Fluorescence Lifetime Imaging Ophthalmoscopy of Mouse Models of Age-related Macular Degeneration

**Background:** Age-related macular degeneration (AMD) is a leading cause of visual impairment, with early metabolic changes in retinal pigment epithelial (RPE) cells and photoreceptors preceding structural damage. Fluorescence lifetime imaging ophthalmoscopy (FLIO) noninvasively measures fluorescence lifetimes (FLT) of retinal intrinsic fluorophores, which can be altered by metabolic changes. This study aimed to investigate FLT changes in two AMD mouse models with different underlying metabolic dysfunctions: ApoE⁻/⁻ mice (impaired lipid transport) and Nrf2⁻/⁻ mice (impaired antioxidant response), compared to wild-type (WT) C57BL/6J mice.

**Methods:** A total of 72 female mice (24 per strain) were studied from 3 to 11 months of age. Monthly examinations included slit lamp, fundoscopy, OCT, and FLIO. FLIO used a 473 nm pulsed laser and detected emission in a short spectral channel (SSC, 498-560 nm) and long spectral channel (LSC, 560-720 nm). Mean fluorescence lifetime (τm) was calculated from biexponential fitting. At euthanasia (6 or 11 months), blood plasma was collected for total antioxidant capacity (TAC) measurement using a colorimetric assay (TEAC), and eyes were enucleated for Oil Red O (ORO) lipid staining of Bruch's membrane, scored 0-3 by masked raters. Statistical analysis used mixed models, Tukey's test, Kruskal-Wallis test, and Mann-Whitney test.

**Key Results:** - **FLIO in SSC:** τm decreased with age in all strains. At 4 months, both AMD models had significantly shorter τm than WT (ApoE⁻/⁻: 1287±60 ps, Nrf2⁻/⁻: 1283±80 ps vs WT: 1373±45 ps; P<0.01). By 11 months, ApoE⁻/⁻ mice showed the shortest τm (1140±71 ps) vs WT (1195±47 ps) and Nrf2⁻/⁻ (1207±85 ps), with significant differences at 9 and 10 months (P<0.05 and P<0.01). - **FLIO in LSC:** τm decreased with age in WT (539±66 ps at 3 months to 407±54 ps at 11 months; P<0.0001). ApoE⁻/⁻ mice showed accelerated shortening from 7 months, while Nrf2⁻/⁻ mice did not shorten as much, leading to significantly longer τm in Nrf2⁻/⁻ vs ApoE⁻/⁻ at multiple time points (e.g., at 11 months: 496±90 ps vs 409±29 ps; P<0.0001). - **OCT:** Cumulative incidence of outer retinal abnormalities was 43% in Nrf2⁻/⁻ and 20% in ApoE⁻/⁻ mice by 11 months; none in WT. The rate of τm decrease in LSC was significantly less in mice with OCT changes than without (P<0.05 at 6 months). - **ORO staining:** At 11 months, ApoE⁻/⁻ mice had significantly stronger staining (mean score 1.6) than WT (0.75, P<0.01) and Nrf2⁻/⁻ (0.83, P<0.05). No significant differences at 6 months. - **TAC:** ApoE⁻/⁻ mice had significantly higher TEAC than WT at both 6 months (16.33±1.74 vs 12.47±1.30 mM, P<0.001) and 11 months (19.03±2.61 vs 14.59±1.83 mM, P<0.001), and also higher than Nrf2⁻/⁻ at 11 months (16.16±2.36 mM, P<0.01).

**Clinical Implications:** The divergent FLT patterns in the two AMD models—shortening in ApoE⁻/⁻ (associated with compensatory hypermetabolism and high TAC) and prolongation in Nrf2⁻/⁻ (associated with oxidative stress and lipofuscin accumulation)—suggest that FLIO can differentiate metabolic states relevant to early AMD. The study provides evidence that FLIO may detect preclinical metabolic changes before structural damage (OCT changes) become apparent. However, direct translation to human AMD is limited by species differences and the short observation period (up to 11 months). Further research is needed to correlate FLT changes with specific metabolic pathways and to validate FLIO as a biomarker for early AMD risk in humans.