Effects of a Non-Hydrolyzable ADP-Ribose Analog on the Gating of the TRPM2 Channel

Abstract

TRPM2 is a non-selective Ca2+ permeable cation channel co-activated by intracellular Ca2+ and ADP-ribose (ADPR). ADPR binds to the C-terminal NUDT9-H domain. Intriguingly, when expressed in isolation, the NUDT9-H domain functions as an active ADPR hydrolase which breaks ADPR into AMP and ribose-5-phosphate. To exploit the relevance of this putative activity for channel gating, we examined the effect of AMPCPR-, a non-hydrolyzable ADPR analogue, on TRPM2 channel gating using macroscopic measurements and steady-state single-channel kinetic analysis. In biochemical experiments purified NUDT9, a mitochondrial ADPR-pyrophosphatase, showed efficient hydrolysis of ADPR, but completely failed to cleave AMPCPR.When applied to inside-out patches the apparent affinity of AMPCPR for TRPM2 activation was two orders of magnitude lower compared to ADPR (K1/2ADPR≈1μM, K1/2AMPCPR≈80μM), and even at a saturating concentration, it could only partially stimulate TRPM2 activity.We also compared macroscopic current relaxation time courses in response to sudden removal of ADPR or AMPCPR. The current decay time constant was almost four times shorter after AMPCPR removal than following ADPR withdrawal.In steady-state single-channel measurements TRPM2 shows bursting behaviour when gated by either nucleotide. We compared mean open times and mean burst durations in the presence of ADPR or AMPCPR. These parameters were independent of the concentrations of the agonists, but both were significantly shorter in the presence of AMPCPR.Our results identify AMPCPR as a low-affinity partial agonist. The fact that channel closing rate is not slowed in the presence of a non-hydrolyzable ligand suggests either that the NUDT9-H domain has no ADPR hydrolase activity, or that the latter is not required for channel closure.