Coumarins give misleading absorbance with Ellman's reagent suggestive of thiol conjugates

In the course of a screening for phytochelatins in cadmium-exposed bryophytes in the terrestrial mosses Polytrichum formosum and Atrichum undulatum we detected compounds with absorption properties and retention times similar to phytochelatins when applying the commonly used standard method RP-HPLC and post-column derivatization with thiol-specific DTNB (Ellman) reagent. Moreover, as with phytochelatins known in other plants, the concentrations of these compounds increased slightly after Cd stress. The concentration of the precursor glutathione (gamma-ECG), however, increased in the presence of Cd. In order to verify the identity of these putative phytochelatins we performed LC-ESI-MS analyses as well as H-1 NMR on extracts from P. formosum and A. undulatum. Spectroscopic investigations indicated that the detected compounds were neither phytochelatins nor other thiol compounds. From the results of HPLC-H-1 NMR and mass spectrometry we concluded that at least one of these substances was a coumarin, probably a 5,8-dihydroxy-7-methoxycoumarin-5-beta-glucopyranoside, which has already been described for A. undulatum and P. formosum. The results of our investigations prove that under the basic pH conditions essential for the Ellman test for thiol compounds, coumarins show comparable UV/VIS absorption properties. Therefore, a positive post-column Ellman reaction cannot unambiguously prove the presence of thiol-containing compounds in plants.