|Home | About | Journals | Submit | Contact Us | Français|
Glycerol monolaurate (GML) is a fatty acid monoester that inhibits growth and exotoxin production of vaginal pathogens and cytokine production by vaginal epithelial cells. Because of these activities, and because of the importance of cytokine-mediated immune activation in human immunodeficiency virus type 1 (HIV-1) transmission to women, our laboratories are performing studies on the potential efficacy of GML as a topical microbicide to interfere with HIV-1 transmission in the simian immunodeficiency virus-rhesus macaque model. While GML is generally recognized as safe by the FDA for topical use, its safety for chronic use and effects on normal vaginal microflora in this animal model have not been evaluated. GML was therefore tested both in vitro for its effects on vaginal flora lactobacilli and in vivo as a 5% gel administered vaginally to monkeys. In vitro studies demonstrated that lactobacilli are not killed by GML; GML blocks the loss of their viability in stationary phase and does not interfere with lactic acid production. GML (5% gel) does not quantitatively alter monkey aerobic vaginal microflora compared to vehicle control gel. Lactobacilli and coagulase-negative staphylococci are the dominant vaginal aerobic microflora, with beta-hemolytic streptococci, Staphylococcus aureus, and yeasts sporadically present; gram-negative rods are not part of their vaginal flora. Colposcopy and biopsy studies indicate that GML does not alter normal mucosal integrity and does not induce inflammation; instead, GML reduces epithelial cell production of interleukin 8. The studies suggest that GML is safe for chronic use in monkeys when applied vaginally; it does not alter either mucosal microflora or integrity.
Glycerol monolaurate (GML), a fatty acid monoester, is generally recognized as safe by the Food and Drug Administration for topical skin and mucous membrane uses at doses up to 100 mg/ml; this is based on many years of experience with GML as an additive to cosmetics and foods. Previously, the compound was shown in vitro to inhibit exotoxin production by gram-positive bacteria, including production of staphylococcal toxic shock syndrome (TSS) toxin-1 and alpha-toxin, at concentrations of ≥20 μg/ml, acting at the level of transcription (24, 28, 35). Streptococci and other gram-positive cocci, which do not produce glycerol ester hydrolases (lipases), are killed by GML at concentrations of ≥10 μg/ml, but production of their exotoxins is inhibited at even lower GML doses (28). Staphylococcus aureus and coagulase-negative staphylococci secrete lipases, and thus, the organisms are resistant to killing by GML except at concentrations approaching 500 μg/ml (28). Studies to assess the effect of GML on normal vaginal microflora lactobacilli have not been performed.
Unlike gram-positive cocci, gram-negative Enterobacteriaceae are resistant to GML at even very high concentrations (>2,000 μg/ml); however, rough mutants, lacking intact lipopolysaccharide (LPS), are highly susceptible to the bactericidal effects of GML (at concentrations of ≥20 μg/ml), indicating that the external components of LPS facilitate GML resistance (28). Interestingly, other gram-negative bacteria, such as Helicobacter, Haemophilus, and Gardnerella spp., which have lipooligosaccharides instead of LPS like Enterobacteriaceae, are susceptible to the killing effects of GML (≥20 μg/ml) (unpublished observations).
Some studies have evaluated the effect of monoglycerides, including GML, on enveloped viruses. Thormar et al. demonstrated in vitro that GML at concentrations as low as 250 μg/ml causes significant reductions in vesicular stomatitis virus, herpes simplex virus, and visna virus infectious titers (33, 34). Other investigators also showed monoglycerides in vitro reduce herpes simplex virus 1 and 2 titers, though in these studies, GML was not tested (26). We have shown in vitro that GML at 100 μg/ml inhibits human immunodeficiency virus type 1 (HIV-1) infection of human T cells (unpublished data).
GML is considered a surfactant, and thus, the compound might be expected to solubilize lipid bilayers (13, 14). However, unlike many surfactants, GML is relatively insoluble at body temperature, with a solubility limit in aqueous solutions of 50 to 100 μg/ml. We previously showed that the major effect of GML is plasma membrane stabilization and interference of signal transduction rather than solubilization (21). For example, GML stabilizes red blood cells, preventing lysis by either hypotonic solutions or bacterial cytolysins (21). GML prevents superantigen and LPS activations of T and B lymphocytes, respectively (36). The compound inhibits superantigen-induced production of proinflammatory cytokines by human vaginal epithelial cells (HVECs) (21) and recently was shown to reduce human vaginal interleukin 8 (IL-8) production when added to tampons (20a). GML does not appear cytotoxic for mammalian cells, but rather their metabolic functions in the presence of GML (100 μg/ml) are slowed (21). These slowed cellular functions result from transitory GML effects on plasma membranes, returning to normal when GML is eliminated by cells, presumably through the action of mammalian lipases.
We hypothesized that GML may be useful in preventing both vaginal pathogen transmission and pathogenesis because of its in vitro ability to (i) inhibit gram-positive bacterial pathogen growth and exotoxin production, (ii) reduce enveloped virus infection of mammalian cells, and (iii) stabilize mammalian cell plasma membranes. The present studies assessed the safety of GML as a potential topical microbicide to prevent HIV-1 transmission to women, with use of a 5% GML gel administered in the simian immunodeficiency virus-rhesus macaque animal model. Studies to assess its effects on lactobacilli in vitro were also undertaken.
Our studies demonstrate that GML neither has adverse effects on normal flora lactobacilli vaginally nor alters mucosal surface integrity due to its surfactant properties. Even though lactobacilli are gram positive, they are not inhibited from growing in vitro by GML. Additionally, daily vaginal application of GML does not alter vaginal microflora lactobacilli of rhesus monkeys over a period of 6 months. This study provides the first quantitative determination of rhesus monkey aerobic vaginal microflora and indicates that GML potentially may be useful as a safe topical microbicide to prevent transmission and pathogenesis due to vaginal pathogens while at the same time maintaining normal flora lactobacilli and mucosal integrity.
A vaginal isolate of Lactobacillus crispatus and a milk isolate of L. casei were used for in vitro analyses. Staphylococcus aureus MN8 is a typical menstrual vaginal TSS organism; the strain produces proinflammatory TSS toxin-1 and cytolysins (2). These organisms are maintained in low passage in the Schlievert laboratory as lyophilized stocks.
Growth curve studies of lactobacilli in the presence of nonformulated GML were performed with Todd-Hewitt broths (Becton Dickinson and Co., Sparks, MD; 25 ml of medium added to 125-ml Erlenmeyer flasks). The cultures were incubated with slow shaking (100 rpm) in a 37°C incubator, with samples removed at designated times for pH and CFU determination on Todd-Hewitt agar plates.
Five percent GML-formulated gel (GML dissolved in K-Y warming gel at the Fairview Compounding Pharmacy, Minneapolis, MN; hereafter referred to as GML-formulated gel) for administration to monkeys was added to square paper disks (30 μl) in vitro. These disks were then placed on petri plates with lawns of L. crispatus and S. aureus MN8 that contained approximately 106 bacteria. The plates were then incubated overnight at 37°C in 7% CO2; the plates were examined for visible inhibition of bacterial growth, indicated by microbes not growing up to the disks. These experiments were performed in quadruplicate with identical results.
The HVEC line was described in detail previously (20). The cells were cultured on plastic in keratinocyte serum-free medium (KSFM; Gibco, Invitrogen, Carlsbad, CA) containing penicillin, streptomycin, and amphotericin B (Gibco) in a 7% CO2 incubator (20). Two to three days prior to experimentation, the HVECs were split 1:4 and cultured in 96-well microtiter plates in the presence of KSFM without antibiotics.
HVECs were used to assess production of the proinflammatory chemokine IL-8 in response to stimulation with L. crispatus and S. aureus MN8 (20). Bacteria were cultured overnight in Todd-Hewitt broth (Becton Dickinson and Company, Sparks, MD). The next day, the bacteria were washed by centrifugation (1,000 × g; 15 min) and resuspended in phosphate-buffered saline (0.005 M NaPO4, pH 7.2, 0.15 M NaCl). Bacteria in triplicate were added to 96-well microtiter plates containing lawns of HVECs (100 μl/well of KSFM). The plates were incubated with bacteria for 6 h, and then IL-8 production was measured by enzyme-linked immunosorbent assay (R and D Systems, Minneapolis, MN).
Twelve female rhesus macaque monkeys 6 to 9 years of age were used in the study. Monkeys were housed in standard stainless steel primate cages (Surburban Surgical, Chicago, IL): eight were singly housed, while four were housed in two pairs of two. All animals had visual and auditory contact with each other in the same room. They were fed twice daily with commercial chow (20% protein primate diet, catalog no. 2050; Harlan Teklad, Madison, WI) and also given a variety of fruit enrichment in the afternoons. Housing rooms were maintained at 65 to 75°F, 30 to 70% humidity, and on a 12:12 light-dark cycle (on, 6:00 a.m.; off, 6:00 p.m.).
For 5% GML-formulated gel and vehicle control gel dosing, as well as collection of vaginal swabs for microbial or GML measurement, the animals were transferred to a tabletop restraint device using a transfer box (25). They were gently restrained in a position to acquire the swab or administer the dose.
Prior to dosing, obtaining vaginal swabs, and performing colposcopy and biopsy, the genital region of each animal was wiped with a dilute (1:1 in saline) chlorhexidine gluconate solution in a single upward motion going from vagina to anus, followed by a clean, dry 2-by-2-inch gauze wiped in the same direction so as not to contaminate the vagina with fecal material. The vagina was manually opened slightly, and a 1-ml syringe without a needle was inserted atraumatically into the vagina until approximately the 0.4-ml mark. A dose of administered agent was delivered into the vagina, and the syringe was removed; animals were dosed either once or twice a day, between 7:00 a.m. and 8:00 a.m. and/or between 7:00 p.m. and 8:00 p.m. Control animals received 1 ml of vehicle control gel, and GML animals received 1 ml of gel containing 50 mg GML formulated in K-Y warming gel.
For collection of vaginal swabs, the swabs, whether for microbial or GML measurement, remained in the vagina for approximately 1 min before they were carefully removed and placed in appropriate containers.
For colposcopy and biopsy, each animal was anesthetized and placed in a sternal position with her posterior elevated approximately 30 degrees from horizontal. The tail of the animal was taped in a position to allow for access to the vagina and to keep the genital area clean; the genital region was cleaned as described above. The vagina was manually opened with a speculum, and a 4-mm lighted rigid endoscope (Linvatec, Utica, NY) was positioned in the vagina to obtain an image of the cervix. The images were projected onto a monitor, and selected images were archived digitally. These images served as a way to evaluate gross inflammation of the cervical area (hyperemia and/or erosion). A scale of no hyperemia/epithelial erosion, moderate hyperemia/epithelial erosion, and severe hyperemia/epithelial erosion was used to classify the range of hyperemia or epithelial erosion seen for each animal at multiple time points. For a result to be classified as no hyperemia or epithelial erosion, there could be no visible redness or areas of cellular disruption on the vagina or cervix. For a classification of moderate hyperemia or epithelial erosion, there could be up to two small areas of redness or cellular disruption on the vagina or cervix. For a classification of severe hyperemia or epithelial erosion, there would be more than two areas of redness or epithelial disruption or a general redness of the vaginal or cervical tissue. We performed cervical biopsy immediately following colposcopy evaluation, while the endoscope was still in place, by utilizing a Baby Tischler punch biopsy device. The scope and biopsy device were used in unison to obtain a cervical biopsy sample (biopsy size, approximately 2 by 2 by 2 mm). All efforts were made to obtain biopsies from different sites each time for each animal. If needed, sterile cotton-tipped applicators were used to place direct pressure on the biopsy site. In rare cases, a larger sterile cotton-tipped applicator was left in the vaginal opening for up to 3 h to ensure hemostasis. Buprenorphine (0.01 to 0.03 mg/kg of body weight) was given immediately after completion of the biopsy. For 5 days following the procedure, animals were monitored for signs of pain or distress and given additional analgesics as needed. The cervical biopsy sample was placed in a cassette (Surgipath, Richmond, IL) into 4% paraformaldehyde for approximately 4 h, transferred to 70% ethanol, and shipped priority overnight to the University of Minnesota, Minneapolis, MN.
Vaginal swabs for microflora determination were obtained from the 12 monkeys over the period from 18 April 2007 to 25 September 2007 (a total of 56 samples from all except one animal); studies with one animal (vehicle control group monkey) began two weeks after those with the other 11 monkeys because the 12th animal was not available until that time (these studies began on 7 May 2007). Nine of the animals per treatment time point were treated vaginally with 1 ml of GML (50 mg/ml)-formulated gel; three of the animals per time point were treated with the same volume of formulated gel in the absence of GML. The protocol for gel administration and sampling for both GML and vehicle control gel-treated groups was as follows. (i) For 8 weeks (18 April 2007 to 19 June 2007), all monkeys were treated with GML or vehicle control twice daily (once in the morning and once in the evening; Fig. Fig.11 shows the timeline for vaginal gel administration and sampling), the animals were then rested for 3 weeks, during which time they received no treatments, and the monkeys were then given GML or vehicle control gel one time daily (morning) for 11 weeks (9 July 2007 to 25 September 2007). (ii) As well, one time in the morning (one hour after administration of GML) and one time in the afternoon (one hour after administration of GML when administered twice daily) on two consecutive designated days (Monday and Tuesday), vaginal swabs from each animal were taken for vaginal microflora determination. These swabs were experimentally determined to absorb on average 0.1 ml of vaginal secretions. The swabs were immediately placed in conditions of 4°C and shipped at 4°C from the Wisconsin National Primate Research Center to the Schlievert laboratory, University of Minnesota, Minneapolis, MN. The swabs were placed in 0.9 ml of 4°C Todd-Hewitt broth and then serially diluted in additional Todd-Hewitt broths for quantitative counts on chocolate agar plates that were incubated aerobically at 37°C in a 7% CO2 incubator. Aerobic microflora were identified as follows: (i) lactobacilli, which were considered gram-positive rods that comprised the predominant gram-positive rod-shaped bacterial microflora; (ii) staphylococci, which were identified as gram-positive cocci that are catalase positive and then subdivided into coagulase positive (S. aureus) and coagulase negative; (iii) streptococci, which were identified only as beta-hemolytic gram-positive cocci that were catalase negative (streptococcal colonies were tiny with large zones of hemolysis on blood agar plates, suggesting that they were group F streptococci); (iv) aerobic gram-negative rods (these organisms were exceptionally rare and thus were not identified more fully); and (v) yeasts. No attempts to determine GML effects on obligate anaerobic microflora were made. When S. aureus was present vaginally, the organisms were tested for superantigen gene content by PCR; these studies were performed to determine the clonal relatedness of isolates.
Differences in in vitro and in vivo counts of vaginal lactobacilli and coagulase-negative staphylococci in monkeys treated with either GML or K-Y warming gel control were assessed by determination of means and standard deviations and through use of unpaired t test analysis of normally distributed data at sampling time points.
Initially, we tested the effects of GML (nonformulated) on the growth and acid production of an L. crispatus strain (Fig. (Fig.2).2). GML (20 μg/ml and 50 μg/ml) had minimal effects on the growth of L. crispatus over the 24-h time period (P > 0.05); GML (100 μg/ml) slowed but did not prevent the growth of the organism (Fig. (Fig.2A).2A). A major end product of lactobacillus metabolism is lactic acid. In the absence of GML and in the presence of GML (50 μg/ml), L. crispatus produced comparable levels of acids, as assessed by comparable reductions in pH of the cultures (P > 0.05 at all time points; Fig. Fig.2B).2B). Lactobacilli are the predominant normal vaginal flora of humans (1, 16, 17; also as shown in this study for rhesus monkeys). As such, it is possible the organisms are present at or near stationary phase, lining mucosal surfaces. Thus, we assessed the effect of GML at 100 μg/ml on stationary-phase cultures (Fig. (Fig.2C);2C); GML blocked the loss of viability of L. crispatus in stationary phase, maintaining the organisms at stationary phase levels for the entire 48-h test period. In contrast, the control cultures, without GML, showed reduced bacterial counts over the 48-h period (P values of <0.001 at 24 and 48 h in comparison to GML-treated cultures). These data were reproduced two times with tests on L. crispatus. In addition, comparable results were obtained when the same tests were performed with L. casei (data not shown).
The 5% GML-formulated gel was also tested for effects on L. crispatus (Fig. (Fig.3).3). The GML-formulated gel did not prevent the growth of L. crispatus as indicated by growth up to the GML-impregnated disk. As a control to ensure that the formulated GML could diffuse into the surrounding medium, the 5% GML gel inhibited the growth of S. aureus MN8 adjacent to the disk. Identical results were obtained for three additional replicates.
Research suggests that the presence of certain spermicides and potentially pathogenic microorganisms leads to production of proinflammatory cytokines vaginally (3-10, 15, 22, 30, 31). The microorganisms may directly cause infections but also may increase host susceptibility to transmission of other infections, including HIV infection (9, 15, 22).
We assessed the ability of normal flora lactobacilli, compared to the ability of potential pathogen S. aureus, present vaginally to induce proinflammatory cytokines from HVECs, with use of IL-8 production as a measure of inflammation (Fig. (Fig.4).4). Interestingly, L. crispatus, a member of the normal vaginal flora, did not induce IL-8 production at either bacterial concentration used. In contrast, comparable doses of S. aureus MN8 either killed HVECs (108 bacteria) or stimulated production of IL-8 (106 and 107 bacteria) (P values of <0.001 for both S. aureus doses in comparison to IL-8 production by untreated HVECs or HVECs induced with either Lactobacillus concentration). GML (100 μg/ml) prevented IL-8 production by HVECs treated with S. aureus MN8 (data not shown), consistent with our previous findings that GML reduces production of IL-8 due to exotoxins secreted by S. aureus MN8 (21).
In anticipation of testing GML as a topical microbicide potentially to manage human vaginal infections and transmission of HIV, we evaluated the safety of GML administered daily to nine monkeys over a 6-month time period (Fig. (Fig.1).1). These animals were compared to monkeys receiving vehicle control. A gel that contained 5% GML was formulated; control gel was the same formulation except GML was omitted. The gel used, including emulsion vehicle, is approved by FDA for human vaginal use. As indicated in Fig. Fig.1,1, for the first 8 weeks, monkeys received two doses of GML (or control) gel daily, and after a three-week rest period, received one dose of GML (or control) gel daily for 11 weeks. Across this time period, quantitative aerobic vaginal microflora, colposcopy, and biopsy studies were performed to assess GML safety.
Like the microflora of humans, the rhesus monkey vaginal microflora was dominated by lactobacilli (Fig. (Fig.5).5). Numbers of lactobacilli remained fairly constant on the vaginal swabs over the entire 6-month time period. Except for sampling points 3, 4, 6, and 10, when there were only two animals receiving vehicle control gel (P < 0.05), there was no significant difference between the lactobacillus counts of GML-treated and vehicle control animals at any remaining time points (P > 0.05). The data indicate that the 5% GML gel did not cause major alterations in normal vaginal lactobacillus counts.
Additional studies to specify the lactobacilli present vaginally in the animals were not performed. However, qualitative analysis of observable morphology and smell suggested that the resident lactobacilli in each monkey remained constant over the entire sampling time period.
Other aerobic vaginal microflora counts were also determined for the same animals (Fig. (Fig.6).6). Surprisingly and unlike humans, rhesus monkeys in general do not harbor aerobic gram-negative rods vaginally. Besides lactobacilli, the next-most-dominant aerobic vaginal microflora organisms in the monkeys were coagulase-negative staphylococci; these organisms remained relatively constant over the 6-month test period. There was no significant difference between the counts of coagulase-negative staphylococci of the GML-treated and vehicle control-treated groups (P values of >0.05 at all time points). The remaining aerobic vaginal microflora consisted of sporadically present S. aureus, beta-hemolytic streptococci, and yeasts. These organisms appeared and disappeared vaginally for unknown reasons. When S. aureus appeared vaginally in a number of monkeys, we hypothesized that this resulted from clonal bacterial spread through monkeys. This was demonstrated not to be the case, however, since the superantigen production profile (as a measure of clonal relatedness) indicated that multiple S. aureus clones were present vaginally among the monkeys (though individual monkeys grew only single S. aureus clones).
Extensive colposcopy studies performed on the animals did not reveal evidence of GML-induced vaginal inflammation or mucosal lesions in cervical vaginal tissues. There was also no evidence from blind evaluations by a pathologist (Stefan E. Pambuccian) of biopsy specimens that chronic GML use caused histopathological changes in cervical vaginal tissue (Table (Table11).
GML is a fatty acid monoester that has been studied for many years, particularly as related to its ability to inhibit growth of gram-positive cocci, such as streptococci, and exotoxin production by S. aureus at concentrations below 100 μg/ml (12-14, 24, 28, 35). These effects result from inhibition of signal transduction in affected bacteria (24, 35). GML also stabilizes eukaryotic membranes, preventing cytokine production by various cell types and reducing the toxic effect of hypotonic solutions and a variety of exotoxins (21). Similar to GML effects on prokaryotic cells, the effects on eukaryotic cells appear to be membrane insertion and stabilization, thereby blocking signal transduction. In vitro studies further suggest that GML reduces infectious titers of enveloped viruses (26, 33, 34; our unpublished observations with inhibition of HIV-1 infection of human T cells).
Because of the effects of GML on exotoxin production by S. aureus, studies have suggested the compound may be useful as a tampon additive to reduce the risk of TSS, an illness associated with use of certain tampons. In vivo studies with healthy volunteer women showed that GML as a tampon additive reduced production of staphylococcal exotoxins vaginally and at the same time reduced the vaginal IL-8 production which was at least partly in response to the presence of S. aureus (20a).
Because of the importance of immune activation and inflammation in HIV-1 transmission and pathogenesis (9, 15, 22), the inhibitory effects of GML on cytokine production, and our in vitro studies showing that GML (100 μg/ml) prevents HIV-1 infection of human T cells (unpublished observation), we hypothesized that GML might be useful in preventing transmission of HIV-1 and have initiated studies with rhesus macaque monkeys both to assess the ability of GML to interfere with simian immunodeficiency virus transmission and its safety for chronic use on vaginal mucosal surfaces. In the safety studies reported here, we first showed that GML (whether nonformulated or formulated) in vitro does not exert adverse effects on normal vaginal flora lactobacilli; we also observed that GML does not prevent acid production and GML blocks loss of viability of the organism in stationary phase. Lactobacilli are considered responsible for maintenance of acidic pH in the vagina at times other than menstruation (1, 3), as seen both for monkeys studied in this research and for humans. In addition to producing lactic acid, lactobacilli also produce hydrogen peroxide, another antimicrobial compound that likely helps maintain the organism as the dominant normal flora organism vaginally (1, 11). We did not test the effect of GML on hydrogen peroxide production, but we believe it is unlikely that GML will selectively inhibit its production.
Next, we showed that lactobacilli do not lead to HVEC chemokine production, as assessed by measuring IL-8 production. In contrast, the potential pathogen S. aureus leads to significant chemokine production, which can be blocked by GML. These data suggest that vaginal GML maintenance of lactobacilli, while at the same time reducing the effects of potential pathogens on the mucosa, is likely to lead to a reduction in inflammation. Prior studies suggest that inflammatory spermicides and vaginal infections may facilitate HIV-1 transmission (7, 9, 15, 22).
To our knowledge, these are the first studies also to quantify monkey vaginal aerobic microflora. Prior studies have demonstrated the presence of the same microbes vaginally as observed in the present study, but quantification was not performed (18, 19). We showed that both lactobacilli and coagulase-negative staphylococci dominate the vaginal aerobic flora, consistent with that seen in humans at times other than menstruation. Although we did not specify the lactobacilli present vaginally in the animals, our qualitative studies (through colony morphology and smell) suggested that resident lactobacilli were maintained across the entire 6-month test period. Also, similar to the case for humans, beta-hemolytic streptococci (though not group B streptococci), S. aureus, and yeasts are sporadically present. It is unclear why these organisms appear and disappear vaginally at various times. However, this appears to be unrelated to the use of GML, as the same thing happened in vehicle control-treated animals. At one point, we hypothesized that S. aureus strains may be clonally spreading through the monkey population, since the strains often appeared in multiple animals at the same time; this was shown not to be the case. A single clonal type of S. aureus colonized individual animals, but this organism did not appear to be shared among the other animals.
One final interesting observation was made: monkeys do not appear to have gram-negative rods vaginally. We had previously noted that animal species that are highly susceptible to the development of TSS have a predominant amount of gram-negative bacteria (Escherichia coli) both vaginally and intestinally (32). Studies have attempted to establish TSS in rhesus macaque monkeys without success (23). Although the reason is not completely clear, we suggest that the millionfold synergy between superantigens that cause TSS and gram-negative LPS may be required for TSS production (27, 29, 32).
In sum, these studies have demonstrated that GML as applied topically to mucosal surfaces is both safe and noninflammatory. Furthermore, the compound appears to promote the survival of lactobacilli, and this may promote vaginal health.
This work was supported by USPHS research grant AI071976 from the National Institute of Allergy and Infectious Diseases, a grant from the University of Minnesota Academic Health Center, and research grant P51 RR000167 from the National Center for Research Resources, a component of the National Institutes of Health, to the Wisconsin National Primate Research Center, University of Wisconsin—Madison.
Published ahead of print on 6 October 2008.