Use of volatile fingerprints to detect Penicillium nordicum strains producing ochratoxin A in ham.

The objective of this study was to evaluate (a) the effect of water availability on growth and ochratoxin A (OTA) production by Penicillium nordicum in a ham-based medium and (b) the potential for discriminating between producing (MPVP 1669) and non-producing strains (MPVP 1446) of P.nordicum using volatile production patterns with a hybrid sensor array system ((NST 3220 Lab Emission Analyser, Applied Sensors, Linköping, Sweden). Initial studies using a 3% ham-based medium were used to identify the water activity (aw) range for growth and OTA production at 25°C. This showed that growth occurred over the range 0.99 to 0.90 aw and was able to produce OTA after 7 and 14 days. Subsequent studies were carried out with both the producing and non-producing strains of P.nordicum to examine the potential for differentiation based on qualitative volatile production patterns. Thus the 3% ham media were inoculated with spores by spread plating onto the surface of 5 replicate treatment at 0.995 and 0.95 aw at 25°C for up to 14 days. Plates were destructively sampled after 1, 2, 3, 7 and 14 days. Agar discs (3 x 2 cm) were placed in vials and sealed, placed in the autosampler of the sensor array unit and the head space analysed after 1 hrs incubation. The maximum response of the sensor array (metal ion and metal oxide arrays) were used in the PCA statistical analyses of the data sets. Up to 72 hrs high associations were observed based on aw more than the capacity for the strains to produce OTA. In fact, the strain were grouped at aw 0.95 and 0.995. However, after 7 days incubation the e-nose was able to discriminate the two strains of P. nordicum grown at aw of 0.995, while strains at aw of 0.95 were associated. The most interesting results were obtained after 14 days incubation. The e-nose was able to discriminate between the producer and non-producer strains at both aw levels. This suggests that qualitative volatile production patterns may differ between these two strains because of the biosynthetic pathways involved. This supports previous work with Fusarium verticilioides where similar discrimination was obtained between fumonisin and non-fumonsin strains. The key volatile biomarkers now need to be identified and these may have potential for use as an early indicator of contamination of such food products by mycotoxigenic strains.