Nuclear and kinetoplast DNA analyses reveal genetically complex Leishmania strains with hybrid and mito-nuclear discordance in Peru

Polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis of the mannose phosphate isomerase (mpi) gene was applied to 134 skin samples collected from patients with cutaneous leishmaniasis (CL) in Peru for identification of the infecting parasite at the species level, and the results were compared with those of cytochrome b (cyt b) gene sequencing obtained in previous studies. Although most results (121/134) including 4 hybrids of Leishmania (Viannia) braziliensis and L. (V.) peruviana corresponded to those obtained in the previous study, PCR-RFLP analyses revealed the distribution of putative hybrid strains between L. (V.) peruviana and L. (V.) lainsoni in two samples, which has never been reported. Moreover, parasite strains showing discordance between kinetoplast and nuclear genes (kDNA and nDNA), so-called mito-nuclear discordance, were identified in 11 samples. Of these, six strains had the kDNAs of L. (V.) braziliensis or L. (V.) peruviana and nDNAs of L. (V.) guyanensis, and three strains had the kDNAs of L. (V.) shawi and nDNAs of L. (V.) braziliensis. The rest were identified as mito-nuclear discordance strains having kDNAs of L. (V.) braziliensis or L. (V.) peruviana and nDNAs of L. (V.) lainsoni, and kDNAs of L. (V.) lainsoni and nDNAs of L. (V.) braziliensis. The results demonstrate that Leishmania strains in Peru are genetically more complex than previously considered.


Introduction
Parasitic protozoa belonging to the genus Leishmania show marked epidemiologic and clinical diversity, causing a wide-range of human and animal diseases extending from localized, selflimiting cutaneous lesions, and severe diffuse and destructive mucocutaneous lesions, to disseminating visceral infection that is fatal in the absence of treatment. The clinical heterogeneity, together with reservoir host spectrum and vector species compatibilities, lead to different parasite species associations, with over 20 species related to human infections. The cause of this heterogeneity, whether by continuous accumulation of mutations via mitotic cell division and/or via sexual recombination promoting admixtures of divergent genomes, is still a matter of debate [1]. Even if we are far from explaining how Leishmania parasites evolve and emerge in natural populations, correct species identification and searching for evidence of genetic recombination can provide key clues to the ecology and transmission patterns of these organisms. Correct diagnosis of leishmaniasis will also help to determine the clinical prognosis and an appropriate species-specific therapeutic regimen [2].
Multi-locus enzyme electrophoresis (MLEE) has been the gold standard for species characterization [3,4]. However, this classification has been challenged based on genetic analysis of molecular targets, as they were much simpler, easier, and more practical [5][6][7][8][9][10][11][12][13][14]. Kinetoplast DNA (kDNA) is frequently used as a target for detection and typing of Leishmania due to its multicopy nature and high sensitivity [15,16]. However, combining nuclear DNA (nDNA) and kDNA markers has improved the power of molecular data to detect unexpected genetically complex strains with characteristics of hybrid and mito-nuclear discordance widely distributed in Ecuador [17]. These recent findings have shown the necessity of updating the available data in its neighbouring country of Peru, with a similar eco-epidemiological situation, and searching for genetic recombination in Peruvian strains that were identified at the species level by kinetoplast cytochrome b (cyt b) gene sequence analysis in a previous study.  [18].
Previous results on the current epidemiological situation in Peru showed that the predominant species were L. (V.) peruviana, L. (V.) braziliensis, and L. (V.) guyanensis in the Andean highlands, tropical lowland rainforests, and northern to central rainforest areas, respectively [19][20][21][22]. These studies also demonstrated the presence of L. (V.) lainsoni and L. (L.) amazonensis in lower-altitude rainforest areas [20][21][22], and the current prevalence in the same areas was confirmed [18]. The identification of hybrid forms was performed using a polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis of a short mannose phosphate isomerase (mpi) gene fragment. Hybrid genotypes have been rarely revealed among natural isolates of Leishmania parasites, which conventionally indicated that members of the genus possess the machinery for genetic exchange [17,19]. A hybrid of L. (V.) braziliensis and L. (V.) peruviana was first reported in the Department of Huanuco [19,23], and was recently identified in the Department of Cajamarca [18]. Additionally, just four cases of infection by L. (V.) shawi were detected in rainforest areas by the Departments of Junin, Madre de Dios, Cusco, and Puno, as previously reported [18,22], and the question remains unanswered about the low rate of infection caused by this species. Verification of the full details of gene property variants will help to understand the eco-epidemiological situation in this endemic area.
The present study utilized PCR-RFLP analysis targeting the mpi gene sequence to identify the infecting parasite at the species level in 134 skin samples collected from patients with cutaneous leishmaniasis (CL) in Peru, and the results were compared with those of cyt b gene sequencing obtained in previous studies. The results demonstrate that genetically complex Leishmania strains are present in Peru, highlighting the need to combine both nDNA and kDNA targets to improve the validity of species identifications for full details of gene property variants.

Sample collection
The clinical samples employed in this study were collected from patients suspected of having CL from 58 sites in 38 provinces administered by 17 departments in Peru using FTA cards (Whatman, Newton Center, MA, USA) and Giemsa-stained smears in previous studies [18,22] (S1 Fig). Only identified specimens in previous studies were used in this study to avoid selection bias. The preparation of FTA Card samples was done as previously described [18]. Briefly, two-mm-diameter disks were punched out from each filter paper. The disks were then transferred to separate tubes, washed two times with FTA Purification Reagent (Whatman) and one time with Tris-EDTA buffer, and then air dried. The filter paper disks were directly subjected to PCR amplification. DNA extraction from Giemsa-stained smears obtained from skin lesions (ulcers and/or nodules) on CL patients was performed as previously described [18]. Briefly, 50 μL of DNA extraction buffer [150 mM NaCl, 10 mM Tris-HCl (pH 8.0), 10 mM EDTA, and 0.1% sodium dodecyl sulfate (SDS)] containing 100 μg/mL of proteinase K was spotted onto each smear sample and mixed well. The samples were transferred to 1.5-mL tubes and incubated at 37˚C for 12 hours. After being heat-inactivated at 95˚C for 5 min, 0.5-μL portions were directly used as the templates for PCR amplification.

PCR-RFLP analysis
PCR products of the mpi gene fragment were digested with the restriction enzymes HaeIII, VpaK11BI, and BstXI. The restriction enzymes were selected with the support program BioEdit. The digested samples were separated by electrophoresis in 2% agarose gel (STAR Agarose RSV-AGRP from Rikaken, Aichi, Japan) for VpaK11BI and BstXI to separate longer fragments and 3% gel for HaeIII to separate shorter fragments in order to produce DNA fragment patterns using the GeneRuler 100 bp Plus DNA ladder (Thermo Fisher Scientific, Waltham, MA, USA) as molecular size markers. The gel was stained with GelRed Nucleic Acid Gel Stain (Biotium, Hayward, CA, USA), and DNA fragments were visualized with a UV transilluminator.

Ethics statement
Informed consent was obtained from all participating adults; for children under the age of 18 years, consent was also obtained from their parents or guardians. Verbal explanations and written information sheets were provided, following the guidelines of the Ethics Committee of the Ministry of Health, Peru. Clinical samples were collected by well-trained local doctors, biologists, and laboratory technicians from health posts and centers of the ministry. The subjects studied were volunteers in routine diagnosis/screening and treatment programs promoted by the ministry. All routine laboratory examinations were carried out free of charge, and treatment with a specific drug, sodium stibogluconate (Pentostan), was also offered free of charge at each health center. The study was approved by the ethics committee of the Graduate School of Veterinary Medicine, Hokkaido University (approval number: vet26-4) and Jichi Medical University (approval number: 17-080) [17,18]. A case report form (CRF) was collected for each sample and unique codes were assigned to ensure confidentiality. All subjects and guardians consented to lesions being photographed anonymously. Permission was obtained from the Ministry of Health and from community leaders in Peru.

Identification of Leishmania species using PCR-RFLP
In a recent study, PCR-RFLP analyses of a long mpi gene fragment (1,130 bp) using the restriction enzymes HaeIII and HpaI differentiated Leishmania species in Ecuador, except for two very closely-related species, L. (V.) guyanensis and L. (V.) panamensis [17]. In addition, using a  [23]. In the present study, new primers were designed to amplify a 807 bp-mpi fragment (S2 Fig), and the PCR-RFLP technique was optimized using three diagnostic restriction enzymes (HaeIII, VpaK11BI, and BstXI) for species identifications in the studied area. The RFLP patterns were predicted according to the sequences obtained from the GenBank database, and were tested on the Leishmania species listed in S1 Table (Fig 1). The application of PCR-RFLP to 134 skin samples collected from patients with CL in Peru (  Eleven samples showing discordance between mpi gene and cyt b gene analyses were systematically analyzed using multiple targets including mpi, hsp70, 6pgd, and COII-ND1 genes to confirm mito-nuclear discordance (Tables 1 and 2). Sequence analysis confirmed the discordance    Unexpectedly, approximately 10% of the samples were identified as strains with mito-nuclear discordance or hybrids caused by genetic exchange, indicating that the genetic structure of Leishmania strains in Peru is more complicated than previously recorded [18].  [31]. The compatibility of different species to construct hybrids is unknown, and our finding of a putative L. (V.) peruviana/L. (V.) lainsoni hybrid expands the known natural compatibility between species. Both species are quite different with respect to these characteristics, and the impact of hybridization between such divergent species is currently unknown. Although the possibility of co-infection cannot be excluded completely, both alleles were proportionately amplified in PCR-RFLP analysis, strongly suggesting infection by hybrid strains in these patients. It is important to note here that isolation of putative hybrid strains as a culture is required for further detailed characterization of these parasites, since the possibility of co-infection cannot be excluded completely. Lutzomyia ayacuchensis is a proven vector of L. (V.) peruviana in the Andean area of southern Peru [34], whereas no transmission of L. (V.) lainsoni by sand fly species has been reported in these areas. Further investigation in relation to isolation of natural hybrids strains may be necessary for basic parasitological research on how and where genetic exchange occurs among Leishmania species. The genetic exchange may affect vector susceptibility, and hybrid strains may be transmitted by both vectors of parental species/strains [35]. In addition, using an animal model, hybrids between L. (V.) braziliensis and L. (V.) peruviana have been suggested to increase the disease severity when compared with the parental strains [36]. The pathological impact of recombinant genotypes in human infections associated with virulence, transmissibility, and drug susceptibility warrants further investigation in relation to improving methods of identification and isolation of natural hybrid strains.
Our results are the first reported evidence of mito-nuclear discordance among Leishmania species from Peru, highlighting the importance of combining both nDNA and kDNA to improve the validity of species identification for full details of gene property variants. Here, we report the first evidence of mito-nuclear discordance having kDNAs of L. (V.) braziliensis or L. (V.) peruviana and nDNAs of L. (V.) guyanensis in Peru, which was reported in Ecuador [17]. In this context, it is important to note that the most mito-nuclear discordant strains were found in the species identified as L. were detected in rainforest areas by the Departments of Junin, Madre de Dios, Cusco, and Puno, as reported previously [18,22] and the question remains unanswered about the rarity of infection caused by this species. L. (V.) shawi was originally identified as a parasite of wild animals, and sand fly species that transmit it may prefer to feed on animals rather than humans, resulting in a steady rate of infections in wild animals via sand fly bites and a lower risk of infection for humans. However, further investigations using both and nDNA and kDNA genes may provide more clarifications about its epidemiological roles and geographical distributions including mito-nuclear discordance, as reported in the present work. Another unexpected finding was the identification of mito-nuclear discordant strains having kDNAs of L.  [18]. Further investigations are required to elucidate the mechanism of mito-nuclear discordance in protozoa. Several mechanisms have been shown to cause mito-nuclear discordance: incomplete lineage sorting, sex-biased asymmetries, introgression, natural selection, and genetic sweeps mediated by Wolbachia infection [37]. Mito-nuclear discordance was firstly reported in systems where early genetic tools were more developed compared with other taxa such us mitochondrial introgression between two species of fruit fly (Drosophila pseudoobscura and D. persimilis) and two species of mouse (Mus domesticus and M. musculus) [38,39]. The number of cases continued to increase slowly until 2001. Following these early discoveries, methods for assaying numerous individuals for their nuclear genotype became more widely available to researchers and, subsequently, the number of cases increased markedly in animals such as mammals, birds, reptiles, amphibians, fish, and insects [37]. More recent studies revealed that mitonuclear discordance occurs in helminth parasites [40][41][42][43][44].
In the present study, PCR-RFLP targeting leishmanial mpi gene fragments was performed, and the results were compared with those obtained by kinetoplast cyt b gene sequence analysis. The PCR-RFLP method was shown to be practical for the identification of Leishmania species in Peru, revealing unexpected genetically complex Leishmania strains with characteristics of hybrid and mito-nuclear discordance. Since hybrid strains were proposed to aggravate the severity of disease and may be transmitted by a larger range of sand fly species, many questions have been raised about the occurrence and frequency of such cross-species genetic exchange under natural conditions, modalities of hybrid transmission, their long-term maintenance, and the consequences of these transfers on phenotypes such as drug resistance or pathogenicity. Further country-wide epidemiological studies will be needed to reveal the characteristics of hybrid and mito-nuclear discordance and provide further insight into the mechanism of genetic exchanges of these parasites.