Transmission of visceral leishmaniasis in a previously non-endemic region of Tunisia: Detection of Leishmania DNA in Phlebotomus perniciosus



Visceral leishmaniasis (VL) has been endemic in northern Tunisia and has occurred sporadically in the center of Tunisia. Recently, there have been several cases from areas known to be free of VL. We report in this work all human and canine cases of VL recorded between 2003 and 2011 and an entomological study of phlebotomine fauna in a previously non-endemic region. Sixty-three cases of VL were diagnosed and identified as L. infantum using several different methods. Eight species of 179 sand flies were caught and identified by both morphological and molecular methods. Two genera were present, Phlebotomus and Sergentomya, with an abundance of the subgenus Phlebotomus (Larrousius) spp., a classic vector of VL in Tunisia. Moreover, Leishmania DNA was detected in seven unfed Phlebotomus pernicousus and L. infantum was identified in three of them. This result confirms the establishment of a transmission cycle of VL in the studied region by the coexistence of infected vectors with infected hosts.


Visceral leishmaniasis (VL) is a parasitic disease of humans, caused by the protozoan Leishmania infantum. In the Mediterranean focus, this parasite is transmitted by female sand flies (Diptera: Phlebotominae) of the subgenus Larroussius (Killick-Kendrick 1990, Killick-Kendrick 1999, Ready 2010) with the domestic dog as the main reservoir. In Tunisia, VL is known to be endemic in the north (Anderson et al. 1938, Vermeil 1956, Chadli et al. 1968) with an incidence of 100 cases/year, and occurs occasionally in the center and the south (Aoun et al. 2009). Recently, an increase of human VL incidence in VL foci has been recorded with many repeated sporadic cases in non-endemic regions (Ayadi et al. 1991, Pousse et al. 1995, Ben Salah et al. 2000, Aoun et al. 2009). Also, an increase in the prevalence of canine VL has been observed in many regions of central Tunisia (Chargui et al. 2007, Chargui et al. 2009). These reviews suggest an increase in the distribution area of VL. However, none of these studies have confirmed the establishment of VL transmission life cycle in previously non-endemic regions by also studying the phlebotomine populations. In our present work, we first report a retrospective study of all cases of VL (human and canine) observed between 2003 and 2011. Epidemiological and clinical analysis has been conducted on cases diagnosed by molecular, in vitro culture, and serological methods based on the detection of Leishmania, Leishmania DNA, and anti-Leishmania antibodies. Secondly, an entomological study of sand flies in regions where positive cases were observed has been conducted, with the aim of understanding the actual state of VL in a previously non-endemic region in central Tunisia.


Study area

Monastir is located in the eastern central region of Tunisia (35°31′/35°36′N; 10°39′/10°47′E) at 1 m above sea level. The terrain of the 1,0242 km region is generally flat with many olive plantations and about 479,000 inhabitants. The climate is semi-arid with annual average rainfall of 346 mm. The mean temperature ranges between 21 and 32° C in summer and between 7 and 16° C in winter. The animal fauna is composed primarily of pet animals (dogs and cats) and livestock (cattle, sheep, horses, and poultry).

Cinical samples

Forty-one patients at the Fatouma Bourguiba Hospital from 2003 to 2011, with symptoms in agreement with leishmaniasis, were hospitalized. For each of them, epidemiological (age, geographical origin, travel history to endemic areas) and clinical (fever, anemia, spleen enlargement) data were collected. Additionally, blood from 22 dogs originating from the Monastir region was sent from veterinary clinics to the parasitology laboratory for leishmaniasis diagnosis. The owners of dogs provided informed consent and information about their pets' age, sex, breed, origin, and travel to endemic regions. Animals were examined for disease signs, including dermatological and ocular changes, loss of weight, apathy, and lymph node and spleen enlargement. Three tubes of blood (3–5 ml) were taken with EDTA-added tubes for molecular methods, with citrate added for in vitro culture, and with tubes without anti-coagulant for the indirect immunofluorescence antibodies test (IFA). After centrifugation at 3500 rpm for 10 min, sera were used in IFA and Buffy coats were tested by PCR and for in vitro culture.

Sand fly sampling and species identification

Sampling of sand flies was carried out in July, 2010, during the period of peak sand fly activity, at eight collection sites within rural regions of the Monastir area where positive VL cases were observed. Phlebotomine sand flies were caught using CDC miniature light traps (John W. Hock, Gainesville, FL). Each trap was set for one night, from 18:00 to 06:00. Sand flies were stored in 96% ethanol. The head and the posterior part of the abdomen of a single sand fly were cut off in a drop of ethanol, then cleared in boiling Marc-André solution and mounted between slide and cover slide for morphological identification by examination of the head and spermathecae (females) or head and genitalia (males). The females belonging to the subgenus Larroussius were identified by observing the base of the spermathecal ducts according to Leger et al. (1983). The males of P. perniciosus have been identified by counting the number of coxite hairs (Pesson et al. 2004). PCR 12S followed by sequencing (Beati et al. 2004) was used for some samples for which morphological identification of the sand fly species was not possible. Molecular analysis was conducted on specimens stored in 50 μl PBS1X at −20° C before DNA extraction.

In vitro culture and identification

IFA was performed using a standard protocol described previously (Chargui et al. 2007). In brief, cell culture-derived promastigote forms of Leishmania infantum MON-1 MHOM/FR/78/LEM75 served as the antigen. Sera were screened at dilutions of 1:20 to 1:2560 in phosphate-buffered saline at pH 7.2. A human or canine anti-immunoglobulin coupled to Fluoline was used as conjugate, at a dilution of 1:30 in PBS with 0.01% Evans blue. The specimens were observed with a fluorescence microscope.

Buffy coat of human and canine samples were inoculated on blood agar base containing 10% rabbit blood and 40 mg/ml of Gentamicin. Cultures were examined at weekly intervals for one month. Isoenzymatic identification was performed for positive cultures as described previously according to Rioux et al. (1990).

Molecular detection and identification of Leishmania

DNA extraction was performed on human and canine samples as described by Lachaud et al. (2001). Briefly, buffy coats from human or canine samples were placed in a lysis buffer of TNNT (0.5 % Tween 20, 0.5 % Nonidet P40, 10 mM NaOH, 100 mM Tris HCl pH 7.6), and Proteinase K was added to a final concentration of 200 μg/ml for human samples and 900 μg/ml for canine ones. The samples were incubated at 56° C for two to 12 h followed by phenol/chloroform purification and ethanol precipitation. Genomic DNA from each female sand fly body (thorax, wings, legs, and abdomen) belonging to subgenus P. (Larrousius) was extracted using the QIAmp DNA Mini Kit (Qiagen, Germany). In addition to classic controls (negative with water and positive with in vitro cultured Leishmania), a negative control (male phlebotomine DNA) and a positive control (male phlebotomine with 20 Leishmania) were added to confirm specificity of PCR with phlebotomine DNA samples. PCR was performed as described previously (Chargui et al. 2012). Two PCR methods were used: PCR-ITS1 (Schonian et al. 1996) was conducted first, and if positive, the Leishmania species was determined by SSCP. If negative, PCR Lei (Spanakos et al. 2002) was used for detection of Leishmania DNA without possible species identification. Also, to confirm results of PCR Lei and PCR ITS, PCR products were sequenced and blasted using the Blastn algorithm against the “non-redundant” GenBank sequence database.


Clinical VL

There were 41 positive human cases determined by at least one method (PCR, IFA, and in vitro culture). Positive result was obtained for 38 cases (92%), 41 cases (100%), and for 18 cases (35%) by PCR, IFA and in vitro culture, respectively (Table 1). Twenty cases were identified at the species level by SSCP as L. infantum. With isoenzymatic methods, two zymodems were identified: MON-1 (eight cases) and MON-24 (two cases). All human positive cases were from the Monastir region without any history of travel to endemic regions. The major clinical signs in human LV cases in this study were fever, anemia, and spleen enlargement in 100%, 70%, and 84.6%, respectively. VL is mostly a childhood disease; the disease was present in children under five years in 73.3% of cases and adults in 7.3% of cases. Furthermore, the mean annual incidence was estimated at 5.1 cases per year, with a minimum of two and maximum of ten. At least one case of human VL has been recorded in all but three areas (77%) since 2003. Concerning canine VL cases, twenty-two dogs were positive by PCR and IFA, and 11 by in vitro culture. The species identification was possible for 15 of them. L. infantum MON-1 was the only species detected. All but two dogs were symptomatic (skin lesions, liver and spleen enlargement, and/or onychogryphosis). The mean age was 2.7 years and the sex ratio of M/F=1.2. Almost half of the positive canine cases were hunting dog breeds (Braque, pointer, English Pointer, Brittany). All but three hunting dogs have been in endemic regions of Tunisia (north) at least once.

Table 1.  Leishmania detection and species identification in clinical and phlebotomine sand fly samples by different methods (in vitro culture, IFA and PCR).
GroupPositive casesSpecies identification
In vitro cultureIFAPCRTotalPCR SSCPIsoenzyme
  1. ND: not done

Human cases18413841 L. infantum (20)MON-1 (8)MON-24 (2)
Canine cases7222222 L. infantum (15)MON-1 (7) -
Sand fliesNDND77 L. infantum (3)
Total25636770 L. infantum (38)MON-1 (15)MON-24 (2)

Phlebotomine sand fly fauna

A total of 179 sand flies (95 males and 84 females) was captured by CDC miniature light traps installed in stations where positive human or canine VL cases had occurred. Female phlebotomine sand flies were unfed (27.3%), gravid (8.9%), or engorged (10.6%). Species identification was conducted morphologically for most (92%) sand flies. In addition, the molecular method with PCR 12S, followed by sequencing, was conducted for 13 females for which morphological identification was difficult or not possible. Sand flies belonged to eight species, including six species in the genus Phlebotomus (P. perniciosus, P. papatasi, P. chabaudi, P. longicuspis, P. riouxi, and P. sergenti) and two belonging to the genus Sergentomyia (S. minuta and S. fallax) (Table 2). Phlebotomus perniciosus was the most common species representing 48% of the total, followed by Phlebotomus (L.) papatasi with 27.3%. Four other Phlebotomus species were present but in less abundance: P. chabaudi (3.3%), P. longicuspis (2.2%), P. sergenti (1.6%), and P. riouxi (1.6%). Furthermore, 15.6% belonged to the genus Sergentomya that have no role in Leishmania transmission.

Table 2.  Phlebotomine sand flies caught in Monastir province by CDC traps in July, 2010. Thumbnail image of

Leishmania-infected sand flies and parasite typing

Twenty-eight sand fly females belonging to the subgenus Phlebotomus (Larroussius) were tested for the presence of Leishmania DNA. Fifty-six percent of them were unfed, 22% gravid, and 19% engorged. Leishmania DNA was detected in seven unfed females by PCR Lei (Table 1). Leishmania species identification was possible in only three P. perniciossus and belonged to L. infantum. To confirm the results of PCRs, PCR products were sequenced and blasted using Blastn algorithm against the “non-redundant” GenBank sequence database. The sequences obtained from PCR Lei70 were similar to the Leishmania spp. sequence (99%) available in the data base (acc no: HQ535858) and those of the PCR ITS product show 93% homology with L. infantum (acc no: FJ497004).


An epidemiological investigation of VL was conducted in a region known to be non-endemic except for a few sporadic cases. This investigation included clinical (human and canine) VL cases as well as the analysis of phlebotomine sand fly fauna. The incidence of VL before 2000 in the Monastir region was less than one case per year (unpublished data). However, a retrospective analysis of cases occurring since 2003 indicate there is an increase in the incidence to an average of 5.1 cases/year. The clinical and epidemiological aspects of these cases are similar to the infantile VL form of Tunisian foci. Indeed, the major clinical signs present in human LV cases in this study were fever, anemia, and spleen enlargement. Only L. infantum has been identified in both human and canine samples, confirming again that only this species is responsible for VL in Tunisia and the presence of both zymodems (MON-1 and MON-24) in this region with a predominance of MON-1 as in the classic focus of LV (Aoun et al. 2008, Kallel et al. 2008). Concerning the geographic distribution, VL was present in almost all districts of Monastir with a higher rate in rural regions (56% of cases). Concerning positive canine cases, at least two cases of canine VL cases have been observed in eight districts (61.5%). Almost half of the dogs that were positive were hunting dogs. Some of them have been in endemic regions for at least one season of transmission where they would have contracted the disease. The peak season of sand flies is between May and November, when females need one or more blood meals to develop their eggs. Sand flies could acquire the parasites when feeding on infected local dogs and transmit Leishmania to non-infected dogs. On the other hand, three other infected dogs that never left Monastir could have contracted the disease by living close to infected dogs transmitted by phlebotomine sand flies. To confirm the non-sporadic aspect of VL in studied regions it was necessary to study the sand fly population. A total of 179 sand flies was caught in eight different stations. Most sand flies belonged to the genus Phlebotomus (82.6%), including P. perniciosus, P. papatasi, P. chabaudi, P. longicuspis, P. riouxi, and P. sergenti that are involved in the transmission of different leishmaniasis in many parts of the world, and to the genus Sergentomyia (S. minuta and S. fallax) that are not vectors of Leishmania. P. perniciosus was captured most often (48%) and the abundance of P. longicuspis was low (2.2%). Furthermore, two new species were detected and not previously observed in this region, P. riouxi and P. sergenti. These differences might be due to the difference in the period of collections and to the type of traps used (Boudabous et al. 2009, Jaouadi et al. 2012). Indeed, Boudabous et al. (2009), caught sand flies during different periods between September and November and used different methods, including both sticky and CDC traps.

We detected Leishmania in seven unfed sand flies by PCR Lei. L. infantum infection was confirmed by PCR ITS, SSCP, and sequencing in three phlebotomine sand flies. In four cases, species identification was not possible due to the difference in sensitivity between tested PCRs (Chargui et al. 2012). Leishmania infantum was detected in three unfed P. perniciossus. In Tunisia, the major vector of VL is Phlebotomus (Larrousius) perniciosus (Killick-Kendrick 1999). P. langeroni, also confirmed as a vector of VL in the north (Guerbouj et al. 2007), seems to be absent in this region (Boudabous et al. 2009), suggesting that there may be different types of transmission cycles in the north. The analysis of the spatial repartition of infected human, canine, and phlebotomine sand flies shows that Leishmania infection has been detected in almost all districts (85%), with at least two cases in each district suggesting the non-sporadic aspect of VL in this region. Furthermore, infection risk was confirmed by detection of Leishmania in unfed phlebotomine sand flies in two districts. Further studies on sand fly populations in all districts and within a longer transmission period could detect more infected flies. Because sand fly dispersal is low and always below 600 m (Killick-Kendrick et al. 1984), and because parasites survive in only permissive sand flies after blood meal defecation (Bates et al. 2007), the detection of Leishmania in unfed Phlebotomine sand flies in a region where human and or canine VL cases were recorded confirms the non-sporadic aspect of VL and the establishment of life cycle transmission of Leishmania in this region by the co-existence of infected phlebotomine sand flies and infected human and dogs. Also, we suggest a new geographic repartition of the disease in Tunisia, as many repeated cases of VL (human and/or canine) have been observed in other regions of the center and south (Chargui et al. 2007). In addition to the analysis of phlebotomine fauna in these regions (Zhioua et al. 2007, Boudabous et al. 2009), Leishmania should be detected in sand flies to confirm the spread of VL from the north to the center and south.