A study to detect Gardnerella vaginalis DNA in interstitial cystitis


R.A. Dixon, Department of Biological Sciences, University of Lincoln, Brayford Pool, Lincoln LN6 7TS, UK.
e-mail: rdixon@lincoln.ac.uk


Objective To investigate the possible role of Gardnerella vaginalis in interstitial cystitis (IC), using molecular methods to avoid difficulties with the culture and recovery of viable organisms, and the problems associated with the recovery of low numbers of culturable organisms.

Materials and methods Thirty-three bladder biopsy samples (29 paraffin-embedded and four freshly frozen) from patients with IC, diagnosed according to National Institute of Diabetes, Digestive and Kidney Diseases criteria, were assessed. Biopsy samples were used as urine samples may be contaminated by normal vaginal flora. A positive control comprised a ‘normal’ biopsy sample from a patient with a previous bladder tumour, seeded with G. vaginalis NCTC 10915. Microbial DNA was extracted from all paraffin-embedded and fresh specimens, and subjected to in vitro amplification by polymerase chain reaction (PCR) with G. vaginalis-specific primers.

Results The anticipated PCR product of 333 base pairs was obtained with the positive control, whereas none of the other biopsy samples showed positive amplification specific for G. vaginalis.

Conclusion As there was no G. vaginalis DNA in any of the samples from patients with IC, it is an unlikely candidate in the pathogenesis of IC.


The causes of IC remain unknown, despite intensive research; the disease mainly occurs in females, with a reported male : female ratio of 1 : 12 [1]. Although bacterial infection is the most likely aetiology, no single organism has so far been confirmed to be the causative agent. Gardnerella vaginalis is found in the vaginal smears from 60% of women [2,3] and it would not be unreasonable therefore to suspect a possible role in IC, as these bacteria can easily gain entry to the bladder through the short female urethra. Others have suggested that infection with fastidious bacteria, particularly G. vaginalis, has a role in the pathogenesis of IC, and that they are often not detected by conventional culture [4,5]. We investigated the possible role of G. vaginalis in IC, using molecular methods to avoid difficulties with the culture and recovery of viable organisms, and the problems associated with the recovery of low numbers of culturable organisms.

Materials and methods

The study comprised 33 bladder biopsy samples (29 paraffin-embedded and four fresh frozen) from patients with IC, diagnosed according to criteria proposed in 1988 [6] with some modifications (criteria based on CMG were not strictly adhered to, as urodynamic studies were not undertaken in all the patients). A ‘normal’ biopsy sample from a patient with a previous bladder tumour was seeded with G. vaginalis culture (NCTC 10915) and included as a positive control. In addition, all paraffin-embedded specimens were investigated for the presence of mitochondrial DNA, as evidence of successful human DNA extraction [7].

Specimens were de-waxed and pretreated with lysozyme and lysostaphin. DNA was extracted from all specimens using the Easy-DNA™ genomic DNA isolation kit, version 1.0 (Invitrogen, Paisley, UK), according to the manufacturers' instructions. Extracted DNA was used for PCR with G. vaginalis specific primers, to amplify a 333 bp product according to the method described previously [8], with some modifications. Primers H and K were obtained from Perkin-Elmer Ltd. (UK) and were specific for G. vaginalis; the sequences were: primer H, TTT ACT GGT GTA TCA CTG TA (bp 381– 400) and primer K, CCG TCA CAG GCT GAA CAG (bp 714–696). The H and K primers (50 pmol of each) were added to a reaction mixture of 5 µL of 10 × PCR buffer (500 mmol/L KCl, 100 mmol/L Tris-Cl, 15 mmol/L MgCl2, 0.1% w/v gelatine, pH 8.3), 8 µL of dNTP mixture (0.02 µmol each of dATP, dCTP, dGTP and dTTP, all 10 mmol/L) and 2 µL of template DNA. Distilled water was added to give a final reaction volume of 50 µL and the tubes sealed with an overlay of 40 µL of autoclaved mineral oil. Taq DNA polymerase (2.5 U of a 5000 U/mL stock solution, Pharmacia Biotech, Amersham, Bucks) was added just before starting the PCR cycle after pre-denaturing the reaction mixture.

G. vaginalis DNA (2 µL, 520 ng) from reference strain NCTC 10915 was used as the positive control. PCR-negative controls were established by replacing the template DNA with distilled water. Sample negative controls were established using 2 µL (250 pg) Escherichia coli DNA (D-4889, Sigma Chem Co., Poole, UK). PCR reactions were performed in a thermocycler after pre-denaturing samples for 2 min at 94 °C, and comprised 40 cycles of denaturation at 94 °C for 1 min, followed by annealing at 60 °C for 1 min and extension at 74 °C for 1 min. This was followed by further extension at 72 °C for 5 min and then the reaction mixture was maintained at 4 °C until the amplified products were analysed. Gel electrophoresis was undertaken with 2% w/v agarose (electrophoresis grade, 15510–019, Gibco BRL, Paisley, UK) concentrations in 0.5% w/v buffer (TBE, 0.5 × 0.045 mol/L Tris-borate, 1 µmol/L EDTA). Wells were loaded with 10 µL of PCR products mixed with 2 µL of gel-loading buffer (0.25% bromophenol blue, 0.25% xylene cyanol FF, 15% Ficoll in water); 1 µL of a 100 bp ladder (27-4001-01, Pharmacia) was mixed with 2 µL of loading buffer and included as a size marker. Electrophoresis was carried out in gel tanks (Biorad, Herts, UK) containing 0.5% w/v TBE buffer at a constant 120 V for ≈ 2 h. Gels were stained with SYBR Green 1 nucleic acid gel stain; a stock solution of SYBR Green 1 nucleic acid gel stain in DMSO (S-7563, Molecular Probes, Eugene, OR, USA) was diluted 1 : 10 000 in buffer (10 mmol/L Tris-HCl, 1 mmol/L EDTA, pH 8.0). Gels were placed in a staining container, covered with staining solution, protected from light by aluminium foil and gently agitated at room temperature for 40 min. Gels were visualized under ultraviolet (302 nm) illumination and photographed on Polaroid 667 black and white film, using a SYBR Green gel-stain photographic filter (S-7569, Molecular Probes) at f12 for 1 s in a dark room.


Amplification of the human mitochondrial gene from DNA extracted from all the paraffin-embedded biopsy samples showed a specific band of 97 bp. The positive control sample (bladder biopsy seeded with G. vaginalis culture) showed the anticipated PCR product of 333 bp and confirmed the presence of G. vaginalis DNA. The sensitivity for detecting G. vaginalis was ≈ 5 pg of DNA. None of the biopsy samples from patients with IC (either fresh-frozen or paraffin-embedded) showed positive amplification specific for G. vaginalis (data not shown).


Urologists in the earlier part of the 20th century, as cited in [9], thought that IC was as a result of chronic or secondary bacterial infection; some still consider this to be the case. Only when numerous research studies failed to identify the agent responsible was it realized that there is more to IC than mere infection. Some still consider that infection is the initial trigger which leads to complex immune cell changes in the bladder wall, culminating in the ‘full picture’ of IC [9].

The importance of fastidious bacteria as pathogens is now well recognized in diseases such as meningitis, endocarditis, liver abscess and infections of the female genital tract [10]. Micro-organisms which are not detected by overnight incubation from aerobic culture on primary isolation media are designated ‘fastidious’. G. vaginalis is a fastidious organism, requiring selective media and special conditions for growth [11]. In a study of the survival of G. vaginalis in human urine, Lam and Birch [12] found that the counts declined by > 99.9% in urine held at 37 °C for 24 h and the viability was nearly lost after 6 h in urine with a pH 5 or 7. Storage at 4 °C and pH 6 were optimal for the survival of the organism. Thus it is not surprising that G. vaginalis is often undetected.

In a series of papers from 1979, Maskell et al. reported the isolation of certain fastidious organisms in the urinary tract of patients with urethral syndrome [13–16]. Hamilton-Miller et al.[17] and Brumfitt et al.[18] disputed Maskell's claim of a role for fastidious organisms in urethral syndrome. Despite much work no one has been able to conclusively confirm or exclude specific bacteria. Maskell has also suggested that urethral syndrome and IC are earlier and later stages of the same disease process, and that infection with fastidious bacteria is implicated in the pathogenesis [4,5].

The present study was undertaken to specifically address the issue of the involvement of G. vaginalis in the pathology of IC. Previous research to assess the involvement of G. vaginalis was based mainly on culture methods which may miss the presence of this organism, because of the difficulty in recovering them. Furthermore, bladder biopsy samples are better than MSU samples, which may be contaminated by normal vaginal flora. The present findings are consistent with the molecular study of Keay et al.[19], who assessed bladder biopsies from six patients with IC and six controls, using a PCR method to amplify DNA encoding bacterial 16S rRNA. All of the biopsies, including the controls, were positive and sequence data suggested the presence of several different genera of bacteria, including Acinetobacter, Propionobacterium, Salmonella and Escherichia. None of the samples tested showed the presence of G. vaginalis.

Archival paraffin-embedded samples used in the present study could be an important resource for similar studies, as microbial DNA can be relatively easily recovered from them. The present study showed no G. vaginalis DNA in bladder biopsy samples, and although it does not totally exclude its involvement in the pathogenesis of IC, it is thus an unlikely candidate.


We are grateful for the gift of paraffin-embedded bladder biopsies from Dr C.L. Parson, Veterans Administration Medical Center, San Diego, USA, and the kind cooperation of Messrs P.A. Hamilton Stewart and G.M. Flannigan, Bradford Royal Infirmary, Bradford. We also thank Mr P.H. O'Reilly (Stepping Hill Hospital, Stockport) for reviewing the manuscript. This work was presented as a poster in the ‘Anaerobe 2000’ meeting in Manchester UK, July 2000.


M. Agarwal, MS, PhD, FRCS(Urol), formerly, Research Fellow, University of Bradford, currently SpR Urology, Glasgow Royal Infirmary.

R.A. Dixon, BSc, MSc, PhD, formerly, Senior Lecturer in Biomedical Sciences, University of Bradford, currently Principal Lecturer, University of Lincoln.