ATP and Sulfonylurea Sensitivity of Mutant ATP-Sensitive K+ Channels in Neonatal Diabetes
http://www.100md.com
糖尿病学杂志 2005年第9期
the Department of Cell Biology and Physiology, Washington University School of Medicine, St. Louis, Missouri
ABSTRACT
The prediction that overactivity of the pancreatic ATP-sensitive K+ channel (KATP channel) underlies reduced insulin secretion and causes a diabetic phenotype in humans has recently been borne out by genetic studies implicating "activating" mutations in the Kir6.2 subunit of KATP as causal in both permanent and transient neonatal diabetes. Here we characterize the channel properties of Kir6.2 mutations that underlie transient neonatal diabetes (I182V) or more severe forms of permanent neonatal diabetes (V59M, Q52R, and I296L). In all cases, the mutations result in a significant decrease in sensitivity to inhibitory ATP, which correlates with channel "overactivity" in intact cells. Mutations can be separated into those that directly affect ATP affinity (I182V) and those that stabilize the open conformation of the channel and indirectly reduce ATP sensitivity (V59M, Q52R, and I296L). With respect to the latter group, alterations in channel gating are also reflected in a functional "uncoupling" of sulfonylurea (SU) block: SU sensitivity of I182V is similar to that of wild-type mutants, but the SU sensitivity of all gating mutants is reduced, with the I296L mutant being resistant to block by tolbutamide (10 mmol/l). These results have important implications for the use of insulinotropic SU drugs as an alternative therapy to insulin injections.
In the pancreatic -cell, the ATP-sensitive K+ channel (KATP channel) couples membrane excitability to insulin secretion (1). Increased glucose metabolism leads to elevated cytosolic ATP/ADP, closure of KATP channels at the plasma membrane, and membrane depolarization. The resulting activation of voltage-sensitive Ca2+ channels causes a rise in [Ca2+]i, which serves as the stimulus for insulin vesicle exocytosis. Sulfonylureas (SUs), which are major hypoglycemic agents used in treating type 2 diabetes, promote insulin secretion by specifically binding the regulatory SUR1 subunit of KATP and inhibiting KATP current (2), thus underscoring the central role of KATP in the regulation of insulin secretion.
Implicit in the paradigm of excitation-secretion coupling is that alterations in KATP currents should disrupt electrical signaling in the -cell and, thereby, alter insulin release. Specifically, decreased metabolic flux, decreased sensitivity of KATP to inhibitory ATP, or increased density of KATP channels should all lead to abnormally high KATP activity and relative hypoinsulinism. In such scenarios, a diabetic phenotype is predicted; indeed, a severe neonatal diabetic phenotype is observed in mice lacking the -cell glucokinase gene (3,4) and transgenic mice expressing -cell KATP channels with reduced sensitivity to inhibitory ATP (i.e., "overactive" KATP) (5). A series of genetic studies have recently demonstrated that heterozygous, missense mutations in the pore-forming Kir6.2 subunit of the KATP channel underlie neonatal diabetes in humans, accounting for both permanent and transient forms of the disease (6eC11). Rare in its occurrence (1:400,000 births), neonatal diabetes is usually diagnosed within the first 3 months of life and requires insulin administration to treat the hyperglycemia (12). In the milder, transient neonatal diabetes (TND), hyperglycemia usually resolves within 18 months after the diagnosis, whereas the equally common permanent neonatal diabetes (PND) requires insulin treatment for life. In both PND and TND, -cell dysfunction is likely the primary lesion. However, in TND, it is proposed that expansion of -cell mass compensates for the -cell defect and underlies the remission phase (13). In one important finding, the administration of SUs to several neonatal diabetic patients carrying Kir6.2 mutations was sufficient to treat the hyperglycemia, thereby implicating KATP "overactivity" as causal in the suppressed insulin release (6,14).
All neonatal diabeteseCassociated Kir6.2 mutations examined to date have been shown to decrease ATP sensitivity of reconstituted KATP channels (7,10,15). Some of the mutations are in putative ATP-binding residues (e.g., R50, I182, R201), but others lie well away from any predicted binding regions. These latter mutations are likely to reduce ATP sensitivity by stabilizing the open state relative to the closed state (16); in fact, this has been demonstrated for some PND mutations (15). Little attention has been paid to the sensitivity of different mutations to insulinotropic SU drugs. This is an important issue given that other mutations that alter open-state stability have been shown to dramatically affect SU sensitivity (17,18) and that SU therapy holds the promise of an alternative to insulin treatment for neonatal diabetes patients (6,14).
In the present study, we characterized Kir6.2 mutations that arguably underlie the mildest form of neonatal diabetes, TND (I182V), as well as the more severe, permanent forms of the disease (V59M, Q52R, and I296L). In all cases, mutations led to decreased ATP sensitivity. The mutations can be mechanistically classified as those that appear to directly alter ATP binding (I182V) and those that indirectly alter ATP sensitivity through changes in gating behavior (V59M, Q52R, and I296L). With respect to the latter group, a concomitant decrease in the sensitivity of reconstituted KATP channels to two clinically prescribed SUs (tolbutamide and glibenclamide) was observed in intact cells and excised patches. These findings have potentially important implications for the treatment of diabetic patients carrying Kir6.2 mutations and suggest that SU dosing may need to be considered on a case-by-case basis.
RESEARCH DESIGN AND METHODS
Molecular biology.
Kir6.2 was cloned into the EcoRI/Cla I site of pCMV6b, and the parental plasmid DNA was used to generate Kir6.2 mutations using the QuickChange Site Directed Mutagenesis kit (Stratagene, La Jolla, CA). SUR1 was cloned into the pECE expression vector. The nucleotide sequences of the mutant Kir6.2 constructs were verified by fluorescence-based cycle sequencing using AmpliTaq DNA polymerase FS (PerkinElmer, Foster City, CA) and an ABI Prism DNA sequencer (PerkinElmer).
Expression of KATP channels in COSm6 cells.
COSm6 cells were plated at a density of 2.5 x 105 cells/well (30-mm, six-well dishes) and cultured in Dulbecco’s modified Eagle’s medium with 10 mmol/l glucose supplemented with FCS (10%). The following day, a transfection cocktail was prepared by adding 6 e蘬 of FuGENE 6 Transfection Reagent (Roche, Indianapolis, IN), 600 ng each of pCMV6b-Kir6.2 and pECEeCgreen fluorescent protein plasmid, and 1 e pECE-SUR1 (per well). After a 45-min incubation at room temperature, the cocktail was added directly to the growth medium on the plated cells. Cells were assayed for KATP currents by patch-clamp measurements 2eC4 days after transfection.
Rubidium flux experiments.
86RbCl (1 mCi/ml) was added in fresh growth medium 48 h after transfection, and cells were incubated for an additional 24 h before Rb+ flux was assayed. For efflux measurements, cells were preincubated for 2 min at 25°C in Krebs-Ringer solution with or without metabolic inhibitors (2.5 e/ml oligomycin plus 1 mmol/l 2-deoxy-D-glucose [Sigma, St. Louis, MO]). At selected time points, the solution was aspirated from the wells and replaced with fresh solution. The 86Rb+ in the aspirated solution was counted in a scintillation solution. Rb+ efflux curves were fit by assuming that a single non-KATP conductance was active in untransfected cells, determined as
1
where exp is an exponential function, t is time, and k1 is the rate constant for efflux through the endogenous, non-KATP pathway.
In transfected cells, efflux curves were fit by assuming that both a KATP and a non-KATP conductance pathway were active in parallel and were determined as
2
where k1 is the rate constant for efflux through the endogenous, non-KATP pathway (obtained from untransfected cells by Eq. 1) and k2 is the rate constant for efflux through the KATP pathway. k2 is then directly proportional to the KATP conductance.
Patch-clamp measurements.
Patch-clamp studies were performed at room temperature in an oil-gate chamber that allowed the solution bathing the exposed surface of the isolated patch to be changed rapidly. The standard bath (intracellular) and pipette (extracellular) solution (K-INT) used in these experiments was comprised of 140 mmol/l KCl, 10 mmol/l K-HEPES, and 1 mmol/l K-EGTA (pH 7.3). Microelectrode resistance was typically 0.5eC1 M when filled with K-INT solution. Membrane patches were voltage clamped with an Axopatch 1B patch-clamp amplifier (Axon, Foster City, CA). Phosphatidylinositol 4,5-bisphosphate (PIP2; 5 e/ml in K-INT) was bath sonicated in ice for 30 min before being used. Glibenclamide was dissolved as a 1 mol/l stock solution in DMSO and diluted to <1% DMSO in K-INT. Tolbutamide was prepared as a 100 mmol/l stock solution in 150 mmol/l KOH and diluted to the appropriate working concentration in K-INT (pH 7.3). PIP2 was obtained from Boehringer Mannheim. Tolbutamide, glibenclamide, and ATP were purchased from Sigma. All currents were measured at a membrane potential of eC50 mV (pipette voltage = +50 mV). Data were normally filtered at 0.5eC3 kHz, and signals were digitized at 22 kHz (Neurocorder; Neurodata, New York, NY). Experiments were digitized into a microcomputer using Clampex 8.2 software (Axon), and off-line analysis was performed using Clampfit 8.2 (Axon) and Microsoft Excel programs. Wherever possible, data are presented as means ± SE. Microsoft Solver was used to fit data by a least-squares algorithm.
Estimation of Po,zero.
Two approaches, one indirect and one direct, were used to estimate the open probability in zero ATP (Po,zero) after isolated membrane patches containing multiple channels were excised (19).
PIP2 method (indirect).
PIP2 was added to the patch until the current reached a saturating level (IPIP2). This was assumed to represent a maximum Po,zero of 0.9 (20). The fractional change in current was calculated as
3
where Iinitial is the initial current. Po,zero was then estimated as
4
Noise analysis method (direct).
In addition, the mean Po,zero was estimated from stationary fluctuation analysis of macroscopic currents (21,22) on short (<1 s) recordings of currents in zero or 5 mmol/l ATP or 10 mmol/l spermine (for estimation of ATP-independent noise). Currents (corresponding to 1eC1000 channels) were filtered at 1 kHz and digitized at 3 kHz with 12-bit resolution. The mean patch current (I) and variance () (2) in the absence of ATP were obtained by subtracting the mean current and variance in 5 mmol/l ATP or 10 mmol/l spermine (i.e., assuming all channels were fully closed or blocked). The single channel current (i) was assumed to be eC3.75 pA at eC50 mV, corresponding to a wild-type single channel conductance of 75 pS (23). Po,zero was then estimated as
5
Three-state model for KATP channel activity.
Detailed kinetic analysis of KATP channel activity predicts complex models, but a simplified three-state model (19,20) can account qualitatively and quantitatively for the essential features of macroscopic KATP channel activity. Accordingly, we examined whether this simple model (Fig. 1), in accounting for the effects of mutations on gating behavior and nucleotide sensitivity, could be used predictively to account for observed changes in drug sensitivity.
The ATP binding affinity (KA) and open-close state equilibrium (KCO) for wild-type channels were assumed to be 3 eol/l and 0.60, respectively. To obtain predictions of behavior for mutant channels (as in Fig. 8), only one or the other parameter was changed. For I182V, it was necessary to change KA (increased by sixfold) without changing KCO. For Q52R, V59M, and I296L channels, it was necessary to change KCO (increased by 40-, 6-, and 600-fold, respectively) without changing KA.
Dose-response curves for SU (glibenclamide and tolbutamide) inhibition were described by the sum of two Hill equations (reflecting two components of inhibition):
6
where Grel is the relative conductance, f is the fraction of high-affinity block, and K1 and K2 are the half-maximal concentration of high- and low-affinity block, respectively.
It is clear that the relative magnitude of the two components was changed by the mutations as was the apparent affinities. We postulated, based on previous findings (17,18), that these changes resulted directly from changes in open-state stability. As shown in Fig. 8, to model the effects of different mutations on SU sensitivity, it was necessary to assume only that the fraction of high-affinity block (f) and the half-maximal blocking concentrations (K1 and K2) of the high- and low-affinity components were each dependent on the open-state stability (defined by KCO), according to the following three formulas:
7
where x = 20 or 1 for glibenclamide (Fig. 2C) or tolbutamide (Fig. 2D) inhibition curves, respectively.
8
where K1(0) is equal to 1 eol/l or 10 nmol/l for glibenclamide or tolbutamide inhibition, respectively.
9
where K2(0) is much greater than 10 or 3 mmol/l for glibenclamide or tolbutamide inhibition, respectively.
RESULTS
Variable on-cell activity correlates with variable ATP sensitivity.
To better understand the spectrum of KATP-induced neonatal diabetes and its underlying mechanisms, we chose to characterize the TND-causing I182V mutation and the three mutations implicated in more severe PND (Q52R, V59M, I296L) (Fig. 2). Mutant Kir6.2 subunits were coexpressed with the wild-type SUR1 subunit, and channel activity in intact cells was first screened by 86Rb flux experiments. As shown in Fig. 3, 86Rb+ efflux from cells transfected with wild-type KATP channels (Kir6.2 + SUR1) was low under basal conditions and was significantly activated by metabolic inhibition with oligomycin and 2-deoxyglucose to lower cellular ATP/ADP. In contrast, Rb+ efflux from cells transfected with mutant Kir6.2 subunits was high under basal conditions and showed very little increase, if any, with metabolic inhibition. To obtain quantitative estimates of conductances, a simple two-pathway (KATP + non-KATP) model for Rb+ efflux was assumed (see RESEARCH DESIGN AND METHODS). Metabolic inhibition reduced the Rb+ efflux rate in untransfected cells; Rb+ efflux time courses in basal and metabolically inhibited conditions for untransfected cells were thus fit by Eq. 1, which assumes a single non-KATP conductance is active in each condition. The rate constants thus obtained were used to estimate rate constants for KATP conductance under basal conditions and metabolic inhibition in transfected cells using Eq. 2. These rate constants were then assumed to be directly proportional to KATP conductance. Figure 4A shows the estimated KATP conductance under basal and metabolically inhibited conditions for each transfected construct. Maximal activated conductance was similar for all constructs except I296L. This mutation gave consistently lower fluxes, a point considered below (see DISCUSSION), but the basal flux was almost maximal.
We next examined KATP channel activity directly in voltage-clamped membrane patches from transfected cells. For each patch, we estimated the current in basal physiological conditions (on-cell current) and the measured maximal KATP current after patch excision (data not shown). In Fig. 4B, relative KATP current activation on-cell from patch-clamp analysis is plotted against the relative KATP activation in intact cells estimated by Rb+ efflux for each mutation. Both measures give correlated activities, suggesting that each method reflects the degree of overactivity in intact cells. To examine the correlation between increased KATP activity in intact cells and ATP sensitivity, the latter was measured directly in excised membrane patches (Fig. 5). All mutant KATP channels exhibited a significant decrease in ATP inhibition, ranging from twofold (V59M) to >100-fold (I296L) (Table 1). Channel overactivity in intact cells correlated reasonably well with reduction in ATP sensitivity, but a direct correlation between ATP insensitivity and severity of the neonatal diabetes was not absolute, as the TND-causing I182V mutation was approximately twofold less sensitive than V59M, a mutation that is associated with severe, syndromic PND (see DISCUSSION). It is conceivable that channel trafficking was altered by neonatal diabetes mutations so that a higher current may partly reflect higher expression levels. As shown in Table 1, patch current density and calculated channel density in this recombinant system were not significantly different between wild-type and mutant channels.
Mutations may affect ATP sensitivity by reduced ATP binding or indirectly by changes in the relative stability of the open state.
Mutations could mechanistically reduce ATP sensitivity directly by reducing ATP affinity or indirectly by affecting the intrinsic opening ability. Previous studies have characterized mutations throughout Kir6.2 that stabilize the open state of the channel relative to the closed state, even in the absence of ATP, and thereby decrease the apparent ATP sensitivity (19,20,24,25). To elucidate potential mechanisms in neonatal diabetes mutations, we measured the open probability in the absence of ATP (Po,zero), using two independent methods: PIP2 activation (19) and nonstationary noise analysis (22) (see RESEARCH DESIGN AND METHODS). Each method gave very similar results (see Fig. 8A). Po,zero of wild-type channels was 0.4 and was not altered by the TND-associated I182V mutation (which reduced ATP sensitivity approximately fivefold), consistent with the mutation directly affecting ATP binding. However, each of the PND-associated mutations caused a significant increase in Po,zero (Table 1), indicative of an indirect effect on ATP sensitivity. This differential mechanistic basis for channel activation could have profound consequences, both for the severity of the resultant neonatal diabetes and for the possibilities for SU therapy (see DISCUSSION).
SU sensitivity is reduced in parallel with increases in open-state stability.
Previously, it has been shown that increasing the open-state stability of KATP, either through mutagenesis or the application of PIP2, functionally "uncouples" the regulatory SUR1 subunit from the ion-conducting Kir6.2 subunit (17,18). This uncoupling is reflected as a reduction or loss of high-affinity SU inhibition through the SUR1 subunit. For Kir6.2 mutations that increase the open-state stability, we may therefore expect to find a concomitant decrease of high-affinity SU block in addition to a decrease in apparent ATP affinity. To test this possibility, we examined SU sensitivity, in intact cells using 86Rb+ flux measurements and in inside-out patch-clamp experiments.
As shown in Fig. 6, glibenclamide almost completely inhibited Rb+ efflux through wild-type channels at clinically effective doses (i.e., in the nanomolar range). It has previously been demonstrated that SU inhibition in excised membrane patches is biphasic, with high- and low-affinity components (17,26). Although a biphasic sensitivity has not been previously demonstrated in intact tissue, it is evident in the SU sensitivity of mutant KATP channels in intact COS cells (Fig. 6), where Rb+ efflux is inhibited to a plateau level by micromolar glibenclamide. For mutants that exhibit increased open-state stability, the high-affinity component decreases (Q52R, V59M, I296L) and the affinity itself decreases (Q52R, I296L). The highest open-state stability mutant (I296L) is almost completely insensitive to glibenclamide in the nanomolar range, but a reduced high-affinity component and lower affinity is also quite clear for the Q52R mutant.
Glibenclamide is a hydrophobic drug (27), and inhibition of KATP channels in excised patches was largely irreversible over the experimental period. However, tolbutamide, a first generation SU, is more hydrophilic, can be rapidly reversed, and has been a useful model SU in excised membrane patch experiments. As shown in Fig. 7, wild-type channels showed typical biphasic block by tolbutamide in inside-out membrane patches (17,26), and sensitivity was unaltered in the I182V mutation (Fig. 7B). This finding is again consistent with the interpretation that the I182V substitution directly alters ATP binding/transduction but does not affect gating (28), and thus the mutant channels remain functionally "coupled" to SUR1.
In sharp contrast, functional uncoupling from the SUR1 subunit is again reflected by a decrease in high-affinity fractional SU block in each of the mutations that increase the open-state stability (Q52R, V59M, I296L) (Table 1). It is interesting that I296L channels, which show the highest open-state stability, are completely insensitive to inhibition by tolbutamide at concentrations as high as 10 mmol/l. However, when wild-type Kir6.2 and I296L subunits were coexpressed to generate heterozygous channels, the SU sensitivity was partially restored, but with a significant decrease in the high-affinity component of block (17% high-affinity tolbutamide block for heterozygous I296L:wild-type channels vs. 55% block for homomeric wild-type channels) (Fig. 7B).
DISCUSSION
Reduced ATP sensitivity underlies both TND and PND.
It is now clear that mutations in the Kir6.2 subunit of KATP represent a common cause of neonatal diabetes in humans (6eC11). To better understand the mechanism(s) by which these mutations alter excitation-secretion coupling and their effect on SU sensitivity, we chose to characterize Kir6.2 mutations implicated in the mildest, transient form of the disease (I182V) (10) and those associated with more severe PND (Q52R, V59M, I296L) (6eC9). In all cases, the sensitivity of homomeric mutant KATP to inhibitory ATP was reduced relative to wild-type KATP, with the following order of sensitivity: WT > V59M > I182V > Q52R > I296L. The syndromic PND-linked V59M mutation exhibited comparable if not greater sensitivity to ATP than did the TND-associated I182V (K1/2ATP = 15 [V59M] and 31 [I182V] eol/l), reflected in a similar degree of "overactivity" in intact cells (Fig. 4B). Assuming the behavior of reconstituted channels reflects KATP activity in the -cell, these data do not demonstrate an absolute correlation between the severity of the diabetes (TND, PND, and syndromic PND) and the degree of ATP sensitivity. Such a positive correlation has been previously suggested based on a comparison of the less severe functional defects of a PND-associated mutation (R201C) with syndromic PND-associated mutations (Q52R, V59G) (15). Taken together, these data suggest a complex phenotype in which the primary defect is at the level of the -cell and altered insulin secretion, but they also indicate that additional genetic and/or environmental factors may modulate the severity of the disease (TND, PND, and syndromic PND). It is interesting that KATP in skeletal muscle (Kir6.2 + SUR2A) is implicated in peripheral insulin sensitivity based on the observation that both Kir6.2- and SUR2A-null mice exhibit increased glucose uptake in skeletal muscle and adipose tissue (29,30). The mouse models, therefore, predict that "overactive" KATP, as observed in neonatal diabetes, may suppress glucose uptake and contribute to the hyperglycemia. Whether neonatal diabetes patients, in addition to having suppressed insulin release, also exhibit insulin insensitivity remains to be determined.
An important distinction between the present study and that of Proks et al. (15) is that the order of sensitivity in Proks et al.’s report was based on the ATP sensitivity of heteromeric channels (generated by the expression of channels from equal amounts of wild-type Kir6.2 and mutant Kir6.2 cRNA), whereas our analysis is of homomeric channels. We have found that in similar experiments with mixed expression of wild-type and mutant Kir6.2 subunits, the rank order of mutant ATP sensitivity is unaltered (data not shown). However, a quantitative analysis of mixed expression is very complicated, given the fact that six distinct combinations of wild-type and mutant subunits will generate six different channel types, each with its own ATP sensitivity. Whether mutant alleles of Kir6.2 will be expressed equally in vivo and in different tissues is also not known. Given these complexities, the analysis of homomeric channels may be the most straightforward approach to understanding mutational effects on channel properties.
Finally, the I296L mutant channel consistently exhibited a lower activity in Rb+ efflux experiments, despite a channel density that was similar to that of wild-type channels in excised patches. This finding may reflect a lower transfection efficiency. However, because the I296L mutation resulted in the highest on-cell activity of all the neonatal diabetes mutants, it is also possible that this elevated K+ conductance was toxic to the cells, so that we were experimentally selecting for a subpopulation of cells that are resistant to elevated K+ conductance in excised patch experiments.
Differential SU sensitivity of neonatal diabetes mutations.
SUs block pancreatic KATP currents and stimulate insulin release. The high-affinity component of block by glibenclamide (IC50 1 nmol/l) or tolbutamide (IC50 1 eol/l) results from binding to the SUR1 subunit (26), whereas the Kir6.2 subunit underlies the low-affinity component (IC50 1,000 x high affinity). A direct correlation between gating behavior of the channel and SU sensitivity has been previously demonstrated (17,18). The consequence of such a relation is that mutations in Kir6.2 that increase the stability of the open state not only reduce the apparent ATP sensitivity, but concomitantly decrease the high-affinity SU block. This is interpreted as a functional "uncoupling" of the regulatory SUR1 subunit from the Kir6.2 channel (17).
We thus predict that PND mutations that increase open-state stability (V59M, Q52R, and I296L) should also reduce SU sensitivity. This hypothesis is consistent with the previous observation that the block of whole-cell K+ currents was not complete in 500 eol/l tolbutamide for PND mutant KATP (15). Indeed, the gating mutations Q52R and V59M but not the ATP-binding mutant I182V exhibited a decrease in the fractional block by the high-affinity component. The Q52R and I296L mutants also exhibited a significant decrease in the sensitivity of the low-affinity block, with I296L mutant channels being completely refractory to inhibition by tolbutamide 10 mmol/l. This implies that in addition to a reduced SU sensing through SUR1, the I296L and Q52R mutations also reduce SU action through the low-affinity site on Kir6.2, either directly or allosterically.
A mechanistic model for nucleotide and SU sensitivity.
Diabetes-causing Kir6.2 mutations can be classified mechanistically as those that directly alter ATP binding and/or transduction of the ATP-mediated signal (I182V) versus those that indirectly affect ATP affinity by increasing the open-state stability (Q52R, V59M, I296L). The data are consistent with recent characterizations of other PND mutations that affect ATP binding (R201C, R201H, I182V) or alter gating behavior, leading to a secondary decrease in ATP sensitivity (V59G, Q52R) (10,15).
Models of KATP channel gating and nucleotide and SU sensitivity have been developed. To have predictive use, such models should account for channel gating (in the absence of ATP), nucleotide sensitivity, and drug sensitivity, and by adjusting relevant parameters, should be able to account for the effects of mutations on channel function. The most complex models (20,28,31) account for the tetrameric nature of the channel and could potentially account for heteromeric channels consisting of wild-type and mutant subunits. However, a relatively simple three-state model (20) (see RESEARCH DESIGN AND METHODS, Fig. 1) can account for all of the pertinent features of macroscopic channel activity and has predictive qualities for the interpretation of nucleotide and drug sensitivities of neonatal diabetes mutants. We therefore postulated that mutational effects on SU sensitivity result directly from changes in open-state stability (17,18).
As shown in Fig. 8, the simplified three-state model quantitatively predicted the effects of neonatal diabetes mutations on each of the three features, Po,zero, ATP sensitivity, and SU sensitivity, by adjusting only two parameters, KA and KCO, which are the equilibrium constants that describe ATP affinity and open-state stability, respectively. By assuming that both the affinity and the fraction of high-affinity SU block are dependent on only KCO, the steady-state SU sensitivities of wild-type and mutant channels can be predicted. As discussed below, this analytical approach may be a useful predictor of SU efficacy in human therapy trials.
Implications for SU therapy of neonatal diabetes.
Recently, SUs have been shown to be effective, at least in the short-term, in maintaining control of blood glucose levels in several PND patients in which Kir6.2 mutations are causal (F331I, V59M, and R201H) (6,14). Given that blood glucose levels and HbA1c values were stable after discontinuation of insulin in these patients, it has been proposed that oral SUs can be an alternative therapy to subcutaneous insulin injections in patients with PND because of Kir6.2-activating mutations. The present results indicate, however, that the efficacy of SU therapy is likely to depend on the nature of the Kir6.2 mutation. Assuming that the SU sensitivity of reconstituted channels reflects the sensitivity of channels in vivo, carriers of the diabetes-causing I296L and similarly behaving Kir6.2 mutations may be less responsive to SU therapy. In one interesting case, a PND patient with the syndromic V59M mutation was responsive to SU therapy and discontinued insulin injections (6), thereby suggesting that despite a significant decrease in sensitivity of homomeric V59M mutant channels to SU inhibition, sufficient high-affinity inhibition was still able to induce insulin release. In the original report by Gloyn et al. (7), it was observed that a parent of the R201H proband was diagnosed as a baby with diabetes and had been effectively treated with tolbutamide for 46 years. The mutation R201H causes a significant loss of ATP sensitivity (7), but it is predicted to be an ATP-binding site mutation (16,32,33) and thus may not show loss of SU sensitivity.
Functional characterization of additional PND-associated mutations in Kir6.2, as well as additional clinical studies, will provide essential data regarding the efficacy of SU therapy in treating KATP-induced diabetes and the possibility of tailoring individual therapy based on the underlying Kir6.2 mutation. The present results suggest that careful attention should be paid to the degree to which neonatal diabetes mutations cause a loss of ATP sensitivity by an increase in open-state stability. The greater the mutational effect on open-state stability is, the higher the SU dose necessary to achieve a given degree of channel inhibition will be. Thus, to achieve the same therapeutic effect (i.e., to obtain sufficient closure of KATP under appropriate physiological conditions), proportionally higher doses of an SU may be clinically required.
ACKNOWLEDGMENTS
This work was supported by National Institutes of Health Diabetes Research and Training Center Grant DK-20579 and National Institutes of Health Grant DK069445 (R01) at Washington University and an American Diabetes Association-Takeda Pharmaceuticals Mentor-Based Minority Postdoctoral Fellowship Program (to M.S.R.).
FOOTNOTES
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
KATP channel, ATP-sensitive K+ channel; PIP2, phosphatidylinositol 4,5-bisphosphate; PND, permanent neonatal diabetes; SU, sulfonylurea; TND, transient neonatal diabetes
REFERENCES
Ashcroft FM, Rorsman P: ATP-sensitive K+ channels: a link between B-cell metabolism and insulin secretion. Biochem Soc Trans18 :109 eC111,1990
Aguilar-Bryan L, Nichols CG, Wechsler SW, Clement JP 4th, Boyd AE 3rd, Gonzalez G, Herrera-Sosa H, Nguy K, Bryan J, Nelson DA: Cloning of the beta cell high-affinity sulfonylurea receptor: a regulator of insulin secretion. Science268 :423 eC426,1995
Terauchi Y, Sakura H, Yasuda, Iwamoto K, Takahashi N, Ito K, Kasai H, Suzuki H, Ueda O, Kamada N, et al.: Pancreatic beta-cell-specific targeted disruption of glucokinase gene: diabetes mellitus due to defective insulin secretion to glucose. J Biol Chem270 :30253 eC30256,1995
Ishihara H, Tashiro F, Ikuta K, Asano T, Katagiri H, Inukai K, Kikuchi M, Yazaki Y, Oka Y, Miyazaki J: Inhibition of pancreatic beta-cell glucokinase by antisense RNA expression in transgenic mice: mouse strain-dependent alteration of glucose tolerance. FEBS Lett371 :329 eC332,1995
Koster JC, Marshall BA, Ensor N, Corbett JA, Nichols CG: Targeted overactivity of beta cell K(ATP) channels induces profound neonatal diabetes. Cell100 :645 eC654,2000
Sagen JV, Raeder H, Hathout E, Shehadeh N, Gudmundsson K, Baevre H, Abuelo D, Phornphutkul C, Molnes J, Bell GI, Gloyn AL, Hattersley AT, Molven A, Sovik O, Njolstad PR: Permanent neonatal diabetes due to mutations in KCNJ11 encoding Kir6.2: patient characteristics and initial response to sulfonylurea therapy. Diabetes53 :2713 eC2718,2004
Gloyn AL, Pearson ER, Antcliff JF, Proks P, Bruining GJ, Slingerland AS, Howard N, Srinivasan S, Silva JM, Molnes J, Edghill EL, Frayling TM, Temple IK, Mackay D, Shield JP, Sumnik Z, van Rhijn A, Wales JK, Clark P, Gorman S, Aisenberg J, Ellard S, Njolstad PR, Ashcroft FM, Hattersley AT: Activating mutations in the gene encoding the ATP-sensitive potassium-channel subunit Kir6.2 and permanent neonatal diabetes [erratum in N Engl J Med351 :1470 ,2004 ]. N Engl J Med350 :1838 eC1849, 2004
Vaxillaire M, Populaire C, Busiah K, Cave H, Gloyn AL, Hattersley AT, Czernichow P, Froguel P, Polak M: Kir6.2 mutations are a common cause of permanent neonatal diabetes in a large cohort of French patients. Diabetes53 :2719 eC2722,2004
Massa O, Iafusco D, D’Amato E, Gloyn AL, Hattersley AT, Pasquino B, Tonini G, Dammacco F, Zanette G, Meschi F, Porzio O, Bottazzo G, Crino A, Lorini R, Cerutti F, Vanelli M, Barbetti F: KCNJ11 activating mutations in Italian patients with permanent neonatal diabetes. Hum Mutat25 :22 eC27,2005
Gloyn AL, Reimann F, Girard C, Edghill EL, Proks P, Pearson ER, Temple IK, Mackay DJ, Shield JPH, Freedenberg D, Noyes K, Ellard S, Ashcroft FM, Gribble FM, Hattersley AT: Relapsing diabetes can result from moderately activating mutations in KCNJ11. Hum Mol Genet14 :925 eC934,2005
Gloyn AL, Edghill EL, Pearson ER, Mackay D, Temple IK, Shield J, Noyes K, Freedenberg D, Gillispie KM, Lambert AP, Gale EA, Ellard S, Hattersley A: Multiple subtypes of diabetes are associated with activating mutations in KCNJ11, which encodes the Kir6.2 subunit of the -cell ATP-sensitive potassium channel [online abstract]. Available at http://scientificsessions.diabetes.org/Abstracts/index.cfmfuseaction=Locator.DisplaysAbstractSearch
Polak M, Shield J: Neonatal and very-early-onset diabetes mellitus. Semin Neonatol9 :59 eC65,2004
Ma D, Shield JP, Dean W, Leclerc I, Knauf C, Burcelin RR, Rutter GA, Kelsey G: Impaired glucose homeostasis in transgenic mice expressing the human transient neonatal diabetes mellitus locus, TND. J Clin Invest114 :339 eC348,2004
Zung A, Glaser B, Nimri R, Zadik Z: Glibenclamide treatment in permanent neonatal diabetes mellitus due to an activating mutation in Kir6.2. J Clin Endocrinol Metab89 :5504 eC5507,2004
Proks P, Antcliff JF, Lippiat J, Gloyn AL, Hattersley AT, Ashcroft FM: Molecular basis of Kir6.2 mutations associated with neonatal diabetes plus neurological features. Proc Natl Acad Sci U S A101 :17539 eC17544,2004
Enkvetchakul D, Nichols CG: Gating mechanism of KATP channels: function fits form. J Gen Physiol122 :471 eC480,2003
Koster JC, Sha Q, Nichols CG: Sulfonylurea and K(+)-channel opener sensitivity of K(ATP) channels: functional coupling of Kir6.2 and SUR1 subunits. J Gen Physiol114 :203 eC213,1999
Reimann F, Tucker SJ, Proks P, Ashcroft FM: Involvement of the N-terminus of Kir6.2 in coupling to the sulphonylurea receptor. J Physiol (Lond)518 :325 eC336,1999
Cukras CA, Jeliazkova I, Nichols CG: The role of NH(2)-terminal positive charges in the activity of inward rectifier K(ATP) channels. J Gen Physiol120 :437 eC446,2002
Enkvetchakul D, Loussouarn G, Makhina E, Shyng SL, Nichols CG: The kinetic and physical basis of K(ATP) channel gating: toward a unified molecular understanding. Biophys J78 :2334 eC2348,2000
Neher E, Stevens CF: Conductance fluctuations and ionic pores in membranes. Annu Rev Biophys Bioeng6 :345 eC381,1977
Sigworth FJ: The variance of sodium current fluctuations at the node of Ranvier. J Physiol307 :97 eC129,1980
Shyng S, Ferrigni T, Nichols CG: Control of rectification and gating of cloned KATP channels by the Kir6.2 subunit. J Gen Physiol110 :141 eC153,1997
Trapp S, Proks P, Tucker SJ, Ashcroft FM: Molecular analysis of ATP-sensitive K channel gating and implications for channel inhibition by ATP. J Gen Physiol112 :333 eC349,1998
Drain P, Li L, Wang J: KATP channel inhibition by ATP requires distinct functional domains of the cytoplasmic C terminus of the pore-forming subunit. Proc Natl Acad Sci U S A95 :13953 eC13958,1998
Gribble FM, Tucker SJ, Ashcroft FM: The interaction of nucleotides with the tolbutamide block of cloned ATP-sensitive K+ channel currents expressed in Xenopus oocytes: a reinterpretation. J Physiol504 :35 eC45,1997
Zunkler BJ, Trube G, Panten U: How do sulfonylureas approach their receptor in the B-cell plasma membrane Naunyn Schmiedebergs Arch Pharmacol340 :328 eC332,1989
Li L, Wang J, Drain P: The I182 region of k(ir)6.2 is closely associated with ligand binding in K(ATP) channel inhibition by ATP. Biophys J79 :841 eC852,2000
Miki T, Minami K, Zhang L, Morita M, Gonoi T, Shiuchi T, Minokoshi Y, Renaud JM, Seino S: ATP-sensitive potassium channels participate in glucose uptake in skeletal muscle and adipose tissue. Am J Physiol Endocrinol Metab283 :E1178 eCE1184,2002
Chutkow WA, Samuel V, Hansen PA, Pu J, Valdivia CR, Makielski JC, Burant CF: Disruption of Sur2-containing K(ATP) channels enhances insulin-stimulated glucose uptake in skeletal muscle. Proc Natl Acad Sci U S A98 :11760 eC11764,2001
Enkvetchakul D, Loussouarn G, Makhina E, Nichols CG: ATP interaction with the open state of the K(ATP) channel. Biophys J80 :719 eC728,2001
Shyng SL, Cukras CA, Harwood J, Nichols CG: Structural determinants of PIP(2) regulation of inward rectifier K(ATP) channels. J Gen Physiol116 :599 eC608,2000
Ribalet B, John SA, Weiss JN: Molecular basis for Kir6.2 channel inhibition by adenine nucleotides. Biophys J84 :266 eC276,2003(Joseph C. Koster, Maria S)
ABSTRACT
The prediction that overactivity of the pancreatic ATP-sensitive K+ channel (KATP channel) underlies reduced insulin secretion and causes a diabetic phenotype in humans has recently been borne out by genetic studies implicating "activating" mutations in the Kir6.2 subunit of KATP as causal in both permanent and transient neonatal diabetes. Here we characterize the channel properties of Kir6.2 mutations that underlie transient neonatal diabetes (I182V) or more severe forms of permanent neonatal diabetes (V59M, Q52R, and I296L). In all cases, the mutations result in a significant decrease in sensitivity to inhibitory ATP, which correlates with channel "overactivity" in intact cells. Mutations can be separated into those that directly affect ATP affinity (I182V) and those that stabilize the open conformation of the channel and indirectly reduce ATP sensitivity (V59M, Q52R, and I296L). With respect to the latter group, alterations in channel gating are also reflected in a functional "uncoupling" of sulfonylurea (SU) block: SU sensitivity of I182V is similar to that of wild-type mutants, but the SU sensitivity of all gating mutants is reduced, with the I296L mutant being resistant to block by tolbutamide (10 mmol/l). These results have important implications for the use of insulinotropic SU drugs as an alternative therapy to insulin injections.
In the pancreatic -cell, the ATP-sensitive K+ channel (KATP channel) couples membrane excitability to insulin secretion (1). Increased glucose metabolism leads to elevated cytosolic ATP/ADP, closure of KATP channels at the plasma membrane, and membrane depolarization. The resulting activation of voltage-sensitive Ca2+ channels causes a rise in [Ca2+]i, which serves as the stimulus for insulin vesicle exocytosis. Sulfonylureas (SUs), which are major hypoglycemic agents used in treating type 2 diabetes, promote insulin secretion by specifically binding the regulatory SUR1 subunit of KATP and inhibiting KATP current (2), thus underscoring the central role of KATP in the regulation of insulin secretion.
Implicit in the paradigm of excitation-secretion coupling is that alterations in KATP currents should disrupt electrical signaling in the -cell and, thereby, alter insulin release. Specifically, decreased metabolic flux, decreased sensitivity of KATP to inhibitory ATP, or increased density of KATP channels should all lead to abnormally high KATP activity and relative hypoinsulinism. In such scenarios, a diabetic phenotype is predicted; indeed, a severe neonatal diabetic phenotype is observed in mice lacking the -cell glucokinase gene (3,4) and transgenic mice expressing -cell KATP channels with reduced sensitivity to inhibitory ATP (i.e., "overactive" KATP) (5). A series of genetic studies have recently demonstrated that heterozygous, missense mutations in the pore-forming Kir6.2 subunit of the KATP channel underlie neonatal diabetes in humans, accounting for both permanent and transient forms of the disease (6eC11). Rare in its occurrence (1:400,000 births), neonatal diabetes is usually diagnosed within the first 3 months of life and requires insulin administration to treat the hyperglycemia (12). In the milder, transient neonatal diabetes (TND), hyperglycemia usually resolves within 18 months after the diagnosis, whereas the equally common permanent neonatal diabetes (PND) requires insulin treatment for life. In both PND and TND, -cell dysfunction is likely the primary lesion. However, in TND, it is proposed that expansion of -cell mass compensates for the -cell defect and underlies the remission phase (13). In one important finding, the administration of SUs to several neonatal diabetic patients carrying Kir6.2 mutations was sufficient to treat the hyperglycemia, thereby implicating KATP "overactivity" as causal in the suppressed insulin release (6,14).
All neonatal diabeteseCassociated Kir6.2 mutations examined to date have been shown to decrease ATP sensitivity of reconstituted KATP channels (7,10,15). Some of the mutations are in putative ATP-binding residues (e.g., R50, I182, R201), but others lie well away from any predicted binding regions. These latter mutations are likely to reduce ATP sensitivity by stabilizing the open state relative to the closed state (16); in fact, this has been demonstrated for some PND mutations (15). Little attention has been paid to the sensitivity of different mutations to insulinotropic SU drugs. This is an important issue given that other mutations that alter open-state stability have been shown to dramatically affect SU sensitivity (17,18) and that SU therapy holds the promise of an alternative to insulin treatment for neonatal diabetes patients (6,14).
In the present study, we characterized Kir6.2 mutations that arguably underlie the mildest form of neonatal diabetes, TND (I182V), as well as the more severe, permanent forms of the disease (V59M, Q52R, and I296L). In all cases, mutations led to decreased ATP sensitivity. The mutations can be mechanistically classified as those that appear to directly alter ATP binding (I182V) and those that indirectly alter ATP sensitivity through changes in gating behavior (V59M, Q52R, and I296L). With respect to the latter group, a concomitant decrease in the sensitivity of reconstituted KATP channels to two clinically prescribed SUs (tolbutamide and glibenclamide) was observed in intact cells and excised patches. These findings have potentially important implications for the treatment of diabetic patients carrying Kir6.2 mutations and suggest that SU dosing may need to be considered on a case-by-case basis.
RESEARCH DESIGN AND METHODS
Molecular biology.
Kir6.2 was cloned into the EcoRI/Cla I site of pCMV6b, and the parental plasmid DNA was used to generate Kir6.2 mutations using the QuickChange Site Directed Mutagenesis kit (Stratagene, La Jolla, CA). SUR1 was cloned into the pECE expression vector. The nucleotide sequences of the mutant Kir6.2 constructs were verified by fluorescence-based cycle sequencing using AmpliTaq DNA polymerase FS (PerkinElmer, Foster City, CA) and an ABI Prism DNA sequencer (PerkinElmer).
Expression of KATP channels in COSm6 cells.
COSm6 cells were plated at a density of 2.5 x 105 cells/well (30-mm, six-well dishes) and cultured in Dulbecco’s modified Eagle’s medium with 10 mmol/l glucose supplemented with FCS (10%). The following day, a transfection cocktail was prepared by adding 6 e蘬 of FuGENE 6 Transfection Reagent (Roche, Indianapolis, IN), 600 ng each of pCMV6b-Kir6.2 and pECEeCgreen fluorescent protein plasmid, and 1 e pECE-SUR1 (per well). After a 45-min incubation at room temperature, the cocktail was added directly to the growth medium on the plated cells. Cells were assayed for KATP currents by patch-clamp measurements 2eC4 days after transfection.
Rubidium flux experiments.
86RbCl (1 mCi/ml) was added in fresh growth medium 48 h after transfection, and cells were incubated for an additional 24 h before Rb+ flux was assayed. For efflux measurements, cells were preincubated for 2 min at 25°C in Krebs-Ringer solution with or without metabolic inhibitors (2.5 e/ml oligomycin plus 1 mmol/l 2-deoxy-D-glucose [Sigma, St. Louis, MO]). At selected time points, the solution was aspirated from the wells and replaced with fresh solution. The 86Rb+ in the aspirated solution was counted in a scintillation solution. Rb+ efflux curves were fit by assuming that a single non-KATP conductance was active in untransfected cells, determined as
1
where exp is an exponential function, t is time, and k1 is the rate constant for efflux through the endogenous, non-KATP pathway.
In transfected cells, efflux curves were fit by assuming that both a KATP and a non-KATP conductance pathway were active in parallel and were determined as
2
where k1 is the rate constant for efflux through the endogenous, non-KATP pathway (obtained from untransfected cells by Eq. 1) and k2 is the rate constant for efflux through the KATP pathway. k2 is then directly proportional to the KATP conductance.
Patch-clamp measurements.
Patch-clamp studies were performed at room temperature in an oil-gate chamber that allowed the solution bathing the exposed surface of the isolated patch to be changed rapidly. The standard bath (intracellular) and pipette (extracellular) solution (K-INT) used in these experiments was comprised of 140 mmol/l KCl, 10 mmol/l K-HEPES, and 1 mmol/l K-EGTA (pH 7.3). Microelectrode resistance was typically 0.5eC1 M when filled with K-INT solution. Membrane patches were voltage clamped with an Axopatch 1B patch-clamp amplifier (Axon, Foster City, CA). Phosphatidylinositol 4,5-bisphosphate (PIP2; 5 e/ml in K-INT) was bath sonicated in ice for 30 min before being used. Glibenclamide was dissolved as a 1 mol/l stock solution in DMSO and diluted to <1% DMSO in K-INT. Tolbutamide was prepared as a 100 mmol/l stock solution in 150 mmol/l KOH and diluted to the appropriate working concentration in K-INT (pH 7.3). PIP2 was obtained from Boehringer Mannheim. Tolbutamide, glibenclamide, and ATP were purchased from Sigma. All currents were measured at a membrane potential of eC50 mV (pipette voltage = +50 mV). Data were normally filtered at 0.5eC3 kHz, and signals were digitized at 22 kHz (Neurocorder; Neurodata, New York, NY). Experiments were digitized into a microcomputer using Clampex 8.2 software (Axon), and off-line analysis was performed using Clampfit 8.2 (Axon) and Microsoft Excel programs. Wherever possible, data are presented as means ± SE. Microsoft Solver was used to fit data by a least-squares algorithm.
Estimation of Po,zero.
Two approaches, one indirect and one direct, were used to estimate the open probability in zero ATP (Po,zero) after isolated membrane patches containing multiple channels were excised (19).
PIP2 method (indirect).
PIP2 was added to the patch until the current reached a saturating level (IPIP2). This was assumed to represent a maximum Po,zero of 0.9 (20). The fractional change in current was calculated as
3
where Iinitial is the initial current. Po,zero was then estimated as
4
Noise analysis method (direct).
In addition, the mean Po,zero was estimated from stationary fluctuation analysis of macroscopic currents (21,22) on short (<1 s) recordings of currents in zero or 5 mmol/l ATP or 10 mmol/l spermine (for estimation of ATP-independent noise). Currents (corresponding to 1eC1000 channels) were filtered at 1 kHz and digitized at 3 kHz with 12-bit resolution. The mean patch current (I) and variance () (2) in the absence of ATP were obtained by subtracting the mean current and variance in 5 mmol/l ATP or 10 mmol/l spermine (i.e., assuming all channels were fully closed or blocked). The single channel current (i) was assumed to be eC3.75 pA at eC50 mV, corresponding to a wild-type single channel conductance of 75 pS (23). Po,zero was then estimated as
5
Three-state model for KATP channel activity.
Detailed kinetic analysis of KATP channel activity predicts complex models, but a simplified three-state model (19,20) can account qualitatively and quantitatively for the essential features of macroscopic KATP channel activity. Accordingly, we examined whether this simple model (Fig. 1), in accounting for the effects of mutations on gating behavior and nucleotide sensitivity, could be used predictively to account for observed changes in drug sensitivity.
The ATP binding affinity (KA) and open-close state equilibrium (KCO) for wild-type channels were assumed to be 3 eol/l and 0.60, respectively. To obtain predictions of behavior for mutant channels (as in Fig. 8), only one or the other parameter was changed. For I182V, it was necessary to change KA (increased by sixfold) without changing KCO. For Q52R, V59M, and I296L channels, it was necessary to change KCO (increased by 40-, 6-, and 600-fold, respectively) without changing KA.
Dose-response curves for SU (glibenclamide and tolbutamide) inhibition were described by the sum of two Hill equations (reflecting two components of inhibition):
6
where Grel is the relative conductance, f is the fraction of high-affinity block, and K1 and K2 are the half-maximal concentration of high- and low-affinity block, respectively.
It is clear that the relative magnitude of the two components was changed by the mutations as was the apparent affinities. We postulated, based on previous findings (17,18), that these changes resulted directly from changes in open-state stability. As shown in Fig. 8, to model the effects of different mutations on SU sensitivity, it was necessary to assume only that the fraction of high-affinity block (f) and the half-maximal blocking concentrations (K1 and K2) of the high- and low-affinity components were each dependent on the open-state stability (defined by KCO), according to the following three formulas:
7
where x = 20 or 1 for glibenclamide (Fig. 2C) or tolbutamide (Fig. 2D) inhibition curves, respectively.
8
where K1(0) is equal to 1 eol/l or 10 nmol/l for glibenclamide or tolbutamide inhibition, respectively.
9
where K2(0) is much greater than 10 or 3 mmol/l for glibenclamide or tolbutamide inhibition, respectively.
RESULTS
Variable on-cell activity correlates with variable ATP sensitivity.
To better understand the spectrum of KATP-induced neonatal diabetes and its underlying mechanisms, we chose to characterize the TND-causing I182V mutation and the three mutations implicated in more severe PND (Q52R, V59M, I296L) (Fig. 2). Mutant Kir6.2 subunits were coexpressed with the wild-type SUR1 subunit, and channel activity in intact cells was first screened by 86Rb flux experiments. As shown in Fig. 3, 86Rb+ efflux from cells transfected with wild-type KATP channels (Kir6.2 + SUR1) was low under basal conditions and was significantly activated by metabolic inhibition with oligomycin and 2-deoxyglucose to lower cellular ATP/ADP. In contrast, Rb+ efflux from cells transfected with mutant Kir6.2 subunits was high under basal conditions and showed very little increase, if any, with metabolic inhibition. To obtain quantitative estimates of conductances, a simple two-pathway (KATP + non-KATP) model for Rb+ efflux was assumed (see RESEARCH DESIGN AND METHODS). Metabolic inhibition reduced the Rb+ efflux rate in untransfected cells; Rb+ efflux time courses in basal and metabolically inhibited conditions for untransfected cells were thus fit by Eq. 1, which assumes a single non-KATP conductance is active in each condition. The rate constants thus obtained were used to estimate rate constants for KATP conductance under basal conditions and metabolic inhibition in transfected cells using Eq. 2. These rate constants were then assumed to be directly proportional to KATP conductance. Figure 4A shows the estimated KATP conductance under basal and metabolically inhibited conditions for each transfected construct. Maximal activated conductance was similar for all constructs except I296L. This mutation gave consistently lower fluxes, a point considered below (see DISCUSSION), but the basal flux was almost maximal.
We next examined KATP channel activity directly in voltage-clamped membrane patches from transfected cells. For each patch, we estimated the current in basal physiological conditions (on-cell current) and the measured maximal KATP current after patch excision (data not shown). In Fig. 4B, relative KATP current activation on-cell from patch-clamp analysis is plotted against the relative KATP activation in intact cells estimated by Rb+ efflux for each mutation. Both measures give correlated activities, suggesting that each method reflects the degree of overactivity in intact cells. To examine the correlation between increased KATP activity in intact cells and ATP sensitivity, the latter was measured directly in excised membrane patches (Fig. 5). All mutant KATP channels exhibited a significant decrease in ATP inhibition, ranging from twofold (V59M) to >100-fold (I296L) (Table 1). Channel overactivity in intact cells correlated reasonably well with reduction in ATP sensitivity, but a direct correlation between ATP insensitivity and severity of the neonatal diabetes was not absolute, as the TND-causing I182V mutation was approximately twofold less sensitive than V59M, a mutation that is associated with severe, syndromic PND (see DISCUSSION). It is conceivable that channel trafficking was altered by neonatal diabetes mutations so that a higher current may partly reflect higher expression levels. As shown in Table 1, patch current density and calculated channel density in this recombinant system were not significantly different between wild-type and mutant channels.
Mutations may affect ATP sensitivity by reduced ATP binding or indirectly by changes in the relative stability of the open state.
Mutations could mechanistically reduce ATP sensitivity directly by reducing ATP affinity or indirectly by affecting the intrinsic opening ability. Previous studies have characterized mutations throughout Kir6.2 that stabilize the open state of the channel relative to the closed state, even in the absence of ATP, and thereby decrease the apparent ATP sensitivity (19,20,24,25). To elucidate potential mechanisms in neonatal diabetes mutations, we measured the open probability in the absence of ATP (Po,zero), using two independent methods: PIP2 activation (19) and nonstationary noise analysis (22) (see RESEARCH DESIGN AND METHODS). Each method gave very similar results (see Fig. 8A). Po,zero of wild-type channels was 0.4 and was not altered by the TND-associated I182V mutation (which reduced ATP sensitivity approximately fivefold), consistent with the mutation directly affecting ATP binding. However, each of the PND-associated mutations caused a significant increase in Po,zero (Table 1), indicative of an indirect effect on ATP sensitivity. This differential mechanistic basis for channel activation could have profound consequences, both for the severity of the resultant neonatal diabetes and for the possibilities for SU therapy (see DISCUSSION).
SU sensitivity is reduced in parallel with increases in open-state stability.
Previously, it has been shown that increasing the open-state stability of KATP, either through mutagenesis or the application of PIP2, functionally "uncouples" the regulatory SUR1 subunit from the ion-conducting Kir6.2 subunit (17,18). This uncoupling is reflected as a reduction or loss of high-affinity SU inhibition through the SUR1 subunit. For Kir6.2 mutations that increase the open-state stability, we may therefore expect to find a concomitant decrease of high-affinity SU block in addition to a decrease in apparent ATP affinity. To test this possibility, we examined SU sensitivity, in intact cells using 86Rb+ flux measurements and in inside-out patch-clamp experiments.
As shown in Fig. 6, glibenclamide almost completely inhibited Rb+ efflux through wild-type channels at clinically effective doses (i.e., in the nanomolar range). It has previously been demonstrated that SU inhibition in excised membrane patches is biphasic, with high- and low-affinity components (17,26). Although a biphasic sensitivity has not been previously demonstrated in intact tissue, it is evident in the SU sensitivity of mutant KATP channels in intact COS cells (Fig. 6), where Rb+ efflux is inhibited to a plateau level by micromolar glibenclamide. For mutants that exhibit increased open-state stability, the high-affinity component decreases (Q52R, V59M, I296L) and the affinity itself decreases (Q52R, I296L). The highest open-state stability mutant (I296L) is almost completely insensitive to glibenclamide in the nanomolar range, but a reduced high-affinity component and lower affinity is also quite clear for the Q52R mutant.
Glibenclamide is a hydrophobic drug (27), and inhibition of KATP channels in excised patches was largely irreversible over the experimental period. However, tolbutamide, a first generation SU, is more hydrophilic, can be rapidly reversed, and has been a useful model SU in excised membrane patch experiments. As shown in Fig. 7, wild-type channels showed typical biphasic block by tolbutamide in inside-out membrane patches (17,26), and sensitivity was unaltered in the I182V mutation (Fig. 7B). This finding is again consistent with the interpretation that the I182V substitution directly alters ATP binding/transduction but does not affect gating (28), and thus the mutant channels remain functionally "coupled" to SUR1.
In sharp contrast, functional uncoupling from the SUR1 subunit is again reflected by a decrease in high-affinity fractional SU block in each of the mutations that increase the open-state stability (Q52R, V59M, I296L) (Table 1). It is interesting that I296L channels, which show the highest open-state stability, are completely insensitive to inhibition by tolbutamide at concentrations as high as 10 mmol/l. However, when wild-type Kir6.2 and I296L subunits were coexpressed to generate heterozygous channels, the SU sensitivity was partially restored, but with a significant decrease in the high-affinity component of block (17% high-affinity tolbutamide block for heterozygous I296L:wild-type channels vs. 55% block for homomeric wild-type channels) (Fig. 7B).
DISCUSSION
Reduced ATP sensitivity underlies both TND and PND.
It is now clear that mutations in the Kir6.2 subunit of KATP represent a common cause of neonatal diabetes in humans (6eC11). To better understand the mechanism(s) by which these mutations alter excitation-secretion coupling and their effect on SU sensitivity, we chose to characterize Kir6.2 mutations implicated in the mildest, transient form of the disease (I182V) (10) and those associated with more severe PND (Q52R, V59M, I296L) (6eC9). In all cases, the sensitivity of homomeric mutant KATP to inhibitory ATP was reduced relative to wild-type KATP, with the following order of sensitivity: WT > V59M > I182V > Q52R > I296L. The syndromic PND-linked V59M mutation exhibited comparable if not greater sensitivity to ATP than did the TND-associated I182V (K1/2ATP = 15 [V59M] and 31 [I182V] eol/l), reflected in a similar degree of "overactivity" in intact cells (Fig. 4B). Assuming the behavior of reconstituted channels reflects KATP activity in the -cell, these data do not demonstrate an absolute correlation between the severity of the diabetes (TND, PND, and syndromic PND) and the degree of ATP sensitivity. Such a positive correlation has been previously suggested based on a comparison of the less severe functional defects of a PND-associated mutation (R201C) with syndromic PND-associated mutations (Q52R, V59G) (15). Taken together, these data suggest a complex phenotype in which the primary defect is at the level of the -cell and altered insulin secretion, but they also indicate that additional genetic and/or environmental factors may modulate the severity of the disease (TND, PND, and syndromic PND). It is interesting that KATP in skeletal muscle (Kir6.2 + SUR2A) is implicated in peripheral insulin sensitivity based on the observation that both Kir6.2- and SUR2A-null mice exhibit increased glucose uptake in skeletal muscle and adipose tissue (29,30). The mouse models, therefore, predict that "overactive" KATP, as observed in neonatal diabetes, may suppress glucose uptake and contribute to the hyperglycemia. Whether neonatal diabetes patients, in addition to having suppressed insulin release, also exhibit insulin insensitivity remains to be determined.
An important distinction between the present study and that of Proks et al. (15) is that the order of sensitivity in Proks et al.’s report was based on the ATP sensitivity of heteromeric channels (generated by the expression of channels from equal amounts of wild-type Kir6.2 and mutant Kir6.2 cRNA), whereas our analysis is of homomeric channels. We have found that in similar experiments with mixed expression of wild-type and mutant Kir6.2 subunits, the rank order of mutant ATP sensitivity is unaltered (data not shown). However, a quantitative analysis of mixed expression is very complicated, given the fact that six distinct combinations of wild-type and mutant subunits will generate six different channel types, each with its own ATP sensitivity. Whether mutant alleles of Kir6.2 will be expressed equally in vivo and in different tissues is also not known. Given these complexities, the analysis of homomeric channels may be the most straightforward approach to understanding mutational effects on channel properties.
Finally, the I296L mutant channel consistently exhibited a lower activity in Rb+ efflux experiments, despite a channel density that was similar to that of wild-type channels in excised patches. This finding may reflect a lower transfection efficiency. However, because the I296L mutation resulted in the highest on-cell activity of all the neonatal diabetes mutants, it is also possible that this elevated K+ conductance was toxic to the cells, so that we were experimentally selecting for a subpopulation of cells that are resistant to elevated K+ conductance in excised patch experiments.
Differential SU sensitivity of neonatal diabetes mutations.
SUs block pancreatic KATP currents and stimulate insulin release. The high-affinity component of block by glibenclamide (IC50 1 nmol/l) or tolbutamide (IC50 1 eol/l) results from binding to the SUR1 subunit (26), whereas the Kir6.2 subunit underlies the low-affinity component (IC50 1,000 x high affinity). A direct correlation between gating behavior of the channel and SU sensitivity has been previously demonstrated (17,18). The consequence of such a relation is that mutations in Kir6.2 that increase the stability of the open state not only reduce the apparent ATP sensitivity, but concomitantly decrease the high-affinity SU block. This is interpreted as a functional "uncoupling" of the regulatory SUR1 subunit from the Kir6.2 channel (17).
We thus predict that PND mutations that increase open-state stability (V59M, Q52R, and I296L) should also reduce SU sensitivity. This hypothesis is consistent with the previous observation that the block of whole-cell K+ currents was not complete in 500 eol/l tolbutamide for PND mutant KATP (15). Indeed, the gating mutations Q52R and V59M but not the ATP-binding mutant I182V exhibited a decrease in the fractional block by the high-affinity component. The Q52R and I296L mutants also exhibited a significant decrease in the sensitivity of the low-affinity block, with I296L mutant channels being completely refractory to inhibition by tolbutamide 10 mmol/l. This implies that in addition to a reduced SU sensing through SUR1, the I296L and Q52R mutations also reduce SU action through the low-affinity site on Kir6.2, either directly or allosterically.
A mechanistic model for nucleotide and SU sensitivity.
Diabetes-causing Kir6.2 mutations can be classified mechanistically as those that directly alter ATP binding and/or transduction of the ATP-mediated signal (I182V) versus those that indirectly affect ATP affinity by increasing the open-state stability (Q52R, V59M, I296L). The data are consistent with recent characterizations of other PND mutations that affect ATP binding (R201C, R201H, I182V) or alter gating behavior, leading to a secondary decrease in ATP sensitivity (V59G, Q52R) (10,15).
Models of KATP channel gating and nucleotide and SU sensitivity have been developed. To have predictive use, such models should account for channel gating (in the absence of ATP), nucleotide sensitivity, and drug sensitivity, and by adjusting relevant parameters, should be able to account for the effects of mutations on channel function. The most complex models (20,28,31) account for the tetrameric nature of the channel and could potentially account for heteromeric channels consisting of wild-type and mutant subunits. However, a relatively simple three-state model (20) (see RESEARCH DESIGN AND METHODS, Fig. 1) can account for all of the pertinent features of macroscopic channel activity and has predictive qualities for the interpretation of nucleotide and drug sensitivities of neonatal diabetes mutants. We therefore postulated that mutational effects on SU sensitivity result directly from changes in open-state stability (17,18).
As shown in Fig. 8, the simplified three-state model quantitatively predicted the effects of neonatal diabetes mutations on each of the three features, Po,zero, ATP sensitivity, and SU sensitivity, by adjusting only two parameters, KA and KCO, which are the equilibrium constants that describe ATP affinity and open-state stability, respectively. By assuming that both the affinity and the fraction of high-affinity SU block are dependent on only KCO, the steady-state SU sensitivities of wild-type and mutant channels can be predicted. As discussed below, this analytical approach may be a useful predictor of SU efficacy in human therapy trials.
Implications for SU therapy of neonatal diabetes.
Recently, SUs have been shown to be effective, at least in the short-term, in maintaining control of blood glucose levels in several PND patients in which Kir6.2 mutations are causal (F331I, V59M, and R201H) (6,14). Given that blood glucose levels and HbA1c values were stable after discontinuation of insulin in these patients, it has been proposed that oral SUs can be an alternative therapy to subcutaneous insulin injections in patients with PND because of Kir6.2-activating mutations. The present results indicate, however, that the efficacy of SU therapy is likely to depend on the nature of the Kir6.2 mutation. Assuming that the SU sensitivity of reconstituted channels reflects the sensitivity of channels in vivo, carriers of the diabetes-causing I296L and similarly behaving Kir6.2 mutations may be less responsive to SU therapy. In one interesting case, a PND patient with the syndromic V59M mutation was responsive to SU therapy and discontinued insulin injections (6), thereby suggesting that despite a significant decrease in sensitivity of homomeric V59M mutant channels to SU inhibition, sufficient high-affinity inhibition was still able to induce insulin release. In the original report by Gloyn et al. (7), it was observed that a parent of the R201H proband was diagnosed as a baby with diabetes and had been effectively treated with tolbutamide for 46 years. The mutation R201H causes a significant loss of ATP sensitivity (7), but it is predicted to be an ATP-binding site mutation (16,32,33) and thus may not show loss of SU sensitivity.
Functional characterization of additional PND-associated mutations in Kir6.2, as well as additional clinical studies, will provide essential data regarding the efficacy of SU therapy in treating KATP-induced diabetes and the possibility of tailoring individual therapy based on the underlying Kir6.2 mutation. The present results suggest that careful attention should be paid to the degree to which neonatal diabetes mutations cause a loss of ATP sensitivity by an increase in open-state stability. The greater the mutational effect on open-state stability is, the higher the SU dose necessary to achieve a given degree of channel inhibition will be. Thus, to achieve the same therapeutic effect (i.e., to obtain sufficient closure of KATP under appropriate physiological conditions), proportionally higher doses of an SU may be clinically required.
ACKNOWLEDGMENTS
This work was supported by National Institutes of Health Diabetes Research and Training Center Grant DK-20579 and National Institutes of Health Grant DK069445 (R01) at Washington University and an American Diabetes Association-Takeda Pharmaceuticals Mentor-Based Minority Postdoctoral Fellowship Program (to M.S.R.).
FOOTNOTES
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
KATP channel, ATP-sensitive K+ channel; PIP2, phosphatidylinositol 4,5-bisphosphate; PND, permanent neonatal diabetes; SU, sulfonylurea; TND, transient neonatal diabetes
REFERENCES
Ashcroft FM, Rorsman P: ATP-sensitive K+ channels: a link between B-cell metabolism and insulin secretion. Biochem Soc Trans18 :109 eC111,1990
Aguilar-Bryan L, Nichols CG, Wechsler SW, Clement JP 4th, Boyd AE 3rd, Gonzalez G, Herrera-Sosa H, Nguy K, Bryan J, Nelson DA: Cloning of the beta cell high-affinity sulfonylurea receptor: a regulator of insulin secretion. Science268 :423 eC426,1995
Terauchi Y, Sakura H, Yasuda, Iwamoto K, Takahashi N, Ito K, Kasai H, Suzuki H, Ueda O, Kamada N, et al.: Pancreatic beta-cell-specific targeted disruption of glucokinase gene: diabetes mellitus due to defective insulin secretion to glucose. J Biol Chem270 :30253 eC30256,1995
Ishihara H, Tashiro F, Ikuta K, Asano T, Katagiri H, Inukai K, Kikuchi M, Yazaki Y, Oka Y, Miyazaki J: Inhibition of pancreatic beta-cell glucokinase by antisense RNA expression in transgenic mice: mouse strain-dependent alteration of glucose tolerance. FEBS Lett371 :329 eC332,1995
Koster JC, Marshall BA, Ensor N, Corbett JA, Nichols CG: Targeted overactivity of beta cell K(ATP) channels induces profound neonatal diabetes. Cell100 :645 eC654,2000
Sagen JV, Raeder H, Hathout E, Shehadeh N, Gudmundsson K, Baevre H, Abuelo D, Phornphutkul C, Molnes J, Bell GI, Gloyn AL, Hattersley AT, Molven A, Sovik O, Njolstad PR: Permanent neonatal diabetes due to mutations in KCNJ11 encoding Kir6.2: patient characteristics and initial response to sulfonylurea therapy. Diabetes53 :2713 eC2718,2004
Gloyn AL, Pearson ER, Antcliff JF, Proks P, Bruining GJ, Slingerland AS, Howard N, Srinivasan S, Silva JM, Molnes J, Edghill EL, Frayling TM, Temple IK, Mackay D, Shield JP, Sumnik Z, van Rhijn A, Wales JK, Clark P, Gorman S, Aisenberg J, Ellard S, Njolstad PR, Ashcroft FM, Hattersley AT: Activating mutations in the gene encoding the ATP-sensitive potassium-channel subunit Kir6.2 and permanent neonatal diabetes [erratum in N Engl J Med351 :1470 ,2004 ]. N Engl J Med350 :1838 eC1849, 2004
Vaxillaire M, Populaire C, Busiah K, Cave H, Gloyn AL, Hattersley AT, Czernichow P, Froguel P, Polak M: Kir6.2 mutations are a common cause of permanent neonatal diabetes in a large cohort of French patients. Diabetes53 :2719 eC2722,2004
Massa O, Iafusco D, D’Amato E, Gloyn AL, Hattersley AT, Pasquino B, Tonini G, Dammacco F, Zanette G, Meschi F, Porzio O, Bottazzo G, Crino A, Lorini R, Cerutti F, Vanelli M, Barbetti F: KCNJ11 activating mutations in Italian patients with permanent neonatal diabetes. Hum Mutat25 :22 eC27,2005
Gloyn AL, Reimann F, Girard C, Edghill EL, Proks P, Pearson ER, Temple IK, Mackay DJ, Shield JPH, Freedenberg D, Noyes K, Ellard S, Ashcroft FM, Gribble FM, Hattersley AT: Relapsing diabetes can result from moderately activating mutations in KCNJ11. Hum Mol Genet14 :925 eC934,2005
Gloyn AL, Edghill EL, Pearson ER, Mackay D, Temple IK, Shield J, Noyes K, Freedenberg D, Gillispie KM, Lambert AP, Gale EA, Ellard S, Hattersley A: Multiple subtypes of diabetes are associated with activating mutations in KCNJ11, which encodes the Kir6.2 subunit of the -cell ATP-sensitive potassium channel [online abstract]. Available at http://scientificsessions.diabetes.org/Abstracts/index.cfmfuseaction=Locator.DisplaysAbstractSearch
Polak M, Shield J: Neonatal and very-early-onset diabetes mellitus. Semin Neonatol9 :59 eC65,2004
Ma D, Shield JP, Dean W, Leclerc I, Knauf C, Burcelin RR, Rutter GA, Kelsey G: Impaired glucose homeostasis in transgenic mice expressing the human transient neonatal diabetes mellitus locus, TND. J Clin Invest114 :339 eC348,2004
Zung A, Glaser B, Nimri R, Zadik Z: Glibenclamide treatment in permanent neonatal diabetes mellitus due to an activating mutation in Kir6.2. J Clin Endocrinol Metab89 :5504 eC5507,2004
Proks P, Antcliff JF, Lippiat J, Gloyn AL, Hattersley AT, Ashcroft FM: Molecular basis of Kir6.2 mutations associated with neonatal diabetes plus neurological features. Proc Natl Acad Sci U S A101 :17539 eC17544,2004
Enkvetchakul D, Nichols CG: Gating mechanism of KATP channels: function fits form. J Gen Physiol122 :471 eC480,2003
Koster JC, Sha Q, Nichols CG: Sulfonylurea and K(+)-channel opener sensitivity of K(ATP) channels: functional coupling of Kir6.2 and SUR1 subunits. J Gen Physiol114 :203 eC213,1999
Reimann F, Tucker SJ, Proks P, Ashcroft FM: Involvement of the N-terminus of Kir6.2 in coupling to the sulphonylurea receptor. J Physiol (Lond)518 :325 eC336,1999
Cukras CA, Jeliazkova I, Nichols CG: The role of NH(2)-terminal positive charges in the activity of inward rectifier K(ATP) channels. J Gen Physiol120 :437 eC446,2002
Enkvetchakul D, Loussouarn G, Makhina E, Shyng SL, Nichols CG: The kinetic and physical basis of K(ATP) channel gating: toward a unified molecular understanding. Biophys J78 :2334 eC2348,2000
Neher E, Stevens CF: Conductance fluctuations and ionic pores in membranes. Annu Rev Biophys Bioeng6 :345 eC381,1977
Sigworth FJ: The variance of sodium current fluctuations at the node of Ranvier. J Physiol307 :97 eC129,1980
Shyng S, Ferrigni T, Nichols CG: Control of rectification and gating of cloned KATP channels by the Kir6.2 subunit. J Gen Physiol110 :141 eC153,1997
Trapp S, Proks P, Tucker SJ, Ashcroft FM: Molecular analysis of ATP-sensitive K channel gating and implications for channel inhibition by ATP. J Gen Physiol112 :333 eC349,1998
Drain P, Li L, Wang J: KATP channel inhibition by ATP requires distinct functional domains of the cytoplasmic C terminus of the pore-forming subunit. Proc Natl Acad Sci U S A95 :13953 eC13958,1998
Gribble FM, Tucker SJ, Ashcroft FM: The interaction of nucleotides with the tolbutamide block of cloned ATP-sensitive K+ channel currents expressed in Xenopus oocytes: a reinterpretation. J Physiol504 :35 eC45,1997
Zunkler BJ, Trube G, Panten U: How do sulfonylureas approach their receptor in the B-cell plasma membrane Naunyn Schmiedebergs Arch Pharmacol340 :328 eC332,1989
Li L, Wang J, Drain P: The I182 region of k(ir)6.2 is closely associated with ligand binding in K(ATP) channel inhibition by ATP. Biophys J79 :841 eC852,2000
Miki T, Minami K, Zhang L, Morita M, Gonoi T, Shiuchi T, Minokoshi Y, Renaud JM, Seino S: ATP-sensitive potassium channels participate in glucose uptake in skeletal muscle and adipose tissue. Am J Physiol Endocrinol Metab283 :E1178 eCE1184,2002
Chutkow WA, Samuel V, Hansen PA, Pu J, Valdivia CR, Makielski JC, Burant CF: Disruption of Sur2-containing K(ATP) channels enhances insulin-stimulated glucose uptake in skeletal muscle. Proc Natl Acad Sci U S A98 :11760 eC11764,2001
Enkvetchakul D, Loussouarn G, Makhina E, Nichols CG: ATP interaction with the open state of the K(ATP) channel. Biophys J80 :719 eC728,2001
Shyng SL, Cukras CA, Harwood J, Nichols CG: Structural determinants of PIP(2) regulation of inward rectifier K(ATP) channels. J Gen Physiol116 :599 eC608,2000
Ribalet B, John SA, Weiss JN: Molecular basis for Kir6.2 channel inhibition by adenine nucleotides. Biophys J84 :266 eC276,2003(Joseph C. Koster, Maria S)