Background

Sand flies (Psychodidae: Phlebotominae) are insects of paramount medical and veterinary significance, mainly due to their competence in transmitting pathogens to many animal species, including humans [1]. Plant-derived carbohydrates (e.g. nectar, honeydew and phloem sap) are part of the diet of both male and female sand flies, being an essential source of energy for various activities, including flight and reproduction [2]. However, adult females are also blood-feeders and require vertebrate blood as a source of protein for oogenesis [3]. Thus, the feeding behaviour of female sand flies may influence their likelihood of acquiring and transmitting pathogens, including Leishmania parasites (Kinetoplastida: Trypanosomatidae), the causative agents of leishmaniases. To date, over 1000 sand fly species have been identified globally, of which more than 50% are exclusively found in the Neotropics [4]. About 98 species of sand flies have been listed as proven or suspected vectors of Leishmania spp. [1].

Leishmaniases are among the top ten neglected tropical diseases causing high levels of morbidity and mortality in endemic areas, mainly in tropical and subtropical regions of the world [5]. Brazil, India, Bangladesh, Sudan, South Sudan and Ethiopia concentrate 90% of the global incidence of visceral leishmaniasis (VL), whereas Afghanistan, Algeria, Colombia, Brazil, Iran, Syria, Ethiopia, Sudan, Costa Rica and Peru concentrate ~ 75% of the global incidence of cutaneous leishmaniasis (CL) [5].

An important factor influencing the dynamics of Leishmania parasite transmission in endemic areas is the feeding behaviour of female sand flies. For example, to be considered a ‘good vector’ for zoonotic Leishmania parasites, females from a given sand fly species should feed frequently on the reservoir host(s) and on humans as well. Therefore, investigations of the blood meals of various species of sand flies are crucial towards a better assessment of the risk of Leishmania transmission in areas where leishmaniasis are endemic. Several methodologies to identify blood meal in sand flies have been used, including enzyme-linked immunosorbent assay (ELISA), mass spectrometry, precipitin test and polymerase chain reaction (PCR) [6,7,8,9,10,11]. More recently, quantitative real-time PCR [12, 13] and PCR followed by amplicon sequencing [14] demonstrated promising results, with high-level sensitivity. Although all these methods generated important information about the feeding behaviour of sand flies, they may present several drawbacks, such as low sensitivity and specificity (e.g. ELISA and precipitin test) and high cost (e.g. mass spectrometry and PCR followed by amplicon sequencing) [12, 15, 16].

In this context, we developed a fast multiplex real-time PCR assay for simultaneous detection of dog and human blood meals and Leishmania parasites in sand flies, with high analytical sensitivity and specificity, as well as relatively low cost.

Methods

Blood samples and Leishmania parasites

Venous blood samples (2 ml) were withdrawn from a dog and a human in EDTA tubes (Vacuette K3EDTA tube; Greiner Bio-One, Kremsmünster, Austria) and frozen at − 20 °C until DNA extraction. Reference strains of Leishmania infantum (MHOM/BR/76/M4192) and Leishmania braziliensis (MHOM/BR/1975/M2903) were obtained from the Leishmaniasis Reference Service of the Aggeu Magalhães Institute (Fiocruz-PE, Recife, Brazil). Both blood samples and Leishmania parasites were used for the preparation of standard curves (see below). Additionally, two non-engorged females of Migonemyia migonei obtained from a laboratory colony established in the Aggeu Magalhães Institute (Fiocruz-PE) were used as negative controls.

Nucleic acid extraction

DNA extraction from sand flies, blood samples and Leishmania parasites was performed using DNeasy Blood & Tissue kit (Qiagen, Hilden, Germany), according to the manufacturer’s instructions. All samples were eluted in 100 µl of buffer AE (10 mM Tris Cl, 0.5 mM EDTA, pH 9.0), properly labelled and frozen at − 20 °C. The quantity and purity (absorbance ratio at 260/280 nm and at 260/230 nm) of the extracted DNA were accessed using a NanoDrop Lite spectrophotometer (Thermo Fisher Scientific, Waltham, USA).

Primer and probe design

Primers and TaqMan hydrolysis probes (Table 1) targeting dog and human DNA were developed based on the sequences of the mitochondrial cytochrome c oxidase subunit 1 (cox1) and cytochrome b (cytb) genes, respectively, available from GenBank (accession numbers: NC_002008.4 and NC_012920.1), using the Primer3 v.0.4.0 (http://bioinfo.ut.ee/primer3-0.4.0/primer3/). Primers and probes were designed following the instructions of the TaqMan Multiplex Optimization User Guide [17] for optimum assay efficiency. In particular, primers should have a GC content of 40–60% and generate amplicons of 50–150 bp. The melting temperature (Tm) should be similar for all primers; the Tm of the probes be ~ 10 °C higher than the Tm of the primers. With that in mind, primers (18–20 bp) and probes (13–25 bp) were designed to have Tm of 58–60 °C and 68–70 °C, respectively (Table 1).

Table 1 Primers and TaqMan probes used in the singleplex and multiplex real-time PCR assays
Full size table

To avoid non-specific amplification, primers were submitted to BLAST/n of the National Centre for Biotechnology Information (NCBI) to verify its specificity. Furthermore, the formation of dimmers, hairpins, and Tm were assessed with the OligoAnalyzer 3.1 software (https://eu.idtdna.com/calc/analyzer). The primers LEISH-1 and LEISH-2 and a TaqMan probe (Table 1) were used to detect a 120 bp fragment of the Leishmania minicircle kinetoplast DNA (kDNA) [18].

Optimization of singleplex real-time PCR assays

Before optimizing the multiplex real-time PCR assay, singleplex real-time PCR assays were optimized to specifically detect dog and human DNA. Additionally, a singleplex real-time PCR assay targeting kDNA was performed as described elsewhere [19]. A dilution matrix was made to determine optimal concentration of primers and probes [17]. The reaction mixture contained 1.35 µl of each primer (final concentration of 900 nM each), 0.3 µl of the probe (200 nM), 2.5 µl of water (DNAse and RNAse free), 7.5 µl of TaqMan Fast Advanced Master Mix (Thermo Fisher Scientific) and 2-µl sample DNA, in a final volume of 15 μl. Positive controls consisted DNA extracted from dog and human blood and from cultured promastigotes of L. infantum, whereas a master mix without DNA and DNA from unfed female sand flies were used as no template control (NTC) and negative control, respectively. The real-time PCR thermal conditions were as follows: 20 s at 95 °C followed by 35 cycles of 1 s at 95 °C and 20 s at 60 °C (estimated running time: 33.5 min). All singleplex real-time PCRs were performed on a QuantStudio 5 Real-Time PCR system (Thermo Fisher Scientific), with automatic baseline and threshold settings. The reactions were performed in triplicate, with inconsistent or undetermined results between the replicates being regarded as negative.

Optimization of the multiplex real-time PCR assay

Sets of primers and probes targeting dog cox1, human cytb and Leishmania kDNA were multiplexed by labelling each probe with a different dye (Table 1). The reaction mixture consisted of 0.675 µl of each primer (900 nM), 0.15 µl of each probe (200 nM), 7.5 µl of TaqMan® Fast Advanced Master Mix (Applied Biosystems) and 3-µl sample DNA, in a final volume of 15 μl. Positive controls consisted of mixed DNA extracted from dog and human blood and cultured promastigotes of L. infantum. Negative controls and thermal cycling conditions employed in the multiplex assay were the same used in the singleplex assays. All the three channels (reporter and quencher) for VIC-QSY, NED-NFQ/MGB and FAM-NFQ were selected. All multiplex real-time PCR assays were run on a QuantStudio 5 Real-Time PCR system (Thermo Fisher Scientific), with automatic baseline and threshold settings. The reactions were performed in triplicate and repeated three times, with inconsistent or undetermined results between the replicates being regarded as negative.

Specificity, sensitivity, linearity and reproducibility

An in silico analysis of the specificity of the primers and probes was checked using program BLAST/n. The analytical specificity was assessed through cross-tests between the targets (dog, human, L. braziliensis and L. infantum) and unfed female sand flies. Standard curves were prepared using nine serial dilutions (108, 107, 106, 105, 104, 103, 102, 101 and 100 fg per reaction) of DNA from dog, human and L. infantum to assess the analytical sensitivity (detection limit) of the assays. The analytical sensitivity was defined as the lowest amount of DNA detectable in a given assay. The cut-off point of an assay was defined as the quantification cycle (Cq) value corresponding to the detection limit [20].

The amplification efficiency (E) was calculated using the slope of the regression line in the standard curve through the equation: E = 10(−1/slope) – 1. A slope close to − 3.33 was considered optimal. The correlation coefficient (R2) value was automatically calculated using measure of the closeness of fit between the regression line and the individual Cq data points of the standard reactions [21]. The y-intercept value also automatically calculated and corresponds to the theoretical Cq value for a single copy of the target molecule. In intra-assays, triplicates were made in the same plate, whereas in inter-assay, triplicates were repeated in three independent assays performed in three different days within a week.

Assay of field-collected sand flies

A total of 97 engorged female sand flies collected in the context of a previous study [22] were tested individually by the newly developed multiplex real-time PCR. These females belonged to two species: M. migonei (n = 95); and Nyssomyia intermedia (n = 2). Details regarding sand fly collection, identification and processing are described elsewhere [22]. Females were collected both indoors (n = 23) and outdoors (n = 74).

Data analysis

Real-time PCR results were analysed using QuantStudio Design and Analysis Software 1.3.1 (Thermo Fisher Scientific). To assess intra- and inter-assay reproducibility, the percent coefficient of variation (% CV) was calculated for each set of triplicate reactions. The positivity rates for Leishmania parasites in female sand flies collected indoors and outdoors was compared using Fisher’s exact test, considering P < 0.05 as statistically significant. Statistical analysis and calculations were performed using GraphPad Prism 5.0 software (GraphPad Software Inc., CA, USA).

Results

Specificity, sensitivity and linearity of the singleplex and multiplex assays

The sets of primers and probes specific for dog cox1, human cytb and Leishmania kDNA detected only the expected target. Moreover, they did not produce any non-specific amplification in the cross-testing with non-target DNA, or with no template and negative controls. Likewise, there were no false positives due to cross-talk between dye signals from each assay. Thus, analytical specificity of the assays was considered to be 100%.

The analytical sensitivity of the singleplex real-time PCR assays for dog cox1 and human cytb was 1000 fg, with Cq values of 34.2 ± 0.4 and 33.2 ± 0.1, respectively. Conversely, the analytical sensitivity for Leishmania kDNA was 1 fg (Cq = 33.5 ± 0.1) (Table 2).

Table 2 Analytical sensitivity and corresponding threshold cycle (Cq) values from singleplex and multiplex real-time PCR assays for each target
Full size table

The linear regression analysis of standard curves confirmed linearity of the singleplex real-time PCR assays for dog cox1 (R2 = 0.999, E = 101.9, slope = − 3.28, y-intercept = 44.1), human cytb (R2 = 0.999, E = 108.9, slope = − 3.12, y-intercept = 42.8) and Leishmania kDNA (R2 = 0.991, E = 103.5, slope = − 3.24, y-intercept = 35.2) (Fig. 1).

Fig. 1
figure 1

Amplification plots (left) and standard curves (right) of the singleplex real-time PCR assays, showing values of slope, correlation coefficient (R2), efficiency (ε) and y-intercept (y) for each target (dog cox1, human cytb and Leishmania kDNA). All DNA samples were tested in triplicate and curves below the threshold line are negative

Full size image

Similar results were found with the multiplex real-time PCR assay, the detection limits for dog cox1 and human cytb being 1000 fg, with Cq values of 33.1 ± 1.0 and 30.6 ± 0.3, respectively. The detection limit for Leishmania kDNA was 1 fg (Cq = 33.1 ± 0.3) (Table 2). Similarly, the linear regression analysis of standard curves confirmed linearity of the multiplex real-time PCR assay for dog cox1 (R2 = 0.996, E = 109.8, slope = − 3.11, y-intercept = 42.5), human cytb (R2 = 0.993, E = 116.0, slope = − 2.99, y-intercept = 39.8) and Leishmania kDNA (R2 = 0.998, E = 109.9, slope = − 3.10, y-intercept = 33.6) (Fig. 2).

Fig. 2
figure 2

Standard curve (top) and amplification plot (bottom) of the multiplex real-time PCR assay, showing values of slope, correlation coefficient (R2), efficiency (ε) and y-intercept (y) for each target (dog, human and Leishmania DNA). All DNA samples were tested in triplicate

Full size image

Reproducibility of multiplex real-time PCR assay

The variability of the multiplex real-time PCR assay was assessed between and within runs based on standard curves. The coefficients of variation values of intra- and inter- assays were in the range of 0.16–4.01% (Table 3) and 0.92–7.44% (Table 4), respectively. Overall, the difference between Cq values intra- and inter-assay was ≤ 2, suggesting that the multiplex real-time PCR assay is reliable.

Table 3 Intra-assay reproducibility of the multiplex real-time PCR assay
Full size table
Table 4 Inter-assay reproducibility of the multiplex real-time PCR assay
Full size table

Evaluation of multiplex real-time PCR assay with field-collected sand flies

Fifty out of 97 (51.6%) engorged female sand flies tested by the multiplex real-time PCR assay were positive. All positive females belonged to the species M. migonei.

Forty (80.0%) females were positive for one target (i.e. human cytb, dog cox1 or Leishmania kDNA) and 10 (20.0%) for two targets. Among females that were positive for one target (n = 40), 85.0% were positive for human cytb, 10.0% for Leishmania kDNA and 5.0% for dog cox1. Among those positive for two targets (n = 10), 80.0% were simultaneously positive for human cytb and Leishmania kDNA and 20.0% were simultaneously positive for human cytb and dog cox1.

Out of 12 Leishmania-positive females, eight were fed on humans and none were fed on dogs. Four positive females were collected indoors and eight outdoors (Fisher’s exact test, P = 0.4704).

Discussion

In this study, we were interested in developing a tool that could generate information about role of sand flies collected inside human houses and surrounding areas in the transmission of Leishmania parasites to dogs and humans. Hence, we developed a TaqMan-based fast multiplex quantitative real-time PCR assay for the simultaneous detection of dog cox1, human cytb and Leishmania kDNA in female sand flies. The addition of multiple primers and probes in a single reaction as well as changes in the number of cycles and annealing temperature can affect the specificity, sensitivity and efficiency of real-time PCR assays [23, 24]. This is in fact one of the main obstacles to overcome while developing a multiplex real-time PCR assay [17]. Although many singleplex assays have been successful in identifying blood meal and Leishmania parasites in sand flies [12,13,14,15, 25], none of them combined the detection of different host and the parasite DNA in a one-step assay. The development of the multiplex real-time PCR assay proposed by the present study resulted in a series of advantages compared to other assays, such as the reduction in reagent consumption, labour time and the ability to provide faster results (considering that regular real-time PCR assays take over 1 hour to complete, e.g. ~ 90 min [12, 18], ~ 78 min [13]). In practice, this assay allows testing a great number of sand flies for both dog cox1, human cytb and Leishmania kDNA in a shorter period of time (~ 34 min), reducing overall costs.

Considering that female sand flies consume a small amount of blood (≤ 1 μl) during blood-feeding [26], one of the main technical challenges while developing a molecular tool for detecting host blood and Leishmania parasites is the necessity to detect and quantify a very limited amount of DNA [16]. In fact, the sensitivity of such an assay depends directly on both the initial quantity of the target DNA in the sample and on the time span from blood ingestion, as there is a progressive degradation of the host DNA during blood digestion [12, 16, 27]. The multiplex real-time PCR assay developed in this study was shown to be specific and highly sensitive, without interference and competition between targets and dyes. Particularly, identical analytical sensitivities were obtained with singleplex and multiplex assays (i.e. 103 fg/reaction for dog cox1 and human cytb, and 1 fg for Leishmania kDNA). These results are similar to those obtained with other assays for detecting blood meals of female sand flies, which reported a detection limit from 102 fg to 104 fg of host DNA [6, 12, 27]. More recently, two SYBR Green-based real-time PCR assays were reported to have a detection limit of 26 fg for dog and 84 fg for human [13]. Despite the good analytical sensitivity of these assays, when non-target DNA samples were used some slight noise was reported in the melting curve analysis, though with a Cq value always higher than 30 [13]. In fact, non-specific signals are a known limiting factor of some SYBR Green-based real-time PCR assays [12, 28], which may eventually obscure the interpretation of the results. The use of hydrolysis probes (e.g. TaqMan probes) may increase the specificity of real-time PCR assays as demonstrated elsewhere [29, 30].

For Leishmania kDNA detection, the analytical sensitivity of our multiplex real-time PCR was similar to a singleplex assay using the same primers and probe [19], allowing the detection of less than a single parasite per sample. This high sensitivity may be partly attributed to the target used (i.e. kDNA), which is present in high number of copies (~ 10,000 copies of minicircle molecules) per parasite [31]. In fact, other real-time PCR assays using the same target gene also reported very good analytical sensitivity [32, 33].

Our multiplex real-time PCR assay was also successfully applied in field-collected samples. A total of 44 females of M. migonei (stored at − 20 °C for ~ 2 years) were positive for human blood. This sand fly species displays a remarkable degree of anthropophily, and it is a proven vector of L. braziliensis and a putative vector of L. infantum in Latin America [34, 35]. Interestingly, eight out of 44 females fed on humans were also positive for Leishmania spp. Leishmania-positive sand flies were collected in human dwellings (four indoors and eight outdoors), where human cases of CL by L. braziliensis were previously recorded [22]. Altogether, these findings strongly suggest that M. migonei is a vector of L. braziliensis for humans in the indigenous villages, where sand flies were collected. While our multiplex real-time PCR assay was tested with sand flies collected from a CL focus, it is also suitable for other epidemiological settings, namely VL endemic regions, considering that the primers and probe used also efficiently detected L. infantum kDNA [18, 19].

The identification of humans as the most frequent host of M. migonei females in this study, also reinforces their high attractiveness for humans [34] as well as the hypothesis that this species may be adapted to feed indoors [22]. It has been shown that human CL patients that have been treated and clinically cured may harbour viable parasites in their scars [36] and it has been suggested that they could eventually act as a source of infection to sand flies [37]. Our multiplex real-time PCR assay may be a useful tool to assess the presence of Leishmania parasites and human blood in sand flies from other CL-endemic areas, ultimately to investigate their possible role in the transmission cycle of L. braziliensis.

The finding of eight Leishmania-positive females which previously fed on humans raises interesting questions regarding whether these females acquired the infection from a previous unknown host (e.g. small rodents) [38] or from humans themselves. Interestingly, four Leishmania-positive engorged female sand flies did not apparently feed on humans or dogs, further suggesting that they probably acquired the parasites from another host.

Conclusions

In conclusion, a novel TaqMan-based fast multiplex real-time PCR assay was developed, optimized and validated herein for simultaneous detection of dog and human blood meals and Leishmania parasites in female sand flies. This assay may represent a tool for assessing Leishmania parasite infection in female sand flies and for investigating whether and how often these females feed on dogs and humans, thereby allowing estimation of the risk of infection in these hosts.