hin et al A Enterobacteriaceae: g f Comamonadaceae: g Methylobacterium g Non-lens wearers Lens wearers Conjunctiva Conjunctiva Contact lens n=52 P Proteobacteria: F. Firmicutes: C, Cyanobacteria: A, Actinobacteria Conjunctiva Non-lens wearers oniunctia vs Contact ans Pseudomonas 旦_ Haemophilus g Streptococcus Sphingobium FIG 4 Differences in relative abundances of phylotypes in ocular microbiota between non-lens wearers and lens wearers. Each phylotype(>1% of average relative abundance in groups)is indicated by a different color at the genus level. (A)Bacterial taxon plots at the genus level. (B)Histogram of unique biomarker bacteria in each group. The LDA effect size(>3.0-fold)was used to detect unique biomarkers. tive effect shown by microbiota against invasions of foreign bac- phyta spp (chloroplast DNA, possibly an environmental contami- teria, has been suggested for the eye(7). This would imply that nant) showed a nonsignificant tendency(nonparametric t test; P some impacts on the eye microbiota, such as those related to wear- 0.26)to be higher in the conjunctiva of lens wearers than in that of ing contacts, might decrease the resilience or self-restoring capac- non-lens wearers(Fig 4; see also Fig. S5 in the supplemental mate ity of the ecosystem. Although the data represent only 3 time rial). We would not have detected differences in the conjunctival mi- points, we found relatively high stability of the eye microbiome. crobiota between lens wearers and non-lens wearers in this study if we The healthy eye is constantly wetted by tears, which may provide a hadincluded samples collected from the ocular surface after the use of constant chemical environment topical proparacaine hydrochloride anesthetic, highlighting the im- The bacterial diversity of the conjunctiva is higher than that in portance of sampling microbiome sites with minimal perturbations. the skin. Moreover, conjunctiva had an alpha diversity similar to The anesthetic may have diluted or washed away bacteria from the that of the oral microbiota(see Fig S3 in the supplemental mate- ocular surface rial). These results are remarkable considering the antimicrobial By comparing the microbiota of ocular conjunctiva to that of effects of tear compound skin under the eye, we found that normal Corynebacterium and This study showed that the eye microbiota of lens wearers is dif- Staphylococcus skin bacteria were detected at higher relative abun ferent from that of non-wearers, resembling more closely the micro- dance in the eye than in the skin, consistent with a recent study biota of the skin. The results are consistent with those previously showing the distribution of skin microbiome by body sites(16) found with culture-dependent methods(20-22). Consistently, the Tear lactoferrin plays a role in reducing levels of Staphylococcus presence of Staphylococcus, Propionibacterium, Corynebacterium, Ba- epidermidis biofilms(23). However, in this study, we found that illus, Micrococcus, Rothia(previously assigned to Stomatococcus), and Staphylococcus levels were lower in lens wearers than in non-lens Pseudomonas in cultures from contact lenses has been reported(1). wearers. There are not many reports of studies using molecular We also detected members of other taxa in contact lenses, such as methods in the diseased eye. Lee et al. (6)studied blepharitis using families Oxalobacteraceae and Enterobacteriaceae at a relative abun- and platforms(Vi to V3 16S rRNA genes with Roche-454). 9 Streptococcus, Methylobacterium, and Acinetobacter, and members of different sampling sites(eyelashes and tears), sequencing region dance of >1% in more than half of the samples. In addition, Strepto- results showed that blepharitis patients(n=7) had increased 4 mBio mbio. asm. org March/April 2016 Volume 7 Issue 2 e00198-16tive effect shown by microbiota against invasions of foreign bac￾teria, has been suggested for the eye (7). This would imply that some impacts on the eye microbiota, such as those related to wear￾ing contacts, might decrease the resilience or self-restoring capac￾ity of the ecosystem. Although the data represent only 3 time points, we found relatively high stability of the eye microbiome. The healthy eye is constantly wetted by tears, which may provide a constant chemical environment. The bacterial diversity of the conjunctiva is higher than that in the skin. Moreover, conjunctiva had an alpha diversity similar to that of the oral microbiota (see Fig. S3 in the supplemental mate￾rial). These results are remarkable considering the antimicrobial effects of tear compounds (19). This study showed that the eye microbiota of lens wearers is dif￾ferent from that of non-wearers, resembling more closely the micro￾biota of the skin. The results are consistent with those previously found with culture-dependent methods (20–22). Consistently, the presence of Staphylococcus, Propionibacterium, Corynebacterium, Ba￾cillus,Micrococcus,Rothia (previously assigned toStomatococcus), and Pseudomonas in cultures from contact lenses has been reported (1). We also detected members of other taxa in contact lenses, such as Streptococcus, Methylobacterium, and Acinetobacter, and members of families Oxalobacteraceae and Enterobacteriaceae at a relative abun￾dance of 1% in more than half of the samples. In addition, Strepto￾phyta spp. (chloroplast DNA, possibly an environmental contami￾nant) showed a nonsignificant tendency (nonparametric t test; P 0.26) to be higher in the conjunctiva of lens wearers than in that of non-lens wearers (Fig. 4; see also Fig. S5 in the supplemental mate￾rial). We would not have detected differences in the conjunctival mi￾crobiota between lens wearers and non-lens wearers in this study if we had included samples collected from the ocular surface after the use of topical proparacaine hydrochloride anesthetic, highlighting the im￾portance of sampling microbiome sites with minimal perturbations. The anesthetic may have diluted or washed away bacteria from the ocular surface. By comparing the microbiota of ocular conjunctiva to that of skin under the eye, we found that normal Corynebacterium and Staphylococcus skin bacteria were detected at higher relative abun￾dance in the eye than in the skin, consistent with a recent study showing the distribution of skin microbiome by body sites (16). Tear lactoferrin plays a role in reducing levels of Staphylococcus epidermidis biofilms (23). However, in this study, we found that Staphylococcus levels were lower in lens wearers than in non-lens wearers. There are not many reports of studies using molecular methods in the diseased eye. Lee et al. (6) studied blepharitis using different sampling sites (eyelashes and tears), sequencing regions, and platforms (V1 to V3 16S rRNA genes with Roche-454). Their results showed that blepharitis patients (n 7) had increased FIG 4 Differences in relative abundances of phylotypes in ocular microbiota between non-lens wearers and lens wearers. Each phylotype (1% of average relative abundance in groups) is indicated by a different color at the genus level. (A) Bacterial taxon plots at the genus level. (B) Histogram of unique biomarker bacteria in each group. The LDA effect size (3.0-fold) was used to detect unique biomarkers. Shin et al. 4 ® mbio.asm.org March/April 2016 Volume 7 Issue 2 e00198-16 mbio.asm.org on June 29, 2016 - Published by mbio.asm.org Downloaded from