Calcium-dependent potassium channels control proliferation of cardiac progenitor cells and bone marrow-derived mesenchymal stem cells

Abstract

Key points: Ex vivo proliferated c-Kit⁺ endogenous cardiac progenitor cells (eCPCs) obtained from mouse and human cardiac tissues have been reported to express a wide range of functional ion channels. In contrast to previous reports in cultured c-Kit⁺ eCPCs, we found that ion currents were minimal in freshly isolated cells. However, inclusion of free Ca²⁺ intracellularly revealed a prominent inwardly rectifying current identified as the intermediate conductance Ca²⁺ -activated K⁺ current (KCa3.1) Electrical function of both c-Kit⁺ eCPCs and bone marrow-derived mesenchymal stem cells is critically governed by KCa3.1 calcium-dependent potassium channels. Ca²⁺ -induced increases in KCa3.1 conductance are necessary to optimize membrane potential during Ca²⁺ entry. Membrane hyperpolarization due to KCa3.1 activation maintains the driving force for Ca²⁺ entry that activates stem cell proliferation. Cardiac disease downregulates KCa3.1 channels in resident cardiac progenitor cells. Alterations in KCa3.1 may have pathophysiological and therapeutic significance in regenerative medicine.

Abstract: Endogenous c-Kit⁺ cardiac progenitor cells (eCPCs) and bone marrow (BM)-derived mesenchymal stem cells (MSCs) are being developed for cardiac regenerative therapy, but a better understanding of their physiology is needed. Here, we addressed the unknown functional role of ion channels in freshly isolated eCPCs and expanded BM-MSCs using patch-clamp, microfluorometry and confocal microscopy. Isolated c-Kit⁺ eCPCs were purified from dog hearts by immunomagnetic selection. Ion currents were barely detectable in freshly isolated c-Kit⁺ eCPCs with buffering of intracellular calcium (Ca²⁺_i ). Under conditions allowing free intracellular Ca²⁺ , freshly isolated c-Kit⁺ eCPCs and ex vivo proliferated BM-MSCs showed prominent voltage-independent conductances that were sensitive to intermediate-conductance K⁺ -channel (KCa3.1 current, I_KCa3.1 ) blockers and corresponding gene (KCNN4)-expression knockdown. Depletion of Ca²⁺_i induced membrane-potential (V_mem ) depolarization, while store-operated Ca²⁺ entry (SOCE) hyperpolarized V_mem in both cell types. The hyperpolarizing SOCE effect was substantially reduced by I_KCa3.1 or SOCE blockade (TRAM-34, 2-APB), and I_KCa3.1 blockade (TRAM-34) or KCNN4-knockdown decreased the Ca²⁺ entry resulting from SOCE. I_KCa3.1 suppression reduced c-Kit⁺ eCPC and BM-MSC proliferation, while significantly altering the profile of cyclin expression. I_KCa3.1 was reduced in c-Kit⁺ eCPCs isolated from dogs with congestive heart failure (CHF), along with corresponding KCNN4 mRNA. Under perforated-patch conditions to maintain physiological [Ca²⁺ ]_i , c-Kit⁺ eCPCs from CHF dogs had less negative resting membrane potentials (-58 ± 7 mV) versus c-Kit⁺ eCPCs from control dogs (-73 ± 3 mV, P < 0.05), along with slower proliferation. Our study suggests that Ca²⁺ -induced increases in I_KCa3.1 are necessary to optimize membrane potential during the Ca²⁺ entry that activates progenitor cell proliferation, and that alterations in KCa3.1 may have pathophysiological and therapeutic significance in regenerative medicine.

Keywords: Calcium; Calcium Signalling; Cardiac progenitor cells; Heart Failure; Ion Channels; Stem cells.

Figure 1. Selective enrichment of c‐Kit⁺ cells following immunomagnetic selection

A, immunostaining of freshly isolated c‐Kit⁺ eCPCs and the non‐myocyte cell fraction from which c‐Kit⁺ eCPCs were purified. Staining shown is nuclear (DAPI, blue) and c‐Kit protein (red). B, fold‐change of KIT, DDR2, CD31, CD34, CD45, CD90 and CD105 mRNA expression level before versus after immunomagnetic sorting. ^* P < 0.05, ^** P < 0.01, ^*** P < 0.001 vs. before sorting, by paired Student's t tests; N = number of dogs.

Figure 2. Functional ion currents endogenously expressed in freshly isolated c‐Kit⁺ eCPCs

A, original ion current recordings obtained from freshly isolated c‐Kit⁺ eCPCs under different experimental conditions (grey/beige recording, Ca²⁺ _i buffered; black recording, 300 nmol l⁻¹ free [Ca²⁺]_i) with 1000 ms voltage ramps from −120 mV to +80 mV with a holding potential of −40 mV. B, I–V relationship of whole‐cell currents recorded in the presence or absence of free‐[Ca²⁺]_i (tight‐seal). C, I–V relationship of whole‐cell ion currents recorded with perforated‐patch configuration (nystatin). D, reversal potential of whole‐cell ion current recorded under various experimental conditions. E, resting potential of c‐Kit⁺ eCPCs under various experimental conditions. ^*** P < 0.001 vs. 300 nmol l⁻¹ free [Ca²⁺]_i, by two‐way repeated‐measures ANOVA (B); one‐way ANOVA with individual‐mean comparisons by Bonferroni‐corrected t tests (D–E); n/N = number of cells/dogs per group. [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 3. Properties of endogenous Ca²⁺‐activated K⁺ current in freshly isolated c‐Kit⁺ eCPCs and responses to various blockers

A, effect of increasing [K⁺]_o to 100 mmol l⁻¹ on Ca²⁺‐activated currents. B and C, responses of currents in freshly isolated c‐Kit⁺ eCPCs to BKCa (B) and KCa3.1‐blockers (C). ^*** P < 0.01 vs. baseline, by two‐way repeated‐measures ANOVA; n/N = number of cells/dogs per group. [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 4. Expression of ion channel subunits in freshly isolated c‐Kit⁺ eCPCs

A, mRNA expression of subunits encoding voltage‐gated (Kv) K⁺ channels and inward‐rectifier (Kir) K⁺ channels in freshly isolated c‐Kit⁺ eCPCs and non‐myocyte fraction. B, Ca²⁺‐activated K⁺ channels (KCa) in freshly isolated c‐Kit⁺ eCPCs and non‐myocyte fraction. Expression levels are relative to cells from the non‐myocyte fraction from which c‐Kit⁺ eCPCs were purified. ^** P < 0.01, ^*** P < 0.001 vs. non‐myocyte fraction, by paired Student's t tests; N = number of dogs. [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 5. Impact of heart disease on KCa3.1 expression and proliferation markers in freshly isolated c‐Kit⁺ eCPCs

A, I _TRAM‐34 (KCa3.1 current; mean ± SEM) density in c‐Kit⁺ eCPCs freshly isolated from control (CTL) and CHF dog hearts. B, KCNN4 gene expression level in CTL and CHF c‐Kit⁺ eCPCs. C, KCa3.1 protein expression level in CTL and CHF c‐Kit⁺ eCPCs. D, population doubling time of CTL and CHF c‐Kit⁺ eCPCs (passages 2–4). E, cyclin expression profiles of CTL vs. CHF c‐Kit⁺ eCPCs. ^* P < 0.05, ^** P < 0.01, ^*** P < 0.001 vs. CTL c‐Kit⁺ eCPCs, by non‐paired Student's t test; n/N = number of cells/dogs per group (A, B); numbers shown on bars are numbers of dogs per group (C–F).

Figure 6. KCa3.1 currents and membrane potential changes during SOCE in c‐Kit⁺ eCPCs

A, I–V curves (mean ± SEM) of whole cell currents before and during SOCE, in the absence or presence of 1 μmol l⁻¹ TRAM‐34. B and C, original current‐clamp recordings in a single cell (B) and mean ± SEM data for V _mem changes induced by SOCE upon re‐exposure to 1.8 mmol l⁻¹ [Ca²⁺]_o (C). 2‐APB, 2‐aminoethoxydiphenyl borate (250 μmol l⁻¹). ^*** P < 0.001 vs. 1.8 mmol l⁻¹ [Ca²⁺]_o, by two‐way repeated‐measures ANOVA (A); one‐way ANOVA with individual mean comparisons by Bonferroni‐corrected t tests (C); n/N = number of cells/dogs per group. [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 7. Intracellular [Ca²⁺] changes during SOCE in c‐Kit⁺ eCPCs

A, representative Fluo‐4 images in freshly isolated c‐Kit⁺ eCPCs. B, data (mean ± SEM) for Fluo‐4 signal intensity acquired before and during SOCE without or with exposure to TRAM‐34. ^** P < 0.01 vs. control (0.1% DMSO), by two‐way repeated‐measures ANOVA with Bonferroni‐corrected t tests to compare differences between individual mean values; n/N = number of fields/dogs per group.

Figure 8. Effect of KCNN4 gene knockdown on KCa3.1 current, membrane potential and intracellular [Ca²⁺] in c‐Kit⁺ eCPCs

A, KCNN4 gene expression (mean ± SEM) in c‐Kit⁺ eCPCs following control and KCNN4 siRNA molecule delivery (numbers shown are numbers of independent experiments). B, I _TRAM‐34 density (mean ± SEM) 48 h after KCNN4 gene knockdown. C, V _mem in control and KCNN4 gene knockdown c‐Kit⁺ eCPCs recorded with the perforated‐patch technique. D, representative Fluo‐4 images in expanded c‐Kit⁺ eCPCs (passage 2) following knockdown of KCNN4. E, data (mean ± SEM) for basal Fluo‐4 signal intensity (indicating [Ca²⁺]_i) acquired 48 h after KCNN4 gene knockdown. ^*** P < 0.001, ^* P < 0.05 vs. CTL‐siRNA, by two‐way repeated‐measures ANOVA (B); non‐paired Student t tests (panels C and E); n/N = number of cells/dogs per group (B and C); n/N = number of fields/dogs per group (E).

Figure 9. KCa3.1 current and membrane potential changes during SOCE in dog BM‐MSCs

A, I _TRAM‐34 density (mean ± SEM) in expanded dog BM‐MSCs. B, V _mem in MSCs before and after KCa3.1 blockade with TRAM‐34. C and D, original current‐clamp recording from one cell (C) and data (mean ± SEM) for V _mem changes induced by SOCE (D). ^*** P < 0.001 vs. corresponding group, by two‐way repeated‐measures ANOVA (A); paired Student's t test (B); one‐way ANOVA with individual‐mean comparisons by Bonferroni‐corrected t tests (D); numbers shown on bars are numbers of cells studied. [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 10. Intracellular [Ca²⁺] changes during SOCE in dog BM‐MSCs

A, representative Fluo‐4 images in expanded dog MSCs (passage 8). B, data (mean ± SEM) for Fluo‐4 signal intensity acquired before and during SOCE without or with exposure to TRAM‐34. ^*** P < 0.001 vs. corresponding control (0.1% DMSO), by two‐way repeated‐measures ANOVA with individual‐mean comparisons by Bonferroni‐corrected t tests; n/N = number of fields/dogs per group.

Figure 11. Effect of cell culture on I _KCa3.1 in c‐Kit⁺ eCPCs

A, I–V curves (mean ± SEM) of TRAM‐34‐sensitive ion currents recorded in freshly isolated and cultured c‐Kit⁺ eCPCs (300 nmol l⁻¹ free‐[Ca²⁺]_i). n/N = number of cells/dogs per group. B, KIT‐gene expression level in freshly isolated and ex vivo proliferated c‐Kit⁺ eCPCs compared to the non‐myocyte fraction (passage 4). ^* P < 0.05 vs. non‐myocyte fraction, by two‐way repeated‐measures ANOVA with individual‐mean comparisons by Bonferroni‐corrected t tests; numbers shown on bars are numbers of dogs per group. [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 12. Effect of KCa3.1 inhibition on proliferation of progenitor cells

A and B, population doubling time of c‐Kit⁺ eCPCs (A) and BM‐MSCs (B) after treatment with 10 μmol l⁻¹ TRAM‐34 for 48 h. C, population doubling time of c‐Kit⁺ eCPCs following KCNN4‐gene knockdown. ^* P < 0.05 vs. vehicle (0.1% DMSO), by one‐way ANOVA with individual‐mean comparisons by Bonferroni‐corrected t tests (A and B); paired Student's t test (C); numbers shown on bars are numbers of dogs per group. [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 13. Effect of KCa3.1 inhibition on cyclin expression in progenitor cells

A and B, cyclin expression profiles of c‐Kit⁺ eCPCs after treatment with 10 μmol l⁻¹ TRAM‐34 for 48 h; C, following KCNN4‐gene knockdown and after treatment with 10 μmol l⁻¹ TRAM‐34 for 48 h (passages 2–4). D and E, cyclin expression profiles of BM‐MSCs after treatment with 10 μmol l⁻¹ TRAM‐34 for 48 h (passages 4–8). ^* P < 0.05, ^** P < 0.01 vs. control (0.1% DMSO), by paired Student's t tests; numbers shown on bars are numbers of dogs per group.