1. Introduction
Throughout evolution, various microorganisms, especially bacteria, colonized the conjunctiva and the cornea as commensals, constituting the so-called OSM
[10][1]. Intriguingly, some studies revealed the capability of the eye to live in complete equilibrium with this community of bacteria. Indeed, the epithelial cells (corneal and conjunctival) of the ocular surface, although constantly in contact with resident commensal bacteria (such as
Staphylococcus epidermidis or
Propionibacterium acnes) and their products, do not trigger any inflammatory response against them; to the contrary, when these cells recognize ocular pathogenic bacteria (such as
Pseudomonas aeruginosa), they produce pro-inflammatory cytokines, including interleukin 1α (IL-1α), tumour necrosis factor α (TNFα), IL-6 and IL-8, chemokines, and interferons. This mode of action suggests the presence of an innate immune system able to distinguish through specific receptors called “pattern recognition receptors”, the commensal from the pathogen-associated molecular patterns, termed, respectively, “microbe-associated molecular patterns” and “pathogen-associated molecular patterns”
[11][2].
2. Ocular Surface Microbiota Composition
Even though the ocular surface is directly exposed to the same external environment, evidence supports the existence of a unique OSM compared to those related of the facial skin and oral mucosa
[12][3]. Nevertheless, the presence of a stable “core ocular surface microbiota” has been extensively debated and is still under discussion. In the past, the OSM’s composition has been characterized using culture-based methods. In this regard, by culturing swabs of healthy conjunctiva, the most commonly found commensal Gram-positive bacteria, although present in low amounts, were
Staphylococcus,
Corynebacterium,
Streptococcus, and
Propionibacterium [13][4]. Furthermore, less frequently, Gram-negative bacteria, including
Haemophilus,
Neisseria, and
Pseudomonas genera, and fungi were also identified
[13,14][4][5]. However, traditional culture methods are nowadays rarely used due to their inability to detect the complete OSM since these techniques cannot identify slow-growing bacteria as well as uncultivable species
[15,16,17][6][7][8]. In this respect, in recent years, the sequencing of bacterial 16S rRNA has been employed to overcome these limitations and to obtain an accurate and complete composition of the healthy human OSM. In 2007, Graham et al. performed the first high-efficiency study in which they identified the bacterial genera present in the conjunctiva samples from 57 healthy individuals using both traditional culture methods and 16S rRNA sequencing
[18][9]. As expected, they found more genera employing the molecular analysis (coagulase-negative
Staphylococcus sp.,
Staphylococcus epidermidis,
Rhodococcus erythropolis,
Corynebacterium sp.,
Klebsiella sp.,
Propionibacterium,
Bacillus sp., and
Erwinia sp.) with respect to the culture procedure (only the coagulase-negative
Staphylococcus and
Bacillus sp.)
[18][9]. However, the composition of this “core” microbiota has been later disputed and other groups proposed a different putative commensal OSM. For instance, Dong et al., by investigating the OSM from the conjunctival swab of four subjects, discovered 12 genera (
Pseudomonas,
Propionibacterium,
Bradyrhizobium,
Corynebacterium,
Acinetobacter,
Brevundimonas,
Staphylococci,
Aquabacterium,
Sphingomonas,
Streptococcus,
Streptophyta, and
Methylobacterium) regularly present in all the examined samples, which may constitute the presumed conjunctival core microbiota. Specifically,
Proteobacteria (64%),
Actinobacteria (19.6%), and
Firmicutes (3.9%) were found to be the most abundant phyla
[17][8]. Another group classified the bacteria from 105 individuals’ healthy conjunctivae
[19][10], confirming the three previously reported predominant phyla (
Actinobacteria (46%),
Proteobacteria (24%), and
Firmicutes (22%)) and six different genera (
Corynebacterium,
Streptococcus,
Propionibacterium,
Bacillus,
Staphylococcus, and
Ralsontia). In another study, Huang et al., analysing 31 conjunctival samples from healthy subjects, corroborated the existence of the three aforementioned phyla as the most abundant (
Proteobacteria (46%),
Actinobacteria (33.9%), and
Firmicutes (15.5%)) and also identified ten different genera (
Corynebacteria,
Pseudomonas,
Staphylococcus,
Acinetobacter,
Streptococcus,
Millisia,
Anaerococcus,
Finegoldia,
Simonsiella, and
Veillonella), which could form the supposed OSM “core”
[20][11]. A further relevant study was conducted by Doan and co-workers, who explored the ocular conjunctival microbiota of healthy individuals by using three different techniques (bacterial culture, 16S rDNA gene deep sequencing, and biome representational in silico karyotyping). They found that
Corynebacteria,
Propionibacteria, and the coagulase-negative
Staphylococci were the most abundant organisms in these samples
[12][3]. Last, always in search of the putative OSM “core”, two more recent studies examined healthy conjunctival samples. In detail, Li et al. observed the
Pseudomonas,
Acinetobacter,
Bacillus,
Chryseobacterium, and
Corynebacterium genera in 54 conjunctival swab samples
[21][12], while Ozkan et al., testing the conjunctiva of 43 healthy individuals, reported again the prevalent presence of the three above mentioned phyla (
Proteobacteria,
Actinobacteria, and
Firmicutes) with the addition of five genera (
Corynebacterium,
Sphingomonas,
Streptococcus,
Acinetobacter, and
Anaerococcus) in one or more subjects. Nevertheless, no specific genus was found in all the individuals, suggesting that the ocular surface does not possess a specific microbiota core signature
[22][13]. However, it should be also underscored that the16S rRNA sequencing has some limitations. For instance, it can detect only bacteria at the level of the genus while it cannot evaluate the functional status of the microbiota
[23][14]. In this regard, some research groups, using next-generation sequencing assays, identified not only bacteria but also other microorganisms, including virus and fungi. For example, Doan et al. reported the presence of the Torque teno virus (belonging to the
Anelloviridae family, mainly transmitted by faecal–oral route and detected in several tissues) on the ocular surface of healthy subjects
[12][3]; furthermore, two fungal phyla (
Basidiomycota and
Ascomycota) and five genera (
Malassezeia,
Rhodotorula,
Aspergillus,
Davidiella, and
Alternaria) were also isolated from >80% of 45 samples analysed
[24][15]. Moreover, other viruses have been found at the ocular surface of healthy individuals employing the neutralization assay
[25][16].
Overall, even if metagenomic sequencing has revealed the presence/absence of specific microbial species not previously identified by the traditional culture-based methods, to date, the existence of a stable and unique OSM “core” remains still unclear. In fact, although distinct studies used similar sequencing techniques, they obtained different results. These discrepancies could be due to several factors, including sample size, methods of sampling, contaminations from the DNA extraction kit and polymerase chain reaction reagents, and depth of sampling
[26,27,28,29][17][18][19][20]. Indeed, in regard of the latter, it should be also emphasized that the OSM’s composition appears to have a vertical stratification: for instance, by swabbing the ocular surface with a light pressure, opportunistic and environmental species, such as
Rothia,
Herbaspirillum,
Leptothrichia, and
Rhizobium, located in the superficial layer, as well as
Firmicutes (
Staphylococci) and
Actinobacteria (
Cornyebacteriae), placed in the mucosal layer, can be isolated. Meanwhile,
Proteobacteria (
Bradyrhizobium,
Delftia, and
Sphingomonas) situated in the conjunctival epithelium require deep swabbing
[17][8]. As a consequence, an accurate analysis of the OSM requires (i) a thorough and deep sampling across the different ocular layers; (ii) a combination of culture-based methods and conventional 16S rRNA techniques; and (iii) the use of alternative approaches, such as transcriptional assays, which can measure the activity of the whole microbiota and identify organisms also according to their species or strain
[4,17,30,31][8][21][22][23].
3. Factors Influencing the Ocular Surface Microbiota Composition
The ocular surface is a dynamic ecosystem and the normal composition of its microorganisms might be hypothetically influenced by different inherent factors, including age, sex, ethnicity, and geographic location, as well as by acquired factors, such as the use of contact lenses, ophthalmic antibiotics, and eye drops
[6,32][24][25].
Regarding age, some studies reported OSM changes from birth to adulthood
[19,33,34,35,36][10][26][27][28][29]. For instance, one of the first studies performed by Isenberg et al. underscored qualitative differences between conjunctival microbes of infants born by vaginal delivery and via caesarean section. In detail, the first group of infants had an OSM similar to the one described in the vagina (
Lactobacillus,
Bifidobacterium,
Diphtheroids,
E. coli,
S. epidermidis, and
Bacteroides); instead, the second group showed bacteria resembling the normal skin flora (
Corynebacterium,
Propionibacterium,
Diphtheroids, and
S. epidermidis)
[34][27]. Moreover, another study reported that the coagulase-negative
Staphylococcus and the anaerobe
Propionibacterium represent the most abundant organisms found in infants regardless of the mode of delivery
[36][29]. Interestingly, it has been observed that already two days after birth the OSM starts to change, where
S. epidermidis,
E. coli, and
S. aureus are the most frequent bacteria
[35][28] and continues to evolve over time. Indeed, some groups found significant differences in the bacterial composition of people at different ages
[19,33][10][26]. For instance, Cavuoto et al., analysing samples from children (<18 years old) and adults, showed a higher abundance of bacteria both at the phylum level (
Proteobacteria,
Firmicutes,
Bacteroidetes,
Fusobacteria) and at the genus level (
Streptococcus,
Staphylococcus, and
Brachybacterium) in paediatric samples as compared to adults; moreover, they found that the phylum
Actinobacteria and the genera
Corynebacterium,
Paracoccus, and
Propionibacterium are more abundant in adults
[33][26]. This evidence is similar to the results found by Zhou et al. when comparing children (<10 years old) and individuals > 10 years old. Indeed, they reported a higher richness and Shannon diversity index (used to measure the species diversity) in the first group as compared to the second one
[19][10]. However, other studies reported contrasting results
[22,37][13][30]. For instance, Ozkan et al. found no effect of age on OSM composition; instead, Wen et al., analysing the OSM of individuals aged between 28 and 84 years, reported a higher Shannon index in the elderly group
[22,37][13][30].
Regarding sex, although some studies did not report OSM differences between males and females
[19[10][26],
33], others described a sex association. In respect of the latter, Shin et al. found a higher abundance of
Acinetobacter and members of
Enterobacteriaceae, as well as a decreased amount of
Anaerococcus, in females
[38][31]; Ozkan observed a higher Shannon diversity index in males as compared to females but no difference in richness
[22][13]. Moreover, Wen reported differences at the genus level, with a significant decrease in
Propionibacterium acnes and
S. epidermidis in males with respect to females and an increase of
E. coli in females
[37][30]. Last, in the context of ethnicity and geographic location, the OSM composition does not seem to be influenced by these two factors
[19][10].
In terms of acquired factors, the use of contact lenses, ophthalmic antibiotics, and eye drops has been associated with an altered OSM. For instance, one study reported a higher abundance, at the level of the skin below the eye, of opportunistic pathogens, including
Pseudomonas,
Acinetobacter,
Lactobacillus, and
Methylobacterium, and a lower amount of typical ocular surface genera, such as
Staphylococcus and
Corynebacterium, in the OSM of contact lens wearers as compared to the non-wearer ones
[38][31]. Furthermore, another study underscored slight microbial variability between orthokeratology lens wearers and no-lens wearers, as well as between soft contact lens wearers and no-lens wearers. In this respect, the first group (orthokeratology lens wearers) had less abundance of
Bacillus,
Tatumella, and
Lactobacillus as compared to the group not using lens; instead, the soft lens wearers group showed less
Delftia and more
Elizabethkingia abundance with respect to no-lens wearers
[39][32]. As expected, the use of ophthalmic antibiotics may also negatively impact OSM. In this regard, Dave et al. reported a significant change in OSM composition after treatment with azithromycin and fluoroquinolonic antibiotics. In detail, they treated 6 patients with azithromycin and 18 patients with fluoroquinolonic antibiotics finding that in azithromycin-treated individuals the amount of the Gram-positive
S. epidermidis and
S. aureus was, at baseline, 54.5% and 18.2%, respectively, and 90.9% and 4.5% after antibiotic exposure; in fluoroquinolone-treated patients, the percentages of these bacteria were, at baseline, 45.7% and 6.5%, respectively, with an increase to 63.4% and 13% after intravitreal antibiotic injection. Moreover, treatment with fluoroquinolonic antibiotics decreased the amount of Gram-negative species from 8.7% at baseline to 1.6%
[40][33]. Other studies described the effects of fluoroquinolone (levofloxacin and moxifloxacin) treatment on OSM. After levofloxacin topical use, Ono et al. found a reduction in ocular bacterial diversity, while after ocular moxifloxacin administration, Celebi et al. reported a decrease in the coagulase-negative
staphylococci (10% vs. 50%),
S. aureus (5% vs. 20%), and
Corynebacterium (5% vs. 15%) in the treated group as compared to the control one
[41,42][34][35]. Regarding moxifloxacin treatment, another clinical study conducted on contact lens wearers found a reduction in Gram-positive commensal bacteria and no change in the amount of Gram-negative bacteria after antibiotic administration
[43][36]. Last, another acquired factor that might influence OSM composition is eye drop use. In this regard, one study described the effect of the eye drops employed to treat glaucoma on conjunctival bacteria, reporting a lower ocular culture-positive rate of bacteria in the treated group as compared to the control one and the presence of Gram-negative bacteria only in the eye drop users
[44][37]. Similar findings were also reported in patients treated with eye drops for curing dry eye syndrome
[45][38].