Cervicovaginal Gardnerella sialidase-encoding gene in persistent human papillomavirus infection

Disturbed vaginal microbiota have a role in the persistence of high-oncogenic-risk human papillomavirus (hrHPV) and Gardnerella spp. is closely related with this condition. Such bacteria are the major source of cervicovaginal sialidases, important for microbiota alterations. The sialidase-encoding gene nanH3 is account for their sialidase activity. Thus, a subset of 212 women positive for hrHPV at the first visit were included in the analysis of the current study aiming to compare the loads of nanH3 in cervicovaginal fluid (CFV) of women with persistent hrHPV infection and with those cleared the infection after a year. Participants were assigned to two study groups named “persistence” (n = 124, 53.22%) or “clearance” (n = 88, 37.77%), according to the HPV status upon enrollment and follow-up. Absolute quantification of nanH3 gene was performed using quantitative real-time PCR (qPCR). Persistence and clearance group did not show statistical difference in the load of nanH3 gene (p = 0.19). When considering the subset of women with HPV16, differences in number of copies of nanh3 gene was observed between the persistent (7.39E+08 copies/μL) and clearance group (2.85E+07 copies/μL) (p = 0.007). Therefore, baseline loads of nanH3 gene is increased in women that persist with cervical HPV16 infection after 12 months.

www.nature.com/scientificreports/ the sialidase-encoding gene nanH3 of Gardnerella spp. in the cervicovaginal fluid of women between persistent hrHPV infection and those who cleared the infection after a 12-month period.

Results
Table 1 displays the sociodemographic, behavioral and clinical characteristics of the participants, according to the study groups.The groups showed differences in few variables such as age (p = 0.04), and the number of sex (p < 0.0001).When considering the number of hrHPV genotypes detected, persistent group had more mixed infection than clearance group (p = 0.002).In relation to the HPV genotypes, HPV52 differed between the groups when considering single or mixed infection (p = 0.01).
The most frequently genotype detected in the population (showed in Supplementary Table S1), was HPV16, followed by HPV31, HPV52, HPV51, HPV58 and HPV45, respectively.Of 233 women included in this study, 21 cleared their HPV genotype detected at baseline and a new genotype was detected at follow-up, therefore these women were not included in the analysis.Thus, clearance group included women without detection of HPV at follow-up (n = 88) and persistence group includes those women showing at least one genotype detected at baseline and follow-up (n = 124).Thus, clearance and persistence rates were 41.50% (n = 88) and 58.49% (n = 124), respectively.For HPV16, rate of clearance was 26.56% (n = 17) and persistence was 73.44% (n = 47).

Discussion
Recently our study group showed that women with persistent HPV16 and HPV18 infection have increased baseline loads of G. vaginalis 27 .Thus, it was hypothesized that gene encoding sialidase, frequently found in Gardnerella spp. 18could have a role in the outcome of HPV infection, being a useful marker for hrHPV persistence.Surprisingly, loads of nanH3 gene differed between clearance and persistence group in the presence of 5 most prevalent genotypes, however such difference remains only in the presence of HPV16.
This study showed high rates of HPV persistence, 58.49% for all hrHPV infection and 73.44% when considered only the HPV16.Such rates were similar to the rates showed in another study conducted in Brazil, with 61.8% of persistence rate 28 .Persistent rates in a large retrospective cohort study in China was 42.7% within 24 months 29 , while a multicentric study showed 54.1% of persistence rate over a year in American, Canadian and Brazilian women 30 .In relation to HPV16, the persistence rate was 52.1% after 12 months in Dutch cohort of young women 31 .
In relation to sociodemographic and behavioral characteristics of the study population, few variables differed between the groups.The median of age was higher in clearance group, agreeing with a study that showed that the highest HR-HPV persistence occurred in the 22-27 years old group, whereas clearance increased in women aged 28-33 years 32 .However, other Brazilian study did not show differences of age when evaluating factors associated with clearance and persistence after 24 months 28 .The number of sex partners and the number of detected genotypes differed between the clearance and persistence group.Furthermore, one of five most prevalent HPV genotypes differed among the groups when considering their detection in single or mixed infection.In fact, a previous study showed that HPV genotypes differ in terms of the period for clearance, as women with mixed infections had a longer time to clear their infections compared to those with a single infection 30 .The increasing in number of sex partners could contribute to persistence since it increases the risk of acquiring new HPV genotypes.
Studies on vaginal microbiota have demonstrated the association between Gardnerella spp.and persistent hrHPV infection, as well as progression of the cervical lesion 33,34 .Of the results of association between Gardnerella spp. with persistence and progression of hrHPV, Usik et al. proposed that in addition to role of Gardnerella spp. in the viral persistence, it also contributes for the disruption on vaginal microbiota (i.e. increased bacterial diversity), which also leads to the persistence and progression of HPV infection 34 .
As recently demonstrated, not all newly described species of Gardnerella produce sialidase 18 .Therefore, studying the presence of sialidase encoding gene of Gardnerella is important for paving the way for further studies focusing in the role of each individual Gardnerella sp. for HPV infection and persistence.Table 1.Sociodemographic, behavioral and clinical characteristics of the participants, according to the study groups.For all analysis P-value < 0.05 was considered as significant.*Chi-square test.# Mann-Whitney test, expressed by median (minimum-maximum).$ Fisher's exact test.Significant values are in bold.To the best of our knowledge, this is the first study to assess the relationship between hrHPV persistence infection and loads of nanH3 gene.It is known that sialidases hydrolyze the sialic acid of the host's epithelial cells, provides a carbon source to Gardnerella allowing their uptake and catabolism and expose binding sites, facilitating bacterial adhesion to the epithelium and further biofilms formation [20][21][22]35 . Siaidases can still cleave mucin oligosaccharides, that decreased the cervicovaginal mucus viscosity, compromising the physical and biochemical barrier against pathogens 36 .Besides that, sialidases compromise the immune barrier by hydrolyzing immunoglobulin A 35,37 .In the present study, results on the quantification of nanH3 did not differ when included   www.nature.com/scientificreports/all hrHPV infection grouped, however loads of nanH3 were higher in the cervicovaginal fluid of women with HPV16 persistent.
Studies have been reporting the association of altered vaginal microbiota with HPV infection [12][13][14]33,34,38 , despite the mechanisms by it can influence in HPV infection remain unknow. It is knwn that Lactobacillus spp.produce several microbicidal factors, including lactic acid 39 , an important agent to acidify the vaginal environment and to control overgrowth of bacteria and preventing STIs 40,41 .The difference of nanh3 loads found in HPV16 infection is intriguing, are needed new approaches involving vaginal microbiota and HPV16, since this genotype is responsible for the most cases of high-grade intraepithelial lesions and invasive cervical carcinomas 42 .
Human papillomaviruses modulate the host immune response by blocking immune-related gene expression and immune signaling pathways and different genotypes may act by different pathways 43 .In addition, cervical microbiome and cytokine profile showed notably different in all stages of the natural history of cervical cancer 44 .Thus, bacterial species may interact with immune signaling during HPV infection contributing to persistence and progression 44 .On the other hand, a consequence of HPV immune evasion is the imbalance in the vaginal microbiota due the reduced amino acid source sustaining the survival of Lactobacillus species 45 .Therefore, the association between BV and HPV seems to be bi-directional.In fact, a study showed that the stability of vaginal microbiome is necessary to maintenance of immune surveillance for HPV16 clearance 46 .In this sense, the differences in the load of nanH3 gene in the presence of HPV16 lead to hypothesize that the vaginal microenvironment would have a different modulation in the presence of different genotypes and the changes in the immune response might create a more permissive microenvironment for bacterial overgrowth.
Gardnerella spp. is present in virtually all cases of BV, the most frequent dysbiosis of the vaginal microbiota 47 .In this study, an association between the nanH3 gene and BV was observed.Robinson et al. suggested the potential use of nanH3 as a molecular diagnostic marker of BV, such PCR test showing 80.95% sensitivity and 78.26% specificity compared with the Nugent score for BV diagnosis 25 .Although the association, in this study values of sensitivity and specificity were lower than the mentioned study.It is worth mentioning that studies differ in terms of prevalence of BV and number of study participants.
Besides the association between nanH3 gene and BV, the load of the gene was significantly increased in BV compared to normal microbiota.As previously discussed, sialidases have deleterious effects in the vaginal microenvironment [20][21][22][35][36][37] . Recenty, our study group showed that sialidase activity in molecular-BV, assessed by V3-V4 16S rRNA sequencing, is associated with changes in bacterial components of the local microbiome as well as increased sialidase levels 48 .Indeed, sialidase activity in cervicovaginal fluid already was demonstrated in the presence of microscopy-detected BV 49,50 .In fact, sialidase activity could promote the growth or colonization of Prevotella, Bacteroides as well as G. vaginalis, vaginal bacterial sialidase producers 51,52 , being the production of sialidase an important step in the biofilm formation 24 .Such mechanism reinforces the role of Gardnerella spp. as the scaffold on vaginal mucosa for attachment of other bacterial species, which can start to form biofilm 53 , a condition already associated with refractory treatment of BV and persistence of BV-associated bacteria after treatment 23,54 .
Although the amount of sialidase-encoding nanH3 gene did not differ between clearance and persistence in the presence of all hrHPV infection, the loads are high in the persistent group and there is a difference among the groups when considering only HPV16 infection.A limitation of this analysis is due the low sample size when stratifying each genotype.Furthermore, such results contribute to better understanding the role of G. vaginalis sialidase-encoding gene in BV, since this condition seems have an important role in the persistence and progression of HPV infection.Thus, microbial products are promising tools for identifying women with increased risk for HPV persistence.

Methods
Data collection, sampling and laboratorial procedures.Between 2012 to 2014, 1635 reproductiveaged HIV-negative women screened for HPV infection in Botucatu, SP, Brazil.A sociodemographic, behavioral, gynecological history and clinical data were obtained during a face-to-face interview using a structured questionnaire and were entered into a Microsoft Excel spreadsheet (Microsoft Corporation, Redmond, WA, USA).HPV positive women were then invited to a follow-up within 12 months.All participants were informed about the procedures of study and informed consent was obtained from all subjects as all reached the age of majority (18 years-old).All methods were carried out in accordance with relevant guidelines and regulations and this current study was reviewed and approved by the Ethics Committee of Botucatu Medical School (Approval number: 5.660.472) and the samples included in the present study belongs to the biorepository of principal study (Biorepository approved number: 3.140.843).
Same sampling procedures during physical examination were conducted by nurses/physicians at enrolment and at follow-up visit.After speculum insertion, vaginal content was obtained with swabs for smearing into microscopic glass slides for Nugent scoring, after gram staining 15 .Mid vaginal wall was utilized for Trichomonas vaginalis culture in Diamond's medium 55 .Samples obtained with endocervical brushes were stored at −20 ºC until Chlamydia trachomatis and Neisseria gonorrhoeae detection, as described previously 55,56 and for HPV detection and genotyping using the Linear Array HPV Genotyping kit (Roche Molecular Systems, Pleasanton, CA, USA).For cervicovaginal fluid sampling, 3 ml of sterile 0.9% NaCl was inserted into vaginal wall, homogenized with cervicovaginal secretion and recovered using a plastic sterile pipette.The cervicovaginal samples were stored at −20 ºC until genomic DNA extraction and G. vaginalis nanH3 gene absolute quantification.
Genomic DNA extraction of cervicovaginal fluid was performed using the UltraClean Soil DNA Isolation kit (MoBio Laboratories, Carlsbad, CA, USA) as recommended by the manufacturer.Extracted DNA was quantified using an Epoch spectrophotometer (Biotek, Winooski, VT, USA) and the quality of the extraction was confirmed by the 260/280 nm absorbance ratio.www.nature.com/scientificreports/A cloning step was performed to obtain the plasmid DNA with Gardnerella nanH3 gene, through polymerase chain reaction (PCR) products from clinical samples using the primers Gardnerella nanH3 F (5′-CAG TTC CAA TGG AAG TGT GC-3′) and Gardnerella nanH3 R (5′-AGC ATC TGG GAA TGC TCT TG-3′) with 51 ºC of annealing temperature.Expected amplicon size was 322 bp for nanH3 gene, confirmed in 1.5% agaroses gel band 25 .Positive samples had the amplicon purified using Illustra GFX PCR DNA and Gel Band Purification Kit (GE Healthcare UK Buckinghamshire, UK) and sequenced by Sanger method in ABI 3500 (Applied Biosystems, Foster City, CA, USA) and confirmed by blast (https:// blast.ncbi.nlm.nih.gov/ Blast.cgi# alnHdr_ 22355 00528).The amplified products were used for the ligation step using the CloneJET PCR Cloning Kit (Thermo Fisher Scientific, Carlsbad, CA, USA), cloned into Escherichia coli DH5-α (Invitrogen, Carlsbad, CA, USA).Plasmid DNA was extracted using GeneJET Plasmid Miniprep Kit (Thermo Fisher Scientific, Carlsbad, CA, USA).
The absolute quantification of nanH3 gene was performed by quantitative real-time PCR (qPCR) using 2× qPCRBIO SyGreen Blue Mix Hi-ROX (PCR Biosystems, London, UK), primers Gardnerella nanH3 F and Gardnerella nanH3 R under the following cycling conditions: 95 °C for 2 min, 25 cycles of denaturation at 95 °C for 5 s, annealing at 60 °C for 25 s, and an extension at 95 °C for 15 s followed by a dissociation stage: 95 °C for 15 s; 60 °C for 1 min and 95 °C for 15 min, according to the manufacturer's recommendations, performed in a StepOnePlus Real-Time System (Thermo Fisher Scientific, Waltham, MA, USA).Samples with a melting temperature value of 81 °C + / − 0.5 °C were considered positive for nanH3 gene.For constructing the standard curve, three plasmid dilutions (3.05E+05 copies/µL, 3.05E+07 copies/µL and 3.05E+09 copies/µL) were utilized for sample cycle threshold interpolation.Loads of nanH3 gene were expressed as the number of copies per volume (µL) of cervicovaginal fluid.

Selection of participants and group assignments.
In the first visit, 544 (33.27%) women were positive for any HPV infection, of which 413 (25.26%) tested positive for hrHPV genotypes and were recruited for the follow-up visit after 12 months.Four hundred sixty-two (27.65%) women returned to the follow-up.For the present study, women positive for Chlamydia trachomatis (138, 8.44%), Trichomonas vaginalis (23, 1.41%), Neisseria gonorrhoeae (2, 0.12%) and only low-risk HPV (131, 8.01%) were not included in the analysis.Five women were excluded of this study due conflicting data in the spreadsheet database.From that, 245 (14.96%) women were hrHPV positive (HPV16, HPV18, HPV31, HPV33, HPV35, HPV39, HPV45, HPV51, HPV52, HPV56, HPV58, HPV59, HPV68, HPV73, HPV82), however, 12 samples were excluded due insufficient volume of cervicovaginal fluid for laboratorial analysis of the current study.Therefore, 233 cervicovaginal fluid samples were included for this study analysis.Participants were assigned at two groups according to the status of hrHPV infection at enrolment and at follow-up: 'clearance group' comprised participants who tested negative for any hrHPV at follow-up and in 'persistence group' were assigned participants positive for hrHPVs genotype at first visit and follow-up, matched with those detected at follow-up.For clearance group, 21 women who showed clearance of baseline genotype and had a new genotype detected at follow-up were not included.

Statistical analysis.
For sociodemographic, behavioral and clinical variables comparisons between the clearance and persistence groups, chi-squared or Fisher's exact were utilized for categorical variables and Mann-Whitney tests was utilized for continuous variables.The loads and presence of nanH3 gene were compared between the study groups using, respectively, Mann-Whitney and chi-squared tests or Fisher's exact test.All analyses considered a P-value < 0.05 to be statistically significant and were performed using GraphPad Prism software (version 6.0, GraphPad, CA, USA).