| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Epimmune Inc., San Diego, CA 92121;
Children’s Research Institute and The Ohio State University, Columbus, OH 43205; and
The Scripps Research Institute, La Jolla, CA 92037
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The organization and structure of MHC genes are also conserved in
humans and both chimpanzee species, P. troglodytes (the
common chimpanzee) and Pan paniscus (the pygmy chimpanzee or
bonobo) (6, 7). All three species have orthologous class I
MHC loci designated A, B, and C that display a remarkable degree of
allelic variation. Most of the nucleotide diversity is clustered in
exons 2 and 3 encoding the 
1 and 2 domains of class I MHC
proteins, and facilitates binding of a constellation of self and
non-self peptides important for CD8+ T cell
control of tumors and intracellular parasites. Diversity is generated
in a two-step process involving point mutation followed by
recombination events that include allele and gene conversions
(8). Of all nonhuman primate species, the extent of class
I allelic diversity is most thoroughly characterized in populations of
captive common chimpanzees. Remarkably, the majority of nucleotide
polymorphisms that define the A, B, and C loci are conserved in humans
and chimpanzees, indicating that they were transmitted from a common
ancestor some 7–10 million years ago (9). Although these
genes have no general species-defining characteristics, it is also true
that identical class I alleles have not yet been found in humans and
chimpanzees. Nevertheless, domains of the class I MHC peptide binding
groove are highly conserved in the most closely related molecules found
in human and chimpanzee species, suggesting the potential for
presentation of an overlapping or identical set of peptides to
CD8+ T cells. This is illustrated by a study of
common and pygmy chimpanzees (10). Although these species
diverged about 2.5 million years ago, their class I complexes,
designated Patr (P. troglodytes) and Papa (P.
paniscus), encode alleles that present a viral peptide across the
species barrier. Thus, Patr-A*0401 and Papa-A*06 class I molecules
differed by six amino acids, but both presented a peptide derived from
HCV with equal efficiency to a CD8+ CTL line derived from
an infected common chimpanzee (10).
This similarity was further underlined by recent studies that demonstrated that HLA supertypes extend to chimpanzees (11). Specifically, it was shown that peptides characterized by cross-reactive binding capacity for multiple common members of the human HLA-A2 or -B7 supertypes also frequently bound with appreciable affinity to chimpanzee-derived BCL lines. In the series of experiments described herein, we sought to confirm and expand these observations, utilizing a larger panel of radiolabeled peptide ligands. Our goals were to identify the specific Patr class I molecules capable of cross-binding human HLA supertype peptides and to establish their peptide binding motifs.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
EBV-transformed B cell lines (BCL) had previously been established from PBMC of nine healthy young adult chimpanzees as described (11). BCL were grown in culture medium consisting of RPMI 1640 medium with HEPES (Life Technologies, Rockville, MD) supplemented with 10% FBS, 4 mM L-glutamine, 50 µM 2-ME, 0.5 mM sodium pyruvate, 100 µg/ml streptomycin, and 100 U/ml penicillin (Gemini BioProducts, Calabasas, CA).
Peptide synthesis
Peptides were either synthesized at Epimmune (San Diego, CA) as previously described (12) or purchased as crude material from Mimotopes (San Diego, CA). Peptides synthesized at Epimmune were purified to >95% homogeneity by reverse phase HPLC. The purity of these synthetic peptides was assayed on an analytical reverse-phase column, and their composition was ascertained by mass spectrometry analysis.
Cloning, sequencing, and analysis of Patr class I cDNA
Isolation of total RNA from BCL, first-strand cDNA synthesis, and Patr class I amplification was performed as described elsewhere (13). Patr class I genes amplified by PCR with primers specific for products of the A, B, or C loci were cloned into the expression plasmid PBJ1neo (14) and bidirectional DNA sequencing was performed by fluorescent dye termination using an ABI 377 automated sequencer (Perkin-Elmer Biosystems, Foster City, CA). A consensus sequence was derived by sequencing at least three clones per allele, and a clone identical with the consensus sequence was then selected for transfection into the class I-deficient human BCL 721.221. Detailed methods for transfection of these cells have been described in detail (15). Briefly, 10–20 µg of plasmid DNA was electroporated into 721.221 cells (Model 600 ECM electroporation system; BTX, San Diego, CA). Cells were then cultured for 2 days in RPMI containing 30% FCS. After that, cells were maintained in RPMI supplemented with Geneticin (1.5 mg/ml of active drug; Life Technologies). Cultures were then screened for actively growing transfectants and sorted by flow cytometry using the fluorescein-conjugated mAb W6/32 (16).
Class I purification
Transfected 721.221 cells expressing single Patr molecules were used as the source of class I molecules. Lysates were prepared from cell pellets, and class I molecules were purified as previously described (17). Briefly, cells were grown in roller bottles and lysed at a concentration of 108 cells/ml in 50 mM Tris-HCl, pH 8.5, containing 1% Nonidet P-40 (Fluka Biochemika, Buchs, Switzerland), 150 mM NaCl, 5 mM EDTA, and 2 mM PMSF. The lysates were passaged through 0.45-µM filters and cleared of nuclei and debris by centrifugation at 10,000 x g for 20 min. MHC molecules were then purified by affinity chromatography. Columns of inactivated Sepharose CL4B and protein A-Sepharose (Amersham Pharmacia Biotech, Piscataway, NJ) were used as precolumns. Subsequently, a column of the anti-HLA (A, B, and C) mAb W6/32 (18) was used to capture Patr molecules. Protein purity and concentration were monitored by SDS-PAGE.
Binding assays
Quantitative assays for the binding of peptides to soluble class
I molecules on the basis of the inhibition of binding of a radiolabeled
standard probe peptide to detergent-solubilized MHC molecules were
performed as previously described (17). The various
peptides used as radiolabeled probes are shown in Table I
. Briefly, 1–10 nM radiolabeled probe
peptide, iodinated (125I; Amersham, Piscataway,
NJ) by the chloramine T method, was coincubated at room temperature
with various amounts of MHC in the presence of 1 µM human
ß2-microglobulin (Scripps Laboratories, San
Diego, CA) and a cocktail of protease inhibitors (containing a final
concentration of 1.07 mg/ml EDTA, 62.5 µg/ml pepstatin A, 325 µg/ml
phenanthroline, 250 µg/ml PMSF, and 60 µg/ml TLCK (Sigma, St.
Louis, MO, and Calbiochem, La Jolla, CA)). At the end of a 2-day
incubation period, the percent of MHC-bound radioactivity was
determined by size exclusion gel filtration chromatography on a TSK
2000 column (TosoHaas, Montgomeryville, PA), or by capture on mAb
W6/32-coated Flash plates (NEN, Boston, MA) counted on a TopCount
instrument (Packard, Meriden, CT).
|
15% binding of the radiolabeled probe peptide were used
for all competitive inhibition assays. Because under these conditions
[label] < [MHC] and IC50 
[MHC], the measured
IC50 values are reasonable approximations of the true
Kd values. Each competitor peptide was
tested in two to four completely independent experiments. Unlabeled
probe peptide was also tested in each experiment, and its
IC50 was measured as a positive control for
inhibition. Live cell binding assay
This assay was performed as a modification of the assay described by Bertoni et al. (11). BCL were washed once in complete RPMI with 5% FBS and incubated overnight at room temperature to increase the class I expression on the cell surface. The next morning, the cells were washed twice in RPMI without FBS, and resuspended in the same medium with human ß2-microglobulin at a final concentration of 3 µg/ml. Ninety-six-well tissue culture plates were prepared containing, in duplicate wells, 105 cpm radiolabeled peptide, 10-fold serially diluted competitor peptides (from 10 µg/ml down to 100 pg/ml final concentration), and a protease inhibitor mixture as described above. Finally, 2 x 106 BCL were added to each well, and the plates were incubated at 20°C for 4 h. At the end of the incubation period, the excess label was removed by washing the cells three times with RPMI, and free and cell-bound peptides were separated by centrifugation of the cells through a FBS gradient. The cells were then counted on a gamma counter.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
As a prelude to the molecular definition of Patr molecules capable
of binding prototype HLA supertype peptides, the specific class I MHC
molecules expressed by BCL derived from three chimpanzees, Wilma
(1558), Hoss (1530), and Barry (1564), were identified by PCR
amplification with Patr A- and B-specific primers, as previously
described (13). Remarkably, 10 of 11 class I A and B
alleles identified in these animals (Table II) were described previously in other
studies of chimpanzees housed at a variety of facilities in Europe and
North America. The only unique allele was Patr-A*0602 (GenBank
accession number AF302080), a subtype that differs from Patr-A*0601 by
one nonsynonymous nucleotide substitution in exon 2.
|
Class I Patr molecules were isolated by standard affinity
chromatography methods from the single Patr transfectants described
above. These purified class I molecules were then tested for their
capacity to bind the prototype HLA supertype peptides listed in Table I
. The peptides chosen represent four of the nine major HLA supertype
binding specificities currently known (22). More
specifically, the A3Con1 peptide (19) was utilized as
prototype A3 supertype binder. In addition, the binding capacity of the
Hu J chain 102–110 and A24Con1 peptides, chosen as representative
ligands of the recently described A1 and A24 supertypes
(22), respectively, were also investigated. Finally, the
previously described B7-specific natural ligand, Hu J chain 5–13
(21), was chosen as the prototype B7 supertype ligand.
Specific signals were detected in the case of several Patr
molecule/radiolabeled peptide combinations, with 10–15% of the
labeled peptide bound by as little as 1–10 nM concentrations of
purified class I molecules in most cases. More specifically,
Patr-A*0101 bound the A3 Con1 peptide (Fig. 1A). Patr-B*0901 bound Hu J
chain 102–110 (Fig. 1B), and Patr-B*1301 selectively bound
the Hu J chain 5–13 ligand (Fig. 1C). Weaker binding was
also noted for the A24Con1 peptide to Patr-A*0701 (Fig. 1D).
In all of these instances, the binding was inhibitable by excess
unlabeled peptide, with IC50 in the range of 0.4 - 165 nM
(data not shown).
|
Definition of the binding motif associated with Patr-A*0101 class I molecules and comparison with the HLA-A*0301 binding specificity
To define the peptide binding motifs of the various chimpanzee-derived class I molecules described above, panels of single amino acid substituted analogues of the various prototype peptide ligands were tested. For this analysis, a main anchor residue is arbitrarily defined as a residue in which >50% of the substitutions tested are associated with a greater than 10-fold change in binding capacity. Likewise, a secondary anchor is defined as a residue in which some (but 50% or less) of the substitutions tested are associated with a >10-fold change in binding capacity.
The specificity of the Patr-A*0101 class I molecule was analyzed first.
Nonconservative amino acid substitutions (K or D) were introduced at
every position of the A3Con1 peptide, and their binding capacity was
quantified in the Patr-A*0101 binding assay (Fig. 2A). Decreases in binding
affinity of >10-fold were detected only at the C terminus.
|
For the sake of comparison, the same set of substitutions was also
tested for their capacity to bind the human class I molecule,
HLA-A*0301, which is a prototype molecule of the HLA-A3 supertype (Fig. 2B). In agreement with previous studies (19),
significant effects were noted at position 2 and the C terminus.
In position two, whereas aromatic or hydrophobic substitutions had no effect or even increased binding capacity, the two charged residues K and D were associated with decreased binding, and modest decreases were also noted with the polar N substitution. At the C terminus, a binding pattern very similar to that observed with Patr-A*0101 was noted. The only significant difference was associated with the A substitution, which was associated with higher binding capacity in the case of Patr-A*0101, but had little effect on HLA-A*0301 binding. In addition, a significant decrease in binding was also noted for the D substitution at position 1, which is a secondary anchor residue for HLA-A*0301 binding (23). Taken together, these results illustrate how a similar, but clearly distinct, fine binding specificity is associated with the two class I molecules Patr-A*0101 and HLA-A*0301.
Definition of the binding motifs associated with the Patr-B*0901 and Patr-A*0602
The binding motifs associated with the two alleles Patr-B*0901 and
Patr-A*0602 were analyzed next, following the same strategy outlined
above. These two alleles are of interest because they both bind Hu J
chain 102–110, a naturally occurring high affinity ligand of
HLA-A*0101, the prototype molecule of the A1-supertype
(22). Nonconservative substitutions (K) were introduced at
every position of the Hu J chain 102–110 peptide, and the binding
capacity of the corresponding peptides was quantified in Patr-B*0901
and Patr-A*0602 binding assays (Fig. 3).
Greater than 10-fold decreases in binding affinity were detected at
positions 2 and the C terminus of both molecules, suggesting their
potential role as main peptide-binding anchors. In addition, >10-fold
effects were also detected in positions 3, 7, and 8 for Patr-B*0901 and
3, 6, and 7 for Patr-A*0602. We speculated that these positions might
act as secondary anchor residues.
|
A), small residues such as S or T were
preferred in position 2. A substitution of A was tolerated but
associated with about 10-fold lower binding capacity. All other
substitutions at this position were associated with >100-fold
decreases in binding capacity. At the C termini, aromatic (W, F, and Y)
and hydrophobic (L) residues were associated with increased binding
capacity, whereas all other substitutions tested (N, D, A, and K) were
associated with decreases in binding capacity of 100-fold or
greater. Position 3 appeared to function as a secondary anchor, with predilection for negatively charged residues (D and E). Significant decreases were also noted for the F and K substitutions. An important role was also ascribed to positions 7 and 8, which appeared to prefer negatively charged or amide (position 7) or uncharged (position 8) residues, respectively. Overall, the Patr-B*0901 peptide binding specificity closely resembles the known binding specificity of the human HLA-A*0101 allele, which is associated with a preference for S or T in 2, D or E in 3, and aromatic residues at the C termini (19).
The specificity of Patr-A*0602 was examined next. For this allele, position 2 was less important, and only D, N, and K substitutions were not accepted. By contrast, position 9 was very selective, and all substitutions tested at this position led to 100- to 10,000-fold decreases in binding. Position 3 was also an important anchor for Patr-A*0602, with most substitutions negatively impacting binding. Furthermore, a very prominent role in determining binding capacity appeared to be played by positions 6 and 7, where most of the substitutions tested also impacted binding. At position 6, only P was tolerated, whereas, at position 7, only L was tolerated.
Based on these results, we conclude that Patr-B*0901 and Patr-A*0602 share overlapping but yet distinct peptide binding motifs. In particular, the motif recognized by the Patr-B*0901 molecule is most similar to the motif associated with the human class I molecule HLA-A*0101. The motif associated with Patr-A*0602 has also several unique features, including the apparent lack of strict dependence on position 2 as an anchor, and the reliance of positions 3, 6, and 7 as additional potentially main anchors.
Definition of the binding motifs recognized associated with the Patr-B*1301 and Patr-A*0701 class I molecules
Additional experiments were designed to define the peptide binding
motifs recognized by Patr-B*1301 and Patr-A*0701. First, the
specificity of Patr-B*1301 was defined following a strategy similar to
the one described above for the other Patr class I molecules (Fig. 4A). Position 2 and the C
terminus were mapped as the main anchor residues, and position 7 and
possibly 6 were identified as secondary anchors. Notably, only P was
allowed at position 2, and all other substitutions were associated with
>100-fold decreases in binding capacity. At the C terminus, a broad
specificity for either aromatic or aliphatic residues was detected. It
was noted that this main anchor specificity is essentially identical
with the known binding specificity of the human class I molecule
HLA-B*0702, in particular, and the B7-supertype in general
(21). Additionally, the Patr-B*1301 motif described is in
accordance with previously described HCV epitopes targeted by
CD8+ T cells from an infected chimpanzee
(16).
|
D). A number of different single
amino acid substitutions of the A24Con1 peptide were synthesized and
tested for binding in the Patr-A*0701 binding assay (Fig. 4B). Greater than 10-fold effects were noted for positions 2
and 6 and the C terminus. The most drastic effects were seen at
position 2 and the C terminus, which were designated as primary anchor
residues, and were associated with a specificity for aromatic/aliphatic
residues in 2, and aliphatic or F at the C terminus. Position 6
appeared to function as a secondary anchor residue, with positive
effects noted in the case of hydrophobic residues (F, L, and A), and
detrimental effects associated with charged substitutions (D and K).
This motif is similar to the previously described HLA-A*2402 motif
(19, 24). Pocket analysis of Patr class I molecules
The data presented in the preceding sections define peptide motifs
specific for several Patr class I molecules. These motifs, summarized
in Table III, share several general
features in common with specific HLA supertype motifs, summarized in
Table IV. To determine whether these
peptide binding specificities could be correlated with structure, the
Patr residues predicted to form the B and F pockets (25)
were tabulated, and compared with the B and F pockets of HLA class I
molecules whose peptide binding specificity was also known (Table V). For this analysis, we examined the
polymorphic residues known to form the B and F pockets of HLA-A*02, as
originally defined by Saper et al. (26). A structural
model of the HLA-A*0201 molecule can be accessed online in the Protein
Data Bank at http://www.rcsb.org/pdb/(PDB ID, 3HLA).
|
|
|
), it was noted that class I
molecules with a preference for S or T, such as Patr-B*0901 and
HLA-A*0101, were associated with the consensus motif of M at residue
45, E at residue 63, N at residue 66, and M at residue 67.
Interestingly, Patr-B*0101, whose peptide binding specificity in
unknown, shares the same consensus residues in the B pocket, and thus
would be predicted to share the same preference for binding small
residues. Alleles with a preference for P (Patr-B*1301 and HLA-B*0702)
were also associated with a common B pocket structural motif (Y 9, E
45, N 63, I 66, Y 67, and Q 70). No other Patr molecules have been
identified with an identical pocket structure. Finally, Patr-A*0701 and
HLA-A*2402 share a similar binding pocket, with identity of S 9, M 45,
E 63, and V 67. The B pocket of Patr-A*0701 lacks the positively
charged residues seen at positions 66 (K) and 70 (H) of HLA-A*2402,
which may account for its ability to accommodate a larger repertoire of
residues. A similar pattern of residues was also observed in
Patr-A*0401, -A*0402, -A*0601, and -A*0901. These alleles would be
predicted to share a preference for hydrophobic or aromatic residues
with Patr-A*0701. In fact, three of four minimal optimal HCV
epitopes presented to CD8+ CTL by either
Patr-A*0601 or -A*0901 have a tryptophan at position 2 (C.M.W.,
unpublished data).
In the case of the F pocket (Table V), alleles with a preference for
positively charged residues at the C terminus (Patr-A*0101 and
HLA-A*0301) share an identical F pocket, lined with D at residue 77, T
at residue 80, L at residue 81, and D at residue 116. Patr-A*0401 and
-A*0402 also share an identical F pocket, and would be predicted have a
preference for peptides with positively charged residues at the C
terminus.
Patr-A*0602 and HLA-A*0101, which have a shared preference for Y at the C terminus, share an identical F pocket with N 77, T 80, L 81, and D 116. This exact pattern is also seen in Patr-A*0601 and -A*1101, which would thus be predicted also to be specific for peptides with Y at the C terminus. A different pattern of F pocket residues is seen in Patr-B*0901 and HLA-A*2402, which share N 77, I 80, and A 81, and have either D or Y at position 116. These alleles have similar specificity in terms of C-terminal residues of their peptide ligands. Patr-B*0101 shares an identical binding pocket with HLA-A*2402, and would be expected to show a preference for similar residues.
Finally, an identical pattern of F pocket residues is seen for Patr-B*1301 and HLA-B*0702, which bind peptides ending with hydrophobic nonpolar residues. The residues lining this pocket are S 77, N 80, L 81, and Y 116. Patr-B*1601 and -B*1701 also share this pocket structure, and would be predicted to bind peptides with hydrophobic residues at the C terminus. It should also be noted that Patr-B*1301 and HLA-B*0702 are identical in both the B and F pockets. In conclusion, the similarities between HLA and Patr molecules at the level of binding motifs and described in the preceding sections are also mirrored by structural similarities in the B and F bonding pockets of the same HLA and Patr molecules.
Partial overlaps in the peptide-binding repertoire of other Patr class I molecules and HLA supertype molecules
Based on the data presented above, it may be hypothesized that the
repertoire of peptides bound by certain chimpanzee and human class I
molecules would overlap significantly. To test this theory, we measured
the capacity of known HLA supertype epitopes to bind appropriate Patr
alleles. The Patr-B*1301 allele, whose peptide binding motif was shown
above to be similar to the one recognized by the human HLA-B*0702
molecule, was investigated first. Specifically, nine HBV-derived
peptides carrying the HLA-B7 supermotif, and capable of binding to
multiple HLA-B7 supertype molecules (27), were tested for
their capacity to bind Patr-B*1301. When the results (Table VI) were compared with previously
published data regarding the capacity of the same peptides to bind
purified HLA-B*0702, a large overlap became apparent. In fact, all of
the peptides that bound HLA-B*0702 with IC50 of <500 nM,
also bound Patr-B*1301 (Table VI).
|
. A
significant degree of overlap was noted in the case of the Patr-A*0101
and the HLA-A3 supertype, where in total 23 of 25 (92%)
HLA-A*0301-binding peptides also bound Patr-A*0101. A high degree of
overlap was also noted in the case of Patr-B*1301 and HLA-B*0702, where
12 of 13 (92%) HLA-B*0702-binding peptides also bound to Patr-B*1301.
The overlap in peptide binding between Patr-B*0901 and HLA-A*0101 was
also considerable, as 12 of 21 (57%) of HLA-A*0101 peptides also bound
Patr-B*0901. An intermediate degree of overlap was noted in the case of
the Patr-A*0701 and HLA-A*2402, where 7 of 14 (50%) HLA-A*2402 binders
also bound Patr-A*0701. Finally, a lower degree of overlap was noted in
the case of the Patr-A*0602 and the HLA-A*0101, where only 8 of 21
HLA-A*0101 binders (38%) also bound Patr-A*0602. This is not
surprising, as the position 2 and 3 specificities of Patr-A*0602 and
HLA-A*0101 are not as tightly correlative as noted for the other
Patr/HLA overlaps described above.
|
A previous study (11) has demonstrated that peptides
characterized by the capacity to bind multiple common members of the
HLA-A2 or -B7 supertypes, also frequently bind, with appreciable
affinity, to chimpanzee-derived BCL. Herein, we sought to expand these
observations to the HLA-A3, -A1, and -A24 supertypes, utilizing a panel
of chimpanzee-derived lines previously characterized for their capacity
to bind HLA-A2 and -B7 supertype peptides by Bertoni and colleagues
(11). Accordingly, radiolabeled HLA-A1, -A3, and -A24
prototype peptides (Table I) were tested for their capacity to bind
live chimpanzee BCL.
The results obtained are summarized in Table VIII. Significant binding was noted for
various peptide/BCL combinations with absolute bound cpm values ranging
from 1032 to 8533 cpm. Binding of the HLA-A3 supertype binder was
detected in 5 of 9 (56%) of the cell lines tested. Likewise, binding
of the HLA-A1 and -A24 supertype binders was detected in 2 of 9 (22%)
and 5 of 9 (56%) of the cell lines tested. The binding was specific,
in that no label bound all lines. Rather, every label and cell line was
associated with a unique binding pattern.
|
, in parentheses.
These values ranged between 1.3 and 385 nM, indicating that a range of
different affinities was detected. A general correlation was also
noted, as expected, between the magnitude of the signal (in terms of
bound cpm) and the affinity of interaction (estimated by
IC50 values). In conclusion, the results presented herein
demonstrate that HLA-A3, -A1, and -A24 prototype supertype peptides
bind with appreciable frequencies to Patr molecules on chimpanzee cells
isolated from chimpanzees in primate centers. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
These results allow, for the first time, a glimpse of the peptide motifs associated with specific Patr class I molecules. The general features of these motifs were similar to those associated with human HLA class I molecules. Position 2 and the C terminus played, in most cases, the role of main anchors, with residues 3, 6, 7, and 8 playing the role of additional, mostly secondary, anchors. Positions 4–5 appeared to be the most permissive in terms of class I binding, and probably correspond to positions protruding away from the class I molecule and accessible for TCR recognition (25, 26).
The definition of peptide-binding motifs for Patr class I molecules will allow a search for chimpanzee-restricted disease-specific peptides predicted to bind to the various Patr alleles, and potentially induce an immune response against the disease of interest. In turn, this will allow more exact monitoring of CTL response, and also provide the molecular basis for the design and testing of epitope-based vaccines in chimpanzees.
Our results confirm the results of Bertoni et al. (11) that HLA supertypes extend to chimpanzees. Furthermore, our study extends these observations to include the HLA-A3, -A1 and -A24 supertypes. The present study unequivocally identifies some of the Patr molecules capable of cross-reacting with human HLA molecules at the level of peptide binding specificity.
In terms of specific motifs, the A3Con1 peptide, which binds several HLA molecules of the HLA-A3 supertype (23), also binds to purified chimpanzee Patr-A*0101. The motif defined for Patr-A*0101 overlaps considerably with the motif defined previously for HLA-A*0301, and differs primarily at position 2, where the chimpanzee class I molecule is associated with a less-stringent specificity.
The HLA-A1 prototype peptide, Hu J Chain 102–110 (20), bound to two purified Patr molecules, Patr-B*0901 and Patr-A*0602. Like HLA-A*0101, Patr-B*0901 preferred serine or threonine at position 2. In the case of Patr-A*0602, position 2 played the role of secondary anchor, and an unusual pattern of main anchors at positions 3 (alanine), 6 (proline), and 7 (leucine) was detected instead. The apparent main anchor role of positions 6 and 7 may be related to conformational constraints related to the presence of a P residue in position 6, and might be a peculiarity of this particular peptide ligand. Indeed, additional high affinity A*06 ligands already identified do not carry P in position 6 (data not shown).
Patr-B*1301 binds the B7 supertype binding peptide Hu J Chain
5–13 L7
Y. The Patr B*1301 molecule was cloned from cells from
chimpanzee 1564, which bound the radiolabeled B7 peptide in the
previous study (11). In this study, the B7 peptide bound
only this single Patr molecule, as would be predicted from the previous
results. The motif defined for Patr-B*1301 is virtually identical with
the motif previously defined for HLA-B*0702 (28, 29), and
consists of P at position 2 and aromatic or hydrophobic residues at the
C terminus of the peptide. Finally, the HLA-A*2402-binding peptide,
A24Con1, bound Patr-A*0701. The motif for this Patr molecule was shown
to be indeed similar to that of HLA-A*2402.
We have also analyzed the residues lining the B and F pockets of
various Patr class I molecules and compared them with the residues
present in the same positions of various human HLA class I molecules.
In many cases, similarities in both F and B pockets were detected, and
these provide a structural basis for the molecular cross-reactivities
detected at the level of binding specificity. This type of analysis
should allow rational prediction of motifs for Patr molecules for which
motifs have not been yet determined. For example, Patr-B*0101, a
molecule utilized in our study, did not bind any of the radiolabeled
HLA-restricted peptides. Examination of the residues lining the B
pocket revealed a very similar pattern with HLA-A*0101 and Patr-B*0901
(Table V), molecules for which a motif of S or T at position 2 is
known. The residues lining the F pocket were identical with HLA-A*2402.
Thus, the motif for Patr-B*0101 would be predicted to have S or T at
position 2 and F, L, I, or W at the C terminus. This prediction is now
being tested experimentally.
In future studies, it might be of interest to elute naturally processed peptides from the same Patr class I molecules. This represents an alternative and complementary method to define MHC-specific motifs. Indeed, in a previous study, the motifs for HLA-A1, -A3, -A11, and -A24 were determined both by acid elution of peptides as well as by the method described herein (19). Equivalent results were reached by both methods, but the method based on peptide-binding assays allows a more accurate quantification of the binding potential of disease-specific epitopes.
It is interesting to interpret the results presented in this study in the context of known epitopes recognized by CD8+ T cells from HCV-infected chimpanzees. For instance, based on similar B and F pocket structures, HLA-A*2402 and Patr-B*0101 should bind peptides with T in position 2 and I in position 9. Patr-B*0101 does indeed present an epitope from the HCV nonstructural 3 (NS3) protein (YTGDFDSVI) that fits this prediction (30). Another NS3 epitope (VPHPNIEEV) has the position 2 (P) and 9 (V) motif that facilitates peptide binding to Patr-B*1301 and HLA-B*0702 molecules (30). Finally, a positively charged amino acid at the COOH terminus of a peptide serves as an anchor for binding to HLA-A*0301, Patr-A*0101, and probably Patr-A*0401. Earlier studies demonstrated that Patr-A*0401-restricted T cells target an HCV E2 epitope containing either K or R at its COOH terminus (30), and two of three Patr-A*0101-restricted HCV epitopes also contain a K residue at this position (data not shown). Thus, not only do human supertype motifs facilitate peptide binding to chimpanzee class I molecules in vitro, they probably also shape the immune response against virus infection in this animal model.
Conservation of peptide binding pockets is probably rooted in the evolutionary relationship between HLA, Patr, and Papa class I molecules. Six families of HLA-A alleles can be defined based on serologic cross-reactivities and shared nucleotide polymorphisms (31, 32, 33). Three families represented by the HLA-A9, -A80, and -A1/A3/A11 molecules probably derived from an ancient A3 lineage, whereas the other three families represented by HLA-A2, -A10, and -A19 derived from an ancient A2 lineage. Perhaps unexpectedly, all Patr-A alleles characterized to date are related to only the HLA-A1/A3/A11 family (6, 7, 34, 35, 36). This supports the concept of transspecies evolution involving transmission of MHC alleles from one species to the next (9), and indicates that chimpanzees may have inherited only part of the human allelic repertoire from a common ancestor. Evolution of MHC molecules in both species occurs by reassortment of point mutations through intra- and interlocus recombination (8). At least for the A locus, this process has been relatively slow (35), and could provide a partial explanation for common peptide binding pockets found in contemporary molecules such as HLA-A*0301 and Patr-A*0101.
Human and chimpanzee class I B alleles evolved somewhat more rapidly
because of a higher rate of recombination at this locus (8, 37). Although this has obscured lineage relationships between
class I B alleles from Homo sapiens and P.
troglodytes, some domains display sequence similarities. This is
most evident for the -1 domain of HLA-B*0702 and related molecules
such as Patr-B*1301 in common chimpanzees and Papa-B*01 and -B*04 in
bonobos (37). A high degree of sequence homology in the
-1 domain, which forms the B pocket of the peptide-binding groove,
indicates that these molecules arose from a common ancestor and that
structural or functional constraints limit mutations. The -2 domain
comprising the F pocket is considerably more divergent in class I B
molecules of chimpanzees and humans, although amino acid residues
governing peptide binding are similar in HLA-B*0702 and Patr-B*1301
molecules (37). Why molecules that diverged several
million years ago have conserved peptide-binding motifs is not clear,
but we speculate that they might mediate effective immune responses
against pathogens common to both species.
Finally, we would like to underline the implications of the present study for the design and testing of vaccines in chimpanzees. The high degree of cross-reactivity of human epitopes for certain Patr molecules may allow the utilization of human epitopes to monitor CTL responses in chimpanzees, as well as for the design of human epitope-based vaccines that could be tested for efficacy in chimpanzees. Based on similarities detected, we have also determined the degree of cross-reactivity between specific Patr and HLA molecules, in terms of their capacity to bind various epitopes, as candidates for inclusion in multiepitope vaccines.
A significant degree of cross-reactivity was detected between several Patr and HLA class I molecules. These data, together with the live cell binding presented here and in the previous study of Bertoni (11), suggest that the degree of cross-reactivity between HLA supertypes and Patr class I molecules will be appreciable. This provides further data to support the concept that the design and testing of experimental vaccines destined for human use in chimpanzees will be a valuable approach.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Denise M. McKinney, Epimmune Inc., 5820 Nancy Ridge, Suite 100, San Diego, CA 92121.
3 Abbreviations used in this paper: HBV, hepatitis B virus; HCV, hepatitis C virus; Patr, Pan troglodytes class I; Papa, Pan paniscus class I; BCL, B cell line.
Received for publication April 24, 2000. Accepted for publication July 19, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. J. Lavender, H. H. Chau, and K. P. Kane Distinctive Interactions at Multiple Site 2 Subsites by Allele-Specific Rat and Mouse Ly49 Determine Functional Binding and Class I MHC Specificity J. Immunol., November 15, 2007; 179(10): 6856 - 6866. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. J. Lavender, B. J. Ma, E. T. Silver, and K. P. Kane The Rat RT1-A1c MHC Molecule Is a Xenogeneic Ligand Recognized by the Mouse Activating Ly-49W and Inhibitory Ly-49G Receptors J. Immunol., March 15, 2004; 172(6): 3518 - 3526. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. H. Shoukry, J. Sidney, A. Sette, and C. M. Walker Conserved Hierarchy of Helper T Cell Responses in a Chimpanzee during Primary and Secondary Hepatitis C Virus Infections J. Immunol., January 1, 2004; 172(1): 483 - 492. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Sidney, S. Southwood, V. Pasquetto, and A. Sette Simultaneous Prediction of Binding Capacity for Multiple Molecules of the HLA B44 Supertype J. Immunol., December 1, 2003; 171(11): 5964 - 5974. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. C. McGuire, S. R. Leib, R. H. Mealey, D. G. Fraser, and D. J. Prieur Presentation and Binding Affinity of Equine Infectious Anemia Virus CTL Envelope and Matrix Protein Epitopes by an Expressed Equine Classical MHC Class I Molecule J. Immunol., August 15, 2003; 171(4): 1984 - 1993. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Mizukoshi, M. Nascimbeni, J. B. Blaustein, K. Mihalik, C. M. Rice, T. J. Liang, S. M. Feinstone, and B. Rehermann Molecular and Immunological Significance of Chimpanzee Major Histocompatibility Complex Haplotypes for Hepatitis C Virus Immune Response and Vaccination Studies J. Virol., May 13, 2002; 76(12): 6093 - 6103. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), Hu J chain 102–110
(
), Hu J chain 5–13 (
), and A24Con1 (
) were incubated with
purified preparations of Patr-A*0101 (A), -B*0901
(B), -B*1301 (C), or -A*0701
(D), as described.
