Maize domestication phenotypes reveal strigolactone networks coordinating grain size evolution with kernel-bearing cupule architecture.

The maize (Zea mays) ear represents one of the most striking domestication phenotypes in any crop species, with the cob conferring an exceptional yield advantage over the ancestral form of teosinte. Remodeling of the grain-bearing surface required profound developmental changes. However, the underlying mechanisms remain unclear and can only be partly attributed to the known domestication gene Teosinte glume architecture 1 (Tga1). Here we show that a more complete conversion involves strigolactones (SLs), and that these are prominent players not only in the Tga1 phenotype but also other domestication features of the ear and kernel. Genetic combinations of a teosinte tga1 allele with three SL-related mutants progressively enhanced ancestral morphologies. The SL mutants, in addition to modulating the tga1 phenotype, also reshaped kernel-bearing pedicels and cupules in a teosinte-like manner. Genetic and molecular evidence are consistent with SL regulation of TGA1, including direct interaction of TGA1 with components of the SL-signaling system shown here to mediate TGA1 availability by sequestration. Roles of the SL network extend to enhancing maize seed size and, importantly, coordinating increased kernel growth with remodeling of protective maternal tissues. Collectively, our data show that SLs have central roles in releasing kernels from restrictive maternal encasement and coordinating other factors that increase kernel size, physical support, and their exposure on the grain-bearing surface.

Maize w3 disrupts homogentisate solanesyl transferase (ZmHst) and reveals a plastoquinone-9 independent path for phytoene desaturation and tocopherol accumulation in kernels.

Maize white seedling 3 (w3) has been used to study carotenoid deficiency for almost 100 years, although the molecular basis of the mutation has remained unknown. Here we show that the w3 phenotype is caused by disruption of the maize gene for homogentisate solanesyl transferase (HST), which catalyzes the first and committed step in plastoquinone-9 (PQ-9) biosynthesis in the plastid. The resulting PQ-9 deficiency prohibits photosynthetic electron transfer and eliminates PQ-9 as an oxidant in the enzymatic desaturation of phytoene during carotenoid synthesis. As a result, light-grown w3 seedlings are albino, deficient in colored carotenoids and accumulate high levels of phytoene. However, despite the absence of PQ-9 for phytoene desaturation, dark-grown w3 seedlings can produce abscisic acid (ABA) and homozygous w3 kernels accumulate sufficient carotenoids to generate ABA needed for seed maturation. The presence of ABA and low levels of carotenoids in w3 nulls indicates that phytoene desaturase is able to use an alternate oxidant cofactor, albeit less efficiently than PQ-9. The observation that tocopherols and tocotrienols are modestly affected in w3 embryos and unaffected in w3 endosperm indicates that, unlike leaves, grain tissues deficient in PQ-9 are not subject to severe photo-oxidative stress. In addition to identifying the molecular basis for the maize w3 mutant, we: (1) show that low levels of phytoene desaturation can occur in w3 seedlings in the absence of PQ-9; and (2) demonstrate that PQ-9 and carotenoids are not required for vitamin E accumulation.

Structure and Origin of the White Cap Locus and Its Role in Evolution of Grain Color in Maize.

Selection for yellow- and white-grain types has been central to postdomestication improvement of maize. While genetic control of carotenoid biosynthesis in endosperm is attributed primarily to the Yellow1 (Y1) phytoene synthase gene, less is known about the role of the dominant white endosperm factor White Cap (Wc). We show that the Wc locus contains multiple, tandem copies of a Carotenoid cleavage dioxygenase 1 (Ccd1) gene that encodes a carotenoid-degrading enzyme. A survey of 111 maize inbreds and landraces, together with 22 teosinte accessions, reveals that Wc is exclusive to maize, where it is prevalent in white-grain (y1) varieties. Moreover, Ccd1 copy number varies extensively among Wc alleles (from 1 to 23 copies), and confers a proportional range of Ccd1 expression in diverse organs. We propose that this dynamic source of quantitative variation in Ccd1 expression was created in maize shortly after domestication by a two-step, Tam3L transposon-mediated process. First, a chromosome segment containing Ccd1 and several nearby genes duplicated at a position 1.9 Mb proximal to the progenitor Ccd1r locus on chromosome 9. Second, a subsequent interaction of Tam3L transposons at the new locus created a 28-kb tandem duplication, setting up expansion of Ccd1 copy number by unequal crossing over. In this way, transposon-mediated variation in copy number at the Wc locus generated phenotypic variation that provided a foundation for breeding and selection of white-grain color in maize.