| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Institut National de la Santé et de la Recherche Médicale, U463, Institute of Biology, and
Department of Pharmacology, College of Pharmacy, Nantes, France
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In a previous study, we have shown that in patients suffering from
rheumatoid arthritis, CD8+ T cells reactive against EBV
proteins were dramatically amplified in synovial fluid compared with
peripheral blood (11). To address the aforementioned question, we
focused our analysis on one patient whose response was predominantly
directed toward a peptide originated from the EBV transactivator BZLF1
presented in the MHC class I B4002 context (12). In fact, this
reactivity could account for most of the observed T cell subset
amplification, as cell lines established after immunosorting with
Vß-specific Abs were reactive against this particular epitope.
Twenty-seven months later, we obtained synovial membrane-infiltrating
lymphocytes from the same patient. We have shown previously by
cytofluorometric analysis that Vß usage among joint-infiltrating T
lymphocytes from this patient was very diverse and stable over time
(11). In the present study, we performed a structural analysis of TCR
- and ß-chains from BZLF1-reactive clones obtained from both
samples to formally document the diversity of the responsive population
and to look for possible shared features between the different TCR
used. Next, we took advantage of this very diverse repertoire to
analyze the contribution of TCR affinities/avidities in the long-term
persistence of the different clonotypes. We demonstrate that despite
significant differences in TCR affinities among the different in vivo
selected clones, these clones displayed very comparable MHC/peptide
density requirements due to a variable contribution of the coreceptor
CD8 in the interaction with the peptide/MHC complex, thus strongly
suggesting that avidity and not affinity is the critical parameter
governing long-term clonotype persistence.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Lymphocytes were isolated from either synovial fluid (first sample) or synovial membrane (second sample) from a 43-yr-old woman who had a typical rheumatoid arthritis lasting for 6 yr at the time of the first sample collection, as previously described (11). Lymphocytes from the first sample were sorted using Vß-specific Abs directed against the subpopulations that were overrepresented in the inflamed joint as compared with peripheral blood (11) and cloned. In the second sample, all CD8-positive lymphocytes from the synovial membrane were sorted out and randomly cloned. Among all of the clones reactive against EBV proteins, only those reactive against the BZLF1 epitope SENDRLRLL presented in the HLA-B4002 were selected for the present study. These clones were maintained in RPMI 1640 10% human serum supplemented with rIL-2 (100 IU/ml) and restimulated every 6 wk under polyclonal activation (irradiated PBL and EBV-transformed B cell lines (BLCL),4 leukoagglutinine (1 µg/ml), and rIL-2), as previously described (13).
Analysis of human T cell clone TCR transcripts
RNA from 5 x 106 T cell clones was extracted
using TRIzol reagent (Life Technologies, Grand Island, NY), according
to the supplier’s instructions, and dissolved in a final volume of 40
µl of water. Reverse transcription was performed on 2.5 µl of the
RNA solution in a final volume of 12.5 µl for 30 min at 45°C in a
mix containing 50 mM Tris-HCl (pH 8.3), 75 mM KCl, 3 mM Mg
Cl2, 10 mM DTT, 10 U rRNasin (Promega, Madison, WI), 1 mM
each dNTP, 100 U Moloney murine leukemia virus (M-MLV) reverse
transcriptase (Life Technologies), and 25 pM C-specific
reverse primer 5'-TGAAGTCCATAGACCTCATGTC-3'. For each clone, five
reverse transcriptions were performed (one for each multiplex PCR).
Each reverse transcription was completed to 50 µl with 1x mix
Taq DNA polymerase (Pharmacia, Piscataway, NJ) (10 mM
Tris-HCl (pH 9), 50 mM KCl, 1.5 mM Mg Cl2), 1.25 U
Taq DNA polymerase, and 25 pM of an equimolar mix of
specific V primers reported in Table I
. Amplification was
performed in a 96-well thermocycler (PTC-100; MJ Research, Cambridge,
MA) with the following cycle: 1x (94°C for 5 min, 45°C for 1.5
min, and 72°C for 1 min); 30x (94°C for 1 min, 45°C for 1.5 min,
and 72°C for 1 min). PCR products were migrated on a 2% agarose gel.
Bands of interest were cut out and incubated overnight in 0.5 ml of
buffered phenol. A total of 50 µl of 10 mM Tris-HCl (pH 7.5) and 1 mM
EDTA was added to each tube, which was then centrifuged at 13,000 rpm
in a microcentrifuge for 1 h. Aqueous phase was recovered,
extracted once with phenol chloroform (1:1), and ethanol precipitated.
DNA pellets were resuspended in 10 µl of 1x Sequenase Rxn buffer
(Amersham, Arlington Heights, IL) with 1 µM of sequencing primer
5'-CTTTGTGACACATTTGTTTGAG-3'.
|
sequence. Analysis of human T cell clone TCR ß transcripts
Reverse transcription was conducted using a CßI reverse primer
(5'-GCAGACAGGACCCCTTGCTGG-3') specific for both Cß1 and Cß2 C
region gene, in the conditions described for C-specific
reverse transcription.
Reverse transcription was completed to 50 µl with 1x mix
Taq DNA polymerase (Pharmacia) (10 mM Tris-HCl (pH 9), 50 mM
KCl, 1.5 mM Mg Cl2); 1.25 U Taq DNA polymerase;
and 12.5 pM of each degenerated primer, VBA,
5'-CAYNRVDMYRTBTMYTGGTA-3', and VBB, 5'-CMYRMHMMYMTKTWYTGGTA-3'. A
first amplification was performed in a 96-well thermocycler (PTC-100;
MJ Research) with the following cycle: 5x (94°C for 1 min, 55°C
for 10 min, and 72°C for 1 min); 5x (94°C for 1 min, 50°C for 10
min, and 72°C for 1 min); 5x (94°C for 1 min, 50°C for 10 min,
and 72°C for 1 min); 5x (94°C for 1 min, 45°C for 2.5 min, and
72°C for 1 min). Since this first PCR resulted in numerous
nonspecific bands due to the use of degenerated primers, a second PCR
was conducted on 0.5 µl of the first PCR completed to 50 µl with
1x Mix Taq DNA polymerase; 12.5 pM of each degenerated
primer, VBA and VBB; and 25 pM of a nested specific Cß primer, CßII
5'-GTGGCCAGGCACACCAGTGTG-3' with the following cycle: 5x (94°C
for 1 min, 60°C for 4 min, and 72°C for 1 min); 35x (94°C for 1
min, 45°C for 2.5 min, and 72°C for 1 min). Bands of interest were
purified and sequenced, as described previously for V sequences.
Cytotoxicity assays
Cytotoxicity was measured in a standard 4-h
51Cr release assay. Briefly, 2 x 106
autologous BLCL were labeled with 50 µCi of 51Cr
(Na251CrO4; Oris Industries, Gif
sur Yvette, France) for 1 h at 37°C, washed five times, and
pulsed with various concentrations of peptide for 1.5 h in RPMI
1640 10% human serum. After two washes, target cells were incubated
with the different T cell clones at an E:T ratio of 10:1 in 150 µl of
RPMI 1640 10% human serum for 4 h at 37°C. Twenty-five
microliters of supernatants were then removed and mixed with 100 µl
of scintillation mixture (Optiphase Supermix, Wallac, U.K.) for liquid
scintillation counting. For CD8 blockage experiments, T cells were
incubated with the anti-CD8 mAb B9.11 (Coultronics, Margency,
France) for 15 min before incubation with the target cells.
ED50 were estimated by fitting the cytotoxicity response
curves with the GraFit software (Erithacus Software, Staines, UK) using
a single binding site model.
For alanine scan analysis, the different clones were tested for cytotoxicity against autologous BLCL loaded with 10, 0.5, or 0.25 µM of each peptide variant (derived from the wild-type BZLF1 9-mer peptide by alanine substitution at each position) at an E:T ratio of 30:1. Results are expressed as percentage of the cytotoxicity obtained with the wild-type peptide in the same conditions.
Cytokine production
The melanoma cell line M88 expressing HLA-B4002 was used as APC in this test because it was easier to exclude M88 cells by size from T cell clones in the following flow-cytometric analysis. M88 cells were pulsed for 1.5 h at 37°C with various concentrations of the BZLF1 peptide in RPMI 1640 medium without FCS and washed twice to avoid autopresentation by T cells. T cell clones (2 x 105) were stimulated by 4 x 105 peptide-pulsed M88 cells in 2 ml of complete medium without IL-2 in the presence of 10 µg/ml of Brefeldin A (Sigma, St. Louis, MO) for 6 h at 37°C in 5% CO2. For intracytoplasmic staining, cells were then fixed for 10 min at room temperature in a solution of PBS 4% paraformaldehyde, washed twice in large volumes of PBS, and stored at 4°C until immunofluorescence analysis.
Flow-cytometric analysis of intracellular cytokines
Fixed stimulated T cell clones were stained for
cytokines using the method described by Jung et al. (37). Briefly,
fixed cells were stained with the different mAbs at a concentration of
5 µg/ml (shown to provide optimal staining for mAbs) for 30 min at
room temperature. Monoclonal FITC-conjugated anti-human IFN-
(4S.B3) and PE-conjugated anti-human IL-2 (MQ1-17H12) Ab were
purchased from PharMingen (San Diego, CA). Reagent dilutions and washes
were done with PBS containing 0.1% BSA (A-9647; Sigma, Saint Quentin
Fallavier, France) and 0.1% saponin (Sigma). After staining, cells
were resuspended in PBS and analyzed on a FACScan flow cytometer using
Cellquest software (Becton Dickinson, Grenoble, France).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ß TCR chains of T cell clones
selected in vivo against a single MHC/peptide complex
To evaluate the diversity and stability of TCR repertoire of T
cells involved in a chronic in vivo response, we performed a structural
analysis of TCR chains from synovial T cell clones derived from a
patient whose response was predominantly directed against a single
viral peptide/MHC complex (BZLF1/HLA-B4002). Since recently published
x-ray crystallographic data of TCR/peptide/MHC demonstrate an
implication of both the - and the ß-chain of the TCR in
peptide/MHC (14, 15, 16, 17), it is crucial to sequence both TCR chains to
analyze structural constraints imposed by recognition of a given HLA
peptide. In the case of clones established after Vß-specific
immunosorting, characterization of CDR3ß was conducted by sequencing
RT-PCR products using specific Vß primers. In the case of clones
randomly isolated from synovial membrane infiltrating lymphocytes, we
developed an RT-PCR approach using degenerated Vß primers. V
transcript amplification was performed by multiplex PCR because absence
of allelic exclusion in the locus prohibits the use of degenerated
primers.
V and Vß sequences of the different clones are presented in Fig. 1. We found that clones from the second
sample expressing BV1S1, BV2S1, BV14S1, and BV22S1 were already present
in the first sample collected 2 yr earlier. The remaining clones whose
Vß had not been analyzed in the first sample (namely BV6S4 and
BV18S1) corresponded to minor subsets for which specific amplification
was nevertheless demonstrated in the first sample by immunoscope
analysis in a previous study (11). Whether this repertoire persistence
results from a process of affinity selection will be discussed below in
light of the functional analysis.
|
, and 11 J represented.
Because of this diversity, it was relevant to look for a selection
process on CDR3 only among clones sharing at least one of these
elements. Among clones using Vß2, two clones, namely A2.3 and A2.8,
shared the same V, J, Vß, and Jß elements. For these two
clones, structural constraints were also apparent on CDR3 regions, with
equal length of both CDR3 and ß. Moreover, these two clones shared
conserved N-encoded residues in the CDR3ß region, suggesting that the
TCR structural constraints in these clones resulted from a stringent
selection process. These CDR3 sequence homologies disappeared when a
different combination of V/Vß was considered, as can be seen with
clone A2.19 (Vß2+), which used different V, J, Jß
than clones A2.3 and A2.8, with clones A13.16 and A13.26, which shared
the same Vß and Jß, but used different -chains, or among the
five Vß22 clones. A similar conclusion could be drawn when
considering clones using the same V, thus suggesting that there was
no obvious predominance of the - or the ß-chain in selecting
reactive clones. Nevertheless, although there was no preferential
association between any particular V and J, a recurrent use of
particular V and J elements was observed: V 28s1 (three
clones); V 25s1, 22s1, and 101s1 (two clones each); J 8 (three
clones); J 28 and 50 (two clones each). In fact, all clones except
one used at least one recurrent V or one recurrent J. This
observation is reminiscent of the overrepresentation of Jß 2s3 and
2s7, regardless of Vß usage that we detected by immunoscope among
B4002/BZLF1 reactive clones in a previous study (11). Thus, the great diversity observed in our panel of clones demonstrates that B4002/BZLF1 is very permissive, i.e., selects a large set of V and J elements, each of which probably engages in different types of contacts with it. In this regard, clones expressing Vß2 and Vß22 that were represented by three and five clonotypes, respectively, were of peculiar interest in that they could be considered as representative of stringent or moderate TCR structural constraints both on combinatorial and junctional diversity. We therefore further investigated their different modes of contact with the selecting HLA/peptide by alanine scan analysis.
Alanine scan analysis
The results of the alanine scan analysis are represented
on Fig. 2. Alanine substitution in
positions 2 and 9, characterized as peptide anchor residues by computer
modelization, abrogated response in the former case, but had little
effect for the latter. This is consistent with the fact that alanine at
position 9 is frequently used as an anchor residue in B4002, as
determined by HLA Peptide Binding Prediction (18). When we tested the
whole panel of clones on peptides substituted at positions 1, 3, 4, 6,
7, and 8, we always observed at least one responsive clone,
indicating that these variants could still bind to B4002 (Fig. 2, and data not shown). In contrast, substitution at position 5, the central
amino acid that has been frequently described as a
crucial contact residue for both CDR3 and ß, abrogated the
reactivity of all clones.
|
-chain in the CDR3 region, thus suggesting
the strong involvement of their CDR3 in contacting the HLA/peptide
complex. Not surprisingly, clone A2.19, which used a different
-chain, showed a distinct pattern of sensivity to substitution at
these positions. However, this clone, which shared Vß2 with A2.3 and
A2.8, but markedly differed in its CDR3ß region, was almost
insensitive to substitution at position 7, thus reinforcing the
critical contribution of CDR3 regions of Vß2 clones in establishing
contact with the HLA/peptide complex. In striking contrast, Vß22
clones displayed a very similar pattern of reactivities despite great
combinatorial and junctional diversity. In particular, it is noteworthy
that they were all very sensitive to alanine substitution at positions
7 and 8, but rather insensitive to substitution at position 6. This
would indicate that contribution of CDR3ß was less critical in clones
Vß22 than in clones Vß2, possibly due to a stronger engagement of
CDR1ß and CDR2ß in the interaction with HLA/peptide. This would be
consistent with crystallographic data of TCR/MCH/peptide complex,
demonstrating variable involvement of CDR1ß and CDR2ß in the
contact (14, 15). Functional analysis of BZLF1-reactive T cell clones
The aim of this analysis was to determine whether we could detect functional differences between the different Vß families of clones, and more specifically between Vß2 and Vß22, which could be related to their different level of structural constraints.
Cytotoxic activity of 11 BZLF1-reactive synovial T cell clones was
evaluated against autologous BLCL loaded with various concentrations of
the relevant 9-mer BZLF1 peptide. As shown on Fig. 3, all clones displayed a very similar
curve of cytotoxicity, except for clone 6.1, which reproducedly showed
a lower reactivity. ED50 were estimated for each clone from
at least three distinct experiments and are reported in Table I
.
ED50 ranged from 9–36 nM, showing that all of these clones
were very similar in terms of peptide/MHC requirements for
cytotoxicity, regardless of their TCR diversity. In a recent report
(19), Schodin et al. demonstrated strong correlations between the
number of MHC/peptide on the target cell required for cytolysis, the
number of interacting TCR complexes, and the affinity of the TCR. Thus,
the avidity of the TCR complex (understood as the sum of TCR and
coreceptor interactions with MHC/peptide) can be estimated by the
density of MHC/peptide required for cytolysis.
|
Numerous studies have demonstrated the stabilizing role of the CD8
coreceptor in TCR/MHC/peptide engagement (20, 21, 22, 23), and evaluation of
CD8 dependency in Ag recognition has been used to indirectly estimate
TCR affinity (24). We therefore investigated the contribution of the
CD8 coreceptor in the recognition of peptide-loaded BLCL by our
different clones to compare their TCR affinity. To this end, we
performed cytotoxicity experiments in the presence of saturating
amounts of anti-CD8 mAb (Fig. 4). For
all but one clone, blocking the CD8 coreceptor resulted in a
significant shift of the cytotoxicity response curve, thus
demonstrating a contribution of CD8-MHC interaction in the
stabilization of the TCR recognition complex.
|
, the level of CD8 dependency was very variable from
one clone to another. Nevertheless, when the ratio between
ED50 in the presence or absence of CD8 mAb was calculated
and clones ranked accordingly (Table I), we noted that CD8 dependency
correlated with Vß usage: most Vß22 clones were weakly CD8
dependent, whereas Vß2 and Vß13 were strongly CD8 dependent. No
significant difference in the level of CD3 or CD8 expression could be
detected by FACS analysis that might have accounted for these
differences in CD8 dependency. Therefore, these variations most likely
reflected differences in TCR affinity, with TCR using Vß2 and Vß13
being of lower affinity for the HLA-B4002/BZLF1 complex than TCRVß6
and most TCRVß22. Interestingly, clones A2.3 and A2.8, which only
differed by 3 aa in their CDR3 and 1 aa in their CDR3ß, showed
marked differences in their CD8 dependency, thus reinforcing the
critical contribution of CDR3s in TCR affinity among Vß2 clones,
whereas clones Vß22 showed rather homogenous CD8 dependency despite
marked differences in their CDR3 sequences. In most cases, a significant cytotoxic response could be restored in the presence of anti-CD8 mAb by increasing peptide-loading concentrations, and the response curves obtained paralleled those in the absence of CD8 blockage. Thus, we show that in most instances, increasing the density of peptide/MHC complexes at the surface of the target cell can compensate for the loss of global avidity due to CD8 blockage, in agreement with previously published data (25). In this regard, clone 22.34 displayed a peculiar pattern of reactivity: its behavior was distinct from that of the other Vß22 T cell clones in that its cytotoxic response curve in the presence of anti-CD8 mAb consistently plateaued at a much lower value than the maximum obtained in the absence of CD8 blockage. This indicates that in the case of clone 22.34, preventing CD8-MHC interaction significantly affected the affinity of its TCR for the MHC/peptide complex, whereas for all of the other Vß22 clones, blockage of CD8 left their TCR affinity intact.
We then asked whether clone 6.1, whose cytotoxic response was almost
unaffected by CD8 blockage, was intrinsically CD8 independent or
whether its CD8 dependency could vary with the affinity of its TCR for
the peptide presented. We therefore tested the cytotoxicity of clone
6.1 against BLCL loaded with the variant peptide A7 (alanine
substitution at position 7 of the BZLF1 9-mer peptide) previously shown
to behave as a weak agonist for this clone. As shown on Fig. 5, the response curve against A7 was
shifted to the right compared with that against the BZLF1 peptide, as
expected for a weak agonist. But the most striking feature was that the
response against A7 was totally abrogated by addition of an
anti-CD8 mAb. Thus, CD8 interaction, which was dispensable for the
recognition of the full agonist BZLF1 peptide by clone 6.1, became
absolutely required when a weaker agonist was presented to this clone.
This clone can thus modulate its CD8 dependency depending on the
affinity of its TCR for the peptide/MHC complex presented. This
observation is in agreement with a recent report from Viola et al. (26)
showing that sensitivity to inhibition by anti-coreceptor Abs was
inversely correlated to the efficiency of TCR-ligand interaction.
|
and IL-2 following stimulation (data
not shown). As shown on Fig. 6, all of
the clones tested were capable of producing IFN- and IL-2 in
response to BZLF1, and although there was some variability in the
percentage of cytokine-producing cells among the different clones,
these variations could not be correlated with Vß usage. When
intensity of fluorescence was plotted instead of percentage of positive
cells, identical results were obtained demonstrating that the amount of
cytokine produced was strictly correlated with the number of responding
cells (not shown). Thus, the efficiency of cytokine production
triggering through the TCR could not distinguish between the different
Vß families.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
- and
ß-chain sequences of B4002/BZLF1-reactive clones that were
specifically amplified in the joint of a rheumatoid arthritis patient.
The repertoire was very diverse, which is consistent with the
recognition of a dominant epitope, as demonstrated in previous reports
(5, 6, 27). In fact, EBV early lytic cycle proteins have been shown to
be frequent targets of CTL in primary and memory responses (28, 29). No
obvious shared structural features could be evidenced among clones
sharing the same Vß, except for some Vß2 clones, in which a
selection process led to structural constraints both at the
combinatorial and junctional diversity level.
We next investigated whether the different stringencies of constraints
among clones were correlated with TCR affinities by evaluating their
CD8 dependency. In this regard, a previous study from Campos-Lima et
al. (7) suggested that stringent structural constraints of the TCR
imposed by recognition of subdominant EBV epitopes were linked to
selection of CD8-independent (i.e., high affinity) clones. Our data do
not support such a strict correlation between structural constraints
and TCR affinity since clones that were tightly constrained, such as
A2.3 and A2.8, turned out to be more CD8 dependent than clones Vß22,
which were loosely constrained. We would rather propose that the use of
a particular Vß has a critical influence on TCR affinity for a given
peptide/MHC complex since, in our model, it was the main parameter that
segregated with CD8 dependency. This would be in agreement with recent
crystallographic data of mouse and human TCR/HLA-peptide complex that
clearly demonstrated that contribution of ß CDRs to the global
binding energy is variable as compared with that of CDRs (14, 15).
In the case of murine 2C TCR bound to H-2Kd-dEV8, only one
residue of the peptide was contacted by a single CDR3ß residue,
whereas CDR1ß and CDR2ß engaged in multiple contacts with peptide
and HLA residues. The opposite situation was seen in human A6 and B7
TCR-contacting HLA A2/Tax that shared the same Vß: in the case of the
A6 TCR, CDR1ß engaged in a single contact with the peptide and none
with the HLA, CDR2ß did not contact the HLA, whereas its CDR3ß
contacted four peptide residues and one HLA residue. In the case of the
B7 TCR, there was no contact between CDR1ß and peptide or HLA, only
two contacts between CDR2ß and HLA, but three peptide residues
contacted by the CDR3 region (30). Therefore, there seemed to be a
balance between the contribution of germline ß CDR1/CDR2 residues and
CDR3ß residues in the stabilization of the TCR/MHC/peptide complex.
We therefore hypothesize that Vß2 has an intrinsically weak ability
to contact the B4002/BZLF1 complex that was compensated in clones
using this Vß by a stringent selection on CDR3 and ß regions,
whereas Vß22 usage conferred a higher basal affinity for the
B4002/BZLF1 complex, thus allowing more degrees of freedom on both
combinatorial and junctional diversity. This hypothesis is further
supported by our alanine scan analysis showing the homogenous
reactivity profiles displayed by Vß22 clones when compared with those
of Vß2 clones.
A striking finding of our study was the narrow range of MHC/peptide densities required to trigger cytolysis in our in vivo selected clones (ED50 between 9 and 36 nM). One explanation would be that a selection took place during in vitro expansion of the clones that resulted in the observed range of reactivities. However, we can formally rule out this hypothesis for the following reasons: first, clones were randomly obtained from CD8- or Vß-sorted joint-infiltrating lymphocytes by polyclonal activation with mitogen, IL-2, and allogeneic feeder cells, and we have documented previously that this procedure does not result in any selection bias, i.e., the panel of recovered clones is representative of that selectively amplified in the joint of the patient (11). Second, clones were further expanded in vitro by rounds of nonspecific (mitogen) stimulations and not by specific restimulations with peptide-loaded APC. In fact, the ED50 of each clone was remarkably constant over time and independent of the number of successive restimulations. Finally, another group from our institute has used the same procedure to obtain tumor-infiltrating lymphocyte clones, and in their case they obtained a panel of clones that varied greatly in the density of MHC/peptide required for cytotoxicity (31). We are therefore confident that the narrow range of reactivities that we observed in our clones resulted from an in vivo selection process and not a culture bias.
This suggests that, in our model, 1) reactive clones were selected in vivo by APC presenting a homogenous density of BZLF1 peptide, and 2) they represented only a fraction of the repertoire potentially reactive against this particular MHC/peptide complex. Moreover, since we have found persisting clones of both weak and high TCR affinity in the two samples taken at a 2-yr interval, we conclude that high TCR affinity per se did not govern persistence of clones in this chronic response, but rather that persisting clones were selected on the basis of their global avidity. Indeed, the homogenous reactivity of all clones, both in terms of cytotoxicity and lymphokine production, could be explained by a differential level of CD8 contribution that compensated for differences in TCR affinities. This homogenous reactivity is not a specific feature of the response against the BZLF1-B4002 complex, since a preliminary analysis of the response against another EBV peptide from the BMLF1 protein presented by HLA-A2 suggests selection of clones with a restricted range of avidities despite variable contribution of CD8 (data not shown). Thus, the critical parameter for long-term selection in vivo being global reactivity to a given number of MHC/peptide complexes, we propose that maximal CD8 engagement (illustrated by clone A13.16 in this study) will define the minimal TCR affinity required to reach the avidity threshold. Once this threshold is reached, an increase in TCR affinity would not lead to stronger activation, and thus clones with higher TCR affinities would have no selective advantage. This would be consistent with the recent report of Busch et al., showing that although reactivation of memory T cells was dependent on appropriate Ag reexposure, the extent and duration of their subsequent in vivo expansion were independent of Ag quantity or stability (32). On the other hand, we were intrigued by the observation of such a variable contribution of CD8 between the different clones. The previous report from Alexander et al. suggested that CD8 dependency was relative and correlated exclusively with the amounts of antigenic determinant densities, i.e., that at low MHC/peptide density, presumed to represent physiologic conditions, all CTL clones needed at least partial CD8 interaction (25). Our data are in discrepancy with this hypothesis since we could find among clones selected in vivo, one clone, clone 6.1, that remained CD8 independent even at low density of MHC/peptide. Moreover, we showed that this clone became very CD8 dependent when confronted to a variant peptide differing by one amino acid substitution in agreement with a previous report (26). We do agree nevertheless that inhibition by anti-CD8 mAb can be reversed by high doses of peptide in CD8-dependent clones. We would thus propose that the level of CD8 contribution is a characteristic of the response of a given clone against a given MHC/peptide complex. Since this contribution resulted in at least a 1000-fold increase in avidity in clones using Vß2 or Vß13, we wondered why such an important contribution was not seen in clones using Vß22 or Vß6, which would have resulted in an even higher avidity. The fact that we could not recover such clones from the joint of our patient suggests that in this in vivo response against EBV, there was a ceiling of maximal avidity above which clones were counterselected. This would be consistent with a recent report from Alexander-Miller et al., showing that proliferation of high avidity cell lines was inhibited by high ligand density and that it could be restored by blocking CD8 interaction (33).
In conclusion, our data suggest that the extent and diversity of the in vivo CTL response are conditioned by two main parameters: 1) the permissivity of the selecting MHC/peptide complex that will determine the size of the potentially reactive repertoire, and 2) the density of MHC/peptide complex presented on the APC in vivo that will define a window of avidity. The coreceptor CD8 by its variable contribution to the stabilization of the TCR/MHC/peptide complex can buffer differences in TCR affinities to maintain clones within this window of avidity, and thus plays a critical role in persistence of the response diversity.
|
|
Footnotes |
|---|
2 C.C. and M.B. contributed equally to this work.
3 Address correspondence and reprint requests to Dr. François Davodeau, Institut de biologie, Institut National de la Santé et de la Recherche Médicale, U463, 9 quai Moncousu, 44093 Nantes Cedex 01, France. E-mail address:
4 Abbreviations used in this paper: BLCL, B lymphoblastoid cell line; CDR, complementarity-determining region.
Received for publication September 28, 1998. Accepted for publication March 10, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ß chains of the human T-cell antigen receptor recognizing HLA-A2 and influenza A matrix peptide. Proc. Natl. Acad. Sci. USA 88:8987.
subset. Eur. J. Immunol. 23:804.[Medline]
ß T cell receptor structure at 2.5 Å and its orientation in the TCR-MHC complex. Science 274:209.ß T cell receptor: mapping the energy of antigen recognition. Immunity 8:413.[Medline]
() and HLA- A2. Nature 387:630.[Medline]
/C) region: organization, sequence, and evolution of 97.6 kb of DNA. Genomics 19:478.[Medline]
This article has been cited by other articles:
|
|
 |
|
  D. Thakral, J. Dobbins, L. Devine, and P. B. Kavathas Differential Expression of the Human CD8{beta} Splice Variants and Regulation of the M-2 Isoform by Ubiquitination J. Immunol., June 1, 2008; 180(11): 7431 - 7442. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Xiao, M. F. Mescher, and S. C. Jameson Detuning CD8 T cells: down-regulation of CD8 expression, tetramer binding, and response during CTL activation J. Exp. Med., October 29, 2007; 204(11): 2667 - 2677. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Laugel, H. A. van den Berg, E. Gostick, D. K. Cole, L. Wooldridge, J. Boulter, A. Milicic, D. A. Price, and A. K. Sewell Different T Cell Receptor Affinity Thresholds and CD8 Coreceptor Dependence Govern Cytotoxic T Lymphocyte Activation and Tetramer Binding Properties J. Biol. Chem., August 17, 2007; 282(33): 23799 - 23810. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. Price, J. M. Brenchley, L. E. Ruff, M. R. Betts, B. J. Hill, M. Roederer, R. A. Koup, S. A. Migueles, E. Gostick, L. Wooldridge, et al. Avidity for antigen shapes clonal dominance in CD8+ T cell populations specific for persistent DNA viruses J. Exp. Med., November 21, 2005; 202(10): 1349 - 1361. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Trautmann, M. Rimbert, K. Echasserieau, X. Saulquin, B. Neveu, J. Dechanet, V. Cerundolo, and M. Bonneville Selection of T Cell Clones Expressing High-Affinity Public TCRs within Human Cytomegalovirus-Specific CD8 T Cell Responses J. Immunol., November 1, 2005; 175(9): 6123 - 6132. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Facchinetti, S. D. Santa, S. Mezzalira, A. Rosato, and G. Biasi A Large Number of T Lymphocytes Recognize Moloney-Murine Leukemia Virus-Induced Antigens, but a Few Mediate Long-Lasting Tumor Immunosurveillance J. Immunol., May 1, 2005; 174(9): 5398 - 5406. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Maile, C. A. Siler, S. E. Kerry, K. E. Midkiff, E. J. Collins, and J. A. Frelinger Peripheral "CD8 Tuning" Dynamically Modulates the Size and Responsiveness of an Antigen-Specific T Cell Pool In Vivo J. Immunol., January 15, 2005; 174(2): 619 - 627. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Couedel, E. Lippert, K. Bernardeau, M. Bonneville, and F. Davodeau Allelic Exclusion at the TCR{delta} Locus and Commitment to {gamma}{delta} Lineage: Different Modalities Apply to Distinct Human {gamma}{delta} Subsets J. Immunol., May 1, 2004; 172(9): 5544 - 5552. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Vely, M.-A. Peyrat, C. Couedel, J.-F. Morcet, F. Halary, F. Davodeau, F. Romagne, E. Scotet, X. Saulquin, E. Houssaint, et al. Regulation of Inhibitory and Activating Killer-Cell Ig-Like Receptor Expression Occurs in T Cells After Termination of TCR Rearrangements J. Immunol., February 15, 2001; 166(4): 2487 - 2494. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
