Nearly one-fourth of reproductive-aged women suffer from bacterial vaginosis (BV), an infection caused by an imbalance in the naturally occurring microbes in the reproductive tract, which can increase the risk of sexually transmitted diseases, pre-term birth, and miscarriage.1,2 Despite being so common, researchers do not fully understand the disease mechanisms, in part because traditional animal models struggle to mirror the complex environment in humans.
Back in 2022, biologist and bioengineer Donald Ingber’s team at the Wyss Institute at Harvard University designed an organ-on-a-chip that modeled the human vaginal microbiome, offering researchers a more reliable system for studying BV.3 However, complications of BV extend beyond the vagina, leading Ingber’s team to build a cervix-on-a-chip system. Their new model, which they described in a paper in Nature Communications, provides a system that captures the complex interactions between the cervical epithelial cells, the mucus they produce, and the microbiome.4
To create a cervix-on-a-chip, the researchers added cells that line the cervical canal to one channel of a microfluidic device. To an adjacent channel, they introduced cervical fibroblasts that form connective tissue. A porous membrane between the channels ensured communication between the two cell types, similar to conditions in the body. The researchers emulated the mechanical cues of mucus flow in the cervix by perfusing the channels with culture medium. This led to the production of mucus, similar to what occurs in a human cervix.
The cervix-on-a-chip resembled the human organ in other ways as well. Hormonal variations during the menstrual cycle influence cervical mucus properties and functions.5 When the researchers mimicked different phases of the menstrual cycle by treating the microfluidic devices with varied levels of the hormones estrogen and progesterone, they observed that hormonal variations influenced the amount of mucus and its properties. High estrogen levels, which mimic the follicular phase, resulted in a thicker mucus layer. In contrast, increased levels of progesterone, which occurs during the luteal phase, resulted in a relative decrease in mucus production. With the help of mass spectrometry, the researchers found that the chemical composition of mucus produced by microfluidic devices exposed to hormones corresponded to previously published observations from clinical samples.
“[The mucus] looked very much like it did in humans, which is really exciting,” said Ingber. “This further validates that organs-on-chips provide a way to mimic organ-level physiology in ways that other systems don’t.”
The cervix-on-a-chip model offers a platform to study bacterial vaginosis (BV). The researchers observed a thicker layer of mucus (green) in a healthy system (top) compared to one exposed to BV-causing microbes (bottom).
Wyss Institute at Harvard University
Once they confirmed that their microfluidic device closely mirrored human cervical tissue, the team used their new model to study the effect of a disrupted microbiome on the integrity and function of the cervical canal. To mimic the cervical microbiome, they introduced Lactobacillus crispatus, a bacterium found in a healthy cervix. While they observed that cells in the channels remained intact, the addition of Gardnerella vaginalis, a bacterium associated with BV, compromised cells’ integrity and altered the production and chemical composition of the mucus.
The team conducted immunoassays and mass spectrometry to analyze the proteins produced by the cervix models. They observed that G. vaginalis exposure led to elevated levels of inflammatory cytokines and proteins like psoriasin that are increased in cervical and ovarian cancers.
The cervix-on-a-chip model is an improvement over existing methods for culturing cervical cells as they allow researchers to control mucus flow and co-culture cells with bacteria to model the microbiome, Ingber said.
“This is really an exciting study and probably one of the best in vitro models that can recapitulate human cervical biology,” said Mala Mahendroo, a reproductive biologist at University of Texas Southwestern Medical Center who was not involved in the study. “This [system] allows us to tease out [microbial] interactions in a systematic way that is difficult to do in an animal or human,” she said. She added that utilizing both animal models and in vitro platforms would be more powerful than studying either in isolation.
Ingber and his team developed the cervix-on-a-chip model using cells from non-pregnant subjects, Mahendroo noted. Her team has shown that cervical epithelial biology changes drastically during pregnancy, “So there probably needs to be a system that might better recapitulate pregnancy,” she said.6 Ingber noted that, in their model, exposure to pregnancy-related hormones gave rise to structures like the mucus plug formed during pregnancy, suggesting the system’s suitability to study pre-term labor.
At present, his team is working on linking the cervix and vagina chips to study how mucus flow between the two systems modulates BV. His team also plans to use the cervix-on-a-chip to study non-hormonal contraceptives.
