• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

Defenses against infections provided by vaginal lactobacilli: Lactic acid versus hydrogen peroxide

Dissertation
Author: Deirdre Elizabeth O'Hanlon
Abstract:
Women with "normal" lactobacilli-dominated vaginal microflora are at significantly reduced risk of a broad range of reproductive tract infections. Unfortunately, most women do not have lactobacilli-dominated vaginal microflora. Lactic acid produced by lactobacilli was once thought to protect against infections, but lactic acid production by lactobacilli in vitro is inadequate for broad-spectrum microbicidal activity against reproductive tract pathogens. Hydrogen peroxide (H2 O2 ) production by certain lactobacilli species is now generally believed to be the primary mechanism of protection, and H2 O2 -producing lactobacilli have been shown in vitro to inactivate several reproductive tract pathogens. However, in vitro aerobic production of lactic acid and H2 O 2 by lactobacilli may not represent the concentrations produced in vivo . Neither lactic acid nor H2 O2 concentrations in cervicovaginal fluid (CVF) have been measured under the hypoxic conditions of the vagina. Additionally, effects of lactic acid and H2 O 2 on vaginal lactobacilli have not previously been considered. Thus, my first aim in this research was to measure the concentrations of H2 O2 and lactic acid in ex vivo (fresh, minimally-diluted, hypoxically-maintained) samples of CVF; I found that H2 O2 was undetectably low. Even with extended aerobic exposure, H2 O2 concentration averaged only 23µM, about 100-times lower than that produced in vitro , due to the high antioxidant activity of CVF. In contrast, the lactic acid concentration in ex vivo CVF averaged 0.97%, about 10-times higher than that produced in vitro . Moreover, the average acidity of ex vivo CVF was pH 3.74, significantly more acidic than reported by numerous prior studies that failed to account for the effects of aerobic exposure. I then measured the microbicidal activities of H2 O2 and lactic acid at physiological concentrations and under conditions consistent with those in vivo against a broad range of reproductive tract pathogens and vaginal lactobacilli. H2 O2 had no detected microbicidal activity under these conditions. In contrast, lactic acid was microbicidal against all but one reproductive tract pathogens tested, and did not harm lactobacilli. I conclude (1) it is improbable that H2 O2 production in vivo by vaginal lactobacilli provides protection against infections, and (2) lactic acid is a broad-spectrum microbicide that likely provides significant protection in vivo .

TABLE OF CONTENTS Abstract of the dissertation ii Acknowledgments iv Table of contents ....vii List of tables xiv List of figures xv Chapter I: The vaginal microflora and infections of the female reproductive tract 1 LA. Abstract... 1 LB. Background 3 I.B.i. The vaginal microflora 3 I.B.ii. Vaginal lactobacilli 6 LB. iii Bacterial vaginosis 9 I.B.iv. Protective mechanisms of vaginal lactobacilli 15 I.C. Significance of the project 18 I.D. Structure of the dissertation 19 vii

I.E. Tables for Chapter I 20 I.F. Figures for Chapter I 24 Chapter II: Hydrogen peroxide concentration and blocking hydrogen peroxide activity by cervicovaginal fluid and semen 55 II.A. Abstract 55 II.B. Background 57 II.B.i. Production of hydrogen peroxide by vaginal lactobacilli as a mechanism of protection against reproductive tract infection 57 II.B.ii. Hydrogen peroxide in vitro and in vivo 60 II.B.hi. The aim of the experiments 62 II.C. Materials and Methods 63 Il.C.i. Participants 63 Il.C.ii.. Collection of cervicovaginal fluid samples 64 Il.C.iii. Microscopy of cervicovaginal fluid and semen samples 66 Il.C.iv. Measurement of hydrogen peroxide 69 II.C.v. Measuring the hydrogen peroxide concentration in CVF under hypoxic and aerobic conditions 71 viii

Il.C.vi. Measuring the blocking of exogenous hydrogen peroxide by cervicovaginal fluid or semen 72 Il.C.vii. Measuring hydrogen peroxide concentration in the culture supernatants of primary isolates and ATCC strains of vaginal lactobacilli species under anaerobic^ aerobic, and vigorously aerated conditions 72 ILC.viii. Statistical analysis 73 II.D. Results 74 II.D.i. Hydrogen peroxide concentration in cervicovaginal fluid under hypoxic and aerobic conditions 74 II.D.ii. Blocking of exogenous hydrogen peroxide by CVF and semen 75 II.D.iii. Hydrogen peroxide concentration in culture supernatants of ATCC and primary strains of vaginal lactobacilli under hypoxic, aerobic, and vigorously aerated conditions ,75 II.E. Discussion 77 II.F. Conclusion 78 II.G. Figures for Chapter II 80 Chapter III: Acidity, lactic acid, and acetic acid in cervicovaginal fluid 92 III.A. Abstract 92 IX

III.B. Background 94 III.B.i. Production of lactic acid by vaginal lactobacilli as a mechanism of protection against reproductive tract infection 94 III.B.ii. Lactic acid in vitro versus in vivo 96 III.B.iii. The aim of the experiments 98 III.C. Materials and Methods 100 IILC.i. Participants 100 III.C.ii. Collection of cervicovaginal fluid samples 100 III.C.iii. Measurement of pH 101 III.C.iv. Measurement of lactic acid and acetic acid concentrations in cervicovaginal fluid 102 III.C.v. Microscopy of cervicovaginal fluid samples 105 Ill.C.vi. Measurement of the effect of carbon dioxide tension on pH of CVF 106 IILC.vii. Measuring pH, lactic acid and acetic acid concentrations in culture supernatants of primary isolates of vaginal lactobacilli 106 Ill.C.viii. Measuring the effect of pH relief on lactic acid concentration produced in culture 107 x

IILC.ix. Statistical analysis 107 III.D. Results 109 III.D.i. The pH of cervicovaginal fluid 109 III.D.ii. The lactic acid and acetic acid concentrations in cervicovaginal fluid 110 III.D.hi. The pH, lactic acid and acetic acid concentrations in culture supernatants of primary isolates of vaginal lactobacilli 110 III.D.iv. The effect of pH relief on in vitro lactic acid production 111 III.E. Discussion 113 III.F. Conclusion 115 III.G. Tables for Chapter III 116 III.H. Figures for Chapter III 118 Chapter IV: The microbicidal activity of hydrogen peroxide and lactic acid against BV- associated bacteria, sexually-transmitted pathogens, and vaginal lactobacilli 128 IV.A. Abstract 128 IV.B. Background 130 IV.B.i. Vaginal lactobacilli provide broad-spectrum protection 130 xi

IV.B.ii. Features of the protection provided by vaginal lactobacilli 131 IV.B.iii. The aim of the experiments 132 IV.C. Materials and Methods 134 IV.C.i. Lactobacilli 134 IV.C.ii. Bacterial vaginosis-associated bacteria 134 IV.C.iii. Sexually transmitted infection pathogens 135 IV.C.iv. Microbicidal activity 136 IV.C.v. Bacterially depleted cervicovaginal fluid 137 IV.C.v. Statistical analysis 140 IV.D. Results 141 IV.D.i. Microbicidal activity of hydrogen peroxide 141 IV.D.ii. Microbicidal activity of acidity, lactic acid, and acetic acid 142 IV.D.iv. Effect of cervicovaginal fluid on pathogen-inactivation by H202-producing lactobacilli in buffer 145 IV.E. Discussion 146 IV.F. Conclusion 148 xii

IV.G. Tables for Chapter IV 149 IV.H. Figures for Chapter IV 150 Chapter V 169 V.A. Summary of the findings 169 V.B. Directions for future research 170 V.B.i. Hydrogen peroxide-producing lactobacilli and bacterial vaginosis 170 V.C. Some final considerations 172 V.C.i. Magnitude of BV-associated risks 173 V.C.hi. Microfloral instability 173 V.C.iv. Acid-dependent metabolism of glycogen 174 V.C.v. Relative growth rates 174 V.C.vi. Probiotics 175 V.D. Figures for Chapter V 177 V.E. References for Chapter V 180 xiii

LIST OF TABLES Table 1: Summary of nineteen studies investigating lactobacillus species isolated from women with lactobacilli-dominated vaginal microflora 20 Table 2: pH, lactic acid and acetic acid concentration measurements made in ex vivo CVF samples from women with healthy, lactobacilli-dominated cervicovaginal microflora 116 Table 3: Effect of 1% bacterially-depleted CVF on the microbicidal activity of H2O2 and lactic acid, against six core species of BV-associated bacteria, and three sexually transmitted pathogens 149 xiv

LIST OF FIGURES Figure 1.1: Micrographs of swab samples from lactobacilli-dominated vaginal microflora (left) and bacterial vaginosis (right) 24 Figure 1.2: Relative frequency of 'normal' vaginal microflora, intermediate vaginal microflora, and bacterial vaginosis in three populations 25 Figure 1.3: Odds ratio of incident infection/adverse event in women diagnosed with BV at study entry, compared to women without BV 26 Figure 2.1: Endogenous H2O2 concentration in CVF in air, CVF in a hypoxic glove-box, and bacterially depleted CVF in air versus time 80 Figure 2.2: Endogenous H2O2 concentration in CVF samples measured at 0, 1, 15 or 240 minutes exposure to air following four hours hypoxic incubation (open circles), or immediately after sample collection entailing ~1 minute exposure to air (closed circles). 81 Figure 2.3: H2O2 concentration detected by Amplex Red® assay in CVF or semen samples, versus the calculated final concentration of exogenous H2O2 added to the samples before assaying 82 Figure 3.1: The correlation of pH and lactic acid concentration in ex vivo CVF samples from women with healthy, lactobacilli-dominated vaginal microflora 118 xv

Figure 3.2: Comparison of the pH's (upper graph) and lactic acid concentrations (lower graph) of ex vivo CVF samples and in vitro hypoxically cultured lactobacilli isolated from the same CVF samples 119 Figure 3.3: Correlation of pH's (upper graph) and lactic acid concentrations (lower graph) of ex vivo CVF samples and in vitro hypoxically cultured lactobacilli isolated from the same CVF samples 122 Figure 4.1.a: The microbicidal activity of H2O2 at pH 7, against four species of vaginal lactobacilli (solid lines) and eighteen species of BV-associated bacteria (dotted lines). 150 Figure 4.1.b: The microbicidal activity of H2O2 with 50mU/mL MPO at pH 7, against four species of vaginal lactobacilli (solid lines) and eighteen species of BV-associated bacteria (dotted lines) 151 Figure 4.1.c: The microbicidal activity of H2O2 at pH 4.5, against three species of vaginal lactobacilli and six species of BV-associated bacteria 152 Figure 4.1.d: The microbicidal activity of H2O2 with 50mU/mL MPO at pH 4.5, against three species of vaginal lactobacilli and six species of BV-associated bacteria 153 Figure 4.2.a: The microbicidal activity of H2O2 alone (dotted lines) or H2O2 with 50mU/mL MPO (solid lines) against HSV-2 154 Figure 4.2.b: The microbicidal activity of H2O2 alone (dotted lines) or H2O2 with 50mU/mL MPO (solid lines) against Trichomonas vaginalis 155 xvi

Figure 4.2.c: The microbicidal activity of H2O2 alone (dotted lines) or H2O2 with 50mU/mL MPO (solid lines) against Haemophilus ducreyi 156 Figure 4.2.d: The microbicidal activity of H2O2 alone (dotted lines) or H2O2 with 50mU/mL MPO (solid lines) against Neisseria gonorrhoeae 157 Figure 4.3.a: The microbicidal activity of lactic acid against four species of vaginal lactobacilli (solid lines) and eighteen species of BV-associated bacteria (dotted lines). 158 Figure 4.3.b: The microbicidal activity of lactic acid at pH 7 against three species of vaginal lactobacilli and species of BV-associated bacteria 159 Figure 4.3.c: The microbicidal activity of acetic acid against four species of vaginal lactobacilli (solid lines) and eighteen species of BV-associated bacteria (dotted lines). 160 Figure 4.4.a: The microbicidal activity of lactic acid (solid line) and acetic acid (dotted line) against HSV-2 161 Figure 4.4.b: The microbicidal activity of lactic acid (solid line) and acetic acid (dotted line) against Trichomonas vaginalis 162 Figure 4.4.c: The microbicidal activity of lactic acid (solid line) and acetic acid (dotted line) against Haemophilus ducreyi 163 Figure 4.4.d: The microbicidal activity of lactic acid (solid line) and acetic acid (dotted line) against Neisseria gonorrhoeae 164 Figure 14: Comparison of lactic acid and H2O2 concentrations of aerobic in vitro cultures xvii

of vaginal lactobacilli and hypoxic ex vivo CVF samples 177 Figure 15: The microbicidal activity of putrescine against Lactobacillus iners and Lactobacillus crispatus 178 Figure 16: The percentage decrease in glycogen in a culture of L. crispatus after 24 hours incubation at 37°C, versus the pH at which the culture medium was maintained. 179 xviii

CHAPTER I: THE VAGINAL MICROFLORA AND INFECTIONS OF THE FEMALE REPRODUCTIVE TRACT I.A. Abstract Huge numbers of microbes, primarily bacteria, live on and in the healthy human body: ten times as many bacterial cells as human cells. Generally, the microflora of the mucosal surfaces are dense and diverse: the gut accommodates l'O10 bacterial cells per milliliter of luminal content, and hundreds of different species coexist in close proximity [1, 2, 3]. During the reproductive years, however, the healthy microflora of the human vagina is quite different, consisting primarily of a relatively sparse monoculture of only one or two species of lactobacilli (105 to 107 bacterial cells per milliliter of luminal content [4]). For over a century, the composition of the vaginal microflora has been recognized as an important element in the health of the female reproductive tract; women with lactobacilli-dominated vaginal microflora are significantly less likely to suffer from reproductive tract disease than women with other kinds of vaginal microflora. Bacterial vaginosis (BV) is a common, transient, but often recurrent disturbance of the vaginal microflora in which the sparse monoculture of lactobacilli is replaced by a dense, polymicrobial mixture of primarily Gram-variable and Gram-negative bacterial species. BV does not appear to be caused by a single bacterial species; Gardnerella vaginalis, Mycoplasma species, Mobiluncus species, and Prevotella species are typically present in BV, together with many other species of Gram-negative anaerobic bacteria. l

The etiology of BV is not clearly understood, but risk factors for BV include menstruation, sexual intercourse, number of sexual partners, use of vaginal lubricants and cleansers, smoking, and Black race. Women with BV are at markedly increased risk of many different reproductive tract diseases, including viral and bacterial sexually- transmitted infections, and complications of conception, pregnancy, childbirth, and abdominal surgery. BV-associated bacteria produce several possible virulence factors, including mucinases and succinic acid, which might contribute to these risks. However, the remarkable number and diversity of reproductive tract diseases associated with BV also suggest the alternative interpretation, namely that vaginal lactobacilli provide broad- spectrum protection against reproductive tract pathogens. Several possible mechanisms by which vaginal lactobacilli could provide this broad-spectrum protection against reproductive tract pathogens have been suggested, including production of lactic acid and H2O2. The production of H2O2 by vaginal lactobacilli as a mechanism of broad-spectrum protection has been consistently and widely emphasized in the literature, supported by data from epidemiological studies and in vitro experiments. Hydrogen peroxide-producing lactobacilli are considered by many to be an essential part of a protective vaginal microflora, and H202-production is probably the most widely used criteria in selecting lactobacilli strains for probiotic vaginal use. 2

I.B. Background I.B.i. The vaginal microflora I.B.i.a. Childbed fever In the mid eighteen-hundreds, childbed fever (puerperal sepsis) killed at least fifteen percent of women giving birth in Europe and America with professional attendance [5], and death rates spiked to one hundred percent during hospital ward epidemics. In 1879, Louis Pasteur published his observations of streptococci in the blood of women with puerperal fever [6]; he conjectured that the mechanical trauma caused by childbirth allowed bacteria already established in the vagina to invade the bloodstream. Thirteen years later, in 1892, Albert Doderlein published his finding that laboring women whose vaginal bacteria consisted solely of Gram-positive, lactic acid-producing bacilli were less likely to develop puerperal sepsis after delivery than were laboring women whose vaginal bacteria consisted partially or completely of other forms [7]. Doderlein also noted that although cervicovaginal fluid (CVF) from all the women he examined showed some degree of acidity, CVF from women with only Gram-positive bacilli was significantly more acidic than CVF from women with other kinds of vaginal bacteria. Doderlein attributed this greater acidity to the lactic acid-producing activity of the Gram- positive bacilli, and conjectured that the bacilli were capable of rendering the vagina inhospitable to all but the most acidophilic microbes, thereby preventing the establishment of the streptococci responsible for puerperal sepsis. Who knew what other pathogenic microbes the beneficial bacilli might be fending off, Doderlein speculated. 3

I.B.i.b. Estrogen, glycogen, and vaginal lactobacilli The Gram-positive lactic acid-producing bacilli Doderlein had observed were quickly identified as lactobacilli, already well-known and well-studied in the context of food fermentation processes (though lactobacilli in the vaginal microflora continued to be called 'Doderlein's bacilli' for many decades, to honor the original observer). The presence of vaginal lactobacilli was found to correlate with the presence of glycogen in the vaginal epithelial cells [8], though in vitro attempts to demonstrate direct fermentation of glycogen by vaginal lactobacilli were inconclusive. Some studies reported that lactobacilli could not ferment glycogen [9, 10], while others reported that lactobacilli could ferment glycogen, though in vitro activity seemed to depend upon the age and density of the lactobacilli culture [11]. The accumulation of glycogen in vaginal epithelial cells was found to depend upon the presence of circulating estrogen [12], produced by the ovaries. Pre-pubescent girls [13], like post-menopausal women [14,15], have much lower levels of estrogen, epithelial glycogen, and vaginal lactobacilli. Neonate girls, transiently circulating maternal estrogen, transiently acquire both the epithelial glycogen and vaginal lactobacilli of reproductively mature women [16], as do pre- pubescent girls treated with exogenous estrogen [17], and post-menopausal women using estrogen-replacement therapies [18]. In the course of a normal menstrual cycle, the concentration of circulating estrogen increases during the proliferative phase and peaks shortly before ovulation; a number of studies have reported a corresponding increase and peak in epithelial glycogen and vaginal lactobacilli [19, 20, 21]. LB.i.e. Vaginal microflora and cervicovaginalfluid Other investigators confirmed and expanded Doderlein's original observations. 4

Women with vaginal lactobacilli were found to be less likely to have symptoms of vaginitis, cervicitis [22], or infection by streptococci than women with other kinds of vaginal microflora. By the early 1920's, standardized systems for assessing the vaginal microflora were in use, and the connection between the composition of the vaginal microflora and the health of the reproductive tract was made explicit in the terminology of Schroder's Reinheitsgrad (Vaginal cleanliness grade') system [23]. Grade I consisted of lactobacilli only, Grade II consisted of lactobacilli mixed with other kinds of bacteria, and Grade III consisted of non-lactobacilli bacteria only. Schroder and others found a dramatic correlation between microflora grade and the acidity, consistency, and color of CVF: Grade I microflora was associated with a highly acidic, viscous, clear or white fluid, while Grade III microflora was associated with a less acidic, watery, yellow or grayish fluid, often with a putrid odor [24, 25, 26]. Other grading systems have elaborated on Schroder's three basic groups, for example subdividing Schroder's Grade I to distinguish among different morphotypes of lactobacilli (primarily between the large, evenly shaped rods typical of Lactobacillus crispatus and the smaller, more irregular rods typical of L. jensenii, L. gasseri, and L. iners), and subdividing Grades II and III to distinguish between non-lactobacilli microflora composed of just one or two types of bacteria, and non-lactobacilli microflora composed of many different kinds of bacteria [27, 28]. A Grade 0 has also been added to some current grading systems to describe the complete or near-complete absence of vaginal bacteria [29], as happens following the use of some antibiotic therapies. 5

I.B.ii. Vaginal Lactobacilli LB.ii.a Frequency and distribution of vaginal lactobacilli species Vaginal lactobacilli were initially identified as Lactobacillus acidophilus, a species already well-studied in milk fermentation. However, subsequent bacteriological and biochemical studies (the comparison of species' macro and micro morphology, growth habits, response to growth conditions, utilization of different fermentation substrates, amount and ratio of different fermentation products produced) showed that other lactobacillus species were more common as vaginal microflora. Molecular studies (generally PCR of 16S rDNA, followed by either Southern blot using probes prepared from reference lactobacilli strains, or by sequence-analysis and comparison with databases of reference lactobacillus sequences) have identified many more vaginal lactobacillus species. Over twenty species of lactobacilli are now known to occur vaginally, including L. brevis, L. buchneri, L. casei, L. cellobiosus, L. crispatus, L. curvatus, L. delbrueckii, L. fermentum, L. gallinarum, L. gasseri, L. helveticus, L. iners, L. jensenii, L. johnsonii, L. mucosae, L oris, L. paracasei, L. plantarum, L.reuteri, L. rhamnosus, L. ruminis, L. salivarius, L. suntoryeus, and L. vaginalis. Table 1 summarizes the locations, methods, and principle findings of nineteen cross-sectional studies analyzing the species-composition of lactobacilli-dominated vaginal microflora The studies summarized in Table 1 fall into two groups: those that use an in vitro culture selection-step on MRS or Rogosa selective medium to isolate vaginal lactobacilli from collected CVF, and those that use a total DNA extraction from collected CVF without attempting to isolate lactobacilli first. Studies that use an in vitro culture 6

selection-step generally report L. crispatus as the most common vaginal lactobacilli species, with L. jensenii and L. gasseri also common; these findings hold true whether the post selection-step species identification is carried out by bacteriological, biochemical, or molecular methods. However, studies that do not use an in vitro culture selection-step generally report L. iners to be the most common vaginal lactobacilli species: L. iners does not grow on either MRS or Rogosa agar media [30]. (Studies that use a total DNA extraction technique without selection on MRS or Rogosa agar medium fail to identify L. iners only when the method of species identification is Southern blot, with L. iners not included in the reference lactobacilli strains used to prepare the probes for use against the amplified DNA.) As noted previously, more than twenty different vaginal lactobacilli species have been identified in studies around the world, but the homogeneity of the most commonly found species is quite striking: just three species, L. iners (where the study methodology permits its detection), L. crispatus and L. jensenii account for approximately 75% of the lactobacillus species isolated in almost all studies, though the relative frequencies of these three species do seem to vary among geographically [31, 32] and racially [33, 34] distinct populations. Intriguingly, a study using the chaperonin-60 gene [35] as the basis for PCR and sequence analysis (rather than 16S rDNA) found considerable strain variation within the one or two species that comprise an individual woman's microflora: on average, L. wer,y-dominated microflora consisted of twelve genetically distinct strains and L. crispatus-dominated microflora consisted of eight genetically distinct strains. I.B.ii.b. Species persistence of vaginal lactobacilli Only a few longitudinal studies of the species-composition of lactobacilli- 7

dominated vaginal microflora have been carried out, compared to a large number of cross-sectional studies. Studies that have analyzed species-composition over the course of several menstrual cycles have concluded that the species-composition of the lactobacilli- dominated vaginal microflora appears quite stable over time, though with some cyclic variations in the total number of bacteria [36, 37] and the proportion of lactobacilli [38, 39]. However, one study that analyzed the species-composition of vaginal microflora in terms of HbOi-producing and non E^Oi-producing lactobacilli over a period of eight months [40] found that about one-third of the women who maintained a lactobacilli- dominated vaginal microflora changed from having IrbCh-producing lactobacilli to having non H2C>2-producing lactobacilli, or vice-versa. This suggests that the species- composition of lactobacilli-dominated vaginal microflora does change over time. I.B.ii.c. Phage infection of vaginal lactobacilli Phage infection of vaginal lactobacilli may be common, even in women with lactobacilli-dominated vaginal microflora. One study found that approximately one-third of the vaginal lactobacilli isolated were infected with phage [41]. Fifty percent of these phage infections were minimally lysogenic, 30% were moderately lysogenic, and the remainder were highly lysogenic. One hundred and fifty two isolates of lactobacilli (representing at least five different species) yielded sixty-seven isolates of phage that fell into four morphologically distinct groups. L. crispatus isolates were less frequently infected with phage than L.jensenii isolates (18% of isolates compared to 43% of isolates, respectively). The host ranges of the phages are strikingly wide: only 13% of the vaginal lactobacilli strains isolated were resistant to all four morphological groups of phage, while 75% of the vaginal lactobacilli strains isolated were susceptible to two or 8

more kinds of phages. I.B.iii Bacterial vaginosis LB.Hi.a. A polymicrobial infection In 1955, Charles Dukes and Herman Gardner published their finding that vaginal microflora dominated by a small Gram-variable bacillus, Haemophilus vaginalis, was associated with grayish-white, thin, copious CVF with a putrid odor, accompanied by no obvious itching, pain, redness, edema, or leucorrhea: a syndrome they referred to as 'Haemophilus vaginalis vaginosis' [42]. H vaginalis alone was found to be ineffective at initiating vaginosis in women with lactobacilli-dominated vaginal microflora: only one of eleven women vaginally inoculated with 1011 H. vaginalis bacterial cells developed symptoms of vaginosis [43]. However, eight of nine women vaginally inoculated with CVF from women with Haemophilus vaginalis vaginosis developed the symptoms, indicating that other factors in addition to H. vaginalis bacteria were necessary to establish infection. Further studies showed that the Gram-negative Mycoplasma [44], Bacteroides [45] and Mobiluncus species [46] were often present in association with H. vaginalis. 'Haemophilus vaginalis vaginosis' was renamed 'bacterial vaginosis' (BV) in recognition of the polymicrobial nature of the infection; H vaginaliswas renamed Gardnerella vaginalis. Bacteriological and biochemical studies show that Gram-negative obligate anaerobes also contribute to the BV microflora, including Prevotella species, Peptostreptococcus micros, and Anaerococcus tetradius [47, 48, 49]. Molecular analysis has revealed even greater diversity within BV microflora, bringing the number of bacterial species associated with BV to over thirty, to include Atopobium vaginae, 9

Leptotrichia amnionii, Megasphaera elsdenii, Megasphaera micronuciformis, Eggerthella hongkongensis, Porphyomonas assacharolytica, Sneathia sanguinegens, Gemella bergeriae, and Peptoniphilus lacrimalis [50, 51]. I.B.iii.b. Diagnosing bacterial vaginosis 'Amsel's criteria1 and 'Nugent scoring' are the primary methods of diagnosing BV infection. Amsel's criteria [52] stipulates that at least three of the following four signs be present for a diagnosis of BV: thin, copious, grayish-white CVF, decreased acidity of CVF (pH > 4.5), production of an amine odor (the 'whiff test') on addition of 10% KOH to a CVF sample; and characteristic 'clue cells':sloughed cells of the vaginal epithelium with abnormally ragged-looking cytoplasmic margins, and thickly coated with small bacilli evident on microscopic examination of CVF mixed with saline, rather than the clean, relatively intact epithelial cells and a sparse monoculture of lactobacilli characteristic of the normal vaginal microflora (Figure 1.1). Nugent scoring [53] is a standardized system for assessing the amount and variety of bacterial morphotypes seen on microscopic examination of a dried, fixed, and Gram- stained sample of CVF. Score-points are added for scarcity or absence of large Gram- positive rods {Lactobacillus morphotype), and for presence and abundance of small Gram-variable bacteria (G. vaginalis morphotype) and Gram-variable curved rods (Mobiluncus morphotype); Nugent's scores of 0 to 3 indicate a healthy, lactobacilli- dominated vaginal microflora, scores of 4 to 6 indicate 'intermediate' microflora similar to Schroder's Grade II, while scores of 7 to 10 indicate BV infection. Although Nugent scores of 0 to 3 are routinely referred to as 'normal', it is worth noting that only a 10

minority of women have lactobacilli-dominated vaginal microflora (Figure 2). Nugent scoring is a more sensitive diagnostic method than Amsel's criteria: approximately one- third of women with BV as assessed by Nugent's scoring have CVF that does not fulfill the requirements for appearance or odor described by Amsel's criteria [54]. A number of proprietary diagnostic tests detect the decrease in metabolic products of lactobacilli, or increase in metabolic products of BV-associated bacteria: the QuickVue Advance® pH and Amines test [55] detects decreased acidity and increased amines; the Osmetech Microbial Analyzer® test [56] detects increased acetic acid; the BVBlue® test [57] detects increased sialidase activity. I.B.iii.c. Prevalence of bacterial vaginosis Bacterial vaginosis is the most common female reproductive tract infection [58]: in the United States, approximately 30% of women between the ages of 14 and 49 are affected [59]; Black women are approximately three times as likely to have BV as non- Hispanic White women, and Asian women are only two-thirds as likely to have B V as non-Hispanic White women [60]. Rates of BV infection differ dramatically among study populations: 13% among sexually inactive adolescent girls [61], 25% among women who have sex with women [62], and 70% among women at high risk of sexually transmitted infections [63]. Rates of BV also differ geographically, with very high prevalence reported in parts of sub-Saharan Africa where HIV-1 infection rates are also very high [64, 65, 66]. Only a minority of women has the lactobacilli-dominated vaginal microflora that is referred to as 'normal'; the majority of women have intermediate microflora or bacterial vaginosis (Figure 1.2) 11

Full document contains 204 pages
Abstract: Women with "normal" lactobacilli-dominated vaginal microflora are at significantly reduced risk of a broad range of reproductive tract infections. Unfortunately, most women do not have lactobacilli-dominated vaginal microflora. Lactic acid produced by lactobacilli was once thought to protect against infections, but lactic acid production by lactobacilli in vitro is inadequate for broad-spectrum microbicidal activity against reproductive tract pathogens. Hydrogen peroxide (H2 O2 ) production by certain lactobacilli species is now generally believed to be the primary mechanism of protection, and H2 O2 -producing lactobacilli have been shown in vitro to inactivate several reproductive tract pathogens. However, in vitro aerobic production of lactic acid and H2 O 2 by lactobacilli may not represent the concentrations produced in vivo . Neither lactic acid nor H2 O2 concentrations in cervicovaginal fluid (CVF) have been measured under the hypoxic conditions of the vagina. Additionally, effects of lactic acid and H2 O 2 on vaginal lactobacilli have not previously been considered. Thus, my first aim in this research was to measure the concentrations of H2 O2 and lactic acid in ex vivo (fresh, minimally-diluted, hypoxically-maintained) samples of CVF; I found that H2 O2 was undetectably low. Even with extended aerobic exposure, H2 O2 concentration averaged only 23µM, about 100-times lower than that produced in vitro , due to the high antioxidant activity of CVF. In contrast, the lactic acid concentration in ex vivo CVF averaged 0.97%, about 10-times higher than that produced in vitro . Moreover, the average acidity of ex vivo CVF was pH 3.74, significantly more acidic than reported by numerous prior studies that failed to account for the effects of aerobic exposure. I then measured the microbicidal activities of H2 O2 and lactic acid at physiological concentrations and under conditions consistent with those in vivo against a broad range of reproductive tract pathogens and vaginal lactobacilli. H2 O2 had no detected microbicidal activity under these conditions. In contrast, lactic acid was microbicidal against all but one reproductive tract pathogens tested, and did not harm lactobacilli. I conclude (1) it is improbable that H2 O2 production in vivo by vaginal lactobacilli provides protection against infections, and (2) lactic acid is a broad-spectrum microbicide that likely provides significant protection in vivo .