Altered Eye Microbiota by Wearing Contact Lenses Staphylococcus, Streptophyta, Corynebacterium, and Enhydrobacter ophthalmology office after utilization of a 50-pl drop of topical 0.5% levels in their eyelashes and tears in relation to controls(n= 4), proparacaine hydrochloride anesthetic Negative-control swabs(n=3 none of which were increased in our lens wearers. In addition to with no sample)were also included. the different sites sampled, their patients were older than ours(59 DNA extraction and sequencing. DNA extraction was carried out 4.5]-year-old patients in our study). More studies are needed to according to the instructions provided by the manufacturer. Extracted g understand which are the opportunistic pathogens associated DNA from samples was stored at-"C until sequencing. The V4 region with infections in lens wearers. Other skin bacteria were enriched of 16s rRNA gene was amplified by PCR using barcoded primers, as pre in contact lenses and in the conjunctiva of lens wearers, which viously described ( 29). To rule out and control for possible reagent con- a considered to be opportunistic pathogens in conjunctivitis, kera- equimolar ratios and purified using a QIAquick PCR purification kit 8 titis, and endophthalmitis(24-26). Enrichment of skin bacteria Illumina Mise platform( Genome Technology Center of NYU Medical y wearing contact lenses suggests that contact lenses could Cente function as a medium to transfer skin bacteria to the ocular sur- Dar NY) using a paired-end technique(2 150-cycle runs) face. Alternatively, contact lenses may be exerting selective pres- the QUIME suite of software tools(v1. 8)(31). The filtered sequence reads sure on ocular bacterial communities in favor ofskin-like bacteria. (Phred 2 Q20)were used to pick the operational taxonomic units Given that people use their fingers(no matter how well (OTUs), with an open-reference OTU picking method based on 97% washed)to put in contact lenses, we tested whether conjunctival identity to entries in the Greengenes database( v13_8). Negative-control. g acterial OTUs in lens wearers originate from hands, and results derived OTUs were discarded from the OTU table using a filtration script suggested that they do not. Consistently, evidence suggests that (filter_otus_ from_otu_table.py)in QIIME. After the chimeric sequences ransplantation of microbiota between body sites changes the bac- were removed using UCHIME (32), all communities were rarefied to terial population structure only temporarily( 27). An experiment 2,090 reads per sample. For comparison of levels of beta diversity between in which lens wearers use or do not use sterile gloves to insert their communities, the unweighted/weighted UniFrac distances (33)were cal- lenses would clarify the origin of the bacteria in contact lenses. culated and PERMANOVA(34)was used to test significance. Linear dis The different taxa in the conjunctiva of lens wearers are re- criminant analysis effect size(LEfSe)(35)was used to detect unique bio- flected in the predicted bacterial gene content, so the differences markers by determinations of the relative abundances of the members of terial gene profiles(mean accuracy =0.99+0.03 standard devia- using PICRUSt(28)with Kegg Orthology(KO)classification(36) tion [SD] for core/bacterial housekeeping functions; minimum To determine the overlap between OTUs of the eye samples and OTUs accuracy=0.82 for membrane-associated functions)(28); fur- found in other skin sites, we used Source Tracker(37)with a previously reported 16S rRNA gene data set derived from samples from human face whether bacterial gene functions confer reduced protection The HMP data set(15)of 16S rRNA (regions V3 to V5)sequences was against ocular infections. downloaded from the NIH HMP website(hmpdacc org) and was Previous studies have suggested that the commensal microbi- trimmed to contain only the v4 region of the 16S rRNA gene using otaoftheocularsurfacecouldinteractwiththehostinimmuneBioperl(http://www.bioperlorg/wiki/main_page).thE16srrNa(V4 system to suppress microbial pathogenicity (4, 6, 7), and the im- region)sequences from the skin of the face(16)were downloaded from pact of wearing contact lenses may affect this protective function. EMBL EBI database(ERP005182). The sequences from the Bouslimani Our report provides novel insights into possible mechanisms by study and the V4-trimmed hMP data set were merged with sequences hich wearing contact lenses increases eye infection risks Further from this study. The QIIME suite(v1. 8)was used to pick OTUs from the research is required to determine if the risk is related to contami- merged sequences using the closed-reference method. Then, all commu- ating the lenses with bacteria from the skin of the finger or if nities were rarefied to 1,000 sequences per sample to calculate bacterial contact lenses exert selective pressures on the eye bacterial com- diversity munity in favor of skin bacteria. Nucleotide sequence accession number. The raw sequences support ing the results of this article are available in the European Nucleotide MATERIALS AND METHODS Archiverepositoryunderaccessionno.PrjEb12498(htTp://www.ebi Sample collection. The study was performed under IRB protocol S12 ac uk/ena/data/view/PRJEB12498) 03905 of the NYU Medical Center. Subjects seeking routine eye care at the ophthalmology practice at the NYU Medical Center from October 2013 to SUPPLEMENTAL MATERIAL June2014weresampledfromtheirleftandrighteyesAtotalof58subSupplementalmaterialforthisarticlemaybefoundathttp://mbio.asm.org/ cts(40 females and 18 males) provided samples; 20 of those subjects lookup/suppl/doi: 10. 1128/mBio. 00198-16/-/DCSupplemental including 9 lens wearers and 11 non-lens wearers) Figure S1, PDF file, 0.4 MB udinally, with biweekly samples collected for 6 weeks(see Table Sl and Figure S2, PDF file, 1.5 MB Table S2 in the supplemental material). Dry sterile cotton swabs(Fishe Figure S3, PDF file, 0.1 MB Scientific)used to swab samplin laced in sterile Figure S4, PDF file, 0.4 MB cryogenic tubes(Thermo Scientific). Sites sampled included right and left Figure S5, PDF file, 1.3 MB. onjunctivas, skin under the right and left eyes, and contact lens(250 total Figure S6, PDF file, 0.7 MB es, including 116 conjunctiva samples, 114 skin samples, and 20 Figure S7, PDF file, 0.1MB contact lens samples). Samples were immediately transported to the lab Table s1. PDF file. 0.1 MB. oratory on ice and frozen at -80C. A total of 38 additional subjects were Table S2. PDF file. 0.1 MB. led once(including 16 lens wearers; n=75)(see Table S3)in an Table S3, PDF file. 0.1 MB. March/April 2016 Volume 7 Issue 2 e00198-16 msio° mbio. asm. org5Staphylococcus, Streptophyta,Corynebacterium, and Enhydrobacter levels in their eyelashes and tears in relation to controls (n 4), none of which were increased in our lens wearers. In addition to the different sites sampled, their patients were older than ours (59 [ 16.6]-year-old patients in the study by Lee et al. and 26 [ 4.5]-year-old patients in our study). More studies are needed to understand which are the opportunistic pathogens associated with infections in lens wearers. Other skin bacteria were enriched in contact lenses and in the conjunctiva of lens wearers, which have increased relative abundances of Pseudomonas, Acinetobac￾ter, and Methylobacterium. Representatives of these genera are considered to be opportunistic pathogens in conjunctivitis, kera￾titis, and endophthalmitis (24–26). Enrichment of skin bacteria caused by wearing contact lenses suggests that contact lenses could function as a medium to transfer skin bacteria to the ocular sur￾face. Alternatively, contact lenses may be exerting selective pres￾sure on ocular bacterial communities in favor of skin-like bacteria. Given that people use their fingers (no matter how well washed) to put in contact lenses, we tested whether conjunctival bacterial OTUs in lens wearers originate from hands, and results suggested that they do not. Consistently, evidence suggests that transplantation of microbiota between body sites changes the bac￾terial population structure only temporarily (27). An experiment in which lens wearers use or do not use sterile gloves to insert their lenses would clarify the origin of the bacteria in contact lenses. The different taxa in the conjunctiva of lens wearers are re- flected in the predicted bacterial gene content, so the differences are not redundant in gene content. We could reliably predict bac￾terial gene profiles (mean accuracy 0.99  0.03 standard devia￾tion [SD] for core/bacterial housekeeping functions; minimum accuracy 0.82 for membrane-associated functions) (28); fur￾ther metagenomic/transcriptomic research is needed to elucidate whether bacterial gene functions confer reduced protection against ocular infections. Previous studies have suggested that the commensal microbi￾ota of the ocular surface could interact with the host in immune system to suppress microbial pathogenicity (4, 6, 7), and the im￾pact of wearing contact lenses may affect this protective function. Our report provides novel insights into possible mechanisms by which wearing contact lenses increases eye infection risks. Further research is required to determine if the risk is related to contami￾nating the lenses with bacteria from the skin of the finger or if contact lenses exert selective pressures on the eye bacterial com￾munity in favor of skin bacteria. MATERIALS AND METHODS Sample collection. The study was performed under IRB protocol S12- 03905 of the NYU Medical Center. Subjects seeking routine eye care at the ophthalmology practice at the NYU Medical Center from October 2013 to June 2014 were sampled from their left and right eyes. A total of 58 sub￾jects (40 females and 18 males) provided samples; 20 of those subjects (including 9 lens wearers and 11 non-lens wearers) were followed longi￾tudinally, with biweekly samples collected for 6 weeks (see Table S1 and Table S2 in the supplemental material). Dry sterile cotton swabs (Fisher Scientific) used to swab sampling sites were immediately placed in sterile cryogenic tubes (Thermo Scientific). Sites sampled included right and left conjunctivas, skin under the right and left eyes, and contact lens (250 total samples, including 116 conjunctiva samples, 114 skin samples, and 20 contact lens samples). Samples were immediately transported to the lab￾oratory on ice and frozen at  80°C. A total of 38 additional subjects were sampled once (including 16 lens wearers; n 75) (see Table S3) in an ophthalmology office after utilization of a 50-l drop of topical 0.5% proparacaine hydrochloride anesthetic. Negative-control swabs (n 3, with no sample) were also included. DNA extraction and sequencing. DNA extraction was carried out using MoBio (CA, USA) PowerSoil-htp 96 well soil DNA isolation plates according to the instructions provided by the manufacturer. Extracted DNA from samples was stored at 20°C until sequencing. The V4 region of 16s rRNA gene was amplified by PCR using barcoded primers, as pre￾viously described (29). To rule out and control for possible reagent con￾tamination, reagents for DNA extraction and for PCR amplification were also sequenced as controls (30). The amplicons were then pooled in equimolar ratios and purified using a QIAquick PCR purification kit (Qiagen Inc., CA, USA). The pooled amplicons were sequenced on an Illumina MiSeq platform (Genome Technology Center of NYU Medical Center, NY) using a paired-end technique (2 150-cycle runs). Data analysis. The 16S rRNA sequence analyses were performed with the QIIME suite of software tools (v1.8) (31). The filtered sequence reads (Phred Q20) were used to pick the operational taxonomic units (OTUs), with an open-reference OTU picking method based on 97% identity to entries in the Greengenes database (v13_8). Negative-control￾derived OTUs were discarded from the OTU table using a filtration script (filter_otus_from_otu_table.py) in QIIME. After the chimeric sequences were removed using UCHIME (32), all communities were rarefied to 2,090 reads per sample. For comparison of levels of beta diversity between communities, the unweighted/weighted UniFrac distances (33) were cal￾culated and PERMANOVA (34) was used to test significance. Linear dis￾criminant analysis effect size (LEfSe) (35) was used to detect unique bio￾markers by determinations of the relative abundances of the members of the bacterial taxonomies. Predictive functional analysis was performed using PICRUSt (28) with Kegg Orthology (KO) classification (36). To determine the overlap between OTUs of the eye samples and OTUs found in other skin sites, we used SourceTracker (37) with a previously reported 16S rRNA gene data set derived from samples from human face and hand (16). The HMP data set (15) of 16S rRNA (regions V3 to V5) sequences was downloaded from the NIH HMP website (hmpdacc.org) and was trimmed to contain only the V4 region of the 16S rRNA gene using BioPerl (http://www.bioperl.org/wiki/Main_Page). The 16S rRNA (V4 region) sequences from the skin of the face (16) were downloaded from EMBL EBI database (ERP005182). The sequences from the Bouslimani study and the V4-trimmed HMP data set were merged with sequences from this study. The QIIME suite (v1.8) was used to pick OTUs from the merged sequences using the closed-reference method. Then, all commu￾nities were rarefied to 1,000 sequences per sample to calculate bacterial diversity. Nucleotide sequence accession number. The raw sequences support￾ing the results of this article are available in the European Nucleotide Archive repository under accession no. PRJEB12498 (http://www.ebi .ac.uk/ena/data/view/PRJEB12498). SUPPLEMENTAL MATERIAL Supplemental material for this article may be found at http://mbio.asm.org/ lookup/suppl/doi:10.1128/mBio.00198-16/-/DCSupplemental. Figure S1, PDF file, 0.4 MB. Figure S2, PDF file, 1.5 MB. Figure S3, PDF file, 0.1 MB. Figure S4, PDF file, 0.4 MB. Figure S5, PDF file, 1.3 MB. Figure S6, PDF file, 0.7 MB. Figure S7, PDF file, 0.1 MB. Table S1, PDF file, 0.1 MB. Table S2, PDF file, 0.1 MB. Table S3, PDF file, 0.1 MB. Altered Eye Microbiota by Wearing Contact Lenses March/April 2016 Volume 7 Issue 2 e00198-16 ® mbio.asm.org 5 mbio.asm.org on June 29, 2016 - Published by mbio.asm.org Downloaded from