Presence of opportunistic bacteria (Rhizobium spp.) with potential for molecular misdiagnosis among canine and feline clinical samples

Abstract

Rhizobium radiobacter was detected in 12 of 187 dogs and 2 of 100 cats using a polymerase chain reaction (PCR) assay formerly designed for the Rickettsia genus. Although PCR primers used for pathogenic infectious agents are specifically assessed to avoid cross-amplification, this retrospective study highlights the importance of sequencing to avoid molecular misdiagnosis.

Molecular techniques have become very popular for diagnosis of infectious diseases. Amplification of DNA of the pathogen is an invaluable tool to identify the possible involvement of these pathogens in disease, helping to overcome substantial challenges for the veterinary clinician regarding diagnosis and medical management.

Commercial amplification kits are not available for many clinically significant pathogens and some of them require researcher-designed methods. An increasing number of laboratories have focused on the development of new molecular diagnostic methods, resulting in the proliferation of PCR tests (1). Specificity depends on design of the primers, since they are chosen based on the DNA sequence targeted to amplify a gene for a selected pathogen. Although primers are tested in silico against genetic data bases, some primer pairs may also amplify phylogenetically related opportunistic organisms which had not been considered in the design. Previous reports have shown the potential amplification of Mesorhizobium DNA from patient samples undergoing testing for Bartonella (1).

Rickettsia belongs to the alpha-2 subgroup of Proteobacteria that includes other organisms such as Bartonella and Brucella, plant pathogens, and symbionts such as Rhizobium and Agrobacterium (2). As these bacteria are phylogenetically closely related, conserved sequences can occur among them resulting in amplification of agents other than the pathogens for which the primers were designed. We have tested the prevalence of opportunistic bacteria (Rhizobium spp.) in clinical samples submitted for diagnosis of vector-borne diseases. The aim was to assess the presence of Rhizobium as a potential cause of misdiagnosis with Rickettsia spp. among clinical samples. Results were stratified with regard to selected historical findings in order to ascertain if Rhizobium could be more prevalent among unhealthy patients, which could suggest a pathogenic role for this organism.

Dogs (n = 187) and cats (n = 100) admitted for various reasons to the Veterinary Teaching Hospital of Barcelona, Spain, between May 2005 and December 2006, were entered into this retrospective study. Archived blood samples from these patients were used.

The medical records of the patients were reviewed retrospectively. Information evaluated included signalment, history, and physical examination findings, complete blood (cell) count (CBC), serum biochemistry results, and other diagnostic tests. These samples were also tested by PCR for Leishmania infantum, Anaplasma spp., Ehrlichia spp., Babesia spp., Bartonella spp., and Hepatozoon spp.

Based on the physical examination status (healthy or unhealthy) and the results of diagnostic tests, the animals in this study were divided into 2 groups. Group 1 included cats (n = 48) and dogs (n = 76) considered to be healthy, and group 2 included cats (n = 52) and dogs (n = 111) with clinical and laboratory findings of illness.

Blood had been collected by cephalic or jugular venipuncture and frozen until DNA extraction. The DNA was obtained from 0.5 mL of EDTA-blood as described elsewhere (3). Polymerase chain reaction (PCR) was carried out in a 20-μL reaction mixture containing: PCR buffer 1×, 0.2 mM of each dNTP, 0.2 μM of each primer, 1.5 mM MgCl₂ and 1 U of Ecotaq DNA polymerase (ECOGEN, S.R.L., Barcelona, Spain). The thermal cycling profile was 94°C for 3 min followed by 40 cycles at 94°C for 30 s, 60°C for 30 s, and 72°C for 30 s with a final extension of 72°C for 7 min. The primer pair used was 5′-ATAAGAAAACTGCCGGTGATAAGCC-3′ and 5′-GTTAGTTTTACCTGAAGGTGGTGAGCT-3′ (originally designed and in-silico tested in genetic data bases for Rickettsia spp. amplification). The eukaryotic 18S RNA Pre-Developed TaqMan Assay Reagents (Applied Biosystems, Foster City, California, USA) were used as internal reference for dog genomic DNA amplification to ensure suitability of each sample for PCR amplification and that negative results corresponded to true negative samples rather than to a problem with DNA loading, sample degradation, or PCR inhibition. The DNA from a healthy specific pathogen-free dog was used as an assay negative control and water was a PCR negative control. The positive PCR control was obtained from a clinical sample which had been previously amplified and sequenced to confirm the pathogen.

All positive samples were characterized at the species level using the MicroSeq 500 16S Bacterial Identification Kit (Applied Biosystems) as described elsewhere (4). Sequences obtained were compared with sequences in the GeneBank data base (http://www.ncbi.nlm.nih.gov/BLAST/) and the Ribosomal Database Project II (http://rdp.cme.msu.edu/seqmatch/seqmatch_intro.jsp) to assign the species.

Statistical analyses were performed using the Statistical Package for the Social Sciencies, software packets for Windows, version 14.0 (2005 SPSS, Chicago, Illinois, USA). Contingency table analysis was performed and statistical significance was set at P ≤ 0.05.

Canine and feline clinical samples submitted for diagnosis of vector-borne diseases were amplified with PCR primers designed for Rickettsia spp. amplification. A PCR product of the expected size for Rickettsia spp. was obtained for 12 canine and 2 feline samples. Four positive dogs were from the healthy group and 8 positive dogs had clinical signs of illness (apathy, lameness, epistaxis, weakness, anorexia, fever, splenomegaly, ascitis, polyuria, and polydipsia). The 2 samples from cats were from asymptomatic patients. These 14 amplicons were sequenced and in all of them, the PCR product was identified as Rhizobium radiobacter. Moreover, the PCR-positive control was Rickettsia rickettsi as expected. The overall prevalence of Rhizobium radiobacter was 6.42% (12 of 187) in dogs and 2% (2 of 100) in cats. There was no statistical association between presence of Rhizobium DNA and age, sex, breed, health condition, and other pathogens. However, it was statistically related with use of parenteral medications and/or fluid therapy (P ≤ 0.001).

Rhizobium species are common environmental bacteria that can be found in soil, plants, and water (5). Although many related alpha-Proteobacteria, as Rhizobium (previously Agrobacterium), are not characterized as primary human or even mammalian pathogens, clinical disease has been reported in immunocompromised human patients with underlying hematological malignancy or solid-organ cancer, and primarily with catheter-associated bacteremia (5,6). Rhizobium species are opportunistic pathogens and have been substantiated as rare causes of bacteremia, endocarditis, peritonitis, and endophthalmitis in humans (7,8) and have been found in respiratory secretions from patients with pulmonary cystic fibrosis (9). There is no statistical association between presence of Rhizobium radiobacter and clinical status in our samples, but the presence of R. radiobacter was statistically related with parenteral medications and/or fluid therapy (P ≤ 0.001). Six of the dogs positive for R. radiobacter PCR amplification had been previously hospitalized with a catheter placed for fluid administration (1 for elective surgery and 5 for treatment of several diseases), 3 were blood donor dogs (with catheter placed for fluid administration after blood extraction), and the remaining 3 dogs had been treated with subcutaneous injections of Meglumine antimoniate (Glucantime; Merial Laboratories SA, Barcelona, Spain) for canine leishmaniosis therapy. Although the number of positive samples is low, these data suggest that Rhizobium radiobacter could be acquired by means of parenteral fluids or medications given to those patients. Conversely, the cats with positive results for Rhizobium radiobacter PCR amplification had not received parenteral medication, and the route of transmission of Rhizobium in these cases is unknown. Previous data from humans suggests that it should be considered a potential community-acquired or hospital-acquired pathogen, especially in immunocompromised patients in whom foreign plastic materials such as prosthetic valves and intravenous catheters have been used (5,6). To the authors’ knowledge, no data are available on the pathogenicity of this species in small animals, but the results herein indicate that Rhizobium radiobacter may be present in clinical samples from small animal patients.

Even though primers for molecular diagnosis may be specifically designed to target a genus or species, amplicons obtained with the selected primers must be sequenced to accurately confirm the identity of the organism whose DNA has been amplified and thereby reduce the chance of misdiagnosis. Previous reports have shown the potential limitations of molecular diagnosis when some regions of the 16S–23S rRNA were selected for detection of Bartonella (1). Those primers had been previously used for Bartonella identification until nonspecific PCR amplification of Mesorhizobium was detected as a result of contamination of molecular-grade water. In our case the presence of Rhizobium radiobacter was detected in clinical samples with primers originally designed for Rickettsia spp. amplification. Therefore, not only new primers designed but also primers previously published for those purposes must be correctly probed and tested for target specificity since new emergent and re-emergent species are continually being investigated (10). For this reason, it is nearly impossible to completely exclude the amplification of species with similarities among DNA sequences targeted. Ideally, therefore, sequencing must be the final step in molecular testing to give an accurate diagnosis since misdiagnosis could result in increased costs, a delay in the diagnosis of more serious diseases, and selection of an inappropriate therapy.

In conclusion, Rhizobium radiobacter DNA was detected in clinical samples in small animal practice, although there was no clear pathogenic potential. Further studies involving isolation in culture should be attempted to prove that the organism, and not just its DNA, is present in clinical samples. Polymerase chain reaction is a powerful tool for the detection of DNA from infectious agents such as Rickettsia, but caution should be exercised when interpreting results because closely related microorganisms such as Rhizobium spp. could be amplified. Ideally, sequencing should be done after PCR amplification of pathogen DNA to ensure specificity.