Characterization of Two-Pore Channel 2 by Nuclear Membrane Electrophysiology

Lysosomal calcium (Ca2+) release mediated by NAADP triggers signalling cascades that regulate many cellular processes. The identification of two-pore channel 2 (TPC2) as the NAADP receptor advances our understanding of lysosomal Ca2+ signalling, yet the lysosome is not amenable to traditional patch-clamp electrophysiology. Previous attempts to record TPC2 single-channel activity put TPC2 outside its native environment, which not reflect TPC2’s true physiological properties. To test the feasibility of using nuclear membrane electrophysiology for TPC2 channel characterization, we constructed a stable human TPC2-expressing DT40TKO cell line that lacks endogenous InsP3R and RyR (DT40TKO-hTPC2). Immunostaining revealed hTPC2 expression on the ER and nuclear envelope. Intracellular dialysis of NAADP into Fura-2-loaded DT40TKO-hTPC2 cells elicited cytosolic Ca2+ transients, suggesting that hTPC2 was functionally active. Using nuclear membrane electrophysiology, we detected a ~220 pS single-channel current activated by NAADP with K+ as the permeant ion. The detected single-channel recordings displayed a linear current-voltage relationship, were sensitive to Ned-19 inhibition, were biphasically regulated by NAADP concentration, and regulated by PKA phosphorylation. In summary, we developed a cell model for the characterization of the TPC2 channel and the nuclear membrane patch-clamp technique provided an alternative approach to rigorously investigate the electrophysiological properties of TPC2 with minimal manipulation.

TPC2 channel is not mediated by NAADP but phosphatidylinositol 3,5-bisphosphate [PI(3,5)P 2 ] 14,15 ; this finding prompts the need for the development of a novel methodology for characterizing the hTPC2 channel in its native membrane. In this regard, we generated a stable DT40 cell line expressing hTPC2 (DT40TKO-hTPC2) that lacked both functional InsP 3 R and RyR (DT40TKO) 16,17 to eliminate the influences by these intracellular Ca 2+ release channels. Using the nuclear membrane patch-clamp technique, we detected a ~220 pS single-channel current activated by NAADP with K + as the permeant ion. The detected single-channel recordings displayed a linear current-voltage relationship, were inhibited by Ned-19, were biphasically regulated by NAADP concentrations, and its channel open probability (P o ) was regulated by PKA phosphorylation.
Taken together, we developed a cell model with minimal manipulation that, combined with nuclear membrane electrophysiology 18 , enabled us to rigorously investigate the biophysical properties of the TPC2 channel in its native membrane.

Results
Generation of a stable hTPC2-expressing DT40TKO line. The single-channel properties of hTPC2 were successfully revealed by the use of lipid bilayer electrophysiology and retargeting channel proteins to the plasma membrane 10,12,13 ; these approaches, however, have major drawbacks that might not truly reflect the biophysical properties of hTPC2 in its native membrane environment. Human TPC2 is predominantly expressed in the late endosome, lysosome, and ER membrane 11 ; therefore, we tested whether nuclear membrane electrophysiology 18,19 could be employed to characterize the electrophysiological properties of the hTPC2 channel. As proposed by Cancela and others 12,20 , NAADP initiates local Ca 2+ release from the TPC2 channel that is amplified subsequently by ER Ca 2+ releasing channel through the Ca 2+ -induced Ca 2+ release. To prevent the influence of other intracellular Ca 2+ release channels (InsP 3 Rs and RyRs), a subline of InsP 3 R deficient chicken B lymphocyte, DT40TKO was used 16,17 . This cell line is insensitive to anti-IgM ligation and caffeine stimulation ( Supplementary  Fig. 2). We established a stable hTPC2 expressing DT40TKO cell line by retroviral infection (DT40TKO-hTPC2). Western blot was used to confirm hTPC2 expression in the DT40TKO-hTPC2 cell line; as shown in Fig. 1a, a band of ~83 kDa was detected by the anti-hTPC2 antibody in cell lysates prepared from the DT40TKO-hTPC2 line but not from the control EGFP-expressing cells. To confirm the location of the expressed hTPC2, a GFPtagged, hTPC2-expressing DT40TKO cell line (DT40TKO-hTPC2-GFP) was generated. Fluorescence microscopy and ER-Tracker Blue-White DPX counterstaining revealed that hTPC2 was expressed in the ER and the nuclear membrane (arrowhead) of an exposed nucleus (Fig. 1b). To verify whether the expressed hTPC2 in DT40TKO-hTPC2 cells formed functional channels, we dialysed NAADP to a DT40TKO-hTPC2 cell using a patch pipette in the whole-cell configuration and monitored the changes in cytosolic [Ca 2+ ]. After the whole-cell configuration was achieved, hTPC2-expressing cells showed potentiated increases in cytosolic [Ca 2+ ]; in contrast, this NAADP-mediated cytosolic [Ca 2+ ] increases was absent in control EGFP-expressing cells (Fig. 2a). The rate of NAADP-elicited Ca 2+ response in hTPC2-expressing cells was significantly higher than those of the control cells (p = 0.0212, n = 3 by unpaired Student's t-test), as shown in Fig. 2b. Taken together, our data demonstrate that functional and stably-expressed hTPC2 localized to the ER and nuclear membrane was generated in DT40TKO cells that lacked functional InsP 3 R and RyR.
NAADP-activated hTPC2 single-channel current in nuclear envelope. Ca 2+ imaging techniques have revealed that NAADP mobilizes lysosomal Ca 2+ via the activation of TPC2 channels 9,11 and Pitt et al. demonstrated, in planar lipid bilayer, that NAADP dose-dependently activates the hTPC2 channel 10 . To test if NAADP can directly activate hTPC2 expressed on the nuclear envelope, we performed nuclear membrane electrophysiology on the DT40TKO-hTPC2 in the "on-nucleus" configuration. Nuclei can be obtained from DT40TKO-hTPC2 cells by mechanical rupture 18,21 (Fig. 3a); gigaohm seals were readily achieved in most of the exposed nuclear membranes. In symmetrical 140 mM KCl (K + as the charge carrier) with 10 nM NAADP, single-channel currents were detected in ~30% of the nuclear patches (n = 412 patches). A 30-second representative current trace of our nuclear patch in symmetric K + recorded at + 60 mV is shown in Fig. 3b. In the presence of 10 nM NAADP in the pipette solution, hTPC2 single-channel currents with a channel open probability (P o ) of 0.39 ± 0.07 (n = 3) were detected from nuclei isolated from DT40TKO-hTPC2 cells; this NAADP-activated single-channel current was not observed when we patched the DT40TKO-EGFP nuclei (data not shown). An amplitude histogram revealed that the detected hTPC2 channels had current amplitudes similar to those obtained from the planar lipid bilayer study ( Fig. 3c; current amplitude at + 60 mV was 13.3 ± 0.29 pA, resulting in a conductance of ~220 pS). Our results show that NAADP activated the hTPC2 channels expressed in the nuclear membrane of DT40TKO-hTPC2 cells and that nuclear membrane electrophysiology is suitable for single-channel recording of NAADP-activated hTPC2 channels.
TPC2 channel on the nuclear membrane is permeable to K + and Cs + . As demonstrated in the reconstituted planar lipid bilayer, hTPC2 forms a large conductance, cation-selective channel permeable to monovalent and divalent ions, similar to other intracellular Ca 2+ release channels 10,12 . We performed analogous experiments by nuclear membrane patch to investigate whether the conductance and permeation properties observed by nuclear membrane patch resemble those observed in the planar lipid bilayer experiments. Current/voltage (I/V) relationships were determined using 2 μ M Ca 2+ in the pipette solution, with symmetrical K + or Cs + as the permeant ion (140 mM pipette:bath). Fig. 4a (K + ) and Fig. 4b (Cs + ) depict representative current sweeps that lasted for 2 s each, with membrane potential steps from − 60 mV to + 60 mV in 20− mV increments. The I/V relationships generated from these experiments are illustrated in Fig. 4c. The detected hTPC2 channels, in both symmetric K + and Cs + , exhibited an ohmic I/V relationship with no discernible voltage dependence. The slope conductance of K + as the charge carrier was 208.4 ± 20.7 pS, which resembles those reported in the lipid Scientific RepoRts | 6:20282 | DOI: 10.1038/srep20282 bilayer 10,12 ; in comparison, the slope conductance of Cs + was 77.5 ± 13.6 pS (Fig. 4c). The conductance and permeation properties of the hTPC2 channel recorded from the nuclear membrane are similar to those reported in previous studies with the lipid bilayer and hTPC2 retargeting to the plasma membrane 10,12 .
NAADP biphasically regulates hTPC2 channel activity. NAADP dose-dependently regulates lysosomal Ca 2+ release and two distinct NAADP binding sites have been proposed 20,22 . To investigate this unique NAADP dependence characteristic, we performed nuclear membrane patches on DT40TKO-hTPC2 nuclei and included different concentrations of NAADP in the pipette solution. As shown in the representative traces in Fig. 5a, no TPC2 channel activity was detected when NAADP was omitted from the pipette solution or when 1 μ M Ned-19, hTPC2 channel activity is regulated by protein kinase A phosphorylation. Apart from regulation by physiological ligands and Ca 2+ , the channel activity of InsP 3 Rs and RyRs are modulated by different protein kinases 6,7 . We analysed the TPC2 protein sequences and found a putative protein kinase A (PKA) phosphorylation site at position 666 in human, which is conserved in several other mammals (Fig. 6a). We therefore investigated the influence of PKA phosphorylation on hTPC2 channel activity using DT40TKO stable cell lines expressing either phosphomimetic (S666E) or unphosphorylatable (S666A) mutants. Western blot analysis confirmed the expression of TPC2 proteins from cell lysates isolated from wild-type or mutant-expressing cells ( Supplementary Fig. 4) and suggests that PKA phosphorylation of hTPC2 did not alter hTPC2 expression. As shown in the representative single-channel current traces in Fig. 6b, channel P o was significantly increased in the phosphomimetic S666E mutant cell line (P o = 0.80 ± 0.09; n = 3) as compared to wild-type hTPC2-expressing cells (P o = 0.63 ± 0.06; n = 3) (p = 0.0151), suggesting that PKA phosphorylation of hTPC2 augments its activity; in contrast, the unphosphorylatable S666A mutation decreased channel activity (P o = 0.33 ± 0.01 vs 0.63 ± 0.06 in wild-type, p = 0.0062). Average changes in channel P o and channel open and closed times in wild-type and mutant-expressing cells are shown in Fig. 6c,d and Supplementary Fig. 5. To characterize how PKA phosphorylation affects hTPC2 channel gating, burst analysis was performed using a T c (the time which separates inter-burst closures from intra-burst closures) of 4 ms; this value was determined from analysis of wild-type TPC2 closed-time histograms containing a single channel studied in the presence of 100 nM NAADP ( Supplementary  Fig. 1). Figure 6e shows the effect of PKA phosphorylation on channel inter-burst duration. Cells with the S666A mutation showed significantly increased inter-burst duration relative to both wild-type ( Fig. 6e; 67.89 ± 4.30 vs 35.23 ± 0.98, p = 0.0149) and S666E ( Fig. 6e; 67.89 ± 4.30 vs 31.03 ± 6.79, p = 0.009). To further validate PKA phosphorylation enhances TPC2 channel activity, we patched DT40TKO-hTPC2 nuclei with the addition of PKA catalytic subunit in the pipette solution. With the presence of PKA catalytic subunit and 10 nM NAADP in the pipette, hTPC2 channel P o was increased significantly as compared to those in the absence of PKA (Fig. 6f, P (Fig. 6d) and increasing the inter-burst duration (Fig. 6e).

Discussion
Ever since the TPC2 channel was proposed to be the NAADP-gated lysosomal Ca 2+ release channel, characterization of TPC2′ s electrophysiological properties has been one of the field's top priorities. Macroscopic electrophysiological properties of TPC2 have been investigated by patch clamping of vacuolin enlarged lysosome 14,15,23 , whereas microscopic TPC2 single-channel currents were recorded by reconstituting immunopurified TPC2 protein into artificial bilayer 10,13 , or by patch clamping of TPC2 channel re-targeted to the plasma membrane 12 . TPC2 properties detected by these approaches may not reflect their real physiological conditions. The whole lysosome patch clamp approach used vacuolin to induce fusion of endosomes and lysosomes 24 . Whether this chemical-induced endosome/lysosome fusion affects TPC2 functions and the luminal ion compositions are uncertain. These may explain why data from this approach are controversial to those obtained from bilayer experiments [activated by NAADP vs PI(3.5)P 2 and non-selective vs Na + -selective 25,26 ]. Furthermore, evidence has been suggested that NAADP may bind to accessory proteins to activate the TPCs 27,28 . The bilayer method requires tedious solubilization and purification procedures; therefore, the accessory protein required for TPC activation may not be able to reconstitute into the bilayer, which may explain why the maximum TPC2 P o recorded from the bilayer method was relatively low 10 . While the plasma retargeting method of hTPC2 successfully demonstrated hTPC2 single-channel activity in the plasma membrane, the maximum detected P o was also low in this system. Moreover, the plasma retargeting approach involves genetic modification of the channel protein at its N-terminus 12 . In a separate study by the same research team, they demonstrated that the N-terminus of TPC1, and possibly TPC2, is critical for NAADP binding 29,30 . Genetic modification of the channel at the N-terminus may adversely affect the physiological properties of the TPC channel, particularly its regulation by NAADP. To overcome these limitations and uncouple the influence of other intracellular Ca 2+ release channels, we generated a stable hTPC2-expressing cell line in a sub-line of DT40 (DT40TKO) that lacks functional InsP 3 R and RyR ( Supplementary Fig. 2) and tested the feasibility of using nuclear membrane patch-clamp to characterize the TPC2 channel in its native membrane environment. As shown by immunocytochemistry and Ca 2+ imaging, we demonstrated that hTPC2 is expressed in the nuclear envelope of DT40TKO cells and our cell model is functionally responsive to intracellular NAADP dialysis (Figs 1 & 2).
Using nuclear membrane electrophysiology in the "on-nucleus" configuration, we demonstrated hTPC2 single-channel recordings without any genetic modification. Inclusion of 10 nM NAADP elicited a single-channel current with ~220 pS conductance when symmetrical K + was the permeant ion (Fig. 3). Similar to the bilayer result, we observed a delayed onset of NAADP-activated channel activity (as shown in the slow Ca 2+ mobilization kinetics in Fig. 2 and progressive increases of P o in Fig. 3). As proposed, NAADP may interact with an unidentified accessory protein associated with the TPC protein complex instead of directly binding to the hTPC channel 27,28 . Our data supports the suggestion that NAADP needs to interact with the accessory protein to fully activate the hTPC2 channel.
The hTPC2 currents detected by nuclear membrane patch-clamp had many similar biophysical properties to those detected by bilayer methods 10,13 . NAADP activates the opening of the hTPC2 channel, which is sensitive to Ned-19 inhibition (Fig. 5). In Fig. 4, the open channel showed a linear I/V relationship and permeable to both K + and Cs + , with slope conductances of 208.4 ± 20.7 pS and 77.5 ± 13.6 pS, respectively, suggesting that the hTPC2 channel is more permeable to K + . Controversial studies suggested that the TPC2 channel is a Na + channel activated by PI(3,5)P 2 but not NAADP 14,15 . Although we did not study Na + permeability using nuclear membrane electrophysiology (due to the low solubility of PI(3,5)P 2 in pipette solution), our permeability results and others have clearly demonstrated that hTPC2 is a non-selective cation channel activated by NAADP. Nevertheless, the use of nuclear membrane patch-clamp for the investigation of hTPC2 has its limitations. All our hTPC2 current traces were recorded in the "on-nucleus" configuration; the luminal nuclear environment could  not be readily controlled under such configuration. As the sensitivity and affinity of hTPC2 to NAADP has been demonstrated to be dependent on luminal [Ca 2+ ] and pH, respectively 10,13 , further nuclear membrane patches in the "luminal-side-out" and "cytoplasmic-side-out" configurations 18,19,31 are required to fully characterize the electrophysiological properties of hTPC channels.
Nonetheless, the "on-nucleus" nuclear membrane patch of hTPC2 can provide some mechanistic insights into the regulation of channel activity by NAADP. Like the data from Ca 2+ imaging studies, our nuclear membrane patch recapitulated the unique bell-shaped regulation of hTPC2 channel activity by [NAADP]. Nanomolar NAADP concentrations led to channel activation, whereas micromolar NAADP concentrations inactivated the channel (Fig. 5a,b). From the channel dwell time analyses (Fig. 5c & Supplementary Fig. 3), the increase in hTPC2 The channel activities of InsP 3 R and RyR are critically regulated by phosphorylation 6,7 ; however, phosphorylation of TPCs by kinases has never been investigated. We identified a putative PKA phosphorylation site in hTPC2 and a phosphomimetic mutation showed significant increase in channel P o (Fig. 6c). The augmentation of channel P o by PKA phosphorylation is caused by stabilization of the channel's open state, possibly due to the shortening of inter-burst intervals (Fig. 6e). Although the inter-burst intervals in the S666E mutant were not significantly different than those of wild-type hTPC2-expressing cells, a one-fold difference in inter-burst interval was observed between the S666E and S666A mutants. Furthermore, wild-type hTPC2 channel P o was significantly increased with the presence of PKA catalytic subunit while the S666A mutation remained unaffected (Fig. 6f,g). These data suggests that the wild-type hTPC2 channel is partially phosphorylated which is agreed with our in vitro PKA activation assay (Fig. 6f,g). Further investigations on NAADP-induced Ca 2+ release by imaging and single-channel patch-clamp experiments in a broad range of NAADP concentration may provide detail mechanistic insights into the regulation of hTPC2 by PKA phosphorylation.
In summary, the permeation and conducting properties of hTPC2 detected by nuclear membrane patch-clamp are consistent with those recorded in the lipid bilayer experiments. More importantly, our approach demonstrated the unique bell-shaped regulation of hTPC2 channel activity by [NAADP] and that channel activity is modulated by PKA phosphorylation at position S666. The application of the nuclear membrane patch-clamp technique to a DT40TKO-hTPC2 cell provides a robust system to characterize the electrophysiological properties of the hTPC2 channel in native nuclear membrane with minimal artificial manipulations.

Methods
Generation of stable human TPC2 (hTPC2)-expressing cell lines. The cDNA encoding hTPC2 9 was kindly provided by Professor Antony Galione (University of Oxford, Oxford, England). The hTPC2 cDNA was sub-cloned into pΔ MX-IRES-EGFP 32 at the EcoRI and NotI sites. Phosphomimetic (Ser-666-Glu) and non-phosphorylatable (Ser-666-Ala) hTPC2 mutants were generated using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA). Verified mutant constructs were inserted into a pΔ MX-IRES-EGFP retroviral vector. Stable hTPC2-expressing cell lines and control EGFP-expressing line were generated by a retroviral expression system using a standard protocol. In brief, retroviral particles were produced by transfecting HEK293T cells with 5 μ g of retroviral vector with the gene of interest, 4.5 μ g of pUMVC (addgene #8449), and 0.5 μ g of pVSV-G (addgene #8454) using the MegaTran transfection reagent (OriGene, Rockville, MD). Viral particles were collected at 48 and 72 hours after transfection. Inositol trisphosphate receptor deficient DT40TKO cells were infected by the retrovirus and the stably transduced cells were selected using flow cytometry by selecting for the GFP-positive cells. Stable cells were then expanded and frozen. Human TPC2 expression was assayed by western blotting and immunocytochemistry. In some experiments, C-terminal GFP-tagged hTPC2 was used. GFP-tagged hTPC2 was generated by sub-cloning hTPC2 into the pENTR1A-GFP-N2 vector (Addgene #19364) and the stable DT40TKO-hTPC2-GFP line was generated using a standard protocol.
Western blot analysis. DT40TKO-hTPC2 and control DT40TKO-EGFP cells were washed with ice-cold phosphate-buffered saline (PBS). Cells were lysed in lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 1% Triton X-100, pH 8.0) with protease inhibitors (Complete Protease Inhibitor Tablets, EDTA-free, Roche, Basel, Switzerland). Cell lysates were centrifuged at 13,200 x g at 4 °C for 30 min and the supernatant was collected. Proteins were separated using 10% SDS-PAGE gels and electrophoretically transferred to polyvinylidene difluoride membranes. Membranes were blocked by incubating the membrane for 1 hour in Tris-buffered saline containing 0.1% Tween-20 (TBST) with 5% non-fat milk at room temperature. Membranes were then incubated with anti-hTPC2 antibody (Cat#: Y158030; 1:500; Applied Biological Materials Inc., Richmond, BC, Canada) overnight at 4°C in TBST with 5% non-fat milk. Membrane was washed with TBST and incubated with HRP-conjugated anti-rabbit IgG secondary antibody (1:10,000, Bio-Rad Laboratories, Hercules, CA, USA) for 2 hours at room experiments and were analysed by one-way ANOVA with Tukey's post hoc test. Individual dwell time analysis was presented in Supplementary Fig. 5. (f) Representative current traces from isolated nuclei of wild-type hTPC2 and S666A stimulated by 10 nM NAADP together with or without PKA catalytic subunit (50 nM) in the pipette solution. (g) Summary of effects of PKA catalytic subunit on single-channel P o detected from wild-type hTPC2 and hTPC2-S666A nuclei. Data were summarized as mean ± SEM from 3 individual experiments and were analysed by one-way ANOVA with Tukey's post hoc test. temperature in TBST with 5% non-fat milk. Chemiluminescence emitted upon the addition of HRP substrate (Millipore, Billerica, USA) was captured by X-ray film.
Fluorescence microscopy. DT40TKO-hTPC2-GFP cells were placed on poly-D-lysine-coated coverslips and stained with 500 nM ER-Tracker Blue-White DPX (Invitrogen) for 30 min at 37 °C. The coverslips were then rinsed with PBS and replaced with fresh ER-tracker-free medium. Stained cells were visualized under a Nikon Eclipse Ti microscope using a 40x oil-immersion objective (Nikon CFI S Fluor Objective, Nikon, Tokyo, Japan). Fluorescence images of ER Tracker Blue staining and GFP fluorescence were captured using a SPOT RT3 CCD digital microscope camera (SPOT Imaging Solutions, Michigan, USA) and analysed using ImageJ (U.S. National Institutes of Health, Bethesda, Maryland).
Isolation of nuclei and patch-clamp recording. Isolation of DT40TKO-hTPC2 nuclei for patch-clamp recordings were prepared by homogenizing cells in nuclei isolation solution (150 mM KCl, 250 mM sucrose, 10 mM Tris-HCl, 1.4 mM β -mercaptoethanol, 0.1 mM PMSF, Complete protease inhibitors, pH 7.3) as previously described 18,21,32 . In brief, cells were washed in ice-cold PBS and centrifuged at 300 × g for 5 min. The cell pellet was resuspended in an appropriate volume of ice-cold nuclei isolation solution. Resuspended cells were transferred to an ice-cold homogenizer (1-mL Duall homogenizer, Kimble-Chase, Vineland, NJ) and were subjected to 12 strokes of homogenization. Forty μ l of cell homogenate were placed in 1 ml of bath solution (140 mM KCl or CsCl, 10 mM HEPES, 200 nM free Ca 2+ , 0.5 mM BAPTA, pH 7.3) and allowed to adhere to a glass-bottom culture dish for 10 min prior to electrophysiological experiments. Isolated nuclei were morphologically distinguishable from intact cells based on their unique morphology (Fig. 3a). The "on-nucleus" patch-clamp configuration was employed to detect the single hTPC2 channel using a HEKA EPC-10 amplifier (HEKA Elektronik) and pipets filled with pipette solution (140 mM KCl, 10 mM HEPES, 0.5 mM Na 2 ATP, 2 μ M free Ca 2+ buffered with 0.5 mM 5,5′ -dibromo BAPTA, pH 7.3). Free [Ca 2+ ] in solutions was adjusted by Ca 2+ chelators with appropriate affinities and verified by fluorometry, as previously described 34 . Pipette resistances generally fell between 10 and 20 MΩ and seal resistances were >1 GΩ , respectively. NAADP was added directly to the patch pipette solution. In some experiments, 50 nM protein kinase A (PKA) catalytic subunit (Cat#: 539576, EMD Millipore Corp., Billerica, MA) was added to the pipette solution together with 10 nM NAADP. Single TPC2 channel traces were sampled at 5 kHz and filtered at 1 kHz. Only recordings that lasted for at least 30s were used for data analysis. Data analysis. Data were acquired using PATCHMASTER software (HEKA Elektronik). Single-channel current traces exhibiting one or two TPC2 channels were used for open probability (P o ) analysis using QuB software 35 . All-points current amplitude histograms were generated from the current records and fitted with normal Gaussian probability distribution functions. The coefficient of determination (R 2 ) for every fit was > 0.95. Slope conductance was determined from linear fitting of current-voltage (I/V) relationships. Open and closed dwell times and burst analysis were performed using the Clampfit 10 Data Analysis Module (pClamp 10, Molecular devices, CA, USA). Channel dwell time constants for the open states were determined from mono-exponential fitting of the open dwell time histograms. Logarithmic plots of dwell closed times revealed two populations, with one larger than the other by an order of magnitude, which indicated the existence of multiple closed states. The burst delimiter (T c ) was defined as 4 ms, as previously described 36,37 . Any closing time that was longer than T c was defined as an inter-burst interval, whereas any time shorter than T c was defined as closed time. These values were calculated after curve fitting of the histograms using the exponential logarithmic probability function. The coefficient of determination (R 2 ) was determined for every theoretical curve fitting and considered acceptable when greater than 0.95.