Conceived and designed the experiments: AMGP AMMJ GJF OTC. Performed the experiments: VHP GJF. Analyzed the data: VHP GJF AMGP AMMJ OC OTC. Contributed reagents/materials/analysis tools: AMMJ OTC. Wrote the paper: VHP GJF AMGP AMMJ OC OTC.
The authors have declared that no competing interests exist.
Methods to determine blood-meal sources of hematophagous Triatominae bugs (Chagas disease vectors) are serological or based on PCR employing species-specific primers or heteroduplex analysis, but these are expensive, inaccurate, or problematic when the insect has fed on more than one species. To solve those problems, we developed a technique based on HRM analysis of the mitochondrial gene cytochrome B (Cyt b). This technique recognized 14 species involved in several ecoepidemiological cycles of the transmission of
Chagas disease is one of the most important tropical diseases in America. This disease is caused by the parasite
Depending on the habitats of triatomines, at least two transmission cycles of
The identification of
Unlike serological tools, PCR-based molecular methods are more specific and usually easier to perform, although the design of specific primers for each species involved in transmission scenarios makes it difficult to identify the vector and mammals from the wild cycle where a wide variety of species exist. Additionally, this technique is wasteful because each sample has to be amplified by PCR many times with different molecular markers
Recently, high-resolution melting (HRM) has become a sensitive genotyping method, based on the characteristics of thermal denaturation of the amplicons. This method has a much higher performance information never before achieved by classical DNA melting curve analysis
All Animals were handled in strict accordance with good animal practice as defined by the Colombian code of practice for the care and use of animals for scientific purposes. Ethical approval (Act N° 53, 30/06/2009) for analyzing animal specimens was obtained from the animal ethics Committee of the University of Antioquia, Medellin, Colombia.
To differentiate among species-specific genotypes of natural blood-meal sources of triatomine bugs, a 383-bp fragment from the Cyt b gene was amplified from DNA samples extracted from tissue or blood of 36 individuals corresponding to 14 species. Species were chosen according to their epidemiological relevance as reservoirs of
Standard species | Common name | Tissue | blood | Total |
|
Opossum | - | 3 | 3 |
|
Mouse | 1 | 2 | 3 |
|
Cow | 2 | 1 | 3 |
|
Goat | - | 3 | 3 |
|
Dog | - | 3 | 3 |
|
Sheep | - | 1 | 1 |
|
Horse | - | 3 | 3 |
|
Pig | 3 | - | 3 |
|
Rabbit | - | 2 | 2 |
|
Cat | - | 3 | 3 |
|
Rat | - | 1 | 1 |
|
Donkey | - | 2 | 2 |
|
Human | - | 3 | 3 |
|
Chicken | 2 | 1 | 3 |
-: sample not used.
To validate HRM-Cyt b profiles, six fifth-stage nymphs per species of
To evaluate the capability of this technique to detect HRM-Cyt b profiles in mixed feeds, five fifth-instar
To apply the HRM technique to triatomines from the field, a total of 20 insects were collected in four localities of Colombia's Caribbean region, which display different eco-epidemiological transmission cycles of
All insects were placed in plastic bottles, marked, transported to the laboratory, and identified according to the classification proposed by Lent and Wygodzinsky
DNA was extracted using the phenol-chloroform method
A 383-bp fragment from the Cyt b gene was real time PCR-amplified from genomic DNA samples using the
HRM analysis was carried out using the Rotor Gene Q software v2.2 with normalization regions between 76.15–78.65°C and 89.50–91.00°C. Genotypes were defined by selecting a sample from each standard species as a reference control to identify unknown samples. The software then auto-called the genotype and melting temperatures of each amplicon and provided a confidence percentage based on the square root of the correlation coefficient between samples and the reference genotypes. Any specimen generating melt curves but having a dF/dT less than 1.0 was considered not subjected to genotyping. Averages of melting temperatures and the confidence percentage of the specimen replicates were assigned to a representative genotype, and the standard deviation, confidence intervals, and variation coefficient were calculated using the Prism GraphPad v4.0 software (GraphPad software, Inc).
To confirm the sequence identity of those samples tested by HRM analysis, a 358-bp fragment of Cyt b was sequenced for the standard samples and ten test samples chosen randomly (2, 3, 4, 5, 10, 12, 15, 16, 19, and 20 (
All samples analyzed amplified a 383-bp product, as previously reported to the Cyt b gene
Tm: melting temperature, Sp: species.
Tm analysis | HRM analysis | |||||||
Sp | Tm | SD | 95% CI of Tm | VC | Tm genotype | HRM genotype | %C | %C SD |
Opossum | 81.02 | 0.09 | 80.97–81.06 | 0.10 | Opossum | Opossum | 81.24 | 15.02 |
Mouse | 81.63 | 0.12 | 81.57–81.69 | 0.14 | Mouse | Mouse | 94.19 | 4.94 |
Cow | 81.76 | 0.49 | 81.51–82.00 | 0.60 | Cow | Cow | 83.07 | 16.63 |
Goat | 81.88 | 0.05 | 81.86–81.91 | 0.06 | Goat | Goat | 88.24 | 13.25 |
Dog | 82.29 | 0.08 | 82.25–82.33 | 0.09 | Dog | Dog | 90.92 | 13.86 |
Sheep | 82.29 | 0.03 | 82.26–82.33 | 0.04 | Sheep | Sheep | 98.69 | 1.50 |
Horse | 82.56 | 0.08 | 82.52–82.60 | 0.10 | Horse | Horse | 93.46 | 9.04 |
Pig | 82.58 | 0.05 | 82.55–82.60 | 0.06 | Pig | Pig | 95.18 | 9.50 |
Rabbit | 82.69 | 0.05 | 82.66–82.72 | 0.06 | Rabbit | Rabbit | 98.55 | 1.33 |
Cat | 83.46 | 0.09 | 83.41–83.50 | 0.11 | Cat | Cat | 78.81 | 22.40 |
Rat | 83.55 | 0.05 | 83.50–83.59 | 0.05 | Rat | Rat | 89.90 | 6.51 |
Donkey | 84.39 | 0.04 | 84.36–84.41 | 0.05 | Donkey | Donkey | 87.85 | 22.14 |
Human | 85.79 | 0.05 | 85.76–85.82 | 0.06 | Human | Human | 96.58 | 2.22 |
Chicken | 86.27 | 0.05 | 86.24–86.29 | 0.06 | Chicken | Chicken | 93.05 | 18.01 |
SD: standard deviation, CI: confidence interval, VC: variability correlation, %C: confidence percentage, %C SD: standard deviation of confidence percentage, Tm Genotype: Genotype according to Tm from all the replicas.
Although some species exhibited the same Tm value, HRM profiles in these species were clearly discriminated and recognized as different genotypes, such as dog and sheep or pig and horse (
The HRM profile of the Cyt b gene obtained from feces and intestinal content from
All samples were successfully amplified by qPCR 1, 5, 15, and 30 days post-feeding (
Tm Analysis | HRM Analysis | ||||||||||
Feed | Sample Source | Time course(days) | Tm1 | SD | Tm1 Genotype | Tm2 | SD | Tm2 Genotype | HRM Genotype | HRM %C | %C SD |
H | IC | 1 | 86.24 | 0.02 | CG | - | - | - | CG | 55.23 | 10.84 |
H | IC | 5 | 86.28 | 0.03 | CG | - | - | - | CG | 89.49 | 8.73 |
H | IC | 5 | 86.3 | 0.03 | CG | - | - | - | CG | 82.15 | 8.77 |
H | IC | 15 | 81.17 | 0.06 | M | 85.89 | 0.06 | H | ND | ND | ND |
H | IC | 15 | 81.43 | 0.04 | M | 86.47 | 0.02 | C | ND | ND | ND |
H | IC | 30 | 86.26 | 0.01 | CG | - | - | - | CG | 61.32 | 3.10 |
H | IC | 30 | 85.79 | 0.01 | HG | - | - | - | HG | 63.65 | 5.66 |
H | F | 30 | 81.54 | 0.01 | M | 85.97 | 0.03 | H | ND | ND | ND |
H | F | 30 | 85.9 | 0.05 | HG | - | - | - | HG | 21.75 | 1.93 |
C | IC | 1 | 86.32 | 0.03 | CG | - | - | - | CG | 94.49 | 3.60 |
C | IC | 1 | 86.32 | 0.03 | CG | - | - | - | CG | 96.83 | 1.80 |
C | IC | 1 | 86.28 | 0.02 | CG | - | - | - | CG | 89.45 | 5.88 |
C | IC | 5 | 86.34 | 0.01 | CG | - | - | - | CG | 93.85 | 1.11 |
C | IC | 5 | 86.34 | 0.04 | CG | - | - | - | CG | 95.56 | 4.01 |
C | IC | 15 | 86.35 | 0.00 | CG | - | - | - | CG | 97.11 | 1.80 |
C | IC | 15 | 86.31 | 0.01 | CG | - | - | - | CG | 96 | 2.75 |
C | IC | 30 | 86.34 | 0.04 | CG | - | - | - | CG | 91.93 | 8.59 |
C | IC | 30 | 86.32 | 0.02 | CG | - | - | - | CG | 98.37 | 0.05 |
C | F | 30 | 86.41 | 0.05 | CG | - | - | - | CG | 33.37 | 18.35 |
H: human, C: chicken, IC: intestinal content, F: feces, CG: chicken genotype, HG: human genotype, M: mixed feeding determined by two peaks in the melting curve, Tm1: melting temperature of peak 1, Tm2: melting temperature of peak 2, SD: standard deviation, HRM %C: confidence percentage HRM, %C SD standard deviation of confidence percentage, ND: not determined, -: No secondary peak.
Therefore, we evaluated the possibility of detecting at least two different blood meal sources. In this case, the melt curve showed two peaks belonging to mouse and chicken amplicons, with their respective Tm. In addition, the dissociation curve had a shape with two falls (
The HRM analysis of Cyt b in
Lowest temperature peak analysis | Highest temperature peak analysis | |||||||||||
Sample | Tm1 | SD | Tm1 genotype | HRM genotype | HRM %C | SD | Tm2 | SD | Tm2 genotype | HRM genotype | HRM %C | SD |
1 | 86.07 | 0.07 | Human | Human | 83.29 | 14.60 | - | - | - | - | - | - |
2 | 85.96 | 0.06 | Human | Human | 90.98 | 5.02 | - | - | - | - | - | - |
3 | 85.96 | 0.05 | Human | Human | 81.84 | 6.48 | - | - | - | - | - | - |
4 | 86.15 | 0.06 | Human | Human | 75.41 | 2.02 | - | - | - | - | - | - |
5 | 86.03 | 0.04 | Human | Human | 90.91 | 3.56 | - | - | - | - | - | - |
6 | 82.04 | 0.08 | Dog | Dog | 56.48 | 3.30 | - | - | - | - | - | - |
7 | 82.02 | 0.15 | Dog | Dog | 53.68 | 25.81 | 86.14 | 0.04 | Human | Human | 64.56 | 14.75 |
9 | 84.74 | 0.02 | ND | ND | ND | ND | 86.13 | 0.03 | Human | Human | 63.11 | 9.50 |
10 | 84.87 | 0.21 | ND | ND | ND | ND | 86.10 | 0.04 | Human | Human | 78.90 | 15.17 |
12 | 81.16 | 0.01 | Opossum | Opossum | 86.02 | 11.19 | 86.10 | 0.03 | Human | Human | 82.24 | 3.30 |
14 | 86.15 | 0.02 | Human | Human | 56.35 | 6.93 | - | - | - | - | - | - |
15 | 84.85 | 0.09 | ND | ND | ND | ND | 86.20 | 0.06 | Human | Human | 40.76 | 11.44 |
16 | 86.19 | 0.06 | Human | Human | 57.23 | 11.38 | - | - | - | - | - | - |
19 | 84.85 | 0.08 | ND | ND | ND | ND | 86.16 | 0.01 | Human | ND | ND | ND |
20 | 86.58 | 0.02 | Chicken | Chicken | 88.71 | 7.08 | - | - | - | - | - | - |
Tm1: melting temperature of peak 1, Tm2: melting temperature of peak 2, SD: standard deviation, HRM %C: confidence percentage of HRM analysis, ND: not determined or not recognized, -: No secondary peak.
As mentioned above, four mixed samples, including one intradomiciliary insect, showed a peak with a Tm = 84.83±0.05, which was inconsistent with any of our standards (
Blastn results showed significant identity values greater than 95% for all samples except for sample number 20 (
Sample | Blastn description | e-Value | Accessionnumber | Identity with the standard used to HRM identification |
2 |
|
0 | AY509658.1 | 98.30% |
3 |
|
0 | AY509658.1 | 98.32% |
4 |
|
0 | AY509658.1 | 98.89% |
5 |
|
0 | AY509658.1 | 98.32% |
10 |
|
0 | AY509658.1 | 98.89% |
12_1 |
|
2.00E-161 | AY509658.1 | 94.69% |
12_2 |
|
2.00E-167 | DQ236278.1 | 98.60% |
15 |
|
0 | AY509658.1 | 98.89% |
16 |
|
9.00E-175 | HQ384199.1 | 97.49% |
19 |
|
9.00E-180 | AY509658.1 | 98.04% |
20 |
|
4.00E-33 | FM205717.1 | 59.78% |
For each sample the identity with the standard species used for HRM identification is shown.
The epidemiological scenario of Chagas disease has become increasingly complex over the years. The natural habitats of some human populations within the forest and deforestation caused by humans are but two of the reasons that may complicate this scenario. Thus, the classical separation of transmission cycles defined for this disease could be different for many places, making it difficult to determine the epidemiological characteristics of particular regions. With this unclear scenario, the determination of blood-meal sources in hematophagous vectors has become essential to surveillance and prevention of potential infection foci.
In this study, HRM analysis of the Cyt b gene made it possible to identify 14 species successfully even when some of them had the same Tm values. Each species was well recognized under a variety of species reaching high confidence percentage values, thus showing the power of this classification using gene amplification and HRM analysis. It is worth noting that this recognition is possible with a single PCR, making it a quick and inexpensive technique.
It is important to highlight that no studies with this number of species standards have been conducted with Chagas disease vectors. Pizarro and Stevens
A similar power is reached when DNA extracted from intestinal content and feces is analyzed with this technique, demonstrating that both can be used for the identification of the blood source from insects. Although no misidentification was recorded with DNA from insect feces, low confidence percentages were reached and other samples did not amplify, showing that the quality of the sample is a limitation of the technique, as reported for real-time PCR
The HRM technique proved to work in both types of samples (intestinal and fecal) at least until 30 days after insects were fed. This showed the applicability of the technique even when the blood-meal sources were limited to the triatomines. Pizarro and Stevens
On the other hand, the cow profile showed two melting peaks, suggesting the amplification of nonspecific fragments. This could stem from nuclear copies of mtDNA (Numts). It is well known that Numts can be amplified in genetic studies based on mtDNA
Unspecific peaks led to testing mixed feedings, showing that the technique detects both Tm sources. This allows identification of many species from which the insects have been fed, even when there are no standards for those species. This is impossible with methods such as ELISA or techniques based on PCR-specific primers. Heteroduplex analysis identifies mixed feeding as well, but the electrophoretic profile becomes very complex, making it difficult to analyze, and sometimes it does not recognize even a single species within the mixture
We processed 20 samples from the field but only 15 amplified. The other five samples were probably of insufficient quality, which is not surprising because it was feces DNA.
In general, the expected genotypes were recognized by the analysis. Those insects captured inside houses showed human feeding and one dog feeding, suggesting a domestic transmission cycle involving humans and to a lesser extent dogs, which has been reported to have epidemiological relevance
We included
There was only one genotype that could not be identified with a Tm = 84.83°C±0.05, indicating that the number of species standards included in the study was adequate. However, this could be improved knowing the fauna and diversity of species of a particular study area. When the sample is not recognized, the next step is sequencing, as reported by other authors
In conclusion, we believe that epidemiological studies involving vectorial incrimination and transmission dynamics must identify the blood-meal source to cover the entire panorama of transmission. HRM analysis of the Cyt b gene is the most powerful technique in this type of study because it can accurately identify the species even when the vector has mixed feeding, it has a high resolution power, and it is fast, easy, and inexpensive. However, it is important to obtain high-quality DNA and be mindful of the fauna of the study area to have an adequate number of standard species.
(TIF)
(TIF)
(PDF)
(DOCX)
(DOCX)