TRPM2 is a member of the melastatin-related TRP protein family. It is expressed in the brain, forms a cation channel that is activated by intracellular ADP-ribose and is associated with induction of cell death (Perraud et al. 2001; Hara et al. 2002). We have further investigated the properties of FLAG-epitope tagged TRPM2 channels stably expressed in HEK-293 cells under the control of a tetracycline inducible CMV promoter. Using Western blotting we found that anti-FLAG antibody detected a single protein band of ~190 kDa only in tetracycline-induced cells. This is the expected size of human TRPM2. Both tetracycline-induced and non-induced cells were studied using calcium imaging and patch-clamp recording. Calcium imaging data are expressed as the change in fura-PE3 ratio (ΔR_340/380) for the 340 and 380 nm excitation wavelengths. Patch-clamp recording conditions were similar to those described in Perraud et al. (2001). There was 0.05 mM EGTA in the standard patch pipette solution and current amplitudes are given for -100 mV. Functional data are expressed as means ± S.E.M. The n values refer to the number of cells.

Functional expression of TRPM2 was first shown in calcium imaging experiments using protocols similar to those described by Hara et al. (2002) but using fura-PE3 as the calcium indicator. A large rise in intracellular calcium occurred in TRPM2-expressing cells after addition of 1 mM hydrogen peroxide to the calcium-containing bath solution (ΔR_340/380 = 0.982 ± 0.43; n = 12). In non-induced cells there was also a large rise in intracellular calcium, although this was about half the size of that in TRPM2 cells (ΔR_340/380 = 0.438 ± 0.65; n = 15; P < 0.05 when compared with TRPM2 cells using Student’s unpaired t test). Because of the size of this background signal we used patch-clamp recording for all other experiments. In non-induced cells with 0.5 mM ADP-ribose in the patch pipette, small currents occurred in the presence of extracellular calcium (-0.073 ± 0.046 nA, n = 3). In induced (TRPM2) cells, larger currents occurred with (-9.73 ± 1.361 nA, n = 5) than without (0.418 ± 0.227 nA, n = 5) ADP-ribose in the pipette. These data suggest there is tonic as well as ADP-ribose-induced TRPM2 function. We also observed a strong dependence of ADP-ribose-induced TRPM2 currents on calcium. If whole-cell recordings were initiated in the presence of barium, rather than calcium, currents were small (-0.236 ± 0.075 nA, n = 5). Moreover, when Ca²⁺ was strongly buffered in the intracellular solution with BAPTA (20 mM) currents were also small (-0.332 ± 0.096 nA, n = 5). Cell blebbing associated with TRPM2 currents in the presence of calcium was absent.

From these data we suggest that TRPM2 has some tonic activity in the absence of ADP-ribose and that the ADP-ribose-induced channel activity, which is associated with cell blebbing, is strongly calcium dependent.

We thank the BHF and Wellcome Trust for support