| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of
*
Microbiology and Immunology and
Cell Biology, Albert Einstein College of Medicine, Bronx, NY 10461
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
- or 
-chain of a VSV8 (unmodified)/H-2Kb-specific
CTL clone as a transgene. Such mice allow independent analysis of one
TCR chain by maintaining the other fixed. The TCR V gene usage of the
responding T cell population was specifically altered depending upon
the presence of the TNP group and its position on the peptide. The CDR3
sequences of the TNP-modified peptide-specific TCRs showed a
preferential J region usage in both the CDR3 and loops,
indicating that the J regions of both CDR3s are critical for
recognition of TNP-modified peptides. In contrast to our previous
observations showing the prime importance of CDR3 residues encoded
by D-segment or N-addition nucleotides for recognition of position 6 of
unmodified VSV8, our studies of TNP-modified peptides demonstrate the
importance of the J region, while the J region was crucial for
recognizing both TNP-modified and unmodified peptides. These data
suggest that different structural strategies are utilized by the
CDR3 and loops to allow interaction with a haptenated
peptide. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
- and -chains,
each containing a V and C domain. The V domain has three
complementarity-determining regions (CDR1, CDR2, and CDR3) that
generally participate in interaction with pMHC. While CDR1 and CDR2 are
germline encoded, CDR3 is derived from genetic recombination events. As
a result of these gene rearrangements, the CDR3 loops have high
diversity and are important for defining TCR specificity for pMHC
(2, 3). Functional (4, 5) as well as
crystallographic studies (6, 7, 8) have demonstrated that
TCRs interact with pMHC in a diagonal orientation (9),
with the - and -chains docking over an area flanking the N
terminus and C terminus of the peptide, respectively. While the CDR1
loop is positioned over the N-terminal end of the peptide, the CDR1
loop is over the C-terminal end of the peptide, with CDR2 and
CDR2 being positioned over the 2 and 1 helices of the MHC
molecule, respectively. The CDR3 and CDR3 loops come in close
apposition, forming a pocket over the central portion of the peptide
capable of accommodating a protruding peptide side chain or presumably
even a bulky hapten group. Glycan- or other hapten-specific T cells play a critical role in certain immune responses (e.g., allergy, contact dermatitis). In some contact hypersensitivity responses, both CD4+ and CD8+ hapten-specific T cells are required (10, 11, 12), with CD8+ T cell-mediated cytotoxicity being mandatory for this phenomenon (13). It has also been reported that T cells specific for certain chemicals (e.g., penicillin, nickel) can be found in peripheral blood of allergic patients (14, 15), while in a collagen-induced arthritis model, a glycan-modified peptide seems to be a dominant CD4+ T cell epitope (16). However, despite the documented ability of T cells to specifically recognize hapten (17, 18) or glycan (19, 20) moieties linked to MHC-bound peptides, the relevant structural features of their TCRs remain ill defined.
The Weltzien group previously established a panel of
H-2b-restricted CTL clones specific for the
trinitrophenyl (TNP) hapten group and found preferential expression of
certain TCR V and J gene elements (21). Upon
derivation of mice transgenic (Tg) for the TCR -chain of one
such TNP-specific CTL clone, they found an increased frequency of
H-2b-restricted TNP-specific CTL precursors, and
TNP-specific lines derived from these mice preferentially expressed
certain TCR V and J gene elements (22). These
studies were critical to the development of the idea that, as for
typical unmodified peptides, recognition of a hapten may also involve
contributions from both the TCR - and -chains. However, since
trinitrobenzenesulfonic acid (TNBS)-modified cells were utilized as Ag
for these studies, the resulting TCRs were likely to represent a
mixture of specificities for TNP-modified peptides in which the TNP
moiety was linked to different positions of diverse peptides, thus
making it difficult to more precisely define the potentially unique
strategies utilized by TCRs that allow recognition of the TNP
hapten.
The CTL response to the vesicular stomatitis virus in C57BL/6 mice is
predominantly directed against a single octapeptide (VSV8; RGYVYQGL)
(23). Our group previously generated TCR single-chain Tg
mice for either the - or -chain of a TCR derived from a
VSV8/H-2Kb-specific CTL clone (24, 42) immunized these mice with VSV8 or singly substituted
variants, and determined the TCR V gene families and CDR3 loop
sequences used by the expanded CTLs. Examination of a large number of
TCR/pMHC interactions in this way enabled us to define the structural
features of TCR - and -chains that are important for in vivo
recognition of VSV8 and its variants (24, 25, 26, 42). To
address the question of how TCRs accommodate haptenated peptides, we
have now immunized these TCR single-chain Tg mice with VSV8 peptides
modified with TNP at either one of two primary TCR contact residues and
examined the TCRs expressed by the expanded CTLs. By using defined
TNP-modified peptides in a system in which the T cell response to the
unmodified peptides is very well characterized, and in which the TCR
- or -chains can be alternately fixed, we were able to identify
distinct structural strategies utilized by the TCR - and -chains
that allow interaction with a haptenated peptide.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Derivation of the N30.7 TCR -chain Tg (Tg) mice was
described previously (24). These mice carry the TCR
-chain of the V2+
V13+ VSV8-specific CTL clone N30.7
(27) as a transgene. They also carry a targeted disruption
of both alleles of the endogenous TCR C locus, and therefore cannot
express their endogenous TCR -chain genes. Derivation of the N30.7
TCR -chain Tg (Tg) mice, in which virtually all peripheral
CD8+ T cells express the V13 transgene on
their surface, will be described in detail elsewhere (42).
C57BL/6 (B6) mice were purchased from The Jackson Laboratory (Bar
Harbor, ME). All mice were maintained in the Institute for Animal
Studies of the Albert Einstein College of Medicine.
Cell line
RMA/s is a TAP-deficient mutant cell line derived from Rauscher leukemia virus-induced C57BL/6 T cell lymphoma, RBL-5 (28). This cell line was maintained in DMEM containing 10% FBS.
Peptides
VSV8 (RGYVYQGL) is the immunodominant peptide derived from vesicular stomatitis virus nucleoprotein (23). All peptides were synthesized by standard solid-phase methods using F-moc chemistry in an automated peptide synthesizer (model 433A; Applied Biosystems, Foster City, CA) at the Peptide Synthesis Facility of the Albert Einstein College of Medicine. Cleavage of the peptide from the resin and removal of the side chain-protecting groups were conducted using trifluoroacetic acid. To prepare peptides modified with a TNP group, we first made F-moc-conjugated VSV8 carrying a lysine substitution at position 4 or 6. These peptides were incubated with TNBS (Pierce, Rockford, IL) to conjugate the TNP moiety to the lysine residue, followed by incubation with piperidine (PerSeptive Biosystems, Hamburg, Germany) to cleave off the N-terminal F-moc. All peptides were purified by reversed-phase HPLC (model HP-1090-M; Hewlett-Packard, Palo Alto, CA) on a Vydac C18 semipreparative column (218TP510; Vydac, Hesperia, CA). The identity of the purified peptides was confirmed by a tandem quadrupole mass spectrometer (TSQ700; Finnigan MAT, San Jose, CA).
In vivo immunization with peptides
Three Tg and three Tg mice were immunized in their hind
footpads with 15 µg peptide emulsified in CFA. One week later, mice
were boosted with 15 µg of the same peptide emulsified in IFA.
Mice were sacrificed and spleens were removed 1 wk following
booster.
Generation and restimulation of CTL in vitro
For CTL generation, spleen cells (5 x 107) obtained from immunized mice were cultured with 1 µM of the immunizing peptide in tissue culture flasks (Falcon 3082; Becton Dickinson, Franklin Lakes, NJ) at 37°C in 9% CO2/air. The culture medium was IMDM (Life Technologies, Gaithersburg, MD) supplemented with 10% heat-inactivated FBS (HyClone, Logan, UT), 2 mM glutamine, 50 U/ml penicillin, 50 µg/ml streptomycin, and 50 µM 2-ME. On day 7, the harvested cells (1 x 105) were restimulated with 50 nM peptide and mitomycin C (Sigma, St. Louis, MO)-treated B6 spleen cells (5 x 106) in complete medium supplemented with 25 U/ml human rIL-2 (Life Technologies) in 24-well culture plates (Falcon 3047; Becton Dickinson). CTL clones were derived from CTL lines by limiting dilution in the presence of 50 nM peptide and mitomycin C-treated B6 spleen cells (1 x 106) in 96-well culture plates (3799; Costar, Cambridge, MA).
Cell-mediated cytotoxicity assay
A conventional 4-h chromium release assay was performed as described previously (25). Briefly, RMA/s cells were labeled with 1.85 MBq Na251CrO4 (Amersham, Arlington Heights, IL) for 1 h at 37°C in 9% CO2/air. They were then washed and used as target cells. For peptide pulsing, 0.1 µM peptide was added to RMA/s cells (2 x 106) during labeling. A total of 3 x 103 target cells (100 µl) was mixed with effector cells (100 µl), and after incubation for 4 h at 37°C in 9% CO2/air, the supernatants (100 µl) were removed and their radioactivity was measured. The percentage of specific lysis was calculated using the following equation: ((a - b)/(c - b)) x 100, in which a is the radioactivity in the supernatant of target cells mixed with effector cells, b is that in the supernatant of target cells incubated alone, and c is that in the supernatant after lysis of target cells with 1% Triton X-100.
Cell sorting
To obtain pure populations of Tg+ CTLs,
peptide-induced CTLs were cultured for 2 wk in vitro and stained with
Abs against CD8 (53-6.7) and either V2 (B20.1; for cultures
derived from Tg mice) or V13 (MR12-3; for cultures derived from
Tg mice), and CD8+ Tg+
cells were collected. All Abs were purchased from PharMingen (San
Diego, CA). Cell sorting of stained cells was performed on a FACStar
(Becton Dickinson).
Determination of the TCR V and V gene family usage
by RT-PCR
mRNA was extracted from sorted cells using the Oligotex Direct
mRNA Kit (Qiagen, Valencia, CA). The mRNA was reverse transcribed into
single-stranded cDNA using Moloney murine leukemia virus reverse
transcriptase and oligo(dT)15 as a primer. For
CTLs derived from Tg mice, V family usage was determined by PCR
using a TCR C primer paired with 1 of 20 V primers, each specific
for a particular V gene family and designed to recognize all known
members of that family (29). For CTLs derived from Tg
mice, V family usage was similarly determined by PCR using a TCR
C primer paired with 1 of 20 V primers (29).
Sequence analysis of CDR3 loops
To determine sequences of TCR CDR3 loops, double-stranded TCR
-chain or -chain cDNA was obtained by PCR amplification of
single-stranded cDNA using Pfu DNA polymerase (Stratagene,
La Jolla, CA) and the appropriate V gene
family-specific primer set. PCR products were purified from gels using
the QIAEX II PCR Purification kit (Qiagen, Valencia, CA), cloned into
pCR2.1 (Invitrogen, Carlsbad, CA), and the TCR CDR3 sequences in
the resulting plasmids were determined at the DNA Sequencing Facility
of the Albert Einstein College of Medicine.
FACS analysis of RMA/s cells pulsed with TNP-modified peptides
RMA/s cells were incubated with 10 µM TNP-modified peptides for 90 min at 37°C in 9% CO2/air. To monitor stabilization of H-2Kb and H-2Db by the TNP-modified peptides, cells were washed and then stained with an Ab against either H-2Kb (AF6-88.5; PharMingen) or H-2Db (28-14-8 purified from culture supernatant), followed by FITC-labeled goat anti-mouse Ig (PharMingen). To determine whether the TNP moiety was solvent exposed, peptide-pulsed cells were washed and then stained with a 1/30 dilution of a rabbit anti-TNP serum (30) (generously provided by H. U. Weltzien, Max-Planck-Institut fur Immunbiologie, Freiburg, Germany), followed by FITC-labeled goat anti-rabbit Ig (PharMingen). FACS analysis was performed on a FACScan (Becton Dickinson).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In previous studies, we utilized TCR single-chain Tg mice
expressing either the TCR - or -chain from a VSV8-specific CTL
clone to define the structural features of TCRs important for
recognition of pMHC (24, 25, 26, 42). In the present
work, we asked whether these Tg and Tg mice might also enable us
to investigate the features of TCRs that allow recognition of
hapten-modified peptides. The TCR contact residues near the center of
VSV8 (positions 4 and 6) (31, 32, 33) were individually
modified with TNP, and the resulting peptides, in which position 4 or 6
was TNP modified, were designated K4-TNP or K6-TNP, respectively (Table I
). The stabilization assay of MHC
molecules using RMA/s cells revealed that these peptides bound well to
H-2Kb, but not to H-2Db
(data not shown).
|
and Tg mice could mount
a CTL response against TNP-modified VSV8 peptides. As shown in Fig. 1, K4-TNP and K6-TNP each elicited good
CTL responses in both Tg and Tg mice. K4-TNP-induced CTLs
recognized K4-TNP, but showed little or no response to K4. Similarly,
K6-TNP-induced CTLs recognized K6-TNP, but not K6. Thus, a strong CTL
response could be induced with TNP-modified VSV8 peptides in TCR
single-chain Tg mice, and the observed cytotoxicity was TNP-specific.
RMA/s cells pulsed with the TNP-modified peptides could be stained
with a polyclonal anti-TNP serum (30), indicating that
the TNP moiety was solvent exposed and thus presumably available for
interaction with the TCR (data not shown).
|
mice (Fig. 2A), nor did K6 in Tg mice
(Fig. 2D), while their TNPylated counterparts readily did
(Fig. 1
, A and D). Because K4 and K6 induced
strong CTL responses in Tg and Tg mice, respectively (Fig. 2, B and C), this lack of CTL generation was not
caused by an inherent immunological defect of these peptides.
|
CTL lines derived from Tg mice immunized with K4-TNP or K6-TNP
showed no cross-reactivity against K6-TNP or K4-TNP, respectively (Fig. 3, C and D). In
contrast, K4-TNP-induced CTL from Tg mice exhibited nearly
equivalent cytotoxicities against both K4-TNP and K6-TNP (Fig. 3A). These patterns of cross-reactivity were confirmed
using CTL clones derived by limiting dilution (Fig. 4). Data for two clones derived from
K4-TNP-immunized Tg mice are shown in Fig. 4A. Clone T4-1
showed equivalent recognition of both TNP-modified
peptides, while T4-2 recognized K4-TNP better, but still reacted
significantly with K6-TNP. A third independently isolated clone, T4-3,
showed a recognition pattern similar to that of T4-2 (data not shown).
In contrast, clones derived from Tg or Tg mice immunized with
K6-TNP showed no reactivity against K4-TNP (Fig. 4, B and
C). A third clone derived from K6-TNP-immunized
Tg mice, T6-3, showed a cross-reactivity pattern indistinguishable
from that of T6-1 and T6-2 (data not shown). Of course, it is possible
that particular clones within the bulk populations might show
cross-reactivity patterns that differ from the population tested at
large and from the clones that we derived by limiting dilution.
|
|
To determine the V gene family usage in responding T cell
populations, we performed an RT-PCR analysis using V or V
family-specific primers. In Table II, we
list families that clearly predominated in cultures from all mice
immunized with a given peptide. Two features of the TCR family usage
were observed. First, the presence of the TNP moiety altered the
specific V gene usage of the expanded T cell population.
K6-TNP-specific CTLs derived from three different Tg mice all used
the V2 and the V8 families, while K6-specific CTLs used the
V13 family. In Tg mice, K4-TNP-specific CTLs commonly utilized
the V18 family, while the K4 peptide did not induce usage of this
particular family. Second, the position of the TNP hapten also appeared
to affect the V gene usage of the responding CTLs. In Tg mice,
K4-TNP-specific CTLs commonly used the V8 family, while
K6-TNP-specific CTLs used the V2 and the V8 families. In Tg
mice, K4-TNP and K6-TNP also induced usage of different V gene families
(V18 and V17, respectively). Thus, the presence of the TNP moiety
and its position in the peptide affected the TCR repertoire of the
responding T cell populations.
|
sequence prevails within the TCR repertoire of CTLs
responding to TNP-modified peptides
To analyze the structural features of CDR3 loops of TCRs
capable of recognizing TNP-modified peptides, we sequenced the CDR3
regions of the several common V gene families utilized by
TNP-specific CTLs obtained from Tg mice. K4-TNP-specific TCRs of the
V8 family showed preferential usage of J1.1 and J2.6
(Table III). Thus, the CDR3 residues just
amino-terminal to the FG motif were conserved (residues VF for J1.1
TCRs and residues EQY for J2.6). The CDR3 loop length was
variable (6–9 residues). For K6-TNP-specific CTLs, we sequenced the
CDR3 loops of V2 and V8 TCRs (Tables IV and V). Similar to
the K4-TNP-specific TCRs, the CDR3 loops of the K6-TNP-specific TCRs
also showed conserved J usage. V2 TCRs used a single J region
(J2.6), and residues just before the FG motif were quite conserved
(EQY; Table IV). Moreover, these CDR3 loops were uniformly quite
short (six residues), likely to provide a cavity sufficiently large to
accommodate the bulky hapten group. Although the TCR CDR3 loops of
V8 TCRs used two different J regions (J2.4 for mouse 1 and 2;
J1.3 for mouse 3), they all showed the same residues (NTLY) just
amino terminal to the FG motif (Table V). Thus, it appears that
residues just N-terminal of the conserved FG motif within the J
region are critical in allowing the TCR -chain to interact with
TNP-modified peptides.
|
|
|
sequence prevails within the TCR repertoire of CTLs
responding to TNP-modified peptides
We next determined the CDR3 sequences of TNP-specific TCRs of
CTLs elicited from Tg mice. For the K4-TNP-specific TCRs, we
sequenced the CDR3 loop for the preferentially expanded V18 family.
These CDR3 loops mainly used the J15 gene segment and had various
lengths (Table VI). For the
K6-TNP-specific TCRs, the CDR3 loops of V17 TCRs showed identical
sequences and a unique J region (J33; Table VII). This J region conservation
underscores the pivotal role that this region of the CDR3 loop plays
in the interaction with TNP-modified peptides. Interestingly, the
K6-TNP-specific TCRs showed longer CDR3 loops (10 aa long) than the
K4-TNP-specific TCRs (largely 8 aa long). This longer length perhaps is
required for facilitating the interaction of the TCR -chain with TNP
at position 6 of VSV8.
|
|
As shown above (Fig. 4), we used limiting dilution to derive seven
independent TNP-specific CTL clones from the CTL lines analyzed in
detail in Tables II–VII. The TCR V gene usage and CDR3 sequences of
these CTL clones are reported in Tables VIII–X.
Most of the clones (those marked with an asterisk in Tables VIII and IX) expressed sequences also identified in bulk populations. Exceptions
were two of the clones derived from K4-TNP-immunized Tg mice (T4-1
and T4-3), both of which expressed V7, rather than V8, the
predominant family identified in all of the CTL lines by RT-PCR and the
family chosen for extensive sequence analysis. Although clone E2,
derived from a K6-TNP-immunized Tg mouse, did not express a TCR
-chain sequence identical with one identified in the CTL lines, it
did express V17, the predominant family identified in bulk cultures
by RT-PCR. Also, while not expressing the predominant
J33, the J22 region expressed by this clone does share the
CDR3 residues QLI with J33.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Direct recognition of the TNP moiety by the TCR is
consistent with our finding that a rabbit anti-TNP serum could
detect the TNP on the surface of RMA/s cells pulsed with the
TNP-modified peptides, thus indicating that the TNP group was solvent
exposed and accessible for direct recognition by TCRs. Although the
crystal structure of the TNP-modified VSV8/H-2Kb
complex has not been solved, molecular modeling suggests projection of
the TNP group above the MHC surface (18). As with TNP,
glycan groups have also been used as haptens in studies of CTL
recognition of chemically modified peptides. Glycan-modified
peptide-specific CTLs can be induced, and recognition is dependent upon
the nature of the glycan (19, 20). The crystal structures
of three different glycan-modified peptides bound to MHC molecules have
all revealed that the glycan moieties protrude from the peptide-binding
groove of the MHC molecule and are available for direct recognition by
TCRs (34, 35).
In our experimental system, hapten-modified peptides were able to
stimulate T cell repertoires that could not be activated by their
unmodified counterparts. Thus, Tg mice were unable to mount a CTL
response to K4 (Fig. 2A), while K4-TNP was strongly
immunogenic in these mice (Fig. 1A). Similarly, Tg mice
did not respond to K6 (Fig. 2D), yet they were able to
respond to K6-TNP (Fig. 1D). These results seem to mimic the
natural situation of hapten-induced hypersensitivity in which
modification of nonimmunogenic self peptides creates new chemical
determinants that can render a self peptide suddenly antigenic.
Our data would suggest that the mobility of the hapten group could also
be a factor increasing the number of TCRs that can respond to a
haptenated peptide. In our previous studies using the N30.7 Tg and
Tg mice as tools to define interaction sites between TCRs and
VSV8/H-2Kb (24, 25, 26, 42) we
found that peptide positions 4 and 6 interact with the TCR - and
-chains, respectively, and that disruption of these critical
interactions leads to nonresponsiveness (24, 26, 42).
Thus, Tg mice are unable to mount a strong CTL response to a wide
variety of position 4 variant peptides (42). Nonetheless,
Tg mice are able to mount a strong CTL response against K4-TNP
(Fig. 1A), and the TCR -chains of the responding CTLs
exhibit preferential V and J gene usage (Tables II and III),
suggesting that in this case, peptide position 4 may be primarily
interacting with the -chain, rather than the -chain. This
observation suggests mobility of the TNP group away from the -chain
and toward the -chain. An analogous situation appears to occur when
Tg mice are immunized with K6-TNP. Tg mice are unable to respond
to most position 6 variants, due to disruption of critical contacts
between position 6 and the Tg-chain (42), but they do
respond to K6-TNP (Fig. 1D). The TCR -chains of the
expanded CTLs exhibit conserved CDR3 sequences (Table VII), suggesting
that the TNP group has shifted away from the -chain and that peptide
position 6 may now be principally interacting with the -chain. The
mobility of the TNP group might allow it to take on different
structures in vivo, thus increasing the number of TCRs able to respond
to a haptenated peptide and perhaps contributing to the increased
immunogenicity of such peptides. Working in a very different system, a
similar idea was recently proposed by the Wilson laboratory
(36). They determined the structure of a rat MHC class I
molecule complexed with an unusually long peptide that bulged out of
the peptide-binding groove centrally. Upon observing that the bulging
residues could adopt different conformations based on their local
environment, they proposed that a greater number of T cells would be
able to respond to such a pMHC complex capable of taking on different
structural identities.
We observed preferential TCR V gene usage among TNP-specific CTLs that
was dependent on the presence of the TNP group (Table II). We
previously showed dramatic alteration of the V gene usage among
peptide-induced CTLs elicited from Tg mice in response to single
substitutions of VSV8 at position 6 (26), leading us to
hypothesize that there might be direct contact between CDR1 and
position 6 of VSV8. Direct interaction between residues in the
C-terminal half of the peptide and the CDR1 loop has also been seen
in crystal structure analyses of TCR/pMHC complexes
(6, 7, 8). While a crystal structure of a TNP-modified
peptide/MHC complex is not currently available, the crystal structures
of glycan-modified peptide/MHC structures showed no conformational
change in the peptide or the MHC molecule due to the peptide
glycosylation (34, 35). Thus, a conformational change of
the MHC molecule at the TCR-binding interface due to the TNP
modification of the peptide, resulting in alteration of V gene usage,
seems unlikely. Instead, we favor the hypothesis that the TNP group
directly contacts the germline-encoded CDR1 or CDR2 loops of the TCRs
and thus alters the V gene family usage of the responding CTLs. In
support of this notion, Luescher et al. (37, 38, 39)
identified TCR/hapten contact sites within the germline-encoded CDR1
or CDR2 regions of TCRs specific for a photoreactive hapten
(4-azidobenzoic acid) conjugated onto the TCR contact residues of the
Plasmodium berghei circumsporozoite peptide PbCS
252–260 (SYIPSAEKI).
If the TCR repertoire of unimmunized Tg and Tg mice were skewed,
this might also account, at least in part, for the preferential V gene
usage observed in response to the TNPylated peptides (Table II). In the
case of Tg mice, we know that this is not the case, as our previous
FACS analysis indicated that the V repertoire of unimmunized Tg
mice is virtually indistinguishable from that of non-Tg mice of the
same background (26). While the unavailability of suitable
reagents precludes FACS analysis of the V repertoire of Tg mice,
the families expanded by TNPylated peptides (i.e., V17 and V18)
are not expanded in response to VSV8 or any of the other position 4 or
position 6 variants that we have tested (42), thus
suggesting they are specifically expanded in response to the haptenated
peptides.
The analysis of the CDR3 sequences of TNP-modified peptide-specific
TCRs showed strong conservation of J and J usage (Tables III–VII). When CDR3 loops of TCRs from unimmunized Tg mice were
sequenced, no predominant J usage was observed, and the occurrence
of J15 and J33 (i.e., those J regions preferentially utilized
by TNP-specific CTLs) was 1/22 and 0/22, respectively
(42). There also does not appear to be intrinsic deviation
of the J usage in Tg mice, as a variety of J gene segments are
utilized in response to VSV8 and its singly substituted variants
(24). Thus, the preferential J region usage observed in
response to K4-TNP and K6-TNP is most probably the result of selection
by the peptide stimulation, suggesting that J and J regions are
critical for enabling the CDR3 and loops to interact with the
TNP hapten moiety. Previous results, in which TNBS-modified cells were
used as an Ag rather than defined TNP-modified peptides, are consistent
with this idea. Kempkes et al. (21) found that 4 of 11 of
their TNP-specific CTL clones utilized J2.6. Iglesias et al.
(22) elicited CTL lines from TCR -chain Tg mice
carrying a transgene derived from a CTL clone specific for
TNBS-modified cells and found preferential usage of J33 (referred to
as JBBM142 in Ref. 22). Although those TCRs might be
specific for the TNP group attached to different positions of various
peptides, we also found preferential uses of these same J regions in
certain cases (Tables III, IV, and VII). Furthermore, by using
photoreactive haptenated peptides, Luescher et al. (37)
identified one J residue (position 97) of the CDR3 loop of the
S14 TCR as a contact residue for the hapten 4-azidobenzoic acid. Thus,
the J regions of the CDR3s may be generally pivotal for enabling TCRs
to interact with hapten moieties.
To evaluate whether different structural strategies might be utilized
by the TCR CDR3 loops for recognition of TNP-modified vs unmodified
peptides, we compared the CDR3 sequences of TCRs specific for
TNP-modified VSV8 peptides with those we previously determined for TCRs
specific for unmodified peptides (24, 42) CDR3
sequence comparison of the K6-TNP-specific TCRs with those of K6- or
other position 6 variant-specific TCRs derived from Tg mice showed a
preferential J usage only in response to the TNP-modified peptide
(Tables IV and V). No such preferential J usage was seen for the
TCRs responding to K6 and most other unmodified position 6 variants
(24). Instead, residue 98 of CDR3, generally encoded by
D-segment and/or N-addition nucleotides, was highly conserved among
CTLs specific for a certain position 6 variant, indicating that the
residue at position 98 of the CDR3 loop is critical for the TCR
-chain to recognize the residue at position 6 of unmodified VSV8. In
contrast, the TNP-specific TCRs examined in this study showed
conservation of J usage, but not necessarily conservation of the
residue at position 98 (Tables III–V). Only the K6-TNP-specific CTLs
of the V2 family appeared to have a conserved amino acid at position
98 of the CDR3 loop (Glu; Table IV). However, in these unusually
short CDR3 loops, this amino acid was encoded by the J gene
segment and not selected by random modification at the V-D or D-J
junction as occurs in the case of the CDR3 loops of unmodified
peptide-specific TCRs (24), and other J regions that
also have Glu at the same position as in J2.6 were not selected by
K6-TNP. Thus, we consider that conservation of the residue at position
98 in this case was a consequence of the selection of a specific J
region and CDR3 loops of a uniform, short length. Selection for a
particular J, rather than a single residue, suggests the possibility
of more numerous contacts between CDR3 and a TNPylated residue, as
compared with an unmodified one.
A different picture resulted from comparison of the CDR3 sequences
of the K4-TNP-specific TCRs (Table IV) with those of K4- or other
position 4 variant-specific TCRs derived from Tg mice
(42). For TCRs responding to the unmodified peptides, we
recently found that the J usage was highly conserved among TCRs
responding to a particular position 4 variant, and that the J usage
changed in response to certain substitutions at position 4 of the
peptide. Similarly, the K4-TNP-specific CTLs also preferentially
utilized a unique J gene segment.
As discussed, while the J region appears crucial for interactions
with TNP-modified VSV8, but not unmodified variants, the J region is
important for recognition of both the modified and unmodified peptides.
Thus, the structural strategies utilized by the TCR CDR3 loops to
interact with a ligand seem to be more conserved for the TCR -chain
than for the -chain. Structural studies provide an explanation for
why this might be so. The crystal structures of complexes of TCRs with
peptide/self MHC class I (but not with peptide/allo-MHC class I) have
shown that the TCR -chain dominates the interaction between the TCR
and pMHC (6, 7, 8, 40, 41). Since the TCR -chain appears
to play an important role for docking the TCR onto pMHC, the structural
features of the TCR -chain should be more conserved than that of the
-chain. Indeed, three crystal structures of TCR/peptide/self MHC
class I complexes indicate that the footprints of the CDR1, 2,
and 3 loops of the TCRs on pMHC are in almost identical positions,
while the CDR1, 2, and 3 loops differ substantially
(6, 7, 8, 41). We believe the greater flexibility in the
positioning of the TCR -chain is also reflected in the
cross-reactivity profiles of the TNP-specific CTLs investigated in this
study. K4-TNP-specific CTLs derived from Tg mice show
cross-reactivity to K6-TNP (Figs. 3A and 4A). As
discussed, the -chain is believed to interact with the TNP group in
the case of CTLs derived from Tg mice. The cross-reactivity
indicates that the -chain can adjust to interact with the TNP group,
at least to some degree, regardless of whether it is at position 4 or
position 6. (The mobility of the TNP group, discussed above, probably
also facilitates this cross-reaction.) In contrast, TNP-specific CTLs
derived from Tg mice, in which the -chain interacts with the
hapten, do not show such cross-reaction (Fig. 3, C and
D).
We previously derived clones from Tg mice specific for VSV8
(unmodified)/H-2Kb (25). We now have
TNP-modified VSV8-specific clones derived from these same mice (Fig. 4, A and B). Thus, both sets of clones express the
same Tg-chain. Structural studies utilizing these reagents should
provide complementary information to that obtained from the
immunobiological data described in this work regarding the unique
strategies utilized by TCR - and -chains that enable them to
interact with nonpeptide components of MHC class I-presented
peptides.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: Department of Immunology, Tsukuba University, 1-1-1 Tennodai, Tsukuba City, Ibaraki, Japan, 305-8575.
3 Current address: Department of Molecular Genetics, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461.
4 Current address: Laboratory of Molecular Genetics and Immunology, Rockefeller University, 1230 York Avenue, Box 98, New York, NY 10021.
5 Current address: Infrastructure Services, Department of Engineering, InfoSpace, Bellevue, WA 98004.
6 Address correspondence and reprint requests to Dr. Stanley G. Nathenson, Department of Microbiology and Immunology, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461. E-mail address: nathenso{at}aecom.yu.edu
7 Abbreviations used in this paper: pMHC, peptide-MHC; CDR, complementarity-determining region; Tg, transgenic; TNBS, 2,4,6-trinitrobenzenesulfonic acid; TNP, 2,4,6-trinitrophenyl; VSV, vesicular stomatitis virus.
Received for publication June 13, 2001. Accepted for publication August 14, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T cell receptors. Annu. Rev. Immunol. 16:523.[Medline]
-producing (Tc1) effector CD8+ T cells and interleukin (IL) 4/IL-10-producing (Th2) negative regulatory CD4+ T cells. J. Exp. Med. 183:1001. chain gene from an H-2b-restricted, TNP-specific cytolytic T cell clone. Eur. J. Immunol. 22:335.[Medline]
chain. Proc. Natl. Acad. Sci. USA 95:5217. repertoire of the responding CD8+ cytotoxic lymphocyte population. J. Immunol. 162:7263. chain gene rearrangement. Proc. Natl. Acad. Sci. USA 95:12538. and chains in ligand binding among H-2Kd-restricted cytolytic T lymphocyte clones specific for a photoreactive Plasmodium berghei circumsporozoite peptide derivative. J. Biol. Chem. 272:8505. and an antigenic peptide bound to MHC class
I. J. Immunol. In
press.This article has been cited by other articles:
|
|
 |
|
  A. Schneeberger, P. Luhrs, R. Kutil, P. Steinlein, H. Schild, W. Schmidt, and G. Stingl Granulocyte-Macrophage Colony-Stimulating Factor-Based Melanoma Cell Vaccines Immunize Syngeneic and Allogeneic Recipients via Host Dendritic Cells J. Immunol., November 15, 2003; 171(10): 5180 - 5187. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Zhang, S. Honda, F. Wang, T. P. DiLorenzo, A. M. Kalergis, D. A. Ostrov, and S. G. Nathenson Immunobiological Analysis of TCR Single-Chain Transgenic Mice Reveals New Possibilities for Interaction between CDR3{alpha} and an Antigenic Peptide Bound to MHC Class I J. Immunol., October 15, 2001; 167(8): 4396 - 4404. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
