| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, MD 21205; and
Department of Chemistry, Massachusetts Institute of Technology, Cambridge, MA 02139
Natarajan, S.K., M. Assadi and S. Sadegh-Nasseri. Stable peptide binding to MHC class II molecule is rapid and is determined by a receptive conformation shaped by prior association iwth low-affinity peptides. J. Immunol. In press.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
/ß heterodimer in SDS. Moreover, the presence of a smaller
hydrophobic residue, such as alanine, as P1 anchor with no contribution
from any other anchor is sufficient to enhance the SDS stability of the
mutant complexes, demonstrating that the basis of SDS resistance may be
localized to P1 interactions. The good correlation between SDS
sensitivity and the exposure of hydrophobic residues provides a
biochemical rationale for the use of this assay to investigate the
maturation of class II molecules and the longevity of the
complexes. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
- and ß-chains in the presence of SDS, as determined from SDS-PAGE
experiments performed without prior boiling of the samples 1, 2, 3, 4 . This
characteristic of MHC class II complexes has been used as a convenient
and sensitive assay to determine the efficiency of in vitro as well as
intracellular peptide binding. This assay has been pivotal in
determining the influence of peptide binding on class II MHC structure
3 as well as in establishing a distinct structural state associated
with stable complexes 5 . Formation of SDS-stable compact dimers has
been suggested to indicate completion of MHC class II folding upon
binding to peptide molecules and thus gaining maturity 6 . Making use
of SDS stability as a peptide binding marker, events in intracellular
trafficking of MHC from synthesis, folding and assembly, sorting to
vesicular compartments, association of invariant chain, peptide
loading, and the following transport to cell surface have all been
characterized 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 . Furthermore, it has been documented that
immunodominant peptide epitopes often form SDS-stable complexes with
MHC molecules, and peptides that form SDS-stable complexes increase the
longevity of class II molecules 22, 23 . Despite the benefits of using the SDS stability assay, a few recent reports have suggested that the SDS stability of the complex is not strictly correlated to the peptide/MHC affinity, thereby questioning the very basis of this assay 22, 24, 25 . Two structures with contrasting properties, long-lived SDS-stable DR1/HA306–3183 complex 26 and short-lived SDS-unstable DR3/CLIP complex 27 , also did not provide sufficient information about the basis of SDS stability. Contradictory interpretations of this assay stem from the lack of a clear understanding of the mechanisms involved at the molecular level.
In this report we describe a study of the roles of specific peptide-MHC interactions and SDS in the behavior of class II complexes in SDS using the well-characterized HLA-DR1 molecule. Pocket 1 (P1) of DR1 plays the most important role in the peptide interactions, as shown by binding studies 28 and x-ray crystal structures of DR1/HA306–318 26 and DR1/A2 29 . P1 is a deep pocket lined with a series of hydrophobic residues that constitute about 85% of the solvent accessible area. The P1 anchor residues, Tyr308 in HA306–318 peptide and Trp307 in A2 peptide, are almost completely buried in pocket 1. A strong preference of pocket 1 for the aromatic side chain residues (Tyr, Trp, Phe) has been reported 28, 30 . Long aliphatic side chains (Met, Leu, Ile, Val) also bind, although less efficiently 30 . In SDS-PAGE experiments with HLA-DR1, only the peptides that had aromatic or long aliphatic side chains as the P1 anchor were able to form SDS-stable complexes 24, 31, 32 .
In this work we investigated the biochemical basis of SDS
stability of peptide-class II complexes using specific peptide
interactions with soluble DR1. Our experiments with several
HA306–318-derived peptides indicate that
hydrophobic interactions between the P1 residues and the bulky P1
anchor and the resulting burial of these residues are primarily
responsible for SDS stability. SDS stability was determined by a new
biosensor-based assay we have developed and by SDS-PAGE. To further
investigate this idea from a different perspective, a site-specific
mutation, ßGly86
Tyr, in pocket 1 was constructed to
increase the contacts between the subunits and make the pocket more
shallow, thereby reducing the contribution from the P1 anchor required
to stabilize the complex. Results with the mutant DR1 help to reaffirm
the correlation between exposed hydrophobicity and SDS sensitivity.
This correlation provides a biochemical rationale for using the SDS
stability assay as a probe for determination of peptide affinity.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
cDNA encoding the extracellular domain of DRB1*0101 4 was mutated to change the ß86 glycine residue to tyrosine. The wild-type gene was cloned into pBluescript II KS vector (Stratagene, La Jolla, CA). The desired mutation was introduced in this clone using two synthetic primers, complementary to opposite strands, containing the nucleotides to be changed. The sequence of oligonucleotide used (only the coding strand is shown, with the altered nucleotides in boldface type and the new silent HindIII restriction site underlined) was 5'-C TAC GGG GTT TAT GAA AGC TTC ACA GTG C-3'.
The primers were extended by thermal cycling with pfu DNA
polymerase to generate the entire plasmid. The clones containing the
mutant gene were screened using the new HindIII site. One
such clone was sequenced throughout the entire coding region to verify
the absence of undesired mutations. The wild-type gene in the
baculovirus transfer vector PVL1393 4 was replaced with the mutant
gene. Baculovirus vector containing both the and mutant ß genes
with two polyhedrin promoters was constructed as described previously
29 .
Production of recombinant soluble DR1 proteins
Soluble DR1 proteins were expressed and purified as originally
described 4 . Baculovirus DNA (BaculoGold; PharMingen, San Diego, CA)
and transfer vectors carrying the wild-type or mutant genes were
cotransfected into Sf9 insect cells to produce recombinant viruses. Hy5
cells were infected with these recombinant viruses, and DR1 proteins
were purified from the culture supernatant using anti-DR1 mAb
(L243) immunoaffinity chromatography columns 33 . Purified wild-type
and mutant DR1 proteins migrated similarly in SDS-PAGE at the expected
sizes of and ß subunits when samples were boiled before
electrophoresis.
Gel filtration
A Superdex 200 fast protein liquid chromatography column (Pharmacia, Piscataway, NJ) was used to analyze wild-type and ßG86Y DR1 protein preparations as well as separate peptide-loaded complexes. The samples were eluted in PBS at a flow rate of 0.4 ml/min. The column was calibrated using a protein m.w. standard (Bio-Rad, Richmond, CA), and the parameters (slope and intercept) of the linear plot between log (m.w.) and elution volume were determined. Using this linear equation and the experimentally determined elution volumes of DR1 and DR1 complexes, their apparent m.w. were calculated.
Biosensor assay to study SDS stability
A new method to study SDS stability was developed using the IAsys Auto+ resonant mirror biosensor (Affinity Sensors, Cambridge, U.K.). A carboxymethyl dextran cuvette (Affinity Sensors) was activated to produce N-hydroxysuccinimide esters using N-hydroxysuccinimide and 1-ethyl-3-(3-diethylaminopropyl)carbodiimide 34 . HLA-DR conformation-specific mAb, L243, was immobilized to this activated surface via primary amine groups. Ab was immobilized in both chambers of the cuvette to a response level of 3000–3500 arc seconds. Neither wild-type nor mutant DR1 showed significant binding to the unmodified surface under experimental conditions. Either 1 µM wild-type or 1 µM mutant DR1 was incubated at 37°C without or with 100-µM concentrations of different peptides for 24 h in PBS buffer (pH 7.4) with 1 mM PMSF and 0.025% NaN3. The reactions were then mixed with SDS such that the concentration of SDS was 0.5% and that of PBS was 80% of the original concentration. After 15 min of incubation at room temperature, 15 µl of the samples were injected using the automated sampler into both cuvettes containing 45 µl of PBS buffer. All samples were allowed to bind to the Ab for 15 min, the time determined to be sufficient for maximum binding. Unbound DR1 was then washed with PBS, and the amount of bound DR1 was determined. The same experiment was performed in two cuvettes side by side, and the results of these independent observations were averaged. Increasing the binding time to more than 15 min did not result in any significant change in the net bound DR1 signal remaining after the wash. Bound DR1 was stripped by exposing the Ab surface to 50 mM 3-(cyclohexylamino)propane sulfonic acid at pH 11.5 for 3 min. The binding capacity of the surface was monitored using the DR1/HA306–318 complex as the first and the last sample in the absence of SDS. No significant decrease in the binding capacity was observed during the course of the experiments. Of note, the concentration of monomeric SDS significantly decreases with increasing sodium ion (counter ion) concentration. Consequently, the concentration of sodium chloride during incubation of samples with SDS significantly influenced SDS stability.
SDS-PAGE
SDS-PAGE experiments were performed essentially as previously described 2 ; 0.5 µM wild-type or ßG86Y DR1 was incubated without any additional peptide or with 100 µM of different peptides at 37°C for 24 h in PBS buffer (pH 7.4) with 1 mM PMSF and 0.025% NaN3. The reaction samples mixed with equal volumes of SDS-PAGE sample buffer containing 0.2% SDS (final concentration) were either incubated for 15 min at room temperature or boiled for 3 min. These samples were then applied to 12.5% PAGE gels, and the gels were silver stained according to standard protocols. With gels for which density measurements had to be made, the gel was scanned on a Agfa Arcus laser scanner, and the intensities of the different protein bands were analyzed using the public domain National Institutes of Health Image program (developed at the U.S. National Institutes of Health and available on the internet at http://rsb.info.nih.gov/nih-image/).
Peptide dissociation kinetics
For this study, we synthesized HA peptide containing 306–318
residues of influenza virus hemagglutinin,
HA306–318 variants, and human
CLIP81–105 (Table I
). The
peptides were purified to apparent homogeneity of >95% by reverse
phase preparative HPLC, and their identities were confirmed by mass
spectrometry. The concentrations of the peptide stock solutions were
determined by ninhydrin assay. For use in kinetic studies, the peptides
were labeled with the fluorophore 7-amino-4-methylcoumarin-3-acetic
acid (AMCA) at the N-terminus using
AMCA-N-hydroxysuccinimide (Pierce, Rockford, IL). Excess
unreacted free label was removed on Sephadex G-10 columns. The labeled
peptides exhibited absorption and emission spectra characteristic of
AMCA. Labeling did not modify the binding characteristics, since
inhibition of binding of labeled peptide by unlabeled peptides
correlated linearly with the concentration of unlabeled peptides. The
labeled peptide complexes migrated similarly to the unlabeled peptide
complexes in both the SDS and nondenaturing gels, further confirming
that the fluorescent labels were nonintrusive.
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We studied the SDS resistance of DR1 complexes with different
HA306–318-derived peptides and CLIP (Table I
).
The HA306–318 variant peptides used fell into
two categories based on substitutions in single (Y308A) or multiple
(anchorless HA, YAK) anchor positions. The anchor residues of
HA306–318 were identified using the
DR1/HA306–318 crystal structure 26 . Radical
substitutions were introduced to study the extreme effects of
side-chain interactions on SDS stability and binding. Anchorless HA
peptide was designed with the aim of constructing a ligand with side
chains that would have no positive interactions with the principal
pockets that accommodate the peptide side-chain anchors. Along with
alanine substitutions at P1, P4, P7, and P9, threonine at P6 was
changed to a glycine instead of an alanine, because alanine
substitution at that position increases the binding affinity by 3- to
10-fold 35, 36 . We also used N312Q, with a substitution at a
nonanchor position, as a positive control. As a negative control, we
used ETEC, a peptide corresponding to 111–123 of CS6 subunit of
pili protein of enterotoxigenic Escherichia coli, which did
not bind to DR1.
Biosensor assay to determine the SDS resistance of complexes
We have developed a new biosensor-based assay that can be used
under controlled conditions to study the effects of different
detergents, such as SDS, on MHC class II/peptide complexes without the
use of an extrinsic label. This assay relies on the availability of the
anti-class II Ab that will recognize only the intact /ß
heterodimer and not the individual subunits. The assay consists of
allowing the MHC/peptide complexes to react with the desired denaturant
(SDS) for a predetermined amount of time (see Materials and
Methods). At this stage, the reaction will contain different
amounts of /ß heterodimers, heterodimeric aggregates, and the
dissociated subunits. The reaction mixture is then injected onto the
Ab-coupled biosensor surface for capture of the SDS-resistant
heterodimers. Once maximal stable binding is reached, a PBS wash
removes the dissociated subunits, leaving behind only the intact
heterodimer bound to the Ab. The amount of complex remaining bound to
the Ab, which is determined from the difference of the optical signal
before binding and after the buffer wash, is a measure of SDS
stability. The time necessary to reach maximal binding of the reaction
mixtures to the Ab surface was empirically determined by correlating
the binding time to the amount of the bound /ß heterodimers
remaining after the buffer wash.
Effect of peptide anchor residues on the SDS stability of DR1 complexes
Representative examples of the biosensor assay using DR1 alone,
DR1 with nonbinding ETEC, and DR1 with HA306–318
are shown in Fig. 1A. The SDS
stability of DR1 complexes with different peptides as determined by
this assay is shown in Fig. 1B. The background level of
binding in these experiments was determined with DR1 in the absence of
any peptide and in the presence of ETEC. Wild-type DR1 complexes with
HA306–318, YAK, and CLIP showed binding
significantly above background. The positive control N312Q/DR1 complex
showed binding similar to that of HA306–318/DR1,
whereas the binding of Y308A/DR1 and that of anchorless HA/DR1 were
similar to that observed at the background level. The samples of DR1
without any added peptides showed a detectable amount of binding to the
L243 Ab. This result is probably due to the presence of strong
multimeric aggregates that do not dissociate into individual subunits
under the conditions used here. At present, the nature of these
aggregates is not clear. The ability of L243 Ab to recognize these
aggregates, however, has been reported previously 4 . Silver-stained
SDS-PAGE gels confirm the presence of such molecular aggregates.
|
C, are consistent with those obtained from the biosensor
assay. Among the HA306–318 variant peptides,
HA306–318 and YAK formed SDS-resistant
complexes, while complexes of DR1 with Y308A and anchorless HA peptides
were SDS sensitive. CLIP/DR1 also formed SDS-stable complexes with DR1,
which, despite higher m.w., migrated more rapidly than the rest of the
complexes. However, the complex with short CLIP peptide
(Ii90–104) migrated similarly to typical antigenic
peptide/DR1 complexes S.K.N. and S. S-N., unpublished
observation). The different electrophoretic mobility of the CLIP/DR1
complex has been attributed to a difference in protein conformation
induced by this peptide 37 . Yet, in the SDS-free nondenaturing gel,
these complexes migrated significantly more slowly than did the
complexes of DR1 with HA306–318 variants (data
not shown). Opposite trends in the nondenaturing and SDS gels indicate
that the different migration patterns do not simply reflect differences
in the hydrodynamic radii (compactness) of the CLIP peptide complexes.
We attribute this unusual migration of CLIP complexes to
electrophoretic artifacts, such as SDS binding to the N-terminal region
of the peptide or differences between the overall charge on the
peptides. Effect of P1 residue mutation on the SDS stability of complexes with HA306–318 peptide variants
Biosensor and SDS-PAGE results with the wild-type DR1 suggested
that both the strong hydrophobic interactions between the P1 anchor and
P1 and the resulting burial of these residues are necessary and
sufficient to result in SDS stability. To account for the dependence of
SDS sensitivity on the exposed hydrophobic regions in the complex, we
hypothesized that the SDS-induced dissociation is initiated by SDS
interacting with P1. To investigate this hypothesis, we mutated
ßGly86 to Tyr, which would potentially reduce the depth
of P1 and the solvent accessibility of hydrophobic residues.
Furthermore, this mutation can be expected to increase hydrophobic
contacts between and ß subunits. Crystal structures of
DR1/HA306–318 26 , I-Ek/Hsp, and
I-Ek/Hb 38 peptide complexes were used as a basis for
designing mutant DR1. We chose to mutate ßGly86 to Tyr
for the following reasons. ßGly86 is the most prevalent
nonconserved residue in P1 among all DR alleles. I-Ek, with
P1 residues that are the same as DR1 except for ßPhe86
and ßLeu90, suggested that introducing an aromatic
residue at ß86 position of DR1 would not cause global structural
changes. The DR1/HA306–318 crystal structure
26 suggested that a tyrosine residue can fill P1. A molecular model
of empty ßG86Y DR1, energy-minimized to relieve bad contacts, also
showed that the tyrosine could be accommodated at ß86 without grossly
perturbing the other residues in the pocket (model not shown). The
hydroxyl group of tyrosine, which is likely to be pointing away from P1
assuming the aromatic ring is accommodated just like
ßPhe86 in I-Ek, could potentially
hydrogen-bond with the solvent, making the pocket more hydrophilic.
ßG86Y DR1 was expressed in insect cells and was purified using the
same procedures as those used for wild-type DR1. The gel filtration
profiles of the wild-type and mutant DR1, under nondenaturing
conditions, were qualitatively the same, with the predominant species
at an apparent molecular mass of 55 kDa and the aggregates, present in
significantly smaller amounts, at higher molecular masses of 150 and
>700 kDa (Fig. 2A). The
predominant species was the /ß heterodimer but behaved like a
protein of a higher molecular mass, reflecting the lack of compactness
in the structure in the absence of a peptide. The amount of the highest
molecular mass aggregate was lower in the mutant than in the wild type,
suggesting that the reduced hydrophobicity of the mutant decreases the
aggregate formation. As expected, this decreased aggregate formation in
the mutant was accompanied by the increased amounts of the other peaks.
|
B). This behavior of the mutant DR1 is in direct contrast
to that of wild-type DR1 that forms SDS-stable complexes with
HA306–318, YAK, and CLIP.
Additional information was obtained from SDS-PAGE experiments with the
mutant protein (Fig. 2C). In the absence of any additional
peptide, the mutant DR1 migrated as a band of approximately the
molecular mass expected for an intact /ß heterodimer and
dissociated subunits. Only Y308A and anchorless HA peptides gave rise
to SDS-stable /ß complex bands that migrated more slowly than did
the heterodimer-like band observed when no additional peptide was
added. We further investigated the identity of the protein in the band
that migrated a little more rapidly than did the peptide complex bands.
In the presence of excess peptide (Y308A) there was a significant
decrease in the intensity of that band in SDS with an accompanying
increase in the Y308A complex band (Fig. 2D), suggesting
that the protein in this band consists of /ß heterodimers. To
further investigate whether this band represents peptide-free protein
or whether it contains /ß heterodimers in complex with endogenous
peptides that may have copurified, we incubated the mutant protein with
a 100-fold molar excess of Y308A peptide and isolated the /ß
complexes by gel filtration. The protein that had an apparent molecular
mass of 55 kDa, upon Y308A binding exhibited a peak characteristic of a
51-kDa species (Fig. 2E). This shift of peaks in the gel
filtration profile, although small, was very consistent and
reproducible, suggesting that the protein achieves a more compact
structure on peptide binding. The formation of mutant/DR1 complex, as
expected, was also accompanied by a decrease in the higher order
aggregates. Half of these isolated complexes were dissociated at 37°C
in the absence of any additional peptide, while the other half were
maintained in an ice-water bath for 1 day. Then both samples were
passed through Sephadex G-50 spin columns to remove any dissociated
peptide, and the samples were further incubated at 37°C for 1 day.
The SDS-PAGE showed that the dissociated sample contained a
significantly increased amount of this band along with a concomitant
decrease in the Y308A complex compared with the undissociated sample
(Fig. 2F). Densitometric analysis of this gel showed that
the 86% of the protein of the dissociated sample was contained in this
band, while 83% of the protein was present as the Y308A complex in the
undissociated sample. This result clearly suggests that this band is
indeed peptide-free mutant /ß mutant DR1, since contaminating
endogenous peptide complexes, if any are present, must have decreased
on incubation with excess Y308A, and furthermore, there was no
possibility for these contaminating complexes to increase during the
experiment. Also, upon boiling in SDS buffer the mutant protein
migrated solely as and ß bands (Fig. 2B), indicating
that there were no other foreign proteins present in these
preparations. We speculate that the peptide-free /ß heterodimer
seen in SDS gels is an intermediate in the dissociation of mutant DR1
to individual subunits. This may be similar to the peptide-free
full-length heterodimer (Floppy) shown to exist in SDS gels in the case
of murine class II molecules 2, 3 . However, the mechanisms involved
in the formation of peptide-free mutant soluble DR1 and full-length
Floppy MHC in SDS may be different. For mutant DR1, enhanced
intersubunit contacts along with the more hydrophilic and shallow P1,
which may result in reduced binding of SDS, leads to an increased
resistance to subunit dissociation. In contrast, in the full-length
murine class II, the hydrophobic transmembrane domain contacts may
resist complete dissociation in SDS even though the extracellular
domains may have disassembled. This is consistent with the observation
that Floppy migrated more slowly than the corresponding compact peptide
complex in the SDS gel, whereas the soluble peptide-free mutant DR1
migrated more rapidly than the peptide-bound species.
Dissociation kinetics of wild-type and ßG86Y DR1 complexes
In an attempt to relate the SDS stability to longevity of
the complexes, we determined the dissociation kinetics of different
wild-type and mutant complexes (Figs. 3,
A and B). Dissociation half-lives of
HA306–318 5 and YAK complexes with the wild
type are 6 and 4.5 days, respectively, at 37°C, pH 7.4. Complexes of
Y308A and anchorless HA with DR1 are relatively fast dissociating, with
half-lives of 36 and 140 min, respectively. Y308A and anchorless HA
complexes with the mutant DR1 also dissociate with half-lives of 43 and
108 min, respectively. Reliable estimates of the half-lives of mutant
DR1 complexes with HA306–318, YAK, and CLIP
could not be obtained because of their fast dissociation rates. The
poor binding of these peptides may be due to the significant negative
steric interactions that can be expected to exist between the P1 anchor
(tyrosine in HA306–318 and YAK and methionine in
CLIP) and tyrosine at ß86 in P1 of the mutant. Thus, among wild-type
DR1/HA306–318 variants, long-lived complexes are
SDS resistant, whereas short-lived complexes are SDS sensitive.
Notwithstanding, mutant DR1 in complex with 308A and anchorless HA,
although SDS stable, are short-lived.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Results with ßG86Y DR1 further reaffirm the basis of SDS stability.
Unlike wild-type DR1, ßG86Y DR1 in the absence of peptide exists as
an ß heterodimer in SDS, suggesting that minimization of exposure
of hydrophobic residues and improved contacts between the subunits can
enhance SDS stability. Furthermore, in direct contrast to the wild
type, the SDS stability of the mutant DR1 is enhanced by peptides,
Y308A and anchorless HA, that have alanine as their first anchor. Such
behavior can be explained by a shallow P1 of the mutant protein that
may be filled by the short alkyl side chain of alanine. With wild-type
DR1, however, failure to fill the deep P1 would clearly expose the
hydrophobic regions, making DR1 heterodimer susceptible to SDS-induced
dissociation.
HA306–318 and YAK peptides, which have tyrosine at P1 as the only common anchor, result in SDS-stable complexes with wild-type DR1, whereas Y308A and anchorless HA peptides, which have alanine at P1 as the only common anchor, result in SDS-stable complexes with ßG86Y DR1. The lack of dependence of SDS stability on anchors other than the first one strongly suggests that P1 along with the main chain interactions are sufficient to result in SDS resistance of DR1 complexes.
Model for SDS stability of DR1 complexes
We propose a model in which the interactions of monomeric SDS with P1 influence the SDS resistance of the complexes. If P1 is already completely occupied by a peptide side chain that can provide better hydrophobic interactions than SDS, then the peptide stays bound, resulting in SDS-stable complexes, but if hydrophobic interactions between SDS and the pocket are more energetically favorable than those between the peptide side chain and the pocket, then SDS preferably binds to the protein, displacing the peptide and triggering subunit dissociation. The consequence of this hypothesis is that the resistance to SDS is a measure of the interaction between the peptide side chain and P1. The binding of SDS and the following dissociation is probably a highly cooperative process, as it is with proteins in general. Notably, under the conditions with which SDS-PAGE is typically performed, the concentration of monomeric SDS is far in excess of that of MHC or the peptide.
The correlation between SDS stability and longevity of the complexes
can be understood by examining the molecular bases for each of these
properties. It is apparent from the crystal structures of complexes of
both human 26, 27, 29, 40 and murine class II alleles 38, 41 that
the interactions between the peptide and the MHC pockets are either
entirely hydrophobic or consist of a combination of electrostatic and
hydrophobic components. For example, the interaction between the
Tyr308 of HA and P1 of DR1 is predominantly hydrophobic,
while some of the better characterized electrostatic charge
interactions between the Lys76 of mouse heat shock protein
and P9 of I-Ek 38 and between the Asp51 of
HEL peptide and P1 of I-Ak 41 are also aided greatly by
the significant hydrophobic contacts made by the aliphatic portion of
these peptide side chains. The extent of hydrophobicity in the binding
sites of class II molecules, in general, is also reflected by self
aggregation in the absence of peptides. Therefore, the hydrophobic
interaction between the peptide and the MHC is an important factor in
determining the longevity of the complexes. SDS serves as a convenient
tool to probe these hydrophobic interactions; hence, it is reasonable
to expect the SDS stability to correlate with the longevity of the
class II/peptide complexes. Consistent with this viewpoint, with
wild-type DR1, the peptides with bulky hydrophobic P1 anchors (HA and
YAK) form long-lived complexes as well as confer resistance against
SDS-induced dissociation. Radical substitutions of the P1 anchor of
HA306–318 with alanines result in the loss of
complex longevity and SDS stability, as observed from results with
Y308A and anchorless HA peptides. With ßG86Y DR1, the mutation
significantly increases the contacts between and ß subunits as
well as presumably decreases the solvent accessibility of the
hydrophobic residues, thereby artificially decreasing the typical
hydrophobicity of a class II binding site. The mutation also
significantly reduces the extent of interactions possible between
peptide and MHC, because pocket 1 is already partially filled.
Therefore, even though the peptides Y308A and anchorless HA do not make
any more positive interactions with the mutant than they do with the
wild type, as evident from similar, short lifetimes of these peptide
complexes with either protein, the mutant complexes are SDS stable.
Mutant DR1 may not represent any naturally occurring MHC protein, since
the mutation severely curtails the binding of the peptide with P1,
which is seemingly the only critical interaction for stabilizing the
DR1 complex, resulting in an MHC molecule incapable of forming a stable
complex with any peptide. Therefore, the dissociation data with the
mutant can be used to further cement the basis of SDS stability being a
measure of buried hydrophobicity and not to be misconstrued as an
exception to the correlation between the dissociation rates and SDS
stabilities observed with the natural alleles.
One consequence of the proposed model is that the SDS stability is a measure of the affinity between the peptide and the naturally occurring MHC molecule. Although our experiments have dealt only with DR1, since prominent hydrophobic interactions between the peptide and the MHC seem to be a common feature of all alleles, we speculate that this model may well be applicable to class II molecules in general. Even small differences in this particular interaction are reflected in SDS-PAGE and biosensor assays, although they are not always detected by typical competition binding assays. A few reports, in which competition assays were used to determine peptide affinities, suggest that there is no strict correlation between SDS stability and complex affinity 22, 24, 25 . This discrepancy may be due to the complexity of the peptide binding and the failure of the typical competition assays always to reflect the intrinsic affinity of the peptide. The affinity measurements are complicated by several technical and conceptual difficulties related to the measurement of association rates with the heterogeneous pool of MHC complexes isolated from the surface of APCs 42, 43 as well as the soluble molecules made in insect cells without their natural ligand 49 . However, the dissociation rates of the class II peptide complexes are a more accurate reflection of the intrinsic affinity of the peptide. Consistent with this viewpoint, the sensitivity of class II complexes to dissociation by HLA-DM, much like SDS stability, also has a stricter correlation with the dissociation rates of the complexes, but exhibits no clear trend with the affinities measured by competition assays 44 .
Correlation of SDS sensitivity with exposed hydrophobic residues
The mechanistic basis for the SDS stability of the class II peptide complex can be better understood if SDS is considered as an active participant in this interaction. SDS primarily binds to hydrophobic regions of proteins 45 . SDS sensitivity has been suggested to indicate incomplete folding of MHC proteins during synthesis and intracellular transport 6 . Incompletely folded proteins, as intermediates in the folding pathway, are often characterized by the presence of exposed nonpolar patches 46 . It is possible that SDS binding to these exposed hydrophobic residues leads to subunit dissociation, thus providing a convenient assay for probing the maturity of the MHC molecules.
The result with wild-type DR1 complex with CLIP peptide is seemingly anomalous. CLIP, which has a methionine anchor, forms a SDS-stable complex that dissociates rapidly 47 . The SDS stability of CLIP complexes is likely to result only from P1 interactions, whereas the affinity of the entire peptide for MHC can be affected by the interactions with other peptide residues. The N-terminal segment (Ii81–89) has been shown to facilitate rapid release of CLIP 47 . The substitution of proline with an alanine at the P6 anchor of CLIP increases binding affinity for DR1 by at least 10-fold 48 , indicating that the peptide makes destabilizing interaction at least with one of the pockets. Therefore, the SDS stability of CLIP/DR1 complexes is probably a measure of strong affinity between the methionine side chain and P1. However, significant destabilizing interactions in the other pockets and regions outside the Ag binding site decrease the overall affinity. Interestingly, CLIP complexes of all the other murine or human alleles reported are shown to be SDS sensitive. This anomalous situation with DR1 can be further explained in terms of Ii function in vivo and the molecular basis of SDS stability discussed here. One of the functions of Ii is to prevent empty class II binding sites from self aggregating 14 , although with the additional requirement that the CLIP region of Ii should dissociate rapidly. P1 of DR1 contains the most extensive array of exposed hydrophobic residues among any binding pocket of known alleles and hence would be the most vulnerable site for self aggregation. The methionine side chain of CLIP with its strong interactions with the P1 residues buries the otherwise exposed hydrophobic regions, thereby protecting against self aggregation and also consequently resulting in SDS resistance. Unfavorable contacts from other peptide residues, however, help satisfy the requirement of short-lived DR1/CLIP complexes.
Juxtaposition of hydrophobic patches and positively charged residues
such as arginine and lysine on the proteins is known to increase SDS
binding significantly 45 . The combination of both ionic and
hydrophobic interactions may have to be invoked to explain the behavior
of different class II alleles toward SDS. For example, P1 Asp is an
extremely critical anchor in the binding of different peptides to
I-Ak 25 . Hen egg lysozyme (HEL) peptide 50-62 with an
alanine substitution at this position leads to a drastic loss in
binding affinity as well as SDS stability, much like the behavior of
Y308A peptide with DR1. Consistent with the solution studies, the
crystal structure of I-Ak with HEL peptide 50-62
shows that the Asp P1 side chain fits exactly into P1, with the
carboxyl group forming a salt bridge with guanidium of Arg 52 while
the aliphatic portion of the side chain makes hydrophobic contacts with
Phe 24 and Phe 54 41 . Even though there are fewer hydrophobic
residues lining P1 in I-Ak than there are in DR1, the
presence of positively charged Arg 52 in close proximity makes it a
potentially good SDS binding site. Peptides with first anchors that do
not fit P1 as well as Asp does can presumably be easily displaced by
SDS, resulting in significantly reduced SDS stability.
The biosensor assay introduced here demonstrates that analysis of the behavior of class II complexes in SDS is not restricted to SDS-PAGE gels and can be complemented by more controlled techniques in solution. The study of interactions of class II complexes with various amphiphiles differing in their hydrophobicity and ionic character will further clarify the issues involved.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Scheherazade Sadegh-Nasseri, Department of Pathology, Johns Hopkins University School of Medicine, 664E Ross Building, Baltimore, MD 21205. E-mail address:
3 Abbreviations used in this paper: HA306–318, peptide containing 306–318 residues of influenza virus hemagglutinin; CLIP, class II major histocompatibility complex-associated invariant chain peptides (residues 81–105); P1, pocket 1; AMCA, 7-amino-4-methyl-coumarin-3-acetic acid; ETEC, a peptide corresponding to 111–123 of CS6 subunit of enterotoxigenic Escherichia coli pili protein.
Received for publication July 8, 1998. Accepted for publication December 10, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ß heterodimers in the absence of antigenic peptide. Cell 68:465.[Medline]
ß dimers and facilitates peptide loading. Cell 82:155.[Medline]
ß heterodimers in endosomes. EMBO J. 11:411.[Medline]
This article has been cited by other articles:
|
|
 |
|
  A. Ferrante and J. Gorski Cooperativity of Hydrophobic Anchor Interactions: Evidence for Epitope Selection by MHC Class II as a Folding Process J. Immunol., June 1, 2007; 178(11): 7181 - 7189. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Marin-Esteban, K. Falk, and O. Rotzschke "Chemical Analogues" of HLA-DM Can Induce a Peptide-receptive State in HLA-DR Molecules J. Biol. Chem., December 3, 2004; 279(49): 50684 - 50690. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Mirshahidi, L. C. K. Ferris, and S. Sadegh-Nasseri The Magnitude of TCR Engagement Is a Critical Predictor of T Cell Anergy or Activation J. Immunol., May 1, 2004; 172(9): 5346 - 5355. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. J. Carven, S. Chitta, I. Hilgert, M. M. Rushe, R. F. Baggio, M. Palmer, J. E. Arenas, J. L. Strominger, V. Horejsi, L. Santambrogio, et al. Monoclonal Antibodies Specific for the Empty Conformation of HLA-DR1 Reveal Aspects of the Conformational Change Associated with Peptide Binding J. Biol. Chem., April 16, 2004; 279(16): 16561 - 16570. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. W. Anderson and J. Gorski Cutting Edge: TCR Contacts as Anchors: Effects on Affinity and HLA-DM Stability J. Immunol., December 1, 2003; 171(11): 5683 - 5687. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. C. Doebele, A. Pashine, W. Liu, D. M. Zaller, M. Belmares, R. Busch, and E. D. Mellins Point Mutations in or Near the Antigen-Binding Groove of HLA-DR3 Implicate Class II-Associated Invariant Chain Peptide Affinity as a Constraint on MHC Class II Polymorphism J. Immunol., May 1, 2003; 170(9): 4683 - 4692. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. S. Arneson, J. F. Katz, M. Liu, and A. J. Sant Hydrogen Bond Integrity Between MHC Class II Molecules and Bound Peptide Determines the Intracellular Fate of MHC Class II Molecules J. Immunol., December 15, 2001; 167(12): 6939 - 6946. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Mirshahidi, C.-T. Huang, and S. Sadegh-Nasseri Anergy in Peripheral Memory Cd4+ T Cells Induced by Low Avidity Engagement of T Cell Receptor J. Exp. Med., September 17, 2001; 194(6): 719 - 732. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.-L. Chou and S. Sadegh-Nasseri Hla-Dm Recognizes the Flexible Conformation of Major Histocompatibility Complex Class II J. Exp. Med., December 18, 2000; 192(12): 1697 - 1706. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. K. Sato, J. A. Zarutskie, M. M. Rushe, A. Lomakin, S. K. Natarajan, S. Sadegh-Nasseri, G. B. Benedek, and L. J. Stern Determinants of the Peptide-induced Conformational Change in the Human Class II Major Histocompatibility Complex Protein HLA-DR1 J. Biol. Chem., January 21, 2000; 275(3): 2165 - 2173. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Carrasco-Marin, S. Petzold, and E. R. Unanue Two Structural States of Complexes of Peptide and Class II Major Histocompatibility Complex Revealed by Photoaffinity-labeled Peptides J. Biol. Chem., October 29, 1999; 274(44): 31333 - 31340. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Rotzschke, K. Falk, J. Mack, J. M. Lau, G. Jung, and J. L. Strominger Conformational variants of class II MHC/peptide complexes induced by N- and C-terminal extensions of minimal peptide epitopes PNAS, June 22, 1999; 96(13): 7445 - 7450. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. C. Korb, S. Mirshahidi, K. Ramyar, A. A. Sadighi Akha, and S. Sadegh-Nasseri Induction of T Cell Anergy by Low Numbers of Agonist Ligands J. Immunol., June 1, 1999; 162(11): 6401 - 6409. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
100-fold molar excess of relevant unlabeled peptides
at 37°C in PBS (pH 7.4) for various times. Then the free peptides
were separated by centrifugation on Sephadex G-50 columns. The amount
of labeled complex present after dissociation was determined by
exciting AMCA at 350 nm and measuring the fluorescence emission at 445
nm in a spectrofluorometer. The fluorescence of the labeled complex
before dissociation (f0) is arbitrarily assigned a value of
100%. The fluorescence of labeled complex after dissociation for
various times is expressed as a percentage of f0. The
dissociation data are fitted to a single exponential curve.