Loss of Protective Bacteria Marks the Microbial Landscape of Cervical Cancer

In the female reproductive tract, a delicate bacterial balance appears to play a far greater role in cervical cancer than previously appreciated. A new study published in MicrobiologyOpen has used high-throughput 16S rRNA gene sequencing to map the vaginal and cervical microbiota of women with cervical cancer and healthy controls in southern China, revealing a dramatic microbial reconfiguration that accompanies malignancy. The findings add weight to a growing body of evidence that the trillions of microbes inhabiting the reproductive tract are not passive bystanders but active participants in the journey from human papillomavirus (HPV) infection to invasive carcinoma.

Cervical cancer remains one of the most significant threats to women’s health worldwide, ranking as the fourth most common malignancy among women globally. More than 600,000 new cases are diagnosed each year, and nearly 90 percent of them occur in resource-limited countries. In China, the disease is increasingly affecting younger women, intensifying the urgency of understanding why some HPV infections resolve quietly while others progress to precancerous lesions and cancer. Persistent infection with high-risk HPV is firmly established as the primary cause, yet only a small fraction of infected women ever develop invasive disease, a discrepancy that has long pointed to additional cofactors.

Among those cofactors, the microbial communities of the vagina and cervix have emerged as compelling suspects. In healthy women, the genus Lactobacillus typically dominates the reproductive tract, maintaining an acidic environment, competitively excluding pathogens, and modulating local immune responses. Epidemiological studies have consistently identified bacterial vaginosis, a condition marked by depletion of Lactobacillus and overgrowth of anaerobic bacteria, as a risk factor for cervical neoplasia. However, the molecular mechanisms linking microbial dysbiosis to carcinogenesis, and the specific functional gene programs expressed by cancer-associated bacteria, have remained poorly defined. Conventional culture-based methods, limited in sensitivity and coverage, have also hindered comprehensive characterization of these complex ecosystems.

To address these gaps, researchers recruited women visiting gynecology departments at Nanchong Hospital, Beijing Anzhen Hospital, and Capital Medical University between January and June 2024. After applying strict inclusion and exclusion criteria, 32 women aged 45 to 65 were enrolled and stratified by vaginal discharge microscopy and cervical biopsy histopathology into 16 cervical cancer patients and 16 healthy controls. Crucially, all cancer patients tested positive for high-risk HPV while all controls were HPV-negative, providing a clean contrast that allowed the team to attribute microbial shifts specifically to HPV-driven disease. Women with recent antibiotic use, vaginal douching, bacterial vaginosis, sexually transmitted infections, prior cervical lesions, or systemic diseases were excluded, minimizing confounding.

The team collected sterile swabs from both the vaginal wall and the cervical os during gynecological examinations, extracted genomic DNA, and amplified the hypervariable V3-V4 region of the bacterial 16S rRNA gene. Sequencing was performed on an Illumina MiSeq platform with paired-end 300-base reads, and the data were processed using the DADA2 algorithm within the QIIME 2 framework to denoise reads, merge pairs, remove chimeras, and construct amplicon sequence variants. Taxonomic annotation relied on a pre-trained Naive Bayes classifier against the Greengenes reference database. After quality filtering excluded three samples from each group, 13 participants per group formed the final analytical cohort, which was well matched for body mass index, pregnancies, menstrual characteristics, and other clinical variables.

The compositional results were striking. Healthy controls were uniformly dominated by Lactobacillus, with a median relative abundance of 72.5 percent and minimal variation between individuals. In cervical cancer patients, Lactobacillus abundance collapsed to a median of just 12.5 percent, a highly significant difference. In its place, anaerobic genera flourished: Prevotella rose from a median of 6.8 percent in controls to 28.3 percent in cancer patients, while Peptostreptococcus, Fusobacterium, Enterococcus, and Corynebacterium were also enriched. The cancer group showed higher overall microbial diversity, greater inter-individual variability, and, notably, site-specific heterogeneity between the vaginal wall and cervical os, whereas healthy women maintained stable communities across both anatomical sites.

Diversity metrics reinforced this picture of ecological disruption. Species richness indices such as Chao1 and observed species did not differ between groups, indicating that cancer samples did not simply gain or lose bacterial types but instead underwent a wholesale shift in community composition. Pielou’s evenness was significantly lower in cancer patients, while Shannon and Simpson indices and Faith’s phylogenetic diversity were all elevated, reflecting colonization by multiple opportunistic pathogens and a fragmented community structure following the collapse of the Lactobacillus-dominated stable state. Analysis of similarities based on Bray-Curtis distances confirmed a statistically significant, though moderate, separation between the groups, with an R-value of 0.218 and a p-value of 0.001. When the cancer patients were subdivided by HPV genotype into HPV16-positive and HPV-positive/non-16 subgroups, community structure remained comparable between the two, suggesting that high-risk HPV infection as a whole, rather than any particular subtype, exerts the dominant selective pressure on the cervical microenvironment.

Network analysis and functional prediction pushed the findings beyond taxonomy into mechanism. Co-occurrence networks built with the SparCC algorithm revealed that healthy samples centered on a stable Lactobacillus-driven network, whereas cancer samples displayed a more complex, Prevotella-centered pathogenic network, with Lactobacillus showing widespread negative associations with Prevotella that suggest competitive exclusion between a health-associated commensal and potential pathogens. Functional prediction using PICRUSt2 mapped gene families onto Kyoto Encyclopedia of Genes and Genomes pathways and revealed that glycolytic flux was significantly elevated in cancer samples, consistent with lactate accumulation and acidification of the tumor microenvironment, while oxidative phosphorylation was diminished, indicating a metabolic shift toward anaerobic fermentation. The aerobactin biosynthesis pathway, which enables bacteria to scavenge iron from the host, was markedly upregulated in the cancer cohort, hinting at a novel link between dysbiosis, iron acquisition, and tumor progression. Meanwhile, reduced representation of antigen-presentation and complement pathways suggested microbial strategies of immune evasion.

The study’s authors caution that several limitations temper these conclusions. The cross-sectional design precludes causal inference, precancerous cases were excluded so microbiota dynamics during lesion progression could not be traced, and 16S rRNA sequencing lacks the species-level resolution needed to fully resolve functional heterogeneity, making metagenomic validation essential. Nevertheless, the results align closely with international findings, including a 2025 meta-analysis confirming elevated Prevotella in cervical cancer patients and cohort studies linking non-Lactobacillus-dominated community state types to persistent high-risk HPV infection and lesion progression. The researchers propose that future work should employ metagenomics to confirm aerobactin genes in enriched taxa, establish HPV-transgenic mouse models colonized with patient-derived Prevotella strains, and test probiotic interventions with Lactobacillus crispatus to restore acidic pH and inhibit pathobionts. If validated in longitudinal cohorts, the microbial signatures identified here, particularly the depletion of Lactobacillus, the enrichment of Prevotella, and the activation of iron-acquisition and inflammatory pathways, could serve as auxiliary biomarkers for early risk prediction and open the door to precision interventions based on microecological remodeling.

Subject of Research: Vaginal and cervical microbiota dysbiosis associated with cervical cancer, characterized by 16S rRNA gene sequencing

Article Title: Exploring Cervical Cancer‐Associated Vaginal and Cervical Microbiota via 16S rRNA Sequencing

Article References: Wang, G., Liu, X., Wang, Z., Tian, H., Li, Y., & Luo, Y. (2026). Exploring Cervical Cancer‐Associated Vaginal and Cervical Microbiota via 16S rRNA Sequencing. MicrobiologyOpen, 15(5), Article e70412. https://doi.org/10.1002/mbo3.70412

Image Credits: AI Generated

DOI: 10.1002/mbo3.70412

Keywords: cervical cancer, microbiome, 16S rRNA sequencing, Lactobacillus, Prevotella, HPV, dysbiosis, PICRUSt2, bacterial vaginosis, aerobactin, QIIME 2, women's health