Oxidative modulation of Piezo1 channels

Abstract

Emerging evidence suggests that mechanosensitive Piezo1 channels play a role in the pathomechanism of various disorders. However, the mechanisms by which accumulating pathologies regulate Piezo1 activation remain unclear. Oxidative stress, a common feature of neurodegenerative diseases, is associated with generation of reactive oxygen species (ROS). While the dependence of Piezo1 channels on temperature, pH, and voltage has been well studied, the redox regulation of these highly mechanosensitive channels remains unknown. We investigated whether oxidative stress modulates the calcium permeability of Piezo1 channels using red blood cells (RBCs) and HEK293T cells transduced with Piezo1 as model systems. Additionally, using the selective H₂O₂ sensor HyPer7, we examined whether Piezo1 activation induces the generation of endogenous ROS. Using flow cytometry, Ca²⁺-imaging, patch clamp and microaspiration techniques we demonstrate that cell-permeable oxidants hydrogen peroxide (H₂O₂) and Chloramine-T, which specifically oxidize cysteines and methionines, inhibited Yoda1-induced activation of Piezo1 in both cell types. In contrast to Chloramine-T, the membrane-impermeable, cysteine-specific oxidant DTNB (5,5'-dithiobis-(2-nitrobenzoic acid)) also inhibited Piezo1, although its inhibitory effect was less pronounced. Mechanical sensitivity of Piezo1 was reduced by H₂O₂ also in RBCs. Scavenging antioxidants N-acetylcysteine and dithiothreitol decreased or eliminated the inhibitory action of H₂O₂ and Chloramine-T. However, overexpression of the antioxidant transcription factor Nrf2 (Nuclear factor erythroid 2-related factor 2) did not prevent the inhibitory effects of Chloramine-T, suggesting a membrane-delimited site of redox modulation. Notably, Piezo1 activation slightly increased endogenous H2O2 production. Our data suggest that the reduced activity of Piezo1 in the oxidative environment is determined by oxidation of both cysteines and methionines, which are enriched in intracellular domains, with methionines playing a predominant role. Given the role of Piezo1 channels in pathophysiology of numerous disorders, we propose that, under conditions associated with oxidative stress, redox modulation of these mechanosensors could be a significant factor contributing to disease pathology.

Keywords: Hydrogen peroxide; Mechanoreceptors; Oxidation; Oxidative stress; Piezo1 channel.

Graphical abstract

Fig. 1

H₂O₂ and Chloramine-T dose-dependently decrease the activity of Piezo1 channels in human RBCs: A, left: flow cytometry with peristaltic pump for time-lapse mode recording. A, middle, right: flow cytometry gating strategy used to identify human red blood cells (RBCs). Light scatter profile for cells based on forward scatter (FSC) and side scatter (SSC) (the region is set to discriminate between RBCs and debris). Singlet gating based on FSC vs. FSC-height (the region is set to discriminate cell doublets). Fluorescence has been detected in FITC channel as Fluo4 AM intensity. Graph showing representative RBC signal from the baseline (dark red) and during Yoda1 application (black). Combining above criteria, RBCs were reliably identified. B: flow cytometry time lapse recording view. The X-axis represents the time (seconds) from the beginning of sample acquisition. Yoda1 was added at 20s; the Y-axis represents the relative fluorescence of Fluo-4 AM. Note decrease in the level of Fluo-4 AM fluorescence after 20min pretreatment in 500 μM H₂O₂ (right record) compared to control (left record). C, left, middle: the time-course of the Yoda1-mediated Ca²⁺ responses (MFI) after 20 min pretreatment in 100 μM (left) and 500 μM (right) of H₂O₂ in comparison to control. Note the significant decrease in MFI after the pretreatment in 100 μM (left) and 500 μM (middle) of H₂O₂. C, right: bar graphs showing decrease in Ca²⁺ response after H₂O₂ pretreatment, % at 240 s of time course recorded in the presence of Yoda1 (5 μM) in comparison to control (H₂O₂ 100 μM: 88.5 ± 1.8 %, n = 22, P < 0.001; H₂O₂ 500 μM: 77.3 ± 3.4 %, n = 11, P < 0.01; n is the number of patients, paired sample t-test). D, left, middle: the time-course of the Yoda1-mediated Ca²⁺ responses after 20 min pretreatment in 100 μM (left) and 500 μM (middle) of Ch-T in comparison to control. Note significant decrease in MFI level. D, right: bar graphs showing decrease in Ca²⁺ response after Ch-T pretreatment (Ch-T 100 μM: 92.2 ± 2.4 %, n = 6, P < 0.05; Ch-T 500 μM: 82.3 ± 2.8 %, n = 8, P < 0.001; n is the number of patients, paired sample t-test). E left, middle: the time-course of the Yoda1-mediated Ca²⁺ response after 20 min pretreatment in NAC 1 mM (left) and NAC + H₂O₂ 500 μM (middle) in comparison to control. E, right: bar graphs showing the level of inhibition (%) of Yoda1-mediated Ca²⁺ responses (H₂O₂ 500 μM: 22.7 ± 3.4 %, n = 11, P < 0.01; NAC 1 mM: 4.5 ± 2.0 %, n = 10, P > 0.05; NAC 1 mM + H₂O₂ 500 μM: 8.6 ± 2.6 %, n = 9, P < 0.05; n is the number of patients, paired sample t-test). Note, that NAC decreases the effect of H₂O₂ (middle) but did not fully prevent inhibition. Statistical analysis between bars was performed using MW test.

Fig. 2

H₂O₂ and Chloramine-T decrease the activity of Piezo1in HEK293T cells: A: schematic presentation of Ca²⁺ imaging protocol with an example of paired Ca²⁺ transients after the activation of Piezo1 channel by Yoda1. B: scheme for the patch clamp set-up. C: representative Ca²⁺ transients with paired-pulse applications of 5 μM Yoda1. Ca²⁺ transients without (left) and after 10 min exposure to H₂O₂ (middle, 500 μM). Note the decrease in the second Yoda1-mediated response after the application of H₂O₂ 500 μM compared to control traces. C, right: corresponding Ca²⁺ ratio during the presence of redox agent (control: 43.2 ± 1.1 %, n = 16, total cell count is 466; H₂O₂ 100 μM: 40.6 ± 1.3 %, n = 7, P > 0.05, total cell count is 213; H₂O₂ 500 μM: 23.8 ± 1.7 %, n = 10, P < 0.0001, total cell count is 213, n is the number of coverslips, MW test). D: representative Ca²⁺ transients after 10 min exposure to reducing agents in the presence of H₂O₂ (left: NAC 1 mM + H₂O₂ 500 μM; middle: DTT 1 mM + H₂O₂ 500 μM). D, right: corresponding Ca²⁺ ratio during the presence of reducing agents with H₂O₂ (NAC + H₂O₂: 47.8 ± 1.3 %, n = 6, P < 0.05, total cell count is 167; DTT + H₂O₂: 48.3 ± 1.5 %, n = 8, P < 0.05, total cell count is 124, n is the number of coverslips, MW test). E, left: representative Ca²⁺ transients without (left) and after 10 min exposure to Ch-T (middle, 100 μM). Note the decrease in the second Yoda1-mediated response after the application of 100 μM Ch-T. E, right: corresponding Ca²⁺ ratio during the presence of redox agent (Ch-T 10 μM: 44.2 ± 3.3 %, n = 7, P > 0.05, total cell count is 227; Ch-T 100 μM: 7.6 ± 1.5 %, n = 10, P < 0.0001, total cell count is 222, n is the number of coverslips, MW test). F: representative Ca²⁺ transients after 10 min exposure to reducing agents in the presence of Ch-T (left: NAC 1 mM + Ch-T 100 μM; middle: DTT 1 mM + Ch-T 100 μM). E, right: corresponding Ca²⁺ ratio in the presence of reducing agents and Ch-T (NAC + Ch-T: 49.9 ± 1.9 %, n = 6, P < 0.05, total cell count is 168; DTT + Ch-T: 49.02 ± 1.5 %, n = 6, P < 0.05, total cell count is 335, n is the number of coverslips, MW test). Note the ability of both the reducing agents to antagonize the impact of oxidants exposure on Piezo1 activity, presented in C, E. G. Yoda1-activated transmembrane current through Piezo1 channel in control and after exposure to H₂O₂ using patch clamp paired-pulse protocol. Note decrease in the current ratio after the exposure to H₂O₂ compared to control (control: 49.52 ± 3.8 %, n = 7; H₂O₂ 500 μM: 11.50 ± 2.6 %, n = 10, P < 0.001, n is the number of cells, MW test).

Fig. 3

Activation of Piezo1 channels lead to endogenous generation of ROS in Piezo1HEK293T cells: A: example traces of HyPer7 fluorescence reporting generation of endogenous H₂O₂ in a fraction of cells after application of Yoda1. Notice comparable HyPer7 response induced by application of 1 mM H₂O₂. Pie diagram shows that 18 % (118 out from 644) of all cells responded after activation of Piezo1 by Yoda1. B: corresponding bar graphs showing changes in the HyPer7 fluorescence after application of Yoda1 and H₂O₂ (control: 0.21 ± 0.03 %, n = 210; non-responder Yoda1: 0.3 ± 0.1 %, n = 511; responder Yoda1: 6.52 ± 0.5 %, P < 0.0001, n = 118; H2O2 1 mM: 6.16 ± 0.6 %, n = 122; P < 0.0001, MW test, n number of cells). C, upper: example traces showing no changes in HyPer7 fluorescence in Ca2+-free solution in Piezo1HEK293T cells and similar lack of responses in non-transfected HEK293T and corresponding pie diagrams.

Fig. 4

Microaspiration experiments with RBCs: A. scheme showing the process of RBC aspiration by glass pipette and RBC's fluorescence after mechanical stimulation. B: bar graphs showing the percentage of responded cells during mechanical stimulation in control and under the application of oxidants. Note dose dependant decrease in the number of responded cells with increasing concentration of oxidants. C. Fluorescent Ca²⁺ signals (F-F_o/F_o) corresponding to channel activation. C, left: experiments with H₂O₂ (control 0.16 ± 0.01 %, n = 92; H₂O₂ 10 μM 0.15 ± 0.01 %, n = 88, P > 0.05; H₂O₂ 100 μM 0.12 ± 0.007 %, n = 71, P < 0.05, MW test, n is number of cells). C, right: experiments with Ch-T (control 0.25 ± 0.01 %, n = 156; Ch-T 100 μM 0.18 ± 0.02 %, n = 99, P < 0.001; Ch-T 500 μM: 0.09 ± 0.004 %, n = 65, P < 0.001, MW test, n is number of cells).

Fig. 5

Piezo1 human AF-model: A, left: cysteines in green from outside part of the channel (left) and cysteines in sideview of the channel (channel blades are hidden). A, right: methionines in orange from outside part of the channel (left) and methionines in sideview of the channel (channel blades are hidden). Note abandon distribution of methionines throughout the whole channel but cysteines mostly from outside membrane side. B:methionines M2467 from the intracellular part of the channel in oxidized state (left) and non-oxidized state (right). Note, decrease in the pore size when methionines are oxidized.