An outline of the Draper's collection method (protocol 1/79) is illustrated in Fig. 1. Infected armadillo liver is homogenized with blender, and the bacterial fraction is collected by centrifugation for 2 times. The sediment is treated with DNase and filtrated through stainless steel mesh. The sediment after centrifugation is suspended in 30% percoll solution and is again centrifuged at 27, 000×g. Bacterial fraction is separated with 2 polymer system such as polyethyleneglycol and dextran. Leprosy bacilli are condenced in upper phase.
A summary of the Mori's collection method is illustrated in Fig. 2. A homogenous suspension of leprosy bacilli is prepared by filtration through 2 sheets of gauze to prevent an agglutination of leprosy bacilli by centrifugation. Percoll gradient is made with 90% percoll 4ml, 80% 5ml, 70% 5ml, 60% 10ml, 50% 5ml and then 8ml of bacterial suspension is placed on the top layer. The tubes are centrifuged with beckman Swinging Bucket SW 27 roter at 27, 000rpm (about 100, 000×g) for 1 hour. Top clear red zone consists of soluble tissue fraction and haemoglobin. Next white zone is insoluble tissue fraction. Middle white zone is rich in leprosy bacilli. Bacterial fraction collected in the bottom of centrifuge tube by Draper's method is more brown than our fraction (Fig. 5 and 6). It may be tissue contaminants. The findings in Ziehl-Neelsen staining of both smear preparations are not different so much. Every bacterial fractions obtained only by centrifugation or 2 polymer system are contaminated with liver tissue components which are stained by anti-armadillo liver rabbit serum and second fluorescent antibody. Host liver tissue contaminant is not found in our alkali treated fraction by the immunofluorescence test. Since the armadillo liver components adhere to the leprosy bacilli fraction, we cannot obtain pure leprosy bacilli without the drastic treatment such as 1 N sodiumhydroxide treatment.
Enzyme treatment was not used in Mori's method, but DNase treatment was used in Draper's method, then it may be unconvenient to extract the deoxynucleic acid from M. leprae. Mori's method finishes in one day, while it takes two days in Draper's method. Yield of leprosy bacilli is 8 % by the Draper's method. On the other hand 56.8 % of leprosy bacilli are collected by Mori's method. A lot of loss in the yield may happen in the 2 polymer system, which is trobublesome, expensive and dangerous to infection. In some improper armadillo livers, brown granules precipitate to the same place of percoll gradient in which the leprosy bacilli precipitate. This armadillo may have some other liver diseases. Therefore, armadillo liver should be treated one by one.