Ocular surface microbiomes of children and adults

Abstract

Purpose : The normal conjunctival surface contains few viable microbes, with roughly 10 colony forming units from a single swab, and 1,000 genomes of bacterial DNA. Given the different spectrum of ocular surface disease observed in children and adults, it is of importance to know how the population of bacterial species on the human conjunctiva changes with age. To address this question, we are using Next Generation DNA sequencing of bacterial 16S rRNA genes to compare the microbial populations of young and old eyes.

Methods : Patients were recruited with informed consent, and swab samples were collected separately from inferior conjunctiva and peri-ocular skin using sterile flocked nylon microfiber swabs wetted with eye drop solution. These were frozen or immersed in ethanol. Because of microbial DNA contamination in commercial kits, we developed a low background mechanical disruption procedure to extract DNA from the samples. PCR pre-amplification of 16S rRNA genes minimized effects of bacterial contaminants in reagents. Bar-coded libraries of samples and controls were run on an Illumina MiSeq sequencer, followed by metagenomic analysis of 16S rRNA genes. Conjunctival cells in each swab were quantified by PCR with primers designed to amplify an invariant Alu repeat class.

Results : Conjunctival swab samples were collected from 195 adult patients with a mean age of 72 years (SD 17.1, range 19 to 96). These were either in good ocular health, undergoing treatment for age related macular degeneration, or affected with diabetic retinopathy. Samples have also been collected from 133 healthy children at the pre-established well child visits. 16S rRNA gene sequence analysis of these is in progress. Initial data from adult, healthy eyes has identified Staphylococcus epidermidis as the most abundant single bacterial species, followed by Corynebacteria and Propionibacter species. Due to our technical innovations, the background bacterial contamination is less than 5% of all sequence reads. Alu-PCR revealed up to 20-fold differences in conjunctival cell yield, and has allowed us to normalize data to epithelial cell yield.

Conclusions : We have developed new techniques to address the challenges involved in analyzing DNA from small numbers of bacterial cells. In addition, we have obtained samples from adult and pediatric patients. These are currently under analysis to determine how the size and species distribution of their bacterial population varies with age.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.