Human papillomavirus infection and cervical intraepithelial neoplasia progression are associated with increased vaginal microbiome diversity in a Chinese cohort

Human papillomavirus (HPV) is regarded as one of the most common sexually transmitted agents in cervical intraepithelial neoplasia (CIN) and cervical cancer [1]. High-risk subtypes of HPV contribute to 99% of cervical neoplasia [1]. However, it is known that high-risk HPV infection is necessary but not sufficient for the development of CINs or cervical cancer [2‐4]. Many other events, such as multiple sexual partners, early initiation of sexual activity and co-infection with other sexually transmitted infections, have been associated with higher risk of HPV infection in the genital tract [5‐7].

In the female genital tract, a healthy vaginal status is commonly associated with low microbial diversity and prevalence of Lactobacillus. Lactobacillus prevents colonization of other bacterial pathogens through production of lactic acid, hydrogen peroxide (H₂O₂) and bacteriocin in the vagina, and therefore keeps integrity of mucosal barriers against virus and opportunistic bacteria [4, 8, 9]. The vaginal microbial profile of women could by classified into five community state types (CSTs) by hierarchical taxonomic clustering, in which CSTs I, II III and V are predominated by L. crispatus, L. iners, L. jensenii and L. gasseri respectively, while CST IV is depleted of Lactobacillus and enriched with anaerobic bacteria like Gardnerella, Megasphera, Sneathia, Prevotella, etc. [10]. Commensal vaginal Lactobacillus species are thought to defend against many pathogens, such as Candida infection [11, 12], sexually transmitted diseases [13], urinary system infections [14, 15] and human immunodeficiency virus (HIV) infection [16]. However, L. iners has many properties different from other Lactobacillus spp., for example unable to produce H₂O_2, and it often predominates in the presence of HPV infection [4, 17, 18] and CIN [18, 19].

Bacterial vaginosis (BV) is a cluster of microbial disorders characterized by a decrease in Lactobacillus and their replacement by high concentrations of other anaerobic bacteria, with a microbial community structure in accordance with CST IV [9]. BV is associated with a higher risk of miscarriage, preterm premature rupture of membranes and a higher susceptibility to sexually transmitted infections, as HPV infection [20‐22]. Some studies to date have reported that vaginal microbiome (VM) plays an important role in the persistence of the HPV infection and the subsequent development of cervical precancerous or cancerous lesions [2, 4, 8, 17‐19, 23‐28]. Increasing VM diversity is associated with advancing CINs severity and viral persistence [26]. The potential mechanisms could be linked to less production of protective lactic acid, H₂O₂ and bactriocin by Lactobacillus, disruption of mucosal integrity which may aid viral entry, higher levels of oxidative stress induced by dybiosis [8]. Particular species like Sneathia spp. have a probable pathological role in HPV acquisition and persistence through cellular targets such as expression of immunosuppressive cytokines [29]. Therefore, it is considerable to take vaginal microbiome as a promising marker not only for HPV infection but also for cervical precancerous lesions.

Nevertheless, the vaginal communities could be influenced by many other factors, including ethnicity, personal hygiene, sexual behaviors and hormonal levels [10, 30]. Ethnicity is key to shape vaginal bacterial communities [4, 10, 31]. Caucasian and Asian women display a significantly greater prevalence of Lactobacillus in the vagina compared to Hispanic and Black women [8, 10, 31]. To our knowledge, data with regard to the vaginal bacterial composition of Chinese populations are inadequate [8, 18, 32]. The analysis, which is performed in a large cohort of women living in a different country and with supposed different hygiene habits [10], is helpful to reinforce the underlying associations. Furthermore, there are few studies about the association between VM and HPV infection and related CIN diseases in Chinese cohorts using high throughput sequencing method [8, 18, 32].

Hence, the objective of this research is to study the role of VM on the HPV infection and the progression of CIN diseases in Chinese populations. We try to identify the microbiological markers related with HPV infection and CINs severity in these cohorts.

Our study addressed a not well-elucidated topic about the association between HPV infection and related CINs or cervical cancer and vaginal microbiome in Chinese cohorts. We observed that HPV infection increased vaginal bacterial richness and diversity regardless of the status of CINs. The vaginal bacterial richness and diversity were further augmented in the women with cervical cancer. Lactobacillus was the most abundant genus in all groups. HPV infection had a negative influence on the abundances of Lactobacillus, Gardnerella and Atopobium. Accordingly, HPV infection increased the relative abundance of Prevotella, Bacillus, Anaerococcus, Sneathia, Megasphaera, Streptococcus and Anaerococcus. The increased proportions of Bacillus, Anaerococcus and the reduced abundance of Gradnerella vaginalis were probably related with the progression of CINs severity. HPV infection without CINs or cancerous lesions was strongly associated with Megasphaera. The most abundant bacterium in the LSIL group was Prevotella amnii. However, Prevotella timonensis, Shuttleworthia and Streptococcaceae at the family level were three taxa related to HSIL. Furthermore, more taxa were associated with the Cancer group including Bacillus, Sneathia, Acidovorax, Oceanobacillus profundus, Fusobacterium, Veillonellaceae at the family level, Anaerococcus and Porphyromonas uenonis. Samples in the Normal group were mostly assigned to CST III. HPV infection converted the vaginal bacterial community structure from CST III to CST IV. Furthermore, the proportions of CST IV were gradually augmented with the progression of the severity of CINs.

Most of the studies have proven that HPV infection can increase vaginal bacterial richness and diversity and lower the percentage of Lactobacillus [4, 8, 17, 18, 23, 24, 40, 41], and our results are in agreement with these previous studies. However, a few studies found no difference between HPV positive and negative groups [25, 42]. HPV infection is thought to alter the acidic environment of the vagina, which might promote outbreaks of bacteria [24]. In addition, HPV infection might lead to changes in the vaginal microbiota by inducing host mucosal immune response and genital infalmmation [41, 43]. High genital inflammation with elevated vaginal PH and non-Lactobacillus-dominant VM have been associated with HPV persistence and progression to cervical cancer [44]. But the underlying biological mechanisms are still unclear. On the other hand, Mitra et al. reported that increasing CINs severity was associated with decreasing relative abundance of Lactobacillus and increasing bacterial diversity [26]. Differently, we only observed a higher bacterial richness and diversity in group Cancer than in groups HSIL, LSIL or HPV, but no differences were detected when compared between each two of groups HPV, LSIL and HSIL. Some other studies also found no connection between the diversity of VM and the CINs progression [45, 46]. It is notable that the study of Mitra et al. [26] did not distinguish the influence of HPV infection from precancerous or cancerous lesions on the diversity of VM.

Similarly to the previous studies [2, 4, 17, 18, 23, 24, 40, 42, 47], we found increased abundances of several anaerobic bacteria such as Prevotella, Bacillus, Anaerococcus, Sneathia, Megasphaera, Streptococcus and Anaerococcus in HPV-infected women. We recognized Megasphaera of Firmicutes phylum as the most significant genus related with HPV infection, while Lee et al. identified Sneathia spp. of Fusobacteria phylum as the microbiological marker of HPV infection [24]. Over all, a microbial environment with a higher proportion of anaerobic bacteria and a lower proportion of Lactobaillus spp. is more likely to HPV infection. A surprising finding in this study was that the proportions of Gardnerella and Atopobium, were reduced in HPV-positive women. Gardnerella vaginalis and Atopobium vaginae were thought to be associated with BV [48]. Gao et al. reported that these two taxa were more frequently detected in HPV-infected women [23]. However, the method used in their study was totally different from this research. Another study in a Caucasian cohort proposed Atopobium spp. and sialidase-encoding gene from Gardnerella vaginalis as microbial markers of HPV persistence [40]. It is hypothesized that the enzyme (sialidase) facilitates the destruction of the mucus layer on the vaginal epithelium and entraps anaerobic bacteria such as Prevotella and Atopobium. The potential reason for this different result remains to explore. To distinguish the influence of HPV infection from precancerous or cancerous lesions on vaginal microbiota, we compared the vaginal bacterial composition between group HPV and groups LSIL, HSIL or Cancer. We found that the proportions of two specific taxa, Bacillus and Anaerococcus, were positively related with the progression of CINs severity. Furthermore, we identified respective taxa for different stages of CIN lesions. Mitra et al. also reported that higher levels of Sneathia sanguinegens, Anaerococcus tetradius and Peptostreptococcus anaerobius were characterized in HSIL compared to LSIL [26]. Corresponding to the impact of HPV infection on the vaginal microbes, we also found that the abundance of Gradnerella vaginalis was gradually reduced with the progression of CINs severity. However, some other studies [19, 27] thought that an enrichment of Gradnerella vaginalis and Atopobium vaginae had a higher CIN risk.

In accordance with the results from an Asian population in the study by Ravel et al. [10], the most abundant CST in Normal group was CST III. The most dominant CST in the HPV positive groups (HPV, LSIL, HSIL and Cancer) was CST IV. We observed HPV infection to be associated with an increased proportion of CST IV, and furthermore its proportion was gradually augmented with the progression of the severity of CINs. It has also been reported by two longitudinal studies that the majority of HPV-positive samples were composed of CST IV (dominated by anaerobic bacteria), and CST IV was related with an increased risk of transitioning to an HPV-positive state [24, 49]. Mitra et al. also found that the rate of CST IV was increased 2 fold in women with LSIL, 3 fold in women with HSIL and 4 fold in women with invasive cancer [26]. CST IV is associated with higher levels of amine production, resulting in carcinogens nitrosamine production [50].

The strength of this study is that it interpreted the vaginal microbial compositions of a large cohort of Chinese women with different stages of HPV-related diseases using high throughput sequencing method, which has not yet been well elucidated. We found that HPV infection increased vaginal bacterial richness and diversity regardless of the status of CINs. The specific microbes and the vaginal bacterial structure were related with the progression of CINs severity in Chinese women. The limitations of this study were that it was a cross-sectional study. Hence, we could not conclude any causal relationship between the VM and HPV infection or CIN diseases. We have to conduct longitudinal studies to study relationships between the dynamics of the VM and the persistence or clearance of HPV infection, and the progression or remission of CIN diseases. In addition, the underlying biological mechanisms also need to be detailed.

This work interpreted the differential vaginal bacteria under HPV infection and various precancerous or cancerous lesions in a Chinese cohort. We distinguished the specific microbes and the vaginal bacterial structure that were related with the progression of CINs severity in Chinese women.