This information is current as
of June 17, 2017.
Limited Diversity of Peptides Related to an
Alloreactive T Cell Epitope in the
HLA-B27-Bound Peptide Repertoire Results
from Restrictions at Multiple Steps Along the
Processing-Loading Pathway
Alberto Paradela, Iñaki Alvarez, Marina García-Peydró,
Laura Sesma, Manuel Ramos, Jesús Vázquez and José A.
López de Castro
References
Subscription
Permissions
Email Alerts
This article cites 40 articles, 22 of which you can access for free at:
http://www.jimmunol.org/content/164/1/329.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2000 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
J Immunol 2000; 164:329-337; ;
doi: 10.4049/jimmunol.164.1.329
http://www.jimmunol.org/content/164/1/329
Limited Diversity of Peptides Related to an Alloreactive T Cell
Epitope in the HLA-B27-Bound Peptide Repertoire Results
from Restrictions at Multiple Steps Along the
Processing-Loading Pathway1
Alberto Paradela, Iñaki Alvarez, Marina Garcı́a-Peydró, Laura Sesma, Manuel Ramos,
Jesús Vázquez, and José A. López de Castro2
M
ajor histocompatibility complex class I molecules constitutively bind and present at the cell surface a large
repertoire of endogenous peptides which can be recognized by CD81 T cells. Class I-bound peptides generally range
in size from 8- to 11-amino acid residues, with a majority of
nonamers. Several factors, including endogenous Ag processing,
peptide transport to the endoplasmic reticulum (ER)3, assisted
loading, peptide affinity, and stability of the MHC-peptide complex determine the nature of the peptides bound to a given class I
molecule (1). However, the precise contribution of these factors in
shaping class I-bound peptide repertoires is insufficiently known.
Proteasomes are multicatalytic protease complexes that mediate
most of the protein degradation in the cytosol. They are the main
Centro de Biologı́a Molecular Severo Ochoa, Universidad Autónoma de Madrid,
Facultad de Ciencias, Madrid, Spain
Received for publication June 25, 1999. Accepted for publication October 19, 1999.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work was supported by Grant SAF97/0182 from the Plan Nacional de I1D and
by a grant from the Comunidad Autónoma de Madrid. We thank the Fundación
Ramón Areces for an institutional grant to the Centro le Biologı́a Molecular “Severo
Ochoa”
2
Address correspondence and reprint requests to Dr. José A. López de Castro, Centro
de Biologı́a Molecular Severo Ochoa, Universidad Autónoma de Madrid, Facultad de
Ciencias, Cantoblanco, 28049 Madrid, Spain. E-mail address: aldecastro@cbm.
uam.es
3
Abbreviations used in this paper: ER, endoplasmic reticulum; TFA, trifluoroacetic
acid; C1R, Hmy2.C1R; MS, mass spectrometry; MALDI-TOF, matrix-assisted laser
desorption/ionization time of flight; MLF, mean linear fluorescence; LLnL, N-acetylL-leucinyl-L-leucinyl-L-norleucinal; LCT, lactacystin; DT50, decay time at 50% of the
maximal fluorescence.
Copyright © 2000 by The American Association of Immunologists
source of HLA class I-bound peptides and can directly generate
natural MHC ligands in vitro. However, the contribution of additional proteases from the cytosol or other cell compartments is
suggested by evidence that proteasome-specific inhibitors do not
totally block class I expression or generation of particular class I
ligands (2, 3). Recently, a subtilisin-like protease with the potential
to substitute for some of the proteasome function has been reported
(4). In addition, peptide trimming could occur in the cytoplasm and
ER, generating peptides not produced directly by the proteasome
(5–10). Furin, a protease in the trans-Golgi network may contribute to the production of class I-restricted viral Ags (11). However,
the actual contribution of nonproteasomal peptidases to shaping
class I-bound peptide repertoires is not known.
Peptides generated in the cytosol are transported into the ER by
the TAP transporter, an heterodimeric protein located in the membrane of the ER. Human TAP deficiency causes a severe drop in
cell surface expression of class I proteins, indicating that transport
of most peptides is TAP dependent (12, 13). Human TAP is less
restrictive for peptides than rat and murine counterparts but may
exhibit large peptide-binding differences, as revealed by combinatorial peptide libraries (14, 15).
Once into the ER, peptide binding to class I molecules occurs
in a process that requires the cooperative activity of at least four
chaperones: calnexin, calreticulin, Erp57, and tapasin. The first
three proteins have a role in assembling and correct folding of the
class I protein and are not known to introduce sequence-dependent
restrictions to peptide loading. Tapasin directly mediates the association of TAP and the class I molecule during peptide loading (16,
17). Allelic variation in MHC class I molecules influences their dependence on tapasin for peptide loading and Ag presentation. In
0022-1767/00/$02.00
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
The influence of various factors along the processing-loading pathway in limiting the diversity of HLA-B27-bound peptides around
a core protein sequence was analyzed. The C5 proteasome subunit-derived RRFFPYYV and RRFFPYYVY peptides are natural
B*2705 ligands. The octamer is an allospecific CTL epitope. Digestion of a 27-mer fragment of C5 revealed that both ligands are
generated from this precursor substrate with the 20S proteasome in vitro in a ratio comparable to that in the B*2705-bound
peptide pool. The C5 sequence allowed to derive a nested set of six additional peptides with 8 –11 residues containing the core
octamer sequence and the Arg2 motif of HLA-B27, none of which was found in the B27-bound pool. Together, low proteasomal
yield, disfavored TAP-binding motifs, and low affinity for B*2705 accounted for the absence of four of the six peptides. The two
remaining differed from the natural octamer or nonamer ligands only by an additional N-terminal Ser residue. Their stability in
complex with B*2705 was lower than the respective natural ligands, raising the possibility that N-terminal trimming might have
favored a shift toward the more stable peptides. The results suggest that the B*2705-bound peptide repertoire has a highly
restricted diversity around a core alloantigenic sequence. This is not explained by a single bottleneck feature, but by multiple
factors, including proteasomal generation, TAP-binding motifs, MHC-binding efficiency, and perhaps optimized stability through
N-terminal trimming. Tapasin-dependent restrictions, although not excluded, were not required to explain the absence in vivo of
the particular peptide set in this study. The Journal of Immunology, 2000, 164: 329 –337.
330
FACTORS LIMITING THE HLA-B27-BOUND PEPTIDE REPERTOIRE
Materials and Methods
Cell lines
The anti-B*2705 alloreactive CTL 27S69 clone and its culture conditions
have been described (20, 21). Hmy2.C1R (C1R) is a human lymphoid cell
line with low expression of its endogenous class I Ags. The transfectant
expressing B*2705 was cultured in DMEM (Life Technologies, Paisley,
U.K.) with 7.5% heat-inactivated FCS. T2 is a TAP-deficient human cell
line of lymphoid origin (12). The B*2705-T2 transfectant was a kind gift
from Dr. David Yu (University of California, Los Angeles, CA). It was
cultured in DMEM supplemented with 5% FCS. RMA-S is a TAP-deficient
murine cell line (22, 23).The B*2705-RMA-S transfectant was cultured in
RPMI 1640 supplemented with 10% FCS. When cultured at 26°C, T2 and
RMA-S transfectants express class I molecules presumably devoid of peptides. These molecules are unstable at 37°C, but their surface expression at
this temperature can be stabilized by exogenous peptide ligands.
Isolation of HLA-B27-bound peptides
This was done as described previously (20), with minor modifications.
Briefly, 1–1.5 3 1010 B*2705-C1R cells were lysed at 4°C in 20 mM
Tris/HCl buffer, 150 mM NaCl, and 1% Nonidet P-40 (pH 7.5) with a
mixture of protease inhibitors. Cell lysates were subjected to affinity chromatography using the W6/32 mAb (IgG2a, specific for a monomorphic
HLA-A, HLA-B, and HLA-C determinant; see Ref. 24). HLA-B27-bound
peptides were eluted from the column with 0.1% trifluoroacetic acid (TFA)
in water at room temperature, filtered through Centricon 3 (Amicon, Beverly, MA), and concentrated to 100 ml for HPLC fractionation. This was
conducted in a Waters alliance system (Waters, Milford, MA), using a
Vydac C18 (0.21 3 25 cm) 5-mm particle size column (Vydac, Hesperia,
CA), at a flow rate of 100 ml/min, as follows: isocratic conditions with
buffer A (0.08% TFA in water) for 15 min, followed by a linear gradient
of 0 – 44% buffer B (80% acetonitrile and 0.075% TFA in water) for 90
min, and a linear gradient of 44 –100% buffer B for another 35 min. Peptide
fractionation was simultaneously monitored at 210 and 280 nm. Fractions
of 50 ml were collected and stored at 220°C.
Mass spectrometry (MS)
The peptide composition of individual HPLC fractions was determined by
matrix-assisted laser desorption/ionization time of flight (MALDI/TOF)
MS. A calibrated Kompact Probe instrument (Kratos-Shimadzu, Manchester, U.K.) operating in the positive linear mode was used. Dried fractions
were resuspended in 5 ml methanol/water (1:1) containing 0.1% formic
acid, and 0.5 ml was applied onto target and dried out. A total of 0.5 ml of
saturated a-ciano-4-hydroxycinnamic acid matrix in water:acetonitrile
(1:1) containing 0.1% TFA was then added and dried out. Sometimes, 1 ml
of these samples was subjected to peptide sequencing in a LCQ electrospray/ion trap mass spectrometer (Finnigan Thermoquest, San Jose, CA),
exactly as previously described (20).
Peptide synthesis and purification
Peptides were synthesized using standard F-moc chemistry and purified by
HPLC to a purity .95%. The correct composition and molecular mass of
purified peptides were confirmed by amino acid analysis using a 6300
amino acid analyzer (Beckman Coulter, Palo Alto, CA), which also allowed their quantification, and by MALDI-TOF and electrospray
ion/trap MS.
Epitope stabilization assays
The quantitative epitope stabilization assay used to measure binding to
B*2705 has been described (25). Briefly, B*2705-RMA-S cells were incubated for 24 h at 26°C. They were then incubated without FCS for 1 h
at 26°C, with 1024–1029 M peptide, transferred to 37°C for 4 h, and
collected for flow microcytometry analysis with the ME1 mAb (IgG1, specific for HLA-B27, HLA-B7, HLA-B22; see Ref. 26). Binding was expressed as the C50, which is the molar concentration of the peptide at 50%
of the maximum fluorescence obtained with that peptide in the concentration range used. When multiple peptides were compared, the C50 value of
a reference peptide was first obtained. Binding of other peptides was expressed as EC50, which is the molar concentration of a given peptide required to obtain the fluorescence value at the C50 of the reference peptide.
EC50 values were calculated as described previously (25). Peptides with
EC50 values ,10 mM were considered to bind with high affinity. EC50
values between 10 and 50 mM were considered to reflect intermediate
affinity, and EC50 . 50 mM indicated low affinity.
A previously described (27) cell surface MHC-peptide complex stability
assay was used. Briefly, T2-B*2705 cells (105 cells/well) were incubated
overnight at 37°C, in serum-free cell culture medium (Serotec, Oxford,
U.K.), in the presence of 100 mM peptide and 100 nM b2-microglobulin.
After intensive washing, cells were incubated for 1 h at 37°C in RPMI 1640
containing 10% FCS and brefeldin A (10 mg/ml) to block egress of newly
synthesized class I molecules. Cells were washed and incubation continued
in the presence of 0.5 mg/ml of brefeldin A at 37°C. Cells were removed at
various times and stained with the ME1 mAb as described above. The decay
of B*2705-peptide complexes was determined as follows: % mean linear fluorescence (MLF) remaining 5 [(MLFt(1pep) 2 MLFt(2pep))/(MLFt 5 0(1pep) 2
MLFt 5 0(2pep))].
Inhibition of peptide presentation assay
N-acetyl-L-leucinyl-L-leucinyl-L-norleucinal (LLnL), which reversibly inhibits proteasomes, cysteine proteases, and calpains (28, 29), was purchased from Sigma (St. Louis, MO). Lactacystin (LCT) was purchased
from E. J. Corey (Harvard University, Cambridge, MA) and irreversibly
inhibits proteasomes (30, 31) and cathepsin A (29, 32). About 5 3 105
B*2705-C1R cells were preincubated for 18 –20 h at 37°C in RPMI 1640
medium supplemented with 10% FCS in the presence of LLnL or LCT.
Then cells were labeled for 90 min at 37°C with 50 mCi of 51Cr in the
continuous presence of proteasome inhibitor, washed (four times) with the
same medium plus 4 mg/ml brefeldin A (Sigma), and seeded in 96-well
plates with or without 10 mM of the RRFFPYYV octamer. CTL 27S69 was
added, and incubation at 37°C for 4 h was conducted in the presence of 2
mg/ml brefeldin A. Specific 51Cr in the supernatants was calculated as
follows: (experimental lysis 2 spontaneous lysis)/(maximum release 2
spontaneous lysis) 3100.
Purification of 20S proteasome and digestion assays
About 3 3 109 B*2705-C1R cells were potter lysed in 50 ml of 50 mM
Tris/HCl and 25 mM KCl (pH 8). The homogenate was centrifuged at
1500 3 g for 10 min. The supernatant was further centrifuged at 100,000 3
g for 1 h, and fractionated on a 35-ml DEAE-cellulose DE52 column
(Whatman, Maidstone, U.K.) equilibrated in homogenization buffer. Proteins were eluted with a linear gradient of 0.025– 0.6 M KCl in 50 mM
Tris/HCl buffer (pH 8). Fractions were analyzed by Western blot using an
anti-C2 polyclonal Ab (33). Proteasome-containing fractions were pooled
and concentrated in an 8-ml DEAE-cellulose DE52 column equilibrated in
homogenization buffer. Bound proteins were eluted with 50 mM Tris/HCl
and 300 mM KCl (pH 8). Protein-containing fractions were scanned using
the Bradford method (Bio-Rad, Munich, Germany) and further subjected to
centrifugation at 200,000 3 g for 18 h in a gradient of 10 –30% glycerol in
1 M urea, 50 mM Tris/HCl, and 50 mM KCl (pH 8). Proteasome-containing fractions were identified by 12% SDS-PAGE (34) and further subjected
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
particular, HLA-B*2705 seemed able to efficiently form peptide
complexes in the absence of tapasin (18).
Finally, peptide binding in the ER will be determined by MHC
affinity. Peptides bind to class I molecules through conserved contacts involving the peptidic main chain and the peptide ends. Additional interactions take place between peptidic anchor residues
and side chain pockets in the peptide-binding site of the class I
molecule (19). Class I polymorphism affects mainly residues located in or close to the pockets, modulating in this way the nature
of class I-bound peptide repertoires.
The experiments described in this study arose from the identification of two closely related natural ligands of B*2705: the
RRFFPYYV octamer and the C-terminally extended nonamer
RRFFPYYVY. These peptides derive from the C5 subunit of the
proteasome. The octamer was the specific epitope of an alloreactive CTL clone raised against B*2705, which recognized the
nonamer much less efficiently (20). Starting from these observations, we have now addressed the following issues: 1) Are both
B*2705 ligands directly generated by the proteasome or require
further processing steps? 2) Does the proteasome generate additional peptides from the same region of the parental protein with
the capacity to bind HLA-B27? 3) If so, do they bind B*2705 in
vivo? 4) What factors impair HLA-B27-binding peptides generated by the proteasome from becoming natural ligands? Together
these questions attempt to define the influence of several structural
and functional factors in shaping the B27-bound peptide repertoire.
The Journal of Immunology
331
FIGURE 2. Amino acid sequence of the region of the C5 proteasome
subunit spanning residues 120 –146, which includes the CTL 27S69
epitope sequence (underlined). The natural octamer and nonamer ligands
containing this sequence and the nested set of related peptides with Arg2
and 8 –11 residues in length are indicated.
HPLC fractionation of the proteasomal digestions was conducted in a
Waters 625LC system using the same column and flow rate as for B*2705bound peptides (see above) and the following chromatographic gradient:
isocratic conditions with buffer A (0.1% TFA in water) for 20 min, followed by a linear gradient of 0 – 44% buffer B (80% acetonitrile and 0.1%
TFA in water) for 80 min, and a linear gradient of 44 –100% buffer B for
another 40 min.
Results
The epitope recognized by CTL 27S69 is proteasome dependent
to anion-exchange chromatography in a MonoQ SR5/5 column (Pharmacia, Uppsala, Sweden), at a flow rate of 0.5 ml/min, as follows: isocratic
conditions with buffer A [50 mM Tris/HCl and 50 mM KCl (pH 8)] for 10
min, followed by a linear gradient of 0 –30% buffer B [50 mM Tris/HCl and
0.5 M KCl (pH 8)] for 5 min and a linear gradient of 30 –100% buffer B for
another 30 min. Purity of the fractions was assessed by denaturing SDS-PAGE
as above. Aliquots of purified proteasome were stored at 280°C.
Proteasome digestions were performed at 37°C in 20 mM HEPES buffer
(pH 7.4). The peptide substrate was incubated with 4 3 106 cell-equivalents of
20S proteasome/mg at a final substrate concentration of 250 mg/ml. Digestions
were stopped by adding 1 vol of 0.1% aqueous TFA and stored at 280°C.
In preliminary experiments, we tested whether lysis of B*2705C1R targets could be affected by the proteasome inhibitors LLnL
and the more specific LCT. As shown in Fig. 1, cells previously
treated with LLnL were not lysed by CTL 27S69 upon adding
brefeldin A to prevent egress of newly synthesized molecules. The
effect was not due to brefeldin A, and lysis was restored upon
addition of the synthetic RRFFPYYV peptide. The same result
was obtained with LCT. This indicates that expression of the
RRFFPYYV epitope is critically dependent on proteasome activity.
FIGURE 3. MALDI-TOF MS spectra of unfractionated proteasome digests of the C5 (120 –146) peptide at 4 h (A) and 24 h (B). About 8 mg of peptide
was incubated with 24 3 106 cell-equivalents of 20S proteasome purified from B*2705-C1R cells. The octamer epitope, its related nonamer, the detected
peptides from the J1-J6 set, and the undigested 27-mer are indicated. Peaks at the lowest mass/charge (m/z) values are due to the matrix. A control showing
the absence of proteasome-derived peptide peaks after incubation of the 20S proteasome for 24 h in the absence of the C5 (120 –146) substrate is included
(B, inset).
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
FIGURE 1. Inhibition of CTL 27S69-mediated lysis of B*2705-C1R
target cells by LLnL or LCT. Cells were treated with the indicated concentrations of proteasome inhibitor and subsequently with brefeldin A as
described in Materials and Methods. f, Percent specific lysis of target cells
either untreated, incubated with 4 mg/ml brefeldin A, or brefeldin A plus
proteasome inhibitor. M, Lysis of the same targets in the presence of
brefeldin A, proteasome inhibitor, and 10 mM RRFFPYYV peptide. Specific lysis of untransfected C1R cells at the E:T ratio used (2:1) was 0%.
Data are means of two (for LLnL) or three (for LCT) independent
experiments.
332
FACTORS LIMITING THE HLA-B27-BOUND PEPTIDE REPERTOIRE
The RRFFPYYV epitope and a nested set of related peptides are
directly generated by the 20S proteasome
der the experimental conditions used, a few peptide bonds were
strongly preferred (Fig. 4B).
An examination of the primary structure of the C5 protein around
the core sequence of the CTL 27S69 epitope revealed a nested set
of sequences ranging from 8 to 11 residues with the HLA-B27
peptide motif R2 (Fig. 2). Thus, we asked whether the known
natural ligands RRFFPYYV and its C-terminally extended
nonamer RRFFPYYVY as well as the other peptides of the nested
set (designated J1-J6) were directly generated in vitro by the 20S
proteasome from B*2705-C1R cells. A precursor peptide spanning
residues 120/146 of the C5 protein was digested for 4, 8, and 24 h
at the same enzyme:substrate ratio. Approximately 30, 82, and
97%, respectively, of the substrate was digested at these three
times on the basis of the absorbance of the HPLC peak corresponding to the undigested 27-mer at 210 nm. Unfractionated digests at both 4 and 24 h were analyzed by MALDI-TOF MS (Fig.
3). At both time points, the maximal signal among the digestion
products corresponded to the RRFFPYYVY nonamer, and two
other major peaks corresponded to RRFFPYYV and to J3. J2 and
J5 yielded minor signals at 4 h, the former of which became more
significant at 24 h. Other peptides, unrelated to the nested set, were
also detectable. As a control for the absence of proteasomal autolysis in vitro, the 20S proteasome was incubated for 24 h in the
same conditions, but in the absence of the 27-mer substrate. No peptide peaks were observed upon MALDI-TOF analysis (Fig. 3B).
A more complete screening of the proteasomal digest was conducted through HPLC fractionation of the digest at 8 h and MS
analysis of the corresponding fractions (Fig. 4). This revealed a
very complex digestion pattern and the direct generation of the two
natural HLA-B27 ligands plus J1-J5, but not J6.
An estimate of individual peptide yields in this digestion was
conducted by computing the peak area of the HPLC fractions in
which each peptide eluted. These fractions usually contained peptide mixtures when analyzed by MS. Thus, each peptide was assigned a percentage of its corresponding HPLC fraction area that
was its corresponding percentage of the added intensity of all of
the MALDI-TOF peptide signals in that HPLC fraction. This procedure is not strictly quantitative, since the relationship between
peptide amount and signal intensity in MALDI-TOF spectra is not
the same for all of the peptides and may be influenced by multiple
factors. In addition, it could not be applied to some peptides that
were detected by electrospray, but not by MALDI-TOF MS. However, synthetic J1-J6 were all easily detectable by MALDI-TOF,
and, in control measurements in which peptides coeluting in this
experiment (nonamer/J3; octamer/J2, J4/J5) were mixed at known
amounts, their molar ratio and the ratio of the peak intensities in
MALDI-TOF spectra were similar (data not shown).
A total of 55 molecular species were detected in the HPLCfractionated proteasomal digest (Fig. 4A), 34 of which were recovered with a yield $0.5% (Fig. 4B). The most prominent peptide
was RRFFPYYVY, accounting for 21.4% of the total digest. The
second most abundant peptide was J3 (4.4%), followed by the
RRFFPYYV octamer (2.5%) and J5 (2.3%). J2 and J4 were below
1% (0.5 and 0.8%, respectively), J1 was obtained with very low
yield (0.05%), and J6 was not detected (Table I). The identity of
the octamer, nonamer, and J1-J5 was confirmed by fragmentation
analysis using electrospray MS/MS.
These results indicate that the natural octamer and nonamer ligands are both directly generated by the 20S proteasome in vitro.
They also suggest that proteasome cleavage limits the number of
putative HLA-B27 ligands within a nested set of peptides by failing to generate some of them and generating others in very low
amounts. Thus, although cleavage occurred at many positions un-
Most peptides from the RRFFPYYV-related set bind B*2705 in
vitro
Binding of J1-J6 to B*2705 in vitro was compared with that of the
octamer and nonamer ligands using a peptide stabilization assay
(Fig. 5). The RRFFPYYVY nonamer, J2, J5, and J6 bound with
the highest affinity (EC50, 1–2 mM). J3 and the RRFFPYYV octamer bound well but somewhat less efficiently (EC50, 6 –7 mM).
J1 and J4 bound weakly (EC50, .50 mM). Thus, four of the six
peptides in the J1-J6 set bound B*2705 in vitro similarly as the two
natural ligands.
The J1-J6 peptides are not found among natural B*2705 ligands
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
On the basis of their efficient binding in vitro (Fig. 5), J2, J3, J5,
and J6 had the potential of being natural B*2705 ligands. To
search for their putative presence in the B*2705-bound peptide
pool, we first determined the retention time of the corresponding
synthetic peptides in HPLC. Then the B*2705-bound peptide pool
from B*2705-C1R cells was fractionated in exactly the same chromatographic conditions, and the HPLC fractions at the retention
time of each peptide, as well as neighbor ones, were analyzed by
MALDI-TOF and, in some cases, also by electrospray ion trap MS
(Figs. 6-8). The natural octamer and nonamer ligands provided a
control for the reproducibility of retention times between the synthetic peptides and the peptide pool.
J3 showed almost the same retention time as the natural
nonamer ligand (Fig. 6A). However, whereas the nonamer was
clearly detected in HPLC fractions 194 –195 of the B*2705-bound
peptide pool, no signal corresponding to J3 was obtained in these
fractions or in the following one (Fig. 6B). This was not due to
deficient detection of J3 by MALDI-TOF MS, since in a mixture
of the synthetic J3 and nonamer both peptides were detected with
similar sensitivity by this technique (data not shown).
Similarly, J2 showed the same retention time as the natural octamer
ligand (Fig. 7A) and both peptides generated similarly intense signals
in MALDI-TOF spectra (data not shown). The octamer was detected
in HPLC fractions 189–193 of the B*2705-bound peptide pool corresponding to retention times of 93.5–95.5 min. A prominent peak
similar to the molecular mass of J2 (M1H1: 1235.4 Da) was seen in
HPLC fraction 192 (Fig. 7B) and, in smaller amounts, in adjacent
fractions (data not shown). When subjected to electrospray MS/MS
fragmentation analysis, the fragmentation pattern was unrelated to J2,
as none of the fragmentation ions of this peptide were observed, but
was compatible with RRFVNVVPTF. This peptide was originally
sequenced from B*2702 (35) and subsequently from B*2705 (36). It
has the same molecular mass as J2 and coelutes with the octamer in
the HPLC conditions used (our unpublished observations). Thus, J2
was not found in this analysis in the B*2705-bound peptide pool.
J4 and J5 also showed very similar retention time in HPLC (Fig.
8A) and similar sensitivity by MALDI-TOF (data not shown), but
they were not present in the corresponding HPLC fractions of the
B*2705-bound peptide pool (Fig. 8B). By the same analysis, we
also failed to find J6 in this pool (data not shown).
The sensitivity of the method used to analyze the putative presence of the J1-J6 peptides in the HLA-B27-bound pool is determined by the features of the MALDI-TOF analysis. In the MS
spectra of Figs. 6B and 8B, the intensity of the signal for each
individual peptide in a given HPLC fraction is relative to the maximal signal from that fraction. For instance, in Fig. 6B, the maximal signal (100%) corresponded to the RRFFPYYVY nonamer
and background noise was about 1%. Thus, absence of a signal
corresponding to J3 suggests that, if present, J3 should be ,1% of
The Journal of Immunology
333
the nonamer. This is considerably lower than the proteasomal yield
of J3, relative to the nonamer, in vitro (see above). Similarly, J4
and J5 would probably be ,1% the amount of the peptides that
give the maximal signals in Fig. 8B. Thus, we cannot exclude that
J1-J6 might be present in amounts ,100-fold lower than the reference nonamer or the most abundant peptides in the corresponding HPLC fractions.
Stability of the J1-J6 peptides in complex with B*2705
Discussion
We have attempted to dissect the factors limiting the peptides naturally presented by HLA-B27 out of the potential ligands derived
from the same region of a protein. The HLA-B27-allospecific
epitope RRFFPYYV and its C-terminally extended nonamer
RRFFPYYVY were both found in the B*2705-bound peptide
pool. Presence of the nonamer, but not the octamer, in B*2701,
despite efficient binding of this peptide to B*2701 in vitro, raised
Table I. Relationship between processing, transport, and binding features of HLA-B27 ligands and their
presence in vivoa
Peptide
RRFFPYYV
RRFFPYYVY
SRRFFPYY
SRRFFPYYV
SRRFFPYYVY
SRRFFPYYVYN
RRFFPYYVYN
RRFFPYYVYNI
a
Name
Proteasome
Yield (%)b
TAP PV
Motifc
Affinityd
Stabilitye DT50
(h)
Found in
vivo
J1
J2
J3
J4
J5
J6
2.5
21.4
0.05
0.5
4.4
0.8
2.3
Not found
1
1
1
1
1
2
2
1
High
High
Low
High
High
Low
High
High
13.2 6 1.8
17.9 6 4.4
7.6 6 0.8
8.5 6 1.3
9.1 6 2.2
9.9 6 3.3
15.2 6 4.1
17.4 6 5.1
Yes
Yes
No
No
No
No
No
No
Features with the potential to impair the presence of the peptide in the B*2705-bound pool in vivo are underlined.
b
Peptide yields (see text) are the means of two digestion experiments at 8 h carried out under the same conditions and with
similar results.
c
TAP-binding motifs are from Refs. (14 and 15).
d
See Fig. 5.
e
DT50 6SD. Data are the means of five to seven independent experiments. See text for an assessment of the DT50 values
for J2 and J3.
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
FIGURE 4. A, HPLC fractionation of the proteasomal digest of the C5
(120 –146) peptide. About 30 mg of substrate was digested for 8 h with 9 3
107 cell-equivalents of 20S proteasome. The elution positions of the octamer epitope, its related nonamer, the detected peptides from the J1-J6 set,
and the undigested 27-mer are indicated. B, Digestion pattern of the C5
(120 –146) peptide by purified 20S proteasome at 8 h. The octamer epitope
sequence is underlined. Peptides found at yields $0.5% of the total digest
are shown as black bars. The octamer (oct), nonamer (non), and J2-J5
peptides are indicated. Arrows indicate cleavage sites that generated peptides with $0.5% yield. Small, medium, and large arrows indicate that 1–3,
4 and 5, and 6 or more peptides, respectively, had N-terminal or C-terminal
ends at that site. Cleavage points that generated peptides with yield ,0.5%
are not shown. The estimated yields, relative to the total digest, of the oct,
non, and J1-J6 are given in Table I.
In a final set of experiments, we analyzed peptide stability in complex with B*2705. The natural octamer and nonamer ligands and
J1-J6 were separately incubated in the presence of human b2-microglobulin with B*2705-T2 transfectant cells. After loading the
exogenous peptides on cell surface-expressed B*2705, the cells
were treated with brefeldin A, to prevent B*2705 egress, and the
decay of HLA-B27-associated fluorescence at the cell surface was
measured as a function of time. The stability of each peptide in
complex with B*2705 was expressed as the time required to obtain
50% of the maximal fluorescence (DT50), measured just before
adding brefeldin A. The results are shown in Table I. The natural
octamer and nonamer ligands had DT50 values of 13.2 and 17.9 h,
respectively, indicating that DT50 values in this range or higher are
appropriate for natural ligands. Thus, absence of J5 and J6 (DT50,
15.2 and 17.4 h, respectively) in vivo was not explained by low
stability. The DT50 value of J4 (9.9 h) was somewhat lower than
for the octamer. However, its binding in vitro was weak (EC50, 61
mM), suggesting that J4 associates inefficiently with B*2705. The
lowest DT50 value corresponded to J1 (7.6 h), a peptide that bound
very weakly in vitro (EC50, .100 mM), J2 (8.5 h), which nevertheless bound efficiently in vitro (EC50, 2 mM), and J3 (9.1 h). It
is unclear whether any of the DT50 values of J1-J6 are outside the
range allowed for natural ligands. However, J2 and J3, which differ
from the octamer and nonamer, respectively, by one additional
N-terminal Ser residue, showed lower stability than the respective
natural ligands. This might be related to the absence of J2 and J3
in vivo (see Discussion).
334
FACTORS LIMITING THE HLA-B27-BOUND PEPTIDE REPERTOIRE
the possibility that the octamer might be generated by trimming of
the nonamer in the ER and that this might be impaired by B*2701
polymorphism (20). However, because both peptides are generated
by the 20S proteasome, trimming is not required to explain the
presence of the octamer in the B*2705-bound peptide pool. This is
in agreement with previous observations that the proteasome directly generates the C-terminal ends of most presented peptides
(2). There are several outstanding aspects of our proteasome digestion experiments. First, the nonamer and the octamer were generated in ;8.5:1 ratio. This is not very different from the estimated
4:1 ratio of these peptides in the B*2705-bound pool (20). It has
been pointed out that in vitro digestions of relatively short peptides
with the 20S proteasome do not resemble the situation in vivo
where ubiquitinated proteins are digested by the 26S proteasome
(2). However, generation of the octamer and nonamer in vitro in a
ratio comparable to that found in the B*2705-bound peptide pool
suggests that their generation in vivo may be similar and a major
determinant of their presence in that pool, with transport and tapasin-mediated loading not having a major influence on their relative abundance.
Second, that the nonamer was the major digestion product after
4 h, when only 30% of the precursor peptide was digested, and the
FIGURE 6. A, Retention times of synthetic RRFFPYYVY and J3. Both peptides were separately subjected to HPLC under the same chromatographic
conditions, and their chromatograms were superimposed. B, MALDI-TOF spectra of HPLC fractions 194 –196 (retention times, 96 –97.5 min) of the
B*2705-bound peptide pool from B*2705-C1R cells. The intensity peak corresponding to RRFFPYYVY in fractions 194 and 195 is marked. The prominent
peak in fraction 196 corresponded to another known B*2705 ligand. Arrows close to mass/charge (m/z) 1400 indicate the expected place for the J3 signal
(M1H1: 1398.6 Da), which was absent.
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
FIGURE 5. Peptide binding to B*2705. Peptides were incubated at various concentrations with B*2705-RMA-S cells at 26°C. HLA-B27-associated
fluorescence was measured 4 h after transfer at 37°C. Binding efficiencies were expressed as EC50 values (see Materials and Methods) and are given for
each peptide. The KTGGPIYKR peptide was used as negative control. Data are means of two independent experiments.
The Journal of Immunology
335
octamer was also in significant amounts indicates that proteasomal
generation of these peptides is strongly favored. That their yields
were maintained after 24 h indicates that their kinetics of generation and stability with the 20S proteasome is similar. Although the
proteasome can destroy class I epitopes (37) and despite cleavage
within the core octamer sequence, the stability of both peptides in
the presence of the proteasome might favor their abundance
in vivo.
The J1-J6 set provided the opportunity to follow, in a defined set
of related peptides, the influence of different factors on shaping the
HLA-B27-bound peptide repertoire. Although our MS analysis
does not rule out that the J1-J6 peptides could be present in very
FIGURE 8. A, Retention times of synthetic J4 and J5. Both peptides were separately subjected to HPLC under the same chromatographic conditions,
and their chromatograms were superimposed. B, MALDI-TOF spectra of HPLC fractions 184 –186 (retention times, 91–92.5 min) of the B*2705-bound
peptide pool from B*2705-C1R cells. Arrows indicate the expected place for the J4 (M1H1: 1512.7 Da) and J5 (M1H1: 1425.6 Da) signals.
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
FIGURE 7. A, Retention times of synthetic RRFFPYYV and J2. Both peptides were separately subjected to HPLC under the same chromatographic
conditions, and their chromatograms were superimposed. B, MALDI-TOF spectrum of HPLC fraction 192 (retention times, 95–95.5 min) of the B*2705bound peptide pool from B*2705-C1R cells. The intensity peak corresponding to RRFFPYYV is marked. The peak at mass/charge (m/z) 1237.4 was
analyzed by quadrupole ion trap electrospray MS/MS. Its fragmentation pattern was compatible with RRFVNVVPTF, a previously reported B*2705 ligand
(see text).
336
FACTORS LIMITING THE HLA-B27-BOUND PEPTIDE REPERTOIRE
for this absence, since in the allowed size range there is room for
nested peptide sets with the potential to bind the class I molecule,
as in the case analyzed here. Rather, besides size and MHC affinity, the specificity features of the processing and transport systems
for class I-bound peptides and the stability of the MHC-peptide
complex are critical limiting factors.
A final word should be devoted to the experimental strategy
used in this study. It was based on in vitro analysis of proteasome
cleavage patterns, peptide binding to “empty” HLA-B27 on the
cell surface, stability of such complexes, and a consideration of
TAP-binding motifs as defined with peptide libraries. Obviously,
in vitro assays do not reproduce the conditions of peptide-processing, transport and MHC binding in vivo. Therefore, our results do
not exclude that processing and binding of the peptides analyzed in
this study may be different in vivo. Despite these limitations, our
analysis explains the presence of two natural ligands of HLA-B27
on the basis of their processing and binding features in vitro. Similarly, failure to detect various related peptides in the B27-bound
pool correlated with one or more unfavorable features related to
their processing, transport, binding, or stability in vitro. This close
correlation cannot be ignored and strongly suggests that some major factors influencing the composition of class I-bound peptide
repertoires can be dissected and characterized in vitro.
Acknowledgments
We thank our colleagues at the Protein Chemistry Laboratory, Anabel Marina and Samuel Ogueta, for their help in MS. We especially thank José G.
Castaño (Instituto de Investigaciones Biomédicas, Madrid) for help and
advice in proteasome purifications, and Daniel López (Instituto de Salud
Carlos III, Madrid) for help in proteasome inhibition studies.
References
1. Deng, Y., J. W. Yewdell, L. C. Eisenlohr, and J. R. Bennink. 1997. MHC affinity,
peptide liberation, T cell repertoire, and immunodominance all contribute to the
paucity of MHC class I- restricted peptides recognized by antiviral CTL. J. Immunol. 158:1507.
2. Rock, K. L., and A. Goldberg. 1999. Degradation of cell proteins and the generation of MHC class I-presented peptides. Annu. Rev. Immunol. 17:739.
3. Serwold, T., and N. Shastri. 1999. Specific proteolytic cleavages limit the diversity of the pool of peptides available to MHC class I molecules in living cells.
J. Immunol. 162:4712.
4. Geier, E., G. Pfeifer, M. Wilm, M. Lucchiari-Hartz, W. Baumeister,
K. Eichmann, and G. Niedermann. 1999. A giant protease with potential to substitute for some functions of the proteasome. Science 283:978.
5. Snyder, H. L., J. W. Yewdell, and J. R. Bennink. 1994. Trimming of antigenic
peptides in an early secretory compartment. J. Exp. Med. 180:2389.
6. Roelse, J., M. Gromme, F. Momburg, G. Hammerling, and J. Neefjes. 1994.
Trimming of TAP-translocated peptides in the endoplasmic reticulum and in the
cytosol during recycling. J. Exp. Med. 180:1591.
7. Elliott, T., A. Willis, V. Cerundolo, and A. Townsend. 1995. Processing of major
histocompatibility class I-restricted antigens in the endoplasmic reticulum.
J. Exp. Med. 181:1481.
8. Hughes, E. A., B. Ortmann, M. Surman, and P. Cresswell. 1996. The protease
inhibitor, N-acetyl-L-leucyl-L-leucyl-L-norleucinal, decreases the pool of major
histocompatibility complex class I-binding peptides and inhibits peptide trimming in the endoplasmic reticulum. J. Exp. Med. 183:1569.
9. Craiu, A., T. Akopian, A. Goldberg, and K. L. Rock. 1997. Two distinct proteolytic processes in the generation of a major histocompatibility complex class
I-presented peptide. Proc. Natl. Acad. Sci. USA 94:10850.
10. Beninga, J., K. L. Rock, and A. L. Goldberg. 1998. Interferon-gamma can stimulate post-proteasomal trimming of the N-terminus of an antigenic peptide by
inducing leucine aminopeptidase. J. Biol. Chem. 273:18734.
11. Gil-Torregrosa, B. C., A. R. Castano, and M. Del Val. 1998. Major histocompatibility complex class I viral antigen processing in the secretory pathway defined by the trans-Golgi network protease furin. J. Exp. Med. 188:1105.
12. Cerundolo, V., J. Alexander, K. Anderson, C. Lamb, P. Cresswell,
A. McMichael, F. Gotch, and A. Townsend. 1990. Presentation of viral antigen
controlled by a gene in the major histocompatibility complex. Nature 345:449.
13. de la Salle, H., D. Hanau, D. Fricker, A. Urlacher, A. Kelly, J. Salamero,
S. H. Powis, L. Donato, H. Bausinger, and M. Laforet. 1994. Homozygous human
TAP peptide transporter mutation in HLA class I deficiency. Science 265:237.
14. Uebel, S., W. Kraas, S. Kienle, K. H. Wiesmuller, G. Jung, and R. Tampe. 1997.
Recognition principle of the TAP transporter disclosed by combinatorial peptide
libraries. Proc. Natl. Acad. Sci. USA 94:8976.
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
low amounts, failure to detect them in the B*2705-bound pool
suggested a significant restriction on the number of putative ligands arising around a core sequence from a given protein, a situation quite different from the nested peptide sets frequent among
class II Ags (38 – 40).
Only the absence of J1 and J4 could be explained on the basis
of low binding to B*2705 in vitro, as all other peptides bound
similarly as natural ligands. Proteasome processing accounted only
partially for further limitation, since only J6 was undetected, and
J1 was generated with very low yield. The possibility that processing in vivo may yield a different peptide pattern is clearly
open. Indeed, since proteasome activity is highly regulated in vivo,
it is possible that the pattern of proteolytic products, especially for
those reflecting minor cleavage specificities, may not be conserved
in vivo. However, on the basis of the correlation between in vitro
processing and in vivo occurrence of the octamer and nonamer
ligands, our results suggest that at least J1-J5 might be generated
in vivo and some (J3 and J5) perhaps in an amount comparable to
the octamer epitope.
The efficiency with which these peptides bind TAP could influence their presence in the B*2705-bound repertoire. TAP-mediated transport was not addressed in this study. However, previous
analyses using combinatorial peptide libraries have established the
importance of the three N-terminal residues and, especially, the
C-terminal one for human TAP binding, and defined the effect of
different amino acid residues at these positions (14, 15). On that
basis, J1-J3 and J6 have favored TAP-binding motifs and would
presumably be transported in a similar way. In contrast, C-terminal
Asn is a disfavored TAP-binding motif, and this might limit availability of J4 and J5 in the ER.
Thus, despite the limitations of in vitro assays, which cannot
obviously reproduce the situation in vivo, it was possible to correlate low proteasomal digestion yields, disfavored TAP-binding
motifs, and/or low B27-binding efficiency with the absence of J1,
J4, J5, and J6 in the B*2705-bound pool. However, the absence of
J3, and perhaps also J2, was not satisfactorily explained by these
criteria.
A possible explanation for the absence of J3 among natural ligands might be its low stability (about 50% in our assay), relative
to the natural nonamer ligand, from which it differs only by an
N-terminal Ser extension. There is evidence for non-proteasomal
cytosolic and ER proteases capable of trimming the N-terminal
residues of proteasomal peptide products (2, 5, 7, 9, 10). It is
conceivable that trimming of J3 to generate the RRFFPYYVY
nonamer might be favored by the higher stability of this peptide in
complex with B*2705. A similar mechanism might explain the
absence of J2, an N-terminal extension of the octamer, which also
binds HLA-B27 in vitro with lower stability (64%) than the octamer ligand.
Thus, of a nested set of eight peptides with size and peptide
motifs appropriate for binding to HLA-B27, only two become
bound in vivo. This limitation cannot be explained by a single
bottleneck feature, but rather by a combination of requirements
including efficient proteasomal processing, appropriate TAP-binding motifs, high affinity for HLA-B27, and sufficient stability of
the B27-peptide complex (Table I). In addition, N-terminal trimming might contribute to further reduce the number of closely
related ligands toward those forming more stable complexes with
HLA-B27. A role of tapasin in limiting loading of these peptides
into HLA-B27, although not ruled out, does not need to be
invoked.
In conclusion, in contrast to class II molecules, nested peptide
sets are unlikely to exist in class I-bound repertoires. The more
strict length requirements for class I-ligands do not fully account
The Journal of Immunology
28. Rock, K. L., C. Gramm, L. Rothstein, K. Clark, R. Stein, L. Dick, D. Hwang, and
A. L. Goldberg. 1994. Inhibitors of the proteasome block the degradation of most
cell proteins and the generation of peptides presented on MHC class I molecules.
Cell 78:761.
29. Lee, D. H., and A. L. Goldberg. 1998. Proteasome inhibitors: valuable new tools
for cell biologists. Trends Cell Biol. 8:397.
30. Craiu, A., M. Gaczynska, T. Akopian, C. F. Gramm, G. Fenteany,
A. L. Goldberg, and K. L. Rock. 1997. Lactacystin and clasto-lactacystin b-lactone modify multiple proteasome b-subunits and inhibit intracellular protein degradation and major histocompatibility complex class I antigen presentation.
J. Biol. Chem. 272:13437.
31. Fenteany, G., R. F. Standaert, W. S. Lane, S. Choi, E. J. Corey, and
S. L. Schreiber. 1995. Inhibition of proteasome activities and subunit-specific
amino-terminal threonine modification by lactacystin. Science 268:726.
32. Ostrowska, H., C. Wojcik, S. Omura, and K. Worowski. 1997. Lactacystin, a
specific inhibitor of the proteasome, inhibits human platelet lysosomal cathepsin
A-like enzyme. Biochem. Biophys. Res. Commun. 234:729.
33. Arribas, J., P. Arizti, and J. G. Castano. 1994. Antibodies against the C2 COOHterminal region discriminate the active and latent forms of the multicatalytic
proteinase complex. J. Biol. Chem. 269:12858.
34. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the
head of bacteriophage T4. Nature 227:680.
35. Rotzschke, O., K. Falk, S. Stevanovic, V. Gnau, G. Jung, and H. G. Rammensee.
1994. Dominant aromatic/aliphatic C-terminal anchor in HLA-B*2702 and
B*2705 peptide motifs. Immunogenetics 39:74.
36. Lamas, J. R., A. Paradela, F. Roncal, and J. A. Lopez de Castro. 1999. Modulation
at multiple anchor positions of the peptide specificity of HLA-B27 subtypes differentially associated with ankylosing spondylitis. Arthritis Rheum. 42:1975
37. Luckey, C. J., G. M. King, J. A. Marto, S. Venketeswaran, B. F. Maier,
V. L. Crotzer, T. A. Colella, J. Shabanowitz, D. F. Hunt, and V. H. Engelhard.
1998. Proteasomes can either generate or destroy MHC class I epitopes: evidence
for nonproteasomal epitope generation in the cytosol. J. Immunol. 161:112.
38. Chicz, R. M., R. G. Urban, W. S. Lane, J. C. Gorga, L. J. Stern, D. A. Vignali,
and J. L. Strominger. 1992. Predominant naturally processed peptides bound to
HLA-DR1 are derived from MHC-related molecules and are heterogeneous in
size. Nature 358:764.
39. Chicz, R. M., R. G. Urban, J. C. Gorga, D. A. Vignali, W. S. Lane, and
J. L. Strominger. 1993. Specificity and promiscuity among naturally processed
peptides bound to HLA-DR alleles. J. Exp. Med. 178:27.
40. Rammensee, H. G., T. Friede, and S. Stevanoviic. 1995. MHC ligands and peptide motifs: first listing. Immunogenetics 41:178.
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
15. Daniel, S., V. Brusic, S. Caillat-Zucman, N. Petrovsky, L. Harrison, D. Riganelli,
F. Sinigaglia, F. Gallazzi, J. Hammer, and P. M. Van Endert. 1998. Relationship
between peptide selectivities of human transporters associated with antigen processing and HLA class I molecules. J. Immunol. 161:617.
16. Pamer, E., and P. Cresswell. 1998. Mechanisms of MHC class I-restricted antigen
processing. Annu. Rev. Immunol. 16:323.
17. Van Endert, P. M. 1999. Genes regulating MHC class I processing of antigen.
Curr. Opin. Immunol. 11:82.
18. Peh, C. A., S. R. Burrows, M. Barnden, R. Khanna, P. Cresswell, D. J. Moss, and
J. McCluskey. 1998. HLA-B27-restricted antigen presentation in the absence of
tapasin reveals polymorphism in mechanisms of HLA class I peptide loading.
Immunity 8:531.
19. Madden, D. R. 1995. The three-dimensional structure of peptide-MHC complexes. Annu. Rev. Immunol. 13:587.
20. Paradela, A., M. Garcia-Peydro, J. Vazquez, D. Rognan, and
J. A. Lopez de Castro. 1998. The same natural ligand is involved in allorecognition of multiple HLA-B27 subtypes by a single T cell clone: role of peptide and
the MHC molecule in alloreactivity. J. Immunol. 161:5481.
21. Lopez, D., R. Garcia Hoyo, and J. A. Lopez de Castro. 1994. Clonal analysis of
alloreactive T cell responses against the closely related B*2705 and B*2703
subtypes: implications for HLA-B27 association to spondyloarthropathy. J. Immunol. 152:5557.
22. Ljunggren, H. G., and K. Karre. 1985. Host resistance directed selectively against
H-2-deficient lymphoma variants. Analysis of the mechanism. J. Exp. Med. 162:
1745.
23. Townsend, A., C. Ohlen, J. Bastin, H. G. Ljunggren, L. Foster, and K. Karre.
1989. Association of class I major histocompatibility heavy and light chains
induced by viral peptides. Nature 340:443.
24. Barnstable, C. J., W. F. Bodmer, G. Brown, G. Galfre, C. Milstein,
A. F. Williams, and A. Ziegler. 1978. Production of monoclonal antibodies to
group A erythrocytes, HLA and other human cell surface antigens: new tools for
genetic analysis. Cell 14:9.
25. Galocha, B., J. R. Lamas, J. A. Villadangos, J. P. Albar, and
J. A. Lopez de Castro. 1996. Binding of peptides naturally presented by HLAB27 to the differentially disease-associated B*2704 and B*2706 subtypes, and to
mutants mimicking their polymorphism. Tissue Antigens 48:509.
26. Ellis, S. A., C. Taylor, and A. McMichael. 1982. Recognition of HLA-B27 and
related antigens by a monoclonal antibody. Hum. Immunol. 5:49.
27. Brooks, J. M., R. A. Colbert, J. P. Mear, A. M. Leese, and A. B. Rickinson. 1998.
HLA-B27 subtype polymorphism and CTL epitope choice: studies with EBV
peptides link immunogenicity with stability of the B27:peptide complex. J. Immunol. 161:5252.
337