Advertisement

Deficiency of the Two-Pore Potassium Channel KCNK9 Impairs Intestinal Epithelial Cell Survival and Aggravates Dextran Sodium Sulfate-Induced Colitis

Open AccessPublished:August 13, 2022DOI:https://doi.org/10.1016/j.jcmgh.2022.08.003

      Background & Aims

      The 2-pore potassium channel KCNK9 regulates intracellular calcium concentration and thus modulates cell survival and inflammatory signaling pathways. It also was recognized as a risk allele for inflammatory bowel disease. However, it remains unclear whether KCNK9 modulates inflammatory bowel disease via its impact on immune cell function or whether its influence on calcium homeostasis also is relevant in intestinal epithelial cells.

      Methods

      Kcnk9-/- mice were challenged with 3% dextran sulfate sodium (DSS) to induce experimental acute colitis. Primary cultures of intestinal epithelial cells were generated, and expression of potassium channels as well as cytosolic calcium levels and susceptibility to apoptosis were evaluated. Furthermore, we evaluated whether KCNK9 deficiency was compensated by the closely related 2-pore potassium channel KCNK3 in vivo or in vitro.

      Results

      Compared with controls, KCNK9 deficiency or its pharmacologic blockade were associated with aggravated DSS-induced colitis compared with wild-type animals. In the absence of KCNK9, intestinal epithelial cells showed increased intracellular calcium levels and were more prone to mitochondrial damage and caspase-9–dependent apoptosis. We found that expression of KCNK3 was increased in Kcnk9-/- mice but did not prevent apoptosis after DSS exposure. Conversely, increased levels of KCNK9 in Kcnk3-/- mice were associated with an ameliorated course of DSS-induced colitis.

      Conclusions

      KCNK9 enhances mitochondrial stability, reduces apoptosis, und thus supports epithelial cell survival after DSS exposure in vivo and in vitro. Conversely, its increased expression in Kcnk3-/- resulted in less mitochondrial damage and apoptosis and was associated with beneficial outcomes in DSS-induced colitis.

      Graphical abstract

      Keywords

      Abbreviations used in this paper:

      A293 (2-(butane-1-sulfonyl-amino)-N-[1-(R)-(6-methoxypyridin-3-yl)-propyl]-benz-amide), DMEM (Dulbecco’s modified Eagle medium), DMSO (dimethyl sulfoxide), DSS (dextran sulfate sodium), ER (endoplasmic reticulum), IBD (inflammatory bowel disease), IEC (intestinal epithelial cell), IL (interleukin), JC-1 (5,5',6,6'-tetrachloro-1,1',3,3'-tetraethylbenzimidazolocarbocyanine iodide), K2P (2-pore potassium (channel)), mRNA (messenger RNA), PBS (phosphate-buffered saline), PK-THPP (1-[1-[6-([1,1′-biphenyl]-4-ylcarbonyl)-5,6,7,8-tetrahydropyrido[4,3-d]pyrimidin-4-yl]-4-piperidinyl]-1-butanone), PCR (polymerase chain reaction), TNF-α (tumor necrosis-factor α)
      Deficiency or blockade of the 2-pore potassium channel KCNK9 aggravates experimentally induced colitis. KCNK9 promotes mitochondrial stability and thus limits caspase-9–dependent apoptosis in intestinal epithelial cells. Lack of KCNK9 cannot be compensated by KCNK3, although the structural similarity is enormous.
      The 2-pore potassium channel (K2P) KCNK9 was identified as a major modulator of intracellular calcium (Ca2+) levels and associated signaling pathways.
      • Enyedi P.
      • Czirjak G.
      Molecular background of leak K+ currents: two-pore domain potassium channels.
      Initially described as a contributor to resting membrane potential in excitable cells such as neurons, evidence for its role in the modulation of signal responses in immune cells has increased in recent years.
      • Bittner S.
      • Budde T.
      • Wiendl H.
      • Meuth S.G.
      From the background to the spotlight: TASK channels in pathological conditions.
      In experimental autoimmune encephalomyelitis, KCNK9 significantly contributes to immune cell activation and disease severity.
      • Meuth S.G.
      • Bittner S.
      • Meuth P.
      • Simon O.J.
      • Budde T.
      • Wiendl H.
      TWIK-related acid-sensitive K+ channel 1 (TASK1) and TASK3 critically influence T lymphocyte effector functions.
      ,
      • Bittner S.
      • Bauer M.A.
      • Ehling P.
      • Bobak N.
      • Breuer J.
      • Herrmann A.M.
      • Golfels M.
      • Wiendl H.
      • Budde T.
      • Meuth S.G.
      The TASK1 channel inhibitor A293 shows efficacy in a mouse model of multiple sclerosis.
      Furthermore, genetic variants in the Kcnk9 gene recently were associated with an increased risk for ulcerative colitis, a variant of inflammatory bowel disease (IBD).
      • Saadati H.R.
      • Wittig M.
      • Helbig I.
      • Hasler R.
      • Anderson C.A.
      • Mathew C.G.
      • Kupcinskas L.
      • Parkes M.
      • Karlsen T.H.
      • Rosenstiel P.
      • Schreiber S.
      • Franke A.
      Genome-wide rare copy number variation screening in ulcerative colitis identifies potential susceptibility loci.
      Previous work indicated various mechanisms by which KCNK9 and other closely related potassium channels such as KCNK3 modulate the intracellular calcium level, involving Gq-proteins and diacylglycerol as direct negative regulators.
      • Wilke B.U.
      • Lindner M.
      • Greifenberg L.
      • Albus A.
      • Kronimus Y.
      • Bunemann M.
      • Leitner M.G.
      • Oliver D.
      Diacylglycerol mediates regulation of TASK potassium channels by Gq-coupled receptors.
      ,
      • Bista P.
      • Pawlowski M.
      • Cerina M.
      • Ehling P.
      • Leist M.
      • Meuth P.
      • Aissaoui A.
      • Borsotto M.
      • Heurteaux C.
      • Decher N.
      • Pape H.C.
      • Oliver D.
      • Meuth S.G.
      • Budde T.
      Differential phospholipase C-dependent modulation of TASK and TREK two-pore domain K+ channels in rat thalamocortical relay neurons.
      Thus, a Ca2+ level increase ultimately decreases KCNK3- and KCNK9-mediated K+ conductance. Conversely, compensatory K+ efflux provides an electrochemical gradient and hence is a prerequisite for Ca2+ influx.
      • Cahalan M.D.
      • Chandy K.G.
      The functional network of ion channels in T lymphocytes.
      Notably, previous work indicated that KCNK3 and KCNK9 share most of their structure and properties, and even form heterodimers. Thus, it repeatedly was assumed that the channels can be substituted for one another.
      • Czirjak G.
      • Enyedi P.
      Formation of functional heterodimers between the TASK-1 and TASK-3 two-pore domain potassium channel subunits.
      Increased intracellular Ca2+ levels are associated with a plethora of signaling responses in intestinal epithelial cells and many of these signals are involved in the pathophysiology of IBD, including the formation and stability of tight junctions and epithelial cell survival.
      • Samak G.
      • Chaudhry K.K.
      • Gangwar R.
      • Narayanan D.
      • Jaggar J.H.
      • Rao R.
      Calcium/Ask1/MKK7/JNK2/c-Src signalling cascade mediates disruption of intestinal epithelial tight junctions by dextran sulfate sodium.
      Dysfunction or even loss of epithelial barrier integrity is a hallmark of IBD,
      • Chang J.T.
      Pathophysiology of inflammatory bowel diseases.
      and drives chronic disease in addition to prerequisites such as increased activation of immune cells or changes in the intestinal microbiome.
      • Torres J.
      • Mehandru S.
      • Colombel J.F.
      • Peyrin-Biroulet L.
      Crohn's disease.
      IBD is characterized by chronic remittent flares of colitis that can present with abdominal pain, diarrhea, and rectal bleeding.
      • Nowacki T.M.
      • Bruckner M.
      • Eveslage M.
      • Tepasse P.
      • Pott F.
      • Thoennissen N.H.
      • Hengst K.
      • Ross M.
      • Bettenworth D.
      The risk of colorectal cancer in patients with ulcerative colitis.
      However, K2P channels do not just share potential interaction points with IBD pathophysiology, but also represent potentially druggable targets. Given its potential link to IBD at a genomic variant level, we decided to investigate the effects of KCNK9 modification in an animal model of IBD and to further elucidate the cellular effects of KCNK9 abrogation on intestinal epithelial calcium homeostasis and cell survival.

      Results

      KCNK9-Deficient Mice Show Exacerbated Morbidity in Acute Dextran Sulfate Sodium–Induced Colitis

      To test whether KCNK9 affects intestinal inflammation, we exposed knockout mice and their respective wild-type controls to 1 course of dextran sulfate sodium (DSS) treatment followed by normal drinking water. Kcnk9-/- mice suffered from increased weight loss (Figure 1A) and showed more severe clinical symptoms, including diarrhea and rectal bleeding, as depicted by the disease activity index (Figure 1B). We examined the intestinal barrier integrity in vivo using fluorescein isothiocyanate–dextran and identified significantly increased barrier disruption in Kcnk9-/- mice (Figure 1E). Accordingly, both inflammatory colon shortening (Figure 1D and F) and histology scores (Figure 1C and G) indicated more severe disease. In line, serum inflammatory cytokine level controls including interleukin (IL)1β, tumor necrosis factor α (TNF-α), and IL6 were found increased in Kcnk9-/- mice compared with controls (Figure 1H). Notably, we were able to document an increase of caspase-9, but not caspase-8, activation in the intestinal epithelium of Kcnk9-/- mice compared with controls (Figure 1I)
      Figure thumbnail gr1
      Figure 1Kcnk9-/- mice show increased susceptibility in DSS-induced colitis. (A and B) Development of body weight (shown as percentage of baseline) and disease activity index (DAI) during acute DSS-induced colitis. DSS was administered from days 1 to 5. (C) Generic images of H&E-stained colon sections from DSS-treated animals and controls. Scale bars: 300 μm. (D) Generic images of explanted colon from DSS-treated wild-type and knockout animals. (E) Fluorescein isothiocyanate (FITC)-dextran levels in peripheral blood on day 9. Mice received FITC-dextran 4 hours before sample collection. (F) Colon length from anus to ileocecal valve measured from ileocecal valve to anus after explantation. (G) Histology score of H&E-stained transverse colon sections. (H) Levels of inflammatory cytokines in peripheral blood on day 9. (I) Assessment of caspase-8 and caspase-9 activity in colon whole-tissue pieces on day 9. Data are shown as (A and B) means ± SD or as (E–I) box–whisker plots (boxes indicate median ± interquartile range, whiskers indicate range). (A and B) n = 10 for DSS-treated and n = 5 for H2O-treated animals, (E–I) n = 5 per group. Statistical significance was tested using 2-way analysis of variance including the (A and B) Bonferroni post hoc test or the (E–I) Mann–Whitney rank-sum test. ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001. KO, knockout; WT, wild-type.

      The Phenotype of KCNK9-Deficient Mice Can Be Reproduced by Pharmacologic Blockade

      To validate the detrimental effects caused by genetic knockout in Kcnk9-/- mice, we specifically blocked KCNK9 using the recently described substance 1-[1-[6-([1,1′-biphenyl]-4-ylcarbonyl)-5,6,7,8-tetrahydropyrido[4,3-d]pyrimidin-4-yl]-4-piperidinyl]-1-butanone (PK-THPP) in wild-type and knockout animals and again acutely induced colitis with DSS treatment.
      • Zeng C.
      • Long X.
      • Cotten J.F.
      • Forman S.A.
      • Solt K.
      • Faingold C.L.
      • Feng H.J.
      Fluoxetine prevents respiratory arrest without enhancing ventilation in DBA/1 mice.
      ,
      • Srisomboon Y.
      • Zaidman N.A.
      • Maniak P.J.
      • Deachapunya C.
      • O'Grady S.M.
      P2Y receptor regulation of K2P channels that facilitate K(+) secretion by human mammary epithelial cells.
      Pharmacologic inhibition of KCNK9 severely affected wild-type animals as indicated by weight loss (Figure 2A) and disease activity index (Figure 2B), but did not further modify the disease course of knockout animals. Barrier disruption also was more substantial in PK-THPP–treated wild-type animals (Figure 2E). Inflammatory colon shortening (Figure 2D and F) and histology scores (Figure 2C and G) again were increased. Similarly, serum cytokine levels for IL1β, TNF-α, and IL6 were increased after PK-THPP treatment in wild-type mice (Figure 2H). In Kcnk9-/- mice, we found no additional effect of PK-THPP treatment compared with injection of the vehicle substance alone. Again, we found increased activation of caspase-9 in the intestinal epithelium of Kcnk9-/- mice, whereas caspase-8 activation remained equal to controls (Figure 2I).
      Figure thumbnail gr2
      Figure 2Pharmacologic blockade of KCNK9 aggravates DSS-induced colitis in wild-type mice. (A and B) Development of body weight (shown as percentage of baseline) and disease activity index (DAI) during acute DSS-induced colitis. DSS was administered from days 1 to 5. (C) Generic images of H&E-stained colon sections from DSS-treated animals and controls. Scale bars: 300 μm. (D) Generic images of explanted colon from DSS-treated wild-type and knockout animals. (E) Fluorescein isothiocyanate (FITC)-dextran levels in peripheral blood on day 9. Mice received FITC-dextran 4 hours before sample collection. (F) Colon length from the anus to the ileocecal valve measured from the ileocecal valve to the anus after explantation. (G) Histology score of the H&E-stained transverse colon sections. (H) Levels of inflammatory cytokines in peripheral blood on day 9. (I) Assessment of caspase-8 and caspase-9 activity in colon whole-tissue pieces on day 9. Data are shown as the (A and B) means ± SD or as (E–I) box-whisker plots (boxes indicate median ± interquartile range, whiskers indicate range). (A and B) n = 10 for DSS-treated and n = 5 for H2O-treated animals, (E–I) n = 5 per group. Statistical significance was tested using 2-way analysis of variance including the (A and B) Bonferroni post hoc test or the (E–I) Mann–Whitney rank sum test. ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001. veh, vehicle; WT, wild-type.

      KCNK9 Deficiency Induces Mitochondrial Dysfunction and Increases Apoptosis in Intestinal Epithelial Cells

      Next, we generated primary culture of murine colonic intestinal epithelial cells (IECs) from KCNK9-deficient mice and wild-type controls. After 24 hours of cultivation, only a minor fraction of Kcnk9-/- IECs underwent apoptosis as indicated by positive caspase-3 staining, and no difference was observed between knockout and wild-type cells (Figure 3A). However, Kcnk9-/- IECs were prone to cell death after treatment with DSS (TNF-α served as positive control) (Figure 3B) and Kcnk9-/- IECs showed pronounced activation of caspase-3 (Figure 3C). Because it is known that DSS and TNF-α induce apoptosis via different pathways, which are dependent on either initiator caspase-8 or caspase-9,
      • Samak G.
      • Chaudhry K.K.
      • Gangwar R.
      • Narayanan D.
      • Jaggar J.H.
      • Rao R.
      Calcium/Ask1/MKK7/JNK2/c-Src signalling cascade mediates disruption of intestinal epithelial tight junctions by dextran sulfate sodium.
      ,
      • Ozoren N.
      • Kim K.
      • Burns T.F.
      • Dicker D.T.
      • Moscioni A.D.
      • El-Deiry W.S.
      The caspase 9 inhibitor Z-LEHD-FMK protects human liver cells while permitting death of cancer cells exposed to tumor necrosis factor-related apoptosis-inducing ligand.
      we repeated the caspase-3 activation assay after adding DSS and respective inhibitors of the initiator caspases. After adding the fluoromethyl ketone-derived caspase-8 inhibitor, a significant difference still was observed in apoptosis induction in Kcnk9-/- IECs after DSS exposure. Conversely, blockade of caspase-9 after adding fluoromethyl ketone-derived caspase-9 inhibitor resulted in a marked reduction in apoptosis and lowering of caspase-3 activation levels in knockout and wild-type IECs (Figure 3D).
      Figure thumbnail gr3
      Figure 3KCNK9 modulates cellular calcium responses and cell survival in IECs. (A) Immunofluorescence staining of isolated IECs showing the fraction of apoptotic cells by determination of caspase-3 activation. (n = 5 per group; legend also applies to BD, G, I–K). (B) Evaluation of cell viability after TNF-α and DSS exposure using the MTS assay (n = 5 per group). (C) Analysis of caspase-3 activation (n = 5 per group). Caspase-3 activation after DSS treatment with the addition of either DMSO (vehicle), fluoromethyl ketone-derived caspase-8 inhibitor (IETD), fluoromethyl ketone-derived caspase-9 inhibitor (LEHD) in IECs (n = 5 per group). (E and F) Fura-2–based assay for detecting cytosolic calcium levels in IECs after DSS exposure. Arrows indicate addition of 5% DSS solution (n = 8 per group). (G) Caspase-3 activation in SM and CFM after DSS exposure (n = 5 per group). (H) Fura-2–based assay for detecting cytosolic calcium levels in IECs after DSS exposure after ER calcium depletion using thapsigargin (n = 8 per group). (I) Assessment of ER stress marker gene Grp78 and Chop expression in IECs after DSS treatment vs controls (n = 5 per group). (J) Quantification of mitochondrial membrane integrity (JC-1), superoxide generation (MitoSOX), and formation of hydroxyl radical via the Fenton reaction (2′,7′-dichlorofluorescin diacetate [DCFDA]) after DSS stimulation (n = 5 per group). (K) Western blot of KCNK9 and respective cell fraction markers (numbers on the right indicate mass in kilodaltons). Statistical significance was tested using the (A–C, G, I, and J) Mann–Whitney U test, (D) Kruskal–Wallis test including the Dunn post hoc test, or (E, F, and H) 2-way analysis of variance including the Bonferroni post hoc test. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, P > .05 was not significant. AU, adsorption unit; CFM, calcium-free medium; Ctrl, control; CY, cytosolic fraction; HPF, high-power field; MITO, mitochondrial fraction; PM, plasma membrane fraction; rRNA, ribosomal RNA; SM, standard medium; RFU, relative fluorescent unit; TPG, thapsigargin; WT, wild-type.
      Because caspase-9 activation is usually the result of organelle damage and the subsequent increase in cytosolic calcium levels, we investigated whether KCNK9 deficiency alters calcium level development over time after DSS exposure of IECs.
      We found that Kcnk9-/- IECs show a significant increase in late-stage cytosolic calcium levels compared with wild-type controls and that increased levels were maintained (Figure 3E). Comparing Kcnk9-/- IECs kept in standard medium with those kept in calcium-free medium, we observed differences in the early phase (minutes 5–10) after stimulation with DSS, but no differences in late-stage calcium levels (minutes 10–15) were seen (Figure 3F). We concluded that the early increase in cytosolic Ca2+ levels results from the influx of extracellular Ca2+, whereas further increases are caused by Ca2+ release from internal storage. Notably, differences in caspase-3 activation remained significant between knockout and wild-type cells even in Ca2+-free medium, suggesting that KCNK9 might predominantly affect the release of Ca2+ from internal stores (Figure 3G).
      Different intracellular calcium storing organelles exist, with the endoplasmic reticulum (ER) often being deemed the most important.
      • Sammels E.
      • Parys J.B.
      • Missiaen L.
      • De Smedt H.
      • Bultynck G.
      Intracellular Ca2+ storage in health and disease: a dynamic equilibrium.
      ER damage and subsequent Ca2+ release are an established consequence of ER stress, which can be induced by various mechanisms such as accumulation of unfolded or misfolded proteins. Furthermore, ER stress is well described in IBD and its animal models including DSS-induced colitis.
      • Das I.
      • Png C.W.
      • Oancea I.
      • Hasnain S.Z.
      • Lourie R.
      • Proctor M.
      • Eri R.D.
      • Sheng Y.
      • Crane D.I.
      • Florin T.H.
      • McGuckin M.A.
      Glucocorticoids alleviate intestinal ER stress by enhancing protein folding and degradation of misfolded proteins.
      Thus, we investigated whether differences in cytosolic calcium levels originate from increased release of ER-stored calcium. We depleted ER calcium in knockout and wild-type IECs by thapsigargin-induced blockade of the ER calcium adenosine triphosphatase (SERCA) and measured cytosolic calcium levels after DSS exposure. Thapsigargin treatment resulted in comparable lowering of peak cytosolic calcium levels in both knockout and wild-type cells (reduction of the peak Fura-2 ratio in knockout cells by 0.167 adsorption units (7.46%) vs 0.245 adsorption units (13.8%) (Figure 3H). However, differences between knockout and wild-type animals persisted and therefore we concluded that increased Ca2+ levels in Kcnk9-/- originate from organelles other than the ER. Furthermore, we observed no differences in expression of ER stress marker genes after DSS treatment (Figure 3F)
      Mitochondria also contain substantial calcium stores and the presence of KCNK9 on mitochondria has been reported previously.
      • Nagy D.
      • Gonczi M.
      • Dienes B.
      • Szoor A.
      • Fodor J.
      • Nagy Z.
      • Toth A.
      • Fodor T.
      • Bai P.
      • Szucs G.
      • Rusznak Z.
      • Csernoch L.
      Silencing the KCNK9 potassium channel (TASK-3) gene disturbs mitochondrial function, causes mitochondrial depolarization, and induces apoptosis of human melanoma cells.
      We explored whether these organelles might be less resistant to DSS exposure in the absence of KCNK9 and used fluorescent dyes to visualize mitochondrial response toward DSS. KCNK9 deficiency boosted both mitochondrial membrane depolarization (measured via 5,5',6,6'-tetrachloro-1,1',3,3'-tetraethylbenzimidazolocarbocyanine iodide [JC-1] assay) and mitochondrial superoxide production (measured via MitoSOX oxidation). Consequently, hydrogen peroxide is increased as measured by an increased Fenton reaction via 2′,7′-dichlorofluorescin diacetate oxidation (all Figure 3J). Because we hypothesized that KCNK9 might regulate mitochondrial integrity, we performed Western blot of isolated cell fractions and found that KCNK9, in addition its presence on the plasma membrane, also was present in the mitochondrial fraction (Figure 3K).

      IECs React to KCNK9 Deficiency via Up-Regulation of KCNK3

      As described earlier, KCNK9 shares most of its properties with its closest relative KCNK3, aside from the formation of heterodimers.
      • Wilke B.U.
      • Lindner M.
      • Greifenberg L.
      • Albus A.
      • Kronimus Y.
      • Bunemann M.
      • Leitner M.G.
      • Oliver D.
      Diacylglycerol mediates regulation of TASK potassium channels by Gq-coupled receptors.
      ,
      • Chokshi R.H.
      • Larsen A.T.
      • Bhayana B.
      • Cotten J.F.
      Breathing stimulant compounds inhibit TASK-3 potassium channel function likely by binding at a common site in the channel pore.
      To test the relevance of this resemblance in IECs, we screened Kcnk9-/- and wild-type control cells for differential expression of the known 2-pore potassium channels. KCNK9 deficiency resulted in a strong increase of Kcnk3 gene expression, whereas expression levels of other K2P channels remained largely unaltered (Figure 4A). Expression of KCNK3 and KCNK9 were visualized at the protein level via immunofluorescence staining (Figure 4B).
      Figure thumbnail gr4
      Figure 4Kcnk9-/- IECs show up-regulation of Kcnk3 mRNA and blockade of KCNK3 normalizes calcium responses. (A) Analysis for gene expression of different K2P channels in untreated Kcnk9-/- and wild-type IECs using quantitative real-time PCR (n = 3 per group). (B) Immunofluorescence analysis for validating channel expression on the protein level in IECs. Columns indicate genotype, rows indicate performed staining. Nuclear staining was performed with 4',6-diamidino-2-phenylindole. Scale bars: 10 μm. (C) Caspase-3 activation after DSS stimulation and addition of KCNK3 and caspase-9 inhibitor A293 (n = 5 per group). (D) Calcium assay after A293 addition (n = 8 per group). (E) Calcium assay after PK-THPP and/or A293 addition and DSS stimulation (n = 8 per group). (F) Calcium assay in Kcnk3-/- IECs (n = 8 per group). (G) mRNA expression of Kcnk3 and Kcnk9 in Kcnk3-/- IECs (n = 5 per group). (H) Evaluation of mitochondrial superoxide generation (MitoSOX) and mitochondrial membrane integrity (JC-1) after DSS stimulation (n = 5 per group). Statistical significance was tested using the (A and G) Mann–Whitney U test, (C and H) Kruskal–Wallis test including Dunn post hoc test, or (D–F) 2-way analysis of variance including the Bonferroni post hoc test. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001; P > .05 was not significant. AU, adsorption unit; Ctrl, control; RFU, relative fluorescent unit; rRNA, ribosomal RNA; WT, wild-type.
      To further elucidate the role of a potential interplay between KCNK3 and KCNK9, we used the experimental inhibitor substance 2-(butane-1-sulfonyl-amino)-N-[1-(R)-(6-methoxypyridin-3-yl)-propyl]-benz-amide (A293), which preferably inhibits KCNK3 (and to a very minimal extent KCNK9) channel function.
      • Bittner S.
      • Bauer M.A.
      • Ehling P.
      • Bobak N.
      • Breuer J.
      • Herrmann A.M.
      • Golfels M.
      • Wiendl H.
      • Budde T.
      • Meuth S.G.
      The TASK1 channel inhibitor A293 shows efficacy in a mouse model of multiple sclerosis.
      Its administration led to no relevant decrease of apoptosis in wild-type IECs after DSS exposure. However, its effect was potentiated in Kcnk9-/- IECs, corroborating the high Kcnk3 gene expression (Figure 4C). Remarkably, we observed a decrease in apoptosis rate even below values found in A293-treated wild-type IECs, and we presume that in the absence of KCNK3 and KCNK9 channel function, Ca2+ influx is largely suppressed in general. In line with this, we detected a decrease of late-stage cytosolic calcium levels in DSS-exposed Kcnk9-/- IECs after incubation with A293 (Figure 4D).
      We then sought to establish whether KCNK3 exerts a specific role in modulation of Ca2+ influx or whether the observed phenomena are owing to transcriptional effects. We silenced channel functions by pharmacologic inhibition in wild-type IECs. First, we specifically blocked KCNK9 by adding PK-THPP before DSS exposure, and 5 minutes after DSS administration we added A293. We found that PK-THPP treatment closely resembled the phenotype observed in Kcnk9-/- IECs, whereas A293 treatment had no demonstratable effect on late-stage Ca2+ influx alone but led to a potent decrease of cytosolic Ca2+ levels when administered subsequently to PK-THPP (Figure 4E).
      Subsequently, we decided to verify our results in a genetic knockout of KCNK3. Examining the DSS-induced increase of cytosolic calcium levels confirmed our previous finding and showed a significant decrease of late-stage calcium influx in Kcnk3-/- IECs (Figure 4F). In those cells, we found a reciprocal increase in Kcnk9 messenger RNA (mRNA) expression, whereas the absence of Kcnk3 mRNA was shown by real-time polymerase chain reaction (PCR) (Figure 4G). We evaluated mitochondrial function via the previously performed assays and detected reduced mitochondrial superoxide production as well as greater persistence of mitochondrial membrane potential (both Figure 4H). However, these effects appear to be caused by increased KCNK9 channel expression because elimination of this aspect by pharmacologic blockade with A293 again yielded no respective phenotype in wild-type IECs.

      KCNK3-Deficient Mice Are Protected From Acute DSS-Induced Colitis

      Lastly, we subjected Kcnk3-/- mice and their respective wild-type controls to DSS-induced colitis. We found that KCNK3 deficiency reduced weight loss and disease activity index compared with wild-type controls (Figure 5A and B). Barrier disruption was reduced, although effects were not statistically significant here (Figure 5E). However, colonic shortening (Figure 5D and F) as well as histologic damage were markedly less pronounced (Figure 5C and G). We also observed lower levels of inflammatory cytokines in peripheral blood samples obtained from our knockout mice compared with control animals (Figure 5I). Lastly, we performed expression analysis of Kcnk9 mRNA from colon samples and confirmed increased expression in Kcnk3-/-, thereby confirming our previous results from IEC studies. Notably, here we found increased expression of KCNK9 in DSS-treated wild-type animals compared with water controls (P = .0317) (Figure 5H). Similar to DSS-induced colitis experiments in Kcnk9-/- mice, we performed activation assays regarding caspase-8 and caspase-9 and found unaltered levels of caspase-8 activation whereas caspase-9 activation was reduced in Kcnk3-/- mice compared with controls (Figure 5J).
      Figure thumbnail gr5
      Figure 5Kcnk3-/- mice show ameliorated courses compared with controls in DSS-induced colitis. (A and B) Development of body weight (shown as percentage of baseline) and disease activity index (DAI) during acute DSS-induced colitis. DSS was administered from days 1 to 5. (C) Generic images of H&E-stained colon sections from DSS-treated animals and controls. Scale bars: 300 μm. (D) Generic images of explanted colon from DSS-treated wild-type and knockout animals. (E) Fluorescein isothiocyanate (FITC)-dextran levels in peripheral blood on day 9. Mice received FITC-dextran 4 hours before sample collection. (F) Colon length from the anus to the ileocecal valve measured from the ileocecal valve to the anus after explantation. (G) Histology score of H&E-stained transverse colon sections. (H) Quantitative assessment of Kcnk9 mRNA expression in colon whole-tissue pieces on day 9. (I) Levels of inflammatory cytokines in peripheral blood on day 9. (J) Assessment of caspase-8 and caspase-9 activity in colon whole-tissue pieces on day 9. Data are shown as (A and B) means ± SD or as (E–J) box-whisker plots (boxes indicate median ± interquartile range, whiskers indicate range). (A and B) n = 10 for DSS-treated and n = 5 for H2O-treated animals, and (E–J) n = 5 per group. Significance was calculated using 2-way analysis of variance including the (A and B) Bonferroni post hoc test or the (E–J) Mann–Whitney U test. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001. FITC, fluorescein isothiocyanate; KO, knockout; rRNA, ribosomal RNA; WT, wild-type.

      Discussion

      Within the presented work, we aimed to establish a deeper understanding of the potential role of KCNK9 in epithelial cell survival and its in vivo impact on DSS-induced colitis. We found that, as well as the genetic knockout, pharmacologic blockade of KCNK9 aggravated DSS-induced colitis including more pronounced weight loss, more severe clinical symptoms, and impairment of intestinal epithelial barrier function with subsequent release of inflammatory cytokines. After establishment of primary colonic IEC cultures from knockout and wild-type animals, we found increased susceptibility of Kcnk9-/- IECs to apoptosis after DSS exposure and, similar to in vivo data, we showed a major role for caspase-9 activation and thus the intrinsic apoptosis cascade. Intracellular calcium levels increased more strongly after DSS treatment in the absence or blockade of KCNK9 and we were able to identify mitochondrial damage as the underlying cause for the excessive increase of cytosolic Ca2+ levels, whereas we found no specific predisposition to ER stress. Although closely related to KCNK9 regarding structure and function, and up-regulated in Kcnk9-/- IECs, KCNK3 could not compensate for the loss of KCNK9 function. Conversely, the beneficial effects of KCNK3 deficiency regarding DSS-induced Ca2+ influx, mitochondrial damage, and apoptosis were accompanied by stronger expression of KCNK9. Similar observations also were made in Kcnk3-/- mice that underwent DSS-induced colitis.
      It generally is accepted that DSS mediates its effects on the intestinal epithelium through a number of mechanisms, including the development of ER stress, Ca2+-dependent breakdown of tight junction proteins, and, ultimately, apoptosis, and DSS therefore at least partially resembles the pathophysiology of ulcerative colitis.
      • Samak G.
      • Chaudhry K.K.
      • Gangwar R.
      • Narayanan D.
      • Jaggar J.H.
      • Rao R.
      Calcium/Ask1/MKK7/JNK2/c-Src signalling cascade mediates disruption of intestinal epithelial tight junctions by dextran sulfate sodium.
      ,
      • Wan Y.
      • Yang L.
      • Jiang S.
      • Qian D.
      • Duan J.
      Excessive apoptosis in ulcerative colitis: crosstalk between apoptosis, ROS, ER stress, and intestinal homeostasis.
      Analysis of cytosolic Ca2+ responses toward DSS exposure showed a 2-step increase of Ca2+. The first step was significantly reduced in Ca2+-free medium and was preserved in thapsigargin-treated IECs and thus most likely represents influx of Ca2+ from the extracellular space. Although KCNK9 was described to modulate this Ca2+ via provision of a compensatory potassium efflux in various cell types,
      • Meuth S.G.
      • Bittner S.
      • Meuth P.
      • Simon O.J.
      • Budde T.
      • Wiendl H.
      TWIK-related acid-sensitive K+ channel 1 (TASK1) and TASK3 critically influence T lymphocyte effector functions.
      ,
      • Nagy D.
      • Gonczi M.
      • Dienes B.
      • Szoor A.
      • Fodor J.
      • Nagy Z.
      • Toth A.
      • Fodor T.
      • Bai P.
      • Szucs G.
      • Rusznak Z.
      • Csernoch L.
      Silencing the KCNK9 potassium channel (TASK-3) gene disturbs mitochondrial function, causes mitochondrial depolarization, and induces apoptosis of human melanoma cells.
      ,
      • Albrecht S.
      • Korr S.
      • Nowack L.
      • Narayanan V.
      • Starost L.
      • Stortz F.
      • Arauzo-Bravo M.J.
      • Meuth S.G.
      • Kuhlmann T.
      • Hundehege P.
      The K2P-channel TASK1 affects Oligodendroglial differentiation but not myelin restoration.
      • Meuth S.G.
      • Herrmann A.M.
      • Ip C.W.
      • Kanyshkova T.
      • Bittner S.
      • Weishaupt A.
      • Budde T.
      • Wiendl H.
      The two-pore domain potassium channel TASK3 functionally impacts glioma cell death.
      • Nakakura S.
      • Matsui M.
      • Sato A.
      • Ishii M.
      • Endo K.
      • Muragishi S.
      • Murase M.
      • Kito H.
      • Niguma H.
      • Kurokawa N.
      • Fujii M.
      • Araki M.
      • Araki K.
      • Ohya S.
      Pathophysiological significance of the two-pore domain K(+) channel K2P5.1 in splenic CD4(+)CD25(-) T cell subset from a chemically-induced murine inflammatory bowel disease model.
      we found no difference in Kcnk9-/- IECs compared with controls. However, the second step of cytosolic Ca2+ increase, which was reduced (but not absent) after thapsigargin treatment and resistant to Ca2+-free medium and thus represents release of Ca2+ from internal storage, was strongly increased in KCNK9 deficiency. Although the majority of Ca2+ released within calcium-induced calcium release comes from the ER,
      • Feske S.
      CRAC channels and disease - from human CRAC channelopathies and animal models to novel drugs.
      ,
      • Concepcion A.R.
      • Feske S.
      Regulation of epithelial ion transport in exocrine glands by store-operated Ca(2+) entry.
      we found no restitution of Kcnk9-/- IECs by thapsigargin. Furthermore, ER stress was rendered as an unlikely cause of the observed phenotype because induction of ER stress upon thapsigargin treatment did not copy the differences between wild-type and KCKN9-deficient mice in terms of cytosolic calcium levels and we observed no increases in ER stress-response genes in knockout IECs compared with wild-type controls.
      However, we were able to show rapid impairment of mitochondrial integrity after DSS exposure in Kcnk9-/- IECs, resulting in permeabilization of the outer membrane and loss of the mitochondrial membrane potential depicted in our JC-1 staining.
      Mitochondrial stress is induced by DSS via different pathways. First, reactive oxygen species are known to disturb the redox equilibrium of mitochondria in DSS-induced colitis.
      • Guo C.
      • Sun L.
      • Chen X.
      • Zhang D.
      Oxidative stress, mitochondrial damage and neurodegenerative diseases.
      Second, regeneration of mitochondria by fission and fusion is impaired after DSS exposure.
      • Mancini N.L.
      • Goudie L.
      • Xu W.
      • Sabouny R.
      • Rajeev S.
      • Wang A.
      • Esquerre N.
      • Al Rajabi A.
      • Jayme T.S.
      • van Tilburg Bernandes E.
      • Nasser Y.
      • Ferraz J.G.P.
      • Shutt T.
      • Shearer J.
      • McKay D.M.
      Perturbed mitochondrial dynamics is a novel feature of colitis that can be targeted to lessen disease.
      Finally, ER stress itself impairs mitochondrial function by interference with Ca
      • Bittner S.
      • Budde T.
      • Wiendl H.
      • Meuth S.G.
      From the background to the spotlight: TASK channels in pathological conditions.
      transport between ER and mitochondria.
      • Malhotra J.D.
      • Kaufman R.J.
      ER stress and its functional link to mitochondria: role in cell survival and death.
      Interestingly, our data also showed that KCNK9 was not restricted to the plasma membrane but also was found on the mitochondrial membrane. Several potassium channels localized on mitochondria have been shown to ameliorate mitochondrial damage and subsequent cell death by limitation of Ca2+ currents.
      • Wrzosek A.
      • Augustynek B.
      • Zochowska M.
      • Szewczyk A.
      Mitochondrial potassium channels as druggable targets.
      ,
      • Bachmann M.
      • Li W.
      • Edwards M.J.
      • Ahmad S.A.
      • Patel S.
      • Szabo I.
      • Gulbins E.
      Voltage-gated potassium channels as regulators of cell death.
      Interestingly, DSS itself induces the MAPK/ERK signaling cascade und thereby enhances mitochondrial Ca2+ release via increased expression of transport proteins ORAI1 and STIM.
      • Johnstone L.S.
      • Graham S.J.
      • Dziadek M.A.
      STIM proteins: integrators of signalling pathways in development, differentiation and disease.
      Notably, DSS also induced increased expression of KCNK9 in wild-type IECs, thus presenting a potential mechanism for fine-tuning mitochondrial Ca2+ efflux.
      It currently only can be speculated how KCNK9 is able to limit mitochondrial toxicity, but recent data have indicated that synthesis of diacyl glycerol after Ca
      • Bittner S.
      • Budde T.
      • Wiendl H.
      • Meuth S.G.
      From the background to the spotlight: TASK channels in pathological conditions.
      + and activation of Gq-proteins inhibits KCNK9 function and thus reduces the rectifying potassium current required for further release of Ca2+.
      • Wilke B.U.
      • Lindner M.
      • Greifenberg L.
      • Albus A.
      • Kronimus Y.
      • Bunemann M.
      • Leitner M.G.
      • Oliver D.
      Diacylglycerol mediates regulation of TASK potassium channels by Gq-coupled receptors.
      Furthermore, KCNK9 is acid-sensitive and closes secondary to reduced pH, which regularly occurs during mitochondrial stress, further reducing conductance.
      • Bittner S.
      • Budde T.
      • Wiendl H.
      • Meuth S.G.
      From the background to the spotlight: TASK channels in pathological conditions.
      In line, previous work has shown impairment of mitochondrial function in the absence of KCNK9 in keratinocytes.
      • Rusznak Z.
      • Bakondi G.
      • Kosztka L.
      • Pocsai K.
      • Dienes B.
      • Fodor J.
      • Telek A.
      • Gonczi M.
      • Szucs G.
      • Csernoch L.
      Mitochondrial expression of the two-pore domain TASK-3 channels in malignantly transformed and non-malignant human cells.
      Mitochondrial permeabilization results in the release of cytochrome c as well as further pro-apoptotic factors and thus is capable of activating caspase-9 and inducing apoptosis, as previously observed.
      • Kalkavan H.
      • Green D.R.
      MOMP, cell suicide as a BCL-2 family business.
      Notably, TNF-α also is capable of inducing mitochondrial disintegration
      • Huai J.
      • Vogtle F.N.
      • Jockel L.
      • Li Y.
      • Kiefer T.
      • Ricci J.E.
      • Borner C.
      TNFalpha-induced lysosomal membrane permeability is downstream of MOMP and triggered by caspase-mediated NDUFS1 cleavage and ROS formation.
      and this explains the increased susceptibility of Kcnk9-/- IECs to TNF-α exposure in our experiments. We assume that these mechanisms are pivotal to the loss of intestinal barrier function and more severe courses of DSS-induced colitis.
      Nonetheless, we were largely surprised to see that KCNK3 was incapable of compensating for KCNK9 deficiency despite its well-known mechanism of compensatory regulation. Protein sequences of both channels are largely similar and previous work has indicated similar properties throughout.
      • Bittner S.
      • Budde T.
      • Wiendl H.
      • Meuth S.G.
      From the background to the spotlight: TASK channels in pathological conditions.
      ,
      • Cotten J.F.
      TASK-1 (KCNK3) and TASK-3 (KCNK9) tandem pore potassium channel antagonists stimulate breathing in isoflurane-anesthetized rats.
      Even more, channels can form functional heterodimers as well.
      • Czirjak G.
      • Enyedi P.
      Formation of functional heterodimers between the TASK-1 and TASK-3 two-pore domain potassium channel subunits.
      Because we observed a reduction of cytosolic Ca2+ increase, mitochondrial damage in Kcnk3-/- IECs, as well as ameliorated DSS-induced colitis in Kcnk3-/- mice, one could assume that both channels mediate opposite functions. However, pharmacologic blockade using the KCNK3 inhibitor A293 restored phenotypes of Kcnk9-/- IECs but did not alleviate findings in wild-type cells. This contrasts blockade of KCNK9 by its specific inhibitor, which copies the disease phenotype of Kcnk9-/- IECs. What indeed was found in vitro as well as in vivo was that KCNK3 deficiency increases Kcnk9 expression, which was not entirely unexpected given the assumption of a common, yet still unknown, promotor.
      • Wilke B.U.
      • Lindner M.
      • Greifenberg L.
      • Albus A.
      • Kronimus Y.
      • Bunemann M.
      • Leitner M.G.
      • Oliver D.
      Diacylglycerol mediates regulation of TASK potassium channels by Gq-coupled receptors.
      ,
      • Cotten J.F.
      TASK-1 (KCNK3) and TASK-3 (KCNK9) tandem pore potassium channel antagonists stimulate breathing in isoflurane-anesthetized rats.
      We thus assume that increased expression of KCNK9 as observed in vitro and in vivo within DSS-induced colitis modulates the phenotype rather than absence of KCNK3.
      It is not clear how A293 selectively restores the phenotype of the further reduction of Ca2+ influx by A293 in both Kcnk9-/- IECs and PK-THPP–exposed wild-type IECs. However, it is likely that a lack of KCNK3/9 potassium conductance largely alters cellular homeostasis beyond mitochondrial function. Unfortunately, it is not possible to further investigate this given the absence of KCNK3-/-/KCNK9-/- double-knockout mice.
      In conclusion, we have shown that KCNK9 positively modulates intestinal epithelial cell survival in the context of intestinal inflammation in vitro and in vivo. Future studies should target strategies to induce overexpression or enhance channel function to boost IEC survival in the context of inflammatory conditions.

      Methods

      Animals

      The development of the Kcnk3-/- and Kcnk9-/- mouse models has been described previously.
      • Bittner S.
      • Bauer M.A.
      • Ehling P.
      • Bobak N.
      • Breuer J.
      • Herrmann A.M.
      • Golfels M.
      • Wiendl H.
      • Budde T.
      • Meuth S.G.
      The TASK1 channel inhibitor A293 shows efficacy in a mouse model of multiple sclerosis.
      Animals were raised on a C57Bl/6J background and crossed-back at least 10 times before use. For experiments, we used knockout animals and their respective wild-type littermates (abbreviated as wild-type throughout text). Animals were kept in specific pathogen-free conditions and fed a standard rodent diet. Local authorities approved all animal experiments (84-02.04.2012.A441; North Rhine-Westphalian Ministry for Environment, Health and Consumer Protection, Department for Animal Protection).

      Experimental Model of Colitis and Barrier Permeability Assay

      Mice (age, 8–10 weeks) with their respective genetic background were exposed to a 3% aqueous solution of DSS (36–50 kilodaltons; MP Biomedicals, Eschwege, Germany) ad libitum via drinking water for 5 consecutive days. Given the reduced breeding efficiency of K2P-deficient mice (a well-known phenomenon), strict local regulations for animal breeding and the need for female animals elsewhere (including strain maintenance), and the known variation of disease severity during the ovarian cycle,
      • Babickova J.
      • Tothova L.
      • Lengyelova E.
      • Bartonova A.
      • Hodosy J.
      • Gardlik R.
      • Celec P.
      Sex differences in experimentally induced colitis in mice: a role for estrogens.
      male animals were used here.
      Furthermore, drinking solution was prepared fresh every 24 hours and consumption per cage was documented for estimation of DSS uptake per mouse (no significant differences were noted here). On days 6–9, mice received normal drinking water. One experimental group received PK-THPP, a potent antagonist of KCNK9,
      • Zeng C.
      • Long X.
      • Cotten J.F.
      • Forman S.A.
      • Solt K.
      • Faingold C.L.
      • Feng H.J.
      Fluoxetine prevents respiratory arrest without enhancing ventilation in DBA/1 mice.
      ,
      • Srisomboon Y.
      • Zaidman N.A.
      • Maniak P.J.
      • Deachapunya C.
      • O'Grady S.M.
      P2Y receptor regulation of K2P channels that facilitate K(+) secretion by human mammary epithelial cells.
      which was administered intraperitoneally once daily throughout the experiment (10 μg/g body weight in 200 μL 1% dimethyl sulfoxide [DMSO]/40% polyethylene glycol 400/phosphate-buffered saline [PBS]) (DMSO; polyethylene glycol 400), Sigma Aldrich, St. Louis, MO). Control animals received 200 μL 1% DMSO/40% polyethylene glycol 400/PBS intraperitoneally. Animals were assessed daily for body weight, stool consistency, and rectal bleeding. Occult rectal bleeding was determined using the Hemoccult test (Beckman Coulter, Brea, CA). On day 9, blood and colon samples were taken for further analysis. The disease activity index score during the experiment and the histology scores were calculated as previously described.
      • Siegmund B.
      • Lehr H.A.
      • Fantuzzi G.
      • Dinarello C.A.
      IL-1 beta -converting enzyme (caspase-1) in intestinal inflammation.
      Serum cytokine levels were assessed using enzyme-linked immunosorbent assays for the respective cytokines (Thermo Fisher Scientific, Waltham, MA) according to the manufacturer’s protocol. For barrier permeability assay, mice received 0.5 mg/g fluorescein isothiocyanate–dextran (average molecular weight, 3–5 kilodaltons; Sigma Aldrich) via oral gavage. After 4 hours, blood samples were taken and sera were measured at 490 nm.

      Isolation of Primary IECs

      For isolation of primary IECs, colons were harvested from 8-week-old naïve male mice. After flushing with ice-cold PBS, the colon was opened longitudinally and cut into small pieces. Tissue pieces were transferred to a 50-mL tube containing 30 mL 1.5 mmol/L EDTA in Hank’s balanced salt solution (Sigma Aldrich) and placed horizontally on an orbital shaker for 10 minutes at 37°C. The supernatant was discarded and 15 mL fresh 1.5 mmol/L EDTA in Hank’s balanced salt solution was added. Tubes then were vortexed for 1 minute, and the supernatant was collected. This procedure was repeated twice. Supernatants then were combined and centrifuged for 10 minutes at 300 × g. Cells were washed and resuspended in Dulbecco’s modified Eagle medium (DMEM) containing 20% fetal calf serum (Thermo Fisher), 1% penicillin/streptomycin cocktail (Thermo Fisher), and 100 ng/mL recombinant murine epidermal growth factor (R&D Systems, Minneapolis, MN). Culture plates were coated with fibronectin/collagen IV (both Thermo Fisher) for 2 hours in advance at 37°C. Cells were counted using a CASY TT cell counter (Omni Life Sciences, Bremen, Germany). Cells were cultured for 24 hours before the respective experiments.

      Cell Viability and Caspase Activation Assay

      For cell viability analysis, cells were seeded on a 96-well plate (0.75 × 105 cells/well) as described earlier. Eight hours after stimulation with either Tnf-α (50 μg/mL) or DSS (3%), media were replaced with 100 μL PBS containing 20% v/v MTS reagent (Abcam, Cambridge, UK), and cells were kept for 45 minutes. Optical density was assessed at 490 nm using a microplate reader (Tecan, Mennedorf, Switzerland).
      To determine caspase-3 activation, a commercial assay (Abcam) was used according to the manufacturer’s recommendations. Plates were read at 405 nm. The caspase inhibitors fluoromethyl ketone-derived caspase-9 inhibitor (tetrapeptide inhibitor for caspase-8, 20 μmol/L; R&D Systems), fluoromethyl ketone-derived caspase-8 inhibitor (tetrapeptide inhibitor for caspase-9, 20 μmol/L; R&D Systems) or vehicle substance (1% DMSO) were added 30 minutes before stimulation. For determination of caspase-8 and caspase-9 activation in the intestinal epithelium after DSS-induced colitis, protein lysates were prepared from colon whole-tissue pieces using denaturing lysis buffer (described by Kwon et al
      • Kwon S.H.
      • Florens L.
      • Swanson S.K.
      • Washburn M.P.
      • Abmayr S.M.
      • Workman J.L.
      Heterochromatin protein 1 (HP1) connects the FACT histone chaperone complex to the phosphorylated CTD of RNA polymerase II.
      ) and protein concentration was evaluated using the DC protein assay kit (Thermo Fisher). Samples were adjusted to a total protein concentration of 200 μg/mL and caspase activation was determined using colorimetric assays for caspase-8 and caspase-9 activation, according to the manufacturer’s protocol (Abcam).

      Mitochondria Isolation and Western Blot

      Mitochondria and cytosolic cell fractions were isolated from primary IECs (5 × 106 cells) using a commercial mitochondria isolation kit (Thermo Fisher Scientific), according to the manufacturer’s recommendations. Plasma membranes were isolated following the previously published protocol by Kwon et al.
      • Kwon S.H.
      • Florens L.
      • Swanson S.K.
      • Washburn M.P.
      • Abmayr S.M.
      • Workman J.L.
      Heterochromatin protein 1 (HP1) connects the FACT histone chaperone complex to the phosphorylated CTD of RNA polymerase II.
      Western blot was performed as described before.
      • Albrecht S.
      • Korr S.
      • Nowack L.
      • Narayanan V.
      • Starost L.
      • Stortz F.
      • Arauzo-Bravo M.J.
      • Meuth S.G.
      • Kuhlmann T.
      • Hundehege P.
      The K2P-channel TASK1 affects Oligodendroglial differentiation but not myelin restoration.
      Antibodies were used as follows: anti-KNCK9 (P5247, 1:1.000; Sigma Aldrich), antivimentin (D21H3, 1:2.000; CST), antimitofilin (ab205718, 1:2.000; Abcam), and anti-CD40 (ab13545, 1:1.500; Abcam). A horseradish-peroxidase–conjugated secondary antibody was used for chemiluminescent labeling (anti-rabbit-IgG, A3682, 1:10,000; Sigma Aldrich).

      RNA Isolation and Quantitative Real-Time PCR

      Cultured intestinal epithelial cells were harvested and RNA was isolated using the Quick-RNA Miniprep kit (Zymo Research, Freiburg, Germany) according to the manufacturer’s instructions, including the optional DNA digestion protocol. Synthesis of complementary DNA was performed using the Maxima H Minus First Strand Synthesis Kit (Fermentas, Waltham, MA) according to the protocol recommended by the manufacturer. Real-time PCR was performed using premanufactured probes (Applied Biosystems, Waltham, MA) according to Table 1. Assays were run on a StepOne Plus Thermocycler (Applied Biosystems).
      Table 1List of TaqMan Probes Used in the Text Including Unique IDs From the Manufacturer’s Database (Thermo Fisher)
      GeneTaqMan assay ID
      Kcnk1Mm00434624_m1
      Kcnk2Mm01323942_m1
      Kcnk3Mm04213388_s1
      Kcnk4Mm00434626_m1
      Kcnk5Mm00498900_m1
      Kcnk6Mm01176312_g1
      Kcnk9Mm02014295_s1
      Kcnk10Mm00504118_m1
      Kcnk12Mm02344857_m1
      Kcnk13Mm00524725_m1
      Kcnk16Mm01279322_m1
      Kcnk18Mm01702237_m1
      Grp78Mm00517691_m1
      ChopMm01135937_g1
      18S rRNAHs99999901_s1
      NOTE. PCR was performed using the StepOne Plus system (Thermo Fisher) with predesigned TaqMan probes (Thermo Fisher). 18S ribosomal RNA probes were used as reference. The prespecified PCR protocol was used as recommended by the manufacturer.

      Detection of Intracellular Calcium Concentration and Mitochondrial Function

      After isolation, 0.75 × 105 cells per well were cultured for 24 hours in the presence of epidermal growth factor in a 96-well plate coated with fibronectin/collagen IV for 24 hours as described earlier. Before the experiment, the medium was removed and replaced with 100 μL DMEM without phenol red containing 5 μmol/L of the Ca2+-sensitive fluorescent dye Fura-2–AM (Abcam). After 30 minutes, the medium was changed to DMEM without phenol red. After 15 minutes, plates were placed in the microplate reader (excitation wavelengths, 340 nm and 380 nm) and emission was determined at 510 nm every 30 seconds. After 5 minutes of measurement, 50 μL DSS dissolved in PBS was added to a final concentration of 5% DSS via the plate reader’s dispenser pumps. Further substances were added as described in the respective experiments (thapsigargin: 1 μmol/L final concentration, treatment 4 hours before the experiment; A293: 10 μmol/L final concentration, kindly provided by Sanofi Aventis; and PK-THPP: 10 μmol/L final concentration) in 50 μL dispensing volume. Mitochondrial membrane potential breakdown was assessed using the cationic dye JC-1 (Thermo Fisher Scientific). Cells were prepared as described earlier but were incubated with 15 μmol/L JC-1. The excitation wavelength was set to 515 nm and emission was detected at 535 nm and 595 nm. The Fenton reaction and superoxide formation were assessed using 2′,7′-dichlorofluorescin diacetate and MitoSOX (both Thermo Fisher Scientific), according to the manufacturer’s recommendations. Assays for function of the mitochondrial electron transport chain were conducted 30 minutes after DSS exposure.

      Immunofluorescence Staining

      Immunohistochemistry was performed as described before.
      • Albrecht S.
      • Korr S.
      • Nowack L.
      • Narayanan V.
      • Starost L.
      • Stortz F.
      • Arauzo-Bravo M.J.
      • Meuth S.G.
      • Kuhlmann T.
      • Hundehege P.
      The K2P-channel TASK1 affects Oligodendroglial differentiation but not myelin restoration.
      The following antibodies were used: rabbit anti-mouse-Task1 (P0981, 1:250; Sigma Aldrich), rabbit anti-mouse-Task3 (P5247, 1:250; Sigma Aldrich), and rabbit anti-mouse–caspase 3 p17 (ab13847, 1:250; Abcam). A goat anti-rabbit IgG–Alexa Fluor 488-conjugate (A27034, 1:500; Thermo Fisher Scientific) was used as the secondary antibody. For visualizing cell nuclei within immunofluorescent stainings, sections were embedded in Fluoromount-G with 4',6-diamidino-2-phenylindole (Thermo Fisher Scientific) before microscopic analysis.

      Statistical Analysis

      Statistical analysis was conducted using IBM SPSS Statistics 28.0 (IBM), and graphs were created using PRISM 9.0.0 (GraphPad Software). Significance was evaluated using 2-way analysis of variance including the Bonferroni post hoc test for clinical data from DSS-induced colitis models. Column statistics were performed using the Mann–Whitney U test for comparison of 2 groups or the Kruskal–Wallis test including the Dunn post hoc test for comparison of multiple groups. A P value less than .05 was considered statistically significant.

      Acknowledgements

      The authors would like to thank Iska Loesmann and Jeannette Budde for their expert technical support.

      CRediT Authorship Contributions

      Steffen Pfeuffer, MD (Data curation: Lead; Formal analysis: Lead; Writing – original draft: Lead)
      Thomas Muentefering, MSc (Formal analysis: Equal; Methodology: Equal; Writing – review & editing: Supporting)
      Leoni Rolfes, MD (Data curation: Supporting; Writing – review & editing: Supporting)
      Frederike Anne Straeten, MD (Data curation: Supporting; Writing – review & editing: Supporting)
      Susann Eichler, PhD (Data curation: Supporting; Formal analysis: Supporting; Writing – review & editing: Supporting)
      Joel Gruchot, MSc (Data curation: Supporting; Formal analysis: Supporting)
      Vera Dobelmann, MSc (Data curation: Supporting)
      Tim Prozorovski, PhD (Data curation: Supporting)
      Boris Görg, PhD (Data curation: Supporting; Formal analysis: Supporting)
      Mihael Vucur, PhD (Data curation: Supporting; Formal analysis: Supporting)
      Carsten Berndt, PhD (Data curation: Supporting; Formal analysis: Supporting; Writing – review & editing: Supporting)
      Patrick Küry, PhD (Writing – review & editing: Supporting)
      Tobias Ruck, MD (Formal analysis: Supporting; Writing – review & editing: Supporting)
      Stefan Bittner, MD (Conceptualization: Supporting; Writing – review & editing: Supporting)
      Dominik Bettenworth, MD (Methodology: Supporting; Writing – review & editing: Supporting)
      Thomas Budde, PhD (Writing – review & editing: Supporting)
      Tom Luedde, MD, PhD (Methodology: Supporting; Writing – review & editing: Supporting)
      Sven G Meuth, MD, PhD (Conceptualization: Lead; Project administration: Lead; Writing – review & editing: Equal)

      References

        • Enyedi P.
        • Czirjak G.
        Molecular background of leak K+ currents: two-pore domain potassium channels.
        Physiol Rev. 2010; 90: 559-605
        • Bittner S.
        • Budde T.
        • Wiendl H.
        • Meuth S.G.
        From the background to the spotlight: TASK channels in pathological conditions.
        Brain Pathol. 2010; 20: 999-1009
        • Meuth S.G.
        • Bittner S.
        • Meuth P.
        • Simon O.J.
        • Budde T.
        • Wiendl H.
        TWIK-related acid-sensitive K+ channel 1 (TASK1) and TASK3 critically influence T lymphocyte effector functions.
        J Biol Chem. 2008; 283: 14559-14570
        • Bittner S.
        • Bauer M.A.
        • Ehling P.
        • Bobak N.
        • Breuer J.
        • Herrmann A.M.
        • Golfels M.
        • Wiendl H.
        • Budde T.
        • Meuth S.G.
        The TASK1 channel inhibitor A293 shows efficacy in a mouse model of multiple sclerosis.
        Exp Neurol. 2012; 238: 149-155
        • Saadati H.R.
        • Wittig M.
        • Helbig I.
        • Hasler R.
        • Anderson C.A.
        • Mathew C.G.
        • Kupcinskas L.
        • Parkes M.
        • Karlsen T.H.
        • Rosenstiel P.
        • Schreiber S.
        • Franke A.
        Genome-wide rare copy number variation screening in ulcerative colitis identifies potential susceptibility loci.
        BMC Med Genet. 2016; 17: 26
        • Wilke B.U.
        • Lindner M.
        • Greifenberg L.
        • Albus A.
        • Kronimus Y.
        • Bunemann M.
        • Leitner M.G.
        • Oliver D.
        Diacylglycerol mediates regulation of TASK potassium channels by Gq-coupled receptors.
        Nat Commun. 2014; 5: 5540
        • Bista P.
        • Pawlowski M.
        • Cerina M.
        • Ehling P.
        • Leist M.
        • Meuth P.
        • Aissaoui A.
        • Borsotto M.
        • Heurteaux C.
        • Decher N.
        • Pape H.C.
        • Oliver D.
        • Meuth S.G.
        • Budde T.
        Differential phospholipase C-dependent modulation of TASK and TREK two-pore domain K+ channels in rat thalamocortical relay neurons.
        J Physiol. 2015; 593: 127-144
        • Cahalan M.D.
        • Chandy K.G.
        The functional network of ion channels in T lymphocytes.
        Immunol Rev. 2009; 231: 59-87
        • Czirjak G.
        • Enyedi P.
        Formation of functional heterodimers between the TASK-1 and TASK-3 two-pore domain potassium channel subunits.
        J Biol Chem. 2002; 277: 5426-5432
        • Samak G.
        • Chaudhry K.K.
        • Gangwar R.
        • Narayanan D.
        • Jaggar J.H.
        • Rao R.
        Calcium/Ask1/MKK7/JNK2/c-Src signalling cascade mediates disruption of intestinal epithelial tight junctions by dextran sulfate sodium.
        Biochem J. 2015; 465: 503-515
        • Chang J.T.
        Pathophysiology of inflammatory bowel diseases.
        N Engl J Med. 2020; 383: 2652-2664
        • Torres J.
        • Mehandru S.
        • Colombel J.F.
        • Peyrin-Biroulet L.
        Crohn's disease.
        Lancet. 2017; 389: 1741-1755
        • Nowacki T.M.
        • Bruckner M.
        • Eveslage M.
        • Tepasse P.
        • Pott F.
        • Thoennissen N.H.
        • Hengst K.
        • Ross M.
        • Bettenworth D.
        The risk of colorectal cancer in patients with ulcerative colitis.
        Dig Dis Sci. 2015; 60: 492-501
        • Zeng C.
        • Long X.
        • Cotten J.F.
        • Forman S.A.
        • Solt K.
        • Faingold C.L.
        • Feng H.J.
        Fluoxetine prevents respiratory arrest without enhancing ventilation in DBA/1 mice.
        Epilepsy Behav. 2015; 45: 1-7
        • Srisomboon Y.
        • Zaidman N.A.
        • Maniak P.J.
        • Deachapunya C.
        • O'Grady S.M.
        P2Y receptor regulation of K2P channels that facilitate K(+) secretion by human mammary epithelial cells.
        Am J Physiol Cell Physiol. 2018; 314: C627-C639
        • Ozoren N.
        • Kim K.
        • Burns T.F.
        • Dicker D.T.
        • Moscioni A.D.
        • El-Deiry W.S.
        The caspase 9 inhibitor Z-LEHD-FMK protects human liver cells while permitting death of cancer cells exposed to tumor necrosis factor-related apoptosis-inducing ligand.
        Cancer Res. 2000; 60: 6259-6265
        • Sammels E.
        • Parys J.B.
        • Missiaen L.
        • De Smedt H.
        • Bultynck G.
        Intracellular Ca2+ storage in health and disease: a dynamic equilibrium.
        Cell Calcium. 2010; 47: 297-314
        • Das I.
        • Png C.W.
        • Oancea I.
        • Hasnain S.Z.
        • Lourie R.
        • Proctor M.
        • Eri R.D.
        • Sheng Y.
        • Crane D.I.
        • Florin T.H.
        • McGuckin M.A.
        Glucocorticoids alleviate intestinal ER stress by enhancing protein folding and degradation of misfolded proteins.
        J Exp Med. 2013; 210: 1201-1216
        • Nagy D.
        • Gonczi M.
        • Dienes B.
        • Szoor A.
        • Fodor J.
        • Nagy Z.
        • Toth A.
        • Fodor T.
        • Bai P.
        • Szucs G.
        • Rusznak Z.
        • Csernoch L.
        Silencing the KCNK9 potassium channel (TASK-3) gene disturbs mitochondrial function, causes mitochondrial depolarization, and induces apoptosis of human melanoma cells.
        Arch Dermatol Res. 2014; 306: 885-902
        • Chokshi R.H.
        • Larsen A.T.
        • Bhayana B.
        • Cotten J.F.
        Breathing stimulant compounds inhibit TASK-3 potassium channel function likely by binding at a common site in the channel pore.
        Mol Pharmacol. 2015; 88: 926-934
        • Wan Y.
        • Yang L.
        • Jiang S.
        • Qian D.
        • Duan J.
        Excessive apoptosis in ulcerative colitis: crosstalk between apoptosis, ROS, ER stress, and intestinal homeostasis.
        Inflamm Bowel Dis. 2022; 28: 639-648
        • Albrecht S.
        • Korr S.
        • Nowack L.
        • Narayanan V.
        • Starost L.
        • Stortz F.
        • Arauzo-Bravo M.J.
        • Meuth S.G.
        • Kuhlmann T.
        • Hundehege P.
        The K2P-channel TASK1 affects Oligodendroglial differentiation but not myelin restoration.
        Glia. 2019; 67: 870-883
        • Meuth S.G.
        • Herrmann A.M.
        • Ip C.W.
        • Kanyshkova T.
        • Bittner S.
        • Weishaupt A.
        • Budde T.
        • Wiendl H.
        The two-pore domain potassium channel TASK3 functionally impacts glioma cell death.
        J Neurooncol. 2008; 87: 263-270
        • Nakakura S.
        • Matsui M.
        • Sato A.
        • Ishii M.
        • Endo K.
        • Muragishi S.
        • Murase M.
        • Kito H.
        • Niguma H.
        • Kurokawa N.
        • Fujii M.
        • Araki M.
        • Araki K.
        • Ohya S.
        Pathophysiological significance of the two-pore domain K(+) channel K2P5.1 in splenic CD4(+)CD25(-) T cell subset from a chemically-induced murine inflammatory bowel disease model.
        Front Physiol. 2015; 6: 299
        • Feske S.
        CRAC channels and disease - from human CRAC channelopathies and animal models to novel drugs.
        Cell Calcium. 2019; 80: 112-116
        • Concepcion A.R.
        • Feske S.
        Regulation of epithelial ion transport in exocrine glands by store-operated Ca(2+) entry.
        Cell Calcium. 2017; 63: 53-59
        • Guo C.
        • Sun L.
        • Chen X.
        • Zhang D.
        Oxidative stress, mitochondrial damage and neurodegenerative diseases.
        Neural Regen Res. 2013; 8: 2003-2014
        • Mancini N.L.
        • Goudie L.
        • Xu W.
        • Sabouny R.
        • Rajeev S.
        • Wang A.
        • Esquerre N.
        • Al Rajabi A.
        • Jayme T.S.
        • van Tilburg Bernandes E.
        • Nasser Y.
        • Ferraz J.G.P.
        • Shutt T.
        • Shearer J.
        • McKay D.M.
        Perturbed mitochondrial dynamics is a novel feature of colitis that can be targeted to lessen disease.
        Cell Mol Gastroenterol Hepatol. 2020; 10: 287-307
        • Malhotra J.D.
        • Kaufman R.J.
        ER stress and its functional link to mitochondria: role in cell survival and death.
        Cold Spring Harb Perspect Biol. 2011; 3: a004424
        • Wrzosek A.
        • Augustynek B.
        • Zochowska M.
        • Szewczyk A.
        Mitochondrial potassium channels as druggable targets.
        Biomolecules. 2020; 10: 1200
        • Bachmann M.
        • Li W.
        • Edwards M.J.
        • Ahmad S.A.
        • Patel S.
        • Szabo I.
        • Gulbins E.
        Voltage-gated potassium channels as regulators of cell death.
        Front Cell Dev Biol. 2020; 8611853
        • Johnstone L.S.
        • Graham S.J.
        • Dziadek M.A.
        STIM proteins: integrators of signalling pathways in development, differentiation and disease.
        J Cell Mol Med. 2010; 14: 1890-1903
        • Rusznak Z.
        • Bakondi G.
        • Kosztka L.
        • Pocsai K.
        • Dienes B.
        • Fodor J.
        • Telek A.
        • Gonczi M.
        • Szucs G.
        • Csernoch L.
        Mitochondrial expression of the two-pore domain TASK-3 channels in malignantly transformed and non-malignant human cells.
        Virchows Arch. 2008; 452: 415-426
        • Kalkavan H.
        • Green D.R.
        MOMP, cell suicide as a BCL-2 family business.
        Cell Death Differ. 2018; 25: 46-55
        • Huai J.
        • Vogtle F.N.
        • Jockel L.
        • Li Y.
        • Kiefer T.
        • Ricci J.E.
        • Borner C.
        TNFalpha-induced lysosomal membrane permeability is downstream of MOMP and triggered by caspase-mediated NDUFS1 cleavage and ROS formation.
        J Cell Sci. 2013; 126: 4015-4025
        • Cotten J.F.
        TASK-1 (KCNK3) and TASK-3 (KCNK9) tandem pore potassium channel antagonists stimulate breathing in isoflurane-anesthetized rats.
        Anesth Analg. 2013; 116: 810-816
        • Babickova J.
        • Tothova L.
        • Lengyelova E.
        • Bartonova A.
        • Hodosy J.
        • Gardlik R.
        • Celec P.
        Sex differences in experimentally induced colitis in mice: a role for estrogens.
        Inflammation. 2015; 38: 1996-2006
        • Siegmund B.
        • Lehr H.A.
        • Fantuzzi G.
        • Dinarello C.A.
        IL-1 beta -converting enzyme (caspase-1) in intestinal inflammation.
        Proc Natl Acad Sci U S A. 2001; 98: 13249-13254
        • Kwon S.H.
        • Florens L.
        • Swanson S.K.
        • Washburn M.P.
        • Abmayr S.M.
        • Workman J.L.
        Heterochromatin protein 1 (HP1) connects the FACT histone chaperone complex to the phosphorylated CTD of RNA polymerase II.
        Genes Dev. 2010; 24: 2133-2145