1 Introduction
The mosquito Anopheles pseudopunctipennis Theobald 1901 (Diptera: Culicidae) is a major vector of human Plasmodium sp. in South America. It is a difficult species to characterize as demonstrated by its variable behavior, habits and ecological needs [1], its inconsistencies as a malaria vector in its wide distribution range [2] and the several morphological subspecies described [3]. Crossmating experiments and cytogenetic studies pointed out evidence that these mosquitoes are comprised of a species complex [4,5]. Although the genetic variability and population structure of An. pseudopunctipennis has been studied biochemically, many aspects remain poorly known [6]. This species lacks specific molecular tools, and because of its high level of variability, more data are urgently needed to better understand how the population genetic structure of this mosquito is related to malaria transmission. Such data will help to better target vector control strategies by Public Health authorities.
Variability of non-coding sequences, particularly intron sequences, is a valuable marker of population variation and subdivision, and can be assayed by PCR amplification using conserved exon primers. Intron-targeted PCR was pioneered by Lessa [7]. This approach, called Exon-Primed Intron-Crossing (EPIC)-PCR [8,9], has been shown to identify substantial variability, mainly from intron length polymorphism. Introns constitute suitable markers for analyzing population structure within a species [10–15] as well as for reconstructing relationships among closely related species [16–18]. EPIC have been also used in gene mapping [19–23], and phylogeography [24–28] where they are the most widely used nuclear markers for such studies [29]. EPIC-PCR has several advantages in populations genetic studies:
- • by using primers from heterologous genes, cloning and sequencing steps can be avoided [30,31];
- • cross-species amplification is easier than when primers are designed from highly conserved exon sequences;
- • for the same reason, within species, PCR artefacts such as null alleles are expected to be less frequent.
Further advantages are that intron systems do not require previous knowledge of the genome to be analyzed; they are generally polymorphic and sometimes hypervariable; are expected to be codominant and selectively neutral; are easily amplifiable by PCR; can be revealed on simple agarose or acrylamide gels; and can be obtained at low cost [32]. Moreover, having both the exon and intron fragments, EPIC can be useful for examining genetic variation at the intraspecific and interspecific level simultaneously, a feature that is particularly useful when studying species complexes [27]. EPIC markers are becoming more popular for use in population genetic studies in insects [13–15,33,34] and do not require assumptions about a particular model of evolution that is often required for microsatellites [14]. In the present article, primers for EPIC amplification of intron sequences for An. pseudopunctipennis are designed, of which several pairs amplify (length)-polymorphic loci that can be used in population genetic studies of this malaria vector.
2 Materials and methods
2.1 Mosquitoes
In October 2006, females An. pseudopunctipennis were captured by one of us (F.L.) from a natural population in Mataral (S 18.6024, W 65.1444, altitude 1500 m), a small village situated in the dry inter-Andean valleys in the centre of Bolivia, using the human bait collection technique outside houses. In the field, collected insects were chloroform killed and stored over desiccant (silica gel) in small vials. In the laboratory, mosquitoes were identified using [35] and kept at –20 °C in their individual vials with silica gel until DNA extraction.
For the various EPIC-PCR (see paragraph below), positive controls consisted in females An. gambiae from a laboratory strain and were provided by our main laboratory at IRD-Montpellier (France). These mosquitoes were stored using the same conditions as for An. pseudopunctipennis until processing.
2.2 Selection of introns and design of primers
Primers for An. pseudopunctipennis were designed from the conserved regions of consecutive exons of different genes from the closely related species Anopheles gambiae Giles. Exon sequences were downloaded in 2005 at http://www.anobase.com in Excel format from the An. gambiae genome database. Gene candidates that were dispersed amongst the An. gambiae genome were selected, and close genes on the same chromosome were avoided. Gene candidates with a higher percentage of similarity with genes from Diptera Apis mellifera and Drosophila melanogaster, were first selected to enhance the chance of similarity to genes from An. pseudopucntipennis.
Intron lengths ranged from 100 to 500 bp. Primers pairs were designed to target the flanking exon sequences taking into account their stability (in terms of CG content and ending with CG or GC), their size (18–20 pb), their close annealing temperatures, and the low probability of primer-dimer formation during the PCR. Possible primers adjacent to the intron sequences were discarded. Fifty-four primer pairs were initially designed and screened by PCR using An. pseudopunctipennis DNA as a template (Table 1).
The fifty-four selected genes and their accession number.
Primer pair no | Gene | Accession number | Primer pair no | Gene | Accession number |
1 | AgaP_AGAP005839 | XM_315863.4 | 28 | AgaP_AGAP011438 | XM_554781 |
2 | AgaP_AGAP003128 | XM_001237495.2 | 29 | AgaP_AGAP011717 | XM_320795.4 |
3 | AgaP AGAP012571 | XM_307301.3 | 30 | AgaP_AGAP010343 | XM_311599.4 |
4 | AgaP_AGAP011936 | XM_320597.4 | 31 | AgaP_AGAP010725 | XM_559235.3 |
5 | AgaP_AGAP001573 | XM_551238.3 | 32 | AgaP_AGAP001813 | XM_321242.4 |
6 | AgaP_AGAP008026 | XM_555438.3 | 33 | AgaP_AGAP003857 | XM_310416.6 |
7 | AgaP_AGAP004780 | XM_318036.4 | 34 | AgaP_AGAP004298 | XM_313573.4 |
8 | AgaP_AGAP005961 | XM_316001.4 | 35 | AgaP_AGAP002956 | XM_311943.3 |
9 | AgaP AGAP012014 | XM_320516.2 | 36 | AgaP_AGAP002301 | XM_312670.1 |
10 | AgaP AGAP011363 | XM_001238009.2 | 37 | AgaP_AGAP003360 | XM_314262.4 |
11 | AgaP_AGAP011166 | XM_309483.4 | 38 | AgaP_AGAP001407 | XM_321726.4 |
12 | AgaP_AGAP001874 | XM_550942.3 | 39 | AgaP_AGAP003437 | XM_311723.4 |
13 | AgaP_AGAP007887 | XM_317605.4 | 40 | AgaP AGAP007738 | XM_001689008.1 |
14 | AgaP_AGAP009824 | XM_318932.3 | 41 | AgaP AGAP008288 | XM_001688954.1 |
15 | AgaP_AGAP004745 | XM_318073 | 42 | AgaP AGAP009200 | XM_319976.3 |
16 | AgaP_AGAP004934 | XM_315024.3 | 43 | AgaP_AGAP008938 | XM_319692.3 |
17 | AgaP_AGAP005622 | XM_315632.3 | 44 | AgaP_AGAP009835 | XM_553715.3 |
18 | AgaP_AGAP005693 | XM_315704.4 | 45 | AgaP_AGAP009785 | XM_318880.4 |
19 | AgaP_AGAP005806 | XM_315822.4 | 46 | AgaP_AGAP009856 | XM_318967.4 |
20 | AgaP AGAP006809 | XM_308938.4 | 47 | AgaP_AGAP004698 | XM_318145.4 |
21 | AgaP_AGAP006825 | XM_308919.3 | 48 | AgaP_AGAP004775 | XM_318039.4 |
22 | AgaP_AGAP007640 | XM_308229.4 | 49 | AgaP_AGAP004841 | XM_314353.3 |
23 | AgaP_AGAP007720 | XM_574504.3 | 50 | AgaP_AGAP004717 | XM_318113.4 |
24 | AgaP_AGAP008526 | XM_316916.3 | 51 | AgaP_AGAP004862 | XM_314327.3 |
25 | AgaP_AGAP008527 | XM_316915.4 | 52 | AgaP_AGAP004915 | XM_315006.4 |
26 | AgaP_AGAP012345 | XM_320207.4 | 53 | AgaP_AGAP004692 | XM_001231109.2 |
27 | AgaP_AGAP011730 | XM_320779.3 | 54 | AgaP_AGAP005948 | XM_315983.4 |
2.3 DNA extraction and amplification
An. pseudopunctipennis and An. gambiae DNA extractions were carried out on mosquito legs using a slightly modified cetyltrimethylammonium bromide (CTAB)-based protocol from Edwards [36]. The protocol was as followed: mosquito legs were homogenized in 200 μl lysis CTAB solution (100 mmol/l Tris HCl pH 8.0; 10 mmol/l EDTA pH 8.0; 1.4 mol/l NaCl and CTAB 2%) in 1.5 ml Eppendorf microcentrifuge tubes. Incubation was carried out at 65 °C for 15 min; the resulting extract was washed with 200 μl chloroform and centrifugated for 5 min at 12 000 rpm. The supernatant was precipitated in 200 μl isopropanol and centrifugated again at 12 000 rpm for 15 min. The pellet was washed with 200 μl 70% ethanol, centrifugated at 12 000 rpm for 5 min, dried at 37 °C for one hour and suspended in 100 μl nuclease-free H2O.
DNA amplifications were carried out immediately after extraction in volumes of 25 μl (1 × Taq buffer, 2.5 mM MgCl2, 0.4 mM dNTPs (Eurogentec, Angers, France), 0.5 UI Taq polymerase (Quiagen, Courtaboeuf, France), 20–25 ng of DNA template, and depending of the locus 0.04 μM or 0.4 μM of each primer (Eurogentec, Angers, France) (Table 2). The optimum annealing temperatures for each primer pair are listed in Table 2. PCR were performed on a Perkin Elmer DNA Thermal Cycler 480 (US Instrument Division, Norwalk, CT, USA) and conditions were: 1 min at 94 °C, followed by 36 cycles of 30 s at 94 °C, 30 s at annealing temperature, 30 s at 72 °C, and a final extension step of 5 min at 72 °C. The amplified products were first visualized on 1.5% agarose gels con ethidium bromide (Fig. 1). Then, for allele size analysis, they were separated by electrophoresis on 8% polyacrylamide gels and visualized by silver-staining. In all PCR, negative control (H2O) and positive controls (An. gambiae) were used.
Sequences of the 14 pairs of primers which successfully amplified 17 loci from Anopheles pseudopunctipennis and their characterization.
Primer pair no | Intron name | Chromosome | Accession number | Primer sequences (5′–3′) | Tm (°C) | No. of alleles | Sizes of alleles (bp) |
Amplification conditions | No. mosquitoes | H E | H O | Fis |
1 | G2LEX1236-1 | 2L | XM_315863.4 | F: TGGCTGGCTTCACGTCCG R: CGAGTGCAGGAACGGTGA |
55 | 4 | 114, 151, 165, 180 | MM1 | 18 | 0.4 | 0.3 | 0.25 |
1 | G2LEX1236-2 | 2L | XM_315863.4 | F: TGGCTGGCTTCACGTCCG R: CGAGTGCAGGAACGGTGA |
55 | 1 | 297 | MM1 | 18 | – | – | – |
2 | GUKEX1858 | 2R | XM_001237495.2 | F: GCCTGTGATCGTGCGTTTCG R:GGCATACCAGCAGCGTGACG |
55 | 4 | 794, 857, 870, 876 | MM1 | 18 | 0.3 | 0.3 | 0.09 |
3 | GUKEX1859-1 | UNKN | XM_307301.3 | F: CGAGGAGGGTGTACAAACGC R: GGTGTCGCCTAGCTCGCCCG |
55 | 2 | 715, 766 | MM1 | 20 | 0.5 | 0.7 | –0.3 |
3 | GUKEX1859-2 | UNKN | XM_307301.3 | F:CGAGGAGGGTGTACAAACGC R:GGTGTCGCCTAGCTCGCCCG |
55 | 2 | 306, 320 | MM1 | 20 | 0.1 | 0.1 | 0.65 |
3 | GUKEX1859-3 | UNKN | XM_307301.3 | F: CGAGGAGGGTGTACAAACGC R: GGTGTCGCCTAGCTCGCCCG |
55 | 3 | 183, 193, 197 | MM1 | 20 | 0.2 | 0.1 | 0.79 |
4 | G3LEX28 | 3L | XM_320597.4 | F:CCAACTACTCGGCCGTGC R:GCCGGCCATCTCCTTCGC |
60 | 3 | 246, 253, 261 | MM2 | 18 | 0.4 | 0.4 | 0.02 |
5 | G2REX47 | 2R | XM_551238.3 | F: GGCACGGTGGGGAAGACG R: CCGTCCACCACCATCGGG |
60 | 4 | 220, 228, 257, 273 | MM2 | 18 | 0.3 | 0.4 | –0.1 |
6 | G3REX1037 | 3R | XM_555438.3 | F: GCAAACGCGAAAGAACCG R: GCCTGGTAGCGCTTCTCG |
60 | 4 | 281, 315, 345, 405 | MM2 | 12 | 0.5 | 0 | 1 |
7 | EX113 | 2L | XM_318036.4 | F: CATCTATCTGCTGAACTCGC R: CGTCGGTCACATTCCACATC |
60 | 3 | 541, 557, 574 | MM2 | 14 | 0.4 | 0 | 1 |
8 | EX1358 | 2L | XM_31600104 | F: CATGCCTCCAATGGTGCC R: CCGTACGTTCCTTCGCCA |
60 | 1 | 249 | MM2 | 20 | – | – | – |
9 | G3LEX3 | 3L | XM_320516.2 | F: CCGAAGATGAGCTCAGAGATGC R: CCTAGCTTGTCGGTGATTTCTG |
55 | 1 | 185 | MM1 | 20 | – | – | – |
10 | G3LEX9 | 3L | XM_001238009.2 | F: CGCCCTGCCTGGCATGGATTCG R: GCAGGCACAGCCACCTTCCGGG |
55 | 1 | 656 | MM1 | 20 | – | – | – |
11 | G3LEX36 | 3L | XM_309483.4 | F: CGCGGCAATCATGAGTGCGCC R: CCACCGGCAGACAGTTGAAGC |
55 | 1 | 191 | MM1 | 20 | – | – | – |
12 | G2REX46 | 2R | XM_550942.3 | F: CCGACGATAGAGGACAGC R: GTTGAAGGTCGACTGTGC |
60 | 1 | 497 | MM2 | 20 | – | – | – |
13 | G3REX491 | 3R | XM_317605.4 | F: CGTTGGAGCAGCAACAACAGC R: GGTAATGATTCCTGATATTGC |
55 | 1 | 136 | MM2 | 20 | – | – | – |
14 | G3REX1062 | 3R | XM_318932.3 | F: CGATCTGCTGGCCGACTTCC R: CCATCGCCCTTGCGCTCACC |
55 | 1 | 177 | MM1 | 20 | – | – | – |
Overall | 0.35 | 0.24 | 0.35 |
2.4 Data analysis
Allele sizes were scored using the numerical procedure implemented in LabImage version 3.0 [37] using a size-standard 100 bp gene ruler. Observed (HO) and expected (HE) heterozygosities, the inbreeding coefficients (Fis statistics) and the R2 coefficient to estimate linkage disequilibrium for each locus pairs were computed using procedures implemented in the GENETIX package [38].
3 Results and discussion
Of the 54 selected pairs of primers, 14 pairs successfully amplified 17 loci (Table 2). All the 54 pairs of primers successfully amplified loci in An. gambiae. Forty pairs of primer did not amplify any loci in An. pseudopunctipennis while they did in An. gambiae. As an example, Fig. 1 presents results of agarose gel electrophoresis for four primers pairs. Fig. 1a shows positive amplifications in both An. pseudopunctipennis and An. gambiae while Fig. 1b gives examples of negative amplifications in An. pseudopunctipennis but positive ones in An. gambiae (positive controls). Because EPIC are universal primers they may amplify introns in other species. Table 3 is based on nucleotide BLAST (NCBI web site) of the 54 studied primers pairs and exhibits those that might amplify in a variety of other species including other mosquitoes (Aedes aegypti and Ae. albopictus, other five species of Anopheles, Culex quinquefasciatus), other Diptera (12 species of Drosophila), other insects (the cricket species Gryllus bimaculatus, the moths of the Noctuidae family Helicoverpa armigera and Heliotis viriplaca), the deer tick Ixodes scapularis, vertebrates (human, dog, mouse), a choanoflagelate Monosiga brevicolis, a cyanobacteria Thermosynechococcus elongates, a Dermatophytes zoophilic fungus Trichophyton verrucosum and even a plant, the common plantain Plantago major. Some primer pairs are more universal than others in the sense that, for example, primer pair no 4 (intron G3LEX28), primer pair no 11 (intron G3LEX36) or primer pair no 21 might amplify in 16, 11 and 7 species respectively. Of all the primer pairs studied, 11 might amplify in Ae. aegypti, the main vector of dengue viruses and 15 in C. quinquefasciatus, an annoying species also vector of human filariasis and West Nile virus (Table 3).
Primer pairs and species in which they may potentially amplify, according to nucleotide sequence alignments using BLAST (NCBI).
Primer pair n° | ||||||||||||||
Species | 1 | 2 | 4 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 16 | 18 | 19 | |
Aedes aegypti | x | x | x | x | x | x | x | |||||||
Ae. albopictus | ||||||||||||||
Anopheles arabiensis | x | x | ||||||||||||
An. bwambae | x | |||||||||||||
An. christyi | x | |||||||||||||
An. merus | x | |||||||||||||
An. quadrianulatus | x | |||||||||||||
An. stephensi | ||||||||||||||
Canis familiaris | x | |||||||||||||
Culex quinquefasciatus | x | x | x | x | x | x | x | x | ||||||
Drosophila melanogaster | x | x | ||||||||||||
D. ananassae | x | x | ||||||||||||
D. erecta | x | |||||||||||||
D. grimshawi | x | x | x | |||||||||||
D. mojavensis | x | x | ||||||||||||
D. persimilis | x | x | ||||||||||||
D. pseudoobscura | x | x | ||||||||||||
D. sechellia | x | x | ||||||||||||
D. simulans | x | x | ||||||||||||
D. virilis | x | |||||||||||||
D. willistoni | x | |||||||||||||
D. yakuba | x | x | ||||||||||||
Gryllus bimaculatus | x | |||||||||||||
Helicoverpa armigera | x | |||||||||||||
Heliotis viriplaca | x | |||||||||||||
Homo sapiens | x | |||||||||||||
Ixodes scapularis | x | x | x | |||||||||||
Monosiga brevicolis | x | x | ||||||||||||
Mus musculus | ||||||||||||||
Plantago major | ||||||||||||||
Thermosynechococcus elongatus | ||||||||||||||
Trichophyton verrucosum | x | |||||||||||||
Number of species that might be recognized | 1 | 2 | 16 | 1 | 2 | 5 | 11 | 1 | 4 | 2 | 2 | 5 | 2 | |
Primer pair no | ||||||||||||||
Species | 20 | 21 | 22 | 29 | 31 | 34 | 35 | 36 | 37 | 38 | 43 | 50 | 54 | Number or primer pairs that might amplify in the species |
Aedes aegypti | x | x | x | x | 11 | |||||||||
Ae. albopictus | x | 1 | ||||||||||||
Anopheles arabiensis | 2 | |||||||||||||
An. bwambae | 1 | |||||||||||||
An. christyi | 1 | |||||||||||||
An. merus | 1 | |||||||||||||
An. quadrianulatus | 1 | |||||||||||||
An. stephensi | x | 1 | ||||||||||||
Canis familiaris | 1 | |||||||||||||
Culex quinquefasciatus | x | x | x | x | x | x | x | 15 | ||||||
Drosophila melanogaster | x | x | x | x | x | x | 8 | |||||||
D. ananassae | x | x | 4 | |||||||||||
D. erecta | x | x | x | 4 | ||||||||||
D. grimshawi | x | 4 | ||||||||||||
D. mojavensis | x | 3 | ||||||||||||
D. persimilis | x | 3 | ||||||||||||
D. pseudoobscura | x | 3 | ||||||||||||
D. sechellia | x | x | 4 | |||||||||||
D. simulans | x | 3 | ||||||||||||
D. virilis | 1 | |||||||||||||
D. willistoni | 1 | |||||||||||||
D. yakuba | x | x | x | x | 6 | |||||||||
Gryllus bimaculatus | x | 2 | ||||||||||||
Helicoverpa armigera | 1 | |||||||||||||
Heliotis viriplaca | 1 | |||||||||||||
Homo sapiens | x | 2 | ||||||||||||
Ixodes scapularis | x | 4 | ||||||||||||
Monosiga brevicolis | 2 | |||||||||||||
Mus musculus | x | 1 | ||||||||||||
Plantago major | x | 1 | ||||||||||||
Thermosynechococcus elongatus | x | 1 | ||||||||||||
Trichophyton verrucosum | 1 | |||||||||||||
Number of species that might be recognized | 2 | 7 | 2 | 4 | 4 | 4 | 1 | 4 | 6 | 1 | 2 | 2 | 2 |
All the 17 loci amplified in An. pseudopunctipennis were tested for their polymorphism level using 20 An. pseudopunctipennis. Nine of the 17 amplified loci were polymorphic, ranging from two to four alleles (Table 2).
Previous studies relying on EPIC-PCR reported that various loci might be scored for a given pair of primers [11,32,39–41]. This phenomenon may reflect former polyploidizations, tandem duplications, and other phenomena occurring during lineage evolution and producing pseudogenes. Since An. pseudopunctipennis is a diploid species, gene duplication processes are more likely involved. The simultaneous amplification of two or three loci (two for G2LEX1236 and three for GUKEX1859) can then reveal the presence of another gene or a pseudogene. These simultaneous amplifications did not seem to disturb the quality of PCR and the reading of the genotypes because: the supplementary profiles did not present parasitic bands; and, the variation of size between band systems was larger than the size polymorphism of each locus. Indeed, when intraspecific allelic variants occur, a locus can be evidenced by the presence of all the allelic combinations following a Mendelian inheritance (homozygotes and heterozygotes); if not, in order to circumvent erroneous allelic assignment, only loci where the allelic size-variation is lower between orthologous than paralogous loci, should be retained [11].
With EPIC-PCR, artifacts such as null alleles are expected to be less frequent than for example, with microsatellites [42]. However, Table 2 exhibits some differences in the number of mosquitoes that correctly responded to amplification, in particular for primer pairs no 6 and 7 (Intron G3REX1037 and EX113, respectively, with 12 and 14 mosquitoes, respectively, instead of the 20 expected). A careful examination of results seemed to indicate the presence of null alleles for these primer pairs and moreover, no heterozygotes were encountered (Fis = 1). Therefore, these two primer pairs may not be suitable for population genetic studies. The lack of amplification in two mosquitoes for primers pairs no 1, 2, 4 and 5 are more likely the fact of PCR manipulation problems (low DNA concentration?) because the failure of amplification with the four above primer pairs occurred with the same two mosquitoes. Null alleles are therefore not suspected and these primer pairs may be proposed for genetic studies.
For the small population under study, the R2 coefficients to estimate linkage disequilibrium for each locus pair are presented in Table 4. Strong linkage disequilibrium appeared between G2REX47 and most of the other loci.
R2 coefficients for linkage disequilibrium computed for each pair of polymorphic loci.
Intron names | G2LEX1858 | GUKEX1859-1 | GUKEX1859-2 | GUKEX1859-3 | G3LEX28 | G2REX47 | G3REX1037 | EX113 |
G2LEX1236-1 | 0.28 | 0.42 | 0.65 | 0.01 | 0.10 | 0.91 | 0.78 | 0.04 |
G2LEX1235-1 | 0.79 | 0.92 | 0.74 | 0.62 | 0.21 | 0.59 | 0.89 | |
GUKEX1859-1 | 0.30 | 0.30 | 0.29 | 0.85 | 0.71 | 0.20 | ||
GUKEX1859-2 | 0.96 | 0.67 | 0.89 | 0.83 | – | |||
GUKEX1859-3 | 0.10 | 0.99 | 0.06 | 0.10 | ||||
G3LEX28 | 0.90 | 0.21 | 0.01 | |||||
G2REX47 | 0.02 | 0.52 | ||||||
G3REX1037 | – |
Compared with other DNA-based techniques such as microsatellites, EPIC-PCR is the only technique based on universal primers that allows a fast screening even for cross-species studies. Because An. pseudopunctipennis is likely a species complex [4,5], with at least five described subspecies [43] and two recognized morphological variants in Bolivian larvae [44], the above EPIC primers may therefore be used to better understand the species status.
Another interesting characteristic of EPIC-PCR is that, likely because gene duplication, one single pair of primer can produce more than one locus and can provide a number of polymorphisms, as it was the case with primer pair no 3 (intron GUKEX1859) in An. pseudopunctipennis.
Primer pairs no 2 and 4 indicate (at least in the small sample of 20 mosquitoes from Mataral village) that the genotype frequencies conform to Hardy-Weinberg equilibrium. Therefore, polymorphism in these introns may be selectively neutral, as predicted for variation in most non-coding DNA sequences. If so, such markers are particularly powerful in population genetics studies.
Population genetics of An. pseudopunctipennis will benefit of the above EPIC markers, and using recently available methods [27,42], other EPIC markers could be isolated more easily.
Disclosure of interest
The authors declare that they have no conflicts of interest concerning this article.
Acknowledgments
This project was partly supported by a PAL+ Grant from the Ministère de la Recherche (France) and a Contrat de Développement Grant from Ministère des Affaires Étrangères (France). The participation of R. Ursic-Bedoya in this study was possible thanks to the Association of Universities and Colleges of Canada (AUCC) and its Canada-Latin America and the Caribbean Research Exchange Grants. The authors would like to thank P. Kengne who encouraged this study and an anonymous referee whose comments helped us to greatly improve the manuscript.