| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Microbiology and Immunology and
William K. Warren Medical Research Institute, University of Oklahoma Health Sciences Center, Oklahoma City, OK 73190
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
Lymphocytes of the T lineage are restricted to seeing peptide Ags strictly in the context of MHC molecules (8). Therefore, triggering a cytotoxic T cell response requires that an Ag first be broken into peptide fragments, intracellularly bound by class I MHC molecules, and subsequently presented to T lymphocytes at the cell surface (9). Because CTL play a critical role in eliminating both infected and neoplastic cells, intensive research efforts have been directed toward stimulating T cells capable of eliminating viruses and tumors (10, 11, 12, 13, 14, 15, 16). To elicit a protective CTL response, one must determine which of the class I-presented peptide epitopes are unique to naturally infected or transformed cells. Once the class I epitopes that distinguish diseased cells are identified, they can be therapeutically applied to elicit protective CTL targeted to these specific epitopes.
The search for epitopes unique to infected/neoplastic cells typically proceeds via generalized, often indirect methods. For example, to begin one typically Edman sequences a pool of class I-eluted ligands containing thousands of individual peptides. The resulting sequence produces a pooled "motif" that summarizes the population of individual peptide epitopes with a consensus amino acid sequence and length (17). Such pooled sequence data can then be used to query a pathogen or tumor-specific protein sequence, from which putative class I-presented vaccine epitopes coincident with given motifs can be selected (12, 18, 19). Motifs derived from pooled sequencing of human class I-eluted peptides are invariably 9 aa in length. The overwhelming majority of these peptide motifs also define dominant or strong "anchors" at position 2 (P2) and P9 (20). One interpretation of this motif data is that endogenous peptide loading requires a nonamer with particular P2 and P9 anchors. Indeed, most searches for putative viral or tumor class I-presented epitopes are predicated upon nonameric templates with appropriate P2 and P9 anchors.
The fractionation of peptides before Edman analysis in a previous report resulted in amino acid sequence data demonstrating that the components of a peptide pool can vary considerably from the overall motif (21). Several studies characterizing individual immunogenic epitopes have also demonstrated that CTL recognize peptide ligands inconsistent with the length or P2/P9 anchors indicated by pooled motifs (22, 23). Because a motif is actually a composite representing thousands of individual constituents, it is not surprising that a fraction of these constituents do not match the composite. However, the extent of deviation around composite motifs remains unrealized, and with accumulating data demonstrating that class I peptide ligands need not fit predicted pooled motifs, the search parameters for identifying putative class I ligands become ambiguous.
An ongoing effort in our laboratory is the systematic characterization of endogenously processed class I peptide ligands. Our approach is to map multiple ligands from divergent class I molecules, after which we directly compare the peptide maps (24). The overall goal of comparatively analyzing peptides from different MHC molecules is to prescribe the rules that allow or prohibit natural peptide binding across class I Ag binding groove polymorphisms. We have sequenced >200 endogenously processed peptides from a series of four HLA-B15 alleles that differ sequentially in their Ag binding grooves. Analysis of these HLA-B15 ligands depicts the range of functional variability around a peptide motif and demonstrates the key role played by a ligand’s C terminus during endogenous peptide loading.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
Transfectants were established and cultured as previously
described with some modifications (24). Briefly, the cloned coding
regions of the four alleles B*1501, B*1503, B*1508, and B*1510, which
differ by up to 7 aa within their Ag binding grooves (Table I
), were truncated by RT-PCR using
primers 5PXI
(5'-GGGCTCTAGAGGACTCAGAATCTCCCCAGACGCCGAG-3') and 3PEI
(5'-CCGCGAATTCTCATCTCAGGGTGAG-3'). This
yielded products that terminated at codon 300 (inserted stop shown in
bold italic) and were flanked by 5' XbaI and 3'
EcoRI restriction sites (underlined). The products were
subcloned into the pcDNA3.1(-) expression vector (Invitrogen,
Carlsbad, CA), verified by DNA sequencing, and electroporated into the
class I-negative cell line 721.221 (25), which was then selected with
1.5 mg/ml G418 (Mediatech, Herndon, VA). Transfectants were screened by
ELISA using W6/32 (26) as the primary Ab and an
anti-ß2m HRP-conjugated secondary Ab (Dako,
Carpinteria, CA), and positive wells were subcloned by limiting
dilution. sHLA-producing clones were then expanded and cultured for
4–6 wk in a Cell Pharm 3000 (Unisyn Technologies, San Diego, CA)
hollow-fiber bioreactor system following inoculation of 4 x
109 viable cells. Basal media was unsupplemented RPMI 1640
with 2 mM L-glutamine (Mediatech), while the extracapillary
space (ECS) feed consisted of RPMI 1640 with L-glutamine
and 10% FCS. Harvests collected from the ECS throughout this time span
averaged 18 L at 6–8 mg/L sHLA; they were centrifuged to remove cell
debris and then passed at 5 ml/min over a 100-ml bed volume column of
W6/32 coupled to cyanogen bromide-activated Sepharose 4B (Amersham
Pharmacia Biotech, Piscataway, NJ) using the GradiFrac System (Amersham
Pharmacia Biotech); separate matrices were prepared and used for each
sHLA type. After washing the column with 1–2 L of 20 mM sodium
phosphate buffer, complexes were eluted with 200 ml of 0.2 N acetic
acid, pH 2.7, at 6.5 ml/min with 4.5 ml fractions automatically
collected.
|
Peptides were extracted from eluted sHLA by pooling the fractions of UV absorbance at 280 nm (A280), adding glacial acetic acid to 10%, and heating for 10 min at 70°C. Upon cooling, peptides were separated from heavy chains, ß2m, and BSA through a stirred cell with a 3-kDa YM membrane (Amicon, Beverly, MA). Isolated peptides were purified of salts and free amino acids by a steep RP-HPLC gradient of 0–100% acetonitrile with trifluoroacetic acid (TFA) (0.06%) in 1 min, held at 100% for 10 min before return to 0% in 1 min on a 2.1 x 100 mm C18 column (Vydac, Hesperia, CA), during which the region corresponding to A214 was manually collected. While 1/100th of this was used for pooled Edman sequencing, aliquots approximating 400 µg of extracted peptides as based upon ELISA estimates of 2% (24) and an assumed handling loss of up to 50% (27) were next injected in a volume of 100 µl of 10% acetic acid and separated by RP-HPLC on a 2.1 x 150 mm C18 column (Michrom Bioresources, Auburn, CA) using a gradient of 2–10% acetonitrile with 0.06% TFA in 0.02 min and 10–60% of the same buffer in 60 min at 180 µl/min with fractions collected automatically every minute. The dye methyl violet base B (Aldrich, Milwaukee, WI) was added in a quantity of 500 ng to the 100-µl injection aliquots before this final RP-HPLC separation to control for gradient consistency between runs. Of the resulting fractions, 1/9th of each was removed and stored at -20°C for later Edman sequencing as desired; the remaining 160 µl of each was concentrated by speed-vac to remove TFA and resuspended in 25 µl of 0.1% acetic acid in 50% methanol for nanoelectrospray mass spectrometry (NanoES-MS). Edman degradation, which was performed on peptide pools to obtain motifs and on fractions accordingly (21), was conducted for 14 cycles on a 492A pulsed liquid phase protein sequencer (Perkin-Elmer Applied Biosystems Division, Norwalk, CT) without cysteine derivatization, and raw data analysis was performed as previously described (28).
NanoES-MS ion map generation, comparison, and nanoelectrospray tandem mass spectrometry (NanoES-MS/MS) of potential ligand matches
To generate ion maps, 1–2 µl from each RP-HPLC was loaded into a gold/palladium alloy-coated borosilicate glass NanoES capillary tip (Protana, Odense, Denmark) by gentle centrifugation and subjected to NanoES-MS on an API III triple quadrupole mass spectrometer (PE SCIEX, Foster City, CA) equipped with a collision cell upgrade (29) and a NanoES ionization source inlet (Protana). Approximately 20–30 scans were collected with the following instrument settings: polarity, positive; needle voltage, 1000 V; orifice voltage, 65 V; N2 curtain gas, 0.6 ml/min; step size, 0.2 atomic mass unit (amu); dwell time, 1.5 ms; and mass range, 325-1400 m/z. Spectra generated from the total ion chromatograms of each fraction were then compared with one another following baseline subtraction (spectral window width, 32 amu) and Kalman smoothing (Kalman gain, 2). Ion matches between spectra were located through data centroiding (minimum distance, 0.3 amu; minimum width, 0.5 amu) and/or direct visual assessment, the latter proving beneficial in confirming matches of comparative intensity extremes. The states of identified matches were then evaluated by collisional dissociation of candidate ions via NanoES-MS/MS. Typical NanoES-MS/MS runs involved gating for an ion with the first quadrupole and scanning a range with the third quadrupole of 30–1000 m/z using a step size of 0.3 amu and a dwell time of 2 ms with lysine underivatized; the collision gas (Ar) was adjusted in each case to optimize fragmentation for the ion examined. NanoES-MS/MS data was evaluated and interpreted using the Predict Sequence algorithm (BioMultiView software, PE SCIEX) as well as PeptideSearch 3.0.2 (30) in instances of low ligand ionization/concentration or poor fragmentation. Advanced BLAST searches (31) were performed against databases available through the National Center for Biotechnology Information (National Institutes of Health, Bethesda, MD) web server to identify homology with currently catalogued sequences.
|
Results and Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
a. Selected ions were then
dissociated by NanoES-MS/MS, and the resulting fragment information was
compared and interpreted (Fig. 1, b and c) to
determine whether the ligands were sequence-identical matches.
|
Individual peptide ligands characterized from the four B15
allotypes are shown in Fig. 2. The number
of ligands for which either complete or partial sequences were obtained
here was as follows: B*1508, n = 80; B*1501,
n = 101; B*1503, n = 36; and B*1510,
n = 35. The pooled motifs of peptides extracted
respectively from B*1508, B*1501, B*1503, and B*1510 describe nonamers
with various P2 and P9 dominant anchors and P3 auxiliary anchor
preferences (Fig. 2), while individual ligand sequences range from 7 to
12 aa in length and demonstrate greater sequence heterogeneity at P2
than at their C-terminal anchors (Figs. 2 and 3). For allotypes B*1508, B*1501, and
B*1503, a dominant C terminus was especially prominent, while
B*1510 exhibited a P2 anchor nearly as strong as its C-terminal anchor
(Fig. 3b). The C-terminal anchor was thus dominant among
ligands bound by all four of the B15 allotypes characterized here.
|
|
a).
Coupling this length variability with P2 and C-terminal sequence
flexibility leaves 
29% of the endogenously loaded peptides
characterized in Fig. 2 as "ideal" nonamers with both P2 and P9
anchors in concordance with their pooled Edman motifs. Overlapping peptide ligands exhibit conserved C termini
Through the comparative mapping and sequencing approach outlined
(Fig. 1), 19 peptide overlaps across B*1508 and B*1501 were defined.
Similarly, seven ligands overlapping across B*1501 and B*1503 were
identified, and four ligands were found to overlap across B*1508,
B*1503, and B*1501. A conservative estimate is that our ion maps for
each of the B15 allotypes represent 2000 individual peptides per
molecule, yet B*1510 was not observed to share overlaps with the other
three B15 allotypes in our ion maps (Fig. 4).
|
depicts, in the context of
the class I peptide binding cleft, the locations of polymorphisms that
individuate the four allotypes and highlights the anchoring residues
for the peptide overlaps as indicated by their presenting molecule’s
respective motif (Fig. 2). Bolding the amino acids of these overlapping
ligands, which are in agreement with the motifs, underscores the trend
whereby a C-terminal anchor sequence is conserved in overlaps while the
N-proximal anchor is more flexible in its location and/or sequence. A
lack of overlaps with B*1510 could be explained by the S
Y
substitution at P116 (Table I). P116 participates in the C, E, and F Ag
binding groove specificity pockets (41) and leaves valine (V) and
leucine/isoleucine (X) as the preferred C-terminal anchors among most
B*1510 ligands characterized (Fig. 2). Thus, the conserved C-terminal
anchors that facilitate the occurrence of B*1508, B*1501, and B*1503
overlaps may fail to preferentially interact with the B*1510 C-terminal
specificity pockets.
Length variation among overlapping ligands tends to mimic that observed
among the total ligands characterized, previously discussed from Fig. 3. Notable exceptions were that overlaps between B*1501 and B*1503 were
primarily octamers and that only one overlapping ligand was longer than
10 aa in length (Fig. 4). However, a limited number of B*1503 overlaps
identified to date makes trends difficult to clearly establish
regarding this B15 allotype.
Distilling the data from the 30 overlapping ligands so far identified
leads us to suggest a model whereby the majority of ligands are first
anchored into the binding groove by their C termini. In order for a
peptide to stably remain in the groove, we propose that following rigid
initial anchoring at the C terminus a ligand might be further tethered
into the class I Ag binding cleft at a more elastic N-proximal
position. Such could argue the case for peptide ligand NQZHGSAEY, a
nonamer that overlaps across B*1508, B*1501, and B*1503 (Fig. 1).
According to this model (Fig. 5), a
C-terminal tyrosine anchors NQZHGSAEY into all three B15 allotypes,
while a glutamine at P2 anchors the peptide into B*1501 and B*1503 and
a glutamine/lysine (Z; most likely a lysine based upon motif
assignments and fractional Edman sequencing data) at P3 provides
additional anchoring for B*1501 and serves as the sole N-proximal
anchor for B*1508. Therefore, length and N-proximal specificity would
appear to play more secondary roles in endogenous ligand binding of B15
peptide epitopes.
|
Edman-derived motifs for the four B15 allotypes shown in Fig. 2
illustrate P2 and P9 dominant anchors and P3 auxiliary anchors, but
these motifs fail to capture trends for auxiliary anchoring at the
C-proximal regions of endogenously bound ligands. Specifically,
additional anchor preferences occur at both P7 and P8 in the case of
nonamers, octamers, and heptamers, while in the cases of ligands longer
than nonamers the two residues internal to the C terminus, which we
refer to as C-1 and C-2, also act as
auxiliary anchors. A glance through the individual B15 ligands shows
that the amino acids threonine (T), serine (S), valine (V), and
glutamine/lysine (Z) predominate at the C-proximal ends (involving
either P7/P8 or C-1/C-2) of these ligands
(Fig. 2). An example of how C-proximal auxiliary anchors impact
endogenous ligand binding is that more than half of the peptides that
overlap across the B*1508 and B*1501 Ag binding grooves bear threonine
at P7 or P8, and in four cases the peptides that bind B*1508/1501 or
B*1508/1501/1503 are heptamers with threonine occupying P7, their
C-terminal positions (Fig. 4). The role of threonine as a
C-proximal/C-terminal auxiliary anchor is best illustrated with
the B*1508/1501/1503 overlapping heptamer CPLSCFT, where threonine
provides a C-terminal anchor not evident in the pooled motif.
Sources of individual B15 ligands reflect cytosolic Ag processing
Examples of peptides from this study with complete homology to
stretches of known proteins are shown in Table II. The majority of ligands are derived
from intracellular proteins, which illustrates the endogenous pathway
is generating the class I-loaded peptides described here (2). Of the
peptide sequences listed, six have also been reported by other
laboratories. It is noteworthy that overlaps across other molecules of
the HLA-B15 group are evident within this data collection: the B*1510
tapasin354–362 ligand HHSDGSVSL has been sequenced
from B*1509 extracts (36), and the B*1508/1501 ubiquitin-protein
ligase83–91-derived ligand ILGPPGSVY was characterized
from endogenously bound B*1502 peptides (36). We identified the
eIF3-p6661–69 nonamer SQFGGGSQY (42, 43) within
B*1508, B*1501, and B*1503 extracts. The decamer YMIDPSGVSY, which is
homologous to proteasome subunit C8150–159, was also
previously described as a ligand for B*1502 (36), B*1508 (36), and
B*4601 (42); we found it presented by B*1501 and by B*1508. Therefore,
overlapping ligands identified here overlap in Ag presentation with the
HLA-B15 allotypes characterized by others, and these overlaps are
likewise determined by both flexible N-proximal anchor positions and
conserved C-terminal anchor/C-proximal auxiliary anchor
preferences.
|
The primary impetus for characterizing class I peptide ligands is to understand the functional impact of class I polymorphism in terms of endogenous ligand presentation. A fundamental realization of how naturally processed peptides bind both individual and multiple class I molecules can then be translated into peptide-based therapies to elicit protective CTL. Therefore, an accurate interpretation of peptide sequence data, either individual or pooled, will in turn further the selection of optimal viral and tumor-associated ligands for therapeutic applications.
Our examination of numerous B15 ligands enhances understanding the rules that govern class I peptide presentation. Building upon the pooled peptide motif, our data indicates that queries for potential epitopes specific to these allotypes might be optimized in three ways. First, although nonamers represent half the ligand population, peptide epitopes can range in length from 7 to 12 aa. Second, effective N-proximal anchors need not be positioned strictly at P2. Third, searches for ligands should most heavily weight C-terminal sequence matches.
In summary, divergent HLA-B15 molecules appear to present similar peptide ligands as long as polymorphisms do not alter C-terminal anchoring pockets and while an N-proximal ligand residue can be subsequently anchored within the Ag binding groove. Supporting data suggests that these principles might extend beyond the four B15 allotypes studied here. Specifically, unpublished results by Ghosh and Wiley (noted in 44) indicate that an octamer has been observed to successfully bind a class I molecule by its C terminus despite being shown through x-ray crystallography to not even reach the N-terminal pocket of the binding groove. In addition, a recently described HIV-gag197–205 CTL epitope presented by murine class I Kd fails to show a motif-prescribed tyrosine at P2 and instead associates stably through its conserved C terminus and an N-proximal preference for glutamine at P3 (23). It will be interesting to see through future studies whether endogenous ligands eluted from other HLA-B, as well as HLA-A and HLA-C, molecules confirm this speculation by exhibiting conserved C termini along with length and N-proximal sequence versatility.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. William Hildebrand, Department of Microbiology and Immunology, 975 NE 10th St, Biomedical Research Center 317, University of Oklahoma Health Sciences Center, Oklahoma City, OK 73104. E-mail address:
3 Abbreviations used in this paper: ß2m, ß2-microglobulin; Pn, position n (where n denotes a ligand residue number given within the text); sHLA, soluble HLA; ECS, extracapillary space; RP-HPLC, reverse-phase HPLC; TFA, trifluoroacetic acid; NanoES, nano-electrospray; NanoES-MS, nano-electrospray mass spectrometry; NanoES-MS/MS, nano-electrospray tandem mass spectrometry; amu, atomic mass unit.
Received for publication November 18, 1998. Accepted for publication April 1, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
1 helix of the HLA-B heavy chain can have an overriding influence on peptide-binding specificity. J. Immunol. 158:1660.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  E. Kim, H. Kwak, and K. Ahn Cytosolic Aminopeptidases Influence MHC Class I-Mediated Antigen Presentation in an Allele-Dependent Manner J. Immunol., December 1, 2009; 183(11): 7379 - 7387. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. R. Manuel, W. W. Yeh, M. S. Seaman, K. Furr, M. A. Lifton, S. L. Hulot, P. Autissier, and N. L. Letvin Dominant CD8+ T-Lymphocyte Responses Suppress Expansion of Vaccine-Elicited Subdominant T Lymphocytes in Rhesus Monkeys Challenged with Pathogenic Simian-Human Immunodeficiency Virus J. Virol., October 1, 2009; 83(19): 10028 - 10035. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Escobar, D. K. Crockett, E. Reyes-Vargas, A. Baena, A. L. Rockwood, P. E. Jensen, and J. C. Delgado Large Scale Mass Spectrometric Profiling of Peptides Eluted from HLA Molecules Reveals N-Terminal-Extended Peptide Motifs J. Immunol., October 1, 2008; 181(7): 4874 - 4882. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Epitope discovery in West Nile virus infection: Identification and immune recognition of viral epitopes PNAS, February 26, 2008; 105(8): 2981 - 2986.
|
|
|
|
|
|
J. T. Loffredo, J. Sidney, S. Piaskowski, A. Szymanski, J. Furlott, R. Rudersdorf, J. Reed, B. Peters, H. D. Hickman-Miller, W. Bardet, et al. The High Frequency Indian Rhesus Macaque MHC Class I Molecule, Mamu-B*01, Does Not Appear to Be Involved in CD8+ T Lymphocyte Responses to SIVmac239 J. Immunol., November 1, 2005; 175(9): 5986 - 5997. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Frahm, S. Adams, P. Kiepiela, C. H. Linde, H. S. Hewitt, M. Lichterfeld, K. Sango, N. V. Brown, E. Pae, A. G. Wurcel, et al. HLA-B63 Presents HLA-B57/B58-Restricted Cytotoxic T-Lymphocyte Epitopes and Is Associated with Low Human Immunodeficiency Virus Load J. Virol., August 15, 2005; 79(16): 10218 - 10225. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. W. Purcell and J. J. Gorman Immunoproteomics: Mass Spectrometry-based Methods to Study the Targets of the Immune Response Mol. Cell. Proteomics, March 1, 2004; 3(3): 193 - 208. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. R. Turnquist, J. L. Petersen, S. E. Vargas, M. M. McIlhaney, E. Bedows, W. E. Mayer, A. G. Grandea III, L. Van Kaer, and J. C. Solheim The Ig-Like Domain of Tapasin Influences Intermolecular Interactions J. Immunol., March 1, 2004; 172(5): 2976 - 2984. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. A. Macdonald, A. W. Purcell, N. A. Mifsud, L. K. Ely, D. S. Williams, L. Chang, J. J. Gorman, C. S. Clements, L. Kjer-Nielsen, D. M. Koelle, et al. A Naturally Selected Dimorphism within the HLA-B44 Supertype Alters Class I Structure, Peptide Repertoire, and T Cell Recognition J. Exp. Med., September 2, 2003; 198(5): 679 - 691. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. D. Hickman, A. D. Luis, W. Bardet, R. Buchli, C. L. Battson, M. H. Shearer, K. W. Jackson, R. C. Kennedy, and W. H. Hildebrand Cutting Edge: Class I Presentation of Host Peptides Following HIV Infection J. Immunol., July 1, 2003; 171(1): 22 - 26. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Admon, E. Barnea, and T. Ziv Tumor Antigens and Proteomics from the Point of View of the Major Histocompatibility Complex Peptides Mol. Cell. Proteomics, June 1, 2003; 2(6): 388 - 398. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. R. Turnquist, S. E. Vargas, A. J. Reber, M. M. McIlhaney, S. Li, P. Wang, S. D. Sanderson, B. Gubler, P. van Endert, and J. C. Solheim A Region of Tapasin That Affects Ld Binding and Assembly J. Immunol., October 15, 2001; 167(8): 4443 - 4449. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Novitsky, N. Rybak, M. F. McLane, P. Gilbert, P. Chigwedere, I. Klein, S. Gaolekwe, S. Y. Chang, T. Peter, I. Thior, et al. Identification of Human Immunodeficiency Virus Type 1 Subtype C Gag-, Tat-, Rev-, and Nef-Specific Elispot-Based Cytotoxic T-Lymphocyte Responses for AIDS Vaccine Design J. Virol., October 1, 2001; 75(19): 9210 - 9228. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Urvater, H. Hickman, J. L. Dzuris, K. Prilliman, T. M. Allen, K. J. Schwartz, D. Lorentzen, C. Shufflebotham, E. J. Collins, D. L. Neiffer, et al. Gorillas with Spondyloarthropathies Express an MHC Class I Molecule with Only Limited Sequence Similarity to HLA-B27 that Binds Peptides with Arginine at P2 J. Immunol., March 1, 2001; 166(5): 3334 - 3344. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
