Metabolism of sphingosine-1-phosphate(S1P) by Chinese hamster ovary cell membranes results in an underestimation of potency at recombinants S1P receptors

Mark R. Dowling, Neil R.F. McGuiness, Michelle E Bradley, John Atkinson, Steven J. Charlton. Novartis Institutes for Biomedical Research, Horsham, West Sussex, U.K. RH12 5AB.

Sphingosine-1-phosphate (S1P) is an important signalling molecule which acts at a family of GPCRs, S1P1-5. A useful assay system for assessing pharmacology at S1P receptors is the [35S]GTPγS assay, particularly as it allows ligand(s) and receptor to be equilibrated prior to initiation of the assay by the addition of [35S]GTPγS. It has been demonstrated, however, that S1P is metabolised by several different membrane-bound enzymes (see Le Stunff et al., 2002 for review), which may impact upon observed potencies in GTPγS assays. The aim of this study was to examine the stability of S1P in the presence of CHO cell membranes and whether any observed metabolism could be inhibited.

Initial experiments examined the effects of increasing incubation time of S1P with CHO cell membranes expressing human S1P2 and S1P3 receptors (1.25 and 2.5 µµg well-1, respectively), prior to the addition of [35S]GTPγS. A time-dependent loss in S1P potency was observed in membranes expressing both S1P2 and S1P3 receptors.

To further examine the effects of incubation of S1P with CHO membranes a ‘bio-assay’ was developed. A range of S1P concentrations were incubated with or without CHO membranes in a 96-well plate, using conditions which mimicked the [35S]GTPγS assay. After a range of incubation times, 50 µL of this reaction mixture was added to CHO cells expressing S1P2 or S1P3 and the subsequent agonist-functional response was assessed by measuring peak agonist-induced Ca2+ response, using a Fluorometric Imaging Plate Reader (FLIPR). The pEC50 of S1P when pre-incubated with CHO cell membranes for 1 hour, compared to that without incubation with membranes, was 7.25 ± 0.20 and 8.36 ± 0.12 for S1P2 cells and 7.62 ± 0.2 and 9.40 ± 0.1 for S1P3 cells (n=3 ± s.e.m), confirming findings in the [35S]GTPγS assay.

The depletion of S1P by CHO membranes was investigated at 3 different temperatures, 4 °C, room temperature (rt; ~22 °C) and 37 °C. The samples were assayed using FLIPR with the S1P3 CHO cells. The pEC50 of S1P incubated with membranes at 4 °C (9.18 ± 0.22) was not significantly different to that without membranes at rt (9.40 ± 0.01). At higher temperatures, however, pEC50 of S1P was significantly (p < 0.05, unpaired Student’s t-test) lower (7.25 ± 0.19 at rt and 6.27 ± 0.12 at 37 °C).

To determine if a phosphatase enzyme was responsible for the loss in S1P potency, a broad spectrum protein phosphatase inhibitor cocktail (PPi - Sigma) was also included in the incubation. The effect of this cocktail on S1P metabolism was then determined using the FLIPR bio-assay at room temperature. At S1P3, the potency (pEC50) of S1P incubated without membranes (9.38 ± 0.13) was not significantly different from that incubated with membranes plus 1 % (v/v) PPi (9.39 ± 0.19), demonstrating that PPi completely inhibited S1P metabolism by CHO membranes.
This study suggests that phosphatase enzymes present on CHO cell membranes metabolise S1P, potentially resulting in a significant underestimation of S1P potency in functional assays.


Le Stunff H et al., (2002) J. Cell. Biol, 158 ; 1039-1049.
Saba J and Hla T (2004) Circ. Res, 94; 724-734.