Polymorphism-specific PCR enhances the diagnostic performance of American tegumentary leishmaniasis and allows the rapid identification of Leishmania species from Argentina
© Marco et al.; licensee BioMed Central Ltd. 2012
Received: 15 June 2012
Accepted: 8 August 2012
Published: 15 August 2012
The diagnosis of the leishmaniases poses enormous challenges in Argentina. The Polymorphism-Specific PCR (PS-PCR) designed and validated in our laboratories has been proven effective for typifying the Leishmania genus from cultured material. Here we evaluated the performance of this method in the diagnosis of American tegumentary leishmaniasis (ATL) and the rapid identification of Leishmania spp. directly from clinical specimens.
A total of 63 patients from northwestern Argentina, with cutaneous or mucocutaneous lesions, underwent an ATL diagnosis protocol which included clinical examination, Leishmanin skin test, and microscopic examination of dermal smears. In addition, we performed PS-PCR on DNA directly extracted from the specimens scraped from the lesions.
Out of the 63 patients, 44 were classified as ATL cases and 19 as non-ATL cases. The diagnostic sensitivity of the microscopic analysis of dermal smears and PS-PCR individually were 70.5% and 81%, respectively. When performing both tests in parallel, this parameter increased significantly to 97.6% (p = 0.0018). The specificities, on the other hand, were 100%, 84.2%, and 83.3% for the combination, respectively (p > 0.05). Using the PS-PCR analysis we successfully identified the Leishmania spp. in 31 out of the 44 ATL cases. Twenty-eight (90.3%) cases were caused by L. (V.) braziliensis, two (6.5%) by L. (V.) guyanensis, and one (3.2%) by L. (V.) panamensis.
The efficacy of the ATL diagnosis was significantly improved by combining the dermal smear examination with a PS-PCR analysis. Our strategy allowed us to reach the diagnosis of ATL with high accuracy regarding the species of the etiological agent in 70.5% of the cases. Moreover, we diagnosed two cases of the disseminated cutaneous form caused by L. (V.) braziliensis and a cutaneous case due to L. (V.) panamensis infection, both findings reported for the first time in Argentina.
KeywordsLeishmaniasis Tegumentary leishmaniasis Leishmania braziliensis Leishmania guyanensis Leishmania panamensis PS-PCR Diagnosis Argentina Species identification Leishmania strains
The leishmaniases are a group of neglected tropical diseases caused by flagellated parasites of the genus Leishmania. According to the World Health Organization (WHO), they remain as category I TDR diseases, signifying their status as emerging and uncontrolled, and highlighting the need for new and better tools for their diagnosis, treatment and prevention .
American tegumentary leishmaniasis (ATL) is endemic in Argentina with an estimated incidence of 8.76 cases/year/106 inhabitants, calculated from 1984 to 2005 . However, these data might not be accurate since the laboratory resources for confirming ATL are not always available, and local physicians have to report suspected cases merely based on clinical evidence without any diagnostic test. Furthermore, the visualization of amastigotes in smears prepared with material scraped from lesions, requires highly trained personnel, is time consuming, and often shows low sensitivity . In the north of Salta province, where 53.1% of all cases reported in the Argentina occur, some institutions enforce applying the Montenegro skin test (MST) together with microscopic examination of smears to diagnose ATL. Although MST detects past infections or previous contact with the parasite, it is a useful tool for screening or as a complementary test for the diagnosis of an active disease .
Leishmania species identification, very important for prescribing the appropriate treatments and evaluating the prognosis, is mainly performed for research purpose, and is rarely applied in clinical practice in Argentina. In a previous work, applying multilocus enzyme electrophoresis (MLEE), we determined that Leishmania (Viannia) braziliensis was the prevalent species causing three different clinical forms of ATL: cutaneous (CL), mucocutaneous (MCL), and recurrent cutaneous leishmaniasis in Salta province. Leishmania (Leishmania) amazonensis and L. (V.) guyanensis were also isolated from patients with CL from this area, however only few isolates were specifically characterized so far [4, 5].
Considering the presence of several dermatological diseases with similar clinical manifestations to ATL, the reduced applicability of serological diagnostic methods because of their cross reactivity with American trypanosomiasis  , and the presence of at least three Leishmania spp. in the area, techniques based on the DNA analysis have been developed as alternative approaches for the diagnosis of ATL and the typing of the Leishmania genus . Among them, the polymorphism-specific PCR (PS-PCR) has been established in our laboratories for the identification of the five main species responsible for ATL . Moreover, this procedure was validated against MLEE, the gold standard method, using Leishmania isolates from local patients cultured in vitro.
In the present study we evaluated the performance of a modified PS-PCR approach applied directly on clinical samples. Our results demonstrate that the method readily permits the detection and species identification of parasites causing ATL in Argentina. Furthermore, we believe that using this approach might have an impact on the clinical-epidemiological management of patients with ATL.
All patients voluntarily requested the tests for the differential diagnosis of their lesions, and consented to participate anonymously in this study. The procedures were approved by the Bioethics Committee of the Health Ministry of Salta, Argentina, and followed the Declaration of Helsinki Principles.
Patients and diagnosis
Sixty-three patients with cutaneous or mucocutaneous lesions suspected of leishmaniasis were included in this study. Patients were diagnosed in three different institutions; the Research Institute of Tropical Diseases, University of Salta, Oran; the Department of Dermatology, Milagro Hospital; and the Department of Dermatology and Otorhinolaryngology, San Bernardo Hospital. All of them are located in Salta, Argentina.
The diagnosis of ATL was performed using a combination of the three criteria currently applied in this area, and described as follows:
Search for Leishmania amastigotes was done on smears of dermal scrapings from the border of the lesions, obtained with sterile wooden toothpicks, stained with May-Grunewald Giemsa and examined under a microscope as described previously .
Montenegro skin test
Leishmanin antigen was prepared from a culture of L. (V.) braziliensis strain MHOM/AR/03/OLO1, following previously described protocols [4, 5]. Four μg (100 μl) of total promastigotes proteins were injected intradermally on the ventral forearm of the patients. Readings were taken 48 hours after the injections, and indurations measuring ≥ 5 mm in diameter were considered reactive .
The inclusion criterion was the presence of compatible tegumentary injuries, which included ulcerative, nodulous, papulous cutaneous or mucocutaneous lesions of two or more weeks of evolution, and a congruent epidemiological history. Previous diagnoses of CL were taken into account for the diagnosis of secondary MCL cases.
All patients diagnosed in the present work as ATL cases were systemically treated with 10–20 mg d−1 kg body wt−1 of pentavalent antimony over 25–30 days. In the cases of incomplete clinical cure, another treatment-cycle of the same extension with antimony or amphotericin B was given. The treatments and clinical follow-up were conducted by local physicians.
DNA samples for PCR assays
Specimens were obtained by scrapping the edge of the lesions with sterile wooden toothpicks and placed in a tube containing 200 μl of TE buffer. The material was heated in a waterbath at boiling temperature for 10 min and then stored at −20°C . DNA was extracted with a phenol-chloroform mixture, precipitated with ethanol and dissolved in 20 μl of TE buffer. A titration assay showed that a 1/20 DNA dilution was optimal for the PS-PCR analyses. Furthermore, the detection limit of the reaction was estimated in 0.3 ng/μl of total Leishmania DNA.
Four Leishmania strains, MHOM/AR/03/OLO1 of L. (V.) braziliensis, MHOM/AR/99/JDM1 of L. (V.) guyanensis, MHOM/PA/71/LS94 of L. (V.) panamensis, and MHOM/BR/73/M2269 of L. (L.) amazonensis, were used as controls. Their DNA was extracted from promastigotes cultures according to described protocols [5, 12], and purified using the GenomicPrep Cell and Tissue DNA Isolation Kit (Amersham Biosciences, NJ) following the manufacturer’s instructions. The DNA concentrations were determined in a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE), and adjusted to 10 ng/μl with Milli Q water.
This analysis took into account of the genomic differences among species of the Leishmania genus. We first performed Arbitrarily-Primed-PCR on total DNA from the five most prevalent species associated with ATL and sequenced those DNA fragments present in single species to design the primers addressing these polymorphisms. In addition, the PCR conditions were adjusted increasing the stringency in order to have the necessary specificity for each reaction . The reactions were performed in duplicate in a GeneAmp PCR System 2400 (Perkin Elmer, Wellesley, MA) using the Roche GeneAmp XL PCR Kit (Applied Biosystems, Foster City, CA) in 15 μl final volume.
Polymorphism-Specific PCR. Primer sequences and expected size of amplicons
The second step addressed the identification of Leishmania species. The reactions were performed with three different sets of primers, b1-b2 for L. (V.) braziliensis, p1-p2 for L. (V.) panamensis and g1-g2 for L. (V.) guyanensis on DNA which gave the expected products with primers V1-V2, under conditions previously reported .
If a sample resulted positive for reaction B, it was assayed again with the same M1-M2 primers, but at 67.5°C or 72°C annealing temperature, which are specific for L. mexicana complex and L. (L.) amazonensis, respectively .
DNA extracted from Leishmania promastigotes cultured in vitro or from confirmed ATL patients was used as a positive control, whereas DNA of eight healthy individuals from non-endemic area (Japanese subjects) or from lesions of non-ATL patients was used as negative control.
Additional experiments to control the DNA preparation procedure, as well as to detect the presence of PCR inhibitors were performed using the TaKaRa Ex Taq DNA polymerase Hot Start Version (Takara-Bio, Shiga, Japan) and specific primers for the human glyceraldehydes 3-phosphate dehydrogenase (GAPDH) gene, following a previously reported protocol .
The PCR products were separated in agarose gels containing ethidium bromide, visualized under UV light and the results were recorded with a Kodak EDAS 290 gel documentation system (Kodak, Rochester, NY).
The diagnostic performance of the tests alone or combined was assessed by calculating several statistical indicators, including sensitivity, specificity and predictive values. First the patients were classified into two groups: ATL cases and non-ATL cases based on the presence or absence of the disease. Then the results of the PS-PCR method for each patient were added to the tables and analyzed independently of the ATL group definition. The sensitivity, specificity, predictive values, and the Cohen’s kappa coefficient were estimated applying Epidat 3.1 software (Pan American Health Organization, Spain) . The predictive values for the laboratory were calculated by applying the Theorem of Bayes assuming a prevalence of 54.82% for the year 2005 with the same software. The proportions were compared by Statistic-z analysis or the Fisher’s exact test . The positive and negative PCR controls, or results from non-endemic healthy controls were not included in any of these calculations.
Results of smear and PS-PCR individually, or combined in parallel, applied to 63 patients
Smear + PS-PCR
Performance of PS-PCR in the diagnosis of American tegumentary leishmaniasis
Performance of smears and PS-PCR analyses, and their combination, in the diagnosis of ATL
Smear + PS-PCR
The PS-PCR has shown a significant higher sensitivity than smears for the diagnosis of MCL. Among nine patients with this clinical form, eight were found positive by PS-PCR, while only two by smear observation (p < 0.05).
Control PCRs with primers for the human GAPDH gene demonstrated that the sampling and purification steps of the protocol were effective and also that the samples were devoid of PCR inhibitors. However, weak bands were obtained with two out of the 44 samples, most likely due to low DNA concentration. Since these samples did not give conclusive results with V1-V2 and M1-M2 primers, they have not been considered for calculations of the sensitivity and the positive predictive values (Table 2).
Identification of Leishmania species by PS-PCR
Leishmania species identified by PS-PCR causing ATL in NW Argentina
L. (V.) braziliensis
L. (V.) guyanensis
L. (V.) panamensis
L. mexicana or donovani complex
The patients with DSCL, caused by L. (V.) braziliensis, had more than 30 lesions each, distributed throughout the body, compromising face, abdomen, extremities, back, and oro-nasal mucosae. The lesions showed different degrees of evolution, from the early papular to the predominant ulcerated form, developing within a two month period following the appearance of the single cutaneous lesion (Figure 1). Leishmania amastigotes were found in the smears from several lesions and the MST was reactive in all cases. All these patients responded well to the antimonial treatment.
The development and/or improvement of diagnostic tools remain a priority in the research field of leishmaniasis. In particular, much work has to be done for increasing the sensitivity and specificity of the diagnostic tests for ATL in endemic areas, such as NW Argentina . In this work, we assessed both the impact of adding a PS-PCR analysis to the diagnosis-methods currently used in Argentina, and its usefulness to the identification of the causal-agent of the disease at the species level.
Indeed, by performing PS-PCR directly on clinical samples, we enhanced significantly the efficacy of the ATL diagnosis, especially for MCL. Appling this method to patients with negative results for the microscopic search of parasites in dermal smears, allowed us to increase substantially the diagnostic sensitivity from 70.5% for smears alone to 97.6% for the combination (Table 3). Thus, these results confirm that the PCR analysis is a valuable and applicable strategy for these cases of suspected ATL with negative microscopic examination. The relevance of this strategy becomes apparent when considering the availability of effective treatments or the tendency of the disease to evolve into severe clinical forms if left untreated. These factors indicate that the sensitivity of the diagnostic method should have the priority for maximization, rather than its specificity . Previous PCR analyses for the diagnosis of leishmaniases displayed sensitivities ranging from 51.7% to 100% [18–20]. This wide variability in the sensitivity may very well be due to the methods employed, but also attributable to differences in the sampling techniques, or the clinical and laboratory criteria applied for defining the presence or absence of the disease . The sensitivity of the PS-PCR analysis estimated in this study (81%) is in close agreement with those reported by other investigators when the method was performed on samples from cutaneous lesions of patients with ATL from endemic areas of Ecuador [21, 22]. Among the reasons for false negative PS-PCR results, the low parasite burden observed in the borders of the lesions might be of importance. Therefore, improving the sampling method, for example, using a syringe to aspirate the sample or collecting the exudates after scraping the lesions, in addition to efficient DNA extraction, would appear to substantially increase sensitivity and diagnostic efficacy .
In the present work, three non-ATL cases were PS-PCR positive, lowering the specificity to 84.2%. Since these patients presented atypical lesions, and were negative for the microscopic parasitological examination and the MST test, we think that these results were due to contamination of the DNA sample.
An identification of Leishmania spp. with PS-PCR applied directly on material scrapped from the lesions was achieved in 70.5% of ATL cases. Leishmania (V.) braziliensis was the most prevalent species (90.3%) among ATL patients, and this is consistent with previous studies aimed to characterize Leishmania stocks isolated in NW Argentina [5, 23]. Since we performed the analysis on samples collected directly from the lesions, the indicators of species prevalence reported here might be more accurate than those published before. Moreover, the method has additional advantages, as it gives the Leishmania spp. identification without further procedures, such as digestion analysis of the PCR products or gene sequencing, and it is applicable in relatively low-resource laboratories.
Leishmania (V.) braziliensis was involved in two cases DSCL reported for the first time in Argentina (Figure 1). Local ATL cases due to L. (V.) guyanensis were also found, confirming our previous report on the presence of this species in this area . These findings are of clinical relevance for the patients, because they require a different treatment, such as a protocol with pentamidine rather than the pentavalent antimonials commonly prescribed in this area [19, 24]. Interestingly L. (V.) panamensis was identified in one sample (Figure 2C). This is the first case reported in Argentina. Although parasites belonging to the Leishmania mexicana or donovani complexes were found involved in cases of leishmaniasis in this area [4, 25], we could not prove their involvement in the present series (Table 4).
In conclusion, by adding the PS-PCR method to the procedures commonly applied in the country, we improved the diagnostic performance of ATL. The strategy also allowed the identification of Leishmania spp. in the vast majority of the cases without isolation of the parasites in vitro.
We are grateful to Keiko Morisawa, Manuel Calvopiña, José Gil, Ruben Cimino, Alejandra Barrio, Kyoko Imamura and Daniel Ribble for their contributions and valuable comments. We also thank the physicians and personnel of the Department of Dermatology, Milagro Hospital and the Department of Dermatology and Otorhinolaryngology, San Bernardo Hospital, Salta, for recruiting the patients.
Financial support was obtained from CONICET (National Scientific and Technical Research Council, Argentina), the National Agency for Science and Technology, Argentina (PICT-2009-0135; PICT-2011-1897). PS-PCR analysis was supported by the Ministry of Education, Science, Culture and Sports of Japan (Grants No. 18256004 and 23256002) and the Japanese Society for the Promotion of Science (Grant KAKENHI 18–06242).
- Report of the Scientific Workving Group meeting on Leishmaniasis. 2004, World Health Organization, Geneva, 5-16.
- Salomon OD, Quintana MG, Zaidenberg M: Urban distribution of Phlebotominae in a cutaneous leishmaniasis focus, Argentina. Mem Inst Oswaldo Cruz. 2008, 103 (3): 282-287. 10.1590/S0074-02762008005000016.View ArticlePubMedGoogle Scholar
- Sosa Estani S, Campanini A, Sinagra A, Luna C, Peralta M, Coutada V, Medina L, Riarte A, Salomon D, Gomez A, et al: Clinical features and diagnosis of mucocutaneous leishmaniasis in patients of an endemic area in Salta. Medicina (B Aires). 1998, 58 (6): 685-691.Google Scholar
- Frank FM, Fernandez MM, Taranto NJ, Cajal SP, Margni RA, Castro E, Thomaz-Soccol V, Malchiodi EL: Characterization of human infection by Leishmania spp. in the Northwest of Argentina: immune response, double infection with Trypanosoma cruzi and species of Leishmania involved. Parasitology. 2003, 126 (Pt 1): 31-39.View ArticlePubMedGoogle Scholar
- Marco JD, Barroso PA, Calvopina M, Kumazawa H, Furuya M, Korenaga M, Cajal SP, Mora MC, Rea MM, Borda CE, et al: Species assignation of Leishmania from human and canine American tegumentary leishmaniasis cases by multilocus enzyme electrophoresis in North Argentina. Am J Trop Med Hyg. 2005, 72 (5): 606-611.PubMedGoogle Scholar
- Gil JF, Hoyos CL, Cimino RO, Krolewiecki AJ, Lopez Quiroga I, Cajal SP, Juarez M, Garcia Bustos MF, Mora MC, Marco JD, et al: Role of three ELISA tests using promastigote homogenates of Leishmania braziliensis, L. amazonensis and L. guyanensis in the diagnosis of tegumentary leishmaniasis. Medicina (B Aires). 2011, 71 (5): 420-428.Google Scholar
- Reithinger R, Dujardin JC: Molecular diagnosis of leishmaniasis: current status and future applications. J Clin Microbiol. 2007, 45 (1): 21-25. 10.1128/JCM.02029-06.View ArticlePubMedGoogle Scholar
- Mimori T, Sasaki J, Nakata M, Gomez EA, Uezato H, Nonaka S, Hashiguchi Y, Furuya M, Saya H: Rapid identification of Leishmania species from formalin-fixed biopsy samples by polymorphism-specific polymerase chain reaction. Gene. 1998, 210 (2): 179-186. 10.1016/S0378-1119(97)00663-X.View ArticlePubMedGoogle Scholar
- Marco JD, Uezato H, Mimori T, Barroso PA, Korenaga M, Nonaka S, Basombrio MA, Taranto NJ, Hashiguchi Y: Are cytochrome B gene sequencing and polymorphism-specific polymerase chain reaction as reliable as multilocus enzyme electrophoresis for identifying Leishmania spp. from Argentina?. Am J Trop Med Hyg. 2006, 75 (2): 256-260.PubMedGoogle Scholar
- Padilla AM, Marco JD, Diosque P, Segura MA, Mora MC, Fernandez MM, Malchiodi EL, Basombrio MA: Canine infection and the possible role of dogs in the transmission of American tegumentary leishmaniosis in Salta, Argentina. Vet Parasitol. 2002, 110 (1–2): 1-10.View ArticlePubMedGoogle Scholar
- Harris E, Kropp G, Belli A, Rodriguez B, Agabian N: Single-step multiplex PCR assay for characterization of New World Leishmania complexes. J Clin Microbiol. 1998, 36 (7): 1989-1995.PubMedPubMed CentralGoogle Scholar
- Evans D: Laboratory maintenance of organisms. Handbook on Isolation, Characterization, and Cryopreservation of Leishmania. World Health Organization Tech Rep Ser. Edited by: Evans D. 1989, World Health Organization, Geneva, 12-14.Google Scholar
- Report of a meeting of the WHO Expert Committee on the Control of Leishmaniases. 2010, World Health Organization, Geneva
- Eresh S, McCallum SM, Barker DC: Identification and diagnosis of Leishmania mexicana complex isolates by polymerase chain reaction. Parasitology. 1994, 109 (Pt 4): 423-433.View ArticlePubMedGoogle Scholar
- Shuto T, Furuta T, Oba M, Xu H, Li JD, Cheung J, Gruenert DC, Uehara A, Suico MA, Okiyoneda T, et al: Promoter hypomethylation of Toll-like receptor-2 gene is associated with increased proinflammatory response toward bacterial peptidoglycan in cystic fibrosis bronchial epithelial cells. FASEB J. 2006, 20 (6): 782-784.PubMedGoogle Scholar
- Santiago Perez MI, Hervada Vidal X, Naveira Barbeito G, Silva LC, Farinas H, Vazquez E, Bacallao J, Mujica OJ: The Epidat program. Rev Panam Salud Publica. 2010, 27 (1): 80-82. 10.1590/S1020-49892010000100012.View ArticlePubMedGoogle Scholar
- Azzimonti Renzo JC: Capítulo 4: Índices Clínicos y Estadígrafos. Bioestadística aplicada a Bioquímica y Farmacia. Edited by: Azzimonti Renzo JC. 2003, Editorial Universitaria de la UNaM, Misiones, Argentina, 2-3. Second editionGoogle Scholar
- Bensoussan E, Nasereddin A, Jonas F, Schnur LF, Jaffe CL: Comparison of PCR assays for diagnosis of cutaneous leishmaniasis. J Clin Microbiol. 2006, 44 (4): 1435-1439. 10.1128/JCM.44.4.1435-1439.2006.View ArticlePubMedPubMed CentralGoogle Scholar
- Rotureau B, Ravel C, Couppie P, Pratlong F, Nacher M, Dedet JP, Carme B: Use of PCR-restriction fragment length polymorphism analysis to identify the main new world Leishmania species and analyze their taxonomic properties and polymorphism by application of the assay to clinical samples. J Clin Microbiol. 2006, 44 (2): 459-467. 10.1128/JCM.44.2.459-467.2006.View ArticlePubMedPubMed CentralGoogle Scholar
- Barrio A, Mora MC, Ramos F, Moreno S, Samson R, Basombrio MA: Use of kDNA-based polymerase chain reaction as a sensitive and differentially diagnostic method of American Tegumentary Leishmaniasis in disease-endemic areas of northern Argentina. Am J Trop Med Hyg. 2007, 77 (4): 636-639.PubMedGoogle Scholar
- Matsumoto T, Hashiguchi Y, Gomez EA, Calvopina MH, Nonaka S, Saya H, Mimori T: Comparison of PCR results using scrape/exudate, syringe-sucked fluid and biopsy samples for diagnosis of cutaneous leishmaniasis in Ecuador. Trans R Soc Trop Med Hyg. 1999, 93 (6): 606-607. 10.1016/S0035-9203(99)90065-2.View ArticlePubMedGoogle Scholar
- Mimori T, Matsumoto T, Calvopina MH, Gomez EA, Saya H, Katakura K, Nonaka S, Shamsuzzaman SM, Hashiguchi Y: Usefulness of sampling with cotton swab for PCR-diagnosis of cutaneous leishmaniasis in the New World. Acta Trop. 2002, 81 (3): 197-202. 10.1016/S0001-706X(01)00215-7.View ArticlePubMedGoogle Scholar
- Krolewiecki AJ, Romero HD, Cajal SP, Abraham D, Mimori T, Matsumoto T, Juarez M, Taranto NJ: A randomized clinical trial comparing oral azithromycin and meglumine antimoniate for the treatment of American cutaneous leishmaniasis caused by Leishmania (Viannia) braziliensis. Am J Trop Med Hyg. 2007, 77 (4): 640-646.PubMedGoogle Scholar
- Minodier P, Parola P: Cutaneous leishmaniasis treatment. Travel Med Infect Dis. 2007, 5 (3): 150-158. 10.1016/j.tmaid.2006.09.004.View ArticlePubMedGoogle Scholar
- Barrio A, Parodi CM, Locatelli F, Mora MC, Basombrio MA, Korenaga M, Hashiguchi Y, Garcia Bustos MF, Gentile A, Marco JD: Leishmania infantum and human visceral leishmaniasis, Argentina. Emerg Infect Dis. 2012, 18 (2): 354-355.View ArticlePubMedPubMed CentralGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2334/12/191/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.