Darwin studied the effects of cross- and self-fertilization in 57 species of flowering plants from 52 genera and 30 families
[9]. He conducted experiments over 11 years, involving up to 10 successive generations of self-pollination and the growing of offspring in pots in his greenhouse or in rows in his garden. His most important conclusion was that cross-fertilization is generally beneficial and self-fertilization (inbreeding) injurious, and hence the reason for the existence of mechanisms to promote cross-pollination. He did, however, find that some species were tolerant to inbreeding, although crosses between cultivars displayed hybrid vigor. One such species was the garden pea (
Pisum sativum), chosen by Mendel for his experiments in plant hybridization that led to his laws of inheritance
[10]. Mendel’s work was unknown to Darwin, and remained unknown to the world until rediscovered in 1900, translated into English in 1901, and promoted by Bateson
[11]. In retrospect, the work of Mendel on the mechanism of inheritance, and of Darwin on the mating system, can be viewed as the foundations of scientific plant breeding from the beginning of the 20th century. The world’s four most important food crops have three different mating systems. Wheat (
Triticum aestivum) and rice (
Oryza sativa) are inbreeding, seed-propagated crops; maize (
Zea mays) is an outbreeding, seed-propagated crop; and the potato (
S. tuberosum) is an outbreeding, but primarily clonally propagated, tuber crop. It can be argued that, as a consequence, maize breeding has benefited the most, and potato breeding the least, from underpinning genetics research
[12]. Darwin was particularly interested in the evolutionary advantages and disadvantages of the variation found in breeding and mating systems, and how one system could evolve into another, such as the shift from crossbreeding to inbreeding
[9][13]. Fisher combined the work of Darwin and Mendel into a theory of inbreeding, with the important addition of an explanation for disadvantageous genes (alleles) tending to be recessive
[14]. Furthermore, he advocated greater use of inbreeding for the practical improvement of domestic plants and animals, given the resounding success of the methods adopted for maize improvement based on inbreeding. Indeed, one of the major achievements in 20th century plant breeding was the production of high yielding, genetically uniform F
1 hybrid cultivars of outbreeding crops from genetically variable, open-pollinated ones. Thus, let researchers return to the breeding of diploid potatoes, starting with diploid potato germplasm.
Since the 1960s, it has also been possible to reduce tetraploid
S. tuberosum to the diploid level through crosses with 2
x Phureja Group ‘pollinators’ that induce dihaploid production by a form of parthenogenesis
[23][24]. Subsequently the frequency of dihaploids (also referred to as haploids) has been increased by the use of elite pollinators, such as IVP35, IVP48, IVP101, and PL-4
[25][26], although the frequency is still low, with reported values of 4.6 to 25.6 dihaploids per 100 berries
[27]. Busse et al.
[27] have described a high throughput method of production based on the original method of Peloquin and Hougas. However, dihaploids are usually male-sterile, unlike two of the first four, US-W1 and US-W4
[28][29], and female fertility can also be a problem in breeding
[27]. Most dihaploids, like the cultivated diploid groups, display gametophytic self-incompatibility
[30][31], which prevents the production of true breeding lines by self-pollination. In other words, diploid potatoes are outbreeders, and the same is true for most of their diploid wild relatives. The self-incompatibility (
S) locus was mapped to chromosome 1 by Gebhardt et al.
[32] and Kaufmann et al.
[33].
Thus, in the selected genotype, as in the parents, deleterious recessive alleles are accommodated without adverse effects through heterozygosity.
Despite being homozygous for some deleterious alleles, the inbred parents must have sufficient vigor and fertility for the maintenance and production of their hybrid.
5. Successful Self-Pollination
Successful self-pollination of diploid potatoes was first made possible by the use of the dominant self-incompatibility inhibitor gene (
Sli) found in
S. chacoense by Hanneman
[35], and mapped to chromosome 12, thus showing that it is independent of the
S locus on chromosome 1
[36][37]. Jansky et al.
[38] explained how inbred clone M6 (homozygous for
Sli) was produced from self-compatible (SC)
S. chacoense by seven generations of self-pollination, and found to be vigorous and fertile, as well as morphologically indistinguishable from
S. chacoense plants that had not been inbred. It possesses a number of desirable agronomic traits, processing quality and resistance to
Pectobacterium carotovorum soft rot and
Verticillium dahlia wilt, as well as undesirable wild species traits, including unacceptably high levels of glycoalkaloids. Interestingly, it produced (small) tubers in a 14 h photoperiod, unlike many other wild species. Genomic information on M6 was provided when Leisner et al.
[39] sequenced and assembled its genome. They anchored 508 Mb (million base pairs), out of the estimated 882 Mb from flow cytometry, into 12 chromosomes (pseudomolecules), and found that their genome annotation represented 37,740 genes. Analysis of single nucleotide polymorphisms (SNPs) across the whole M6 genome revealed 1,414,890 biallelic SNPs from a total of 208 Mb of assayable nucleotides, an SNP frequency of 0.68%, compared with 4.8% loci heterozygous, out of 8303 represented in the SolCAP SNP array
[40]. There was enhanced residual heterozygosity on chromosomes 4 (1.73%), 8 (2.37%), and 9 (2.10%) compared with the overall figure of 0.68%. Endelman and Jansky
[41] crossed M6 as the male parent with the male-sterile doubled monoploid
S. tuberosum Phureja Group clone DM1-3 516 R44 (DM1-3), which was used to provide the first published sequence of the potato genome
[42]. They self-pollinated the F
1 to produce an F
2 population for quantitative trait locus (QTL) analysis. Kaiser et al.
[43] explored self-fertility and resistance to the Colorado potato beetle (
Leptinotarsa decemlineata) in a diploid
S. chacoense recombinant inbred line population derived from 308 F
2 individuals after crossing M6 (self-compatible) and resistant accession USDA8380-1 (80-1, self-incompatible). Fifty-five F
5 families were analyzed, and all individuals contained at least one copy of the
Sli gene; however, this was not sufficient for selfed fruit and seed production. Loci on chromosomes 3, 5, 6, and 12 were identified as novel targets for self-fertility improvement, and a major QTL for foliar leptine glycoalkaloid biosynthesis and Colorado potato beetle resistance was mapped to chromosome 2. Self-compatibility and resistance to Colorado potato beetle were introgressed into diploid breeding material with desirable tuber traits. Thus, germplasm useful to practical breeders is emerging from such genetics research.
From 2008, Lindhout et al.
[5][44] used the
Sli gene for the production of diploid inbred lines that could be used to produce true F
1 hybrid cultivars for maximum heterosis and genetic uniformity. They started by producing inter-species hybrids between diploid potato germplasm and a homozygous accession of the wild species
S. chacoense carrying the
Sli gene. The hybrids were extremely vigorous, and about half of them produced many berries upon self-pollination.
Clot et al.
[45] found that the
Sli allele (SC haplotype) is in fact widespread in the cultivated gene pool of the potato plant, including the tetraploid cultivars Garnet Chili, Irish Cobbler, Early Rose, Kennebec, Russet Norkotah, Sierra Gold, Yukon Gold, Snowden, Atlantic and Mountain Rose; with Garnet Chili, Irish Cobbler, and Early Rose probably, and Russet Norkotah definitely, possessing two copies of the
Sli allele. They concluded that cultivar Rough Purple Chili, introduced into the USA in 1851 by Goodrich
[46], is the origin of the SC haplotype. Interestingly, Haynes and Guedes
[21] found that, in their long-day-adapted diploid population of Phureja/Stenotomum Groups, out of 42 clones evaluated, 32 flowered, and of these, 20 were successfully self-pollinated. Clot et al.
[45] also developed KASP (Kompetitive allele-specific PCR) markers that can be used in breeding programs for the marker-assisted selection of self-compatibility (SC). Using a subset of six of these
Sli KASP™ markers, Kaiser et al.
[47] assessed the contribution of
Sli to SC in the Michigan State University diploid germplasm, which represents diverse clones derived from multiple North American breeding programs. Although the
Sli markers predicted SC in some germplasm, there were discrepancies, which emphasized the need to identify other genomic regions critical to SC and the role of the environment in the expression of genes involved in the SC reaction. For example, although M6 was confirmed as being homozygous for the SC
Sli genotype at all six loci, the self-compatible clones 1S1 and DMRH-89, and a plant of
S. chacoense accession PI 133664–40, were homozygous for the SI
Sli genotype at the six loci.
Ma et al.
[48] have confirmed that the self-compatible, heterozygous, diploid potato clone RH89-039-16 (RH) contains the
Sli gene, and that it is capable of interacting with multiple allelic variants of the pistil-specific S-ribonucleases (S-RNases), and thus functions as a general S-RNase inhibitor to impart self-compatibility to RH and other self-incompatible potatoes. RH has a pedigree derived from dihaploidized tetraploid commercial cultivars.
Ye et al.
[49] created self-compatible diploid potatoes through knock out mutations (loss-of-function) in the self-incompatibility
S-RNase gene (at
S locus) of the diploid self-incompatible
S. tuberosum Phureja Group clone S15-65, using the CRISPR–Cas9 system. The growth vigor and plant morphology of the mutant lines did not differ from those of clone S15-65, indicating that they could be used directly for breeding. The percentage of plants in the selfed families without the Cas9 cassette varied from 3.6% to 24.5%, thus demonstrating that self-compatibility can be achieved without introducing any exogenous DNA. Using the same method, the researchers also obtained
S-RNase mutants in two more
S. tuberosum Phureja Group clones (S15-47 and S15-76) and two
S. tuberosum Stenotomum Group clones (S15-48 and S15-107). Hence, improved diploid clones from a population improvement scheme could be made self-fertile for the purpose of inbreeding and crossbreeding to produce F
1 hybrid cultivars. Enciso-Rodriguez et al.
[50] also used the CRISPR–Cas9 system to generate targeted knockouts in conserved coding regions of the
S-RNase gene. They achieved nine knockout lines (deletions/insertions) which transmitted self-compatibility to their progeny.
Self-Pollination in Other Outbreeding Crops
Diploid hybrid breeding occurred much earlier in other outbreeding crops. The six examples given in my book on plant breeding
[51] fell into three groups: (1) no incompatibility system to prevent self-pollination, (2) a sporophytic system, and (3) a gametophytic system.
Maize (
Zea mays) naturally reproduces by wind cross-pollination, primarily, but not exclusively, as a result of having separate male and female flowers on the same plant (monoecious). However, it is also easy to self-pollinate, and hence proved suitable for both genetics and plant breeding research. The breeding research on the effects of inbreeding and crossbreeding in maize by East
[52], Shull
[53][54], and Jones
[55] resulted in double-cross (DC) hybrids being grown in the USA from the 1930s, and single-cross (SC = F
1) hybrids from the 1960s
[56]. Onions (
Allium cepa) are predominantly a cross-fertilizing, insect-pollinated, diploid species, although self-pollination occurs when the inflorescences of breeding material are simply bagged. F
1 hybrids started to appear in the USA from the early 1950s, and in the Netherlands from the late 1960s
[57], but are still to be developed in India
[58]. Carrots (
Daucus carota) are an outcrossing, insect-pollinated, diploid species with hermaphrodite flowers that are usually protandrous, although self-pollination is not restricted by incompatibility. Carrot hybrids started to become available in the 1970s and, since the 1980s, have been replacing open-pollinated cultivars worldwide
[59].
Since the 1960s, F
1 hybrid production has been the driving force behind the breeding of vegetable Brassicas (
Brassica oleracea), such as Brussels sprouts, cabbage, calabrese/broccoli, and cauliflower. Self-pollination is normally prevented by a sporophytic self-incompatibility system, as was first demonstrated and explored by Thompson
[60] in the context of fodder kale (
B. oleracea var.
acephala) breeding, which started at the former Plant Breeding Institute in Cambridge in 1951
[61]. Inbred lines can be produced and maintained by pollinating a bud with pollen from another flower on the same plant at least two days before the bud opens; a key difference from a gametophytic self-incompatibility system.
Rye (
Secale cereale) is a diploid, cross-pollinated cereal with an effective gametophytic self-incompatibility system
[62]. Dominant self-fertility genes were detected in various European germplasm
[63][64][65] and transferred into breeding pools by repeated backcrossing. In Germany, hybrid breeding started around 1970
[66], and hybrid (single-cross × restorer synthetic composed of two inbred lines) cultivars predominated in most Western and Central European rye-growing areas by 2009
[67]. Sugar beet (
Beta vulgaris) is another outbreeding crop in which self-fertilization is normally prevented by a gametophytic self-incompatibility system. Again, self-fertility was provided by a dominant gene (
Sf)
[68]. In recent years, diploid sugar beet hybrids have become prevalent in Europe and elsewhere, most of which are three-way cross hybrids (2
x CMS
mm F
1 × 2
x N MM), where CMS is cytoplasmic-genetic male sterility, N is normal cytoplasm, and
mm is the genotype required for the monogerm seed
[69].
In conclusion, in the three crops with a gametophytic self-incompatibility system, potato, rye, and sugar beet, reliable and successful self-pollination required the discovery and use of a dominant self-fertility gene.