Abstract
Purpose :
To explore differences in gut microbiome between ARM cases and controls in a twin cohort and analyse discordant twins.
Methods :
Fundus photographs from the TwinsUK cohort were graded for signs of early ARM using the Rotterdam Classification.
DNA extraction from faecal samples was performed. Amplification of the 16S ribosomal RNA (rRNA) V4 region was performed, followed by amplicon sequencing using a multiplexed approach on Illumina MiSeq platform. After demultiplexing sample reads paired-ends were merged and de novo chimera removal was performed on the 16S rRNA gene sequencing per sample using UCHIME. Remaining reads were collapsed to de novo operational taxonomic units (OTUs) at 97% identity in QIIME v1.9.0. Greengenes v13_8 was used to assign OTU taxonomy via QIIME.
Samples were rarefied to a depth of 5000 reads prior to diversity and abundance analysis. α-diversity indices (Chao1, Shannon & observed OTUs) and β-diversity metrics (weighted & unweighted UniFrac) were calculated using QIIME default methods. Categories were compared by analysing distance matrices using adonis. Differences in mean proportions of genera between cases and controls were analysed using White’s non-parametric test and application of false discover rate (FDR) correction in STAMP.
Results :
100 early ARM cases and 261 controls were included. 64 twin pairs were discordant for early ARM (39 monozygotic, 25 dizygotic). All participants were female. Mean age(SD) and BMI(SD) of cases vs controls was 67.3(8.9) vs 63.9(9.7) years, and 24.8(5.0) vs 26.0(4.6) kg/m2, respectively.
Weighted UniFrac β-diversity was significantly different between cases and controls for the whole cohort (p<0.05) but not discordant twins (p=0.09). The biggest overall differences were enrichment of genera Succinivibrio and Serratia in cases, and Succiniclasticum in controls (corrected q-values <1e-15). All α-diversity indices were reduced in cases vs controls irrespective of concordance, but did not reach statistical significance (p>0.1).
Conclusions :
Significant differences in gut microbiome β-diversity and differential abundance of genera exist between early ARM cases and controls. These preliminary findings suggest that gut microbiome may be altered in ARM and its modification may be targeted for treatment. Further work is needed to understand the role of gut microbiome in early ARM pathogenesis.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.