• donovanosis
• Calymmatobacterium granulomatis
• molecular diagnosis
• colorimetric detection

Donovanosis is a chronic genital ulcerative disease with endemic foci in socially disadvantaged populations throughout the world.¹ In recent years there has been a resurgence of interest in this disease owing to the association of genitoulcerative disease with an increased risk of HIV-1 transmission and acquisition. While unproved, there is a highly likely synergistic link between HIV-1 and donovanosis.² The Australian government has a target to eradicate donovanosis by 2003.³

The causative organism of donovanosis, Calymmatobacterium granulomatis, has been associated with the genus Klebsiella on the basis of morphological, ultrastructural, and antigenic studies.^{4, 5} Furthermore, we have recently reported phylogenetic evidence for its reclassification as Klebsiella granulomatis comb nov.⁶

Previously we reported the development of the first polymerase chain reaction (PCR) assay for the detection of C granulomatis in clinical specimens.⁷ The specificity of the PCR assay is based on two unique base changes in the phoE gene of C granulomatis which eliminate HaeIII restriction sites, enabling clear differentiation from closely related species of Klebsiella. A major advantage of this test is that it can be utilised with swab rather than biopsy/tissue smear specimens. It was suggested in this report that it should now be possible to incorporate this PCR/HaeIII digest into a colorimetric detection system for use in routine diagnostic laboratories. We now report a colorimetric detection system for the detection of C granulomatis from clinical specimens. There is no intent to present this work as a validation trial. Rather, our intent is to make this methodology available to as many researchers/medical practitioners as possible so that appropriate validation trials may be carried out.

DNA extractions and amplification of 188 bp phoE products were carried out as previously reported,⁷ with the exception that a biotin group was attached to the 5` PCR primer. The PCR products were then cleaved by addition of 1 unit of HaeIII and incubation at 37°C for 1 hour. Tosyl activated magnetic beads were obtained (Dynal, Carlton, Victoria, Australia) and washed as recommended by the manufacturer. A capture oligonucleotide with 5` amine group was designed internal to the HaeIII restriction sites of the C granulomatis phoE product (5`NH<sub>2</sub>CCAGCAGGTTCTGATC GTT3`) and added to the beads at 150 pmol/mg beads in 0.1 M borate buffer pH 9.5. Incubation at 65°C for 15 minutes allowed covalent bond formation between the tosyl groups and 5` amine groups. Free tosyl groups were blocked with 0.2 M ethanolamine pH 7.0 containing 0.1% bovine serum albumin for 4 hours at 37°C. Before use the beads were diluted 1 in 4 with Probe Suspension 2, pH 4.0 (sodium phosphate solution containing <0.2% solubiliser and <26% chaotrope, contained within Cobas Amplicor Chlamydia trachomatis Detection Kit, Roche Diagnostic Systems, Inc, Branchburg, NJ, USA). PCR/HaeIII digest products (25 μl) were denatured with 0.1 M sodium hydroxide (25 μl). The denatured solution (30 μl) was incubated with 100 μl of prepared beads for 30 minutes at 37°C to allow for neutralisation and binding with the capture oligonucleotide. All subsequent steps of detection were carried out using the standard EIA methodology (that is, wash, add avidin-HRP conjugate, wash, add substrate). A magnetic particle concentrator (Dynal, Carlton, Victoria, Australia) was utilised at each step to separate the beads from the supernatant and allowed all steps to be carried out in one tube.

A schematic representation of this system is provided (fig 1).

The sample set tested with this detection system was identical to that previously reported,⁷ and included 14 positive samples from patients with donovanosis, eight samples from patients with unrelated genital conditions (herpes genitalis, syphilis, gonorrhoea, human papillomavirus infection, and mixed anaerobic infection—there is no reported chancroid or lymphogranuloma venereum in the Northern Territory of Australia), and isolates of K pneumoniae, K ozaenae, and K rhinoscleromatis. Owing to the cleavage of the fragment containing the biotin group from the capture oligonucleotide target region it was expected that the three Klebsiella digest products would give negative readings while the intact C granulomatis fragment would give positive readings.

Strong colorimetric readings were obtained for all 14 positive samples from patients with donovanosis. All eight samples from patients with unrelated genital conditions were negative as were PCR digests obtained from the three Klebsiella isolates (fig 1).

This methodology has advantages over agarose gel electrophoresis in that it can be utilised by routine diagnostic laboratories and incorporates a second level of specificity via the presence of an internal capture oligonucleotide.

All 14 positive samples from patients with donovanosis included in this report were obtained from the Top End of the Northern Territory of Australia over a 2 year period. However, the test has been/is being successfully utilised by doctors in Central Australia, South Australia, and Western Australia (data not shown as ethical clearance applies only to Top End of the Northern Territory) and at the very least should be particularly valuable if there is a serious attempt to eradicate donovanosis in Australia.

It is interesting to note that the base change at position 532 of the coding nucleotide sequence (C→T) results in the amino acid change Arg157→Cys157. Arg157 is located near the top of one of the surface exposed loops in the Klebsiella PhoE protein and corresponds to Arg158 of the Escherichia coli PhoE protein.⁸ Korteland et al reported that all E coli K 12 strains tested in their laboratory with resistance to the PhoE specific bacteriophage TC45 showed a mutation at Arg158.⁹ While unproved, it is plausible that the C granulomatis isolates present in Australia have also acquired this selective advantage at some point in their evolution.

This gives some strength to the future stability of this mutation in the Australian context though the question still remains as to whether the organism is clonal in an international context. The specificity of the PCR assay still needs to undergo trials and be validated using a larger number of samples from geographically diverse locations within and outside Australia in order to test the stability and applicability of the two unique base changes on which it relies.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Left: Schematic representation of the colorimetric detection system for C granulomatis. Arrows represent HaeIII restriction sites within the 188 bp phoE product for C granulomatis (above) and other klebsiellae (below). The black box within the 188 bp product represents the target region for the capture oligonucleotide. Right: Results from left to right: K pneumoniae; K rhinoscleromatis; positive swab sample from patient with donovanosis; positive swab sample from patient with donovanosis; K ozaenae.