Background: The incidence of invasive mycoses is increasing worldwide. PCR-RFLP was applied to the
identification of 10 reference strains and 90 cultures of agents of invasive mycoses. In addition, the new
approach was applied to detect fungal agents in 120 biological samples (blood, cerebrospinal fluid and bone
marrow). PCR-RFLP results were compared with the ones obtained with conventional methods (culture,
microscopy, and biochemical testing).
Results: The assays carried out with the reference strains (
Candida albicans
,
Candida parapsilosis
,
Candida
tropicalis
,
Candida krusei
,
Candida guilliermondii
,
Cryptococcus neoformans
,
Cryptococcus gattii
and
Histoplasma
capsulatum
), demonstrated that the RFLP profiles were correctly predicted by the in silico investigation and
allowed unequivocal identification of all chosen reference strains. The PCR-RFLP also identified 90 cultures of
agents of invasive mycoses correctly, 2.5 times faster than the conventional assays. Evaluating PCR-RFLP with
biological samples it was observed that the PCR was found to be 100% accurate and the RFLP profiles allowed
the identification of the etiological agents:
C. neoformans (n = 3) and
C. gattii (n = 1) in CSF samples,
H.
capsulatum (n = 1) in bone marrow and
C. albicans (n = 2) in blood cultures. The detection and identification
by PCR-RFLP were found to be between two to ten times faster than the conventional assays.
Conclusion: The results showed that PCR-RFLP is a valuable tool for the identification of invasive mycoses that can
be implemented in hospital laboratories, allowing for a high number of clinical analyses per day.