Factorial design for optimization of production and extraction conditions of volatile organic compounds from Colletotrichum sp.

Colletotrichum is among the ten most important genera of pathogenic fungi in the world. Species of this genus are difficult to identify, even by investigating their DNA using multi locus phylogenetic analysis, configuring fungal complexes. Thus, this work aims the use of Volatile Organic Compounds (VOCs) produced by different isolates of Colletotrichum spp., associated with açai (Euterpe oleracea), as possible chemotaxonomic markers of the genus. The study of VOC production, extraction and characterization was performed using headspace solid phase microextraction (HS-SPME) coupled to gas chromatography coupled to mass spectrometry (GC-MS). From a set of 10 fungal isolates, one of them was selected to investigate the optimal conditions aiming at the maximum production of VOCs. The optimization study was performed from a factorial design 24, from which the variables culture medium (Potato Dextrose Agar-BDA, acai fiber), adsorption fiber (PDMS / DVB, DVB / CAR / PDMS), extraction temperature (30 ° C and 50 ° C) and time of the extraction (15 min and 30 min) at the two different levels were evaluated. From the responses (number of peaks) observed additional experiments such as central point and central composite planning (axial points) were performed to evaluate the error associated with the experiment and to verify the presence of quadratic interactions between the studied parameters. The planning data were treated using the response surface methodology: BDA culture medium, DVB / CAR / PDMS volatile extraction adsorption fiber and extraction time of 30 min resulting in a predicted total of 49 peaks (Fig. 1). This result was confirmed by performing the quadruplicate optimal point experiment, resulting in a number of peaks identical to that predicted by the model with 95% confidence. There was no influence of extraction temperature in the system under study, thus, the variation of the responses from the lowest to the highest level was not significant compared to the random error of the model. From the optimum conditions the other fungal isolates will have their VOC profiles explored to propose the chemical markers.