Zoledronic Acid Blocks Overactive Kir6.1/SUR2-Dependent KATP Channels in Skeletal Muscle and Osteoblasts in a Murine Model of Cantú Syndrome

Cantú syndrome (CS) is caused by the gain of function mutations in the ABCC9 and KCNJ8 genes encoding, respectively, for the sulfonylureas receptor type 2 (SUR2) and the inwardly rectifier potassium channel 6.1 (Kir6.1) of the ATP-sensitive potassium (KATP) channels. CS is a multi-organ condition with a cardiovascular phenotype, neuromuscular symptoms, and skeletal malformations. Glibenclamide has been proposed for use in CS, but even in animals, the drug is incompletely effective against severe mutations, including the Kir6.1wt/V65M. Patch-clamp experiments showed that zoledronic acid (ZOL) fully reduced the whole-cell KATP currents in bone calvaria cells from wild type (WT/WT) and heterozygous Kir6.1wt/V65MCS mice, with IC50 for ZOL block < 1 nM in each case. ZOL fully reduced KATP current in excised patches in skeletal muscle fibers in WT/WT and CS mice, with IC50 of 100 nM in each case. Interestingly, KATP currents in the bone of heterozygous SUR2wt/A478V mice were less sensitive to ZOL inhibition, showing an IC50 of ~500 nM and a slope of ~0.3. In homozygous SUR2A478V/A478V cells, ZOL failed to fully inhibit the KATP currents, causing only ~35% inhibition at 100 μM, but was responsive to glibenclamide. ZOL reduced the KATP currents in Kir6.1wt/VMCS mice in both skeletal muscle and bone cells but was not effective in the SUR2[A478V] mice fibers. These data indicate a subunit specificity of ZOL action that is important for appropriate CS therapies.


Introduction
Cantú Syndrome (CS) is a rare autosomal-dominant, a multi-organ condition characterized by cardiomegaly, vascular dilation, low blood pressure, hypertrichosis, neuromuscular symptoms, and skeletal malformations [1][2][3]. CS is caused by gain-of-function (GOF) mutations in the ABCC9 and KCNJ8 genes [4,5], encoding the SUR2 and Kir6.1 subunits, respectively, of ATP-sensitive potassium (KATP) channels. To date, >70 genetically confirmed individuals have been reported with CS and associated with >30 missense ABCC9 or KCNJ8 mutations [6,7], yet there is currently no recommended therapy for CS. Sulfonylurea inhibitors are obvious candidate therapies. Glibenclamide concentrations dependently reduce the KATP channel current of wild-type (WT) and CS mutant channels in heterologous expression systems, as well as in cardiac, smooth, and skeletal muscles cells of CS mutant mice in which the SUR2 A478V mutation is engineered into the mouse locus using CRISPR/Cas9 [8,9]. The reversal of cardiovascular phenotypes in SUR2 A478V mice [10] and the apparent benefit of glibenclamide, with minimal disruption of glycemic control

Drugs and Solutions
The KATP channels blocker glyburide/glibenclamide cat. N • PHR1287 and BaCl 2 cat. N • 449644 were purchased from Sigma (SIGMA Chemical Co., Milan, Italy) as well as all chemicals. The zoledronic acid (ZOL) was synthesized, purified in our labs, and dissolved in phosphate-buffered saline (PBS) stock solution at 1 or 10 × 10 −3 M concentrations.
CaCl 2 was added to the pipette solutions to give a free Ca 2+ ion concentration of 1.6 × 10 −6 M in whole-cell experiments. The calculation of the free Ca 2+ ion concentration in the pipette was performed using the MaxChelator software (Stanford University; Stanford, CA, USA). These intracellular Ca 2+ ion concentrations are required for a stable seal formation in whole-cell experiments.
The culture of primary bone cells from adult mice was obtained from long bone such as femora or calvaria, as previously described [9,31] Bones were collected, cleaned, and flushed to remove the internal bone marrow cells. Small bone pieces of around 1 mm 3 were treated with trypsin-EDTA 0.25% (w/v) for 1 h and with 0.2% collagenase solution for an additional hour in a shaking water bath to remove all remaining soft tissue and adherent cells. Clean bone pieces have been cultured in a basal medium enriched with 50 µg/mL of ascorbic acid. As previously described, osteoblasts from calvaria chips were positive for Alizarin red staining and positive for PCR gene expression on cell pellets [32].

Patch-Clamp Experiments
Whole-cell patch-clamp experiments were performed in asymmetrical K + ion concentration in physiological conditions using pipettes with a resistance of 3-5 MΩ. A drug solution was applied on the extracellular side. KATP currents of native skeletal muscle fibers were recorded in excised macropatches using pipettes of 0.9-1.2 MΩ of resistance on isolated skeletal muscle fibers in symmetrical 150 mM K + ions concentrations on both Cells 2023, 12, 928 4 of 12 sides of the membrane during pulse going from 0 mV to −60 mV (Vm) and drug solution directly applied on the internal side of the patches.
Drug actions on channel currents recorded during instantaneous whole-cell I/V relationships were investigated by applying a depolarization protocol in response to voltage pulses from −180 mV to +100 mV (Vm) in 20 mV steps. All experiments were performed at room temperature (20-22 • C) and sampled at 2 kHz (filter = 1 kHz) using an Axopatch-1D amplifier equipped with a CV-4 head stage (Axon Instruments, Foster City, CA, USA). Drug solutions were applied using a fast perfusion system during the continuous monitoring of the seal resistance at 18-22 • C room temperature in excised patches. In whole-cell patch-clamp experiments, before recording, cells were equilibrated for 10 s with the drug solution. The fast perfusion system is programmed at this time interval for seal stability. Data acquisition and analysis were performed using the pCLAMP 10 software suite (Axon Instruments, Foster City, CA, USA), as previously described. Seal resistance was continuously monitored during the experiment.

Statistical Analysis
The data are expressed as an average ± S.E.M. unless otherwise specified. The significance between data pairs was calculated by the paired Student's t-test for p < 0.05. One-way ANOVA was used to evaluate the significance within and between data with a variance-ratio F > 1 at significance levels of p < 0.05.
The percentage of KATP current inhibition induced by ZOL was calculated as (I CTRL-I drug)/(I CTRL-I BaCl 2 ) × −100. BaCl 2 solution at high millimolar concentrations fully blocks all types of Kirs channels, including the Kir6.1 and Kir6.2 channels. The concentration-response data of ZOL against different KATP channel currents were fitted by the equation: ligand binding, four-parameter logistic function (linear) of Sigma Plot, 10.0.

Patch-Clamp Experiments
Whole-cell patch-clamp experiments were performed in asymmetrical K + ion concentration in physiological conditions using pipettes with a resistance of 3-5 MΩ. A drug solution was applied on the extracellular side. KATP currents of native skeletal muscle fibers were recorded in excised macropatches using pipettes of 0.9-1.2 MΩ of resistance on isolated skeletal muscle fibers in symmetrical 150 mM K + ions concentrations on both sides of the membrane during pulse going from 0 mV to −60 mV (Vm) and drug solution directly applied on the internal side of the patches.
Drug actions on channel currents recorded during instantaneous whole-cell I/V relationships were investigated by applying a depolarization protocol in response to voltage pulses from −180 mV to +100 mV (Vm) in 20 mV steps. All experiments were performed at room temperature (20-22 °C) and sampled at 2 kHz (filter = 1 kHz) using an Axopatch-1D amplifier equipped with a CV-4 head stage (Axon Instruments; Foster City, CA, USA). Drug solutions were applied using a fast perfusion system during the continuous monitoring of the seal resistance at 18-22 °C room temperature in excised patches. In wholecell patch-clamp experiments, before recording, cells were equilibrated for 10 s with the drug solution. The fast perfusion system is programmed at this time interval for seal stability. Data acquisition and analysis were performed using the pCLAMP 10 software suite (Axon Instruments; Foster City, CA, USA), as previously described. Seal resistance was continuously monitored during the experiment.

Statistical Analysis
The data are expressed as an average ± S.E.M. unless otherwise specified. The significance between data pairs was calculated by the paired Student's t-test for p < 0.05. Oneway ANOVA was used to evaluate the significance within and between data with a variance-ratio F > 1 at significance levels of p < 0.05.
The percentage of KATP current inhibition induced by ZOL was calculated as (I CTRL-I drug)/(I CTRL-I BaCl2) × −100. BaCl2 solution at high millimolar concentrations fully blocks all types of Kirs channels, including the Kir6.1 and Kir6.2 channels. The concentration-response data of ZOL against different KATP channel currents were fitted by the equation: ligand binding, four-parameter logistic function (linear) of Sigma Plot, 10.0.

Results
Previous studies showed that zoledronic acid (10 −12 -10 −4 M) inhibited the KATP currents recorded in excised macropatches from isolated wild-type Extensor digitorum longus (EDL) and Soleus skeletal muscle fibers [31]. As shown in Figure 1, zoledronic acid also potently inhibited KATP channels in Flexor digitorum brevis (FDB) fibers from Kir6.1 wt/V65M mice, with an IC50 of ~1 μM and slope of ~0.6.  Intriguingly, the KATP currents of the heterozygous SUR2 wt/A478V muscles were less sensitive to zoledronic acid inhibition at a concentration above the IC 50 but much lower slope (~0.3) but more sensitive at the lower range of concentrations, zoledronic acid failed to fully inhibit KATP currents in homozygous SUR2 A478V/A478V muscle, even at the highest concentrations ( Figure 1). These results are in striking contrast to the effects of the sulfonylurea glibenclamide, which shows similar action on FDB fibers from WT and SUR2 wt/A478V but was less active on Kir6.1 wt/V65M as also seen in recombinant and native vascular smooth muscle channels [8,[33][34][35] (Figure 2). The ability of glibenclamide to block KATP channels in Kir6.1 wt/V65M FDB muscle is in contrast to the lack of inhibitory effect in the soleus muscle [9,12,32,36,37] (Table 1). The KATP channels of SUR2 A478V/A478V were also less responsive to glibenclamide action. age steps going from 0 mV to −60 mV (Vm). The drug solution was applied on the intracellular of the patches. Zoledronic acid fully reduced KATP channel currents in either heterozy SUR2 wt/AV and Kir6.1 wt/V65M but not in homozygous SUR2 A478V/A478V muscle fibers. Each point o concentration-response curves was obtained on 11-19 patches.
Intriguingly, the KATP currents of the heterozygous SUR2 wt/A478V muscles were sensitive to zoledronic acid inhibition at a concentration above the IC50 but much lo slope (~0.3) but more sensitive at the lower range of concentrations, zoledronic acid fa to fully inhibit KATP currents in homozygous SUR2 A478V/A478V muscle, even at the hig concentrations ( Figure 1). These results are in striking contrast to the effects of the fonylurea glibenclamide, which shows similar action on FDB fibers from WT SUR2 wt/A478V but was less active on Kir6.1 wt/V65M as also seen in recombinant and native cular smooth muscle channels [8,[33][34][35] (Figure 2). The ability of glibenclamide to b KATP channels in Kir6.1 wt/V65M FDB muscle is in contrast to the lack of inhibitory effe the soleus muscle [9,12,32,36,37] (Table 1). The KATP channels of SUR2 A478V/A478V were less responsive to glibenclamide action. In primary cultured bone calvaria cells, zoledronic acid was a very potent blocke both WT and Kir6.1 wt/V65M KATP channels in whole-cell membrane patches when app to the outside of the cell membrane ( Figure 3). With an apparent IC50 of ~100pM, this e is considerably more potent than glibenclamide, which also inhibits both WT and mu channels with IC50 of ~10 nM (Figure 3).   Fitting parameters of the concentration-response relationships of the currents recorded at −60 mV (Vm) in inside-out (I-O) and whole cell (W-C) patches vs. drug concentrations were calculated by using the equation: Ligand Binding; four-parameter logistic function (linear) of Sigma Plot, 10.0. Data significantly differ between * fiber data and ** osteoblasts data with one-way ANOVA for p < 0.05 and F > 1.
In primary cultured bone calvaria cells, zoledronic acid was a very potent blocker of both WT and Kir6.1 wt/V65M KATP channels in whole-cell membrane patches when applied to the outside of the cell membrane ( Figure 3). With an apparent IC 50 of~100 pM, this effect is considerably more potent than glibenclamide, which also inhibits both WT and mutant channels with IC 50 of~10 nM (Figure 3).

Discussion
Taken together with our previously published work [23], these data indicate that zoledronic acid inhibits KATP channels in the musculoskeletal system with varying potency in different cell types, where potency in femoral osteoblasts ≈ calvaria osteoblasts ≈ soleus muscle > EDL muscle ≈ FDB muscle (Table 2). Our previous gene expression analysis showed high expression of Kir6.1 and SUR2 in osteoblasts and soleus muscle [12,23,38] and we previously showed that ZOL inhibits recombinant KATP channels composed of Kir6.1/SUR2B subunits with increased potency relative to Kir6.2/SUR2B or Kir6.2/SUR1 channels [9]. Therefore, we propose that increased ZOL sensitivity in osteoblasts and soleus muscle likely reflects increased Kir6.1/SUR2B contributions to native KATP channels in these tissues relative to FBD and EDL muscle. Importantly, this is also consistent with our previous finding that the Kir6.1 V65M mutation, which reduces glibenclamide inhibition [11], has a greater effect on GLIB sensitivity of soleus KATP channels versus FDB KATP channels [9]. The high expression ratio of KCNJ8-Kir6.1/KCNJ11-Kir6.2 relates to the potency of the zoledronic acid in the musculoskeletal tissues (Table 2).
In contrast to reduced GLIB inhibition in native cells from Kir6.1 wt/V65M mice, here we show that the Kir6.1 V65M mutation does not significantly affect ZOL inhibition of KATP in either osteoblasts or FBD muscle. This suggests that the Kir6.1 V65M mutation does not markedly reduce ZOL effects and points to different molecular mechanisms of action for ZOL and GLIB inhibition. This finding might have potential implications for the treatment of CS patients: glibenclamide reverses CS-associated cardiovascular pathologies in SUR2 wt/A478V mutant mice but has reduced efficacy in Kir6.1 wt/V65M mice, leading to the suggestion that clinical efficacy of glibenclamide may be limited in Kir6.1-variant CS patients [10,39,40]. Our data here suggest that if ZOL is safely tolerated, or if other safe drugs can be identified that work via a similar mechanism, they might be more effective than sulfonylureas in Kir6.1-variant patients. Curiously, we observed a marked reduction of ZOL inhibition in muscle from SUR2 A478V mutant mice. The mechanism for this effect is unknown and requires further study, but it might include direct effects on the binding site or allosteric effects of the mutation. In this respect, the KATP channels of the heterozygous SUR2 wt/A478V fibers showed a much lower slope than the channel in the WT/WT mice fibers. This can be related to the high-affinity interaction of the drug on the high-affinity sites on the channel complexes and the loss of interaction with lower-affinity sites, as supported by the fact that the channels were less sensitive to zoledronic acid inhibition at a concentration above the IC 50 . Lower affinity sites Cells 2023, 12, 928 9 of 12 of zoledronic acid are either the nucleotide sites and sulfonylureas sites on Kir6.2-SUR1 and Kir6.1-SUR1 complexes as observed by docking analysis suggesting that conformational changes of these sites are responsible for the reduced sensitivity of the channels to zoledronic acid in the heterozygous SUR2 wt/A478V .
In the present work, we demonstrate that zoledronic acid effectively and potently inhibits KATP currents in different skeletal muscle types and bone cells of Kir6.1 wt/V65M CS mice but not in the skeletal muscle of SUR2 A478V mice. It is a particularly potent inhibitor of KATP currents in soleus fibers and osteoblast cells. The effects of zoledronic acid in osteoblast and soleus muscle fibers of CS mice can be explained by the high expression/activity of the target channel subunits in these tissues vs. the fast-twitching fibers. Gene expression analysis indeed showed that osteoblasts and soleus fibers from WT/WT mice and rats express high Kir6.1-SUR2 channel proteins supporting this hypothesis [11,[41][42][43].
One important point to consider is the very different sensitivities to zoledronic acid that are observed in skeletal muscle and osteoblasts. In WT EDL and FDB skeletal muscle, KATP channel inhibition IC 50 is~1 µM, when applied directly to the cytoplasmic face of the membrane in excised patch experiments. Conversely, in bone calvaria, the drug is~3 orders of magnitude more potent, with IC 50~1 nM, when applied outside the membrane. We suggest that zoledronic acid likely interacts with KATP channels on a cytoplasmic site. High uptake activity into osteoblast may raise the sub-membrane concentration substantially beyond bath concentrations in whole-cell studies. In some cells that did not belong to the musculoskeletal apparatus, the action of the drug applied on the extracellular side can be limited by the influx rate.
Previously, our docking simulation analysis [23,24] identified several potential binding sites for zoledronic acid on SUR2 subunits in the sulfonylurea and nucleotide binding regions and an additional site in the ATP binding region of Kir6 subunits. We identified in silico zoledronic acid binding at a site of SUR2 common to both SUR2A and SUR2B splice variants, composed of S401, T400, T687, W681, and S713 residues, at which zoledronic acid bound with binding energy and RMSD values comparable to those for ATP binding. Residue A475 (equivalent to the human A478 numbering) is located in a nearby region. Previous recombinant channel studies indicate that the SUR2 A478V mutation increases channel activity by promoting activation of the channel by MgADP binding to the SUR2 nucleotide binding folds (NBFs) [33,35], whereas the Kir6.1 V65M mutation acts to increase the intrinsic open state stability of the channel itself [5]. Most evidence points to Kir6.2/SUR2 being the primary subunit composition for KATP channels in skeletal muscle [14,30,39]. It also suggests that the primary action of zoledronic acid may be on the Kir6 subunit rather than the SUR2 subunit since the GoF is preserved in AV skeletal muscle channels, i.e., zoledronic acid inhibition is less effective when there is SUR2-dependent activation.

Conclusions
In conclusion, the inhibitory effects of zoledronic acid on KATP currents in Kir6.1 wt/VM CS mice observed in different skeletal muscle phenotypes, and bone cells but not in skeletal muscle of SUR2 A478V mice can help improve drug prescription in CS patients to treat musculoskeletal symptoms. Zoledronic acid is in the WHO's List of Essential Medicines used as a primary therapy for skeletal disorders caused by imbalanced bone homeostasis [38], where osteoblast and osteoclast activities are not perfectly coupled, leading to excessive bone resorption like osteoporosis and Paget's disease and Multiple Myeloma [40][41][42][43][44]. Therefore, being an antiproliferative and anticancer drug [45], it can be prescribed in those CS patients also affected by cancer. It can be a case for a novel drug repurposing in rare diseases in specific CS patients with Kir6.1 wt/VM mutations. It should be of note that zoledronic acid is a medication approved by both the FDA and EMA for bone diseases and muscular dystrophy in pediatric patients. It was shown that 24 months of e.v. zoledronic acid treatment, in addition to vitamin D and calcium, leads to improvements in bone mass scores compared with vitamin D and calcium alone in Duchenne Muscular Dystrophy patients treated with corticosteroids. It was generally found to be safe and well tolerated apart from marked acute-phase reaction on first exposure in these people with some benefit in reducing pain [45][46][47]. Similarly, it is effective in glucocorticoid-induced osteoporosis in chronic pediatric illnesses associated with osteoporotic fractures [46], but the mechanism of action is not known. In our opinion, it is likely that the reported case reports of rhabdomyolysis and myoglobinuria following the administration of zoledronic acid in Duchenne Muscular Dystrophy can be associated with the interactions with muscle KATP channels [47,48].