TM3 and STM3 Promote Flowering Together with FUL2 and MBP20, but Act Antagonistically in Inflorescence Branching in Tomato

Abstract

The moment at which a plant transitions to reproductive development is paramount to its life cycle and is strictly controlled by many genes. The transcription factor SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (SOC1) plays a central role in this process in Arabidopsis. However, the role of SOC1 in tomato (Solanum lycopersicum) has been sparsely studied. Here, we investigated the function of four tomato SOC1 homologs in the floral transition and inflorescence development. We thoroughly characterized the SOC1-like clade throughout the Solanaceae and selected four tomato homologs that are dynamically expressed upon the floral transition. We show that of these homologs, TOMATO MADS 3 (TM3) and SISTER OF TM3 (STM3) promote the primary and sympodial transition to flowering, while MADS-BOX PROTEIN 23 (MBP23) and MBP18 hardly contribute to flowering initiation in the indeterminate cultivar Moneyberg. Protein-protein interaction assays and whole-transcriptome analysis during reproductive meristem development revealed that TM3 and STM3 interact and share many targets with FRUITFULL (FUL) homologs, including cytokinin regulators. Furthermore, we observed that mutating TM3/STM3 affects inflorescence development, but counteracts the inflorescence-branching phenotype of ful2 mbp20. Collectively, this indicates that TM3/STM3 promote the floral transition together with FUL2/MBP20, while these transcription factors have opposite functions in inflorescence development.

Keywords: FRUITFULL; MADS-box transcription factor; SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1; Solanum lycopersicum; floral transition; inflorescence architecture; reproductive meristems.

Figure 1

Characterization of tomato SOC1-like genes during the floral transition. (a) Phylogenetic tree of SOC1-like proteins from several Solanaceae species and Arabidopsis thaliana. Numbers at the branches indicate bootstrap values inferred from 5000 replicates. (b) Stereomicroscope images of reproductive meristem development. Dashed lines indicate hand-dissected tissue for transcriptome analysis. The scale bar is 200 µm. (c) Expression of the SOC1-like genes MBP23, TM3, STM3, MBP14, MBP13, MBP19 and MBP18 relative to CAC in WT VM, TM and FM/IM tissue (dCT values are shown per gene relative to the reference). Shown are the mean values ± SE of three biological replicates. L, leaf; VM, vegetative meristem; TM, transition meristem; FM/IM, floral and inflorescence meristem; SYM, sympodial shoot meristem.

Figure 2

Functional characterization of SOC1 homologs in flowering control. (a) Y2H screening showing interactions of SOC1-like proteins with several MADS-domain TFs. Reciprocal interaction is shown by dark grey shading, one-way interaction by light grey shading, and blank cells indicate a failure to interact. MBP24 was only tested as prey (AD), not as bait (BD). (b) Quantification of the number of leaves to the first inflorescence, i.e., the primary transition to flowering, in higher-order slsoc1 mutants under greenhouse conditions. (c) Quantification of the primary transition to flowering in single and double slsoc1 mutants under growth-chamber conditions. (d) Quantification of the sequential sympodial shoot floral transitions in higher-order slsoc1 mutants under greenhouse conditions. Shown are mean values of the cumulative number of leaves per sympodial unit (SU) ± SE. The average size of the first five SUs per plant was used to calculate significance. (e) Representative sympodial shoots were from WT and tm3 stm3 plants. Leaves were bent to the right for visualization. L, leaf number. In (b–d), letters indicate statistical significance.

Figure 3

TM3/STM3 share targets with the SlFULs that also independently contribute to the floral transition. (a) Venn diagram showing the overlap of DEGs of tm3 stm3 and q-ful in VM, TM and FM/IM stages. Genes with p_adj < 0.01 in at least one stage were selected as DEG. (b) Expression of genes downstream of tm3 stm3 and q-ful during reproductive meristem development. (c) MBP13 and MBP14 expression in VM, TM and FM/IM of tm3 stm3 and q-ful. (d) CKX4 and CKX6 expression in VM, TM and FM/IM of tm3 stm3 and q-ful. (e) Quantification of the primary transition until flowering of tm3 stm3, mbp23 tm3 stm3 mbp18, ful2 mbp20 and tm3 stm3 ful2 mbp20 under greenhouse conditions. (f) Expression of TFs downstream of tm3 stm3 but not q-ful during reproductive meristem development. In (b–d,f), bars show mean FPKM values ± SEM. Asterisks indicate statistical significance according to the DESeq2 P_adj values from the mutant compared to its respective WT (* P_adj < 0.05, ** P_adj < 0.01, *** P_adj < 0.001). B3 domain TF, Solyc01g108940; AP2-like ERF, Solyc06g068570; Zinc finger TF, Solyc06g065440. In (e), different letters indicate statistically significant differences. FPKM, fragments per kilobase million; VM, vegetative meristem; TM, transition meristem; FM/IM, floral and inflorescence meristem.

Figure 4

Functional characterization of SlSOC1 and SlFUL genes in inflorescence development. (a) Expression of tomato SOC1, FUL and SEP homologs in the first sFM or sIM of WT plants. Shown are the average FPKM values of four biological replicates. AN and UF are shown as marker genes as a control for sampling stages. The stereomicroscope image at the bottom shows the harvested tissue. The scale bar is 200 µm. (b) Quantification of inflorescence phenotypes. A normal inflorescence was defined as unbranched, having no revertance to vegetative growth and 7 to 10 flowers. Inflorescences deviating from this were categorized and representative inflorescences are shown at the bottom with color-coded borders. The mutants were phenotyped in two independent screenings with separate WT controls, which had a consistent fraction of normal inflorescences.

Figure 5

Model showing the proposed functions of SlSOC1 and SlFUL proteins in reproductive meristem development. White TFs interacting with FUL2/MBP20 are possibly J2/EJ2.