Studies using stably transformed tobacco plants containing very low levels of PI-PLC in their guard cells show that this enzyme plays a role in the events associated with the inhibition of stomatal opening by ABA, but not in the cellular reactions that are responsible for ABA-induced stomatal closure. However, Commelina communis guard cells microinjected with the InsP3 antagonist, heparin, fail to close on addition of ABA. There are three possible explanations for this apparent data mismatch. The differences may be indicative of species-specific signalling pathways, the presence of a PI-PLC isoform(s) that is not down-regulated in these transgenic lines and/or they may reflect differences between short-term (acute) administration of an inhibitor and long-term (chronic) effects of gene manipulation. It is possible that the guard cell is a robust signalling system that is able to adapt or compensate for the chronic loss of PI-PLC, but which is unable to adjust quickly to acute loss of this component. It would be interesting to investigate this possibility further using either transient manipulation of gene expression or through the use of an inducible promoter.