GABAA receptors in mouse cortical homogenates are phosphorylated by endogenous protein kinase A

MOLECULAR BRAIN RESEARCH Molecular Brain Research 24 (1994) 55-64

ELSEVIER

Research Report

GABA A receptors in mouse cortical homogenates are phosphorylated by endogenous protein kinase A Mohammad H. Jalilian Tehrani, Eugene M. Barnes Jr. * Verna and Marrs McLean Department of Biochemistry, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA

(Accepted 16 December 1993)

Abstract

Biochemical, molecular, and electrophysiological studies suggest that phosphorylation of fl subunits of the G A B A A receptor (GaR) by exogenous protein kinase A inactivates the receptor channels. We have developed a method which for the first time allows the study of GaR phosphorylation in brain tissues by endogenous PKA. Desalted homogenates or crude synaptic membranes from mouse cerebral cortex were incubated with [y-32p]ATP and 8-Br-cAMP or chlorophenylthio-cAMP. Extracts from these incubations were immunoprecipitated by polyclonal antibodies against native GaR and analyzed by SDS-gel electrophoresis and autoradiography. In both homogenates and membranes, cAMP-dependent incorporation of 32p was observed for a 57-kDa peptide, and to a lesser extent 51- to 53-kDa peptides. Phosphorylation of affinity-purified GaR by the catalytic subunit of PKA also produced a major 57-kDa phosphopeptide and a minor 51-kDa phosphopeptide. Limited digestion by S. aureus V-8 protease of the 57-kDa phosphopeptide from the desalted homogenates or from purified receptors produced a major 32p-labeled fragment of 11 kDa, suggesting that the phosphorylation site is similar to that shown previously to reduce GaR function. The phosphorylation of GaRs in homogenates was time dependent and blocked by H-89 or protein kinase inhibitor 5-24, specific inhibitors of protein kinase A. Prolonged incubations resulted in dephosphorylation of the 57-kDa phosphoprotein by a microcystin-LR sensitive phosphatase. In cortical homogenates the level of cAMP-dependent phosphorylation of the 57-kDa GaR peptide was more than 5 times that obtained with washed synaptic membranes. However, assays o f PKA using the heptamer kemptide as substrate showed that the specific activity in the particulate fraction was 57% that of the homogenate. This suggests that GaRs on synaptic membranes are preferentially phosphorylated by a cytoplasmic form of protein kinase A. By comparing the [3H]flunitrazepam-photolabeled 53-kDa GaR subunit with the 51-57 kDa [32p]peptides from cortical homogenates, the molar ratio of [32p]/[3H] was estimated at 0.43, suggesting that a substantial fraction of the GaR pool is phosphorylated under these conditions. Key words: GABA A receptor; Phosphorylation; Protein kinase A

1. Introduction

The bicuculline-sensitive y-aminobutyric acid receptor, commonly referred to as the G A B A A receptor (GaR), is the major transducer of fast inhibitory neurotransmission in the vertebrate central nervous system. G a R s represent a target for common psychoactive drugs including benzodiazepines, barbiturates, and alcohol [11]. G a R s are hetero-oligomeric proteins which belong to the super family of ligand-gated ion channels [32]. Molecular cloning has revealed diverse subunits for G a R s which, on the basis of sequence homology, are

* Corresponding author. Fax: (1) (713) 798-7854. 0169-328X/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0169-328X(93)E)232-0

grouped into four major classes ( a , / 3 , 3', and 3). T h r e e of these subunit classes have families of isoforms; so far, 6 distinct a, 4 /3, and 3 3' variants have b e e n identified (see Burt and Kamatchi [7] for a review). Protein phosphorylation is thought to play an important role in regulating the physiological function of ligand-gated ion channels. T h e activity of nicotinic acetylcholine receptors is regulated by several protein kinases [18,19]. Kainate-evoked currents through glutam a t e receptors are increased by application of protein kinase A [14,42]. Initial studies of G a R s suggested that activation of endogenous c A M P - d e p e n d e n t protein kinase (protein kinase A, PKA) reduces G A B A - g a t e d chloride flux in brain m e m b r a n e vesicles and intact cortical neurons [16,36]. Subsequently, it was shown

56

M.H. Jalilian Tehrani, E.M. Barnes Jr./Molecular Brain Research 24 (1994) 55-64

that application of exogenous PKA to several cellular and subcellular preparations produces a decline in GaR function [24,27,31]. GaR currents are also reduced by activation of protein kinase C [20]. On the other hand, an unidentified protein kinase appears to be involved in maintaining GaR in an active state [9,15]. Potential phosphorylation sites on GaR subunits were first recognized by means of consensus motifs. All four /3 subunit isoforms have a single motif for PKA phosphorylation, while ~/2L, a5, and a6 subunits have consensus sites for protein kinase C [29,32,35]. Direct phosphorylation of peptides from purified native or recombinant GaRs has been demonstrated using exogenous protein kinases [4,5,21,26]. Although the fll subunit of an a l f l l y 2 recombinant GaR was phosphorylated by native PKA in intact transfected kidney cells [27], it is not yet clear if similar biochemical events occur in excitable tissues. In order to approach this question, we have developed a method which permits the study of GaR phosphorylation in brain homogenates by endogenous protein kinase A. By application of this method, we show that GaR peptides on crude synaptic membranes, including a predominant substrate tentatively identified as a /3 subunit, are preferentially phosphorylated by a cytoplasmic form of PKA. Peptide mapping of the phosphorylation site indicates similarity to those shown previously to regulate GaR function. Some of these findings have appeared in preliminary reports [37,38].

2.3. Preparation of mouse cortical fractions Adult C57BL/6J mice were sacrificed by cervical dislocation. The cerebral cortex was rapidly dissected out and homogenized in 20 volumes of ice-cold 10 mM Hepes buffer (pH 7.4) containing 2 mM EDTA, 1 mM dithiothreitol (DDT), and protease inhibitors (2 mM benzamidine, 0.1 m g / m l bacitracin, 0.1 m g / m l soybean trypsin inhibitor, 0.3 m g / m l phenylmethylsulfonyl fluoride, and 1 mM N-ethylmaleimide). Fractions were isolated by centrifugation as shown in Fig. 1. Where indicated, endogenous ATP and cAMP were removed by desalting on spun columns [28] of Sephadex G-50 which had been equilibrated with Hepes buffer. Protein was determined by the method of Markwell et al [25]. ATP was determined by luciferase assay [23]. 2.4. Phosphorylation assays The general procedure of Walaas et al [44]. was used for endogenous phosphorylation experiments. Aliquots of cortical bomogenates eluted from G-50 spun columns or washed membranes were added to reaction mixtures (final volume 200 tzl) containing 25 mM Hepes (pH 7.4), 10 mM MgCI2, 1 mM EGTA, 1 mM EDTA, 1 mM DDT and 10 IzM [y-32p]ATP (5-10 Ci/mmol). Where indicated, cAMP analogs or inhibitors were included. The mixtures were incubated for 1 rain (or as specified) at 30°C. The reaction was terminated by addition of an equal volume of ice-cold pyrophosphate (PPi) buffer (10 mM Na PPi, 10 mM Na Pi, 100 mM NaF, 10 mM EDTA, 10 mM EGTA, pH 8.0). This was followed by centrifugation at 48,000x g for 30 min. Phosphorylation of affinity purified GaR from mouse brain was carried out using the catalytic subunit of protein kinase A. The reaction mixtures (final volume 100/xl) contained 20 mM Tris Cl (pH 7.5), 20 mM MgCll, 1 mM EDTA, 1 mM EGTA, 10 mM /3mercaptoethanol, 2 tzM [y-32p]ATP (17 Ci/mmol), and 1 /.~g GaR. The reaction was initiated by addition of PKA catalytic subunit (10 activity units or as indicated), carried out for 15 min at 30°C, and

2. Materials and methods

Cortex Homogenized in Buffer 10 strokes, 3000 rpm

2. l. Materials [3H]flunitrazepam (74.1-79.4 Ci/mmol)was obtained from New England Nuclear and [y-32p]ATP (4500 Ci/mmol) from ICN Biochemicals. ATP, 8-Br-cAMP, and 8-(4-chlorophenylthio)-cAMP were purchased from Boehringer Mannheim. Staphylococcus aureus V-8 protease was from Pierce, kemptide from Sigma, protein kinase inhibitor 5-24 (PKI 5-24) from the Howard Hughes protein synthesis core facility at Columbia University, N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinolinesulfonamide (H-89) from Biomol Research, microcystin-LR from Gibco BRL, pansorbin (S. aureus cells) from Calbiochem, and clonazepam and RO7-1986 from HoffmannLaRoche (Nutley, NJ). Z2. GaR isolation and production of polyclonal antibodies

Sephodex G 50

Homogenate 2 m[n, 750 rpm

Eluted-Homogenate

Super.

Pellet

(E-S) GABA A receptors were purified according to Sigel et al. [33] using benzodiazepine (Ro7-1986) affinity chromatography. From 100 g (wet weight) of chicken brain or 50 g of mouse brain approximately 300 tzg and 120 ~g receptor protein was obtained, respectively. Chicken GaR preparations (75/.~g/100/~l) were emulsified with an equal volume of Freund's complete adjuvant and injected subcutaneously into New Zealand white rabbits. Booster injections (40-75 l~g protein) were provided at biweekly intervals. Development of antibody titer for GaR was determined by solid-phase assays.

I 4s,0o0 g

I Super.

30 rain

(H)

48,000 g 30 rain

Pellet

Super.

Supernatant

Pellet (M)

Super. ($)

Pellet (E-M)

Fig. 1. Scheme for fractionation of mouse cortical tissue. Mouse cerebral cortex was homogenized and fractionated by centrifugation under the conditions shown. Where indicated, the fractions were desalted by elution from spun columns [28] of Sephadex G-50.

M.H. Jalilian Tehrani, E.M. Barnes Jr./Molecular Brain Research 24 (1994) 55-64 then terminated by addition of 400 ~1 PPi buffer containing 1% Triton X-100 and 0.1% SDS. The protein kinase A activity in individual fractions (Fig. 1) was determined according to the method of Klann et al. [22] with some modifications. Aliquots (0.2-2 /xg protein) were added to assay mixtures (final volume 50 /zl) containing 25 mM Hepes (pH 7.0), 5 mM MgCI2, 0.5 mM EDTA, 0.5 mM EGTA, 100 /xM [y-SZP]ATP (0.5 Ci/mmol), 250 mM NaCI, 10 mM Na pyrophosphate, 25 p.M kemptide, and 10 /zM 8-Br-cAMP. Assays were carried out for 10 rain at 37°C and then terminated by addition of 25/zl of ice-cold stop solution (3 mM ATP and 300 mM EDTA). Aliquots were applied to phosphocellulose paper which was washed 3 times for 20 rain in 75 mM H3PO4, dried, and counted. Kemptide or 8-Br-cAMP was omitted from control assays with equivalent results. All activities were corrected by subtraction of these control values. A series of dilutions from each fraction was used to establish that the phosphorylation rate was proportional to the amount of protein added.

2.5. Immunoprecipitation and SDS-PAGE The 32p-labeled membrane pellets were solublized for 30-45 min at 4°C in pyrophosphate buffer containing 1% Triton X-100, 0.1% SDS, and protease inhibitors. Pellets labeled with [3H]flunitrazepam were solubilized in 10 mM Tris C1 (pH 7.5), 150 mM NaCI, 2 mM EGTA, 2 mM EDTA, 1% Triton X-100, 0.1% SDS, and protease inhibitors. The mixtures were clarified by centrifugation at 32,000 g for 90 min; centrifugation at 100,000 g for 60 rain gave similar results. The clear extracts were incubated with polyclonal antibody RA4 (i #1 or as indicated) overnight at 4°C with continuous agitation. Staph A cells (20-30 #1 of a 10% suspension) were added, the mixtures were incubated for 30-45 min at 4°C, and then centrifuged at 10,000 g for 1 min. The pellets were washed four times with buffer by resuspension and centrifugation. The final pellets were resuspended in 30/xl of SDS PAGE sample buffer, incubated at 100°C for 3 rain, and centrifuged. Aliquots of the supernatants and molecular weight standards were applied to 10% polyacrylamide-SDS gels which were prepared, electrophoresed, and stained with Coomassie blue as described by Garfin [12]. After drying the gels, 32p-proteins were detected by autoradiography using Kodak XAR-5 film. Gels with 3H-labeled proteins were treated with EN3HANCE Autoradiography Enhancer (New England Nuclear) for 1 hr, dried, and subjected to fluorography. Densitometric analysis of the autoradiograms was performed using an AMBIS optical imaging system.

2.6. One-dimensional phosphopeptide mapping Limited V-8 protease digestion and electrophoresis was carried out according to Cleveland et al. [10]. Phosphopeptide bands identified by autoradiography were excised from 10% gels and rehydrated for 30 min in 125 mM Tris-Cl (pH 6.8), 0.1% SDS, and 1 mM EDTA. The swollen gel pieces were loaded on 15% polyacrylamide gels with 5 p.g S. aureus V-8 protease per lane. The gels were run as described and analyzed by autoradiography.

2. 7. Photoaffinity labeling with [SH]flunitrazepam GaRs were photoaffinity labeled with [3H]flunitrazepam according to the method of Vitorica et al. [41] with some modifications. Mouse cortical membranes (1 mg/ml) were incubated for 30 rain at 4°C with 2 nM [3H]flunitrazepam in 50 mM Tris-Cl (pH 7.4), 1 mM EDTA, 1 mM EGTA, 0.1 m g / m l bacitracin, 2 mM benzamidine, and 0.l m g / m l trypsin inhibitor. Control experiments to define nonspecific labeling also contained 1 # M clonazepam. The mixtures were irradiated with short wave UV light for 10 min and then centrifuged at 48,000 X g for 10 rain.

57

3. Results

3. I. Quantitative immunoprecipitation of GaR Polyclonal antibody RA4, raised against affinitypurified GaR from chicken brain, was capable of immunoprecipitating mouse brain GaR as shown in Fig. 2. Triton X-100 extracts from crude synaptic membranes were prepared, immunoprecipitated, and assayed for specific binding of [3H]muscimol and [3H]flunitrazepam as decribed by Vitorica et al. [40]. 120 ,-.

100

[SH]Muscim°l

0 .~

-.,o~

80

0.-

"5 60 D,. 0

:~

E E

/

40

/ ,~

• v

Chicken/PA Chicken/RA4

Mouse/PA [] Mouse/RA4

//

•

20

- -

0

0.01

0.1

1

AnHsera

o

100

."2--

80

L. o

60

=

40

(/~I)

/)

120 v c

10

3 [ H ]~F l u n i f r a z e p a r n v ~ - - - - - v

•

/?

E

//

v Mouse/PA D Mouse/RA4

/, ,v/ / /

20 0.Ol

0.1 Anfisero

Chieken/PA

V Chicken/RA4

1

10

(/.1,1)

Fig. 2. Immunoprecipitation of chicken and mouse brain OaR. Triton X-100 extracts from chicken and mouse brain were prepared as described under Methods and incubated overnight at 4°C with the amount shown of preimmune serum (PA) or immune serum (RA4). Staph A cells were added and the mixtures were incubated for 30 rain at 4°C. After centrifugation and washing, the pellets and original supernatants were assayed [40] for ligand binding with 40 nM [3H]muscimol (upper panel) or 5 nM [3H]flunitrazepam (lower panel) for 30 min at 4°C. The results shown are for a typical assay carried out in duplicate. Values for ligand binding in the pellets are expressed as a percentage of that present in the unprecipitated brain extract.

58

M.H. Jalilian Tehrani, E.M. Barnes Jr./Molecular Brain Research 24 (1994) 55-64

O f the solubilized sites for ligand binding, nearly 100% for [3H]muscimol and 80% for [3H]flunitrazepam binding were recovered in the immunoprecipitates from mouse brain. Radioligand binding was barely detectable in control precipitates using preimmune serum. Similar antibody titers were observed with chicken brain G a R (Fig. 2), a result consistent with a sequence identity of more than 95% between chicken and rodent G a R subunit peptides [1]. Immunoblots of affinity purified mouse brain G a R with R A 4 serum (not shown) demonstrated cross-reactivity with two peptide bands, one of 40-45 kDa and the other 50-54 kDa, similar to those observed with other polyclonal G a R antisera [40]. 3.2. Endogenous phosphorylation o f GaRs

H o m o g e n a t e s of mouse cerebral cortex were desalted on spun columns and incubated with [y-32p]ATP in the presence of c A M P analogs. After extraction and immunoprecipitation, autoradiographic analysis of precipitates on SDS gels showed cAMP-dependent incorporation of 32p into a 57-kDa peptide and to a lesser extent, 51- to 53-kDa peptides (Fig. 3A). These phosphopeptides were not observed in precipitates with p r e i m m u n e serum (Fig. 3B). When the level of specific peptide phosphorylation in homogenates was com-

-

Br Homo.

CPT

-

Br CPT Memb.

pared with that in an equivalent amount of protein from crude synaptic membranes (Fig. 3A), labeling was consistently found to be much higher in the former. Densitometric analysis of the autoradiograms (e.g., Fig. 3A) shows that the labeling in homogenates was 4-5 times that in membranes. However, prior to extraction, binding sites for G a R ligands as well as the phosphopeptides were completely recovered in m e m b r a n e pellets and essentially non-detectable in the soluble fraction derived from the homogenates. Since the membrane fraction represents only 55% of the protein in the homogenate, the assays of phosphorylation in homogenates actually contain only about half of the G a R present in the m e m b r a n e assays. Thus the actual amount of G a R phosphorylation in homogenates is likely to be greater than 5-fold that in the m e m b r a n e fraction. Luciferase assays showed that residual A T P in the desalted homogenates or washed m e m b r a n e s corresponded to less than 1% of the A T P added to the phosphorylation assays. In order to compare the results of the endogenous phosphorylation experiments with a more defined system, we incubated G a R which had been affinity-purified from mouse brain with the catalytic subunit of P K A and [y-32p]ATP. These immunoprecipitates also showed major and minor phosphopeptides of 57 and 51 kDa, respectively (Fig. 4). The phosphorylation sites on

-

-I-

i

P

i

-I-

-

I

Fig. 3. Endogenous phosphorylation of GaRs in cortical homogenates and crude synaptic membranes. A (left): aliquots (400/xg protein) from eluted homogenates (Homo.) or washed membranes (Memb.) were incubated as described under Methods with 10 #M [y-32pIATP for 1 min at 30°C. The assays also contained, as indicated, 10/zM 8-Br-cAMP (Br) or chlorophenylthio-cAMP(CPT). cAMP was omitted from controls (-). The assays were terminated by dilution and the mixtures were centrifuged. Extracts from the pellets were immunoprecipitated with polyclonal antibody RA4, run on SDS-PAGE, and analyzed by autoradiography. B (right): phosphorylation reactions containing eluted homogenates were incubated in the absence (±) or presence (+) of 10 /zM 8-Br-cAMP as in A. Preimmune (P) or immune serum RA4 (I) was used for immunoprecipitation. The apparent M~ of the major phosphoproteins is shown in kDa.

59

M.H. Jalilian Tehrani, E.M. Barnes Jr./Molecular Brain Research 24 (1994) 55-64

the 57-kDa peptides were examined further by excision of these bands from the gel and limited digestion with S. aureus V-8 protease. Resolution of these digests on polyacrylamide gels (Fig. 5) revealed a major 32p_ labeled l l - k D a fragment derived both from the endogenous phosphorylation experiments and those with the purified components. An l l - k D a fragment (not shown) was also generated by V-8 protease treatment of the minor 51- to 53-kDa phosphopeptides from cortical homogenates. Since the amount of 32p incorporation into the 51- to 53-kDa peptides was much more variable than that for the 57-kDa subunit, it appears likely that the former may be derived by degradation of the latter. The time dependence of the endogenous cAMP-dependent G a R phosphorylation in cortical homogenates is illustrated in Fig. 6. Autoradiographs of the immunoprecipitates (Fig. 6A) show that the 32p labeling of the the 57-kDa and 51- to 53-kDa polypeptides increased with incubation times up to 1 min and then declined markedly by 5 min. Densitometric analysis of the 51- to 57-kDa region of autoradiographs from additional experiments (Fig. 6B) reveals that incorporation of label

11

8

A

B

Fig. 5. One-dimensional peptide maps of the major 57-kDa phosphopeptide from cortical homogenates and from purified components. The 57-kDa phosphopeptide was produced by phosphorylation in cortical homogenates (A) as in Fig. 3, or from incubations of purified GaR and PKA (B) as in Fig. 4. Phosphopeptide bands were excised from the gel, rehydrated, loaded on a 15% SDS gel, and then digested with 5 /zg/lane of S. aureus V-8 protease as described by Cleveland et al. [10]. The approximate Mr of the phosphopeptides is shown in kDa.

0.1 0.5 1

2

5

10

PKA (units) Fig. 4. Phosphorylation of affinity-purifiedGaR by the PKA catalytic subunit. GaR (0.85 ~zg)purified from mouse brain was incubated for 15 rain at 30°C with 2/~M [7-32p]ATP and PKA catalytic subunit in the amount shown (activity units). The reaction was terminated by dilution and the products were analyzed by immunoprecipitation as described in the legend of Fig. 3.

reached a m a x i m u m at 1 min and then declined progressively. The rate of this decline was approximately 20% per min. The decrease in 32p incorporation in prolonged incubations could be prevented by addition of 300 nM microcystin-LR, a phosphatase inhibitor (not shown). The presence of microcystin-LR had no significant effect on the G a R phosphorylation observed in the 1-min incubations with homogenates or m e m branes. T a k e n together, the data indicate that phosphatase activity does not compromise the assay u n d e r standard conditions (e.g., those in Fig. 3). Consistent with the c A M P - d e p e n d e n c e (Fig. 3), the endogenous phosphorylation of G a R s in homogenates was blocked by 10 /zM P K I 5 - 2 4 or 10 /~M H-89, both specific inhibitors of protein kinase A (Fig. 7). 3.3. Comparison o f 32p incorporation with [3H]flunitrazepam photolabeling

I n order to provide an independent m e t h o d for quantifying G a R polypeptides on SDS gels, we photoaffinity labeled mouse cortical m e m b r a n e s with [3H]flunitrazepam. This procedure has b e e n previously

M.H. Jalilian Tehrani, E.M. Barnes Jr./Molecular Brain Research 24 (1994) 55-64

60

120

lOO

g c

(

q

60

40

2°V

10 20 30 60 3O0

5

Ov 0

Time (sec)

I

I

I

I

I

I

60

120

180

240

500

560

Time

(see)

Fig. 6. Time-dependence of GaR phosphorylation in cortical homogenates. Reactions were carried out with 8-Br-AMP as described in the legend of Fig. 3 for the times indicated. GaR was immunoprecipitated and analyzed as described under Materials and methods. A (left): autoradiographs of immunoprecipitates. B (right): relative autoradiographic intensity of the 51- to 57-kDa region expressed as a function of incubation time. The autoradiographs were quantified by densitometry, and the film densities obtained from each time point were expressed relative to that from the 60 sec incubation. The data represent the mean + S.E. from three homogenate preparations.

N

51

-

+

C

-

+

W

--

+

H

Fig. 7. Effects of protein kinase inhibitors on endogenous phosphorylation of GaR in cortical homogenates. Eluted homogenates were incubated for 3 min at 4°C without added inhibitor (C), or with 10 ~tM PKI 5-24 (W), or 10/~M H-89 (H). GaR phosphorylation assays were carried out as described in the legend of Fig. 3 in the absence ( - ) or presence ( + ) of 8-Br-cAMP.

shown to label G a R a subunits in o t h e r preparations [6,34]. After extraction of the m e m b r a n e s and precipitation with antiserum RA4, fluorograms of SDS gels showed 3H-labeling of 53- and 47-kDa polypeptides (Fig. 8). These peptides were not d e t e c t e d in labeling mixtures which contained 1 /xM c l o n a z e p a m to block specific binding of [3H]flunitrazepam or in p r e i m m u n e controls without c l o n a z e p a m (not shown). The a m o u n t of [3H]flunitrazepam which b e c a m e irreversibly b o u n d during the p h o t o r e a c t i o n was d e t e r m i n e d by subsequent addition of clonazepam. This showed that 34% of the specifically-bound 3H-label was irreversibly coupled to the membranes. Excised regions of gels containing the 53-kDa [3H]polypeptide (e.g., Fig. 8) and the 51- to 57-kDa [32p]polypeptides from h o m o g e n a t e s (e.g., Fig. 3A) were c o u n t e d by liquid scintillation. A f t e r correction for the efficiency of [3H]photolabeling and normalization for m e m b r a n e protein, the molar ratio of [32p]/[3H] was f o u n d to be 0.43. This indicates that a substantial fraction of native G a R receptors in these preparations is subject to phosphorylation by e n d o g e n o u s PKA. 3.4. Protein kinase A activity in cortical fractions

Because of the striking differences in G a R p h o s p h o rylation in h o m o g e n a t e s and m e m b r a n e s (Fig. 3A), it was important to determine the relative distribution of

M.H. Jalilian Tehrani, E.M. Barnes Jr./Molecular Brain Research 24 (1994) 55-64

the soluble and particulate forms of PKA. For the cortical fractions described in Fig. 1, the PKA specific activity was assayed using the heptamer kemptide substrate (Fig. 9). The specific activities in homogenates, homogenates eluted from spun columns, supernatants, and eluted supernatants were similar. The PKA specific activity in the crude membrane fraction represented 57% of that found in the eluted homogenate. It is these two fractions which are compared for GaR phosphorylation capacity in Fig. 3A. The PKA specific activity of eluted membranes (Fig. 9) was actually lower than that for the uneluted membranes, rendering unlikely the possibility that the elution procedure removed an endogenous inhibitor of PKA from this fraction.

61

51:) °_

E o

4o

13)

E

3o

-8 E E

>.

20

T

-,i,°_ > °_

G <
10

4. Discussion H

It is now generally accepted that phosphorylation of GABA A receptors by PKA reduces GABA-gated chloride currents [16,27,31,36]. Exogenous PKA is clearly

" 53 " 47

N T Fig. 8. Photoaffinity labeling of GaR with [JH]flunitrazepam. Mouse cortical membranes were incubated with 2 nM [3H]flunitrazepam in the absence (T) and presence (N) of 1/~M clonazepam for 30 min at 4°C, and then irradiated with short wave UV light for 10 min. Membranes were isolated by centrifugation and GaR was extracted, immunoprecipitated, and analyzed by SDS-PAGE and fluorography as described under Materials and methods.

S

M

E-H

E-S

E-M

Fig. 9. Specific activity of protein kinase A in mouse cortical fractions. Aliquots (0.2-2 ~xg protein) from the fractions shown in Fig. 1 were added to assay mixtures containing 10/zM 8-Br-cAMP, 25/xM kemptide, and 100 p~M [~,-32p]ATP. These were incubated for 10 min at 37°C and and analyzed by binding to phosphocellulose paper as described under Materials and methods.

capable of incorporating phosphate into GaR peptides, as demonstrated with purified receptors [4,5,21,26], native receptors in brain membrane vesicles [24] and recombinant receptors in kidney cells [27]. However, it has not previously been demonstrated that GaRs are phosphorylated in brain tissues by endogenous PKA. As an initial step toward this goal, we have shown in this study :that the PKA present in cortical homogenates is capable of phosphorylating a native population of GaRs. This was made possible by applying the general approach of endogenous phosphorylation [43,44] to desalted cortical homogenates. We initially identified GaR phosphopeptides by specific immunoprecipitation with polyclonal antibodies raised against affinity-purified native GaR. Although monoclonal or site-directed polyclonal antibodies may provide greater specificity, the identity of the GaR subunit isoforms which are phosphorylated in cortical homogenates is not yet known. Furthermore, the existence of several members of a subunit family (for example, ot2o~3 subunit pairs), in the same GaR complex [30] further obscures selection of appropriate immunoreagents. The polyclonal antiserum RA4 used in the present experiments has the advantage of quantitatively immunoprecipitating the major ligand binding sites for GaRs in mouse and chick brain. The properties of this antibody are comparable to those of previously characterized antisera [40]. RA4 immunoprecipitation revealed cAMP-dependent phosphorylation of a 57-kDa peptide in cortical homogenates. The 57-kDa

62

M.H. Jalilian Tehrani, E.M. Barnes Jr./Molecular Brain Research 24 (1994) 55-64

phosphopeptide from homogenates was further identified by comparison with the products from phosphorylation of affinity-purified GaR by the catalytic subunit of PKA. A peptide of similar apparent M r was produced from these purified components. In addition, digestion of both 57-kDa phosphoproteins with V-8 protease generated a ll.5-kDa phosphopeptide. While this supports the notion that the 57-kDa protein is a GaR subunit, establishing which subunit is much more difficult. We have photolabeled the major a subunits in mouse cortex with [3H]flunitrazepam, revealing 47and 53-kDa peptides. By this criterion, the 57-kDa protein is not likely to be an a subunit. In addition, Browning et al. [4] found a 58-kDa phosphopeptide was generated from incubations of the PKA catalytic subunit with GaR purified from bovine brain. A combination of affinity-labeling and immunological studies led these investigators to suggest that the phosphoprotein is a/3 subunit. On digestion with V-8 protease, an ll-kDa phosphopeptide was released from the 58-kDa protein. Since our results are in agreement with those of Browning et al. [4], we tentatively identify the 57-kDa phosphoprotein from cortical homogenates as a GaR/3 subunit. Reductions in GaR currents have previously been associated with PKA-dependent phosphorylation of a 58-kDa /31 polypeptide [27]. In these experiments, we established that protein kinase A is the enzyme responsible for generation of the GaR phosphopeptides in cortical homogenates. This process was cAMP dependent and blocked by PKI 5-24 or by H-89, both specific inhibitors of PKA. We were surprised by the finding (Fig. 3A) that phosphorylation of GaRs by endogenous PKA is much more efficient in cortical homogenates than in crude synaptic membranes. Several possibilities have been considered which could account for this observation. (i) A limitation on substrate availability seems unlikely since more than 95% of the GaR pool is recoverable in the synaptic membrane fraction [39] and no GaR phosphopeptides were found in the soluble fraction. (ii) Dissociation of PKA during membrane isolation does not appear to be a serious problem. PKA assays using kemptide showed that the specific activity of the membrane fraction represented 57% that of the soluble fraction, in good agreement with published data [17]. (iii) Significant contamination of the membrane fraction by free ATP was ruled out by luciferase assays. (iv) Phosphatase activity does not appear to place a severe restraint on accumulation of GaR phosphopeptides under the conditions of Fig. 3. Accordingly, we suggest that the most likely explanation is that GaR is preferentially phosphorylated by a cytoplasmic form of PKA. Futhermore, a substantial fraction of the GaR pool is phosphorylated under these conditions. When the phosphorylation of GaR polypeptides by PKA in cortical homogenates was compared to photolabeling of an

equivalent amount of membranes with [3H]flunitrazepam, a [32p]/[3H] molar ratio of 0.43 was obtained. Both soluble and particulate forms of PKA are found in many tissues but the brain has a uniquely high fraction of the bound form [17]. The subcellular localization of PKA is determined by a family of regulatory subunit (R) isoforms, while substrate selectivity is dictated by a family of catalytic subunit variants [2]. Most RI subunits are cytoplasmic, while RII subunits are usually associated with cytoskeletal elements by RII anchoring proteins. The best characterized RII anchoring proteins, microtubule associated protein 2 (MAP-2) and A-kinase anchoring protein 150 (AKAP-150) are associated primarily with dendritic microtubules [3,13]. An interesting hypothesis which emerges from these studies is that RII anchoring proteins could facilitate neurotransmission by targeting PKA to specific subcellular regions. Although AKAP 150 and RII were found in a mouse brain fraction enriched in postsynaptic densities, [8] a role for these anchoring molecules in targeting phosphorylation of postsynaptic membrane proteins has not been established. Our studies suggest that the GaR, a major element in postsynaptic signal transduction, is preferentially phosphorylated by a soluble, rather than a particulate form of PKA. With respect to the in vitro approach used here, it should be pointed out that the isolation procedure could have perturbed the normal distribution of subcellular sites for PKA binding. Studies of the localization of PKA and GaRs and of their interactions in vivo, will be necessary to address these questions more definitively. As a prelude to GaR phosphorylation experiments in intact neurons, we believe that studies of endogenous phosphorylation in tissue homogenates have some critical advantages. In particular, this approach avoids some of the problems inherent in the metabolic labeling of cells to constant specific activity, particularly, the large isotope dilution of 32p and the accumulation of background 32p-proteins. Although current methods do not permit evaluation of the functional consequences of GaR phosphorylation in this system, these have been amply demonstrated for apparently quite similar PKA phosphorylation sites in recombinant receptors [27]. Nevertheless, the methods described here can be useful for defining roles for a variety of protein kinases and phosphatases, establishing the chemical nature of their GaR peptide substrates, as well as evaluating immunological reagents. We suggest that these will be important steps toward examining GaR phosphorylation in living cells.

Acknowlegements This work was supported by Grants DK17436, MH 47715, and NS 11535 from the National Institutes of

M.H. Jalilian Tehrani, E.M. Barnes Jr./Molecular Brain Research 24 (1994) 55-64

Health. The authors thank Patricia A. Calkin for assistance with radioimmunoassays and immunoblots.

References [1] Bateson, A.N., Lasham, A. and Darlison, M.G., 3,-Aminobutyric acid A receptor heterogenity is increased by alternative splicing of a/3 subunit, J. Neurochem., 56 (1991) 1437-1440. [2] Beebe, S.J., Salomonsky, P., Jahnsen, T. and Li, Y., The C,/ subunit is a unique isoenzyme of the cAMP-dependent protein kinase, J. Biol. Chem., 267 (1992) 25505-25512. [3] Bernhardt, R. and Matus, A., Light and electron microscope studies of the distribution of microtubule-associated protein 2 in rat brain: a difference between dendritic and axonal cytoskeletons, J. Comp. Neurol., 266 (1984) 203-221. [4] Browning, M.D., Bureau, M., Dudek, E.M. and Olsen, R.W., Protein kinase C and cAMP-dependent protein kinase phosphorylate the/3-subunit of purified ~/-aminobutyric acid A receptor, Proc. Natl. Acad. Sci. USA, 87 (1990) 1315-1318. [5] Browning, M.D., Endo, S., Smith, G.B., Dudek, E.M. and Olsen, R.W, Phosphorylation of the GABA A receptor by cAMP-dependent protein kinase and by protein kinase C: Analysis of the substrate domain, Neurochem. Res., 18 (1993) 95-100. [6] Bureau, M. and Olsen, R.W., 3,-Aminobutyric acid/benzodiazepine receptor protein carries binding sites for both ligands on both two major peptide subunits, Biochem. Biophys. Res. Commun., 153 (1988) 1006-1011. [7] Burt, D.R. and Kamatchi, G.L., GABA A receptor subtypes: From pharmacology to molecular biology, FASEB J., 5 (1991) 2916-2923. [8] Carr, D.W., Stofko-Hahn, R.E., Fraser, I.D.C., Cone, R.D. and Scott, J.D., Localization of the cAMP-dependent protein kinase to the postsynatic densities by A-kinase anchoring proteins, J. BioL Chem. 267 (1992) 16818-16823. [9] Chen, Q.X., Stelzer, A., Kay, A.R. and Wong, R.K., GABA A receptor function is regulated by phosphorylation in acutely dissociated guinea-pig hippocampal neurons, J. Physiol., 420 (1990) 207-221. [10] Cleveland, D.W., Fischer, S.G., Kirschner, M.W. and Laemmli, U.K., Peptide mapping by limited proteolysis in sodium dodecyl sulfate and analysis by gel eleetrophoresis, J. Biol. Chem., 252 (1977) 1102-1106. [11] DeLorey, T.M. and Olsen, R.W., 3,-Aminobutyric acid A receptor structure and function, J. Biol. Chem., 267 (1992) 1674716750. [12] Garfin, D.E., One-dimensional gel electrophoresis, Meth. Enzytool., 182 (1990) 425-441. [13] Glantz, S.B., Amat, J.A. and Rubin, C.S., cAMP signalling in neurons: Patterns of neuronal expression and intracellular localization for a novel protein, AKAP 150, that anchors the regulatory subunit of cAMP-dependent protein kinase I1/3, Mol. Biol. Cell 3 (1992), 1215-1228. [14] Greengard, P., Jen, J., Nairn, A.C. and Stevens, C.F., Enhancement of glutamate response by cAMP-dependent protein kinase in hippocampal neurons, Science, 253 (1991) 1135-1138. [15] Gyenes, M.F., Farrant, M. and Farb, D.H., "Run-down" of ~/-aminobutyric acid A receptor function during whole-cell recording: A possible role for phosphorylation, Mol. Pharmcol., 34 (1988) 719-723. [16] Heuschneider, G. and Schwartz, R.D., cAMP and forskolin decrease 3~-aminobutyric acid-gated chloride flux in rat brain synaptosomes, Proc, Natl. Acad. Sci. USA, 86 (1989) 2938-2942. [17] Hofmann, F., Bechtel, P.J. and Krebs, E.G., Concentrations of cyclic AMP-dependent protein kinase subunits in various tissues, J. Biol. Chem., 252 (1977) 1441-1447.

63

[18] Hopfield, J.F., Tank, D.W., Greengard, P. and Huganir, R.L., Functinal modulation of the nicotinic acetylcholine receptors by tyrosine phosphorylation, Nature, 336 (1988) 677-680. [19] Huganir, R.L., Delcour, A.H., Greengard, P. and Hess, G.P., Phosphorylation of the nicotinic acetylcholine receptor regulates its rate of desensitization, Nature, 321 (1986) 774-776. [20] Kellenberger, S., Malherbe, P. and Sigel, E., Functions of c~l, /32 and 3,2S 3,-aminobutyric acid A receptor is modulated by protein kinase C via multiple phosphorylation sites, J. Biol. Chem., 267 (1992) 25660-25663. [21] Kirkness, E.F., Bovenkerk, C.F., Ueda, T. and Turner, A.J., Phosphorylation of 3,-aminobutyrate GABA A/benzodiazepine receptors by cyclic AMP-dependent protein kinase, Biochem. J., 259 (1989) 613-616. [22] Klann, E., Chen, S.-J. and Sweatt, D., Persistent protein kinase activation in the maintenance phase of long term potentiation, J. Biol. Chem., 266 (1991) 24253-24256. [23] Leach, F.R. and Webster, J.J., Commercially available firefly luciferase reagents, Methods Enzymol., 133 (1986) 51-70. [24] Leidenheimer, N.J., Machu, T.K., Endo, S., Olsen, R.W., Harris, R.A. and Browning, M.D., Cyclic-AMP dependent protein kinase decreases 3,-aminobutyric acid A receptor-mediated 36C1 uptake by brain microsacs, J. Neurochem., 57 (1991) 722-7254. [25] Markwell, M.A.K., Haas, S.M., Bieber, L.L. and Tolbert, N.E., A modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples, Anal. Biochem., 87 (1978) 206-210. [26] Moss, S,J., Doherty, C.A. and Huganir, R.L., Identification of the cAMP-dependent protein kinase and protein kinase C phosphorylation site within the major intracellular domain of /31, 3,2S, and ~,2L subunits of the ),-aminobutyric acid type A receptor, J. Biol. Chem., 267 (1992) 14470-14476. [27] Moss, S,J., Smart, T.G., Blackstone, C.D. and Huganir, R.L., Functional modulation of GABA A receptors by cAMP-dependent phosphorylation, Science, 257 (1992) 661-665. [28] Penefsky, H.S., A centrifuged-column procedure for the measurement of ligand binding by beef heart FI, Meth. Enzymol., 56 (1979) 527-530. [29] Pritchett, D.B. and Seeburg, P.H., 3,-Aminobutyric acid A receptor a5 subunit creates novel type II benzodiazepine receptor pharmacology, J. Neurochem., 54 (1990) 1802-1804, [30] Pollard, S., Duggan, M.J. and Stephenson, F.A., Further evidence for the existence of an a subunit heterogenity within discrete ~/-aminobutyric acid A receptor subpopulations, J. Biol. Chem., 268 (1993) 3753-3757. [31] Porter, N.M., Twyman, R.E., Uhler, M,D. and Macdonald, R.L., Cyclic AMP-dependent protein kinase decreases GABA A receptor current in mouse spinal cord neurons, Neuron, 5 (1990) 789-796. [32] Schofield, P.R., Darlison, M.G., Fujita, N., Burt, D.R., Stephenson, F.A., Rodreguez, H., Rhee, L.M., Ramachandron, J, Reale, V., Glencorse, T.A., Seeburg, P.H. and Barnard, E.A., Sequence and functional expression of the GABA A 3,-aminobutyric acid receptor shows a ligand-gated receptor super-family, Nature, 328 (1987) 221-223. [33] Siegel, E., Stephenson, F.A., Mamalaki, C, and Barnard, E.A., A 3,-aminobutyric acid/benzodiazepine receptor complex of bovine cerebral cortex, J. Biol. Chem., 258 (1983) 6965-6971. [34] Stephenson, F.A., Duggan, M.J. and Casalotti, S.O., Identification of a3 subunit in the GABA A receptor purified from bovine brain, FEBS Lett., 243 (1989) 358-362. [35] Swope, S.L., Moss, S.J., Blackstone, C.D. and Huganir, R.L., Phosphorylation of ligand -gated ion channels: a possible mode of synaptic plasticity, FASEB J., 6 (1992) 2514-2523. [36] Tehrani, M.H.J., Hablitz, J.J. and Barnes, E.M., Jr., cAMP increases the rate of GABA A receptor desensitization in chick cortical neurons, Synapse, 4 (1989) 126-131.

64

M.H. Jalilian Tehrani, E.M. Barnes Jr./Molecular Brain Research 24 (1994) 55-64

[37] Tehrani, M.H.J. and Barnes, E.M., Jr., Endogenous phosphorylation of GABA A receptor subunits in mouse brain, Soc. Neurosci. Abstr., 18 (1992) 1161. [38] Tehrani, M.H.J. and Barnes, E.M., Jr., Cytoplasmic protein kinase A plays a major role in phosphorylation of cortical GABA A receptors, J. Neurochem., 61 (1993) $76. [39] Tehrani, M.H.J. and Barnes, E.M., Jr., Identification of GABA A/benzodiazepine receptors on clathrin-coated vesicles from rat brain, J. Neurochem., 60 (1993) 1755-1761. [40] Vitorica, J., Park, D., Chin, G. and de Bias, A.L., Monoclonal antibodies and conventional antisera to GABA A receptor/benzodiazepine receptor/C1- channel, J. Neurosci., 8 (1988)615622. [41] Vitorica, J., Park, D., Chin, G., and de Bias, A.L., Characteriza-

tion with antibodies of the 3,-aminobutyric acid A/benzodiazepine receptor complex during development of the rat brain, J. Neurochem., 54 (1990) 187-194. [42] Wang, L.-Y., Taverna, F.A., Huang, X.-P., MacDonald, J.F. and Hampson, D.R., Phosphorylation and modulation of a kinate receptor (GIuR6) by cAMP-dependent protein kinase, Science, 259 (1992) 1173-1175. [43] Walaas, S.I., Nairn, A.C. and Greengard, P., Regional distribution of calcium and cyclic adenosine 3':5'-monophosphate-regulated protein phosphorylation systems in mammalian brain, J. Neurosci., 3 (1983) 291-301. [44] Walaas, S.I., Perdahl-Wallace, E., Winblad, B. and Greengard, P., Protein phosphorylation systems in postmortem human brain, J. Mol. Nearosci., 1 (1989) 105-116.