Optimization of donation collection
We recruited a pilot donor (donor 0) to initially optimize sample collection and processing procedures, and to confirm the feasibility of the protocol. Vaginal fluid was collected using a disposable menstrual cup and donor material 0 was homogenized with sterile saline solution (see Methods). The donation material was then divided into several “analytical aliquots”, which were stored at -80°C for testing Lactobacillus viability over time and a remaining “donation aliquot” to be used for possible transplantation. Donor 0 provided five donations over nine days, with a median of 400 μL of vaginal secretions per donation (range Lactobacillus The number of colony forming units (CFU) was similar in both types of aliquots: median 3.7 × 10seven CFU/mL in donation aliquot relative to 5.3 × 10seven CFU/mL in an analysis aliquot (Fig. 1a).
The microbial community composition of the donation and analysis aliquots by bacterial 16S rRNA amplicon sequencing revealed that the bacterial community profile of the two types of aliquots was nearly identical at all time points (Fig. 1b ). The relative abundances of L. crispatus and L.iners varied in time between the five donations, but L. crispatus was consistently at higher relative abundance (66.9–96.0%) than L.iners (0.35–24.9%). Nope-Lactobacillus taxa represented L.inerswhich generally does not grow on MRS (deMan, Rogosa and Sharpe) agar used for L. crispatus cultivate.
After demonstrating the feasibility of fundraising, we received FDA approval for our Investigational New Drug Protocol (#018173). Between March 2019 and November 2020, we screened 49 additional women by telephone and conducted 8 in-person screening visits to recruit three donors (donors 1–3) (Fig. 2a). Women were considered for the study if they were premenopausal, had a Nugent score of 0–3, agreed to abstain from sexual activity for the entire donation period, and denied any history of BV. A comprehensive set of inclusion and exclusion criteria were used for final selection (see Methods). Four potential donors failed screening due to a Nugent score >3, and one person was unable to have a screening lab sample due to a recent blood donation and did not return for a follow-up visit to complete screening.
All three donors identified as white, cisgender women and reported a preference for male sex partners. Donors 1 and 3 were using oral contraceptives and donor 2 was using a Mirena IUD for contraception. A complete list of screening tests is presented in Table 1, which includes tests to screen for underlying medical comorbidities (acceptable ranges in Table 1 and Table S1). Donor demographics, including age, race, and BMI, are included in Table S2. Registered donors underwent extensive testing for potentially transmissible infections at registration, at last donation visit, and 30–45 days after final donation (Fig. 2B). This ensured identification of incident infection throughout the donation period. Participants agreed to remain sexually abstinent during the donation period. Each individual donation was tested for prostate specific antigen (PSA) at the time of processing to confirm the absence of sperm, and assay aliquots for each donation were tested for HPV and again for sperm using a Y-chromosome-specific PCR test. As of June 2020, all donors underwent a nasal swab PCR test for SARS-CoV-2 before enrollment, were screened for symptoms before each donation, and each donation was tested for SARS-CoV-2 by RT-PCR.
Fundraising for potential VMTs
Donor 1 provided eight donations over 12 days, with a median volume of 0.4 ml (range 0.1 to 0.8 ml) and a median weight of 1.42 g (range 1.0 to 1, 8g) (Fig. S1). Donor 2 provided 20 donations over 37 days, with a median volume of 0.75 ml (range 0.3 to 1.1 ml) and a median weight of 1.3 g (range 1.1 to 2, 0g). Donor 3 provided 14 donations over 40 days, with a median volume of 0.55 ml (range 0.4 to 0.9 ml) and a median weight of 1.1 g (range 0.5 to 1, 5g). All donations had a pH
Characteristics of the donation for a potential VMT
For donor 1, Nugent scores for all donations ranged from 0 to 1 (Fig. 3a). For most Gram stain slides, there were no white blood cells (WBC) or 1 WBC/epithelial cell was observed in 2 donation samples (1 and 6) (Fig. 3a). A high white blood cell count often indicates a vaginal infection, but this donor tested negative for known infections. For donor 2, the Nugent score ranged from 0 to 3. As with donor 1, there were no white blood cells present or 1 WBC/epithelial cell.
To determine the bacterial community composition of TMV donors, we used 16S rRNA amplicon sequencing along with species-specific qPCR assays to L. crispatus and L.iners. We found that donor 1 and donor 3 were L. crispatus dominant throughout the duration of the donation (Fig. 3a). Donor 2, however, had varying levels of L. jensenii and L.inerswhich together constituted almost the entire community for all donation samples.
Using species-specific qPCR tests, L. crispatus and or L.iners were found to be almost mutually exclusive in these donors. In donor 2, qPCR results show elevated amounts of L.iners and not detectable L. crispatus demonstrated that despite normal Nugent scores, the primary Lactobacillus species was not L. crispatusbut rather L. jensenii. Moreover, despite L. jensenii having a higher relative abundance in 16S rRNA sequencing, the absolute amount of L.iners in donations from donor 2 was high (Fig. 3b).
Lactobacillus viability after untreated storage at −80°C
Although a previous study used fresh donations for VMT, for long-term use and scalability of donated hardware, optimal storage protocols are required. We sought to maximize the viability of bacteria in the samples over time and to minimize any changes to the material. We therefore frozen assay aliquots from each donation without the addition of glycerol or other cryoprotectants at −80°C and used the aliquots to quantify the stability of Lactobacillis. To measure viable Lactobacillis CFU after storage at −80°C, aliquots of each donation were serially diluted and plated on Lactobacillus-MRS selective agar plates (Fig. 3c). For Donor 1, Lactobacillus viability (CFU/mL of sample) was stable between 30 days (median 2.8 × 108range 4.5 × 10seven–1.0 × 109), 3 months (median 5.2 × 108range 5.3×10seven–9 × 108), 6 months (median 3.7 × 108 range 1.2 × 108–1.5 × 109), and 11 months (median 1.7 × 108 range 5.5 × 10seven–1.1 × 109) after collection. For Donor 2, Lactobacillus viability (CFU/mL) was initially stable between 30 days (median 2.0 × 108range 4.7×10seven–5.6 × 108) and 3 months (median 1.1 × 108range 1.9 × 10seven–2.9 × 108) but decreased at 9 months (median 5.2 × 106range 8 × 105–1.1 × 10seven). This decline can be attributed to the difference in community composition and associated differences between L. crispatus compared to the mixture of L. jensenii and L.iners, the latter generally not growing on MRS. Donor 3 was similar to donor 1, with stable CFU counts when comparing data at collection (median 1.8 × 108 CFU/mL, range 1.5 × 10seven–5.8 × 108 CFU/mL), 3 months (median 1.6 × 108 CFU/mL, range 3.6 × 10seven–2.8 × 108 CFU/mL) and 6 months (median 1.2 × 108 CFU/mL, range 3.4 × 10seven at 3.8×108 CFU/mL). This suggests that donations with L. crispatus dominant communities can be stored without the need for cryoprotectants.