Detection of Atopobium vaginae in Postmenopausal Women by Cultivation-Independent Methods Warrants Further Investigation

Abstract

We sequenced 16S rRNA genes from the vaginal swab contents of a postmenopausal woman with asymptomatic bacterial vaginosis (BV). Sequences from Atopobium vaginae were the most commonly detected. In a survey of 35 other postmenopausal women, this organism was detected in 44% with BV but not in any subjects deemed healthy.

CASE REPORT

A 48-year-old postmenopausal (as determined by a follicle-stimulating hormone assay) female with no history of urinary tract infection was studied. The subject was selected on the basis of microscopic analysis of vaginal swabs smeared on microscope slides, Gram-stained, and scored by the Nugent method (16). The subject did not report symptoms, though the Nugent score of 9 indicated the presence of bacterial vaginosis (BV) (no lactobacilli morphotypes present; gram-negative or -variable rods and clue cells observed). Yeast was not observed microscopically but was detected in the sample by PCR (5). Total DNA was obtained from the vaginal swab contents as previously described (2).

Bacterial 16S rRNA genes containing hypervariable regions suitable for the identification of bacteria were amplified by using PCR primers based within conserved regions of the genes. The majority of the 16S rRNA gene was amplified by PCR using eubacterial primers pA and pH* (8). The PCR products were cloned using pGEM-T according to the instructions of the manufacturer (Promega, Madison, Wis.). Investigation of the partial 16S rRNA sequences (n = 65 clones) was conducted using BLAST analysis. The majority (61%) of the sequences from the clones had homology to Atopobium vaginae (39 sequences had 98 to 99% homology to the sequence at GenBank accession no. AF325325 [800 bp typically searched] and 1 sequence had 98% homology to the sequence at AY269023). Also, sequences that had homology to sequences of Slackia (AF101241), Atopobium (AB015945), Eggerthella (AY321961), Megasphaera (AY271940, AY271947, and AY278622), Dialister (AY162469), Prevotella (AF385519 and L16476) and Lactobacillus (Y116329) species and other sequences from non-culture-grown microorganisms (AB100481 and AB034086) were detected less frequently.

Discussion.

Postmenopausal women are more susceptible to BV, a condition that can lead to other adverse health events such as urinary tract infections. Many postmenopausal women harbor a microbial consortium consistent with BV but without the induction of symptoms (3). Indeed, many potential pathogens (such as Gardnerella vaginalis and Mobiluncus) in the vaginal tract can exist as commensal inhabitants. It is thought that a shift to a symptomatic BV state may simply be due to a decline in the levels of “beneficial” lactic acid and hydrogen peroxide-producing lactobacilli and/or an increase in the levels of gram-negative anaerobes.

General studies of the vaginal bacterial microbiotas to date have largely been conducted using bacteriological culture-based techniques. However, certain microorganisms from the vaginal tract are difficult to cultivate (2, 3, 13). The aim of the study was to characterize the vaginal microbiotas of a case subject with BV by non-culture-based methods and to compare the results to observations from routine microscopic analysis. Further investigation regarding the presence of A. vaginae in other subjects was also carried out after the organism was found to be predominant in the case subject.

Specific PCR primers were also designed for the most commonly occurring cloned bacterial sequence (that of A. vaginae) to determine the frequency of these sequence types among other postmenopausal subjects. Oligonucleotide probe Atopobium 291 was used as the forward primer in this study (10). A reverse primer specific for A. vaginae (5′-CTC CTG ACC TAA CAG ACC-3′ [positions 568 to 586]) was designed. All PCR products (295 bp) were sequenced to confirm identity. Primers were tested on DNA from 18 urogenital tract isolates and other microorganisms (3), including A. vaginae (ATCC BAA-55). Other than that from A. vaginae, only DNA from Bacteroides vulgatus (DSM 3289) amplified a product. The sequence of this product was in excess of 100 bp larger than the target sequence in size and could be easily differentiated in a 2% agarose gel.

Vaginal DNA samples from two groups of subjects were tested for the presence of A. vaginae by specific PCR. The first group consisted of 20 postmenopausal subjects who were known to have a high incidence of BV (as determined by a previous investigation [3] in which 8 subjects were graded as having BV, 6 subjects were graded as having intermediate-grade bacterial colonization, and 6 were healthy. This group (n = 21, including the case subject) had five Atopobium-positive subjects, all with Nugent scores indicative of the presence of BV. Of the five positive-testing subjects, four had had a G. vaginalis sequence previously detected (3). The second group of subjects (namely, 15 postmenopausal subjects on hormone replacement therapy [HRT]) were recruited for this study as it was postulated that due to therapy, they would have a lower incidence of BV (21). Postmenopausal subjects on HRT therapy (n = 15) ranged in age from 41 to 64 years (mean, 53.26 years) and had been receiving HRT for 3 months to 15 years (average, 4.6 years). None of the subjects had symptoms of yeast vaginitis or BV. Each gave informed consent. By Nugent analysis, 1 subject was graded as having BV, 4 were graded as having intermediate-grade colonization, and 10 had normal Nugent scores. Analysis of the data showed that the Nugent scores were significantly different (P = 0.0277) between groups. The HRT group had only one A. vaginae-positive subject (according to the Nugent score), whose Nugent score indicated the presence of an intermediate-grade bacterial colonization. The Nugent scores of the subjects with A. vaginae detected were significantly higher than those of the subjects without A. vaginae detected (Table 1).

TABLE 1.

Detection of A. vaginae in vaginal samples compared to Nugent score

Nugent scorea	Total no. of subjectsb	No. of subjects whose samples werec:
		Atopobium negative	Atopobium positive
Normal (0-4)	16	16	0
Intermediate (5-7)	10	9	1
BV (8-10)	10	5	5

^aNugent data (except those from the case study subject) for the postmenopausal group of women are from a previous study (3).

^bIncludes case subject.

^cNugent values for A. vaginae-positive subjects were 9 (case subject), 5 (HRT subject), and 7 to 8, 8, 9, and 10 (postmenopausal subjects). The mean (± standard deviation) Nugent score (nearest whole number) for the Atopobium-negative subjects was 3 (±2.78); the mean Nugent score for the Atopobium-positive subjects was 8 (±1.744) (P = 0.0024 [Mann-Whitney nonparametric two-tailed test]).

Nugent scoring relies upon the Gram stain characteristics of vaginal swab contents as smears on microscope slides. The presence of Atopobium (which is a member of the Coriobacteriaceae family) in vaginal samples has the potential to add confusion to the Nugent scoring system, as the genus contains organisms formerly classified as Lactobacillus species (14, 17) and could be interpreted as normal. In addition, some related species are thought to readily decolorize and may appear to be Gram negative or variable (6). In terms of culture-based diagnosis, Atopobium are generally slow growing, nonreactive in conventional biochemical tests (6), and may not stand out as being particularly deleterious to a host during a polymicrobial infection. While the A. vaginae type strain was isolated from the vagina of a “healthy” individual (22), there are reports of A. vaginae causing an ovarian abscess (9) and other isolates are listed in the culture collection at the University of Göteborg, Sweden, that originated from blood and amniotic membrane infections. Atopobium species and other members of the related Coriobacterium group have been isolated from various oral infections (6, 15, 17-19). In summary, there is support for the idea of the ability of these organisms to be pathogenic to the host.

Interestingly, the major metabolic products of Atopobium include lactic acid (7, 10, 22) (viewed as a component of the physiochemical vaginal environment favorable for the prevention of the growth of undesirable microorganisms). In addition to producing lactic acid, however, some species are also known to have dipeptidyl peptidase activity and produce significant amounts of ammonium in other environments (7), the latter potentially acting as a substrate for BV microorganisms, including G. vaginalis (20). While the commensal relationship between the ammonium-producing Prevotella bivia and the utilization of ammonium by G. vaginalis has been demonstrated previously (20), it is also possible that other associations exist. However, it remains to be determined whether Atopobium species, Atopobium-related species, and other species detected here are involved directly in BV or simply react opportunistically when vaginal conditions change.

Atopobium species have been detected in habitats in which it is thought that protein is generally more available than sugars as a source of energy (7, 11, 12). This may also be the case in postmenopausal women, as there is a decrease in the available fermentable carbohydrate glycogen from vaginal epithelial cells during the onset of menopause (1, 4), which may result in a shift among the vaginal microbiotas to become more reliant on energy from other metabolic sources. HRT subjects had a lower incidence of detected A. vaginae. It has been demonstrated that the use of vaginal estriol can lead to a recovery of the normal lactobacilli flora and a reduced risk of infection (21); this may explain their lower rate of detection here.

From these studies it is evident that there may be other microbial species of importance to the vaginal ecosystem; however, studies with larger numbers of women other than postmenopausal subjects (as tested here) are required to determine whether organisms such as A. vaginae are truly abnormal. If such is the case, this raises the question of appropriate treatment of BV and the need to reevaluate culture and microscope findings with respect to adverse clinical outcomes.