Average values are reported as mean ± SEM. Statistical analyses were performed using IgorPro and InStat v3 (GraphPad Software Inc., La Jolla, CA). Animals were continuously superperfused with normal external saline during all recordings. For most experiments, solution was delivered by a gravity-fed perfusion system and removed using a peristaltic pump. For experiments involving the application of channel blocking drugs or ion substitution, we designed and fabricated a microfluidic chip to generate laminar flow in a 1 ml chamber under the water immersion

MEK inhibitor side effects objective. In this system, solutions were delivered with a peristaltic pump (flow Luminespib clinical trial rate: 2.4 ml per minute) and inflow was changed between control and experimental

solutions via a manually controlled HPLC valve (Rheodyne, Rohnert Park, CA). Amiloride and Na+-free saline were applied for at least one minute of continuous superfusion. Controlled, mechanical stimuli were delivered using a calibrated glass probe whose movement was recorded on analog s-video tape during each experiment, as described (O’Hagan et al., 2005). The probe was moved using a piezoelectric bimorph (Piezo Inc, Boston, MA) driven by a custom-designed, low-noise, high-voltage amplifier and controlled by voltage pulses delivered via the patch-clamp amplifier (EPC-10), a buffer amplifier and filter (120 Hz), and control software (Patchmaster, HEKA, Bellmore, NY). Probes were fabricated from borosilicate glass Thymidine kinase rods (O.D. 1.2 mm) on a pipette puller (Sutter Instruments, Novato, CA) and mounted on the bimorph using beeswax to hold the probe inside a small glass sleeve. In initial experiments, spring constants were measured by two independent methods. The first involved fabricating a set of known masses from a length of metal wire and measuring the displacement produced by hanging that mass from the tip of the probe. The

effective spring constant, k, was found by fitting a plot of force (= mg) versus displacement with a line. The second used a microelectromechanical system (MEMS) based force-sensor that was fabricated and calibrated (k = 12.9 N/m) as in Park et al. (2007). The sensor was mounted on a piezoelectric actuator (PIHera P-622.Z; Physik Instrumente) and the tip of the sensor was brought into contact with the tip of the glass probe. The deflection of the glass probe for a given force was calculated from the difference between the movement of the piezoelectric actuator and the deflection of the force sensor. The spring constant of the glass probe was calculated from the measured force-displacement curves. The second method is more accurate and was used for all later probes.