Title of the abstract: Circadian rhythms of retinal microglia in early diabetes

Authors – Varun Pathak, Jasenka Guduric-Fuchs, Hanagh Winter, Cara Crady, Beth Grazier, Tom Friedel, Eleni Beli

Study design: Preclinical study using control and STZ-induced diabetic CXCR1^gfp/+ reporter mice.

Purpose: This study investigated how diabetes affects the circadian and daily rhythms of microglial spatial distribution and morphology in the retina. Microglia are dynamic and exhibit circadian rhythms, but their retinal behaviour in health and disease is poorly understood.

Methods: Retinas from control and diabetic CX3CR1^gfp/+ mice were collected at different times within a 24hr Light/Dark and Dark/Dark cycle. Retinal flat mounts stained with collagenIV underwent microglial soma counts, ramification index, and vessel coverage analysis across layers using Imaris and ImageJ. Statistical significance and circadian rhythmicity were assessed using Two-Way ANOVA and Cosinor analysis (p<0.05).

Results: Under DD conditions, microglia displayed circadian rhythmicity with spatial and morphological changes over 24 hours. In controls, microglia numbers and ramification index peaked in the early morning and dropped at night by 45.1% and 53.8%, respectively. In early diabetes, these rhythms persisted, but soma numbers were reduced by 29.7% and ramification index dropped by 50%, with the deep plexus being mostly affected (54.3%). In LD conditions microglial numbers in controls peaked at midday and during night, while diabetic mice displayed altered rhythms. Furthermore, microglia were least active at night in diabetic mice, with light exaggerating circadian disruptions. Perivascular microglia were also quantified under LD conditions. Controls showed circadian rhythmicity, peaking during the day and dipping at night. Diabetic mice lacked rhythmicity, with perivascular microglia elevated by 43.1% compared to controls at midday and consistently high at all nighttime points.

Discussion: This study highlights significant circadian alterations in microglial activity under diabetic conditions. While rhythmicity persisted under constant darkness (DD), diabetes reduced microglial soma numbers and ramification index, suggesting heightened activation. Under LD cycles diabetic mice exhibited disrupted rhythms and elevated perivascular microglia particularly during the nighttime. These findings underscore heightened circadian dysregulation in diabetes and microglial sensitivity to light exposure.

Conclusions: Retinal microglial dynamics are regulated by an endogenous circadian clock, which remains functional in diabetes. However, light significantly influences microglial behaviour  in diabetes, disrupting circadian rhythmicity and overriding normal dynamics.