Interlaboratory testing of methods for assay of xylanase activity

Twenty laboratories participated in a collaborative investigation of assays for endo-1,4-β-xylanase activity based on production of reducing sugars from polymeric 4-O-methyl glucuronoxylan. The substrates and methods already in use in the different laboratories were first recorded and the apparent activities obtained using these methods in the analysis of a distributed enzyme sample were compared. The standard deviation of the results reported in this analysis was 108% of the mean. Significant reduction in interlaboratory variation was obtained when all the participants used the same substrate for activity determination, each with their own assay procedure. The level of agreement was further improved when both the substrate and the method procedure were standardized. In a round robin testing of a single substrate and method, including precise instructions for enzyme dilution, the standard deviation between the results after the rejection of two outliers was 17% of the mean. This figure probably reflects the inherently poor reproducibility of results when using only partially soluble, poorly defined and rather impure polymeric substrates. The final level of variation was however low enough to allow meaningful comparison of results obtained in different laboratories when using the standardized assay substrate and method procedure. Fifteen laboratories also participated in preliminary testing of an assay based on the release of dyed fragments from 4-O-methyl glucuronoxylan dyed with Remazol Brilliant Blue dye. High values of the coefficients of correlation indicated good linearity between the amount of dyed fragments released and enzyme concentration. The relative standard deviations of the results obtained by fifteen laboratories were about 30% for an optimum range of xylanase activity in the reaction mixture.