NEWS IN PHYSIOLOGICAL SCIENCES Channels With Single Transmembrane Segments

To the Editor: In reviewing recent studies on the Isk K+ channel, Toru Takumi (4, who discovered this intriguing protein, suggests that together with the M2 protein of influenza virus and phospholemman, Isk defines a new group of voltage-dependent channels with single transmembrane segments, His argument ignores the sequence homology between phospholemman and two other membrane proteins, the y-subunit of Na+-K+-ATPase and the cyhemolysin of Staphylococcus aureus. As pointed out by Palmer and colleagues (3), the y-subunit of Na+-K+ATPase resembles phospholemman in the putative transmembrane domain (Fig. 1). Na+-K+-ATPase is an ATP-dependent ion transporter, rather than a voltage-gated channel, in which the functional role of the y-subunit is unclear (2). In contrast, cu-hemolysin is a prokaryotic protein that forms hexameric pores in eukaryotic target cells (1). The pores are large, exhibit weak ion selectivity, and are open at zero transmembrane potential. The homology between phospholemman and cu-hemolysin extends over a restricted region and might be dismissed had not this region (the glycine-rich loop) been implicated in pore function. During assembly, the loop becomes occluded and is inaccessible to proteases (6). Mutagenesis experiments suggest that the loop is either essential for a late step in assembly or for directing the formation of the transmembrane channel, conceivably by forming its lining (5). Therefore, phospholemman, the y-subunit of Na+-K+-ATPase, and cyhemolysin are apparently functionally diverse but, at least at the level of amino acid sequence, structurally related membrane proteins. While grouping the Isk channel, the M2 protein, and phospholemman is reasonable on functional grounds, the extrapolation to structural similarity is unwise in the absence of compelling experimental evidence. In any event, the comparison made by Dr. Takumi is proving highly stimulating.