|Home | About | Journals | Submit | Contact Us | Français|
A murine model of cutaneous leishmaniasis with green fluorescent protein positive (GFP+) L. major enables the monitoring of parasitic load via measurements of GFP fluorescence intensity, allowing for a faster and more efficient way of monitoring the clinical outcome of photodynamic therapy (PDT). This model may provide new insights on the phototoxic aspects in PDT. Although PDT regimens may be somewhat different in humans, it is expected that the developed model will facilitate the optimization and clinical translation of PDT as a therapy for cutaneous leishmaniasis and the eventual development of topical PDT treatments for other granulomatous infections.
Green fluorescence from GFP+ L. major in cutaneous leishmaniasis lesions on the ear of BALB/c mouse.
Cutaneous leishmaniasis (CL) is an infectious parasitic disease widely spread in the developing world, affecting people in 88 countries with 1.5 million new cases reported each year [1–3]. In recent years, CL has caused growing concern since 1 out of 100 American soldiers have contracted this disease during current US operations in the Middle East . Ulcerative lesions on exposed skin caused by L. major may be present for decades, leading to blood vessel and nerve damage, rough scar formation, and secondary infection [5, 6]. Although numerous treatment modalities are available, there is no ideal therapy for CL . Current treatment options for leishmaniasis include physical therapy and chemotherapy , but results observed are less than satisfactory. Therapeutic approaches that involve systemic treatments with antiparasitic drugs have a variety of side effects including an overall toxic effect  on the organism and infection recurrence, which is partly due to treatment resistance . In addition, use of physical therapies such as thermotherapy, surgery, or laser removal often causes disfiguring scars that are aesthetically undesirable . Given the disappointing results of current treatment methods, the development of a more effective and conveniently administered treatment that is easily tolerated will be of significant benefit.
Research in this area has shown that photodynamic therapy (PDT) offers an effective alternative treatment for CL [10–13]. Studies have revealed that, unlike systemic treatments with antiparasitic drugs, PDT can be applied to infected tissue in a localized manner [12, 14], enabling treatment of CL lesions without damaging healthy tissues . Furthermore, PDT shows no sign of systemic toxicity [13, 15], eliminating one of the biggest health issues related to existing CL treatments. A clinical study of photodynamic action on cutaneous parasitic infections conducted by Enk et al.  demonstrated that PDT offers relatively fast localized healing of infected lesions. Additional studies have found that PDT flattens, reduces the size, and selectively eliminates inflammatory infiltrate of CL lesions, leading to rapid healing without scarring [16, 17]. The excellent cosmetic outcome of PDT makes this alternative an ideal treatment modality in modern dermatology .
Despite the positive results obtained by PDT to treat CL, current PDT treatments can only decrease, but not eradicate, parasitic load in CL lesions . Many studies in this area have been done using 3,7-bis(di-n-butylamino) Phenothiazin-5-ium Bromide (PPA904), a photosensitizer (PS) that has demonstrated great therapeutic efficacy against L. major [10, 17, 19]. A study by our group  led to the conclusion that in order to achieve optimal parasite eradication and improve the therapeutic outcome of the treatment several subsequent PDT administrations are necessary. These findings led to the search for an effective and simple way of monitoring the outcome of PDT. Efficient monitoring of parasitic load in the CL lesions after PDT will enable us to determine when subsequent treatment administrations would be more effective in achieving complete parasite eradication.
In the search for an effective way to monitor the outcome of PDT we developed a murine model of CL infected with GFP+ L. major. It has been demonstrated that transfected Leishmania promastigotes expressing modified GC-rich S65T mutant GFP (termed GFP+) display a bright green fluorescence distributed throughout the cell . The goal of this study is to determine if measuring fluorescence intensity of GFP+ L. major present in CL lesions can enable us to monitor parasitic load before and after PDT application. Compared to current parasite quantification methods, which involve mice sacrifice, tissue homogenization, serial dilutions, and 7 days of incubation to be able to see parasite growth [10, 17, 21], the use of fluorescence intensity measurements for parasitic load determination represents a faster, easier, and more efficient way of monitoring the outcome of PDT. Furthermore, because fluorescence intensity measurements can be done in vivo in a non-invasive manner, mice sacrifice is not necessary, thus minimizing the number of animals needed for the study. Establishing GFP+ L. major murine models as a feasible way of monitoring the efficacy of this treatment will represent a significant benefit and a big step towards the optimization of PDT as a treatment modality for CL.
PPA904 (3,7-Bis(N,N-dibutylamino) phenothiazinium bromide (MW = 532.7) was obtained from Photo-pharmica (Leeds, UK). Cream was used for photodynamic therapy. PPA904 is formulated at a low concentration (500µM/0.266mg/ml) in Unguentum M cream (contains: purified water, white soft paraffin, cetostearyl alcohol, polysorbate 40, propylene glycol, glyceryl monostearate 40–50, liquid paraffin, medium-chain an-hydrous silica, sodium hydroxide).
The NIH Friedlin V1 strain (MHOM/IL/80/FN) was used in this study. Leishmania parasites were cultured at 24°C without CO2 in medium 199 (M199) supplemented with 20% heat-inactivated fetal calf serum, 60 ng/mL penicillin G sodium salt, 100 ng/mL kanamycin sulfate, 50 ng/mL flucytosine, 10 ng/mL chloramphenicol, 2 mM L-glutamine, 40 mM HEPES, 0.1 mM adenine (in 50 mM HEPES), 5 mg/mL hemin (in 50% triethanolamine) and 1 mg/mL 6-biotin (medium 199 complete; M199-C).
Infective-stage metacyclic promastigotes were isolated from stationary cultures (4–5 days old) by using a uniform procedure based on a modification of a method of density gradient purification . Briefly, 10 ml of parasite suspension in M199-C containing approximately 1×1010 stationary-phase promastigotes were layered on a discontinuous density gradient in a 50-ml conical Falcon tube consisting of 10 ml of 20% Ficoll stock solution made in distilled water and 10 ml of 10% Ficoll diluted in M199-C. The gradient was centrifuged for 15 min at 2,000 × g at room temperature, and the parasites in the upper 10% Ficoll were collected and washed by centrifugation at 3,000 × g.
Parasites expressing GFP+ were obtained by transfection of L. major with the large SwaI targeting fragment from pIR1SAT-GFP+(b) (Strain B3538), by electroporation and plating on semisolid media as described . Numerous clonal lines were obtained, and correct integration into the rRNA locus was confirmed by Southern blotting and strong GFP+ expression was confirmed by flow cytometry. Several were used to infected BALB/c mice and one showing normal virulence was used in these studies (SSU::IR1SAT-GFP+(b)).
Female BALB/c mice of 6 to 8-week-old were obtained from Charles River Laboratories (Wilmington, MA) for this study. All animal procedures were performed according to protocols approved by the Massachusetts General Hospital Subcommittee on Research Animal Care.
For in vivo infections, 1 × 106 metacyclic parasites (this dose in Leishmania research has been used to induce the rapid development of CL lesions) in 20 µl of PBS were inoculated intradermally into the ear of 6 to 8-week-old BALB/c female mice using a 27.5-gauge needle. The evolution of the lesion was monitored for 3 to 5 weeks.
Mice were infected with L. major 3 weeks prior to use. A PPA904 cream was applied topically (2mg/cm2) to lesions under dark conditions (topical application time of PPA904 cream was 60 minutes). The CL lesions were then irradiated for 11.6 minutes with a LC-122A non-coherent light source (LumaCare® Ltd, CA) equipped with a 665 ± 15 nm fiber optic bundle at a fluence of 21 J/cm2 (fluence rate of 30 mW/cm2). Mice were anaesthetized during the whole time of topical cream application and irradiation (2–2.5h) with an intramuscular injection of ketamine (90 mg/kg)-xylazine (10 mg/kg) cocktail. Mice with no procedure served as the infected non-treated (control) group.
Wavelength-resolved fluorescence hyperspectral images of the CL lesions were acquired with the Maestro imaging system (CRI). A 460- to 480-nm excitation filter and a 515-nm long-pass emission filter (were used to acquire fluorescence images of GFP. A 640 nm excitation filter and a 700-nm long-pass emission filter (were used to acquire fluorescence images of PPA904. Hyperspectral images between 515 and 530 nm at 5-nm intervals for GFP and between 700 and 750 nm at 5-nm interval for PPA904 were acquired on anesthetized animals. A cuvette containing 1×107 GFP+ L. major was imaged with the Maestro imaging system before each imaging session and subsequently acquired images were normalized by the fluorescence intensity of GFP. GFP fluorescence in tissue was quantified by spectrally unmixing GFP fluorescence spectrum with the tissue auto fluorescence spectrum using linear least-squares optimization.
The statistical analysis was based on the calculation of arithmetic mean and standard deviation. The difference between two means was compared by a two-tailed un-paired Student's t test assuming equal variances or Mann-Whitney test in unequal variances. One-way analysis of variance techniques were used for multiple group comparisons. Bonferroni’s multiple comparison test was chosen when certain pairs had to be analyzed. The relationship between two or more variables was measured by correlation analysis (Pearson coefficient). A p-value of less than 0.05 was considered statistically significant.
Leishmania major parasites expressing GFP+ integrated into the small subunit ribosomal RNA locus were obtained as described in the methods. Previous studies have shown that expression from this site yields stable high and uniform expression . In order to establish if parasitic load could be determined by measuring fluorescence intensity of GFP+ L. major parasites, we evaluated if fluorescence intensity correlates with number of parasites (Figure 1A).
Different numbers of metacyclic parasites were plated in 96-well plates and fluorescent intensity was measured spectrophotometrically. Linear regression analysis demonstrated a direct correlation between number of parasites and GFP fluorescence (r = 0.972) (p < 0.001).
Before progressing to animal models, we had to establish that fluorescence signals were due only to viable parasites. After several freeze-thaw cycles to ensure parasites were dead, fluorescence intensity of both live and dead Leishmania parasites was measured at 520nm. As was expected, live GFP+ L. major parasites showed a bright green florescence signal, while dead GFP+ L. major did not fluoresce.
Once we established that only viable parasites were capable of fluorescing, a murine model of CL infected with GFP+ L. major was developed. Several weeks after intradermal inoculation of GFP+ L. major into mouse ears as described in Materials and Methods section, a visible lesion very similar to those seen in non-GFP murine models of CL [10, 17] was observed (Figure 1B). In vivo fluorescence measurements of mouse ears infected with GFP+ Leishmania show a bright green fluorescent signal in the area of the CL lesion and no fluorescence in the surrounding healthy tissue. Our results demonstrate that we have obtained a murine model of CL where area of infection can be easily visualized by fluorescence.
In order to establish the use of CL murine models infected with GFP+ L. major as an efficient way of monitoring the outcome of PDT and determining when a subsequent treatment would be more effective, fluorescence intensity in CL lesions with GFP+ L. major was measured before, right after, and every day after PPA904-PDT treatment (Figure 2).
GFP fluorescence measurements before PPA904 application were used as 100% of parasitic load. Fluorescence measurements of GFP+ CL lesions show that parasitic load was significantly decreased right after PPA904-PDT (about 80% decreases). After PDT, parasitic load underwent a biphasic pattern. This pattern correlates with results observed in previous studies of PDT in CL murine models using PPA904  and EtNBSe , another phenothiazine compound. Subsequent treatments might be most effective when applied at day 4 after first PDT. The aforementioned studies, instead of using fluorescence measurements to determine parasitic load, used typical parasite quantification methods, which involve mice sacrifice, tissue homogenization, serial dilutions, and 7 days of incubation to be able to measure parasite growth. Thus, the fact that similar results were obtained with both parasite quantification methods, establishes that CL murine models infected with GFP+ L. major is a simpler, more efficient, and cost-effective way of monitoring the outcome of PDT.
Being able to monitor fluorescence decay of GFP, and thus changes in parasitic load, during the process of PDT enables us to study more deeply the relationship between the photosensitizer (PS) and the observed decrease in parasitic load. PPA904-PDT was performed as described in the Materials and Methods section and fluorescence decay of both GFP and the PS were measured during the process (Figure 3).
By observing GFP fluorescence decay we could establish that right after PS application there is a decrease in parasitic load. The fact that this was observed even before light irradiation had begun suggests that PPA904 has a dark toxic effect on the Leishmania parasites. This dark toxicity has also been reported by previous studies of PPA904-PDT on CL murine models  and has been considered a positive effect.
Once PDT has started, GFP fluorescence decays during the first minute of the treatment, but thereon after remains generally constant. This suggests that parasite death occurs during the first minute of treatment, and after that, parasitic load remains unaffected. When this is compared to fluorescence decay of PPA904, we can see that fluorescence decay of both GFP and PPA904 hit a plateau after the first minute of irradiation.
In our search for a better way to monitor the outcome of PPA904-PDT a CL murine model with GFP+ L. major was used. It was established that live GFP+ L. major parasites fluoresce whereas dead GFP+ L. major parasites do not, and it was determined that fluorescence intensity of GFP has a positive correlation with parasitic load. Efficacy of PDT on the outcome of CL was monitored via measurements of fluorescence intensity of GFP before, right after, and every day after PPA904-PDT. Results show that parasitic load underwent biphasic changes after PDT, a pattern previously reported in our studies of PDT on classic CL murine models [10, 17]. The fact that the same results were obtained using CL murine models with GFP+ L. major shows that the developed model is in fact a feasible way of monitoring the outcome of PDT. In addition to monitoring the outcome of PDT, CL murine models with GFP+ L. major enable us to study changes in parasitic load during the PDT treatment and correlate fluorescence decay of GFP and PPA904.
The development of a CL murine model infected with GFP+ L. major represents a significant advantage when monitoring the clinical outcome of PDT and a big step towards the optimization of this treatment. Studies of PPA904-PDT  have demonstrated that more than one treatment is needed to achieve complete parasite eradication. To determine when the application of a successive treatment would be best, it is necessary to monitor parasitic load every single day after the first treatment has been performed. Currently used methods for monitoring parasitic load after PDT are tedious and time-consuming [10, 17, 21]. Unlike these currently used methods, the use of GFP+ CL murine models represents a much faster, easier way of monitoring parasitic load after PDT treatment. Since parasitic load in CL murine models infected with GFP+ L. major is determined in vivo through fluorescence measurements, mice sacrifice is not necessary, thus minimizing the number of animals needed for the study and the time it takes to monitor the outcome of PDT. Another advantage of using CL murine models infected with GFP+ L. major is the capability of monitoring parasitic load during the process of PDT by measuring GFP fluorescence decay. Being able to monitor decay of GFP fluorescence during PDT allows for a better understanding of the relationship and correlation between PS and parasite death. Furthermore, monitoring fluorescence decay of GFP during PDT may provide new insights on the photophysic aspects of antimicrobial PDT.
The main problem with using GFP as an optical reporter for viability of microbial pathogens in order to monitor infections in animal models is correlating green fluorescence with microbial viability. We used a set of freeze thaw cycles to kill the parasites and when this procedure also produced loss of green fluorescence, we asserted that GFP correlates with viability. Previous studies have successfully demonstrated the parasiticidal effect of various commonly-used antileishmanial drugs using GFP-transfected Leishmania . Alternative killing strategies that do not permeabilize the cell membrane and allow the GFP to leak out were explored on example of miltefosine . The drug has been shown to induce apoptosis-like death of Leishmania parasites without permeabilization the cell membrane. Furthermore, work by Mehta et al. has established GFP+ murine leishmaniasis models as a feasible way of monitoring the outcome (and, hence, parasite burden) of experimental immunotherapy against leishmaniasis . Thus, along with our observation, there are some evidences of GFP being a suitable optical reporter of parasite viability.
PDT has emerged as a very promising and effective treatment for CL [12, 13, 16], but despite the positive results obtained, current PDT treatments only decrease, but do not eradicate, parasitic load . Optimizing the outcome of PDT in murine models of CL is a key step towards the clinical translation of this therapy. Previous studies  have shown that after PPA904-PDT parasitic load was significantly reduced but not completely eradicated. Thus, in the need for an effective way to monitor the outcome of PDT we used a murine model of CL infected with GFP+ L. major. Using this model, parasitic load after PDT was monitored via fluorescence measurements. The same pattern observed in all of these studies suggests that several subsequent sessions of PDT are definitely necessary to achieve parasite eradication with PPA904.
PDT is based on the photo-oxidation of biological materials induced by a PS, which selectively localizes it-self in certain tumorous cells or tissues. When exposed to light of the appropriate wavelength and at a sufficient dose, cells in contact with PS are destroyed. In principle, a photodynamic response leading to cytotoxicity happens whenever a PS and light occur simultaneously. The extent of the response is modulated by PS concentration and by light . Nevertheless, it is important to note that in the majority of PS under investigation, PDT efficacy is also oxygen dependent due to the generation of singlet oxygen during treatment . Results from our studies of fluorescence decay of PPA904 and GFP during PDT have provided new insights on the relationship and correlation between PS and parasite death. It has been observed that during PDT, fluorescence decay of both GFP and PPA904 hit a plateau after the first minute of irradiation, suggesting that fast photobleaching of PS may be responsible for insufficient killing activity of current PDT regimens. We have previously emphasized the importance of adequate PS distribution in the CL lesions in order to achieve an effective PDT treatment . Our current investigation demonstrated the importance of using an adequate fluence rate to prevent fast photobleaching of PS and achieve an effective PDT regimen.
As previously mentioned, the fluence rate used to execute the PDT process is of utmost importance. The use of a lower fluence rate leads to slower PS decay (slower photobleaching), and in consequence, to slower oxygen consumption. On the other hand, the use of higher fluence rates leads to faster oxygen consumption. Oxygen depletion in tissue can be a limiting factor for the PDT process, since under anoxic conditions, the PDT effect of certain PS can be abolished . It has been thought that in clinical settings the use of higher fluence rates to perform PDT would be favorable because the total irradiation time would be reduced; however, reduced efficacy of tumorous tissue has been reported [28, 29]. This lower efficacy has been attributed to oxygen depletion during irradiation, given that during the photochemical reaction oxygen is consumed at a rate higher than the rate of reperfusion. Nevertheless, the use of fluence rates that are too low could end up making the process too time consuming. Due to the importance that fluence rate has on the successful outcome of PDT, the use of a model such as the GFP+ CL murine model that allows the study of the correlation between PS and parasite death and the clinical effects of diverse experimental parameters is of great advantage when it comes to optimizing treatment protocols.
Our studies with GFP+ L. major also suggest that there was not enough PS to achieve L. major eradication. It is possible that in combination with a lower fluence rate, a higher concentration of PS can retard photobleaching effects. Another solution could be reapplication of PS during the PDT treatment. Being able to study the cytotoxic effects of the dynamics of light and oxygen consumption may eventually lead to the development of a shorter and more effective PDT regimen.
Besides contributing to the optimization of current PDT treatments for CL, the developed murine model with GFP+ L. major can serve as the conceptual platform for the development of other GFP+ models that can monitor the outcome and contribute to enhance topical PDT regimens for other granulomatous infections.
This work was supported by the Department of Defense Medical Free Electron Laser Program Grant No. FA9550-04-1-0079 (to T. Hasan) and NIH AI029646 (to S. Beverley). We thank Andreas Hubel for providing constructs and transfected L. major. For this work, Elena Latorre-Esteves was presented with an outstanding poster presentation award at the 2008 Annual Biomedical Research Conference for Minority Students (ABRCMS) in Orlando, Florida.
Elena Latorre-Esteves is a Senior Student in the Department of Biology at the University of Puerto Rico, Mayagüez. She is currently a Minority Access to Research Careers (MARC) Scholar and her main interests reside in fields related to Biomedical Research. After graduation, Ms. Latorre-Esteves plans on pursuing an M.D., Ph.D.
Oleg E. Akilov, M.D., Ph.D. is an Instructor at the Department of Dermatology, University of Pittsburgh. His interdisciplinary expertise spans the fields of dermatology, immunology, immunogenetics, parasitology, and photomedicine, with a special emphasis on cellular aspects of immunology of skin diseases of parasitic and neoplastic origin. For the last six years, he has worked on the optimization of therapeutic strategies for the management of leishmaniasis and cutaneous lymphomas. Dr. Akilov is a co-author of 1 book chapter, 2 patents, and over 30 papers.
Prakash Rai, Ph.D. is currently a Postdoctoral Fellow at the Wellman Centre of Photomedicine, Massachusetts General Hospital, Harvard Medical School, Boston, USA. His expertise ranges in interdisciplinary fields of chemical and biological engineering, chemistry and nanotechnology with a focus on bio-driven self-assembly of nanomaterials, hydrogel-based transdermal drug delivery technology and nanocarrier-based disease theranostics. Dr. Rai is a co-author on 2 patents and 8 papers.
Stephen M. Beverley received his Ph.D. while working with Alan Wilson at the University of California, Berkeley, using molecular phylogenetic approaches to study the colonization of the Hawaiian Islands by Drosophila, where this genus has undergone a spectacular adaptive radiation. He was introduced to Leishmania during postdoctoral work with Bob Schimke at Stanford University, California, in the course of studies of extrachromosomal DNA amplification. He then joined the faculty at Harvard Medical School, Massachusetts, in 1983 where he became the Hsien Wu and Daisy Yen Wu Professor in Biological Chemistry and Molecular Pharmacology. In 1997, he was recruited to Washington University Medical School in St Louis, Missouri, as the Head of the Department of Molecular Microbiology. His lab has focused on the molecular genetics of Leishmania, including the development of DNA transfectional methods and their application to questions such as drug resistance, virulence and the synthesis of the parasite surface glycocalyx.
Tayyaba Hasan, Ph.D., is a Professor in the Department of Dermatology at Harvard Medical School at the Wellman Center for Photomedicine at Massachusetts General Hospital (MGH), and an independent researcher with close to 20 years of experience in cancer research, expertise in photobiology and photodynamic therapy and the mechanisms and strategies for targeted applications, and specifically, in site-directed photochemistry and fluorescence diagnostics. Dr. Hasan is also the Director of the Office for Research Career Development at MGH, and serves on the Subcommittee of Professors and the Faculty Council at Harvard Medical School. In addition, she is an affiliated faculty in the Harvard-Massachusetts Institute of Technology (MIT) Division of Health Sciences and Technology in Cambridge, Massachusetts.
PACS: 00.00.Xx, 11.11.Yy