SLOW WALKER3, Encoding a Putative DEAD-box RNA Helicase, is Essential for Female Gametogenesis in Arabidopsis


  • Man Liu,

    Corresponding author
    1. Key Laboratory of Molecular and Developmental Biology, National Centre for Plant Gene Research, Institute of Genetics and Developmental Biology, the Chinese Academy of Sciences, Beijing 100101, China
    2. Graduate University of the Chinese Academy of Sciences, Beijing 100039, China
    Search for more papers by this author
  • Dong-Qiao Shi,

    Corresponding author
    1. Key Laboratory of Molecular and Developmental Biology, National Centre for Plant Gene Research, Institute of Genetics and Developmental Biology, the Chinese Academy of Sciences, Beijing 100101, China
    Search for more papers by this author
  • Li Yuan,

    1. Key Laboratory of Molecular and Developmental Biology, National Centre for Plant Gene Research, Institute of Genetics and Developmental Biology, the Chinese Academy of Sciences, Beijing 100101, China
    2. Graduate University of the Chinese Academy of Sciences, Beijing 100039, China
    Search for more papers by this author
  • Jie Liu,

    1. Key Laboratory of Molecular and Developmental Biology, National Centre for Plant Gene Research, Institute of Genetics and Developmental Biology, the Chinese Academy of Sciences, Beijing 100101, China
    Search for more papers by this author
  • Wei-Cai Yang

    Corresponding author
    1. Key Laboratory of Molecular and Developmental Biology, National Centre for Plant Gene Research, Institute of Genetics and Developmental Biology, the Chinese Academy of Sciences, Beijing 100101, China
    Search for more papers by this author

These authors contributed equally to this work.

Corresponding author
Tel(Fax): +86 10 6480 7761; E-mail:


RNA helicases are adenosine tri-phosphatases that unwind the secondary structures of RNAs and are required in almost any aspect of RNA metabolism. They are highly conserved from prokaryotic to eukaryotic organisms. However, their precise roles in plant physiology and development remain to be clarified. Here we report that the mutation in the gene SLOW WALKER3 (SWA3) results in the slow and retarded progression of mitosis during megagametogenesis in Arabidopsis. SWA3 is a putative RNA helicase of the DEAD-box subfamily. Mutant megagametophyte development is arrested at four- or eight-nucleate stages, furthermore, one of the synergids in about half of the mutant embryo sacs displays abnormal polarity, with its nucleus locating at the chalazal end, instead of the micropylar end in the wild-type. Transmission of the mutation through female gametophytes is severely reduced in swa3. However, a small portion of mutant embryo sacs are able to develop into mature and functional female gametophytes when pollination was postponed. The SWA3 in Arabidopsis is a homolog of Dbp8 in yeast. Dbp8 interacts with Efs2 and is essential for biogenesis of 18S rRNA in yeast. Our data suggest that SWA3 may form a complex with AtEfs2 and take roles in ribosomal biogenesis as RNA helicase during megagametogenesis in Arabidopsis.


The plant life cycle alternates between a diploid sporophytic and a haploid gametophytic phase. Successful development of female gametophytes in higher plants is the consequence of perfect coordination of cell specification, proliferation, and differentiation. The female gametophytes of flowering plants are largely reduced in size and complexity compared with those of the “naked seed” plants, the gymnosperms. The female gametophyte of higher plants is regarded as an ideal system, since the basic and critical developmental steps, such as meiosis, mitosis, determination of cell fate and polarity, nuclear migration and fusion, cell differentiation, cellularization and communication, are all covered during gametogenesis. Moreover, the haploidy of gametophytes makes it possible to investigate the mutation in essential genes, as this cannot be carried in diploid sporophyte cells when the homozygote is not available. Female gametophyte development occurs over two subsequent phases, referred to as megasporogenesis and megagametogenesis. The pattern of female gametophyte development in Arabidopsis represents the Polygonum-type program, which is exhibited by >70% of flowering plants. In Arabidopsis, the archesporial cell differentiates from a hypodermal cell of the nucellus and then turns to megaspore mother cell (MMC). MMC is distinct because of its large size, big nucleus and dense cytoplasm in ovule primodium cells. Four megaspores are produced after the meiosis of MMC and the chalazal-most one, the functional megaspore, will undergo three rounds of mitosis without cytokinesis, giving rise to a coenocyte with eight nuclei. These nuclei are spaced out by a large central vacuole with four nuclei at the chalazal end and the other four at the micropylar end. Then, two nuclei, one from each pole, migrate toward the micropylar half of the embryo sac to form the polar nuclei, and later fuse to each other to form a diploid central cell nucleus or secondary nucleus. Meanwhile, cellularization takes place in the embryo sac and results in the formation of a seven-celled female gametophyte. Finally the mature female gametophyte is featured as four cell types, seven cells, namely two synergids at the micropylar end, an egg next to synergids, a central cell, and three antipodals at the chalazal end (Drews et al. 1998; Grossniklaus and Schneitz 1998; Yang and Sundaresan 2000; Yang et al. 2010).

During the last 10 years, quite a few of mutants and genes that are involved in female gametophyte development were identified (Drews et al. 1998; Grossniklaus and Schneitz 1998; Yang and Sundaresan 2000; Pagnussat et al. 2005; Liu and Qu 2008; Yang et al. 2010). Among these mutants, many showed aberrant mitosis during megagametogenesis. This suggests that mitotic division is crucial to female gametophyte development. These mutants are characterized with their abnormal number and/or location of nuclei in the embryo sac, such as nomega (Kwee and Sundaresan 2003), hadad (hdd) (Moore et al. 1997), prolifera (prl) (Springer et al. 1995; Springer et al. 2000), retinoblastoma related1 (rbr1) (Ebel et al. 2004), anaphase-promoting complex2 (apc2) (Capron et al. 2003), slow walker1 (swa1) (Shi et al. 2005), swa2 (Li et al. 2009), and eostre (Pagnussat et al. 2007). There is no doubt that the production and positioning of the nuclei, such as the accurate schedule of nuclei division, the migration of the nuclei and the formation of polarized embryo sac, are elaborately tuned during the mitotic cell cycle of female gametophyte. Regarding the genes that are involved in gametophytic cell progression, some encode or are linked with cell cycle regulators, for example, APC6/CDC16 (encoded by NOMEGA), APC2, RHF1a (RING-finger E3 ligase) and RHF2a (Liu et al. 2008) are E3 ubiquitin-protein ligases that are targeting cyclins, cyclin-dependent kinase inhibitor or other cell cycle regulators, RBR functions as an important regulator of G1/S transition, and PRL is a subunit of DNA replication licensing factor. However in other mutants that are defective in mitosis during female gametophyte development, such as swa1, swa2, the target genes are involved in rRNA processing or ribosomal biogenesis. This suggests that ribosomal biogenesis is crucial for female gametophyte development in plants.

Recently, the role of RNA processing and ribosome biogenesis has come to light during embryo sac development. The components of the RNA splicing machinery are shown to regulate gametic cell fate. LACHESS (LIS) (Gross-Hardt et al. 2007), a homolog of the yeast splicing factor PRP4, participates in gametic cell fate determination by preventing the differentiation of accessory cells. GFA1/CLO (Coury et al. 2007; Moll et al. 2008; Liu et al. 2009), a gene encoding a putative splicing factor, is essential for the development of female gametophyte and embryo, and is also involved in regulating gametic cell identity. These data imply that RNA processing plays an important role in megagametogenesis in plants.

To further reveal the close link of RNA processing and female gametogenesis, here we report the functional analysis of SWA3, an RNA helicase gene encoding a member of the DEAD-box protein family. Our data show that the mutation in SWA3 gene results in the arrest of female gametophyte development at four- or eight-nucleate stage and irregular polarity of synergid cell in Arabidopsis. The Arabidopsis SWA3 protein, a homolog of yeast Dbp8p, which can interact with Esf2 protein and is essential for biogenesis of the 18S rRNA in yeast, can physically interact with AtEsf2 protein. These results indicate that SWA3 may have similar function in rRNA biogenesis as its yeast counterpart, and thus further implies that rRNA biogenesis is critical to female gametophyte development in Arabidopsis.


Genetic analysis of swa3 mutant

The swa3 mutant was identified from a large scale screening of the gene trap lines, which was generated by the Ds transposon insertion system in Arabidopsis thaliana ecotype Landsberg erecta (Springer et al. 1995; Sundaresan et al. 1995). Since the Ds element harbors an NPTII gene, the reduced transmission efficiency of kanamycin resistance indicates that the mutation is either gametophytic or embryo lethal. The swa3 mutant (since the homozygous mutant cannot be recovered, the swa3 mutant in this paper refers to plants of genotype swa3/SWA3) shows distorted Mendelian segregation and reduced seed set compared with wild type plants (Table 1). In the self-pollinated progeny of the swa3 plants, the segregation ratio of kanamycin-resistant (KanR) to kanamycin-sensitive (KanS) seedlings is 0.95:1 (564:595), instead of the typical 3:1 Mendelian ratio. This suggests that it is likely a gametophytic mutation. Further reciprocal cross between swa3 and wild type plants were carried out. When the wild type plants were pollinated with pollen grains from swa3 anthers, the F1 progeny displayed a KanR:KanS ratio of 0.83:1 (471:566), indicating that swa3 mutation has 83% transmission efficiency through the male. However, the KanR:KanS ratio of F1 seedlings is 19:874 when swa3 stigmas were pollinated with wild type pollen grains, which suggests only 2.2% transmission efficiency through the female. This indicates that swa3 is a female sterile mutant. In the swa3 siliques of 6 or 7 d after pollination (dap), approximately 50% (526:526) of seeds are aborted, compared with the full seed sets in the wild type siliques (Figure 1).

Table 1.  Transmission efficiency of swa3 mutants
Parental genotypes Female × MaleProgenyKanR (expected)Transmission efficiency (%)
  1. Plants were crossed manually. The crossing seeds were collected, and grown on selective Murashige and Skoog (MS) plates with Kanamycin. The results of each group were from three independent plants.

SWA3/SWA3×swa3/SWA347155645.9% (50%)83.2%
swa3/SWA3×SWA3/SWA3 19874 2.1% (50%) 2.2%
swa3/SWA3×swa3/SWA356459548.7% (50%) 
Figure 1.

Seed development in the swa3 mutant.

(A) A wild-type silique with full seed set.
(B) A swa3 silique with aborted seeds. White arrows indicate the aborted seeds.

Alexander staining for viable pollen showed that only 1.1% of pollen grains (n= 1 539) from swa3 plants are non-viable, suggesting that the pollen viability is not affected dramatically in swa3 mutant. Taken together, these data indicate that the mutation has a vital effect on the female, rather than the male gametophyte.

Female gametophyte development is severely affected in the swa3 mutant

Female gametophyte development is a precisely regulated and programmed process. In Arabidopsis, steps of female gametophyte development have been studied intensively and the whole process is divided into seven stages according to cell cycle progression and cellularization in embryo sac (Figure 2A–G) (Christensen et al. 1997). The mature embryo sac is a highly polarized structure: the egg nucleus locates at the chalazal end of the cell and close to the central nucleus, while the vacuole of the egg cell locates at the micropylar end. However, the synergid cells display an opposite polarity to the egg, with the nuclei at the micropylar end and vacuoles at the chalazal end of the cell (Figure 2).

Figure 2.

Phenotype of mutant female gametophyte in swa3 plants.
Wild-type female gametophytes of different stages: one-nucleate stage (FG1, A), two-nucleate stage (FG2, B), late two-nucleate stage with a central vacuole (FG3, C), four-nucleate stage with two nuclei at each pole (FG4, D), eight-nucleate stage with four nuclei at each pole (early stage FG5, E), seven-celled stage with unfused polar nuclei (late FG5, F), and four-celled female gametophyte (FG7, G). HM: Mutant female gametophytes are arrested at FG4 (H), early FG5 with mispositioned nuclei (I), late FG5 (J), late FG5 with mispositioned nuclei (K, L, M). Bar, 10 μm.

To further characterize the mutant phenotype, we observed swa3 gametophytes using confocal laser scanning microscope (CLSM) (Shi et al. 2005). The flowers of stage 12c from swa3 were emasculated, and the pistils were fixed in glutaraldehyde solution 36 h after emasculation. All the wild type embryo sacs are mature and develop to the final stage (stage FG7, Figure 2G) at this moment (Christensen et al. 1997). However, among the 274 ovules from swa3 plants, about half (n= 132) have mature embryo sacs (FG7, four-cell stage) and exhibit no defect compared with the wild-type. This suggests that these ovules represent the wild type ovules in the swa3 siliques. The other half of the female gametophytes show diverse defects from four-nucleate to eight-nucleate stage (stage FG5), meanwhile, some display abnormal morphology of displaced nuclei at the micropylar end. In the “retarded” embryo sacs, 8.4% (n= 12) are at the four-nucleate stage (Figure 2H), suggesting that the third mitosis is disrupted in these female gametophytes. Interestingly, except the slow mitotic progression, these arrested ones have no detectable morphological difference compared with the wild-type of stage FG4. This shows that these mutant ovules are arrested at the FG4 stage. 14.8% (n= 21) and 27.5% (n= 39) mutant female gametophytes are arrested at the early and late FG5 stage (Figure 2I–M), respectively. However, about half of the mutant female gametophytes of eight-nucleate stage show irregular nuclear positioning at the micropylar end compared with the wild-type. Just as showed in Figure 2K and L, only one nucleus is present at the micropylar end in some mutant embryo sacs, other four nuclei is at the central portion of the embryo sac. There might be some defect with the cell fate determination of one of the synergids in these two embryo sacs, since the synergid nucleus is positioned similarly to that of the egg. As in the embryo sac shown in Figure 2M, two nuclei locate at the chalazal pole, and the other six nuclei are present at the micropylar pole, indicating that nuclear positioning and cell polarity are disrupted. These findings suggest that the cell cycle progression, nuclear migration and fusion, and cell polarity of the mutant embryo sacs are disordered, implying that SWA3 plays multiple roles during female gametophyte development.

Synchrony of female gametophyte development is impaired in swa3 pistils

Our previous observation shows that the abnormalities of swa3 female gametophyte development are detectable from stage FG4. The synchrony of female gametophyte development in swa3 mutant pistils was investigated carefully. In wild type plants, female gametophyte development is highly synchronous (Christensen et al. 1997; Shi et al. 2005). A detailed study on the stages of female gametopohyte in the same pistil was carried out. The inflorescences were fixed from different plants, and configurations of developmental stages of megagametophyte in the same pistil are recorded with CLSM. The results are presented in Tables 2 and 3. In the wild-type pistils, the female gametophyte development takes a narrow range of two sequentially developmental stages, which is consistent with the previous reports (Christensen et al. 1997; Shi et al. 2005). The development of the female gametophytes within a pistil is fairly synchronous, and usually one developmental stage is predominant. As shown in Table 3, asynchronous development is found in swa3 pistils. From stage FG4 onward, the female gametophytes within the same pistil are dispersed among four or five stages, and it seems that no stage is predominant. These data indicate that synchrony of female gametophyte development is impaired in swa3, and the defect is detectable from stage FG4, which is consistent with previous observations.

Table 2.  Synchrony of female gametophyte development in wild-type plants
Pistil numberNumber of female gametophytes at development stagesTotal ovules
S131       31
S2 8826     42
S3   527    32
S4   1415   29
S5    339   42
S6    1126 8 45
S7      4181941
S8       14243
Table 3.  Synchrony of female gametophyte development in swa3 mutant
Pistil numberNumber of female gametophytes at development stagesTotal ovules
  1. aFG5 includes eight-nuclei embryo sacs with proper or irregular position of nuclei.

Ss13510      45
Ss2 1528 2    45
Ss3  22215    37
Ss4   8221511  56
Ss5   12516a 2  44
Ss6   114111016 52
Ss7   1 915a 19 44
Ss8   1811a 319 51
Ss9    321a 16 949
Ss10    124a  2045

To investigate whether the “delayed” embryo sacs in swa3 ovules are able to develop into the mature and functional female gametophytes, the postponed pollination experiment (Shi et al. 2005) was performed. Pistils of swa3 were emasculated at floral stage 12c (Smyth et al. 1990), and pollinated with wild type pollen after 12 h, 24 h and 48 h, respectively. The seeds were collected according to different pollination time points, and germinated on MS medium with kanamycin to test the ratio of KanR to KanS. The result indicates that the KanR:KanS ratio is increased with the delay of pollination, which indicates that a small part of the swa3 ovules are able to be fertilized when pollination is delayed (Figure 3): the longer time the pollination is delayed, the higher KanR:KanS ratio will be obtained. This indicates that a small fraction of mutant embryo sacs can develop into functional female gametophytes if given longer time. However, the slow development causes mismatch of immature female gametophyte with mature pollen and the failure of fertilization when naturally pollinated. Since this mutant exhibits similar phenotype as swa1 reported earlier in our lab, we named it as slow walker3 (swa3).

Figure 3.

KanR:KanS ratio of F1 progeny from the delayed pollination test.

Flowers of swa3 plants are emasculated at floral stage 12c and pollinated with pollen at 12 h, 24 h, and 48 h after emasculation, respectively.

Isolation of the SWA3 gene

Thermal asymmetric interlaced polymerase chain reaction (TAIL-PCR) (Liu et al. 1995) was used to obtain the flanking sequence of the Ds insert in swa3 mutant. The flanking sequence cloned reveals that the Ds element is inserted into the third intron of gene AT1G16280 (Figure 4A), a single-copied gene in the Arabidopsis genome. A DNA product of predicted size was obtained from genomic DNA of swa3, but not from that of the wild-type, when gene-specific and Ds primers are selected for the PCR, confirming the Ds insertion that was unveiled with TAIL-PCR. Sequence analysis indicates that the Ds insertion results in a 8-bp duplication at the integration site.

Figure 4.

Molecular features of SWA3 protein.

(A) Diagram showing Ds insertion in the third intron of the SWA3 gene caused an 8 bp duplication. The open and bold boxes show exons, the short lines in between boxes show introns. The nucleotide numbers are consistent with those in the Arabidopsis genome.
(B) Phylogenetic tree of SWA3 with its homologs from other organisms.
(C) Alignment of the SWA3 protein to its homologs with high similarity from other organisms. Identical amino acids are shown with white letters in bold boxes, and similar amino acids are shown with grey boxes. The nine conserved motifs of DEAD box are underlined. Os07g0633550 (rice), Hs143187 (human), Mm388127 (mouse), cel14864 (Caenorhabditis elegans), Dbp8 (S. cerevisiae) and AtSWA3 (Arabidopsis) and AT5G60990 (Arabidopsis).

To verify whether the Ds insertion leads to the phenotype of swa3 mutant, a complementation experiment was carried out. A 7.1 kb genomic sequence, including 2.9 kb fragment upstream from the start codon, the coding frame and 3.1 kb fragment downstream from the stop codon, was cloned into pCAMBIA1300 and introduced into swa3 plants by Agrobacterium-mediated infiltration. Totally, 13 independent transgenic lines were obtained. The seed sets of these lines are restored to 73.3–93.4% of the wild type level, and the progeny of these lines show 1.57–2.84:1 KanR to KanS ratio, which indicates that the distorted ratio in swa3 mutant is partially recovered. Furthermore, two plants homozygous for both the swa3 allele and the introduced transgenic copy were obtained, and they show full seed sets. It indicates that the disruption of AT1G16280 results in the defects of female gametophyte development in swa3, thereafter we refer AT1G16280 as the SWA3 gene.

SWA3 is a DEAD-box RNA helicase and homologous to the yeast splicing factor Dbp8

SWA3 is predicted to encode a protein of 491 amino acid residues with an estimated pI of 8.86 and molecular weight of approximately 54.8 KD. Structure analysis reveals that this protein contains a DEAD domain at N-terminal, which is a major functional domain of DEAD/H box helicase. SWA3 also contains conserved C-terminal domain belonging to the helicase family. All nine classical conserved motifs essential in DEAD protein family (de la Cruz et al. 1999; Rocak and Linder 2004), as well as the core component of RNA helicase (Tanner 2003; Cordin et al. 2004), are present in SWA3 protein (Figure 4C). Generally, these motifs are functional in RNA substrate binding and adenosine tri-phosphate (ATP) binding, or exhibit activity of ATPases. Previous study shows that there are over 53 DEAD-box RNA helicases in the Arabidopsis genome classified as AtRH1 to AtRH53 (Aubourg et al. 1999; Boudet et al. 2001; Mingam et al. 2004). SWA3 is identical to AtRH36 based on the sequence analysis. Sequence analysis indicates that homologs of SWA3 protein show high similarity with each other in eukaryotes, such as rice, mouse, Caenorhabditis elegans, human and yeast (Figure 4B). Dbp8, its homolog in Saccharomyces cerevisiae, shares high identity and similarity to SWA3, with 42% identity and 59% similarity at the amino acid level. In yeast, Dbp8 is a DEAD/H box protein of ATP-dependent RNA helicases, which is required for the synthesis of the 18s rRNA and biogenesis of 40S small ribosomal subunit (Daugeron and Linder 2001). It suggests that SWA3 may have similar function in Arabidopsis.

Many DEAD-box RNA helicases can interact with proteins involved in RNA biogenesis (Linder 2006). It is reported that, in yeast ribosome biogenesis, Esp2 interacts directly with Dbp8, and then binds to pre-rRNAs to mediate ATP hydrolysis (Granneman et al. 2006). The AT3G56510 is the only homolog to Esf2 in the Arabidopsis genome. To verify whether the SWA3 plays a similar role as Dbp8, yeast two-hybrid assay was applied to test whether SWA2 interacts with AtEsf2. Full length SWA3 and AtEsf2 were subcloned into pGBKT7 and pGADT7, respectively. Yeast cells co-transformed with the bait pBD-SWA3 and the prey pAD-AtEsf2 grow well on minimal medium lacking tryptophan, leucine, histidine and adenine, indicating that SWA3 can physically interacts with AtEsf2 in yeast cells (Figure 5A, B). Expression level analysis according to microarray data from GENVESTIGATOR ( shows that the expression pattern of SWA3 is correlated with that of AtEsf2, and they can overlap with each other spatially and temporally (Figure 5C). It means that the two proteins may interact with each other in vivo and involved in pre-rRNA biogenesis.

Figure 5.

Interaction of SWA3 and AtEsf2 in yeast cells.

Yeast cells are transformed with pBD-SWA3 and pGADT7 (sector 1), pBD-SWA3 and pAD-AtEsf2 (sector 2), and pGBKT7 and pAD-AtEsf2 (sector 3), respectively and grow on SD-Trp/-Leu (A) and SD-Trp/-Leu/-His/-Ade (B). Note SWA3 interacts with AtEsf2 in yeast cells (sector 2). (C) Correlation of the expression profiles of SWA3 and AtEsf2 in organs. Data were obtained from Genevestigator ( microarray dataset in this analysis.

SWA3 is expressed constitutively in plant

To further detect the function of SWA3, the expression pattern of SWA3 was examined by semi-quantitative reverse transcription-polymerase chain reaction (RT-PCR) and RNA in situ hybridization experiment. In the RT-PCR experiment, transcript is detected in roots, stems, leaves, inflorescences, siliques and seedlings (Figure 6A). The result shows that the SWA3 gene is expressed strongly in inflorescences and seedlings. This is consistent with the microarray data from GENEVESTIGATOR. With RNA in situ hybridization, strong hybridization signals of the SWA3 gene are detected in all organs examined, including female gametophytes in all developmental stages, embryos at early stage, and pollens (Figure 6B–H), and this is coincident with RT-PCR result.

Figure 6.

Expression pattern of SWA3 revealed by reverse transcription-polymerase chain reaction (RT-PCR) and RNA in situ hybridization.

(A) Expression pattern of SWA3 revealed by RT-PCR analysis. ACTIN2 gene was used as an internal reference.
(B to H) Micrographs showing SWA3 expression in the inflorescence (B), ovule primordium (C), ovule (D), early globular embryo (E and F) and pollen grains (G). (H) No signal was detected with sense probe. Bar, 10 μm.


Regulation of RNA processing and ribosomal biogenesis is critical for female gametogenesis in plants

Regulation of gene activities can be accomplished at transcriptional, translational, and post-translational levels. So far, many genes controlling female gametophyte development have been identified (Liu and Qu 2008; Yang et al. 2010). Female gametophyte is a highly specified structure building up from fundamental steps of cell division and differentiation. The genes involved in megagametogenesis can be classified into two groups, the special group only for female gametophyte development, for example, the genes essential in processes that particularly evolved for female gametophyte structure or function, to some extent, this group is mostly related with differentiation; and the general group for some cellular phases common to many cells in the organism, such as cell cycle, metabolism. These two groups contribute to the complexity of female gametophyte development. The megagametogenesis may be disturbed because of the mutation in the gene of either group. Genes involved in mRNA splicing, pre-rRNA maturation, and ribosomal biogenesis belong to the general group without doubt. However, these genes have become more and more distinctive in female gametophyte development. Quite a few mutations in these genes have been reported in the last 10 years (Chekanova et al. 2002; Christensen et al. 1998; Shi et al. 2005; Bove et al. 2006; Li et al. 2009; Liu et al. 2009). This implies that RNA processing is crucial to plant development. In our lab, several genes that are functional in RNA biogenesis and necessary to normal cell cycle progression during female gametogenesis in Arabidopsis have been isolated (Shi et al. 2005; Li et al. 2009; Liu et al. 2009). The mutants share similar phenotype of slow mitosis progression of the female gametophyte; accordingly, they are named as slow walker (swa). SWA1 encodes a nucleolar protein with six WD repeats and plays a role in rRNA biogenesis in Arabidopsis. The mutation of SWA1 disrupts the processing of 18S pre-RNA and impairs the mitotic division cycles of embryo sac development (Shi et al. 2005). SWA2 is the homologous of NOC1 in yeast and is involved in ribosome export. These data suggest that the progression of cell cycle during female gametophyte development is inevitably linked with the components that are involved in the ribosome biogenesis. This implies that the regulation of ribosomal biogenesis is one of the key aspects for the mitotic progression of the embryo sac in plants.

It is well known that RNA helicases are involved in almost all cellular processes of RNA processing machinery. SWA3 encodes a putative DEAD-box RNA helicase, and the closest homolog of SWA3 with known function is the yeast Dbp8, which is required for ribosomal biogenesis in yeasts (Daugeron and Linder 2001). SWA3 physically interacts with AtESf2 in the yeast two-hybrid system, demonstrating that SWA3 is likely to be involved in ribosomal biogenesis as well. The loss of SWA3 function results in irregular gametic cell specification at the micropylar end and retarded progression of cell cycle. Taken together, we speculate that SWA genes take roles in cell cycle progression during female gametophyte development via ribosomal biogenesis.

In addition, RNA processing has also been shown to play an important role in gametic cell specification and differentiation during embryo sac development (Gross-Hardt et al. 2007; Moll et al. 2008; Liu et al. 2009). To date, only a small number of genes for gametic cell fate have been identified, including LIS (Gross-Hardt et al. 2007), GFA1/CLO and ATO (Moll et al. 2008). Remarkably, LIS, AFA1/CLO and ATO are all putative RNA splicing factors. In swa3, changes in synergid cell polarity have also been observed, suggesting the defeat of cell fate identity. It would be interesting to clarify whether the cell fate determination in synergid is normal or not with the aid of egg- or synergid-specific molecular markers.

There is only a single copy of SWA3 in the Arabidopsis genome. Therefore, it is expected to function constitutively in plants; nevertheless, mutation in SWA3 affects female gametogenesis severely and shows little impact on male gametophyte development. This may be a result of genetic redundancy of RNA helicases, as there are as many as 53 RNA helicases in Arabidopsis. In addition, both SWA3 and AtEsf express weakly in pollen, suggesting they might not be essential during male gametogenesis.

In summary, studies on SWAs provide a comprehensive view: RNA processing and ribosomal biogenesis are critical for cell cycle progression and cell fate during female gametophyte development in Arabidopsis.

Materials and Methods

Plant materials and plant transformation

The Arabidopsis wild type plants used in this paper are all Landsberg erecta ecotype. The swa3 mutant was obtained from Ds insertion lines (Springer et al. 1995) in Landsberg erecta ecotype. The Ds insertion in mutant plants was confirmed with Ds flanking sequence primer (5′-CGTTCCGTTTTCGTTTTTTAC-3′) and gene-specific primers 77Bseq5 (5′-ATCAAGGTTTTGCTTGAAAAC-3′), 77BgenMR2 (5′-CCCAGGTGTACTAGAACCTA-3′). Plants were grown in growth room with 16 h light at 22 °C and 8 h dark at 18 °C cycles. Seeds harvested from transgenic lines via the Agrobacterium-mediated infiltration method (Bechtold and Pelletier 1998) were sterilized with 20% bleach and then 70% ethanol, 5 min for each step, rinsed five times with sterile water, and plated on Murashige and Skoog plates with appropriate antibiotics (50 μg/mL kanamycin, or 20 μg/mL hygromycin). The resistant plants were transferred to soil after 10 d of incubation.

Confocal laser scanning microscopy

Confocal laser scanning microscopy analysis was carried out according to the method described previously (Christensen et al. 1997; Shi et al. 2005; Liu et al. 2009). To analyze the female gametophyte phenotype of swa3, flowers of stage 12c (Smyth et al. 1990) were emasculated and pistils were then fixed 48 h later. The ovules were imaged using Zeiss LSM510 META laser scanning microscope with a 488 nm argon laser and an LP530 filter.

Genetic complementation and phylogenetic analysis

The genomic sequences flanking Ds are identified with TAIL-PCR (Liu et al. 1995). For construction of the SWA3 genomic clone for complementation, three genomic fragments are amplified using primers 77BproF (5′-CAAGGGATCCTGATGGTCAGCCTTA-3′)/77BproMR (5′-AAAGCTAAATCCGTTAAT-3′), 77BproMF (5′-GGAAGCTGCGTTATCCTATG-3′)/77BgenMR2 (5′-CCCAGGTGTACTAGAACCTA-3′), 77BgenMF (5′-CCTTGTCATCAATTATGACATTCCGAG-3′)/77BterRSac (5′-CCTCAGATGAAGGTCACAGA-3′). The amplified fragments were digested by BamHI and SacI, NdeI and NcoI, KpnI and SacI respectively, and subsequently subcloned into pCAMBIA1300 (Cambia, Brisbane, Australia).

The homologous sequences from different organisms were obtained from the National Center for Biotechnology Information ( with BLASP. Phylogenetic analysis was carried out as previously described (Liu et al. 2009). ClustalW2 ( and MEGA 2.1 software were used for multiple sequence alignments and the phylogentic tree with the neighbor-joining method.

RNA isolation and expression analysis

Total RNA was isolated from roots, stems, rosette leaves, inflorescences, siliques and seedlings of wild type plants using TRIzol reagent (Invitrogen, Carlsbad, CA, USA), and digested with RNase-free DNase to remove genomic DNA (TaKaRa, Shiga, Japan). RNA was then used as a template for PCR amplification to confirm there was no genomic DNA contamination. One microgram total RNA was reverse-transcribed using oligo(dT)18 primers and reverse transcriptase XL (TaKaRa). Primer combinations kd77BcDNAMF (5′-GATGTTCCTCCAGTTTTCTCTA-3′) and kd77BcDNAMR (5′-GTCGCATAGACTGAGAATTCA-3′), ACTIN2-F (5′-GAAGATTAAGGTCGTTGCACCACCTG-3′)/ACTIN2-R (5′-ATTAACATTGCAAAGAGTTTCAAGGT-3′) were used to detect the expression of SWA3 and ACTIN2. The expression of ACTIN2 was used as an internal control.

RNA in situ hybridization was carried out as described previously (Yang et al. 1999; Liu et al. 2009). Inflorescence was fixed with 4% formaldehyde in 0.01 M phosphate-buffered saline (PBS) buffer (pH 7.0) and dehydrated through ethanol series and embedded in paraplast (Sigma, St. Louis, MO, USA). Sections as thick as 8 μm were prepared with a Leica microtome (Leica, Wetzlar, Germany) and dewaxed with xylene. For probe preparation, a 588-bp SWA3 cDNA fragment was amplified with primer 77B-up (5′-CCATCATGTCTAAATCCCGC-3′) and primer 77B-down (5′-ATCAGGATTGTTTTCAAGCAA-3′), and the product was subcloned into pGEMT-easy vector (Promega, Madison, WI, USA). To obtain antisense probe, the plasmid containing 588-bp cDNA fragment was linearized with SalI, and transcribed in vitro with T7 RNA polymerase (Promega). For the sense probe, the plasmid was digested by NcoI and transcribed with SP6 RNA polymerase (Promega). The hybridization and signal detection were carried out according to previous reports (Yang et al. 1999; Shi et al. 2005; Cordin et al. 2006).

Yeast two-hybrid analysis

The yeast transformation was carried out according to Xie et al. (Xie et al. 1999). The yeast strain used was Saccharomyces cerevisiae AH109 (Clontech, Palo Alto, CA, USA). Transformed cells were plated on -Trp-Leu-His-Ade DO supplement for screening of positive colonies, and incubated at 30 °C for 4 d.

The full-length SWA3 cDNA was amplified by PCR with primers BD77B-upBa (5′-CGGGATCCGTATGGAGGAACCAACGCCGGAGGAGG-3′) and BD77B-downPs (5′-AACTGCAGTCAATTTTCTGTTGATTTCTGCCTT-3′), and cloned into the pGBKT7 (Clontech) in between BamHI and PstI sites to create pBD-SWA3 as a bait in yeast two-hybrid system. The full-length AtEsf2 cDNA was amplified with primers AtEsf2-up-AD-Eco (5′-CGGAATTCATGCAGAGTGAGGAATCTCACGAGC-3′) and AtEsf2-down-AD-Xh (5′-CCGCTCGAGTTAAGAACCGCCAAATACCGATGCT-3′). AtEsf2 cDNA was cloned into pGADT7 (Clontech) at EcoRI and XhoI sites, to generate pAD-Esf2 as a prey.

Accession numbers

Sequence data from this article can be found in the Arabidopsis Genome Initiative database under the following accession numbers: SWA3, At1g16280; AtEsf2, At3g56510.

(Co-Editor: Chun-Ming Liu)


This research was supported with funds from the National Natural Science Foundation of China (30921003 and 30830063) and the Chinese Academy of Sciences (KSCX2-YW-N-048).