Identification and Biological Characterization of Leishmania (Viannia) guyanensis Isolated from a Patient with Tegumentary Leishmaniasis in Goiás, a Nonendemic Area for This Species in Brazil

The aim of this study was to characterize clinical field isolates of Leishmania spp. obtained from patients with American Tegumentary Leishmaniasis (ATL) who live in Goiás state, Brazil. The presumed areas of infection were in Goiás, Tocantins, and Pará states. Three isolates of parasites were identified as L. (Viannia) braziliensis and one as L. (V.) guyanensis. The in vitro growth profiles were found to be similar for all parasites. Nevertheless, in C57BL/6 mice, L. (V.) guyanensis infection was better controlled than L. (V.) braziliensis. Yet in C57BL/6 mice deficient in interferon gamma, L. (V.) guyanensis lesions developed faster than those caused by L. (V.) braziliensis isolates. In BALB/c mice, the development of lesions was similar for isolates from both species; however, on the 11th week of infection, amastigotes could not be observed in macrophages from L. (V.) guyanensis-infected mice. Thus, L. (V.) guyanensis can be circulating in Goiás, a state where autochthonous cases of this species had not yet been reported. Considering the difficulties to differentiate L. (V.) guyanensis from L. (V.) braziliensis at the molecular, morphological, and clinical (human and murine models) levels, the presence of L. (V.) guyanensis infections is possibly underestimated in several regions of Brazil.


Introduction
Leishmaniases are protozoan diseases caused by more than 20 Leishmania species, which are transmitted by about 30 species of phlebotomine sand flies. Human infections cause three strikingly different clinical presentations and numerous clinical varieties ranging from asymptomatic to disfiguring forms of tegumentary and potentially fatal visceral leishmaniasis. American Tegumentary Leishmaniasis (ATL) presents a spectrum of clinical manifestations characterized by cutaneous (CL), mucosal (ML), disseminated (DL), and diffuse cutaneous leishmaniasis (DCL) [1,2]. Brazil together with other nine countries accounts for 70-75% of estimated CL cases in the world [3]. A report of the Brazilian Secretary's Office of Surveillance in Health showed a geographic expansion of ATL during the 1980s from the Northern

Materials and Methods
2.1. Mice. Female C57BL/6 (wild-type [WT]) or C57BL/6 IFNgamma knockout (IFN KO C57BL/6) and BALB/c mice, six to eight weeks old, were obtained from the breeding animal facilities of the IPTSP/Federal University of Goiás, Goiânia, Brazil. All the animal handling and procedures were approved by the Ethical Committee from Clinical Hospital of the Federal University of Goiás on the ethical handling of research animals.

Patients.
Four patients were assisted at the Tropical Disease Hospital (Anuar Auad, Goiânia, Goiás, Brazil) with the diagnostic hypothesis of leishmaniasis. All of them live in Goiás, but the presumed areas of infection were Goiás, Tocantins, and Pará/Maranhão border. Diagnosis of ATL was confirmed by epidemiological, clinical, and laboratory analyses. The patients' data are presented in Table 1. The protocols on this investigation relative to human patients and animals were approved by the Ethical Committee from Hospital das Clínicas, Universidade Federal de Goiás, and all patients signed an informed consent form.

Parasite Isolation and Cultures.
A fragment of the cutaneous lesion biopsy was macerated and inoculated in mice footpads (IFN KO C57BL/6) or directly cultured at 26 ∘ C in Grace's Insect Cell Culture Medium (Gibco-BRL Life Technologies, Grand Island, NY, USA) containing 20% heatinactivated fetal bovine serum (Cripion, Andradina, SP, Brazil), 2 mM glutamine, penicillin (100 U/mL), and streptomycin (100 g/mL) (supplements and antibiotics were purchased from Sigma Chemical Co., St. Louis, MO, USA). After 1-2 months draining lymph nodes from the animals were processed as described [21] and cultured in complete Grace's medium. After being expanded in culture, the parasite isolates were cryopreserved in liquid nitrogen. The parasite isolates were thawed and expanded once in complete Grace's medium before use in all experiments. All these procedures were previously described [21].  (PCR). The identification of the isolates was based on three strategies: (1) small subunit ribosomal RNA (SSU rDNA) was sequenced as previously described, using primers S12/S4 [24]. Positive control reactions were performed using a reference genomic DNA purified from axenic cultures of L.  MCEB/BR/84/M8408, while in negative controls no genomic DNA was added. The amplified product was analyzed in a 2% agarose gel electrophoresis stained with ethidium bromide. The nucleotide sequence of the 520 bp fragment was obtained directly by automatic sequencing using an ABI Big-Dye kit as described [23].
The amplified product was also purified from agarose gels using the QIAquick PCR purification kit (Qiagen, Valencia, USA) and cloned in pGEM-T easy (Promega Corporation, Madison, WI, USA). The nucleotide sequence of three independent positive clones, confirmed previously by restriction analysis, was determined as described above using pUC/M13, IR1, IR2, 5.8F (5 GCAGTAAAGTGCGATAAGTGG 3 ), and 5.8R (5 GGAAGCCAAGTCATCCATC 3 ) primers. Nucleotide sequence analyses were performed using Lasergene Software (DNASTAR) and Clone Manager 9.0 Software. Phylogenetic analysis was performed using RAxML [28].

In Vitro Growth of Leishmania
Isolates. The parasite isolates extracted from draining lymph nodes of infected IFN KO C57BL/6 mice were cultured at an initial concentration of 5 × 10 5 /mL in 24-well culture plates (TPP, Techno Plastic Products, Trasadingen, Swizerland) in complete Grace's medium at 26 ∘ C. Samples of parasites were counted daily for 13 days in a hemocytometer after dilution in 2% formaldehyde solution in PBS, under light microscopy.

Infection of Mice.
Groups of four mice were injected subcutaneously (s.c.) with 5 × 10 6 live promastigotes (50 L) in stationary phase of growth into the left hind footpad. Lesion development was followed by measuring the thickness of the infected paw with a dial caliper at weekly intervals and expressed by the arithmetic mean and standard error mean (SEM) of the net thickness increase (infected minus control contralateral paw thickness). Following ethical procedures, when the paw lesion reached 5 mm in thickness or presented ulceration, the mice were euthanized.

Tissue Processing for Optical Microscopy.
To analyze the local inflammatory reaction and presence or absence of parasites, footpads were removed postmortem on the 5th or 6th (IFN KO C57BL/6) or 11th (C57BL/6 WT, BALB/c) week after infection, excised, and prefixed with 10% formalin, followed by fixation in Bouin solution (picric acid 75%, glacial acetic acid 5%, and formaldehyde 10%) prior to paraffin embedding. Five m sections from the material were stained with hematoxylin and eosin (H&E) and examined under light microscopy.

Statistical
Analysis. Data are presented as mean ± standard error of the mean (SEM). Two way ANOVA/Bonferroni was used to compare the data, and the differences were considered significant when < 0.05.

Patient Profiles.
The age of patients varied from 19 to 46 years and they had one to three cutaneous lesions located in the limbs that appeared from two to eight months before the diagnosis. All patients were diagnosed with ATL, presenting the cutaneous localized clinical form (LCL) according to clinical and laboratory analyses. In all patients, the lesions were ulcerated. Patients' data are presented in Table 1.

Molecular Characterization of Leishmania
Isolates. SSU rDNA amplification was performed on four clinical field isolates and controls, using primers S12/S4 and the PCR products were analyzed by automatic sequencing. The nucleotide sequence of the four isolates identified HPV6, UAF5, WSS5, and PLR6 as species of the Viannia subgenus (data not shown). All samples were also analyzed by PCR of the G6PD gene with primers specific for L. (V.) braziliensis or "nonbraziliensis" Viannia species. This analysis confirmed the identity of HPV6, UAF5, and WSS5 as L. (V.) braziliensis and of PLR6 as a "non-braziliensis" isolate (data not shown). The identification of the PLR6 isolate was based on the analysis of ribosomal ITSs 1 and 2. Approximately 1 kb fragment was amplified and digested with Hae III. The analysis of restriction fragment polymorphisms indicated that PLR6 displayed a pattern compatible with L. (V.) guyanensis ( Figure 1). This was confirmed by nucleotide sequencing of the 1 kb fragment encompassing ITS1, 5.8S rDNA, and ITS2. The sequence obtained (Genbank number AJ000299.1) showed 99% identity with L. (V.) guyanensis (MHOM/BR/75/M4147).

Behavior of the Isolates in In Vitro
Culture. Replication rates of the four isolates were similar in complete Grace's medium at 26 ∘ C during 13 days. The growth curves exhibited typical logarithmic and stationary phases. The parasites formed large clumps at the stationary phase (data not shown). The maximum number of parasites occurred within 4 to 6 days, ranging from around 5 × 10 7 to 1 × 10 8 /mL (Figure 2). After 10 days of culture, parasites of all isolates began to die.

Course of Infection in Mice.
In order to compare the outcome of infection caused by all isolates, stationary-growthphase promastigotes (the 6th day of culture) were inoculated into C57BL/6 WT and BALB/c mouse footpads. Infection was successfully established for all L. (V.) braziliensis isolates in C57BL/6 WT and BALB/c mice and the lesions increased to a size of approximately 1.0-1.5 mm (Figures 3(a), 3(b), and 3(c)). The infection with the L. (V.) guyanensis PLR6 isolate caused a lesion more severe in BALB/c mice than in C57BL/6 ( < 0.05), which completely controlled the infection by 11 weeks (Figure 3(d)).  (Figure 4(a)) and dissemination of parasites to the contralateral paw (increased thickness) was apparent on the 6th week after infection when the infected footpad began to ulcerate (data not shown). No parasites could be seen at higher magnification (1000x, data not shown).
In comparison, BALB/c mice infected with L. (V.) braziliensis also presented intact epidermis and superficial dermis but, in the deep dermis, a mononuclear inflammatory infiltrate rich in vacuolated macrophages was located close to and infiltrating the muscle bundles ( Figure 5(c)); most macrophages were parasitized with L. (V.) braziliensis WSS5 ( Figure 5(e)). On the other hand, in mice inoculated with L. (V.) guyanensis PLR6, hypertrophic scar and accentuated fibrosis were seen in the dermis and a mononuclear inflammatory infiltrate ( Figure 5(d)) with vacuolated macrophages free of intact parasites was observed in the deep dermis ( Figure 5(f)).
A marked difference in the histology was seen in IFN KO C57BL/6 on the 6th week after inoculation with the same isolates ( Figure 6). These mice, inoculated with the L. (V.) braziliensis WSS5 isolate, did not have epidermal ulceration of the paw and from the plantar to dorsal side of

Discussion
This report characterizes the Leishmania (Viannia) species isolated from skin biopsies of four patients assisted at the Tropical Diseases Hospital (Anuar Auad, Goiânia, Goiás), with a diagnosis of localized cutaneous leishmaniasis. In our previous studies, three of these isolates were identified as L.
(V.) braziliensis (data not shown and [29]) and one remained unidentified. Here, characterization of the ribosomal ITS allowed identification of the latter isolate as L. (V.) guyanensis.
It is important to stress here the difficulties in correctly identifying this isolate as L. (V.) guyanensis. To achieve this characterization, we have used three strategies: small subunit ribosomal RNA (SSU rDNA) was sequenced [24]; sets of primers for G6PD were used in PCR assays to discriminate L. (V.) braziliensis from other organisms of the Viannia subgenus [25]; and ITS was amplified and cloned and the nucleotide sequence of three independent positive clones was phylogenetically analyzed [28]. It is crucial to identify the Leishmania species in order to define which parasites are circulating in a geographic area, to establish the transmission cycles of ATL and to implement the best possible treatment. These points, especially the last one, together with our results, indicate the need of more suitable molecular techniques to define the Leishmania species in the diagnosis of ATL. in the Northern region of Brazil, and there are no reports about autochthonous cases of ATL caused by this species in Goiás state or another Brazilian regions [7,8]. Thus, our findings confirm the fact that there is a high probability of L. (V.) guyanensis being introduced in Goiás state due to the migratory behavior of patients infected with these parasites from the Northern to Central Western region. This possibility is reinforced by the fact that the Tropical Disease Hospital/Anuar Auad assists several patients from the states of the Northern region (around 25% of the assisted patients, personal communication), pointing out the need of parasite species identification in ATL patients. The distribution of Leishmania spp. is dependent on vectors and reservoir hosts present in a geographic area. Thus, in Goiás, there are 47 different species of phlebotomine sand flies [14], with a predominance of L. intermedia and L. whitmani [8,13,14]  ) guyanensis has been reported, but the response to antimonial treatment can be different [15,17]. The isolate L. (V.) guyanensis PLR6 has not been tested for antimonial susceptibility, but the other isolates L. (V.) braziliensis HPV6, UAF5, and WSS5 were uniformly susceptible in vitro to meglumine antimoniate and amphotericin B [22].
Corroborating the difficulties in identifying the Leishmania species relying on the clinical findings, our results did not show any difference in the monophasic-culture replication rates among the different isolates. In these cultures, Leishmania species were morphologically similar, and the in vitro growth profiles were similar to those previously described for L. (V.) braziliensis [34].
Our group has previously confirmed that the isolate L. (V.) braziliensis HPV6 is able to infect C57BL/6 mice and the J774 murine macrophage cell line [33]. In the present study, we confirmed the infection capacity of this isolate in C57BL/6 and BALB/c mice. All four isolates infected C57BL/6 and BALB/c mice. In contrast to C57BL/6 WT mice infected with L. (V.) braziliensis isolates, those infected with L. (V.) guyanensis PLR6 showed fast regression of the lesion, which almost disappeared after 11 weeks. BALB/c mice also showed nonulcerative skin swelling when infected with all isolates. However, in the deep dermis of L. (V.) guyanensis PLR6infected mice, no parasites were detected inside macrophages whereas in L. (V.) braziliensis WSS5-infected footpads we observed a large number of parasites. The size of lesions caused by (L. (V.) braziliensis) HPV6, UAF5, and WSS5 was similar to those described by Pereira et al. [35] but larger than the size found in murine models of infection with this species [36][37][38]. de Moura et al. [39] reported that inoculation of L. (V.) braziliensis into the ear dermis of BALB/c mice leads to the development of an ulcerated lesion. The discrepancies with our results could be related to the site of inoculation or the virulence of the parasite strain. Considering the size and time course of the infection caused by L. (V.) guyanensis PLR6, our results were similar to those obtained by Sousa-Franco et al. [40], and like these authors we did not find parasites inside macrophages after 11 weeks of infection.
In this study, high susceptibility of IFN KO C57BL/6 mice to all four isolates confirms a close association between resistance and production of Th1 cytokines (IFN ) during the course of L. (Viannia) spp. infection as has been described by de Souza-Neto et al. [41]  spp. led us to use this mouse strain for the process of parasite isolation from lesions of ATL patients [21,43].
ATL has been considered as a social and economic problem of the poor population which has resulted largely from an intense migratory movement into rural areas and the forested hillsides that are close to the outskirts of the urban centers. However, due to the intense international travel and the large contingents of displaced and migratory populations, Tegumentary Leishmaniasis has to be considered as a diagnosis of nonhealing indolent ulcers also in nonendemic areas. Moreover, infection by L. (V.) guyanensis has been diagnosed in Europe in a soldier who denied travelling to endemic areas or having blood transfusions, raising some uncommon possibilities of contagion [44].
Nowadays, migration among Brazilian regions has largely increased, both inside forested areas and in urban areas.

Conclusions
We have isolated and characterized three clinical field isolates of Leishmania spp., from patients probably infected in Goiás, Tocantins, and Pará states, Brazil, as L. (Viannia) braziliensis and one as L. (V.) guyanensis. The latter species had not yet been described in Goiás. Infection of mouse strains BALB/c, C57BL/6 wild-type, and C57BL/6 lacking gammainterferon (IFN KO C57BL/6) showed differences in lesion development among the Leishmania strains. In addition, better infection control of L. (V.) guyanensis than L. (V.) braziliensis was observed in mice in the presence of IFN but not in the absence of this cytokine. Molecular identification of L. (V.) guyanensis in a patient resident in Goiás stresses the importance of correct species identification and suggests that the presence of this species is possibly underestimated in several areas of Brazil.