Morphology does not allow differentiating the species of the Phlebotomus perniciosus complex: Molecular characterization and investigation of their natural infection by Leishmania infantum in Morocco

Morphological and DNA-based complemented approaches were applied for characterization of sympatric populations of Phlebotomus longicuspis and Phlebotomus perni-ciosus in Morocco. Both sand fly species are generally recorded in sympatry in North Africa but on few occasions have been molecularly characterized. The diagnostic confusion of these species has


| INTRODUC TI ON
Cutaneous leishmaniasis (CL) is a parasitic disease that is endemic in more than 70 countries, with an incidence of 1.2 million cases per year. The World Health Organization included four Maghrebi countries (Algeria, Tunisia, Libya and Morocco) among the countries with a high CL burden in its latest report (WHO, 2021). The sand fly fauna is very diverse in these countries and many species are incriminated in the transmission of the three endemic Leishmania species: Leishmania major, L. tropica, (both dermotropic) and Leishmania infantum. The latter is a causative agent of visceral, cutaneous and mucosal human leishmaniasis as well as canine leishmaniasis.
Phlebotomus longicuspis is only present in North Africa where it is most frequent in the semi-arid, arid and peri-arid bioclimatic zones (Benallal et al., 2022;Martín-Sánchez et al., 2000;Pesson et al., 2004). In the Maghreb region, both species are generally found in sympatry from the subhumid to Sahara bioclimatic zones (Benallal et al., 2022). Morphological examinations of P. perniciosus males revealed the existence of two morphotypes: a typical form with bifid copulatory valves and an atypical form with single tipped copulatory valves sharply curved at their apex; some specimens can show mixed features as well (Benabdennbi et al., 1999;Martín-Sánchez et al., 2000;Morillas Márquez et al., 1991;Pesson et al., 2004). Atypical specimens have been misidentified as P. longicuspis until morphological diagnosis had been resolved by isoenzyme and genetic studies (Martín-Sánchez et al., 2000;Pesson et al., 2004). This misinterpretation has likely led to mistakes in the geographical distribution of these species and their involvement in the transmission of L. infantum in the area (Guernaoui et al., 2005;Martín-Sánchez et al., 2000).
Furthermore, isoenzyme and mitochondrial DNA analysis of P.
longicuspis specimens from Morocco have revealed that this species does not have the genetic characteristics of a single, reproductively isolated species, but two sibling species, named P. longicuspis s.s. and P. longicuspis LCx Pesson et al., 2004).
Phlebotomus perniciosus and P. longicuspis females are differentiated morphologically through the examination of the dilatation in the distal part of the spermathecal ducts, finding a relatively thick-or thin-walled bulb respectively. No morphological variation and intermediate aspects have been found (Berchi et al., 2007;Guernaoui et al., 2005;Léger et al., 1983). Therefore, the females of two forms of P. perniciosus and those of two P. longicuspis sibling species are considered, respectively, morphologically indistinguishable.
Atypical P. perniciosus and P. longicuspis s.l. are widespread in Morocco, Tunisia and Algeria whereas typical P. perniciosus is limited to the north of these countries (Benallal et al., 2017;Boudabous et al., 2012;Zarrouk et al., 2016). Depaquit et al. in 2005 recorded for the first time the presence of P. longicuspis s.l. south of the Sahara. The distribution of P. longicuspis LCx is unknown (Zarrouk et al., 2016). According to these authors, the typical form of P. perniciosus and P. longicuspis s.l. would be involved in L. infantum transmission whereas the epidemiological role of the atypical P. perniciosus and P. longicuspis LCx must be put into question.
Given the possibility of phenotypic differences of biomedical importance between these taxa, more extensive characterization of their populations could help assess their vectorial capacity.
Mitochondrial DNA fragments such as cytochrome oxidase I (COI) and cytochrome b (Cytb) are efficient markers to distinguish the two 2 | MATERIAL S AND ME THODS

| Study area
El Borouj (coordinates 07°36′W-32°29′N) is situated at an altitude of 410 m above sea level in the Settat province, central Morocco. The population is around 20,000 inhabitants and the average growth rate is close to 2%. Anthroponotic cutaneous leishmaniasis (ACL) by L. tropica is endemic in El Borouj while no cases of human visceral or cutaneous leishmaniasis due to L. infantum have been described (Ministry of Health, Morocco, 2016). In 2017 we detected a case of cutaneous leishmaniasis due to L. infantum in the Oulad Bouchair neighbourhood using a Granaleish Multiplex qPCR (University of Granada, Spain, Trade Mark Number 3667362/5).

| Ethical approval statement
The project was approved by the Ethics Committee of the University of Granada and by the Moroccan Ministry of Health. The study and its procedures were explained to the mayor and heads of local healthcare centre and households in El Borouj. Written consent was obtained from the families who agreed to let us place the traps in their homes.

| Sand fly collection and morphological identification of the species
Sand flies were caught using CDC light traps inside households in 2014-2015. Houses with and without ACL cases distributed in nine neighbourhoods were sampled. CDC traps were set in each selected house (one to two per house) for one night under favourable weather conditions. Males and females were separated and morphologically identified using taxonomic keys (Gijón-Robles et al., 2018).
For molecular studies, the genitalia of Larroussius specimens were individually removed and mounted on slides under a coverslip for morphological identification whereas the rest of the body was stored at −20°C for DNA extraction. The number of setae on the two inner faces of the coxites was used to differentiate P. longicuspis LCx (LCx, 19-22 setae) and P. longicuspis s.s (LC, 21-31 setae) whereas male copulatory valves were used to differentiate atypical and typical form P. perniciosus specimens, both with 10-16 coxite setae (PN) (Pesson et al., 2004).
Density (sand flies/trap/night), abundance (% specimens of a given species/total sand flies) and frequency (% positive sampling stations for a given species) data were calculated. Mean density values, 95% confidence interval (CI 95%) of the mean density, and minimum and maximum values of density in the sampled households were calculated for the peripheral neighbourhood of Oulad Bouchair compared to the other neighbourhoods in El Borouj. Statistical significance (p value) and odds ratio were determined by logistic regression. Software package IBM SPSS Statistics version 21.0 was used for the statistical analysis.

| Sand fly DNA extraction
Genomic DNA was extracted from the head, thorax and attached anterior abdomen of individual Larroussius males and females (Martín-Sánchez et al., 2000). A commercially available kit was used (RealPure kit from REAL, Ref. RBMEG01), according to the manufacturer instructions. Each sand fly was individually placed in a sterile 1.5 mL Eppendorf tube and kept in liquid nitrogen for a few seconds to facilitate the mechanical rupture of the tissues using a pestle. The DNA was resuspended in 20 μL bidistilled water and kept at −20°C until use.

| Sequencing and comparative sequence analysis
Amplified PCR products were eluted from agarose gel using Real

| Blood meal identification
A PCR-RFLP method using universal primers complementary to the conserved region of the mtDNA Cytb gene in vertebrates was applied for the identification of the meal source in sandflies; selective restriction enzyme cleavage of a short variable region of 359 bp with HaeIII and HinfI enzymes was performed (González et al., 2015).
DNA samples from potential hosts of different taxa were included as reference controls. When necessary, sequencing and comparative sequence analysis were also used to identify the hosts.

| Abundance and density of Larroussius species
Overall 2227 sand flies were intradomiciliary captured of which 1158 were male specimens (52.0%) and 1069 females (48.0%). Phlebotomus sergenti was the most abundant species captured (47.42%). Table 1 shows The third cluster is only composed of two morphologically identified P. longicuspis LCx males and supported by a bootstrap value of 79.9% (Figure 1). A similar topology including the same three clusters containing the same individuals was found through ITS2 analysis ( Figure 2) and concatenate trees but supported by lower bootstrap values (Appendix S1); for all three trees, cluster 1 and 2 are sister groups. COI analysis adding more Larroussius species showed clusters 2 and 3 as sister groups. (Figure 3 and Appendices S2 and S3). Genetic distances within clusters and between clusters ranged from 0 to 0.019 and 0.045 to 0.072, respectively whereas P. langeroni displayed genetic distances values ranging from 0.060 to 0.079 when compared to clusters, and pairwise distance between P. ariasi and P.

F I G U R E 1
chadlii was 0.038 (Appendix S4).
The comparative sequence analyses allowed the identification of eight Cytb, 10 COI and 28 ITS2 haplotypes. The fixed differences that characterize each cluster are shown in Table 3. The phylogenetic relationships among the 16 haplotypes of the P.
perniciosus complex are shown in Figure 4 and Appendices S5-S7.

| Leishmania infantum DNA detection
Leishmania infantum was detected in one of nine females morphologically classified as P. perniciosus and 7 of 56 females identified as P.
longicuspis s.l. based on morphology. The female P. perniciosus found infected showed 100% homology with other P. perniciosus specimens in the subcluster 1.2 of the tree of Figure 1. Five P. longicuspis females that were found infected showed 100% homology with the Cytb haplotype of cluster 2 from Figure 1. The sixth infected P. longicuspis female differed in two to three bases from the two male specimens included in cluster 3 of Figure 1, showing as a new Cytb haplotype in Figure 4.
The seventh infected P. longicuspis female belonged to subcluster 1.2.
Parasite loads were low in all of them (<10 parasites/sand fly). determine the vertebrate host source of the blood meal. The origin of blood meals was identified as: Homo sapiens for three P. perniciosus and four morphological P. longicuspis from subcluster 1 (1), cluster 2 (2) and cluster 3 (1) from Figure 4; Homo sapiens + Felis catus for one P. perniciosus and two morphological P. longicuspis from cluster 2; Felis catus, Bos taurus and Equus asinus for, respectively, 2, 1 and 1 morphological P. longicuspis from cluster 2; and Ovis aries for 2 P.

| DISCUSS ION
Phlebotomus sergenti was the most abundant and densest species within households in the recently endemic focus of ACL in El Borouj higher altitudes Franco et al., 2010;Mahamdallie et al., 2011;Rioux, 1995;Zarrouk et al., 2016) as opposed to the semi-arid or arid character of the sampled areas of El Borouj.
Phlebotomus langeroni is found from Lebanon to Spain (Esseghir et al., 2000;Haddad et al., 2003;Sáez et al., 2018) and is poorly Note: Underlined nucleotides are synapomorphic characters diagnostics of the three main lineages after comparing all the Cytb sequences of the specimens studied with the Cytb sequences published by Boudabous et al. (2012) and Depaquit et al. (2005) (16 Cytb haplotypes compared), and the data provided by Pesson et al. (2004) for which sequences are not published. Population genetics will help to assess the threat of the geographical spread of L. infantum in Morocco by determining the density, abundance and vector role of the species of the P. perniciosus complex identified correctly. Although the three genes tested are efficient markers to differentiate the taxa involved, we highlighted the use of Cytb which has been more intensively studied and has proven capable of defining more precisely its synapomorphic characters. Furthermore, Cytb is known to be phylogenetically informative for Phlebotomus species and useful for dating speciation events because of its clock-like rate of nucleotide substitution (Esseghir et al., 1997(Esseghir et al., , 2000.

| CON CLUS ION
Our study of a sympatric population of these Larroussius using three different genes evidence that some females with typical morphology of P. longicuspis are genetically homologous to P. perniciosus; these females and also those with typical morphology of P. perniciosus were naturally infected by L.infantum. Moreover, it shows the existence of P. longicuspis s.s. and one undescribed morphologically indistinguishable species, both found naturally infected by L. infantum. These taxa cannot be differentiated by morphological methods but characterized by a distinctive genetic lineage for which synapomorphic diagnostic characters are described.

ACK N O WLE D G E M ENTS
This study was funded by the University of Granada (Centro de Iniciativas de Cooperación al Desarrollo, CICODE, 2013).

FU N D I N G I N FO R M ATI O N
This study was funded by the University of Granada (Centro de Iniciativas de Cooperación al Desarrollo, CICODE, 2013). Funding for open access charge: Universidad de Granada/CBUA.

CO N FLI C T O F I NTER E S T S TATEM ENT
We (all authors) declare no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.