| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Immunology, Division of Medicine, Imperial College of Science, Technology, and Medicine, Hammersmith Hospital, Du Cane Road, London, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and H2-Ek cDNAs were
generated with mutations in the sequences encoding the interface
regions of the MHC class II double dimer. Stable DAP.3 transfectants
expressing these cDNAs were generated and characterized biochemically
and functionally. Substitutions in either interface region I or III did
not affect T cell activation, whereas combinations of amino acid
substitutions in both regions led to substantial inhibition of
proliferation or IL-2 secretion by human and murine T cells. Because
the amino acid-substituted molecules were serologically
indistinguishable from wild type, bound antigenic peptide with equal
efficiency, and induced Ag-dependent CD25 expression indicating TCR
recognition, the reduced ability of the mutants to induce full T cell
activation is most likely the result of impaired double dimer
formation. These data suggest that MHC class II molecules, due to their
structural properties, actively contribute to TCR
cross-linking. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MHC class II molecules are heterodimers consisting of an 
- and a
-chain that bind peptides for presentation to
CD4+ Th lymphocytes (5, 6). Class II
heterodimers bind peptides in endosomal/lysosomal compartments, and,
following peptide binding, the complexes are transported to the cell
surface. T cell activation occurs when TCRs of the appropriate
specificity bind to the membrane-distal domains of the MHC molecule and
its bound peptide, and CD4 molecules bind to sites on the
membrane-proximal domains (7, 8). The crystal structures
of MHC class II molecules from both human and mouse (9, 10) revealed that the molecules were packed as double dimers,
and in such a way that two TCR and two CD4 molecules could bind
simultaneously. However, MHC class I molecules have not been
shown to crystallize as double dimers. A study using mutational
analyses of the mouse MHC class II molecule,
H2-Ek, has yielded data that are consistent with
the possibility that MHC class II molecules form double dimers
(11).
Whether double dimers exist as preformed complexes at the cell surface of the APC or whether the dimerization/oligomerization is induced by ligation of the TCR is not yet clear. High-molecular-mass bands containing MHC class II molecules in Western blots of murine and human cell lysates have been interpreted as preformed MHC class II double dimers (12, 13, 14, 15). Furthermore, using biophysical techniques, preformed double dimers have been described in MHC class II-transfected mouse fibroblasts (16). However, the conclusion from experiments involving soluble MHC class II and TCR molecules was that dimerization/oligomerization was driven by cognate interactions between the TCR and MHC class II:peptide complexes (4).
There are three contact areas between the HLA-DR1 molecules in the
double dimers seen in the crystal structure (Ref. 9 , and
see Ref. 17 for review). Interface regions I and II are
located in the membrane-distal part of the molecules, and interface
region III is located in the membrane-proximal domains. Amino acids
49–55 in the 1 domain constitute interface region I, and residues
88 and 111 of the 1 domains are part of interface region II. Amino
acids from both the - (E158, H177, and E179) and the -chain
(H111, H112, and E162) contribute to interface region III.
To address the importance of MHC class II dimerization in T cell
activation, we have generated mutations in HLA-DR1 and
H2-Ek -chain-encoding cDNA clones at positions
encoding for amino acids in interface regions I and III of the MHC
class II molecule. The altered MHC class II molecules were
characterized biochemically and functionally.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All reagents, if not otherwise stated, were purchased from Sigma (St. Louis, MO).
Mutagenesis of HLA-DRB1*0101 cDNA and construction of tagged HLA-DRA
Mutagenesis of the HLA-DRB1*0101 cDNA was performed
using a cassette-cloning strategy, overlap-extension PCR, or a
combination of the two, as previously described (18, 19).
In brief, two cloning sites were introduced in the cDNA by PCR,
flanking the region corresponding to interface region I in the 1
domain to facilitate the cassette-cloning strategy. The restriction
sites AatII and BglII were chosen because these
sites were absent in the HLA-DRB1*0101 cDNA, and
introduction of these sites did not alter the translated amino acid
sequence. Synthesized oligonucleotides (Department of Virology,
Hammersmith Hospital) with sequences corresponding to wild-type
(wt)2 and mutant amino
acid sequences (Table I
) were annealed to create dsDNA for ligation
into the AatII- and BglII-cut HLA-DR1B1 cDNA. PCR
mutagenesis using splicing by overlap extension was used to create the
cDNAs with mutations in interface region III. The final constructs were
all cloned into the expression vector plasmid cytomegelovirus U (pCMU)
(20). The procedure of epitope tagging of cDNA constructs
has previously been described (21). Briefly, the HLA-DRA
fusion constructs, HLA-DRA-hemagglutinin (HA) and HLA-DRA 6 histidine
(6HIS), were made by ligating synthesized oligonucleotides with
sequences encoding a HA epitope recognized by the mAb 12CA5 and 6HIS,
respectively, at the 3' end of the HLA-DRA cDNA. The annealed oligos
were ligated between XbaI and KpnI sites
previously introduced by overlap PCR. The final constructs contained
the sequences encoding the tags after the cytoplasmic coding sequence,
followed by the stop codon. All oligonucleotide sequences and PCR
amplification conditions are available upon request. To confirm the
fidelity of PCR amplification and the cassette-cloning procedure,
nucleotide substitutions were confirmed by dsDNA sequencing (Department
of Virology, Hammersmith Hospital). In addition to the wt and mutant
HLA-DR1B, all other cDNAs (HLA-DMA, HLA-DMB, HLA-DRA, and human
invariant chain (Ii) p31) were subcloned into the expression vector
pCMU. The HLA-DRA, HLA-DRB1, and human Ii p31 cDNA clones were provided
by Dr. L. Karlsson (R.W. Johnson Pharmaceuticals Research Institute,
San Diego, CA). HLA-DMA and -B were kindly provided by Prof. J.
Trowsdale (Department of Pathology, Cambridge University, Cambridge,
U.K.).
|
-chain
and the one-letter code of the amino acid present in the novel molecule
in parentheses, e.g., HLA-DR(51A). Generation of a panel of transfectants expressing wt and mutant MHC class II molecules
Transfections were conducted as described (20),
except that 25 µg of DNA and 5 x 105
cells were used per transfection. Each transfection was performed in a
100-mm Petri dish. For Figs. 1 and 2, HLA-DRA, HLA-DRB1 (or mutants),
HLA-DMA, HLA-DMB, and human Ii p31 were transfected in molar ratios
6:6:2:2:9. For Fig. 3, 12 µg each of HLA-DRA and HLA-DRB1 were used
together. For Fig. 8, 12 µg each of H2-EA and
H2-Ebk cDNA were used together with 1 µg of the
neomycin-resistant gene. Empty vector DNA was used in mock
transfections.
|
|
|
|
Transfected cells were detached from the Petri dishes with 2 mM EDTA in PBS 72 h after transfection. For Ni-Agarose precipitation, the cells were spun down and lysed in 1 ml of 1% Nonidet P-40 in TBS (pH 7.4), PMSF (1 mM), and Trasylol (Sigma-Aldrich, Dorset, U.K.) (1:1000). DNA and cell debris were removed by high-speed centrifugation, and the cell lysates were than precleared two times for 1 h at +4°C with Sepharose CL4B (Pharmacia Biotech, Uppsala, Sweden). The lysates were then mixed with 25 µl of a 50% slurry of Ni2+-Agarose (Clontech, Hampshire, U.K.) and incubated for 4 h with end-over-end rotation at +4°C. The lysates were washed three times in lysis buffer and once with lysis buffer with 15 mM Imidazole (Sigma-Aldrich). The bound material was eluted with lysis buffer containing 120 mM Imidazole. The eluted material was mixed with an equal amount of SDS sample buffer containing 4% SDS and boiled for 5 min before SDS-PAGE (see below). In experiments in which MHC class II SDS-stable dimers and high-molecular-mass complexes were studied, the cells were lysed in 1% Nonidet P-40 in PBS, nuclei were spun down, and the supernatants were directly diluted with an equal amount of SDS sample buffer containing 4% SDS. Samples were split into two portions. One portion was boiled for 5 min, and the other was left at room temperature. Samples were separated on 10% polyacrylamide gels and blotted onto Immobilon P membranes (Millipore, Bedford, MA). Blots were prehybridized with 3% defatted skim milk in PBS and hybridized with the mouse mAbs TAL.1B5 and L243, or 12CA5 in the case of the Ni-Agarose-precipitated samples. The bound Ab was detected using HRP-labeled sheep anti-mouse Ab (Amersham, Arlington Heights, IL) together with ECL (Amersham).
Cells and mAbs
HeLa cells were grown in DMEM (Life Technologies, Rockville, MD) supplemented with 10% FBS, 2 mM glutamine, 100 U/ml streptomycin, and 100 U/ml penicillin at 37°C in an atmosphere of 5% CO2, 95% air and passaged every other day.
DAP.B7 cells were generated by transfecting a subclone of the fibroblastoid L cell line, DAP.3, with a human CD86 cDNA together with a hygromycin-resistant gene as previously described (22). The cells were cultured as above in medium supplemented with 200 U/ml hygromycin.
The human T cell clones HC6 and NF4, which are HLA-DR1 restricted and specific for influenza HA peptide HA307–319, were generated and maintained as previously described (23). They were used in proliferation assays at least 1 wk after their last Ag stimulation. The mouse T cell hybridoma, 2B4, specific for pigeon cytochrome c and restricted by H2-Ek, was passaged twice weekly.
The B cell hybridomas secreting MHC class II-specific Abs L243 and 14.4.4S were obtained from American Type Culture Collection (ATCC; Manassas, VA) TalB.1.5 has been described previously (24). The mAb 12CA5 (Boehringer Mannheim, Indianapolis, IN) was used to detect the HA-tagged constructs.
Functional assays of T cell activation
For assays using human T cell clones, stimulator cells were prepulsed overnight in the presence of a titration of the HA peptide before being treated with mitomycin-C (50 µg/ml; Kyowa Hakko Kogyo, Tokyo, Japan) for 45 min. The cells were washed three times in tissue culture medium before being added to 96-well plates at a concentration of 3 x 104 cells/well. The responder T cell clones (1 x 104 cells/well) were cocultured with the APC in a total volume of 200 µl tissue culture medium and incubated for 48 h at 37°C in a 5% CO2 atmosphere. Cultures were pulsed with 1 µCi [3H]thymidine (ICN Biomedicals, Oxfordshire, U.K.) per well and incubated for 18 h. Cells were harvested onto glass fiber filter mats (Wallac, Turku, Finland) using an automated cell harvester (Skatron, Sterling, VA). T cell proliferation was measured as [3H]thymidine incorporation as determined by liquid scintillation spectroscopy using a Wallac Betaplate liquid scintillation counter. Results were expressed as the total mean cpm for triplicate samples.
When the 2B4 hybridoma was used, stimulator cells were prepared as described above and plated out at 2 x 104 cells/well. T cells were added at 2 x 104 cells/well. Culture supernatants (100 µl per well) were removed after 48 h and transferred into the wells of a 96-well plate that was frozen and thawed to eliminate any live cells. CTLL2 cells (3 x 103 cells/well) were then added. Cultures were pulsed with [3H]thymidine and harvested 8 h later for counting as above.
Peptide binding assay
Cells were plated in a 24-well plate at 3 x 105 cells/well and left to adhere to the plastic for 4 h. The cells were then incubated for 16 h with biotinylated HA306–319 peptide at the desired concentration. The cells were trypsinized and washed once in ice-cold tissue culture medium (all subsequent incubations were performed on ice) and twice in FACS buffer (PBS, 1% FBS, and 0.01% NaN3). The cells were then incubated for 30 min with FITC-avidin (10 µg/ml; Vector Laboratories, Peterborough, U.K.) in FACS buffer. The cells were washed three times with FACS buffer followed by 30 min incubation in biotinylated anti-avidin (10 µg/ml; Vector Laboratories), then washed again three times in FACS buffer followed by a final incubation with FITC-avidin (10 µg/ml) and three additional washes. Finally, the cells were analyzed by flow cytometry.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To evaluate the importance of MHC class II double dimer formation
for T cell activation, we generated a set of HLA-DRB*1 cDNAs with
mutations at positions encoding residues predicted from the crystal
structure of HLA-DR1 to be important for double dimer formation (Table I
). Amino acids from the HLA-DR
polypeptide are involved in two of the three interface regions between
the two heterodimers in the putative double dimer (9, 17). To determine whether the mutants were expressed and
correctly folded, we performed transient transfections into HeLa cells
of the mutated cDNAs together with HLA-DRA, HLA-DMA, HLA-DMB, and
Ii p31 cDNAs. The transient transfectants were analyzed for HLA-DR
expression by flow cytometry after staining the cells with the mouse
anti-DR mAb L243 and FITC-conjugated goat anti-mouse
antiserum, both at saturating concentrations. The amino
acid-substituted HLA-DR molecules were expressed at the cell surface at
similar levels, as shown in Fig. 1. Flow
cytometric histograms of the mutants showed that a >90% transfection
efficiency was achieved, and the histogram shapes of the HLA-DR wt
transfectant and the mutants were similar (data not shown). Similar
results were obtained when the cells were stained with three other
mAbs, DA6.231, 7.4.1, and CA2 (data not shown). One of the mutants, in
which a codon for N at position 49 was introduced to create a
glycosylation site in the dimer interface, was not expressed and will
not be discussed further. We concluded from these experiments that the
amino acid-substituted -chains were able to fold correctly and were
transported to the cell surface.
To determine whether amino acid substitutions in the putative double
dimer interface prevented the action of HLA-DM and Ii, accessory
molecules needed for peptide loading of HLA-DR1 (25),
total cell lysates of the transfected HeLa cells were subjected to SDS-
PAGE, Western-blotted, and analyzed for the presence of SDS-stable
HL-DR dimers. Formation of HLA-DR1 SDS-stable dimers has been
shown to be dependent on the expression of HLA-DM and Ii and to
indicate high-affinity binding of peptides to the groove of MHC class
II molecules (25, 26, 27, 28). As presented in Fig. 2, probing with a combination of L243 and
TAL.1B5 mAbs showed that the altered molecules were comparably stable
to HLA-DR1 wt molecules in SDS. This indicates that the amino
acid-substituted HLA-DR1 molecules can bind peptides endosomally and
implies that they associated normally with Ii and HLA-DM. Similar
results were obtained with the stable transfectants in the DAP.3 cell
line. A representative example for one of the double mutants is shown
in Fig. 2E.
It has been proposed that high-molecular-mass bands at 120 kDa and
above consist of preformed double dimers of MHC class II molecules
(12, 13, 14, 15). Several high-molecular-mass bands were observed
in the lysates of most of the transfectants. However, the HLA-DR(54A)
variant lacked a distinct band of about 180 kDa, the HLA-DR(57A)
variant containing a substitution just outside the interface region I
lacked a distinct band at about 180 kDa, and the band at 120
kDa showed up only weakly following prolonged exposure, indicating a
biochemical phenotype that differed from HLA-DR1 wt molecules. Similar
results were obtained with DAP.3:B7 cells expressing HLA-DR wt or a
selected set of HLA-DR molecules with amino acid substitution in both
region I and III of the MHC class II double dimer interface (Fig. 2E).
To address whether HLA-DR1 molecules spontaneously associate as double
dimers, we generated HLA-DRA cDNAs, encoding the wt sequence fused with
a sequence encoding 6HIS or with an amino acid sequence from the HA Ag
recognized by the mAb 12CA5. These cDNAs were transfected in
combination with cDNAs encoding HLA-DRB, HLA-DMA, HLA-DMB, and Ii
p31. Cell lysates from the transfectants were precipitated with
Ni-Agarose, and the bound material was subjected to SDS-PAGE and
Western blotting using the 12CA5 mAb. We assumed that this assay would
detect complexes of MHC class II molecules stable in SDS because the
Ni-Agarose precipitation was performed in Nonidet P-40, a weaker
detergent than SDS. If MHC class II dimers/oligomers exist in the
absence of TCR ligation, HA-tagged molecules would be expected to
coprecipitate with the 6HIS-tagged MHC class II molecules and to be
detectable on the Western blot. In cells transfected with all the
components needed for peptide loading and Ag presentation (Fig. 3, lane 1), HA-tagged
HLA-DR-chains were barely detectable. In contrast, in the absence of
HLA-DM, Ii p31, or both (Fig. 3, lanes 5, 7, and
6, respectively), aggregates of MHC class II molecules were
formed, showing that the assay can reveal the binding between MHC class
II molecules. However, we concluded that these aggregates cannot be
functional peptide-presenting dimers because in the absence of HLA-DM
and Ii, HLA-DR1 does not form SDS-stable dimers, an indication
of peptide loading (13).
Combined amino acid substitutions in regions I and III led to a reduction in the ability to activate a CD4-dependent T cell clone
To investigate the capacity of the amino acid-substituted HLA-DR
molecules to induce Ag-specific T cell proliferation, stable
transfectants expressing the mutant HLA-DR cDNAs were generated in a
clone of DAP.3 cells previously transfected with a human CD80 cDNA
(DAP.3/B7). Stable transfectants expressing high levels of HLA-DR1
expression were selected using goat anti-mouse mAb-coated magnetic
beads after incubating the cells with mAb L243. The mAb L243 stained
all the transfectants after sorting to a similar degree (Fig. 4). The expression levels of the MHC
class II molecules were regularly assessed, and the cells were resorted
if necessary.
|
A. In the crystal structure
of HLA-DR1, amino acids 52E and 55R appeared to form two salt
bridges with the opposite -chain in the double dimer. We speculated
that the HLA-DR(55E), which has a negatively charged amino acid at
position 55 in place of a positively charged arginine in wt DR1,
would cause a repulsive force between the 1 domains of the double
dimer. This mutant, as well as the mutant HLA-DR(52R55E), was as
efficient as HLA-DR1 wt in stimulating HC6 (Fig. 5B). The
stable transfectants expressing HLA-DR molecules with alanine
substitutions in interface region III, HLA-DR(162A) and HLA-DR(111A,
112A), were also fully able to stimulate HC6 (Fig. 5C). The
lack of inhibition due to region III substitutions suggests that these
amino acid changes did not interfere with the efficiency of the CD4
interaction for this T cell clone, which was highly sensitive to
anti-CD4 mAb inhibition (data not shown).
|
25% of the
response induced by HLA-DR1 wt (Fig. 5D). The HLA-DR(52A,
55A, 162A) and HLA-DR(111A, 112A, 162A) mutants were slightly superior;
however, these transfectants expressed almost 5-fold higher levels of
HLA-DR than the HLA-DR1 wt transfectant. It is of note that the shape
of the dose-response curves was conserved for most of the
transfectants, and higher peptide concentrations did not overcome the
inefficiency of presentation by the doubly substituted molecules. The
same pattern of reactivity was obtained with a second clone with the
same specificity (data not shown). The amino acid-substituted HLA-DR molecules bound peptide with comparable efficiency to wt
One potential variable that it was crucially important to account for was the peptide-binding ability of the different HLA-DR1 molecules. Although most of the amino acid substitutions were at positions that would not be predicted to alter peptide binding, it was possible that indirect effects might occur. To address this, a peptide-binding assay was performed using a biotinylated form of the peptide used for the functional assays, namely, the HA307–319 peptide. Cells were incubated with the peptide for 16 h, and then peptide binding was revealed by the addition of a three-layer streptavidin complex, including the fluorochrome, FITC. The extent of peptide binding was determined using flow cytometry.
The results of this assay are shown in Fig. 6. None of the transfectants bound
peptide significantly less efficiently than HLA-DR1 wt when binding was
corrected for the level of HLA-DR expression. Only the transfectant
expressing the variant HLA-DR(55E) diverged by more than two SDs from
wt, and this molecule bound peptide more, not less, efficiently. These
results suggested that any functional differences that were seen in
response to the substituted molecules could not be accounted for by
differences in peptide binding.
|
Suboptimal stimulation of T cells can lead to up-regulation of
activation markers, such as CD25, but fails to cause proliferation
(29). To determine whether the variant MHC class II
molecules could cause such suboptimal stimulation, HC6 cells were
incubated with peptide-pulsed DAP.3/B7 transfectants expressing HLA-DR1
wt or HLA-DR(52A, 55A, 162A), one of the region I- and III-substituted
molecules that induced minimal proliferation (Fig. 7). After 48 h, the T cells were
analyzed by flow cytometry for CD25 expression. The mixed cell
population of HC6 T cells and DAP.3/B7 transfectants was stained with
directly conjugated anti-CD4 and anti-CD25 mAbs. The expression
levels of CD25 on the CD4+ cells were
up-regulated to a similar degree after incubation with transfectants
expressing HLA-DR1 wt or HLA-DR(52A, 55A, 162A) (Fig. 7, B
and D). These results imply that the region I and III
variant molecules did cause TCR occupancy and CD3-transduced signals,
although they were insufficient to induce IL-2 secretion and maximal
proliferation.
|
To test the generality of these findings, several of the same
amino acid substitutions were introduced into the mouse MHC class II
-chain, H2-Ek. Stable transfectants were
generated in DAP.3/B7 cells and selected for high levels of cell
surface expression, as described previously. One additional mutation
was introduced, leading to the introduction of a tryptophan at position
54 in the 1 domain, H2-Ek(54W). Tryptophan has
a particularly bulky side chain, so that this change alone might be
sufficient to sterically inhibit double dimer formation.
The transfectants were tested for their ability to induce IL-2
secretion by the pigeon cytochrome-c-specific,
H2-Ek-restricted T cell hybridoma, 2B4. The
results are shown in Fig. 8. As can be
seen, a pattern of reactivity was seen similar to that observed in the
human system. Alanine substitutions in regions I or III alone had no
effect on the response of 2B4. In contrast, combining substitutions in
both regions caused substantial inhibition. In addition, the single
amino acid substitution of W at position 54 led to a significant
reduction of IL-2 secretion.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-chain of the HLA-DRB1*0101 molecule in two of the three
regions involved in the interface between putative MHC class II double
dimers. These altered HLA-DR1 molecules were serologically
indistinguishable from wt and appeared to be assembled and transported
to the cell surface normally. They also acquired stability in SDS,
implying that HLA-DM and Ii acted normally on these molecules and that
they were occupied with naturally processed peptides. Additionally,
similar levels of peptide occupancy were observed when the
transfectants expressing native or altered HLA-DR1 molecules were
incubated with exogenous HA peptide. When the transfectants were tested
as APCs for a HLA-DR1-restricted, peptide-specific human T cell clone,
amino acid substitutions involving only one of the two regions had no
effect on Ag-induced proliferation. In contrast, when substitutions at
the two interface regions were combined, this led to a marked reduction
in T cell proliferation despite marked up-regulation of the activation
Ag, CD25. Introduction of amino acid substitutions into
H2-Ek had similar effects on the production of
IL-2 by a mouse T cell hybridoma.
One of the questions that has surrounded the issue of MHC class II
dimerization is whether double dimers exist in the absence of the TCR
or whether the TCR:CD4 complex is instrumental in their formation.
Several groups have presented data supporting spontaneous MHC class II
dimerization. Immunochemical evidence of MHC class II double dimers was
first reported using lysates of a mouse B cell lymphoma where both a
60-kDa and a 120-kDa form of H-2Ek was detected
using immunoprecipitation and Western blotting techniques with
Ek-specific mAbs (12). One of these
mAbs, Y17, was reported to bind preferentially to the 120-kDa form and
also to recognize an identical molecular mass complex in lysates from
mouse splenocytes. Both the 60-kDa and the 120-kDa
H-2Ek complexes were shown to be present on the
cell surface, and neither contained Ii, as indicated by the failure of
an Ii-specific mAb to immunoprecipitate the complex. The 120-kDa
complex immunoprecipitated using Y17 was stable at 37°C, but
dissociated at 50°C to either the 60-kDa form or to the individual
- and -chains. At higher temperatures, the 120-kDa complexes
completely dissociated to free - and -chains. Additional
cross-linking experiments showed that these complexes were present at
the cell surface (13). Similar data for human MHC class II
molecules have been reported (14, 15). However, it has
recently been shown that some of these high-molecular-mass complexes
are formed between MHC class II heterodimers bound to the Abs used in
the assays (30). Furthermore, similar high-molecular-mass
complexes have been shown to contain MHC class II molecules bound to
proteins of varying length captured in the endocytic pathway. Peptide
mapping studies by Pierce et al. attempted to address that possibility
(13). The proteins in the 60-kDa and 120-kDa bands were
proteolytically cleaved and reanalyzed by SDS-PAGE. The results showed
that both bands contained a similar set of polypeptides. However, it is
possible that the peptide maps reflected the composition of the MHC
class II molecules and not that of MHC class II-associated peptides.
Spontaneous MHC class II association has also been suggested by
single-particle fluorescence imaging and scanning force microscopy
using living cells (16, 31). However, the data reported
here (Fig. 3) highlights the influence of Ii and DM. In the absence of
either of these molecules, spontaneous HLA-DR association was seen.
This raises the possibility that MHC class II oligomerization, as seen
in these systems, may reflect either a deficiency in these other
components of the class II pathway or associations between class
II-associated Ii peptide-occupied MHC class II molecules, which are
irrelevant in T cell recogition.
The most direct evidence for dimerization/oligomerization of class II molecules has recently been described by Reich et al. (4). They observed structurally ordered aggregation of MHC class II molecules and T cell Ag receptors and found that this played a key role in triggering T cell activation. However, the oligomerization was driven by cognate recognition of MHC:peptide complexes by the TCR. The limitation of this study was that it involved soluble molecules, and these may behave differently from molecules arrayed in the cell membrane. The aim of the study described here was to examine dimerization in a cellular system. Two of the altered HLA-DR1 molecules generated in this study lacked high-molecular-mass bands when cell lysates were run on SDS gels. However, the presence or absence of high-molecular-mass bands showed no correlation with the ability of these HLA-DR molecules to induce T cell activation. Therefore, it seems more likely that the strong complexes detected in SDS gels are either aggregates of incompletely folded MHC class II molecules or complexes of a class II molecule and other proteins of undefined nature that have been shown to form SDS-stable aggregates (32, 33). However, the failure to detect preformed MHC class II complexes here could be explained if the MHC class II intermolecular avidity is too low to isolate them in the presence of the detergent Nonidet P-40.
Several explanations can be offered for the inhibititon of T cell
proliferation seen with the combined region I and III substitutions.
First, it could be due to impaired CD4 binding. We consider this to be
unlikely for two reasons. First, the amino acids in the 2 domain
that have been implicated in interacting with CD4 are between positions
134 and 148 (7, 8). The changes introduced into the 2
domain in these experiments were spatially distant from the CD4
interaction site in the 2 domain. In addition, if impaired CD4
contact was responsible for these effects, the region III substitutions
would be predicted to inhibit T cell activation independently of the
changes in the 1 domain, which was not the case.
A second possibility is that the amino acid substitutions impaired
peptide binding. The data obtained was inconsistent with this
explanation, in that the altered HLA-DR dimers exhibited the
same degree of stability in SDS as seen with wt HLA-DR1 following
transfection with Ii- and DM-encoding cDNAs. As has been illustrated by
the analysis of MHC class II molecules coexpressed with the Ii in cell
lines that lack DM expression, failure to exchange the class
II-associated Ii peptide for naturally processed peptides leads to
dimers that spontaneously dissociate in SDS. Furthermore, all the
HLA-DR molecules expressed by stable transfectants bound comparable
amounts of the promiscuous HA peptide. Although the kinetics of peptide
binding was not analyzed, the final level of peptide occupancy is most
relevant to the functional experiments in which the transfectants were
prepulsed with the same HA peptide for a similar time period.
A third explanation for these results is that the altered HLA-DR molecules differed from wt HLA-DR conformationally. This is difficult to fully exclude without resorting to crystallization and structural analysis. However, there is nothing to suggest that these molecules were conformationally modified. As mentioned above, they were readily expressed, and at similar levels to native HLA-DR1 molecules, implying efficient folding and transport. In addition, there was no obvious serological change as a consequence of the amino acid changes, judging from the binding of four anti-HLA-DR mAbs (data not shown). The same argument applies as was rehearsed for CD4 binding, in that some negative effect would be expected due to changes in region I or III alone if conformational effects were responsible, in that the two domains would be expected to fold independently. Additionally, the increase in CD25 expression induced by peptide presented by the HLA-DR(52A, 55A, 162A) molecule provides independent evidence that the altered molecule led to TCR recognition and CD3-transduced signals.
By this process of elimination, the most plausible explanation for these results is that the combined alterations in regions I and III compromised T cell activation due to the inhibition of double dimer formation.
These findings were extended into a murine system by the introduction
of alanine residues into some of the same amino acid positions in
H2-Ek. As for HLA-DR, alanine substitutions at
positions 54 or 162 alone had no negative effect on 2B4 IL-2 secretion.
However, exactly as applied in the human system, combining these
changes caused profound inhibition of the 2B4 response. One additional
substitution was made in H2-Ek that provided
further support for the double dimer model. A tryptophan was introduced
at position 54 of the -chain. The rationale for the choice of this
amino acid was that tryptophan has a particularly bulky side chain, and
therefore might sterically impede the association of double dimers
without any change in region III. This prediction was fulfilled, in
that the 54W H2-Ek molecule was significantly
less efficient than wt in stimulating the 2B4 hybridoma.
Taking all the available data into account, we believe the most
coherent hypothesis is that the avidity of MHC molecule homotypic
association is low and leads to the rapid association and dissociation
of dimers and oligomers at the cell surface. The capture by the TCR of
MHC class II:peptide complexes at the cell surface, for which it has
specificity, is not dependent on dimerization. However, downstream
events do seem to be. One attractive model is that the TCR "sorts"
through a large number of MHC:peptide complexes to capture the single
species for which it has specificity, and the dimerization event
initiates the formation of the T cell:APC synapse. TCR signaling is
thought to be sustained by a serial triggering mechanism (34, 35). According to this theory, only a few agonist molecules can
trigger many TCRs. This does not contradict our data. One possibility,
which is now under investigation in our laboratory, is that the TCR
down-regulation is dependent on MHC class II-driven oligomerization.
The existence of binding sites for CD4 on both the - and the
-chain of MHC class II molecules (36) supports the idea
that CD4 serves to stabilize these laterally associated complexes. The
results described here suggest that MHC class II molecules, due to
their structural properties, actively contribute to the formation of
these higher-order complexes that contribute to T cell activation.
|
Footnotes |
|---|
2 Abbreviations used in this paper: wt, wild type; HA, hemagglutinin; Ii, invariant chain; MFI, mean fluorescence intensity; pCMU, plasmid cytomegalovirus U; 6HIS, 6 histadines.
Received for publication April 26, 2000. Accepted for publication October 11, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2 domain of HLA-DR molecules. Nature 356:799.[Medline]
)2 superdimers. J. Immunol. 161:2307.-chain subunits of human ia alloantigens. Immunology. 50:613.[Medline]
heterodimers in the absence of antigenic peptide. Cell 68:465.[Medline]
and chains in CD4 function indicates a role for ordered oligomerization in T cell activation. J. Exp. Med. 182:779.This article has been cited by other articles:
|
|
 |
|
  O. Dushek and D. Coombs Analysis of Serial Engagement and Peptide-MHC Transport in T Cell Receptor Microclusters Biophys. J., May 1, 2008; 94(9): 3447 - 3460. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Li, Y. Zhao, Y. Guo, Z. Li, L. Eisele, and W. Mourad Zinc Induces Dimerization of the Class II Major Histocompatibility Complex Molecule That Leads to Cooperative Binding to a Superantigen J. Biol. Chem., March 2, 2007; 282(9): 5991 - 6000. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. P. Mycko, H. Cwiklinska, J. Szymanski, B. Szymanska, G. Kudla, L. Kilianek, A. Odyniec, C. F. Brosnan, and K. W. Selmaj Inducible Heat Shock Protein 70 Promotes Myelin Autoantigen Presentation by the HLA Class II J. Immunol., January 1, 2004; 172(1): 202 - 213. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. Wetzel, T. W. McKeithan, and D. C. Parker Live-Cell Dynamics and the Role of Costimulation in Immunological Synapse Formation J. Immunol., December 1, 2002; 169(11): 6092 - 6101. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
