Random amplification of polymorphic DNA for tracing and molecular epidemiology of Listeria contamination in a cheese plant

The objective of this study was to evaluate a modified random amplification of polymorphic DNA (RAPD) technique as a tool to determine routes of Listeria contamination in dairy production facilities. Fifty-two strains of L. monocytogenes and L. innocua were isolated from an Austrian cheese factory, and the isolates were analyzed by the RAPD technique. Initially, DNA from all the Listeria strains was subjected to RAPD with 20 arbitrary 10-mer oligonucleotide primers, and the strains were grouped into 10 RAPD groups. In a further step, 200 more primers were evaluated by comparing RAPD patterns of several pools of DNA from each of the subgroups. This approach established that (i) the RAPD was capable of identifying a single polymorphic strain in a pool of 10 closely related strains, and (ii) that no more polymorphic strains of Listeria could be identified with 200 additional primers. We also compared our RAPD differentiation to previously published RAPD methods of Listeria typing, and found them less accurate. Forty-eight of the isolates were serotyped as L. monocytogenes serotypes 1/2a, 1/2b, 3c and 4b; of these, 41 strains clustered into a single group with identical RAPD fingerprints comprising both 1/2a and 1/2b serotypes; six more strains were grouped into four more RAPD groups. The single L. monocytogenes serotype 4b isolate also constituted a unique RAPD group. The remaining four isolates were all L. innocua 6b, and each represented a separate RAPD group. Thus, different serovars of L. monocytogenes composed a single RAPD group, and the same serovars were found in different RAPD groups, indicating a possible recombinational mechanism for generations of serovars. Conversely, a similar mechanism might not be operational in L. innocua, since serovar and RAPD typing was concordant in this species. Initially, PCR detection of the hly locus encoding listeriolysin O, the major virulence factor, in all strains of L. monocytogenes but not in L. innocua suggests that all L. monocytogenes field isolates are potentially pathogenic.