Lipoprotein Protocols

1 PCR Methodology Applied to Genetic Studies of Lipoprotein Metabolism and Atherosclerosis Shelley A. Cole and James E. Hixson

1. Introduction During the 1980s molecular genetics became widely applied to hpoprotein research. The cloning of candrdate genes for atherosclerosis (apohpoprotems, lipid-processing enzymes, lrpid transfer proteins, and hpoprotem receptors) allowed the identrficatron of mutattons responstble for rare dysltpoproteinemtas, as well as more common genettc variants. The cloned candidate genes were used as probes on Southern blots to find polymorphtc nucleottde substituttons at restriction enzyme sates(restrtction fragment-length polymorphisms or RFLPs) that could be used as genetic markers m linkage and assoclatron studtes. Such molecular-genetic studies have shown that genetic variation 1sa component m mdivrdual variation m lipoprotein metabohsm and rusk of atherosclerosts. Southern blotting IS a laborrous and time-consummg method to genotype large groups of indtviduals for genetic studies. With the advent of the polymerase chain reaction (PCR) to selectively amphfy specific DNA sequences (I), there came a much faster way to type indrviduals for genetic mutations involved m atherosclerosis, and to type polymorphic markers at candidate genes. Though most widely applied to RFLPs at first, PCR was soon applied to detect nucleotide substitutions and other polymorphrsms that were not associated with changes in restriction enzyme sttes. Although PCR has Impacted all areas of molecular-genetic studies of lipoprotem metabolism mcludmg clonmg, sequencing, and gene expression, this chapter will concentrate on the application of PCR m genetic-marker studies as a method to study the effect of inter-indrvtdual genetic varration on lipoprotein metabolism and atherosclerotic From

Methods in Molecular B/o/ogy, Vol 110 Llpoprotem Protocols Edlted by J M Ordovas 0 Humana Press Inc , Totowa, NJ 7

Co/e and Hixson

2

risk. Though not an exhaustrve presentation, it will include an overview of the basic approaches afforded by the PCR method. For a more detailed discussion of the PCR process itself, the reader is referred to the many excellent publications dedicated to this subject (i.e , refs. 2 and 3). 2. Materials 2.1, Equipment

and Supplies

1, Thermal Cycler (see Note 1). 2 Vertical electrophoresis apparatus (for acrylamide-gel electrophoresls, see Note 2) 3. Horizontal electrophoresis apparatus (for agarose-gel electrophorests, see Note 3) 4 Temperature-gradient gel electrophoresis (TGGE) apparatus (Biometra, Gottmgen, Germany or TGGE System, Diagen GmbH, Hilden, Germany) 5. Gradient maker (Bto-Rad Laboratories, Hercules, CA). 6 Dot- or slot-blottmg apparatus (Mmlfold I or II Mamfold, S&S, Keene, NH) 7 Ultraviolet transillummator (UVP, Upland, CA, or UVP, Cambrtdge, UK) 8 Recording camera (Gel Dot 1000 Video Gel Documentation System, Bto-Rad Laboratories, or Foto/System 1000 Camera System, Fotodyne, Hartland, WI). 9. UV chamber (GS Gene Lmker, Bio-Rad Laboratories) 10 Stertle thin-wall 0 5-mL microfuge tubes 11 96-Well Microtiter dishes. 12 Type 57 high-speed Polaroid film. 13 Whatman 3MM chromatographtc paper. 14. Hybond N filters (Amersham Life Science, Arlington Heights, IL, or Amersham Internattonal plc, Little Chalfont, Bucks, UK). 15. BtoMax MS mtenstfymg screen (Kodak Scientific Imaging Systems, Rochester, NY). 16. Autoradiographic film (Kodak XAR-5, Kodak Scientific Imaging Systems or Beta-max from Amersham).

2.2. Reagents 2.2.7. Restriction-Enzyme Cleavage of PCR Products to Detect RFLPs and Protein lsoforms 1 Synthetic olrgonucleotrdes (primers) flanking the sequenceto be amplified (see Note 4). 2. Tuq DNA polymerase: Thermus aquatlcus DNA polymerase or AmplitaqTM (Perkm Elmer Cetus, Norwalk, CT). Tuq DNA polymerase is currently available from several other manufacturers. 3 Mixture of dNTPs (dGTP, dATP, dTTP, dCTP each at 2.5 mM) Most suppliers provide premix nucleotldes; alternatively, the workmg mixture can be prepared by mixing equal volumes of a 10 mM solution of each of the four dNTPs 4 5X Tuq buffer (or 5X PCR buffer), whtch includes 250 & KCl, 50 mM TrisHCl, pH 8 3, and 7 5 mM MgCl* Thts buffer can also be prepared without the

PCR Methodology

5 6 7 8.

9

10

11 12

3

MgC12. In that case, the optimal amount of MgCl, should be added separately as suggested in Notes 5 and 6. MgCl, (see Notes 5 and 6) Dlmethyl sulfoxlde (DMSO). Filter-stenhze. Light mineral 011(see Note 7) Suitable restriction endonuclease enzymes and buffers: Suppliers generally ship the enzymes with optimized digestion buffers and they should be used as mdlcated by the manufacturer. AlternatIvely, One-Phor-All Buffer (Pharmacla Biotech, Piscataway, NJ) can be used with a wide range of reactlons. Appropriate gels either agarose or acrylamide for checking the reactlon and the DNA fragments (see Note 8) 5 mg/mL Ethldmm bromide (EtBr)* 500 mg EtBr (Sigma E-875 l), add ddH,O to 100 mL The working solution of EtBr (1 pg/mL) should be prepared by diluting the stock solution with water or gel buffer. 10X TBE buffer. 0 90 MTns-borate, pH 8 0,0.02 M ethylenedlamme tetra-acetlc acid EDTA). Dilute the stock to 1X TBE or 0 5X TBE for use 20X TAE buffer. 0 8 M Tns, 0.4 M NaOAc, and 0 04 M Na2EDTA, and glacial acetic acid to pH 8 3 m ddH,O Mix 96.9 g Tris base, 32 8 g NaOAc-3H,O and 14.9 g Na*EDTA Dissolve m approx 700 mL of ddH*O, adjust the pH to 8.3 with glacial acetlc acid, and bring to 1 L with ddH20. To prepare 1X TAE, &lute 1:20 with ddH*O. This buffer can be mltlally sterilized by autoclaving and it can be stored at room temperature. If needed, it can be sterilized agam by filtration through a 0 22-w filter

2.2.2. Gel Analysis of Physical Properties that Distinguish Allelic Substitutions in PCR Products Temperature Gradient Gel Electrophoresis Denaturing-Gradlent Gel Electrophoresis (DGGE), and Single-Stranded Conformation Polymorphism (SSCP)

(TGGE),

In addition to those reagents listed in Subheading 2.2.1., the followmg specific reagents are used for TGGE, DGGE, and SSCP. 1 Acrylamide (electrophoresls grade) 2. N,N’Methylenebisacrylamlde (electrophoresis grade). 3. 10% (w/v) Ammonium persulfate: preferably made fresh prior to use. Alternatively, it can be kept at 4°C for up to 1 wk. 4. TEMED (N,N,N’,N’ Tetramethylenediamme). Store at 4°C. 5 SDS 6 Urea 7 [Y-~~P]ATP 5000 Cl/mm01 (Amersham) 8. [Y-~~P]ATP: 3000 Wmmol. 9 [a-32P]dCTP. >3000 Wmmol. 10 T4 DNA polynucleotlde kmase (10,000 U/n&, New England Blolabs, Beverly, MA)

4

Cole and Hixson

11 T4 buffer, 50 mM Tris-HCl, pH 7 510 mMMgSO,, 5 mA4 DTT, 50 pg/mL bovine serum-albumm (BSA) 12 2X TGGE loading buffer 40% sucrose. 13. Glycerol. 14 Formamide dye. 950 mL/L formamide, 20 mmol/L EDTA, 0 5 g/L bromophenol blue, 0 5 g/L xylene cyan01 FF. 15. TGGE gel support films (Diagen GmbH, Hrlden, Germany)

2.2.3. Typing Known Nucleotide Substitutions Using PCR and Hybridization with Allele-Specific Oligonucleotides (ASO) In addition to those reagents listed in Subheading specific reagents are used for hybridization with nucleotides (ASO).

2.2.1., the followmg allele-specific oligo-

1 5X SSPE. 0.75 M NaCl, 50 mMNaH2P0,, 5 mA4 EDTA, pH 7 4 2 50X Denhart’s 1 g F~coll, 1 g polyvmylpyrohdone, 1 g BSA, and ddHzO to 100 mL Pass through a Nalgene filter and keep m refrigerator 3 20X Standard salme-citrate (SSC): 17 53 g NaCI, 8.82 g sodmm citrate Dissolve m approx 80 mL ddHzO, adjust pH to 7 0 wrth hydrochloric acid (HCl), and bring final volume to 100 mL. To prepare 1X SSC ( 15 mM NaCI, 1.5 mA4 Na citrate), dilute 1.20 with dHzO.

2.2.4. Batch Analysis of PCR Products 1 Gelatin 2. Nomdet P-40 3. Tween-20.

2.2.5. PCR-Based Solid-Phase Methods for Genotyping of Known Nucleotide Substitutions 1. 5’-Biotinylated primers (see Note 4). 2. 5% (w/v) Suspension of avidm-coated polystyrene particles 3. Phosphate buffered salme (PBS)* 140 mMNaC1,2 6 mMKC1, 10 mMNa2HP04, 1.8 mA4 KH2P04, pH 7 4. Store at 4°C. 4. Sequencing kits (1 e , Sequenase, US Brochemicals, Cleveland, OH)

3. Methods 3.7. PCR for Detection of Nucleotide Substitutions as Genetic Markers for Linkage and Association Studies For the detection of polymorphic nucleottde substitutions, PCR is an efficient process where results on many samples can usually be obtamed in a single day. However, the DNA sequence of the gene region containing the polymorphism must be known. This sequence is used to design synthetic ohgonucleottde primers that flank the polymorphtsm. PCR involves multtple

PCR Methodology

5

cycles of denaturation of the template DNA, annealing of the primers to the template, and extension wtth Tuq DNA polymerase to amplify the regron flanked by the primers Several different procedures can then be used to detect the presence or absence of the nucleotide substitution m the amplified fragment. For those that change restriction-enzyme sites, the amplified fragment is digested with the appropriate restriction enzyme and electrophoresed to determme the fragment sizes. For those substitutions that do not alter restrtcttonenzyme sues, several methods can be used to detect the presence or absence of the change. In some Instances, restriction sites can be created to distmgutsh the nucleotide substitution by using a mismatched primer (4). PCR products can be probed under stringent condttions with AS0 that only remain hybridized tf there are no sequence mismatches. Other techniques rely on the altered electrophorettc mtgration of DNA with different base compositions. Newer technologies are also discussed that depend on the initial amplificatton of the region of interest by PCR. 3.1.1. Restrrctlon-Enzyme Cleavage of PCR Products to Detect RFLPs and Protein

lsoforms

One of the first apphcattons of PCR m lipoprotein research was to type previously identified genetic markers. When a polymorphic nucleotide substitution alters a restriction-enzyme site m a PCR product, amplified DNA samples will show an RFLP on digestion with the enzyme and subsequent gel electrophoresis and staining with ethidmm bromide. If the restrtction sateis absent in both alleles of a DNA sample, there will be one fragment on the gel corresponding m size to the uncut PCR product. If the restriction site is present m both alleles, there will be two fragments whose sizes are determined by the position of the enzyme-cleavage site m the amplified fragment. If the DNA sample is heterozygous, both uncut and cut fragments will be present When protein polymorphisms are caused by nucleotide substttutions that also alter restrtctton-enzyme sites, they can be typed at the DNA level by restrtction-enzyme tsoform genotyping (restriction isotyping) (5). Restriction tsotypmg provides a rapid method of typing the common apohpoprotem E (apo E) tsoforms (E2, E3, E4) that are important determinants of plasma ltptd levels (6). Synthetic ollgonucleotides are used to amplify apo E gene sequences that span ammo acid postttons 112 and 158. The amplification products are digested with HhaI and subjected to electrophorests on polyacrylamide gels. Each of the isoforms is unambtguously distingutshed by a unique combination of HhaI fragment sizes.HhaI cleaves at G CGC encoding Argt12 (E4) and Arg158 (E3, E4), but does not cleave at G TGC encoding Cys’12 (E2, E3) and cyst58 (E2). Mutations m apo A-IV are responsible for isoforms that have effects on lipid and lipoprotein levels (7), and that may be associatedwith differential response

Cole and Hixson

6

to diet (8,9). Restrictton isotypmg can also be used to dlstmgulsh apo A-IV lsoforms at the DNA level (10). In this instance, a mismatched primer IS used m the PCR reaction to introduce a PvuII restriction site when the primer amphfies an apo A-IV-2 allele, but not an apo A-IV-1 allele. An alrquot from the same PCR reaction can be used to type the Thr347+Ser polymorphism by digestion with Hi&. 3.1 .l .l. BASIC PCR PROTOCOL

The protocols descrtbed m this chapter include one or more steps requiring DNA amplification. To simplify the description of specific protocols, we delineate below a basic PCR protocol. It should be noted that each laboratory may require slight modifications depending on the DNA template characteristics, DNA quality, source of reagents, and equipment available, to name some of the factors that may affect the PCR reaction (see Notes 5 and 6). The PCR reaction IS usually performed m volumes ranging from 5-l 00 $. Because of the small volumes used per reaction, we recommend making a cocktall with all the reaction components (allow for a slight excess)with the exception of the DNA template: To prepare a cocktail for 10 PCR reactions of 100 $ each, mix the following components: 1 200 $5X PCR buffer 2 80 pL dNTPs (each at 2 5 mM) This results m a final concentration of each dNTP m the reaction of 200 w. 3 40 & primer pair (each primer at 25 pM. The primer solution 1sprepared by mixing 1: 1 mixture of the 50-mprlmer solutions) This provides a final concentratlon of 1 pmol/pL. of each primer m the reactlon mixture 4 20 pL Taq polymerase 5. 650 pL ddH20 (some reactions require additional reactlon components, such as DMSO or additional MgC12, to achieve optimal results. This is usually specified in the reaction conditions. In those cases, the volume of the addltlonal component should be substrated from that of ddH20 to mamtam the final concentration of all other reagents) Keep the cocktail on ice

To perform the PCR reaction, follow the following steps. 1 Pipet 100 pL of the cocktail mto a sterile 0 5-mL Eppendorf tube (the amounts described should be scaled down to fit other usual reactlon volumes, i e , 50, 2075 a). 2 Add 1 p.L of DNA (contammg about 100 ng genomlc DNA). 3 Mix and quick spin to bring the reaction mixture to the bottom of the tube 4. If the thermal cycler does not have a hot bonnet, overlay the reaction mixture with one drop of mineral oil

PCR Methodology

7 A2A2

A7A2

AlA 4 255 bp

785 bp * 70 bp + FORWARD -

7.

PRIMER

REVERSE

4 Avall

785

PRIMER -

I

255 bp

Fig. 1. PCR-based typing of the AvaII RFLP in exon 13 of the LDL receptor gene. The schematic below the gel shows the portion of the LDL Receptor gene that is amplified for typing of the AvaII RFLP in exon 13 (primers shown by horizontal arrows, polymorphic AvuII site shown by a vertical arrow). The sizes of the uncut PCR product and the AvaII fragments are given below the map. The gel shows typing of DNA samples for each genotype (fragment sizes shown adjacent to the gel). 5. Cap the tubes and place them in the temperature block, making sure that each tube makes good contact with the heater walls. 6. Program the thermal cycler for the reactions conditions (see Note 6): Denaturation: 93-94°C for 1.5 min; Annealing: 5065°C for 2 min; Extension: 72°C for 2 min.

Below are presented some specific examples where the PCR technique is used to detect polymorphisms that may be relevant as genetic markers. (Primer sequences, PCR, and cutting conditions for other genetic markers frequently studied in lipid research are presented in Appendix 1.) 3.1 .1.2. DETECTION OF THE AvAII RFLP OF THE LDL RECEPTOR

Using PCR amplification, we have developed a protocol to detect the AvaII RFLP in exon 13 of the LDL receptor (12). The results are shown in Fig. 1. We used published sequence data (13) to design primers to amplify the region containing the RFLP. 1. Prepare the reaction mix as described in Subheading 3.1.1.1. (see Note 5), using the following primers: forward primer from exon 13 (5’-CAGTGCCAA CCGCCTCACAGG-3’) and reverse primer from exon 14 (5’-CCTCTCACA CCAGTTCACTC-3’). Use DMSO in the reaction at a final concentration of 10% (see Note 5).

8

Cole and Hixson 2 Add 0.5-1.0 1.18genomlc DNA. 3. Set the thermocycler to the following condltlons (see Notes 6 and 9): denaturation: 94’C for 1 mm, annealmg and elongation, 64’C for 1 5 mm Use 30 cycles of amplification. 4 After successful amphficatlon, digest the 255-bp amplification product with AvaII (see Note 10) by addmg 4-8 U of the restrlctlon enzyme directly to the tube containing the PCR reaction (see Note 11) 5 Incubate the reaction for 2-6 h at 37’C. 6 Analyze the results of the dlgestlon on a 1 5% agarose gel (see Note 12) The digestion of the amplification product with AvaII produces 185- and 70-bp fragments when an mdivldual is homozygous for presence of the site (designated A2A2), a 255-bp fragment if an mdlvidual is homozygous for absence of the site (designated AlAl), and all three fragments in heterozygotes (AlA2)

3.1.2. Gel Analysis of Physical Properties that Distinguish Allelic Substrtutions in PCR Products (TGGE, DGGE, and SSCP) Polymorphic nucleotide substitutions can be detected m PCR fragments even when they do not alter a restriction site. In TGGE, a double-stranded PCR fragment dlssoclates into a partially single-stranded fragment at a temperature specific for its DNA sequence, thus slowing its migration through a denaturing gel (19). Sequence differences as small as a single basepalr substitution are distinguished by distinct denaturatlon temperatures and migration to different positions in the gel. TGGE 1sa simplified version of DGGE, m which the temperature of the gel remains constant throughout electrophoresls, but the gel contains a gradient of denaturant (28). In SSCP analysis (20), radlolabeled primers are used to amplify fragments that are denatured, electrophoresed through nondenaturmg gels, and exposed to X-ray film. Polymorphic smglestranded fragments assume distinctive conformations and migrate differently owing to the substttutlons m their sequence (20). These techniques are simple and sensitive methods able to detect most nucleotide substitutions. The basic principles of primer design and optimization of the PCR reaction are the same as previously described for RFLP analysis. The primers should be chosen to amplify DNA fragments from 300-500 bp in length. In some cases, the primers may include GC-rich segmentson their 5’ ends called “GC-clamps,” which improve the detection of polymorphisms by altering conditions for fragment dissociation (21). The TGGE apparatus uses a standard vertical-gel electrophoresis chamber, or an apparatus specially designed for this purpose, with a denaturing polyacrylamlde gel contaimng urea and formamide. A temperature gradient is applied either perpendicular or parallel to the direction of electrophoresls. To establish the temperature gradient, aluminum plates with internal channels that circulate cooled and heated water are attached to the surfaces of the glass plates

PCR Methodo/ogy

9

of the vertical polyacrylamrde gel (19). For a more detailed descrrptron of this technique, see ref. 22 To type polymorphisms using TGGE, it is first necessary to determine the temperature range and condrtrons for denaturtng the amplified fragment. This is carried out using a denaturing 6.5% polyacrylamide gel (37.5:1, acrylamrde:bis-acrylamide; 0.8 mm thick) with one preparative well spannmg the width of the gel. 1 Prepare two acrylamide stock soluttons, one with 100% denaturant (40% formamide, 7 M urea), and one wtthout denaturant The final volume percentage of denaturant m the gel (usually 40-60%) IS determmed by the proportion of the denaturing acrylamtde solutron that 1s added to the nondenaturmg solutton In thts way, only two solutions need to be prepared and mtxed to generate a range of gels differing by their final denaturant concentrations 2 Establish a wide temperature gradient perpendicular to the dtrectton of electrophoresis, with cold temperature (I e., 16’C) on the left of the gel and hot temperature (1 e., 46°C) on the right of the gel 3. MIX the PCR reaction (100 pL volume) of the desired fragment with an equal volume of 2X TGGE loading buffer (40% sucrose), and load into the preparative well 4 Carry out the electrophorests in 0.5X TBE (70-80 V for 22 h). 5 Stain the gel using EtBr to reveal a sigmoidally shaped curve that results from denaturatton of the double-stranded fragment into more slowly migrating single-stranded structures. The narrow temperature range correspondmg to the posmon of the mflectton pomts of the sigmoidal curve represents the denaturmg temperature of the fragment This temperature range, along wtth the percent denaturant used m the gel, IS used to tdenttfy and type polymorphtsms m the fragment Once the denaturing condttions of the PCR fragment have been determmed, genomrc DNA samples can be routmely screened for the presence of polymorphisms. 1 Amplify the fragments of interest from DNA samples m PCR reactions (20 pL volume) 2 Mix the amplificatton product with an equal volume of 2X TGGE loading buffer 3 Load each sample mto separate wells on a 6.5% polyacrylamtde gel havmg the appropriate denaturant concentratron, determined as aforementtoned The alummum plates are used to establish a temperature gradient parallel to the dtrectton of electrophoresrs (cooler on top, warmer on the bottom) using the narrower temperature range as determined from the perpendicular gradient gel prevtously descrtbed. 4. Stain with EtBr. Polymorphtsms are resolved because each allele will denature at slightly different temperatures and migrate to distinctive positions during electrophorests

Cole and Hixson

10 3.1.2 1. IDENTIFICATION OF POLYMORPHISMS IN THE ANGIOTENSINOGEN (AGT) GENE USING TGGE

Recent studies have identified polymorphisms in AGT that are associated with essential hypertension in humans (for review, see ref. 23). One of these, the Met235+Thr substttution m exon 2 of AGT, was originally identified using SSCP (24). We developed a TGGE assayto type the Met235+Thr substitution, and identified four other variants m exon 2 of AGT (25). Two of these, a Tyr24*+Cys ammo actd substitutton and a T+C nucleotide substnution, had been previously characterized (24), whereas two new variants were identified m a US black populatton (Thr209+Ile and Leu2i’+Arg). To amplify exon 2 of the AGT gene use the following protocol: 1 MIX 19 & of the standard PCR reaction mixture (containmg the followmg primers forward prtmer 5’-GATGCGCACAAGGTCCTGTC-3’ and reverse primer. 5’-GCCAGCAGAGAGGTTTGCCT-3’) with 0 5 pg of DNA in 1 pL of ddHzO or the appropriate DNA storage buffer. 2. Carry out the amphficatlon usmg the following condmons Imttal denaturatron at 95°C for 5 mm followed by 30 ampbfication cycles (60°C for 1 min, 70°C for 2 mm, and 95°C for 1 mm) 3 Load the PCR products onto a denaturmg polyacrylamtde gel (5 6 A4 urea, 32% formamrde) with a temperature gradient of 33°C at the top of the gel and 41 “C at the bottom 4 Electrophorese the gel at 70 V for 17 h 5. Stam the gel wrth EtBr and photograph under UV light 3.1.2.2.

TGGE ANALYSIS OF ENDOTHELIAL CELL LEUKOCYTE ADHESION MOLECULE 1 (ELAM-1)

ELAM- 1 (also called selectm E) is an adhesion molecule that mediates bmding of circulatmg leukocytes to endothehal cells m the arterial wall. Figure 2 shows the results of TGGE analysis of intron 2 of the gene encoding ELAM- 1. PCR primers were destgned usmg published sequence data from exons 1 and 2 (26) to amplify the 631-bp intron (forward primer, 5’CAGTTTCTC TGAGCTCTCACTTTGG-3’; reverse primer, 5’- CCAGGATCCACTCTC TTTAATGA-3’). 1. Mix 19 pL of the standard PCR reaction mixture (contammg the primers mdtcatedabove) with 0.5 ug of DNA in 1& of ddHzOor DNA storage buffer (27) 2. Carry out the PCR reaction using the following condltlons denaturatton for 5 mm at 95’C, followed by 30 cycles of denaturatton at 95°C for 1 mm, annealing at 56°C for 1 mm, and elongatton at 70°C for 1 mm. 3 To determine the denaturatton temperature, electrophorese the PCR fragment at 70 V for 22 h m a 6 5% denaturing polyacrylamide gel (40% denaturant) with a temperature gradient (16-46”(Z) perpendicular to the direction of electrophoresrs. 4. Stam using EtBr.

11

PCR Methodology fLAM7 TGGE

22°C

I 29°C

16°C

-46°C

Fig. 2. PCR-based temperature gradient gel electrophoresis (TGGE) analysis of intron 2 of the ELAM-I gene. Panel A, TGGE analysis of a single PCR reaction of ELAM-1 intron 2 using a perpendicular temperature gradient. The denaturation temperatures of the fragment correspond to the inflection points of the sigmoidal curve. Panel B, PCR fragments from five different DNA samples using a horizontal gradient with temperatures derived from panel A. Samples with single fragments are homozygotes and samples with two fragments are heterozygotes for TGGE alleles.

Figure 2A shows the sigmoidal melting curve produced upon EtBr staining, with the double-stranded fragment on the left (16°C) denaturing into the slower-migrating single-stranded structures on the right (46°C). The presence of two inflection points indicates that this fragment has at least two melting domains. The major inflection point of denaturation occurred between 22°C and 29°C. Figure 2B shows detection of allelic variants in intron 2 of the ELAM- 1 gene. Intron 2 was amplified from 5 different genomic DNA samples, and electrophoresed on a 40% denaturing, 6.5% polyacrylamide gel with a temperature gradient parallel to the direction of electrophoresis (22°C at the top to 29°C at the bottom). Homozygous individuals yielded PCR fragments that migrate identically, and heterozygous individuals yielded two fragments that migrate differently owing to allelic nucleotide differences (see Note 13). 3.1.2.3.

DGGE

ANALYSIS

OF APO

E

ALLELES

In DGGE, PCR-amplified DNA fragments are denatured and migrate to different positions on a gel containing a linear gradient of denaturant based on sequence differences as small as a single bp (18). DGGE is therefore analogous to TGGE (described above). In fact, TGGE is a simplified procedure derived from theories originally used to develop DGGE. The PCR amplification strategy and interpretation of results are the same for both procedures. The difference between the two methods involves the gel electrophoresis. Whereas a temperature gradient is used to denature PCR fragments in TGGE, the gradi-

12

Co/e and Hixson

ent in DGGE is generated by lmearly Increasing the concentratton of a denaturmg solvent m the gel. DGGE uses a vertical gel submerged in a 5-L chamber contammg electrophoresis buffer (1X TAE; 0.04 M Tris-acetate, 0.00 1 M EDTA, pH 8 0). The buffer 1s stirred and mamtamed at 60°C. The 6.5% polyacrylamrde gel (1 X TAE; 30:0.8 acrylamide:bis-acrylamide) contams a lmearly increasing concentratton of urea/formamide, prepared using a gradrent maker or syringe pump (30). The denaturing gradient IS arranged perpendicular to the dtrectton of electrophoresrs to determine the denaturation profile, or parallel to the dn-ectron of electrophoresrs for detection and typing of polymorphisms The gel IS electrophoresed at 6 V/cm, then stained with EtBr and analyzed as prevtously described for TGGE Parker et al. (31) have developed a method of distmgmshmg apo E tsoforms by their melting behavior on denaturing gradient gels. They designed primers to amplify a 64 1-bp fragment from exon 4 of the apo E gene. One of the prtmers contamed a “GC clamp” consisting of 15 repeats of GC on its 3’ end to modify the melting behavior of the fragment. 1 Amplify 1 pg of DNA m 100 pL usmg a modtficatlon of the standard PCR reaction mixture that includes glycerol to a final concentration of 15% (v/v) The followmg primers are used (5’-TCGCCCGCCCCATCCCAGCCCTTC-3’ and 5’GCGCGCACCTCCGCCACCTGCTCCgcgcgcgcgcgcgcgcgcgcgcgcgcgcgc-3’) The amphticatton condtttons are as follows: 96’C for 30 s, 55’C for 1 mm and 72°C for 1 mm, 30 cycles 2. Drgest the fragment with StyI. This releases the 467-bp domain contaming the sequence differences that dtstmgursh the 3 common apo E isoforms, along with the arttfictal GC clamp. 3 Electrophorese the fragment (16 h at 75 V) on a 6 5% polyacrylamrde gel wrth a lmearly mcreasmg concentratron of denaturant from 40 to 90% (100% denaturant, 40% formamtde, 7 M urea m 1X TAE) All three lsoforms are dtstingutshed by therr drstinct mobrhty differences m the gel This method also has the potential to identtfy other rare known mutations or new sequence differences m this region of apo E

3 1 2.4 SSCP ANALYSIS The SSCP technique permits the identtficatron of most sequence varrations m DNA fragments between 150 and 250 nucleotrdes long This techmque takes advantage of the fact that under nondenaturaing conditions a smgle strand of DNA will adopt a conformation that IS determined by tts sequence composmon. Even a smgle base difference may induce a conformattonal change enough to be detectable by its different electrophoretrc mobility m acrylamrde gels The basic prmctples of prtmer design and optimization of the PCR reaction are the same as prevtously descrtbed. Primers are destgned to amplify 150- to

PCR Methodology

13

250-bp fragments m the gene of Interest (amplified DNA fragments greater than 300 bp should be avoided). To produce a radiolabeled PCR product, either the 5’ ends of the primers are labeled with [Y~~P]ATP or [Y-~~P]ATPusmg T4 DNA polynucleottde kinase prior to PCR, or [a-32P]dCTP is added directly to a standard PCR reaction usmg nonlabeled primers (20,32). The amplified, radrolabeled PCR product IS denatured usually by the addrtion of SDS (0.1%) and formamtde (50%), heating (95°C 2 min), and coolmg tmmedrately on ice. The samples are electrophoresed on a 5 or 6% nondenaturing polyacrylamtde gel (0.3-0.4 mm thick) at 4°C at -30 W, or at room temperature at 10 W (gel must contam 10% glycerol to mamtam secondary structures). The gel is dried and exposed to X-ray film (20,32). Polymorphrsms are distmguished by the different mobilrues of the radmlabeled single-stranded fragments. The mobility difference is owing to the distmctrve three-drmenstonal conformatrons assumed by the denatured, smglestranded fragments as they migrate through the gel. The unique conformattons are dependent on sequence drfferences as small as a single nucleotide substrtutron. 3 I 2 4 I SSCP Analysis of Apollpoproteln(a) [ape(a)]. Plasma levels of the atherogemc hpoprotem(a) [Lp(a)] are largely determined by the gene for ape(a), a unique protein constituent of Lp(a) particles. Lp(a) levels are inversely correlated with the number of krmgle 4 repeats m the associated ape(a) (33). However, ape(a) alleles of the same size are not always assocratedwith slmrlar plasma levels of Lp(a). In studres to find sequences other than kringle 4 that affect Lp(a) levels, Cohen et al. (34) used SSCP to identify sequence variation m the 5’ flanking region and in introns of the ape(a) gene. The authors used these SSCP polymorphrsms to distinguish ape(a) alleles in a family with a parent that appeared homozygous according to ape(a) size. They showed that each allele, although identical in size, segregated with different Lp(a) concentrations. Analysis of 23 mdtviduals whose ape(a) alleles were identical m size revealed that 91% were heterozygous at one or more of the SSCP sites, and showed that the ape(a) gene was more polymorphtc than originally believed. 1 Perform each PCR reaction m a 20-& volume of the standard PCR reaction mrxture contammg 0 1 pg genomrc DNA, and 3 3 pmol [a3*P]dCTP (3000 Wmmol). a For the polymorphrsm located m the 5’ flankrng regton (SSCPI), use the following prtmers S’-TGACATTGCACTCTCAAATATTTTA-3’ and 5’-CATATACAAGATTTTGAACTGGGAA-3’, and the followmg amplrfication condrttons: 30 cycles of 95°C for 1 mm, 55°C for 1 mtn, and 72°C for 1 mm. b For the polymorphism located m the mtron between the two exons that encode the first krmgle 4 repeat (SSCP 2) use the followmg primers 5’TTGGTCATCTATGACACCACATCAA-3’ and 5’-GGCTCGGTTGAT

14

Cole and Hixson

TTCATTTTTCAGC-3’ The ampltficatton conditrons are as follows+ 30 cycles of 95°C for 1 mm and 68°C for 3 mm. c For the polymorphism located m the mtron between the second exon of K433 and the first exon of K4-34 (SSCP3), use the following primers’ 5’-CTT GGTGTTATACCATGGATCCCAA-3’ and 5’CACCATGGTAGCAGT CCTGGACTGT-3’. The amphficatton conditions are as for SSCP2. d. For the polymorphism located in the mtron between the second exon of K435 and the first exon of K4-36 (SSCP4), use the followmg primers. 5’-ATC GAGTGTCCTCACAACTCCCACA-3’ and 5’-GGGGTCCTCTGATGC CAGTGTGGTA-3’ Amphticatton condmons as those descrtbed for SSCP2 2 Digest 10 pL of each reaction with 5 U of the appropriate restrtctton enzyme (Hue111 for SSCP2, H&I for SSCP3, and and HI&I for SSCP4) m a total volume of 50 pL at 37’C for 2 h. The SSCP 1 fragment does not need enzymatic dtgestton prior to electrophoretic separatton 3. Dilute 5 pL of the digestion reactton wtth 20 p.L of formamlde dye (for the SSCP 1 product use 1 pL in 20 pL of dye) 4. Denature at 95°C for 3 mm. Place on Ice munedlately 5. Load 3 uL onto a 6% nondenaturmg polyacrylamlde gel m 2X TBE with 10% glycerol. 6. Electrophorese at 300 V for 14 h 7 Dry the gel and expose to Kodak XAR-5 film for 12 h at -7O’C The samples bearing polymorphic alleles will be clearly identified by then dtfferent mobilmes as compared to the weld type. 3.1.2.4.2. SSCP Analysis of ELAM-I Figure 3 shows the results of SSCP

analysis of intron 2 of ELAM-1 This IS the same PCR fragment that was analyzed by TGGE as previously described, but m this case the fragment was digested with Hz&I prior to analysis to give fragment sizes (263 and 368 bp) that would be more sensitive for SSCP analysis. 1. Label both prtmers on their 5’ ends m a reaction contammg 1.5 ug of prtmers, 10 U T4’DNA polynucleotlde kmase, [Y-~~P]ATP (50 $I, 2000Ct/mmol), and T4 buffer for 30 min at 37°C 2. Carry out the PCR reaction as previously descrtbed for TGGE (Subheading 3.1.2.2.), except the first 20 cycles use only cold prtmers, and the last 10 cycles mclude the additton of 3 x 1O5total cpm of each labeled prrmer 3 Digest the labeled PCR fragment with HrnfI 4. Denature the fragments by addttion of SDS to a final concentration of 0.1%. 5 Add an equal amount of loadmg buffer (contaming 95% formamide), and heat at 95°C for 2 mm 6 Quench on ice and load the reaction onto a 8% nondenaturmg polyacrylamtde gel for electrophorests (35 W at 4°C) 7 Expose the gel to X-ray film to reveal the polymorphisms (see Fig. 3A,B). Samples m which two fragments can be dlstingmshed represent heterozygotes for SSCP alleles, and samples with single fragments represent homozygotes

15

PCR Methodology ELAMI

SSCP

Fig. 3. PCR-based single-strand conformation polymorphism (SSCP) analysis of intron 2 of the ELAM-1 gene. Panel A, SSCP analysis of 368 bp Hi& fragments, and Panel B, 263 bp Hinfl fragments from PCR amplified ELAM- 1 intron 2 from six DNA samples. Samples with single bands are homozygotes and samples with two bands are heterozygotes for SSCP alleles.

3.1.2.4.3. SSCP of Exon 3 of the LDL Receptor Gene 1. Amplify the LDLR exon 3 using the following flanking primers: 5’ primer; STGACACTTCAATCCTGTCTCTTCTG-3’ and 3’ primer; 5’-ATAGCAAAG GCAGGGCCACACTTAC-3’. The fragment is labeled by the inclusion of 0.03 pCi/PCR reaction of [a-32P]deoxycytosine triphosphate (800 pCi/mmol, 10 pCi/pL) in the PCR mix. 2. Use the following PCR cycling conditions: 96°C for 5 min and 68°C for 5 min for one cycle; 96°C for 1 min, 68°C for 1 min for 35 cycles; and one cycle of 5 min at 68°C. The resulting fragment should be 172 bp. 3. Dilute 5 & of the PCR mixture with 25 pL of 1 g/L SDS and 10 mmol/L EDTA. 4. Mix 5 pL of this solution with 5 pL of formamide dye. 5. Denature the mixture by boiling at 100°C for 3 min, and chill immediately on ice. 6. Load 4 & of each sample onto a 7.5% polyacrylamide nondenaturing gel (acrylamide:bisacrylamide, 49: 1) in 1X TBE buffer containing 50 mL/L glycerol. Gels are 40 cm x 30 cm x 0.4 mm. 7. Run the electrophoresis at 200 V for 16 h at room temperature. 8. Transfer the gel onto a Whatman 3 MM chromatographic paper, dry, and expose to X-ray film for 24 h at -70°C. 9. SSCP variants will be apparent in the autoradiograph by the presence of extra bands in the corresponding lanes (see Note 14).

3.1.3. Typing Known Nucleo tide Substitutions Using PCR and Hybridization with AS0 Previously identified single nucleotide substitutions can also be detected by hybridization of DNA samples immobilized on nylon or nitrocellulose membranes with 32P- labeled ASOs. Under stringent hybridization and wash conditions, ASOs will only anneal to sequences that match perfectly, with a single mismatch being sufficient to prevent hybridization (38). Before PCR, AS0 techniques were limited by high background, and required large amounts of genomic DNA. The sensitivity of AS0 hybridization was greatly improved

76

Cole and Hixson

by PCR amplification of the target sequence prior to hybndlzatlon, making it a powerful technique for detecting nucleotide substitutions (39). For the PCR-based AS0 assay, primers are designed to amplify the region containing the nucleotlde substitution. A small amount of the PCR reaction is denatured and applied to a nylon or nitrocellulose filter. This “dot blot” IS hybridized under strmgent condltlons with an AS0 that has been labeled at its 5’ end with [T-~~P]ATP and T4 DNA polynucleotlde kmase. The dot blot 1s washed under high-stringency condltlons to remove unmcorporated probe, and exposed to X-ray film The AS0 will only hybrldlze to a perfectly complementary sequence, so that ASOs can be designed to detect both wild-type and mutant sequences. 3.1.3.1.

PCR AND AS0 HYBRIDIZATION TO DETECT APO E ISOFORMS

Several protocols have been developed to type the three common isoforms of apo E by PCR and AS0 hybridization (40-42). In each case, primers were designed to amplify the apo E gene region encoding the polymorphic residues at ammo acid positions 112 and 158. Four different ASOs were designed (19-2 1 nucleotldes m length) to detect the nucleotide differences that code for Arg’ l2 (E4) or Cys I’* (E2, E3), and Arg158(E3, E4) or CYS’~~(E2). The protocols differed in that some ASOs were designed to hybridize to the coding and others to the noncodmg DNA strands, so that mismatches were A with C rather than G with T, which IS more thermodynamlcally stable. Each AS0 was labeled at its 5’ end with T4 DNA polynucleotlde kinase and [Y-~~P]ATP.Hybridlzatlon condittons for each AS0 were determined empirically using PCRamplified apo E DNA from mdivlduals with known genotypes. A small amount of the amplified apo E DNA was dotted on several filters for hybrldlzatlon and washing under the appropriate stringent conditions for each AS0 probe. The pattern of hybridization of the ASOs to the dot blots identified the apo E isoform genotype. In one casewhere a specific PCR product was not obtained, the PCR reaction was electrophoresed and the gel was blotted to detect hybrldlzatlon of the ASOs with PCR products of the correct size (41). 1 Mix 0.5 pg of DNA with 50 pL of a modified PCRreaction containing in addition 10mA4P-mercaptoethanoland 10%DMSO The primers for this reactlon are the following 3’-ACGCGGCGGACGTCGAGGAACCT-5’ and 3’-GAGGGA CCGGCCGGTCGGCGAT-5’ 2. Carry out the amplification using the followmg condltlons Initial denaturatlon at 95%C for 7 mm followed by 30 ampllficatlon cycles(95°C for 1 mm and 65°C for 2 mm). 3 Load 5 pL of the PCR products onto a 1 5% agarosegel to check for amphficatlon. Stain the gel with EtBr and examineunder UV light. A 330-bp amphficanon band should be visible.

PCR Methodology

17

4. The following allele-specific probes are labeled at their 5’ ends with [Y~~P]ATP and T4 polynucleotlde kmase (see Subheading 3.1.2.4.2.): E3( 112)*3’-CTCCTGCACACGCCGGCGGAC-5’ E4( I 12): 5’-GAGGACGTGCGCGGCCGCCTG-3’ E3( 158): 5’-CTGCAGAAGCGCCTGGCAGTG-3’ E2( 158): 3’-GACGTCTTCACGGACCGTCAC-5’ 5. MIX 1 pL of the PCR reaction with 200 pL of 15X SSC. 6. Heat to 95-1OO“C for 7 min to denature the DNA. 7. Apply to a Hybond N filter using a slot-blotting apparatus. Four Identical filters should be prepared. Label each of them with the name of the probe that will be used for hybridlzatlon E2( 158); E3( 158); E3( 112); and E4( 112). 8. Bmd the DNA to the filters by 3 mm ultraviolet uradtatton. 9. Prehybridize each filter mdrvidually in 4 mL 5X SSPE, 5X Denhart’s, 0.5% SDS, for 30 min at 65°C. 10. Add 5 pmol of the corresponding probe to each filter. 11. Hybridize at 65°C for 1 h 12. Rinse quickly with 2X SSPE, 0 1% SDS at room temperature. 13. Wash with 5X SSPE, 0 1% SDS for 15 mm. Use 69°C for the E2( 158) and E3(158) probes and 73°C for the E3(112) and E4(112). 14 Expose to X-ray film for 16 h at -70°C using a single intensifying screen, 15. The results should be interpreted as follows: E2(158)

E3(158)

-t

-

+

-

+ + -

+ + + + -t

-I+ + -I-

+ + +

(+, posrttve hybridization

E3(112)

E4( 112)

signal; -, hybridization

Genotype E22 E23 E33 E34 E24 E44 signal negative)

3.1.4. Batch Analysis of PCR Products In addition to typing common polymorphisms, PCR can be used to type rare disease mutations in populations. However, rapid screening of large numbers of DNA samples requires an efficient method that combines pooling of DNA samples followed by PCR amplificatton. This allows batch screening of many samples, with further analysts performed on only those batches that contam samples with the mutatton. 3.1.4.1.

APOLIPOPROTEIN 6 ARG~~OO-+GLN

Batch analysts has been applied to the detection of carrters for the Arg3500+Gln mutation in the receptor binding domain of apohpoprotein B (apo B). This mutation is responsible for familial defective apo B-100, a disorder

18

Cole and Hixson

characterized by diminished LDL receptor bindmg and high plasma cholesterol and LDL levels (for review, see ref. 43). 1 Pool up to 50 blood samples(1 mL of each). 2. Extract the DNA from the pooled material. 3. Amplify 250 ng of the pooled DNA sample m a total reactlon volume of 50 pL containing 12.5 mMTris-HCl, pH 8 3,62.5 mMKCI,2 25 mMMgCl*, 10% (v/v) DMSO, 0.0 1% gelatme, 0.11% (v/v) Nomdet P-40, 0 11% (v/v) Tween-20, 100 ng of each primer (BCF. 5’-CCAACACTTACTTGAATTCCAAGA GCACAC-3’ and BCR: 5’-GAATATATGCGTTGGAGTGTGGCTTCTCC-3’), 200 r-in/rof each dNTP and 1.5 U of Taq polymerase The thermal cycler condltlons are as follows* one step at 97°C for 1 min, followed by 30 cycles of denaturatlon at 95°C for 30 s, annealing at 40°C for 30 s, and extension at 72°C for 30 s This step amplifies both the wild-type and the possible mutant sequences 4 Check for successful amplification of the DNA by agarose gel electrophoresls (3%) using a 5-pL allquot. 5 Examine the ampllficatlon product, containing both wild-type and possibly mutant sequences, using a second round of ampllficatlon. Take a 5-pL allquot, dilute 1.100 with distilled water, and add 2.5 $ to a SO-@, PCR reactlon mixture, similar to the onepreviously described, but contammg primers (BMSF) (5’-TTC CAAGAGCACAAA-3’) and (BCR) The amplification condltlons are as follows: one step at 97°C for 1 min, followed by 30 cycles of denaturation at 95°C for 30 s, annealing at 60°C for 30 s, and extension at 72’C for 30 s The second amplification uses an allele-specific primer that produces an amplified fragment only if the apo B Arg3500+Gln mutation 1spresent (44) 6 For those pools that produce a second-round PCR fragment, individual DNA samples are isolated 7. Add 2.5 & contaming 250 ng of DNA to a 50-pL PCR mixture similar to the ones previously described but containing the primers BRSF (5’-CCAACACTT ACT TGAATTCCAAGAGCACCC-3’) and BCR Amplification condxtlons are one step at 97°C for 1 mm, followed by 30 cycles of denaturatlon at 95°C for 30 s, annealing at 40°C for 30 s, and extension at 72%C for 30 s The BRSF 1s a mismatched primer that creates an MspI restriction site m the mutant allele (45)) thereby identlfjrmg indlvlduals who are carriers of the apo B Arg3500+ Gln mutation 8. Mix 5 pL of the PCR product with 5 pL of “One-phor-all” restriction endonuclease buffer containing 10 U of MspI and incubate at 37°C for 3 h. 9. Examine the digestion products by electrophoresis rn a 3% agarose gel The normal allele produces a band of 104 bp and the mutant allele results in a band of 133 bp.

3.1.5. PC/?-Based Solid-Phase Methods for Genotyping of Known Nucleotide Substltutlons New PCR-based solid-phase technologies are being developed that may facilitate automated typing of known nucleotide substitutions. Automated tech-

PCR Methodology

19

nologies would be particularly useful m the clmlcal setting, which requires rapld and accurate typing of DNA samples from patients. These new technologies exploit binding of biotm-labeled PCR products with a solid-phase matrix contaimng avidin. 3.151.

SOLID PHASE NUCLEOTIDE SEQUENCING-APO

E

Automated direct sequencing of PCR products allows the unambiguous typing of nucleotide substitutions. An example 1s the direct sequencing of affinity-captured PCR-amplified apo E DNA fragments (46). The region of apo E encoding the common isoforms is amplified with one set of primers, and then subjected to a second round of amplification with a nested set of primers to increase the speclficlty of the PCR product. One of the primers m the second amplification reaction 1slabeled with blotm. The biotmylated amplified fragment is captured on an avldm solid matrix, taking advantage of the biotmavldm interaction. The nonblotmylated strand of the captured fragment 1s released by denaturation, the remaining strand is purtfied, and Its nucleotlde sequence IS determmed directly by the chain-termination method. This technique allows the routme typing of all combmatlons of the apo E lsoforms, as well as the detection of rare mutations m the region. Carry out the first PCR reactlon using the standardproceduredescribed in Sub2

3. 4 5 6. 7. 8. 9 10 11 12 13

heading 3.1.1.1., and the following primers (Pl*5’-AAGGAGTTGAAGGCC TACAAAT-3’ and P4. 5’-GGATGGCGCTGAGGCCGCGCTC-3’) Take a small ahquot of the product of the previous PCR reactlon (3 pL of a 1* 100 dilution) and carry out a second PCR using the same amphficatlon condltlons and the followmg nested primers (P2. S-TCGCGGGCCCCGGCCTGGTACA-3’, and P3: S-GAACAACTGAGCCCGGTGGCGG-3’), one of whxh IS 5’-blotmylated Take a 25-pL aliquot of the second amphficatlon and dilute 1: 1 with PBS Add 5 pL of a 5% (w/v) suspension of avldm-coated polystyrene particles Leave the samples at room temperature (-2O’C) for 1 h. Precipitate the particles m a microfuge for 2 min at the highest speed settmg. Wash the preclpltate with 1 mL of 15 mMNaC1, 1.5 n-uI4 Na citrate. Repeat the wash twice with PBS containing 0 1% Tween-20. Incubate the particles with 1 mL of 0.15 A4 NaOH for 15 mm at 37°C Centrifuge as aforementioned and repeat the incubation with 1 mL of 0.15 M NaOH under the same condltlons Wash twice with 1 mL 0.1% Tween-20 in 50 mMNaCl,40 mA4Tns-HCI, pH 7.5 Carry out a last wash with 0.01% Tween-20 m 50 m&I NaCl, 40 mM Tns-HCI, pH75. Suspend the particles carrying the DNA template m 10 pL of 50 mMNaC1, 20 mM MgCl,, 40 mM Tns-HCl, pH 7.5, contaimng 0 5-l pmol of the sequencmg primer (nonbiotmylated P2 or P3 as aforementioned, depending on which of the primers has been blotmylated for the second PCR reactlon)

Cole and Hixson 14. Proceedto the sequenceof the fragment using the chain-termination method of Sanger(47), usmg one of several sequencingkits available 3.1 5.2. PRIMER-GUIDED NUCLEOTIDE INCORPORATION ASSAY-APO

E

This technique uses the same strategy of immobihzmg PCR products on a solid phase for characterization of the nucleottde substitutton As aforementioned, the region of apo E encoding the common isoforms is subjected to two rounds of amplification, the second round using a nested biotm-labeled PCR primer. The biotinylated fragment 1sthen captured and purified on an avidm matrix (46). In this protocol, the captured strand is used as a template for parallel one-step primer-extension reactions by DNA polymerase. Each reaction uses a labeled nucleoside triphosphate complementary to one of the two possible nucleotides present at the variable site. A primer anneals to the template immediately upstream of the nucleotide substitution, and is only elongated by addition of the labeled nucleoside triphosphate tf it is complementary to the nucleotide at the variable site. In a sample from a heterozygote, a signal is obtained in both reactions. Samples from homozygous mdividuals produce signal from only one reaction. This method is particularly suitable for automation because the results are quantitative (based on total mcorporated signal) and electrophoretic separation IS not required. Nonradtoactive detection methods are also possible, which renders the technique more suttable for climcal diagnostics (48). 3.2. PCR to Defect Hypervariable Repeat Polymorphisms Although nucleotide substitutions are a useful tool for genetic analyses, their small number of alleles (usually two) and low allele frequencies often limit their use for linkage analyses m families. For linkage studies, the markers of choice are hypervartable repeat polymorphisms. Hypervariable repeats are regions in the genome containing many copies of repeats of simple sequences m tandem arrays. Hypervariable repeats are highly polymorphic with respect to number of repeats, typically yielding large numbers of alleles and high levels of heterozygostty, which facilitates tracking of specific alleles within families. Although repeats of varying sizes can be found near candidate genes, the more common repeats of two, three, or four nucleotides (dt-, tri-, and tetranucleotide repeats) are located throughout the genome and have proven useful m genome-wide screens. Hypervariable repeat polymorphisms are typed using PCR primers from unique sequences that flank the tandem repeat region, followed by electrophoresis that separates the PCR fragments by size.

PCR Methodology

21

3.2 1 PCR Typing of Hypervariable Repeat Polymorphisms in Candidate Genes of Lipid Metabolism Hypervariable repeats that are located close to known genes can be used to study the influence of candidate loci on quantitative phenotypes m family studies. Such repeat polymorphisms have been reported near genes for the LDL receptor (49), apohpoprotem CIII (50), lipoprotein hpase (51), and apo B (52-55,. As previously stated, PCR primers are designed from unique sequences that flank the hypervariable repeat The PCR reaction must first be optimized to efficiently amplify the desired fragments, and to limit the amount of nonspecific PCR products. One strategy IS to optimize the reaction usmg unlabeled primers, and to check the products using a 1.5-2% agarose gel, or 4-8% polyacrylamide gel with EtBr staining Although every reaction has to be optimized for each primer pan, the followmg conditions are suggested to begin the optimization procedure: 1 Mix 0 5 pg of genomrcDNA with the standardreactioncocktail in a final volume of20 pL 2 Use the followmg standard PCR condttlons: mittal denaturatton at 95°C for 5 mm, 25 cycles of denaturatton at 95°C for 30 s, annealing at 55°C for 30 s, and elongation at 72°C for 30 s, and a final IO-mm elongation pertod at 72°C 3 Once the opttmal PCR condmons have been determmed, radiolabel one of the prtmers at the 5’ end using T4 DNA polynucleotrde kmase and [T-~*P]ATP or [Y-~~P]ATP, and add it to each sample m the PCR reaction to generate radrolabeled fragments 4 Denature the amplified DNA by addmon of gel-loadmg buffer containing 95% formamtde and trackmg dye. 5 Heat at 95°C for 2 mm and remove immedtately to Ice 6 Electrophorese the samples through a 48% denaturing, polyacrylamtde gel (19 1, acrylamide*bis-acrylamtde, 7 M urea) m a standard sequencmg gel apparatus 7. Dry the gel under vacuum at 80°C and expose rt to X-ray film for several hours to overnight. Each ampltfied allele appears as a band on the autoradrogram. The size of each allele IS proportional to the number of repeats present In Its hypervarlable regton

3.2.1 .l

ANALYSIS OF THE APO B 3’ HYPERVARIABLE REGION

Perhaps the best example of a hypervariable polymorphism at a candidate locus of lipid metabolism is the hypervariable region located near the 3’ end of the apo B gene (called apo B 3’ HVR). It contains multiple copies of AT-rich, 15-bp tandem repeats located 75 bp downstream from the second polyadenylation signal of apo B (52,53). Many studies of different populattons have identified large numbers of apo B 3’ HVR alleles. In the study entitled “Pathobiological Determmants of Atherosclerosis in Youth (PDAY),” up to 22

22

Cole and Hixson

different HVR alleles were identified among 232 mdlvlduals (56). Nucleotide sequencing of these apo B 3’ HVR alleles revealed extensive variation with respect to both numbers and specific sequences of the AT-rich repeats. The apo B 3’ HVR was amplified m a PCR reactlon as follows: 1 MIX 0.5 pg genomlc DNA with the standard PCR reaction cocktail contammg 1 pmol/& of each primer (forward primer 5’-ATGGAAACGGAGAAATTATG3’; reverse primer S-CCTTCTCACTTGGCAAATAC-3’), m a final volume of 20 pL 2 Denature the reactlon mixture at 97°C for 10 mm, and carry out the amphficatlon reaction using the following condltlons 25 cycles of denaturatlon at 94°C for 1 mm, annealing at 55°C for 1 mm, and extension at 72°C for 5 mm. 3. Radiolabel the forward primer at its 5’ end m a reaction usmg 1.5 ~18of primer, 10 U T4 DNA polynucleotlde kmase, [Y-~*P]ATP (50 &I, 6000 Cl/mmol), and T4 hgase buffer for 30 mm at 37’C 4 RadIolabel the amphticatlon products by adding 1 ng of the 32P-labeled forward primer and contmue the amphficatlon reaction for 5 addItIona cycles 5 Electrophorese the amphficatlon products through 4% denaturing (7 A4 urea) polyacrylamlde sequencmg gels for 5 h at 100 mA 6 Dry the gels and expose for autoradlography with mtenslfymg screens Figure 4 presents an autoradlogram from the PDAY study that shows the large numbers of apo B 3’ HVR alleles (Note 15)

3.2.2. PCR Typing of Random Hypervariable Repeats for Genome Searches to Find New Genes that influence L/p/d Metabolism In addition to hypervarlable repeats near known candldate genes, PCR can be used to type hypervarlable repeats dlstrlbuted throughout the genome (called short tandem repeats or STRs). The STRs are used for genome searches to detect linkage wtth quantltatlve risk factors of hpid metabolism and atherosclerosis (60,61). Genome searches have been successful m identifying novel genes that cause smgie gene disorders such as cystic fibrosis, retinoblastoma, Duchenne muscular dystrophy, neurofibromatosls type 1, Huntington’s dlsease, and familial forms of several different cancers. Recent advances in gene mapping now permit investigators to conduct genome searches m standard laboratory settmgs. A crItIca development has been the identification of thousands of STR polymorphisms that are eastly typed by PCR (62). Most of the STR polymorphisms

bemg used for genome

searches consist of tandemly-

associated repeats of short sequences (dl-, tri-, or tetranucleotldes) (63,64). The STR markers can be typed by methods described m Subheading 3.2.1. using radlolabeled

primers,

or by recently

developed

automated

methods

dlscussed below. A crltlcal development

for conducting

genome searches IS the automation

of

PCR-based methods for high-throughput typmg of STRs usmg automated DNA

23

PCR Methodology

APO B 3’ HYPERVARIABLE

REGION

Fig. 4. PCR-basedtyping of apo B 3’ hypervariable region (apo B 3’ HVR) alleles. This gel shows different apo B 3’ HVR alleles identified in DNA samplesfrom the PDAY study. Each lane contains amplified apo B 3’ HVR fragments from a single DNA sample,and the final lane on the right showsa DNA sizestandard. sequencers (65,66). The STR fragments are labeled using PCR with fluorescently tagged primers, allowing automated detection after electrophoresis by sensors in the DNA sequencer. Associated computer programs automate estimation of STR fragment sizes,allele-calling according to size, and genotype entry into the computer database for further analyses. A great advantage of this system is that many different STRs can be simultaneously typed in a single lane during electrophoresis (called multiplexing). STR fragments labeled with the same dye, but that differ in size, can be distinguished in the same lane. The DNA sequencer also can detect each dye separately, so STRs of the same size,but that differ in dyes, can be combined in the same lane. The result is that large numbers of markers can be multiplexed, separated by both dye and size. A recent study has shown that fluorescence-based genotyping is at least as accurate as autoradiography-based methods with respect to variation both within and between gels (67). Although dinucleotide repeats occur with the highest frequency (68), increased allelic size differences make tri- and tetranucleotide repeats better for automated genotyping. Recent efforts have yielded large numbers of tri- and tetranucleotide repeat markers, including their assembly for genome-wide linkage maps (69). DNA samples from family members are placed into 96-well microtiter dishes for separate PCR reactions of each STR marker to avoid preferential

Cole and Hixson

24

amplification of particular markers in combined reactions. A typical PCR reaction (total volume 15 p.L) IS carried out as follows: 1 Plpet mto each of the microtlter wells 100 ng DNA, 5 pmol of each fluorescently labeled primer, 0.5 U Taq polymerase, 0.25 mM dNTPs, and 5X PCR buffer 2 Place the plates m a thermal cycler for PCR amphficatlon. Standard PCR condotions include initial denaturatlon at 95°C for 5 mm, 25 cycles of denaturatlon at 94°C for 30 s, annealing at 55°C for 30 s, elongation at 72°C for 30 s, and a final 1O-mm elongation period at 72°C 3 After PCR, pool the products of separate primer pairs mto another microtiter dish for multiplexing. 4 Add a red-labeled DNA size standard to each mixture. 5. Load the pooled PCR products mto one of the 36 wells on a sequencing gel mounted in the DNA sequencer for electrophoresls The STR fragments are detected and quantltated by fluorescent emissions, and their sizesare estimated by comparison with the DNA size standard (see Note 16). Figure

5 shows fluorescent

emlsslon

profiles (called electrophoretograms)

for 9 multiplexed STRs (three each of blue, yellow, and green) using the automated DNA sequencer. Each electrophoretogram IS a single color, and all have been measured from a single lane. The peaks show the positions and amounts of the STR fragments on the gel. The boxed numbers under the peaks show the sizes that are estimated from a size standard that is coelectrophoresed

in every

lane. These sizes provide the labels for each allele, and are automatically entered into the database for subsequent lmkage analyses. 3.3. Conclusions PCR has become one of the most widely used techniques in molecular blology, and has allowed a more efficient apphcatlon of molecular genetic studies to a range of disciplmes. Although it ISimpossible to cover the complete extent of appltcatlons of PCR technology, its impact on all aspects of the genetic study of lipid metabolism is evidenced by this chapter. From typing of RFLPs and nucleotlde substltutlons to automation of genome-wide screens, PCR plays an integral part in the ldentificatlon of genes important in hpld metabolism and atherosclerosis. 4. Notes 1. Thermal cyclers for PCR are now avallable from multiple vendors The followmg list represents, m alphabetical order, the most commonly used brands according to a recent Cell’96 survey: a Eppendorf North America (Madison, WI) and Eppendorf-Netheler-Hmz GmbH (Hamburg, Germany) b Erlcomp (San Diego, CA)

25

PCR Methodology

Sample 0071 Blue

Yellow

I

-30000 -20000 -10000

&jij dsh

-1500

-1000

Green

-500

r,ll ,l ,,“‘,“‘,‘“,‘“,“‘,“‘,‘“,“’,”’,

120

bo

vSamplelD 0071 Mother 0532Father 0194 0197 0533 0522

140

160

D2S13601(Blue) 144 152 136 136 136 144 136 152 136 144 136 152

180

200

220

240

DlS5181 (Blue) 199 211 207 215 207 211 199 215 199 207 207 211

260

280

DlS15971 163 171 171 175 163 171 163 175 163 171 171 171

300

(Green)

320

340

D2S17801(Yeilowl 320 320 316 316 316 320 316 320 320 316 316 320

Fig 5. PCR-based multtplexmg of fluorescent STR polymorphisms for automated genotypmg. An electrophoretogram from an automated DNA sequencer is shown for rune multiplexed trt- and tetranucleottde STRs (three for each color) for a smgle indtvtdual (sample 007 1). The peaks show fluorescent emissions, and the boxed numbers under the peaks show allele sizes m base pairs. The table under the electrophoretogram shows allele sizes for a nuclear family (0071 is the mother) for four of the STRs.

c Hybaid (In England: Teddmgton, Middlesex; m the US: Franklm, MA). d Lab-Line Instruments (Melrose Park, IL) e MJ Research (Watertown, MA).

26

Cole and Hixson

f. Perkm-Elmer/Applied Biosystems (Foster City, CA) g Techne (m England. Techne, Duxford, Cambridge; m the US* Techne, Prmceton, NJ) 2. Vertical electrophorettc units and power supplies are available from several sources 3. Standard horizontal agarose gel boxes and electrophoresrs power supplies are available from several sources A peristaltic pump might be also recommended to recirculate the buffer. 4 The basic rules of primer design for PCR in general also apply to the design of PCR primers that are used to amplify polymorphic regions of candidate genes for atherosclerosis The primers should have similar G+C content, and not be self-complementary, or complementary to each other, especially at then 3’ ends (3) Synthetic ohgonucleotide primers are available through many commercial sites Although more expensive than shorter primers, longer primers (25-30 nucleotides) provide more consistent amplification. The primers should be chosen so that the target fragment can be easily amplified (1 e , less than 1 kb m length), and the digestion products can be easily separated on standard agarose or polyacrylamide gels When designing primers, we sometimes create a restriction site for later cloning experiments by replacing one or two nucleotides in the primer’s 5’ region. For specific applicatrons brotmylated or fluorescent labeled primers are used. These primers they can be custom-made (Gibco-BRL) or the label can be mcorporated usmg available commercral kits: FluoroAmpTMT4 Kinase Brotm Ohgonucleotide labeling system (Promega). 5 A typical PCR reactron will have a final volume of 20-30 pL, and will contain 0 5-l .O pg of genomic DNA sample, 1 pmol/pL of each primer, 0.025 U/pL Tuq DNA polymerase, 200 $4 of each dNTP, and the Tuq manufacturer’s buffer, which includes 50 rnA4 KC1 and 10 mMTris-HCl, pH 8 3. MgCIZ is necessary at a final concentration of usually 1 5-2.0 r&f, but can vary by several mM The reaction’s final MgC12 concentration is also affected by the amount of EDTA rn the sample’s storage buffer. In addition, the PCR reaction may be optimized by adding the cosolvent, DMSO (10% final concentration). Inmally, amphfication is tried with various MgCl, concentrations, and with DMSO present or absent, to determme the optimum buffer conditions for amplification A PCR Optimizer@ Kit is available from Invitrogen (San Diego, CA) 6 There are many strategies to achieve successful amplification and to optimize PCR reactions. Optimization should begin with determmation of the best MgC12 concentrations, followed by varying annealing temperatures, and trymg the reaction with and without a cosolvent If a set of primers still does not give successful amplification, other parameters ofthe reaction can be varied (e.g., times and temperatures for PCR cycles) If a set of primers does not work after many attempts, a new set should be designed using a different nucleotide sequence Different approaches to improve sequence specificity of PCR reactions have been developed. Smce Tug polymerase has some activity at lower temperatures, combmmg

PCR Methodology

7.

8

9.

10

11

12.

27

the PCR reaction components at room temperature can lead to nonspecific hybridization and spurious amplification products that consume PCR reagents and decrease the yield of the specific product. To address this problem, the PCR reaction can be assembled at a temperature above the annealing temperature of the primers (i e , 80°C) and then cycled, a procedure referred to as “hot start PCR” (24,15). Another method to prevent nonspecific priming IS called “touchdown” PCR (16). In this procedure, initial PCR cycles have a high annealing temperature, and the annealing temperature IS lowered through subsequent cycles. This protocol has been shown to increase both the specificity and yield of the desired PCR product (17). If the Thermal Cycler used does not have a heated bonnet then a drop of light mineral 011should be used to prevent evaporation of the sample. Several supphers provide mineral oil for PCR applications or wax beads. Invitrogen provides a HotWax OptiStartTM Kit for PCR Optimization using HotWaxTMMg2+ beads Several types and sources of agarose are currently available. For an m-depth description of the techniques to separate different DNA size fragments see Chapter 2 The temperatures and lengths of time for each step m the amplification cycle (primer annealing, extension, and denaturation), as well as the number of cycles, is empirically determined for each candidate gene These are not only gene-specific, but also depend on the type of PCR instrument that IS being used For amplification from a genomic sample using thin-walled PCR tubes in a thermocycler, we usually start with mitral denaturation at 9S’C for 5 mm, then 30 cycles with denaturation at 95°C for 30 s, annealing at 55-6O”C for 30 s, and extension at 72°C for 45 s, with one additional final extension at 72°C for 10 mm. The times and temperatures, as well as other reaction parameters, are varied depending on the results of the mnial amplification One way to treat large numbers of samples is to make a dilution of the restriction enzyme (1: 10 in the manufacturer’s 1X reaction buffer), and then add aliquots of the diluted restriction enzyme to each reaction. In some instances, the restriction enzyme does not work well m the PCR reaction mix The addition of some additional enzyme (a “boost”) after the mitral mcubation can sometimes help complete the digestion. Another approach is to remove half of the amplified fragment (-300 ng) to digest with the addition of 1 pL of undiluted enzyme (1 O-20 U), 1 pL of the manufacturer’s 1OX buffer, and H20 to a final volume equal to the original PCR reaction. Electrophoresis conditions are chosen to provide efficient resolution of the amplified fragment and its digestion products. In general, for larger fragments (> 500 bp) the digested samples and a DNA size standard are electrophoresed on a l-2% agarose gel containing EtBr using 0.5X TBE buffer (0.045 MTris-borate, pH 8 0, 0 001 M EDTA), and visualized with UV light. A 4-8% nondenaturing polyacrylamtde gel (19.1, acrylamide.bis-acrylamide) with 1X TBE can be used for smaller fragments (< 500 bp). The gel is stained with EtBr, and the products visualized with UV light (II)

28

Cole and Hlxson

13 Although shorter fragments allow caster mterpretation of electrophorettc results and easier detection of melting domams, longer fragments can still be analyzed by TGGE. If the longer fragments have an Internal restriction sue, the fragment can be digested and both fragments analyzed on the same gel The mmal determination of the denaturation profile of a fragment may Involve several attempts with dtfferent denaturant concentrattons and temperature ranges before a satisfactory sigmoidally shaped melting curve is obtained One strategy IS to run two perpendicular gradient gels, one wtth 40% and one with 60% denaturant, and both with a temperature gradient of 16-46”C. Based on the results from these gels, one can adjust the temperature range and/or denaturant concentrations to reveal the best denaturing conditions There are a number of computer programs, both freeware (1 e., refs 28 and 29) and commercially available (MacMelt software; Bio-Rad Laboratories), that will predict the denaturatton profile of a DNA fragment based on Its base composition. These programs facihtate decisions about the placement of prtmers and length of fragments to be amplified and analyzed by TGGE Although TGGE has proven to be a sensitive method of detectmg nucleotide substitutions, TGGE is a more difficult method for typing polymorphisms m a large group of samples than digestion with diagnosttc restriction-enzymes It may be beneficial to sequence the differing alleles, and identify a restriction enzyme recogmtion sate that mtght be altered by the substitution responsible for TGGE vartation. 14 In order to resolve the smgle-stranded fragments, the PCR-amphfied product should not be longer than 300 bp It is also important to maintain the secondary structure in the fragments by keepmg the gel cold during electrophorests, and/or using 10% glycerol m the gel, and runnmg at low wattage (32). High resolution gel matrix solutions have been developed that increase the separation of DNA based on conformatlonal differences (Hydrolmk-MDE gels; FMC BioProducts, Rockland, ME) The SSCP procedure previously described uses radioactive nucleotides for labeling PCR products However, several recent protocols have been developed for nonradioactive SSCP analysis, including the use of fluorescently labeled primers, EtBr staining, and silver-staining (for discussions, see refs. 32 and 35-37) 15 A key to typmg hypervariable repeats IS optimization of the PCR reaction, so that specific amplrficatron of the repeat is favored over nonspectftc ampllfication In some instances, bands appear that mirror the repeat polymorphism, but are clearly of the wrong size. These can somettmes actually help dtstmguish alleles In most cases, however, nonspecific amplification can obscure the polymorphic bands and consume PCR reagents, thus interfering with the amplification of the desired alleles One solutton that has been recommended IS transferrmg the gel contammg the amplified products to nylon or mtrocellulose membranes, and hybridtzmg with locus-specific oligonucleotide probes (57) Other suggestions include using touchdown PCR (58) as previously descrtbed (Note 6)

PCR Methodology

29

A common problem in typing dmucleotide repeat polymorphisms IS the stuttering or shadow bands that occur below the true allellc bands These occur at 2-bp increments and are usually owing to slipped-strand mlspalrmg (59) One way to overcome this problem is to add formamide (32%) to the polyactylamide gel (57). 16 In order to take full advantage of the multlplexmg capabilities of the automated system, many PCR reactions of different STRs are pooled for each mdlvldual and loaded mto a single well. The amount of each reactlon that ~111give an optimal signal must be determined empirically If too much of one reaction IS loaded, bleedthrough artifacts may occur which appear as emission peaks m a different color spectrum. The STR sizes are estimated using an internal size standard m each lane The exact size estimated for each allele will vary slightly from lane to lane m a single gel and vary more from gel to gel. Therefore, the original computer-based estlmatlons of the allele sizes must be modified in a process called “binning ” For example, all alleles that range from 99 7 to 100 3 bp m the origmal size estlmatlon will be binned and given a size of 100 bp The next larger allele may range from 101 7 to 102 3 bp, and will be binned at 102 bp This exercise requires empmcal evaluation of the allele size data by the operator, who then establishes the size ranges for each binned allele. Although binning works well wlthm gels, the process often falls for more extensive varlablllty between gels, where increased allehc size ranges cause overlap between bms for different alleles Binning can be further comphcated by the presence of shadow bands (described in Note 15), and the nontemplate addition of single nucleotldes to the 3’ ends of PCR products A number of strategies have been applied to correct bmnmg problems, mcludmg development of computer algorithms (I e , ref. 70), use of longer gels to provide better electrophoretlc separation of alleles, and use of a long final extension to increase the proportion of amplified alleles with an added base Perhaps the best solution IS to use tn- and tetranucleotlde STRs rather than dinucleotide STRs. The size differences are larger between alleles for trland tetranucleotlde STRs, thus avoldmg overlap m size ranges for bmmng of different alleles

Acknowledgments The authors wish to thank Cerise Rieper for her assistance in preparing the chapter. This work was supported by NIH grants HL45522, HL28972, and HL39913. References 1, Saikl, R. K., Scharf, S., Faloona, F., Mullis, K B., Horn, G T., Erltch, H A., and Arnhelm, N. (1985) Enzymatic amplification of P-globin genomlc sequences and restrIction site analysis for diagnosis of sickle cell anemia. Sczence 230, 1350-1354. 2 Innis, M A , Gelfand, D H , Smnsky, J. J., and White, T J (eds.) (1990) PCR Protocols. A Guide to Methods and Appltcatzons Academic, San Diego, CA

Cole and Hixson

30

3. McPherson, M J , Quirke, P., and Taylor, G. R., (eds.) (1991) PCR A Practzcal

Approach Oxford Uruverstty Press, New York. 4 Hallassos, A., Chomel, J. C , Tesson, L., Baudis, M , Kruh, J , Kaplan, J. C , and

Knzis, A (1989) Modification of enzymattcally amplified DNA for the detectton of point mutations. Nuclerc Acids Res 17,3606. 5 Htxson, J. E. and Vernier, D T (1990) Restrtctton tsotypmg of human apolipoprotein E by gene amphfication and cleavage with HhaI J. Llpld Res 3 1, 545-548. 6. Davtgnon, J , Gregg, R. E., and Smg, C F. (1988) Apohpoprotem

8.

9.

10 11 12. 13 14

15 16.

17 18

19.

E polymorphism and atherosclerosis Arterzoscferoszs 8, 1-21. Tenkanen, H and Ehnholm, C (1993) Molecular biology of apohpoprotem AIV. Curr. Oprn Llpldol 4,95-99. McCombs, R. J , Marcadts, D. E., Ellis, J , and Wemberg, R. B. (1994) Attenuated hypercholesterolemic response to a high-cholesterol diet m subJects heterozygous for the apolipoprotem A-IV-2 allele N Engl J Med 331,70&7 10 Mata, P., Ordovas, J. M , Lopez-Miranda, J., Lichtenstem, A H , Clevidence, B , Judd, J T , and Schaefer, E J. (1994) ApoA-IV phenotype affects diet-induced plasma LDL cholesterol lowering. Arterioscler Thromb 14, 884-89 1 Hixson, J E and Powers, P. K. (1991) Restriction isotyping of human apohpoprotem A-IV rapid typing of known tsoforms and detection of a new tsoform that deletes a conserved repeat. J Llpld Res 32, 1529-l 535 Sambrook, J , Frttsch, E F., and Mamatts, T (1989) Molecular Clonzng A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Sprmg Harbor, NY Hobbs, H H., Esser, V , and Russell, D W (1987) AvaII polymorphism m the human LDL receptor gene. Nuclezc Aczds Res 15, 379 Yamamoto, T , Davis, C G , Brown, M S., Schnetder, W. J., Casey, M L , Goldstein, J L., and Russell, D. W (1984) The human LDL receptor a cystemerich protein with multiple Alu sequences m its mRNA Cell 39,27-38. D’Aqmla, R T., Bechtel, L J , Vtdeler, J A , Eron, J J , Gorczyca, P , and Kaplan, J C (1991) Maxtmtzmg senstttvtty and spectftctty of PCR by preampllfication heating. Nuclex Acids Res 19, 3749. Mulhs, K B (1991) The polymerase chain reaction m an anemic mode. how to avoid cold ohgodeoxyribonuclear fusion PCR Methods Appl 1, l-4. Don, R. H., Cox, P. T., Wamwrtght, B J , Baker, K , and Matttck, J S (1991) Touchdown PCR to circumvent spurious priming during gene amplification Nucleic Acids Res 19,4008. Hecker, K H. and Roux, K H (1996) High and low annealing temperatures increase both specificity and yield m touchdown and stepdown PCR Biotechniques 20,478485 Fischer, S G. and Lerman, L S (1983) DNA fragments differing by single basepair substttutions are separated m denaturing gradient gels: correspondence with melting theory Proc Nat1 Acad Scz USA 80, 1579-1583. Wartell, R. M., Hosseini, S. H., and Moran, Jr , C P (1990) Detecting base pair substitutions m DNA fragments by temperature-gradient gel electrophorests Nucleic Acids Res 18,2699-2705

PCR Methodology

31

20. Ortta, M., Suzukt, Y., Sektya, T., and Hayasht, K. (1989) Rapid and sensitive detection of point mutations and DNA polymorphtsms using the polymerase chain reaction. Genomlcs 5, 874-879. 2 1 Sheffield, V. C , Cox, D. R., Lerman, L. S., and Myers, R. M. (1989) Attachment of a 40-base-pair G+C-rich sequence (GC-clamp) to genomtc DNA fragments by the polymerase chain reaction results in improved detection of single-base changes Proc Nat1 Acad Sci USA 86,232-236. 22. Hence, K , Harders, J., Wiese, U , and Riesner, D. (1994) Temperature gradient gel electrophoresis (TGGE) for the detection of polymorphic DNA and RNA, Mol Btol m Protocols for Gene AnaZysls (Harwood, A. J., ed ), Humana, Totowa, NJ, pp. 2 1l-228 23 Thtbonnier, M. and Schork, N J. (1995) The genetics of hypertension. Curr Opm Genet Dev 5,362-370. 24 Jeunemattre, X., Soubrter, F , Kotelevtsev, Y V., Lifton, R. P., Wtlllams, C S., Charm, A., Hunt, S C., Hopkins, P. N., Williams, R R., Lalouel, J.-M., and Corvol, P (1992) Molecular basis of human hypertension role of angtotensmogen Cell 71, 169-180. 25 Hixson, J. E. and Powers, P K (1995) Detection and charactertzation of new mutations m the human angiotensmogen gene (AGT) Human Genet 96, 110-I 12. 26. Collins, T , Willlams, A., Johnston, G. I., Kim, J., Eddy, R , Shows, T , Gtmbrone, Jr., M. A., and Bevilacqua, M. P (199 1) Structure and chromosomal location of the gene for endothehal-leukocyte adhesion molecule 1 J Bzol Chem 266, 2466-2473. 27 Powers, P. K. and Hixson, J. E. (1993) Ban1 and PvuII polymorphisms

28.

29

30 31

32. 33 34 35

m mtron 2 of selectin E (SELE). Human A401 Genet 2, 1082. Lerman, L. S. and Silverstem, K (1987) Computational simulation of DNA meltmg and its application to denaturing gradient gel electrophoresis. Methods Enzymol. 155,482-501. Steger, G. (1994) Thermal denaturation of double-stranded nucleic acids: prediction of temperatures critical for gradtent gel electrophorests and polymerase cham reaction. Nuclex Acids Res 22,2760-2768. Fischer, S G and Lerman, L. S. (1979) Length-independent separation of DNA restriction fragments m two-dtmensional gel electrophorests. Cell 16, 19 l-200. Parker, S., Angelico, M. C , Laffel, L., and Krolewski, A. S. (1993) Apphcation of denaturing gradient gel electrophoresis to detect DNA sequence differences encoding apohpoprotein E isoforms. Genomics 16,245-247 Hayashi, K. (1991) PCR-SSCP: a simple and sensitive method for detectton of mutations in the genomic DNA. PCR Methods Appl 1,34-38. Utermann, G (1989) The mysteries of lipoprotein(a). Science 246,904-9 10 Cohen, J. C., Chiesa, G., and Hobbs, H. H. (1993) Sequence polymorphisms m the apolipoprotem(a) gene J Clin Invest. 91, 163&1636 Hongyo, T., Buzard, G. S., Calve& R. J., and Weghorst, C. M. (1993) Cold SSCP. a simple, rapid and non-radtoactive method for opttmized single-strand conformation polymorphism analyses Nucletc Acids Res. 21, 3637-3642.

32

Cole and Hixson

36 Vtdal-Puig, A and Moller, D. E (1994) Comparative sensmvity of alternative single-strand conformation polymorphtsm (SSCP) methods Bzotechnzques 17, 490-496. 37. Iwahana, H , Adzuma, K , Takahashi, Y., Katashima, R., Yoshimoto, K., and Itakura, M. (1995) Multiple fluorescence-based PCR-SSCP analysis with postlabeling PCR Methods Appl. 4,275-282 38 Wallace, R B., Shaffer, J , Murphy, R F., Bonner, J., Hirose, T , and Itakura, K ( 1979) Hybridization of synthetic oligdeoxyrtbonucleotides to 4x-O 174 DNA the effect of single base patr mismatch Nucleic Acids Res 6,3543-3557 39 Salki, R. K., Bugawan, T L , Horn, G T , Mulhs, K B , and Erhch, H. A. (1986) Analysts of enzymatically amphfied P-globin and HLA-DQa DNA with allelespecific ohgonucleotide probes. Nature 324, 163-166 40. Weisgraber, K H., Newhouse, Y. M , and Mahley, R W. (1988) Apohpoprotem E genotypmg usmg the polymerase chain reaction and allele-specific ohgonucleottde probes, Brochem Bzophys Res Commun 157, 1212-1217. 41 Smeets, H J M , Poddighe, J , Stuyt, P M. J., Stalenhoef, A. F. H , Ropers, H H , and Wiermga, B (1988) Identification of apolipoprotem E polymorphism by using synthetic oligonucleottdes J Lipid Res 29, 123 1-1237 42 Houlston, R. S., Snowden, C , Green, F , Alberti, K. G M M , and Humphries, S E. (1989) Apohpoprotem (apo) E genotypes by polymerase chain reaction and allele-specific ohgonucleottde probes no detectable linkage disequilibrmm between apo E and apo CII Human Genet. 83,364-368 43 Innerartty, T L , Mahley, R W , Wetsgraber, K. H , Bersot, T. P., Krauss, R M , Vega, G L , Grundy, S M., Frtedl, W , Davignon, J., and McCarthy, B. J (1990) Familial defective apohpoprotem B-100. a mutation of apohpoprotem B that causes hypercholesterolemia J Llpzd Res 31, 1337-1349 44. Ruztcka, V , Marz, W , Russ, A., and Gross, W. (1992) Apolipoprotem B(Arg3500+Gln) allele specific polymerase chain reaction. large-scale screening of pooled blood samples. J Lipzd Res 33, 1563-l 567. 45 Hansen, P S , Rudtger, N , TybJaerg-Hansen, A., Faergeman, O., and Gregersen, N (199 1) Detection of the apoB-3500 mutation (glutamine for arginine) by gene amphfication and cleavage with A4spI J Llpld Res 32, 1229-1233 46. Syvanen, A. -C., Aalto-Setala, K., Kontula, K., and Soderlund, H. (1989) Dtrect sequencing of affimty-captured amphfled human DNA apphcation to the detection of apohpoprotem E polymorphism FEBS Lett 258, 7 l-74. 47 Sanger, F., Nicklen, S., and Coulcon, A. R. (1977) DNA sequencing with chamtermmatmg mhtbitors Proc Nat1 Acad Scl USA 74, 5463-5467. 48 Syvanen, A. -C , Aalto-Setala, K., Harm, L., Kontula, K., and Soderlund, H (1990) A primer-guided nucleotide mcorporation assay m the genotyping of apohpoprotein E Genomlcs 8,684--692 49 Zuliani, G. and Hobbs, H. H. (1990) Dmucleotide repeat polymorphism at the 3’ end of the LDL receptor gene. Nucleic Acids Res. 18,430O. 50. Zuliani, G. and Hobbs, H. H. (1990) Tetranucleotide repeat polymorphism m the apohpoprotem C-III gene Nuclezc Aczds Res l&4299.

PCR Methodology

33

51. Zuham, G and Hobbs, H H. (1990) Tetranucleotide repeat polymorphtsm m the LPL gene. Nuclezc Acids Res l&4958. 52 Knott, T J , Wallis, S C , Pease, R J , Powell, L M , and Scott, J (1986) A hypervarlable region 3’ to the human apohpoprotem B gene. Nucleic Acids Res. 14,9215-9216

53 Huang, L.-S and Breslow, J L (1987) A unique AT-rich hypervarlable mmlsatellite 3’ to the apoB gene defines a high information restrtctlon fragment length polymorphtsm J BloI Chem 262,8952-8955. 54. Zuham, G and Hobbs, H. H. (1990) Tetranucleotide repeat polymorphtsm in the apohpoprotein B gene Nucleic Acids Res l&4299. 55 Ludwig, E H., Haubold, K., and McCarthy, B. J. (199 1) Analysts of two dlfferent tandem repetmve elements within the human apohpoprotem B gene J Lipzd Res 32,374-379. 56 Hixson, J E , Powers, P K , and McMahan, C A (1993) The human apolipoprotem B 3’ hypervariable region: detection of eight new alleles and comparisons of allele frequencies in blacks and whites. Human Genet 91,475-479 57 Lttt, M , Hauge, X , and Sharma, V. (1993) Shadow bands seen when typing polymorphtc dmucleotrde repeats. some causes and cures Biotechnlques 15, 28&284 58 Mellersh, C. and Sampson, J (1993) Stmphfymg detection of mtcrosatelhte length polymorphisms. Bzotechniques 15,582-584 59. Hauge, X Y and Lttt, M (1993) A study of the ortgm of ‘shadow bands’ seen when typing dmucleotrde repeat polymorphisms by the PCR Human A401 Genet 2,411-415 60. Paterson, A. H., Lander, E. S., Hewitt, J. D., Peterson, S , Lmcoln, S E , and Tanksley, S. D. (1988) Resolutton of quantitative traits mto Mendehan factors by using a complete linkage map of restriction fragment length polymorphtsms Nature 335,72 l-726 61. Lander, E. S. and Botstem, D. (1989) Mapping mendelian factors underlying quantttattve traits usmg RFLP linkage maps Genetzcs 121, 185-l 99 62. Cooperattve Human Linkage Center (CHLC): Murray, J C., Buetow, K. H., Weber, J. L., Ludwtgsen, S., Scherpbrer-Heddema, T., Mamon, F , Qurllen, J , Sheffield, V C , Sunden, S., Duyk, G. M.; Genbthon: Welssenbach, J., Gyapay, G , Dtb, C., Morrissette, J , Lathrop, G. M , Vtgnal, A , Umverslty of Utah White, R., Matsunamt, N., Gerken, S., Mehs, R., Albertsen, H , Plaetke, R., Odelberg, S ; Yale University Ward, D.; Centre d’Etude du Polymorphrsme Humain (CEPH): Dausset, J., Cohen, D., and Cann, H. (1994) A comprehensrve human linkage map with centtmorgan density. Science 265,2049-2054 63. Weber, J L and May, P E. (1989) Abundant class of human DNA polymorphisms which can be typed using the polymerase chain reaction Am J Human Genet 44,388-396 64 Edwards, A., Clvltello, A., Hammond, H A, and Caskey, C T (1991) DNA typing and genetic mapping wrth trtmerlc and tetramerrc tandem repeats Am. J Hum Genet 49,746756

34

Cole and Hixson

65. Mansfield, D. C , Brown, A. F., Green, D. K., Carothers, A. D., Morris, S. W , Evans, H. J., and Wright, A. F. (1994) Automation of genetic linkage analysts using fluorescent mrcrosatelhte markers. Genomlcs 24,225-233 66 Reed, P W., Davies, J L., Copeman, J. B., Bennett, S. T., Palmer, S. M , Pritchard, L E., Gough, S. C L , Kawaguchr, Y., Cordell, H J., Balfour, K. M., Jenkins, S. C., Powell, E. E., Vignal, A., and Todd, J. A. (1994) Chromosomespecific mtcrosatellite sets for fluorescence-based, semr-automated genome mappmg Nat Genet 7,390-395. 67 Schwengel, D. A , Jedlicka, A. E , Nanthakumar, E. J , Weber, J L., and Levitt, R. C. (1994) Comparrson of fluorescence-based semt-automated genotypmg of multiple mtcrosatellite 1oc1 wtth autoradiographtc techmques Genomzcs 22, 46-54. 68 Weissenbach, J., Gyapay, G., Drb, C., Vignal, A , Mortssette, J., Mtllasseau, P , Vaysseix, G , and Lathrop, M (1992) A second-generatron lmkage map of the human genome. Nature 359,794-801 69. Scheftield, V C., Weber, J L., Buetow, K. H., Murray, J C , Even, D A , Wiles, K., Gastter, J. M., Pulido, J. C , Yandava, C., Sunden, S L , Mattes, G., Busmga, T , McClam, A., Beck, J , Scherprer, T , Grlliam, J , Zhong, J., and Duyk, G M (1995) A collectron of trt- and tetranucleottde repeat markers used to generate high quality, high resolutron human genome-wade lmkage maps. Human A401 Genet 4, 1837-1844 70. Perlin, M. W., Lancta, G., and Ng, S. -K. (1995) Toward fully automated genotypmg. genotyping mrcrosatellite markers by deconvolutton. Am J Human Genet 57, 1199-1210

Separation of Small-Size DNA Fragments Using Agarose Gel Electrophoresis Jose M. Ordovas 1. Introduction Submarine horrzontal agarose gel electrophoresis has been a workhorse for the molecular brologrst. Initrally, it was primarrly used as a quack check for many drfferent molecular biology procedures, and for the separation of largesize nucleic acid molecules (1); however, agarose gel electrophorests 1s now becoming a tool of choice to separate small DNA-size fragments when the difference between the fragments of interest IS greater than l%, a task that has traditionally been performed using polyacrylamide gel electrophoresrs. The characterrstics of some newly available agarose permits one to carry out DNA separations comparable to those obtamed using <8% acrylamide gels (2,5). The reasons for the success of agarose come from their easy handlmg, economy, versatility, lack of toxicity, and speed (I,.?). This chapter describes a detailed protocol to separate DNA fragments between 50 and 250 bp using Metaphor agarose and TBE running buffer. The steps described can be easily adapted to other DNA size ranges, agarose types, and running buffers. Alternative protocols using vertical agarose gel electrophoresls are less commonly used m the laboratory and they will not be discussed here (61, 2. Materials 2.1. Equipment 1. Microwave oven. 2. Hot plate with stirrer 3. Horizontal electrophoresis chamber. Many supphers and models are available, these specific units can be obtained from Bra-Rad (Hercules, CA) and are ltsted accordmg to their increase sample number capacity. Mmt-Sub Cell GT, Wade From Methods in Molecular Rology, Vol 7 IO Llpoprotem Protocols Edited by J M Ordovas 0 Humana Press Inc , Totowa, NJ

35

Ordovas

36

Mint-Sub Cell GT, Sub-Cell GT DNA Electrophoresis Cell, Sub-Cell Model 96 Cell, and Sub-Cell Model 192 Cell. Well-forming combs are also available m several sizes allowmg the runnmg of as few as 8 samples or as high as 51 per comb. The most recently designed combs (available for Sub-Cell models 96 and 192) are compatible with the use of multichannel plpets to load the gels Gel caster (Bto-Rad) (see Note 1) PowerPac 300 Power supply (Bto-Rad) Rectrculator-chiller water bath (Lauda Digital Refrigerated Ctrculatmg Water Bath, Brinkmann RM Series, VWR Sctentific Products, Bridgeport, NJ) Gel Dot 1000 Video Gel Documentatton System, Bra-Rad, or Foto/System 1000 Camera System, Fotodyne (Hartland, WI) For low-volume laboratories, the Polarotd Gel documentation system (Bto-Rad) constitutes a feastble and more economical alternative Glass or clear polypropylene stammg contamers (VWR Scientific Products, Bridgeport, NJ) 2.2. Reagents 1 Metaphor agarose (FMC BtoProducts, 2 Ethidium bromide (EB) stock solution St Louts, MO) m 100 mL of dtsttlled totally m solutton Keep at room

Rockland, ME) (10 mg/mL)* 1 g of EB (Sigma no E7637, water. Stir for several hours until the dye is temperature m a dark glass container

(see Note 2).

3 SYBR Green I nucleic-acid gel stam (FMC Btoproducts, cat no 505 12) 4 5X TBE buffer 0 45 A4 Trts-borate, pH 8 0, 0 01 M ethylenediamtne tetra-acetic actd (EDTA), 54 g Tris-base, 27.5 boric acid, and 3 72 g NazEDTA * 2H20 AdJust to 1 L with distilled water Dilute the stock to 1X TBE (89 mA4 Tris-borate, 2 m/r4 EDTA) or 0 5X TBE (45 mM Trisborate, 1 nu’r4 EDTA) for use 5. 10X Ficoll-gel loadmg solution. 25% Ficoll (type 400, Sigma no F4375) m dtstilled water; 0 4% bromophenol blue (BPB) (w/v), and 0 4% Xylene cyan01 (they can be obtained as a dye mtxture from Sigma #B3269). Keep at room temperature (see Note 3)

6. Whatman no 1 filter paper (VWR Sctenttfic Products) 7. Powered activated charcoal (Sigma no. C5260) 8. 5% hypophosphorus acid: Dilute 1: 10 with water using the commerctally avatlable 50% solution (J T Baker, Purified Grade no. JT0178-1 VWR Sctenttfic Products). 9 0.5 MSodium mtrtte: Dtssolve m water 3 45 g of sodium nitrite to a final volume of 100 mL, prepare freshly before use (Sigma no. S2252). 10 1 M Sodium bicarbonate 84 g/L of distilled water (Sigma no S 6014). 11 DNA Markers for nucleic-acid gel electrophoresis. The following size standards are recommended for size asstgnment within the range described m this protocol pBR322/HaeIII (22 fragments, 8-587 bp, Sigma no. D8397); pUC 18IHaeIII (11 fragments, 1 l-587 bp, Sigma D 6293), 20-bp ladder (20 bp to 1 kb, Invttrogen, no R340-0 1, Carlsbad, CA).

DNA Separation

3 % Metaphor

4 Agarose

5

Fig. 1. Schematic representation of the optimal separation range (bp) for different concentrations of Methaphor agarose using 0.5X TBE as running buffer (shaded vertical bars). The black horizontal bar represents the mobility of the dye Xylene cyan01 and the gray bar represents the mobility of the dye bromophenol blue. The y-axis is in log scale.

3, Fine Resolution of DNA Fragments Less Than 1 kb Using Metaphor Agarose 3.1. Gel Preparation The protocol described uses an agarose concentration of 3% in TBE buffer. This combination is optimal for the separation of DNA fragments in the 50- to 250-bp range (the most common range reported in the literature for DNA markers). For other size ranges, see Fig. 1 for optimal agarose concentrations. This protocol illustrates the preparation of standard-size mini gels, the same principles apply to the preparation of larger gel sizes(see Note 4). 1. Weight 1.5 g of Metaphor agarose. 2. Add 50 mL of 1X TBE electrophoresis buffer into a 200-mL beaker. coated magnetic stir bar on the beaker and place it over a magnetic 3. Stir the buffer at high speed and add slowly the agarose powder solution. Continue the stirring for 1-2 min to allow the agarose room temperature.

Put a Teflonstirrer. to the buffer to hydrate at

Ordovas 4. Cover the beaker with plastic wrap and punch several small holes m it using a needle Use a marker to place a signal on the wall of the beaker mdtcatmg the mitral level of buffer. 5. Heat the agarose suspension m a microwave oven until bubbles start to appear (see Note 5). 6. Remove the beaker from the oven and swirl the suspension slowly. 7. Place the beaker again m the microwave and boil the solution for 1 mm 8 Repeat step 6 and check that all agarose particles are m solution. 9 Add enough distilled water (prewarmed at 6O’C) to compensate for the lost of volume owing to boiling Use as a reference the mark from Subheading 4. 10 At this time add ethidium bromide to a final concentration of 0 5 pg/mL (see Note 6). 11. Mix thoroughly to avoid differences m the concentration of agarose and dye between different regions of the gel once it solidifies 12 Place the beaker m a water bath at 60°C to cool the solution prior to casting the gel 13. Prepare the gel-casting tray on a level surface 14 Verify that the combs are perfectly clean and that no dried agarose residues are present from previous gel castings 15. Place the comb over the tray and check that the clearance between the bottom of the teeth and the surface of the tray is between 0 5 and 1 mm. Remove the comb 16 Pour the agarose solutton into the tray and replace the comb (see Note 7). 17 Leave the agarose to gel at room temperature for at least 30 mm. Once the agarose is solid, place the gel for an additional 30 mm at 4°C. 18. After the gel is ready, place the gel-containing tray mto the electrophoresis chamber and cover the gel with running buffer (use 0.5X TBE containing 0.5 pL/mL of ethidmm bromide, to a level approx l-2 mm above the gel surface) 19 Remove the comb very slowly (see Note 8) 20 If needed, add 0.5X TBE running buffer containing EB to the chamber until its level reaches approx 2-3 mm over the surface of the gel (Note 9)

3.2. Sample Preparation

and Electrophoretic Separation A gel-loading solution is always added to the samples containing the DNA fragment(s). The purpose of this step is twofold: First, it increases the density and adds color to the sample, thus simplifying the loadmg of sample into the gel; second, it allows to follow the progression of the electrophoresis owing to the known mobility of the dyes used m the loading solutron. 1 Mix 10 pL of sample (see Note 10) with 1 pL of 10X Fmoll 10X loading solutton (see Note 3) 2. Load the samples mto the well usmg a micropipetor Make sure that the tip of the prpet does not reach the bottom of the well because this may puncture the gel, resulting m sample lost (see Note 11) 3 Place the hd over the electrophoresis chamber (remember that the DNA migrates toward the anode [red lead] and that the wells containing the samples should be

DNA Separation

39

closer to the cathode [black leads]). Run the electrophoresis at 5 V/cm and follow the progression of the bromophenol blue (fast-movmg band, which m the conditions described m this protocol, has an electrophorettc mobility equivalent to approx 30 bp; see Fig. 1) (see Note 12). Stop the run by turning off the power supply and disconnectmg the leads when the BPB dye reaches approx 1 cm above the bottom of the gel. 3.3. DNA Defection EB is by far the most common fluorescent dye used to visualize DNA

fragments separated by agarose gel electrophoresis. The limit of DNA detection using this dye IS approx 10 ng per band (see Note 2). 1. Remove the gel from the electrophoresis chamber and the castmg tray (some trays are UV transparent and the gel does not need to be removed from the tray) and Immerse it m distilled water for 15 min to wash excess EB 2 Dtscard the solution and repeat the washing procedure again (see Note 2) 3 Vtsuahze the bands on a transillummator and create a permanent record of the results using any of the gel documentation systems currently avatlable 4. Notes 1 Tradttionally, the casting of agarose gels was carried out sealing the open ends of the gel trays with autoclave tape and pouring the gel solution into the mold. Once the agarose solidified, the tape was removed and discarded Nowadays, gel casters permit a more convenient and leak-free system to cast agarose gels. 2. EB IS a potent mutagen Extreme care should be taken when preparmg the solutions. Safer alternatives are to purchase pre weighted EB tables or previously prepared IO-mg/mL solutions (Bio-Rad or Sigma). Gloves should be always used when handling any material containing EB, such as running buffers, gels, and discarded washing solutions. All EB containing solutions should be decontaminated before disposal For concentrated solutions of EB, this may be carried out as follows: a. Dilute the EB solution to less than 0.5 mg/mL with water b. Add 0 2 volumes of 5% hypophosphorus acid and 0.12 volumes of fresh 0 5 h4 sodium nitrite. It is important to mamtam an acidic pH below 3.0. c. Leave the mixture for 1 d at room temperature and add 3-4 volumes of 1 A4 sodium bicarbonate, then discard the solution For those solutions containing 50.5 pg/mL of EB, such as electrophoresis running buffer or destaining washes, the followmg procedure 1srecommended. a Add 1 g of powered activated charcoal/L of EB-containing solution. b. Leave 1 h at room temperature with frequent swirlmg of the container to expose the solution to the charcoal c Filter the solution through a Whatman no, 1 filter paper and discard the filtrate d. Take the filter contammg the charcoal, place it m a sealable plastic bag, and dispose of tt m the hazardous waste

40

Ordovas Several other stammg alternatives are available to visualize DNA bands m agarose gels. The fluorescent dye SYBR Green I may be up to 100 times more sensttive than EB. Furthermore, this dye fluoresces only after binding the DNA, thus decreasing the background stain observed with EB. This effect is specially problematic m gels with a high percent of agarose at those described here. On the other hand, the SYBR Green I stam requires dedicated polypropylene contamers; tt IS more ltmited m its diffusion mto the gels, limitmg their thickness This dye requtres the use of photographic filters (SYBR Green photographic filter, or Wratten no 15 gelatm filter) and transillummators different to those used for EB. Several other loading solutions are commonly used containing sucrose or glycerol instead of Ficoll; however, they result in more diffuse and U-shaped DNA bands. Moreover, glycerol interacts with the TBE running buffer altering its buffering capacity The volume needed to prepare a 3-mm-thick standard mmigel is 45 mL. For gel trays designed for a higher sample throughput, such as the Sub-Cell Model 96 or Model 192 cells, the typical volumes are 75 and 112 mL An alternative to the microwave oven IS the more traditional hot plate with stirrer The only problem associated with this method of solubthzmg the agarose IS that requtres a much longer time to prepare the gels Sharper bands and consequently increased resolution can be achieved by running the electrophesis wtthout EB m both the gel and the runnmg buffer If this alternattve is selected, then, followmg the electrophoretic separation, the gel should be nnmersed m a solution contammg 0 5 pg/mL of EB m disttlled water during 2@40 mm, depending on the thickness of the gel. Afterwards, the gel should be washed with distilled water as descrtbed m Subheading 3.1.3. The agarose gel should not be more than 4 mm thtck Thtcker gels will seriously compromise Its resolution, specially for the smaller DNA fragments Moreover, thicker gels result on increased background staining if using EB stain, or poorer staining if usmg SYBR Green I stain, resultmg m both cases m reduced sensittvity It IS important to cover the surface of the gel with runnmg buffer prior to removmg the combs to prevent damage of the wells. The high concentration of agarose used in this protocol results in gels that are more brittle than normal and more prone to breakage. It IS recommended that the comb be lifted at an angle rather than completely horizontal This helps the entrance of buffer into the wells and prevents the formation of a partial vacuum while hftmg the comb, which could result m a leaky well. The running buffer has to be high enough to cover the surface of the gel, mcludmg the meniscus that usually forms around the wells (protrudmg about 1 mm over the surface of the gel); however, excess of buffer should also be avoided because this will reduce the voltage gradient gomg across the gel, resultmg m lower DNA mobility and loss of band sharpness. A good compromise is to have about 2 mm of runnmg buffer covering the entire surface of the gel.

DNA Separa t/on

41

Another factor to take mto constderatton is buffer depletion. Signals of buffer depletion are* gel meltmg (overheating may be detected by excesstve or uneven water condensatton on the chamber lid) and smearing of the DNA bands Another hmt of this problem may be the change of color of the BPB to yellow mdtcatmg excesstve actdtficatton of the anodic chamber Recirculatton of the buffer reduces its depletion A 0.5X TBE buffer can be used for up to 50 watt-hours before replacement when usmg a large stze electrophoretic chamber (-2 L total volume capacity), but only up to 15 watt-hours when using the tradtttonal mmigels. 10 The most common use of agarose gel electrophoresis is the assessment of PCR amphficatton products. The amount of DNA IS rarely assessed prior to loading mto the gel and usually a 10 pL load may contain around 100 ng of DNA, which is adequate for detection using EB. However, remember that the number of fragments and their size distribution are important factors to take into constderation Few larger fragments tend to overload the capacity of the gel faster that several small fragments, resulting m poor resolution. Conversely, small fragments have less EB chromogemctty and this may compromtse their detection. Each laboratory should determme empmcally the best compromise between volume (amount) of DNA loaded, well volume, and gel thtckness to achieve optimal results for each spectfic application 11 Remember to run DNA size markers on all gels. A variety of DNA ladders are available from different vendors that allow the precise asstgnment of size for the DNA fragments present m the samples. 12 When runnmg time IS a maJor factor, the electrophorests can be carried out at higher voltages (15-l 7 V/cm) Moreover, when used in combmatton with Metaphor as separation medium and TBE as running buffer, this higher voltage contributes to increased band sharpness and better resolution; however, the buffer needs to be maintained at -15’C during the run to prevent the gel from melting This 1s achieved usmg a recirculator water bath to recirculate cold electrophoreSIS buffer through the electrophoresis chamber. It 1s essential not to start the rectrculatton until the samples have entered the gel, otherwise the flow of a rapid moving buffer will remove the samples from the gel

Acknowledgments This work was supported by grants HL54776 from the National Institutes of Health and contract 53-K06-5-10 from the US Department of Agriculture Research Service.

References 1 Sugden, B , De Troy, B , Roberts, R. J., and Sambrook, J. (1975) Agarose slab-gel electrophorests equipment. Anal Blochem 68,36-46 2. Hoch, H. and Lewallen, C. G. (1977) High concentration agarose gel. a new medium for high resolution electrophoresis. Anal Blochem. 78,3 12-3 17.

Ordovas

42

3. Smith, D. R (1996) Agarose gel electrophoresis. Methods A401 Bzol 58, 17-2 1 4 Upcroft, P and Upcroft, J A. (1993) Compartson of properties of agarose for electrophoresis of DNA. J Chromat 618,79-93 5 Calladme, C. R., Collms, C M., Drew, H. R , and Mott, M R (1991) A study of electrophorettc mobthty of DNA m agarose and polyacrylamtde gels. J Mol. Bzol 221,981-1005.

6 Pascali, V L., Pescarmona, M , and Dobosz, M., and D’AloJa E (1991) Efficient, Small scale electroelutton of hrgh molecular weight DNA from agarose gels by a miniature vertical electrophoresis cell Electrophoreszs 12, 3 17-320

3 Quantification of mRNA by Polymerase Chain Reaction (PCR) Using an Internal Standard and a Nonradioactive Detection Method William H. Karge III, Ernst J. Schaefer, and Jose M. Ordovas 1. Introduction Quantification of the level of expression for a specific gene m samples obtained during various experimental conditions is becoming increasingly important as attempts are made to gam knowledge relating to how these altered conditions affect cells at the molecular level. Traditionally, an estimate of the copy number of a gene transcription product in a sample was made by employmg traditional Northern blot, dot-blot hybridization, or solution hybridization techniques. These techniques are labor-intensive and usually require microgram quantities of total RNA in order to detect specific mRNAs. The low sensitivity of these assaysmakes it difficult to estimate differences m expression less than fourfold. In addition, genes with low levels of expression often cannot be detected at all by these assays. Polymerase chain reaction (PCR) 1sbeing employed as a feasible alternative to determine mRNA expression for specific genes, PCR offers the advantage of greatly increased sensitivity over the traditional methods previously used to estimate the level of expression for specific genes. The increased sensitivity of PCR allows for the detection of mRNA transcripts present in low copy numbers followmg reverse transcription (RT). Initial attempts to quantitate PCR amplification of mRNA for specific genes employed the use of genes that were thought to vary little with experimental conditions, such as p-actin, or an unrelated sequence as an internal standard (1). Due to potential differences in amphfication efficiency of the primers for the standard and the gene of interest, these techniques achieved qualitative rather than quantitative results. From

Methods m Molecular Edited by J M Ordovas

B/ology, Vol 110 Lfpoprotem Protocols 0 Humana Press Inc , Totowa, NJ

43

44

Karge, Schaefer, and Ordovas

Wang, Doyle, and Mark (2) described a method m which a synthetic RNA 1s used as an internal standard for quanttfymg the amount of a specific mRNA by PCR. The target mRNA IS coamplified m the same tube with a synthetic RNA standard. The standard uses the same primer sequences as the target mRNA, but the PCR yields a different stze product from the standard sequence. The two products can be separated easily by standard nondenaturmg polyacrylamlde gel electrophoresls (PAGE). Wang et al. used radrotsotopes with high specific activity as a label to detect the product from then PCR assay. Although this label produces an assaywtth high sensrttvrty, tt mtroduces problems related to radiation safety, hazardous waste disposal, and the use of a label with a short half-life. Powell and Kroon (3) adapted the method of Wang et al. (2) for use of a nonradloacttve label. Detection and quanttficatlon of PCR products required the mcorporatton of a dtgoxigenm label, Southern blotting, and the use of a chemtlummescence detectton system We modttied both methods to use a standard silver staining techmque (4). Neither the mcorporatton of a radloactive nor a nonradroactrve label or blotting 1s required for detection and quantification of products. Size differences between native and standard RNA products allow for easy dtscrlmmatlon and quantlticatton on gels stained wtth silver nitrate. 2. Materials

2.1. Equipment 1 Tissue homogenizer (for soft tissues, such as hver, use a Potter-ElvehJem Tissue Grinder, Wheaton VWR cat no 62400 For connecttve tissues use a PotterElvehJem Tissue Grinder with glass Pestle, Kontes, VWR cat no KT885500) 2 Balance (Model AC 100, Mettler-Toledo, OH) 3 Weighing boats 4 Sterile scissors 5 Sterile tweezers 6. Sterile ~-CC glass pipets. 7 Sterile 1~-CC polypropylene tubes 8. Eppendorf 1.5-mL tubes 9 Microcentrifuge 10 Vortex. 11. Kimwipes 12. Micropipetes (P20, P- 100, P200- 1000, Ramm, Wobum, MA) 13. Spectrophotometer (GeneQuant II, Pharmacia Biotech, Piscataway, NJ) 14 Electrophoresis chamber, minisub cell (cat. no. 170-4307 Bio-Rad, Hercules, CA) 15 Power supply, 25-200 V (Model 20012 0 Bto-Rad) 16 Thermocycler (DNA Thermal Cycler II, Perkm-Elmer Cetus, Norwalk, CT) 17 Thin-walled tubes for PCR, 0.2 or 0.5 mL (Perkm-Elmer)

Quantification

of mRNA by PCR

45

18 Glass plates (20 x 20 cm, 1 5-mm spacers, Bio-Rad) 19 Electrophorests chamber, vertical (Model Protean II XI, Bio-Rad). 20. Laser densitometer (LKB 2202, LKB Instruments, Paramus, NJ, with LKB 2400 GSXL software for mtegratton of peak areas)

2.2. Reagents 1. TRIZOL (cat no 18038-042 Gtbco-BRL, Grand Island, NY). 2 Ltqutd Nttrogen 3 DEPC water (Prepare under a fume hood) Keep DEPC on ice and add 0.1 mL of DEPC per 100 mL of dtsttlled water Leave overnight with the cap open and autoclave the next day Autoclavmg is critical to destroy the DEPC, which can degrade RNA. Store at 4°C DEPC-water should be used to bring reagents to required volumes 4 0 1% DEPC solutton (prepare under hood). Keep DEPC on rce and add 1 mL DEPC per 1 L distilled water Do not autoclave Used for treating glassware, electrophorests chamber, and so on Store at room temperature, m a fume hood 5 GITC solutton (tissue guanidmm solutton): MIX 59 08 g guamdtum isothiocyanate, 2 0 mL sterile 0 5 A4 EDTA (5 mM), and 2 5 mL sterile 2 M TrisHCl, pH 7 5 Brmg volume to 95 mL with DEPC-water Heat at 65°C until dissolved Sterilize by filtratton with Nalgene disposable filter (0 2 pm). Add 5 mL of P-mercaptoethanol. Do not autoclave. Store at 4°C m a dark bottle. 6 Chloroform 7 Isopropanol. 8 Ethanol, 100,75,70, and 10% (v/v) Sterthze by filtermg with nonaqueous filter (0 22 pm) and syrmge mto a 15-mL tube and store at 4°C 9 3 A4 Sodium acetate (cat no S-2889, Sigma, St LOUIS, MO). Hydrated Na acetate 40.8 g m 100 mL DEPC-water. Autoclave after preparation and after each use Store at 4°C 10 Agarose, ultra pure (cat. no. 55 10 UB Gibco-BRL) 11 2 A4 Tris-HCl, pH 7 5 Dissolve 24.2 g Trizma base in 100 mL DEPC-water Autoclave after preparation and after each use Store at 4°C 12. 0 5 M NazEDTA, pH 8.0. Dissolve 18.61 g Na*EDTA m 100 mL DEPC-water Autoclave after preparation and after each use. Store at 4°C 13 50X TAE buffer: 2 A4 Tris Base, 2 A4 glacial acetic actd, and 50 mM EDTA MIX 242.2 g Trts base, 57.1 mL glacial acetic acid, and 18.6 g EDTA. AdJust to 1 L with dtsttlled water Dilute I:50 to make the 1X workmg solutton. The pH should be 8 18 If this 1snot the final pH, use acetic acid for pH admstment and brmg to final volume wtth DEPC-water. Autoclave and store at room temperature m a dark bottle. 14 1X TAE + EB (Note 1): Add 50 pL ethidmm bromtde (cat no E 15 10, Sigma) per 1 L sterile 1X TAE Store at room temperature m a dark bottle 15 1X TE, pH 7 5-8 0 Dissolve 1.576 g Trts-HCl (10 rmW) and 0 336 g Na,EDTA (1 n-&Q m DEPC-water (-900 mL), adjust the pH (between 7.5 and 8 0) Bring volume to 1 L with DEPC-water. Autoclave after preparation and after each use Store at 4°C.

46

Karge, Schaefer, and Ordovas

16 RNA loading buffer (cat no G-7654, Sigma) 17 GeneAmp Thermostable rRTth Reverse Transcriptase Kit (cat. no N808-0069, Perkm-Elmer). 18. Primers (Notes 2-5). 5’ LDL-receptor 5’-CAATGTCTCACCAAGCTCTG-3’ 3’ LDL-receptor 5’-TCTGTCTCGAGGGGTAGCTG-3’ 5’ HMG-CoA 5’-TACCATGTCAGGGGTACGTC-3’ 3’ HMG-CoA 5’-CAAGCCTAGAGACATAATCATC-3’ 5’-GTCTCTGAATCAGAAATCCTTCTATC-3’ 5’ IL- 1a (control) 3’ IL- 1a (control) 5’-CATGTCAAATTTCACTGCTTCATCC-3’ 19. Water, RNA grade (cat. no. W-4502, Sigma). 20 Mineral 011(cat. no 0186-2302, Perkin-Elmer) 21 bzs-Acrylamlde, 19 1 (cat no 0496, AMRESCO, Solon, OH) 22. 30% Acrylamlde (a potent neurotoxm in liquid form). 29 g acrylamlde + 1 g bzs-acrylamlde + 100 mL distilled water and heat at 37°C until dissolved 23. TEMED (cat. no. 076 1, AMRESCO) 24 5X TBE buffer. 0 445 A4Trl.s Base (Trizma base, cat no T1503, Sigma), 0.445 M boric acid (cat. no. B-6768, Sigma) and 10 mMEDTA (Na2EDTA * H,O; cat no T1503, Sigma). MIX 54 0 g Trls base, 27 5 g boric acid, and 3 72 g EDTA Adjust to 1 L with distilled water The resultmg pH should be 8.0. Dilute 1.5 or 1.10 to make 1X or 0 5X workmg solution 25. Ammonium persulfate (cat. no A-9164, Sigma). 26 PBR322 DNA digest using HaeIII (cat. no D8397, Sigma). 27 10% ETOH and 0 5% acetlc acid* 400 mL ETOH, 20 mL acetic acid; bring up to 4 L with distilled water 28 0 1% Silver mtrate (cat no. S-0139, Sigma) 300 mg silver nitrate m 300 mL dHz0 29 Developing solution (1.5% NaOH, 0 0 1% sodium borohydrate, 37% formaldehyde) 300 mL 500 mL 1.5% NaOH 4.5 g 7.5 g 0.0 1% Na borohydrlde 30.0 mg 50.0 mg 37% formaldehyde 12mL 2.0 mL 30 0.75% Sodium carbonate (cat. no. S-2127, Sigma) 3.75 g in 500 mL dlstilled water 3. Methods

3.7. Isolation

of Total RNA

Tissue specimens for mRNA analysis can be homogemzed m TRIZOL reagent, a monophasic mixture of guanidine lsothiocyanate and phenol. RNA 1s Isolated by a modlflcation of the rapid lsolatlon method of Chomenczynskl and Sacchl(5). This method was mltlally evaluated against the ultracentrifugatlon method described by Chirgwm et al. (6), previously employed in this laboratory and found to provide similar yields of RNA in much less time.

Quantification

of mRNA by PCR

47

3.1.1. Tissue Homogenization 1 Redissolve tissue guamdmm solution (GITC) by heating m a 65%C water bath (GITC solutions when kept m the cold tend to crystalize part of their GITC It is important to redissolve this solution prior to its use to homogenize tissue samples). GITC is also used to wash the homogenizer between specimens 2. Bring an amount of TRIZOL reagent sufficient for all your samples to room temperature (1 mL/2&100 mg tissue). 3. Place frozen tissue specimens in a contamer of liquid nitrogen and remove the needed sample using sterile tweezers, work quickly to prevent sample melting. 4. Place sample on weigh boat with a little liquid nitrogen in it. 5 Weigh the sample, making sure not to record the final weight until the liquid nitrogen has evaporated. 6. Transfer the sample to a homogenizer containing 0.8 mL of TRIZOL reagent and homogenize nnmediately using a loose insert 7 Transfer this volume of homogenate, using a ~-CCglass pipet, to a sterile 1 5-mL Eppendorf tube and rinse the homogemzer with 0 2 mL of TRIZOL Use the ~-CC glass ptpet to transfer this volume to the 1.5-mL tube and mix by gentle inversion 8 Wash the homogenizer between samples with DEPC water and 2.0 mL GITC to avotd carry-over The GITC, whrch mhrbrts RNases, will ensure that the water does not dilute out the GITC concentration m the TRIZOL.

3.7 2. Phase Separation 1 Incubate the homogenized samples for 5 mm at room temperature for complete dissociation of nucleoprotem complexes 2 Add 0 2 mL of chloroform per 1 mL of TRIZOL reagent. 3 Cap the tubes securely and shake them vigorously by hand for 15 s 4. Incubate samples at room temperature for 2-3 min 5 Centrifuge m a microfuge at 4°C at no more than 12,OOOgfor 15 mm Followmg centrrfugatron, the mixture separates mto a lower red phenol-chloroform phase, an interphase, and a colorless upper aqueous phase. RNA is in the aqueous phase

3.1.3. RNA Precipitation 1. Transfer the aqueous phase to a fresh 1.5-mL Eppendorf tube mrcropipet. 2. Precipitate the RNA by mixing with cold (4°C) isopropanol 3. Add 0.5 mL of sterile 100% isopropanol per 1 mL of TRIZOL homogenization, and mix by inversion. 4 Incubate at room temperature for 10 mm 5 Centrifuge for 10 mm at 12,OOOg at 4°C. The RNA precipitate pellet on the side and bottom of the tube. Depending on the isolated, the pellet may not be vmble, or may appear as a small 6. Carefully decant the supernate.

using a 1000-pL mmally used for

forms a gel-like amount of RNA white pellet.

48

Karge, Schaefer, and Ordovas

7 Wash the pellet once with 75% ethanol, addmg at least 1 mL of ethanol per 1 mL of TRIZOL reagent used for homogemzatton 8 Vortex and centrifuge at no more than 7500g for 5 min at 4°C 9. Decant the supernate and briefly dry the RNA pellet (invert over clean Kimwipes). 10 Dissolve the RNA in 100 pL of DEPC water by gently mixing the pellet up and down using a lOOO+L microptpet. If the pellet 1svery small, dissolve m only 50 pL. 11, Incubate for 10 mm at 55-6O”C if the pellet does not dissolve easily

3.1.4. Measure the Amount of Isolated Total RNA 1 Add 400 pL sterile 1X TE buffer and 4 pL of sample to an Eppendorf tube on ice. 2 Read optical density (OD) in UV spectrophotometer at wavelength of 260 nm after blanking with sterile 1X TE. Reset the wavelength to 280 nm, blank with 1X TE, and read the OD. Good quality RNA samples should have an 2601280 absorbance ratio > 1 6. Lower ratios can be caused by incomplete washing, DNA contamination, or protein contamination. 3. Calculate the yield using the formula 1 OD unit (260 nm) = 40 pg/mL of RNA Remember the sample was diluted l/100 (4 pL m 400 pL)

3.7.5. Storing the Sample Add 10 pL 3 M sodium acetate and 200 $., of 100% ethanol to each sample and store at -7OOC. Label each sample with the followmg: date, amount of RNA available (in pg), final concentration after adding 3 M Na Acetate and ethanol, total volume, type of sample, and the experimental condition.

3.1.6. Assessing RNA Quality by Electrophoresis The integrity of isolated RNA 1sevaluated by electrophoresmg on 1.0% agarose gels in a Tris/Acetate/ bromide (TAE + EB).

EDTA

2-pL alrquots buffer with 0.01% ethidium

1 Prepare 1% RNA grade agarose (0 5 g agarose per 50 mL sterile 1X TAE + EB per gel). 2 Microwave the agarose for l-2 mm while watching for botlmg (see Chapter 2) 3 Let the agarose cool for 15 min and pour mto a DEPC-treated mmtgel chamber (Note 6).

4. Place 6 pL sterile 1X TE, 1 pL RNA loading buffer, and 4 pL sample mto a sterile 1 5-mL Eppendorf tube on ice. (Use 2 pL sample and 8 pL TE if the pellet recovered was large, mdtcatmg a high RNA concentratton.) 5. Add lo-& lambda marker tn one well and IO-pL sample per well to the 1% agarose gel in sterile 1X TAE + EB runnmg buffer. 6. Electrophorese for 5 mm at 25 V to load the samples, and then run the gel for 30 mm at 75 V.

Quantification

of mRNA by PCR

49

Ethrdium bromide-stained products can be seenusing short wavelength UV. Distinct bands indicate good quality RNA. A 2: 1 or greater, a ratio of 28s to 18s RNA components as vtsualized by ultraviolet light is consrdered acceptable for further analysis, Strong bands at the top of the gel, near the well, suggest DNA or protein contamination. Smearing of the stain is a sign of RNA degradation.

3.2. Amplification

by PCR

3.2.7. Hot-Start and Reverse TranscrIption Reverse transcribe aliquots of total hepatic RNA mto cDNA using the Perkin-Elmer GeneAmp Thermostable rTth Reverse Transcrtptase PCR kit. 1 Place a rack with 0 5-mL thin-walled PCRtubeson Ice andaliquot the approprr-

ate volumes of samplesand controls. 2. Wash RNA stored m 0 1 A4 sodium acetate, 66% ethanol Centrifuge samples at 12,OOOgfor 15 mm at 4°C and decant Next, wash the pellets with 300 p.L of 70% ethanol, mix by mversion (do not vortex), centrifuge for 6 mm at 12,000g at 4°C decant, and dry on Kimwipes. 3. For “Hot Start” (Note 7), prepare the following reagent mrxes with enough volume for l-2 more samples than you are running to allow for sufficient volume to microprpet Mix each component well prior to adding it to the reaction mix (volumes listed are per sample) MIX A MIX B RNA grade water

11 40 pL to 12.15 pL

MnCl* ATP CTP GTP TTP rTth Polymerase

2 0 uL 041.LL RT buffer 0 4 l.lL 2olJ3’ primer 0.25 pL to 1 00 pL 04cls 0.4 pL 2.0 uL Total volume/sample 14 4 pL 5.6 p.L The amount of primer required depends on its concentration, with a final concentration of 0 1-O 5 @4 usually sufficient (Note 8). The sum of the total volumes of reaction mixes A and B should 20 pL per sample tube. Adjust the volume of water in mix A to compensate for changes m the amount of primer used per sample Keep polymerase frozen until ready to use. Keep all other reagents on ice once thawed The positive control tube should have 10.4 p.L RNA grade water, 2.0 pL RT buffer, 1.O pL 3’ control primer (cat no. DM152), and 2.0 pL of the Perkm-Elmer control material (5 x lo3 copies/pL) m place of Mix A. The negative control tube should have 11 4 p.L RNA grade water, 2.0 pL RT buffer, and 1 0 @ 3’ control primer (cat no DM152) in place of mix A (total volume still 14 4 pL)

Karge, Schaefer, and Ordovas

50

4. Vortex each reaction mix and spm for 1 mm m the mtcrofuge usmg 1 5-mL Eppendorf tubes as holders prior to pipetting the solutions mto each sample Tubes are centrifuged to bring the small volumes of reactants back to the bottom of the tube to ensure optimal reaction conditions. 5 Add the calculated amount of mix A to each tube, microfuge for 30 s, resuspend each pellet 50 times using a micropipet, and spin agam for 30 s. 6 Place tubes contammg samples and mtx A only in the thermal cycler for “Hot Start Reverse Transcrtptton.” Enter the approprtate parameters to program a 5 mm cycle at 90°C (step 1) followed by a 5 mm cycle at 70°C (step 2) and a 15 mm cycle at 70°C (step 3) Durmg step 1, the secondary structure of RNA IS denatured and the 3’ pnmer binds to the RNA Fmahze the preparatton of mix B during this time by adding the polymerase. (It 1scrttical to keep the polymerase on ice before use ) Vortex mix B and spm for 30 s Step 2 provides enough time to add mix B prior to the start of the next phase, reverse transcription (RT), which takes place during step 3 Open each tube during step 2, and add 5.6 pL of mix B to each tube Next, add 65 pL (2 drops) of RNA-grade mmeral 011to each tube to decrease evaporation 7 Remove each tube from the thermal cycler and spur m the microfuge for 30 s This step must be done rapidly to decrease the time the tubes are not at 70°C 8. Replace each tube before the start of step 3, RT cDNA IS the end product of step 3 The final volume of 20 pL of reaction mtxture contams reverse transcrtptase buffer (10 mA4Tris-HCl, 90 mA4KC1, pH 8 3), 1 mA4MnC12, 0 2 mA4of each NTP, 0.1 M 3’ prtmer, and 5 U of rTth polymerase overlatd wtth 65 Ccs,of mineral 011 9 Prepare the PCR buffer during step 3 by combmmg the followmg (Note 9). 61.0-61 75 l.tL RNA-grade water 8.0 pL Chelatmg buffer 15 mA4 MgC12 10.0 l.lL 5’ primer 0.25-l .O pL Total volume/sample 80.0 pL For the positive and negative Perkin-Elmer controls, add 1 pL of 5’ primer (DM15 I), 6 1 0 pL RNA grade water, 8 0 pL chelatmg buffer, and 10 0 pL MgCl, 10. The reverse transcriptton product is removed from the thermal cycler and quick chilled on ice. The cDNA can be stored at 4°C for several days before performing PCR

3.2.2. PCR Perform PCR using a final volume of 100 pL contammg 2.0 mA4MgCl,, 5 U rTth polymerase, 0.2 &of each NTP, 0.8X chelatmg buffer (1OX buffer contams 50% v/v glycerol, 100 mM Tns-HCl, pH 8.3, 1 M KCl, 7.5 M EGTA, 0.5% Tween-20), and 0.1 A4of 5’ and 3’ prtmers. The chelating buffer chelates MnC12 (Notes 10-12). 1 After reverse transcription 1scomplete, remove samples from the thermal cycler and add 80 & of PCR buffer mix to each tube. Each reaction should contain a

Quantification

2 3.

4 5

of mRNA by PCR

57

tube with all the above buffers and enzymes without any RNA to exclude PCR product contammation (the negative control) (Note 13) Vortex each tube and spm for 1 mm. Place the tubes back m the DNA cycler and enter the appropriate parameters for denaturing, annealing, and extension Standard PCR parameters are as follows a 95“C for 1 mm, 0 s (initial denaturation). b 60°C for 1 min, 30 s (annealing). c. 70°C for 1 mm, 30 s (extension) d Program this sequence for 25 cycles e Lmk to a file programming for 70°C for 5 mm, 0 s (final extension) followed by 4”C-hold (To keep the DNA cold until removed from the thermal cycler) The temperatures for annealing and extension vary depending on which mRNA you are quantifying; see example protocols (Subheading 3.3., Notes 14-21) At the end of PCR, press STOP and store the samples at 4°C.

3.2.3. PCR Product Detection by Silver Staining of Polyacrylamide Gels 1 Prepare the gel for electrophorests a Wash 20 x 20-cm glass plates with 100% ethanol, place spacers between the plates, and attach the clamps. b Place assembled plates onto the Bio-Rad apparatus with a sheet of Parafilm on top of the rubber seal and tighten the clamps on both sides. c. Check for leaks by fillmg at least half-way with 100% ethanol Pour out the ethanol and allow to an-dry 2 Prepare an 8% polyacrylamide gel using the followmg reagents to make up 50 mL per gel. for 50 mL (1 gel) 100 mL (2 gels) 30% acrylamide 13.30 mL 26.60 mL distilled water 3135mL 62 70 mL 500mL 1OOOmL 10X TBE (Note 22) 10% ammonmm persulfate (Note 23) 0.35 mL 0.70 mL *Place the solution m a clean sidearm flask and deaerate before adding the last reagent’ TEMED 35 ClL 50 PJ3. Draw the solution mto the barrel of a IO-mL syringe using an 18-gage needle. Invert the syringe and expel any air m the barrel Place the needle between the two plates and till the space almost to the top. Place a comb m place and use the remaining acrylamlde to fill the plates completely Allow 60 mm for the gel to polymerize at room temperature 4 Prepare samples and marker a Marker 1 pL HAE Digest I-kb DNA Ladder + 20 pL TE+ 2 uL RNA loading buffer

Karge, Schaefer, and Ordovas

52

5.

6

7. 8.

9 10 11 12

b Samples from 10-40 pL may be used dependmg on the expected level of expression m the tissue of interest. c Ratio of sample.dye should always be 10 1. Place plates into electrophorests umt, and fill the inner and outer chambers with 1X TBE, making sure to check for leaks The outer chamber should be filled halfway up the gel Place the entrre volume of samples and marker mto their respecttve wells Run at 3OCL400 V and no more than 60 A per gel (800 V, 120 A for 2 gels) until the second dye front has mtgrated half-way down the gel (1.5-2 h) When electrophoresrs 1s complete, pour the inner buffer down the smk and remove the clamps from the glass plates Carefully remove the gel from the plates and fix the cDNA wtth a solutton of 10% ETOH and 0 5% acettc acid, usmg two appltcattons of thts solutton for 3 mm each time a While fixing the cDNA, prepare 0 1% stlver nitrate stammg solution. b Also, prepare developing solutton during this or the next step Stain with silver mtrate for 10 mm Rinse the gel two or three times wtth distilled water and place m developmg solutton for 20 mm on the gel shaker Durmg this ttme, prepare the solutton to fix the gel, 0 75% Na carbonate FIX the gel for 10 mm and store m the 10% ETOH, 0 5% acetic acid solutron until the gel is scanned

3.2.4. Quantification of Products by Densltometry Scan each pan of bands correspondmg to the target mRNA and the internal standard cRNA m the one-drmensional mode by a laser densitometer. Quanttficatton of the target mRNA is calculated by compartson wtth the peak area of absorbance of the cRNA standard, and expressed as copies of RNA per pg of total hepattc RNA. The followmg formula proposed by Powell and Kroon (31, and later verified by Wang et al. (7) IS used for these calculattons: mRNA molecules/Clg total RNA = R x (copies of cRNA added) pg of total hepattc RNA added

where R is the ratio of the target and standard peak areas

3.3. Example 1: RT-PCR of LDL Receptor mRNA 3.3.1 Protocol 1 The first protocol

1s designed

to determine

whether

the amplification

efficiency IS srmtlar for the monkey target mRNA and the standard cRNA manufactured from an analysts of human gene sequences.As demonstrated by Scheuermann and Bauer (8), generating PCR products with varymg numbers of PCR cycles will allow for the detection of problems relating to efficiency differences.

Quant/fication of n-H/VA by /‘CR l

-

53

mRNA 1=0.946 Linear (mRNA ~0.946)

0 cRNA 1=0.892 - - - Linear (cRNA r=O.892) ---___--

LDL Receptor PCR Verification 0.4 8 0.35 + 0.3 z 0.25 8 0.2 &I 0.15

0.1 0.05 18

20

22

24

26

28

30

CYCLES

Ftg 1 (Continued on pp 54 and 5.5). Quanttficatton to tdentrfy eftictency drfferences between the control and target RNA molecules from cynomolgus monkeys The amount of product generated by each RNA was measured by silver stammg and laser desnsrtometry (A) Prrmers for LDL-receptor were used m each PCR reactron The data represent the average of two PCR runs. LDL-R RT-PCR IS performed by coamphfymg 100 ng of total cellular RNA and 5 x 1O4molecules of AW 109 cRNA for 18 to 30 cycles to compare the amplrficatron efficiency of the standard and target RNAs This cRNA comes m the RTPCR krt commercrally available from Perkm-Elmer/Cetus and contains prrmer sates for LDL-R m addrtron to HMG-CoA reductase and 10 other genes Prepare seven tubes wrth 100 ng RNA and 5 x lo4 molecules of AW 109 cRNA m each for PCR Include one posrttve and one negative control tube in the reactton Remove sample tubes from PCR at 18, 20, 22, 24, 26, 28, and 30 cycles a few seconds prior to the end of the extension phase for each cycle (70%(Z) Remove the control tubes at the end of 30 cycles. Run 20 u.L of each sample on an 8% polyacrylamrde gel and sliver stam Scan the detectable silver stained bands correspondmg to the predicted size of the internal standard cRNA and target LDL-Receptor mRNA products, 30 1 and 258 bp, respectrvely Plot the data on a graph of absorbance peak area versus cycle number The slope of the curve is related to the reaction efficiency (Fig. 1A) The plot reveals vn-tually parallel slopes of the curves for the target mRNA and standard cRNA products between 18 and 30 cycles. Since the data fit straight lmes, the PCR reactions appear to still be m the exponential phase of amphficatton, during which reactron efficrencres should be constant.

54

Karge, Schaefer, and Ordovas 1.4

LDL Receptor PCR Validation

18

B -r-------

I

0.20 / 20

22

24

26

28

30

Cycles

n

1.2

mRNA ~0.957 Linear (mRNA

1=0.957) - - - Linear (cRNA

--1

-0.2

24

26

28

30

CYCLES

Fig 1. (B) The data from (A) was analyzed m terms of the ratio of target mRNA to control cRNA to illustrate PCR reaction efficiency dtfferences. (C, D) same as (A) and (B), respectively, except primers for HMG-CoA reductase were used for PCR

7 Express the data m terms of the ratio of target mRNA to standard cRNA (M/S) versus cycle number should result m lmes that are almost hortzontal (Fig. IS) These results indicate that the effkrencres of the reactions are essenttally the same over the entire range of amplification cycles for LDL-R.

Quantification

55

of mRNA by PCR

1.4 1.2 -

is 0.l $ 0.6. E 0.4 0.2 -

D

HMG CoA Reductase PCR Validation

0 Lr

, 20

22

, 24

I 26

I-

I 28

30

32

Cycles

Fig 1 (contrnued)

3.3.2. Protocol 2 The second protocol is designed to determine whether quantitatlon IS lmear with respect to the amount of target mRNA present. Based on results from the first series of experiments described above, 24 cycles is selected as an optimal number of cycles within the exponential phase of amphficatlon for LDL-R. 1 A range of 25-150 ng total hepatic RNA and 1 67 x lo4 to 10 x lo4 molecules cRNA IS coamplified for 24 cycles to quantltate copies of LDL-R mRNA m the monkey liver sample evaluated 2 Prepare master mix with 25 ng mRNA/pL in a total volume of 59 pL Add 1 p.L of 1 x lo6 AW 109 standard for a final volume of 60-a tubes, with 1 67 x lo4 copies cRNA per & Prepare tubes contammg 25, 50, 75, 100, 125, and 150 ng total hepatlc RNA by ahquoting the appropriate volumes from the Master Mix per sample (1 & = 25 ng, 2 pL = 50 ng, and so on). The tubes will contam 1.67, 3 33, 5.00, 6.67, 8.35, and 10.0 x lo4 copies of standard cRNA, respectively 3. Plots of the results as shown in Fig. 2A to demonstrate that the intensity of the silver-stained bands is directly proportional to the amount of RNA used for RT-PCR

3.3.3. Protocol 3 Finally, to evaluate the reproduclblllty of the silver stammg method, a range of RNA concentrations from one sample 1s amplified as m the second protocol

56

Karge, Schaefer, and Ordovas

0 ngTotal RNA 2 1.8 1.6 * 1.4 4 1.2 w 1 2 0.8 .P 0.6 zo.4 0.2 6

12

18

24

30

36

ng Total RNA Fig. 2. Quantification to determine linearity of PCR detectlon m relation to mcreasmg amounts of total hepatlc RNA. (A) Primers for LDL-receptor were used m each PCR reaction The data represent the average of two PCR runs. (B) Same as (A), except primers for HMG-CoA reductase were used for PCR.

described above. The copy numbers per microgram of total hepatic RNA amplified for LDL-R mRNA are calculated from the ratio of target mRNA to standard cRNA. LDL-R mRNA levels varied by 11.2% for 12 samples with a mean of 6.15 x 10’ copies per pg total RNA.

Quantification of mRNA by PCR

57

3.4. Example 2: RT-PCR of HMGCoA Reducfase 3.4 1. Optimizing Number of Amplification Cycles The efficiency of amphficatron during HMG-CoA reductase RT-PCR IS evaluated by coampltfymg cRNA for 18-32 cycles.

24 ng of total cellular RNA and 5 x 1O4 molecules

1 Prepare eight tubes with 24 ng RNA and 5 x lo4 molecules of AW 109 cRNA n-r each for PCR. Include one positive and one negative control tube m the reactton 2 Remove sample tubes from PCR at 18,20,22,24,26,28,30, and 32 cycles a few seconds prior to the end of the extenston phase for each cycle (7O%C) Remove the control tubes at the end of 32 cycles 3 Run 30 pL of each sample on an 8% polyacrylamide gel and silver stain 4. Scan the detectable silver-stained bands correspondmg to the predicted size of the internal standard cRNA and target HMG-CoA reductase mRNA products, 303 bp and 246 bp, respectively 5 Plot the data on a graph of absorbance peak area versus cycle number (Fig. 1C). The plot reveals vn-tually parallel slopes of the curves for the target mRNA and standard cRNA products between 20 and 32 cycles for HMG-CoA reductase 6 Express the data m terms of the ratto of target mRNA to standard cRNA (M/S) versus cycle number should result m lmes that are almost hortzontal (Fig. 1D) These results indicate that the efficiencies of the reactions are essentially the same over the entire range of amphtication cycles for HMG-CoA reductase

3.4.2. Optimizing Amount of RNA Based on results from the first series of experiments described above, 26 cycles are optrmal for HMG-CoA reductase for the next series of experrments. 1. A range of 636 ng total hepatrc RNA and 1 875 x IO4 to 1 I 25 x IO4 copies cRNA is coamphfied for 26 cycles for quantitation of the copies of HMG-CoA reductase mRNA 2. Prepare tubes contammg 6, 12, 18, 24, 30, and 36 ng total hepatic RNA and 1 875,3 75,4.68,6.5,7 375, and 11 25 x lo4 copies of cRNA from a master mix similar to the LDL receptor protocol,

3 Plotsthe resultsasshownm Fig. 2B to demonstratethat the intensity of the silver stained bands IS directly proportional to the amount of RNA used for RT-PCR

3.4.3. Evaluating Reproducibility To evaluate the reproductbrlity of the srlver stammg method, a range of RNA concentratrons from one sample 1samplified as m the second protocol descrrbed above. The copy numbers per microgram of total hepattc RNA amplified for HMG-CoA reductase mRNA are calculated from the ratro of target mRNA to standard cRNA. HMG-CoA reductase mRNA levels varied by 15.5% for SIX samples with an average of 5.15 x lo5 copies per pg total RNA.

58

Karge, Schaefer, and Ordovas

3.5. Discussion The quantrtative PCR method described in thuschapter can accurately determine the amounts of low copy number mRNAs m limited amounts of sample (less than 100 pg of tissue). Silver staining allows for drrect comparrson of the final PCR products, confirmatron of the lack of nonspecific products, and verification of the size of the specific products m one step. The detectron technique IS rapid and safe, and eliminates the need for radioactivity, special labels, or generation of probes. The technique IS sensitive enough to require as little as 6 ng of total RNA for HMG-CoA reductase or 25 ng of RNA for LDL receptor analysts, compared to the need for samples m the microgram range for conventional blotting or hybridization techniques frequently employed for analySISof mRNAs present m low copy numbers. Stamed gels can be dried and stored mdefimtely for comparrson to other PAGE gels from the same or future experiments Reverse transcrtptton and coamphfication of the target mRNA and the standard cRNA m the same tube controls for variability due to differences in sample preparation and conditions during reverse transcription and amplification. Usmg the same primer for target mRNA and standard cRNA opttmrzes condttions for comparable amplification efficiencies for the mRNA and cRNA. The parallel lines obtained when plotting densttometrrc peak areas for target monkey mRNA and Perkm-ElmerKetus human cRNA standard versus the number of cycles of amplificatron mdtcate that the prrmers designed by Wang et al. (2) yield equivalent reaction effcrencres for the two cynomolgus monkey genes analyzed here and the respective standard sequences. Therefore, the commercially available RT-PCR kit and standard can be utilized for accurate and reproducrble investigation of the expression of these two important parameters of cholesterol metabolism m this animal model. Levels of expression for mRNA m this example protocol were found to fall within ranges prevrously reported for thesemessagesusmg similar PCR techniques. Powell and Kroon (3) calculated that human liver tissue has between 7 and 35 x lo4 copies for LDL-R mRNA and 5 to 22 x lo5 copies for HMG-CoA reductase mRNA per cogtotal RNA. Wang et al. (2) esttmated that there are 1.3 x lo4 copies of LDL-R mRNA m a normal human coronary artery. Gebhardt et al. (9) reported much greater levels of expression ranging from 2.3 to 5.4 x lo6 copies for LDL-R mRNA and from 1.7 to 6.6 x 1O6copies for HMG-CoA reductaseusing competitive PCR and a human gastric tumor cell line. The variability of results in this study compared favorably with those reported by Powell and Kroon (3). Those mvestrgatorsreported CVs ranging from 2 to 30% for LDL-R and HMGCoA reductase mRNA levels, whereas our protocol resulted m CVs of 11.2% and 15.5%, respectively.

Quantification

of mRNA by PCR

59

The ability to obtain reltable results from small sample sizes makes it possible to conduct experiments in which sufficient material for analysis can be acquired by liver biopsy. This has important implications in that animals that are expensive to mamtain and limited in supply do not have to be sacrtficed for studies pertaining to gene expression; and the potential exists for multiple samples from the same animal over time as various expertmental conditions are introduced. In addition, the technique could also be applied to human biopsy samples in clinical laboratories m the future for studies of human pathology. The replacement of radioactive labels by a conventional chemical detection method allows this technique to be easily performed by laboratories desiring to limit the use of radtoactivrty and the generation of hazardous waste 4. Notes 1 Ethidmm bromide is a powerful mutagen Handle the gels carefully and clean any spills so that nobody IS accidentally exposed to this reagent Dispose the used EB solutions as suggested m Chapter 1 2 Primers used m PCR are generally 15-30 nucleotides in length, which allows the use of higher annealing temperatures, Primers ~15 nucleotides can yield increased amounts of nonspecific primmg. When designing primers to detect mRNA, it is advisable to select the sequences from regions that are interrupted by mtrons in the genomic DNA In this way, any DNA contammation from the isolation procedure will run as a htgher band m the polyacrylamide gels that will not Interfere with the band of interest Moreover, if the mtron is large m size, the PCR wrll not produce band at all 3 GC content should be m the range of 40-60%, resulting m more energy required for the reaction (higher annealing temperatures). 4 Primer pairs should be designed so that there is no complementartty of the 3’ ends Thts will reduce the mctdence of primer-dimer pairs, 5 Avoid repetitive sequences in the primers (i e., GGGG), along with stretches of polypurines or polypyrimtdmes. 6. When assessing RNA quality (Subheading 3.1.6.), treat the electrophoresrs chamber, gel plate, and comb with 0.1% DEPC in water Rinse the equipment with sterile DEPC-water. The 0.1% DEPC ~111destroy any potential RNase contammants. The sterile DEPC water will rinse away the 0.01% DEPC so the DEPC will not degrade RNA 7. Hot-start PCR is performed to mnumtze nonspectfic priming while Increasing specificity and yield Imtial denaturation IS performed m the absence of polymerase or primers Assemblmg all components of the reaction mixture before thermal cycling begms can increase the potential for misprimmg. 8 Primer concentrations should be m the range of 0. l-l ccn/r.Too htgh a concentratton can Increase prtmer-dimer generation. Too low a concentratton can decrease yield.

60

Karge, Schaefer, and Ordovas

9 MIX MgCl, very well for PCR reagent mtxes The magnesmm precipitates at cold temperatures. 6-10 pL can be added to yield 1 5-2.5 mM MgClz depending on optimal reaction conditions (adjust RNA-grade water volumes accordingly) 10. Most PCR 1s carried out m the pH range of 8 3-8.8. I1 Mg2+ concentration has a major effect on the efficiency of PCR because of its forming complexes wtth dNTPs and Mg2+ requtrements of the polymerase enzyme Excesses of Mg2+ increases nonspecific priming while low Mg2+ levels reduce the yield of the reaction 12. Polymerase concentrations are usually optimal m the range of 0 005-O 025 lJ1p.L Higher concentrations can produce increases m nonspecific product formation 13 Reaction volume and tube wall thickness are significant variables for ensuring efficient thermal eqmlibrmm of the reaction mix Most reactions are performed m 25-100 pL volumes usmg 0 5- or 0 2-mL reaction tubes 14 The optimal annealing temperature (r,.,,) can be calculated usmg the followmg formula. T,,, = [(A-tT) x 2°C + (G+C) x 4”C] - 5°C 15 The T, derived from the formula IS a starting pomt for optimizing reaction condttions Optimal annealing temperatures can be determmed by runnmg optlmlzation experiments m whtch the annealing temperature IS changed by 2°C at a ttme until the greatest product yield is produced by PCR 16 Annealing temperatures lower than the calculated T,, will increase the probability of nonspecific amphtication products Raising the temperature above the calculated T,,, will increase the specificity, however product yield will decrease 17 Adding extension cycles of 15-20 s/cycle increases yield as enzyme concentration becomes a rate-hmmng condition 18 To Increase speclflcity and eliminate nonspecttic products. Increase annealing temperature, change Mg2+ concentration, change primer design, titrate the number of cycles, change the type of polymerase, or add a PCR enhancer 19. To improve fidelity (decrease error rate)* Decrease number of cycles, increase target RNA concentratron, change the type of polymerase, or remove any PCR enhancers 20 To increase yield of final product Increase number of cycles, increase target RNA concentratron, change the type of polymerase, decrease annealing temperature, or increase the extension time for larger size products 21 Optimal dNTP concentratton depends on the product size, primer concentration, Mg2+ concentration, and reaction conditions 22 It IS Important to use new TBE The borate m this buffer tends to precrpttate and the silver nitrate will bind to tt resulting m a dark background 23. Ammonmm persulfate may be stored at 4°C for 1 mo

References 1 Gllhland, G , Perrm, S., Blanchard, K., and Bunn, H F (1990) Analysts of cytokme mRNA and DNA. detection and quantitation by competitive polymerase cham reaction Proc Nat1 Acad Scl USA 87,2725-2729

Quantlficabon of mRNA by PCR

61

2 Wang, A M., Doyle, M V , and Mark, D. F (1989) Quantrtatron of mRNA by polymerase chain reactton Proc Nat1 Acad Scl USA 86, 97 17-972 1 3 Powell, E E and Kroon, P A (1992) Measurement of mRNA by quantltatrve PCR with a nonradioactive label J Llprd Res 33,609-614 4 Hames, B D (1990) One-dimenstonal polyacrylamrde gel electrophorests, m Gel Electrophoresls of Proteuw A Practical Approach, 2nd ed (Hames, B D and Rtckwood, D eds ), Oxford Umversity Press, New York, pp 60-67. 5. Chomenczynskt, P (1993) A reagent for the single-step stmultaneous rsolatron of RNA, DNA and proteins from cell and ttssue samples Blotechntques 15, 532-535 6. Chngwm, J M , Przyblya, A E , MacDonald, R J , and Rutter, W J. (1979) Isolation of btologrcally acttve rrbonuclerc acid from sources enriched In ribonuclease. Bzochemutry 18, 5294-5299. 7 Wang, Z , Chang, S Y , Macmullen, G , Huang, D P , Kwok, S , and Spadoro, J (1994) Use of internal standards to quantify target nucleic acid sequences by PCR (Abstract). Clan Chem 40,2335. 8 Scheuermann, R H and Bauer, S. R (1993) Polymerase chain reaction-based mRNA quanttftcatlon using an internal standard* analysis of oncogene expression. Meth Enzymol 218,44&473. 9. Gebhardt, A., Peters, A., Gerdmg, D., and Nrendorf, A. (1994) Rapid quantttatton of mRNA species m ethrdmm bromide-stamed gels of competttlve RT-PCR products. J. Lzpzd Res 35, 976-981

Generation

of Transgenic

Mice

Marten H. Hofker and Marco Breuer 1. Introduction Transgenic animals can be obtained by direct injection of DNA into one of the pronuclel of a fertilized oocyte. This DNA will typically Integrate at one site of the genome, as a multicopy msertlon, arranged in a head-to-tall fashion. Transgenesls 1s applicable to a wide range of mammahan species, including mice, rats, rabbits, sheep, pigs, and cattle. The focus of this chapter will be on mice, because mice are the most widely used laboratory animals for transgemc studies in cardiovascular research. In addition to the conventional transgemc technology described here, mice are the only mammalian species suitable for gene targeting by homologous recombmation in embryonic stem cells. Gene targeting allows gene rearrangements at predetermined sites and 1s a very important extension of the posslbihties to modify the mouse germline. This topic is described in Chapter 5. Genetically modifikd mice serve as excellent models for multifactorial diseases, owing to the possiblhty of studying a partlcular mutation on an inbred background. Further, combmmg different transgemc mice allows studying genetic interactions. Owing to these opportunities, research m transgenic animals has become one of the standard approaches m biomedical research. Transgenesls primarily serves to change the expression of a particular gene and examme the resulting phenotyplc alterations. In general, gene expression will be increased. The expression of the transgene may follow the endogenous pattern, or can be limited to distinct cell types or particular developmental stages. Alternatively, the gene can be expressed in cell types where the gene 1s normally inactive (ectopic gene expression). The desired expression pattern can be obtamed through the use of specific promoters. Furthermore, a mutant form of the gene, either natural or engineered, can be From

Meethods m Molecular E&ted by J M Ordovas

Bology, Vol 110 Lfpoprotem Protocols 0 Humana Press Inc , Totowa, NJ

63

64

Hofker and Breuer

used that should be capable of exerting a dominant effect. In addition to studies on gene function and pathology, transgenesis IS an ideal tool for augmenting m vitro gene-expression studies armed at identifying tissue-specific regulatory elements (1). In contrast to expression studies using tissue culture, transgenes are present m all tissues, and exposed to all mtra- and mtercellular signals. In the following sections we will describe some basic guidelines for construct design, micromjectmg oocytes, and screening transgemc founder mice. For other aspectswe would like to recommend two additional textbooks. Basic molecular biology technology, such as gene cloning, Southern blot hybridizattons, and plasmid tsolattons fall outside the scope of this chapter and can be found in ref. 2. Mouse husbandry, and microsurgical technologies for isolating oocytes and generating sterile male studs is well-described m ref. 3. Particularly, the maintenance of a breeding colony supporting transgenesis 1sa specialized task that is usually carried out by a core facility. Hence, when an animal facility and a microinjection setup are available, the strategtes and protocols described here wtll simplify the designing of experiments to generate transgenic mice. 2. Materials 2.1. Reagents

for the Preparation

of DNA for Microinjection

21.1. Clones Propagated in E. coli 1 Standard equipment for gel electrophoresls The agarose to be used 1s electrophoresls agarose standard low-m, (I e., Bio-Rad, Hercules, CA, cat.

2 3

4

5.

no 162-0100). TAE running buffer: 0.04 M Trts-Acetate, 1 mM ethylenediamme tetra-acetic acid (EDTA); ethidmm bromide stock solution. 10 mg/mL ethtdmm bromide, use 30 l.tL/L in the gel and running buffer for DNA stammg Electra elution buffer. 25 mA4 Tris, 1.5 mA4 EDTA; adjusted to pH 8.0 with boric acid. For DNA purttication. 3 A4 NaAc, pH 5 6, neutral phenol, pH 8 0 (Gibco, Gatthersburg, MD, cat no. 35730-019); phenol chloroform/chloroform/ isoamylalcohol24 24.1 (Gibco cat no 15593-03 1) TE“: 10 mMTris-HCI, 0.1 mMEDTA, pH 7 6 To avoid glass, pH is determined using a paper dipstick, and the water used 1s Analar water (Glbco, embryo qualtty) The buffer needs to be filtered (pathogen andparticle-free, to avoid blocking the microinlectton needle). Special equipment: electroelution supplies. We use microtrap cups (Isco, Lincoln, NE) as part of the electrophoresis system “Little Blue Tank ” Protocols depend on the exact equipment used, and are usually provided by the manufacturer The dialysis filter to be used IS Micropore

Generation of Transgenic Mice 2.1.2. Clones Propagated as Yeast Artificial Chromosomes

65 (YACs)

1 A setup for pulsed field gel electrophorests (PFGE) can be obtained from various commerctal supphers (for Instance, Bio-Rad CHEF DRII or DRIII) TBE runmng buffer, 0.045 A4 Tns-Borate, 0 001 M EDTA, Seakem Gold agarose (FMC). 2 Standard electrophoresis equipment, TAE running buffer 0 04 MTris-acetate, 1 mM EDTA; MP agarose (Boehringer-Mannhelm, Mannhelm, Germany); NuSteve GTG agarose (FMC). (FMC BtoProducts, Rockland, ME) 3. Agarase (Gelase, 0 2 U/L, Epicentre, Madison, WI); agarase buffer 0.2 mMEDTA, 7.5 mMTrts-HCl, pH 6 5, 100 mA4NaCI. 4. Injection buffer 0 2 mA4EDTA, 7.5 mA4Tris-HCl, pH 7 5, 100 mMNaC1 5 Dialysis filter Spectra/Par MWCO, 3500; (Spectrum Medical Industries, Houston, TX)

(The d@erent types of agarose listed above are highly recommended to use in combination with the protocols provided. However, other brands may offer equivalent types of agarose that will work equally well.) 2.2. Tail-Tip DNA Purification 1. Tall mix*100 mM Tris-HCI, 10 nnV EDTA, 0.2% SDS, 200 mM NaCl (Tris IS obtained from a 1 M Trts-HCl, pH 8.5 stock solutton, and the EDTA from a 0.5 M, pH 8.0, stock solutton adjusted with 5 N NaOH) 2 Protemase K stock solutton at 20 mg/mL 3 Isopropanol and 70% ethanol, TE -4* 10 mMTris-HCl, 0 1 mMEDTA, pH 7 6. 4 Special equipment* (hybridization) oven, capable of rotating. In our laboratory, we Insert the mtcrofuge tubes m the hybridization tubes provided with the hybridization oven, which allows rotating the microfuges tubes along their axis. Just rocking the tubes may be less efficient

2.3. Mouse Technology 2.3.1. Superovulation 1 2 3 4.

Folhgonan 1000 I E. (Intervet, Boxmeer, The Netherlands). Chorulon 1500 I E./I.U (Intervet). sterile water of physiological salt solution I-mL Syringes with 26G needle.

2.3.2. Oocyte Isolation and Handling: Materials 1 2. 3 4.

Mineral Oil (Sigma, St Louis, MO, cat no. M8416). M2 medium (Sigma cat no M7167). Ml6 medium (Sigma cat no M7292). Pemcrllium and Streptomycin solution, 10.000 U/mL (Gtbco-BRL, Gatthersburg, MD, cat no. 15140-l 14) 5. Pyruvate (Sigma cat no. P-2256). 6 Hyaluromdase type II (Sigma cat no. H-2 126).

66

Hofker and Breuer

7. 8 9 10 11

Depression sltdes 35-mm Petri dash for mtcrodrops Fire-polished Pasteurs pipet. Transfer ptpet (drawn Pasteurs prpet with an Internal diameter off 200 pm) Transplantation pipet (drawn Pasteurs pipet with an internal diameter shghtly wider as an egg cells)

2.3 2. Oocyte Isolation and Handling. Preparation of the Media M2-medium add to the 50 mL M2-medmm 0.5 mL Pemcilltum and Streptomycin solution If necessary, adJust pH to pH 7.3-7 4 with 5 NNaOH Sterthze with a 0 2-pm filter Discard the first few mL Divide in 2-mL ahquots m sterile tubes and store at 4“C until use These ahquots can be used for 1 mo M16-medium. add to the 50 mL M 16-medium: 0 5 mL Pemctllmm and Streptomycin solutton. Stenllze with a 0 2-pm filter. Discard the first few mL Drvlde m 2-mL aliquots m sterile tubes and store at 4°C until use. These ahquots can be used for 1 mo Hyaluromdase solutron: Dissolve 50 mg hyaluronidase in 50 mL M2-Medium (Sigma) If necessary, adjust to pH 7 3-7 4 with 5 NNaOH Sterilize with a 0 2-pm filter Discard the first few milliliters. Divide m 0.5~mL ahquots and store at -20°C Pyruvate Dtssolve 36 mg pyruvate into 10 mL of water Sterilize with a 0.2~pm filter. Discard the first few mL. Divide m 0.5-mL aliquots and store at -2O’C

3. Methods 3.1. Construct

Design

Before deciding on the actual construct design, one should choose the origin of the gene. In many cases, researchers use a human gene as the transgene. This choice offers the followmg advantages: Cloned human genes are often readily available in a wide variety of forms (cDNA, genomic). Further, the presence of human genes (and other nonmouse genes) can usually be detected easily m the mouse at the DNA, mRNA, and protein level. However, when studying a nonmouse gene product, tt should be noted that unwanted side effects may happen. For instance, the biological activity of the gene product could be dependent on protein-protein interactions, or homodimerizatton, which may no longer occur in the presence of proteins with a different ortgm (mouse and man). In other cases, the transgene may downregulate the mouse homolog, either at the transcription level, or durmg synthesis or catabolism of the protein. One should be aware of these processes, because they will affect the outcome of subsequent experiments.

3.1 1. Gene Constructs with Endogenous Regulatory Elements In many cases, taking a full-length mammalian gene, including 5’ and 3’ regulatory regions, results in transgenic mice showing a bona fide expression

Generation of Transgenic Mice

67

pattern, stmilar to the endogenous mouse gene. Apparently, regulatory sequences have often been conserved evolutionarily. For small genes (up to 15-20 kb) taking the genomic portion of the gene 1ssample when using a h-phage or cosmid clone as the source for the construct. For larger genes, up to approx 150 kb, tt 1s still fairly easy to obtain the entire gene, owing to the availability of large insert clones m E. coli (BACs, PACs, P 1). Such clones are being made available through commercial suppliers. Alternatively, tt 1s possible to use overlappmg genomic fragments, that will reconstitute to a full-length gene upon microinjection (#). A protocol to prepare DNA, (propagated through E. co&based vector systems), for micromjectton 1sprovided m Subheading 3.2. Larger genes require the use of YACs, which are typically maintained in the yeast strain Saccharomyces cerewsiae. Another reason for choosmg a YAC 1s that the use of a larger DNA fragment improves the chance of obtaining an integration site independent expression pattern, and a copy number dependent expression level (5). In principle, all genes can be cloned in YACs, and purified through the scheme given Subheading 3.2.2. However, inserts exceeding 500 kb become hard to handle, and are more difficult to use for micromjectton. Several other procedures have been developed to obtain transgenic mice with large YACs. For instance, the YAC can be transfected into an ES cell first (via sferoplast fusion or lipofectton), providing the opportumty to select for clones containing the entire YAC (6,7). These ES cells can be injected mto blastocysts (see Chapter 5) for generating transgemc mice. Alternatively, when the size of the YAC is too large, it is possible to generate a shorter YAC by inserting a new telomer at a random position m the YAC vta alu-mediated homologous recombination (8). When using a genomic clone as a starting point for transgenesis, the cloned gene should be tested for the integrity of the desired locus by subsequent digestron with several restrictton enzymes, and hybridization to a full-length cDNA clone. The hybridization pattern of the clone has to match with the hybridization pattern of genomic DNA. In parallel, one needs to map the 5’ and 3’ gene relative to the insert of the clone, m order to determine the amount of DNA that will be included to provide tissue-specific regulatory elements. How much DNA needs to be present on either side of the gene 1sdifferent for each individual gene. In a first experiment, 10-40 kb can be added to either side of the gene. The amount of flanking sequences included should be m proportion to the size of the gene, and also depends on the availability of flanking DNA. Although a large amount of flanking sequenceswill reduce the risk of omtttmg important regulatory elements, too much DNA may introduce unknown and unwanted genes.

68 3.12. Gene Constructs with Heterologous

Hofker and Breuer Regulatory Elements

When it is impossible or undesirable to use a full length genomic clone, a construct can be made by combimng a promoter and a gene, each from different origin (see, for example, ref. 9). Another good reason for using these constructs is to limit the expression of the gene to a certain cell type, or to a particular developmental stage. The construct should contain the followmg elements: promoter, gene (cDNA), transcriptron start site, translation start site (ATG), stop codon, and polyadenylatton signal. Further, mtrons and other regulatory elements such as enhancers can be mcluded to increase mRNA levels. Although the successof these heterologous constructs cannot be guaranteed, some general guidelines for construct design are as follows. In prmctple, one should take a promoter that has proven effectivity in transgemc mace. To date, a wide range of promoters for almost any cell type has been used successfully m transgenesis. In addition to mammahan genes, reporter genes such as P-galactosrdase or luciferase are very suitable to test promoter specificity and strength m vivo (10). The gene used is generally m the form of a cDNA. Although it would be advantageous to use the genomic form of the gene, this choice is usually hampered by the large size of most genes. Any full-length cDNA clone will, by defimtion, contam a transcriptional start site (ATG) and a stop codon. The heterologous promoter should still contam a small portion of exon 1, m order to keep the transcripttonal start site intact. The gene of interest should be cloned in this first exon. When the appropriate cloning sites have been determined, the sequence needs to be screened for the presence of transcriptional start sites, 1.e , the sequence ATG. It is important that the first ATG m the transcribed sequence belongs to the gene and that additional ATGs are not accidentally mtroduced between the transcriptional start site and the translational start site. Furthermore, 6 bp of the original 5’ sequence, termed the Kozak sequence (II), should precede the ATG to ensure a strong translational start signal. The presence of at least one intron m the construct aids in getting a high mRNA level. Introns can be derived from heterologous genes (12), or can be derived from the gene Itself. Although the use of an endogenous mtron may sometimes lead to more extensive clonmg steps, the chance of obtaining appropriate splicing may improve. The polyadenylation signal may be present already m the cDNA sequence or can be added by cloning. Alternatively, it is possible to use the 3’ end of the gene, mcluding some introns and the entire 3’ noncoding region. Because this region determines mRNA stability, and is often not completely present in full-length cDNA clones, the use of this part of the gene should be considered. However, in some cases, inserting the 3’-untranslated region will be counterproductive, because it may lead to mcreased mRNA turnover.

Generation of Transgenic Mice

69

3.2. Preparation of the DNA for Microinjection DNA of clones propagated in E coli, includmg plasmids, h-phages, cosmids, PACs, BACs, and Pls can be isolated followmg routine purification procedures. DNA from yeast requires a different approach (13). After DNA isolatton, the insert fragment should be liberated from the vector sequences by restriction enzyme digestion, particularly m case of relatively small (<40 kb) Inserts. In caseof larger inserts (BACs, PACs, Pls), removing the entire vector 1s still recommended. The release of the insert from the vector DNA is tmportant, because prokaryotic DNA has a tendency to methylate, leading to the

silencmg of the transgene. The purified DNA can be prepared for micromjectton followmg the protocol given m Subheading 3.2.1. In case of YACs, the complete clone can be used without further digestions, since the vector arms are quite small as compared to the insert size. The procedure in Subheading 3.2.2. describes the purificatton of the YAC DNA from gel, which can be used for microinJection. 3.2 I Purification of Constructs Propagated m E. co11 I Digest 50-100 pg DNA, leaving aslittle vector sequencesaspossibleattachedto the insert 2 Separate the fragments through agarose gel electrophorests The percentage of agarose used IS dependent on the Insert size. Run the gel m TAE buffer with ethtdmm bromide (0 3 mg/L) 3 Cut out the band from the gel 4. Apply electroelution. Protocols depend on the exact equipment used The elutton buffer 1s 12 5 mMTris, 2 5 mMEDTA, adjust to pH 8.0 with boric acid

After the electroelution: 5 Recover the DNA solution in a mtcrofuge tube, and add l/l0 volume 3 MNaAc, pH56 6. Add 1 volume of neutral phenol (pH 8 0; Gibco), gently swirl the tube for 30 s, and centrtfuge for 5 mm at full speed m a microfuge. 7. Recover the supernatant, and repeat step 6 8. Recover the supernatant and add 1 volume Phenol/Chloroform/Isoamylalcohol 24.24: 1 (Gibco), gently swirl for 30 s, and centrtfuge for 5 mm at full speed m a mtcrofuge. 9. Recover the supernatant, and add 1 volume chloroform. Gently mix the sample and centrifuge for 2 mm at full speed m a microfuge. 10. Recover the supernatant and add 1 volume ether. Gently mix and dtscard the supernatant (continue wtth the bottom fraction). 11. Repeat the ether extraction (IO). 12. Add 2 volumes of ethanol and precipitate the DNA overnight at -20°C

70

Hofker and Breuer

During subsequent stages, none of the chemicals may have been exposed to glass. Not observing this recommendation may lead to loss of viability of the oocytes. 13 Centrifuge down the DNA at full speed in a mlcrofuge Dissolve the pellet m 150 & TEA 14. Dlalysls of the DNA solution against 2 L TEA for 2 h at room temperature 15 Determine the DNA concentration by taking a small allquot and comparing this sample against a h DNA standard sample. Dilute the DNA to a concentration of l-4 pg/mL, and make ahquots of 15 & each. These aliquots can be stored at -80°C. Recurrent thawing and freezing should be avoided

3.2.2, Purification of Constructs Propagated rn Yeast YACs can be separated from the other yeast chromosomes usmg PFG It 1s desirable that the YAC has a different size than any of the yeast chromosomes. In that case, the YAC can be purified from the yeast DNA without using restriction enzyme digestions. Otherwise, a restrlctlon enzyme (combmatlon) needs to be determmed that allows separating the YAC from the yeast DNA However, finding those enzymes can be difficult, particularly in the case of bigger YACs. A good alternatlve is to transfer the YAC to “wmdowed” yeast strain, m which the size of the overlapping chromosome has been altered (14). The protocol here has been adapted from Wutz et al. (15). 1 Separate the YAC through PFGE using 0 6% Seakem Gold agarose m 0 5X TBE overnight at 4’C. Running conditions and gel density will depend on the size of the YAC. Do not use ethidlum bromide m the gel or buffer durmg electrophoresls. 2. Cut l-cm-wide strips from each side of the preparative gel, and stain the strips with ethidium bromide to locate the YAC positlon m the gel Using the stained parts as mdlcators a slice 0.5 cm wide containing the YAC can be cut from the unstained part of the gel. Cut out one of the adjacent yeast chromosomes as well. Eqmlibrate the gel shces m 1X TAE for 2 x 30 mm. 3. Embed the gel slices, turned 90’ mto a regular 0 5% MP agarose gel m 1X TAE. Posltlon the yeast chromosome parallel to the YAC. Cut out a 1-cm2 hole m front of each slice, and fill this hole with 4% NuSieve GTG agarose m 1X TAE. Run the gel overnight at 50 V at 4°C Do not use ethldmm bromide m the gel or buffer durmg electrophoresls. 4 Determme the posItion of the yeast chromosome m the NuSieve GTG agarose by staining with ethldium bromide. Do not stain the entire gel, but recover a small part the NuSleve agarose block, which should contain the YAC based on the observation with the yeast chromosome Do not take any MP agarose, because It will be hard to remove from the DNA solution during the later stages. Stam the remamder of gel with ethldium bromide to check for DNA mlgratlon. Equilibrate the agarose block in agarase buffer at 50°C for 1 h

Generation of Transgenic Mice

71

5 Transfer the agarose block to a 1.5-mL microfuge tube and heat to 70°C for 20 min. Keep the melted agarose sample at 45°C for 10 mm and incubate with 5 pL of agarase (Gelase) for 1 h Mix carefully using a ptpet with a cutoff I-mL (blue) tip 6 Add 3 pL of agarase, and continue the digestion during I h. 7 Place the sample on Ice No agarose should become visible. Dialyze overnight at 4°C against the micromjection buffer. Note that the composition of the mjectlon buffer is different from above (see Subheading 3.2.1.) Also, do not use glass for handling liquids at this stage. The water should be of tissue culture quality 8 Centrifuge the sample for 5’ at 13,000 rpm to remove remainmg debris. Collect the supernatant, and test 10 pL by PFGE to check the integrity and concentration of the YAC preparation Dilute the DNA in mlcroinJectlon buffer to 2-5 ng/mL

3.3. Mouse Husbandry,

Choice of Strains, and Oocyfe Isolation

As stated earlier, this manual assumes that the mvesttgator has access to a professional animal house where the mice can be housed under approprtate

conditrons and where ample experience exists wtth mace breedmg and operation techniques involved m transgenesis. All mace involved have ad lzbitum accessto water and standard chow and are housed under 12-h day/night cycles. For a detailed description for setting up a colony for transgenic purpose (oocyte donor, fertile studs and sterile studs for generating pseudopregnant female) as well as the transgenic techniques we refer to “Mampulatmg the Mouse Embryo” by Hogan et al (3), and the video guide “Transgemc Techniques m Mice” (21), Before the injections can start, one has to consider which strain has to be used for transgenesis. If a defined genetic context 1s important, for example for rmmunological studies or for genetic studies on polygemc diseases, transgenesis should be preferably performed m inbred strains. Although there are a few inbred strains that can be used, the most widely used strain is C57BL/ 65, probably because of histortcal reasons. The common disadvantage of inbred strains is then relative poor yield of fertilized eggs after superovulation as compared to Fl hybrids. Furthermore, “inbred” eggs have an attenuated viability upon in vitro culture, mrcroiqection,

and transplantation.

In addition,

most

inbred strams have a reduced reproduction capacity. Therefore, the use of inbred strains should be avoided, if a homogeneous genetic background 1snot an absolute requirement. One exception may be the recently introduced FVB/ N inbred strain, which gives a high number of fertilized egg cells upon superovulatron (16). However, m other cases, Fl hybrrds are being used to obtain a high number of oocytes (up to 20-25), of which a large fraction will develop to term. Several Fl hybrids can be chosen, for instance the widely used hybrid of C57BL/6J and CBA (BCBA). Other combinations that have been used are C57BL/6J x SJL, C3H/HeJ x C57BL/6J, C3H/HeJ x DBA/2J, and C57BL/6J x DBA/2J.

72

Hofker and Breuer

3.3.1. Superovulation 1 Add the 5-mL solvent to the freeze-dried follrgonan and mix well Thus stock can be used for a period of2 mo when stored at 4’C 2 Add the 5-mL solvent to the freeze-dried folhgonan and mix well. This stock can be used for a period of 2 mo when stored at 4°C Day 1 3 Dilute folligonan stock solution IOX with sterile water or physiologrcal solution 4 At 4 PM, inject each female mouse tp wrth 0.2 mL

salt

Day 2 5 Dilute Chorulon stock solutton 10X with sterile water or physlologlcal salt solution, 6 At 1 PM, inject each female mouse rp with 0 2 mL and transfer to a fertile stud. Use one male per female To ensure the maximum number of fertrllzed eggs, these male mice are only used once a week At 8 mo of age, or earlier when the plugging ratio IS dropped below 70%, these male mice are replaced. Day 4 7 Check females for vaginal plugs early m the morning able to use for isolation of the oocytes.

Mace with plugs are smt-

3.3.2. Procedure for Isolation of Oocytes 1 One day before the rsolatron procedure* Incubate one ahquot of M2, Ml6 (loosen cap), the paraffin oil, and the depression slides ovemrght m the 37°C 5% CO, mcubator before use 2 At the day of isolation: Add 20 & of the pyruvate solutron to the 2 mL M2 and M 16 aliquots and mix by gently swrrlmg. 3. Prepare the depression slides with 0.2 mL M2, two depression shdes with 0 2 mL Hyaluromdase solution, and three depression slides with 0 2 mL Ml6 medium Put each depression slide rn a 9-cm Petri dish and incubate them briefly m a 5%, 37°C incubator 4 Prepare a 35-mm Petri dish with 5-8 mrcrodroplets (3-5 pL) of Ml6 under paraffin 011and incubate brtefly m the 5%, 37°C incubator 5 Collect females with vaginal plugs

Start dissecting female mice at 10 AM since at a later time point the cumulus cells begin to fall of and the egg cells are more difficult to isolate. 6. Sacrrtice the mouse by cervmal dtslocatron 7 Place the mouse at her back and drsmfect abdomen wrth 70% ethanol. 8 Remove skm from the abdomen

Generation of Transgenic Mice

73

9. Open the abdominal cavity by cutting through the body wall in such a way that the complete abdominal cavity 1svisible 10 Collect the oviducts and place them m an depression slide filled with 0 2 mL M2 medium. 11 Transfer one oviduct to a new depressron slide with M2 medium 12. Disrupt the oviduct at the swellmg with a fine needle The cumulus will be expelled in the medium 13. Repeat steps 8-9 for all other oviducts. 14. Collect all curnull and single egg cells with a fire-polished Pasteurs ptpet and transfer to a depression slide filled with hyaluronidase solution. 15 Incubate m CO, mcubator until the cumulus cells begin to fall off 16 Transfer the egg cells to a new depression slide wtth hyaluromdase solutton and Incubate again until the egg cells are free of cumulus cells Take care not to overexpose to hyaluromdase, because this may be harmful to the egg cells 17. Transfer the eggs cells to M2 medium, prpet up and down the cells several times until most cells are free of any cumuli cells. 18. Transfer the egg cells to an depression slide with Ml6 medium and wash cells by pipetting up and down. 19 Repeat this step twice more. 20. Transfer cells (free of any debris) to a microdroplet of Ml6 under paraffin 011. 2 1. Store m a CO, incubator until the inJections can start.

3.4. Pronuclear

Injection

As mentioned

previously, this protocol assumes that one has access to an qection unit. If this is not the case several types of microscopes, mtcromanipulators, injectors, and needle pullers are commercially available. We refer

to Hogan et al. (3) for a detailed descrtption of an injection setup. Injection of the egg cells and subsequent transplantation are in principle as described by Hogan et al. (3). For injection, we make use of two injection chambers. These chambers are made of a perspex ring (with an outer diameter slightly smaller of the slide and 1.5 mm high), which is fixed on a siltconized microscopic slide with 2% agarose. In each chamber is a droplet of 10 pL of Ml 6 and a 3-5 pL DNA droplet under paraffin oil, These chambers are prepared just before use and kept m a 90-mm Petri dish m the CO* incubator until use. The two chambers are used alternately. Injection procedure: 1. Just before inJection, 2B-40 1-cell eggs are transferred into the Ml 6 droplet of the injection chamber. Take no more cells than that can be inJected within 20-30 min 2. Place injection-chamber on the microscope 3. Fill mjection needle with DNA solution from the DNA droplet. 4. Place holding pipet and injection needle mto position.

74

Hofker and Breuer

For the egg cell mampulatron and mjectron, a x300 or x400 magmficatron is used. 5. Fixate one egg cell wtth the holding prpet. 6. Introduce the injection needle carefully into one of the pronuclei In most cases the male pronucleus IS the best accessible, owmg to its larger srze Take care not to hit the nucleolus. 7 Expel the DNA solution into the pronucleus Inlect just enough to see the pronucleus swell under the microscope 8 Withdraw the mjection needle gently out of the oocyte. If any material IS accidentally pulled out of the nucleus, or when the oocytes tend to become leaky too often, the needle has to be replaced. 9 Place the correctly Injected cell on one side (top) of the Ml6 mtcrodroplets Inappropriately injected cells, mcludmg cells that have become leaky after micromjection are placed on the opposite (bottom) side. 10 Repeat steps 4-8 until all cells are mjected 11. After all cells have been injected, the correctly and viable cells are transferred mto a new Ml6 microdroplet of the 35-mm Petri dish. Incubate m a CO, mcubator until transplantation.

In general 60-80% of the egg cells will survive this injection and can be transferred (3) to the oviduct of a pseudopregnant female. Transplantation can be done on the same day as the mjectron (l-cell stage). However, we prefer to culture the l-cell eggs overnight, during which time they will develop to the 2-cell stage, and select only the properly developed embryos for transplantation. 3.5. Generafing Pseudopregnant Females (Fosters) For pseudopregnant females, any strains may be used that are known for being good mothers, who ~111also take care of small litters with only a few pups. In general, most Fl hybrids or outbred strains can be used. For practical reasons, we use females from the same strain (BCBA) that were used as oocyte donors. The females should be at least 5-6 wk old and have a weight below 30 g. The pseudopregnant females are obtained by mating with a sterile male. This male can be either genetically sterile or vasectomized (3). To have enough pseudopregnant females available, at least 10 females per oviduct transplantation are being mated. Each sterile male (which has been housed individually to avoid fighting) is placed m a cage with two females. The females with vagmal plugs can only be used at the day of transplantation. The surplus of pseudopregnant fosters can be reused after 2 wk. The vasectomized male mice can be used twice a week for matmgs. The foster mice are checked every other day for pregnancy. Depending on the strain, the litter will be delivered approx 19 d after transplantation.

Generation of Transgenic Mice

75

3.6. Analysis of Founders The first Inters, produced by the fosters, typically will harbor one transgenic mouse per 5-l 0 mice tested. These transgemc founder mice can be identified through Southern blot analysis of the tail-tip DNA. It is also possible to mvesttgate the first litters through polymerase chain reaction (PCR) analysis. However, it is highly recommended to follow up any PCR-positive mice using Southern blot analysis. The problem with PCR analysis is that its high sensitivity allows detecting spurious positives. Such mice may never produce transgemc offspring owing to extremely low numbers of positive cells carrymg the transgene. In contrasts, chimeric mice identified by Southern blotting are often successful m generating transgemc offspring, albeit sometimes m low numbers owing to chimeras. However, the next generation should show a normal Mendelian segregation pattern, Southern blot analysis and PCR analySIShas been described m detail elsewhere (2). Here, we will provide a rapid procedure for tail-tip analysis, adapted from Laird et al. (17). In addition to DNA analysis, it is sometimes possible to screen for the presence of the newly introduced protein. Thus approach is useful in case of apohpoproteins, for instance, which can be detected m the blood provided that an mnnunological assay IS available. 3.61. Tail- T/p DNA isolation 1 Cut the tip (1 cm) of a mouse tail, and store the tail in a 1 5-mL mtcrofuge tube on ice During this procedure, it is convenient to sedate the mouse hghtly usmg halothane or isofluorane m a suitable setup (Both compounds are toxic for the liver upon chrome exposure.) At this ttme, the mouse needs to be marked (3) as well Furthermore, it is possible to take a small blood sample from the mouse for biochemical analysis. The tail can be stored at -20°C for a longer period, or processed munedrately. 2 Add 740 $, Tail mrx and 10 pL Protemase K (20 mg/mL). Incubate overnight at 5S’C under gentle agitation. Without agitation, the tails will dissolve, but the subsequent procedures (Southern blot analysis, PCR) become less robust 3. Centrifuge the remammg debris (hairs, bone) down in a mtcrofuge for 10 min at full speed, and transfer the supematant to a fresh micromge tube 4 Add an equal volume of tsopropanol, and mix gently, allowing the DNA to become visible as small threads 5 Harvest the DNA using a Pasteur pipet, of whtch the glass ttp has been closed by meltmg. If there IS too vigorous action in the preceding step, it makes this step unnecessartly difficult. 6. Keep the DNA sticking to the glass tip, and rinse the DNA m 70% ethanol. Let the DNA dry briefly m the air, and transfer the DNA to 75 pL TEd 7. Dissolve the DNA thoroughly by incubating for 15 min at 6S’C, vortexmg briefly 8. Store the DNA samples at 4°C until needed.

Hofker and Breuer

76

3.7. Analysis of Offspring All mace carrying the transgene will be used m a breeding scheme. It is important to quickly obtain a large number of offsprmg, allowing identification of the mouse strains that may become useful. Hence, it is advisable to use productive female mice such as the BCBA hybrids used for the superovulation. The first litters produced by transgentc founder mice can be tested by PCR. In addition, a representative number of transgenic mice needs to be analyzed by Southern blot analysts as well. The banding patterns observed on the Southern blots should be equal to the founder mouse, ensuring that the transgene has integrated at one site, and that the offsprmg has the same transgemc msertion The signal intensity among the offsprmg should also be similar, but may be higher than m the founder An estimate can be made about the copy number, because at this stage all cells should carry a copy of the transgenic msertron on one of their chromosome pairs. Multiple transgemc insertions m a single mouse strain will segregate during the breedmg, and should be maintained as two or more different mouse lmes. Multiple insertions are difficult to discern from an unstable integration of the transgene. In caseof doubt, znsztu hybridtzation can rapidly provide the answer. Once a transgenic founder mouse has produced sigmficant numbers of offsprmg, a typical first analysis includes a complete analysts by a mouse pathologist, mRNA analysis of all accessible organs, and biochemical analysis to obtain a preliminary impression of the effect of the transgene. 4. Notes 1 The product of the transgene should be responsible for the phenotype of the mouse However, the transgene may also affect the expressron of (nearby) endogenous genes, which may Influence the phenotype in an unforseen manner To exclude that an mappropriate phenotype IS ascrrbed to the transgene, tt 1s important to have more than one transgemc lme avatlable

2 Breeding transgenesto homozygosityis sometimesimpossrble owing to lethal recessivemutations mtroduced by the transgemc msertton 3 Some transgemc constructs may fat1 to work properly, t.e , all four or more transgemc lines do not express the gene at stgmficant levels The problem may be that the construct is not destgned properly However, it IS also possible that the transgene dertved protein interferes with embryogenests In the latter case, the use of a different (mductble) promoter is ltkely to be more successful

4. The behavior of recessivemutationscan be studiedin mice when the transgeneIS expressed at very high levels (10, or when the transgenic mouse IS crossed with a

knockoutmouseto eliminate the endogenousgene(29) 5. In the absence of suitable regulatory elements such as a locus control region, a high copy number of the transgene IS associated with a decrease m transcnptron (20).

Generation of Transgenic Mce

77

5. Information on the World Wide Web A wealth of information is available through the Internet. The Jackson laboratories provide almost any kind of mouse-related information (http://www.jax.org/). Another source providmg a good database contammg knockout and transgenic mice is TBASE, provided by Johns Hopkins Umversity (http://www.brs.med.Jhml.edu/Dan/tbase/tbase.html). Unfortunately, databases for conventional transgemc mice have been made at a much smaller scale than databases for knockout mice. Important mformation regardmg the status of the mouse genome map, including available YAC clones, is provided by the Whitehead Institute (http://www-genome.wi.mit.edu/cgi-bin/mouse/mdex) and by The European Collaborattve Interspectfic Mouse Backcross (EUCIB) consortium (http:l/www.hgmp.mrc.ac.uk/MBxHomepage.html). References 1 Taylor, J. M , Stmonet, W S , Lauer, S J , Zhu, G , and Walker, D (1993) Regulatton and expression of the human apollpoprotem E gene m transgemc mice Cur-r Opm Llpld 4, 84-89 2 Sambrook, J , Frttsch, E F , and Maniatis, T. (1989) Molecular Clonzng A Laboratory Manual, 2nd ed. (Ford, N., ed.), Cold Spring Harbor Laboratory, Cold Sprmg Harbor, New York. 3 Hogan, B., Beddmgton, R , Constantim, F., et al (1994) Mampulatmg the mouse embryo A laboratory manual 2nd ed Cold Sprmg Harbor Laboratory, Cold Spring Harbor, New York 4 Pteper, F R , de Wrt, I C M , Pronk, A. C J., Kooiman, P M., StrtJker, R , Krimpenfort, P J A , Nuyens, J. H., and de Boer, H. A . (1992) Effictent generation of functtonal transgenes by homologous recombmatton m murme zygotes Nucl Acids Res 20, 1259-1264 5 Schedl, A., Montolm, L , Kelsey, G., and Schutz, G. (1993) A yeast arttfictal chromosome covering the tyrosmase gene confers copy number-dependent expression in transgenic mice Nature 362(6417), 258-261 6 Jakobovtts, A , Moore, A L , Green, L L., Vergara, G J , Maynard Currte, C E , Austm, H. A , and Klapholz, S. (1993) Germ-lme transmission and expression of a human-derrved yeast artrfictal chromosome Nature 362(6417), 255-258 7. Strauss, W. M., Dausman, J., Beard, C , Johnson, C., Lawrence, J B., and Jaemsch, R (1993) Germ lme transmtsston of a yeast artificial chromosome spanning the murme alpha l(I) collagen locus. Science 259(5103), 1904-1907 8. Heus, J J , de Wmther, M P , van de Vosse, E., van Ommen, G. J., and den Dunnen, J. T (1997) Centromerrc and noncentromertc ADE2-selectable fragmentation vectors for yeast arttficial chromosomes m AB 1380 Genome Res. 7(6), 657-660 9 Palmiter, R D , Chen, H. Y., and Brinster, R. L. (1982) Dtfferenttal regulation of

metallothtonein-thymtdme kmase fusion genes in transgemc mace and then offspring Cell 29, 70 1-7 10

78

Hofker and Breuer

10 Korhonen, J , Lahtinen, I, Halmekyto, M., Alhonen, L., Janne, J , Dumont, D , and Ahtalo, K. (1995) Endothebal-specific gene expression dtrected by the tze gene promoter m VIVO. Blood 86, 1828-l 835 11. Kozak, M. (1987) At least six nucleottdes preceding the AUG mttiator codon enhance translation m mammalian cells. J MOE Biol 196(4), 947-950. 12. Palmiter, R. D., Sandgren, E. P , Avarbock, M. R., Allen, D. D , and Brinster, R L. (1991) Heterologous mtrons can enhance expression of transgenes m mice Proc Natl Acad Scl USA 88,478-482. 13 Schedl, A , Larin, 2 , Montoliu, L , Thtes, E., Kelsey, G., Lehrach, H., and Schutz, G (1993) A method for the generation of YAC transgemc mice by pronuclear mtcroinjection Nuclerc Aczds Res 21(20), 4783-4787 14. Hamer, L., Johnston, M., and Green, E. D (1995) Isolation of yeast arttfictal chromosomes free of endogenous yeast chromosomes. Construction of alternate hosts with defined karyotyptc alterattons. Proc Nat1 Acad SCI USA 92(25), 11,706-l 1,710 15 Wutz, A , Smrzka, 0 W., Schwelfer, N., Schellander, K., Wagner, E. F., and Barlow, D. P (1997) Imprinted expression of the Igf2r gene depends on an mtromc CpG Island Nature 389(6652), 745-749 16 Taketo, M., Schroeder, A. C., Mobraaten, L E., Gunning, K. B., Hanten, G , Fox, R. R , Roderick, T H., Stewart, C. L., Lilly, F., Hansen, C T , et al. ( 199 1) FVB/ N: An inbred mouse stram preferable for transgenic analyses Proc Nat1 Acad Scz USA 88,2065-2069 17 Land, P. W , ZiJderveld, A , Lmders, K., Rumckt, M A , Jaemsch, R., and Berns, A (1991) Stmpltfied mammaltan DNA lsolatton procedure. Nucl Acids Res. 19,4293. 18 Huang, Y. D, Schwendner, S. W , Rall, S C , and Mahley, R W.-(ReprintAuthor) (1996) Hypohpidemtc and hyperlipidemlc phenotypes m transgemc mice expressing human apoltpoprotem e2. J Biol Chem 271(46), 29,146 29,151 19 Van Vlqmen, B J M., Wtllems van DiJk, K., Van? Hof, H B., Van Gorp, P J. J , Van der Zee, A, Van der Boom, H , Breuer, M L., Hofker, M. H., and Havekes, L. M. (1996) In the absence of endogenous mous apolipoprotem E, apoltpoprotem E2(Arg-158 -> Cys) transgemc mice develop more severe hyperltpoproteinemia than apobpoprotein E3leiden transgemc mice. J B~ol Chem 271(48),30,595-30,602 20. Garrick, D., Ftermg, S., Martin, D .I K , and Whitelaw, E (1998) Repeat induced gene silencing in mammals. Nature Genet. 18(l), 919-923 21. Pedersen, R. and Rossaut, Y. (mstructors) (1989) m Transgenzc Techniques zn Mrce A Vzdeo Gwde Cold Spring Harbor Laboratory, Cold Spring Harbor Laboratory, New York.

5 Preparation

of Knockout

Mice

Jestis Osada and Nobuyo Maeda 1.

Introduction

The analysis of the in vivo role of a specific apolipoprotein can be ascertamed in two extreme conditions. In one condition, protein is missmg; in the other, it is produced in a large amount. Both conditions can be easily achieved using genetlc engineering techmques. In the first case, the gene coding for the specific protein is inactivated (gene targeting) (I), whereas, m the second case, multiple copies of the gene are added (transgenesls) (2). The latter posslbllity has a wide range of animals to be used, whereas gene targeting has, so far, been restricted to mice. The application of gene targeting to these animals results in mice lackmg certain genes, called knockout mice. That gene targetmg may be used to replace genes 1s a promising aspect of gene therapy, allowmg replacement of damaged genes wlthout affecting other genes. Development of knockout mice is a multistep process that requires. 1 Clonmg of mouse genomic gene. 2 Deslgmng an exogenous construct to disrupt the gene once the exogenous DNA IS Inserted mto the gene via homologous recombmation 3 Transfection of the construct mto mouse embryomc stem (ES) cells. 4 Selection of the clones that have undergone homologous recombmatlon 5 Injection of ES cells into blastocysts. 6 Transference of blastocysts mto pseudopregnant mothers to obtain chimeras. 7 Germline transmission of chimeras and born of heterozygous animals. 8. Breeding of heterozygous animals to obtain homozygous ammals.

2. Materials The following

reagents should be prepared:

1. Chloroform*lsoamyl alcohol (CIA): 24 vol of chloroform are mlxed with 1 volume of lsoamyl alcohol. From

Methods /n Molecular Bology, Vol 7 10 Lpoprotem Protocols Edlted by J M Ordovas 0 Humana Press Inc , Totowa, NJ

79

80

Osada and Maeda

2 100X Denhardt’s solution* 2% polyvmyl pyrrohdone, 2% ficoll-400, and 2% bovine serum albumin (BSA) (fractton V) 3 1X TE. 10 mMTrrs-HCI, pH 8 0, 1 mMEDTA Autoclave 4. 10 mg/mL Salmon sperm DNA solution (ssDNA). 1 g of ssDNA is embedded m 100 mL of 1X TE overnight at 65”C, then autoclaved for 20 min Once cold, the solutron IS extracted with an equal volume of neutral phenol and centrifuged for 10 mm at 1SOOg.The upper phase is recovered and extracted with an equal volume of CIA and centrifuged as noted Upper phase 1s the ssDNA ready to be used. Store ahquoted at -20°C 5. Prehybrtdatton solution: 50% formamide, 3X saline sodmm citrate (SSC), 10X Denhardt’s, 2% sodium dodecyl sulfate (SDS), and 20 clg/mL denatured ssDNA. Store at -20°C 6. Loading buffer: 0 25% bromophenol blue, 0 25% xylene cyan01 FF, and 30% glycerol. 7 Hybridanon solutron 50% formamrde, 3X SSC, 1X Denhardt’s, 5% Dextran sulfate, 2% SDS, and 20 pg/mL denatured ssDNA Store at -20°C 8 SM elution buffer 50 mMNaC1, 10 mA4 MgS04, 50 mMTrts-HCl, pH 7 5, and 0 01% gelatin. Autoclave for 20 mm 9 Sterile STE* 0 lMNaC1, 10 mMTrrs-HCl, pH 8 0, 10 mA4 ethelenedtamme tetraacetic acid (EDTA), pH 8.0, Autoclave before using. 10 GTE: 50 mM glucose, 25 mM Tris-HCl, pH 8 0, 10 mA4 EDTA, pH 8 0. Autoclave. Store at 4°C. 11. Phosphate-buffered saline (PBS): 8 g/L NaCl, 0 2 g/L KCI, 1.44 g/L Na*HPO,, 0 24 g/L KH,P04, pH 7 2 Autoclave 12 Trypsin solution: 0 04% trypsm (Grbco-BRL, Gatthersburg, MD, No. 610-5050 AG), 0.02% EDTA, 0.8% NaCl, 0 04% KCl, 0 1% glucose, 0.035 % NaHCO,, pH 7.0. Sterthze by filtration Store at 4°C. 13 ES medium* 15% fetal bovine serum (FBS), 115 M P-mercaptoethanol, 2.4 mM L-glutamine; 4.5 g/L glucose in Dulbecco’s modified Eagle’s medium (DMEM) H (Gibco-BRL). Sterilize by filtration. Stable 2 wk at 4°C. 14. Freezing medmm for tissue culture. 11% dimethyl sulfoxtde (DMSO); 50% FBS, and 39 % DMEM-H medium. Sterihze by filtratton. Stable 2 wk at 4°C. 15 M2 medium. 94.66 mMNaCl,4 78 mMKC1, 1 71 mMCaC12, 1.19 mMKH,PO,, 1.19 MgS04, 4 15 m,W NaHC03, 20.85 mJ4 HEPES, pH 7.4,23 28 mA4 sodmm lactate, 0 33 mM sodium pyruvate, 5.56 mM glucose, 4 g/L BSA, 60 mg/L pemcrllm G potassium, 50 mg/L streptomycin sulfate, and 10 mg/L phenol red. Sterilrze by filtration Check osmolartty Stable 1 wk at 4°C 16 2.5% (w/v) avertm’ Prepare a 100% solutton of 2,2,2-trtbromoethyl alcohol m tertamyl alcohol. Make a dilution of 100% to a 2.5% m PBS. Store at 4°C protected from light. 17 ES Lysrs buffer. 10 mMTris-HCl, pH 8.2,2 WEDTA, 400 mMNaC1, 1% SDS and 100 pg/mL proteinase K. Store frozen at -20°C 18. Tail lysts buffer: 50 nuI4 Trts-HCl, pH 7 5, 100 m&J EDTA, 100 mM NaCl, and 1% SDS.

Preparation of Knockout AdIce

81

3. Methods

3.1. Cloning of the Mouse Genomic Gene 3. I. 7. Genomic Library We are currently using genomic libraries from the ES cell lme E14TGa m h phage vectors. This ES cell lme 1sderived from strain 129 mice. Isolation of genomic clones from the ES cell lme used for targeting is recommended to mmimize polymorphic differences between the targeting DNA and the endogenous gene. Other authors have successfully prepared animals using commercial mouse genomic libraries from Stratagene (La Jolla, CA) or Clontech (Palo Alto, CA). If the gene is not polymorphic among different mouse strains, libraries can be purchased from the aforementioned compames, otherwise, they should be prepared as described m refs. 3 and 4. 3.1.2. Finding Wodwtg Condrtions To isolate a mouse genomic clone, several strategies are possible. If the mouse cDNA is available, the best option is to prepare a cDNA probe and use to screen However, it is desirable to verify first that this cDNA probe does not hybridize to repetitive sequences m the selected hybridization condtttons. To verify such a possibility, the most appropriate procedure 1sto carry out a Southern analysis. 1, 5 c(gof mousegenomic DNA are digestedby five common restriction enzymes (for example,EcoRI, BarnHI, HlndIII, &I, and SacI) m 25 pL of total reaction volume for 3 h at 37°C Followmg digestion, 2 pL of loading buffer are added to the reaction tube, the complete volume IS loaded into a 0 8% agarose gel m 1X TBE and run in 1X TBE for 20 h at 2 1 V. In a separated lane, load 0 5 Pg of a DNA marker (generally h digested by HindIII) After runnmg, the gel IS stamed in 50 l.tg/mL ethidmm bromide for 20 min and destained in dtsttlled water for 5 mm 2 The gel IS denatured m 0 5 NNaOH, 1.5 MNaCl for 1 h and transferred to nylon membrane (Hybond-N, Amersham, Arlmgton Hetghts, IL) in 20X SSC as described in ref. 3. Next day, membrane is dried at room temperature and crosslinked for 2 mm by exposure to ultraviolet (UV) light. 3. Membrane is transferred to a plastic bag and prehybrtdized m 10 mL of prehybridatton solution at 42’C for 1 h. Random prtmmg of 20 ng of probe is simultaneously carried out using Boehrmger Manheim (Indianapolis, IN) Ktt and 30 PCI of [32P]- dCTP (Amersham) at 37°C One hour later, 200 p.L of 10 mg/mL ssDNA are added to the labeling reaction and denatured wtthout purification at 95°C for 10 mm Followmg this denaturation, reaction is chilled on ice, mixed with 8 mL of hybrtdtzatton solutton, and added to the bag containing filter once prehybrtdization solution had been poured The bag is sealed without bubbles and left at 42°C for 18 h.

82

Osada and Maeda

4. Filter is washed twice m 3X SSC, 0 5% SDS at 68’C Once washed, filter 1s covered with Saran wrap and checked with a Geiger counter to confirm washing. In 1X scale, radioactivity should not be htgher than 1 Otherwtse, further washmg is required. Falter is exposed to autoradiographic film (Kodak X-omat, Kodak, Rochester, NY) with intensifiers for several days. 5 Film is developed, following Kodak processing mstructions, and analyzed to confirm presence of spectfic bands and size Smears mean that the complete cDNA is not specific because of the presence of repetmve sequences The cDNA should be cleaned of poly (A) tails and repetitive sequences by restrictton analysts and separation of fragments until single bands appear on the genomic Southern A careful selection of probe may save a lot of time. If mouse cDNA IS not available, the next choice IS rat cDNA When both are not available, human may be used However, m this case, smtabilny of the probe should be checked first on a human genomic Southern analysis and later tested on that of a mouse A positive result with the first and a negative result wtth the second m the aforementtoned conditions are mdtcative of genetic heterogeneity, and less restrtctive conditions m hybrtdtzatton and washing temperatures should be considered A series of experiments are necessary to find the conditions when a certain probe recognizes specific mouse genomic sequences

3.1 3. Screening Procedure A million independent phages infect bacteria and lysis plaques are immobilized on nylon membranes following procedures described m ref. 3; this represents 20 filters of 15-cm disks. These filters are hybridized to the aforementioned prepared probe in the experimentally defined conditions. A second set of copy filters may be prepared to avoid false-positive. Only positrve spots m original and copy filters should be considered and plugs of the area correspondmg to those spots m the plate should be eluted in 1mL of SM. Successive rounds of mfection, transference, and hybridation are carried out until isolated positive clones are obtained (3). DNA from these phages will be prepared as described m ref. 3, digested with several restriction enzymes, and blotted to membranes. Hybrrdizatron to different fragments of cDNA will provide mformation to locate coding regions within each clone. Fragments will be subcloned into plasmids for further characterization. A restrrctron analysis with all enzymes of plasmid polylinker will provide a valuable map and a sequence analysis of a coding fragment will definitively prove the tdentity of the clone. 3.2. Designing the Targeting Plasmid Most constructs (see Note 1) use the positive/negative selection developed by Mansour et al (7). In a replacement strategy, the bacterial nedR gene, flanked by homology

fragments,

IS Inserted mto one of the codmg regions to disrupt

the gene and as positive selectable marker. The promoter of this gene is gen-

Preparation of Knockout Mice NotlH

RH

K

83 R

B

BH

B

Sal I

Not I

Fig. 1 Design of a construct in a single tube legation All fragments have cohesive and complementaryends H, HindIII; K, KpnI; R, EcoRI; B, BamHI

erally the synthetic mutant polyoma-enhanced HSVtkl’ (MCI) (6), which is a weak promoter. The plasmld pMC1 neopolA may be purchased from Stratagene. As negative selection, the herpex virus thymidzne kinase gene contaming the polyoma virus enhancer (MC 1) is used. This plasmid, pMC 1tk, may be also obtained from Stratagene. This gene has to be placed outside the homology fragments. We currently prepare constructs m a single step of hgation, as shown in Fig. 1. All DNA pieces, short- and long-length fragments of homology, neo and tk genes are cut to have complementary and different restriction sites. All those inserts should be purified previously. 1. Run digestion reactions in a 1% Sea-plaque(FMC, Rockland, ME) preparative agarose gel m 1X TBE Recovery of fragments IS performed by melting the corresponding agarose at 65°C. Once melted, add a l/10 th of volume of 5 MNaCl and an equal volume of neutral phenol and let equilibrate to this temperature for 5 mm Vortex and spm at 12,OOOgfor 10 mm. Take upper layer and extract with an equal volume of chloroform: isoamyl alcohol Vortex and spin as noted Take upper layer and extract with an equal volume of washed ether. Once spun, the ether, top layer IS discarded. Aqueous phase is precipitated by adding an equal volume of isopropanol; let precipitate at -7O“C overnight. A centrifugatron of 12,OOOgfor 15 min ~111recover inserts as pellets. Decant supernatants, and wash pellets with 100 pL of 70% ethanol. A second spinning 1s requrred to reattach pellets to the Eppendorf tubes. Once decanted, pellets are dried and dissolved m 5 pL of 1X TE Amount of DNA in each insert IS determined 2. Equal molartty of all mserts is put together with 1 & of 10X ltgatton buffer and 1 pL of T4 DNA hgase(Amersham) in a total volume of 10 pL and left at 14°C overnight.

84

Osada and Maeda

3. Ligation reactton IS diluted l/5. Five pL of diluted hgatton are added to 50 @., of competent DH5a and kept 30 mm on ice (Competent DH5a can be purchased from several companies: Gibco-BRL, Stratagene.) Shock for 2 min at 42’C and 2 mm on ice. Plate 50 pL on a LB agar plate contammg 50 pg/mL ampictllm. 4. Growing colonies are analyzed by picking them with a tooth pick and dissolvmg m 20 pL of sterile STE Once the colony IS dispersed, a security streaking of the colony ts carried out on a square of a multisquare LB agar with ampicillm plate The bacterial suspension IS extracted with 10 pL of phenol and 10 pL of CIA Vortex and spm 10 min at 12,OOOg Upper phase is transferred to a new Eppendorf tube and to this is added 1 uL of 200 pg/mL RNase A free of DNase Tubes are incubated at 55°C for 30 mm 5 Analysis of supercoil plasmids is carried out by runnmg 15 pL on a 1% agarose gel m 1X TBE Those plasmids which supercools have the half size ofthe expected for the whole plasmid are the colonies that are grown on a midi scale to obtain enough DNA to charactertze the plasmid and to prepare the targeting material 6 One colony is grown m 100 mL of LB containmg 50 pg/mL ampicillm overnight at 37°C Then, medmm IS spun for 15 mm at 15OOgand decanted Pellet is resuspended m 5 mL of GTE until a homogenous suspension is formed 5 mL of 0.4 N NaOH and 5 mL of 2% SDS are added and mixed carefully by gentle mversions Lyse bacteria for 15 mm on ice. Add 7 5 mL 5 A4 potassium acetate, mix gently, and leave on ice another 15 mm. Spm for 15 mm at 12,OOOg Carefully transfer supernatant to a clean centrifuge tube, add 23 mL isopropanol, and leave for 30 min on ice. Spur for 15 min at 12,OOOg. Decant and wash pellet with 10 mL 70% ethanol. Spin for 5 mm at 12,OOOg.Decant and dry pellet. Resuspend m 300 pL of 1X TE and transfer to an Eppendorf, add 5 pL. of 200 pg/mL RNase A free of DNase and incubate 30 mm at 55°C. Precipitate with 300 pL of 5 M NaCl and 350 pL 30% (PEG), mix, incubate for 30 mm on ice, and spin for 15 min at 12,OOOg.Decant and wash with 500 p.L 70% ethanol Spin for 5 mm at 12,OOOg Decant and dry Dissolve m 450 pL 1X TE, add 50 l.& 5 MNaCl, extract 250 pL phenol and 250 pL CIA Vortex Spin for 10 mm at 12,OOOg Take upper phase, extract with 500 pL of CIA, vortex, spin for 5 mm at 12,OOOg Take upper phase, add 1 mL ethanol, and precipitate 30 min on ice Spur for 15 mm at 12,OOOg Spin, decant, wash with 200 pL 70% ethanol, spin for 3 mm at 12,OOOg, decant, dry, and dissolve in 200 pL 1X TE Estimate concentration by measuring absorbance at 260 nm of an aliquot 7 Linearize 20 pg of plasmid DNA with the approprtate restriction enzyme for 1 h at 37’C Precipttate with l/lOth volume of 3 Msodmm acetate and 2 vol of ethanol for 15 mm at -70°C Spin for 15 mm at 12,OOOg,decant, wash with 100 pL of 70% ethanol, spur for 5 min, decant, and dry inside a tissue-culture hood DISsolve m 20 & of sterile 1X TE prepared for ttssue culture

3.3. Transfection of the Construct into Mouse ES 3.3.7. Growing of Mouse ES Cells To maintain ES cells m their undifferentiated state (see Notes 2-5), we currently grow them on a feeder of primary culture of mouse embryomc fibro-

Preparation of Knockout Mice

85

blasts transgenic for the neORgene (Available from The Jackson Laboratory, Bar Harbor, ME). 3 3.1 .l

ISOLATION OF MOUSE EMBRYONIC FIBROBLASTS

1 Sacrifice a 14- to 16-d pregnant mouse. Aseptically remove uterus and place it m a tube contaimng sterile PBS Next manipulations should be carried out m a trssue culture hood 2 Rmse blood of the uterus by transferring uterus to a clean tube contammg sterile PBS 3 Place each piece of uterus contammg an embryo on a Petri dish with PBS DISsect the piece discarding amniotic sac and placenta Transfer the embryo to a clean Petri dish with PBS 4. Dissect mouse embryo and remove all internal organs If embryos are older than 16 d, remove the head Rinse carcass several times in fresh PBS 5. Place carcass m a Petri dish and using scalpel, mince the carcass. Add 1O-l 5 mL trypsm solution and transfer tissue suspension to sterile Erlenmeyer flask containing 25 mL of trypsin solution. 6 Incubate flask at 37°C for 30 mm with gentle shaking 7. Transfer cell suspension to two 50-mL Falcon centrifuge tubes and add equal volumes of DMEM-H medium containing 10% FBS (Gibco-BRL) 8. Centrifuge at 3008 for I5 mm Aspirate supernatant and resuspend both tubes m 5 mL of fresh DMEM-H containing 10% FBS Pool all cells into one tube 9 Seed the 10 mL m one loo-mm Petrt dish; one embryo per Petri dish. 10 Incubate at 37°C m a humtdttied incubator m 5% COZ, 95% an. 11 Change medium 24 h later and let grow until they reach almost confluency 3.3 1 2. MOUSE EMBRYONIC FIBROBLAST FREEZING PROCEDURE 1 Remove the medium and wash two times the plate with 1X PBS 2 Add 1 mL of trypsm solution. Gentle swirl to cover whole plate Incubate at 37°C for 2-3 mm. 3. Inactivate trypsm by addmg 4 mL of DMEM-H medium with 10% FBS. Transfer cell suspension to a 15-mL Falcon tube and spin 3008 for 10 mm. Aspirate supernatant 4 Resuspend pellet m 1 mL of DMEM-H with FBS. Complete volume to 10 mL with the same medium 5 Seed 1 mL per Petri dish (100 mm) containing 10 mL DMEM-H with FBS This is passage number 1 (PI) Once cells are confluent, trypsmize, repeating steps l-3 6 Resuspend pellet from each plate m 1 mL of freezing medium and storage m one cryovtal in a liquid-nitrogen contamer 3 3.1.3

PREPARINGFIBROBLASTFEEDERLAYERS

1. Thaw one vial m a 37°C bath until a little bit of ice is on top 2 Dilute each vial into 10 mL of DMEM with 10% FBS. Spin for 10 mm at 300g Aspirate supernatant.

Osada and Maeda

86

3 Resuspend pellet m 1 mL of DMEM with FBS Complete volume to 10 mL and seed m one loo-mm Petri dash This 1s considered passage 2 4. Spht cells in ratro 1.5 Thts new passage is done followmg steps l-5 of Subheading 3.3.1.2. This is passage 3. 5. Grow to confluence and split m ratto 1.4. This gives passage 4 and a total of 20 Petri dishes 6 Aspnate medra from the feeders Wash twice cells with 5 mL of PBS. 7. Add 1 mL trypsin solutton/plate. Incubate at room temperature for 3-5 mm 8 Add 8 mL of 5% FBS DMEM-H medium to one of the plate 9 Pipet cells and medium mto the next plate and resuspend cells completely 10 Pool all cells into a 50-mL Falcon tube. Spin at 300g for 10 min. 11 Aspirate medium Resuspend pellet first m 1 mL of 5% FBS DMEM-H medium and complete volume to 20 mL with the same medium Use an ahquot to determine cell concentration 12 Irradiate cells by exposure to 3000 rads of y-nradratton 13 Seed 2-2.5 x lo6 cells/100 mm plate. Appropriate volume is added to each Petri dish containing 10 mL 5% FBS DMEM-H medium Irradiated tibroblasts should be used wrthm 7 d Contaminated feeders should not be used. 3.3.1 4. MANIPULATION OF ES CELLS

We currently use BK4, a subclone of E14TGa that comes from 129 mouse stram. Other ES cell lmes are avallable from different research groups (ref. 5) or from Amerxan Type Culture Collection (ATCC). 1 Transfer frozen vial to a 37°C bath until ice appears on top 2 Dilute the vial m 10 mL of ES culture medium m a 15mL tube. 3 Spin for 10 mm at 300g Aspirate medium. Resuspend m 5 mL of ES culture medium 4 Split m ratto 1 10 by seedmg 0 5 mL m 10 mL of ES culture medium added to a loo-mm Petri dish containing a fibroblast feeder layer. In total, 10 plates are seeded. Plates should be labeled with type of cell Ime, passage, and date. 5. Twenty-four hours later, change medium. After 2-3 d of culture, cells become confluent and need to be passaged 6 Aspirate medium and wash dish three times with 5 mL PBS. 7 Add 1 mL trypsm solution Swrrl to cover entu-e plate Incubate at 37’C for 3-5 mm. 8 Add 4 mL ES medium to stop trypsm action and gently resuspend cells 9. Split in ratio 1: 10 0 5 mL of resuspended cells are added to a new 100~mm Petri dish containing 10 mL of ES culture medium and a fibroblast feeder layer 3.3.1 5 ES CELLS FREEZING PROCEDURE

To save cell line, confluent 100~mm Petri dishes should be frozen. Follow steps 6-8 of Subheading

3.3.1.4.

1 Transfer cell suspension to a centrtfuge Falcon tube and spm at 300g for 10 min. 2. Aspirate medium Resuspend cells m 1 mL of freezmg medmm.

Preparation of Knockout Mice

87

3 Transfer cell suspenston mto cryovtals. 4. Place vials into Styrofoam box and freeze at -70°C for 24 h 5 Transfer vials to hquid nitrogen for long-term storage Label each veal with cell line, date, and passage.

3.3 1.6. ELECTROPORATION OF ES CELLS 1 One or two 100~mm Petrt dishes are trypsmized as described in steps 6-g from Subheading 3.3.1.4. 2. Inhibit trypsin reactton by adding 10 mL ES cell medium and resuspend cells 3. Combme medta from both plates 4. Take an aliquot and dilute l/l0 in PBS to count cells 5. Spin for 10 min at 300 g to pellet cells. 6 Resuspend cells in 1 mL of PBS and complete volume to 10 mL with PBS. 7 Spm again at the aforementtoned condtttons Aspirate medium 8 Resuspend pellet m 0.5 mL PBS. Usually 3-7 x 1O’cells are used to electroporate 9 Transfer the 0 5 mL cell suspension together with 20 pL lmeanzed plasmtd (1 pg/uL) to the electroporation cuvet (Bto-Rad) and cover with the lid Mtx well. 10 Apply a pulse of 300 V, 200 pF for 999 ms 11. Mtx cells wtth 12 mL ES culture medium 12 Transfer the 12 mL to twelve 100-mm Petri dishes containing 10 mL of ES culture medium and a fresh fibroblast feeder layer 13. Allow cells to recover for 24 h. 14 Aspirate media and replace by ES culture medmm contammg 200 pg/mL G4 18 and 2 @4 ganciclovn. Let grow m this medmm for 10 d. Change medmm if tt starts turnmg yellow Generally 100-200 colornes/plate may be observed as survivors of double selectton.

3.4. Selection of C/ones that Underwent Homologous Recombination 3.4.1. Preparation of a Fibroblast Feeder Layer To prepare the fibroblast feeder layer for a 24-well plate: 1. 2. 3 4.

Take a loo-mm plate dash covered with trradtated fibroblast. Aspirate medium Add 1 mL trypsm solutron and swirl gently to cover whole plate Incubate 2-3 min at room temperature Inactivate trypsm with 12 mL ES culture medium containmg G-41 8 and ganciclovtr Mix well. 5 Transfer 0.5 mL to each well of a 24-well plate.

3.4.2. Picking and Expansion of Colonies 1 Place the 100-n-m Petri dash containing double survtvors on the surface of an Inverted mtcroscope 2 Observe with a low-power objective the position of a colony.

88

Osada and Maeda

3. Remove plastic lid 4 Sterilize a finely drawn glass mtcroptpet (Becton Dickinson, Rutherford, NJ), attached to a tube connectmg to a mouthpiece, by aspirating 70% ethanol, rinse with sterile PBS 5 Using the microscope and mtcroptpet, dissect the colony m two parts, half of colony is transferred to a well of a 24-well plate wrth a feeder layer and the other half portion to an Eppendorf tube containing 0 5 mL PBS. 6 Repeat the process of picking colonies. In the 0.5 mL of PBS, five Independent colonies are pooled to analyze by polymerase chain reaction (PCR). Generally, 100 colomes are picked m one experiment 7. The portion of colony m the 24-well plate is left to attach for 1 d To avoid dtfferentiatron and facilitate growth, colony 1s trypsmrzed. 8. Remove medium. 9. Wash twice wrth 1 mL PBS 10. Add 0.1 mL trypsm solutton, swirl to cover. Incubate for 2-3 mm at 37’C 11 Stop trypsin by adding 0 9 mL ES culture medium containing G-418 and gancrclovn. It will take approx 7 d to become confluent

3.4.3. PCR Screening of Pools 1 Spin Eppendorf tubes containmg five independent colonies m PBS at 300g for 10 mm. 2 Aspirate PBS 3. Resuspend cell pellet m 1 mL of PBS. Spin again m the same conditions 4 Resuspend cells m 50 pL of sterile drstrlled water 5 Freeze cells at -70°C overnight. 6 Boil tubes at 95°C for 10 mm 7 Add 1 pL of 10 mg/mL proteinase K in 1% SDS solution Incubate for 3 h at 65°C 8 Inactivate protemase K by placing tubes in a boilmg bath for 10 mm. 9 Chill on ice Spin 5 for mm to recover volume. Keep on ice 10 Take 10 pL of cell homogenate and add 15 pL containing PCR buffer, nucleottdes, primers, and Tuq polymerase. Fmal concentrations: PCR buffer (16.6 mA4ammonmm sulfate, 67 mM Tris-HCl, pH 8 8, 6 7 mA4 MgCl,, 5 mM P-mercaptoethanol, and 6 7 @4 EDTA), 1 rnA4 each nucleotide, 1 luV each primer, and 0 2 U Taq polymerase Conditions used for this particular PCR product were at 93°C for 1 mm, then 40 cycles of 93°C for 30 s, and 65°C for 10 min Primer 1 should correspond to a region outside the region of homology that is m the target locus and not in the construct, and primer 2 should correspond to neo gene from the incoming DNA, but not in the genome Only when homologous recombmation takes place, these primers will be next to each other, allowing PCR amphficatron of a predicted-size DNA fragment. 11 Analyze PCR products by running a 1 2% agarose gel m 1X TBE. Use approprrate DNA markers 12 Stain gel m 500 pg/mL ethtdium for 10 min and destam for 10 mm m d&led water. 13 Only wells growing ES cells correspondmg to positive pools will be expanded and re-analyzed (see Note 6).

Preparation of Knockout Mice

89

3.4.4. Expansion of Clones and DNA Preparation for Southern 1 Remove medmm of positive and confluent wells of the 24-well plate descrtbed m Subheading 3.4.2., step 11. 2 Add 0.1 mL of trypsin solution and incubate for 3-5 min at 37°C. 3 Inhibit trypsin with 0 9 mL ES culture medmm wrthout selection anttbtottcs. 4 Splat in ratio I * 10 to grow 0 1 mL are placed for growing m a new 24-well plate with feeder fibroblasts 5 Spin the remaining 0.9 mL at 300g for 10 min. 6 Aspirate medium 7 Resuspend pellet m 1 mL PBS 8 Spm at 300g for 10 min Aspirate supernatant. 9. Repeat wash procedure 10. Resuspend pellet m 100 pL of ES lys~s buffer 11. Incubate at 55°C overnight 12. Next day, add 33 pL 6 MNaCl MIX well by inverting. 13 Spin at 12,OOOg for 15 mm to sediment cellular detritus. 14. Carefully transfer supernatant to a clean Eppendorf tube 15. Add 200 pL of ethanol and mtx well. 16. Ftsh the filamentous DNA and transfer to a clean Eppendorf tube. 17 Let dry and dissolve m 25 pL 1X TE by heating at 65°C for 3 h 18 Analyze 1 pL by PCR 19 Posttive PCR clones are analyzed by digesting 12 @+with the appropriate restrtcnon enzyme and Southern analysis as described m Subheading 3.1.2. 20 Positive wells of Subheading 3.4.4., step 4, when confluent, are split m ratio 1:3, repeating steps l-3 2 1 Seed 0.3 mL m 6-well plate with feeders and 3 mL ES medium m each well The triphcate wells of each positive clone are prepared for analyzing DNA, mtcromJection, and storage in hquid nitrogen. 22. Remove media from confluent 6-well plate. 23 Wash three times with 3 mL PBS. Aspirate PBS 24 Add 0.4 mL trypsm solutions and incubate for 3-5 min at 37°C 25. Stop trypsin with 1.1 mL of ES culture medmm 26 Pellet cells at 300g for 10 mm. 27. Resuspend cells* (I) wells prepared to analyze DNA, step 21, m 100 pL of lys~s buffer and proceed as m steps 11-17 Digest 1 pL wtth the appropriate restriction enzyme and Southern analysis as descrtbed m Subheading 3.1.2. 28. Resuspend cells: (II) wells containing cells to storage, step 21, m 1 mL of freezmg medium, place m a Styrofoam container, and leave -70°C overnight Transfer vial to a liquid nitrogen container for prolonged storage (see Notes 4-5). 29. Resuspend cells: (III) wells containing cells to be injected, step 21, in 5 mL of ES culture medium.

3.5. Microinjection

of ES Cells into Biastocysts

1 On d 3 5 after postcoitus, C57BW6 naturally mated females are euthamzed with 0 5 mL of 2 5% avertm (w/v)

Osada and Maeda

90 2 Open the abdommal

3

4

5 6 7. 8 9

10 11 12. 13

cavity. Cut the uterus at the cervix and pull the uterus upward while cutting mesometrmm. Liberate uterus by cutting the uterus below the junction with the oviduct Useful diagrams explammg the procedure may be found in ref. 5 Place uterus in 0.5 mL M2 medium m a 35-mm Petri dish. Flush each horn with 0.2 mL M2 medium using a 25-gage needle attached to a 1-mL syringe to liberate blastocysts Collect blastocysts with the aid of a mouth-controlled ptpet and a dtssectmg microscope and transfer to clean drops of M2 medium covered with liquid paraffin in a 35-mm Petri dish. Incubate blastocysts for 2-3 h at 37°C m a humidified incubator to increase blastocoele cavity. Transfer ES cells into ES medium drops of the 35-mm Petri dash. Place the Petri dish contammg ES medium drops and blastocysts onto the microscope/micromampulator setup. Using the mjection pipet, pick 1O-l 5 ES cells with suction Immobtltze a blastocyst with the holding ptpet placing the embryo mass along the end of the pipet Bring the injection pipet mto the same focal plane of the blastocyst equatortal plane Penetrate the trophectoderm and Insert the tip between two cells of the inner mass of the embryo Expel 7 ES cells and withdraw the injecting pipet Transfer the inJected blastocysts to a clean drop of M2 medmm and incubate l-2 h at 37°C m a humidified mcubator.

3.6. Transference

of Blastocysts

into Pseudopregnant

Mothers

1 Mate CD 1 females to vasectomized males. 2 Check everyday for the presence of vaginal plug. Use pseudopregnant females 2.5 d later. 3 Anesthetize female by an intraperitoneal injection of 0 3 mL of 2 5 % (w/v) avertin solution 4 Wipe the back of the animal with 70% ethanol and shave 5 Make a small transverse mctsion about 1 cm left of the spinal cord at the level of the first lumbar vertebra. 6. Retract skm to visualize ovary and fat pad and make an inctsion through the abdominal wall. 7. Exteriorize ovary and one-third of the uterus and clamp fat pad to hold. 8 Load a transfer ptpet with two air bubbles, 12-20 blastocysts m M2 medium, and finally another air bubble. 9. Puncture the tip of the uterus near the uterotubal junction avotdmg small blood vessels until the lumen has been reached 10 Blow the transfer pipet until the two bubbles have entered the lumen Withdraw transfer pipet. 11 Replace reproductive organs mto the abdominal cavity.

Preparation of Knockout Mice

91

12. 13 14. 15. 16.

by

Suture body wall and close skm with wound clips. Leave recover the animal from anesthesia m an isolated cage. Remove wound clips 7-10 d after surgery. Check cage 17-l 8 d postsurgery for the berth of the offspring On d 21 postpartum, wean pups, identify them, and analyze the chimerism checkmg the extent of the agouti color on the coatmg skin

3.7. Germline Transmission of Chimeras 1. Two months after partum, mate 50% or more chimertc animals wtth normal C57BL/6 animals. Usually, one chimeric male with two normal females 2. Check cages daily on d 20-25 after mating for the presence of dellvery. 3. Euthanize black pups and save the agouti ones. 4. Wean agouti pups when they are 21 d old Identify them, separate males from females m different cages, and do not leave more than five animals together m the same cage 5 Cut a small piece of tail to analyze DNA.

3.7.1. Preparation of Tad DNA 1 Place the tail tip in 200 pL of tail lysrs buffer, add 10 & of 10 mg/mL protemase K, and incubate 24 h at 42’C with gentle inversion 2 Twenty-four hours later, add another 10 & of fresh protemase K and contmue mcubation for 24 h 3 Add 50 + of 5 M NaCl and mix vtgorously without vortexmg by mvertmg several times 4. Spur for 15 mm at 12,OOOgto pellet detritus 5 Transfer supernatant to a clean Eppendorf. 6. Add 200 pL of ethanol and fish precipitated DNA to a clean Eppendorf tube 7 Dry filamentous DNA and dtssolve in 100 & of 1X TE by leaving tubes at 65°C for 4 h. 8. Dtgest 3 pL wtth the appropriate restriction enzyme and perform a Southern analysis 9 Check autoradiogram for the presence of a mutated allele. Animals having normal and mutated alleles are heterozygous ammals of F 1 generation.

3.8. Obfention 1. 2. 3. 4 5 6

of Homozygous

Animals

Breed 2-mo-old heterozygous animals. Place 1 male wtth 2 females m the same cage Check cages daily on d 20-25 for the presence of babies. Follow carefully survival of all animals. Wean animals at d 2 1. Identify them and separate males from females m different cages. Cut a small piece of tail to genotype as described in Subheading 3.7.2. Analyze frequency distribution of homozygous, heterozygous, and wild-type animals 7. Characterize phenotype of homozygous ammals to establish the influence of the absence of this particular gene on hpoprotem metabolism

92

Osada and Maeda

4. Notes 1 Different types of constructs are possible A careful descrrption of this point IS beyond the goal of this manuscript. More detailed information may be found m refs. 1, 6, and 7. 2 To maintain ES cells in then undifferentiated state is the most crucial step. ES cells should be seen as small colonies where no mdividual cells are distmguishable Other presentation may indicate differentiated cells or a different cell culture Growmg cells on feeder cells is not the only possrbihty They can be cultured in the presence of Leukemia mhibitory factor (LIF) The latter procedure requires more experience to recognize the morphology of cells Some workers combine both procedures to achieve a higher level of security in keepmg cells undifferentiated (see refs. 5 and 6). 3 ES cells are sensitive to trace contammants m media Glassware should be absolutely clean; alternatively, plastic disposable materials can be used 4 Although ES cells can theoretically be cultured mdefimtely, their karyotypes may change m long-term culture Assessment of number of chromosomes should be carried out 5 ES cells should be cultured m antibiotic-free media For this reason, mycoplasma contammation should be tested m ES cell lines before micromJection 6 Extreme and mettculous procedure should be carried out to avoid contammation of PCR reagents, pipets, and plasticware with previous amphfied DNA Positrve and negative controls are recommended for use m any screenmg procedure

References 1 Koller, B. H and Smithies, 0. (1992) Altering genes m animals by gene targeting Ann Rev Inmunol

10,705-730.

2 Palmiter, R. D and Brinster, R E (1986) GermLme transformation of mice Ann Rev Genet 20,465-499

3. Sambrook, J., Fritsh, E. F , and Mamatis, T. (1989) Molecular Clonzng. A Laboratory Manual 2nd ed Cold Sprmg Harbor Laboratory, Cold Sprmg Harbor, NY 4 Asubel, F. M., Brent, R., Kingston, R. E , Moore, D 0 , Seldman, J. G , Smith, J. A , and Struhl, K (1989) Current Protocols zn Molecular Bzology Greene Publishing and Wiley-Interscience, NY 5 Hogan, B., Beddington, R., Constantim, F., and Lacy, E (1994) Manzpulatzng the Mouse Embryo A Laboratory Manual Cold Sprmg Harbor Laboratory, Cold Spring Harbor, NY 6 Joyner, A L (1993) Gene Targeting A Practical Approach IRL, Oxford, UK 7. Mansour, S. L., Thomas, K. R., and Capecchi, M R (1988) Disruption of the proto-oncogene mt-2 in mouse embryo-derived stem cells. a general strategy for targeting mutations to non-selectable genes. Nature 336,348-352.

Fast Ultracentrifugation Methods for the Separation of Plasma Lipoproteins Jose M. Ordovas 1. Introduction The presence of liptds wrthm the hpoprotem particles confers thesemacromolecular complexes with a lower density compared with other serum protems. With the arrival m the 1940s of the analytical ultracentrifuge, these partrcles could be separated from other plasma protems as a dtscrete peak. During the followmg years, it was shown that this peak was made up of a wide spectrum of particle stzesand denstties ranging from 0.92 to 1.21 g/mL (I). Ltpoprotems were classified mto four major classes:chylomrcrons (exogenous trtglycerrde rich particles of d < 0.94 g/mL), very low density hpoprotems (VLDL, endogenous triglyceride rich particles of d = 0.94-1.006 g/mL), LDL (cholesteryl ester rich particles of d = 1.006-1.063 g/mL), and HDL (particles contammg approx 50% protein of d = 1.063-l .21 g/mL) With subsequent tmprovements to the ultracentrimgation techniques, further heterogeneity was detected within each of those major lipoprotein classesthat resulted m the need for further subdtvtston into several density subclassessuch as HDL2a (d = 1.1O1.125 g/mL), HDL2b (d = 1.063-l. 10 g/mL), and HDL3 (d = 1.125-I .2 1 g/mL). The separation of lipoproteins by ultracentrifugation has been essential for the advances in the field; however, the traditional ultracentrifugation techniques were time-consuming, costly, and required considerable hands-on ttme. Moreover, they usually required relatively large amounts of plasma or serum, usually between 5 and 10 mL. Some additional problems associated with these techniques were the long exposure of samples (up to 72 h) to very htgh g-forces and salt concentrations. These circumstances led to considerable alteration of the hpoprotem particles resulting m loss of several apohpoproteins, From

Methods Edited by

m Molecular J M Ordovas

Bology,

Vol 110 Lpoprotetn

0 Humana

93

Press

Inc , Totowa,

Protocols NJ

94

Ordovas

primarily apo AI, apo AIV, and apo E from the respective lipoprotein fractions (2,3). The long periods of time and the conslderable mampulatlon of the samples may also result in bacterial contamination and m oxidation of the lipid and protein components of the lipoprotein particles, especially those m LDL For these reasons, several improvements have been Introduced m the last few years, conslstmg primarily m the use of much shorter run times and lesser amount of sample (&5j. This progress has been made possible due to the advances m hardware, primarily from faster ultracentrifuges and rotors designed for higher speeds and with more optimal geometry. Preparative ultracentrifugation of lipoproteins has been used for several purposes, Large-scale separation of hpoprotems has been carried out to isolate, purify, and characterize hpoprotem particles. At a more reduced scale, ultracentnfugation has been used for many years m speclahzed clmical laboratories to carry out the measurement of lipids within each lipoprotein fraction to help m the dlagnosts of lipoprotein abnormalities. Nowadays the latter apphcatlon has been substituted for the most part by alternative methods that are less tlmeconsuming and more amenable to automation. However, separation of llpoprotem fractions continues to be an essential procedure in laboratories devoted to basic hpoprotem research. This chapter will describe the methodology used to isolate hpoprotem fracttons using ultracentrifugatlon protocols that reduce time and sample requirements. 2. Materials 2.1. Ulfracenfrifuge (see Note 7)

(Beckman Instruments,

Palo Alto, CA)

The protocols described m Subheadings 3.1. and 3.3.1. are designed to use the L8 80M ultracentrifuge. For that described in Subheading 3.3.2., we recommend the Optima TLX bench-top ultracentrifuge. For all other protocols described, the TL- 100 1srecommended. 2.2. Rotors (Beckman 1. NVT90 or NVTl 00. 2 3

Insfrumenfs)

TLA-1202 TLA-1004

2.3. Ultracentrifuge Tubes (Beckman Instruments) 1 Beckman l/2 x 2-m. optlseal polyallomer tubes, Beckmanno 362185, 4 9-mL nominal tube capacity 2 Thick-wall polycarbonatetubes(cat. no. 343778), 1-mL nommal tube capacity. 3 Quick-Seal Polyallomer (cat. no. 362248), 5.1-mL nominal tube capacity

Plasma Lipoprotein Separation Methods 2.4. Other Equipment 1. 2. 3. 4 5. 6 7 8 9.

95

and Materials

Mark II Abbe 10494 Refractometer MISCO Products Division (Cleveland, OH). Quick-Seal Tube Sealmg Kit (Beckman cat. no. 345529). CentriTube Slicer (Beckman cat. no. 347960). Fractton recovery system with flat-bottomed or hollow-cone caps (Beckman cat no. 343890 or 342023 Through 342026). Fraction collector system (GradiFrac System, Pharmacia Biotech) 0 2-p low-protein bindmg filters (Millex-GV, M&pore Corporation, Bedford, MA) PD- 10 columns (Pharmacia Biotech) Cholesterol enzymatic kit (Sigma, St LOUIS, MO). Triglycertde enzymatic kit (Sigma).

2.5. Reagents 1. Potassmm bromide (KBr, mol wt 119.0, P 0838, Sigma) 2. 10% EDTA: 10 g of Na,EDTA * 2H20 (mol wt 372.2, E1644, Sigma) adjust the volume to 90 mL with double-distilled (dd) H,O, adJUSt the pH to 7 0 and fimsh the volume adlustment to 100 mL. 3. KBr (d 1.006 g/mL). 10 1 g of KBr, 5 mL of 10% EDTA, pH 7.0 Adjust the volume to I L with ddH,O (see Note 2). 4 KBr (d 1 019 g/mL). 30.6 g of KBr, 5 mL of 10% EDTA, pH 7 0 Adjust the volume to 1 L with ddH20 5. KBr (d 1 063 g/mL) 90 3 g of KBr, 5 mL of 10% EDTA, pH 7 0. Adjust the volume to 1 L with ddHzO 6 KBr (d 1.21 g/mL) 316.6 g of KBr, 5 mL of 10% EDTA, pH 7 0. Adjust the volume to 1 L with ddH*O 7 KBr (d 1.35 g/mL). 5 13 g KBr, 5 mL 10% EDTA, pH 7 0. Adjust the volume to 1 L with ddHzO 8 0 154 MNaCl(9 g/L) To prepare 250 mL of this solution weight 2.25 of NaCl (mol wt 58.44, S 7653, Sigma), dissolve m ddH,O and adjust to 250 mL using a volumetric flask. Add approx 1 g of Chelex (Sigma) and stir for 2-3 h After this period, allow the resin to settle down for a couple of hours (it can be left overnight) and filter the solution through a 0.2-u filter. Store this solution a 4°C. 9. PMSF stock solution 250 mM In isopropyl alcohol. 0.435 g PMSF (Phenylmethylsulfonylfluoride, mol wt 174.2, P 7626, Sigma) to 10 mL in isopropyl alcohol (mol wt 60 10,10398, Sigma) requires gentle heat and mixing to dissolve crystals Solution will recrystalhze over time It should be gently warmed before use m order to dissolve the crystals. 10. PMSF salme. Put 50 mL saline in 50 mL orange-top tube, add 100 p.L of PMSF stock solution and 250 pL of 10% EDTA. White precipitate will form, mix on a rocker overnight to dissolve. PMSF final concentratton. 0.5 nul4. 11. PBS, phosphate buffered salme. 0 15 MNaCl, 10 mMNaHP, pH 7.4 Dissolve 8 g NaCl, 0.2 g KCl, 1.44 g Na2HP04, and 0.24 g KH2P04 m 900 mL of distilled water, check the pH (adJust if necessary), and bring the volume to 1 L

Ordovas

96 3. Methods 3.1. Density-Gradient Ultracentrifugation Using a Near- Vertical Rotor (N VT-90) 3 1.1. Sample Preparation

1 Obtain the blood by vempuncture with EDTA (1 mg/mL blood) as anttcoagulant. 2 Separate the plasma by low speed centrifugation (25OOg, 10 min, 1S’C) (see Note 3) The presence of EDTA partially inhibits some proteases, thus decreasmg protein degradation, however, for several applications, such as the rdentrficatron of variants or truncated forms of apoltpoproteins, It IS essential to prevent degradation and rt is recommended to treat the samples wrth protease mhtbttors as soon as possrble (see Note 4).

3 7.2. Density Adjustment The followmg protocol is designed to be used with Opttseal polyallomer tubes and the NVT-90 rotor. The nomtnal capacity of each tube ts 4.9 mL Optrmal separatton of lipoproteins is achteved with plasma volumes ranging between 0.8 and 1.0 mL. The preparation of the samples varies slightly depending on then volume. For plasma samples with volumes greater than 0.8 mL, the following twostep density gradients should be used: 1. Adjust plasma densrty to 1.35 g/mL by addmg 0 574 g KBr per mL of plasma 2 Subtract the final sample volume from 4.9 mL (maximum tube capacity), and place this volume of PMSF-saline mto the tubes. 3. Wrth a syringe carrying a blunt-tipped, large-gage needle, underlay the density adjusted plasma under the saline. Care should be taken to avoid blowing any bubbles through the solution. Rinse the syringe and needle. For sample recommended:

volumes

less than 0.8 mL, a three-step

density

gradient

is

4 Adjust plasma density to 1.2 1 by adding 0.327 g KBr per mL sample 5 Subtract the plasma sample volume plus 1 mL from the total volume of the tube (4 9 mL), and place this volume of PMSF saline into the ultracentrtfugatton tubes With a syringe and blunt-tipped, large-gage needle, slowly underlay the densityadjusted plasma under the saline, careful not to blow any bubbles through the solution. 6. Underlay 1 mL of 1.35 density solution under the density-adjusted plasma Rinse the syringe and needle 7. Balance those tubes that are going to be on opposite srdes of the rotors very carefully usmg a precisron balance. If needed, adjust weight m the tubes with a Pasteur prpet and PMSF saline

Plasma Lipoprotein Separation Methods

97

Fraction Collector

LN Monitor and recorder

Gradient fractionator

Pump

Fig. 1. Schematic representation of the system used for fraction recovery using the protocol described in Subheading 3.1.4.

3.1.3. Separation 1. Put black caps on tubes, and place tubes symmetrically into the NVT 90 rotor. Put on metal spacers and plastic screw sealers and tighten to 120 in./lb with a torque wrench. 2. Place the rotor into the centrifuge (L8 80M or TL- 100) and set up the following settings: Speed = 70,000 rpm Temperature = 7°C Time = 1 h and 30 min Acceleration = 9 (slowest) Deceleration = 0 (no brake) 3. Start the run and check the progress after 10 or 15 min to be sure the centrifuge is working properly and no warning lights are on.

The entire spin takes approx 2 h. Be ready to fraction the tubes as soon as the centrifuge has stopped, as the gradients are not stable. 3.1.4. Fraction Collection There are several choices for collecting the lipoprotein fractions separated using density gradient ultracentrifugation. In the one described below, a highdensity solution pumped from the bottom of the tube pushes the contents of the tube up to a fraction collector (see Fig. 1).

Ordovas

98

All the reagents and materials used for the collection of the fractions should be prepared in advance and they should be in place when the centrifuge stops. The fraction collector should be programmed to collect about 200~p.L fractions. Plastic eppendorf tubes are suitable to collect and store the fractions. The protocol below describes fraction collection using a fraction recovery system with a hollow-cone cap and a GradiFrac collection System. 1. Remove the tubes carefully not to disturb the lipoprotein layers. Place the tube in the holder. Position top gasket and tighten to create a perfect seal between the rim of the tube and the upper gasket. 2. Turn on the pump at low speed. Wait until there is a continuous flow of the highdensity solution from the needle (Note 5). 3. Quickly screw the needle through gasket and bottom of the tube, making sure the needle aperture clears the bottom of the tube. Turn the pump speed up. 4. Once the bottom fraction has been pushed above, the needle the speed can be increased further. 5. Watch for the sample front to come down through the tubing. When in gets close to the end, start the fraction collector. 6. Once all the contents of the tube have been collected, turn off the pump and stop the fraction collector. It is important to flush the needle and tubing with distilled water as the high concentrations of salt to increase the plasma density tend to crystallize and plug the tubing. Moreover, these are also very corrosive to the equipment (needle, fraction collector).

3.2. Sequential

Micro-Ultracentrifugation

This method takes advantage of the high g-forces that can be reached using the TLlOO ultracentrifuge and the TLl00.2 fixed-angle rotor. Moreover, the sample size required for the separation is reduced to 0.4 mL. As in the classical sequential ultracentrifugation, the several lipoprotein fractions are isolated following a series of sample density adjustments, centrifugation, and infranate/supernate separations.

3.2.1. Preparation of VLDL 1. Place 400 pL of serum or plasma into a 1-mL thick-walled polycarbonate tube. Overlay carefully with 600 pL of 1.006 KBr solution. 2. Place the rotor (TL100.2 fixed-angle) containing the loaded tubes into the TL- 100 and centrifuge with the following settings: Speed = 100,000 rpm Temperature = 16°C Time = 2 h Acceleration= 5 Deceleration = 7 Be ready to fraction the tubes as soon as the centrifuge has stopped.

Plasma Lipoprotein Separation Methods

99

3. Once the rotor has reached full stop, remove it carefully from the centrifuge and get the tubes out trying not to disturb the layers. 4. Set up the Beckman CentriTube Slicer to cut the tube at 0.4 mL point below the top layer (containing the VLDL fraction) (see Note 6). 5. Remove the upper fraction using a transfer pipet or a micropipet trying to recover as much as possible. If some material appears to get stuck to the walls of the tubes, rinse carefully using some of the upper fraction. Minimize the foaming during the rinse. 6. To recover the infranatant (containing the other lipoprotein fractions plus all the other plasma proteins), release the cutter blade and pipet out the remaining solution using the same approach used for the supernate. The material at the bottom of the tube gets very viscous and it should be homogenized with the less viscous material part of the infranatant fraction to reduce losses. Like for the supernatant, this process should be done avoiding the production of foam.

3.2.2. Preparation of Intermediate Density Lipoproteins (IDL) IDL represents a minor lipoprotein fraction in healthy fasting individuals; however, its isolation may be of interest in metabolic studies or in those subjects with lipoprotein abnormalities. 1. Adjust the density of the lower fraction (obtained during the steps described in Subheading 3.2.1.) to 1.019 g/mL by adding 24 & of d 1.35 g/mL solution. 2. Adjust volume to 1 mL using the d 1.O19 g/mL solution. 3. Repeat the centrifugation under the same conditions described for VLDL. 4. Recover the top fraction (containing IDL) and the bottom fraction (containing LDL, Lp(a) and HDL plus plasma proteins) as described for VLDL.

3.2.3. Preparation of LDL 1. Add 94.6 pL of the 1.35-g/mL KBr solution to the bottom fraction from the IDL isolation (Subheading 3.2.2.). The volume of this fraction should be approx 600 p.L. 2. Mix well and bring the volume to 1 mL using the d 1.063 solution. 3. The centrifugation conditions are similar to those described before for VLDL and IDL; however, the length of the run should be increased to 2.5 h. 4. Cut and recover the top (LDL) and bottom (HDL plus plasma proteins) fractions as indicated above.

3.2.4. Preparation of HDL 1. Add to the bottom fraction (600 pL from Subheading 3.2.3.) 675 pL of 1.35 g/mL solution. Place 1 mL into the centrifuge tube and spin for 3 h, maintaining all other conditions as described for the isolation of the other lipoprotein fractions (see Note 7). 2. Recover the top fraction containing HDL as described for the other lipoprotein fractions. The bottom fractions should be free of lipoprotein fractions.

Ordovas

100 3.3. Preparation of LDL for Measurement of Oxidation

Parameters

Ultracentrifugation is frequently used to isolate exclusively the LDL fraction m order to measure Its susceptibility to oxidation. Although the above described protocols can be used successfully for this purpose, for those studies defimng very early stages of oxidation it IS necessary to avoid ltpoprotein peroxidatlon during the purification steps; therefore, stmplified procedures aimed to reduce separation time and sample manipulation are recommended

3.3.1. Isolation of LDL Using the NW90

Rotor

This is similar to the method described above, but it has been optimized simphfied for the isolation of LDL.

and

1 Add 0 490 g of KBr mto a 12 x 75 glass tube, followed by 1.5 mL of plasma 2 Mix carefully with a stirrer or by pipettmg up and down with a mlcroplpet, avolding the formation of foam. This will brmg the density of the plasma to 1.2 1 g/mL The final volume of this solution will be approx 1.7 mL. Samples should be kept on ice during all procedures to diminish hpoprotem modification 3 Add -3 2 mL of the 9-g/L NaCl solution to a 4.9-mL Optlseal tube. 4. Using a ~-CC syringe with a 3-m round-bottom 20-gage needle, lay carefully the density adjusted plasma underneath the NaCl solution. Be careful not to bubble any air through the solutions. After adding the plasma, the tube should be filled already to the top shoulder If this 1snot the case, add more of the 9-g/L NaCl solution 5 Run the ultracentrlfugatlon as described m Subheading 3.1.3. 6 Once the rotor reaches full stop, remove the tubes, carefully trymg not to disturb the layers. Work expeditiously as the lipoprotem bands begin to diffuse once the rotor stops resulting on poor band ldentlficatron and crosscontammation between lipoprotein fractions 7 Locate the LDL band usually found between 0.5 and 0 75 m from the neck of the tube (dlstmgulshable by its intense yellow color) and mark usmg a felt pen the position of the upper and lower boundary of the band on the tube wall. 8 Usmg a Pasteur plpet, remove the solution above the upper mark (contaming the VLDL). With a new Pasteur plpet, remove the LDL layer (to the lower mark). 9 Transfer the LDL to a prechilled ~-CC test tube The LDL solution should be totally clear with no signs of self-aggregation. If the latter 1sapparent, then filter the solution through a 0.2-p low-protein bmdmg.

3.3.2. Rapid isolation of Plasma Lipoproteins with a Benchtop Ultracentrifuge This section describes an alternate procedure to that m Subheading 3.3.1., using a Beckman TLX Optima benchtop centrifuge. Separations are carried out m a TLA 100.4 rotor holding eight tubes with a volume of 5.1 mL each, enablmg lipoprotein preparation from a total volume of 13.6 mL plasma m a single run.

Plasma Lipoprotein Separation Methods

101

1 Obtain the blood by vempuncture with EDTA (1 mg/mL blood) as anticoagulant. 2. Isolate plasma by low speed centrifugation at 2500g for 10 mm at 15°C. 3. Adjust the density of the plasma to 1.24 g/mL by the addition of 38 1 6 mg solid KBr to 1 mL of plasma 4. Underlay the appropriate denstty solution (see Table 1 for the conditions used for rapid isolation of VLDL, LDL, HDL, and HDL subclasses) with the densityadjusted plasma using a syringe or a Pasteur pipet. All density solutions contain 200 1.18EDTA/mL and are prepared in 50 &phosphate buffered saline (PBS), pH 7.4. 5. Seal the tubes usmg the Quick-Seal system. 6 Carry out the ultracentrifugation according to the condittons described m Table 1 for each speciftc application The single lipoprotem fractions are well-visible in the tubes and can be recovered by direct syringe aspiration. Use the 120-mm spin for VLDL, LDL, and HDL isolation; VLDL is located at the top, LDL approximately m the middle, and HDL m the lower third of the tube. Using the isolation protocols shown m Table 1, the recovered hpoprotein fractions are virtually free of albumin, except for HDL3, which contains approx 1O-l 5% albumin as assessed by densrtometry of 12% SDS-PAGE. Lipoprotein fractions are desalted by size-exclusion chromatography on PD10 columns (Pharmacia) equilibrated with PBS (50 mM, pH 7.4) prror to peroxidation studres The chemical composition of the various hpoprotem classes is analyzed using commercially available enzymatic test-kits; the protein content is measured by the Lowry method (7). To analyze the apolipoprotetn composition of the various lipoprotein fractions, I recommended the use of 3.75-10% or 3.75-20% gradient SDS-PAGE with subsequent Coomassie blue staining (see Note 8).

4. Notes 1 Two major brands of ultracentrifuges (Beckman and Sorvall) are available m the market. In this chapter, the conditions are described for Beckman ultracentrifuges and rotors, however, the protocols can be adapted for the equivalent Sorvall equipment. 2. KBr is a hygroscopic salt; for this reason the KBr used m the preparation of the density solutions should be thoroughly dried overnight in an oven at 115Y To avoid repeating this procedure every ttme new solutions are made, the dried KBr should be maintained m a desiccator Moreover, the precise density of all KBr solutions should be verified using a refractometer. 3. Both plasma and serum can be used as the starting material to separate hpoproteins by ultracentrifugation; however, it should be noted that slightly different results are obtained using one or the other. Moreover, plasma is preferred if the hporotem fractions are going to be used for measurements of oxidation parameters. 4 PMSF could be used for this purpose using the following procedure* Add 2 pL of the alcoholic PMSF stock solution to the 1-mL plasma sample and set on a rocker for a few minutes

Table 1 Protocols

for Rapid Isolation

of Different

Lipoprotein

Classes

in a TLX Optima

Benchtop

Ultracentrifuge

Bottom (plasma) Fraction isolated VLDL, LDL, HDL HDL2, HDL3

Temp (“C)

Centnfugatlon mm

15 15

Centrhgatlon 1scamed out at 100,000 x-pmat 15%C ‘50 mk4 PBS (pH 7 4) contammg 0 9 % NaCl (w/v) b83 4 mg KEklmL PBS

120 240

ttme,

Dens@, g/d 1.24 1.24

Volume, l-d 1.7 1.7

Top Density, g/mL

Volume, mL

1.006 1.063b

3.3 3.3

Plasma Lipoprotein Separation Methods

103

5 During this step it is absolutely essential to avoid bubbling when the needle penetrates the bottom of the tube For this reason it is necessary to have a postttve flow of the high-denstty solutton while the needle 1spuncturmg the bottom of the tube Moreover, thts step should be done quickly to avotd burlding up back pressure m the needle 6 Manual pipetting can also be used to separate the different hpoprotem fracttons; once the user has gamed some experience with the fractionation this becomes a faster alternative However, manual recovery may compromise the purity of the fractions. 7. When the hpoprotem fractions separated using those techniques are used to measure the concentratton of their different protein and hptd components, it is necessary to remember that a correctron factor should be applied to these results to adjust for the differences between the original plasma volume and the volume m which the fraction was recovered followmg ultracentrtfugatton For HDL, the results should be multtphed by 1 275 (only 1 mL of the I 275 mL that constttute the final volume of the sample are loaded mto the tube for HDL tsolatton) 8 Detailed protocols for SDS-PAGE are described in Chapter 8 of this work.

Acknowledgments Thts work was supported by grants HL54776 from the National Institutes of Health and contract 53-K06-5-10 from the US Department of Agriculture

Research Service References 1 Have], R. J., Eder, H A , and Bragdon, J. H (1955) The dtstributton and chemtcal composition of ultracentrifugally separated lipoproteins m human serum J Clan Invest 34, 1345-1353. 2. Castro, G R and Fielding, C J. (1984) Evidence for the distnbution of apohpoprotem E between lipoprotein classes m human normocholesterolemrc plasma and for the ongm of unassociated apohpoprotein E (Lp-E) J Llpld Res 25,5%-67 3 Kunitake, S. T and Kane, J. P (1982) Factors affecting the mtegnty of htgh density lipoproteins m the ultracentrifuge J Lipld Res 23, 936-940. 4 Ptetzsch, J , Subat, S , Nttzsche, S., Leonhardt, W., Schentke, K. U., and Hanefeld, M (1995) Very fast ultracentnfugation of serum hpoprotems: influence on lipoprotem separation and composition. Btochzm Bzophys Acta 1254, 77-83. 5 Brousseau, T., Clavey, V., Bard, J M., and Fruchart, J C. (1993) Sequential ultracentrifugatton micromethod for separation of serum llpoprotems and assays of lipids, apollpoprotems, and ltpoprotem particles. Clan Chem. 39,960-964. 6. Napoli, C , Mancim, F. P., Corso, G., Malorm, A., Crescenzr, E , Postrghone, A., and Palumbo, G. (1997) A simple and raped purtficatton procedure minimizes spontaneous oxrdattve modtticattons of low density lipoprotem and lipoprotein (a). .J. Bzochem 121, 1096-l 101 7. Lowry, 0. H., Rosebrough, N. J., Fan-, A. L , and Randall, R J (1951) Protein measurement with the Folm phenol reagent. J. Blol Chem 193,265-275

Preparative Isolation of Plasma Lipoproteins Using Fast Protein Liquid Chromatography (FPLC) Jose M. Ordovas

and Doreen Osgood

1. Introduction

The presence of lipids wtthin the hpoprotein particles confers these complexes with a lower density compared with other serum proteins. Based on thts property, lipoproteins were classically separated into four major classesdesrgnated as chylomicrons, very low-denstty lipoproteins (VLDL), low-denstty lipoprotems (LDL), and high-density lipoproteins (HDL). In addition to then densities, these macromolecules differ from other plasma proteins because of their size (see Table 1). This characteristic makes these particles suitable to be separated using size-exclusion chromatography (I). Size-exclusion chromatography using agarose columns was for many years an alternative to ultracentrtfugatton for the isolation of lipoprotein particles (2). This method could be carried out without a major investment m equipment. Moreover, the samples were not exposed during the separation process to the high g-forces and salt concentrations characteristic of the ultracentrifugatton procedure. On the downside, the classical separation using conventional agarose columns was time-consummg and resulted in large sample dilutions. Later, high-pressure liquid chromatography (HPLC) was also used, primarily for analytical purposes (3); however, the original materials were reported to interact with the lipoprotein particles resulting m poor recoveries as well as clogging of the columns (4) (it has been recently reported that newly developed packmg materials have circumvented this problem see ref. 5). In 1982, Pharmacia LKB (Uppsala, Sweden) developed “fast protein liquid chromatography” (FPLC). This procedure allows resolution of the major lipoprotein fractions and quantitation of both the lipid and apohpoprotein constituents m as From

Methods m Molecular Bfology, Vol 110 Llpoprotem Protocols Edlied by J M Ordovas 0 Humana Press Inc , Totowa, NJ

105

106

Ordovas

Table 1 Physical Characteristics Lipoprotein class Chylomtcrons Remnants VLDL IDL LDL HDL2 HDL3

of Plasma

and Osgood

Lipoproteins

Diameter, run

Density, g/mL

80-500 >30 30-80 25-35 18-28 9-12 5-9


Electrophoretrc mobthty (agarose gels) Origin Orrgin we-P pre-P and 0 P a a

little as 20 pL of sample (67). Thus, thustechnology has become the method of chorce among those mvestrgators working with small animal models (i.e., transgenic mice), on which the volume of plasma 1susually a major limrtmg factor. Another Important advantage of this technique relates to the possibiltty of analyzing minor components of the hpoprotems fractrons that may be loosely bound to these particles. As previously indicated, the mild condmons used in this separation procedure allow recovery of the lrpoprotem particles n-ra way that may be more srmrlar to then- native state m the bloodstream. 2. Materials 2. I. Equipment 1 Basic FPLC System (Amersham Pharmacia Biotech, Prscataway, NJ). This system consrsts of a program controller (LCC 501 Plus), two P-500 pumps, a mixer, prefilter, seven-port M-7 valve, sample loops, ultravtolet (UV) momtor UV- 1, and a Frac-200 fraction collector To improve the throughput of the system, It is convenient to have a superloop attached to an etght-port M-8 valve and a P-l peristaltic pump for automatic multiple-sample loading (Pharmacia also offers a fully automated FPLC with computer control ) 2 Two HR 16/50 columns (16~mm internal diameter, 500 mm m length; - 100-mL gel bed) (see Note 1) (Amersham Pharmacra Brotech) 3 Packing Equipment HR (Pharmacia). 4. Superose 12 preparative grade (Pharmacta) 5. Superose 6 preparative grade (Pharmacra). 6 0.22~pm Low-protein binding filter (4 mm Mtllex, Durapore; Millrpore Corporatton Bedford, MA).

2.2. Reagents 1. Phosphate buffered salme (PBS) (containing ethylene diamine tetra-acetrc acrd [EDTA] and NaN,). 0 15 MNaCl, 10 n-&phosphate buffer, pH 7.4 Dtssolve 8 g NaCl, 0 2 g KCl, 144 g Na*HPO, and 0 24 g KH,PO,, 1.861 g Na,EDTA;

107

Lipoprotein Isolation Using FPLC Valve 2

..z

I

=-

1

fl.MV-7’d\/

Recorder ‘,f

III

Superose 6

I

Fraction collector Fig. 1. Schematic representation of the FPLC system used in this protocol for multiple automated preparative separations of lipoproteins from plasma.

2. 3.

4. 5.

and 0.2 g sodium azide (sodium azide is important to prevent bacterial growth in the columns) in 900 mL of distilled water. Check the pH (adjust if necessary) and bring the volume to I L. Isolated lipoprotein fractions for calibration. Gel filtration HMW Calibration kit (aldolase [158kDa]; catalase [232kDa]; ferritin [440kDa]; thyroglobulin [669kDa]; blue dextran 2000 [-2,OOOkDal) (Amersham Pharmacia Biotech; cat no. 17-0441-01). Reagent kit for cholesterol measurement. Reagent kit for triglyceride measurement.

3. Methods The FPLC format used in this protocol for separation of lipoprotein fractions is illustrated in Fig. 1. The most simple version uses a single Superose 6 packed 16/50 column. Improved resolution can be obtained by using two Superose 6

108

Ordovas and Osgood

16/50 columns or a Superose 6 16150 and a Superose 12 16150 columns m tandem configuration, to provide better separation of the HDL fraction. 1. For first-time operation, or if the system is not runnmg all the time, it is necessary to warm up the system for approx 90 min. 2 During this time, bring buffers and columns from the cold room to allow them to equilibrate at room temperature. 3 Connect the shorter preflanged tubmg (the outlet) to the detector. 4. Connect the longer preflanged tubing (the inlet) to the valve, whtch IS used for sample injection and elutton (see Note 2). 5 Degas the eluent solution and filter through a 0.22~pm filter (see Note 3). 6. Equilibrate the columns with 2-column volumes of H,O, followed by 2-column volumes of eluent buffer. 7. Remove all particulate material from the plasma samples by centrifugatron at 12,OOOgfor 10 min or by filtration through a 0.22~pm low-protein binding syringe filter (see Note 4) 8. Place the samples mto a tube rack It is recommended that the samples stt on an ice bath to prevent degradation (if seven samples are prepared for loading, the last sample will not be injected unttl -23 h after the begmnmg of the run). 9 Program the system to inject 2 mL of each sample every 200 mm (100 mL) (see Note 5 and Fig. 2). 10. Using PBS as eluent, start the separation usmg a flow rate of 0 5 mL/min under a pressure of 1 O-l.1 mPa (see Note 6) 11. Collect fractions of 500 ,uL (see Note 7). 12. Proceed with the analyttcal procedure(s) of choice. m lme cholesterol measurement; manual or automated lipid determination or apolrpoprotein quantitatlon 13 Pool the fractions contaming discrete lipoprotem fracttons. Use the direct UV reading as a guide, or alternatively, the mformatlon from the lipid measurement, in order to determme the cut points (see Fig. 3 for a representattve separation profile)

4. Notes 1, For analytical separations, the Superose HR 6 lo/30 and/or the Superose HR 12 1O/30 columns are recommended. The sample volumes used for mlection m these columns are between 20 and 200 pL (maximum protein load -10 mg) All other conditions are similar to those described for the preparative protocol described here. These columns can be purchased ready to use from Pharmacia The columns used m this protocol have to be packed in the laboratory using the mstructions provtded with the packing material (Superose 6 and Superose 12 prep grade) The only addtttonal equipment needed to pack the columns in the laboratory are the Packing Equipment HR and a pump, such as the Pharmacia high precision Pump P-500 2. It IS recommended to run the columns oppostte of gravity to increase column packing life

Lipoprotein /so/a tion Usrng FPL C

109 Valve

Volume (mL) 0 0

%B Valve

0 0

CItlhL

0 0

valve pump c-mL/mln

22 05

25 27 30

Pump C-mL/mm Port set Valve

0 61 12

50

Valve

II

60

Call Method 2

30 31 33 5

Valve pump C-mL/mm Pump C-mL/mm

23 05 0

100 102

Valve Valve

12 11

105

130 131 133 5

Call Method 2 Valve Pump C Pump C

24 0 5 mL/mm 0

200

Call method3 Call method 2

mL/mm

Repeat thesestepsthe necessarynumber of cyclesdepeting on the number of samples to be mjectedmto the columns

Flow rate Recorder ON Fdl sample loop w dh sample#l

1

T”mFrac ON 1

j

InJectmg sample #I mto cohmln

wash sample loop

>

FdI sample loop With 2 5 mL of sample #2 Inject sample #2 (2 ml) Into colluM

wash sample loop I

Fti sampleloop wtul25nLof sample #3

Inject sample Into column Wash sample loop

Method 2 Volume 0 05 15

Wash sampleloop Functton valve Pump C- mL/mm Pump C-mL/mm

Valve 21 05 0

Fig. 2. Example of the steps used to program the FPLC system for a preparative run using two HR 16/50 columns m tandem as illustrated in Fig. 1 3 Although the working pressures used in FPLC systems are lower than those used m HPLC separations, degassmg the solvents IS also a recommended practice. Pressure reductions can cause dissolved gases to come out of solution. The two

Ordovas and Osgood

770

0

40

80

120

160

ml

Fig 3 Chromatographtc profile of a normal serum using the condmons described in thts protocol The peaks (from left to right) represent VLDL, LDL, and HDL

locattons where this occurs are the suction side of the pump and at the column outlet. Water usually has a higher dtssolved-gas content The followmg approaches can be used to degass the solvents a Subject the solvent to vacuum for 5-10 min to remove the gases b. Subject the solvent to ultrasonics for 10-l 5 mm to remove the gases c Sparge the solvent with a gas that has a very low solubility compared to the oxygen and nitrogen from the atmosphere. Helmm is the preferred chotce (5 min of gentle bubbling from a 7-pm sinter IS usually sufficient) Note that most aqueous-based solvents usually have to be degassed every 24 h Also, remember that solubility of gases Increases as temperature decreases, so ensure eluents are at instrument temperature prior to degassing. 4. If the sample is very lipemic, some of the lipid may float to the top during centrtfttgatton. It is important to homogenize the sample again before mjectton Moreover, this type of sample may need to be diluted 1.2 (or more if needed) with PBS to prevent the clogging of the column. 5 When using analytical columns, sample injections can be spaced every 20-25 mm Cahbrattons of the runs should be done m each laboratory because conditions may very slightly from lab to lab.

Lipoprotein Isolation Using FPLC

111

6. It IS not recommended to exceed 0.5 mL/min, as this will pack the Superose within the columns and the life of the columns ~111 shorten. If the operating pressure of the columns during a run increases (i.e., >l. 1 Mpa), this IS an indication that the columns are getting clogged. At this point, it 1srecommended to change the filters in the columns. It is also recommended during the maintenance procedures to run the columns wtth reversed flow. 7. The major limitation ofunattended operation of the FPLC system maybe on the capactty of the fraction collector. The recommended instrument (FRAC-200) may accommodate up to 175 tubes; however, this number of tubes may not be enough when all fractions are collected for multiple samples. In this case, the fraction collector should be programmed to send the unwanted fractions to a waste container

References 1 Tadey, T and Purdy, W C (1995) Chromatographtc techniques for the isolation and purification of lipoprotems. J Chromat 671,237-253 2 Sata, T., Havel, R. J., and Jones, A L. (1972) Characterization of subfractions of triglyceride-rich hpoproteins separated by gel chromatography from blood plasma of normolipidemic and hyperhpidemic humans. J Lzpid Res 68,757-768 3. Okazakt, M , Itakura, H , Shiraisht, K., and Hara, 1. (1983) Serum lipoprotem measurement. Liquid chromatography and sequential flotation ultracentrifugatton compared. Clin Chem 29,768-773. 4. Carrol, R M. and Rudel, L. L (1983) Lipoprotem separatton and low density lipoprotem molecular weight determination using high performance gel-filtratron chromatography J Lqxd Res, 24,200-207 5 Okazaki, M., Sasamoto, K., Muramatsu, T , and Hosaki, S. (1997) Analysis of plasma lipoproteins by gel permation chromatography. Handbook of lipoprotein testing (Rifat, N., Warmck, G. R., and Dominiczak, M H , eds ), AACC Pres Washington DC, pp 53 l-548 6. Van Gent, T. and van Tol, A (1990) Automated gel permeatton chromatography of plasma lipoprotems by preparative fast protein liquid chromatography J. Chrom 525,433-44 1 7 Marz, W., Siekmeier, R , Schamagl, H., Seiffert, U B., and Gross, W (1993) Fast lipoprotein Chromatography: mew method of analysis for plasma lipoproteins. Clan. Chem. 39,2276-228 1,

Separation of Apolipoproteins by Polyacrylamide Gel Electrophoresis Jose M. Ordovas 1. Introduction Polyacrylamide gel electrophoresis (PAGE) has been for several decades the workhorse of protein and nucleic acid separation (1). Several of the reagents involved in this technique can be modified, added, removed, or combined with the others to provide the optimal resolution for the specific interest of each researcher. Over the years, contmuous Improvements have been made to these techmques. More recently, the mtroductlon of capillary electrophoresls has added further resolution to this electrophoretic technique (2,3). However, most laboratories are still using the conventional PAGE systemsand this will be the focus of the present chapter. The separation of apohpoprotems represents an analytical challenge even for a method as flexible as PAGE. Apohpoprotems display a wide range of molecular

weights from above 500,000

Dalton

for apo B and some ape(a)

lsoforms to about 6000 Dalton for the apo C proterns. This diversity is also evldent in their isoelectric pomts and in their different solubihties

m aqueous and

orgamc solvents. 2. Materials 2.7. Equipment 1. Vertical electrophoresls system: Protean II xi vertical electrophoresls cell and gel castmg set (Blo-Rad, Hercules, CA) or Hoefer 600 Series (Pharmacla Biotech, Uppsala, Sweden, or Pharmacia Blotech, Plscataway, NJ). 2. Electrophoresis Power supply: PowerPac 1000 (Blo-Rad) or Hoefer PS 500XT and EPS 600 (Pharmacla Blotech) From

Methods m Molecular Biology, Vol I10 Ltpoprotem Protocols EdIted by J M Ordovas 0 Humana Press Inc , Totowa, NJ

713

114

Orcio vas

3 Gradient former: Model 385 from Bto-Rad allows the casting of two lmear 1 5-mm-thick gradient gels used m the Protean II xi vertical electrophoresis umt Alternatively, the Model 395 allows the casting of up to 12 lmear 1.5-mm-thtck gradient gels. 4 General purpose centrtfuge GS- 15 (Beckman Instruments, Fullerton, CA). 5. Spectra/Por*2, dialyses membrane, MWCO 12,000 to 14,000 10 or 25 mm flat width (VWR cat no 252 18-402 and 252 18-435). 6 Polypropylene Cornmg Brand Comcal Centrifuge tubes, 15 mL 6000 maximum RCF g (VWR cat no 21008-678 or Corning Costar no 430052) 7 Polypropylene Cornmg Brand Conical Centrtfuge tubes, 50 mL 6000 maximum RCF g (VWR cat no 2 1008-7 14 or Cornmg Costar no 430290) 8. 15-mL Borostllcate glass centrifuge tubes with screw caps (VWR cat no 2 1020684; Ktmble cat no 73785-15) 9 Parafilm 10 Lyophihzer (Vn-tts, Gardmer, NY). 11 Mtcroptpets 12 Dyla-Dual Hot Plate-Sttrrer (VWR cat. no. 58849-001). 13 Orbital shaker (VWR, Bel-Art cat. no. F37041-0000) 14 Gel documentatton system (Hoefer PH Serves PhotoMan Direct, Pharmacta) 15 Gel Drier (Hoefer SE 1200 Easy Breeze An Gel-Drymg System, Pharmacta) or Gel Drying Ktt (Promega cat no V7 120)

2.2. Reagents 2.2.1. Sample Preparation 1 Dialysis tubing boilmg solution 2% (w/v) sodmm bicarbonate and 1 mA4 Na2EDTA, pH 8 0 m double-dtsttlled water (ddHzO) 2 1 mMNa,EDTA, pH 8 0 3. Dehptdatton mixture 1: ethanol/diethyl ether (3.1, v/v) 4 PBSE (PBS + 10 m/I4 EDTA) 5 Dtethyl ether 6 4.9 A4 Trichloroacettc acid (TCA) Drssolve 80 g of TCA (T 9159, Sigma, St Louts, MO) m ddH,O. Bring to a final volume of 100 mL 7 3 6 mM Sodium deoxycholate. Dissolve 0 156 g of deoxychohc acid, sodium salt monohydrate (Sigma cat. no D-5670) m 100 mL of ddHzO

2.2.2. Sodium Dodecyl Sulfate (SDS)-PAGE 1 Stackmg gel buffer, 0.5 M Tris-HCl, pH 6 8. Dtssolve 6 g of Trts m 40 mL of distilled water. Titrate with 2 M HCL (-25 mL) Bring the volume to 100 mL wnh ddHzO 2 Separating gel buffer; 3 0 M Tns-HCl, pH 8 8. Dissolve 36 3 g of Tns m disttlled water and titrate with -25 mL of 2 MHCl Bring the volume to 100 mL with ddHzO 3 10X Reservoir buffer; 0.25 MTrts, I 92 Mglycme, 1% SDS, pH 8 3 Dissolve 30.3 g of Trts, 144 g of glycine, and 10 g of SDS; add dtsttlled water to 1000 mL. Dtlute l/10 prior to use

Apolipro tein Separation

115

4 Stock Acrylamide

bzs-solutlonU 30 g Acrylamlde, 0 8 g blsacrylamlde (see Note 1); dissolve and adjust the volume to 100 mL with ddH,O. Add 1 g Dowex 1 amon exchange resm (Sigma) Remove the resin by filtration before use (see Note 2) This solution should be maintained m a dark bottle at 4°C When stored under these conditions, the stock solution can be used for 3-4 wk (see Note 3) 5 10% Ammomum persulfate (AP) 1 g AP (Sigma cat no A-3678) In 10 mL of ddH,O. In order to get consistent polymerlzatlon results from day to day, this solution should be prepared freshly and used wlthm the same day (see Note 4) 6

10% SDS* Dissolve 1 g of SDS (Sigma cat no. L-4509) m ddH,O Bring the final volume to 10 mL Store at room temperature Solutions contammg SDS preclpltate m the cold 7 N,N,N’,N’-tetramethylethylenedlamme (TEMED, cat no. T928 1, Sigma) (see Note 4). 8 SDS-PAGE sample buffer: 125 n&I Tns-HCl,

9

10 11 12 13 14. 15. 16. 17. 18. 19

dissolve 0 3 g m about 15 mL distIlled water and adjust the pH to 6 8 Add 4.0 g Glycerol (20% m the final volume) and 0.8 g SDS (4% tn the final buffer volume) 2.5 mg bromophenol blue (0.05% final concentration m buffer) can be added to color the sample. Adjust the volume with ddHzO to a final volume of 20 mL. Immediately before use, add 40 PL of P-MercaptoethanoVmL of sample buffer Coomassle blue stam: Dissolve 600 mg Coomassle brilliant blue G-250 (Sigma cat. no. B-5 133) mto a mixture containing 260 mL ddH20, 240 mL of ethanol, and 100 mL acetic acid (usmg mcreasmg ratios of methanol to dlstilled water results m gel shrinkage). Prepare ahead of time (at least 1 h) and filter through Whatman no 1 filter paper to remove nonsolublhzed Coomassle blue that could deposit on the surface of the gel This may create stammg artifacts that ~111affect protein ldentlfication and will give false readings when scannmg IS used to quantltate the protein bands. Destainmg solution. 260 mL ddH,O, 240 mL of ethanol and 100 mL acetic acid. Silver stam fixing reagent 200 mL ethanol, 50 mL acetlc acid, and 250 mL ddHzO Silver stam washmg reagent: 400 mL ethanol and 600 mL ddH*O. Silver-stain fix/sensltlzmg reagent: 200 mL 25% glutaraldehyde, 10 mL 37% formaldehyde, 40 mL ethanol, and ddHzO to 1000 mL. Silver-stam sensitizing reagent: 40 mg sodium thlosulfate m 200 mL ddHz0 200X stock sliver mtrate solution. 2 g silver nitrate (Sigma cat no S-6506) m 10 mL of ddH*O Store m a dark glass bottle Silver-stain developer solution: 5 0 g sodium carbonate, 40 mL of formaldehyde and ddHzO to 200 mL Prepare immediately before use. Silver-stam stop solution 5% (v/v) acetic acid. Silver-stam gel storage solution. 0.03% (w/v) sodmm carbonate Molecular Weight Standards (M 3913 [6,500-66,000 mol wt], M 3788 [36,000-205,000 mol wt] Sigma).

116

Ordovas

2.2.3. Tricine-SDS-PAGE 1 Trtcme Anode (lower) buffer: 200 mM Tns pH 8.9. 96 8 g Tns; 185 mL 2 M HCI, ddHaO to 4 L. (Note* The pHs for the anode and gel buffers are essential to get optimal performance from this protocol ) 2. Tricine cathode (upper) buffer: 100 mM Tris, 100 mM Tncme (N-tris[Hydroxymethyl]methylglycme, Sigma cat. no. T 7911) and 0 1% SDS, pH 8.25). Dissolve 12 1 g Tris, 17 9 g Tricine and 1 g SDS in ddH,O and bring the final volume to 1 L 3 Trtcme Gel Buffer 3 0 M Tris, 0.3% SDS, pH 8.45. Dissolve 36 3 g Tris and 300 mg SDS m ddH*O, add 50 mL HCl and bring the final volume to 100 mL with ddH,O 4. Stock acrylamtde solution. 48 g of acrylamide and 1.5 g bisacrylamide, dissolve m ddH*O, and bring the final volume to 100 mL 5. Molecular weight standards (Mark12 Wide-range protein standard [2,500-200,000], Novex, San Diego, CA).

2.2.4. Gradient-SDS-PAGE 1 Lower gradient gel acrylamide solution (similar to the one described in Subheading 2.2.2.) 29 2 g acrylamtde and 0 8 g bisacrylamtde* Dissolve m 100 mL ddH,O on a magnetic stirrer with low heat Stir until the acrylamide and btsacrylamide are completely dissolved (about 20 mm) This stock solution of 30% acrylamtde should be stored at 4°C and protected from the light 2 Upper gradient gel acrylamide solution. 8 g acrylamtde and 42 1 mg btsacrylamtde dissolved m ddHzO to a final volume of 150 mL 3 Lower gradient gel buffer 182 g Trizma base, 3 g Na*EDTA, and 4 g SDS Drssolve in 1 L ddHs0 on a stirrer plate Adjust the pH to 8 8 with HCl Store at 4°C. 4. Upper gradient gel buffer: 67.5 g Tris-HCl; 2.5 g Trizma base; 3 g EDTA and 4 g SDS. Dissolve m 1 L of ddHzO on a magnetic stirrer plate. Adjust the pH to 6 8 wtth HCI Store at 4°C.

2.2.5 Isoelectric Focusing (IEF) 1 IEF dialysis solution. 10 mM ammomum bicarbonate 2 Ampholme (preblended ampholytes pH 4 O-6 5, Stgma cat. no A-4799) 3 IEF Sample buffer 100 pL of 10 mMTris, 10 mM m-dithiothrettol (DTT, Sigma cat. no. D-9 163), 8 M urea, pH 8 2. 4. IEF upper electrophorests buffer (5 mMNaOH)+ Add 5 mL 1 MNaOH to 1 L ddH,O 5. IEF lower electrophoresis buffer (10 mMphosphortc acid)* 10 mL 1 Mphosphortc acid (Sigma cat no. P-0180) m 1 L ddHaO 6 IEF fixing solutton: 30 g of trrchloroacetic acid and 13 g 5-sulfosahcyhc acid (Sigma cat. no. S-3 147) in 500 mL of distilled water 7. IEF destaining solutton 200 mL of glacial acetic acid and 600 mL methanol m 1200 mL of distilled water 8. IEF staining solution. 1 g of Coomassie bnlhant blue m 450 mL methanol, 100 mL glacial acetic acid and 450 mL distilled water 9 IEF Markers (IEF-M2, IEF range 3 6-6.6, Sigma).

Apoliprotein Separation

117

2.2.6 Two-Dimensional

(2-D) Gel Electrophoresis 1. 2-D sample buffer-l: 50 mA4 Trts-HCI, pH 6 8, 6 M urea, 30% v/v glycerol,

2% w/v SDS, and 2% w/v DTE. 2 2-D sample buffer-2: 50 mM Tris-HCl, pH 6.8,6 M urea, 30% v/v glycerol, 2% w/v SDS, 2.5% w/v todoacetamtde, and a trace of bromophenol blue 3 2-D agarose solution. 0.5% w/v agarose and 25 nnI4 Trts, 198 mM glycme and 0 1% SDS (w/v), pH 8 3.

3. Methods 3.1. Sample Preparation The specific protocol for sample preparation depends on the method used to Isolate the lipoprotein fractions. Samples isolated usmg ultracentrrfugatron require a dialysis step prior to delipidatron in order to remove the KBr or NaBr used to raise the density of the sample during ultracentrifugation (see Note 5).

3.1.1. Dialysis 1. Cut the dtalysis tubmg mto pieces of the proper stze for the volume of the samples to be dialyzed 2 Place the tubmg mto a beaker containing about 1 L of dtalysts tubing bodmg solutton and boll for 10 mm Make sure that the tubing remains submerged durmg the entire period. This step removes additives used to preserve the integrity of the membrane durmg storage and rmpurtttes that could affect the composttton or integrity of the lipoprotein fraction being dralyzed. 3. Discard the buffer and wash the tubing extensively with dtsttlled water 4. Boil again for 10 mm, this time using only 1 nut4 EDTA, pH 8 0. 5. Let the tubing to cool before storing it at 4°C The same recommendation indicated during boiling applies to storage. Make sure the tubing IS always immersed m the fluid. Rinse with dtsttlled water before use 6 Dtalyze the isolated hpoprotems against 2 L PBSE overmght at 4°C.

3.1.2. Delipidation If the lipoprotern fractions have been obtained by gel-filtration chromatography, then the dialysis step IS not required and deltptdatton can

be carried out without prior processmg. Delrpidation accomplrshes two purposes. Ftrst, tt concentrates the apolipoprotems by precrpttatton. This 1s specially advantageous when gel-filtration chromatography 1s used for ltpoprotem separation because constderable sample drlutton may occur. Second, It removes the lipids, thus provtdmg a much cleaner and reproducible separation profile of the apoltprotems (see Note 6). The following protocols are suggested.

Ordovas

118 3.1.2

1. PROTOCOL

1

1 Add the sample (containing at least 100 g protein) to 10 vol of the ice-cold ethanol/dietyl ether dehpldatmg mixture (delipldatlon mixture 1). 2 Delipldate overnight at -2O’C This step is more efficiently carried out placing the tubes m a rotary mixer. 3 Centrifuge the tubes at 1OOOgmaintaining the temperature of the centrifuge between -10 and 0°C 4. Remove the orgamc solvent using a Pasteur plpet, making sure that the tip does not touch or disrupt the pellet. 5 Wash the pellet with Ice-cold ethyl ether. Use the same volume as for step 1 6. Centrifuge again at lOOOg, maintaining the temperature of the centrifuge between -10 and 0°C. 7 Remove the ethyl ether as before and leave the protein pellet to air-dry to remove visible solvent left behind by the Pasteur plpet. If the protem pellet dries completely, it ~111 be very hard to redissolve in the sample buffer. 8 Solubihze the pellet usmg between 100 and 500 @. of SDS-PAGE sample buffer. The volume depends on the initial amount of protein used m the dehpldatlon procedure The final concentration of the sample should be between 50 and 100 PgllOO pL, and it will depend on the capacity of the gel and the complexity of the samples (number of different protems m the sample). 9. Leave tubes uncapped for about 30 min at room temperature to facilitate the evaporation of any remaming ether. 10. Cap the tubes and incubate the samples at 100°C for 3-5 min

3.1.2.2.

PROTOCOL

2

A variation of Protocol 1 has been reported that shortens consrderably the actual deliptdation process, avolds the dlalysls step, and improves apohpoprotein recovery (4). 1 MIX 1 mL isolated lipoprotein fraction with 10 mL diethyl ether for 2 mm 2 Add 0 2 mL of 4.9 M TCA and 0 2 rnL of 3 6 mM sodium deoxycholate and mix brlefly Let settle for 1 h and remove the top layer of dlethyl ether by aspiration. 3 Pellet the precipitated apohpoprotems by centrlfugation at 3000g for 20 min. 4 Discard the supernant and proceed as indicated m steps 8-10 from Subbeading 3.1.2.1.

3.2. SDS-PAGE The protocol described here is based on the tradltlonal Laemmli system (I). 3.2.1. Preparation of Separating Gel 1. Clean glass plates (16 x 18 or 20 x 20 cm) with water and detergent Wipe with ethanol to ensure they are not marked with fingerprints 2. Using 1.5-mm spacer bars, assemble glass sandwiches and clamp mto position on the gel casting set

Apolipro tein Separation

719

Table 1 Preparation of Separating Gels for the Laemmli System (volumes 5% Separation range, 50-250 kDa

Reagents Acrylamidel bisacrylamide H20

Separation buffer 10% SDS 10% AP TEMED

5 6

7. 8 9.

10.

11.

12.

5

in mL) 7 5% Separation range, 30-l 00 kDa 75

10% Separation range, 15-70 kDa 10

mix 20 8 3 75 03 0 15 0.015

183 3.75 0.3 0.15 0.015

15 8 3 75 0.3 0 15 0.015

Fill the gel cassette with dlsttlled water, m order to ensure that an efficient seal IS obtamed between the glass cassette and the castmg platform Mark the level of water on the glass plate with a marker and wait 15 min If the level remains the same, this is an indication that no leaks exist in the gel cassette. Empty the gel cassette and dry it with a hair dryer Mix m an erlenmeyer the acrylamlde/bisacrylamlde, separating buffer, SDS solution, and H20 suited to prepare the volume and concentration of gel more adequate for your apphcatlon Table 1 provides the volumes to prepare 30 mL of gel, which is the usual volume for a 18 x 16-cm plate with 1.5-mm spacers Leave the gel mix to equilibrate at room temperature Degas (see Note 7) the solution using a vacuum of less than 125 torr for at least 15 mm This time can be shortened if continuous swirling of the solution is carried out to accelerate the release of the air bubbles. Add quickly the ammomum persulfate and TEMED Mix the solution gently to avoid the remtroduction of oxygen, but thoroughly to avoid uneven polymerlzatlon Pour the acrylamide solution mto the gel mold. Use a 50-mL syringe without the needle and pour the solution m a contmuous stream to reduce the mtroductlon of oxygen m the solution Leave enough space for the stacking gel (approx 1 cm below the bottom of the well-forming comb). Overlay carefully with water saturated lsobutanol to exclude oxygen from the gel surface A syringe without the plunger and with the needle bent mto a U-shape provides a contmuous and gentle flow of the overlay solution Let the gel polymerize for about 30 mm. Polymerlzatlon can be ascertained by the presence of a characteristic schlearmg pattern at the interface between the gel and the overlay solution Discard the overlay solution and wash the top of the gel with distilled water to remove unpolymenzed acrylarmde. Remove any remammg water using a paper towel

120

Ordovas

3.2.2. Preparation of the Stacking Gel 2 3. 4 5 6

7

8.

9.

10. 11 12

In a disposable 50-mL plastic tube, mix 1 25 mL acrylamide/brsacryIamtde solution, 2.5 mL stackmg buffer, 0.1 mL 10% SDS, and 5.65 mL of HZ0 Degas the solution as described m Subheading 3.2.1. Add 0 1 mL of 10% AP and 10 L TEMED Mix the solutton gently to avoid the remtroductron of oxygen, but thoroughly to avoid uneven polymertzatron Pour the acrylamrde solution directly into the gel mold Leave a small amount (about 1 mL) m the tube in case tt is needed later. Insert a clean comb (Note 8) into the stacking gel solutron Be careful not to trap any an under the comb teeth If bubbles are present, remove partrally the comb and reintroduce tt m a slight angle and add more stackmg gel if needed to fill entirely the gel mold Leave the gel to polymenze for about 30 mm. Remove the comb very carefully The stackmg gel 1snot very consrstent and can break or deform very easily. If the teeth of the wells become twisted, they should be straightened (Note 9) Remove the gel cassettes from the gel castmg set and place them into the upper reservoir tank Frll the lower reservon wrth about 3-4 L of lower runnmg buffer and place the upper set (upper reservorr tank and gel cassettes attached) onto the lower reservoir tank At thus time make sure that there is no bubbles on the bottom of the gel This may cause uneven conductrvny and dtstorted bands If bubbles are present, they can be removed by flushing this area with a 5-mL ptpet loaded with the lower electrophoresrs buffer and fitted at the end with a syringe needle bent to a U-shape Wash the sample wells with upper buffer to remove unpolymerrzed acrylamrde and fill the wells and the upper tank with about 1 L of upper runnmg buffer Load the samples prepared as described m Subheading 3.1.2. (see Notes 10 and 11) Use a long-tipped micropipet tip, bemg careful not to prick the bottom of the well Load at least 1 lane of molecular weight markers on each gel

3.2.3. Electrophoretic

Run

1. Place the cover over the electrophoresis unit and connect the cables to the power SUPPlY2. Set the voltage to 100 V untrl the samples have entered the stacking gel as ascertained by the absence of blue color m the sample application well. This takes about 30 min. 3. Increase the voltage to 250 V The run is complete when the bromophenol blue band is reachmg the bottom of the gel (- 5 mm from the bottom) Running trmes may vary widely depending on the concentratron of acrylamtde used in the preparation of the gel. 4 Remove the upper reservoir tank with gel cassettes attached and pour the electrophoresis buffer from the upper tank down the smk 5. With a plastic spatula, pry open the cassette to remove the gel. Do not use metal spatulas because they tend to chip the glass plates, which will result m breakmg or in possible acrylamide solution leaking during later gel castmgs.

Apoliprotein

Separation

121

6. Remove carefully the gel from the glass plates. With practtce the gel can be lifted from the plate with both hands (remember always to wear gloves) and placed into the stainmg tray. An alternative technique is to place the glass plate contammg the gel resting on an angle onto one of the walls of the staining tray, and to shde the gel gently mto the staming solutron while removing the glass plate from the tray

3.2.4. Protein Staining There are several alternattves to vrsualize protein bands separated by PAGE Two common protocols mvolving Coomassie blue and stlver-stammg are described. 3 2.4.1

COOMASSIE BLUE STAINING

Thts stain 1s easy, reproducible, and does not require much hand on ttme; moreover, the resulting color produces good signal response for the gel scanners used in the laboratory The major drawback of this stain 1sthe lack of sensmvity when samples are very diluted or when the aim of the separatton is to detect minor bands (like when detecting contamination of an apolipoprotem preparation or when exammmg rare mutants). 1. Stain the gel for at least 4 h It can also be left overnight; however, the stainmg tray should be covered to prevent evaporation of the solvents that may result m prectpitatton of the dye The staining is more homogeneous If the staining takes place wrth gentle shaking using an orbital shaker at low speed. 2. Discard the staining solution (follow the mstitution regulations regarding disposal of organic solvents) 3. Add destaining solution and rock gently for 1 h Discard the used destammg solution and repeat this procedure until a clear and even background is obtained in the gel 3 2.4 2. SILVER-STAINING

Silver staining is more time consummg and more prone to artifacts; however, it has the advantage of its high sensitivity. 1 Place the gel m a glass tray and approx 200 mL silver stain fixing reagent for 10 mm Discard the solution according to local regulations. 2 Rinse with 200 mL distilled water for 10 min. Discard the water afterwards 3 Add 100 mL fix/sensitizing reagent for 5 min. Discard the solution with the hazardous waste. 4 Rinse with 200 mL wash reagent for 20 mm. Discard the solution. 5 Rinse with 200 mL distilled water for 20 mm Discard the rmse 6. Add 100 mL sensmzmg reagent for 1 mm. Discard with the hazardous waste 7 Rmse twice with distilled water for 1 mm each time.

8. Add the silver nitrate solutron (rememberto dilute the stock solution 19200with distilled water) and incubate for 20 mm

122

Ordovas

9 Rinse with 200 mL of distilled water for 1 min. 10. Add 100 mL of the developmg solution. The background of the gel turns light yellow. At this point (about 1 mm) discard the developmg solution and add 100 mL fresh developing solution Develop until the desired protein stammg IS achieved (this may take between 5 and 30 min) 11. Stop the development by incubating with 200 mL of stop solution for 5 mm 12 Keep the gels immersed m storage solution

3.2.5. Gel Storage and Analysis (see Note 72) It is recommended to document the results of the electrophorests possible. The followmg steps should be car-r-red out:

as soon as

1 Take a picture of the wet gel using a gel documentation system. 2 Scan the gel using a laser densitometer and analyze the data 3 Dry the gel to keep a permanent record

3.3. Tricine PAGE Resolutron of low-molecular-weight apolipoprotems such as the apo Cs or the monomer of apo A-II could be achieved in theory using the traditional SDS-PAGE system described before using high acrylamrde concentrations (-20%). However, m practice, the resolutron of these proteins using this approach tends to be poor. The following protocol has been optimized to resolve protein bands as small as 1 kDa (5). 1 Set the gel casting mold with the cassettes m place 2. Mix u-r an erlenmeyer adequate volumes of acrylamide/bisacrylamide, tricme separating buffer, SDS solution, and HZ0 to suit your specific application and gel size. Table 2 provides the volumes to prepare two 18 x 16-cm plates with 1 5-mm spacers. 3 Prepare the gel and load the samples as described above but using the recommended trrcine upper and lower runnmg buffers instead of those used for the traditional Laemmh method (1) 4. Different protocols have been used for runnmg these gels In our experience, constant voltage at 90 V during 16 h or until the leading dye front reaches about 1 cm from the bottom of the gel provides adequate separation of the apo Cs bands Check carefully during the run that enough buffer is mamtamed m the upper reservoir. If the level drops to the point that the flow of electricity is interrupted or it becomes uneven, the electrophoresis should be stopped and the additional upper buffer added Another possibility is to begin with 90 V for the first hour and Increase the voltage to 250 V for the remainder of the run. This is a feasible alternative, however, it is usually not optimal m terms of timing durmg the workday unless the gels have been prepared the day before 5. Staining of the bands is carried out as previously described

Apohprotem Separation

123

Table 2 Preparation of 12% Separating and 4% Stacking SDS-PAGE Gels (volumes Reagents Acrylamidelsolutlon bisacrylamlde stock solution Gel buffer Glycerol ddHzO 10% AP TEMED

3.4. Gradient

in mL)

12% separating gel I1 0

4% stacking gel 10

150 5.0 14.0 0.150 0 015

3.1 8.4 0 100 0015

SDS-PAGE

The preparation of gradient gels is more cumbersome and time-consummg, however, for the separation of the whole spectrum of apohpoproteins in a single run, these gels are the only approach.

3.4.7. Gel Preparation 1 Prepare gel castmg set and cassettes as described m Subheading 3.2.1. 2. Mount the gradient gel maker on a stirring block supported about 24 m above the benchtop using an adjustable platform 3. Posltlon the outlet tubing over the gel cassette. Ensure that the casting unit is level 4 Place a small Teflon-coated stirring bar m the right chamber of the gradient maker (the chamber closer to the outlet tubing). Set the stirring bar at a moderate rate to ensure complete mlxmg of the solutions while pouring the gel without mtroducmg too much air into the solution 5 If the purpose IS to separate the entire range of apolipoprotem sizes, use the protocol m Table 3 to prepare a 4-22.5% gradient gel: m two 50-mL tubes, prepare the 4% and 22 5% acrylamide solutions. 6 If the interest 1s to resolve apohpoprotems m the medium to small size range, then the followmg volumes should be used to prepare a 7-20% gradient gel* m two 50-mL tubes, prepare the 7 and 20% acrylamlde solutions (see Table 4) 7 As soon as the final two reagents are added, the solutions must be speedily transferred to the gradient maker. Pour the 4% solution into the left-hand chamber of the graident marker. Open the stopcock of the left-hand chamber momentarily, allowing air in the passage connecting the two chambers to be driven out. Close the stopcock. 8. With a Pasteur plpet, transfer any 4% acrylamide solution m the right-hand chamber back into the left-hand chamber.

Orciovas

124 Table 3 Preparation

of 4-22.5%

Gradient

Separating

Percent gel solution Reagents Lower gradient gel acrylamide solution Lower gradient gel buffer ddHzO Ammonmm persulfate (10%) TEMED

Gels (volumes

4% 16 x 18-cm Gels

in mL) 22.5%

20 x 20-cm

Gels

16 x 18-cm Gels

20 x 20-cm Gels

26

39

7.5

113

5

75

2.5

3.8 mL

12.3 0.100

18.5 0 150

-

-

0.050

0075

0 010

0 015

0.005

00075

Table 4 Preparation of 7-20% Gradient Separating Gels (volumes In mL) Percent gel solution 7% 20% 16 x 18-cm 20 x 20-cm 16 x 18-cm 20 x 20-cm Reagents Gels Gels Gels Gels Lower gradient gel acrylamide solutron Lower gradient gel buffer ddH*O Ammonium persulfate (10%) TEMED

3.3

495

96

14.4

36

54

3.6

5.4

75 0.075

11.25

0.1125

1.2 0075

0 1125

0.075

00113

0.075

00113

The amrnonnu-n persulfate and the TEMED being poured

18

should be added last, tmmedlately prior to the gel

9. Wrth the right-hand stopcock closed (the one connected to the outlet tubing), pour the 22.5% solution mto the right-hand chamber 10 Open the right-hand stopcock and mittate the flow by applying slight pressure (with a plunger) to the chamber. 11. Direct flow down the srde of the gel cassette 12. As soon as the gel solution reaches the bottom of the plates, open the left-hand stopcock, and make sure there is adequate mixing of the two gel solutions m the right-hand chamber. In order to reach proper equihbrmm the fillmg speed should be very slow.

Apohpro kin Separation Table 5 Preparation for Gradient

of Stacking Gels Gel Electrophoresis

Reagents Upper gradient gel acrylamtde solution Upper gradient gel buffer Ammonium persulfate (10%) TEMED

125

(volumes 16 x 18-cm Gels

in mL) 20 x 20-cm Gels

6

9

2 0.080 0.040

3 0.120 0.060

13. Allow all of the gel soluttons to run into the sandwich. Overlay the acrylamtde gradient with water-saturated isobutanol, using a 5-mL syringe without the plunger and with the needle hunched 180”; this will ensure that the gel will polymerrze with a level upper edge. Gel polymertzation occurs withm l-2 h, as evldented by the appearance of a distinct interface between the alcohol layer and the top of the gel 14. Posttlon the well comb at the top of the glass cassette. Make sure that combs are clean and free of debris 15 Prepare the stacking gel by adding mto a 50-mL comcal tube the followmg reagent volumes (see Table 5) 16 As soon as the ammonmm persulfate and the TEMED have been added, fill the gel cassette with this solutton using a Pasteur plpet. Polymertzatlon should occur within 10 mm. 17 Proceed with the removal of the comb, sample loading, and electrophoresis as described in Subheading 3.2.2. 18 Record the results, scan and dry as mdtcated m Subheading 3.2.5. (For an example of the separation, see Fig. 1.)

3.5. IEF IEF uses a pH gradient formed by small amphoteric molecules (ampholytes) to separate apolipoproteins according to their isoelectric points (PI). Different from the previously described PAGE methods, IEF is an end point method. In polyacrylamide gels containmg ampholytes, a linear pH gradient builds up when the electric field 1sapplied. Ampholytes migrate in the electric field until they reach the position of their respective p1 and then they stop moving. The apohpoproteins behave Just like ampholytes and migrate to the posttron in the gradient where the pH matches their respective p1 and then they stop moving and “focus” rn a very narrow band, where they stay, whatever the duration of the electrophoretic run. This technique produces very sharp and well-defined bands, and separatesthe apolipoprotems using a different physical characteristic (PI) than the previously described protocols (size) (6).

126

Ordovas

BlOO 205 lcD 116kD 97kD 66 kD

29 m

Fig. 1. 4-22.5% gradient SDS-PAGE showing the separation of the major apoliproteins in VLDL (lane 2), LDL (lane 3), and HDL (lane 4). Lane 1, molecularweight markers.

3.5.1. Sample Preparation 1. Dialyze the isolated lipoprotein fraction overnight against IEF dialysis solution using dialysis tubing of 12,000-14,000 molecular-weight cutoff. 2. Transfer the sample to a 15-mL conical tube (use 50-mL conical tubes if the sample volumes are over 5 mL) and freeze at -70°C by placing the sample in an ultrafreezer (see Note 13). 3. Lyophilize overnight or longer if large volumes are processed. 4. Delipidate as described in Subheading 3.1.2. 5. Resolubilize the protein pellet in 100 pL of IEF sample buffer. This process is more efficient if the samples are incubated at 37°C in an oven or a water bath. 6. Spin the samples to remove any nonsolubilized material and transfer the clean supernate to an Eppendorf tube. 7. Measure the protein content by Lowry.

3.5.2. Gel Preparation 1. Dissolve 1.5 g of acrylamide and 40 mg of bisacrylamide in 19 mL of 8 M urea and filter through a 0.45~pm filter (Millex-GS, Millipore, Bedford, MA). 2. Add 1 mL of ampholines (LKB, pH 4-6.5) and degas as described in Subheading 3.2.1. 3. Add 25 L of 10% AP and 3 L of TEMED, mix well, and pour into the gel cassette as described in Subheading 3.2.1., but, in this case, fill the gel cassette with the solution, as no stacking gel is used in this procedure.

Apoliprotein

Separation

4 Insert a clean comb mto the gel solution. Be careful not to trap any air under the comb teeth If any bubble IS present, remove partially the comb and remtroduce it m a slight angle and add more gel solution if needed to fill entirely the gel mold 5 Leave the gel to polymerize for about 30 min. Carefully remove the comb 6. Remove the gel cassettes from the gel casting set and place them mto the upper reservoir tank. 7 Fill the lower reservoir with about 3-4 L of lower IEF running buffer and place the upper set (upper reservon tank and gel cassettes attached) onto the lower reservotr tank. At this time make sure that there are no bubbles m the bottom of the gel. This may cause uneven conductlvlty and distorted bands. If bubbles are present, they can be removed by flushing this area with a 5-mL pipet loaded with the lower electrophoresis buffer and fitted at the end with a syringe needle bent to a U-shape 8 Wash the sample wells with upper IEF buffer to remove unpolymerized acrylamlde and fill the wells and the upper tank with about 1 L of upper running buffer. The sample loading should be done immediately after washing to avoid diffusion of the urea from the gel into the well. This will pose some problems when trymg to load the samples because they may not go readily to the bottom of the well. 3.5.3.

Sample

Loadrng

and Nectrophoretic

Run

1. Load about 50 g of protein per lane using a Hamilton syrmge or a mlcropipet with a flat long-tipped plpet tip (it 1sconvenient to run one lane of pI markers on each gel) 2 Run the gels overnight at constant current with the initial voltage set at 100 V Alternatively, these gel can be run for 4 h with an mitral voltage set at 400 V 3 After the run, remove the gel from the cassette using the technique suggested m Subheading 3.2.3. and place it m IEF fixing solution for 30 mm If possible, use an orbital shaker set at 25 cycles/min 4. Discard the fixing solution and equihbrate the gel m IEF destain solution for 10 min prior to the stammg step. 5. Stain the gel with IEF staining solution for 3 h. Remove the solution and add destaining solution until the background 1sclear.

3.6.2-D Gel Electrophoresis 2-D gel electrophoresls is the most powerful analytical tool for characterization of complex protein mixtures in whole-cell extracts as well as in plasma, and the concept of “proteome” is emerging to reflect the protein content of a cell. 1. Prepare a standard IEF gel for the first dimension and follow the steps mvolvmg sample preparation, gel casting, and electrophoretic run as shown in Subheading 3.5. After concluding the IEF electrophoretic run, cut out the lane contaming the sample of interest and remove approx 1 cm from the anodlc and cathodic sides. 2. Place the lane mto a 15-mL polypropylene tube filled with 10 mL 2-D sample buffer-l for 15 mm. 3 Discard the solution and replace with 10 mL 2-D sample buffer-2 for 15 mm.

Ordovas 4 Prepare an SDS-PAGE gel as described in Subheading 3.2. (For standard SDSPAGE) or 3.4. (for gradient PAGE) with the following modtfications. Pour the separating gel until it reaches about 1 cm from the top of the plate Overlay with water-saturated butanol. Leave the gel polymerize for >30 mm 5 Remove the butanol and blot dry the gel surface wtth a paper towel 6 Overlay with a 2-D agarose solution heated at about 70°C. 7. Load the equilibrated gel strip from the first dimension through thus solutron This has to be done very qurckly whrle the overlymg solution is still hqutd. 8 Run the second drmensron using the same condtttons as those described m Subheadings 3.2. or 3.4. 9 Proceed with fixing and staining as previously described.

4. Notes 1. Acrylamrde IS a potent neurotoxm. The absorption can be through the skin, mgestton, or inhalation. Strict guideline6 should be used to avotd personal and coworkers’ exposure a. Always wear gloves when handling unpolymerized acrylamide b Work m a well-ventdated area and avord the formatron of aerosols. c Avoid creating acrylamrde “dust.” This IS specially common while weighing the acrylamrde. d When skin contact 1s detected, wash immediately with soap and rinse the affected area with water. Less is known about the toxic characteristics of btsacrylamide. They should be handled with the same precauttons as those descrtbed for acrylamide 2 The purity of acrylamide IS crucial to achieve optimal and reproducible results The presence of contaminants can create poor results and usually much time is wasted trackmg the problem m other dnectrons. The major contammant wrthm acrylamide IS acrylic acid The presence of acrylic acid at levels above 0 00 1% can create serious problems in the polymerization and performance of the gels Resins may be used to trap this and other tome contaminants. 3. Under normal storage condmons, dry acrylamtde and brsacrylamrde keep well for years, however, the stock soluttons should not be kept for more than 1 mo owing to the hydrolysis of acrylamtde to acrylic actd. 4. Ammonium persulfate is a highly hygroscoprc substance and must be stored m tightly sealed containers to prevent self-oxidatron and decompositron. Even under optrmal condmons, ammonium persulfate should be discarded after 1 yr Working solutrons must be prepared daily and if possible just before the preparation of the gels. The reduced shelf-life also applies to TEMED. Like ammonmm persulfate, TEMED is highly hygroscopic and loses activity with time Yellowish coloratton of TEMED IS a clear indication of lost activity 5 If the high concentration of KBr IS not removed by dialysis, the samples ~111 precipitate u-r the well. Moreover, the high salt concentratton wrll produce an aberrant sample mob&y and excessrve heatmg of the gel during the electrophoretrc separation,

Apoliprotein Separation

729

Table 6 Polyacrylamide Gels: Suggested Maximum Volume and Protein Loads by Comb Type and Gel Thickness

Comb type 10 Wells 15 Wells 2-D gel

Maximum Gel thickness, load volume mm per well in & 1.0 1.5 1.0 1.5 1.0 1.5

25 40 15 25 500 750

Maximum protein load per band m pg 0.5-2 0 0 5-2.0 0.5-2 0 0 5-2.0 lo-20 1O-20

6 In our experience the deltptdatton step can be skipped when just a quick assessment is needed The SDS present m the sample buffer and in the gel act as dehpidation agent for the hpoprotems. However, when precise separation and band quantitation are needed the full dehpidation procedure is recommended. 7 Degassing 1s essential, especially when ammonium persulfate and TEMED are used as initiators. Oxygen IS an inhibitor of polymerizatton Induced by free radicals and its excluston from the solutions 8 Combs should be cleaned with water and dried with ethanol Material left over m the combs may mhtbit polymerization m the adjacent areas, resultmg in uneven or mcomplete wells 9. Covering the gel with Saran Wrap is recommended to get complete polymerization of the upper region m contact with the air. After polymerizatton, gently remove the comb while addmg upper reservoir buffer with a syrmge. This avotds bubbles and maintains the wells stratght. If, after takmg these precautions, the wells’ walls are stall twisted, these can be straightened usmg a syrmge wtth a blunt needle by pushing the walls mto the correct posttion. 10. The loading of the samples may be eased by marking the bottom of the wells with a water-resistant marker. 11. The optimal and maximum amounts of the sample to apply into the wells depend on multtple condmons, such as the number of wells per gel, the size and thickness of the gel, the complextty of the sample, and the method of stammg, these should be determined empirtcally. Table 6 can be use as an mittal guide to dectde the starting conditions 12. In some cases, the purpose of the run is to subject the separated protems to blotting and immunodetectton. Thus, the gel should not be fixed or stained and it should be immediately equilibrated in electroblotting transfer buffer. 13. Place the tubes in the freezer at approx a 30” angle to maximtze the surface area of the solutton within the tube This provides a more effictent and faster lyophilization. A minimum of 1 h should be allowed for the freezmg process When the samples are frozen, uncap the tubes and seal them with paratilm Using

130

Ordovas a needle, prick the parafilm three or four times to allow the lyophilization of their contents. This process should be done expeditiously to prevent the thawing of the samples.

References 1. Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227,680-685. 2. Tadey, T. and Purdy, W. C. (1995) Chromatqgraphic techniques for the isolation and purification of lipoproteins. J. qf Chromatograp. B: Biomed. Appli. 671, 237-253. 3. Lehmann, R., Liebich, H., Grubler, G., and Voelter, W. (1995) Capillary electrophoresis of human serum proteins and apolipoproteins. Electrophoresis 16,998-l 00 1. 4. Mindham, M. A. and Mayes, P. A. (1992) A simple and rapid method for the preparation of apolipoproteins for electrophoresis. J. Lipid. Rex 33, 1084-1088. 5. Schagger, H. and von Jagow, G. (1987) Tricine-Sodium dodecyl sulfate-Polyacrylamide gel electrophoresis for the separation of proteins in the range of 1 to 100 kDa. Anal. Biochem. 166,368-379. 6. Ordovas, J. M., Litwack-Klein, L. E., Wilson, P. W. F., and Schaefer, E. J. (1987) Apolipoprotein E isoform phenotyping methodology and population frequency with identification of apoE and apoE isoforms. J. Lipid. Res. 28,37 l-380.

Isolation and Purification of Serum Lipoprotein (a) Using Lectin Affinity Leo J. Seman, Carl de Luca, and Jennifer

L. Jenner

1. Introduction

Lipoprotein (a) [Lp (a)] is a lipoprotein first discovered by K&-e Berg in 1963 in an attempt to study variants of j3 lipoproteins. (11. In 1973, Berg et al. discovered that Lp (a) correlated with coronary heart disease (CHD) in a case control study (2). Today, Lp (a) is considered by most as an independent risk factor in CHD. The mechanism by which it is atherogenic is not fully understood. Lp (a), structurally, is similar to low-density lipoprotein (LDL) in that it possessesa lipid core and apolipoprotein (apo) B-100, and its dissimilarity is that it contains an additional glycoprotein called apo (a) (3-5). Apo (a) shares homology with plasminogen, and both proteins contain cysteine-rich protein domains termed “kringles.” The kringles of apo (a) have the greatest homology with kringle IV of plasminogen and one copy of kringle V of plasminogen. The multiple copies of kringle IV vary in their homology with each other and are categorized as kringles Type l-10. Apo (a) kringle Type 2 is tandemly repeated anywhere from 15-40 times with only one copy of kringles type 1 and 3-10. These tandem repeats of kringle Type 2 give rise to many isoforms of apo (a) of varying molecular weight (3004300 kDa) (6-10). Because apo (a)‘s mode of inheritance and expression is autosomal codominant, one person can have two different isoforms (3). This nature of apo (a) has made the standardization of the quantification of Lp (a) difficult. Isolation techniques commonly employed for other lipoproteins such as ultracentrifugation are not optimal for Lp (a) because of apo (a)‘s many isoforms, which result in the particle being found throughout the density spectrum. Other techniques such as immunoFrom: Methods in Molecular Biology, Vol. 170: Lipoprotein Protocols Edited by: J. M. Ordovas 0 Humana Press Inc., Totowa, NJ

131

132

Seman, de Luca, and Jenner

affinity chromatography can be too specific for some epitopes found on some lsoforms but not others. The lectm-affimty method described in this chapter 1s lsoform Independent (11). The basis of this methodology takes advantage of the glycoprotem characteristic of apo (a). Wheat-germ agglutmm (WGA) 1s a lectm that has high affinity for N-acetyl-D-glucosamme (GlcNac) and iV-acetyl-neuraminic acid (NANA), which are two of the monosaccharides that are well represented on apo (a). Because there are very few other plasma protems that are as heavily glycosolated as apo (a), upon exposure to WGA bound to Sephacryl S-1000@, Lp (a) as a particle bmds to the WGA with minimal competition from other plasma protems. Following the removal of the unbound fraction of plasma, the Lp (a) IS released when the WGA is exposed to GlcNac at a higher molar concentration than 1s found on apo (a), thus releasing the particle in the eluate (II).

2.1. Equipment 1. 2. 3 4 5 6 7 8. 9. 10 Il. 12.

pH Meter, Accumet Basic (Fisher Scientific, Pittsburgh PA, cat no. 13-63%BAA) Scale, AC100 (Mettler, Highstown, NJ). Stir plate, Nuova II, (Thermolyne, Dubuque IA). Stir bar, 1” octagonal (FisherBrand, Pittsburgh, PA, cat no 14-51 l-63) Beaker, 100 mL Pyrex (Fisher, cat no 02-540H) Centrifuge Tubes, 50 mL (Falcon, Franklin Lakes, NJ). Centrifuge, RT6000 (Sorvall, Newtown, CT). Buchner ground-glass filter (FIsherBrand, Pittsburgh, PA, cat no 09-753-2). Vortex, Geme 2 (Sclentlfic Industries, Bohemia, NY) Rocker, Specs-Mix (Thermolyne, Dubuque, IA) MlcrofugeTM 12, (Beckman, Palo Alto, CA). Microcentrifuge tube filter, 0 45 pm (Thomas Sclentlflc, Swedesboro, NJ, cat no. 4596-606). 13 Pipetman P200, PlOOO (Ramm, France)

2.2. Materials 1 2 3 4. 5 6 7 8

Sephacryl S-1000 (Pharmacia, Plscataway, NJ, cat. no. 17-0476-01). Cyanogen Bromide (Sigma, St. Louis, MO, cat. no. C-6388) Wheat-germ agglutinin (WGA) (Sigma, cat no. L9640). 1 MNaOH (Sodium Hydroxide, Sigma, cat. no. S8045). 1 M HCl (Hydrochloric Acid, Sigma, cat. no H7020). 0 1 MNaHC03 (Sodmm Bicarbonate, Sigma, cat no S6297) buffer, pH 11 .O 1 MTrizma base (Sigma, cat no T-1503), pH 10 0 0.1 A4 Sodium acetate (Sigma, cat. no. S95 13), 0 5 M NaCl (Sodium Chloride, Sigma, cat no S9265), pH 4.0. 9 0.1 MNaHCO,, 0.5 MNaCl, pH 9 5. 10 0.1 MNaN3 Sodium Azide, (Sigma, cat. no. S-2002), 2 MNaCl, 0 05M NaHC03, pH 7.0.

Lp(a) Purification Using Lectin Affinity

133

11 Phosphate buffered saline (PBS)* 11 3 mM NaH2P04 (Sodium Phosphate monobasic anhydrous, Sigma, cat. no SO75 I), 15 3 mMNa,HPO, (Sodium Phosphate dlbaslc anhydrous, Sigma, cat. no. SO876), 200 m&I-NaC1, pH 7.2. 12. PBS + 200 mML-proline (PBS-p) (Sigma, cat. no. P8449). 13 PBS -t 200 nuI4 GlcNac (PBS-g) (Sigma, cat no. A8625). 14. Protein measurement assay (Bio-Rad, Hercules CA, cat. no 500-0006)

3. Methods 3.1. WGA is Immobilized to Sephacryl S-1000 (see Note 7) 3. I. 1. Sephacryl S- 1000 Is Washed and Allowed to Swell Wash 10 mL Sephacryl S-1000 three times with 4 volumes each of distilled water, incubate on rocker for 20-30 min and centrifuge (35OOg, 15min at 1OT) and decant supernatant. Wash overnight at 4°C m 4 volumes distilled water. It 1s ready for use or It may be stored for future use in one volume of a 0.1 M sodium azide, 2 h4 sodium chloride solution (see Notes 2-4). 3.1.2. Sephacryl S- 1000 Is Activated with Cyanogen Bromide 1. Make a 1.1 mixture of washed Sephacryl S-1000 and dlstllled water Adjust the pH of the mixture to 11 0 with a 1 A4 sodmm hydroxide solution and pour the mixture mto a 100 mL beaker with a 1-m. stir bar. 2. Measure 1 g of cyanogen bromide. Add a few drops of 1 MNaOH to cyanogen bromide to solubllize It (see Note 5). 3. Add solubilized cyanogen bromide to the Sephacryl S-1000 constantly stlrrmg and monitoring the pH as to maintam It between 10 5 and 11.5, by adding 1 M sodium hydroxide solution dropwise. This reaction will reach completion m approximately 20 mm and IS noted by reaching a pH plateau of 10 5-11.5 for at least 10 mm (see Note 6) 4. When the reaction is complete, transfer the activated Sephacryl S-1000 to a Buchner ground glass filter and use a low pressure vacuum to filter, filtering with 300 mL Ice-cold 0.1 M sodmm bicarbonate, pH 11 .O.

3.1.3. WGA is Immobilized to Cyanogen Bromide Activated Sephacryl S- 1000 1. While filtering the Sephacryl S-1000, prepare 100 mg WGA. To each 25-mg bottle of WGA, add 1 drop of 1 M HCl (WGA goes into solution better at acidic pH). Add 500 pL of a 0 1 M sodium bicarbonate solution, pH 11 .Oto each 25 mg bottle of WGA 2. When the Sephacryl S-1000 is filtered to a moist cake, transfer it to a beaker, using approx 1O-l 5 mL of 0.1 M sodium bicarbonate solutton, pH 11 0 To this add the solubihzed WGA and stir at 4’C for 16-l 8 h 3. Transfer the mixture to a 50-mL centrifuge tube and centrifuge (35OOg, 15 min at 1O’C) remove the supernatant and measure its protein content (see Note 7).

134

Seman, de luca, and Jenner

4 After this overnight incubation, a blocking step IS necessary to react with lmldlzoles created m the activation step. Incubate activated Sephacryl S- 1000 with 40 mL of 1 M Trizma base solution at 4°C for 4 h. 5 After 4 h, pour WGASephacryl S-1000 into a Buchner ground glass filter and filter with a 0 1 Msodium acetate, 0.5 MNaCl solution, pH 4 0. (We used approx 300-500 mL.) 6. Filter with a 0.1 M sodium bicarbonate, 0 5 MNaCl solution, pH 9 5 (We used approx 300-500 mL ) 7. When almost dry, transfer the WGA Sephacryl S-1000 to a 50 mL centrifuge tube, add 40 mL of the PBS + 200 mML-proline solution (PBS-p). Resuspend by rocking or vortexmg Store at 4’C (see Note 8)

3.2. Lipoprotein (a) Isolation and Purification 3.2.7. Lipoprotein (a) is Isolated and Purified from Serum, Macromethod 1 In a 50-mL centrifuge tube add 5 mL of WGA.Sephacryl S- 1000 and up to 10 mL of serum, from which Lp (a) will be isolated (see Note 9) 2. Rock mixture at room temperature for 20 min 3 Pour mixture mto wide diameter column where WGA Sephacryl S-1000 will be less than 1 cm column height. Allow the elutlon of the PBS-p and the unbound fraction of the serum (see Note 10) 4. To the column add 2-3 volumes of PBS-p equivalent to the amount of serum added. For example add 20-30 mL PBS-p if 10 mL of serum was originally added. Allow this to elute completely 5 To the column add 1 volume of PBS + 200 mA4GlcNac (PBS-g) equivalent to the serum added For example, add 10 mL PBS-g if 10 mL of serum was originally added. Collect this elutlon as Lp(a) (see Note 11).

3.2.2. Lipoprotein (a) Is Isolated and Purified from Serum, Micromethod 1 In the top chamber ofthe mlcrocenttlfuge tube filters add 150 (L WGA.Sephacryl S1000 and up to 300 pL of serum. Close and place on rocker for 20 min (see Note 12). 2 Centrifuge the mlcrocentrlfuge tube filter m the mlcrofuge at top speed (12,000g) for 3 min 3 Replace lower chamber of mlcrocentnfnge tube filter with a new one and to the top chamber add 300 pL of PBS-p. Place on rocker for 5 mm 4 Centrifuge the mlcrocentrlfuge tube filter in the mlcrofuge at top speed (12,OOOg) for 3 mm 5 Replace lower chamber of mlcrocentnf%ge tube filter with a new one and again to the top chamber add 300 pL of PBS-p. Place on rocker for 5 min. 6. Centrifuge the mlcrocentrlfuge tube filter m the mlcrofuge at top speed (12,000g) for 3 mm 7. Replace lower chamber of microcentrifuge tube filter wrth a new one and to the top chamber add 300 clr, of PBS 200-g Place on rocker for 15 min

Lp(a) Purification Using Lectin Affinity

135

8 Centnmge the microcentrifuge tube filter in the mtcromge at top speed (12,OOOg) for 3 mm. 9 Collect eluate as Lp (a) (see Note 11)

4. Notes 1 A commercrally avatlable product, LtpoPro TM Lp(a)-Cholesterol reagent, produced by Genzyme Corporation (Cambridge, MA, cat. no 80-4383-02) can be used in place of WGA nnmobilized to Sephacryl S-1000. 2 Sephacryl S-1000 is usually very heavily packed m the bottle and it IS recommended to agttate the bottle to resuspend it. When resuspended, pour destred amount into a 50-mL centrifuge tube. Centrifuge (3SOOg for 15 min at 10°C) to repack and assess volume. 3. If using previously swelled and stored Sephacryl S-1000, first wash two times with 4 volumes each of dtsttlled water; incubate 20-30 min for each wash. If Sephacryl was stored m distilled water only, wash first with 0.1 M sodium aztde, 2 A4 sodium chloride for 20 mm, then wash twice with 4 volumes distilled water. 4 One volume is equal to the volume of the packed Sephacryl. For example, if 10 mL of Sephacryl is used, then one volume would equal 10 mL, 2 volumes would equal 20 mL, and so on Warnzng* The activatton step must be performed in the hood, wearing the proper safety attire. Monitor the temperature of the mixture and maintain tt at 20°C by the periodtc addition of crushed ice. Measuring protein content of supernatant gives an estimate the amount of WGA bound to Sephacryl S- 1000. It is desired that, mmimally, 5 mg of WGA be bound to 1 mL of Sephacryl S-1000 m order to ensure adequate binding of Lp (a) to WGA.Sephacryl S-1000. 8. If the WGA:Sephacryl S- 1000 will not be used immediately, a 0.1 MNaNs, 2 A4 NaCl, 0 05 A4 NaHCOs solutron, pH 7.0, should be used m lieu of the Indicated PBS-p solution. When ready to use the WGA:Sephacryl S-1000, it is necessary to replace the storage solution with PBS-p which can be accomplished by centrtfugation (35OOg, 15 mm at 1O’C) and decanting the storage solution and adding 40 mL of PBS-p Resuspend. 9 The volume of serum can vary depending on the concentratton of Lp (a) m the serum. One mL of packed WGA:Sephacryl S-1000 can bind up to 100 mg of Wa) 10. The following procedure IS best performed using a fraction collector, ultravtolet (UV) spectrophotometric momtor measuring absorbance at 254h, and a recordmg instrument. The UV monitor should be zeroed with PBS-p. II The GlcNac elutton will contam other plasma glycoprotems of varying concentration m addition to Lp (a) Depending on the desired purity of Lp (a) these other plasma glycoprotems can be removed by size-exclusion chromatography with an upper limit less than IO6 molecular weight. I would suggest Sephacryl S-300 HR or equivalent for this purpose This is best achieved by devising a column to have

136

Seman, de Luca, and Jenner

a void volume V, equal to the total N-Acetyl-o-Glucosamine eluate applied. Pure Lp(a) will elute in void volume (V,) Run on PBS-p at 20°C 12 For the addltlon of the PBS-p and PBS-g in the followmg steps, 300 pL is listed for the volumes, but that is determined by the volume of serum added For example, if 150 pL of serum 1sadded, the volume for the followmg steps IS 150 pL. This allows the concentration of the Lp(a) in the GlcNac eluate to be same as the concentration m the serum orlgmally added. To concentrate or dilute the Lp(a), adjust the volume of the PBS-g accordingly

References 1 Berg, K. (1963) A new serum type system in man- the Lp system. Acta Pathol Mwroblol &and 59, 369-382. 2. Berg, K., Dahlen, G , and Frock, M. H. (1974) Lp (a) lipoprotein and pre-/3-hpoprotem in patients with coronary heart disease. Clrn. Genet 59, 230-235 3. McLean, J. W, Tomlinson, J. E., Kuang, W. -J., Eaton, D. L., Chen, E. Y., Fless, G. M., et al (1987) cDNA sequence of human apohpoprotem (a) 1shomologous to plasminogen. Nature 330, 132-l 37. 4 Fless, G. M., Zummalen, M. E., and Scanu, A. M (1986) Physlologlcal properties of apolipoprotem (a) and hpoprotem (a)- derived by the dissociation of human plasma lipoprotein (a) J Bzol Chem 261, 87 12-8718 5. Seman, L J. and Breckenndge, W. C. (1986) Isolation and partial charactenzatlon of human apohpoprotem (a) from lipoprotein (a) Biochem Cell Biol 64, 999-l 009 6 Gavlsh, D., Azrolan, N., and Breslow, J. L. (1989) Plasma Lp (a) concentration IS inversely correlated with the ratio of Krmgle IV/Kringle V encoding domains m the apo (a) gene. J Clua Invest. 84, 2021-2027. 7 Kochinsky, M. L., Beislegel, U., Henne-Bruns, Eaton, D. L , and Lawn, R. M (1990) Apohpotem (a) size heterogemcity IS related to the variable number repeat sequences m its mRNA Blochemlstry 29, 640-644 8. Lindahl, G., Gersdorf, E., Menzel, H. J., Seed, M., Humphnes, S., and Utermann, G. (1990) Varlatlon m the size of human hpoprotem (a) is due to a hypervarlable region in the gene Human Genet 87, 2 153-2 16 1. 9 Lackner, C , Cohen, J C , and Hobbs, H H (1993) Molecular defimtlon of the extreme size polymorphism m apollpoprotem (a). Human Mol Genet 2, 933-940 10. Lackner, C., Boerwmkle, E., Leffert, C. C , Rahmlg, T., and Hobbs, H. H. (199 1) Molecular basis of apolipoprotem (a) lsoform size heterogeneity revealed by pulse field gel electrophorerls J Clm Invest 87, 563-567. 11. Seman, L. J , Jenner, J L , McNamara, J R , and Schaefer, E. J. (1994) Quantlfication of lipoprotein (a) in plasma by assaying cholesterol in lectm-bound plasma fraction. Clan Chem. 40(3), 40@403.

Electrophoretic Separation of LDL and HDL Subclasses David L. Rainwater 1. Introduction An extensive literature supports the hypothesis that measurements of lipoprotein subclasses provide a more detailed reflection of lipoprotein metabolism and a more accurate prediction for risk of cardiovascular disease. Lipoprotein subclasses have been resolved on the basts of density fractionation, gel filtration, and immunological and electrophorettc procedures. Perhaps one of the most widely used methods has been the resolution of lrpoproteins on the basis of parttcle size using nondenaturmg pore gradient gel electrophoresis (GGE). GGE procedures have been based on highly reliable gradient gels supplied m two formats: PAA2/16 for resolving low-density lipoproteins (LDLs) and PAA4/30 for resolvmg high-density lipoprotems (HDLs) (Pharmacia, Piscataway, NJ). These gel formats produce repeatable and detailed separations of the two major classes of lipoprotem particles, appropriate for quantitative analyses of lipoprotein phenotypes. Following are descriptions of methods we use to separate and quantrtate lipoprotein subclasses.Because of the uncertain supply of gradient gels from commercial sources, a protocol for making gradient gels is described below (Subheading 3.1.). In addition, several methods for staining different lipoprotein constituents are described and strategies for measuring lipoprotein phenotypes are discussed. In describing our methods, I have attempted to provide the details necessary for an interested investigator to adapt the protocols to specific research objectives.

From

Methods m Molecular Biology, Vol 110 Llpoprofem Protocols Edited by J M Ordovas 0 Humana Press Inc , Totowa, NJ

137

138

Rainwater

2. Materials 1 Each of the four acrylamtde stock solutions used to construct the three gradients are made up m GGE buffer, which contains 90 mM Tns, 80 mM borate, 2.5 mA4 ethylenedlamme tetra-acetic acid (EDTA), pH 8 3. The four stock solutions are made as follows a 292.95 g/L Acrylamlde and 17 05 g/L bu-acrylamlde (3 1% total, 5 5% crosslinker) b 32 g/L Acrylamlde (3 2% total, 0% crosslmker) c 172.8 g/L Acrylamide, 7 2 g/L bzs-acrylamlde, and 50 g/L sucrose (18% total, 4% crosslinker) d 28.8 g/L Acrylamlde and 1.2 g/L bzs-acrylamlde (3.0% total, 4% crosslmker) Solutions are filtered through scmtered glass, degassed by somcatlon, and stored at room temperature for as long as 6 wk Each stock solution 1s made at least 1 d m advance of use Working solutions are made Just before casting. e. 3 1% To each L of stock solution a, add 1.5 mL freshly prepared ammonium persulfate (100 g/L) and 0.25 mL 3-dlmethylammoproplomtrlle f 3.2%: To each L of stock solution b, add 4 6 mL ammonium persulfate and 0 6 mL 3-dlmethylaminoproplomtrlle g. 18%. To each L of stock solution c, add 1 5 mL ammonium persulfate and 0 25 mL 3-d1methylammoprop1omtrlle h. 3%: To each L of stock solution d, add 4 6 mL ammomum persulfate and 0 6 mL 3-dlmethylammoproplomtrlle 2. The gradients are formed using a WKZ dual-pump gradient controller system (ISCO, Lincoln, NE). In precisely controllmg the flow rates of two peristaltic pumps, one pumping a high-limit solution and the other a low-limit solution, the system allows generation of any gradient that can be adequately specified by less than 20 linear gradient segments. The two streams of acrylamide solution must be mixed before entering the casting chamber; specifications for an external mixmg chamber are given elsewhere (I). We have used two types of casting chamber, the GSC-8 Gel Slab Casting Apparatus (Pharmacia), which holds 8 cassettes, and one holding 16 cassettes made for us by a local plastics firm (Taylor Plastics, San Antonio, TX) The glass cassettes for holding the gradient gels are made from two acid-washed sheets of glass (82 mm x 82 mm x 1 0 mm, VWR) and two Delrm plastic spacers (82 mm x 4 0 mm x 2 7 mm, VWR). The cassettes are held together by use of thm plastic tape (3/4 m wide, Lemac Akron, Talladge, OH) Making tight cassettes will help prevent the development of bubbles between gel and glass (I). 3. The gradient gels are run m a Pharmacla GE-2/4 Electrophoresls Apparatus (also supplied by Taylor Plastics; see Note 1). The apparatus has a pump that clrculates buffer between electrode vessels and it has coolmg coils m the lower reservoir that permit use of a circulating coolmg bath to control temperature 4 Commonly, the high molecular weight standards cahbratlon kit (Pharmacla), a mixture containing thyroglobulm (17 nm diameter), ferritm (12 2 nm), lactate

Separation of LDL and HDL Subclasses

139

dehydrogenase (8.16 nm), and albumin (7.1 nm) are used as calibrators for HDL subclasses. Thyroglobulin and ferritm, together with carboxylated polystyrene microspheres (38 nm diameter, Duke Scienttfic Corporation, Palo Alto, CA), are used to calibrate gels for the larger lipoproteins. In addition, a laboratory-standard lipoprotein sample should be run in each gel to assessconsistency of separation and staining characteristics.

3. Methods

3.7. Nondenaturing

Gradient Gels

For many years, nondenaturmg polyacrylamrde gradient gels were commercially available from Pharmacra, and there are now several commercral sources for similar gradient gels (see Note 2). The gels, designed to fit the Pharmacra GE-2/4 Gel Electrophorests Apparatus, are nearly 3 mm thick and m then cassettesmeasure approx 75 x 75 x 5 mm. Such gels are able to accommodate the large protein loads necessary for detecting hpoprotem constituents that may occur in relatively low concentrations in plasma (2). In addition, lipoprotein size properties have been measured m other electrophoresrs systems,both commercial and homemade Following are given our procedures for making gradient gels that duplicate the characteristics of those originally made by Pharmacra. Some detarls of the following methods for making HDL and LDL gradient gels have been pubhshed (I,31 and will not be repeated here. The purpose of this section IS to summarrze m one place a protocol for casting gradients and to provide detarls for making a new gel, the composite gradient gel, which enables separattons of LDLs and HDLs in the same gel 1. Each day, the peristaltic pumps are calibrated as suggested by the manufacturer. The plastic tubing from each pump is connected to the external mixmg chamber, which m turn is connected to a three-way stopcock at the bottom of the castmg chamber by a short length of plastic tubmg. 2. The gradient is made at room temperature (about 25 + 1’C) and each casting chamber is protected from temperature fluctuations (including the heat generated by polymerizing acrylamide m a nearby chamber) by use of foam msulators 3. After leveling the castmg chamber, 30 mL of 20% ethanol IS uqected into the chamber via the three-way stopcock, followed by mitiation of the gradient 4 Table 1 presents the characteristics of three acrylamide gradients, two were designed for specialized studies of LDLs and HDLs, and one, the composite gradient, was designed to permit analyses of both types of lipoprotems m the same gel The gradient-making system holds constant the summed flow rates from the two pumps; what varies is the proportion of total flow that is provided by each pump. Thus, we need only specify the proportion of total flow that is provided by the high hmit solution pump to completely describe the gradient characteristics.

Table 1 Descriptions Nondenaturing the high-limit

of Gradient Gradient acrylamide

Segments for the HDL, LDL, and Composite Acrylamide Gradients Used in Making Gels for Lipoprotein Separations (Given are the percents of total flow that come from solution at the start and end of each gradient segmenta) LDL Gel

Gradient Segment Number $

1 2 3 4 5 6 7 8 9 IO

Duratron (ml@ 2.0 2.0 20 20 20 2.0 2.0 2.0 2.0 2.7

% High (S-0 00 54 112 17 1 23.3 30 1 37 6 46.2 56.3 69 5

HDL Gel % Hrgh (End) 54 II 2 17 I 23 3 30 1 37 6 46 2 56 3 69 5 100.0

Duratton (ml@ 17 17 25 25 24 24 24 24 10 1.5

% High wm 0.0 5.0 10.8 I6 6 22 6 29 0 36 9 45 7 56 7 71 9

Compostte Gel % Htgh (End) 5.0 10.8 16.6 22.6 29.0 36.9 45.7 56 7 71.9 100 0

Duratron (ml@ 2.0 20 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0

=For the LDL gradient, the high- and low-hmtt solutions are 18 and 3% acrylamlde, respectively (see Subheading they are 3 1 and 3 2%, respectively, and for the composite gradient, they are 3 1 and 3%, respectively

% Hrgh Pw 0.0 2.5 57 96 13.8 18.7 24.6 31.4 41.9 59.3

% High (End) 25 57 96 13 8 18 7 24 6 314 419 59.3 100.0

2.); for the HDL gradlent,

Separation of LDL and HDL Subclasses

141

The actual volume of a gradient required (and of the two hmtt solutions) varies with type of gradient and with mdtvtdual castmg chamber charactertsttcs and must be determined empmcally. The gradient volume is controlled by changmg the pump flow rates. 5. Following completion of the gradtent, a volume of 60% sucrose IS pumped mto the chamber bottom to drsplace exactly the acrylamrde gradient mto the glass cassettes, and the stopcock IS closed Allow polymerization to proceed for at least 3 h. 6. Before dtsturbing the casting chamber, first check that a sharp interface line has appeared between the polymerized gel and the overlayermg solutions, and touch the stde of the chamber to assure that the temperature has returned to room temperature The block of polyacrylamide, whtch contams the gel cassettes, IS slowly removed from the castmg chamber under water (to ensure an IS not aspirated mto the cassettes). Indtvtdual gels are cleaned and carefully inspected before sealmg them m small plastic bags with 2 mL of GGE buffer (see Note 3). Sealed gels may be stored in the refrtgerator for as long as two months with no apparent adverse effects 7. Thts IS a very reliable protocol, but several aspects of the procedure should be monitored m order to assure an exactly predictable gradrent gel before any samples are commrtted (see Note 4). Careful records of each gel castmg will help identify potential sources of problems that may develop and other gels possibly affected.

3.2. Electrophoresis 1, Plasma samples are cleared of partrculate materials by a brief centrtfugatron prior to use. To 20 pL plasma, add 30 pL of 1 g/L bromophenol blue m a 40% sucrose solution Load 10 pL (i e ,4 pL plasma) in each lane (see Note 5) 2. Prechill the GGE buffer and keep it cold during electrophoresrs by runnmg in a cold room or by use of a circulating cooler bath Carefully insert gel cassettes m the slots m the upper electrode vessel making sure that the gel surface IS exactly parallel to the bottom of the vessel. The well-former is inserted so that about 1 mm of gel surface pushes mto each well. Remove any trapped au bubbles with a prpet. Assemble the upper and lower electrode vessels with the safety cover and prerun the gel at 120 V for 1 h 3 Load all samples, control products, and cahbrators (except the carboxylated polystyrene microspheres) m the well-formers. Subject the gels to electrophoreSISm the following sequence. 15 V for 15 mm, 70 V for 20 min, and 125 V for 24 h (1 e , 3,000 Vsh) Because proteins will bind to the microspheres and thus reduce mobthty, load this calibrator separately: 3 h after imttation of electrophoresis, dilute microspheres 20-fold (in a 1: 1 mixture of 1 g/L bromophenol blue m 40% sucrose and GGE buffer) and load 3 & in the lane with the high molecular weight standards during a brief interruption of the electrophoresrs protocol

142

Rainwater

3.3. Staining Choice of stain depends primarily on the questions to be answered by the data. However, limitations of equipment may also play a role in selecting an appropriate staining procedure. For example, the LKB laser densitometer scans at 632.8 nm and, therefore, cannot be used with Oil Red 0 staining. Below (Subheadings 3.3.1. and 3.3.2.) are given two protocols we routmely use to stain lipoproteins followmg GGE, one a lipid and one a protein stain 3.3.1. Sudan Black B (SETS) One of the more commonly used stains for lipoproteins, SBB stains a number of purified lipids, mcludmg cholesteryl esters and triglycerides (4). However, several studies have shown that SBB stammg of lipoprotems predicts cholesterol drstributlons well and that this IS not affected m hypertriglycerrdemrc samples (5-7). Methods also have been descrtbed for stammg hpoproteins wrth SBB prior to electrophorests (7); this approach enables densitometry to be done while the gel remains m the cassetteand reduces exposure to Cellosolve (see Note 6). 1 Make the SBB stain solutron as described (8) by slowly addmg 5 g SBB to a zmc acetate solutron (8 g zmc acetate dissolved in 1 L water plus 0.4 L Cellosolve) whtch is reduced gradually to 450 mL by heating at 100°C. Filter the SBB stain solutton whtle hot, after cooling to room temperature, and again just before use 2. With gentle (70 rpm) shaking of the gel at room temperature, fix m 10% trtchloroacetic acid for 1 h, soak in 30 mL 50% Cellosolve m water for 1 h, soak overnight in 30 mL of SBB stain solution, and destain wtth multtple changes of 50% Cellosolve in water (takes about 6 h). Occastonally stain parttcles adhere to the gel surface, remove these by careful rubbing with a cotton-tipped applicator 3 Stop the destaimng process when lipoprotems and microspheres are clearly VIStble and most of the background color IS removed; some additional destaming occurs during the subsequent steps. 4. Soak the gels overmght rn Coomassre destain solutton (methanol:H20:acettc acid, 20: 17:3) along with a piece of polyurethane sponge foam to facthtate complete destaining 5 Some gel calibrators, including the hpoprotem-standard solution and the carboxylated polystyrene mrcrospheres, are visualized by SBB stammg procedure. However, the protems in the high molecular weight standards must be stained separately The standards lane 1s identified by the stained mtcrospheres and by the absence of stained hpoprotems characterrsttc of nearby sample lanes. Center a narrow (2 mm wide) strip of filter paper over the standardslane and soak by dropwise addmon (use a 25pL Hamilton syringe) of Coomasste stam (0 7.5 g Coomassre brrlhant blue R-250/L of HzO:methanol.acettc acid, 5.4.1). Keep the filter paper wet-appearing by continued apphcatron of stain for about 45 mm (see Note 7) Destam the gel m several changes of the Coomassre destain solutton

Separatron of LDL and HDL Subclasses

143

6 After destaming, restore the original gel shape by soakmg overnight m several changes of GGE buffer contammg sodium azide (20 mg/L) Gels may be stored with this buffer m sealed plastic bags for several years wrth no apparent loss of stain.

3.3.2. Immunological

Stains for Apolipoproteins

We have described methods for the immunological detection of apohpoprotein distributions among baboon and human plasma lipoproteins (9,10). The prmcipal advantage of this approach is that once the immunological detection system is estabhshed, then any protein for which a specific antibody is avatlable, may be probed, Distributions of some protems, such as apo E or apo CII, may reflect unique aspects of lipoprotein metabolism independent of the general distributions of lipids among lipoprotems. Although we can expect the distributions of apo AI among HDL subclassesto be similar to the distributions of cholesterol, the distributions of many other apolipoproteins are quite distmctive. 1 Following electrophoretic resolutton, transfer plasma hpoproteins m the gel to nitrocellulose paper (0 2 pm BA83, Schliecher and Schuell, Keene, NH) by holdmg 50 V constant for 24 h in a TE22 transfer unit (Hoeffer Scientific Instruments, San Francisco, CA) with GGE buffer. This procedure generates heat, which must be removed by a clrculatmg cooling bath We established that thus protocol ensures nearly quantitatrve transfer of hpoprotems: about 90% of the radioactivity m a plasma sample spiked with radtoiodmated lrpoproteins was recovered on the mtrocellulose paper after electrophoretic transfer and the immunochemical procedures described m the followmg paragraphs. 2. After transfer, fix the paper with 2% glutaraldehyde and block other protein bmding sites on the paper with 5% milk proteins dissolved in phosphate-buffered saline (PBS) wrth 1 g/L Tween-20 and 0.5 g/L Antifoam A 3 The distributions of apoltpoproteins among lipoproteins on the paper are detected by sequential binding wrth a prrmary antibody (e.g., sheep antrhuman apo AI) and with an appropriate secondary antibody (e.g , donkey antisheep IgG) both diluted in the blocking solution The secondary anttbody is radiorodinated using the chloramme T method (II) and an autoradlogram 1s made to record apolipoprotem dtstrrbutions. 4 Calibration is made difficult by the fact that the calibrator protems are not vtsible on the autoradrogram. However, the locations of the calibrator proteins are readily visualized on the mtrocellulose paper by use of Ponceau S staining (12) Pamting the corners of the paper with a phosphorescent dye (e.g., autoradiography pen, Bel-Art Products, Pequannock, NJ) before autoradiography ~111 leave marks on the autoradrogram that permit an exact alignment of caltbrators measured on the mtrocellulose paper wrth data obtained from the autoradrogram. Peroxrdaselabeled antrbodres coupled with chemrlummography (e.g., ECL Western Blot-

144

Rainwater tmg, Amersham InternatIonal, Little Chalfont, Buckinghamshlre, UK) 1s an alternatlve to the use of radloactlvlty for detectingantibody locations.

3.3.3. Other Stains 3 3.3.1. COOMASSIE

In contrast to the other stains described, this protein stam cannot be used with plasma samples owmg to lack of specificity. Therefore, hpoprotems must first be isolated from plasma, usually by ultracentrifugation in a high-density salt solution, prior to electrophoresls. To avoid problems of streakmg and spreading, all lanes are loaded with the same strength salt solution. Because apo B 1sthe predominant protein among p lipoproteins, one can assume that Coomassie stammg profiles correspond to distrlbutlons of apo B. However, among HDLs, a significant portion of stain is bound by proteins other than apoA1. GGE-resolved lipoprotems may be stained with both types of Coomassle brilliant blue: R-250 and G-250 (13,147. 3.3 3.2. LIPID STAINS

Oil Red 0 1sa commonly used stain for lipids m plasma llpoprotems (Z4J Soak gels in 0.4 g/L 011 Red 0 dissolved m 60% ethanol at 55-60°C for 24 h, followed by destaining with 10% acetic acid. The fluorescent dye, fillpin, has been used to stain unesterlfied cholesterol. After electrophoresis, fix the gel with 10% trxhloroacetlc acid, soak with filipin solution, destain with PBS, and scan with a fluorometer (15). Alternatively, the gel may be illumlnated with ultraviolet (UV)-A lamps and photographed using high-resolution negative film (4 x 5 in., Type 55 film, Polaroid Corporation, Cambridge, MA). Slightly larger than the original gel, the negative Image gives a positive correlation between fluorescence and density and 1s readily scanned by any densitometer. Load protein molecular weight standards at twice recommended concentration. These are vlsuahzed as very slight peaks and valleys on the negative film. 3.4. Densitometry and Analysis In densitometry, restoration of original gel shape is essential (monitored as straight gel edgesand lanes), as is proper alignment of the gel. In order for gel calibration to be correct for all lanes,eachabsorbanceprofile must start and end at the samepoints (1.e, pore sizes)for all lanes on a gel. One way to ensure correct alignment 1s to load the same sample (i.e., the hpoprotem-standard sample) in the outside lanes, when the two absorbance profiles are superimposable,then all lanesin between are properly aligned. 2. An absorbance profile, representing the size dlstrlbutlons of hpoprotem constltuents, can be produced by any densitometer capable of scannmg wet gels or film.

Separation of LDL and HDL Subclasses Many densitometersarepackagedwith software capable of analyzmg absorbance profiles Ideally, the densitometeralso generatesan ASCII file, which can be evaluated by a variety of other programs, both commercial and private, to overcome any limitations of the densitometer software. Densitometer performance should be monitored on a routme basis(seeNote 8) 3 Calibrate HDL gels using the relative mobilities of thyroglobulm, femtin, lactate dehydrogenase, and albumin in the high molecular weight standards mixture (16). Calibrate LDL gels using the relative mobllltles of thyroglobulm and ferritin from the high molecular weight standards mixture plus the

mlcrospheres (161. 4. To identify the baselme, inspect all absorbance profiles from a gel for consensus low points (often occurring at 22 and 32 nm in the LDL size region and at 7 2 and 14 nm in the HDL sizeregion). Theseconsensuslow points canbe usedto specify a gel baseline or, alternatively, a gel lane(s) can be set aside to measure the absorbance profile for background stain (1 e., do not load plasma mto one or more lanes). The baseline absorbance is subtracted from the absorbance profile for each lane to generate the absorbance owing to staining of lipoprotein constituents 5 A properly calibrated gel IS the cornerstone of high-quahty data It is therefore useful to calculate diameters for several consistently occurring features of the lipoprotem standard sample This will test the calibration curve for the gel. It 1s important to compare carefully features of each sample lane and the respective absorbance profile, requiring several points of correspondence (e.g., relative peak heights and mobdlties both within and across sample lanes). Occasionally, such comparisons need to be done long after the initial evaluation of the gel, so It IS wise to keep as much of the original data as possible (see Note 9)

3.5. Evaluation

of the Data

3.5.1. Comparison of Gel Formats Figure 1 presents a selection of human and baboon samples run on the three HDL, LDL, and composite. As can be seen, the mobilities and resolutions of the different LDL bands are quite similar for the LDL and the composite gel formats. Although the HDL patterns are somewhat compressed m the composite gel format when compared to the HDL gel format, albumin, a key standard in gel calibration for HDL profiles, is retained. Based on the gradient characterlstlcs given in Table 1, it is posstble to calculate acrylamlde concentrations along the length of the gel. Figure 2 diagrams the percent acrylamide as a function of gel length (i.e., relative mobility or R,) for each of the three gel formats. Also presented are the calculated acrylamlde concentrations based on measured mobilities of specific particles run m each of the three gel formats. Although mobllitles differed, particles came to rest in similar pore sizes(i.e., percent acrylamlde) across the three gel gradients. Thus, the mobllity of lipoproteins m different gel gradients is predictable. gel formats,

146

Rainwater

Fig. 1, Human and baboon samples run on the HDL (A), composite, (B), and LDL (C) gel formats. Nine lanes containing the same sequence of samples are shown for each gel; the four lanes on the left contain human-plasma samples and the four lanes on the right contain baboon-serum samples. The center lane contains standards whose locations are indicated to the sides; in order of decreasing diameter, the standards are: M, microspheres; T, thyroglobulin; F, ferritin; C, catalase; L, lactate dehydrogenase; A, albumin. Lipoproteins were stained with Sudan black B and proteins in the standards lane were stained with Coomassie brilliant blue as described in the text (Subheading 3.3.1.).

3.5.2. LDL Phenotypes Human LDL particles fall, generally, within the range of 22-29 nm in diameter. Several nomenclatures have been devised to describe LDL subclasses (I 7), but there is no general consensus,resulting perhaps in part from an inadequate understanding of the metabolic bases for LDL heterogeneity. Therefore, several approaches to describing LDL sizephenotypes derived from GGE have been used. Apparently simplest is to dichotomize LDL absorbance patterns into type A (peak diameters larger than 25.5 nm) and type B (peak diameters smaller than 25.5 run). This is a subjective procedure that is best done by developing a consensus phenotype among several independent readers (IS). Of course, LDL peak diameters occur as a continuum, making it difficult to type some profiles with LDL peaks in the intermediate size region. Thus, some researchers measure diameters of the predominant LDL species,yielding a continuous variable. By comparison with the categorical variables, the exact mode of inheritance for a major gene controlling LDL size phenotype could not be established biologically

when LDL diameters relevant information.

were used (19), suggesting a degradation of Frequently, there are two (or more) distinct

147

Separation of LDL and HDL Subclasses

Acrylamide

(%)

Fig. 2. Percent acrylamlde correspondmg to different relative mobWles (R,) for each of the three gel formats. Total acrylamlde concentration at each point along the gel length was calculated from mformatlon m Table 1 assuming that gel length 1s lmearly related to the gradient. The lines represent calculated acrylamlde concentrations along the gel length for (A) HDL gel, (B) composite gel, (C) LDL gel The symbols represent the acrylamlde concentrations calculated from relative moblllties of the followmg particles circle, baboon LDL in the lane to the right of the standards lane m Fig 1; square, thyroglobulm; triangle, ferritm; diamond, lactate dehydrogenase

LDL bands m a sample, only one of which can be scored as predominant (even though bands may have nearly equal Intensity). Therefore, a limltatlon of measurmg just predommant LDL peak diameter IS a necessity to ignore the contnbutlon to LDL phenotype resulting from other minor LDL bands. This hmltation has led to strategies devoted to generating a variable that better describes the size dlstrlbutlons of all LDLs. For example, Campos et al. (20) calculated an LDL particle score, which IS a size estlmate weighted by the proportional absorbance of the different LDL peaks. Variations on this strategy include expressing the proportion of total LDL absorbance m larger particles (e.g , larger than 25.5 nm) (3) and estimating

the median LDL diameter

(I e , the diameter of LDL particles for which summed absorbance of larger LDLs equals the summed absorbance of smaller LDLs). Finally, a promising strategy has been developed that analyzes the entire LDL absorbance profile on a coordinate by coordinate basis (21). Although computatlonally difficult, this approach does not require knowledge of subclassesand has the potential to identify

LDL size regions influenced

by specific metabolic

parameters.

148

Rainwater

3.5.3. HDL Phenotypes Inspection of frequency histograms for HDL peak diameters suggested the existence of five HDL subclasses,two falling m the srze region of HDL, density particles and three falling in the size region of HDL:, density particles (13). Approximate mean diameters and corresponding mean hydrated denslties for the five subclasses were found to be 10.57 nm and 1.085 g/mL for HDLZ,,, 9 16 nm and 1.115 g/mL for HDL*,, 8 44 nm and 1.136 g/mL for HDL3,, 7.97 nm and 1.154 g/mL for HDLjb, and 7.62 nm and 1.17 1 g/mL for HDL3, (13). Proportional absorbance occurring m each subclass may be estimated by summing absorbances between the boundaries for the subclasses (i.e., the diameters, 13, 9.7, 8.8, 8.2, 7.8, and 7.2 nm form boundaries for the five HDL subclasses /16J or by fitting curves that describe each subclass (10,22). Other strategies to describe HDL size phenotypes are similar to those used for LDL phenotypes: (a) HDL particle score (2) or median diameter, (b) percent of absorbance on particles larger than HDL3 (3); and (c) raw absorbance at each coordinate of the absorbance profile (23). Although presently undefined, there also are HDL particles larger than HDL,, often termed HDL, . Particularly prominent m some lines of baboons, HDL, particles contain cholesterol, apo AI, apo AII, and apo E (7,9). Human HDLs stained for cholesterol or apo AI generally do not extend beyond 13 nm m diameter (i.e., HDL&, although a high proportion of apo E occurs on such HDL,-size particles (unpublished data). 3.5.4. Conclusion Although a number of techniques have been developed to separate classes of hpoprotems, perhaps none surpasses the advantages of GGE, in terms of repeatability, resolution, and ease of use. That lipoprotein size phenotypes provide useful mformatlon about cardiovascular disease is evident m results from many laboratories. The challenges for the future will be to understand better the metabolic bases for variation in hpoprotem size phenotypes and to determine how they are related to onset and progression of cardlovascular disease 4. Notes 1. Because the sources of speclahzed electrophoresls chambers have become uncertain, we have evaluated an alternatlve, the Mim-PROTEAN II (Bio-Rad, Hercules, CA) Two gradient gels can be subjected to electrophoresis m each unit. This inexpensive unit has no pump to mix the reservoir buffers and It must be run m the cold because there are no cooling coils. Nevertheless, the llpoprotein patterns for gels run in this unit are virtually identical to those run m the more expensive electrophoresls chambers.

Separation of LDL and HDL Subclasses 2 At the present there are two commercial sources of gels designed as substttutes for the Pharmacta PAAY16 and PAA4/30 gradient gels: ISOLABS (Akron, OH) and Alamo Gels (San Antonio, TX). 3 Routinely, we store gels m plastic bags made from 4 mil4-in tubular plastic rolls that are heat-sealed at each end using a thermal bag sealer (National Bag Company, Hudson, OH) We also have used 4 mil 4 x 4 in reclosable zipper bags (National Bag), which are less convenient because of the smaller opening, for the same purpose. 4 To aid m troubleshootmg, we keep a data sheet for each lot of gels cast On the data sheet we record. calibration values for pumps, dates stock soluttons were made, room temperature, volumes remaining m the high- and low-hmtt solutton containers, and polymerized gel length Variation from normal m either of the latter two measures signals a problem has occurred m making the gradient In the event gels do not exhibit expected properties, then mformation from thts data sheet can be used to help Identify the source of the problem and the gel lots that might be affected. 5. A wide range of sample volumes may be loaded into these gels. One group has reported loading nearly 30 uL of plasma (m three sequential aliquots) per sample lane (2) Assuming more than 65 g/L of plasma protems, this load volume would contam more than 2 mg proteins Of course, the lower hmtts of load volumes are dictated primarily by stammg efficiency for the lipoproteins of interest, for example, apohpoprotem dtstrlbutlons may be detected m submtcrohter volumes of plasma using immunoblottmg and autoradtography 6 The method described here for SBB staming uses the solvent Cellosolve (ethylene glycol monoethyl ether), which is a hazardous chemical Personnel should be properly protected; this includes wearing gloves and a full-face respirator mask, and working in a chemical fume hood with adequate air-flow. 7. Gel regions with higher acrylamtde concentrations tend to crack as they dry out during the staining process To avoid thts, apply to the gel surface a plastic mask that covers the edges of the gel and bathe the gel edges m a puddle of water 8. Densitometer performance should be monitored on a routme basis (e.g., monthly). For an LKB UltroScan XL Laser Densitometer, we monitor performance using a Densitometer Test Plate (Pharmacia) and a Step Tablet No 2 (Kodak, Rochester, NY). The Test Plate has grids that test beam resolution and a gel image that can be used to test that the laser beam exactly traverses the area indicated by the ruler. The step Tablet has steps of known density that are used to identify the range of absorbances for which the densitometer has acceptable accuracy. Comparing current performance with records of previous assessments will help tdentify fairly subtle deviations before they can affect large amounts of data 9 Records, such as printouts and computer files containing absorbance profiles, should be kept m case data need to be re-evaluated m the future In addttton, tt often proves useful to keep records of the gel as well. The gel itself 1s available for comparisons with absorbance profiles for many years tf sealed with sodmm aztde-containing buffer m plastic bags and refrigerated. Also, a photographic record of the gel can be made using Polaroid 667 film with a yellow filter

Rainwater

150

Acknowledgments The author wishes to acknowledge important contributions made by Perry H. Moore, Jr., David W. Andres, Amareshwar T. K. Singh, Mahmood Poushesh, Wendy R. Shelledy, Allen L. Ford, and Tom D. Dyer to the development of these methods. This work was supported m part by NIH grants HL-28972, HL-45522, and RR-09919. References 1 Rainwater, D L , Andres, D. W., Ford, A L , Lowe, W. F , Blanche, P J , and Krauss, R. M. (1992) Production of polyacrylamide gradient gels for the electrophoretic resolutton of lipoproteins. J Lzpzd Res 33, 1876188 1. 2 Li, Z., McNamara, J R , Ordovas, J. M., and Schaefer, E. J. (1994) Analysis of high density hpoprotems by a modified gradtent gel electrophoresis method J Lzpzd Res 35, 1698-1711 3. Singh, A T K , Rainwater, D. L , Haffner, S M , VandeBerg, J. L , Shelledy, W R., Moore, P. H , Jr , and Dyer, T D. (1995) Effect of diabetes on hpoprotem size Arterioscler Thromb. Vusc BzoZ 15, 1805-1811 4 SchJeide, 0 A , Rivm, A U , and Yoshmo, J (1963) Uptake of hpid stains by lipids and serum lipoproteins. Am J Clan Path01 39,329341 5 Callats, F , Roche, D., and Andreux, J. P. (1987) Value of polyacrylamide gradient gel electrophoresis of lipoproteins for determmmg HDL cholesterol C2zn Chem 33,1266. 6. Gambert, P , Farmer, M , Bouzerand, C., Athras, A , and Lallemant, C. (1988) Direct quantitation of serum high density hpoprotem subfractions separated by gradient gel electrophoresis Clzn Chum Actu 172, 183-190 7. Cheng, M.-L., Kammerer, C. M., Lowe, W. F., Dyke, B , and VandeBerg, J. L. (1988) Method for quantitatmg cholesterol m subfractions of serum hpoprotems separated by gradient gel electrophoresis. Bzochem Genet 26,657-68 1. 8 McNamara, J R., Campos, H , Ordovas, J. M , Peterson, J., Wilson, P W F , and Schaefer, E. J (1987) Effect of gender, age, and lipid status on low density hpoprotem subfraction distribution ArterzoscZeroszs 7,483+90 9. Rainwater, D. L., Kammerer, C. M , Cheng, M -L , Sparks, M. L., and VandeBerg, J. L. (1992) Distribution of specific apohpoprotems detected by immunoblottmg of baboon hpoprotems resolved by polyacrylamide gradtent gel electrophoresis Blochem Genet 30, 143-158. 10. Rainwater, D. L., Blangero, J , Moore, P. H., Jr., Shelledy, W R., and Dyer, T. D (1995) Genetic control of apolipoprotem A-I distribution among HDL subclasses Atheroscleroszs 118,307-3 17. 11. Greenwood, F. C., Hunter, W. M., and Glover, J. S (1963) The preparation of i3’I-labelled human growth hormone of high specific radioactivity Bzochem J 89,114--123

12. Montelaro, R. C. and Salinovich, 0 (1986) Reversible staining and peptide mappmg of protems transferred to mtrocellulose after separation by sodium dodecyl sulfate-polyacrylamide gel electrophoresis Anal Biochem 156, 34 l-347

Separation of LDL and HDL Subclasses

151

13 Blanche, P. J , Gong, E. L., Forte, T. M., and Nichols, A. V (1981) Charactertzatton of human htgh-density hpoprotems by gradient gel electrophorests Bzochlm Bzophys Acta 665,408-4 19 14 Krauss, R. M and Burke, D J (1982) Identificatton of multiple subclasses of plasma low density hpoproteins in normal humans J Lipid Res 23,97-104 15. Lefevre, M. (1988) Localization of lipoprotein unesterified cholesterol in nondenaturmg gradient gels with tilipin. J Llpld Res 29, 8 15-8 18 16. Wtlhams, P. T , Krauss, R. M., Nichols, A. V., Vranizan, K. M., and Wood, P. D S (1990) Identifymg the predominant peak diameter of high-denstty and low-denstty lipoproteins by electrophoresis. J. Lzpld Res 31, 113 l-l 139 17. Krauss, R. M. and Blanche, P. J. (1992) Detection and quantitatton of LDL subfractions Curr Open Llpldol 3,377-383. 18 Austin, M A , Brunzell, J D , Fitch, W. L , and Krauss, R M (1990) Inheritance of low density lipoprotem subclass patterns m famihal combined hyperliptdemia Arterzoscleroszs

10, 520-530.

19 Austin, M A., Pan-itz Jarvik, G., Hokanson, J E , and Edwards, K (1993) Complex segregation analysis of LDL peak particle diameter. Genet Eprdemlol 10, 599-604

20. Campos, H., BliJlevens, E., McNamara, J. R , Ordovas, J M , Posner, B. M , Wilson, P W F , Castelh, W P., and Schaefer, E. J. (1992) LDL particle size distributton: results from the Frammgham Offspring Study. Arterzoscler Thromb 12, 141&1419 2 1. Wtlhams, P. T , Vramzan, K M , and Krauss, R. M. (1992) Correlations of plasma hpoprotems with LDL subfractions by particle size m men and women J Lzpzd Res. 33, 765-774.

22. Verdery, R. B., Benham, D. F , Baldwin, H. L., Goldberg, A P , and Nichols, A V (1989) Measurement of normative HDL subfractton cholesterol levels by Gaussian summation analysis of gradient gels J, Lipzd Res 30, 1085-1095 23. Williams, P T , Krauss, R. M , Vranizan, K. M , Stefanick, M L , Wood, P D. S , and Lmdgren, F. T (1992) Associations of hpoprotems and apohpoprotems with gradient gel electrophorests estimates of high density hpoprotem subfractions m men and women Arterzoscler Thromb 12,332-340.

11 Measurement Catherine

of Lipoprotein

Fievet and Jean-Charles

Particles Fruchart

1. Introduction Lrpoprotems have been traditionally classified based on their hydrated densities or electrophoretic mobilities. In the 1960s Alaupovic (1) proposed that plasma lipoprotems could be viewed as a system of particles characterized by their apolipoprotein composition. This concept evolved m the proposal of using “apolipoprotem profiles” for typing dyslipoprotememias. Along these lines, several recent reports have shown that certain apohpoprotem profiles may be better predictors of lipoprotem disorders than the traditional lipid profiles (2). During the last decade, immunological methods have taken advantage of the specificity provided by polyclonal or monoclonal antibodies (PAbs, MAbs) to identify, separate, and measure specific lipoprotem particles directly m plasma (3). The apphcation of these techniques to clinical and epidemiological studies has shown that these apohpoprotein-specific subclasses may be important diagnostic tools for dyslipoproteinemia, as well as useful predictors for coronary-artery disease (2,46). Moreover, they may provide useful clues about the chemical nature of the corresponding lipoprotem particles Various procedures have been described for the measurement of apolipoproteins such as radial immunodiffusion, electroimmunoassays, radioimmunoassays, enzyme immunoassays, mununoturbidimetric methods, and immunonephelometry ((i-11), each having its own advantages and limitations (74. Some of them can be performed directly on total plasma For others, a previous separation of apo B- or apo A-containing lipoproteins might be needed. For many years, madequate standardization of the methodological approaches, absence of reference methods, and lack of normal intervals have hampered the climcal use of apolipoprotem determmations. This problem has been circumFrom

Methods Edtted

in Molecular

by J M Ordovas

Bology, 0 Humana

153

Vol 110 Llpoprotem Press

Inc , Totowa.

Protocols NJ

154

Fievet and Fruchart

vented for the apohpoprotems (apo) A-I and B by usmg a certified International Reference Material (9,1/J). The methods described here allow reproducible apohpoprotem measurements, and they are designed for easy integration into the routine workload of specialized research and clmlcal laboratories. However, it should be noted that, owing to the lack of standardization, the results obtained by one laboratory may not be directly compared with those obtained by other laboratories using different methodology and standards. More recently, the development of highly specific MAbs has added an even greater specificity, and It 1s now possible to separate particles exhrbttmg the epltope at their surface from particles where epltope IS masked or altered m

conformation. 2. Materials

2.1. Analysis of Apolipoproteins Within Specific Lipoprotein Fractions

Using PAbs

1 2 3 4.

Hydragel LpC-III (Laboratolre Stbla, 92130 Issy-les-Moulmeaux, France) Hydragel LpE (Laboratolre Sebla, 92130) Hydragel LpA-I particles (Laboratolre SCbla, 92 130) Electrophoretlc power supply (Laboratolre SCbla, GD 6 1D or GD 25 1 D) (see Note 1) 5. Horizontal electrophoresis tank (Laboratolre SCbla, S60 or K20) (see Note 2) 6 Incubator Drier (Laboratoire Stbla, IS 80)

2.2. Measurement of Lipoprotein Particles Using Two-Site Differential lmmunoenzymometric

Assays

1, Flexible PVC 96-well mlcrotlter plates (Costar, Cambndge, MA) 2 Automatic enzyme-llmlted lmmunosorbent assay (ELISA) Processor (Behrmg Institute, Marburg, Germany) 3. Electronic dilutor (Dllutrend, Boehrmger-Mannhelm, Mannhelm, Germany) 4 Polyclonal anti-apo A-I 5 Polyclonal anti-apo A-II. 6. Peroxldase-labeled antibodies. 7 Coating and washing buffers* Na phosphate-buffered salme (phosphate, 10 mM, NaCl, 150 mM, pH 7 4) (PBS) 8. PBS plus bovine serum albumin (1%, w/v) (PBS-BSA) 9. Peroxldase substrate solution 3 g of o-phenylenediamme-dlhydrochlorlde (Sigma, St-Louis, MO)/L of 0 1 A4 phosphate citrate buffer, pH 5 5, contammg hydrogen peroxide (3.5 mM)

2.3. Methods Based on the Use of MAbs 1. Flexible PVC 96-well microtiter plates (Costar) 2 Automatic ELISA Processor (Behrmg Institute)

Measurement 3 4. 5 6. 7 8

of Lipoprotein Partxles

155

Electrontc dtlutor (Dtlutrend, Boehrmger-Mannhelm) MAbs. BL3, BLS, BL7 Goat polyclonal affimty-purified antibodies peroxldase conjugated Coattng and washing buffers (PBS): 10 tiphosphate, 150 mMNaC1, pH 7.4 PBS-BSA* PBS contamtng 1% BSA Peroxtdase substrate solution. 3 g of o-phenylenedtamme, dthydrochlorlde (Sigma)/L of 0.1 Mphosphate citrate buffer, pH 5.5, contammg hydrogen peroxide (3 5 mM)

3. Methods 3.1. Analysis of Apo E and Apo C-Ill Within Specific Lipoprotein Fracfions

Using PAbs

Sequential immunoprectpitation of specific lipoprotein particles in free solution, using consecuttve antibody treatments was mtttally described to separate and measure specific subpopulattons of particles wtthm very low and low

density lipoproteins (VLDLs and LDLs) (12). This approach was also used as early as 1982 to evaluate apo A-I-containing out apo A-II (13)

ltpoprotem

particles wtth or wtth-

A modification of this procedure consists m preparation of immunoaffinity columns containing

antibodies

linked to a solid phase. The quantitattve

deter-

mination of apohpoprotem 1s then performed m retained and unretamed column fractions, both fractions being recovered m soluble forms, therefore allowmg a direct measurement of hpoprotem particles (14). Although this approach has been used for years as reference method to measure hpoprotem particles, this technique 1snot practical for routme measurement m a clmtcal laboratory. This 1sa very time-consuming procedure, difficult to standardize, and not easily amenable to automation. Another shortcommg of this approach relates to the high cost associated with the large amounts of antibodies needed In this section, we describe the use of a combmation of tmmunoprectpttatron m free solutton and electrotmmunodtffusion to measure the distrtbutton of apo C-III and apo E among apo B- and non-apo B-containing hpoprotems.

Lipoproteins containing apo B (LpB) (chylomicrons, VLDL, LDL) and hpoproteins without apo B (Lp non-B) (high-density lipoproteins) (HDL) seem to exert opposite effects as risk factors for coronary heart disease. Apo Cs and

apo E are components of both classesof lipoproteins, and several studies have shown that their distribution among them could be of great relevance in determining their metabolic fate and thus their plasma concentrations (15-20). Several simple procedures have been developed for the separation of LpB and Lp non B. Procedures based on the use of polyamomc reagents have been used as acceptable alternatives to the classical ultracentrifugal fractionation. These include heparm-Mn*+ preclpitatton of hpoprotems (21), phosphotungstate-

Fievet and Fruchart

156

Mg2+ (22), a mixture of polyethyleneglycol and dextran and MgC12 (231, or concanavalin A (24). They are cheap and simple to carry out; however, they suffer from lack of specificity. In contrast, the unrnunoprecipitation by antiapo B antibodies seemsto be the optimal procedure for separating lipoprotein particles based on the presence or absence of apo B of a specific apolipoprotem composition. The procedures descrtbed below are straightforward and all reagents can be obtained as part of commercial kits (Hydragel LpC-III, LpE). These measurements have been recently applied to assessCHD risk m a large population study (25). Each kit includes 10 agarose gel foils (16-wells, of which 14 can be used) contammg polyclonal antiserum to apo C-III or apo E, plus buffer concentrate, standard plasma (calibrated for total apo C-III and apo E), anti-apo B tmmunoglobulms,

and acid violet stain and destaining

m concentrated soluttons

(see Note 3). 3.1.1. lmmunoseparatlon of Apo B and non-apo B-Containing Lipoproteins This step is common for both apo E and apo C-III measurements. 1 One volume of whole plasma 1s mixed with one volume of a ready-to-use solution of apo B mrmunoglobulms 2 The mtxture IS incubated at room temperature for 5 mm and the resultmg tmmunoprecipitate 1sseparated by centrifugation at 3000g for 10 min 3 The supematant is recovered for measurement of apo C-III and apo E m the Lp non-B fraction,

3.12. Electroimmunodiffusion to Measure Apo E in Whole Plasma and in the Lp non-B Fraction 1, For the measurement of total apo E, plasma is diluted sevenfold m isotonic saline 2 For the measurement of apo E m the Lp non-B fractton (obtained as described above), the supemate 1sused without any further ddution. 3 Four calibration points are prepared in isotonic salme by diluting a lyophthzed pool of human plasma (provided with the kit) that serves as a secondary standard This pool has been calibrated for apo E contents using purified apo E. 4. A quality-control sample (lyophdlzed plasma-calibrated to apo E) is analyzed as a whole plasma in each series to ensure the reproducibdlty of the assay and to correct for inter assay variation 5 Five microliters of each calibratton point, diluted whole and control plasma, and Lp non-B supernates are applied into the wells of the agarose gel provided within the Hydragel apo E kit. 6 The electrophoresis is carried for 4 h at a constant voltage of 50 V. 7. After migration, the proteins not reacting with the antibody are absorbed by placing over the gels one thm filter paper previously soaked with isotomc

Measurement

of Lipoprotein Particles

157

salme and two thtck filter papers, all held m place under a pressure of 1 kg for 20 min. 8 Thereafter, the gels are washed in isotonic saline for 60 min and pressed again for 10 mm. 9. The plates are then dried and stained. 10 After destaining, the heights of the rockets are measured, and calibration curves are drawn with the concentrations of apo E of the different dilutions of the standard on the abscissa and the hetghts of the rockets on the ordinates. Samples and control values of apo E are read from each respectrve standard curve The difference between the apo E concentrations obtained for whole (total apo E) and Lp non-B supernate (apo E-Lp non-B) plasma corresponds to apo E*LpB values Figure 1 shows typical rockets obtained by this procedure (see Notes 4 and 5).

3.1.3. Electroimmunodiffusion to Measure Apo C-Ill in Whole Plasma and in the Lp non-B Fraction 1 For the measurement of total C-III, plasma is diluted 15-fold in isotonic salme 2 For the measurement of apo C-III m the Lp non-B fraction (obtained as described above) the supernate 1s used wtthout any further dilution. 3 Four calibration pomts are prepared in isotonic saline by diluting a lyophthzed pool of human plasma (provided with the kit) that serves as a secondary standard This pool has been calibrated for apo C-III contents usmg purttied apo C-III. 4 A quality-control sample (lyophthzed plasma-cahbrated to apo C-III) 1sanalyzed as a whole plasma m each series to ensure the reproductbthty of the assay and to correct for interassay variatton. 5. Five mtcrohters of each calibratton pomt, diluted whole and control plasma, and Lp non-B supernates are apphed mto the wells of the agarose gel provided wtthm the Hydragel apo C-III kit 6 The electrophoresis is carrted for 4 h at a constant voltage of 50 V 7. After migratton, the proteins not reacting with the anttbody are absorbed by placing over the gels one thin filter paper previously soaked with tsotomc saline and two thick filter papers, all held m place under a pressure of 1 kg for 20 min 8 Thereafter, the gels are washed in isotonic saline for 60 mm and pressed agam for 10 min. 9. The plates are then dried and stamed. 10. After destaining, the hetghts of the rockets are measured. The calibratton curves are drawn with the concentrations of apo C-III of the different dilutions of the standard on the abscissa and the heights of the rockets on the ordinates. Samples and control values of apo C-III are read from each respective standard curve. The difference between the apo C-III concentrations obtained for whole (total apo C-III) and Lp non-B supernate (apo C-III-Lp non-B) plasma corresponds to apo C-1II:LpB values. Figure 2 shows typical rockets obtamed by this procedure (see Notes 5 and 6).

Fig. 1. Typical rockets obtained by electroimmunoassay

Fig. 2. Typical rockets obtained by electroimmunoassay 158

on Hydragel Lp-C-III.

on Hydragel LpE.

Measurement of Lipoprotein Particles

159

3.2. Measurement of LpA-1 and LpA-I/A-II Particles in Who/e Plasma The application of crossed-immunoelectroassay to plasma apohpoprotems has allowed further characterization of various lipoprotein species according to their apohpoprotem content (26). Through this procedure, Atmeh et al. (27) evaluated lipoprotein particles which contain apo A-I, but not apo A-II (LpA-I). This method employed electrophoresis through two agarose gel layers, each containing different antibodies. The first layer, containing PAbs against apo A-II, retained those lipoprotems containing apo A-II, whereas those containing apo A-I without apo A-II migrated into the upper layer, which was embedded with a PAb against apo A-I. As a routme assay for the direct measurement of LpA-I particles in plasma, we are usmg a simpler procedure developed m collaboration with Sebia (28). This method utilizes ready-to-use agarose plates containing both anti apo A-I and anti-apo A-II. Anti-apo A-II is incorporated m high excess,resultmg m the retention of LpA-I/A-II particles within an immunoprecipitation peak close to the origin. LpA-I migrates further mto the gel and produces a second tmmunoprecipitation

peak The height of this peak is used to determine

the LpA-I con-

centration in the sample. Each Hydragel LpA-I agarose gel contains 16-wells (of which 14 can be used). In addition, each kit contains buffer concentrate, lyophtlized standard

serum (calibrated for LpA-I), and acid violet stain and destammg in concentrated solutions. 1 Plasma samples should be diluted 1: 100 with saline (10 pL of sample plus 990 pL of salme) 2. Four calibration points are prepared in isotonic saline by diluting the standard serum provided as suggested m the kit’s instructions 3 A qualtty-control sample (lyophilized serum calibrated for LpA-I) is analyzed m each serves m a similar fashion to the plasma samples m order to control for interassay variations 4. Five microliters of each standard calibration point, dtluted control serum, and samples are applied mto the wells of a gel foil and then run for 4 h at a constant voltage of 50 V 5 After migration and washing in isotonic salme, gels are dried as described above for LpC-III and LpE determinations, and stained 6. Two rockets corresponding to LpA-I/A-II particles (lower peak) and to LpA-I particles (less intensively colored higher peak) are visible for each sample (Fig. 3) The height of the rockets 1smeasured, and calrbratton curves are drawn with the concentrations of LpA-I of the different dilutions of the standard on the abscissa and the heights of the rockets on the ordinates Samples and control values of LpA-I are read from standard curve (see Notes 7 and 8).

160

Fievet and Fruchart

Fig. 3. Separation of apo A-I containing lipoproteins (LpA-I and LpA-I/A-II) electroimmunoassay (Hydragel LpA-I).

by

3.3. Measurement of Lipoprotein Particles Using Two-Site Differential lmmunoenzymometric Assays Two-site differential immunoenzymometric assayshave been proposed for the direct measurement in plasma of different apolipoprotein associations (3). The principle of these assaysis shown on Fig. 4. Antigens (plasma) react with immobilized antibodies (Abl). The solid phase is washed to remove unbound antigens. An excess of peroxidase-labeled antibodies (Ab2-POD) are allowed to react with bound antigens. The solid phase is washed again to remove unbound labeled antibodies. In the enzymatic indicator reaction, solid phase-bound enzyme activity is measured. Color development is directly related to the amount of antigen in the sample. Utilizing this general scheme, it is possible to measure lipoprotein particles that contain those apolipoproteins specifically recognized by antibodies Abl and Ab2. (i.e., in the determination of LpA-I/A-II particles, Abl will be anti-apo A-I and Ab2 will be anti-apo A-II). Our laboratory routinely measures lipoprotein particles containing at least A-I and A-II (LpA-I/A-II) (29), C-III and B (LpC-III/B) (301, apo E and B (LpE/B) (301, and (a) and B (Lp(a)/B) (31). Other combina-

Measurement

Solid phase-bound antlapolipoprotein-Ab,

of Lipoprotein

Partdes

161

POD-AbZ conjugate

Soltd phase-bound Ab,Ag complex

Ilgand, antigen, analyte, apolipoprotein

POD-labelled

sandwich

complex

1 2 1,2-Phenylenediamine

+ 2Hz02

-

2,2’-diamino-azobenzene

Fig. 4. Two-site differential tmmunoenzymometric

+ 4 HZ0

assay for hpoprotem particles

tlons including those hpoprotem particles that contam apo C-II, apo C-III, or apo E among apo B- or apo A-I containing lipoprotems have similarly been described (19). 1. As a first step, the sohd phase (generally a 96-well microtiter plate) 1s coated with affinity-purified antibody Abl (or immunoglobulin fraction if the antiserum has a sufficient titre) (500 ng per well) and incubated overnight at 4’C 2. After three successive washings, diluted samples and standards are added into the wells (see Note 9). 3 The plate is covered and then incubated for 2 h at 37°C. 4 After washing three times and aspirating the wells dry, the diluted conjugate (peroxidase-labeledAb2) is plpetted into eachwell and the plate ISagain covered and incubated for 2 h at 37OC 5. Finally, the substrate solution is added mto the wells and coloration is allowed to develop at room temperature for 30 mm m the dark. 6 The reaction 1sthereafter stoppedby adding 1 M HCl. 7 The absorbance at 492 nm 1srecorded In each series, a blank test without antigen is run under the same conditions than the samples

162

Fievet and Fruchart

The optimal quantities and concentrations of sample and reagents are dependent on the characteristics of the antibodies used for the test. The working range is also dependent on the assay,but the general procedure described here is applicable to any lipoprotein particle measurement. The sensitivity of such methodology is very high, and in our hands, the lowest quantity that can be measured in 100 p.L of sample applied per well is 1 ng of LpC-III/B and 2 ng of LpE/B. This high sensitivity also represents the greatest source of error becausevery high sample dilutions are used. Nevertheless, when utilizing robotic equipment to dilute the samples, the variation of the testsis less than 10%. In summary, these immunoenzymometric methods are fast, specific, and sensitive to measure lipoprotein complexes in plasma without previous lipoprotein separation. 3.4. Methods Based on the Use of MAbs The development of techniques to produce highly specific antibodies (MAbs generated using purified apolipoproteins or using synthesized short fragments of an apolipoprotein) may offer the possibility of separating lipoprotein particles with an epitope expressed on the surface from particles in which the epitope is masked or altered in conformation. Depending on the molecular structure of lipoproteins and the apolipoprotein conformation on their surface, it is therefore possible to separate discrete particles according to an immunological criterion. Their measurement is performed by an immunoenzymometric assay as described above, in which the immobilized antibody is a MAb. The peroxidase-labeled antibody may be obtained from a polyclonal antiserum, becausethe first step leads to the immobilization of lipoprotein particles, which present the relative epitopes on their surface. Using this methodological approach (32), we evaluated the potential correlation between specific epitopes on apo B-containing lipoprotein subpopulations and predisposition to coronary-artery disease. Three lipoprotein particles were specifically measured in normal subjects and patients using three distinct MAbs. The immunological accessibility was assessedthrough the separate apo B concentrations. Thus, we obtained a more-accurate assessment of coronaryartery disease risk as obtained with the measurement of total apo B levels (33). 1. Coat each well of a flexible PVC plate with 0.1 mL (2 pg protein) of a MAb (BL3, BL5, or BL7) diluted in 0.1 M phosphate buffer, pH 7.2. 2. Incubate overnight at 4°C. 3. Wash three times with PBS and aspirate the wells dry. 4. Add 100 pL of samples and standards diluted in PBS into the wells (see Note 10). 5. Cover the plate with a plastic sealer and incubate for 2 h at 37°C. 6. Wash three times with PBS and aspirate the wells dry. 7. Add 0.1 mL of conjugate (diluted 15,000-fold in BSA-PBS) to each well.

Measurement of Lipoprotein Particles

163

8. Cover the plate with a plastic wrap and incubate at 37°C for exactly 2 h. 9. Repeat the washing steps as indicated in step 6. 10. Incubate each well with 100 pL of a freshly prepared substrate solution Finally, the substrate solution is added into the wells and allow the coloration to develop at room temperature for 30 min in the dark. 11. The reaction is thereafter stopped by adding 1 M HCl. 12. The absorbance at 492 nm is recorded. In each series, a blank test without antigen is run under the same conditions than the samples. 4. Notes 1, Any low-voltage power supply is adequate to run this type of electrophoresis. 2. As an alternative to the electrophoresis tanks distributed by Sebia, one may use the horizontal electrophoresis apparatus traditionally used to run DNA. These are available from several manufacturers. 3. Detailed description regarding buffer compositions and dilutions are included with the kit instructions. 4. The working range for the assay is from 2.0 to 12.0 mg/dL. The assay is precise and the coefficient of variation is less than 10%. 5. Analyses must be performed on fresh samples because a systematic bias has occurred when the samples have been frozen. 6. The working range for the assay is from 1.3 to 7.8 mg/dL. The assay is precise and the coefficient of variation is less than 10%. 7. The described method can reliably quantify LpA-I at 10 mg/dL. Its reproducibility is good, reaching a maximal between-run coefficient of variation of 4.56%. 8. It is recommended to carry out analyses on pure fresh samples or stored at 4°C for no longer than 1 wk. Nevertheless, our own experience has shown that samples kept at -20°C for 4 mo gave very similar results to fresh ones. 9. We perform coating, washing, addition of conjugate, substrate, and HCl, and spectrophotometer reading by use of an automatic ELISA Processor (Behring Institute). Samples and standard are diluted with an electronic dilutor (Dilutrend, Boehringer-Mannheim). The assay buffer (coating and washing) is Na PBS (phosphate, 0.1 M, NaCl, 0.137 IV, pH 7.4). In order to minimize nonspecific binding to microtiter wells, BSA (I%, w/v) is added into the assay buffer for dilution of antigen and conjugate. The peroxidase-labeled Ab2 is prepared with polyclonal affinity-purified antibodies coupled with the enzyme, following previously described detailed protocol (34). For the peroxidase substrate solution, we use 3 g of o-phenylenediamine, dihydrochloride (Sigma)/L of 0.1 M phosphate citrate buffer, pH 5.5, containing hydrogen peroxide (3.5 mM). 10. The apoB content of LDL used as primary standard is estimated by a modified Lowry protein assay (35) before and after precipitation of apo B with isopropanol (36). For everyday assays, a secondary standard is prepared by pooling frozen normal sera and adding antibiotics, protease inhibitors, and antioxidants. This cocktail contains: 0.27 mMEDTA, 0.9 mMcis-amino-caproic acid, 0.6 mA4 chloramphenicol, and 0.3 mM gluthatione.

164

Fievet and Fruchart

References 1 Alaupovic, P. (1968) Recent advances in metabolism of plasma lipoprotems chemical aspects. Prog. Biochem Pharmacol 4,9 l-l 09. 2 Gotto, A. M. and Levy, R I (1983) Dyslipoproteinemias characterized by absent, excess, or aberrant apohpoprotems, m Proceedzngs of the Workshop on Apollpoprotezn Quantzjcation (Lippel, K., ed.), NIH Pubhcatton No. 83-1266, Chevy Chase, MD, pp 25147 3. Fruchart, J C (1990) Lipoprotein heterogeneity and its effect on apolipoprotein assay &and J Urn. Lab Invest 50,51-57. 4 Fievet, C., MeJean, L., Drouin, P., and Fruchart, J C (1988) Molecular analysis of atherogenic lipoprotein particles and atherosclerosis, in Ezcosanozds, Apolrpoproteins, Lipoprotew Partxles and Atherosclerosrs (Malmendier, C L and Alaupovic, P , eds.), Plenum, New York, pp. 279-282. 5 Parra, H J , Cachera, C., Equagoo, K , Dracon, M., Fruchart, J C , and Tacquet, A. (1988) Quantitative abnormalittes of lipoprotein particles in chrome hemodialysis patients, in Elcosanozds, Apollpoprotelns, Lipoprotein Particles and Atherosclerosis (Malmendier, C L , Alaupovm, P., eds.), Plenum, London, pp 283-287 6. Parra, H J , Arveiler, D., Evans, A. E , Cambou, J P., Amouyel, P., Bmgham, A , McMaster, D , Schaffer, P., Douste-Blazy, P., Luc, G., Richard, J. L., Ducimetibre, P , Fruchart, J C., and Cambien, F. (1992) A case control study of lipoprotem particles in two populations at contrasting risk for coronary heart disease the ECTIM study. Arterloscler Thromb 12,701-707. 7. Steinberg, K K , Cooper, G. R., Grasier, S R., and Rosseneu, M. (1983) Some considerations of methodology and standardization of apolipoprotem A-I unmunoassays Clan Chem 29,415-426. 8 Rosseneu, R , Vercaemst, R. Steinberg, K. K , and Cooper, G R (1983) Some considerations of methodology and standardization of apohpoprotem B unmunoassays. Clm Chem. 29,427433. 9. Marcovma, S. M., Albers, J J., Dati, F., Ledue T. B., and Ritchie, R. F. (1991) Internatronal Federation of Clinical Chemistry Standardtzatton ProJect for measurements of apohpoprotems A-I and B Clzn Chem 37, 16761682. 10. Marcovina, S. M., Albers, J. J., Henderson, L 0 , and Hannon W. H (1993) International Federation of Clinical Chemistry Standardization ProJect for Measurements of Apohpoprotems A-I and B. III* comparability of apolipoprotein A-I values by use of international reference material CEzn Chem 39, 773-781. 11. Alaupovic, P., McConathy, W. J., Currry, M. D., and Fesmire, J D (1982) Characterization of dyshpoproteinemias by apolipoprotein profiles, in Lzpoproterns and Coronary Atheroscleroszs (Noseda, G., Fragiacomo, C , Fumagalh, R , and Paoletti, R , eds.), Elsevier, Amsterdam, pp. 135-144 12. Alaupovic, P., Tavella, M., and Fesmue, J (1987) Separation and identification of apo B-containing hpoprotem particles m normohptdemic subjects and patients with hyperlipoproteinemias. Adv Exp Med. Blol 210,7-14.

Measurement

of Lipoprotein

Particles

165

13. Cheung, M. C and Albers, J J (1982) Distribution of high density lipoprotem particles with different apoprotein composition. particles with A-I and A-II and particles with A-I but no A-II. J Lipid Res 23,747-753 14 Bekaert, E. D., Alaupovlc, P., Knight-Gibson, C., Norum R. A., Laux, M J., and Ayrault-Jarrter, M (1992) Isolation and partial characterization of lipoprotein A-II (LpA-II) particles of human plasma Biochim Biophys Acta 1126, 105-l 13 15. Alaupovic, P (1981) David Rubmstein Memorial Lecture: the biochemical and chmcal significance of the interrelattonship between very low denstty and high density lipoprotems Can J Blochem. 59,565-579. 16. Parsy, D , Dracon, M , Cachera, C., Parra, H J , Vanhoutte, G , Tacquet, A., and Fruchart, J. C (1988) Lipoprotein abnormalities m chronic hemodialysis patients. Nephrol Dial Transplant 3,5 l-56. 17 Blankenhorn, D H , Alaupovic, P., Wickam, E , Chm, H P., and Azen, S P. (1990) Prediction of anglographic change in native human coronary arteries and aortocoronary bypass grafts. Cuwlatlon 81,470-476. 18. Chivot, L , Mainard, F , Bigot, E., Bard, J M , Auget, J. L , Madec Y , and Fruchart, J. C. (1990) Logisttc dlscrimmant analysts of lipids and apolipoprotems in a population of coronary bypass patients and the significance of apohpoprotems C-III and E. Atheroscleroszs 82, 205-2 11. 19 Alsayed, N and Rebourcet, R (1991) Abnormal concentrations of C-II, C-III, and E apolipoprotems among apohpoprotem B-containing, B-free, and A-I-containing hpoprotem particles m hemodialysis patients. Clin. Chem 37, 387-393 20. Genes& J. J., Bard, J. M., Fruchart, J. C., Ordovas, J M., Wilson, P F. N., and Schaefer, E J (1991) Plasma apohpoprotem A-I, A-II, B, E and C-III containing particles m men with premature coronary artery disease. Atheroscleroszs 90, 149-157. 2 1. Warmck, G. R. and Albers, J J (1978) A comprehensive evaluation of the heparm- manganese precipitation procedure for estimating high density hpoprotern cholesterol. J Llprd Res 19,65-76. 22 Puchois, P., Luley, C , and Alaupovic, P. (1987) Comparison of four procedures for separating apohpoprotein A- and apohpoprotein B-contammg lipoproteins m plasma Clin Chem 33, 1597-1602. 23 Graziam, M. S., Zanolla, L., Righettt, G., Marchetti, C , Zanotto, G., and Lupo, A (1994) Distribution of C-II and C-III pepttdes in hpoprotem classes. methods and clinical significance Clw Chem 40, 240-244. 24. McConathy, W. J. and Alaupovlc, P. (1974) Studies on the interaction of concanavalm A with major density classes of human plasma lipoproteins. Evidence for the specific binding of hpoprotem B m its associated and free forms. Febs Letters 41, 174-178 25. Luc, G , F&et, C , Arveiler, D., Evans, A E., Bard, J. M., Cambien, F., Fruchart, J C., and Duclmetiere, P (1996) Apolipoproteins C-III and E in apo B- and nonapo B-containing lipoproteins m two populations at contrasting risk for myocardial infarction: the ECTIM study. J. Lipid Res 37,508-5 17. 26. McConathy, W. J., Alaupovlc, P., and Fesmire, J. C. (1983) Various applications of electroimmunoassay for plasma apolipoprotein, in Proceedings of the Workshop

Fievet and Fruchart

27.

28.

29.

30

31

32.

33

34 35

36

on Apolzpoprotem Quantzjkatron (Lippel, K., ed ), NIH Publication No. 83-1266, Chevy Chase, MD, pp. 276- 28 17. Atmeh, R F , Shepherd, J , and Packard, C J. (1983) Subpopulations of apohpoprotein A-I in human htgh-denstty lipoprotems Their metabolic profiles and response to drug therapy Btochlm. Biophys Acta 751,175-l 88. Parra, H. J., Mezdour, H., Ghalim, N., Bard, J. M , and Fruchart, J C. (1990) Differential electroimmunoassay of human LpA-I lipoprotein particles on readyto-use plates. Clan Chem 36, 143 l-1435. Koren, E., Puchots, P , Alaupovic, P., Fesmue, J , Kandousst, A , and Fruchart, J C (1987) Quantification of two types of apolipoprotein A-I contammg hpoprotem parttcles m plasma by enzyme-linked differential antibody immunosorbent assay Clin Chem 33,38-43 Kandoussi, A., Cachera, C., Parsy, D , Bard, J M , and Fruchart, J. C (1991) Quantitative determination of different apoltpoprotem B containing hpoprotems by an enzyme lmked tmmunosorbent assay: apo B with apo C-III and apo B with apo E J Immunoassay 12,305-323 Vu-Dac, N., Mezdour, H., Parra, H J , Luc, G., Luyeye, I, and Fruchart, J C (1989) A selective bt-site mununoenzymattc procedure for human Lp(a) hpoprotem quanttficatton using monoclonal annbodies against apo (a) and apo B. J Lzpzd Res. 30, 1437-1444 Luyeye, I., Fievet, C., DuPont, J. C , Durteux, C , Shmane, N., Lecocq, J. F , Demarqmlly, C., and Fruchart J.C. (1988) Human apohpoprotein B . evidence for its tmmunochemtcal heterogeneity using monoclonal anttbodies and an tmmunoenzymometrtc assay. Cltn Biochem. 21,255-261 F&et, C., Nuttens, M. C , Ducimetitre, P., Fruchart, J C , Bertrand, M., and Salomez, J. L (1991) Relation of arteriographically defined coronary artery dtsease to serum hpoprotem pattcles mapped with monoclonal antibodtes. Czrculatzon 84, 153-159 Nakane, P and Kawaat, J. (1974) Peroxidase labelled antibody, a new method of conJugation. J. Hutochem Cytochem. 22, 1084-1091. Markwell, M. A K , Haas, S. M , Bteber, L L., and Tolbert, N E (1978) A modtfication of the Lowry procedure to stmphfy protein determination m membrane and lipoprotein samples. Anal Biochem 87,206-210. Egusa, G , Brady, D W., Grundy, S. M., and Howard, B. V (1983) Isopropanol prectpttation method for the determinatton of apolipoprotein B specific activity and plasma concentrations during metabolic studies of very low density hpoprotem and low density lipoprotem apohpoprotem B. J Lzpzd Res 24, 126 l-l 267.

12 Methodological Approaches for Assessing and Protein Oxidation and Modification in Plasma and Isolated Lipoproteins Wolfgang

Lipid

Sattler, Ernst Malle, and Gert M. Kostner

1. Introduction There is increasing evidence that oxtdatively modified low-denstty hpoprotems (LDL) play an important role during atherogenesis (I). LDL oxtdation can be mltiated by reactive oxygen species and is accompanied by characteristic changes m all of the major lipid subclasses as well as the apohpoprotem moiety of the LDL particle. These changes do not occur simultaneously, but rather at different time scales as a result of mitiation and propagation of the peroxrdation cycle and subsequentdecomposition of primary lipid-peroxidatton products. Therefore, a method applied at a given time point may give different results if applied at another ttme. The different techniques described here for the measurement of hptd and protem modification (for review, see refs. 2-5) mclude measurement of thiobarbituric actd reactive substances(TBARS), conjugated dienes, lipid hydroperoxides by photometric or high-performance hquid chromatography (HPLC) methods, consumption of polyunsaturated fatty acids (PUFAs) and antioxidants, cholesterol oxidatton products, aldehydes, or oxygen uptake during LDL oxidation The degree of protein modification is usually assessedby agarose gel electrophoresis, photometric measurement of free s-amino groups, sodium dodecyl sulfate-polyacrylamide gel electrophorests (SDS-PAGE), or more advanced analytical techniques, e.g., electrosprayliqutd chromatography-mass spectrometry (LC-MS). More recently, a variety of different enzyme-lmked immunosorbent assay (ELISA) procedures utihzing monoclonal antibodies (MAbs), which were generated toward oxidationspecific epitopes m apo B-100, have been established. The most important From

Methods m Molecular Edlted by J M Ordovas

B/ology, Vol 110 L/poprotem Protocols 0 Humana Press Inc , Totowa, NJ

167

Sattler, Ma//e, and Kostner

168 Table 1 Assessment

of Protein

Modification

Type of measurement Surface charge Free-amino groups Fragmentation of apo B- 100 TBARS Dttyrosme formation HNE, MDA, or HOC1 modified epitopes in apo B-100 FINE-epitopes

Macrophage uptake Liver uptake by rats Fluorescence

Method REM, agarose-gel electrophoresis Calorimetric with trinitrobenzene-sulfomc (TNBS) SDS-PAGE Colortmetric HPLC

Refs acid

or HPLC

6 6, 7 8 9 10

Specific monoclonal antibodies, suitable for ELISAs and immunohistochemtstry of plaque material

11-14

Electrospray-LC-MS; demonstrated preponderance of HNE-Michael adducts over Schiffs base formatton m B-lactoglobulin and hemoglobm Degradation by macrophages as a measure for the biological activity of a modified lipoprotein preparation One of the few methods where the biological properties of oxtdtzed lipoprotems are assessed in vivo Owmg to newly formed epitopes m apo B- 100 LDL fluorescence strongly increases during oxtdatton (Ex360/Em430 nm). Useful for contmuous monitoring of apo B modtficatton Dttyrosine formation

15

16

17

18

19

methods used to assesslipid and protein oxrdation or modificatron are summarized in Tables 1 and 2. During oxidation of a lipoprotem particle, a complex mixture of (per)oxrdatron products can be formed m the liprd and/or apohpoprotem domam in a trmedependent manner and this makes it extremely difficult to analyze the “degree of oxrdatron” or “modrfication” m vivo. Owmg to these drfficulties, often quantitatively undefined terms like “mmimally modified,” “heavily 0x1dized,” and so on are used m the literature. According to the complex nature of lipid peroxrdatton products a bulk of different techmques for the assessmentof lipid and protein oxidatton have been described and rt IS impossrble to cover fully all of these methods in this chapter. A frequently asked question is also:

Lip/d and Protein Oxidation Table 2 Assessment of Lipid (per)Oxidation Type of measurement Method Lipid Colorlmetric measurements. Levels of LOOH hydroperoxldes measured m plasma are consistently two to three orders of magnitude higher as compared to HPLC methods Conjugated dlenes Continuous measurement of 234 nm absorption; calculation of 2nd derivative gives addItiona mformation Oxygen Contmuous measurement with an oxygen electrode, consumption comparable oxldatlon characteristics as obtained during diene measurement PUFA consumption During lipoprotein oxidation PUFA consumption can be followed by GC analysis of Fatty acid methyl esters (FAMES) Antioxidant Antloxidants are lost during oxidation; however, consumption considerable amounts of LOOH are formed m the presence of a-TocH CholesterylesterHPLC with chemilummescence or electrochemical hydroperoxldes detection, highly selective and sensitive, applicable to plasma and isolated hpoprotems, demonstrated the confinement of the majority of circulating LOOH to HDL PhosphohpidHPLC with chemolummescence (CL) detection hydroperoxldes Oxidized GC/MS after hydrolysis or direct analysts by cholesterol UV-HPLC (esters) HydroperoxyAnalysis of trimethylsllyl-derivatives (TMS)-Me-FAs and hydroxybefore and after NaBH4 reduction fatty acids Aldehydes Headspace-GC, GC/MS Pentane exhalation Volatile breakdown products of lipid peroxidation, can be analyzed by GC Could be a measure for lipid peroxldatlon in vivo

169

Refs. 2&23, 62

24,25

26

8

27 28-31

31,32 33,34 35

36-38 39

“Which parameters should be apphed to assessthe oxldative stress in vivo?” Because there is no straightforward answer to this question, we describe here some methods that are commonly used in our laboratory that are suitable to analyze protein modification and/or lipid (per)oxidation in plasma and/or ISOlated lipoproteins (see Note 1).

170

Sattler, Male, and Kostner

2. Materials 2.1. Measurement

of Relative Electrophoretic

Mobility

1 2 3 4.

Electrophoresis equtpment I % Agarose-gels, Ltptdophor system (Immuno, Vtenna). Agarose gel-protem fixing solutton* sulfosallcyhc acid (0 5 %, w/v) Agarose gel-protem stains. Thiazm-Red or Coomassie Blue (1% m HZ0 contammg 1% acetic acid) 5 Agarose gel-lipid stain: Sudan-Black (1 g Sudan-Black m 40 mL H20 and 60 mL ethanol)

2.2. Measurement 1 2. 3 4 5 6 7 8 9. 10 11

of TBARS in Plasma

Mtcroplate reader. 96-Well microttter plates, An Eppendorf centrifuge Water bath 50 mM Butylated hydroxy toluene (BHT) m methanol. 200 mA4 Orthophosporic acid. n-Butanol Saturated aqueous NaCl solution 0 11 ii4 TBA reagent 800 mg thtobarbiturtc acid (TBA) m 50 mL 0.1 M NaOH 10% Trtchloroacetic acid (w/v) 0 67% TBA.

2.3. ELISA Assays Mtcroplate reader 2 96-Well mtcrottter plates (Maxtsorp, NUNC, Rosktlde, Denmark) 3 Phosphate buffered salme (PBS) pH 7.4. 1 43 g Na2HP04 x H20, 0.2 g KH2HP04, 0.2 g KCl, and 8 g NaCl/L. ELISA coating buffer: PBS, 1 mg/mL EDTA, and 10 @4 Trolox-C ELISA wash PBS plus 21 2 g NaCl and 0.5 mL Tween-20/L. ELISA dtluent: PBS contammg 1% bovine serum albumin (BSA). Hybrtdoma culture supertants Horseradish peroxidase-conjugated goat antimouse IgG (AG 18 1P Chemtcon, Temecula, CA) 9 Chromogen solution: 5 1 g citric acid x HZO, 0 15 g Na2HP04 x 2 H20, adjust the pH to 5 0, and add 0 1 g tetramethylbenztdme and 1.8 mM H202, adjust the final volume to 1 L (although the amounts gtven are to prepare 1 L of chromogen solution, usually the volumes needed m a given day are much less and the preparation ofthe solution should be scaled down as needed) 10. 2 A4 H,S04

2.4. Photometric

Evaluation

of Lipid Peroxidation

Products

1. Hitachi U-2000 spectrophotometer equipped with a 6-posmon cell holder and a personal computer for data acqutsmon.

Llpici and Protein Oxrdation 2 0.1 nuI4 cuso+ 3 Methylene blue reagent kit (Determmer Thousand Oaks, CA)

2.5. lodomefric

171 LPO, Kamiya Bromedrcal Company,

Assays

1 CHOD-iodrde reagent (Merck, Darmstadt, Germany) 2 Phosphohpase A2 from NUJU mocambzque (1580 U/mg protem) (Sigma, St Louis, MO). 3 Cholesterol esterase from Pseudomonas sp. (400 U/mg protem) (Sigma) 4. Triglyceride hpase (Calbrochem, Lucerne, CH) 5. Glutathrone peroxrdase (GPx) (Srgma). 6. 1.5% Sodmm taurocholate. 7 Ethyl acetate 8. 1 mA4HCl 9 Nitrogen 10 Ethanol 11 Acetic acrd/methylene chloride (3 2, v,v) (contammg 0 001% BHT) 12 Aqueous potassium rodrde (2 4 g/2 mL water) 13 0.35% CdClz 14. 0.2 MNa2S203 15, 100 m1I4 potassmm phosphate buffer, pH 7 2

2.6. HPLC Analysis 1 2 3 4. 5 6. 7 8 9. 10 11. 12 13 14 15. 16 17 18

2,2’-Azobrs(2-drammopropane)drhydrochlorrde (AAPH). 0.02% Acetic acid in methanol Hexane. Tabletop centrrfuge. TurboVap sample concentrator Mobile phase* methanol/2-propanol(1’ 1, v/v). LC-18 reversed phase (RP) column (25 x 0 4 cm, equipped wrth a 5-cm guard column) Waters 490E programmable multrwavelength detector. HPLC system. Cholesteryl lmoleate Toluene 2 mM 2,2’-azobrs-2,4-drmethylvaleromtrtle (AMVN) Argon A120, solid-phase extractron column Hexane. t-Butyl-methyl ether Methanol NaBH4

172

Sattler, Male, and Kostner

3. Methods

3.1. Methods for the Assessment of (Apolipo) Protein Modification in Plasma and Isolated Lipoproteins 3.1.1. Measurement of Relative Electrophoretic Mobility (REM) Treatment of LDL wrth a variety of different agents leads to modifications m some of the ammo acids of apo B-100. Among these modifying agents are acetic anhydride (40), secondary lipid peroxidation products like 4-HydroxyZ-nonenal (HNE) (42) and Malondialdehyde (MDA) (42), transition metals (43), myoglobin (44), hypochlorite (reagent and generated m vivo by the H202hallde/myeloperoxidase system; refs. 14 and 49, and cells m the presence of transition metals (1,4,5). A common feature of these agents is modification of the apohpoprotem moiety of the LDL particle, resultmg m an increased negative net charge, mainly owing to adduct formation of reactive aldehydes with posttively charged ammo acids. The increased negative net charge of these modified hpoprotem particles 1sreflected by a higher electrophoretic mobility in agarose gels as compared to native, unmodified LDL, and is correlated with the degree of apo B- 100 modification Measurement of REM is a relatively reliable method indicating changes m the apoprotein domain of LDL. In our laboratory, measurement of REM is routmely carried out with 1% agarose-gels of the Liprdophor system (Immuno). 1 2 3. 4 5

On thesegels load -50 pg LDL-protein to eachwell Run at a constantcurrent of 25 mA/gel(60 mm). Fix the gels in sulfosalicylic acid (0 5 %, w/v, 60 mm) Rinse in water (30 mm) Stain for proteins with Thiazm-Red or Coomasste Blue (1% m Hz0 contammg

1% acettc acid). Alternatively the staining can be done for lipids with SudanBlack (1 g Sudan-Black m 40 mL Hz0 and 60 mL ethanol). 6 Calculate the REM as the ratio of migration distances of the modified and the native LDL sample An example for Increasing REM of an LDL sample followmg MDA treatment IS shown m Fig. 1 (see Note 2).

3.1.2. Measurement of TBARS in Plasma Among the different analytical methods established for the assessmentof lipid and lipoprotein oxidation, measurement of TBARS is one of the most popular method to assess the degree of oxidation The test IS simple, rapid, and

inexpensive and (under certain circumstances) a relatively reliable method to determine MDA levels (either free or protein bound; see below) during lipid peroxidation in vitro. During this procedure, the sample(s) under investigation is (are) heated in the presence of TBA at low pH resulting in the formation of a pink chromophore ([TBA],-MDA) with an absorption maximum near 532 nm

Lipid and Protein Oxidation

173

Fig. 1. Electrophoretic mobility of LDL samples modified with increasing MDA concentrations. LDL (1 mg/mL in PBS) was incubated in the presence of a freshly prepared MDA solution for 3 h at 37°C under nitrogen. Final MDA concentrations in the incubation mixture: lane I,0 mrnol/L (control LDL); lane 2, 1 mmol/L; lane 3,5 rnmol/L; lane 4, 10 mmol/L; lane 5,25 mmol/L; and lane 6,50 mmol/L. Electrophoresis of the samples was performed on agarose gels using the lipidophor system. (or fluorescence at 552 nm). There are, however, some major shortcomings inherent to this assay and one should be aware of: 1. Aldehydes other than malondialdehyde (MDA) can form TBA adducts. 2. If no antioxidants are included during the assay procedure, a major part of MDA can be generated during sample treatment. 3. Sialic acid(s) (46), bilirubin, DNA, sugars, and cyclooxygenase products are reactive toward TBA (reviewed in refs. 3 and 9). To circumvent these problems, plasma or lipoprotein samples are heated in the presence of an antioxidant (usually BHT) or HPLC separation of the chromophores with subsequent fluorescence detection is applied. In the following section, two procedures for photometric measurement of total (free and protein bound) and free MDA in plasma are described. 3.1.2.1.

PHOTOMETRIC MEASUREMENT OF TOTAL MDA IN PLASMA

The following procedure was established by Jentzsch et al. (47). The assay is easy to perform and was designed for a high sample throughput in short time (e.g., during clinical studies). The assay includes 3 mMBHT during the TBARS measurement to avoid inadvertent lipid peroxidation during sample treatment. When BHT concentrations <3 mM are included in the assay, Jentzsch et al. recommend argon purging of the samples prior to the heating step. It is impor-

174

Sattler, Male, and Kostner

tant to note that the authors have observed MDA recoveries from plasma, which mrght be as low as 50% (47). The assay is performed as follows: 1 MIX 200 pL of sample (or standards, see below) with 25 uL BHT (54 mM m ethanol) and 200 pL orthophosphoric actd (200 mA4) m 2-mL Eppendorf tubes and vortex for 10 s. 2 Add 215 pL of TBA reagent (see Subheading 2.2., item 9) and vortex again. 3. Heat the sample m a water bath at 90°C for 45 mm 4 Cool at room temperature. 5 Extract TBARS with 500 pL n-butanol m the presence of 50 pL of a saturated NaCl solutron. 6. Separate phases by centrrmgation (lO,OOOg, 1 mm). 7. Transfer 250 pL of the upper organic layer to a 96-well microtrter plate and measure the absorbance at 535 nm (TBA-MDA adduct) and 572 nm to correct for baseline absorptron m a plate reader 8 Calculate the concentratron of TBARS by plotting A535-A572 vs concentratron of a freshly prepared MDA solutton The MDA stock is prepared by hydrolySIS of tetra-methoxypropane (tMP, 10 nuI4) with 10 m&f HCl for 10 mm at room temperature. Jentzsch et al. (47) recommend hydrolysis of 50 pL of tetramethoxypropane with 10 mL of HCl (10 mA4) for the preparation of MDAstandards. 3 1.2.2. PHOTOMETRIC METHOD FOR FREE MDA MEASUREMENT

IN PLASMA

This procedure is a slight modification (i.e., mclusion of BHT as an antioxidant) of a method described by Esterbauer and Cheeseman (9). 1. MIX 100 pL of plasma with 200 pL of me-cold Trrchloroacetrc acid (TCA) ( lo%, w/v, prectprtatron of plasma proteins) and 20 pL of BHT (stock solutron 50 mM, final BHT concentration in the assay is 3 3 mM) 2. Centrifuge at 10,500g in an Eppendorf centrifuge 3. MIX 200 pL of the supernatant with 200 pL of 0.67% TBA. 4 Heat m a water bath (95’C, 10 mm). 5. Cool at room temperature 6 Read the absorbance at 532 nm. Calibration curves are established with freshly prepared MDA stocks as described above.

In an attempt to improve the selectivtty of MDA measurements, numerous chromatographic methods for qualitative and quantitative analysis (including HPLC, gas chromatography flame ionization detection [GC-FID], GC-electron capture detection [ECD], and GC-mass spectrometry [GC-MS] analysis) in plasma and lipoproteins have been described and were recently reviewed by Kmter (48) (see Note 3).

Lipid and Protein Oxidation

175

3.1.3. ELlSAs Although there is still an ongoing debate on whether all oxidized lipids present m atherosclerotic lesions may originate from LDL oxidation (49,50), the in vivo modification of LDL must be considered a prerequisite for uptake of these particles by macrophage scavenger receptors and formation of cholesterol-engorged foam cells. Therefore, different immunochemical approaches have attempted to characterize the type and extent of modificatton m circulating LDL particles. However, not all antibodies that have been raised against modified forms of apo B have been rigorously demonstrated to be specific for this protein, but rather exhtbit a residual activity against native apo B or modified proteins other than apo B (II). Cu+2-oxidized LDL has been widely used as an antigen for the production of MAbs (51-53). Preobrazhensky et al. (54) reported a MAb-based immunoassay for evaluation of lipoprotein oxidation in isolated LDL and apo B- 100 modification m unfractioned serum samples. Measurement of human plasma “LDL oxidation levels” by sandwich ELISA assay revealed eightfold higher levels in isolated LDL fractions from hemodialysis subjects when compared to controls (55). In both assays,sampleswere apphed to microtiter plates that were precoated with MAb and detected by polycional anti apo B-100 antibodies. Kotam et al. (56) reported a sensttive and specific ELISA for localization of MDA-LDL m lipoprotein fractions separated by density-gradient ultracentrifugation. Holvoet and coworkers (13) designed a competitive ELISA for measurement of plasma levels of MDA-modified LDL. The significantly higher concentrations of MDA-LDL in patients with acute myocardtal infarction (1.4 + 0.1 mg/dL, y1= 60), carotid atherosclerosis (0.86 _+ 0.11 mg/dL, n = 22), and subacutely occluded carotid arteries (1.3 f 0.18 mg/dL, n = 11) m comparison to age-matched controls (0.19 + 0.02 mg/dL, n = 44) clearly revealed elevated plasma levels of MDA-LDL as a marker for unstable atherosclerotic cardiovascular disease. In a promismg approach to study another lipid-peroxidation product, Waeg and coworkers (12) reported the generation of MAbs suitable for specific detection of HNE-modified protems by indirect and competitive ELISA. These Abs, which could also detect HNE-protein-adducts in ox-LDL, exhibited no detectable crossreaction with proteins modified by MDA, nonanal, nonenal, and 4-hydroxyhexenal, and appear to be suitable tmmunologic tools for HNEmodified histidine determination in oxidized (1ipo)protems and peroxidized liver microsomes. For immunologic detection and measurement of hypochlorite (HOCl)modified LDL (II)-another modificatton of LDL leading to cholesterol enriched foam cells in a similar way as ox-LDL or acetylated LDL (57)-we

Sattler, Malle, and Kosher

176 200

T -o-

150

--

N-LDL

b

HOCL.LDL

*

HNE-LDL

- - * - - MDA-LDL e

--&--ox-LDL

9100-2 050

--

000

7

h-v_ 1

.

.t..-;

3 10

-w___

‘I 1000

100

-.

““‘:

* 10000

“‘34 1oDooo

dilution

Fig 2 Supernatants containing MAb clone 2DlOG9 do not crossreact wtth native-, aldehyde-modified, and Cu-oxidized LDL Antigens (1 pg/mL) were coated onto microttter plates before serial diluttons (from l/10 to l/100,000) of the hybndoma supematants were added. The antigens used were: native (N)-LDL; HOCl-LDL (800 HOC1 molecules/LDL particle); HNE-LDL (5 nuI4 HNE), MDA-LDL (20 rmI4 MDA); and Cu-ox-LDL (2.5 mMCu2+-tons). Antigen binding was detected using peroxtdase-conjugated rabbrt antimouse IgGs with tetramethylbenzidme as a substrate (optical densities are grven tn millt-absorbance units; mAU). have raised MAbs that do not crossreact with native LDL, MDA-LDL, HNE-LDL (Fig. 2) or even peroxynitrite-modified using the following indirect ELISA.

Cu*+-ox-LDL, LDL. We are

1. Coat ELISA plates with different LDL-modifications in PBS, pH 7.4, 1 mg/mL ethelenedtamine tetraacetic acid (EDTA), and 10 pMol/L Trolox-C for 18 h at 4°C. Use 100 pL/welI. 2 Wash unbound LDL four times with ELISA-wash (200 pL/well each wash) 3 Block unspecific binding sites with ELISA-dtluent (200 pL/well) 4 Incubate the plates wrth hybridoma supematants (50 l&/well) at an appropriate dilution (1: 10 to 1 100 000) at 37°C for 2 h Wash plates as noted. Add the secondary Ab* peroxidase-conjugated goat antimouse IgG. Develop color with 100 pL chromogen solution Stop the development by adding 50 uL H2S04 (2 M). Read the absorbance at 450 nm using a Hamilton microplate reader (see Note 4).

3.2. Measurement of Conjugated Dienes in Plasma and Isolated Plasma Lipoprotein During (per)oxtdatton of PUFAs, the isolated double-bond system is rearranged to a conjugated diene system with an absorption maximum at 234 nm.

177

Lipid and Protein Oxidation 0.85 Lagg c

0.6-

P~OpWJC.llO~-

Decompoaitlon-phase ,F---‘--

f c! E

0.4"

2

0.2-

'G ?? %

,./'

Time

(mm)

Fig 3 Contmuous momtormg of Cu2”-initiated LDL oxidatton by the conjugated diene method. LDL (50 pg protein/ml) was incubated m the presence of 1 67 wCu2+ at 30°C and the increase in 234 nm absorption was contmuously monitored The arrow mdtcates the duration of the lagphase of this parttcular LDL preparatton The continuous measurement of conjugated drene formatron IS routinely used to establish the suscepttbilrty of a given LDL preparation toward copperInitiated llprd peroxtdation. The advantages/disadvantages of thus method have been outlined m detail by Puhl et al. (24). Our standard oxidatton assay IS routinely performed with a Hitachi U-2000 spectrophotometer equipped with a &position cell holder and a personal computer for data acquisrtron. The cuvets are kept at 30°C by means of a Peltier element. 1, Dtlute LDL solutrons (see Note 5) wrth PBS to a final concentration of 250 pg of total LDL (corresponding to -50 pg LDL protein) to give a final volume of 3 mL The contents are mixed by gentle mversron of the cuvet. 2. Subsequently, add 50 pL of CuS04 (0 1 mM, prepared freshly every week). 3 Start data acquisition at 234 nm lmmedtately upon addition of copper ions. Record absorptron every 2 min for 5 h. A typical example of a contmuous dtene measurement of a peroxrdrzmg LDL sample is shown m Fig. 3. 4 Calculate the different oxidation parameters, The characteristtcs of an oxrdative profile (dtene formation vs time) consists of three consecutive phases, i.e., a lag, propagation, and decomposition phase. The lag phase (intercept of the tangent with the time axrs) is an estimate for the susceptrbrhty of a given LDL preparation toward copper-mediated oxrdatron (the phystological determinants and stgmficance of the length of the lag phase tn thts assay are not well-understood). The slope of the tangent reflects

178

Sattler, Maile, and Kostner

the rate of diene formation (AE/mm). Finally, the total diene concentration may be estimated by the maximum 234 nm absorbance using the molar absorptivity of 2.8 x 104/M/cm. It is important to keep in mmd that both the lag phase and the rate of propagation exhibit an exponenttal dependency on the temperature used during these experiments (59). In additton the LDL:Cu*+ ratio affects the lag phase and the rate of propagation (42). It is therefore important to perform oxidation experiments at constant temperatures and LDLCu2+ ratios when different LDL preparations are compared with regards to then susceptibility towards Cu*+ Initiated oxidation. Instead of Cu2’, a water-soluble peroxyl generator, AAPH at concentrations ranging from 0.5-2 mM is frequently used to elicit LDL peroxidation. This procedure yields very similar results as compared with Cu*+ oxidation (see Note 6). 3.3. Photometric Evaluation of Lipid Peroxidation Products in Plasma In this assay, the photometric methods described below are routmely used by different laboratories to assessthe degree of lipid peroxidation m plasma or isolated hpoprotems. A common feature of these photometric assaysis a redox reaction between lipid hydroperoxides and the chromophore of the particular assay.The major disadvantage of these photometric methods is the fact that the levels of lipid-peroxtdation products found in plasma are about two to three orders of magnitude higher as compared to results obtained with highly sensitive and selective analytical techniques like HPLC wtth chemilummescence or electrochemical detection (EC), owing to interference of various substances 3.3.1. Methylene Hue Method In this assay,lipid-hydroperoxide (LOOH) measurement 1sperformed with a commercially available methylene blue reagent kit (Determiner LPO, Kamya Biomedical Company). During the assay,serum or plasma samples are treated with a mixture of ascorbate oxidase and lipoprotein hpase m a first step. Subsequently, plasma LOOH are reduced to Lipid-Hydroxide (LOH) m the presence of hemoglobm and methyl-carbamoyl-methylene blue, resulting m the concomitant formation of methylene blue (22). As pointed out by the manufacturers, the followmg precautions have to be taken to avoid erroneously high LOOH values. test tubes have to be free of contaminating traces of detergents, abnormally high levels of bihrubm ~111interfere with the assay, and immediate separation of plasma or serum from blood samples is required. Mix 20 & of plasmawith 200 ).ILof reagent 1 (contammgascorbateoxrdaseand lipoprotein lipase). 2 Incubate at 30°C for 5 mm under nitrogen or argon atmosphere 1

179

Lipid and Protein Oxidation 3 Add 400 ene blue) 4 Incubate 5. Measure 6 Establish peroxide

pL of reagent 2 (contammg hemoglobm and methyl-carbamoyl-methylfor 15 mm at 30°C. the absorptton of liberated methylene blue. the calibration curve with Increasing concentrations

of cumene hydro-

3.3.2. Direct lodometric Assay of Plasma LOOHs This assay is based on the capacity of LOOHs to convert free I, to I,-, which can be measured photometrically at 365 nm (20). However, this assay is also likely to result m an overestimation of the actual content of LOOH owmg to its reactivity toward other compounds. Plasma samples are spiked with EDTA and BHT (25 and 20 pJ4, final concentrations, respectively) to inhibit madvertent lipid peroxidation during sample handling and durmg the assayprocedure The ortgmal assay is performed as follows: 1 2 3 4

Mix 100 pL of plasma wtth 1 mL of the Merck CHOD-iodide reagent Incubate for 30 mm at room temperature in the dark Measure the absorbance at 365 nm Calculate the concentrattons of hpid hydroperoxtdes by the molar extmctton coefficient of I,- (E = 2 46 x lO,/M/cm) Cumene hydroperoxtde can be used to test the linearity of the assay; tt is, however, not recommended as a standard because tt was shown that Cumene-OOH 1s only partially reduced under the assay condtttons described above (59)

3.3.3. lodometnc Assay of Free Fatty Acid-Hydroperox/des (FFAOOH) Following Enzymatic Hydrolysis of Plasma Lipids The calorimetric assay described here was established to measure free FFAOOH m plasma followmg enzymatic hydrolysis of all major hpid subclasses. Owing to some important control steps,this iodometric procedure should allow a more specific and sensitive measurements of plasma LOOH as compared to the previously described iodometric measurement. However, thts procedure can not be easily applied during studies where high sample throughput is required. The method outlined below was originally described by Cramer et al. (21). Because it requires a time-consuming hydrolyses step at 37°C mcreasmg the risk of inadvertent sample oxidation (and therefore overestimation of the true hpidhydroperoxide content), we work under a nitrogen atmosphere whenever possible. In a first step, enzymatic hydrolysis of total plasma hpids is performed on one mL aliquots of plasma. 1 Hydrolyze 1 mL plasma for 6 h (37°C) m the presence of 10 IU of phosphohpase A,, 1 IU of a nonspectfic Pseudomonas esterase, 100 IU trtglyceride lipase, and 1 5% sodium taurocholate (thm layer chromatography [TLC] revealed vntually complete hydrolysis of estertfied plasma lipids)

180

Sattler, Ma//e, and Kostner

2 Acidify the samples to pH 3.5 with 90 pL crtrtc acid (2 M) 3 To extract fatty acids and then peroxtdation products, add 2 mL ethyl acetate (containing 0 01% BHT) and 1 mL 100% ethanol. 4. Centrifuge at 5OOg for 10 mm at 4°C 5 Aspirate the top organic layer and transfer to another Pyrex tube Sometimes phase separation IS only achieved when the samples are cooled to -20°C prior to centrifugation 6. To this tube, add 2 mL ethyl acetate and 1 mL deiornzed water to achieve phase separation 7. Remove the upper orgamc layer and wash with 2 mL HCl(1 mM) 8. Aspirate the washed organic layer 9 Dry with a stream of nitrogen and resuspend n-r50 p.L of ethanol 10 Add to the samples 350 pL acetic actd/methylene chloride (3.2, v v) (contammg 0 001% BHT) and 15 Ccs,aqueous potassium iodide (2 4 g/2 mL water) 11. MIX and Incubate for 3 min in the dark. 12 Add 1.O mL 0 35% CdC12. 13 Centrifuge at 200g for 5 min at room temperature 14. Remove the aqueous phase for measurement of the absorbance at 353 nm Measure the absorbance after treating the contents of the spectrophotometer cuvet with 10 pL of 0 2 M Na2S203 to correct for non-Is- absorbance. All measurements are made against solvent blanks, prepared by substitutmg 1.O mL of 100 mM potassium phosphate buffer, pH 7.2, for 1 mL of plasma Liberated fattyacid hydroperoxides can be reduced by GPx (60), thus for blank measurements, each sample is treated with 5 U of GPx and the hydroperoxtde content is calculated as the difference m absorbance obtained from samples incubated in the presence and absence of GPx. Concentrations of fatty acid hydroperoxides estimated by the absorbance of trtiodide are calculated usmg the molar absorptivity of Is- at 353 nm (E = 2.3 x 104/h41cm).

This modified iodometric assay corrects for two major sources of possible interferences, i.e., backtitration of 11 to I- with Na2S203 corrects for absorbance caused by nonhydroperoxide material, whereas reduction of hydroperoxides with GPx eliminates absorbance caused by nonhydroperoxtde oxidants. In an attempt to validate the usefulness of the three calorimetric methods previously described for routine use, we have measured the LOOH concentration

present in plasma obtamed from 30 different donors. During this study, LOOH concentrations measured by the modified iodometric procedure (Subheading 3.1.3.) were significantly lower as compared to the umnodtfied procedure (Subheading 3.1.2.): 5.8 f 1.9 l.tA4vs 4.2 + 2.7 p.M,p C 0.02 and even lower as

compared to the methylene blue procedure (4.2 + 2.2 pA4 vs 8.56 f 5.82 pA4). However, as mentioned earher, the (circulatmg) LOOH concentrations m plasma estimated with the three calorimetric kits was still approximately two orders of magmtude higher as compared with data obtained by plasma analysis with selective HPLC methods, i.e., 3-230 nMas outlined in ref. 61 (see Note 7).

Lip/d and Protein Oxidation

161

3.4. HPLC Analysis of Lipidhydroperoxides in Lipoproteins This chapter describes some HPLC assaysparttcularly suitable to detect different classes of lipid hydroperoxides in lipid extracts of lipoproteins. Exposure of lipoproteins to oxidizing conditions results in core and surface lipid oxidation. Depending on the chemical composttton of the lipoprotem class under investigation, cholesterylester-hydroperoxides (CEOOH), triglyceridehydroperoxides (TGOOH), or phosphohpid-hydroperoxide (PLOOH) may be preferenttally formed during radical-initiated lipoprotem oxidation. For LDL, the ratio of CEOOH:PLOOH is approx 4: 1, whereas it 1s -1: 1 when highdensity lipoproteins (HDL) was subjected to oxidation experiments. However, in addition to CEOOH formation during AAPH-mediated oxidation of LDL and HDL, we have observed the formation of considerable amounts of cholesterylester-hydroxides (CEOH) even during early stagesof oxidation (62). Therefore a specific HPLC assayshould allow detection of very small concentrations of LOOHs and should also distinguish between LOOH and LOH. Our routme oxidation experiments are performed with a thermolabile, watersoluble peroxyl radical generator. Lipoprotem lipids are extracted m a biphasic solvent system consisting of MeOH/hexane (1:5, v/v) containing 0.02% acetic acid. In this biphasic extraction system, a preseparation of polar and neutral lipid(hydroperoxide)s is achieved. Phospholipids (PLs), PLOOHs as well as water-soluble antioxidants (e.g., urate and ascorbate) can be directly analyzed from the methanoltc phase, whereas cholesterol (Ch), triglycerides (TGs), CEs, and the correspondmg hydroperoxides are confined to the hexane phase. Lipldsoluble antioxidants (tocopherols, ubiquinols, carotinoids) are also found m the hexane phase. 3.4. I. Analysis of Choles teryles ter-Hydro(pero)xides by Simultaneous UV Detection at 234 and 210 nm The HPLC method described below was originally developed for a combination of UV and postcolumn chemiluminescence detection as described by Frei et al. (28) and Yamamoto (29,63) and modified as described elsewhere (31). These procedures apply a combination of UV-detection (210 nm, nonoxidized lipids) and postcolumn chemilummescence detection (CEOOH or PLOOH). The UV-detector efflux is passed into a chemiluminescence detector, where lipidhydroperoxides (but not lipid-hydroxides; see below) are detected by the microperoxtdase/isolummol reaction. Because we have currently no chemtluminescence detector at our disposal, we compromtse by simultaneous UV-detection at 210 and 234 nm. There, nonoxidized lipids are detected at 2 10 nm and the conjugated diene system of the corresponding cholesterylesterhydro(pero)xides is detected at 234 nm, respectively. This allows separation of different lipid classes,the mam disad-

182

Sattler, Male, and Kostner

vantage of 234~nm detection is tts low sensitivity, being several orders of magnitude less sensitive as compared to the HPLC postcolumn chemiluminescence detection. Consequently, 234-nm detection is not very well-suited for the assessment of lipid peroxtdation during very early stages, where concentrations of LOOHs are in the low nanomolar range. In the following a HPLC procedure allowing the simultaneous detection of cholesterol, CEO(O)Hs, and CEs wtthm a 30-min HPLC run is described. It is noteworthy, however, that, unlike HPLC-CL, detection at 234 nm does not allow selective detection of CEOOHs, because CEOOHs and CEOHs are not separated under the HPLC conditions used. Both of these lipid-peroxidation products contam a conjugated double-bond system m the molecule and contribute to the 234 nm absorption m a single peak. This is important because during peroxyl radical mediated peroxtdation of LDL and HDL, considerable amounts of CEOHs are formed (62). 1. Oxrdrze LDL or HDL (0 5 and 2 mg protein/ml m PBS, respectively) with a water-soluble peroxyl generator (e g , AAPH) 2 Transfer 400 pL of the hpoprotein solution to a Pyrex tube contammg 2 mL of methanol (0.02% acettc acid) and 10 mL of hexane 3. Extract on a vortex mixer. 4. Centrifuge at 2,000g for 5 mm (4’C) to achieve phase separation 5. Withdraw 9 mL of the upper hexane layer to a comcal glass tube. 6. Dry under argon m a TurboVap sample concentrator 7 Redissolve m 180 pL of the mobile phase (methanol/2-propanol, 1 1, v/v) and analyze on a LC-18 RP column (25 x 0.4 cm, equrpped wrth a 5-cm guard column) eluted with methanol/2-propanol (1 1, v/v) at a flow rate of 1 mL/mm as the mobile phase. Detection 1sperformed srmultaneously at 2 10 and 234 nm with a Waters 490E programmable multrwavelength detector, allowing simultaneous absorptron measurement at four different wavelengths.

Figure 4 shows a typical UV trace of the hexane extracts of native and oxidized HDL (1.8 mg protein/ml, 2 mM AAPH, 2 h, 37”(Z), using the chromatographic

conditions

described

above. Under these experimental

conditrons,

cholesterol and the main cholesterylesters are clearly separated from each other and can be quantitatively analyzed by peak area comparison with external standards of known concentration (for an internal standard procedure, see ref. 31). The preparation of cholesteryllmoleate-hydroperoxide and cholesteryllinoleate-hydroxide

standards is described below.

A modification of the mobile phase described above allows the separation of cholesterylester hydroperoxides and cholesterylester hydroxides (64). This mobile phase consists of acetonitrile/2-propanol/H20 (44:54:2, v/v/v). These chromatographtc conditions allow separation of CEOOH and CEOH, respectively; the elutron times of the unoxrdtzed cholesterylesters

are as long as 60 mm,

Lipid and Protein Oxidation

183

.016.

B 1

0 0

5

IO Retention

I5 Time

20

25

30

(min)

Fig. 4. HPLC separatron of hexane-extracted HDL lipids with simultaneous detection at 210 nm (unoxldlzed liplds) and 234 nm (Ch18 20(O)H) HDL (2 mg protem/ mL) was oxldlzed in the presence of AAPH (2 mM) at 37°C. Lipids were extracted m hexanelmethanol, the hexane phase was recovered, dried under argon, and redissolved m 180 $ mobile phase LIpIds were separated on a LC- 18 column (25 x 0 46 cm) with methanol/2-propanol as the mobile phase (1 mL/min) The HPLC traces shown represent detection of nonoxldlzed lipids at 2 10 nm obtained from native HDL (A) and detection of Chl8:20(O)H in lipid extracts obtained from native (B) and oxidized (C) HDL. Note the different scaling factors. Peak assignment: 1, cholesterol; 2, Chl8.2; and 3, Ch18:20(O)H however. Figure 5 shows an HPLC chromatogram obtained from HDL llpld extracts oxidized in the presence of 2 mMAAPH. From the HPLC trace shown in Fig. 5, it IS evident that during peroxyl radical induced oxldatlon of HDL, considerable amounts of CEOH are generated, and that rt IS important to dtstmguish between these oxtdatlon products (see Note 8).

184

Sattler, Male, and Kostner

A

2

l.Ch182OCH 2=Chi8’2-OH

6

1

4 2 5. A 2 2

-I

0 0

5

10 Retention

15

20

25

Time (mln)

Ftg 5 HPLC separatton of Chl8 200H and Chl8:20H (A) shows separation of a Ch18:200H (1) and a Ch18.20H (2) standard. The trace shown m (B) represents separation of Ch18.200H and Ch18:20H present m hexane hprd extracts obtained from AAPH-oxidized HDL. Analysts was performed on a LC-18 column (25 x 0.46 cm) eluted with acetomtrtle/2-propanol/water (22.27.1, v/v/v) at 1 mL/mm and momtored at 234 nm

3.4.2. Analysis of Phospholipid-Hydroperoxides by HPLC-UV-CL Detection

(PLOOH)

PLOOH from peroxrdizing HDL samples can be analyzed directly from the lower methanolic phase of the biphastc lipid-extraction system described above. PLOOH are analyzed by ion-exchange chromatography on a SupelcostlNH2 column (25 x 0.46 cm) with methanol/50 mA4NaH2P04 (95.5, v/v) as the mobile phase (flow rate of 1 mL/min) with subsequent UV-CL detection (32). Phosphatidylcholine-hydroperoxide (PCOOH) and phosphatidylethanolamme-

L/p/d and Protein Oxidation

185

hydroperoxide (PEOOH) are not separated under these conditions. Miyazawa et al. (32,66) have described an improved HPLC-UV-CL method for phosphatidylcholine-and phosphatidylethanolamine-hydroperoxide analysis in lipid extracts obtained from rat liver and brain specimens. These authors recommend a mobile phase consisting of hexane/:!-propanol/methanol/water (5:7:2:1, all vol) to separate PCOOH and PEOOH on a Jasco Finepak (25 x 0.46 cm) N-propylamine column. Separation of PCOOH and PEOOH is achieved within a 20 min HPLC run. 3.43. Preparation of Cholesteryllinoleate and -Hydroxide

Hydroperoxide

1 20 mg of cholesteryllmoleateis oxidized in 1 mL of toluene with 2 mM AMVN

(5 h, 37°C) and dried under argon 2. Redissolve the dried restdue m 1 mL of hexane andapply to an A1203solid-phase extraction column (bed volume 1 mL preconditioned with hexane). 3. Elute the unoxtdrzed Ch18.2 with hexane and the Chl8.2-OOH with 15 mL of t-butyl-methyl ether (Merck) 4 Dry the resulting Ch18:2-OOH standard and redisolve the residue in hexane (dilution factor - 1 100). 5. Determine the concentration of Ch18:2-OOH by its 234 nm absorbance, using a molar-absorbancy coefficient of 29.500 M-l x cm-‘.

These standards are stable for up to 6 mo at -70°C when stored m glass ampoules. Ch18:20H is prepared by chemical reduction of Ch18:200H with NaBH4 in methanol at 4°C essentially as described in (35). Briefly: 1 2 3 4. 5. 6. 7

Dry under nitrogen 5-10 mg Ch18.200H in hexane Redissolve m 1 mL methanol Add 20 mg NaBH,. Leave the sample on ice for 60 mm Add 1 mL of water to destroy excess NaBH+ Extract Chl8:20H mto hexane. Determine the concentratron by 234 nm absorption as described above.

9- and I3-cholesteryllinoleate-hydroxides are also commercially available from Cayman Chemicals (Ann Arbor, MI). Another point worth mentioning is that neutral A1203 can considerably vary in activity, therefore, sometimes hexane/diethylether (95:5, v/v) has to be used to achieve elution of the unoxidized Ch18.2 from the A1203 columns. Alternatively the Chl8:2/Ch18:200H mixture can be purified on a preparative LC- 18 column with MeOH/2-propanol as mobile phase.

Sattler, Male, and Kostner 4. Notes 1 Momtormg plasma hprd oxidation IS not an easy task Thus there IS not a single test reflecting all modrfications of plasma constttuents by “oxrdattve stress.” 2. REM IS only a rather crude measure to momtor for lipid oxrdatron because modrficatrons not changing the charge of hpoprotems will not be detected. 3 TBARS comprise a variety of compunds, and only part of them are MDA or HNE. Thus tf one wants to measure a specific substance, the use of chromatographic methods IS unavoidable 4. There are numerous MAbs described m the literature, most of them recogmzmg different epltopes In many cases, the exact eprtope IS not even characterized Special care must be grven to the charactertstms of the antibodies, whether they show crossreactlvrty with nonmodtfied hpoprotems or oxldatrvely modrfied plasma proteins. 5. For studies definmg very early stages of oxrdatton, tt IS necessary to avotd hpoprotein peroxidation during the purification steps We therefore isolate very lowdensity hpoprotems (VLDL), LDL, and HDL wrthm a 120- to 240-mm run using m a Beckman TLX Optima bench-top centrifuge (31). Preparations are carried out with a TL 100 4 rotor holding eight tubes with a volume of 5.1 mL each, enablmg hpoprotem preparation from a total volume of 13.6 mL plasma m a smgle run. The method described for the rapid tsolatron of plasma hpoprotems of different denstty was optlmrzed to obtain hpoprotem preparations vntually free of contammatmg albumin (except for the HDL, fraction) It IS noteworthy that shorter centrtfugation times than those recommended can result in lipoprotem preparations heavily contaminated with albumm The small amounts of watersoluble antioxidants such as ascorbate or urate present m the hpoprotem preparations are removed together with EDTA and KBr by dialysis or size-exclusion chromatography These hpoprotem preparations can be readtly utthzed for OXIdatton experiments. Once the hpoprotems are purified, they should be used at the earliest possible time point The methods described differ m then- in selectrvtty and sensmvrty However, a feature common to all of these methods is the requtrement to avoid inadvertent oxldatron during sample handlmg or work-up. This IS normally achieved by unrnedrate tsolatron of plasma or serum and mcluslon of antroxtdants where necessary In addmon, sample handlmg under nitrogen or argon mmimtzes the risk of inadvertent sample oxrdatron, e.g., exclusion of BHT during measurement of total TBARS m plasma or tissue samples results m the (artrfactual) generation of huge amounts of lipid peroxrdatlon products m the actual assay procedure durmg the heating step. 6 In contrast to Cu2+-mediated oxtdatton, thermo-labile peroxyl-radical generators like AAPH (water soluble) or AMVN (lipid soluble) allow the generation of peroxyl-radicals at controlled rates Online measurement of dtene formation durmg oxtdatton experiments IS easy, reliable, and reproducible, given that some important precauttons are taken Measurement of dtene formation during AAPH-nuttated oxtdatlon experiments

Lip/d and Protein Oxidation

187

requires the mcluston of AAPH m the reference cuvet owmg to the (thermal) formation of AAPH decomposttton products contributing to 234 nm absorptron During studies where LDL samples obtained from different donors are compared with regard to their oxtdtzabthty, it 1sextremely important to perform the oxldation experiments at the same temperature and the same copper LDL ratios One should also avoid the mcluston of substances that are not removed during desaltmg of LDL samples (e g., chloramphemcol) and contribute to 234 nm absorption (for details, see ref. 24) 7 The method outlined by Cramer et al (21) appears to be the most reliable one because tt corrects for all possible interferences. Owing to tts lengthy protocol, tt may be not applicable for routme analysts For routine use, the Methylene Blue method is recommended 8 As outlined above, HPLC analysis of CEOOH and PLOOH by postcolumn chemtlummescence detection IS htghly sensitive and selective It is important to note that under the chromatographic condmons described (1 e , methanol/2-propanol, 50 50, v/v), CEOOH and CEOH coelute, and this could lead to arttfactual results when UV detection at 234 nm 1sused instead of CL detection However, the mobile phase described by Krithartdes et al (64) allows separation of CEOOH and CEOH. Cholesteryl lmoleate-hydroperoxtde 1s used as a standard for quantttatlon. The preparation of this compound 1s a relatively straightforward procedure (see above) Because the neutral A1203 solid phase extraction columns can vary tn then activity, it is sometimes necessary to elute the nonoxrdtzed Ch18.2 with hexane containing 5% diethyl ether.

Acknowledgment This work was supported by grants from the Austrian Research Foundation (P12000 to WS, P 11276 to EM and P 11691 -Med to GMK).

References 1. Steinberg, D., Parthasarathy, S , Carew, T. E., Khoo, J. C., and Wttztum, J. L. (1989) Beyond cholesterol: modtficattons of low-density hpoprotem that increase its atherogemcity N Engl J Med 320,915-924. 2 Esterbauer, H. and Ramos, P. (1995) Chemistry and pathophystology of oxtdation of LDL. Rev Physzol Biochem Pharmacol 127,3 l-64. 3 Chait, A (1992) Methods for assessing lipid and lipoprotein oxidation Curr Opmlon Llpzdol 3,389-394

4. Stembrecher, U. P., Zhang, H., and Lougheed, M (1990) Role of oxidattvely modified LDL m atherosclerosts. Free Radzc Bzol Med 9, 155-168. 5 Berliner, J A and Heinecke, J. W (1996). The role of oxtdtzed hpoprotems m atherogenesis. Free Radlc Btol Med. 20,707-727 6. Malle, E., Ibovntk, A., Leis, H. J., Kostner, G. M., Verhallen, P. F. J., and Sattler, W. (1995) Lysme modification of LDL or hpoprotem (a) by 4-hydroxynonenal or

Sattler, Ma//e, and Kostner

188

7

8

9

10. 11.

12 13

14

15. 16.

malondialdehyde decreases platelet serotonm secretion without affecting platelet agg&gability and eicosanoid formation Artenoscler Thromb Vusc Blol l&377-384. Stembrecher, U. P. (1987) Oxidation of human low density lipoprotein results m derivatization of lysme residues of apolipoprotein B by hpide peroxide degradation products. J. Blol Chem. 262,3603-3608. Sattler, W., Kostner, G. M., Waeg, G., and Esterbauer, H. (1991) Oxidation of lipoprotem (a) A comparison with low-density lipoproteins. Biochzm Bzophys Actu 1081,65-74. Esterbauer, H and Cheeseman, K H. (1990) Determination of aldehydic lipid peroxidation products: malonaldehyde and 4-hydroxynonenal. Methods Enzymol 186,40742 1 O’Connell, A. M., Gieseg, S. P., and Stanley, K. K (1994) Hypochlorite oxidation causes cross-linking of Lp(a) Blochim Bzophys Acta 1225, 180-l 86 Malle, E , Hazell, L., Stocker, R., Sattler, W., Esterbauer, H., and Waeg, G (1995) Immunologic detection and measurement of hypochlorite-modified LDL with specific monoclonal antibodies Artenoscler. Thromb Vast. Blol 15,982-989. Waeg, G., Dimsity, G , and Esterbauer, H. (1996) Monoclonal antibodies for detection of 4-hydroxynonenal modified proteins Free Rad Res. 25, 149-l 59 Holvoet, P., Perez, G , Zhao, Z., Brouwers, E., Bernar, H , and Collen, D (1995) Malondialdehyde-modified low density hpoprotems in patients with atherosclerotic disease J CEm Invest 95,261 l-2619. Hazell, L , Arnold, L , Flowers, D , Waeg, G., Malle, E , and Stocker, R (1996) Presence of hypochlorite-modified protems m human atherosclerotic lesions. J Clan Invest. 97, 1535-1544. Bruenner, B A , Jones, A D , and German, J. B. (1995) Direct characterization of protein adducts of the lipid peroxidation product 4-hydroxy-2-nonenal using electrospray mass spectrometry Chem. Res Tox~ol 8,552-559 Goldstein, J. L , Ho, Y. K., Basu, S K., and Brown, M S (1979) Binding site on macrophages that mediates the uptake and degradation of acetylated low density hpoprotem, producing massive cholesterol deposition. Proc N&l. Acad. Scr USA 76,333-337.

17. Van Berkel, T. J., De Rijke, Y. B., and Krurt, J. K (1991) Different fate in vivo of oxidatively modified low density lipoprotem and acetylated low density lipoprotein m rats. Recognmon by various scavenger receptors or Kupffer and endotheha1 liver cells. J Blol. Chem 266,2282-2289. 18. Koller, E , Quehenberger, O., Jurgens, G., Wolfbeis, 0 S., and Esterbauer, H (1986) Investigation of human plasma low density hpoprotem by three-dimensional fluorescence spectroscopy FEBS Lett. 198,229-234. 19 Hemecke, J. W., Ll, W., Daehnke,H. L., and Goldstein,J. A. (1993) Dltyrosme,a specific marker of oxidation is synthesized by the myeloperoxidase-hydrogen peroxide system of human neutrophils and macrophages J Bzol Chem 268,406MO77 20. El-Saadam, M., Esterbauer, H., El-Sayed, M., Goher, M , Nassar, A. Y., and Jtirgens, G. (1989) A spectrophotometric assay for lipid peroxides in serum hpoproteins using a commercially available reagent. J. Lipid Res 30,627-630.

Lipid and Protein Oxidation

169

2 1 Cramer, G L., Miller, J. F , Pendleton, R B., and Lands, W. E. M (1991) Iodometric measurement of lipid hydroperoxldes m human plasma Anal Blochem 193, 204-2 11 22. Tatelshi, T., Yoshimine, N., and Kuzuya, F (1987) Serum hpld peroxide assayed by a new colorlmetrlc method. Exp. Gerontol 22, 103-l 11. 23 Jlang, Z. Y., Hunt, J. V., and Wolff, S. P. (1992) Ferrous iron oxidation m the presence of xylenol orange for detection of llpld hydroperoxlde in low density hpoprotem. Anal Bzochem 202,384-389. 24. Puhl, H., Waeg, G , and Esterbauer, H (1994) Methods to determine oxldatlon of low-density hpoprotems. Methods Enzymol. 233,425-441 25 Corongiu, F P and Banm, S (1994) Detection of conjugated dlenes by second derivative ultraviolet spectrophotometry. Methods Enzymol 233,303-3 10 26 Sato, K., Nlki, E., and Shimasaki, H. (1990) Free radical-mediated chain oxldatlon of low density hpoprotem and its synergistic mhlbltlon by vitamin E and vitamin C Arch Blochem Blophys 279,402-405. 27 Bowry, V W and Stocker, R. (1993) Tocopherol-mediated peroxldatlon. The prooxldant effect of vitamin E on the radical-initiated oxidation of human low density lipoprotein. J Am Chem Sot 115,6029-6044 28. Frel, B., Yamamoto, Y , Nlclas, D , and Ames, B N (1988) Evaluation of an lsoluminol chemllummescence assay for the detectlon of hydroperoxldes m human blood plasma Anal Biochem 175,120-l 30 29. Yamamoto, Y. (1994) Chemiluminescence-based high-perfomance liquid chromatography assay of lipid hydroperoxides. Methods Enzymol 233,3 19-324 30 Bowry, V. W., Stanley, K K , and Stocker, R. (1992) High density hpoprotem 1s the maJor carrier of lipid hydroperoxides m human blood plasma from fasting donors. Proc Nat1 Acad. Scz USA 89, 10,316-10,320 3 1 Sattler, W , Mohr, D , and Stocker, R (1994) Rapid isolation of lipoproteins and assessment of their peroxldatlon by high-performance liquid chromatography postcolumn chemuluminescence. Methods Enzymol. 233,469+89. 32. Mlyazawa, T , Suzuki, T., FuJimoto, K., and Yasuda, K (1992) Chemllummescent simultaneous determination of phophatidylcholine hydroperoxlde and phophadltylethanolamme hydroperoxide m the liver and brain of the rat. J Llpzd Res 33, 1051-1058. 33. Carpenter, K L , Taylor, S E., van der Veen, C , Wllhamson, B. K., Ballantine, J A , and Mitchinson, M. J. (1995) Lipids and oxidised lipids in human atherosclerotic lesions at different stages of development. Blochim. Bzophys Acta 1256, 141-150 34 Brown, A J , Dean, R. T., and Jessup, W. (1996) Free and esterlfied oxysterol* formation during copper-oxldatlon of low density hpoprotem and uptake by macrophages J. Lipid Res. 37,320-335. 35. VanKuijk, F J G. M , Thomas, D W , Stephens, R J., and Dratz, E A (1990) Gas chromatography-mass spectrometry assays for lipid peroxides Methods Enzymol 186,388-398

Sattler, Ma//e, and Kostner

190

36. Frankel, E. N., Hu, M. L., and Tappel, A. L. (1989) Rapid headspace gas chromatography of hexanal as a measure of lipid peroxidation m biological samples. Lzpids 24,97&98 1 37. Thomas, M J , Robison, T W , Samuel, M , and Forman, H J (1995) Detecting and identifiymg volatile aldehydes as dinitrophenylhydrazones using gas chromatography mass spectrometry Free Radtc BloE Med 18, 553-557. 38. Luo, X P , Yazdanpanah, M , Bhoor, N., and Lehotay, D. C. (1995) Determmation of aldehydes and other lipid peroxidation products m biological samples by gas chromatography-mass spectrometry. Anal Brochem 228,294-298 39 JeeJeebhoy, K. N. (1991) In viva breath alkane as an index of hpid peroxidation Free Radlc. Bzol Med 10, 191-193 40 Basu, S K , Goldstem, J. L , Anderson, R. G. W., and Brown, M. S. (1976) Degradation of catiomzed low density lipoprotem and regulation of cholesterol metabolism in homozygous famihal hypercholesterolemia fibroblasts Proc Nat1 Acad Scl USA 73,3179-3182 41 Esterbauer, H , Dieber-Rotheneder, M., Waeg, G., Striegl, G., and Jurgens, G (1990). Biochemical, structural, and functional properties of oxidized low-density hpoprotem Chem Res Toxrcol 3,77-92. 42. Haberland, M , Fong, D , and Cheng, L (1988) Malondraldehyde-altered protein occurs m atheroma of Watanabe heritable hyperhpidemic rabbits Sczence 241, 215-218. 43 Ramos, P , Gieseg, S P , Schuster, B., and Esterbauer, H (1995) Effect of temperature and phase transition on oxidation resistance of low density lipoprotem J Llpld Res 36, 2113-2128. 44. Hogg, N., Rice-Evans, C., Darley-Usmar, V , Wilson, M. T , Paganga, G., and Bourne, L (1994) The role of lipid hydroperoxides m the myoglobm-dependent oxidation of LDL. Arch Blochem Bzophys 314,3%-44 45. Heinecke, J W. (1994) Cellular mechanisms for the oxidative modification of hpoprotems* implications for atherogenesis. Coronary Artery Dzs 5,205-210. 46 Warren, L. (1959) The thiobarbituric acid assay of sialtc acids J Bzol Chem 234, 1971-1975 47. Jentzsch, A. M., Bachmann, H , Furst, P., andBiesalski, H K (1996) Improved analysis of malondialdehyde in human body fluids Free Radlc. Blol. Med. 20,25 l-256 48 Kmter, M. (1995) Analytical technologies for lipid oxidation product analysrs J Chromatogr B 671,223-236. 49. Boyd, H. C., Gown, A. M , Wolfbauer, G , and Chait, A (1989) Direct evidence for a protein recognized by a monoclonal antibody against oxidatively modified LDL m atherosclerotic lesions from a Watanabe heritable hyperhpidemrc rabbit Am J Path01 135, 815425. 50 Rosenfeld, M E , Palmski, W , Yla-Herttuala, S , Butler, S , and Witztum, J L (1990) Distribution of oxidation specific lipid-protein adducts and apohpoprotem I3 m atherosclerotic lessons of varying severity of WHHL rabbits. ArterzoscleroSlS

10, 336-349

51 Palmski, W , Yla-Herttuala, S., Rosenfeld, M E , Butler, S. W., Socher, S. A, Parthasarthy, S , Curt~ss, L K , and Witztum, J (1990) Antisera and monoclonal

Lipid and Protein

52

53

54

55

56

57

58.

59 60 61

62

63

64

65.

Oxidation

191

antibodies specific for epitopes generated durmg oxtdative mod&cation of low density lipoprotem Arterzoscleroszs 10, 325-335 Magil, A B , Frohhch, J. J , Inms, S M., and Stembrecher, U. (1993) Oxidized low-density lipoprotem m experimental focal glomeruloscerosis K&zey Znt 43, 1243-1250 Hammer, A., Kager, G , Dohr, G., Rabl, H , Ghassempur, I, and Jurgens, G (1995) Generation, characterization, and histochemrcal application of monoclonal antibodies selectively recognizmg oxidatrvely modified apo B-contammg serum hpoproteins. Arterzoscler Thromb Vast Biol 15,704-7 13 Preobrazhensky, S , Trakht, I., Chestkov, V , and Wentz, M (1995) Monoclonal antibody-based immunoassay for evaluation of lrpoprotem oxrdatlon Anal Blochem 227,225-234 Itabe, H , Yamamoto, H , Imanka, T , Shrmamura, K , Uchiyama, H , Klmura, J , Sanaka, T , Hata, Y , and Takano, T (1996) Sensitive detection of oxidatively moditied low density hpoprotem using a monoclonal antibody J Lzpzd Res 37,45-53 Kotam, K., Maekawa, M., Kanno, T., Kondo, A , Toda, N , and Manabe, M ( 1994) Distribution of unmunoreactive malondialdehyde-modified low-density hpoprotem m human serum. Blochzm Bzophys Acta 1215, 121-125. Hazel& L. J and Stocker, R (1993) Oxidation of low-density lipoprotein wtth hypochlorite causes transformation of the lipoprotem into a high-uptake form for macrophages Bzochem J 290, 165-172. Klemveld, H A , Hak-Lemmers, H L , Stalenhof, A F., and Demacker, P N (1992) Improved measurement of low-density-lipoprotem susceptibility to copper-induced oxidation. apphcation of a short procedure for isolating low-density hpoprotem C11n Chem 38, 20662072 Darrow, R A and Orgamsciak, D T. (1994) An improved spectrophotometric trnodide assay for lipid hydroperoxides Lzpzds 29,591-594 Ursmi, F (1987) The role of selenium peroxidases m the protection agamst oxidative damage of membranes Chem Phys Llprds 44,255-276 Chajbs, V , Sattler, W., Stultschmg, M , and Kostner, G. M (1996) Photometric evaluatton of lipid peroxidation products m human plasma and copper oxidized low density hpoprotems* correlation of different oxidation parameters. Atherosclerosis 121, 193-203. Sattler, W., Christison, J., and Stocker, R (1995) Cholesterylester hydroperoxtde reducing activity associated with isolated high- and low-density hpoprotems Free Radic Blol Med 18,421-429 Yamamoto, Y., Frer, B., and Ames, B N (1990) Assay of lipid hydroperoxldes using high-performance hquid chromatography with isolummol chemilummescence detection. Methods Enzymol 186,371-380 Kritharrdes, L., Jessup, W , Gifford, J , and Dean, R T. (1993) A method for definmg the stages of low-density-hpoprotem oxidation by the separatton of cholesterol and cholesteryl ester-oxidation products Anal Blochem 213, 7S89 Miyazawa, T. (1989) Determmation of phosphohpid hydroperoxides m human blood plasma by a chemilummescence assay. Free Radtc Blol Med 7,209-2 17

13 Separation and Quantitation of Phospholipid Classes by HPLC George M. Patton and Sander J. Robins

1. Introduction The major lipids of human lipoproteins are cholesterylester (CE), trtacylglycerol (TG), phosphatidylcholme (PC), cholesterol, and sphingomyelm (SM). Lrpoproteins also contain lesser amounts of glycosphingoliprds, phosphattdylethanolamme (PE), phosphattdylinosltol (PI), lyso-PC (LPC), and varrous forms of fat-soluble vitamins and cofactors (see Note 1). The proportion of these components varies over a wide range, depending on the type of hpoprotein. The PE fraction IS composed of roughly equal proportions of alkenylacyl and diacyl hprds. The PC fraction is composed primarily of dtacyl PC, but there are also small amounts of alkenylacyl and alkylacyl PC. However, because of the large amount of choline phosphoglycerrdes in hpoproteins, alkenyacyl, and alkylacyl PC are present in quantities roughly equivalent to that of the PEs and of PI. The purpose of this chapter is to describe hrgh-pressure liquid chromatography (HPLC) procedures to separate and quantitate the major phosphohpid species of lipoproteins. All the chromatography procedures described here are isocratic procedures, As indicated below, several of the procedures are readily amenable to gradient HPLC, but others are not, especially where quantitatron is by integration of peak areas. The initial fractionation of lipid classes IS performed by normalphase (silica) chromatography. The properties of silica columns vary widely depending on the manufacturer. All the normal-phase procedures descrrbed here use LiChrospher St 100 columns (see Note 2). Subsequent molecular species fractionation of the lipid classesis performed by reversed-phase (RP) chromatography. The RP procedures described here use primarily Altex Ultrasphere From

Methods m Molecular Btology, Vol 110 Llpoprotern Protocols Edhxl by J M Ordovas 0 Humana Press Inc , Totowa, NJ

193

Patton

and Robms

ODS columns. These columns have a unique selectivtty that makes them particularly useful for separating lipids based on their fatty actd content. Purified hpoprotems generally contam no lipid or nonhptd material that interferes with the lipid fracttonation. Therefore, a precolumn is generally not necessary. However, a screen, type, zero dead-volume filter should be installed between the mlector and the column as a precaution (see Note 3). A problem common to all lipid analysts is the spontaneous oxtdation of the unsaturated fatty acids. The oxidation of lipids by molecular oxygen mvolves the addition of oxygen across the carbon-carbon double bonds, prtmarily by free-radical mechanisms. Exposure to light, particularly ultraviolet (UV) light, facilitates the formatton of oxygen radicals. In addmon, m the presence of some transition metal tons (particularly Cu2+), apo B-contammg lipoprotems are subject to oxidation by a protein mediated process. Fortunately, lipoproteins contam natural antioxidants (i.e., compounds that oxidize more readily than fatty acids) (see Note 4). Once the fractionatton of the lipids begms, the endogenous antioxidants are separated from most of the other lipids and the problem of oxidation becomes more severe. Thus, the fractionation should proceed expedttiously. In addition, the purified lipids should always be kept under an inert atmosphere and stored m a few milhliters of an appropriate solvent, usually chloroform/methanol (2: 1, v/v) (see Note 5).

2. Materials 2.7. Equipment 1 2. 3 4 5 6 7 8. 9 10.

One or more HPLC systems with Integrator and variable wavelength detector Gas chromatograph (GC) or other equipment to quantitate TG and CE Two LiChrospher Si 100 (5 pm, 4 x 250 mm) columns One or two RP Ultrasphere ODS (5 pm, 2 x 250 mm) columns One Spherisorb ODS (3 or 5 q, 2 x 250 mm) column Screen filters 0.45 pm [or finer) filter system for mobile phases Screw-cap culture tubes with Teflon-lined caps. Low-speed centrifuge Nitrogen evaporator

2.2. Solvents 1 2 3 4 5. 6. 7.

Acetonitrile Benzene Chloroform Cyclohexane Diethyl ether Ethanol. Hexane.

Phospholipid Classes and HPLC 8 9 10 11

195

Methanol. Methyl t-butyl ether 2-Propanol Tetrahydrofuran

2.3. Chemicals 1e 2 3. 4

Acetlc acid (glacial) Benzolc anhydride. N,N-Dlmethyl ammopyrldme Nitrogen gas

2.4. Enzymes and Buffers 1 Baczllus cereus Phosphohpase C (Boehrmger Mannhelm, Type 1, Mannhelm, Germany) 2 B cereus PI specific phosphohpase C (Boehrmger Mannhelm) 3. B cereus Sphingomyelmase (Sigma, St. LOUIS, MO) 4 50 mM Potassium phosphate buffer, pH 7 0 5 50 mk! Tris-HCl, pH 7 4, and 1 mA4calcmm chloride 6. 50 mMBorate buffer, pH 7.0.

2.5. Internal 1 2 3 4 5. 6 7 8.

Standards

Cholesteryl heptadecanoate Dl-tetradecanoyl phosphatidylcholme (14.0-14.0 PC) Dl-hexadecenoyl phosphatldylcholme (16*1- 16.1 PC) Heptadecanoic acid (17.0). Heptadecanoyl lysophosphatldylcholine (17:0 LPC) N-Lignoceroyl-okdlhydrosphmgosme. Stigmasterol Tn-heptadecanom (17.0-l 7 0- 17 0 TG)

3. Methods 3.1. Quantitation The quantitation of liplds is difficult because there IS no simple and sensltwe means to quantltate all the components. The quantitatlon of the phospholipids and neutral lipids pose different problems. The major focus of this chapter IS the quantitation of iipoprotem phospholiplds. A number of automated and/or enzyme-linked assaysare available to quantitate the major neutral lipids (CE, TG, and cholesterol). To determine the fatty acid composition of the CE, TG, and free fatty acid fractions, gas chromatography 1sgenerally

the method of choice. Gas chromatography is also applicable to the diacyl and alkenylacyl phosphohplds, but it provides only partial Information about the alkylacyl lipids and sphmgomyelm. For the purpose of this chapter, we assume

Patton and Robins

196

that CE and TG will be quantitated by gas chromatography and the phospholipids by HPLC. Cholesterol can be quantitated either by GC, HPLC, or other appropriate methods. The first step in the analysis of lipoprotein lipids is to extract the lipids from the lipoprotems. To quantttate the lipids, internal standards are then added to the lipid extract before any further fractionation. The suggested internal standards used for the quantitation of each lipid class are indicated* Lipid class

Internal standard added

CE

Cholesteryl heptadecanoate Tri-heptadecanoate Stigmasterol (HPLC) or cholestanol (GC) 16:1-16 1 PC 17:o 17:o LPC

TG Cholesterol Phospholipids Free fatty acid LPC

There are at present no suitable commercially available internal standards for PE, PI, and SM. These phospholipids will be quantitated by adding a second internal standard, 14:0-14:0 PC (for PE and PI) or N-lignoceroy-mdihydrosphmgosine (for SM), to the phospholiptd fractions after they have been separated into individual lipid classes. 14:0-14:0 PC and ceramlde can be used as internal standards for the other phospholipid classes because the next step m the procedure is to convert the glycerophospholiplds to diacylglycerldes (DGs) and SM to ceramide before further fractionation. The DGs and ceramides are then benzoylated (mtroducmg a chromophore) and the PE and PC fractions are separated into dtacyl, alkenylacyl, and alkylacyl fractions, which are quantitated by integration of peak areas. The first internal standard (16. l-l 6: 1 PC) will determine the amount of dtacyl-PC; the second internal standard (14:0-14:0 PC) will relate the amount of the other diacyl-phosphohpids and SM to diacyl-PC.

3.2. Extraction of Lipids 3.2.1. Method of Folch The most reliable procedure to extract lipids from tissues and lipoproteins the method of Folch et al. (I).

is

1. An aqueous suspensron of the lipoprotein is transferred to a screw-capped culture tube. 2. Nineteen volumes of chloroform/methanol (2:l) are added to one volume of sample, i.e , 1 mL of sample plus 19 mL of chloroform/methanol (2: 1). 3. The sample is sealed under nitrogen and shaken vigorously. 4. Internal standards are added at this point.

Phospholipid

Classes and HPLC

197

5 The sample is allowed to sit at room temperature for several hours. 6 Then 0.2 vol (i.e., 4 mL) of saline (0.9% NaCl) are added. 7 After vigorous shaking, the sample is centrifuged (10 min) to separate the phases. 8 The upper phase (methanol/water) is discarded and the lower phase (chloroform) is transferred to a clean tube with a Pasteur pipet, being careful to avoid transferrmg any of the protein that partitions at the interface or any of the residual upper phase

3.2.2. Method of Bligh and Dyer When extracting large volumes, i.e., either dilute samples or large amounts of lipoproteins, the Folch procedure can be cumbersome. Therefore, with large volumes, the method of Bhgh and Dyer (2) is frequently used. 1. Two milliliters of an aqueous lipoprotem suspension are transferred to a 20-mL screw-cap culture tube and 7.5 mL of chloroform/methanol (1.2) is added 2. The sample is sealed under nitrogen and shaken vrgorously 3 If the lipids are being quantitated by internal standards, the appropriate internal standards are added at this point 4. Then, 2 5 mL of chlorofom is added, followed by vigorous shakmg. 5. Finally, 2.5 mL water IS added, the sample is again shaken vtgorously, and centrifuged to separate the phases. 6 As with the Folch extraction, the upper phase is removed and the lower phase is transferred to a clean tube, being careful not to transfer any of the protein at the interface

Alternatively, lipids can be extracted using hexane/2-propanol/water (3) instead of chloroform/methanol. However, to the best of our knowledge it has not yet been demonstrated that the procedure yields quantrtatrve recoveries. 3.3. Separation

of the Major Lipid C/asses (Fig. 1)

1 Column. LiChrospher Si 100 (5 pm), 4 x 250 mm When inittally conditionmg a column for this application, several days are required to completely equilibrate the column with mobile phase 2. Mobile phase* hexane/2-propanoll25 mM potassium phosphate buffer, pH 7.01 ethanol/acetic acid (369:475*56.100:0 1). Mix and let stand overnight When the organic solvents are added to the phosphate buffer, some of the potassium phosphate precipitates out. Therefore, this mobile phase must be filtered through a 0 45-pm (or finer) filter. 3 InJection solvent mobrle phase or hexane/2-propanol/water (6:s: 1) 4. Flow rate: 0.5-l .O mL/mm. 5. Detection: absorbance at 205 nm

Patton and Robins

198

PC

‘E I

I

0

I

I

30

f

60

TIME

1

96

I

120

150

(min)

Fig. 1. Separationof LDL lipid classesby normal-phaseHPLC The flow rate was initially 0.5 mL/min and was increased to 1 0 mL/mm as indtcated by the arrow NL, neutral lipid; PE, phosphatidylethanolamine;PI, phosphatidylinositol; PC, phosphatidylcholme, SM, sphmgomyelm,LPC, lysophosphatidylcholtne. 3 3.1. Procedure 1 The lipid extract is dried under nitrogen and dissolved in a small volume of injection solvent or mobile phase and all or part of the sample is injected onto the column. The flow rate is generally 1 mL/min, but when separating large amounts of lipoprotem lipids, especially very low-density lipoproteins (VLDL) or chylomicrons, rt may be necessary to use a lower flow rate (0.5 mL/min) to separate the neutral lipid (NL) and PE fractions. After the PE fraction is eluted, the flow rate can be increased to 1mL/mm. A representative chromatogram (low-density lrpoprotem [LDL] liptds) is shown

199

Phosphollpld Classes and HPLC

x 76

---I TIME

60

9'0

(min)

Fig 2 Accelerated separation of LDL lipid classes by normal-phase HPLC. Frfteen milliliters of water were added to 1 L of the mobile phase used in Fig. 1. Flow rate was 1.O mL/mm The elutron volume for LPC IS 1.6 times that of PC and IS not shown. Peaks are identified m the legend to Fig. 1

in Fig. 1. Although there are large differences in the lipid to protein ratio of the lipoprotems and in the amount of NLs, the relative proportion of the phospholipid classes of the major human lipoproteins is similar. 3.3.2. Procedure 2 If only the major lipid classesare to be quantitated, considerably less material would be required. Under those circumstances, the time required to separate the lipids can be reduced by adding more water to the mobile phase (Fig. 2). In the example shown m Fig. 2, 15 mL of water were added to 1 L of mobile phase after filtermg.

The amount of water added will depend on the amount of

Patton and Robms

200

PC

-0

lb

2’0

TIME

i0

4b

(min)

Fig. 3 Separation of the choline containing phosphohplds of LDL by normal-phase HPLC. Peaks are identified m the legend of Fig. 1.

material being injected, the particular characteristics of the column, and the resolutron required. The more difficult separations are resolving the NL from PE and PC from SM. Thus, this lipid-class separation procedure is amenable to gradient elutton using an mcreasing proportron of water. 3.3.3. Procedure 3 When only the choline contammg phosphohprds are to be separated and quantitated, an alternative procedure is available (Fig. 3). 1. 2. 3, 4. 5.

Column. LlChrospher Si 100 (5 pm), 4 x 250 mm. Mobile phase acetomtrile/methanol/water (560*330.60). Injection solvent: hexane/2-propanol/water (6.8: 1). Flow rate: 0.5 mL/mm. Detection* absorbance at 205 nm

201

Phospholipid Classes and HPLC

CE

Chol

J, I 0

I

,

10

20

30

TIME

_ I

1

40

50

60

70

(min)

Fig. 4 Separation of LDL neutral lipid classes by normal-phase HPLC. The neutral lipid fraction of LDL was isolated as shown m Fig. 1. CE, cholesterol ester; TG, triacylglycerol, Chol, cholesterol

As can be seen in Fig. 3, this procedure does not resolve the NLs from the glycosphmgolipids, PE or PI, but it does clearly resolve the two major phospholipids, i.e., PC and SM, and is the prefered method for the recovery of LPC. With this chromatography system a precolumn is generally advisable, because the acetonitrile leaches some material from the polytetrafluroethylene (PTFE) or polyethylene in the pump (piston seals),pulse damper, and most injectors, which results in severe tailing.

3.4. Separation of Neutral Lipid Classes (Fig. 4) 1. Column* LlChrospher Si 100 (5 pm), 4 x 250 mm. 2 Mobile phase: hexaneltetrahydrofitraniacetlc acid (500 2O:O.l)

202

Patton and Robins

3 Injection solvent. hexaneor mobile phase 4 Flow rate. 1 0 mL/min 5 Detection absorbanceat 205 nm 3.4.1. Procedure The NL fraction of ltpoprotems contains primarily CE, TG, and cholesterol (Fig. 4), but it also contains the antloxldants, vltamms, vrtamm esters, and free

fatty acids. If the NL fraction 1s obtained using the hexane/2-propanol/water system (Figs. 1 and 2), the NL fraction IS dried under mtrogen, and extracted by the method of Folch et al. (I) to remove the phosphate buffer. The dried lower phase 1sdissolved m hexane and applied to the column. The TG and CE fractions are collected, and dried under nitrogen. They are then ready for conversion to the appropriate derlvattve for gas chromatography. If the cholesterol remaining after derlvatlzmg the CE fatty acids Interferes with GC of the fatty acids, it can be removed by using the NL separation procedure with a moblle phase consisting of hexane/tetrahydrofuran/acetrc acid (500:50.0.1). The derlvatlzed fatty acids elute with the solvent front and the cholesterol elutes between 17 and 19 mm. The cholesterol can also be quantltated by GC (with cholestanol or other aproprlate internal standard), or by HPLC after conversion to the benzoyl ester as described below (with stlgmasterol as internal standard). If the NL fraction 1sobtained using the acetomtnle/methanoI/water system (Fig. 3), the dried NL fraction is dissolved m hexane and applied to the column. The glycosphingoliplds, PI, and much of the PE are not soluble m a small volume of hexane and, therefore, there will be considerable msoluble material which can be removed by centnfugatlon. Some of the PE and a trace of PI will be dissolved m the hexane and applied to the column. That material will not elute from the column in this solvent system, but will accumulate on the top of the column. As it accumulates, the resolution of the column will decrease. Therefore, at some point the column will have to be washed with acetonitnle/ methanol/water (or hexane/2-propanol/water) to remove the accumulated phosphohplds. After the column ts washed with a polar mobile phase, it 1snecessary to dry the column using a sequence of solvents to remove the water from the column. For this purpose, wash the column with at least 10 vol each of methanol, methanol/chloroform (1: l), chloroform, tetrahydrofuran, and mobile phase. The NL fractionation procedure 1sparticularly amenable to gradient elutlon usmg an increasing proportion of tetrahydrofuran. The limiting separation 1s that of CE from TG. Therefore, after the CE 1seluted, the proportion of tetrahydrofuran can be increased from approx 4-10% to elute the sterol fraction However, the absorbance at 205 nm will increase rapidly as the proportion of tetrahydrofuran Increases.

Phospholipid

Classes and HPLC

3.5. Phospholipid

Hydrolysis

203

and Derivatization

In order to quantitate phospholipids (and cholesterol) by HPLC, it is necessary to convert the phospholipids into a derivative, which improves the chromatographic properties of the lipids and provides a chromophore to permit quantitation by integration of peak areas. Although there are a number of schemes to derivatrze phospholipids, we routinely hydrolyze the lipids with phospholipase C or sphmgomyelinase and convert the resulting DG or ceramide to the benzoyl ester. The subsequent RP separation of molecular species is based on differences in the interaction of the fatty acids with the stationary phase (C18) and solubility in the mobile phase. Most chromophores are either bulky or highly polar moieties (especially those that have very high extinction coefficients) that have large effects on the partitioning of the lipids between the column and the mobile phase. The greater the effect of the chromophore, the less the influence of the fatty acids and the more difficult fractionation becomes. Benzoylation is generally an acceptable compromise between sensitivity and chromatographability. 1 The purified phospholipld classes, PE, PI, PC, SM, and LPC, are dried under mtrogen 2 If the phosphohpid classes were obtained using the hexane/2-propanollwater system, they should be extracted using the Folch procedure to remove the potassium phosphate 3 PE and PC are hydrolyzed using B cereus phosphohpase C (12U/mL) (BoehrmgerMannhelm, Type I) m 50 mA4potassium phosphate buffer, pH 7 0 PI IS hydrolyzed with B cereus PI specific phospholipase C (0.5 U/mL) (Boehrmger-Mannhelm) in 50 mMpotassmm phosphate buffer, pH 7.0 SM IS hydrolyzed usmg sphmgomyelmase (1 U/mL) (Sigma) m 50 mA4 Tris buffer, pH 7.4, and 1 mM calcium chloride. LPC is hydrolyzed using B. cereusphospholipase C (12 U/mL) m 50 mA4 borate buffer, pH 7.0. 4 To each phospholipid (m an 8-mL screw cap culture tube), add 0.5 mL of the appropriate hydrolysis reagent and 2 mL of dlethyl ether. 5. Seal the tube under mtrogen and shake on a vortex mixer for 1 min 6 Then, except for LPC, add 4 mL of hexane, shake again, and centrifuge to separate the phases. For LPC, add 4 mL of diethyl ether, shake, and centrifuge. 7. Remove the upper phase to a clean screw cap culture tube and dry under nitrogen 8 Repeat the extraction twice more with 3 mL hexane (or dlethyl ether for LPC) Using this procedure, PE, PI, and PC are converted to the 1,3-DG, SM to ceramide, and LPC to 1-monoacyl glycerol (MG) (see Note 6). The hydrolyzed phospholtptds (and cholesterol) are then derivatized by the method of Blank et al. (5).

Patton and Robins

204

1 To each sample of cholesterol, DG, MG, or ceramide m an 8-mL screw-capped tube, add 4 mg dtmethyl ammopyridme and 10 mg of benzoic anhydride m 0.3 mL benzene. 2. Seal under nitrogen and let stand at room temperature for 1 h 3 Then, dry the reaction mixture under nitrogen. 4. When dry add 2 mL of methanol/water (8 2) saturated with sodium carbonate and then 4 mL of hexane 5 Mix well, and centrtfuge to separate the phases 6. Remove the bottom phase (methanol/water) and discard. 7. Wash the hexane phase three more times with 2 mL of methanol/water/sodtum carbonate, and then twice with methanol/water (8.2). 8 Transfer the washed hexane phase to a clean tube and dry under nitrogen The derivatized

lipids can be stored in a small volume

(l-2

mL) of chloro-

form/methanol (2: 1). The derrvatrzed cholesterol, PI, SM, and LPC can then be subjected to RP chromatography to separate the molecular species from the internal standard for quantttation. However, if very small quantities of material are being analyzed, tt 1sfrequently advisable to further purify the samples by normal-phase chromatography before RP chromatography. 3.6. Purifying the Benzoylafed /Icy/glycerides by Normal-Phase HPLC (Fig. 5) 1 2 3 4 5.

Column. LiChrospher Si 100 (5 pm), 4 x 250 mm. Mobile phase hexane/tetrahydrofuran/acetic acid (500.20 0 1) InJectton solvent hexane Flow rate: 1 .OmL/mm. Detectton. absorbance at 230 nm

The derivatized acylglycertdes are dissolved in hexane and apphed to the column. The alkylacyl and alkenylacyl benzoyl DG if present would elute between 4 and 7 mm. The benzoyl DG (PI) elutes between 7 and 9 mm The dtbenzoyl MG (LPC) elutes between 9 and 20 min. 3.7. Purifying

the Derivafized

Ceramides

(Fig. 6)

1. Column. LiChrospher St 100 (5 pm), 4 x 250 mm 2 Mobile phase: hexane/tetrahydrofuran/acetic acid (500*35.0.1)

3. Injection solvent: hexane. 4. Flow rate: 1.O mL/min. 5. Detection. absorbance at 230 nm.

The procedure is the same as for the acylglycertdes. The SM m human hpoproteins and most other species contains only nonhydroxy fatty acids, which elute between 18 and 30 mm. Some species also contain SM with hydroxy

205

Phosphollpid Classes and HPLC

b

TIME

(min)

Fig. 5 Purification of benzoylated acylglycerldes by normal-phase HPLC Benzoylated acylglycerides were derived from LDL-PI as described m the text a-Alkenylacyl and alkylacyl;b-diacyl, c-dibenzoylatedmonoacyl(lysophospholipids)

fatty acids, which elute between 13 and 17 mm. The benzoylated mternal standard (N-lignoceroyl-m-dihydrosphmgosine, dlS.O-24:0) resolves mto two peaks m this system (Fig. 6A). The first peak elutes before the human hpoprotem SM (Fig. 6B) Therefore, the amount of SM can be quantitated at this stage by integration of peak areas. However, the lipoprotein SM peaks must be corrected for the amount of the second peak of the internal standard in the lipoprotem SM peak. 3.8. Separation of Diracfy/ Forms of Benzoyl PE and PC (Fig. 7) The ethanolamme phosphoglycerides of lipoproteins are composed of about 30% alkenylacyl and 70% dlacyl PE. The choline phosphoglycerides are composed overwhelmmgly of diacyl PC (95%), but also contam both alkenylacyl (1%) and alkylacyl (3%) PCs at levels comparable to the PEs and PI. These

Patton and Robins

206

B

-

c

IS

,u

1

TIME (minutes) Fig. 6. Purtficatton of benzylated ceramide by normal-phase HPLC (A) Benzoylated iV-hgnoceroyl-nL-dihydrosphmgosine (mternal standard). (B) Benzoylated ceramtde derived from LDL-SM with Internal standard

diradyl forms of PE and PC can be quantitated and purified by normal phase chromatography (6) 1. 2 3. 4 5.

Column LtChrospher Si 100 (5 pm), 4 x 250 mm. Mobile phase cyclohexane/hexane/methylt-butyl ether/aceticacid (375:125 IO.0 1) Injection solvent: hexane. Flow rate. 1.0 mL/mm Detection. absorbanceat 230 nm

The relative amount of each diradyl form is determined by integration of the peak areas. However, the dlacyl peak also contams the internal standard(s) whose areas must be subtracted from the dracyl peak(s) after the proportron of

internal standard(s) IS determined by RP chromatography. 3.9. Separation of Benzoylated Lipids into Molecular Species The benzoylated lipids can be separated mto molecular species and quantltated by integratron of peak areas by RP HPLC. All the various forms can be

207

Phospholipid Classes and HPLC

A

C

B

T--

a C

!2

0

10

L-

AlLiL

I, 20

30

40

I30

30

70

0

10

20

30

40

30

TIME (minutes) Fig. 7. Separation and purification of benzoylated dtradyl forms of LDL-PE and PC (B) by normal-phase HPLC a-alkenylacyl, b-alkylacyl, c-diacyl.

(A)

separated usmg the same column(s) and mobile phase (methanol/acetonitrile/ water), but in general it IS more convenient to use slightly different mobile phases for the various forms of lipids, i.e., cholesterol, dlacyl, alkylacyl, alkenylacyl, and ceramlde. 1. Column. Ultrasphere ODS (5 pm), 2 x 250 mm 2 Mobile phase: Lipid class

Mobile phase

Benzoyl dtacyl lipids (Fig. 8)

Methanol/water/acetomtrile (94 25:4.0.1.75) Methanol/water/acetomtrile (94 75:3.5*1.75) Methanol/water/acetomtrile (95.25:3 0 1 75) MethanoVacetomtrile (50:50) Methanol

Benzoyl alkenylacyl lipids (Fig. 9) Benzoyl alkylacyl lipids (Fig. 10) Benzoyl ceramide (Fig. 11) Benzoyl cholesterol (Fig. 12) 3 InJection solvent Methanol 4 Flow rate’ 0.3-O 5 mL/mm 5 Detection absorbance at 230 nm

9

\I f 0

1

2

3

4

5

6

7

9

C IS

II

0

1

2

3

(hour:1

4

TIh4E Fig 8

6

7

Phosphollpid Classes and HPLC

209

Table 1 Identification of the Major Molecular of Glycerophospholipidse

Species

Peaknumber

Molecular species

1 2 3 4 5 6 7 8 9 10 11

16-O - 22.6 18:l - 22.6 16:0 - 20.4 18:l - 20.4 16:O - 18:2 18.1 - 18 2 18:0 - 22 6 16 0 - 18.1 18:O - 20.4 18:0 - 18 2 18:0 - 18.1

OSee ref. 7 for a detailed ldentlficatlon phosphohpld molecular species

of glycero-

For quantitation of the phospholiplds and cholesterol, a single RP column can be used, and only the cholesterol and dlacyl fractions need to be separated. The elutlon time (and resolution) can be reduced by decreasing the amount of water m the mobile phase. However, the resolution has to be sufficient to resolve completely the internal standard(s) from all other components To determine the molecular species composition of the lipid classes, two columns connected in series are generally required to adequately resolve the major molecular species. However, even with two columns, the major molecular species are contaminated with minor components. When necessary, almost all of these minor components can be separated by rechromatography of the fractions on a Spherisorb ODS (3 rim) column, 2 x 250 mm, with acetonitrlle as the mobile phase (Fig. SC, insert). Acetonttrile forms a complex with the double bonds of the fatty acids, which increases their solubllity in the mobile phase and, therefore, decreasesthe retention time. Thus, molecular

Fig. 8. (prevzouspage) Separation of benzoylated dlacyl molecular species of LDL-PE (A), PI (B), and PC (C) by reversed-phase HPLC The numbered peaks are tdentlfied In Table 1. The insert m (C) shows the separation of peak 9 into Its two major components, 18:0-20 4 and 18: 1-l 8: 1 by RP HPLC on a Spherisorb ODS

(3 p) column with acetonitrlle as the mobile phase (0.3 mL/min) as described m the text.

210

Patton and Robins

TIME (hours) Fig. 9 Separation of benzoylated alkenylacyl molecular species of LDL-PE and PC (B) by RP HPLC. The numbered peaks are identified m Table 1.

(A)

species with the same polartty can be separated based on the number of double bonds. With pure acetomtrile as the mobile phase, chemical degradatron of alkenylacyl ltptds 1ssometimes observed. Therefore, for alkenylacyl ltptds, the mobile phase 1sacetomtrtle/2-propanoI/water (90:8.5: I .5). When using 3-q particle size columns, a precolumn (3-p particle size) is generally necessary to avoid a raprd increase m back pressure. Those molecular specres that contam any given fatty acid and 16.0 or 18.1 (i.e., 16:0-18:2 and 18: l-18:2) are reasonably well-resolved by the Ultrasphere ODS column (Figs. S-10). On a Sphertsorb ODS column (and most other ODS columns), these molecular species would elute as a single peak. Usmg the Ultrasphere ODS column with acetomtrtle as the mobile phase, the 16:0- 18:2

Phospholipid

Classes

0

1

211

and HP1 C

2

3

4

5

6

7

6

s

TIME (hours) Fig. 10. Separation of benzoylated alkylacyl molecular RP HPLC. The numbered peaks are identtfied in Table 1

species of LDL-PC

by

and 18: l- 18:2 elute as a single peak. However, on the Spherisorb ODS column with acetonitrile as the mobile phase, the 18. l- 18:2 elutes before the 16*0- 18 2. Thus, by taking advantage of differences in the selectivity of the columns, virtually all the components of a given fractton can be resolved. The only exceptions are for those pairs of molecular species that contam the same number of carbon atoms and the same number of double bonds of the same family, i.e., 18:0-20:4 (n-6) and 16:0-22:4 (n-6). Pairs of molecular species with the same number of carbons and double bonds in which the unsaturated fatty acids are m different famtltes (such as 18:0-20:5 [n-3] and 16:0-22:5 [n-6])

can be resolved. 4. Notes 1 It ts not altogether clear whether free fatty acid and LPC, which are present only m trace amounts in hpoprotems, are actually components of the hpoprotems m vivo or whether they are formed during isolation of the lipoproteins, or are bound to albumin remainmg m the lipoprotein fractions 2 Owmg to the difficulty of equilibrating slhca columns when using mobile phases of widely different polarity, practtcally speaking the same column cannot be used for the separation of phosphohpid classes and neutral lipid classes. Therefore, at least two silica columns are required. 3 As a further precaution, all mobile phases except those prepared exclusively with HPLC grade solvents should be filtered through a 0.45~pm (or finer) filter.

Patton and Robins

I

1

I

1

I

1

1

2

3

4

5

6

TIME (hour) Fig. 11. Separation of benzoylated ceramide molecular species of LDL-SM by RP HPLC The numbered peaks are identified in Table 2. IS, internal standard (N-lignoceroyl-DL-dihydrosphingosme)

4 To mirnmtze hpid oxidation during the isolation of lipoprotems, whenever possible limit exposure to oxygen by keeping the lipoprotems under an inert atmosphere (nitrogen or argon), use EDTA to chelate transition metal ions, and keep the ltpoprotems cold and out of the light 5. Never store lipids dry. The organic solvents used to extract and fractionate lipids dissolve most plastics and leach a number of compounds from virtually all plastics Therefore, use only glass ptpets and screw capped glass tubes with Teflon or PTFE-lined caps 6 When extracting the MG and DG, be very careful not to transfer any of the aqueous phase and do not heat the hexane (ether) phase when drying under nitrogen. Water and other protic solvents (mcludmg alcohols and chloroform) facilitate the isomerization of the MG and DG. Isomertzation of the glycerides

Chol

Stig

TIME (min) Fig. 12. RP HPLC of benzoylated cholesterol (Chol) and the internal standard, stigmasterol (Stig)

214

Patton and Robins Table 2 Identification of SMa Peak number 1 2 3

4 5 6 7

8 9

10 11 12

of the Major Molecular

Species

Molecular species d18:l - 14.0 d18 1 - 15 0 d18.1 - 16.0 dl8 1 - 17 0 d18.1 - 18.0 d18.1 - 24.2, d18:2 - 24.1 d18.1 - 20:0 d18.1 - 21.0 d18:l - 24 1, d18 2 - 24 0 d18.1 - 22 0 d18 1 - 23 0 d18 l-240

‘ISee ref. 8 for a detailed analysts of the molecular speciesof LDL-SM

makes subsequent analysis of the lipids extremely dlfflcult, because the positional isomers are resolved into separate peaks by the RP procedures. Therefore, if the hydrolyzed lipids must be stored before denvatlzatlon, store them m hexane References 1 Folch, J., Lees, M., and Sloane-Stanley, G. H. (1957) A simple method for the isolation and purification of total lipids from animal tissues. .I Biol. Chem 226, 497-509 2 Bligh, E G. and Dyer, W. J (1959) A rapid method of total llpld extraction and

purification Can J Bzochem Physlol 37,911-914. 3 Radm, N. S. (1981) Extractlon of liplds with solvents of low toxlclty Methods Enzymol. 72,5-7 4 Jungalwala, F B , Evans, J. E., and McCluer, R H. (1976) High-performance llquld chromatography of phosphatidylcholme and sphmgomyelm with direct detection m the region of 200 nm. Blochem J 155,55-60. 5 Blank, M L , Robinson, M., Fitzgerald, V., and Snyder, F (1984) Novel quantitative method for determination of molecular species of phospholipids and dlglyerides. J Chromatogr 298,473-482 6 Nakagawa, Y and Horrocks, L. A. (1983) Separation of alkenylacyl, alkylacyl, and diacyl analogues and their molecular species by high performance llquld chromatography. J, Lzpzd Res 24, 1268-1275.

Phospholipid Classes and HPLC

215

7 Patton, G M. and Robins, S. J. (1987) HPLC of molecular spectes of glycerophosphohptds m studies of lipoprotems and lipid transport, m Journal of Chromatography Ldmzry, vol 37 (Kuksis A., Elsevier, ed.), NY, pp 3 1 l-347. 8. Myher, J J., Kuksis, A., Shepherd, J , Packard, C. J., Mortise& J. D., Taunton, 0 D., and Gotto, A. M (198 1) Effect of saturated and unsaturated fat diets on molecular species of phosphattdylcholme and sphingomyelin of human plasma hpoprotems Bzochzm Biophys. Acta 664, 110-l 19.

Assays of Lecithin Cholesterol (LCAT) Milada Dobi&sova

Acyltransferase

and Jiri J. Frohlich

1. Introduction Lecithin cholesterol acyltransferase (LCAT, EC 2.3.1.43), a plasma enzyme secreted by hepatocytes, catalyzesthe transfer of acyl group of fatty acids from the 2-sn position of lecithin to the 3-hydroxy group of cholesterol. Ordmarily the reaction takes place mostly on the surface of high-density lipoprotems (HDL). In human plasma, thts reaction is the major source of cholesteryl esters (CE) and has a crucial role in the remodelmg of plasma lipoproteins and in reverse cholesterol transport A number of revtews describe the properties, mechanism of action, and methods of LCAT determination (1-5). Whereas there is a definite potential for the clinical use of LCAT assays,its use 1snot widespread, being now ltmited to the diagnosis of LCAT deficiency syndromes, such as LCAT deficiency and Fish Eye Disease. After almost 30 yr of efforts to reach consensus on the best assessmentof LCAT activity and/or the rate of cholesterol esteritication in plasma, there is still considerable ambiguity and mtsunderstandmg surrounding the assays used today. It 1s of particular importance to find a method that reflects the physiologtcal process and indicates the amount of produced CE. The LCAT reaction in plasma consists of three phases. Each of them can be influenced by reaction conditions during rn vitro assay and thus affect the fmal result. The stages of the reaction are as follows: 1. Activation of the phosphohpid bilayer by a protein or pepttde. 2. Releaseof a fatty acid by hydrolysis of lecithin (mostly in the ~2-2 position) by phosphohpaseA2-like activrty. 3. Transfer of fatty acyl to the acyl acceptor,3l3-hydroxygroup of cholesterol. From

Methods KI Molecular Edlted by J M Ordovas

Biology, Vol 110 Lpoprotern Protocols 0 Humana Press Inc 1 Totowa, NJ

217

218

Dobi&ov&

and Frohlich

A reverse LCAT reaction with independent acylation of lysoleclthm to lecithm (acyl donor is lecithin and the activator 1sLDL) is less Important because of its low rate under physlologlcal conditions. The activation of the phosphohpid bilayer (the first stage of the reaction) that takes place primarily in HDL IS mostly owing to the presence of apollpoprotem AI (apo AI). This activation by apo AI 1soptimal only if HDL particles with apo AI, i.e., LpAI are involved (m the absence of apo AII). In addition, the suitability of these particles as substrates for LCAT 1sdetermined by their size and, possibly, their charge (6-Q. The esterification rate depends on the composition of lecithin, particularly on the acyl-chain lengths and their degree of saturation, which in turn determines (in a way similar to the ratio of phosphollplds to cholesterol) the optlmum fluidity of the phosphohpid bilayer and thus the optimal physical chemical state for the reaction. Availability of unesterlfied cholesterol (UC) and its transfer from various compartments of the plasma pool may also affect the esterificatlon rate. Despite the fact that the reaction takes place on the surface of HDL particles it has been proven that practically all the cholesterol substrate for LCAT originates from very low-density lipoproteins (VLDL) and low-density lipoprotein (LDL) particles (9,10) The current methods of measuring LCAT can be divided into those estlmatmg LCAT protein (mass) in plasma or LCAT activity (using artificial substrates or heat-inactivated plasma) and those measuring the rate of cholesterol esterification in plasma using the subJect’s own plasma or plasma depleted of apo B-containing lipoproteins (HDL-plasma). It 1s important to realize that each of the aforementioned methods yields different mformatlon: Some reflect the mass of the LCAT protein; others, such as the esterlficatlon reactions m whole or HDL-plasma, reflect the effects of various lipoprotein particles on the reaction rate (see Note 1). Because Albers et al (1,11) has been published, which is the best available assayof LCAT protein mass(the competitive double-antibody radlolmmunoassay) we will not further elaborate on it. The focus of this chapter 1son the most reproducible and most accepted methods m estimation of LCAT activity and cholesterol-estenfication rate with emphasis on the methods that have been used in clinical studies and yielded results consistent with our understanding of the LCAT reaction and lipoproteins as risk factors for atherosclerosis. 1.1. LCAT Assays 1.1.1. LCATActuty LCAT activity refers to the activity of the enzyme measured independently of its natural (autologous) substrate. LCAT activity is expressed as the molar

219

Assays of LCA T

esteritication rate (MER) and under ideal circumstances represents the quantity of UC esterif’ied by the source of LCAT (m most casesa given volume of plasma or plasma devoid of LDL and VLDL particles) in units of time (nmol/ mL/h). However, MER can also be expressed as nmoles of cholesterol esterified per mg of LCAT protein per hour, i.e., specific activity. Activity of LCAT is usually measured by counting the amount of UC labeled with 3H or 14Cthat has been mcorporated into CE. The substrates most frequently used are welldefined liposomes composed of egg yolk lecithin, apo AI purified from human plasma, and radiolabeled 3H or 14CUC (12). Because heat-inactivated plasma has limited stability and low reproducibility, this source of the reaction substrate has been mostly abandoned. A number of laboratories have investigated the most appropriate substrate and labeling procedures for the LCAT reaction. Proteoliposomes composed of apoA1, lecithin, and cholesterol prepared by cholate-dialysis technique have also been employed (13). However, m our and others’ experience the best-tried and best-defined methods use labeled lecithin-cholesterol ethanolosome substrate and human apo AI as an activator of the reaction (see Note 2). 1.1.2. Cholesterol-Esterification

Rate (CER)

Esterification rate of cholesterol is the term use to describe the LCAT reaction in an autologous substrate (see Note 3). Theorettcally, the measurement of the esterification rate m plasma, a natural substrate containing both the subject’s enzyme and lipoprotein particles, should best reflect the process m vivo. The measurement can be carried out using two distmct methods (or their modifications). In the first one, labeled UC is mcorporated mto lipoproteins of whole plasma and the amount of the label in CE after the incubation period is measured (14). The second method simply estimates (using enzymatic or gasliquid chromatography [GLC] assay) the difference m plasma concentration of UC before and after incubation at 37OC (25). Both of these approaches have their limitations. The radioassay assumesa complete equilibration of the label with UC of the plasma lipoprotein bilayers and an equal rate of exchange of UC among various lipoprotems, whereas the measurement of UC mass is limited by the very small differences in the pre- and postincubation samples and the coefficient of variation of the assaysof UC currently in use. The CER used to be expressed as the fractional esterification rate (FER), i.e., percentage of increase of 3H-CE/h of incubation or percentage of decrease in UC in the mass assay. Another value, which takes into account the subject’s actual concentration of UC, is the MER or, more appropriately, CER, i.e., nmol CE esterified m a mL plasma/h (nmol CE/mL/h). It is calculated by multiplymg FER with the concentration of UC or, if mass is measured, simply as a decrease m nmol UC m 1 mL/h. Confusion still exists m the use of the terms “MER” and “activity of

220

Dobi&ov$

and Frohlich

LCAT” in assays using either artificial or autologous substrate. We strongly recommend the use of the term CER only to describe sltuatlons m which the subject’s plasma is used for the assay;the terms “MER” and “activity of LCAT” should be preserved for LCAT activity estimation m artificial substrate. 1.1.3. FER in HDL-Plasma (FE&& Doubts about the physiological significance of the aforementioned measurements of LCAT protein, LCAT activity, and CER lead to a search for a more relevant type of assay (see Note 4). We considered the following issues: 1. That the bulk of cholesterol esterlficatlon occurs on the surface of HDL particles 2 That larger (HDLZb) particles are known to inhtblt, whereas the smaller ones (HDLsb,,) increase the rate of esterification (26,Z 7) 3. That LDL 1sthe source of most of UC wlthm the a-migrating HDL (IO). To separate the effects of the apo B-containing lipoproteins, and m partlcular to avoid the flow of UC from LDL and of HDL particle size, we measure the FER (I&19) m apo B lipoprotem-depleted plasma (HDL-plasma). In the absence of VLDL and LDL, the other plasma proteins (particularly albumin) facilitate the diffusion of UC among lipoprotein particles (3) and also bmd lysolecithm, thus preventmg product mhlbltlon of the reaction We assume that FERHDL reflects the ability of the mdlvldual HDL-LCAT complex to esterify UC that has been transported by diffusion from surfaces of other hpoproteins.

2. Materials 2.1. Measurement 2.7.7. Reagents

of L CA T Activity

1 Egg yolk phosphatldylcholme (EYPC): (Sigma, St LOUIS, MO) Type II-E, 5 mg/mL m absolute ethanol, store at -20°C. 2. UC: (Sigma) CH-S, 1 mg/mL m absolute ethanol; store at 20°C. 3 [7(n) 3H]-cho1estero1 (specific activity 5 Wmmol) was purchased from Amersham (Amersham, UK) 4 Purified apo AI at a concentration of 1 mg/mL m 0.15 A4 NaCl, 1 mA4 EDTA, 0.03% azlde Store at 4°C (see Subheading 3.1. for isolation procedure). 5. Thm-layer chromatography (TLC) plates (plastic sheets silica gel 60,,,, layer thickness 0.2 mm) (Merck, Darmstadt, Germany) 6 Assay buffer: 10 mA4 Tris-HCl, pH 7.4 (1 211 g/L), 5 mM ethylenedlamme tetra-acetic acid (EDTA) (1.861 g/L), 0.15 MNaCl(8 783 g/L). Store at 4°C 7 0 1 A4 P-Mercaptoethanol. 35 pL in 5 0 mL of assay buffer (make fresh before each assay) 8. UCXE standard. 20 mg cholesterol and 20 mg cholesteryl oleate m 10 mL chloroform

221

Assays of L CA T 9. Quahty control (QC): aliquots of 0.5 mL of normolipidemic -20°C for 3 mo 21.2.

Preparation of Apo A-l by Chromatofocusing

plasma stored at

(20)

2 1 2 1. MATERIALS 1 2. 3 4. 5 6. 7

Bio-Rad econocolumns 5 x 75,15 x 75, 1.5 x 10 cm. Pertstaltic pump Fraction collector 13 x 100~mm Test tubes Rexyn I-300 (Fisher Scientific, Pittsburgh, PA, No. R-208-500). Brogel HTP Hydroxyapatite (Biorad, Hercules, CA, No 130-0420). Polybuffer exchanger (PBE) 94 200 mL (Pharmacia, Uppsala, Sweden, No 170712-01). 8. Polybuffer 74 250 mL (Pharmacta No. 17-07 13-O 1) 9. Crude urea. 2.1 2.2. STOCK SOLUTIONS 1 10X Concentrate tmidazole/HCl buffer 0 25 M, pH 7 4 (4 26 g/250 mL). 2 8 A4 Crude urea (480 g/L) purified by passage down rexyn I-300 column at least twice and must be used within 48 h 3 Equilibration buffer: 0.025 M imidazole, 7.2 M urea, 1 mA4 EDTA, pH 7 4 (50 mL 0 25 A4 imidazole, 450 mL 8 Murea, 0 186 g/L EDTA). 4 Solubilization buffer: 0 025 A4 imidazole, 7 2 M urea, 20 mM dithiothreitol (DTT), pH 7.4. 5 Elutton buffer. 1 part Polybuffer, 74 to 8 parts 8 M urea, pH 4 0, and degas (111 mL polybuffer, 889 mL 8 Murea, and adJust the pH to 4 0 with HCl). 6. HTP equthbration buffer. 0 01 M NaH2P04, 0.1% SDS, pH 7.2 7. HTP elution buffer: 0.5 MNaH2P04, 0.1% sodium dodecyl sulfate (SDS), pH 6.8.

2.2. Measurement of FER in HDL-Plasma 2.2.1.

Reagents

1 [7(n) 3H]-cholesterol (spec. activity 5 Ci/mmol) (available from Amersham, UK) 2. The kit for assay of UC is available from Wako (Richmond, VA). 3 Solvent for liquid scintillation counting fluid (PPO scmtillator, available from Fisher Scientific, Vancouver, Canada). 4 Whatmann no. 1 filter paper. 5 TLC plates (plastic sheets silica gel 60 rZs4layer thickness 0.2 mm) were available from Merck, Darmstadt, Germany. 6 Ethanol (98%). 7. Petroleum ether. 8. Diethylether. 9. Acetic acid. 10. Iodine crystals.

DobiAsovA and Frohlich

222 All chemicals were of analytical grade.

11. Buffer Trts-saline (TBS): 10 mM Tris-HCl, pH 7.4 (1 21 lg/L), 5 mM EDTA (1 861 g/L), 0 15 MNaCl(8.783 g/L) Store at 4’C 12 Phosphotungstate solution (PTA)* Dissolve 4 g phosphotungstic acid and 16 mL 1 MNaOH m a total of 100 mL detomzed water 13, 2 M MgC12 solutton m detomzed water. 14 QC* Aliquots of 0 5 mL of normoliptdemtc plasma are stable if stored at -20°C for 3 mo. Duplicate of QC are precipitated and their FERun, estimated m each assay run

3. Methods 3.1. Preparation

of Apo A-l by Chromatofocusing

(20)

1. Isolate VLDL and HDL by sequential ultracentrtfugatton at den&es 1.006 g/mL for VLDL and 1 085-l 210 g/mL for HDL, as described m Chapter 6 and ref. 21 2 Purify all hpoprotems by a second ultracentrifugation step at the upper density limit and dialyze extensively against 0 15 M sodmm chloride containing 1 n&f EDTA and 0.03% sodium azide. Total protein was measured by the method of Lowry et al (22) using bovine serum albumin (BSA) as standard and chloroform extraction to remove turbidity (23). 3. Deltptdate with ethanol-diethyl ether (3: 1) and dry the final pellet under a stream of nitrogen (24). If the protein pellet 1s not going to be used munedtately, then store at -20°C until use 4. Solubtltze the apoprotem pellets m 10 mM Trts-HCI, pH 7 4, containing 7 2 A4 urea and 10 mA4 DTT. 5 Stir for 16 h at 4°C 6. Remove insoluble apo B aggregates by centrifugation (5OOOg,20 mm) and adjust the supernatant to 10 mg/mL protein with solubtltzmg buffer 7. All urea-containing soluttons must be prepared from twice-detomzed (Rexyn 1300, Fisher Scientific) 8 M urea, stored at 4°C and used within 48 h of preparation Chromatographtc procedures are performed at 4°C. 8. Pour and eqmltbrate the column Mix approximately 100 mL of PBE 94 gel with 100 mL of equtlibratton buffer overnight in a vacuum flask at 4’C 9. Discard approx 50 mL of supernatant and degas the slurry for 1 h 10 Pour the slurry mto the column and measure the pH of the eluant 11. Wash with equtltbration buffer (minus the urea) at approx 50 mL/h until the eluant is at pH 7 4 12. Once the column 1s equilibrated to the correct pH value, run approx 200 mL of equtbbratton buffer with the urea added before the sample is applied 13 Solubthzatton of the sample. The apo HDL sample, after dehptdation and removal of any residual solvent wtth nitrogen gas, can be solubthzed m solubthzation buffer at approx 5 mg/mL. After stirring overnight if possible at 4°C the sample may be spun at 5000g for 20 mm to remove any aggregation that may have occurred The time between solubilization of the sample and column loading

223

Assays of LCAT

14. 15

16 17.

18 19. 20 21 22 23

should be kept to a mmimum to maximize resolution. A small ahquot of the sample should be kept back for future purilicatton analysis. Running the column: Load 5 mL of elutton buffer just prior before loading the sample. Load the sample itself and then continue with more elutton buffer. The column may be run as high as 50 mL/h, collectmg 150 drops/tube (about 4.5 ml/tube) for about 240 tubes. This works out to be approx 1 L for the entire gradient to be run Recovery of apoprotems. Measure the optical densities (ODs) at 280 of the fractions and measure the pH of about every fifth fraction. Pool the protein peaks Equilibrate the Biogel HTP (5-10 mg protein/g HTP) with equilibration buffer m a small econocolumn. All this column work should be done at room temperature Apply each focused protein peak directly to the column and check eluant for protem (OD at 280). If there is a substantial amount of protein eluting from the column at this point, reload the sample. After the sample has been loaded, wash the column with 100 mL of the equilibration buffer until the OD readings are very low. Elute the protein wtth the elution buffer, collectmg 2.5-mL fractions, 85 drops/ tube at 25 mL/r. Measure the ODs at 280 and pool the protein peaks Measure the protein usmg the Lowry method Peaks should be dialyzed agamst standard dialysts buffer 1 x 4L at room temperature for mmimum of 4 h and then 2x4Lat4”Cfor4heach. Analysis and storage of apoproteins. Analyze each protein peak for purity on both sodmm dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and IEF gels. Store pure protems m 1-mg ahquots in the freezer. Regeneration of the column Wash the column with 200 mL 0.025 M imidazole buffer, 1 A4 NaCl (29 44g/L), pH 7 4, and wash with 0.025 M imidazole alone until pH of the eluant is back to pH 7 4

3.2. Production

of Ethanolosomes

1. Dry the followmg mtxture under nitrogen: 260 pL EYPC, 150 @L UC, and 12 uL 3H-UC (lmCl/mL) in a 12 x 75-mm test tube. This gives a molar ratio of PC*UC of4:l. 2. Dissolve the dry residue m 125 pL of absolute ethanol Take up the solution m a 1-mL syringe wtth a small-gage needle (>25) and inject rapidly m 10 mL of assay buffer. Vortex well. 3. Concentrate the solution on an Amicon YM-30 membrane to a volume of no more than 2.0 mL. Adjust to a final volume of 2.5 mL assay buffer. This solution should be clear or only very slightly turbid. 4. Store ethanolosomes at 4°C and use within 14 d

3.3. Measurement

of LCAT Activity

1. Run duplicates for each sample, including blank and QC. 2. After determining the number of samples to be run, prepare a mix of hposomes (ethanolosomes) (30 l&/sample), apo A-I (5-10 pL/sample; the actual amount

Dobitisova and Frohlich

224

will vary from batch to batch, 1.e , use that amount associated with optimal acttvation), and assay buffer (SO-85 pL/sample; volume will depend on amount of apo AI, or volume of sample to give final volume of 125 pL). Calibrate each new batch of apo AI using the QC samples, i e , use 4-15 uL to determine the optimal concentration of apo AI to achieve the maximal rate of LCAT reaction 3 Plpet equal volumes of the mixture mto test tubes (125 pL /sample) and Incubate at 37°C for 30 mm Leave for 30 mm at room temperature. 4 Prepare a solutton contaming 5 parts of 8% BSA solution with 1 part of the 0 1 M S-mercaptoethanol sufficient for all samples. Pipet 60 pL into each tube. 5 Finally, pipet into each tube 15 pL of the sample and mcubate at 37°C for 30 min

3.3.1. Extraction and TLC Separation 1 After mcubation, the tubes are placed mto crushed ice and the contents mixed immediately with 1 5 mL of 98% ethanol. 2. The mixture is stirred and left standmg for 2 h at room temperature 3 Samples are then centrifuged at 15OOg for 10 mm and the supernatants are dried in a stream of air at 40°C 4. After samples are dry, add to each sample 50 p.L chloroform and 10 pL UCiCE standard to make vtsuallzation of labeled components m the diluted sample possible 5 Apply 6 samples and 10 pL of UC/CE standard to one TLC plate 20 x 10 cm m a fume hood 6 Develop plates in a glass tank contaming a mixture of petroleum ether-diethyl ether-acetic acid (105/18/l 5 v/v/v) till solvent front reaches the top lute (this usually takes 8-I 0 mm) 7 Dry the plates m fume hood at room temperature 8 Place the dry TLC m a glass tank containing iodme crystals until the spots are visible Outlme the contours of the resultmg spots with pencil; any remammg traces of iodme are then allowed to subhmate m the hood The spots are then cut out with scissors and put mto scintillation vials.

3.3.2. Measurement of Radioactivity The toluene scintillatton fluid (PPO scinttllator, available from Ftsher SCIenttfic, Vancouver, Canada) is poured over the cuts from TLC, shaken, and allowed to stand for at least 3 h before countmg tn p-scmttllation counter

3.3.3. Calculations FER = %CE/h = (% CE,,, ,&

x2

MER = nmol of cholesterol esterified /mL/h = FER x 4.66 nmol of UC x lOOO/vol of sample in pL

Assays of LCA T

225

3.4. Measurement of FER in HDL-Plasma 3.4.1. Blood Collection Blood should be collected into EDTA-containing tubes after a 12 h overnight fast, kept at 4’C, and centrifuged within 2 h at 175Og for 10 mm to separate cells from plasma (see Notes 5-7).

3.4.2. Preparation of HDL-Plasma (18) 1 2 3 4

To 500 pL plasma m a 1-mL plastic centrifugatton vial, add 50 pL PTA solutton Stir and add 12.5 pL MgC& solution Stir again and allow to stand for 20 min at 4°C. Centrifuge the suspension for 20 min at 10,OOOgat 4°C

3.4.3. Labeling of Lipoproteins with 3H-Cholesterol 3.4.3.1.

PREPARATION OF 3H-UC DISKS (78)

1 Dry 10 $ of solution of [7[n]-3H]-cholestero1 (250 pCr m 250 pL of toluene) under N, to remove traces of toluene 2 Dtssolve the residue m 400 pL of ethanol 3. Apply 3 pL of this solution to the paper disks (5 mm m diameter) 4 Cut out the disks from Whatman no. 1 filter paper usmg a letter punch and place them horrzontally on thin hypodermic needles fixed on a stand 5 After evaporation of the solvent, place the disks separately m mdtvtdual stoppered 12 x 75-mm glass test tubes and keep m a refrigerator, Preprepared dtsks can be used for at least 3 mo. 3.4.3.2.

LABELING OF LIPOPROTEIN SAMPLE

1. TO 75 $ of TBS add 50 pL of precooled (4’C) HDL- plasma (or plasma or other lipoprotein solution) and immerse the 3H-cholesterol labeled paper disk mto each solution 2 Close the tubes with Para-film and leave them on ice overnight Spontaneous transfer of the label proceeds at low temperature and Iabelmg homogeneity IS attained after 18 h (19). 3. Run duphcates for each sample including QC. 3.4.3.3.

SAMPLE INCUBATION AND EXTRACTION

1. Discard the disks with remaining 3H-cholesterol label (usmg forceps), 2 Place the test tubes with labeled HDL-plasma m a shaking water bath and mcubate at 37°C (see Note 8) for 30 min (see Note 9). 3. Place the tubes on ice, and rmx the contents immedrately with 1.OmL of 98% ethanol 4. Stir the mixture and leave standing for 2 h at room temperature. 5. Centrifuge the samples at 15OOg for 10 mm. 6 Dry the supernatants by evaporation in a stream of air at 40°C

Dobi&ovA

226

and Frohlich

3.4 3.4. TLC SEPARATION AND MEASUREMENT OF RADIOACTIVITY Follow the procedure described m 3.3.1. and 3.3.2. 3 4 3.5. CALCULATIONS (SEE NOTES 10-13)

Fractional esterification rate in VLDL/LDL-depleted plasma (FER& is calculated as the difference between the percentage of labeled esterified cholesterol before and after mcubatton. In human VLDL/LDL-depleted plasma, the percentage of cholesterol esterified during the labelmg and incubation on ice (blank) was always 0.3%: FER,,,

(%/h) = [(% CE,,, ,&

x 2]- 0.3

4. Notes 1. The activity of the enzyme determined m the arttfictal substrate 1s highly correlated with LCAT mass In exceptional cases, such as genetic LCAT defects, enzyme activity may not correspond to its concentratton 2. When ethanolosomes are used as a substrate and the source of LCAT 1s either whole plasma or HDL-plasma the ensumg LCAT acttvtttes are remarkably stml= 26.3 + 5.9 and MERHDL+,tasma= 25 7 f 6 7 nmol/mL/h, r = 0 499 lx WQma in 225 subjects). This strongly suggests that the enzyme activity is not influenced by the phosphotungstate-magnesium chloride precipttatton of apo B-containing lipoproteins from plasma 3. Accordmg to Albers et al. (25), the concentration of LCAT protein influences the production of CE in plasma by approx 60% However, there 1sa weak but stgrnficant correlation (r = 0.34 m 225 SubJects studied) between LCAT activity in HDL-plasma (m absence of LDL and VLDL) and FER,oL Slmtlarly to the lack of relation between LCAT mass and the productton of plasma CE (25), LCAT activity in plasma does not correlate with FER,,, The average values of FERHDL m fresh and stored for 2 mo at -20°C plasma (n = 3 1) were very close. 2 1.4 f 6 8 before and 2 1.8 _+6.4%/h after storage m plasma samples (n = 18) stored at (r = 0.923) Average values of FER,,, -70°C for 2 yr were 23 3 f 4.9%/h and 21.6 + 5 9%/h m the fresh and stored samples, respectively (r = 0.895). In both esttmatlons, there was no slgmficant difference at 10% level between pairs. Short term (wtthm 3 mo) FER,,, was measured twice 3 mo apart m 3 1 apparently healthy subjects There was no significant difference between the first and second estimation (FERHDL 1 vs FERnm 2 = 97.5 f 2.6% m a paired t-test). Long term (within 5 yr) there were no differences (at 10% level) over a period of 2-5 yr (n = 12) The effect on the reaction rate of increasmg temperature (between 35°C and 39°C) 1s very significant and its magnitude differs among indtvtduals. In four different samples, FER,,, increased by 38-59% at mcubatton temperatures between 37°C and 39’C, but only by about 20-22% between 35°C and 37°C

227

Assays of LCAT Table 1 FE&,‘ and Particle Size Distribution in Subjects at Increased Risk of CAD and Controls

(79,27-N)

Subjects

n

FERHDL, %/h

HDL,,, %

HDh,,c, %

Normal female Normal male HT female HT male NIDDM female NIDDM male

53 63 32 40 33 57

1063~ 36 168+ 45 180* 64 286+ 86 26 3 3110.8 35.3 z!I 3 3

266_+ll 1 158f 84 175+124 8.6 k 4 9 -

149* 71 22.3 + 11 4 252k107 37.4 k 12.0 -

HT, hypertension,

NIDDM,

non-msuhn-dependent

diabetes melhtus

9 The esterlficatlon of 3H-UC m HDL plasma is lmear for up to 60 mm m all samples 10. The mterassay varlablllty was assessed m three HDL-plasma samples with values of 8 5, 18 7, and 26 8%/h The measurement of FERHDL was FERHDL repeated 12 times over a period of 1 mo. The coefficient of variation for the 12 measurements ranged from 4 5 to 7 3% 1 I Normal values of FERHDL and the HDL subspecies dlstrlbutlon (using gradient gel electrophoresis) were established in 116 apparently healthy mdlvlduals The average values of FERHDL are significantly different m men and women with comparable plasma lipid concentrations (16.8 + 4.5%/h and 10 6 + 3 6%/h, resp , p > 0 001) and correlate well with the HDL partxle size dlstributlon m the mdlvldual plasma. Plasma samples with high relative content of HDLzb (typical for women and in general healthy people) have low FERHDL while men and persons with major risk factors for coronary artery disease (CAD) have high FERHDL, low HDLZb and high HDL3b,c (I9,27-31) 12. As particle-size distribution of plasma HDL determines FER HDL, we have suggested that this parameter can be used not only as a measure of actlvlty of LCAT in the specific HDL pool, but also as an estimate of HDL particle-size dlstnbutlon. The mformatlon about particle size dlstrlbutlon 1s important because the small HDL partxles may have atherogemc potential, whereas the large HDLZb appears to have “protective” properties (27,32,33). Our previous studies showed that FERHDL dlscrlmmates well between patrents with the major risk factors for atherosclerosis, those with symptomatic coronary artery disease (CAD) and risk factor-free controls (Table 1). We also found that FERHDL value correlates well with the known lipid risk factors for CAD such as the plasma level of tnglycerides and HDL-cholesterol concentration, as well as with body mass 13 FER,,, radloassay is a reproducible and relatively simple method (compared with other more laborious methods, such as the gradient gel electrophoresls method when combmed with ultracentnfugation), which may be smtable for both retrospective and prospective studies of patients at risk or with symptomatic CAD

228

Dobidsovd and Frohkh

Acknowledgment This work has been supported by grants from Grant Agency of Czech Republic 306/96/k220 and B.C. Heart and Stroke Foundation and Medlcal Research Council (Canada).

References 1. Albers, J. J., Chen, C., and Lack0 A. G. (1986) Isolation, characterization, and assay of lecrthm-cholesterol acyltransferase. Methods Enzymol 129, 763-783 2. DoblBsovB, M (1983) LCAT. regulation of endogenous transport of cholesterol Adv Lipid Res 20, 107-194. 3 Fielding, C J. and Fieldmg, P. E (1995) Molecular physiology of reverse cholesterol transport J Llpzd Res. 36,21 l-228. 4 Jonas, A (1991) Lecithin-cholesterol acyltransferase m the metabolism of highdensity lipoproteins. Blochim Biophys Acta 1084,205-220. 5 Von Eckardstein, A , Huang, Y., and Assmann, G (1994) Physlologlcal role and clmlcal relevance of high density lipoprotem subclasses. Curr. Opznlon Lzpidol 5,4OU16 6. Cheung, M. C and Albers, J. J. (1984) Characterization of hpoprotem particles isolated by mununoaffimty chromatography: particles containing A-I and A-II and particles containing A-I but no A-II. J Biol. Chem 259, 12,201-12,209. 7. Puchols, P , Kandoussl, A , Fievet, P , Fourrier, J L., Bertrand, M , Koren, F , and Fruchart, J C (1987) Apolipoprotein A-I containing particles m coronary artery disease Atherosclerosis 68,35-40 8 Ohta, T , Hatton, S., Nlshiyama, S., and Matsuda, I (1988) Studies on the lipid and apohpoprotem compositions of two species of apoA-I containing lipoprotems m normohpidemic males and females. J Llpld Res. 29,721-728 9 Kostner, G M., Kmpping, G , Groener, J. E , Zechner, R , and Dleplmger, H. (1987) The role of LCAT and cholesteryl ester transfer proteins for the HDL and LDL structure and metabolism. Adv. Exp A4ed Blol 210,79-86. 10. Fieldmg, P E., Mnda, T , and Fielding, C J (1991) Metabolism of low-density hpoprotem free cholesterol by human plasma lecithin-cholesterol acyltransferase. Biochemistry 30,855 l-8557 11 Albers, J J., Adolphson J. L., and Chen, C (1981) Radlolmmunoassay of human plasma leclthm-cholesterol acyltransferase J Clan Invest 67, 14 1-148 12 Jonas, A. (1986) Synthetic substrates of lecithin cholesterol acyltransferase J Lipid Res 27,689-698 13. Chen, C. and Albers, J. J (1982) Characterization of proteoliposomes containing apoprotem A-I. a new substrate for the measurement of lecithin cholesterol acyltransferase activity J Lipid Res 23,680-89 1. 14 Stokke, K T and Norum, K R. (1971) Stand J Clin Lab Invest. 27,21-27 15. Dieplinger, H. and Kostner, G M (1980) The determination of lecithin: cholesterol acyltransferase in the clinical laboratory: a modified enzymatic procedure Clan Chrm Acta 106,319-324

Assays of LCAT

229

16. Barter, P J., Hopkins, G. J , Gorjatschko, L., and Jones, M. E (1984) Competitive inhibition of plasma cholesterol esterificatron by human high density hpoprotemsubfraction 2. Bzochlm Bzophys Acta 793,260-268. 17. Barter, P J , Hopkins, G J., and Gorjatschko, L. (1985) Lipoprotein substrates for plasma cholesterol esterification: influence of particle size and composition of the high density lipoprotein subfraction 3. AtheroscEeroszs 58,97-107 18 Dobtasova, M. and Schutzova, M. (1986) Cold labelled substrate and estimation of cholesterol esterificatron rate in lecithin cholesterol acyltransferase radroassay. Physlol

Bohemoslov 35,3 19-327.

19. Dobiasovb, M. and Frohlich, J. (1996) Measurement of fractional esterrficatton rate of cholesterol in plasma depleted of apoprotem B containing lipoprotem. methods and normal values. Physlol. Res 45,65-73 20 McLeod, R , Lacko, A G., Pritchard, P H., and Frohlich, J (1986) Purrficatron of brologically active apohpoprotems by chromatofocussing. J Chromatogr 381, 271-283. 2 1 Wills, G. L , Lane, P. A , and Weech, P. K (1984) A guidebook to lipoprotein technique, m Laboratory Technzques in Biochemistry and Molecular Biology, vol 14 (Burdon, R. H and van Krtppenbe, P. H., eds.), Elsevter, Amsterdam, p 18. 22 Lowry, 0. H., Rosebrough, N J., Farr, A. L., and Randall, R L. (1951) Protein measurement with the Folin phenol reagent. J. Blol Chem 193,265-275 23 Kane, J. P., Sata, T., Hamilton, R. L., and Havel, R. J (1975) Apoprotem cornpositron of very low density lipoproteins of human serum. J. Clm Invest 56, 1622-l 634 24. Scanu, A. M. and Edelstem, C. (197 1) Solubihty m aqueous soluttons of ethanol of the small molecular werght peptrdes of the serum VLDL and HDL. Anal Biochem 44,576588.

25. Albers, J. J , Chen, C. H , and Adolphson, J. L (1981) lecithmcholesterol acyltransferase (LCAT) mass. its relationship to LCAT activity and cholesterol esterrfication rate. J Llptd Res. 22, 1206-1213. 26. Murakamr, T , Mrchelagnoh, S., Longhr, R , Grafranceschi, G , Pazzucom, F , Calabresi, L., Sntori, C. R., and Franceschmi, G. (1995) Trrglycertdes are maJor determmants of cholesterol estertfication/transfer and HDL remodeling m human plasma. 15, 1819-1828. 27 Dobrasova, M., Strrbrna, J., Sparks, D L , Pritchard, P H., and Frohhch, J J (199 1) Cholesterol esterrfication rates m very low density lipoprotein and low density lipoprotein-depleted plasma: relation to high density hpoprotem subspecies, sex, hyperlipidemia, and coronary artery disease. Arterlosclerosls Thromb. 11,64-70

28. Dobiasova, M , Strrbrna, J., Pritchard, P. H., and Frohhch, J J. (1992) Cholesterol esterificatron rate in plasma depleted of very low and low densrty hpoprotem is controlled by he proportion of HDLZ and HDL, subclasses. study m hypertensive and normal middle aged and septuagenarian men J Llptd Res 33,1411-1418. 29. Dobrasova, M. and Frohhch, J. J. (1994) Structural and functional assessment of high density hpoprotem heterogeneity. Clw. Chem. 40, 1554-l 558.

DobiAsovA and Frohkh 30. Dobiasova, M , Stribrna, J., Frohhch, J. J (1995) Relation of cholesterol esterification rate to the plasma distribution of high-density lipoprotem subclasses m normal and hypertensive women Chn Invest Med l&449-454. 3 1. Tan, M H., Loh, C , Dobiasova, M., and Frohlich, J (1998) Fractional estenficatton rate of HDL particles m patients wtth type 2 diabetes. Dzabetes Care 21, 139-142 32 Drexel, H., Aman, F W., Rentsch, K., Neunschwander, C , Leuthy, A, Khan, S I , and Foltath, F. (1992) Relation of the level of high-density lipoprotem subfractions to the presence and extent of coronary artery disease Am J. Cardzol 70,436-440. 33. Willlams, P T , Krauss, R. M , Vramzan, K M., Stefamck, M. L , Wood, P. D S , and Lmdgren, F T (1992) Assoctations of hpoprotems and apohpoprotems with gradient gel electrophorests estimates of high density hpoprotem subfraction m men and women Arterloscleroszs Thromb. 12,332-340

15 Determination of the Mass Concentration and the Activity of the Plasma Cholesteryl Transfer Protein (CETP)

Ester

Laurent Lagrost 1. Introduction In human plasma, cholesteryl esters and triglycerides can exchange between various hpoprotem fractions through the action of one specific lipid transfer protein: the cholesteryl ester transfer protem (CETP) (1,2). Since m VIVOCETP activity results in the redistribution of neutral lipid species between pro- and antiatherogemc lipoprotein particles, a marked interest has been given to the evaluation of CETP mass and activity in human plasma; and durmg the last decade, several specific assays---Including net mass transfer assays, isotopic transfer assays,and immunoassays--have been proposed. In fact, the cholesteryl ester transfer reaction is a complex process that integrates a sequence of events and is dependent on a number of factors, among them the CETP mass concentration, as well as the amount and the composition of the plasma lipoprotein substrates (2). Thus, the evaluation of plasma CETP activity should ideally integrate several methods in order to obtam a clear picture of what IS occurring. Two distinct methods are described m this chapter: (1) a competitive enzyme-linked immunosorbent assay(ELISA) of the CETP protein; and (2) an isotopic transfer assay, which evaluates the activity of CETP as modulated by endogenous plasma factors. 1.1. Enzyme-Linked lmmunosorbent Assay (ELISA) of CETP Several radiom-munoassays and enzyme immunoassays have been developed for the determination of CETP mass concentrations (3-6). In particular, the first radioimmunoassay-described by Marcel and coworkers (31, and using specific monoclonal antibodies (MAbs)-provided much information on the From

Methods m Molecular Sro/ogy, Vol 110 Lpoprotern Protoco/s Edlted by J M Ordovas 0 Humana Press Inc , Totowa, NJ

231

232

Lagros t

mass concentration and distribution of CETP m human plasma. However, the latter immunoassay of CETP presents the disadvantage to require anti-CETP antibody labeling and radioisotope handlmg, making it unusable by some laboratories. A competitive ELISA using anti-CETP MAbs similar to those previously used by Marcel and coworkers for CETP radioimmunoassay (3) is described m Subheading 3.1. This assay is based on the competttive binding of specific mouse anti-CETP MAbs to CETP present m the human plasma sample, and to partially purified CETP bound to a 96-microwell immunotitration plate (7) (see Fig. 1). The enzyme immunoassay does not require antibody labeling, and anti-CETP antibodies bound to the CETP-coated microwells are detected by usmg peroxidase-conlugated antimouse antibodies. The absorbance measured in microwells after the addition of an enzyme substrate solution is mversely proportional to the amount of CETP m the human plasma sample. All the steps-including pipetting, dilutmg, dispensing, washmg, and photometry-can be automated.

1.2. Isotopic Cholesteryl Ester Transfer Assay During the last 15 yr, several expertmental protocols have been described to determine plasma CETP activity, and it is possible to distinguish between two groups of methods: (1) the measurement of the net mass transfer of cholesteryl esters among high-density hpoproteins (HDL), very low-density lipoprotein (VLDL), and low-density lipoprotem (LDL) plasma fractions; (2) the measurement of the rate at which radiolabeled cholesteryl esters are exchanged between HDL and apo B-contaimng lipoproteins. Whereas the former assay allows evaluation of the net lipid flux resulting from multiple lipid exchange reactions in plasma, it must be carefully conducted. Indeed, m plasma samples with elevated CETP activity, most of the net mass transfer reactions may have already occurred in vivo, reaching virtually an eqmlibrmm state and making the plasma lipoprotems as poorer substrates m the cholesteryl ester transfer reaction as compared with lipoproteins from plasma samples with low CETP activity (8,9). In contrast, the isotopic transfer assay allows evaluation of the ability of plasma CETP to interact with lipoprotem substrates and to exchange neutral lipid species even in the absence of significant net lipid flux. In this particular case, two distinct approaches have been used, with the evaluation of CETP activity etther in total, native plasma, or m lipoprotem-deficient plasma. It is noteworthy that in the latter case CETP activity is measured independently of endogenous llpoprotems, reflecting mainly CETP mass concentration, and very good correspondences were reported in the literature between CETP activity measured in lipoprotein-deficient plasma and CETP mass concentrations measured with specific mrrnunoassays. In other words, when a CETP immunoassay is available, determination of isotopic cholesteryl ester transfer

Mass Concentration

and Activity of Plasma CETP

Sample

Plate coating 4”C, overnight

233

+ anti-CETP antibodies 4”C, overnight

I

+ albumin room temperature,

30 min

I

37°C.

3 hours

washing + peroxidase-conjugated 37”C, 1 hour

A

CETP

d

Albumin

),

Anti-CETP

f

Pe$&ynjugated

+ enzyme substrate room temperature, antibody

antibodies

solution 15 min

+

reaction

stopped

and absorbance

reading

Fig. 1. Sequentialprocedurefor the competitive ELISA of CETF. rates in lipoprotein-deficient plasma does not add much information on CETP activity in plasma, unless a specific lipid transfer inhibitor protein is present in the lipoprotein-deficient fraction. In contrast, determination of isotopic transfer rates in total, native plasma reflects not only the plasma CETP mass levels, but also the ability of endogenous lipoproteins to act as donors or acceptors in the CETP-mediated lipid transfer process. In fact, a number of studies revealed that abnormal plasma cholesteryl ester transfer activity in various diseases

234

Lagros t

associated with lipoprotem disorders relates at least m part to alterations in the concentration and composltlon of plasma lipoproteins (2). Based on these observations, determination of CETP activity m human plasma should integrate at least two methods to give a clear picture of what IS actually occurring m vivo: (1) the determination of CETP mass concentration by using a specific tmmunoassay as described m Subheading 3.1.; and (2) the determination of isotopic cholesteryl ester transfer rates m total, native plasma, which reflects plasma CETP mass as modulated by endogenous plasma factors (among them the concentration and composition of plasma llpoprotem substrates). Subsequently, comparative analysis of data obtained with the two distinct expenmental approaches allows evaluation of the relative contribution of hpoprotem components and CETP mass m determming cholesteryl ester transfer activity in total plasma. An experlmental protocol for the determmatlon of the lsotoplc transfer of radiolabeled cholesteryl esters from either a tracer dose of trltlated htgh-density hpoprotems-subfraction 3 (HDL3) toward plasma apo B-contammg lipoproteins, or from a tracer dose of trltiated LDL toward plasma HDL, 1s described m Subheading 3.2. (see Fig. 2).

2. Materials 2.1. ELISA

2 3 4

5

Blomek 1000 BloRobotics System with the Immunofit EIA/RIA Data Analysis Software (Beckman Instruments, Palo Alto, CA) Titer plates 96-deep-well titer plates (Beckman) Microwell plate. F96 CERT-Maxlsorp from Nunc (Kamstrup, Denmark) TP 1 anti-CETP MAbs* Commerclaly avallable from University of Ottawa Heart Institute-Research Corporation (Dr R W Mllne and Dr Y Marcel) TP 1 antibodies recognize an epltope that is localized m the carboxy-terminal region of the CETP molecule (10). Partially purified CETP Preparation from cttrated human plasma by a sequential procedure including ammonium sulfate precipitation, ultracentrifugatlon, hydrophobic mteractlon chromatography, cation-exchange chromatography, and anlonexchange chromatography (21) Storage of CETP-contammg fractions at -80°C (see Note 1)

6. Peroxidase-conjugated antimouse antibodies (Blo-Rad, Hercules, CA) diluted In albumm-phosphate buffer 7 Coatmg buffer 15 mA4Na2C0,, 35 mMNaHC03, 3 mMNaN3, pH 9 6 8 Washmg solution 150 mA4 NaCI, 0 025% (v/v) Tween-20 (Merck, Darmstadt, Germany). 9. Albumm-phosphate buffer 10 mM Na,HPO,, 5 mA4 NaH*PO,, 150 mA4 NaCl, 1% bovine albumm (w/v), pH 7.2 10 Anti-CETP antibodies dilution buffer: albumin-phosphate buffer containing 1% Trlton X- 100 (Pierce, Rockford, IL)

Mass Concentration and Activity of Plasma CETP

235

Ultracentrifugation

de 1.066 g/ml fraction

pc d > 1.066 g/ml fraction .:: F1

dz

I.666 g/ml fraction

1

::. I( gi .Y

d c 1.066 g/ml fraction

Radioactivity

Transfer

from

Radioactivity

counting

radiolabeled HDk ma apoB-containing

Transfer towards

counting

from radiolabeled plasma HDL.

LDL

Fig. 2. Sequential procedures for the isotopic transfer assay of plasma cholesteryl ester transfer activity. 11. Enzyme substrate solution: 0.4 g/L o-phenylenediamine, 0.68 g/L hydrogen peroxide in 6.6 mA4 sodium phosphate, 3.4 n&I citrate, pH 5.2 buffer; freshly prepared and kept in darkness. 12. 2.5 MH2S04 solution.

2.2. Isotopic

Cholesteryl

Ester Transfer Assay

1. Ultracentrifugation: TL-100 (Beckman); Rotor: 100.2 with 2-mL Quickseal tubes (Beckman). 2. Test tubes: Micro Test Tubes 3810 (Eppendorf, Hamburg, Germany).

236

Lagrost

3 [ la,2a-3H]cholesterol m toluene (Amersham France SA, Courtaboeuf, France); specific acttvtty, 46 CtImM 4. Trts-buffered saline (TBS). 10 nuI4 Trts-HCl, 150 r&I NaCl, 5 mA4 Na,-EDTA, 3 mA4 sodium aztde, pH 7.4 5. Iodoacetate solution. 15 n&f todoacetic acid m TBS 6 Solid KBr. 7. KBr solutron. 0.102 g/L potassium bromide m water, d = 1 07 g/mL 8 Scmttllation fluid. OptiScmt Htsafe 3 (Pharmacia)

3. Methods

3.1. ELISA 3.1.1. Plate Coating 1. Add a 100~)IL volume of parttally purified CETP diluted m the coatmg buffer (protem concentratron, approx 10 ug/mL) mto each well of a polystyrene microwell plate (see Notes 2 and 3) 2. Cover with an autoadhesrve plastic film and Incubate overnight at 4°C 3. Rinse four ttmes wrth the washmg solutron, and block nonspectfic absorptton by incubating each well for 30 mm at room temperature with 200 pL of albummphosphate buffer 4. Empty plate mrcrowells.

3.1.2 Sample Treatment 1. Dilute CETP-containing samples in albumm-phosphate buffer 2 Dilute monoclonal antt-CETP anttbodtes (TPl) m Trtton-contammg albummphosphate buffer 3 MIX equal volumes of diluted sample and diluted TPl soluttons m titer plates, cover, and incubate overmght at 4°C (see Note 4). 4. Add 100-pL ahquots of incubated mixtures to microwell plates, and incubate for 3 h at 37% 5 Rinse four times with washing solution.

3.1.3. Detection of Bound Anti-CETP Antibodies 1 Add a loo-& volume of peroxidase-conjugated antimouse antibodies, and mcubate for 1 h at 37’C 2 Rinse four times with washing solutron 3. Add a 100~& volume of enzyme substrate solutron, and incubate for 15 mm at room temperature in the dark (see Note 5). 4 Stop the reaction by the addition of 50 pL of H$O,, and read the absorbances at 492 nm (see Notes 6-8)

3.2. Isotopic Cholesteryl Ester Transfer Assay 3.2. I. Preparation of d > I. 13 g/mL Human Plasma Fraction 1. Adjust normohptdemtc human plasma (20 mL) to density 1 13 g/n& with solid KBr (0.188 g of dehydrated KBr per milhhter of plasma)

Mass Concentratron and Activity of Plasma CETP

237

2. Ultracentrifuge for 7 h at 350,OOOg. 3. Recover the d > 1.13 g/mL infranatant by tube slicing, and dialyze overnight against TBS.

3 2.2 Preparation of Human Plasma Low-Density Lipoproteins (LDL) 1. Adjust normohpidemlc human plasma (10 mL) to density 1 019 g/mL with solid KBr (0.0 18 g of dehydrated KBr per milliliter of plasma). 2. Ultracentrifuge for 4 h at 350,OOOg. 3. Recover the d > 1,019 g/mL mfranatant by tube slicing. 4. Adjust the d > 1.019 g/mL plasma fraction to density 1 055 g/mL with solid KBr (0.052 g of dehydrated KBr per milliliter of plasma). 5 Ultracentrlfuge for 5 h at 350,OOOg. 6. Recover the d < 1.055 g/mL LDL fraction, and dialyze overnight against TBS

3.2.3. Radiolabeling

of HDL3 and LDL

1. Plpet 500 p.L of [ la,2ct-3H]cholesterol mto a 20-mL glass tube, and evaporate toluene under a stream of nitrogen. 2 Add 50 pL of ethanol, and vortex gently to resuspend trltlated cholesterol. 3. Add the isolated d > 1 13 g/mL plasma fraction to glass tube containing trltlated cholesterol, quickly vortex, and incubate for 24 h at 37°C in a shaking water bath to allow the esterlficatlon of trltlated cholesterol through the reaction catalyzed by the 1eclthin:cholesterol acyltransferase (LCAT), which IS contained m the d > 1 13 g/mL plasma fraction 4 Add LDL to the incubated mixture, and prolong the incubation for an additional 6-h period to allow the CETP-mediated exchange of biosynthetically radlolabeled cholesteryl esters between HDL, and LDL fractions (see Note 9). 5. Isolate sequentially HDL3 and LDL by using ultracentrlfugatlon Isolate HDL, as the 1 13 <: d < 1 21 g/mL fraction, with one 7-h, 350,OOOg spin at the lowest density, and two successive 10-h, 350,OOOg spins at the highest density Isolate LDL as the 1.019 < d < 1 055 g/mL fraction, with one 4-h, 350,OOOg spm at the lowest density, and two 5-h, 350,OOOg spins at the highest density 6 Dialyze trltiated HDL, and LDL fractions overnight against TBS

3.2.4. Determination

of Plasma Cholesteryl Ester Transfer Activity

1 Mix plasma (25 pL), radlolabeled hpoprotein substrate (2 5 nmol of cholesterol), iodoacetate solution (5 pL), and TBS m a final volume of 50 & in Eppendorf test tubes. Iodoacetic acid is added to block the activity of the plasma lecithin: cholesterol acyltransferase (LCAT). 2 Vortex, and incubate for 3 h at 37°C m a shaking water bath. Keep control mixtures at 4OC 3. Place incubated tubes on ice for 5 min, and then remove condensed water from tube wall by one 5-mm, low-speed (6000g) centrifugation step 4. Add a 45-w volume of each incubated mixture to 1.95 mL of a d = 1.07 g/mL KBr solution m 2-mL Quickseal centrlfugatlon tubes.

238

Lagrost

5 Ultracentrtfuge for 5 h at 350,OOOg (100,000 r-pm), and recover the VLDL+LDLcontammg supernatant, and the HDL-contammg infranatant in 1-mL volumes by tube shcmg (see Notes 10 and 11) 6. Transfer ultracentrtfugally isolated fractions into counting vials containing 2 mL of scmtillatton fluid. Vortex. 7. Radioactrvtty counting for 5 min m a liquid scmtillatton counter (see Notes 12 and 13)

4. Notes Detailed description of the methodology used to prepare partially purified CETP IS beyond the scope of this chapter and can be found m ref. 11. When setting up the CETP ELISA, the amounts of anti-CETP antibodies, peroxtdase-conjugated anttmouse antrbodres, and CETP antigen reacting m the ELISA must be successtvely determined according to the general procedure prevtously described (12). Briefly, serial dtluttons of anti-CETP anttbodtes are added to coated microwells, and the plates are incubated for 3 h at 37°C. After washing, various dtlutions of peroxrdase-conjugated antibodies are added to each anttCETP dtlutton series, plates are Incubated for 1 h at 37°C and colortmetrtc reaction is conducted by adding the enzyme substrate solutton The latter procedure allows one to obtain stgmotdal titration curves on a semilogaritmtc scale, and absorbance values at 492 nm are directly related to the amount of anti-CETP antibodies added. The selected optimal amount of peroxtdase-conjugated second antibodies must allow one to obtain dose-response curves (1) with an absorbance background value below 0.2, and (2) with an absorbance range of approx 2 0 The antKETP antibody dilution must be a hmitmg factor in the assay, and the working drlutron must be chosen before the stgmotdal trtratton curve reaches the maximal absorbance value. Finally, following the selection of antibody dtlutions, a standard inhibition curve is obtained with serial diluttons of a CETP-containing fraction m order to determine the detectable amount of CETP as well as the workmg range of the assay It is not necessary to use a pure CETP preparation to coat plate microwells, and partially purified, CETP-enriched fractions obtained through different expertmental ways may be used However, a sufficiently high degree of purtficatton, approxtmatmg 3000-fold compared with normal human plasma, must be achieved m order to obtain accurate and reproductble competmon curves. The calibration of CETP ELISA IS achieved by using a frozen plasma standard stored m ltqutd nitrogen. CETP concentratron in the plasma standard can be determined as compared either with a pure CETP preparation of known protem concentration or with a plasma standard from another laboratory A cahbratton curve for each plate of CETP tmmunoassay 1s constructed with eight duplicate dilutions of the standard plasma of known CETP concentratton Standard curves are fit to the data points, and calculation of CETP concentrations m biologtcal samples are realized by using a data analysis software. The tangerine coloration of oxtdtzed o-phenylenediamme chromogen 1s lightsensitive, and mrrnunotitratton plates must be kept in dark during mcubation of plate mtcrowells with the enzyme substrate solution and until absorbance reading

Mass Concentration

and Actwity of Plasma CETP

239

6 The minimum detectable content is about 1 ng of CETP per well, with a working range of l-10 rig/well The CETP assay is independent of plasma lipid levels, and similar displacement curves can be obtamed with normolipidemic, hypercholesterolemm, and hypertriglyceridemtc plasmas 7 Intra- and interassay coefficients of variations are 4 and 6%, respectively 8. The mean CETP concentration in normohpidemic plasma is approx 2.8 pg/mL, with a range of 1.9-4.2 ugImL. 9 Typically labeled preparations of LDL and HDLs have specific activities of approx 4000 and 12,000 cpmnmol of cholesterol, respectively More than 95% of the total radioactivity of both lipoprotem substrates reside in the cholesteryl ester moiety Thts can be checked after separation of lipid species by using thmlayer chromatography 10. Prior to radioactivity counting, acceptor lipoprotems can be separated from donor ones by usmg additional techniques, includmg selective precipitation (13,14), solid-phase-bound lipoproteins (15), or polyacrylamtde gradient gel electrophoresis (16). In particular, the latter technique allows one to determine simultaneously cholesteryl ester transfers not only toward one smgle hpoprotein fraction but toward the various plasma lipoprotein classes, reflecting what IS actually susceptible to occur m VIVO (16) 11. A major concern with isotopic assays of cholesteryl ester transfer activity is the requirement for separation of donor from acceptor lipoprotem substrates before radioactivity counting. Interestingly, the Scmtillation Proximity Assay (SPA) technology (Amersham) can overcome the latter problem by evaluatmg the radioactivity content of one given substrate m homogenous medium through the use of fluoromicrospheres, which can react specifically with radtolabeled hgands Both streptavidine/biotm (Cholesteryl ester transfer protem [3H]SPA assay system, Amersham) and apohpoprotem/anttapolipoprotem antibody (2 7) mteractions between fluoromicrospheres and lipoprotems have been used m CETP activity assays. The SPA technology is of particular Interest because of its high potential for automation 12. Cholesteryl ester transfer rates are calculated as the percentage of total cholesteryl ester radioactivity transferred from the tritiated lipoprotein donor to either the VLDL + LDL or the HDL plasma acceptors compared with control mixtures kept at 4°C. Results can be expressed m percentage of trrtrated cholesteryl esters transferred per hour per milhhter of plasma (%/h/mL) 13 Specific CETP activity can be calculated as the ratio of plasma CETP activity to plasma CETP mass concentration, and it can be expressed m percentage of total radiolabeled cholesteryl esters transferred per hour per microgram of CETP (%/h/pg) 14. The experimental system described above can also be used to measure cholesteryl ester transfer from one radiolabeled lipoprotem donor toward one isolated hpoprotein acceptor m the presence of a CETP source (purified CETP, lipoprotemdeficient plasma, and so on)

240

Lagrost

References 1. Tall, A. R. (1995) Plasma lipid transfer proteins Annu Rev Bzochem 64, 235-257. 2. Lagrost, L. (1994) Regulation of cholesteryl ester transfer protein (CETP) activity Review of in vttro and in vtvo studies. Bzochzm Biophys Acta 1215, 209-236 3 Marcel, Y L , McPherson, R , Hogue, M , Czarnecka, H , Zawadzkr, Z , Weech, P. K., Whitlock, M E , Tall, A R., and Mime, R W (1990) Distrrbutton and concentratton of cholesteryl ester transfer protein m plasma of normoliptdemtc subjects. J Clan. Invest 85, 10-17 4 Mezdour, H , Kora, I , Parra, H J , Tartar, A., Marcel, Y. L., and Fruchart, J C (1994) Two-site enzyme immunoassay of cholesteryl ester transfer protein with monoclonal and ohgoclonal antibodies Clm Chem 40, 593-597 5. Kukasawa, M , Arai, H , and Inoue, K (1992) Establishment of arm-human cholesteryl ester transfer protein monoclonal anttbodtes and radrounmunoassaying of the level of cholesteryl ester transfer protein in human plasma. J Bzochem 111,996-998.

6 Ritsch, A., Auer, B., Foger, B., Schwarz, S., and Patsch, J R (1993) Polyclonal antibody-based mnnunoradiometric assay for quanttficatton of cholesteryl ester transfer protein J Llpld Res 34,673-679. 7. Guyard-Dangremont, V , Lagrost, L , Gambert, P , and Lallemant, C. (1994) Competitive enzyme-linked mnnunosorbent assay of the human cholestetyl ester transfer protein (CETP). Clm Chum. Acta 231, 147-160 8 Son, Y S and Ztlversmrt, D B (1986) Increased lipid transfer activittes m hyperliptdemtc rabbit plasma Arterloscler 6, 345-35 1 9. Van Tol, A (1993) CETP-catalysed transfer of cholesteryl esters from HDL to apo B-containing hpoprotems in plasma from diabetic patients. Eur. J Clzn Invest 23, 856 10 Swenson, T. L., Hesler, C. B , Brown, M L., Quinet, E , Trotta, P P., Haslanger, M. F , Gaeta, F C. A., Marcel, Y L , Milne, R W., and Tall, R (1989) Mechanism of cholesteryl ester transfer protein mhibmon by a neutrahzmg monoclonal antibody and mapping of the monoclonal antibody epitope. J BIOI Chem 264, 14,318-14,326 11 Lagrost, L. and Barter, P J. (199 1) Effects of various non-esterified fatty acids on the particle stze redistribution of high density hpoprotems Induced by the human cholesteryl ester transfer protein. Biochzm. Bzophys Acta 1082, 204-2 10 12 Engvall, E (1980) Enzyme immunoassay ELISA and EMIT Methods Enzymol 70,419-439. 13. Morton, R. E and Zilversmit, D B (1981) A plasma mhtbrtor of triglyceride and cholesteryl ester transfer activities J Bzol. Chem. 256, 11,992-l 1,995. 14. Groener, J. E. M., Pelton, R. W , and Kostner, G. M (1986) Improved estimation of cholesteryl ester transfer/exchange acttvity m serum or plasma. Clan Chem 32,283-286.

Mass Concentration

and Activity

of Plasma

CETP

15. Sparks, D L , Frohhch, J , Culhs, P., and Pritchard, P. H. (1987) Cholesteryl ester transfer activity m plasma measured by using solid phase-bound high-densrty lipoproteins Clan Chem 33, 390-393. 16 Florentin, E , Lagrost, L , Lallemant, C., and Gambert, P (1994) Polyacrylamide gradient gel electrophoresis as a method to measure transfers of radiolabeled cholesteryl esters between several plasma lipoprotein fractions. Anal Bzochem 216,352-357

17 Lagrost, L., Loreau, N , Gambert, P., and Lailemant, C. (1995) Immunospecitic scmtillation proximity assay of cholesteryl ester transfer protein activity Clzn Chem 41,914-919

In Vitro Measurement of Lipoprotein and Hepatic Lipases Elisabet Vilella and Jorge Joven 1. Introduction Two genetically related hpases perform the hydrolysis of fatty acids from triglycertdes and phosphohpids m plasma lipoproteins; lipoprotein lipase (LPL) and hepatic hpase (HL) (1). The former catabolizes the hydrolyses of triglycerides m chylomicrons and very low-density lipoproteins (VLDL) producing the so-called remnant particles. The later, with higher affinity for phospholiptds, may modify the conformation of lipoproteins acting on the surface of mtermediate-density lipoproteins (IDL) and high-density lipoprotem (HDL) particles. Both enzymeshave a high affimty for the proteoglycans that anchor the hpases to the endothelial surface, then resident site. LPL has a peripheral distribution being more abundant m tissues m which fatty acids are stored (adipose ttssue) or predominantly used as metabolic fuel (muscle), whereas HL is confined to the endotheha of the liver. When an heparm bolus is mjected mto the blood stream, a release of lipases from their anchors to the ctrculatton is observed because hpases have higher affimty for heparm than for proteoglycans. After heparin mlection, the amount of LPL and HL cnculatmg m plasma is increased several hundred folds and the amount of hpases in postheparm plasma is taken as an indrrect measure of the lipase content in the mdlvidual(2). LPL and HL are glycoproteins of a molecular mass of approx 4.5 kDa by sedimentation equilibrium ultracentrifugation and an apparent monomertc molecular mass of 60 kDa on sodium dodecyl sulfate-polyacrylamtde gel electrophoresis (SDS-PAGE); LPL is only active as homodtmer and needs apolipoprotein CII as a cofactor. They are members of a family of related enzymes that are serme esterases. A high degree of homology exists between the genes for LPL, HL, and pancreatic hpase. The mRNA for LPL m humans 1s highly From

Methods E&ted

by

m Molecular J M Ordovas

Bro/ogy,

Vol 110 Ltpoprotern

0 Humana

243

Press

Inc , Totowa.

Protocols NJ

Vile/la and Joven

244

homologous with that of mice, rats, and cows. It also has 47% homology with the mRNA for HL, which may represent a similar catalytic, heparm-, and lipidbinding sites, but a major difference respect to the apo CII-bmdmg site (3). Recently, it has been shown that lipases act as bmdmg proteins, bridging between lipoproteins and several cell surface sites such as unspecific bmdmg to proteoglycans and a more specific binding to the low-density llpoprotem receptor or to the low-density lipoprotein receptor-related protein receptor. This noncovalent crosslink between lipoproteins and cell surfaces mediated by llpases is possible because hpases have heparin- and lipid-bmdmg domains. This concept is still controversial with respect to HL, but tt appears demonstrated that the homodlmerlc nature of LPL allows the protein to actually conform blochemlcal bridges. LPL can therefore retain lipoproteins to the cell surface and facilitate their internallzatlon and catabolism, a phenomenon that has been observed m vitro m several experimental models but Its physiological importance has not yet been completely evaluated (4). LPL total deficiency causes severe hypertriglyceridemia, the so-called chylomlcronemia syndrome that is a rare autosomal recessive disorder characterized by a massive accumulation of chylomicrons m plasma and a correspondmg increase of plasma triglyceride concentration. Partial deficiency of LPL 1s also related to familial combined hyperlipidemia and other forms of mild hypertrlglyceridemla. Hepatlc lipase deficiency, however, is only associated to mild dislipemia, supportmg its role as a modulating enzyme (5). Measurement of LPL and/or HL activities may be useful for diagnostic purposes, but the techniques described are basically devoted to research because they are still tedious and require a careful standardization to keep imprecision as low as possible. Moreover, the measurement of the enzyme mass alone or m combination with its activity may broaden the field of application. Unfortunately, there are no commercially available reference materials, calibrators, or control materials, and the assistance of colleagues for provldmg skillful and materials is mandatory when a laboratory is being imtlated in the field. Antibodies against human lipases are also difficult to rise and mamtain because the high homology with the lipases present in the experimental animals; again, using well-established antlbodles 1shighly recommended. Therefore, we presume that the interlaboratory reproducibility 1slow. 2. Materials 2.7. Equipment 1 2. 3. 4.

Centrifuge, refrigerated with a swing-out rotor. Sonicator. Liquid scintillation a counter Multichannel spectrophotometer (for 96-well mlcrotiter plates).

In Vitro Measurement of LPL and HL

245

5. Microtlter plate washer (optional). 6. 96-Well microtlter plates and lids and sealing tapes (MaxlSorp, Nunc, Roskllde, Wlesbaden-Biebnch, Denmark, Germany). 7 8-12 multlchannel mlcroplpets (50-250 pL) and tray

2.2. Buffers and Solutions 1. 0.1 M Carbonate-bicarbonate buffer, pH 10.5. Mix 800 mL of 0.1 A4 sodium carbonate with 200 mL of 0.1 M sodrum bicarbonate. 2. Phosphate buffered saline (PBS): 0.15 MNaCl, 10 mMphosphate buffer, pH 7.4. Dissolve 8 g NaCl, 0.2 g KCl, 1.44 g Na2HP04 and 0 24 g KH2P04, in 900 mL of distilled water, check the pH (adjust if necessary), and bring the volume to 1 L. (Preparing a stock 10 X solution is convement.) 3. PBS, 0.05% (v/v) Tween-20 (PBS-T). Dissolve 0.5 mL ofTween-20 in 1 L of PBS. 4 PBS, 0 1% (v/v) Tween-20, 4% (w/v) BSA, 0.1% (w/v) heparm (PBS-T-A-H) Dissolve 0 1 mL of Tween-20 in 100 mL of PBS. Add 4 g of bovine serum albumm (BSA), essentially fatty acid free, Sigma, St. Louis, MO, No A-75 11) and 0.1 g of heparm (Sigma, No. H-7005) on the top of the solution and let it dissolve without stirring. It should be freshly prepared, because albumin solutions are easily contaminated 5 0 05 MCitrate-phosphate buffer, pH 5 0. Dissolve 7 3 g citric acid and 23 88 g Na2HP04 12 HZ0 m 1 L of water, pH 5.0. 6 0.2 M Tns-HCl buffer, pH 8 2 (Trizma, Sigma, No. T-5003) 7 0.2 M Tris-HCl buffer, pH 8.8 (Tnzma, Sigma, No. T-5753). 8. Methanol:chloroform:heptane (56.50.40 by vol, Belfrage mixture) 2.3. Glass ware 1. 2. 3 4

Glass lithium-heparmate blood-collection IO-mL Glass test tubes. Glass scmtillatlon vials. Cryotubes.

tubes

3. Methods

3.7. LPL and HL Activities Assuming there is not a significant amount of circulating pancreatic lipase, LPL and HL are released from their active sites in the endothelmm by the mJectlon of an heparin bolus. It 1simportant to remember that LPL needs apo

CII as a cofactor (normally present m its substrates,VLDL and chyllomlcrons) and has an optimum pH of 8.8. HL does not need apo CII, and the optimum pH 1s8.2. Both enzymes need albumin to capture the free fatty acids released by the hydrolysis (see Note 1). For the assay, the sample 1s mixed with an emulsion-containmg cold and labeled [‘4C]trlolem

After incubation

and selective extraction

of the [‘4C]oleic

acid produced from the enzyme-catalyzed hydrolysis, liquid-scintlllatlon

Vile/la and Joven

246

counting IS performed. To measure LPL activity selectlvlty, HL is inactivated by incubating the sample with SDS Conversely, the LPL activity may be selectively inhibited in the absence of apo CII with high NaCl concentrations and/or specific antibodies. 3 1 1. Pre- and Postheparin Plasma Fasting for at least 12h ISmandatory. Prepare 2 hthmm-heparmate tubs (5 mL) labeled as pre- and postheparin sample. During the procedure, plasma should be on ice. 1. Draw blood into a 5-mL hthmm-heparinate tube (preheparin plasma). 2. Inject a bolus of 70 IU sodium heparmate/kg body weight

3. Wait 10min. 4. Draw blood from the opposite arm into a refrigerated 5-mL hthmm-heparmate tube (postheparm plasma). 5. Centrifuge nnmedlately to separate plasma from cells m a refrigerated centnfuge. Aspirate and allquot (250 p,L) plasma m cryotubes Freeze immediately at -7O’C. Samples may be stored several months but thawing should be avoided

3.1.2. LPL Standard from Bovine Milk (Optional) Unpasteurized bovine milk can be obtained fresh from a dairy farm. 1 Remove cream by centrifugatlon at 4°C 2 Stir the skim milk (500 mL) for 30 mm at 4°C with 0 1 A4 trlsodmm citrate dlhydrate and then dialyze against PBS (0 15 M NaCI, 10 mM sodium phosphate, pH 7 4) Three changes of 6 h each one and IO-fold excess buffer 3 Add glycerol to 30% (v/v) and freeze the standard m 1-mL ahquots at -70°C 4 Validate each standard preparation

3.1.3. Source of Apo C/I 1 2 3. 4

Pool serum from several volunteers. Inactivate by heating at 56°C for 1 h Ahquot (1 mL) and freeze at -20°C. Validate each serum pool

3.1.4. Substrate Preparation For LPL assay use the followmg

procedure.

1. Into a 20-mL glass scintlllatlon vial, add 100 mg of unlabeled tnolem, [14C]triolem (Amersham), 5 68 mL of gum arabic (Sigma, No G-9752) in 0.2 M Tris-HCl buffer, pH 8.2). 2. Somcate under the following condltlons. 24 cycles of 20 s on, 10 s off vial content by inversion after every 4th cycle 3. Add and mrx by mverslon. 2.75 mL of 200 g/L BSA m 0 2 MTris-HCI pH 8.2, and 0.8 mL of 1.42 MNaCl m 0.2 M Tris-HCl buffer, pH 8 2.

5 @I of (90 g/L Mix the buffer,

In Vitro Measurement

247

of LPL and HL

For HL assay use the same procedure with the following modlficatlon: Add 2.73 mL of 200 g/L BSA and 2.58 mL of 3.24 MNaCl, both prepared m 0.2 M Tris-HCl buffer, pH 8.8, to the triolein-emulsifier mixture.

3.7 5 Substrate Preincubation for LPL Assay 1. Prepare glass test tubes, with proper stoppers, for at least three experiments for each sample. 2 Pipet 0 42 mL of substrate emulsion m each tube. 3. Add 80 pL of heat inactivated pooled human serum (source of apo CII). 4 Incubate 90 min at 37°C (transfer of apo CII from pooled serum to substrate micelles).

3.7.6. Sample Preincubation (30 min before completion of substrate preincubation) 1. Plpet 0.20 mL of sample (PHP or preHP), saline as blank and skim milk standard mto an Eppendorf tube. 2. Add 0 2 mL of 70 mA4 SDS in 0.2 MTris buffer, pH 8.2 (to inactivate HL) 3 Incubate at 20°C for 30 min (see Note 2)

3.7 7. Incubation of Sample with Substrate 1 Add 10 pL of sample to each of the three tubes of activated substrate. 2. Incubate 1 h, exactly at 28°C mto a water bath with smooth shaking. 3 Stop triglyceride hydrolysis by adding 5.33 mL of methanol:chloroform:heptane (Belfrage mixture)

3.1.8. HL Assay Incubation Proceed as for the LPL assay incubation I 2. 3. 4 5.

but with the following

differences:

Preparation of the substrate (see Subheading 3.1.4.). No premcubatlon of substrate with apo CII source. All buffers have pH 8.8. No sample pretreatment High NaCl concentration m the reaction mixture to inactivate LPL

3.7.9. Fatty Acid Extraction 1 Vigorously shake the hydrolysis mixture with the methanol:chloroform.heptane for 30 s 2. Add 1.5 mL freshly prepared 0.1 A4 carbonatebicarbonate buffer, pH IO.5 3. Shake 10 s. 4. Centrifuge for 45 mm at 15008 using a swing-out rotor. 5. Prepare glass scmtillation vials with 50 pL of glacial acetic acid containing 0.05% (w/v) ferric chloride. 6. To each scmtillation tube, add 2 mL of the upper phase and swirl 7. Add 16 mL Ecoscmt*toluene (7.1)

248

Vile/la and Joven

3.1.10. Radioactivity measurement and calculations Count [t4C]oletc acid setting the counter to the followmg conditions: 1 Channel 397-655. 2 5 mm of counting 3. Automatic quench correctron (based on Compton curve).

Lipase activity is calculated from the difference m counts/min between the blank and the sample vials. LPL bovine-milk standard, if added to the assay,is used to calculate intra-assay variability. Activity values are given m prnol/h/mL, which corresponds to the amount of oleic acid released per hour and per milhhter of plasma sample. Expected values m normal subjects using this method are, for LPL, 15 zk12 unrol/h/mL in men and 14 f 9 pmol/h/mL in women, and for HL 25 + 15 Itmol/h/mL in men and 10 If: 8 pmol/h/mL in women The expected interassay coefficient of vartatton should be between 5 and 15% (see Note 3) and the mtra-assay coeffictent of vartation between 2 and 10% (see Note 4). Values will differ for different serum pools (source of apo CII) and emulsions (substrate). 3.2. LPL Mass ELISA Assay 3.2.1. Assay Principle A sandwich ELISA assay IS used A polyclonal antibody (PAb) to capture LPL from samples is coated to the wells of a microtiter plate. The unbound antibody is washed away. Then samples are incubated and, after washing off the free material, a second antibody, horseradish-peroxidase-conjugated, is added to detect the molecules of LPL bound to the plate well. The development is carried out with Orthophenylenediamine (OPD, 4 mg/tablet, Sigma, P-8787). Bovine LPL (or human LPL) is used as standard. 3.2.2. Pre- and Postheparin Plasma Proceed as in Subheading 3.1.1. 3.2.3. Plate Coating 1 Prepare the coating buffer drlutmg the PAb m PBS (10 mM phosphate, 150 rnA4 NaCl, pH 7 4) at 10 pg/mL into a final volume of 10 mL 2. Pour the antibody solutton mto a tray and dispense 100 &/well wtth the multtchannel pipet. 3. Cover the plate with a lid and incubate 4 h at 37°C m a humid atmosphere. 4. Rinse off the coating buffer and wash the wells three times with PBS-T washing buffer. 5 Leave the empty plates on ice.

In Vitro Measurement of LPL and HL

249

3.2.4. Sample Dilution and Incubation Start before the coating mcubatton IS completed. Preparing samples is a tedious procedure that needs attention. 1. Prepare a large tray with crushed ice, where m samples and plates must always be kept. 2 Thawed samples should be mixed and centrifuged in a refrigerated centrifuge to remove the fibrm clot. If the sample contains chylomtcrons, these need to be removed: spin 25,OOOg, 30 min, 20°C Discard the upper phase. 3. Dilute standard (bovine or human LPL) m sample buffer (PBS-T-A-H) to obtain concentrations of 0, 1, 10,25, 50,75, 100, and 200 ng/mL. 4 Dilute samples 1 2 for preH and 1 10 for PHP m sample buffer. 5 Dispense 100 pL diluted standard or sample/well Prepare duplicates for each standard concentration and triphcates for each sample dilution. 6 Cover plates with sealing tape and incubate for 12-16 h at 4°C with constant movement

3.2.5. Conjugated Antibody lncuba tion 1 Dilute horseradish-coqugated monoclonal antibody (MAb) 1.5000 (usually between 1.3000 and 1* 10000) in PBS, 0 2% (v/v) Tween-20,4% (w/v) BSA, into a final volume of 10 mL. 2. Wash plates four times with PBS-T 3. With the multichannel micropipet, dispense 100 L/well. 4 Seal the plate and incubate 4 h at room temperature, protected from light and with constant movement.

3.2.6. Development Prepare the buffer for the substrate during the previous incubation: to 10 mL of 0.05 A4 citrate-phosphate, pH 5.0, add 4 mg of OPD and 4 p.L of 30% H202. The temperature of the buffer should be 20-25°C when added to the plates. 1. Wash plates four times with PBS-T. 2 With the multichannel micropipet dispense 100 pL/well and leave the plates on a mixer to develop for 20-30 mm. When the htghest standard achieves an intense yellow, stop the reaction by adding 50 pL/well of 3 M HzS04 Read the plate at

492 nm (see Note 5) 3. Calculatethe regression curve with the standard and extrapolate concentrations for samples. Once corrected for the dilution factor, values are given m ng of LPL/mI. of plasma. Normal values for males are 70 f 20 ng/mL m preheparm plasma and 540 + 105 ng/mL in postheparm plasma.

3.3. Additional Procedures 3.3.1. Human or Bovine LPL

In orderto have a good standardfor the ELISA assay,a purified preparation of human or bovine LPL 1s required. Bovine LPL can be easily obtained m

250

Vilella and Joven

great amounts (6). ELISA to measure method described methods to purify 3.3.2.

Antibodies

However, human LPL that is the perfect choice to set up an human LPL, is tedious to purify m large amounts. The same for bovine-milk LPL can be used for human milk. Other human LPL from plasma have also been described (7). Against

L PL

The PAb, purified, used to coat the wells, can be an antlbody raised against hLPL or bLPL m goat or hen (8). We recommend using a monoclonal antibody like 5D2 (9) horseradish peroxidase conjugated. The ELISA must be readjusted

for each batch of antibodies. 3 3.3. Conjugation

of Antibodies

with Peroxidase

Several methods have been described to conjugate peroxldase

to antibodies.

The periodate method works very well m our hands (for protocols, see ref. IO). 4. Notes I. More specific hpase selectivity can be achieved using antlbodles that specifically inhibit LPL or HL activity (9,12). 2 If PHP activity values do not differ from PreHP, the most common problem is that samples are not properly drawn, perhaps owing to the fact that heparm IS not adequately injected An other posslblhty 1s that samples are degraded during manipulation 3 If mterassay variation 1svery high, then attention should be paid to increase the accuracy m pipetting Plpet cahbratlon shold be checked Alternatively, substrate not properly prepared may be a source of error 4 If the problem 1swith that mtra-assay variation 1svery high, then the problem may be owing to slight differences in buffer pH, substrate, or source of apo CII If the mcubatlon trmmg 1snot respected, it may also originate high intra-assay variation 5. The following problems may be encountered when reading the plates. a. If no color develops, the reaction has not occurred This may be owing to the following* development buffer has not been properly prepared, H,O, has not been added, substrate has expired, or antibodies may not be working. They should be checked b If the problem 1spoor color development, this may be owing to the following conjugated antibody 1stoo diluted, the peroxldase/IgG conjugation ratio 1stoo low, or substrate has not been properly prepared or added without tempermg c. Blanks too high. unspecific bmdmg of peroxldase (free or ConJugated antibody) to the microtiter well, or low concentration of detergent m washing and sample buffer d Slmllar values m pre- and postheparm samples require mvestlgatlon If both values are very low for normal samples, check the postheparm plasma sample If both values are very high, consider inactive LPL mass and check for activity

In Wtro Measurement

of LPL and HL

References 1 Kern, P. A (199 1) Ltpoprotem hpase and hepatic ltpase Curr Opznlon Llpzdol 2, 162-169 2 Henderson, A. D., Rtchmond, W., and Elkeles, R. S. (1993) Hepattc and hpoprotem hpases selecttvely assayed m postheparm plasma Clan Chem 39,2 18-223 3. Santamarina-FoJo, S. and Dugi, K. A. (1994) Structure, function and role of ltpoprotein hpase m hpoprotem metabolism Curr. Opznzon Lzpzdol 5, 117-125 4 Oltvecrona, G and Ohvecrona, T (1995) Triglyceride hpases and atheroscleroSIS. Curr Opmlon Llpldol 6, 291-305. 5. Hegele, R. A., Ltttle, J. A , Vezina, C., Magutre, G F , Tu, L , Wolever, T S., Jenkins, D J , and Connely, P W (1989) Hepatx lipase deficency Clmtcal, btochemical, and molecular genetic characteristics Arterzoscler Thromb 13,72&728. 6. Iverms, P.-H. and Ostlund-Lmdqvist, A -M (1986) Preparation, charactertzatton and measurement of lipoprotein hpase. Methods Enzymol 129,69 l-704 7. Hayasht, R., Tajtma, S., and Yamamoto, A (1986) Purtficatton and charactertzatton of hpoprotem hpase from human postheparm plasma and its compartson with purified bovine milk lipoprotein lipase. J Bzochem 100,3 19-33 1 8. Vilella, E , Joven, J , FernBndez, M , Vllaro, S , Brunzell, J D , Olivecrona, T , and Bengtsson-Oltvecrona, G (1993) Lipoprotein ltpase m human plasma 1s mamly inactive and associated with cholesterol-rich hpoprotems J Llpzd Res 34,1555-l 564 9. Babirak, S P , Iverms, P -H , FuJimoto, W Y., and Brunzell, J D Detection and charactertzatron of the heterozygote state for lipoprotem lrpase deficrency Arterlosclerosrs 9,326-334 10 Crowther, J R (1995) ELISA theory and practice Methods A401 Blol 42,207-2 18 11 Karpe, F , Ohvecrona, T., Walldms, G., and Hamsten, A (1992) Ltpoprotem hpase m plasma after an oral fat load relation to free fatty acids J Lzpld Res 33, 975-984

17 Cellular

Assays for Lipoprotein

Receptors

Susan Acton and Attilio Rigotti 1. Introduction The activity of a lipoprotein receptor can be measured by analyzmg the bmdmg, mternahzation, or degradation of hpoprotems by cells expressmg the receptor m culture. Receptor activity can also be evaluated mdirectly by measuring the consequences of receptor-mediated cholesterol uptake on cellular cholesterol metabolism using enzyme assays such as those for acylCoA cholesterol acyltransferase (ACAT) and 3-hydroxy-3-methylglutarylcoenzyme A (HMGCoA) reductase (I). In addition, lipoprotem-receptor interactions can be analyzed by evaluating the binding of lipoproteins to cellmembrane preparations, detergent-solubilized receptors, and to receptors immobilized on inert membranes (e.g., ligand-blotting analysis) (2,3). This chapter will focus on lipoprotein-receptor interactions on intact cells. Because of differences in the cellular mechanisms for lipid delivery of the various lipoprotein receptors, not all hpoprotem receptors can be measured with the same cell-based assays. Therefore, depending on the receptor of interest, one or more types of studies will be appropriate and one or more types of tracers will be used. For example, the activity of the low-density lipoprotem receptor (LDLR) can be followed by measuring degradation of the protein component of low-density lipoprotem (LDL) (apo B) because LDLR recognition of the particle leads to internalization and degradation of the particle. However, the high-density lipoprotem (HDL) receptor activity of scavenger receptor class B, type I (SR-BI) does not lead to degradation of the protein component of HDL (most of which is apo A-I) and therefore the HDL-related activity of this receptor cannot be measured using an HDL degradation assay (4). The choice of labeled tracer used will also be determmed by the receptor and type of study being done. The two most commonly used From

Methods Edlted by

,n Molecular J M Ordovas

B/o/ogy, Vol 110 Lpopfofem Protocols 0 Humana Press Inc , Totowa, NJ

253

254

Acton and Rigotti

tracers to measure the activity of lipoprotem receptors are modifications to the protein component of the lipoproteins (usually by iodination), and llpophtllc tracers, which are generally noncovalently added to lipoprotems. It should also be noted that onoff rates can differ dramatically from receptor to receptor, therefore, incubation and wash conditions can affect the results of each assay.Therefore, for each receptor activity that is being measured, it is prudent to tailor the binding and washing condttions appropriately. It is also necessary to match the cell culture conditions to the cell type being used. When studies are being performed on transfected cells, one can clearly determine the role of the receptor of interest in cellular lipoprotem metabolism by comparing the findings to those using untransfected cells. However, for primary cells that may express many different types of lipoprotein receptors, it becomes much more difficult to interpret the role of an mdividual receptor m ltpoprotein metabolism. For example, HDL bmdmg to hepatocytes could be at least explained by SR-BI, LDLR, and LRP if there IS some apoE present m the HDL preparation. In this case,it would be necessary to prepare apoE-free HDL and utilize receptor-specific blocking agents (selective competitors, antibodies, and so on) to differentiate between the contributions of the various receptor types to lipoprotem metabolism. Even with all possible precautions, one may still be measuring activities of unknown receptors This chapter gives general guidelines with examples of cell culture conditions for transfected (CHO) cells. For culture conditions of other cell types the reader is referred to refs. I and 3. This chapter will focus on the methods used for cell-based lipoprotem receptor assays. 2. Materials 2. I. Equipment 1. 2 3 4. 5. 6. 7 8. 9. 10. 11.

CO, incubator N, tank y-Counter (Beckman Instruments, Fullerton, CA) Vortex. Rotary tube mixer. Iodmation hood Beckman TJ6 centrifuge (Beckman) Wallac mtcrobeta counter (Wallac Oy, Turku, Finland) Fluorskan 96-well fluortmeter plate reader (Labsystems, Rochester, NY) Vacuum desiccator (Labconco, Fisher Scientific, Hampton, NH). SANYO MSE Somprep 150 Ultrasonic Disruptor (Integrated Services, TCP, Palisades Park, NJ) 12. Spectrophotometer. 13 Orbital Shaker. 14 Ultracentrifuge (for hpoprotem isolation after labeling with lipid tracers).

Cellular Assays for Lipoprotein Receptors

255

2.2. Reagents and Supplies 2.2.1. Labeling of Isolated Lipoprotein Fractions. 1. LDL (PerImmune, Rockville, MD, or Molecular Probes, Eugene, OR, cat no. L3486) Lipoprotein fractions can also be isolated as described in Chapter 6 of this volume. 2 HDL (PerImmune). 3 1.O M Glycine, pH 10 0 (it makes the solution basic wtthout denaturing the protein, this mmlmizes lipid labelmg). 4 2.0 mCi of carrier-free Na1251 (Amersham Life Science, Arlmgton Heights, IL). 5 1:20 Iodine monochloride (ICI) solution (950 pL of 2 MNaCl plus 50 pL of 0.02 A4 ICI) 6. Sephadex G-25 columns (1 x 10 cm PDlO, Pharmacla) for desalting and separation of free iodine. 7 0.22~pm Filter (Mrlhpore). 8 Bolton-Hunter reagent (2 mC1, N-succin~mrdyl-3,4-hydroxy-5-[‘251]rodophenyl propionate) -2000 Ci/mmol (Amersham) 9 Sodium borate buffer, 0 1 M, pH 8 5 10 Saline-ethylenediamme tetra-acetic acid (EDTA). 0 15 M NaCl, 0 0 1% EDTA, pH74

2.2.2. Following the Fate of the Lipid Component of Lipoproteins 11 12 13 14 15. 16

Egg-yolk phosphatidylcholme (Fluka, Busch, Swrtzerland) Cholesteryl oleate (BDH Chemicals, Poole, Dorset, UK). Chloroform/methanol (1: 1, v/v) [ la,2a(n)-3H] cholesteryl oleoyl ester, 3@-60 Ci/mmol (Amersham) [ 1a,2a(n)-3H] cholesterol oleyl ether, 30-60 C~/mrnol (Amersham) DiI (l,l’-dloctadecyl-3,3,3’,3’-tetramethyl~ndocarbocyan~ne perchlorate, Molecular probes, Eugene, OR, cat. no D-282) in dimethyl sulfoxlde (DMSO) (3 mg/mL)

2.2.3. Following the Fate of the Protein Component of Lipoproteins 17. Radiolabeled lipoprotems (labeled as described here or purchased from PerImmune). 18. CHO cells (American Type Culture Collectron, Rockvllle, MD) 19. 6-Well dishes. 20 Adequate tissue-culture medium. 2 1. Phosphate buffered saline, calcium-and magnesium-free (CMF-PBS) 140 mMNaC1, 2 6 mM KCl, 10 mMNa2HP04, 1 8 mM KH2P04, pH 7.4 Store at 4°C 22. Newborn calf lipoprotem deficient serum (NCLPDS) (PerImmune) 23 10 mM HEPES. 24. PBS with calcium and magnesium. 200 mL 0.90% (w/v) NaCl, 8 mL 1.15% (w/v)KCl, 6 mL 1 22%(w/v) CaCl,; and 2 mL 3.82% (w/v) MgS04.H20 Add carefully and with constant stirring 40 mL of 0 1 M phosphate buffer, pH 7 4 (17 8 g of Na2HP04.H20 + 20 mL of 1 N HCI, diluted to 1 L).

256

Acton and Rigott/

25 Trts wash buffer. 50 mM Tris-HCl, 0.15 MNaCl, pH 7.4. Use at 4°C 26. Bovine serum albumin (BSA) wash buffer: 50 mM Tris-HCl, 0 15 M NaCl, 2 mg/mL BSA, pH 7 4. Use at 4’C. 27 Dextran sulfate LDL release buffer: 4 mg/mL dextran sulfate, 50 mM NaCl, 10 rnA4 HEPES, pH 7.4 28 0 1 NNaOH 29 12 x 75-mm Borosihcate tubes. 30 20% Trmhloroacetic acid (v/v) m water 3 1. 40% Potassmm iodide (KI) 32. Hydrogen peroxide. 33 Chloroform. 34 Isopropyl alcohol 35. 0.1 A4 KCl, 10 mA4 Tris-HCl, 1 mM EDTA, 0.025% NaN,, pH 8 0 36. Phospholtpids kit (Boehrmger Mannhelm, Germany). 37. DMSO. 38. Fatty acid-free BSA (FAF-BSA) (Sigma). 39 Liquid scmtillation mixture (Scintiverse, Fisher) 40. Potassium bromide (KBr) 41. Lipoprotem deficient human serum (LPDS): d >l 2 1 g/mL ultracentrifugated portion of serum. 42 PBSE (PBS + 10 mMEDTA) 43 0 45-pm Filter (Milhpore Corporation) 44. FAF-BSA medium* serum-free medium containing 0.5% FAF-BSA and 10 mA4 HEPES 45 96-Well plates

3. Methods 3.7. Following

the Fate of the Protein Componenf

The protein components of lipoproteins can generally be labeled by todination, which 1s easily detected tn small quantities. A reliable and reproducible method for iodinating LDL is the iodine monochloride method (I). HDL can also be rodinated by this method for cell-association studies of SR-BI, or can be iodinated using Iodobeads (Pierce, Rockvrlle, IL) although there tends to be some batch-to-batch varratron m background binding to untransfected cells (4). Apo E-rich lipoproteins (high density lipoprotein subfraction 1 [HDLt] and high density ltpoprotem subfractron c [HDL,]) are not well-iodinated on tyrosmes (5). Thus, these proteins should be labeled using the Bolton Hunter method (2,6). Currently, lipoprotems can also be purchased m todinated form from PerImmune (Rockville, MD) (see Note 1).

3.1.1. Radioiodination of LDL Using the Iodine Monochloride Method 1 Dialyze isolated LDL (density cut. 1.030-l ,050 kg/L) against saline-EDTA 2 Bring the concentration of LDL to -5 mg/mL.

257

Cellular Assays for Lipoprotein Receptors

3 Equthbrate PD- 10 columns, First, remove the cap and pour off the preservative; second, cut bottom off column and mount m stand with an empty beaker underneath; third, run 25 mL of water through, allow to dry, and recap. 4 Place 1 mL of the LDL solution (or the required amount for the specific experiment) into a 15mL conical tube and label approprtately (make sure LDL fractions are fresh and have never been frozen, otherwise the efficiency of the iodmation will decrease dramatically). 5. Add 0.5 mL of 1.O M glycme, pH 10.0, to the LDL sample. 6. Prepare the iodmation hood. All iodmations must be done m air tight, negative pressure hoods equipped with charcoal filters that trap volatile radioactive iodme (a byproduct of the radioiodmatton reaction). The hood must also be vented mto an accessory laboratory hood to the outside. The hood should contam a. Plastic-backed paper on bottom. b. Triple layer trash bag in small box. c. Ring stand for PD-10 column. d Rotary tube mixer e. Test tube rack for samples and reagents

7. Add half of the 5 pL required of Iodine monochlorlde solution (seeNote 2) to the vial containing the 2 0 mCt of carrier-free Na*251 isotope. 8 Rapidly withdraw isotope-ICl solution with a tuberculin syringe and add to the lipoprotein fraction 9 Add the other 2.5 pL of the ICI solution to the isotope vial to wash the remaining isotope. 10 Add this wash to the hpoprotem fractton. 11 Cap the tube tightly and invert on rotator for 1 mm. 12 Uncap and dispose of cap 13 Allow the reaction sample to sit for 10 min at room temperature for the free iodme to escape. 14. Add require volume of PBS to the sample to raise the volume to 2 5 mL 15 Separate bound from free radiotodide by runmng the reaction mixture over the PD- 10 column, Add sample to column with a disposable plastic ptpet. 16 Allow the sample to run mto the column, while collectmg the eluate Discard this first eluate. 17 Place a new collection tube on stand and add 3.5 mL of saline-EDTA to the column. Collect desalted eluate 18 Sterilize the labeled LDL by membrane filtration through a 0.22-pm filter. Keep it at 4°C and do not freeze tt at any time. Usually specific acttvmes of 5OMOO cpmng protein are obtained

3.1.2. Radioiodination of HDL by the Bolton-Hunter Method The protocol describes the lodination of HDL. The same procedure can be applied to other hpoprotem fractions. 1 Dialyze the HDL preparatton (l-5 mg of protein/ml)

agamst borate buffer

2. Dry the Bolton-Hunter reagent m the shipping vial using mtrogen (keep on Ice).

258

Acton and Rigotti

3. Add 1.5 mL of the dialyzed HDL directly to the vial containing the dried reagent. 4. Incubate for 30 min on ice. Mix using a syringe inserted through the vial seal. 5. Transfer the solution to a dialysis bag and dialyze against saline-EDTA for 48 h in the cold room. Sterilize labeled HDL through a 0.22-m filter. Usually specific activities of 800 cpm/ng of protein are obtained.

3.1.3. Lipoprotein Binding at 4 “C One of the simplest ways to measure the binding of lipoproteins to cells expressing the receptor of interest is to stop the process of receptor-mediated

endocytosis by cooling the cells to 4°C. The cells are then incubated with prechilled radiolabeled ligand. After equilibrium is reached, the cells are then washed several times to remove unbound ligand and then lysed and counted.

Because this study is done at equilibrium, this type of analysis allows one to measure the dissociation constants of binding sites on the surface of cells by Scatchard analysis (7). Note, however, that the values obtained will be dissociation constants at 4”C, which may differ dramatically from dissociation constants at 37°C (see Note 3). 1. Set cells in a 6-well dish in the appropriate growth medium so that they are confluent and well-attached for the assay period. For transfected CHO cells, this would be equivalent to setting 250,000 cells/well in a 6-well dish in 2 mL of growth medium followed by incubation for 48 h in a humidified, CO, incubator. The number of cells to be plated and the time will vary depending on the type of cell being used as well as the growth conditions. There may also be experiments in which the investigator may not wish the cells to be confluent during the assay period (see Note 4). 2. Transfer the dishes onto ice to cool for 30 min. Remove the medium and wash twice with 2 mL ice-cold PBS to remove unlabeled lipoproteins that were in the growth medium. Add 1 mL of labeled lipoprotein at 4°C (generally between 0.1 and 50 pg protein/ml is a good range) in medium containing 5% NCLPDS and 10 mM HEPES. NCLPDS can be made (1,8) or purchased from PerImmune. To determine the amount of nonspecific binding in the assay, add 40X unlabeled lipoproteins to parallel wells. This unlabeled lipoprotein will compete with the labeled lipoprotein for specific sites, but will not have an effect on the nonspecific binding of label. Incubate between 1 and 6 h at 4°C with slow shaking. The length of time needed will depend on the rate at which equilibrium is reached for a particular receptor-ligand interaction which should be determined empirically. 3. After incubation, remove the medium from the wells. 4. Wash the wells gently. For the LDL receptor and the class-A scavenger receptors, this is equivalent to three rapid washes with 2 mL cold BSA wash buffer, followed by two IO-min washes with 2 mL cold BSA buffer, followed by a final rapid wash with 2 mL cold Tris-buffer. 5. Lyse the cells with 1 mL 0.1 N NaOH for 15 min at room temperature and count 500 mL in a y-counter. Determine the protein content of the lysate by Lowry analysis.

Cellular Assays for Lipoprotein Receptors

259

6. Specific binding is determined by the difference between the total binding and the nonspecific binding. It is reported as ng lipoproteitimg cell protein and in this case is calculated as 2 x [(cpm of sample)/(mg cell protein) - (cpm of nonspecific sample)/(mg cell protein)]/specific activity of the ligand.

3.1.4. Lipoprotein Binding at 37°C Binding at 37OC can be determined if a reagent is identified that will release the cell-surface-bound ligand at 4°C so that it can be separately determined from internalized ligand. For example, LDL can be released from the LDL receptor using dextran sulfate (I). 1. Set cells as in step 1 of Subheading 3.1.3. 2. Remove medium, wash once with 2 mL PBS and add 1 mL labeled lipoprotein (between 0.1 and 50 ~18protein/ml) in medium containing 5% NCLPDS. For determining nonspecific binding, add 40X unlabeled lipoprotein to parallel wells. Incubate for the desired period of time at 37°C (5 h is sufficient for achieving equilibrium with the LDL receptor at 37”C, but the time required should be determined empirically for each receptor). 3. After incubation, remove the medium from the wells (this can be discarded, or examined for degraded ligand as indicated in Subheading 3.1.7.). 4. Wash the wells gently as in step 4 of Subheading 3.1.3. 5. Add 1.5 mL of release solution at 4°C (for LDL binding to the LDL receptor: 4 mg/mL dextran sulfate, 50 mM NaCl, 10 mM HEPES, pH 7.4). Incubate for 1 h on a shaker at 4°C. 6. Transfer the release solution to 12 x 75-mm tubes and count 0.75 mL to determine binding. 7. Lyse the cells in 1 mL of 0.1 N NaOH and determine the protein content by Lowry analysis. 8. Specific binding is determined as in step 6 of Subheading 3.1.3.

3.1.5. Lipoprotein Internalization at 37°C To determine the amount of internalized ligand, cells that have had cell-surface ligand released (as indicated above) are washed, lysed, and counted. 1. Follow steps l-6 in Subheading 3.1.4. Continue by washing the cells twice with 2 mL Tris-wash buffer. 2. Add 1.5 mL of 0.1 N NaOH and incubate for 15 min at room temperature. 3. Resuspend the lysate and count 0.75 mL to determine the amount of ligand internalized. 4. Determine the amount of protein in the cell lysate by the method of Lowry. 5. The amount of internalized ligand is determined by the difference between the total internalized and the amount internalized in the presence of excess unlabeled ligand (nonspecific internalization). It is reported as ng lipoprotein/mg cell protein and in this case is calculated as 2 x [(cpm of sample)/(mg cell protein) (cpm of nonspecific sample)/(mg cell protein)]/specific activity of the ligand.

Acton and Rigotti

260

3.7.6. Lpoprotern Cell Assocratron at 37°C Cell association (which does not separate cellsurface from internalized hgand) at 37°C can be determmed by incubating the cells with radiolabeled ligand at 37°C for different periods of time, followed by washing, lysis, and counting. Follow steps l-2 of Subheading 3.1.4. After mcubatton, remove medium from wells. For analysis of the LDL receptor and the class A scavenger receptors, wash the wells gently but raptdly three ttmes wtth 2 mL cold BSA wash buffer, followed by two lo-min washes with 2 mL cold BSA buffer, followed by a final rapid wash with 2 mL cold Tris-wash buffer For SR-BI, wash rapidly twtce wtth 2 mL cold Tris-wash buffer and once with 2 mL BSA-wash buffer. Lyse the cells with 1 mL 0.1 N NaOH for 15 mm at room temperature and count 500 $ of the lysate m a gamma counter. Determine cell protein content by Lowry analysis. The amount of spectfic cell-associated ltgand is determined by the difference between the total cell-associated hgand and the amount of cell-associated ligand in the presence of excess unlabeled ligand (nonspecific cell-associated llgand). It IS reported as ng hpoprotem/mg cell protem and m thts case IS calculated as 2 x [(cpm of sample)/(mg cell protein) - (cpm of nonspecific sample)/ (mg cell protem)]/specrfic activity of the hgand

3.1.7. Lipoprotein Degradation at 37°C Degradation is a measure of the amount of hgand that IS degraded as a result of receptor-mediated uptake into the cells. In the case of lipoproteins that are labeled by iodinatron of their protein moiety by the iodine monochloride method, one measures the amount of radtolabeled tyrosme that is released by the cells. This is a slightly more complex assay than either binding or uptake because one must remove the undegraded radiolabeled lipoprotein as well as the free iodide that is m solutron m order to measure only degraded hgand. 1. Follow steps l-2 of Subheading 3.1.4. 2 Add labeled hpoprotem in medium to two wells that do not contam cells and allow these to Incubate along with the wells contammg the cells. These will serve as controls for degradation m the absence of cells (no cell blanks). If the no-cell blank

controls

are not mcluded,

it ~111 not be possible

to dlstmgulsh

between

degradation that has occurred during storage of the iodinated lipoprotem, and that caused by the cells

3. After Incubation (5 h for LDL receptor), transfer the medtum from the wells mto 12 x 75-mm glass tubes contammg 1 mL of 20% trtchloroacettc actd, vortex, and

incubate 30 min at 4°C

Cellular Assays for Lipoprotem Receptors

261

4. During thts mcubation, wash the cells on the plate twice with 2 mL Tris-wash buffer. Add 1 mL of 0 1 NNaOH/well and incubate for 15 min at room temperature to allow the cells to lyse. Assay the protein content (e.g , Lowry assay) of 50-100 pL. 5 After the TCA precipitation, spin the tubes for 10 mm at SOOg in a Beckman TJ-6 centrifuge at 4°C. 6 Transfer 700 pL of the supernatant to 12 x 75-mm glass tubes containing 7 pL of 40% KI. While vortexing, add 28 mL hydrogen peroxide to each tube. After all tubes are done, revortex all Incubate at room temperature until supernatant IS clear and a black precipttate is on the bottom (generally takes 5-l 0 min). Revortex if the solution contmues to be cloudy 7. Add 1 mL of chloroform to each tube and vortex thoroughly Incubate for 5 mm at room temperature. 8 Spur the tubes for 5 min at 8008 m Beckman TJ6 centrifuge 9 Very carefully so as to not disturb the interface, remove 0.5 mL of supernatant to 10 x 75 tubes. Respin if layers are disturbed and try again. 10 Cork and count in a y-counter. 11. Data IS reported as ng ligand degraded/5 h/mg cell protein. Calculations are done as follows: a. Subtract the no-cell blank cpm from the sample cpm b Multiply the product by four (to correct for the relative amounts of sample taken at each step) c. Divide the product by the specific activity of the iodmated ligand (in cpm/ng protein) d. Divide the product by the amount of cell protein in the well to get the final amount of total degraded hgand/5 h/mg cell protein To calculate specific degradation, subtract the value calculated for degradation in the presence of excess unlabeled hgand from the value calculated for degradation m the absence of excessunlabeled ligand.

3.2. Determining

the Fate of the Lipid Component

The fate of the lipid components of HDL can be followed separately from the protein components by using lipid-labeled tracers, Two commonly used tracers are the lipid-soluble fluorescent dye, DiI, and radiolabeled cholesterol ester (or ether). To compare the fates of the lipid and protein components, one can perform parallel experiments with the protein component labeled m one experiment and the lipid-labeled m the other, or use dual-labeled hpoprotems m a single experiment. These studies are particularly valuable when examinmg selective lipid uptake into cells, for example, as has been demonstrated for SR-BI (4). The lipoproteins labeled with cholesteryl ester can also be used for assays of hydrolysis of the hpid moiety of the lipoprotems and for assays of reesterification of the fatty acids (see Note 5).

262

Acton and Rigotti

3.2.1. Preparation of 3H-Cholesteryl Ester Labeled HDL 1. Add egg-yolk phosphatidylcholine and cholesteryl oleate dissolved m chloroform methanol (1.1, v/v) in a mass ratio of 60: 1 to a 25-n& scintillation vial. 2 Dry the mixture under a nitrogen stream Carry out this evaporation mstde a solvent hood. 3. Add 250 pCt of [ 1,2,6,7-3H] cholesteryl oleoyl ester (or [ 1,2,6,7-3H] cholesteryl oleoyl ether) 4 Dry again under nitrogen and place overnight rn a vacuum desiccator to remove any trace of solvent. 5 Add to the vial 7.5 mL of 0 1 MKCl, 10 mMTns-HCl, 1 WEDTA, 0 025% NaN,, pH80 6 Maintaining the contents of the vial under a constant nitrogen stream, somcate using a MSE Somprep 150 for 40 mm using an amplitude of 12 kHz. The temperature should be mamtamed at 52°C 7 Adjust the emulsron to a density of 1 21 g/mL wtth KBr and separate the donor 3H-labeled partmles by density gradient ultracentrifugation (as described m Chapter 6) The density of these particles is about 1 03 g/mL 8. Measure the phospholipid content of the particles using a calorimetric assay 9 Mix 2 mL of human hpoprotem-detictent serum (dtalysed extensively against salme-EDTA) with 1 mL of human HDL (containing 3.5 mg of protein) and with the previously 3H-labeled donor particles The ratio of HDL protein mass to donor particle phospholipids should be 8.1 10 Incubate the mixture for 5 h at 37°C m a shaking water bath, mamtammg the mixture under nitrogen atmosphere 11 Reisolate the HDL by gradient density ultracentrifugation (as described in Chapter 6) after raising the density of the solution to 1 2 1 g/mL with KBr 12 Dialyze exhausttvely the isolated HDL against PBSE and filter through 0 22- or 0.45~pm filters. In some protocols, a final step could include dtalysts against culture medium or any other appropriate solution.

3.2.2. 3H-Cholesteryl Ester/Ether Labeled HDL Uptake Assay: 6-Well Plate Method 1 Set cells as in step 1 of Subheading 3.1.3. 2. Remove medium, wash once with 2 mL PBS and add labeled HDL (between 1 and 20 pg protein/ml) m 1 mL FAF-BSA medium For nonspecific background, add 40X unlabeled HDL to parallel wells. Incubate the desired length of time at 37°C (we suggest trying 5 h first) 3. Chill cells on ice and wash rapidly with three times with 2 mL Trts buffer (twice with and once without BSA [2 mg/mL]). 4. Extract cholesteryl oleate (or cholesteryl ether) with 1 mL of tsopropyl alcohol (30 mm at room temp wtth gentle shaking) Remove 0 5 mL and count in liquidscintillation mixture 5. The amount of specific cholesteryl ester (ether) uptake is determined by the dtfference between the total uptake and the amount of uptake m the presence of

263

Cellular Assays for Lipoprotein Receptors

excess unlabeled hgand (nonspecific uptake). It is reported as ng cholesterol ester (ether)/mg cell protem and in this case is calculated as 2 x [(cpm of sample)/(mg cell protein) - (cpm of nonspecific sample)/(mg cell protem)]/specitic activity of the cholesteryl ester (ether).

3.2.3. 3H-Cholesteryl Ester/Ether-Labeled 96-Well Plate Method

HDL Uptake Assay:

1, Seed transfected cells at 50,000 cells/well in a 96-well plate. Incubate for > 16 h at 37°C in a humidified, CO* incubator. 2. Remove medium and wash two times in FAF-BSA. 3 Add 3H-cholesteryl ester HDL at 2.5 to 30 pg/mL (80 &/well) in FAF-BSAmedium. For nonspecific background, add 40X unlabeled HDL to parallel wells. Incubate desired length of time at 37°C (5 h is reasonable to try first) 4. Remove medmm and wash once with PBS contammg 2 mg/mL FAF-BSA Add 100 & Scmtiverse and read the plate on a Wallac microbeta counter 5. The amount of specific cholesteryl ester (ether) uptake is determined by the difference between the total uptake and the amount of uptake in the presence of excess unlabeled &and (nonspecific uptake). It is reported as ng cholesterol ester (ether)/well and in this case 1s calculated as [(cpm of sample)/well - (cpm of nonspecific sample)/well]/specific activity of the cholesteryl ester (ether)

3.2.4. Analyzing Lipid Uptake Using Dil-Labeled Lipoproteins DiI-282 is a lipid compound that integrates noncovalently into the outer lipid layer of lipoprotems. Lipoproteins labeled with DiI have several applications: 1 Analysis of the cellular mechanism of lipid uptake by comparison with results using lipoproteins labeled in the protein component 2 Expression cloning of new lipoprotein receptors. 3. Fluorescence-activated cell sorting (FACS) for selectton of a subpopulatton of cells with different lipoprotein receptor activities. 4. Confocal mtcroscopy and video imaging analysis. DiI-labeled lipoprotems can be prepared by a simple (2,3) or can be purchased from PerImmune.

procedure

in the lab

3.2 4 1. LABELING OF HDL WITH THE FLUORESCENT PROBE DII The protocol here describes the fluorescence labeling procedure can be applied to other lipoprotein fractions. 1. 2. 3. 4 5

of HDL.

The same

Dialyze the HDL preparation (l-2 mg of protein/ml) against saline-EDTA Add 2 mL of LPDS per mg of HDL Filter the solution through a 0.45-pm filter. Add 50 pL of DiI per mg of HDL protein. Mix gently Incubate the mtxture for 8-15 h at 37°C.

264

Acton and Rigotti

6. Adjust density at 1.21 g/mL with KBr and reisolate the HDL by ultracentrifugatton (as described in Chapter 6). 7. Dialyze the HDL fraction against saline-EDTA. 8. Filter through a 0.45+m filter. 3.2.4.2.

LIPID UPTAKE USING A ~-WELL FORMAT

1. Set cells as in step 1 of Subheading

2. 3 4. 5 6. 7 8

3.1.3. Some wells will be processed m parallel (except for the DMSO extraction) and used for NaOH lysts and Lowry analysis. Remove medium and add 750 &/well of DtI-HDL (l-20 &mL) m FAF-BSA medium. Incubate 5 h at 37’C m humidified incubator with 5% CO, Wash twice rapidly with 2 mL/well of PBS. Drssolve cells m DMSO (100%) 2 n&/well. Make a standard curve of DII-HDL (0.1-10 mg/mL) in DMSO Read the standard curve and 1 mL of each sample on a fluonmeter (550 nm excn, 565 nm emtss). Calculate the amount of DtI-HDL taken up by comparing the fluorescence readings to those of the standard curve which should be approximately linear Uptake values are expressed as ng lipoprotem/mg cell protein

3.2.4.3.

LIPID UPTAKE USING A 96-V&u

FORMAT

1. Seed transfected cells at 50,000 cells per well m a 96-well plate. Incubate for >16 h at 37°C in a humidified CO2 incubator. 2. Remove medium and wash 2 times m FAF-BSA-medium 3 Add DiI-HDL at 2.5-20 pg/mL (80 &/well) m FAF-BSA-medium. To determine the nonspecific background, add 40X unlabeled HDL to parallel wells Incubate desired length of trme at 37°C (5 h is sufftctent to give a posmve signal) 4 Remove medium and wash once with PBS contammg 2 mg/mL FAF-BSA, then once with PBS (no BSA). Solubthze with 50 $0 1 NNaOH and read the plate on a 96-well fluonmeter plate reader at 550 nm excitatton, 565 nm emission 5. Calculate the amount of DtI-HDL taken up by comparing the fluorescence readmgs to those of a standard curve of DiI-HDL dtssolved m 0 1 N NaOH, whtch should be approximately linear Values are expressed as described m step 8 of Subheading

3.2.4.2.

4. Notes 1 One should be aware that most iodination

procedures cause some oxidation,

therefore, if one wishes to mmlmlze oxldatlon of lipoprotems,

some precautions

should be taken. In general, hpoprotems should be isolated and stored under anttoxtdtzmg conditions This includes the use of antioxidants (EDTA, butylated hydroxytoluene [BHT]) when isolating LDL, the gassmg of buffers with N2 before use, and, where possible, all steps should be done m the dark. LDL should be iodmated as soon as possible after it is isolated, taking similar anttoxtdation

Cellular Assays for Lipoprotem Receptors

265

precautions in the iodmation procedure. The iodinated LDL should be used within a few days if oxidation is of great concern, but can m general be used up to 1 mo before degradation of the ligand becomes excessive 2. The amount of ICI to add depends on the amount of apohpoprotem wtthin the hpoprotein fraction and its “average” molecular wetght The calculation IS as follows: ICl (mg) = [(mg protein) x 62 S]/[(mol wt of the protein) x 0.501 3 Many receptors have shown quite dtfferent affinittes for their ligands at 4°C vs 37°C so that binding studies at 4°C can only be used to determine the Kd of the receptor at 4°C. The affinity of SR-BI for HDL at 4°C appears to be very weak and it is therefore difficult to measure by this method, whereas it is relatively straightforward to measure the affimty of LDL for the LDL receptor at 4°C. In addition, overwashmg of lipoprotems bound to SR-BI expressing cells causes loss of specific binding. 4. It is crucial to mamtam consistent cell growth and assay conditions to obtain valid conclusions between cell strams from different patients and from the same strain on different days. Different batches of reagents (e-g , lipoprotems, NCLPDS, FAF-BSA) can also markedly influence receptor acttvmes, particularly m primary lmes; therefore, care should be taken to control for possible heterogeneity m assay conditions. 5. It is not trivial to achieve a high level of specific activity when labeling hpoprotems wtth cholesterol ester or cholesterol ether There are a number of laboratortes with shghtly dtfferent methods (P-II). In additton to exchange reactions, lrpoprotems can be labeled by reconstitution methods (22,13) Conversely, one can purchase 3H-cholesteryl ester-labeled HDL from Amersham as part of their CETP kit This cholesteryl ester is shipped in a solution contammg albumin and it IS not possible to get an accurate determmation of the amount of HDL protein in this kit from the manufacturer

Acknowledgments We would like to thank Monty Krieger and members of his laboratory for guidance m the use of some of the techniques presented here. Attilio Rigotti was supported by a Howard Hughes Medical Institute Postdoctoral Fellowship. References 1 Goldstem, J. L., Basu, S. K., and Brown, M. S. (1983) Receptor-mediated endocytosis of low density hpoprotem m cultured cells. Meth Enzymol 98,241-260. 2 Innerarity, T. L , Pitas, R E , and Mahley, R. W. (1986) Lipoprotein-receptor mteractions. Meth Enzymol 129, 542-565 3 Arnold, K. S., Innerartty, T L , Pitas, R E., and Mahley, R. W (1992) Lipoprotein-receptor interactions, m Lrpoproteln Analysts A Practical Approach (Converse, C. A and Skmner, E. R , eds.), Oxford University Press, London, pp. 145-168.

266

Acton and Rigotti

4 Acton, S , Rigotti, A , Landschulz, K T , Xu, S., Hobbs, H. H , and Krueger, M

5

6

7. 8 9

10

11

12 13

(1996) Identification of scavenger receptor SR-BI as a high density lipoprotem receptor. Science 271,5 18-520. Innerarity, T. L , Pitas, R. E , and Mahley, R W (1979) Bmdmg of argmine-rich (E) apoprotem after recombmatton with phosphollpid vesicles to the low density hpoprotem receptors of fibroblasts. J Blol Chem 254,4 18-690. Bolton, A. E and Hunter, W. M (1973) The labelling of proteins to high specific radioactivities by conJugation to a ‘251-contaming acylatmg agent. Blochem J 133,52%539 Scatchard, G. (1949) Ann N Y Acad Scz 51,660 Krueger, M (1986) Isolation of somatic cell mutants wtth defects m the endocytosis of low-denstty lipoprotein. Meth Enzymol 129,227-237. Pittman, R. C., Glass, C. K., Atkinson, D., and Small, D. M. (1987) Synthetic high density lipoprotem particles Application to studies of the apoprotem specificity for selective uptake of cholesterol esters. J Bzol. Chem 262,243-542. Pieters, M. N., Schouten, D., Bakkeren, H. F., Esbach, B., Brouwer, A , Knook, D L , and van Berkel, T J. (199 1) Selective uptake of cholesteryl esters from apohpoprotem-E-free high-denstty lipoprotems by rat parenchymal cells m viva is efficiently coupled to bile acid synthesis. Bcochem J 280, 359-365 Stem, Y., Dabach, Y , Hollander, G., Halperm, G , and Stem, 0 (1983) Metabolism of HDL-cholesteryl ester m the rat, studied with a nonhydrolyzable analog, cholesteryl lmoleyl ether Blochtm. Blophys Acta 752,98-105 Jonas, A. (1986) Reconstitution of high-density hpoprotems. Meth Enzymol 128, 553-582 Walsh, M. T and Atkmson, D (1986) Reassembly of low-density lipoprotems Meth Enzymol 128,582-608

Measurements of Proteoglycarwlipoprotein Interaction by Gel Mobility Shift Assay Eva Hurt-Camejo,

German Camejo, and Peter Sartipy

1. Introduction Proteoglycans (PGs) of the arterial intima form complexes with apo B- 1OOcontaining lipoproteins. Several lines of evidences indicate that this process IS a key pathogenic event for the development of atherosclerosrs (for review, see ref. 2). First, retained apo B m early as well as advanced lesions IS closely associated with arterial PGs (2,3). Second, purified arterial PGs, partrcularly those from lesion-prone sites, bmd low-density lipoprotein (LDL) in vitro (4). Third, LDL from patrents with coronary-artery disease has a higher affinity for arterial PGs than LDL from apparently healthy subjects (5). The interaction between apo B-lOO-contaimng lipoproteins and PGs is of romc nature and 1smodulated by structural features of both types of macromolecules (6-9). PGs are macromolecules that are found predommantly on the cell surface and in the extracellular matrix, and that contam sulfated polysaccharides called glycosammoglycans (GAG). GAGS are polymers of drsaccharrde repeats, which are mostly highly sulfated and negatively charged. The main GAG in PGs of the arterial wall are chondrortm sulfate 4 and 6 (C45,C65), dermatan sulfate, heparan sulfate, and hyaluromc acid. Three properties of the PGs appear to control their association with apo B-lOO--containing lipoproteins: nature of the GAG chains, degree of sulfation of the disaccharide subunits, and length of the chains (9). The C6S-rich Versrcan IS the most abundant PG m the extracellular arterial intima (I&II). Versican can form large aggregates with hyaluronic acid and the monomers can contain several GAG chains with different length and proportions of C6S, C4S, and dermatan. The susceptibility to atherosclerosis of human arteries correlates with the in vitro affinity of then chondroitin sulfate-GAG (CS-GAG) for LDL, and the chains of GAG From

Methods /n Molecular EUology, Vol 1 IO Lpoproteln Protocols Edlted by J M Ordovas 0 Humana Press Inc , Totowa, NJ

267

268

Hurt-Camejo,

Camejo, and Sartipy

with high LDL affinity are longer than those of low affinity (4). We found that proliferatmg human artertal smooth-muscle cells (HASMC) synthesize CS-rich PG (versican) with longer CS-GAG chains than when quiescent. Furthermore, this CS-PG showed two to three times higher affinity for LDL at physiological tonic conditrons than the PG and GAG of resting cells. The segments m the apo B- 100 that are responsible for the assoclatton of LDL with CS-GAG rich arterial verslcan are: 3 147-3 157 (SVKAQYKKNKHRH), 3359-3377 (RLTRKRGLK LATALSLSNK), and 2 106-2 121 (RQVSHAKEKLTALTKK). These sequences have a high probability of residing on the lipoprotein surface and have an excess of positive amino acid residues (12,13). The distribution of posmve charges m the apo B-100 segments coincides reasonably well wtth the distribution of negative charges on an octasacchartde segment of C6S (14). The experimental results prevrously descrrbed suggest that high affimty of LDL for arterial CS-PG m vitro could reflect tendency of the lipoprotein to interact with the intima in vivo, and that this property could contribute to atherogenesis. This hypotheses is supported by clmical studies. Results with patients that suffered a myocardial infarction before 50 yr of age showed that high affinity of LDL for human CS-PG was independently associated with the disease (15). A recent study found that small, dense LDL from patients with the “atherogenic llpoprotem phenotype” have high affimty for human arterial CS-PG (16). Furthermore, apo B hpoprotem-lowering drugs, e.g., simvastatm and gemtibrozil, also decrease the LDL affinity for CS-PG, suggestmg that part of then antlatherogemc action may be associated with a reduced entrapment of LDL m the PGs of the arterial mtima (17). The possible contrtbution of LDL-PG mteractton to atherogenesis and its clmical consequenceshas prompted the development of methods for measuring the affinity of LDL for PGs ex viva. In this chapter, we describe the apphcatton of gel mobility shift assay to evaluate the interaction between LDL and PGs or GAGS synthesized by HASMC m vitro. Gel mobility shaft assayshave mostly been used for evaluation of mteracttons between protems and DNA. The physicochemical similarities between DNA and PGs/GAGs make possible the use of techniques originally developed for DNA to be applied with PGs and GAGS This technique is based on the observatton that the movement of a PG or GAG molecule through a nondenaturmg agarose gel is hindered when apo B- 1OOcontaining lipoproteins (LDL), are bound to it. After the electrophoresls of solutions containing LDL and PGs or LDL and GAGS, tt is possible to detect bands corresponding to the migration of complexes, as well as those for the free PGs and GAGS. The amount of free and bound components can be measured after electrophoresls if the PGs and GAGS have been metabolically labeled with 35S.In case of GAGS, the amount of free polysaccharide can be evaluated by staining with toluidine blue. The gel mobility shift assay has

PG-Lipopro tein Interaction

269

advantages compared with other methods used to study LDL-PG/GAG mteraction. First, the PGs, GAGS, and LDL are initially free m solution and they do not need to be mnnobrltzed on any surface, therefore avordmg unnecessary modificattons that may alter the interaction between these molecules; second, rt requires small amounts of PGs, GAGS, or LDL; and, last but not least, tt IS an easy, rapid, and inexpensive method to perform. 2. Materials 2.7. hstruments 1. Horizontal submerged electrophoresis system, gel tray 15 cm wade, 10 cm long with buffer recuculatron/cooling system (Bio-Rad, Hercules, CA) 2 Electrophoresrs Power Supply EPS 500/400 (Pharmacra, Uppsala, Sweden). 3 A computer-controlled digrtal autoradiograph system (Berthold, Willbans, Germany) was used for the analysis of the agarose gel mobility shift with radroactive-labeled PGs and GAGS.

4 The data were analyzedwith aGraphPadPrrsmasoftware, version 2 0 (GraphPad Software, SanDiego, CA) 2.2. Reagents 1 Solvents and reagents of analytical grade were purchased from Sigma (St Lotus, MO) and Merck (Rahway, NJ). 2 Sample buffer: 10 nuI4 HEPES buffer, pH 7 4, contammg 140 mA4 NaCl, 5 nuI4 CaC&, and 2 mA4 MgCl* (see Note 1) 3 Runnmg buffer: 10 mA4 HEPES buffer, pH 7 2, contammg 2 mM CaCl, and 4 mM MgC12 (see Note 1). 4. Frxmg solution: 0 1% cetylpyrrdmmm chloride dissolved in 70% ethanol 5 PD-IO gel-exclusron columns to equilibrate LDL m the desired buffer were from Pharmacla. 6 Agarose NuSreve 3.1 and gel-bond films were from FMC BroProducts (Rockland, ME). 7 Autoradiography boxes and hyperfilm MP were from Amersham Life Scrence (Amersham, Sweden AB, Sweden)

2.3. Biological Materials 2.3.1. Plasma Lipoprotems Human LDL (D 1.019-l ,063 g/mL) was Isolated by KBr sequenttal ultracentrifugatron of fresh plasma from two healthy donors. LDL was stored at 4°C m the KBr buffer solution containing 1 mg/mL of Na-ethylenediamme tetra-acetic acid (EDTA) in sterile glass tubes and used wtthin 2 wk (18). Protein was determined by Bradford procedure (Bio-Rad Protein Assay, BioRad Chemical Dtvrsron, Richmond, CA).

Hurt-Camejo,

270

Camejo, and Sartipy

2.3.2. Proteoglycan isolation Radioactive biolabeled PGs and GAGS synthesizedby HASMC m vitro were isolated and characterized as described (19,20). 3. Methods 3.1. Incubation

of LDL with PG

The experimental conditions as well as the evaluation of the interaction with LDL described here are similar for both unlabeled or blolabeled PG or GAGS. The experiments described were performed with 3sS-and 3H-biolabeled CS-PG (verslcan) isolated from cell-culture media of HASMC. 1. Equilibrate the 35S-/3H-PG m dIstIlled water. 2. Lyophlhze and dissolve m sample buffer. 10 mM HEPES buffer, pH 7.4, containmg 140 mMNaCl,5 mMCaC12 and 2 mM MgC12 3. Measure the amount of cpm and the content of GAG (22) m order to calculate the amount of cpm/pg GAG m the PG preparation. 4. Dilute the PG or GAG preparation with sample buffer and divide it m samples of 20 pL containing approx 10,000 cpm, 0 2 pg GAG m 20 & (50,000 cpm/pg GAG) m 0 5-mL Eppendorf tubes 5 Eqmhbrate LDL in sample buffer using PD-10 columns (Pharmacla) and adjust the concentration to 1 mg protein/ml. 6. Add increasing concentrations of LDL* 0, 2, 4, 6, 10, 15, and 20 pg to each Eppendorf tube containing the PG or GAG samples and complete the volume to 40 pL with sample buffer. Do all samples m duplicate 7 Mix gently the samples by flipping the tubes and incubate for 1 h at room temperature. 8 Add 4 pL glycerol (10% final concentration) to increase the density of the samples prior to loadmg them m the agarose gel.

3.2. Gel Mobility Shift Assay 3.2.1. Agarose Gel Electrophoresis 1. Prepare a 0.7% agarose gel by adding 50 mL of running buffer (10 mM HEPES buffer, pH 7 2, containing 2 mA4 CaC12 and 4 mM MgC12) to 0.35 g of NuSleve 3.1 agarose m a 250-mL glass beaker (see Note 2). 2 Boll the agarose suspension for 2-3 min in a microwave oven at maximum effect to dissolve the agarose. 3. Adjust the final volume to 50 mL with distilled water 4. Cool down the agarose solution to 5@-6O”C. 5. Cast the gel over the hydrophilic side of a gel-bond film 6. Place a comb of the size of choice in the gel and leave the gel to solidify at room temperature for 1 h. In the experiments described here we used a 15 teeth-comb.

PG-Lipoprotein

271

Interaction

7 Remove the comb carefully and allow the gel to solidify completely overnight at 4°C m a humid chamber. This step improves significantly the resolution of the gel and allows the preparation of several gels m advance. 8 The next day, place the gel-bond film with the agarose gel in the hortzontal submarme electrophoresis equipment with buffer recirculation. 9 Fill the electrophoresis chamber with runnmg buffer. 10 F111the wells in the gel wrth running buffer with help of a prpet m order to remove possible air bubbles that can disturb sample loading. 11 Load 20 pL of each sample m the wells in duplicate. 12. Run the electrophoresis at 60 V, constant voltage, for 2 h Allow the samples to migrate into the gel without buffer recnculatron for the first 10 mm

3.2.2. Gel Flxa tion, Autoradiography,

and Staining

1 Remove the agarose gel from the electrophoresrs chamber and place it m a contamer for at least 2 h at room temperature wtth 200 mL of fixing solution* 0.1% cetylpyrldmium chloride dissolved in 70% ethanol To be able to compare results from gels run at different time points, it IS Important always to follow the same fixing conditions 2 Remove the gel from the fixing solution and air-dry it m a hood overnight. The gel can also be dried with help of a harr-dryer In thus case, a 50-mL agarose gel will take 45-60 mm to dry 3. Quantitatively evaluate the dried gel m a Berthold digital autoradiograph (Berthold, Willband, Germany). The electrophoretic mobrlity shift of the bands can also be vrsuahzed by placing the dried agarose gel m an autoradiography box with a hyperfilm at -70°C for 3 or more days, depending on the total amount of cpm present m the 35S/3H-brolabeled PG or GAG samples. If nonradlolabeled PGs or GAGS are used they can be visualized by staining the dried gel with 0 1% toluidine blue in acetic acid.ethanol.water (0.15:5) for 30 mm at room temperature and destained with the solvent mixture until the background IS clear The LDL-PG and LDL-GAG complexes and the free LDL can be vrsualized by staining the drred gel with 0.06% (w/v) Fat Red 7B, 0.015% (w/v) 011 Red 0 m 60% ethanol for 1 h at room temperature. Excess of stammg solution can be washed away with water

3.3. Quantitative Evaluation of LDL-PG Interaction by Agarose Gel Mobility Shift Assay The diagram in Fig. 1 illustrates a gel mobility shift assay for evaluation of LDL-PGs and LDL-GAGs interaction. Lane A shows a solution m the origin of the agarose gel contammg only PGs or GAGS. Lane B shows what happen in lane A after 2 h of electrophoresis: because free PGs or GAGS are very negatively charged molecules, they will move raptdly as a band at the top of the gel. Lane C shows a solutton containing LDL-PGs or LDL-GAGs complexes and some free PGs or GAGS and LDL molecules that was loaded mto the well of an

272

Hurt-Camejo,

A)

Oh

6)

2h

c)

Oh

D)

2h

Camejo, and Sartipy

e?

I

: 0 Free LDL

LDL-PGIGAG complexes

Free PGIGAG

Fig 1 Diagram lllustratmg gel mob&y shift assay. (A) A solution in the origin of the agarose gel contanung only PGs or GAGS. (B) Result m A after 2 h of electrophoresls (C) A solution containing LDL-PG or LDL-GAG and some free PGs or free GAGS loaded into the well (D) Result m C after 2 h of electrophoresls. agarose gel. Lane D represents the situation in lane C after 2 h of electrophoreSE: free PGs or GAGS molecules move rapidly through the gel to the top as a

band the LDL-PGs or LDL-GAGs complex also move mto and through the gel an appreciable distance but more slowly. This protocol was designed for the mteractlons between LDL and PG or GAG to take place at or near to physiological ionic-strength conditions (sample buffer) and the electrophoretic separation to take place at conditions that maxlmlze them (running buffer). 3.3.1. Analysis of LDL-PG Apparent Affinity Constant by Gel Mobility Shift Assay The electrophoretic

running buffer used in this gel mobility

shift assay has a

low ionic strength with pH 7.2. In this system, PGs and GAGS, which are very negatively charged molecules (isoelectric point I 1.O),migrate rapidly through the gel from the origin (cathode) to the upper top of the gel (anode). In contrast, LDL particles with an average isoelectric point of 5.8 migrate very little m this buffer system. The band of free PG or GAG that migrate through the gel reflects the amount of free PG or GAG present in the sample loaded m the agarose well. By measuring the radioactivity or optical density of this faster moving PGs or GAGS band, the fraction of free PGs or free GAGS is directly measured

PCS-Lipoprotein Interaction

273

0 0 2 2

4 4

6 6

10 10 15 15 20 20

pg LDL

Fig. 2. Autoradiography of a gel mobility shift assay using 35S/3H-biolabeled proteoglycans synthesized by proliferating human arterial smooth-muscle cells. The arrows at the left side of the picture indicate the regions where the measuring of cpm were performed corresponding to free PG and bound PG or LDL-PG complex bands.

and the bound fraction calculated by difference from the total free PGs or GAGS measured in samples containing no LDL. Another alternative is to measure directly the bound cpm in the gel. This is illustrated in Fig. 2, where a line is used to define the area of cpm corresponding to free and bound. These measurements allow then the calculation of the apparent affinity constant (K,) according to an equation reported previously for DNA-protein interaction analysis: K, = [FI(Lo(l

-F)]

(1)

where F is the fraction of radioactive PCs or GAGS bound and Lo is the total molar concentration of LDL, in large excessover that of PGs or GAGS (22). In this case a precise K, can be calculated if the value ofF is not very close to zero (high concentrations of LDL) or unity (low concentration of LDL). This condition is accomplished by choosing different concentrations of LDL and PGs or GAGS and following the change in the amount of free PGs or GAGS after gel mobility shift assays (see Note 3). Equation 1 described is based on the assumption that a small fraction of LDL binds to PGs. In these situations, you can assume that the concentration of free ligand is approximately equal to the concentration added. This assumption simplities the analysis of the binding experiments, and the standard analysis methods depend on this assumption. However, in some LDL-PGs binding experiments, LDL may have a high affinity for the PGs, and the PGs or GAGS are capable of binding several molecules of LDL per GAG chain and the above assumption cannot be applied in this type of situation. Complex formation depletes both freeLDL and free PC-binding sites. In these cases,equilibrium-binding equations that take into account these ligand depletion should be applied (23): Y = [f& + Lo + B,,,ax) + 1 (Kd + Lo + Bm&

- (4x Lo x Bmax)]/2

(2)

274

Hurt-Camelo,

Camejo, and Sartipy

The data from this algorithm can conveniently be analyzed by several commercially available nonlinear regression fits. In practice, experiments are carrted out as explained above by using a constant concentration of biolabeled PGs or GAGS and increasing concentrattons of LDL (Lo). The calculated Kd value will be in molar concentration of LDL. To calculate B,,, = total moles of LDL bound per unit of PG or GAG, a separate experiment needs to be performed where a constant concentration of LDL is mixed wrth increasing concentrations of blolabeled PGs or GAGS. The aim of with this experiment is to work with a spectrum of band shift mobility that covers from maximum mobility, total free PGs or GAGS, to maximum retardation or mobility shift of the biolabeled PC or GAG induced by a known molar concentration of LDL. This optimal condition is achieved by trying different concentrattons of LDL and PGs or GAGS. This B,,, value can then be used to transform the cpm bounds measured in the gel mto moles of LDL per unit (cpm or pg) of PGs or GAGS. It should be indtcated that lack of knowledge of the state of aggregation of the PGs and GAGS in the gel does not allow accurate stotchiometry determmatton m molecular terms. If the mformation about the stoichrometry is not needed, then the analysis of the binding may be simplified considerably by choosmg experimental conditions that guarantee that LDL IS always in great excesswith respect to PGs or GAGS. In this case, either Eq. 1 or Eq. 2 can be used to quantitative evaluate complex formation, both analytical models will give similar results. 3.4. Results Figure 2 shows the results of an association reaction between biolabeled PGs from proliferating HASMC and unlabeled LDL analyzed by gel mobility shift assay.The figure shows an autoradiography of a dried agarose gel. It can be observed that as the concentration of LDL increases (from 0. l-l .O p.iU or 2-20 pg/pL) the amount of free PG in the reaction dlmmtshes, whereas the level of LDL-PG complexes (bound) increases. LDL-PG content of cpm can be determined directly by measuring with a digital autoradtograph. This can also be done by cutting out the radioactive bands and assaying by liqutd-scmtillation counting (after appropriate corrections for possible quenching by the gel), or by running the autoradlography film in a densitometer and measurmg the areas of the bands corresponding to free PGs and LDL-PG bound. This last option can also be applied when detecting unlabeled PGs or GAGS with toluidine blue staining. In the absence of LDL (0 in Fig. 2), some cpm from PGs that do not migrate may be found in the origin of the agarose gel. These cpm probably come from large aggregates of PGs and should be subtracted from the total cpm measured corresponding to LDL-PG complexes (cpm bound). Aggregation often occur when using high molecular-weight PGs type versican

275

PG-Lipoprotein Interaction

Table 1 Analysis of the Data from Gel Mobility Shift Assay Showed in Fig. 2 to Evaluate the Interaction Between Biolabeled PGs Synthesized by Proliferating HASMC and LDL According to Eq. 1 Described in the Text LDL nM

0 100 200 300 500 750 1000

cpma free

cpm bound

F

1-F

Lo( I-F)

K, p/W’

Kd nMb

69 40 20 15 10 12 8

0 29 49 55 59 57 59

0.420 0.710 0.797 0 855 0826 0855

0.580 0290 0.203 0 145 0.174 0145

58 58 61 73 131 145

72 12 2 13 1 117 6 30 5 89

138 82 76 85 159 170

K mean f StD & 0%. 2)

11+3

1

95f28 62f 15

%pm values represent average of duplicate measurements m the same agarose gel hKd = UK,

Table 2 Analysis of the Data from Gel Mobility Shift Assay to Evaluate the Interaction Between Biolabeled PGs Synthesized by Resting HASMC and LDL According to Eq. 1 Described in the Text LDL nA4 0 100 200 300 500 750 1000 K mean f StD KI (Eq. 2)

cpma cpm free bound 70 62 52 34 20 16 10

0 8 18 36 50 54 49

F

1-F

Lo( 1-F)

K, /.uW’

K,, rut@

0.114 0 257 0.515 0 714 0.771 0.700

0 886 0 743 0.486 0.286 0.228 0 300

88.6 149 0 146.0 143 0 172.0 300 0

129 1.72 3 53 4 48 4 48 2.33

775 581 283 223 223 429

3.3 k 1.3

348 f 155 367 rk 200

%pmvaluesrepresentaverageof duphcatemeasurements m the sameagarosegel hKd= l/K,

(25.0 106). Lyophihzation and concentrations steps during the isolation of PGs may also Induce aggregation of PGs. Table 1 and Table 2 present the data from gel mobility shift electrophoresis assay to evaluate the interaction of LDL with biolabeled PGs from prohferat-

276

Hurt-Camejo, l

75

0

1

Camejo, and Sartipy

PGs Proliferating Cells PGs Resting Cells

25 0 260

460

660

860

1 1000

LDLnM

Fig. 3. Binding curves with data from Table 1 (PGs from proliferating cells) and Table 2 (PGs from resting cells). bound cpm (LDL-PG complex cpm) vs LDL concentration.

ing and resting HASMC, respectively. These tables show the raw data corresponding to the bands of free PGs (cpm free) and LDL-PG complex (cpm bound) as indicated m Fig. 2 measured with a computer-controlled dtgital autoradiograph. The tables also show the quantitative analysis of the data according to Eq. 1 described prevtously. This equation allow the evaluation of binding in terms of the apparent affmty constant (K,) as LDL molartty in the presence of excess of LDL at or near to physiological ionic conditions (sample buffer). LDL showed a three- to fourfold higher affinity with PGs from prohferatmg HASMC (Table 1) than with PGs from restmg HASMC (Table 2). These results can also be observed m Fig. 3 that is a graph of cpm bound vs LDL concentration for both types of PGs. It can be observed that the binding curve reached saturation at lower concentrations of LDL with PGs from prohferatmg HASMC than with PGs from restmg cells. This experiment represent a good example of how the gel mobility shift assay can be used to evaluate quantitatively the mteraction between LDL and different PGs or GAGS. The association of phospholipase A2 (sPLA2) with PGs and GAGS have also been studied using this technique (20). Tables 1 and 2 also compare the Kd (dissociatton constants) values obtained by analyzing the data according to Eq. 2. This equation as explained above takes into account ligand depletion owing to complex formation. The Kd values obtained with Eq. 1 and Eq. 2 were similar. This indicates that no depletion of LDL occurred during complex formation m the experimental conditions used, and that the molar concentrations of LDL were always m excess (0.1-I pJ4, apo B 500,000 mol wt) with respect to biolabeled PGs (OS-1 nM,

PG-Lipoprotein

Interaction

approx 5.0 x 1O6mol wt). The similarity in results obtained with two different analysis techniques also indicates the K, and & values were determined accurately using gel mobility shift assay.

4. Notes 1. The sample buffer and running buffer described are critical both for the quality of the results and their interpretation. The sample buffer was selected to provide a near-to-physiological ionic strength, pH and Ca2+ and Mg2+concentratlons where the interaction of LDL or other lipoprotems with PGs and GAGS take place. The running buffer 1sselected to provide a pH and iomc strength where the complex 1s separated from the free molecules. This buffer should retain the associations formed during mcubatlons. This is achieved by lowering the lomc strength but retammg the pH and Ca*+ and Mg2+ concentrations. HEPES has a pKa closer to the pH used and thus provides a good buffer capacity. The low ionic strength also reduces the current requirement and therefore gives a low heat generating system. The runnmg buffer used in this system is of low ionic strength. This buffer contains cofactors (Mg2+ and Ca*‘) needed for the LDL-PG or LDL-GAG interaction The electrostatic component of LDL-PC or LDL-GAG mteractlons generally leads to stronger binding at low salt concentrations. Thus the stability of the complexes will tend to be enhanced as they move from the sample buffer (wtth 140 mMNaC1) placed m the wells into the gel buffer (runnmg buffer, wlthout NaCI). This means that complexes persist as a band throughout the electrophoresls run, and dlssoclatlon of the complexes barely occurs In case that dissoclatlon of complexes does occur during the electrophoresls, the PGs or GAGS released ~111 not reach the band of free PGs or GAGS migrating first. Thus, the levels of free PGs or GAGS and the complexes seen in the gel reflect those in the solution that was loaded onto the gel This is explained with the following observations. In running buffer without NaCl (low ionic strength), LDL IS known to bmd tightly to PGs or GAGS, forming highly cooperative nonspec~fic complexes in which many LDL particles are associated with each PG or GAG molecule. Such complexes migrate very httle in a 0 7% agarose gel. However, if LDL and PGs or GAGS are Incubated m sample buffer contammg 140 mA4 NaCl, the formatlon of these large nonspeclfic complexes 1s mmlmal. Placing LDL-PG or LDL-GAG m sample buffer mto the wells and running the electrophoresls in runnmg buffer leads to only specific complexes m the gel. These observations indicate that whatever changes m buffer concentration occur during the start of the electrophoresis, they do not affect the complexes It appears that this gel mobility shift system yields information about the complex of interest, with few artifacts from the agarose gel or the gel buffer. 2. A variety of gels can be used m these experiments Horizontal submarine gels are most convenient: these gels permit rapld loading of samples Agarose gels (0 71 2%) can be used with PG, GAG, and large macromolecules such as LDL The quality of the agarose matrix is important. The gel matrix should not affect the properties of the complexes as measured m electrophoresls. If, for example,

278

Hurt-Camejo,

Camejo, and Sartipy

the gel prevents the separation of free PG, then the apparent affinity constant would be higher than the true value. The critical qualities for agarose are low levels of electroendosmosis (10.13%), and good clarity at the working concentrations New, high-quality agaroses are commercially available for most molecular biology applications We recommend testing and comparing new gel matrtces m then particular experimental conditions 3 In order to calculate affimty constant (K,), tt IS necessary to measure the formation of LDL-PG or LDL-GAG complexes. This can be done usmg other methods such as affinity chromatography, or by coating plastic surfaces with one of the reactants (enzyme-lmked m-nnunosorbent assay [ELISA] type of method). These methods are easy to use; however, one must be concerned that the immobthzation of LDL or PG to filters, agarose beads, or plastic surfaces may mduce modifications in the macromolecules that disturb the LDL-PG or LDL-GAG interaction under study An advantage of using the gel mobrlity shift assay 1s that free unmodified LDL and PGs or GAGS macromolecules can form complexes in solution.

References 1 Hurt-Camejo, E , Olsson, U., Wrklund, 0 , Bondjers, G , and CameJo, G (1997) Cellular consequences of the assocratron of apoB lipoprotems with proteoglycans Arterioscler. Thromb Vast Blof 17, 101 l-1017. 2 Yla-Herttuala, S., Solaktvi, T., Hirvonen, J., et al. (1987) Glycosammoglycans and apolrpoproteins B and A-I m human aortas. chemical and immunological analysis of lesion-free aortas from chtldren and adults. Arterloscleroszs 7,333-340 3. Nievelstem-Post, P., Mottino, G., Fogelman, A, and Frank, J (1994) An ultrastructural study of lipoprotein accumulation m cardiac valves of the rabbit Arterzoscler. Thromb 14, I 15 l-1 16 1. 4 Cardoso, L. and Mourao, P. (1994) Glycosammoglycan fractions from human arteries presenting diverse susceptrbihty to atherosclerosis have different bmdmg affinities to plasma LDL. Arterzoscler Thromb Vast B~ol 14, 115-l 24. 5. CameJo, G., Acquatella, H., and LaLaguna, F. (1980) The interaction of low density lipoprotein with arterial proteoglycans: an additional risk factor? Atherosclerosls 36,55-65. 6. CameJo, G , Lopez, A, Lopez, F., and Quiiiones, J. (1985) Interaction of low density lipoproteins wtth arterial proteoglycans the role of charge and siahc acid content. Atheroscleroszs 55,93-105. 7. Anber, V., Miller, J S., McConnell, M , Shepherd, J., and Packard, C. J (1997) Interaction of very-low-density, intermediate-density, and low-density hpoproteins with human arterial wall proteoglycans. Arterzoscler Thromb. Vast Biol 17,2507-25 14 8. Wagner, W. D , Salisbury, B. G., and Rowe, H A. (1988) A proposed structure of chondronin 6-sulfate proteoglycans of human normal adJacent atherosclerotic plaque. Arterlosclerosrs 6,407-4 17

PG-Lipoprotein

Interaction

279

9 Jackson, R , Busch, S., and Cardm, A. (1991) Glycosammoglycans.molecular properties, protein Interactions and role m physiological processes Physiolog. Rev 71,481-539 10. Volker, W , Schmidt, A., and Buddecke, E (1987) Mapping of proteoglycans in human arterial tissue. Eur J Ceil. Biol 45, 72-79. 11. Wight, T N (1996) The vascular extracellular matrix, m Atherosclerosrs and Coronary Artery Dlseuse (Fuster, V., Ross, R , and Topol, E. J , eds.), Philadelphia: Lippmcott-Raven, pp 42 l-440 12. CameJo, G , Olofsson, S -O., Lopez, F , and BondJers, G. (1989) Identification of apoB-100 segments mediating the mteraction of low density lipoproteins with arterial proteoglycans Arterlosclerosls 8, 368-377. 13. Hurt-CameJo, E. and CameJo, G. (1997) Potential involvement of type II phospholipase A2 m atherosclerosis. Atheroscleroszs 132, l-8. 14. Olsson, U., CameJo, G., and BondJers, G. (1993) Binding of a synthetic apohpoprotein B-100 peptide and peptide analogues to chondroitm-&sulfate effect of the lipid environment. BEochemzstry 32, 1858-1865. 15. Linden, T., BondJers, G., CarrreJo, G., Bergstrand, R., Wilhelmsen, L , and Wiklund, 0 (1989) Affimty of LDL to a human arterial proteoglycan among male survivors of myocardial infarction. Eur J Clan Invest 19,38-44 16. Anber, V., Griffin, B., McConnell, M., Packard, C , and Sheperd, J (1996) Influence of plasma lipid and LDL-subfraction profile on the interaction between low density lipoprotem with human arterial wall proteoglycans. Atherosclerosis 124, 261-271. 17 Wtklund, O., Bondjers, G., Wright, I., and CameJo, G. (1996) Insoluble complex formation between LDL and arterial proteoglycans m relation to serum lipid levels and effect of lipid lowermg drugs. Atheroscleroszs 119, 57-68 18 Havel, R , Eder, H , and Bragdon, J H. (1995) The distribution and chemical compositton of ultracentrifugally separated lipoprotems m human serum J Clzn Invest 34, 1345-l 353 19 Camejo, G , Fager, G , Rosengren, B. E. H -C., and Bondjers, G (1993) Binding of low density ltpoprotems by proteoglycans synthesized by proliferating and qutescent human arterial smooth muscle cells. J. Biol Chem 268, 14,13 l-14,137. 20 Sartipy, P., Johansen, B., CanneJo, G., Rosengren, B., BondJers, G , and Hurt-Camejo, E (1996) Binding of human phosphohpase A2 type II to proteoglycans. differential effect of glycosammoglycans on enzyme activity. J, Bzol Chem 271,26,307-26,3 14 21. Goldberg, R. and Kolibas, L. (1990) An improved method for determining proteoglycans synthesized by chondrocytes m culture Connect Tissue Res 24, 265-275 22 Revzm, A. (1989) Gel electrophoresis

assays for DNA-protein mteractrons BloTechnlques 7,346-355. 23. Epps, D. E., Raub, T J., and Kezdy, F. J. (1995) A general, wide-range spectrofluorometric method for measuring the site-specific affinities of drugs toward human serum albumin Anal Blochem. 227,342-350.