Calmodulin-Dependent Regulation of Overexpressed but Not Endogenous TMEM16A Expressed in Airway Epithelial Cells

Regulation of the Ca2+-activated Cl− channel TMEM16A by Ca2+/calmodulin (CAM) is discussed controversially. In the present study, we compared regulation of TMEM16A by Ca2+/calmodulin (holo-CAM), CAM-dependent kinase (CAMKII), and CAM-dependent phosphatase calcineurin in TMEM16A-overexpressing HEK293 cells and TMEM16A expressed endogenously in airway and colonic epithelial cells. The activator of the Ca2+/CAM-regulated K+ channel KCNN4, 1-EBIO, activated TMEM16A in overexpressing cells, but not in cells with endogenous expression of TMEM16A. Evidence is provided that CAM-interaction with TMEM16A modulates the Ca2+ sensitivity of the Cl− channel. Enhanced Ca2+ sensitivity of overexpressed TMEM16A explains its activity at basal (non-elevated) intracellular Ca2+ levels. The present results correspond well to a recent report that demonstrates a Ca2+-unbound form of CAM (apo-CAM) that is pre-associated with TMEM16A and mediates a Ca2+-dependent sensitization of activation (and inactivation). However, when using activators or inhibitors for holo-CAM, CAMKII, or calcineurin, we were unable to detect a significant impact of CAM, and limit evidence for regulation by CAM-dependent regulatory proteins on receptor-mediated activation of endogenous TMEM16A in airway or colonic epithelial cells. We propose that regulatory properties of TMEM16A and and other members of the TMEM16 family as detected in overexpression studies, should be validated for endogenous TMEM16A and physiological stimuli such as activation of phospholipase C (PLC)-coupled receptors.


Introduction
In our earlier study we analyzed the regulation of the Ca 2+ -activated Cl − channel (CaCC) TMEM16A. We found that the channel can be activated by activators of Ca 2+ /calmodulin (CAM)-regulated KCNN4 K + channels, such as 1-EBIO and riluzole [1]. In addition, we demonstrated that TMEM16A physically interacts with CAM. A subsequent report indicated that the anion selectivity of TMEM16A is dynamically regulated by the Ca 2+ /CAM complex [2]. This study showed that CAM reversibly increases the permeability ratio P HCO3-/P Cl − . Further support for a CAM-dependent regulation of TMEM16A came from an interesting study by the Colecraft team [3]. They described a Ca 2+ -unbound form of CAM (apo-CAM), which is pre-associated with TMEM16A or the paralogue TMEM16B channel complexes. Apo-CAM mediates a Ca 2+ dependent sensitization of activation and Ca 2+ dependent inactivation of TMEM16 channels [3,4]. Moreover, CAM-dependent activation and inactivation of TMEM16A had also been found in an earlier study [5]. In contrast to these studies, other laboratories did not find evidence for regulation of TMEM16A by CAM or alteration of the bicarbonate permeability of TMEM16A by CAM [6,7].
Earlier studies describe a crosstalk between Ca 2+ /CAMKII-dependent regulation and inositol (3,4,5,6)-phosphate dependent regulation of CaCC, which also appears to be cell-type-dependent, possibly explaining some of the controversial findings outlined above [21,22]. In our earlier study, we observed that regulation of TMEM16A by INO-4995, a lipid inositol phosphate, differs depending on whether the channel is overexpressed in mammalian cells such as HEK293, or whether regulation of endogenous TMEM16A is examined [23]. While overexpressed TMEM16A was potently activated by the inositol phosphate INO-4995, endogenous TMEM16A expressed in airway and colonic epithelial cells was not. Activation of endogenous TMEM16A by purinergic receptor stimulation was augmented by INO-4995 [23]. Along the same line, the membrane permeable dioctanoyl-PIP 2 phosphatidylinositol 4,5-bisphosphate-diC8 activated overexpressed TMEM16A, but did not touch endogenous TMEM16A [24]. Furthermore, other previously reported activators of TMEM16A such as melittin or cinnamaldehyde failed to activate endogenous TMEM16A. These differences prompted us to compare the role of CAM for acute receptor-mediated activation of TMEM16A in overexpressing cells and epithelial cells with endogenous expression of TMEM16A. Here, we made use of four different cell lines. HEK293 cells were used to study regulation of overexpressed TMEM16A, while airway epithelial cells (BCi_NS1, CFBE) and colonic HT 29 cells were used to study regulation of endogenous TMEM16A. We demonstrate the absence of a significant CAM-dependent regulation for endogenous TMEM16A, solidifying the conclusion that control by CAM and other regulatory properties are indeed different for endogenous and overexpressed TMEM16A. Moreover, we made use of independent techniques (patch clamp, Ussing chamber, YFP-quenching) to reduce the chance for misinterpretations due to methodological artefacts. The data suggest only a minor role of CAM and the CAM-dependent proteins CAMKII and calcineurin for activation of endogenous TMEM16A, while CAM is likely to enhance Ca 2+ sensitivity of overexpressed TMEM16A.

Cell Culture
All cells were grown at 37 • C in a humidified atmosphere with 5% (v/v) CO 2 . Culture conditions for CFBE and HT 29 cells have been described earlier [25]. In brief, airway epithelial cells were grown in DMEM/Ham's F-12 with L-Glutamine medium supplemented with 10% (v/v) fetal bovine serum (FBS), 1% (v/v) L-glutamine 200mM and 1% (v/v) HEPES 1M (all from Capricorn Scientific, Ebsdorfergrund, Germany). CFBE parental cells were grown in MEM with Earle's Salts with L-Glutamine medium (Capricorn Scientific, Ebsdorfergrund, Germany) supplemented with 10% FBS. The airway epithelial cell line H441 was grown in RPMI and DMEM media. The immortalized human airway basal cell line BCi-NS1 (kindly provided by Prof. Ron Crystal, Weill Cornell Medical College, New York City, NY, USA) was maintained in Bronchial Epithelial Growth Media (Lonza). Cells were differentiated by growing on permeable supports (Snapwell, Corning, NY, USA) in an air-liquid interface (ALI) for up to 30 days in PnemaCult-ALI medium supplemented with PneumaCult-ALI 10X Supplement, PneumaCult-ALI Maintenance Supplement, hydrocortisone and heparin (all from StemCell Technologies, Vancouver, BC, Canada), and 1% penicilin-steptomycin (10 000 U/mL; Gibco, ThermoFisherScientific, Waltham, MA, USA).

Patch Clamping
Cells were patch-clamped after growing on coated glass coverslips for 2-3 days. Whole cell patch clamp techniques and data analysis have been described earlier [27]. In brief, patch pipettes were filled with a cytosolic-like solution containing (in mM): KCl 30, K-Gluconate 95, NaH 2 PO 4 1.2, Na 2 HPO 4 4.8, EGTA 1, Ca-Gluconate 0.758, MgCl 2 1.03, D-Glucose 5, ATP 3; pH 7.2. The intracellular (pipette) Ca 2+ activity was 0.1 µM. Fast whole cell current recordings were performed as described recently [28]. The bath was perfused continuously with standard bicarbonate-free Ringer's solution (in mM: NaCl 145, KH 2 PO 4 0.4, K 2 HPO 4 ·3 H 2 O 1.6, Glucose 5, MgCl 2 ·6 H 2 O 1, Ca-Gluconate·1 H 2 O 1.3) at a rate of 8 mL/min. Patch pipettes had an input resistance of 2-4 MΩ and whole cell currents were corrected for serial resistance. Currents were recorded using a patch clamp amplifier (EPC 7, List Medical Electronics, Darmstadt, Germany), the LIH1600 interface, and PULSE software (HEKA, Lambrecht, Germany), as well as Chart software (AD Instruments, Spechbach, Germany). Cells were stimulated with 1, 10, or 100 µM ATP in standard bicarbonate-free Ringer's solution. Cells were current-clamped for most of the time. In regular intervals, membrane voltage (Vc) was clamped in steps of 20 mV from −100 to +100 mV. The inhibitor of Ca 2+ -activated KCNN4 K + channels, if indicated TRAM-34 (100 nM), was present in the patch clamp and other experiments to avoid contributions of Ca 2+ -activated K + channels.

Ussing Chamber
Short circuit Ussing chamber measurements were performed on permeable supportgrown BCi-NS1 cells. Experiments were performed under short circuit Ussing chamber conditions (Physiological instruments) in the presence of 5% CO 2 and 25 mM HCO 3 − at 37 • C, as described earlier [29].

Materials and Statistical Analysis
All compounds used were of highest available grade of purity and were obtained from Sigma-Aldrich (St. Louis, Missouri, USA), unless indicated otherwise. Data are shown as individual traces/representative images and/or as summaries with mean values ± SEM, with the respective number of experiments given in each figure's legend. The population-effective sizes for the experiments were typically around 0.87-0.94. As we did not want to rely on results obtained in one cell line only, we compared results obtained in overexpressing HEK293 cells with those obtained in cell expressing endogenous TMEM16A channels (CFBE, BCi-NS1, HT 29 cells). For statistical analysis, paired or unpaired Student's t-test or ANOVA were used where appropriate. A p-value of < 0.05 was accepted as a statistically significant difference.

Overexpressed TMEM16A Is Spontaneously Active Due to Enhanced Ca 2+ Sensitivity
We observed in earlier studies that TMEM16A, when overexpressed in HEK293 cells, produces whole cell currents in the presence of basal cytosolic Ca 2+ concentrations of 0.1 µM and with physiological concentrations of K + , Na + , and Cl − in the patch pipette filling solution [30,31] (Figure 1C,D). Such basal currents were not observed in cells expressing TMEM16A endogenously, such as HT 29 colonic epithelial or CFBE airway epithelial cells [31] ( Figure 1A,B). Here, TMEM16A was activated only after increase in the intracellular Ca 2+ concentration, e.g., by application of the purinergic agonist ATP [31] ( Figure 1A,B). In overexpressing cells, application of ATP activated whole cell currents in addition to the already enhanced basal currents. Enhanced and activated currents were inhibited by different blockers of TMEM16A such as CaCCinhAO1 (AO1) or niflumic acid [1,30]. Moreover, enhanced basal currents were inhibited by application of AO1 or in the absence of pipette (cytosolic) Ca 2+ , indicating that TMEM16A is in charge of both enhanced basal and ATP-activated currents ( Figure 1E-H). These data strongly suggest a higher Ca 2+ sensitivity of overexpressed TMEM16A when compared with endogenous TMEM16A. ATP-activated currents ( Figure 1E-H). These data strongly suggest a higher Ca 2+ sensitivity of overexpressed TMEM16A when compared with endogenous TMEM16A. To further demonstrate enhanced Ca 2+ sensitivity of overexpressed TMEM16A, we performed patch clamp experiments in which we loaded the patch pipette with pipette filling solution containing different free Ca 2+ concentrations. At 0 µM pipette Ca 2+ no TMEM16A, whole cell currents were observed in overexpressing HEK293 cells or CFBE cells expressing endogenous TMEM16A. Whole cell currents were activated in overexpressing HEK293 cells at Ca 2+ concentrations as low as 0.01 µM (10 −8 M), while TMEM16A currents in CFBE cells were only observed at 1 µM (10 −6 M) Ca 2+ , again strongly suggesting a higher Ca 2+ sensitivity for overexpressed TMEM16A (Figure 2). At 10 µM Ca 2+ , whole To further demonstrate enhanced Ca 2+ sensitivity of overexpressed TMEM16A, we performed patch clamp experiments in which we loaded the patch pipette with pipette filling solution containing different free Ca 2+ concentrations. At 0 µM pipette Ca 2+ no TMEM16A, whole cell currents were observed in overexpressing HEK293 cells or CFBE cells expressing endogenous TMEM16A. Whole cell currents were activated in overexpressing HEK293 cells at Ca 2+ concentrations as low as 0.01 µM (10 −8 M), while TMEM16A currents in CFBE cells were only observed at 1 µM (10 −6 M) Ca 2+ , again strongly suggesting a higher Ca 2+ sensitivity for overexpressed TMEM16A (Figure 2). At 10 µM Ca 2+ , whole cell currents were smaller due to the well-known Ca 2+ -dependent inactivation of TMEM16A [23]. No whole cell currents were activated in mock-transfected HEK293 cells.

Enhanced Ca 2+ Sensitivity of Overexpressed TMEM16A is Modulated by CAM
We previously reported regulation of TMEM16A by CAM [1], which, however, has been discussed controversially (c.f. Introduction). Here, we examined activation of overexpressed TMEM16A by the activator of the CAM/KCNN4 complex, 1-EBIO. No effects were seen for 1-EBIO in mock-transfected HEK293 cells. In contrast, TMEM16A-overexpressing cells showed enhanced basal currents, which were further augmented by 1-EBIO, clearly suggesting CAM-regulation of TMEM16A. Application of the CAM-inhibitor trifluoperazine (TFP) inhibited enhanced basal currents and attenuated activation by 1-EBIO. Moreover, the TMEM16A-inhibitor CaCCinhAO1 (AO1) [32] strongly inhibited enhanced basal currents and abolished the effect of 1-EBIO ( Figure 3A-E). Finally, activation by 1-EBIO was examined in overexpressing HEK293 cells treated with scrambled RNA or with siRNA for CAM. Activation by 1-EBIO was attenuated in cells in which expression of CAM was knocked down ( Figure 3F,G). Knockdown of CAM was assessed by semiquantitative RT-PCR and was 94 ± 2 % (n = 5). A similar number of CAM-transcripts are expressed in CFBE and HEK293 cells ( Figure 3H). Notably, in contrast to TMEM16A-overexpressing HEK293 cells, no effect of siRNA-CAM was observed on the activation of TMEM16A in CFBE cells. These results suggest that enhanced basal TMEM16A currents are due to enhanced Ca 2+ sensitivity, probably modulated by association of CAM with TMEM16A.

Enhanced Ca 2+ Sensitivity of Overexpressed TMEM16A Is Modulated by CAM
We previously reported regulation of TMEM16A by CAM [1], which, however, has been discussed controversially (c.f. Introduction). Here, we examined activation of overexpressed TMEM16A by the activator of the CAM/KCNN4 complex, 1-EBIO. No effects were seen for 1-EBIO in mock-transfected HEK293 cells. In contrast, TMEM16A-overexpressing cells showed enhanced basal currents, which were further augmented by 1-EBIO, clearly suggesting CAM-regulation of TMEM16A. Application of the CAM-inhibitor trifluoperazine (TFP) inhibited enhanced basal currents and attenuated activation by 1-EBIO. Moreover, the TMEM16A-inhibitor CaCCinhAO1 (AO1) [32] strongly inhibited enhanced basal currents and abolished the effect of 1-EBIO ( Figure 3A-E). Finally, activation by 1-EBIO was examined in overexpressing HEK293 cells treated with scrambled RNA or with siRNA for CAM. Activation by 1-EBIO was attenuated in cells in which expression of CAM was knocked down (Figure 3F,G). Knockdown of CAM was assessed by semiquantitative RT-PCR and was 94 ± 2 % (n = 5). A similar number of CAM-transcripts are expressed in CFBE and HEK293 cells ( Figure 3H). Notably, in contrast to TMEM16A-overexpressing HEK293 cells, no effect of siRNA-CAM was observed on the activation of TMEM16A in CFBE cells. These results suggest that enhanced basal TMEM16A currents are due to enhanced Ca 2+ sensitivity, probably modulated by association of CAM with TMEM16A.

No Activation of Endogenous TMEM16A by 1-EBIO and No CAMKII-Dependent Regulation of TMEM16A in Airway Epithelial Cells
We examined whether CAM-dependent regulation can also be detected for TMEM16A expressed endogenously in CFBE airway epithelial cells. Expression of TMEM16A in CFBE cells has been shown previously [25,27]. We also found expression of Ca 2+ -activated KCNN4 channels in CFBE (and H441) airway epithelial cells, but not in HEK293 cells ( Figure 4A). Both CFBE and HT 29 colon epithelial cells express the splice variant TMEM16Aa,b,c ( Figure 4B). Patch clamp experiments were performed in the absence or presence of TRAM-34, a potent inhibitor of Ca 2+ -activated KCNN4 K + channels. In the absence of TRAM-34, 1-EBIO activated a whole cell current and hyperpolarized the membrane voltage ( Figure 4C,D). In contrast, in the presence of TRAM-34, no hyperpolarization was observed, and 1-EBIO did not activate whole cell currents ( Figure 4E,F). Under both conditions, ATP activated whole cell currents. The data suggest that in airway epithelial cells with endogenous TMEM16A, 1-EBIO activates KCNN4 but not TMEM16A. assessed by semiquantitative RT-PCR in both CFBE cells expressing endogenous TMEM16A and in HEK293 cells overexpressing TMEM16A. Similar number of CAM-transcripts are expressed in CFBE and HEK293 cells. Mean ± SEM (n = number of experiments). *Significant activation by 1-EBIO and inhibition by AO1 or TFP, respectively (p < 0.05; paired t-test). # Significant difference when compared to activation in the absence of AO1, TFP, or siCAM, respectively (p < 0.05; unpaired ttest).

No Activation of Endogenous TMEM16A by 1-EBIO and No CAMKII-Dependent Regulation of TMEM16A in Airway Epithelial Cells
We examined whether CAM-dependent regulation can also be detected for TMEM16A expressed endogenously in CFBE airway epithelial cells. Expression of TMEM16A in CFBE cells has been shown previously [25,27]. We also found expression of Ca 2+ -activated KCNN4 channels in CFBE (and H441) airway epithelial cells, but not in HEK293 cells ( Figure 4A). Both CFBE and HT29 colon epithelial cells express the splice variant TMEM16Aa,b,c ( Figure 4B). Patch clamp experiments were performed in the absence or presence of TRAM-34, a potent inhibitor of Ca 2+ -activated KCNN4 K + channels. In the absence of TRAM-34, 1-EBIO activated a whole cell current and hyperpolarized the membrane voltage ( Figure 4C,D). In contrast, in the presence of TRAM-34, no hyperpolarization was observed, and 1-EBIO did not activate whole cell currents ( Figure 4E,F). Under both conditions, ATP activated whole cell currents. The data suggest that in airway epithelial cells with endogenous TMEM16A, 1-EBIO activates KCNN4 but not TMEM16A.  Because CAM-dependent kinase (CAMKII) and the serine/threonine-phosphatase calcineurin require CAM to operate, we examined the effects of other CAM-activators (DCE-BIO, riluzole), CAMKII-inhibitors (KN-62, KN-93), and the calcineurin-inhibitor tacrolimus (FK-506) on basal and ATP-activated endogenous currents in airway epithelial cells. Notably, DCEBIO enhanced basal currents, and both DCEBIO and riluzole somewhat augmented ATP-activated currents but failed to reach significance ( Figure 5A,B). Tacrolimus and inhibition of CAMKII attenuated ATP-activated currents. Taken together, apart from CAMKII-dependent activation, there is no clear evidence for CAM or calcineurin-dependent regulation of endogenous TMEM16A in CFBE airway epithelial cells.
Because CAM-dependent kinase (CAMKII) and the serine/threonine-phosphatase calcineurin require CAM to operate, we examined the effects of other CAM-activators (DCEBIO, riluzole), CAMKII-inhibitors (KN-62, KN-93), and the calcineurin-inhibitor tacrolimus (FK-506) on basal and ATP-activated endogenous currents in airway epithelial cells. Notably, DCEBIO enhanced basal currents, and both DCEBIO and riluzole somewhat augmented ATP-activated currents but failed to reach significance ( Figure 5A,B). Tacrolimus and inhibition of CAMKII attenuated ATP-activated currents. Taken together, apart from CAMKII-dependent activation, there is no clear evidence for CAM or calcineurin-dependent regulation of endogenous TMEM16A in CFBE airway epithelial cells.

No Evidence for CAM, CAMKII, or Calcineurin-Dependent Regulation of Endogenous TMEM16A Expressed in Bci-NS1 Airway Epithelial Cells
In order to exclude cell-specific or methodological factors, we re-examined CAM, CAMKII, or calcineurin-dependent regulation in Ussing chamber recordings with human BCi-NS1 airway epithelial cells. When grown on permeable supports, these cells differentiate into a polarized epithelium with airway epithelial typical properties [25,33]. Cells were measured under control conditions or after exposure to KN-93, DCEBIO, or tacrolimus. Compounds were added subsequently to luminal or basolateral sides of the epithelium, and the effects of TRAM-34 (T-34), the epithelial Na + channel inhibitor amiloride (amil), the inhibitor of the cystic fibrosis transmembrane conductance regulator Cl − channel CFTRinh172 (Inh172), and ATP on short circuit currents (Isc) were examined. Neither Na + absorption by ENaC nor basal Cl − currents by CFTR or ATP-activated TMEM16A/CFTR currents were significantly affected by any of the inhibitors (Figure 6). The time course for ATP-activation of TMEM16A remained unchanged by the drugs. (B) Calculated current densities. All experiments were performed in the presence of the KCNN4-inhibitor TRAM-34. Mean ± SEM (n = number of experiments). * Significant activation by ATP (p < 0.05; paired t-test).

No Evidence for CAM, CAMKII, or Calcineurin-Dependent Regulation of Endogenous TMEM16A Expressed in Bci-NS1 Airway epithelial cells
In order to exclude cell-specific or methodological factors, we re-examined CAM, CAMKII, or calcineurin-dependent regulation in Ussing chamber recordings with human BCi-NS1 airway epithelial cells. When grown on permeable supports, these cells differentiate into a polarized epithelium with airway epithelial typical properties [25,33]. Cells were measured under control conditions or after exposure to KN-93, DCEBIO, or tacrolimus. Compounds were added subsequently to luminal or basolateral sides of the epithelium, and the effects of TRAM-34 (T-34), the epithelial Na + channel inhibitor amiloride (amil), the inhibitor of the cystic fibrosis transmembrane conductance regulator Cl − channel CFTRinh172 (Inh172), and ATP on short circuit currents (Isc) were examined. Neither Na + absorption by ENaC nor basal Cl − currents by CFTR or ATP-activated TMEM16A/CFTR currents were significantly affected by any of the inhibitors (Figure 6).

No Evidence for CAM, CAMKII, or Calcineurin-Dependent Regulation of Endogenous TMEM16A in HT 29 Colonic Epithelial Cells
We finally examined if any of the inhibitors affect ATP-activated TMEM16A whole cell currents in HT 29 colonic epithelial cells. Here, we used iodide quenching of halidesensitive yellow-fluorescent protein (YFP), stably expressed in HT 29 cells. ATP-induced quenching in HT 29 cells transfected with scrambled RNA was impressive ( Figure 7A) and was completely abolished by siRNA-knockdown of TMEM16A ( Figure 7B), indicating that TMEM16A was fully in charge of ATP, i.e., Ca 2+ -activated whole cell currents in HT 29 cells. None of the inhibitors DCEBIO, KN-62, KN-93, riluzole, or tacrolimus, applied at different concentrations, affected current activation by ATP ( Figure 7C,D). Taken together, the results suggest that CAM-dependent regulation of TMEM16A is present in overexpressing HEK293 cells but shows little effects on TMEM16A expressed endogenously in airway and colonic epithelial cells.

No Evidence for CAM, CAMKII, or Calcineurin-Dependent Regulation of Endogenous TMEM16A in HT29 Colonic Epithelial Cells
We finally examined if any of the inhibitors affect ATP-activated TMEM16A whole cell currents in HT29 colonic epithelial cells. Here, we used iodide quenching of halidesensitive yellow-fluorescent protein (YFP), stably expressed in HT29 cells. ATP-induced quenching in HT29 cells transfected with scrambled RNA was impressive ( Figure 7A) and was completely abolished by siRNA-knockdown of TMEM16A ( Figure 7B), indicating that TMEM16A was fully in charge of ATP, i.e., Ca 2+ -activated whole cell currents in HT29 cells. None of the inhibitors DCEBIO, KN-62, KN-93, riluzole, or tacrolimus, applied at different concentrations, affected current activation by ATP ( Figure 7C,D). Taken together, the results suggest that CAM-dependent regulation of TMEM16A is present in overexpressing HEK293 cells but shows little effects on TMEM16A expressed endogenously in airway and colonic epithelial cells.

Discussion
Regulation of the Ca 2+ -activated Clchannel TMEM16A has been studied extensively. In most patch clamp studies, the channel is examined as overexpressed protein activated by high pipette Ca 2+ in the presence of unphysiological salt concentrations and at 20 °C [34,35]. Our previous work showed that activation of TMEM16A by phosphatidylinositol 4,5-bisphosphate or inositol 3,4,5,6-tetrakisphosphate is well detected for overexpressed TMEM16A, but not for endogenous TMEM16A [23,24]. Moreover, other activators or inhibitors also show differential effects between overexpressed and endogenous channels [24]. The overexpressed channel is already active at basal intracellular (pipette) Ca 2+ as low as 0.01 μM. This enhanced basal current is reduced at temperatures lower than 37 °C [30,31]. Notably, similar is observed for TMEM16F, which is a phospholipid scramblase and a nonselective ion channel [31]. Interestingly, the calcium-hypersensitive aspartic acid 408-to-glycine 408 (D408G) TMEM16F mutant showed spontaneous scrambling and ion channel activity when overexpressed in HEK293 cells, but both scrambling and ion currents were not enhanced in macrophages isolated from knockin-mice expressing D408G-

Discussion
Regulation of the Ca 2+ -activated Cl − channel TMEM16A has been studied extensively. In most patch clamp studies, the channel is examined as overexpressed protein activated by high pipette Ca 2+ in the presence of unphysiological salt concentrations and at 20 • C [34,35]. Our previous work showed that activation of TMEM16A by phosphatidylinositol 4,5-bisphosphate or inositol 3,4,5,6-tetrakisphosphate is well detected for overexpressed TMEM16A, but not for endogenous TMEM16A [23,24]. Moreover, other activators or inhibitors also show differential effects between overexpressed and endogenous channels [24]. The overexpressed channel is already active at basal intracellular (pipette) Ca 2+ as low as 0.01 µM. This enhanced basal current is reduced at temperatures lower than 37 • C [30,31]. Notably, similar is observed for TMEM16F, which is a phospholipid scramblase and a nonselective ion channel [31]. Interestingly, the calcium-hypersensitive aspartic acid 408-to-glycine 408 (D408G) TMEM16F mutant showed spontaneous scrambling and ion channel activity when overexpressed in HEK293 cells, but both scrambling and ion currents were not enhanced in macrophages isolated from knockin-mice expressing D408G-TMEM16F instead of wt-TMEM16F [31,36] (data not shown). For both TMEM16A and TMEM16F, we proposed a role of lipid-dependent regulation modulating their Ca 2+ sensitivity and activity at basal Ca 2+ activity. Lipid-dependent regulation may depend on the ambient temperature and other factors such as temperature-sensitive phospholipase A2, reactive oxygen species, and the cytosolic (pipette) ion composition [31,37]. Moreover, we speculated earlier that i) massively overexpressed TMEM16A may end up in non-raft compartments, which are more accessible to other cellular Ca 2+ sources, ii) overexpressed TMEM16A may translocate its activating Ca 2+ source, i.e., the endoplasmic reticulum, by membrane tethering, and iii) unknown antagonistic accessory proteins may be stoichiometrically underrepresented in TMEM16A-overexpressing cells [23,24,38,39]. It is also entirely possible that temperature-sensitive Ca 2+ -permeable transient receptor potential (TRP) channels are involved in the temperature-sensitive activation of TMEM16A [40][41][42]. Notably, TRP channels have also been shown to be regulated by PIP 2 and CAM [43,44]. Alternatively, other accessory proteins may interact differentially under overexpressed or endogenous conditions [24].
Pre-association of Ca 2+ -free CAM (apoCAM) may preferentially occur with overexpressed TMEM16A and thus sensitize the channel for intracellular Ca 2+ [3,4]. Accordingly, the benzimidazolone 1-EBIO may then promote the association between apoCAM and TMEM16A. 1-EBIO is a known activator of Ca 2+ sensitive SK-channels [45,46]. SK channels, e.g., KCNN4, consist of a α-subunit that requires CAM as ß-subunit to gate the channel upon binding of Ca 2+ to CAM. 1-EBIO enhances the interaction between the α-subunit and CAM and thereby shifts the [Ca 2+ ] i /channel activity-relationship to lower [Ca 2+ ] i [47]. Notably, 1-EBIO did not activate any SK currents in the absence of [Ca 2+ ] i [47], and similarly, TMEM16A was not spontaneously active and was not activated by 1-EBIO in the absence of cytosolic (pipette) Ca 2+ (Figure 1).
Association of a ß-subunit such as holoCAM or apoCAM with TMEM16A may not be surprising, given the recent findings of an association of the KCNQ1-K + channel ß-subunit KCNE1 with TMEM16A [48]. Nevertheless, activation of TMEM16A by benzimidazolone compounds and the effects of CAM-regulated CAMKII or calcineurin were very moderate in epithelial cells with endogenous expression of TMEM16A. Taken together, it appears indispensable to examine whether activation by CAM and other accessory proteins/ßsubunits [48], phosphatidylinositols, temperature or small molecule activators, respectively, does exist for endogenous channels expressed in non-cultured cells and tissues under physiological conditions. Author Contributions: Conceptualization, K.T., J.O., R.C., R.S., and K.K.; methodology, K.T., J.O., R.C., and R.S.; formal analysis, K.T., J.O., R.C., R.S., and K.K.; investigation, K.T., J.O., R.C., R.S., and K.K.; data curation, K.T., J.O., R.C., R.S., and K.K.; writing-original draft preparation, K.K.; writing-review and editing, K.K. All authors have read and agreed to the published version of the manuscript.