Investigation of associations between Piezo1 mechanoreceptor gain-of-function variants and glaucoma-related phenotypes in humans and mice

Abstract

Glaucoma disproportionately affects individuals of African descent. Prior studies of the PIEZO1 mechanoreceptor have suggested a possible role in glaucoma pathophysiology. Here, we investigated associations between a Piezo1 gain-of-function variant common in individuals of African descent with glaucoma-related phenotypes. We analyzed whole genome sequences to identify Piezo1 variants and their frequencies among 1565 human participants. For the most common variant (e756del), we compared phenotypes between heterozygotes, homozygotes, and wildtypes. Longitudinal mixed effects models of visual field mean deviation (MD) and retinal nerve fiber layer (RNFL) thickness were used to evaluate progression. Based on trends in the models, further investigation was conducted using Piezo1 gain-of-function mice. About 30% of African descent individuals had at least one e756del allele. There were trends suggesting e756del was associated with higher IOPs, thinner RNFLs, lower optic nerve head capillary densities, and greater decreases in MD and RNFL thickness over time, but these did not reach statistical significance. Among mice, increased Piezo1 activity was not significantly associated with IOP or retinal ganglion cell density. Our study confirms that the Piezo1 e756del gain-of-function variant is a frequent polymorphism present in African descent individuals but is unrelated to examined differences in glaucoma phenotypes. Ongoing work is needed to elucidate the role of Piezo1-mediated mechanotransduction in glaucoma.

Figure 1

Intraocular pressure (IOP) in Piezo1^GOF mice. IOP was measured via tonometer in anesthetized mice (11 wildtype: Piezo1^+/+, 25 heterozygotes: Piezo1^GOF/+/+, and 11 homozygotes: Piezo1^GOF/GOF). (A) No significant differences were observed between groups (one-way ANOVA with a Tukey post-test) (mean +/− s.d.). (B) The IOP measurements at various ages. A simple linear regression for each genotype is plotted for visualization. No obvious differences in the impact of age were noted.

Figure 2

Ganglion cell density in Piezo1^GOF mice. Retinal ganglion cells were stained with an anti-RBPMS antibody (Millipore: ABN1376). Cell density was calculated in three 150 µm² sample areas in each retina and the average was used per mouse. Example wildtype and homozygous staining is shown (RBPMS = magenta)(right). (A) No significant differences were observed between groups (one-way ANOVA with a Tukey post-test)(mean + /− s.d.). (B) No obvious differences in the impact of age were noted, but increased numbers would be required to make any robust comparisons. A simple linear regression per genotype is shown for visualization.