Inside Out Integrin Activation Mediated by PIEZO1 Signaling in Erythroblasts

Abstract

The non-selective mechanosensitive ion channel PIEZO1 controls erythrocyte volume homeostasis. Different missense gain-of-function mutations in PIEZO1 gene have been identified that cause Hereditary Xerocytosis (HX), a rare autosomal dominant haemolytic anemia. PIEZO1 expression is not limited to erythrocytes and expression levels are significantly higher in erythroid precursors, hinting to a role in erythropoiesis. During erythropoiesis, interactions between erythroblasts, central macrophages, and extracellular matrix within erythroblastic islands are important. Integrin α4β1 and α5β1 present on erythroblasts facilitate such interactions in erythroblastic islands. Here we found that chemical activation of PIEZO1 using Yoda1 leads to increased adhesion to VCAM1 and fibronectin in flowing conditions. Integrin α4, α5, and β1 blocking antibodies prevented this PIEZO1-induced adhesion suggesting inside-out activation of integrin on erythroblasts. Blocking the Ca²⁺ dependent Calpain and PKC pathways by using specific inhibitors also blocked increased erythroid adhesion to VCAM1 and fibronectins. Cleavage of Talin was observed as a result of Calpain and PKC activity. In conclusion, PIEZO1 activation results in inside-out integrin activation, facilitated by calcium-dependent activation of PKC and Calpain. The data introduces novel concepts in Ca²⁺ signaling during erythropoiesis with ramification on erythroblastic island homeostasis in health and disease like Hereditary Xerocytosis.

Keywords: PIEZO1; calcium signaling; erythroblastic island; integrin α4β1; integrins.

FIGURE 1

Erythroblasts integrin expression. (A) Integrin mRNA expression (count per million) in CD71+/CD235a+ erythroblasts; data mined from Heshusius et al. (2019). Gene expression was found for 11 integrin subunits: six α sub-units (ITGA) and five β sub-units (ITGB; n = 4). (B) Dot plot showing a representative CD71/CD235 expression plot depicting the erythroblast stage of analysis. Right histograms and bar graph indicate expression of α4 integrin and β1 integrin in CD71+CD235a+ erythroblasts in 3 donors as measured by flow cytometry.

FIGURE 2

Increased erythroblast binding to VCAM-1 and fibronectin after Yoda1 mediated PIEZO1 stimulation. (A,B) Erythroblasts were treated or not with 1 μM of Yoda1 for 10 min and subjected to a flow assay as described in Materials and Methods (7.5 ml/h flow rate). Adhesion of erythroblasts to VCAM (A) or fibronectin (B) was quantified using imageJ (n = 9 pictures of different donors; data depicted as mean ± SD; ^{* * **}P < 0.0001; C,D) Erythroblasts were pre-treated with anti-α4, anti-α5, or anti-β1 and stimulated or not with 1 μM Yoda1 (10 min) as indicated and flowed over a surface coated with VCAM (C) or fibronectin (D; data depicted as mean ± SD mean n = 9 images was used to calculate statistic; ^∗P < 0.05, ^∗∗P < 0.01, ^∗∗∗P < 0.001, and ^{* * **}P < 0.0001).

FIGURE 3

PIEZO1 stimulation with Yoda1 results calpain and PKC dependent integrin activation. (A) In untreated and Yoda1 treated erythroblasts, the calpain activity assay measured the absorbance of cleaved calpain substrate at 505 nm, which is quantified using a fluorescence plate reader. Active calpain and inhibited calpain where used as positive and negative controls, respectively (n = 3, ^∗∗P < 0.01). (B) Expression of Talin (±225 kDa) and the cleavage of Talin (±190 kDa; indicated by arrow) in untreated erythroblasts and erythroblasts incubated with 1 μM Yoda1 for 10, 30, and 60 min. The same amount of protein was loaded and Tubulin (±50 kDa) was used as loading control (representative of n = 2). (C,D) Erythroblasts were pretreated with PKC inhibitor (500 nM) and calpain inhibitor (1 mM) and stimulated with and without 1 μM Yoda1 (10 min) as indicated and flowed over a surface coated with VCAM (C) or fibronectin (D; n = 2 donors; data depicted as mean ± SD; mean of n = 18 images (of two different donors) was used to calculate statistic ^{* * **}P < 0.0001).