PIEZO ion channel is required for root mechanotransduction in Arabidopsis thaliana

Abstract

Plant roots adapt to the mechanical constraints of the soil to grow and absorb water and nutrients. As in animal species, mechanosensitive ion channels in plants are proposed to transduce external mechanical forces into biological signals. However, the identity of these plant root ion channels remains unknown. Here, we show that Arabidopsis thaliana PIEZO1 (PZO1) has preserved the function of its animal relatives and acts as an ion channel. We present evidence that plant PIEZO1 is expressed in the columella and lateral root cap cells of the root tip, which are known to experience robust mechanical strain during root growth. Deleting PZO1 from the whole plant significantly reduced the ability of its roots to penetrate denser barriers compared to wild-type plants. pzo1 mutant root tips exhibited diminished calcium transients in response to mechanical stimulation, supporting a role of PZO1 in root mechanotransduction. Finally, a chimeric PZO1 channel that includes the C-terminal half of PZO1 containing the putative pore region was functional and mechanosensitive when expressed in naive mammalian cells. Collectively, our data suggest that Arabidopsis PIEZO1 plays an important role in root mechanotransduction and establish PIEZOs as physiologically relevant mechanosensitive ion channels across animal and plant kingdoms.

Keywords: Arabidopsis; PIEZO; ion channels; mechanosensation; root.

Fig. 1.

Expression pattern of PZO1::GUSPlus reporter line in Arabidopsis root. (A) Expression pattern of the GUS reporter protein under 2,000 bp of the PZO1 promoter in a 7-d-old seedling. (B and C) Expression in the upper root and root tip when the plant is grown on the surface of standard MS media. Red arrowhead indicates that PZO1 is no longer expressed in the oldest root cap cells that are most distal and are known to be sloughed off. (D and E) Expression in the upper root and root tip grown inside the MS media. Black arrows indicate the cross-section of root displayed in G and H. (F) qRT-PCR for PZO1 in upper root of plants grown on the surface (“surface”) or within the MS media (“inside”). **P < 0.01, n = 4 (mean ± SD). (G and H) Cross-sections of the root cap in GUS reporter lines, which indicate expression in columella and LRC cells. The GUS staining images are representative from three different transgenic lines and four to five independent seedlings.

Fig. 2.

Penetration of roots into hard media is compromised in pzo1 mutants. (A) Root length of Arabidopsis roots grown 9 d vertically at 60° to stimulate growth into the MS media. Black arrowheads indicate roots within the MS media and white arrowheads indicate root growth on the surface of MS media. (B) Root length of WT and pzo1 mutants 9 d after germination (n = 30 to 36; mean ± SD). (C) Representative plate of 18-d-old Arabidopsis seedlings challenged by root barriers. The more visible roots have not penetrated the barrier and are on the surface of the MS media. WT and mutants were grown vertically at 90° on normal MS media (0.5× MS + 8.5 g/L agar) before reaching the barrier (0.5× MS + 12 g/L agar). (D) The percentage of roots that penetrate barriers of different concentrations of agar as indicated (n = 8 to 11 plates) and each plate consists of 7 to 12 seedlings. (E) Representative plate of 2-wk-old Arabidopsis seedlings challenged by root barriers indicating the different root growth length in the denser media (imaged brightness was adjusted for clarity). White arrowheads indicate roots within the MS media and black arrowheads indicate root growth on the surface of MS media or roots that did not penetrate the barrier. (F) Root lengths of 2-wk-old seedlings that have penetrated the barriers (indicated) (n = 25 to 28, mean ± SD). The results were analyzed with two-way ANOVA and Tukey’s multiple comparison test. **P < 0.01, ***P < 0.001.

Fig. 3.

A chimeric channel that includes putative PZO1 pore sequences is activated by negative pressure applied to cell-attached patches. (A) The region of PZO1 in the CH is highlighted in the mPiezo1 structure; gray is mPiezo1 sequence (577 to 1,185 aa) and red is PZO1 sequence (1,228 to 2,485 aa). The first 577 aa of mPiezo1 are not resolved in the structure. (B) Stimulus-response curves are shown for all mPiezo1 (black) and mPiezo1/PZO1 CH (red) stretch-activated currents (SACs, V_pipette = +80 mV in cell-attached configuration). Small high threshold responses are occasionally observed in HEK P1KO cells transiently transfected with the empty IRES-GFP vector control (blue). The maximal current observed from vector-transfected cells was −5.4 pA (dotted line) and this value is used as a cutoff for identifying mPiezo1- and CH-mediated SAC. (C) I_max is shown for mPiezo1 (black), CH (red), and control cells (blue). (D and E) SACs recorded in a patch from cells expressing either mPiezo1 (D) or CH (E); current amplitudes increase with increasing negative pressure (shown below each family of currents). (F) The negative pressure (mmHg) at which the first response to stretch is observed (threshold) when patches are challenged with −5 mmHg increments is plotted. (G) The pressure producing half-maximal currents (P₅₀, determined using GraphPad Prism) is shown. (H and I) A stretch stimulus eliciting a submaximal response is applied during a voltage ramp protocol in order to record SAC currents in WT mPiezo1- (H) and CH- (I) expressing cells between ±60 mV and determine the apparent reversal potential (V_rev; shown by arrows). ns, no statistical significance.

Fig. 4.

PZO1 and calcium influx in response to mechanical indentation to the Arabidopsis root tip. (A) PZO1 expression was knocked down in the root cap using an artificial microRNA under the PIN3 promoter (PIN3::amiR-PZO1) such that PZO1 is specifically knocked down in columella cells (CCs) (blue) but PZO1 is still expressed in lateral root cap (LRC) cells (red). Black dashed circle indicates the area of stimulation by the blunt-tip pipette and the ROI for Ca²⁺ signal intensity measurement. (B) Displacements of the root cap by the stimulating pipette. Mechanical stimulation was initiated 30 to 35 s after GCaMP3 signal recording; a ramp (1 µm/ms) and hold (150 ms) stimulus was applied in 20-µm increments every 15 s. (C and D) Representative image of the GCaMP3 Ca²⁺ response in a WT root tip before (C) and after (D) a 80-µm mechanical stimulation (images are from Movie S2). (C and F) Representative image of the Ca²⁺ responses in PZO1 knockdown PIN3::amiR-PZO1 before and after an 80-µm mechanical stimulation (images are from Movie S4). (G) Ca²⁺ responses in the LRC cells and columella cells in response to mechanical stimulation in 7-d-old seedlings; n = 15 (mean ± SD). Arrowheads indicate the mechanical stimulation (in micrometers). (H) Ca²⁺ responses after mechanical stimulation in PZO1 knockdown of PIN3::amiR-PZO1 root cap in 7-d-old seedlings. Ca²⁺ transients are significantly reduced in columella cells compared to LRC cells; n = 14 (mean ± SD). (I) Area under the curve from starting the stimulation by an 80-μm indentation until the GCaMP3 signals returned to baseline (total of 15 s). (J) Maximum peak of fluorescence in WT and PIN3::amiR-PZO1 (pzo1 knockdown) (n = 14 to 15, mean ± SD). **P < 0.01, ***P < 0.001.