Long-term Systemic Pentosan Polysulfate Administration Causes Retinal Function Decrease with Changes in Cell Markers

Abstract

Purpose : Pentosan Polysulfate Sodium (PPS) is prescribed to treat bladder pain or discomfort. Pearce et al. [2018] reported a maculopathy in adults taking this medication. We then showed that systemic treatment with PPS causes retinal function diminution in a common laboratory mouse [Girardot et al. 2019]. Here we report further ex-vivo features of this toxicity.

Methods : 10 male and 10 female 129S2/SvPasCrl mice were given intraperitoneal injections of 0.1X Hanks’ Balanced Salt Solution (vehicle) or PPS (12 PPS, 8 vehicle mice). Injections were given several days weekly for 31 weeks. Concentration of drug was 10 mg/kg. Mice regularly had retinal function assessed by electroretinography (ERG). After sacrifice, one eye was fixed using freeze substitution for sagital paraffin sectioning and histology. The contralateral globe was fixed in z-fix for RPE-flatmounts. Flatmounts were dissected within 1 week of sacrifice and stained for ZO-1 and α-catenin.

Results : At 2.5 months we reproduced findings of decreased ERG a- and b-wave amplitudes in the PPS-treated animals, (p=0.0195 for b-waves, 0.0084 for a-waves beginning at 3.5 months, two-way ANOVA). We observed several discrepancies in post-mortem immunostaining. In transverse sections, arrestin and rhodopsin staining was increased. Staining was also increased for synaptic markers PSD95 and synaptophysin, and PKCα, a marker of bipolar and retinal ganglion cells. En-face RPE flatmounts showed trends of decreased cytosolic and nuclear α-catenin, which has been associated with DNA damage responses [Serebryannyy et al. 2017]. Gender differences were observed and are the subject of ongoing experiments.

Conclusions : PPS treatment has nuanced effects on the retina. Functional decline is accompanied by increases in a variety of cell markers, as well as a trend of differential α-catenin protein translocation. Maybe this is all indicative of damage-induced remodeling (Marc et al. 2003). We are still optimizing our model to simulate a phenomenon that is the result of years, and sometimes decades, of drug treatment. Perhaps cell death and more apparent morphological disruption comes with longer durations of treatment. The link between our work and the clinical observations may become clearer as we continue to improve our model and investigate the underlying biochemical mechanisms.

This is a 2021 ARVO Annual Meeting abstract.