| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Medical Microbiology and Immunology, Faculty of Medicine, University of Alberta, Edmonton, Alberta, Canada
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1 and 2
domains, respectively. A third glycosylation site in the 3 domain is
also found on some class I molecules (4). In addition to a polymorphic
polypeptide backbone, murine class I heavy chains also possess
substantial microheterogeneity primarily attributed to the variable
size, branching, and charge of N-linked oligosaccharides (5, 6, 7). In
contrast, the single N-linked oligosaccharide on human class I
molecules appears to have a more uniform structure (8). Glycosylation
of class I molecules is not required for CTL recognition (9, 10).
However, certain modifications on class I MHC oligosaccharides may
interfere with class I-restricted TCR interactions (11). Recently, various human and murine NK cell receptors for class I molecules have been described (12). The Ly-49 family, consisting of Ly-49A through -I, are type II transmembrane disulfide-linked glycoproteins expressed on murine NK cells and small subsets of T cells (13, 14). The extracellular domain of these glycoproteins bear homology to animal C-type lectins (15, 16). Some Ly-49 family members, such as Ly-49A, Ly-49C, and Ly-49G2 have been shown to negatively regulate NK cell lytic activity upon interacting with specific class I MHC products (17, 18, 19). The H-2Dd molecule is a ligand for Ly-49A and probably Ly-49G2 (19, 20, 21). The interaction of Ly-49A with Dd is suggested to involve the N-linked oligosaccharides on Dd, since binding of Dd-expressing cells to Ly-49A-expressing cells is inhibited following treatment with tunicamycin, an antibiotic that inhibits N-linked glycosylation (22). In addition, Ly-49A interaction with Dd, and Ly-49C interaction with an H-2s ligand, can be inhibited by sulfated polysaccharides (22, 23). These studies suggested that carbohydrate moieties, possibly sulfated oligosaccharides, on Dd may be important for its interaction with Ly-49A. However, there is no evidence to date that class I molecules are sulfated.
In the present report, we demonstrate that H-2Dd expressed on a variety of cells is indeed sulfated. Furthermore, we provide evidence that the sulfate modification can occur on N-linked oligosaccharide(s) of Dd.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Eight- to twelve-week-old female BALB/c mice were obtained from the University of Alberta mouse breeding facility.
Enzymes and chemicals
Endoglycosidase F (Endo F)/peptide N-glycosidase F (PNGase
F) and PNGase F purified from Flavobacterium meningosepticum
were purchased from Boehringer Mannheim (Laval, Quebec, Canada) and
Sigma (St. Louis, MO), respectively. Recombinant Streptomyces
plicatus endoglycosidase H (Endo H) was purchased from Boehringer
Mannheim. Con A was from Pharmacia Fine Chemicals (Uppsala, Sweden).
Immobilized protein A on Trisacryl GF-2000 was obtained from Bio-Rad
(Rockford, IL). Fischer’s sulfate-free medium, sulfate-free DMEM, and
RPMI 1640 Select-Amine Kits were obtained from Life Technologies
(Burlington, Ontario, Canada). An autoradiography enhancer,
EN3HANCE, was purchased from DuPont NEN (Richmond, B.C.,
Canada). Tunicamycin and BSA (fraction V) were purchased from Sigma
Immunochemicals. Sodium chlorate (NaClO3) was purchased
from Fisher Scientific (Fair Lawn, NJ).
Na235SO4 in water (
43 Ci/mg S)
and Tran35S label (>1000 Ci/mmol) were purchased from ICN
Pharmaceuticals (Irvine, CA).
Monoclonal Abs
34-5-8S (IgG2a), anti-Dd (24), 11-4.1 (IgG2a), anti-Kk (25), B22.249 (IgG2a), anti-Db (26) and A1 (IgG2a), anti-Ly-49A (27) have been characterized. These hybridomas were obtained from the American Type Culture Collection (ATCC; Rockville, MD), except A1, which was obtained from Dr. James Allison (University of California, Berkeley, CA), and B22.249, which was obtained from U. Hammerling (Memorial Sloan-Kettering Cancer Center, New York, NY). Hybridoma supernatants were concentrated by NH4SO4 precipitation.
Tumor cell lines
The S49.1 and Yac-1 T lymphomas were obtained from ATCC. The A20.Cy B lymphoma was provided by Dr. A. O’Rourke (Scripps Clinic, La Jolla, CA). The NZB1.1 T lymphoma was kindly provided by Dr. Hanne Ostergaard (University of Alberta, Edmonton, Canada). The A20.Cy, S49.1, and NZB1.1 cell lines were maintained in DMEM, whereas Yac-1 was grown in RPMI with 5% heat-inactivated FCS, 20 mM HEPES, 2 mM L-glutamine, 100 U penicillin, and 100 µg/ml streptomycin.
Isolation of resting T cells
Spleen cells were harvested from 8- to 12-wk-old female BALB/c mice. After lysing RBC with 0.14 M NH4Cl in Tris, pH 7.3, spleen cells were incubated at 5 x 106/ml on tissue culture-treated plates for 1 h at 37°C in DMEM/10% FCS. Nonadherent cells were passaged at a density of 1 x 107 cells/ml through a nylon wool column. Unbound cells were used for radiolabeling.
Activation of T cells by Con A
Upon removal of RBC, BALB/c spleen cells were cultured in DMEM/10% FCS and 3 µg/ml of Con A in 24-well plates at 5 x 106/ml. After 48 h of culture, the activated cells were harvested and used for radiolabeling.
Radiolabeling
Cells were metabolically labeled at 2 x 106 cells/ml for 7 to 8 h at 37°C. For sodium sulfate (sulfate ion; SO4-2) labeling, cells were incubated with 1 mCi/ml of Na235SO4 in Fischer’s medium or sulfate-free DMEM supplemented with 10% dialyzed FCS. For Tran35S ([35S]methionine and [35S]cysteine) labeling, cells were incubated with 200 µCi/ml of radioactive label in methionine-free RPMI with 10% dialyzed FCS.
Inhibition of Dd sulfation by NaClO3 and N-linked glycosylation by tunicamycin
Cells at a concentration of 2 x 106/ml were labeled with Na35SO4 in Fischer’s medium supplemented with 10% dialyzed FCS for 7 h in the presence of 10 or 20 mM of NaClO3. To inhibit N-linked glycosylation, cells were labeled with Na35SO4 or Tran35S label in Fischer’s medium or methionine-free RPMI supplemented with 10% dialyzed FCS in the presence of 1 µg/ml of tunicamycin, for 7 h.
Immunoprecipitation
Five to ten million metabolically labeled cells were lysed at 1 x 107 cells/ml in PBS containing 1% Nonidet P-40, 1 mM PMSF and 5 mM iodoacetamide, pH 7.2, at 4°C. Cell lysates were centrifuged at 16,000 x g for 15 min and supernatants collected. The lysate was precleared twice with 30 µl of protein A-packed beads and 5 µl of BALB/c normal serum, followed by immunoprecipitation with 30 µl of protein A-packed beads and 10 µg of 11-4.1 or 34-5-8S mAb per 107 cells for 2 h. Beads were washed with PBS containing 1% Nonidet P-40, 1 mM PMSF, 5 mM iodoacetamide, and 1.0% deoxycholate, pH 7.2, and eluted with 60 µl of 2% SDS reducing sample buffer at 100°C for 4 min.
SDS-PAGE, fluorography, and autoradiography
Radiolabeled proteins were resolved by 10% SDS-PAGE under reducing conditions (28). After intensification by autoradiography enhancer (EN3HANCE), the gels were vacuum-dried and exposed to XAR5 Kodak films.
Two-dimensional gel electrophoresis of radiolabeled class I MHC
In some experiments, 35SO4-2 and Tran35S-labeled Dd were eluted from mAb-bearing beads and subsequently acetone precipitated with 5 µg of BSA carrier to remove SDS before two-dimensional gel electrophoresis. Precipitates resuspended in isoelectric focusing (IEF) sample buffer were separated by IEF for 4800 V-h using a 4 to 1 ratio of Bio-Lyte 5/7 and Bio-Lyte 3/10 (Bio-Rad, Mississauga, Ontario, Canada). The second dimension separation was then conducted in 10% SDS-PAGE reducing gels (29).
Removal of N-linked oligosaccharides
Radiolabeled proteins from Ab-bearing beads were alkylated as described (5), and acetone precipitated at -20°C overnight with BSA carrier. For Endo F/PNGase digestion, acetone precipitates were resuspended and heated for 4 min at 100°C in 50 µl of sample buffer containing 80 mM sodium phosphate, 50 mM EDTA, 1% Nonidet P-40, 0.2% SDS, and 1% 2-ME, pH 6.0. For Endo H treatment, the buffer was 0.1 M sodium acetate, 0.2% SDS, and 1% 2-ME, pH 6.0. Samples were then digested by 0.2 mU Endo F/PNGaseF, 2 U PNGase F, or 2 mU Endo H at 37°C for 10 h in the presence of 0.2 mg/ml aprotinin and 2 mM PMSF. After 10 h of incubation, half of the original amount of enzyme was added again and the digestion was conducted for another 10 h.
Cell adhesion assay
Ly-49A was isolated as described previously (30), and the cell adhesion assay has been described (30). Briefly, 5 x 104 Na51CrO4-labeled cells were added to each well of microtiter plates immobilized with purified Ly-49A at various densities or a control protein BSA, following preincubation for 20 min with medium or Abs at 5 µg/ml. The plates were centrifuged at 400 rpm for 5 min, and incubated at 37°C for 1 h. The unbound cells were removed, and the percentage of bound cells was calculated as follows: [cpm bound/(total cpm - spontaneous cpm)] x 100.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It has been demonstrated that sulfated mono- and polysaccharides
interfere with Dd binding by Ly-49A in cell-cell
binding assays (22). Furthermore, we have observed that such soluble
sulfated sugar structures are capable of inhibiting
H-2Dd-expressing cells such as the NZB1.1 tumor cell line
from binding solid phase, purified Ly-49A (data not shown). We
investigated the possibility that Dd molecules themselves
display sulfate moieties, perhaps explaining the binding competition
offered by soluble sulfated carbohydrates. The NZB1.1 T lymphoma was
labeled with Tran35S in methionine-free RPMI or
35SO4-2 in Fischer’s medium for 7 to
8 h. As expected, H-2Dd, with a molecular mass of
approximately 45 kDa, was immunoprecipitated from
Tran35S-labeled cell lysates by the mAb 34-5-8S, but not
the isotype control mAb 11-4.1 (Fig. 1
A, lanes
1 and 2). The autoradiograph also
indicated that a sulfated form of Dd was immunoprecipitated
from the 35SO4-2-labeled cell lysate using
the same Dd-specific mAb (Fig. 1A,
lane 4). No sulfo-labeled band was
detectable using the isotype control mAb, 11-4.1 (Fig. 1A,
lane 3). The sulfation of Dd
was not specific for cells that were labeled in Fischer’s medium,
since sulfated Dd was also detected in cells labeled in
sulfate-free DMEM (Fig. 1B). In addition, no change
in Dd expression was detected in cells that had been grown
in Fischer’s medium (data not shown). Thus, these results indicate
that H-2Dd expressed by NZB1.1 is sulfated.
|
). No sulfated bands were detected in
all isotype control immunoprecipitates (data not shown). It is
interesting to note that Dd expressed by different cells is
not sulfated to the same degree when comparing Tran35S
label and 35SO4-2 label in
immunoprecipitates, and sulfation did not have a direct correlation to
the relative level of Dd cell surface expression. For
instance, A20.Cy and S49.1, which express relatively high levels of
Dd at the cell surface (data not shown) and incorporated
substantial Tran35S label (Fig. 2, lanes
2 and 3), yielded the least
35SO4-2-labeled Dd (Fig. 2, lanes 6 and 7). Alternatively,
NZB1.1 and Yac-1 have lower levels of Dd cell surface
expression (data not shown) and incorporated Tran35S label
to a lesser extent (Fig. 2, lanes 1 and
4), yet have higher amounts of sulfated
Dd (Fig. 2, lanes 5 and
8). In addition, within the m.w. ranges of the
Dd observed by Tran35S labeling for each cell
line, the Mr of the sulfated Dd
species can differ between cell lines, possibly due to oligosaccharide
microheterogeneity and sulfation of specific Dd glycoforms
expressed by the cell lines (Fig. 2). For example, the S49.1 cell line
has sulfated Dd of the highest Mr
among all of the tumors cell lines analyzed (Fig. 2, lane
7), with a corresponding portion of the S49.1
Tran35S-labeled Dd having an identical
Mr (Fig. 2, lane
3).
|
Since all H-2Dd-expressing tumor cells that were
examined expressed sulfated H-2Dd, we wished to determine
whether Dd expressed by normal resting and Con A-activated
T cells were also sulfated. Normal resting T cells express a lower
level of Dd compared with their Con A-activated
counterparts and incorporate labeled amino acids at a substantially
reduced rate. Only a very small amount of Tran35S label was
incorporated into the H-2Dd immunoprecipitated from resting
BALB/c splenic T cells, and no detectable
35SO4-2 was incorporated (Fig. 3, lanes 2 and
6). It is still possible, however, that
Dd expressed in normal resting cells does incorporate some
sulfate, but their low metabolic rate does not allow us to resolve this
issue with radiolabeled sulfate. In contrast, a strong incorporation of
Tran35S label and a low but detectable incorporation of
35SO4-2 was observed in Dd
immunoprecipitated from Con A-activated T cells (Fig. 3, lanes 4 and 8). Although
incorporation of sulfate into the Dd of Con A blasts is
low, we have observed this result consistently in several experiments.
Thus, sulfation of the Dd molecule occurs in normal
metabolically active cells and is not restricted to transformed cell
lines.
|
Due to the variable action of glycosyltransferases in the
processing and maturation of N-linked oligosaccharides, murine
class I MHC molecules are known to possess a high degree of
microheterogeneity (5, 6, 7). Therefore, to identify how many
Dd glycoforms undergo sulfation, two-dimensional gel
electrophoresis was conducted on Dd immunoprecipitated from
NZB1.1 cells labeled with either Tran35S or
35SO4-2 for 8 h. The
Tran35S-labeled Dd was separated into various
subpopulations by IEF followed by SDS-PAGE, with isoelectric points
(pIs) ranging from 5.5 to 6 (Fig. 4A). The detected pI
range for the separated Dd glycoforms is in agreement with
that published previously (7). The pattern of separation detected with
35SO4-2-labeled Dd (Fig. 4B) corresponds to the mature glycoform species of
Dd (toward the acidic end of the gel) observed by
Tran35S label (Fig. 4A). This suggests
that all of the detectable mature glycosylation variants of
Dd from NZB1.1 are modified with sulfate.
|
Sulfation is a posttranslational modification that can occur on
tyrosine residues or carbohydrate moieties (31, 32). Since there are
perhaps three potential tyrosine sulfation sites (33) and two known
N-linked glycosylation sites on Dd, the polypeptide
backbone, the N-linked carbohydrate(s), or both, may be modified with
sulfate. NZB1.1 cells were labeled with Tran35S or
35SO4-2 in the presence or absence of
tunicamycin, an inhibitor of N-linked glycosylation (34). The
carbohydrateless Dd polypeptide backbone was detected in
Tran35S-labeled Dd immunoprecipitates (Fig. 5A, lane
1, arrow), although relatively poorly, possibly due to the
generally rapid turnover of carbohydrateless class I molecules, which
are deficient in associating with the ER chaperon, calnexin (35). No
radiolabeled material was detected in Dd immunoprecipitates
of 35SO4-2-labeled NZB1.1 treated with
tunicamycin (Fig. 5A, lane 5).
This result could imply that 35SO4-2 is
added to the N-linked carbohydrate moieties on Dd. However,
it remains a possibility that in the presence of this inhibitor,
Dd does not efficiently transit the trans-Golgi
network, where sulfate addition is thought to occur, and thus tyrosine
as well as oligosaccharide sulfation cannot take place. To address this
issue more directly, Dd immunoprecipitated from
Tran35S or 35SO4-2-labeled
NZB1.1 was treated with a mixture of Endo F and PNGase F or PNGase F
alone. Upon digestion with these enzymes, which remove N-linked
oligosaccharides, the 35SO4-2 label on
Dd was no longer detectable by autoradiography (Fig. 5A, lanes 7 and 8),
in contrast to the untreated control (Fig. 5A,
lane 6). The disappearance of the sulfate
label was due to removal of the N-linked oligosaccharides since the
migration of the Tran35S-labeled Dd treated in
parallel with the same N-glycosidases shifted to a position consistent
with it lacking N-linked carbohydrates (Fig. 5A,
lanes 3 and 4). An additional
fainter band of slightly reduced Mr is observed
in the untreated Tran35S-labeled control (Fig. 5A, lane 2) and following Endo
F/PNGase F treatment of Tran35S-labeled Dd
immunoprecipitates (Fig. 5A, lanes 3
and 4). This probably represents the protein product
of a Dd mRNA splice variant that lacks 13 amino acids in
its cytoplasmic tail (36). Assuming this is the case, it too may be
sulfated on N-linked oligosaccharide(s), observed as the fainter
sulfated band just below the major sulfated Dd band (Fig. 5A, lane 6), but is missing
following enzyme treatment (Fig. 5A, lanes
7 and 8). Taken together, these data
indicate that the N-linked oligosaccharide(s) on Dd
expressed on NZB1.1 cells is sulfated, yet the tyrosine residues of
Dd are not detectably sulfated. When
35SO4-2-labeled Dd was treated
with Endo H, an endoglycosidase that removes immature high mannose
N-linked oligosaccharides (37), it was resistant to this enzymatic
treatment (Fig. 5B, lanes 1 and
2). Thus, sulfate modification(s) occurs on mature
(postmedial Golgi) forms of N-linked carbohydrate moieties on
Dd.
|
Since we demonstrated that sulfation of Dd on
NZB1.1 cells was located on Asn-linked oligosaccharide(s), this
sulfation process should be inhibited by NaClO3. This
compound is a potent inhibitor of the formation of 3'-phosphoadenosine
5'-phosphosulfate (PAPS), an active sulfate donor in vivo (38). When
sulfo-radiolabeling of NZB1.1 was conducted in the presence of
NaClO3 in Fischer’s medium (Fig. 6, lanes 2 and
3) or sulfated-free DMEM (data not shown), sulfation
of Dd was not detectable, in contrast to the untreated
control (Fig. 6, lane 1). Following
incubation with NaClO3, no reduction in Dd
expression was observed by either FACS analysis or immunoprecipitation
of Trans35S-labeled material (data not shown). This
indicates that chlorate ions inhibit sulfation of Dd
N-linked oligosaccharides.
|
We have previously demonstrated that adhesion of cells of
H-2d origin to isolated Ly-49A is blocked by 34-5-8S,
an Ab specific for Dd, or A1, an Ab specific for Ly-49A
(30). This is also the case for NZB1.1 binding to immobilized Ly-49A
(Fig. 7A), indicating
that the cell adhesion to Ly-49A is mediated by Dd. Since
soluble sulfated polysaccharides can bind the carbohydrate recognition
domain of Ly-49A (22), and we have demonstrated that the N-linked
oligosaccharide(s) on Dd is (are) sulfated, and
NaClO3 is able to block this posttranslational
modification, we reasoned that pretreatment of the the
Dd-expressing cells with chlorate ions might inhibit cell
adhesion to Ly-49A. When NZB1.1 were cultured in sulfate-free DMEM,
there was a small reduction in their binding to Ly-49A relative to when
they are grown in the same medium supplemented with sulfate ions (Fig. 7B). The substantial residual binding under these
conditions could be due to a long-lived intracellular pool of
SO4-2 still available to the cells for modification of
Dd. In addition, the cells might also be able to derive
free sulfate from sulfur-containing amino acids (39, 40). Perhaps as a
result, sulfation of Dd can still take place in
sulfate-free medium, and thus only a small reduction in cell binding to
Ly-49A is observed. However, when the cells were treated with 20 mM
NaClO3 for 7 h, preventing active sulfate donor (PAPS)
formation, which is an essential reaction for glycoprotein sulfation,
there was a significant reduction in the percentage of cells bound to
Ly-49A (Fig. 7B). The reduction in adhesion became
particularly apparent when the density of the immobilized Ly-49A was
limiting (Fig. 7B), whereas at a high density of
Ly-49A, the chlorate-treated cells were still able to interact with
immobilized Ly-49A. The results shown in the representative assay of
Figure 7 were consistently observed in four separate experiments. These
results suggest that sulfate-modified N-linked oligosaccharide(s) of
Dd can participate in the interaction of Dd
with Ly-49A, with sulfation becoming essential for cell adhesion when
the Ly-49A density is relatively low.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1 and
2 domains. In addition, it has two conserved N-linked glycosylation
sites at residues 86 and 176. When treated with PNGase F, the sulfate
label of the NZB1.1 Dd was removed, and no residual
sulfo-radiolabel was detected on the polypeptide backbone. This
strongly suggests that N-linked oligosaccharides on NZB 1.1
Dd are sulfated. However, we cannot rule out the
possibility that some tyrosine sulfation occurs to a small extent, and
that this sulfation might not be detected by the methods employed in
this study. In addition, it remains possible that Dd
expressed by other tumor cell lines are sulfated on tyrosine residues
and/or N-linked oligosaccharide(s). We observed that Dd expressed by various tumor cells is apparently not sulfated to the same extent. Among the tumor cell lines analyzed, NZB1.1, a spontaneous tumor originating from NZB mice, has the highest level of sulfated Dd in absolute terms and relative to the amount of Dd expressed by these cells. In contrast, A20.Cy and S49.1 have the lowest relative sulfation of Dd, by comparing 35S04-2-labeled Dd with Tran35S-labeled Dd in immunoprecipitates and the amount of Dd expressed at the cell surface. Thus, the degree of Dd sulfation is not directly correlated with the level of Dd expression. This difference might be accounted for by the origin, growth, and potential survival selection that the individual tumors may have undergone in vivo. For example, tumor cells may differ in the level of expression of ATP sulfurylase, the enzyme involved in the transfer of sulfate to glycoprotein acceptor molecules, or some tumor cell lines may have a larger intracellular pool of sulfate or obtain sulfate from sulfur-containing amino acids more efficiently than others (39, 40). In addition, the sulfated Dd immunoprecipitated from a variety of tumors can differ slightly in their relative mobilities. These differences are likely to be attributable to the variable complexity of N-linked oligosaccharides that can be expressed on Dd (6). Whether other class I MHC allelic products can be sulfated similar to Dd remains to be determined.
Sulfation of Dd is not a phenomenon restricted to tumor cell lines, since we found that Dd immunoprecipitated from Con A-activated T cells is also sulfated. However, all of the tumor cells we tested appear to have a higher degree of Dd sulfation compared with Con A-activated T cells. Perhaps transformation can lead to overexpression of sulfotransferases or other factors, which favor hypersulfation of Dd. It remains to be determined whether enhanced sulfation of Dd or other class I molecules might provide a means for some tumor cells to escape destruction by NK cells, possibly by enhancing interaction with NK inhibitory receptors such as Ly-49 family members.
Two-dimensional SDS-PAGE analysis of NZB1.1 Dd indicated that sulfated forms of Dd were found to comigrate with all major mature glycosylation variants, within the pI range of 5.5 to 6.0. Therefore, sulfation may contribute to Dd microheterogeneity. This possibility is supported by the fact that neuraminidase treatment of Dd only partially reduces its isoelectric heterogeneity (6). Whether all mature glycoforms of Dd are sulfated when expressed by tumor cells other than NZB1.1 remains to be determined. Based on the sulfation patterns of Dd expressed by the tumor cell lines A20.Cy and S49.1, it seems that only certain subsets of Dd expressed by these cells are likely to be sulfated.
We found the sulfo-radiolabeled Dd of NZB1.1 to be Endo H resistant, but PNGase F sensitive, indicating that Dd sulfation is a postmedial Golgi modification. These results agree with previous studies indicating that sulfation of both carbohydrate and tyrosine residues occurs in the trans-Golgi network (42). Thus sulfation may be the last modification of Dd before its expression at the cell surface. The data presented in this report do not distinguish which saccharide residue(s) on Dd is (are) sulfated. Sulfate modification can be complex, since it has been reported to occur on a variety of saccharide residues, including penultimate or terminal galactose/N-acetylgalactosamine, as well as peripheral or core N-acetylglucosamine (GlcNAc) on N-linked oligosaccharides (32, 43, 44). It has been proposed that a hexose-6-SO4 is part of the oligosaccharide structure recognized by Ly-49A because soluble sulfated hexoses can inhibit this interaction (22). Since all major glycoforms of Dd expressed by NZB1.1 are sulfated, it can be speculated that the sulfate group is added to a carbohydrate moiety that is found in all of them. Thus, one possibility is that sulfate is added to one of the N-linked oligosaccharide core GlcNAc residues. It was observed that after treatment with 1-deoxymannojirimycin or swainsonine, inhibitors of mannosidase I and II, respectively (which interfere with oligosaccharide processing events required for conversion to complex oligosaccharide forms), Dd-expressing cells were still able to bind efficiently to Ly-49A-expressing cells (22). It has been demonstrated that swainsonine treatment of MDCK cells still allows sulfation of an N-linked core GlcNAc (45). This might explain why swainsonine and perhaps 1-deoxymannojirimycin have no effect on Dd interaction with Ly-49A, since they might still allow sulfation of an N-linked core GlcNAc. However, the exact carbohydrate unit(s) that is sulfated on the Dd oligosaccharides expressed by various cells remains to be determined.
It was previously demonstrated that the interaction of Ly-49A and Dd is inhibited by sulfated mono- and polysaccharides. In addition, Ly-49A has C-type lectin properties, in that it is capable of binding sulfated oligosaccharides in a calcium-dependent manner (22). These results indirectly suggested that binding by Ly-49A could be dependent on sulfated carbohydrates expressed on Dd. Oligosaccharide sulfation requires the sulfate donor PAPS and specific sulfoglycosyltransferases, and is sensitive to chlorate ions, which inhibit PAPS formation catalyzed by ATP-sulfurylase (38). We have shown here that Dd sulfation is susceptible to NaClO3 inhibition and that NZB1.1 cultured in the presence of NaClO3 exhibited reduced Dd-dependent binding to purified Ly-49A, especially when the Ly-49A density was limiting. These results indicate that sulfated carbohydrate moieties on Dd can play a role in this interaction. Although the density of Ly-49 molecules on NK cells has not been determined, it is likely to be relatively low, based on FACS analyses (17, 21, 46). Our solid phase adhesion assay employing isolated Ly-49A allowed us to titrate Ly-49A through low densities that may be in the range expressed on NK cells (30). Since cell adhesion was most sensitive to NaClO3 at such a density of Ly-49A, it is possible that sulfation of Dd may play a significant role in the interaction of Dd with Ly-49A molecules expressed on NK cells.
It is likely that the membrane distal carbohydrate recognition domain of Ly-49A is involved in its interaction with the sulfated carbohydrate(s) of Dd; however, the precise role that sulfated oligosaccharide(s) on Dd play(s) in Ly-49A binding remains unclear. The sulfated oligosaccharides may enhance the affinity, but perhaps not provide class I allele specificity to the interaction. It is possible that class I sulfated carbohydrate moieties strengthen initial contact, and subsequent interaction with Ly-49A is determined by the class I MHC polypeptide backbone. The interaction may be analogous to class I interaction with calnexin in the endoplasmic reticulum (1). Here, initial calnexin interaction appears to be dependent on an oligosaccharide intermediate expressed on class I in the ER compartment, and subsequent stable contact is based on protein-protein interaction (35). Alternatively, and probably less likely, sulfated oligosaccharides expressed on class I may play a prominent role in allele specificity of Ly-49 recognition, possibly due to allele-specific oligosaccharide microheterogeneity (47).
Preliminary results from several laboratories suggest that individual Ly-49 family members recognize subsets of class I MHC alleles, and their specificities can partially overlap. For example, Dd is a ligand for Ly-49A and probably Ly-49G2 (19, 20, 21). It will be of interest to determine whether sulfation affects class I interactions with different Ly-49 members, e.g., Ly-49A and G2 to the same extent or uniquely, depending on the specific Ly-49 receptor. In addition, Ly-49 family members may efficiently recognize only certain subsets of sulfated Dd. Although we have shown that Dd expressed by Yac-1 is sulfated, this cell line is sensitive to Ly-49A+ NK lysis (17), and is reported to be unable to stably bind Ly-49A-expressing cells (46). Since Ly-49A interaction with Dd is peptide dependent but not peptide specific, bound peptide is not likely to account for these results. Instead, perhaps in addition to expressing relatively low levels of Dd, Yac-1 expresses sulfated Dd subset(s) or glycoforms that cannot stably interact with Ly-49A. Given the complexity of N-linked oligosaccharides on murine class I MHC molecules, the type and extent of sulfation may provide another level of regulation for Ly-49 member interaction with class I MHC molecules.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Kevin P. Kane, Department of Medical Microbiology and Immunology, 1-41 MSB, The University of Alberta, Edmonton, Alberta, Canada T6G 2H7. E-mail address:
3 Abbreviations used in this paper: N-linked, asparagine-linked; Endo F, endoglycosidase F; Endo H, endoglycosidase H; N-acetylglucosamine, GlcNAc; NaClO3, sodium chlorate; PAPS, 3'-phosphoadenosine 5'-phosphosulfate; PNGase F, peptide-N-glycosidase; SO4-2, sulfate ion; Tran35S, 35S-labeled methionine and cysteine; pI, isoelectric point; IEF, isoelectric focusing.
Received for publication October 16, 1997. Accepted for publication December 22, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
/ß antigen receptor. J. Immunol. 138:815.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  B. D. Spiegelberg, J. dela Cruz, T.-H. Law, and J. D. York Alteration of Lithium Pharmacology through Manipulation of Phosphoadenosine Phosphate Metabolism J. Biol. Chem., February 18, 2005; 280(7): 5400 - 5405. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Sundback, A. Achour, J. Michaelsson, H. Lindstrom, and K. Karre NK Cell Inhibitory Receptor Ly-49C Residues Involved in MHC Class I Binding J. Immunol., January 15, 2002; 168(2): 793 - 800. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Matsumoto, K. Tajima, M. Mitsuki, and K. Yamamoto H-2 allele specificity of the NK cell C-type lectin-like MHC class I receptor Ly49A visualized by soluble Ly49A tetramer Int. Immunol., May 1, 2001; 13(5): 615 - 623. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Matsumoto, M. Mitsuki, K. Tajima, W. M. Yokoyama, and K. Yamamoto The Functional Binding Site for the C-Type Lectin-Like Natural Killer Cell Receptor Ly49a Spans Three Domains of Its Major Histocompatibility Complex Class I Ligand J. Exp. Med., January 15, 2001; 193(2): 147 - 158. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
