Optimal flowering time is of central importance to a

Transcription

1 o ri ginal research Ghd7 ( Represses Sorghum Flowering in Long Days: Ghd7 Alleles Enhance Biomass Accumulation and Grain Production Rebecca L. Murphy, Daryl T. Morishige, Jeff A. Brady, William L. Rooney, Shanshan Yang, Patricia E. Klein, and John E. Mullet* Abstract Sorghum, a strong repressor of flowering in long days (LDs, was identified as the CONSTANS, CO-like, and TOC1 (CCT- domain protein encoded by SbGhd7. Sorghum Ghd7 increases photoperiod sensitivity and delays flowering by inhibiting expression of the floral activator SbEhd1 and genes encoding FT. SbGhd7 expression is light-dependent and gated by the circadian clock. In LDs when flowering is repressed, SbGhd7 mrna abundance peaks in the morning and evening. In short days (SDs when floral initiation occurs, the evening phase of SbGhd7 expression occurs in darkness, reducing SbGhd7 mrna abundance. In energy sorghum hybrids, dominant alleles of SbGhd7 and SbPRR37 act in additive fashion to delay floral initiation for ~175 d until daylengths decrease below 12.3 h. In contrast, photoperiod-insensitive grain sorghum genotypes containing recessive alleles of SbGhd7 and SbPRR37 flower in 60 to 80 d. Recessive alleles of SbGhd7 and SbPRR37 are present in historically important sorghum germplasm introduced into the United States in the 1800s that was used to produce early flowering grain and sweet sorghum. Recessive alleles ghd7-1 and prr37-1 are present in maturity standards such as SM100, and in BTx406, a genotype used in the Sorghum Conversion Program to convert late-flowering photoperiod-sensitive sorghum accessions into early flowering photoperiod-insensitive germplasm useful for grain sorghum breeding. The results show that alleles of SbGhd7 confer differences in photoperiod sensitivity and flowering times that are critical for production of late-flowering high-biomass energy sorghum and early flowering grain sorghum. Published in The Plant Genome 7 doi: /plantgenome Crop Science Society of America 5585 Guilford Rd., Madison, WI USA An open-access publication All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher. Optimal flowering time is of central importance to a plant s reproductive success. Early flowering is useful for grain production in regions where growing seasons are short and in environments where terminal drought or adverse temperatures for grain production occur frequently during the latter part of growing seasons. Longer duration of vegetative growth in forage crops increases leaf and stem biomass production (Rooney et al., In sweet sorghum (Sorghum bicolor subsp. bicolor, delayed flowering increases the size of stems and the potential for sucrose accumulation (Burks et al., In energy sorghum, delayed flowering and long duration of vegetative growth is a key trait associated with high biomass yield and nitrogen use efficiency (Rooney et al., 2007; Olson et al., 2012, In Olson et al., 2012, biomass yield was reported as 25 to 40 dry Mg/ha (or 25 to 40 dry metric t/ ha. Late-flowering, high-biomass energy sorghum and other C4 energy grasses with long vegetative growth duration could help meet the U.S. Energy Independence and Security Act of 2007 mandate to produce 21B gals of cellulosic biofuels per year (U.S. House, Flowering time is regulated in a complex manner by development (autonomous pathway, daylength R.L. Murphy, Centenary College of Louisiana, Shreveport, LA 71104; R.L. Murphy, D.T. Morishige, S. Yang, and J.E. Mullet, Dep. of Biochemistry and Biophysics, Texas AM Univ., College Station, TX J.A. Brady, Texas Agrilife Research and Extension Center, Stephenville, TX W.L. Rooney, Dep. of Soil and Crop Sciences, Texas AM Univ., College Station, TX P.E. Klein, Dep. of Horticultural Sciences and Institute for Plant Genomics and Biotechnology, Texas AM Univ., College Station, TX Received 28 Nov *Corresponding author (jmullet@tamu.edu. Abbreviations: CCT, CONSTANS, CO-like, and TOC1 domain; Ct, cycle threshold; DTF, days to flowering; INDEL, insertion-deletion polymorphism; LD, long day, 14-h light/10-h dark; LL, constant light free-running conditions; QTL, quantitative trait loci; SD, short day, 10-h light/14-h dark; SNP, single nucleotide polymorphism; SSR, simple sequence repeat. the plant genome july 2014 vol. 7, no. 2 1 of 10

2 (photoperiod, temperature, gibberellins, shading, and other factors (Imaizumi and Kay, 2006; Andrés and Coupland, 2012; Sanchez et al., 2011; Higgins et al., 2010; Tsuji et al., Many of the pathways that control flowering time converge on the regulation of floral integrator FT-like genes of the PEBP gene family that encode peptides (florigens that move from the leaf to the shoot apex where they bind to FD and induce floral meristem transition. In Arabidopsis, photoperiod regulation of flowering time is mediated by circadian clock and light regulation of CONSTANS (CO activity, an activator of FT in LDs (Imaizumi and Kay, In rice (Oryza sativa L., sorghum, and other grasses, FT-like genes (OsHd3a/OsRFT1, ZCN8 are also regulated by the timing and extent of EARLY HEADING DATE 1 (Ehd1 expression, an activator of FT expression unique to grasses, and by a large number of other genes that modulate Ehd1 and FT expression (Doi et al., 2004; Itoh et al., 2010; Tsuji et al., Sorghum is a drought-tolerant C4 grass that is cultivated widely as a source of grain, forage, sugar, and biomass, especially in drought-prone regions of the world (Rooney, 2004, 2007; Vermerris, 2011; Smith and Frederiksen, 2000; Dahlberg et al., Optimal flowering time of sorghum varies from ~50 to >200 d, depending on growing location and crop type. Grain sorghum is generally selected for early flowering to enhance grain yield and yield stability, whereas energy sorghum is selected for late flowering to enhance biomass yield (Rooney, 2004, The sorghum germplasm collection (n = 40,000 varies extensively in flowering time and photoperiod sensitivity (Quinby and Karper, 1945; Quinby, 1966; Craufurd et al., For example, a survey of sorghum accessions revealed critical photoperiods ranging from 10.2 to 17.5 h (Craufurd et al., Sorghum flowering time is modified by temperature (Major et al., 1990, gibberellins (Morgan et al., 1986, and photoperiod (Major et al., 1990; Childs et al., 1997 among other factors. More than 40 QTL for flowering time have been identified in sorghum (Mace et al., The degree of photoperiod sensitivity in sorghum, a SD plant, depends in part on alleles of the maturity loci Ma 1 through (Quinby and Karper, 1945; Quinby, 1966; Rooney and Aydin, Light signaling through phytochrome B (Ma 3 is required for photoperiod sensitivity in sorghum as in other plants (Childs et al., Ma 1 encodes PRR37 an inhibitor of flowering in LDs that represses expression of the floral activator Ehd1 and SbCN8 (FT; Murphy et al., In this study, we report the identification of, a locus that causes very late flowering in LDs (Rooney and Aydin, 1999, as SbGhd7, a floral repressor regulated by the circadian clock and light signaling. Materials and Methods Plant Materials The ATx623 R BC 1 population (n = 1821 was grown under field conditions in the summer of 2003 in College Station, Vega, and Plainview, TX. For flowering date determinations, days to midanthesis (pollen shed were evaluated. BTx623 F 2 plants (n = 124 were grown under LDs in greenhouse conditions in 2008 and phenotyped for DTF (anthesis. Tissue was collected from individuals for genotyping, and seed was collected for planting. Following F 2 family genotyping, selected F 3 progeny (n = 255 from that population were grown under the same LD greenhouse conditions in 2010, and scored for flowering at anthesis. Tissue was collected from this generation as well, and 185 combined progeny from both populations were used in downstream QTL mapping. Map-Based Cloning of Ghd7 Genomic DNA was isolated from each mapping line using the FastDNA Spin Kit (MP Biomedicals, Solon, OH. High throughput genotyping by 77 bp single-end sequencing on an Illumina Genome Analyzer IIx was used to identify novel SNP markers of parental lines and F 2 /F 3 offspring. Analysis of raw sequencing data was performed according to Morishige et al. (2013. Genetic map construction was performed using Mapmaker/EXP 3.0 (Lander et al., 1987, and the resulting map was used in combination with flowering time phenotypic data as input in QTL Cartographer. The log of odds threshold was set using permutations of the data at 1000 iterations. Quantitative trait loci above this threshold were considered statistically significant. The locus was refined further in the BTx623 F 2 / F 3 and ATx623 R BC 1 populations using SSR markers (SSRIT, Temnykh et al., 2001, cleaved amplified polymorphic sequence markers, SNPs, and insertiondeletion polymorphisms (INDELS. The SNP and INDEL markers were discovered by de novo sequencing of regions surrounding predicted open reading frames. Capillary sequencing of these markers was performed on the ABI 3130xl Genetic Analyzer, and sequence assembly and analysis was performed using Sequencher v. 4.8 (Gene Codes Corporation, Ann Arbor, MI. All markers used for fine mapping are listed in Supplemental Table S3. Ghd7 Allele Sequencing Genomic DNA from each line in Table 1 was extracted using the FastDNA Spin Kit (MP Biomedicals, Solon, OH, and amplified using primers flanking the Ghd7 region. Ghd7 gene and coding sequences were amplified, and sequencing reactions were performed using the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA. These reactions were sent for capillary sequencing on the Applied Biosystems 3130xl Genetic Analyzer. The results were analyzed using Sequencher v of 10 the plant genome july 2014 vol. 7, no. 2

3 Table 1. Analysis of Ghd7 alleles in historically important sorghum accessions. Genotype Ma 1 DTF Description Energy sorghum R Ghd7-2 PRR37 >175 very photoperiod sensitive R Ghd7-5 PRR37 >175 very photoperiod sensitive R Ghd7-6 PRR37 >175 very photoperiod sensitive Historical genotypes Red Kafir ghd7-2 prr grain sorghum (1876 Tx Blackhull Kafir ghd7-1 prr grain sorghum (1890 Dwarf Yellow Milo ghd7-1 PRR37 grain sorghum (1905; ma 2 ma 3 Double Dwarf Feterita ghd7-2 prr grain sorghum (1906 Hegari ghd7-2 PRR grain sorghum (1908; ma 4 D ouble Dwarf Yellow Milo ghd7-1 PRR37 grain sorghum (1918; ma 2 ma 3 Early White Milo ghd7-1 prr grain sorghum (1920 Standard Blackhull Kafir ghd7-1 prr grain sorghum (1924 Kalo ghd7-2 prr grain sorghum (1933 Caprock (Tx7000 ghd7-1 prr grain sorghum (1941 Combine 7078 ghd7-1 prr grain sorghum (1944 CK60 (BTx3197 ghd7-1 prr grain sorghum (1950 Tx623 ghd7-1 prr grain sorghum (1960s Rio ghd7-2 prr sweet sorghum (1972 Conversion program BTx406 ghd7-1 prr conversion program IS12646 Ghd7-3 PRR37 >175 exotic accession IS12646c ghd7-1 prr converted accession Maturity genotypes (Tx623 R Ghd7 PRR37 >175 Ma 1 Ma 2 Ma 3 Ma 4 100M ghd7-1 PRR Ma 1 Ma 2 Ma 3 Ma 4 SM100M ghd7-1 prr ma 1 Ma 2 Ma 3 Ma 4 Days to flowering (DTF were determined under long day (LD summer greenhouse conditions. The relative order of flowering times among the genotypes studied were similar in LD summer and fall greenhouse conditions, although all plants flowered later under the higher temperature conditions in the summer. DTF data obtained from the USDA-ARS Germplasm Resource Information Network National Plant Germplasm System. For the ghd7-2 allele, a protocol for flanking PCR was used to delimit the border of a repetitive sequence inserted into the intron (Siebert et al., Genomic DNA (5 µg from control genotypes (ghd7-1 and Ghd7, as well as from a ghd7-2 line, was digested using blunt end cutting restriction enzymes (New England BioLabs, Inc.. Adapters (Supplemental Table S4 were then ligated on to the end of the fragments with T4 DNA ligase (Promega, and PCR was performed and subjected to capillary sequencing, followed by analysis with Sequencher v Expression Analysis The expression of Ghd7 was analyzed in leaf tissue collected from 39-d-old 100M and ATx623 R07007 plants treated with LD or SD light conditions at 3-h intervals, as previously described (Murphy et al., Raw cycle threshold (Ct values were collected for each gene, and calibrated to 18S ribosomal RNA to obtain Figure 1. Phenotypic analysis of field-grown sorghum. The ATx623 R (center flowers later, is taller, and accumulates more biomass than ATx623 (left and R (right parental lines used to generate the. Plants were planted on 16 Apr in College Station, TX, and photographed in August. Ct values. Relative expression was calculated using the Comparative Ct ( Ct method (Bookout and Mangelsdorf, 2003 with the most highly expressed sample used as the calibration sample between both LD and SD samples. Mean values are based on three technical replicates and three biological replicates for both reference and target genes, ±SEM (Bookout and Mangelsdorf, Accession Codes Ghd7 allele sequences have been deposited at GenBank, Accession No. JX448383, JX448384, JX448385, JX448386, JX448387, JX448388, JX448389, JX Results To analyze how various combinations of Ma 1 alleles affect flowering time, late-flowering hybrids were produced by crossing Tx623 (ma 1,, ; ~56 d to flowering, DTF to R (Ma 1, ma 5 ; 74 DTF; Fig. 1. When planted in College Station, TX, in early April, plants initiate flowering in September, after ~175 d of growth when daylengths decrease below 12.3 h, reaching anthesis in mid-to-late October (Rooney and Aydin, The long duration of vegetative growth of the energy sorghum hybrids results in more than twice as much biomass accumulation compared with grain sorghum hybrids that flower in June (Olson et al., To examine the role plays in determining flowering time, plants were selfed to create F 2 and F 3 populations (n = 182 that were grown and phenotyped for flowering time in a greenhouse under 14 h LD conditions during the fall. Genetic analysis revealed three major quantitative trait loci (QTL in this population that modified flowering time; (SBI-01 (SBI-06, ~0.7 Mbp, and Ma 1 (SbPRR37; SBI-06 at Mbp; Murphy et al., 2011; Rooney and Aydin, 1999; Fig. 2A. Under LD murphy et al.: ghd7 represses flowering in energy sorghum 3 of 10

4 Figure 2. Quantitative trait loci (QTL mapping of (blue triangle. (A Statistically significant QTL are those with log of odds (LOD scores that surpass the threshold value (horizontal red line, determined by permutation test at 1000 iterations. Chromosome numbers are given along the abscissa. Ma 1 (purple triangle and (green triangle are noted above their respective QTL peaks. (B Fine mapping of. The interval was refined to 267kb between SNP1 and SSR5 on the end of chromosome 6. Close up of the region. IND1 lies in the promoter of the strongest candidate gene, Grain number, plant height, and heading date (Ghd7; Supplemental Table S3. Recombinants are given in parentheses. (C Ghd7 allele sequences. The dominant Ghd7-1 allele from R contains a CONSTANS, CO-like, and TOC1 (CCT domain. The ghd7-1 allele from ATx623 contains a five base insertion in the first exon. This causes a frameshift and a premature stop. The ghd7-2 allele contains a large insertion in the intron (black triangle, which affects the integrity of the second exon. Exons are shown as boxes, introns as solid lines. Golden yellow color is the protein coding sequence; red indicates CCT domain; pale yellow indicates extensive missense. 100-bp scale is given. The pink area represents the portion of the DNA that would code for the CCT domain if not for the insertion in ghd7-2. greenhouse conditions, F 3 plants recessive at ma 1 and (ma 1 ma 1, flowered in 65 d, plants dominant for either Ma 1 or (ma 1 ma 1 _; Ma 1 _, flowered in 96 to 100 d and plants dominant for both Ma 1 (Ma 1 flowered in ~120 d. These results indicate that Ma 1 (PRR37 act in an additive fashion to repress flowering under LD conditions. Differences in flowering time in this population conferred by inheritance of various Ma 1 allelic combinations were examined under field conditions in Selected genotypes were planted in College Station, TX, in early April, and harvested in mid-september. During this period, daylengths increase from 12.5 h at Figure 3. Flowering time in BTx623 R derived lines with various Ma 1 allelic combinations. In the parental lines Tx623 and R.07007, and in other lines recessive for both ma 1 and (lines 04, 07, 7 to 10 internodes were produced. In lines recessive for either ma 1 or (lines 09 and 10, 15 to 16 internodes were produced. In plants dominant for both Ma 1 (as well as, 40 to 43 internodes were produced. Selected genotypes were planted under field conditions in College Station, TX, in early April 2012 and harvested in late September. The number of stem nodes was used as an indirect measure of floral initiation. emergence to 14.3 h in July, and then decrease to 12.3 h in mid-september. All genotypes were harvested in late September, and variation in the number of stem nodes (equivalent to leaf number was used as an indirect measure of differences in days to floral initiation (Major et al., The parental lines Tx623 and R.07007, and progeny genotypes that were recessive for both Ma 1, produced 7 to 10 internodes (Fig. 3. Plants producing this number of internodes initiate flowering in late May or early June, and reach anthesis in late June or early July. Lines with genotypes that were recessive for either Ma 1 or in otherwise dominant maturity allelic backgrounds produced 15 to 16 internodes (Fig. 3. plants that were dominant for both Ma 1 had 40 to 43 internodes at the end of September, consistent with the additive repressing action of Ma 1 on flowering time in LD field environments. 4 of 10 the plant genome july 2014 vol. 7, no. 2

5 Figure 4. Relative expression of Ghd7 in the 100M genotype for a full 72-hour time course. Plants were treated under 14-h light/10-h dark (long day, LD, solid line or 10-h light/14-h dark (short day, SD, red dashed line conditions. The ordinate represents expression normalized to 18S ribosomal RNA expression and relative to a calibrator sample and is based on three biological and three technical replicates ± SEM (Bookout and Mangelsdorf, The black bar above the plot indicates the dark period for LD-treated plants; the gray bar indicates subjective dark during constant light free-running (LL conditions. Red bars indicate darkness for SD-treated plants; pink indicates subjective dark during LL conditions. Open bars denote light periods. Light-gray shading within the plot area indicates darkness for SD treated plants only; dark-gray shading indicates darkness for both LD- and SD-treated plants. Ma 1 was recently identified as PRR37 (Murphy et al., 2011, but the molecular nature of was not known before this study. The gene underlying was identified using a BC 1 population (n = 1821 created by backcrossing plants derived from Tx623 R to the male sterile genotype ATx623 (ma 1. This mapping population segregated for flowering time caused by alleles of Ma 1 (PRR37. Markers for Ma 1 alleles enabled variation in flowering time due to this locus to be accounted for. The remaining variation in flowering time and markers spanning the locus were used to delimit the interval encoding to a region between markers SNP1 and SSR5 (Fig. 2B; SNP = single nucleotide polymorphism, SSR = simple sequence repeat, covering 267 kb and 14 genes (Supplemental Table S1; Phytozome v. 8.0, verified 25 Mar The best candidate for was Sb06 g000570, an ortholog of rice Ghd7 as determined by sequence similarity and comparative analysis of Ghd7 in grass genomes (Yang et al., The functional Ghd7 allele in R consists of a 741 bp coding-region translating to a 246 amino acid product, which contains a CCT domain spanning residues 187 to 231 (Fig. 2C. The sequence of ghd7-1 from Tx623 ( contained a five base insertion, creating a frameshift and premature termination of the protein (Fig. 2C; Supplemental Table S2. No obvious functional polymorphisms were detected in the other genes in the delimited locus. Sorghum PRR37 mrna abundance peaks in the morning and evening in LDs, but only in the morning in SDs, through coordinate regulation by light and the circadian clock (Murphy et al., SbPRR37 and SbGhd7 both encode repressors that inhibit flowering in LD, suggesting that expression of these genes might be regulated in a similar way. To examine this possibility, the expression of SbGhd7 was analyzed by qrt-pcr in leaves of plants entrained in LD or SD, followed by constant light free-running conditions (LL; Murphy et al., 2011; Itoh et al., 2010; Samach and Coupland, 2000; Millar and Kay, In 100M (Ma 1 plants grown in LD conditions, Ghd7 mrna abundance peaks 3 h (Fig. 4; arrow and 15 h (Fig. 4, arrowhead after initial light exposure in the morning, similar to SbPRR37 (Murphy et al., By contrast, in SD, the evening peak of SbGhd7 expression is reduced relative to LD (Fig. 4, red dashed line, indicating differential expression in LD and SD photoperiods. Regulation of Ghd7 expression was further examined by quantifying expression in 100M under constant illumination (Fig. 4, 24 to 72 h LL. The peaks in SbGhd7 mrna abundance observed in the morning and evening in LD plants continued for at least 2 d under LL conditions, indicating that SbGhd7 is controlled by the circadian clock. Furthermore, when SD-treated plants that show only a morning peak of Ghd7 RNA abundance are allowed to free run in LL, the evening peak of Ghd7 expression reappears (Fig. 4, 39 and 63 h. Taken together, these data indicate that differential Ghd7 expression in LD and SD photoperiods depends on exposure to light and gating by the circadian clock, consistent with prior observations on PRR37 in sorghum (Murphy et al., To further examine the light dependence of SbGhd7 expression, leaf tissue was collected from plants (Ma 1 grown in LD and then released into constant light (Fig. 5A or dark (DD; Fig. 5B. Ghd7 mrna abundance varied as observed in 100M during the 24 h LD cycle, (Fig. 5A/B; arrow and arrowhead, but abundance was reduced during DD, indicating that light is required for high levels of Ghd7 expression (Fig. 5B. In rice, Ghd7 represses flowering by inhibiting expression of Ehd1, an activator of the FT-genes Hd3a and RFT1 (Xue et al., In sorghum grown in LD, PRR37 was found to reduce the expression of SbEhd1 and to act murphy et al.: ghd7 represses flowering in energy sorghum 5 of 10

6 Figure 5. Ghd7 expression is clock- and light-dependent. Plants were treated under14-h light/10-h dark (long day, LD, solid line or 10-h light/14-h dark (short day, SD, red dashed line conditions. Relative expression of Ghd7 was analyzed at 3-h intervals by quantitative RT-PCR. (A ATx623 R plants Ghd7 expression increased in the morning (arrow and evening (arrowhead of LDs. (B The expression of Ghd7 is diminished in ATx623 R plants grown in 14-h/10-h LD and then released into DD at time 24 h. The ordinate represents expression normalized to 18S ribosomal RNA expression and relative to a calibrator sample and is based on three biological and three technical replicates ± SEM (Bookout and Mangelsdorf, The black bar above the plot indicates the dark period for LD-treated plants; the gray bar indicates subjective dark during constant light freerunning (LL conditions. Red bars indicate darkness for SD-treated plants; pink indicates subjective dark during LL conditions. Open bars denote light periods. Light-gray shading within the plot area indicates darkness for SD treated plants only; dark-gray shading indicates darkness for both LD- and SD-treated plants. together with to inhibit flowering (Murphy et al., 2011; Rooney and Aydin, To determine if SbGhd7 inhibits expression of the floral activator SbEhd1, the expression of this gene in leaves of plants (Ma 1 and 100M (Ma 1, plants grown in LDs was quantified and compared. In SDs, SbEhd1 mrna abundance peaked twice during the night in both genotypes (Fig. 6A, red line. In the, expression of SbEhd1 was greatly repressed in LD relative to SD and at times was undetectable (Fig. 6A, black line. SbEhd1 expression was also repressed in 100M in LD compared with SD (Murphy et al., Comparison of SbEhd1 RNA levels in 100M and plants grown in LD showed that Ehd1 RNA abundance was >10-fold higher in leaves of 100M (Ma 1, compared with (Ma 1 plants (Supplemental Fig. S1A; 100M, green line vs., orange line. In rice, Ehd1 up-regulates the expression of Hd3a (FT and RFT1, members of the PEBP-gene family that act as florigens in this species (Doi et al., Sorghum lacks RFT1 but possesses a gene that is collinear with Hd3a (SbCN15, as well as the FT-like gene SbCN8, a Figure 6. Ehd1 and SbCN8 expression is repressed in LD in Ghd7 plants. Plants were treated under14-h light/10-h dark (long day, LD, solid line or 10-h light/14-h dark (short day, SD, red dashed line conditions. Relative expression of Ehd1 and SbCN8 was analyzed at 3-h intervals by quantitative RT-PCR. In the ATx623 R.07007, (A Ehd1 and (B SbCN8 are strongly repressed in LDs, but not in SDs. The ordinate represents expression normalized to 18S ribosomal RNA expression and relative to a calibrator sample and is based on three biological and three technical replicates ± SEM (Bookout and Mangelsdorf, The black bar above the plot indicates the dark period for LD-treated plants; gray bars indicate subjective dark during constant light freerunning (LL conditions. Red bars indicate darkness for SD-treated plants; pink indicates subjective dark during LL conditions. Open bars denote light periods. Light-gray shading within the plot area indicates darkness for SD treated plants only; dark-gray shading indicates darkness for both LD- and SD-treated plants. predicted ortholog of ZCN8, the florigen in maize (Meng et al., 2011; Lazakis et al., 2011; Supplemental Table S3. In plants (Ma 1, SbCN8 is repressed in LD relative to SD (Ma 1, ; Murphy et al., 2011; Fig. 6B and repression occurs to a greater extent in plants compared with 100M (Supplemental Fig. S1B. Analysis of SbCN12, SbCN15, and CO expression in the and 100M indicated that SbGhd7 in Ma 1 genetic backgrounds modifies expression of these genes to a more limited extent (Supplemental Fig. S2, and the expression patterns of SbCN8 and SbCN12 are similar (each has two peaks but not identical (Fig. 6B and Supplemental Fig. S2A; Murphy et al., Five additional functional alleles of SbGhd7 were identified in highly photoperiod-sensitive sorghum genotypes (Supplemental Table S2. Sorghum accessions encoding functional alleles of Ghd7 and PRR37 (Ma 1 flowered in >175 d in LD summer greenhouse conditions (Table 1. Many of the genotypes that flowered earlier in LDs contained the ghd7-1 null allele, including historically 6 of 10 the plant genome july 2014 vol. 7, no. 2

7 important progenitors of the U.S. grain sorghum-breeding program, such as Texas Blackhull Kafir (circa 1890, Table 1. A second recessive allele, ghd7-2, was identified in grain sorghum genotypes such as Red Kafir (circa 1876, Double Dwarf Feterita (1906, Hegari (1908, and the sweet sorghum genotype Rio (1972; Table 1, Supplemental Table S2, Supplemental Fig. S3. The ghd7-2 allele is characterized by insertion of a large repetitive DNA at position 268 bp within Ghd7 s only intron (Fig. 2C, Supplemental Fig. S4. The size of this insertion was too large to characterize precisely; however, BLAST analysis of sequences from the insertion revealed related repetitive sequences throughout the sorghum genome. Populations derived from BTx623 Rio segregated for Ma 1 but not for, indicating that the recessive ghd7-2 allele in Rio is null or very weak (Murray et al., In the 1940s, sorghum maturity standards that varied in alleles of Ma 1 through Ma 4 were constructed from Double Dwarf Yellow Milo and Early White Milo (Quinby, M and SM100, two of the maturity standards that differ in Ma 1 (SbPRR37 alleles, were used previously to determine how PRR37 modifies the expression of genes in the flowering time pathway (Murphy et al., Sequence analysis revealed that the maturity standards 100M and SM100 and their progenitors possess ghd7-1 (Table 1. Rooney and Aydin (1999 showed that 100M and SM100 are dominant for. When grown in 14-h daylengths in a greenhouse in summer conditions, the relative flowering times of the (Tx623 R.07007, >175 d, Ma 1 ma 1, 100M (~140 d, Ma 1 Ma 1,, and SM100 (~69 d, ma 1 ma 1, were consistent with the additive repressing action of Ma 1 on floral induction. In 1963, the USDA established the Sorghum Conversion Program to convert late-flowering exotic sorghum accessions into early flowering genotypes useful for grain production (Dahlberg, The genotype BTx406 was developed as a source of recessive ma 1 (prr37-1 and dw 2 for use in the conversion program (Stephens et al., This genotype was crossed to tall late-flowering exotic sorghum accessions to substitute these alleles for Ma 1 and Dw 2 in exotic germplasm, converting them to early flowering, short genotypes, respectively (Supplemental Fig. S3. Sequence analysis showed that BTx406 contains a null allele of (ghd7-1 in addition to the null allele of Ma 1 (prr37-1, Table 1. Conversion of the late-flowering line IS12646 to early flowering IS2646c involved substituting ghd7-1 and prr37-1 from BTx406 for the corresponding dominant alleles present in IS12646 (Table 1. These results indicate that the molecular basis for the success of the Sorghum Conversion Program based on BTx406 is that recessive alleles for, ma 1, and dw 2 are all present in this genotype at one end of LG-06, resulting in the efficient conversion of tall (Dw 2, photoperiod-sensitive exotic genotypes that contained either Ma 1, ma 1, or Ma 1 into short, early flowering lines (Klein et al., Discussion, a repressor of flowering in sorghum, was identified in this study as SbGhd7. Ghd7 also functions as a repressor of flowering in rice (Xue et al., 2008, wheat (Triticum aestivum L.; Yan et al., 2004, barley (Hordeum vulgare L.; Trevaskis et al., 2006, and maize (Zea mays L.; Ducrocq et al., 2009; Hung et al., In sorghum, Ghd7 confers photoperiod sensitivity and delays flowering in LDs, even in genotypes that are null for PRR37 (Ma 1, another repressor of flowering in LDs (Murphy et al., Therefore, as observed in rice (Koo et al., 2013, Ghd7 can repress flowering in sorghum through a pathway independent of PRR37. In sorghum genotypes that contain functional PRR37 alleles, Ghd7 acts in an additive fashion with PRR37 to enhance photoperiod sensitivity and delay flowering until daylengths are <12.3 h under field conditions (Fig. 7. In rice, Ghd7 and PRR37 were also reported to repress flowering in an additive fashion (Koo et al., In LDs, SbGhd7 and SbPRR37 both reduce expression of the floral activator SbEhd1 and SbCN8 (ortholog of ZCN8, a source of florigen in maize and flowering (Murphy et al., 2011; this study. SbGhd7 further reduced the expression of the floral activator SbEhd1 in Ma 1 (PRR37 genotypes consistent with the additive repressing action of Ma1 and Ma6 on floral induction. The molecular basis of the additive repressing action of Ghd7 and PRR37 on Ehd1 expression in sorghum remains to be elucidated. However, studies of the interaction of wheat ZCCTs (Ghd7-like factors, CO, and NF-Y transcription factors that modulate flowering time gene expression may provide insight into the biochemical basis of repression (Li et al., In sorghum, Ghd7 and PRR37 decrease Ehd1 expression in LDs but not SDs, consistent with their ability to modulate flowering in response to photoperiod. Expression of both genes peaks in the morning and evening in LDs, and evening phase expression is attenuated in SDs coincident with floral induction. Expression of Ghd7 and PRR37 is regulated by the circadian clock and light, suggesting common upstream regulation of the two genes. Regulation of Ghd7 expression by photoperiod is observed in rice (Xue et al., 2008, wheat (Yan et al., 2004; Distelfeld and Dubcovsky, 2010, and barley (Trevaskis et al., In rice, high Ghd7 expression requires the activity of GIGANTEA (GI, a component of the circadian clock (Itoh et al., The two peaks of Ghd7 RNA observed in sorghum during LDs have not been reported in rice. In barley, Ghd7 (VRN2 genes show two to three peaks of expression in LDs (Trevaskis et al., The significance of these differences is not yet clear; however, in sorghum, variation in the amplitude of the evening peak of Ghd7 expression in LDs vs. SDs is highly correlated with differences in flowering time. Sorghum genotypes that are dominant for both and Ma 1 are highly photoperiod sensitive and exhibit strong repression of Ehd1 and SbCN8 (FT-like gene in maize and sorghum in LDs resulting in very murphy et al.: ghd7 represses flowering in energy sorghum 7 of 10

8 Figure 7. A simplified model depicting the co-regulation of flowering by SbPRR37 (Ma 1 and SbGhd7 (. Both light and circadian clock contribute to the regulation of floral repressors SbGhd7 and SbPRR37, which in turn downregulate expression of SbCN genes in the leaves, thereby delaying the floral transition in the shoot meristem. late flowering under field conditions in College Station, TX (~200 d. Knowledge of the action of these floral repressors and the presence of alleles of Ma 1 in sorghum germplasm has aided the construction of early flowering pollen and seed parent inbreds that can be used to generate energy sorghum hybrid seed in locations also used to produce grain sorghum hybrids (Rooney et al., The resulting energy sorghum hybrids have enhanced photoperiod sensitivity resulting in long duration of vegetative growth when grown in most locations in the United States, greatly increased yield of biomass for biofuels production (Olson et al., 2012, and enhanced nitrogen use efficiency (Olson et al., The enhanced biomass yield of energy sorghum hybrids and other C4 grasses with long duration of vegetative growth would reduce the acreage needed for biomass production for biofuels by up to 50% compared with biomass generated by a grain sorghum crop (Olson et al., 2012; Somerville et al., 2010; Carroll and Somerville, 2009; Byrt et al., Historically important sorghum accessions used for development of grain sorghum were found to encode recessive alleles of Ghd7 and, in most cases, recessive alleles of PRR37. The genotypes from Africa with recessive ghd7 and prr37 alleles came from South Africa (Blackhull Kafir, Red Kafir and Egypt (Feterita, Hegari, latitudes where daylengths may be too long during the growing season to be compatible with grain genotypes that have high photoperiod sensitivity. In contrast, highly photoperiod-sensitive genotypes often originate within 20 degrees of the equator where increased photoperiod sensitivity could be useful to delay flowering when daylengths differ to only a small extent. Craufurd et al. (1999 identified 16 sorghum accessions originating within 20 degrees of the equator, with critical photoperiods <12.5h (10.2 to 12.5 h similar to the energy sorghum hybrids (Ma 1 described in this study. In contrast, the photoperiod insensitive sorghum accessions, with one exception, were adapted to higher latitudes (20 to 45 N or S. Grain sorghum grown in the United States is largely photoperiod insensitive and selected to flower in ~60 to 80 d to minimize exposure to drought, insect pressure, or cold temperature during the reproductive phase depending on location. The finding that grain sorghum used in the United States contains recessive alleles of and Ma 1 is consistent with the need for early flowering at the latitudes (26 to 45 N and daylengths (>12.3 h during the season where most sorghum grain production occurs in the United States. Similarly, alleles of Ghd7 and PRR37 have been found to be important for growing rice at different latitudes (Xue et al., The USDA Sorghum Conversion Program was established in the 1960s to convert late-flowering exotic germplasm into photoperiod insensitive early flowering sorghum genotypes by introgressing recessive alleles of both Ma 1 (prr37-1 (ghd7-1 from BTx406 into these accessions. The discovery that is closely linked to Ma 1 and Dw 2 on SBI-06 suggests that this linkage aided the success of the USDA Sorghum Conversion Program before DNA marker-based selection. This program converted ~800 late-flowering accessions into short early flowering lines useful for grain sorghum breeding, greatly increasing the genetic diversity of sorghum germplasm and grain sorghum hybrids used in the United States and elsewhere. Knowledge of the flowering time regulatory pathway and alleles of Ghd7 and PRR37 is now enabling marker-assisted selection of germplasm for production of early flowering grain sorghum hybrids and late-flowering energy sorghum hybrids (Rooney et al., In summary, alleles of Ghd7 have had an important impact on sorghum, wheat, barley, maize, and rice breeding programs, demonstrating the special significance of this gene family for cereal production by enabling optimized flowering time for grain and energy crops. 8 of 10 the plant genome july 2014 vol. 7, no. 2

9 Supplemental Information Available Supplemental information is included with this article, see file Supplementary Information. Acknowledgments This research was supported by the Office of Science (Biological and Environmental Research, U. S. Department of Energy, Grant No. DE-FG02-06ER64306, the Perry Adkisson Chair (J.E. Mullet, and Ceres Inc. The authors do not declare a conflict of interest. Author contributions: J.E. Mullet was the principal investigator who conceived the study. R.L.Murphy, D.T. Morishige, J.A. Brady, S. Yang, and P.E. Klein carried out QTL mapping and fine mapping studies. R.L. Murphy and D.T. Morishige conducted allele sequencing and expression analysis. P.E. Klein performed bioinformatic analysis. W.L. Rooney created the populations used and directed field studies. R.L. Murphy and J.E. Mullet were primarily responsible for writing the paper. References Andrés, F., and G. Coupland The genetic basis of flowering responses to seasonal cues. Nat. Rev. Genet. 13: doi: /nrg3291 Bookout, A.L., and D.J. Mangelsdorf Quantitative real-time PCR protocol for analysis of nuclear receptor signaling pathways. Nucl. Recept. Signal. 1:e012. doi: /nrs Byrt, C., C. Grof, and R. Furbank C4 plants as biofuel feedstocks: Optimising biomass production and feedstock quality from a lignocellulosic perspective. J. Integr. Plant Biol. 53: doi: / j x Burks, P., T.J. Felderhoff, H.P. Viator, and W.L. Rooney The influence of hybrid maturity and planting date on sweet sorghum productivity during a harvest season. Agron. J. 105: doi: /agronj Carroll, A., and C. Somerville Cellulosic Biofuels. Annu. Rev. Plant Biol. 60: doi: /annurev.arplant Childs, K.L., F.R. Miller, M.M. Cordonnier-Pratt, L.H. Pratt, P.W. Morgan, and J.E. Mullet The sorghum photoperiod sensitivity gene, Ma(3, encodes a phytochrome B. Plant Physiol. 113: doi: /pp Craufurd, P.Q., V. Mahalakshmi, F.R. Bidinger, A. Qi, E.H. Roberts, R.H. Ellis, R.J. Summerfield, and G.L. Hammer Adaptation of sorghum: Characterisation of genotypic flowering responses to temperature and photoperiod. Theor. Appl. Genet. 99: doi: / s Dahlberg, J.A Collection, conversion, and utilization of sorghum. In: C.W. Smith and R.A. Frederiksen, editors, Sorghum: Origins, history, technology, and production. John Wiley & Sons, New York. p Dahlberg, J.A., J. Berenji, and V.C. Sikora Assessing sorghum [Sorghum bicolor (L. Moench] germplasm for new traits: Food, fuels and unique uses. Maydica 56: Distelfeld, A., and J. Dubcovsky Characterization of the maintained vegetative phase deletions from diploid wheat and their effect on VRN2 and FT transcript levels. Mol. Genet. Genomics 283: doi: /s Doi, K., T. Izawa, T. Fuse, U. Yamanouchi, T. Kubo, Z. Shimatani, M. Yano, and A. Yoshimura Ehd1, a B-type response regulator in rice, confers short-day promotion of flower and controls FT-like gene expression independently of Hd1. Genes Dev. 18: doi: /gad Ducrocq, S., C. Giauffret, D. Madur, V. Combes, F. Dumas, S. Jouanne, D. Coubriche, P. Jamin, L. Moreau, and A. Charcosset Fine mapping and haplotype structure analysis of a major flowering time quantitative trait locus on maize chromosome 10. Genetics 183: doi: /genetics Higgins, J.A., P.C. Bailey, and D.A. Laurie Comparative genomics of flowering time pathways using Brachypodium distachyon as a model for the temperate grasses. PLoS ONE 5(4:e doi: / journal.pone Hung, H.-Y., L.M. Shannon, F. Tian, P.J. Bradbury, C. Chen, S.A. Flint- Garcia, M.D. McMullen, D. Ware, E.S. Buckler, J.F. Doebley, and J.B. Holland ZmCCT and the genetic basis of day-length adaptation underlying the postdomestication spread of maize. Proc. Natl. Acad. Sci. USA 109(28:E1913 E1921. doi: /pnas Imaizumi, T., and S.A. Kay Photoperiodic control of flowering: Not only by coincidence. Trends Plant Sci. 11: doi: /j. tplants Itoh, H., Y. Nonoue, M. Yano, and T. Izawa A pair of floral regulators sets critical day length for Hd3a florigen expression in rice. Nat. Genet. 42: doi: /ng.606 Klein, R.R., J.E. Mullet, D.J. Jordan, F.R. Miller, W.L. Rooney, M.A. Menz, C.D. Franks, and P.E. Klein The effect of tropical sorghum conversion and inbred development on genome diversity as revealed by high-resolution genotyping. The Plant Genome [A Supplement to Crop Sci.] 48:S12 S26. doi: /cropsci tpg Koo, B.-H., S.-C. Yoo, J.-W. Park, C.-T. Kwon, B.-D. Lee, A. Gynheung, Z. Zhang, J. Li, Z. Li, and N.-C. Paek Natural variation in OsPRR37 regulates heading date and contributes to rice cultivation at a wide range of latitudes. Mol. Plant. 6(6: doi: /mp/sst088 Lander, E.S., P. Green, J. Abrahamson, A. Barlow, M.J. Daly, S.E. Lincoln, and L.A. Newberg MAPMAKER: An interactive computer package for constructing primary genetic linkage maps of experimental and natural populations. Genomics 1: doi: / ( Lazakis, C.M., V. Coneva, and J. Colasanti ZCN8 encodes a potential orthologue of Arabidopsis FT florigen that integrates both endogenous and photoperiod flowering signals in maize. J. Exp. Bot. 62(14: doi: /jxb/err129 Li, C., A. Distelfeld, A. Comis, and J. Dubcovsky Wheat flowering repressor VRN2 and promoter CO 2 compete for interactions with NUCLEAR FACTOR-Y complexes. Plant J. 67: doi: / j x x Mace, E.S., C.H. Hunt, and D.R. Jordan Supermodels: Sorghum and maize provide mutual insight into the genetics of flowering time. Theor. Appl. Genet. 126: doi: /s z Major, D.J., S.B. Rood, and F.R. Miller Temperature and photoperiod effects mediated by the sorghum maturity genes. Crop Sci. 30: doi: /cropsci x x Meng, X., M.G. Muszynski, and O.N. Danilevskaya The FT-Like ZCN8 gene functions as a floral activator and is involved in photoperiod sensitivity in maize. Plant Cell 23: doi: / tpc Millar, A.J., and S.A. Kay Circadian control of cab mrna accumulation in Arabidopsis. Plant Cell 3: Morgan, P.W., L.W. Guy, and C.-I. Pao Genetic regulation of development in Sorghum bicolor. Plant Physiol. 83: doi: / pp Murphy, R.L., R.R. Klein, D.T. Morishige, J.A. Brady, W.L. Rooney, F.R. Miller, D.V. Dugas, P.E. Klein, and J.E. Mullet Coincident light and clock regulation of pseudoresponse regulator protein 37 (PRR37 controls photoperiodic flowering in sorghum. Proc. Natl. Acad. Sci. USA 108:16,469 16,474. doi: /pnas Murray, S.C., A. Sharma, W.L. Rooney, P.E. Klein, J.E. Mullet, S.E. Mitchell, and S. Kresovich Genetic improvement of sorghum as a biofuel feedstock: I. QTL for stem sugar and grain nonstructural carbohydrates. Crop Sci. 48: doi: /cropsci Olson, S.N., K. Ritter, J. Medley, T. Wilson, W. Rooney, and J. Mullet Energy sorghum hybrids: Functional dynamics of high nitrogen use efficiency. Biomass Bioenergy 56: doi: /j. biombioe Olson, S.N., K. Ritter, W. Rooney, A. Kemanian, B.A. McCarl, Y. Zhang, S. Hall, D. Packer, and J. Mullet High biomass yield energy sorghum: Developing a genetic model for C4 grass bioenergy crops. Biofuels Bioproducts Biorefining 6: doi: /bbb.1357 Quinby, J.R Fourth maturity gene locus in sorghum. Crop Sci. 6: doi: /cropsci x x murphy et al.: ghd7 represses flowering in energy sorghum 9 of 10

10 Quinby, J.R The maturity genes of sorghum. Adv. Agron. 19: doi: /s ( Quinby, J.R., and R.E. Karper Inheritance of three genes that influence time of floral initiation and maturity date in milo. Agron. J. 37: doi: /agronj x Rooney, W.L Sorghum improvement Integrating traditional and new technology to produce improved genotypes. Adv. Agron. 83: doi: /s ( Rooney, W.L., and S. Aydin Genetic control of a photoperiod-sensitive response in Sorghum bicolor (L. Moench. Crop Sci. 39: doi: /cropsci x x Rooney, W.L., J. Blumenthal, B. Bean, and J.E. Mullet Designing sorghum as a dedicated bioenergy feedstock. Biofuel. Bioprod. Bior. 1(2: doi: /bbb.15 Samach, A., and G. Coupland Time measurement and the control of flowering in plants. Bioessays 22: doi: /(sici ( :13.0.co;2-l Sanchez, S.E., J.I. Cagnola, M. Crepy, M.J. Yanovsky, and J.J. Casa Balancing forces in the photoperiodic control of flowering. Photochem. Photobiol. Sci. 10: doi: /c0pp00252f Siebert, P.D., A. Chenchik, D.E. Kellogg, K.A. Lukyanov, and S. Lukyanov An improved PCR method for walking in uncloned genomic DNA. Nucleic Acids Res. 23: doi: / nar/ Smith, C.W., and R.A. Frederiksen History of cultivar development in the United States: From Memoirs of A.B. Maunder-Sorghum Breeder. In: C.W. Smith and R.A. Fredrikson, editors, Sorghum: Origin, history, technology, and production. John Wiley & Sons, New York. p Somerville, C., H. Youngs, C. Taylor, S.C. Davis, and S.P. Long Feedstocks for Lignocellulosic Biofuels. Science 329: doi: /science Stephens, J., F.R. Miller, and D.T. Rosenow Conversion of alien sorghums to early combine genotypes. Crop Sci. 7:396. doi: / cropsci x x Temnykh, S., G. DeClerck, A. Lukashova, L. Lipovich, S. Cartinhour, and S. McCouch Computational and experimental analysis of microsatellites in rice (Oryza sativa L.: Frequency, length variation, transposon associations, and genetic marker potential. Genome Res. 11: doi: /gr Trevaskis, B., M.N. Hemming, W.J. Peacock, and E.S. Dennis HvVRN2 responds to daylength, whereas HvVRN1 is regulated by vernalization and developmental status. Plant Physiol. 140: doi: /pp Tsuji, H., K.-I. Taoka, and K. Shimamoto Regulation of flowering in rice: Two florigen genes, a complex gene network, and natural variation. Curr. Opin. Plant Biol. 14: doi: /j.pbi U.S. House th Congress. Energy Independence and Security Act of U.S. Gov. Print. Office, Washington, DC. Vermerris, W Survey of Genomics Approaches to Improve Bioenergy Traits in Maize, Sorghum and SugarcaneFree Access. J. Integr. Plant Biol. 53: doi: /j x Xue, W., Y. Xing, X. Weng, Y. Zhao, W. Tang, L. Wang, H. Zhou, S. Yu, C. Xu, X. Li, and Q. Zhang Natural variation in Ghd7 is an important regulator of heading date and yield potential in rice. Nat. Genet. 40: doi: /ng.143 Yan, L., A. Loukoianov, A. Blechl, G. Tranquilli, W. Ramakrishna, P. San Miguel, J.L. Bennetzen, V. Echenique, and J. Dubcovsky The Wheat VRN2 Gene is a Flowering Repressor Down-Regulated by Vernalization. Science 303: doi: /science Yang, L., T. Liu, B. Li, Y. Sui, J. Chen, J. Shi, R.A. Wing, and M. Chen Comparative sequence analysis of the Ghd7 orthologous regions revealed movement of Ghd7 in the grass genomes. PLoS ONE 7:e doi: /journal.pone of 10 the plant genome july 2014 vol. 7, no. 2