Ne’~ron, Vol. 6, 869-878, June, 1991, Copyright
@ 1991 by Cell Press
Agrin Induces Phosphorylation of the Nicotinic Acetylcholine Receptor Bruce G. Wallace,* Zhican Qu,+ and Richard 1. Huganirt *Department of Physiology University of Colorado Health Sciences Center Denver, Colorado 80262 +Howard Hughes Medical Institute Department of Neuroscience Johns Hopkins University School of Medicine Baltimore, Maryland 21205
Summary Agrin causes acetylcholine receptors (AChRs) on chick myotubes in culture to aggregate, forming specializations that resemble the postsynaptic apparatus at the vertebrate skeletal neuromuscular junction. Here we report that treating chick myotubes with agrin caused an increase in phosphorylation of the AChR B, y, and 6 subunits. H-7, a potent inhibitor of several protein serine kinases, blocked agrin-induced phosphorylation of they and 6 subunits, but did not prevent either agrin-induced AChR aggregation or phosphorylation of the B subunit. Experiments with anti-phosphotyrosine antibodies demonstrated that agrin caused an increase in tyrosine phosphorylation of the B subunit that began within 30 min of adding agrin to the myotube cultures, reached a plateau by 3 hr, and was blocked by treatments known to block agrin-induced AChR aggregation. Anti-phosphotyrosine antibodies labeled agrin-induced specializations as they do the postsynaptic apparatus. These results suggest that agrin-induced tyrosine phosphorylation of the B subunit may play a role in regulating AChR distribution. Introduction Agrin, a protein isolated from the electric organ of Torpedo californica, induces the formation of specializations on chick myotubes in culture at which several components of the postsynaptic apparatus are concentrated, including acetylcholine receptors (AChRs), acetylcholinesterase, butyrylcholinesterase, a heparan sulfate proteoglycan, and a 43 kd AChR-associated protein (Godfrey et al., 1984; Nitkin et al., 1987; Wallace, 1986,1989; McMahan and Wallace, 1989). Several lines of evidence suggest that the formation of the postsynaptic apparatus at developing neuromuscular junctions is triggered by the release of agrin from motor axon terminals (Nitkin et al., 1987; McMahan and Wallace, 1989). For example, the formation of agrin-induced specializations on myotubes in culture mimics events at developing neuromuscular junctions (Wallace, 1988), agrin, or a similar protein, is found in the cell bodies of motor neurons and is transported down their axons (Magill-Sole and McMahan, 1988, 1989), and antibodies to agrin block nerveinduced AChR aggregation (Reist and McMahan,
1990, Sot. Neurosci., abstract). AChRs accumulate into agrin-induced specializations through lateral migration of receptorsalreadyon the surfaceof the myotube by a process that does not require protein synthesis (Godfrey et al., 1984; Wallace, 1988). Thus, agrin-induced AChR aggregation must be mediated by posttranslational modification(s) of existing proteins. Protein phosphorylation is one of the primary mechanisms for the posttranslational regulation of protein function. Recent studies have demonstrated that the nicotinic AChR from electric organ and skeletal muscle is phosphorylated by several protein serine and tyrosine kinases (Huganir and Miles, 1989; Huganir and Greengard, 1990). The nicotinic AChR is a pentamerit complex of four types of subunits, a, 8, y, and 8, in the stoichiometry a&G (Changeux et al., 1984). When AChRs are isolated from the electric organ of T. californica each of the four receptor subunits is phosphorylated (Vandlen et al., 1979; Hopfield et al., 1988). In addition, several protein kinases that phosphorylate AChRs are found in postsynaptic membranes isolated from the electric organ: a CAMPdependent protein kinase that phosphorylates the y and 8 subunits (Huganir and Greengard, 1983), protein kinase C, which phosphorylates the a and 8 subunits (Huganir and Greengard, 1983; Safran et al., 1987), and atyrosine-specific protein kinase that phosphorylates the 8, y, and 8 subunits (Huganir et al., 1984; Hopfield et al., 1988). AChRs in skeletal muscle are also phosphorylated by similar protein kinases (Miles et al., 1987, 1989; Smith et al., 1987; Ross et al., 1987, 1988). Many of the phosphorylation sites have been identified and are located on the major intracellular loop of each subunit between the third and fourth transmembrane a-helices. These phosphorylation sites are conserved in the amino acid sequences of most AChR subunits from all species sequenced so far (Huganir and Miles, 1989). These findings suggest that phosphorylation of AChRs may modulate receptor function in several ways. Phosphorylation of AChRs has been shown to increase the rate of receptor desensitization (Eusebi et al., 1985; Huganir et al., 1986; Hopfield et al., 1988; Albuquerque et al., 1986; Middleton et al., 1986,1988; Mulle et al., 1988). Recent studies have suggested that phosphorylation of AChRs may also be involved in regulating receptor distribution. Myotubes transformed with the Rous sarcoma virus do not have spontaneously occurring AChR aggregates, nor do they respond to aggregating factors such as agrin; these effects of transformation are mediated by the SK gene product, a protein tyrosine kinase (Anthony et al., 1984). In addition, the formation of agrin-induced AChR aggregates is blocked by phorbol 12-myristate 13-acetate (TPA), an activator of protein kinase C (Wallace, 1988). TPA also causes spontaneous and agrin-
induced AChR aggregates to disperse (Wallace, 1988; Ross et al., 1988). In chick myotubes it has been shown that one of the proteins phosphorylated by protein kinase C is the AChR itself (Ross et al., 1988). Moreover, recent studies have demonstrated that tyrosine phosphorylation of AChRs in muscle is regulated by innervation, apparently through the release of a nerve-derived factor (Qu et al., 1990). To investigate whether agrin may be the factor that regulates tyrosine phosphorylation of AChRs at the neuromuscular junction and whether agrin-induced phosphorylation may be involved in receptor aggregation, we analyzed the effect of agrin on AChR phosphorylation in cultured chick myotubes. A preliminary account of some of this work has been presented (Wallace, 1991).
%leucine
3’P-phosphate
immunoblot
!?!P!m!!?? 180 kD-
116.
-
-
84. -58
-
-
-j9 -r!
48.5.
Results isolation of AChRs Agrin causes AChRs already on the surface of myotubes in culture to aggregate (Godfrey et al., 1984; Wallace, 1988). Accordingly, we designed a protocol to isolate surface AChRs selectively. Intact myotubes were incubated with biotinylated a-bungarotoxin, rinsed to remove unbound toxin, and extracted with detergent. The solubilized toxin-receptor complexes were precipitated with streptavidin-conjugated Sepharose beads. Nonspecific binding was assessed by including unconjugated a-bungarotoxin or curare during the incubation with biotinylated a-bungarotoxin. Four polypeptides were precipitated specifically by this procedure (Figure 1, left panel); these were considered to represent the four different subunits of the chick AChR. If the AChR-toxin-streptavidin beads were boiled in SDS sample buffer to remove and dissociate the toxin-receptor complexes for electrophoresis, the four AChR polypeptides had apparent molecular masses of 40,42,48, and 52 kd, as illustrated in Figure 1. On the other hand, if the toxin-receptor complexes were eluted at room temperature, the two intermediate bands ran at 45 and 50 kd, respectively; the 50 kd partially overlapped the 52 kd polypeptide (data not shown). This shift in migration, apparently due to proteolysis, is a known characteristic of they and 6 subunits of AChRs isolated from many species (Huganir and Racker, 1980; Merlie and Sebbane, 1981). To maximizetheseparation betweenthefourAChRsubunits, samples were routinely boiled prior to electrophoresis. Identification of the AChR Subunits Based on their apparent molecular masses and changes in migration with boiling, the AChR polypeptides were tentatively identified as follows: the 40 kd polypeptide as the a subunit, the 42 kd polypeptide as y, the 48 kd as 6, and the 52 kd as p. The assignments of a, P, and 6 were confirmed using subunit-specific monoclonal antibodies (Figure 1, center panel). The 42 kd polypeptide, which did not stain with any of the
36.5
26.6.
con Figure 1. Agrin Cultured Chick
BTx
cl
p
E
Induces Phosphorylation Myotubes
con
ag
of Surface
13Tx AChRs
in
Left panel: Fluorograms of AChRs isolated from myotubes incubated in leucine-deficient medium supplemented with 1 mCi/ml [3H]leucine. The isolation of polypeptides of 40,42,48, and 52 kd (con) was blocked by including 2 x 10m6 M unlabeled a-bungarotoxin (BTx) or 5 x 10e4 M curare (data not shown) during the incubation with 4 x IO-* M biotinylated a-bungarotoxin. Accordingly, these four polypeptides were identified as AChR subunits. Center panel: lmmunoblot analysis of AChR subunits. AChRs were isolated and probed with subunit-specific monoclonal antibodies raised against T. californica or eel AChRs. The 40 kd polypeptide bound anti-a antibodies (#61), the 48 kd polypeptide bound two different anti-b antibodies (#I37 and 88B), and the 52 kd polypeptide bound two different anti-b antibodies (#Ill and #124). None of the eight monoclonal antibodies assayed, including two anti-y/s antibodies, recognized the 42 kd polypeptide. a = #61, p = #124, 6 = #137. Right panel: Autoradiograms of AChRs isolated from myotube cultures incubated overnight with 0.5 mCi/ml [“P]H1P04. They, S, and to a lesser extent p AChR subunits were consistently labeled in control cultures (con); no label was ever seen associated with the a subunit. Overnight treatment with agrin caused an increase in phosphorylation of they, 6, and especially 0 subunits (ag). Specificity in the isolation of phosphopolypeptides from agrin-treated cultures was confirmed by including unlabeled a-bungarotoxin (BTx) or curare (data not shown) during the incubation with biotinylated a-bungarotoxin. The same patterns of phosphorylation were seen in unboiled samples; however, the label associated with the 6 subunit tended to obscure that assocfated with the p subunit.
monoclonal antibodies tested (including two anti-y/& subunit antibodies), will be referred to as the y subunit, although the possibility that it is a fragment of the b or 6 subunits cannot be excluded. None of a variety of additional proteinase inhibitors or modifications
of the
isolation
protocol
that
were
explored
al-
Agrin Induces Phosphorylation 871
of AChRs
Table 1. ARrin-Induced AChR Sub;nits
Increase
in Phosohorvlation ,
of
Subunit
Controla
Agrin-Treated
Differenceb
pc
B Y F
145 1071 1000
429 1366 1190
284 :t 129 295 :t 311 190 t 291
@ 1991 by Cell Press
Agrin Induces Phosphorylation of the Nicotinic Acetylcholine Receptor Bruce G. Wallace,* Zhican Qu,+ and Richard 1. Huganirt *Department of Physiology University of Colorado Health Sciences Center Denver, Colorado 80262 +Howard Hughes Medical Institute Department of Neuroscience Johns Hopkins University School of Medicine Baltimore, Maryland 21205
Summary Agrin causes acetylcholine receptors (AChRs) on chick myotubes in culture to aggregate, forming specializations that resemble the postsynaptic apparatus at the vertebrate skeletal neuromuscular junction. Here we report that treating chick myotubes with agrin caused an increase in phosphorylation of the AChR B, y, and 6 subunits. H-7, a potent inhibitor of several protein serine kinases, blocked agrin-induced phosphorylation of they and 6 subunits, but did not prevent either agrin-induced AChR aggregation or phosphorylation of the B subunit. Experiments with anti-phosphotyrosine antibodies demonstrated that agrin caused an increase in tyrosine phosphorylation of the B subunit that began within 30 min of adding agrin to the myotube cultures, reached a plateau by 3 hr, and was blocked by treatments known to block agrin-induced AChR aggregation. Anti-phosphotyrosine antibodies labeled agrin-induced specializations as they do the postsynaptic apparatus. These results suggest that agrin-induced tyrosine phosphorylation of the B subunit may play a role in regulating AChR distribution. Introduction Agrin, a protein isolated from the electric organ of Torpedo californica, induces the formation of specializations on chick myotubes in culture at which several components of the postsynaptic apparatus are concentrated, including acetylcholine receptors (AChRs), acetylcholinesterase, butyrylcholinesterase, a heparan sulfate proteoglycan, and a 43 kd AChR-associated protein (Godfrey et al., 1984; Nitkin et al., 1987; Wallace, 1986,1989; McMahan and Wallace, 1989). Several lines of evidence suggest that the formation of the postsynaptic apparatus at developing neuromuscular junctions is triggered by the release of agrin from motor axon terminals (Nitkin et al., 1987; McMahan and Wallace, 1989). For example, the formation of agrin-induced specializations on myotubes in culture mimics events at developing neuromuscular junctions (Wallace, 1988), agrin, or a similar protein, is found in the cell bodies of motor neurons and is transported down their axons (Magill-Sole and McMahan, 1988, 1989), and antibodies to agrin block nerveinduced AChR aggregation (Reist and McMahan,
1990, Sot. Neurosci., abstract). AChRs accumulate into agrin-induced specializations through lateral migration of receptorsalreadyon the surfaceof the myotube by a process that does not require protein synthesis (Godfrey et al., 1984; Wallace, 1988). Thus, agrin-induced AChR aggregation must be mediated by posttranslational modification(s) of existing proteins. Protein phosphorylation is one of the primary mechanisms for the posttranslational regulation of protein function. Recent studies have demonstrated that the nicotinic AChR from electric organ and skeletal muscle is phosphorylated by several protein serine and tyrosine kinases (Huganir and Miles, 1989; Huganir and Greengard, 1990). The nicotinic AChR is a pentamerit complex of four types of subunits, a, 8, y, and 8, in the stoichiometry a&G (Changeux et al., 1984). When AChRs are isolated from the electric organ of T. californica each of the four receptor subunits is phosphorylated (Vandlen et al., 1979; Hopfield et al., 1988). In addition, several protein kinases that phosphorylate AChRs are found in postsynaptic membranes isolated from the electric organ: a CAMPdependent protein kinase that phosphorylates the y and 8 subunits (Huganir and Greengard, 1983), protein kinase C, which phosphorylates the a and 8 subunits (Huganir and Greengard, 1983; Safran et al., 1987), and atyrosine-specific protein kinase that phosphorylates the 8, y, and 8 subunits (Huganir et al., 1984; Hopfield et al., 1988). AChRs in skeletal muscle are also phosphorylated by similar protein kinases (Miles et al., 1987, 1989; Smith et al., 1987; Ross et al., 1987, 1988). Many of the phosphorylation sites have been identified and are located on the major intracellular loop of each subunit between the third and fourth transmembrane a-helices. These phosphorylation sites are conserved in the amino acid sequences of most AChR subunits from all species sequenced so far (Huganir and Miles, 1989). These findings suggest that phosphorylation of AChRs may modulate receptor function in several ways. Phosphorylation of AChRs has been shown to increase the rate of receptor desensitization (Eusebi et al., 1985; Huganir et al., 1986; Hopfield et al., 1988; Albuquerque et al., 1986; Middleton et al., 1986,1988; Mulle et al., 1988). Recent studies have suggested that phosphorylation of AChRs may also be involved in regulating receptor distribution. Myotubes transformed with the Rous sarcoma virus do not have spontaneously occurring AChR aggregates, nor do they respond to aggregating factors such as agrin; these effects of transformation are mediated by the SK gene product, a protein tyrosine kinase (Anthony et al., 1984). In addition, the formation of agrin-induced AChR aggregates is blocked by phorbol 12-myristate 13-acetate (TPA), an activator of protein kinase C (Wallace, 1988). TPA also causes spontaneous and agrin-
induced AChR aggregates to disperse (Wallace, 1988; Ross et al., 1988). In chick myotubes it has been shown that one of the proteins phosphorylated by protein kinase C is the AChR itself (Ross et al., 1988). Moreover, recent studies have demonstrated that tyrosine phosphorylation of AChRs in muscle is regulated by innervation, apparently through the release of a nerve-derived factor (Qu et al., 1990). To investigate whether agrin may be the factor that regulates tyrosine phosphorylation of AChRs at the neuromuscular junction and whether agrin-induced phosphorylation may be involved in receptor aggregation, we analyzed the effect of agrin on AChR phosphorylation in cultured chick myotubes. A preliminary account of some of this work has been presented (Wallace, 1991).
%leucine
3’P-phosphate
immunoblot
!?!P!m!!?? 180 kD-
116.
-
-
84. -58
-
-
-j9 -r!
48.5.
Results isolation of AChRs Agrin causes AChRs already on the surface of myotubes in culture to aggregate (Godfrey et al., 1984; Wallace, 1988). Accordingly, we designed a protocol to isolate surface AChRs selectively. Intact myotubes were incubated with biotinylated a-bungarotoxin, rinsed to remove unbound toxin, and extracted with detergent. The solubilized toxin-receptor complexes were precipitated with streptavidin-conjugated Sepharose beads. Nonspecific binding was assessed by including unconjugated a-bungarotoxin or curare during the incubation with biotinylated a-bungarotoxin. Four polypeptides were precipitated specifically by this procedure (Figure 1, left panel); these were considered to represent the four different subunits of the chick AChR. If the AChR-toxin-streptavidin beads were boiled in SDS sample buffer to remove and dissociate the toxin-receptor complexes for electrophoresis, the four AChR polypeptides had apparent molecular masses of 40,42,48, and 52 kd, as illustrated in Figure 1. On the other hand, if the toxin-receptor complexes were eluted at room temperature, the two intermediate bands ran at 45 and 50 kd, respectively; the 50 kd partially overlapped the 52 kd polypeptide (data not shown). This shift in migration, apparently due to proteolysis, is a known characteristic of they and 6 subunits of AChRs isolated from many species (Huganir and Racker, 1980; Merlie and Sebbane, 1981). To maximizetheseparation betweenthefourAChRsubunits, samples were routinely boiled prior to electrophoresis. Identification of the AChR Subunits Based on their apparent molecular masses and changes in migration with boiling, the AChR polypeptides were tentatively identified as follows: the 40 kd polypeptide as the a subunit, the 42 kd polypeptide as y, the 48 kd as 6, and the 52 kd as p. The assignments of a, P, and 6 were confirmed using subunit-specific monoclonal antibodies (Figure 1, center panel). The 42 kd polypeptide, which did not stain with any of the
36.5
26.6.
con Figure 1. Agrin Cultured Chick
BTx
cl
p
E
Induces Phosphorylation Myotubes
con
ag
of Surface
13Tx AChRs
in
Left panel: Fluorograms of AChRs isolated from myotubes incubated in leucine-deficient medium supplemented with 1 mCi/ml [3H]leucine. The isolation of polypeptides of 40,42,48, and 52 kd (con) was blocked by including 2 x 10m6 M unlabeled a-bungarotoxin (BTx) or 5 x 10e4 M curare (data not shown) during the incubation with 4 x IO-* M biotinylated a-bungarotoxin. Accordingly, these four polypeptides were identified as AChR subunits. Center panel: lmmunoblot analysis of AChR subunits. AChRs were isolated and probed with subunit-specific monoclonal antibodies raised against T. californica or eel AChRs. The 40 kd polypeptide bound anti-a antibodies (#61), the 48 kd polypeptide bound two different anti-b antibodies (#I37 and 88B), and the 52 kd polypeptide bound two different anti-b antibodies (#Ill and #124). None of the eight monoclonal antibodies assayed, including two anti-y/s antibodies, recognized the 42 kd polypeptide. a = #61, p = #124, 6 = #137. Right panel: Autoradiograms of AChRs isolated from myotube cultures incubated overnight with 0.5 mCi/ml [“P]H1P04. They, S, and to a lesser extent p AChR subunits were consistently labeled in control cultures (con); no label was ever seen associated with the a subunit. Overnight treatment with agrin caused an increase in phosphorylation of they, 6, and especially 0 subunits (ag). Specificity in the isolation of phosphopolypeptides from agrin-treated cultures was confirmed by including unlabeled a-bungarotoxin (BTx) or curare (data not shown) during the incubation with biotinylated a-bungarotoxin. The same patterns of phosphorylation were seen in unboiled samples; however, the label associated with the 6 subunit tended to obscure that assocfated with the p subunit.
monoclonal antibodies tested (including two anti-y/& subunit antibodies), will be referred to as the y subunit, although the possibility that it is a fragment of the b or 6 subunits cannot be excluded. None of a variety of additional proteinase inhibitors or modifications
of the
isolation
protocol
that
were
explored
al-
Agrin Induces Phosphorylation 871
of AChRs
Table 1. ARrin-Induced AChR Sub;nits
Increase
in Phosohorvlation ,
of
Subunit
Controla
Agrin-Treated
Differenceb
pc
B Y F
145 1071 1000
429 1366 1190
284 :t 129 295 :t 311 190 t 291
