Gating the mechanical channel Piezo1: a comparison between whole-cell and patch recording

Abstract

Piezo1 is a eukaryotic cation-selective mechanosensitive ion channel. To understand channel function in vivo, we first need to analyze and compare the response in the whole cell and the patch. In patches, Piezo1 inactivates and the current is fit well by a 3-state model with a single pressure-dependent rate. However, repeated stimulation led to an irreversible loss of inactivation. Remarkably, the loss of inactivation did not occur on a channel-by-channel basis but on all channels at the same time. Thus, the channels are in common mechanical domain. Divalent ions decreased the unitary conductance from ~68 pS to ~37 pS, irrespective of the cation species. Mg and Ca did not affect inactivation rates, but Zn caused a 3-fold slowing. CytochalasinD (cytoD) does not alter inactivation rates or the transition to the non-inactivating mode but does reduce the steady-state response. Whole-cell currents were similar to patch currents but also had significant differences. In contrast to the patch, cytoD inhibited the current suggesting that the activating forces were transmitted through the actin cytoskeleton. Hypotonic swelling that prestressed the cytoskeleton and the bilayer greatly increased the sensitivity of both control and cytoD cells so there are two pathways to transmit force to the channels. In contrast to patch, removing divalent ions decreased the whole-cell current. The difference between whole cell and patch properties provide new insights into our understanding of the Piezo1 gating mechanisms and cautions against generalization to in situ behavior.

Figure 1. Inactivation in cell-attached patches. Panel (A) shows the results of a two-step staircase stimulus where the multichannel currents are shown to inactivate rather than adapt. The first step activates the channels and they proceed to desensitize. If desensitization were caused by adaptation of the stimulus the second step would produce a transient current similar to the first step. In general the second step produced very little current and in the example shown, the second step produced a complete conversion of the transient current to a non-inactivating current. Panel (B) shows the state model used to fit the transient and the optimal rate constants were obtained from the MAC routine in QUB. The rate constant k₁₂ = k¹² *exp(q*P) where q is the pressure sensitivity shown in parentheses after the constant k₁₂ and k¹² is the rat at zero pressure. For the fitted data, note that once the stimulus ends the recorded current trace is smaller than that of the model for many seconds as drop in pressure wrinkled the membrane and it took time to reanneal.

Figure 2. Irreversible transition of inactivating Piezo1 to a non-inactivating form. Panel (A) is the initial response in cell-attached mode showing a transient with no steady-state response while the next stimulus has a significant steady-state indicated by an arrow. After repeated stimuli of high pressure and voltage, all channels lose inactivation simultaneously and irreversibly (B). Panels (C) and (D) are single channel recordings of the inactivating and non-inactivating forms, respectively (average data of individual pulses are shown just below the stimulus trace). Treatment with cytoD (4–6 μM) reduces the ratio of peak to steady-state current (E) but does not change the transition from inactivating to non-inactivating (F), and the inactivation rate is unaltered.

Figure 3. Conductance of Piezo1 in the presence and absence of divalent ions. (A) Without divalent ions the unitary conductance is approximately 68 pS. (B) In the presence of 1 mM Mg the conductance is approximately 37 pS. Panel C shows inactivation with the falling phase fit to a single exponential indicated. The time constant of the fit (τ) is shown for different divalent ions. Mg²⁺and Ca²⁺ have a small effect on the inactivation rate but Zn slows it 3-fold.

Figure 4. Whole-cell currents require divalent ions. (A) Cells were perfused with bath solution containing Mg²⁺. Holding potential was -60 mV and the stimulus is the indicated ramp. Removal of divalent ions reduced the whole-cell currents and the addition of Ca²⁺ restored them. Panel (B) shows the average response of the data from (A), and (C) shows the aggregate response (mean ± SD).

Figure 5. Disruption of the cytoskeleton and swelling alters Piezo1 sensitivity. Panel (A) shows typical current traces from two cells (control and cytoD) with and without swelling. Hypotonic swelling greatly enhanced the indentation sensitivity (-60 mV holding potential and 4 μm indentation) resulting in a 6-fold increase in current compared with control (B). Treatment with cytoD reduced the mean response by 11-fold but that response was potentiated by hypotonic swelling suggesting that there are at least two modes of stress transfer. Panel (B) summarizes these results (squares) and shows that treatment of cells with cytoD suppresses the activity of Piezo1 (triangles) compared with the control (circles). Panel (C) is the mean peak current (± SD) from many cells subjected to swelling or cytoD.