In vivo expression analysis of Microsporum canis secreted subtilisin‐like serine proteases in feline dermatophytosis

Abstract

The cellular and molecular pathophysiology of dermatophytosis is largely unknown in both humans and animals. During the past 5 years, we developed a broad research programme to better understand the host‐fungus relationship in Microsporum canis ringworm. This allowed us to isolate and characterize several putative fungal virulence factors, especially keratinolytic proteases, belonging to two gene families encoding three subtilisin‐like serine proteases (SUBs) and three metalloproteases (MEPs). We demonstrated that all the SUBs and two MEPs were produced in hair of experimentally‐infected guinea pigs. Using the same model, we also demonstrated the immunogenicity of at least two proteases, SUB3 and MEP3. However, we recently showed that SUB3 and MEP3 produced as recombinant proteases and used as subunit vaccines failed to protect against a M. canis experimental infection in guinea pigs. As a prelude to investigation of the role of SUBs in M. canis infection in natural hosts and to their potential use as vaccines in cats, this study was designed to establish their presence in in vivo clinical conditions in the cat and human. RT‐nested PCRs using specific primers of internal fragments of SUB1, SUB2 and SUB3 were performed on total RNA extracted from infected hair or scales from naturally infected cats (n = 4) and humans (n = 2). Two specific primer pairs were also designed for the amplification of an internal fragment of M. canis actin mRNA, which was used as a positive control for M. canis RNA extraction and RT‐nested PCR. Fungal strains from which one or more SUB mRNA(s) could not be detected from infected samples were checked by PCR for the presence of the correspondent SUB gene(s) in their genomes. Results indicated that M. canis actin mRNA was detected in all samples, indicating that both M. canis RNA extraction and RT‐nested PCR conditions were successful. The three SUBs were transcribed in two cats while mRNAs encoding two of them were detected in the other two cats. The SUB1 gene was found to be transcribed in humans. All SUB genes where in vivo transcription was not detected were demonstrated to be present in the genomes of the corresponding strains. These results suggest that M. canis SUBs are strongly involved in the cat‐fungus relationship, whereas SUB1 could play a significant role in M. canis dermatophytosis in humans. Indeed, although being a very difficult technique to perform in filamentous fungi, the RT‐nested PCR used in this study demonstrated that all SUB genes could be transcribed in vivo in the cat. It suggests strongly that the corresponding encoded proteases are expressed by the fungus during invasion of feline keratinized structures. Although these data were obtained from a rather small number of samples, they are the first ones dealing with the molecular pathogenesis of dermatophytosis during natural infections. Moreover, they must be included in the frame of our previous and present studies focusing on different aspects of the host‐fungus relationship, including both the characterization of M. canis virulence factors and the host response to infection. Funding: Fonds de la Recherche Scientifique Médicale.