The existence of gap junctions coupling neurones and astroglial cells was initially suggested by Nedergaard (1994), who observed propagated Ca²⁺ waves between astrocytes and neurones in mixed cultures. This observation was somehow neglected until very recently, when gap junctional coupling between co-cultured embryonic neurones and astrocytes was confirmed by both dye-transfer assay and direct measurement of junctional currents (Froes et al. 1999). Direct coupling between astrocytes and neurones was further substantiated by experiments in situ, in brainstem slices, which found dye transfer between astrocytes and neurones and demonstrated electrical synchronisation between spontaneously firing neurones and astroglial cells (Alvarez-Maubecin et al. 2000). In the present study we directly demonstrated electrical coupling between Bergmann glial cells (BG) and Purkinje neurones (PN) in acutely isolated cerebellar slices. Electrical coupling between these two cells was identified by dual whole-cell voltage clamp (Clark & Barbour, 1997), which allowed direct recording of junctional current.

Sagittal cerebellar slices (250 µm) were prepared from 14- to 21-day-old Sprague-Dawley rats as described previously (Kirischuk et al. 1995). Animals were humanely killed according to UK legislation. Whole-cell recordings from PN-PN, PN-BG and BG-BG pairs were made using Nomarski optics and infrared visualisation, which allowed precise morphological identification of cells. The pipettes used had a resistance of 3-4 mV when filled with intracellular solution (mM: KCl 130, MgCl₂ 2, Hepes/KOH 10, ATPNa₂ 4, Fura-2K₅ or Fluo-3K₅ 0.2, pH 7.3). Slices were continuously superfused by artificial extracellular solution containing (mM): NaCl 130, KCl 2.5, CaCl₂ 2.5, MgCl₂ 1, NaHCO₃ 26.3, KH₂PO₄ 1.6, glucose 10, pH 7.4 when equilibrated with 95 % O₂-5 % CO₂ at 22-24 °C. Current recordings were made with EPC-9/2 amplifier (HEKA, Lambrecht, Germany); electrical stimulation, data aquisition and off-line analysis were controlled by PULSE software (HEKA, Lambrecht, Germany). After establishing dual whole-cell configuration cells were held at -80 mV. Junctional currents were recorded by applying hyper- and depolarising voltage sequences ranging from -120 to +40 mV (voltage step 10 mV) to one of the cells in the pair, while ion currents were measured from both cells.

As has been shown before (Clark & Barbour, 1997), junctional currents were frequently observed in BG-BG pairs: we found electrical coupling in 27 out of 34 pairs analysed. When the similar protocol was applied to the PN-BG pairs junctional currents were found in 32 out of 48 pairs analysed. The electrical coupling was bi-directional as similar junctional currents were observed in PN when the voltage step protocol was applied to BG. To correlate the appearance of these currents with gap junctions we treated slices with octanol (0.2 mM) or halotane (0.1 mM) – known inhibitors of gap junction conductance. Both agents applied for 5 min resulted in a complete inhibition of junctional currents in PN-BG pairs (n = 5 for octanol and 5 for halotane). The washout (15 min) led to a complete recovery of junctional currents after treatment with octanol (n = 5 out of 5); the action of halotane was irreversible.

We conclude therefore that cerebellar neurones and glial cells are directly connected via gap junctions.

This study was supported by The Wellcome Trust (grant to A.V.).