Glossy15 Plays an Important Role in the Divergence of the Vegetative Transition Between Maize and Its Progenitor, Teosinte
Dingyi Xu,Xufeng Wang,Cheng Huang,Guanghui Xu,Yameng Liang,Qiuyue Chen,Chenglong Wang,Dan Li,Jinge Tian,Lishuan Wu,Yaoyao Wu,Li Guo,Xuehan Wang,Weihao Wu,Weiqiang Zhang,Xiaohong Yang,Feng Tian
DOI: https://doi.org/10.1016/j.molp.2017.09.016
IF: 27.5
2017-01-01
Molecular Plant
Abstract:The timing of developmental transitions is important for plant growth and environmental adaptation. All plants undergo a series of developmental transitions during their life cycles, and each of these phases is characterized by unique morphological and physiological attributes (Bäurle and Dean, 2006Bäurle I. Dean C. The timing of developmental transitions in plants.Cell. 2006; 125: 655-664Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar). In maize, the vegetative transition from juvenile-to-adult vegetative development occurs in a coordinated manner and is marked by the production of leaves that differ in a suite of morphological and physiological traits, many of which contribute to fitness and crop productivity (Moose and Sisco, 1994Moose S.P. Sisco P.H. Glossy15 controls the epidermal juvenile-to-adult phase transition in maize.Plant Cell. 1994; 6: 1343-1355Crossref PubMed Scopus (109) Google Scholar). Significant advances have been made in understanding the molecular mechanisms regulating the juvenile-to-adult vegetative transition. Glossy15 (Gl15), an APETALA2 (AP2)-like transcription factor, plays a key role in promoting juvenile leaf identity traits and suppressing adult leaf identity traits (Moose and Sisco, 1996Moose S.P. Sisco P.H. Glossy15, an APETALA2-like gene from maize that regulates leaf epidermal cell identity.Genes Dev. 1996; 10: 3018-3027Crossref PubMed Scopus (193) Google Scholar). microRNA172 (miR172) is a negative regulator that antagonizes the activity of Gl15 by mediating Gl15 mRNA degradation (Lauter et al., 2005Lauter N. Kampani A. Carlson S. Goebel M. Moose S.P. microRNA172 down-regulates glossy15 to promote vegetative phase change in maize.Proc. Natl. Acad. Sci. USA. 2005; 102: 9412-9417Crossref PubMed Scopus (363) Google Scholar). The dominant Corngrass1 (Cg1) mutant initiates more juvenile leaves than wild-type plants do, and this phenotype was found to be caused by the ectopic overexpression of two tandem microRNA156 (miR156) genes (Chuck et al., 2007Chuck G. Cigan A.M. Saeteurn K. Hake S. The heterochronic maize mutant Corngrass1 results from overexpression of a tandem microRNA.Nat. Genet. 2007; 39: 544-549Crossref PubMed Scopus (483) Google Scholar). In Cg1 mutants, overexpression of miR156 causes a decrease of miR172, which then leads to an increase in expression of Gl15 (Chuck et al., 2007Chuck G. Cigan A.M. Saeteurn K. Hake S. The heterochronic maize mutant Corngrass1 results from overexpression of a tandem microRNA.Nat. Genet. 2007; 39: 544-549Crossref PubMed Scopus (483) Google Scholar). This regulatory circuit of miR172, miR156, and AP2 transcription factor appears to be a general mechanism for regulating vegetative phase changes in higher plants. After the vegetative transition, plants switch to reproductive growth. The correct timing of the transition to flowering is of the utmost importance, as it ensures the reproductive success of plants in changing environments and therefore has an important impact on plant fitness. Maize (Zea mays ssp. mays) was domesticated from the wild ancestor teosinte (Zea mays ssp. parviglumis), an endemic species in southern and western Mexico (Matsuoka et al., 2002Matsuoka Y. Vigouroux Y. Goodman M.M. Sanchez G.J. Buckler E. Doebley J. A single domestication for maize shown by multilocus microsatellite genotyping.Proc. Natl. Acad. Sci. USA. 2002; 99: 6080-6084Crossref PubMed Scopus (888) Google Scholar). From its Meso-American tropical origin, maize has adapted to geographically widespread environments, accompanied by dramatic changes in its flowering time (Hung et al., 2012Hung H.Y. Shannon L.M. Tian F. Bradbury P.J. Chen C. Flint-Garcia S.A. McMullen M.D. Ware D. Buckler E.S. Doebley J.F. et al.ZmCCT and the genetic basis of day-length adaptation underlying the postdomestication spread of maize.Proc. Natl. Acad. Sci. USA. 2012; 109: E1913-E1921Crossref PubMed Scopus (225) Google Scholar). Several studies have been conducted to identify the genetic factors that control the difference in flowering time between maize and teosinte, and a few key genes have been cloned (Hung et al., 2012Hung H.Y. Shannon L.M. Tian F. Bradbury P.J. Chen C. Flint-Garcia S.A. McMullen M.D. Ware D. Buckler E.S. Doebley J.F. et al.ZmCCT and the genetic basis of day-length adaptation underlying the postdomestication spread of maize.Proc. Natl. Acad. Sci. USA. 2012; 109: E1913-E1921Crossref PubMed Scopus (225) Google Scholar, Yang et al., 2013Yang Q. Li Z. Li W. Ku L. Wang C. Ye J. Li K. Yang N. Li Y. Zhong T. et al.CACTA-like transposable element in ZmCCT attenuated photoperiod sensitivity and accelerated the postdomestication spread of maize.Proc. Natl. Acad. Sci. USA. 2013; 110: 16969-16974Crossref PubMed Scopus (241) Google Scholar). In contrast to the significant advances in understanding the reproductive transition, much less is known about the changes in the vegetative transition that occurred during maize domestication. It remains unclear whether the early developmental timing observed in maize also underwent significant divergence during its domestication and how these changes were regulated and associated with local adaptation. To examine the differences in the vegetative transition between maize and teosinte, we grew a panel of 50 diverse maize inbreds and 13 teosintes (Supplemental Table 1) in a winter nursery and investigated their variation regarding the timing of the juvenile-to-adult vegetative transition. The most visually obvious phenotypic marker that distinguishes juvenile from adult development is the production of leaf epicuticular wax (Moose and Sisco, 1994Moose S.P. Sisco P.H. Glossy15 controls the epidermal juvenile-to-adult phase transition in maize.Plant Cell. 1994; 6: 1343-1355Crossref PubMed Scopus (109) Google Scholar). Juvenile maize leaves are coated with epicuticular wax and lack epidermal hairs, whereas adult maize leaves are pubescent and glossy. Following a previous method (Foerster et al., 2015Foerster J.M. Beissinger T. de Leon N. Kaeppler S. Large effect QTL explain natural phenotypic variation for the developmental timing of vegetative phase change in maize (Zea mays L.).Theor. Appl. Genet. 2015; 128: 529-538Crossref PubMed Scopus (11) Google Scholar), we scored the last leaf with epicuticular wax (LLEW) as an indicator trait for measuring the timing of the juvenile-to-adult transition in the maize and teosinte lines. As shown in Figure 1A, the maize lines exhibited an average of 6.2 leaves with juvenile epicuticular wax, which was 2.5 leaves fewer than in the teosinte plants (P < 0.01), indicating that the juvenile identity ends much earlier in maize than that in teosinte. This result suggests that maize may have evolved an early juvenile-to-adult vegetative phase change from its wild ancestor, teosinte. To identify the genetic factors controlling the divergence of the vegetative transition between maize and teosinte, we conducted quantitative trait locus (QTL) mapping for the timing of the juvenile-to-adult vegetative phase change using a large maize-teosinte experimental population, including 866 BC2S3 recombinant inbred lines (RILs) that were previously genotyped based on 19 838 SNPs (Hung et al., 2012Hung H.Y. Shannon L.M. Tian F. Bradbury P.J. Chen C. Flint-Garcia S.A. McMullen M.D. Ware D. Buckler E.S. Doebley J.F. et al.ZmCCT and the genetic basis of day-length adaptation underlying the postdomestication spread of maize.Proc. Natl. Acad. Sci. USA. 2012; 109: E1913-E1921Crossref PubMed Scopus (225) Google Scholar). LLEW was scored in this maize-teosinte BC2S3 population. A total of 12 QTLs for LLEW were mapped, which collectively explained 43% of the total phenotypic variation (Figure 1B and Supplemental Table 2). Notably, a large-effect QTL explaining 13.7% of the observed phenotypic variation was detected on chromosome 9 (qVT9-1) (Figure 1B and Supplemental Table 2). None of the other QTLs could explain more than 6% of the phenotypic variation (Supplemental Table 2). These results suggest that the vegetative transition is controlled by a single large-effect QTL and many QTLs with small effects, which is a typical characteristic of genetic architecture observed for classical domestication traits. We further found that among the 12 QTLs for LLEW, 10 loci (83.3%) acted in the same direction with the maize allele promoting the juvenile-to-adult vegetative phase change (Figure 1C and Supplemental Table 2). This consistent direction of the allelic effect across the majority of mapped QTLs provides a genetic explanation for the directional divergence of the early juvenile-to-adult vegetative phase change during maize domestication. Given the predominant role of qVT9-1 in controlling the difference in the vegetative transition between maize and teosinte, we performed fine mapping to identify the candidate gene underlying qVT9-1 using near-isogenic lines (NILs) developed from a heterogeneous inbred family (HIF) that segregates only for the qVT9-1 region. A total of 35 recombinants were identified from an F2 population containing 1500 plants using markers M1 and M5. Additional markers were developed to further resolve the breakpoints of the recombinants (Supplemental Table 5), and the 35 recombinants can be classified into six genotype groups according to their genotypes across markers. For each genotype group, one recombinant family with enough F3 seeds was planted for progeny phenotypic testing. Following the previously described fine mapping strategy (Hung et al., 2012Hung H.Y. Shannon L.M. Tian F. Bradbury P.J. Chen C. Flint-Garcia S.A. McMullen M.D. Ware D. Buckler E.S. Doebley J.F. et al.ZmCCT and the genetic basis of day-length adaptation underlying the postdomestication spread of maize.Proc. Natl. Acad. Sci. USA. 2012; 109: E1913-E1921Crossref PubMed Scopus (225) Google Scholar), homozygous recombinant and homozygous nonrecombinant plants within each F3 family were identified and compared for the difference in LLEW. qVT9-1 was finally delimited to a 263-kb region between markers M3 and M4 that included one intact gene, GRMZM2G160730, and 600 bp of GRMZM2G145104 (Figure 1D). The GRMZM2G160730 gene corresponds to the AP2-like transcription factor Gl15, which has been identified as a qualitative mutation and found to play a primary role in the maintenance of the juvenile phase (Moose and Sisco, 1996Moose S.P. Sisco P.H. Glossy15, an APETALA2-like gene from maize that regulates leaf epidermal cell identity.Genes Dev. 1996; 10: 3018-3027Crossref PubMed Scopus (193) Google Scholar, Lauter et al., 2005Lauter N. Kampani A. Carlson S. Goebel M. Moose S.P. microRNA172 down-regulates glossy15 to promote vegetative phase change in maize.Proc. Natl. Acad. Sci. USA. 2005; 102: 9412-9417Crossref PubMed Scopus (363) Google Scholar). It was clear that Gl15 was most likely the underlying gene of qVT9-1. To examine how qVT9-1 affects the vegetative phase change, we developed NILs that were homozygous for the maize allele or the teosinte allele across the qVT9-1 region (designated NILmaize and NILteosinte, respectively) and compared the differences in the production of macrohairs across leaf stages in these plants. As shown in Figure 1E, NILmaize initiated the production of epidermal hairs in leaf 5, whereas NILteosinte initiated the production of epidermal hairs in leaf 6, showing a one-leaf delay in the onset of adult vegetative development. Despite the significant difference in the vegetative transition, NILmaize and NILteosinte exhibited the same number of leaves at maturity and flowered a similar number of days after sowing, indicating that qVT9-1 has no effect on the onset of reproductive development (Supplemental Figure 1). We next measured Gl15 mRNA expression in the shoot apices of NILmaize and NILteosinte during early shoot development. At 8 days after sowing (DAS), Gl15 was expressed at similarly high levels in NILmaize and NILteosinte (Figure 1F). After 8 DAS, Gl15 expression declined in both NILteosinte and NILmaize, but the Gl15 mRNA level decreased more rapidly in NILmaize than in NILteosinte. At 13 DAS, the Gl15 mRNA level in NILteosinte remained significantly higher than in NILmaize (P < 0.05) (Figure 1F). This significant expression difference extended until 16 DAS. At 21 DAS, the Gl15 mRNA level declined to a minimum level in both NILmaize and NILteosinte, with no significant difference in Gl15 expression being detected between these lines (Figure 1F). The differential Gl15 expression pattern during early shoot development correlated well with the early vegetative phase change in NILmaize and the late phase change in NILteosinte, supporting that Gl15 was most likely the candidate gene for qVT9-1. To identify potential functional variants in the Gl15 region, we conducted an association analysis by sequencing Gl15 and its upstream and downstream regions in a panel of 517 diverse maize lines scored for LLEW. A total of 286 variants (SNPs and InDels) with MAF ≥0.05 were identified across the 6-kb sequenced region around Gl15 (Figure 1G). Using a model that corrects for familiar population structure, we tested the sequence variants for associations with LLEW in the maize association panel. A total of 32 sequence variants showed significant associations with LLEW (0.01/286 = 3.5 × 10−5, Bonferroni multiple test correction) (Figure 1G and Supplemental Table 3). None of these associated variants were located in the conserved AP2 domain or miR172 target region. The maize and teosinte parents of the BC2S3 population showed allelic polymorphism at 24 of the associated variants (Supplemental Table 3). One of the most significant variants was SNP2154 (G/A) (P = 8.1 × 10−8), which corresponded to a stop codon polymorphism (TGG versus TGA), at which the B73 reference exhibited the G nucleotide (Figure 1H and Supplemental Table 3). The maize parent of the BC2S3 population carried a G nucleotide, which promotes the vegetative transition, while the teosinte parent carried an A nucleotide, which delays the vegetative transition (Figure 1H). The A nucleotide at SNP2154 corresponds to a stop codon, while the G nucleotide encodes tryptophan, causing a downstream stop codon (position 2175) 21 bp from SNP2154 to be used (Figure 1H). Thus, relative to the maize parent of the BC2S3 population, in which the stop codon is located at position 2175 (B73 reference allele), the teosinte parent exhibits a stop codon at position 2154. Interestingly, a recent genome-wide association study for LLEW in maize nested association mapping (NAM) population also revealed significant associations in the Gl15 region (Foerster et al., 2015Foerster J.M. Beissinger T. de Leon N. Kaeppler S. Large effect QTL explain natural phenotypic variation for the developmental timing of vegetative phase change in maize (Zea mays L.).Theor. Appl. Genet. 2015; 128: 529-538Crossref PubMed Scopus (11) Google Scholar). This previous study found that an SNP located in the 9th exon of Gl15 showed the most significant association with LLEW in the NAM population. This SNP corresponds to SNP2017 in our study and is in strong linkage disequilibrium (LD) with SNP2154 (r2 = 0.53). Due to the stop codon variant SNP2154, the maize and teosinte parents of the BC2S3 population exhibited seven amino acids difference at the C-terminus of Gl15 protein. To determine whether this difference has led to functional alteration of Gl15 protein, we performed transcriptional activation assay in yeast for the maize and teosinte Gl15 protein and further analyzed their subcellular localization in epidermal cells of Nicotiana benthamiana leaves. The results showed that the maize and teosinte Gl15 protein exhibited similar transcriptional activation activity (Supplemental Figure 2) and were both localized in the nucleus (Supplemental Figure 3), indicating that SNP2154 might not alter the protein function of Gl15. Although we cannot completely exclude the possibility that the altered C-terminal length might affect the interaction ability of Gl15 protein with other protein factors, our current findings suggest that SNP2154 more likely causes the differences in Gl15 expression level that have been observed between NILmaize and NILteosinte (Figure 1F). To further verify our hypothesis, we performed a dual-luciferase transient expression assay in maize protoplasts. The miR172 precursor (pre-miR172) driven by the CaMV 35S promoter was used as an effector, and luciferase (LUC) gene fused with the Gl15-3′UTR sequence containing the miR172 target site from the maize or the teosinte parent was used as the reporter (Figure 1I). Overexpression of miR172 significantly repressed the LUC activity in the construct containing the Gl15-3′UTR sequence from the maize or the teosinte parent (P < 0.05) (Figure 1I), which was consistent with the negative regulatory role of miR172 on Gl15 (Lauter et al., 2005Lauter N. Kampani A. Carlson S. Goebel M. Moose S.P. microRNA172 down-regulates glossy15 to promote vegetative phase change in maize.Proc. Natl. Acad. Sci. USA. 2005; 102: 9412-9417Crossref PubMed Scopus (363) Google Scholar). More importantly, LUC fused with Gl15-3′UTR from the maize parent exhibited significantly lower luciferase activity than LUC fused with Gl15-3′UTR from the teosinte parent when co-transfecting with the miR172 effector (Figure 1I). This result was consistent with the lower expression of Gl15 in NILmaize than in NILteosinte during early shoot development described above. Based on these findings, we speculate that SNP2154 most likely acts as a regulatory variant of Gl15 to control the natural variation in maize vegetative transition. To examine how Gl15 contributed to the divergence of the vegetative transition during maize domestication, we analyzed the nucleotide diversity of the region flanking SNP2154 in diverse teosinte and maize lines. Interestingly, SNP2154 was found to be a standing genetic variation in teosinte, with the G allele occurring as a minor allele (Figure 1J). In contrast, in the maize lines, the frequency of the G allele increased to 65%, and became a dominant allele (Figure 1J). Given the effect of the SNP2154-G allele in promoting the juvenile-to-adult vegetative transition, the significant increase in its frequency in maize might have resulted from strong selection for an earlier vegetative phase change during maize domestication, which is expected to leave a footprint of selection at the nucleotide level. To address this question, we compared the nucleotide diversity of the maize and teosinte lines in the region flanking SNP2154 (Supplemental Table 4). Overall, we did not find evidence of a selective sweep in this region, as the maize lines retained 56% of the nucleotide diversity of teosinte, in line with the average genomic reduction of nucleotide diversity due to the bottleneck effect of domestication. However, the maize lines carrying the SNP2154-G allele retained only 23% of the nucleotide diversity found in teosinte (Figure 1K and Supplemental Table 4), in contrast to the 49% of retained nucleotide diversity in the maize lines carrying the SNP2154-A allele. A coalescent simulation incorporating the maize domestication bottleneck (Eyre-Walker et al., 1998Eyre-Walker A. Gaut R.L. Hilton H. Feldman D.L. Gaut B.S. Investigation of the bottleneck leading to the domestication of maize.Proc. Natl. Acad. Sci. USA. 1998; 95: 4441-4446Crossref PubMed Scopus (278) Google Scholar) showed that the severe loss of genetic diversity in the maize lines carrying the SNP2154-G allele could not be explained by the maize domestication bottleneck alone, indicating strong past selection for the SNP2154-G allele. In conclusion, cultivated maize has evolved an early vegetative phase transition from its ancestor, teosinte. We provided genetic evidence showing that Gl15, a known AP2-like transcription factor, might play a critical role in regulating this divergence of the vegetative transition during maize domestication. Association analysis within diverse maize lines detected a stop codon polymorphism SNP2154, which showes the most significant association with the vegetative transition. Further analyses indicate that SNP2154 most likely functions through altering the expression level of Gl15. Population genetic analyses revealed that SNP2154 was the target of a partial selective sweep that drove a pre-existing low-frequency allele in teosinte to sweep throughout the maize population during domestication. Similar selective sweep from standing genetic variation has also been observed for the tb1 and gt1 genes, which were critical for the morphological domestication of maize (Studer et al., 2011Studer A. Zhao Q. Ross-Ibarra J. Doebley J. Identification of a functional transposon insertion in the maize domestication gene tb1.Nat. Genet. 2011; 43: 1160-1163Crossref PubMed Scopus (481) Google Scholar, Wills et al., 2013Wills D.M. Whipple C.J. Takuno S. Kursel L.E. Shannon L.M. Ross-Ibarra J. Doebley J.F. From many, one: genetic control of prolificacy during maize domestication.PLoS Genet. 2013; 9: e1003604Crossref PubMed Scopus (86) Google Scholar). These results suggest that selection from standing genetic variation might be pervasive and have played important roles in plant evolution. This research was supported by the National Key Research and Development Program of China (2016YFD0100303 and 2016YFD0100404), the Recruitment Program of Global Experts and the Fundamental Research Funds for the Central Universities.