Supporting information

Supplemental Figure S1: Expression of DASH in different plant organs and in seed tissues. (A) Expression data were downloaded from the Medicago truncatula Gene Expression Atlas (MtGEA) using the probe set Mtr.21255.1.S1_at. (B) qRT-PCR on RNA extracted from 12 DAP seeds and pods (without seeds). Expression levels were obtained by normalization with the ACTIN and PDF2 genes (Verdier et al., 2008). Data are average values ± SD from three biological replicates carried out in technical replicates.

Supplemental Figure S2: Distance analysis of DOF proteins. The tree was created using full-length DOF coding sequences of 36 proteins from Arabidopsis thaliana (At), 27 from Glycine max (Gm), 22 from Medicago truncatula (Medtr), 8 from Pisum sativum (Ps) and 3 PBF sequences from cereals: Zea mays (ZmPBF), Hordeum vulgare (BPBF) and Triticum aestivum (WPBF). These 96 protein sequences were aligned and organized based on sequence similarities of the full-length DOF protein sequences by ClustalW algorithm using CLC genomics workbench software (CLC bio -CLC sequence viewer 6.2, Aarhus, Denmark). Protein sequences are provided in Table S3.

Supplemental Figure S3: Position of the EMS mutation and of the Tnt1 insertions within the DASH sequence.

Supplemental Figure S4: Genetic and molecular analysis of the dash mutants. (A) PCR-based segregation analysis of the progeny of self-fertilized heterozygous EMS109 or NF6042 mutants. (B) Frequency of small seeds within immature heterozygous pods of EMS109 and NF6042 mutants. Normal and small seed distributions are compared to a theoretical 3:1 distribution. (C) Seeds dissected from heterozygous NF6042 (top) and WT (bottom) pods at 8 DAP. Arrows indicate putative homozygous seeds. (D-E) qRT-PCR analysis of DASH expression in homozygous seeds from EMS109 (D) or NF5285 (E) mutants compared to their respective WT. Expression levels were obtained by normalization with the MSC27 and PDF2 genes. For the NF6042 line (F), RNA was extracted from seeds isolated from heterozygous pods at 8 DAP. Data were normalized with the GAPDH and PDF2 reference genes.

Supplemental Figure S5: Complementation of EMS109 mutant. Mature seed weight was measured for WT (A17), EMS109 mutant (at the end of plant life cycle) and two complemented lines (EMS109 compl A and EMS 109 compl B).

Supplemental Figure S6: Microscopic observations of embryo and endosperm at 8 DAP from WT (A, C) and EMS109 (B, D) seeds. C and D are close-up images highlighting degenerating endosperm in the dash mutant. Eo = Endosperm, Emb = Embryo, SC = Seed Coat. Scale Bars = 50µm (C-D) and 100µm (A-B).

Supplemental Figure S7: Delay in embryo growth in NF5285 mutant seeds. Cleared pictures of embryo and endosperm through early seed development.

Supplemental Figure S8: IAA dosage on pods of NF5285 vs WT from 8 to 12 DAP.

Supplemental Table S1: List of differentially expressed genes at 8 DAP in dash pods compared to WT. Log2 ratio of mutant data versus WT from 3 biological replicates using the Affymetrix GeneChip. To complement our dataset, data from the Medicago gene expression atlas (Benedito et al., 2008, Pang et al., 2008) were added (columns H to U). The colour scale ranged from red (max expression) to green (minimum expression) in the tissue.

Supplemental Table S2: List of differentially expressed genes belonging to the “auxin” functional class. Log2 ratio of mutant data versus WT from 3 biological replicates using the Affymetrix GeneChip is presented. To complement our dataset, data from the Medicago gene expression atlas (Benedito et al., 2008, Pang et al., 2008) were added (columns H to U). The colour scale ranged from red (max expression) to green (minimum expression) in the tissue.

Supplemental Table S3: List of the 96 DOF protein sequences used to compute the distance analysis in Supplemental Fig. S2.