Microbial Signatures in The Rodent Eyes With Retinal Dysfunction and Diabetic Retinopathy

Abstract

Purpose: The gut microbiome has been linked to disease pathogenesis through their interaction in metabolic, endocrine, and immune functions. The goal of this study was to determine whether the gut and plasma microbiota could transfer microbes to the retina in type 1 diabetic mice with retinopathy.

Methods: We analyzed the fecal, plasma, whole globe, and retina microbiome in Akita mice and compared with age-matched wild-type (WT) mice using 16S rRNA sequencing and metatranscriptomic analysis. To eliminate the contribution of the ocular surface and plasma microbiome, mice were perfused with sterile saline solution, the whole globes were extracted, and the neural retina was removed under sterile conditions for retinal microbiome.

Results: Our microbiome analysis revealed that Akita mice demonstrated a distinct pattern of microbes within each source: feces, plasma, whole globes, and retina. WT mice and Akita mice experienced transient bacteremia in the plasma and retina. Bacteria were identified in the retina of the Akita mice, specifically Corynebacterium, Pseudomonas, Lactobacillus, Staphylococcus, Enterococcus, and Bacillus. Significantly increased levels of peptidoglycan (0.036 ± 0.001 vs. 0.023 ± 0.002; P < 0.002) and TLR2 (3.47 ± 0.15 vs. 1.99 ± 0.07; P < 0.0001) were observed in the retina of Akita mice compared to WT. Increased IBA+ cells in the retina, reduced a- and b-waves on electroretinography, and increased acellular capillary formation demonstrated the presence of retinopathy in the Akita cohort compared to WT mice.

Conclusions: Together, our findings suggest that transient bacteremia exists in the plasma and retina of both cohorts. The bacteria found in Akita mice are distinct from WT mice and may contribute to development of retinal inflammation and barrier dysfunction in retinopathy.

Figure 1.

Alpha diversity in Akita mice. Alpha diversity violin plots compared the Chao-1 index, observed species, and Heip's evenness between feces (A), plasma (B), globes (C), and retina (D) from Akita and WT mice. Each dot represents individual sample.

Figure 2.

Beta diversity in Akita mice. Beta diversity PLS-DA plots showing phylogenetic community compositions within each genotype between feces (A), plasma (B), globes (C), and retina (D) from Akita and WT mice. Each dot represents individual sample.

Figure 3.

Phylogenic variation and composition in feces, plasma, globes, and retina from Akita and WT mice. Counts per million–normalized counts of MetaPhlAn displayed differential abundances of prominent taxa (A). The ratio of Firmicutes and Bacteroidetes was calculated in feces and plasma, of Akita mice and compared with WT mice (B). Phylogenic differences in globes and retina of Akita and WT mice. For preparation of ocular globes, the 16S rRNA of the ocular surface microbiome was subtracted from analysis of the globe. For the preparation of the retinas, Mice were perfused with sterile saline solution, the ocular surface microbiome was remove using chemical and mechanical washes and then the retinas were dissected and removed under sterile conditions.

Figure 4.

Identification of bacterial species in the feces, plasma, globes and retina in Akita mice. The 16S rRNA sequencing identified top 10 genera in plasma and globes of both cohorts (A). Only six genera were detected in retina of Akita and WT mice (B).

Figure 5.

Identification of functional pathways in the globes and retina of Akita mice. The functional pathways influenced by microbial imbalance in the globes of diabetic mice (A). Akita mice appeared to have three defined functional expression profiles: Environmental information processing, nitrogen metabolism, and infectious diseases. In the retina, two different pathways, PST system and protein PsiE, were noted in Akita mice (B).

Figure 6.

Levels of PGN, TLR-2, and inflammation in the retina of Akita mice. The levels of PGN (A) and TLR-2 (B) were significantly higher in the retina of Akita mice compared with the WT cohort. The number of activated Iba-1⁺ cells in the retina of both cohorts were quantitated (C). The expression of VCAM-1 was immunohistochemically detected in retina (D) and VCAM-1 fluorescence was quantified (E). Data are presented as mean ± SEM. Each circle represents n number in the cohorts. Each dot represents individual sample.

Figure 7.

Assessment of visual function and acellular capillary in Akita and WT littermates. Visual function in Akita mice was assessed by ERG and was compared with WT mice. The amplitude of a-waves and b-waves was measured in both scotopic (A-B) and photopic (C-D) conditions after nine months of diabetes. Representative images of acellular capillaries in Akita and WT mice. Red arrows indicate acellular capillaries in the retinas (E). Enumeration of acellular capillaries is summarized and presented as mean number of acellular capillaries/mm³ ± SEM. Each circle represents n number per group (F). Each dot represents individual sample.