Genital and Oral Microbiome and Behçet’s Disease Activity

The aetiopathogeneses of Behçet’s Disease (BD) remains elusive with multifactorial genetic and epigenetic factors resulting in multisystemic disease. Oral and genital ulceration are common and inuences disease outcome. We hypothesised that dysregulation of genital and oral microbial communities contributes to BD disease activity. 153 BD patients’ samples, 70 matched oral and genital (Female: Male, 58:12; mean age, 42±13.9: 39.3±10.3), 12 unmatched samples; 16s rRNA sequencing utilised and V1/V2 and V3/V4 regions analysed. BD outcomes: oral and genital ulcer severity and BD activity scores, Psychological and Social Well-being scales, Headache Impact Test-6 (HIT-6) were included. All the analyses were performed with R software. The alpha and beta diversity had anatomical specicity, with signicant differences between genital and oral samples; p values<0.05 irrespective of presence or absence of ulcers. Interestingly, in the genital area Bacteroidota were present (G_U: 29% - 10%) and (G_nU: 27% - 14%) compared to less than 1% oral area of V1/V2 and V3/V4. Proteobacteria were uniquely present with (O_U: 9%) and (O_nU: 12%) in oral, and less than 0.01% in genital area for V3/V4 region. Gender anatomical specic communities were noted: females with genital ulcers Gardnerella, Lactobacillus, Atopobium were signicantly increased compared to than males, with V3/V4 analysis indicating that Lactobacillus and Gardnerella were signicantly increased by 20 times in females than males (p-adj <0.05). In contrast Peptoniphilus and Corynebacterium were signicantly increased in males than females. Streptococcus was signicantly increased with oral ulceration, while Veillonella was signicantly decreased in patients without oral ulceration. Colchicine had a signicant effect on the bacterial abundance irrespective of the presence or absence of ulceration. In this cohort, the WSAS (Work and Social Adjustment Scale) values were higher in active disease. Our results suggest that dysregulated microbial communities occur in BD. V1/V2 demonstrates that during episodes of ulceration the pathogenic bacteria genus Escherichia-Shigella appear in both oral and genital ulcers. V3/V4 outcomes show that ulceration in both regions is assigned to genus; Lachnospiraceae, Saccharimonidales, Coriobacteriales. Streptococcus is related to the presence of oral ulcers, while Veillonella is presence when patients are ulcers free may be a useful marker of disease regression. international instrument for the assessment BD activity in the of clinical features; mouth ulcer, genital ulcer, skin lesions, eye joint involvement, blood vessel involvement, gastrointestinal and CNS complications, preceding prior day of assessment [17]. Other factors: patients’ medication, and lifestyle (e.g., stress, smoking, alcohol intake, and diet), included.


Abstract Background
The aetiopathogeneses of Behçet's Disease (BD) remains elusive with multifactorial genetic and epigenetic factors resulting in multisystemic disease. Oral and genital ulceration are common and in uences disease outcome. We hypothesised that dysregulation of genital and oral microbial communities contributes to BD disease activity. 153 BD patients' samples, 70 matched oral and genital (Female: Male, 58:12; mean age, 42±13.9: 39.3±10.3), 12 unmatched samples; 16s rRNA sequencing utilised and V1/V2 and V3/V4 regions analysed. BD outcomes: oral and genital ulcer severity and BD activity scores, Psychological and Social Well-being scales, Headache Impact Test-6 (HIT-6) were included. All the analyses were performed with R software.

Results
The alpha and beta diversity had anatomical speci city, with signi cant differences between genital and oral samples; p values<0.05 irrespective of presence or absence of ulcers. Interestingly, in the genital area Bacteroidota were present (G_U: 29% -10%) and (G_nU: 27% -14%) compared to less than 1% oral area of V1/V2 and V3/V4. Proteobacteria were uniquely present with (O_U: 9%) and (O_nU: 12%) in oral, and less than 0.01% in genital area for V3/V4 region. Gender anatomical speci c communities were noted: females with genital ulcers Gardnerella, Lactobacillus, Atopobium were signi cantly increased compared to than males, with V3/V4 analysis indicating that Lactobacillus and Gardnerella were signi cantly increased by 20 times in females than males (p-adj <0.05). In contrast Peptoniphilus and Corynebacterium were signi cantly increased in males than females. Streptococcus was signi cantly increased with oral ulceration, while Veillonella was signi cantly decreased in patients without oral ulceration. Colchicine had a signi cant effect on the bacterial abundance irrespective of the presence or absence of ulceration. In this cohort, the WSAS (Work and Social Adjustment Scale) values were higher in active disease.

Conclusion
Our results suggest that dysregulated microbial communities occur in BD. V1/V2 demonstrates that during episodes of ulceration the pathogenic bacteria genus Escherichia-Shigella appear in both oral and genital ulcers. V3/V4 outcomes show that ulceration in both regions is assigned to genus; Lachnospiraceae, Saccharimonidales, Coriobacteriales. Streptococcus is related to the presence of oral ulcers, while Veillonella is presence when patients are ulcers free may be a useful marker of disease regression.

Background
The main entrances to the body from the external environment are the oral and genital cavities. Each has a unique microbiome and any alteration in the status of microbiome composition may trigger disordered function of the mucosal barrier 'dysbiosis' [1][2][3][4][5]. orogenital ulceration occurs in 85% or more patients with Behçet's Disease (BD), which is a complex in ammatory vasculitic disease which involves the mucous membranes, eyes, joints, central nervous system, vascular and gastro-intestinal systems. Oral ulceration can be seen in buccal and labial mucosa, tongue (all surfaces), and gingivae. They also appear in the soft and hard palate, the faeces and tonsillar bed and oropharynx; these ulcers cause a break in the epithelium, and disruption of the occlusive IgA1 mucosal barrier function which allows organisms to gain access to the immune system [3,6]. Genital ulceration in males occurs most frequently on the scrotum while ulcers on the shaft and glans penis are infrequent. In females, genital ulcers commonly occur on the labia majora and labia minora, vulva mucosa and rarely the cervix. There is complex interaction between environmental triggers and immune system of the patients own the patients' underlying genetic background.
Together these interactions can both increase mucosal ulceration, and the ulcers in turn may cause exacerbation of the systemic problems of BD. This observation has led us to believe that the composition and diversity of genital and oral microbiome may play a signi cant role in the aetiopathogenesis of BD. Next-generation sequencing (NGS) offered the opportunity of cultivation-independent assessment of microbial communities and therefore revealed a multitude of thus far unknown bacteria.
In healthy women, for instance, the vaginal microbiome, shows a predominance of Lactobacilli which promotes the maintenance of the vaginal homeostasis and prevents the colonisation and growth of adverse microorganisms. Lactobacilli protect against several urogenital and sexually transmitted infections. Therefore, disruption of this normally Lactobacillus-dominated microbial composition to a more complex, diverse polymicrobial community is thought to be one of the common causes of vaginal diseases [7], While a variety of bacterial species including Streptococcus spp. and Helicobacter pylori has been implicated in recurrent oral ulceration [8,9], and their role remained controversial. Our group study in 2015 showed increase in the colonisation of Rothia dentocariosa is found in the non-ulcer area of BD and Recurrent Aphthous Stomatitis (RAS). Moreover, S. salivarius and S. sanguis heavily colonises on active oral ulcers in BD. In contrast, there is an increase in the colonisation of Neisseria and Veillonella are found in healthy controls [10]. Other study found an increase in Haemophilus parain uenzae and decrease in Alloprevotella rava. Leptotrichia are also observed in salivary microbiome in BD. Interestingly, after dental and periodontal treatment, the salivary microbiome pattern stabilises to a more symbiotic pattern in short-term [11]. It is widely accepted that a variety of medications, not just antibiotics could affect and alter equilibrium of microbiome. The impact of several drugs on gut microbiome has been widely studied [12]. Similarly in Behçet's treatment with azithromycin decreased cutaneous folliculitis lesions and accelerated the healing of oral ulcers [13]. Additionally, minocycline successfully reduced the pro-in ammatory cytokines in vitro by their peripheral blood mononuclear cells when stimulated with Streptococcal antigen [14].
To date, the genital microbial pro le and the differences between ulcer and non-ulcer microbiome in genital mucosa has not been investigated in BD cohort. In this study, we perform a comparative analysis of genital and oral microbiome pro le from ulcerated and non-ulcerated mucosa of active and inactive BD patients.

Subjects and samples collection
This cross-sectional study of 153 BD samples were collected from the London Behçet's Centre, Royal London Hospital, Barts Health after ethical approval was granted [15]. Male and female patients were recruited with age range between 19 to 73 years old, those patients were divided into four groups: orally active (ulcerated) and orally inactive (non-ulcerated), genitally active (ulcerated) and genitally inactive (non-ulcerated). The BD patients' exclusion criteria were as follows: not ful lling the International Criteria of Behçet's Disease (ICBD 2014), pregnancy, age under 18 and using antibiotics in the last three months. Each sample was subjected to amplicon metagenomic sequencing followed by pro ling of microbial community taxonomic composition.
In total, 70 matched oral and genital samples: (F:M, 58:12; mean age, 42±13.9: 39.3±10.3), and extra 12 samples were not matched; 3 samples were only orally, and 9 samples were only genitally. One genital sample was excluded. Samples were coded as orally active, having oral ulcers (n=35; 23%) during the time of sampling; and (n=38; 24%) were orally inactive. In addition, those who had genital ulcers (who were genitally active, n=46; 30%) during the time of sampling, and samples from those with no genital ulcers (n=33; 21%) during the time of sampling.
Oral and Genital Ulcer Severity Score (OUSS and GUSS) forms These forms are an integral part of the clinical information collected by clinicians or senior nurses on the clinical assessment day. The OUSS and GUSS forms were developed and validated in the London Behçet's Centre. The data collected were ulcer characteristics (number, size, duration, ulcer-free period, pain and site, evidence of scars and discharge) were used to evaluate the severity of oral and genital ulceration. The oral health related quality of life (OHRQoL) and genital health related quality of life (GHRQoL) were also assessed [3,16].

Behçet's Disease Current Activity Form (BDCAF)
The BDCAF is a valid international instrument for the assessment BD activity in the clinic which scores the history of clinical features; headache, mouth ulcer, genital ulcer, skin lesions, eye symptoms, joint involvement, blood vessel involvement, gastrointestinal and CNS complications, which present over the preceding 4 weeks prior to the day of assessment [17]. Other factors: patients' medication, and lifestyle (e.g., stress, smoking, alcohol intake, and diet), were included.

Psychological and Social Well-being Scales
The Work and Social Adjustment Scale (WSAS) is valid and reliable self-report ve-item questionnaire of the functional impairment attributable to an identi ed problem or disorder on a 0 to 8 scale. The patient health questionnaire-9 and risk assessment (PHQ-9) evaluates depression and passive thoughts of death or self-injury within the last two weeks, score can range from 0 to 27, since each of the 9 items can be scored from 0 (not at all) to 3 (nearly every day). Generalised Anxiety Disorder Assessment (GAD7) measures severity of anxiety over the last 2 weeks, total score for the seven items ranges from 0 to 21. The Warwick-Edinburgh Mental Wellbeing Scale (WEMWBS) measures mental well-being and it is scored by summing responses to each item answered on a 1 to 5 Likert scale. Patients completed these questionnaires on the day of their clinical assessment.
Headache Impact Test-6 (HIT- 6) This questionnaire consists of six items used to measure a wide spectrum of the factors contributing to the burden of headache. The scores range between 36 and 78, the larger scores re ecting greater impact.

Experimental Design
Collection of oral and genital samples: Samples were collected as a part of the clinical assessment during patients' routine examination. The base of the ulcerated and non-ulcerated (oral mucosa, and genital mucosa/skin) from BD patients were biopsied with a nylon bristle cytology brush (Flowgen, Nottingham, UK). The brushes were dipped up and down 10 times in a 1.5ml Eppendorf containing 600µl of cell lysis solution (5 PRIME bacterial DNA Collection and Extraction Kit, 5 PRIME Inc., Gaithersburg, USA). The specimens were kept at room temperature and transported from the clinic to the laboratory.

Bacterial DNA extraction, PCR ampli cations and sequencing
The bacterial DNA of oral and genital swabs was extracted using the 5 PRIME bacterial DNA Collection and Extraction Kit (5 PRIME Inc., Gaithersburg, USA), according to the manufacturer's instructions. Brie y, after adding the proteinase K to the Eppendorf's containing the swabs were incubated for 3 hours 55 ºC.
Thereafter, the lytic enzyme was added to the samples, which were then incubated at 37 ºC for one hour. After this step, the RNase was added, and the samples were incubated for another 1 hour at 37 ºC. Finally, each DNA sample was puri ed and dissolved in 25µl of DNA hydration solution by incubating it overnight at room temperature. The concentration of DNA was quanti ed by using 1µl tested by CLARIOstar microplate (BMG LABTECH, Germany) using ELISA reader (OPTIMA software version 2, 00R3, UK). The DNA samples were stored at -80ºC until PCR ampli cations and sequencing applied.
Amplicon sequencing libraries were prepared for the Illumina MiSeq System following the Illumina 16S Metagenomics Sequencing Library Preparation guideline. 4 sets of primers were used (Table 1)  16s rRNA data analysis The raw paired-end FASTQ reads were imported into the Quantitative Insights into Microbial Ecology 2 program [19] and demultiplexed with the native plugin.
Raw reads were subsequently deposited into the European Nucleotide Archive (ENA).
Quality lter, trimming, denoise, mergepairs, and chimera removal of the data was carried out using the Divisive Amplicon Denoising Algorithm 2 (DADA2) [20]. To lter out true biological Amplicon Sequence Variants (ASVs) from spurious sequences generated during PCR ampli cation and sequencing, a minimum of 50X read was required for each ASVs.
The taxonomic assignment of the ASVs was performed using the feature-classi er 2 plugin [19,20], implemented in QIIME2 against the SILVA 138 SSU nonredundant database. A consensus con dence threshold of 0.6 was chosen.

Evaluation of variable regions information
All the information from the 4 regions were analysed, and alpha diversity was calculated for each region using the Shannon diversity index and the "Observed features" with native QIIME2 plugin.
The Variance In ation Factor (VIF) (is a measure of the amount of multicollinearity in a set of multiple regression variables) was calculated using the Python package 3.6 'statsmodels' [21] to identify collinearity among the alpha diversity indexes. A VIF of 1 indicates no correlation between the indexes. Instead, VIFs greater than 5 indicate that the indexes are correlated between them.

Statistical analysis
All analyses described were performed in R v.3.5.2 (R Development Core Team, 2018). Rarefaction curves were calculated to evaluate if the sequencing efforts generated enough data to well represent the overall microbial diversity in samples. The corresponding plot was generated with the phyloseq R package [22].
The Shannon index and Inverse Simpson α-diversity metrics were calculated to estimate the variation of bacterial diversity.
The Kruskal-Wallis H test for all and pairwise tests were used to compare the groups (Oral ulcer: O_U; Oral no ulcers: O_nU; Genital ulcers: G_U; Genital no ulcers: G_nU). Benjamini and Hochberg correction were used.
The Venn diagrams were created using the 'Venn Diagram' R package [23]. To represent the list of ASVs shared between oral and genital samples (ulcer and no ulcer) status. Before conducting the statistical analysis, a rarefying step was carried out based on a random sampling of each library. Community analyses (beta diversity) were performed with the Bray-Curtis quantitative distance metric (evenly sampled at 2,000 reads per sample) using the 'vegan' R package implemented in the 'phyloseq' R package [22].
The structure of microbial communities was explored using a Non-Multidimensional Scaling (NMDS) with 'k=3' as an ordination method. Distance-based redundancy analysis (db-RDA) was conducted to correlate the effects of different types of dispensed medications on microbial community structure and disease activity. This analysis was carried out using the "capscale" function implemented in the 'vegan' R package. The best db-RDA model was selected by using the vegan's 'ordistep' function.
The statistical signi cance of the models (see Supplementary materials for formula) was determined by the Adonis test (permutation-based ANOVA, PerMANOVA) [24] with 999 permutation-based Bray-Curtis distance matrices. PerMANOVA Pairwise contrast was performed, and the Benjamini-Hochberg FDR correction was used to calculate q-values. The test was performed using the "adonis" function in the vegan R package [25,26].
Pairwise comparisons for all pairs of levels of a factor was tested by using Permutational MANOVA with the adonis. pair function in the 'EcolUtils' R package [26,27]. Differential abundance analysis was completed using the DESeq2, R package [28]. Benjamini and Hochberg [29] procedure was used for multiple testing P-value adjustment, and a False Discovery Rate (FDR) of 10% was selected to denote statistical signi cance [28,30]. Heatmaps that represent the counts of the most abundant ASVs were produced with the 'pheatmap' R package [31].
The variation in terms of abundance of each ASV was estimated using the Deseq2 R package [28] performed between two groups both for oral and genital areas in different conditions.

Results
To characterize the microbiome composition in BD, we collected and analysed a total of 152 samples: oral and genital samples from each BD patient. The average age of BD patients (42.2±13.9 years) was not signi cantly different between males and females (t-test P value = 0.32). In total 61 (83%) patients were females, while 12 (17%) were males for the oral samples. 66 patients (85%) were females and 12 (15%) were males for the genital samples. During the sampling, 49 patients (68%) had evidence of active disease 26 patients (36%) were on systemic medication. At the time of oral samples collection, 20 patients (28%) from this cohort were using Triorasol mouthwash (TMW) ( Table 2).
Sequence Analysis and evaluation of variable primer regions outcomes About 1,001,682 million reads were obtained from 152 samples. After quality ltering, merging reads and chimera removal of the two Illumina runs, we got 970,790 sequences for 147 samples, with a median frequency of 3,715 reads per sample (were the max number of reads per samples obtained was 26,393 and minimum was 751). Also, it obtained 2,785 ASVs (amplicon sequence variants) [20].
The most abundant Phyla detected were similar in all four regions sequenced. The observed Genera distribution was slightly different but not signi cant. In addition, there was a high correlation between the alpha diversity values of the four regions. After the fragmentation and ampli cation steps in the library's preparation, all four hypervariable regions could not be traced back to the same original molecule [32] due to the ampli cation bias some bacteria will get counted twice, and others not. Due to abundant phyla similarity and alpha diversity correlated, statistical analyses for investigating the microbial community structure of BD, were not performed using all 4 regions simultaneously. The reads obtained from V1/V2 and V3/V4 showed a higher level of reading quality that suggested a greater probability of correctly identifying ASVs at taxonomic level regions. We therefore, decided to focus on the V1/V2 and V3/V4 hypervariable regions to compare the relative performance and analyses. In total we sequenced the V1/V2 and V3/V4 regions. Sequence classi cation was performed using the same approach for both data sets and we subsequently compared the resulting taxa identi ed.
Bacterial diversity of genital and oral samples in BD patients using V1/V2 and V3/V4 regions The Shannon diversity index and Inverse Simpson index were used to calculate and describe the bacterial diversity within samples for each sample's group, which show that genital samples (ulcerated and non-ulcerated) had signi cantly less bacterial diversity compared to the oral samples (ulcerated and nonulcerated). In addition, the Figure 1, demonstrates that the genital samples are mainly dominated by few ASVs while the oral samples have more distribution of more species present. Statistically, the Pairwise Wilcox test con rms that alpha diversity of genital and oral was signi cantly different; p value= 0.015. However, in case of considering the groups: (O_U vs O_nU) and (G_U vs G_nU) the pairwise Kruskal-Wallis H test based on Shannon was not signi cant ( Figure  1).
Ordinate analysis Non-Metric Multidimensional Scaling (NMDS) on Bray-Curtis distances among samples was used to visualise the microbial community structure. 'k=3' to state that the variation should be condensed into 3 axes with a stress value= 0.10. The beta diversity analysis showed that the microbial community's composition of oral and genital samples was signi cantly different, and they shared a few bacterial taxa (pseudo-F =12.6892, p value< 0.001).
The PERMANOVA test shows there is no clear separation between the microbial composition of the samples attributable to the "ulcer" and "no ulcer" status (ADONIS test p-value>0.05) (Figure 2).
The analysis of the microbial distribution of the 50 most abundant ASVs, showed that those associated with the oral mucosal areas belonged to different genera compared to the genital mucosa/skin areas. Focusing on speci c ASVs of V1/V2, we observed that the most abundant ASVs dominating genital area was assigned to Lactobacillus, Prevotella, Dialister, Atopobium, Escherichia-Shigella and Staphylococcus. While the oral samples were Streptococcus, Veillonella, Rothia, Lautropia, Haemophilus and Actinomyces (Figure 3).

Analysis of V1/V2 region compared to V3/V4 region based on patients' age
The result ndings of the Wilcoxon rank-sum test and Mann-Whitney test demonstrated that there was no clear effect of the age on the BD patients on microbial diversity by V1/V2 and V3/V4 analysis, except in the older patients (60 years old and above) genital no ulcer (G_nU) samples have high values but the number of cases were not su cient to have a signi cant effect (n = 4).
Analysis of V1/V2 region compared to V3/V4 region based on patients' gender Figure S1 highlights the distribution of the 50 most abundant ASVs, assigned to the taxonomic rank of genus. The gender data of (O_U vs O_nU) and (G_U vs G_nU) on V1/V2 and V3/V4 are summarised in (Table 3a, b, c, and d). Table 3a reveals that the relative abundant genera in males with genital ulcers were in ascending order were Corynebacterium, Dialister, Streptococcus, Prevotella, but Lactobacillus, Dialister, Prevotella, Atopobium, Escherichia-Shigella, and Veillonella were highly abundance in females with genital ulcers.
Differential expression analysis (DESeq2) was used to compare the bacterial abundance between males and females. V1/V2 analysis failed to show any signi cant impact of gender on the bacterial abundance of oral and genital area. V3/V4 analysis of genital area demonstrated that Lactobacillus and Gardnerella were signi cantly 20 times more in females than males in general; with p-adj <0.05. However, when the DESeq2 analysis used to compare male vs the female ulcer group with ulcer group, indicated that the Gardnerella, Lactobacillus, Atopobium were signi cantly increased in females than males, while Peptoniphilus and Corynebacterium were signi cantly decreased in females than males.
The effect of Triorasol mouthwash (TMW) on the oral and genital microbiome based on V1/V2 and V3/V4 region analysis A constrained ordination test was used to investigate the impact of Triorasol mouthwash (TMW) on the bacteria abundance. The analysis of V1/V2 showed that about 2% variation between different bacterial abundance when Triorasol mouthwash was used. There was a signi cant variation between patients who have ulcers using TMW compared to patients with no ulcers and no mouthwash (V1/V2; R 2 = 0.021; P value= 0.03). Also, TMW in uences the bacterial abundance of patients with no oral ulcers who use TMW compared to patients with no oral ulcers and no mouthwash (R 2 = 0.074; P value= 0.04) ( Figure 5A).
However, V3/V4 region analysis of the oral area shows no signi cant differences of bacterial abundance between samples with or without ulcers considering TMW.
The V1/V2 region analysis demonstrated that the oral samples in patients with no-ulcer have a reduction of Prevotella, Actinomyces, Haemophilus, Streptococcus and Porphyromonas, gure 5B shows that Streptococcus is signi cantly increased when the oral ulcers are present while Veillonella is signi cantly decreased in patients of oral ulcer.
Additionally, the V1/V2 and V3/V4 regions analysis demonstrate that TMW did not have a signi cant effect on bacteria abundance in genital area (ulcerated and non-ulcerated samples).

The effect of systemic medication and clinical outcomes on oral and genital microbiome abundance based on V1/V2 and V3/V4 region analysis
In the genital area, Figure 6 and Table S1 revealed that the use of Colchicine has a signi cant effect on the microbial abundance (V1/V2; P= 0.020, V3/V4; P= 0.003) but there was no relationship with Colchicine and the presence or absence of the ulcers in the genital area. It is also clear that there was a cluster of samples that have higher WSAS (V1/V2; P= 0.004, V3/V4; P= 0.012), WSAS was not linked to the presence or absence of ulcers. Only V3/V4 region analysis showed that gender has a signi cant impact on microbial abundance (P= 0.001). However, the V1/V2 analysis results suggested that there was an effect of Azathioprine on bacterial abundance, just failed to be signi cant (P= 0.058), in addition V3/V4 analysis results showed that MMF fails to be signi cant (P= 0.078).
Most patients treated with Colchicine and Azathioprine also have lower WSAS values. Additionally, performed analyses showed that most of the patients with a higher WSAS value are patients who have active disease, and that the bacterial presence differs signi cantly from the non-active patients (ADONIS, R2= 0.05, p value =0.029).
In the oral area, V1/V2 and V3/V4 demonstrated that was no effect of systemic medication or clinical outcomes on the oral bacterial abundance in BD patients

Discussion
The present study is the rst using V1/V2 and V3/V4 regions sequences to identify bacterial species that are associated with the presence and absence of orogenital ulcers and disease activity. Included is a well-de ned cohort of BD patients, clinical disease outcomes, and therapeutic regimens to investigate whether there is any effect on the bacterial diversity and abundance.
Genital samples had signi cantly less Alpha and beta diversity bacterial diversity compared to the oral samples. However, the alpha and beta bacterial diversity were not signi cantly impacted by the presence of oral or genital no ulcer samples. This nding corroborates some data associated with the salivary microbiome in a recent study of Lichen planus where no difference in bacterial diversity with or without erosive lesions. They were unable to demonstrate a signi cant difference in bacterial species between two oral Lichen planus groups (erosive and non-erosive) [33], while a previous BD study was found signi cant differences between oral ulcers and no ulcers samples [34] .This discrepancy could be due to the advanced sequencing techniques used to process and analysed our samples. However, in all 3 studies genetic factors, medication, and length of time of the disease process may in uence the microbial community. A state of chronic in ammation in BD may affect the composition of the organisms at the mucosal barrier allowing less perturbation and a more stable mucosal barrier microbial community.
The core microbiome analysis based on ASVs displayed that the patients' group samples shared a few bacterial taxa; V1/V2 result demonstrates that oral ulcer samples and genital ulcer samples are assigned to genus Escherichia-Shigella. In addition, all the study groups regardless of the presence of ulcers, they were sharing respectively the following genera: Fusobacterium, Cutibacterium, Lactobacillus, and Streptococcus. However, V3/V4 outcomes show that oral ulcer samples and genital ulcer samples are assigned to genus; Lachnospiraceae, Saccharimonadales, and Coriobacteriales. In addition, all the study groups regardless of the presence of ulcers, they are belong respectively to the genera: Escherichia-Shigella, Neisseria, Atopobium, Atopobium, Dialister, Veillonella, Gemella, Granulicatella, Lactobacillus, Rothia, Renibacterium, Bi dobacteriaceae, and Streptococcus. In summary, differential abundance analysis was identi ed that Streptococcus is related to the presence of oral ulcers while Veillonella is related to absence of oral ulcers, as a potential microbial marker of disease activity.
Diversity analysis failed to show that BD patients' age factor had a clear effect on microbial diversity by V1/V2 and V3/V4 regions analysis, and it could be as a result of our collection samples cohort, which it did not has same number of patients in each age groups. Although, Xu, C, et al., [35] found a steady increase of the alpha diversity as the age increase, and there was a steep drop in the extremely elderly age group; however, the beta diversity were notably larger in childhood age than the other age group [35,36].
BD patients' gender did not have any signi cant impact on the bacterial diversity or on the bacteria abundance in oral and genital areas based on the V1/V2 region analysis. However, V3/V4 region analysis of genital samples showed that Lactobacillus and Gardnerella were signi cantly 20 times more in females than males in all genital samples regardless the presence and absence of ulceration. In case of genital ulcer groups, the Gardnerella, Lactobacillus, Atopobium were signi cantly increased in females than males, while Peptoniphilus and Corynebacterium were signi cantly increased in males than females. Troccaz M, et al., [37] reported that Peptoniphilus and Corynebacterium were more represented in males than in females, and Corynebacterium was associated with body odour. In addition, during the analysis of our BD male samples, Rothia was found in both oral and genital samples which is usually present in oral samples commonly [37]. Therefore, hygiene habits such as shaving and hand washing are important, which it might has an in uence on their immune system.
Moreover, cofactors such as; balanced/unbalanced diet, alcohol consumption, and smoking did not have effect on the bacterial pro le in this study.
Triorasol mouthwash (TMW) has a signi cant impact on bacterial abundance in our patients' samples. V1/V2 region analysis showed reduction in the abundance of oral bacterial composition such as; Prevotella, Actinomyces, Haemophilus, Streptococcus and Porphyromonas in BD patients who used TMW. In addition, Streptococcus is signi cantly increased when the oral ulcers are present while Veillonella is signi cantly decreased in patients with oral ulcer, these ndings might help us to control oral ulcers and disease activity. Moreover, Staphylococcus was signi cantly less in non-active BD patients. However, V3/V4 region analysis in the oral samples shows no signi cant differences of bacterial relative abundance variations between samples with or without ulcers considering TMW. Also, it was reported that TMW did not have a direct effect on bacteria abundance in genital area.
Medication used for treating BD patients' have diverse effects on the oral and genital microbiome. Interestingly, all the prescribed disease modifying medications such as prednisolone or biologics did not show any impact on the microbiome. In addition, Azathioprine and MMF failed to show a signi cant effect on the microbiome. Colchicine which is used as a rst line management for BD patients alone or combined with other immunomodulatory medications and acts as an anti-in ammatory and decreases neutrophil nets, had no impact on the oral microbiome. However, Colchicine had a signi cant effect on the microbial abundance in genital area irrespective of the presence or absence of the ulcers. Shimizu J, et al., [38] reported that Colchicine (daily doses of 0.79 ± 0.26 mg) hardly affected gut microbes and host intestinal mucosa. They also suggested that Azathioprine (a daily dose of 75 mg) had marginal effects on the gut microbes [38].
Work and Social Adjustment Scale (WSAS) assesses the impact of a person's mental health di culties on their ability to function in terms of work, home management, social leisure, and personal or family relationships. Interestingly, most of our patients treated with a combination of Colchicine and Azathioprine, they have lower WSAS scores. Additionally, most of the patients with a higher WSAS value were patients who have active disease.
Limitations of this study are; the imbalance of sampling size between genital and oral may have affected the result. A larger BD and healthy population spread throughout the BD lifecycle are important to be included in future studies. Our future work will concentrate on a longitudinal study of BD patients.

Conclusion
Our results indicate a signi cant difference between the mucosal genital and oral microbial community composition of BD patients. There were no differences of bacterial diversity between ulcerated and non-ulcerated sites. Moreover, we veri ed that Streptococcus is related to the presence of oral ulcers while Veillonella is related to absence of oral ulcers (which may be a potential microbial marker of disease activity). In addition, the presence of Staphylococcus is associated with BD disease activity. In the genital area, Lactobacillus and Gardnerella were more abundant in females than males; the Gardnerella,Lactobacillus, Atopobium were higher in females than males with genital ulcers, while Peptoniphilus and Corynebacterium were signi cantly increased in males than females with genital ulcers. Clinically, Triorasol mouthwash (TMW) shows effective managements to mucosal diseases including BD which presents with mucosal ulceration [39].

Declarations
Ethics approval and consent to participate Obtained.

Availability of data and materials
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Figure 4
Page 15/16 Number of unique bacterial genera in Oral and Genital area with or without ulcers samples. Venn diagrams show the number of shared genera in the four different groups. The plots were created using an online tool (http://bioinformatics.psb.ugent.be/webtools/Venn/).

Figure 5
A: The effect of TMW combined with the presence of the ulcer in the oral area on the microbial community structure.
Constrained Analysis of Principal Coordinates (CAP) ordination method was used to analyse. Data considered are related to V1/V2 region. B: DESeq2 differential abundance analysis expressed as Log2FC comparison of the oral samples with or without.
Negative Log2FC represents Genus enhanced in the two groups. Each point represents an individual AVS assigned at the Genus level. To enhance clarity, only those AVSs with p-adj<0.0001 are shown. Streptococcus is signi cantly more present in samples with ulcers. However, Veillonella is more present in non-ulcer samples.

Figure 6
The effect of systemic medication and clinical outcomes on bacterial abundance in genital area.