br Discussion CHO K cell lines stably expressing each of
Discussion CHO-K1 cell lines stably expressing each of the NH2-terminal isoforms of the human glycine reuptake transporter GlyT1 have been developed. GlyT1 uptake assays have been validated for two lines for each isoform, one low and high expressor. Expression levels in the lines have been confirmed by two independent measures: Vmax and active site titration. Both measures rank the cell lines in nearly the same order. This report is believed to be the first use of active site titration to measure site density in a cell-based transporter assay. At first glance the site densities measured by active site titration might seem implausible, as high as 1.1×107 transporters per cell. There is precedent, however, for transporter site densities this high. Aubrey et al. (2005) measured transporter currents in a CHO-K1 line stably expressing GlyT1. Site density was a model-dependent parameter that ranged in their data from 1.2×107 to 8.5×107 transporters per cell. Low- and high-expressing cell lines were selected to test whether 8-Bromo-cGMP, sodium salt sale level affects measured inhibitor potency, as reported for the serotonin transporter (Ramsey and DeFelice, 2002). No dramatic difference emerged. There is a weak trend for higher IC50 values (lower potency) in high-expressing cell lines. The two-way analysis of variance conducted for ALX-5407 IC50 values indicates that the difference between expression levels is statistically significant. This comparison is only suggestive, however, as “high” and “low” expressors are only a qualitative description of the cell lines. Actual expression levels have not been matched (Table 1). The data presented here confirm the observation of Williams et al. (2004) that select antipsychotics are weak inhibitors of GlyT1, with IC50 values ranging between 5μM and 50μM (except for olanzapine whose IC50 is greater than 100μM). One antipsychotic not available to Williams et al., aripiprazole, also inhibits GlyT1 with an IC50 near 10μM. These data consider the pharmacology of GlyT1 isoforms in a recombinant expression system, neglecting native protein-protein interactions. For example, neuronal GlyT1 interacts through its carboxyl terminus with the neuronal scaffold protein PSD-95. It is speculated that this interaction blocks recognition of neuronal GlyT1 by carboxyl-terminal antibodies, whereas these antibodies detect glial GlyT1. In contrast, amino-terminal antibodies detect both neuronal and glial GlyT1 (Cubelos et al., 2005, Raiteri and Raiteri, 2010). Amino-terminal isoforms of GlyT1 may enable differential formation of complexes with other proteins, with possible consequences for GlyT1 function or pharmacology. Before attempting to detect such complex interactions, it seemed prudent to characterize any differences in intrinsic isoform pharmacology. The inhibitors tested in this validation study barely distinguish among isoforms of GlyT1. The largest span observed is for ALX-5407, but even here there is only a 10-fold difference between the lowest and highest observed IC50 values (6-fold within the high-expressing cell lines). While this difference is statistically significant, it would be difficult to use a difference of this magnitude to distinguish isoforms pharmacologically. The cell lines reported here are tools for discovering inhibitors that might distinguish among GlyT1 isoforms.
Acknowledgments We thank Cephalon Medicinal Chemistry for synthesis and characterization of reference inhibitors, Beth Ann McKenna for experimental assistance, Karen L. Milkiewicz for preparation of Fig. 3, and Mark Ator and Rita Raddatz for comments on the manuscript.
Introduction Glycine is a well-characterized, simple amino acid with a unique role in the CNS. It can act either as an inhibitory neurotransmitter or as a potentiator of NMDA-dependent glutamatergic excitatory neurotransmission when acting as a necessary co-agonist for the NMDA receptor. The former function is mediated by ligand-gated, strychnine-sensitive glycine binding sites (often referred to as glycineA sites) in the spinal cord, cerebellum and brainstem (Betz, 1990). The latter function is mediated by ligand-gated, strychnine-insensitive glycine binding sites (often referred to as glycineB sites) in the thalamus, hippocampus and cortex (Johnson and Ascher, 1987, Bergeron et al., 1998, Danysz and Parsons, 1998).