Ion channels in vertebrate glia.

More than 20 years ago, Kuffler (1967) began a review of glial electrophysiology with a summary of hypotheses of glial functions; these functions were much the same as those proposed by Nageotte, Golgi, Lugaro, and Ram6n y Cajal 70 years before him (reviewed by Somjen 1988) and are not substantially changed today. Yet the timelessness of these views may result more from our continued ignorance than from the prescience of these pioneers. Ram6n y Cajal (1909) suggested that this poor understanding of function originated because physiologists did not have the tools to study glial cells directly. Only recently has this changed. The last 30 years of glial electrophysiology can be divided into three periods of study. Classical studies (prior to 1970, beginning most prominently with the work of Kuffler & Potter 1964, and reviewed by Kuffler 1967 and Somjen 1975) primarily consisted of microelectrode voltage recording from the large glial cells of invertebrates and lower vertebrates in situ. During the second period (beginning with Dennis & Gerschenfeld 1969; reviewed by Kuffler et al 1984), the microelectrode technique was used to study mammalian glia, either in situ or in vitro. Since 1982 (Kettenmann et al 1982), the patch-clamp technique has made small cells accessible to voltage-clamp recording. Among the many findings of the first two periods, three of the general principles to emerge, thought to hold for all glial cells in all species, were: (a) glial membranes are mainly permeable to potassium, (b) glial