An a-MAIZE-ing development

Maize was first domesticated by the indigenous peoples in southern Mexico about 10,000 years ago and is now a staple food in many parts of the world with a total production of about 1.2 billion tons. The wild-grass ancestor of modern maize is teosinte, which has tiny seed heads, in contrast to the ear of modern maize, the corn cob. The transition between these two plants involved the retraction of fruit cases, which allowed a stronger, thicker, and higher-yield cob. Although this is a remarkable change, the only gene identified to contribute to this has been the transcription factor Teosinte glume architecture 1 (Tga1) (Dorweiler et al., 1993). Tga1, however, only contributes to a fraction of the different ear phenotypes between teosinte and corn.

In a new publication, Jiahn-Chou Guan and colleagues (Guan et al., 2022) reveal that strigolactones (SLs) regulate many aspects of maize domestication phenotypes by modulating the accumulation of TGA1. Ten years ago, it was reported that the SL-deficient carotenoid cleavage dioxygenase 8 (ccd8) maize mutant had a smaller ear (Guan et al., 2012). Now, a closer look at ear features revealed that in the absence of SLs, glumes and rachids that form the protective cover around the teosinte kernel were enlarged—features that are very reminiscent of the tga1 phenotype. By creating a ccd8-specific revertant mutant, using exogenous SL, and disrupting the SL receptor genes DWARF14 (D14) -a and -b, the authors demonstrated that these phenotypes were caused by SL deficiency. To investigate the overlap between the TGA-attributed features and those caused by SL signaling, Guan and coworkers then created a ccd8/tga1 double mutant. They noticed that some of the features, such as the length of double-recessive ears, were intermediate between the two mutant phenotypes. Other features were complementary; for example, the cupules were longer in the double mutant than in either of the single mutants.

The authors performed RNA-Seq experiments comparing ear transcriptomes of wild-type, ccd8, tga1, and ccd8/tga1 plants. They found a 53% overlap in gene upregulation between the ccd8 and tga1 mutant samples, supporting the idea that TGA1 acts in the SL network. What are the interaction points between the SL signaling pathway and TGA1? The authors used transient gene expression in Nicotiana benthamiana to check for interaction of TGA1 with the SL transcriptional repressor DWARF53 (D53) and the SL receptor D14a. To assess the interaction, they made use of TGA1's toxic effect, causing necrosis at the infiltrated area. Interestingly, the necrotic effect was mitigated when TGA1 was expressed with both D14a and D53 but not when it was expressed with either one alone. Guan et al. used yeast-two-hybrid assays to confirm that TGA1 interacts with D14 and D53, providing strong evidence that TGA1 is a direct SL response factor. However, there are SL-regulated traits that are independent of TGA1. Can those be explained as a consequence of the SL system's sequestration capability? In teosinte, TGA1 is an activator of its own close paralog NOT1 that binds to the same SL signaling components. In this situation, high levels of both TGA1 and NOT1 would overwhelm the SL signaling system, in which TGA1 and NOT1 do not get entirely sequestered by D14 and D53. This would create a free and hence SL-independent fraction of TGA1 and NOT1, severely curtailing the plant's response to SL (see Figure 1). In modern maize, however, a single amino-acid change has turned TGA1 into a repressor of NOT1 (Wang et al., 2015). Here, the TGA1 system is SL-dependent, which has opened the door for breeding various SL-based traits that gave rise to today's corn cob.

Figure 1.

Interactions of TGA1 and NOT1 with the SL signaling components in modern maize and in teosinte. In maize, TGA1 represses NOT1, and the entire pool of TGA1 and NOT1 gets sequestered by the SL signaling components D14 and D53, making the system SL-dependent. In teosinte, TGA1 activates NOT1, leading to higher TGA1 and NOT1 protein levels, overwhelming the SL sequestration, and creating an SL-independent pool of TGA1 and NOT1. Adapted and modified with Biorender.com from Guan et al., 2022, Figure 7.

The work by Guan et al. is a major step in our understanding of maize domestication. The authors not only present a model of how SLs can mediate TGA1, but they also explain how this could have helped shape the vast phenotypic difference between modern maize and its wild-grass ancestor teosinte. Other phytohormones such as auxin and cytokinins are important in the regulation of kernel shape and size. Now the authors have identified SLs as a crucial player in maize ear morphology, adding to the growing number of plant developmental processes controlled by this group of hormones.