Mycobacterium ulcerans, a bacterium belonging to the same family as M. tuberculosis and M. leprae, is the causative agent of Buruli ulcer (BU). BU has been described as a neglected tropical disease and is the third most common mycobacterial infection after tuberculosis and leprosy 1 . It is a necrotizing, painless, cutaneous infection mainly localized on the limbs of affected individuals and causes extensive damage to the skin, its associated tissues and even the bone. The pathology of the disease is due to mycolactone, a necrotizing and immunosuppressive lipid toxin produced by the bacterium. The genes that code for the production of the toxin are located on a 174 kb plasmid. In endemic countries like Ghana, Cote d'Ivoire and Benin, the disease is prevalent in rural poor communities. In the most endemic district in Ghana a prevalence of up to 150.8 per 100,000 individuals has been reported 2 . However epidemiological studies suggest underreporting and improper diagnosis as some of the factors hindering the determination of the exact disease burden in endemic areas. In all endemic areas, the infection burden has long term socio-economic impacts on infected individuals and their communities.

Current reporting of cases of BU infection is based on the presentation of the symptoms. This represents a challenge due to the vast number of other skin infections or conditions that may exhibit symptoms similar to that of BU 3 4 5 . The WHO has therefore directed that all clinically diagnosed or suspected cases of BU be confirmed. Currently, this can only be done in reference laboratories, since the current methods are not amenable to point of care diagnosis.

Laboratory confirmation of BU is complex and has evolved over the years. Mycobacterium ulcerans stains red (acid-fast bacilli, AFB) in the Ziehl-Neelsen staining procedure but this method has a low sensitivity 6 . Swabs taken from lesions often do not show AFB by microscopic examination. Culturing M. ulcerans from clinical samples is difficult and has a low sensitivity of about 35-60% 6 7 . The bacterium is notoriously slow-growing (6-8 weeks) and culture media are frequently contaminated with other faster growing species 8 . PCR methods have been developed for BU diagnosis based on the 16S rRNA gene 6 , the hsp65 gene 9 , or the insertion sequence IS-2404 10 . Although the sensitivity of PCR is high (98%), this method is expensive and requires technical expertise in terms of DNA extraction and equipment needed.

Notwithstanding the fact that a combination of these methods can lead to an accurate diagnosis, the highly technical and expensive nature of the techniques confines them to reference laboratories. Thus, there is no simple, rapid test that is appropriate for early point of care diagnosis in the low-resourced laboratory settings where the disease is most prevalent. This represents a huge gap between early diagnosis before ulcers and deformities occur and the critical need for treatment and prevention of associated deformities. We report the development of a simple and relatively inexpensive test for M. ulcerans diagnosis, which could easily be applied in basic healthcare facilities, without recourse to expensive, complex and time-consuming methods. This new method is based on a novel DNA amplification method, developed by Notomi and colleagues 11 .

The methodology used is termed loop mediated isothermal amplification (LAMP) technique and has been applied for the molecular diagnosis of a variety of diseases including; P. falciparum malaria, West Nile Virus, Influenza, Tuberculosis, Trypanosomiasis and filarial detection in mosquitoes 12 13 14 15 16 .

The diagnosis of BU relies primarily on culture 8 and, conventional and real-time PCR methods 6 10 18 19 . However, these methods are time consuming, expensive and limited to well-resourced laboratories. The BU-LAMP assay on the other hand is amenable to point of care diagnosis in rural health facilities, due to the ease, simplicity and low technical expertise required.

The BU-LAMP assay specifically amplified DNA from M. ulcerans but not from the closely related M. marinum or any of the other Mycobacterium species tested. These results clearly demonstrate the high specificity of the BU-LAMP assay in detecting M. ulcerans, compared to other methods, such as the IS-2404 PCR 10 20 , which amplifies other environmental mycobacteria, especially those that produce mycolactone 21 22 . This new assay also has a very high sensitivity, as it proved to be 10 times more sensitive than the most used PCR method, based on IS-2404 10 20 . We noted that the high level of sensitivity of the LAMP assay at amplifying minute amounts of DNA was one of the things that needed fine tuning in subsequent research, particularly for adaptation to point of care facilities.

The evaluation of the BU-LAMP method in animal biopsy samples showed successful amplification of DNA by the Qiagen extraction and boil preparation methods, and the use of raw samples. However, the failure of raw samples in producing visibly-detectable results may be because there was not enough DNA that could be amplified, or due to the presence of inhibitors, which may have been inactivated through heat incubation (boil preparation) or the pure Qiagen DNA extraction method. Nonetheless, the fact that heat-treated samples produced visibly detectable results is of significance for the applicability of this method in resource-poor settings. The use of heat-treated samples in low-resource settings eliminates the requirements for time-consuming and expensive DNA extraction methods. The time involved from DNA extraction to result was considerably lower, since the BU-LAMP results can be obtained within 1.5 hours compared to about 6 hours for the conventional PCR. In addition, costing of this new assay was estimated at US$ 4 per sample, compared to the conventional PCR method of approximately US$ 6.0 per sample. The assay also eliminates the use of thermal-cyclers and expensive UV illuminators. The need for UV lights can be removed completely through the use of SYBR green dye 23 .

Of the human samples analyzed, 59% and 27% were positive using the BU-LAMP method on Qiagen and boil-prepared- DNA respectively. Only 14% of samples were positive using the conventional PCR method on Qiagen extracted DNA. The discrepancy between BU-LAMP positive and PCR positive results could be attributed to the high sensitivity of the BU-LAMP method and its ability to detect DNA quantities 10 times lower than the sensitivity limit of 0.1 M. ulcerans genome equivalent established for the IS-2404 method 20 . These results may have been affected by various factors such as the time from sample collection to processing, the medium of storage and the reliability of the collections. The samples tested had been in storage for at least 5 months; therefore, future implementation research studies should focus on analyzing samples immediately after they have been collected. Although the total sample size was small, fine needle aspirates, punch biopsies and tissue in medium provided the best results and may be considered as the best sample types for field application of the BU-LAMP assay. The extent of disease will also have to be considered in the acquisition of samples. Despite all the possible limitations and the low number of samples analyzed, the results show that in our study the LAMP method performed better than the conventional PCR method.

In conclusion, the BU-LAMP method shows promise as a diagnostic tool at point of care facilities in BU endemic communities. This new method requires very little manipulation of BU samples, and amplifies DNA with high specificity and rapidity under constant temperature conditions (in a water bath). Unlike the conventional PCR method that requires the use of a thermal cycler, purified DNA samples are not a requirement. DNA amplification occurs within 1 hour and the resulting product is a turbid solution, indicative of product amplification. Sample confirmation can therefore be done visually with the naked eye, and the intensity of the fluorescence observed is indicative of amount of DNA present. In addition to these characteristics, other basic requirements such as cold storage facility for reagents and water bath are usually available at point of care facilities and therefore make the BU-LAMP method feasible as a point of care test. However, there is the need for further studies to determine the sensitivity of the BU-LAMP assay, on a larger sample size, of freshly collected sample types (skin biopsy, punch biopsy, swabs and fine needle aspirates), and on the efficacy and applicability of the assay in low-resourced laboratory settings.

The authors declare that they have no competing interests.

DKS conceived the study, participated in the study design, the experiments and drafting of the manuscript. CQ and LM designed and performed the experiments and drafted the manuscript. PA provided M. ulcerans strains and animal biopsy samples and drafted the manuscript. DAB designed the study and drafted the manuscript. All authors read and approved the final manuscript.