Class I PI3Ks activate stretch-induced autophagy in trabecular meshwork cells

Abstract

Elevated intraocular pressure (IOP) is the primary risk factor for glaucoma, a leading cause of irreversible blindness worldwide. IOP homeostasis is maintained through a balance between aqueous humor production and its drainage through the trabecular meshwork (TM)/Schlemm's Canal (SC) outflow pathway. Prior studies by our laboratory reported a key role of mechanical forces and primary cilia (PC)-dependent stretch-induced autophagy in IOP homeostasis. However, the precise mechanism regulating this process remains elusive. In this study, we investigated the upstream signaling pathway orchestrating autophagy activation during cyclic mechanical stretch (CMS) in primary cultured human TM cells, using biochemical and cell biological analyses. Our results indicate that TM cells express catalytic subunits of class IA PI3Ks (PIK3CA, B, and D), and that inhibition of class IA isoforms, but not class II and III, significantly prevent CMS-induced autophagy. Importantly, PIK3CA was found to localize in the PC. Furthermore, we identified a coordinated action of Class IA PI3Ks along INPP4A/B, a 4' inositol phosphatase, responsible for the formation of PI(3,4)P₂ and PI(3)P and stretch-induced autophagy in TM cells. These findings contribute to a deeper understanding of the molecular mechanisms underlying IOP homeostasis.

Keywords: Autophagy; Glaucoma; Mechanical stress; PI3K; Primary cilia; Trabecular meshwork.

Fig. 1

Expression profile of PI3Ks in HTM cells. A mRNA expression levels of PI3K catalytic subunits in TM cells. mRNA expression level of each PI3K catalytic subunit were analysed using single-cell RNA sequencing data obtained from outflow pathway cells previously published [30]. The expression (%) in TM cells is represented in the right panel, and expression above 1% is considered significant. Five PI3K catalytic subunits (class IA: PIK3CA, B and D; class II: PIK3C2A; Class III: PIK3C3) are considered significantly expressed in TM cells. B Western blot (WB) analyses of protein expression levels of each PI3K catalytic subunit

Fig. 2

Inhibition of class IA PI3Ks prevents CMS-induced autophagy in HTM cells. A The effect of PIK-75 treatment on LC3 II expression levels during CMS. HTM cells were treated with varying concentrations of PIK-75 (ranging from 20 nM to 2 µM) and subjected to CMS for 24 h. LC3 II expression levels were monitored by WB analysis, with ACTB used as a loading control. Data are presented as the mean ± S.D. (n = 3). *, p < 0.05; **, p < 0.01 (one-way ANOVA with Tukey’s post hoc test). B The effect of triple knockdown of class IA PI3K catalytic subunits on LC3 II expression levels during CMS. HTM cells were transfected with three siRNA against PIK3CA, PIK3CB and PIK3CD together, and incubated for 3 days, followed by CMS for 24 h. LC3 II expression levels were monitored by WB analysis. Band intensities were quantified using Image Lab™ touch software, normalized with ACTB, and fold changes were calculated and graphed. Data are presented as the mean ± S.D. (n = 6). *, p < 0.05; **, p < 0.01 (two-way ANOVA with Tukey’s post hoc test). CNT: siCNT, NS: non-stretch, CMS: cyclic mechanical stretch. C The effect of class I PI3K inhibition on autophagy flux. HTM cells were transduced with 10 pfu/cell of AdtfLC3 for 48 h and subsequently treated with 2 µM PIK-75, followed by CMS for 24 h. DAPI was used to stain nuclei. Autophagosomes (yellow puncta) and autolysosomes (red puncta) were observed using confocal microscopy. The images were processed using Fiji software. Scale bars: 10 µm

Fig. 3

PIK3C2A-mediated PI(3)P formation does not affect CMS-induced autophagy in HTM cells. A The effect of PIK3C2A knockdown on LC3 II expression levels during CMS. HTM cells were transfected with siRNA against PIK3C2A, and incubated for 3 days, followed by CMS for 24 h. LC3 II expression levels were monitored by WB analysis. Band intensities were quantified using Image Lab™ touch software, normalized with ACTB, and fold changes were calculated. Data are presented as the mean ± S.D. (n = 4). ns, not significant; **, p < 0.01 (two-way ANOVA with Tukey’s post hoc test). CNT: siCNT, NS: non-stretch, CMS :cyclic mechanical stretch. B The effect of exogeneous PI(3)P treatment on PC-mediated autophagy during CMS. HTM cells were deciliated by treating with 2 mM CH for 3 days and then treated with 1 µM PI(3)P, followed by CMS for 24 h. PI(4)P is used as negative control. LC3 II expression levels were monitored by WB analysis. Band intensities were quantified using Image Lab™ touch software, normalized with ACTB, and fold changes were calculated and graphed. Data are presented as the mean ± S.D. (n = 3). ns, not significant (one-way ANOVA with Tukey’s post hoc test in each group). CNT: control; DC: deciliated. vehicle: carrier only, NS: non-stretch, CMS: cyclic mechanical stretch

Fig. 4

Localization of PIK3CA on the PC and its dissociation from PC upon CMS in HTM cells. A Immunocytochemistry staining showing the localization of PIK3CA on the PC. After 24 h of CMS, HTM cells were fixed and immunostained with PIK3CA (green) along with Ac-TUBA4A (red), a PC marker. DAPI was used to stain nuclei. Images were acquired with a confocal microscope and processed by using Fiji software. Scale bars: 10 µm. B Quantification of the colocalization of PIK3CA on the PC. The signal intensity of PIK3CA on the PC in each cell was measured using the Cilia Q plugin in Fiji. Data are presented as the mean ± S.D. (n = 55 and 38 for NS and CMS, respectively). *, p < 0.05 (Student’s t-test). NS: non-stretch, CMS: cyclic mechanical stretch

Fig. 5

Triple KD of PIK3CA, B and D decreases the intensity of PI(3)P on the PC in HTM cells. A Immunocytochemistry staining showing the localization of PI(3)P on the PC. HTM cells were treated with siRNA for 3 days followed by 24 h of CMS treatment. Cells were then fixed and immunostained for PI(3)P (red) along with ARL13B (green), a PC marker. Arrows indicate PI(3)P on the PC. Nuclei were stained with DAPI. Images were acquired using a confocal microscope and processed with Fiji software. Scale bars: 10 µm. B Quantification of PI(3)P intensity on the PC. The signal intensity of PI(3)P on the PC in each cell was measured using the Cilia Q plugin in Fiji. Data are presented as the mean ± S.D. (n = 106, 69, 70 and 67for siCNT-NS, siPIK3CABD-NS, siCNT-CMS and siPIK3CABD-CMS, respectively). *, p < 0.05, ****, p < 0.0001, ns, not significant (one-way ANOVA with Tukey’s post hoc test in each group Anova). NS: non-stretch, CMS: cyclic mechanical stretch

Fig. 6

Regulation of autophagy by PI(3,4)P₂-positive vesicles during CMS in HTM cells. A Increase in AKT-PH-GFP positive vesicle by CMS. HTM cells were transfected with the AKT-PH-GFP plasmid and incubated for 48 h, followed by CMS for 24 h. The GFP positive signals were observed using confocal microscopy. DAPI was used to stain nuclei. Curly bracket symbols and arrows indicate AKT-PH-GFP positive signals in plasma membrane and vesicles, respectively. Scale bars: 10 µm. The images were processed, and the number of AKT-PH-GFP positive vesicles per cells were quantified using Fiji software. Data are presented as the mean ± S.D. (n = 13 and 20 for NS and CMS, respectively). ***, p < 0.001 (Student’s t-test). NS: non-stretch, CMS: cyclic mechanical stretch. (B) GFP-AKT-PH positive vesicles partially overlap with clathrin heavy chain 1 (CLTC). HTM cells were transfected with pGFP-AKT-PH and incubated for 3 days, followed by CMS for 24 h. Cells were then immunostained with CLTC. DAPI was used to stain nuclei. Images were acquired with a confocal microscope and processed by using Fiji software. Scale bars: 10 µm. (C) Effect of inhibition of 5’ or 4’ inositol phosphatases on LC3 II expression levels during CMS. HTM cells were transfected with two siRNA against SHIP1/2 or INPP4A/B together, respectively, and incubated for 3 days, followed by CMS for 24 h. LC3 II expression levels were monitored by WB analysis. Band intensities were quantified using Image Lab™ touch software, normalized with ACTB, and fold changes were calculated. Data are presented as the mean ± S.D. (n = 3 and 7 for siSHIP1/2 and siINPP4A/B, respectively). ns not significant; *, p < 0.05; **, p < 0.01 (one-way ANOVA with Tukey’s post hoc test in each group). CNT: siCNT, NS: non-stretch, CMS: cyclic mechanical stretch

Fig. 7

PIPs and autophagosomes biogenesis. HTM cells expressing AKT-PH-GFP (green) were stained with LC3 (magenta) with or without CMS. Note AKT-PH-GFP partially co-localizing (arrowheads within bracket) or in contact with LC3-labeled puncta (arrows). Images were acquired with a confocal microscope and processed by using Fiji software. Scale bars: 10 µm