Supporting Online Material for

Transcription

1 Supporting Online Material for Synchronization of the flowering transition by the tomato TERMINATING FLOWER gene. Cora A. MacAlister 1, Soon Ju Park 1, Ke Jiang 1, Fabien Marcel 2, Abdelhafid Bendahmane 2, Yinon Izkovich 3, Yuval Eshed 3 and Zachary B. Lippman 1 * * To whom correspondence should be addressed. lippman@cshl.edu (Z.B.L.) This PDF file includes Supplemental Note Supplemental Figures 1 to 9 Supplemental Table 4 1

2 Supplemental Note Coordination of flowering When and where flowers form on a plant, as well as inflorescence organization, varies widely between related species. In annual monopodial plants like the Arabidopsis model, flowering occurs after a short vegetative stage, followed by seed set and death. In contrast, in perennial plants like trees and their tomato model, recurring flowering events during sympodial growth enable the development of new vegetative shoots, inflorescences, and flowers during a single growing season and over many years. Although both monopodial and sympodial plants undergo a primary flowering transition, sympodial plants have the added complication of coordinating floral termination with the initiation and growth of sympodial meristems. The primary flowering event in tomato can be broken down to successive stages: i) induction of a gradual vegetative to reproductive transition following germination, ii) formation of the specialized sympodial meristems, the floral sympodial bud (SIM) and vegetative sympodial shoot (SYM), iii) termination of the primary shoot in a flower, and iv) rapid release of apical dominance allowing outgrowth of the sympodial meristems. When endogenous flowering signals are limiting, as occurs when SFT is mutated, the first stage is severely delayed, floral termination is weak, and the sympodial program fails to activate 1. Likewise, overexpression of SFT (35S:SFT) causes rapid flowering, and inflorescence architecture and sympodial cycling are largely unaffected unless the flowering repressor SP is also lost. Thus, inductive flowering signals from SFT, counterbalanced by SP, impose tight coordination among all flowering stages such that induction drives floral termination, and termination is necessary and sufficient for sympodial growth 2. By analyzing the tmf mutant, we have made the surprising discovery that the process of flowering in the primary shoot can be genetically partitioned and desynchronized. Despite precocious activation of a portion of the floral specification program - specifically the FA-AN complex - in the vegetative stage of tmf mutants, the terminating primary meristem lacks many developmental hallmarks of more advanced flowering stages, and the sequential activation of floral transition (FUL) and floral identity (SEP) genes is reversed (Fig. 4). That TMF blocks precocious floral termination in the primary shoot suggests TMF transcriptional dynamics may have evolved to permit the synchronization of a gradual flowering transition with floral termination and the activation of sympodial growth. It is possible that the shortened vegetative stage of tomato axillary meristems could explain the unique sensitivity of the primary meristem to loss of TMF, and that other factors functioning redundantly with or independently of TMF might exist to prevent precocious activation of AN-FA in sympodial shoots. Our observation of recurring single flower inflorescences in plants expressing AN under the S promoter is consistent with this hypothesis. Although the AN-FA complex is highly conserved 3-5, its regulation differs between species. Based on a current model 4, Arabidopsis UFO is expressed in the SAM from embryogenesis onward 6-8, and activation of the LFY-UFO complex depends on transcriptional upregulation of 2

3 LFY during the reproductive transition. In contrast, these relations seem to be reversed in petunia, in which activation of the UFO ortholog, DOT, in the incipient floral meristem defines when flower formation begins. Unexpectedly, we find that overexpression of either FA or AN can accelerate flowering and cause single flower inflorescences. It may be that FA activation is capable of driving termination by feedback regulation between FA and AN. Importantly, recent transcriptome analysis of flowering transition dynamics in Arabidopsis SAMs revealed that UFO is upregulated 8-fold in response to floral induction, suggesting a similar relationship between LFY and UFO expression dynamics might also exist in Arabidopsis 9. 3

4 Supplementary Figure 1. Loss of tmf does not cause an accelerated transition from juvenile to adult growth. (a) Representative final leaf produced by the PSM before floral termination of WT (left) and tmf (right). (b) Quantification of leaf complexity (mean number of primary, secondary and intercalary leaflets). Leaf complexity increases over developmental time, but tmf does not exhibit precociously increased complexity, suggesting it is not undergoing a more rapid transition from juvenile to adult growth (phase change), despite precocious flower formation. Error bars are s.e.m. and different letters mark statistically significant differences (student s t-test, p-value <0.05). 4

5 Supplementary Figure 2. Double mutants between tmf and members of the florigen pathway are additive. (a-c and e) Shoot architecture diagrams of the indicated genotypes. White circles mark inflorescences and lines represent shoots producing the indicated number of leaves 5

6 (L). (a) In wild type plants the primary inflorescence is formed after approximately 8 leaves at which point sympodial cycling begins with the reiteration of a three-leaf plus inflorescence sympodial unit. (b) tmf is early flowering (after approximately 5 leaves) and fails to transition to sympodial cycling. Lateral shoots originating from canonical axillary meristems develop later and successfully enter sympodial growth. (c) In mutants of tomato florigen, single flower truss (sft/ft), the primary flowering transition is extremely late (approximately 20 leaves) and the primary inflorescence undergoes vegetative reversion to produce a leafy inflorescence interspersed with flowers. (d) Flowering time for primary and axillary meristems of the indicated genotypes. tmf partially suppresses the late flowering of sft in both the primary and axillary meristems, though tmf alone does not alter axillary meristem flowering time. (e) In mutants of the floral repressor and florigen antagonist self pruning (sp/tfl1), sympodial cycling is initiated normally, but the successive sympodial units become progressively shorter until the shoot terminates in two consecutive inflorescences. (f) Flowering time for the primary meristem and the average number of leaves produced in the first three sympodial units of each axillary shoot. tmf is early flowering for the primary shoot in both WT and sp mutant backgrounds. Similarly, lateral shoots of tmf and tmf sp double mutants initiate normal sympodial cycling, but terminate like sp in tmf sp double mutants. (g-h) The tmf sp double mutant produces the tmf single flower primary inflorescence with enlarged sepals (g) and undergoes termination as in sp (h). White dots mark the inflorescence from each successively formed sympodial unit. Note the single leaf (L) separating each unit. Error bars in d and f are s.e.m. and * marks significant differences (Wilcoxon-Mann-Whitney test * p-value <0.05, ** p-value <0.005 and *** p<0.0005). 6

7 Supplementary Figure 3. The tmf lesion is a Rider Ty1-copia-like retrotransposon insertion in Solyc09g (a) Semi-quantitative RT-PCR of Solyc09g and flanking genes from the TM of tmf (left) and WT (right). Solyc9g expression was not detected in either WT or tmf, and there is no evidence for expression at other stages of meristem maturation 10. DNA ladder is shown to the left. (b) The coding region of Solyc09g cannot be PCR-amplified from tmf genomic DNA, unlike a control sequence within the tmf mapping interval. 2-Log DNA ladder is shown from 500-1,000bp. (c) A Southern blot probed with the complete coding sequence of Solyc09g shows a genomic rearrangement in tmf mutants. For each restriction enzyme the lane order is: 1. M82 (a processing type tomato), 2. Break o Day (BOD: the WT progenitor of the tmf mutant), 3. tmf, 4. M99 (a fresh market tomato variety), and 5. Heinz 1706, which was sequenced by the tomato genome sequencing consortium 11. White arrows mark the expected fragment sizes according to the tomato genome sequence, while white asterisks highlight the structural deviation observed in tmf. (d) Rider element insertion site mapping in tmf by PCR. Chromosome 9 physical positions are given in kilobases along the top. Fragments in green are amplifiable from both WT and tmf, whereas fragments in red can only be amplified in WT suggesting a disruption in the region spanning the fragment. The blue dashed lines mark the region of genomic disturbance in tmf. Red triangle marks the observed Rider transposon insertion. 7

8 Supplementary Figure 4. The tmf-2 TILLING allele phenocopies tmf and carries a mutation that disrupts a highly conserved amino acid. (a) tmf-2 produces a primary inflorescence with a single flower with enlarged leaf-like sepals (arrow) like the original allele of tmf. (b) Flowering time in tmf and tmf-2 is accelerated in both penetrant and non-penetrant individuals. * marks statistically significant difference from WT (t-test p<0.05), error bars are s.d. (c) Diagram of the TMF protein with the site of the missense threonine105 to isoleucine mutation in tmf-2 marked in red and the conserved ALOG family domain in blue. Shown below is a partial multiple sequence alignment of TMF and the predicted protein sequences of ALOG family members in Arabidopsis thaliana, Selaginella moellendorffii (Smo) and Physcomitrella patens (Pp) showing the highly conserved amino acid disrupted in tmf-2 and flanking residues. 8

9 Supplementary Figure 5. Phylogenetic tree of the ALOG gene family. Neighbor joining phylogenetic tree based on the complete predicted protein coding sequences of the ALOG family members of tomato, Arabidopsis thaliana (LSH1-10), Selaginella moellendorffii (Smo) and Physcomitrella patens (Pp). Bootstrap values for 100 replicates are given inside nodes. 9

10 Supplementary Figure 6. The TMF protein shows transcriptional activity and interacts physically with transcriptional factors. (a) The construct used to test TMF for transcriptional activation activity and the corresponding control (left) show that the TMF full-length ORF induces His and LacZ synthesis in yeast in -His medium (middle) and liquid culture -gal assay (right) respectively. GAL4 BD (empty vector), control; -gal, -galactosidase. (b) Confirmation of interactions between TMF and selected transcription factors identified by yeast-two-hybrid screen (Supplementary Table 1). Left column is non-selective media (+His) and right column is selective media (-His) supplemented with 0.5 mm 3-aminotriazole (3AT) to reduce autoactivation of TMF. 10

11 Supplementary Figure 7. Transcriptome profiling of tmf vegetative meristems. (a-b) Representative vegetative apices of WT (a) and tmf (b) as collected for mrna sequencing. Dashed lines mark the limit of collected tissue, and leaf numbers (L) are indicated. (c-d) K- means cluster analysis of 674 differentially expressed genes between tmf and WT (Supplementary Table 2). The differentially expressed genes were divided into up-regulated (532 genes) and down-regulated (142 genes) in tmf compared to WT and compared to genes dynamically expressed during PSM maturation 10. The WT expression patterns of the 389 upregulated and 112 downregulated dynamically expressed genes were clustered by K-Means clustering using cluster numbers optimized by minimization of within-group sum of squares and 11

12 hierarchical clustering dendrograms. (c) The two clusters of genes that were down-regulated in tmf. (d) The four clusters of genes that were up-regulated in tmf. Number of genes in each cluster is given in the lower right corner. (e) A sampling of gene expression changes in tmf vegetative meristems showing that floral meristem and organ identity genes like the tomato homolog of Arabidopsis APETALA1 (SlAP1/MC) are activated precociously, while other flowering transition markers genes like S are not. The Digital Differentiation Index (DDI) algorithm quantifies unknown tissue maturation states based on expression of marker genes from known calibration tissues. (f) DDI predictions of tmf and control vegetative meristems based on known WT meristem stages (EVM: Early Vegetative Meristem, MVM: Middle Vegetative Meristem, LVM: Late Vegetative Meristem) 10. Curves represent density distribution of maturation states predicted by marker genes. Colored curves, calibration stages of a second biological replicate predicted on top of a first biological replicate; black curves, predictions of tmf and WT progenitor vegetative meristems. Heatmaps represent P values (in the form of scaled 1/(-logP); darker color indicates higher P-values and higher similarity of maturation) from pairwise t-tests between calibration stages and WT and tmf meristems. Supplementary Figure 8. Constitutive overexpression of FA/LFY produces a single flower primary inflorescence like tmf and tmf2-d. (a) Semi-quantitative RT-PCR of FA expression in the cotyledons of WT and tmf2-d (Fig. 5). (b and c) Transgenic constitutive overexpression of FA under the cauliflower mosaic virus 35S promoter recapitulates the tmf2-d early flowering and single flower primary inflorescences, frequently with enlarged sepals (white arrows) with incomplete penetrance when introgressed into the MicroTom genetic background (see Online Methods). 12

13 Supplementary Figure 9. Transcription of the Nicotiana benthamiana ortholog of AN (NbAN) is activated earlier than tomato AN during meristem maturation. (a) Stereoscope images showing matched developmental stages of PSM maturation between S. lycopersicum (cv. M82) and N. benthamiana. Dashed lines indicate tissue dissected for qrt-pcr, L=leaf number counted from 1st leaf. (b) Dynamic expression of the tomato AN and N. benthamiana AN ortholog (NbAN) by qrt-pcr showing that NbAN is already activated in the TM stage compared to tomato and is also activated earlier in the SYM (Student t-test assuming equal variance, tomato TM vs tomato LVM, P=0.5834, tomato SYM vs tomato LVM, P=0.5505; N. benthamiana TM vs N. benthamiana LVM, P=0.0029**, N. benthamiana SYM vs N. benthamiana LVM, P=0.0087**. ** indicates statistical significance at P<0.01). 13

14 Supplemental Table Legends Supplementary Table 1. List of TMF-interacting proteins as determined by Yeast Two- Hybrid analysis. [This table is provided as a separate file in MS Excel (.xlsx) Format] Tomato gene identifiers are shown along with annotated descriptions. Supplementary Table 2. Genes differentially expressed between tmf mutants and WT vegetative apices. [This table is provided as a separate file in MS Excel (.xlsx) Format] Genes showing greater than two fold change and P value < 0.05 are listed (P-values and log fold changes are shown in column C and D, respectively). The total mean RPKM expression values over two biological replicates of both genotypes are also shown (column E and F for tmf mutants and WT, respectively). Additional columns show whether a particular gene belongs to a Transcription Factor (TF) family (column F) and functional annotation of each gene (column G). Supplementary Table 3. Primer sequences. [This table is provided as a separate file in MS Excel (.xlsx) Format] Supplementary Table 4. Read number and mapping rate for mrna sequencing libraries. genotype replicate total reads mapped reads multiple mapping mapping rate multi rate Tmf 1 29,344,526 16,813, , Tmf 2 39,558,459 19,674, , WT 1 44,161,216 28,310,492 1,220, WT 2 44,205,239 22,977,411 1,055,

15 Supplemental References 1. Lifschitz, E. & Eshed, Y. Universal florigenic signals triggered by FT homologues regulate growth and flowering cycles in perennial day-neutral tomato. Journal of Experimental Botany 57, (2006). 2. Shalit, A. et al. The flowering hormone florigen functions as a general systemic regulator of growth and termination. Proc. Nat.l Acad. Sci. USA 106, (2009). 3. Chae, E., Tan, Q.K.-G., Hill, T.A. & Irish, V.F. An Arabidopsis F-box protein acts as a transcriptional co-factor to regulate floral development. Development 135, (2008). 4. Souer, E. et al. Patterning of Inflorescences and Flowers by the F-Box Protein DOUBLE TOP and the LEAFY Homolog ABERRANT LEAF AND FLOWER of Petunia. The Plant Cell 20, (2008). 5. Ikeda-Kawakatsu, K., Maekawa, M., Izawa, T., Itoh, J.-I. & Nagato, Y. ABERRANT PANICLE ORGANIZATION 2/RFL, the rice ortholog of Arabidopsis LEAFY, suppresses the transition from inflorescence meristem to floral meristem through interaction with APO1. The Plant Journal 69, (2012). 6. Long, J.A. & Barton, M.K. The development of apical embryonic pattern in Arabidopsis. Development 125, (1998). 7. Lee, I., Wolfe, D.S., Nilsson, O. & Weigel, D. A LEAFY co-regulator encoded by UNUSUAL FLORAL ORGANS. Current Biology : CB 7, (1997). 8. Weigel, D. & Nilsson, O. A developmental switch sufficient for flower initiation in diverse plants. Nature 377, (1995). 9. Torti, S. et al. Analysis of the Arabidopsis Shoot Meristem Transcriptome during Floral Transition Identifies Distinct Regulatory Patterns and a Leucine-Rich Repeat Protein That Promotes Flowering. The Plant cell 24, (2012). 10. Park, S.J., Jiang, K., Schatz, M.C. & Lippman, Z.B. Rate of meristem maturation determines inflorescence architecture in tomato. Proc. Nat.l Acad. Sci. USA 109, (2012)