Tungstate activates BK channels in a β subunit- and Mg2+-dependent manner: relevance for arterial vasodilatation

MS # CVR-2011-1135R1 Tungstate activates BK channels in a β subunit- and Mg 2+ -dependent manner: relevance for arterial vasodilation. Ana I. Fernández-Mariño 1 , Cristina Porras-González 2 , Patricia González-Rodríguez 2 , Jana Selent 3 , Manuel Pastor 3 , Juan Ureña 2 , Antonio Castellano 2 , Miguel A. Valverde 1 , and José M. Fernández- Fernández 1 * 1 Laboratori de Fisiologia Molecular i Canelopaties, Departament de Ciències Experimentals i de la Salut, Universitat Pompeu Fabra, Barcelona, Spain; 2 Instituto de Biomedicina de Sevilla, Hospital Universitario Virgen del Rocío/CSIC/Universidad de Sevilla, Spain; and 3 Programa d'Investigació en Informàtica Biomèdica (GRIB), IMIM-Hospital del Mar, Universitat Pompeu Fabra, Barcelona, Spain. *Corresponding Author: Dr. José Manuel Fernández-Fernández, Laboratori de Fisiologia Molecular i Canelopaties, Department de Ciències Experimentals i de la Salut, Universitat Pompeu Fabra, C/ Dr. Aiguader 88, Barcelona 08003, Spain. Phone: 34 93 316 08 54. Fax: 34 93 316 09 01. Email: [email protected] Words: 6496. Running title: BK channels role in tungstate-induced vasodilation. MS # CVR-2011-1135R1 2 Tungstate activates BK channels in a β subunit- and Mg 2+ -dependent manner: relevance for arterial vasodilation. Ana I. Fernández-Mariño, Cristina Porras-González, Patricia González-Rodríguez, Jana Selent, Manuel Pastor, Juan Ureña, Antonio Castellano, Miguel A. Valverde, and José M. Fernández- Fernández Abstract Aims: Tungstate reduces blood pressure in experimental animal models of both hypertension and metabolic syndrome, although the underlying mechanisms are not fully understood. Given that the large-conductance voltage- and Ca 2+ -dependent K + (BK) channel is a key element in the control of arterial tone, our aim was to evaluate whether BK channel modulation by tungstate can contribute to its antihypertensive effect. Methods and Results: Patch-clamp studies of heterologously expressed human BK channels ( 1-4 subunits) revealed that cytosolic tungstate (1 mM) induced a significant left-shift (~ 20 mV) in the voltage-dependent activation curve only in BK channels containing β 1 or β 4 subunits but reduced the amplitude of K + currents through all BK channels tested. The β 1 -dependent activation of BK channels by tungstate was enhanced at cytosolic Ca 2+ levels reached during myocyte contraction, and prevented either by removal of cytosolic Mg 2+ or by mutations rendering the channel insensitive to Mg 2+ . Lower concentration of tungstate (0.1 mM) induced voltage-dependent activation of the vascular BKβ 1 channel without reducing current amplitude, and, consistently, exerted a vasodilatory action on wild-type but not β 1 -knockout mouse arteries precontracted with endothelin-1. Conclusion: Tungstate activates BK channels in a β subunit- and Mg 2+ -dependent manner and induces vasodilation only in mouse arteries that express the BK β 1 subunit. Keywords: BK channel; tungstate; arteries; vasodilation; hypertension. MS # CVR-2011-1135R1 3 Introduction Arterial resistance, which is mainly controlled by the basal tone of the vascular smooth muscle cells (VSMC), determines the rate of blood flow to tissues. Both circulating factors and those resident at the vessel wall participate in the establishment of this tone and their malfunction is typically associated to functional and structural vessels’ alterations in patients suffering from essential hypertension 1 . Therefore it is important to identify the molecular and genetic elements participating in the control of vascular tone and their possible use as targets with therapeutic potential. Among the factors controlling vascular tone, voltage-gated ion channels mediating Ca 2+ entry into the VSMC and their regulatory mechanisms are central to this process 2 . One such key regulatory mechanism involves the large-conductance voltage- and Ca 2+ -activated K + (BK) channel. BK channel couples local increases in intracellular Ca 2+ to augmented channel activity, membrane hyperpolarization (preventing the activation of L-type Ca 2+ channels) and vascular relaxation 3 . Medications already exist that target different mechanisms involved in blood pressure regulation 4 although its normalization is not always satisfactory, and none of them target BK channels. Compared to currently used hypertension treatments, such as Ca 2+ antagonists, BK channel openers are expected to potentiate an existing physiological regulatory system that acts as a negative feedback mechanism to limit membrane depolarization, L-type Ca 2+ channel opening and the subsequent VSMC contraction 3 . BK channels, formed by α subunit tetramers, are encoded by a single gene, SLO (KCNMA1), and are activated by membrane depolarization and binding of intracellular Ca 2+ and Mg 2+ , being their effect additive and involving independent pathways 5,6 . Accordingly, distinct binding sites for physiological concentrations of Ca 2+ (in the micromolar range) and Mg 2+ (in the millimolar range) have been identified in the pore-forming α channel subunit 6-8 . Mg 2+ blocks the pore of BK channels in a voltage-dependent manner 6,9 but also enhances BK channel activation by stabilizing the open states. This results in a shift of the conductance-voltage (G-V) relationship to a more negative voltage range, an effect that is independent of voltage-sensor movement or Ca 2+ binding to the channel 6,10 . Channel activation by Mg 2+ requires a divalent cation binding site that consists of amino acid side chains from MS # CVR-2011-1135R1 4 both the voltage-sensor domain (VSD) (D99 and N172) and the first cytoplasmic regulator of conductance for K + (RCK1) domain (E374 and E399) of the BK α channel subunit 8 . Depending on the tissue, BK α subunits associate with different auxiliary β subunits (β 1 - β 4 ), modifying BK subcellular distribution and gating, thereby contributing to channel diversity and involvement in different physiological processes 11 . The molecular composition of the vascular BK channel includes the regulatory β 1 subunit, whose expression increases channel sensitivity to Ca 2+ favoring its activation by low Ca 2+ sparks 12 . Moreover, the presence of the β 1 subunit is relevant to smooth muscle contraction induced by vasoactive peptides and the regulation of arterial tone and blood pressure in both animal models of hypertension and humans 13-18 . Altogether makes the vascular BKαβ 1 channel an attractive target for the development of new antihypertensive therapies. Tungstate, a compound with antidiabetic 19 and antiobesity properties 20 , has been reported to reduce blood pressure in experimental animal models of both hypertension 21,22 and metabolic syndrome 23 . The antihypertensive action of tungstate has been associated to the inhibition of the increased xanthine oxidase activity in endothelial cells of hypertensive animals 21 . Given the key role of BK channels in arterial tone control and the reported antihypertensive effect of tungstate we assessed the tungstate effect on BK channel activity and arterial contractility. MS # CVR-2011-1135R1 5 Methods Cell transfection HEK293 cells were transfected using polyethylenimine ExGen 500 (Fermentas Inc., Hanover, Maryland, USA), following the manufacturer’s instructions (seven equivalents PEI per 3.3 μg of cDNAs expressing the wild-type (WT), D99A or the N172A BK channel human α subunit alone or with one of the four different human β subunits (β 1-4 ) (1:2 ratio), together with the transfection reporter pEGFPN1). Electrophysiology Inside-out ionic currents were recorded in macropatches from EGFP-positive cells, 2-3 days after transfection 17 . For current-activation studies, membrane macropatches were clamped at 0 mV, pulsed for 150 milliseconds (or 20 ms for expressed BK channels containing the inactivating β 2 regulatory subunit) from –100 mV to +200 mV in 10-mV steps, and repolarized to –80 mV for 20 milliseconds. Experiments were performed at room temperature (22–26°C). Relative conductance was determined by measuring tail current amplitudes at –80 mV. For each patch the conductance-voltage (G-V) relationship was fitted with the following Boltzmann equation: G/Gmax = 1/(1+exp(-(V-V 1/2 act )/k act ) where G is the value of the instantaneous tail current at each test voltage, Gmax is the maximum obtained tail current, V is the test voltage applied to the membrane, V 1/2 act is the voltage for halfmaximal current activation, and k act (an index of the minimum number of elementary charges that move through the electric field to gate the channel) is the slope factor of the Boltzmann term. The showed G-V curves represent averages from a set of patches at the indicated experimental condition. Putative tungstate binding sites at the human pore-forming α subunit of the BK channel To obtain some insight into structural features of tungstate-protein interactions, we analyzed the residue composition in the tungstate binding site of 45 available X-ray structures (see Supplementary material online for a complete list of the PDB IDs and Table S1). These studies supported the role of MS # CVR-2011-1135R1 6 divalent cations as cofactors for tungstate binding and they allowed us to suggest putative tungstate binding sites at the human BKα subunit based on an identified aspartate-lysine/lysine-aspartate (DK/KD) motif. Interestingly, the two BKα regions that include the Mg 2+ binding site 8 contain a DK/KD motif within or next to the relevant D99, N172 and E374 residues (Supplementary material online, Figures S1-S3 and Table S2). Arterial tension measurements Femoral arteries were isolated from adult male, age-matched (18-24-week-old) +/+ (WT) and −/− BKβ 1 (β 1 -knockout) C57BL/6 mice. Arterial rings were obtained after surgical procedures from animals of weight 26-28gr sacrificed by i.p. administration of a lethal dose of sodium thiopental (200 mg/kg). Arterial rings were cleaned of connective tissue, cut in rings and mounted on a small-vessel myograph (Danish Myo Technology, Aarhus, Denmark) to measure tension connected to a digital recorder (AcqKnowledge 3.8.1, Biopac System Inc.). All experiments were conducted in accordance with the Spanish legislation on protection of animals, approved by the local animal care committee (Comité Ético de Experimentación de la Universidad de Sevilla) and conformed to the Directive 2010/63/EU of the European Parliament. Statistics Data are presented as the means ± S.E.M. Statistical tests included Student’s t test, Mann-Whitney test, One Way Analysis of Variance (ANOVA) followed by a Tukey posthoc test, or Nonparametric ANOVA (Kruskal-Wallis Test) followed by a Dunn posthoc test, as appropriate. Differences were considered significant if P < 0.05. See also Supplementary material online for further details of experimental procedures. MS # CVR-2011-1135R1 7 Results Effect of tungstate on BK channels of different molecular composition To evaluate whether tungstate modulates BK channel activity, we expressed the human α subunit alone or with one of the four human regulatory β subunits (β 1-4 ) in HEK293 cells. Figure 1 (A, C, E, G, I) shows representative BK currents, recorded before and after the addition of 1 mM tungstate to a nominally Ca 2+ -free bath (intracellular) solution containing 0.7 mM Mg 2+ (a physiological concentration found in most mammalian cells 24 ). Tungstate reduced (29-39%) the amplitude of K + currents through all BK channels tested (Figure 1K). Changes in BK channel activity are best evaluated by the analysis of the G-V relationship of the measured K + currents, before (control) and after exposure to tungstate (Figure 1B, D, F, H, J). From these curves we estimated the voltage for channel half-activation (V 1/2 act ) (Figure 1L), which is a convenient parameter to study the effect of BK channels modulators since it is directly related to the energy required to open the channel. We found that tungstate only produced a significant decrease of the V 1/2 act values (by ~22 mV) for β 1 - or β 4 -containing BK channels, favoring their activation. Tungstate did not change the slope factor of the fitted G-V curves (k act ), suggesting no interference with the voltage-sensing process of BK gating (Figure 1M). We also studied whether tungstate-induced activation of BKαβ 1 channels was modified by increasing [Ca 2+ ] up to values detected near BKαβ 1 channels during VSMC contraction (~10 µM) 25 . The inhibitory effect of tungstate (1 mM) on BKβ 1 current amplitude was significantly smaller (P < 0.05) at 10 µM Ca 2+ (Figure 2A, C). BKαβ 1 V 1/2 act was significantly lower at 10 µM Ca 2+ (as previously reported) and further reduced by tungstate (~40 mV versus ~20 mV at 0 Ca 2+ ; P < 0.05) (Figure 2B, D). Tungstate-induced activation of the BKαβ 1 channel depends on its Mg 2+ -sensitivity The analysis of available structural data on tungstate-protein complexes supported the role of divalent cations as cofactors for tungstate binding (see Methods). Given the effect of tungstate on BK channel V 1/2 act in the absence of cytosolic Ca 2+ , we can discard this divalent cation as essential cofactor for MS # CVR-2011-1135R1 8 tungstate action. The other divalent cation of physiological relevance in the modulation of BK channel activity is Mg 2+ . Therefore, we next investigated its possible role as tungstate cofactor. We observed a dual modulation of human BKαβ 1 channels by Mg 2+ in the absence of cytosolic Ca 2+ similar to that reported for murine BK channels 6,10 . Millimolar concentrations of cytosolic Mg 2+ dose-dependently reduced the current amplitude at positive voltages (Figure 3A, G (white bars)). Mg 2+ also left-shifted the G-V curve of the human BKβ 1 channel (Figure 3B), thereby reducing the channel V 1/2 act , without affecting the curve slope (Figure 3H, I open circles). Tungstate (1 mM) produced similar reduction of BKαβ 1 current amplitude in the absence or presence of cytosolic Mg 2+ (Figure 3C, E, G (black bars)). However, removal of cytosolic Mg 2+ prevented the tungstate-induced decrease of BKαβ 1 channel V 1/2 act (Figure 3D, H). In the presence of 5 mM Mg 2+ , tungstate was not able to produce a further V 1/2 act reduction beyond the one induced by the divalent cation (Figure 3F, H). The Mg 2+ requirement for the tungstate effect was further tested on human BKαβ 1 channels containing mutations in residues of the  subunit VSD that contribute to Mg 2+ coordination. D99A mutation abolish and N172A mutation substantially reduce Mg 2+ -induced activation of mSlo channels 8 . Supplementary Figure S3 shows the position of these mutations and alignment of human and mouse BK channel VSD linkers containing the mutations. In accordance with previous reports on mouse BK channel, the introduction of the D99A and N172A mutations slightly modified the V 1/2 act for human BKαβ 1 channels in the absence of divalent cations (Supplementary material online, Figure S4E), and the addition of 0.7 or 5 mM cytosolic Mg 2+ still reduced the amplitude of K + currents through mutant D99A and N172A human BKαβ 1 channels at positive voltages without significant effects on the G-V curves (Supplementary material online, Figure S4). This is consistent with the idea that Mg 2+ -induced blockade and activation of BK channels occur via different binding sites. When exposed to 1 mM tungstate both D99A and N172A human BKαβ 1 channels presented a reduction in current amplitude similar to that observed for wild-type (WT) channels (Figure 4A, C, E) but lose the tungstate-induced channel activation (Figure 4B, D, F). MS # CVR-2011-1135R1 9 Tungstate-induced vasodilation of precontracted mouse arteries relies on the expression of the BK channel β 1 subunit To determine the impact of BK channel modulation by tungstate on vascular function, we tested the action of tungstate on the contractility of femoral arterial rings from both WT and BK β 1 -knockout mice. Application of 1 mM tungstate induced similar constriction of arterial rings from both WT and β 1 -knockout mice (Supplementary material online, Figure S5), an effect consistent with the considerable degree of BK current inhibition produced by such concentration of tungstate, independently of the presence of the β 1 subunit (see Figure 1A, C, K). However, lowering tungstate to 0.1 mM did not significantly affect the basal tension (Supplementary material online, Figure S5). Therefore, we evaluated whether 0.1 mM tungstate could yet modulate heterologously expressed BK channels. In the absence of β 1 , 0.1 mM tungstate produced a small reduction of BK currents (Figure 5A, E) but was ineffective in left-shifting the G-V relationship (Figure 5B) or the V 1/2 act (Figure 5F). In the presence of β 1 , 0.1 mM tungstate did not inhibit BK currents (Figure 5C), actually there was a slight increase in the overall current (Figure 5E), and induced a significant left-shift in the G-V curve (Figure 5D) resulting in a ~10 mV decrease of V 1/2 act (Figure 5F). Thus, 0.1 mM tungstate activates BKαβ 1 channels without negatively altering BKαβ 1 current amplitude. Next, we tested the effect of 0.1 mM tungstate on femoral arterial rings. Exposure of femoral arterial rings to K + depolarizing solutions produced marked and reversible contractile responses in both WT (Figure 6A) and β 1 -knockout (Figure 6B) arteries. Arterial rings were then precontracted with endothelin-1 (ET-1, 50 nM), an efficacious vasoconstrictor peptide produced by endothelial cells that increases cytosolic Ca 2+ in VSMC and has been related to the pathogenesis of hypertension (reviewed in 26 ). No significant difference (P = 0.21, Student’s t test) was found in the contractile response to ET-1 between WT (10.35 ± 0.73 mN, n = 21) and β 1 -knockout arteries (11.79 ± 0.79 mN, n = 12). Consistent with the observed β 1 -dependent induced-activation of BK channels, 0.1 mM tungstate significantly reduced arterial tension after 5 minutes of application on precontracted arterial rings from wild-type mice, a vasodilatory effect that was maximal after 20 minutes of tungstate exposure (Figure 6A, C). Such tungstate effect was reverted by the addition of the BK channel blocker paxilline (Figure 6A) and was not mimicked by the xanthine oxidase inhibitor allopurinol MS # CVR-2011-1135R1 16 References 1. Touyz RM. Molecular and cellular mechanisms regulating vascular function and structureimplications in the pathogenesis of hypertension. Can J Cardiol 2000;16:1137-1146. 2. Martens JR & Gelband CH. Ion channels in vascular smooth muscle: alterations in essential hypertension. Proc Soc Exp Biol Med 1998;218:192-203. 3. Ledoux J, Werner ME, Brayden JE, Nelson MT. Calcium-activated potassium channels and the regulation of vascular tone. Physiology (Bethesda ) 2006;21:69-78. 4. The sixth report of the Joint National Committee on prevention, detection, evaluation, and treatment of high blood pressure. Arch Intern Med 1997;157:2413-2446. 5. Cui J & Aldrich RW. Allosteric linkage between voltage and Ca 2+ -dependent activation of BK-type mslo1 K + channels. Biochemistry 2000;39:15612-15619. 6. Shi J, Cui J. Intracellular Mg 2+ enhances the function of BK-type Ca 2+ -activated K + channels. J Gen Physiol 2001;118:589-606. 7. Xia XM, Zeng X, Lingle CJ. Multiple regulatory sites in large-conductance calcium-activated potassium channels. Nature 2002;418:880-884. 8. Yang H, Shi J, Zhang G, Yang J, Delaloye K, Cui J. Activation of Slo1 BK channels by Mg 2+ coordinated between the voltage sensor and RCK1 domains. Nat Struct Mol Biol 2008;15:1152-1159. 9. Ferguson WB. Competitive Mg 2+ block of a large-conductance, Ca 2+ -activated K + channel in rat skeletal muscle. Ca 2+ , Sr 2+ , and Ni 2+ also block. J Gen Physiol 1991;98:163-181. 10. Zhang X, Solaro CR, Lingle CJ. Allosteric regulation of BK channel gating by Ca 2+ and Mg 2+ through a nonselective, low affinity divalent cation site. J Gen Physiol 2001;118:607-636. MS # CVR-2011-1135R1 17 11. Torres YP, Morera FJ, Carvacho I, Latorre R. A marriage of convenience: beta-subunits and voltage-dependent K + channels. J Biol Chem 2007;282:24485-24489. 12. Tanaka Y, Meera P, Song M, Knaus HG, Toro L. Molecular constituents of maxi KCa channels in human coronary smooth muscle: predominant alpha + beta subunit complexes. J Physiol (Lond) 1997;502:545-557. 13. Brenner R, Perez GJ, Bonev AD, Eckman DM, Kosek JC, Wiler SW et al. Vasoregulation by the beta1 subunit of the calcium-activated potassium channel. Nature 2000;407:870-876. 14. Pluger S, Faulhaber J, Furstenau M, Lohn M, Waldschutz R, Gollasch M et al. Mice with disrupted BK channel beta1 subunit gene feature abnormal Ca 2+ spark/STOC coupling and elevated blood pressure. Circ Res 2000;87:E53-E60. 15. Amberg GC, Bonev AD, Rossow CF, Nelson MT, Santana LF. Modulation of the molecular composition of large conductance, Ca 2+ activated K + channels in vascular smooth muscle during hypertension. J Clin Invest 2003;112:717-724. 16. Amberg GC & Santana LF. Downregulation of the BK channel beta1 subunit in genetic hypertension. Circ Res 2003;93:965-971. 17. Fernández-Fernández JM, Tomás M, Vázquez E, Orio P, Latorre R, Sentí M et al. Gain-of- function mutation in the KCNMB1 potassium channel subunit is associated with low prevalence of diastolic hypertension. J Clin Invest 2004;113:1032-1039. 18. Sentí M, Fernández-Fernández JM, Tomás M, Vázquez E, Elosua R, Marrugat J et al. Protective effect of the KCNMB1 E65K genetic polymorphism against diastolic hypertension in aging women and its relevance to cardiovascular risk. Circ Res 2005;97:1360-1365. 19. Barberà A, Gomis RR, Prats N, Rodríguez-Gil JE, Domingo M, Gomis R et al. Tungstate is an effective antidiabetic agent in streptozotocin-induced diabetic rats: a long-term study. Diabetologia 2001;44:507-513. MS # CVR-2011-1135R1 18 20. Claret M, Corominola H, Canals I, Saura J, Barcelo-Batllori S, Guinovart JJ et al. Tungstate decreases weight gain and adiposity in obese rats through increased thermogenesis and lipid oxidation. Endocrinology 2005;146:4362-4369. 21. Suzuki H, Delano FA, Parks DA, Jamshidi N, Granger DN, Ishii H et al. Xanthine oxidase activity associated with arterial blood pressure in spontaneously hypertensive rats. Proc Natl Acad Sci U S A 1998;95:4754-4759. 22. Swei A, Lacy F, Delano FA, Parks DA, Schmid-Schonbein GW. A mechanism of oxygen free radical production in the Dahl hypertensive rat. Microcirculation 1999;6:179-187. 23. Peredo HA, Zabalza M, Mayer MA, Carranza A, Puyo AM. Sodium tungstate and vanadyl sulfate effects on blood pressure and vascular prostanoids production in fructose-overloaded rats. Clin Exp Hypertens 2010;32:453-457. 24. Flatman PW. Mechanisms of magnesium transport. Annu Rev Physiol 1991;53:259-271. 25. Perez GJ, Bonev AD, Nelson MT. Micromolar Ca 2+ from sparks activates Ca 2+ -sensitive K + channels in rat cerebral artery smooth muscle. Am J Physiol Cell Physiol 2001;281:C1769- C1775. 26. Kawanabe Y & Nauli SM. Endothelin. Cell Mol Life Sci 2011;68:195-203. 27. Jaggar JH, Porter VA, Lederer WJ, Nelson MT. Calcium sparks in smooth muscle. Am J Physiol Cell Physiol 2000;278:C235-C256. 28. Standen NB & Quayle JM. K + channel modulation in arterial smooth muscle. Acta Physiol Scand 1998;164:549-557. 29. Alioua A, Tanaka Y, Wallner M, Hofmann F, Ruth P, Meera P et al. The large conductance, voltage-dependent, and calcium-sensitive K + channel, Hslo, is a target of cGMP-dependent protein kinase phosphorylation in vivo. J Biol Chem 1998;273:32950-32956. MS # CVR-2011-1135R1 19 30. Hofmann F, Ammendola A, Schlossmann J. Rising behind NO: cGMP-dependent protein kinases. J Cell Sci 2000;113:1671-1676. 31. Jiang Z, Wallner M, Meera P, Toro L. Human and rodent MaxiK channel beta-subunit genes: cloning and characterization. Genomics 1999;55:57-67. 32. McManus OB, Helms LM, Pallanck L, Ganetzky B, Swanson R, Leonard RJ. Functional role of the beta subunit of high conductance calciumactivated potassium channels. Neuron 1995;14:645-650. 33. Tomás M, Vázquez E, Fernández-Fernández JM, Subirana I, Plata C, Heras M et al. Genetic variation in the KCNMA1 potassium channel alpha subunit as risk factor for severe essential hypertension and myocardial infarction. J Hypertens 2008;26:2147-2153. 34. Valverde MA, Rojas P, Amigo J, Cosmelli D, Orio P, Bahamonde MI et al. Acute activation of Maxi-K channels (hSlo) by estradiol binding to the beta subunit. Science 1999;285:1929- 1931. 35. Dick GM & Sanders KM. (Xeno)estrogen sensitivity of smooth muscle BK channels conferred by the regulatory beta1 subunit: a study of beta1 knockout mice. J Biol Chem 2001;276:44835-44840. 36. Bukiya AN, Liu J, Toro L, Dopico AM. Beta1 (KCNMB1) subunits mediate lithocholate activation of large-conductance Ca 2+ -activated K + channels and dilation in small, resistancesize arteries. Mol Pharmacol 2007;72:359-369. 37. Nardi A & Olesen SP. BK channel modulators: a comprehensive overview. Curr Med Chem 2008;15:1126-1146. 38. Braun M, Ramracheya R, Bengtsson M, Zhang Q, Karanauskaite J, Partridge C et al. Voltage-gated ion channels in human pancreatic beta-cells: electrophysiological characterization and role in insulin secretion. Diabetes 2008;57:1618-1628. MS # CVR-2011-1135R1 20 39. Houamed KM, Sweet IR, Satin LS. BK channels mediate a novel ionic mechanism that regulates glucose-dependent electrical activity and insulin secretion in mouse pancreatic betacells. J Physiol 2010;588:3511-3523. 40. Bukoski RD, Shearin S, Jackson WF, Pamarthi MF. Inhibition of Ca 2+ -induced relaxation by oxidized tungsten wires and paratungstate. J Pharmacol Exp Ther 2001;299:343-350. 41. Hanzu F, Gomis R, Coves MJ, Viaplana J, Palomo M, Andreu A et al. Proof-of-concept trial on the efficacy of sodium tungstate in human obesity. Diabetes Obes Metab 2010;12:1013- 1018. MS # CVR-2011-1135R1 21 Figure Legends Figure 1. Effect of tungstate on BK channels composed by different combinations of α and β subunits. Representative currents recorded from excised inside-out macropatches obtained from HEK293 cells expressing the BK α subunit alone (A), α + β 1 subunits (C), α + β 2 subunits (E), α + β 3 subunits (G) or α + β 4 subunits (I). Currents were recorded at cytosolic 0 Ca 2+ and 0.7 mM Mg 2+ before (top panels) and 5-10 minutes after (bottom panels) cytosolic application of 1 mM tungstate (WO 42- ). The voltage protocol was as described in Methods (applies also to Figures 2-5). Average G- V curves for the different combinations of BK channels: α subunit only (panel B, n = 8); α + β 1 subunits (panel D, n = 7); α + β 2 subunits (panel F, n = 5); α + β 3 subunits (panel H, n = 5); α + β 4 subunits (channels, panel J, n = 5). Normalized conductance and solid curves obtained by fitting to the Boltzmann equation are shown in this and Figures 2-5. K, Reduction of the BK peak current at the test potential of +200 mV (in %) produced by 1 mM WO 42- for BK channels constituted by different α/β combinations, as indicated. No significant differences were found among the distinct BK channels studied (P = 0.67, ANOVA). L, V 1/2 act plot for the different BK channels tested, before (control, ) and after (1 mM WO 42- , ) tungstate application (number of experiments for each experimental condition are as indicated above for the G-V curves). Differences in V 1/2 act before and after addition of 1 mM WO 42- were tested by Student’s t test: BKα channels, P = 0.24; BKαβ 1 channels, P < 0.0001; BKαβ 2 channels, P > 0.91; BKαβ 3 channels, P = 0.64; BKαβ 4 channels, P = 0.0025. M, No significant differences were found for average slope factor (k act ) of the fitted G-V curves, before (control, ) and after (1 mM WO 42- , ) tungstate application in the presence of cytosolic 0 Ca 2+ and 0.7 mM Mg 2+ (tested by Student’s t test: BKα channels, n = 8, P = 0.6; BKαβ 1 channels, n = 7, P = 0.66; BKαβ 2 channels, n = 5, P = 0.46; BKαβ 3 channels, n = 5, P = 0.11; BKαβ 4 channels, n = 5, P = 0.87). Figure 2. Effect of tungstate on BKαβ 1 channels in the presence of physiological cytosolic Ca 2+ levels. A, BKαβ 1 currents recorded in the presence of cytosolic 10 µM Ca 2+ and 0.7 mM Mg 2+ , before (top panels) and 5-10 minutes after (bottom panels) cytosolic application of 1 mM WO 42- . B, Average G-V curves for BKαβ 1 currents in the presence of 10 µM Ca 2+ / 0.7 mM Mg 2+ , before (control, ) and after (1 mM WO 42- , ) tungstate application (n = 3). C, Reduction of the BKαβ 1 peak current at +200 MS # CVR-2011-1135R1 22 mV (in %) produced by 1 mM WO 42- in the presence of 500 nM or 10 µM Ca 2+ , as indicated. Significant differences were found when comparing the 10 µM Ca 2+ condition to the other experimental situations (P < 0.05, ANOVA followed by a Tukey posthoc test) (data for 0 Ca 2+ are replotted from Figure 1 for comparison). D, Average left-shifts of the BKαβ 1 G-V relationship induced by 1mM tungstate in the absence (n = 7) and in the presence of 500 nM (n = 6) or 10 µM cytosolic Ca 2+ (n = 3). Differences in V 1/2 act before and after addition of 1 mM WO 42- were tested by paired Student’s t test: 0 Ca 2+ condition, P < 0.0001; 500 nM Ca 2+ condition, P < 0.01; 10 µM Ca 2+ condition, P < 0.01. Control V 1/2 act values (in mV) at different cytosolic Ca 2+ levels (before tungstate application), were in the range of those previously reported for BKαβ 1 channels 17 : 0 Ca 2+ condition, 138.76 ± 2.46; 500 nM Ca 2+ condition, 115.64 ± 5.31; 10 µM Ca 2+ condition, -9.41 ± 5.72. Figure 3. Activation of BKαβ 1 channels by tungstate requires physiological levels of cytosolic Mg 2+ . A, BKαβ 1 currents were recorded from the same patch at cytosolic 0 Ca 2+ and 0 (left panel), 0.7 mM (middle panel), or 5 mM Mg 2+ (right panel). B, Average G-V curves for BKαβ 1 currents in the presence of increasing concentrations of cytosolic Mg 2+ (n = 9) (, 0 Mg 2+ ; , 0.7 mM Mg 2+ ; and , 5 mM). Representative currents recorded at 0 Ca 2+ and 0 Mg 2+ (C) or 5 mM Mg 2+ (E), before (top panels) and 5-10 minutes after (bottom panels) cytosolic application of 1 mM WO 42- . Average G-V curves for BKαβ 1 currents at 0 Mg 2+ (D, n = 7) and 5 mM Mg 2+ (F, n = 3), before (control, ) and after (1 mM WO 42- , ) tungstate application. G, Reduction of the BKαβ 1 peak current at +200 mV (in %) produced by 0.7 (n = 5) and 5 mM Mg 2+ (n = 9) (white bars) or by 1 mM WO 42- in the presence of the indicated levels of cytosolic Mg 2+ (black bars; 0 Mg 2+ , n = 5; 0.7 mM Mg 2+ , n = 7; 5 mM Mg 2+ , n = 3). Current amplitude reduction by 5 mM Mg 2+ was larger than the one induced by 0.7 mM Mg 2+ (P < 0.05, Student’s t test). Not significant differences were found when comparing the effect of tungstate under the different Mg 2+ conditions (P = 0.47, Nonparametric ANOVA (Kruskal-Wallis Test)) (data for current amplitude reduction by tungstate obtained at 0.7 mM Mg 2+ are re-plotted from Figure 1 for comparison). H, V 1/2 act plot for the BKαβ 1 channel at different cytosolic Mg 2+ levels, before (control, ) and after (1 mM WO 42- , ) tungstate application (0 Mg 2+ , n = 7; 0.7 mM Mg 2+ , n = 7; 5 mM Mg 2+ , n = 3). Differences in V 1/2 act before and after addition of 1 mM WO 42- were tested by MS # CVR-2011-1135R1 23 Student’s t test or Mann-Whitney test, as appropriate: 0 Mg 2+ condition, P = 0.53; 0.7 mM Mg 2+ condition, P < 0.0001; 5 mM Mg 2+ condition, P = 0.6 (0.7 mM Mg 2+ data are re-plotted from Figure 1 for direct comparison). I, Not significant differences were found for average slope factor of the fitted G-V curves (k act ) neither among the different Mg 2+ concentrations at control condition (tested by One- Way ANOVA: n = 9, P = 0.89) not before (control, ) and after (1 mM WO 42- , ) tungstate application (tested by Student’s t test or Mann-Whitney test, as appropriate: 0 Mg 2+ condition, n = 7, P = 0.68; 0.7 mM Mg 2+ condition, n = 7, P = 0.66; 5 mM Mg 2+ condition, n = 3, P = 0.73). Figure 4. D99A and N172A mutations in the VSD of the BK  subunit prevent activation of BKαβ 1 channels by tungstate. BKαD99Aβ 1 (A) or BKαN172Aβ 1 (C) currents were recorded at cytosolic 0 Ca 2+ and 0.7 mM Mg 2+ before (top panels) and 5-10 minutes after (bottom panels) cytosolic application of 1 mM WO 42- . Average G-V curves for BKαD99Aβ 1 (B, n = 4) and BKαN172Aβ 1 (D, n = 3) currents at 0 Ca 2+ / 0.7 mM Mg 2+ , before (control, ) and after (1 mM WO 42- , ) tungstate application. E, Reduction (in %) by 1 mM WO 42- of the peak current (measured at +200 mV) through WT BKαβ 1 , BKαD99Aβ 1 or BKαN172Aβ 1 channels. Not significant differences were found among WT and mutants BKαβ 1 channels (P = 0.31, Nonparametric ANOVA (Kruskal- Wallis Test)). F, Mean V 1/2 act before (control, ) and after (1 mM WO 42- , ) tungstate application for WT (n = 7; P < 0.0001, Student’s t test; data re-plotted from Figure 1 for direct comparison), D99A mutant (n = 4; P = 0.49, Mann-Whitney test) and N172A mutant (n = 3; P = 0.2, Mann-Whitney test) BKαβ 1 channels. Figure 5. Effect of micromolar tungstate (0.1 mM) on BKα and BKαβ 1 channels. BKα (A) or BKαβ 1 (C) currents were acquired at cytosolic 0 Ca 2+ and 0.7 mM Mg 2+ before (top panels) and 5-10 minutes after (bottom panels) cytosolic application of 0.1 mM WO 42- . Average G-V curves for BKα (B, n = 5) and BKαβ 1 (D, n = 3) currents at 0 Ca 2+ / 0.7 mM Mg 2+ , before (control, ) and after (0.1 mM WO 42- , ) micromolar tungstate application. E, Changes (in %) by 0.1 mM WO 42- of the peak current (at +200 mV) trough WT BKα or BKαβ 1 channels. F, Mean V 1/2 act before (control, ) and MS # CVR-2011-1135R1 24 after (0.1 mM WO 42- , ) micromolar tungstate application for BKα (n = 5; P = 0.11, Student’s t test) and BKαβ 1 (n = 3; P < 0.05, Mann-Whitney test) channels. Figure 6. Tungstate-induced vasodilation of precontracted mouse arteries requires BK channel β 1 subunit expression. Tension recordings of femoral arterial rings obtained from either a WT (A) or BK β 1 -knockout (KO) (B) mouse that were contracted first by K + depolarizing solutions (as indicated) and then by 50 nM endothelin-1 (ET-1). Subsequent application of 0.1 mM WO 42- exclusively reduced arterial tension in the WT ET-1-precontracted ring (effect reverted by paxilline (Pax, 0.5 µM)). C, Time course of average arterial tension changes (in %) after application of 0.1 mM WO 42- to ET-1-precontracted WT (, n = 21 (at 20 min n = 6)) and β 1 -KO rings (, n = 12 (at 20 min n = 4)). Application of the xanthine oxidase inhibitor allopurinol (100 µM) did not modify the tension of WT ET-1-precontracted femoral arterial rings over 20 minutes (▲, n = 5). MS # CVR-2011-1135R1 Figure 1. MS # CVR-2011-1135R1 2 Putative tungstate binding sites at the human pore-forming α subunit of the BK channel To obtain some insight into structural features of tungstate-protein interactions, we studied available structural data on tungstate complexes. The analysis of the residue composition in the tungstate binding site (4.5Å radius) of 45 available X-ray structures reveals the residues and type of contacts which are commonly found in tungstate complexes (Supplementary material online, Figure S1A). The most preferred are polar residues (as serine, arginine, and lysine) interacting with the tungstate ion mainly by their polar side chain. Interestingly, hydrophobic residues also contact the polar tungstate ion by using their backbone atoms. Polar and hydrophobic residues in the binding pocket can stabilize the tungstate ion by direct as well as indirect (water-mediated) interactions. This is supported by the finding of numerous water molecules in the tungstate binding site of the studied X-ray structures (Supplementary material online, Figure S1A). There is evidence that cations can facilitate tungstate binding. Indeed, 20% of the 45 available tungstate complexes show a cation-mediated tungstate binding. In one structure (1J9K 3 ) the tungstateprotein interaction is stabilized by a Ca 2+ ion, in another structure (3KBP 4 ) a Zn 2+ and Fe 3+ ion stabilize tungstate. A total of 7 structures are in complex with a Mg 2+ ion (1FEZ 5 , 1U7P 6 , 2E6B 7 , 2I34 (Felts R, Tanner J. The crystal structure of Class C acid phosphatase from Bacillus anthracis with tungstate bound, unpublished data [2011]), 3E84 8 , 3ET5 9 and 3NVB (Fedorov A, Federov E, Toro R, Burley S, Almo S. Crystal structure of N-terminal part of the protein BF1531 from Bacteroides fragilis containing phosphatase domain complexed with Mg 2+ and tungstate, unpublished data [2011]) suggesting that Mg 2+ ions could also play a role in tungstate action on the BK channel. Those proteins belong to different types of enzymes mainly found in bacteria: phosphonoacetaldehyde hydrolase (EC = 3.11.1.1, PDB 1FEZ), protein-tyrosine-phosphatase (EC = 3.1.3.48, PDB 1U7P), 5'-nucleotidase (EC = 3.1.3.5, PDB 2E6B), acid phosphatase (EC = not classified in UniProt DB, PDB 2I34), a putative acylneuraminate cytidylyltransferase (EC = not classified in UniProt DB, 3E84), NADP phosphatase (EC = not classified in UniProt DB, PDB 3ET5) and the so far uncharacterized protein PDB 3NVB with a phosphatase domain (Supplementary material online, Table S1). We carried out a Blast search for all 7 tungstate-Mg 2+ protein complexes under default settings in order to detect human homologues. This search reveals that the majority of tungstate-Mg 2+ protein complexes (PDBs 1FEZ, MS # CVR-2011-1135R1 3 2E6B, 2I34, 3E84, 3ET5, and 3NVB) have no human homologues with the only exception of the protein-tyrosine-phosphatase (EC = 3.1.3.48, PDB 1U7P) (Supplementary material online, Table S1). An analysis of the residue composition in the binding pocket of available Mg 2+ -tungstate complexes shows a high occurrence of aspartate and lysine (Supplementary material online, Figure S1B). Moreover, threonine, serine, asparagine, arginine, and glutamate are involved in the tungstate binding (Supplementary material online, Figure S1B). A structural alignment of all Mg 2+ -tungstate complexes reveals four aspartate clusters (C1, C2, C3 and C4, Supplementary material online, Figure S2) around the tungstate ion, conserved in different types of hydrolases 1-4 (Felts R, Tanner J. The crystal structure of Class C acid phosphatase from Bacillus anthracis with tungstate bound, unpublished data, [2011]) and transferases 8 . The role of these aspartates is twofold (Supplementary material online, Figure S2): (i) the C1 and C2 interact indirectly with the tetrahedral tungstate ion through a Mg 2+ ion mediating strong ionic contacts between protein and tungstate, and (ii) a direct aspartate interaction of C1 with tungstate suggests that aspartates in C1 can be present in a protonated state. Moreover, the common occurrence of a lysine in addition to an aspartate in Mg 2+ -tungstate complexes seems to be structurally favorable. On the one hand, a lysine can stabilize directly the tungstate ion through interaction of its positively charged NH 3+ side chain with one of the negatively charged tungstate oxygens. On the other hand, a lysine can also promote an optimal orientation of aspartates in C3 allowing the coordination of the Mg 2+ ion via a water molecule (Supplementary material online, Figure S2). Finally, the studied Mg 2+ -tungstate complexes are stabilized by numerous polar interactions with water molecules which mediate indirect contacts between the Mg 2+ ion, protein side chain/backbone and the tungstate ion (Supplementary material online, Figure S2). The available experimental data on tungstate-protein complexes described above provides valuable information for the location of putative tungstate binding sites at the human pore-forming α subunit of the BK channel. Assuming a Mg 2+ -dependent tungstate binding, we can expect a simultaneous occurrence of aspartates and lysine in the binding pocket (Supplementary material online, Figures S1B and S2). It is worth to note that the two cytoplasmic regions (belonging to the voltage-sensor domain and the RCK1 domain) including the binding site for Mg 2+ involved in Mg 2+ - dependent BK channel activation 10 contain a DK/KD motif within or next to the relevant residues D99 MS # CVR-2011-1135R1 4 (site 1), N172 (site 2) and E374 (site 3) (Supplementary material online, Table S2 and Figure S3A). The inspection of the full cytoplasmic part of the human BK channel α subunit also revealed two additional sites containing DK/KD motifs: site 4 in the cytoplasmic region between S9 and S10 and site 5 in the cytoplasmic tail (Supplementary material online, Table S2 and Figure S3A). Overall, residues in vicinity of the DK/KD motif in all sites 1 to 5 (Supplementary material online, Table S2) belong to polar (Glu, Arg, Asn, Ser, Thr, and Asp) and non-polar residues (Ala, Ile, Trp, Gly, and Leu) that have been identified in the studied Mg 2+ -tungstate complexes (Supplementary material online, Figure S1B), thus being reasonable interaction partners for the Mg 2+ -tungstate complex. Solutions and software for electrophysiological studies Borosilicate glass patch pipettes had 2-3 M3 3resistance and were filled with a solution containing (in mM): 140 KCl, 1.2 MgCl 2 , 0.15 CaCl 2 , 5 EGTA, 10 HEPES (300 mOsm/L, pH 7.35). In experiments performed at 0 Mg 2+ , pipette solution contained (in mM): 140KCl, 0.15 CaCl 2 , 5 EGTA, 10 HEPES (300 mOsm/L, pH 7.25). 0 Ca 2+ solution (nominally Ca 2+ -free) bathing the cytoplasmic face of the patch membrane, to which 0.1 mM or 1 mM WO 42- was added after some control recordings, contained (in mM): 140 KCl, 5 EGTA, 10 HEPES (0 Mg 2+ condition). The cytosolic levels of Mg 2+ were increased when appropriate by adding 0.7 or 5 mM Mg 2+ to this 0 Ca 2+ cytosolic solution. All these bath solutions were adjusted to 300 mOsm/L and pH 7.25. In some experiments the intracellular free Ca 2+ concentration (calculated using EqCal; Biosoft, Cambridge, United Kingdom) was adjusted to either 500 nM or 10 µM by different combinations of CaCl 2 and EGTA added to the bath solution indicated above containing 0.7 mM Mg 2+ . pClamp8, PatchMaster and FitMaster softwares were used for pulse generation, data acquisition and subsequent analysis. Currents were acquired at 10 kHz and low-pass-filtered at 1 kHz. Arterial tension measurements Arterial rings derived from adult male, age-matched (18-24-week-old) +/+ (WT) and −/− BKβ 1 (β 1 - knockout) C57BL/6 mice, were cleaned of connective tissue, cut in rings (≈2 mm length) and MS # CVR-2011-1135R1 5 mounted on a small-vessel myograph (Danish Myo Technology, Aarhus, Denmark) to measure tension connected to a digital recorder (AcqKnowledge 3.8.1, Biopac System Inc.). The rings were placed on a chamber filled with a standard Krebs solution that contained (in mM): 119 NaCl, 2.5 CaCl 2 , 4.7 KCl, 1.17 MgSO 4 , 24 NaHCO 3, 1.17 KH 2 PO 4 , 5.5 glucose, bubbled with 95% O 2 and 5% CO 2 at pH 7.4. Before the experiments, the segments were subjected to an optimal tension (90% of tension equivalent to an intramural pressure of 100 mm Hg) and stabilized for at least 45 min. After the equilibration period, we discarded arterial rings that developed a tension of < 2-4 mN in response to 20-35 mM KCl (when extracellular K + was increased, KCl substituted equimolarly NaCl). As reported previously, application of a low-depolarizing KCl solution (20 mM) produced a significant (P < 0.05, Student’s t test) increased contractile response in β 1 -knockout (7.31 ± 0.88 mN, n = 12) in comparison to WT arterial rings (4.9 ± 0.58 mN, n = 21), an effect that has been linked to a reduced BK channel activity, most likely due to absence of BK β 1 subunit 11 . No significant difference was found in the arterial constriction induced by 35 mM KCl among genotypes (P = 0.36, Student’s t test: WT, 9.12 ± 0.79 mN, n = 21; β 1 -knockout, 10.28 ± 0.93 mN, n = 12). All the drugs used were added directly to the chamber while vessel tension was monitored. Experiments were performed at 30ºC. Measurements of changes in arterial tension induced by tungstate or allopurinol on endothelin-1 (ET- 1) precontracted arterial rings were taken at the indicated times (Figure 6). Results obtained with tungstate correspond to 9 WT and 8 β 1 -knockout animals (except at 20 minutes: 6 WT and 4 β 1 - knockout animals). Results obtained with allopurinol correspond to 2 WT animals. Sodium tungstate, ET-1 and allopurinol were provided by Sigma, and paxilline by Alomone. Supplementary References 1. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, et al. The Protein Data Bank. Nucleic Acids Res. 2000;28:235-242. 2. Molecular Operating Environment (MOE) software, version 2010.10, (2010) http://www.chemcomp.com/software.htm. MS # CVR-2011-1135R1 6 3. Lee JY, Kwak JE, Moon J et al. Crystal structure and functional analysis of the SurE protein identify a novel phosphatase family. Nat Struct Biol 2001;8:789-794. 4. Klabunde T, Strater N, Frohlich R, Witzel H, Krebs B. Mechanism of Fe(III)-Zn(II) purple acid phosphatase based on crystal structures. J Mol Biol 1996;259:737-748. 5. Morais MC, Zhang W, Baker AS, Zhang G, Dunaway-Mariano D, Allen KN. The crystal structure of bacillus cereus phosphonoacetaldehyde hydrolase: insight into catalysis of phosphorus bond cleavage and catalytic diversification within the HAD enzyme superfamily. Biochemistry 2000;39:10385-10396. 6. Peisach E, Selengut JD, Dunaway-Mariano D, Allen KN. X-ray crystal structure of the hypothetical phosphotyrosine phosphatase MDP-1 of the haloacid dehalogenase superfamily. Biochemistry 2004;43:12770-12779. 7. Iwasaki W, Miki K. Crystal structure of the stationary phase survival protein SurE with metal ion and AMP. J Mol Biol 2007;371:123-136. 8. Lu Z, Wang L, Dunaway-Mariano D, Allen KN. Structure-function analysis of 2-keto-3- deoxy-D-glycero-D-galactonononate-9-phosphate phosphatase defines specificity elements in type C0 haloalkanoate dehalogenase family members. J Biol Chem 2009;284:1224-1233. 9. Felts RL, Ou Z, Reilly TJ, Tanner JJ. Structure of recombinant Haemophilus influenzae e (P4) acid phosphatase reveals a new member of the haloacid dehalogenase superfamily. Biochemistry 2007;46:11110-11119. 10. Yang H, Shi J, Zhang G, Yang J, Delaloye K, Cui J. Activation of Slo1 BK channels by Mg 2+ coordinated between the voltage sensor and RCK1 domains. Nat Struct Mol Biol 2008;15:1152-1159. MS # CVR-2011-1135R1 7 11. Pluger S, Faulhaber J, Furstenau M, Lohn M, Waldschutz R, Gollasch M et al. Mice with disrupted BK channel beta1 subunit gene feature abnormal Ca 2+ spark/STOC coupling and elevated blood pressure. Circ Res 2000;87:E53-E60. MS # CVR-2011-1135R1 8 Supplementary Figures Figure S1. MS # CVR-2011-1135R1 9 Figure S2. MS # CVR-2011-1135R1 10 S0-S1 Linker Human RTLKYLWTVCCHCGGKTKEAQKINNGSSQADGTLKPVDEKEEAVAAEVGWMTSVKDWAGVMISAQTLTGR 113 Mouse RTLKYLWTVCCHCGGKTKEAQKINNGSSQADGTLKPVDEKEEVVAAEVGWMTSVKDWAGVMISAQTLTGR 113 S2-S3 linker Human AAND 173 Mouse AAND 173 Figure S3. A B MS # CVR-2011-1135R1 11 Figure S4.