Louis H. Philipson, Donald F. Steiner

K

[Ca²⁺]

Topologies of the inward rectifier K channel family and ATP-binding cassette proteins. NBF, nucleotide-binding fold; R, regulatory domain.

Sulfonylureas are insulin secretagogues used to treat diabetes mellitus, which depolarize cells by blocking K channels (3, 6). Cloning of the receptor for these important drugs and the clinical result of its malfunction are described in two reports in this issue of SCIENCE (7), which also shed new light on the nature of the K_ATP channel in islets.

K_ATP channels are also present in cardiac myocytes, where they were first described (8), and in muscle, neuronal, and mitochondrial membranes (9). These channels participate in a multitude of processes that link energy metabolism to regulation of the membrane potential. Modulated by G proteins and protein kinases, K_ATP channels have a rich pharmacology; medicinal applications include hypertension, asthma, heart disease, diabetes, and epilepsy (10). The differences in single-channel conductance, ATP sensitivity, and pharmacology are such that it is likely that K_ATP channels are encoded by a family of related genes (9).

K_ATP itself, the ion permeation subunit, may belong to a new family of K channel genes encoding inward rectifier-like K channels (11). Their transmembrane topology, unique among ion channels, seems to include only two membrane-spanning domains and an intervening region homologous to the pore domain of Shaker-like channels (Fig. 1A) (12). Three such genes are candidate K_ATP channels: ROMK1, rcKATP-1/CIR, and uK_ATP-1 (11, 13). ROMK1 has special importance in the renal tubule but seems to generate a current dissimilar to the cell IKATP. The cardiac rcKATP-1 and related genes (KATP-2, GIRK3) are the likely potential channel subunit genes. However, rcKATP-1/CIR is critical for a different K current, the G protein-coupled inwardly rectifying K channel (I_KACh) (13). I_KACh thus seems to be heteromultimeric, possibly with a large number of subunits, comprised of both rcKATP-1/CIR along with GIRK1, originally thought to encode this channel by itself (13). The rcKATP-1/CIR channel messenger RNA is expressed in brain and other tissues, and the protein may heteromultimerize to generate several K currents (11, 14). None of the two-membrane-spanning domain channels has the sensitivity to sulfonylureas seen in native cell K_ATP channels. Now we see that these or similar channels may associate in some way with a separate protein that does bind sulfonylureas, newly described in the reports in this issue (7).

The sulfonylurea receptor (SUR) has been wooed and won by a more traditional strategy: The protein was purified through the use of covalently bound sulfonylurea derivatives and partially sequenced. Complementary DNA (cDNA) probes were then employed to screen multiple cDNA libraries to obtain the entire cDNA. The two reports in this issue identifying the cDNA encoding SUR from Aguilar-Bryan et al. and Thomas et al. are thus the products of almost 10 years of work, begun in the laboratory of Chris Boyd. Surprisingly, they have found that the binding protein (no ion permeation has yet been found with the expressed protein) is a member of the growing family of ATP-binding cassette proteins. This family includes the P-glycoprotein and multidrug resistance (MRP) proteins, which cause chemotherapeutic drug resistance when overexpressed, and the CFTR protein, mutations of which cause cystic fibrosis [see figure (14)]. These proteins all have multiple membrane-spanning domains and two nucleotide-binding folds; at least some of them are ATP-hydrolyzing pumps, while the CFTR behaves as a channel. A surprising feature of the SUR is the predicted orientation of the amino-terminal domain in the extracellular space, in the absence of an identifiable leader sequence. A cryptic cleavage site may explain why the mass of 177 kilodaltons predicted from the SUR cDNA sequence differs markedly from the apparent mass of 140 kilodaltons for the photolabeled protein, as the similar MRP protein has a highly sensitive carboxyl-proteolysis site (15, 16).

The report by Thomas and co-workers shows that a well-known, if rare, disease of newborns, familial hyperinsulinemic hypoglycemia of infancy (FHHI), is associated in several families with mutations in the second nucleotide-binding domain of SUR, suggesting that this fold may have something to do with insulin secretion by analogy with a similar CFTR mutation (7, 17). The mechanism of this defect, a putative increase in the ability of ATP to close K_ATP channels, suggests that the cells would be chronically depolarized and therefore hypersecretory. If the SUR mutations actually cause a loss of open K_ATP channels, this explains why the syndrome is autosomal recessive, because only a few open K_ATP channels are required to set the resting membrane potential (18). [It also explains the well-known leucine sensitivity of these patients, because leucine stimulates cell oxidative phosphorylation, which would produce more ATP to further depolarize the membrane (19).] Gene therapy may therefore afford an interesting opportunity for treatment of FHHI. Mental retardation in FHHI, ascribed to repeated hypoglycemia, should now be reconsidered to be complicated by defective SUR expression in the nervous system, and cardiac abnormalities might be uncovered in these patients as well. Insulin hypersecretion due to an SUR mutation raises the possibility that insulin hyposecretion syndromes related to defects in SUR might also exist, although diabetes genes have not been localized to the region of FHHI.

Whither K_ATP? Inward rectifier-related channels and SUR certainly seem destined for each other, assuming the SUR is not itself a channel, but the right match remains elusive. An exhaustive combinatorial approach with pairs of inward rectifier-like channels may be illuminating. Additional bridging subunits may yet turn up to couple sulfonylurea binding to channel inhibition. The isolation of cDNAs for these putative membrane proteins will, we sense, fuel future studies of metabolic coupling to cell excitability.

References

1. P. Dean and E. Mathews, Nature 219, 389 (1968); J. C. Henquin and H. P. Meissner, Experentia 40, 1043 (1984); J. F. Worley III et al., J. Biol. Chem. 269, 14359 (1994).

2. C. B. Wollheim and G. W. G. Sharp, Physiol. Rev. 61, 914 (1981); A. E. Boyd III, Diabetes 37, 847 (1988); P. Gilon, R. M. Shepherd, J. C. Henquin, J. Biol. Chem. 268, 22265 (1993).

3. S. J. H. Ashcroft and F. M. Ashcroft, Cell. Signalling 2, 197 (1990).

4. B. E. Corkey et al., J. Biol. Chem. 263, 4254 (1988); I. D. Dukes et al., ibid. 269, 10979 (1994).

5. T. D. Plant, J. Physiol. (London) 404, 731 (1988); W. Pralong, C. Bartley, C. B. Wollheim, EMBO J. 9, 53 (1990); M. Valdeolmillos et al., FEBS Lett. 259, 19 (1989); M. W. Roe et al., J. Biol. Chem. 268, 9953 (1993).

6. N. C. Sturgess et al., Lancet 2, 474 (1985).

7. P. M. Thomas et al., Science 268, 426 (1995); L. Aguilar-Bryan et al., ibid., p. 423.

8. A. Noma, Nature 305, 147 (1983).

9. F. M. Ashcroft, Annu. Rev. Neurosci. 11, 97 (1988); M. L. J. Ashford, P. R. Boden, J. M. Treherne, Br. J. Pharmacol. 101, 531 (1990); A. D. Beavis, Y. Lu, K. D. Garlid, J. Biol. Chem. 268, 997 (1993).

10. J. R. de Weille et al., Pfluegers Arch. 414, s80-s87 (1989); M. Gopalakrishnan, R. A. Janis, D. J. Triggle, Drug Dev. Res. 28, 95 (1993).

11. Y. Kubo et al., Nature 362, 127 (1993); Y. Kubo et al., ibid. 364, 802 (1993); K. Ho et al., ibid. 362, 31 (1993); M. L. J. Ashford et al., ibid. 370, 456 (1994); F. Lesage et al., FEBS Lett. 353, 37 (1994). N. Inagaki et al., J. Biol. Chem., in press.

12. K. M. Hurley et al., Biophys J. 68, A448 (1995); M. Pessia et al., ibid., p. A31; Yang et al., ibid., p. A362; N. Dascal et al., ibid., p. A362; K. W. Chan et al., ibid., p. A353.

13. G. Krapivinsky et al., Nature 374, 135 (1995).

14. S. P. Cole et al., SCIENCE 258, 1650 (1992); M. Drumm and F. Collins, Mol. Gen. Med. 3, 33 (1993).

15. H. Bernardi, M. Fosset, M. Lazdunski, Proc. Natl. Acad. Sci. U.S.A. 85, 9816 (1988); K. L. Gaines, S. Hamilton, A. E. Boyd, III, J. Biol. Chem. 263, 2589 (1988); D. A. Nelson, L. Aguilar-Bryan, J. Bryan, ibid. 267, 14928 (1992); S. E. Ozanne et al., Diabetologia 38, 277 (1995).

16. N. Krishnamachary et al., Oncology Res. 6, 119 (1994).

17. L. Smit et al., Proc. Natl. Acad. Sci. U.S.A. 90, 9963 (1993).

18. D. L. Cook et al., Diabetes 37, 495 (1988).

19. M. J. MacDonald et al., Arch. Biochem. Biophys. 269, 400 (1989).