Bacterial Distribution On The Ocular Surface of Patients With Primary Sjögren’s Syndrome

Abstract

Many studies have shown that gut microbial dysbiosis is a major factor in the etiology of autoimmune diseases but none have suggested that the ocular surface (OS) microbiome is associated with Sjögren’s syndrome (SS). In this prospective study, we analyzed bacterial distribution on the OS in patients with primary SS. Among the 120 subjects included in this study, 48 patients (group A) had primary SS, whereas 72 subjects (group B) had dry eye symptoms that were unrelated to SS. We evaluated clinical dry eye parameters such as the OS disease index, ocular staining score, Schirmer’s test, and tear break-up time. Conjunctival swabs were used to analyze the microbial communities from the two groups. Bacterial 16S rRNA genes were sequenced using the Illumina MiSeq platform, and the data were analyzed using the QIIME 1.9.1 program. The Shannon index was significantly lower in group A than in group B microbiota. An analysis of similarity using the Bray–Curtis distance method found no difference in beta-diversity between the two groups. In group A, Actinobacteria at the phylum level and Corynebacteria at the genus level exhibited low abundance, but the differences were not statistically significant. SS apparently decreases the diversity of the OS microbial community and may affect the abundance of some bacterial strains at the phylum and genus levels. These observations may be important for the pathophysiology of SS and should be investigated in future studies.

Introduction

Sjögren’s syndrome (SS) is an autoimmune disease that is characterized by lymphocytic infiltration of exocrine glands such as salivary and lacrimal glands, resulting in dry eyes and a dry mouth.¹ Patients with SS and dry eyes exhibit particular glandular and mucosal immunopathological changes. CD4⁺ T cells and dendritic cells infiltrate conjunctival tissue, and inflammatory cytokines such as interferon gamma and interleukin (IL)-17, secreted from these T cells, induce apoptosis of epithelial and goblet cells, exacerbating dryness of the ocular surface (OS).^2,3 Therefore, SS changes the immunological characteristics of the corneal and conjunctival surfaces.

Commensal microorganisms are present on the uninfected human OS, and several studies have used traditional microbiological culture techniques to study the microbiota of the OS.^4–6 Deep sequencing of bacterial DNA has also shown that various species of commensal microorganisms are present on the OS.⁷

However, it is unclear whether the species distribution of commensal microorganisms on the OS is altered in patients with SS. Unusual microbial flora distribution patterns are associated with some immunological diseases, and normalizing these distribution patterns is one approach to treatment. Crohn’s disease is a chronic inflammatory disorder of the gastrointestinal tract that results from an excessive innate and adaptive immune response to environmental factors in genetically susceptible subjects.⁸ Crohn’s disease suppresses the function of regulatory T cells, which play an important role in immunological homeostasis.⁹ Importantly, bacterial dysbiosis may be an important pathological consequence of Crohn’s disease, and restoring the microbial balance is one method of treating the disease.^10,11 In fact, patients with Crohn’s disease have more pathogenic bacterial species and fewer normal commensal bacterial species.¹² Zaheer et al. attempted to demonstrate an association between gut dysbiosis and SS in a CD25 knockout mouse model. The study revealed that lack of commensal gut bacteria aggravated dacryoadenitis and generated autoreactive CD4⁺ T cells, increasing pathogenicity in the knockout mice. Therefore, in this SS mouse model, the commensal gut bacteria and their metabolites may have an immunoregulatory function that protects the exocrine glands.¹³

However, it is difficult to establish a clear link between gut dysbiosis and pathology of the OS in patients with SS. Therefore, it is necessary to study the OS microbiota. Unfortunately, as Zaheer et al. reported, OS microbiota are sparse and difficult to analyze. Therefore, we used next-generation sequencing (NGS) to analyze numerous microbial populations distributed across the OS, similar to a previous study.¹⁴ Here, we assess the composition of the OS microbial communities and evaluate the significance of the diversity and abundance of bacterial strains.

Materials And Methods

Study design and subjects

This prospective case–control study was conducted in accordance with the Declaration of Helsinki, and the study protocol was approved by the Institutional Review Board of Incheon St. Mary’s Hospital (Incheon, South Korea). Also, written informed consent was obtained from all participants. Patients were recruited at the Cornea Service of Incheon St. Mary’s Hospital. Patients who wore contact lenses or had any of the following conditions or treatments that could affect eye dryness were excluded: meibomian gland dysfunction above grade 2, allergic conjunctivitis, ocular surgery within the last 6 months, punctal plug insertion, or topical treatments other than artificial tears in the 3 months before this study. Patients who were taking medication for a systemic disease such as autoimmune diseases, diabetes, hypertension, thyroid disease, allergic disease, or a depressive disorder were also excluded.

A total of 120 subjects were included in this study. In total, 15 subjects were male and 115 subjects were female. All of the study subjects had been referred to ophthalmology departments to investigate a diagnosis of primary SS and reported having dry eyes (e.g., dryness, itching, or the sensation of a foreign body) and a dry mouth (e.g., frequent thirst, dry feeling in the mouth and throat, burning or tingling sensation in the mouth and tongue, or cracked lips). We used the American College of Rheumatology/European League Against Rheumatism diagnostic criteria announced in 2016 to diagnose SS.⁴³ In total, 48 patients (group A) were diagnosed with primary SS and the remaining 72 subjects (group B) were classified as having dry eyes and dry mouth symptoms that were unrelated to SS. Patients who had at least two of the following three characteristics were diagnosed as having SS: positive serum anti-SSA/Ro and/or anti-SSB/La (i.e., positive rheumatoid factor and antinuclear antibodies titer ≥ 1:320), labial salivary gland biopsy exhibiting focal lymphocytic sialadenitis with a focus score ≥ focus/4 mm², or keratoconjunctivitis sicca with OSS.

Assessment of dry eye clinical parameters

OSS that included corneal and conjunctival staining scores, Schirmer’s test, and TBUT were evaluated by the same ophthalmologist (HBH). Dry eye symptoms were graded numerically from 0 to 4 using the OSDI, and the sum of these scores was used in the analyses. The OSDI, developed by the Outcomes Research Group at Allergan (Irvine, CA, USA), is a 12-item questionnaire that assesses vision-related functioning. Each item is scored on a 5-point scale, resulting in a total OSDI score ranging from 0 (no symptoms) to 100 (maximal symptoms).⁴⁴ To measure TBUT, fluorescein was instilled from a fluorescein strip in a drop of saline (Haag-Streit, Koeniz, Switzerland). The fluorescein was applied to the superior-temporal bulbar conjunctiva with participants instructed to gaze infero-nasally, and the TBUT was measured using a slit-lamp under a cobalt blue light.⁴⁵ The mean value of three TBUT measurements was used in this study. The OSS was measured in accordance with the SICCA registry ocular examination protocol.⁴⁶ Corneal punctate epithelial erosions (PEEs) were evaluated and scored after staining with a fluorescent dye. Corneal scores were assigned as follows: PEEs absent = 0 points; one to five PEEs = 1 point; six to 30 PEEs = 2 points; and more than 30 PEEs = 3 points. An additional point was assigned if more than one patch of confluent aggregated staining was found in the pupillary area or if filaments present on the cornea were stained. The maximum possible score for each cornea was 6 points. To assess conjunctival staining scores, fluorescein was washed out with non-preserved saline solution, after which 1% lissamine green dye (Leiter’s Pharmacy, San Jose, CA, USA) was applied to the inferior conjunctival fornix. After several blinks, conjunctival staining scores in the temporal and nasal bulbar conjunctiva were evaluated separately as follows: up to nine dots = 0 points; 10 to 32 dots = 1 point; 33 to 100 dots = 2 points; and more than 100 dots = 3 points. The maximum possible score for the conjunctiva (i.e., temporal and nasal regions) was 6 points. For Schirmer’s test, standard Schirmer strips (Eagle Vision, Memphis, TN, USA) were placed in the lateral one-third of the lower eyelid, without topical anesthesia. After 5 min, the length of the wet portion of the strip was measured.

Sample collection and DNA extractions

The sampling and DNA extraction procedures used in this study were the same as those used in our previous research.¹⁴ In the absence of topical anesthesia, sterile dry cotton swabs (MEDIUS Corp., Tokyo, Japan) were used to sample four different locations: left upper bulbar conjunctiva, left lower bulbar conjunctiva, right upper bulbar conjunctiva, and the right lower bulbar conjunctiva. Each cotton swab was placed into a 1.5 mL sterilized Eppendorf microtube and frozen at –70℃ until DNA extraction. Nucleic acids were extracted from the swabs using an i-genomic Soil DNA Extraction Mini Kit (iNtRON, Seoul, South Korea) with a bead-beating apparatus, in accordance with the manufacturer’s instructions with the following method alterations. To enhance cell lysis, approximately 1.5 × lysis buffer was added to the samples for 30 min. DNA concentrations were measured using a Qubit fluorometer (Invitrogen, Gaithersburg, MD, USA).

16S rRNA gene library preparation

The hypervariable region 4 (V4) of the 16S rRNA gene was amplified from the extracted DNA using the 515F (5ʹ–GTGYCAGCMGCCGCGGTAA–3ʹ) and 806R (5ʹ–GGACTACNVGGGTWTCTAAT–3ʹ) primers.⁴⁷ The polymerase chain reaction (PCR) conditions were 94°C for 3 min, followed by 35 cycles of 94°C for 45 s, 50°C for 60 s, and 72°C for 90 s, followed by a final extension step at 72°C for 10 min. Each sample was amplified in three replicate 25-µL PCR reactions, purified using the UltraClean PCR clean-up kit (Mo Bio Laboratories, Solana Beach, CA, USA), and combined into a single tube. Purified DNA was quantified by incorporating Picogreen (Invitrogen), in accordance with the manufacturer’s instructions.

Sequencing data analyses

Libraries for the V4 region alone were generated, and 16S rRNA-based MiSeq sequencing was conducted using standard Earth Microbiome Project protocols (http://www.earthmicrobiome.org/emp-standardprotocols/) at the Argonne National Laboratory (Lemont, IL, USA). QIIME version 1.9.1 was used for processing all sequencing reads.⁴⁸ Sequences were de-multiplexed according to their barcodes, merged, and quality-filtered using the default parameters. OTUs were clustered using the latest Greengenes 13_8 reference sequences at 97% similarity (approximately corresponding to species-level OTUs), using the UCLUST algorithm. To remove chimera sequences, UCHIME within USEARCH was used.⁴⁹ Then, sequences were aligned using PyNAST software.⁵⁰ Representative OTU sequences were evaluated taxonomically using the ribosomal database project classifier,⁵¹ retrained on the Greengenes database (13_8),⁵² and mitochondrial sequences were filtered out of the OTU table. OTU reads found in negative control samples were removed from the OTU table.

To calculate species diversity and richness, alpha diversity analyses using the Shannon and Chao1 indexes were processed using the QIIME script. To measure the similarity among communities, beta diversity was calculated and MFA plots were constructed.

Negative controls

All experimental procedures including sampling, DNA extraction, preparation, and sequencing were implemented in a consistent manner. DNA extraction and PCR amplification of samples from eyes and unused sterile cotton swabs were all performed using exactly the same procedures. When DNA was extracted from the unused sterile cotton swabs, the quantity of DNA produced was very low (i.e., less than 0.05 ng DNA µL^-1), and no DNA was amplified by the bacteria-specific primers (i.e., 27F and 1494R did not generate a PCR band). To minimize variability from any batch effects during different sequencing runs, all samples were sequenced using the same sequencing run. Negative controls were generated in reaction mixtures containing no template. During the sequence data analyses, OTU reads identified in DNA extractions of negative controls were removed from the sequencing data using metadata-based filtering.

Diagnostic testing procedure

The testing procedure for ophthalmic examinations was as follows:

1. Subjective interview regarding symptoms of dry eyes (i.e., OSDI) and recording the patient’s medical history (HBH and KDK).
2. Slit-lamp evaluation of the cornea, conjunctiva, eyelids, and Meibomian glands (HBH).
3. Schirmer’s test (without topical anesthesia; HBH).
4. TBUT test using fluorescein dye, repeated 3 times (HBH).
5. Fluorescein and lissamine green staining of the cornea and conjunctiva (HBH).

Thereafter, subjects returned after 2 weeks and were classified into two groups according to whether the rheumatologist confirmed a diagnosis of SS.

6. Conjunctival swab sampling (HBH, KDK and JMY).
7. DNA extraction and sequencing analysis (BNK).

Statistical analyses

All data are expressed as means ± SD. Statistical analyses were performed using JAMOVI open-source (http://www.jamovi.org) and XLSTAT (Addinsoft, Paris, France) software. Group differences in age and dry eye parameters were evaluated using Student’s t-test after confirming that the data were normally distributed using the Kolmogorov–Smirnov test. Group differences in alpha diversity (i.e., Shannon and Chao1 indexes) were analyzed using a GLM by controlling for age and dry eye parameters as covariates. Before the GLM analysis, we confirmed that the data were normally distributed using the Q–Q plotting method. MFA was based on PCA and was used to create a visual representation of the data, with enrolled subjects and the various microbial strains arranged into two groups. To compare the microbial communities associated with the two groups, an ANOSIM was performed using Bray–Curtis dissimilarities after 9999 permutations and the “vegan” package in ANOSIM. We used Student’s t-test to compare species richness between the groups. A p-value < 0.05 was considered statistically significant.

Results

Basic characteristics and dry eye parameters

Of the 120 subjects included in this study, 48 subjects (group A) were diagnosed with SS according to American College of Rheumatology/European League Against Rheumatism diagnostic criteria; the remaining 72 subjects (group B) did not have SS. Basic characteristics of all subjects are described in Table 1. In both groups, there were more females than males. Kolmogorov–Smirnov tests were used to confirm that the data were normally distributed (p > 0.05). The mean ages of groups A and B were 51.71 ± 9.46 years and 54.50 ± 13.60 years, respectively, and these ages were not significantly different (independent t-test, p > 0.05). There were no significant differences between the two groups in terms of clinical parameters for dry eyes such as ocular staining score (OSS), OS disease index (OSDI), tear break-up time (TBUT), and Schirmer’s test (independent t-tests, all p > 0.05).

Sequencing data

The 16S rRNA V4 amplicon libraries were sequenced from 120 individual samples. We obtained a total of 1890438 high-quality reads, and the number of observations (distinct operational taxonomic units [OTUs]) was 10668. There was a mean ± standard deviation (SD) of 15753.7 ± 6354.1 sequences per sample, with a median of 15928 reads (range, 5001 to 37333). Additionally, all the raw reads were classified using a naïve Bayesian ribosomal database project classifier for taxonomic assignment at 97% similarity. The mean (with SD) number of OTUs was 412.27 ± 160.87, the minimum number of OTUs was 117, and the maximum number of OTUs was 1013 (Table 2).

We constructed rarefaction curves to check the sequencing depth¹⁵ and confirmed that these samples were sequenced deeply enough when viewed as plots of observed species (Fig. 1A) and phylogenetic diversity (PD) of the whole tree (Fig. 1B).

Alpha diversity of the microbial community

There were no statistically significant differences in OTUs between the two groups (p = 0.470, Fig. 2A). However, the PD of the whole tree was significantly different in each group (p = 0.016, Fig. 2B). In addition, we used Student’s t-tests to analyze the alpha diversity of the OS microbiome in each group. The Chao1 index showed that the species richness between the two groups was not significantly different (p = 0.924, Fig. 2C). However, the Shannon index showed that there was less diversity among bacterial species in group A than in group B (p = 0.017, Fig. 2D). We also analyzed the alpha diversity of the two groups in terms of controlling variables such as age, gender, and dry eye severity markers (i.e., Schirmer’s test, OSS, TBUT, and OSDI) using a general linear model (GLM). First, we confirmed that the data were normally distributed using the Q–Q plotting method. Among all the covariates, only age was significantly different between the two groups (p = 0.015). Therefore, the GLM was used to compare OTUs, PD of the whole tree, the Chao1 index, and the Shannon index for each group. The OTU (p = 0.158) and Chao1 index (p = 0.686) were not significantly different. However, for groups A and B, the PD of the whole tree (p = 0.002) and the Shannon index (p = 0.001) were significantly different.

Taxonomic composition and dominant genera and phyla

16S rRNA gene sequences were classified by phyla and genera to analyze the taxonomic composition of the microbial community. When the swab samples were analyzed, 32 bacterial phyla were detected and these were grouped into 14 major phyla. The most abundant four bacteria (> 1%) were listed for groups A and B. The Proteobacteria were most abundant in both groups (means with SD, 38.19% ± 24.72 and 34.24% ± 21.66, respectively), followed by the Actinobacteria (32.81% ± 23.06 and 37.43% ± 29.04, respectively), the Firmicutes (16.56% ± 18.26 and 15.70% ± 18.28, respectively), and the Bacteroidetes (5.39% ± 5.44 and 5.31% ± 5.87, respectively; Fig. 3A). However, there were no statistically significant differences in the relative abundance of these strains between the two groups (Student’s t-test, p > 0.05).

A total of 698 bacterial genera were detected and these were grouped into 15 major genera. The most abundant 14 bacteria (> 1%) were listed for groups A and B. The Corynebacteria were most abundant in both groups (means with SD, 24.54% ± 25.34 and 30.07% ± 31.91, respectively), followed by the Neisseriaceae (14.19% ± 24.67 and 10.77% ± 20.76, respectively), the Staphylococci (10.91% ± 16.22 and 8.07% ± 16.86, respectively), the Streptococci (1.87% ± 4.29 and 3.21% ± 5.18, respectively), the Acinetobacter (3.27% ± 5.37 and 1.81% ± 2.70, respectively), the Cryocola (2.72% ± 4.73 and 1.54% ± 4.64, respectively), the Xanthomonads (0.35% ± 0.62 and 2.46% ± 6.63, respectively), the Bradyrhizobia (2.17% ± 4.02 and 0.95% ± 2.95, respectively), the Sphingomonads (1.60% ± 2.72 and 1.25% ± 2.22, respectively), the Pseudomonads (0.98% ± 2.13 and 1.33% ± 5.05, respectively), the Enhydrobacter (0.98% ± 2.38 and 1.17% ± 1.95, respectively), the Commamonads (1.94% ± 3.78 and 0.43% ± 1.10, respectively), and the Erythrobacter (1.46% ± 3.11 and 0.65% ± 1.19, respectively; Fig. 3B). The Acinetobacter were more abundant in group A (Student’s t-test was borderline significant, p = 0.05), and the Xanthomonads were significantly more abundant in group B (Student’s t-test, p = 0.03).

Microbial communities on the OS

Figure 4 shows a multiple factor analysis (MFA) based on global principal component analysis (PCA) of microbial communities at the phylum (Fig. 4A) and genus (Fig. 4B) levels. At both levels, the microbial communities in each group were different but not significantly so. An analysis of similarity (ANOSIM) using the Bray–Curtis method and 9999 permutations found that the microbial communities were not significantly separated at the phylum (p = 0.673) or genus (p = 0.637) level.

Discussion

Physiological interactions between the gut microbiota and immune system occur during human development,¹⁶ and gut dysbiosis plays an important role in the pathophysiology of autoimmune diseases such as systemic sclerosis, rheumatoid arthritis, ankylosing spondylitis, and SS.¹⁷ These autoimmune diseases are often associated with ophthalmic disorders such as uveitis and keratoconjunctivitis sicca.^18–20 Keratoconjunctivitis sicca is caused by inflammatory cell infiltration of the lacrimal glands and occurs frequently in patients with SS. Many studies have investigated the association between gut dysbiosis and OS lesions.^19,21 Moon et al. reported that the gut microbiome of patients with SS differed from that of control subjects,²¹ and Paiva et al. showed that the severity of SS was inversely correlated with microbial diversity.¹⁹ However, the pathological manifestations of keratoconjunctivitis sicca occur on the OS, and there are fewer microbes on the OS than in the gut. The cornea and conjunctival mucosa are directly exposed to the environment. However, due to defense mechanisms such as antimicrobial proteins in the tears, the protective corneal epithelium, and the presence of various immune cells, few microorganisms flourish on the OS of the uninfected eye.^22–24 Consequently, few studies have analyzed microbial communities on the OS of patients with autoimmune diseases such as SS. Pavia et al. devised a desiccated-stress SS animal model that exhibited significant differences in the microbiota found in the oral mucosa compared with the gut. However, no meaningful differences were observed in conjunctival swab samples because few microbes were present on the OS.¹⁹ Therefore, as demonstrated by our previous research,¹⁴ NGS may be required to analyze these microbiomes.

After controlling for age and dry eye clinical markers using the OSS, TBUT, and Schirmer’s test, we analyzed microbial diversity on the OS of patients with SS (group A) and subjects who did not have SS (group B). The Shannon index was significantly lower in the patients with SS. The Shannon index evaluates microbial diversity by describing the entropy of an ecosystem or in representative samples. Therefore, patients with SS exhibited a decrease in microbial diversity on the OS. In the animal model studied by Pavia et al., the severity of ocular and systemic SS was inversely correlated with gut microbial diversity.¹⁹ Moon et al. also found differences in gut microbial diversity between patients with SS and control subjects.²¹ In addition, studies of the OS microbiota in subjects with dry eyes, contact lens wearers, and patients with blepharitis or trachoma identified characteristic differences in the abundance of some microbial species.^25–28 Our study is the first to analyze microbial communities on the OS of patients with and subjects without an autoimmune disease. In addition, we tried to exclude subjects who had systemic or ocular diseases, as well as those who had taken systemic agents (e.g., hydroxychloroquine) and topical drugs that could affect the OS microbiota. Finally, we focused on differences in the microbiota that were caused by SS by controlling for age and clinical markers for dry eyes.

We found that the alpha diversity of the microbial community decreased in patients with SS. Previous studies reported the presence of a gut–eye lacrimal-gland microbiome axis in patients with SS and that commensal bacteria were involved in protecting the OS and lacrimal glands.^19,29 However, we suggest that protection of the OS may also be compromised if the homeostasis of commensal bacteria is disrupted. Previous studies have shown that both Gram-positive and Gram-negative bacteria use quorum sensing to regulate biofilm formation, expression of virulence factors, and motility.³⁰ Quorum sensing is based on the density of bacterial populations within a particular environment.³¹ Therefore, if the homeostasis of commensal bacterial on the OS is disrupted and bacterial diversity decreases, pathogenicity may occur.

We also analyzed species abundance. At the phylum level, Proteobacteria and Actinobacteria were the predominant microorganisms on the OS, regardless of group. This observation is consistent with previous research.^7,14,32,33 By contrast, Firmicutes are generally the predominant microorganisms present in the gut. Moon et al. reported that Firmicutes were the predominant microorganisms in the gut of patients with SS.²¹ At the genus level, Corynebacteria were the predominant microorganisms on the OS. This observation is also consistent with previous research. These results suggest that a core community of commensal bacteria are present on the OS and that the species abundance of OS bacteria differs from that of gut bacteria.

There were more Proteobacteria and fewer Actinobacteria on the OS of patients with SS, compared to control subjects. However, these differences were not statistically significant. Previous studies have suggested that increased prevalence of Proteobacteria is a marker for an unstable microbial community and may also be associated with several inflammatory metabolic disorders.^34,35 On the other hand, Actinobacteria play important roles in regulating gut permeability, the immune system, metabolism, and the gut–brain axis.³⁶Actinobacteria can stimulate the production of IL-4 and IL-13, thereby reducing inflammation.³⁷Actinobacteria may also prevent the formation of biofilm and the growth of pathogenic bacteria.³⁸ Therefore, these potential changes in the prevalence of Proteobacteria and Actinobacteria on the OS of patients with SS may be important. Interestingly, as in our previous study on diabetic patients, genera related to the Acinetobacter were abundant on the OS of patients with SS.¹⁴Acinetobacter strains are frequently associated with nosocomial infections, especially in intensive care units, and particular species are reportedly associated with urinary tract infections, burn infections, and hospital acquired pneumonia.³⁹Acinetobacter colonies on the skin are also associated with allergic reactions.⁴⁰ In view of these features, the presence of Acinetobacter strains on the OS of patients with primary SS, which is an autoimmune disease, may be important. However, the abundance of Acinetobacter was low compared to other genera. We also observed a decreasing trend in the abundance of Corynebacteria on the OS of patients with SS. Interestingly, Corynebacteriummastitidis, which is frequently found on the OS, plays an important role in the local immune response of the corneal and conjunctival mucosa. St Leger et al. found that this organism elicited a commensal-specific IL-17 response from γδ T cells in the ocular mucosa and drove neutrophil recruitment and the release of antimicrobial agents into the tears. In addition, this Corynebacteria species was able to protect the OS from infection by pathogenic Candida albicans and Pseudomonas aeruginosa.⁴¹ Further research will be needed to understand how these observations may be linked and whether ocular commensal bacteria can train cells involved in the immune response to resist pathogens.

Our study had some limitations, and our results should be interpreted with caution due to the limited sample size and the paucity of bacteria present on the OS. Although differences in the bacterial populations were observed in the MFA plots at the phylum and genus levels, significant differences in the Bray–Curtis distances were not found. In addition, there are practical limitations inherent to analysis of bacterial species on the OS using DNA sequencing. However, we were able to exclude the possibility of contamination of our sterile cotton swab samples with bacterial DNA. In addition, the proportions of microorganisms that we identified on the OS were similar to those described by previous studies. Although 16S rRNA sequencing with sparse bacteria from the OS may generate some false positive reads and sequencing errors, NGS has many advantages over conventional culture methods.^33,42 This is the first study to investigate the microbiota of the OS in patients with SS using NGS, while controlling for dry eye clinical markers.

In conclusion, SS decreases the diversity of the microbial community on the OS of patients and affects the abundance of particular bacterial strains at the phylum and genus levels. It is unclear whether these changes result from or cause immunological disorders, but these possibilities may be investigated in future studies to understand more fully the pathophysiology of SS.

Declarations

Acknowledgments

We thank Dr. Dion Antonopoulos and Mr. Jason Koval from Argonne National Laboratory (Lemont, IL, USA) for sequencing the samples. The authors acknowledge financial support from the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (2017R1C1B5017408). This research was also supported by a Grant of Translational R&D Project through the Institute for Bio-Medical Convergence, Incheon St. Mary’s Hospital, The Catholic University of Korea. The authors alone are responsible for the content and preparation of this article.

Author contributions

HBH and KDK designed the research and collected the samples. HBH and HSK wrote and edited the manuscript. BNH and MJK analyzed the data. YCK, HSK, JMY and KDK performed the experimental work. All authors read and approved the manuscript.

Data availability

We declare that all experimental data are described accurately in this manuscript. For further information, please contact the corresponding author.

Competing interests

The authors declare that they have no competing interests.

References

1. Argyropoulou, O. D. et al. One year in review 2018: Sjögren's syndrome. Clin Exp Rheumatol. 36 (Suppl 112), 14–26 (2018).
2. De Paiva, C. S. et al. Dry eye-induced conjunctival epithelial squamous metaplasia is modulated by interferon-gamma. Invest Ophthalmol Vis Sci. 48, 2553–2560 https://doi.org/10.1167/iovs.07–0069 (2007).
3. Zhang, X. et al. Interferon-gamma exacerbates dry eye-induced apoptosis in conjunctiva through dual apoptotic pathways. Invest Ophthalmol Vis Sci. 52, 6279–6285 https://doi.org/10.1167/iovs.10–7081 (2011).
4. Fernandez-Rubio, M. E., Rebolledo-Lara, L., Martinez-Garcia, M., Alarcon-Tomas, M. & Cortes-Valdes, C. The conjunctival bacterial pattern of diabetics undergoing cataract surgery. Eye (Lond). 24, 825–834 https://doi.org/10.1038/eye.2009.218 (2010).
5. Hori, Y. et al. Bacteriologic profile of the conjunctiva in the patients with dry eye. Am J Ophthalmol. 146, 729–734 https://doi.org/10.1016/j.ajo.2008.06.003 (2008).
6. Sankaridurg, P. R. et al. Lid and conjunctival micro biota during contact lens wear in children. Optom Vis Sci. 86, 312–317 https://doi.org/10.1097/opx.0b013e318199d20c (2009).
7. Dong, Q. et al. Diversity of bacteria at healthy human conjunctiva. Invest Ophthalmol Vis Sci. 52, 5408–5413 https://doi.org/10.1167/iovs.10–6939 (2011).
8. Clough, J. N., Omer, O. S., Tasker, S., Lord, G. M. & Irving, P. M. Regulatory T-cell therapy in Crohn's disease: Challenges and advances. Gut. https://doi.org/10.1136/gutjnl–2019–319850 (2020).
9. Zundler, S. & Neurath, M. F. Immunopathogenesis of inflammatory bowel diseases: Functional role of T cells and T cell homing. Clin Exp Rheumatol. 33, S19–28 (2015).
10. Kostic, A. D., Xavier, R. J. & Gevers, D. The microbiome in inflammatory bowel disease: Current status and the future ahead. Gastroenterology. 146, 1489–1499 https://doi.org/10.1053/j.gastro.2014.02.009 (2014).
11. Khanna, S. & Raffals, L. E. The microbiome in Crohn's disease: Role in pathogenesis and role of microbiome replacement therapies. Gastroenterol Clin North Am. 46, 481–492 https://doi.org/10.1016/j.gtc.2017.05.004 (2017).
12. Sartor, R. B. Microbial influences in inflammatory bowel diseases. Gastroenterology. 134, 577–594 https://doi.org/10.1053/j.gastro.2007.11.059 (2008).
13. Zaheer, M. et al. Protective role of commensal bacteria in Sjögren syndrome. J Autoimmun. 93, 45–56 https://doi.org/10.1016/j.jaut.2018.06.004 (2018).
14. Ham, B. et al. Distribution and diversity of ocular microbial communities in diabetic patients compared with healthy subjects. Curr Eye Res. 43, 314–324 https://doi.org/10.1080/02713683.2017.1406528 (2018).
15. Lozupone, C. A. & Knight, R. Species divergence and the measurement of microbial diversity. FEMS Microbiol Rev. 32, 557–578 https://doi.org/10.1111/j.1574–6976.2008.00111.x (2008).
16. Lynch, S. V. & Pedersen, O. The human intestinal microbiome in health and disease. N Engl J Med. 375, 2369–2379 https://doi.org/10.1056/NEJMra1600266 (2016).
17. Mao, K. et al. Innate and adaptive lymphocytes sequentially shape the gut microbiota and lipid metabolism. Nature. 554, 255–259 https://doi.org/10.1038/nature25437 (2018).
18. Horai, R. & Caspi, R. R. Microbiome and autoimmune uveitis. Front Immunol. 10, 232 https://doi.org/10.3389/fimmu.2019.00232 (2019).
19. De Paiva, C. S. et al. Altered mucosal microbiome diversity and disease severity in Sjögren syndrome. Sci Rep. 6, 23561 https://doi.org/10.1038/srep23561 (2016).
20. Cavuoto, K. M., Banerjee, S. & Galor, A. Relationship between the microbiome and ocular health. Ocul Surf. 17, 384–392 https://doi.org/10.1016/j.jtos.2019.05.006 (2019).
21. Moon, J., Choi, S. H., Yoon, C. H. & Kim, M. K. Gut dysbiosis is prevailing in Sjögren's syndrome and is related to dry eye severity. PLoS One. 15, e0229029 https://doi.org/10.1371/journal.pone.0229029 (2020).
22. Akpek, E. K. & Gottsch, J. D. Immune defense at the ocular surface. Eye (Lond). 17, 949–956 https://doi.org/10.1038/sj.eye.6700617 (2003).
23. Zegans, M. E. & Van Gelder, R. N. Considerations in understanding the ocular surface microbiome. Am J Ophthalmol. 158, 420–422 https://doi.org/10.1016/j.ajo.2014.06.014 (2014).
24. Ozkan, J. & Willcox, M. D. The ocular microbiome: Molecular characterisation of a unique and low microbial environment. Curr Eye Res. 44, 685–694 https://doi.org/10.1080/02713683.2019.1570526 (2019).
25. Lee, S. H., Oh, D. H., Jung, J. Y., Kim, J. C. & Jeon, C. O. Comparative ocular microbial communities in humans with and without blepharitis. Invest Ophthalmol Vis Sci. 53, 5585–5593 https://doi.org/10.1167/iovs.12–9922 (2012).
26. Graham, J. E. et al. Ocular pathogen or commensal: A PCR-based study of surface bacterial flora in normal and dry eyes. Invest Ophthalmol Vis Sci. 48, 5616–5623 https://doi.org/10.1167/iovs.07–0588 (2007).
27. Zhou, Y. et al. The conjunctival microbiome in health and trachomatous disease: A case control study. Genome Med. 6, 99 https://doi.org/10.1186/s13073-014-0099–x (2014).
28. Shin, H. et al. Changes in the eye microbiota associated with contact lens wearing. mBio. 7, e00198 https://doi.org/10.1128/mBio.00198–16 (2016).
29. Trujillo-Vargas, C. M. et al. The gut-eye-lacrimal gland-microbiome axis in Sjögren syndrome. Ocul Surf. 18, 335–344 https://doi.org/10.1016/j.jtos.2019.10.006 (2020).
30. Bassler, B. L. How bacteria talk to each other: Regulation of gene expression by quorum sensing. Curr Opin Microbiol. 2 (99), 582–587 https://doi.org/10.1016/s1369–5274 (1999).
31. Miller, M. B. & Bassler, B. L. Quorum sensing in bacteria. Annu Rev Microbiol. 55, 165–199 https://doi.org/10.1146/annurev.micro.55.1.165 (2001).
32. Ozkan, J. et al. Temporal stability and composition of the ocular surface microbiome. Sci Rep. 7, 9880 https://doi.org/10.1038/s41598-017-10494–9 (2017).
33. Huang, Y., Yang, B. & Li, W. Defining the normal core microbiome of conjunctival microbial communities. Clin Microbiol Infect 22, 643 e647–643 e612, doi:10.1016/j.cmi.2016.04.008 (2016).
34. Rizzatti, G. et al. A common factor in human diseases. Biomed Res Int 2017, 9351507, doi:10.1155/2017/9351507 (2017).
35. Shin, N. R., Whon, T. W., Bae, J. W. & Proteobacteria Microbial signature of dysbiosis in gut microbiota. Trends Biotechnol. 33, 496–503 https://doi.org/10.1016/j.tibtech.2015.06.011 (2015).
36. Binda, C. et al. Actinobacteria: A relevant minority for the maintenance of gut homeostasis. Dig Liver Dis. 50, 421–428 https://doi.org/10.1016/j.dld.2018.02.012 (2018).
37. Cani, P. D. et al. Changes in gut microbiota control inflammation in obese mice through a mechanism involving GLP-2-driven improvement of gut permeability. Gut. 58, 1091–1103 https://doi.org/10.1136/gut.2008.165886 (2009).
38. Azman, A. S., Mawang, C. I., Khairat, J. E. & AbuBakar S. Actinobacteria-a promising natural source of anti-biofilm agents. Int Microbiol. 22, 403–409 https://doi.org/10.1007/s10123-019-00066–4 (2019).
39. Dent, L. L., Marshall, D. R., Pratap, S. & Hulette, R. B. Multidrug resistant Acinetobacter baumannii: A descriptive study in a city hospital. BMC Infect Dis. 10, 196 https://doi.org/10.1186/1471-2334-10–196 (2010).
40. Hanski, I. et al. Environmental biodiversity, human microbiota, and allergy are interrelated. Proc Natl Acad Sci USA. 109, 8334–8339 https://doi.org/10.1073/pnas.1205624109 (2012).
41. St Leger, A. J. et al. An ocular commensal protects against corneal infection by driving an interleukin-17 response from mucosal gammadelta T cells. Immunity 47, 148–158 e145, doi:10.1016/j.immuni.2017.06.014 (2017).
42. Doan, T. et al. Paucibacterial microbiome and resident DNA virome of the healthy conjunctiva. Invest Ophthalmol Vis Sci. 57, 5116–5126 https://doi.org/10.1167/iovs.16–19803 (2016).
43. Shiboski, C. H. et al. 2016 American College of Rheumatology/European League Against Rheumatism classification criteria for primary Sjögren's syndrome: A consensus and data-driven methodology involving three international patient cohorts. Ann Rheum Dis 76, 9–16, doi:10.1136/annrheumdis–2016–210571 (2017).
44. Schiffman, R. M., Christianson, M. D., Jacobsen, G., Hirsch, J. D. & Reis, B. L. Reliability and validity of the ocular surface disease index. Arch Ophthalmol. 118, 615–621 https://doi.org/10.1001/archopht.118.5.615 (2000).
45. Korb, D. R., Greiner, J. V. & Herman, J. Comparison of fluorescein break-up time measurement reproducibility using standard fluorescein strips versus the dry eye test (DET) method. Cornea. 20, 811–815 https://doi.org/10.1097/00003226-200111000-00007 (2001).
46. Whitcher, J. P. et al. A simplified quantitative method for assessing keratoconjunctivitis sicca from the Sjögren's Syndrome International Registry. Am J Ophthalmol. 149, 405–415 https://doi.org/10.1016/j.ajo.2009.09.013 (2010).
47. Caporaso, J. G. et al. Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. ISME J. 6, 1621–1624 https://doi.org/10.1038/ismej.2012.8 (2012).
48. Caporaso, J. G. et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods. 7, 335–336 https://doi.org/10.1038/nmeth.f.303 (2010).
49. Edgar, R. C. Search and clustering orders of magnitude faster than BLAST. Bioinformatics. 26, 2460–2461 https://doi.org/10.1093/bioinformatics/btq461 (2010).
50. Caporaso, J. G. et al. PyNAST: A flexible tool for aligning sequences to a template alignment. Bioinformatics. 26, 266–267 https://doi.org/10.1093/bioinformatics/btp636 (2010).
51. Wang, Q., Garrity, G. M., Tiedje, J. M. & Cole, J. R. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol. 73, 5261–5267 https://doi.org/10.1128/AEM.00062–07 (2007).
52. DeSantis, T. Z. et al. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl Environ Microbiol. 72, 5069–5072 https://doi.org/10.1128/AEM.03006–05 (2006).

Tables

Table 1. Ages and dry eye clinical parameters of the two groups

Table 2. Sequencing results for collected samples