Structural Basis for HLA-A2 Supertypes

10 Structural Basis for HLA-A2 Supertypes Pandjassarame Kangueane and Meena Kishore Sakharkar Summary The human leukocyte antigen (HLA) alleles are extremely polymorphic among ethnic population, and the peptide-binding specificity varies for different alleles in a combinatorial manner. However, it has been suggested that majority of alleles can be covered within few HLA supertypes, where different members of a supertype bind similar peptides, yet exhibiting distinct repertoires. Nonetheless, the structural basis for HLA supertype-like function is not clearly known. Here, we use structural data to explain the molecular basis for HLA-A2 supertypes.

Key Words: HLA; alleles; peptide; binding; supertypes; structural basis

1. Introduction The human leukocyte antigen (HLA) alleles are highly polymorphic among ethnic population. Today, more than 1,800 HLA alleles are known and about a 1,000 of them refer to the class 1 loci (1). Class I alleles bind peptides of length 8–10 residues during T-cell-mediated immune response (2). Therefore, the possible combination of specific HLA–peptide binding is large. However, it has been suggested that a majority of alleles can be grouped into few “HLA supertypes,” where the members of a supertype bind similar peptides, yet exhibiting distinct binding repertoires (3). The functional overlap between different alleles within defined supertypes will significantly reduce peptidebinding diversity. A catalog of functional overlap is critical for grouping alleles into supertypes from sequence information. In recent years, a number of supertypes have been defined by comparing peptide-binding data. Thus, HLA-A1 (4), HLA-A2 (3,5), HLA-A3 (5), HLA-A24 (4), HLA-B7 (5), HLA-B27 (4), HLA-B44 (6), HLA-B58 (4), and HLA-B62 (4) supertypes have been defined. From: Methods in Molecular Biology, vol. 409: Immunoinformatics: Predicting Immunogenicity In Silico Edited by: D. R. Flower © Humana Press Inc., Totowa, NJ

155

156

Kangueane and Sakharkar

Alternatively, Chelvanayagam et al. (7), Zhang et al. (8), Zhao et al. (9), and Doytchinova et al. (10) grouped HLA alleles into functionally overlapping clusters from sequence data. Chelvanayagam et al. (7) identified interaction pockets from HLA–peptide crystal structures; Zhang et al. (8) defined A–F structural binding pockets; Zhao et al. (9) defined functional pockets made of critical polymorphic functional residue positions (CPFRP); Doytchinova et al. (10) used molecular interaction fields (MIF), hierarchical clustering (HC), and principal component analysis (PCA); and Lund et al. (11) used clustering procedures for grouping HLA alleles into putative supertypes. However, the structural basis for “supertype-like” HLA function is not clearly known. Here, we use structural complexes of HLA–peptide structures to explain HLA supertypes. 2. Methodology 2.1. HLA Supertype Data HLA-A2 supertype data are obtained from literature (Table 1). These data describe the binding/nonbinding of 25 peptides to A∗ 0201, A∗ 0202, A∗ 0203, A∗ 0206, and A∗ 6802. Table 1 summarizes six peptides binding to all members of the A2 supertypes (A∗ 0201, A∗ 0202, A∗ 0203, A∗ 0206, and A∗ 6802). The functional overlap between different members of the supertype is interesting. It also shows several peptides binding to some members but not all members of the A2 supertypes (Table 1). 2.2. HLA Sequences The protein sequences of HLA-A (295 alleles) were obtained from IMGT/HLA (release 2.5) for this analysis (1). 2.3. Functional Pockets in HLA Structures The CPFRPs were used to define functional pockets in HLA structures (9). HLA-A allele sequences are polymorphic but homologous among themselves. Hence, they have a similar 3D structure in space. However, the polymorphic residues are discontinuously distributed in structure. The residues at CPFRP demonstrated a change in solvent accessibility (ASA) of >0 Å2 upon complex formation in a set of HLA–peptide structures (9) and at least one amino acid polymorphism among 295 HLA-A alleles (IMGT/HLA release 1.14). The 21 CPFRPs thus identified are then classified into virtual pockets for each peptide residue, as shown in Fig. 1.

Structural Basis for HLA-A2 Supertypes

157

Table 1 List of known HLA supertypes Peptide LLFNILGGWV YLVAYQATV KVAELVHFL FLWGPRALV FLLLADARV IMIGVLVGV KIFGSLAFL CLTSTVQLV RLIVFPDLGV YLQLVFGIEV LLTFWNPPV VLVGGVLAA WMNRLIAFA DLMGYIPLV ILHNGAYSL YLSGANLNL VMAGVGSPYV ILAGYGAGV LMTFWNPPV YLVTRHADV HMWNFISGI YLLPRRGPRL LLFLLLADA LLTFWNPPT ALCRWGLLL

Supertypes A∗ 0201 A∗ 0202 A∗ 0203 A∗ 0206 A∗ 6802 Reference A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2 A2

b b b b b b b b b b b b b b b b b b b b b b b b b

b b b b b b b b b b b b b nb b b b b nb nb nb nb b nb nb

b b b b b b b b b b b b b b b b b b b b b b nb b b

b b b b b b b b b b b b nb b nb nb nb nb b b b b nb nb nb

b b b b b b nb nb nb nb nb nb b b nb nb nb nb nb nb nb nb nb nb nb

(12) (12) (13) (13) (12) (13) (13) (13) (12) (13) (13) (12) (12) (12) (13) (13) (13) (12) (13) (12) (13) (12) (12) (13) (12)

Peptides with known binding or nonbinding information are available for five HLA-A alleles. b, binder; nb, nonbinder.

2.4. HLA Supertypes and Virtual Binding Pockets of CPFRP The HLA-A2 supertype data for 25 peptides covering A∗ 0201, A∗ 0202, A∗ 0203, A∗ 0206, and A∗ 6802 are mapped manually to virtual pockets made of CPFRP in Fig. 1. Specific residue at CPFRP is assigned for each HLA allele with known supertype data. The visual representation of supertypelike function for known A2 supertypes along with structurally meaningful virtual pockets consisting of CPFRP provides structural insight into HLA supertype-like function.

158

Kangueane and Sakharkar

Fig. 1. A graphical representation of human leukocyte antigen (HLA)-A2 supertypes with peptides binding to specific HLA alleles is mapped to critical polymorphic functional residue positions (CPFRP) for each of these alleles (Red, nonbinder; Green, binder).

3. Results Table 1 summarizes 25 peptides with known binding/nonbinding data to ∗ A 0201, A∗ 0202, A∗ 0203, A∗ 0206, and A∗ 6802. These peptides bind to more than one HLA allele, and thus, they show overlapping function with two or

Structural Basis for HLA-A2 Supertypes

159

more HLA alleles. Six of these peptides bind to all the five alleles, and eight of these peptides bind to any four of these alleles (Table 1). Table 1 also summarizes that eight peptides bind any three of these alleles and three peptides bind any two of these alleles. Overall, all the 25 peptides show overlapping function with at least two alleles. Figure 1 shows the graphical representation of 25 peptides binding or nonbinding with A∗ 0201, A∗ 0202, A∗ 0203, A∗ 0206, and A∗ 6802. Figure 1 also shows the mapping of each HLA allele to the 21 CPFRPs with their corresponding residues. Among alleles, A∗ 0201, A∗ 0202, A∗ 0203, A∗ 0206, and A∗ 6802, 33% (seven) of CPFRP show variations. However, 66% (14) show no variation among A∗ 0201, A∗ 0202, A∗ 0203, A∗ 0206, and A∗ 6802 at the CPFRP. The virtual pockets formed by the CPFRP are shown for each peptide residue position in Fig. 1. This comprehensive mapping between peptides, alleles, function, CPFRP, and virtual pockets is aimed at explaining the overlapping functional property in HLA-A2 supertypes. 4. Discussion More than 1,800 HLA alleles have been defined (1). Therefore, the number of theoretically possible combinations of HLA–peptide complexes is extremely large. However, the immune system maintains a homogenous balance by specific selection, degeneration, and discrimination (self/non-self) of short peptides using HLA molecules. Although, HLA molecules are polymorphic in ethnic population, they exhibit a substantial amount of functional overlap through the phenomenon of “HLA supertypes,” where members bind similar peptides and yet display distinct repertoires. A number of “HLA supertypes” have already been defined using binding data (Table 1). Table 1 shows six peptides binding to all members of the A2 supertype (A∗ 0201, A∗ 0202, A∗ 0203, A∗ 0206, and A∗ 6802). The functional overlap between different members of the supertype is interesting. These also show several peptides binding to some members but not all members of the A2 supertypes (Table 1). The concept of HLA supertypes is that alleles belonging to supertypes bind a highly shared set of peptides; in principle, it should be possible to predict peptide binding of other members of a supertype using experimental results based on just one member of the type. However, as illustrated in Table 1, this promise does not hold true in the major supertype A2. Hence, the binding of peptides to different members of the A2 supertype is combinatorial in selection and degeneration. Moreover, this grouping is inconclusive, given the known number of HLA alleles. If the molecular basis for supertype-like function of

160

Kangueane and Sakharkar

HLA molecules is known, extrapolation of supertype function to other HLA molecules will be trivial. In Fig. 1, the A2 supertypes (A∗ 0201, A∗ 0202, A∗ 0203, A∗ 0206, and ∗ A 6802) are mapped to their corresponding CPFRP residues. This provides a graphical visualization of four groups of HLA supertypes with the CPFRP residues. For the five HLA alleles, the residues at seven CPFRP (9, 63, 66, 70, 74, 152, and 156) show residue-level changes. However, residues at 14 CPFRPs (7, 24, 35, 73, 76, 77, 80, 81, 97, 99, 114, 116, 163, and 167) are identical among A∗ 0201, A∗ 0202, A∗ 0203, A∗ 0206, and A∗ 6802. The functional overlap among these alleles is partly due to the conservation at the 14 CPFRP. The difference in function among these alleles for some of the peptides given in Fig. 1 is due to the variations at the CPFRP. Figure 1 also shows the virtual pockets defined for each peptide residue position using CPFRP residues. Data show that each of the eight virtual pockets (1, 2, 3, 4, 5, 6, 7, and 9) have at least one mutating residues at the CPFRP. This accounts for the subtle changes associated with the peptide-binding function. In an attempt to explain the difference in peptide-binding function of four groups of peptides with the five alleles, we mapped functional information with CPFRP residues and virtual pockets. Six peptides in G1 (group 1) bind all the five alleles (Fig. 1). These peptides show functional overlap with these alleles, exhibiting supertype-like property. This implies that the residue-level changes at the seven CPFRPs are insensitive to peptide binding in these peptides. However, this is not strictly true for peptides in G2 (group 2), G3 (group 3), and G4 (group 4), as shown in Fig. 1. HLA-A∗ 0201 and HLA-A∗ 0202 show 156L→156W mutation and residue 156 is involved in virtual pockets 3, 4, and 6. The involvement of 156 is deterministic for peptide binding in one peptide in G2, four peptides in G3, and two peptides in G4 (Fig. 1). Comparison of A∗ 0202 and A∗ 0203 with A∗ 0201 shows 156L→156W between A∗ 0201 and A∗ 0202 and 9F→9Y between A∗ 0201 and A∗ 0203. Between A∗ 0201 and A∗ 0206, 152V→152E and 156L→156W changes are observed. These changes at 9 [AND, OR] 152 [AND, OR] 156 affect binding of peptides to A∗ 0202, A∗ 0203, and A∗ 0206 despite their binding to A∗ 0201. Comparison of A∗ 6802 with A∗ 0201 in Fig. 1 shows changes at six positions (9, 63, 66, 70, 74, and 156). Thus, 17 peptides that bind to A∗ 0201 are nonbinders to A∗ 6802. These data indicate that residues at CPFRP determine functional overlap between alleles in A2 supertype. However, it is important to generate mapping matrices incorporating functional overlap at multiple layers for gathering a more clear picture of “supertype-like” function in future investigations.

Structural Basis for HLA-A2 Supertypes

161

5. Conclusion HLA–peptide binding is useful in the design of peptide vaccine candidates, immunotherapeutic targets, and diagnostics agents. The theoretically possible combinations are overwhelmingly large. However, the functional overlap between alleles occurs at the level of supertypes. An understanding of their structural principles has a significant role in generating supertypes from sequence. Here, we show that the 21 CPFRPs have a role to play in determining overlapping function between two or more alleles. The 14 conserved CPFRPs explain overlapping function, and the 7 nonconserved CPFRPs explain nonoverlapping function. We hope to create a much clear picture of this phenomenon in future studies. References 1. Robinson J., M.J. Waller, P. Parham, N. de Groot, R. Bontrop, L.J. Kennedy, P. Stoehr & S.G.E. Marsh: IMGT/HLA and IMGT/MHC – sequence databases for the study of the major histocompatibility complex. Nucleic Acids Res. 31(1) 311–314 (2003). 2. Yewdell J.W., E. Reits & J. Neefjes: Making sense of mass destruction – quantitating MHC class I antigen presentation. Nat. Rev. Immunol. 3(12) 952–961 (2003). 3. Del Guercio M.F., J. Sidney, G. Hermanson, C. Perez, H.M. Grey, R.T. Kubo & A. Sette: Binding of a peptide antigen to multiple HLA alleles allows definition of an A2-like supertype. J. Immunol. 154(2) 685–693 (1995). 4. Sette A. & J. Sidney: Nine major HLA class I supertypes account for the vast preponderance of HLA-A and -B polymorphism. Immunogenetics 50(3–4) 201–212 (1999). 5. Sette A. & J. Sidney: HLA supertypes and supermotifs – a functional perspective on HLA polymorphism. Curr. Opin. Immunol. 10(4) 478–482 (1998). 6. Sidney J., S. Southwood, V. Pasquetto & A. Sette: Simultaneous prediction of binding capacity for multiple molecules of the HLA B44 supertype. J. Immunol. 171(11) 5964–5974 (2003). 7. Chelvanayagam G.: A roadmap for HLA-A, HLA-B, and HLA-C peptide binding specificities. Immunogenetics 45(1) 15–26 (1996). 8. Zhang C, A. Anderson & C. DeLisi: Structural principles that govern the peptidebinding motifs of class I MHC molecules. J. Mol. Biol. 281(5) 929–947 (1998). 9. Zhao B., A.E.H. Png, E.C. Ren, P.R. Kolatkar, V.S. Mathura, M.K. Sakharkar & P. Kangueane: Compression of functional space in HLA-A sequence diversity. Hum. Immunol. 64(7) 718–728 (2003). 10. Doytchinova I.A., P. Guan & D.R. Flower: Identifying human MHC supertypes using bioinformatics methods. J. Immunol. 172(7), 4314–4323 (2004). 11. Lund O., M. Nielsen, C. Kesmir, A.G. Petersen, C. Lundegaard, P. Worning, C. Sylvester-Hvid, K. Lamberth, G. Roder, S. Justesen, S. Buus & S. Brunak:

162

Kangueane and Sakharkar

Definition of supertypes for HLA molecules using clustering of specificity matrices. Immunogenetics 55(12), 797–810 (2004). 12. Scognamiglio P., D. Accapezzato, M.A. Casciaro, A. Cacciani, M. Artini, G. Bruno, M.L. Chircu, J. Sidney, S. Southwood, S. Abrignani, A. Sette & V. Barnaba: Presence of effector CD8+ T cells in hepatitis C virus-exposed healthy seronegative donors. J. Immunol. 162(11), 6681–6689 (1999). 13. Kawashima I., S.J. Hudson, V. Tsai, S. Southwood, K. Takesako, E. Appella, A. Sette & E. Celis: The multi-epitope approach for immunotherapy for cancer: identification of several CTL epitopes from various tumor-associated antigens expressed on solid epithelial tumors. Hum. Immunol. 59(1), 1–14 (1998).