To investigate the molecular variability of Mycobacterium avium subsp. paratuberculosis(MAP) in Italy, 8 different strains, collected in different regions and provinces, were analysed with Amplified Fragment Length Polymorphism (AFLP) technique. ATCC strain 19698, isolated in USA, was added to the analysis as a reference and to act as an "outgroup" in the analysis. The AFLP technique produces a large number of molecular markers through selective PCR amplification, with no need of prior sequence information or probe isolation. The information obtained from this analysis will improve the knowledge of the molecular variability of MAP, for which information on genomic sequence polymorphism is poor, and as a tool for epidemiological studies.

Eight different AFLP primer combinations produced a total of 193 polymorphic fragments (88%) with an average of 24.12 fragments per primer pair and a range between 1 and 85. The PstG-MseTT primer combination produced 86 fragments, 85 of which were polymorphic (99%), and was able to differentiate all 8 strains in a single experiment. These preliminary data indicated that the AFLP approach is able to produce a sufficient number of polymorphic markers to distinguish between different MAP strains.

A genetic distance matrix between MAP strains was generated using PHYLIP software package and UPGMA cluster analysis was performed on Nei-Li (1979) model. To obtain a robust tree, bootstrap was set at 1000 and the CONSENSE option selected.

The UPGMA tree showed three different clusters of strains: the major one included five strains, the second grouped two strains and the third comprised an Italian strain, isolated in Cuneo, and ATCC 19698. There was no relationship between clusters and geographic locations where isolates were obtained. Additional epidemiology information identifying contacts between farms may explain the UPGMA data. Ongoing work is examining the variability of MAP strains within a restricted geographical region.