to the res­pec­tive xenobiotic (Martinoia et al. 1993). Hence, it may be concluded that the rate limiting step in xenobiotic removal is con­ju­gation, whereas the beneficial sequestration reactions proceed very fast and effectively.

The fact that many structurally diverse xenobiotics and natural compounds are transported via this GS-X pump into the vacuole leads to the point that glutathione conjugation may act as a „tag“. It may be a general signal for the sequestration of structurally differing but functionally similar molecules from the cytosol. According to the above mentioned three phase model of xenobiotic metabolism, the GS-X pump in the tonoplast would be a con­stituent of the „excretion“-part of the metabolism in plants. Some authors have claimed the vacuole as final storage pool for all kinds of molecules (Marrs 1996, Martinoia et al. 1993, Li et al. 1995). As such, they characterize phase III as „storage excretion“ in plants. However, this is not in accordance with the results of decades of research on xenobiotic metabolism in plants.

Moreover, work performed in onion indicates that GS-conjugates might be first transported to the nucleus before they get sequestered in the vacuole (see above).

On the background of these metabolism data, it is logical that glutathione conjugates may be intermediates rather than end products of detoxification. This topic has been reviewed extensively elsewhere (Schröder 1997). Elaborate investigations of herbicide metabo­lism in several plant cell cultures have elucidated that GS-conjugates have only short life­times in the cultured cells, and that they are rapidly further metabolized. First evidence for the enzymatic background of these cleaving reactions has been obtained only recently. Wolf et al. (1996) identified a specific carboxypeptidase for the cleavage of xenobiotic glutathione con­jugates in the vacuoles of barley. Carboxypeptidases are exopeptidases clea­ving terminal amino acids from polypeptides, whereas the physiological role of endopeptidases would be the regulation of enzyme activities via the cleavage of internal peptide bonds in poly­pep­ti­des (Zuber and Matile 1968). A natural function for exopeptidases has been found in the clea­vage of peptides in malt. Barley contains up to five distinct carboxypeptidases (Mikola 1983), three of which have been characterized recently (Breddam et al. 1983, 1985, Breddam and Sörensen 1987). However, none of these carboxypeptidases is identical to the enzyme puri­fied by Wolf et al. (1996). After 1080fold purification via cation exchange chro­mato­graphy the enzyme was shown to be a 56 kD monomer. This enzyme has activity for the cleavage of glycine from several xenobiotic glutathione conjugates (Table 2) with good affinity. The carboxypeptidase does not accept conjugates of ethacrynic acid nor S-(alachlor)-g-glutamylcysteine as a substrate, although the latter intermediate has been shown to be present in the vacuoles of [¹⁴C]-alachlor treated barley leaves (Schröder and Wolf unpublished).

Only few glutathione conjugates have been shown to accumulate in the plant vacuole for more than few hours. The sequestration as such, however, makes a lot of sense because it efficiently lowers the bioavailability of the conjugate. Assuming vacuolar sequestration was a general mechanism for intermediate storage, a model for the compartimentation and enzy­mology of glutathione conjugates would look like the one depicted in Figure 7.

Carboxypeptidase seems not to be responsible for the cleavage of g-glutamylcysteine residues. Consequently a dipeptidase or a g-glutamyltranspeptidase should be present in the vacuole. These enzymes have been described for the catabolism of glutathione but have never been assayed with xenobiotic dipeptideconjugates and their localization is not known.