Florigen-Encoding Genes of Day-Neutral and Photoperiod-Sensitive Maize Are Regulated by Different Chromatin Modifications at the Floral Transition

Abstract

The activity of the maize (Zea mays) florigen gene ZEA CENTRORADIALIS8 (ZCN8) is associated with the floral transition in both day-neutral temperate maize and short-day (SD)-requiring tropical maize. We analyzed transcription and chromatin modifications at the ZCN8 locus and its nearly identical paralog ZCN7 during the floral transition. This analysis was performed with day-neutral maize (Zea mays ssp. mays), where flowering is promoted almost exclusively via the autonomous pathway through the activity of the regulatory gene indeterminate1 (id1), and tropical teosinte (Zea mays ssp. parviglumis) under floral inductive and noninductive photoperiods. Comparison of ZCN7/ZCN8 histone modification profiles in immature leaves of nonflowering id1 mutants and teosinte grown under floral inhibitory photoperiods reveals that both id1 floral inductive activity and SD-mediated induction result in histone modification patterns that are compatible with the formation of transcriptionally competent chromatin environments. Specific histone modifications are maintained during leaf development and may represent a chromatin signature that favors the production of processed ZCN7/ZCN8 messenger RNA in florigen-producing mature leaf. However, whereas id1 function promotes histone H3 hyperacetylation, SD induction is associated with increased histone H3 dimethylation and trimethylation at lysine-4. In addition, id1 and SD differently affect the production of ZCN7/ZCN8 antisense transcript. These observations suggest that distinct mechanisms distinguish florigen regulation in response to autonomous and photoperiod pathways. Finally, the identical expression and histone modification profiles of ZCN7 and ZCN8 in response to floral induction suggest that ZCN7 may represent a second maize florigen.

Figure 1.

Photographs of line B73 maize and teosinte plants induced and uninduced for flowering. A, Wild-type (wt) plant at V7 stage. The inset shows a meristem at the floral transition stage. B, Maize id1 mutant plant at V7 stage. The inset shows a meristem at the vegetative stage. C, Teosinte grown under NB regimen for 27 d and induced for 10 d under SD conditions. The inset shows an inflorescence meristem. D, Teosinte plant grown under noninductive NB conditions for 37 d. The inset shows an uninduced meristem. Teosinte plants exhibited extensive tillering, but only leaves from the main shoot were harvested for analysis. Bars = 0.25 mm for all inset photographs.

Figure 2.

RT-PCR analysis of ZCN7 and ZCN8 transcripts in line B73 and teosinte plants. A, RT-PCR was performed with oligo(dT)-primed cDNA prepared with RNA extracted from mature (ML) and immature (IL) leaves and with (+) or without (−) the addition of reverse transcriptase (RT). Schemes of ZCN7 and ZCN8 genes and positions of primers used for RT-PCR are reported. The structures of both genes are conserved in line B73 and teosinte. B, Strand-specific RT-PCR carried out using forward (ZCN7-for and ZCN8-for) and reverse (ZCN7-rev and ZCN8-rev) primers for synthesizing cDNA produced by the antisense and sense RNA strands, respectively. Subsequent RT-PCR with ZCN7/8-1 and ZCN7/8-2 primers permits the detection of both spliced and unspliced RNAs. C, Diagram schematizing the three transcript isoforms produced by ZCN7 and ZCN8 genes. RNA length and position with respect to the gene structure are based on strand-specific RT-PCR results with line B73 plants.

Figure 3.

Strand-specific qRT-PCR analysis of ZCN7 and ZCN8 transcript isoforms in B73 and teosinte plants. Real-time qRT-PCR quantification is shown for ZCN7 and ZCN8 sense and antisense transcripts from mature (ML) and immature (IL) leaves of wild-type (wt) and id1 mutant B73 plants (A) and of teosinte plants grown under inductive SD and inhibitory NB flowering conditions (B). Bar diagrams show mean values of transcript amounts for one biological replicate, normalized to glyceraldehyde-3-phosphate dehydrogenase2 (gapc2) sense mRNA. Wild-type and SD values are set to 1, and the amount and direction (increase or decrease) of change for each RNA isoform in id1 mutants versus the wild type (A) or in SD versus NB conditions (B) was calculated using the 2^−ΔΔCt method as described by Rossi et al. (2007; for values less than 1, the negative value was obtained by applying the formula −1/2^−ΔΔCt). Asterisks indicate statistically significant changes (P ≤ 0.01) achieved in separate analyses of the two biological replicates.

Figure 4.

Analysis of ZCN8 histone modifications in line B73 wild-type (wt) and id1 mutant plants. A, Schematic depiction of the ZCN8 gene, with black boxes representing exons. The positions of the regions analyzed in ChIP assays are indicated. B, Bar diagrams represent real-time PCR quantification of ChIP DNA, reported as percentage of the chromatin input, from assays performed using the indicated antibodies. The data are average values from two independent ChIP assays and from three PCR repetitions for each ChIP assay and are reported by subtracting the background signal, measured by omitting antibody during the ChIP procedure. Asterisks indicate statistically significant changes (P ≤ 0.01) in the id1 mutant versus the wild type. The grouping of histone marks on the basis of how its variation associates with the activity of the id1 gene is indicated in parentheses. Similar results were obtained after correction for nucleosome occupancy measured as reported by Rossi et al. (2007). H3Cter, Histone H3 C-terminal region.

Figure 5.

Analysis of ZCN8 histone modifications in teosinte SD and NB plants. A, Schematic depiction of the ZCN8 gene, with black boxes representing exons. The positions of the regions analyzed in ChIP assays are indicated. B, Bar diagrams represent real-time PCR quantification of ChIP DNA, reported as percentages of chromatin input, from assays using the indicated antibodies. The data are average values from two independent ChIP assays and from three PCR repetitions for each ChIP assay and are reported by subtracting the background signal, measured by omitting antibody during the ChIP procedure. Asterisks indicate statistically significant changes (P ≤ 0.01) in SD versus NB plants. The grouping of histone marks on the basis of how its variation associates with SD floral induction is indicated in parentheses. Similar results were obtained after correction for nucleosome occupancy measured as reported by Rossi et al. (2007). H3Cter, Histone H3 C-terminal region.

Figure 6.

Summary of ZCN7 and ZCN8 epigenetic pattern variations in line B73 and teosinte plants. The diagram is based on data reported in Figures 4 and 5 and in Supplemental Figures S6 and S7. The direction of each arrow indicates the direction of the change for the histone modification listed at left and for cytosine methylation (mC) measured by means of restriction with the MspJI enzyme. Variations of epigenetic mark levels that occur in immature leaf and that are maintained in mature leaf are highlighted by black arrows. WT, Wild-type.

Figure 7.

Model of florigen regulation in autonomous maize and photoperiod-induced teosinte. Autonomous maize (left) requires ID1 regulatory protein activity (orange stars) in developing leaves to establish chromatin modifications that allow the expression of florigen genes (ZCN7 and ZCN8). Thus, the id1 gene acts in immature leaves to establish a chromatin signature and prime the leaf for florigen synthesis as the leaf develops. Active chromatin is specified by H3ac. Once the distal portion of the leaf develops, another signal (unknown) activates florigen production in leaf vasculature, which then migrates to the shoot apical meristem (SAM) to activate flowering genes (purple dotted line). The autonomous signal may consist, partly, of changes in metabolic activity. Metabolic changes could also indirectly activate florigen production (?). In teosinte (right), floral induction is dependent on SD photoperiods and the circadian clock to activate florigen production. Similar to id1, the photoperiod pathway also establishes chromatin modifications in immature leaves, which enable florigen synthesis in mature leaves, but the pattern of histone modifications related to its activity is different from the one created by id1 in the autonomous pathway (i.e. open chromatin is specified by H3K4me2/H3K4me3). The horizontal dashed lines across the mature maize and teosinte leaves delineate the regions of the immature, developing leaf zone (bottom) from the mature leaf blade (top).