(Received for publication, October 4, 1996, and in revised form, November 14, 1996)
From the Department of Immunological Diseases,
Bristol-Myers Squibb Pharmaceutical Research Institute,
Seattle, Washington 98121 and ¶ Fred Hutchinson Cancer Research
Center, Seattle, Washington 98125
Cross-linking of the Fc receptor (FcR) to surface
immunoglobulin (sIg) on B cells inhibits the influx of extracellular
calcium and abrogates the proliferative signal. The mechanism by which this occurs is not well understood. In this report we show that co-cross-linking the FcR to the antigen receptor gives rise to very
selective modulation of signal transduction in B cells.
Co-cross-linking sIg and the FcR enhanced the phosphorylation of the
FcR, the adapter protein, Shc, and the inositol 5-phosphatase Ship.
Furthermore, phosphorylation of the FcR induced its association with
Ship. Cross-linking of the FcR and sIg decreased the tyrosine
phosphorylation of CD19, which led to a reduction in the association of
phosphatidylinositol 3-kinase. In addition, the phosphorylation of
several other proteins of 73, 39, and 34 kDa was reduced. Activation of
the cells with either F(ab
)2 or intact anti-IgG induced
very similar changes in levels of tyrosine phosphorylation of most
other proteins, and no differences in the activation of several protein
kinases were observed. These results indicate that the inhibitory
signal that is transmitted through the FcR is not mediated by a global shutdown of tyrosine phosphorylation but is, rather, a selective mechanism involving localized changes in the interactions of adapter proteins and the enzymes Ship and phosphatidylinositol 3-kinase with
the antigen receptor complex.
Activation of B cells through the antigen receptor gives rise to a sequence of intracellular signals that leads to the proliferation and/or differentiation of the cells (reviewed in Refs. 1 and 2). In contrast, concomitant activation of B cells through both the sIg1 and Fc receptors leads to a dominant negative signal that inhibits activation of the cells (3-6). Thus the role of the FcR on B lymphocytes is to provide a powerful mechanism for the cells to distinguish between free antigen and antigen-antibody immune complexes.
The inhibitory pathway triggered following co-ligation of the sIg
and FcR is not well understood. Studies have recently been undertaken
to try to distinguish the molecular events that give rise to the
differences in signaling mediated through the antigen receptor alone
from that following the cross-linking of sIg to the FcR. Stimulation of
the cells with F(ab)2 anti-IgG, which mimics antigen
binding to the receptor, gives rise to the mobilization of calcium
within the cell from both intracellular and extracellular pools. Uptake
of extracellular calcium appears to be critical for the outcome of
activation. Co-cross-linking of the FcR to sIg with intact anti-Ig,
which mimics the binding of antibody-antigen complexes, inhibits the
influx of extracellular calcium (7, 8). Additionally, co-ligation of
the sIg and Fc receptors subsequent to the opening of the plasma
membrane calcium channel stimulates the closure of the channel (8), and
the level of intracellular calcium rapidly declines. In A20 cells, it
is possible to restore the influx of extracellular calcium and to
overcome the FcR-mediated inhibition of cell activation and IL-2
production by low concentrations of calcium ionophore (8).
Activation of B cells through the antigen receptor also results in the
tyrosine phosphorylation of many proteins, including the cell surface
antigens CD19 and CD22 that are associated with the antigen receptor
(9-12), the components of the antigen receptor complex, Ig and
Ig
(13-15), the tyrosine kinases Syk, Lyn, and Btk (16-18), the
phospholipases PLC
1 and PLC
2 (19-21), and the adapter and
signaling proteins Cbl, Vav, Shc, and Gap (3, 22-29). Phosphorylation
and activation of PLC
1 and PLC
2 results in the generation of
inositol 1,4,5-trisphosphate, which in turn stimulates the release of
calcium from intracellular stores. However, the signaling pathway that
regulates the opening and closing of the plasma membrane calcium
channel has not been well established.
Co-cross-linking of the FcR to sIg also gives rise to the phosphorylation of many intracellular proteins, including the FcR itself, on a tyrosine residue in the intracellular domain (30). The region containing this tyrosine residue is essential for the inhibitory activity of the FcR (30). Subsequently, it has been found that phosphorylation of the FcR on this tyrosine residue results in the interaction of the receptor with the Src homology 2-containing tyrosine phosphatase SHP-1 (31, 32). This association, which may give rise to activation of the phosphatase, is then thought to provide the signal that shuts off activation of the receptor complex and thus to abrogate the proliferation of the cells.
SHP-1 has been found to be associated with the antigen receptor in unstimulated cells (32, 33) and with CD22 in cells activated through sIg (34). Cross-linking of CD22 with antibodies can modulate signaling through the antigen receptor (35), and it has been proposed that one mechanism for this is that ligation of CD22, by antibodies or counter receptors on adjacent cells, gives rise to the removal of the SHP-1 from the immediate locale of the sIg signaling complex, thus removing the inhibitory enzyme (35, 36). Thus phosphorylation of CD22 and the FcR and their interactions with SHP-1 may provide opposing regulatory control on the strength or duration of the signal transmitted through the antigen receptor (33, 35, 36).
The events that follow the phosphorylation of the FcR and its association with SHP-1 have not been well characterized. The phosphorylated proteins that are the targets of the phosphatase in this regulatory pathway have not been identified, and the influence of the SHP-1 and FcR interaction on the activation of downstream signal transduction pathways has not been determined. Additionally, it is not known how the phosphorylation of the FcR and its subsequent interaction with SHP-1 leads to the rapid closure of the calcium channel.
In this report we have compared the signaling events that occur following triggering of the antigen receptor either alone or with co-ligation with the FcR. We have found that the phosphorylation and activation of many proteins in the signaling pathways were not significantly different between the two modes of stimulation; however, there were marked changes in the phosphorylation states of a few proteins, including CD19, FcR, Ship, and Shc. These changes resulted in the association of the Ship with the FcR complex and a loss of association of PI3-kinase with CD19.
These results indicate that the inhibitory signal mediated by the FcR on B cells is a very selective modulation of intracellular signaling and is not simply a consequence of the global dephosphorylation of the signaling components of the antigen receptor by the association of SHP-1 with the receptor complex. Rather, the co-ligation of sIg and the FcR gives rise to changes in the phosphorylation state of a limited set of proteins, which then alters their ability to interact with components of the antigen receptor. The results also suggest that the local alterations in phosphatidylinositol phosphate metabolism by enzymes associated with the antigen receptor complex, namely PI3-kinase and Ship, may be critical in the regulation of the opening and closure of the plasma membrane calcium channel.
While this manuscript was in preparation, Chacko and co-workers (41) and Ono and co-workers (51) also reported finding the enhanced phosphorylation of Ship and its association with Shc and the FcR following co-ligation of sIg and the FcR.
The murine B cell line A20
was obtained from the American Type Culture Collection. The cells were
cultured in RPMI 1640 medium containing 10% fetal calf serum and 50 µM -mercaptoethanol. Rabbit F(ab
)2 and
intact anti-mouse IgG were obtained from Cappel (Durham, NC).
Horseradish peroxidase-labeled reagents (sheep F(ab
)2
anti-mouse IgG, sheep F(ab
)2 anti-rabbit IgG, and protein
A) and ECL reagents were obtained from Amersham Corp. Gamma-Bind Plus
was from Pharmacia Biotech Inc. Monoclonal antibodies to
phosphotyrosine (RC20H) and to the signaling intermediates Shc, Vav,
Gap, Erk1, Erk2, Grb2, PI3-kinase, and PLC
1 were obtained from
Transduction Laboratories (Lexington, KY). Monoclonal antibodies to
CD19, CD32 (2.4G2), and CD22 were obtained from Pharmingen (San Diego,
CA). Polyclonal antibodies to Cbl, Shc, Grb2, Erk2, PLC
2, JAK1,
JAK2, and c-Jun amino-terminal kinases 1 and 2 were from Santa Cruz
Biotechnology (Santa Cruz, CA). Polyclonal antibodies to Ship (5340 and
5367) were provided by Dr. Larry Rohrschneider (Fred Hutchinson Cancer Research Center) and have been described previously (40). Polyclonal antibodies to PLC
1 and PLC
2 were obtained from within
Bristol-Myers Squibb; antibodies to Fyn, Btk, Syk, Lyn, Ig
, and
Ig
were obtained from Dr. Joe Bolen (Bristol-Myers Squibb). The
biotinylated peptides CAENTITY(p)SLL and CAENTITYSLL were synthesized
by SYNPEP (Dublin, CA) with a biotin molecule followed by an aminohexyl
spacer group linked through the amino-terminal cysteine. All other
reagents were from Sigma.
Cells at 2-3 × 108/ml were stimulated
with either F(ab)2 or intact anti-mouse IgG for the
indicated times. The reaction was stopped by adding 10 ml of ice-cold
PBS to each tube. The cells were pelleted by centrifugation at 4000 rpm
for 30 s; the supernatants were removed, and the cells were lysed
at 1 × 108/ml in CHAPS lysis buffer (8). The lysates
were either used immediately or stored frozen at
70 °C.
Immunoprecipitations were carried out at 4 °C using 5 µg of
antibody/1 ml of cell lysate. After 1 h, 80 µl of a 50%
suspension of protein G or protein A was added to the samples, and
these then rocked continuously for an additional 3 h at 4 °C.
The samples were washed four times with 1 ml of PBS containing 10 mM CHAPS, 500 µM sodium orthovanadate, and
200 µM phenylmethylsulfonyl fluoride. After the last wash the pellets were resuspended in 100 µl 2 × SDS sample buffer. Lysates and immunoprecipitates were analyzed by SDS-polyacrylamide electrophoresis (PAGE) and transferred to polyvinylidene difluoride membranes The membranes were blocked with PBS containing 5% bovine serum albumin, 1% ovalbumin, and 1 mM sodium
orthovanadate, and the blots were then probed with the indicated
antibodies and visualized by ECL (Amersham).
Affinity precipitations with peptides were carried out by adsorbing the biotinylated peptides to strepavidin agarose (50 µg of peptide and 50 µl of packed resin/5 × 107 cells) in 1 ml of PBS for 2 h at 4 °C. The beads were washed three times with PBS and once with CHAPS lysis buffer prior to addition to cell lysates. The affinity precipitates were then treated in the same way as the immunoprecipitates as outlined above.
IL-2 ProductionThe stimulation of IL-2 release from A20 cells was measured by enzyme-linked immunosorbent assay as described previously (8). The cells were stimulated with antibodies and reagents as indicated for 24 h. The cell supernatants were collected and assayed immediately or stored frozen. The viability of the cells after treatments was determined by exclusion of trypan blue and found to be greater than 94% at the highest concentration of wortmannin that was used (100 nM).
Calcium MobilizationThe mobilization of both intracellular and extracellular calcium was measured on an SLM 8000 spectrofluorometer using the indicator dye Indo-1 (Molecular Probes, Eugene, OR), as described previously (8). Briefly, the cells (1 × 106/ml) in RPMI 1640 medium containing 10% fetal bovine serum were loaded with Indo-1 AM for 45 min at 37 °C. The cells were washed three times in Hank's balanced salt solution containing 10 mM HEPES and 1% fetal calf serum. The cells were then resuspended at 1 × 106/ml in the same buffer for analysis.
Co-cross-linking the
FcR to sIg abrogates the influx of extracellular calcium within 1-2
min of activation of the cells (Ref. 8 and Fig. 7A). To
address the changes that occur on activation of the cells through sIg
alone or when sIg is cross-linked to the FcR, we studied the early
cellular responses that occur within this time frame. As previously
reported, stimulation of the cells through sIg with F(ab)2
rabbit anti-mouse IgG gave rise to a rapid increase in the tyrosine
phosphorylation of many intracellular proteins. Furthermore,
co-cross-linking the FcR with the sIg using intact anti-mouse Ig did
not markedly alter the overall increase tyrosine phosphorylation (Fig.
1A) compared with stimulation with F(ab
)2 rabbit anti-mouse IgG. However, on closer
examination some differences could be observed. Immunoprecipitation of
the FcR confirmed that, in contrast to stimulation of the cells with F(ab
)2 rabbit anti-mouse Ig, activation of the cells with
intact anti-mouse Ig induced the phosphorylation of the FcR (Fig.
1B and Ref. 30). In addition, several changes in the levels
of tyrosine phosphorylation of other proteins that co-precipitated with
the FcR could also be seen (Fig. 1B). These changes included a decrease in the tyrosine-phosphorylated proteins that run with molecular weights of about 34,000, 39,000, and 73,000. On longer exposure of the blots, an increase in the presence of a
tyrosine-phosphorylated protein of 145-150 kDa could also be seen
(Fig. 1B, lane 3).
We next examined the effect of co-cross-linking the FcR to sIg on the
tyrosine phosphorylation of various intracellular proteins that may be
involved in the signal transduction pathways. Immunoprecipitations were
carried out to isolate specific proteins from lysates of cells
activated either through the antigen receptor alone or by the
co-cross-linking of sIg and FcR. As described earlier (Fig. 1B and Ref. 30), concomitant activation of the cells through the sIg and Fc receptors induced the tyrosine phosphorylation of the
FcR, but there were also increases in other proteins.
Immunoprecipitation of either Grb2 or Shc from cell lysates revealed
that there was an increase in the tyrosine phosphorylation of a protein
doublet centered around 52 kDa, corresponding to Shc, when the cells
were activated with intact anti-IgG (Fig. 2A, lanes 3 and
6). Western analysis of immunoprecipitates of
either Shc or Grb2 revealed that there was a weak association between
Grb2 and Shc in cells activated with F(ab)2 anti-IgG (Fig.
2B, lanes 5 and 8), and this was enhanced in
cells activated with the intact anti-IgG (Fig. 2B, lanes 6 and 9). In addition, co-cross-linking of the sIg and Fc
receptors gave rise to a significant increase in the presence of a
tyrosine-phosphorylated protein of 150 kDa in the immunoprecipitates of
either Grb2 or Shc (Fig. 2A, lanes 3 and 6).
Subsequent analysis of these immunoprecipitates revealed that this
phosphoprotein was recognized by antibodies to the inositol 5
-phosphatase Ship and that activation of cells by co-ligation of sIg
and FcR gave rise to an increase in the levels of Ship found associated
with Shc and Grb2 (Fig. 2B, lanes 6 and
9).
Immunoprecipitation of Ship from cell lysates using the two different
polyclonal antibodies revealed that there was a small increase in
tyrosine phosphorylation of the enzyme following stimulation of the
cells with F(ab)2 anti-IgG, but this was markedly enhanced on co-cross-linking sIg to the FcR (Fig. 3A, lanes 3 and
6). In the same immunoprecipitates there was
an additional highly phosphorylated band of about 55 kDa that strongly
resembled the phosphorylated FcR. Since antibodies that recognize the
FcR on Western blots are not available, it was not possible to directly
test this. However, Western analysis, using the polyclonal anti-Ship
antibodies, after stimulation of the cells with intact anti-IgG and
immunoprecipitation of FcR, revealed that Ship was indeed associated
with the FcR when the receptor was phosphorylated (Fig. 2B, lane
3). Thus Ship was the 150-kDa tyrosine phosphoprotein that was
originally detected on antiphosphotyrosine blots of FcR
immunoprecipitates (Fig. 1B, lane 3).
To confirm the association between Ship and the FcR, affinity
precipitations were carried out using immobilized peptides
corresponding to the unphosphorylated and phosphorylated forms of the
cytoplasmic domain of FcRIIb. The immobilized phosphorylated peptide
was able to bring down Ship from the lysates of unactivated A20 cells (Fig. 3B lane 3), whereas the strepavidin beads alone or the
beads containing the unphosphorylated peptide did not (Fig. 3B,
lanes 1 and 2). Activation of the cells with either
F(ab
)2 or intact anti-IgG did not significantly change the
interaction of the peptides with Ship (data not shown).
In contrast to the increase in phosphorylation of FcR, Ship, and Shc, a
significant reduction in the tyrosine phosphorylation of CD19 were
observed at all times following stimulation of the cells with intact
anti-IgG. Stimulation of the cells through sIg alone gave rise to a
marked increase in the tyrosine phosphorylation of CD19 (Fig. 4A,
lane 2). This was greatly reduced when the cells were activated by co-cross-linking sIg to the FcR. (Fig. 4A, lane 3). It has been shown that phosphorylation of CD19 results in its
association with PI3-kinase (28, 43). In agreement with this, analysis
of immunoprecipitates of CD19 from cells activated with
F(ab)2 anti-IgG revealed that PI3-kinase was associated with the phosphorylated CD19 (Fig. 4B, lane 5); however, the
enzyme was not detected in immunoprecipitates from cells that had been activated by co-cross-linking sIg to the FcR (Fig. 4B, lane
6). It was not possible to reprobe these blots with an anti-CD19
antibody, since those available do not Western blot.
Stimulation of A20 cells with F(ab)2 anti-IgG gives rise
to an initial mobilization of calcium from intracellular stores, which
lasts for about 60 s, followed by a second phase, which involves
the influx of extracellular calcium (Fig. 5B, top
panel, and Ref. 8). Activation of the cells with
F(ab
)2 anti-IgG in the presence of wortmannin, which has
been shown to inhibit PI3-kinase (37, 38), resulted in the
dose-dependent inhibition of the later, extracellular,
component of the calcium mobilization within the cell (Fig.
5A). At higher concentrations of wortmannin (>60
nM) slight inhibition of mobilization of calcium from
intracellular stores was observed. When wortmannin was added after the
influx of extracellular calcium had been initiated, this influx was
rapidly curtailed (Fig. 5B, bottom panel). The inhibitor
also blocked the F(ab
)2 anti-IgG-induced production of
IL-2 in a dose-dependent fashion (Fig. 5C).
Wortmannin at concentrations up to 200 nM had no effect on
the patterns of tyrosine phosphorylation of whole cell lysates and
immunoprecipitates of CD19, FcR, and Shc induced by F(ab
)2
anti-IgG (data not shown).
As shown in Fig. 1, the phosphorylation of many of the proteins in the
cell lysates was not altered with stimulation of the cells with intact
anti-IgG. When this was analyzed in more detail, no significant
differences in the ability to induce the tyrosine phosphorylation
of Syk, Cbl, PLC1, CD22, Ig
, PI3-kinase, and JAK2 (Fig. 6,
A and B) or Vav, Gap, Tyk, and
JAK1 (data not shown) were observed between cells activated by
the two different modes of triggering. The changes in the other bands
visible in the immunoprecipitates in Fig. 6 reflect the change in
phosphorylation of p73 (g) and p150 (h). It was
not possible to detect changes in the phosphorylation of Lyn or Btk
following activation of the cells through sIg or sIg with the FcR (data
not shown). Fyn was constitutively phosphorylated in unactivated cells
(Fig. 6A, g); on activation of the cells with either
F(ab
)2 or intact anti-IgG there was a slight reduction in
mobility of the phosphoprotein, suggesting that there was a change in
the posttranslational modification of the enzyme. Stimulation of the
cells with intact anti-IgG also induced marked increases in the
tyrosine phosphorylation of PLC
2 and Ig
; however, this phosphorylation appeared to be slightly less than that induced with
stimulation of the cells through sIg alone using F(ab
)2 anti-IgG (Fig. 6). Overall, these results indicate that
co-cross-linking the FcR to sIg does not simply cause a global decrease
in tyrosine phosphorylation of intracellular proteins.
Kinetics of Fc Receptor-induced Changes
Addition of intact
anti-IgG to A20 cells that have previously been activated with
F(ab)2 anti-Ig causes a very rapid drop in intracellular
calcium by closing the plasma membrane channel, thereby halting the
influx of extracellular calcium (Fig. 7A and Ref. 8). We also found that the addition of intact anti-IgG to
previously activated cells stimulated the phosphorylation of FcR, Shc,
and Ship, the association of p150 with Shc and Grb2, and a decrease in
the phosphorylation of CD19. To determine whether the induction of
these changes in tyrosine phosphorylation had kinetics similar to the
closing of the calcium channel, cells were activated with
F(ab
)2 anti-Ig for 1 min, and then sIg was co-cross-linked
to the FcR by the addition of intact anti-Ig to the ongoing reaction.
Within 15 s following the addition of intact anti-Ig to the cells
an increase in the tyrosine phosphorylation of the FcR could be seen;
after 45-60 s phosphorylation of the FcR was maximal (Fig. 7B,
a). Similarly, the increases in the phosphorylation of Shc and the
associated p150 Ship could be detected in immunoprecipitates of Shc or
Grb2 within 15-30 s after the addition of intact anti-Ig and reached a
maximal level after about 60 s (Fig. 7B, c and
d). Co-cross-linking the FcR to sIg stimulated the
dephosphorylation of CD19 with similar kinetics (Fig. 7B, b). The results indicate that the changes in tyrosine
phosphorylation and protein associations of the FcR, CD19, and Shc are
consistent with the kinetics of closing of the plasma membrane calcium
channel, and one or more of these may induce the change in ion
mobilization.
The FcR plays a crucial role in the regulation of activation of B cells (3-6). Co-cross-linking the antigen and Fc receptors gives rise to the phosphorylation of the receptor and the generation of a dominant inhibitory signal that abrogates the proliferation or differentiation of the lymphocytes (30). FcR phosphorylation promotes its association with other intracellular proteins such as SHP-1 (31-33) via Src homology 2 interactions. It has been proposed that the association between SHP-1 and the FcR activates the phosphatase, which then shuts down the signaling through the antigen receptor (31, 32). It has also been shown that co-cross-linking sIg to the FcR prevents the influx of extracellular calcium by closing the plasma membrane calcium channel (8). However, it is not clear how co-cross-linking the FcR to sIg abrogates the influx of calcium and whether the association of activated SHP-1 with the receptor complex curtails all of the activation signals mediated by the antigen receptor or if the effect is more restricted.
In this study we show that co-cross-linking sIg to the FcR gives rise
to very selective modulation of the signaling pathways stimulated by
the antigen receptor. Co-ligation of sIg and the FcR neither inhibits
nor reverses the stimulation of phosphorylation of many of the
intermediate proteins thought to be involved in signaling through the
antigen receptor. There was no detectable difference in the
phosphorylation of Cbl, Vav, Fyn, Gap, Syk, PI3-kinase Ig, and CD22
between activation of the cells through either receptor.
Phosphorylation of the Ig
chain of the antigen receptor complex was
also induced by triggering either with F(ab
)2 or intact
anti-IgG, although the levels of phosphorylation stimulated by the
latter treatment were slightly diminished. The significance of this is
unclear at present.
It has been shown that both PLC1 and PLC
2 become phosphorylated
and activated following stimulation of the cells through the antigen
receptor (19-21). This then generates the formation of inositol
trisphosphate, which then leads to the release of calcium from
intracellular stores. In addition, co-cross-linking sIg to the FcR
curtails the influx of extracellular calcium but does not significantly
inhibit the mobilization of calcium from intracellular stores (7, 8).
In agreement with these observations, our experiments showed that
co-cross-linking sIg to the FcR had no significant effect on the
phosphorylation of PLC
1. Phosphorylation of PLC
2 was still
significantly enhanced over unstimulated cells, although the levels
were slightly decreased. These results are in contrast to a recent
report using murine splenic B cells, in which activation of PLC
2 was
significantly inhibited upon co-ligation of the FcR to sIg (48). The
reason for this disparity is not clear. It may reflect the differences
in cell populations that were used or the conditions of activation.
However, it is difficult to directly compare the two studies, since in
the other report the influence of the signals on calcium mobilization
from the intracellular and extracellular stores was not described.
Stimulation of B cells through sIg gives rise to the rapid ordered activation of a cascade of protein kinases (16-18, 44-47). No differences were observed in the activation of the kinases Syk, Lyn, Fyn, Btk, Erk1, Erk2, and c-Jun amino-terminal kinase 1, between stimulation of the cells through either sIg alone or co-ligation of the FcR to sIg (data not shown). This suggests that these enzymes are not the target for FcR-associated SHP-1.
Co-cross-linking Fc and antigen receptors did give rise to some very
specific changes in the signal transduction pathway. As reported
previously (30), stimulation of the lymphocytes with intact anti-Ig
gave rise to a marked increase in the phosphorylation of the FcR.
Immunoprecipitation of the FcR from activated cells also revealed other
changes in the cells. A decrease in the phosphorylation of three
proteins that co-precipitated with the immune complexes was
consistently observed. These proteins, of about 34, 39, and 73 kDa,
could also be detected in other immunoprecipitates (see Figs.
2A and 4A). The identity of these proteins is not
known, but preliminary experiments have indicated that they are not the Ig and Ig
chains, Syk, SHP-1, SHP-2, or Raf-1. In addition to these changes, it was also possible to detect an increase in the level
of a phosphoprotein of about 150 kDa in the immunoprecipitates of the
FcR from cells activated with intact anti-IgG.
Immunoprecipitation of Shc or Grb2 following activation of the B cells
by co-ligation of sIgG and FcR revealed that there was enhanced
phosphorylation of a doublet centered around 52 kDa, corresponding to
Shc. In addition, immunoprecipitates of both Grb2 and Shc from cells
activated by co-cross-linking sIg and FcR showed an increase in the
levels of an associated phosphoprotein of about 150 kDa. Analysis of
the immunoprecipitates of Shc, Grb2, and FcR indicated that this
protein was Ship, the inositol 5-phosphatase (39, 40). Subsequently,
immunoprecipitation of Ship revealed that the phosphorylation of the
enzyme increased following activation of cells through sIg alone;
however, the level of phosphorylation was further significantly
enhanced when the cells were activated by co-ligation of sIg and FcR.
Ship can associate with Shc through the phosphotyrosine binding domain
of the latter (39, 40), and there are also several proline-rich
sequences within Ship that may allow interaction with the SH3 domains
of other proteins such as Grb2 (40). Since Shc interacts with Grb2,
Ship could also indirectly associate with Grb2 via a mutual interaction
with Shc. Stimulation of an increase in the phosphorylation of Ship could give rise to an enhanced association with Shc and thereby Grb2.
These interactions would account for the increased level of Ship seen
in immunoprecipitates of Shc and Grb2 in this study. Alternatively the
phosphorylation of Ship may promote the interaction of the enzyme with
the Src homology 2 domains of Grb2 or Shc. At present the factors
regulating these interactions have not been elucidated.
Ship was also found present in the immunoprecipitates of the FcR after
stimulation of the cells by co-ligation of antigen and Fc receptors. It
could not be detected in immunoprecipitates of FcR from cells activated
through sIgG alone. Additionally, immunoprecipitates of Ship from cells
stimulated with intact anti-IgG revealed the presence of a
phosphoprotein that closely resembled the FcR. These results suggest
that the phosphorylated FcR can associate with Ship, probably through
interaction with the Src homology 2 domain of the enzyme. A
biotinylated peptide corresponding to the sequence surrounding the
phosphorylated tyrosine in the cytoplasmic domain precipitated Ship
from the lysate of unstimulated cells, whereas the nonphosphorylated
form of the peptide did not. This shows that the phosphorylation of
this tyrosine in the cytoplasmic domain of the FcR is sufficient to
promote the interaction between the receptor and Ship. Affinity
precipitates from cells stimulated with F(ab)2 or intact
anti-IgG did not significantly change the ability of the phosphorylated
peptide to interact with Ship. This indicates that phosphorylation of
an intermediary protein such as Shc is not essential for the
interaction.
Ship catalyzes the hydrolysis of the 5-phosphate of inositol tetraphosphate and phosphatidylinositol 3,4,5-trisphosphate (39). Our data indicate that interaction of this enzyme with the antigen receptor complex, through its association with the phosphorylated FcR, is likely to alter locally the metabolism of either inositol phosphates or phosphatidylinositol phosphates. This may directly regulate the activity of the plasma membrane calcium channel. It is interesting to note that retroviral expression of Ship in FD-fms cells results in strong inhibition of growth (40). This inhibition may be due to the alteration of the ability of cells to mobilize calcium.
In contrast to the effect on FcR, Ship, and Shc, co-ligation of sIg and
FcR resulted in the decrease of phosphorylation of several proteins,
including CD19 and other, as yet uncharacterized, proteins. Activation
of cells through the antigen receptor alone results in an increase in
the tyrosine phosphorylation of both CD19 and CD22 (9, 10, 12, 49). In
our experiments, co-ligation of the FcR to sIg greatly decreased the
phosphorylation state of CD19 but did not significantly alter the level
of phosphorylation of CD22. Several signaling molecules are found
associated with CD19 following activation of B cells through the
antigen receptor; these include PI3-kinase, Vav, and Fyn (28). Our
results show that on co-cross-linking the antigen and Fc receptors, the
reduction in tyrosine phosphorylation of CD19 gave rise to a
corresponding decrease in the level of associated PI3-kinase. This
suggests that PI3-kinase activity in the receptor complex may play a
crucial role in the regulation of influx of extracellular calcium.
Addition of wortmannin to the cells prior to activation with
F(ab)2 anti-IgG markedly abrogated the secondary influx of
extracellular calcium although having only a modest effect on the
mobilization of calcium from intracellular stores. The inhibitor also
induced a rapid decrease in the levels of intracellular calcium in
cells that were already mobilizing extracellular calcium and blocked
sIg-induced IL-2 production. Wortmannin inhibits PI3-kinase activity at
concentrations down to 10
8 M (37, 38, 50);
therefore, its effect on the mobilization of calcium is consistent with
(but not proof of) PI3-kinase activity, in association with CD19,
playing a role in the regulation of the calcium channel.
Analysis of cells that were stimulated initially through sIg and then
subsequently restimulated by co-ligation of sIg and the FcR indicated
that the changes in tyrosine phosphorylation of the proteins occurred
rapidly. The kinetics of these biochemical changes are of the right
time frame to be able to account for the changes in influx of
extracellular calcium, which diminishes over the first few minutes
following co-ligation of the receptors. It should be pointed out that
the effect of intact anti-Ig on tyrosine phosphorylation of all of
these proteins was not simply due to the abrogation of an influx of
extracellular calcium. Removal of extracellular calcium with 2.5 mM EGTA did not alter the tyrosine phosphorylation state of
any of the proteins when the cells were activated with
F(ab)2 anti Ig (data not shown).
In summary, co-cross-linking of the FcR to sIg gives rise to a very
discrete set of changes in signal transduction. Many of the signaling
pathways that were stimulated following ligation of the antigen
receptor alone were unaltered when sIg was co-cross-linked to the FcR.
These include the activation of several different protein kinases, the
phosphorylation of many intracellular proteins, and the activation of
the transcription factor NFB (42) (data not shown). Thus the role of
SHP-1 in the negative regulation of B cell receptor signaling is very
specific and is not a global shutdown of tyrosine phosphorylation at
the antigen receptor. Our data suggest that dephosphorylation of CD19
(perhaps by SHP-1) and the subsequent loss of association of PI3-kinase
are crucial steps in altering calcium mobilization. This may be
mimicked by the inhibition of PI3-kinase activity with wortmannin. In
addition, our studies have shown that the phosphorylation of the FcR
promotes its interaction with Ship, an inositol 5
-phosphatase. Thus
co-ligation of the FcR and sIg markedly changes the overall activity of
lipid-modifying enzymes that are associated with the receptor complex.
In particular, it is possible that the local levels of
phosphatidylinositol 3,4,5-trisphosphate are critical to the regulation
of opening and closing of the calcium channel.
We thank Dr. Bruce Cohen (Bristol-Myers Squibb) for critical reading of the manuscript and helpful discussions and Dr. Joe Bolen (DNAX) for providing antibodies.