Supplemental Methods

Plant material

Plants were grown in growth chambers at 70% humidity and daily cycles of 16h light at 21°C and 8 h darkness at 18°C.

The mutants were genotyped as follows:

For fis2-1 (Luo et al. 1999) a 360 bp fragment was amplified using the primers 5'-ACCGCTCTGCATGTAACTCTTTTCT-3’ and 5'-ATGATGAAAATGTATCATCGACACCAAG-3’ and digested with Bsu36I to distinguish the wild type pattern (180 + 180 + 180 bp) from the mutant pattern (180 bp + 360 bp). For fie-2 (Ohad et al. 1999) (1 bp G to A substitution at position 1781 of the gene), a 140 bp fragment was amplified using the primers 5‘-GTATAATCTGTGCTCGCAAATCTGC-3’ and 5’-ACTTTAGGAAGAAAAGAAAGCCAAGAG-3‘using an annealing temperature of 58°C and digested wit PstI to distinguish the wild-type fragment (115 bp) from the mutant fragment (140 bp). For msi1-2 (Guitton et al. 2004) (1bp deletion at the position 304 in the gene), a 122bp PCR fragment was amplified using the primers 5‘-ATACCTCCGAGAGCGGGCC-3’ and 5‘-CACAGCCAAAGCCACCAAAT-3’ using an annealing temperature of 58°C and digested for 3 h with AvaI to distinguish the wild-type fragment (103 bp) from the mutant fragment (122 bp).

Construction and analysis of pMEA::GUS reporter lines

A 5 kb promoter was amplified from genomic Ler DNA using the primers 5'-CGGAATTCGCTCTCCAGCCACTTGCAGCAGC-3’ and 5'-TCGAGCTCCATGGCCTCACCATCGTCCTCATGG-3’ introducing EcoRI and NcoI/SacI restriction sites respectively. This fragment spans the full intergenic region upstream of the open reading frame of MEA as well as 200 bp downstream the transcription start. The PCR product was subcloned into pBluescriptSK (Stratagene) with EcoRI/SacI and subsequently cloned into pCAMBIA 1381Z using the EcoRI/NcoI restriction site to create a translational fusion with the GUS reporter gene. The binary plasmid pDP45 generated was introduced into Agrobacterium strain GV3101; pMP90 and used for transforming Arabidopsis thaliana accession Landsberg erecta by floral dip without vacuum infiltration (Bechtold,et al. 1998). The transformants were selected on germination medium (4g/L Murashig & Skoog, Sigma) supplemented with 25 mg/L hygromycin (Duchefa).

For histochemical analysis of GUS reporter gene expression, flowers were emasculated and collected after 24 h (Fig. 1C-E) or pollinated and collected 5-6 days after pollination (Fig. 1D-F). Unfertilized gynoecia were opened and directly transferred to the reaction buffer for 72 h at 37°C (2 mM 5-bromo,4-chloro,3-indolyl-D-glucuronide (Biosynth-AG), 10 mM EDTA, 0.1% Triton X-100, 0.1 mM KFeCN, 50 mM phosphate buffer pH 7.2). Developing siliques were opened and fixed 1 h in acetone 90% at –20°C, washed three times in phosphate buffer 50 mM pH 7.2 before incubation in the reaction buffer as above. After incubation, stained material was rinsed briefly in water. Unfertilized gynoecia were dissected and mounted in 80% glycerol. Developing siliques were cleared in 8:2:1 chloral hydrate:water:glycerol 1 h then dissected and mounted in the same solution. Microscopic inspection of ovules and seeds was carried out under Nomarski optics using a light transmission microscope (DMR Leica, Germany). Pictures were recorded using a Magnafire CCD camera (Optronics).

RNA extraction and cDNA preparation

For the time-course experiments in Table 1 and Supplemental Table S1, flowers were emasculated and the gynoecia were collected 24 h later (BF, before fertilization) or hand-pollinated and harvested one, two or three days after pollination (1 dap, 2 dap, 3 dap). Typically, 10 gynoecia or 8 siliques at 1 dap and 6 siliques at 2 and 3 dap were collected in a single tube for RNA extraction. For the experiment shown Fig.1 A-B, three emasculated gynoecia or siliques were used per extraction. The tissue was frozen in liquid N and grinded 6 sec with autoclaved glass beads in a mixer (Ivoclar Vivadent, Schaan, Fürstentum Liechtenstein). Total RNA was extracted using TRIzol® (Invitrogen Life Technology) following the supplier's recommendation. The extract was treated with 2 units RNase-free DNaseI (Amersham Pharmacia Life Sciences) and purified with phenol:chloroform (1:1) before ethanol precipitation. The RNA was reverse transcribed 1 h at 42°C using 0.5µg of oligo dT primers (Invitrogen), 0.25 mM of each deoxynucleotide triphosphate (dNTP), 5 mM dithiothreitol (DTT), and 200 units of Superscript II reverse transcriptase (GIBCO-BRL), followed by heat inactivation at 72°C for 15 min.

Quantitative real-time RT-PCR

FIS2, FIE and MSI11 transcripts were detected using a Sybr Green Assay together with the reference gene ACT11 (see below). 1/12 of the cDNA preparation was used for amplification using specific primers (see below) at a concentration of 384nM in 26µl reaction volume. The PCR reaction and quantitative measurements were carried out with an ABI Prism 7700 Sequence Detection System (Applied Biosystems) using 2 min at 50°C, 10 min at 95°C, 40 cycles of 15 s at 95°C, and 60 s at 60°C. Three replicates were performed for each sample. The specificity of the unique amplification product was determined by melting curve analysis according to the manufacturer’s instructions. The primers were designed to span an exon-exon junction to ensure a cDNA-specific amplification. The sequences of the forward and reverse primers are given thereafter, and their genomic position relative to the start codon is indicated in brackets: FIS2 5‘-AACGCCTGAGACTTGAACGTC-3’ (+1840) and 5‘-GAACATGTCCATCCGCGATC-3’ (+1950); FIE 5‘-TCAAAGAGTGTGGACAACGAGATC-3’ (+752) and 5‘-ATCACACATTGGAACCGGGTA-3’ (+875); MSI1 5‘-TGGTTTGGGACCTTAGCAGG-3’ (+1103) and 5‘-TTTGCTAGTGTGACCACCGTG-3’ (+1203); FWA 5‘-CGAAAGCCTCTCGACCCTT-3’ (+1488) and 5‘-TCATTCGCTGTGCTAGCTTCA-3’ (+1588); ACT11 5’-GGAACAGTGTGACTCACACCATC-3’ and 5’-AAGCTGTTCTTTCCCTCTACGC-3’.

MEA and mea-1 or mea-2 transcripts were quantified using a Taqman assay (Perkin Elmer - Applied Biosystems) together with the internal control ACT 11. Primers and probes were all designed using the PrimerExpress software and according to the guidelines of Applied Biosystems. We developed an assay distinguishing transcripts derived from the MEA wild-type mea-1 mutant allele (Table 1) or the mea-2 mutant allele (Fig.1). Both alleles carry a Ds transposon but differ with respect to the insertion site and a footprint in mea-2 (Grossniklaus et al, 1998). The PCR fragments and Taqman probes were designed as depicted in Supplemental Figure S3. With the mea-1/MEA heterozygous mutant the wild-type MEA transcript was detected using the primers 5'-TCTGATGTTCATGGATGGGG-3’ (S11) and 5'-GGTAGGAAGAACCAATCCGATCT-3’ (reti Rev2) and the VIC labeled MGB probe 5'-TCACTCATGATGAAGCTAA-3’; the mea-1 transcript was detected using the primers Ds5-2 5'-CGTTCCGTTTTCGTTTTTTACC-3' (Grossniklaus et al. 1998) and 5'-AAGGTAAAGAAGTAGGAAGAACCAATC-3’ (reti AS3) and the FAM labeled MGB probe 5'-CCGACCGTTATCGTATAA-3’ binding in the Ds element. For the mea-2/MEA heterozygous mutant the wild-type MEA transcript was detected using the primers 5'- GGATTGCAACAATCGCTTTG-3’ (S1) and 5'- CACCAAGAGTGCCATCTCCA-3’ (AS1) and the VIC labeled MGB probe 5'- TGCTGCTAATCGTGAATG-3’; the mea-1 transcript was detected using the primers Ds5-1 5'- CCGTTTACCGTTTTGTATATCCCG-3' (Grossniklaus et al. 1998) and 5'-CCAATGCACAAATCGACAATG-3’ (S2) and the same FAM labeled MGB probe in the Ds element 5'-CCGACCGTTATCGTATAA-3’.

The master mix contained each gene specific primer in a concentration of 900nM and the MGB probe in a concentration of 190 nM.

For normalization we chose ACTIN11 (ACT11), a gene specifically expressed in the embryo sac and young developing seeds (Huang et al. 1997) unlike other standard reference genes (GADPH, 18S). While the ACT11 mRNA level is constant during the first 4-5 days after pollination (dap), it drops dramatically afterwards such that ACT11 cannot be used for normalization at later stages (Grossniklaus et al. 1998 and our data not shown).

Calculation of transcript levels and statistical analysis

The value of the CycleThreshold (CT) was averaged for each triplicate PCR reaction. The assay to detect ACTIN11 (ACT11) was done on each cDNA sample. We preferRed ACT11 to other reference genes as it is specifically expressed in the embryo sac and its zygotic products (Huang et al. 1997). We found the ACT11 transcripts to be stably expressed around fertilization and during early seed development (up to 4-5 dap, data not shown). For normalization we subtracted the CT value of ACT11 from the CT value obtained with each FIS gene detection assay, resulting in a DCT value.

For comparison of the transcript levels in the mutants and the wild type we took the DCT value of one of the wild-type sample (one of the three or more biological replicates, rep. in Supplemental Table S1) and subtracted it from the DCT value of each other samples (wild-type and mutants) to give a DDCT value. This value was in turn used to express the relative transcript levels as 2DD CT. The average and standard deviation was calculated with three to six biological replicates (number in column 'rep.' in Supplemental Table S1). For Table 1 we expressed the fold increase or decrease by dividing the average 2DD CT in the mutant with the average 2DD CT from the wild type.

The significance of the differences between relative levels in wild type and mutant samples was validated using a t test with the calculation


m, mean; s, variance; n, number of measurements; 1, wild type; 2, mutant. The level of significance was read in a t table using the appropriate degree of freedom (df= n1+n2-2). P>0.05 was considered as not significant.

For expressing the changes of the relative levels of MEA, FIS2, FIE and MSI1 transcripts along with seed development (Fig. 3A), the DCT values from wild type samples at each time point (BF, 1, 2 and 3 dap) were subtracted with one DCT value at time point BF giving a new DDCT value. As a result, the average transcript level (2DD CT) before fertilization is 1 for each gene. For expressing the relative abundance (Fig. 3B) the DCT were directly compared.

ChIP PCR

Ten primers pairs were designed to amplify partially overlapping PCR products with a size of 300 to 570bp that span 3.8 kb promoter region of the MEDEA gene (Fig. 2A). PCR assays were performed in 25mL reactions using 2mM MgCl2 (except for fragment #2, 2.5 mM) and the cycling conditions were: 4 min at 94°C, followed by 30–36 cycles of 94°C (15 s), 50-59°C (20 s), and 72°C (45 s). The annealing temperature was 59 °C for all fragments except the fragment #1 (54°C) and #2 (50°C). The PCR reactions were tested on serial dilutions of the input material to ensure a linear amplification. The sequence of the primer pairs is given below: Fragment#9: 5‘-AAAACGAGGATGGTCCATCAGC-3’ and 5‘-GCCTGATGAAGAAGGGAGAATACTC-3’ ; Fragment #8: 5‘-TATTCTCCCTTCTTCATCAGGCC-3’ and 5‘-CGTTTAACCTCTGCAACCACCA-3’; Fragment #10: 5‘-GGTGGTTGCAGAGGTTAAACGT-3’ and 5‘-GAGTTCGTTATAAATCCTTGTGTTAAAACG-3’; fragment #11: 5‘-TTACGTTTTAACACAAGGATTTATAACGAACTC-3’ and 5‘-CAGATGGATAATCGTCGTGTGTTTT-3’; Fragment #7: 5‘-AAAACACACGACGATTATCCATCTG-3’ and 5‘-TCGCAAAGATTGAAACACTGAGA-3’; Fragment #6: 5‘-TCAGTGTTTCAATCTTTGCGAATT-3’ and 5‘-TGATCCCAAAGGCACAAGCT-3’; Fragment #3: 5‘-ATCGCCCAAGCTTGTGCC-3’ and 5‘-GGAGGTTAAGTCTATCCGCCGTAA-3’; Fragment #5: 5‘-CGGGATCCGAGAGCCGTTGTGGCAGTGACC-3’ and 5‘-CGGGATCCCGATATACGGGTTAAATTCCTAGC-3’; Fragment #2: 5‘-ATATATGTCACAAAAATAGTAGAATATCAGAAAGC-3’ and 5‘-TGTCCGGTAAGACTTATATTAATGTCATTAT-3’; Fragment #1: 5‘-GCTAGGAATGAATACTAATATAT-3’ and 5‘-CTCGTCTTCTCTGATGTTGG-3’.

As controls we amplified the fragment I of PHE1 and ACTIN3 (Köhler et al. 2003).

References

Bechtold, N. and Pelletier, G. 1998. In planta Agrobacterium-mediated transformation of adult Arabidopsis thaliana plants by vacuum infiltration. Methods Mol. Biol. 82: 259-266.

Grossniklaus, U., Vielle-Calzada, J.P., Hoeppner, M.A., and Gagliano, W.B. 1998. Maternal control of embryogenesis by MEDEA, a Polycomb group gene in Arabidopsis. Science 280(5362): 446-450.

Guitton, A.E., Page, D.R., Chambrier, P., Lionnet, C., Faure, J.E., Grossniklaus, U., and Berger, F. 2004. Identification of new members of FERTILISATION INDEPENDENT SEED Polycomb group pathway involved in the control of seed development in Arabidopsis thaliana. Development 131(12): 2971-2981.

Köhler, C., Hennig, L., Spillane, C., Pien, S., Gruissem, W., and Grossniklaus, U. 2003. The Polycomb group protein MEDEA regulates seed development by controlling expression of the MADS-box gene PHERES1. Genes Dev 17(12): 1540-1553.

Luo, M., Bilodeau, P., Koltunow, A., Dennis, E.S., Peacock, W.J., and Chaudhury, A.M. 1999. Genes controlling fertilization-independent seed development in Arabidopsis thaliana. Proc Natl Acad Sci U S A 96(1): 296-301.

Ohad, N., Yadegari, R., Margossian, L., Hannon, M., Michaeli, D., Harada, J.J., Goldberg, R.B., and Fischer, R.L. 1999. Mutations in FIE, a WD Polycomb group gene, allow endosperm development without fertilization. Plant Cell 11(3): 407-416.

8