ION CHANNELS IN SPINAL-CORD ASTROCYTES INVITRO .2. BIOPHYSICAL AND PHARMACOLOGICAL ANALYSIS OF 2 NA+ CURRENT TYPES

1. Na+ currents expressed in astrocytes cultured from spinal cord were studied by whole cell patch-clamp recording. Two subtypes of astrocytes, pancake and stellate cells, were morphologically differentiated and showed expression of Na+ channels at densities that are unusually high for glial cells (2-8 channels/mum2) and comparable to cultured neurons. 2. Na+ currents in stellate and pancake astrocytes were comparable to neuronal Na+ currents with regard to Na+-current activation (tau(m)) and inactivation (tau(h)) time constants, which were equally fast in both astrocyte types. However, they differed with respect to voltage dependence of activation, and current-voltage (I-V) curves were approximately 10 mV more positive in stellate cells (-11.1 +/- 5.6 mV, mean +/- SD) than in pancake cells (19.7 +/- 4.5 mV). Steady-state activation (m(infinity) curves) was 16 mV more negative in pancake (mean V1/2 = -48.8 mV) than in stellate cells (mean V1/2 = -32.7 mV). 3. Steady-state inactivation (h(infinity) curves) of Na+ currents was distinctly different in the two astrocyte types. In stellate astrocytes h. curves had midpoints close to -65 mV (-64.6 +/- 6.5 mV), similar to most cultured neurons. In pancake astrocytes h(infinity)-curves were -25 mV more negative, with midpoints close to -85 mV (84.5 +/- 9.5 mV). 4. The two forms of Na+ currents were additionally distinguishable by their sensitivity to tetrodotoxin (TTX). Na+ currents in stellate astrocytes were highly TTX sensitive [half-maximal inhibition (K(d)) = 5.7 nM] whereas Na+ currents in pancake astrocytes were relatively TTX resistant, requiring 100- to 1,000-fold higher concentrations for blockage (K(d) = 1,007 nM). 5. Na+ currents were fit by the Hodgkin-Huxley (HH) model. In pancake astrocytes, as in squid gigant axons, Na+-current kinetics could be well described with an m3h model, whereas in stellate astrocytes Na+ currents were better described with higher-order power terms for activation (m). On average, best fits were obtained using an m4h model. 6. Pancake astrocytes were capable of generating action-potential (AP)-like responses under current clamp whereas stellate astrocytes were not. The h(infinity) curve for APs shows that membrane potentials more negative than -70 mV are required to allow these responses to occur. Because the cells' resting potential was much more depolarized (-40 mV), this mismatch between the steady-state inactivation (h(infinity) curve) and resting potential, which results in Ne current inactivation, does not allow these responses to occur spontaneously and makes it unlikely that AP-like responses occur in spinal cord pancake astrocytes in vivo. 7. Although Na+ currents expressed in stellate astrocytes resembled, in all their features, Na+ currents observed in cultured neurons, the combination of negative h(infinity) curve and TTX insensitivity characterizing Na+ currents in pancake astrocytes has not been observed in excitable cells, suggesting the existence of an astrocyte-specific Na+ channel in these cells.