Numerical analysis of PAGE protein patterns and the taxonomic relationships within the 'Mycoplasma mycoides cluster'.

Twenty-six isolates belonging to the 'Mycoplasma mycoides cluster' have been characterized by one-dimensional SDS-PAGE of their cellular proteins. A numerical classification based on the resulting patterns and using a correlation coefficient revealed four distinct phenons at a similarity (S) level of 70%, comprising: (a) bovine group 7 strains; (b) M. capricolum and F38-like strains; (c) M. mycoides subsp. capri and LC strains ('subsp. mycoides'); (d) M. mycoides subsp. mycoides (SC). At the 75% S level, they could be divided further to give eight phenons. The composition of the clusters at both levels was in good agreement with their previous classification, except for M. mycoides subsp. mycoides LC and M. mycoides subsp. capri, which were clustered in a single phenon at 70% S and could not be clearly separated at 75% S. We conclude that high-resolution SDS-PAGE, combined with computerized analysis of protein patterns, provides an extremely effective approach to the investigation of taxonomic relationships within this group of mycoplasmas.