- Email: [email protected]
Requirement for cysteine in the color silver staining of proteins in polyacrylamide gels
ANAL\‘TICAI
156, 136-139
RIOCHEMISTRY
(1986)
Requirement for Cysteine in the Color Silver Staining of Proteins in Polyacrylamide Gels PAULJ.CHUBAANDSUNILPALCHAUDHURI Ci ?q’ne
State
L’nivcwitv
Sdml
Received
yf Medicine.
December
Detroit.
Michigan
48201
23. 1985
To determine whether cysteine residues have a contribution to the mechanism of color silver staining, we silver stained sodium dodecylsulfate polyacrylamide gel electrophoresis separations of proteins which have few or no cysteines. Proteins without cysteine stained negatively (yellow against a yellow background) with silver. Proteins with one or more cysteines stained orange. red. brown, or green/gray depending on the mole percentage of cysteine and whether they contained covalently attached lipids. The colors could not be correlated with the mole percentages of cysteine of these proteins indicating that some components other than cysteine affect the staining color of cysteine-containing proteins. Silver staining of amino acids. sugars. nucleotide bases, or lipopolysaccharide dot-blotted onto nitrocellulose paper implicated adenine, lipids, the basic amino acids, and glutamine. but not sugars or other amino acids in silver/protein complexes. ‘51 1986 Academic KEY
Press, Inc. WORDS:
gel staining;
gel electrophoresis;
proteins;
Silver staining of polypeptides in acrylamide gels is being increasingly used for visualization of banding patterns, particularly when radiolabeling is undesirable. A number of silver staining protocols have been described (l-4) and several are commercially available. In many cases colored images have been observed. Since each protein stains a characteristic color, there exists a great potential for protein identification by color. Although images resulting from the differential reduction of silver halide crystals are quite familiar from photographic processes, the mechanism of color silver staining of proteins, lipids (5), and nucleic acids (6) remains unclear. Based on an experiment in which autoradiography of gels using radioactive silver tracers was performed, Merril and Goldman (7) eliminated the possibility that differential exclusion of silver ions in the region of the gel occupied by protein causes staining and suggested that charges on protein molecules and/ or direct chemical reactions may have some effect. The use of tunicamycin in tissue culture to selectively block the addition of N-linked 0003-2697/86
$3.00
Copyright 0 1986 by Academic Press. Inc. All rights of reproduction in any form reserved.
amino
acids; silver:
cysteine.
oligosaccharides to mammalian cell glycoproteins or the use of neuraminidase to cleave sialic acid residues from glycoproteins did not alter the staining color or intensity of proteins separated on 2-D gels (8) even though mobility in either the isoelectric focusing or molecular weight dimension was altered; therefore it seemed unlikely that carbohydrate or negatively charged groups alone caused silver staining. Recently, Nielsen and Brown (9) suggested that the color of protein/silver complexes is related to the binding of silver to the side chains of charged amino acids. The data presented here show that although the basic amino acids as well as cysteine bound silver in their free form, only proteins containing cysteine stain positively with color silver stain. MATERIALS
AND
METHODS
Outer membrane preparations from Eschcoli K- 12 strain AB2463 grown in modified M9 medium were electrophoresed on 11.5%polyacrylamide gelsasdescribed( 10) rrichiu
136
CYSTEINE
REQUIRED
and silver stained by the method of Sammons et al. (3). Fixation was performed by incubating the gels in a solution of 50% ethanol, 7.5% acetic acid overnight. The gels were washed in three changes of distilled water for 1 h each and then incubated in 1.9% silver nitrate for I h. The image was developed in 0.75 M NaOH, 7.5% (v/v) formaldehyde for 5 min at room temperature. Color enhancement was achieved by incubating the gel in two changes of 7.5% (w/v) sodium carbonate for 1 h each. Gels were photographed immediately or stored in 7.5% acetic acid. A solution of 1% Kodak Rapid Fix “Solution A” can be used for destaining the gels: however, the colors may be altered. Protein bands were identified by their mobilities relative to the protein standards: lactalbumin ( 14.2 kDa), trypsin inhibitor (20.1 kDa), trypsinogen (24 kDa), carbonic anhydrase (29 kDa). glyceraldehyde-3-phosphate dehydrogenase (36 kDa), egg albumin (45 kDa). or bovine albumin (66 kDa) (Sigma). For silver staining of free amino acids. nucleotide bases, sugars (Sigma), or lipopolysaccharide (a gift from M. Aqua), 5 ~1 of each of the solutions listed in the legend to Fig. 2 were “dot-blotted” onto nitrocellulose paper (Schleicher anld Schuell) using a capillary pipette. The nitrocellulose was then incubated in 0.95% silver nitrate (Sigma) for 5 min followed by reduction in 0.75 M NaOH containing 0.75% formaldehyde for 30 sec. The reaction was stopped by washing in distilled water. RESULTS
AND
DISCUSSION
In initial observations, we tried to eliminate the two artifactual horizontal lines often encountered in silver staining by including 0.25 M iodoacetamide in the sample buffer as suggested by Hashimoto et a/. (1 I ). No proteins could be detected by silver staining these preparations. This wasa strong indication that alkylation of the sulthydryl groups of cysteinecontaining proteins abrogated the reduction of silver which would have occurred in those
FOR
SILVER
137
STAIN
e
66.2
4-45 c-36 t-29
0mpC Y&PCompA --w -
t
24
-IPP--+
a
b
FIG. 1. Photograph of silver stained SDS-PAGE separation of K. coli outer membranes (lane a), or Sigma molecular weight markers (lane b). The migration positions of the orq~A. optnF. ~I~wzC, or lipoprotein lpp components are indicated.
areas of the gels containing protein. On the basis of this observation we predicted that proteins which lack cysteine residues would not stain with silver. A short search of the literature presentedseveral examples of proteins with low cysteine content which were reported to stain anomalously with silver stains. These included histone H 1 (9), calmodulin. and troponin C (13) of which histone Hl and calmodulin lack cysteine and troponin C contains a single cysteine. We silver stained SDS-PAGE’ separated outer membrane proteins (OMPs) of E. co/i known to contain low mole percentages of cysteine (Fig. I, lane a). As expected, the OIM~F and OIMVCgene products which are devoid of cysteine ( 13) stained negatively with the color silver stain with the same yellow color as the background (Fig. 1). The E. co/i ompA gene product (or its conformational isomer) which hastwo cysteines per 329 residues(14) stained ’ Abbreviations used: SDS-PAGE. sodium dodecylsulfate polyacrylamide gel electrophoresis: OMP. outer membrane proteins.
138
CHUBA
AND
a slightly darker yellow color than the background or orange. Bacterial lipoprotein (lpp) is an outer membrane component which has fatty acid covalently linked to it through glyceryl-cysteine ( 15) but is otherwise devoid of cysteine. This protein stains green, a color which is characteristic of lipopolysaccharides as analyzed by the silver staining method of Tsai and Frasch (5), indicating that the lipid portion of lipoproteins does contribute to silver staining. A yellow or negatively staining band was also observed for the 3500 Da form of the protein hormone glucagon which lacks cysteine. The 4500 Da form of glucagon stained green, suggesting some post-translational modification. The amino acid compositions of the molecular weight marker proteins are mostly known (16). Carbonic anhydrase which has the lowest mole percentage cysteine (0.4%) stains with the lowest intensity (Fig. 1, lane b). Other marker proteins which contain higher amounts of cysteine stain more intensely. We were unable to correlate the color of the protein bands with the cysteine content except in the case of negatively staining pro-
1
14
2
3
15
4
16
5
6
16
17
PALCHAUDHURI
teins described above. For example, trypsinogen has a high cysteine content (5.2%) and stains brown, whereas lactalbumin which also has a high (6.5%) mole percentage of cysteine, stains red. The amounts of cysteine in other standard proteins are albumin (6.0%), trypsin inhibitor (9.2%), and glyceraldehyde-3-phosphate dehydrogenase (1.4%). By silver staining a number of amino acids. nucleotide bases, sugars, or lipopolysaccharides which had been dot-blotted to nitrocellulose paper, we observed that spots corresponding to the basic amino acids or glutamine turned brown, whereas cysteine produced an intense blue color which turned yellow and faded within a few minutes. The brown histidine and arginine spots also faded within 30 min, but the lysine spot was stable. Lipopolysaccharide (brown) or adenine (white with a brown halo) also stained under these conditions. Other components (listed in the legend to Fig. 2) did not differ appreciably in staining from distilled water which gave a faint grey halo. Therefore it seems likely that the presence of basic amino acids. glutamine, or lipids may influence the silver staining color or intensity of proteins in acrylamide gels. None-
7
19
20
6
9
21
22
10
11
12
13
‘. *‘j
23
24
25
26
27
28
29
30
31
FIG, 2. Dot-blotting preparations of the following solutions, silver stained as described under Materials and Methods: (1) 1% met, (2) 1% pro, (3) 1% thr. (4) 1% ser, (5) 1% his, (6) 1% val. (7) 1% arg, (8) 1% asp. (9) I%lys.(lO) l%gly,(ll) I%cys.(l2) I%ala.(l3) l%tyr,(l4) l%trp.(l5) I%glu,(l6) l%leu.(17) 1% ile, (18) 1% asn, (19) 1% gin. (20-22) distilled water. (23) 25% xylose, (24) 25% ribose, (25) 25% dextrose, (26) 25% rhamnose, (27) 25% sucrose. (28) 0.25% adenine. (29) 0.25% thymine, (30) 0.25% uracil. (31) lipopolysaccharide (from Salmonella).
CYSTEINE
REQUIRED
theless, the presence of the basic amino acids alone is not sufftcient for staining if these proteins do not contain cysteine, as the ompF and ompC proteins (Fig. 1) have appreciable amounts of lysine, arginine, and histidine (lys + arg + his = 29%) (13), yet do not stain. The results presented above provide an explanation for some of the difficulties which investigators have reported in silver staining proteins. In particular, it is clear that quantitation of proteins in silver stained gels must be approached with extreme caution. It is interesting in retrospect that the keratin proteins which as it turns out are responsible for the artifactual horizontal lines in silver staining ( 17) have extraordinarily high mole nercentagesof cysteine (greater than 17%) (-16). We predict that other cysteine-free proteins will have anomalous silver staining characteristics. ACKNOWLEDGMENTS Our thanks go to Dr. David Sammons for his encouragement. This research was supported by NIH Grant RR 08167-08.
REFERENCES 1. Merril. C. R., Switzer. R. C., and Van Keuren. M. L. ( 1979) Proc. NatI. .lcad. Sci. US.A 76. 23 l-137. 2. Oakley. B. R.. Kirsch. D. R., and Morris, N. R. ( 1980) Anal. Biochun. 105, 36 l-363.
FOR
SILVER
STAIN
139
3. Sammons, D. W., Adams, L. D.. and Nishizawa, E. E. (1981) Electrophoresis 2, 135-141. 4. Friedman, R. D. (1982) .&ra/. Biochem. 126. 346349. 5. Tsai. C. M.. and Frasch. C. E. (1982) Anal. Biochem. 119, 115-119. 6. Sommerville, L. L.. and Wang, L. (1981) Biochem. Biophy.7. Re.7. Common. 102, 53-58. 7. Merril. C. R.. and Goldman, D. (1984) in Two Dimensional Electrophoresis of Proteins: Methods and Applications (Celis, J. E., and Bravo. R.. eds.), pp. 93-109, Academic Press, Orlando. Fla. 8. Sammons, D. W., Adams. L. D., Vidmar, T. J.. Hatheld, C. A., Jones. D. H.. Chuba, P. J.. and Crooks, S. W. ( 1984) in Two Dimensional Gel Electrophoresis of Proteins: Methods and Applications. (Celis. J. E., and Bravo. R., eds.), pp. I I I-126, Academic Press. Orlando. Fla. 9. Nielsen. B. L., and Brown. L. R. ( 1984).dnal Biochem. 141,31 l-315. IO. Schnaitman, C. A. (I 97 I) J. Bacterial. 108, 545-552. Il. Hashimoto. F.. Horigome, T.. Kanbayashi, M.. Yoshida. K., and Sugano. H. (1983) Anal. Biochem. 129. 192-199. 12. Schleicher, M., and Watterson. M. ( 1983). Anul. Biwhem. 131, 3 12-317. 13. Mizuno. T., Chou, Mei-Yin, and Inouye. M. (1983) J. Blol. Chem. 258, 6932-6940. 14. Braun. G.. and Cole, S. T. ( 1984) .&fool. Gen. Gcner. 195, 321-328. IS. Braun. V. (1975) Biochim Biuphj:~. .3c/u 415, 335377. 16. Dayhotf M. 0. (1972) Atlas of Protein and Sequence and Structure, National Biomedical Research Foundation, Silver Spring, Maryland. 17. Ochs. I>. (1983) Anal. Biochem. 135, 470-474.
156, 136-139
RIOCHEMISTRY
(1986)
Requirement for Cysteine in the Color Silver Staining of Proteins in Polyacrylamide Gels PAULJ.CHUBAANDSUNILPALCHAUDHURI Ci ?q’ne
State
L’nivcwitv
Sdml
Received
yf Medicine.
December
Detroit.
Michigan
48201
23. 1985
To determine whether cysteine residues have a contribution to the mechanism of color silver staining, we silver stained sodium dodecylsulfate polyacrylamide gel electrophoresis separations of proteins which have few or no cysteines. Proteins without cysteine stained negatively (yellow against a yellow background) with silver. Proteins with one or more cysteines stained orange. red. brown, or green/gray depending on the mole percentage of cysteine and whether they contained covalently attached lipids. The colors could not be correlated with the mole percentages of cysteine of these proteins indicating that some components other than cysteine affect the staining color of cysteine-containing proteins. Silver staining of amino acids. sugars. nucleotide bases, or lipopolysaccharide dot-blotted onto nitrocellulose paper implicated adenine, lipids, the basic amino acids, and glutamine. but not sugars or other amino acids in silver/protein complexes. ‘51 1986 Academic KEY
Press, Inc. WORDS:
gel staining;
gel electrophoresis;
proteins;
Silver staining of polypeptides in acrylamide gels is being increasingly used for visualization of banding patterns, particularly when radiolabeling is undesirable. A number of silver staining protocols have been described (l-4) and several are commercially available. In many cases colored images have been observed. Since each protein stains a characteristic color, there exists a great potential for protein identification by color. Although images resulting from the differential reduction of silver halide crystals are quite familiar from photographic processes, the mechanism of color silver staining of proteins, lipids (5), and nucleic acids (6) remains unclear. Based on an experiment in which autoradiography of gels using radioactive silver tracers was performed, Merril and Goldman (7) eliminated the possibility that differential exclusion of silver ions in the region of the gel occupied by protein causes staining and suggested that charges on protein molecules and/ or direct chemical reactions may have some effect. The use of tunicamycin in tissue culture to selectively block the addition of N-linked 0003-2697/86
$3.00
Copyright 0 1986 by Academic Press. Inc. All rights of reproduction in any form reserved.
amino
acids; silver:
cysteine.
oligosaccharides to mammalian cell glycoproteins or the use of neuraminidase to cleave sialic acid residues from glycoproteins did not alter the staining color or intensity of proteins separated on 2-D gels (8) even though mobility in either the isoelectric focusing or molecular weight dimension was altered; therefore it seemed unlikely that carbohydrate or negatively charged groups alone caused silver staining. Recently, Nielsen and Brown (9) suggested that the color of protein/silver complexes is related to the binding of silver to the side chains of charged amino acids. The data presented here show that although the basic amino acids as well as cysteine bound silver in their free form, only proteins containing cysteine stain positively with color silver stain. MATERIALS
AND
METHODS
Outer membrane preparations from Eschcoli K- 12 strain AB2463 grown in modified M9 medium were electrophoresed on 11.5%polyacrylamide gelsasdescribed( 10) rrichiu
136
CYSTEINE
REQUIRED
and silver stained by the method of Sammons et al. (3). Fixation was performed by incubating the gels in a solution of 50% ethanol, 7.5% acetic acid overnight. The gels were washed in three changes of distilled water for 1 h each and then incubated in 1.9% silver nitrate for I h. The image was developed in 0.75 M NaOH, 7.5% (v/v) formaldehyde for 5 min at room temperature. Color enhancement was achieved by incubating the gel in two changes of 7.5% (w/v) sodium carbonate for 1 h each. Gels were photographed immediately or stored in 7.5% acetic acid. A solution of 1% Kodak Rapid Fix “Solution A” can be used for destaining the gels: however, the colors may be altered. Protein bands were identified by their mobilities relative to the protein standards: lactalbumin ( 14.2 kDa), trypsin inhibitor (20.1 kDa), trypsinogen (24 kDa), carbonic anhydrase (29 kDa). glyceraldehyde-3-phosphate dehydrogenase (36 kDa), egg albumin (45 kDa). or bovine albumin (66 kDa) (Sigma). For silver staining of free amino acids. nucleotide bases, sugars (Sigma), or lipopolysaccharide (a gift from M. Aqua), 5 ~1 of each of the solutions listed in the legend to Fig. 2 were “dot-blotted” onto nitrocellulose paper (Schleicher anld Schuell) using a capillary pipette. The nitrocellulose was then incubated in 0.95% silver nitrate (Sigma) for 5 min followed by reduction in 0.75 M NaOH containing 0.75% formaldehyde for 30 sec. The reaction was stopped by washing in distilled water. RESULTS
AND
DISCUSSION
In initial observations, we tried to eliminate the two artifactual horizontal lines often encountered in silver staining by including 0.25 M iodoacetamide in the sample buffer as suggested by Hashimoto et a/. (1 I ). No proteins could be detected by silver staining these preparations. This wasa strong indication that alkylation of the sulthydryl groups of cysteinecontaining proteins abrogated the reduction of silver which would have occurred in those
FOR
SILVER
137
STAIN
e
66.2
4-45 c-36 t-29
0mpC Y&PCompA --w -
t
24
-IPP--+
a
b
FIG. 1. Photograph of silver stained SDS-PAGE separation of K. coli outer membranes (lane a), or Sigma molecular weight markers (lane b). The migration positions of the orq~A. optnF. ~I~wzC, or lipoprotein lpp components are indicated.
areas of the gels containing protein. On the basis of this observation we predicted that proteins which lack cysteine residues would not stain with silver. A short search of the literature presentedseveral examples of proteins with low cysteine content which were reported to stain anomalously with silver stains. These included histone H 1 (9), calmodulin. and troponin C (13) of which histone Hl and calmodulin lack cysteine and troponin C contains a single cysteine. We silver stained SDS-PAGE’ separated outer membrane proteins (OMPs) of E. co/i known to contain low mole percentages of cysteine (Fig. I, lane a). As expected, the OIM~F and OIMVCgene products which are devoid of cysteine ( 13) stained negatively with the color silver stain with the same yellow color as the background (Fig. 1). The E. co/i ompA gene product (or its conformational isomer) which hastwo cysteines per 329 residues(14) stained ’ Abbreviations used: SDS-PAGE. sodium dodecylsulfate polyacrylamide gel electrophoresis: OMP. outer membrane proteins.
138
CHUBA
AND
a slightly darker yellow color than the background or orange. Bacterial lipoprotein (lpp) is an outer membrane component which has fatty acid covalently linked to it through glyceryl-cysteine ( 15) but is otherwise devoid of cysteine. This protein stains green, a color which is characteristic of lipopolysaccharides as analyzed by the silver staining method of Tsai and Frasch (5), indicating that the lipid portion of lipoproteins does contribute to silver staining. A yellow or negatively staining band was also observed for the 3500 Da form of the protein hormone glucagon which lacks cysteine. The 4500 Da form of glucagon stained green, suggesting some post-translational modification. The amino acid compositions of the molecular weight marker proteins are mostly known (16). Carbonic anhydrase which has the lowest mole percentage cysteine (0.4%) stains with the lowest intensity (Fig. 1, lane b). Other marker proteins which contain higher amounts of cysteine stain more intensely. We were unable to correlate the color of the protein bands with the cysteine content except in the case of negatively staining pro-
1
14
2
3
15
4
16
5
6
16
17
PALCHAUDHURI
teins described above. For example, trypsinogen has a high cysteine content (5.2%) and stains brown, whereas lactalbumin which also has a high (6.5%) mole percentage of cysteine, stains red. The amounts of cysteine in other standard proteins are albumin (6.0%), trypsin inhibitor (9.2%), and glyceraldehyde-3-phosphate dehydrogenase (1.4%). By silver staining a number of amino acids. nucleotide bases, sugars, or lipopolysaccharides which had been dot-blotted to nitrocellulose paper, we observed that spots corresponding to the basic amino acids or glutamine turned brown, whereas cysteine produced an intense blue color which turned yellow and faded within a few minutes. The brown histidine and arginine spots also faded within 30 min, but the lysine spot was stable. Lipopolysaccharide (brown) or adenine (white with a brown halo) also stained under these conditions. Other components (listed in the legend to Fig. 2) did not differ appreciably in staining from distilled water which gave a faint grey halo. Therefore it seems likely that the presence of basic amino acids. glutamine, or lipids may influence the silver staining color or intensity of proteins in acrylamide gels. None-
7
19
20
6
9
21
22
10
11
12
13
‘. *‘j
23
24
25
26
27
28
29
30
31
FIG, 2. Dot-blotting preparations of the following solutions, silver stained as described under Materials and Methods: (1) 1% met, (2) 1% pro, (3) 1% thr. (4) 1% ser, (5) 1% his, (6) 1% val. (7) 1% arg, (8) 1% asp. (9) I%lys.(lO) l%gly,(ll) I%cys.(l2) I%ala.(l3) l%tyr,(l4) l%trp.(l5) I%glu,(l6) l%leu.(17) 1% ile, (18) 1% asn, (19) 1% gin. (20-22) distilled water. (23) 25% xylose, (24) 25% ribose, (25) 25% dextrose, (26) 25% rhamnose, (27) 25% sucrose. (28) 0.25% adenine. (29) 0.25% thymine, (30) 0.25% uracil. (31) lipopolysaccharide (from Salmonella).
CYSTEINE
REQUIRED
theless, the presence of the basic amino acids alone is not sufftcient for staining if these proteins do not contain cysteine, as the ompF and ompC proteins (Fig. 1) have appreciable amounts of lysine, arginine, and histidine (lys + arg + his = 29%) (13), yet do not stain. The results presented above provide an explanation for some of the difficulties which investigators have reported in silver staining proteins. In particular, it is clear that quantitation of proteins in silver stained gels must be approached with extreme caution. It is interesting in retrospect that the keratin proteins which as it turns out are responsible for the artifactual horizontal lines in silver staining ( 17) have extraordinarily high mole nercentagesof cysteine (greater than 17%) (-16). We predict that other cysteine-free proteins will have anomalous silver staining characteristics. ACKNOWLEDGMENTS Our thanks go to Dr. David Sammons for his encouragement. This research was supported by NIH Grant RR 08167-08.
REFERENCES 1. Merril. C. R., Switzer. R. C., and Van Keuren. M. L. ( 1979) Proc. NatI. .lcad. Sci. US.A 76. 23 l-137. 2. Oakley. B. R.. Kirsch. D. R., and Morris, N. R. ( 1980) Anal. Biochun. 105, 36 l-363.
FOR
SILVER
STAIN
139
3. Sammons, D. W., Adams, L. D.. and Nishizawa, E. E. (1981) Electrophoresis 2, 135-141. 4. Friedman, R. D. (1982) .&ra/. Biochem. 126. 346349. 5. Tsai. C. M.. and Frasch. C. E. (1982) Anal. Biochem. 119, 115-119. 6. Sommerville, L. L.. and Wang, L. (1981) Biochem. Biophy.7. Re.7. Common. 102, 53-58. 7. Merril. C. R.. and Goldman, D. (1984) in Two Dimensional Electrophoresis of Proteins: Methods and Applications (Celis, J. E., and Bravo. R.. eds.), pp. 93-109, Academic Press, Orlando. Fla. 8. Sammons, D. W., Adams. L. D., Vidmar, T. J.. Hatheld, C. A., Jones. D. H.. Chuba, P. J.. and Crooks, S. W. ( 1984) in Two Dimensional Gel Electrophoresis of Proteins: Methods and Applications. (Celis. J. E., and Bravo. R., eds.), pp. I I I-126, Academic Press. Orlando. Fla. 9. Nielsen. B. L., and Brown. L. R. ( 1984).dnal Biochem. 141,31 l-315. IO. Schnaitman, C. A. (I 97 I) J. Bacterial. 108, 545-552. Il. Hashimoto. F.. Horigome, T.. Kanbayashi, M.. Yoshida. K., and Sugano. H. (1983) Anal. Biochem. 129. 192-199. 12. Schleicher, M., and Watterson. M. ( 1983). Anul. Biwhem. 131, 3 12-317. 13. Mizuno. T., Chou, Mei-Yin, and Inouye. M. (1983) J. Blol. Chem. 258, 6932-6940. 14. Braun. G.. and Cole, S. T. ( 1984) .&fool. Gen. Gcner. 195, 321-328. IS. Braun. V. (1975) Biochim Biuphj:~. .3c/u 415, 335377. 16. Dayhotf M. 0. (1972) Atlas of Protein and Sequence and Structure, National Biomedical Research Foundation, Silver Spring, Maryland. 17. Ochs. I>. (1983) Anal. Biochem. 135, 470-474.