SEARCH

SEARCH BY CITATION

Keywords:

  • endosperm;
  • cell division;
  • WRKY;
  • yeast two-hybrid;
  • transcription factor

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Proceduers
  7. Constructs
  8. Acknowledgements
  9. References
  10. Supporting Information

Arabidopsis seed size is regulated by the IKU pathway that includes IKU2 (a leucine-rich repeat kinase) and MINI3 (a WRKY transcription factor). We report the cloning of the IKU1 (At2g35230) gene. iku1 mutants cause reduced endosperm growth and the production of small seeds. IKU1 encodes a protein containing a VQ motif, which is a motif specific to plants. IKU1 is expressed in the early endosperm and its progenitor, the central cell. Restoration of IKU1 function in the endosperm is sufficient to rescue seed size. A genomic construct carrying mutations in the VQ motif failed to complement the iku1 mutation, suggesting an essential role for the VQ motif. IKU1 interacts with MINI3 in the yeast two-hybrid system, consistent with an IKU1 function in the IKU-MINI pathway. Our data support the proposition that endosperm development is an important determinant of seed size.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Proceduers
  7. Constructs
  8. Acknowledgements
  9. References
  10. Supporting Information

In flowering plants, the developing seed contains three structurally distinct components: seed coat, endosperm and embryo. The seed coat is derived from the maternal ovule integuments, and the embryo and the endosperm are produced by the double fertilization event typical of flowering plants. Endosperm development is coordinated with the development of the embryo and seed coat (Garcia et al., 2005), and is initiated by a series of nuclear divisions, without cellularization, resulting in a large multinucleate syncytium (Boisnard-Lorig et al., 2001). After a number of rounds of mitosis without cellularization the syncytium is partitioned by cytokinesis (cellularization), beginning in the region surrounding the embryo and proceeding towards the chalazal region (Sorensen et al., 2002; Olsen, 2004). In many species, including Arabidopsis, the cellular endosperm undergoes cell death and is gradually consumed during seed development, leaving only the peripheral aleurone-like cell layer in mature seeds (Olsen, 2004).

Arabidopsis seed size is controlled by an interaction between endosperm growth and integument proliferation and elongation (Garcia et al., 2005; Jofuku et al., 2005; Ohto et al., 2005). Integument growth is controlled by a pathway where TRANSPARENT TESTA GLABRA 2 (TTG2) appears to play a controlling role (Garcia et al., 2005; Dilkes et al., 2008). Epigenetic regulation mediated by DNA METHYLTRASFERASE 1 is involved in integument proliferation and elongation (FitzGerald et al., 2008). The early growth phase of the endosperm is under the control of the HAIKU (IKU) pathway (Garcia et al., 2003), in which IKU2 encodes a leucine-rich repeat kinase (At3g19700) and MINI3 encodes a WRKY family protein, a putative transcription factor regulating endosperm growth (Luo et al., 2005). Both MINI3 and IKU2 show decreased expression in the loss-of-function mutant iku1. IKU2 expression is reduced in a mini3 background, whereas MINI3 expression is unaltered in the iku2 mutant. MINI3:GUS, a promoter GUS fusion, is upregulated in mini3-1 mutant siliques, indicating that MINI3 is autorepressive. These data suggest that IKU1, IKU2 and MINI3 act in the same pathway of seed development (Luo et al., 2005). IKU2 and MINI3 are both regulated by SHORT HYPOCOTYL UNDER BLUE 1 (SHB1), which binds to their promoters (Zhou et al., 2009).

In this study, we cloned and characterized the function of the IKU1 gene, providing evidence for a role of IKU1 in the regulation of endosperm growth.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Proceduers
  7. Constructs
  8. Acknowledgements
  9. References
  10. Supporting Information

At3g35230 encodes the plant-specific conserved protein IKU1

IKU1 is localized in a 120-kb region of chromosome 2, containing 33 annotated genes (Figure 1a). We sequenced 11 of these 33 coding sequences and identified a mutation in At2g35230, a single nucleotide substitution (G [RIGHTWARDS ARROW] A) forming a stop codon at amino acid residue 143. At2g35230 encodes a VQ motif-containing protein of unknown function (http://www.arabidopsis.org). A segment of At2g35230, spanning from 2145 bp upstream of the ATG start codon to 1065 bp downstream of the stop codon, was introduced into homozygous iku1 plants. Stable transgenic plants derived from T1 seeds produced both wild-type sized and small seeds in the T2 generation, mostly in a 3:1 ratio (Tables 1 and S4), indicating that the iku1 mutant phenotype was complemented by the wild-type At2g35230 sequence, and confirming that At2g35230 encodes IKU1. (More transgenic plants showing complementation are listed in Table S4.)

image

Figure 1.  Cloning of the IKU1 gene. (a) Detailed mapping of the iku1 locus. Small seeds were isolated from F2 seeds of iku1 (in Landsberg erecta, Ler) crossed with Columbia (Col). DNA from 1750 plants (about 3300 chromosomes) producing small seeds was PCR-amplified using markers around the iku1 mutation (Garcia et al., 2003). The numbers of genetic crossovers between each marker and the iku1 mutation are marked, and indicate that iku1 is located in a region of 120 kb with 33 annotated genes. In total, 11 genes were sequenced. In iku1, At2g35230 shows a mutation from G to A that introduces a stop codon. Other sequenced genes did not show any mutation. (b,c) Complementation of the iku1 small-seed phenotype using the At2g35230 genomic sequence. (b) T1 seeds contain about 0.5% wild-type size seeds after dipping iku1 homozygous flowers in Agrobacterium carrying the At2g35230 genomic sequence. (c) Seeds produced by a stable T1 transgenic plant of iku1 dipped with At2g35230 showing segregation of wild-type size seeds and small seeds at a ratio of 3:1. Arrows show the seeds of wild-type size. Scale bars: 1 mm.

Download figure to PowerPoint

Table 1.   T1 plants produced wild-type (WT) sized and small seeds when iku1 was transformed with the At2g35230 genomic sequence
T1 seed phenotypeNo. of seeds for germination on selective mediumNo. of surviving plantsProgeny analysis
WT-sizedSmallχ2Expected ratio
  1. χ2(0.05) = 3.841.

Wild-type sized10041 (4 plants for analysis)54140.7063:1
69220.033
60200
30110.073
Small∼10 00020 (3 plants for analysis)601513:1
33140.574
16090.24715:1

We have identified regions along the IKU1 sequence that are conserved in homologues in other plant species, suggesting specific functions of these motifs (Figure 2). IKU1 contains a conserved short VQ motif FXhVQChTG (pfam05678), where X is any amino acid and h is a hydrophobic amino acid (Morikawa et al., 2002; Andreasson et al., 2005), located between 41 and 71 aa (http://www.arabidopsis.org) (Figure 2). An EQRP sequence is repeated five times between 177 and 221 aa, followed by three repeats of amino acid residues SQ and four repeats of PQ. The IKU1 gene does not contain an intron. Searches for protein sequence homology suggest that VQ domain proteins similar to IKU1 are found only in plants (Figures 2 and 3). There is weak similarity throughout the sequence between IKU1 and several proteins from the moss Physcomitrella patens. In flowering plants, a number of proteins show a high degree of similarity to IKU1 (At2g35230). IKU1 (At2g35230) forms a clade with At1g32610 and similar proteins from Populus, Castor bean (Ricinus communis), cotton (Gossypium hirsutum) and grape (Vitis vinifera) (Figure 3).

image

Figure 2.  Sequence alignment of 13 proteins similar to IKU1 identified in plants using TBLASTN. As well as the conserved VQ motif region, there are other regions that are conserved. The mutation that introduces a stop codon to iku1 is indicated by an arrow. The positions of mutations introduced into the constructs for complementation have also been indicated.

Download figure to PowerPoint

image

Figure 3.  Phylogenetic tree of IKU1-like proteins. IKU1 (At2g35230) is grouped with At1g32610, and then forms a bigger clade with proteins from Populus, Castor bean, cotton, and grape. The proteins from maize, sorghum and rice are grouped together, and form a bigger clade with proteins from dicotyledonous species.

Download figure to PowerPoint

There are at least 34 genes encoding proteins with a VQ motif in the Arabidopsis genome (http://www.arabidopsis.org). These proteins show similarity to the four IKU1-like proteins, only in the VQ motif region. In rice (Oryza sativa), Populus trichocarpa, maize (Zea mays), Sorghum, Castor bean, grape, and moss, there are also proteins containing the VQ motif that do not have sequence similarity to the other regions of IKU1, suggesting that the VQ motif is important in the function of these different proteins.

The VQ motif is important for the function of IKU1

To test the function of the VQ motif and other conserved motifs of IKU1 in seed development, we introduced constructs bearing mutations in each of the conserved regions into the iku1 mutant line, and screened for wild-type sized seeds. The T1 seeds of iku1 transformed with the VQ mutant construct did not include any wild-type sized seeds (Figure 4, column 2; Table S4). We replaced IVQQ in the VQ motif with EDLE (Figures 2 and 4), and the transgenic plants also produced only small seeds, indicating that the construct with the mutation in the VQ motif was not able to rescue the iku1 mutant seed phenotype. In contrast, when we introduced mutations and a deletion into three other conserved regions (12–15, 154–158 and 328–340 aa) (Figure 4, column 1, 3, 5; Table S4), transformation restored the wild-type size in a proportion of T1 seeds. Following selection, the next generation transgenic plants produced both wild-type sized seeds and small seeds, indicating that these constructs were able to complement the iku1 seed phenotype. When the repeat region 177–221 aa was deleted, the iku1 small seed phenotype was rescued by the construct, suggesting that this repeat region is also not important for IKU1 function (Figure 4, column 4; Table S4). Similarly, when we replaced PSQG with EDLE spanning 369–372 aa in a region that is not conserved between IKU1 and other similar proteins, this construct was able to complement the iku1 mutation (Figure 4, column 6; Table S4). We conclude that among these conserved regions only the VQ motif is important for the seed development function of IKU1.

image

Figure 4.  Functional domain analysis of IKU1. (a) IKU1 N terminus mutation: VNRI (12–15 aa) was replaced with ELLL. (b) VQ motif mutation: IVQQ (58–61 aa) was replaced with EDLE. No complementation was observed. (c) VQ downstream mutation: YMRYL (154–158 aa) was replaced with ELLL. (d) Repeat region deletion: an amino acid repeat region (177–221 aa) was deleted. (e) IKU1 C terminus deletion: a conserved region (328–340 aa) was deleted. (f) Non-conserved region replacement: four non-conserved amino acids (PSQG) were replaced with EDLE. Following floral dipping of the iku1 mutant, all constructs except that carrying the mutation in VQ motif, produced T1 seeds containing wild-type sized and small seeds; the stable transgenic plants also produced segregating seeds. Arrows show the big seeds in T1 seeds. Scale bars: 1 mm.

Download figure to PowerPoint

IKU1 expressed in the endosperm directly regulates seed size

Transcriptome analysis of the three fractions (endosperm, embryo and seed coat) of dissected seeds indicates that IKU1 is expressed in endosperm during the early stages of seed development [up to the heart stage embryo; Arabidopsis expression data set (GSE11262) from NCBI, http://www.ncbi.nlm.nih.gov]. An IKU1 genomic sequence with a GFP tag fused at the N terminus (PIKU1:GFP:IKU1) under the control of the IKU1 promoter was introduced into the iku1 mutant background. Some of the T1 seeds were wild-type sized. T2 seeds from the subsequent transgenic plants showed complementation with a 3:1 or 15:1 segregation ratio of wild-type sized to iku1 seeds, indicating that the GFP:IKU1 protein fusion expressed under the IKU1 promoter was functional (Figure 5k; Table S4). GFP:IKU1 was expressed both in the unfused polar nuclei and in the fused polar nucleus in the central cell before fertilization (Figure 5a,b), and in the syncytial endosperm after fertilization (Figure 5c–h). Following self-fertilization of transgenic plants carrying a single copy of the GFP:IKU1 construct, GFP fluorescence was observed in three-quarters of the seed population, suggesting that IKU1 is expressed from both parental alleles, consistent with the IKU1 gene having sporophytic expression (Garcia et al., 2003). Furthermore, GFP expression was observed in wild-type ovules pollinated by GFP:IKU1 pollen (Figure S1C). GFP:IKU1 was localized in the endosperm nuclei (Figure 5c–h) and nuclear localization was confirmed in onion epidermal cells (Figure S1A). The green fluorescence signal detected throughout plant tissues probably reflects either background autofluorescence or a low level of GFP:IKU1 fusion expression that does not localize functionally to the nucleus. We conclude that IKU1 is expressed preferentially in the endosperm, and localizes to the nuclei of the syncytial endosperm.

image

Figure 5.  Expression of IKU1. (a,b) The GFP protein was localized to the nucleus and expressed in the unfused central cell nuclei and fused nucleus before fertilization. (c–h) GFP fluorescence was observed in early endosperm nuclei: some ovules have GFP in the cells of the seed coat. (i) An iku1 ovule directly transformed by a PIKU1:GFP:IKU1 construct showing GFP in two endosperm nuclei. (j) An ovule directly after being transformed showing dividing endosperm nuclei with GFP. (k) Seeds from a stable T1 transgenic plant of iku1 transformed with the PIKU1:GFP:IKU1 construct, showing the segregation of wild-type size and small seeds. Arrows show a normal sized and a small seed in (k) Scale bars: 0.05 mm in (a–j); 1 mm (k).

Download figure to PowerPoint

Among T1 seeds harvested from the iku1 plants transformed with the At2g35230 genomic sequence construct there were approximately 0.5% seeds of wild-type size (Figure 1b). Non-transformed iku1 plants produced only small seeds. We also saw wild-type sized T1 seeds when the iku1 plants were transformed with the PIKU1:GFP:IKU1 construct, and observed 7 out of 10 000 developing seeds showing GFP expression in the endosperm (Figure 5i,j). These observations suggest that endosperm had been transformed in these T1 seeds. About 40% of plants derived from the T1 seeds of wild-type size produced both large and small T2 seeds. These plants were carrying the transgene, as proved by progeny testing on selective media, suggesting that the embryos had also been transformed (Table 1). However, 60% of the T1 plants, although derived from wild-type sized seeds, produced only small T2 seeds, which did not carry the transgene. This indicates that in these seeds the endosperm but not the embryo had been transformed, and that the transformed endosperm is able to restore wild-type seed size (Table 1). The complementation of the iku1 seed phenotype shows that IKU1 functions in endosperm, and directly regulates seed size.

In order to test whether any T1 small seeds contain only transformed embryos, we germinated about 10 000 on selective medium. There were a small number of transgenic survivors (20 plants), these produced both wild-type sized and small seeds, and three plants were used to check the segregation of seed phenotypes. These plants produced both wild-type sized seeds and small seeds (in a 3:1 or 15:1 ratio), showing the complementation of the iku1 phenotype and the transformation of the embryo (Table 1). All the small T1 seeds without a transgene resulted in plants producing only small seeds. These results suggest that transformation of the embryo alone does not restore seed size in T1 seeds. T-DNA transformation with the floral-dipping method targets the female gametophyte and either the central cell (the endosperm progenitor) or the egg cell (the embryo progenitor) (Bechtold et al., 2003). We concluded that the endosperm had been transiently transformed to restore seed size, and that the control of seed size by IKU1 originates from its function in the endosperm and not in the embryo.

IKU1 and MINI3 interact in both the yeast two-hybrid system and the bimolecular fluorescence complementation (BiFC) assay

The similar seed phenotypes of iku1, mini3 and the mini3/iku1 double mutant suggest that the genes operate in the same pathway (Figure S2). MINI3 has an overlapping expression pattern with IKU1 in early endosperm (Luo et al., 2005), and both proteins are localized to the nucleus (Figure S1A,B). We investigated if IKU1 and MINI3 interact physically using the yeast two-hybrid system. Yeast, co-transformed with pACT-MINI3 (PREY) and pAS-IKU1 (BAIT), grew without added histidine, indicating that the MINI3 polypeptide interacts with the IKU1 protein (Figure 6a). A deleted IKU1 (deletion from 256 aa to the end of the protein) in the pAS vector showed auto-activation, as did a C-terminal MINI3 construct (deletion of 1–200 aa at the N terminus) fused to pAS. Because of the auto-activation we could not identify the interacting domain of IKU1. However, an N-terminal fragment of MINI3 fused to pAS (without auto-activation) interacted with IKU1 fused to the pACT domain (Figure 6b), suggesting that the N terminus of MINI3 (which does not contain the WRKY domain) is important for the interaction with IKU1.

image

Figure 6.  MINI3 and IKU1 interact in a yeast two-hybrid system. (a) When yeast was co-transformed with MINI3 fused to an activation domain and IKU1 fused to a DNA binding domain, yeast was able to grow on selective medium. Controls show limited growth on the same plate. (b) When yeast was co-transformed with a C-terminus-deleted MINI3 fused to a DNA binding domain and IKU1 fused to an activation domain, yeast was able to grow on selective medium. Controls showed limited growth in the same plate.

Download figure to PowerPoint

The association of the two proteins was also tested in a BiFC (split EYFP) assay (Citovsky et al., 2006). We detected EYFP signals in the onion epidermal cells when two constructs, 35S:MINI3:cEYFP and 35S:nEYFP:IKU1, were co-delivered through a particle system (Figure S3Ba). No EYFP signal was found with the combination of the original BiFC vector (cEYFP/pUGW2) and the construct 35S:nEYFP:IKU1 as a negative control (Figure S3Bb). To test the specificity of interaction between IKU1 and MINI3, we used a deleted IKU1 (deletion of 1–151 aa at the N terminus, including the VQ motif) fused to nEYFP. No EYFP signal was observed when the MINI3:cEYFP fusion and the nEYFP:IKU1Δ1–151 fusion were co-delivered (Figure S3Bc), suggesting that the N terminus of IKU1 is important for the interaction between IKU1 and MINI3. When we replaced nEYFP:IKU1Δ1–151 with the nEYFP:GUS fusion, again, no positive signal was detected (Figure S3Bd). No positive signals were found when the two original vectors, cEYFP/pUGW2 and nEYFP/pUGW0, were co-expressed (Figure S3Bf). EYFP signals were observed in the combination of the BiFC vector (nEYFP/pUGW0) and the construct 35S:MINI3:cEYFP (Figure S3Be). There may have been a nonspecific association between the MINI3:cEYFP fusion and nEYFP in the original vector nEYFP/pUGW0.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Proceduers
  7. Constructs
  8. Acknowledgements
  9. References
  10. Supporting Information

We have identified IKU1 as At2g35230, encoding a VQ motif protein, and have shown that the VQ motif is essential for IKU1 function. There are at least 34 Arabidopsis VQ motif-containing proteins: only three have sequences homologous with IKU1 outside the VQ motif. The seed-specific phenotype of the iku1 mutant suggests that it functions in seed development, and that the other three genes do not function redundantly with IKU1 in the seed. This is supported by the lack of expression of these three genes in the endosperm, whereas IKU1 is expressed in the endosperm of the developing seed before cellularization (Arabidopsis expression data set GSE11262). Through the complementation of the iku1 phenotype with genomic sequences with or without a GFP tag, we found that wild-type IKU1 is able to restore seed size. Mutations in specific regions of the IKU1 protein indicated an important function of the VQ motif, and only this motif, in endosperm growth.

The endosperm and embryo can be transformed independently by Agrobacterium using the floral-dipping technique (Bechtold et al., 2003). We used this ability to complement the loss of function iku1 mutation in the endosperm only. Wild-type sized T1 seeds are produced when the endosperm of the iku1 mutant is transformed with the IKU1 genomic sequence, but T1 seeds with only the embryos transformed remained small. The direct transformation of endosperm in T1 seeds was visualized by GFP expression of a PIKU1:GPF:IKU1 construct. IKU1 is preferentially expressed in the early endosperm before cellularization, and the only phenotype shown by the iku1 mutant is small seed size resulting from reduced endosperm proliferation (Garcia et al., 2003). IKU2 and MINI3 are regulated by IKU1, and also function in the early endosperm (Luo et al., 2005). This observation further emphasizes that the IKU-MINI pathway acts in an endosperm-autonomous manner, and that the endosperm is the key determinant of seed size.

The seed phenotypes of the double mutants iku1/mini3, iku1/iku2 (Garcia et al., 2003) and iku2/mini3 (Luo et al., 2005) are the same as the single mutants, suggesting that these genes operate in the same pathway. IKU1 and MINI3 are both localized to the nucleus, and are expressed at the same stage of early developing endosperm before cellularization (Luo et al., 2005). IKU2 has low expression but appears to be endosperm specific. Using the yeast two-hybrid system and a BiFC assay, we demonstrated that MINI3 (a WRKY protein) and IKU1 are likely to interact. Two other WRKY proteins interact with a VQ motif protein, and the VQ motif has been shown to be important for the physical association (Andreasson et al., 2005). Recent results have shown that SHB1, although involved in other developmental pathways, regulates seed size by binding to the promoters of MINI3 and IKU2 (Zhou et al., 2009). SHB1 may serve as an upstream modulator involved in the regulation of seed development, perhaps via the IKU-MINI pathway in the endosperm.

The yeast two-hybrid data indicate that IKU1 and MINI3 form a complex and regulate the downstream gene IKU2: mutations in IKU1 or MINI3 result in reduced IKU2 expression and cause small seed formation. We propose that the genes of the IKU–MINI pathway have important functions during normal seed growth via the expression of MINI3, IKU2 and IKU1 in early endosperm development (Luo et al., 2005). Mutation in IKU1 may result in the failure of the IKU1–MINI3 complex to form, and, as a consequence, the unassociated MINI3 protein represses the expression of the MINI3 gene by a feedback mechanism (Luo et al., 2005). Our results show that the IKU–MINI pathway is an important developmental control affecting early endosperm growth and seed size.

Experimental Proceduers

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Proceduers
  7. Constructs
  8. Acknowledgements
  9. References
  10. Supporting Information

Material and growth conditions

Columbia (Col), Landsberg erecta (Ler), mini3 (Luo et al., 2005), iku1 (Garcia et al., 2003) and iku1 mini3 were used in this study. iku1 mini3 double mutants were constructed especially for this study. All plants were grown in pots (13 cm in diameter) with compost (Debco Seed Raising Mix, http://www.debco.com.au/products/potting/seed.php) under continuous artificial light at 20°C in CSIRO-designed growth chambers. The artificial light was achieved by using an incandescent light source producing a fluence of 150 μE (1 E = 1 mol of photons). Primary transformants were selected on MS medium supplemented with 50 μg ml−1 Kanamycin.

Mapping and sequencing the genomic region covering the IKU1 locus

IKU1 had previously been located 1.6 cM south of the marker Cop1a (Garcia et al., 2003). We designed indel markers (Table S1) between Col and Ler based on the sequence information of BACs T14G11, T31E10 F19I3, T4C15, T32F12 and T20F21 (http://www.arabidopsis.org). Primers were used for genotyping F2 plants derived from small seeds from a cross between iku1 (in the Ler background) and Col. The PCR products from Ler and Col have a minimum 8-bp difference in length, which can be visualized on a 4% agarose gel. In total, around 1750 small seeds were isolated from the F2 seeds of iku1 crossed with Col. The plants derived from those seeds should be homozygous iku1 iku1. Most of the plants derived from small seeds had a Ler genotype at marker T14G11 64K, and only seven were heterozygote. No heterozygous plants were detected by marker T4C15 76K, indicating that this marker was very close to the iku1 mutation. Other markers between T14G11 64K and T4C15 76K could detect different numbers of heterozygous plants (Figure 1). Another two markers, T4C15 13K and T32F12 22K, detected more heterozygotes, indicating that IKU1 was located between F19I3 62K and T4C15 13K (Figure 1). This region is about 120 kb, with 33 annotated genes. We sequenced the coding region of 11 genes in this region.

Transformations in plants and onion cells

Arabidopsis was transformed using the floral-dip technique (Clough and Bent, 1998). Transformation for transient expression studies in onion epidermal peels was performed with the BIO-RAD PDS-1000/He System (Bio-Rad, http://www.bio-rad.com; method refined by Upadhyaya et al., 1998). For BiFC experiments, various plasmids encoding cEYFP and nEYFP fusion proteins were mixed at a 1:1 (w/w) ratio and adsorbed onto gold particles (Citovasky et al., 2006).

Phylogenetic analysis

The amino acid sequences containing a VQ motif were obtained from the NCBI database (http://www.ncbi.nhn.nih.gov). mega 4.1 (http://www.megasoftware.net) was used to make sequence alignments and construct phylogenetic trees using the neighbour-joining method. The protein IDs or accession numbers are indicated in the phylogenetic tree.

Microscopy

Mature T1 or T2 seeds of transformed plants were photographed and captured as digital images under a Leica DMR upright microscope (http://www.leica.com). GFP images were captured using Zeiss AxioImager M1 fluorescence microscope. EYFP images were captured using a Leica TCS SP2 confocal microscope.

Constructs

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Proceduers
  7. Constructs
  8. Acknowledgements
  9. References
  10. Supporting Information

Complementation of iku1 with an At2g35230 genomic sequence

To create a genomic IKU1 construct for complementation, we amplified At2g35230 using forward 5′-ACGCGTCGACTTTTTGTGCAGATAAA-ACAAAACA and reverse 5′-TGCTCTAGACCATGACCGAATCAGAAATGT primers. The PCR product was cloned into a pGEM-T Easy vector (Promega, http://www.promega.com; note that this product was used as the template DNA for PCR when we generated mutations in the IKU1 gene and constructed PIKU1:GFP:IKU1), and then transferred into pART-27 using a NotI digest (Fermentas, http://www.fermentas.com). This construct was introduced to the iku1 mutants by floral dipping (Clough and Bent, 1998). The resulting T1 seeds from the iku1 mutants were scored under microscopy for the evaluation of seed size.

Construction of fusion proteins using a 35S:GFP vector (35S:GFP:IKU1 and 35S:GFP:MINI3)

Using primers attB1, 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTTGGATAGGCCTAGACAAAATGATC, and attB2, 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTCTAGTAATCATTCCATCTTGGACTT, we generated a PCR product of IKU1 cDNA and then cloned the PCR fragment into pDONR201 using the BP reaction (Invitrogen, http://www.invitrogen.com). The cloned IKU1 fragment was then inserted between 35S and GFP in the GFP GW N(pUC119) vector (kindly provided by Ben Trevaskis, CSIRO) using the LR reaction (Invitrogen, http://www.invitrogen.com). The vector that hosted the 35S:GFP:IKU1 construct was used for transient GFP assays in onion epidermis cells. Similarly, the MINI3 cDNA fragment generated using primers attB1, 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTTCGAAATGACGTCGTATTGGGCT, and attB2, 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTCTACATGTCGACACCAAACTTAA, was cloned into the GFP GW N(pUC119) vector for transient GFP assays.

The above IKU1 PCR product with attB1 and attB2 sites (Clontech, http://www.clontech.com) was also cloned into a modified T easy vector with attP sites (ampicillin resistant, kindly provided by Chris Helliwell, CSIRO), and subsequently cloned into BinGFP GW N(Pbi101) vector (kindly provided by Ben Trevaskis, CSIRO) using the LR reaction. This vector was used for constructing an IKU1 GFP fusion driven by the IKU1 native promoter (see next section).

IKU1-GFP fusion driven by the IKU1 native promoter (PIKU1:GFP:IKU1)

A fragment was obtained by PCR amplification from the previous pGEM-T Easy vector with the At2g35230 genomic sequence as the template DNA (see the section for constructing the vector for the iku1 complementation). The primers were 5′-CTGTACGATGGTAGCAGCTCCCGCTCGAGCTCCACAAAAATCCAATACCT (XhoI) and 5′-TCTACCACCACCCGGGTTAGT (Sma I). This fragment contains the original vector sequence, the whole IKU1 promoter, and part of the IKU1 coding sequence downstream of the SmaI site. Another fragment containing GFP and part of IKU1 was amplified using the previous construct [35S:GFP: IKU1 in vector BinGFP GW N(Pbi101)] as the template. The primers were 5′-CTGTACGATGGTAGCAGCTCCCGCTCGAGATGACCCGTGGTTCTCATCAC (XhoI) and 5′-ACTAACCCGGGTGGTGGTAGA (SmaI). The two fragments were digested with XhoI and SmaI, and then fused together. The resulting construct contained the pGEM-T Easy vector backbone and GFP fusion with IKU1 driven by the IKU1 native promoter, of which PIKU1:GFP:IKU1 was transferred to the pART27 vector for plant transformation.

Introduction of mutations into constructs for complementation

The pGEM-T Easy vector that carries the IKU1 genomic sequence was used as a template for PCR amplification. The primers (Table S2) were designed to re-amplify the whole plasmid, and allowed for introducing mutations into IKU1 because the primers had altered sequences (underlined). The PCR products were digested by XhoI. The self-ligated plasmids were then amplified and sequenced. The mutated IKU1 genomic sequences were introduced into pART27 with a NotI digest. The first pair of primers replaced IVQQ (58–61 aa) in the VQ motif with EDLE, the second pair resulted in the deletion of 45 amino residues (177–221 aa) containing repeats, the third pair replaced the VNRI (12–15 aa) in the conserved region of the N terminus with ELLL, the fourth pair resulted in the deletion of 12 amino residues (328–340 aa) in the conserved region of the IKU1 C terminus, the fifth pair of primers replaced YMRYL (154–158 aa) downstream of the VQ motif with ELLL and the last pair replaced the non-conserved PSQG (369–372 aa) with EDLE (Figures 2 and 4). We tried to maximize the chance of losing IKU1 function by substituting amino acids with new amino acids with different properties, such as replacing acidic amino acids with basic ones. The positions of amino acids are the numbers in the IKU1 protein sequence (At2g35230; http://www.arabidopsis.org).

Constructs for the BiFC assay (split EYFP)

The vectors (nEYFP/pUGW0 and cEYFP/pUGW2) used for the BiFC assay were kindly provided by Tsuyoshi Nakagawa, Shimane University, Japan (http://podb.nibb.ac.jp/Organellome/bin/browseFunctionalPDF.php?downloadFunctional=BiFC%20vector_Nakagawa_0201.pdf). The constructs for the BiFC assay were obtained using the BP and LR reactions as follows (Table S3). With the cDNAs as templates, the ORF of IKU1 or MINI3 was PCR amplified using the primers listed in Table S3. Primers contained either an attB1 or an attB2 site. Using the BP reaction, the IKU1 or MINI3 fragment was cloned into pDONR201. Using the LR reaction, the IKU1 or MINI3 was cloned into the BiFC vectors, which contain the attR sites (Figure S3; Table S3). The resulting plasmid contains a recombinant gene, which was confirmed by sequencing. 35S:nEYFP:GUS was generated with an LR reaction using the vector nEYFP/pUGW0 and pDONR:GUS (kindly provided by Masumi Robertson, CSIRO). 35S: nEYFP:IKU1Δ1–151, 35S:nEYFP:GUS and the two original BiFC vectors (nEYFP/pUGW0 and cEYFP/pUGW2) were used as negative controls. 35S:nEYFP: IKU1Δ1–151 encodes a recombinant protein of nEYFP fused to a deleted IKU1 (deletion of 1–151 aa at the N terminus of IKU1, including the VQ motif).

Yeast strains, transformation, and two-hybrid constructs and assays

The lithium acetate procedure (James et al., 1996) was used to transform yeast strain PJ69-4A (MATa, trp1-901, leu23112, Ura352, his3–200, gal4Δ, gal80Δ, GAL2-ADE2, LYS2:GAL1-HIS3, met2:GAL7-lacZ; Rose et al., 1990; Gietz and Woods, 1994).

The Matchmaker two-hybrid system was used (Clontech). Using a cDNA clone as a template, a full-length MINI3 cDNA clone was amplified by PCR. The primers are 5′-GGGACAAGTTTGTACAAAAAAGCAGGCTTCGAAATGACGTCGTATTGGGCT and 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTTCAGGCGTTCCATGATCCCATC, containing either an attB1 or an attB2 site. The PCR product was cloned into pDONR201 using the BP reaction. Then the subsequent MINI3 hosted in pDONR201 was cloned using the LR reaction into a modified transcriptional activation domain (AD) vector pACT2, which contains attR sites. We cloned a C terminus deleted MINI3 fragment into the DNA-binding domain (DBD) vector pAS2 using primers: 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTTCGAAATGACGTCGTATTGGGCT and 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTGTCGCTTTCCATCTGAAGTATG.

A full-length IKU1 cDNA clone was amplified by PCR using primers 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTTGGATAGGCCTAGACAAAATGATC and 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTGTAGTAATCATTCCATCTTGGACTTG. The PCR product was cloned into pDONR201 using the BP reaction. Then the IKU1 in pDONR201 was cloned using the LR reaction into the modified pACT2 and pAS2 vectors, which contain attR sites.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Proceduers
  7. Constructs
  8. Acknowledgements
  9. References
  10. Supporting Information

AW was funded by a PhD grant from Yangzhou University and conducted these experiments at CSIRO. DG was funded by the Ecole Normale Superieure and a PhD grant from the Institut National de la Recherche Agronomique. The project was initiated at the Ecole Normale Superieure de Lyon (mutant isolation and map-based cloning). KF is funded by Yangzhou University. HZ is funded by Sichuan Agricultural University. FB is funded by Temasek Life Science Laboratory. ML, AC, JP and ED are funded by CSIRO. We thank Rosemary White, Carl Davies and Mark Talbot for their assistance in image processing, and Narayaya Upadhyaya and Kerrie Ramm for their assistance in onion cell transformation.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Proceduers
  7. Constructs
  8. Acknowledgements
  9. References
  10. Supporting Information
  • Andreasson, E., Jenkins, T., Brodersen, P. et al. (2005) The MAP kinase substrate MKS1 is a regulator of plant defence responses. EMBO J. 24, 25792589.
  • Bechtold, N., Jolivet, S., Voisin, R. and Pelletier, G. (2003) The endosperm and the embryo of Arabidopsis thaliana are independently transformed through infiltration by Agrobacterium tumefaciens. Transgenic Res. 12(5), 509517.
  • Boisnard-Lorig, C., Colon-Carmona, A., Bauch, M., Hodge, S., Doerner, P., Bancharel, E., Dumas, C., Haseloff, J. and Berger, F. (2001) Dynamic analyses of the expression of the HISTONE:YFP fusion protein in Arabidopsis show that syncytial endosperm is divided in mitotic domains. Plant Cell 13, 495509.
  • Citovsky, V., Lee, L.Y., Vyas, S., Glick, E., Chen, M.H., Vainstein, A., Gafni, Y., Gelvin, S.B. and Tzfira, T. (2006) Subcellular localization of interacting proteins by bimolecular fluorescence complementation in Planta. J. Mol. Biol. 36, 11201131.
  • Clough, S.J. and Bent, A.F. (1998) Flower dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J. 16, 735743.
  • Dilkes, B.P., Spielman, M., Weizbauer, R., Watson, B., Burkart-Waco, D., Scott, R.J. and Comai, L. (2008) The maternally expressed WRKY transcription factor TTG2 controls lethality in interploidy crosses of Arabidopsis. PLoS Biol. 6(12), 27072720.
  • FitzGerald, J., Luo, M., Chaudhury, A. and Berger, F. (2008) DNA methylation causes predominant maternal controls of plant embryo growth. PLoS One 3(5), e2298.
  • Garcia, D., Saingery, V., Chambrier, P., Mayer, U., Jurgens, G. and Berger, F. (2003) Arabidopsis haiku mutants reveal new controls of seed size by endosperm. Plant Physiol. 131, 16611670.
  • Garcia, D., FitzGerald, J.N. and Berger, F. (2005) Maternal control of integument cell elongation and zygotic control of endosperm growth are coordinated to determine seed size in Arabidopsis. Plant Cell 17, 5260.
  • Gietz, R.D. and Woods, R.A. (1994) Molecular Genetics of Yeast: practical Approaches. London: Oxford Univ. Press, pp. 121134.
  • James, P., Halladay, J. and Craig, E.A. (1996) Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast. Genetics 144(4), 14251436.
  • Jofuku, K.D., Omidyar, P.K., Gee, Z. and Okamuro, J.K. (2005) Control of seed mass and seed yield by the floral homeotic gene APETALA2. Proc. Natl Acad. Sci. USA, 102(8), 31173122.
  • Luo, M., Dennis, E.S., Berger, F., Peacock, W.J. and Chaudhury, A. (2005) MINISEED3 (MINI3), a WRKY family gene, and HAIKU2 (IKU2), a leucine-rich repeat (LRR) KINASE gene, are regulators of seed size in Arabidopsis. Proc. Natl Acad. Sci. USA, 102, 1753117536.
  • Morikawa, K., Shiina, T., Murakami, S. and Toyoshima, Y. (2002) Novel nuclear-encoded proteins interacting with a plastid sigma factor, Sig1, in Arabidopsis thaliana. FEBS Lett. 514, 300304.
  • Ohto, M.A., Fischer, R.L., Goldberg, R.B., Nakamura, K. and Harada, J.J. (2005) Control of seed mass by APETALA2. Proc. Natl Acad. Sci. USA, 102(8), 31233128.
  • Olsen, O.A. (2004) Nuclear endosperm development in cereals and Arabidopsis thaliana. Plant Cell 16(Suppl.), S214S227.
  • Rose, M.D., Windston, F. and Hieter, P. (1990) Methods in Yeast Genetics. Plainview, NY: Cold Spring Harbor Lab.
  • Sorensen, M.B., Mayer, U., Lukowitz, W., Robert, H., Chambrier, P., Jurgens, G., Somerville, C., Lepiniec, L. and Berger, F. (2002) Cellularisation in the endosperm of Arabidopsis thaliana is coupled to mitosis and shares multiple components with cytokinesis. Development 129, 55675576.
  • Upadhyaya, N.M., Li, Z., Ramm, K., Gaudron, J., Waterhouse, P.M., Wang, M.B., Kositratana, W. and Gong, Z.X. (1998) Production of Transgenic rice with rice ragged stunt virus synthetic resistance genes. In: Proc. Int. Symp. Biotechnology Tropical & Subtropical Species. Ed RA Drew. Acta Hortic. 461, 393400.
  • Zhou, Y., Zhang, X., Kang, X., Zhao, X., Zhang, X. and Ni, M. (2009) SHORT HYPOCOTYL UNDER BLUE1 associates with MINISEED3 and HAIKU2 promoters in vivo to regulate Arabidopsis seed development. Plant Cell 21, 106117.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Proceduers
  7. Constructs
  8. Acknowledgements
  9. References
  10. Supporting Information

Figure 1. Using an onion transient expression system both GFP:IKU1 and GFP:MINI3 driven by a 35S promoter showed expression in nuclei. a. GFP:IKU1 is nuclear localised with speckles. The formation of speckles within the nuclei of onion cells by GFP:IKU1 fusion protein may be due to the action of the strong 35S promoter which drives GFP:IKU1 expression. In the plants transformed by an IKU1 promoter-driven GFP:IKU1 fusion construct no speckles were found. b. GFP:MINI3 is nuclear localised. c. Developing seeds of a wild type plant pollinated with PIKU1:GFP:IKU1 pollen showed GFP activity. (Bars = 5 μm in a & b, Bar = 0.05mm in c).

Figure 2. Seed weights (μg) of two iku1/mini3 double mutant plants, an iku1 plant, a mini3 plant and a L.er plant.

Figure 3. Assay for MINI3 and IKU1 interaction in a BiFC system. a. Vectors used for BiFC assay. a. Schematic drawing of the original BiFC gateway vectors. Ori, ColE1 origin; Ampr, ampicillin resistant marker; Cmr, chloramphenicol resistant marker. b. Constructs used in a BiFC assay. b. BiFC assay showing a likely interaction between MINI3 and IKU1. a. EYFP signals in the onion epidermal cell nuclei were seen when the vector 35S:MINI3:cEYFP and the vector 35S:nEYFP:IKU1 were co-delivered. b. No EYFP signal was detected in the combination of cEYFP and nEYFP:IKU1, nor in the combination of cEYFP and nEYFP (in f). e. The combination of MINI3:cEYFP and nEYFP showed EYFP signals. c&d. The combination of MINI3:cEYFP and nEYFP:IKU1Δ1-151 (or nEYFP:GUS) did not show EYFP signal. (Bars = 150 μm).

Table 1. Markers for genetic mapping of iku1.

Table 2. Primers for introducing mutations into IKU1.

Table 3. Constructs and primers for BiFC assay.

Table 4. Selected iku1 stable transgenic plants transformed with various genomic sequences.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
TPJ_4271_sm_Legends.doc27KSupporting info item
TPJ_4271_sm_Table-S1.doc37KSupporting info item
TPJ_4271_sm_Table-S2.doc34KSupporting info item
TPJ_4271_sm_Table-S3.doc31KSupporting info item
TPJ_4271_sm_Table-S4.doc61KSupporting info item
TPJ_4271_sm_Figure-S1.tif1892KSupporting info item
TPJ_4271_sm_Figure-S2.tif816KSupporting info item
TPJ_4271_sm_Figure-S3.tif7118KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.