Novel mechanisms of PIEZO1 dysfunction in hereditary xerocytosis

Abstract

Mutations in PIEZO1 are the primary cause of hereditary xerocytosis, a clinically heterogeneous, dominantly inherited disorder of erythrocyte dehydration. We used next-generation sequencing-based techniques to identify PIEZO1 mutations in individuals from 9 kindreds referred with suspected hereditary xerocytosis (HX) and/or undiagnosed congenital hemolytic anemia. Mutations were primarily found in the highly conserved, COOH-terminal pore-region domain. Several mutations were novel and demonstrated ethnic specificity. We characterized these mutations using genomic-, bioinformatic-, cell biology-, and physiology-based functional assays. For these studies, we created a novel, cell-based in vivo system for study of wild-type and variant PIEZO1 membrane protein expression, trafficking, and electrophysiology in a rigorous manner. Previous reports have indicated HX-associated PIEZO1 variants exhibit a partial gain-of-function phenotype with generation of mechanically activated currents that inactivate more slowly than wild type, indicating that increased cation permeability may lead to dehydration of PIEZO1-mutant HX erythrocytes. In addition to delayed channel inactivation, we found additional alterations in mutant PIEZO1 channel kinetics, differences in response to osmotic stress, and altered membrane protein trafficking, predicting variant alleles that worsen or ameliorate erythrocyte hydration. These results extend the genetic heterogeneity observed in HX and indicate that various pathophysiologic mechanisms contribute to the HX phenotype.

Figure 1.

HX-associated PIEZO1 mutations. (A) The position of the 6 PIEZO1 mutations identified in this report and the first 2 HX-associated PIEZO1 mutations described, R2456H and M2225R, are shown on a model of PIEZO1 derived from the cryoelectron microscopy structure. The locations of the channel anchor, the peripheral helices (PH), the inner helices (IH), and the outer helices (OH) are shown. (B) Conservation of HX-associated mutant amino acids. The 4 amino acids located in the COOH-terminal domain containing the putative PIEZO1 channel exhibit high degrees of conservation.

Figure 2.

Time course and IF analysis of wild-type PIEZO1 expression in HEK293 cells. (A) Time-course experiments in an HEK293 cells doxycycline-inducible model of PIEZO1 demonstrated significant protein expression beginning at 16 hours and lasting for 96 hours postinduction. The time frame between 24 and 48 hours postinduction was used for all experiments unless otherwise indicated. (B) IF studies localized PIEZO1 expression to the cell membrane after doxycycline induction. 4′,6-Diamidino-2-phenylindole (DAPI) stain of nuclei (blue), phalloidin AF488 stain of F-actin (green), and ATT0647N FLAG-tagged PIEZO1 (red) images in uninduced and induced cells are shown. Scale bars, 10 µM. IAB, primary antibody, mouse anti-FLAG tag ATTO647N; IIAB, secondary antibody, goat anti-mouse; In, induced; Un, uninduced.

Figure 3.

MA currents in HEK293 cells expressing wild-type or mutant PIEZO1. (A) Representative traces of MA inward currents from uninduced (Un) or induced (Ind) stably transfected HEK293 cells expressing wild-type (WT) PIEZO1. Using a patch-clamp electrophysiological technique, functionality of recombinant PIEZO1 channel activity was examined in the whole-cell configuration. Upon a series of mechanical stimulations of a glass probe (1-µm increments, 150-ms step duration), currents were recorded in the whole-cell patch-clamp configuration at holding potential −80 mV. The stimulus waveforms are shown above the current traces. The red lines are MA currents fitted with monoexponential function. (B) Average of inactivation time constant (τ, milliseconds) for wild-type, HX-associated mutant, and known HX-associated mutant (R2456H and M2225R) controls. *P < .05 ***P < .001 (Student t test). (C) Threshold of activation. Several of the mutants studied displayed greater sensitivity to mechanical stimulus as indicated by minimal indentation activation threshold. ***P < .001 (Student t test). n.s., not significant.

Figure 4.

Altered channel kinetics exhibited by PIEZO1 mutants. MA currents in HEK293 cells expressing wild-type or mutant PIEZO1 were examined using a whole-cell patch-clamp technique. (A) K1877del mutant. Left panel, representative traces of peak current amplitude of MA wild-type and PIEZO1-K1877del cells. Both faster channel kinetics and smaller peak current amplitude were observed for the K1877del mutant. Right panel, a comparison of averaged peak currents amplitude recorded at 6 μm from wild-type and K1877del mutant cells, with the K1877del mutant showing a significant decrease of peak current. (B) E2496ELE mutant. Representative traces of MA currents in wild-type (left) and PIEZO1-E2496ELE mutant (right) cells, as a function of depth of the indenting probe. The stimulus waveforms used in activation protocol are shown above the traces with the colors corresponding to each depth of indentation. The E2496ELE insertion mutant displayed increased inactivation rate compared with wild type. In addition, whereas wild-type channel inactivated completely, in 77% of records from the mutant, persistent mechanocurrent was observed, indicating incomplete channel closure. Bottom panel, a current recorded from an E2496ELE mutant at 11 µm displaying increased steady-state availability of the channel shown by the interruption in inactivating current. For presentation, a zoomed in view of the interrupted region is shown with the current peak cutoff.

Figure 5.

Osmotic stress and PIEZO1 activity. The sensitivity of PIEZO1 expressed in HEK293 cells to different environmental osmolality was analyzed. (A) Changes in cell diameter were measured after solutions of different tonicity were applied to the cell surface; n = 61/Osm, ***P < .001. (B-C) MA currents applied at 6 µm were recorded in the whole-cell patch-clamp configuration, at holding potential −80 mV in HEK293 cells expressing wild-type PIEZO1 (uninduced and induced). Recordings were obtained from isotonic (Iso; 300 mOsm) physiological solution, after perfusion with hypotonic (Hypo; 200 mOsm) solution, and after washout with isotonic (Iso; 300 mOsm) solution. The same experiments were performed only with isotonic (300 mOsm) physiological solution as a control. There was a dramatic increase in the average peak current of MA currents in hypotonic conditions (P < .005, n = 6). To determine the involvement of PIEZO1 on recorded current responses, 30 µM of the inhibitor gadolinium were added to hypotonic solution. The same experiments were performed only with isosmotic (300 mOsm) physiological solution, as a control. A typical response after the addition of gadolinium is shown. (D) MA currents were similarly measured in HEK293 cells expressing HX-associated mutant PIEZO1 (uninduced and induced), with recordings obtained at isotonic (300 mOsm) and hypotonic (200 mOsm) physiological solutions (Table 3).

Figure 6.

Cellular localization of wild-type and mutant PIEZO1 determined by IF confocal microscopy. (A) Wild-type and mutant FLAG-tagged PIEZO1-expressing HEK293 cells were transfected with ER-GFP, a plasmid that expresses GFP fused to the ER signal sequence of calreticulin and KDEL (ER retention signal). Cells were stained with anti-FLAG AF555 (red) and anti-GFP (green). (B) Wild-type and mutant FLAG-tagged PIEZO1-expressing HEK293 cells were stained with anti-FLAG ATT0647N (red), DAPI (blue), and an antibody against Golgi apparatus, anti-GM130, conjugated with Alexa Fluor 488 (green). Three arginine mutants, R2488Q, R2302H, and R2088G, demonstrated partial intracellular retention, colocalizing with ER and Golgi markers. Scale bars, 10 μm.

Figure 7.

Studies of the E2496ELE PIEZO1 mutant. (A) HEK293 cells expressing the E2496ELE mutant displayed morphological changes upon induction. Top panels: compared with wild type, the mutant cells demonstrated swelling as early as 16 hours postinduction with rounding up of edges, followed by wide-scale detachment of cells from the plate by 48 hours. Bottom panels: none of the known PIEZO1 inhibitors, including the G spatulata mechanotoxin (GsMTx4), gadolinium, or ruthenium red, protected cells from this effect. Original magnification ×100. (B) Live cell numbers and annexin V staining. Beginning at ∼48 hours postinduction, the number of live, mutant cell numbers determined by CCK8 staining decreased dramatically (left graph) and cells began to undergo apoptosis as assayed by annexin V staining (right). Camptothecin (CPT) was used for artificial induction of apoptosis in wild-type cells. Scale bars, 1 μm. FITC, fluorescein isothiocyanate.