Molecular identification of two newly identified human pathogens causing leishmaniasis using PCR-based methods on the 3′ untranslated region of the heat shock protein 70 (type I) gene

PCR-based methods to amplify the 3′ untranslated region (3′-UTR) of the heat shock protein 70 (type I) gene (HSP70-I) have previously been used for typing of Leishmania but not with Leishmania (Mundinia) martiniquensis and L. (Mundinia) orientalis, newly identified human pathogens. Here, the 3′-UTRs of HSP70-I of L. martiniquensis, L. orientalis, and 10 other species were sequenced and analyzed. PCR-Restriction Fragment Length Polymorphism (RFLP) analysis targeting the 3′-UTR of HSP70-I was developed. Also, the detection limit of HSP70-I-3′-UTR PCR methods was compared with two other commonly used targets: the 18S small subunit ribosomal RNA (SSU-rRNA) gene and the internal transcribed spacer 1 region of the rRNA (ITS1-rRNA) gene. Results showed that HSP70-I-3′-UTR PCR methods could be used to identify and differentiate between L. martiniquensis (480–2 bp) and L. orientalis (674 bp) and distinguished them from parasites of the subgenus Viannia and of the subgenus Leishmania. PCR-RFLP patterns of the 3′-UTR of HSP70-I fragments digested with BsuRI restriction enzyme successfully differentiated L. martiniquensis, L. orientalis, L. braziliensis, L. guyanensis = L. panamensis, L. mexicana = L. aethiopica = L. tropica, L. amazonensis, L. major, and L. donovani = L. infantum. For the detection limit, the HSP70-I-3′-UTR PCR method could detect the DNA of L. martiniquensis and L. orientalis at the same concentration, 1 pg/μL, at a similar level to the SSU-rRNA PCR. The PCR that amplified ITS1-rRNA was more sensitive (0.01 pg/μL) than that of the HSP70-I-3′-UTR PCR. However, the sizes of both SSU-rRNA and ITS1-rRNA PCR amplicons could not differentiate between L. martiniquensis and L. orientalis. This is the first report of using HSP70-I-3′-UTR PCR based methods to identify the parasites causing leishmaniasis in Thailand. Also, the BsuRI-PCR-RFLP method can be used for differentiating some species within other subgenera.


Introduction
Leishmaniasis is a newly emerging disease in Thailand. To date two species of Leishmania have been reported from Thailand: L. (Mundinia) martiniquensis [1] and L. (Mundinia) orientalis [2]. There has been one report of L. infantum but this may have been an imported case [3]. The clinical signs of L. martiniquensis infection include both cutaneous and visceral leishmaniasis in both HIV-infected and immunocompetent patients [1,4]. One case of L. orientalis presented as simple cutaneous leishmaniasis in an immunocompetent patient, and two others as disseminated cutaneous/visceral leishmaniasis in HIV-infected patients [2].
Identification of the species responsible for a Leishmania infection is important for clinical management, treatment, and epidemiology. Although there can be substantial overlap in symptoms, each species will cause its own spectrum of disease manifestations that require their own supportive measures and have different prognoses. For example, L. major and L. tropica both cause cutaneous leishmaniasis in the majority of patients and can occur sympatrically in many regions of North Africa and the Middle East but have different epidemiology [5]. L. major is exclusively zoonotic in origin and skin lesions infection will typically resolve, whereas L. tropica tends to occur in epidemics driven by anthroponotic transmission, with lesions that are more persistent or reactivate (leishmaniasis recidivans). Another example is cutaneous leishmaniasis in Central and South America where infections caused by L. amazonensis and L. braziliensis can occur in the same areas, but the consequences of infection are potentially very different. L. amazonensis mostly causes simple cutaneous leishmaniasis but can also progress to disseminated cutaneous leishmaniasis with multiple lesions, whereas L. braziliensis will sometimes progress to destructive mucocutaneous disease [6,7].
Although multilocus enzyme electrophoresis (MLEE) of cultured parasites was the best approach for Leishmania species identification for many years [8], it is a time-consuming procedure and requires mass parasite culturing. However, more critically, it has now become apparent that strains with apparently the same enzyme phenotype can have distinct amino acid sequences, and putative heterozygous phenotypes are also difficult to interpret [9]. Therefore, MLEE has become superseded by molecular approaches such as PCR and DNA sequencing, which have the potential to be more sensitive, rapid and easier to perform. Results are available within days versus weeks or months, and in the case of sequencing provide a permanent record available to all researchers. So far, several target genes have been used in PCR and sequencing approaches for Leishmania species identification, such as the SSU-rRNA gene [10], the RNA polymerase II gene [11], the DNA polymerase α gene [12], gp63 genes [13], the cytochrome oxidase II gene [14,15], spliced leader mini-exon genes [16], the glucose-6-phosphate dehydrogenase (G6PD) gene [17], the heat shock protein 70 (HSP70) gene [18], 18S ribosomal RNA [19], the cysteine protease b (cpb) gene [20], the N-acetylglucosamine-1-phosphate transferase gene [21], the internal transcribed spacer 1 (ITS1) region of the SSU-rRNA gene [22,23], the cytochrome b (cyt b) gene [24], 7SL RNA genes [25], and triose-phosphate isomerase (tim) genes [26].
For the species found in Thailand, so far only one PCR-based method, using a pair of primers for minicircle kinetoplast DNA [27], has been used to discriminate L. martiniquensis from 'L. siamensis' (MHOM/TH/2010/TR) (syn. L. orientalis), but the size of the PCR products could not differentiate these parasites from other subgenera [28]. In 2012, Requena and colleagues analyzed the 3 0 untranslated region (3 0 UTR) of Leishmania HSP70-type I (HSP70-I) genes from 24 strains representing eleven Leishmania species following PCR amplification [29]. It was observed that there was sufficient sequence conservation in the primer targets to enable amplification of DNA from species of both the subgenus Leishmania and Viannia. This particular region of the HSP70-I gene appears to have the potential to discriminate between Leishmania subgenera by direct visualization of different sizes of PCR amplification products [29]. The main objective of this study was to use the 3 0 UTR region of the HSP70-I gene to identify and diagnose L. martiniquensis and L. orientalis, thereby expanding the species range to the subgenus Mundinia [30] and including the species found in Thailand. In addition to direct PCR, HSP70-I-3 0 -UTR restriction fragment length polymorphism (HSP70-I-3 0 -UTR PCR-RFLP) was also performed to investigate the potential of this method to identify Leishmania species. In addition, the detection limit of the HSP70-I-3 0 -UTR PCR was compared with that of two other widely used targets: the SSU-rRNA gene and the ITS1 of the rRNA gene.

Ethics statement
The study was approved by the ethics committee of the Faculty of Medicine, Chulalongkorn University (IRB number: 051/64). No patient information is presented in this study.

Parasite isolates and culture
The following species isolates were grown as promastigotes in vitro as previously described [1]

Isolation of DNA
Leishmania species and human blood DNAs were extracted using a QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. DNA concentration was measured using a Nano Drop spectrophotometer (ND-1000 model, Fisher Scientific, Loughborough, UK).

In silico restriction analysis
In silico analyses of predicted BsuRI restriction fragments of the 3 0 -UTR sequences of HSP70-I genes were performed using RestrictionMapper version 3 (www.restrictionmapper.org/) and theoretical fragment sizes were determined manually. Microsatellite distribution in the 3 0 -UTR of HSP70-I genes was analyzed manually.

PCR-RFLP analysis
PCR amplifications targeting the 3 0 -UTR sequences of HSP70-I genes were performed as described above. The PCR products were digested with a restriction enzyme, BsuRI, (Thermo Fisher Scientific Baltics UAB, Vilnius, Lithuania) according to the manufacturer's instruction. BsuRI is an isoschizomer of HaeIII, which has been previously used to analyze Leishmania by PCR-RFLP [16,18], both cutting the target sequence GGCC. Patterns of digestion products were analyzed by 3% agarose gel electrophoresis. The GeneRuler 100 bp DNA ladder (Thermo Fisher Scientific Baltics UAB, Vilnius, Lithuania) was used as a DNA size marker. Gels were stained with GelRed (Thermo Fisher Scientific, Loughborough, UK), and visualized using a GelDoc imaging system (Ultra-Violet Products Ltd., Cambridge, UK).

Detection limits of the PCR method
Ten-fold serial dilutions of the extracted Leishmania DNAs were prepared to generate standard concentrations at 100,000; 10,000; 1,000; 100; 10; 1; 0.1; 0.01; 0.001; and 0.0001 pg/μL. To mimic a real situation of the detection of DNA of parasites extracted from blood or tissue samples, human DNA extracted from blood at a concentration of 18 ng/μL was used as a background DNA and diluent in the 10-fold dilution series of the Leishmania DNA. One microliter of each concentration was used as a DNA template for PCR amplification in a final volume of 25 μL. Each PCR reaction contained 100,000 to 0.0001 pg of Leishmania DNA and 19.8 ng of human DNA. PCR was performed on the 3 0 -UTR sequences of HSP70-I genes as described above and these results compared with two other commonly used identification targets as follows. Leishmania-specific primers R221 (5 0 -GGTTCCTTTCCTTGATTTAGC-3 0 ) and R332, (5 0 -GGCCGGTAAAGGCCGAATAG-3 0 ) were used to amplify a region of the 18S rRNA gene, generating a product of 603 bp [31,32]. Primers LeR (5 0 -CCAAGTCATCCATCGCGACACG-3 0 ) and LeF (5 0 -TCCGCCCGAAAGTTCACCGATA-3 0 ) were used to amplify an internal transcribed spacer 1 region (ITS1) of the rRNA gene, generating a product of 379 bp [22]. To confirm the presence of a standard concentration of background human DNA universal primers UNFOR403 (5'-TGAGGACAAATATCATTCTGAGG-3') and UNREV1025 (5'-GGTTGTC CTCCAATTCATGTTA-3') were used to amplify the human DNA [33]. PCR products were run on 1.5% agarose gels, stained with GelRed (Thermo Fisher Scientific, Loughborough, UK), and visualized using a GelDoc imaging system (Ultra-Violet Products Ltd., Cambridge, UK).

Evaluation of the assay on DNA from clinical samples
To test this PCR assay on DNA extracted directly from clinical samples, six DNA samples extracted from saliva, blood, and skin of L. martiniquensis cases [34,35] were used.

PCR amplification and sequencing of 3 0 -UTR regions of Leishmania HSP70-I genes
The 70-IR-D and 70-IR-M primers were used to amplify the HSP70-I-3 0 -UTR DNA from five previously characterized isolates of Leishmania from northern Thailand. Four of them were L. martiniquensis (LSCM1, LSCM2, LSCM3, LSCM5) and all of these produced a band of~480 bp on agarose gel electrophoresis, similar to the reference isolate LEM2494 (S1 Fig). Another isolate was L. orientalis (LSCM4; reference isolate). It produced a band of~670 bp. The PCR products of the five Thai isolates were sequenced. The sequences of the four isolates of L. martiniquensis were aligned and showed a high degree of similarity with each other and the reference isolate LEM2494 (99.5-100% identity) (S2 Fig).
In addition to these, HSP70-I-3 0 -UTR PCR products were generated from 10 other species of Leishmania: L. aethiopica, L. amazonensis, L. braziliensis, L. donovani, L. guyanensis, L. infantum, L. major, L. mexicana, L. panamensis, and L. tropica. These PCR products were variable in size, but all could be clearly discriminated from the small~480 bp product of L. martiniquensis using gel electrophoresis (Fig 1). The~670 bp product of L. orientalis was also a different size to those of the other species but would be quite difficult to discriminate by electrophoresis alone. Parasites of the subgenus Viannia (L. braziliensis, L. guyanensis, and L. panamensis) produced products of~550-630 bp, whereas the remaining species all in subgenus Leishmania produced larger products of~750-780 bp (Fig 1). To confirm these results, these HSP70-I-3 0 -UTR PCR products were all sequenced, and the exact sizes of these sequences are shown in Table 1, together with their GenBank accession numbers.

The HSP70-I-3 0 -UTR PCR assay on DNA extracted from clinical samples
The HSP70-I-3 0 -UTR PCR was applied to six DNA samples extracted from clinical specimens: three from saliva; two from blood; and one from skin of L. martiniquensis cases [34,35]. PCR products were successfully generated in each case and, as expected, the size of these products from saliva, blood, and skin specimens completely matched with the size of the L. martiniquensis PCR product (Fig 2).

The 3 0 -UTRs of HSP70-I genes contain various microsatellites
One source of size variation in untranslated regions is the occurrence of microsatellites. Since the 3 0 -UTRs of HSP70-I genes in L. martiniquensis, L. orientalis, and L. mexicana have not been analyzed for microsatellites, a search for structural motifs along the sequenced fragments of the three species was performed. The results show that the 3 0 -UTR of HSP70-I fragments contained many microsatellites. The most common microsatellite was CA repeats (as TG or GT on the complementary strand). However, TGC-microsatellites were not found in L. orientalis (Fig 3).

In silico prediction of BsuRI restriction fragments of 3 0 -UTR sequences of HSP70-I genes and PCR-RFLP analysis
To explore the potential for enhanced discrimination between species by PCR-RFLP, in silico analyses of BsuRI restriction fragments of the 3 0 -UTR sequences of HSP70-I genes of L. martiniquensis, L. orientalis and the 10 other species were performed. This revealed a variety of different predicted patterns and fragment sizes (Table 1). We also analyzed L. mexicana for the first time, and the in silico analysis showed different fragment sizes from most other species, but in this respect was similar to L. aethiopica and L. tropica.
To test the in silico prediction, PCR-RFLP analysis targeting the 3 0 -UTR of HSP70-I following by digesting with BsuRI restriction enzyme was performed for the 12 Leishmania species.  Table 1). The PCR-RFLP patterns of L. guyanensis (M4147) and L. panamensis (LS94) were similar and one extra band (>500 bp) was observed differentially from the in silico predicted patterns. Results also confirm that L. mexicana, L. aethiopica and L. tropica had similar PCR-RFLP patterns (Fig 4 and Table 1). Likewise, L. donovani and L. infantum had similar  banding patterns. Some of the fragments that were similar in size were below the resolution limit and could not be discriminated by gel electrophoresis, appearing as merged bands. These are indicated in Table 1.
Of these results perhaps the most surprising is the similarity in banding pattern of L. mexicana to L. aethiopica/L. tropica and difference to L. amazonensis. To confirm this was a genuine result and not due to a mix up of cultures further sequence analysis was performed on three L. amazonensis and two L. mexicana isolates (S3 Fig). This analysis confirmed the close sequence similarity between all five of these isolates and the reason for the banding patterns observed. The two L. mexicana isolate sequences were identical to each other and the three L. amazonensis isolates showed 95-99% identity. All five isolates had a conserved BsuRI site, but due to single nucleotide polymorphisms and insertions, an additional site found in all three L. amazonensis isolates was absent in both L. mexicana isolates, and conversely an additional site in both L. mexicana isolates was absent from all three L. amazonensis isolates. These account for the difference in banding pattern between L. amazonensis and L. mexicana. The analysis shows that the similar sized bands observed in L. mexicana and L. aethiopica/L. tropica is a result of co-incidence rather than overall sequence similarity.

Detection limits of three PCR methods to differentiate between L. martiniquensis and L. orientalis
Finally, for L. martiniquensis and L. orientalis we investigated the detection limit of the HSP70-I-3 0 -UTR PCR method and compared this target with two other widely used sequences, SSU-rRNA and ITS1-rRNA. The HSP70-I-3 0 -UTR method could detect L. martiniquensis and L. orientalis at the same DNA concentration which was 1 pg/μL (Fig 5A). Compared to this, the PCR amplified SSU-rRNA regions showed similar detection limits in both Leishmania species (Fig 5B). The ITS1-rRNA PCR method could detect the DNA of both Leishmania species at a lower concentration of 0.01 pg/μL (Fig 5C). However, note that the sizes of both SSU-rRNA and ITS1-rRNA PCR amplicons could not differentiate between L. martiniquensis and L. orientalis.

Discussion
Species identification in leishmaniasis is important in diagnostics, epidemiology, and clinical studies. To that end, DNA sequencing is an effective method used to identify Leishmania species. However, in some remote areas of Thailand and other low income or developing countries in Southeast and East Asia or elsewhere, it is not a cost-effective method and is not feasible for routine diagnosis, and sending samples elsewhere for sequencing has other problems, especially in transportation. Simple PCR methods would be preferred for diagnosis and identification of the pathogenic parasites and also medical follow-up. The current study demonstrated that PCR of the HSP70-I-3 0 -UTR region was able to generate products that could readily distinguish between L. martiniquensis (480-2 bp) and L. orientalis (674 bp), the two major causative agents of leishmaniasis in Thailand, on the basis of their size alone, providing this is accurately measured. To confirm the application of the HSP70-I-3 0 -UTR PCR method to clinical samples, DNA extracted from saliva, blood, and skin, were used. The positive results indicate that the PCR method can be applied to DNA extracted from clinical samples in general. Although we did not test with DNA extracts from clinical samples of L. orientalis because no materials were available, the sensitivity testing shows that the PCR was able to identify DNA down to 1 pg/uL, which in practice is certainly useful for diagnosis. Such information is clinically useful, as from the cases known to date L. martiniquensis appears capable of causing more serious disease than L. orientalis. However, information on possible intra-specific diversity of L. orientalis HSP70-I-3 0 -UTR is limited as only one L. orientalis isolate [2] was cultured and available for this study. If more L. orientalis isolates are obtained in the future, intraspecific diversity of the region can be investigated. Therefore, PCR alone can be used to confirm a suspected autochthonous case of leishmaniasis in Thailand and could provisionally identify the species.
Analysis of the HSP70-I sequences generated in this and previous studies has shown that the 3 0 -UTR region in Leishmania species of the subgenus Leishmania was clearly the longest. For the subgenus Mundinia, the UTR regions of all species except L. martiniquensis were longer than species of the subgenus Viannia. The UTR region of L. martiniquensis was the shortest. The biological significance of this size variation, if any, is unclear, but it might influence adaptation of the parasites in different environments/hosts as the 3 0 -UTR contains various sequences of regulatory regions that are involved in post-transcription of HSP70 expression. The HSP70 protein assists the Leishmania parasites in survival within host macrophages and adaptation to the new environment following infection and tolerance against different stresses during their life cycle [36]. Further study on the regulation of the HSP70-I expression in the new species in the subgenus Mundinia should be performed, especially L. martiniquensis, as it can be a pathogen in various hosts in several regions of the world [1,[37][38][39].
Microsatellites are repeated motifs of about one to six non-coding nucleotides ubiquitously distributed in all eukaryotic genomes [40]. Microsatellites of the HSP70-I-3 0 -UTR fragments of L. martiniquensis, L. orientalis, and L. mexicana were reported here for the first time. The most common microsatellite found, CA repeats, was observed in the 3 0 -UTR of all studied species. This is consistent with the reports in L. infantum, L. major and L. braziliensis genomes [29,41]. However, another abundant repeat in Leishmania genomes, TGC-microsatellite [41], was absent in L. orientalis, as also reported in the species of the subgenus Viannia [29]. Many other microsatellites such as C, GC, GTG, GGC, and GCG that might be categorized as uncommon for Leishmania [29] were noted in all samples of Leishmania species studied. In this study, size variants of the 3 0 -UTR fragments were partially due to a consequence of variations in the repeat number of the microsatellite sequences, the longer UTRs having a greater number of microsatellites. Although not seen within the isolate sequences examined in this study, such variations could also contribute to infraspecific variations in PCR product sizes.
Regarding species found outside Thailand, the HSP70-I-3 0 -UTR PCR generated products for L. martiniquensis and L. orientalis that were different in size from other tropical Leishmania as confirmed by sequencing. However, by gel electrophoresis alone the sizes of these PCR products could not reliably identify them, in particular L. orientalis. Therefore, PCR-RFLP analyses were carried out to attempt an improved level of discrimination. PCR-RFLP methods have been reported to be sometimes useful in providing improved Leishmania species discrimination compared to PCR alone, for example, using HaeIII restriction fragments of a conserved region of the Leishmania minicircle kinetoplast DNA to distinguish L. braziliensis from L. amazonensis species [42]. Similarly, PCR-RFLP assays of ITS1 genes were used for direct identification of L. major and L. tropica [43]. In the current study the results of PCR-RFLP analyses in general corresponded to the prediction by in silico analyses of BsuRI restriction fragments of the HSP70-I-3 0 -UTR amplified region. The PCR-RFLP patterns could clearly distinguish L. martiniquensis, L. orientalis, L. braziliensis, L. amazonensis and L. major species. Although different to the other species, the PCR-RFLP still could not distinguish two pairs of However, previous studies have reported that PCR-RFLP patterns of the HSP70 gene fragments digested with BccI restriction enzyme can differentiate between L. guyanensis and L. panamensis if required [44,45]. In this study, an unexpected band of >500 bp in both L. guyanensis and L. panamensis PCR-RFLP patterns was noted. The origin of this band is unclear but may have resulted from sequence variation in the HSP70-I-3 0 -UTR from these species. For L. mexicana, its pattern was similar to L. aethiopica and L. tropica but different from other species in the subgenus Viannia and Leishmania, including L. amazonensis, which is otherwise generally considered closely related to L. mexicana. However, in this case the results are due to coincidence rather than any underlying relationship between L. mexicana and L. aethiopica or L. tropica. Overall, the combination of the PCR and PCR-RFLP analysis enabled clear identification of the two species that cause leishmaniasis in Thailand and discrimination of these from species found elsewhere in the world.
In Thailand, Sriworarat et al. (2015) have demonstrated a colorimetric loop-mediated isothermal amplification (LAMP) technique for the direct detection of Leishmania DNA. The LAMP assay could also detect DNA from multiple Leishmania species other than 'L. siamensis (MHOM/TH/2010/TR)' and L. martiniquensis, including L. aethiopica, L. braziliensis, L. donovani and L. tropica [46]. However, LAMP is prone to contamination due to the large amount of DNA that it can generate, and its capability to amplify minute amounts of DNA [46]. One PCR-based method using a pair of primers for minicircle kinetoplast DNA gene [27] can be used to discriminate L. martiniquensis from 'L. siamensis' (MHOM/TH/2010/TR) but the size of PCR products could not differentiate parasites in other subgenera [28]. For the ITS1-PCR method, it is not suitable for the discrimination of 'L. siamensis' (syn L. orientalis) and L. martiniquensis infection as it generates the same size of products [1,28,47].
The results of the detection limits in this study showed that the ITS1-rRNA PCR was the most sensitive method which could detect DNA of L. martiniquensis and L. orientalis at a concentration of 0.01 pg/μL compared to the HSP70-I-3 0 -UTR PCR and SSU-rRNA PCR methods which could detect DNAs of both species at the same DNA concentration (1 pg/μL). This can be explained in that the HSP70 genes in Leishmania species studied so far are all arranged in a single genomic cluster that contains five or six HSP70-I copies followed by one HSP70-II copy [48], whereas copies of the SSU rRNA gene and the ITS1 region of the rRNA gene have been observed at between 20-40 copies per cell in Leishmania species [31,49].
The PCR-based methods used in this study can now be applied to the identification of Leishmania species obtained from vectors and reservoirs in Thailand to investigate their epidemiological significance. The technique is simple to perform and can be implemented in all settings where PCR-RFLP is available. For definite autochthonous cases, the PCR product alone may be sufficient for identification. However, where an imported case is suspected or cannot be eliminated, the L. orientalis PCR product is quite similar in size to that of L. panamensis and L. guyanensis, in which case the BsuRI-PCR-RFLP method can be used for differentiating these species and some other species in other subgenera. However, where identification of species is essential or the infection is likely to have been acquired outside Thailand, sequencing of the HSP70-I-3 0 -UTR product or a similar discriminating target sequence is recommended. In conclusion, our data show that the 3 0 -UTR of HSP70-I region is a suitable target for PCRbased identification of and discrimination between L. martinquensis and L. orientalis.