of MW-accelerated decal. From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ 101
102
Madden MATERIALS A N D M E T H O D S
Three different MW-accelerated decal protocols are described. Method 1 is the basic protocol for single samples, method 2 is the basic protocol for multiple samples, and method 3 is the basic protocol with direct specimen cooling. All three methods were performed using a Model 3450 800W Microwave Processor (Ted Pella, Redding, CA), with the following features: load cooler, for water recirculation and cooling within the MW cavity; air bubbler, for solution agitation within the sample processing vial; Teflon-coated temperature (temp) probe, to prevent corrosion of the metal by EDTA; and multiple inlet/outlet ports, for external access to the MW cavity, for recirculation and cooling of fluids during MW procedures. Sample fixation is done prior to starting decal. For ultrastructural preservation, tissues requiring decal should be fixed in an aldehydecontaining fixative. For LM, 10% neutral buffered formalin, 4% buffered paraformaldehyde, or Bouin' s fixative are good choices. For routine EM, a fixative containing at least 1% glutaraldehyde is optimal. Fixation concentrations and times are tissue- and application-specific; therefore, the appropriate fixation method should be determined and carried out prior to decal. Processing after the decal step can be done either by routine or MW-assisted methods for LM (see Chapter 16) or EM (Giberson et al., 1997). For all three decal methods, a buffered 0.1 M EDTA solution (F.W. 292.3) was used.
Recipe for 1 L 33.8 g Na2PO4.7H20, 3.3 g Na(PO4)2.H20, 29.2 g EDTA; mix in 950 mL deionized water heated to 50°C in the MW, to speed dissolution; adjust pH to 7.4 with either 1 N NaOH or 1 N HC1, after the final volume has been adjusted to 1000 mL. Filter the solution with a 0.22 ~tm filter before use.
MW Calibration Prior to beginning MW-accelerated decal, a determination of MW energy intensity is undertaken, using a procedure first described by Login and Dvorak (1994)(Fig. 1). A neon bulb array (available from most EM suppliers) is used for this procedure. 1. Fill an 800 mL (Pyrex®or plastic) beaker with 700 mL water (water load). 2. Place the beaker in the left front corner of the MW oven cavity, and insert the load cooler hoses into the beaker, and turn the pump on. Adjust water volume in the beaker to 700 mL.
MW-Accelerated Decalcification
103
Fig. 1. Calibration of MW energy distribution was performed using a neon bulb array (Login and Dvorak, 1994). The array was placed in the center of the oven cavity, and the water load moved, until a large cold spot (no bulb illumination) was created. 3. Set the temp control on the load cooler to 30°C. 4. Place the neon bulb array in the center of the MW chamber. Note: If a platform will be under the specimen during the processing cycles, be sure to place the platform under the neon bulb array for the calibration procedure. 5. Irradiate the neon bulb array for 1 min at full power (100% power), and observe the illumination pattern of the bulbs. The bulbs that light indicate hot spots; bulbs that do not light indicate cold spots. 6. Move the water load to create a large cold spot (Fig. 1). The cold spot is the area where the processing containers will be placed (Giberson and Demaree, 1995).
Method 1: Basic Protocolfor Single Samples (Fig. 2) The following materials are required: 800 mL (Pyrex or plastic) beaker; 20-mL glass scintillation vial(s) with cap (liner removed); 100-ram polypropylene (PP) Petri dish bottom (cat. no. 36136, Ted Pella, Inc.) inverted for use as a specimen platform; 25-mm round biopsy sponges (cat. no. 62325-01, Electron Microscopy Sciences); decal solution. The decalcification process proceeds as follows:
Fig. 2. Decalcification method 1. (A) 20-mL glass scintillation vial with modified cap, which is used for M W decalcification. One hole in the cap is for the temp probe and the other for the air bubbler hose. (B) Recirculated water load was positioned in the left front corner of the oven cavity. The specimen vial and platform were placed in the cold spot, with the temp probe (opposite the specimen and touching the bottom of the vial) and air hose (0.25 in. below fluid surface) inserted into the vial.
MW-Accelerated Decalcification
105
1. Fill the 800-mL beaker with 700 mL cold water, and place the load cooler hoses in the beaker. 2. Calibrate the MW oven. 3. Place an inverted Petri dish (100-mm PP) on the oven floor in the cold spot. The inverted dish serves as a platform to elevate the specimen vials, and to isolate them from heat that is retained in the oven floor. 4. Place two biopsy sponges in the bottom of a 20-mL scintillation vial, and fill with 10 mL buffered EDTA. 5. Remove the liner from the cap of the scintillation vial, and make two holes in the top, with a hot dissecting needle or small drill bit, to accommodate the temp probe and air tubing. 6. Put the specimen in the vial, and cap the vial. Note: If a cap is not available for the vial being used, cover the vial with any lab plastic sealing film that clings tightly to glass. Make two slits in the top, for the temp probe and air line. 7. Place the temp probe to the bottom of the vial, away from the specimen. 8. Insert the air tubing into the decal solution, where the bubbles will agitate the fluid, without introducing bubbles into the sample. 9. Set the temp on the water load cooler to 30°C, turn on the pump, and check to be sure the water level is at 700 mL. 10. Turn on the air-bubbler pump. 11. Close the oven door, and set the power to the maximum wattage of the oven (100%). 12. Set the temp restriction on the MW to 50°C, and the processing time to 90 min. 13. Turn the oven on, and irradiate the sample. When the processing time is finished, check the sample, to determine if it is decalcified. Note: To determine the decal end point, several methods may be used. The most common methods are to probe the tissue with a thin needle, or by touch (Flower, 1951), contact radiography of the sample using a low-kV X-ray unit (Ng and Ng, 1992), checking the decal solution for Ca by Ca precipitation methods, and visibly checking the sample. Some tissues, such as cochleae, will become translucent when decal is complete. 14. If the sample needs further decal, change the solution, and repeat steps 7-13, until the sample is fully decalcified. The decal process usually requires multiple cycles, sometimes over several days, depending on the size of the specimen and density of bone calcification.
Method 2: Basic Protocol for Multiple Specimens (Fig. 3) The following materials are required: 800-mL (Pyrex or plastic) beaker; tissue baskets with Teflon lids (cat. nos. 36142, 36129, Ted Pella); two each 100-ram PP Petri dishes with lids. The decal process proceeds as follows:
Fig. 3. Decalcification method 2. (A) For processing multiple specimens, flowthrough capsules (Ted Pella, cat. no. 36142) with Teflon lids are used, as a convenient way to separate samples. A 100-mm-diameter PP Petri dish (Ted Pella, cat. no. 36136), with two holes in the lid, served as the container for the decal solution and specimen capsules. (B) Placement of the water load, platform, specimen container, temp probe, and air tubing within the M W cavity (identical to Method 1).
MW-Accelerated Decalcification
107
1. Perform steps 1-3 of Method 1. 2. Fill a Petri dish with decal solution, and fully submerge the flowthrough capsules, with the sample in the capsule and the lid in place. Note: A small piece of paper with specimen identification can be placed in the capsule with the sample. 3. Turn the capped capsules on their sides. This will float the specimens in the decal solution. Make sure that the decal solution covers the capsules, and remove any air bubbles trapped in the capsules. 4. Pierce or drill two small holes in the Petri dish lid, to accommodate the temp probe and air tubing. 5. Cover the Petri dish with the lid, and place it on the platform, in the cold spot. 6. Perform steps 7-14 of Method 1.
Method 3: Processing with Direct Cooling of the Specimen (Fig. 4) The following materials are required: PP container with lid, such as Rubbermaid ® brand, with minimum dimensions of 6 × 6 × 2 in. high, holding a vol of 600 mL or more; PP barbed tubing connectors; plastic support racks for scintillation vials (Method 1). The decal process proceeds as follows: 1. Modify the PP container into a circulating water bath by adding two hose connectors to the container. Drill two holes opposite one another, at a height of 1 in. from the bottom of the container. Make the diameter of the hole a close fit for the flush end of the barbed fitting. Apply clear silicone sealant around the rim of the hole in the container, and insert the fitting into the hole. With the fitting in place, seal around both sides, taking care not to clog the openings. Let the silicone sealant cure for 24 h, before using the water bath. 2. Place the container in the middle of the oven cavity, and connect the inlet and outlet tubing from the load cooler directly to the water bath. 3. Place a plastic vial rack in the bottom of the container. The rack should be open enough to allow for circulation of the water around the sides and bottom of the vials. It should also provide enough support to keep the vials from tipping, once the water is added to the bath. If necessary, secure the plastic rack to the bottom of the water bath, beforehand, with a few dabs of silicone adhesive. 4. Fill the water bath almost full with cold water, and start the load cooler pump. Let the water level stabilize, and add or remove water, so that the level is approximately three-fourths of an inch above the inlet and outlet holes in the container. 5. Set up the specimen vials according to step 3 of Method 1, and place the vials in the water bath.
Fig. 4. Decalcification m e t h o d 3. (A) Specimen vials were cooled with circulating water from the load cooler. A water bath, with added inlet and outlet ports, was constructed from aPP storage container and two pp barbed tubing connectors. A plastic rack was placed in the bath, to provide support for the scintillation vials. (B) The inlet and outlet tubes running from the load cooler were h o o k e d directly to the water bath. The bath served as the water load for the oven cavity and as the cooling bath for the samples. The temp p r o b e was placed in a vial containing E D T A solution only.
MW-Accelerated Decalcification
109
Note: Use the same fluid volume for all samples. To prevent evaporation of the decal solution, cover the vials with plastic film or an unlined plastic cap. To prevent pressure buildup caused by heating during the decal process, pierce the plastic film, or loosely cap the vial. 6. Set up a "dummy" scintillation vial, containing the same volume of decal solution as the samples, and place it in the water bath. Use this vial for placement of the temp probe. 7. Perform steps 7-14 of Method 1. RESULTS When the times required for decal, using routine methods vs MW-assisted methods with EDTA, are compared, there are considerable time savings associated with MW-assisted methods (Table 1). Based on the results listed in Table 1, MW methods using EDTA are 3-84× faster than routine processes. Decal times varied, depending on the types and sizes of tissue, with larger and denser bones, such as human and monkey temporal bones, taking a longer period to decalcify than mouse femurs and vertebrae. In this study, tissue ultrastructure was well-preserved in a variety of tissues from different animals, after MW-assisted decal using EDTA. The bone marrow from mouse femurs (Fig. 5), decalcified according to methods 1 and 3, was compact and presented well-preserved megakaryocytes, sinusoidal granulocytes, and nucleated red blood cells in the hematopoietic cords. Inner ear tissues from several animal species remained intact, with excellent cellular preservation, following MW-accelerated decal (Figs. 6A,B; 7; 8A,B; 9A,D). Figure 6A, a crosssection of a rabbit cochlear turn, exhibits a well-preserved organ of Corti, spiral ligament, and vestibular membrane (Reissner' s membrane). The limbal region of a bat cochlea (Fig.6B), decalcified using method 3, demonstrated the close relationship of the bony tissue with the cells and supporting matrix of the limbus. Figures 7 and 8A,B are electron micrographs of an inner hair cell and associated organelles from a macaque cochlea decalcified by method 1. These images demonstrate excellent ultrastructural preservation (results from Madden and Henson, 1997). The hair cell nucleus is round, and the cytoplasm is rich in organelles, such as mitochondria and smooth endoplasmic reticulum. Swelling or shrinkage of the organelles, and extraction of the cytoplasm, frequently seen in routinely processed hair cells, were not observed. Junctional complexes between the hair cell and the adjacent supporting cells are well defined. Figures 8A,B are higher-magnification
23 d 14 m
19-25 d 3-4 wk 2-3 wk
17 190-400
19 <6 24
1.5-2
14-20 9
5
Rat hemimandibles Rat skulls Rabbit temporal bones
Mouse cochleae
Guinea pig cochleae Mouse vertebrae (spinal column) Mouse femurs
3d
1560 h 10 d 6-8 wk
2 mo
Routine decal time
15
Microwave decal time (h)
Macaque monkey temporal bones Guinea pig tibias Human temporal bones
Tissue
0.1 M EDTA in Na phosphate buffer, pH 7.4 10% EDTA, pH 6.8 10% EDTA, pH 6.8 w/2% paraformaldehyde 7% EDTA, pH 7.3 EDTA 0.1 M EDTA in Na phosphate buffer, pH 7.4 0.1 M EDTA in Na phosphate buffer, pH 7.4 10% EDTA, pH 6.8 0.1 M EDTA in Na phosphate buffer, pH 7.4 0.1 M EDTA in Na phosphate buffer, pH 7.4
EDTA Solution
Method used
Methods 1 and 3
Keithley et al., 2000 Method 3
Methods 1 and 2
Faria et al., 1992 Hellstrom and Nilsson, 1992 Method 3
Keithley et al., 2000 Keithley et al., 2000
Madden and Henson, 1997
MW'-Accelerated vs Routine Decal Time Using EDTA as Decalcifying Solution
Table 1
Fig. 5. Bone marrow from mouse femur. 1-ktm semithin sections, stained with toluidine blue, were MW-fixed and decalcified according to method 3. Samples, prior to sectioning, were embedded in LR White resin (London Resin) using M W methods (Giberson et al., 1997). (A) The marrow is compact, exhibiting well-preserved sinusoids (S), hematopoietic cords (*), and megakaryocytes (M). Bar = 40 ktm. (B) A higher magnification of a megakaryocyte possessing a multilobed nucleus (N), with a sharply delineated demarcation membrane (arrowheads) surrounded by platelets (P). Bar = 20 ktm.
b~
i
iii £;~i~iiii~;ili~iii~i~ ~ i~
~:~,:;~.:~..,:.~.:.~........................ ,:
....
. . . .
~,~d ¸~'~¸'•
~ ........................
~
•
' ....
~!~ii!~i!~iiiiiiiiiiiii~ ¸
~,~ * ~, •i ~•~¸~ii~!!~!~i ~ ~i
~:i~ ~iiiiiI~
~'
~~i'~,~,i~iii ~;~'III ~i~~ii ~
,~,~
. ;~!~!il-i~- :i~!!ii!:{~!i i:i:~ik~ Ii:!i!:
Fig. 6. Mouse and bat cochleae. The samples were decalcified, using method 1. (A) Cross- section of the basal turn of a mouse cochlea displays a well-preserved organ of Corti (OC), spiral ligament (SL), and an intact Reissner' s membrane (RM). 1 ~ m epoxy resin section, toluidine blue stain. Bar = 40 ~tm. (B) The spiral limbus of a bat cochlea (Mormoops blainvilli). The microprojections of bone (B) from the osseous spiral lamina, which protrude into the limbus (L) (Henson and Henson, 2000b),
"
L
~i~i~
MW-Accelerated Decalcification
113
Fig. 7. Inner hair cell. An electron micrograph of an inner hair cell (IHC) and supporting cells (S) from a Japanese macaque monkey. The arrows indicate regions shown at higher magnification in Fig. 8. Bar = 3.0 l.tm. Reprinted with permission from Madden VJ and Henson MM, 1997. images of the Golgi complex in the hair cell and the rough endoplasmic reticulum in the supporting cell, respectively. They illustrate the excellent membrane preservation possible following decal in the MW. Nerve tissue surrounded by bone is one of the most difficult tissues to preserve without artifact. Figures 9A-D demonstrate that rapid decal
114
Madden
Fig. 8. Inner hair cell. A higher magnification of the organelles marked in Fig. 7. (A) The Golgi apparatus in this micrograph shows the high quality of membrane structure possible following MW decal. Bar = 0.5 ~m. (B) In the supporting cells next to the inner hair cell, the rough endoplasmic reticulum and microtubules are well-preserved. Bar = 0.5 ~tm. Reprinted with permission from Madden VJ and Henson MM, 1997. using MWs is not deleterious to the structure of the nerves (Figs. 9A,C), spiral ganglion cells (Fig. 9B), and surrounding bone (Fig. 9D). Myelin structure is well-preserved after decal in the MW. Figure 10 demonstrates successful postembedding immunolabeling with a polyclonal antibody (Ab), to peroxidase ofa neutrophil (Fig. 10A) and an eosinophil (Fig. 10B) from mouse bone marrow. Figure 11 demonstrates the distribution of collagen II in the basilar membrane of a rabbit cochlea using a monoclonal Ab to collagen II. In both figures, the ultrastructure of the tissues is well preserved, and the immunolabeling is distinct and specific. DISCUSSION Based on the results shown here and in previous publications, the use of the MW oven accelerates the rate of decal dramatically, without compromising the structure of the tissue at the LM and EM level (Mad-
MW-Accelerated Decalcification
115
Fig. 9. Nerves and ganglion. Nerves, and associated ganglion cells surrounded by bone tissue, are not damaged by MW-accelerated decal. (A) Cross-section of the modiolus of a rabbit cochlea, showing a bundle of myelinated axons from the cochlear nerve (N) surrounded by bone (*). The ground substance and vascular elements of the bone sinuses (arrowheads) are well-retained, with no shrinkage, and the nerve bundle is free from artifacts. 1 ~tm LR White resin section, toluidine blue stain. Bar= 20 ~tm. (B) Rabbit cochlea, showing a section through the spiral ganglion of the cochlear nerve, with well-preserved myelinated nerve fibers (N) and ganglion cells (arrowheads). 1 ~tm LR White resin section, toluidine blue stain. Bar = 20 ~tm. (C) An electron micrograph of a cross-section of myelin sheath from a Japanese macaque cochlear nerve axon. The major dense lines (arrowheads) and intraperiod lines (arrows) are sharply defined. Bar = 50 nm. (D) An electron micrograph of a macaque temporal bone containing an osteocyte (O), with the surrounding bone matrix (B), demonstrates excellent ultrastructural morphology, following MW decal. The cytoplasm and nucleus of the osteocyte are well-preserved, and the cell membrane is prominent. Cytoplasmic processes of the osteocyte maintain close association with the bone (arrow), without shrinkage. Bar = 0.5 ~tm. den and Henson, 1997; Faria et al., 1992; Hellstrom and Nilsson, 1992; Kok and Boon, 1992; Ng and Ng, 1992; Roncaroli et a1.,1991). In addition, MW-accelerated decal does not diminish the antigenicity of the tissue, or make it unusable for in situ hybridization techniques
Fig. 10. Granulocytic cells. Electron micrographs of granulocytes from mouse bone marrow illustrate that antigenicity was retained after M W decal. Labeling was specific with minimal background and no evidence of diffusion of the antigen. (A) Neutrophil stained by the indirect postembedding immunogold method with a rabbit antihuman myeloperoxidase primary Ab and goat antirabbit immunoglobulin G 15 nm gold secondary antibody. The gold label was present in the azurophilic granules of the neutrophil. Bar = 1 ktm (B) An eosinophil from the same preparation as (A). The polyclonal myeloperoxidase Ab crossreacted with the eosinophilic peroxidase, shown here in the eosinophilic granules (arrows) of the cell. Note that the basic protein crystalline core of the granules (asterisk) did not react with the Ab. Bar = 1 ktm.
-.q
Fig. 11. Basilar membrane. (A) A low-magnification electron micrograph of a rabbit basilar membrane. Collagen II is a structural protein found in fiber bundles (arrowheads) located in the basilar membrane, which supports the organ of Corti in the inner ear (Dreiling, et al., 1999). Bar = 5.0 ~tm. (B) Higher-magnification electron micrograph showing fiber bundles (arrows) labeled with a three-step postembedding immunogold procedure using a monoclonal primary Ab to collagen II, biotinylated antimouse immunoglobulin G, followed by tertiary labeling with goat antibiotin immunoglobulin G 10 nm gold. Bar = 1.0 ktm.
118
Madden
(Dreiling et al., 1999; Keithley et al., 2000; Kaneko, et al., 1999; Madden, unpublished research). Also, artifactual shrinkage of the tectorial membrane, a commonly seen artifact (Henson and Henson, 2000a), was not present, and the cellular integrity of inner sulcus cells was also retained (see Fig. 6B). Researchers and clinicians may become frustrated initially with MW decal, because it is not as "instantaneous" as MW fixation. For example, although MW-accelerated fixation of a mouse cochleae is complete after 45-60 min (Madden and Henson, unpublished results), MW decal of the same cochleae, using EDTA, will take 1.5-2.0 h. An understanding of the factors that influence decal is essential for developing a successful MW protocol that is tissue- and application-specific. Decal time, by routine and MW methods, is dependent on several factors: the formulation and temp of the decalcifying solution, the degree of calcification of the tissue, and the size and density of the specimen. For clinical specimens that necessitate fast turnaround for pathology review, EDTA may be substituted by stronger acids, such as nitric and hydrochloric, which will shorten the decal time in the MW as compared to EDTA (Balaton and Loget, 1989). For better morphology and staining properties at the LM level, treatment with slower-acting and lessharsh acids than nitric and hydrochloric acid, such as acetic and formic acid, may be used (Athanasou et al., 1987; Louw et al., 1994; Roncaroli et al., 1991; Ng and Ng, 1992). However, if the tissue is to be used for EM or immunocytochemistry, and if the optimal preservation of ultrastructure and antigenicity is desired, EDTA is a better choice for a decalcifying agent. Aside from shortening decal times, the use of MW irradiation, in conjunction with Ca chelators, such as EDTA, may serve as an antigen retrieval method for aldehyde-fixed tissues (Keithley et al., 2000; Morgan et al., 1994). Higher concentrations of EDTA will chelate more Ca than lower concentrations at the same volumes, but without greatly increasing the decal rate (Table 1). This could result from the decrease in penetration depth of MWs in higher-salt solutions, which can limit the amount of MW irradiation the tissue receives (Kok and Boon, 1992). Higher salt concentrations will also heat more rapidly (Login and Dvorak, 1994), which could lengthen the period between MW cycles, and cause the decal to proceed at a slower rate, despite the greater Ca chelation potential of the higher concentration of EDTA. The volume of decal solution used is also important. Enough solution volume for the longer cycle times required for MW decal, without exhaustion of the decalcifying agent, is necessary, yet the volume should
MW-Accelerated Decalcification
119
be low enough so that MWs can penetrate through the fluid to reach the tissue. Changing the decal solution between each 90-min cycles is one way of ensuring that enough acid remains in the solution to complete the decal process. Frequent replenishment of the EDTA will also allow the investigator to use smaller volumes of the solution, which will result in slower heating and more rapid heat dissipation in the MW, thereby increasing the amount of time that the sample is irradiated. Temp of the decal fluid plays a role in shortening MW decal times. At a temp of 45°C, decal proceeds at a faster rate than at 37°C (Madden and Henson, 1997). Keithley et al. (2000) found that guinea pig tibia bones, decalcified by MW methods using EDTA, took 17 h at 45°C compared to only 3 h at 100°C. However, high temps, above 60°C, may be deleterious to tissue morphology (Kok and Boon, 1992) and cause the denaturation of temp-sensitive antigens. The temp set-point of 50°C, used in the methods here, is optimal, both for the speed of MW decal and the retention of ultrastructure and antigenicity, for a wide variety of tissues. Specimen size, shape, and composition also effect the decal rate. For example, larger samples will take longer to decalcify than smaller ones of the same density. Tissues of the same density and mass, which are thick and cuboidal in shape, will take longer to fully decalcify than samples that are rectangular and thin in one dimension (Madden, unpublished observations). Compact bone from an adult animal will take much longer to decalcify than the same-size sample of cancellous, trabecular bone from a young animal, as well. Decal time in the MW can be shortened by keeping the specimen size as small as possible, thin in one dimension, and by trimming away any decalcified tissue that is not of interest, after each irradiation cycle. Factors that are important to the reproducibility of MW-accelerated decal are calibration of the MW energy distribution within the oven cavity (for Methods 1 and 2), agitation of the decal solution, cooling of the water load or sample vial with a load cooler, and control of the MW magnetron with a temp probe. The period of time required for decalcifying tissues in the MW can range from less than 1 h, for very small samples, to many hours, for large samples, such as monkey temporal bones. For this reason, a temp probe, interfaced with the MW, is necessary to control the cycling of the magnetron, and prevent to overheating of the decal solution and the specimen. The thermistor temp probe, supplied with the Pelco 3450 MW Processor, is designed to cause minimal contributory heating of the sample solution. The probe is also coated in Teflon, allowing for immersion in the EDTA without the risk of corrosion. The response time
120
Madden
of the thermistor probe is approx 0.5-1.0 s, which is sufficiently fast to control the rate of heating, if the sample is placed in the cold spot, as determined by the MW oven calibration procedure described by Giberson and Demaree (1995). The temp, volume, and position of the water load are critical for the reproducibility of methods 1 and 2. Since the MW-absorbing properties of water change with both temp and volume (Login and Dvorak, 1994; Giberson and Demaree, 1995), these parameters should be optimized and held constant throughout the process of MW irradiation. The use of a water recirculation system, such as the load cooler supplied with the Pelco 3450 MW Processor, provides for continual cooling of the waterload beaker or water bath during the decal process, and maintains the temp of the water at a constant level. The use of an air bubbler, as a means of agitation, is also important in controlling the temp of the decal solution and specimen during prolonged irradiation. Air bubbled through the EDTA minimizes the temp gradient within the solution, which in turn contributes to more accurate temp measurement and control of the MW oven by the thermistor probe. The agitation of the solution surrounding the specimen also minimizes uneven thermal-conductive heating of the tissue, and serves to expose the tissue to unexhausted EDTA. During the longer cycles required for MW decal, heat is generated and builds up in the oven floor, from the MWs that are not dampened by the water load or the specimen load. When the sample vial is resting directly on the oven floor, heat conducts from the floor to the vial, raising the temp of the decal solution. The rate of heat dissipation from the sample takes longer, lengthening the time between magnetron cycles. Because of this, the total time for decal is longer. A simple solution to this problem is to elevate and isolate the specimen container, by placing it on top of a MW-transparent platform, such as the inverted plastic Petri dish used in the methods presented here (see Fig. 2B). Polyethylene biopsy sponges, placed between the tissue and the bottom of the sample container, also serve to isolate the tissue from conductive heating effects from the sample vial. More importantly, elevating the tissue off the vial bottom allows for the tissue to be exposed to the decal fluid on all sides, shortening the time required for decal. Placement of the vials directly in the water bath as in method 3, circumvents the problem of heat retention in the oven floor. The water bath fulfills several purposes: It serves as the water load, effectively absorbing extraneous MWs that would otherwise strike the oven floor;
MW-Accelerated Decalcification
121
the turbulence from the water recirculation loop gently agitates the sample vials; and placement of the vials, directly in the continually cooled water, conductively cools the decal solution within the vials, allowing for longer irradiation cycles and a faster decal process. The greatest benefit in MW-accelerated decal is the time savings that the method provides. Months of waiting for a human temporal bone to decalcify is no longer necessary, if the MW methods presented here are employed as part of the histological processing procedure. Research experiments involving calcified tissues can proceed more rapidly, and clinical diagnoses that involve pathology review of bony tissues, such as bone marrow core biopsies, may be turned around in a day. The added benefit of decal in the MW is that the tissue retains ultrastructural integrity, as well as, or better than, tissues decalcified by routine methods. In fact, when comparing routinely decalcified tissues to ones that have undergone MW decal, the major dense and intraperiod lines of the myelin lamellae are often more distinct in the MW-decalcified tissues (Madden and Henson, 1997; and Fig. 9C). Tissues that are decalcified using MW methods, and EDTA as the decalcifier, are usable for a wide range of applications, both clinical and experimental, such as routine LM and EM, immunocytochemistry, cytochemical staining, and in situ hybridization techniques. When used in conjunction with other MW techniques mentioned in this book for fixation, embedding, and staining, the processing of calcified tissues for histology and EM need no longer be seen as a cumbersome and timeconsuming task. MW-accelerated decal is a useful method that can be incorporated routinely in clinical and research labs as part of the histological technique, with excellent results. ACKNOWLEDGMENTS The author would like to thank Dr. C. Robert Bagnell, Jr., director of the Microscopy Services Laboratory, University of North Carolina, Department of Pathology and Laboratory Medicine, and Dr. Miriam M. Henson, UNC Department of Cell Biology and Anatomy, for giving encouragement, support, and time to pursue the work presented here. The author would also like to thank Drs. Miriam Henson and O. W. Henson, Jr., and Mr. Rick Dreiling, UNC at Chapel Hill, and Drs. David Smith and Michiya Sato, Duke University Medical Center, for providing the specimens shown in the figures. Many thanks to Richard Giberson, Ted Pella, for providing technical advice, and for his patient editorial assistance.
122
Madden
REFERENCES Athanasou NA, Quinn J, Heryet A, Woods CG, McGee JO (1987) Effect of decalcification agents on immunoreactivity of cellular antigens. J Clin Patho140: 874-878. Baird IL, Winbom WB, Bockman DE (1967) Technique of decalcification suited to electron microscopy of tissues closely associated with bone. Anal Rec 159:281-289. Balaton AJ, Loget P (1989) Decalcification acceleree par les micro-ondes.Annal Pathol 9: 140-141. Dreiling FJ, Henson MM, Henson OW Jr (1999) Immunoelectron microscopic demonstration of collagen II in the basilar membrane and spiral ligament of New Zealand white rabbits (Abstract). Twenty Second Midwinter Res. Meeting, Assoc. Res. Otolaryngol. 22:77. Faria MR, Dechichi P, de Mello Rode S (1992) Decalcification of mineralized rat mandible tissues with the aid of microwaves. Rev Paul Med 110:283-284. Flower GC (1951) Notes on electrophoretic decalcification of bone. J Med Technol 9:106-107. Giberson RT, Demaree RS Jr. (1995) Microwave fixation: understanding the variables to achieve rapid and reproducible results. Microsc Res Tech 32:246-254. Giberson RT, Demaree RS Jr., Nordhausen RW (1997) Four-hour processing of clinical/ diagnostic specimens for electron microscopy using microwave technique. J Vet Diagn Invest 9:61-67. Hellstr6m S, Nilsson M (1992) Microwave oven in temporal bone research. Acta OtoLaryngo1493(Suppl): 15-18. Henson MM, Henson OW Jr. (2000a) Spiral limbus: specializations in Mormoopidae (Abstract) Twenty Third Midwinter Res. Meeting, Assoc. Res. Otolaryngol. 23:201. Henson OW Jr., Henson MM (2000b) Tympanic membrane: Highly developed smooth muscle arrays in the annulus fibrosus of mustached bats. JAssoc Res Otolaryngol 1:25-32. Kaneko M, Tomita T, Nakase T, Takeuchi E, Iwasaki M, Sugamoto K, Yonenobu K, Ochi T (1999) Rapid decalcification using microwaves for in situ hybridization in skeletal tissues. Biotech Histochem 74:49-54. Keithley EM, Truoung T, Chandronait B, Billings PB (2000) Immunohistochemistry and microwave decalcification of human temporal bones. Hearing Res 148:192-196. Kok LP, Boon ME (1992) Microwave Cookbook for Microscopists. Coulomb, Leiden, The Netherlands. Login GR, Dvorak A (1994) The Microwave Toolbook: A Practical Guide for Microscopists. Beth Israel Hospital, Boston, MA. Louw I, De Beer DP, Du Plessis MJ (1994) Microwave histoprocessing of bone marrow trephine biopsies. Histochem J 26:487--494. Madden VJ, Henson MM (1997) Rapid decalcification of temporal bones with preservation of ultrastructure. Hearing Res 111:76-84. Morgan JM, Navabi H, Schmid KW, Jasani B (1994) Possible role of tissue-bound calcium ions in citrate-mediated high-temperature antigen retrieval. J Pathol 174:301-307. Ng KH, Ng LL (1992) Microwave-stimulated decalcification of compact bones. Eur J Morpho130:150-155. Roncaroli F, Mussa B, Bussolati G (1991) Microwave oven for improved tissue fixation and decalcification. Pathologica 83:307-310.
10
Microwave Processing ' of Sediment Samples Dawn Lavoie, Janet Watkins, and Yoko Furukawa
INTRODUCTION Microfabric is the three-dimensional spatial arrangement of sediment grains and pore space in marine sediments. In addition to the mineral particles, consideration of the fluid-filled voids is essential for understanding sedimentological properties, such as permeability. Traditional tools and techniques limit the study of sediment microfabric. With the advent of high-resolution electron microscopy, microfabric relationships can be directly imaged and quantified. Unfortunately, transmission electron microscope (TEM) sample preparation techniques are generally time-consuming processes requiring the replacement of pore fluid (seawater) with hydrophobic resin, through multiple serial dilutions. During the preparation and handling process, some degree of microfabric disturbance is unavoidable. To alleviate this problem, the authors have adapted a clinical, biological microwave (MW) technique for the preparation of sediment samples. The MW technique is rapid, and introduces little disturbance during processing. This chapter discusses this new MW technique for the preparation of sedimentological samples for TEM observation and analysis. To evaluate this new technique, TEM samples were prepared from sediment collected from North Key Harbor, FL, and from the Chesapeake Bay, off the mouth of the Patuxent River, MD, by both traditional and MW methods, for comparison. From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ 123
124
Lavoie, Watkins, and Furukawa
The adaptation of MW preparation allows serial dehydration and polymerization to be completed in a few hours, with minimal microfabric disturbance; traditional methods of sample preparation, which involve replacement of fluids under ambient laboratory conditions, take several days to several weeks, depending on the permeability of the sediments being prepared. During dehydration, infiltration, and polymerization of the samples, utilizing a specialized MW (Giberson et al., 1997), temperature in the MW is computer-controlled to maintain an even temperature throughout, for uniform processing during these steps. LR White resin (LRW) is used, instead of the more traditional Spurr's resin (Baerwald et al., 1991), because it is relatively hydrophilic, has low toxicity and low viscosity, is a convenient, ready-to-use single component mix, and does not result in any measurable sediment distortion after polymerization. Sediment samples, prepared using the MW and LRW during processing, do not require the tedious exchange of solvents nor are as many serial dilutions required for traditional processing of samples. The resulting resin-embedded sediments are satisfactory for TEM imaging.
Samples NORTH KEY HARBOR
Sediment samples recovered from North Key Harbor, FL (Fig. 1) originate primarily from the breakdown and subsequent transport of biogenic carbonate material from nearby reefs and keys, which include plates of aragonitic green algae (Halimeda, Penicillus, and Udeota), molluscan shells, benthic and planktonic foraminifera, echinoid spines, sponge and coral fragments, and diatoms (Bentley and Nittrouer, 1997; Stephens et al., 1997). The microfabric in the study area was determined by the depositional regime, as well as the postdepositional alteration caused by early diagenesis (Furukawa et al., 1997), and is comprised of grains of various shapes and sizes (Fig. 2). Porosity and permeability, which are key physical parameters influencing sediment behavior, are functions of this arrangement of grains and pore space; therefore, the maintenance of the pore network during handling and preparation procedures is especially important. CHESAPEAKEBAY
Sediment samples recovered from the mouth of the Patuxent River in the Chesapeake Bay, MD (Fig. 3), are silts and clays, predominantly illite, smectite, and chlorite, in decreasing order of abundance, and trace amounts of kaolinite and quartz (Bennett et al., 1995). Smectite has a high affinity for water and a large surface area:mass ratio, which contribute to high water content and porosity. In addition, the microfabric of the surficial sediment
125
MW Processing of Sediments
Y
OF M E X ~ O Dept~ in Famoms
83*
GULF
45'
OF
MEXICO
-
* #
~,
DryT ~ p s 24* 30'N 83 °
40'
Key
o~,~ Fa~s 20'
82 ° W
Fig. 1. Location of North Key Harbor sediments. Multishaped, multisized, coarse-grained calcium carbonate sediments were derived from nearby reefs and keys in the Dry Tortugas, 60 nautical miles west of Key West.
has a very open structure, which is maintained by limited particle contact of edge-to-face and edge-to-edge associations of clay mineral plates, rather than the stronger face-to-face structure that develops with burial pressures (Fig. 4). The fragility of the clay microfabric in this type of open structure is easily disturbed during sample preparation processes, and thus is a good fabric for testing the new MW preparation technique. METHODS Diver-collected sediment cores were used in both the North Key Harbor and Patuxent River study sites. The sediment was subsampled away from disturbed zones near plastic core tube walls. Subsamples were prepared in one of two ways: by traditional methods of preparation, using Spurr's resin and multiple serial dilutions; or using a new MW processing technique with LRW.
126
Lavoie, Watkins, and Furukawa
Fig. 2. Scanning electron micrograph illustrating the multisized nature of grains, primarily Halimeda plates, in this view. Although the plates appear solid in SEM observation, they are actually comprised of hundreds of aragonite submicron-sized needles that become abraded from the edges of the plates, and form the micrite-sized matrix material (see Fig. 7).
Traditional Method of Preparation Subsamples were carefully critical-point dried, following replacement of seawater (sediment institial fluid) with a series of transition fluids: ethanol (ETH), amyl acetate, and liquid CO2 (Baerwald et al., 1991; Bennett 1976). Dried samples were embedded with Spurr's resin under vacuum in a BEEM TM capsule, polymerized at 60°C for 24 h, and ultrathin-sectioned by microtome (Baerwald et al., 1991) for TEM analysis. During sample preparation, samples were found to be sensitive to handling, and critical-point-dried samples would collapse under slight jarring. Thus, extreme care was used during preparation. Computer image analysis of TEM micrographs, recorded from these samples, showed no change in pore volume between samples, indicating minorto-no disturbance of microfabric (Bennett, 1976; see below).
127
MW Processing of Sediments
i,
t t
0,,
V
~
~
%
• ,,- 1=
' '
Chesapeake
t ~
".;
, 0
~
~
°
t,
~
S a m p l e site at ~ e m o u t r t of the P a t u x e n i l ; : l i v e r •. 3 8
41
' ", , ,
If
- -f
P
20'
/~ /
_
J,
P~
P/¢
/,"
;, Depth in feet
ql,; ,.
,
I
¢,
", "~'~
119
5'
' t',,~.
¢
t"' , d
~
" .* .,.
Bay • • ,.
-1
(
- 10' "
-
I
38 ° 05'N
76°10'W
Fig. 3. Silty clays, used to test the new MW processing technique, were collected from the mouth of the Patuxent River in the Chesapeake Bay.
Microwave Preparation 1. Description of equipment: The MW has been adapted for the lab, with a vented MW chamber to effectively remove fumes during processing and when the MW chamber door is opened (Fig. 5). The automatic magnetron prewarming allows the oven to preheat and start the timer when it is fully prewarmed. The MW allows numbered keys to be independently programmed for varying time and power levels. The temp probe, located inside the MW cavity, monitors and maintains the desired sample temp. The load cooler, located above the oven, circulates the water load through beaker A, and maintains its temp (Fig. 6). This temp maintenance is critical for reproducibility, because, as water is heated, it absorbs less MW energy, thus creating another
128
Lavoie, Watkins, and Furukawa
Fig. 4. Clay fabric typical of that recovered from the mouth of the Patuxent River. Smectite looks almost amorphous; illite is identified as the dark black particles. Note the open fabric and the edge-to-face (e-f) and edge-to-edge (e-e) orientation of clay particles. This sample was prepared using the new MW technique. Image analysis quantification of porosity was 49%
variable to be considered. The RS-232 port on the oven allows the temp to be examined on a real-time basis, and/or stored to a file on the computer. 2. Processing using the MW system: Samples, while in BEEM capsules and held in a Teflon ® sample holder, were processed according to the protocols set out in Table 1, which illustrates the authors' adaptations
MW Processing of Sediments
t,---=
:
..........U
"
129
. . . .
"
•
U
~1
/
l
Fig. 5. Schematic illustrating the arrangement of(A) load cooler, (B) MW processor, and (C) temp controller. (Reproduced with permission from Ted Pella, 1997).
Fig. 6. Interior of MW showing location of beakers A and B. of the basic Ted Pella (1997) MW processing procedure. Both acetone and ethanol were used as exchange fluids, in order to test the hypothesis that organic matter might survive fluid exchanges of ethanol more easily than fluid exchanges of acetone.
130
Lavoie, Watkins, and Furukawa Table 1
Protocol for MW Processing of Sediment Samples LRWhite M W / Solvent a
Dehydration
Infiltration
Polymerization
50% Solvent 37°C 1 x 40 s 60% Solvent 37°C 1 x 40 s 70% Solvent 37°C 1 x 40 s 80% Solvent 37°C 1 x 40 s 90% Solvent 37°C 1 x 40 s 100% Solvent 37°C 2 x 40 s 1:1 Solvent:resin 45°C 2 x 15 min 100% resin 45°C 3 x 15 min 100% resin 60°C 10 min 100% resin 70°C 10 min 100% resin 80°C 25 min
aSolvent may be either acetone or ethanol. MW preparation: This procedure involves the use of a lab MW to accelerate chemical processing of sediment for TEM. It is important to try this technique, and adjust accordingly, for the type of sediment being prepared. For example, marine clay sediments require more graded solvent exchanges than carbonate sediments for dehydration and infiltration, because of their low coefficient of permeability.
a. Dehydration: The temp restriction was set to 37°C. The temp probe was immersed in the solvent (ethanol or acetone) in the B E E M capsule. The sample was microwaved, after each solvent change, for 40 s at 100% power. The water was renewed in beaker B, when it became warm to the touch. If the sample solution was cloudy, indicating water was present in the sample, another solvent exchange was completed.
MW Processing of Sediments
131
b. Infiltration: The temp restriction was set to 45°C. The temperature probe was immersed in the solvent in the BEEM capsule. The sample was microwaved at each solvent-resin (LRW) exchange for 15 rain at 100% power. Again, the water in beaker B was changed when warm to the touch. At this point, each BEEM capsule was sealed with a Parafilm®-lined cap, placed in the Teflon sample rack, then set in a polypropylene polymerization container. The Teflon rack with the samples was then totally immersed in 1000 mL water. c. Polymerization: The temp restriction was set to 60, 70, and 80°C for 10 rain, 10 rain, and 25 rain, respectively. The temp probe must remain immersed in the water bath. The samples were microwaved at 100% power for the indicated restriction times. The steps outlined above require approx 2 h, after which the samples can be removed from the BEEM capsules and ultramicrotomed for TEM examination. RESULTS Sediments from North Key Harbor depict examples of Halimeda plates prepared for TEM examination in the traditional manner (Fig. 7A), and using the new MW technique (Fig. 7B). Halimeda plates, composed of numerous aragonite needles, although fragile after death, retain their approximate orientation, after preparation using either technique. A close examination of the edges of the Halimeda plate shows no substantial loss of needles in either image, suggesting that both techniques retain satisfactory orientation of particles. The fracturing of brittle grains, as illustrated by the remains of a biogenic test in Fig. 7A, is apparent in samples prepared by both techniques, and is an artifact of microtome sectioning. No differences in preservation or orientation of matrix carbonate particles were evident between samples prepared by the two techniques. In contrast to the North Key Harbor sediments, the silty clays from the Patuxent River site are comprised of small, submicron-sized smectite and illite grains, arranged in loose domains of edge-to-face and edge-to-edge orientation, which are maintained during the processing procedures. Samples prepared from the traditional (Fig. 8A) and MW (Fig. 8B) methods appear to contain the same high volume of pore space. Submicron-sized clay particles form an onion skin (Fig. 8B) arrangement around voids that are probably remnants of precursor cellular material. Quantitative measurement of pore space in the most dense clay matrix found in this sediment (Fig. 9, traditional preparation; and Fig. 4, MW preparation) using image analysis indicates image porosity to be 49 and 54%, respectively.
132
Lavoie, Watkins, and Furukawa
Fig. 7. TEM micrograph of carbonate sediment recovered from North Key Harbor. The Halimeda plate, illustrated in both (A), prepared using traditional processing, and (B), prepared using MW-assisted processing, has retained its shape and orientation of the numerous aragonite needles. DISCUSSION The authors' primary criterion for successful sample preparation is that sediment microfabric, the orientation of grains and pore space, be maintained, in a rapid and reproducible manner. Traditional preparation techniques, utilizing the hydrophobic Spurr's resin, required that all water be removed from the sediment sample, through a series of fluid exchanges and critical point drying. The permeability of the sediment determined the rate at which these exchanges could be completed. In a carbonate sediment, such as the North Key Harbor samples, in which Darcy's coefficient of perme-
133
MW Processing of Sediments
Fig. 7B.
ability is about 10-5 cm/s, the traditional preparation process can be completed in 3-4 d. In clay sediments, such as those recovered from Chesapeake Bay, in which Darcy's coefficient is about 10-7-10-8 cm/s, the traditional preparation process requires several weeks to complete. In addition to time concerns, the probability of fabric disturbance increases with time required for preparation and amount of handling (e.g., number of fluid exchanges). The new MW technique allows preparation of samples in 2-4 h using either Spurr' s or LRW. Although some edge distortion with LRW, using the traditional preparation technique, was observed, no edge disturbance is apparent in samples embedded in LRW, when MW-prepared. The ability to substitute LRW for Spurr' s resin is advantageous, since Spurr' s resin is carcinogenic. At this time, although the authors have not quantitatively measured the amount of sedimentary organic matter preserved for TEM obser-
134
Lavoie, Watkins, and Furukawa
Fig. 8. Clay microfabric from Patuxent River clays prepared (A) by traditional methods, including Spurr's resin impregnation, and (B) using MW-assisted processing and LRW impregnation. The onion skin arrangement of clay particles marks traces of organic matter remnants. Note the open structure and biogenic remains preserved by both preparation techniques. vation, more organic matter seems to be preserved using the faster MW technique, because organic matter has less time to be degraded. The authors also suspect that ethanol is more likely to result in organic matter preservation than with acetone, although more work is needed to quantitatively determine the difference. In the absence of any subjective or quantitative differences in microfabric preservation, the time-saving potential of the MW technique makes the authors' sample preparation effective for use in sediment samples.
135
MW Processing of Sediments
Fig. 8B.
ACKNOWLEDGMENTS This work was accomplished with the support of the Naval Research Laboratory core research program, Program Element Number 0601153N. Nancy Carnaggio and Kinsley McCrocklin of NRL helped in traditional sample preparation. The authors acknowledge the help and support of Richard Giberson of Ted Pella. REFERENCES Baerwald RJ, Burkett PJ, Bennett RH (1991) Techniques for the preparation of submarine sediments for electron microscopy.In Bennett RH, BryantWR, Hulbert MH,
136
Lavoie, Watkins, and Furukawa
Fig. 9. Microfabric of the most dense clay matrix from Patuxent River sediment, prepared in the traditional manner. Image analysis quantification of porosity is 54%, similar to that measured in clay fabric prepared using the new MW processing technique (see Fig. 4). eds. Microstructure of Fine-Grained Sediments, From Mud to Shale. SpringerVerlag, New York, pp. 309-320. Bennett RH (1976) Clay fabric and geotechnical properties of selected submarine sediment cores from the Missippi river delta. Ph.D. thesis, Texas A&M University, College Station, TX. Bennett RH, Lavoie DL, Sawyer W, Hunter N, Meyer M, Kennedy C, et al. (1995) Biogeotechnical and biogeochemical properties of sediments from Chesapeake Bay: mass physical and mechanical properties of sediments near the mouth of the Patuxent River, NRL/PU/7430-95-0008. SSC, MS 39529.
M W Processing of Sediments
137
Bentley SJ, Nittrouer CA (1997) Environmental influences on the formation of sedimentary fabric in a fine-grained carbonate-shelf environment: Dry Tortugas, Florida Keys. Geo-Marine Lett 17:268-275. Furukawa Y, Lavoie D, Stephens K (1997) Effect of biogeochemical diagenesis on sediment fabric in shallow marine carbonate sediments near the Dry Tortugas, Florida. Geo-Marine Lett 17:283-290. Giberson RT, Demaree RS Jr., Nordhausen RW (1997) Four-hour Processing of clinicaldiagnostic specimens for electron microscopy using microwave technique. J Vet Diagn Invest 9:61-67. Stephens KP, Fleischer P, Lavoie DL, Brunner C (1997) Scale-dependent physical and geoacoustic property variability of shallow-water carbonate sediments from the Dry Tortugas, Florida. Geo-Marine Lett 17:299-305. Ted Pella (1997) Microwave-assisted rapid processing protocol for electron microscopy, Redding, CA 96003-1448. Available from Ted Pella, Redding, CA.
11
Microwave Polymerization in Thin
Layers of London Resin White Allows Selection of Specimens for Immunogold Labeling Jennifer E. Lonsdale, Kent L. McDonald, and Russell L. Jones
INTRODUCTION Preparing a specimen for immunogold labeling is a time-consuming process. The specimen must first be fixed, dehydrated, and infiltrated with resin, without destroying protein antigenicity. The specimen is then sectioned, and the sections labeled with antibodies (Abs) conjugated to 5-15 nm gold particles. The most time-consuming step in immunogold labeling does not begin until the sections are viewed inside the transmission electron microscope (TEM), specifically, locating regions that have been labeled with gold particles. Labeled sections must be examined at high magnifications (in excess of x16,000), in order to visualize the gold particles. At these magnifications, the field of view is reduced to only a few microns, making a scan of the surface area of an entire section impractical. A more efficient way to locate immunogold label is to focus on regions of the specimen predicted to contain the protein that is recognized by the Ab. Finding these regions can be challenging, however. If an Ab recognizes a protein that is celltype- or developmental-stage-specific, the presence of that cell type or stage must be confirmed by thick sectioning, before ultrathin sections are taken for immunogold labeling. In some cases, locating the appropriate cells for immunogold labeling means sectioning through several blocks of specimen before finding the right cells. From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ
139
140
Lonsdale, McDonald, and Jones
Increasing the efficiency of locating target cells requires protocols that allow specimens to be examined with a light microscope (LM) before sectioning. During thin-layer embedding (TLE), specimens are embedded in very thin (0.16-0.19 mm), optically clear layers of resin on a microscope slide (Howard and O'Donnell, 1987; Lonsdale et al., 1999; McDonald, 1994; Reymond and Pickett-Heaps, 1983; Taylor, 1984). After polymerization, embedded specimens can easily be examined by LM to assess the degree of preservation and specimen orientation before sectioning. Well-preserved specimens of the target-cell type can then be specifically selected and remounted for sectioning. Thus far, most methods for TLE have been designed for epoxy resins, which can denature proteins, making most samples infiltrated with epoxy resins unsuitable for immunogold labeling. The resins of choice for immunogold labeling are either the Lowicryl series or London Resin White resins (LRWs) (Hobot and Newman, 1993). Lowicryls are challenging to work with, because they require low temperatures (-35 to-70°C) for infiltration and embedding, and must be kept away from moisture (Griffiths, 1993). Despite these obstacles, methods for TLE in Lowicryl have been developed for some cell types (Kiss and McDonald, 1993; McDonald, 1994). This chapter, describes two procedures for embedding samples in thin layers of LRW. Infiltration in LRW does not require low temperature (temp), and has the additional advantage that the volatile fumes are less hazardous than those of Lowicryls. MATERIALS A N D M E T H O D S
Protocol 1. Specimen preparation: a. Prepare specimen for embedding in LRW. 2. Slide mold preparation: Time: -5 min/slide. a. Treat slide with lubricant. b. Glue Thermanox ® spacer to slide. c. Allow slide to dry 15 min. d. Treat Thermanox cover slip with lubricant. Note: Slide molds can be made in advance. 3. Embedding: Time:--5 min/slide. a. Place specimen in slide mold. b. Fill mold with resin (--300 ktL). c. Orient specimen. d. Carefully place cover slip over mold.
MW Polymerization in Thin-Layer LRW
141
4. Rapid polymerization: Time:--1 h. a. Place slide molds on floor of microwave (MW) vacuum (vac) unit. b. Place vac unit in center of MW. c. Insert temp probe into vac chamber. Tip of probe should be --5 mm above slides. d. Pull 25 mmHg vac on the vac chamber. e. Set MW to 60°C and the variable wattage to level 1. f. MW for 60 min. Raise the wattage one level every 5 min. The last 35 min will occur at wattage level 6. 5. Remounting: Time:-- lh. a. Remove cover slip from the surface of the polymerized resin. b. Use a LM to select sample. c. Use a scalpel to excise selected samples. d. Glue the excised sample to a dummy block with two-part epoxy. e. Polymerize the two-part epoxy at 60°C for 1 h or at room temperature overnight. Total time:-- 2.5 h.
Specimen Preparation All specimens in this chapter were prepared for TLE, with high-pressure freezing, followed by freeze substitution: barley (Hordeum vulgare cv. Himalaya) aleurone protoplasts, MW-infiltrated with LRW as described by Lonsdale et al. (1999); nematodes (Caenorhabditiselegans; provided by R. Aroian, University of California, San Diego, CA); fly embryos (Drosophilamelanogaster;provided by D. Sharp, University of California, Davis, CA); and sea urchin micromeres (Strongylocentrotous purpuratus; provided by F. Wilt, University of California, Berkeley, CA), were prepared as described by McDonald (1999). Methods of tissue fixation, dehydration, and infiltration in LRW, which do not require freezing or MW preparation, can also be used (see Hobot and Newman [1993] for sample protocols).
Slide Mold Preparation MATERIALS 1. Glass microscope slides. 2. MS-122 Release Agent Dry Lubricant (Miller Stephenson, Danbury, CT). 3. Clean cloth or lab tissue. 4. 22 x 60 mm Thermanox cover slips (cat. no. 72271, Electron Microscopy Sciences, Fort Washington, PA). 5. Scalpel. 6. Aron Alpha 201 Industrial Krazy Glue (cat. no. 14435, Ted Pella, Redding, CA).
142
Lonsdale, McDonald, and Jones
PROCESS 1. Spray microscope slides with dry release agent. This lubricant allows the resin to be easily removed from the slide after polymerization. Allow the lubricant to dry to an opaque film. Wipe excess lubricant from the slide with a clean cloth (the slide should be clear and feel smooth). 2. Use a scalpel to remove the center of Thermanox cover slip, leaving a 3-4 mm border. Attach the cover slip border to the lubricated microscope slide with Krazy Glue. The cover slip border forms a shallow well on the microscope slide (Fig. 1A). 3. Allow the glue to dry for at least 15 min, before loading the specimen. 4. Spray the cell-growth side of an intact Thermanox cover slip with dry release agent. Allow the lubricant to dry to an opaque film, and remove the excess lubricant as described in slide mold preparation, step 1.
Embedding MATERIALS 1. Slide mold (see above). 2. Treated Thermanox cover slip 3. LRW (Ted Pella).
(see above).
PROCESS 1. Place fully infiltrated sample in a drop of fresh LRW in the center of the slide mold. 2. Fill the well with fresh LRW. The well will hold approx 300 ~tL. 3. Disperse, and orient samples, as necessary with fine needles. 4. Use forceps to carefully place the cover slip, lubricant-side down, over the resin, without introducing air bubbles. Troubleshooting: If air bubbles are present in the well, the resin surrounding these bubbles will not polymerize, resulting in sample loss. To reduce the number of air bubbles that form, degas the resin by placing it in a beaker inside a chamber attached to a laboratory vac line or a vac pump. Pull a constant vac (approx 20 mmHg) on the chamber, for at least 15 rain. Degassing is complete when bubbles are no longer present on the surface of the resin. After filling the slide mold with degassed resin, apply the cover slip by carefully rolling it over the surface of the well. If air bubbles do form, remove them by gently lifting the corners of the cover slip, and teasing out the bubbles with a needle or excess LRW in a hypodermic syringe.
A Thermanox cover
slip
~,/ ....jLRW containingspecimen / ~ ~ ~ :///~i|--- Microscopeslide / ' / / • " eb-~ i ~ "--C/~i .........~ T h e r m a n o x spac er
B
'~
\ /
~\ \ A"
~nc,,
Tungsten-carbide
/sSp~cle(:tn~edn
~.. ~ . . ~ C
~-.~. -~/~ D
Excised LRW containing //2 s p e c ~ ~Forceps o partepoxy
E
Block mounted in microtome
ExcisedLRW
with selected
/specimen
~ e,ox,m l ~
,
K!lfe
Fig.
1. TLE in LRW. (A) Specimens infiltrated with LRW are placed in the center of a slide mold. The slide mold consist of a Thermanox spacer glued onto a microscope slide, and is covered with an intact Thermanox cover slip. (B) After polymerization, the cover slip is removed, and specimens are selected with LM. Selected specimens are marked with a tungsten-carbide pencil. (C) Marked specimens are excised with a scalpel. (D) Excised specimens are remounted on a dummy block coated with two-part epoxy. (E) Once the epoxy hardens, the block is shaped and sectioned on an ultramicrotome.
144
Lonsdale, McDonald, and Jones
Rapid Polymerization MATERIALS 1. 2. 3. 4.
Prepared slide molds containing sample and resin. Variable-wattage MW oven (Model 3450, Ted Pella). MW vac chamber (Model 3435, Ted Pella). Vac pump.
PROCESS 1. Place slides containing thin layers of LRW and sample on the floor of a MW vac chamber. Place the vac chamber in center of a variablewattage MW containing only the recirculating water bath. 2. Place the temp probe through temp probe hole, and adjust the height, so that the probe is approx 5 mm over the slides. 3. Pull 25-mmHg vac on the chamber. 4. Set the temp restriction to 60°C and the variable-wattage setting to level 1. 5. Program the MW to run for 60 min. Every 5 min, increase the variable wattage by one level. Level-6 wattage will be reached 25 min after starting the MW. Continue microwaving at level 6 wattage for the remaining 35 min. 6. After 60 min, the resin should be polymerized. Release the vac and check the slides. If the resin is not completely polymerized, return the slides to the chamber, and MW under vac for another 15-20 min. Troubleshooting: Watch the vac strength throughout the polymerization process. Should the vacuum chamber develop a leak during the polymerization procedure, and the vac strength falls below 20 mmHg, pull the vac again. If the vac pump is hooked directly to the MW, and the stopcock on the vac chamber remains open during processing, the vac can be adjusted without stopping the MW.
Slow Polymerization MATERIALS 1. Airtight chamber that will fit into an embedding oven (food storage container, such as Rubbermaid ® or Tupperware®). 2. Two stopcocks (cat. no. 28, 644-3, Aldrich, Milwaukee, WI). 3. Gaseous N 2 tank with tubing to fit stopcock. 4. Embedding oven (60°C). PROCESS 1. Place the slides in a small, airtight chamber, with a stopcock attached to each side. An example of such a container is a food storage container with a hole drilled in each side. Stopcocks can then be glued into the holes with two-part epoxy. When the lid is attached and the stopcocks are closed, the container is airtight.
MW Polymerization in Thin-Layer LRW
145
2. Place the lid on the container, and open the two spigots. 3. Attach one spigot to a source of gaseous N 2. Allow the N 2 to slowly fill the chamber. Turn off the gas source, and close the stopcocks. 4. Place the N2-filled chamber in a 60°C oven for 2 d.
Remounting MATERIALS 1. Dissecting or compound microscope (depending on specimen size and magnification needed). 2. Tungsten carbide pencil (Fisher). 3. Emery board. 4. "Dummy" blocks (made of an epoxy resin in a flat mold). 5. Two-part epoxy (Epoxy 907, Miller Stephenson). 6. Scalpel or razor blade. 7. Forceps.
PROCESS 1. Remove the Thermanox cover slip from the thin layer of resin. 2. Examine the slides with a microscope. 3. Circle selected specimens with a tungsten pencil (Fig. 1B). Another option for marking selected specimens is an inscribing objective (Leica, Deerfield, IL) that will encircle the specimen with a diamond scribe directly below the objective. 4. Use a scalpel to cut the circled specimen from the thin layer of resin (Fig. 1C). 5. Use an emery board to create a rough surface on a dummy block. Coat the surface with a thin layer of two-part epoxy. 6. Use forceps to pick up the specimen excised from the thin layer of resin, and place the specimen on an epoxy-coated block (Fig. 1D). Use the forceps to gently push the specimen into the epoxy, and, if necessary, to orient it on the block. Note: During embedding, samples tend to sink to the bottom of the well. When remounting the excised sample, check the location of the specimen. Be sure to place the excised sample "specimen-side up" on the block. 7. Place the remounted blocks in a 60°C oven for 1 h. Troubleshooting: Excised resin that is not completely polymerized will curl after 60 min in the oven. Test for complete polymerization, by placing a blank piece of resin in the oven for 60 min, to see if it curls. Alternatively, let the remounted epoxy harden overnight at room temperature.
146
Lonsdale, McDonald, and Jones
8. Face and section blocks (Fig 1E). Note: The two-part epoxy used for remounting will not damage a diamond knife during sectioning, and is stable under the electron beam.
Sectioning, Immunocytochemistry, and Electron Microscopy Following remounting, specimens were sectioned to a thickness of 70 nm, with a diamond knife (Diatome) on a Reichert-Jung Ultracut E (Leica). Sections were placed on Formvar and carbon-coated Gilder 100-mesh Ni grids (Ted Pella). Immunogold labeling and staining was performed as described by Lonsdale et al. (1999). Abs against malate synthase (a generous gift of Dr. R. Trelease, Arizona State University, Tempe, AZ) were diluted 1:20; Abs against the aquaporin c~-TIP (a generous gift of Dr. K Johnson, San Diego State University, San Diego, CA) were diluted 1:100. Grids were examined with a JEOL- 100CX electron microscope (Jeol, Peabody, MA) at an accelerating voltage of 80 kV. RESULTS
TLE Protocols TLE in LRW was used to prepare barley aleurone protoplasts for sectioning and immunogold labeling for TEM. Two methods were used to create an anaerobic environment for the polymerization of thin layers of LRW. Although both methods were successful in polymerizing the resin, each has disadvantages to consider. In the first method, molds containing thin layers of resin were placed in a vac chamber in a MW oven. Polymerization was established by gradually increasing the wattage of a MW over 25 min, and maintaining full wattage for 35 min. A 25-mmHg vac was maintained during polymerization. The entire process was completed in 60 min. This method provided a rapid and reliable way to polymerize a thin layer of LRW, but two artifacts were identified. In some instances, large air bubbles formed in the center and around the edges of the mold. Air bubbles were surrounded by zones of unpolymerized resin (approx 3-5 mm wide). Specimens present in the zone of unpolymerized resin could not be sectioned. Large air bubbles formed from dissolved gases that coalesced when the resin was placed under vac. Bubble formation could be somewhat reduced by thoroughly degassing the resin before embedding. The introduction of a few air bubbles into the mold during the embedding procedure was difficult to avoid.
MW Polymerization in Thin-Layer LRW
147
In cases in which air bubbles were not present, the polymerized layer of resin was sometimes uneven. Generally, the layer was thin at the center of the mold and thicker toward the edges. Samples present in the thin regions of the resin layer could not be remounted and sectioned. This artifact may also be the result of dissolved gases coalescing in the resin. Although liquids do not compress in a complete vac, gases can be greatly reduced in volume. If air pockets cannot escape from polymerizing resin, the air may be compressed. In this case, the effects of air compression may be compounded by the fact that Thermanox cover slips are flexible. A combination of the vac-driven compression of dissolved gases in the resin and the flexibility of the cover slip may allow the center of the cover slip to dip toward the bottom of the slide mold, creating an uneven layer of resin. This artifact might be prevented by adjusting the strength of the vac applied, or by using a smaller surface area for polymerization (but these techniques have not been tested). The second method used to create an anaerobic environment for the polymerization of LRW in a thin layer involved the use of a Nz-filled chamber in a 60°C oven. Polymerization took place over 2 d. Polymerization took longer using the N 2 chamber, but the above mentioned problems with air bubbles and uneven layers of polymerized resin were observed less frequently. Uneven layers ofresin were occasionally found with specimens polymerized in the N 2 gas chamber, suggesting that uneven polymerization may result partly from resin shrinkage. One problem unique to the N 2 chamber method of polymerization concerned the removal of the cover slip from the resin, following polymerization. Long-term polymerization of LRW in a N 2 chamber altered the properties of the Thermanox cover slip, causing it to bond tightly to the resin. When this occurred, the cover slip could not be separated from the resin layer. To overcome this problem, Thermanox cover slips were treated with the same lubricant that prevented the resin from bonding to the glass microscope slide. Lubricated cover slips did not bond to LRW after 2 d in a N 2 chamber at 60°C.
Sample Selection The thin layer of LRW resin produced with both protocols was optically clear, and could easily be examined by LM. The developmental stage and quality of preservation of barley aleurone protoplasts, embedded in thin layers of LRW, were easy to assess. Figure 2A shows a cluster of barley aleurone protoplasts, embedded in LRW, and viewed with compound microscope equipped with differential interference
148
Lonsdale, McDonald, and Jones
Fig. 2. Barley aleurone protoplasts embedded in a thin layer of LRW. (A) Cluster of protoplasts prepared by TLE, as viewed through a LM equipped with differential interference contrast optics. Protoplasts 1 and 2 are intact cells of different developmental stages. Protoplasts 3 and 4 are not well preserved. (B) Highermagnification view of protoplast 1. C, cytoplasm; V, vacuole. Bar = 10 gm.
contrast optics. Several protoplasts, at different developmental stages and levels of preservation, are visible. Protoplasts 1 and 2 in Fig. 2A are of different developmental stages, as determined by the number and size of the vacuoles within the cell (Bush et al., 1986). These protoplasts were spherical, indicating that the plasma membrane was not damaged during processing. Protoplasts 1 and 2 contain the same degree of subcellular detail that is visible in living protoplasts observed with differential interference contrast optics (Lonsdale et al., 1999), suggesting that the cell contents are well preserved. At higher magnifications, structural details, such as vacuolar and cytoplasmic retention, could be resolved (Fig. 2B). Protoplasts 1 and 2 were selected and remounted for sectioning and immunogold labeling. Further analysis of the cluster of protoplasts embedded in a thin layer of LRW, in Fig. 2, showed that protoplasts 3 and 4 (Fig 2A) were damaged during processing. The plasma membrane of protoplast 3 ruptured, releasing the internal organelles into the media. The lack of subcellular detail in protoplast 4 suggested that, although the cell was spherical and the plasma membrane was intact, the cytoplasm and organelles were extracted. Protoplasts 3 and 4 were not selected for remounting and sectioning. The ability to select specific samples for sectioning and immunogold labeling was not limited to the barley aleurone system. TLE of several other cell types, including C. elegans (Fig. 3A,B), D. melangaster
MW Polymerization in Thin-Layer LRW
149
Fig. 3. TLE in other organisms and cell types. (A) C. elegans, (B) C. elegans embryos, (C) D. melanogaster embryo, and (D) S. purpuratus micromeres, embedded in thin layers of LRW, and viewed with a LM equipped with differential interference contrast optics. embryos (Fig. 3C), and S. purpuratus micromeres (Fig. 3D), was also performed. Regardless of the method of LRW polymerization, individual specimens of C. elegans, embedded in thin layers, remained intact during processing, and could be selected for whole mount sections. The developmental stage of embryos within C. elegans could be also determined by LM, allowing embryos of the desired developmental stage to be selected for sectioning (Fig. 3B). LM was also used to determine the developmental stages of D. melanogaster embryos (Fig. 3C) and S. pupuratus micromeres (Fig. 3D) embedded in thin layers of LRW.
Immunocytochemistry The process of TLE in LRW preserved the antigenicity of intercellular proteins, and these proteins could be localized with immunogold labeling. The authors chose Abs against a soluble organellar enzyme (malate synthase) and an integral membrane protein (a-TIP), to test the effectiveness of immunogold labeling in barley aleurone protoplasts prepared for TEM with TLE (Fig. 4). Malate synthase, a soluble enzyme involved in the glyoxylate cycle, is known to reside in the lumen of a type of plant peroxisome known as
•"..........
....>~:::~<~%~:!.i '~'~il....
~.....
~:.:~:!i~:~:i~, ,~i~:~!i~.~
....~
i~
~~i~
~i,.~.:..:.~!!,~
~..,~~:,. ,:~, .~,~!i~!!i~ ......................... ~:~.~i
~iii~,,,~ ~,~.~................ ~..~,.............. .~ ~,,... "-..~:~:~,
.....~':i~
.i:~::~Lijl ~:~ ..,~i~,~: :::~ .~:~:~i~i{!~!,,.~,!..... .i~
"
:~.~.:k~l~:~,
.-,~i~:~'~~ ~..~. !, y::.~
.~,,,~::i~.
~B~!
i.... ~.~:...~,~LIIII :~!!~.............
.::~::~:~::.~:~<. ~,,,,:~:..~ii ~i::,~. .... :.. i~:: -.,,:~, ......:.-.::::< Uii ~::..,~:~:..... ~g.:...:~:~i!, ~ ......~!ii:~:::.: ~ G : ....F:: ...... ~%.. ~
©
Dm~
©
pml
o
MW Polymerization in Thin-Layer LRW
151
a "glyoxysome" (Huang et al., 1983; Trelease et al., 1971). This organelle is similar to animal peroxisomes, because it is surrounded by a single lipid bilayer, and has an electron dense appearance when viewed with TEM (Trelease et al., 1971). Polyclonal Abs against malate synthase accurately labeled glyoxysomes in barley aleurone protoplasts prepared by TLE (Fig. 4A, B). Polyclonal Abs against a-TIP, an aquaporin that resides on the vacuolar membrane of plant cells (Johnson et al., 1989), were used to test for the retention of antigenicity of membrane-associated proteins in barley aleurone prepared for TEM with TLE. Abs against (x-TIP labeled vacuolar membranes, but not other membranes in barley aleurone protoplasts prepared by TLE (Fig. 4C,D). The degree of Ab labeling did not differ between the two methods of thin-layer polymerization (MW vs 60°C oven), and did not appear to be significantly different from the degree of labeling in samples prepared by traditional LRW embedding procedures (data not shown). DISCUSSION TLE is a useful technique for the preparation of specimens for TEM. Using this method, specific specimens can be selected and oriented for sectioning. The benefits of TLE have been extensively discussed (Howard and O'Donnell, 1987; Reymond and Pickett-Heaps, 1983; Taylor, 1984). This chapter presents two methods for TLE in LRW. The requirement for an anaerobic environment during polymerization was the major obstacle to polymerizing LRW in a thin layer. This obstacle was overcome by using either a vac chamber in a clinical MW oven or a Nz-filled chamber in an embedding oven. Both of these methods effectively and efficiently polymerized LRW in a thin layer, while maintaining protein antigenicity for immunogold labeling. Embedding
Fig. 4. (opposite page) Immunogold labeling of barley aleurone protoplasts prepared for TEM with TLE. (A) Protoplast labeled with Abs against malate synthase, a marker of plant glyoxysomes. The outlined area in (A) is enlarged in (B) to show the gold particles (arrowheads). (C) Protoplast labeled with Abs against or-TIP, a marker of the plant vacuolar membrane. The outlined area in (C) is enlarged in (D) to show the gold particles (arrowheads). G, glyoxysome; M, mitochondria; N, nucleus; O, lipid storage bodies; V, vacuole. Bar in (A) and (C) = 1 gm. Bar in (B) and (D) = 0.5 gm.
152
Lonsdale, McDonald, and Jones
specimens in thin layers of LRW allowed samples to be examined and specifically selected for sectioning, based on morphological characteristics, removing the time-consuming task of sectioning random specimens to determine if the cell or tissue of interest is present. The ability to polymerize thin layers of LRW in the MW oven, rather than an embedding oven, provides another way to increase the efficiency of specimen preparation for TEM. As with any protocol involving the clinical MW, polymerization of thin layers of LRW in the MW is rapid. When used in combination with previously described methods for MW fixation, dehydration, and infiltration (Giberson et al., 1997; Lonsdale et al., 1999), the total time for block preparation for TEM can be as short as 5 h. Living specimens can be taken through the tissue preparation, embedding, sectioning, and immunolabeling steps in just 2 d. When combined with the time-saving advantages of TLE, MW polymerization of thin layers of LRW greatly increases the efficiency of specimen preparation for TEM and immunogold labeling. ACKNOWLEDGMENTS The authors thank Paula Sicurello and Reena Zalpuri for technical assistance during the development of this technique. This research was supported by grants from the National Science Foundation and the Competitive Research Grants Program of the US Department of Agriculture. REFERENCES Bush DS, Comejo M-J, Huang C-N,Jones, RL (1986) CaZ+-stimulatedsecretionof m-amylase during development in barley aleurone protoplasts. Plant Physio182:566-574. Giberson RT, Demaree RS, Nordhausen RW (1997) Four-hour processing of clinical/ diagnostic specimens for electron microscopy using microwave technique. J Vet Diagn Invest 9:61-67. Griffiths G (1993) Embedding media for section immunocytochemistry. In Fine Structure Immunocytochemistry. Springer-Verlag, Heidelberg, Germany, pp. 90-136. Hobot JA, Newman GR (1993) Resin Microscopy and On-Section lmmunocytochemistry. Springer-Verlag, Berlin, Germany. Howard RJ, O'Donnell KL (1987) Freeze substitution of fungi for cytological analysis. Exp Mycol 11:250-269. Huang AHC, Trelease RN, Moore TS (1983) Plant Peroxisomes. Academic, New York. Johnson KD, Herman EM, Chrispeels MJ (1989) Abundant, highly conserved tonoplast protein in seeds. Plant Physio191:1006-1013. Kiss JZ, McDonald KL (1993) Electron microscopy immunocytochemistry following cryofixation and freeze substitution. In Asai, DJ, ed, Antibodies in Cell Biology, vol. 37. Academic, San Diego, CA, pp. 311-341.
M W Polymerization in Thin-Layer LRW
153
Lonsdale JE, McDonald KL, Jones RL (1999) High pressure freezing and freeze substitution reveal new aspects of fine structure and maintain protein antigenicity in barley aleurone cells. Plant J 17:221-229. McDonald K (1999) High-pressure freezing for preservation of high resolution fine structure and antigenicity for immunolabeling. In Hajibagheri N, ed, Electron Microscopy Methods and Protocols, vol. 117. Humana, Totowa, NJ, 77-97. McDonald KL (1994) Electron microscopy and EM immunohistochemistry. Methods Cell Bio144:411-444. Reymond O, Pickett-Heaps J (1983) Routine fiat-embedding method for electron microscopy of microorganisms allowing selection and precisely oriented sectioning of single cells by light microscopy. J Microsc 130:79-84. Taylor J (1984) Correlative light and electron microscopy with fluorescent stains. Mycologia 76:462-467. Trelease RN, Gruber PJ, Becker WM, Newcomb EH (1971) Microbodies (glyoxysomes and peroxisomes) in cucumber cotyledons. Plant Physio148:461-475.
2
In Vivo Microwave-Assisted Labeling of Allium and Drosophila Nudei MarkA. SandersandDavidM. Gartner
INTRODUCTION Ultrastructural studies usually require that the sample be sacrificed and the tissue of interest be prepared by specialized fixation procedures. The goal of any given ultrastructural study is to observe and record the distribution and organization of a cellular constituent(s), in a way that most closely represents its in vivo distribution. Chemical fixations with aldehydes or cold solvents are often employed to preserve specimens for ultrastructural studies. However, these studies are only "snapshots" in time of the chemical termination of the sample. There can often be difficulties in interpreting the results, because of the harsh manipulations that chemical fixations can induce upon the specimen. Another practical disadvantage of chemical fixation in the preparation of biological specimens for microscopy is the time required to complete the process. In recent years, advances in stain technology, microscope environmental controls, and molecular techniques have permitted the study of numerous biological model systems under in vivo conditions. This is particularly true of specimens genetically modified to express green fluorescent protein, which has given researchers the ability to observe specific labeled structures by in vivo fluorescence microscopy (Tornehave et al., 2000). To this end, this report introduces a method of using a microwave irradiation (MWI) system to enhance, accelerate, and increase the penetration depth of nuclear DNA staining in two model systems in vivo: Drosophila melanogaster and Allium sp. root meristems. The organisms used in this study were selected because of From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ 155
156
Sanders and Gartner
their broad use as model systems in studying various genetic, structural, and cell-biology techniques, and the difficulty of conventional specimen preparation. Even for vital stains, these specimens required enzymatic treatment, dissection, or microinjection. The nucleic acid (NA) probes used in this study have been successfully used as vital nuclear labels in numerous studies on model systems that have readily accessible plasma membrane surfaces (Cooper et al., 1999; Frey, 1995). However, these probes have met with limited success in plant and insect systems in which there are cell barriers to the plasma membrane. The authors' previous experience indicated that Syto 13 (Molecular Probes, Eugene, OR) (488 nm excitation, 518 nm emission) had been effective in Allium root-tip nuclear staining at 10-50 ~tM/mL. Syto 62 (Molecular Probes) at 20-50 ~tM/mL (652 nm excitation, 676 nm emission), had proven effective in Drosophila staining, using 647 nm line on krypton/argon laser systems for observation. These observations lead the authors to attempt to facilitate vital staining of difficult specimens using MWI. This study demonstrates that Syto vital NA stains will label nuclear DNA in vivo, and the labeling is far superior to that of control treatments made in the absence of MWI. Also, benefits of this technique can be tremendous. Under the light microscope, the observer can select cells of different types, developmental stages, and ultrastructural permutations, in advance of additional experimental paradigms. MATERIALS AND METHODS
Reagents Highest purity reagents and deionized water should be used. All reagents are from Sigma, unless otherwise indicated. Be certain to adjust pH to the indicated value at room temperature (temp), and filter all solutions through a 0.2-~t filter. 1. Nuclei-staining buffer (NSB): Combine 10 mM MES-KOH (pH 5.5), 0.2 M sucrose, 2.5 mM ethylenediamine tetraacetic acid, 2.5 mM dithiothreitol, 0.1 mMspermine, 10 mMNaC1,10 mMKC1,1% glycerol. 2. Syto staining solution: Working dilution of 10-50 ~tM/mLSyto fluorescent NA stains (Molecular Probes,) in NSB and (optional) 1 mg/mL DNase free RNase A. Note: Too much RNase will not effect the cells, however, lower concentration of RNase will effect the quality of analysis by preventing good NA staining.
In Vivo MW-Assisted Labeling
157
MW Setup The MW processing portion of the procedure was carried out using a Model 3451,900-W programmable MW oven (Ted Pella, Redding, CA). The unit came with MW power controller, to set different power levels; load cooler, to recirculate and cool water within the MW cavity; software for temp and time monitoring; and a temp probe, to monitor and control temp maximums during processing. Inside the MW samples were placed on a Cold Spot TM (Ted Pella) (Fig. 1). This device served as the water load and processing surface for the samples. The load cooler recirculated and cooled the water within the Cold Spot, TM and maintained the temp below 25°C during processing. The samples were processed using either a MW-transparent, Teflon ® spot plate, containing 100 ktL staining solution per tissue-containing well (Allium), or on 24 × 50 mm no. 1 glass cover slips (Drosophila). Samples were MW-irradiated at ~250 W (power setting 1) for 2-4 min, allowed to rest for 2 rain, and once again processed for an additional 2-4 rain. During the MW procedure, tissue temp was allowed to rise no higher than a preprogrammed 25°C level, then was held at that level by means of a rapid on-off cycling of the magnetron power source, controlled by feedback from the temp probe. The temp probe was inserted into a well containing 100 ktL NSB, which provided thermal data throughout the procedure. Temp rise was recorded and plotted over time. In this fashion, the authors were able to generate MW energy capable of allowing access of the stain, without significant thermal energy, i.e., heating or boiling.
Specimen Preparation ALLIUMSP. ROOT MERISTEM
Allium sp. seeds were allowed to germinate on a moist paper towel inside a ziplock plastic bag stored at room temp. When root meristems had elongated to 3-5 mm (2-4 d), the distal 2 mm was excised and immediately placed into NSB for processing. The specimen was then placed into a 12-cavity Teflon embedding mold (Ted Pella), but any Teflon container (i.e., Teflon spot plate ®) should work. A 100-ktL NSB/ Syto stain was used to cover the specimen.
DROSOPHILAEMBRYOS Virgin wild-type (WT) Oregon R females were mated with WT Oregon R males for 3 d at 25°C, on standard cornmeal media. To avoid embryo crowding and lethality caused by anoxia, embryos were collected forup to 6 h, at 3-h intervals, on grape juice agar plates at 25°C.
158
Sanders and Gartner Temperature Probe Port Load In
~\
Load Out ~
/
Fig. 1. Cold SpotTM drawing, detailing the water in-flow (Load In) and out-flow ports (Load Out). The body of the Cold SpotTM is filled with water, which is recirculated and cooled by the load cooler from the MW. After the collection of embryos, the total number of embryos was determined. From 30-36 embryos were exposed to --250 W of MW energy for the time indicated in MWI Procedure. The embryos were transferred to glass vials containing cornmeal agar media, and incubated at 25°C. The number of hatched first instar larvae and empty chorions was determined. Embryos from WT Oregon R females were collected on agar culture media containing grape juice, at 20-45-min intervals. Embryos were collected and mounted onto glass 24 x 50-mm cover slips (no. 1 thickness) coated with a thin film of glue that was prepared by dissolving doublesided tape adhesive in heptane (Minden et al., 1989). Embryos were dechorionated by hand, and, depending on the relative humidity, embryos were briefly desiccated for 4-8 min, using a covered dish containing anhydrous CaSO 4. The embryos were then covered with 30 gL 20 gM Syto stain in NSB, and covered with oxygenated halocarbon oil (Series 700; Halocarbon Products).
M W I Procedure The sample holder containing the specimen was placed on the glass surface of the cold spot. The MW vacuum (vac) chamber, less the base
In Vivo MW-Assisted Labeling
159
(cat. no. 3435, Ted Pella) was placed over the sample holder on the cold spot (Fig. 2). The MW temp probe was placed through the probe hole of the chamber, so that it rested in a sample holder well that contained only NSB (no sample). A 10-15 mmHg vac was pulled on the chamber with an oil-less vac pump (cat. no. 3436, Ted Pella). Samples were then exposed to MW treatment for 2 min at a MW power setting of--250 W (no. 1 power level), 2-min pause, followed by 2 min at a MW power setting of approx 250 W (no. 1 power level). The temp restriction was set at 25°C for the entire process. Total incubation time was 6 min. The chamber was slowly vented, and the samples removed. A gas-permeable gasket was made, using Carolina Observation Gel (cat. no. 63148-62-9, Carolina Supply, Burlington, NC). The specimen was then placed into the well made by the gasket, and NSB added to fill the cavity. A 22 x 22 mm no. 1 cover slip was then placed over the sample and pressed gently to seal (cover slips can be viewed from either side). Sample viability can be maintained for up to 1 wk under these conditions.
Microscopy and Image Acquisition Images were collected, using a B io-Rad MRC 1024 scanning confocal system mounted on a Nikon Diaphot 300 microscope, equipped with a 15-mW krypton/argon laser. A x60/1.4 NA Plan apochromatic and objective lens was used for all analyses. Four-dimensional images (three-dimensional over time) were recorded every 20 s for up to a 6-h period, using Bio-Rad Lasersharp software time-lapse features, at slowscan mode. Images were stored to disk for later observation/analysis. Digital files were processed in NIH Image. Individual still frames were saved as PICT or TIFF files, and Adobe PhotoShop was used to adjust image size, contrast, and pseudocoloring of images. Prints were made, using the Fujix Pictography 3000 (Fuji, Elmsford, NY). RESULTS Although these methods produce successful results, some modification of the procedures, found through empirical trial and error, may be required to obtain optimal results.
Allium sp. Nuclear Staining Patterns Allium root meristems have approx 20% of cells in the mitosis stage of the cell cycle, with the mitotic event taking approx 1-4 h to complete, to form two new daughter cells. This system allows for the observation of DNA during higher plant mitotic events, under conditions that require minimal environmental modifications.
160
Sanders and Gartner
Fig. 2. Pelco 3450 Laboratory Microwave System, as used for in vivo labeling. Samples are placed on the surface of the Pelco Cold Spot, T M which is cooled to <25°C using the Pelco 3420 Microwave Load Cooler. A Pelco 3435 MW vac chamber is placed over the sample and a vac of 10-15 mmHg is obtained. The temp probe is inserted and placed into a blank sample that contains only buffer. Samples were then exposed to MW treatment for 2 min at a MW power setting of approx 250 W (no. 1 power level), then a 2 min pause, followed by 2 min at a MW power setting of approx 250 W (no. 1 power level).
The majority of mitotic cells typically lie between 20 and 200 ktm below the epidermal layer of cells, which makes both labeling and visualization difficult. MWI permitted the NA Syto stains to penetrate and label cells throughout the thickness of the specimen in vitro. Staining with Syto 13 provided excellent nuclear labeling, without background interference, with confocal scanning laser microscopy (Fig. 3A). In control samples treated identically, without MWI, there was sporadic labeling of nuclei in the epidermal layer, but no detectable labeling of the nuclei of the cortical layer, where the mitotic cells reside (Fig. 3B). Time-lapse observations have shown that the nuclear morphology is maintained for at least 2 h, following root excision and MWI-assisted labeling (data not shown).
In Vivo MW-Assisted Labeling
161
Fig.3. Allium root tip meristem labeling, using Syto 13 (Molecular Probes) vital NA stain. MW-assisted labeling (A) consisted of 6 min total incubation time. Samples were incubated in 10 gM Syto 13 for 2 min MW treatment at 250 W, 2 min post-MW incubation, 2 min MW treatment at 250 W. Optical sections of groups of cells were collected on a Bio-Rad 1024 laser-scanning confocal microscope, using 488 nm and 647 nm (autofluorescence of cell wall constituents) excitation wavelengths. Digital maximum projections were made from the --20 gm thick-stack of optical sections. Note mitotic profiles of nuclei. (B) Demonstrates labeling in absence of MWI. Incubation time, 15 min. Scale bar = 10 t.tm. Viability of Drosophila Embryos Following MWI Initial experiments, using MWI of living Drosophila embryos, revealed that treatments for up to 2 rain at low power (--250 W at 25°C), did not lead to detectable losses of viability (Fig. 4). After MWI, WT embryos progressed normally through several rounds of mitosis (Fig. 5A-F). Nuclear divisions were synchronous, and proceeded in well-organized waves across the embryo. This orderly progression of nuclear cycles results in an evenly spaced monolayer of nuclei at the surface of the syncytial blastoderm. As noted by others (e.g., Sullivan et al., 1990), the authors did occasionally observe nuclei that failed to complete mitosis. Such nuclei lose their association with the cell cortex and rapidly depart into the interior of the embryo: This event is termed "nuclear fallout." The authors did not see an increase in the number of nuclear fallouts, compared to other reports (Sullivan et al., 1990), following MWI. Time-lapse observations have shown that viability is maintained for several rounds of nuclear division following MWI-assisted labeling (Fig. 4).
% Viability of Drosophila Embryos following MWI 100 90 80 70 60 50 40 30 20 10 0 1
1.5
2
2.5
3
3.5
4
4.5
5
5.5
6
6.5
7
Fig. 4. Graph of percent of viable Drosophilaembryos following MWI at a MW power setting of approx 250 W (no. 1 power level). A total of 94 embryos were examined.
Fig. 5. Living Drosophila embryos, labeled with Syto 62 (Molecular Probes). Time-course of Drosophila WT embryos in vivo by confocal microscopy. The embryos were collected from WT Oregon females, and exposed to MWI in the presence of 10 pm Syto 62. Shown are successive confocal images (A-F), selected from time-lapse collections taken at 20-s intervals. A - F follow a sequence of a mitotic cycle in a representative WT embryo. Shown are fields of nuclei progressing through mitosis. Scale bar = 10 pm.
In Vivo MW-Assisted Labeling
163
DISCUSSION Following the early work of Mayers (1970), MWI has been utilized to accelerate biological specimen processing for light and electron microscopy. Because it can both raise the temp and significantly increase molecular motion (Boon et al., 1990), MWI enables chemicals and other small molecules, such as stains or fixatives, to diffuse rapidly into cells and tissues. Additionally, MW technology can be used for antigen retrieval and antibody labeling on sections of fixed and processed tissue. Performing the antibody reaction, under controlled MW conditions, results in significant improvements, such as shorter incubation times and lower antibody concentrations (Stone et al., 1999). MWI also facilitates detection of structural details difficult to localize with conventional methods (Boon et al., 1990; Gu et al., 1995). These observations encouraged the authors to attempt to facilitate vital staining of difficult specimens, using MWI. The use of MWI greatly shortened the processing and labeling times, without apparent compromise of quality ultrastructural preservation or the specificity of labeling, in the two model systems used in this study. When samples for in vivo microscopy are processed, the temp must be critically controlled, because temp that are too high could alter the integrity and viability of the tissues. Temp control can be achieved in several ways. A temp probe with a feedback mechanism is used to regulate the energy output of the MW to maintain the desired temp. Additional control can be achieved by placing a water load in the chamber of the MW. The water helps absorb extra energy, and distributes the energy evenly throughout the chamber (Boon et al., 1990; Login and Dvorak, 1993; Gu, 1994; Giberson et al., 1995). The authors' results demonstrate that microwaving is a useful method for in vivo staining on difficult specimens, such as Drosophila melanogaster and Allium sp. It should be noted that excessive MWI could lead to overheating and cell death. The details may not necessarily apply to other ovens, stains, or buffers. Thus, for individual applications, different durations of MWI should be tested. In addition, appropriate temp control is essential. The cooling capacity of the recirculating water load of the cold spot and the rapid cycling of the magnetron power supply, utilizing sample temp feedback, can control the temp. Additionally, this setup allows energy from MWI to be evenly distributed throughout the chamber. The use of 10-15 mmHg vac to the samples is not essential, but does contribute to more consistent and uniform labeling. Successful nuclear staining of these in vivo or in vitro model systems by a standard 2-6-min
164
Sanders and Gartner
microwaving at 250 W is possible. The increased availability of green fluorescent protein-expressing living model systems, paired with the use of vital probes, make this procedure highly desirable. REFERENCES Boon ME, Hendrikse FCJ, Kok PG, Bolhuis P, Kok LP (1990) Practical approach to routine immunostaining of paraffin sections in the microwave oven. Histochem J 22:347-352. Cooper MS, D' Amico LA, Henry CA (1999) Confocal microscopic analysis of morphogenetic movements. Methods Cell Bio159:179-204. Frey T (1995) Nucleic acid dyes for detection of apoptosis in live cells. Cytometry 21:265-274. Giberson RT, Demaree RS Jr. (1995) Microwave fixation: understanding the variables to achieve rapid reproducible results. Microsc Res Tech 32:246-254. Gu J (1994) Microwave Immunocytochemistry. In Gu J, Hacker GW, eds, Modern Methods in Analytical Morphology. Plenum, New York, pp.67-80. Gu J, Choi T-S, WhittleseyM, Slap S, Anderson V (1995) Development of microwave immunohistochemistry. Cell Vision 2:257-259. Login GR, Dvorak AM (1993) Review of rapid microwave fixation technology: its expanding niche in morphological studies. Scanning 15:58-66. Mayers CP (1970) Histological fixation by microwave heating. J Clin Patho123:273-275. Minden JS, Agard DA, Sedat JW, Alberts BM (1989) Direct cell lineage analysis in Drosophila melanogaster by time-lapse, three-dimensional optical microscopy of living embryos. J Cell Biol 109:505-516. Stone JR, Walker WA, Povlishock JT (1999) Visualization of a new class of tramatically injured axons through the use of a modified method of microwave antigen retrieval. Acta Neuropatho197:335-345. Sullivan W, Minden JS, Alberts BM (1990) Daughterless-abo-like, a Drosophila maternal-effect mutation that exhibits abnormal centrosome separation during the late blastoderm divisions. Development 110:311-323. Tornehave D, Hougaard DM, Larsson L-I (2000) Microwaving for double indirect immunofluorescence with primary antibodies from the same species and for staining of mouse tissues with mouse monoclonal antibodies. Histochem Cell Biol 113:19-23.
3
Microwave-AssistedCytochemistry Accelerated Visualization ofAcetylcholinesterase at Motor Endplates John R Petrali and Kenneth R. Mills
INTRODUCTION Acetylcholinesterase (ACHE) is the modulating enzyme of cholinergic systems. Study of its morphological distribution and its pathophysiological disposition following disease or exposure to anticholinesterase compounds such as pesticides, organophosphates, and chemical (chem) warfare nerve agents, has been through the use of a multitude of specific cytochem reactions which use metal-capturing methods of thiocholine, following hydrolysis of acetylthiocholine by ACHE. The first of these methods to be recognized and used in many investigations was the Koelle technique (Koelle and Friedenwald, 1949). This in situ procedure used copper (Cu) as a capturing metal which without any further treatment yielded dense precipitates localized at sites of enzyme activity. These precipitates could be visualized at the light-microscopy and electron microscopy (EM) level. The empirical reaction for the primary cytochem Cu- capturing reaction is~:as follows" Acetylthiocholine iodide + AChE ~ Thiocholine iodide + Acetate Thiocholine iodide + Cu sulfate ~ Cu thiocholine iodide *The opinions or assertions herein are the private views of the authors, and are not to be construed as official or as reflecting the views of the Army or the Department of Defense. In conducting this research, the investigators adhered to the "Guide for Care and Use of Laboratory Animals of the Institute of Laboratory Animal Resources National Research Council." From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ 165
166
Petrali and Mills
The intermediate white precipitate, Cu thiocholine iodide, could be darkened by sequential incubation with a variety of secondary reactants, such as, potassium ferricyanide (Karnovsky and Roots, 1964), sodium sulfide/silver nitrate (Namba et al. 1967) and phosphomolybdic acid (Tsuji and Fournier, 1984). In the case of the secondary reactants Na sulfide/Ag nitrate the final reaction product is a dark-toning step similar to photographic development which yields an insoluble dark precipitate easily seen at reactive sites. The time required by most laboratories for conventional cytochem processing of AchE, which includes fixation of selected tissues, followed by primary and secondary incubations, is typically 8-24 h. With the advent of microwave (MW)-assisted tissue processing and MW-assisted cytochem incubations, times required for visualization of AChE can be shortened significantly. Here is presented a MW-accelerated modified Koelle-Friedenwald-Na sulfide/Ag nitrate protocol for ACHE, useful for diagnostic and investigative pathology (PATH). Guinea pig diaphragm is used as the subject tissue. The resultant histological and ultrastructural presentations of the enzyme at motor endplates are evaluated and compared with a companion study of the same AChE method using conventional procedural times. In this study, time required for the conventional procedure was 7 h. Time required for the accelerated procedure was 1.7 h. MATERIALS A N D M E T H O D S Since conventional methodology and procedural times for AChE visualization are well documented, this chapter presents only materials and methods for the MW-accelerated protocol of AChE localization used in this study. All tissues incubated for AChE localizations were eventually processed for routine hematoxylin and eosin histopathological study and transmission EM. The lab MW oven utilized for this study was the Pelco 3440, 800 W. The animal used was the haired guinea pig, which was euthanatized by an overdose of Na pentobarbital (1 mL/kg ip), followed by the induction of a pneumothorax. The diaphragm was removed immediately, placed in a Petri dish with saline and processed according to the following protocol.
MW Aldehyde Fixation MATERIALS
1. Buffered fixative (1/2 Karnovsky/0.1 M Na cacodylate, pH 7.4, 190 mOsm). 2. Two 250-mL polypropylene beakers. 3. Temperature (temp) probe.
MW-Assisted AChE Labeling
167
PROCESS 1. Place two 250-mL beakers containing 200 mL water (23°C) in MW, as heat sinks. 2. Remove saline from Petri dish, and add 10 mL fixative. 3. Place sample in MW at a predetermined cold spot. 4. Set temp set point at 37°C, and place temp probe in Petri dish with sample. 5. Irradiate for 10 s at 100% power, 20 s at 0% power, and 10 s at 100% power. 6 Continue fixation outside of MW for 5 min.
B uffe r Was h MATERIAL REQUIRED Hank' s buffer.
PROCESS 1. Remove fixative. 2. Add buffer to Petri dish, and wash (outside MW) 3 x 10 min.
Preincubation Medium MATERIAL Medium (12.5 rnL 0.1 M Na citrate, 25.0 mL Hank's buffer, 12.5 mL 0.06 M Cu sulfate). PROCESS 1. Remove buffer. 2. Incubate tissue in medium (outside MW) for 10 min.
Primary Incubation Medium MATERIAL Medium (12.5 mL 0.1 M Na citrate, 25.0 mL Hank's buffer, 12.5 mL 0.06 M Cu sulfate, 50 mg acetylthiocholine iodide).
PROCESS 1. Remove preincubation medium from Petri dish, and add 10 mL primary medium. 2. Place sample in MW at a predetermined cold spot. 3. Set temp set point at 37°C, and place temp probe in Petri dish. 4. Irradiate for 30 s at 100% power. 5. Continue incubation outside of MW for 5 min.
168
Petrali and Mills
B uffe r Was h MATERIALS Hank's buffer. Paocrss 1. Remove primary incubation medium. 2. Add buffer to Petri dish, and wash (outside MW) 3 x 5 rain.
Secondary Incubation Medium MATERIALS 1% Na sulfide, deionized water, 0.5% Ag nitrate, 1% Na thiosulfate. PROCESS 1. Remove buffer, and transfer tissue to a 20-mL scintillation vial. The following steps are done outside the MW. 2. Incubate tissue in 1% Na sulfide for 1 min. 3. Remove Na sulfide. 4. Wash with deionized water 3 × 5 min. 5. Remove water. 6. Incubate in 0.5% Ag nitrate for 1 min. 7. Remove Ag nitrate. 8. Wash with deionized water 3 x 5 min. 9. Remove water. 10. Incubate in 1% Na thiosulfate for 5 min. 11. Remove Na thiosulfate. 12. Wash with deionized water 3 × 5 min, and process for light microscopy or EM. RESULTS Histopathological examination of MW-processed diaphragm revealed AChE reaction products strongly specific for motor endplates to the exclusion of other muscle sites (Fig. 1). Except for focal areas of rarefaction of some muscle fibers, the structural presentation of the diaphragm and localization of the enzyme were unchanged from that seen with conventional fixation and incubations (Fig. 2). At the ultrastructural level, with the exception of occasional dilatation of mitochondria and some minor presynaptic swelling surrounding presynaptic vesicles, the subcellular details of the muscle fibers at motor endplates were mostly unchanged from that seen with conventional processing. Reaction products for AChE were specific for the primary cleft and junctional folds of motor endplates (Fig. 3), and were mostly replicate, compared with conventional processing (Fig. 4).
MW-Assisted AChE Labeling
169
Fig. 1. Histological section of MW-processed diaphragm, showing AChE reaction product (arrows) at a motor endplate.
Fig. 2. Histological section of conventionally processed diaphragm, showing AChE reaction product (arrows) at a motor endplate.
170
Petrali and Mills
Fig. 3. Ultrastructural section of MW-processed diaphragm, showing AChE reaction product (arrows) in the primary cleft and junctional folds of a motor endplate, m, dilated mitochondrium; psv, presynaptic vesicles. Fig. 4. Ultrastructural section of conventionally processed diaphragm, showing AChE reaction product (arrows) in the primary cleft and junctional folds of a motor endplate.
MW-Assisted AChE Labeling
171 DISCUSSION
Results of this study add to the growing value of the lab MW oven in diagnostic and investigative PATH. The 79% decrease in time required, to achieve results that closely approach or are equal to those of conventional methods for the characteristic localization of ACHE, makes MW processing for this enzyme now rapid and reliable. Although requiring standardized time for hematoxylin and eosin paraffin processing, light microscopy evaluations in this study were especially useful for quick diagnostic evaluations of large expanses of diaphragm for the presence or inhibition of the enzyme at motor endplates. The ultrastructural portion of this study attests to the ability of MW processing to shorten response times, when addressing the investigative subcellular disposition of the enzyme. The authors' lab, which now embraces the lab MW oven as an on-line integral part of tissue processing to include special staining and embedding, has now significantly reduced the extended times needed by most EM labs to determine ultrastructural localizations of ACHE. However, still lacking are standardized, uniform lab MW setups for any given procedure. MW literature is resplendent with personalized strategies regarding, among others, cold spots, hot spots, specimen size, temps, sensor-probe placements, fixing solutions, position of heat sinks, power applications, and so on. This ambiguity is reminiscent of the halcyon days of the EM sciences, when all was in formative stages. When MW processing does become standardized, as it must, then it might be suggested, at that juncture, that the lab MW oven might well prove to be the one instrument that finally places the EM sciences in the mainstream of diagnostic and investigative PATH. REFERENCES Karnovsky MJ, Roots L (1964) Direct-coloring thiocholine method for cholinesterase. J Histochem Cytochem 12:219-221. Koelle GB, Friedenwald JS (1949) Histochemical method for localizing cholinesterase activity. Proc Soc Exp Biol Med 70"617-622. Namba T, Nakamura T, Grob D (1967) Staining for nerve fiber and cholinesterase activity in fresh frozen sections. Am J Clin Patho147:74-77. Tsuji S, Fournier M (1984) Ultrastructural localization of acetylcholinesterase activity by means of the electron dense precipitate derived from Koelle's cuprous thiocholine iodide by treatment with phosphomolybdic acid and osmium tetroxide. Histochemistry 80:19-21.
Immunoelectron 14 Microwave-Assisted Microscopy of Skin Localization of Laminin, TypeIV Collagen, and BullousPemphigoidAntigen John R Petrali and Kenneth R. Mills
INTRODUCTION Chemical (chem) fixatives and conventional fixation times, used for standardized preservation of tissues, can result in serious alterations to morphology, as a consequence of solubilization and conformational changes of proteins and lipids (Login and Dvorak, 1985). These untoward changes typically result in compromised antigenicity of many tissue proteins and loss of enzyme specificity. Microwave energy (MWE), used in conjunction with, or as an alternative to, routine chem processing, is gaining increasing support for use in diagnostic and investigative laboratories. With MWE processing, tissue antigens can be rapidly and distinctly better preserved, antigen retrieval made more replicate, and histochem and immunochem reactions made ultrafast (van Vlijmen-Willems and van Erp, 1993). These desirous applications, used in concert now with dedicated lab MW ovens, make the biotechnology of MWE a valuable addition to diagnostic immunoelectron microscopy (IEM). The opinions or assertions herein are the private views of the authors, and are not to be construed as official or as reflecting the views of the Army or the Department of Defense. In conducting this research, the investigators adhered to the "Guide for Care and Use of Laboratory Animals of the Institute of Laboratory Animal Resources National Research Council." From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ 173
174
MW-Assisted IEM of Skin
The present study had two objectives: to use emerging MWE processing techniques for determining immunoultrastructural locations of selected skin basement-membrane zone (BMZ) proteins of special interest in the study of vesicating skin lesions, and to compare MWEprocessed skin with conventionally processed skin, for ultrastructural integrity. Proteins selected were bullous pemphigoid antigen (BPA), a protein of hemidesmosomes, known to be antigenically altered in some bullous diseases; laminin, the major glycoprotein of the lamina lucida of the BM used immunochemically for determining extent of bullous lesions; and type IV collagen, the ubiquitous protein of the lamina densa of BM, used immunohistochemically to demarcate dermal penetration of vesicating lesions. All three proteins are considered labile to routine chem processing, usually requiring cryoprocedures or enzymatic digestion for their presentation (Moran et al., 1988; Monteiro and Inman, 1995; Petrali and Oglesby-Megee, 1997). MATERIALS A N D M E T H O D S
MW Processing (Fig. 1) Full-thickness skin samples were extracted by scalpel from site-shaved normal-haired guinea pigs euthanatized with sodium pentobarbital (80 mg/kg). Skin specimens were dissected to 4.0 mm cubes and immersed in a 20-mL polyethylene plastic fixing vial, containing 5 mL 2% EM-grade formaldehyde and 0.05% glutaraldehyde buffered with 0.1 M Na cacodylate (pH 7.4, 190 mOsm). The fixing vial was placed on a polystyrene platform in a lab MW oven (Pelco 3440, 800 W), 20 mm above the oven floor, and positioned at a cold spot, as determined by a neon bulb array (Login and Dvorak, 1994). A 500-mL glass beaker, containing 300 mL distilled water at 23°C, was placed in the rear left comer of the oven chamber, as a heat sink. Specimens were irradiated according to the following sequence: 10 s at 100% power, 20 s at 0% power, 10 s at 100% power, all to a temperature (temp) of 50°C, as monitored by a temp probe placed within the specimen vial (Giberson et al. 1997). Total time in the MW oven was 40 s. Specimens were immediately removed from the vial and washed 3x with cacodylate buffer. Replicate skin specimens fixed for 2 h at room temp by immersion in the same combined aldehyde fixative, but not microwaved, were processed as controls.
Immunocytochemistry Razor blade sections of MWE-fixed and control skin specimens were processed for pre-embedding immunocytochem, using the indirect
Petrali and Mills
175
Extract X4k Tissue
Immersion Fluid (aldehydes)
Microwave to 50°C (40 sec)
Process for EM
Fig. 1. Processing steps, from the animal to embedded block. immunoperoxidase method. Primary antibodies (BPA [1:20], laminin [1:100], and type IV collagen[l: 100] were sequenced with peroxidaseconjugated goat antihuman for BPA, and goat antirabbit for laminin and for type IV collagen, all at dilutions of 1:20. All immunoreagents were diluted in phosphate buffered saline. The immunostaining sequence was 3% normal goat serum, 20 min; primary antibody, 90 min; secondary antibody, 45 rain; diaminobenzidine, 8 min; 1% cacodylate-buffered osmium tetroxide, 1 min. Immunostained sections were washed, stored overnight in cacodylate buffer at 4°C, dehydrated in graded ethanol, and embedded in epoxy resins. Ultrastructural analysis was performed on ultrathin epoxy sections, with and without differentiation with uranyl acetate. RESULTS Immunoultrastructural reactivity for all subject proteins in MWEprocessed skin was specific to the BMZ to the exclusion of other skin sites. BPA immunoreaction products were localized to hemidesmosomal plaques within the epidermal basal cell and basal cell membrane, and to subadjacent regions of the lamina lucida of the BM (Fig. 2A). The lamina densa was free of reaction products. No staining for BPA was evident in nonirradiated control skin (Fig. 2B). Laminin was localized to the lamina lucida of the BM of MWE-processed skin (Fig. 3A). Basal cell membranes and hemidesmosomal plaques were free of reaction products for laminin. Control sections were not reactive for laminin (Fig. 3B). Localization of type IV collagen was strongly specific to the lamina densa of the BM (Fig. 4A). Collagen fibers of the dermis immediately subadjacent to the lamina densa were also reactive. No staining
176
MW-Assisted IEM of Skin
~ii~i~¸~!)~i~
.... ii~!iiiii~
•
.....
:. :.•:~:~.~i¸::~!~ /!iiiiiill
i¸¸ ~i~i~,~i :~!
~
Fig. 2. (A) MWE-processed skin shows BPA (arrows) localized to hemidesmosomal plaques, basal cell membrane, and subadjacent regions of the lamina lucida, ep, epidermis; d, dermis. (B) No staining for BPA is evident in nonirradiated control skin. ep, epidermis; d, dermis. Magnification: x30,000.
Petrali and Mills
177
Fig. 3. (A) MWE-processed skin shows laminin (arrows) localized to the lamina lucida of the BM. ep, epidermis; d, dermis; bc, basal cel. (B) Control section shows no reactivity for laminin, ep, epidermis" d, dermis.Magnification: x30,000.
178
MW-Assisted IEM of Skin
Fig. 4. (A) MWE-processed skin shows type IV collagen (arrows) strongly specific for the lamina densa of the BM. Collagen fibers (arrowheads) immediately subadjacent to the lamina densa are also reactive, ep, epidermis; d, dermis. (B) Control section shows no staining for type IV collagen, ep, epidermis ; d, dermis. Magnification: x30,000.
Petrali and Mills
179
was evident in any control section for type IV collagen (Fig. 4B). In all cases, localizations were most evident along the free margins of the embedded specimens. Immunoreaction products for each subject protein of the BMZ could be traced for several millimeters from free margins of the embedded specimen, with staining intensity waning toward the deeper recesses of the section. Control sections were never reactive for any of the proteins tested. DISCUSSION The results of this comparative study demonstrate the value of MWE for rapid preservation of ultrastructure and the retention of antigenicity for IEM presentation of skin BMZ proteins. Because of their instability during routine chem processing, immunohistochem demonstrations of BM proteins are usually restricted to cryosections, or to staining processes that require enzymatic digestion. In many cases, these procedures seriously compromise or actually preclude the use of IEM in their analyses. In the present study, 40-s MWE-fixed skin presented acceptable ultrastructure and precise anatomical localizations of the B MZ proteins, B PA, laminin, and type IV collagen. BPA localizations to hemidesmosomal plaques of the basal cell, and to subadjacent areas of the lamina lucida, are in agreement with, and reflect the purported specific sites of, the two antigenic epitopes of BPA: BPA 230 and BPA 180. BPA 230, also identified as BPAG l, has been shown to be associated with intracellular domains of the hemidesmosome; BPA 180, also identified as B PAG 2, binds extracellularly along the basal cell plasma membrane, in proximity to the lamina lucida of the BM (Regnier et al., 1985; Ishiko et al., 1993; Mutasim et al., 1989). Laminin immunoreactivity was selective for the lamina lucida of the BM, with some reaction products overlying the lamina densa. In no case was there internal staining of basal cells or of dermal collagen. Laminin, the major glycoprotein of the lamina lucida, has been reported to be altered to specific antisera during the development of vesicating lesions in some skin toxicities (Petrali and Oglesby-Megee, 1997). The specificity of type IV collagen for the lamina densa of the BM was clearly presented. In some sections, staining extended to dermal collagen immediately subjacent to the lamina densa, perhaps indicating some displacement of collagen fibers of the lamina densa during processing or crossreactivity with components of anchoring fibrils. The observation that aldehyde-fixed control skin specimens, fixed conventionally for 2 h at room temp, were devoid of any immunospecificity, attests to the damaging effect of the fixative when used without the benefit of MWE.
180
MW-Assisted IEM of Skin
This study allows the conclusion that MWE-assisted processing is a viable method for the immunoultrastructural presentation of skin BMZ proteins usually considered labile to conventional ultrastructural processing. This study appears in J. Toxicol Cut and Ocular Toxicol (1999) 18(4):341-348. ACKNOWLEDGMENT The authors acknowledge Dr. Grant Anhalt, Department of Dermatology, Johns Hopkins University, Baltimore, MD, for his guidance with, and generous supplies of B PA antisera. REFERENCES Giberson RT, Demaree RS Jr., Nordhausen RW (1997) Four-hour processing of clinical/ diagnostic specimens for electron microscopy using microwave technique. J Vet Diagn Invest 9:61-67. Ishiko A, Shimizu H, Kikuchi A, Ebihara T, Hashimoto T, Nishikawa T (1993) Human autoantibodies against the 230-kD bullous pemphigoid antigen (BPAG 1) bind only to the intracellular domain of the hemidesmosome, whereas those against the 180-kD bullous pemphigoid antigen (BPAG2) bind along the plasma membrane of the hemidesmosome in normal human and swine skin. J Clin Invest 91:1608-1615. Login GR, Dvorak AM (1985) Microwave energy fixation for electron microscopy. Am J Pathol 120: 230-243. Login GR, Dvorak AM (1994) Methods of microwave fixation for microscopy. Progr Histochem Cytochem 27: 72-94. Monteiro-Riviere NA, Inman AO (1995) Indirect immunohistochemistry and immunoelectron microscopy distribution of eight epidermal-dermal junction epitopes in the pig and in isolated perfused skin treated with bis (2-chloroethyl) sulfide. Toxicol Patho123: 313-325.
Moran RA, Nelson F, Jagirdar J, Paronetto F (1988) Application of microwave irradiation to immunohistochemistry: preservation of antigens of the extracellular matrix. Stain Techno163: 263-269. Mutasim D, Morrison L, Takahashi Y, Labib R, Skouge J, Diaz L, Anhalt G (1989) Definition of bullous pemphigoid antibody binding to intracellular and extracellular antigen associated with hemidesmosomes. J Invest Dermato192: 225-230. Petrali JP, Oglesby-Megee S (1997) Toxicity of mustard gas skin lesions. Microsc Res Techno137: 221-228. Regnier M, Vaigot P, Michel S, Prunieras M (1985) Localization of bullous pemphigoid antigen (BPA) in isolated human keratinocytes. J Invest Dermato185:187-190. van Vlijmen-Willems I, van Erp P (1993) Microwave irradiation for rapid and enhanced immunohistochemical staining: application to skin antigens. Biotechnol Histochem 68: 67-74.
15
i Microwave Paraf~n Techniques for Botanical Tissues Denise Schichnes, Jeffrey A. Nemson, and Steven E. Ruzin
INTRODUCTION The popularity of in situ hybridization (ISH) of nucleic acids and immunolocalization of proteins has caused a resurgence of interest in paraffin microtechnique by the plant biology community. However, the amount of time required for proper anatomical preservation of plant tissues results in degradation of nucleic acids and proteins in the sample (Jackson, 1991). There are protocols that may be used to reduce degradation, but they often result in poor anatomical preservation and target (nucleic acid, protein) conservation (Jackson, 1991; Kouchi and Hata, 1993; Ruzin, 1999). The authors became interested in developing a protocol for microwave (MW) paraffin embedding of plant tissue after reviewing the benefits that MW ovens have brought to the field of electron microscopy (Kok and Boon, 1989; Login and Dvorak, 1994; Giberson et al., 1997), and have developed a protocol that, in 5 h, yields embedded plant tissue in paraffin for ISH, immunolocalization, and standard anatomical observation (Schichnes et al., 1999). With this technique, the quality of tissue preservation for ISH and immunolocalization is superior to traditional procedures, which usually require 7 d to complete. The tissue preservation for anatomical study is equivalent to the traditional protocol (Johansen, 1940; Berlyn and Miksche, 1976; Ruzin, 1999), which requires a minimum of 9 d to complete. In addition, the authors have developed a MW protocol to mount paraffin ribbons to gelatin-coated From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ 181
182
Schichnes, Nemson, and Ruzin
slides (Haupt, 1930; Sass, 1958) and Fisher Probe-on Plus ® slides (Fisher, Pittsburgh, PA). The MW method requires 30 min, as opposed to the standard protocol, which requires a minimum of 6 h at 42°C. Finally, the authors have adapted a classic plant microtechnique staining protocol, Johansen's Safranin-O and Fast Green (FCF) protocol (Johansen, 1940), for the MW oven. This adaptation yields staining equivalent to the traditional procedure in 40 min, rather than 2 d. The authors' protocol was originally developed using Zea mays shoot apices. This tissue is difficult to embed in paraffin, because of variance in tissue densities. Additionally, the developing leaf primordia often trap air, which impedes the infiltration of solutions (Freeling and Lane, 1994). The MW protocol has been successfully applied to many other plant tissues, ranging from grasses to gymnosperms, and even Quercus suber (cork oak), with some modifications for particularly delicate plant tissues, such as Arabidopsis thaliana and friable callus. The benefits of this MW protocol have been threefold for this laboratory, the Biological Imaging Facility, a core facility for the University of California Berkeley campus: An increased quality of tissue has been achieved for ISH and immunolocalization studies, and the ability to process tissue is much more rapid; delicate tissue can now be embedded in paraffin, which was previously only usable if embedded in resin, saving an enormous expense in time and chemical cost; the combination of MW fixation, paraffin embedding, ribbon mounting, and staining protocols has proved an invaluable teaching tool, allowing the authors' plant microtechnique class to be condensed from one semester into a 1-wk workshop, and to cover more material. MATERIALS A N D M E T H O D S 1. 2. 3. 4. 5. 6. 7.
Materials Pelco 3440 MAX lab MW oven, with variable wattage control (Ted Pella, Redding, CA). PolyTemp Polysciences load cooling water bath (Polysciences, Warrington, PA). Static water load: 400 mL Tri-Corner beaker (Fisher Scientific) no. 2-593-50D with 400 mL water. Glass scintillation vial, 15-mL size (Fisher) no. 3-338-E. Sample water bath, plastic, measuring 8.5 x 12 x 5 cm. Fixative (FAA) (Ruzin, 1999): 50 mL 100% ethanol, 5 mL acetic acid, 10 mL 37% formalin, and 35 mL water. 10 x Phosphate-buffered saline (PBS) (Ruzin, 1999): 1.3 M NaC1, 0.07 M Na2HPO 4 , 0.03 M NaH2PO 4, and pH to 7.0. using NaOH.
MW Paraffin Techniques for Plants
183
8. Fixative PFA (Ruzin, 1999): 100mL 1 x phosphate-buffered saline; pH to 11.0 with NaOH; heat to 60°C; add 4 g paraformaldehyde; and stir until solution is clear; pH to 7.5; cool on ice. 9. 10 mM SCrensen's buffer, pH 7.2 (SCrensen, 1909): 5.6 mL 0.2 M NaH2PO 4, 14.4 mL 0.2 M Na2HPO 4, and 380 mL deionized water. 10. Paraplast Extra paraffin (Fisher) no. 12-646-113. 11. Modified Haupt's gelatin (Ruzin, 1999). 12. Glass slides. 13. Probe-On Plus ® slides (Fisher). 14. Johansen's Safranin-O and Fast Green (FCF) staining procedure (Johansen, 1940).
Process 1. Preparation of tissue: Dissect tissue into glass scintillation vial, with 10 mL fixative at 4°C, and keep on ice until ready to begin MW protocol. Use PFA as a fixative for immunolocalizations, use SCrensen' s buffer, with no chemical fixative, for delicate tissues, and use FAA for all other tissues. 2. Determining sample location: Time: --2 min. Using a neon bulb array, determine sample and water-load placements. Locations are marked on a sheet of paper taped to the oven floor. 3. MW fixation: Time:--45 min. a. Place glass vials in plastic water bath. Fill water bath until water level is equal to fixative level. b. Set the MW variable wattage to 650 W for most samples; 450 W for delicate tissue. Place temperature (temp) sensor in a vial with the samples. c. MW for 15 min, with temp limit set to 37°C. Replace with fresh fixative, and cool on ice to 12°C. Repeat twice. d. Change the static water load after each cycle. 4. Alcohol dehydration: Time: -10 min. a. MW sample at 67°C for 1 min, 15 s each step in the ethanol dehydration series. 1 × 50%, 1 × 70%, 1 x 70% with Safranin-O (0.1%), 2 × 100%. b. Change static water load. c. MW sample at 77°C for 1 min, 30 s each step in the isopropanol dehydration series. 1 × 50% ethyl alcohol (EtOH): 50% isopropanol, 1 × 100% isopropanol. d. Change static water load. 5. Infiltration: Time:-3 h. a. MW sample in 50% isopropanol:50% molten paraffin for 10 min at 77°C. b. MW sample in 100% molten paraffin for 10 min at 67°C. c. MW sample in 100% molten paraffin for 30 min at 67°C. Repeat 4x. d. Change water load after each cycle. e. Embed sample in aluminum or paper boats, and cool to room temp.
184
Schichnes, Nemson, and Ruzin
6. Mounting paraffin ribbons: Time: 30 min. a. Do not put slides directly on MW floor. Place them on cardboard, or stack in a glass staining dish placed on its side. b. Place temp sensor in a 100-~tL drop of water on an adjacent slide. Use a hydrophobic slide (Fisher Probe-On Plus) to retain the water drop. c. MW slides for 30 min at 43°C. 7. Staining sections: Time: -1.5 h. a. Deparaffinize slides in xylene 2 × 10 min. Hydrate sections to 70% EtOH. b. Place slides in staining dish, and cover with Johansen's Safranin-O solution. Place staining dish in a water bath. Loosely cover staining dish with plastic wrap to prevent spattering. Insert temp probe, through the plastic wrap, into the staining dish. MW at 60°C for 40 min. c. Dehydrate slides for 5 s in 95% EtOH, with 0.5% picric acid. d. Wash slides for 5 s in 95% EtOH with ammonium hyroxide (4 drops/ 100 mL EtOH). e. Dehydrate slides for 5 s in 100% EtOH. f. Counterstain for 10-15 s in Fast Green FCF staining solution. g. Wash for 5 s in used Fast Green clearing solution. Wash 5 s in fast green cleating solution. h. Clear in xylene 2x for 10 min, keep slides in xylene, until mounting cover slips. ~ ~ : RESULTS Typical examples of MW and traditionally processed Zea mays shoot apices are shown for comparison in Fig. 1. MW fixation and embedding required 5 h; the traditional protocol took 9 d. The overall quality of the tissue prepared using the MW is comparable to the tissue prepared through the traditional protocol. The overall morphology of the sample is preserved, as well as the internal anatomy. There are no indicators of poor fixation and infiltration, such as holes, tips, or tears in the samples. However, when looking at the highly magnified region of the meristem in Fig. 1B-D, the difference in quality becomes apparent. The meristem is a delicate, densely cytoplasmic structure that is easily damaged, and therefore a good indicator of the quality of tissue preservation. The two outer layers of the MW-prepared specimen (Fig. 1B) are plump and intact. The cytoplasm has not shrunken away from the cell wall, and mitotic figures (arrowhead) are found frequently. The sample processed traditionally (Fig. 1D) is not as wellpreserved. The two outermost layers of cells are shrunken and damaged. The nuclei of these cells are large and not well shaped, and it is difficult to find any mitotic figures in this general area.
M W Paraffin Techniques for Plants
185
Fig. 1. Examples of Zea mays shoot apical meristem and surrounding leaf primordia in longitudinal section. The sample shown in (A) was fixed, dehydrated, and embedded using the MW protocol. A detailed view of the same sample is shown in (B). The arrowhead points to a mitotic figure. The sample shown in (C) was processed using traditional methods. A detailed view of the same sample is shown in (D). All samples were stained using Johansen's Safranin-O and Fast Green FCF protocol. Sample A was stained using the MW protocol outlined in this chapter; sample B was stained conventionally. Bar = 50 ~tm.
186
Schichnes, Nemson, and Ruzin
Fig. 2. Examples of Zea mays shoot apical meristem and surrounding leaf primordia in longitudinal section. In situ localization experiments were performed using a DIG-labeled antisense probe, which hybridized to knotted (Smith and Hake, 1992). Sample shown in (A) was fixed, dehydrated, and embedded using the MW protocol. (B) was processed for in situ analysis using traditional methods (Jackson, 1991). Bar = 100 gm. Figure 2 compares Zea mays shoot apices processed for ISH studies using the MW (Fig. 2A) and traditional (Fig. 2B) protocols. The MW protocol required 5 h; the traditional protocol required 7 d. The traditional protocol for in situ studies is shorter than the protocol for standard anatomical preservation, to minimize the time in which mRNA degrades (Jackson, 1991). Because of the short exposure to fixative, the overall anatomical preservation of the tissue is poor. Notice that the shoot apical meristem in Fig. 2B is sunken and not dome-shaped, as it is in its native state. The outer cell layers of the meristem (the 11 and 12 layers) are completely crushed. There is a large tear in the tissue, indicative of poor infiltration and embedding. This example is typical of the quality of samples that are embedded according to the traditional protocol. In contrast, the MW-prepared sample (Fig. 2A) is of excellent quality. The apical dome is well-preserved, and there are no tears in the tissue. The staining is darker, indicating better preservation of the mRNA, and the pattern is more tightly localized than in traditional preparations. Color figures could not be included in this publication, but refer to Schichnes et al. (1999) for a detailed analysis of in situ localization results using MW techniques.
MW Paraffin Techniques for Plants
187
w
Fig. 3.Arabidopsis thalianarosette leaves shown in transverse section. The sample in (A) has been processed according to the MW protocol; (B) has been processed according to traditional protocols. The samples were stained using Johansen's Safranin-O and Fast Green FCF MW modified protocol. Bar = 100 }.tm.
The final figure, Fig. 3, details results from a variation of the authors' M W protocol, to accommodate delicate tissues, such as Arabidopsis thaliana and friable callus. Traditionally, these tissues are fixed with a high concentration of glutaraldehyde and are embedded in resin or meth-
188
Schichnes, Nemson, and Ruzin
acrylate for sectioning. Paraffin embedding nearly always yields poor results with these tissue types (Ruzin, 1999). To accommodate delicate tissue, two major changes were made to the MW technique. The authors used 10 mM SCrensen's buffer (no chemical fixative) and MWs to stabilize these delicate samples. The power of the oven was adjusted to 450 W (normally, 650 W). Figure 3A shows an Arabidopsis leaf in transverse section, after having been processed using the alternative MW method. The leaf internal structure is well-preserved, the vascular bundles are well-defined, and the phloem (delicate and easily crushed) is visible. The epidermal cells are expanded and rounded, as in the native state. Stomatal complexes, including air spaces in the mesophyll, are open and not crushed. Finally, the chloroplasts are present and intact. Although the leaf in Fig. 3B was the same size and shape as the leaf in Fig. 3A before fixation and paraffin processing, it shrank during the procedure. The epidermal layer is crushed and not distinguishable from the other cell layers. It is not possible to distinguish stomatal complexes or mesophyll air spaces, and it is difficult to find the vascular bundles. Of the large midvein that is discernable, only the xylem elements, which are secondarily supported by lignin, and therefore some of the strongest parts of the leaf, are visible. DISCUSSION The MW samples are superior to theft traditional counterparts, when comparing corresponding indicator regions in the sample, such as the outer cell layers of the meristem. The samples prepared for ISH reflect this, as well. Not only are MW-processed in situ samples better-preserved, but they show a stronger and more tightly localized signal pattern than their traditionally prepared counterparts. MW and traditional mounting and staining are comparable, with less time required for the MW procedure. Another important benefit of the MW technology is the ability to embed delicate botanical tissues, such as Arabidopsis and friable callus, in paraffin. Paraffin microtechnique is less expensive, less technically demanding, and generates less toxic waste than resin embedding. Before the variable-wattage option on the MW, it was not possible to adequately prepare these tissues in paraffin, either traditionally or in the MW. The authors have placed the tissue in an isotonic solution, and used MW energy to stabilize the tissue, which has proved effective for these delicate botanical tissues.
M W Paraffin Techniques for Plants
189
Quick processing times do not mean poor sample quality. Nothing was sacrificed by using the MW to process botanical tissues. We have gained increased tissue quality and the ability to process delicate samples that were previously only available through the time and expense of resin embedding. ACKNOWLEDGMENTS This technique was developed in the College of Natural Resources Biological Imaging Facility. The authors dedicate this work to the memory of Wally Porter. REFERENCES Berlyn G, Miksche J (1976) Botanical Microtechnique and Cytochemistry. The Iowa State University Press, Ames, Iowa. Freeling M, Lane B (1994) The maize leaf. In: Freeling M, Walbot V, eds, The Maize Handbook. Springer-Verlag, New York, pp. 17-28. Giberson RT, Demaree RS Jr., Nordhausen RW (1997) Four-hour processing of clinical/ diagnostic specimens for electron microscopy using microwave technique. J Vet Diagn Invest 9:61-67. Haupt A (1930) A gelatin fixative for paraffin sections. Stain Technol 5:97-98. Jackson D (1992) In situ hybridisation in plants. In Bowles DJ, Gurr SJ, McPherson, M, eds, Molecular Plant Pathology: A Practical Approach. Oxford University Press, New York, pp. 163-174. Johansen DA (1940) Plant Microtechnique. McGraw-Hill, New York. Kok L, Boon M (1990) Microwaves for microscopy. J Microsc 158:291-322. Kouchi H, Hata S (1993) Isolation and characterization of novel nodulin cDNAs representing genes expressed at early stages of soybean nodule development. Mol Gen Genet 238:106-119. Login G, Dvorak A (1994) Methods of microwave fixation for microscopy. Progr Histochem Cytochem 27:72-94. Ruzin SE (1999) Plant Microtechnique and Microscopy. Oxford University Press, Cambridge. Sass J (1958) Botanical Microtechniqe. The Iowa State University Press, Ames, IA. Schichnes D, Nemson J, Ruzin, SE (1999) Microwave protocols for paraffin microtechnique and in situ localization in plants. Microsc Microanalysis 4:491-496. Smith LG, Hake S (1992) Initiation and determination of leaves. Plant Cell 4:1017-1027. SCrensen, S (1909). Enzymstudien. II. Mitteilung. Uber die messung und die Bedeutung der Wasserstoffionen-kunzentration bei enzymatischen Progressen. Biochem Z 21:131-200.
16
Microwave-Assisted Formalin Fixation of Fresh Tissue
A ComparativeStudy Richard T. Giberson and Douglas E. Elliott
INTRODUCTION Over the past two decades, there has been a growing interest in reducing the turnaround time required to preserve and process specimens into paraffin blocks for sectioning, and subsequent evaluation by light microscopy (LM) for surgical pathology (PATH) or research (Login, 1978; Leong, 1991, 1993, 1994; Hopwood, 1993; Boon et al., 1986; Kok and Boon, 1990a; Battifora, 1999; Ruijter et al., 1997; Gamble, 1998). Tissue samples are typically placed in plastic cassettes for all processing steps. Equipment has been on the market for a number of years to facilitate rapid processing. Processing, after sample preservation, is done either by automatic tissue processors (Leong, 1991, 1993, 1994; Hopwood, 1993; Gamble, 1998; Battifora, 1999) or by microwave (MW)-assisted processing (Boon et al., 1986; Kok and Boon, 1992, 1996, 1997; Crowder and Giberson, 1998). Either process cycles the tissue cassettes, after preservation, through dehydration (usually a graded series of ethanol), intermedium (usually xylene, isopropanol, chloroform, or other xylene substitute), and molten paraffin. MW-assisted methods are generally considered to result in quicker turnaround times (Boon et al., 1986; Kok and Boon, 1992; Crowder and Giberson, 1998) than those obtained with automated processors (Leong, 1991, 1994; Kok and Boon, 1992). Formalin and aldehyde fixation, for either LM or transmission electron microscopy, has not benefited from automation as a method to From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. Demaree Jr. © Humana Press Inc., Totowa, NJ 191
192
Giberson and Elliott
speed up the process. The only piece of equipment that the authors are aware of, that has been exploited to accelerate this step, is the MW (Login, 1978; Leong, 1991, 1993, 1994; Hopwood, 1993; Login and Dvorak, 1993; Giberson and Demaree, 1995, 1999; Giberson et al., 1997; Crowder and Giberson, 1998). However, the primary application of the MW has been to accelerate the fixation of tissues for electron microscopy (Login and Dvorak, 1993; Giberson and Demaree, 1995, 1999, Giberson et al., 1997). For the sake of clarity, when MW radiation is used to accelerate tissue preservation in a solution, there are essentially two strategies: final solution temperatures (temps) >45°C (Leong, 1991, 1993, 1994; Boon and Kok, 1994; Kok and Boon, 1997; Login and Dvorak 1993), and final solution temps <40°C (Giberson and Demaree, 1995, 1999; Giberson et al., 1997). Tissue preservation is, by default, the single most important step, because it precedes all subsequent processing, and is responsible for stabilization or fixation of proteins and other cellular constituents against the rigors of subsequent processing steps. Without adequate preservation, the evaluation of the results, by staining, immunohistochemistry and/or morphologic detail, can be compromised and/or misinterpreted. Formalin remains the most popular fixative for LM and PATH (Leong, 1991, 1993, 1994; Battifora, 1999; Ruijter et al., 1997), and defines the morphological basis for diagnostic PATH (Leong, 1993; Battifora, 1999; Boon and Kok, 1994; Kok and Boon, 1992). MW stabilization (heating tissue in normal saline to approx 60°C in the MW) is a method to circumvent the use of formalin in the laboratory (Leong, 1991, 1993, 1994). Formalin, in the hydrated form of methylene glycol, requires 4-6 h just to penetrate tissue blocks up to 5-mm-thick, and a minimum of 24 h for enough formaldehyde (HCHO) to be created to fix a tissue biopsy (Leong, 1993; Kok and Boon, 1990a, 1997; Battifora, 1999; Helander, 1994). The most accepted method of MW-assisted formalin fixation is to let the samples (thickness >1 mm) soak in formalin (i.e., 10 % neutral buffered formalin (NBF) -4% HCHO, and ~ 1% methanol in 0.1 M phosphate buffer) for at least 4 h prior to microwave irradiation (MWI). The samples, in formalin, are irradiated at high power to a final temp of 55°C (Kok and Boon, 1990a, 1992, 1997). An intriguing technique to get around the 4-h soak is to inject a large surgical specimen with formalin, prior to MWI (Ruijter et al., 1997). Successful MW-assisted formalin fixation of fresh tissue for surgical PATH has not been reported (Leong, 1991, 1993, 1994; Kok and Boon, 1990a, 1997; Battifora, 1999; Lemire, 2000). This chapter outlines, for the first time, a methodology for rapid MW-assisted formalin fixation of fresh tissue. Four tissue types were cho-
MW-Assisted Formalin Fixation
193
sen for this study. Two processing groups of tissue were fixed on the bench for 24 h in 10% NBF, and two groups were fixed in 10% NBF in the MW. Subsequent processing of samples from each fixation group was done in either a Sakura VIP Processor (groups B and D), or by MW-assisted methods (groups A and C). The 24-h formalin-fixed tissue, which was transferred to the automatic tissue processor for subsequent processing into paraffin (group D), served as the standard for all comparisons. The role of temp and wattage during MW-assisted formalin fixation was also examined. MATERIALS A N D M E T H O D S A Model 3451 Microwave Processor with vacuum chamber (cat. no. 3435, from Ted Pella, Redding, CA) was used for all MW-assisted processing steps. The MW had the following features: water-load recirculation and cooling inside the MW cavity, using a load cooler; adjustable continuous power output from 250 to 750 W, in 100-W increments; temperature-restriction (TR) temp probe, set to control solution temp maximums (maxs) during MW-assisted processing steps; and MW-compatible vacuum chamber for use during the paraffin infiltration step. Prior to beginning MW-assisted processing (see the next two subheadings), two 600-mL plastic beakers were filled with 500 mL tap water each. One was placed in the left front (~4 in. back from the front ) of the MW cavity, and the other was placed in the right rear (--6 in. from back wall of the MW). Enough space was left between the two beakers to comfortably place a container holding the cassettes. The water in the left front beaker was recirculated and cooled to 30°C by the load cooler during processing. Plastic histology cassettes were used for all processing steps for all trials and groups.
Effect of Temp and MW Power (Wattage) on MW.Assisted Formalin Fixation Fresh beef kidney or liver was used to examine the effects of MW wattage and final reagent temps on MW-assisted formalin fixation and MW stabilization. Tissue slices, from 3.0 to 5.0 mm thick, were placed in plastic histology cassettes and MW irradiated in a 325-mL Rubbermaid TM container, containing either 0.9% saline or 10% NBF. The TR temp probe was set to control max reagent temps, between 40 and 60°C, for the different experimental trials. The influence of temp on MW stabilization in 0.9% saline and MW-assisted formalin fixation was evaluated. Next, the effects of low wattage only, high wattage only,
194
Giberson and Elliott
and a combination of low-to-high wattage settings were examined for MW-assisted formalin fixation. The experimental design for each trial is detailed in Table 1. Cassettes were processed in a Rubbermaid 325 mL container inside the MW. Three runs were done for each trial. An evaluation of the results from all the trials was used to determine what MW and temp parameters, if any, would make direct MW-assisted formalin fixation of fresh tissue possible (see next subheading).
Four Processing Methods for Evaluation of MW-Assisted Formalin Fixation Four processing groups were designated: Group A (MW fix, MW process), Group B (MW fix, automatic processor process), Group C (24-h bench fix, MW process) and Group D (24-h bench fix, automatic processor process) (see Table 2). Groups A-D were processed, using the following reagents: 10% NBF as the fixative; absolute ethanol for dehydration; xylene for the intermedium; Fisher Tissue Prep 2 for paraffin infiltration. Chicken liver, lung, kidney, and intestine were processed for each group. After harvest, whole organs were placed in cold (--4°C) pH 7.0, 0.2 M Sorensen' s phosphate buffer, prior to processing by either routine or MW methods. The organs were removed from the buffer, cut into _< 2-mm-thick slices, and placed in cassettes, to begin fixation by bench or MW-assisted methods. Cassettes were placed in a Rubbermaid 470-mL container for MW-assisted processing. Table 3 describes in detail the four processing methods used in this study. Groups A (MW fix) and C (bench fix) were processed into paraffin, using the MW processor. Groups B (MW fix) and D (bench fix) were processed into paraffin, using a Sakura VIP Processor set on a 12-h cycle. The tissue was placed in the processor, at the beginning of the ethanol dehydration step. The reagents, times, TRs, and power levels for MW processing are indicated in Table 3, as well as the times and reagents used in the Sakura VIP Processor. The noncirculated water load (right rear) was removed from the MW, for the vacuum-assisted paraffin infiltration step. The recirculated water load was repositioned to the center rear of the MW, to accommodate the vacuum chamber. Group D (bench fix to VIP Processor) was considered the standard of comparison for the four processing groups. Paraffin sections, 2-3-~tm thick, were cut, then stained with hematoxylin and eosin (H&E), using a Sakura DRS 601 automatic stainer. Slides were cover-slipped and examined by LM.
MW-Assisted Formalin Fixation
195 Table 1
Trials Used to Evaluate Effects of Temperature and Wattage on M W Stabilization and MW-Assisted Formalin Fixation Trial no. and description 1. Temp Effect 2. Temp 3. Temp 4. Temp 5. Temp 6. Temp 7. Temp 8. Wattage Effect (Low Only)/) 9. Wattage Effect (High Only) c 10. Wattage Effect (Low to High) d
MW Temp probe setpoint °C
Total time a (min)
MW irradiation time (min)
Reagent
MW wattage setting (W)
40 55 40 55 40 50 60 40
10 10 5 5 15 15 1 40
10 10 5 5 10 10 10 30
0.9% Saline 0.9% Saline 10% NBF 10% NBF 10% NBF 10% NBF 10% NBF 10% NBF
750 750 750 750 650 650 650 <450
40
30
30
10% NBF
>650
40
30
30
10% NBF
350-650
aTotal time equals the total time the fresh tissue was in the reagent, in or out of the MW. /)"Low only" refers to a continuous power output setting during MWI of 450 W or less. C"High only" refers to a continuous power output setting during MWI of 650-750 W. d"Low to high" refers to a continuous power output setting during MWI of <450 W for 30 min, followed by 10 min at 650 W.
Table 2 Processing Groups Used to Compare MW-Assisted Formalin Fixation to 24-h (Bench) Fixation in Formalin Processing method MW-assisted formalin fixation MW-assisted processing into paraffin MW-assisted formalin fixation Automatic tissue processor into paraffin 24-h bench fixation in formalin MW-assisted processing into paraffin 24-h bench fixation in formalin Automatic tissue processor into paraffin
Group A
Group B
Group C
Group D a
X X
aStandard for comparison of the different processing groups.
X X
196
Giberson and Elliott
Table 3 Processing Steps for MW-Assisted and Standard Methods of Tissue Processing into Paraffin
Step
MW-assisted
Standard processing
Fixative Fixation: Steps: reagent not changed between steps Time each step (mins): MW power setting: MW TR setting: Total time (min): groups:
10% Neutral buffered formalin First Second 30 350 W 40°C 40
10 650 W 40°C Groups A and B
Dehydration Solvent: Steps: reagent changed at each step Time each step (min): MW power setting: MW TR setting: Total time (min): groups:
Absolute ethanol 70% 90% 3 × 100%
10% Neutral buffered formalin
Room temp 1440 Groups C and D Tissue to Sakura VIP Processor Absolute ethanol 70% 80% 2 x 9 5 % 2x100%
5 5 3×5 650 W 650 W 650 W 40°C 40°C 40°C 25 Groups A and C
45
45
300
Groups B and D
Xylene 100% a 100% b
Xylene 100% 100%
10
2 x 45
2 × 60
Intermedium
Solvent: Steps: reagent changed at each step Time each step (min): MW power setting: MW TR setting: Total time (min): groups:
60
60
450/550 W 650 W 45°C 55/65°C 20 Groups A and C
120
Groups B andD
Fisher tissue prep 2 100% c 100% c 100% c
Fisher tissue prep 2 100% 100% 100%
10 10 10 650W 6650W 650W 65°C 65°C 65°C 30 Groups A and C
60
60
180
Groups B and D
10
Paraffin infiltration Paraffin: Steps: reagent changed at each step Time each step (rain): MW power setting: MWTR setting: Total time (min): groups: Total processing time (min/h)
115/1.92
60
2040/34
aFirst 5 min of the step is at 450 W; the second 5 min at 550 W. bFirst 5 min of the step, the TR is at 55°C • the second 5 min, it is 65°C. CVacuum (20" of H g - 50.8 torr) is used inthe following cycle for each step: 2 min, on, 2 min, off, 2 min, on, 4 min, off.
MW-Assisted Formalin Fixation
197 RESULTS
Effect of Temp and Wattage on Tissue Preservation A summary of the results from the runs for trials 1-10 can be found in Table 4. Trials 1-7 evaluated the effects of high-wattage settings (650 or 750 W) and different temp max on the visual appearance of the processed tissue. The visual appearance of the tissue from trials 1-4 are shown in Fig. 1A,B and those from trials 5-7, in Fig. 2-C. Tissue processed at 40°C, in either 0.9% saline or 10% NBF, exhibits a rawmeat appearance, when contrasted to tissues processed at temps 50°C or higher (trials 2, 4, 6, 7). The results from trials 8-10 are summarized in Fig. 3A,B. Low wattage alone (<450 W) (Fig. 3A) clearly demonstrates a raw-tissue appearance in the center. Tissue in the MW for the same time as in Fig. 3A, but irradiated from 350 to 650W (Fig. 3B), appears fixed throughout. The results of the high only wattage effect (>650 W, trial 9) are not shown. The tissue, however, was fixed only at the periphery, and the center was raw in appearance. The results from the runs for trial 9 also produced a detectable outer layer or skin on the tissue sample. Regardless of the duration of MWI at high wattage, the penetration of the fixative (10% NBF) was always limited to the outer 0.75 mm of the tissue, when the temp was maintained at 40°C (results not shown).
Comparison: MW.Assisted Formalin Fixation (Low to High Wattage) to 24-H Formalin Fixation, by Four Processing Methods The results from groups A-D indicate that the four tissues from each group (Figs. 4A-D to 7A-D) could be successfully processed by any of the four methods. The contrast of the H&E staining is different for each of the groups. Groups A and C generally have the most contrast for each of the tissues processed, and Groups B and D the least. The former were processed by MW-assisted techniques, after the fixation step, and the latter were processed, in the Sakura automatic tissue processor after the fixation step. Each grouping had one tissue fixed in the MW and one on the bench. In Fig. 5, (MW-fixed kidney) (Fig. 5A,B) demonstrate the least amount of shrinkage between Bowman's capsule and the underlying glomerulus, compared to the 24-h bench-fixed samples (Fig. 5C,D). Group A, which underwent complete MW-assisted processing, required, in actual time, just over 2 h to have paraffin-embedded tissue ready to section. Contrast that to group D, the standard for comparison
i,,,,,a
10
Trial
is raw in appearance appears stabilized throughout appears raw except at edge appears raw in the center appears raw in appearance appears mostly fixed throughout appears fixed throughout appears raw in the center
Temp effect/40°C/0.9% Saline/10 min MW Temp effect/55°C/0.9% Saline/10 min MW Temp effect/40°C/10% NBF/5 min MW Temp effect/55°C/10% NBF/5 min MW Temp effect/40°C/10% NBF/10 min MW Temp effect/50°C/10% NBF/10 min MW Temp effect/60°C/10% NBF/10 min MW Wattage Effect (low only): 40°C/10% NBF/30 min MW + 10 min = 40 min total time Wattage effect (high only): 40°C/10% NBF/30 min MW Wattage effect (low to high): 40°C/10% NBF/30 min MW
Tissue appears fixed throughout
Results not shown
Results
Description Tissue Tissue Tissue Tissue Tissue Tissue Tissue Tissue
Table 4 Visual Evaluation of M W Processing Results from Trials 1-10
3B
1A 1A 1B 1B 2A 2B 2C 3A
Fig.
MW-Assisted Formalin Fixation
199
Fig. 1. (A) M W stabilization at 750 W oftissue in a 0.9% saline solution for 10 min at two different TRs. The tissue irradiated at a TR of 40°C still appears raw when contrasted to the tissue irradiated at a TR of 55°C, which appears stabilized throughout. Scale divisions = 0.5 mm. (B) M W I at 750 W of tissue in 10% NBF for 10 min at two different TRs. The tissue irradiated at a TR of 40°C still appears raw in the center, when contrasted to the tissue irradiated at a TR of 55°C, which appears to be fixed throughout. Scale divisions = 0.5 mm.
200
Giberson and EUiott
Fig. 2. (A) M W I at 650 W of tissue in 10% NBF for 10 min at a TR of 40°C. The tissue was left in 10% NBF for an additional 5 min in warm fixative, after MWI. Note the bloody appearance of the tissue and the thin outer layer, approx 0.5 mm, that appears fixed. Scale divisions = 0.5 mm. (B) MWI at 650 W of tissue in 10% NBF for 10 min at a TR of 50°C. The tissue was left in 10% NBF for an additional 5 min in warm fixative after MWI. Note the tissue appears fixed throughout on the left half. Scale divisions = 0.5 mm. (C) M W I at 650 W of tissue in 10% NBF for 10 min at a TR of 60°C. The tissue was left in 10% NBF for an additional 5 min in warm fixative, after MWI. Note that the tissue appears fixed throughout and darkened at the right edge. Scale divisions = 0.5 mm.
MW-Assisted Formalin Fixation
201
Fig. 3. (A) MWI at 350 W (10 min), followed by irradiation at 450 W (20 min) of tissue in 10% NBF, at a TR of 40°C. Note the raw appearance of the tissue in the center and the outer layer, approx 1.0mm, that appears fixed. Scale divisions = 0.5 mm. (B) Sequential MWI steps of 350, 450, and 650 W, for 10 min each, of tissue in 10% NBF, at a TR of 40°C. Note the upper (2-mm thick) piece appears completely fixed, and the lower (3-mm-thick) piece appears fixed, as well. Scale divisions = 0.5 mm. with no M W component, which required just over 34 h to have paraffinembedded tissue ready to section. Group B ( M W fix, automatic processor process) required approx 11 h, and group C (24-h bench fix, M W process) required approx 26 h to have embedded tissue ready to section.
~ill
'
~:~
Fig. 4. (A) Results of the four processing groups for chicken intestine. Group A (A) and group B (B) were MW-fixed, and group C (C) and group D (D) underwent a 24-h bench fix in 10% NBF. Groups A and C were processed, after fixation, by MW-assisted methods (see Table 3), and groups B and D were processed, after fixation, in the Sakura VIP Processor.
Fig. 5. Results of the four processing groups for chicken kidney. Group A (A) and group B (B) were MW-fixed, and group C (C) and group D (D) underwent a 24-h bench fix in 10% NBF. Groups A and C were processed, after fixation, by MW-assisted methods (see Table 3), and groups B and D were processed, after fixation, in the Sakura VIP Processor. 202
Fig. 6. Results of the four processing groups for chicken lung. Group A (A) and group B (B) were MW-fixed, and group C (C) and group D (D) underwent a 24-h bench fix in 10% NBF. Groups A and C were processed, after fixation, by MWassisted methods (see Table 3), and groups B and D were processed, after fixation, in the Sakura VIP Processor.
i c ...
!
~!iD
Fig. 7. Results of the four processing groups for chicken liver. Group A (A) and group B (B) were MW-fixed, and group C (C) and group D (D) underwent a 24-h bench fix in 10% NBF. Groups A and C were processed, after fixation, by MW-assisted methods (see Table 3), and groups B and D were processed, after fixation, in the Sakura VIP Processor. 203
204
Giberson and Elliott
When the slides from groups A-D (Fig. 4A-D to 7A-D) were evaluated by pathologists at the University of California, Davis, the overall processing for each group was judged adequate for use in diagnostic evaluation. DISCUSSION The effect of temp alone on MW-assisted formalin fixation and stabilization is evident from an examination of Fig. 1A,B and 2A-C. The results from the first seven trials indicated that reagent temps (10% NBF or 0.9% saline) of 50°C or higher, for 15 min or less, would result in the tissues having a distinct color and texture change from that of the fresh tissue (Figs. 1A,B and 2A-C). When the temps during processing were maintained at 40°C, and the MW power output was set at 650 or 750 W, only the presence of 10% NBF produced a distinct color and texture change at the periphery of the tissue, after MW exposure (trials 1-5 and Figs. 1B and 2A). Max and/or optimum solution temps (usually for 0.9% saline) for MW stabilization have been reported to between 45-55°C (Boon and Kok, 1994) and 50-70°C (Leong, 1993, 1994). MW-assisted formalin fixation has been reported for tissue no thicker than 5 mm, which had been immersed in formalin for at least 4 h prior to MW exposure, at wattages above 600 and final fixative temps of 55°C (Kok and Boon, 1992, 1997). All attempts made to uniformly fix tissue at 650 or 750 W in 10% NBF at 40°C, in the MW (trial 9), resulted in the outer periphery of the tissue becoming well-fixed, while the center remained raw in appearance. This result agreed with that of other authors (Kok and Boon, 1992, 1997). The influence of MW energy during processing is thought to be an acceleration of the diffusion process caused by MW heating (Boon et al., 1986; Kok and Boon, 1990a, 1990b, 1997). When formalin is heated by MW (Kok and Boon 1990b, 1994, 1997) or other methods (Erlich and Lazarus, 1898), HCHO is formed: HCHO is the fixative component (Fox et al., 1985) that crosslinks to tissue proteins (Fox et al., 1985; Helander, 1994). Trials 8 and 9 (low wattage only/high wattage only), in which 10% NBF temps were maintained at 40°C in the MW for 30 min, demonstrated only peripheral fixation of the tissue. These results could suggest that MW heating, at temps as low as 40°C, can create the fixative component, HCHO. The time of these trials was sufficient for fixation, based on the results of trial 10. It is presumed that MW heating and
MW-Assisted Formalin Fixation
205
HCHO formation refers to temps of 45°C, based on the trial results and published temps used for MW stabilization (Boon and Kok, 1994, Leong, 1993, 1994). However, 55°C, as a final temp is used for MW-assisted fixation of tissues immersed in formalin for at least 4 h, prior to MW exposure (Kok and Boon, 1990a, 1997; Boon and Kok, 1994). The results of trial 10 (10 min at 350 and 450 W + 10 min at 650 W) suggested that fixation had taken place, and also implied that the formation of HCHO may be wattage-dependent, rather than temp-dependent at 40°C. The four processing groups were designed to test the results of trial 10. The use of a 24-h bench fix in 10% NBF (groups C and D) was considered the fixation benchmark. Group D was considered the overall benchmark, and adequate for 2-mm-thick tissue (Kok and Boon, 1990a, 1997; Battifora, 1999; Helander, 1994). Groups A and B were placed in 10% NBF and MW-irradiated for 30 min at 350 W, followed by 10 min at 650 W. From the authors' experience (unpublished research and this study), it is recommended that at least 75 % of the total MW time be spent at low wattage, and the remaining 25% at high wattage. When other wattage settings were employed (e.g., 550 W as either a high or low wattage), tissue quality, after processing into paraffin and H&E staining, was not consistent (unpublished research). Even though the evaluation of H&E-stained paraffin-embedded tissue is a subjective process, the results from the four processing groups suggest that it is difficult to differentiate between MW and bench-fixed tissues (Figs. 4A-D to 7A-D). Tissues from groups A and C stained with H&E, similarly. The same could be said for groups B and D. Groups A (MW fix) and C (bench fix) were processed into paraffin, using the MW, and groups B (MW fix) and D (bench fix) were processed into paraffin, using the Sakura VIP. H&E staining appears to be more dependent on tissue processing after the fixation step (see Figs. 4A-D to 7A-D). The mechanism of formalin fixation is believed to proceed sequentially in three steps (Kok and Boon, 1992, 1997): diffusion of methylene glycol into the tissue (HCHO solutions contain little HCHO; instead the primary chemical species is methylene glycol, formed by the reaction of HCHO and water) (Walker, 1944; Bumett, 1982); formation of HCHO, the fixative component (Fox et al., 1985), by the dehydration reaction of methylene glycol in the tissue; binding of HCHO to the proteins, by chemical crosslinking of the HCHO to tissue proteins and other cellular components (Helander, 1994; Fox et al., 1985). This three-step process, when done outside the MW, requires at least 24 h or longer to complete (Leong, 1993; Kok and Boon, 1990, 1997; Battifora, 1999; Helander, 1994).
206
Giberson and Elliott
Can the three steps of formalin fixation be explained within the context of the results of trials 1-9, the results of trial 10, and the comparison of processing groups A-D? The authors believe the answer is "yes." The explanation is also supported, if the following are considered, as well. First, MW ovens, household or lab, are designed to heat. The continuous power (wattage) output of these instruments is the max (usually always above 700 W). The other power settings on MWs are full power for part of the time (e.g., at 50% power, a 700-W MW will produce 700 W for 30 s out of every minute). Second, no variable-power-output MWs were commercially available from vendors in the microscopy or histology field for use in tissue processing, until just recently. Based on trials 1-10 and the processing results from groups A-D, a two-part mechanism is proposed for MW-assisted formalin fixation of fresh tissue. First, the diffusion of methylene glycol, vs the dehydration of methylene glycol to HCHO, is promoted at a temp of 40°C and MW power output of 350 W. HCHO formation and crosslinking does not become an inhibiting factor to the continued diffusion of methylene glycol into tissue slices, under these conditions. At a MW power output of 650 W and formalin temp of 40°C, the dehydration reaction of methylene glycol to HCHO is favored. The formation of HCHO results in the crosslinking of tissue proteins and other cellular components. Fixation is adequate to withstand the rigors of processing into paraffin by either routine (automatic tissue processors) or MW methods. The reports of crust formation at the periphery of the tissue when fresh tissue, is irradiated in formalin (Kok and Boon, 1990a, 1997), agrees with the findings of trial 9, and supports part 2 above. Also, it has been reported that MW energy can influence the dehydration of aldehydes in organic chemistry (Majetich and Wheless, 1997), and that MW exposure, in the presence of aldehydes, will cause crosslinking (Hopwood, 1988). The visual evidence that trial 8 tissue (wattage <450) was raw in the center, vs trial 9, which had a crust on the periphery, supports part 1 above. The presence of erythrocytolysis is thought to be a good measure of the degree of formalin fixation (Kok and Boon, 1990a). In all the tissue samples from each of the processing groups, there was no evidence of erythrocytolysis. This finding supports adequate formalin fixation in all groups.
MW-Assisted Formalin Fixation
207
To the authors' knowledge, the use of low, continuous wattage during formalin fixation has not been tried until now, because the equipment was unavailable. The influence of temp in the dehydration reaction of methylene glycol should be re-examined. The results from trials 8-10 suggest that MW power, rather than MW heating, may be a more important component when temps are maintained at 40°C. Formalin temps less than 40°C (e.g., 35 and 30°C) have been attempted for MW-assisted fixation. The results were similar to those seen in this study (unpublished research). The results of this study indicate that MW-assisted formalin fixation of fresh tissue will proceed when maximum fixative tempes of 40°C are maintained during MWI, and the MW power remains at 350 W for the first 75% of the time. It is then elevated to at least 650 W for the remaining 25 %. When 2-mm-thick tissues are fixed by this method, the results are comparable to routine methods, and result in a 97% time savings, compared to 24-h formalin fixation. REFERENCES Battifora H (1999) Quality assurance issues in immunohistochemistry. J Histotechno 22:169-175. Boon ME, Kok LP, Ouwerkerk-Noordan E (1986) Microwave-stimulated diffusion for fast processing of tissue: reduced dehydrating, clearing and impregnating times. Histopathology 10:303-309. Boon ME, Kok LP (1994) Microwaves for immunohistochemistry. Micron 25:151-170. Burnett MG (1982) Mechanism of the formaldehyde clock reaction. J Chem Educ 59:160-162. Crowder CH, Giberson R (1998) Microwave processing: clinical laboratory methods for paraffin, special stains, decalcification and electron microscopy. National Society for Histotechnology, Salt Lake City, UT. Workshop 57. Erlich P, Lazarus A (1898) Die anamie I Abt Holder, Wien. Fox CH, Johnson FB, Whiting J, Roller PP (1985) Formaldehyde fixation. J Histochem Cytochem 33:845-853. Gamble M (1998) Guide to automating the histology laboratory. Lab Med 29:497-501. Giberson RT, Demaree RS Jr. (1995) Microwave fixation: understanding the variables to achieve rapid reproducible results. Microsc Res Tech 32:246-254. Giberson RT, Demaree RS Jr., Nordhausen RN (1997) Four-hour processing of clinical/ diagnostic specimens for electron microscopy. J Vet Diagn Invest 9:61-67. Giberson RT, Demaree RS Jr. (1999) Microwave processing techniques for electron microscopy: a four-hour protocol. In: Hajibagheri N, ed, Electron Microscopy Methods and Protocols. Humana, Totowa, NJ, pp. 145-158. Helander KG (1994) Kinetic studies of formaldehyde binding in tissue. Biotechnol Histochem 69:177-179. Hopwood D (1993) Microwaves and tissue processing. USA Micros Anal 1:23-25.
208
Giberson and Elliott
Hopwood D, Yeaman G, Milne G (1988) Differentiating the effects of microwave and heat on tissue proteins and their cross-linking by formaldehyde. Histochem J 20:341-346. Kok LP, Boon ME (1990a) Microwaves for microscopy. J Microsc 158:291-322. Kok LP, Boon ME (1990b) Physics of microwave technology in histochemistry. Histochem J 22:381-388. Kok LP, Boon ME (1992) Microwave Cookbook for Microscopists: Art and Science of Visualization. Coulomb, Leydon, The Netherlands, pp. 137-175. Kok LP, Boon ME (1997) Microwave methods for sample preparation in pathology. In: Kingston HM, Haswell SJ, eds, Microwave-Enhanced Chemistry: Fundamentals, Sample Preparation, and Applications. American Chemical Society, Washington, DC, pp. 641-654. Kok LP, Boon ME (1996) New developments of microwave technology in pathology: Combining vacuum with microwave irradiation. Cell Vision 3:224. Lemire TD (2000) Microwave irradiated canine and feline tissues: Part 1. Morphologic evaluation. J Histotechno123:113-120. Leong AS-Y (1991) Microwave fixation and rapid processing in a large throughput histopathology laboratory. Pathology 23:271-273. Leong AS-Y (1993) Microwave techniques for diagnostic laboratories. Scanning 15:88-98. Leong AS-Y (1994) Microwave technology for morphological analysis. Cell Vision 1:278-288. Login GR (1978) Microwave fixation versus formalin fixation of surgical and autopsy tissue. Am J Med Techno144:435-437. Login GR, Dvorak AM (1993) Reviw of rapid microwave fixation technology: Its expanding niche in morphologic studies. Scanning 15:58-66. Majetich J, Wheless K (1997) Microwave heating in organic chemistry: an update. In: Kingston HM, Haswell SJ, eds, Microwave-Enhanced Chemistry: Fundamentals, Sample Preparation, and Applications. American Chemical Society, Washington, DC, pp. 481-482. Ruijter ET, Miller GJ, Aalders TW, van de Kaa CA, Schalken JA, Debruyne FM, Boon ME (1997) Rapid microwave-stimulated fixation of entire prostatectomy specimens. Biomed-II MPC study group. J Patho1183:369-375. Walker JF (1944) Formaldehyde. Reinhold, New York.
i
1 7 Microwave-, sisted,ocessing of Biological Samples for Scanning Electron Microscopy Richard S. Demaree, Jr.
INTRODUCTION Microwave (MW)-assisted processing of biological (biol) samples for transmission electron microscopy (TEM) has been actively studied, beginning with the pioneering work by Bernard (1974). The benefits of rapid MW-assisted processing, with reduction in chemical volumes, and hence wastes, are beginning to be widely accepted (Leong, 1994). The author's laboratory has played a role in the development of MWassisted techniques for TEM, since the mid-1990s (Giberson and Demaree, 1995; Demaree, Giberson, and Smith, 1995; Giberson, Demaree, and Nordhausen, 1997; Giberson and Demaree, 1999), and now utilizes MW-assisted biol-tissue processing for TEM, on a routine basis, for research, as well as for graduate and undergraduate class projects. Recently, this research was extended to include MW-assisted preparations of biol tissues for scanning electron microscopy (SEM) (Fox and Demaree, 1999). These also are now routinely incorporated into research and class projects. MATERIALS AND M E T H O D S Protocol Outline
1. Find cold spot in MW oven, using neon bulb array. 2. Cool primary fixative to 10-15°C. From: Microwave Techniques and Protocols Edited by: R. T. Giberson and R. S. DemareeJr. © HumanaPress Inc., Totowa, NJ 209
210
Demaree
3. Place small samples in 1.7-mL polypropylene (pp) microcentrifuge tubes (larger samples in 60 x 15 mm PP Petri dishes). 4. Primary fixation: 2.5% buffered glutaraldehyde (GA), 40 s at 100% power, temperature restriction (TR) at 45°C. 5. Temperature (temp) change (start temp to final temp) should be at least + 10°C. 6. Buffer rinse: 3 × 5 rain outside MW oven at room temp. 7. Optional secondary fixation: 1-2% aqueous or buffered osmium tetroxide (OsO4), 40 s at 100% power, TR to 45°C. 8. Dehydration in either ethanol or acetone: 50, 70, 95, 3 x 100%, 40 s each, 100% power, TR 45°C. 9. Dry: critical-point drying or hexamethyldisilazane (HXDS) 3 x 100%, 40 s at 100% power, TR 45°C, and dry in conventional over at 60°C for 15 min (after 5 min, remove excess HXDS). 10. Mount on stub, metal-coat, and view.
Detailed Protocol FIXATION Materials: 1. Neon bulb array. 2. 2.5% GA. Process: 1. Find cold spot in MW (adjust with water loads). 2. Place 0.6 mL fixative in 1.7-mL pp microcentrifuge tube. Cool to 10-15°C. 3. Completely immerse small sample(s) in fixative (3:1 [v/v] vol fixative:tissue). Place tube in cold spot, and MW for 40 s at 100% power with a TR of 45°C. Notes: Place temp probe in a blank tube. Temp change (starting to finish) must be at least 10°C. If not, repeat step 3. Tissue culture (TC) cells are best preserved in fixatives at room temp or at incubator temp. Use 0.5 % buffered GA for TC cells, for ciliates, use Parducz' fixative. If clumping is a problem, omit aldehyde, and use only OsO 4. TC cells often are best preserved with shorter MW fixation times (try 10 or 20 s), and use a reduced power level (540 W vs 725 W).
BUFFER RINSE Materials: Buffer. Process: 3 × 5 min at RT outside the MW.
MW Processing of Biol Samples for SEM
211
SECONDARYFIXATION Materials" Cooled (10-15°C) 1-2% aqu or buffered OsO 4. Process: Place sample(s) in secondary fix in a cold spot, and MW for 40 s at 100% power, with a TR of 45°C. Note: TC cells often are best preserved with shorter MW fixation times (try 10 or 20 s), and use a reduced power level (540 W vs 725 W). DEHYDRATION Materials: 50, 70, 95, 100% ethanol or acetone. Process: 1. Place sample(s) in 50% solvent in a cold spot, and MW for 40 s at 100% power, with a TR of 45°C. 2. Replace with 70%, and repeat. 3. Replace with 95%, and repeat. 4. Replace with 100%, and repeat 3x. Note: Acetone is the author' s dehydrating agent of choice, but ethanol is satisfactory for most tissues. TC cells are best dehydrated in 10-s steps, using reduced power (540 W). DRYING Material: HXDS, 60 x 15 mm pp Petri dish. Process: 1. Cover sample with 100% HXDS in a pp Petri dish, and MW in a cold spot for 40 s at 100% power, with a TR of 45°C. 2. Replace with 100% HXDS, and repeat. 3. Replace with 100% HXDS, and repeat. 4. Replace (cover) with 100% HXDS in pp Petri dish, and place in conventional drying oven at 60°C for 5 min. 5. Remove excess HXDS, and again place in oven for 10 rain. 6. When completely dry, mount on stub, metal-coat, and view. Note: Conventional critical point drying may be substituted for HXDS drying. Do not use glass Petri dishes. RESULTS Biological samples processed for SEM, using the protocol described above are presented in Figs. 1-3. The protists, Spirostomum sp. and Peranema sp. (Fig. 1A), were fixed in Parducz' fixative (saturated mercuric chloride and OsO4), and were dehydrated using MW processing, then critical point dried.
212
Demaree
Fig. 1. (A) Spirostomum sp. (large cell) and Peranema (small cell). Parducz' fixative. Critical point drying. Bar = 50 ~tm. (B) MODE-K (murine duodenal epithelial) cell with Salmonella typhimuium infection. GA. HXDS dry. Bar = 5 ~tm.
MW Processing of Biol Samples for SEM
213
Fig. 2. (A) Human lymphocyte. GA. HXDS dry. Bar = 3.75 gm. (B) Same cell. Bar = 300 nm. The TC cells were fixed in either 0.5% GA (Fig. 1B) or 2.0% GA (Figs. 2A,B and 3A,B), then dehydrated using MW protocols. OsO4 has also been successfully used as the only fixative in MW processing (not illustrated here). HXDS drying, using the author's MW protocol (Figs. 1B and 2A,B), was as successful as critical point drying (Figs. 1A and 3A,B), but
214
Demaree
Fig. 3. (A) Human T-4 lymphocyte. GA. Critical point dry. Bar = 3.75 gm. (B) Same cell. Bar = 300 nm. needed n o C O 2 tank, no critical-point drier, and was slightly faster. With magnifications as high as x100,000, using field emission SEM (Figs. 2B and 3B), both drying methods appear to be artifact-free. DISCUSSION Although MW processing for TEM has received considerable attention (see Login and Dvorak, 1993; Giberson and Demaree, 1999),
M W Processing of Biol Samples for SEM
215
research into MW processing for SEM has been virtually ignored. There have been scattered reports in the literature about using MW-enhanced processing for one step of SEM tissue preparation. For example, Argall and Armati (1990) MW primary fixed-cell cultures in saline alone, followed by conventional secondary fixation in OsO4 and subsequent processing; Hotta et al. (1990) used MW maceration of frozen tissues after freeze-cracking for SEM; and Walze (1993) used MW processing in OsO4, followed by standard processing. The author et al. have reported studies on biol SEM preparations, using MW technology at every step in processing (Fox and Demaree, 1999). Now, MW processing is routinely utilized for SEM for both research and student class projects. To date, a tissue has not been found that could not be successfully processed for SEM, using the MW protocol described above. Tissues have been examined tissues using fieldemission SEM, and no artifacts could be found at magnifications up to x l00,000. One drawback to using field emission SEM is the lengthy chromium-coating process, which is longer than the entire MW processing. At this lab, MW-enhanced processing ofbiol samples for SEM is now the protocol of choice for the following reasons: It is faster than standard protocols, generates significantly less chemical waste, requires smaller amounts of reagents, yields morphology equal to standard processing, and needs no critical-point dryer. ACKNOWLEDGMENTS Appreciation is expressed to Dr. Dave Dorward at National Institutes of Health, Rocky Mountain Labs, for providing the field emission micrographs, and to Norm Fox for Fig. lB. REFERENCES Argall K, Armati P (1990) Use of microwave fixation in the preparation of cell cultures for observation with the scanning electron microscope. J Electron Microsc Tech 16:347-350. Bernard GR (1974) Microwave irradiation as a generator of heat for histological fixation. Stain Techno149:215-224. Demaree RS Jr., Giberson RT, Smith RL (1995) Routine microwave polymerization of resins for transmission electron microscopy. Scanning 17(Suppl. 5)'25-26. Fox NE, Demaree RS Jr. (1999)Quick bacterial microwave fixation technique for scanning electron microscopy. Microsc Res Tech 46"338-339. Giberson RT, Demaree RS Jr.(1995) Microwave fixation: understanding the variables to achieve rapid reproducible results. Microsc Res Tech 32"246-254.
216
Demaree
Giberson RT, Demaree RS Jr (1999) Microwave processing technique for electron microscopy: a four hour protocol. In: Hajibagheri, N, ed, Electron Microscopy Methods and Protocols. Humana, Totowa, NJ, pp. 145-158. Giberson RT, Demaree RS Jr, Nordhausen RN (1997) Four hour processing of clinical/ diagnostic specimens for electron microscopy using microwave technique. J Vet Diagn Invest 9:61-67. Hotta Y, Kato H, Watari N (1990) Simple and rapid maceration method for scanning electron microscopy using microwave. J Electron Microsc 39: 63-66. Leong S-Y (1994) Microwave technology for morphological analysis. Cell Vision 1:278-288. Login GR, Dvorak AM (1993) Review of rapid microwave fixation technology: its expanding niche in morphologic studies. Scanning 15:58-66. Walzl MG (1993) Microwave-enhanced osmium tetroxide fixation and processing of mite embryos for scanning electron microscopy. Eur J Morpho131:151-155.
Index
Bowman's capsule, 197 Brain, 76 Bulbous pemphigoid antigen, 173, 176
A Acetonitrile, 90, 99 Acetylcholinesterase, 165, 168-170 Adenocarcinoma, 68 Adenovirus, 60, 63, 64 Agar, 78 AIDS, 7 Air bubbler, 120 Air cut surface, 80 Allium, 155, 157, 159, 161,163 Amyloidosis, 68, 70, 71 Animal diseases, 49-66 Antigen preservation, 90 Antisense probe, 186 Aquaporin, 146, 151 Arabidopsis thaliana, 32, 182, 187 Aragonite, 126, 131,132 Araldite, resin, 39 Astrocytoma, 68 Autoradiography, 27, 34 Axon, 115
C Calcium chloride, 86 Calibration, microwave, 15 Callus, 182,187, 188 Capillary loops, 19 Capsicum frutescens, 32 Cell cultures, 93 Cell wall, 25, 26, 32 Chesapeake Bay, 127 Chicken, 50, 57, 62 Chlamydia, 57, 61, 63 Chlamydomonas, 19, 22 Chlorite, 124 Cilia, 43, 46, 99 Clay, 124 Clinical samples, 89 Cochlea, 109, 112, 115 Cold spot, 3, 29, 47, 53, 77, 157, 158,210 Collagen II, 117 Collagen IV, 173, 175, 178, 179 Confocal microscopy, 159, 160 Coral, 124 Critical point drying, 212, 214 Cucurbita pepo, 32 Cytochemistry, 165-171 Cytomegalovirus, 5, 7, 28, 68
B
Barley, Hordeum vulgare, 141,146, 148, 151 Basement membrane, kidney, 42 Basilar membrane, 117 Biopsies, 37 Blastoderm, 161 Bone marrow, 109, 111, 116 Botanical tissues, see plant Bouin' s fixative, 102 Bowman's capsule, 19, 22
D
Decalcification, 93, 101 Scintillation vial, 104 217
218 Decalcification times, 110 Dehydration, baskets, tissue-processing, 17 Dendrites, 76 Deparaffinize, 42, 184 Desmosomes, 70-72 Diaphragm, 166, 168-170 Diatom, 124 Dragon lizard, liver, 60, 64 Drosophila melanogaster, 141,155, 157, 161-163 Dry tortugas, 125 Duck, plague, 56, 59 Dummy tube, 38, 52 Dynein arms, 46 E
EDTA, 93, 101 ELISA, 50 Embedding, difficult specimens, 89 Embryonal carcinoma, 68 Endosperm, persimmon, 31, 32 Ependymoma, 68 Epon, resin, 39 Epoxy resin, 10 Escherichia coli, 57, 62, 64 F
Femur, mouse, 111 Field Emission SEM, 213, 214 Fixation, ice-encased, 1 immersion, 76 Fixative temperatures, 197-207 Foraminifera, 124 Formalin, 18, 53, 102
Index Formalin, fixation of flesh tissue, 191-208 Freeze substitution, 141 Freeze-drying, 27 G Gill, 51 Glomerulus, see kidney Glue preparation, 158 Glycine max, see soybean, 27 Granulocyte, 116 Green algae, Halimedia, Penicllus, Udeota, 124 Green fluorescent protein, 155, 164
H
Hair cell, 109, 113, 114 Halimeda, 126, 131, 132
Hemidesmosome, 176 Hepatocytes, see liver Herpes virus, 50, 56-59 Hexamethyldisilazane (HXDS), 210,213 Hexamita sp., 5, 8 High pressure freezing, 141 Hippocampus, rat, 59, 75-88 Hot spots, 48 Hypoxia, 76 I
Illite, 131 Immunogold, 33, 47, 90, 94, 116, 130, 173 In situ hybridization, 27, 181, 186 In vivo, 155, 163 Intestine, 94 Intestine, chicken, 202 J,K
Jejunum, turkey, 8 Kaolinite, 124
Index
Karnovsky's fixative, 73 Kidney, chicken, 202 rat, 5, 6, 19, 95, 98 stones, 95, 96 turkey, 60 Koelle technique, 165 Koi, 515,658 Kupffer cell, 57, 61 Lamellar bodies, 19 Laminin, 173, 175, 177, 179 Laryngeal trachetis, 50 Leukoencepholopathy, 69 Lipid, 5, 10, 22 Liquid crystal sheets, 91, 98 Liver, chicken, 203 lovebird, 57 rat, 19, 20, 94 Load cooler, 157, 163 Low wattage, 18, 193,207 Lowicryl, 140 LR White, 10, 33, 34, 124, 140 Lung, chicken, 203 rat, 5, 10, 19 Lymphocyte, human, 213 Lyophilize, see freeze-drying M
Magnesium Chloride, 86 Malate Synthase, 146, 151 Maraglas, resin, 39 Megakaryocyte, 109, 111 Membraneous Glomerulonephritis, 41, 44 Meningioma, 68 Meristem, 184-186 Mesothelioma, 68, 70, 71
219
Methylene Glycol, 192, 206 Microcentrifuge Tube, 210 fixation in, 17 Microfabric, 123, 134, 136 Microtubules, 31, 32 Microvilli, 70, 71 Mitotic Figures, 184, 185 Mitotic Cells, 160 Mold Release, 141 Mollusk Shells, 124 Motor endplate, 166, 168, 169 Murine duodenal cell, 212 Muscle, 42, 45, 95, 97 Myelin, 45, 97, 115 Myeloperoxidase, 116 N
Negative-stain, 51 Nematodes, 19 Nematodes, Caenorhabditis elegans, 141
Neon bulb array, 3, 15, 38, 53, 98, 102, 103,210 Neoplasm, renal, 42 Nerve, 94, 95, 97, 113, 115 Nerve, sural, 42 Net cut surface, 80, 84 Neuroendocrine carcinoma, 68 Neuropil, 83 Nucleus, 163 O,P
Oncocytoma, 42 Organ of corti, 109, 112, 117 Pacheco's disease, 56 Paraffin, animal protocol, 196 plant protocol, 181-189 Paraformaldehyde, 102 Parducz' fixative, 210 Parrot, 56, 60, 61
220
Index
Pathology, archived samples, 67-74 diagnostic, 37 Peranema sp., 211, 212 Pheochromocytoma, 68, 70, 72 Phloem, 31, 32 Pioloform grids, 81 Plant, 25, 29, 94, 181-189 Plant vacuole, 151 Plasma cell, 54, 58 Plasmolysis, 31 Plastid, 25, 26, 32, 188 Podocytes, 19 Post-staining, uranyl acetate and lead, 41 Potassium ferrocyanide, 53, 54, 83, 85, 90, 91,166 Processing protocol, decalcification, 10 l - 122 SEM, 209-216 TEM, 2, 3, 4, 91-95 Protoplast, 31, 32, 141,146, 148, 151
Service lab, 89-100 Sieve-tube, 31, 32 Silt, 124 Skin, 95, 173 Smectite, 124, 128, 131 Soybean, 26, 95, 96 Spinal cord, 76 Spirostomum sp., 211, 212 Spurr' s, resin, 34, 54, 64 Stomate, 188 Subepithelial deposits, 42 Synapse, 75, 76, 82 postsynaptic densities, 83 Syto 13, 156, 158, 160-162
Q, R Quartz, 124 Quercus suber, 182 Reduced osmium, see potassium ferrocyanide Reissner' s membrane, 109, 112 Resin, infiltration, 9 polymerization, 1, 18, 65 Rhabdoid tumor, 68
U,V Unicryl (TM), 10 Vacuole, 25, 26 Vacuum chamber, 144 Vacuum processing, animal, 13, 163 immunocytochemistry, 155-164 plant, 25-35, 163, 181-189 Variable wattage, 183 Veterinary, 49-66 Virus, 49, 50 Visual maps, 82
S
Saline stabilization, 199 Salmonella typhimurium, 212 Scanning EM, 2, 209-216 Sea urchin, Stron gylocentrotous purpuratus, 141 Sediment, 123 Serial sections, 76, 79, 82
T Temperature probe, 48, 119, 163 Thymoma, 68, 70, 72 Tonofilaments, 71, 72 Turkey, kidney, 60, 63 Type II alveolar cells, 19, 20
W-Z Water immersion, 10 Water load, 3, 16, 91, 108 Xylem, 188 Yeast, 19 Zea mays, 182, 184-186 Zoonotic diseases, 49-66