3,5-Diaminobenzoic acid reacted rapidly with the product from HNO2 deamination of heparin, heparan sulphate and 2-amino-2-deoxyhexoses under very mild conditions (pH 3.0 and 37°C) to give stable fluorescent derivatives. The fluorescence yield was rectilinearly related to the concentration of heparin etc. Less than 0.1 μg of 2-amino-2-deoxyhexose was easily measurable in standard cuvettes. The deamination products of glucosamine and (particularly) galactosamine were labile in the HNO2 reagent, with half-lives of 20-40 min at room temperature. At 0°C they were much more stable. The analogous product from heparin was not so labile. Under the standard conditions, and at room temperature, relative fluorescence yields (D-glucosamine = 1.0) were: D-galactosamine, 0.75; D-gulosamine, 0.38; D-mannosamine, approx, 0.20. Neutral sugars, chondroitin sulphates, DNA and N-acetylneuraminic acids did not react, nor did N-acetylamino sugars or non-deaminated hexosamines. It is suggested that the Dische-Borenfreund indole method, the Kissane-Robins DNA assay and the proposed amino sugar method are all examples of simple aldehyde reactions. The specificity of the proposed method is considerably greater than that of the Dische-Borenfreund procedure, partly because of the much milder reaction conditions. The proposed method is very reproducible, about 50-100 times as sensitive as the Elson-Morgan reacton, and 10-50 times as sensitive as the Dische-Borenfreund procedures. It is also convenient; acid hydrolysates of amino sugar-containing compounds can be directly neutralized with sodium acetate solution.
