TGFβ2 induces a fibrotic phenotype in pTM cells and increases expression and membrane insertion of the TRPV4 channel.

(A-B) Five-day TGFβ2 treatment (1ng/mL) significantly altered expression of TGFβ pathway effectors, cytoskeletal machinery, and canonical fibrotic markers. (C) TGFβ2 treatment significantly increased TRPV4 and PIEZO1 expression, but not TREK1 and TRPC1 expression. Mean ± SEM shown. N = 4 - 8 experiments, each gene tested in 3-7 different pTM strains (See Table 1). Two-tailed one sample t-test of TGFβ2-induced gene expression levels as a percent of control samples. (D) Isolation of membrane proteins from two separate pooled pTM samples suggests TGFβ2 treatment drives increased TRPV4 membrane insertion. N = 2 independent pooled samples, 3 pTM strains were pooled per sample. * P < 0.05, ** P < 0.01.

Donor information for primary human trabecular meshwork (pTM) strains used in this study.

TRPV4-mediated Ca2+ influx is potentiated by five-day TGFβ2 treatment.

(A) Five-day TGFβ2 treatment (1 ng/mL) increased TRPV4 agonist-induced (GSK101, 10 nM) Ca2+ influx in pTM cells compared to serum-free media alone treated cells tested on the same day (N = 5 pTM strains, n = 3 - 5 slides/condition/day, individual data points over mean ± SEM). Two-tailed one sample t-test of TGFβ2-treated cell average GSK101 response as a percent of control samples from the same pTM strain on the same day. (B) Violin plots showing the distribution of GSK101-induced Ca2+ responses for each pTM strain tested in A. Thick dashed line indicates mean, while light dashed line indicates quartiles. (C) Representative traces showing TRPV4 agonist-induced Ca2+ influx (seen as an increase in F340/F380) in pTM (mean ± SEM of 4 representative cells/ group), alongside example Fura-2-loaded pTM cells before (i), during (ii), and after (iii) GSK101 application. Scale bar = 50 µm. ** P < 0.01

TGFΒ2-induced TRPV4 potentiation is not seen at a shorter time period, regardless of treatment strength.

(A) TGFβ2 treatments for 24 hours at 1 ng/mL (N = 6 pTM strains, n = 3 - 5 slides/condition/day) or 5 ng/mL (N = 5 pTM strains, n = 3 - 5 slides/condition/day) did not show potentiation of GSK101-evoked TRPV4 Ca2+ influx (SI Appendix Figure S2) and were significantly lower than cells treated with TGFβ2 for 5d at 1ng/mL (5d TGFβ2 results from Figure 2A). Individual data points over mean ± SEM. One-way ANOVA with Tukey’s multiple comparisons test, statistics for individual 1d treatment groups compared to control groups shown in Figure S1. (B) Representative traces for GSK101 response following 24-hour TGFβ2 treatment, traces show mean ± SEM of 3-4 cells. (C) Average current density in response to GSK101 (24-hour control: n = 11 cells, 24-hour TGFβ2: n=10 cells) shows generally increased current in TGFβ2-treated cells. Data shows mean ± SEM (D - E) Violin plots of individual cell strains shown in A. Thick dashed line indicates mean, while light dashed line indicates quartiles. ** P < 0.01, *** P < 0.001

Effects of TRPV4 inhibition/activation on TGFβ2-induced contraction of TM cells.

(A) Representative longitudinal 24-well plate scans of collagen type I hydrogels seeded with pTM subjected to the different treatments (dashed lines outline size of contracted constructs). (B) Longitudinal quantification of hydrogel construct size compared to the control group at the 0 minute time point. (C) Detailed comparisons between groups at each experimental time point. n = 6 hydrogels/group. One-way ANOVA with Tukey’s multiple comparisons test, data in (B & D) shows individual data points over mean ± S.D. One pTM strain shown: TGFβ2-induced contractility induction, HC-06-mediated rescue of hypercontractility, and GSK101-induced transient (15 min) contraction were consistent across (3/3) pTM strains tested (SI Appendix Figure S3). ** P < 0.01, *** P < 0.001, **** P < 0.0001.

TRPV4 activation is necessary to maintain LV-TGFβ2-induced ocular hypertension.

(A) Intravitreal injection of LV-TGFβ2 (week 1), but not LV-Control, elevates IOP in WT mice (N = 5 eyes/group) as early as one-week post-injection. Injection of TRPV4 antagonist HC-06, but not PBS, produced multiday IOP reduction in LV-TGFβ2 treated eyes. HC-06 and PBS injections did not affect IOP in LV-Control injected eyes. Two-way ANOVA with Bonferroni post-hoc analysis (B) Direct comparison of the results of PBS and HC-06 injections in the eyes shown in A. Two-way ANOVA with Bonferroni post-hoc analysis (C) Intravitreal injection of LV-TGFβ2 in Trpv4-/- mice (N = 6 eyes/group) resulted in only mild OHT; plotted against WT eyes at matching timepoints (3 WT cohorts including the 5 WT eyes shown in A-B, N = 8-15 eyes/group). (D) Statistical comparison of the IOP values shown in C. The IOP in LV-TGFβ2 WT eyes was significantly elevated compared to the LV-TGFβ2 Trpv4-/- eyes from 2 weeks post-injection. LV-Control injected eyes in WT or Trpv4-/- eyes remain close to the baseline value and are not significantly different. Two-way ANOVA with Bonferroni post-hoc analysis. (A, C) shows mean ± SEM. Data in (B, D) shows individual data points over mean ± SEM, * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001

TRPV4 inhibition inhibits nocturnal IOP elevation in control and TGFβ2 overexpressing eyes.

(A-B) Post-LV injection daytime (12-2:00 P.M) and nocturnal (9-10:00 P.M.) IOP compared in WT mice (N = 4-6 eyes/group) before drug treatment. LV-TGFβ2 eyes were elevated at daytime, but nocturnal OHT was not significantly different between LV-Ctrl and LV-TGFβ2 eyes in two separate cohorts of mice measured under isoflurane anesthesia (A) or while awake (B). (C-D) Anesthesia had no significant effect on measured nocturnal IOP. One-way ANOVA with Tukey’s multiple comparisons test. (E-F) PBS-injected eyes did not exhibit changes in daytime or nighttime intraocular pressure; however, HC-06 injection reduced TGFβ2-induced IOP elevations during the day and LV-Ctrl and LV-TGFβ2 nocturnal IOPs (N = 4 Eyes/Group); Two-way ANOVA with Bonferroni post-hoc analysis. Figures show datapoints over mean ± SEM, *P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001

TGFβ2-TRPV4 interactions in TM remodeling and ocular hypertension.

Chronic exposure to TGFβ2 induces upregulation of functional TRPV4 channels alongside the autoinhibitory canonical modulator SMAD7. TRPV4-mediated Ca2+ influx, canonical, and non-canonical TGFβ2 signaling stimulate the Rho/ROCK pathway to augment cytoskeletal contractility, and stimulate ECM release. Actomyosin contractility promotes outflow resistance and drives OHT and underpins a vicious feedforward TRPV4-dependent loop that maintains ocular hypertension. Created with BioRender.com/umbwthh.

Sequences, product size, and reference numbers for PCR Primers used in this study.

No significant difference was seen in TRPV4 or PIEZO1 expression between pTM samples treated with TGFβ2 (1 ng/mL) alone or TGFβ2 + TRPV4 antagonist HC-06 (5µM) for five days.

N=3-4 independent experiments. Within each gene, symbols indicate paired samples. Wilcoxon matched-pairs signed rank test and paired t-test used respectively.

TGFβ2 concentrations of 1 ng/mL (A) and 5 ng/mL (B) did not significantly increase TRPV4-induced calcium influx with respect to control cells.

Individual statistical analysis of experiments shown in Fig. 3 A (1 ng/mL: P = 0.138, 5ng/mL: P = 0.095), one sample t-test.

TRPV4 activation is obligatory for TGFβ2-induced TM cell contractions.

(A, D) Representative longitudinal 24-well plate scans of collagen type I hydrogels seeded with two distinct pTM strains (pTM 1: A-C, pTM 2: D-F) subjected to the different treatments as in Fig. 4. (B, E) Longitudinal quantification of hydrogel construct size. (C, F) Detailed comparisons between groups at each experimental time point (N = 6 experimental replicates/ pTM strain). One-way ANOVA with Tukey multiple comparisons test, data in (B, D) shows individual data points over mean ± SEM * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001.

Detailed view of pTM seeded collagen constructs used in Figure 4 and S3.

High resolution representative image of collagen gels used for contractility experiments without circle around periphery of gel.

Expansion of Fig. 5D.

IOP in LV-TGFβ2-injected eyes was significantly elevated compared to both LV-Ctrl injected WT and Trpv4-/- eyes, as well as LV-TGFβ2-injected Trpv4-/- eyes (N = 6 eyes/condition).

Nocturnal IOP is not significantly affected by LV-TGFβ2 overexpression.

Expanded time series of IOP measured weekly from a second cohort of mouse eyes (Fig. 6B) injected with LV-Ctrl (n = 6 eyes) or LV-TGFβ2 (n = 4 eyes). LV-TGFβ2 resulted in elevated diurnal IOP which gradually approached the IOP seen in nocturnal measurements, but did not further elevate IOP above nocturnal values. In this cohort, both diurnal and nocturnal measurements were made in awake animals.

Uncropped Western blots.