br Materials and methods br Results and discussion We recent

Materials and methods
Results and discussion We recently found that the GlyT1 N-terminal and C-terminal sequences could be modified by phophorylation, or partially lost under pathological conditions by calcium-dependent proteolysis (Baliova and Jursky, 2005, Baliova and Jursky, 2010). While recombinant GlyT1 N-terminal and C-terminal sequences contain several calpain cleavage sites, it is not clear if all these sites are actually cleaved in vivo. We previously purified two sets of Laminin 925-933 mg (antiGlyT1C603-626 and antiGlyT1C626-638) from a polyclonal serum against the GlyT1 C-terminus. They are excellent tools for studying the two most distal C-terminal calpain cleavage sites of GlyT1 in vivo (Baliova and Jursky, 2010). In this work we further divided the antiGlyT1C626-638 antibodies into two distinct pools depending mostly on the presence of phenylalanine 633 and serine 636. The results showed that later bursts of antibodies purified on a mouse GST-SNGSSRFQDSRI fusion protein, anti-GlyT1C626-638FS, were able to recognize both wild-type and S636D mutated mGlyT1; surprisingly they also recognized the upstream epitopes of the G626/S627 calpain cleavage site. This prevented them from being used to detect C-terminal GlyT1C calpain cleavage. In contrast, antibodies purified on the rat fusion protein GST-SNGSSRLQDSRI, anti-GlyT1C626-638S, stained only the terminal 12 amino-acid long calpain fragment; unfortunately they were completely dependent on the presence of an unmodified serine 636, meaning that they marked indistinguishingly both C-terminal cleavage by calpain and potential phosphorylation of serine 636 in vivo. Our previous experiments showed that pathological calcium increase in synaptosomal preparations caused the rapid decline of the epitopes closest to the C-terminus while the more internal epitopes were spared. There are several possible reasons for this differential effect. First, calpain might cleave the very C-terminal sequences at the previously identified calpain cleavage site G626/S627. Second, the calcium dependent decrease in antiGlyT1C626-638S immunoreactivity might reflect phosphorylation of Ser-636, since the phosphomimetic mutant S636D losses about 80% of the antiGlyT1C626-638 antibody signal (Baliova and Jursky, 2010). Finally, phosphorylation of some amino-acids might protect GlyT1C from calpain cleavage, explaining the differential sensitivity of the C-terminal region to calpain cleavage observed in our previous experiments (Baliova and Jursky, 2010). To investigate the effect of phosphorylation on calpain cleavage we performed a bioinformatics search for potential phosphorylation sites in the GlyT1 C-terminus using NetPhos (www.cbs.dtu.dk/services/NetPhos/; Blom et al., 1999). Since it was previously reported that GlyT1 is not significantly phosphorylated on tyrosine residues (Vargas-Medrano et al., 2011), we focused mainly on threonine and serine phosphorylation. We also investigated the conservation of predicted phophorylation sites by aligning the species-specific GlyT1 C-terminal sequences. Those potentially phosphorylated residues which were not conserved or were substituted with amino-acids that are not negatively charged were regarded as less likely to be phophorylated (Fig. 1A). Two approaches are available to investigate the effect of phosphorylation on the regulation of proteins. The first involves the inhibition of phosphatases or kinases by specific inhibitors, while the second uses phosphomimetic mutations which replace potentially phosphorylated serine residues with aspartic or glutamic acid. While the phosphomimetic approach is not always effective, it does have the advantage that only the molecule of interest is affected, and other cellular processes are unaltered. The GlyT1 C-terminus has several potentially phosphorylable threonine and serine residues. Serine 605 in the mouse GlyT1b transporter is highly conserved between species and it has a high probability of being phosphorylated (Fig. 1A). When this residue is replaced by a S605D phophomimetic mutation in a recombinant fusion protein calpain cleavage at T602/T603 is completely blocked (Fig. 1B). Interestingly, no other calpain cleavage sites are affected, which indicates that the overall structure of GlyT1C has likely not been modified extensively by this mutation. In addition a S605A neutral mutation does not block calpain cleavage at this site (Fig. 1B).