Intracellular ATP does not inhibit Slo2.1 K+ channels

Abstract Under normal physiological conditions, the open probability of Slo2.1 K+ channels is low. Elevation of cytosolic [Na+] and [Cl−] caused by ischemia or rapid electrical pacing of cells increases the open probability of Slo2.1 channels and the resulting outward current can stabilize the resting state of cells. Initial characterization of heterologously expressed human Slo2.1 indicated that these channels were inhibited by physiological levels of intracellular ATP. However, a subsequent study found that intracellular ATP had no effect on Slo2.1 channels. Here, we re‐examine the effects of intracellular ATP on cloned human Slo2.1 channels heterologously expressed in Xenopus oocytes. Our studies provide both direct and indirect evidence that changes in intracellular [ATP] have no effect on Slo2.1 channels. First, we directly examined the effects of intracellular ATP on Slo2.1 channel activity in excised inside‐out macropatches from Xenopus oocytes. Application of 5 mmol/L ATP to the intracellular solution did not inhibit Slo2.1 currents activated by niflumic acid. Second, we lowered the [ATP]i in whole oocytes using the metabolic inhibitor NaN3. Depletion of [ATP]i in oocytes by 3 mmol/L NaN3 rapidly activated heterologously expressed KATP channels, but did not increase wild‐type Slo2.1 channel currents activated by niflumic acid or currents conducted by constitutively active mutant (E275D) Slo2.1 channels. Third, mutation of a conserved residue in the ATP binding consensus site in the C‐terminal domain of the channel did not enhance the magnitude of Slo2.1 current as expected if binding to this site inhibited channel function. We conclude that Slo2.1 channels are not inhibited by intracellular ATP.


Introduction
Intracellular Na + -activated K + (K Na ) channels were first described in guinea pig cardiomyocytes 30 years ago (Kameyama et al. 1984) and later in cultured trigeminal ganglion neurons from quail embryos (Bader et al. 1985;Haimann et al. 1990). K Na channels in neurons can either be activated by Na + ions that are conducted by voltageactivated Na + channels during the rapid upstroke phase of action potentials (Bader et al. 1985) or by a persistent inward Na + current (Budelli et al. 2009). However, in the heart it appears that K Na channels are only activated under pathological conditions such as hypoxia or ischemia (Kameyama et al. 1984). The molecular basis of K Na channels was discovered in 2003 with the cloning of Slo2.1 (Bhattacharjee et al. 2003) and Slo2.2 (Yuan et al. 2003). Based on RT-PCR, Slo2.1 is expressed in most tissues, including the brain, heart, skeletal muscle, lung and liver, whereas Slo2.2 is highly expressed within specific regions of the brain, and to a lesser extent in the kidney, testis, and very weakly in the heart (Yuan et al. 2003). The Slo2.1 channel was first described as a functional hybrid between a K Na and K ATP channel because it was activated by intracellular Na + (and Cl À ) and inhibited by intracellular ATP (Bhattacharjee et al. 2003). Activation of large conductance Slo2.1 K + channels in response to elevated [Na + ] i and depletion of intracellular ATP as occurs during hypoxia or ischemia might protect cardiomyocytes or neurons from damage such as Ca 2+ -overload by stabilizing the resting (hyperpolarized) state of the cell. Consistent with its proposed sensitivity to changes in [ATP] i , Slo2.1 channels contain a consensus ATP binding site just C-terminal to the second cytoplasmic RCK (regulate conductance of K + ) domain. Inhibition of Slo2.1 currents by intracellular ATP (~80% reduction at 5 mmol/L) in whole cells or excised membrane patches of CHO cells transfected with Slo2.1 cDNA was reported to be dependent on the integrity of this motif (Bhattacharjee et al. 2003). However, subsequent studies on striatal interneurons and HEK293T cells transfected with Slo2.1 cDNA showed no effect of intracellular ATP on Slo2.1 channel activity (Berg et al. 2007), consistent with the early studies of cardiac K Na channels (presumably Slo2.1) that were recorded in the presence of 2 mmol/L K 2 ATP in the intracellular solution (Kameyama et al. 1984). Moreover, we have found that heterologously expressed human Slo2.1 channels can be robustly activated by either niflumic acid (NFA) or elevation of [NaCl] i (Dai et al. 2010;Garg et al. 2013;Garg and Sanguinetti 2012) in intact Xenopus oocytes where [ATP] i is estimated to be 4.6 mmol/L (Gribble et al. 2000). As inhibition of Slo2.1 by intracellular ATP has been cited as one of the defining features of these channels (Bhattacharjee and Kaczmarek 2005;de Los Angeles Tejada et al. 2012;Garg et al. 2013;Garg and Sanguinetti 2012;Kaczmarek 2013;Paulais et al. 2006), in this study we have investigated the effects of altering cytosolic [ATP] on heterologously expressed human Slo2.1 channels in intact Xenopus oocytes and in excised inside-out macropatches.

Oocyte isolation and cRNA injection
Methods for isolation of oocytes from Xenopus laevis were approved by the Institutional Animal Care and Use Committee of the University of Utah. Briefly, adult female frogs were anesthetized by immersion in a 0.2% tricaine methanesulfonate solution. A small surgical incision was made to remove the ovarian lobes. To separate the individual oocytes from follicle cells, ovarian lobes were manually dispersed using tweezers and placed into a Ca 2+ -free saline solution containing 2 mg/ml each of type I and II collagenase (Worthington Biochemical Corporation, Lakewood, NJ) and gently shaken for 1-1.5 h. The Ca 2+ -free saline solution contained (in mmol/L): 96 NaCl, 2 KCl, 1 MgCl 2 , and 5 HEPES; pH was adjusted to 7.6 with NaOH.

Two-electrode voltage-clamp protocol
Standard two-electrode voltage-clamp (TEVC) techniques (Goldin 1991;Stuhmer 1992) were used to record whole cell Slo2.1 K + channel current (I Slo2.1 ) at room temperature. Oocytes were placed in a small chamber (0.3 mL volume, RC-1Z; Warner Instruments) and superfused at a rate of~2 mL/min with recording solution that contained (in mmol/L): 98 NaCl, 2 KCl, 1 CaCl 2 , 1 MgCl 2 , and 5 mmol/L HEPES; pH adjusted to 7.6 with NaOH. To study I KATP , this solution was modified by omitting NaCl and increasing the [KCl] to 104 mmol/L. Agarose-cushion microelectrodes used for TEVC recordings were fabricated as described previously (Schreibmayer et al. 1994) and had tip resistances ranging from 0.2 to 0.8 MO after back-filling with 3 M KCl. A Dell 2400 computer, GeneClamp 500 amplifier, Digidata 1322A data acquisition system, and pCLAMP 8.2 or 9.0 software (Molecular Devices, Sunnyvale, CA) were used to produce voltage protocols and to record digitized current and voltage signals.

Macropatch recordings
Multi-channel patch recordings of I Slo2.1 were performed using the standard excised inside-out configuration of the patch clamp technique (Hamill et al. 1981). The vitelline membrane from each oocyte was removed manually after 2014 | Vol. 2 | Iss. 9 | e12118 Page 2 treatment with hypertonic solution (400 mmol/L sucrose) for 2-6 min. Patch pipettes were fabricated from 1 mm OD borosilicate glass capillaries (World Precision Instruments, Sarasoto, FL) using a Sutter Instruments P-97 puller (Sutter Instrument, Novato, CA). Pipette tips were heat polished after pulling and had a resistance of 2-4 MΩ when filled with a pipette (extracellular) solution that contained (in mmol/L): 90 NaCl, 10 KCl, 1 CaCl 2 , 1 MgCl 2 , 10 HEPES and 2 niflumic acid (NFA); pH was adjusted to 7.2 with NaOH. NFA was added to the pipette solution to activate I Slo2.1 . The bath (intracellular) solution contained (in mmol/L): 90 KCl, 10 NaCl, 2 EGTA, 2 MgCl 2 , 10 HEPES and 10 lM phosphatidylinositol 4,5bisphosphate diC8 (PI(4,5)P 2 diC8); pH was adjusted to 7.2 with KOH. During the day of recording, the stock solution of PI(4,5)P 2 diC8 was kept on ice until just before use when it was added to the intracellular solution.
A Dell computer, Axopatch 200B patch clamp amplifier, Digidata 1322A data acquisition system and pCLAMP 9.0 software (Molecular Devices) were used to produce command voltage steps and to record currents. Currents were digitized at 5 kHz after on-line filtering at 1 kHz with an eight-pole low-pass Bessel filter. After patch excision, macroscopic I Slo2.1 was recorded for at least 4 min to ensure stability before 5 mmol/L K 2 ATP was added to the intracellular solution. Currents were recorded during 1-s voltage ramps applied from +80 to À140 mV. The holding potential was À80 mV and the interval between voltage ramps was 5 s.

Data analysis
pCLAMP9 (Molecular Devices) and Excel (Microsoft, Redmond, WA) software were used to analyze digitized data, and Origin 8.5 (OriginLab, Northampton, MA) was used to prepare figures. Results are expressed as mean AE SEM. (n = number of oocytes). Statistical analysis was performed using Prism 5 software (GraphPad Software, Inc., La Jolla, CA). Statistical significance was evaluated using paired Student's T-test or ANOVA (one-way or two-way) with Tukey's multiple comparison post-hoc test where appropriate. A P-value <0.05 was considered significant.

Chemicals
Sodium azide (NaN 3 ), niflumic acid (NFA) and K 2 ATP were purchased from Sigma-Aldrich (St. Louis, MO). PI (4,5)P 2 diC8 (Echelon Biosciences Inc., Salt Lake City, UT) was dissolved in the macropatch intracellular solution and stored in glass vials as a 230 lM stock at À80°C. All working solutions were prepared immediately before use by dilution of stock solutions previously stored at À20°C.

Results
Intracellular ATP has no effect on I Slo2.1 recorded from inside-out macropatches of oocyte membrane The effect of ATP applied directly to the intracellular side of inside-out membrane patches excised from Xenopus oocytes was determined. In most patches, the magnitude of I Slo2.1 decreased rapidly soon after patch excision, complicating analysis of the effects of intracellular ATP. However, the rate of current run-down was greatly reduced by including 10 lM PI(4,5)P 2 diC8 to the solution bathing the inside (cytosolic side) of the membrane patch. Slo2.1 channels can be activated by elevated [NaCl] i delivered via intracellular pipettes in intact oocytes (Garg et al. 2013). However, Slo2.1 channels were not activated in excised inside-out macropatches when [NaCl] i was elevated from 3 mmol/L to 30 or 300 mmol/L, even in the presence of 10 lM PI(4,5)P 2 diC8. Therefore, heterologously expressed Slo2.1 channels were activated by 2 mmol/L NFA in the pipette solution. I Slo2.1 was recorded under voltage clamp in response to repetitive ramping of the transmembrane voltage from +80 to À120 mV (Fig. 1A, inset). Under control conditions, large outward currents were activated almost instantly in response to stepping the voltage from À80 mV to +80 mV, and current magnitude declined as the voltage was ramped to À120 mV, becoming inward at potentials negative to À80 mV (Fig. 1A, black trace). After several voltage ramps were applied, 5 mmol/L K 2 ATP was added to the intracellular solution and the voltage ramps were continued for another 5 min. There was no difference between the currents recorded in the presence of 5 mmol/L ATP (Fig. 1A, red trace) compared to the control trace. Currents were rapidly and almost fully blocked by addition of 100 lM N-methyl verapamil (D890) to the intracellular solution (Fig. 1A, blue trace), confirming that the membrane patch was in the inside-out configuration as the permanently charged D890 only blocks I Slo2.1 when added to the intracellular side of the membrane (Garg et al. 2013). Figure 1B shows currents from the same patch after averaging multiple traces under each condition. The effect of 5 mmol/L K 2 ATP was determined in a total of 4 patches, and no significant change in I Slo2.1 was detected (Fig. 1C). simple intracellular salt solution. To determine if changes in [ATP] i under conditions where the membrane is not stretched and the cytoplasmic contents not altered, we used NaN 3 to lower [ATP] i in intact oocytes. Extracellular application of NaN 3 (3 mmol/L) was previously reported to lower [ATP] i from 4.6 mmol/L to 1.2 mmol/L and activate heterologously expressed K ATP channels in Xenopus oocytes (Gribble et al. 1997). We first confirmed the effect of NaN 3 on K ATP channels. From a holding potential of 0 mV, voltage ramps of 0.4 s duration were applied from +40 to À80 mV once every 10 s. As reported previously (Gribble et al. 1997), extracellular application of 3 mmol/L NaN 3 produced a gradual increase in whole cell currents in oocytes co-injected with Kir6.2 and SUR1 cRNAs ( Fig. 2A). I KATP started to activate within 3-5 min after NaN 3 application, reached a steady state level after about 10 min and was completely reversible upon washout ( Fig. 2A and B). There was an initial transient increase in current level after NaN 3 removal which could result from relief of direct azideinduced block of I KATP (Gribble et al. 1997). NaN 3 did not increase currents during the voltage ramps in oocytes not injected with Kir6.2 and SUR1 cRNAs (not shown). Figure 2C compares the average current amplitudes for 10 oocytes recorded at the start (+40 mV) and end (À80 mV) of the voltage ramps before, and after treatment with 3 mmol/L NaN 3 .
Oocytes injected with low amounts of Slo2.1 cRNA (~1 ng) elicit negligible currents under control conditions. As reported previously (Dai et al. 2010;Garg and Sanguinetti 2012), application of 1 mmol/L NFA to the extracellular solution causes an increase in I Slo2.1 that reaches a peak in about 4 min, followed by a slow decline over 20 min (Fig. 3A, black circles). The EC 50 for NFA activation of I Slo2.1 is 2.1 mmol/L (Dai et al. 2010). We purposely used a non-saturating concentration of NFA (1 mmol/L) for these experiments to allow for a possible further increase in response to lowering [ATP] i . In another group of oocytes, the effect of 3 mmol/L NaN 3 (in the continued presence of 1 mmol/L NFA) on I Slo2.1 was determined (Fig. 3A, red circles). After NFA-mediated activation of I Slo2.1 reached a peak in 4 min, the control recording solution was switched to a solution that contained 3 mmol/L NaN 3 plus 1 mmol/L NFA. I Slo2.1 was rapidly inhibited by NaN 3 within seconds after its application, and this effect was rapidly reversible upon washout of the NaN 3 as the magnitude of NFA-activated I Slo2.1 was the same at 20-25 min as the control group (Fig. 3A). The rapid inhibition of NFA-activated Slo2.1 current may result from NaN 3 competition with NFA binding to the channel, or from direct physical block of the channel by NaN 3 . E275D Slo2.1 channels are constitutively active in the presence of physiological [Na] i and [Cl] i and absence of NFA (Garg et al. 2013). The magnitude of E275D Slo2.1 channel currents can be increased 2.5-fold by NFA with an EC 50 of 192 lM (Garg et al. 2013), indicating that these mutant channels are not fully activated under normal physiological conditions. We examined the effects of NaN 3 on these channels to address the possibility that activation of WT channels by NFA might somehow prevent inhibition of I Slo2.1 in response to NaN 3 -induced reduction in [ATP] i . Extracellular application of 3 mmol/ L NaN 3 inhibited these mutant channels to a lesser extent  Slo2.1 contains a consensus ATP binding motif (GPKHSGKT) in the C-terminus of the channel. Mutation of the first Gly to a Ser in this motif was reported to prevent ATP from inhibiting Slo2.1 channel currents (Bhattacharjee et al. 2003). We mutated the highly conserved second Lys of this motif to Ala (K1031A). If as previously proposed, ATP inhibits I Slo2.1 by binding to this putative binding site, then mutant channels harboring a disrupted motif would be expected to have larger currents than WT channels under similar conditions. We tested this prediction by comparing the magnitude of 1 mmol/L NFA-activated I Slo2.1 conducted by WT and K1031A Slo2.1 channels in whole oocytes. Oocytes from a single isolation were injected with 1 ng of WT or K1031A Slo2.1 cRNA and recorded after 2 days of incubation to A B Figure 3. NaN 3 does not activate Slo2.1 channels heterologously expressed in Xenopus oocytes. (A) Plot of time-dependent effect of 1 mmol/L NFA (•, n = 3) or 1 mmol/L NFA co-applied with 3 mmol/L NaN 3 (•, n = 4) for 13 min (indicated by red bar) on whole-cell I Slo2.1 recorded using TEVC from oocytes expressing WT Slo2.1 channels. Currents were recorded during repetitive pulses to 0 mV and normalized relative to the peak current measured in response to NFA (at~4 min). Oocytes were injected with 1 ng WT Slo2.1 cRNA and studied 2 days later. (B) NaN 3 does not enhance currents conducted by constitutively active E275D Slo2.1 channels. Currents recorded during repetitive pulses to 0 mV were normalized relative to the initial current and plotted as a function of time after start of voltage clamp. At 5 min, 3 mmol/L NaN 3 was added to the bathing solution for a total of 13 min (indicated by red bar), then washed out (n = 7). Oocytes were injected with 0.3 ng E275D Slo2.1 cRNA and recorded 2 days later. Figure 2. NaN 3 activates K ATP channels heterologously expressed in Xenopus oocytes. (A) Voltage ramp protocol (upper panel) used to elicit whole-cell currents (lower panel) measured by TEVC. Currents were recorded under control conditions, after 10 min exposure to 3 mmol/L NaN 3 and after washout of NaN 3 . Arrow indicates 0 current level. (B) Time-dependent changes in the inward current recorded at À80 mV in a single representative oocyte induced by application and subsequent washout of 3 mmol/L NaN 3 . (C) Bar graphs comparing I KATP measured at À80 and +40 mV during voltage ramps before (control), at peak of NaN 3 effect and after washout of NaN 3 (n = 10). *P < 0.001 relative to control, # P < 0.001 relative to NaN 3 -peak current (one-way ANOVA and Tukey's multiple comparison test). achieve comparable levels of expression (Fig. 4A). The I-V relationships for both currents recorded in the absence and presence of 1 mmol/L NFA are plotted in Fig. 4B. In the absence of NFA, WT and K1031A Slo2.1 channel currents were undetectable and only very small endogenous oocyte currents were measurable (0.47 AE 0.03 lA at +80 mV). NFA application provoked a marked and similar increase of current magnitude at all potentials examined between À130 and +80 mV in oocytes expressing WT or K1031 Slo2.1 channels (Fig. 4B). These findings suggest that the putative ATP binding site in the C-terminus of Slo2.1 is non-functional.

Discussion
Due to conflicting reports in the literature, we re-examined the effects of intracellular ATP on human Slo2.1 K + channels heterologously expressed in Xenopus oocytes. Previously, Bhattacharjee et al. (2003) reported that intracellular ATP at a concentration of 5 mmol/L caused >80% inhibition of heterologously expressed human Slo2.1 channels, whereas Berg et al. (2007) reported no inhibition by intracellular ATP of native K Na channels or cloned human Slo2.1 channels. Specifically, neither constitutively expressed Cl À -activated K + currents in striatal interneurons, nor human Slo2.1 heterologously expressed in HEK293T cells measured using the whole cell voltage clamp were inhibited by 3 mmol/L intracellular ATP (Berg et al. 2007). In the present study, we found that application of 5 mmol/L K 2 ATP to the cytosolic side of excised inside-out macropatches did not inhibit I Slo2.1 , and that mutation of a consensus ATP binding motif did not enhance I Slo2.1 . Unlike the previous studies, we included 10 lM PIP 2 in the cytosolic solution to prevent run-down of current and used NFA in most experiments to activate Slo2.1. It is possible that these compounds may have affected the response of Slo2.1 channels to ATP in excised inside-out patch recordings. To control for this possibility, we showed that lowering [ATP] i with NaN 3 also did not enhance E275D channel currents. Although these mutant channels are constitutively active, they can be further activated by NFA or elevated [NaCl] i (Garg et al. 2013). Overall, our findings are in agreement with Berg et al. (2007) that intracellular ATP does not inhibit heterologously expressed human Slo2.1 channels.
We performed experiments to determine if lowering [ATP] i by treatment of intact oocytes with NaN 3 might enhance Slo2.1 channel activity. I Slo2.1 was recorded from whole oocytes before and after treating oocytes with the metabolic inhibitor, NaN 3 that inhibits cytochrome C oxidase in the mitochondrial electron transport chain and causes depletion of intracellular ATP (Trapp and Ashcroft 2000). At physiological pH, NaN 3 is almost completely dissociated in solution and azide ions (N À 3 ) combine with extracellular protons to form HN 3 , a weak acid that can penetrate the plasma membrane (Trapp and Ashcroft 2000). In Xenopus oocytes, 10 min exposure to 3 mmol/L NaN 3 reduced the [ATP] i from 4.6 AE 0.3 mmol/L to 1.2 AE 0.1 mmol/L (Gribble et al. 2000). We confirmed that treatment of oocytes with 3 mmol/L NaN 3 for 13 min activated heterologously expressed K ATP (Kir6.2/SUR1) channels, but found that the same treatment did not activate either constitutively active E275D Slo2.1 channels or WT Slo2.1 channels partially activated by NFA.
Slo2.1 contains a consensus ATP binding site (1025-GPKHSGKT-1032) that is located distal to the second RCK domain in the C-terminal. The sequence GXXXXGKT, known as Walker motif A (W-motif) has been shown to bind nucleotides in a wide variety of proteins [e.g. aand b-subunits of F1-ATPase, myosin and other ATP-requiring enzymes (Walker et al. 1982)]. The W-motif forms a loop around the nucleotide and the highly conserved Lys and Thr residues bind the phosphate group of ATP. An analysis of the crystal structure of 92 proteins that have a W-motif revealed that in about half of these proteins, the motif did not form a loop or did not bind ATP (Ramakrishnan et al. 2002). Our finding that K1031A Slo2.1 channel currents were of the same magnitude as WT Slo2.1 channel currents indicates that the W-motif is nonfunctional in Slo2.1. Together our findings indicate that intracellular ATP does not inhibit human Slo2.1 channels heterologously expressed in Xenopus oocytes, consistent with the lack of effect of ATP on these channels when heterologously expressed in HEK293T cells (Berg et al. 2007). This finding implies that during ischemia it is the elevation of [Na + ] i and [Cl -] i , but not the reduction of [ATP] i that causes activation of Slo2.1 K + channels.