Channel expression was determined using standard whole cell patch clamp recordings, as described previously [9 (link)]. Briefly, cells were removed from the incubator, and incubation media was replaced with a bath solution containing, in mM: 145 N-methylglucamine (NMG), 2 CaCl2, 1 MgCl2, 10 HEPES, 10 glucose, pH adjusted with HCl to 7.3, osmolality 330. The intracellular solution contained, in mM: 100 KCl, 35 NMG, 10 EGTA, 4 MgCl2, 1 CaCl2, 10 HEPES, pH adjusted with HCl to 7.3, osmolality 290. Once seal formation was achieved, cells were superfused by constantly flowing bath solution. Cells were clamped at a holding potential of -80 mV and currents activated by depolarization to 0 mV. In Fig. 4B and 4C, currents were recorded in the cell-attached patch configuration. In these experiments, the pipet contained intracellular solution, except that 100 mM NMG-Cl was substituted for KCl. The bath solution had 100 mM KCl substituted for NMG-Cl. This held the whole cell resting potential near 0 mV (data not shown), which enabled us to know the membrane potential across the patch. In cell attached patches, the pipet potential with respect to ground was held at +80 mV, and channels were activated by stepping the pipet potential to 0 mV. All currents were filtered at 2 KHz (Axopatch 200B amplifier, Axon Instruments, Union City, CA) and digitized at 10 KHz for whole cell recordings and 20 KHz for single channel recordings (pClamp 9, Axon Instruments). Recording pipets were fabricated from 8250 glass (Garner Glass, Claremont, CA), coated with Sylgard and firepolished. Pipet resistance was between 1.5 and 2.0 MΩ.