It is seven years since the first reports of cDNAs encoding pistil glycoproteins that segregated with particular S-alleles. During this time, the S-glycoproteins of the Solanaceae have been identified as RNases. This enzymatic activity relies on the presence of histidine residues at the putative active site of the RNase, and these are conserved in all S-glycoproteins so far characterized. The proteins also contain "hypervariable" regions that may have some role in allelic specificity. It is particularly interesting that putative S-glycoproteins from Japanese pear, which is from a different family, the Rosaceae, are also RNases. To counter the temptation to extrapolate to other families with gametophytic self-incompatibility, there is evidence that in another family, the Papaveraceae, poppy S-glycoproteins are not RNases. The current evidence is consistent with a process in which the S-RNase moves into the incompatible pollen tube and degrades RNA, including rRNA. As rRNA genes are not transcribed in pollen, the resulting degradation would lead to the death of the cell. But still we are left with some important gaps in our knowledge. How does the S-RNase move across the wall and membrane and into the pollen tube? How is the specificity of the interaction controlled? What is the mechanism of signal transduction? A major bottleneck in unraveling the story is understanding the nature of the S-locus product in pollen. Is it related to the stylar S-locus product or is it the product of a different gene in the same locus? Each question underlines the sketchiness of our knowledge of many plant processes that are not specific to pollination, but that we need to understand if we are to work out the details of self-incompatibility. For example, we have a very incomplete understanding of cell wall synthesis generally and pollen wall synthesis in particular. How do macronutrients move through cell walls to the cytoplasm of cells generally and pollen tubes in particular? What is the nature of the receptor-ligand interaction in plant cells generally and pollen tubes in particular? A similar range of questions and gaps in our knowledge exist in the sporophytic system, exemplified by studies in Brassica spp. In this case, we have no known enzymatic or other function for the stigmatic S-glycoproteins. We do, however, know that the S-locus in Brassica includes at least two genes, one encoding a S-glycoprotein and the other encoding a protein kinase.(ABSTRACT TRUNCATED AT 400 WORDS)