| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* National Centre for Cell Science, Maharashtra, India;
Institute of Microbial Technology, Chandirgarh, India; and
Department of Biotechnology, Devi Ahilya Vishwavidyalaya, Indore, India
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and IL-2), which can
take part in cell-mediated immunity, or they can strengthen the
humoral arm of the immune system by secreting Th2-type cytokines
such as IL-5, IL-4, IL-6, and IL-13 (3). Thus, the
nature of the costimulatory molecules and their regulation appears to
be of considerable interest. Available data suggest that various
receptor ligand pairs, expressed on the surface of T cells and APCs,
have shown to possess costimulatory properties. These include
ICAM-1/LFA-1 (4), VCAM-1/VLA-4 (5), HSA
(6), B7-CD28 (7), B7.2-CD28 (8),
and LFA-3/CD2 (9). Although it is considered that B7-CD28
interaction is central in the steps of T cell activation, studies on
CD28-deficient mice (10) have shown that this may not be a
limiting factor in T cell response. Furthermore, compelling evidence is
being rapidly added to the existing literature about the involvement of
hitherto unknown molecules with costimulatory functions, which also
have the ability to initiate and propel the T cell response
irrespective of the B7-CD28 pathway. To name a few, the 4-1BB
(11), OX 40 (12), B7 RP (13),
and B7-DC (14) are the latest addition to this
list. Working on strategies to find out additional molecules with T cell costimulatory properties, which not only activate the T cells but also show a distinct bias in Th response, we have earlier reported the identification and characterization of few novel costimulatory molecules present on the surface of APCs. Interestingly, a 150-kDa protein (M150) isolated from the membrane fraction of activated macrophages, the APC that is known to stimulate Th1 response, was shown to function as a Th1-specific costimulatory molecule (15, 16). Furthermore, two new costimulatory molecules, viz B1 (17) and B3 (18), were also identified on the surface of LPS-activated B cells and were found to elicit Th2 response. In continuation of our ongoing studies, we demonstrate in this study that the protein of 99–105 kDa, termed as B2, is expressed on LPS-activated B cells and specifically induces Th2 responses from cultured CD4+ T cells. This protein (B2) was also found to be expressed on thioglycolate-elicited macrophages and bone marrow-differentiated dendritic cells (data not shown). Peptide sequence homology between B2 and gp96, detailed biochemical characterization of B2, and codistribution of B2 and gp96 on LPS-activated B cell surface have revealed the identity of B2 as gp96. The significant role of gp96, in immunotherapy of cancers, has been well documented as it activates tumor-specific CTL responses (19). However, subsequent studies have also revealed that the gp96-engineered tumor rejection is not due to gp96 per se, but rather due to tumor-derived peptides, which are found to be associated with it (20). Furthermore, recently it has also been demonstrated that the tumor-derived peptide-stripped gp96 activates CD8+ T cells both in vivo and in vitro (21). However, a definite role of gp96 in T cell activation is yet to be offered in detail. In the foregoing account, we demonstrate for the first time that gp96 binds on anti-CD3-activated CD4+ T cells, through CD91, costimulates, and further skews the Th differentiation toward Th2 phenotype. Furthermore, the MHC-independent expression of gp96 on the APC has also been revealed by our confocal microscopic studies. Taken together, our study not only highlights a novel feature of the stress-related protein, but also emphasizes a direct link between stress and the regulation of immune responses.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Female BALB/c mice were obtained from the National Institute of Nutrition (Hyderabad, India). Mice were maintained in the experimental animal facility of National Center for Cell Science, India, and used at 6–8 wk of age.
Medium
CD4+ T cells, B cells, and A20 (B cell lymphoma) cells were cultured in RPMI 1640 (Life Technologies, Grand Island, NY) supplemented with penicillin (70 µg/ml), streptomycin (100 µg/ml), glutamine (4 mM), 2-ME (50 µM), sodium pyruvate (1 mM), HEPES (20 µM), and 10% heat-inactivated FCS (Life Technologies).
Preparation of primary cells
B cells. A single cell suspension of splenocytes was prepared in balanced salt solution (pH 7.2). The RBCs were depleted by hemolytic Gey’s solution. After incubating the cells with B cell enrichment Ab mixture (Stem Cell Technologies, British Columbia, Canada), the cells were negatively separated as per manufacturer’s protocol. Cells obtained in this manner were 95% pure. These splenic B cells of different densities were separated by discontinuous Percoll gradient centrifugation. The cells collected from 70–66% interface layer of 50, 60, 66, and 70% gradient were resting B cells. For LPS activation, these cells were cultured at a concentration of 5 x 106 cells/3 ml/well of a six-well plate (Corning, Corning, NY) with 10 µg/ml LPS from Salmonella typhosa (Sigma-Aldrich, St. Louis, MO) for 72 h in the presence of 5% CO2.
T cells. RBC-deprived splenocytes were incubated in CD4+ T cell enrichment Ab mixture (Stem Cell Technologies) and were prepared by following manufacturer’s guidance. The purity of T cells was over 95%, as analyzed by FACSVantage (BD Biosciences, San Jose, CA). The Th1 and Th2 subsets were prepared as described by Flavell and colleagues (22). In brief, for the generation of Th1 cells, purified splenic CD4+ T cells were cultured for 7 days in the presence of 4 µg/ml Con A (Sigma-Aldrich), 1 ng/ml IL-2 (BD PharMingen, San Diego, CA), and 5 ng/ml IL-12 (BD PharMingen). Th2 cells were generated by culturing the CD4+ T cells 7 days in the presence of 4 µg/ml Con A (Sigma-Aldrich), 1 ng/ml IL-2 (BD PharMingen), and 10 ng/ml IL-4 (BD PharMingen). After 7 days, cells were washed twice with RPMI 1640 and further cultured for 48 h in the presence of cytokines specific for the generation of either Th1 or Th2 subsets. Th1 subsets were reactivated with 1 ng/ml IL-2 and 5 ng/ml IL-12, whereas Th2 subsets were reactivated with 1 ng/ml IL-2 and 10 ng/ml IL-4. The identity of Th1 and Th2 phenotype was established by RT-PCR of subset-specific signature cytokines (data not shown).
Isolation of B2
LPS-activated B cell surface proteins were prepared, as described elsewhere (17, 18). In brief, cells were homogenized in the homogenizing buffer along with protease inhibitor mixture (10 µg/ml leupeptin, 10 µg/ml aprotinin, 10 mM iodoacetamide, 10 µg/ml antipain, 10 µg/ml pepstatin, 10 µg/ml chymostatin, and 1 mM PMSF) and then were subjected to a brief centrifugation at 700 x g for 10 min at 4°C to remove the nuclear fraction. The supernatant was centrifuged at 110,000 x g for 2 h at 4°C. The pellet was solubilized in buffer containing 1% Triton X-100, 20% glycerol, and 20 mM Tris-HCl (pH 7.5) for 8 h at 4°C, followed by centrifugation at 100,000 x g for 1 h at 4°C. The supernatant, containing membrane proteins, was separated using 10% SDS-PAGE. Isolation of the protein band of the interest was done as described in our earlier studies (17, 18). Briefly, protein band was located by staining a strip of the gel with silver nitrate, and the appropriate unstained portion was crushed and eluted with 1% SDS, 100 mM NH4HCO3, 50 mM Tris-HCl, 0.1 mM EDTA, and 150 mM NaCl (pH 8.0) at 37°C for 48 h. After filtration and centrifugation to remove the polyacrylamide particles, the solution was dialyzed against 0.1% SDS, 10 mM NH4HCO3, 10 mM Tris-HCl, 0.1 mM EDTA, and 50 mM NaCl (pH 8.0) at 4°C for 24 h. SDS was removed using Extracti-D gel column (Pierce, Rockford, IL), and the protein was estimated using bicinchoninic acid kit (Pierce).
Two-dimensional electrophoresis and reverse-phase HPLC
The reagents for two-dimensional gel electrophoresis were prepared essentially according to Amory et al. (23) and performed as advocated by Penin et al. (24). For reverse-phase HPLC, isolated B2 molecule was diluted with 0.5% trifluoroacetic acid and loaded onto a microbore HPLC column (C8) (Aquapore RP 300, Brownlee Columns; Applied Biosystems, Foster City, CA) and monitored.
Protein sequencing
Isolated B2 molecule was blotted onto ProtBlot (Applied Biosystems) in 3-[cyclohexamino]-1-propane sulfonic acid buffer (pH 11.0), followed by staining in 0.2% Ponceau S and 1% acetic acid, and washed with PBS. The protein was then digested with trypsin, and a sample equivalent of 2 pmol from one of the digested fractions was subjected to internal sequencing up to 13 and 15 residues on Applied Biosystems (model 492 A) Procise Sequencer (at Protein Sequencing Facility, Worcester Foundation for Experimental Biology, Shrewsbury, MA).
Negative staining of liposomes
Liposomes were prepared by a reverse-phase evaporation method, as described elsewhere (17), and a sample droplet was picked up by touching a carbon-coated grid from 20 µl of aqueous liposome suspension. After allowing the excess liquid to drain off, the grid was gently dipped in 1% phosphotungstic acid (pH 7.0) and dried by blotting on a filter paper. Grids were observed in a transmission electron microscope (JEOL 1200 EXII, Tokyo, Japan), and representative fields were photographed.
Coupling of iodinated B2 to liposomes
125I labeling of B2 was performed using IODO bead method with PD-10 column (Pharmacia Biotech, Uppsala, Sweden). Iodinated samples were then precipitated by TCA. A 1:500 ratio of protein samples to liposomes was placed in a dialysis bag and dialyzed for 96 h at 4°C against 10 mM Tris (pH 8.0), 0.01 mM EDTA, and 10 mM NaCl with at least 12 changes. The contents of the dialysis bag were spun down for 2 h at 4°C at 178,000 x g. The pellet was then dissolved in PBS and later passed through a Sephadex G-50 mini column to remove unliposomized protein. Obtained sample (2 ml) was layered on top of a discontinuous gradient of 5–40% sucrose (w/v) in 10 mM Tris-HCl (pH 6.8), 150 mM NaCl, and 0.1 mM EDTA. The sample was then centrifuged overnight at 98,000 x g in a Beckman Coulter (Fullerton, CA) SW28 rotor at 4°C. A 2-ml sample was collected from the 5 and 10% interface, washed in 5 vol PBS by centrifuging at 178,000 x g for 2 h at 4°C. The pellet was dissolved in PBS and used. The protein content coupled to liposomes was estimated, after lysis with 1% SDS, by the bicinchoninic acid method.
Binding assay using electron-microscopic autoradiography
The radiolabeled (25) and liposomized samples were incubated with CD4+ T cells (preactivated for 30 min with plate-bound anti-CD3 at 10 µg/ml concentration) or A20 cells for 30 min at 37°C. After incubation, the cells were washed thrice with cold PBS. To the pellet, an equal volume of low melting agarose (Sigma-Aldrich) was added and allowed to gel, and the latter was cut into 1-mm3 pieces. Trapped cells were fixed in 1% paraformaldehyde, 1% glutaraldehyde in PBS at 4°C for 1 h, followed by washing with PBS and postfixing in 1% osmium tetraoxide for 90 min at 4°C in the dark. PBS-washed samples were passed through graded acetone series, and then were embedded in epoxy resin (Bio-Rad, Hercules, CA). Sections cut on Ultracut S (Vienna, Austria) were coated in darkness with photographic emulsion (Ilford Imaging (Paramus, NJ) nuclear L4 emulsion) and incubated in darkness for 2–3 wk in a dessicator at 4°C. The autoradiographs (25) were developed and stained with aqueous uranyl acetate and lead citrate, and were observed in a transmission electron microscope (JEOL 1200 EXII).
Competitive binding assay
For competition binding, a range of nonradioactive (cold) but liposomized dilutions of B2 was allowed to bind on the anti-CD3-activated CD4+ T cells. After washing, 125I-labeled B2 was added to a final volume of 200 µl and incubated for 2 h at 4°C with gentle agitation at 15-min interval. After extensive washing, the cell pellets were subjected to gamma counting (Beckman Coulter).
Raising and purification of anti-B2 polyclonal Ab
B2-specific peptide (GVVDSDDLPLNVSRE) as denoted by protein sequencing was linked with keyhole limpet hemocyanin (KLH5; Sigma-Aldrich) in the presence of 0.5% gluteraldehyde (1 mol peptide/50 aa of carrier proteins) in 2 ml reaction volume. After linking, the reaction was stopped with 1 M glycine. The peptide-KLH mixture then dialyzed for 12 h against PBS with repeated changes. A total of 150 µg of KLH-linked peptide was mixed with equal volume of CFA (Bangalore Genei, Pune, India) and immunized s.c. in 12-mo-old female rabbits. After 4 wk, the rabbits were boosted with 75 µg of Ag. Five days after first boost, anti-B2 antisera was collected and anti-KLH Abs were removed by using anti-KLH Ab removal column (Bangalore Genei). The sera were further affinity purified using IgG purification kit (Bangalore Genei).
Immunoblotting
Cell lysate (15 µg) or immunoprecipitated samples were separated by 10% SDS-PAGE and transferred onto a nitrocellulose membrane (Bio-Rad). The immunodetection was performed with an anti-B2 (affinity-purified polyclonal rabbit anti-B2 IgG) or anti-gp96 Ab (goat anti-gp96 C-terminal specific polyclonal Ab; Santa Cruz Biotechnology, Santa Cruz, CA) and with control Abs (normal rabbit IgG and normal goat IgG; Bangalore Genei), followed by HRP-coupled secondary Ab incubation (anti-goat HRP, Santa Cruz Biotechnology, and anti-rabbit HRP, Bangalore Genei), and visualized using chemiluminescence substrate (Pharmacia Biotech). For our initial studies, anti-gp96 Ab was kindly provided by M. Green from St. Louis University School of Medicine (St. Louis, MO).
Immunoprecipitation
Ice-cold PBS-washed cells were lysed in lysis buffer containing 0.5% Triton X-100, 10 mM Tris-HCl (pH 8.0), and 140 mM sodium chloride at 4°C for 1 h. The cell lysate was centrifuged at 10,000 x g for 30 min at 4°C to remove the nuclear debris. To 100 µl of supernatant, related Ab was added either at the final concentration of 15 µg/ml or at mentioned dilutions. The Ab-supernatant mixture was incubated for 2 h at 4°C, followed by adding 20 µl of protein G-agarose (Life Technologies), and further incubated for 2 h at 4°C on a rocker. At 4°C, the slurry was washed twice with TSA solution (10 mM Tris-HCl, pH 8.0, 140 mM sodium chloride, 0.01% sodium azide) with 0.1% Triton X-100, followed by once with TSA solution and once with 50 mM Tris-HCl (pH 6.8). The slurry was boiled for 5 min in Laemmli sample buffer. After a short spin, the supernatant was electrophoresed on the 10% SDS-PAGE, and the bands were visualized either by silver staining or by immunodetection on a nitrocellulose.
Surface labeling and confocal microscopy
B cells (1 x 105) were harvested from cultures, washed, and suspended in FACS buffer (PBS with 2% FCS and 0.1% sodium azide). Cells were first treated with normal hamster sera to prevent any nonspecific binding of the Abs. Then, in FACS buffer cells were incubated with anti-gp96 (Santa Cruz Biotechnology) and/or anti-I-Ad PE Ab (BD PharMingen) for 45 min at 4°C. After two washes with FACS buffer, followed by staining for 30 min at 4°C with required fluorescence-conjugated secondary Ab, cells were washed twice and fixed with 1% paraformaldehyde and then were visualized with a confocal laser-scanning microscope (Zeiss, Jena, Germany) at a magnification of x63.
For competitive binding between B2 and gp96, plate-bound anti-CD3-activated (10 µg/ml, 30 min) Th cells (1 x 105) were first fixed with 4% buffered paraformaldehyde for 20 min at room temperature, washed, and then incubated along with FITC-conjugated B2 (B2-FITC) and unlabeled gp96 (Immatics Biotechnologies, Tubingen, Germany). In parallel experiments, activated Th cells were also incubated with FITC-conjugated B2 along with anti-CD91 Ab (Progen, Heidelberg, Germany). Both the competitors were used at 35-fold molar excess concentration. OVA and B2 were conjugated to FITC following standard laboratory protocol.
Additional studies were performed to ascertain whether B2 is codistributed with gp96. For this, 1 x 105 LPS-activated B cells were incubated simultaneously with rabbit anti-B2 and biotinylated goat anti-gp96 polyclonal Ab (10 µg/ml) for 45 min at 4°C. Cells were washed and stained with a mixture of anti-rabbit FITC and streptavidin-PE for 30 min at 4°C. After incubation, cells were washed and were visualized with a confocal laser-scanning microscope at a magnification of x63.
Inhibitory binding studies between anti-B2 and anti-gp96 Ab were also performed on LPS-activated B cells. For this, cells (1 x 105) were first incubated with biotinylated goat anti-gp96 Ab (10 µg/ml) for 45 min at 4°C. After washing, the cells were then incubated with the rabbit anti-B2 Ab (10 µg/ml) and stained with anti-rabbit FITC. In parallel experiments, cells were first incubated with anti-B2 rabbit Ab (10 µg/ml) for 45 min at 4°C. After washing, the cells were then incubated with the biotinylated goat anti-gp96 Ab (10 µg/ml) and stained with streptavidin-PE.
T cell proliferation assay and quantitation of cytokines
Purified CD4+ T cells (1 x
105 cells/well) were cultured in the presence or
absence of anti-CD3 (1 µg/ml; BD PharMingen) in 96-well
round-bottom plate (Corning) in 200 µl final volume. The wells were
precoated with anti-CD3 only when liposomized proteins were used in
the culture. Otherwise, while using soluble anti-CD3, before adding
the cells, wells were precoated with active or heat-inactivated gp96
(Immatics Biotechnologies) at various concentrations and preincubated
with or without rabbit anti-B2 polyclonal Ab or normal rabbit IgG
(28 µg/ml). The culture was pulsed with 1 µCi/well
[3H]thymidine (NEN, Boston, MA) for last
12 h of 60-h culture, followed by assessing the radioactivity in a
liquid scintillation counter (Packard, Downers Grove, IL). The
supernatants collected at 48 h and were measured for the level of
different cytokines by using Quantikine M kit (R&D Systems,
Minneapolis, MN). Lymphokine bioassay was done as described elsewhere
(17, 18). In brief, the supernatants from the control and
experimental wells were collected after 22 h and tested for IL-2
and IL-4 using HT-2 cell line (1 x 104
cells/well), IFN- on WEHI-279 (2 x 104
cells/well), and IL-5 on resting splenic B cells (1 x
105 cells/well). The cells were incubated in
RPMI, 10% FCS (100 µl/well in the case of IL-2, IL-4, and IFN-,
and 200 µl/well in the case of IL-5). The cells were then pulsed with
1 µCi [3H]thymidine during the last 6 h
for IL-2, IL-4, and IFN-, and last 16 h in the case of IL-5
culture. The incorporated radioactivity was monitored by liquid
scintillation counting. The values are expressed in terms of units per
milliliter derived from the standard curve.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our approach for identifying the possible existence of
Th2-specific costimulatory molecules has been to resolve proteins from
the membrane of activated APCs by gel electrophoresis, followed by
elution of individual protein bands. To provide native context, each
protein was reconstituted into lipid vesicles, and these reconstituted
(liposomized) proteins were then tested for their costimulatory
activity. B2 was identified, as a 99- to 105-kDa protein band, when
LPS-activated membrane proteins of B cells were subjected to analyze by
10% SDS-PAGE (Fig. 1
A). The
homogenity of the B2 was checked by two-dimensional gel
electrophoresis, in which it appeared as a single spot (Fig. 1B). In addition, the same protein, when subjected to
reverse-phase HPLC, showed a single prominent peak at 32.9 min in the
chromatogram (Fig. 1C), demonstrating thereby that this
protein preparation was homogeneous in its present form. Purified B2
was then reconstituted in liposomes, following which it was added in
varying amounts to CD4+ T cells being cultured on
anti-CD3-coated plates. As shown in Fig. 2A, liposomized B2 was indeed
capable of inducing a dose-dependent proliferative response from T
cells. Optimum proliferation, which was 20-fold more than the basal
level, was observed with B2 concentration of 1 µg/ml, while higher
doses of B2 showed an inhibitory effect (Fig. 2A). The
specificity of this response is particularly evident from the fact that
no such effect was observed in the several control groups that were set
up in parallel and are shown in Fig. 2B. Also, this activity
was not contingent upon its presentation in liposomized form, as
similar results were obtained when the protein was used in immobilized
form in T cell assays (data not shown). Controls such as anti-CD3
alone, liposomized B2 alone, or control liposome (not containing B2)
did not influence the CD4+ T cell proliferation
and lymphokine secretion over the basal level. However, as shown in
Fig. 2C, mean of three independent experiments,
costimulation with B2 resulted in a specific (near 10-fold) increase in
the production of IL-4 and IL-5. As opposed to this, levels of IL-2 and
IFN- in the same culture supernatants were virtually unaffected
(Fig. 2C). These experiments, therefore, suggest that B2 is
a costimulatory molecule that specifically induces the naive
CD4+ T cells toward a Th2 phenotype.
|
|
Our above observations of the costimulatory activity of B2 also
suggested that the receptor for it must exist on the surface of
T cells. To test this, we labeled B2 with 125I
before its reconstitution in liposomes (Fig. 3, A and B), and
the preparation was incubated with T cells, and binding, if any, was
monitored by electron-microscopic autoradiography. As shown in Fig. 3C, the radiolabeled B2 incorporated in liposomes was indeed
found to bind anti-CD3-activated T cells, being distributed
uniformly along the cell surface, as denoted by arrows. As opposed to
this, in the control experiment, no specific labeling of the T cells
could be detected when liposome-incorporated
125I-labeled anti-mouse IgM was used for the
binding studies (Fig. 3D). However, this later preparation
was found to bind to A20 cells, as denoted by arrows (Fig. 3E). This result confirmed that the absence of such binding
to T cells (Fig. 3D) was not due to any defect at the level
of the liposome preparation, but the specificity of the interactions
itself.
|
). These
cumulative results, therefore, suggest the existence of receptor of B2
on the surface of the T cells. Another interesting feature of the data
shown in Fig. 4 is the fact that the binding of B2 is optimal only in
anti-CD3-activated T cells. In contrast, B2 binding to resting T
cells was reduced by nearly 20-fold (Fig. 4). These findings therefore
seem to imply that the putative receptor for B2 is up-regulated upon
cross-linking of the TCR/CD3 complex.
|
To further characterize B2, we performed a proteolytic digestion
with trypsin (as its N terminal was found to be blocked), and two of
the fragments were subjected to the amino acid sequence analysis, as
shown in Fig. 5. Interestingly, both of
the obtained sequences exhibited complete identity with murine
glucose-regulated protein of 94 kDa, or more commonly referred as gp96.
The alignment of amino acid sequences of gp96 and B2 is shown in Fig. 5. This unanticipated relationship between B2 and gp96 has prompted
further analysis. Therefore, to confirm the relationship between B2 and
gp96, we next performed criss-cross immunoprecipitation experiments
using lysates derived from a B cell lymphoma (A20). For these
experiments, either a commercially available anti-gp96 Ab
preparation or those generated against B2 peptide were used. When the
A20 cell lysates were immunoprecipitated with either anti-gp96 or
anti-B2 Ab, a protein band migrating at an identical molecular mass
was obtained in both cases (Fig. 6A, lanes 1 and
7), whereas this band was absent when control IgG was used
instead (Fig. 6A, lanes 3 and 8). The
identical molecular mass of B2 or gp96 could further be evaluated by
the position of standard protein molecular mass marker of 97.4
kDa (Fig. 6A, lane 6). Furthermore, when the
supernatant of protein G-agarose-bound Ag-Ab (anti-gp96 polyclonal)
complex was analyzed on the SDS-PAGE, any protein band equivalent to B2
mass could not be visualized even by silver staining (Fig. 6A, lane 2). As opposed to this, control Ab
failed to immunoprecipitate the protein equivalent to B2 mass during
the experiment (Fig. 6A, lane 4). The position of
B2, as one of the SDS gel-separated proteins of A20 cell lysate, is
shown in lane 9 of Fig. 6A. Importantly,
anti-B2 Ab failed to immunoprecipitate any protein band from A20
lysates that had been first incubated or precleared with saturating
concentration of anti-gp96 (Fig. 6A, lane 5).
Conversely, preclearing the cell lysate with saturating concentration
of anti-B2 was also found to deplete the gp96 protein, as
determined in a subsequent immunoprecipitation analysis with
anti-gp96 (Fig. 6A, lane 10). The striking
homology between gp96 and B2 was further confirmed by
immunoprecipitation, followed by Western blot experiments. Thus, the
immunoprecipitated product obtained with anti-B2 was cross-reactive
with anti-gp96 (Fig. 6B). Similarly, the protein product
immunoprecipitated with anti-gp96 was also recognized by
anti-B2 Ab (Fig. 6C).
|
|
and 6 strongly support that the
protein B2, isolated from the membrane of LPS-activated B cells, was in
fact the stress protein gp96. Cell surface localization of gp96 and B2
As B2 was isolated from the surface of LPS-activated B cells and
the essential prerequisite of a costimulatory molecule is that it
should be expressed on surface of APCs, we therefore examined the
expression of gp96 on activated B cells. LPS-regulated, nearly 10 times
higher, cell surface expression of gp96 had previously been observed on
murine B cells (26); however, it was essential to check
its expression in the context of MHC molecule, as it had also been
reported that gp96 being a tumor-specific transplantation Ag, carrying
immunogenic peptide of the tumor cells, comes to the cell surface
through MHC I (27). In addition, it had also been
described that B cells were unable to present the gp96-Ag complex
through MHC I (19). These observations caused us to
examine the expression of gp96 on B cells in the context of MHC II. Our
confocal microscopic studies revealed that gp96 could be expressed on
the surface of the splenic B cells and is dependent on LPS stimulation
(Fig. 7, upper and lower
panel). We next investigated whether B2 and gp96 are codistributed
on LPS-activated B cells. Images obtained through confocal microscope
showed the codistribution of B2 with gp96 (Fig. 8, A–D). Furthermore, the
cell surface binding of anti-B2 Ab (Fig. 8E) was
efficiently blocked (almost 80%) by anti-gp96 Ab, when the cells
were preincubated with the anti-gp96 Ab (Fig. 8F).
Similarly, preincubation with anti-B2 Ab considerably prevented the
binding of anti-gp96 Ab to the LPS-activated B cells (Fig. 8, G and H). Experiments with control Abs did not
show any effect on such competitive binding. Moreover, preincubation of
cells with higher concentration of anti-B2 Ab completely abrogated
the binding of anti-gp96 and vice versa. Thus, our results not only
show the identical expression pattern of B2 and gp96 on LPS-activated B
cells, but also depict colocalization of both the molecules, which
unequivocally proves the identity between B2 and gp96. Furthermore, the
expression of gp96 on the LPS-activated B cells was independent of the
expression of MHC molecules, as most of the gp96 molecules were not
colocalized with MHC II (Fig. 9). Thus,
the results of confocal imaging, i.e., activation-dependent and
MHC-independent surface expression of gp96 on B cell and its
colocalization with B2, render indirect support to the possible role of
gp96 as a costimulatory molecule.
|
|
|
As B2 and gp96 exhibited identical behavior and the B2 effectively
costimulated the CD4+ T cells, we therefore were
interested to know the possible role of gp96 in
CD4+ T cell activation. Thus, to further
elucidate the role of gp96 in T cell activation, commercially available
plate-bound gp96 was tested for its ability to stimulate
CD4+ T cells when anti-CD3 was used as the
source of first signal. Interestingly, we found that gp96 is a potent
costimulatory molecule and can activate the naive
CD4+ T cells in a dose-dependent manner. An
optimum level of activation, nearly 6-fold increment over the basal
level was noted when gp96 was used at 10 µg/ml concentration.
Furthermore, heat-inactivated gp96 failed to costimulate the Th cells,
suggesting thereby that the costimulatory activity of gp96 is sensitive
to thermal denaturation (Fig. 10A). Furthermore, blocking
of the optimum costimulatory activity through gp96 by affinity-purified
anti-B2 polyclonal Ab, at a concentration of 28 µg/ml, diminished
the CD4+ T cell proliferation to the basal level.
As opposed to this, control Ab (normal rabbit IgG), when used at the
same concentration, was unable to abrogate such potency of gp96 (Fig. 10B). Thus, our results not only confirm the functional
identity between B2 and gp96, but also assign a new costimulatory role
to gp96.
|
) and Th2 type (IL-4, IL-10, and
IL-13) as well. Interestingly, it was found that gp96 signaling was
skewing the differentiation of naive Th cells only toward Th2 type; as
IFN-, the hallmark of Th1 differentiation was not detected at all in
the culture supernatants. Moreover,diminishing the
costimulation through gp96 either by using anti-B2 polyclonal Ab or
by heat inactivating the protein also abolished the cytokine levels
(Fig. 10C). Therefore, our experiments unequivocally prove
that gp96 functions as a Th2-specific costimulatory molecule. Expression of CD91 on Th cells and its binding to B2/gp96
The receptor of gp96 has been identified as CD91
(28), and results described in preceding sections show
that B2 and gp96 are the same molecule. Also, electron-microscopic
studies reveal that B2 binds to acitvated Th cells (Fig. 3). Therefore,
we next examined whether such binding of gp96/B2 on T cells was
mediated through CD91. Our results show that in comparison with a small
population of resting Th cells (5–8%), FITC-tagged B2 binds to a
95–98% population of the anti-CD3-activated
CD4+ T cells. Anti-CD91 Ab or unlabeled purified
gp96 protein, at a 35 times higher molar concentration, abrogated this
binding by 80% (Fig. 11), suggesting
thereby that: 1) activated Th cells exhibit CD91, and 2) B2 and gp96
compete out each other for the same receptor (CD91).
|
, inset). |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although it is difficult to catalogue the inventory of stress-mediated danger signals, however, a simplified view suggests that it could be just an invasion and physiological disturbances induced by pathogen, release of its metabolites, or endotoxin. Recently, the role of danger signal in the form of endotoxin (LPS) has been proven in the survival of memory T cells (32). Thus, to mimic the danger signal of the in vivo system, we cultured B cells with LPS and analyzed the capacity of such B cells to activate the Th cells. We found that Th cells cultured with anti-CD3 and LPS-activated B cells predominantly secrete the lymphokines representative of Th2 phenotype. Further analysis of such selective induction of Th cells by LPS-activated B cells led us to identify new Th2-specific costimulatory molecules (17, 18) from its membrane fraction.
In the present study, we describe the isolation and characterization of yet another molecule of 99–105 kDa, named B2, which provides Th2-specific costimulatory signal to the Th cells. It may also be emphasized in this work that the B2 alone at any given concentration does not stimulate the T cells, indicating that the biological effect of B2 is mediated only after cross-linking of the TCR. The homogeneity of the preparation of B2 protein not only was validated by reverse-phase HPLC and two-dimensional gel electrophoresis, but also had been categorically proven by the sequencing data, as the sequence analysis of two of the B2 peptides showed full identity with the sequence of a single protein, i.e., gp96. Compelling literature is available indicating the role of gp96 in immunotherapy of cancers. Retarded progression of primary cancers, reduced metastatic load, as well as a prolonged life span have been observed in tumor-bearing animals when treated with gp96 purified from autologous cancer cells (33). Furthermore, expansion of NK cell population and gp96-mediated elicitation of MHC I-restricted, tumor-specific CD8+ T lymphocytes had been noticed when the patients were immunized with the autologous cancer-derived gp96 preparation (34). In these studies, the role of gp96 has largely been attributed to its ability as carrier of immunogenic cancer cell-derived peptides. Of particular importance is the fact that the peptides associated with gp96 have been shown not to be MHC restricted, thereby being capable of cross-priming for CTL responses (35). It is also demonstrated that both macrophages and dendritic cells, when pulsed with the gp96 derived from tumor cells, have become potent activators of tumor-specific CTL responses, which may be again through an activation of CD4+ T cells (33). However, it may also be emphasized here that a tumor-derived gp96-Ig secretory fusion protein can directly activate specific CD8+ T cells, in the absence of any assistance from either macrophages or CD4+ T cells (36). Moreover, peptide-stripped gp96 purified from tumor cells has also been shown to be capable of directly activating CD8+ T cells both in vivo and in vitro (21). Thus, these results collectively suggest that gp96 indeed plays an important role in T cell activation, the mechanism of which is yet to be established. So, as the internal sequences of B2 showed full identity with gp96, we have used classical immunological approaches to verify whether or not B2 and gp96 are identical. Our approach was based on the protocol described by Schaiff et al. (37), in which the authors identified a 97-kDa protein as gp96 not only by its sequence homology with gp96, but also through immunodetection by anti-gp96 Ab. Interestingly, results of our immunodetection experiments with anti-gp96 Ab coupled with sequence analysis of B2 conclusively proved that B2 is indeed gp96.
Our experiments regarding the surface localization of gp96 (the
probable costimulatory molecule) on APC also show that, only after LPS
activation, gp96 tends to accumulate at the cell periphery and also on
the surface of B cells. Glycoprotein 96 has been categorized as a
member of heat shock protein (hsp) 90 and hsp83 family, respectively,
in mice and Drosophila, and fluorescent microscopic studies
with heat-stressed Drosophila cells also revealed parallel
observation with respect to the localization of hsp83
(38). As far as negligible expression of gp96 on resting B
cells is concerned, our data are strengthened by the fact that other
authors have also reported negligible expression of gp96 on
unstimulated splenic B cells of Xenopus and catfishes
(39). Furthermore, negligible expression of gp96 on the
surface of resting B cells is in confirmation with the fact that
resting B cells are poor APCs and thus do not express the threshold
level of most of the costimulatory molecules unless activated. In the
foregoing account, confocal microscopic studies on the expression
patterns of B2 and gp96 on LPS-activated B cells reveal the
colocalization of these two molecules, which further supported the
identical nature of gp96 and B2. Validating the results of confocal
microscopy, results obtained by FACS analysis also showed near about
same expression pattern of B2 and gp96 molecule on various cell lines
viz A20, J774, and P815. However, no detectable level of expression of
either of these two molecules was found on EL-4 cell line (data
not shown). One of the pilot studies regarding the surface localization
of gp96 has revealed its expression only on transformed (Meth A induced
fibrosarcoma) cells (40). However, it has also been
postulated that due to the presence of C-terminal KDEL (lysine-aspartic
acid-glutamic acid-leucine) sequence, which is the common retention
signal for endoplasmic reticulum proteins (41), the
majority of this protein stays as a residual form in the endoplasmic
reticulum (42). Only under stressful conditions, gp96
tends to redistribute to the Golgi apparatus (43), found
to be enriched to some extent in the nucleus (44) and also
to the outer surface of the plasma membrane (45).
Costimulatory molecules are largely characterized by: 1)
MHC-independent surface expression on APCs, and 2) ability to work
crossing the MHC barrier. Significantly, our finding regarding the
expression of gp96 on murine B cells in the context of MHC suggests
that the expression of this stress protein is also independent of the
expression of MHC molecules. Moreover, validating the second property
of costimulatory molecule, interestingly enough, the surface expression
of gp96 has also found to be phylogenetically conserved
(39). Furthermore, the ancestral immunological role of
gp96 as danger-signaling molecule (39, 46) has recently
been outlined in the literature. Thus, time and again, the expression
of gp96 on cell surface (40, 45), especially on B
cells (26, 39), had been shown. However, no definite
role to this molecule had been assigned yet. So, based on our
previous results and reported data, we hypothesize that being a
family member of one of the most primitive hsp, i.e., hsp90, gp96 may
have originated as the first ever known danger-signaling molecule, as
it is also found on some specific cells of the agnathan (hagfish),
which lacks an adaptive immune system (39). Furthermore,
this function of danger signaling (46), as a signal
between innate and adaptive immunity, the molecule has been possibly
retained through the evolution. Thus, to further elucidate the
role of gp96 in the T cell activation, we have examined the role of
this molecule as the source of accessory signal. Interestingly,
gp96-mediated proliferation of CD4+ T cells
was only seen when the cells were cultured in the similar
condition of TCR occupancy and gp96, which also lead the
differentiation of Th cells toward Th2 phenotype by inducing the
secretion of IL-4 and IL-13. Furthermore, it would be also worth
mentioning in this work that the secretion of IFN- in undetectable
level could be due to the inhibitory effect of IL-10 on IL-12 and
IFN- production (47).
For an effective signal transduction, a costimulatory molecule should bind to its counter ligand on the target cell. Therefore, further studies were undertaken to explore this possibility, and the results obtained through electron-microscopic autoradiographic and competitive inhibition studies indicate the higher expression of a putative receptor of this molecule only on anti-CD3-activated T cells because no binding of 125I-labeled B2 beyond the basal level had been found when unstimulated naive T cells were incubated with it. Furthermore, parallel experiment by Singh-Jasuja et al. (48), in the search for binding of fluorescent-labeled gp96 on unstimulated splenic T cells, has also resulted in similar observations. Because the receptor of gp96, designated as CD91 (28), has recently been described on dendritic cells and macrophages (49, 50) and costimulation of Th cells requires receptor-ligand interaction, a question therefore arises as to whether Th cells express the same receptor for gp96 or not. Our experiments to answer this question resulted in an observation that, like the expression of inducible costimulator (51) and CTLA-4 (52), the expression of gp96 receptor (CD91) on Th cells was also found to be activation dependent. This observation, regarding the expression of CD91 on T cells, is further supported by the fact that the expression of CD91 has also been recently reported on a population of T cell thymoma, i.e., EL-4 cells (53). Interestingly, expression of CD91 was found to be present on Th1 as well as Th2 subsets with only a marginal increase (10–15%) in the Th2 population expressing CD91. Our observation regarding no preferential expression of CD91 on Th1 and Th2 subsets could possibly be explained by the fact that generation of a Th response and its maintenance and/or regulation are two immunologically distinct events. The former solely depends upon the strength of signal, i.e., the level of TCR interaction along with the level of CD28 signaling and the cytokine milieu (54), which denote the mode of activation and in turn the fate for the differentiation of naive Th cells. To exemplify, for activation of Th cells, an increasing concentration of plate-bound anti-CD3 had proven to have the synergic effect with the increasing concentration of anti-CD28 Ab, and simultaneous signaling through ICAM and B7.1 had also resulted in a more potent Th response (55). Moreover, it had also been found that the generation of Th2 subset from naive Th cells needs higher strength of signal in comparison with the strength required the generation of Th1 cells (54, 56), which does not hold true for the maintenance of either of these Th subsets. For example, although signaling through CD28 to naive Th cells and the generation of a Th2-biased response, by the production of Th2-specific cytokines such as IL-4 (57), had been documented in the literature (54, 58, 59), nevertheless, it had also been demonstrated that after polarization CD28-mediated signaling is not essential for the maintenance of Th2 cells (60, 61). Furthermore, engagement of the first signal (ionomycin) alone was proven to be sufficient to activate the Th2 cells, whereas for the activation of Th1 cells a combined signal by ionomycin and PMA was found to be essential (62). In continuation, it has also been demonstrated that IL-1, but not IL-4, plays a pivotal role for initiating the maintenance of Th2 subset and could either be secreted by the macrophages or even could be produced by the signaling through CD28 to the Th2 subsets (63). This IL-1-mediated signaling is also demonstrated to be independent of B7-CD28 interaction (63) and proven to be Th2 specific, as the Th1 cells do not express the IL-1R on them. In contrast, costimulation through CD28 to Th1 subsets generates IL-2 production, which in turn participates in their maintenance (63).
Thus, even though CD28 may be found to be present on both the polarized Th subsets (63), however, the requirement and/or outcome of signaling through it might be entirely different (60, 63, 64). In light of these observations, it is not surprising to find the ligand of gp96 on Th1 and Th2 cells. However, it is a matter of further investigation to reveal the extent of requirement of gp96-mediated signaling and its outcome on differentiated Th1 and Th2 cells, more importantly in a scenario of stress-mediated skewing of immunoresponse from Th1 to Th2 kind (65). Because stress stimuli such as heat shock, glucose starvation, or LPS activation (26) induce the up-regulation of gp96 and the costimulatory activity of gp96 also governs the Th response toward Th2, therefore, it would be pertinent to postulate that the molecule like gp96 might regulate this shift of Th response. Moreover, during in vitro CD4+ T cell proliferation assay (data not shown), glucose-deprived or heat-shocked APCs also generated a more potent Th response than the unstressed control group. Thus, the present observations depicting the role of gp96 in T cell activation probably denote a sort of parallelism between stress and costimulation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 P.P.B. and D.S.V. contributed equally to this work.
3 Current address: Division of Transplantation, Department of Surgery, University of Michigan, Ann Arbor, MI 48109.
4 Address correspondence and reprint requests to Dr. Gyan C. Mishra, National Centre for Cell Science, Ganesh khind Road, Pune 411007, Maharashtra, India. E-mail address: gcmishra{at}nccs.res.in
5 Abbreviations used in this paper: KLH, keyhole limpet hemocyanin; hsp, heat shock protein.
Received for publication August 28, 2001. Accepted for publication July 23, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
stimulated monocytes: a potential mechanism for amplification of T cell activation through the CD28 pathway. Cell. Immunol. 137:429.[Medline]
production and proliferation of ovalbumin-specific CD4+ T cells. Eur. J. Immunol. 24:2092.[Medline]
-producing CD4+ T cells following activation of naive CD4+ T cells. J. Immunol. 158:1085.[Abstract]
