Mechanisms of self-incompatibility and unilateral incompatibility in diploid potato (Solanum tuberosum L.)

In chapter 1 an overview is given of the major mechanisms operating in Angiosperms that prevent or limit the degree of inbreeding. The two major systems that function on the basis of interaction between pollen and stigma/style, are the sporophytic and the gametophytic self-incompatibility systems (SSI and GSI). The plant is called the sporophyte and pollen and egg cells are called gametophtytes. In the sporophytic system, the pollen grains carry the information about the pollen donor in their coating. Thus, the pollen coating does not reflect the pollen genotype but deposits in it reflects the genotype of the pollen donor and the dominance relationships between the self-incompatibility alleles ( S -alleles). When the recipient has incompatibility characteristics in common with the pollen coating, the combination will be incompatible and pollen germination and pollen tube growth will be arrested on or in the stigma. In the Brassicaceae, a group displaying SSI, signal transduction seems to be an important mechanism for triggering an SI response.In the gametophytic self-incompatibility system, the pollen reflects the genotype of the pollen grain itself. When the incompatibility allele(s) of the pollen grain are met by a similar allele in the recipient, the pollen tube growth will be arrested. Thus, selfing provokes a gametophytic self-incompatibility (SI) response. Non-matching of S -alleles between plants of the same species results in a compatible combination. Most diploid Solanaceous species display GSI. The styles contain extracellularly the products of the style-expressed S -alleles, the S- glycoproteins. About the pollen components, contributing to SI, little is known, but S -heterozygosity in the pollen causes self-compatibility. The cultivated potato, Solanum tuberosum L . ( tbr ), is a tetraploid and behaves, due to S -heterozygosity in the pollen, as self-compatible species, whereas diploid potato generally possesses an active operating GSI system.There exist, however, also diploid species that are self-compatible (SC). Frequently, regardless of SSI and GSI, the SI species can be crossed with related SC species only when the latter are used as female parents. This means, the SC species can be used as the pollen acceptor (acceptance), but the SI species rejects the pollen of the SC species (non-acceptance). This phenomenon, in which interspecific hybridisation can occur in only one direction, is called Unilateral Incompatibility or Unilateral Incongruity (UI).In chapter 2 it is described how the basic plant material, used for SI research, was developed and selected. Vigour, abundant flowering and a good fertility were prerequisites for this material, but the most important characteristic was a reliable SI reaction in pollen and style. The combination of these characteristics is rarely found in diploid tbr . From a diploid tbr population, expressing four different S -alleles, plants could be selected for all six S-heterozygosity classes, that met all the afore mentioned criteria. S -allele composition could be tested by performing test crosses, but in addition to this, stylar extracts were analysed by iso-electric focusing, followed by silver staining. The S -glycoproteins, also called S-RNases because of their RNase properties, focus in the basic part of the gels. The selected material was used for the creation and selection of SI plants that were homozygous for the S -alleles. Normally, the SI system will prevent S -homozygotisation, unless the SI system is weakened by pollen- or style expressed minor or major SI-suppressor genes. A weakening of the SI response can cause seed set after selfing. This is called pseudo-compatibility (PC).Occasionally, however, some pollen tubes manage to penetrate the style , even when the SI system is fully functional and PC can be excluded. The seed set will then, however be too low to establish a sink-source relationship that is strong enough to cause berry formation: the flowers will drop. The S.phureja ( phu ) clones IvP35 and IvP48 are normally compatible with diploid tbr , but the hybrid seed has the remarkable and useful characteristic, that the embryo's have a nodal band, which is visible through the seed coat as a seed spot at the first node between hypocotyl and the cotyledons. Pollination with those phu clones after making crosses that were incompatible, caused berry formation. This additional pollination is called "counterfeit pollination". Spotless seed, harvested from those berries, yielded both S -heterozygotes and S -homozygotes. Analysis on the seed set and the strength of the SI reaction in this offspring showed that, even when the original parents were selected for their good SI reaction, weakened SI was present, that could be expressed in either the pollen or the style. It was shown that this had a heritable character. From this material, S -homozygotes could be selected that were reliable in their SI reaction and that served as tester clones, as described in the chapters 3, 4 and 5.The selected material, described in chapter 2, was poor in its transformation efficiency. For the functional analysis of, for instance, S -allele based constructs, an efficient transformation system is essential. It was decided, therefore, to select for this trait. Transformation efficiency was introduced from other unrelated sources. Well transformable clones with a reliable stylar SI expression could be selected from this material (Appendix 2), that were used for a gain-and loss of function approach. Sense (chapter 3) and antisense S-RNase constructs (chapters 3 and 5) were introduced by genetic transformation. Indeed, sense S2 transgene constructs, driven by the promoter of the style-specific endochitinase SK2 , were able to cause an incompatibility reaction against S2 pollen in plants that did not contain the S2 allele when not transformed. Some of those constructs showed such a high level of expression, that due to some mechanism, the endogenous S -alleles were down-regulated and became compatible for the endogenous S- alleles, whilst remaining incompatible for the transgene S- allele. The antisense S1-RNase and S2-RNase constructs were able to reduce the expression of the corresponding S1 and S2 -alleles, which resulted in a break-down of the incompatibility reaction against the corresponding S1 and S2 -pollen. Thus, the gain and loss of function approach showed the key role of the S-RNases in the stylar side of the self-incompatibility reaction.In chapter 4 it is described why ver is self-compatible and how this is expressed in hybrid offspring, when crossed with self-incompatible tbr . When the former species is used as recipient, the hybrids suffer from cytoplasmic male sterility, thus disabling a further analysis of inheritance and expression of SI, SC and UI. The reciprocal cross fails normally, as already stated, due to UI. However, some of the potato clones, described in chapter 2, accepted ver -pollen and yielded male and female fertile hybrid offspring. Those particular potato clones are called "acceptors"for ver pollen, as an exception to the rule of UI. Plants that show UI are thus called "non-acceptors". It was shown that the species ver can be SC due to at least two different reasons: 1) there is no stylar S-glycoprotein and a stylar SI response is therefore disabled, and 2) there is a pollen-expressed self-compatibilizing factor, SCver . This SCver -factor was linked with the S -locus of ver , at an estimated distance of 18 cM. SCver is also capable of suppressing the SI reaction against pollen-expressed tbrS -alleles. This suppression depends, however, on the genotype of the pollen recipient. Acceptors allow for the penetration of SCver carrying pollen, but specific non-acceptors can inhibit this type of pollen. It was shown that there exist differential reactions against ver pollen, and in particular, also against the SCver factor. Experiments with somatically doubled hybrids showed that where the stylar part of the S- locus of ver is inactive, that the pollen part of the S- complex is not only capable of triggering a UI reaction, but also in causing the SI-based phenomenon of mutual weakening. Mutual weakening is the phenomenon that when two different S- alleles are expressed in a pollen grain, the GSI reaction in the style will not take place anymore, even when those S -alleles are expressed in the style as well. Thus, a dual function of the pollen part of the S -locus is made likely.In chapter 5 the gain and loss of function approach, as described in chapter 3, was used to test whether the stylar part of the S -locus is involved in the UI response too. The sense approach failed, due to the absence of transgenic ver regenerants, but the loss-of-function was successful. Both a transgenic non-acceptor tbr x ver hybrid and a transgenic non-acceptor tbr clone, both expressing only the S2 allele in the style, showed a collapse of the UI reaction that coincided with the antisense S2 caused break-down of the SI response against S2 pollen. The S -locus complex shows thus a dual function for both the pollen part and the stylar part, both contributing to the SI and the UI response. It was made likely that ver can have a putative non-acceptor background for self pollen, but that its expression requires S-glycoproteins to be expressed. In this chapter it is discussed why the most important hypotheses about UI are not necessarily conflicting with each other. An explaining and predictive model with interactions of a range of genes and alleles is presented. The most important genes and their properties are:the acceptance gene A, which causes acceptance ( aa genotypes being non-acceptors), but knows different alleles that cause differential reactions against ver pollen;the inhibitor gene I, which causes non-acceptance and is epistatic over A ;the pollen-expressed SC factor SCver , which is in weak linkage with the S- locus, causes pollen to be compatible in any style, except those with the genotypes aaII and aaIi, in which it is inactive or even causes a UI reaction;the S -locus complex with both a pollen component and a style component, in which the pollen ver factor triggers a UI response, and the active stylar part is needed for a UI reaction in aaii non-acceptor genotypes. The latter explains why the introduction of an active S -allele in a SC species (such as reported by Murfett et al., 1996) can bring about a sudden SI or UI response and why ver can be a putative non-acceptor for self pollen, without becoming self-incompatible.As a consequence of this, the expression "Unilateral Incompatibility" cannot completely be replaced by the expression "Unilateral Incongruity". The latter expression is valid in cases where the S -locus does not contribute to the UI response at all.In the final chapter some of the results are discussed in a broader context. The last part stresses that the dual function of the S -locus implicates that, within the existing model of S-RNase activity in the SI system, a second function of the S -glycoproteins must be postulated. This can be triggering a signal transduction, resulting in a SI like response, resulting in the arrest of the pollen tube, but which may be independent of the RNase properties of the S -glycoproteins.