This article describes a protocol for creating a microfluidic vagina-on-a-chip (Vagina Chip) culture device that enables the study of human host interactions with a living vaginal microbiome under microaerophilic conditions. This chip can be used as a tool to investigate vaginal diseases as well as to develop and test potential therapeutic countermeasures.
Women’s health, and particularly diseases of the female reproductive tract (FRT), have not received the attention they deserve, even though an unhealthy reproductive system may lead to life-threatening diseases, infertility, or adverse outcomes during pregnancy. One barrier in the field is that there has been a dearth of preclinical, experimental models that faithfully mimic the physiology and pathophysiology of the FRT. Current in vitro and animal models do not fully recapitulate the hormonal changes, microaerobic conditions, and interactions with the vaginal microbiome. The advent of Organ-on-a-Chip (Organ Chip) microfluidic culture technology that can mimic tissue-tissue interfaces, vascular perfusion, interstitial fluid flows, and the physical microenvironment of a major subunit of human organs can potentially serve as a solution to this problem. Recently, a human Vagina Chip that supports co-culture of human vaginal microbial consortia with primary human vaginal epithelium that is also interfaced with vaginal stroma and experiences dynamic fluid flow has been developed. This chip replicates the physiological responses of the human vagina to healthy and dysbiotic microbiomes. A detailed protocol for creating human Vagina Chips has been described in this article.
A vaginal microbiome dominated by Lactobacillus spp. that helps to maintain an acidic microenvironment plays an important role in maintaining female reproductive health1. However, at times there can be a change in the composition of microbial communities that comprise the microbiome, which results in an increase in the diversity of vaginal bacteria. These dysbiotic changes, which often result in a switch from a Lactobacillus-dominated state to one dominated by more diverse anaerobic bacterial species (e.g., Gardnerella vaginalis), are associated with various diseases of the reproductive system, such as bacterial vaginosis, atrophic vaginitis, urinary tract infection, vulvovaginal candidiasis, urethritis, and chorioamnionitis2,3,4,5. These diseases, in turn, increase a woman's chances of acquiring sexually transmitted diseases and pelvic inflammatory disease6,7,8,9. They also pose a higher risk for pre-term birth and miscarriages in pregnant women10,11,12 and have also been implicated in infertility13,14,15,16.
Although efforts have been made to model vaginal dysbiosis using vaginal epithelial cells cultured in static, two-dimensional (2D) culture systems17,18, they do not effectively mimic the physiology and complexity of the vaginal microenvironment19. Animal models also have been used to study vaginal dysbiosis; however, their menstrual phases and host-microbiome interactions differ greatly from that in humans, and thus, the physiological relevance of results from these studies remains unclear19,20,21. To counteract these issues, organoids and Transwell insert models of human vaginal tissue also have been used to study host-pathogen interactions in the FRT19,22,23,24. But because these are static cultures, they can only support co-culture of human cells with living microbes for a short period of time (<16-24 h), and they lack many other potentially important physical features of the human vaginal microenvironment, such as mucus production and fluid flow22.
Organ Chips are three-dimensional (3D) microfluidic culture systems that contain one or more parallel hollow microchannels lined by living cells cultured under dynamic fluid flow. The two-channel chips enable the recreation of organ-level tissue-tissue interfaces by culturing different cell types (e.g., epithelium and stromal fibroblasts or epithelium and vascular endothelium) on opposite sides of a porous membrane that separates the two parallel channels (Figure 1). Both tissues can be independently exposed to fluid flow, and they can also experience microaerobic conditions to enable co-culture with a complex microbiome25,26,27,28. This approach was recently leveraged to develop a human Vagina Chip lined by hormone-sensitive, primary vaginal epithelium interfaced with underlying stromal fibroblasts, which sustains a low physiological oxygen concentration in the epithelial lumen and enables co-culture with healthy and dysbiotic microbiomes for at least 3 days in vitro29. It was demonstrated that the Vagina Chip could be used to study colonization by optimal (healthy) L. crispatus consortia and detect inflammation and injury caused by non-optimal (non-healthy) G. vaginalis containing consortia. Here, we describe in detail the methods that are used to create the human Vagina Chip as well as to establish healthy and dysbiotic bacterial communities on-chip.
This research was performed in compliance with institutional guidelines for the use of human cells. The cells were obtained commercially (see Table of Materials). All steps should be performed aseptically in a biosafety cabinet (BSC). Use only filter (or barrier) pipette tips for this protocol.
1. Culturing human vaginal epithelial cells
2. Culturing human uterine fibroblast cells
3. Chip activation and channel coating
4. Seeding chip basal channel with HUFs
5. Seeding chip apical channel with vaginal epithelial cells
6. Connecting chips to pods and differentiating vaginal epithelial cells
7. Bacterial inoculation of differentiated chips
NOTE: Perform the following steps in a Lab and BSC that comply with regulations to handle microbes.
8. Analysis of chip effluents and digests
The human vagina is lined by a stratified epithelium that overlies a fibroblast-rich collagenous stroma. To model this, a tissue interface was created by culturing primary human vaginal epithelium and fibroblasts on opposite sides of a common porous membrane within a two-channel microfluidic Organ Chip device. Formation of the vaginal epithelium is monitored using bright field microscopic imaging, which reveals the formation of a continuous sheet of cells that progressively forms multiple cell layers (Figure 2A). Previous reports confirmed that this morphology correlates with the development of a fully stratified epithelium when viewed in cross-section29. However, if the epithelial layer appears patchy and discontinuous (Figure 2B), the Vagina Chip may not be fit for use in experiments.
Figure 3 shows a schematic representation of the generation of the Vagina Chip. To validate the functionality of the Vagina Chip, the chips were inoculated with L. crispatus and G. vaginalis to model healthy and dysbiotic vaginal environments, respectively. G. vaginalis is the bacterium primarily involved in bacterial vaginosis. To check if healthy and dysbiotic bacteria engraft on the Vagina Chips, the bacterial load was quantified in the Vagina Chips inoculated with the different bacterial populations by plating channel effluents and digested cell layers on selective bacterial growth media (De Man-Rogosa-Sharpe (MRS) agar for L. crispatus and Brucella blood agar (BBA) for G. vaginalis)(see Table of Materials) and comparing them to similar cultures using the original inoculum. Colonies of L. crispatus and G. vaginalis were detected within 48 h of plating (Figure 2C), confirming that both healthy and dysbiotic bacteria engrafted in the Vagina Chips. However, if bacterial colonies are observed on the plates containing the inoculums but not observed on plates containing the effluent or digest after the required incubation, then it can be concluded that the bacteria did not engraft.
A healthy vaginal environment is acidic, and dysbiosis results in an increase in vaginal pH31. Therefore the pH of the effluent of the apical epithelial channel of the Vagina Chip was also analyzed. The pH of Vagina Chips inoculated with L. crispatus was similar to that of uninoculated control chips, and when co-cultured with G. vaginalis they experienced significantly increased pH (Figure 2D). If the pH of an uninfected Vagina Chip is observed to be high, it indicates that there is a problem, and these chips should not be used for experiments.
The inflammatory state of vaginal tissue is also sensitive to the composition of the vaginal microbiome, with a dysbiotic microbiome stimulating inflammation. Upon analysis of the pro-inflammatory cytokines in the apical channel of the Vagina Chip 3 days after inoculation with either L. crispatus or G. vaginalis, the pro-inflammatory response was similarly found to be higher with G. vaginalis compared to uninfected chips and chips inoculated with L. crispatus (Figure 2E). Taken together, these results show that the Vagina Chip closely mimics the human vaginal microenvironment in both healthy and dysbiotic states.
Figure 1: A two-channel chip and its pod. (A) Image of a two-channel PDMS chip depicting its channels and ports. (B) Image of a pod depicting the reservoirs and ports for apical and basal channels. (C) Schematic diagram showing the cross-section view of a Vagina Chip infected with microbes. Please click here to view a larger version of this figure.
Figure 2: Vagina Chip mimics healthy and dysbiotic human vaginal microenvironments. (A) Vaginal epithelial cells in the apical channel of a robust Vagina Chip. Scale bar represents 1 mm of the chip in the top image and 500 µm in the bottom image. (B) Vaginal epithelial cells in the apical channel of an inadequate Vagina Chip. Scale bar represents 1 mm of the chip in the top image and 500 µm in the bottom image. (C) Engraftment of L. crispatus (LC) and G. vaginalis (GV) in the Vagina Chip. (D) pH of Vagina Chips after 72 h incubation with L. crispatus (LC) and G. vaginalis (GV) as compared to uninfected (control) Vagina Chips. (E) Pro-inflammatory response of Vagina Chips to L. crispatus (LC) and G. vaginalis (GV) after 72 h incubation, as compared to uninfected (control) Vagina Chips. (C–E) Graphs depict mean ± SD for 4-6 chips; *p < 0.05, **p < 0.01, ***p < 0.001 as compared to control Vagina Chips; Each (●) represents data from 1 chip. Please click here to view a larger version of this figure.
Figure 3: Schematic representation of the protocol for the generation of the Vagina Chip. Schematic representation of the steps involved in seeding cells and generation of the Vagina Chip. HVECs – Human vaginal epithelial cells; HUFs – Human uterine fibroblasts; VEM – Vaginal epithelial medium; FM – Fibroblast media; DM – Differentiation medium; PS – Penicillin Streptomycin. Please click here to view a larger version of this figure.
Past in vitro models of the human vagina do not faithfully replicate vaginal tissue structures, fluid flow, and host-pathogen interactions19,22. Animal models are also limited by inter-species variation in microbiome and differences in the estrous or menstrual cycle19,22. This manuscript describes a protocol to create an Organ Chip model of the human vagina that can effectively mimic human responses to healthy and dysbiotic microbial communities.
This protocol involves seeding vaginal epithelial and fibroblast cells on opposite sides of a shared porous membrane that separates parallel microchannels in a two-channel Organ Chip device that is commercially available (see Table of Materials). The porous membrane allows for the migration of growth factors and other forms of intercellular communication. However, the collagen coating and presence of the cell monolayers prevent the mixing of media between channels. Upon the formation of a vaginal epithelial cell monolayer in the apical channel, differentiation factors are introduced into the medium flowing in the basal channel, which passes through the interstitial space and thereby promotes differentiation of the vaginal epithelial cells to form a stratified epithelium. The density of the vaginal epithelial cells at the time of seeding is a crucial determinant of the health of the Vagina Chip at the end of the differentiation phase. Thus, the density of vaginal epithelial cells should be assessed before initiating differentiation, which should not be initiated until a monolayer is established. Exposure to the differentiation factors can be continued until the desired density of the vaginal epithelial cells is obtained and before bacterial inoculation. Further, it should be noted that the growth rate may vary for primary vaginal epithelial cells from different donors (or commercial sources), which could affect the quality of the Vagina Chip generated. In all microfluidic Organ Chip studies, it is of utmost importance to remove any bubbles that might form in the channels throughout the chip culture as they interfere with the medium flow and will eventually result in reduced nutrient availability and a loss of cell viability.
This protocol also describes how to use the Vagina Chip to establish bacterial communities on-chip that mimic either the healthy vaginal state or bacterial vaginosis. The Vagina Chip also can be used to study other vaginal diseases or disorders; however, care should be taken to understand the characteristics of each disease and the best means for correlating results with clinical findings when carrying out these studies. In summary, the human Vagina Chip opens new avenues to study a plethora of diseases and conditions related to the FRT, and it can be a valuable tool for investigating potential therapeutics.
The authors have nothing to disclose.
This research was sponsored by funding from the Bill and Melinda Gates Foundation (INV-035977) and the Wyss Institute for Biologically Inspired Engineering at Harvard University. We also thank Gwenn E. Merry, Wyss Institute, for editing this manuscript. The diagram in Figure 1 has been created with BioRender.
0.22 µm Steriflip | Millipore | SCGP00525 | To degas media |
2 channel chip | Emulate | BRK-S1-WER-24 | Part of the two-channel Chip kit |
200 μL barrier tips (or filter tips) | Thomas Scientific, SHARP | 1159M40 | Tips used for chip seeding |
Activation Reagent 1 (or ER-1 powder) | Emulate | Chip S1 Basic Research kit-24PK | Part of the two-channel Chip kit; Storage temperature -20 °C |
Activation Reagent 2 (or ER-2 solution) | Emulate | Chip S1 Basic Research kit-24PK | Part of the two-channel Chip kit; Storage temperature 4 °C |
Adenine | Sigma Aldrich | A2786 | Component of the Differentiation media |
Brucella blood agar plates | VWR International Inc. | 89405-032 | with Hemin and Vitamin K; For the enumeration of Gardnerella vaginalis |
Ca2+ and Mg2+ free DPBS (DPBS (-/-) | ScienCell | 303 | For washing cells |
Calcium Chloride | Sigma Aldrich | C5670 | Component of the Differentiation media |
Calcium chloride (anhyd.) | Sigma Aldrich | 499609 | Component of HBSS (LB/+G) |
Collagen I | Corning | 354236 | For the coating solution for HVEC |
Collagen IV | Sigma Aldrich | C7521 | For the coating solution for HVEC |
Collagenase IV | Gibco | 17104019 | For the dissociation of cells from the Vagina Chips |
Complete fibroblast medium | ScienCell | 2301 | Media for the culture of HUF |
Complete vaginal epithelium medium | Lifeline | LL-0068 | Media for the culture of HVEC |
D-Glucose (dextrose) | Sigma Aldrich | 158968 | Component of HBSS (LB/+G) |
DMEM (Low Glucose) | Thermofisher | 12320-032 | Component of the Differentiation media |
Dynamic Flow Module (or Zoë) | Emulate | Zoë-CM1 | Regulates the flow rate of the chips |
Ham's F12 | Thermofisher | 11765-054 | Component of the Differentiation media |
Heat inactivated FBS | Thermofisher | 10438018 | Component of the Differentiation media |
Human uterine fibroblasts | ScienCell | 7040 | HUF |
Human vaginal epithelial cells | Lifeline | FC-0083 | HVEC |
Hydrocortisone | Sigma Aldrich | H0396 | Component of the Differentiation media |
ITES | Lonza | 17-839Z | Component of the Differentiation media |
L-glutamine | Thermofisher | 25030081 | Component of the Differentiation media |
Magnesium chloride hexahydrate | Sigma Aldrich | M2393 | Component of HBSS (LB/+G) |
Magnesium sulfate heptahydrate | Sigma Aldrich | M1880 | Component of HBSS (LB/+G) |
MRS agar plates | VWR International Inc. | 89407-214 | For enumeration of Lactobacillus |
O-phosphorylethanolamine | Sigma Aldrich | P0503 | Component of the Differentiation media |
Pen/Strep | Thermofisher | 15070063 | Component of the Differentiation media |
pH strips | Fischer-Scientific | 13-640-520 | For measurement of pH |
Pods (1/chip) | Emulate | BRK-S1-WER-24 | Part of the two-channel Chip kit |
Poly-L-lysine | ScienCell | 403 | For the coating solution for HUFs |
Potassium chloride | Sigma Aldrich | P3911 | Component of HBSS (LB/+G) |
Potassium phosphate monobasic | Sigma Aldrich | P0662 | Component of HBSS (LB/+G) |
Sterile 80% glycerol | MP Biomedicals | 113055034 | For freezing bacterial samples |
Triiodothyronine | Sigma Aldrich | T6397 | Component of the Differentiation media |
Trypan Blue Solution (0.4%) | Sigma Aldrich | T8154 | For counting live/dead cells |
TrypLE Express | Thermofisher | 12605010 | For the dissociation of cells from the Vagina Chips |
Trypsin Neutralizing Solution (TNS) | ScienCell | 113 | For neutralization of Trypsin |
Trypsin/EDTA Solutiom (0.25%) | ScienCell | 103 | For cell dissociation |
β-estradiol | Sigma Aldrich | E2257 | Hormone for differentiation media |