A conserved genetic architecture among populations of the maize progenitor, teosinte, was radically altered by domestication.

Very little is known about how domestication was constrained by the quantitative genetic architecture of crop progenitors and how quantitative genetic architecture was altered by domestication. Yang et al. [C. J. Yang et al., Proc. Natl. Acad. Sci. U.S.A. 116, 5643-5652 (2019)] drew multiple conclusions about how genetic architecture influenced and was altered by maize domestication based on one sympatric pair of teosinte and maize populations. To test the generality of their conclusions, we assayed the structure of genetic variances, genetic correlations among traits, strength of selection during domestication, and diversity in genetic architecture within teosinte and maize. Our results confirm that additive genetic variance is decreased, while dominance genetic variance is increased, during maize domestication. The genetic correlations are moderately conserved among traits between teosinte and maize, while the genetic variance-covariance matrices (G-matrices) of teosinte and maize are quite different, primarily due to changes in the submatrix for reproductive traits. The inferred long-term selection intensities during domestication were weak, and the neutral hypothesis was rejected for reproductive and environmental response traits, suggesting that they were targets of selection during domestication. The G-matrix of teosinte imposed considerable constraint on selection during the early domestication process, and constraint increased further along the domestication trajectory. Finally, we assayed variation among populations and observed that genetic architecture is generally conserved among populations within teosinte and maize but is radically different between teosinte and maize. While selection drove changes in essentially all traits between teosinte and maize, selection explains little of the difference in domestication traits among populations within teosinte or maize.

Domestication reshaped the genetic basis of inbreeding depression in a maize landrace compared to its wild relative, teosinte.

Inbreeding depression is the reduction in fitness and vigor resulting from mating of close relatives observed in many plant and animal species. The extent to which the genetic load of mutations contributing to inbreeding depression is due to large-effect mutations versus variants with very small individual effects is unknown and may be affected by population history. We compared the effects of outcrossing and self-fertilization on 18 traits in a landrace population of maize, which underwent a population bottleneck during domestication, and a neighboring population of its wild relative teosinte. Inbreeding depression was greater in maize than teosinte for 15 of 18 traits, congruent with the greater segregating genetic load in the maize population that we predicted from sequence data. Parental breeding values were highly consistent between outcross and selfed offspring, indicating that additive effects determine most of the genetic value even in the presence of strong inbreeding depression. We developed a novel linkage scan to identify quantitative trait loci (QTL) representing large-effect rare variants carried by only a single parent, which were more important in teosinte than maize. Teosinte also carried more putative juvenile-acting lethal variants identified by segregation distortion. These results suggest a mixture of mostly polygenic, small-effect partially recessive effects in linkage disequilibrium underlying inbreeding depression, with an additional contribution from rare larger-effect variants that was more important in teosinte but depleted in maize following the domestication bottleneck. Purging associated with the maize domestication bottleneck may have selected against some large effect variants, but polygenic load is harder to purge and overall segregating mutational burden increased in maize compared to teosinte.

The genetic architecture of the maize progenitor, teosinte, and how it was altered during maize domestication.

The genetics of domestication has been extensively studied ever since the rediscovery of Mendel's law of inheritance and much has been learned about the genetic control of trait differences between crops and their ancestors. Here, we ask how domestication has altered genetic architecture by comparing the genetic architecture of 18 domestication traits in maize and its ancestor teosinte using matched populations. We observed a strongly reduced number of QTL for domestication traits in maize relative to teosinte, which is consistent with the previously reported depletion of additive variance by selection during domestication. We also observed more dominance in maize than teosinte, likely a consequence of selective removal of additive variants. We observed that large effect QTL have low minor allele frequency (MAF) in both maize and teosinte. Regions of the genome that are strongly differentiated between teosinte and maize (high FST) explain less quantitative variation in maize than teosinte, suggesting that, in these regions, allelic variants were brought to (or near) fixation during domestication. We also observed that genomic regions of high recombination explain a disproportionately large proportion of heritable variance both before and after domestication. Finally, we observed that about 75% of the additive variance in both teosinte and maize is "missing" in the sense that it cannot be ascribed to detectable QTL and only 25% of variance maps to specific QTL. This latter result suggests that morphological evolution during domestication is largely attributable to very large numbers of QTL of very small effect.

The genetic architecture of teosinte catalyzed and constrained maize domestication.

The process of evolution under domestication has been studied using phylogenetics, population genetics-genomics, quantitative trait locus (QTL) mapping, gene expression assays, and archaeology. Here, we apply an evolutionary quantitative genetic approach to understand the constraints imposed by the genetic architecture of trait variation in teosinte, the wild ancestor of maize, and the consequences of domestication on genetic architecture. Using modern teosinte and maize landrace populations as proxies for the ancestor and domesticate, respectively, we estimated heritabilities, additive and dominance genetic variances, genetic-by-environment variances, genetic correlations, and genetic covariances for 18 domestication-related traits using realized genomic relationships estimated from genome-wide markers. We found a reduction in heritabilities across most traits, and the reduction is stronger in reproductive traits (size and numbers of grains and ears) than vegetative traits. We observed larger depletion in additive genetic variance than dominance genetic variance. Selection intensities during domestication were weak for all traits, with reproductive traits showing the highest values. For 17 of 18 traits, neutral divergence is rejected, suggesting they were targets of selection during domestication. Yield (total grain weight) per plant is the sole trait that selection does not appear to have improved in maize relative to teosinte. From a multivariate evolution perspective, we identified a strong, nonneutral divergence between teosinte and maize landrace genetic variance-covariance matrices (G-matrices). While the structure of G-matrix in teosinte posed considerable genetic constraint on early domestication, the maize landrace G-matrix indicates that the degree of constraint is more unfavorable for further evolution along the same trajectory.

ZmCCT9 enhances maize adaptation to higher latitudes.

From its tropical origin in southwestern Mexico, maize spread over a wide latitudinal cline in the Americas. This feat defies the rule that crops are inhibited from spreading easily across latitudes. How the widespread latitudinal adaptation of maize was accomplished is largely unknown. Through positional cloning and association mapping, we resolved a flowering-time quantitative trait locus to a Harbinger-like transposable element positioned 57 kb upstream of a CCT transcription factor (ZmCCT9). The Harbinger-like element acts in cis to repress ZmCCT9 expression to promote flowering under long days. Knockout of ZmCCT9 by CRISPR/Cas9 causes early flowering under long days. ZmCCT9 is diurnally regulated and negatively regulates the expression of the florigen ZCN8, thereby resulting in late flowering under long days. Population genetics analyses revealed that the Harbinger-like transposon insertion at ZmCCT9 and the CACTA-like transposon insertion at another CCT paralog, ZmCCT10, arose sequentially following domestication and were targeted by selection for maize adaptation to higher latitudes. Our findings help explain how the dynamic maize genome with abundant transposon activity enabled maize to adapt over 90° of latitude during the pre-Columbian era.

Defining the Role of the MADS-Box Gene, Zea Agamous-like1, a Target of Selection During Maize Domestication.

Genomic scans for genes that show the signature of past selection have been widely applied to a number of species and have identified a large number of selection candidate genes. In cultivated maize (Zea mays ssp. mays) selection scans have identified several hundred candidate domestication genes by comparing nucleotide diversity and differentiation between maize and its progenitor, teosinte (Z. mays ssp. parviglumis). One of these is a gene called zea agamous-like1 (zagl1), a MADS-box transcription factor, that is known for its function in the control of flowering time. To determine the trait(s) controlled by zagl1 that was (were) the target(s) of selection during maize domestication, we created a set of recombinant chromosome isogenic lines that differ for the maize versus teosinte alleles of zagl1 and which carry cross-overs between zagl1 and its neighbor genes. These lines were grown in a randomized trial and scored for flowering time and domestication related traits. The results indicated that the maize versus teosinte alleles of zagl1 affect flowering time as expected, as well as multiple traits related to ear size with the maize allele conferring larger ears with more kernels. Our results suggest that zagl1 may have been under selection during domestication to increase the size of the maize ear.

The role of cis regulatory evolution in maize domestication.

Gene expression differences between divergent lineages caused by modification of cis regulatory elements are thought to be important in evolution. We assayed genome-wide cis and trans regulatory differences between maize and its wild progenitor, teosinte, using deep RNA sequencing in F1 hybrid and parent inbred lines for three tissue types (ear, leaf and stem). Pervasive regulatory variation was observed with approximately 70% of ∼17,000 genes showing evidence of regulatory divergence between maize and teosinte. However, many fewer genes (1,079 genes) show consistent cis differences with all sampled maize and teosinte lines. For ∼70% of these 1,079 genes, the cis differences are specific to a single tissue. The number of genes with cis regulatory differences is greatest for ear tissue, which underwent a drastic transformation in form during domestication. As expected from the domestication bottleneck, maize possesses less cis regulatory variation than teosinte with this deficit greatest for genes showing maize-teosinte cis regulatory divergence, suggesting selection on cis regulatory differences during domestication. Consistent with selection on cis regulatory elements, genes with cis effects correlated strongly with genes under positive selection during maize domestication and improvement, while genes with trans regulatory effects did not. We observed a directional bias such that genes with cis differences showed higher expression of the maize allele more often than the teosinte allele, suggesting domestication favored up-regulation of gene expression. Finally, this work documents the cis and trans regulatory changes between maize and teosinte in over 17,000 genes for three tissues.

Mapping Prolificacy QTL in Maize and Teosinte.

Teosinte, the ancestor of maize, possesses multiple ears at each node along its main stalk, whereas maize has only a single ear at each node. With its greater ear number, teosinte is referred to as being more prolific. The grassy tillers 1 (gt1) gene has been identified as a large-effect quantitative trait locus underlying this prolificacy difference between maize and teosinte, and the causal polymorphism for the difference was mapped to a 2.7kb control region 5' of the gt1 ORF. The most common maize haplotype (M1) at the gt1 control region confers low prolificacy. A prior study reported that 29% of maize varieties possess the teosinte haplotype (T) for the control region, although these varieties are nonprolific. This observation suggested that these maize lines might possess an additional factor, other than gt1, suppressing prolificacy in maize. We discovered that the factor suppressing prolificacy in maize varieties with the gt1 T haplotype mapped to a 3.20 cM interval, which includes gt1 Subsequent DNA sequence analysis revealed that the maize varieties with the apparent T haplotype actually possess a distinct maize haplotype (M2) that is similar, but not identical, to the T haplotype in sequence but is associated with a nonprolific phenotype similar to the M1 haplotype. Our data indicate that the M2 haplotype or a closely linked factor confers a nonprolific phenotype. Our data suggest that 2 different alleles or haplotypes (M1 and M2) of gt1 were selected during domestication, and that nonprolificacy in all maize varieties is likely a result of allele substitutions at gt1.

From many, one: genetic control of prolificacy during maize domestication.

A reduction in number and an increase in size of inflorescences is a common aspect of plant domestication. When maize was domesticated from teosinte, the number and arrangement of ears changed dramatically. Teosinte has long lateral branches that bear multiple small ears at their nodes and tassels at their tips. Maize has much shorter lateral branches that are tipped by a single large ear with no additional ears at the branch nodes. To investigate the genetic basis of this difference in prolificacy (the number of ears on a plant), we performed a genome-wide QTL scan. A large effect QTL for prolificacy (prol1.1) was detected on the short arm of chromosome 1 in a location that has previously been shown to influence multiple domestication traits. We fine-mapped prol1.1 to a 2.7 kb "causative region" upstream of the grassy tillers1 (gt1) gene, which encodes a homeodomain leucine zipper transcription factor. Tissue in situ hybridizations reveal that the maize allele of prol1.1 is associated with up-regulation of gt1 expression in the nodal plexus. Given that maize does not initiate secondary ear buds, the expression of gt1 in the nodal plexus in maize may suppress their initiation. Population genetic analyses indicate positive selection on the maize allele of prol1.1, causing a partial sweep that fixed the maize allele throughout most of domesticated maize. This work shows how a subtle cis-regulatory change in tissue specific gene expression altered plant architecture in a way that improved the harvestability of maize.

ZmCCT and the genetic basis of day-length adaptation underlying the postdomestication spread of maize.

Teosinte, the progenitor of maize, is restricted to tropical environments in Mexico and Central America. The pre-Columbian spread of maize from its center of origin in tropical Southern Mexico to the higher latitudes of the Americas required postdomestication selection for adaptation to longer day lengths. Flowering time of teosinte and tropical maize is delayed under long day lengths, whereas temperate maize evolved a reduced sensitivity to photoperiod. We measured flowering time of the maize nested association and diverse association mapping panels in the field under both short and long day lengths, and of a maize-teosinte mapping population under long day lengths. Flowering time in maize is a complex trait affected by many genes and the environment. Photoperiod response is one component of flowering time involving a subset of flowering time genes whose effects are strongly influenced by day length. Genome-wide association and targeted high-resolution linkage mapping identified ZmCCT, a homologue of the rice photoperiod response regulator Ghd7, as the most important gene affecting photoperiod response in maize. Under long day lengths ZmCCT alleles from diverse teosintes are consistently expressed at higher levels and confer later flowering than temperate maize alleles. Many maize inbred lines, including some adapted to tropical regions, carry ZmCCT alleles with no sensitivity to day length. Indigenous farmers of the Americas were remarkably successful at selecting on genetic variation at key genes affecting the photoperiod response to create maize varieties adapted to vastly diverse environments despite the hindrance of the geographic axis of the Americas and the complex genetic control of flowering time.

A distant upstream enhancer at the maize domestication gene tb1 has pleiotropic effects on plant and inflorescent architecture.

Although quantitative trait locus (QTL) mapping has been successful in describing the genetic architecture of complex traits, the molecular basis of quantitative variation is less well understood, especially in plants such as maize that have large genome sizes. Regulatory changes at the teosinte branched1 (tb1) gene have been proposed to underlie QTLs of large effect for morphological differences that distinguish maize (Zea mays ssp. mays) from its wild ancestors, the teosintes (Z. mays ssp. parviglumis and mexicana). We used a fine mapping approach to show that intergenic sequences approximately 58-69 kb 5' to the tb1 cDNA confer pleiotropic effects on Z. mays morphology. Moreover, using an allele-specific expression assay, we found that sequences >41 kb upstream of tb1 act in cis to alter tb1 transcription. Our findings show that the large stretches of noncoding DNA that comprise the majority of many plant genomes can be a source of variation affecting gene expression and quantitative phenotypes.

A unified mixed-model method for association mapping that accounts for multiple levels of relatedness.

As population structure can result in spurious associations, it has constrained the use of association studies in human and plant genetics. Association mapping, however, holds great promise if true signals of functional association can be separated from the vast number of false signals generated by population structure. We have developed a unified mixed-model approach to account for multiple levels of relatedness simultaneously as detected by random genetic markers. We applied this new approach to two samples: a family-based sample of 14 human families, for quantitative gene expression dissection, and a sample of 277 diverse maize inbred lines with complex familial relationships and population structure, for quantitative trait dissection. Our method demonstrates improved control of both type I and type II error rates over other methods. As this new method crosses the boundary between family-based and structured association samples, it provides a powerful complement to currently available methods for association mapping.

Defining the Role of prolamin-box binding factor1 Gene During Maize Domestication.

The prolamin-box binding factor1 (pbf1) gene encodes a transcription factor that controls the expression of seed storage protein (zein) genes in maize. Prior studies show that pbf1 underwent selection during maize domestication although how it affected trait change during domestication is unknown. To assay how pbf1 affects phenotypic differences between maize and teosinte, we compared nearly isogenic lines (NILs) that differ for a maize versus teosinte allele of pbf1 Kernel weight for the teosinte NIL (162mg) is slightly but significantly greater than that for the maize NIL (156mg). RNAseq data for developing kernels show that the teosinte allele of pbf1 is expressed at about twice the level of the maize allele. However, RNA and protein assays showed no difference in zein profile between the two NILs. The lower expression for the maize pbf1 allele suggests that selection may have favored this change; however, how reduced pbf1 expression alters phenotype remains unknown. One possibility is that pbf1 regulates genes other than zeins and thereby is a domestication trait. The observed drop in seed weight associated with the maize allele of pbf1 is counterintuitive but could represent a negative pleiotropic effect of selection on some other aspect of kernel composition.

grassy tillers1 promotes apical dominance in maize and responds to shade signals in the grasses.

The shape of a plant is largely determined by regulation of lateral branching. Branching architecture can vary widely in response to both genotype and environment, suggesting regulation by a complex interaction of autonomous genetic factors and external signals. Tillers, branches initiated at the base of grass plants, are suppressed in response to shade conditions. This suppression of tiller and lateral branch growth is an important trait selected by early agriculturalists during maize domestication and crop improvement. To understand how plants integrate external environmental cues with endogenous signals to control their architecture, we have begun a functional characterization of the maize mutant grassy tillers1 (gt1). We isolated the gt1 gene using positional cloning and found that it encodes a class I homeodomain leucine zipper gene that promotes lateral bud dormancy and suppresses elongation of lateral ear branches. The gt1 expression is induced by shading and is dependent on the activity of teosinte branched1 (tb1), a major domestication locus controlling tillering and lateral branching. Interestingly, like tb1, gt1 maps to a quantitative trait locus that regulates tillering and lateral branching in maize and shows evidence of selection during maize domestication. Branching and shade avoidance are both of critical agronomic importance, but little is known about how these processes are integrated. Our results indicate that gt1 mediates the reduced branching associated with the shade avoidance response in the grasses. Furthermore, selection at the gt1 locus suggests that it was involved in improving plant architecture during the domestication of maize.

Genetic signals of origin, spread, and introgression in a large sample of maize landraces.

The last two decades have seen important advances in our knowledge of maize domestication, thanks in part to the contributions of genetic data. Genetic studies have provided firm evidence that maize was domesticated from Balsas teosinte (Zea mays subspecies parviglumis), a wild relative that is endemic to the mid- to lowland regions of southwestern Mexico. An interesting paradox remains, however: Maize cultivars that are most closely related to Balsas teosinte are found mainly in the Mexican highlands where subspecies parviglumis does not grow. Genetic data thus point to primary diffusion of domesticated maize from the highlands rather than from the region of initial domestication. Recent archeological evidence for early lowland cultivation has been consistent with the genetics of domestication, leaving the issue of the ancestral position of highland maize unresolved. We used a new SNP dataset scored in a large number of accessions of both teosinte and maize to take a second look at the geography of the earliest cultivated maize. We found that gene flow between maize and its wild relatives meaningfully impacts our inference of geographic origins. By analyzing differentiation from inferred ancestral gene frequencies, we obtained results that are fully consistent with current ecological, archeological, and genetic data concerning the geography of early maize cultivation.

The role of teosinte glume architecture (tga1) in coordinated regulation and evolution of grass glumes and inflorescence axes.

• Hardened floral bracts and modifications to the inflorescence axis of grasses have been hypothesized to protect seeds from predation and/or aid seed dispersal, and have evolved multiple times independently within the family. Previous studies have demonstrated that mutations in the maize (Zea mays ssp. mays) gene teosinte glume architecture (tga1) underlie a reduction in hardened structures, yielding free fruits that are easy to harvest. It remains unclear whether the causative mutation(s) occurred in the cis-regulatory or protein-coding regions of tga1, and whether similar mutations in TGA1-like genes can explain variation in the dispersal unit in related grasses. • To address these questions TGA1-like genes were cloned and sequenced from a number of grasses and analyzed phylogenetically in relation to morphology; protein expression was investigated by immunolocalization. • TGA1-like proteins were expressed throughout the spikelet in the early development of all grasses, and throughout the flower of the grass relative Joinvillea. Later in development, expression patterns differed between Tripsacum dactyloides, maize and teosinte (Z. mays ssp. parviglumis). • These results suggest an ancestral role for TGA1-like genes in early spikelet development, but do not support the hypothesis that TGA1-like genes have been repeatedly modified to affect glume and inflorescence axis diversification.

MADS-box genes of maize: frequent targets of selection during domestication.

MADS-box genes encode transcription factors that are key regulators of plant inflorescence and flower development. We examined DNA sequence variation in 32 maize MADS-box genes and 32 randomly chosen maize loci and investigated their involvement in maize domestication and improvement. Using neutrality tests and a test based on coalescent simulation of a bottleneck model, we identified eight MADS-box genes as putative targets of the artificial selection associated with domestication. According to neutrality tests, one additional MADS-box gene appears to have been under selection during modern agricultural improvement of maize. For random loci, two genes were indicated as targets of selection during domestication and four additional genes were indicated to be candidate-selected loci for maize improvement. These results suggest that MADS-box genes were more frequent targets of selection during domestication than genes chosen at random from the genome.

The origin of the naked grains of maize.

The most critical step in maize (Zea mays ssp. mays) domestication was the liberation of the kernel from the hardened, protective casing that envelops the kernel in the maize progenitor, teosinte. This evolutionary step exposed the kernel on the surface of the ear, such that it could readily be used by humans as a food source. Here we show that this key event in maize domestication is controlled by a single gene (teosinte glume architecture or tga1), belonging to the SBP-domain family of transcriptional regulators. The factor controlling the phenotypic difference between maize and teosinte maps to a 1-kilobase region, within which maize and teosinte show only seven fixed differences in their DNA sequences. One of these differences encodes a non-conservative amino acid substitution and may affect protein function, and the other six differences potentially affect gene regulation. Molecular evolution analyses show that this region was the target of selection during maize domestication. Our results demonstrate that modest genetic changes in single genes can induce dramatic changes in phenotype during domestication and evolution.

Fine scale genetic structure in the wild ancestor of maize (Zea mays ssp. parviglumis).

Analysis of fine scale genetic structure in continuous populations of outcrossing plant species has traditionally been limited by the availability of sufficient markers. We used a set of 468 SNPs to characterize fine-scale genetic structure within and between two dense stands of the wild ancestor of maize, teosinte (Zea mays ssp. parviglumis). Our analyses confirmed that teosinte is highly outcrossing and showed little population structure over short distances. We found that the two populations were clearly genetically differentiated, although the actual level of differentiation was low. Spatial autocorrelation of relatedness was observed within both sites but was somewhat stronger in one of the populations. Using principal component analysis, we found evidence for significant local differentiation in the population with stronger spatial autocorrelation. This differentiation was associated with pronounced shifts in the first two principal components along the field. These shifts corresponded to changes in allele frequencies, potentially due to local topographical features. There was little evidence for selection at individual loci as a contributing factor to differentiation. Our results demonstrate that significant local differentiation may, but need not, co-occur with spatial autocorrelation of relatedness. The present study represents one of the most detailed analyses of local genetic structure to date and provides a benchmark for future studies dealing with fine scale patterns of genetic diversity in natural plant populations.

The role of barren stalk1 in the architecture of maize.

The architecture of higher plants is established through the activity of lateral meristems--small groups of stem cells formed during vegetative and reproductive development. Lateral meristems generate branches and inflorescence structures, which define the overall form of a plant, and are largely responsible for the evolution of different plant architectures. Here, we report the isolation of the barren stalk1 gene, which encodes a non-canonical basic helix-loop-helix protein required for the initiation of all aerial lateral meristems in maize. barren stalk1 represents one of the earliest genes involved in the patterning of maize inflorescences, and, together with the teosinte branched1 gene, it regulates vegetative lateral meristem development. The architecture of maize has been a major target of selection for early agriculturalists and modern farmers, because it influences harvesting, breeding strategies and mechanization. By sampling nucleotide diversity in the barren stalk1 region, we show that two haplotypes entered the maize gene pool from its wild progenitor, teosinte, and that only one was incorporated throughout modern inbreds, suggesting that barren stalk1 was selected for agronomic purposes.