Challenges in the characterization of the ocular surface microbiome

Abstract

Purpose : Since the ocular surface is continuously exposed to the environment, an important role of its microbiome is to protect the underlying tissue against colonization of pathogens. If left unchecked, the ocular surface microbiome (OSM) may itself cause disease by uncontrolled immune responses resulting in inflammatory processes. Due to the low abundance of microbial relative to host DNA, the generation of stable and reproducible sequencing data to characterize the human OSM is challenging.

Methods : To characterize the OSM by whole-metagenome sequencing, swabs of the lower conjunctiva were collected using standard cotton swabs compared to flocked nylon swabs. Total DNA was extracted by either the E.Z.N.A. MicroElute Genomic DNA (Omega) or the QIAamp DNA Microbiome (Qiagen) kit. The latter includes a mechanical lysis and host DNA depletion step. Samples were prescreened for bacterial DNA using RT-qPCR of 16S rRNA genes.

Results : More DNA was isolated using flocked swabs (p=0.0378, n=21). The majority of the identified taxa were bacteria (96.60 ± 1.74%), whereas 1.86 ± 1.38% and 1.54 ± 0.75% (SD) were viruses and eukaryotes, respectively. Using the Qiagen kit, a total of 112 species belonging to 71 genera in eight phyla were found compared to eleven species belonging to nine genera in three phyla isolated with the Omega kit (Fig. 1). The most abundant phyla were Proteobacteria and Actinobacteria, additionally Firmicutes if using the Qiagen kit. Extraction kits output differed in both, microbial composition and resolution (Fig. 2).
Fig. 1: Detected microbial phyla, genera and species using the E.Z.N.A. MicroElute Genomic DNA kit (Omega) vs. the QIAamp DNA Microbiome kit (Qiagen). The same three patients were sampled on consecutive days.
Fig. 2: The microbial composition on phylum (A) and species (B) level. The lower six samples are the same three subjects sampled on consecutive days extracted with either with the Omega (yellow) or Qiagen (green) kit.

Conclusions : Despite low microbial abundance of the OSM, preliminary data showed that robust sequencing data can be generated using flocked swabs to increase the DNA input and without 16S rRNA amplification to eliminate amplification bias and to include viral and fungal in addition to bacterial DNA. Microbial diversity depended on the DNA extraction kit, therefore positive controls with known microbial composition will be included in validation studies.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.