A Luciferase-Expressing Leishmania braziliensis Line That Leads to Sustained Skin Lesions in BALB/c Mice and Allows Monitoring of Miltefosine Treatment Outcome

Background Leishmania braziliensis is the most prevalent species isolated from patients displaying cutaneous and muco-cutaneous leishmaniasis in South America. However, there are difficulties for studying L. braziliensis pathogenesis or response to chemotherapy in vivo due to the natural resistance of most mouse strains to infection with these parasites. The aim of this work was to develop an experimental set up that could be used to assess drug efficacy against L. braziliensis. The model was tested using miltefosine. Methodology/Principal Findings A L. braziliensis line, originally isolated from a cutaneous leishmaniasis patient, was passaged repeatedly in laboratory rodents and further genetically manipulated to express luciferase. Once collected from a culture of parasites freshly transformed from amastigotes, 106 wild type or luciferase-expressing stationary phase promastigotes were inoculated subcutaneously in young BALB/c mice or golden hamsters. In both groups, sustained cutaneous lesions developed at the site of inoculation, no spontaneous self- healing being observed 4 months post-inoculation, if left untreated. Compared to the wild type line features, no difference was noted for the luciferase-transgenic line. Infected animals were treated with 5 or 15 mg/kg/day miltefosine orally for 15 days. At the end of treatment, lesions had regressed and parasites were not detected. However, relapses were observed in animals treated with both doses of miltefosine. Conclusions/Significance Here we described experimental settings for a late-healing model of cutaneous leishmaniasis upon inoculation of a luciferase-expressing L. braziliensis line that can be applied to drug development projects. These settings allowed the monitoring of the transient efficacy of a short-term miltefosine administration.


Introduction
Leishmania spp. are the etiological agents of leishmaniasis, a complex of vector-borne infectious diseases transmitted by sand flies. The disease is widespread in the world in tropical and subtropical regions. Approximately 12 million people are affected worldwide and 1.2 million new cases occur each year [1]. Leishmaniasis is responsible for a spectrum of clinical manifestations including visceral, cutaneous, mucosal, disseminated and diffuse cutaneous disease that are determined, at least partially, by the species of the parasite [2]. In South America, Leishmania (Viannia) braziliensis is the most prevalent species isolated from patients displaying either cutaneous or mucocutaneous leishmaniasis [3][4][5]. This species is also associated with disseminated leishmaniasis [5].
In spite of its epidemiological importance, the study of L. braziliensis infections is limited by the difficulty of establishing good animal models of disease. L. braziliensis infects hamsters and induces a chronic cutaneous disease with no spontaneous healing [6,7]. However, inbred strains of mice are generally resistant to L. braziliensis or develop a transient and self-healing cutaneous lesion [7][8][9][10]. When the parasites are inoculated into partially susceptible strains such as BALB/c mice, a non-ulcerative lesion develops initially and evolves to spontaneous healing in about 30 days [8,10]. A mouse model of sustained skin lesions with L. braziliensis would be very useful for studies on the pathogenesis of the disease as well as a tool to develop new control strategies, such as vaccines and drugs.
The treatment of leishmaniasis relies on a few drugs, most of them displaying some inadequacy to present requirements. They are expensive, toxic and require prolonged and parenteral administration. Additionally, drug resistance has been reported [11,12]. More than a decade ago, miltefosine (MF) was described as an oral drug against visceral leishmaniasis (VL) and has been widely used in the Indian subcontinent for that application [13][14][15]. On the other hand, MF efficacy for the treatment of cutaneous leishmaniasis (CL) seems to be dependent on the Leishmania species and even on intraspecies heterogeneity. Some clinical studies in South America identified cure rates varying from 53% to 91% in CL patients [16][17][18]. Two clinical studies performed in Brazil detected cure rates of 71.4% and 75%, respectively, in infections due to L. guyanensis [19] and L. braziliensis [20]. However, given the lack of other new treatment alternatives, further understanding of MF's potentialities and limitations is necessary.
We have recently described the use of luciferase as a reporter for drug efficacy evaluation in cutaneous and visceral leishmaniasis models [21,22]. In both models, luciferase has proven to be an accurate measure of parasite burden.
In this study, we generated a transgenic line of L. braziliensis expressing luciferase that was adapted to induce chronic lesions in BALB/c mice. Using this experimental model, we evaluated the efficacy of MF to reduce lesion size and parasite load.

Methods Drugs
Amphotericin B sodium deoxycholate and MF were obtained from Sigma-Aldrich (St. Louis, MO, USA). Stock solutions of amphotericin B and MF for in vitro experiments were prepared in sterile water (10 mM). Miltefosine used for the treatment of infected mice was prepared daily in sterile saline (0.9% NaCl) from a 6 mg/mL stock solution.

Parasites
The L. braziliensis strain MHOM/BR/94/H3227 was kindly provided by Dr. Maria Jania Teixeira, from Universidade Federal do Ceará. These parasites were isolated from a cutaneous ulcer from a patient in Ceará State, Brazil and the strain was previously typed as L. braziliensis [10]. Promastigotes were grown in medium M199 (Sigma-Aldrich) supplemented with 10% heatinactivated foetal calf serum, 2% urine, 0.25% hemin, 12 mM NaHCO 3 , 50 U/mL penicillin and 50 μg/mL streptomycin at 25°C. For in vivo experiments, female BALB/c mice (4-5 weekold) were inoculated with 10 6 stationary-phase promastigotes injected subcutaneously in the right hind footpad. Amastigotes were obtained from infected mice, as described [23] and differentiated back as promastigotes in M199. Promastigotes were counted in a Neubauer haemocytometer.
Generation of a transgenic line of L. braziliensis expressing luciferase A construct containing the luc2 gene was previously built [22] taking advantage of the vectors pSSU-int and pSPαHYGα which were kindly provided by Dr. Tony Aebischer (Robert Koch Institute, Berlin, Germany) and Dr. Marc Ouellette (Universite Laval, Quebec, Canada). The linear cassette purified upon Pac I and Pme I digestion of the construct described previously, was used to transfect L. braziliensis H3227 promastigotes. Briefly, the cassette contains the luc2 gene followed by a Leishmania 3' UTR, the hygromycin phosphotransferase gene and fragments of the L. mexicana small subunit (SSU) ribosomal DNA (rDNA) at the cassette extremities to promote homologous recombination. Transfection was performed as described [24] using 5 μg of linear DNA. Twenty-four hours after transfection, 32 μg/mL hygromycin B was added for selection of mutants. The L. braziliensis transfectant line was kept for three passages in M199 containing hygromycin B and then plated on semi-solid M199 medium/1% agar supplemented with 1.2 μg/mL biopterin, 2% urine and 32 μg/mL hygromycin B for cloning [24]. Four independent clones were analysed for integration of the cassette into the rDNA locus by PCR amplification, using primers complementary to sequences inside and outside of the linear cassette. Genomic DNA of these clones was purified as described [25] for confirmation by PCR. Primers used were S1 (5'-GATCTGGTTGATTCTGCCAG-3') and S4 (5'-GATCCAGCTGCAGGTTCACC-3') [26] that anneal to the SSU rDNA sequence flanking the insertion sites and primers luc2F (5'-GCGGGATCCATGGAAGATGCCAAAAACAT TAAG-3'), luc2R (5'-CACGCGCATACATTCACGGCGTTACACGGCGATCTTGCCGC-3') and luc2i (5'-GACCGACTACCAGGGCTTCC-3') that anneal to the luc2 gene contained in the linear cassette ( Fig 1A) [22].

Parasite susceptibility to drugs in vitro
Susceptibility of L. braziliensis wild type and transgenic lines to amphotericin B and MF was evaluated by a [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide] (MTT, Sigma-Aldrich) viability test assay as described [27]. For this assay, promastigotes (2×10 6 parasites per well) were incubated in the presence of increasing concentrations of amphotericin B (0.037 to 0.3 μM) or MF (25 to 200 μM) for 24 h. MTT cleavage was measured in a microplate reader (POLARstar Omega, BMG Labtech, Ortenberg, Germany) with a test wavelength of 595 nm and a reference wavelength of 690 nm.
For intracellular amastigotes, bone marrow-derived macrophages (BMDM) from BALB/c mice were used as previously described [21]. BMDM were plated on round glass coverslips in 24-well culture dishes, at a density of 3 × 10 5 cells in 400 μL of RPMI 1640 medium (Gibco, Invitrogen Corporation) supplemented with 10% FCS (Gibco, Invitrogen Corporation) in a 5% CO 2 atmosphere for 16 h at 37°C allowing macrophages to adhere. Macrophages were then infected with stationary-phase promastigotes using ratios varying from 20 to 35 parasites per macrophage for 3 h at 33°C. Non-internalized parasites were removed by washing with warmed PBS, followed by the addition of fresh medium containing increasing concentrations of amphotericin B (0.025 to 0.2 μM) or MF (0.5 to 24 μM). After 48 h, the cells were fixed in methanol and stained using the panoptic haematological labelling method (Instant Prov kit, Newprov, Pinhais, PR, Brazil). The percentage of infected macrophages was determined by counting 100 cells in three independent experiments. The half-maximal effective concentrations (EC 50 ) were determined from sigmoidal regression of the concentration-response curves. Assays were performed in triplicate and results are expressed as the mean and standard deviation (SD) of at least three independent experiments.

Luciferase assay in vitro
The activity of luciferase was measured in recombinant L. braziliensis promastigotes at logarithmic and stationary growth phases. Parasites were washed twice with PBS and then serially diluted for the luciferase assay, which was performed using the One Glo Luciferase Assay System (Promega Corporation) according to the manufacturer's instructions and as described previously [22]. Bioluminescence was measured in a microplate reader (POLARstar Omega, BMG Labtech, Ortenberg, Germany). Each point was tested in triplicate in two independent experiments.

Animal infections
Male golden hamsters (3 to 5 weeks-old) were obtained from the Instituto de Medicina Tropical, University of São Paulo. Female BALB/c mice (3 to 5 weeks-old) were from the Instituto de Ciências Biomédicas, University of São Paulo. Animals were kept in cages and received unlimited food and water.
Given that the wild type (Lb-WT) and luciferase transgenic parasites (Lb-LUC) were in different passages as promastigote cultures, they were initially used to inoculate highly susceptible hamsters. Initially, 10 6 stationary phase promastigotes were injected into hamsters' hind footpads in a volume of 30 μL. Once the infections were established, parasites of both lines were recovered from the inoculated footpads and differentiated back into promastigotes in M199 at 25°C. These promastigotes were then used to inoculate BALB/c mice using the same protocol described above, except for the host. When the lesions developed, parasites were recovered, transformed into promastigotes and used to inoculate mice and hamsters with the results described in the Results section. Disease progression was followed for 11 weeks in hamsters and for 24 weeks in BALB/c mice.

Treatment of mice with MF and evaluation of treatment efficacy
BALB/c mice inoculated with Lb-WT or Lb-LUC were submitted to treatment with MF. Treatment was initiated 4 weeks post-inoculation, using two different doses of the drug: 5 or 15 mg/ kg/day, in groups of 5 animals. MF was prepared daily from a stock solution of 6 mg/mL and was administered by oral gavage for 15 consecutive days. An untreated group was used as a control for each line studied. Lesion sizes were evaluated once a week by measuring the difference in thickness between the infected and the contralateral uninfected footpad using a caliper (Mitutoyo Corporation, Kawasaki, Kanagawa, Japan). Parasite burden in Lb-LUC inoculated BALB/c mice was determined 6 weeks (at the end of MF treatment), 17 and 23 weeks post-inoculation. The parasite load was quantified by measuring luciferase activity through bioimaging (IVIS Spectrum, Caliper Life Sciences, Inc. MA/ USA) as described [21]. Briefly, before luminescence detection animals received 75 mg/kg luciferin (VivoGlo Luciferin, Promega) intraperitoneally followed by anesthesia in a 2% isoflurane atmosphere (Cristália). Images were collected after 20 minutes using a high-resolution mode with 2 minutes exposure from a fixed-size region of interest. Results were quantified with Living Image software version 4.3.1 (Caliper Life Sciences) and were expressed as photons/second/square centimeter/steradian (ph/sec/cm 2 /sr).

Susceptibility to MF after in vivo treatment
Lb-WT parasites were recovered from infected mice that had been treated with 5 mg/kg/day (n = 5) or with 15 mg/kg/day (n = 5) MF at the 19 th and 23 th week post-inoculation, respectively. These parasites were differentiated into promastigotes in M199 medium and their susceptibility to MF was determined by the MTT assay as described above. As controls, parasites were also rescued from an untreated infected animal at the 19 th week post-infection.

Statistical analysis
Unpaired two-tailed Student's t tests were used to compare the EC 50 determined in vitro on promastigotes and intracellular amastigotes of wild type parasites and luciferase expressing lines and for comparing disease progression. Parasite burden was analysed for statistical significance by One Way ANOVA, followed by the Tukey post-test. All statistical analyses were performed using GraphPad Prism 6.0 software (GraphPad Software, Inc., La Jolla, CA USA). Results were considered significant at P < 0.05.

Generation and characterisation of a luciferase expressing line of L. braziliensis
To generate a luciferase expressing L. braziliensis line (Lb-LUC), promastigotes of the H3227 strain were transfected with a linear cassette containing the luc2 gene flanked by SSU rDNA sequences (Fig 1A). After transfection, parasites were selected in the presence of hygromycin B and cloned. Four independent clones were tested for the presence of the luc2 gene by PCR. Amplification of a 1.6 kb fragment with oligonucleotides complementary to the 5' and 3' ends of the luc2 open reading frame was observed in the four selected clones while the same reaction led to the absence of amplification from wild type parasites (Fig 1B). Additional pairs of primers were used to confirm the correct integration into the ribosomal DNA locus in one of the selected clones (clone 3, lane 3 in Fig 1B). Primers S1 and S4 [26], complementary to the rDNA outside the linear cassette, were used in combination with primers luc2R and luc2i, respectively (Fig 1A). These two combinations of primers amplified the expected fragments of 2.4 and 6.7 kb in the transgenic line but not in wild type parasites (Fig 1A and 1C).
Bioluminescence in the transgenic line (Lb-LUC) was confirmed after incubation of parasites with luciferin. A linear correlation between the number of promastigotes and the production of light was observed (Fig 1D).
Growth curves for transgenic Lb-LUC and wild type (Lb-WT) promastigotes were indistinguishable (S1A Fig). Logarithmic and stationary phase promastigotes presented similar light emission (S1B Fig) indicating that luciferase expression was stable along the growth curve. The stability of the transgene was also evaluated by removing the drug pressure from promastigote cultures. Luciferase expression was stable with the same levels of light production in parasites kept in the absence of hygromycin B for 25 passages in culture (S1C Fig).
No differences were observed in the infectivity of Lb-LUC and Lb-WT to macrophages after 24 or 48 hours evaluated by the morphology of infected cells, percentage of infected cells or number of amastigotes per infected macrophage (S2A-S2D Fig). In addition, a direct correlation was observed between the number of Lb-LUC intracellular amastigotes per macrophage and bioluminescence (S2E Fig). Finally, in vitro susceptibility of Lb-LUC and Lb-WT promastigotes and intracellular amastigotes to amphotericin B and MF showed no significant differences (Table 1).

Disease progression in hamsters and BALB/c mice inoculated with Lb-LUC
Hamsters and BALB/c mice were inoculated with 10 6 Lb-WT or Lb-LUC promastigotes subcutaneously in the footpad. In both animal models, there were no significant differences in the kinetics of lesion development or in lesion size when the two parasite lines were compared (Fig  2A and 2B). In both models, lesions were non-ulcerative (Fig 2C and 2D). In hamsters, lesions were detectable from two weeks post-inoculation, progressed steadily for the next five weeks and then the oedema stabilized till the 11 th week post-inoculation (Fig 2A), when animals were euthanized. In BALB/c mice, lesions were clearly detected after two weeks, progressed in size until weeks 9-10 and then decreased in the following weeks (Fig 2B and S3 Fig), but the local oedema remained detectable until 23 weeks post-inoculation (Fig 2B and S3 Fig). Therefore, no spontaneous healing was observed in the animal models studied during the period evaluated.
BALB/c mice inoculated with Lb-LUC were examined by in vivo imaging after 6, 17 and 23 weeks (Figs 3 and 4). Luminescence was detected from the lesion site in all mice at the 6 th week post-inoculation (Figs 3B and 4). By week 17 th , light emission was not detected in 2 out of 5 mice (Figs 3C and 4). No bioluminescence was detected in untreated animals at the 23 th week post-inoculation (Figs 3D and 4), although measurable oedema was still observed in this group at the 24 th week (Fig 3A).

In vivo efficacy of MF in Lb-LUC inoculated BALB/c mice
The late-healing infection model of L. braziliensis H3227 in BALB/c mice was used to test MF efficacy. Treatment was initiated 4 weeks post-inoculation in animals injected with the transgenic line Lb-LUC (Fig 3) or with the wild type parasites (S3 Fig). Animals were treated with MF 5 or 15 mg/kg/day for 15 consecutive days. Both doses of MF resulted in regression of lesions at the end of the treatment (6 weeks post-inoculation) (Fig 3A and S4 Fig). Luminescence was not detected by bioimaging at the end of the treatment (Figs 3B and 4).
However, 8 weeks post-inoculation a relapse was observed in animals treated with 5 mg/kg/ day MF, with lesions progressing in size for the following weeks (Fig 3A and S3 Fig). These findings were confirmed by quantification of bioluminescence 120 days post-inoculation (17 th week), time when animals treated with 5 mg/kg/day presented a higher parasite burden than untreated animals (Figs 3C and 4).   A clinically detectable relapse was also observed in animals treated with 15 mg/kg/day from the 15 th week post-inoculation that progressed in the following weeks (Fig 3A). Although no luminescence was detected in these animals at week 17 th (120 days) post-inoculation (Figs 3C  and 4), parasites were clearly detected by in vivo imaging at the 23 th week post-inoculation (Figs 3D and 4).
Animals inoculated with Lb-WT and treated with both doses of MF presented the same pattern of disease progression, response and relapse to treatment observed with the Lb-LUC transgenic line (Fig 3A and S3 Fig).
To investigate whether these relapses were due to MF resistance acquired during the therapy regimen, parasites were recovered from 5 animals treated with MF 5 mg/kg/day (at the 19 th week post-inoculation) and with 15 mg/kg/day (at the 23 th week post-inoculation) (S3 Fig). After differentiation to promastigotes, MF susceptibility was determined in parasites isolated from untreated and treated animals and no significant changes in drug susceptibility were found between these parasites ( Table 2).

Discussion
To the best of our knowledge, this is the first report of a luciferase-expressing L. braziliensis, the causative species of cutaneous and mucocutaneous leishmaniasis in humans and the most prevalent species in South America [2]. The importance of research on L. braziliensis infections cannot be overstated. However, it faces multiple difficulties largely derived from the distinct (6 th week post-inoculation) (B), at the 17 th (C) and at the 23 th week post-inoculation (D) (40, 120 and 160 days post-inoculation respectively). Statistical analysis was performed with One Way ANOVA, followed by the Tukey post-test. * p < 0.05; **, p < 0.001. Ph/sec/cm 2 /sr, photons per second per square centimetre per steradian.
doi:10.1371/journal.pntd.0004660.g003 behaviour of this Leishmania species, in vitro and in vivo. L. braziliensis promastigotes do not grow easily in axenic cultures and amastigotes are scarce in the lesions. The lack of an appropriate animal model mimicking the human infection in all of its potential clinical presentations is perhaps one of the main drawbacks. Most L. braziliensis strains are infective to mice but lead to asymptomatic or short-lived, self-healing disease [28,29]. The infection is more sustained in hamsters [6,29], but do not progress to a mucosal or disseminated disease.
The development of lesions at the site of inoculation with the H3227 strain of L. braziliensis in BALB/c mice was previously described [10]. However, in the original description, lesions were self-limited and healed 30 days post-inoculation. This strain was kindly given to us by Dr. Maria Jania Teixeira (Universidade Federal do Ceará, Brazil) and, before being submitted to transfection, was passaged in mice repeatedly. Most likely as a result of these repeated in vivo passages, the line obtained after transfection, together with the parental wild type parasites were then able to induce a late-healing disease in BALB/c mice, as well as being pathogenic in hamsters. A late-healing animal model of L. braziliensis infection is likely to become very useful as a tool in drug development or in the evaluation of the disease pathogenesis.
It is unclear at present whether other strains of L. braziliensis could be adapted to cause latehealing disease in BALB/c mice. We have attempted to do so with three other isolates as well as with the type strain of L. braziliensis M2903 without success but cannot rule out that possibility. It will be interesting to investigate whether mutations on this Leishmania isolate (H3227) are present and could be correlated with the altered phenotype.
These parasites were transformed to express the luciferase gene integrated into the ribosomal locus. Luciferase expression was stable throughout the life cycle. As previously reported for other luciferase-expressing Leishmania species [21,22,30], this L. braziliensis transgenic line presented the same biological properties in vitro and in vivo of the parental line and was useful for parasite quantification of promastigotes and amastigotes, in vitro and in vivo. Luciferase as a reporter has become instrumental in pre-clinical drug development projects, especially in chronic infections. Our previous results with L. amazonensis [21] and L. chagasi [22]  lines expressing luciferase clearly demonstrated that luciferase detection correlates well with parasite burden in vivo. In the case of H3227 parasites, the sensitivity of detection in vivo has proven to be insufficient to detect low parasite burdens, as was the case in Lb-LUC in untreated animals or when relapses were noticed in the groups treated with the higher dose of MF. This limitation in sensitivity should be taken into account in drug efficacy studies. Nevertheless, bioimaging was used successfully to document infection and as a semi-quantitative method for comparing parasite burden between different groups. The model of infection with Lb-LUC in BALB/c mice was put to test using MF, the newest addition to leishmaniasis chemotherapy. The susceptibility of the H3227 strain to MF in vitro can be considered comparable to other L. braziliensis strains [31,32]. Accordingly, at the end of MF treatment, Lb-LUC inoculated mice showed regression of lesions and absence of luminescence. However, although the doses used here were higher than the WHO recommended scheme for CL and VL (2.5 mg/kg/day) [15], a clinical relapse was evident in animals treated with 5 mg/kg/day as soon as two weeks after the end of treatment. In these animals, lesions progressed in the following weeks reaching sizes similar to the ones observed in untreated animals. In animals treated with 15 mg/kg/day, oedema in the footpad reappeared 10 weeks after the interruption of treatment.
Relapses in VL patients treated with MF do not seem to be due to acquired resistance [33]. Likewise, our data showed similar MF susceptibility in parasites rescued from untreated and treated animals, indicating that L. braziliensis did not acquire resistance to MF during treatment.
In VL patients, MF failure may be correlated with low drug exposure due to pharmacokinetic-pharmacodynamics factors [34]. MF has been shown to reach high levels in the skin [15] but poor drug distribution to cutaneous lesions may be another reason for the lower efficacy of the drug against CL when compared with VL. Furthermore, the greater genetic diversity of L. braziliensis compared to L. donovani [35,36] may increase the number of elements that contribute to treatment failure. Moreover, other factors, unrelated to drug resistance, should be considered as quiescence and even the presence of the LRV1 virus [37,38].
The persistence of viable parasites after treatment with 5 and 15 mg/kg/day MF per 15 days indicated that, even for the highest dose used, clinical response did not reflect sterile cure. In fact, sterile cure does not seem to be required for long-term control of the disease, as has been shown by the presence of parasites in healthy skin in animals [39,40]. On the other hand, relapses have been reported in patients treated with MF for all clinical forms of the disease [33,[41][42][43][44][45][46]. Therefore, we believe the data presented herein stress the need for long-term follow up of MF treated patients and suggest that higher doses and/or longer treatment regimens should be evaluated in order to avoid relapses. In addition, the combined use of MF with other antileishmanial drugs should be considered as an alternative for the chemotherapy of the disease.