Block of inhibitory junction potentials and TREK-1 channels in murine colon by Ca2+ store-active drugs

Abstract

Post-junctional enteric inhibitory responses are composed of at least two components attributed to the release of a purine and nitric oxide (NO). The nitrergic component is characterized by membrane potential hyperpolarization; however, the conductances involved and the role of Ca(2+) stores in regulating these conductances are controversial. Conventional microelectrode recordings were performed in intact muscle strips and whole-cell voltage clamp experiments were performed on freshly dispersed cells and COS7 cells stably transfected with TREK-1 channels. Here we show that several Ca(2+) store-active compounds, including caffeine, ryanodine, and cyclopiazonic acid, reduce inhibitory junction potentials and responses to sodium nitroprusside in murine colonic muscles. We previously proposed that two-pore K(+) channels of the TREK family mediate a portion of the hyperpolarization response to NO in colonic muscles. We tested the effects of Ca(2+) store-active drugs in COS cells expressing murine TREK-1 channels and found these compounds block TREK-1 currents. These effects were greatly attenuated by dialysing cells with protein kinase A inhibitory peptide (PKAI). Caffeine also blocked stretch-dependent K(+) (SDK) channels, thought to be due to expression of TREK channels, in colonic myocytes, but these effects were not apparent in excised patches. Taken together our data show that Ca(2+) store-active compounds inhibit TREK-1 channels, native SDK channels, and nitrergic inhibitory junction potentials. These effects appear to be due, in part, to the cAMP/PKA stimulatory actions of these drugs and inhibitory effects of TREK channels.

Figure 1. Inhibitory junction potentials in murine colonic circular muscles

IJPs were evoked by single pulses of EFS (0.3 ms duration). Post-junctional responses (recorded from circular muscle cells) consisted of a rapid hyperpolarization, partial repolarization, and a 2nd longer duration hyperpolarization phase. Top line of traces shows control IJPs elicted in the presence of nifedipine (1 μm). The 2nd component of the IJP is nitrergic and blocked by l-NNA (100 μm) as shown in the first traces on the 2nd line. Apamin (0.3 μm) reduced the 1st component, suggesting it was primarily due to activation of SK channels (2nd traces on 2nd line). A small portion of the 1st component was not completely blocked by this concentration of apamin. Joint application of apamin and l-NNA (3rd trace on 2nd line) blocked most of both components. Traces for the bottom line of traces were lined up for display purposes, but in fact treatment with l-NNA caused depolarization. Thus, the voltage scales in brackets refer to the 1st and 3rd traces, and the scale not in brackets refers to the 2nd trace.

Figure 2. Effects of caffeine on IJPs

In this experiment IJPs were evoked by EFS at 3 pulse durations (0.1, 0.3, 0.5 ms; A) and after treatment with several concentrations of caffeine (0.1–10 mm; B–F). Caffeine caused concentration-dependent reduction in the nitrergic (slow) component of the IJPs. Caffeine also reduced the 1st component of the IJPs, but this was not the main focus of the present study. G and H, summaries of the effects of caffeine on the nitrergic component of the IJP in 5 experiments of this type.

Figure 3. Effects of CPA and ryanodine on IJPs

IJPs were evoked by EFS at 3 pulse durations (0.1, 0.3, 0.5 ms; A). CPA (1–10 μm; B and C) greatly reduced the amplitude of the nitrergic IJPs and essentially blocked this component at 10 μm. CPA also reduced the amplitude of the 1st component of the IJP, but the mechanism of this effect was not investigated in the present study. D and E, IJPs evoked by EFS at 3 pulse durations (0.1, 0.3, 0.5 ms) under control conditions (D) and after ryanodine (10 μm; E). Ryanodine caused a significant decrease in the amplitude of the nitrergic component of the IJP. Recordings in panels A–C and D and E were made from the same cell in separate experiments.

Figure 4. Effects of caffeine on IJPs in the presence of l-NNA or l-methionine

Effects of caffeine on IJPs after the nitrergic component of the IJPs were blocked by l-NNA or l-methionine. A and C, IJPs evoked by EFS at 3 pulse durations (0.1, 0.3, 0.5 ms) in the presence of caffeine which blocked the 2nd (nitrergic) component of the IJPs. The effects of caffeine were repeated after pretreatment with l-NNA (100 μm; B) or l-methionine (1 mm) to block the nitrergic component of the IJP. Application of caffeine after the nitrergic component was blocked did not reveal additional effects of caffeine.

Figure 5. Effects of sodium nitroprusside (SNP) before and after treatment with caffeine or CPA

SNP (1 μm) caused reversible depolarization of muscles (A; addition at downward arrow; washout at upward arrow, all panels). After caffeine pretreatment, SNP did not affect membrane potential (2nd trace in A). The hyperpolarization response was restored after washout of caffeine (3rd trace in A). B, the 1st trace shows another example of hyperpolarization in response to SNP. CPA pretreatment reduced the hyperpolarization in response to SNP (2nd trace). The hyperpolarization in response to SNP was restored after CPA was washed out. Traces for B are lined up for display purposes, but in fact treatment with CPA caused depolarization. Thus, the voltage scales in brackets refer to the 2nd trace and the scales not in brackets refer to the 1st and 3rd traces.

Figure 6. Effects of caffeine on TREK-1 current expressed in COS-7 cells

Currents were recorded under whole cell voltage clamp. The membrane potential was stepped from −80 to +70 mV in 10 mV increments. A, representative traces demonstrating concentration-dependent reduction in TREK-1 currents by caffeine (1–5 mm). B, normalized I–V relationship before and in the presence of caffeine (1–5 mm). C, concentration–response curve. IC₅₀ was 2.4 mm. D, representative single channel recordings in response to ramped potential from −80 to +80 mV. Note significant outward current under control conditions, and reduction in this current in the presence of caffeine (5 mm).

Figure 7. Effects of theophylline on mTREK-1 currents

Membrane potential was stepped from −80 to +70 mV in 10 mV increments. A, representative control currents recorded from COS-7 cells expressing TREK-1 channels. B, the inhibitory effects of theophylline (5 mm) on the TREK-1 currents. C, the current–voltage (I–V) relationship before (^) and after (•) theophylline application.

Figure 8. Effects of Ca²⁺ store-active compounds TREK-1 currents

Membrane potential was stepped from −80 to +70 mV in 10 mV increments. A, D and G, representative TREK-1 currents in 3 cells. B, E and H, the inhibitory effects of thapsigargin (1 μm), CPA (10 μm) and ryanodine (20 μm) on TREK-1 currents, respectively. C, F and I, current–voltage relationship before and after thapsigargin, CPA and ryanodine, respectively.

Figure 9. Effects of caffeine on SDK channels of murine colonic myocytes in single channel recordings

Patches of membrane were held at 0 mV in asymmetrical K⁺ (5/140 mm) gradients. A, application of negative pressure (−30 cmH₂O) to patch pipettes increased channel activity in cell-attached patches. Caffeine decreased the open probability of SDK channels. B, expanded traces from regions of the record in A (a: after negative pressure application; and b: after caffeine). In some records small conductance channels were activated by caffeine. These channels are likely to be K_ATP because these channels are present in colonic myocytes and activated by cAMP-dependent mechanisms (see Sanders et al. 2006). C and D, amplitude histograms during application of negative pressure and caffeine treatment.

Figure 10. Effects of caffeine on single channel conductances in colonic myocytes

On cell patches were held at 0 mV in asymmetrical K⁺ (5/140 mm) gradients. In A a pipette lacking ChTX was used. Caffeine (5 mm) increased the open probability of channels with conductance features consistent with BK channels (A and C). Aa, expanded traces from region of the record denoted in A (a: after caffeine). B and C show amplitude histograms before (B) and after (C) addition of caffeine. In an experiment using a pipette containing ChTX (200 nm), SNP (10 μm) increased the open probability of SDK channels in a cell attached-patch (D). In the continued presence of SNP application of caffeine (5 mm) dramatically decreased open probability of SDK channels. Db, an expanded trace from the region of the record in D (b: after caffeine in the presence of SNP). E and F, amplitude histograms in the presence of SNP (E) and after addition of caffeine (F). G, results from a ramp depolarization protocol to test the direct effects of caffeine on SDK channels in excised patches in the inside-out configuration. Ramp protocols were used for these studies because of the high open probability of SDK channels in excised patches (Koh & Sanders, 2001). Membrane potential was ramped repetitively from −80 to +80 mV. To the right of the slow sweep record in G a record of averaged current from 10 ramps is shown (black trace). Addition of caffeine (5 mm) had no effect on the SDK conductance (red trace).

Figure 11. Effects of forskolin and caffeine on TREK-1 currents in COS-7 cells

Membrane potential was stepped from −80 to +70 mV in 10 mV increments. A–C, currents and summary I–V curves for COS-7 cells expressing TREK-1 channels before (•) and in the presence of forskolin (1 μm) (^). Forskolin, like caffeine inhibited TREK-1 currents. D–F, that the effects of forskolin are blocked by dialysis of cells with protein kinase A inhibitor (PKAI). G–I, the effects of caffeine (5 mm) were greatly reduced in cells dialysed with PKAI. I, current–voltage (I–V) relationships in control (•) and after caffeine (^) in cells dialysed with PKAI.

Figure 12. Effects of CPA and thapsigargin on TREK-1 currents

Membrane potential was stepped from −80 to +70 mV in 10 mV increments. A and D, control currents in cells dialysed with protein kinase A inhibitor (PKAI). B and E, the effects of CPA (10 μm) and thapsigargin (1 μm) on TREK-1 currents in cells dialysed with PKAI. Note that the inhibitory effects of CPA are abolished and CPA becomes an agonist for TREK-1 currents, and the effects of thapsigargin are preserved in spite of the presence of the inhibitor for PKA. C and F, average normalized current-voltage (I–V) relationships in control (•) and after CPA (^) or thapsigargin (^).