A Closely Associated Phospholipase C Regulates Cation Channel Function through Phosphoinositide Hydrolysis

Abstract

In the hemaphroditic sea snail, Aplysia californica, reproduction is initiated when the bag cell neurons secrete egg-laying hormone during a protracted afterdischarge. A source of depolarization for the afterdischarge is a voltage-gated, nonselective cation channel, similar to transient receptor potential (TRP) channels. Once the afterdischarge is triggered, phospholipase C (PLC) is activated to hydrolyze phosphatidylinositol-4,5-bisphosphate (PIP₂) into diacylglycerol (DAG) and inositol trisphosphate (IP₃). We previously reported that a DAG analog, 1-oleoyl-2-acetyl-sn-glycerol (OAG), activates a prominent, inward whole-cell cationic current that is enhanced by IP₃ To examine the underlying mechanism, we investigated the effect of exogenous OAG and IP₃, as well as PLC activation, on cation channel activity and voltage dependence in excised, inside-out patches from cultured bag cell neurons. OAG transiently elevated channel open probability (P_O) when applied to excised patches; however, coapplication of IP₃ prolonged the OAG-induced response. In patches exposed to OAG and IP₃, channel voltage dependence was left-shifted; this was also observed with OAG, but not to the same extent. Introducing the PLC activator, m-3M3FBS, to patches increased channel P_O, suggesting PLC may be physically linked to the channels. Accordingly, blocking PLC with U-73122 ablated the m-3M3FBS-induced elevation in P_O Treatment with m-3M3FBS left-shifted cation channel voltage dependence to a greater extent than exogenous OAG and IP₃ Finally, OAG and IP₃ potentiated the stimulatory effect of PKC, which is also associated with the channel. Thus, the PLC-PKC signaling system is physically localized such that PIP₂ breakdown products liberated during the afterdischarge modulate the cation channel and temporally influence neuronal activity.SIGNIFICANCE STATEMENT Using excised patches from Aplysia bag cell neurons, we present the first evidence of a nonselective cation channel physically associating with phospholipase C (PLC) at the single-channel level. PLC-mediated breakdown of phospholipids generates diacylglycerol and inositol trisphosphate, which activate the cation channel. This is mimicked by exogenous lipids; furthermore, these second messengers left-shift channel voltage dependence and enhance the response of the channel to protein kinase C. PLC-mediated lipid signaling controls single-channel currents to ensure depolarization is maintained for an extended period of firing, termed the afterdischarge, when the bag cell neurons secrete egg-laying hormone to trigger reproduction.

Keywords: afterdischarge; diacylglycerol; inositol trisphosphate; mollusk; neuroendocrine cell; voltage dependence.

Figure 1.

A voltage-dependent cation channel from cultured Aplysia bag cell neurons. A, Cation channel activity in an excised, inside-out patch, voltage-clamped at indicated membrane potentials (left) with Na⁺-based nASW plus 20 mm tetraethylammonium at the extracellular face and a K⁺-aspartate-based intracellular saline with 1 μm Ca²⁺ at the cytoplasmic face. The closed state is at the top of each trace, indicated by −C, and the open-state, designated by −O, is at the bottom (at right of −60 trace). Openings are seen as unitary inward (negative) current deflections. As the patch is depolarized from −90 (bottom trace) to 30 mV (top trace), the channel opens more often, and more time is spent in the open state and less time in the closed state. Scale bars apply to all traces. B, Voltage dependence of cation channel P_O expressed by normalizing activity within each experiment to the P_O at 30 mV. A half-maximal voltage of activation (V₅₀) of 38 mV and a k of 23 are determined from a Boltzmann sigmoidal curve fit to the data. The n value refers to the number of patches. Inset, Phase-contrast micrograph of a bag cell neuron in vitro (top). The pipette is coated with dental wax to reduce noise. After seal formation, the pipette is excised from the soma and current is recorded across the patch at the tip. Schematic of a cation channel in a membrane patch (bottom). C, All-points histograms of recordings in A represent current flow during channel closed (C) and open (O₁ or O₂) states at different voltages. Gaussian fits to these peaks are used to determine channel amplitude plotted in D. D, Single-channel I/V relationship shows subtle rectification at potentials >0 mV. A conductance (g) of 24 pS is derived from a linear fit of −90 to −15 mV; this represents the physiological range of the channel and avoids the region of rectification. E, The cation channel is part of a regulatory complex that includes PKC and calmodulin (CaM). The voltage-sensing domains open the channel in response to depolarization (ΔV_m), whereas PKC-dependent phosphorylation and CaM binding intracellular Ca²⁺ activate the channel, allowing Na⁺ and Ca²⁺ to enter the cell and K⁺ to leave during the afterdischarge.

Figure 2.

Kinetic analysis of true single channels shows that the cation channel stays open longer and favors reopening with depolarization. A, Closed-state (left column) and open-state (right column) dwell times at 30, 0, and −30 mV from a one-channel-only single-channel recording are plotted as histograms. Each histogram is best fit with a probability density function (three closed-state time constants: τ_C1, τ_C2, and τ_C3; two open-state time constants: τ_O1 and τ_O2). The longest closed-state time constant decreases with depolarization (−30: 8.83 ± 6.7 s vs 30: 3.3 ± 2.5 s), whereas the longer open-state time constant subtly increases in duration (−30: 23.7 ± 4.6 ms vs 30: 44.6 ± 18.8 ms). B, The duration of the longest closed-state time constant, τ_C3, decreases linearly with depolarization (left). Data for τ_C3 are normalized to −60 mV, at which τ_C3 is the longest. Similarly, the length of τ_O2 increases linearly with depolarization. Data are normalized to 30 mV, at which τ_O2 is the longest.

Figure 3.

A diacylglycerol analog transiently activates the cation channel, and this is prolonged by IP₃. A, Application of 25 μm OAG, a diacylglycerol analog, to the cytoplasmic face of a cation channel-containing patch held at −60 mV increases activity for the 3 min following addition (middle) compared with the initial recording (top). Nine minutes after the addition, the P_O has returned to the initial level (bottom). Scale bars apply to all traces in A, B, and D. B, Excising patches into intracellular saline containing 5 μm IP₃, and then delivering OAG, increases the extent and duration of the OAG-induced elevation of cation channel activity. Before OAG addition, channel activity is minimal (top). Application of OAG to patches in the presence of IP₃ increases P_O (middle), and this effect is prolonged (bottom). C, Summary data show that the P_O goes up following addition of OAG (gray bars), compared with delivery of DMSO (white). In patches excised into IP₃ (black), the OAG-induced change in P_O is almost twofold higher than untreated cells. The increased P_O is significantly longer in IP₃-treated patches than in untreated cells (6 min: p = 0.0082; 9 min: p = 0.0019; 12 min: p = 0.0171, KW-ANOVA). **p < 0.01 (Dunn's multiple-comparisons test). *p < 0.05 (Dunn's multiple-comparisons test). For this and all subsequent bar graphs, the numbers within the bars refer to the number of patches. In this instance, certain n values change starting at the 6 or 9 min bin due to some patches not lasting the full 12 min recording period. D, The synergy between OAG and IP₃ is blocked by heparin. For an inside-out patch excised into 100 μm heparin, an IP₃-receptor blocker, plus IP₃, the subsequent application of OAG does not alter P_O (bottom) compared with the initial activity (top). E, Summary data of the OAG-induced change in cation channel P_O show that 10 μm (gray bars) or 100 μm heparin (black bars) prevents the activity increase induced by OAG in IP₃ compared with IP₃ alone (white) (p = 0.0010, KW-ANOVA). *p < 0.05 (Dunn's multiple-comparisons test). **p < 0.01 (Dunn's multiple-comparisons test). F, Left, Intracellular Ca²⁺ measured from the soma of a cultured bag cell neuron in nASW under whole-cell voltage-clamp at −60 mV using K⁺-aspartate-based intracellular saline supplemented with 1 mm fura. Bath application of 5 μm IP₃ has no real impact on Ca²⁺, yet introducing 20 μm CPA later on causes a clear Ca²⁺ elevation reflective of store depletion. Right, The average response to IP₃ is not significantly different from a theoretical mean of zero (p > 0.05; one-sample t test), whereas the difference between the mean Ca²⁺ change to CPA versus IP₃ readily meets significance (*p < 0.0001; unpaired Student's t test).

Figure 4.

The voltage dependence of the cation channel is left-shifted by OAG and IP₃. A, Cation channel activity in excised, inside-out patches exposed to 25 μm OAG alone (top traces at each voltage) or excised into 5 μm IP₃ followed by OAG (bottom traces) and held at the potentials indicated on the left. Compared with OAG alone, the P_O of the channel is higher in the presence of OAG + IP₃, particularly at −60 and −15 mV. Ordinate at bottom right applies to −60 mV, whereas that beside the −15 mV trace applies to −15 and 15 mV; abscissa applies to all traces. B, Voltage dependence of cation channel P_O in OAG (half-filled circles), OAG and IP₃ (filled circles), or PMA (open circles). A Boltzmann curve fit to the P_O activity normalized to 30 mV in OAG + IP₃ (thick black) is left-shifted (V₅₀ = −12 mV, k = 23) compared with OAG (dotted gray) (V₅₀ = −3 mV, k = 13) and PMA (thin black) curves. The PMA curve is similar to naive channels (compare with Fig. 1D) and is shown here as a control. OAG with IP₃ results in a larger left-shift in the voltage dependence than OAG alone (−90 mV: p > 0.05, KW-ANOVA; −75 mV: p > 0.05, KW-ANOVA; −60 mV: **p = 0.0046, KW-ANOVA with p < 0.05 for PMA vs OAG + IP₃ and OAG vs OAG + IP₃; −45 mV: **p = 0.0087, KW-ANOVA with p < 0.05 PMA vs OAG + IP₃ and OAG vs OAG + IP₃; −30 mV: *p = 0.0154, KW-ANOVA with p < 0.05 PMA vs OAG + IP₃ and OAG vs OAG + IP₃; −15 mV: **p = 0.0046, KW-ANOVA with p < 0.05 PMA vs OAG + IP₃ and OAG vs OAG + IP₃; 0 mV: p > 0.05, KW-ANOVA; 15 mV: **p = 0.0048, KW-ANOVA with p < 0.05 PMA vs OAG + IP₃; all post hoc tests Dunn's multiple-comparisons).

Figure 5.

A PLC activator increases cation channel activity in excised, inside-out patches. A, Addition of 25 μm o-3M3FBS (bottom), an inactive control molecule, to the cytoplasmic face of a patch held at −60 mV, results in no change in P_O from the initial level (top). B, In a separate patch from a different neuron, applying 25 μm m-3M3FBS, a PLC activator, increases channel activity (bottom trace) compared with the initial period (top). C, In a patch excised from a neuron treated with 5 μm of the PLC inhibitor, U-73122, delivery of m-3M3FBS has no effect on P_O. Scale bars apply to all traces in A–C. D, Summary data show that application of m-3M3FBS to excised patches increases channel activity over the course of ∼6 min and reaches a peak by ∼9 min (gray bars), and this effect can be blocked by U-73122 treatment before patch excision (black). Adding the inactive form, o-3M3FBS, has no effect on channel activity (white) (p = 0.0083, KW-ANOVA). *p < 0.05 (Dunn's multiple-comparisons test).

Figure 6.

PLC activation left-shifts the voltage dependence of the cation channel. A, Cation channel activity increases as holding potentials are depolarized from −60 (bottom) to 15 mV (top) when excised from cells treated with 25 μm of either the active (m-3M3FBS) or inactive (o-3M3FBS) form the PLC activator. However, at both −60 and −15 mV (middle), channel P_O is higher with the active m-3M3FBS. Ordinate at bottom right applies to −60 mV, whereas that beside the −15 mV trace applies to −15 and 15 mV; abscissa applies to all traces. B, Exposure to 25 μm m-3M3FBS (filled circles) results in a left-shifted voltage dependence compared with the inactive o-3M3FBS (open circles). A Boltzmann curve fit to the P_O activity normalized to 30 mV in m-3M3FBS (thick black) is left-shifted (V₅₀ = −24 mV, k = 17) compared with the o-3M3FBS (thin black) curve (V₅₀ = 26 mV, k = 18) (−90: p > 0.05; −75 mV: **p = 0.0095; −60 mV: **p = 0.0087; −45 mV: *p = 0.0260; −30 mV: **p = 0.0012; −15 mV: *p = 0.0350; 0 mV: **p = 0.007; 15 mV: p > 0.05; all Mann–Whitney unpaired U test).

Figure 7.

PKC-induced cation channel activation is increased following addition of PLC breakdown products. A, Addition of 1 mm ATP to the cytoplasmic face of a cation channel held at −60 mV in an excised, inside-out patch increases P_O ∼2-fold (bottom trace) compared with initial activity (top). Prior work shows that this is due to PKC-mediated phosphorylation (i.e., it is sensitive to PKC inhibition and reversed by protein phosphatases) (Wilson et al., 1998). B, Following treatment of an intact bag cell neuron with 25 μm m-3M3FBS, a cation channel-containing patch is excised (top). Delivery of ATP to this patch provokes a marked P_O increase (bottom). C, Moreover, in a different patch, after exposure to 25 μm OAG (top trace), introducing ATP results in an even larger elevation in channel activity. Scale bars apply to all traces in A–D. D, When excised into 5 μm IP₃ followed by 25 μm OAG (top), the response to ATP is a very substantial increase in P_O (bottom). E, Summary data of excised patches show that, compared with pre-ATP, after addition of OAG, with or without IP₃, ATP significantly increases channel activity in these patches. This P_O change is far greater when both OAG and IP₃ are present versus OAG alone (p < 0.0001, KW ANOVA). *p < 0.05 (Dunn's multiple-comparisons test). ***p < 0.001 (Dunn's multiple-comparisons test).

Figure 8.

Schematic representation of the Aplysia cation channel regulatory complex in the bag cell neuron plasma membrane. A, The cation channel is found with closely associated PKC (Wilson et al., 1998), CaM (Lupinsky and Magoski, 2006), and PLC. At rest, the voltage sensors are located closer to the inner leaflet. PIP₂ is located in the plasma membrane. A representative single-channel recording from the channel in an excised patch held at −60 mV indicates that the channel is predominantly in the closed (C) state, with few transitions to the open (O) state. B, In response to the afterdischarge, PLC is activated (lightning bolt) and PIP₂ is hydrolyzed to DAG and IP₃. The combined liberation of DAG and IP₃, along with PKC-dependent phosphorylation and Ca²⁺/CaM, gates the channel in a positive manner. The bag cell neuron membrane also depolarizes, resulting in further channel activation and cation influx. Given these regulatory elements, channel activity increases dramatically, as evidenced by the representative single-channel recording.