Calcium dependent modification of distal C-terminal sequences of glycine transporter GlyT1

Neurochemistry International 57 (2010) 254–261

Contents lists available at ScienceDirect

Neurochemistry International journal homepage: www.elsevier.com/locate/neuint

Calcium dependent modification of distal C-terminal sequences of glycine transporter GlyT1 Martina Baliova, Frantisek Jursky * Laboratory of Neurobiology, Institute of Molecular Biology, Slovak Academy of Sciences, Dubravska cesta 21, 842 51 Bratislava, Slovakia

A R T I C L E I N F O

A B S T R A C T

Article history: Received 20 December 2009 Received in revised form 21 May 2010 Accepted 4 June 2010 Available online 11 June 2010

Glycine transporter GlyT1 plays important role in maintaining accurate glycine concentration in local brain microenvironment. Transporting efficiency of GlyT1 is strongly affected by the state of its distal Cterminus, which regulates transporter trafficking and cellular surface density. Using selected range of antibody epitopes against C-terminal region of GlyT1 we investigated its changes during calcium overload, the ubiquitous phenomena of several brain pathologies. We show that immunoreactivity against the last 12 amino acids of GlyT1C-terminal region exhibits robust calcium dependent decline, while the immunoreactivity of closely located region shows relatively small changes. Process is fully blocked by calcium chelation and inhibited by cysteine proteases inhibitors as well as inhibitors of protein kinase C. Distal GlyT1C-terminal end contains PDZ binding motif responsible for GlyT1 interaction with trafficking and clustering proteins. Its removal/modification could be part of the mechanism changing glycine homeostasis during physiological/pathological conditions characterized by elevated calcium. ß 2010 Elsevier Ltd. All rights reserved.

Keywords: Neurotransmitter Transporter Glycine Protein kinase C PKC GlyT1 Calpain Caspase Calcium

1. Introduction Amino acid glycine acts as a brain neurotransmitter, influencing both inhibitory and excitatory brain activity (Betz, 1992; Johnson and Ascher, 1987). Two glycine transporters GlyT1 and GlyT2, members of the family of sodium dependent neurotransmitter transporters play major role in the regulation of brain neurotransmitter glycine pools (Nelson, 1998). They control local glycine concentration in various brain regions, where their localization has been described using immunohistochemical and pharmacological methods (Borowsky et al., 1993; Jursky and Nelson, 1995, 1996; Zafra et al., 1995; Cubelos et al., 2005a; Luccini et al., 2008). Further complexity and compartmentalization is reflected by the existence of species specific GlyT1a, b, c as well as GlyT2a, b, c transporter subtypes (Kim et al., 1994; Ponce et al., 1998; Jursky and Baliova, 2002; Ebihara et al., 2004). Similarly as in the case of other transporter family members, acute regulation of glycine concentration during brain activity is achieved by coupling the glycine transporters cycle to sodium/ chloride gradient across the membrane and cycling of the transporters molecules between membrane and intracellular

* Corresponding author. Tel.: +421 2 5930 7437; fax: +421 2 5930 7416. E-mail address: [email protected] (F. Jursky). 0197-0186/$ – see front matter ß 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuint.2010.06.003

compartments (Nelson, 1998; Masson et al., 1999; Robinson, 2002). Cell surface localization is influenced by the transporter Nterminal interactions with Syntaxin 1A and its C-terminal interactions with trafficking and clustering proteins (Geerlings et al., 2001; Cubelos et al., 2005a,b). To study the glycine transporter function in vivo, genes for GlyT1 and GlyT2 has been deleted in mouse (Gomeza et al., 2003a,b). Despite that both knock-outs die soon after they are born, several important facts about the function of both transporters have been learned from these models. While mice deficient in GlyT1 resemble symptoms of non-ketonic hyperglycinemia, GlyT2 absence leads to hyperplexia (Gomeza et al., 2003a,b; Rees et al., 2006; Eulenburg et al., 2006; Harvey et al., 2008). Experiments further suggest that GlyT1 transporter is responsible for termination of inhibitory glycinergic transmission on strychnine sensitive glycine receptor and GlyT2 rather serves for refilling the vesicles in glycinergic presynaptic terminals. NMDA receptor is colocalized with GlyT1 (Cubelos et al., 2005a) and contains glycine/D-serine co-agonist site. It was generally assumed that glycine acts as major co-agonist at this site in vivo. Recent studies however indicate that glial and neuronal-derived Dserine rather than glycine acts as endogenous ligand of NMDA receptor (Mothet et al., 2000; Panatier et al., 2006; Wolosker, 2007). Glycine concentration in cerebrospinal fluid is low micromolar and it should saturate glycine site on NMDA receptor.

M. Baliova, F. Jursky / Neurochemistry International 57 (2010) 254–261

GlyT1 inhibitors however potentiate NMDA responses, which indicate that transporters lower the glycine concentration under saturating level. Recent discovery of specific glycine transporters inhibitors allowed modification of the local glycine concentration in vivo. Such strategy seems to be very promising for treatment of schizophrenia and other psychoses caused by hypothetical hypofunction of NMDA (Sur and Kinney, 2007; Boulay et al., 2010; Javitt, 2009). Heterozygous GlyT1 knock-out mice are viable and have doubled extracellular glycine concentration. It was however recently reported, that they exhibit abnormal redistribution of hippocampal synaptic NMDA receptors into extra synaptic sites (Imamura et al., 2008). Thus despite mice seems to be phenotypically normal and even some increase of memory retention is observed, elevated brain glycine concentration may introduce unexpected pathological side-effects. The facts mentioned above indicate that glycine function in brain is tightly regulated or pathologically altered mostly by glycine transporters. Modification of cytosolic regions of transporters, which interconnect them with intracellular regulatory pathways, could be one of the ways to provoke brain glycine concentration changes. We recently described N-terminal truncation of GlyT1 with calpain protease (Baliova and Jursky, 2005). Because the major epitopes of previously used antibodies were located outside the short C-terminal fragment of GlyT1 removed by calpain, we were unable to detect the modification of GlyT1 on distal C-terminal region reported here. We hypothesize that such truncation/modification during the pathological calcium overload could either additionally contribute to abnormal GlyT1 function or it represents certain compensation regulatory feedback. 2. Materials and methods 2.1. Materials Peroxidase linked anti-rabbit antibodies were from Millipore (Temecula, CA, USA). ECL reagents, Calpain inhibitor I, caspase inhibitor Z-VAD-FMK, Chelerythrine were from Sigma (St. Loius, MO, USA). Oligonucleotides were synthesized by VBC Genomics Bioscience Research (Vienna, Austria). All other chemicals used were of the purest grade available. 2.2. Construction of GST-GlyT1C, GST-GlyT1CD3, GST-GlyT1CD12, GST-GlyT1C(12), meth-6Xhis-GlyT1C-GST fusion proteins and screening of calpain cleavage sites For fusion proteins GST-GlyT1C, GST-GlyT1CD3, GST-GlyT1CD12, regions containing amino acids F561-I638, F561-D635 and F561-G626 of mouse glycine transporter GlyT1b were amplified by PCR reaction using forward EcoRI primer 50 ttgtacgcagaattccagctctgccgc-30 identical for all three fragments and reverse SalI primers 50 -caacaagtcgactcatatccgggagtcctg-30 for GST-GlyT1C, 50 -ctcatgtcgactagtcctggaagcggctgg-30 for GST-GlyT1CD3 and BamHI primer 50 -ctggaggatcctcagcccacgatggggatc-30 for GST-GlyT1CD12. Fragments were inserted into pGEX5X-1 (GE Healthcare, Freiburg, Germany). Plasmids were transformed into BL21 (DE3) (Novagen, Merck, Darmstadt, Germany) and fusion proteins were purified after 2 h of induction of cultures with 0.3 mM IPTG according to manufacturer instructions. For the fusion protein Meth-GlyT1C-GST, the GlyT1C-terminal region was amplified using forward NdeI primer 50 -ccattgtaccatatgttccagctctgccgc-30 and reverse BamHI primer 50 -gcaacaacggatcccctatccgggagtcc-30 . Following digestion with restriction enzymes NdeI and BamHI the DNA was in inserted into pET21a (Novagen, Merck, Darmstadt, Germany) together with downstream in frame DNA sequences coding for glutathione-S transferase (Baliova and Jursky, 2005). Finally forward 50 -catatgcatcaccatcaccatcacc-3 and reverse 50 -atatggtgatggtgatggtgatgc-30 Meth-His-Tag primers were annealed together and ligated into NdeI site resulting in introduction of methionine initiated Meth-6Xhis-Tag upstream of Meth-GlyT1C coding sequence. Plasmid was transformed into Escherichia coli BL21 and overexpresion was achieved after 2 h of induction at 37 8C using 0.3 mM IPTG. Protein was distributed into soluble fraction as well as inclusion bodies. Soluble protein was isolated using Ni2+/NTA-agarose under native condition using homogenization and wash buffer (50 mM sodium phosphate pH 8.0, 0.3 M NaCl, 14 mM 2-ME, 20 mM Imidazole). Elution was achieved by supplementing the same solution with 10 mM EDTA. Eluate was diluted 5 with 25 mM sodium phosphate pH 7.2, 0.150 M NaCl, 1% Triton X-100 and recovered on GST-sepharose. All fusion protein used in this work were eluted with 10 mM glutathione and calpain cleavage as well as cleavage sites determination were performed as previously described (Franekova et al., 2008).

255

2.3. Selective affinity purification of epitopes Polyclonal rabbit antibodies were raised against whole mouse GlyT1C-terminal region (amino acids 554–638) using pMAL fusion protein as antigen and purified on affinity column as previously described (Jursky and Nelson, 1996). For separation of distal GlyT1C epitopes, DNA primers coding for last 12 amino acids of mouse GlyT1 flanked by EcoRI and HindIII restriction sites 50 -aattcagtaacggctccagccgcttccaggactcccggatatgaa-30 and 50 -agctttcatatccgggagtcctggaagcggctggagccgttactg-30 were synthesized, annealed and ligated into modified (BamHI+1 frame shifted) EcoRI/HindIII digested pET34(+) (Novagen, Merck, Darmstadt, Germany) downstream in frame with cellulose binding protein Tag. Plasmid was transformed into BL21, induced with IPTG and fusion protein was isolated on microcrystalline cellulose. Epitopes against last 12 amino acids of GlyT1C (anti-GlyT1C626–638) were isolated by affinity purification from polyclonal serum using affinity column with immobilized fusion protein using method described previously (Baliova et al., 2004). Fresh affinity column was also used for removal of these epitopes from whole anti-glyT1C antibodies. This resulted in antibodies directed against amino acids 554–625 of GlyT1C (antiGlyT1C554–626). The specificity of all antibodies was verified by cross absorption with corresponding fusion proteins. 2.4. Site directed oligonucleotide mutagenesis To replace potentially phosphorylated serine in position S636 with asparagine, GST-GlyT1C construct was mutated using primers P2-F 50 -cgcttccaggacgatcggatatgagtc-30 , P2-R 50 -gactcatatccgatcgtcctggaagcg-30 . For mutagenesis we used Quick-change mutagenesis kit (Agilent, Stratagene Products, La Jolla, CA, USA) according to manufacturer’s suggestions. 2.5. Isolation of synaptosomes and internal calpain activation Crude mouse hindbrain synaptosomes were isolated as described previously (Baliova et al., 2004) with the following modifications: homogenate was prepared in ice cold 0.35 M sucrose/5 mM Hepes–NaOH, pH 7.4, 10 mM EDTA, 2 mM EGTA. Following 5 min centrifugation at 1000  g to remove nuclei, synaptosomes were recovered from the supernatant by 10 min centrifugation at 12 000  g and washed once with sucrose solution. To remove excess of metal chelators, synaptosomes were washed two times with sucrose solution without EDTA and EGTA. Aliquots of synaptosomes were resuspended in 10 fold volume of 25 mM Hepes/NaOH pH 7.4, with and without the presence of 0.2 mM CaCl2 and incubated for 15 min at 37 8C. Two additional calcium-containing samples were incubated in presence of 50 mM calpain inhibitor I, or general caspase inhibitor 25 mM Z-VAD-FMK. In certain samples 5 mM MgCl2 was added as the necessary component of phosphorylation. Phosphorylation was blocked addition of 17 mM chelerythrine. Additionally, synaptosomes were incubated with ZnCl2, MnCl2 and CoCl2 (final 2 mM each) with and without presence of 2 mM CaCl2. Samples were dissolved in SDS sample buffer, resolved in 7.5% PAGE and transferred to imobilon. Blot was then probed with antibodies anti-GlyT1C554–625 and anti-GlyT1C626–638. 2.6. Western blot data analysis The relative intensity of immunostaining was quantified with UN-SCAN-IT, Silk Scientific Inc. Utah, USA and plotted with GraphPad Prism 4.00 for Windows, GraphPad Software, San Diego, CA, USA.

3. Results Despite the presence of several calpain cleavage sites in the recombinant GlyT1C fusion protein sequence, our previous screening of spinal cord synaptosomes did not show marked changes of GlyT1C-terminal immunoreactivity following calcium increase. Polyclonal antigenic determinants are often distributed heterogeneously and successful detection of proteolytic truncation in vivo could depend on position of major antibody epitopes (Baliova et al., 2009). To verify this possibility we decided to determine the exact position of calpain cleavage sites in recombinant GlyT1C fusion protein and to separate possible groups of antibody epitopes directed against peptide fragments released by calpain cleavage. Because protein sequencing by Edman degradation proceeds through the protein N-terminal end, it was possible to use only one set of the proteolytic fragments to determine the calpain cleavage sites. Some of the released fragments were too small to be separated by PAGE. To overcome this limitation, we took advantage of the fact that calpain specificity is mostly determined

[(Fig._1)TD$IG]

256

M. Baliova, F. Jursky / Neurochemistry International 57 (2010) 254–261

Fig. 1. Localization of most distal calpain cleavage site in C-terminal region of mouse glycine transporter GlyT1. Arrows points to the most distal cleavage site determined by protein sequencing of indicated fragment released by calpain cleavage. Lower part of the figure shows alignment of C-terminal domains of GlyT1 and GlyT2 and protruding sequence of GlyT1 removed by calpain.

by 11 amino acids (Tompa et al., 2004). In an alternative approach, which we used previously (Baliova et al., 2009), we introduced methionine at the beginning of the GlyT1C-terminal fragment and removed the stop codon from its end. This allowed us to fuse the GlyT1C-terminus with a bigger protein tag in the opposite orientation. Following calpain cleavage, the fragments bearing the N-terminal part of GlyT1C remain attached to the larger protein tag and allowed their isolation using the standard 12% SDS gel. When we made this construct, we found that GlyT1C-protein sequence in Meth-GlyT1C-GST fusion protein was cleaved by unknown endogenous E. coli proteolytic system (not shown). This activity interfered with subsequent calpain cleavage and isolation of proteolytic fragments for purpose of protein sequencing. In order to recover full-length protein we inserted initiating methionine and 6Xhistidine tag coding DNA linker upstream of GlyT1C-GST DNA coding sequence. Two steps isolation procedure using both 6Xhistidine and GST tag allowed as recovering the fulllength protein. Calpain cleavage of Meth-6Xhis-GlyT1C-GST fusion protein released three major proteolytic fragments indicating the existence of minimally three calpain cleavage sites (Fig. 1). Predicted size deduced from the gel mobility of the fragments suggested that

all cleaved sites are located in GlyT1C moiety of the fusion protein. The Edman protein sequencing of these fragments showed that cleavage sites are located in positions equivalent to R567/T68, T603/T604 and G625/S626 of whole mouse GlyT1B protein sequence. Location of the calpain cleavage site in position G625/ S626 revealed that calpain removes 12 amino acids from GlyT1Cterminus (Fig. 1). To test if antibody epitopes against this short amino acid region are present in our polyclonal serum, we constructed fusion of cellulose binding protein (CBD) with the last 12 amino acids of GlyT1C and probed it with total anti-GlyT1C antibody. Following the detection of specific antibody staining (not shown), epitopes anti-GlyT1C626–638 were purified on immobilized CBDGlyT1C626–638 fusion protein. Using the same fresh immobilized fusion protein, the epitopes anti-GlyT1C626–638 were removed from polyclonal anti-GlyT1C554–638 antibodies, resulting in anti-GlyT1C554–625 antibodies. In subsequent experiments we verified the position of immunoreactivity on individual bands of calpain cleaved GST-GlyT1C fusion protein for these two sets of antibodies (Fig. 2). Appearance of second band in close vicinity of intact GST-GlyT1C fusion protein in calpain cleaved sample, indicated that calpain removes short stretch of amino acids on its distal end, which corroborated the results obtained by protein sequencing of cleavage products of Meth-6Xhis-GlyT1C-GST fusion protein. As expected, removal of these amino acids correlated with disappearance of anti-GlyT1C626–638 epitopes in one of the closely located bands. Interestingly 12 amino acid truncated GST-GlyT1CD12 runs higher than intact protein, which indicates that distal 12 amino acid peptide has significant influence on fusion protein secondary structure. Fig. 2 also shows that following removal 12 amino acids and elimination of antiGlyT1C626–638 epitopes, there are still comparably strong antiGlyT1C554–625 epitopes suitable for detection of upstream located calpain cleavage sites. Epitopes were however completely eliminated by the next calpain cleavage site in position T603/T604 indicating that these epitopes reside between amino acids 604 and 626. Fig. 2 also indicates that in GST-GlyT1C additional calpain cleavage site exists, previously not observed in Meth6Xhis-GlyT1C-GST which is located between position R567/T568 and T603/T604. Mapping and separation of the anti-GlyT1C epitopes allowed us to verify the existence of calpain cleavage sites in isolated mouse synaptosomes. In order to allow calcium to enter the intrasynaptosomal compartment and activate endogenous calpain, we incubated synaptosomes in hypoosmotic high (200 mM) calcium buffer. As shown on upper part of the Fig. 3, high calcium showed only small effect on distribution of anti-GlyT1C554–625 epitopes,

[(Fig._2)TD$IG]

Fig. 2. Localization of anti-GlyT1C epitopes on peptide fragments released by calpain. Intact and calpain cleaved GST-GlyT1C fusion proteins were resolved in 12% PAGE. Samples were transferred to immobilon and positions of all cleaved fragments were visualized either with coomassie staining or anti-GST antibodies (figure two left panels). Right two panels indicate localization of anti-GlyT1C554–625 and anti-GlyT1C626–638 epitopes on protein fragments released by calpain.

[(Fig._3)TD$IG]

[(Fig._4)TD$IG]

M. Baliova, F. Jursky / Neurochemistry International 57 (2010) 254–261

257

Fig. 4. Effects of expected modifications in the last 12 amino acids region of GlyT1 on GlyT1C626–638 immunoreactivity. Potential calpain and caspase cleavage of GlyT1C was simulated by removing the DNA coding for the last 3 or 12 amino acids of GST-GlyT1C fusion construct. Mutational replacement of GlyT1 serine 636 codon with asparagine coding triplet was chosen to mimics the phosphoserine. Fusion proteins were isolated, transferred to immobilon and probed with anti-GlyT1C554–625 or anti-GlyT1C626–638 antibodies. All these modifications strongly and selectively influenced antiGlyT1C626–638 immunoreactivity.

638 epitopes, similarly as we observed in calpain mediated removal of last 12 amino acids of GST-GlyT1C (Fig. 4). The presence of general caspase inhibitor Z-VAD-FMK in synaptosomal samples indeed inhibited the decrease of distal GlyT1C immunoreactivity, with slightly lesser extend as observed with calpain inhibitor (Fig. 3). Some of the previous reports however indicated that general caspase inhibitor Z-VAD-FMK acts as a calpain inhibitor (Bizat et al., 2005), therefore we cannot exclude the sole effect of Z-VAD-FMK on calpain.

[(Fig._5)TD$IG]

Fig. 3. Differential influence of high calcium on proximal and distal immunoreactivity of GlyT1C-terminal region. Mouse synaptosomes were incubated for 15 min at 37 8C in 25 mM Hepes–NaOH, pH 7.4 containing either 2 mM EGTA or 0.2 mM calcium. Two additional calcium samples were incubated in presence of calpain inhibitor I (50 mM) or general caspase inhibitor Z-VAD-FMK (25 mM). Synaptosomes were dissolved with SDS sample buffer, resolved in 7.5% PAGE and transferred on immobilon. Cleavage of Cterminal regions of GlyT1 in mixture of synaptosomal proteins was visualized with anti-GlyT1C554–625 and anti-GlyT1C626–638 antibodies. The quantification data show row means SEM from three separate experiments.

indicating that calpain cleavage sites detected in recombinant fusion protein upstream of anti-GlyT1C626–638 epitopes are not extensively cleaved in synaptosomal preparations. On the contrary there was a significant decrease of anti-GlyT1C626–638 antibodies following calcium increase (lower part of Fig. 3). Decrease of immunoreactivity was strongly suppressed by calcium chelation and inhibited by calpain inhibitor. As it is shown on lower part of the Fig. 3 inhibition of calpain did not fully recover decrease of distal GlyT1C immunoreactivity, which suggested that some additional calcium dependent modifications could be involved. The presence of asparagine (D) in position 635 indicated possibility that distal end of GlyT1C could be a substrate of other calcium-induced enzymes, caspases. Removal of last three amino acids (-SRI), which mimics potential caspase cleavage, resulted in complete loss of anti-GlyT1C626–

Fig. 5. Effect of phosphorylation on calcium dependent modification of distal GlyT1C626–638 immunoreactivity. Mouse synaptosomes were incubated for 15 min at 37 8C either in 25 mM Hepes– NaOH pH 7.4 containing either 2 mM EGTA or supplemented with 0.2 mM calcium in the presence or absence of 5 mM magnesium. In additional magnesium and calcium-containing sample 17 mM chelerythrine was added to inhibit the phosphorylation. Synaptosomes were solubilized with SDS sample buffer resolved in 7.5% PAGE and transferred on immobilon. Influence of inhibitors on calcium dependent modification of last 12 amino acids of GlyT1 was screened with anti-GlyT1C626–638 antibodies (CH – abbreviates chelerythrine). The quantification data show row means SEM from three separate experiments.

[(Fig._6)TD$IG]

258

M. Baliova, F. Jursky / Neurochemistry International 57 (2010) 254–261

Fig. 6. Sensitivity of calcium dependent modification of distal GlyT1C626–638 immunoreactivity to certain metals. Mouse synaptosomes were incubated for 15 min at 37 8C in 25 mM Hepes–NaOH pH 7.4 alone or supplemented with indicated 2 mM bivalent metal cations (upper part of the figure). Alternatively metals were supplied together with 2 mM calcium (lower part of the figure). Synaptosomes were solubilized with SDS sample buffer, resolved in 7.5% PAGE, transferred on immobilon and stained with anti-GlyT1C626–638 antibodies.

Another calcium-induced modification, which can significantly alter the immunoreactivity is phosphorylation (Susarla et al., 2004). Phosphorylation prediction analysis of distal 12 amino acids peptide removed by calpain by software NetPhos 2.0 Server (www.cbs.dtu.dk/services/NetPhos/; Blom et al., 1999) showed that serines in positions 627, 630, 631 and 636 in GlyT1C could be potentially phosphorylated (score 0.124; 0.290; 0.118; 0.461, on scale 0–1). Replacement of most probably phosphorylated serine 636 with asparagine, which mimics phosphoserine, resulted in significant loss of anti-GlyT1C626–638 immunoreactivity in GSTGlyT1C fusion protein (Fig. 4). The presence of magnesium in synaptosomal samples stimulated effect of calcium mediated decrease of anti-GlyT1C626–638 immunoreactivity, which indicated that phosphorylation in this region indeed can occur (Fig. 5). Presence of chelerythrine suppressed magnesium stimulated decrease of anti-GlyT1C626–638 immunoreactivity, indicating that in addition to proteolysis, PKC mediated phosphorylation can take place in region of last 12 amino acids of GlyT1 during the calcium overload (Fig. 5). Activity of cysteine proteases such as calpain and caspases is influenced by various metal ions. To investigate their influence, we substituted calcium with Zn2+, Mn2+ and Co2+ (Fig. 6). Neither of these metals induced decrease of GlyT1C626–638 immunoreactivity in synaptosomes. When ions were supplemented together with calcium, manganese did not interfere with calcium-induced decrease of immunoreactivity. In contrast zinc and cobalt totally inhibited decrease of GlyT1C626–638 immunoreactivity induced by calcium overload. 4. Discussion Intracellular calcium concentration is important regulatory event. It is however poisonous when calcium signals exceed physiological range, which is almost ubiquitous event during cell injuries. In both normal and pathological conditions calcium frequently causes activation of calcium dependent proteases, kinases leading to proteolytic cleavage and phosphorylation of various substrates. We previously showed that cytosolic domains of several neurotransmitter transporters are substrates of calpain protease (Baliova et al., 2004, 2009; Baliova and Jursky, 2005; Franekova et al., 2008). In glycine transporter GlyT1C recombinant fusion protein we observed several calpain cleavage sites, which we were unable to sufficiently confirm in vivo. Absence of cleavage in synaptosomes could be explained via protein interactions, protective phosphorylation, but detection of calpain cleavage in

vivo also depends on position of antibody epitopes used for detection of cleaved fragments in complex mixture of proteins. Here we investigated whether separation of different anti-GlyT1C polyclonal antibody epitopes will allow us to detect in vivo some of the calpain cleavage sites, found in recombinant GlyT1Cterminus. Due to virtually identical rat and mouse GlyT1C-terminal sequences, we previously used mouse anti-GlyT1C antibodies for detection of calpain cleavage in rat spinal cord synaptosomes. The rat GlyT1C protein sequence however contains single 633F/L substitution when compared with corresponding mouse GlyT1Cterminal region. We found that this amino-acid change in the middle of calpain cleaved 12 amino-acid region has significant effect of interspecies affinity of anti-GlyT1C626–638 antibodies resulting in up to 20 fold difference in intensity of obtained immunoreactive signal. Above fact reflects the additional contribution to our previous inability to detect C-terminal calpain cleavage in rat spinal cord synaptosomes. We found that most antigenic regions in GlyT1C are located in the second half of the GlyT1C-terminus, which is in concert with previously published results (Olivares et al., 1994). We however further show that in contrast to epitopes 554–625 which are affected by calcium only minimally, the epitopes against the last 12 amino acids (626–638) of GlyT1C exhibits up to 95% decrease of immunoreactivity within 15 min upon calcium overload in mouse synaptosomal preparations. This time window is within the range of 5–30 min observed previously (Bi et al., 1997; Guttmann et al., 2002). Potential close association of transporter C-terminus with NMDA/calpain signaling complexes (Adamec et al., 1998; Cubelos et al., 2005a) as well as the relatively high abundance of both GlyT1 and calpain in spinal cord white matter (Borowsky et al., 1993; Chakrabarti et al., 1989; Jursky and Nelson, 1996; Ray et al., 2002) could be additionally responsible for relatively high extend of GlyT1C-terminal cleavage. Calpain inhibitor I significantly but not completely inhibited calcium-induced loss of distal GlyT1C immunoreactivity. This prompted us to screen if other calcium modulated proteases or kinases inhibitors could affect change of immunoreactivity. Application of general caspase inhibitor Z-VAD-FMK inhibited decrease of immunoreactivity, which was slightly lesser but comparable to that observed with calpain inhibitor, indicating possible involvement of caspases in this process. Since caspases exhibits asparagine specificity in their cleavage site, direct cleavage could occur after asparagine 635 in GlyT1C protein sequence. Such cleavage will remove last three amino acid representing PDZ binding motif. This potential caspase cleavage

M. Baliova, F. Jursky / Neurochemistry International 57 (2010) 254–261

site is located nine amino acids downstream of calpain cleavage site. Calpain and caspase frequently share substrates. Similar distance between one of the calpain cleavage sites and nine amino acid apart caspase cleavage sites is located in a-spectrin molecule (Wang, 2000). It was however previously reported that caspase general inhibitor Z-VAD-FMK could inhibits calpain (Bizat et al., 2005). For this reason we cannot exclude that the effect observed with Z-VAD-FMK is caused by calpain inhibition. Additionally caspase cleaves calpain inhibitor (Porn-Ares et al., 1998) and it could this way indirectly potentiate calpain cleavage. Caspase inhibition could then paradoxically result in inhibition of calpain activity. Above indicate that such cleavage could have complex regulation. Investigated GlyT1C-terminal region contains several potential phosphorylation sites and phosphorylation itself might cause significant decrease of anti-GlyT1C626–638 immunoreactivity. In order to distinguish between the phosphorylation and proteolytic truncation, we isolated synaptosomes in presence of high concentration of both EDTA and EGTA. We expected that this treatment will remove magnesium from ATP-Mg2+ complex, which is the real substrate of kinases and this will separate potential effect of phosphorylation. Subsequent addition of magnesium to calcium-containing samples indeed stimulated calcium mediated decrease of anti-GlyT1C626–638 immunoreactivity and this stimulation was suppressed by PKC inhibitor chelerythrine. Similar results were obtained with other PKC inhibitor bisindolylmaleimide II (not shown). Above indicate that during calcium overload in vivo, when magnesium is present, distal C-terminal region of GlyT1 might undergo calpain mediated truncation as well as PKC mediated phosphorylation. In consent with this, it was previously reported that GlyT1 is mainly regulated by protein kinase Ca in glioma C6 cells (Morioka et al., 2008). Involvement of cysteine proteases was supported also by influence of certain metals on decrease of anti-GlyT1C626–638 immunoreactivity. We used metals typically influencing cysteine proteases such as zinc, manganese and cobalt. When calcium was replaced with these metals, neither of them was able to induce decrease of anti-GlyT1C626–638 immunoreactivity. Additionally, while zinc and cobalt abolished the effect of calcium, manganese did not prevent calcium-induced decrease of immunoreactivity. These effects correspond with the potential involvement of cysteine proteases (Suzuki and Ishiura, 1983; Pontremoli et al., 1985; Perry et al., 1997). Interestingly amino acids similarity alignment of two glycine tranporters GlyT1 and GlyT2 shows that C-terminal domain of GlyT1 transporter contains 13 amino acids extension. Here we found that, almost exactly this extension is removed by calpain cleavage. Paradoxically truncated fusion protein runs higher in SDS-PAGE when compared with intact protein. Above indicate that removal/modification of this region could have profound influence on structure of whole GlyT1C-terminus. Glycine regulates several brain functions and glycine transporters significantly contribute in such processes. In previous work we reported calpain sensitive regions in N-terminal domains of GlyT1a and GlyT1b (Baliova and Jursky, 2005). Here we show that distal part of glycine transporter GlyT1C-terminus is cleaved by calpain protease. Region removed by calpain is involved in high density PDZ clustering of GlyT1, which results in higher capacity for glycine clearance. Artificial construct of GlyT1 transporter missing the last 12 amino acids removed by calpain exhibited decrease of glycine transport upon expression in HEK293 cell, which was about half of the Vmax observed with wild type of GlyT1 and truncation in the N-terminal domain even further decreased the uptake (not shown). This is in agreement with previous mutational studies (Cubelos et al., 2005b).

259

Similarly, it has been demonstrated in Muller glia (Gadea et al., 2002), which express the glycine GlyT1 transport system, that while moderate increase of intracellular calcium with ATP or caffeine resulted in stimulation of glycine uptake, the higher calcium levels induced by ionophores, which was inhibited by calpain inhibitor decreased the uptake. Similar effect on endogenous GlyT1 uptake we observed with ionophore induced calpain activation in rat C6 neuroblastoma cells (not shown). We did not make an attempt to study this model in more detail, since nonspecific calpain over activation leads to modification of many cellular substrates. It is therefore difficult to attribute change in GlyT1 uptake to specific GlyT1 calpain cleavage. However taken together above results, during the pathological insult resulting in calcium-induced calpain activation, one of the reasons for change of glycine transporter uptake efficiency might be modification of its cytosolic regions and especially its PDZ binding motif containing distal C-terminus. Pathways activated through calcium permeable NMDA receptor stimulation frequently lead to calpain activation (Waxman and Lynch, 2005). As a matter of fact NMDA induced calpain activation caused its truncation and downregulation providing negative feedback to hamper NMDA receptor mediated toxicity (Yuen et al., 2008). GlyT1 is colocalized with the NMDA receptor in the same protein complex (Cubelos et al., 2005a) and calpain mediated truncation of GlyT1 with subsequent decrease in glycine uptake capacity might cause glycine overload. If such overload increases over the 0.1 mM threshold, which is required for rapid NMDA internalization (Nong et al., 2003) it might contribute to receptor downregulation and protection of neurons against NMDA induced neurotoxicity. Protective function of glycine during brain insults has been previously reported (Tijsen et al., 1997; Tonshin et al., 2007). However, local glycine overload could potentionaly reach the millimolar concentrations when glycine exhibits toxic effect on neurons (Barth et al., 2005). Indeed, previously studied effect of various GlyT1 inhibitors showed negative, neutral or positive effects on development of cerebral ischemic infarction (Szabo, 2005). Interestingly treatment with the calpain inhibitor in the zymosan-induced paw inflammation provided anti-inflammatory and anti-hyperalgesic effects (Kunz et al., 2004). Above suggests that the activation of calpain is involved in the sensitization of nociceptive neurons and calpain activity and may present an interesting novel drug target in the treatment of pain and inflammation. It was demonstrated that intravenous or intrathecal administration of GlyT1 and GlyT2 inhibitors produced a profound anti-allodynia effect in a partial peripheral nerve ligation model and other neuropathic pain models in mice mediated through spinal glycine receptor a3 (Morita et al., 2008). Calpain activation and cleavage of C-terminal region of GlyT1 might lead to decrease of GlyT1 mediated glycine uptake. This could lead to local glycine concentration increase, which has virtually the same effect as application of GlyT1 inhibitors. Additionally in sciatic nerve ligation model of neuropathic pain, GAT1 is upregulated and GAT1 inhibitors have analgesic effects (Daemen et al., 2008). Since GAT1 is also truncated by calpain in C-terminal region (Baliova et al., 2009), calpain activation during inflammation or tissue damage with subsequent GlyT1C and GAT1C-terminal cleavage could therefore represent natural analgesic process during healing of damaged tissue. In summary in this article we provide the evidence that during the calcium overload and calpain activation, in addition to previously reported N-terminal truncation, calpain mediated cleavage and PKC mediated phosphorylation can occur within the last 12 amino acid of glycine transporter GlyT1. Such modifications, triggered by pathological or physiological stimuli modify/ remove important regulatory sequences of GlyT1, including PDZ

260

M. Baliova, F. Jursky / Neurochemistry International 57 (2010) 254–261

binding motif and it will likely result in rapid change of local glycine homeostasis. Acknowledgements This work was supported by Slovak Academy of Sciences grants VEGA 2/0052/10 and 2/0045/10. References Adamec, E., Beermann, M.L., Nixon, R.A., 1998. Calpain I activation in rat hippocampal neurons in culture is NMDA receptor selective and not essential for excitotoxic cell death. Brain Res. Mol. Brain Res. 54, 35–48. Baliova, M., Knab, A., Franekova, V., Jursky, F., 2009. Modification of the cytosolic regions of GABA transporter GAT1 by calpain. Neurochem. Int. 55, 288–294. Baliova, M., Jursky, F., 2005. Calpain sensitive regions in the N-terminal cytoplasmic domains of glycine transporters GlyT1A and GlyT1B. Neurochem. Res. 30, 1093– 1100. Baliova, M., Betz, H., Jursky, F., 2004. Calpain-mediated proteolytic cleavage of the neuronal glycine transporter, GlyT2. J. Neurochem. 88, 227–232. Barth, A., Nguyen, L.B., Barth, L., Newell, D.W., 2005. Glycine-induced neurotoxicity in organotypic hippocampal slice cultures. Exp. Brain Res. 161, 351–357. Betz, H., 1992. Structure and function of inhibitory glycine receptors. Q. Rev. Biophys. 25, 381–394. Bi, X., Chen, J., Dang, S., Wenthold, R.J., Tocco, G., Baudry, M., 1997. Characterization of calpain-mediated proteolysis of GluR1 subunits of alpha-amino-3-hydroxy5-methylisoxazole-4-propionate receptors in rat brain. J. Neurochem. 68, 1484–1494. Bizat, N., Galas, M.C., Jacquard, C., Boyer, F., Hermel, J.M., Schiffmann, S.N., Hantraye, P., Blum, D., Brouillet, E., 2005. Neuroprotective effect of zVAD against the neurotoxin 3-nitropropionic acid involves inhibition of calpain. Neuropharmacology 49, 695–702. Blom, N., Gammeltoft, S., Brunak, S., 1999. J. Mol. Biol. 294, 1351–1362. Borowsky, B., Mezey, E., Hoffman, B.J., 1993. Two glycine transporter variants with distinct localization in the CNS and peripheral tissues are encoded by a common gene. Neuron 10, 851–863. Boulay, D., Bergis, O., Avenet, P., Griebel, G., 2010. The glycine transporter-1 inhibitor SSR103800 displays a selective and specific antipsychotic-like profile in normal and transgenic mice. Neuropsychopharmacology 35, 416–427. Cubelos, B., Gimenez, C., Zafra, F., 2005a. Localization of the GLYT1 glycine transporter at glutamatergic synapses in the rat brain. Cereb. Cortex 15, 448–459. Cubelos, B., Gonzalez-Gonzalez, I.M., Gimenez, C., Zafra, F., 2005b. The scaffolding protein PSD-95 interacts with the glycine transporter GLYT1 and impairs its internalization. J. Neurochem. 95, 1047–1058. Chakrabarti, A.K., Banik, N.L., Powers, J.M., Hogan, E.L., 1989. The regional and subcellular distribution of calcium activated neutral proteinase (CANP) in the bovine central nervous system. Neurochem. Res. 14, 259–266. Daemen, M.A., Hoogland, G., Cijntje, J.M., Spincemaille, G.H., 2008. Upregulation of the GABA-transporter GAT-1 in the spinal cord contributes to pain behaviour in experimental neuropathy. Neurosci. Lett. 444, 112–115. Ebihara, S., Yamamoto, T., Obata, K., Yanagawa, Y., 2004. Gene structure and alternative splicing of the mouse glycine transporter type-2. Biochem. Biophys. Res. Commun. 317, 857–864. Eulenburg, V., Becker, K., Gomeza, J., Schmitt, B., Becker, C.M., Betz, H., 2006. Mutations within the human GLYT2 (SLC6A5) gene associated with hyperekplexia. Biochem. Biophys. Res. Commun. 348, 400–405. Franekova, V., Baliova, M., Jursky, F., 2008. Truncation of human dopamine transporter by protease calpain. Neurochem. Int. 52, 1436–1441. Gadea, A., Lo´pez, E., Herna´ndez-Cruz, A., Lo´pez-Colome´, A.M., 2002. Role of Ca2+ and calmodulin-dependent enzymes in the regulation of glycine transport in Mu¨ller glia. J. Neurochem. 80, 634–645. ˜ ez, E., Lo´pez-Corcuera, B., Arago´n, C., 2001. Calcium- and syntaxin Geerlings, A., Nu´n 1-mediated trafficking of the neuronal glycine transporter GLYT2. J. Biol. Chem. 276, 17584–17590. Gomeza, J., Hulsmann, S., Ohno, K., Eulenburg, V., Szoke, K., Richter, D., Betz, H., 2003a. Inactivation of the glycine transporter 1 gene discloses vital role of glial glycine uptake in glycinergic inhibition. Neuron 40, 785–796. Gomeza, J., Ohno, K., Hulsmann, S., Armsen, W., Eulenburg, V., Richter, D.W., Laube, B., Betz, H., 2003b. Deletion of the mouse glycine transporter 2 results in a hyperekplexia phenotype and postnatal lethality. Neuron 40, 797–806. Guttmann, R.P., Sokol, S., Baker, D.L., Simpkins, K.L., Dong, Y., Lynch, D.R., 2002. Proteolysis of the N-methyl-d-aspartate receptor by calpain in situ. J. Pharmacol. Exp. Ther. 302, 1023–1030. Harvey, R.J., Carta, E., Pearce, B.R., Chung, S.-K., Supplisson, S., Rees, M.I., Harvey, K., 2008. A critical role for glycine transporters in hyperexcitability disorders front. Mol. Neurosci. 1, 1–6. Imamura, Y., Ma, C.L., Pabba, M., Bergeron, R., 2008. Sustained saturating level of glycine induces changes in NR2B-containing-NMDA receptor localization in the CA1 region of the hippocampus. J. Neurochem. 105, 2454–2465. Javitt, D.C., 2009. Glycine transport inhibitors for the treatment of schizophrenia: symptom and disease modification. Curr. Opin. Drug. Discov. Devel. 12, 468–478.

Johnson, J.W., Ascher, P., 1987. Glycine potentiates the NMDA response in cultured mouse brain neurons. Nature 325, 529–531. Jursky, F., Baliova, M., 2002. Glycine neurotransmitter transporters. Biologia 57, 689–694. Jursky, F., Nelson, N., 1995. Localization of glycine neurotransmitter transporter (GLYT2) reveals correlation with the distribution of glycine receptor. J. Neurochem. 64, 1026–1033. Jursky, F., Nelson, N., 1996. Developmental expression of the glycine transporters GLYT1 and GLYT2 in mouse brain. J. Neurochem. 67, 336–344. Kim, K.M., Kingsmore, S.F., Han, H., Yang-feng, T.L., Godinot, N., Seldin, M.F., Caron, M.G., Giros, B., 1994. Cloning of the human glycine transporter type 1: molecular and pharmacological characterization of novel isoform variants and chromosomal localization of the gene in the human and mouse genomes. Mol. Pharmacol. 45, 608–617. Kunz, S., Niederberger, E., Ehnert, C., Coste, O., Pfenninger, A., Kruip, J., Wendrich, T.M., Schmidtko, A., Tegeder, I., Geisslinger, G., 2004. The calpain inhibitor MDL 28170 prevents inflammation-induced neurofilament light chain breakdown in the spinal cord and reduces thermal hyperalgesia. Pain 110, 409–418. Luccini, E., Romei, C., Raiteri, L., 2008. Glycinergic nerve endings in hippocampus and spinal cord release glycine by different mechanisms in response to identical depolarizing stimuli. J. Neurochem. 105, 2179–2189. Masson, J., Sagne´, C., Hamon, M., El Mestikawy, S., 1999. Neurotransmitter transporters in the central nervous system. Pharmacol. Rev. 51, 439–464. Morioka, N., Abdin, J.M., Morita, K., Kitayama, T., Nakata, Y., Dohi, T., 2008. The regulation of glycine transporter GLYT1 is mainly mediated by protein kinase Ca in C6 glioma cells. Neurochem. Int. 53, 248–254. Morita, K., Motoyama, N., Kitayama, T., Morioka, N., Kifune, K., Dohi, T., 2008. Spinal antiallodynia action of glycine transporter inhibitors in neuropathic pain models in mice. J. Pharmacol. Exp. Ther. 326, 633–645. Mothet, J.P., Parent, A.T., Wolosker, H., Brady Jr., R.O., Linden, D.J., Ferris, C.D., Rogawski, M.A., Snyder, S.H., 2000. D-serine is an endogenous ligand forthe glycine site of the N-methyl-D-aspartate receptor. Proc. Natl. Acad. Sci. USA 97, 4926–4931. Nelson, N., 1998. The family of Na+/Cl neurotransmitter transporters. J. Neurochem. 71, 1785–1803. Nong, Y., Huang, Y.Q., Ju, W., Kalia, L.V., Ahmadian, G., Wang, Y.T., Salter, M.W., 2003. Glycine binding primes NMDA receptor internalization. Nature 422, 302–307. Olivares, L., Aragon, C., Gimenez, C., Zafra, F., 1994. Carboxy terminus of the glycine transporter GlyT1 is necessary for processing of the protein. J. Biol. Chem. 269, 28400–28404. Panatier, A., Theodosis, D.T., Mothet, J.P., Touquet, B., Pollegioni, L., Poulain, D.A., Oliet, S.H., 2006. Glia-derived D-serine controls NMDA receptor activity andsynaptic memory. Cell 125, 775–784. Perry, D.K., Smyth, M.J., Stennicke, H.R., Salvesen, G.S., Duriez, P., Poirier, G.G., Hannun, Y.A., 1997. Zinc is a potent inhibitor of the apoptotic protease, caspase3. A novel target for zinc in the inhibition of apoptosis. J. Biol. Chem. 272, 18530–18533. Ponce, J., Poyatos, I., Arago´n, C., Gime´nez, C., Zafra, F., 1998. Characterization of the 50 region of the rat brain glycine transporter GLYT2 gene: identification of a novel isoform. Neurosci. Lett. 242, 25–28. Pontremoli, S., Sparatore, B., Salamino, F., Michetti, M., Melloni, E., 1985. The reversible activation by Mn2+ ions of the Ca2+-requiring neutral proteinase of human erythrocytes. Arch. Biochem. Biophys. 239, 517–522. Porn-Ares, M.I., Samali, A., Orrenius, S., 1998. Cleavage of the calpain inhibitor, calpastatin, during apoptosis. Cell Death Differ. 5, 1028–1033. Ray, S.K., Neuberger, T.J., Deadwyler, G., Wilford, G., DeVries, G.H., Banik, N.L., 2002. Calpain and calpastatin expression in primary oligodendrocyte culture: preferential localization of membrane calpain in cell processes. J. Neurosci. Res. 70, 561–569. Rees, M.I., Harvey, K., Pearce, B.R., Chung, S.K., Duguid, I.C., Thomas, P., Beatty, S., Graham, G.E., Armstrong, L., Shiang, R., Abbott, K.J., Zuberi, S.M., Stephenson, J.B., Owen, M.J., Tijssen, M.A., van den Maagdenberg, A.M., Smart, T.G., Supplisson, S., Harvey, R.J., 2006. Mutations in the gene encoding GlyT2 (SLC6A5) define a presynaptic component of human startle disease. Nat. Genet. 38, 801–806. Robinson, M.B., 2002. Regulated trafficking of neurotransmitter transporters: common notes but different melodies. J. Neurochem. 80, 1–11. Sur, C., Kinney, G.G., 2007. Glycine transporter 1 inhibitors and modulation of NMDA receptor-mediated excitatory neurotransmission. Curr. Drug Targets 8, 643–649. Susarla, B.T., Seal, R.P., Zelenaia, O., Watson, D.J., Wolfe, J.H., Amara, S.G., Robinson, M.B., 2004. Differential regulation of GLAST immunoreactivity and activity by protein kinase C: evidence for modification of amino and carboxyl termini. J. Neurochem. 91, 1151–1163. Suzuki, K., Ishiura, S., 1983. Effect of metal ions on the structure and activity of calcium-activated neutral protease (CANP). J. Biochem. 93, 1463–1471. Szabo, L., 2005. Differential effect of GlyT1 inhibitors on the development of ischemic cerebral infarction. J. Cereb. Blood Flow Metab. 25, S15. Tijsen, M.J., Peters, S.M., Bindels, R.J., van Os, C.H., Wetzels, J.F., 1997. Glycine protection against hypoxic injury in isolated rat proximal tubules: the role of proteases. Nephrol. Dial. Transplant. 12, 2549–2556. Tonshin, A.A., Lobysheva, N.V., Yaguzhinsky, L.S., Bezgina, E.N., Moshkov, D.A., Nartsissov, Y.R., 2007. Effect of the inhibitory neurotransmitter glycine on slow destructive processes in brain cortex slices under anoxic conditions. Biochemistry (Mosc) 72, 509–517.

M. Baliova, F. Jursky / Neurochemistry International 57 (2010) 254–261 Tompa, P., Buzder-Lantos, P., Tantos, A., Farkas, A., Szila´gyi, A., Ba´no´czi, Z., Hudecz, F., Friedrich, P., 2004. On the sequential determinants of calpain cleavage. J. Biol. Chem. 279, 20775–20785. Wang, K.K., 2000. Calpain and caspase: can you tell the difference? Trends Neurosci. 23, 20–26. Waxman, E.A., Lynch, D.R., 2005. N-methyl-D-aspartate receptor subtypes: multiple roles in excitotoxicity and neurological disease. Neuroscientist 11, 37–49.

261

Wolosker, H., 2007. NMDA receptor regulation by D-serine: new findings and perspectives. Mol. Neurobiol. 36, 152–164. Yuen, E.Y., Ren, Y., Yan, Z., 2008. Postsynaptic density-95 (PSD-95) and calcineurin control the sensitivity of N-methyl-D-aspartate receptors to calpain cleavage in cortical neurons. Mol. Pharmacol. 74, 360–370. Zafra, F., Arago´n, C., Olivares, L., Danbolt, N.C., Gime´nez, C., Storm-Mathisen, J., 1995. Glycine transporters are differentially expressed among CNS cells. J. Neurosci. 15, 3952–3969.