The cervicovaginal microbiome (CVM) plays a crucial role in women's health and is typically structured in microbial community state types (CSTs), which are characterized by the dominance of particular bacterial species (I, II, III, and V) or the assemble of a diverse bacterial population (IV). CSTs I, II, III, and V are dominated by the species Lactobacillus crispatus, Lactobacillus gasseri, Lactobacillus iners, and Lactobacillus jensenii, respectively, while CST IV is described by the co-occurrence of pathogenic anaerobic bacteria such as Gardnerella vaginalis, Atopobium vaginae, and Dialister micraerophilus1. In healthy cervical conditions, the CVM is associated with an increased abundance of Lactobacillus species and low microbial diversity2. Changes in this composition results in bacterial vaginosis (BV, CST IV), which associates with adverse outcomes such as pre-term birth and a higher susceptibility to HIV and high-risk human papillomavirus (hrHPV) infections3.
Evaluating CSTs during health and disease is currently challenging because they have been mostly classified with amplicon-based sequencing technologies, which generally do not achieve species-level microbiome profiling4. Likewise, several studies have proposed the existence of up to nine CST groups5, and such non-uniformity in CSTs classification makes it difficult to examine their clinical associations in cross-sectional and longitudinal studies. Recent reports suggest a reclassification of CSTs into the classical five groups with novel communities assigned as subgroups6. Nevertheless, these observations have not been confirmed in follow-up studies, and there is still the need to assess microbial communities by applying high-resolution sequencing.
Our group recently developed the sequencing technology circular probes-based RNA sequencing (ciRNAseq) for high-resolution profiling of the cervicovaginal microbiome at DNA and RNA levels7. Through ciRNAseq profiling of cervical smears, we observed the well-known associations of CST I with hrHPV-negative conditions and CST IV with hrHPV-positive women and high-grade cervical lesions, which corroborated its potential for microbiome sequencing and particularly for CSTs classification7. Additionally, we identified the species Lactobacillus acidophilus and Megasphaera genomosp type 1 in the CVM and determined their association with health and disease conditions, respectively7. In general, CVM studies do not annotate these species, likely due to their high level of sequence identity with other species in the microbiome. Thus additional studies applying high-resolution microbiome profiling are needed to fully evaluate the CVM composition in women3.
We therefore aimed to investigate CSTs in a large cohort of cervical smears from 541 women participating in the Dutch population-based cervical cancer screening program and applied ciRNAseq for targeted sequencing of their CVMs.
Through unsupervised clustering analyses on 341 microbiomes, we observed novel subgroups of CSTs I, III, and IV that correlated with the abundance of specific bacterial species. CST I-A and CST I-B were both characterized by L. crispatus dominance, but I-B exhibited a higher abundance for L. acidophilus and L. iners. In addition, CST III-A and III-B were both characterized by L. iners dominance, and III-B showed an exclusive occurrence of L. acidophilus and increased abundance for L. jensenii. Alternatively, CSTs IV was classified into subgroups A, B, and C, with IV-A and IV-B showing similar microbial composition but the latter characterized by dominance for M. genomosp type 1. CSTs IV-A exhibited dominance for several microbial species, including Lactobacillus. CST IV-C is characterized by the dominance of non-Lactobacillus species. We also analyzed Species richness and diversity indices for all CSTs and observed that the B subgroups showed a significantly higher number of species when compared to the A subgroups, which further supported their classification.
To validate our findings, we compared the prevalence of CSTs in a second cohort that consisted of an additional 200 hrHPV-positive women with either low- or high-grade cervical lesions and previously analyzed hrHPV-negative and positive women without cervical abnormalities. We observed the subdivisions of CSTs in all study groups and found that CST V was associated with hrHPV-negative conditions and CST IV-A with hrHPV infection and cervical disease.
Altogether, our results demonstrate the existence of novel microbial communities in the cervicovaginal microbiome that associate with cervical health and disease in the context of hrHPV infections and which require high-resolution microbiome profiling for adequate classification8.
- Ravel J, et al. Vaginal microbiome of reproductive-age women. Proceedings of the National Academy of Sciences. 2011;108(Supplement 1):4680.
- Godha K, Tucker KM, Biehl C, Archer DF, Mirkin S. Human vaginal pH and microbiota: an update. Gynecological Endocrinology. 2018;34(6):451-5.
- Molina Mariano A, Coenen Britt A, Leenders William PJ, Andralojc Karolina M, Huynen Martijn A, Melchers Willem JG. Assessing the Cervicovaginal Microbiota in the Context of hrHPV Infections: Temporal Dynamics and Therapeutic Strategies. mBio. 2022;0(0):e01619-22.
- Graspeuntner S, Loeper N, Künzel S, Baines JF, Rupp J. Selection of validated hypervariable regions is crucial in 16S-based microbiota studies of the female genital tract. Sci Rep. 2018;8(1):9678.
- Brooks JP, et al. Changes in vaginal community state types reflect major shifts in the microbiome. Microb Ecol Health Dis. 2017;28(1):1303265.
- France MT, et al. VALENCIA: a nearest centroid classification method for vaginal microbial communities based on composition. Microbiome. 2020;8(1):166.
- Andralojc KM, et al. Novel high-resolution targeted sequencing of the cervicovaginal microbiome. BMC Biology. 2021;19(1):267.
- Molina, M.A., Andralojc, K.M., Huynen, M.A. et al. In-depth insights into cervicovaginal microbial communities and hrHPV infections using high-resolution microbiome profiling. npj Biofilms Microbiomes 8, 75 (2022). https://doi.org/10.1038/s41522-022-00336-6