Genetic Control and Comparative Genomic Analysis of Flowering Time in Setaria (Poaceae)

We report the first study on the genetic control of flowering in Setaria, a panicoid grass closely related to switchgrass, and in the same subfamily as maize and sorghum. A recombinant inbred line mapping population derived from a cross between domesticated Setaria italica (foxtail millet) and its wild relative Setaria viridis (green millet), was grown in eight trials with varying environmental conditions to identify a small number of quantitative trait loci (QTL) that control differences in flowering time. Many of the QTL across trials colocalize, suggesting that the genetic control of flowering in Setaria is robust across a range of photoperiod and other environmental factors. A detailed comparison of QTL for flowering in Setaria, sorghum, and maize indicates that several of the major QTL regions identified in maize and sorghum are syntenic orthologs with Setaria QTL, although the maize large effect QTL on chromosome 10 is not. Several Setaria QTL intervals had multiple LOD peaks and were composed of multiple syntenic blocks, suggesting that observed QTL represent multiple tightly linked loci. Candidate genes from flowering time pathways identified in rice and Arabidopsis were identified in Setaria QTL intervals, including those involved in the CONSTANS photoperiod pathway. However, only three of the approximately seven genes cloned for flowering time in maize colocalized with Setaria QTL. This suggests that variation in flowering time in separate grass lineages is controlled by a combination of conserved and lineage specific genes.

LOD graphs for each of the eight trials. Each graph shows the LOD scores for each trial along each chromosome in Setaria, with chromosomes lined up end to end (I-IX). The horizontal blue line represents the P < 0.05 significance level, so that QTL are declared significant if the peaks are above that line. In several cases there are multiple peaks underlying a single QTL. GH = Greenhouse, GC = Growth Chamber, F = Field. OK = Oklahoma State University, Stillwater, OK; BT = Boyce Thompson Institute, Ithaca, New York; GA = University of Georgia, Athens, GA. The second greenhouse trial (GH2-OK) was grown in 2010 with additional metal halide and high-pressure sodium lighting at Oklahoma State University during the months of July to September. Day length, consisting of natural light with supplemental lighting before dawn and after dusk, was 16 hours, with an average daily temperature of 26.5 °C, and light intensity of 1400 μmol cm -2 s -1 . In this trial only 107 RILs were evaluated due to limitations in space. The experimental design consisted of a RCBD with four blocks, two planted at high density (11 cm apart) and two at low density (31 cm apart), with one replicate of each RIL per block. RILs were planted in 11x11x14 cm square pots (approximately 1573 cm 3 ), filled with Metro-Mix 366 (Sun Gro Horticulture Canada Ltd.). Plants were irrigated as needed with an aqueous complete fertilizer mix (Jack's mix: Nitrogen,). consisted of a RCBD with two blocks, with three replicates of each RIL per block. RILs were planted 25 cm apart from each other in rows. In order to mitigate the variable competition effects stemming from RILs of different sizes growing together, RIL rows were separated by rows of border plants, with rows 25cm apart. A commercial foxtail millet variety (German millet) was used for the borders and plants were watered as needed. One nitrogenous fertilizer application (as urea) was done in the sixth week of growth at a rate of 50 lb/acre (56 kg/ha). The second trial (F2-OK) was also a RCBD with four blocks, two planted at high density (15 cm apart) and two at low density (60 cm apart), with four replicates of each RIL per block. German millet was used as border plants, as in the first field trial.
Two field trials were also conducted at the University of Georgia, Athens, in June-July 2011 and July-August 2011 (14.3 hours day length, average temperature of 28 0 C). These trials are labeled F1-GA and F2-GA. For the first trial (F1-GA), seeds were sown in soil filled peat pots (1 seed/pot). The pots were placed at 4 0 C for 1 week after which they were placed in the glasshouse under natural light conditions. Twenty days later, the plants were planted in the field in hill plots in a RCBD with two blocks. For the second trial, ~100 seeds per RIL were sown in soil in pots, placed at 4 0 C for 4 days and then in the glasshouse under 14.2 hours natural day light for four days. Two-day old seedlings were transplanted to soil-filled peat pots (1 plant/pot), and 10-day old seedlings were planted in hill plots in a RCBD with two blocks. In both trials, hill plots were separated by 36" and each hill plot contained a single RIL in four replicates. Plots were irrigated when necessary.

Growth Chamber Trials: The first growth chamber experiment (GC-BT) was conducted at the Boyce Thompson
Institute. Plants were grown under artificial light with 12 hours of light (750 µmol.m -2 .s -1 ) and 12 hours of darkness.
Average temperature was 31˚C during the light period and 25˚C during the dark. Five plants per RIL were planted in 220.5 cm 3 pots filled with metro mix 360 (Sun Gro Horticulture), and thinned to two plants 10 days after sowing.

Molecular Marker Development and Genotyping
Almost two hundred published SSR primer pairs (JIA et al. 2007;JIA et al. 2009;GUPTA et al. 2012) were tested on the parents and a total of 126 informative markers were chosen to genotype the RIL population. PCR fragment separation for SSR markers was done via agarose gels (1 to 3 % depending on fragment sizes) or with an ABI PRISM 3730 Genetic Analyser (Applied Biosystems). Amplification reactions for SSR fragments separated in agarose gels were done in a 10 μl volume containing 0.5 Units of Taq DNA polymerase (Promega Corp), 1.5 mM MgCl 2 , 100 μM dNTPs, 20 ng of genomic DNA, 2.0 μl of 5x GoTaq buffer (Promega Corp) and 0.32 μM of forward and reverse primer.
Some primers required the addition of 5% DMSO (indicated DMSO in Supplementary Table1) to consistently amplify.
PCR cycling conditions consisted of an initial denaturation step at 94 °C for 2 min followed by 35 amplification cycles with an initial touchdown, and a final 5 min extension cycle at 72 °C. Each of the 35 cycles had a denaturation step at 94 °C for 40 s followed by 40s at the annealing temperature which was started 4 °C above the primers' melting temperature and decreased every two cycles by 2 °C to reach the primers' annealing temperature by the fifth cycle, and an extension step of 1 min at 72 °C. PCR fragments were visualized in 1 to 3 % agarose gels stained with ethidium bromide. Amplification reactions for SSR markers analyzed with the ABI PRISM 3730 Genetic Analyser (Applied Biosystems) were done using the "economic method" developed by Schuelke (SCHUELKE 2000) and further improved by Missiaggia and Grattapaglia (MISSIAGGIA and GRATTAPAGLIA 2006). PCR reactions were done essentially as described above, except that primer concentrations were 0.10 μM for the forward tailed primer, 0.3 μM for the reverse primer and 0.2 μM for the fluorescently labeled forward primer  Locus, linkage group, primer combinations, reaction conditions (where different from general description in text), polymorpism detection method, and Genbank accession number for STS markers (first number in each case is for S. italica B100 and the second is for S. viridis A10). An asterisk (*) after the locus name denotes an RFLP marker (WANG et al. 1998).