Feature Essay 7.1 Self and non-self: recognition processes in flowering plants

Printer-friendly version

R.B. Knox


Figure 1 Professor Bruce Knox, Professor of Botany, University of Melbourne

In 1969, while at the Australian National University in Canberra, I spent a sabbatical period at the Institute of Plant Development, University of Wisconsin, Madison, USA, with Dr Jack Heslop-Harrison and discovered that pollen grain walls are loaded with a range of extracellular hydrolytic enzymes. Certain enzymes occur in cavities of the patterned exine wall, while others are incorporated into the smooth inner intine wall layer, which forms the surface at the germinal apertures. When pollen is moistened for germination, these proteins diffuse rapidly from their wall sites into the sur-rounding medium. The outflow of pollen proteins is considered part of a general ‘dialogue’ between pollen and stigma. The question arose as to the possible function of these proteins. Could they play a role in pollen transfer to the stigma by wind or water currents or by animal pollinators? Could they have a defence function, preventing attack by microorganisms? Could they act as recognition molecules permitting pollen germination and tube growth on the stigma surface? Or could they be involved in degrading the stigma surface to permit pollen tube penetration?

These questions were finally answered by a series of experiments done at the Australian National University in the early 1970s. Our first experiments were carried out on poplar trees as part of a tree breeding program developed by Professor Lindsay Pryor and involved transfer of desired characteristics from one species (white poplar) to another (black poplar) by cross-pollination. Unfortunately, the cross between the two species did not set seed. So we used the ‘mentor’ pollen technique developed by Dr Reinhold Stettler at the University of Washington, Seattle, USA, for other species, to see if we could obtain hybrid seed. This method is based on the work of earlier plant breeders, such as Michurin in Soviet Russia, who had successfully obtained hybrids from crosses between species by mixing the species’ own pollen which sets seed readily (black poplar) with the pollen of the other species (white poplar) and applying it to stigmas. The problem with this method is that large numbers of the progeny would be selfs, with only a few hybrid seeds resulting. Stettler showed that mentor pollen killed by high doses of gamma radiation could be used, so that the resulting progeny were all hybrid. When this was tested on the black X white poplar system, we obtained many hundreds of hybrids, which formed the basis for a breeding program and which has been highly successful in providing fast-growing but high-quality timber in subtropical regions of Australia. The hybrids also provided the answer to our question, because we were able to intervene in the pollination process and determine if pollen proteins could replace mentor pollen in the interaction and successfully generate hybrid seeds.

We were able to show that diffusible molecules from mentor pollen (black poplar) would enable white poplar pollen to set hybrid seed on black poplar stigmas. These molecules included a range of proteins and glycoproteins which were obtained by extracting living pollen grains for 5 min. Extract was painted on the stigma, followed by dusting with the white poplar pollen. Although seed set was much lower than after self-pollination, all seeds were again hybrids. We checked the specificity of the response, including self diffusate (black poplar) followed by self-pollination (black poplar, which gave expected high seed set) and white poplar diffusate with white poplar pollen (no seed set). Our inter-pretation is that the mentor diffusate altered the recognition response of the stigma to white poplar pollen, so that successful seed setting occurred.

In a large genus such as Populus, the existence in the breeding system of barriers to reproduction between species is known as interspecific incompatibility. Pollen from the incompatible species (white poplar) is perfectly viable but is unable to set seed on the other species (black poplar). All intraspecific pollinations are compatible, leading to high rates of seed set. However, many families of flowering plants also show intraspecific self-incompatibility, in which pollen from an individual plant, even though perfectly viable, is unable to set seed on its own stigmas, but can set seed on the stigmas of most other individuals of the species (see Section 7.2.4). It seemed worthwhile to carry out similar pollinations with a plant showing such an intraspecific incompatibility system, with the goal of increasing numbers of self seeds.

Dr Barbara Howlett and I extended the experiments to self-recognition in the daisy Cosmos bipinnatus (family Asteraceae). In this case, pollen is inhibited on the stigma surface following self-pollination. This is a rapid process, taking just 40 min for the entire fertilisation events from pollen touchdown to gamete fusion. The mentor technique gave good results: gamma-irradiated mentor pollen mixed with viable self pollen gave 27% of the seed set expected from a compatible mating. Self-matings had a low rate of seed set, with gamma-irradiated self pollen mixed with viable self pollen giving seed sets of up to 4%. When aqueous extracts of compatible mentor pollen followed by viable self pollen were applied, seed sets of 12–15% were achieved. In these experiments, the pollen diffusate had been partially purified, so that the recognition responses are more likely to be caused by proteins or glycoproteins, but the key elicitor protein remained to be identified.

Today, cloning of genes encoding diffusible proteins from allergenic types of pollen has provided some clues. Many proteins have proved to be enzymes associated with the degradation or synthesis of plant cell walls, others are expressed during a period of stress such as pathogen attack. In ragweed, a close relative of Cosmos, small proteins, each made up of only 45 amino acids, have a region with eight cysteine residues, which form four pairs of disulphide bonds and give these proteins a series of distinctive loops, making them ideal for performing roles in recognition and specificity. These proteins resemble toxins from snake venom and recognition factors in fungi, which can trigger the host defence response and so limit their plant range. This tells us that, at the pollen grain surface, there is a range of molecules that possess defence or recognition capacities, making them the likely arbiters of recognition and specificity in pollen–stigma interactions. The future may reveal more of the relationships between recognition of spores and pollen.