Detection of Chlamydiaceae and Chlamydia-like organisms on the ocular surface of children and adults from a trachoma-endemic region

Trachoma, the leading infectious cause of blindness, is caused by Chlamydia trachomatis (Ct), a bacterium of the phylum Chlamydiae. Recent investigations revealed the existence of additional families within the phylum Chlamydiae, also termed Chlamydia-like organisms (CLOs). In this study, the frequency of Ct and CLOs was examined in the eyes of healthy Sudanese (control) participants and those with trachoma (case). We tested 96 children (54 cases and 42 controls) and 93 adults (51 cases and 42 controls) using broad-range Chlamydiae and Ct-specific (omcB) real-time PCR. Samples positive by broad-range Chlamydiae testing were subjected to DNA sequencing. Overall Chlamydiae prevalence was 36%. Sequences corresponded to unclassified and classified Chlamydiae. Ct infection rate was significantly higher in children (31.5%) compared to adults (0%) with trachoma (p < 0.0001). In general, 21.5% of adults and 4.2% of children tested positive for CLOs (p = 0.0003). Our findings are consistent with previous investigations describing the central role of Ct in trachoma among children. This is the first study examining human eyes for the presence of CLOs. We found an age-dependent distribution of CLO DNA in human eyes with significantly higher positivity in adults. Further studies are needed to understand the impact of CLOs in trachoma pathogenicity and/or protection.

Broad-range Chlamydiae real-time polymerase chain reaction and sequence classification. Among 189 swab samples in total, 68 (36%) were positive for Chlamydiae DNA. All the nucleic acid extraction negative controls and real-time polymerase chain reaction (PCR) negative controls remained negative when tested with Chlamydiae real-time PCR. Although no swab has been taken as a field control, the risk of contamination can be ruled out because of existing differences in the distribution of positive samples for Chlamydiae real-time PCR and various profiles of sequences corresponding to different Chlamydiae families among children and adults as well as cases and controls. There were no statistically significant differences in positivity for Chlamydiae or C. trachomatis between male and female participants (p > 0.05). The average threshold cycle (C T ) value for Chlamydiae real-time PCR was 34 ± 9.4. A standard curve was prepared to evaluate the efficiency of real-time PCR and to achieve a comparative concentration (inclusion forming units (IFUs)) for the samples based on their C T values ( Supplementary Fig. S1). Among the 68 positive samples, 55 (81%) yielded assembled forward and reverse sequences (~200 bp) that were used for BLAST-n analysis, Naïve Bayesian classification in RDP, and phylogenetic analysis (  Tables S1 and S2). We were not able to assemble the sequences of the 13 remaining samples owing to multiple peaks in chromatogram lanes, which either might suggest the presence of more than one member of Chlamydiae in the ocular swab, or low amounts of DNA in the sample, which can cause PCR-generated artefactual species variation [38][39][40][41] .
BLAST-n results for sequences were categorized based on the first ( Table 2) and second BLAST-hit identification (excluding uncultured bacteria) (Supplementary Table S2) and closest 16 S ribosomal RNA (16S rRNA) classification match (Table 2). Of 16 samples that were detected as Chlamydiales bacterium by first BLAST-hit identification, nine sequences had highest similarity to Chlamydiales bacterium CRIB 32 and CRIB 33, six to Chlamydiales bacterium NS11, and one to both Chlamydiales bacterium CRIB 32/NS11 (Table 2). Sequences for Chlamydiales bacterium CRIB 32/33 and Chlamydiales bacterium NS11 were previously identified in domestic water and human nasal samples, respectively 42,43 .
Based on this 16S rRNA classification, four sequences corresponded to new species levels, 21 to new genus levels, and one to a new family level of the phylum Chlamydiae. Altogether, 54 out of 55 assembled sequences were identified as the nearest neighbor to four families including Chlamydiaceae (55.6%), Parachlamydiaceae (29.6%), Simkaniaceae (13%), and Criblamydiaceae (1.9%) ( Table 2). Additionally, the retrieved 16S rRNA gene sequences were classified using the Naïve Bayesian Classifier of RDP, by which the majority of the sequences were also placed within Chlamydiaceae (52.7%) and Parachlamydiaceae (27.3%). Owing to length, the remaining sequences could only be assigned as unclassified Chlamydiales (20%) ( Table 2). In general, the prevalence of Chlamydiaceae sequences was significantly higher in children than in adults (p = 0.0069) and CLO sequences were more frequent in adults than in children (p = 0.0003) (Figs 4 and 5).
Phylogenetic analysis of sequences. Phylogenetic analyses were performed with two different datasets to obtain better insights into the taxonomic placement of the 16S rRNA sequences from this study. The sequences obtained in this study were added to a Bayesian inference tree of full-length chlamydial 16S rRNA gene sequences representing the major known chlamydial lineages using Parsimony (Fig. 2) (Supplementary  Table S1 and Supplementary Fig. S2). Similar to BLAST-n results, the majority of sequences were assigned to the Chlamydiaceae (52.7%). The sequences affiliating with CLOs grouped together with sequences of members of the Parachlamydiaceae (9.1%) and Criblamydiaceae (1.8%). A total of 36.3% of the sequences were closely related to as yet unclassified chlamydial sequences and reflect the high diversity within the Chlamydiae. A maximum likelihood phylogenetic tree (Fig. 3) revealed the phylogenetic relationship of the obtained 16S rRNA gene sequences to their closest BLAST-hit including these sequences along with some reference sequences.
Prevalence of Chlamydiae DNA in children. Of 96 samples from children (54 in the trachomatous inflammation-follicular (cTF) group and 42 in the control (cC) group), 32 (33.3%) yielded positive results for Chlamydiae (25 in the cTF group and 7 in the cC group). The prevalence of Chlamydiae DNA in the cTF group (45.3%) was significantly higher than that in the cC group (16.7%) (p = 0.0032). The average C T values in the cTF and cC groups tested using Chlamydiae real-time PCR were 30.8 ± 9.2 and 35.7 ± 0.7, respectively (p = 0.0154) (Fig. 6a).  Among the 32 positive samples, 27 sequences were assembled. Sequences from 26 samples corresponded to various families: 22 (84.6%) were positive for Chlamydiaceae, three (11.5%) for Parachlamydiaceae, and one for (3.8%) Criblamydiaceae (Table 2 and Fig. 4a). In addition, one sample (cTF092) exhibited 87% similarity to the order Chlamydiales and appeared to be phylogenetically close to unclassified Chlamydiales (Table 2 and Fig. 2).
In the cTF group, among the 19 sequences classified as Chlamydiaceae, 17 (89.5%) showed 99-100% homology to C. trachomatis (p < 0.0001). These results were confirmed using omcB real-time PCR. In general, C. trachomatis  Table 2. BLAST-n analysis and RDP classification of Chlamydiae 16 S rRNA sequences detected in ocular swabs of children and adults. * Uncultured bacteria were excluded from final BLAST hits by choosing the "Exclude" option in BLAST-n page for "Uncultured/environmental sample sequences". # Accession numbers for submitted sequences to the NCBI database (https://www.ncbi.nlm.nih.gov/genbank/). § GenBank accession number was not provided for the cTF-092 because the BLAST query coverage was less than 90% to other Chlamydiales 16S rRNA.
sequences had a frequency of 31.5% (17/54) among the cTF group. The remaining two positive samples showed sequence homology with Chlamydia caviae/Chlamydia felis and Chlamydia gallinacea and remained negative when tested with omcB real-time PCR. Among three assembled sequences in the cC group, two revealed highest similarity to both C. caviae/C. felis and one to only C. caviae ( Table 2). None of the samples in the cC group exhibited similarities to C. trachomatis or had positive testing results using omcB real-time PCR. Moreover, among the three sequences classified as Parachlamydiaceae in the cTF group, two sequences showed highest similarity to Parachlamydia and one to Neochlamydia. In addition, one sample (cTF051) was classified as Criblamydiaceae (Table 2). This is also reflected in the phylogenetic trees (Figs 2 and 3). Altogether, C. trachomatis comprised the majority of 16S rRNA sequences compared to non-trachomatis Chlamydiaceae (NTC) and all CLOs among children (C. trachomatis vs. NTC and CLOs, p = 0.0005 and p = 0.0001, respectively) (Fig. 7a). The frequency of Chlamydiaceae sequences was significantly higher than that of all classified or unclassified CLOs among the children samples (p < 0.0001) (Figs 4a and 5a). No significant differences were found between cTF and cC groups with respect to the affiliation to different CLO families (p > 0.05) (Fig. 7a).

Prevalence of Chlamydiae DNA in adults.
Overall, 36 out of 93 (38.7%) swabs taken from adults tested positive for Chlamydiae DNA. Prevalence of Chlamydiae in the adult control (aC) group (20/42) was higher than that in adults with trachomatous trichiasis (aTT) (16/51); however, this difference was not significant (p > 0.05). The average C T values in the aTT and aC groups tested using Chlamydiae real-time PCR were 36.0 ± 0.45 and 36 ± 0.8, respectively (p > 0.05) (Fig. 6a). The majority of sequences from adults were identified by BLAST search as being similar to the Parachlamydiaceae family (13 sequences, 46.4%), eight (28.6%) similar to the Chlamydiaceae and seven (25%) to the Simkaniaceae family ( Table 2 and Fig. 4b). Phylogenetic analysis of the sequences from adults, including many lineages of yet uncultured chlamydial species, revealed three main groups including unclassified Chlamydiales (64.3%), Chlamydiaceae (25%), and Parachlamydiaceae (10.7%) (Fig. 5b).
The distribution of Chlamydiaceae sequences within the aTT group consisted of one sequence that showed highest similarity to genus Chlamydia, one to C. pecorum, and three that showed equivalent identification as either C. caviae or C. felis. With regard to the three controls positive for the Chlamydiaceae family, two sequences revealed highest similarity to C. trachomatis and also tested positive in omcB real-time PCR. The remaining sample exhibited equivalent similarity to C. caviae/C. felis (Table 2).
In general, the number of sequences classified as CLOs within the adult group was significantly higher than that classified as Chlamydiaceae (p = 0.0137) (Figs 4b and 5b). Nevertheless, no significant differences were found between aTT and aC groups with respect to the affiliation to different Chlamydiae families (p > 0.05) (Fig. 7b).

Discussion
This study provided evidence for the presence of Chlamydiaceae and CLO DNA in human ocular samples of patients with trachoma and healthy controls in Sudan. Overall, 36% of all samples tested positive for Chlamydiae DNA. Whereas the prevalence of C. trachomatis was significantly higher in children than in adults, a significantly higher distribution of CLO DNA was detected in adult samples compared to those from children. However, based on differences in the occurrence of identified CLO sequences among the patients with trachoma and healthy controls in children and adults, a pathogenic role for these bacteria in trachoma is not evident.
To our knowledge, this study is the first to examine human ocular samples for the presence of all members of the Chlamydiae. Whereas 22.9% of children' and 8.6% of adults' samples corresponded to Chlamydiaceae sequences, amplicons of 21.5% of adults' samples and 4.2% of children samples corresponded to sequences originating from different CLOs. These results are comparable with a prior study on cats with and without (controls) ocular diseases 16 . Specifically, C. felis, the known pathogen for feline chlamydial conjunctivitis 44 , was found among the majority of diseased eyes in animals under the age of five. The distribution of positive samples for non-C. felis Chlamydiales increased from 30% for cats under the age of one to 54% for cats older than 10 years in animals with and without ocular diseases 16 . Furthermore, in two separate studies, in nasopharyngeal samples from healthy children and adults with and without pneumonia, 11.4% of children samples and 36% of samples from adults tested positive for members of various families in the phylum Chlamydiae 43,45 . It is not clear why the distribution of CLOs in adults was more than five times higher than that in children. Although this difference might constitute a bias resulting from the small sample size in each group, this does not appear likely.
In the present study, we found a higher diversity of CLOs in adults than in children. Some of the analyzed sequences belong to members of the Parachlamydiaceae, which have previously been detected in other mammalian tissues 46 . Parachlamydia spp. have been found directly and indirectly in human corneal samples and disposable contact lenses, respectively 19,47 . To our knowledge, this is the first report of Criblamydiaceae in human eyes. The majority of sequences obtained, however, were affiliated with various as yet uncharacterized chlamydial lineages. This is unexpected, because these lineages are generally to date represented only by environmental clone sequences, with members of such lineages not having been considered to be also associated with humans. Although the sequences belonging to these divergent CLOs have mainly been detected in healthy participants and may also originate from other sources, it is still noteworthy to find such a diversity of CLOs sequences in human eyes. Considering that similar profiles of CLOs were obtained among the aTT and aC groups, we cannot suggest any association of these bacteria with trachoma in adults. In children, the distribution of CLOs did not significantly differ between the cTF and cC groups.
Several studies have reported the pathogenic impacts of Parachlamydia spp. in respiratory disorders 6,20,21,26 and S. negevenesis in patients with bronchiolitis, although S. negevenesis pathogenic potential remains controversial 6,20,21,[48][49][50] . Notably, the possible role of some CLOs in respiratory tract infections was mainly suspected in pediatric populations; similarly, in the present study, the only limited evidence suggesting a role of Parachlamydiaceae and Criblamydiaceae in trachoma pathogenesis was documented in children. In our study, no strong evidence of the pathogenic impact of CLOs on adult human eyes was obtained, which supports previous results by Vidgen et al. suggesting CLOs as commensals of the ocular microbiota in koala 51 . Estimated loads of CLOs in this study are comparable with the reported load of the bona fide microbiome: 100 to 1000 colony-forming unit (CFU) per mL in human tear fluid 52 . These results are in line with several studies reporting a non-significant distribution of Parachlamydiaceae and other CLOs among diseased and healthy eyes of sheep, pigs, and cats [16][17][18]53 . Moreover, CLOs have been detected in different samples such as skin, nose, and cervicovaginal samples of healthy individuals 22,38,43,45 .
The high prevalence of Chlamydiae DNA in the present study is comparable with findings of human skin and nasal samples, which support the idea that ocular microbiota colonizes in parallel with microbiota of the upper respiratory system and skin 22,43 . This high prevalence compared to results from human respiratory or cervicovaginal samples can be explained through the frequent contact of the ocular surface directly or indirectly with a source of bacteria in the environment, infected animals, and humans. In particular, the majority of unclassified chlamydial sequences in this study demonstrated highest homology to previously isolated sequences from water. Additionally, members of Parachlamydiaceae are known to naturally infect amoeba, which they utilize as a replicative niche 23,42,[54][55][56] . Moreover, Simkania and Criblamydia have been detected repeatedly in domestic waters 14,42,54,55,57,58 , which highlights the importance of water as a rich source of CLOs 5 and possibly a main source of ocular infection with Chlamydiae. These results also emphasize the validity of environmental improvement as an essential factor of the SAFE strategy (Surgery for trichiasis, Antibiotics to clear ocular infection, promotion of Facial cleanliness, and Environmental improvement).

Figure 2.
Phylogenetic affiliation of 16S rRNA gene sequences detected in this study. A Bayesian tree built from full-length chlamydial sequences is shown, to which the partial sequences generated in this study were added using Parsimony. Only sequences closely related to the sequences found in this study are depicted. The full dendrogram including all used sequences is available as Supplementary Fig. S2. Branches belonging to the Chlamydiaceae and CLOs are marked (light and dark purple, respectively). Sequences of cases (purple) and controls (dark purple) of children (cTF and cC, respectively), and cases (dark blue) and controls (cyan) of adults (aTT and aC, respectively), are marked.  The profile of the Chlamydiaceae community present in ocular mucosa differed significantly between adults and children. Whereas C. trachomatis was only detected in two (2.1%) adult samples, 17 samples (31.5%) of children with TF were found to be positive for C. trachomatis. These results are consistent with previous data suggesting a central role for C. trachomatis in active trachoma among children 1,59,60 . Notably, the only two samples from adults that yielded positive results for C. trachomatis infection in the present study were in the aC group. The absence of infection with C. trachomatis in adults with trachoma can be explained through: [1] the effects of non-C. trachomatis bacteria and mechanical irritation on the eyes [30][31][32]34,61,62 ; [2] the development of acquired immunity to C. trachomatis and its impact on the duration of ocular infection, which decreases significantly with age 28,34,59,61,63 ; and [3] the kinetics of the disease, which can be explained by the recovery phase after repeated infections with C. trachomatis 60 . During this time, patients can be diagnosed positive for trachoma although Chlamydia is no longer present in the eyes 60 .
Classification of the Chlamydiaceae positive sequences into C. trachomatis and NTC provided an opportunity to investigate the prevalence of NTC bacteria. Overall, we found no association between NTC species and trachoma in this population. Although NTC species in the aTT group (9.8%) compared to the aC group (2.4%) were more frequent, this difference was not statistically significant. This is in contrast to previous studies, which found a high prevalence of NTC species in trachoma samples from Nepal 28,29 . Differences in NTC prevalence between present data and previous studies by Dean et al. 28,29 could be due to the small sample size in all of these studies and regional variations.
Unfortunately, BLAST results for the majority (8/11, 72.7%) of sequences positive for NTC species could not differentiate between C. caviae and C. felis as the same identification score was obtained for both species. Moreover, the non-significant distribution of sequences positive for C. caviae/C. felis between cases compared to controls was unable to provide a clear indication of their role in this study. C. caviae and C. felis cause keratoconjunctivitis in guinea pigs and cats, respectively 16,19,28,64,65 . Previous studies have also reported the presence of C. caviae and C. felis in human eyes 19,66,67 . In our study, one sequence obtained in the aTT group and one in the cTF group demonstrated highest homology to C. pecorum and C. gallinacea, respectively. The presence of C. pecorum in an ocular sample from a patient with trachoma has also been previously reported 29 . To the best of our knowledge, this is the first molecular evidence of the presence of C. gallinacea DNA in humans, a species recently added to the Chlamydiaceae that was reported for the first time in cattle by Li et al. 68 and subsequently described as a causative agent of avian chlamydiosis with zoonotic potential [69][70][71] .
C T values revealed important information about the bacterial load in each sample. The average C T value for children infected with C. trachomatis corresponds to 1.000 to 10.000 IFU per swab (Fig. 6c), whereas those for NTC and CLO positive samples correspond to 10 to 100 IFU per swab (Fig. 6b-d). These data support prior findings that active infection (higher loads) of C. trachomatis is associated with active trachoma 61,[72][73][74] . Furthermore, consistent with previous data, we found higher loads of bacteria in samples from children 59,72,73 , suggesting this age group as the major source of C. trachomatis infection 59 . Higher C T values, which are corresponding to lower loads of bacteria in positive samples for CLOs might be due to the persistent form of infection in these samples as it has been described earlier 74,75 . Nevertheless, high C T values may also be the consequence of an early infection, the presence of residual DNA from a previous infection, or inoculation of the eye with dead microorganisms 28,61 . Our study also provides evidence that the presence of CLOs and C. trachomatis in human eyes was influenced by age. Previously, 16S rRNA gene sequencing revealed a highly diverse bacterial community in human eyes [76][77][78] . Moreover, age-dependent changes in the ocular microbiota have been reported in prior studies 79,80 . It has been shown that the ocular microbiome of individuals with healthy conjunctiva has a significantly different bacterial abundance and diversity in children aged ≤10 compared to older participants 80 . In addition, recent studies have uncovered strong evidence regarding the impacts of ocular commensal microbiota on regulating local immune responses and developing defense mechanisms against colonization of pathogens in the eyes 52,79,81,82 . The absence of CLOs and C. trachomatis in children and adults can thus be partially explained owing to age-dependent changes in the microbiota of eyes and its impacts on the development of different local immune responses. This may in turn influence the presence of Chlamydiae and multiplicity of infection. For example, a study on the cervical microbiota of women infected with C. trachomatis and healthy controls have shown a different diversity of microbiota in those with asymptomatic infection of C. trachomatis 83 . Additionally, the abundance of bacterial taxa was different in the cervical microbiota of women infected with C. trachomatis compared to that in healthy controls 83   This study has several limitations. As the sample size was relatively small, the use of a larger sample size within this study might have minimized the bias in biodiversity of detected CLOs in each group and provided a better understanding of the possible associations of CLOs and NTC species with eye health or disease. In addition, the length of the amplicons used for sequencing was not sufficient to resolve the species (i.e. C. felis/C. caviae) or genus level in every case, due to the sequence identity or low representation of many chlamydial lineages in public databases, respectively.
In summary, C. trachomatis is associated with active trachoma in Sudanese children. We presented evidence demonstrating an age-dependent distribution of divergent groups of CLOs in human eyes. However, the biology of the CLOs identified in this study remains unknown. Our findings do not support a significant association between the presence of CLOs and NTC species and active trachoma in children or TT in adults. Comparative studies with a larger sample size on the prevalence of CLOs in human eyes from different geographic regions may provide insight into the possible role of these microorganisms in pathogenicity and/or protection. Further studies are needed to understand the impacts of ocular microbiota and local immune responses on the presence of C. trachomatis and CLOs in the eyes.

Material and Methods
Ethical approval. The study was conducted in accordance with the Declaration of Helsinki. The National Ethics Authorities in Sudan (Sudanese National Research Ethics Review Committee (NRERC) (No. 174-8-12)) and the ethical committee of the Medical University of Vienna approved the study and all the included procedures. Written informed consent was obtained from all adult individuals at the time of sample collection. For those participants aged <18 years that wished to take part in the study, consent was obtained from a parent/ guardian. All samples were anonymized. Trachoma grading. Clinical phenotypes were assessed in the field by an experienced ophthalmologist trained in trachoma grading using the World Health Organization simplified grading system 85 . Individuals presenting with follicular trachoma were classified as TF, individuals with pronounced papillary hypertrophy and inflammatory thickening of tarsal conjunctiva were classified as "Trachomatous Inflammation − Intense (TI)" and individuals with tarsal conjunctival scarring as "Trachomatous scarring (TS)". Individuals with inturned eyelashes were graded as TT. Individuals with no clinical signs of follicles, papillary hypertrophy, or conjunctival scarring were classed as healthy controls. Cases and controls were matched by age and gender.
Study population and sampling. Participants for this study were recruited in trachoma endemic areas in Sudan 86 . For adults with TT, sample collection took place in a Field Surgery Clinic in the Al Qadarif region (Fig. 1). A total of 51 aTT scheduled for trichiasis surgery and 42 matched aC without history and present signs of trachoma were enrolled. In addition, 54 cTF and 42 cC subjects aged 1 to 9 years were recruited at two Quran Schools in the Al Qadarif region. Conjunctival samples were taken from the upper tarsal conjunctiva from both eyes using polyester flocked swabs (UTM-RT collection kits, Copan USA, Murrieta, CA, USA) using standard methodology 87,88 . The sampling ophthalmologist wore gloves that were changed after each individual. Swabs were stored in universal transport medium and frozen immediately in liquid nitrogen cryogenic shipping containers. In Austria, samples were subsequently stored at −80 °C. DNA preparation. Genomic DNA was extracted from the conjunctival swab samples taken from left eyes. As a nucleic acid extraction negative control, one empty swab was included during each DNA extraction run. Each swab was vigorously vortexed for 5 min inside the universal transport medium, then the entire transport medium was transferred to a new tube. Cells in the swab transport medium were pelleted by centrifugation at 17,000 × g for 30 min. DNA was extracted from the resuspended pellet in ATL buffer of the QIAamp DNA Investigator Kit (QIAGEN GmbH, Hilden, Germany), according to manufacturer instruction, and DNA was stored at −20 °C until further use.

Broad-range Chlamydiae real-time PCR and sequencing.
A Chlamydiae-specific real-time PCR was performed on a final reaction volume of 20 µL targeting a fragment of approximately 207 to 215 bp of the 16S rRNA gene as described by Lienard et al. 45 . Cycling conditions were 5 min at 95 °C, followed by 50 repetitions of 3-step cycles of 15 s at 95 °C, 20 s at 67 °C, and 15 s at 72 °C, all carried out in a PikoReal real-time PCR System (Thermo Fisher Scientific). All samples were tested in duplicate. After optimization of the PCR, a C T value of ≤37 has been suggested as a reliable cutoff to provide reproducible results and samples with this C T value were considered as positive samples. In each run, 2 wells were dedicated to negative controls (distilled water) and 2 wells were used as nucleic acid extraction negative controls. A standard curve was generated for each real-time PCR run using decimal serial dilutions of extracted genomic DNA from 1 × 10 6 to 1 × 10 1 IFUs of C. trachomatis ( Supplementary Fig. S1).
Purification of amplicons for positive samples was performed using the QIAquick PCR Purification Kit (QIAGEN GmbH) according to the manufacturer's protocol. Sanger sequencing of purified PCR products was performed by Eurofins Genomics AT (Vienna, Austria). Inner primers (panFseq and panRseq) resulted in an approximately 200 bp amplicon as explained previously by Lienard et al. 45 .
Specific TaqMan real-time PCR for the omcB gene of C. trachomatis. To assess C. trachomatis prevalence in particular, a specific real-time PCR was carried out on all samples to target the omcB gene of C. trachomatis. The reaction was performed in a final volume of 20 µL using iTaq Supermix (BioRad, Reinach, Switzerland), 0.3 µM concentration of each primer, 0.1 µM concentration of probe, and 5 µL sample DNA 89 . The program was set at 95 °C for 5 min, followed by 45 cycles of 15 s at 95 °C and 30 s at 60 °C, all carried out in a PikoReal real-time PCR System. All samples were tested in duplicate. Samples with a C T value of ≤37 were considered as positive. Distilled water (a negative control) and C. trachomatis DNA (a positive control) were included in each experiment. Data analysis. The obtained partial forward and reverse sequences of 16S rRNA gene were trimmed, aligned and assembled in the Geneious R10.2.2 software package. Of the resulting 55 sequences, 50 (90.9%) were of high quality with average phred quality scores above 30 and 5 (9.1%) of the sequences had average phred quality scores above 23 (cTF-025, aC-259, aC-261, aC-268 and aC-285). All consensus sequences were checked for the presence of chimera with DECIPHER 90 and compared with sequences available in the GenBank database of the National Center for Biotechnology Information via the BLAST server (https://www.ncbi.nlm.nih.gov/blast/). 16S rRNA gene identification cut-offs of 97, 95, and 90% were applied for the achieved sequences to approximately classify members of the phylum Chlamydiae to the species, genus, and family levels, respectively 91 . In addition, the 16S rRNA sequences were classified with the Naïve Bayesian classifier of RDP using a 70% confidence interval 92 . Obtained nucleotide sequences have been submitted to the NCBI database (https://www.ncbi.nlm.nih.gov/genbank/) under accession numbers (MH119764-99).
Based on the results from sequencing of 16S rRNA gene sequences and C. trachomatis omcB real-time PCR, samples assigned to the Chlamydiaceae were divided into two groups: C. trachomatis and NTC.
Statistics. Variables were compared by t test. A p-value < 0.05 was taken as statistically significant. Data analysis was performed using GraphPad Prism 6.0 (GraphPad Inc., La Jolla, CA, USA) software.