Identification of Hemotropic Mycoplasmas in an Eastern Box Turtle (Terrapene carolina carolina) and a Yellow-bellied Slider (Trachemys scripta scripta) from North Carolina, USA

Identification of vector-borne diseases in reptiles has largely been unexplored. Our project was inspired by the documentation of Bartonella spp. in sea turtles in North Carolina, US (Valentine et al. 2007), because subsequent investigations of reptiles, including terrestrial turtles, as a possible reservoir host remained lacking. Various Bartonella spp. have been identified in North Carolina cetaceans (Harms et al. 2008; Maggi et al. 2008) and river otters (Chinnadurai et al. 2010). Also, known and novel Mycoplasma spp. have been documented in reptiles, the majority of which are associated with clinical signs involving the respiratory tract (Brown et al. 2001; Feldman et al. 2006; Ossiboff et al. 2015). Hemotropic mycoplasmas (hemoplasmas) are small, epicellular, cell wall-deficient bacteria encountered in the blood of a variety of animal species that can contribute to pathogenic outcomes such as anemia, ill-thrift, and infertility (Messick 2004). Although it has been determined that hemoplasmas share phenotypic and genetic similarities with other Mycoplasma spp., variability in the genomes of these bacteria continue to complicate exact classification at the species level (Guimaraes et al. 2014). Different hemoplasma species are zoonotic and cause the same effects on red blood cells in human patients as they do in other mammals (Sykes et al. 2010; Steer et al. 2011). Clinical signs and pathogenic properties (if any) associated with hemoplasmas in reptiles are unknown.

The North Carolina State University College of Veterinary Medicine Turtle Rescue Team is a student-run organization of volunteers who rehabilitate wild reptiles (Lewbart et al. 2005). Between May and July 2014, samples of 53 spleens were collected from turtles admitted to the Turtle Rescue Team that either died or were euthanized, mostly due to trauma (mainly vehicular). The 53 specimens included 23 eastern box turtles (Terrapene carolina carolina), 14 yellow-bellied sliders (Trachemys scripta scripta), six river cooters (Pseudemys concinna), five eastern painted turtles (Chrysemys picta), three common snapping turtles (Chelydra serpentina), one common mud turtle (Kinosternon subrubrum), and one common musk turtle (Sternotherus odoratus). Turtles originated from central North Carolina. After death, specimens were stored at 4 C for up to 2 d until tissue harvest. We performed necropsies with the guidance of a sea turtle necropsy manual (Work 2000) using appropriate modifications made for each species. If DNA extraction could not be performed immediately, spleen specimens (up to 5 g, depending on the size of turtle) were harvested and stored frozen at 0 C in 5 mL screw-cap centrifuge tubes.

Spleen specimens were thawed or cut directly after necropsy into small pieces using a scalpel blade. Pathogen DNA extraction was performed with a standardized kit (DNeasy Blood & Tissue Kit, Qiagen, Valencia, California, USA), following the manufacturer's instructions for DNA extraction from tissues. Pathogen DNA concentrations and absorbance ratios at 260 nm and 280 nm were determined using the Nanodrop® ND-1000 (Thermo, Wilmington, Delaware, USA) to assess the concentration and purity of the DNA samples collected. Repeated DNA extractions were performed in eight cases to ensure that yields were >15 ng/μL.

We used pathogen DNA amplification, through PCR, to detect the presence of Bartonella spp. and hemotropic Mycoplasma spp. DNA in each sample. We amplified hemotropic Mycoplasma by targeting a conserved region of the 16SrRNA using oligonucleotides HemMycop16S-322s: 5′ GCC CAT ATT CCT ACG GGA AGC AGC AGT 3′ and HemMyco16S-938as: 5′ CTC CAC CAC TTG TTC AGG TCC CCG TC 3′ as previously described (Maggi et al. 2013b). Detection of Bartonella DNA was performed by targeting the 16S–23S intergenic spacer region using oligonucleotides 325s (5′ CCTCAGATGATGATCCCAAGCCTTTTGGCG 3′) and 1000as (5′ CTGAGCTACGGCCCCTAAATCAGG 3′) as forward and reverse primers, respectively (Valentine et al. 2007). Amplification for each genus was performed in a 25 μL final volume reaction. The 25 μL reaction mix contained 12.5 μL of MyTaq™ Red Mix (Bioline, Taunton, Massachusetts, USA), 0.2 μL of 50 μM of each forward and reverse primer (IDT® DNA Technologies, Coralville, Iowa, USA), 7.3 μL of molecular grade water, and 5 μL (75–150 ng total) of template DNA. We prepared positive controls using either 5 μL Mycoplasma hematoparvum 16SrRNA or Bartonella henselae intergenic spacer plasmid vectors (pGEM-T Easy Vector Systems, Promega, Madison, Wisconsin, USA) at a final concentration of five copies per microliter, for both Mycoplasma and Bartonella detection. Amplifications were performed in an Eppendorf Mastercycler EPgradient S® (Eppendorf, Hauppauge, New York, USA).

We used 2% agarose gel electrophoresis to analyze PCR products, with detection using ethidium bromide under ultraviolet light. Bartonella spp. DNA was not amplified from any turtle spleen. Two turtle spleens contained Mycoplasma spp. DNA. Amplicon products were sequenced by Genewiz (Research Triangle Park, North Carolina, USA) to establish species strain identification using chromatogram and alignment analysis (ContigExpress® and AlignX software, Vector NTI® v10, Invitrogen, Carlsbad, California, USA). After sequencing, phylogenetic and taxonomic comparisons were made using GenBank. The DNA sequence from the eastern box turtle number 24 (GenBank accession no. MG649987) was most similar to the following Mycoplasma sequences: 87% (490/564 base pairs [bp]) similar to Mycoplasmataceae bacterium T2110 (GenBank: DQ318957.1), an unclassified mycoplasma found in Atlantic Cod (Gadus morhua); 87% (484/557 bp) similar to Candidatus Mycoplasma haemohominis, found in humans; and 86% (483/564 bp) similar to Candidatus Mycoplasma haemominutum, found in cats (Fig. 1). The DNA sequence from yellow-bellied slider number 47 (GenBank: MG519832) was most similar to the following Mycoplasma sequences: 92% (514/557 bp) similar to Mycoplasma insons, previously found in green iguanas (Iguana iguana); 92% (514/557 bp) similar to Mycoplasma cavipharyngis, found in guinea pigs (Cavia porcellus); 92% (512/557 bp) similar to Mycoplasma fastidiosum, found in equines (Fig. 2).

Hemotropic Mycoplasma spp. have not previously been amplified from a reptile. The potential clinical impact that hemotropic Mycoplasma spp. have on reptiles is unknown because their pathogenicity in reptile species remains undetermined. Many turtles enter a rehabilitation facility already in compromised health, and stresses associated with trauma and rehabilitation can exacerbate what would otherwise be nonclinical or minor health issues. If hemotrophic Mycoplasma spp. are normally nonpathogenic in chelonian species, hemoplasmosis could have a negative impact on an animal's recovery and chance for ultimate release for those animals in rehabilitation facilities where concurrent diseases can exacerbate nonpathogenic conditions.

We do not know if the Mycoplasma spp. in chelonians are contagious to humans or to other animal species. Hemotrophic Mycoplasma spp. amplified from reservoir hosts, including foxes (Vulpes lagopus; Mascarelli et al. 2015), deer (Odocoileus virginianus; Maggi et al. 2013a), raccoon (Procyon lotor; Frerichs and Holbrook 1971), rodents (Elko and Cantrell 1968), and other animals (Messick 2004), might serve as sources for human infections. Zoonotic diseases such as Salmonella spp. and Eastern Equine Encephalitis virus have occurred following contact with nonclinical reptiles (White et al. 2011; Middleton et al. 2014). We used splenic extracts in our analyses and presumed that the Mycoplasma spp. was present in erythrocytes and not in splenic stroma.

Although we did not amplify DNA from a Bartonella spp. from terrestrial turtles in this study, a negative PCR result does not rule out a potential Bartonella spp. infection. In nonreservoir hosts, Bartonella spp. are most often found in very low concentrations in blood and splenic tissues, which could result in false negative PCR amplification. The role, if any, which reptiles play in transmission of Mycoplasma is unknown, and further investigation into hemotropic Mycoplasma and other blood pathogens is necessary. Our study provided additional information supporting the presence in reptiles of hemotropic Mycoplasmas, which were genetically very similar to pathogenic strains found in humans and other mammals. Until further investigation has evaluated the pathogenicity and spread of hemotropic Mycoplasma in reptiles, isolation of turtles with compromised health status, maintaining cleanliness and sterility of surgical and medical rehabilitation supplies or use of separate supplies for infected turtles, and having personal protective equipment for individuals handling potentially infected turtles might be indicated.

We thank the Robert J. Koller Aquatic Animal Medicine Research Endowment for financial support and the Turtle Rescue Team students who assisted in the procurement of specimens and samples.