| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Medical Department I, Klinikum Benjamin Franklin, Freie Universität Berlin, Berlin, Germany;
Department of Immunology, Deutsches Rheuma Forschungszentrum Berlin, Berlin, Germany;
Department of Biochemistry, Institute for Biochemistry, Charité, Berlin, Germany;
Institute of Molecular Medicine, John Radcliffe Hospital, Oxford, United Kingdom; and
¶ Institute for Anthropology and Human Genetics, Ludwig Maximilians Universität München, Munich, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Among the HLA-B27 subtypes, HLA-B*2705 and HLA-B2704 are the most strongly associated with spondyloarthropathies, whereas subtypes HLA-B*2706 and B*2709 are associated either weakly or not at all (7, 8). Besides HLA-B27, there is strong evidence that bacteria are also crucial for pathogenesis (9, 10). HLA-B27 transgenic rodents develop spondyloarthropathy-like manifestations only if raised in a nonsterile environment (11, 12). In humans the clearest evidence for the role of bacteria in the pathogenesis of spondyloarthropathies is found for Chlamydia trachomatis or certain enterobacteriae in the pathogenesis of ReA. Furthermore, 20–40% of HLA-B27-positive ReA patients move on to ankylosing spondylitis after 10–20 years (13), suggesting that the ReA-associated bacteria can cause ankylosing spondylitis, possibly through a cross-reaction between a bacterial and a self Ag (10).
The main hypothesis advanced for the association between HLA-B27 and spondyloarthropathies is the arthritogenic peptide theory. It states that some B27 subtype alleles, due to their unique amino acid residues, bind a specific arthritogenic peptide(s) that is recognized by CD8-positive T cells (6, 14, 15, 16). Thus, for further investigation and verification of this hypothesis it is important to identify HLA-B27-specific peptides able to stimulate CD8-positive T cells.
Highly selective screens using, for instance, lysteriolysin for Listeria (17) and bacteria-specific T cell clones or lines are of limited value (18, 19) because they might miss the relevant epitopes. Therefore, we have studied Chlamydia trachomatis, the genome of which has recently been sequenced, using a more comprehensive approach (20). A huge number of potential HLA-B27-binding peptides occur in the genomes, indeed too many to test in functional assays such as proliferation or CD8 T cell cytotoxicity assays. To overcome this problem we combined two computer programs that allow preselection of potentially relevant HLA-B27 peptides together with flow cytometry (21) to identify HLA-B27-restricted CD8+ stimulatory chlamydial peptides. The SYFPEITHI epitope prediction program identifies peptides on the basis of anchor residues required for HLA-B27 binding (22), while the proteasome program (23, 24) is based on the proteasomal peptide cleavage properties responsible for the generation of most MHC class I peptides. Using this novel approach we screened the entire C. trachomatis proteome and identified a number of B27-restricted MHC-class I CD8+ epitopes that might be potentially relevant for disease pathogenesis.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CBA/J mice (H-2k), 6–12 wk old and
transgenic for HLA-B*2705 and human
2-microglobulin (9, 25), and nontransgenic CBA/J-mice were maintained in the barrier
facilities of the Klinikum Benjamin Franklin (Berlin, Germany), where
all breedings were performed. The transgenic mice were immunized i.v.
with 3 x 107 IFU purified infectious
elementary bodies of C. trachomatis serovar L2. Three weeks
later the mice were boosted with 3 x 107
inclusion forming units elementary bodies. One week after the second
immunization, spleen cells from three to five mice were harvested,
pooled, and used for either cytokine secretion experiments or
cytotoxicity assays. The elementary bodies were a gift from Dr. M.
Hartmann, Friedrich Schiller University (Jena, Germany) and were
prepared as previously described (26). The local review
board for animal studies approved these experiments.
RMA cell lines and culture media
RMA is a Rauscher leukemia virus-induced cell line of C57BL/6
(H-2b) origin; RMA-S is a mutant of this cell
with a defect in peptide loading of class I molecules due to a mutant
TAP-2 locus (27). RMA-S-B27 cells are
transfectants expressing the HLA-B*2705 subtype; the 
1 and 2
domains are human, whereas the 3 domain is of the
H-2b haplotype. The cells were a gift from Dr.
H.-G. Rammensee (Institute for Cell Biology, Tübingen,
Germany). RMA-S cells and their transfectants express very low amounts
of class I molecules, which are unstable on the cell surface when
cultured at 37°C due to their TAP defect. However, when cultured at
28°C RMA-S cells and their B27 transfectants can express the empty
class I heavy chains, or, alternatively, the heavy chains may be
occupied by low affinity peptides (28). At 37°C surface
expression of the heavy chains can only be stabilized by adding
exogenous peptide ligands. The cells were cultured at 37°C in 5%
CO2 in RPMI 1640 culture medium, supplemented
with 10% FCS, 2 mM L-glutamine, 100 U/ml
penicillin, 100 µg/ml streptomycin sulfate, and 5 x
10-5 M 2-ME (Sigma, St. Louis, MO). For stable
transfection cells were periodically supplemented with 0.5 mg/ml G-418
sulfate. If not stated otherwise, all media and supplements were
obtained from Life Technologies (Karlsruhe, Germany).
Epitope stabilization assay
For peptide loading, RMA-S-B27 cells were preincubated in RPMI 1640 culture medium with 2% FCS at 28°C for 6 h and cultured for an additional 8 h with 100 µM selected chlamydial or control peptides. The cells were then transferred to 37°C for 10 h. Cell surface staining of HLA-B27 on the RMA-S-B27 cells was performed by direct staining in 45 µl of RPMI 1640 medium and 5 µl of mouse anti-human HLA-B27-FITC (HLA-ABCm3; Serotec, Eching, Germany) for 20 min at room temperature; subsequently, the cells were washed twice in PBS/1% BSA, resuspended in FACS buffer, and analyzed by flow cytometry using a FACScan flow cytometer (BD Biosciences, San Jose, CA).
Search for HLA-B27-binding peptides derived from C. trachomatis
A search for HLA-B27 binding nonamer peptides from C.
trachomatis proteins was conducted with an epitope prediction
program described by Rammensee and colleagues (22). The
SYFPEITHI program, (http://www.uni-tuebingen.de/uni/kxi.) combines
knowledge about MHC class I binding patterns and functional data (T
cell responses to MHC class I/peptide complexes). A scoring system for
HLA-B27 binding was used, giving highest numbers to anchor positions
with arginine (R) at position 2, second highest numbers for amino acids
at auxiliary anchor positions (position 9 for HLA-B27), and lower
numbers for preferred amino acids (29, 30). Peptides were
selected for a binding score of 
25 and in an extra round for a
binding score between 20 and 24.
Prediction of cleavage-determining amino acid motifs of the 20S proteasome
A second mathematical program mimicking functions of the cleavage process and based on calculations of particular cleavage-determining amino acid motifs in 20S proteasomes (23, 24) was used to investigate whether peptides selected by the epitope prediction program would be cut by proteasomes. The main assumption of this program is that the mutual recognition of peptide substrates and active sites of the proteasome is determined by the presence of distinct amino acid residue motifs in the vicinity of the scissile bond. For this reason, each nonamer of interest was enlarged by the 10 natural amino acids at the N- and the C-terminal ends next to the presumptive cleavage site. On this basis, the likelihood for the preselected nonamer peptides to be cleaved by the 20S proteasome was analyzed.
Peptide synthesis
Nonamer peptides were synthesized according to standard F-moc solid phase synthesis methods on a Syro-Synthesizer (MultiSynTech, Witten, Germany), purified by HPLC (Shimadzu LC-10; Shimadzu Scientific Instruments, Duisborg, Germany), and identified by mass spectroscopy (LCQ ion trap; Thermoquest, Eberbach, Germany). The purity of the peptides was >95%. Peptides were dissolved in DMSO and further diluted with serum-free culture medium at a concentration of 5 mg/ml. The peptides were stored frozen at -20°C.
Identification of B27-restricted chlamydial nonamer peptides by Ag-specific flow cytometry
Peptide-specific CD8+ T cells from HLA-B27
transgenic mice primed in vivo with C. trachomatis were
stimulated in vitro with pools of peptides. The 199 peptides were
created in two peptide matrixes comprising 10 pools (pools 1–10 and
11–20 each) in a horizontal direction, while the respective pools
21–30 and 31–40 were arranged in the vertical direction. Each pool
contained a mixture of 10 different peptides. Thus, both matrixes
comprised 40 pools in total to ensure that each peptide was represented
by a horizontal and a vertical pool. An individual stimulating peptide
was identified by crossing of two stimulating pools. To determine
whether selected peptides could stimulate CD8+
splenic T cells from Chlamydia-primed HLA-B27-transgenic
mice, intracellular cytokine staining was used after Ag-specific T cell
stimulation (21). Briefly, 2 x
106 primed cells were stimulated for 6 h in
1 ml of culture medium in 10-ml tissue culture tubes (Greiner,
Nurtingen, Germany) with anti-CD28 Ab (1 µg/ml) for optimal
stimulation (31) plus mixed peptide pools (5 µg/ml
peptide each), individual peptides (10 µg/ml), or no antigenic
peptide (negative control). Stimulation was stopped with brefeldin A (5
µg/ml) after 2 h, and after an additional 4 h cells were
harvested and stained with anti-CD69-FITC (5 µg/ml; BD
PharMingen, San Diego, CA) and anti-CD8-APC (1 µg/ml; BD
PharMingen). Next, the cells were fixed in 2% formalin for 15 min and
washed, and the pellet was resuspended for 10 min in saponin buffer
followed by staining with PE-conjugated rat anti-mouse IFN-
(1
µg/ml; BD PharMingen). After gating on CD8-positive T cells, cells
that were double positive for the early activation marker CD69 and for
intracellular IFN- were regarded as Ag specific. Analysis was
performed using a BD Biosciences FACScan flow cytometer with CellQuest
software.
In vitro expansion of Chlamydia-specific CTL
To investigate CD8+ T cells for cytotoxicity against the selected chlamydial peptides, splenocytes from three to five mice, primed in vivo with Chlamydia, were resuspended in RPMI 1640 culture medium. To increase the number of Ag-specific cytotoxic CD8+ T cells, 1 x 107 primed splenic cells in 7 ml of culture medium were transferred into 50-ml upright flasks each (Nunc, Wiesbaden, Germany) and restimulated in vitro with 10 µg/ml respective chlamydial peptide or control peptide. After 6 days of incubation at 37°C in 5% CO2, cells were harvested, Ficoll-Paque (Pharmacia, Freiburg, Germany)-purified, resuspended in culture medium, counted, and used as effector cells for cytotoxicity assays.
Cytotoxicity assay against chlamydial peptides
Cytotoxic activity was measured on peptide-loaded RMA-S-B27 cells. For this, RMA-S-B27 cells were incubated and loaded with different peptide fractions each at 28°C in RPMI 1640/2% FCS and then shifted to 37°C as described. The cells were then washed with PBS, and the loaded cell pellets were labeled at 37°C with 100 µCi of Na51Chromat (Amersham, Braunschweig, Germany) for 1 h. After washing three times in PBS the labeled target cells were resuspended in RPMI 1640 supplemented with 1% FCS, 2 mM L-glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin sulfate, and 5 x 10-5 M 2-ME (Sigma) For the assay, 5 x 103 targets were seeded in 96-well U-bottom plates (Nunc) along with different numbers of effector cells in a total volume of 200 µl. Spontaneous release was determined in wells with target cells and medium alone. Maximum release was determined by adding 100 µl of 2% Triton X-100 (Sigma) to wells containing 100 µl of target cells without CTL. Target cells were tested for lysis in triplicate in a standard 6-h chromium release assay. Specific lysis was calculated as: 100 x (cpm experimental release - cpm spontaneous release/cpm maximal release - cpm spontaneous release). Data presented are the means of triplicate determinations.
Identification of B27-restricted chlamydial peptides in patients with Chlamydia-induced ReA
In three HLA-B27+ patients with
Chlamydia-induced ReA (patient 1: male, 44 years old, 2-wk
duration of the arthritis; patient 2: male, 34 years old, 4-wk duration
of arthritis; patient 3: male, 36 years old, 4-wk duration of
arthritis) synovial fluid mononuclear cells (SF) were investigated as
described for B27-transgenic mice. Briefly, SF of the patient were
separated through Ficoll gradient. Cells were then stimulated for
6 h with pools of peptides and analyzed as indicated for the mouse
experiments. Cells were stained with self-conjugated anti-CD69-PE
(5 µg/ml), anti CD8-Cy5 (1 µg/ml), and self-conjugated
FITC-anti-human IFN- (1 µg/ml) Ab. After gating on
CD8-positive T cells, cells that were double positive for the early
activation marker CD69 and intracellular IFN- were regarded as Ag
specific. HLA-B27 subtyping was performed by PCR and hybridization. All
patients were positively typed for HLA-B*2705. Approval of the local
ethical committee of the Benjamin Franklin Hospital was achieved for
the investigation of patients’ synovial fluid.
FACS analysis of Chlamydia peptide-specific T cells with HLA-B27 tetramers
HLA-B27 tetramers were generated as previously described (32) with some modifications. The expression vector pLM1-HLA-B27 (a gift from E. Collins, D. Garboczi, and D. Wiley, Boston University, Boston, MA) was modified by tagging with the BirA recognition sequence as previously described (33) and by mutating the Cys67 to Ser67 (34). Correct folding and biotinylation of HLA-B27 were analyzed by gel filtration and gel electrophoresis, HLA-B27 monomers were tetramerized with PE-labeled streptavidin for FACS analysis. We generated HLA-B27 tetramers with the EBV peptide nucleoprotein (NP; aa 258–266) (35), the rat peptide L5 (aa 36–44) (36), and chlamydial peptides 146 and 195. Mononuclear cells from a B27+ donor with EBV infection were stained with the HLA-B27/EBV NP (aa 258–266) tetramer immediately after harvesting the cells and 14 days after stimulation with 1 µg/ml EBV NP (aa 258–266) peptide in the presence of IL-2 and IL-7.
For B27/tetramer analysis SF from patient 3 were available. Furthermore, SF from two other patients (patient 4: female, 22 years old, B27+, 4-wk duration of Chlamydia-induced ReA; patient 5: male, 22 years old, B27+, 5-wk duration of oligoarthritis, arthritis not triggered by Chlamydia) that had been stored in liquid nitrogen were available for tetramer staining, For the latter two patients, we did not have enough cells to investigate intracellular cytokine staining in parallel. SF from frozen samples of these patients were incubated overnight in culture medium. After performing a Ficoll gradient to get rid of the dead cells, mononuclear cells were incubated with anti-CD3-FITC, anti-CD8-PerCP, and tetramers for 30 min at room temperature in 0.5% BSA/0.05% azide, followed by flow cytometry. CD3-positive T cells were gated. As a negative control the HLA-B27/rat L5 (aa 36–44) tetramer was used.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Analysis of the chlamydial genome by Stephens and colleagues
(20) resulted in the identification of 894 likely
protein-coding genes. The protein sequences deduced from this database
were the basis for our screening approach for candidate nonamer
peptides with binding to HLA-B*2705. In the first step all nonamer
peptides with arginine (R) at position 2 (29, 30) were
selected. This selection step resulted in about 18,000 nonamer
peptides. Selecting for only those peptides with an arbitrary binding
score of 25 to HLA-B*2705 reduced the number of peptides to 1881. The
biological reason for selecting a binding score of 25 or greater was
that most of the published natural peptides identified in the
HLA-B*2705 molecule showed a binding score of 25 when we checked
these natural ligands with the binding program.
Peptide generation by the proteasome program
In a second step we asked which of the 1881 selected peptides
would also be processed by the proteasome system (Fig. 1
). One hundred and ninety-nine of the
1881 nonamer peptides fulfilled this criterion (cutting probability,
4.5). All of these nonamer peptides were then synthesized, and their
function tested.
|
To identify peptide-specific CD8+ T cells
derived from HLA-B27 transgenic mice that were infected in vivo with
C. trachomatis, primed splenocytes were stimulated with
pools of the selected 199 peptides. Table I summarizes the results of four
different mouse experiments using pools of peptides for the stimulation
of CD8+ T cells. Eleven of 40 pools were
stimulatory in all four experiments, and the CD8+
T cell response could be attributed at the level of a single peptide to
nine peptides (Table I). In Fig. 2 two
examples of single positive peptides (131 and 133) with their
corresponding pools (14/31 and 14/33, respectively) are shown. The same
experiment performed in parallel at the same time gave nearly identical
results (data not shown), proving that the results are reproducible.
For the remaining 28 pools the percentage of CD69/IFN-
double-positive CD8 T cells was comparable to the control value
(
0.02%) in the presence of anti-CD28 but without Ag. The
sequences of these stimulatory peptides, their source in the database,
and the proteins from which they are derived are shown in Table II. When we performed the same experiment
with nontransgenic mice after immunization with Chlamydia
and in vitro peptide stimulation, the percentage of IFN--positive
cells was not higher than that with the negative controls (without
peptide; data not shown).
|
|
|
To confirm that the nine selected peptides were able to bind to
HLA-B27, they were tested in an assembly assay with the TAP-defective
RMA-S-B27 transfectant mouse cell line (27). RMA-S-B27
transfectants were incubated with 100 µM of each peptide at 28°C,
then incubated at 37°C for 10 h. HLA-B27 molecules stabilized by
peptides were monitored by direct staining with an FITC-coupled,
anti-HLA-B27-specific mAb. Fig. 3
shows the different features of the transfectant cells with and without
added peptides at different incubation temperatures. RMA-S-B27 cells
incubated at 37°C without peptide showed no HLA-B27 expression, as
expected (Fig. 3A), but did so at 28°C (Fig. 3B). However, when the temperature was shifted from 28 to
37°C (Fig. 3C) we found that after 10 h the cell line
no longer showed empty heavy chain expression (thermolabile). In
contrast, when the RMA-S-B27 line was incubated with exogenous peptides
at 28°C, stable expression of the heavy chains was maintained at
37°C (thermostable). Incubation with a control peptide unrelated to
HLA-B27 binding revealed no stabilization of HLA-B27 on the surface
(Fig. 3D), whereas the B27 undecamer self-peptide (aa
169–179) derived from a polymorphic region of the 1 domain
(37) and the influenza nonamer NP (aa 383–391)
(38), both known to bind B27, showed stabilization of the
molecule (Fig. 3, E and F). As indicated by the
binding of the anti-B27 Ab, all nine selected peptides, shown to
stimulate IFN- secretion by CD8+ T cells, were
also found to stabilize HLA-B27 on the cell surface after shifting the
cells from 28 to 37°C. The binding efficiency of the nine peptides
was comparable to that of the B27 undecamer self-peptide or the
influenza nonamer peptide NP (aa 383–391).
|
We asked next whether peptides that induce IFN- release also
stimulated a CD8+ cytotoxic T cell response.
Eight of the nine peptides tested triggered Ag-specific lysis of target
cells by CD8+ cytotoxic T cells (Fig. 4). Lysis at 10% for peptide 194 was
weak, but was still regarded as positive, because lysis obtained with
the control peptide was negligible. The self-derived B27 control
peptide did not induce cytotoxicity, although it is a good B27 binder.
No lysis was observed for peptide 80. When this experiment was
performed with either unprimed splenocytes from B27-transgenic mice or
splenocytes derived from Chlamydia-primed nontransgenic CBA
mice, no lysis was observed after in vitro restimulation of splenic
cells with the respective peptides of RMA-S-B27 peptide-loaded cells
(data not shown).
|
We next investigated whether the same approach would allow
identification of peptide-specific CD8+ T cells
from the SF of Chlamydia-induced ReA
patients. SF of patients were stimulated
with pools of peptides as described in the mouse experiments.
Using the peptide matrix we could identify 11 human peptides that
proved also to be stimulatory on the single peptide level: four
peptides (8, 68, 145, and 195) in patient 1, six peptides (8, 68, 133,
144, 146, and 194) in patient 2 and eight peptides (68, 80, 133, 138,
146, 194, 195, and 196) in patient 3 (Fig. 5). One of the peptides was
seen by all three patients (68), whereas peptides 133, 146, 194, and
195 were seen by two patients (Table III). Eight of 11 human peptides were
also seen by murine CD8+ T cells (Table III),
whereas peptide 131 was only recognized by murine, but not human, T
cells (Tables II
and III).
|
|
We used tetramers of HLA-B27 with Chlamydia peptides
146 and 195 to stain the SF of two patients with
Chlamydia-triggered ReA and one with
B27+ control arthritis, not triggered by
Chlamydia (Fig. 6A). The HLA-B27/rat L5 (aa
36–44) peptide tetramer was used for control experiments. For these
studies frozen cells were examined after 1 night of incubation in
culture medium without further expansion and were directly analyzed by
flow cytometry. In Chlamydia-induced arthritis, two
patients (3 and 4), but not the B27+
control patient (5), had CD8+ T cells specific
for the B27/Chlamydia peptide 195 tetramer (Fig. 6A). No staining was observed with the
B27/Chlamydia peptide 146 complex or with the control rat L5
peptide tetramer. The B27/tetramer of the rat L5 peptide did not bind
to any T cells in all three patients, indicating the specificity of
this experiment. The HLA-B27/EBV NP (aa 258–266) complex specifically
detected T cells in a patient with previous EBV infection (Fig. 6B). Tetramer-specific T cells could be expanded from 0.12
to 56.3%, indicating that the HLA-B27 NP258–266
tetramer had a biological function (Fig. 6B). Some tetramer
staining was observed for CD8-negative T cells. This has also been
reported in the past by other groups (39, 40), especially
when cells were analyzed without prior in vitro T cell expansion by
peptide stimulation (41).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The two biomathematical computer programs (23, 24), which
were combined, allowed us to reduce the number of HLA-B27-restricted
peptides to a workable amount of 199 from an initial number of
18,000 nonamers with an arginine at position 2. When we set a
binding score of 25, we still obtained a number of 1881 nonamer
peptides. In the initial computerized screening step (22)
peptides were selected according to the presence of canonical HLA-B27
anchor residues and the probability of their binding to HLA-B27
molecules. Although this program allows the identification of
HLA-B27-binding peptides with high fidelity, it does not allow us to
predict which of these, due to the given sequence environment within a
protein, are actually immuno-relevant and generated by the cellular
proteasome system. Therefore, we screened the 1881 theoretical peptides
with a proteasome-based cleavage site prediction program based on the
enzymatic properties of the 20S proteasome and the established
efficiency of the system to generate MHC class I molecules
(42). Using the proteasome program it is important to
realize that the predicted epitopes will be generated with high
probability. However, it cannot be excluded that the program, due to a
still limited database, is restrictive against peptides that are
generated by the proteasome but do not obey the presently known
cleavage parameters. Nevertheless, both programs have a predictive
accuracy of 90% (22, 23). Thus, the first experimental
combination of the two computerized epitope prediction programs allowed
us to tackle an entire C. trachomatis proteome and to
overcome the pitfalls of biochemical approaches to elute
pathogen-derived peptides from the MHC molecules on the cell surface in
sufficient amounts for analysis.
To test the selected peptides for their immunogenicity we used
Ag-specific flow cytometry based on the staining of cell surface
expression markers and intracellular cytokine secretion (21, 43, 44). This method identifies Ag-specific T cells directly ex vivo
after short-term stimulation, preventing a change in the Ag specificity
by long term cultures of T cells. Furthermore, Ag-specific
CD8+ T cells can be detected even at low
frequencies of 0.1%, a sensitivity that is not reached by assays
previously used. A recent study used a related approach for the
identification of peptides derived from the tumor-associated Ag
preferentially expressed Ag of melanoma presented by HLA-A*0201.
However, this investigation did not use Ag-specific flow cytometry as a
readout for CD8 response and was limited to the analysis of just one
protein (45).
The chlamydial peptides from proteins with known function are mainly derived from either secretory or conserved proteins (20). Secretory proteins are of special interest because they might obtain easier access to the MHC class I pathway. Conserved Ags shared by various bacterial species might also be of relevance because not only C. trachomatis, but also enterobacteriae such as Yersinia, Salmonella, or Shigella, can induce ReA. Furthermore, conserved bacterial Ags might be crucial for cross-reactivity with human self Ags. Some of the peptides identified are derived from hypothetical proteins with unknown function. Immunodominant peptides from hypothetical proteins have also been reported for Borrelia burgdorferi (19).
Until now there have been no data about how many peptides derived from
a whole bacterium are stimulatory to CD8+ T cells
when presented by a specific MHC class I molecule. About 10 peptides
found in our study present a rather small number. However, for viruses,
although containing 100-fold fewer proteins compared with bacteria,
it had been found that only a single peptide can be immunodominant for
the CD8 response in the context of an MHC class I molecule (46, 47).
Because the chosen restriction scores for both programs were arbitrary, it appeared possible that we might have missed potentially positive peptides due to the applied stringent score values. Therefore, we searched in an additional approach for peptides with a binding score between 20 and 24. To restrict the resulting number of peptides two additional selection steps had to be used, choosing groups of candidate proteins and excluding good binders to the HLA-B*2709 subtype, which is not associated with disease (8). In support of the initially applied restrictive selection parameters none of the selected 69 additional found peptides turned out to be stimulatory (data not shown).
By using the HLA-B27-transgenic mouse model we could show that
all peptides were indeed B27 binders, and we confirmed the
immunogenicity of eight of nine identified peptides in a cytotoxic T
cell assay (Fig. 4), while the B27 self-derived peptide was tolerated
and thus did not induce cytotoxicity. Earlier studies reported a good
HLA-B27 binding of chlamydial heat shock protein 60-derived
peptides (48), and a recent study demonstrated HLA-A2- and
HLA-B51-restricted CD8+ CTLs for the major outer
membrane protein of C. trachomatis in humans
(49). However, no stimulatory, HLA-B27-restricted
chlamydial major outer membrane protein- or 60-kDa heat shock
protein-derived peptide was identified.
Our first and main aim in this study was to establish a method for the identification of HLA-B27-restricted, bacteria-derived peptides that are immunogenic for CD8+ T cells using the HLA-B27-transgenic mouse, which gave us reproducible and stable experimental conditions. Although similar epitopes in mice and humans had been described in the past, it has been doubted whether peptide processing and T cell repertoire are the same in mice and humans, because auxiliary molecules that help to load MHC class I might be not identical (50). Therefore, it was of great importance that we could also investigate three HLA-B27-positive patients with Chlamydia-induced ReA. The access to synovial fluid, directly from the site of inflammation, is an advantage compared with studies using peripheral blood, because we have shown recently that the Ag-specific frequency is about 10 times higher in synovial fluid than in peripheral blood (43). In these patients we identified 11 peptides, three of which were only seen by human T cells, one only by mice, and eight by both human and murine T cells. The relevance of the single peptides for the pathogenesis has to be investigated in future studies.
The arthritogenic peptide hypothesis is currently the one most likely to explain the presentation of peptides to CD8-positive T cells and the striking association of HLA-B27 with human diseases. Evidence for this is found in transgenic animal systems (51, 52) as well as in human studies (15, 16, 53, 54). Furthermore, the fact that the disease not-associated subtypes HLA-B*2706 and B*2709 differ from the associated subtypes by only one or two amino acids located in the peptide binding groove (7, 8) can best be explained by this hypothesis. A CD8+ HLA-B27-restricted T cell response against Chlamydia or against ReA-associated bacteria might not be essential for an effective anti-bacterial immune response. However, we have postulated (6, 10) that this small response might be sufficient to induce an autoimmune response in HLA-B27-positive patients who move on to ankylosing spondylitis (13). Thus, although the number of Ag-specific CD8+ T cells in our assay is relatively small compared with the background, we think that this response is relevant. It has been a general problem in the past to identify Ag-specific T cells of low frequencies, for example directed against autoantigens (55). The immunogenicity of the peptides identified in this study has to be confirmed in future experiments. For mice, we already did this by using a cytotoxic assay, and for humans we presented data using B27/peptide-specific tetramers. We will expand on these approaches in the future by separating Ag-specific CD8+ T cell using B27/chlamydial peptide-specific tetramers, as described recently for HLA-A2- and HLA-B51-restricted chlamydial major outer membrane protein peptide motifs (56), and by using the cytokine secretion assay (57) to prove the relevance of the peptides identified here for the pathogenesis of disease. In this study we present a feasible approach for the identification of immunodominant peptides from a complex bacterium.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Joachim Sieper, Department of Medicine, Division of Rheumatology, Klinikum Benjamin Franklin, Freie Universität Berlin, Hindenburgdamm 30, 12200 Berlin, Germany. E-mail address: hjsieper{at}zedat.fu-berlin.de
3 Abbreviations used in this paper: ReA, reactive arthritis; NP, nucleoprotein; SF, synovial fluid mononuclear cells.
Received for publication April 17, 2001. Accepted for publication August 3, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
His116, differentiating the ankylosing spondylitis-associated HLA-B*2705 from the non-associated HLA-B*2709 subtype, influences peptide-specific CD8 T cell recognition. Eur. J. Immunol. 28:2508.[Medline]
2-microglobulin: a model of human spondyloarthropathies. J. Exp. Med. 182:1153.2-microglobulin control cell surface expression of HLA-B27 in transgenic mice with a 25 kb B27 fragment. I. K. Egorov, and C. S. David, eds. Transgenic Mice and Mutants in MHC Research 205. Springer Verlag, Berlin.
-encoded tyrosine for T cell recognition of an HLA B27/peptide complex. Eur. J. Immunol. 28:2704.[Medline]
2-microglobulin-free heavy-chain homodimer structure. J. Immunol. 162:5045.12 domain of HLA-B27 contributes to the disease pathogenesis in transgenic mice. Hum. Immunol. 60:116.[Medline]
This article has been cited by other articles:
|
|
 |
|
  C Wright, M Edelmann, K diGleria, S Kollnberger, H Kramer, S McGowan, K McHugh, S Taylor, B Kessler, and P Bowness Ankylosing spondylitis monocytes show upregulation of proteins involved in inflammation and the ubiquitin proteasome pathway Ann Rheum Dis, October 1, 2009; 68(10): 1626 - 1632. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. J. Cragnolini, N. Garcia-Medel, and J. A. Lopez de Castro Endogenous Processing and Presentation of T-cell Epitopes from Chlamydia trachomatis with Relevance in HLA-B27-associated Reactive Arthritis Mol. Cell. Proteomics, August 1, 2009; 8(8): 1850 - 1859. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. T. Loffredo, J. Sidney, A. T. Bean, D. R. Beal, W. Bardet, A. Wahl, O. E. Hawkins, S. Piaskowski, N. A. Wilson, W. H. Hildebrand, et al. Two MHC Class I Molecules Associated with Elite Control of Immunodeficiency Virus Replication, Mamu-B*08 and HLA-B*2705, Bind Peptides with Sequence Similarity J. Immunol., June 15, 2009; 182(12): 7763 - 7775. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J. Cragnolini and J. A. L. de Castro Identification of Endogenously Presented Peptides from Chlamydia trachomatis with High Homology to Human Proteins and to a Natural Self-ligand of HLA-B27 Mol. Cell. Proteomics, January 1, 2008; 7(1): 170 - 180. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Holland, N. Faal, I. Sarr, H. Joof, M. Laye, E. Cameron, F. Pemberton-Pigott, H. M. Dockrell, R. L. Bailey, and D. C. W. Mabey The Frequency of Chlamydia trachomatis Major Outer Membrane Protein-Specific CD8+ T Lymphocytes in Active Trachoma Is Associated with Current Ocular Infection Infect. Immun., March 1, 2006; 74(3): 1565 - 1572. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J Zou, H Appel, M Rudwaleit, A Thiel, and J Sieper Analysis of the CD8+ T cell response to the G1 domain of aggrecan in ankylosing spondylitis Ann Rheum Dis, May 1, 2005; 64(5): 722 - 729. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Appel, W. Kuon, M. Kuhne, M. Hulsmeyer, S. Kollnberger, S. Kuhlmann, E. Weiss, M. Zeitz, K. Wucherpfennig, P. Bowness, et al. The Solvent-Inaccessible Cys67 Residue of HLA-B27 Contributes to T Cell Recognition of HLA-B27/Peptide Complexes J. Immunol., December 1, 2004; 173(11): 6564 - 6573. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Kuon, M. Kuhne, D. H. Busch, P. Atagunduz, M. Seipel, P. Wu, L. Morawietz, G. Fernahl, H. Appel, E. H. Weiss, et al. Identification of Novel Human Aggrecan T Cell Epitopes in HLA-B27 Transgenic Mice Associated with Spondyloarthropathy J. Immunol., October 15, 2004; 173(8): 4859 - 4866. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kollnberger, L. A. Bird, M. Roddis, C. Hacquard-Bouder, H. Kubagawa, H. C. Bodmer, M. Breban, A. J. McMichael, and P. Bowness HLA-B27 Heavy Chain Homodimers Are Expressed in HLA-B27 Transgenic Rodent Models of Spondyloarthritis and Are Ligands for Paired Ig-Like Receptors J. Immunol., August 1, 2004; 173(3): 1699 - 1710. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Hulsmeyer, M. T. Fiorillo, F. Bettosini, R. Sorrentino, W. Saenger, A. Ziegler, and B. Uchanska-Ziegler Dual, HLA-B27 Subtype-dependent Conformation of a Self-peptide J. Exp. Med., January 20, 2004; 199(2): 271 - 281. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Sesma, I. Alvarez, M. Marcilla, A. Paradela, and J. A. L. de Castro Species-specific Differences in Proteasomal Processing and Tapasin-mediated Loading Influence Peptide Presentation by HLA-B27 in Murine Cells J. Biol. Chem., November 21, 2003; 278(47): 46461 - 46472. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Zhong, P. A. Reche, C.-C. Lai, B. Reinhold, and E. L. Reinherz Genome-wide Characterization of a Viral Cytotoxic T Lymphocyte Epitope Repertoire J. Biol. Chem., November 14, 2003; 278(46): 45135 - 45144. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Stebbing, D. Bourboulia, M. Johnson, S. Henderson, I. Williams, N. Wilder, M. Tyrer, M. Youle, N. Imami, T. Kobu, et al. Kaposi's Sarcoma-Associated Herpesvirus Cytotoxic T Lymphocytes Recognize and Target Darwinian Positively Selected Autologous K1 Epitopes J. Virol., April 1, 2003; 77(7): 4306 - 4314. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. H. Boyle and J. S. Hill Gaston Breaking the rules: the unconventional recognition of HLA-B27 by CD4+ T lymphocytes as an insight into the pathogenesis of the spondyloarthropathies Rheumatology, March 1, 2003; 42(3): 404 - 412. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Rodel, H. Vogelsang, K. Prager, M. Hartmann, K.-H. Schmidt, and E. Straube Role of Interferon-Stimulated Gene Factor 3{gamma} and Beta Interferon in HLA Class I Enhancement in Synovial Fibroblasts upon Infection with Chlamydia trachomatis Infect. Immun., November 1, 2002; 70(11): 6140 - 6146. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Constantin, E. E. Bonney, J. D. Altman, and O. L. Strickland Major Histocompatibility Complex (MHC) Tetramer Technology: An Evaluation Biol Res Nurs, October 1, 2002; 4(2): 115 - 127. [Abstract] [PDF]
|
|
|
|
|
|
I. Popov, C. S. Dela Cruz, B. H. Barber, B. Chiu, and R. D. Inman Breakdown of CTL Tolerance to Self HLA-B*2705 Induced by Exposure to Chlamydiatrachomatis J. Immunol., October 1, 2002; 169(7): 4033 - 4038. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Ramos, I. Alvarez, L. Sesma, A. Logean, D. Rognan, and J. A. Lopez de Castro Molecular Mimicry of an HLA-B27-derived Ligand of Arthritis-linked Subtypes with Chlamydial Proteins J. Biol. Chem., September 27, 2002; 277(40): 37573 - 37581. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J Zou, M Rudwaleit, A Thiel, R Lauster, J Braun, and J Sieper T cell response to human HSP60 and yersinia 19 kDa in ankylosing spondylitis and rheumatoid arthritis: no evidence for a causal role of these antigens in the pathogenesis Ann Rheum Dis, May 1, 2002; 61(5): 473 - 474. [Full Text]
|
|
|
|
|
|
R. Holtappels, D. Thomas, J. Podlech, and M. J. Reddehase Two Antigenic Peptides from Genes m123 and m164 of Murine Cytomegalovirus Quantitatively Dominate CD8 T-Cell Memory in the H-2d Haplotype J. Virol., January 1, 2002; 76(1): 151 - 164. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Alvarez, M. Marti, J. Vazquez, E. Camafeita, S. Ogueta, and J. A. Lopez de Castro The Cys-67 Residue of HLA-B27 Influences Cell Surface Stability, Peptide Specificity, and T-cell Antigen Presentation J. Biol. Chem., December 21, 2001; 276(52): 48740 - 48747. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
