View Report Details

Development of an assay for the quantitative determination of the white fish content of commercial products.

Project Code: Q01051

31/05/2003

Central Science Laboratory
Hird, H ;
Food and Environment Research Agency, Sand Hutton, York YO41 1LZ
Chisholm, J;
Central Science Laboratory
Hold, G

1. The aim of the project was to develop an assay that would allow the accurate

detection and quantification of cod or haddock in food products. The basis of the assay was the hypothesis that the weight of fish tissue would correlate closely with the copy number of a single copy genomic gene and the variation in this measurement would be equal or less than the variation in the nitrogen content of the same tissue. If the hypothesis was proved the project would move to a second phase where an assay to identify and quantify cod or haddock in a complex food product would be developed.

2. The hypothesis was tested by comparing the nitrogen content and copy number of 10 cod and 10 haddoc July ‘01, October ‘01 and January ‘02.

3. Statistical analysis of the data was performed using a nested ANOVA, where the individual data sets were subjected to an estimation of the relative uncertainty associated with the use of the measurement method. The measurement of uncertainty (MU) associated with the analysis of the cod nitrogen, and the haddock copy number and nitrogen showed similar uncertainty (13, 15 and 14% respectively). However, the measurement of cod copy number was much smaller (5%). The variation associated with sampling and analysis of DNA in cod and haddock are approximately equal. The low measurement uncertainty associated with DNA from cod appeared to be due to lower variation between fish and season-ground compared to haddock.

4. On the basis of the initial results and after consultation with the project officers the project was adapted to the development of Taqman based assays for cod and haddock, a strategy more likely to produce a quantitative method than the one proposed in the ROAME.

5. Taqman primer and probe sets were designed to be specific for cod or haddock. Unfortunately the cod primer and probe sets failed to efficiently Page 3 of 55 amplify cod DNA but were subsequently found to amplify cod cDNA in an rtPCR reaction. This indicated that the cod sequence in the NCBI data base is not appropriate for the cod caught in the North East Atlantic and the North Sea.

6. A primer and probe set specific for haddock was optimised for primer length and concentration, and cycling conditions.

7. A supermarket shelf survey of popular products was conducted to ascertain the level of heat processing of the fish incorporated into complex seafood products. The survey found that the fish in commercial products was either raw or heavily processed.

8. The limit of detection of the haddock primer and probe set was assessed using dilutions of haddock template DNA in water and was determined to be a dilution of 10 linear portion of the dilution curve was also at a dilution of 10

9. The effect of processing on the quality of DNA was assessed using DNA extracted from haddock tissue which had been poached, autoclaved, or left raw. There was no significant difference between poached and untreated haddock tissue, however, degradation the DNA by the heat and pressure was evident for autoclaved haddock.

10. Fish tissue in commercial products is either lightly cooked or raw, or highly processed. The closeness of the untreated and poached results indicate that this assay would be applicable to raw or lightly processed food products, but would not however be applicable to highly processed products, other than for haddock identification.

11. Samples were prepared to examine the matrix effects associated with complex food products. Haddock tissue was added in varying quantities to the following matrices: white sauce for lasagne, breadcrumbs from turkey escalopes, cod or Page 4 of 55 water at 90, 75, 50, 25, 10 and 0%. The standard curves show no marked matrix effects at 90, 75, 50 and 25% haddock, however at the 10% level there were statistically significant effects of matrix. The standard curve was found to be relatively flat and therefore inaccurate at low percentages of haddock.

12. Model samples were prepared by spiking commercial products, including high fat, high carbohydrate, high DNA and acidic matrices, with known quantities of haddock tissue. The DNA was then extracted and analysed using the optimised conditions for the haddock primer and probe set. The samples were quantified against a standard curve of haddock in water, since matrix effects had been previously shown to be statistically insignificant above 10% haddock. The calculated percentages are given in Table 3 and appear to be very different from the known haddock composition. The apparent inaccuracy of the calculated percentage does not agree with the data found for matrix effects. These samples were, however, particularly challenging and in hindsight, did not reflect the range of commercial samples which contain fish.

13. In conclusion, an assay specific for haddock which is suitable for quantitation of haddock in raw fish or lightly processed products has been optimised. This assay is now ready for full validation using commercial products.

Main File

Some of the files on this site may be in a format that your computer can't read. However, you can download Readers and Viewers for the following document types below: