NAD⁺ and H₂O₂ activate a non-selective cation channel in striatal neurones (Smith et al. 2003). This conductance exhibits many similarities to the recently cloned TRPM2 channel and the NSNAD channel expressed by rat CRI-G1 insulinoma cells (Herson et al. 1999; Perraud et al. 2001). Here, we examine whether ADP-ribose, an activator of TRPM2 and NSNAD, induces a similar current in striatal neurones.

Cultured striatal neurones were prepared from humanely sacrificed E18 rat embryos. Patch clamp recordings were made at ~23°C after 7-14 days in vitro. Cells were usually voltage-clamped at -70 mV in either the whole-cell or outside-out patch configurations. The recording pipette contained (mM) kgluconate 140, NaCl 10, Na₂ATP 4, Na₂GTP 0.1, Hepes, 10, pH 7.2. As required this was supplemented with 0.5 mM ADP-ribose. The bath solution contained (mM) NaCl 140, KCl 5, MgCl₂ 2, CaCl₂ 1, Hepes 10, pH 7.4. CaCl₂ was replaced with 1 mM BaCl₂ when required. Data are expressed as mean ± S.E.M.

Whole-cell recordings with ADP-ribose in the pipette exhibited a time-dependent development of inward current (Imax = -340 ± 60 pA, n = 23). In all cells replacement of extracellular Ca²⁺ with Ba²⁺ lead to a substantial decrease in inward current; a similar reduction was also produced by buffering extracellular Ca²⁺ with 5 mM EGTA. In control experiments, without ADP-ribose in the recording pipette, a comparable Ba²⁺-sensitive current consistently failed to arise (n = 18). The current-voltage relationship of the ADP-ribose activated current was linear, with a reversal potential of -6 ± 3 mV (n = 8). This is indicative of a non-selective cation conductance, like that of TRPM2 and most other TRP channels. This single channel currents underlying the ADP-ribose response could be seen in both whole cell recordings and excised patches. These exhibited openings that often lasted seconds. The mean single channel slope conductance was 90 ± 5 pS (n = 10). Channel activity could be reversibly blocked by extracellular application of 0.5 mM LaCl₃ (n = 4). In both whole-cell and excised patch recordings, the ADP-ribose-activated current slowly declined in amplitude. This decay developed with a mean time constant of about 400 seconds (n = 6) and was incomplete, plateauing at around 55 % of the peak response.

The current described here shares many properties with those evoked by H₂O₂ and NAD⁺ in striatal neurones. Moreover these properties match those of heterologously expressed recombinant TRPM2. These parallels suggest TRPM2 channels may be functionally expressed in the CNS neurones.

KH is in receipt of EU Framework V Postdoctoral Fellowship