Differentiation of Leishmania (L.) infantum, Leishmania (L.) amazonensis and Leishmania (L.) mexicana Using Sequential qPCR Assays and High-Resolution Melt Analysis

Leishmania protozoa are the etiological agents of visceral, cutaneous and mucocutaneous leishmaniasis. In specific geographical regions, such as Latin America, several Leishmania species are endemic and simultaneously present; therefore, a diagnostic method for species discrimination is warranted. In this attempt, many qPCR-based assays have been developed. Recently, we have shown that L. (L.) infantum and L. (L.) amazonensis can be distinguished through the comparison of the Cq values from two qPCR assays (qPCR-ML and qPCR-ama), designed to amplify kDNA minicircle subclasses more represented in L. (L.) infantum and L. (L.) amazonensis, respectively. This paper describes the application of this approach to L. (L.) mexicana and introduces a new qPCR-ITS1 assay followed by high-resolution melt (HRM) analysis to differentiate this species from L. (L.) amazonensis. We show that L. (L.) mexicana can be distinguished from L. (L.) infantum using the same approach we had previously validated for L. (L.) amazonensis. Moreover, it was also possible to reliably discriminate L. (L.) mexicana from L. (L.) amazonensis by using qPCR-ITS1 followed by an HRM analysis. Therefore, a diagnostic algorithm based on sequential qPCR assays coupled with HRM analysis was established to identify/differentiate L. (L.) infantum, L. (L.) amazonensis, L. (L.) mexicana and Viannia subgenus. These findings update and extend previous data published by our research group, providing an additional diagnostic tool in endemic areas with co-existing species.


Introduction
Leishmaniasis is caused by many Leishmania species belonging to subgenera Leishmania (Leishmania) and Leishmania (Viannia), creating a global public health problem with 0.2-0.4 million cases of visceral leishmaniasis (VL) and 0.7-1.2 million cases of cutaneous leishmaniasis (CL) per year [1]. In specific 1 not available.

ITS1-PCR RFLP
The L. (L.) mexicana strains and clinical samples were typed using ITS1-PCR RFLP according to Monroy-Ostria et al. [22]. Briefly, the PCR was performed using primers LITSR and L5.8S, following the amplification protocol-94 • C for 4 min followed by 36 cycles of 94 • C for 40 s, 54 • C for 30 s and 72 • C for 1 min and a final extension at 72 • C for 6 min. PCR products were nested using the same PCR conditions for 18 cycles. PCR products were digested with HaeIII for 3 h at 37 • C and for 20 min at 80 • C to inactivate the enzyme. The restriction fragments were subjected to electrophoresis on a 4% agarose gel.

DNA Sequencing and Phylogenetic Analysis
The alanine aminotransferase (ALAT) gene was amplified in clinical samples px2, px3, px9, px10, pxJLC and in L. (L.) mexicana MHOM/MX/2011/Lacandona according to Marco et al. [23] using primers ALAT.F and ALAT.R. The amplification conditions were-94 • C for 3 min followed by 40 cycles of 94 • C for 30 s, 55 • C for 30 s and 72 • C for 30 s. PCR products were purified using the Agencourt AMPure XP kit (Beckman Coulter, Brea, CA, USA) and sequenced using the BigDye Terminator v3.1 Cycle Sequencing Kit (Thermo Fisher Scientific, Waltham, MA, USA) on ABI 3730xL DNA Analyzer (Thermo Fisher Scientific, Waltham, MA, USA). Chromatograms were visualized with ApE software and consensus sequences were generated and compared between them and with other validated species of L. (L.) mexicana deposited in GenBank using the Blastn tool available in the same platform. A phylogenetic reconstruction based on the Maximum Likelihood (ML) method was generated and a phylogenetic tree was constructed with 10,000 bootstrap replications, using the close-neighbor interchange method in Mega 6.0.

qPCR Assays
The qPCR-ML (amplifying kDNA minicircle subclass more represented in L. (L.) infantum) and qPCR-ama (amplifying kDNA minicircle subclass more represented in L. (L.) amazonensis) were performed as previously described [18]. The new assay qPCR-ITS1 was performed using the new primers ITS1mexama_F (5′-GGATCATTTTCCGATGATTACACC-3′) and ITS1mexama_R (5′-CTGCAAATGTTGTTTTTGAGTACA-3′), flanking a portion of ITS1 sequence containing differences between L. (L.) amazonensis and L. (L.) mexicana (Figure 1). The primers were designed using Primer BLAST and were verified against the ITS1 sequences of L. (L.) amazonensis (n = 32) and L. (L.) mexicana (n = 30) encompassing forward and/or reverse primers, available in the Genbank database. For all assays, PCR reactions were carried out in a 25 μL volume with 1-3 μL of template DNA using SYBR green PCR master mix (Diatheva srl, Fano, Italy) or TB Green premix ex TaqII Mastermix (Takara Bio Europe, France) and 200 nM of each primer in a Rotor-Gene 6000 instrument (Corbett Life Science, Mortlake, Australia). The amplification conditions were-94 °C for 10 min, 40 cycles at 94 °C for 25 s, 60 °C for 20 s and 72 °C for 20 s. At the end of each run, a melting curve analysis was performed from 78 to 92 °C with a slope of 1 °C/s, and 5 s at each temperature. The reactions were performed in duplicate. Dilution curves (from 1.0 to 1 × 10 −5 ng/reaction) were established using L. (L.) mexicana MHOM/MX/2011/Lacandona DNA for qPCR-ML, qPCR-ama and qPCR-ITS1. The threshold cycles (Cq) were determined using the quantitation analysis of the Rotor-Gene 6000 software, setting a threshold to 0.15. To evaluate the potential interference of host DNA as a background in the qPCR analysis, 30 ng of human DNA was spiked in the reaction tubes.

High-Resolution Melt (HRM) Analysis
The qPCR-ML, qPCR-ama and qPCR-ITS1 amplicons were analyzed by HRM protocol on a Rotor-Gene 6000 instrument as described previously [24] with few modifications. Briefly, HRM was carried out over the range from 80 to 90 °C (qPCR-ML, qPCR-ama) or 75 to 85 °C (qPCR-ITS1), rising at 0.1 °C/s and waiting for 2 s at each temperature. Each sample was run in duplicate, and the gain was optimized before melting on all tubes.

Ethics Approval
This study was conducted according to the principles expressed in the Declaration of Helsinki.  For all assays, PCR reactions were carried out in a 25 µL volume with 1-3 µL of template DNA using SYBR green PCR master mix (Diatheva srl, Fano, Italy) or TB Green premix ex TaqII Mastermix (Takara Bio Europe, France) and 200 nM of each primer in a Rotor-Gene 6000 instrument (Corbett Life Science, Mortlake, Australia). The amplification conditions were-94 • C for 10 min, 40 cycles at 94 • C for 25 s, 60 • C for 20 s and 72 • C for 20 s. At the end of each run, a melting curve analysis was performed from 78 to 92 • C with a slope of 1 • C/s, and 5 s at each temperature. The reactions were performed in duplicate. Dilution curves (from 1.0 to 1 × 10 −5 ng/reaction) were established using L. (L.) mexicana MHOM/MX/2011/Lacandona DNA for qPCR-ML, qPCR-ama and qPCR-ITS1. The threshold cycles (Cq) were determined using the quantitation analysis of the Rotor-Gene 6000 software, setting a threshold to 0.15. To evaluate the potential interference of host DNA as a background in the qPCR analysis, 30 ng of human DNA was spiked in the reaction tubes.

High-Resolution Melt (HRM) Analysis
The qPCR-ML, qPCR-ama and qPCR-ITS1 amplicons were analyzed by HRM protocol on a Rotor-Gene 6000 instrument as described previously [24] with few modifications. Briefly, HRM was carried out over the range from 80 to 90 • C (qPCR-ML, qPCR-ama) or 75 to 85 • C (qPCR-ITS1), rising at 0.1 • C/s and waiting for 2 s at each temperature. Each sample was run in duplicate, and the gain was optimized before melting on all tubes.

Ethics Approval
This study was conducted according to the principles expressed in the Declaration of Helsinki. Authorities were strictly followed. All patients received treatment and clinical care by health authorities and signed a written informed consent for the collection of samples and subsequent analysis.

Statistical Analysis
Statistical analysis was performed with GraphPad Prism 5.0 (GraphPad Software, San Diego, CA, USA). Normal distribution of data was assessed by D Agostino and Pearson omnibus normality test (alpha = 0.05). Difference between Tm mean values was evaluated using the nonparametric Mann-Whitney test.  (Table 3). Table 3. qPCR-ML and qPCR-ama results in strains/clinical isolates.  (Table 3). In the qPCR-ML, the presence of 30 ng of purified human DNA delayed the limit of detection to 1.0 × 10 −1 ng ( Figure S1). With regard to qPCR-ama, the efficiency and detection limit were evaluated using 10-fold L. (L.) mexicana MHOM/MX/2011/Lacandona DNA serial dilutions (from 1.0 to 1×10 −5 ng) in three independent experiments. There was a linear correlation between the log of DNA concentration and Cq value (slope = −3.3909, R 2 = 0.9716) with a reaction efficiency of 97%. In order to evaluate the interference of host DNA, the DNA dilutions were spiked with 30 ng of purified human DNA, showing a delay on the Cq values but with comparable efficiency and limit of detection ( Figure 2). The efficiency and detection limit obtained with L. (L.) mexicana DNA were in agreement with previous results obtained using DNA template from L. (L.) amazonensis [18].

Results
The qPCR-ML/qPCR-ama approach was also applied to 11 clinical samples. These samples were characterized as L. (L.) mexicana by ITS1-PCR RFLP ( Figure S2), with the exception of pxCMU, for which a digestion profile could not be obtained. Moreover, the genotype of five clinical samples (px2, px3, px9, px10, pxJLC) were further confirmed as L. (L.) mexicana by sequencing and phylogenetic analysis of the alanine aminotransferase (ALAT) gene ( Figure S3). All samples showed Cq qPCR-ama <Cq qPCR-ML (Table 4) (Figure 2). The efficiency and detection limit obtained with L. (L.) mexicana DNA were in agreement with previous results obtained using DNA template from L. (L.) amazonensis [18]. The qPCR-ML/qPCR-ama approach was also applied to 11 clinical samples. These samples were characterized as L. (L.) mexicana by ITS1-PCR RFLP ( Figure S2), with the exception of pxCMU, for which a digestion profile could not be obtained. Moreover, the genotype of five clinical samples (px2, px3, px9, px10, pxJLC) were further confirmed as L. (L.) mexicana by sequencing and phylogenetic analysis of the alanine aminotransferase (ALAT) gene ( Figure S3). All samples showed Cq qPCR-ama <Cq qPCR-ML (Table 4) (Table S1). However, three clinical samples failed to amplify (Px7, PxGSF, PxCMU). Overall, the qPCR-ITS1 HRM assay for amazonensis/mexicana species discrimination showed 84.2% sensitivity and 100% specificity.  (Table S1). However, three clinical samples failed to amplify (Px7, PxGSF, PxCMU). Overall, the qPCR-ITS1 HRM assay for amazonensis/mexicana species discrimination showed 84.2% sensitivity and 100% specificity.  (n = 6). Line within the box represents the median and the red dots above and below the whiskers represent the outliers that are either greater than 95th or less than 5th percentile. ** p < 0.01.    6). Line within the box represents the median and the red dots above and below the whiskers represent the outliers that are either greater than 95th or less than 5th percentile. *** p < 0.001.

Discussion
The identification of Leishmania species is an important diagnostic aspect, especially in Latin America, not only for epidemiological studies but also for the accurate monitoring of clinical disease evolution. In fact, the only species causing VL in this geographical region is L. (L.) infantum (syn.   ) amazonensis (n = 6). Line within the box represents the median and the red dots above and below the whiskers represent the outliers that are either greater than 95th or less than 5th percentile. *** p < 0.001.

Discussion
The identification of Leishmania species is an important diagnostic aspect, especially in Latin America, not only for epidemiological studies but also for the accurate monitoring of clinical disease evolution. In fact, the only species causing VL in this geographical region is L. (L.) infantum (syn. Line within the box represents the median and the red dots above and below the whiskers represent the outliers that are either greater than 95th or less than 5th percentile. *** p < 0.001.

Discussion
The identification of Leishmania species is an important diagnostic aspect, especially in Latin America, not only for epidemiological studies but also for the accurate monitoring of clinical disease evolution. In fact, the only species causing VL in this geographical region is L. (L.) infantum (syn. chagasi), while cutaneous or mucocutaneous (MCL) manifestations can also be generated by Viannia subgenus and L. (L.) mexicana complex. In this epidemiological and clinical context, the species discrimination appears pivotal, e.g., to monitor a cutaneous lesion that could evolve in VL, MCL or disseminated CL, depending on the species. In this view, molecular diagnostic tools allowing species discrimination can be helpful. The kDNA minicircles are ideal targets for highly sensitive molecular detection of Leishmania spp. since they are present in thousands of copies per cell [25]. Since the pioneering work of Nicolas et al. [26], many qPCR assays have been designed on conserved regions of minicircles to detect/quantify Leishmania parasites. Moreover, several authors investigated the possibility to exploit minicircle sequences to discriminate Leishmania parasites at the species level, reaching only partial results due to the variability of minicircle subclasses [15,16]. Previously, we proposed an SYBR Green qPCR-based approach to distinguish L. (L.) infantum from L. (L.) amazonensis, exploiting the different abundance of minicircle subclasses rather than targeting a species-specific sequence. Using this approach, which relies on two qPCR assays (qPCR-ML and qPCR-ama) and evaluation of Cq values, we were able to distinguish the two species adequately [18].
In this work, we tested this approach with L. (L.) mexicana, which is genetically close to L. (L.) amazonensis. The comparison of Cq values of qPCR-ML and qPCR-ama confirmed results previously obtained with L. (L.) amazonensis, allowing us to include L. (L.) mexicana among the Leishmania (Leishmania) species that can be differentiated from L. (L.) infantum, therefore extending the conclusion of our previous work. Importantly, this approach was successfully applied to cutaneous lesions of 11 patients diagnosed with diffuse or localized cutaneous leishmaniasis. Notably, the clinical sample pxCMU, which was negative in ITS1-PCR RFLP, was identified as L. (L.) mexicana/amazonensis, evidencing the highest sensitivity of our qPCR assays targeting minicircles. These results further support the possibility of exploiting the relative abundance of minicircles for Leishmania species discrimination. Moreover, we confirmed the use of an adequate diagnostic approach based on consecutive qPCR assays to define species [18], as also proposed by other authors [27].
The distinction between L. (L.) amazonensis and L. (L.) mexicana is important for epidemiological studies and disease monitoring, but it can be challenging [28]. For instance, hsp70 analysis by Fraga et al. [29] did not resolve between these species. On the other hand, other authors were able to separate these species based on multilocus sequence typing (MLST) [30] or sequential real-time PCR assays [27].
The qPCR coupled with HRM analysis is considered as a good option in molecular diagnostics since it avoids the use of modified oligonucleotides, it is accurate, allows high-throughput applications and is faster and cheaper than other types of analysis such as MLST, RFLP or single-gene DNA sequencing. Moreover, since the qPCR is a closed-tube system, the potential for carryover contamination will be reduced. In the attempt to discriminate between L. (L.) amazonensis and L. (L.) mexicana, HRM profiles of amplicons from qPCR-ama were investigated; however, their heterogeneity did not us allow to distinguish these two species reliably. Since Schönian et al. demonstrated the possibility to discriminate the two species using ITS1-PCR RFLP [31], we designed an HRM-based assay exploiting differences in L. (L.) amazonensis and L. (L.) mexicana ITS1 sequences, in order to avoid restriction digestion and electrophoretic analysis. This process allows saving a considerable amount of time to perform the analysis and avoids possible difficulties in restriction fragment identification. As expected from the in silico sequence analysis, the observed HRM Tm values of all L. (L.) mexicana samples were significantly higher as those of all L. (L.) amazonensis samples, allowing a robust distinction between these two species. The fact that three clinical samples did not amplify (Px7, PxGSF, PxCMU) was probably due to the lower sensitivity of qPCR-ITS1 as compared to the assay targeting kDNA minicircles.

Conclusions
In the attempt to use a qPCR-based approach to differentiate Leishmania species co-existing in the New World, sequential qPCR assays and HRM analysis have been implemented. The results showed that-(i) L. (L.) infantum can be distinguished from L. (L.) mexicana comparing the Cq values of qPCR-ML and qPCR-ama, as previously shown for L. (L.) amazonensis; (ii) this distinction was possible not only using strains/isolates but also in clinical samples; (iii) the differentiation between L. (L.) amazonensis and L. (L.) mexicana was achieved by qPCR-ITS1 HRM analysis. Therefore, it was possible to design/update an algorithm that allows us to identify/differentiate L. (L.) infantum, L. (L.) amazonensis, L. (L.) mexicana and Viannia subgenus with sequential qPCR assays coupled with HRM analysis targeting minicircle kDNA and ITS1 sequence (Figure 6), which further extends our previous work.  Supplementary Materials: The following are available online at www.mdpi.com/2076-2607/8/6/818/s1, Figure  S1: Electrophoretic analysis of qPCR-ML products, Figure S2: Digestion of ITS1 amplicons of clinical samples and reference Leishmania strains with the restriction endonuclease HaeIII, Figure S3: Maximum likelihood phylogenetic tree of ALAT amplicons, Figure S4: HRM analysis of qPCR-ama amplicons, Table S1: