Rhodococcus spp. as ubiquitous contaminants of paraffin‐embedded tissues in PCR analysis for Mycobacterium spp. skin infections

Abstract

Molecular techniques have greatly enhanced the sensitivity of the laboratory diagnosis of infectious diseases. Highly sensitive polymerase chain reaction (PCR) analysis of tissue DNA is being used to characterize and speciate infectious agents. The use of PCR is increasing in the diagnosis of Mycobacterium spp. infections. Frequently, the primers used are not specific for the genus and they may amplify other closely related organisms such as Rhodococcus spp. Some of these organisms are ubiquitous and, as for organisms of the genus Mycobacterium, they are usually isolated from soil. For this reason, they may be potential contaminants of instruments, reagents or even skin. As part of a study to characterize the causative organisms of feline leprosy, nodular pyogranulomatous lesions from cats with no recognizable organisms in tissue sections, using acid‐fast stains or the BCG technique, were analysed by nested‐PCR. Briefly, samples from 12 cats were obtained retrospectively from paraffin‐embedded tissues, deparaffinized and DNA extracted by standard techniques. Extracted DNA was amplified with previously characterized sets of primers for a fragment of the 16S rRNA gene. Amplified fragments were sequenced in an automated sequencer, analysed and compared by computer analysis with known sequences of related organisms. In addition, 14 feline and canine control tissue samples were analysed. These included mildly inflamed (psychogenic alopecia and allergic skin disease, n = 5) and noninflammatory (neoplastic skin masses, n = 3) skin lesions, and internal organs (liver, spleen and lung, n = 6). The results revealed DNA sequences consistent with Rhodococcus spp. (99–100% homology) in nine of 12 (75%) of the acid‐fast bacteria‐negative nodules. The remaining two samples were negative. None of the samples yielded amplicons with homology to Mycobacterium spp. Additionally, nine of 14 (64%) unrelated control tissues showed similar amplification of a bacterium with high homology (98–99%) to Rhodococcus erythropolis and other Rhodococci; two of 14 amplified a fragment of DNA with homology to Gordonia spp. (soil organism), and one sample amplified a fragment of DNA with homology to Mycobacterium sphagni (soil organism). The remaining two samples were negative. Parallel controls (PCR master mix inoculated with sterile water and PCR master mix alone) were consistently negative. Based on these results, two main scenarios may be in play: (1) in lesions with histologically undetectable acid‐fast organisms, the amount of mycobacterial DNA may be very low, and other dominant contaminant organisms such as Rhodococcus spp. may be amplified preferentially; (2) a negative acid‐fast result obtained histologically may be accurately predictive for an absence of infection with Mycobacterium spp., and only contaminants are amplified. In summary, the significance of Rhodococcus spp. as a causative agent of nodular dermal lesions with no recognizable acid‐fast organisms, when results are obtained exclusively by PCR analysis, should be questioned. In this event, the aetiology of the lesions should be further investigated using other complementary diagnostic procedures (e.g. DNA capture, in situ hybridization, in situ PCR). Funding: Self‐funded.