PIEZO1 mechanical insensitivity in generalized lymphatic dysplasia with the potential for pharmacological rescue

Abstract

PIEZO1 variants have been associated with generalized lymphatic dysplasia (GLD) through mechanisms involving reduced PIEZO1 expression. Here, we report variants where the mechanism involves reduced channel mechanical sensitivity. Two of the variants encode amino acid changes in the channel's cap structure (Ile2270Thr and Arg2335Gln), one in the ninth transmembrane helical unit (THU) below the cap (Gly1978Asp) and one in the fifth THU distant from the cap (Glu829Val). Patch-clamp studies of the cap and sub-cap variant channels revealed abolished or reduced channel mechanical sensitivity with the possibility to activate the channels and partly rescue mechanical sensitivity by the small molecule Yoda1. The potency of Yoda1 at the variant channels was less than at the wild-type channel, but chemical synthesis of Yoda1 analogs revealed a molecule with improved potency. The data suggest cases of GLD in which there is decreased channel mechanical sensitivity and the potential to reduce dysfunction pharmacologically.

Keywords: Cell biology; Pharmacology.

Graphical abstract

Figure 1

Pedigrees of GLD families GLD07–09 Affected individuals are indicated with filled circles or squares. PIEZO1 genotypes are indicated for individuals who underwent Sanger sequencing. The wild-type allele of the genotype is indicated by minus sign (−), and plus (+) represents the alternative allele. Arrows indicate the proband. IBD indicates the unconfirmed identity by descent.

Figure 2

Expression but mechanical resistance of I2270T hPIEZO1 (A) Western blot data for untagged and hemagglutinin (HA)-tagged wild-type (WT) hPIEZO1 and its E829V and I2270T variants transiently expressed in modified HEK293 (T-REx-293) cells and immunoblotted with anti-HA antibody (upper) and anti-β-actin antibody (lower, loading control). Data are shown for 3 independent experiments (experiments 1, 2, 3). The numbers below the lower gel are for the anti-HA band intensity divided by the β-actin band intensity normalized to WT-HA (mean ± SD values for E829V-HA and I2270T-HA are 0.59 ± 0.17 and 1.18 ± 0.22, respectively). (B) Left: example ionic currents in outside-out patch recordings from T-REx-293 cells transiently expressing WT or I2270T hPIEZO1 exposed to the 105-mmHg pressure pulse shown schematically at the top. Right: summary data for the types of experiment shown on the left. Amplitude of the peak current is represented as mean ± SD and each independent data point is superimposed (WT n = 11, I2270T n = 4). (C) Left: increase in intracellular Ca²⁺ concentration indicated by increase in the fura-2 fluorescence (F) ratio above baseline (ΔF_ratio) in T-REx-293 cells transiently transfected with WT hPIEZO1, I2270T hPIEZO1, or empty vector (vector control). Cells were stimulated with 1 μM Yoda1. Example data are shown for a single representative 96-well plate experiment (mean ± SEM, N = 4–5 wells each). Right: summary data for experiments of the type shown on the left for the signal measured between 30 and 60 s after Yoda1 application (n = 4 independent experiments). Data are mean ± SD normalized to the respective WT channel data and subtracted for the amplitude in the vector control (vc) group.

Figure 3

Mechanical resistance but pharmacological activation of I2286T mPIEZO1 (A–D and I) Data for outside-out patch recordings from T-REx-293 cells stably transfected with empty vector (vector control), wild-type (WT) mPIEZO1, or I2286T mPIEZO1. The voltage across each membrane patch was −80 mV. (A) Upper: pressure pulse protocol in which a 200-ms pulse was applied to 15 mmHg and then incremented (Δ) every 12 s in steps of 15 mmHg up to a maximum of 105 mmHg. Lower: example ionic currents from patches excised from cells transfected with empty vector (vector control), WT mPIEZO1, or I2286T mPIEZO1. Currents evoked by 75 and 90 mmHg are colored in green and olive, respectively. (B–D) For experiments of the type exemplified in (A), peak current amplitudes plotted against pressure and shown as mean ± SD with individual data points for each experiment superimposed (n = 9 for vector control, n = 15–17 for WT, and n = 7 for I2286T). The smooth curve in (C) is a fitted Boltzmann function with mid-point (P₅₀) at 42.4 mmHg. (E and F) Example data for the increase in intracellular Ca²⁺ concentration indicated by increase in the fura-2 fluorescence (F) ratio above baseline (ΔF_ratio) in T-REx-293 cells stably transfected with empty vector (vector control) (E) or I2286T mPIEZO1 (F). Mean ± SEM and N = 4–5 wells each. (G) Summary mean ± SEM concentration-response data for experiments of the type shown in (F) with a Hill equation fitted to the WT mPIEZO1 data (EC₅₀ 0.24 μM) and data points for I2286T mPIEZO1 joined by straight lines (n = 3–6). (H) As for (G) but I2286T mPIEZO1 only and for long (20 min) exposure to Yoda1. Data are normalized to the response to 3 μM Yoda1. The curve is a fitted Hill equation, generating an EC₅₀ of 0.41 μM (n = 5–6). (I) Left: data for a patch exposed to 75 mmHg pressure pulses in the presence of DMSO (vehicle control) and then 5 μM Yoda1 (+Yoda1). Right: for experiments of the type exemplified on the left, the maximum amplitude of current evoked by the pressure pulse. Mean ± SD with each independent data point superimposed (n = 7). ∗p < 0.05 (paired Student’s t test). (J) Chemical structure of Yoda2b (CHR-1871-032). (K) Similar to the approach of (F) but side-by-side comparison of the effects of Yoda2b and Yoda1 on I2286T mPIEZO1 in the same 96-well plate (mean ± SEM, N = 4–5 wells each). (L) Similar to (G) except using Yoda2b instead of Yoda1 and for I2286T mPIEZO1 only (n = 3). The fitted Hill equation yielded an EC₅₀ of 0.71 μM. The dashed curve is the fitted Hill equation from Figure S6B for WT mPIEZO1 (EC₅₀ 0.14 μM).

Figure 4

Mechanical resistance but pharmacological activation of R2351Q mPIEZO1 (A and B) Data for outside-out patch recordings from T-REx-293 cells stably transfected with R2351Q mPIEZO1. The voltage across each membrane patch was −80 mV. (A) Upper: 200-ms pressure pulse protocol applied to 15 mmHg and then incremented (Δ) every 12 s in steps of 15 mmHg up to a maximum of 105 mmHg. Lower: example ionic currents from a patch excised from a cell transfected with R2351Q mPIEZO1. Currents evoked by 75 and 90 mmHg are colored in green and olive, respectively. (B) For experiments of the type shown in (A), quantification of peak current amplitude plotted against pressure and shown as mean ± SD with individual data points for each experiment superimposed (n = 6 for R2351Q). The dashed curve is the fitted Hill equation from Figure 3C for WT mPIEZO1. (C, D, F, and G) Data for the increase in intracellular Ca²⁺ concentration indicated by the increase in the fura-2 fluorescence (F) ratio above baseline (ΔF_ratio) in T-REx-293 cells stably transfected with R2351Q. (C) Summary concentration-response data with data points joined by straight lines. Mean ± SEM and N = 4–5 wells each (n = 3–6). The dashed curve is the fitted Hill equation from Figure 3G for WT mPIEZO1. (D) As for (C) but with long (20 min) exposure to Yoda1. Data are normalized to the response to 3 μM Yoda1. The curve is a fitted Hill equation, generating an EC₅₀ of 0.42 μM (n = 5–6). (E) Left: data for a patch exposed to 75 mmHg pressure pulses in the presence of DMSO (vehicle control) and then 5 μM Yoda1. Right: for experiments of the type shown on the left, maximum amplitude of current evoked by the pressure pulse. Mean ± SD with each independent data point superimposed (n = 7). ∗p < 0.05 (paired Student’s t test). (F) Side-by-side comparison of the effects of Yoda2b and Yoda1 on R2351Q mPIEZO1 on the same 96-well plate (mean ± SEM, N = 4–5 wells each). (G) Similar to (C) except using Yoda2b instead of Yoda1 (n = 3). The dashed curve is the fitted Hill equation from Figure S6B for WT mPIEZO1.

Figure 5

Reduced mechanical sensitivity and pharmacological activation of G1994D mPIEZO1 (A–C) Data for outside-out patch recordings from T-REx-293 cells stably transfected with G1994D mPIEZO1 (A and B) or WT mPIEZO1 (C). The voltage across each membrane patch was −80 mV. (A) Upper: 200-ms pressure pulse protocol applied to 15 mmHg and then incremented (Δ) every 12 s in steps of 15 mmHg up to a maximum of 105 mmHg. Lower: example ionic currents from a patch excised from a cell transfected with G1994D mPIEZO1. Currents evoked by 75 and 90 mmHg are colored in green and olive, respectively. (B and C) For experiments of the type shown in (A), quantification of peak current amplitude plotted against pressure for G1994D mPIEZO1 (B) and WT mPIEZO1 (C) normalized to the maximum I_peak value and shown as mean ± SD with individual data points for each experiment superimposed (n = 7–8 for G1994D, n = 15–17 for WT). A single Boltzmann function is fitted to the G1994D data, but no P₅₀ is indicated because current saturation did not occur. The fitted Boltzmann function to the WT data had a mid-point (P₅₀) at 49.4 mmHg. (D, E, and G) Data for the increase in intracellular Ca²⁺ concentration indicated by increase in the fura-2 fluorescence (F) ratio above baseline (ΔF_ratio) in T-REx-293 cells stably transfected with G1994D mPIEZO1). (D) Summary concentration-response with data points joined by straight lines. Mean ± SEM and N = 4–5 wells each (n = 3–6). The dashed curve is the fitted Hill equation from Figure 3G for WT mPIEZO1. (E) As for (D) but using long (20 min) exposure to Yoda1. Data are normalized to the response to 3 μM Yoda1. The curve is a fitted Hill equation, generating an EC₅₀ of 0.31 μM (n = 5–6). (F) Left: data for a patch exposed to incrementing pressure pulses from 15 to 105 mmHg pressure pulses in the presence of 5 μM Yoda1. Right: for experiments of the type shown on the left, mean and individual data of the type in (B) but in 5 μM Yoda1. The smooth curve is a fitted Boltzmann function with mid-point (P₅₀) of 64.0 mmHg. The dashed curve is the Boltzmann fit to data for G1994D without Yoda1 (−Yoda1) from (B). Mean ± SD with each independent data point superimposed (n = 9–10). (G) Similar to (D) except using Yoda2b (n = 3). The dashed curve is the fitted Hill equation from Figure S6B for WT mPIEZO1.

Figure 6

Model for PIEZO1 variant effects and the potential for a therapeutic strategy targeted to PIEZO1 Left: physiological lymphatic endothelial cell (LEC) with wild-type sodium and calcium (Na⁺ and Ca²⁺)-permeable PIEZO1 channels (P1_WT) that have active caps that are important for channel activation by mechanical force. The ion fluxes through the channels stimulate physiological activities of the LECs. Middle: lymphedema LEC containing PIEZO1 variants of 3 types: P1_v1 (reduced expression, e.g., E829V), P1_v2 (disruption to the cap, strongly reducing mechanical activation—I2270T or R2335Q), and P1_v3 (disruption to the cap-associated blade, partially reducing mechanical sensitivity—G1978D). Because of these defects, there is less Na⁺ and Ca²⁺ entry and thus reduced physiological activities of the LECs, leading to lymphatic dysfunctions that are seen as pleural and pericardial effusions and other features of GLD. Right: improvement in P1_v2 and P1_v3 functions due to the presence of a Yoda small-molecule agonist (e.g., Yoda2b) that stimulates these channels and has the potential to at least partially restore physiological LEC and lymphatic activities. Left, middle, and right: a functional cap is indicated by green color and a partially functional cap by orange and a loss of functional cap by red. Deeper blue indicates a trapped or disrupted channel in intracellular compartments.