Vol. 186, No. 3, 1992 August 14, 1992

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 1449-]454

M A P P I N G T H E O R I E N T A T I O N OF AN A N T I G E N I C P E P T I D E BOUND IN T H E A N T I G E N B I N D I N G G R O O V E OF H - 2 K b U S I N G A M O N O C L O N A L ANTIBODY

Sebastian Joyce , Rui Sun #, and Stanley G. Nathenson *,# * # Departments of Cell Biology and Microbiology & Immunology, Albert Einstein College of Medicine, Bronx, NY 10461

Received June 26, 1992

SUMMARY The major histocompatibility complex class I molecules are receptors for intracellular peptides, both of self and non-self origin. When non-self peptides (eg., pathogen derived) are bound to the class I molecules, they form ligands for T cell receptors resulting in antigen specific lysis of the infected cells by cytotoxic T lymphocytes. Therefore, an understanding of the process of antigen recognition requires the precise definition of the structural features of the bimolecular complex formed by a single well defined antigenic peptide bound to the class I molecule. A strategy using antibodies was developed to probe the structural features of the H-2K b containing a defined peptide in the antigen cleft. We report that the binding surface area of a K b specific monoclonal antibody (28-13-3s) includes residues in the a l (Gly56 and Glu58) and c~2 (Trp167) helices of Kb thus, binding across the antigen binding groove. When cells treated with the antigenic peptide of vesicular stomatitis virus, N52-59, and its alanine substituted analogs were tested for 28-13-3s binding, it was found that position 1 of the peptide also forms a part of the antibody binding site. This finding strongly supports the positioning of the N-terminus of N52-59 proximal to pocket A, thus, assuming an orientation parallel to the a l helix. ~ 1992 Academic Press,

Inc.

Class I molecules of the major histocompatibility complex (MHC) play a crucial role in immune surveillance which entails the continuous presentation of self and non-self antigens on the cell surface. When cells present non-self antigens, such as those derived from intracellular pathogens or mutant self proteines, antigen specific cytotoxic T lymphocytes (CTL) are activated resulting in the specific lysis of the cells that are recognized as non-self (1-3). The ligand recognized by the CTL can be mimicked in vitro by peptides added to cells that express the appropriate class I molecule (4-6). At the molecular level, peptide presentation occurs when the peptide is bound in the antigen binding groove of the class I molecules (4). The isolation of a naturally processed viral peptide from vesicular stomatitis virus nucleoprotein, N52-59 (ArgGlyTyrValTyrGlnGlyLeu), recognized by H-2Kb restricted CTL (5)

1449

0006-291x/92 $4.00 Copyright © 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

Vol. 186, No. 3, 1992

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

has allowed a detailed analysis o f the features o f the peptide in class I binding (7). Optimal binding of N 5 2 - 5 9 to K b requires the presence o f Tyr at positions 3 and 5, and Leu at position 8 as K b anchor residues and a length o f eight amino acids, properties which correlate with the physicochemical features o f the g r o o v e (7). The residues at positions 1Arg, 2Gly, 4Val, 6Gln, and 7 G l y are putative T cell receptor contact residues (7). This arrangement of T cell receptor and K b contact residues in an interspersed m a n n e r suggests that the N52-59 peptide is bound to K b in an extended c o n f o r m a t i o n (7).

Here, w e report the m a p p i n g o f a K b specific m o n o c l o n a l antibody (mAb)

epitope and the use o f this m A b to probe the structural features of the K b molecule occupied by a single defined peptide, N52-59.

MATERIALS

AND METHODS

R8 is an Abelson leukemia virus transformed pre-B cell line from which the R8.8, R8.9, and R8.10 lines were derieved by ethyl methyl sulfonate (EMS) mutagenesis followed by the selection of 2g-13-3s (8) epitope loss mutants. RMA is a Rauscher murine leukemia virus transformed T cell line from which H-2b class I loss mutant, RMA-S, is derieved by EMS mutagenesis (9). Cells were grown in RPMI-1640 (Gibco, NY) supplemented with 10% fetal calf serum (Bioproducts, Inc.), L-glutamine, pennicilin, and streptomycin (Gibco) at 37°C to a density of 5xl06.ml -I. Total cellular RNA was extracted from ~5xl~ cells (R8.8, R8.9, and R8.10) and poly(A)÷ RNA was isolated by poly(dT)-cellulose column chromatography as described (10, 11). cDNAs from each mutant cell line were synthesized from the poly(A) ÷ RNA using K b specific oligoneucleotides in the presence of AMV reverse transcriptase under stringent conditions (Tin-2; 1t). The full length cDNAs were further amplified by polymerase chain reaction (PCR) using K b specific forward and reverse primers (12). The uncloned PCR products were sequenced using end labelled Kb specific primers by the chain termination method (13) using the Sequenase kit (U S Biochemicals) according to the procedures previously established in the laboratory (11). The deduced amino acid changes were then mapped on the model Kb structure (14) using the computer graphics program QUANTA on a personal Iris workstation. The N52-59 peptide was synthesized at the Laboratory of Macromolecular Analysis and the Ala substituted analogs were synthesized by the RAMPS method using Fmoc chemistry as described (7). All the peptides were >99% pure by mass spectrometric analysis (7). The induction of cell surface expression of H-2Kb molecules was performed by incubating 100BM of each peptide with 5x105 RMA-S cells for 16 hours at 37°C as described (15). Antibody binding studies were done by indirect immunofluorescence as described earlier (16). Five million cells were incubated with the first antibody at 1:1(30 dilution on ice for 30 min. First antibody included anti-Kb mAb Y3 [protein A (Pharmacia) purified; cx2 domain (10,17)], K9-136 [ascites; al,et2 domain(ll)[, and 28-13-3s (purified by 33% ammonium sulphate precipitation) as well as the anti-Db mAb, B22-249 [protein A purified; cxl domain (17), as the negative control. (The concentration of the purified mAb was 1 mg.m1-1.) The first antibody binding was resolved by using fluorescin isothiocyanite conjugated anti-mouse immunoglobulin (heavy and light chain specific; Pierce) at 1:100 dilution in a reaction on ice for 30 min. Fluorescence was monitored by FACScan analysis (Becton Dickenson) on a log scale amplification. Standard error of the mean of fluorescence intensity were calculated using the Statistics programme for the Macintosh (StatWorksI~, version 1.2) where the sample size for Y3, K9-136, 28-13-3s, and B22249 binding studies were 6, 6, 13, and 4, respectively.

RESULTS

AND

DISCUSSION

R8.8, R8.9, and R 8 . 1 0 are in vitro mutants o f R8 selected for the loss o f binding o f 28-13-3s m A b as described earlier (16). Cytoftuorometric analysis with Y3 and K 9 - 1 3 6 revealed that the 1450

Vol. 186, No. 3, 1992

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Figure 1. The 28-13-3s epitope includesresiduesin both, the col and cc2 heliciesof Kb as well as the residue at position 1 in the peptide. (a) Ribbondiagramof the ctl and ~2 domainsof H-2Kb as seen by the T cell receptor. The positionsin Kb that contributeto the epitopeof 28-13-3s are indicatedand the wild type and the mutantresidue at that site are shown. (b) A schematicmodelof the N52-59 peptidein the antigenbindinggrooveof Kb molecule.

binding of these m A b were not altered in the three mutants, while the 28-13-3s binding was lost (16 and not shown). Nucleotide sequencing of the entire K b gene in these mutants revealed single point mutations resulting in amino acid substitutions leading to Gly56Glu (R8.8) and Glu58Lys (R8.9) changes in residues o f the col helix. Significantly, a residue in the (x2 helix was affected in the other mutant, R8.10 (Trp 167Arg; 18). Thus, the 28-13-3s epitope included residues situated in both, the o~1 and or2 helices (Fig. la). This pattern of binding raised the possibility that the m A b would bind a region of K b encompassing a part of the peptides bound to K b and hence, the m A b would be a powerful tool to study the orientation o f K b associated peptides. This idea was examined using the vesicular stomatitis virus antigenic octapeptide, N52-59, as a model peptide and employing an approach to induce the cell surface expression of K b molecules on RMA-S cells (15), thus allowing the mapping of the 28-13-3s epitope on a homogeneous preparation of a MHC+peptide complex.

When the naturally processed N52-59 peptide (5) was incubated with RMA-S, up regulation of K b molecules occurred as judged from the increased expression of the Y3 and K9-136 epitopes (Fig. 2b,c). However, a corresponding increase in the 28-13-3s epitope was not seen (Fig. 2d). This is consistent with the probability that the amino acid residue P1Arg (P1 refers to the position and residue number in the peptide) must in some manner obstruct the binding of the 28-13-3s. Therefore,

the

induction

of

surface

expression

of

K b was performed using peptide analogs that had single substitutions of Ala at positions reported to be TCR contact residues in N52-59 peptide (7; Fig. 2a) and tested with the three mAb. All the I

1451

Vol. 186, No. 3, 1992

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

a ~eptides

PI P2 P3 P4 P5 P6 P7 P8

/P~-8

R

|P]A

A

/

G

Y

v

Y

Q

G

L

A A

| P6A b

20

. v~

c

.....

LL

_

0

13 z d

RMA-S(37)

40

~

• 28-13-3s

e

RMA-S(37)

301--~ -

-

-

• B22-249

2.Jl

11

I

O.76

0.21

0.22

0

0.21

0.49

0.06 0.32

=-

=-

,~ ~-

13

~ _~ gz

r

=.

=- =. =-

~;

RMA-S(37)

¢~

gz

r

RMA-S(37)

Figure 2. Cytofluorometric analysis of the antibody epitopes upon binding of N52-59 and its Ala substituted analogs (a) to H-2K b molecule. The data are represented as the ratio of the mean fluorescence intensity (MFI) in the presence of N52-59 or its analogs to the MFI of RMA-S in the absence of peptides. The values above the bar is the standard error of the mean, where n = 6, 6, 13, and 4 for Y3 (b), K9-136 (c), 28-13-3s (d) and B22-249 (e), respectively.

peptide analogs (P1Ala, P2Ala, P4AIa, and P6Ala) induced surface expression of K b as identified by Y3 and K9-136 (Fig. 2b,c). However, the 28-13-3s epitope was identifiable only when the peptide with Ala at P1 was bound to K b and not when the remaining analogs (P2Ala, P4Ala, and P6Ala) which have the wild type Arg residue at P1 were bound to K b (Fig. 2d). Similar loss of 28-13-3s binding was also observed when P5Ala, P7Ala, and P8Ala analogs (which also have P1Arg) of N52-59 were used in the above assay (not shown). Thus, the P1 residue in the peptide forms a part of the 28-13-3s epitope (Fig. lb) consistent with an orientation in which N52-59 is bound to K b along the length of the antigen binding groove with the N-terminus of the peptide proximal to pocket A (pockets are as defined in ref. 14, 19), thus, assuming an orientation similar to that of the c~1 helix of the class I molecule (Fig. lb). 1452

Vol. 186, No. 3, 1992

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

P1Arg in N52-59 probably interferes with the 28-13-3s epitope either because of the size or because of the positive charge of its side chain. The size of the side chain seems less critical to the 28-13-3s epitope since the Sendai virus antigenic nonapeptide which has P1Phe (20) does not affect the binding of the antibody (not shown). Thus, the restoration of the 28-13-3s epitope by P1Ala in the N52-59 analog may be due to the removal of the positive charge from the P1 side chain.

The positioning of the P1 residue of the peptide proximal to pocket A (14, 19) of the antigen binding groove of the K b molecule has several implications.

First, it corroborates our earlier

findings that the side chain of P1Arg is solvent exposed and forms a potential contact site for the TCR (7). Furthermore, the fact that 28-13-3s binds to a majority of native K b molecules expressed on the cell surface also suggests that Arg is very rarely represented at P1 in the self peptides constitutively bound to K b (18).

Second, since N52-59 is a prototypical peptide which includes

the amino acids of the major K b binding motif (7, 21, 22), the orientation of the peptide described here may be representative of most, if not all, peptides bound in the K b antigen binding groove. This proposed model is consistant with the orientation of the generic peptide modeled from the crystallographic data of HLA-B27 (23) and that suggested for antigenic peptides bound to HLA-A2 (24, 25). Furthermore, the recent solution of the three dimensional structure of K b with the N5259 peptide bound to it confirms the findings reported here (26).

ACKNOWLEDGMENTS. We thank Dr. K-i. Shibata for the Ala substituted N52-59 peptide analogs; the Laboratory of Macromolecular Analysis for the synthesis of N52-59 and the mass spectrometic analysis of the analogs; Ms.V. Warren and Ms. C. Rapelje for assistance with the FACScan; and Ms. R. Spata for the secretarial help. SJ is a recipient of the Cancer Research Institute postdoctoral fellowship. This work was supported by grants from the NIH and Irvington Institute for Medical Research to SGN and the NIH training grants.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8.

9. 10. 11. 12. 13.

Zinkernagal, R.M. & Doherty, P.C.(1979)Adv. ImmunoL 27, 51-177. Townsend, A.R.M., Gotch, F.M. & Davey, J. (1985) Cell 42,457-67. Townsend, A.R.M. et al. (1986) Cell 44, 959-968. Townsend, A.R.M. & Bodmer, H. (1989)Annu. Rev. Immunol. 7, 601-624. van Bleek, G.M. & Nathenson, S.G. (1990)Nature (London) 348, 213-216. Falk, K., Rotzschke, O., & Rammensee, H.-G. (1990) Nature (London) 348,213-216. Shibata, K.i., Imarai, M., van Bleek, G.M., Joyce, S. & Nathenson, S.G. (1992) Proc. natl. Acad. Sci. U.S.A. 89, 3135-3139. Hammerling, G.J. et al. (1982) Proc. natL Acad Sci. U.S.A. 79, 4737-4741. Ljunggren, H.G. & Karre, K. (1986) J. Exp. Med. 162, 1745-1759. Ajitkumar, P. et al. (1988) Cell 54, 47-56. Geliebter, J. (1987)Focus 9,5-8. McCabe, P.C. (1990) In PCR Protocols: A Guide to Methods and Applications. Academic Press, New York. Sanger, F., Niclden, S. & Coulson, A.R. (1977)Proc. natl. Acad. Sci. USA 74, 5463-5467. 1453

Vol. 186, No. 3, 1992

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

14. Joyce, S., Garrett, T.P.J., Geliebter, J., Sun, R. & Nathenson, S.G. (1991) In Antigen processing and Recognition (J.McCluskey,Ed.), pp 109-120. CRC Press, Boca Raton. 15. Townsend, A.R.M. et al. (1989) Nature (London) 340,443-448. 16. Geier, S.S., Zeff, R.A., McGovem, D., Rajah, T.V. & Nathenson, S.G. (1986)J. Immunol. 137, 1239-1243. 17. Allen, H., Wraith D., Pala, P., Askonas, B. & Flavell, R.A. (1984)Nature(London)309,279-281. 18. Williams, D.H. Borriello, F., Zeff, R.A., Nathenson, S.G. (1988) J. Biol. Chem. 263, 4549-4560. 19. Saper, M.A., Bjorkman, PJ. & Wiley, D.C. (1991) J. molec. Biol. 219,277-319. 20. Schumacher, T.N.M. et al. (1991) Nature (London) 350, 703-706. 21. Falk, K., Rotzschke, O., Stevanovic, S., Jung, G. & Rammensee, H.-G. (1991) Nature(London) 351, 290296. 22. van Bleek, G.M. & Nathenson, S.G. (1991) Proc. natl. Acad. Sci. U.S.A. 88, 11032-11036. 23. Madden,D.R., Gorga, J.C., Strominger, J.L. & Wiley, D.C. (1991) Nature (London)3 53, 321-325. 24. Latron, F. et al.(1991)Proc, natl. Acad. Sci. U.S.A. 88, 11325-11329. 25. Rotzschke, O. & Falk, K. (1991) Immunol. Today 12, 447-455. 26. Zhang, W., Young, A., Imarai, M.C., Nathenson, S.G. & Saccheteni, J.C. (1992) Proc. natl. Acad. Sci. U.S.A., in press.

1454