Dr. Ourlad Alzeus G. Tantengco
The University of Texas Medical Branch

Ascending genital infections are a key contributor to preterm birth (PTB), a leading cause of neonatal morbidity and mortality. Ureaplasma parvum, a common genital mycoplasma, is frequently isolated from the reproductive tract of pregnant women with adverse outcomes, yet the mechanisms of its ascent and interaction with maternal-fetal tissues remain poorly understood. Traditional animal and in vitro models have limitations in replicating the dynamic and multicellular architecture of the cervix and fetal membranes. Thus, there is a need for physiologically relevant platforms to study host-pathogen interactions across reproductive barriers. 

This study aims to (1) develop a physiologically relevant microfluidic organ-on-a-chip (OOC) model that recapitulates the human vaginal-cervical-decidual interface, (2) investigate the dynamics of U. parvum ascent and inflammatory response in maternal and fetal cells, and (3) validate in vitro findings using a CD-1 pregnant mouse model of ascending infection. 

A six-chamber vagina-cervix-decidua organ-on-a-chip (VCD-OOC) was engineered to mimic the human lower genital tract and decidua using immortalized human epithelial and stromal cells. U. parvum was inoculated into the vaginal chamber and its ascent monitored via immunofluorescence, cytotoxicity (LDH assay), and cytokine profiling. Media from the decidua chamber was transferred to a feto-maternal interface OOC (FMi-OOC) to model fetal exposure. To validate these findings, CD-1 pregnant mice were intravaginally or intraamniotically inoculated with U. parvum and monitored for PTB and pup mortality. 

U. parvum ascended through the cervix to the decidua within 48 hours without causing significant cytotoxicity or inflammation alone. However, when co-treated with lipopolysaccharide (LPS) to mimic polymicrobial infection, massive pro-inflammatory cytokine surges (e.g., IL-8, GM-CSF) were observed in both OOC models. In vivo, vaginal inoculation alone rarely induced PTB (≤20%), while intraamniotic inoculation resulted in PTB in 67% of cases, supporting OOC findings.

OOC models offer a sophisticated, ethical alternative to study complex reproductive pathologies and bridge gaps between cell culture and animal research. Our findings underscore the limited pathogenicity of U. parvum alone and highlight the need for polymicrobial insults to elicit strong inflammatory responses. This platform may accelerate discovery of therapeutic targets and reduce reliance on animal use by refining preclinical infectious disease models.