Ecology, feeding and natural infection by Leishmania spp. of phlebotomine sand flies in an area of high incidence of American tegumentary leishmaniasis in the municipality of Rio Branco, Acre, Brazil

Background Phlebotomine sand flies (Diptera: Psychodidae) are insects of medical importance due to their involvement in the zoonotic transmission of Leishmania spp. to vertebrates. The aim of this work was to study the ecology of the sand fly fauna of two types of environments, a rural environment (the Transacreana Road) and an urban park (Horto Florestal Park), both located in the municipality of Rio Branco in the state of Acre, Brazil. Additionally, this study intended to investigate Leishmania infection and blood meal sources of these sand flies using molecular techniques. Methods The sand fly fauna was studied in different environments (i.e. forest and peridomestic environments in a rural area, and an urban forest) using Shannon traps and HP light traps to collect sand fly specimens over 13 consecutive months (December 2014 to January 2016). For investigating natural infection by Leishmania and the source of sand fly blood meals, DNA samples were extracted from female sand flies and subjected to polymerase chain reaction targeting ITS1 and cytb genes. DNA sequencing was subsequently used to identify species of Leishmania and the source of blood meals. Results A total of 2515 individual sand flies of 43 species were collected and identified, Trichophoromyia auraensis (839; 33.35%), Trichophoromyia spp. (537; 21.35%) and Evandromyia saulensis (187; 7.43%) were more abundant in the rural area (S = 41 species) than in the urban forest. No significant differences were found in species richness between forest and peridomestic environments in the rural area (H = 0.04; P > 0.05), but a larger number of species was found in the forest. Leishmania DNA was sequenced in 13 samples, confirming the presence of L. (V.) braziliensis in Th. auraensis (n = 1), Ev. saulensis (n = 2), Ev. walkeri (n = 1), Ps. llanosmartinsi (n = 1), Pi. nevesi (n = 2), Ps. davisi (n = 1), Ps. ayrozai (n = 1), Pa. aragaoi (n = 1), Ny. antunesi (n = 1) and Ev. infraspinosa (n = 1). Only Ps. ayrozai possessed a sequence similar to that of L. (V.) guyanensis (99%). Through microscopic analysis, five specimens of Ev. saulensis were found to possess flagellate forms in the hindgut, with an infection rate of 2.4%. Samples from 33 fed females were submitted to cytb gene amplification, for which sequencing determined that all were similar to the sequence deposited on GenBank for Gallus gallus (domestic chicken). Conclusions The high abundance of Trichophoromyia auraensis and Ev. saulensis, and the detection of L. (V.) braziliensis DNA, suggests that both species may be vectors of American tegumentary leishmaniasis. Psychodopygus ayrozai was found to be infected by L. (V) braziliesnsis and L. (V.) guyanensis, and although collected in low abundance, it may be a potential vector in the region. The sand fly fauna was found to be rich and diverse with predominance of the genus Psychodopygus. Identification of food sources of fed females showed that 100% amplified a gene region compatible with the domestic chicken, which although considered refractory in the disease transmission cycle, may have an influence on the population dynamics of sand flies.


(Continued from previous page)
Conclusions: The high abundance of Trichophoromyia auraensis and Ev. saulensis, and the detection of L. (V.) braziliensis DNA, suggests that both species may be vectors of American tegumentary leishmaniasis. Psychodopygus ayrozai was found to be infected by L. (V) braziliesnsis and L. (V.) guyanensis, and although collected in low abundance, it may be a potential vector in the region. The sand fly fauna was found to be rich and diverse with predominance of the genus Psychodopygus. Identification of food sources of fed females showed that 100% amplified a gene region compatible with the domestic chicken, which although considered refractory in the disease transmission cycle, may have an influence on the population dynamics of sand flies.

Background
Leishmaniasis is a complex of vector-transmitted zoonotic diseases, of great clinical importance and epidemiological diversity. It is caused by protozoan parasites of the genus Leishmania (order Kinetoplastida, family Trypanosomatidae) in vertebrate hosts including man. Transmission is performed by hematophagous insects of the subfamily Phlebotominae (family Psychodidae), known as sand flies, which have complex ecological behavior [1][2][3]. There are two main clinical types of leishmaniasis, which are caused by different protozoan species of the genus Leishmania: visceral leishmananiasis (VL) in which the etiological agent is Leishmania infantum and tegumentary leishmaniasis (TL) that affect the skin and mucous membranes [1,4]. American tegumentary leishmaniasis (ATL) is a significant disease in Brazil due in part to the great abundance of sand fly vectors of the various species of Leishmania that occur in the country [3][4][5][6]. With 20,187 cases reported in 2015 [7], ATL has been reported in all states, suggesting adaptation of the parasites and their vectors to anthropic environments [8][9][10][11]. They have adapted to urban environments that are surrounded by to forest remnants, also because of the greater supply of food sources for the vectors in the form of domestic animals in peridomestic areas [11,12]. In the North Region of Brazil, ATL is endemic with 45% (9278) of the confirmed cases for the country in 2015 [7]. This region is characterized by human occupation associated with agricultural activities, mineral extraction and intense deforestation. It is also a region possessing low human development indices, with the absence of basic services such as sanitation in most residences, which are considered potential sites for the occurrence of tropical diseases [12][13][14].
In the state of Acre, located southwest of the Amazon in Brazil, ATL is considered endemic with high rates of incidence and prevalence [11,15,16]. Of the few studies that have been undertaken in the region, there have been reports of Leishmania (Viannia) guyanensis, Leishmania (V.) braziliensis, Leishmania (V.) lainsoni and Leishmania (Leishmania) amazonensis [17][18][19]. Likewise, these studies have reported the sand fly species Bichromomyia flaviscutellata, Migonemyia migonei, Nyssomyia whitmani, Ny. antunesi, Ny. umbratilis, Psychodopygus ayrozai, Ps. davisi, Ps. paraensis and Trichophoromyia ubiquitalis as the main vectors of the etiological agents of leishmaniasis [11,16,[20][21][22]. Rio Branco is a municipality in Acre with the greatest occurrence of the disease, with 500 ATL cases reported between 2013 and 2014 [7]. However, studies of leishmaniasis are still incipient, especially with regard to sand fly ecology and food sources, and the transmission cycle of Leishmania spp. In addition, little information has been acquired regarding natural infection of possible vectors by dissection and observation of trypanosomatids [23][24][25], or by molecular methods such as polymerase chain reaction (PCR), which are capable of identifying Leishmania DNA, such the studies carried out in Assis Brasil [22] and Rio Branco [26].
The objective of this study was to investigate the sand fly fauna of rural and urban environments, and determine their food source and natural infection by Leishmania in an endemic area of ATL in Rio Branco.

Study area
The municipality of Rio Branco (09°59′11"S, 67°49′ 52"W), capital of the state of Acre (AC), is located in the southwestern part of the Amazon Region of northern Brazil (Fig. 1a). It encompasses an area of 8835.52 km 2 [27], and possesses an estimated population of 377,057 inhabitants [27]. The vegetation is composed of dense and open tropical forest with predominance of bamboo and palm trees. The climate is equatorial, with temperatures varying from 24°C to 32°C and annual rainfall of between 1877 and 1982 mm. The hottest period of the year corresponds to the months of July to August, marked by intense fires due to the conversion of forest to pastures [28].
To determine sand fly anthropophily, nocturnal sampling (06:00 pm to 08:00 pm) was performed by 2 persons using a Shannon trap in an area of peridomiciliar forest at Dom Joaquim Settlement during 13 nights. The females collected during this sampling were killed with ethyl acetate and dissected on slides with a drop of sterile saline solution to expose and examine the digestive tract and spermathecae for flagellate research and species identification, respectively. The specimens were then stored, in microtubes containing absolute ethanol, in pools of up to 10 females according to the month of collection and species, for subsequent molecular detection of Leishmania spp.
Fed females collected in HP light traps had only their genitalia exposed for species identification, and then were stored in 90% absolute ethanol for food source analysis.
All the remaining specimens collected in the Shannon trap were clarified according to Forattini [29] and identified according to the taxonomic key proposed by Galati [30]. The abbreviation of species names followed Marcondes [31].

DNA extraction and PCR conditions
After being identified, female sand flies were subjected to DNA extraction using the Gentra Puregene® Cell and Tissue Extraction Kit (QIAGEN, Hilden, Germany), following the manufacturer's protocol.
After DNA extraction, PCR was performed targeting the internal transcribed spacer 1 (ITS1) using the primers LITSR -Forward (5′-CTG GAT CAT TTT CCG ATG-3′) and L5.8S -Reverse (5′-TGA TAC CAC TTA TCG CAC TT-3′) [32], to detect natural infection by Leishmania spp. In non-engorged females, the region amplifies a 350 bp fragment by means of the following reaction: 1× buffer solution (200 mM Tris-HCl, pH 8.4; 500 mM KCl), 1.5 mM MgCl 2 , 0.2 mM dNTPs, 0.5 pmol of the LITSR primer, 0.5 pmol of the L5.8S primer, 1 U of Taq DNA Polymerase Platinum (Invitrogen, California, USA) of 5 μl template DNA in a total volume of 25 μl. Samples were amplified in an automatic thermocycler (MaxyGene Gradient -AXYGEN, Corning, USA) with 33 cycles of denaturation at 95°C for 30 s, annealing at 53°C for 1 min and extension at 72°C for 1 min. In negative control, pure water was used with DNA from male sand flies. The amplified product was analyzed on a 2% agarose gel stained with GelRed (Nucleic Acid Gel Stain -Biotium, Fremont, USA) and compared with amplified material from positive samples from reference strains of Leishmania braziliensis (MHOM/BR/75/M2903). Positive samples were cloned into competent Escherichia coli (DH5-α) cells using the CloneJet PCR cloning kit (Thermo Scientific, California, USA) following the manufacturer's protocol. The colonies of transformed bacteria were selected, re-submitted to PCR of the ITS1 gene and sequenced.
Blood-feeding study and PCR conditions DNA extracted from blood-fed female sand flies was submitted to PCR for amplification of the Cytochrome B gene, which amplifies a fragment of approximately 359 bp using the cytb primers (forward: 5′-CCA TCC AAC ATC TCA GCA TGA TGA AA-3′; and reverse: 5′-GCC CCT CAG AAT GAT ATT TGT CCT CA-3′) [33]. DNA extraction from the females followed the same protocol as described previously but in a laminar flow booth and adopting disinfection measures with care not to contaminate samples with human DNA. The PCR reaction was performed using 1× buffer solution (200 mM Tris-HCl, pH 8.4; 500 mM KCl), 1.5 mM MgCl 2 , 0.2 mM dNTPs, 0.5 pmol LITSR primer, 0.5 pmol L5.8S primer, 1 U of Taq DNA polymerase platinum (Invitrogen, California, USA) and 5 μl template DNA, in a total volume of 50 μl. Samples were amplified in an automatic thermocycler (MaxyGene Gradient -AXY-GEN, Corning, USA) with 32 cycles of 95°C for 20 s (denaturation), 53°C for 30 s (annealing) and 72°C for 1 min (extension). As negative control, pure autoclaved water and DNA of male and female sand flies without the presence of intestinal blood were used. Five μl of the amplified product was analyzed on a 2% agarose gel stained with 0.1 μl of GelRed (Biotium Acid Gel Stain -Biotium, Fremont, USA) and compared to the band obtained by positive control (DNA amplified from canine clot).

Sequencing
For both the study of natural infection and that of food source, amplified PCR products were sent to Macrogen (South Korea) for sequencing using a 3730XL Applied Biosystems. The sequences obtained were edited using the program Sequencher, compared by BLAST with sequences available in GenBank.

Data analysis
Kruskal-Wallis was used to test the hypothesis of equality among the ecological descriptors in the different ecotopes studied. The measures of diversity among the environments were obtained using the Shannon-Wiener's (H′) diversity index, the Pielou's equitability index (J'), Margalef's richness index and the Berger-Parker's dominance index [34], followed by comparisons and Spearman's correlation tests between species and climatic variables (temperature, relative humidity and rainfall). The analyses were performed using the Past 2.17c program and IBM SPSS 20 software, with a 5% level of significance. The minimum natural infection rate was calculated considering the number of positive pools multiplied by 100 and divided by the number of insects tested [35].
Evandromia saulensis (n = 66) was the most frequently captured species in Shannon traps, indicating the anthropophilic behavior of this species, followed by Psychodopygus davisi (n = 48). A total of 2304 specimens of 41 species were collected at rural sites, with the most abundant being Th. auraensis, Trichophoromyia sp., Ev.saulensis, Pr. calcarata and Ps. carrerai carrerai. In the urban park, 213 specimens of 26 species were collected, with the most abundant being Th. auraensis, Trichophoromyia sp. and Ny. antunesi. Sand fly abundance was higher in the forest and peridomestic environments in the rural area than in the urban forest (H = 17.9, df = 42, P < 0.05) (Fig. 2). In the rural area, mean species richness was 5.16 (S = 41) and mean equitability (J') 0.61, whereas in the urban park the mean richness was lower at 4.66 (S = 26), and equitability higher (J' = 0.72), since it is an environment with a more uniform pattern of abundance. The diversity index was higher in the urban forest, but not significantly different from that of the rural area according to the Kruskal-Wallis test (H = 1, df = 1, P > 0.05) for two samples. Species dominance, according to the Berger-Parker's index, averaged 0.33 and 0.30 for the rural area and urban forest, respectively. There was a significant correlation between the total number of sand flies collected and mean temperature on the days of collection (Spearman's rho = 0.58, n = 13, P = 0.03) ( Table 3); however, there was no significant correlation between the number of sand flies collected by HP light traps and relative humidity (Spearman's rho = -0.66, n = 13, P = 0.831) or monthly rainfall (Spearman's rho = -0.38, n = 13, P = 0.45) during the study period. Species richness for forest and peridomestic environments in the rural area did not differ significantly (H = 0.04, df = 1, P > 0.05), although there was a greater number of species in the forest. The greatest diversity (Shannon index, H′) of species was in the primary forest of the urban forest, with 17 species (H′ = 2.39), followed by rural peridomestic environments (H′ = 2.11) with 26 species, rural forest (H′ = 1.99) with 32 species and secondary forest in the urban forest (H′ = 1.85) with 20 species (Table 2).

Discussion
In the state of Acre, ATL and environmental impacts have historically been related to the processes of human occupation, which from the 1970's experienced a gradual conversion from the extractive model of the region to agricultural activity [28]. These environmental changes, associated with the construction of rural settlements, urbanization and massive logging with a view to human occupation, played a major role in the dissemination of this disease in the region and the state. The impacts resulting from the dynamics of land use in the Amazon have had environmental consequences such as high rates of deforestation, loss of biodiversity, adaptation of vector species to peridomestic environments, potentially leading to extinction of rare species or increasing others.
The increase in ATL cases reported in the North Region of Brazil suggests an adaptation of sand flies to transmit ATL agents in the peridomestic environment [11], also as a consequence of the impact of anthropic activities on the environment, including animal farming near houses. This may have favored the adaptation of the vectors to human dwellings and modified the transmission cycle of ATL [36]. Studies of geographical distribution [16], ecological indices of the fauna and rates of natural infection rates are scarce in Brazil. In this study, a total of 2515 sand flies were captured and the most abundant species sampled were Th. auraensis, indistinguishable females of Trichophoromyia sp., Ev. saulensis and Pressatia calcarata, altogether representing 73.8% of the collections. Of these, L. (V.) braziliensis was detected in Th. auraensis and Ev. saulensis, corroborating other studies in the state that point to Th. auraensis as an abundant species in the sampled environments in Acre [20][21][22]26] and its potential involvement in the  ) guyanensis by PCR-RFLP [22], indicating its importance in maintaining the circulation of Leishmania spp., although its vectorial competence has not yet been proven [37]. In Madre de Dios, Peru, Th. auraensis has been found infected with L. (V.) lainsoni and L. (V.) braziliensis [38], reinforcing the importance of this species and its possible capacity as a vector in these regions of the Amazon Basin. Among potential vectors of ATL agents to humans, Ev. saulensis has not been considered of medical importance. Although it has a large distribution from central Brazil to Central America, this species is not commonly collected in sand fly studies. However, this species has been found infected with L. (V.) braziliensis in Rio Branco [26], and in the present study was the third most frequent species in light-trap collections, the most abundant species in the collections using Shannon, being present in all environments sampled and with dominance in the forest environment, confirming its anthropophilic behavior. We also found flagellated forms in five specimens of Ev. saulensis, reinforcing its participation in the enzootic transmission cycle of ATL and the need for studies of the vectorial capacity and competence in transmitting Leishmania to humans. Furthermore, the potential participation of Ev. saulensis in the transmission of ATL agents, in addition to the vectors already incriminated needs to be investigated within the state of Acre. The genus Psychodopygus was the most dominant genus in the collections, with a total of seven species collected, of which Ps. carrerai carrerai and Ps. davisi were the most abundant in the study. Species of this genus are frequent in faunal studies of this region and

Note:
Equal letters indicate a non-significant correlation between the total number of sand flies collected and temperature, relative humidity and rainfall (P > 0.05); and different letters indicate a significant correlation (P < 0.05) have been found naturally infected in the Amazon Region [39][40][41][42]. In our studies, Psychodopygus davisi was found infected with L. (V.) braziliensis, which corroborate a previous study carried out in Assis Brasil where parasite of the L. (V.) braziliensis complex was found [22]. Reinforcing the importance of this species in the region is that it has been found naturally infected with Leishmania sp. in the state of Rondônia [43,44], and also infected by L. (V.) braziliensis in the Serra dos Carajás, Pará and in the neighboring country of Peru [38,42]. Psychodopygus davisi has been frequently found in peridomesitc environments, suggesting its adaptation to anthropic environments. Based on its anthropophilic behavior, density and occurrence of natural infection of Leishmania, this species is incriminated in the transmission cycle of ATL in the Amazon Region [37,[43][44][45]. The species Ny. whitmani, Ny. antunesi, Bi. flaviscutellata, Ps. hirsutus hirsutus, Ps. ayrozai and Migonemyia migonei were found in low abundance in the different environments studied, yet these are considered vectors of species of Leishmania in the region [20]. Psychodopygus ayrozai was found naturally infected with L. (V.) guyanensis, reinforcing its role in the etiological agent's circulation of ATL in the state. In other studies, L. (V.) guyanensis was isolated from human cases and in sand flies already described in other epidemiological studies [18,22]. Psychodopygus ayrozai is common in the North Region of Brazil [6], especially in peridomiciliar areas, indicative of its high degree of anthropophilia [14]. In the Amazon Region, species of Psychodopygus are associated with ATL transmission and can be important vectors in zoonotic and enzootic cycles [46]. Thricophoromyia ubiquitalis was described as a vector of L. (V.) lainsoni in the state of Pará [16,46]. In a study carried out near the Solimões River (AM), Th. ubiquitalis was found to be the predominant species, accounting for 57.94% of flies sampled, and was also found naturally infected by L. (V.) lainsoni [47]. In our study, only one specimen of Th. ubiquitalis was captured in a Shannon trap. Faunal studies of sand flies in three municipalities of Acre identified 52 species, among them Th. auraensis, Ny. antunesi, Ny. whitmani, and Ps. davisi were responsible for 66.95% of the collected flies, with Ny. whitmani being the most abundant [20]. In the present study, these species represented 40.7% of the species collected, with the most frequent being Th. auraensis (33.3%). In 2008, Silva-Nunes et al. [16] investigating the epidemiology of ATL and surveying the phlebotomine fauna in order to identify possible vectors in the municipality of Acrelândia, collected 14 species, three of which are known vectors of the disease: Ny. antunesi (59.1%), in the peridomestic environment and forests edges; Ny. whitmani, more frequent in peridomestic environment (15%) and the only one within a house; and Th. ubiquitalis, also in peridomestic environment. In this study, Ny. whitmany and Ny. antunesi were more frequent in forest environments.
In a faunal survey of three areas of the city of Rio Branco during the course of a single year, Araújo-Pereira et al. [21] found high diversity of sand flies, with the most abundant being Th. auraensis and Ny. whitmani, representing 72% of the collected specimens. Some of the species collected by these authors were already known as vectors of the parasite that causes ATL (i.e. Ny. whitmani, Ny. antunesi and Bi. flaviscutellata). These data corroborate the high frequency of Th. auraensis (33.3%) and other species known to be potentially vectors of Leishmania parasites to humans.
In Assis Brasil, Teles et al. [22] evidenced a rich and diverse sand fly fauna of 67 species, of which Th. auraensis/ruifreitasi and Ps. davisi were dominant. These species were collected in domestic and peridomestic forest environments, demonstrating their potential as vectors of ATL agents in the region.
In the present study, species diversity was lower in the rural area than in the urban forest, with a lower index of equitability due to high species dominance. Thus, sand fly frequency seems to be related to levels of environmental degradation and food supply provided by animals in peridomestic environments, with changing the behavior of those populations compared with sand flies in unaltered environments [48]. Thus, an environmental park interconnected to an area of environmental protection, even with low frequency and lower richness, suffers less anthropic effects, and thus tends to maintain the balance on species distribution in the ecosystem. The ecological indices in the different studied ecotopes did not show any significant differences, having the greatest diversity for primary forest in the urban park. Silva et al. [48] suggested that areas with vegetation in different successional stages may influence the dominance and frequency of sand flies. Also, high species diversity is common in the Amazon Region, as observed by other authors [37,43,47].
There was an increase in the abundance of sand flies collected with light traps at the beginning of the rainy season, from December to April, with smaller quantities during the dry months with low humidity and the occurrences of Amazonian cool periods. The species frequency had a significant positive correlation with temperature variation (Spearman's rho = 0.58, P = 0.030), indicating that with increasing temperature there is an increase for abundance of collected sand flies. Other studies have recognized a similar pattern of seasonal variation in sand flies such in Campo Grande, Mato Gross do Sul, with an increase in the abundance of captured specimens after the months of highest rainfall and humidity [49]. It should be noted that there are distinct seasonal patterns for sand flies in different regions of Brazil. In Rio Grande do Norte, a study found that sand flies were most abundant during the dry season, and declined in abundance in the rainy season when temperatures decreased and relative humidity was high [50].
All fed female sand flies submitted to analysis of blood sources exhibited an amplified gene region compatible with Gallus gallus (domestic chicken). Although chickens do not act as reservoirs of Leishmania spp., their presence may influence on the population dynamics of sand flies with increasing population density in the peridomestic environment, which could contribute to increase the risk of the transmission of Leishmania to humans [51]. The presence of animal shelters (e.g. stables and chicken pen) in the peridomestic environment represents a permanent food supply and provides shelter for sand flies, as observed in studies in the State of Maranhão, where sand fly sampling was unsuccessful in environments without the presence of domestic animals [52]. Again, the presence of large numbers of chickens in peridomestic environments can be highly attractive to sand flies and may contribute to the maintenance of their life cycle, and thus constitute an epidemiological problem by keeping the vector near the human environment [52,53]. It is worth noting that the presence of infected sand flies and the presence of synanthropic reservoirs can increment the transmission cycle in the peridomestic environment of the present study. Thus, in this study, domestic chicken appears to be a common blood source for sand flies, probably because it does not offer resistance and are kept in large number in the study area [53].

Conclusions
The sand fly fauna found in the present study was composed of 43 species and included known vectors of ATL, such as Nyssomyia whitmani, Ny. antunesi, Bichromomyia flaviscutellata and Mg. migonei. The high frequency of Trichophoromyia auraensis and Evandromyia saulensis, and the detection of L. (V.) braziliensis DNA, and Ps. ayrozai with L. (V.) guyanensis DNA, indicate that these species could be putative vectors for ATL in this Amazonian region. Investigation of blood sources of sand flies revealed a preference among female sand flies collected in this area for domestic chicken, which may be participating in the population dynamics of these insects.