Address correspondence and reprint requests to Dr. Ferenc Erdődi, Department of Medical Chemistry, Medical and Health Science Center, University of Debrecen, H-4032 Debrecen, Nagyerdei krt. 98. Hungary. E-mail: firstname.lastname@example.org
Protein phosphatase-1M (PP1M, myosin phosphatase) consists of a PP1 catalytic subunit (PP1c) and the myosin phosphatase target subunit-1 (MYPT1). RhoA-activated kinase (ROK) regulates PP1M via inhibitory phosphorylation of MYPT1. Using multidisciplinary approaches, we have studied the roles of PP1M and ROK in neurotransmission. Electron microscopy demonstrated the presence of MYPT1 and ROK in both pre- and post-synaptic terminals. Tautomycetin (TMC), a PP1-specific inhibitor, decreased the depolarization-induced exocytosis from cortical synaptosomes. trans-4-[(1R)-1-aminoethyl]-N-4-pyridinylcyclohexanecarboxamide dihydrochloride, a ROK-specific inhibitor, had the opposite effect. Mass spectrometry analysis identified several MYPT1-bound synaptosomal proteins, of which interactions of synapsin-I, syntaxin-1, calcineurin-A subunit, and Ca2+/calmodulin-dependent kinase II with MYPT1 were confirmed. In intact synaptosomes, TMC increased, whereas Y27632 decreased the phosphorylation levels of MYPT1Thr696, myosin-II light chainSer19, synapsin-ISer9, and syntaxin-1Ser14, indicating that PP1M and ROK influence their phosphorylation status. Confocal microscopy indicated that MYPT1 and ROK are present in the rat ventral cochlear nucleus both pre- and post-synaptically. Analysis of the neurotransmission in an auditory glutamatergic giant synapse demonstrated that PP1M and ROK affect neurotransmission via both pre- and post-synaptic mechanisms. Our data suggest that both PP1M and ROK influence synaptic transmission, but further studies are needed to give a full account of their mechanism of action.
The Ca2+-dependent release of neurotransmitters from synaptic vesicles is a crucial event in neurotransmission (Sudhof 2004). Neurotransmitter release occurs after the Ca2+-mediated fusion of synaptic vesicles with the target membrane, involving the formation of a complex between soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins. This complex includes the 25 kDa synaptosomal-associated protein (SNAP-25), syntaxin-1, and synaptobrevin. The Ca2+ sensor synaptotagmin completes the SNARE-induced fusion by promoting fusion pore formation. Following exocytosis, vesicles are recycled and refilled with transmitters to form newly releasable pools.
Ser/Thr-specific phosphorylation of SNAREs and their interacting partners are important tools in these regulatory processes (Turner et al. 1999; Snyder et al. 2006). For example, phosphorylation of SNAP-25 by protein kinase A (PKA, EC 22.214.171.124) or protein kinase C (PKC, EC 126.96.36.199) (Nagy et al. 2002, 2004), and phosphorylation of tomosyn (a syntaxin-1-binding protein) by PKA (Baba et al. 2005) facilitated the SNARE complex formation and neurotransmitter release. In contrast, syntaxin-1 phosphorylation by RhoA-activated kinase (ROK, EC 188.8.131.52) had a negative impact (Sakisaka et al. 2004) on these events. Synapsins, the major vesicle-binding phosphoproteins, may play crucial roles in synaptic vesicle mobilization and recycling (Chi et al. 2001). Distinct protein kinases catalyze the phosphorylation of synapsin-I at multiple sites (Cesca et al. 2010), but the physiological significance of this process is not fully understood yet.
Although less is known of the roles of protein phosphatases in neurotransmission, some aspects of their significance have already been described. Protein phosphatase-1 (PP1, EC 184.108.40.206), -2A (PP2A, EC 220.127.116.11), and calcineurin (PP2B, EC 18.104.22.168) appear to be responsible for the dephosphorylation of > 90% of neuronal phosphoproteins (Mansuy and Shenolikar 2006). PP2B initiates synaptic vesicle endocytosis by dephosphorylation of a group of proteins termed dephosphins (Cousin and Robinson 2001). Moreover, PP1 and PP2A catalytic subunits (PP1c and PP2Ac, respectively), as well as PP2B exhibited certain specificities in the dephosphorylation of different phosphorylation sites on synapsin-I (Jovanovic et al. 2001). However, the dephosphorylation of synapsin-I and other pre-synaptic phosphoproteins has been largely unexplored with respect to the regulatory (targeting) subunits of the protein phosphatases involved.
Protein phosphatase-1M (PP1M or myosin phosphatase) consists of the 38 kDa PP1cδ (also termed PP1cβ) isoform, a 130/133 kDa myosin phosphatase target subunit-1 (MYPT1), and a 20 kDa subunit of unknown function (Hartshorne et al. 2004). Phosphorylation of MYPT1 by ROK at Thr696 and/or Thr853 results in the inhibition of PP1M (Feng et al. 1999; Murányi et al. 2005). We have shown previously that PP1M is present in cortical synaptosomes where it is associated with ROK (Lontay et al. 2004), suggesting that both PP1M and ROK may be involved in the regulation of synaptic transmission. In this work, we used specific inhibitors of ROK and PP1 to study their functions in synaptosome exocytosis and neurotransmitter release, sought novel targets of PP1M in synaptosomes, and identified several MYPT1-interacting neuronal proteins. We conclude that both ROK and PP1M influence neurotransmitter release via interacting with and phosphorylation/dephoshorylation of several pre-synaptic proteins (e.g., myosin-II, synapsin-I, syntaxin-1). Finally, using a thin-slice preparation, we demonstrate that ROK and PP1M act on both pre- and post-synaptic targets when regulating neurotransmission.
Materials and methods
Synaptosomes were prepared from Wistar rat cerebral cortex as described earlier (Dunkley et al. 1986; Lontay et al. 2004) and used in the experiments as detailed in Supplementary Information. The protocol for animal handling was authorized by the Animal Care and Protection Committee of the University of Debrecen, and it was in accordance with the guidelines of the European Union Council and with the Hungarian regulations. The rats were bred in the departmental animal house. For the preparation of brain slices, 10–14-day-old Wistar rats of both sexes were used (n = 23).
The synaptosomes were isolated, concentrated, and prepared for ultrathin sectioning using an established protocol (Moring et al. 1990). The pellets were fixed and immunogold (IG) labeling of the ultrathin sections applying anti-MYPT1 and anti-ROK antibodies was conducted using a previously described protocol (Brorson and Nguyen 2001). Further details of the procedure are given in Supporting information.
High-throughput fluorescent plate assay of exocytosis
Synaptic vesicle exocytosis was assessed by monitoring the release of the fluorescent dye FM 2-10 from synaptosomes using a previously described technique (Baldwin et al. 2003) that was modified to adopt it to a high-throughput assay. Synaptosomes loaded with FM 2-10 styryl dye were applied onto 96-well plates (Multifluor-1, Thermo Labsystem, Philadelphia, PA, USA) and samples (in triplicates) were pre-incubated in Krebs buffer — either alone (control) or in the presence of the kinase and phosphatase inhibitors. Each effector was applied for 5 min before inducing depolarization. In control experiments, only the buffers that were used to dissolve the effectors were applied. The experiments were conducted in the presence of 1.2 mM Ca2+ or 1.2 mM EGTA, so that the EGTA-treated samples could be used as controls for their parallel (Ca2+-containing) wells. Neurotransmitter release was induced by adding KCl solution with the automatic injector of Fluoroskan FL to achieve a final K+ concentration of 30 mM. Exocytosis from the synaptosomes was measured as the decrease of fluorescence (excitation: 488 nm; emission: 540 nm) using the kinetic measurement feature of the Ascent Software 2.6. The extent of the Ca2+-dependent exocytosis was calculated as the difference between the release measured in the presence of Ca2+ and EGTA (i.e., without added Ca2+). The relevant data were determined between 0 and 6 min after the stimulation at 30-s intervals and expressed in arbitrary units.
Protein determination, sodium dodecyl sulfate–polyacrylamide gel electrophoresis, and immunoblotting were carried out as described before (Lontay et al. 2004). Details of other procedures and the antibodies used are given in Supporting information. To assess and compare changes in the extent of phosphorylation, densitometric analysis of the blots was performed (ImageJ, NIH, Bethesda, MD, USA). The density of the phosphorylated proteins was normalized to that of the loading controls.
In vitro phosphorylation/dephosphorylation assays
The endogenous PP1 and PP2A were irreversibly inactivated by pre-incubation with 1 μM microcystin-LR (MC-LR) for 5 min in synaptosome lysates (0.33 mg/mL) containing 20 mM Tris-HCl (pH 7.4), 1 mM (2S,3S)-1,4-bis(sulfanyl)butane-2,3-diol, and 5 mM MgCl2. Phosphorylation of synapsin-ISer9 and syntaxin-1Ser14 was initiated with the addition of 0.2 mM ATP in the absence (control) or presence of 0.4 mU/mL ROK (Upstate, Lake Placid, NY, USA) and incubated for 15 min. Buffer exchange was performed using Protein Desalting Spin Columns (Pierce, Rockford, IL, USA) to remove MC-LR, ATP, and Mg2+. The ROK-phosphorylated samples were divided into three equal parts and were incubated for 10 min alone, or with adding either 5 nM purified skeletal muscle PP1c (Tóth et al. 2000) or 5 nM PP1c plus 25 nM Flag-MYPT1. Flag-MYPT1 was expressed in HEK293 cells and purified on anti-Flag-peptide-coupled Sepharose (Sigma, St. Louis, MO, USA) according to the manufacturer’s recommendations. The reactions were stopped by adding hot SDS sample buffer followed by boiling. The samples were analyzed by immunoblotting using anti-synapsin-IpSer9 and anti-syntaxin-1pSer14 antibodies.
Immunoprecipitation and GST-MYPT pull-down assay
Immunoprecipitations of synaptosome lysate (100 μg protein) were carried out using anti-MYPT11−296, anti-ROK, anti-syntaxin (Sigma), or anti-synapsin (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) antibodies covalently coupled to Protein A Sepharose using Seize X Protein A immunoprecipitation kit (Pierce) (Lontay et al. 2004). Details of immunoblotting procedures and isolation of glutathion-S-transferase (GST)-MYPT1 bound synaptosomal proteins are described in Supporting information.
Protein identification by mass spectrometry
Desalted and concentrated protein samples (20 μL) obtained from the pull-down experiments were digested with 100 ng trypsin overnight. The reaction was stopped by addition of 10% formic acid and the mixture was injected to an Agilent 1100 nano-HPLC system. The peptides were eluted from a Zorbax 300SB-C18 column (Agilent Technologies, Santa Clara, CA, USA) at 300 nL/min flow rate using a 2-h gradient (0–30% acetonitrile in 0.1% formic acid) and subjected to mass spectrometry (MS) using an ESI-4000 QTRAP mass spectrometer (Applied Biosystems MDS Sciex, Carlsbad, CA, USA). MS/MS spectra were acquired in an information-dependent acquisition mode that automatically selected and fragmented the two most intensive peaks from each MS spectrum. Protein identification was based on the MS/MS spectra using the MASCOT (http://www3.interscience.wiley.com/journal/68500773/abstract?CRETRY=1&SRETRY=0) search engine and the SwissProt/NCBI database. Proteins that were present in three repetitive experiments and gave 99% or higher confidence in the sequence analysis were selected. Those synaptosome proteins were considered as MYPT1-binding partners that were identified from the GST-MYPT1 pull-down fractions, but were not present in the GST ones.
Immunolabeling and confocal microscopy
Detailed description of the preparation and the major steps of the immunolabelling have been presented elsewhere (Pál et al. 2005; Rusznák et al. 2008) and detailed in Supporting information. Immunolabelling was performed with the combined application of monoclonal synaptophysin (1 : 2000; BioGenex, San Ramon, CA, USA) and either MYPT11−296- (1 : 100) or ROK-specific (1 : 50) antibodies. The immunolabeling was visualized using a confocal microscope (Zeiss LSM 510; Oberkochen, Germany) equipped with a 40× oil immersion objective. The optical thickness was 1 μm. For the colocalization analysis, the ‘Co-localization highlighter’ and ‘Intensity Correlation Analysis’ plug-ins of ImageJ were used.
Recording and analysis of excitatory post-synaptic currents
All electrophysiological experiments were performed in artificial cerebrospinal fluid (for its composition and other details, see Supporting information). The experiments were conducted on 200-μm-thick brain slices. Neurons were viewed using an Axioskop microscope (Zeiss) equipped with differential interference contrast optics and a 63× water immersion objective. Slices were incubated with tautomycetin (TMC; 5 μM), trans-4-[(1R)-1-aminoethyl]-N-4-pyridinylcyclohexanecarboxamide dihydrochloride (10 μM), or both drugs for 1 h. The recordings were performed within the following 2 h. All experiments were conducted at a holding potential of −60 mV. Excitatory post-synaptic currents (EPSCs) were evoked using a monopolar stimulating electrode that was connected to a BioStim STC-7a stimulator (Supertech, Pécs, Hungary). During stimulation, the pulses were delivered with a frequency of 50 Hz. ‘Minimal stimulation’ was employed in all experiments. In some cases, miniature EPSCs (mEPSCs) were recorded in the presence of 1 μM tetrodotoxin.
Normalized data were analyzed by t-tests (for two groups) or by analysis of variance (anova, for > two groups). Parametric statistical tests were used if the assumptions of such tests were met. Normality of the data was tested by Kolmogorov–Smirnov tests. All normalized variables used in statistical analyses were found to be normally distributed. Tests were conducted using spss (IBM, Armonk, NY, USA) for Windows. All data represent mean ± SEM.
Electron microscopic localization of MYPT1 and ROK in synaptosomes
Synaptosome fractions were satisfactorily pure and dense for performing immunolabeling (Fig. 1). In most cases, synaptosomes were situated in close proximity of each other (Fig. 1a, b and d) or, less frequently, at specialized appositions. In the latter case, synaptosomes were in contact with typical—symmetric or asymmetric—synaptic appositions (Fig. 1a, c and e). Anti-MYPT11−38 and anti-MYPT11−296 antibodies were equally effective for immunolabeling. The quality of the immunolabeling was enhanced when the glutaraldehyde content of the fixative was reduced to 0.1% and an appropriate antigen-retrieval technique (Brorson and Nguyen 2001) was employed. MYPT1-immunogold (MYPT1-IG) particles appeared as highly electron-dense spheres connected to electron-dense material within the synaptosomes or in their membranes (Fig. 1b). Similar to the MYPT1-IG particles, ROK-IG particles appeared as electron-dense dots in the synaptosomes (Fig. 1d and e). In single-labeled preparations, 27% of the synaptosomes (n = 304) were associated with other synaptosomes at specialized synaptic appositions. MYPT1-IG particles were localized to pre-synaptic terminals in 62% and to both sites (i.e., to pre- and post-synaptic terminals) in 2–3% of these ‘associated’ synaptosomes. In double-labeled preparations (n = 125), ROK-IG particles were found at pre-synaptic and at both terminals in 37% and 44% of the associated synaptosome pairs, respectively. In colabeled synaptosomes, MYPT1-IG and ROK-IG particles were detected together in 38% of the labeled synaptosomes. They were also found separately (Fig. 1d and e) or at neighboring positions (Fig. 1d). In most synaptosomes containing both MYPT1-IG and ROK-IG particles, the number of MYPT1-IG particles exceeded that of the ROK-IG ones, with a typical MYPT1-IG/ROK-IG ratio of 4 : 2. Possible coupling of MYPT1 and ROK was assessed on the basis of the number of MYPT1- and ROK-IG particles being closer to each other than 10 nm. This way, the MYPT1-/ROK-IG particles were coupled in 13% of the double-labeled synaptosomes (n = 125).
Effects of PP1 and ROK inhibitors on synaptosomal exocytosis
To determine the role of PP1M and ROK in the exocytosis from synaptosomes and in neurotransmitter release, TMC and Y27632 were applied as specific inhibitors of PP1 and ROK, respectively. Y27632 has been shown to be a selective ROK inhibitor when tested against a plethora of kinases (Davies et al. 2000), whereas TMC and okadaic acid (OA) selectively inhibited PP1 and PP2A, respectively, in COS-7 cells (Mitsuhashi et al. 2003). In our experiments, TMC (5 μM) and OA (50 or 100 nM) selectively inhibited synaptosomal PP1 and PP2A (Figure S1a, Supporting information). PP1 activity was ∼63% of the total phosphatase activity. Phosphatase activity was also determined in the presence of a PP1M-specific inhibitor—thiophosphorylated C-kinase-activated phosphatase inhibitor protein of 17 kDa (P-CPI-17; (Erdődi et al. 2003), Eto et al. 2004). These experiments suggested that ∼50% of the total phosphatase activity is attributable to PP1M, implying that PP1M has predominance over both other PP1 holoenzymes and PP2A in synaptosomes. The relative amounts of PP1 and PP2A in the synaptosomal lysate corresponded to an estimated concentration of PP1 and PP2A of 3.2 and 1.6 μM, respectively, in intact synaptosomes (Figure S1b). These amounts were also in accordance with the MC-LR-induced, dose-dependent inactivation pattern of phosphatases observed in synaptosomal lysates (Figure S1c).
To assess synaptosome exocytosis, a high-throughput assay was employed. The synaptosomes were loaded with FM 2-10 and the reduction of the fluorescence intensity was recorded, which was the result of the exocytosis of the dye. Application of TMC (i.e., inhibition of PP1) markedly suppressed, whereas the ROK inhibitor Y27632 profoundly increased synaptosome exocytosis (Fig. 2a). When applied simultaneously, Y27632 prevented the inhibitory effects of TMC, but this was accompanied with a substantial decrease of its own stimulatory effect (Fig. 2a). The time course of the kinetic curves characterizing the decrease of FM 2-10 fluorescence (i.e., the increase of the rate of the synaptosome exocytosis) had a rapidly declining part and a slower phase leading to a plateau, but this biphasic character was obvious in slow exocytotic processes only. The effects of other phosphatase and kinase inhibitors on synaptosomal exocytosis were also assessed by examining the relative decrease of fluorescence recorded at 30 s into the stimulation (Fig. 2b, lower part). Tautomycin (TM; 0.5–1 μM), which relates structurally to TMC and is also considered as a PP1-specific inhibitor in vivo (Favre et al. 1997), inhibited exocytosis in a concentration-dependent manner, although its inhibitory effect was less pronounced than that of the TMC. OA did not have a significant influence on this rapid phase of exocytosis at the concentrations tested (10–100 nM). However, 100 nM OA exerted a moderate (∼28%)—but statistically significant—inhibitory effect at 180 s (Fig. 2b, upper part). Y27632 (10 μM) was slightly more effective than (S)-(+)-2-methyl-1-[(4-methyl-5-isoquinolinyl)sulfonyl]homopiperazine 2HCl (10 μM H1152; another ROK inhibitor) in stimulating the exocytosis from synaptosomes (Fig. 2b). As it has been reported that Y27632 might inhibit smooth muscle PKCδin vitro (Eto et al. 2001), two established PKC inhibitors, GF109303 (Bisindolylmaleinimide I) and Gö6976 (5,6,7,13-Tetrahydro-13-methyl-5-oxo-12H-indolo[2,3-a]pyrrolo[3,4-c]carbazole-12-propane-nitrile) were also tested to assess the significance of PKC inhibition in synaptosomal exocytosis. Neither inhibitor stimulated the exocytosis or exerted appreciable effect on the rapid phase, while a ∼27% suppression of exocytosis was observed at 180 s when 1 μM Bisindolylmaleinimide I was applied (Fig. 2b, upper part).
Identification of MYPT1-binding proteins and potential substrates of PP1M and ROK in synaptosomes
We carried out pull-down assays with GST-MYPT1 and GST (as control) to separate potential MYPT1-interacting proteins from synaptosome lysate. The potential interacting proteins were subjected to mass spectrometry analysis for identification. Table 1 lists the MYPT1-interacting proteins identified by the peptide sequences obtained from the mass spectrometry analysis. These proteins included neuron-specific (synapsin-I, syntaxin, syntaxin-binding protein, synaptotagmin-1, and -2, SNAP25, or brain-specific polypeptide PEP19) or ubiquitously expressed proteins (sodium/potassium-transporting ATPase 1-3, ATP synthase, and the guanine nucleotide-binding protein G0) as well as enzymes with important roles in neuronal signal transduction pathways and in neurotransmitter release [Ca2+/calmodulin-dependent kinase II (CaMKII) and PP2B)]. Known MYPT1-interacting proteins, such as isoforms of the 14-3-3 protein (Koga and Ikebe 2008) and the microtubule-associated protein tau (Amano et al. 2003), also showed up in the bound fraction. Many of these neuronal MYPT1-binding proteins are thought to function as mediators of neurotransmitter release and synaptic vesicle trafficking (e.g., synapsin-I, syntaxin-1, and SNAP-25) and are regulated by phosphorylation/dephosphorylation. We conducted reciprocal immunoprecipitation experiments with antibodies specific for synapsin-I, syntaxin-1 as well as MYPT1 and ROK. MYPT1 reciprocally coprecipitated with ROK and synapsin-I (Fig. 3a). MYPT1 and syntaxin-1 coprecipitated when anti-MYPT1 was used for immunoprecipitation, but not when anti-syntaxin was the precipitating antibody. ROK reciprocally coprecipitated with all of these proteins. The binding of CaMKII (presumably the α isoform) and PP2B to MYPT1 was confirmed by positive identification of these proteins on western blots in the GST-MYPT1 pull-down fractions (Fig. 3b). These results confirmed some of the major interactions identified by mass spectrometry analyses of the GST-MYPT1 pull-down fractions.
Table 1. GST-MYPT-binding proteins of rat cortical synaptosomes
aThe MASCOT peptide scores ranged from 31 to 253 for the interacting proteins identified.
Protein phosphatase 1 regulatory subunit 12A (MYPT1)
The presence of PP1M and ROK in synaptosomes and their interactions with synaptosomal proteins suggest that common substrates of these two enzymes may exist among the neuronal proteins. To confirm this, PP1-specific dephosphorylation was selectively inhibited (by TMC) followed by an assessment of the phosphorylation patterns of proteins present in synaptosome lysate (Fig. 3c). PP1 inhibition led to a substantial increase of the phosphorylation of many proteins—catalyzed by endogenous kinases present in the synaptosome lysate—as judged by autoradiography of the gels containing separated proteins. Four bands at ∼75 kDa, ∼45–50 kDa, ∼37 kDa, and ∼25 kDa (bands 1–4, respectively) were selected for further analysis, as their apparent molecular masses were very similar to those of certain neuronal MYPT1-interacting proteins that were identified by mass spectrometry (e.g., synapsin-I, synaptotagmin, syntaxin-1, and SNAP-25). When proteins excised from these bands were subjected to mass spectrometry analysis, specific phosphopeptides were not unambiguously detected, but bands 1, 3, and 4 included the MYPT1-interacting synapsin-I, syntaxin-1, and SNAP-25, respectively (data not shown). Selective inhibition of PP1M (by P-CPI-17) increased the phosphorylation level of the proteins in all bands (Fig. 3d), suggesting that PP1M was involved in their dephosphorylation. The basal phosphorylation in synaptosome lysate was profoundly suppressed by Y27632, indicating that ROK is one of the major endogenous kinases that phosphorylate the proteins in bands 1–4.
ROK and PP1M mediate the phosphorylation levels of MLC20Ser19, synapsin-ISer9, and syntaxin-1Ser14 in cortical synaptosomes
We assessed how TMC and Y27632 influence the phosphorylation status of MYPT1Thr696 and the 20 kDa light chain of myosin-II (MLC20) at Ser19 in intact synaptosomes during KCl depolarization (Fig. 4a and b). The depolarization slightly decreased both phosphorylation levels compared with control (data not shown). In depolarized synaptosomes, phosphorylation of MYPT1Thr696 (yielding MYPT1pThr696) was moderately increased by TMC and profoundly decreased by Y27632 (Fig. 4a). Similarly, direct inhibition of PP1c (by TMC) resulted in a significant increase in the amount of phosphorylated MLC20Ser19 (MLC20pSer19). ROK inhibition (by Y27632) almost completely diminished the amount of MLC20pSer19 when TMC was absent (Fig. 4b), implying that ROK is responsible for the phosphorylation of MLC20Ser19. The relative decrease of the amounts of both MYPT1pThr696 and MLC20pSer19 upon combined application of Y27632 and TMC was substantially less than that induced by Y27632 alone. This observation suggests that inhibition of PP1 (by TMC) permits the activity of other kinase(s), distinct from ROK, which could also participate in MYPT1Thr696 and MLC20Ser19 phosphorylation. These unidentified kinase(s) may correspond to the integrin-linked kinase or zipper-interacting protein kinase, especially as these kinases have been shown to phosphorylate both MYPT1Thr696 and MLC20Ser19 (Hartshorne et al. 2004). Proving their actual roles in synaptosomes, however, is yet to be achieved.
As ROK was previously reported to phosphorylate syntaxin-1Ser14 (Sakisaka et al. 2004), changes in the phosphorylation status of candidate pre-synaptic substrates of PP1M and ROK (e.g., synapsin-I and syntaxin-1) were also determined (Fig. 4c and d) during KCl-evoked depolarization of synaptosomes. A search for consensus ROK site(s) in synapsin (using the Kinexus PhosphoNET database—www.phosphonet.ca) indicated Ser9 as a possible phosphorylation site for this kinase with a relatively high score. Depolarization resulted in 2- and 1.5-fold increase in the levels of phosphorylated synapsin-ISer9 (synapsin-IpSer9) and syntaxin-1Ser14 (syntaxin-1pSer14), respectively (data not shown). This elevated synapsin-IpSer9 level was further increased upon treatment with TMC (∼two-fold; Fig. 4c), whereas ROK inhibition markedly decreased the amount of synapsin-IpSer9. The pattern of changes of the amounts of syntaxin-1pSer14 and synapsin-IpSer9 was similar upon the different treatments (Fig. 5d). ROK inhibition decreased both the basal and TMC-induced synapsin-IpSer9 and syntaxin-1pSer14 amounts to similarly low levels, indicating the predominant involvement of ROK in these phosphorylation processes.
The identity of the phosphatase holoenzyme that may act on synapsin-IpSer9 and syntaxin-1pSer14 is not known. To assess the possible roles of PP1c and PP1M, we generated phosphosubstrates by phosphorylating synapsin-ISer9 (Fig. 4e) and syntaxin-1Ser14 (Fig. 4f) in synaptosome lysates. In these experiments, the endogenous PP1 and PP2A were irreversibly inactivated by covalently coupling PP1c and PP2Ac with MC-LR (Fig. 4e–f, controls). Addition of exogenous ROK to these lysates resulted in a 3–3.5-fold increase of the extent of synapsin-ISer9 and syntaxin-1Ser14 phosphorylation when compared with control (i.e., without ROK), providing further evidence that these sites are phosphorylated by ROK. Following buffer exchange to remove MC-LR, ATP, and Mg2+, purified PP1c was added to the ROK-phosphorylated samples with or without purified Flag-MYPT1. When PP1c was added alone, it dephosphorylated both synapsin-IpSer9 and syntaxin-1pSer14. The extent of dephosphorylation of both phosphoproteins increased when PP1c and Flag-MYPT1 were added together (i.e., PP1M holoenzyme was formed), suggesting that MYPT1 had a PP1c-targeting role under these conditions.
Localization and function of MYPT1 and ROK in the ventral cochlear nucleus
Brain slices accommodating the glutamatergic ‘giant synapse’ formed between the acoustic nerve (endbulbs of Held) and the cell bodies of the bushy cells were also examined. The expression and localization of PP1M (by identifying MYPT1) and ROK were studied using confocal microscopy. The possible pre- or post-synaptic localization of either protein was assessed by simultaneous synaptophysin-specific immunolabeling. As demonstrated in Fig. 5, MYPT1-specific immunoreaction was present within the cell bodies of the bushy neurons (Fig. 5a1), indicating its post-synaptic expression. Colocalization with the synaptophysin-specific labeling (Fig. 5a2) was also noted (Fig. 5a3 and a4). To assess the degree of colocalization, parameters allowing objective characterization of the extent of colocalization were determined. In the presented case, the Pearson’s coefficient was 0.121, the overlap coefficient was 0.443, while the intensity correlation quotient (ICQ) was 0.121. Similar results were obtained in four more images (taken from different locations), indicating a relatively high degree of overlap between the synaptophysin- and MYPT1-specific immunolabelings. Based on its morphology and size (∼5 μm in diameter), the circular structure marked with an asterisk in Fig. 5a1–a4 was identified as the cross-section of a capillary (Sposito and Gross 1987). As the endothelium expresses MYPT1 (e.g., Hirano et al. 1999; Birukova et al. 2004), its strong immunopositivity was regarded as an internal positive control.
The ROK-specific immunoreactivity (Fig. 5b1) was not prominent within the cell bodies of the bushy neurons, although immunopositive puncta corresponding to intracellular ROK could be identified (arrows in Fig. 5b1). Rather, the ROK-specific immunolabeling mainly appeared as a ring-like pattern around the cell bodies and demonstrated partial colocalization with synaptophysin (Fig. 5b3 and b4). In the presented case, the Pearsons’ coefficient, overlap coefficient, and ICQ were 0.264, 0.461, and 0.236, respectively. Similar values were yielded in all other cases (n = 4), indicating partial but definite colocalization.
Functional experiments were also conducted, in which minimal stimulation of the acoustic nerve fibers was performed in combination with paired-pulse protocols (ν = 50 Hz). The resulted EPSCs were recorded from the cell bodies of the bushy neurons under control conditions or after the application of TMC and/or Y27632 (Figure S2a1). To quantify the observed effects, the amplitude of the first EPSC (A1) and the paired-pulse ratio (PPR) was determined (Figure S2a2 and a3). Under control conditions, A1 and PPR were 0.87 ± 0.18 nA and 0.95 ± 0.08, respectively (n = 10). After 1-h pre-incubation with TMC, A1 was reduced to 0.52 ± 0.22 nA, whereas PPR was increased to 1.47 ± 0.32 (n = 6). In contrast, when Y27632 was applied, A1 and PPR were 1.04 ± 0.27 nA and 0.91 ± 0.03, respectively (n = 7). If the two inhibitors were simultaneously applied, they seemed to counter-balance each other’s action (A1: 0.59 ± 0.18 nA; PPR: 1.15 ± 0.10; n = 6). Although the trends of the changes were noteworthy—TMC decreased A1 and increased PPR, whereas inhibition of ROK exerted the opposite effect—only the effect of TMC on the PPR was statistically significant.
Trains of stimuli (ν = 50 Hz) were also applied. This stimulus protocol is more sensitive than the application of paired-pulse stimuli (Srinivasan et al. 2008), thus it was more likely that any effect of TMC or Y27632 could be revealed. To quantify the changes, the relative amplitudes of consecutive stimuli were monitored (Fig. 6). TMC reduced the amplitude of the first EPSC and increased the relative amplitudes of the second and subsequent EPSCs (Fig. 6a1). Pre-incubation with Y27632 caused the opposite effect: it increased the amplitude of the first EPSC and decreased the relative amplitudes of the third and further EPSCs (Fig. 6a2). When the slices were pre-incubated with both inhibitors simultaneously, no significant change occurred (Fig. 6a3).
TMC and Y27632 altered the time course of the EPSCs as well. This effect was quantified by fitting the falling phase of the EPSCs to a single exponential function. TMC markedly slowed down the falling phase: the decay time constant changed from 1.91 ± 0.35 ms (n = 8) to 3.39 ± 0.67 ms (n = 6; Figure S2b1 and b2). Y27632 reduced the time constant to 0.820 ± 0.095 ms (n = 6). When both inhibitors were present, the decay time constant became 4.86 ± 1.26 ms (n = 6)—a finding that cannot be explained by the possible antagonistic effects of the PP1 and ROK inhibitors. However, as detailed characterization of the kinetic changes was outside the scope of this work, no further experiments were performed addressing the possible background of this phenomenon.
mEPSCs were also recorded and analyzed (Fig. 6b and c). mEPSCs are the consequences of non-action potential-related, stochastic vesicular release from the pre-synaptic terminal. TMC pre-incubation significantly reduced the frequency (from 2.46 ± 0.39 Hz to 1.15 ± 0.45 Hz), but did not change the amplitude of the mEPSCs. In contrast, when Y27632 was applied, the frequency remained unchanged but the amplitude was significantly reduced (from 18.7 ± 1.0 pA to 12.1 ± 1 pA). TMC significantly reduced the coefficient of variation of the mEPSC amplitudes (from 0.54 ± 0.07 to 0.37 ± 0.03; p < 0.05), whereas pre-incubation with Y27632 had no appreciable effects (0.58 ± 0.02).
Our results indicate that PP1M and ROK are present in both cortical synaptosomes and in a giant synapse of the rat auditory brain stem. PP1M and ROK affect synaptosomal exocytosis and influence neurotransmission by acting on both pre- and post-synaptic targets. Selective inhibition of PP1 and ROK exert opposite effects: the former suppresses, the latter stimulates exocytosis and neurotransmitter release. As PP1c is assumed to exert its effect complexed with regulatory/targeting proteins (Cohen 2002; Mansuy and Shenolikar 2006), we assessed the significance of PP1M (i.e., the PP1c-MYPT1 complex) in synaptic transmission. PP1M is the predominant holoenzyme in synaptosomes (see Figure S1a) and it may be the major determinant of PP1-dependent dephosphorylation processes. It has already been suggested that PP2A regulates neurotransmission: PP2A-specific inhibitors (OA, fostreicin) reduced glutamate release or depolarization-induced exocytosis of synaptosomes (Baldwin et al. 2003) and OA blocked basal synaptic transmission in hippocampal slices (Koss et al. 2007). In this study, inhibition of PP1 decreased the extent of KCl-induced exocytosis, indicating that PP1-type phosphatases are involved in the regulation of synaptic exocytosis. PP1M activity is regulated via a ROK-mediated inhibitory phosphorylation of Thr696 in MYPT1 (Hartshorne et al. 2004). This mechanism is present in synaptosomes too, as indicated by the Y27632-mediated suppression of MYPT1Thr696 phosphorylation. We have previously shown that PP2A-specific phosphatase inhibition increases the phosphorylation level of MYPT1 at Thr696 and Thr850 (Lontay et al. 2005; Kiss et al. 2008), resulting in an inhibition of PP1M. We propose that the inhibitory effect of OA on synaptosomal exocytosis is exerted via suppression of PP2A (a MYPT1 phosphatase) thereby increasing the inhibitory phosphorylation of MYPT1.
Beside MLC20, a well-known substrate of both ROK and PP1M, synapsin-I, and syntaxin-1 have emerged as possible novel interactive partners and substrates for PP1M and ROK in synaptosomes. An obvious question is how the phosphorylation of these substrates relates to synaptic exocytosis and neurotransmitter release. The information available in the literature is controversial about the effects of myosin phosphorylation on neurotransmitter release. Myosin phosphorylation increased acetylcholine release in cultured superior cervical ganglion neurons (Mochida 1995) and it also contributed to fusion pore expansion during exocytosis in chromaffin cells (Neco et al. 2008). Moreover, it has also been indicated that actomyosin-based contraction coupled with MLC20 phosphorylation may stabilize the cortical actin network, leading to inhibition of exocytosis (Vitale et al. 1995; van Leeuwen et al. 1999). In addition, ROK-dependent MLC20 phosphorylation is involved in synapse loss under neuropathological conditions (Sunico et al. 2010). In this work, the phosphorylation levels of MYPT1 and MLC20 were diminished by the inhibition of ROK during KCl-evoked depolarization of synaptosomes, whereas PP1 inhibition increased the phosphorylation levels of both proteins. These results suggest that myosin phosphorylation and ROK activity correlate inversely with synaptosomal exocytosis.
We provide evidence that ROK phosphorylates synapsin-ISer9 and syntaxin-1Ser14, and these phosphorylated proteins are the substrates of PP1. MYPT1 increased the activity of PP1c toward these substrates in vitro, suggesting that MYPT1 has a targeting role in the PP1M-mediated dephosphorylation of synapsin-IpSer9 and syntaxin-1pSer14. It has been shown that phosphorylation of Ser9 in an N-terminal peptide of synapsin-I by PKA counteracts the suppressive effect of the dephosphorylated peptide on synaptic transmission (Hilfiker et al. 2005). Synapsin-ISer9 phosphorylation by PKA influences synaptic vesicle life cycle during synaptogenesis (Bonanomi et al. 2005) and promotes synapsin-I dissociation from vesicles, allowing its dispersion (Chi et al. 2001) and enhancing the rate of synaptic vesicle exocytosis (Menegon et al. 2006). Our results show that the amount of synapsin-IpSer9 increases during KCl-evoked exocytosis. However, further increase in the amount of synapsin-IpSer9 upon PP1 inhibition or its decrease by ROK inhibition correlates inversely with the observed changes in exocytosis and this controversy cannot be resolved at this point. Nevertheless, it has been shown that caseine kinase-2-dependent phosphorylation of syntaxin-1Ser14 suppresses glutamate release from synaptosomes (Gil et al. 2011). ROK-dependent phosphorylation of syntaxin-1Ser14 has been shown to increase the affinity of syntaxin-1pSer14 to tomosyn and this interaction impairs SNARE complex formation, thereby it could contribute to reduced exocytosis (Sakisaka et al. 2004). This regulatory feature of ROK was not observed in cells over-expressing a Ser14→Ala mutant of syntaxin-1, indicating the significance of syntaxinSer14 phosphorylation in this process. Our results show that increase in syntaxin-1pSer14 by PP1 inhibition and decrease of syntaxin-1pSer14 upon ROK inhibition are accompanied with suppressed or stimulated synaptosomal exocytosis, respectively. Although more work is needed to determine the significance of the changes in the phosphorylation levels of different proteins (e.g., MLC20, synapsin-I, and syntaxin-1) in synaptic transmission, we propose that the regulation of syntaxin-1Ser14 phosphorylation by ROK and PP1M is a major factor controlling synaptic exocytosis and neurotransmitter release.
PP1 and ROK inhibition affected synaptic transmission in a thin-slice preparation. The giant synapse studied here is often utilized for the investigation of synaptic transmission (Cant and Morest 1979; Ryugo and Sento 1991; Isaacson and Walmsley 1996; Ryugo et al. 1996; Oleskevich and Walmsley 2002; Oleskevich et al. 2004). Our results demonstrate that MYPT1 and ROK are present pre- and post-synaptically in the endbulb of Held-bushy cell synapse. Consistent with these data, both of them were found in pre-and post-synaptic locations in synaptosomes, too (Fig. 1). In addition, MYPT1 was found in association with PP1c and ROK in the pre-synaptic fraction (Lontay et al. 2004), while other neuronal-targeting subunits of PP1 (e.g., neurabin, neurofilament-L) were identified in the post-synaptic densities of synaptosomes (Terry-Lorenzo et al. 2002). Our present findings indicating that PP1M interacts with and may be involved in the dephosphorylation of several pre-synaptic proteins are in accord with the conclusion that PP1M is the major phosphatase in the pre-synaptic terminal.
We demonstrate that inhibition of either PP1M or ROK influences neurotransmission. To identify the most likely mechanism(s) of action, the following considerations should be made.
(a) Several bodies of evidence indicate that pre-synaptic mechanisms are involved in the mediation of the effects. (i) Both TMC and Y27632 altered the relative amplitudes of the third and subsequent evoked EPSCs. Considering the direction of the changes, we conclude that TMC reduces, whereas Y27632 increases the release probability. (ii) TMC changed the paired-pulse ratio (the slight depression was transformed to facilitation) and it caused a significant reduction of the frequency of the mEPSCs. These effects provide additional support to the hypothesis that TMC is capable of reducing the release probability, although they do not give information about possible changes of the size of the readily releasable pool (Stevens and Wesseling 1999). (iii) TMC significantly reduced the coefficient of variation of the mEPSC amplitudes. As reduced release probability is associated with a reduced variance (Silver et al. 1998; Oleskevich et al. 2000), this observation is another argument for the release probability reducing effect of TMC.
(b) There is strong evidence for the involvement of a post-synaptic mechanism in mediating the effects of both TMC and Y27632, as the former increased, whereas the latter decreased the decay tau. An alteration of the decay of the EPSCs may have three reasons: change of the amount of neurotransmitter released, change of the rate of neurotransmitter removal from the synaptic cleft, and change of post-synaptic receptor desensitization/deactivation. However, under point (a) it was concluded that TMC reduces, whereas Y27632 increases the release probability—in other words, TMC reduces the amount of glutamate released into the synaptic cleft and Y27632 has the opposite effect. If only the amount of neurotransmitter was reduced in the synaptic cleft, the decay tau would be reduced (Takahashi et al. 1995). Therefore, our observation (i.e., the combination of reduced release probability and increased decay tau) can only be explained if an effect on the post-synaptic receptors is assumed. We conclude that TMC reduces, whereas Y27632 increases the rate of desensitization (or deactivation) of post-synaptic non-NMDA receptors.
(c) Y27632 caused a significant decrease of the mEPSC amplitude, which may be explained by either pre- or post-synaptic mechanisms. Decreased mEPSC amplitude may be the consequence of an effect on the post-synaptic receptors, resulting in, inter alia, their reduced sensitivity. Alternatively, reduced mEPSC amplitude may also occur because of pre-synaptic changes (e.g., reduced neurotransmitter content of the vesicles). However, as Y27632 did not change significantly the coefficient of variation of the mEPSC amplitudes, a strictly pre-synaptic mechanism of reducing mEPSC amplitude is unlikely.
Taken together, we conclude that both PP1M and ROK participate in the regulation of neurotransmission by acting via both pre- and post-synaptic mechanisms. Our observations provide further evidence that phosphorylation/dephosphorylation processes play important roles in the regulation of neurotransmission by, inter alia, determining the response features of the non-NMDA receptors (Swope et al. 1992; Wang et al. 2005; Lee 2006; Derkach et al. 2007; Santos et al. 2009) or controlling the phosphorylation status of post-synaptic proteins (Lee 2006). In fact, post-synaptic localization of ROK and its involvement in the control of dendritic spine morphology (Schubert et al. 2006) and protrusive motility (Tashiro and Yuste 2004) were evident in previous studies, although the phosphorylation targets of this kinase have remained elusive. The involvement of PP1M in post-synaptic processes is also supported by a previous article, demonstrating that the PP1M-specific inhibitor CPI-17 is present post-synaptically and it is required for long-term depression (Eto et al. 2002). Our present data reveal binding of CaMKII to MYPT1. The facts that—beside their pre-synaptic occurrence—both CaMKII and MYPT1 are present post-synaptically, suggests functional significance of this interaction. Binding of autophosphorylated CaMKII to post-synaptic densities switches the dephosphorylating phosphatase from PP2A to PP1 (Strack et al. 1997), but the targeting subunit for PP1c in this process has not been identified yet. Our data support the hypothesis that the PP1M holoenzyme acts as a potential CaMKII phosphatase and MYPT1 targets PP1c toward the CaMKII autophosphorylation site. However, further studies are required to uncover the exact targets and functions of both ROK and PP1M in post-synaptic phosphorylation.
This work was supported by grants from the Hungarian Scientific Research Fund OTKA K68416, K72812, PD 75276, CNK 80709, from the Ministry of Health of Hungary ETT 244/2006, by TÁMOP 4.2.2.-08/1-2008-0019 DERMINOVA, TÁMOP-4.2.2/B-10/1-2010-0024 and TÁMOP-4.2.1./B-09/1/KONV-2010-0007. B.L. and Z.S. are recipients of the J. Bolyai Fellowship. The authors are indebted to Krisztina Pocsai, Ágnes Németh, and Zsuzsanna N. Fekete for their excellent technical assistance.