- Email: [email protected]
A possible role for phosphorylation of the acetylcholine receptor
Medical Hypotheses 12 : 185 - 190, 1983
A POSSIBLE ROLE FOR PHOSPHORYLATION OF THE ACETYLCHOLINE RECEPTOR M.E. Carstens, J.J.F. Taljaard and A.C. Neethling. Neurochemistry Research Group, Department of Chemical Pathology, University of Stellenbosch, P.O. Box 63, Tygerberg 7505, South Africa. ABSTRACT The acetylcholine receptor (AChR) from Torpedo fuscomaculata can be phosphorylated. This reaction was fully characterised. In addition, the influence of factors with possible modulatory effects on the phosphorylation of this receptor were investigated. In order to suggest the possible role of phosphorylation in neurotransmission, the effects of this reaction on the binding of ligands to the AChR were investigated. It was found that phosphorylation enhanced the binding of the antagonist, a-bungarotoxin to the AChR. Due to experimental difficulties the effect on binding of the agonist, acetylcholine, to the receptor, could not be determined. However, a hypothesis is proposed regarding the role of phosphorylation of the AChR in neurotransmission. INTRODUCTION The electric organs of Torpedo and Electrophorus electricus provide pure nicotinic cholinergic systems. In situ phosphorylation of the AChR has been demonstrated in both species (1, 2). This reaction was fully characterised in our laboratory (3). Over the years, several workers in the field of neurochemistry have argued for a role of phosphorylation in neurotransmission. This necessitated an investigation into the site of phosphorylation as well as possible modulation of the reaction by factors known to be involved in postsynaptic events. Candidates $0' such a proposed modulatory role are the monovalent cations, Na and K (4), as well as the divalent cation, Ca*+, known to influence postsynaptic events (5). In addition, several kinase catalysed phosphorylation reactions are stimulated by cyclic AMP and cyclic GMP (6) and some neurotransmitters are also known to influence postsynaptic phosphorylation (7). Concerning the role of phosphorylation, there is evidence that the reaction causes changes in physical properties of the AChR (8) as well as its micro-environment (9). Since it is possible that phosphorylation could also cause subtle changes in the affinity of the AChR for endogenous neurotransmitter, agonists and antagonists its effect on the AChR binding of these agonists as well as antagonists was investigated in our laboratory. We hereby propose a possible role for phosphorylation of the AChR. 185
EXPERIMENTAL FINDINGS 1.
Characteristics of the phosphorylation Reaction
In the course of the investigation it was found that a large percentage of radio-activity incorporated into AChR-enr' ithed membranes was due to binding of ATP. This was confirmed using(8- C)ATP Measuring kinase activity in the presence of both (y-32P)ATP and(8-j4C)ATP enabled correction for this phenomenon. After purification of the AChR on affinity columns, SDS-gel electrophoretic studies revealed the presence of four subunits with molecular weights 37 000, 45 000, 52 000 and 65 000 daltons. One protein with a molecular weight corresponding to the 37 000 dalton subunit was shown to bear the site for phosphorylation (submitted for publication) as well as the o-bungarotoxin binding site (3). The phosphorylation reaction followed classical Michaelis-Menten kinetics. A Lineweaver-Burk plot of the data yielded a Km of 0,16 mM and an ATP concentration of 0,5 mM was employed in subsequent studies (3). GTP could not replace ATP as a phosphate donor. A rapid linear incorporation of phosphate into the membrane preparation occurred and maximal incorporation of 32P was observed after 3 min. Phosphorylation reactions were carried out in 100 mM Tris-HCl buffer at a pH of 8,6 at **oc. Mn*+ was required as cofactor, and optimal activity was found at a concentration of 60 mM. Thin layer chromatography and high voltage paper electrophoresis of the enzymatic hydrolysates of phosphorylated AChR preparations indicated serine to be the only phosphorylated amino acid. 2.
Modulation of the Phosphorylation Reaction
It has previously been shown that Na+ and Kt can affect phosphorylation of the AChR (1, 10). However, thes5 cations did not exert any effects in our hands. On the other hand Ca + (10 uM) was shown to inhibit the reaction (submitted for publication). Neu otr nsmitter release is crii Ca9'. However, modulatory tically dependent upon the presence of free effects of Ca*+ can also be postsynaptic. Ca*+ can modify the characteristics of the AChR-channel complex at the frog neuromuscular junction (11). The Ca*+ concentration in the synaptic cleft is in the mM range. As we have shown that 1 mM Ca*+ inhibited the phsophorylation reaction completely, this would argue for an intracellular location of the kinase enzyme. There the factors controlling cytosolic free Ca*+ concentrations, e.g. calmodulin (12), could act as a fine control mechanism regulating the level of phosphorylation. Another factor with a possible modulatory effect was investigated, i.e. the cyclic nucleotides. Cyclic AMP has no effect on the phosphorylation reaction, whereas cyclic GMP (20 uM), caused stimulation, supporting the second-messenger hypothesis c7). Results found with the AChR agonists, acetylcholine and carbamylcholine (1 and 10 uM) showed stimulation of phosphorylation of AChR-enriched membrane breparations. On the other hand antagonists d-tubocurarine,
186
hexamethonium and decamethonium at 1 uM concentrations caused as much as 40% inhibition of the reaction. 4 he opposing effects of agonists and antagonists lent further support for a distinct role of phosphorylation in the postsynaptic modulation of neurotransmission at the nicotinic cholinergic receptor. 3.
Phosphorylation and ligand binding to the AChR
We investigated the effects of phosphorylation on the binding of a-bungarotoxin (antagonist) and acetylcholine (endogenous neurotransmitter and agonist) to the AChR. The results (submitted for publication) showed that, before in vitro phosphorylation, AChR-enriched membrane preparations (ex Torpedo fuscomaculata) bound a-bungarotoxin with a KU of 8,4 nM and a B,,, of16 pmoles / mg protein. However, after in vitro phosphorylation, the Bmax was increased to 91 pmoles/mg protein, although the affinity was not changed (KU 11 nM). Control binding studies, utilising different concentrations of ATP, showed that the increase in the number of binding sites for the a-toxin could be correlated with the level of phosphorylation. On the other hand, binding of acetylcholine to AChR-enriched membrane preparations before in vitro phosphorylation, revealed the presence of two binding components: one with high affinity (KU 8,4 nM; B,,, 1,35 pmoles/mg protein) and one with lower affinity (KU 520 nM; B,,, 12 pmoles/mg protein). These results were in agreement with those reported by Boyd and Cohen (13), who proposed that acetylcholine binds to a single receptor site in two different conformational states. Unfortunately, due to experimental difficulties, the effect of phosphorylation on the binding of acetylcholine to in vitro phosphorylated AChRenriched membrane preparations could not be measured, although it was assumed that the effect was similar to that found with a-bungarotoxin. CONCLUSION Hypothesis regarding the role of phosphorylation of the AChR A model has been proposed by Boyd and Cohen (13), assuming the existence of a single class of AChR. In the absence of cholinergic ligands, these receptors exist in two interconvertible conformations, one binding agonist with low and the other with high affinity. The high affinity conformation is thought to be the receptor conformation stabilised by the agonist at equilibrium and is presumably involved in desensitisation (14). Another model concerning different states of the AChR, has been proposed by Heidmann and Changeux (15), who postulated that the receptor exists in three different states: a resting, an active and a desensitised state. Taking into consideration the evidence of the studies from our laboratory and the two postulated models, the following hypothesis is proposed, illustrated as follows:
187
Acetylcholine AChR
-
non-phosphorylated non-functional
AChR
k
AChR
partially phosphorylated
partially phosphorylated
functional
functional
RESTING
ACTIVE
(low affinity agonists)
(low affinity)
(high affinity antagonists)
Acetylcholine
Increased phosphorylation
-1
-1
AthR completely phosphorylated
AChR
BLOCKED
DESENSITISED (high affinity)
The primary role of phosphorylation at the postsynaptic membrane is to create a specific conformation of the receptor, required for the binding of cholinergic ligands. The limited number of binding sites for agonists and antagonists in the membrane preparation, prior to in vitro phosphorylation, can be explained by a partial in vivo phosphorylation and activation of receptors. The most important evidence for this statement stems from the observations that phosphorylation of AChR-enriched membrane preparations enhanced a--bungarotoxinbinding almost 6-fold. Although this stimulation could not be shown with acetylcholine, it can be presumed that it would also apply in this case, since it was shown that there was no difference in the number of binding sites for acetylcholine B,,, (14 pmoles/mg protein) and a-bungarotoxin (16 pmoles/mg protein). In addition, it is also proposed that phosphorylation has a second function in the process of desensitisation. The functional AChR in the resting state binds acetylcholine with low affinity and, in addition to channel activation, phosphorylation is stimulated. This acetylcholinestimulated phosphorylation could be the phosphorylation, specifically required for desensitisation, leading to a state of the receptor which binds more agonist with high affinity. This event would correspond to the time required for phosphorylation (3 min.). It is tempting to speculate further that the stimulatory effect of cyclic GMP on phosphorylation suggests the possible involvement of guanyl cyclase.
188
The functional receptor also binds antagonists, but with high affinity and in this case, channel opening does not occur. Further phosphorylation is also inhibited, resulting in the receptor being in a blocked state. The role of ATP binding to the AChR is unknown. However, it is plausible that the binding observed, could form part of an enzyme-substrate complex, and that the bound ATP is required for the kinase to act upon. If this is the case, the observation that antagonists also inhibit binding of ATP could be the underlying mechanism for the observed inhibition of phosphorylation. ACKNOWLEDGEMENTS We are indebted to the South African Medical Research Council for financial support and the Cape Provincial Administration for the use of facilities. This work was performed by one of the authors (M.E. Carstens) at the University of Stellenbosch in partial fulfilment of the Ph.D. degree. REFERENCES 1.
Teichberg VI, Changeux J-P. Evidence for protein phosphorylation and dephosphorylation in membrane fragments isolated from the electric organ of Electrophorus electricus. FEBS lett. 74: 71, 1977.
2.
Gordon AS, Diamond I. Reversible phosphorylation of the membranebound Acetylcholine receptor. J. Supramol. Struct. 14: 163, 1980.
3.
Carstens ME, Weller M, Neethling AC, Taljaard JJF. Characterisation of protein phosphorylation in acetylcholine receptor-enriched membrane preparations from Torpedo fuscomaculata. Mol. Cell. Biochem. 42: 161, 1982.
4.
Takeuchi A, Takeuchi N. On the permeability of endplate membrane during the action of transmitter. J. Physiol. 154: 52, 1960.
5.
Miyamoto E, Kuo JF, Greengard P. Cyclic nucleotide-dependent protein kinases. J. Biol. Chem. 244: 6395, 1969.
6.
Greengard P. Science 199:
7.
Dolphin AC, Greengard P. Serotonin stimulates phosphorylation of Protein I in the facial motor nucleus of rat brain. Nature 289: 76, 1981.
8.
Saitoh T, Wennogle LP, Changeux J-P. Factors regulating the susceptibility of the acetylcholine receptor protein to heat inactivation. FEBS lett. 108: 489, 1979.
9.
in vitro of membrane fragSaitoh T, Changeux J-P. Phosphorylation -ments from Torpedo marmorata electric organ. Eur. J. Biochem. 105: 51, 1980.
Phosphorylated proteins as physiological effecters. 146, 1978.
10. Gordon AS, Davis CG, Diamond I. Phosphorylation of membrane proteins at a cholinergic synapse. Proc. Natl. Acad. Sci. USA 74: 263, 1977.
189
11. Magleby KL, Weinstock MM. Nickel and calcium ions modify the characteristics of the acetylcholine receptor-channel complex at the frog neuromuscular junction. J. Physiol. 299: 203, 1980. 12. Gevers W. Calmodulin: 76: 439, 1980.
the law of increasing returns. S.A.J. Sci.
13. Boyd ND, Cohen JB. Kinetics of binding of 3H-acetylcholine and 3Hcarbamoylcholine to Torpedo postsynaptic membranes: slow conformational transitions of the cholinergic receptor. Biochemistry 19: 5344, 1980. 14. Boyd ND, Cohen JB. Kinetics of binding of 3H-acetylcholine to Torpedo postsynaptic membranes: association and dissociation rate constants by rapid mixing and ultrafiltration. Biochemistry 19: 5353, 1980. 15. Heidmann T, Changeux J-P. Structural and functional properties of the acetylcholine receptor protein in its purified and membranebound states. Ann. Rev. Biochem. 47: 317, 1978.
190
A POSSIBLE ROLE FOR PHOSPHORYLATION OF THE ACETYLCHOLINE RECEPTOR M.E. Carstens, J.J.F. Taljaard and A.C. Neethling. Neurochemistry Research Group, Department of Chemical Pathology, University of Stellenbosch, P.O. Box 63, Tygerberg 7505, South Africa. ABSTRACT The acetylcholine receptor (AChR) from Torpedo fuscomaculata can be phosphorylated. This reaction was fully characterised. In addition, the influence of factors with possible modulatory effects on the phosphorylation of this receptor were investigated. In order to suggest the possible role of phosphorylation in neurotransmission, the effects of this reaction on the binding of ligands to the AChR were investigated. It was found that phosphorylation enhanced the binding of the antagonist, a-bungarotoxin to the AChR. Due to experimental difficulties the effect on binding of the agonist, acetylcholine, to the receptor, could not be determined. However, a hypothesis is proposed regarding the role of phosphorylation of the AChR in neurotransmission. INTRODUCTION The electric organs of Torpedo and Electrophorus electricus provide pure nicotinic cholinergic systems. In situ phosphorylation of the AChR has been demonstrated in both species (1, 2). This reaction was fully characterised in our laboratory (3). Over the years, several workers in the field of neurochemistry have argued for a role of phosphorylation in neurotransmission. This necessitated an investigation into the site of phosphorylation as well as possible modulation of the reaction by factors known to be involved in postsynaptic events. Candidates $0' such a proposed modulatory role are the monovalent cations, Na and K (4), as well as the divalent cation, Ca*+, known to influence postsynaptic events (5). In addition, several kinase catalysed phosphorylation reactions are stimulated by cyclic AMP and cyclic GMP (6) and some neurotransmitters are also known to influence postsynaptic phosphorylation (7). Concerning the role of phosphorylation, there is evidence that the reaction causes changes in physical properties of the AChR (8) as well as its micro-environment (9). Since it is possible that phosphorylation could also cause subtle changes in the affinity of the AChR for endogenous neurotransmitter, agonists and antagonists its effect on the AChR binding of these agonists as well as antagonists was investigated in our laboratory. We hereby propose a possible role for phosphorylation of the AChR. 185
EXPERIMENTAL FINDINGS 1.
Characteristics of the phosphorylation Reaction
In the course of the investigation it was found that a large percentage of radio-activity incorporated into AChR-enr' ithed membranes was due to binding of ATP. This was confirmed using(8- C)ATP Measuring kinase activity in the presence of both (y-32P)ATP and(8-j4C)ATP enabled correction for this phenomenon. After purification of the AChR on affinity columns, SDS-gel electrophoretic studies revealed the presence of four subunits with molecular weights 37 000, 45 000, 52 000 and 65 000 daltons. One protein with a molecular weight corresponding to the 37 000 dalton subunit was shown to bear the site for phosphorylation (submitted for publication) as well as the o-bungarotoxin binding site (3). The phosphorylation reaction followed classical Michaelis-Menten kinetics. A Lineweaver-Burk plot of the data yielded a Km of 0,16 mM and an ATP concentration of 0,5 mM was employed in subsequent studies (3). GTP could not replace ATP as a phosphate donor. A rapid linear incorporation of phosphate into the membrane preparation occurred and maximal incorporation of 32P was observed after 3 min. Phosphorylation reactions were carried out in 100 mM Tris-HCl buffer at a pH of 8,6 at **oc. Mn*+ was required as cofactor, and optimal activity was found at a concentration of 60 mM. Thin layer chromatography and high voltage paper electrophoresis of the enzymatic hydrolysates of phosphorylated AChR preparations indicated serine to be the only phosphorylated amino acid. 2.
Modulation of the Phosphorylation Reaction
It has previously been shown that Na+ and Kt can affect phosphorylation of the AChR (1, 10). However, thes5 cations did not exert any effects in our hands. On the other hand Ca + (10 uM) was shown to inhibit the reaction (submitted for publication). Neu otr nsmitter release is crii Ca9'. However, modulatory tically dependent upon the presence of free effects of Ca*+ can also be postsynaptic. Ca*+ can modify the characteristics of the AChR-channel complex at the frog neuromuscular junction (11). The Ca*+ concentration in the synaptic cleft is in the mM range. As we have shown that 1 mM Ca*+ inhibited the phsophorylation reaction completely, this would argue for an intracellular location of the kinase enzyme. There the factors controlling cytosolic free Ca*+ concentrations, e.g. calmodulin (12), could act as a fine control mechanism regulating the level of phosphorylation. Another factor with a possible modulatory effect was investigated, i.e. the cyclic nucleotides. Cyclic AMP has no effect on the phosphorylation reaction, whereas cyclic GMP (20 uM), caused stimulation, supporting the second-messenger hypothesis c7). Results found with the AChR agonists, acetylcholine and carbamylcholine (1 and 10 uM) showed stimulation of phosphorylation of AChR-enriched membrane breparations. On the other hand antagonists d-tubocurarine,
186
hexamethonium and decamethonium at 1 uM concentrations caused as much as 40% inhibition of the reaction. 4 he opposing effects of agonists and antagonists lent further support for a distinct role of phosphorylation in the postsynaptic modulation of neurotransmission at the nicotinic cholinergic receptor. 3.
Phosphorylation and ligand binding to the AChR
We investigated the effects of phosphorylation on the binding of a-bungarotoxin (antagonist) and acetylcholine (endogenous neurotransmitter and agonist) to the AChR. The results (submitted for publication) showed that, before in vitro phosphorylation, AChR-enriched membrane preparations (ex Torpedo fuscomaculata) bound a-bungarotoxin with a KU of 8,4 nM and a B,,, of16 pmoles / mg protein. However, after in vitro phosphorylation, the Bmax was increased to 91 pmoles/mg protein, although the affinity was not changed (KU 11 nM). Control binding studies, utilising different concentrations of ATP, showed that the increase in the number of binding sites for the a-toxin could be correlated with the level of phosphorylation. On the other hand, binding of acetylcholine to AChR-enriched membrane preparations before in vitro phosphorylation, revealed the presence of two binding components: one with high affinity (KU 8,4 nM; B,,, 1,35 pmoles/mg protein) and one with lower affinity (KU 520 nM; B,,, 12 pmoles/mg protein). These results were in agreement with those reported by Boyd and Cohen (13), who proposed that acetylcholine binds to a single receptor site in two different conformational states. Unfortunately, due to experimental difficulties, the effect of phosphorylation on the binding of acetylcholine to in vitro phosphorylated AChRenriched membrane preparations could not be measured, although it was assumed that the effect was similar to that found with a-bungarotoxin. CONCLUSION Hypothesis regarding the role of phosphorylation of the AChR A model has been proposed by Boyd and Cohen (13), assuming the existence of a single class of AChR. In the absence of cholinergic ligands, these receptors exist in two interconvertible conformations, one binding agonist with low and the other with high affinity. The high affinity conformation is thought to be the receptor conformation stabilised by the agonist at equilibrium and is presumably involved in desensitisation (14). Another model concerning different states of the AChR, has been proposed by Heidmann and Changeux (15), who postulated that the receptor exists in three different states: a resting, an active and a desensitised state. Taking into consideration the evidence of the studies from our laboratory and the two postulated models, the following hypothesis is proposed, illustrated as follows:
187
Acetylcholine AChR
-
non-phosphorylated non-functional
AChR
k
AChR
partially phosphorylated
partially phosphorylated
functional
functional
RESTING
ACTIVE
(low affinity agonists)
(low affinity)
(high affinity antagonists)
Acetylcholine
Increased phosphorylation
-1
-1
AthR completely phosphorylated
AChR
BLOCKED
DESENSITISED (high affinity)
The primary role of phosphorylation at the postsynaptic membrane is to create a specific conformation of the receptor, required for the binding of cholinergic ligands. The limited number of binding sites for agonists and antagonists in the membrane preparation, prior to in vitro phosphorylation, can be explained by a partial in vivo phosphorylation and activation of receptors. The most important evidence for this statement stems from the observations that phosphorylation of AChR-enriched membrane preparations enhanced a--bungarotoxinbinding almost 6-fold. Although this stimulation could not be shown with acetylcholine, it can be presumed that it would also apply in this case, since it was shown that there was no difference in the number of binding sites for acetylcholine B,,, (14 pmoles/mg protein) and a-bungarotoxin (16 pmoles/mg protein). In addition, it is also proposed that phosphorylation has a second function in the process of desensitisation. The functional AChR in the resting state binds acetylcholine with low affinity and, in addition to channel activation, phosphorylation is stimulated. This acetylcholinestimulated phosphorylation could be the phosphorylation, specifically required for desensitisation, leading to a state of the receptor which binds more agonist with high affinity. This event would correspond to the time required for phosphorylation (3 min.). It is tempting to speculate further that the stimulatory effect of cyclic GMP on phosphorylation suggests the possible involvement of guanyl cyclase.
188
The functional receptor also binds antagonists, but with high affinity and in this case, channel opening does not occur. Further phosphorylation is also inhibited, resulting in the receptor being in a blocked state. The role of ATP binding to the AChR is unknown. However, it is plausible that the binding observed, could form part of an enzyme-substrate complex, and that the bound ATP is required for the kinase to act upon. If this is the case, the observation that antagonists also inhibit binding of ATP could be the underlying mechanism for the observed inhibition of phosphorylation. ACKNOWLEDGEMENTS We are indebted to the South African Medical Research Council for financial support and the Cape Provincial Administration for the use of facilities. This work was performed by one of the authors (M.E. Carstens) at the University of Stellenbosch in partial fulfilment of the Ph.D. degree. REFERENCES 1.
Teichberg VI, Changeux J-P. Evidence for protein phosphorylation and dephosphorylation in membrane fragments isolated from the electric organ of Electrophorus electricus. FEBS lett. 74: 71, 1977.
2.
Gordon AS, Diamond I. Reversible phosphorylation of the membranebound Acetylcholine receptor. J. Supramol. Struct. 14: 163, 1980.
3.
Carstens ME, Weller M, Neethling AC, Taljaard JJF. Characterisation of protein phosphorylation in acetylcholine receptor-enriched membrane preparations from Torpedo fuscomaculata. Mol. Cell. Biochem. 42: 161, 1982.
4.
Takeuchi A, Takeuchi N. On the permeability of endplate membrane during the action of transmitter. J. Physiol. 154: 52, 1960.
5.
Miyamoto E, Kuo JF, Greengard P. Cyclic nucleotide-dependent protein kinases. J. Biol. Chem. 244: 6395, 1969.
6.
Greengard P. Science 199:
7.
Dolphin AC, Greengard P. Serotonin stimulates phosphorylation of Protein I in the facial motor nucleus of rat brain. Nature 289: 76, 1981.
8.
Saitoh T, Wennogle LP, Changeux J-P. Factors regulating the susceptibility of the acetylcholine receptor protein to heat inactivation. FEBS lett. 108: 489, 1979.
9.
in vitro of membrane fragSaitoh T, Changeux J-P. Phosphorylation -ments from Torpedo marmorata electric organ. Eur. J. Biochem. 105: 51, 1980.
Phosphorylated proteins as physiological effecters. 146, 1978.
10. Gordon AS, Davis CG, Diamond I. Phosphorylation of membrane proteins at a cholinergic synapse. Proc. Natl. Acad. Sci. USA 74: 263, 1977.
189
11. Magleby KL, Weinstock MM. Nickel and calcium ions modify the characteristics of the acetylcholine receptor-channel complex at the frog neuromuscular junction. J. Physiol. 299: 203, 1980. 12. Gevers W. Calmodulin: 76: 439, 1980.
the law of increasing returns. S.A.J. Sci.
13. Boyd ND, Cohen JB. Kinetics of binding of 3H-acetylcholine and 3Hcarbamoylcholine to Torpedo postsynaptic membranes: slow conformational transitions of the cholinergic receptor. Biochemistry 19: 5344, 1980. 14. Boyd ND, Cohen JB. Kinetics of binding of 3H-acetylcholine to Torpedo postsynaptic membranes: association and dissociation rate constants by rapid mixing and ultrafiltration. Biochemistry 19: 5353, 1980. 15. Heidmann T, Changeux J-P. Structural and functional properties of the acetylcholine receptor protein in its purified and membranebound states. Ann. Rev. Biochem. 47: 317, 1978.
190