| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Pathology and Laboratory Medicine, Emory University School of Medicine, Atlanta, GA 30322
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
heterodimers initially
assemble with the chaperone protein, invariant chain
(Ii),3 in the
endoplasmic reticulum of APCs. Ii stabilizes the newly formed MHC
heterodimers through multiple interactions involving sites both inside
and outside the peptide-binding groove. A flexible region of Ii encoded
by exon 3 is believed to act as a surrogate peptide, occupying the
peptide-binding groove in Ii complexes. A targeting signal in the
cytoplasmic domain of Ii directs transport and localization of the
complexes to compartments in the endosomal pathway, where Ii is
released through sequential proteolytic cleavage events. However, the
Ii segment that occupies the peptide-binding groove, class
II-associated Ii chain peptide (CLIP), is inaccessible and remains
associated with class II , blocking peptide loading
(2). HLA-DM functions to accelerate the rate of
dissociation of CLIP, freeing the binding site and promoting the
loading of MHC class II molecules with heterogeneous peptides present
in the endosomal environment (3, 4, 5) Crystallographic studies demonstrated that the structure of HLA-DM is very similar to that of conventional MHC class II molecules (6, 7). However, the peptide-binding groove is largely closed, with only a single pocket remaining that is too small to accommodate CLIP or other peptides in a conventional manner. DM promotes CLIP dissociation and peptide exchange through a mechanism involving direct interaction with MHC class II peptide (3, 4, 5). The specific nature of the binding interaction (8, 9) remains unknown but mutational studies and other lines of evidence suggest that DM may interact with the lateral surface of class II molecules below the end of the peptide-binding groove that accommodates the N terminus of bound peptide (3, 10, 11). It was initially thought that DM might selectively recognize structural features of class II-CLIP complexes. However, subsequent studies showed that it can accelerate the dissociation of peptides with sequences unrelated to CLIP (12, 13, 14). Indeed, current evidence suggests that DM increases the rate of dissociation of all peptide complexes such that the rate observed in the presence of a given concentration of DM is approximately proportional to the intrinsic peptide dissociation rate observed in the absence of DM (12). Newly synthesized MHC class II molecules have a relatively brief exposure to DM in endosomal compartments during trafficking to the cell surface. During this brief exposure, less stable complexes, such as those bearing CLIP, are much more likely to dissociate than are complexes with greater stability. Thus, DM appears to edit the repertoire of peptides presented by class II molecules by catalyzing multiple rounds of peptide exchange, favoring survival of the most stable complexes.
Studies with soluble recombinant DM demonstrated that micromolar concentrations were required for half-maximal activity in assays measuring peptide dissociation or exchange with purified HLA-DR class II molecules (4). By contrast, similar activity was observed with nanomolar concentrations of detergent-solubilized full-length DM purified from B cell lines (3, 5). It was subsequently shown that substoichiometric concentrations of HLA-DO copurify with DM isolated from B cells (15). The function of HLA-DO remains controversial. It clearly has a capacity to inhibit DM function in a variety of experimental conditions (15, 16, 17, 18) but it also has been reported to promote the peptide loading and chaperone functions of DM (19, 20). The possibility that contaminating HLA-DO was responsible for the enhanced activity of purified DM was ruled out by Busch et al. (21), who showed that full-length recombinant DM is much more potent than the soluble protein lacking transmembrane and cytoplasmic domains. In addition, these authors demonstrated that much higher concentrations of full-length DM are required to promote peptide binding to soluble DR molecules as compared with full-length DR. Thus, the transmembrane and/or cytoplasmic domains of DM and DR molecules were demonstrated to be required for efficient DM-mediated peptide loading. In the present study, the role of the transmembrane and cytoplasmic domains is further investigated.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
DRA and DRB1 cDNAs were generously provided by Dr. E. Long (National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD); DMA and DMB gene constructs have been previously described (5). B7.1 (CD80) cDNA in pcDNA3.0 was a kind gift from Dr. C. Larsen (Department of Surgery, Emory University, Atlanta, GA).
For expression in insect cells: DRA and DRB1 cDNA complete sequences or
truncated fragments were subcloned into pAcUW51 BamHI and
BglII sites, respectively. To generate C-terminal truncated
sequences, stop codons and restriction sites were incorporated into the
reverse primers annealing to the connecting peptide sequence, excluding
the transmembrane regions of the molecules. The truncated fragment was
amplified by PCR and cloned into the same vector. Each construct was
checked for orientation and sequenced. Recombinant baculoviruses were
produced in Sf9 cells using the Baculovirus Transfection kit (BD
PharMingen, San Diego, CA). Several cycles of viral stock amplification
were done to obtain viral titers of
109–1010 infectious
particles/ml. Proteins were expressed by infection of Hi-5 cells with
the recombinant viruses at a multiplicity of infection of 10.
Viral constructs for HLA-DM complete or truncated segments were
generated similarly in pAcUW51 (BD PharMingen). Three days after
infection, cells were harvested or supernatants collected. Full-length
DR or DM was purified by affinity chromatography from 0.5% Nonidet
P-40 (NP40; U.S. Biochemical, Cleveland, OH) cell lysates using
mAbs LB3.1 (22) and WT-1 (W. P. Thayer and
P. E. Jensen, unpublished data), respectively. Supernatants
containing soluble forms of DR1 and DM were cleared by centrifugation
at 14,000 x g for 1 h before affinity
chromatography. Both full-length and soluble proteins were eluted in
100 mM glycine-NaOH buffer at pH 11.5 containing 0.5% octyl
-D-thioglucopyranoside (SOG; Sigma-Aldrich,
St. Louis, MO) and maintained in 50 mM Tris buffer (pH 7.5) and 0.5%
SOG. Protein concentrations were determined by sandwich ELISA: DR1 was
captured by mAb L243 (23) and detected with biotinylated
mAb IVA12 (ATCC HB145) followed by incubation with
europium-streptavidin; DM molecules were captured with mAb WT-1
(W. P. Thayer and P. E. Jensen, unpublished data) and
detected with biotinylated MAP-DM (a kind gift from Dr. P.
Cresswell, Yale University, New Haven, CT). Protein purity was assessed
by SDS-PAGE and Coomassie staining.
For expression in COS-7 cells, the same fragments of DM and DR1 genes were subcloned into pcDNA3.0 (Invitrogen, San Diego, CA). Chimeric constructs to express DR1 molecules with transmembrane-cytoplasmic domains of either DM or CD80 were generated by overlapping PCR (24) and cloned into pcDNA3.0.
Peptide dissociation assays
CLIP peptide was biotinylated as previously described
(25). All binding and dissociation reactions were done in
binding buffer (0.1 M citrate-phosphate buffer at pH 5.0, 0.2% NP40,
containing protease inhibitors) (5). DR1-biotin CLIP
complexes were formed by incubation of 500 nM DR1 with 5 µM
biotin-CLIP for 18 h at 37°C. For dissociation, preformed
complexes were diluted 10-fold and incubated in 20-µl volumes with
100 µM unlabeled MAT(17–31). After neutralization, the complexes
were captured on L243-coated microtiter wells at 4°C and biotin
complexes were measured with streptavidin-europium as previously
described (5). Peptide binding and dissociation reactions
of preformed fluorescein-labeled complexes were conducted as previously
described (12). In brief, 100 µM fluorescein-labeled
MAT(17–31) peptide and 2 µM full-length or soluble DR1 were
incubated in binding buffer for 24 h at 37°C. Unbound
fluorescein-labeled MAT(17–31) peptide was separated from the class
II-peptide complex using Sephadex G-50 columns (Boehringer Mannheim,
Indianapolis, IN). Samples were diluted 1/10 to DR1 concentrations of
0.2 µM in binding buffer and divided into smaller aliquots to which
various amounts of full-length or soluble DM were added together with
an excess (100 µM) of unlabeled MAT(17–31) peptide. Peptide
dissociation was monitored using high performance size exclusion
chromatography. At various times, 20-µl sample aliquots were removed.
DR1-FITC-peptide was quantified with a QC-PAK GFC 200 size exclusion
column (7.8 x 150 mm; Tosohaas, Montgomeryville, CA) and a
Shimadzu Scientific Instruments (Columbia, MD) RF-10A fluorometer with
490 nm excitation and detection at 520 nm. The column buffer was 50 mM
phosphate buffer (pH 7.0) with 0.2 mM dodecyl
-D-maltoside. Peptide binding was assessed as a function
of fluorescence peak height.
Peptide binding to chimeric molecules expressed in COS-7 cells
Cells were transfected by the DEAE-dextran method as described
elsewhere (26). Briefly, 4 x
105 cells were plated on 60-mm tissue culture
dishes in DMEM-10% FBS 24 h before transfection. Five dishes were
prepared for each transfection and pooled at the end of the experiment.
The following day cells were washed twice without serum in DMEM, 10 mM
HEPES, and incubated in 2 ml DMEM, 10 mM HEPES containing 500 µg
DEAE-dextran, 100 µM chloroquine, and DNA: 1 µg each class II -
and -chain constructs for 3 h at 37°C. DNA was removed by
aspiration and 2 ml 10% DMSO in DMEM and10 mM HEPES was added to each
dish for 2 min at room temperature, followed by incubation in DMEM,
10% FBS, and 10 mM HEPES at 37°C. Cells were harvested by
trypsinization at 42 h after transfection. An aliquot of cells was
analyzed for surface expression of DR1 by staining with mAb L243
followed by FITC-labeled goat anti-mouse Ig (whole molecule;
Sigma-Aldrich) and analyzed on a FACScan cytofluorometer (BD
Biosciences, Mountain View, CA). The remaining cells were lysed in 1%
NP40 in the presence of protease inhibitors. Lysates were cleared of
cellular debris by centrifugation. To each lysate, citrate-phosphate
buffer (pH 5.0) was added to a final concentration of 0.1 M and
biotin-MAT peptide to a final concentration of 2 µM. Each lysate was
then divided into aliquots and varying concentrations of DM
were added followed by a 3.5-h incubation at 37°C. Samples were then
neutralized and DR1 molecules in the lysates were captured on
microtiter plates coated with L243 mAb. DR1 molecules in the lysates
were quantified by sandwich ELISA as described above. Biotinylated
peptide bound to DR1 was quantified with europium-streptavidin
fluorescence as described previously (27).
Peptide binding to murine MHC class II molecules
IAd and IE d were purified from detergent-solubilized A20 B lymphoma (28) membrane preparations using mAb immunoaffinity columns: MK-D6 for IAd (29) and 14-4-4 for IEd (30) as previously described (25). 14-4-4 was also used in the purification of IEk from CH27 cells (31).
Purified murine class II molecules (50 nM) were added to1 µM biotin-labeled peptide on ice (IAd, Myo(106–118); IE d, HEL(104–120); IEk, MCC(82–103) in 100 mM citrate-phosphate buffer (pH 5.0) with protease inhibitors in the presence or absence of purified HLA-DM (100 nM). Varying concentrations of NP40 were added to duplicate samples. After mixing, the samples were equilibrated on ice for 20 min followed by a 3-h incubation at 37°C. After neutralization, class II-biotin-peptide complexes were captured on assay plates precoated first with goat anti-mouse Ig (H and L chains; Sigma-Aldrich) followed by the appropriate mAbs: MK-D6 (IAd) or 14-4-4S (IEd and IEk). Biotin-peptide complexes were quantified by europium-streptavidin fluorescence as mentioned above (27).
Liposome experiments
A lipid solution containing dipalmitoyl phosphatidyl choline
(C16:0; Sigma-Aldrich) and cholesterol in an 8:2 molar ratio in
chloroform was dried to form a thin film in glass tubes under a stream
of nitrogen. The lipid was solubilized on ice in 1%
n-octyl--D-glucopyranoside in the
presence of purified full-length HLA-DM or HLA-DR1 to form solutions
containing 1 mM lipid with or without 1 µM DR and/or 88 nM DM.
Samples were dialyzed for 24 h at 4°C against three changes of
PBS using dialysis membranes with a molecular mass cutoff of
12,000 Da to remove detergent and generate unilamellar liposomes
(32, 33). Peptide binding to DR1 in liposomes was measured
by mixing equal volumes of liposomes containing DR1 and liposomes
containing DM or equal volumes of liposomes containing both DM and DR1
with liposomes containing neither protein. Samples, containing a final
concentration of 125 nM DR1 and 11 nM DM (as indicated) were incubated
for 3 h at 37°C in 100 mM citrate-phosphate (pH 5) with 1 µM
biotin-MAT(17–36). The samples were placed in ice and liposomes were
solubilized in buffer containing 1% NP40, 0.1% Tween 20, 0.1 M NaCl,
0.4 M Tris (pH 8), and 1 mg/ml BSA. Biotin-peptide DR1 complexes were
quantified with a europium fluorescence immunoassay as described above.
Control samples were solubilized in 1% NP40 before incubation with
biotin-peptide at pH 5, neutralization, and assayed for peptide-DR1
complexes.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our initial studies, soluble recombinant HLA-DM lacking both
transmembrane and cytoplasmic domains (truncated DM) (Fig. 1
) was observed to have little or no
activity in assays measuring enhanced binding by purified DR1 molecules
in detergent solution. The recombinant protein appeared to be properly
folded as judged by reactivity in immunoassays using mAbs with
specificity for nonoverlapping conformational determinants in DM (data
not shown). Subsequent experiments demonstrated that peptide binding
was efficiently catalyzed in the presence of high (micromolar)
concentrations of the soluble truncated protein. Full-length
detergent-solubilized recombinant DM was generated using a baculovirus
expression system to investigate this phenomenon further. Since the
full-length protein requires the presence of detergent for
solubilization, experiments were done in standard assay buffer
containing 0.2% NP40. The full-length recombinant protein was observed
to be as potent as DM purified from B cells in assays measuring peptide
binding to purified DR1 molecules. The potency of the recombinant
full-length DM vs truncated DM was compared in assays measuring
enhancement of dissociation of CLIP from preformed DR1-CLIP complexes.
An approximately 400-fold greater concentration of truncated DM was
required to induce the same degree of CLIP release during a 3-h
incubation at pH 5 (Fig. 2a).
Remarkably, a very similar dose relationship was observed when
comparing the capacity of full-length DM to enhance the dissociation of
CLIP from full-length vs truncated DR1 molecules (Fig. 1) lacking
transmembrane and cytoplasmic domains (Fig. 2b). In
these experiments, all components were generated as recombinant
proteins, ruling out any contribution from HLA-DO or cofactors that
might copurify with DM or DR from B cell lines. These results suggested
that interactions involving the transmembrane or cytoplasmic domains of
both DM and DR proteins were required to promote maximal DM catalytic
activity.
|
|
a). In comparison, the effect
observed with a 16-fold higher concentration of truncated DM was
considerably less. A linear relationship was observed between the rate
of peptide dissociation and DM concentration for both forms of DM (Fig. 3b). However, the slope of rate vs DM concentration, a
measure of DM potency, was 240-fold lower for truncated compared with
full-length DM. This difference is in reasonable agreement with that
observed in assays measuring CLIP dissociation.
|
A requirement for the cytoplasmic domains in DM catalytic function
was excluded through analysis of L cell transfectants. Membrane
preparations from L cells expressing comparable levels of full-length
DM or a truncated form lacking only the cytoplasmic domains of both
- and -chains was observed to have similar activity in assays
measuring DM-catalyzed peptide binding to purified DR1 molecules (data
not shown). The possibility that specific interactions between the
transmembrane domains of DM and DR are required for optimal activity
was addressed by generating chimeric molecules with altered
transmembrane domains. We initially attempted to do this by replacing
the transmembrane domains of DM and with corresponding domains
from DR1. However, expression of the mutant DM molecules could not be
detected in transiently transfected COS cells, suggesting that they did
not fold properly. As an alternative approach, constructs encoding
chimeric molecules were generated in which the transmembrane domains of
DR and DR were replaced with the corresponding domains from DM
(Figs. 1
and 4). Surface expression of
the mutant protein was similar to wild-type DR1 in COS cells after
transient transfection as determined by flow cytometry using the
DR-specific mAb L243 (data not shown). The capacity of DM to promote
peptide binding to the mutant DR molecule was analyzed by incubating
detergent lysates of the transfected cells with various concentrations
of purified full-length DM in the presence of biotin-labeled
MAT(17–31) peptide. Biotin-peptide-DR1 complexes were measured using a
europium fluorescence immunoassay. DM was fully capable of catalyzing
peptide binding to the mutant DR molecules with substituted
transmembrane domains (Fig. 4). In the experiment shown, maximal
peptide binding by the mutant DR was somewhat greater than wild type.
This can be attributed to variation in the DR expression levels and it
was not a consistent finding. The transmembrane domains of DM and DR
are reasonably similar. To fully exclude a requirement for specific
interactions between these domains, the transmembrane domains of DR
and were each replaced with a sequence from the unrelated molecule,
CD80 (B7-1). This mutant heterodimer was also expressed efficiently in
COS cell transfectants and was fully susceptible to DM-catalyzed
peptide loading (Fig. 4). Thus, we found no evidence for
sequence-specific interactions between the transmembrane domains of DM
and DR required to promote catalytic function.
|
A puzzling observation obtained in peptide-binding experiments
with murine class II molecules may be relevant to the experiments
described above, demonstrating a role for transmembrane domains in DM
function. Using standard assay buffer, containing 0.2% NP40 detergent,
purified HLA-DM was observed to have little activity in catalyzing
peptide binding by certain purified full-length mouse class II
molecules. However, DM activity was revealed by reducing the NP40
concentration and optimal activity was observed at concentrations
approximating or below the critical micelle concentration for this
detergent (Fig. 5). Variable results were
obtained under conditions where no NP40 was added, probably due to
protein aggregation. The assay samples also contained low
concentrations of SOG (
0.3 mM), the detergent used to maintain
solubility of the purified protein stocks. Avva and Cresswell
(34) previously reported that detergents containing 8- to
10-carbon unbranched hydrocarbons, such as octylglucoside, can
facilitate CLIP release from DR molecules at acidic pH. It is possible
that SOG could mediate a similar effect at limiting concentrations of
NP40 by promoting the dissociation of previously bound peptides from
mouse class II molecules, freeing more binding sites for biotin-labeled
peptide. However, several points weigh against this explanation. The
detergent concentrations used in the CLIP release assays
where 100-fold higher than the concentration of SOG present in our
experiments. In addition, peptide binding in the absence of DM was not
affected by varying NP40 concentrations (Fig. 5). Finally, reduction in
NP40 concentrations below 0.2% had little effect on DM-catalyzed
peptide binding to DR1 molecules under parallel conditions. Several
studies based on transfection experiments have demonstrated that HLA-DM
can modulate peptide loading by mouse IAk,d
molecules (35, 36, 37). However, the possibility that there
may be a partial species incompatibility between human DM and mouse
class II molecules has been raised (38). One explanation
for our results is that, at limiting detergent concentrations, DM and
mouse class II molecules are codistributed into a smaller volume of
detergent micelles and their effective concentrations are therefore
increased. This effect could compensate for a partial species
incompatibility. The interaction between DM and full-length DR may be
sufficiently optimized so that NP40 detergent concentrations below
0.2% have little effect. Consistent with this idea, we have observed
that DM activity is reduced at NP40 concentrations >1% in assays with
DR1 (data not shown).
|
The results described above support the hypothesis that the
transmembrane domains of DM and DR optimize their functional
interaction in detergent solutions by facilitating their colocalization
in detergent micelles. However, the possibility that detergent
interactions with the transmembrane domains may have some complex
effect in promoting DM activity could not be excluded. DM and DR were
reconstituted into lipid bilayers to directly determine the effect of
membrane localization on DM-mediated peptide exchange. Purified
full-length DM and DR were reconstituted together or separately into
unilamellar liposomes containing phosphatidylcholine and cholesterol
generated by detergent dialysis. Care was taken to ensure that all
samples contained the same total concentration of DM, DR, and liposomes
such that they differed only in the distribution of DM and DR molecules
among the liposomes. The samples were incubated at pH 5 with
biotin-MAT(17–31) peptide, followed by pH neutralization, detergent
solubilization, and assay for DR1-biotin-peptide complexes. Peptide
binding was only slightly above background in samples containing DM and
DR independently distributed into different liposomes (Fig. 6). In contrast, peptide binding was
markedly enhanced under conditions where DM and DR were localized in
the same liposomes. As a control, parallel liposome samples were
solubilized on ice in 0.5% NP40 detergent before incubation with
labeled peptide. Under these conditions, the proteins were randomly
redistributed into micelles and rapid peptide-binding activity was
restored in all samples containing both DM and DR. These results
demonstrate directly that membrane colocalization is a critical
parameter in controlling DM-mediated peptide exchange activity.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It is possible that the transmembrane domains of DM or DR somehow affect the conformation of these molecules promoting functional interaction. This possibility is supported by a previous observation that membrane localization can alter the pH dependence of peptide binding by DR molecules (39). In addition, specific sequences in the transmembrane domain of HLA-Cw6 were reported to be required for functional recognition by inhibitory receptors on natural killer cell clones through a mechanism that might involve alterations in membrane mobility or conformation (40). However, the fact that transmembrane domains in both DM and DR are required makes it unlikely that these domains indirectly facilitate functional interaction between the molecules by subtly altering the conformation or conformational flexibility of both molecules.
Experiments with chimeric DR1 molecules containing transmembrane
sequences from DM or a completely unrelated molecule, CD80, ruled out
the possibility that a specific interaction between the transmembrane
domains (41) promotes the functional interaction of DM
with DR. The mutant DR1 protein was efficiently assembled and expressed
in transiently transfected COS cells, indicating that the DM and
transmembrane regions readily substitute for the corresponding
sequences in DR1 molecules. It is interesting that the converse was not
true. We were unsuccessful in expressing chimeric DM molecules
containing transmembrane sequences from DR1. The observation that
truncated soluble DM or DR molecules can be efficiently expressed and
secreted by insect cells (at 26°C) demonstrates that there is no
obligate requirement for interactions between the and
transmembrane domains for pairing or formation of stable molecules.
However, assembly of IAk MHC class II
heterodimers in COS cell transfectants was previously reported to
involve interactions between the transmembrane domains
(42). It is possible that transmembrane domain
interactions play a role in assembly of DM heterodimers at
physiological temperatures in mammalian cells. Notably, Calafat et al.
(43) demonstrated that the transmembrane and cytoplasmic
domains of MHC class II have the potential to specifically induce
multilaminar morphology in lysosomal compartments of human embryonal
kidney cells, similar to MHC class II compartment peptide
loading compartments of professional APC. Thus, in addition to membrane
anchoring, the transmembrane/cytoplasmic domains in MHC class II
molecules have the potential to influence the organization of endosomal
compartments.
Our findings are most compatible with the conclusion that the
transmembrane domains of DM and DR are required to facilitate
colocalization in membrane or detergent micelles. The role of membrane
colocalization was clearly illustrated in experiments involving
reconstitution of full-length DM and DR molecules into lipid bilayers.
DM and DR were observed to cooperate efficiently when localized
together in unilamellar liposomes, whereas little or no activity was
observed when the two proteins were segregated into separate liposomes
incubated together with labeled peptide Ag. Full activity was recovered
in the latter samples when DM and DR were randomly redistributed into
micelles by detergent solubilization of the liposomes. Thus, DM-DR
interactions occur efficiently in cis- but not
trans-configuration. In addition, we observed that the
capacity of purified full-length HLA-DM to enhance peptide binding by
several mouse class II molecules was increased by reducing the
concentration of NP40 detergent in the assays. There are a number of
potential explanations for this. For example, the conformation of some
or all murine class II might be differentially affected by exposure to
detergent, indirectly influencing interactions with DM. Alternatively,
by reducing detergent concentration, the volume of micelles is reduced
with a concomitant reduction in the effective volume in which the
transmembrane-containing protein is distributed. The predicted effect
is to increase the effective concentration of DM and class II,
increasing their rate of interaction. In contrast to the results with
the mouse class II proteins examined (IEk,
IAk, IAd), reducing the
concentration of NP40 below 0.2% had little effect on DM-catalyzed
peptide binding to DR1. However, high concentrations of NP40 (
1%)
partially inhibit DM activity on DR1. It is possible that there is a
partial species incompatibility between human DM and mouse class II
molecules that accounts for this difference (38). Notably,
HLA-DM can efficiently catalyze peptide binding to purified
IAb molecules in 0.2% NP40 (44). It
is interesting to consider the possibility that polymorphisms may
influence the ability of HLA-DM to interact with mouse class II
molecules. The effect of varying detergent concentrations has not yet
been studied with IAb. Detergents can influence
peptide binding by MHC class II molecules through a variety of
potential mechanisms, including direct effects on peptide release or
exchange (34). In our studies, however, the concentration
of NP40 selectively modulated DM activity with no direct effect on
peptide binding in the absence of DM. Thus, we believe that the best
explanation for the effect of detergent concentration in our
experiments relates to its effect on the volume of micelles into which
DM and DR partition.
The effect of transmembrane anchoring on DM-DR interaction provides an important example of how membrane localization can affect the kinetics of protein-protein binding and enzymatic reactions. Soluble proteins in solution are freely tumbling with random orientation, whereas transmembrane proteins have relatively constrained orientations with respect to the plane of the membrane. It is likely that productive interactions between DM and DR occur through a lateral interface in both molecules (10) and that this topology is favored by colocalization in lipid bilayers or micelles. The high local concentration of interacting molecules in biological membranes promotes interactions through mass action whereas conventional concentrations are defined by total solution volume. Volume exclusion by irrelevant integral membrane proteins has also been proposed to contribute toward enhanced association of membrane-associated molecules (45). This effect is not likely to be important in our experiments with purified molecules. Grasberger et al. (45) reported model calculations for the contributions of these parameters, estimating that these factors can increase the likelihood of forming dimers by six orders of magnitude as compared with soluble proteins in solution. Under the experimental conditions used in the present study, with purified proteins variably partitioned into detergent micelles through transmembrane domains, we empirically measured an enhancement of somewhat greater than two orders of magnitude based on relative DM functional potency.
Our findings underscore the potential importance of the local
distribution of DM and substrate class II molecules in the membranes of
peptide loading compartments. Schafer et al. (46)
estimated that the ratio of DR:DM was 23:1 in a B cell line. This was
reduced to 5:1 in isolated subcellular fractions enriched for MHC class
II compartment peptide loading vesicles. DM-catalyzed peptide
loading occurs in a series of late endosomal compartments enriched for
both class II molecules and DM (MIIC). These endosomes have complex
internal membrane structures including vesicles or sheets. DM has been
reported to be enriched in the limiting membrane of these compartments,
whereas DR is distributed in both limiting and internal membranes
(47, 48). Of relevance, it has been reported that the
cytoplasmic tail of HLA-DO can influence the distribution of both DM
and DR in multivesicular MIIC by preferentially localizing them to the
limiting membrane (49). Through mechanisms such as this,
extrinsic factors might influence DM peptide exchange function by
controlling DM-DR colocalization. Of further interest is the potential
role of tetraspan proteins such as CD82 which coprecipitate with DM,
DR, and other proteins (47, 48). These molecules may form
scaffolds that regulate protein interactions. In contrast to DM,
tetraspan proteins have been reported to be enriched in internal
membranes (47, 48). The limiting and internal membranes
also differ in the distribution of lipids, defining specialized domains
(50). Clearly, the local distribution of DM and DR and
dynamics of trafficking in these complex structures could have a major
impact on the regulation of peptide loading and editing in the MHC
class II Ag presentation pathway.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Peter E. Jensen, Department of Pathology, Emory University School of Medicine, Room 7313 WMB, 1639 Pierce Drive, Atlanta, GA 30322. E-mail address: pjensen{at}emory.edu
3 Abbreviations used in this paper: Ii, invariant chain; CLIP, class II-associated Ii chain peptide; NP40, Nonidet P-40; SOG, octyl -D-thioglucopyranoside.
Received for publication June 7, 2001. Accepted for publication August 31, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
dimers and facilitates peptide loading. Cell 82:155.[Medline]
heterodimers in the absence of antigenic peptide. Cell 68:465.[Medline]
This article has been cited by other articles:
|
|
 |
|
  W. Y. Lee, D. A. Weber, O. Laur, E. A. Severson, I. McCall, R. P. Jen, A. C. Chin, T. Wu, K. M. Gernet, and C. A. Parkos Novel Structural Determinants on SIRP{alpha} that Mediate Binding to CD47 J. Immunol., December 1, 2007; 179(11): 7741 - 7750. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. M. Grotenbreg, M. J. Nicholson, K. D. Fowler, K. Wilbuer, L. Octavio, M. Yang, A. K. Chakraborty, H. L. Ploegh, and K. W. Wucherpfennig Empty Class II Major Histocompatibility Complex Created by Peptide Photolysis Establishes the Role of DM in Peptide Association J. Biol. Chem., July 20, 2007; 282(29): 21425 - 21436. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Xing, T. Yu, Y. N. Wu, M. Roy, J. Kim, and C. Lee An expectation-maximization algorithm for probabilistic reconstructions of full-length isoforms from splice graphs Nucleic Acids Res., June 6, 2006; 34(10): 3150 - 3160. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Kralovicova, S. Houngninou-Molango, A. Kramer, and I. Vorechovsky Branch site haplotypes that control alternative splicing Hum. Mol. Genet., December 15, 2004; 13(24): 3189 - 3202. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. H. Koonce and E. K. Bikoff Dissecting MHC Class II Export, B Cell Maturation, and DM Stability Defects in Invariant Chain Mutant Mice J. Immunol., September 1, 2004; 173(5): 3271 - 3280. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Stratikos, D. C. Wiley, and L. J. Stern Enhanced Catalytic Action of HLA-DM on the Exchange of Peptides Lacking Backbone Hydrogen Bonds between their N-Terminal Region and the MHC Class II {alpha}-Chain J. Immunol., January 15, 2004; 172(2): 1109 - 1117. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. P. Belmares, R. Busch, K. W. Wucherpfennig, H. M. McConnell, and E. D. Mellins Structural Factors Contributing to DM Susceptibility of MHC Class II/Peptide Complexes J. Immunol., November 1, 2002; 169(9): 5109 - 5117. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Stratikos, L. Mosyak, D. M. Zaller, and D. C. Wiley Identification of the Lateral Interaction Surfaces of Human Histocompatibility Leukocyte Antigen (HLA)-DM with HLA-DR1 by Formation of Tethered Complexes That Present Enhanced HLA-DM Catalysis J. Exp. Med., July 15, 2002; 196(2): 173 - 183. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
). As a control, parallel samples
were solubilized in detergent before incubation with peptide to
redistribute DM and DR together in micelles (
).