Variation in restriction fragment length polymorphisms among serial isolates from patients with Trichophyton rubrum infection

Molecular genotyping, of strains of Trichophyton rubrum and T. mentagrophyt es from patients with onychomycosis of the toes was performed to ascertain whether the fungal genotype changes over the course of time as sequential s amples were obtained from patients receiving antifungal therapy and during follow-up. Sixty-six serial strains of T. rubrum and 11 strains of T. menta grophytes were obtained from 20 patients (16 patients with T. rubrum, 4 wit h T. mentagrophytes) who were treated with oral antifungal therapy and obse rved over periods of up to 36 months. These strains were screened for genet ic variation by hybridization of EcoRI-digested genomic DNAs with a probe a mplified from the small-subunit (18S) ribosomal DNA and adjacent internal t ranscribed spacer regions. A total of five restriction fragment length poly morphism (RFLP) types were observed among 66 strains of T. rubrum. Two majo r RFLP types, differentiated by one band shift, represented 68% of the samp les. None of the patients had a unique genotype. More than one RFLP type wa s often observed from a single patient (same nail) over a period of 1, 2, o r 3 years, even in cases that did not appear cured at any time. Samples tak en from different nails of the same patient had either the same or a differ ent genotype. The genotypic variation did not correspond to any detectable phenotypic variation. Furthermore, no correlation was observed between the efficacy of the treatment administered and the genotype observed. While the DNA region studied distinguished among T. rubrum, T. mentagrophytes, and T . tonsurans, intraspecific RFLP variation was observed for T. rubrum and T. mentagrophytes strains. While independent multiple infection and coinhabit ation of multiple strains may explain the presence of different genotypes i n a nail, microevolutionary events such as rapid substrain shuffling, as se en in studies of repetitive regions in Candida species, may also produce th e same result. The recovery of multiple strains during the course of sequen tial sampling of uncured patients further suggests that the typing system i s not able to distinguish between relapse or reinfection, ongoing infection , and de novo infection.