Validation of molecular typing for Mycobacterium avium subsp. Paratuberculosis (MAP)

Some molecular epidemiological studies about MAP showed the need of a combination of different typing techniques because of limited genetic diversity of MAP. The aim of the study was the validation of such a combined protocol. The typing technique should provide a high discriminatory power to characterize the isolates most comprehensively for epidemiological studies as reliable as possible. About 200 bovine isolates from 56 herds of nine Federal States of Germany, and 28 isolates of other hosts were typed by three established techniques: IS900-RFLP (BstEII and PstI), MIRU-VNTR (8 markers), and MLSSR analysis (4 loci). Application of combined typing revealed a relatively large heterogeneity of MAP in Germany (49 profiles). The results of individual typing methods did not correlate with each other. Some of the MIRU-VNTR and MLSSR markers showed a very low discriminatory power, characterized the same target in the genome or were suggested to be not stable enough for epidemiological studies. Removal of these markers could reduce the effort of the method. Two MAP genotypes were more prevalent than others in Germany. It was possible to validate the combined typing protocol by typing of isolates from well characterized herd and population origin of one federal state. Identical profiles of some cattle and red deer isolates around the Eifel national park suggested a transmission between these species. The same was found for one sheep, one goat and one cattle isolate or between cattle isolates from different herds of another area. When patterns of one technique (RFLP, MIRUVNTR or MLSSR) were omitted, the resulting apparent epidemiological links changed completely. The different kinds of typing techniques (fingerprint-, PCR- und sequence based methods) describe different targets in the genome. Therefore, at present, considering the validity of epidemiological investigations, it is not possible to omit one of the detected typing results randomly.