Novel Rht-1 dwarfing genes: tools for wheat breeding and dissecting the function of DELLA proteins

the number of seminal roots in maize co-map with the root developmental genes rtcs and rum1. Journal of Experimental Botany 67, 1149–1159. Salvi S, Tuberosa R. 2015. The crop QTLome comes of age. Current Opinion in Biotechnology 32, 179–185. SanMiguel P, Gaut BS, Tikhonov A, Nakajima Y, Bennetzen JL. 1998. The paleontology of intergene retrotransposons of maize. Nature Genetics 20, 43–45. Schnable JC. 2015. Genome evolution in maize: from genomes back to genes. Annual Review of Plant Biology 66, 329–343. Schnable JC, Freeling M. 2011. Genes identified by visible mutant phenotypes show increased bias toward one of two subgenomes of maize. PLOS ONE 6, e17855. Schnable JC, Springer NM, Freeling M. 2011. Differentiation of the maize subgenomes by genome dominance and both ancient and ongoing gene loss. Proceedings of the National Academy of Sciences, USA 108, 4069–4074. Schnable PS, Ware D, et al. 2009. The B73 maize genome: complexity, diversity, and dynamics. Science 326, 1112–1115. Singh V, van Oosterom EJ, Jordan DR, Messina CD, Cooper M, Hammer GL. 2010. Morphological and architectural development of root systems in sorghum and maize. Plant and Soil 333, 287–299. Tai H, Lu X, Opitz N, Marcon C, Paschold A, Lithio A, Nettleton D, Hochholdinger F. 2016. Transcriptomic and anatomical complexity of primary, seminal, and crown roots highlight root type-specific functional diversity in maize (Zea mays L.). Journal of Experimental Botany 67, 1123–1135. Tai H, Opitz N, Lithio A, Lu X, Nettleton D, Hochholdinger F. 2017. Non-syntenic genes drive RTCS-dependent regulation of the embryo transcriptome during formation of seminal root primordia in maize (Zea mays L.). Journal of Experimental Botany 68, 403–414. Wei F, Coe ED, Nelson W, et al. 2007. Physical and genetic structure of the maize genome reflects its complex evolutionary history. PLoS Genetics 3, e123. Xu X, Liu X, Ge S, et al. 2011. Resequencing 50 accessions of cultivated and wild rice yields markers for identifying agronomically important genes. Nature Biotechnology 30, 105–111. Zhang E, Yang Z, Wang Y, Hu Y, Song X, Xu C. 2013. Nucleotide polymorphisms and haplotype diversity of RTCS gene in China elite maize inbred lines. PLoS One 8, e56495. Zhu J, Kaeppler SM, Lynch JP. 2005. Topsoil foraging and phosphorus acquisition efficiency in maize (Zea mays). Functional Plant Biology 32, 749–762.

Reduced height-1 (Rht-1) dwarfing alleles have provided an essential breeding tool for increasing wheat grain yields and providing lodging resistance under high inputs. In this issue (pages 443-455), Van De Velde et al. demonstrate the potential of novel Rht-1 alleles for allowing more flexible control of stature and preharvest sprouting resistance in wheat breeding programmes. On a fundamental level, these alleles provide an important opportunity to uncover the signalling events that are responsible for improving these traits.
During the Green Revolution, the benefits of intensive agronomic practices to increase wheat grain yields could only be fully achieved when combined with varieties containing Reduced height (Rht) dwarfing genes (Hedden, 2003). The beneficial effects of these dwarfing alleles on grain yields are twofold: first, they prevent excessive stem elongation in response to high nitrogen fertilizer regimes that are prone to make the crop susceptible to environmental damage through lodging. Second, they allow a higher proportion of photosynthate to be partitioned to the grain by increasing the number of grains within the spikelets of the spike. The most widely utilized Rht dwarfing genes in worldwide wheat breeding programmes are those containing lesions at the Rht-1 locus. These include the Rht-B1b and Rht-D1b semi-dwarfing alleles which, at around the turn of the last century, were estimated to be present in over 70% of wheat cultivated worldwide (Evans, 1998). Remarkably, both of these Rht-1 homoeoalleles originated from the same Japanese variety, Norin-10, and were successfully exploited in US wheat breeding programmes in the 1950s (Hedden, 2003;Wilhelm et al., 2013).
Gibberellins (GAs) are plant hormones that promote many aspects of vegetative and reproductive development, including stem elongation and germination (Sponsel, 2016). The central regulators of the GA signalling pathway are nuclear-localized DELLA proteins (DELLAs) that act to repress GA-responsive growth through their physical association with transcription factors and other down-stream components (Thomas et al., 2016). Bioactive GAs relieve the DELLA growth repression by targeting their rapid degradation through a GID1-GA receptor-mediated signalling pathway (Nelson and Steber, 2016). The Rht-1 dwarfing alleles are known to reduce stem extension by causing partial insensitivity to GAs (Gale and Marshall, 1973). This altered response is caused by mutations in the homoeologous DELLA genes Rht-B1 and Rht-D1 (Peng et al., 1999;Pearce et al., 2011).
DELLAs are members of the GRAS family of transcriptional regulators (Thomas et al., 2016), containing two distinct domains: an N-terminal regulatory domain and a C-terminal functional GRAS domain (Box 1). The N-terminal domain is required for binding the GID1-GA receptor complex, a process which ultimately triggers DELLA degradation and promotes GA-responsive growth. The agronomically important Rht-B1b and Rht-D1b semi-dwarfing alleles contain mutations that introduce premature stop codons in the region of these genes encoding the N-terminal GID1-GA binding domain (Peng et al., 1999;Pearce et al., 2011;Box 1). It is believed that the effect of these mutations is to produce an N-terminally truncated protein which cannot be bound by the GID1-GA receptor, therefore resisting GA-mediated degradation and acting to constitutively repress GA-responsive growth and development. The severe dwarfing allele, Rht-B1c, also contains a lesion in the N-terminal coding region, which is predicted to have the same effect on perturbing GA signalling. However, in this case, the increased stability of RHT-B1C is due to a 30-amino acid insertion within the GID1-GA binding domain (Pearce et al., 2011;Wu et al., 2011). Although conclusive biochemical evidence is lacking, the current consensus of opinion regarding the milder GA-insensitive phenotype observed in Rht-B1b and Rht-D1b The upper panel shows an amino acid alignment of a region of the N-terminal regulatory domains of the wheat, rice and Arabidopsis DELLAs. The blue bars and asterisks below the alignment indicate conserved residues required for binding to the GID1-GA receptor (Murase et al., 2008), targeting them for degradation and relieving growth repression. The positions of mutations are shown in Rht-B1c (purple triangle indicating the site of a 30 amino acid insertion) and Rht-B1b (green triangle indicating the introduction of a premature stop codon; Q64*). Translational reinitiation (potential sites are indicated by a green bar) is thought to produce an N-terminally truncated RHT-B1B protein. The current model suggests that RHT-B1C and RHT-B1B constitutively repress GA-responsive growth because they do not bind to the GID1-GA complex and are therefore not degraded in response to the GA signal. The lower panel shows a schematic diagram illustrating conserved domains within RHT-B1 and the predicted effect of Rht-1 dwarfing mutations on the encoded proteins. The positions of amino acid substitutions caused by the ovg missense mutations are indicated within the GRAS domain (vertical blue lines) (Chandler and Harding, 2013;Van de Velde et al. 2017).
compared to Rht-B1c is due to a lower level of accumulation of the N-terminally truncated proteins produced by a process of translational reinitiation (Peng et al., 1999).

Novel Rht-1 dwarfing alleles
Despite the success of Rht-1 dwarfing alleles in controlling wheat stature and grain yields, the current range of genetic and phenotypic diversity is limited. An elegant screen performed by Chandler and Harding (2013) has greatly extended this diversity (Box 2). In this study they identified taller suppressor lines, designated overgrowth (ovg) alleles, following mutagenesis of a severely dwarfed line containing Rht-B1c. A total of 35 intragenic Rht-B1c suppressor alleles were identified, the majority of mutations resulting in amino acid substitutions within conserved regions of the functional GRAS domain (Chandler and Harding, 2013;Van De Velde et al., 2017;Box 1). As well as providing alleles that have the potential to allow breeders to precisely control stature over a wider range, the nature of these mutations provides us with the tools for understanding how DELLAs act to control diverse developmental processes. These studies demonstrate the exciting potential of these alleles on both of these levels.
To establish whether the ovg alleles represented useful alternatives to the classical dwarfing gene, Rht-B1b, Van De Velde et al. (2017) selected fourteen for further phenotypic characterization. Based on stem height these were classified as semidwarfing or tall depending on whether they were shorter or taller than Rht-B1b, respectively (Box 2). This allowed the selection of three alleles (two semi-dwarfing and one tall) whose agronomic potential was established following introgression into several elite spring wheat cultivars that have markedly different continental growing environments. Highly consistent effects on stem length and other architectural characteristics were observed in the presence of the ovg alleles. Importantly, no obvious detrimental effects on grain yields were observed under field conditions. The stability that they exert on these crop traits in different genetic backgrounds and under diverse environments clearly demonstrates their potential in wheat breeding programmes.

Rht-1 alleles with differential effects on GA-responsive traits
During the past decade there have been many studies demonstrating that DELLAs interact with a multitude of different classes of proteins to repress aspects of GA-responsive growth (Thomas et al., 2016). The regulatory mechanisms that these associations affect also differ depending on the interaction partner. Despite the large variety of DELLA partners and the diverse regulatory mechanisms involved, the majority of these studies have identified the GRAS domain as responsible for mediating these associations and exerting control. This is in agreement with the position of ovg mutations causing amino acid substitutions, which

Box 2. Wheat overgrowth mutants
The wheat overgrowth mutants (var. Maringa) were identified by mutagenizing the severe dwarf line Rht-B1c and screening for M2-suppressor lines that displayed increased leaf elongation or final height (Chandler and Harding, 2013). The screen resulted in the identification of many new Rht-B1 dwarfing alleles that can be used for more precise control of wheat stem height. Illustrating this, the image shows two ovg mutants (Rht-B1c.32 and Rht-B1c.3) compared to wild-type Rht-B1a (tall control), and the classical Rht-1 dwarfing lines Rht-B1b (semi dwarf) and Rht-B1c (extreme dwarf). Picture courtesy of Karel Van De Velde.
were identified solely within the region encoding the GRAS domain of RHT-B1C (Chandler and Harding, 2013;Van De Velde et al., 2017). Van De Velde et al. (2017) have used the recently resolved crystal structure of the rice GRAS protein OsSCL7 to model the effect of ovg mutations (Li et al., 2016). Many of these substitutions are suggested to occur within the interior of the DELLA protein, potentially affecting interacting residues (Van De Velde et al., 2017). It is tempting to speculate that the impact of these is to alter the 3D structure and perturb interactions with down-stream GA signalling components. To establish whether this is the case, it will first be necessary to identify RHT-1-interacting partners and then determine the impact of the ovg mutations on these interactions.
Based on our current understanding of how Rht-1 dwarfing genes cause GA insensitivity it might be reasonable to expect that the impact of a particular allele has an equivalent effect across different developmental processes controlled by GAs. However, there is evidence that this is not the case: a recent study of Rht-1 near-isogenic lines has indicated that those alleles containing premature stop codons, including the severe dwarf Rht-D1c, do not dramatically affect grain dormancy whereas Rht-B1c enhances it strongly (Gooding et al., 2012). The characterization of germination responses in the ovg mutants derived from Rht-B1c clearly confirms these differences between alleles, with the majority of them producing a level of dormancy that is consistent with their effect on stature (Van De Velde et al., 2017). In contrast, this correlation was not observed in the Rht-B1b semi-dwarf controls, in which dormancy was unaffected. This suggests an uncoupling of GA-regulated developmental responses that are controlled by the RHT-1 repressors lacking 70 amino acids of the N-terminus (Van De Velde et al., 2017). It is interesting to note that there is evidence supporting a functional role for the N-terminus of the rice DELLA SLR1 (Hirano et al., 2012). If the DELLA N-terminus does have a functional role, what is perhaps surprising is the absence of ovg missense mutations in this region. A plausible explanation for this could be the design of the original screen, which involved identifying enhanced leaf or stem elongation (Chandler and Harding, 2013). If feasible, a similar screen focusing on germination responses may deliver alternative intragenic suppressor alleles. Nevertheless, from the perspective of crop improvement, it is clear that the enhanced dormancy observed in the ovg mutants provides the potential to increase resistance to preharvest sprouting beyond that obtained with the currently deployed Rht-1 alleles.

Wheat: a new model plant
Attempts aimed at manipulating specific signalling pathways to improve traits in bread wheat have been severely hindered by its hexaploid nature and the lack of a reference genome sequence. The recent availability of a near-complete genome sequence now raises exciting possibilities for quickly identifying genetic elements underlying these traits (Borrill et al., 2015). An obvious problem with the hexaploid genome is that the presence of functionally redundant homoeologues impedes the identification of recessive lossof-function alleles, probably obscuring genetic diversity that can be exploited by wheat breeders. The recent development of robust reverse genetics platforms for functional genomics in wheat, including TILLInG populations (King et al., 2015) and CRISPR/Cas-based mutagenesis (Wang et al., 2014), raises the exciting prospect of unlocking this genetic diversity. An advantage of polyploid genomes is that mutagenesis events can be chemically induced at much higher frequencies than those achieved in diploid species, allowing the generation of increased genetic diversity (Parry et al., 2009). The unparalleled collection of DELLA mutations identified by Chandler and Harding (2013) clearly emphasizes these benefits. It also illustrates the power of novel screening strategies for generating improved genetic diversity that is unavailable with conventional breeding approaches.