Growth factors stimulate the Na-K-2Cl cotransporter NKCC1 through a novel Cl--dependent mechanism

The Na-K-2Clcotransporter NKCC1 is an important volume-regulatory transporter thatis regulated by cell volume and intracellular Cl. Thisregulation appears to be mediated by phosphorylation of NKCC1, althoughthere is evidence for additional, cytoskeletal regulation via myosinlight chain (MLC) kinase. NKCC1 is also activated by growth factors andmay contribute to cell hypertrophy, but the mechanism is unknown. Inaortic endothelial cells, NKCC1 (measured as bumetanide-sensitive⁸⁶Rb⁺ influx) was rapidly stimulated by serum,lysophosphatidic acid, and fibroblast growth factor, with the greateststimulation by serum. Serum increased bumetanide-sensitive influxsignificantly more than bumetanide-sensitive efflux (131% vs. 44%),indicating asymmetric stimulation of NKCC1, and produced a 17%increase in cell volume and a 25% increase in Cl contentover 15 min. Stimulation by serum and hypertonic shrinkage wereadditive, and serum did not increase phosphorylation of NKCC1 or MLC,and did not decrease cellular Cl content. When cellularCl was replaced with methanesulfonate, influx via NKCC1increased and was no longer stimulated by serum, whereas stimulation by hypertonic shrinkage still occurred. Based on these results, we proposea novel mechanism whereby serum activates NKCC1 by reducing itssensitivity to inhibition by intracellular Cl. Thisresetting of the Cl set point of the transporter enablesthe cotransporter to produce a hypertrophic volume increase.

     

Effect of the Na-K-2Cl cotransporter NKCC1 on systemic blood pressure and smooth muscle tone

Studies in rat aorta have shown that the Na-K-2Cl cotransporter NKCC1 is activated by vasoconstrictors and inhibited by nitrovasodilators, contributes to smooth muscle tone in vitro, and is upregulated in hypertension. To determine the role of NKCC1 in systemic vascular resistance and hypertension, blood pressure was measured in rats before and after inhibition of NKCC1 with bumetanide. Intravenous infusion of bumetanide sufficient to yield a free plasma concentration above the IC(50) for NKCC1 produced an immediate drop in blood pressure of 5.2% (P < 0.001). The reduction was not prevented when the renal arteries were clamped, indicating that it was not due to a renal effect of bumetanide. Bumetanide did not alter blood pressure in NKCC1-null mice, demonstrating that it was acting specifically through NKCC1. In third-order mesenteric arteries, bumetanide-inhibitable efflux of (86)Rb was acutely stimulated 133% by phenylephrine, and bumetanide reduced the contractile response to phenylephrine, indicating that NKCC1 influences tone in resistance vessels. The hypotensive effect of bumetanide was proportionately greater in rats made hypertensive by a 7-day infusion of norepinephrine (12.7%, P < 0.001 vs. normotensive rats) but much less so when hypertension was produced by a fixed aortic coarctation (8.0%), again consistent with an effect of bumetanide on resistance vessels rather than other determinants of blood pressure. We conclude that NKCC1 influences blood pressure through effects on smooth muscle tone in resistance vessels and that this effect is augmented in hypertension.     

Na-K-2Cl cotransporter inhibition impairs human lung cellular proliferation

The widespread presence of the Na-K-2Cl (NKCC) cotransporter protein suggests that chronic administration of inhibitors may result in adverse effects. Inhibition of the NKCC cotransporter by loop diuretics is felt to underlie the diuretic and the pulmonary smooth muscle relaxant effects of this drug class. However, the fundamental regulation of salt and water movement by this cotransporter suggests that it may also mediate cell volume changes occurring during cell cycle progression. Thus we hypothesized that NKCC cotransporter inhibition by loop diuretics would decrease cellular proliferation. Normal human bronchial smooth muscle cells (BSMC) showed a significant concentration-dependent decrease in cell counts after 7 days of exposure to both bumetanide (n=5-10) and furosemide (n=6-16) compared with controls. Proliferation was similarly inhibited in normal human lung fibroblasts (n=5-9). To determine whether this was due to loss of cells, we performed apoptosis assays on BSMC. Both annexin V-propidium iodide staining (n=5-10) and single cell gel electrophoresis assays (n=4) were negative for necrosis and apoptosis in BSMC exposed to 10 microM bumetanide. Subsequent analysis of the cell cycle by flow cytometry showed that bumetanide-exposed BSMC were delayed in G1 phase compared with controls (n=4-8). This is the first evidence for loop diuretic inhibition of airway smooth muscle cell proliferation. NKCC cotransporter inhibition impeded G1-S phase transition without facilitating cell death. Thus although inhibition by loop diuretics relaxes airway smooth muscle, the NKCC cotransporter may have a more important role in cell proliferation regulation.     

Kinetics of hyperosmotically stimulated Na-K-2Cl cotransporter in Xenopus laevis oocytes

A detailed study of hypertonically stimulated Na-K-2Cl cotransport (NKCC1) in Xenopus laevis oocytes was carried out to better understand the 1 K(+):1 Cl(-) stoichiometry of transport that was previously observed. In this study, we derived the velocity equations for K(+) influx under both rapid equilibrium assumptions and combined equilibrium and steady-state assumptions and demonstrate that the behavior of the equations and curves in Lineweaver-Burke plots are consistent with a model where Cl(-) binds first, followed by Na(+), a second Cl(-), and then K(+). We further demonstrate that stimulation of K(+) movement by K(+) on the trans side is an intrinsic property of a carrier that transports multiple substrates. We also demonstrate that K(+) movement through NKCC1 is strictly dependent upon the presence of external Na(+), even though only a fraction of Na(+) is in fact transported. Finally, we propose that the larger transport of K(+), as compared with Na(+), is a result of the return of partially unloaded carriers, which masks the net 1Na(+):1K(+):2Cl(-) stoichiometry of NKCC1. These data have profound implications for the physiology of Na-K-2Cl cotransport, since transport of K-Cl in some conditions seems to be uncoupled from the transport of Na-Cl.     

Characterization of SPAK and OSR1, regulatory kinases of the Na-K-2Cl cotransporter

Our recent studies demonstrate that SPAK (Ste20p-related Proline Alanine-rich Kinase), in combination with WNK4 [With No lysine (K) kinase], phosphorylates and stimulates the Na-K-2Cl cotransporter (NKCC1), whereas catalytically inactive SPAK (K104R) fails to activate the cotransporter. The catalytic domain of SPAK contains an activation loop between the well-conserved DFG and APE motifs. We speculated that four threonine residues (T231, T236, T243, and T247) in the activation loop might be sites of phosphorylation and kinase activation; therefore, we mutated each residue into an alanine. In this report, we demonstrate that coexpression of SPAK (T243A) or SPAK (T247A) with WNK4 not only prevented, but robustly inhibited, cotransporter activity in NKCC1-injected Xenopus laevis oocytes. These activation loop mutations produced an effect similar to that of the SPAK (K104R) mutant. In vitro phosphorylation experiments demonstrate that both intramolecular autophosphorylation of SPAK and phosphorylation of NKCC1 are significantly stronger in the presence of Mn2+ rather than Mg2+. We also show that SPAK activity is markedly inhibited by staurosporine and K252a, partially inhibited by N-ethylmaleimide and diamide, and unaffected by arsenite. OSR1, a kinase closely related to SPAK, exhibited similar kinase properties and similar functional activation of NKCC1 when coexpressed with WNK4.     

Phosphorylation and transport in the Na-K-2Cl cotransporters, NKCC1 and NKCC2A, compared in HEK-293 cells

Na-K-2Cl cotransporters help determine cell composition and volume. NKCC1 is widely distributed whilst NKCC2 is only found in the kidney where it plays a vital role reabsorbing 20% of filtered NaCl. NKCC2 regulation is poorly understood because of its restricted distribution and difficulties with its expression in mammalian cell cultures. Here we compare phosphorylation of the N-termini of the cotransporters, measured with phospho-specific antibodies, with bumetanide-sensitive transport of K(+) ((86)Rb(+)) (activity) in HEK-293 cells stably expressing fNKCC1 or fNKCC2A which were cloned from ferret kidney. Activities of transfected transporters were distinguished from those of endogenous ones by working at 37 °C. fNKCC1 and fNKCC2A activities were highest after pre-incubation of cells in hypotonic low-[Cl(-)] media to reduce cell [Cl(-)] and volume during flux measurement. Phosphorylation of both transporters more than doubled. Pre-incubation with ouabain also strongly stimulated fNKCC1 and fNKCC2A and substantially increased phosphorylation, whereas pre-incubation in Na(+)-free media maximally stimulated fNKCC1 and doubled its phosphorylation, but inhibited fNKCC2A, with a small increase in its phosphorylation. Kinase inhibitors halved phosphorylation and activity of both transporters whereas inhibition of phosphatases with calyculin A strongly increased phosphorylation of both transporters but only slightly stimulated fNKCC1 and inhibited fNCCC2A. Thus kinase inhibition reduced phosphorylation and transport, and transport stimulation was only seen when phosphorylation increased, but transport did not always increase with phosphorylation. This suggests phosphorylation of the N-termini determines the transporters' potential capacity to move ions, but final activity also depends on other factors. Transport cannot be reliably inferred solely using phospho-specific antibodies on whole-cell lysates.     

Immunolocalization of a Na-K-2Cl cotransporter in human tracheobronchial smooth muscle.

Inhibition of the Na-K-2Cl (NKCC) cotransporter by loop diuretics is associated with airway relaxation, but there has been no direct evidence for the expression of this protein in airway smooth muscle. Thus we hypothesized that a NKCC cotransporter is present and functional in airway smooth muscle cells. Monoclonal and polyclonal antibodies were used first to demonstrate the presence of a NKCC cotransporter protein in isolated human fetal trachea and normal human bronchial smooth muscle cells (BSMC) by Western blotting. The cotransporter protein was then localized by immunohistochemical staining to airway smooth muscle cells in culture and in situ. The localization was confirmed by indirect immunofluorescence and laser confocal microscopy in the BSMC. Cotransporter function in BSMC was also confirmed in vitro by bumetanide-mediated inhibition of rubidium uptake. Our present findings thus document the presence of a functional NKCC cotransporter in human airway smooth muscle, providing a basis for defining the role of this ion cotransporter in airway smooth muscle function.     

JNK is a volume-sensitive kinase that phosphorylates the Na-K-2Cl cotransporter in vitro

Cell shrinkagephosphorylates and activates the Na-K-2Cl cotransporter (NKCC1),indicating the presence of a volume-sensitive protein kinase. Toidentify this kinase, extracts of normal and shrunken aorticendothelial cells were screened for phosphorylation of NKCC1 fusionproteins in an in-the-gel kinase assay. Hypertonic shrinkage activateda 46-kDa kinase that phosphorylated anNH₂-terminal fusion protein, withweaker phosphorylation of a COOH-terminal fusion protein. Thiscytosolic kinase was activated by both hypertonic and isosmoticshrinkage, indicating regulation by cell volume rather than osmolarity.Subsequent studies identified this kinase as c-JunNH₂-terminal kinase (JNK).Immunoblotting revealed increased JNK activity in shrunken cells; therewas volume-sensitive phosphorylation ofNH₂-terminal c-Jun fusion protein;immunoprecipitation of JNK from shrunken cells but not normal cellsphosphorylated NKCC1 in gel kinase assays; and treatment of cells withtumor necrosis factor, a known activator of JNK, mimicked the effect ofhypertonicity. We conclude that JNK is a volume-sensitive kinase inendothelial cells that phosphorylates NKCC1 in vitro. This is the firstdemonstration of a volume-sensitive protein kinase capable ofphosphorylating a volume-regulatory transporter.

     

Transcriptional repression of human cad gene by hypoxia inducible factor-1alpha

De novo biosynthesis of pyrimidine nucleotides provides essential precursors for DNA synthesis and cell proliferation. The first three steps of de novo pyrimidine biosynthesis are catalyzed by a multifunctional enzyme known as CAD (carbamoyl phosphate synthetase-aspartate carbamoyltransferase-dihydroorotase). In this work, a decrease in CAD expression is detected in numerous cell lines and primary culture human stromal cells incubated under hypoxia or desferrioxamine (DFO)-induced HIF-1alpha accumulation. A putative hypoxia response element (HRE) binding matrix is identified by analyzing human cad-gene promoter using a bioinformatic approach. Promoter activity assays, using constructs harboring the cad promoter (-710/+122) and the -67/HRE fragment (25-bases), respectively, demonstrate the suppression of reporter-gene expression under hypoxia. Suppression of cad-promoter activity is substantiated by forced expression of wild-type HIF-1alpha but abolished by overexpression of dominant-negative HIF-1alpha. A chromatin immunoprecipitation assay provides further evidence that HIF-1alpha binds to the cad promoter in vivo. These data demonstrate that the cad-gene expression is repressed by HIF-1alpha, which represents a functional link between hypoxia and cell-cycle arrest.     

Reduced Na-K pump but increased Na-K-2Cl cotransporter in aorta of streptozotocin-induced diabetic rat

The activities of Na-K-ATPase and Na-K-2Cl cotransporter (NKCC1) were studied in the aorta, heart, and skeletal muscle of streptozotocin (STZ)-induced diabetic rats and control rats. In the aortic rings of STZ rats, the Na-K-ATPase-dependent (86)Rb/K uptake was reduced to 60.0 +/- 5.5% of the control value (P < 0.01). However, Na-K-ATPase activity in soleus skeletal muscle fibers of STZ rats and paired control rats was similar, showing that the reduction of Na-K-ATPase activity in aortas of STZ rats is tissue specific. To functionally distinguish the contributions of ouabain-resistant (alpha(1)) and ouabain-sensitive (alpha(2) and alpha(3)) isoforms to the Na-K-ATPase activity in aortic rings, we used either a high (10(-3) M) or a low (10(-5) M) ouabain concentration during (86)Rb/K uptake. We found that the reduction in total Na-K-ATPase activity resulted from a dramatic decrement in ouabain-sensitive mediated (86)Rb/K uptake (26.0 +/- 3.9% of control, P < 0.01). Western blot analysis of membrane fractions from aortas of STZ rats demonstrated a significant reduction in protein levels of alpha(1)- and alpha(2)-catalytic isoforms (alpha(1) = 71.3 +/- 9.8% of control values, P < 0.05; alpha(2) = 44.5 +/- 11.3% of control, P < 0.01). In contrast, aortic rings from the STZ rats demonstrated an increase in NKCC1 activity (172.5 +/- 9.5%, P < 0.01); however, in heart tissue no difference in NKCC1 activity was seen between control and diabetic animals. Transport studies of endothelium-denuded or intact aortic rings demonstrated that the endothelium stimulates both Na-K-ATPase and Na-K-2Cl dependent (86)Rb/K uptake. The endothelium-dependent stimulation of Na-K-ATPase and Na-K-2Cl was not hampered by diabetes. We conclude that abnormal vascular vessel tone and function, reported in STZ-induced diabetic rats, may be related to ion transport abnormalities caused by changes in Na-K-ATPase and Na-K-2Cl activities.     

A volume-sensitive protein kinase regulates the Na-K-2Cl cotransporter in duck red blood cells

When Na-K-2Cl cotransport is activated in duckred blood cells by either osmotic cell shrinkage, norepinephrine,fluoride, or calyculin A, phosphorylation of the transporter occurs ata common set of serine/threonine sites. To examine the kinetics andregulation of the activating kinase, phosphatase activity was inhibitedabruptly with calyculin A and the subsequent changes in transporterphosphorylation and activity were determined. Increases in fractionalincorporation of ³²P into thetransporter and uptake of ⁸⁶Rb bythe cells were closely correlated, suggesting that the phosphorylation event is rate determining in the activation process. Observed in thismanner, the activating kinase was 1)stimulated by cell shrinkage, 2)inhibited by cell swelling, staurosporine, orN-ethylmaleimide, and3) unaffected by norepinephrine orfluoride. The inhibitory effect of swelling on kinase activity wasprogressively relieved by calyculin A, suggesting that the kinaseitself is switched on by phosphorylation. The kinetics of activation bycalyculin A conformed to an autocatalytic model in which thevolume-sensitive kinase is stimulated by a product of its own reaction(e.g., via autophosphorylation).

     

A single binding motif is required for SPAK activation of the Na-K-2Cl cotransporter.

BACKGROUND: SPAK (Ste20p-related proline alanine-rich kinase) phosphorylates and activates NKCC1 (Na-K-2Cl cotransporter) in the presence of another serine/threonine kinase WNK4 (With No lysine (K)). However, whether or not the docking of SPAK to NKCC1 is a requirement for cotransporter activation has not been fully resolved. METHODS: We mutated both SPAK binding motifs in the amino-terminal tail of NKCC1 and tested the interaction between SPAK and NKCC1 using a semi in vivo yeast two-hybrid assay, (32)P-ATP in vitro phosphorylation assays, and (86)Rb(+) uptake (a K(+) congener) assays in heterologously expressed Xenopus laevis oocytes. We also used site-directed mutagenesis to identify the principle phospho-regulatory threonine residues in the amino-terminal tail of NKCC1. RESULTS: A single SPAK binding motif is necessary for isotonic NKCC1 activation. Mutation of the phenylalanine (F) residue within the motif abrogates binding and function. Phosphorylation of the cotransporter is markedly reduced in the absence of SPAK docking to NKCC1. Truncations of internal regions of the amino-terminus of NKCC1 do not disrupt protein structure enough to affect cotransporter function. Threonine residues (T(206) and T(211)) are both identified as phospho-regulatory sites of NKCC1 function. CONCLUSION: We demonstrate that physical docking of SPAK to NKCC1 is necessary for cotransporter activity under both baseline and hyperosmotic conditions. We identify T(206) and T(211) as major phospho-acceptor sites involved in cotransporter function, with T(206) common to two separate regulatory pathways: one involving SPAK, the other involving a still unknown kinase that is responsive to forskolin/PKA stimulation.     

Blood pressure regulates the activity and function of the Na-K-2Cl cotransporter in vascular smooth muscle

The Na-K-2Cl cotransporter (NKCC1) is one of several transporters that have been linked to hypertension, and its inhibition reduces vascular smooth muscle tone and blood pressure. NKCC1 in the rat aorta is stimulated by vasoconstrictors and inhibited by nitrovasodilators, and this is linked to the contractile state of the smooth muscle. To determine whether blood pressure also regulates NKCC1, we examined the acute effect of hypertension on NKCC1 in rats after aortic coarctation. In the hypertensive aorta (28-mmHg rise in mean blood pressure), an increase in NKCC1 activity (measured as bumetanide-sensitive (86)Rb efflux) was apparent by 16 h and reached a plateau of 62% greater than control at 48 h. In contrast, there was a slight decrease in NKCC1 activity in the hypotensive aorta (21% decrease in mean blood pressure). Measurement of NKCC1 mRNA by real-time PCR revealed a fivefold increase in the hypertensive aorta compared with the hypotensive aorta or sham aorta. The inhibition by bumetanide of isometric force response to phenylephrine was significantly greater in the hypertensive aorta than in the control aorta or hypotensive aorta. We conclude that NKCC1 in rat aortic smooth muscle is regulated by blood pressure, most likely through changes in transporter abundance. This upregulation of NKCC1 is associated with a greater contribution to force generation in the hypertensive aorta. This is the first demonstration that NKCC1 in vascular smooth muscle is regulated by blood pressure and indicates that this transporter is important in the acute response of vascular smooth muscle to hypertension.     

Novel molecular variants of the Na-K-2Cl cotransporter gene are responsible for antenatal Bartter syndrome.

Antenatal Bartter syndrome is a variant of inherited renal-tubular disorders associated with hypokalemic alkalosis. This disorder typically presents as a life-threatening condition beginning in utero, with marked fetal polyuria that leads to polyhydramnios and premature delivery. Another hallmark of this variant is a marked hypercalciuria and, as a secondary consequence, the development of nephrocalcinosis and osteopenia. We have analyzed 15 probands belonging to 13 families and have performed SSCP analysis of the coding sequence and the exon-intron boundaries of the NKCC2 gene; and we report 14 novel mutations in patients with antenatal Bartter syndrome, as well as the identification of three isoforms of human NKCC2 that arise from alternative splicing.     

Regulation of erythrocyte Na-K-2Cl cotransport by threonine phosphorylation

A method is described to measure threonine phosphorylation of the Na-K-2Cl cotransporter in ferret erythrocytes using readily available antibodies. We show that most, if not all, cotransporter in these cells is NKCC1, and this was immunoprecipitated with T4. Cotransport rate, measured as 86Rb influx, correlates well with threonine phosphorylation of T4-immunoprecipitated protein. The cotransporter effects large fluxes and is significantly phosphorylated in cells under control conditions. Transport and phosphorylation increase 2.5- to 3-fold when cells are treated with calyculin A or Na+ arsenite. Both fall to 60% control when cell [Mg2+] is reduced below micromolar or when cells are treated with the kinase inhibitors, 4-amino-5-(4-methylphenyl)-7-(t-butyl)pyrazolo[3,4-d]pyrimidine or staurosporine. Importantly, these latter interventions do not abolish either phosphorylation or transport suggesting that a phosphorylated form of the cotransporter is responsible for residual fluxes. Our experiments suggest protein phosphatase 1 (PrP-1) is extremely active in these cells and dephosphorylates key regulatory threonine residues on the cotransporter. Examination of the effects of kinase inhibition after cells have been treated with high concentrations of calyculin indicates that residual PrP-1 activity is capable of rapidly dephosphorylating the cotransporter. Experiments on cotransporter precipitation with microcystin sepharose suggest that PrP-1 binds to a phosphorylated form of the cotransporter.     

Regulation of erythrocyte Na-K-2Cl cotransport by threonine phosphorylation

A method is described to measure threonine phosphorylation of the Na-K-2Cl cotransporter in ferret erythrocytes using readily available antibodies. We show that most, if not all, cotransporter in these cells is NKCC1, and this was immunoprecipitated with T4. Cotransport rate, measured as ⁸⁶Rb influx, correlates well with threonine phosphorylation of T4-immunoprecipitated protein. The cotransporter effects large fluxes and is significantly phosphorylated in cells under control conditions. Transport and phosphorylation increase 2.5- to 3-fold when cells are treated with calyculin A or Na⁺ arsenite. Both fall to 60% control when cell [Mg²⁺] is reduced below micromolar or when cells are treated with the kinase inhibitors, 4-amino-5-(4-methylphenyl)-7-(t-butyl)pyrazolo[3,4-d]pyrimidine or staurosporine. Importantly, these latter interventions do not abolish either phosphorylation or transport suggesting that a phosphorylated form of the cotransporter is responsible for residual fluxes. Our experiments suggest protein phosphatase 1 (PrP-1) is extremely active in these cells and dephosphorylates key regulatory threonine residues on the cotransporter. Examination of the effects of kinase inhibition after cells have been treated with high concentrations of calyculin indicates that residual PrP-1 activity is capable of rapidly dephosphorylating the cotransporter. Experiments on cotransporter precipitation with microcystin sepharose suggest that PrP-1 binds to a phosphorylated form of the cotransporter.     

The Na-K-2Cl cotransporter is in a permanently activated state in cytoplasts from Ehrlich ascites tumor cells

Brief incubation of Ehrlich ascites tumor cells with cytochalasin B causes the formation of blebs in the surface membrane. Gentle homogenization removes the blebs as intact cytoplasts which contain neither mitochondrian or nucleus, nor other cytoplasmic membranous organelles. The Na-K-2Cl cotransporter is present in the cytoplasts in a permanently activated state, whereas the Na-K-2Cl transport system in unperturbed intact cells is silent. Pretreatment of intact cells with cytochalasin B for l min stimulates the bumetanide-inhibitable K⁺ influx fivefold. The influx into purified cytoplasts when expressed per g protein is three- to fourfold higher than the influx into cytochalasin B-treated intact cells. Thus, the membrane vesicles are enriched with the cotransporter, and the cotransporter is present in an activated state. The K influx into cytoplasts is inhibited about 40% by Na-free, Cl-free or bumetanide-containing media and to a similar extent by Fab fragments prepared from antiserum against purified proteins of the cotransporter. The K _I for bumetanide was 0.19±0.06 m for the cytoplasts as compared to 0.67±0.11 m for the intact cells. SDS gel electrophoresis of membrane proteins from the cytoplast membranes compared to the membranes of intact cells shows a reduced number of bands and a majority of bands showing reduced staining, whereas a few bands are stained more intensely. Particularly notable is a band at 80 kD, which is similar to the molecular weight previously reported for the main membrane protein isolated from intact cells using a bumetanide-Sepharose affinity column. An immunoblot of the cytoplast preparation using antibodies against the purified bumetanide binding proteins showed strong immunodetection of the 80 kD protein.We are grateful to Marianne Schiødt, Birgit Blytmann Jørgensen, Thomas Krarup and Beverley Dyer for expert assistance. This work was supported by grants from the Danish Natural Science Research Council (11-6835 to E.K.H.) and the National Institutes of Health (DK 33640 to P.B.D.) and by a Carlsberg Foundation research fellowship (to F.J.).