The protocol presents a mouse model of vaginal colonization with anaerobically cultured human vaginal bacteria. We focus on Gardnerella vaginalis, while including suggestions for Prevotella bivia and Fusobacterium nucleatum. This protocol can also be used as a guide for vaginal inoculations and viable recovery of other anaerobically grown bacteria.
The mammalian vagina can be colonized by many bacterial taxa. The human vaginal microbiome is often dominated by Lactobacillus species, but one-in-four women experience bacterial vaginosis, in which a low level of lactobacilli is accompanied by an overgrowth of diverse anaerobic bacteria. This condition has been associated with many health complications, including risks to reproductive and sexual health. While there is growing evidence showing the complex nature of microbial interactions in human vaginal health, the individual roles of these different anaerobic bacteria are not fully understood. This is complicated by the lack of adequate models to study anaerobically grown vaginal bacteria. Mouse models allow us to investigate the biology and virulence of these organisms in vivo. Other mouse models of vaginal bacterial inoculation have previously been described. Here, we describe methods for the inoculation of anaerobically grown bacteria and their viable recovery in conventionally raised C57Bl/6 mice. A new, less stressful procedural method for vaginal inoculation and washing is also described. Inoculation and viable recovery of Gardnerella are outlined in detail, and strategies for additional anaerobes such as Prevotella bivia and Fusobacterium nucleatum are discussed.
In mammals, the vagina is home to a consortium of bacterial species. The human vaginal microbiome is unique among mammals in its abundance of members of the genus Lactobacillus and corresponding low vaginal pH (3.8-4.5)1,2,3,4. Disruption of this Lactobacillus dominance is associated with a variety of negative health outcomes.
In bacterial vaginosis (BV) there are fewer lactobacilli and an increased abundance of diverse anaerobic bacteria, such as Gardnerella vaginalis and Prevotella bivia5,6. Women with BV are at increased risk of sexually transmitted infections7,8,9, infertility10, pregnancy losses11, preterm birth12,13,14, intrauterine infections15, cervical infections16, and cancer16,17,18. BV is also associated with a higher likelihood of vaginal colonization by potentially pathogenic bacteria, such as Fusobacterium nucleatum1,19,20, a common isolate from amniotic fluid infections21.
Due to the importance of anaerobic bacteria in human vaginal health, there is a need for animal models that can be used to investigate the biology and pathogenesis of these organisms. Here, we describe methods for vaginal inoculation and viable recovery of Gardnerella vaginalis in estrogenized mice and suggest additional strategies for Prevotella bivia and Fusobacterium nucleatum. Other models of murine vaginal colonization have previously been described, but these have focused on the inoculation and recovery of facultative anaerobic bacteria such as Group B Streptococcus22 and Neisseria gonorrhoeae23 that were cultured aerobically. Inoculation and recovery of obligate anaerobes can be successfully accomplished with appropriate experimental strategies. We discuss approaches for the viable recovery of several bacterial taxa and suggest empirical evaluation of conditions for the viability of additional species/strains of interest.
The classical method of estimating colonization by live bacteria is the recovery of colony-forming units (CFUs). In conventionally raised mice with their own endogenous microbiota, this requires recovery on agar media that selects against members of the endogenous vaginal microbiota while still allowing the inoculated strain of bacteria to grow. Here, we use a streptomycin-resistant isolate of G. vaginalis24 that can be selectively recovered on a streptomycin-containing agar medium. To ensure the medium is sufficiently selective and, conversely, that streptomycin-resistant bacteria are not present in the endogenous microbiome, vaginal lavages collected just prior to inoculation should be plated on selective (streptomycin-containing) agar.
The best method for inoculum preparation may vary among species and strains of bacteria. Prior to the start of experiments in mice, preliminary work should be performed to determine preferable conditions for the culture medium, growth endpoint, and preparation of inoculum, as well as the susceptibility to oxygen and viability in PBS. In the case of more oxygen-sensitive bacteria, alternate preparations of the inoculum can be considered (e.g., in an anaerobic culture medium with an appropriate vehicle control group)25,26.
All animal experiments were approved and conducted in accordance with institutional guidelines and the Guide for the Care and Use of Laboratory Animals of the University of California San Diego (UCSD, Protocol Number: S20057, 2020-onward) and before that at Washington University, St. Louis (Protocol 20110149, up to 2020).
NOTE: The protocol below utilizes conventionally raised female C57Bl/ mice, age 6-11 weeks at the time of inoculation. Please note vendor information and specific age ranges used previously in the relevant cited work.
1. Preparation and administration of β-estradiol 17-valerate
NOTE: β-estradiol 17-valerate is a carcinogen and a reproductive toxin that can be absorbed through the skin27,28. Wear proper PPE during the handling of powder and liquid solution. It is also an aquatic toxin29; dispose of it properly in an appropriately sealed and labeled container.
2. Preparation of inoculum
NOTE: This section contains the general steps for the preparation of inocula from anaerobically grown streptomycin-resistant Gardnerella vaginalis JCP8151B-SmR24,25,31,32. All steps in this section should be done in an anaerobic chamber.
3. Vaginal pre-lavage and inoculation with Gardnerella
4. Collection of vaginal lavages to determine viable Gardnerella colonization
5. Sacrifice, dissection, and tissue homogenization
A schematic representation of an example experiment shows some of the ways one might carry out the described protocol (Figure 1).
Some animals will carry endogenous microbes in their vaginal microbiomes that will grow on nonselective media. The methods described here rely on streptomycin-resistant isolates of the studied bacterial strains, together with selective media containing streptomycin, to select against endogenous bacteria (Figure 2A,B). It is possible for streptomycin resistance to develop spontaneously, so checking mice prior to infection (day 0 washes) can help establish if this may be a problem.
Recovery of viable bacteria from vaginal washes is accomplished by serial dilution and plating on agar media. A standard petri dish can hold up to a 6 x 6 array of 5 µL spots, allowing enumeration of bacterial titers from six replica spots across six orders of magnitude for each sample if using 10-fold serial dilutions (Figure 3A). This method can be used to enumerate titers of bacteria ranging from Gardnerella auf Prevotella and Fusobacterium (Figure 3B–D).
Together, these methods allow one to test hypotheses and make interpretations about bacterial colonization/infection of the female reproductive tract. For example, a comparison of Gardnerella titers in vaginal washes and vaginal tissue homogenates demonstrated concordance between these methods (Figure 4A). Likewise, in a vaginal co-inoculation model, titers of Prevotella in uterine tissue were significantly higher (approximately 20-fold) during Gardnerella co-infection compared to Prevotella mono-infection (Figure 4B). Finally, wild-type versus mutant bacterial strains can be compared in these models to establish the roles played by specific genes and their products in the processes of reproductive tract colonization/infection. For example, a mutant strain of Fusobacterium lacking the ability to transport and consume the carbohydrate sialic acid led to lower levels of colonization by 3 days post-inoculation (Figure 4C).
Figure 1: Schematic of an example experiment33. Please click here to view a larger version of this figure.
Figure 2: Growth of endogenous murine vaginal microbes on selective versus nonselective agar. Vaginal lavages from five female C57Bl/6 mice (1-5) were streaked onto two different NYC III agar plates and incubated anaerobically. (A) The plate on the left does not contain any antibiotics and permits the growth of endogenous anaerobic bacteria from the murine vaginal tract. (B) The plate on the right contains 1 mg/mL streptomycin and selects against the growth of endogenous bacteria. Please click here to view a larger version of this figure.
Figure 3: Recovery of viable G. vaginalis, P. bivia, and F. nucleatum CFUs from vaginal washes. (A) CFUs of G. vaginalis JCP8151B-SmR inoculum, replica-plated as a dilution series on NYC III agar. (B–D) Recovery of viable (B) G. vaginalis24, (C) P. bivia25, and (D) F. nucleatum26 from the vaginal lavages of mono-infected animals. For each graph (B–D), the data is combined from two independent experiments, with 10-15 mice per experiment24,25,26. Please click here to view a larger version of this figure.
Figure 4: Further analysis of anaerobic bacterial colonization in a murine vaginal inoculation model. (A) G. vaginalis titers in vaginal washes correlate with G. vaginalis titers recovered from vaginal tissue homogenates at 24 h and 72 h post-inoculation (combination of two independent experiments, n = 10 each. Significance determined by Spearman's rank correlation test)24. (B) Co-infection with G. vaginalis leads to increased P. bivia titers in uterine horn tissue when compared to P. bivia mono-infected animals at 48 h post-inoculation (combination of two independent experiments, each with 6-10 mice per group. Significance determined by Mann-Whitney U-test, **p = 0.0017)25. (C) F. nucleatum mutant with disrupted sialic acid transporter (Ω SiaT) has decreased titers in vaginal lavages of mono-infected mice as compared to F. nucleatum wild-type at 72 h post-inoculation (combination of two independent experiments, each with 10 mice per group. Significance determined by Mann-Whitney U-test, **p < 0.01)26. Please click here to view a larger version of this figure.
Supplementary File 1: Examples of methods used for different vaginal bacteria grown under anaerobic conditions. Please click here to download this File.
The most significant way the model described here differs from previously published vaginal inoculation models is the use of anaerobically cultured bacteria, which require special considerations for the preparation of viable inoculum and the recovery of bacteria from the mice. Specific steps in the protocol above may vary slightly depending on the species/strain of bacteria used, as suggested in Supplementary File 1. Additionally, this manuscript describes a new procedural method for the administration of bacteria and vaginal washes that requires less experience on the part of the investigator and less restraint for the mice. If the experimenter is not comfortable with the form of restraint described in step 3.2. and step 3.3., the mice can instead be scruffed as previously described34 with or without anesthesia according to the experimental design and institutional IACUC guidelines.
An important caveat to the use of anaerobically cultured bacteria in murine models is that certain steps (such as any involving live animals) must be done outside of the anaerobic chamber. Therefore, the most critical steps in this protocol are those in which the bacteria of interest will be exposed to an aerobic environment, such as during inoculation of the mice. The use of any new anaerobe in this model will require preliminary investigation into its ability to maintain viability upon exposure to oxygen (such as the comparison of pre- and post-inoculation CFUs as described in step 2.5. and step 3.4.). Depending on the degree of oxygen and PBS sensitivity of the organism, different methods of inoculum preparation should be considered (Supplementary File 1, step 2.4.5.). For inoculations using G. vaginalis JCP8151B, we have found that a single tube of inoculum can be prepared in PBS and used to infect all of the mice. If a different anaerobic bacterium is being used that is more sensitive to oxygen, the inoculum can instead be prepared in separate tubes for each mouse so that repeated opening/closing of the tube is not required. Likewise, if the bacterial species of interest is more fastidious/less capable of surviving in PBS than G. vaginalis, inoculation can instead be prepared in culture media. If the medium used contains reducing agents, such as the CDC anaerobe medium used to culture Prevotella bivia (Supplementary File 1, step 2.1. and step 2.2.), then the bacteria will also have increased protection from oxygen exposure.
Alternative methods for homogenization can also be considered, such as bead beating22,35, which is done with the tube cap closed and may, therefore, introduce less oxygen than the handheld homogenizer. Additionally, more oxygen-sensitive organisms may require a longer equilibrium time for media. An oxygen indicator such as resazurin can be used to check that anaerobic conditions have been reached (step 2.1.). Alternative methods such as anaerobic jars may affect results and should be vetted for bacterial viability prior to use.
The oxygen sensitivity and fastidiousness of the experimental organism may also play a role in the successful recovery of viable bacteria from the mouse during lavages/homogenization (Supplementary File 1, step 3.2. and step 5.1.). For highly sensitive organisms, the time between the lavage being taken and plated may lead to a loss of viability and cause an artificially low estimate of bacterial vaginal colonization. To mitigate this effect, collected lavages can immediately be diluted into fresh anaerobic culture media (with reducing agent). Similarly, organ homogenization can be done in fresh culture media rather than PBS to increase bacterial viability and can also be done in the anaerobic chamber itself to minimize oxygen exposure.
Depending on the purpose of a given experiment, the experimenter must pay attention to control groups. To test hypotheses, baseline measures of uninfected control mice that are mock-treated with vehicle alone will help establish if there is induction of chemokines/cytokines or other specific outcomes of infection. The control vehicle used must be the same as the vehicle used for inoculation, so if the bacteria are inoculated in culture media, then uninoculated culture media must be used as the control (Supplementary File 1, step 2.4.5.).
The methods described in this protocol could also be applied to facultative bacteria capable of growing anaerobically but that are traditionally studied using aerobic culture conditions (such as Escherichia coli or Group B Streptococcus). This could be especially relevant for infections by these organisms in body sites that are believed to contain low levels of oxygen, such as the cervix. It may be that a facultative organism more successfully infects sites with less oxygen when the inoculum is prepared anaerobically as compared to aerobically. In such an experiment, the recovery of viable bacteria for CFU enumeration may not require anaerobic conditions.
The authors have nothing to disclose.
We acknowledge Lynne Foster for technical assistance during the development of these models. We appreciate startup funds from the Department of Molecular Microbiology and the Center for Women's Infectious Disease Research (to AL), and the March of Dimes (Basil O'Connor award to AL) which helped support early-stage experiments. The National Institute of Allergy and Infectious Diseases (R01 AI114635) also supported the development of these models.
β-estradiol 17-valerate | Sigma | E1631 | estrogenization of mice |
0.22 µm, 50 mL Steriflip vacuum filter | Fisher | SCGP00525 | sterile filtering sesame oil |
14 mL tube | Falcon | 352059 | estrogenization of mice |
190-proof ethanol | Koptec | V1105M | dissection, tissue homogenization, CDC and Columbia media |
1x PBS | Fisher | MT21040CM | vaginal lavage, G. vaginlalis inoculum prep, tissue collection and homogenization |
25 G × 5/8 -in needle | BD | 305122 | estrogenization of mice |
5 mL tube | Falcon | 352063 | Tissue collection and homogenization |
Anaerobic chamber | Coy | 7200000 | Growth of anaerobic bacteria |
Bacto agar | Fisher | DF0140074 | G. vaginalis, F. nucleatum, and P. bivia agar plates |
Bacto peptone | Fisher | DF0118170 | CDC anaerobic media for P. bivia growth |
Columbia broth | Fisher | DF0944170 | Columbia media for F. nucleatum growth |
Defibrinated sheep blood | Hemostat | DSB500 | CDC anaerobic media for P. bivia growth |
Forceps | Fine Science Tools | 11008-13 | mouse/tissue dissection |
Glucose | Sigma | G7528 | NYCIII media for G. vaginalis growth |
Handheld homogenizer | ISC Bio | 3305500 | Tissue homogenization |
Hemin | Sigma | 51280 | CDC anaerobic media for P. bivia growth, Columbia media for F. nucleatum growth |
HEPES | Cellgro | 25-060-Cl | NYCIII media for G. vaginalis growth |
Horse serum | Hemostat | SHS500 | NYCIII media for G. vaginalis growth |
Hydrogen gas mix (5% hydrogen, 10% CO2, 85% nitrogen) | Matheson | Gas for anaerobic chamber | |
Isofluorane | VetOne | 502017 | mouse anaethesia at sacrifice |
L-cysteine | Sigma | C1276 | CDC anaerobic media for P. bivia growth |
L-glutamate | Sigma | G1626 | CDC anaerobic media for P. bivia growth |
Milli-Q Water Purifier | Millipore | IQ-7000 | for making media and homogenization |
NaCl | Sigma | S3014 | NYCIII media for G. vaginalis growth |
NaOH tablets | Sigma | S5881 | CDC anaerobic media for P. bivia growth |
Nitrogen gas | Airgas | Gas for anaerobic chamber | |
p-200 filter tips | ISC Bio | P-1237-200 | vaginal lavages and bacterial inoculations |
pancreatic digest of casein | Sigma | 70169 | CDC anaerobic media for P. bivia growth |
Proteose peptone #3 | Fisher | DF-122-17-4 | NYCIII media for G. vaginalis growth |
Scissors | Fine Science Tools | 14084-08 | mouse/tissue dissection |
Sesame oil | Fisher | ICN15662191 | estrogenization of mice |
Single edge surgical carbon steel razor blades | VWR | 55411-050 | tissue dissection |
Sodium chloride | Sigma | S3014 | CDC anaerobic media for P. bivia growth |
Spectrophotometer | BioChrom | 80-3000-45 | measuring bacterial OD600 |
Streptomycin | Gibco | 11860038 | add to agar media to make selective plates |
Tuberculin slip tip 1ml syringe | BD | 309659 | estrogenization of mice |
Vitaim K3 | Sigma | M5625 | Columbia media for F. nucleatum growth |
Vitamin K1 | Sigma | 95271 | CDC anaerobic media for P. bivia growth |
Yeast extract | Fisher | DF0127-17-9 | NYCIII media for G. vaginalis growth |