| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,
,,,,,,**
*
Neurology Service, and
Research Service, Veterans Affairs Medical Center, Portland, OR 97207;
Department of Neurology, Oregon Health Sciences University, Portland, OR 97201;
§
Virginia Mason Research Center, Seattle, WA 98101 and the Departments of Rheumatology and Immunology, University of Washington, Seattle, WA 98195;
¶
Department of Clinical Pathology, St. Paul’s Hospital, Seoul, Korea;
||
Connetics Corporation, Palo Alto, CA 94303;
#
Council for Tobacco Research, New York, NY 10022; and
**
Department of Molecular Microbiology and Immunology, Oregon Health Sciences University, Portland, OR 97201
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MS is believed to be a T cell-mediated autoimmune disease and therefore might be amenable to TCR peptide therapy (1, 9, 10). The Ag specificities of the disease causing T cells in MS are uncertain, but there is considerable evidence to suggest that MBP is one of the Ag relevant to the pathogenesis of MS. Given the appropriate genetic background, MBP-specific T cells can induce chronic relapsing EAE, which clinically and pathologically mimics MS (11, 12). Most patients with MS have evidence of T cell sensitization to human MBP (13). MS patients have an increased frequency of MBP-specific T cells in their blood and cerebrospinal fluid (14), and their MBP-specific T cells appear to be activated (14, 15, 16). T cells from HLA-DR2+ MS patients usually recognize MBP in association with the HLA-DR2 molecule, which is associated with the disease in North American patients (17, 18). Finally, T cells present in the brains of some MS patients bear a CDR3 homology to that of encephalitogenic MBP-reactive rat T cell clones (19). MBP-specific T cells thus may participate in the pathogenesis of MS and are reasonable candidates to target for anti-TCR therapy.
We previously demonstrated the overutilization of TCR BV5S2 (Vß5.2) among MBP-specific T cell clones from some MS patients (20), a finding that has been more generally confirmed in a review of more than 600 clones reported in more than 20 studies worldwide (21). Subsequently, we showed that low doses of slightly modified synthetic peptides from the CDR2 of BV5S2 and BV6S1 could safely vaccinate patients with progressive MS, inducing significant changes in the frequency of circulating CD4+ TCR peptide-specific T cells in 7 of 11 and 6 of 11 subjects, respectively (22, 23). In contrast, peptide-specific Ab responses were found in only 1 of 11 subjects. Recently, we completed a double blind, placebo-controlled trial using both the germline BV5S2 CDR2 peptide and a Y49T-substituted version of the peptide for vaccination (24). The results of this study demonstrated a decreased T cell response to MBP and possible clinical benefit in patients who responded to vaccination. Moreover, when data from the two clinical studies were combined, we found a highly significant correlation between the degree of T cell response to vaccination and clinical benefit (25). Among DR2+ donors, three of four of the strongest responders to the (Y49T)BV5S2-38–58 peptide experienced clinical improvement, suggesting that the immunogenicity of the peptide for T cells may contribute to its therapeutic effect. In the present study, we compared the immunogenicity and binding properties to DR2 alleles of the BV5S2 CDR2 peptides with other regions of the BV5S2 sequence. We also assessed the immunogenicity of CDR2 peptides from three other BV gene families in two patients who had MBP-specific T cell V gene biases other than BV5S2.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Nine men and two women with clinically or laboratory-supported
definite MS by Poser criteria (26) participated in our study (Table I
). Ten had progressive disease and one
(patient No. 11) had relapsing/remitting MS. MBP-specific T cells from
two patients (No. 1 and 2) previously were shown to utilize BV5S2 and
BV6S1 (21). Five patients (No. 1, 2, 3, 4, and 7) had participated in
the initial trial using BV5S2 and BV6S1 peptides (21), and one patient
(No. 8) had participated in a phase two trial (24).
|
Peptides were synthesized following published deduced amino acid
sequences for the TCR V regions (27, 28, 29) and were numbered using the
system of Kabat et al. (30). Peptides were synthesized by the
Merrifield solid phase technique and purified by HPLC as described
previously (31). Organic solvents (acetonitrile and methanol) were
removed by rotary evaporation. Peptide remaining as solute was frozen
and lyophilized overnight. Lyophilized peptide was dissolved in
lactated Ringer’s solution, and the pH adjusted to 7.0 to 7.5 with
sodium hydroxide. Peptide solutions were filter sterilized, aliquoted
in single-dose vials, frozen at -20°C, and thawed immediately before
use. In this state, peptides retain biologic activity for at least 2
yr. All lots given to patients lacked pyrogenicity as determined by a
commercial laboratory (North American Science Associates, Irvine, CA).
Amino acid sequences for the peptides are given in Table II.
|
Previous experience with TCR peptide immunization indicated that the optimal dose of peptide for most patients was 100 µg and that responders usually had evidence of a T cell response within the first 5 wk of therapy (22). Based on this experience, patients received 100 µg of peptide for each treatment, given as an intradermal injection in the forearm in one site with 0.1 ml of peptide solution at a concentration of 1 mg/ml. Injections were repeated weekly for a total of four injections. In some individuals, a specific peptide was then given every 4 wk at a dose of 100 to 200 µg. Four of the BV5S2 peptides (BV5S2-1–22, -13–33, -25–42, and -33–52) caused acute, self-limited inflammatory reactions at the injection site in some patients. The reactions occurred within a few minutes of injection, typically resolved in 15 to 20 min, and were not associated with any systemic signs or symptoms such as shortness of breath, hypotension, or tachycardia. If a patient developed an immediate-type inflammatory reaction at the injection site of a peptide, no further injections of that peptide were given to that patient.
Patients were injected with one to three of the overlapping BV5S2 peptides at a time. When using more than one peptide, peptides from nonoverlapping regions of the BV5S2 molecule were used. Postimmunization frequencies were obtained for 3 to 52 wk to determine whether there were significant T cell responses to the peptides. Five patients (No. 1, 2, 3, 4, and 7) were immunized with (Y49T)BV5S2-38–58 as part of the initial trial of this peptide and were then subsequently immunized with the remaining BV5S2 peptides (22). For these patients, comparisons of responses to the (Y49T)BV5S2-38–58 peptide were made with the other peptides using frequencies obtained 3 to 12 wk after the initial series of four injections of 100 µg of this peptide.
Determination of Ag-specific T cell frequencies
Limiting dilution assays (LDA) were used to estimate the
circulating frequencies of TCR peptide-specific T cells as previously
described (22). Blood for LDA was obtained two to three times before
injection of each peptide and then immediately before and 1 wk after
each injection for the first four injections. PBMC were separated by
Ficoll gradient centrifugation and subjected to LDA in microtiter
plates. Cell dilutions were 5, 2.5, 1.25, and 0.625 x
105 cells/well. At cell concentrations of 
0.125
x 105, 2 x 105 irradiated autologous
PBMC were added to each well to serve as APC. Ten to twenty-four
replicate wells were cultured at each cell concentration with Ag.
Proliferation after 5 days in vitro was measured by
[3H]thymidine uptake. Individual wells were scored as Ag
responders if the cpm exceeded 2 SD of the mean cpm of 9 to 24 control
wells cultured at the same cell concentration without Ag.
By using the percentage of nonresponding wells at each concentration,
Ag-specific T cell frequencies and their 95% confidence intervals were
estimated by the 
2 minimization method (32), employing a
program adapted for use with a personal computer containing a math
coprocessor chip. This method of analysis gives an estimated frequency
with a 95% confidence interval (mean ± 1.96 SEM). The SEM
reflects both the effects of the cell dilutions of the LDA and the
number of replicate wells used. This method of analysis of LDA, which
gives an estimated frequency with SEM, allows for statistical
comparison of frequencies obtained at different times from the same
subject.
Definition of TCR peptide responder
The frequency of TCR peptide-specific T cells obtained on the
day a patient later received the first peptide injection was used as
the reference "preimmunization" frequency for that peptide.
Patients who had two or more mean postimmunization T cell frequencies
to a TCR peptide that were higher than the 95% confidence level of the
preimmunization frequency, at least one of which was 
2 cells/million
PBMC, were considered "responders" and to have had a "positive"
response to the peptide. Patients whose postimmunization frequencies
did not meet both of these requirements were considered
"nonresponders" and to have had a "negative" response to the
peptide.
Quantitation of relative binding affinities of TCR peptides for DR2 alleles
Binding assays were performed with affinity-purified HLA class II molecules. To isolate DRB1*1501 from DRB5*0101, both dimers were affinity purified from mouse cell lines expressing one of these alleles. DRB1*1501 was obtained from the line L466.1, provided by Dr. Robert Karr (Searle, St. Louis, MO) and DRB5*0101 was purified from L2a.1.4.21, provided by Dr Sandra Rosen-Bronson (Georgetown University Medical Center, Washington, DC). Affinity purification of the class II dimers was performed as previously described by Kwok et al. (33), using L243 as the capture Ab.
Direct binding assays were done by coincubating 25 nM of purified DR molecule with biotinylated peptides in citrate-phosphate buffer at pH5.5 containing 0.75% O.G. (N-octyl-ß-D-glucopyranoside; Sigma, St. Louis, MO), and l mM PMSF. Reactions were performed in 96-well plates for 18 h at 37°C. The samples were then transferred to 96-well plates precoated with L243 and blocked with PBS 5% FCS, the samples were neutralized with 50 ml of Tris, pH 8.0, containing 0.75% O.G. and incubated for 18 h at 4°C. Plates were developed by addition of europium-labeled streptavidin (Delfia, Turku, Finland) for 4 h, followed by enhancement buffer for 1 h. Fluorescence was measured in a Delfia 1232 fluorometer.
Competition assays were performed as described above, but the nonbiotinylated peptides of interest were coincubated at a concentration of 0.001–10.0 µM with Bi-hMBP-84–102 at 0.1 or 0.05 µM. Plates were incubated overnight, transferred to L243-labeled 96-well plates, and developed as in the direct binding assay. Relative binding was calculated by dividing the concentration of Bi-hMBP used in the assay by the IC50 (the concentration of nonbiotinylated peptide at which 50% of Bi-hMBP binding was inhibited).
Biotinylated peptides were synthesized with an Applied Biosystems
432 Peptide Synthesizer (Foster City, CA). For biotinylation, excess
F-moc, 
-amino caproic acid from Novachem (San Diego, CA) in
N,N-dimethylformamide (DMF) was added to the
reaction vessel after deprotection of the N-terminal amino acid
residue. Two caproic acid residues were added and the final coupling
reaction was then conducted with excess biotin in 50% DMSO/50% DMF.
The amino acid sequences of peptides used in this study were:
Bi-hMBP-84–102, NPVVHFFKNIVTPRTPPPS; Bi-(Y49T)BV5S2-38–58,
ALGQGPQFIFQTYEEEERQRG; Bi-BV5S2-38–58, ALGQGPQFIFQYYEEEERQRG.
Selection of MBP- and TCR peptide-specific T cell lines
MBP-specific T cell lines were selected from the blood of patients 7 and 11, and TCR peptide-specific T cell isolates were selected using a modification of previously described methods (34). PBMC were separated by Ficoll density gradient sedimentation, then washed and incubated at 5 x 105 cells/0.2 ml in complete medium (RPMI 1640 plus 2% human AB+ serum, 25 mM HEPES buffer, sodium bicarbonate, sodium pyruvate, glutamine, antibiotics, and antimycotics) with 50 µg/ml of human MBP or TCR BV5S2 peptides for 5 days at 37°C in a humidified 5% CO2 atmosphere. Up to 5 ng/ml of human rIL-2 (R&D Systems, Minneapolis, MN) was then added to cultures for an additional 3 to 5 days to expand MBP- or TCR peptide-reactive T cells. T cells were restimulated with 50 µg/ml of Ag presented by autologous irradiated (4500 rad) PBMC at a ratio of 1:4 (T cells:PBMC) for 3 days and then expanded in rIL-2-containing medium. Cycling between Ag stimulation and rIL-2 expansion was continued until the proliferative response to Ag approximated or exceeded the response to Con A. The specificity of T cell response to Ag was evaluated by incubating 2 x 104 T cells with 2 x 105 irradiated PBMC in 0.2 ml triplicate microtiter wells in the absence or presence of Ag. Assay plates were incubated for 3 days at 37°C in 5% CO2 and were pulsed with 0.5 µCi [3H]TdR for 18 h. Cells were harvested on glass fiber filters and incorporated [3H]TdR was counted using a scintillation counter.
Human MBP was extracted and purified from snap-frozen brain (35) supplied by the National Disease Research Interchange (NDRI Philadelphia, PA). The 84–102 peptide of human MBP was synthesized using the same methods described above for the TCR peptide. The amino acid sequence of the 84–102 peptide given as single-letter codes was NPVVHFFKNIVTPRTPPPS. Some of the overlapping BV5S2 peptides used for establishing the specificity of T cell lines and mapping the CDR2 epitope were obtained from Chiron (Emeryville, CA).
TCR V gene expression
TCR V gene expression of MBP-specific T cell lines was
determined as previously described (23). Total RNA was isolated using a
kit (Stratagene, La Jolla, CA) based on the method of Chomczynski and
Sacchi (36). First-strand cDNA was synthesized in a 50-µl reaction
using AMV reverse transcriptase and a downstream BC-specific primer,
H3Cb3 (CTGCTCAGGCAGTATCTGGAG). The presence of BV-specific PCR
products was determined as follows: 0.1 to 0.2 µl cDNA was used in a
15-µl reaction including 0.35 µM of specific BV primer, 0.35 µM
of H3Cb5 (CTGCTTCTGATGGCTCAAACAC), a BC primer internal to that
used for cDNA synthesis (2–3% of which was radioactively labeled with
[
-32P]ATP), 200 µM dNTPs, and 0.5 U
Taq polymerase in 1x buffer (50 mM KCl, 10 mM Tris-HCl, pH
9.0, 0.1% Triton X-100, 2 mM MgCl2). BV primer sequences
were from the set described by Choi et al. (29) except for BV12 (27),
BV13.1 (27), BV13.2 (27), BV17 (37) and BV21 (38), as previously
published (23). A negative control reaction was run with each sample to
verify that no PCR bands appeared in the absence of BV primer. All BV
primers were checked in control reactions to insure that no PCR bands
appeared in the absence of cDNA. Amplification was conducted for 22 to
26 cycles in a thermocycler (Perkin-Elmer, Norfolk, CT): denaturation
was conducted at 94.5°C x 20 s; annealing at 60°C
x 30 s; extension at 72°C x 60 s. Ten milliliters of
the PCR reaction was run on a 6% polyacrylamide gel which was then
dried for 1 h and exposed to x-ray film overnight. PCR BV products
of the appropriate size were excised from the gel and quantified by
liquid scintillation counting.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A major goal of this study was to evaluate the relative
immunogenicity and characteristics of the BV5S2 CDR2 peptide, residues
38–58, vs other potential epitopes within the BV5S2 sequence.
Moreover, we wished to compare the activity of a slightly modified
germline peptide, (Y49T)BV5S2-38–58, that was strongly immunogenic in
most patients vaccinated in two previous clinical trials, vs the
germline BV5S2-38–58 sequence. Our strategy was to first identify
patients who responded (No. 1–4) vs those who did not respond (No.
5–7) to the Y49T-substituted CDR2 sequence, and then to vaccinate
these patients with overlapping peptides representing the remainder of
the BV5S2 sequence. As in the previous studies, patients were injected
intradermally with 100 µg of synthetic peptide in buffer, weekly for
4 wk and then monthly, and responses were quantified by evaluating
changes in the frequency of Ag-specific T cells from blood drawn before
and during the vaccination protocol. A representative response to
vaccination with the BV5S2-33–52 peptide, but not two other peptides,
is illustrated in Figure 1 for patient 5,
by a significant and consistent increase in the frequency of
peptide-specific T cells beginning 1 wk after the first injection.
|
, three of the four
patients who responded to vaccination with the (Y49T)BV5S2-38–58
peptide did not respond to any additional peptide in the BV5S2
sequence. The fourth (No. 1) responded to only one other region, the
CDR1 peptide encompassed by residues 25–42. Of note, three of these
responders to the (Y49T)BV5S2-38–58 peptide were
DR2+, and one (No. 3) was DR2-. Two of
the three patients shown in Figure 2 who did not respond to initial
immunization with (Y49T)BV5S2-38–58 (No. 5 was DR2+ and
No. 6 DR2-) were responsive to a subsequent course of
vaccination with the native BV5S2-38–58 peptide, as well as to the
contiguous overlapping BV5S2-33–52 peptide (Figs. 1
and 2). The third
patient (No. 7, DR2-), however, did not respond to any of
the overlapping BV5S2 peptides. Two additional patients (not screened
with overlapping BV5S2 peptides), who initially did not respond to
(Y49T)BV5S2-38–58, were responsive to vaccination with other CDR2
peptides, including patient 8, to BV5S2-38–58; and patient 9, to
BV5S2-33–52 (Table III). Patients 10 and
11 were unresponsive to (Y49T)BV5S2-38–58 peptide, and patient 10 also
failed to respond to the native BV5S2-38–58 peptide. Other BV5S2
peptides were not used in these patients, however. These data indicate
that most of the immunogenicity of the BV5S2 molecule resides within
the CDR2 and that significant T cell responses to BV5S2 epitopes can be
induced both in the presence or absence of DR2.
|
|
To compare the rate of vaccination, we compiled results from all
of the patients treated to date (May 1997) with BV5S2 peptides,
including those involved in our previous phase I and phase II trials
and those presented for the first time in Figure 2 and Table III. As
shown in Figure 3, the (Y49T)BV5S2-38–58
peptide was the most widely immunogenic, inducing responses in 12 of 25
(48%) vaccinated patients. The homologous germline BV5S2-38–58
peptide and the overlapping BV5S2-33–52 peptide were slightly less
active but comparable to each other, vaccinating, respectively, 33 and
38% of the patients. In contrast, only one other peptide,
BV5S2-25–42, induced a response in 1 of 7 (14%) patients.
|
, there appeared to be a
bimodal distribution of frequencies among patients responding to
vaccination with the (Y49T)BV5S2-38–58 peptide, with a group of four
strong responders above 20 cells/million and a group of moderate
responders with 2 to 8 cells/million. All of the responders to the
native BV5S2-38–58, BV5S2-33–52, and BV5S2-25–42 peptides had
frequencies in the moderate range. Because of the wide range of
positive values, there were no significant differences in the mean
responses among the different groups. However, as reviewed previously,
we found a highly significant correlation between the strength of
response and a favorable clinical outcome, with three patients
exhibiting the strongest responses experiencing clinical
improvement (25).
|
Binding properties of overlapping BV5S2 peptides for DR2 alleles
HLA-DR2 is expressed in 
60% of Caucasian MS patients and was
present in a majority of patients enrolled in our studies. Since most
of the T cell clones responsive to TCR CDR2 peptides are DR restricted
(34), we sought to determine whether there was a relationship between
the binding properties of the overlapping BV5S2 peptides for DR2
alleles and their immunogenicity in vivo. Table IV shows the binding affinities of each
BV5S2 peptide for the two alleles most commonly coexpressed in
DR2+ MS patients, DRB1*1501 and DRB5*0101, as assessed
by competition with biotinylated MBP peptide (known to bind strongly to
DR2). Of the native BV5S2 peptides tested, the moderately immunogenic
BV5S2-38–58 peptide clearly had the highest binding affinity
(comparable with the MBP-84–102 peptide) for both DR2 alleles. In
contrast, the overlapping BV5S2-33–52 peptide, which was also
moderately immunogenic, had a weak but measurable binding for both DR2
alleles. The BV5S2-25–42 peptide, which induced T cells in only one
DR2 homozygous patient, had a relative strong binding affinity for
DRB1*1501, but no measurable binding for DRB5*0101. The BV5S2-1–22
peptide also had a relatively strong binding to DRB1*1501, but was not
immunogenic in any of the patients tested. Other peptides (BV5S2-58–78
and BV5S2-80–95) bound weakly to DRB5*0101, but were not
immunogenic.
|
, the Y49T-substituted peptide
demonstrated a slightly greater binding to both DR2 alleles than the
native BV5S2-38–58 peptide. Taken together, these data indicate the
order of binding of immunogenic peptides to DR2 as
(Y49T)BV5S2-38–58 > BV5S2-38–58 > BV5S2-25–42 >>
BV5S2-33–52, a result generally consistent with the in vivo
vaccination studies in mostly DR2+ patients.
|
To establish that responses induced by vaccination with
BV5S2-38–58 peptide were specific, T cell lines from two donors (RM
and RT) were selected in vitro, and then tested with overlapping 17-mer
peptides from the BV5S2 sequence. As shown in Table V, T cell lines from donor 1 and donor 5
responded selectively to the BV5S2-36–52 and -41–57 peptides, but
were not reactive against any other overlapping BV5S2 peptide. These
data establish the peptide specificity of BV5S2-38–58-reactive clones
and indicate a lack of cross-reactivity between the vaccinating peptide
and nonhomologous peptides within the BV5S2 sequence.
|
To map the T cell determinants within the immunogenic CDR2
peptides, we selected T cell lines and clones from two responders, No.
1 and No. 5, and tested their proliferation response to a series of
overlapping or truncated peptides from the CDR2. In total, we tested 4
different T cell isolates from patient 1, who responded initially to
vaccination with the (Y49T)BV5S2-38–58 peptide; and 20 different T
cell isolates from patient 5, who responded to vaccination with
BV5S2-38–58 and BV5S2-33–52 peptides, but not to the initial
vaccination with the (Y49T)BV5S2-38–58 peptide. In Table VI, we
present data from four representative isolates, including 5A, blood T
cells used directly from patient 5 without selection in vitro; 5B, a T
cell clone selected in vitro with the overlapping BV5S2-33–52 peptide;
and 5C and 1A, T cell lines selected in vitro with the native
BV5S2-38–58 peptide. It is noteworthy that no matter how the isolates
were selected, all of the T cells tested recognized both the native and
Y49T-substituted BV5S2-38–58 peptides. Although optimal responses
occurred with longer peptides that included most of the residues in the
native 38–58 sequence, significant responses were detected with
peptides that included both Q44 and E52, suggesting a core epitope of
residues 44–52 (QFIFQYYEE). It is noteworthy that the 5C T cell line
also responded to the highly homologous BV5S3-38–58 peptide, which
contains only three differences, including an E52K substitution in the
core region.
|
As shown in Figure 2 and Table III, a few patients failed to
respond to vaccination with any of the BV5S2 peptides tested. To
identify whether CDR2 peptides from V gene families other than BV5S2
might be targeted for vaccination, MBP-specific T cell lines were
selected from patient 7 and a second patient, No. 11, who did not
respond to vaccination with the (Y49T)BV5S2-38–58 peptide. Using the
RT-PCR method for determining BV gene expression, our data suggested
that patient CJ’s MBP-specific T cells overexpressed message for BV9
compared with unselected blood cells (Fig. 6), and the MBP-specific T cells of
patient 11 overexpressed BV3, BV13S1, and BV15 (Fig. 7).
|
|
, vaccinating patient 7 with
BV9-39–59 successfully induced a significant response, boosting the
preimmunization T cell frequency of <0.2 cells/million to a peak T
cell frequency of 2.4 cells/million.
For vaccination of patient 11, whose MBP-specific T cells expressed a
more complex array of BV genes, we prepared two peptides from BV3
(residues 11–30 and 38–58) and a CDR2 peptide from BV12S2 (residues
38–58). BV12S2-38–58 was chosen because of its partial homology in
the CDR2 with BV3, BV13S1, and BV15 that were overexpressed by
MBP-specific T cells (Table VII).
Vaccination with BV3-11–30 did not induce any significant changes in T
cell frequency (Table III). Similarly, vaccination with BV3-38–58 did
not induce a T cell frequency >2 cells/million, although the peak T
cell frequency of 1.7 cell/million was significantly increased over
baseline (Table III). After vaccination with BV12S2-38–58 for 8 wk,
patient 11 developed a significant T cell response with an initial peak
T cell frequency of 3.3 cells/million. Patient 11 continued to receive
BV12S2-38–58 for an additional 8 mo and eventually developed a T cell
frequency of 11.6 cells/million. Upon continued vaccination with the
BV12S2-38–58 peptide, patient 11 eventually developed a significant
response to the BV3-38–58 peptide, even without further vaccination
with that peptide (Table III), suggesting a minor degree of
cross-reactivity between these two partially homologous sequences
(Table VII). Of additional importance, a
patient 11 T cell line responsive to the BV12S2-38–58 peptide (7841
net cpm) did not respond to the BV5S2-38–58 peptide (0 net cpm),
indicating no cross-reactivity between these weakly homologous CDR2
sequences sharing only five identical residues. Interestingly, the
MBP-specific T cell frequency in patient 11 was initially 6.0
cells/million, but fell to 0.3 cells/million after 10 mo of vaccination
with BV12S2-38–58 peptide (not shown), in a manner similar to that
observed in our previous trials (22, 24).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Two aspects related to the immunogenicity of TCR peptides include the rate of vaccination among different patients and the strength of the vaccination response in each individual. Based on extensive evaluations in inbred rodents, we believe that the ability of a given TCR peptide to stimulate a specific T cell response is dependent upon the natural expression of a similar version of the peptide in the context of self MHC I or II molecules on T cells bearing the parent TCR protein (39). Thus, a set of epitopes from the BV5S2 sequence would be presented on the surface of T cells using the BV5S2 molecule in its TCR for Ag, including MBP. The rapid expansion during disease episodes of activated autoreactive T cells expressing BV5S2 or other TCR determinants is then postulated to induce anti-BV-specific regulatory T cells through network interactions.
Although one might expect a predominance of MHC I restricted TCR epitopes facilitated by intracellular processing pathways, we have consistently found TCR peptide-reactive T cells to be CD4+ and restricted by MHC II molecules (34), suggesting that TCR determinants indeed can enter the class II processing pathway. In recent experiments, we have found that the TCR peptide-reactive T cells can be stimulated by autologous MBP-reactive T cells expressing the appropriate BV gene (unpublished data), indicating recognition of class II-restricted TCR peptides on whole cell targets. However, we also anticipate the presence of regulatory TCR peptide-specific MHC I-restricted T cells, which might well be amplified better by MHC class I allele-specific nonamers than by the longer 21-mer peptides that we are currently injecting. Indeed, cytotoxic CD8+ T cells specific for undefined V region determinants can be induced readily in vitro by stimulating blood cells with autologous, irradiated CD4+ T cells (40). Moreover, whole cell vaccination in vivo with irradiated MBP-specific T cell clones can induce cytotoxic CD8+ T cells specific for the injected clonotype, and it is probable that at least some of this response is directed at TCR V gene determinants (41).
Clearly the efficiency of this initial network-priming phase could be influenced by peptide-MHC interactions, as well as by a degree of tolerance that might well be expected to self TCR sequences. Injection of low concentrations (100–300 µg) of soluble peptide intradermally, as per our vaccination protocol, serves to expand the primed T cells using professional tissue APC beyond the natural levels induced by limited T-T interactions. However, successful vaccination with BV5S2 peptides would be expected only in those patients in whom there was a substantial expansion of BV5S2+ T cells specific for MBP or other disease-associated autoantigens that could prime the TCR-specific regulatory T cells. The strength of the vaccination response would further depend on the efficiency of the priming event, a process that would likely depend upon the degree of involvement of BV5S2 in the response to autoantigen. In general, the extent of priming, as assessed by the maximum T cell frequencies attained after vaccination, was not predicted by the initial prevaccination frequency, except for the group of four patients in whom postvaccination frequencies reached >20 T cells/million PBMC. We are currently assessing activation markers on precursor TCR peptide-specific T cells that might better reflect the initial priming events.
Another variable that would be expected to influence the strength of vaccination is the immunogenicity of the peptide itself, a property related to the degree of foreignness and to the ability of the peptide to bind to available MHC molecules. One might expect there to be some degree of tolerance to self TCR sequences, especially since there must be presentation of naturally processed TCR determinants by T cells within the thymus. It is clear that this tolerance is relatively limited, given the ability of self TCR sequences to boost T cell and Ab responses in rodents as well as in some MS patients. However, extended or slightly modified TCR sequences appear to be more immunogenic than germline sequences: Falcioni et al. reported tolerance to a TCR peptide represented in a larger recombinant V gene construct, but T cell responsiveness to a longer version of the peptide that was apparently not cross-reactive with either the larger molecule or the shorter peptide (42). We have reported strongly immunogenic and EAE-protective TCR peptides in rats that do cross-react with a recombinant V gene protein (43), indicating that the ability to circumvent tolerance while maintaining cross-reactivity is a function of the peptide. Moreover, we have observed stronger immunogenicity and better protection against EAE in mice with a highly homologous rat recombinant V gene protein vs the mouse germline recombinant protein (manuscript in preparation).
Taken together, these data suggest that tolerance may indeed exist to peptides that correspond exactly to naturally processed TCR sequences, but this tolerance can be broken relatively easily by even minor changes in sequence or peptide length. These observations in rodents appear to extend to humans, in that both germline and altered TCR peptide sequences are capable of inducing specific T cells (24), with an enhanced ability of the substituted vs the native BV5S2 CDR2 peptide to induce high frequencies of CD4+ Th2 cells that we have associated with clinical benefit in MS (25). Although these CD4+ Th2 cells would also be expected to promote production of anti-TCR Abs, elevated Ab titers to the injected 21-mer TCR peptides were only occasionally observed in vaccinated patients with increased T cell frequencies (22). We did not assess Ab responses in the current study, and it is not known if peptides outside the CDR2 can induce specific Abs.
It is now well established that positive selection of T cells
occurs most efficiently with antigenic peptides having
intermediate binding affinities for MHC molecules (44). Consistent
with this idea, we found that the most immunogenic peptides,
(Y49T)BV5S2-38–58 and native BV5S2-38–58, had the highest relative
binding affinities for both DRA:DRB1*1501 and DRA:DRB5*0101 (Table IV
and Fig. 5), DR2 alleles commonly expressed in our MS patient
population. The T for Y substitution not only retained strong T cell
responses, but also promoted binding to DR2 alleles. This suggests that
position 49 is an MHC II contact residue, and in this orientation one
might predict that residues I46, Y(T)49, E52, and R55 would function as
MHC binding residues for DRA:DRB1*1501, and residues Y(T)49, E52, and
R55 for DRA:DRB5*0101 (45), with either E or K being permissive at
residue 52 (see Table V). These two BV5S2 CDR2 peptides were also able
to vaccinate DR2- patients (No. 3, DR5 and DR7; No. 6, DR3
and DR6), suggesting that they may be promiscuous binders (46). The
BV5S2-33–52 peptide had a much lower affinity for DR2, but was able to
induce moderate T cell responses in three patients primed initially
with the substituted BV5S2-38–58 peptide. Once induced, T cells raised
by vaccination with the BV5S2-33–52 peptide were fully reactive to
both native and substituted BV5S2-38–58 peptides, indicating
recognition of a common epitope, with residues 44–52 constituting an
essential core (Table VI). Other BV5S2 peptides with moderate affinity
for DR2 alleles (residues 1–22 and 25–42) had no detectable or
infrequent immunogenic activity, supporting the dominance of the BV5S2
CDR2 epitope.
Patients 7 and 11, who did not respond to BV5S2 peptides, represent individuals who apparently have little or no involvement of BV5S2 T cells in response to neuroantigens, and thus a low degree of natural priming to BV5S2 determinants. However, MBP-specific T cells expressing other BV genes could be identified, and with this information, both patients were successfully vaccinated with the corresponding (BV9) or cross-reactive (BV12S2) CDR2 peptides. These data have two important implications. First, it would appear that CDR2 peptides may have certain general properties that promote immunogenicity, both as naturally processed determinants displayed on autoreactive T cells and as soluble Ags given therapeutically. One such property may be the consistent amphipathic orientation of amino acids in this region (which predicts immunogenicity; 47 , including a grouping of three hydrophobic residues at positions 45–47, and two more hydrophobic residues at positions 49 and 50, surrounded by solvent soluble residues. The abundance of hydrophobic residues in a variety of spacings may serve to anchor the peptide to different MHC molecules, with the more hydrophilic residues extending into the solvent phase and providing T cell specificity. These properties that contribute to the immunogenicity of CDR2 peptides undoubtedly have functional importance in stabilizing interchain interactions within the intact TCR, as well as providing an external, solvent-exposed loop that has been demonstrated to interact with MHC/peptide.
The second important implication raised by the successful vaccination of patients 7 and 11 with non-BV5S2 CDR2 peptides is that broader efficacy of the vaccination approach will likely require multiple TCR peptide components. In addition to the BV5S2, BV9, and BV12S2 CDR2 peptides described above, we have demonstrated vaccinating activity with a CDR2 peptide from BV6S1 (22) and others with a CDR2 peptide from BV6S5 (48, 49), in total implicating five different antigenic CDR2 peptides in MS patients. Moreover, some of the known antigenic peptides (e.g., human BV5S2-38–58 and BV6S1-38–58) have rather extensive sequence homologies (1), and have been shown to be partially cross-reactive for T cells (22, 23). In rats, we have documented cross-reactivity as well as cross-protection against EAE among BV8S2, BV8S6, and BV6 CDR2 peptides (1). In this study, we confirmed cross-reactivity for T cells between the highly homologous CDR2 sequences in BV5S2 and BV5S3 subfamily members, detected a possible cross-reactivity between partially homologous BV12S2-38–58 and BV3-38–58 peptides in patient 11 and demonstrated the lack of cross-reactivity between the weakly homologous BV5S2 and BV12S2 CDR2 peptides. These data support the idea that some degree of cross-stimulation might well be expected among BV CDR2 peptides that have a relatively high degree of homology. However, a more comprehensive study will be required to establish how such cross-reactivity might influence TCR peptide responses. Such CDR2 homologies are limited among human BV gene families, but could serve to functionally expand the pool of activator and target T cells involved in TCR network regulation.
We believe that the known antigenic TCR peptides may be part of a limited set of V genes that are commonly included in the MS response to MBP, and it is likely that most patients will express one or more of these V genes, with concurrent priming of regulatory T cells. Because TCR peptide-specific Th2 cells release soluble inhibitory factors that may locally affect both target and bystander Th1 cells (24, 32), vaccination with a mixture containing TCR peptides corresponding to even one of the overrepresented V genes or a cross-reactive sequence might provide sufficient regulation to inhibit CNS inflammation by neuroantigen-specific T cells. Additional vaccine components would thus provide a greater probability for therapeutic benefit.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 This work was supported by the Department of Veterans Affairs, Grants NS23221, NS23444, and NS21466 from the National Institutes of Health, the Nancy Davis Center Without Walls, and private donors.
3 Address correspondence and reprint requests to Arthur A. Vandenbark, Neuroimmunology Research R&D-31, Portland VA Medical Center, 3710 SW U.S. Veterans Hospital Road, Portland, OR 97201. E-mail address:
4 Abbreviations used in this paper: EAE, experimental autoimmune encephalomyelitis; LDA, limiting dilution assay; MBP, myelin basic protein; MS, multiple sclerosis; CDR, complementarity-determining region; Bi-hMBP, biotinylated human MBP.
Received for publication October 20, 1997. Accepted for publication March 11, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and Vß chain genes even though the major histocompatibility complex and encephalitogenic determinants being recognized are different. J. Exp. Med. 169:27.8 and
Vß5 in MBP-specific clones. Int. Rev. Immunol. In
press.
This article has been cited by other articles:
|
|
 |
|
  D N Bourdette, E Edmonds, C Smith, J D Bowen, C R. Guttmann, Z P Nagy, J Simon, R Whitham, J Lovera, V Yadav, et al. A highly immunogenic trivalent T cell receptor peptide vaccine for multiple sclerosis Multiple Sclerosis, October 1, 2005; 11(5): 552 - 561. [Abstract] [PDF]
|
|
|
|
 |
|
 E. A. E. van Tienhoven, C. P. M. Broeren, A. Noordzij, J. P. A. Wagenaar, W. van Eden, and M. H. M. Wauben Nasal application of a naturally processed and presented T cell epitope derived from TCR AV11 protects against adjuvant arthritis Int. Immunol., December 1, 2000; 12(12): 1715 - 1721. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R N S Heard, S M Teutsch, B H Bennetts, S D Lee, E M Deane, and G J Stewart Lack of restriction of T cell receptor beta variable gene usage in cerebrospinal fluid lymphocytes in acute optic neuritis J. Neurol. Neurosurg. Psychiatry, November 1, 1999; 67(5): 585 - 590. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
