Intracluster Restriction of Fc Receptor (cid:103) -Chain Tyrosine Phosphorylation Subverted by a Protein-tyrosine Phosphatase Inhibitor*

This study shows that aggregation of U937 cell high affinity IgG Fc receptor (Fc (cid:103) RI) results in the transient tyrosine phosphorylation of Fc (cid:103) RI (cid:103) -chain but not the phosphorylation of (cid:103) -chains associated with nonaggregated IgA Fc receptors (Fc (cid:97) R) on the same cells. Thus, normally, tyrosine phosphorylation of (cid:103) -chains is limited to FcR in aggregates. In contrast, aggregation of Fc (cid:103) RI in the presence of vanadate induced the sustained tyrosine phosphorylation of Fc (cid:103) RI (cid:103) -chains and the rapid and extensive phosphorylation of nonaggre- gated Fc (cid:97) R (cid:103) -chains and low affinity IgG Fc receptors (Fc (cid:103) RII). This global phosphorylation of motifs on nonaggregated FcR was also detected upon aggregation of Fc (cid:97) R or Fc (cid:103) RII, which induced the phosphorylation of nonaggregated Fc (cid:103) RI (cid:103) -chains. Vanadate prevented dephosphorylation of proteins and increased kinase activ- ity in stimulated cells. Evidence failed to support alternative explanations such as acquisition of phospho- (cid:103) through subunit exchange or a coalescence of nonaggre- gated with aggregated FcR. It is likely, therefore, that activated kinases interacted with nonaggregated FcR in stimulated cells. Pervanadate induced the tyrosine phosphorylation of (cid:103) -chains in the

Aggregation controls signaling through Fc receptors (FcR) 1 (1,2). Cross-linking of high affinity IgG Fc receptors (Fc␥RI) or IgA Fc receptors (Fc␣R) on monocytic U937 cells results in the rapid generation of oxygen radicals (3)(4)(5)(6) and tyrosine phosphorylation of their respective ␥-chains (6,7). Cross-linking of the low affinity Fc receptor for IgG (Fc␥RII), which lacks ␥-chains (8,9) results in the phosphorylation of tyrosine motifs in the cytoplasmic domain of the receptor (10). The ␥-chain immunoreceptor tyrosine activation motifs containing YXXL sequences (11) are substrates for Src family kinases (12) that are capable of reacting with the nonphosphorylated motif and of binding through SH2 domains to the tyrosine-phosphorylated product Y*XXL. Activity of Src family kinases Hck and Lyn associated with Fc␥RI are increased upon Fc␥RI crosslinking (13). Binding of Lyn to high affinity IgE Fc receptor (Fc⑀RI) ␥-chains is increased by Fc⑀RI aggregation (14).
For Fc⑀RI, phosphorylation of tyrosine motifs is restricted to aggregates, with little involvement of "bystander" nonaggregated receptors (15)(16)(17). Limitation to intracluster units is also evidenced by sustained binding of Lyn to Fc⑀RI in isolated aggregates and phosphorylation of aggregate subunits in preference to an exogenously supplied substrate (14,18). Both kinds of evidence suggest a spatial restriction of kinase activity to aggregates. It has been suggested that restriction is due to a requirement for aggregated receptors as sites for kinase activation (14,19) and to a requirement that substrate be in the aggregated state (18). Another possibility is that kinase activity is under a positive control preventing activity in nonaggregated receptors. If this is the case, inhibition of the control should allow tyrosine phosphorylation of nonaggregated as well as aggregated FcR.
Evidence presented from an earlier report (20) and the present report is consistent with the second model. We used an assay system that allowed us to examine the effect on one FcR type of cross-linking another FcR type on the same cell. Western blots of precipitated FcR showed that normally there is no detectable kinase activity for nonaggregated FcR. However, in the presence of a phosphatase inhibitor, aggregation of one FcR type induced rapid and extensive phosphorylation of tyrosine motifs on noncross-linked FcR. The data in this report implicate phosphatases as necessary to prevent global FcR involvement and suggest that normal intracluster restriction of ␥-chain phosphorylation may be due to this vanadate-sensitive mechanism.
FcR Activation and the Respiratory Burst-U937 10.6 cells in superoxide (O 2 . ) assay medium were added to an equal volume of a second medium containing 10 g/ml control antibodies or mAb 197, which cross-links Fc␥RI through Fc and Fab trivalent binding (23). Alternatively, cells were reacted with 5 g/ml control or anti-FcR antibody (24) for 20 min at 22°C, centrifuged, resuspended in O 2 . assay medium, and added to an equal volume of a second medium containing 40 g/ml anti-murine antibody. FcR aggregations were done at 37°C. For assaying tyrosine phosphorylations or O 2 . , the second medium was 10 Ϫ4 M luminol in phosphate-buffered saline (24). Orthovanadate (Na 3 VO 4 ), buffered and at a concentration of 200 M, was present during FcR aggregation except where indicated. To measure respiratory bursts, luminol-mediated chemiluminescence was monitored on a Pharmacia 1250 luminometer and is expressed in mV, as described previously (24). Cellular Tyrosine Phosphoproteins-Cells reacted with antibodies were rapidly chilled, washed twice with cold phosphate-buffered saline, and boiled for 20 min in nonreducing SDS sample buffer. For reduction, boiling was continued for 3 min following the addition of 4% 2-mercaptoethanol. Proteins were separated by SDS-PAGE and analyzed by Western blot. phospho-␥ in cellular proteins was distinguished from nonphosphorylated ␥-chains through the migration pattern on nonreducing gels. Unreduced phosphorylated ␥-chains migrate to a broad ϳ28-kDa position compared with unphosphorylated bands at ϳ22 kDa (6).
Assay for in Vitro Subunit Exchange-Cells were incubated with 197 or HB63 in activation medium, and solubilized at a concentration of 5 ϫ 10 6 /ml in 0.5% Nonidet P-40, 0.5% digitonin lysis buffer (6). Duplicate Fc␣R precipitations were conducted on lysates from 1.5 ϫ 10 6 cells on A77-conjugated beads. Also prepared were Fc␥RI-and Fc␣R-depleted lysates from 2 ϫ 10 7 cells. FcR were depleted through successive adsorptions by bead-conjugated antibodies: goat anti-murine for 3 h, goat anti-murine for 1 h, A77 for 16 h, goat anti-murine for 1 h, hIgG for 1 h, and A77 for 1 h. Preadsorption was verified through anti-phosphotyrosine immunoblot of adsorbed proteins to detect FcR-associated phospho-␥. Duplicate Fc␣R precipitates were either kept on ice as controls or rotated with preadsorbed lysates from 4 ϫ 10 6 cells for 1 h. Control precipitations were executed on lysate aliquots representing 2 ϫ 10 6 cells to assess the presence of free phospho-␥ following FcR depletion. The control precipitations were performed on bead-conjugated PY20 or on anti-␥ antibody-coated protein A-Sepharose for 1 h. All precipitates were washed three times with lysis buffer, and boiled in SDS-sample buffer. Precipitates were separated by SDS-PAGE and analyzed by Western blot.
Cyofluorographic Assay for Measuring FcR Aggregation/Internalization-Cells were incubated at 37°C for 18 min with RPMI alone or containing 197 or HB63 (5 g/ml). These cells were centrifuged, washed briefly, and incubated an additional 10 min with 5 g/ml A77 or P3. Vanadate (200 M) was present or absent throughout. All samples were washed three times with 0.1% bovine serum albumin in phosphatebuffered saline, stained with FITC-conjugated F(ab)Ј 2 goat anti-mouse IgG at 4°C, and analyzed by cytofluorography as described previously (23,25). Results are expressed as FITC-antibody binding sites/cell.

Transient Tyrosine Phosphorylations and the Effect of Vanadate-An early response by monocytic cells triggered through
Fc␥RI is the tyrosine phosphorylation of several proteins including Fc␥RI ␥-chains (6,7). Triggering also induces a transient respiratory burst that is tightly coupled to de novo receptor cross-linking (23). To determine whether induced tyrosine phosphorylations were also transient, we reacted U937 cells with anti-Fc␥RI mAb 197, which effectively cross-links because of trivalent binding. The time course of induced tyrosine phosphorylations was measured by Western blot. Under normal conditions ( Fig. 1, right panel), tyrosine phosphorylations of pp72 (Fig. 1A), and ␥-chains (Fig. 1B) were transient, peaking by ϳ3-5 min. Additional transiently phosphorylated proteins were detected with longer exposures. The lack of a sustained phosphorylation suggests that phosphatase activity is present in aggregated receptors. In the presence of vanadate, however, phosphorylations of 72-kDa proteins (Fig. 1A), ␥-chains ( . production (Fig. 1C). Respiratory burst kinetics were similar to the tyrosine phosphorylation response in the absence of vanadate. Absence of Bystander Involvement during FcR Cross-linking-According to reports, Fc⑀RI aggregation does not result in the phosphorylation of nonaggregated (bystander) Fc⑀RIassociated ␥-chains (15)(16)(17). To determine whether nonaggregated FcR in monocytic cells become phosphorylated, we used an assay system in which FcR of one class were aggregated and nonaggregated FcR of another class were examined for phosphorylation of their associated ␥-chains. We aggregated Fc␥RI in the absence of vanadate for an optimal time (5 min) (Fig. 1) and examined aggregated Fc␥RI␥ and nonaggregated Fc␣R␥ by immunoprecipitating the receptors from lysates of the cells. As shown in Fig. 2, nonaggregated Fc␣R contained only a trace of phosphotyrosine compared with aggregated Fc␥RI. In the converse experiment, Fc␣R were aggregated with little effect on Fc␥RI␥ (Fig. 2). Longer incubations did not increase phosphorylation of nonaggregated receptors (not shown). These results indicate that nonaggregated bystanders were not significantly targeted by aggregation-activated kinases.
Tyrosine Phosphorylation of Bystander FcR in the Presence of Vanadate-Because Fc␥RI triggering in the presence of vanadate resulted in extensive phosphorylation of cellular ␥-chains ( Fig. 1), we examined the possibility that this may have included the phosphorylation of nonaggregated FcR. We crosslinked Fc␥RI and examined Fc␣Rg in receptor immunoprecipitates. As shown by Western blot (Fig. 3A), ␥-chains coprecipitating with nonaggregated Fc␣R were extensively phosphorylated. Blotting with anti-␥ antibodies confirmed this and demonstrated similar intensities of phospho-␥ bands in nonaggregated Fc␣R and aggregated Fc␥RI (Fig. 3A). In the same experiment, we cross-linked Fc␣R and examined nonaggregated Fc␥RI␥ in receptor immunoprecipitates (Fig. 3B). As shown (Fig. 3B), nonaggregated Fc␥RI␥ was phosphorylated in anti-Fc␣R-activated but not in nonactivated cells. Recoveries of receptors in precipitates in all cases were assessed by anti-␥chain blots.
Cross-linking of Fc␣R was also executed in the presence of hIgG1 to block a potential Fc interaction of anti-Fc␣R with Fc␥RI. The results show that ␥-chains in the hIgG1-Fc␥RI complexes had become phosphorylated (Fig. 4, A and B, lane 1). Furthermore, the possibility of anti-murine antibody co-crosslinking and stimulating via bound hIgG1 was also eliminated by an oxidase assay in which cells preincubated with hIgG1 or not and incubated with the same set of antibodies were found to be activated only through IgA receptors. Values from the oxidase assay (in mV) were 4451 Ϯ 228 for A77-hIgG1-coated cells, 4414 Ϯ 752 for A77-coated cells, and 19 Ϯ 9 and 9 Ϯ 2 for P3-hIgG1-and P3-reacted cells, respectively. All received second antibody. These results eliminated Fc bridging as the source of nonaggregate involvement.
In the same experiment, Fc␥RI were cross-linked (Fig. 4, A  and B, lane 5), causing the extensive phosphorylation of Fc␣R. Anti-␥ immunoblots of aggregated (lane 6) compared with nonaggregated Fc␥RI (lane 1) show similar intensities of phospho-␥ bands, suggesting that similar numbers of chains in nonaggregated ligand-occupied Fc␥RI were phosphorylated. Noticeable decreases in unphosphorylated ␥ and increases for phospho-␥ within individual samples (lanes 1 and 6) imply an efficient shift in state. Similar mobility patterns for phospho-␥ in each case are consistent with equivalent site modifications by kinases. These results demonstrate the efficient phosphorylation of tyrosines on nonaggregated FcR ␥-chains.
To examine phosphorylation kinetics, we measured the onset  1 and 2), aggregated Fc␥RI (lanes 3 and 4), and nonaggregated Fc␥RI (on 32.2-conjugated beads, lanes 5 and 6) were precipitated, and the nonreduced precipitates were separated by SDS-PAGE on 16% gels and analyzed by anti-phosphotyrosine Western blot. The bracket denotes nonphosphorylated ␥, and bars denote phosphorylated ␥.  6) were precipitated via protein G-Sepharose. Fc␥RI was precipitated via hIgGconjugated beads (lanes 7 and 8). Nonreduced precipitates were electrophoresed and analyzed by sequential anti-phosphotyrosine and anti-␥ Western blot. Anti-␥ blots of aggregated Fc␣R show that this precipitate was inefficiently recovered (lane 5). Brackets denote phospho-␥. and maximal phosphorylation times for ␥-chains of aggregated Fc␥RI and nonaggregated Fc␣R. As shown by anti-phosphotyrosine and anti-␥ Western blot (Fig. 5), phosphorylation of Fc␣R␥ was detectable by 6 min and plateaued by 18 min (Fig.  5). Fc␥RI␥ phosphorylation was detectable by 3 min and plateaued between 12 and 18 min. Similar maximal intensities were observed, and anti-␥ blots show similar amounts of FcR in precipitates. These results show the rapid and prolonged phosphorylation of aggregated and nonaggregated FcR ␥-chains.
Phospho-␥ Is Not Acquired through Subunit Exchange-To determine whether nonaggregate phosphorylation could be an artifact of immunoprecipitation in which phospho-␥ exchanged for unphosphorylated ␥, or vice versa, immunoprecipitates of nonaggregated Fc␣R (containing phospho-␥-chains) were incubated for the usual time with an unstimulated lysate precleared of endogenous FcR ␣-chains. Exchange was judged by comparing the original with lysate-incubated precipitates. As shown in Fig. 6A, these two were identical, indicating that Fc␣R did not exchange its associated phospho-␥ during immunoprecipitation. In the converse experiment (Fig. 6B), unphosphorylated Fc␣R in precipitates were incubated in lysates of Fc␥RI-stimulated cells. The lysates had been precleared of Fc␥RI and Fc␣R (Fig. 6B, lower panel) but contained free phospho-␥ chains (lane 7). Exchange was again judged by comparing original with lysate-incubated precipitates. The results show that unphosphorylated ␥ in Fc␣R precipitates was not exchanged for phospho-␥. Collectively, the results show that phospho-␥ was not acquired through subunit exchange in vitro.
To determine whether phospho-␥ exchange in vivo explains the appearance of phospho-␥ in nonaggregated FcR, we triggered Fc␥RI and precipitated from the cell lysate nonaggregated Fc␥RII. Fc␥RII lack ␥-chains but are phosphorylated in cytoplasmic domain motifs upon cross-linking (10). Fig. 7A shows that nonaggregated Fc␥RII in Fc␥RI-activated but not in nonactivated cells were phosphorylated on tyrosines. Similarly, upon triggering through Fc␥RII, ␥-chains for Fc␥RI and Fc␣R became phosphorylated (Fig. 7B). These data indicate that FcR lacking exchangeable ␥-chains are phosphorylatable bystanders and, with cross-linking, are able to induce ␥-chain phosphorylation. This suggests that direct kinase activity rather than  3 and 4)) or kept on ice (lanes 1 and 2). Precipitates were washed and separated by nonreducing SDS-PAGE. phospho-␥ retained by Fc␣R precipitates was assessed by anti-phosphotyrosine and anti-␥ (not shown) Western blot. B, lack of Fc␣R ␥ exchange with phospho-␥ in the lysate. 197-stimulated and HB63-reacted nonstimulated cell precipitates were exposed to FcRcleared lysates from stimulated cells (S/lysate) (lanes 1-4) or kept on ice (lanes 5 and 6). Precipitates were washed and separated by reducing SDS-PAGE. Analysis for exchange of Fc␣R ␥-chains for phospho-␥ in the lysate was by anti-phosphotyrosine Western blot. A control precipitate to assess free phospho-␥ remaining in the FcR-depleted lysate is shown in lane 7. Depletion from the stimulated lysate (S-lysate) was verified by anti-phosphotyrosine Western blot for phospho-␥ in preadsorbed proteins (lower panel). subunit exchange in vivo explains bystander ␥-chain phosphorylation.
Bystander Phosphorylation Is Not Due to FcR Co-aggregation-We investigated the possibility that vanadate may have induced co-aggregation of nonaggregated with aggregated FcR, making nontargeted FcR␥ available to aggregate-docked kinases. Aggregation was assessed by measuring internalization of receptors. Following aggregation and a predetermined interval for internalization of Fc␥RI, the cells were fluorescently labeled to quantitate Fc␥RI and Fc␣R remaining on the surface. The results (Fig. 8) show that 197-Fc␥RI aggregates were effectively internalized (Ͼ60%) without a concomitant reduction in surface Fc␣R. As similar results were obtained in the presence and absence of vanadate (Fig. 8), the data do not support a vanadate-induced co-aggregation.
Other indirect evidence suggests a lack of co-aggregation. As shown in Figs. 3, 4, and 5, nonaggregated FcR were deficient in tyrosine phosphoproteins that co-precipitate with aggregated FcR. In several experiments, the co-precipitating panel consisted of 32-, 52-66-, and 72-kDa (Syk kinase) 2 bands. Discrete co-precipitations are consistent with a lack of co-aggregation of FcR types.
Tyrosine Phosphorylation of Nonaggregated ␥ by Treatment of Cells with Pervanadate-Treatment of nonactivated cells with vanadate prereacted with H 2 O 2 in order to produce pervanadate induced the tyrosine phosphorylation of ␥. This did not occur upon treatment with vanadate or H 2 O 2 alone. These data (Fig. 9) show that ␥-chains can become phosphorylated in the absence of any FcR cross-linking. Based on this, it appears that the kinases that interacted with nonaggregated FcR were negatively regulated by tyrosine phosphatases.  (26) for aggregated Fc⑀RI in rat basophilic leukemia cells. In our experiments, peaks of phosphorylation occurred by 3-5 min. Importantly, even at the peak of this activity, phosphorylation of ␥-chains triggered by Fc␥RI aggregation occurred on subunits of the aggregated receptors but was absent from noncross-linked Fc␣R on the same cells. Similarly, cross-linking of Fc␣R did not cause phosphorylation of ␥-chains on Fc␥RI. This absence of nonaggregated FcR involvement indicates that tyrosine phosphorylation in monocytes is normally restricted to FcR in aggregates or clusters. As previously mentioned, this lack of bystander involvement is normal for nonaggregated Fc⑀RI ␥-chains in suboptimally Fc⑀RI-triggered basophils (15)(16)(17).
Restriction of Kinase Activity to Clusters Subverted by Vanadate-Interestingly, we found that aggregation of Fc␥RI in the presence of vanadate resulted in tyrosine phosphorylations not only of ␥-chains associated with aggregated Fc␥RI but also of ␥-chains associated with nonaggregated Fc␣R and of the cytoplasmic domain of nonaggregated Fc␥RII. Phosphorylation of nonaggregated Fc␣R began shortly after the onset of phos-2 L. C. Pfefferkorn and S. L. Swink, manuscript in preparation.  1 and 2) with vanadate present. Fc␥RI and Fc␥RII were precipitated and electrophoresed under reducing conditions. B, Fc␥RII aggregation induces phosphorylation of ␥-chains associated with Fc␥RI and Fc␣R. Cells were preincubated with Fab IV.3 (lanes 5-8) or medium (lanes 1-4) and incubated for 18 min with sheep anti-mouse (lanes 1-8) in the presence of vanadate. Cholate lysis buffer extracts were prepared and Fc␥RI were precipitated on hIgG (lanes 2 and 6) or 32.2 beads (lanes 4 and 8), or Fc␣R were precipitated (lanes 3 and 7) from 2 ϫ 10 6 cell equivalents. SDS extracts representing 10 5 cells were also separated for total cellular proteins (lanes 1 and 5). Proteins were electrophoresed under nonreducing conditions and analyzed by sequential anti-phosphotyrosine (upper panels) and anti-␥ (lower panels) Western blot.
FIG. 8. Lack of vanadate-induced co-aggregation of Fc␣R. Aggregation was assessed by measuring induced internalization, and all incubations were conducted in the presence or absence of 200 uM vanadate. Cells were incubated for 18 min with HB63 or 197 to occupy (Ϫ) or cross-link (ϩ) Fc␥RI, and then for an additional 10 min with P3 or A77. After washing, cells were stained with FITC-second antibody and analyzed by cytofluorography. Cells reacted with HB63, then P3 or with 197, then P3 were stained for surface Fc␥RI sites. Surface Fc␣R sites were obtained by subtracting data from 197 or HB63, then P3 reacted cells from that of 197 or HB63, then A77, reacted cells. Data represent the mean of FITC-second antibody binding sites/cell Ϯ the standard deviation.
phorylation of tyrosines in aggregated Fc␥RI␥, suggesting the rapid activation or association of activated kinases with nonaggregated FcR. Phosphorylation of nonaggregated FcR was efficient, comparing well with phosphorylated ␥-chains in overtly aggregated FcR. It was extensive, as demonstrated by dramatic shifts of total cellular and FcR-associated ␥-chains from the nonphosphorylated to the phosphorylated state. Collectively, the results show an activation-dependent phosphorylation of nonaggregated FcR ␥-chains and motifs under conditions that inhibit phosphatase activity.
In the presence of vanadate, there was an inhibition of protein tyrosine phosphatase activity, and also an increased or sustained tyrosine kinase activity. This was demonstrated in the observation that FcR aggregation-dependent increase in cellular phosphotyrosine occurred over a longer time than the normal peak activity would have predicted. Though normally transient FcR-triggered tyrosine phosphorylations would have declined after the first 3-5 min of stimulation, phosphorylation in the presence of vanadate was greater during this initial period and it continued to accumulate for 12-18 min. Thus, vanadate prevented dephosphorylation of ␥-chains, and it appears to have either prevented deactivation (and promoted release) of kinases in aggregates or activated kinases preassociated with nonaggregated FcR.
To further support this conclusion, evidence is presented that argues against alternative explanations. We show that ␥-chain phosphorylation did not occur as a result of (i) anti-FcR bridging of nontargeted FcR or (ii) vanadate-induced co-aggregation of nonaggregated with aggregated FcR. We also demonstrated that (iii) phospho-␥ on nonaggregated receptors was not acquired by subunit exchange in vitro. As for in vivo, (iv) ␥-chain exchange between aggregated and nonaggregated FcR would not account for the ability of bystander Fc␥RII to be phosphorylated or to induce the phosphorylation of Fc␥RI and Fc␣R ␥-chains when cross-linked. Collectively, the evidence is consistent with direct kinase activity on bystander component chains.
This conclusion is further supported by the observation that pervanadate treatment of cells resulted in the phosphorylation of ␥-chains without any FcR cross-linking. Pervanadate is a potent inhibitor of phosphotyrosine dephosphorylation (27,28). It increases tyrosine phosphorylations in a number of cell types (28 -30), including T lymphocytes (31,32). The effect of pervanadate in this study suggests that relevant kinase activation can occur independently of aggregation and, conversely, that ␥-chains need not be aggregated to be substrates. Since ␥-chains do not require the aggregated state for phosphorylation, this state is unlikely to dictate restriction of kinase activity to clusters.
Thus, it is clear from results in this report that kinase activity for ␥-chains is not obligatorily limited to clusters of FcR, although clustering is the normal mechanism for kinase activation. It is also clear that aggregation of ␥-chains is not a physical requirement for kinase interaction with their tyrosines. Pribluda et al. (18) have described clustered Fc⑀RI␥ motifs as the normal and necessary state of substrate for activated kinases to phosphorylate in trans their nearest neighbor Fc⑀RI ␥-chains. The same group has also shown Fc⑀RI dimers are a sufficient size to satisfy the requirements for transphosphorylation (33). Therefore, our observations are not inconsistent with the model of Pribluda et al., since nonaggregated Fc␥RI and Fc␣R may exist as dimers 3 prepared to transphosphorylate paired chains but needing something more for kinase activation.
Our central conclusion is that there is a vanadate-sensitive mechanism that prevents kinase activation and the tyrosine phosphorylation of the nonaggregated FcR component chains. Normal intracluster restriction of ␥-chain phosphorylation may be due to this mechanism, and phosphatases as regulatory molecules are implicated in the process.
Vanadate subverted the normal cellular mechanism, but it is not clear how that occurred. One difficulty in interpreting molecular events is that we do not know in sufficient detail how kinases that phosphorylate FcR tyrosine motifs become activated, with what proportion of receptors they are preassociated, and what other regulatory molecules are present. One interpretation of our data is that vanadate blocked the deactivation (dephosphorylation) of kinases in aggregates and caused their release. However, in view of the effect of pervanadate, it is more likely that kinases are sufficiently preassociated with, or recruited to, nonaggregated FcR and phosphorylate receptors once activated. Yamashita et al. (14) estimate that ϳ25% of resting Fc⑀RI are associated with Lyn kinase in rat basophilic leukemia cells. Wang et al. (13) identified Lyn and Hck associated with resting Fc␥RI. Kent et al. (34) found phosphatase activity in Fc⑀RI aggregates and Swieter et al. (26) found it in monomers. In our experiments, FcR-mediated oxygen radical production may have converted some vanadate to pervanadate or, alternatively, stimulation may have caused channeling of orthovandate intracellularly. Either way, it would appear that phosphatases were inhibited that were functionally associated with kinases in nonaggregated FcR. Thus, a reasonable hypothesis is that nonaggregated FcR phosphorylation is normally negatively regulated by phosphotyrosine phosphatases and that aggregation induces FcR␥ phosphorylation by transiently inactivating the phosphatases.