D2. MATERIALS AND METHODS. Mouse strains. B10.D2 (M504) (H-2 a) breeding pairs, kindly sent by Dr. Jan Klein, were maintained in the breeding colony.
0022-1767/78/1203-0726502.00/0 THE JOURNAL OF IMMUNOLOGY Copyright © 1978 by The Williams & Wilkins Co.
Vol. 120,No. 3, March 1978 Printed in U.S.A.
S T R U C T U R A L D I F F E R E N C E S B E T W E E N THE M O U S E H-2D P R O D U C T S OF THE M U T A N T B10.D2.M504 (H-2 da) A N D THE P A R E N T A L N O N M U T A N T S T R A I N B10.D2 (H-2d) 1 J. L Y N N E BROWN, 2 R O D E R I C K N A I R N , 3 AND S T A N L E Y G. N A T H E N S O N 4 From the Departments of Microbiology and Immunology, and Cell Biology, Albert Einstein College of Medicine, Bronx, New York 10461
The H-2K and H-2D glycoproteins from the mutant and H-2D da suggest that a complex mutation has taken m o u s e strain B10.D2.M504 (H-2Kd,D da) were compared place. with the H-2K and H-2D molecules from the parent nonmutant strain B10.D2 (H-2Kd,Dd). Whereas in every reThe mouse major histocompatibility complex (MHC) 5 on spect the K molecules of the mutant and n o n m u t a n t cells chromosome 17 is of interest to immunologists because m a n y were identical, the D damolecule from the mutant s h o w e d of its gene products are involved in the processes of immune notable differences from the D d molecule of the nonmurecognition and reaction. At the present level of analysis five tant parent. ~First, smaller a m o u n t s (ca. 25%) of the D da major regions of this chromosomal segment have been defmed: glycoprotein appeared to be present in M504 cell extracts K , / . S, G, and D (1-3). T h e S region controls the level of the as compared to the amount of D d glycoprotein present C4 component of C (4, 5) and the G region, the presence of an in nonmutant cell extracts. Second, a proportion of the erythrocyte antigen H-2.7 (6, 7). The K and D regions contain D da glycoprotein molecule~ s h o w e d an abnormal interthe K and D genes, respectively, whose gene products are cell action with the lectin from lentil beans (LcH), s u g g e s t i n g surface glycoproteins of approximately 45,000 m.w. (8). These an altered carbohydrate chain on some of the molecules. Third, comparative tryptic peptide analysis s h o w e d a products are associated noncovalently in a 1:1 ratio with the significant degree of diversity b e t w e e n the mutant D da 11,600 m.w. protein fl2-microglobulin (9-12) and are present in the membranes of many cell types but most densely on those molecule and the n o n m u t a n t D d molecule. Thus, analysis of the lymphoid system. The K and D gene products are of arginine-labeled peptides s h o w e d at least t w o peptides unique to H-2D d and at least three peptides unique to involved as the targets of ceU-mediated graft rejection in the H-2D d~. Analysis of lysine-labeled peptides s h o w e d at allogeneic rejection response, and in the in vitro counterpart, least three peptides unique to H-2H d and greater than the cell-mediated lympholysis (CML) test (1-3). These products t w o peptides unique to the mutant H-2D da. These data are also thought to play a role in the recognition process inherent in the killingof virus-infected target cellsby immune thus document that the D da glycoprotein has an altered primary amino acid sequence as compared to the non- T cells (13).The I region contains immune response (Ir) genes which control responses to certain alloantigens (14, 15) as well mutant D d molecule. The finding of differences in greater as genes whose products are serologically detected as the la than one-third of the total peptides recovered correlates antigens. with the strong serologic and histogenic reactivity beA n extreme polyrnorphism is one of the most notable propt w e e n the cells of the M504 mutant and the cells of the B10.D2 parent. This peptide diversity is in contrast to erties of the H-2 K and D glycoproteins. However, very little the small degree of v a r i s t i o n (ca. 10%) noted for the is known about the extent or nature of this genetic variability, of the mechanisms for generating and maintaining the polymutants of the H-2K b locus (e.g., H-2K b~, H-2K bd) which morphism, or of its role, if any, in the function of the H-2 have only histogenic reactivity and essentially no seroproducts. One approach to these questions is through the study logic differences b e t w e e n mutant and nonmutant cells. The multiple structural differences noted b e t w e e n H-2D d of the structure of the H-2 products of mouse mutants having alterations in biologic responses controlled by the MHC, as such information will bear on the relationships of structure to biologic properties. A number of such mutants in the K or D Received for publication July 25, 1977. genes have been isolated (16-18). For example, two mutants in Accepted for publication December 5, 1977. the H - 2 K b gene, B6.C-H-2 ba (Hzl) (18) and B6-H-2 bd (M505) The costs of publication of this article were defrayed in part by the (17), show histogenic reactions such as skin graft rejection and payment of page charges. This article must therefore be hereby marked graft-vs-host reaction (GVHR), as well as positive reactions in advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. the in vitro mixed lymphocyte reaction (MLR) and CML This work was supported by National Institutes of Health Grants assays. Structural studies have shown that these two H-2K b AI-07289 and AI-10702, and Grant IM-77 from the American Cancer mutants have small but discrete primary sequence differences Society. in their K gene products relative to the parental strains, thus 2 Recipient of fellowships from the National Institutes of Health establishing structural evidence for a mutation in the H - 2 K and the Damon Runyon Memorial Fund. Present address: Department gene (19). The observations suggest that, at the most simple of Genetics, Washington University School of Medicine, 660 S. Euclid Avenue, St. Louis, Missouri 63110. 5 Abbreviations used in this paper: MHC, major histocompatibility 3 Recipient of Fellowship DRG-120-F from the Damon Runyon- complex; CML, cell-mediated lympholysis; GVHR, graft-vs-host reacWalter Winchell Cancer Fund. tion; SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electro4 To whom reprint requests should be addressed. phoresis; LcH, Lens culinaris hemagglutinin or lectin. 726
1978]
STRUCTURE OF H-2Dd` MUTANT GLYCOPROTEIN
level of explanation, a K gene mutation can provide both a CML and MLR reactivity. Other mouse strains have been isolated with D gene mutations. One of these mutants is designated B10.D2 (M504) (H-2da). Unlike the previously discussed K mutants which were not chemically induced, this strain came from a parent treated with the mutagen diethyl sulfate (an alklyating agent) (20). The mutation was localized to the H-2D region by the use of skin grafts and Fl hybrid crosses (21). Serologic,analysis indicated that H-2da had lost two specificities, 40 and 49, associated with the H-2d haplotype and gained a new specificity, 50, not present in the H-2a haplotype (22). In addition, there was evidence (22) that the H-2.4 specificity expressed by the allele H-2da in the recombinant R106 differed quantitatively from H2.4 of the allele H-2Da by quantitative absorption analysis. In addition to these serologically detectable differences, there was also a mutual stimulation in MLR, cytotoxicity in CML, skin graft rejection, and GVHR reactivity between M504 and the parental B10.D2 (23-25). Thus, M504 seems to be an unusual mutation in which the new D gene product is characterized by serologically detectable alterations, as well as histogenic reactivity when compared to its parental haplotype
H.2 d. Because of the unusual biologic reactions reported above, we have examined some of the biochemical and primary structural features of the H-2D gene product from M504 in comparison to that of its parent, B10.D2. Our studies reveal a relatively extensive structural difference between the H-2D glycoprotein of M504 as compared to that of the parent molecule from B10.D2. MATERIALSAND METHODS
Mouse strains. B10.D2 (M504) (H-2 ~a) breeding pairs, kindly sent by Dr. Jan Klein, were maintained in the breeding colony of Dr. Frank Lilly of the Albert Einstein College of Medicine. B10.D2, new (H-2 d) mice were purchased from the Jackson Laboratory, Bar Harbor, Maine. Antisera. The following antisera were prepared as described previously (26). 1) To detect H-2 specificity 4 and 4':6 (C3H × C57BL/6)F~ anti-A/Jax or (C3H × C57BL/6)F~ anti-B10.A {(H-2* × H-2b)F~ anti-H-2a}; 2) To detect H-2 specificity 31: A/Jax anti-Meth A: (H-2~ anti-H-2d). Radiolabeling, solubilization, lentil lectin affinity chromatography, and immunoprecipitation. Radiolabeling of spleen cells with radioactive amino acids and the subsequent solubilization of the cell surface membrane with 0.5% NP-40 have been described in detail {27). Partial purification of the NP-40 lysate by affinity chromatography with the lectin from Lens culinaris (LcH) was performed as described previously (19). The radioactive material that passed through the LcH column without binding was called the "protein pool" whereas that which bound and was then eluted with buffer containing amethyl mannoside was termed the "glycoprotein pool." Immunoprecipitation of the specific H-2K or H-2D gene product from both the "glycoprotein pool" and "protein pool" was carried out as described previously (19).
Reduction and alkylation, Bio-Gel A chromatography, trypsin digestion, and ion exchange chromatography. These procedures were carried out essentially as described previously (27) except that the elution buffer for the Bio-Gel A column contained 0.002 M dithiothreitol (A grade, Calbiochem, San H-2.4 is the private specificity of the H-2Dd product of H-2d; H2.4' is used to denote the cross-reactive private specificity of the H-2D product of H-2 da.
727
Diego, Calif.) and the trypsi~ digestion buffer was 0.1 M ammonium bicarbonate, pH 8.55.
Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE). SDS-PAGE on gels of 10 to 15% acrylamide (Eastman Chemical Co., Rochester, N. Y.) and 0.1% sodium dodecyl sulfate (Sequenol grade, Pierce Chemical Co., Rockford, Ill.) was carried out with the discontinuous buffer system described by Maizel (28). Assay: inhibition of immune precipitation. Radiolabeled "titering antigen" (H~-leucine-labeled NP-40 lysate of B10.D2 spleen cells) was partially purified by LcH affinity chromatography. A constant amount of "titering antigen" (an aliquot to give 20,000 cpm) was tested with serial dilutions of the antisera to be used. The size of the precipitate was normalized with appropriate amounts of BALB/c normal mouse serum (Jackson Laboratory) and the equivalent amount of goat anti-mouse antibody was added to precipitate the antigen-aUoantibody complex. The sensitivity of the assay was adjusted by choosing an amount of aUoantibody to react with 70 to 80% of the maximum specific H-2 radioactivity in the aliquot of "titering antigen." Serial dilutions of nonradioactive inhibitor (either the original spleen preparation to be assayed or aliquots of the protein or glycoprotein pools produced after passage through the LcH affinity column) in a 500-pl ~,olume were incubated with the appropriate amount of antiserum and BALB/c normal mouse serum for 2 hr at 4°C. Then the predetermined amount of "titering antigen" was added and the incubation was continued for 2 hr at 4°C. The specified equivalent amount of goat antimouse antibody was then added to cause precipitation and the incubation was continued an additional 3 hr at 4°C, after which the precipitate was recovered by centrifugation, washed, solubilized, and counted in a Beckman 230 liquid scintillation counter. Appropriate full precipitation controls {containing antisera and normal mouse sera but no inhibitor) and nonspecific zero precipitation controls (containing only normal mouse sera but no inhibitor and no aUoantisera) were also included in each assay. Percentage of inhibition was determined using the formula: cpm in full precipitate - cpm in inhibitor cpm in full precipitate - cpm in zero precipitate Corrections were made for volume differences in the glycoprotein and protein pools. Pronase digestion. An immune precipitate of 3H-fucose-labeled M504 or B10.D2 H-2D or H-2K glycoprotein was prepared from the NP-40 spleen cell lysate with the appropriate monospecific alloantisera and goat anti-mouse antibody and washed with 0.01 M Tris, 0.15 M NaC1, pH 7.4. Pronase, B grade, 45,000 PUK/g (Calbiochem) was made 50 mg/ml in digestion buffer (0.25 M Tris, pH 7.4, + 10 #M CaCl2), predigested for 2 hr at 37°C according to Sefton and Keegstra (29). Fifty-microliters of this were added to the sample in 200 ~l of digestion buffer which was incubated for 24 hr at 37°C. Then a h ~ h e r 50 ~1 of Pronase were added, and the incubation was continued for an additional 24 hr. Finally the reaction was stopped by boiling the reaction mixture for 5 rain and centrifuging at 2000 rpm for 10 min in the 253 rotor in the International PR-2 centrifuge, after which the supernatant was removed and analyzed on a P-6 column. Bio-Gel P-6 gel filtration. Glycopeptide samples were chromatographed on Bio-Gel P-6 (200 to 400 mesh, Bio-Rad Laboratories, Richmond, Calif.). The column (0.9 x 112 cm) was equilibrated and eluted with 0.1 M Tris, pH 8.0, containing
J. L. BROWN, R. NAIRN, AND S. G. NATHENSON
728
0.001 M pentachlorophenol (Eastman Kodak, Rochester, N. Y.) at 25°C. Fractions of 0.8 ml were collected and 80-pl aliquots were counted in Toluene Omnifluor scintillation fluid (New England Nuclear, Boston, M~ss.). DEAE-Sephadex chromatography. A D E A E - S e p h a d e x A25 column {Anion exchanger, cap. 3.5 -+ 0.5 mEq/g, particle size 40 to 120 # (Pharmacia Fine Chemicals, Inc., Piscataway, N. J.) was poured (0.5 x 10 cm) and equilibrated with 0.01 M Tris, p H 8.4, containing 0.00] M pentachtorophenol. T h e pool from the Bio-Gel P-6 column was diluted 1:10 with distilled water and the p H was adjusted to 8.4 with 1 M NaOH. This neutralized pool was applied to the column and the column was then washed with 0.01 M Tris, pH 8.4. T h e radiolabeled glycopeptides were eluted with a linear salt gradient (110 ml of 0.0 M NaCl and 110 ml of 0.3 M NaCl, both in 0.01 M I r i s , p H 8.4). This was followed by a 50-ml wash of 1.0 M NaC1. Fractions of 2.0 ml were collected. Appropriate dilutions of every 5th sample were made and the conductivity was determined. T h e samples were analyzed for radioactivity by scintillation counting in Aquasol (New England Nuclear). RESULTS
Quantitative aspects of H-2.4' and H-2.4 on the cell surface and in NP-40 extracts. T h e D da molecule from M504 reacted with the anti-D d antiserum against H-2.4 {see Materials and Methods, Antisera). Hence detection of the altered molecule carrying H-2.4' in the M504 m u t a n t was accomplished with the same H-2.4 antiserum as t h a t used for the parent cell B10.D2. When the amount of antigenic reactivity on the cell surface of M504 and B10.D2 spleen cells was assayed, it was found t h a t the reactivity for H-2.4' was approximately 25% of t h a t found for H-2.4. On the other hand, the reactivity of the K d molecule (H-2.31), as judged by the inhibition of the immune cytolysis assay, was essentially identical on the spleen cells of both the M504 m u t a n t and the B10.D2 parent. Since these quantitative differences might be due to an altered availability of the H-2D da glycoprotein on the cell surface, we also examined the total cellular H-2 content in cell extracts prepared with the nordonic detergent NP-40. For such measurements we used a method devised by J. H. Freed for the quantitative inhibition of precipitation of radiolabeled H2.4 antigen. Lysates from M504 and B10.D2 when tested for H-2.4 activity by this method {Table I) showed t h a t the 50% inhibition point for B10.D2 was one-fourth t h a t of M504. Thus, as was also found for intact cells, M504 spleen cell lysates
TABLE I Quantitative inhibition of immune precipitates by M504 and B IO.D2 extracts a Extract
M504 B10.D2
No. of cell equivalents to produce 50% inhibition of b
H-2A (H-2D)
H-2.31 (H-2K)
5.5 x 107 1.5 X 107
4% x 106 5.5 x IOS
The operationally defined soluble extract is prepared from radioactively labeled spleen cells solubilized in 0.5% NP-40 (1 hr, 2°C) at a concentration of 5 × l0 s cells/ml and then subjected to centrifugation at 2°C in the SW-40.3 rotor at 30,000 rpm for 90 rain in the Beckman L2-65 ultracentrifuge. b Data given as equivalent numbers of spleen cells in extract to produce 50% inhibition of indirect immune precipitation of radiolabeled, partially purified H-2 antigen prepared by the procedure outlined in Materials and Methods section.
[VOL.
120
contain roughly one-fourth the H-2.4 activity of B10.D2 spleen cell lysates. In contrast, when these same cell lysates were tested for H-2.31 activity, the 50% inhibition points for M504 and B10.D2 extracts were essentially the same, although M504 appeared to require 10 to 20% fewer equivalents than B10.D2. T h u s the results with the inhibition of precipitation m e t h o d are essentially identical to those found by absorption experiments on whole cells. A t t e m p t s to compare total precipitable radioactivity of H2D d~ and H-2D d molecules labeled in their protein moiety were unsatisfactory due to problems of contaminating materials present in unpurified extracts. Hence we were unable to assess directly the question of whether the decreased reactivity for H-2.4 antibody in the M504 cell extracts reflected fewer molecules or reduced affinity by the antibodies for the H-2.4' site on the H-2D da glycoprotein. Properties of H-2.4' and H-2.4-bearing molecules during LcH affinity chromatography Affmity chromatography on the Sepharose-linked LcH separates a glycoprotein-enriched fraction of the crude Npo40 cell lysate from a protein-enriched fraction. Previous studies have shown t h a t H-2K and H-2D glycoproteins interact with the lectin presumably through their carbohydrate moieties and can be specifically eluted with a buffer containing a-methyl mannoside. An examination of the properties of H-2.4' and H-2.4 during LcH affinity chromatography showed remarkable differences. As measured by the inhibition of precipitation method (see Materials and Methods), the H-2.4 activity of a B10.D2 lysate (Fig. 1A, U p p e r Panel) was found mainly in the glycoprotein fraction with some
j 50
A) BIO.D2 H-24 100
5O Z O
G
B} B]OD2 H-23}
rf100
I Z A) M504 H-24'
w
B} M504 H-231
loo
100
.50
10
" J ~
i
, L i , ,i,, 100
50
i ~J.~. 500 M.I INHIBITOR
I00
500
Figure 1. Inhibition of precipitation assays of the protein and glycoprotein pools produced after lentil lectin affinity chromatography of nonradioactive NP-40 spleen lysates. NP-40 lysates were made of 5 × l0s spleen cells of either M504 or BI0.D2, and these were initially assayed by inhibition of immune precipitation for H-2.4 and H-2.31 activity. Each spleen cell lysate was fractionated by LcH chromatography and the material in the initial effluent (the protein enriched pool) and that eluted with buffer containing 0.1 M a-methyl mannoside (the glycoprotein pool) were then assayed for H-2.4 and H-2.31 activity by inhibition of immune precipitation as described in Materials and Methods. The results have been corrected for volume differences. The panels marked A show the H-2.4 activity and the panels marked B show the H-2.31 activity. O O indicates the activity in the protein pool {P). H indicates the activity in the glycoprotein pool (G).
1978]
729
STRUCTURE OF H-2Dda MUTANT GLYCOPROTEIN
residual activity (10 to 20%) in the protein pool, as expected. However, the H-2.4' activity from M504 {Fig. 1A, Lower Panel} was partitioned between the protein and glycoprotein pools in almost the reverse manner. That is, approximately 60% of the activity was found in the protein pool and 40% in the glycoprotein pool. On the other hand, as shown in Figure 1B, the H2Ka (H-2.31) activity behaved identically in both M504 and B10.D2 spleen cell lysates. Thus in both cases, the majority of H-2.31 activity was found in the glycoprotein pool with approximately 20% in the protein pool. The reduced binding of the M504 H-2.4' antigen as compared to the B10.D2 H-2.4 antigen is not a matter of overloading the LcH column since rechromatography of the material showed that no additional activity bound to the LcH column. Carbohydrate moiety. It was considered possible that the difference in binding properties of the H-2.4 and H-2.4' molecules to the LcH affinity column was caused by an altered or absent sugar side chain. Accordingly we examined some of the properties of the carbohydrate moieties of these glycoproteins. NP-40 extracts of spleen cells from B10.D2 and M504, radiolabeled in 3H-fucose, were used for immune precipitates with H2.4 and H-2.31 antisera. SDS-PAGE analysis showed that approximately equal amounts of fucose radioactivity were precipitated with the anti-4 serum in the B10.D2 and M504 NP-40 extracts. Separation into the glycoprotein and protein fractions by LcH chromatography also showed parallel results for parent and mutant extracts in terms of partitioning of 3H-fucose radioactivity into the protein and glycoprotein pools. When immunoprecipitation analyses of H-2.4 and H-2.31 in the B10.D2 NP-40 extracts were performed, specific radioactivity was precipitated from the glycoprotein pool only, and no specific radioactivity was precipitated from the protein pool. The same results were obtained for M504. Thus, in the normal B10.D2 parent cell most of the activity as measured by the inhibition of binding assay was in the glycoprotein pool and all of the fucose-labeled radioactivity was in the glycoprotein pool. However, in the case of M504 whereas nearly 60% of the inhibitory activity was found in the protein pool, only the glycoprotein pool carried fucose-labeled material precipitable by antiserum to H-2.4. From this experiment we can tentatively conclude that that portion of the H-2.4' activity partitioning in the glycoprotein pool contains fucose whereas that portion of the H-2.4' activity detected in the protein pool does not contain fucose. Accordingly the H-2.4' molecules found in the protein pool may have a part or all of the carbohydrate moiety lacking. We examined the fucose-labeled H-2.4' molecule further in order to determine if the fucose-containing carbohydrate moiety found on this molecule w~.s altered from that found on the B10.D2 glycoprotein. A small aliquot of a specific precipitate made with anti-4 antiserum from 3H-fucose-labeled B10.D2 and M504 NP-40 extracts was examined by SDS-PAGE; 95% of the radiolabel migrated as a single sharp peak at the m.w. of H-2D. The remainder of each precipitate was digested exhaustively with Pronase for 48 hr as described in Materials and Methods. Such a procedure is known to remove nearly all of the protein potypeptide backbone except for a few amino acids attached to the sugar side chain itself. The glycopeptides produced by Pronase digestion were then subjected to gel filtration chromatography on a Bio-Gel P-6 column. The glycopeptides from both the mutant and parent eluted as a single radioactive peak with a m.w. of approximately 3100 to 3200. In view of the identical size of the mutant and parent glycopeptides, the charge heterogeneity was then examined by applying
the pool of the material from the P-6 column to a DEAESephadex A-25 column and eluting with a linear gradient of sodium chloride. The elution patterns from the aH-fucose-labeled M504 H-2.4' and from the 3H-fucose-labeled B10.D2 H2.4 on DEAE chromatography are shown in Figure 2. The patterns are essentially identical. Therefore it appears that at least by the criteria of size and charge the carbohydrate moieties of the LcH-Sepharose bound H-2.4 and H-2.4' molecules from parent and mutant, respectively, are identical. Comparative tryptic peptide profiles. For unequivocal results in the peptide chromatographic comparison technique, purification of specific immune precipitates was necessary. Immune precipitates of H-2.4 and H-2.4' were prepared from the glycoprotein pools generated by the LcH affinity chromatography procedure. These precipitates were combined in an appropriate ratio and after solubilization in SDS were reduced and alkylated, as described previously (27), and applied to a Bio-Gel A0.5 M column. The peak of radioactivity eluting from the BinGel A-0.5 M column of the correct m.w. for the H-2D glycoprotein was pooled, precipitated with trichloroacetic acid, and after washing digested with trypsin, and the soluble peptides were applied to an ion exchange column as described previously (27). Figure 3 shows the tryptic peptide profiles for the argininelabeled H-2.4' and H-2.4 molecules from M504 and B10.D2, and Figure 4 shows the lysine-labeled tryptic peptide patterns for the H-2.4' and H-2.4 glycoproteins. The arginine-labeled
FUCOSE B10.D2 H-24
//~100
j
°B°
/
_Too
%)
100 sI.O0
3H FUCOSE M504 H-2 4'
o c~
500
.
//
~
~-
/
t o,o
~00~ o
20
40
60 BO 100 FRACTION NUMBER
120
14o
Figure2. DEAE-Sephadex A-25, anion exchange chromatography of aH-fucose-labeled H-2.4 glycopeptidesfrom B10.D2 and from M504. Specific immune precipitates of H-2.4 from BI0.D2 and H-2A' from M504 aH-fucose-labeled NP-40 spleen cell lysates were digested with Pronase as described in Materials and Methods. The pronase digested H-2 glycopeptides were purified from small amounts of contaminants by gel f'dtration on Bio-Gel P-6 and the material in the 3300 m.w. radioactive peak from each column applied to a 0.9- x 10-cm DEAESephadex A-25 column and washed in with the column buffer, 0.01 M Tris, pH 8.4. The radiolabeled glycopeptideswere eluted with a linear salt gradient from 0 M NaCl to 0.3 M NaCl followed by a 1.0 M NaCl wash. The solid line denotes aH-fucose radioactivity and the lighter continuous line denotes the salt gradient measured by conductivity.
peptide maps reveal a considerable amount of similarity with, however, certain striking and significant differences. T h e r e appear to be peptides eluting a) at p H 4.25 (fraction 120), and b) at p H 4.51 (fraction 146) which are present in the B10.D2 H-2.4 and absent in the M504 H-2.4'. In turn, three new lowyield peptides are present at p H 3.53 (fraction 47), p H 4.62 (fraction 158), and p H 4.84 (fraction 193) in the map of H-2.4' from M504. Thus, one can conclude that there are at least two soluble arginine-labeled peptides t h a t are unique to H-2D d, and at least three peptides unique to H-2D da, I000,
80C
....
3H ARG M504 H-2.4'
--
14C ARG BIO.D2 H-2,4
5.0
J
y-
4.0 L) < O
pH
40(
rr
5.0 200
4~0
A
8'o
200
FRACTION NUMBER
Figure 3. Comparative arginine-labeled tryptic peptide profile of H2.4' from M504 (ZH-arginine) and H-2.4 (14C-arginine) from B10.D2. The specific immune precipitate of H-2.4 from the B 10.D2 glycoprotein pool and of H-2.4' from the M504 glycoprotein pool were combined, solubilized in SDS, reduced with 0.11 M dithiothreitol, alkylated with 0.25 M iodoacetamide and applied to a Bio-Get A-0.5 M column equilibrated with 0.05 M Tris-HCl, pH 7.5, containing 0.5% SDS and 0.002 M dithiothreitol. The appropriate H-2 pools were precipitated with 15% TCA, washed, and digested with TPCK-trypsin. The acid soluble peptides were separated on a 25-cm column of Spherix XX860-0 resin by using a linear pyridine acetate gradient from 0.05 M pyridine, pH 3.13, to 2.0 M pyridine, pH 5.0, as described in detail in Reference 27. The 200 3-ml fractions were evaporated to dryness and radioactivity determined by liquid scintillation counting in Aquasol.
Lysine-labeled maps. Figure 4 shows the 3H-tysine-labeled soluble peptides of H-2.4' precipitated from the glycoprotein pool of M504 and 14C-lysine-labeled soluble peptides from H2.4 precipitates from the glycoprotein pool of B10.D2. There appear to be peptides eluting at p H 3.84 (fraction 80), p H 4.16 (fraction 108), and p H 4.50 (fraction 141) which are present in the B10.D2 H-2.4 and absent in the M504 H-2.4'. There are clearly peptides eluting at p H 3.91 (fraction 87) and p H 4.19 (fraction 111) that are unique to M504 H-2.4'. In addition, there are clearly M504 H-2.4' unique peptides present in the complex regions of the gradient bounded by fractions 118 to 126 and fractions 160 to 170. Control tryptic peptide comparisons. T h e H-2K d gene products (H-2.31) were also isolated from NP-40 spleen cell preparations from M504 and B10.D2. T h e tryptic peptide profdes of appropriately labeled glycoprotein mixtures were examined as described above. Both the arginine-containing {Fig. 5) and lysine-containing {data not shown) soluble peptide prof'des are identical. DISCUSSION
Notable polymorphism is one of the most striking properties of the H-2K and H-2D glycoproteins. This is exemplified serologically by the numerous widely divergent antigenic specificities associated with H-2D and H-2K alleles of different haplotypes, and by comparative peptide mapping (30) and N-terminal sequence analyses (31) of different H-2 glycoproteins from several haplotypes. Peptide comparisons show that products of alleles of the same gene, for example K b and K d, share only 30 to 35% of peptides or have differences of about 65 to 70% (Table II). Further, products of alleles of either the K gene or ...... ] H ARG M504 H-2,31
5°°f
. . . . . 5H Ly$ M504 H-2.4'
14C Lys BIO.D2 H-2.4
--
5.0
H
>
4.0 pH
c.)
