Animals continued breathing independently. The right bulla was opened fully using a forceps; a hole was made in the left bulla to prevent pressure buildup in the left middle ear. Based on cranial landmarks, an ∼1 mm diameter craniotomy was created by carefully scraping the bone between the bulla and the brainstem with a small handheld drill, exposing the brain surface slightly laterally from the MSO. Dura, arachnoids, and pia mater were removed locally. In some experiments, recording locations were marked with biocytin (0.5%), which was added to the pipette solution, or with postrecording injection of saturated Alcian Blue at the recording position (Figure S1). In these

experiments, animals were sacrificed with a lethal dose of Nembutal and subsequently perfused intracardially with saline, followed by a 4% paraformaldehyde solution. Brains were further INCB018424 in vivo processed as described in Horikawa and Armstrong (1988) with minor modifications. Histology confirmed MSO as the recording location in 6 of 6 animals. Thick-walled borosilicate glass micropipettes with filament had a resistance of 3.5–6 MΩ when filled

with recording solution. Pipettes were filled with Ringer solution for juxtacellular recordings, which contained NaCl 135, KCl 5.4, MgCl2 1, CaCl2 1.8, HEPES 5 mM; for whole-cell recordings the pipette contained (in mM): 138 K-gluconate, 8 KCl, 0.5 EGTA, 10 HEPES, 10 Na2Phosphocreatine, 4 MgATP, 0.3 NaGTP (pH 7.2 selleck chemicals with KOH). Electrodes were typically inserted laterally (and ventrally) from the cell layer and advanced in dorsomedial direction at an angle of 20–30 degrees with the vertical. The thin somatic layer (Rautenberg et al., 2009) was identified based on the polarity reversal Bumetanide of the local field potential response (“neurophonics”) during alternating monaural click stimuli to the left and right ear (Figure S1; Biedenbach and Freeman, 1964; Clark and Dunlop, 1968; Galambos et al., 1959). Pipettes had a high positive pressure (>300 mbar) when crossing the brain surface, which was lowered to 10–30 mbar when approaching the cell layer (located

at 400–1,000 μm from the surface). Juxtacellular (loose-patch) or whole-cell recordings were made by slowly advancing the pipette while monitoring both its resistance and the presence of EPSP or spike activity. For juxtacellular recordings, pressure was released if a neuron was approached, and slight negative pressure was briefly applied while moving the electrode another 2 to 10 μm toward the cell until pipette resistance increased to a value of typically 30 MΩ. Because physical contact with a cell is essential for the large size of the juxtacellular potentials (Lorteije et al., 2009), we consider it very unlikely that another, nearby cell contributed significantly to the measured potentials.