IP6 | $29.95 120 Capsules, 500 mg | Order |
Version 3.1 Updated: July 5 1999
What are the pathways of InsP6 synthesis and metabolism? How are they regulated? Where in the cell is InsP6 located? Does InsP6 have some biological function inside cells, or does it serve only as a precursor reservoir pool for the diphosphorylated polyphosphates? Although some of these questions may appear rather simple, they have in fact proved rather difficult to resolve.
Take, for example, the problem of determining the subcellular localization of InsP6. There is more than one good reason for doubting that all the cellular InsP6 is free in the cytosol. For example, there is the well-known tendency of InsP6 to form insoluble complexes with divalent cations in vitro (1). This property does have its biological uses. In plant seeds, insoluble InsP6-cation complexes are deposited as globular inclusions in membrane-bound storage bodies (for reviews, see (2;3)6 may comprise 50-80% of plant seed total phosphorous (2;3)6 that is in the corn they are fed with; the InsP6 they excrete is a major source of phosphorous pollution that occurs during leaching and "runoff" from soil treated with excess animal wastes. This problem is driving attempts to generate corn mutants with decreased InsP6 levels (4).
One group of animals is known to contain InsP6-cation deposits; they may comprise half the weight of the mature dispersal larvae of most species of dicyemid mesozoans. This appears to be another example of a rather specialized function for InsP6. The deposits may impart negative bouyancy that may keep these minute parasites close to the sea bottom where they can encounter their invertebrate hosts (5). Since there is no evidence for precipitates of InsP6 in any other animal, how could this be prevented? It has been suggested that InsP6 might be sequestered into a vesicular compartment with an ionic composition that might be more compatible with the polyphosphate remaining in solution (1). However, after a careful search for such a compartment, it was concluded that one probably did not exist (6). Instead, there is evidence from in vitro studies that much of the cell's InsP6 is "wallpapered" to cellular membranes (7). The polyphosphate is proposed to be held in place by the formation of an electrostatically bonded InsP6 - cation - phospholipid sandwich (7). Consistent with this idea, the binding of InsP6 to cellular membranes is disrupted by chelation of divalent and trivalent cations by EDTA (6;7)6 exists in a complex with cations in vivo, but in a manner that does not lead to any precipitation.
Cation-binding properties of InsP6 have been suggested to have additional physiologically-important uses: In vivo at least, the remarkable affinity of InsP6 for iron totally inhibits this metal's ability to catalyze the formation of hydroxyl radicals (8) . Thus, it has been suggested that a physiological function of InsP6 is to transport iron within the cell in a form that protects against the potentially lethal consequences of free radical formation (8). This proposal extends an earlier observation that InsP6 is a powerful anti-oxidant in vitro (9). Although these chemical properties of InsP6 are indisputable, there is no direct evidence that the polyphosphate either transports iron or acts as an antioxidant in vivo. Ion-chelation and anti-oxidant properties of InsP6 have also been suggested to contribute to an apparent anti-neoplastic effect of InsP6, when administered in high (mM) concentrations in a number of model systems (10;11)6 to cancer prevention in humans (12). Also, if any anti-neoplastic benefits do result from an increase in dietary InsP6, they could be offset by the ion-chelating properties of this polyphosphate leading to a decrease in mineral availability (12). The binding of Ca2+ to InsP6 has also been suggested to be advantageous, since this association can block the formation of crystals of calcium oxalate and calcium phosphate (13). It has therefore been proposed that InsP6 is excreted in urine (in amounts that are in direct proportion to its dietary intake), where it is suggested to prevent kidney stone formation (13). However, the published descriptions of InsP6 being a constituent of urine are not conclusive: samples of urine have been batch-chromatographed on ion-exchange resin, under conditions where any InsP6 would have been separated from inorganic phosphate (14). Yet the "InsP6 fraction" could have contained many other phosphorylated compounds. At the very least, it would be useful to further purify and analyze this "InsP6 fraction", perhaps by dephosphorylating it and demonstrating that inositol is recovered, in an appropriate proportion to the amount of Pi released.
Total cellular levels of InsP6 might be sufficient to provide an average concentration of up to100 µM (15;16)6 may be bound to membranes (see above), and to proteins (see below), it is not clear how much is left to be freely soluble. This lack of a good understanding of the free cellular InsP6 concentration creates great uncertainty concerning what is the relevance in vivo of some effects of InsP6 obtained in vitro, particularly those that require µM levels.
There also continues to be some dispute over what is the pathway of InsP6 synthesis. This problem has been most carefully studied in model systems outside the animal kingdom: in the slime-mould, Dictyostelium (17), in yeasts (18;19)5 was identified as the immediate precursor of InsP6. The rate of InsP6 synthesis in S. pombe is considerably accelerated during osmotic stress (18); it will be of interest to understand the significance of this response. In plants, at least, it has been suggested the Ins(1,3,4,5,6)P5 2-kinase may act "in reverse" to directly support ATP synthesis (21). In support of this idea, the reaction catalyzed by this enzyme (isolated from soybean seeds) has been determined to have an equilibrium constant of 14 (22). Ins(1,3,4,5,6)P5 2-kinase activity in Saccharomyces cerevisiae was recently found to be encoded by the YDR315C gene product (19).
Two published studies with animal cells have also found them to contain an Ins(1,3,4,5,6)P5 2-kinase (23;24)5 isomers, albeit in relatively small quantities (23). There is also a bewildering array of InsP5 kinase activities: Ins(2,3,4,5,6)P5 1-kinase (23), Ins(1,2,4,5,6)P5 3-kinase (23;25-27)5 5-kinase (23) and D/L-Ins(1,2,3,4,6)P5 4/6-kinase (23). Thus, in principle, any InsP5 isomer might be the de novo precursor for InsP6. However, there is no evidence in animal systems for the existence of the necessary InsP4 kinases that could synthesize these "alternative" InsP5 isomers (i.e. those with a 2-phosphate). Only Ins(1,3,4,5,6)P5 has the benefit of being in an established metabolic pathway by which it can be the de novo precursor for InsP6. In contrast, all the other InsP5 isomers are only known to be formed by dephosphorylation of InsP6 itself (28), through the catalytic activity of the Multiple Inositol Polyphosphate Phosphatase ("MIPP" - see the Section that deals with this enzyme).
The presence in cells of MIPP and multiple InsP5 kinases suggests that cells contain a complex network of substrate cycles that interconvert InsP5 and InsP6. The classical interpretation of this situation would be that it provides the cell with a particularly sensitive means of regulating the levels of these metabolites (29). While flux through multiple substrate cycles would seem to be a particularly extreme means of regulating InsP6/InsP5 interconversion, some control over InsP6 metabolism may be particularly important at key stages in the cell cycle, or during differentiation (15;30-32)6 dephosphorylation is regulated.
A large number of laboratories have been drawn to the idea that InsP6 might be a cellular signal. However, one factor that has counted against this concept is that cellular levels of InsP6 - at least in the short term - are, at best, only slightly affected by extracellular stimulii. For example, in HL-60 cells, a 10% increase in InsP6 mass was observed upon stimulation with chemotactic peptide (33). Most other workers have tended only to assay [3H]InsP6 levels indirectly, in [3H]inositol-labeled cells. Even in those few, isolated cases where effects of a reasonable size have been seen (30% increases in [3H]InsP6 levels in carbachol-stimulated neuroblastoma cells (34); 56% increase in [3H]InsP6 levels in cerebellar granule cells upon depolarization with high [K+] (35)), the data cannot be considered definitive. The problem with such experiments is that the InsP6 pool incorporates [3H]inositol so slowly that it may take more than a week to label to equilibrium (36). Unless equilibrium labeling can be demonstrated, it is very possible that it is specific radioactivity and not mass that is responding to cell stimulation. Arguably the most careful effort to define an acute stimulus-dependent change in [3H]InsP6 levels was published only recently (37). Even then, after a 7 day radiolabeling protocol, the effect was quite small; levels of [3H]InsP6 in glucose-challenged pancreatic beta-cells increased by a little over 10% (37). Of course, it is possible that stimulus-dependent effects upon InsP6 might be more subtle than we can currently measure. For example, a large local change in a metabolic sub-pool of InsP6 could be missed in global InsP6 assays (see (38) for an apparent example of this phenomenon). It is also conceivable that extracellular agonists might target the free cytosolic InsP6 concentration without there being much of a change in total cellular InsP6 levels (analogous to the receptor-dependent increases in cytosolic Ca2+ which occur without substantial impact upon total cellular [Ca2+]). For example, perhaps the free InsP6 concentration is manipulated by regulation of its distribution between cellular membranes and the cytosol.
InsP6 has been reported to have a number of biochemical and physiological effects in vitro, but considerable caution is required when interpreting them. One problem is that InsP6 has considerable prowess as a chelator of divalent and trivalent cations (39;40)6 was probably responsible for what was originally thought to be an action of InsP6 as a neurotransmitter (41). Subsequent recognition that ion-chelation is a non-specific action of InsP6 (39;40)6 actions in vitro is that dose/response curves are of limited value. We know that total cellular InsP6 is around 15-100 µM, (15;16;42)6 is almost certainly not uniformly distributed throughout the cell (see above). It is therefore difficult to predict its concentration at a candidate effector site in vivo. Thus, in experiments with InsP6, it is particularly crucial to demonstrate specificity. Any response to InsP6 should be shown to have a preference for that polyphosphate, and other inositol phosphates should be shown to be less active.
InsP6 is, of course, highly phosphorylated; this is also a problem for investigators, as it can influence protein function by low affinity, non-specific charge effects that may not exist in vivo, where they can be neutralized by divalent and trivalent cations (43). One way to exclude this possibility is to show that an effect of InsP6 in vitro is neither reversed by adding excess divalent cations, nor imitated by the non-physiological analogue which also has high negative charge density, namely, inositol hexasulphate (InsS6)(43). Unfortunately, these negative controls are only useful when they yield just such a result. Instead, when InsS6 mimics an effect of InsP6, this outcome is equivocal, such as was the case when both compounds activated L-type Ca2+ channels and insulin secretion in the H1T T15 insulinoma (37;44)6 was imitating a physiological or non-physiological effect of InsP6. In such circumstances, it is especially important to determine if the effects of InsP6 are preserved in a physiologically-relevant ionic milieu. This proposed action of InsP6 upon Ca2+ channels has been attributed to inhibition of protein phosphatase activity; InsP6, with IC50 values of 4-13 µM, inhibited three different species of serine/threonine protein phosphatases in vitro (37). It is difficult to envisage how this effect could have a specific signalling consequence. It is a broad spectrum effect upon the activities of several different types of protein phosphatases, and was not even specific for InsP6, since Ins(1,3,4,5,6)P5 acted with nearly the same potency (37).
A recent study has indicated that InsP6 may have a role in regulating mRNA export from the nucleus (19). A genetic screen revealed three new mutants with mRNA export deficiencies that also had in common an impaired ability to synthesize InsP6. The defective gene in one of these mutants expressed Ins(1,3,4,5,6)P5 2-kinase activity, an enzyme which immunofluorescence microscopy revealed to be concentrated at the nuclear periphery (19). One route forward may now be to determine whether InsP6 modifies the functions of any of the proteins that regulate mRNA export, although, as mentioned above, this approach will also necessitate that any regulatory potential for InsP6 be placed in a context whereby changes in its cellular concentration can be seen to have an impact. Another possibility is that the requirement for InsP6 is as a precursor for the diphosphoinositol polyphosphates, which may actually be the functionally active compounds.
Another approach that has been taken in order to determine if InsP6 is physiologically active has been to isolate and characterize proteins which preferentially bind this polyphosphate. Several such proteins have been discovered which have ligand affinities in the nanamolar range: InsP6 binds tightly to vinculin, a component of platelet cytoskeleton (45) and to myelin proteolipid protein, which participates in myelin deposition (46) and to coatomer, which regulates vesicle traffic between the ER and the Golgi (47;48)6 was nearly 200-fold weaker a ligand for coatomer than was InsP6 (48), suggesting (see above) that the latter's effect is of some significance. However, there is little information in any of these studies to suggest that ligand binding alters the function of any of these target proteins.
Recombinant synaptotagmin has also been reported to bind inositol phosphates, with InsP6 being the most potent of the tested ligands (49). Type I synaptotagmin is part of the synaptic vesicle complex, and appears to promote exocytosis by "sensing" local changes in [Ca2+] and by interacting with several other proteins (50). There are also some non-neuronal synaptotagmins that share this inositol phosphate binding ability (51). It appears that InsP6 and inositol lipids compete for the same ligand binding site; an IC50 value of 10 µM describes InsP6 competition with PtdIns(4,5)P2 binding to recombinant synaptotagmin (52). The binding of synaptotagmin to PtdIns(4,5)P2 in the plasma membrane may help docking and fusion of exocytic vesicles. Perhaps InsP6 antagonizes this process. Another way InsP6 may inhibit vesicle docking is by preventing an association of synaptotagmin with AP-2 (53). However, it should be noted that, unlike recombinant synaptotagmin, preparations of native synaptotagmin do not have such a high affinity for InsP6, and instead prefer to bind Ins(1,3,4,5)P4 and Ins(1,3,4,5,6)P5 (54). Likewise, InsP6 is at least 10-fold less effective at competing PtdIns(4,5)P2 from native synaptotagmin than is the case with the recombinant protein (52). Are these data telling us that the high affinity of recombinant synaptotagmin for InsP6 may not be physiologically relevant? Or alternatively, is it possible that the ligand binding differences between the native and recombinant proteins are a manifestation of a physiologically-relevant regulatory process, covalent modification for example, that can "switch" ligand affinity and/or specificity? This could be a fruitful area for future research.
Studies with two other InsP6-binding proteins, AP-2 (55-57)6 binding to these "adaptor" proteins is inhibition of clathrin cage assembly (55;58;59)6 has been shown not to substitute for the ability of InsP6 to inhibit AP-3 mediated clathrin assembly (59). This has led to the suggestion that InsP6 might be a general antagonist (or "clamp") of endocytic vesicle traffic (60). Demonstrations that microinjection of InsP6 into cells can inhibit vesicle traffic (61) are consistent with the participation in this process of an inositide-binding site. However, such experiments do not prove that this is a physiologically relevant function of InsP6 itself, rather than some other inositide. There is also little evidence (see above) that short-term, acute changes in InsP6 levels, such as those imposed by the microinjection protocol, have any physiological relevance.
There is more scope for believing that InsP6 has some longer-term regulatory function as a "clamp". Cellular mass levels of InsP6 do rise and fall quite substantially at certain points in the cell-cycle and during cellular differentiation (15;30-32)6 levels occur at a time when there are dramatic alterations in the rates of many vesicle trafficking processes, and a redistribution of vesicular organelles (62). There is a need for more experiments that might probe the physiological significance of such changes. The inositol lipid, PtdIns(3,4,5)P3, has recently been found to be more potent than InsP6 at inhibiting clathrin assembly by AP-2 (63) and AP-3 (64). However, cellular levels of PtdIns(3,4,5)P3 are extremely low in the absence of receptor-coupled activation of tyrosine kinases, so it is under these conditions where a role for InsP6 still seems possible.
1. Irvine, R.F., Moor, R.M., Pollock, W.K., Smith, P.M., and Wreggett, K.A. (1988) Philos.Trans.R.Soc.Lond.[Biol] 320, 281-298
2. Gibson, D.M. and Ullah, A.B.J. (1990) in Inositol metabolism in plants (Morre, D.J., Boss, W.F., and Loewus, F.A., eds) pp. 77-92, Wiley-Liss, New York
3. Raboy, V. (1990) in Inositol metabolism in plants (Morre, D.J., Boss, W.F., and Loewus, A.L., eds) pp. 55-76, Wiley-Liss, New York
4. Ertl, D.S., young, k.a., and Raboy, V. (1998) J.Environ.Qual. 27, 299-304
5. Lapan, E.A. (1975) Exp.Cell Res. 94, 277-282
6. Stuart, J.A., Anderson, K.L., French, P.J., Kirk, C.J., and Michell, R.H. (1994) Biochem.J. 303, 517-525
7. Poyner, D.R., Cooke, F., Hanley, M.R., Reynolds, D.J.M., and Hawkins, P.T. (1993) J.Biol.Chem. 268, 1032-1038
8. Hawkins, P.T., Poyner, D.R., Jackson, T.R., Letcher, A.J., Lander, D.A., and Irvine, R.F. (1993) Biochem.J. 294, 929-934
9. Graf, E. and Empson, K.L. (1987) J.Biol.Chem. 262, 11647-11650
10. Vucenik, I. and Shamsuddin, A.M. (1994) j.nutr. 124, 861-868
11. Shamsuddin, A.M. and Yang, G.-Y. (1995) Carcinogen. 16, 1975-1979
12. Zhou, J.R. and Erdman, J.W. (1995) Crit.Rev.Food Sci.Nutr. 35, 495-508
13. Grases, F., Garcia-Gonzalez, R., Torres, J.j., and Llobera, A. (1998) Scand.J.Urol.Nephrol. 32, 261-265
14. March, J.G., Simonet, B.m., Grases, F., and Salvador, A. (1996) Analytica Chimica Acta 367, 63-68
15. French, P.J., Bunce, C.M., Stephens, L.R., Lord, J.M., McConnell, F.M., Brown, G., Creba, J.A., and Michell, R.H. (1991) Philos.Trans.R.Soc.Lond.[Biol] 245, 193-201
16. Bunce, C.M., French, P.J., Allen, P., Mountford, J.C., Moor, B., Greaves, M.F., Michell, R.H., and Brown, G. (1993) Biochem.J. 289, 667-673
17. Stephens, L.R. and Irvine, R.F. (1990) Nature 346, 580-583
18. Ongusaha, P.P., Hughes, P.J., Davey, J., and Michell, R.H. (1998) Biochem.J. 335, 671-679
19. York, J.D., Odom, A.R., Murphy, R., Ives, E.B., and Wente, S.R. (1999) Science 285, 96-100
20. Brearley, C.A. and Hanke, D.E. (1996) Biochem.J. 314, 227-233
21. Biswas, S., Maity, I.B., Chakrabarti, S., and Biswas, B.B. (1978) Arch.Biochem.Biophys. 185, 557-566
22. Phillippy, B.Q., Ullah, A.H.J., and Ehrlich, K.C. (1994) J.Biol.Chem. 269, 28393-28399
23. Stephens, L.R., Hawkins, P.T., Stanley, A.F., Moore, T., Poyner, D.R., Morris, P.J., Hanley, M.R., Kay, R.R., and Irvine, R.F. (1991) Biochem.J. 275, 485-499
24. Ji, H., Sandberg, K., Baukal, A.J., and Catt, K.J. (1989) J.Biol.Chem. 264, 20185-20188
25. Rudolf, M.T., Kaiser, T., Guse, A.H., Mayr, G.W., and Schultz, C. (1997) Liebigs Ann./Recueil 9, 1861-1869
26. Guse, A.H. and Emmrich, F. (1991) J.Biol.Chem. 266, 24498-24502
27. Craxton, A., Erneux, C., and Shears, S.B. (1994) J.Biol.Chem. 269, 4337-4342
28. Nogimori, K., Hughes, P.J., Glennon, M.C., Hodgson, M.E., Putney, J.W., Jr., and Shears, S.B. (1991) J.Biol.Chem. 266, 16499-16506
29. Crabtree, B. and Newsholme, E.A. (1985) Curr.Top.Cell.Regul. 25, 21-76
30. Barker, C.J., French, P.J., Moore, A.J., Nilsson, T., Berggren, P.-O., Bunce, C.M., Kirk, C.J., and Michell, R.H. (1995) Biochem.J. 306, 557-564
31. Guse, A.H., Greiner, E., Emmrich, F., and Brand, K. (1993) J.Biol.Chem. 268, 7129-7133
32. Balla, T., Sim, S.S., Baukal, A.J., Rhee, S.G., and Catt, K.J. (1994) Molecular Biology of the Cell 5, 17-28
33. Pittet, D., Lew, D.P., Mayr, G.W., Monod, A., and Schlegel, W. (1989) J.Biol.Chem. 264, 7251-7261
34. Sasakawa, N., Nakaki, T., Kashima, R., Kanba, S., and Kato, R. (1992) J.Neurochem. 58, 2116-2123
35. Sasakawa, N., Nakaki, T., Kakinuma, E., and Kato, R. (1993) brain research 623, 155-160
36. Michell, R.H., King, C.E., Piper, C.J., Stephens, L.R., Bunce, C.M., Guy, G.R., and Brown, G. (1988) J.Gen.Physiol. 43, 345-355
37. Larsson, O., Barker, C.J., Sjöholm, A., Carlqvist, H., Michell, R.H., Bertorello, A., Nilsson, T., Honkanen, R.E., Mayr, G.W., Zwiler, J., and Berggren, P.-O. (1997) Science 278, 471-474
38. Singh, J., Hunt, P., Eggo, M.C., Sheppard, M.C., Kirk, C.J., and Michell, R.H. (1996) Biochem.J. 316, 175-182
39. Sasakawa, N., Sharif, M., and Hanley, M.R. (1995) Biochem.Pharmacol. 50, 137-146
40. Sun, M., Wahlestedt, C., and Reis, D.J. (1992) Eur.J.Pharmacol. 215, 9-16
41. Vallejo, M., Jackson, T., Lightman, S., and Hanley, M.R. (1987) Nature 330, 656-658
42. Szwergold, B.S., Graham, R.A., and Brown, T.R. (1987) Biochem.Biophys.Res.Commun. 149, 874-881
43. Palmer, R.H., Lodewijk, V.D., Woscholski, R., Le Good, J.A., Gigg, R., and Parker, P.J. (1995) J.Biol.Chem. 270, 22412-22416
44. Efanov, A.M., Zaitsev, S.V., and Berggren, P.-O. (1997) Proc.Nat.Acad.Sci.USA 94, 4435-4439
45. O'Rouke, F., Matthews, E., and Feinstein, M.B. (1996) Biochem.J. 315, 1027-1034
46. Yamaguchi, Y., Ikenaka, K., Niinobe, M., Yamada, H., and Mikoshiba, K. (1996) J.Biol.Chem. 271, 27838-27846
47. Fleischer, B., Xie, J., Mayrleitner, M., Shears, S.B., Palmer, D.J., and Fleischer, S. (1994) J.Biol.Chem. 269, 17826-17832
48. Ali, N., Duden, R., Bembenek, M.E., and Shears, S.B. (1995) Biochem.J. 310, 279-284
49. Fukuda, M., Aruga, J., Niinobe, M., Aimoto, S., and Mikoshiba, K. (1994) J.Biol.Chem. 269, 29206-29211
50. Südhof, T.C. (1995) Nature 375, 645-653
51. Ibata, K., Fukuda, M., and Mikoshiba, K. (1998) J.Biol.Chem. 273, 12267-12273
52. Schiavo, G., Gu, Q.-M., Prestwich, G.D., Söllner, T.H., and Rothman, J.E. (1996) Proc.Nat.Acad.Sci.USA 93, 13327-13332
53. Mizutani, A., Fukuda, M., Niinobe, M., and Mikoshiba, K. (1998) Biochem.Biophys.Res.Commun. 240, 128-131
54. Niinobe, M., Yamaguchi, Y., Fukuda, M., and Mikoshiba, K. (1994) Biochem.Biophys.Res.Commun. 205, 1036-1042
55. Beck, K.A. and Keen, J.H. (1991) J.Biol.Chem. 266, 4442-4447
56. Timerman, A.P., Mayrleitner, M.M., Lukas, T.J., Chadwick, C.C., Saito, A., Watterson, D.M., Schindler, H., and Fleischer, S. (1992) Proc.Nat.Acad.Sci.USA 89, 8976-8980
57. Voglmaier, S.M., Keen, J.H., Murphy, J.-E., Ferris, C.D., Prestwich, G.D., Snyder, S.H., and Theibert, A.B. (1992) Biochem.Biophys.Res.Commun. 187, 158-163
58. Norris, F.A., Ungewickell, E., and Majerus, P.W. (1995) J.Biol.Chem. 270, 214-218
59. Ye, W., Ali, N., Bembenek, M.E., Shears, S.B., and Lafer, E.M. (1995) J.Biol.Chem. 270, 1564-1568
60. Shears, S.B. (1996) in myo-Inositol phosphates, phosphoinositides and signal transduction (Biswas, B.B. and Biswas, S., eds) pp. 187-226, Plenum Press, New York and London
61. Llinįs, R., Sugimori, M., Lang, E.J., Morita, M., Fukuda, M., Niinobe, M., and Mikoshiba, K. (1994) Proc.Nat.Acad.Sci.USA 91, 12990-12993
62. Warren, G. and Wickner, W. (1996) Cell 84, 395-400
63. Gaidarov, I., Chen, Q., Falck, J.R., Reddy, K.K., and Keen, J.H. (1996) J.Biol.Chem. 271, 20922-20929
64. Hao, W., Tan, Z., Prasad, K., Reddy, K.K., Chen, J., Prestwich, G.D., Falck, J.R., Shears, S.B., and Lafer, E.M. (1997) J.Biol.Chem. 272, 6393-6398