Feruloyl-polysaccharides can be oxidatively coupled in isolated cell walls by peroxidase plus exogenous H(2)O(2) in vitro, but the extent to which similar reactions may occur in the apoplast in vivo was unclear. Numerous cellular factors potentially control feruloyl coupling in vivo, and their net controlling influence is not readily studied in vitro. Therefore, we have monitored apoplastic feruloyl coupling in cultured maize cells in vivo using a radiolabelled model substrate, 5-O-feruloyl-alpha-L: -arabinofuranosyl-(1-->3)-beta-D: -xylopyranosyl-(1-->4)-D: -xylose (FAXX). FAXX was expected to permeate the wall and to undergo reactions analogous to those normally exhibited by apoplastic feruloyl-polysaccharides in vivo. Little difference was found between the fates of [feruloyl-(14)C]FAXX and [pentosyl-(3)H]FAXX, indicating negligible apoplastic hydrolase or transferase activities. Very little radioactivity entered the protoplasm. Maize cells that had recently been washed in fresh medium were able to bind most of the FAXX (90%) in their cell walls, regardless of the age of the culture. During wall-binding, the [(14)C]feruloyl groups were converted to [(14)C]dehydrodiferulates and larger coupling products, as revealed by TLC after alkaline hydrolysis. As expected for an oxidative reaction, wall-binding was delayed by added anti-oxidants (ascorbate, ferulate, sinapate, chlorogenate or rutin). It was also completely inhibited by iodide, an H(2)O(2)-scavenger, indicating a role for peroxidase rather than oxidase. The observations indicate that oxidative coupling of feruloyl groups occurred within the cell wall, dependent on endogenous apoplastic H(2)O(2) and wall-localised peroxidase, in vivo. Cells that had not recently been washed in fresh medium were much less able to bind FAXX, indicating the presence in the apoplast of an endogenous inhibitor of oxidative coupling. This inhibitor was of low M(r), was destroyed by heating, and remained in the aqueous phase (pH approximately 3.5) when shaken with ethyl acetate. Its effectiveness was not altered by ascorbate oxidase. It is thus a small, heat-labile, hydrophilic inhibitor (not ascorbate) which we suggest plays a natural role in the control of wall cross-linking, and thus potentially in the control of cell growth.