SEARCH

SEARCH BY CITATION

Keywords:

  • anther;
  • meiosis;
  • tapetum;
  • pollen mother cell (PMC)

Summary

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

Tapetum development and meiosis play crucial roles in anther development. Here we identified a rice gene, DEFECTIVE TAPETUM AND MEIOCYTES 1 (DTM1), which controls the early stages of that development. This gene encodes for an endoplasmic reticulum (ER) membrane protein that is present only in cereals. Our T-DNA insertion mutations gave rise to abnormal tapetal formation. Cellular organelles, especially the ER, were underdeveloped, which led to hampered differentiation and degeneration of the tapetum. In addition, the development of pollen mother cells was arrested at the early stages of meiotic prophase I. RNA in-situ hybridization analyses showed that DTM1 transcripts were most abundant in tapetal cells at stages 6 and 7, and moderately in the pollen mother cells and meiocytes. Transcripts of UDT1, which functions in tapetum development during early meiosis, were reduced in dtm1 anthers, as were those of PAIR1, which is involved in chromosome pairing and synapsis during meiosis. However, expression of MSP1 and MEL1, which function in anther wall specification and germ cell division, respectively, was not altered in the dtm1 mutant. Moreover, transcripts of DTM1 were reduced in msp1 mutant anthers, but not in udt1 and pair1 mutants. These results, together with their mutant phenotypes, suggest that DTM1 plays important roles in the ER membrane during early tapetum development, functioning after MSP1 and before UDT1, and also in meiocyte development, after MEL1 and before PAIR1.


Introduction

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

Anther development begins with the differentiation of the hypodermis into archesporial cells (Maheshwari, 1950). They then divide periclinally to form inner primary sporogenous cells and outer primary parietal cells. Those sporogenous cells differentiate into pollen mother cells (PMCs) and primary parietal cells, which further differentiate into secondary parietal cells and the endothecium. Secondary parietal cells divide to acquire a middle layer and determine tapetal layer fate. The completed anther wall comprises four cell layers: epidermis, endothecium, middle layer and tapetum (Davis, 1966; Scott et al., 2004).

Defects in early anther formation have been studied in rice. In multiple sporocytes 1 (msp1) mutants, extra meiocytes are produced instead of the tapetum and middle layers (Nonomura et al., 2003). This suggests that the receptor kinase MSP1 is important for determining the fate of parietal cells. Similar phenotypes have been observed from Arabidopsis mutants in EXTRA SPOROGENOUS CELLS/EXCESS MICROSPOROCYTES 1 (EXS/EMS1), which is an ortholog of MSP1 (Canales et al., 2002; Sorensen et al., 2002; Zhao et al., 2002). Mutations in TAPETUM DETERMINANT 1 (TPD1), coordinating with EXS/EMS1 receptor protein kinase, also cause similar phonotypes (Yang et al., 2003). In rice, two TPD1-like genes have been identified; OsTDL1A encoded from one of these genes binds to MSP1 (Zhao et al., 2008). In meiosis arrested at leptotene 1 (mel1) mutants, chromosome condensation is arrested in early meiosis, causing PMCs to vacuolate in developing anthers (Nonomura et al., 2007). MEL1 appears to modulate the division of pre-meiotic germ cells by affecting the modification of meiotic chromosomes via small RNA-mediated gene silencing.

Several genes are involved in the pairing of homologous chromosomes. For example, knock-out plants of HOMOLOGOUS PAIRING ABERRATION IN RICE MEIOSIS 1 (PAIR1) and PAIR2, and knock-down plants of DISRUPTION OF MEIOTIC CONTROL 1 (OsDMC1), do not form bivalents because of a failure of homologous chromosome pairing and synapsis (Nonomura et al., 2004a,b; Deng and Wang, 2007). Those mutants produce abnormal multispores rather than tetrads. PAIR1 contains putative coiled-coil motifs, and PAIR2 encodes a HORMA-domain protein that is homologous to Saccharomyces cerevisiae HOP1 and Arabidopsis ASY1 (Caryl et al., 2000). OsDMC1 is homologous to RecA, which is needed for cytokinesis as well as chromosome pairing (Bishop et al., 1992). Mutants defective in PAIR3 also do not form bivalents because of failures in homologous chromosome pairing and synapsis at diakinesis (Yuan et al., 2009). PAIR3 also encodes a putative coiled-coil protein that appears to control such pairings. In addition, mutants in ZEP1, a gene encoding a transverse filament protein as homolog of Arabidopsis thaliana ZYP1, do not assemble the synaptonemal complexes in early prophase I (Wang et al., 2010). In pollen semi-sterility 1 (pss1) mutants, abnormal spindle formation causes unusual chromosome behavior, occasionally forming triads instead of tetrads (Zhou et al., 2011).

Several genes involved in tapetum development have also been identified. Defects in Undeveloped Tapetum 1 (UDT1) generate a vacuolated tapetum, precluding further development of the tapetum and middle layer during early meiosis. However, meiocytes go through normal meiotic cell division in udt1 (Jung et al., 2005). Mutants in Arabidopsis DYSFUNCTIONAL TAPETUM 1 (DYT1), an ortholog of UDT1, show phenotypes similar to those of udt1 (Zhang et al., 2006). These genes encode basic helix–loop–helix (bHLH) family transcription factors that appear to be one of the earliest players in tapetum development. Mutants in another bHLH gene, TAPETUM DEGENERATION RETARDATION (TDR), exhibit delayed degeneration of tapetum and abnormal formation of pollen walls. In particular, the aliphatic composition of the anthers is greatly changed in the tdr mutant (Zhang et al., 2008). TDR directly associates with the promoter regions of OsC6 and Cysteine Proteases 1 (OsCP1) (Li et al., 2006). Downregulation of the former causes the pollen exine to be disorganized and fewer orbicules to develop along the locular sides (Zhang et al., 2010). OsC6 belongs to a group of lipid transfer proteins, together with OsC4. Both are abundantly expressed in anthers at the early microspore stages (Tsuchiya et al., 1992). OsCP1, encoding a putative Cys protease, appears to be involved in the release of microspores from tetrads, with the first detectable sign of abnormality in oscp1 being observed at the uni-nucleated pollen stage, just before leaving the microspore stage (Lee et al., 2004). Here, we describe a male-sterile rice mutant defective tapetum and meiocyte 1 (dtm1) that has defective tapetal cells and meiocytes.

Results

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

Identification of a male-sterile mutant defective in early anther development

We have previously generated T-DNA insertion mutants in japonica rice (Jeon et al., 2000; Jeong et al., 2002; An et al., 2003; Ryu et al., 2004) and determined their insertion sites (Ryu et al., 2004; An et al., 2005; Jung et al., 2008). From T2 progeny of the population, we isolated a sterile mutant defective in anther development. Mutant anthers were white and smaller than those from segregating wild-type (WT) siblings (Figure 1a,b). To determine whether pollen grains were present, we cleared mature anthers with benzyl-benzoate-four-and-a-half fluid (Herr, 1982). This experiment showed that no grains existed in the mutant anthers (Figure 1c). Nonetheless, mutant spikelets pollinated with WT pollen grains produced fertile seeds (Table S1). When their heteroprogeny was crossed with WT pollen, two types of offspring resulted: the WT and one that carried T-DNA at an approximately 1:1 ratio (Table S1). These results indicated that the female organs were normal. When the heterozygous progeny was selfed, normal and sterile plants were produced at a 3:1 ratio (i.e. 69 fertile:25 sterile; χ2 = 0.128 for 3:1), thereby demonstrating that this sterile phenotype was the result of a male defect caused by a single recessive mutation. Plants of this mutant, dtm1, had no other developmental abnormalities.

image

Figure 1.  Phenotypes of wild-type (WT) and dtm1 anthers. (a) Comparisons among WT (left), dtm1-1 (middle) and dtm1-2 (right) spikelets after removing palea. (b) Comparison between WT and dtm1 anthers. (c) Magnified anthers after clearing. Abbreviations: a, anther; c, carpel; g, glume; l, lemma; lo, lodicule. Scale bars: 1 mm (a, b), or 100 μm (c).

Download figure to PowerPoint

DTM1 encodes an unknown protein

Sequence analysis of the flanking region identified from our mutant showed that T-DNA was inserted into the first intron of LOC_Os07g43010, annotated as a hypothetical protein with no close homologs in the rice genome. This gene consists of two exons and an intron, with the T-DNA being located 1424 bp downstream from the ATG start codon (Figure 2a). From our rice flanking sequence tag database (An et al., 2003; Jeong et al., 2006; Jung et al., 2008), we isolated another T-DNA-tagged line, dtm1-2, in which T-DNA was inserted into the first intron, 284 bp upstream of the mutant dtm1-1 insertion site (Figure 2a). These dtm1-2 mutant plants exhibited the same male-sterile phenotype as dtm1-1 (Figure 1). No transcripts of DTM1 were found in the dtm1-1 or dtm1-2 mutant plants (data not shown), indicating that both are null alleles.

image

Figure 2.  Characterization of DTM1. (a) Scheme for DTM and T-DNA insertion positions in dtm1 mutants: black boxes, exons; connecting line, introns; white boxes, untranslated regions. The ATG start codon and TAG stop codon are both shown. T-DNAs are represented as triangles. P1 and P2 are primers for genotyping and semi-quantitative RT-PCR analysis of DTM transcripts. LB is the left-border primer of T-DNA. (b) Comparisons among DTM1 and homologous proteins from maize and sorghum. The dotted line indicates putative signal peptide; the two black lines indicate the transmembrane regions. (c) Schematic diagrams of fusion constructs DTM1:sGFP (upper) and AtBiP:mRFP (lower); PUbi, maize Ubiquitin promoter; P35S, 35S promoter of cauliflower mosaic virus; Tnos, nopaline synthase terminator. (d–g) Subcellular localization experiment using rice mesophyll protoplasts. Transient expression of control construct containing PUbi::sGFP (d), DTM1:sGFP fusion protein (e), AtBiP:RFP fusion protein as an endoplasmic reticulum (ER) marker (f), and merged image (g). Scale bars: 10 μm. (h–k) Subcellular localization experiment using maize mesophyll protoplasts. Transient expression of control construct containing PUbi::sGFP (h), DTM1:sGFP fusion protein (i), AtBiP:RFP fusion protein as ER marker (j), and merged image (k). Scale bars: 10 μm. (l) ER localization of DTM1. Mitochondria and ER fractions were obtained from protoplasts transformed with DTM1-Myc together with AtBiP-RFP (ER marker) or OGR1-GFP (mitochondria marker). Immunoblotting was performed with anti-Myc Ab, anti-RFP Ab or anti-GFP Ab. Abbreviations: ER, ER fraction; M, mitochondrial fraction; T, total protein. (m) Membrane localization of DTM1. Protein extracts from protoplasts transformed with DTM1-Myc together with AHA2-RFP (membrane protein marker) or mRFP (cytosolic protein marker) were separated into soluble (S) and membrane (M) fractions. Immunoblotting was conducted using anti-Myc Ab or anti-RFP Ab.

Download figure to PowerPoint

DTM1 encodes a peptide sequence of 112 amino acids, which corresponds to KOME clone AK059960. A blast search identified ACF83128, ACG25864 and ACG29558 of maize, and Sb02g039690 of sorghum, as having >97% identify with DTM1 (Figure 2b). Because no obvious homolog was found in dicotyledonous species, including Arabidopsis, we propose that these conserved genes may play a unique role in cereal crop species.

DTM1 was predicted to have an N-terminal signal peptide and two transmembrane segments, suggesting that it is a membrane protein. Therefore, we performed a transient subcellular localization experiment with the DTM1:sGFP fusion protein in protoplasts from etiolated rice (Figure 2d–g) and maize leaves (Figure 2h–k). As a control, we co-transferred the AtBiP:mRFP fusion construct, which was previously reported to be localized to the endoplasmic reticulum (ER) (Kim et al., 2001). A GFP signal from the DTM1:sGFP fusion coincided with the RFP signal driven by AtBiP:mRFP (Figure 2g,k). This result suggested that DTM1 is an ER protein.

To further confirm this, we performed immunoblot analyses using proteins extracted from mesophyll protoplasts transformed with DTM1-Myc, together with either AtBiP-RFP (ER protein) or OGR1-GFP (mitochondrial protein). As expected, DTM1 and AtBiP proteins were detected in the ER fraction, whereas OGR1-GFP was localized to the mitochondrial fraction (Figure 2l). To determine whether DTM1 is associated with membranes, mesophyll protoplasts were transformed with DTM1-Myc together with AHA2-RFP (membrane protein) or RFP (soluble protein). After 12 h of transient expression by the introduced genes, the protoplasts were ruptured by pipeting, and cell debris and chloroplasts were removed by centrifugation. The supernatant was sonicated to disrupt the organelles, and membrane and soluble fractions were separated by ultracentrifugation. As shown in Figure 2m, DTM1 was detected at the same location as AHA2 in the membrane fraction. Therefore, these experiments supported our conclusion that DTM1 is an ER protein located at the membrane.

DTM1 is preferentially expressed in tapetal cells

Although DTM1 was expressed in all organs examined, transcripts were more abundant in young panicles (Figure S1a). In developing anthers, expression was higher during meiosis (Figure S1b). Laser microdissection transcriptome analyses also demonstrated its high expression in tapetal-layer cells and meiocytes (Hobo et al., 2008). To examine further the spatial and temporal patterns of this expression, we performed RNA in-situ hybridization analyses of the anthers at stages 6–8 (Figure 3). Transcripts of DTM1 were most abundant in the tapetal cells at stages 6 and 7, and moderately in the PMCs and meiocytes. The transcripts were also weakly detected in most anther walls, including the middle layer, endothecial cells, and epidermal cell layer. The level of transcript was significantly reduced at stage 8.

image

Figure 3.  RNA in-situ hybridization analyses of DTM1. (a, b) Wild-type (WT) anthers at stage 6 hybridized with antisense probe (a) and sense probe (b). (c, d) WT anthers at stage 7 hybridized with antisense probe (c) and sense probe (d). (e, f) WT anthers at stage 8 hybridized with antisense probe (e) and sense probe (f). Abbreviations: MC, meiocyte; PMC, pollen mother cell; T, tapetum cell; Tds, tetrads. Scale bars: 10 μm.

Download figure to PowerPoint

Microscopic observations of dtm1 anthers

To determine where and when defects occur in the dtm1 mutant, we compared thin sections of anthers from both WT and mutant samples. Anthers develop over 12 stages (Zhang and Wilson, 2009). In our WT, the primordium was formed via cell division in the L1, L2 and L3 layers of the floral meristem at stage 1. At the four corners of the hypodermis, archespores began to differentiate at stage 2, before developing to sporogenous cells and parietal cells during stages 3 and 4. At stage 5, the parietal layers developed endothecium, middle and tapetal layers, creating four layers of anther walls (Figure 4a). The sporogenous cells generated PMCs within the locule at stage 6 (Figure 4b). During stages 7 and 8, The PMCs then underwent meiosis and formed dyads and tetrads (Figure 4c,d). By that time, the tapetal-cell layers showed considerable development and differentiation, whereas cells from the middle layer had started to shrink. Afterward, the cytoplasm of the tapetal cells pulled away from the peritapetal region and condensed toward the middle layer (Figure 4d).

image

Figure 4.  Transverse sectional comparison between WT and dtm1-1 anthers. Cross sections of WT (a–d, i–l) and dtm1-1 (e–h, m–p) anthers at stage 5 (a, e), stages 6–7 (b, f), stage 8a (c, g), stage 8b (d, h), stage 9 (i, m), stage 10 (j, n), stage 11 (k, o) and stage 12 (h, p). DMC, degenerative meiocyte; En, endothecial cell; Ep, epidermal cell; MC, meiocyte; ML, middle layer; MP, mature pollen; MSP, microspore; PMC, pollen mother cell; SC, sporogenous cell; SPC, secondary parietal cell; T, tapetum cell; Tds, tetrads. Scale bars: 10 μm.

Download figure to PowerPoint

The dtm1 mutant anthers did not differ from the WT at stages 5–7 (Figure 4e,f). That is, the four layers of the anther wall were normally formed and PMCs had developed as expected. However, differences between the WT and mutant started to emerge at stage 8a (Figure 4g). That is, whereas the middle cell layer in the WT was almost entirely diminished, no change was seen for that layer in dtm1, remaining, instead, in the condition associated with stage 7. Aniline blue staining showed the irregular stacking of callose on the mutant meiocytes (Figure S2). We monitored the expression levels of OsGSL5 (putative glucan synthase gene, AP003454), which encodes an anther-specific callose synthase (Yamaguchi et al., 2006), and Osg1 (glucanase, AB070742), which is required for callose degradation (Wan et al., 2011). During stages 6–8a, OsGSL5 expression was not altered in the mutants compared with the WT, whereas Osg1 transcripts were reduced in the mutants (Figure S2c,d). Such a decrease in Osg1 expression would have caused a hyperaccumulation of callose in dtm1.

At stage 8b, the dtm1 tapetum, unlike that of the WT, did not differentiate, but instead became enlarged (Figure 4h). Meiocytes were degenerated and crushed meiotic products were present in the locule. The middle layer appeared to remain the same as at the previous stage, without degeneration.

Stage 9 for the WT began with tapetal cells being condensed and young microspores detaching from the tetrads (Figure 4i). In stage 10, microspores enlarged and had expanded vacuoles (Figure 4j). By stage 11, the tapetal cells had almost disappeared and young microspores became bi-nucleus pollen grains, with both a generative nucleus and a vegetative nucleus (Figure 4k). The vacuole shrank and cytoplasm contents increased. At stage 12, the tapetal cell layer was completely empty and vacuoles in the grains had disappeared, being filled instead with starch granules and other substances (Figure 4l).

In contrast, dtm1 anthers at stage 9 had more swollen tapetal cells (Figure 4m). At the later stages, those swollen cells and remnants of degenerated meiotic cells were still evident (Figure 4n–p).

Transmission electron microscopic observations of dtm1 mutant and wild-type anther walls

To observe subcellular alterations, we performed transmission electron microscopy (TEM) analysis with ultrathin transverse sections of anther walls and microsporocytes. At stage 8a, the WT middle layer became thinner than the other layers (Figure 5a). In contrast, that of the mutant remained almost the same as before (Figure 5e). The WT tapetal cell cytoplasm was rich with cellular organs, such as the elaioplast, mitochondria and stacks of ER, which occupied most of the cytoplasm (Figure 5b,c). By comparison, the mutant tapetal cell cytoplasm was filled with globular structures lacking any obvious organelle development (Figure 5f); these phenotypes appeared from late stage 7 (Figure S3). We also observed spiral forms that appeared to be undeveloped ER stacks (Figure 5g). Because dilatation of the ER cisternae is the most outstanding feature in tapetal differentiation (Mamun et al., 2005; Parish and Li, 2010), the dtm1 sections appeared to be defective in that regard. The WT cytoplasm began to detach from the peritapetal regions and radial cell wall, making spaces obvious between tapetal cells (arrowheads in Figure 5b). In the mutant, however, the tapetum cytoplasm was not so detached (Figure 5f). At this stage, a callose layer was deposited over the cell walls of meiocytes (Figure 5d). However, the cell wall and callose wall of dtm1 meiocytes were irregular (Figure 5h).

image

Figure 5.  TEM analyses of anthers. (a, e) Anther wall layer of the wild type WT (a) and dtm1 (e) at stage 8a; En, endothecial cell; Ep, epidermal cell; ML, middle layer; T, tapetum cell. (b, f) Magnified images of tapetal cell from the WT (b) and dtm1 (f) at stage 8a. Arrowheads (b, f, i) show area generated by cytoplasm condensation; el, elaioplat; ER, endoplasmic reticulum; m, mitochondria; n, nucleus; PT, peritapetal region. (c, g) Magnified images of tapetal cytoplasm from WT (c) and dtm1 (g) at stage 8a. Arrows indicate ER-like stacks. (d, h) Meiocytes of WT (d) and dtm1 (h) at stage 8b; Ca, callose wall; W, cell wall. (i) Tapetum of WT at stage 8b. Arrowheads indicate tapetal cell walls toward locule. Arrows point to radial cell walls. (j) Tapetum of dtm1 at stage 8b. Arrowheads indicate vesicles. Arrows point to dense bodies. (k, l, m) Magnified images of cytoplasmic areas from dtm1 tapetal cell at stage 8b. Arrows (k) indicate abnormal ER stacks; V, vacuole; Ve, vesicle. (n, o) Tetrad of WT (n) and degenerated meiocytes of dtm1 (o) at stage 8b. Arrows point to the primexine (microspore wall) surrounding the tetrad; Ca, callose wall. Scale bars: 1 μm (a–c, e–g and i–k); 5 μm (d, h, l–o).

Download figure to PowerPoint

At stage 8b, these morphological differences were more evident. The WT tapetal cell cytoplasm was shrunken, with the tapetum cell wall toward the peritapetal region remaining, but the radial cell walls starting to disintegrate (Figure 5i). Such dissolution generally follows cytoplasm condensation (Parish and Li, 2010). In contrast, the tapetal cell cytoplasm from dtm1 did not shrink (Figure 5j). Instead, mutants retained the original cell walls, which contained irregularly shaped, small, dense bodies (arrows in Figure 5j) and expanded vesicles (arrowheads in Figure 5j). Some vesicles also had spiral deposits (Figure 5k). Cells were occasionally occupied by numerous small vesicles and vacuoles (Figure 5l), whereas others had large vacuoles (Figure 5m). For the WT, tetrads formed at this stage and were enveloped by newly produced primexine walls (Figure 5n). However, mutant meiocytes did not develop to tetrads. These degenerated meiocytes contained numerous vesicles and, instead of primexine, a callose wall was present around them (Figure 5o).

To examine whether the leaves had any defects, we performed TEM analysis using the third leaves, where the DTM1 transcripts were detected. Morphology of the ER and other organelles of the mutant was almost the same as for the WT (Figure S4).

Development of pollen mother cells was arrested during meiosis I in dtm1

One aspect of the dtm1 phenotype is the failure to form tetrads. Therefore, to observe defects within the meiotic process, we stained chromosomes with 4′,6-diamidino-2-phenylindole (DAPI) and counted PMCs at each step in that process (Figure 6a,b). We also used anther length as a criterion for assessing stage progression. In WT anthers of 0.5–0.6 mm in length, most PMCs were found at prophase I, being almost evenly distributed from the leptotene to the diakinesis stage, but with a few also having entered the metaphase. When anthers followed to the next stages (0.7–0.9 mm in length), they rapidly progressed to metaphase and later phases. In contrast, for dtm1 anthers that were 0.5–0.6 mm in length, most of the cells were at the pre-meiosis stage. In more developed anthers (0.7–0.9 mm in length), some had entered early prophase I, but none had further progressed to the metaphase. In total, 73.1% of the mutant PMCs in the 0.5–0.9-mm stages were at pre-meiosis or leptotene, compared with only 13.8% of the WT, with the remainder having already advanced to later stages. These observations indicated that meiosis in the dtm1 mutant anthers was slow, and was eventually arrested at prophase I.

image

Figure 6.  Distribution of chromosomes during meiosis. (a, b) Histograms showing number of pollen mother cells (PMCs) at each meiotic stage from wild-type (WT) (a) and dtm1-1 (b) anthers that are 0.5–0.9 mm in length. (c–f) Chromosome spreads from WT male meiocytes at leptotene (c), zygotene (d), pachytene (e) and diplotene (f). Arrowheads indicate chiasmata. Scale bars: 10 μm. (g–j) Chromosome spreads from dtm1-1 meiocytes at leptotene (g), zygotene (h), pachytene (i) and diplotene (j). Scale bars: 10 μm.

Download figure to PowerPoint

We further examined whether chromosomes were altered during prophase I. In the WT, chromosomes began to condense at leptotene (Figure 6c), being clustered at one end of the nucleus at the zygotene stage (Figure 6d). At pachytene, homologous chromosomes were arranged in a line after the formation of the synaptomal complex was complete (Figure 6e). At diplotene, none of the bivalents were in close contact along the entire length except at the chiasmata (Figure 6f). These WT chromosomes then further condensed and completed meiosis I (Figure S5).

In dtm1 mutants, a majority of meiotic cells remained in leptotene or zygotene, and only a few reached pachytene or diplotene. Those chromosomes began to condense normally at leptotene (Figure 6g), but were less compact than the WT at zygotene (Figure 6h). At pachytene, the mutant chromosomes seemed to aggregate, and alignment was not as clear as with the WT (Figure 6i). This aggregating pattern was more severe at diplotene (Figure 6j). We did not observe any condensed chromosome fragments, e.g. univalents or bivalents. These observations suggested that the loss of function of DTM1 affects PMCs entrance to meiosis, and hinders further progress in the early stages of prophase I.

DTM1 acts later than MSP1 and prior to UDT1 and PAIR1

To investigate the relationship between DTM1 and previously studied genes, we monitored the expression of genes associated with anther development (Figure 7). We first examined MSP1 because that gene plays crucial roles at the stage of entering into male and female sporogenesis. It was also the earliest gene to be investigated in detail (Nonomura et al., 2003). Our real-time PCR analysis showed that expression of MSP1 was not altered in dtm1 (Figure 7a). We then monitored UDT1, which functions in early tapetum development (Jung et al., 2005). Our analyses showed that its expression was decreased in the mutant (Figure 7b). Therefore, these results suggest that DTM1 functions later than MSP1, but prior to UDT1.

image

Figure 7.  Expression analysis of anther wall development marker genes in the dtm1-1 mutant compared with the wild type (WT). Quantitative RT-PCR of MSP1 (a), UDT1 (b), GAMYB (c), TDR (d), OsC4 (e), OsC6 (f) and CP1 (g) in dtm1-1 and segregating WT control. Y-axis, gene expression relative to rice Ubiquitin (Os03g0234200) transcript level. Results are averages of three independent experiments.

Download figure to PowerPoint

We also analyzed the expression of GAMYB and TDR, which are related to tapetal programmed cell death (PCD) (Zhang et al., 2008; Aya et al., 2009). However, unexpectedly, transcript levels of these genes were not significantly altered (Figure 7c,d). By contrast, those of OsC4, OsC6 and CP1 (Lee et al., 2004; Huang et al., 2009; Zhang et al., 2010) were significantly reduced in the dtm1 anthers (Figure 7e–g). OsC6 and CP1 are direct targets of TDR: the former is related to pollen wall formation and the latter functions in tapetal PCD.

Because dtm1 anthers also showed phenotypic changes in meiocytes, we compared transcript levels of the genes involved in meiosis (Figure 8). OsMEL1, which acts in premeiotic germ cell division, was not significantly changed in dtm1 (Figure 8a). Among PAIR1, PAIR2, PAIR3, DMC1 and ZEP1, which are known to function in chromosome pairing or synapsis (Nonomura et al., 2004a,b; Deng and Wang, 2007; Yuan et al., 2009; Wang et al., 2010), the transcript level of PAIR1 was diminished in the mutants, whereas the other genes were not affected (Figure 8b–f). In addition, the expression of PSS1, which encodes microtubule-stimulated ATPase activity, was reduced (Figure 8g).

image

Figure 8.  Expression analysis of meiosis-related genes in the dtm1-1 mutant compared with the wild type (WT). Quantitative RT-PCR of MEL1 (a), PAIR1 (b), PAIR2 (c), PAIR3 (d), DMC1 (e), ZEP1 (f) and PSS1 (g) in dtm1-1 and segregating WT control. Y-axis, gene expression relative to rice Ubiquitin (Os03g0234200) transcript level. Results are averages of three independent experiments.

Download figure to PowerPoint

To verify this genetic hierarchy, we evaluated DTM1 transcripts in msp1, udt1, tdr and pair1 mutants. DTM1 levels were lower in msp1 anthers, supporting our conclusion that DTM1 plays a role after MSP1 (Figure 9a). However, those levels were unchanged in the other mutants (Figure 9b–d), demonstrating that DTM1 functions before those genes.

image

Figure 9.  Expression analysis of DTM1 in other anther mutants compared with the wild type (WT). The mutants are as follows: msp1 (a), udt1 (b), tdr (c) and pair1 (d). Y-axis, gene expression relative to rice Ubiquitin (Os03g0234200) transcript level. Results are averages of three independent experiments.

Download figure to PowerPoint

Discussion

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

DTM1 is essential for rice tapetum development and meiosis

The phenotypic characteristics of dtm1 mutants are: (i) underdeveloped tapetal cells blocked at the early developmental stage; (ii) meiocytes arrested at the pre-meiosis stage; and (iii) no degeneration of the middle layer. Although these phenotypes appear in three different cell types, they can all be attributed to a defect in the tapetum because DTM1 is more preferentially expressed in tapetal cells when the mutants begin to show those flaws. The phenotype for aborted middle-layer degeneration is common in tapetum-defective mutants, such as udt1, wda1 and tdr in rice (Jung et al., 2005, 2006; Li et al., 2006). Arabidopsis tapetum mutants ems1, tpd1, dyt1 and tdf1 also maintain a thick middle layer (Zhao et al., 2002; Yang et al., 2003; Zhang et al., 2006; Zhu et al., 2008).

The tapetum is also necessary for the development of meiocytes and pollen. In maize ms23 and ms32 mutants, an undifferentiated tapetum results in aborted meiocytes after the onset of prophase I (Chaubal et al., 2000). In the Arabidopsis ems1 mutant, meiocytes do not complete cytokinesis (Zhao et al., 2002). Similarly, the meiocytes fail cytokinesis in the tapetum mutant dyt1 (Zhang et al., 2006). For the rice msp1 mutant, in which tapetal cells are replaced by meiocytes, meiotic nuclear division is arrested at prophase I (Nonomura et al., 2003). These reports indicate that tapetal and meiocyte activities may be coordinated during early anther development.

However, it is also possible that these defects may be independent events in different cell types. The development of PMCs in dtm1 is arrested in early prophase I, unlike most tapetum mutants that exhibit defects in cytokinesis. This hypothesis is supported by our in-situ RNA analyses that revealed DTM1 is also expressed in meiocytes. This is consistent with an earlier conclusion that is a meiocyte-expressed gene in laser-captured PMCs (Tang et al., 2010). We identified DTM1 as a PMC-preferential gene with 78-fold higher expression in PMC compared with tricellular pollens and 5.5-fold higher expression than that found in seedlings. Therefore, we propose that the mutant phenotypes are caused by defective DTM1 expression in both the tapetum and meiocytes.

DTM1 functions between MSP1 and UDT1 in tapetal development primarily at late stage 7

To study the relationship between DTM1 and previously identified genes that function in the tapetum, we analyzed their expression levels in dtm1 mutant anthers. We also measured levels of DTM1 transcript in mutant anthers when available.

Whereas DTM expression was reduced in msp1, that of MSP1 was not affected in dtm1, suggesting that the latter functions before the former. Here, Affymetrix microarray data was also used with developing anthers at different stages (Fujita et al., 2010; Jung et al., 2011). MSP1 was highly expressed during stages 3–7, before declining at stages 8 and 9 (Figure 10). However, levels of DTM1 transcript were higher between stages 6 and 8a, demonstrating that DTM1 functions after MSP1. The latter gene plays an important role in the specification of anther wall layers (Nonomura et al., 2003). Mutations result in an excessive number of both male and female sporocytes. In addition, the tapetum layer is completely lost. Because we found that all four layers were properly developed in dtm1, we can conclude that DTM1 should function later than MSP1. Although DTM1 expression was decreased in msp1, the reduction was not severe. This was probably because of gene expression in the meiocytes of the msp1 mutant anthers.

image

Figure 10.  Expression patterns of anther developmental genes and time when mutant phenotypes begin to occur. Microarray analyses of tapetum developmental genes (a) and pollen mother cells (PMCs) and microspore developmental genes (b) from Fujita et al. (2010). Red arrows indicate stages when mutant phenotypes begin to appear. (a), (b), (c) and (d) show biological replicates of anthers at each stage. Stages 3–5 coincide with anther length 0.10–0.15 mm; stage 6, 0.20–0.45 mm; stage 7, 0.40–0.55 mm; stage 8a, 0.45–0.80 mm; stage 8b, 0.75–1.10 mm; stage 9, anther with uninucleate pollens; stage 10–11, bi-nucleate; stage 12, tri-nucleate.

Download figure to PowerPoint

We previously reported that UDT1 plays an essential role during early tapetum development (Jung et al., 2005). There, the udt1 tapetum failed to differentiate and became vacuolated during meiosis (Jung et al., 2005). Meiocytes of the udt1 mutants were normal during early pre-meiosis, but did not develop to tetrads. In this study, we showed that the transcripts of UDT1 are reduced in dtm1, whereas those of DTM1 are unaffected in udt1. Therefore, these phenotypes and expression data indicate that UDT1 functions later than DTM1.

TAPETUM DEGENERATION RETARDATION acts downstream of GAMYB and UDT1 during early anther development (Liu et al., 2010). The gamyb mutants exhibit defects in tapetum PCD and the formation of exine and Ubish bodies. Mutants in tdr have similar defects in tapetum PCD and pollen wall formation. Compared with them, udt1 show earlier defects in tapetum development and PCD. These findings indicate that UDT1 functions before GAMYB and TDR. Because dtm1 affects UDT1 expression, we expected that GAMYB and TDR expression would be changed. However, neither was significantly altered in dtm1. Because GAMYB is expressed highly throughout anther development, stopping that process at any stage would not change the level of its expression (Figure 10a; Fujita et al., 2010). Low levels of expression by TDR were found in the early stages of anther development, but transcripts began to accumulate at late stage 7 (Figure 10a). Because such transcription was not altered in dtm1, the mutant anthers should have developed to stage 7.

Both OsC6 and OsCP1 are direct targets of TDR (Li et al., 2006). Knock-down plants of OsC6, which is related to lipid transfer from the tapetum to the pollen exine through the orbicules, produce irregular pollen walls, causing grains to abort (Zhang et al., 2010). Mutations in cp1, which encodes Cys protease, also exhibit collapsed microspores (Lee et al., 2004). C4 is considered a putative lipid transfer protein, like C6. Here, CP1 and OsC6 were expressed at very low levels before stage 8, but were then greatly increased at late stage 8 (Figure 10a). In contrast, the OsC4 transcript began to accumulate at early stage 8. Because transcripts of these three genes were not detectable in dtm1, the mutant anthers should have been defective at late stage 7 or at early stage 8.

DTM1 functions between MEL1 and PAIR1 in meiocytes

We studied the relationship between DTM1 and Mel1 because mutation in the latter caused defects in the meiocytes at prophase I. MEL1 expression was not affected in dtm1, suggesting that it functions before DTM1. Although both mel1 and dtm1 PMCs were arrested at prophase I, mel1 presented fused, aberrant PMCs, whereas those of dtm1 were normal. The MEL1 gene is expressed specifically at stages 5–7 in archesporial and sporogenous cells, before the pre-meiosis stage (Nonomura et al., 2007). However, we found that DTM1 expression was low in those cells but reached high levels at stage 7 in the PMCs (Figure 10b). Therefore, we can conclude that DTM1 functions later than MEL1.

We noted that transcript levels of PAIR1 were significantly reduced in dtm1, whereas DTM1 transcript levels were not altered in pair1 anthers. The dtm1 meiocytes did not progress beyond diplotene, whereas the pair1 PMCs eventually developed into, albeit abnormal, tetrads (Nonomura et al., 2004a). Therefore, DTM1 appeared to function before PAIR1. Because both DTM1 and PAIR1 encode unknown proteins, it is difficult to speculate on their relationship. A transient assay using the 35S-GFP–PAIR1 fusion construct has revealed that the fusion protein is localized to the nucleus, suggesting that PAIR1 acts there (Nonomura et al., 2004a). As DTM1 is an ER protein, the two genes do not appear to interact directly with each other.

Because our dtm1 mutants did not form dyads or tetrads, dtm1 preceded other meiosis mutants such as pair2, pair3, dmc1 and zep1. Mutants of pair2 and dmc1 developed multispores instead of tetrads because of defects in homologous pairing. Those of pair3 produced abnormal microspores because of a lack of bivalent formation, whereas zep1 exhibited partial abortion caused by increased crossover during meiosis. However, transcript levels of these genes were not affected in the dtm1 mutants. Similarly, PAIR2 expression is not changed in mel1 (Nonomura et al., 2007). Mutations in PAIR3 do not affect the expression of other meiotic genes, including MEL1, PAIR1, PAIR2 and DMC1 (Yuan et al., 2009). Thus, it is likely that these meiosis genes begin their expression in anthers much before the mutants show their phenotypes. This hypothesis is supported by our expression profiles that indicated the transcript levels of these genes were higher at stages 5 and 6, before the dtm1 mutant phenotype became apparent (Figure 10b). By contrast, expression of PSS1 was reduced in dtm1. The pss1 mutants manifested abnormal spindle formation because of diminished microtubule interaction; PSS1 is reported to act during later stages of the meiotic process (Zhou et al., 2011).

DTM1 is localized to the ER

DTM1 is unique to cereals: no homologous gene occurs in dicot species, including Arabidopsis. Our BLASTP search indicated that DTM1 has 17% identity with yeast SPC1p, a subunit of the signal peptidase complex (SPC), which is located in the ER. Purified SPC complex from dog contains five subunits: SPC12, SPC18, SPC21, SPC22/23 and SPC25 (Paetzel et al., 2002). SPC12 has two transmembrane segments and the molecular weight is approximately 12 kDa, which is similar to DTM1. A mammalian protein homologous to the yeast SPC complements a yeast mutant sec11 that is defective in signal peptidase I (Fang et al., 1996). Because DTM1 is also a small ER protein, we examined whether it functions as an SPC. In performing complementation experiments with yeast mutant sec11, we were unable to rescue the mutant by overexpressing DTM1.

The failure of DTM1 to rescue the yeast mutant might result from the diversity between yeast and plants. Therefore, we investigated whether DTM1 interacts with other SPC subunits. We performed yeast two-hybrid experiments and co-immunoprecipitation assays to test the relationship between DTM1 and putative SPC subunits encoded by LOC_Os05g23260, LOC_Os06g16260 and LOC_Os09g38370. These code for proteins with homology to yeast SPC18, SCP21 and SPC25, respectively. Here, DTM1 did not interact with any of those proteins (data not shown), indicating that it is unlikely to be a transpeptidase.

Although the DTM1 transcript was detected more preferentially in developing panicles, especially in tapetum cells, it was also expressed in both vegetative and reproductive cells. However, we observed the mutant defects only in anthers. In the tapetum, the ER was underdeveloped in dtm1. A large quantity of cellular energy is needed in the early stages of anther development (Liu and Dickinson, 1989). Dilatation of the ER cisternae is the most outstanding feature in tapetal differentiation (Mamun et al., 2005; Parish and Li, 2010). Because DTM1 is an ER protein and the lack of the protein affects ER development, it can be postulated that DTM1 is an essential ER component that is needed for ER development.

Experimental Procedures

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

Plant materials

The dtm1-1 (Oryza sativa var. japonica cv. Dongjin, line 2C-40285) and dtm1-2 (O. sativa var. japonica cv. Hwayoung, line 2B-00145) mutants were isolated from T-DNA tagging lines generated from O. sativa var. japonica (Jeon et al., 2000; Jeong et al., 2002; Ryu et al., 2004). Insertion mutant lines 9-142-08 for udt1-1, 3A-09404 for msp1, 3A-08673 for tdr, and 3A-05160 and 3A-60500 for pair1 were also used. Lines 3A-09404, 3A-05160 and 3A-60500 were generated by Tos17, whereas 3A-08673 was generated by T-DNA. The development of line 9-142-08 has previously been reported (Jung et al., 2005, 2006).

Subcellular localization of DTM1 protein

The DTM1 coding region was amplified without its stop codon, using two primers, 5′-ggatccGCTTGGTGAGCCACC-3′ and 5′-actagtGCTAGATACGTACATCC-3′ (italicized nucleotides are restriction enzyme sites). An amplified fragment was digested with EcoRI and SpeI, and inserted into the pGA3452 vector (Kim et al., 2009b) to generate the DTM1:sGFP fusion molecule. A vector containing the mRFP fusion with AtBiP has been reported previously (Kim et al., 2001). Protoplast preparation and transformation procedures were conducted, as described by Park et al. (2011).

Subcellular fractionation and immunoblotting

Subcellular fractions containing mitochondria or ER were obtained from protoplasts, as described earlier (Hawes and Satiat-Jeunemaitre, 2001; Lee et al., 2011). Briefly, transformed protoplasts were homogenized and incubated on ice in HMS buffer (330 mM sorbitol, 50 mM HEPES-KOH, pH 7.6, and 3 mM MgCl2) and centrifuged at 900 g for 5 min to remove chloroplasts. The supernatant was then loaded onto a two-step sucrose gradient (from bottom to top: 2 ml of 60% and 4 ml of 36% sucrose in 50 mM HEPES-KOH, pH 7.6) and centrifuged at 40 000 g for 90 min. The ER and mitochondrial fractions were collected at the top of the 36% sucrose layer and at the 36–60% sucrose interface, respectively. The procedure to identify soluble and membrane-bound proteins was as described by Iwata et al. (2008). Transformed protoplasts were homogenized in LE buffer (80 mm Tris-HCl, pH 7.5, 12% sucrose and 1 mm phenylmethylsulfonyl fluoride, PMSF) and centrifuged at 100 000 g for 1 h. The resulting pellet was re-suspended on the LE buffer and sonicated. Proteins were separated by SDS-PAGE and subjected to immunoblot analysis using anti-RFP (Abcam, http://www.abcam.com), anti-GFP (Santa Cruz Biotechnology, http://www.scbt.com) or anti-myc (Cell Signaling Technology, http://www.cellsignal.com) antibodies.

Histochemical analyses

Spikelets at different developmental stages were fixed with 3% (w/v) paraformaldehyde and dehydrated in an ethanol series. The samples were embedded in Technovit 8100 resin (Heraeus Dental International, http://heraeus-dental.com), then sectioned to 5–10 μm in thickness, stained with 0.1% toluidine blue and observed with an Olympus microscope BX61 (Olympus, http://www.olympus.com).

Ultrastructural analyses by transmission electron microscopy

Anthers at various stages of development plus the third leaves from 9-day-old seedlings were fixed in sodium phosphate buffer (pH 7.2) that contained 3% glutaraldehyde (Sigma-Aldrich, http://www.sigmaaldrich.com). They were then post-fixed using 2% osmium tetroxide (Pelco International, http://www.pelcoint.com). After dehydration, the specimens were placed in Spurr’s low-viscosity embedding mixture (EMS, http://www.emsdiasum.com/). Ultrathin sections (40–60 nm in thickness) were stained with 2.5% uranyl acetate and 2.5% lead citrate aqueous solutions, and examined with a transmission electron microscope JEOL 1200 (Jeol Ltd, Tokyo, Japan).

RNA extraction and quantitative real-time RT-PCR

Total RNA was isolated at different developmental stages (Huang et al., 2009), and synthesized cDNAs were used for quantitative real-time RT-PCR. All experiments were conducted at least three times, with three or more independent samples each. Primers used for analyzing transcript levels are listed in Table S2.

Meiotic chromosome observation

Young spikelets containing anthers of 0.5–0.9 mm in length were fixed in Carnoy’s fixative (3:1 ethanol:glacial acetic acid) and stored at 4°C. Chromosome spreads of PMCs were prepared according to the method of Chen et al. (2005). They were stained with DAPI (1 μg ml−1 DAPI in a buffer containing 50% glycerol and 10 mm citrate, pH 4.5). Images were photographed under an Olympus BX61 microscope.

RNA in-situ hybridization

Spikelets at different developmental stages were fixed, dehydrated, embedded, sliced and attached to slides as previously described (Lee et al., 2007). For preparation of digoxigenin-labeled RNA probes, we PCR amplified the DTM1 coding region with primer set 5′-ATGGGGAGGGACGAGATG-3′ and 5′-CTAGCTAGATACGTACATCC-3′. Images were photographed with an Olympus BX61 microscope.

Acknowledgements

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

We express our thanks to Cheol Woong Ha for yeast complementation experiments, Kyungsook An for generating the T-DNA insertional lines and handling the seed stock, Woo Taek Kim, Jong-Jin Park and Jin Ping for valuable discussion, and Priscilla Licht for her critical reading of the manuscript. This work was supported in part by grants from the Next-Generation BioGreen 21 Program (grant no. PJ008215), the Rural Development Administration, Republic of Korea; the Basic Research Promotion Fund, Republic of Korea (KRF-2007-341-C00028); and Kyung Hee University (20110269).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Supporting Information
  • An, S., Park, S., Jeong, D.H. et al. (2003) Generation and analysis of end sequence database for T-DNA tagging lines in rice. Plant Physiol. 133, 20402047.
  • An, G., Lee, S., Kim, S.H. and Kim, S.R. (2005) Molecular genetics using T-DNA in rice. Plant Cell Physiol. 46, 1422.
  • Aya, K., Ueguchi-Tanaka, M., Kondo, M., Hamada, K., Yano, K., Nishimura, M. and Matsuoka, M. (2009) Gibberellin modulates anther development in rice via the transcriptional regulation of GAMYB. Plant Cell, 21, 14531472.
  • Bishop, D.K., Park, D., Xu, L. and Kleckner, N. (1992) DMC1: a meiosis-specific yeast homolog of E. coli recA required for recombination, synaptonemal complex formation, and cell cycle progression. Cell, 69, 439456.
  • Canales, C., Bhatt, A.M., Scott, R. and Dickinson, H. (2002) EXS, a putative LRR receptor kinase, regulates male germline cell number and tapetal identity and promotes seed development in Arabidopsis. Curr. Biol. 12, 17181727.
  • Caryl, A.P., Armstrong, S.J., Jones, G.H. and Franklin, F.C. (2000) A homologue of the yeast HOP1 gene is inactivated in the Arabidopsis meiotic mutant asy1. Chromosoma, 109, 6271.
  • Chaubal, R., Zanella, C., Trimnell, M.R., Fox, T.W., Albertsen, M.C. and Bedinger, P. (2000) Two male-sterile mutants of Zea mays (Poaceae) with an extra cell division in the anther wall. Am. J. Bot. 87, 11931201.
  • Chen, C.B., Xu, Y.Y., Ma, H. and Chong, K. (2005) Cell biological characterization of male meiosis and pollen development in rice. J. Integr. Plant Biol. 47, 734744.
  • Davis, G.L. (1966) Systematic Embryology of the Angiosperms. New York: Wiley.
  • Deng, Z.Y. and Wang, T. (2007) OsDMC1 is required for homologous pairing in Oryza sativa. Plant Mol. Biol. 65, 3142.
  • Fang, H., Panzner, S., Mullins, C., Hartmann, E. and Green, N. (1996) The homologue of mammalian SPC12 is important for efficient signal peptidase activity in Saccharomyces cerevisiae. J. Biol. Chem. 271, 1646016465.
  • Fujita, M., Horiuchi, Y., Ueda, Y. et al. (2010) Rice expression atlas in reproductive development. Plant Cell Physiol. 51, 20602081.
  • Hawes, C.R. and Satiat-Jeunemaitre, B. (2001) Plant Cell Biology: A Practical Approach, 2nd edn. Oxford, New York: Oxford University Press.
  • Herr, J.M. Jr (1982) An analysis of methods for permanently mounting ovules cleared in four-and a-half type clearing fluids. Stain Technol. 57, 161169.
  • Hobo, T., Suwabe, K., Aya, K. et al. (2008) Various spatiotemporal expression profiles of anther-expressed genes in rice. Plant Cell Physiol. 49, 14171428.
  • Huang, M.D., Wei, F.J., Wu, C.C., Hsing, Y.I. and Huang, A.H. (2009) Analyses of advanced rice anther transcriptomes reveal global tapetum secretory functions and potential proteins for lipid exine formation. Plant Physiol. 149, 694707.
  • Iwata, Y., Fedoroff, N.V. and Koizumi, N. (2008) Arabidopsis bZIP60 is a proteolysis-activated transcription factor involved in the endoplasmic reticulum stress response. Plant Cell, 20, 31073121.
  • Jeon, J.S., Lee, S., Jung, K.H. et al. (2000) T-DNA insertional mutagenesis for functional genomics in rice. Plant J. 22, 561570.
  • Jeong, D.H., An, S., Kang, H.G., Moon, S., Han, J.J., Park, S., Lee, H.S., An, K. and An, G. (2002) T-DNA insertional mutagenesis for activation tagging in rice. Plant Physiol. 130, 16361644.
  • Jeong, D.H., An, S., Park, S. et al. (2006) Generation of a flanking sequence-tag database for activation-tagging lines in japonica rice. Plant J. 45, 123132.
  • Jung, K.H., Han, M.J., Lee, Y.S., Kim, Y.W., Hwang, I., Kim, M.J., Kim, Y.K., Nahm, B.H. and An, G. (2005) Rice Undeveloped Tapetum1 is a major regulator of early tapetum development. Plant Cell, 17, 27052722.
  • Jung, K.H., Han, M.J., Lee, D.Y. et al. (2006) Wax-deficient anther1 is involved in cuticle and wax production in rice anther walls and is required for pollen development. Plant Cell, 18, 30153032.
  • Jung, K.H., An, G. and Ronald, P.C. (2008) Towards a better bowl of rice: assigning function to tens of thousands of rice genes. Nat. Rev. Genet. 9, 91101.
  • Jung, K.H., Jeon, J.S. and An, G. (2011) Web tools for rice transcriptome analyses. J. Plant. Biol. 54, 6580.
  • Kim, D.H., Eu, Y.J., Yoo, C.M., Kim, Y.W., Pih, K.T., Jin, J.B., Kim, S.J., Stenmark, H. and Hwang, I. (2001) Trafficking of phosphatidylinositol 3-phosphate from the trans-Golgi network to the lumen of the central vacuole in plant cells. Plant Cell, 13, 287301.
  • Kim, S.R., Lee, D.Y., Yang, J.I., Moon, S. and An, G. (2009a) Cloning vectors for rice. J. Plant. Biol. 52, 7378.
  • Kim, S.R., Yang, J.I., Moon, S., Ryu, C.H., An, K., Kim, K.M., Yim, J. and An, G. (2009b) Rice OGR1 encodes a pentatricopeptide repeat-DYW protein and is essential for RNA editing in mitochondria. Plant J. 59, 738749.
  • Lee, S., Jung, K.H., An, G. and Chung, Y.Y. (2004) Isolation and characterization of a rice cysteine protease gene, OsCP1, using T-DNA gene-trap system. Plant Mol. Biol. 54, 755765.
  • Lee, D.Y., Lee, J., Moon, S., Park, S.Y. and An, G. (2007) The rice heterochronic gene SUPERNUMERARY BRACT regulates the transition from spikelet meristem to floral meristem. Plant J. 49, 6478.
  • Lee, J., Lee, H., Kim, J., Lee, S., Kim, D.H., Kim, S. and Hwang, I. (2011) Both the hydrophobicity and a positively charged region flanking the C-terminal region of the transmembrane domain of signal-anchored proteins play critical roles in determining their targeting specificity to the endoplasmic reticulum or endosymbiotic organelles in Arabidopsis cells. Plant Cell, 23, 15881607.
  • Li, N., Zhang, D.S., Liu, H.S. et al. (2006) The rice tapetum degeneration retardation gene is required for tapetum degradation and anther development. Plant Cell, 18, 29993014.
  • Liu, X.C. and Dickinson, H.G. (1989) Cellular energy levels and their effect on male cell abortion in cytoplasmically male sterile lines of Petunia hybrida. Sex. Plant Reprod. 2, 167172.
  • Liu, Z., Bao, W., Liang, W., Yin, J. and Zhang, D. (2010) Identification of gamyb-4 and analysis of the regulatory role of GAMYB in rice anther development. J Integr Plant Biol 52, 670678.
  • Maheshwari, P. (1950) An Introduction to the Embryology of Angiosperms, 1st edn. New York: McGraw-Hill.
  • Mamun, E.A., Cantrill, L.C., Overall, R.L. and Sutton, B.G. (2005) Cellular organisation in meiotic and early post-meiotic rice anthers. Cell Biol. Int. 29, 903913.
  • Nonomura, K., Miyoshi, K., Eiguchi, M., Suzuki, T., Miyao, A., Hirochika, H. and Kurata, N. (2003) The MSP1 gene is necessary to restrict the number of cells entering into male and female sporogenesis and to initiate anther wall formation in rice. Plant Cell, 15, 17281739.
  • Nonomura, K., Nakano, M., Fukuda, T., Eiguchi, M., Miyao, A., Hirochika, H. and Kurata, N. (2004a) The novel gene HOMOLOGOUS PAIRING ABERRATION IN RICE MEIOSIS1 of rice encodes a putative coiled-coil protein required for homologous chromosome pairing in meiosis. Plant Cell, 16, 10081020.
  • Nonomura, K.I., Nakano, M., Murata, K., Miyoshi, K., Eiguchi, M., Miyao, A., Hirochika, H. and Kurata, N. (2004b) An insertional mutation in the rice PAIR2 gene, the ortholog of Arabidopsis ASY1, results in a defect in homologous chromosome pairing during meiosis. Mol. Genet. Genom. 271, 121129.
  • Nonomura, K., Morohoshi, A., Nakano, M., Eiguchi, M., Miyao, A., Hirochika, H. and Kurata, N. (2007) A germ cell specific gene of the ARGONAUTE family is essential for the progression of premeiotic mitosis and meiosis during sporogenesis in rice. Plant Cell, 19, 25832594.
  • Paetzel, M., Karla, A., Strynadka, N.C. and Dalbey, R.E. (2002) Signal peptidases. Chem. Rev. 102, 45494580.
  • Parish, R.W. and Li, S.F. (2010) Death of a tapetum: a programme of developmental altruism. Plant Sci. 178, 7389.
  • Park, J.J., Yi, J., Yoon, J., Cho, L.H., Ping, J., Jeong, H.J., Cho, S.K., Kim, W.T. and An, G. (2011) OsPUB15, an E3 ubiquitin ligase, functions to reduce cellular oxidative stress during seedling establishment. Plant J. 65, 194205.
  • Ryu, C.H., You, J.H., Kang, H.G., Hur, J., Kim, Y.H., Han, M.J., An, K., Chung, B.C., Lee, C.H. and An, G. (2004) Generation of T-DNA tagging lines with a bidirectional gene trap vector and the establishment of an insertion-site database. Plant Mol. Biol. 54, 489502.
  • Scott, R.J., Spielman, M. and Dickinson, H.G. (2004) Stamen structure and function. Plant Cell, 16(Suppl), S46S60.
  • Sorensen, A., Guerineau, F., Canales-Holzeis, C., Dickinson, H.G. and Scott, R.J. (2002) A novel extinction screen in Arabidopsis thaliana identifies mutant plants defective in early microsporangial development. Plant J. 29, 581594.
  • Tang, W.H., Tang, X.A., Zhang, Z.Y., Zhang, W.J., Zhao, X.M., Li, X.A., Zhang, D. and Liu, Q.Q. (2010) Global gene profiling of laser-captured pollen mother cells indicatesmolecular pathways and gene subfamilies involved in rice meiosis. Plant Physiol. 154, 18551870.
  • Tsuchiya, T., Toriyama, K., Nasrallah, M.E. and Ejiri, S. (1992) Isolation of genes abundantly expressed in rice anthers at the microspore stage. Plant Mol. Biol. 20, 11891193.
  • Wan, L., Zha, W., Cheng, X., Liu, C., Lv, L., Wang, Z., Du, B., Chen, R., Zhu, L. and He, G. (2011) A rice beta-1,3-glucanase gene Osg1 is required for callose degradation in pollen development. Planta 233, 309323.
  • Wang, M., Wang, K., Tang, D., Wei, C., Li, M., Shen, Y., Chi, Z., Gu, M., and Cheng, Z. (2010) The central element protein ZEP1 of the synaptonemal complex regulates the number of crossovers during meiosis in rice. Plant Cell 22, 417430.
  • Yamaguchi, T., Hayashi, T., Nakayama, K. and Koike, S. (2006) Expression analysis of genes for callose synthases and rho-type small GTP-binding proteins that are related to callose synthesis in rice anther. Biosci. Biotechnol. Biochem. 70, 639645.
  • Yang, S.L., Xiea, L.F., Mao, H.Z., Puah, C.S., Yang, W.C., Jiang, L.X., Sundaresan, V. and Ye, D. (2003) TAPETUM DETERMINANT1 is required for cell specialization in the Arabidopsis anther. Plant Cell, 15, 27922804.
  • Yuan, W., Li, X., Chang, Y., Wen, R., Chen, G., Zhang, Q. and Wu, C. (2009) Mutation of the rice gene PAIR3 results in lack of bivalent formation in meiosis. Plant J. 59, 303315.
  • Zhang, D.B. and Wilson, Z.A. (2009) Stamen specification and anther development in rice. Chinese Sci. Bull. 54, 23422353.
  • Zhang, W., Sun, Y., Timofejeva, L., Chen, C., Grossniklaus, U. and Ma, H. (2006) Regulation of Arabidopsis tapetum development and function by DYSFUNCTIONAL TAPETUM1 (DYT1) encoding a putative bHLH transcription factor. Development, 133, 30853095.
  • Zhang, D.S., Liang, W.Q., Yuan, Z., Li, N., Shi, J., Wang, J., Liu, Y.M., Yu, W.J. and Zhang, D.B. (2008) Tapetum degeneration retardation is critical for aliphatic metabolism and gene regulation during rice pollen development. Mol. Plant, 1, 599610.
  • Zhang, D., Liang, W., Yin, C., Zong, J. and Gu, F. (2010) OsC6, encoding a lipid transfer protein, is required for postmeiotic anther development in rice. Plant Physiol. 154, 149162.
  • Zhao, D.Z., Wang, G.F., Speal, B. and Ma, H. (2002) The excess microsporocytes1 gene encodes a putative leucine-rich repeat receptor protein kinase that controls somatic and reproductive cell fates in the Arabidopsis anther. Genes Dev. 16, 20212031.
  • Zhao, X.A., de Palma, J., Oane, R., Gamuyao, R., Luo, M., Chaudhury, A., Herve, P., Xue, Q. and Bennett, J. (2008) OsTDL1A binds to the LRR domain of rice receptor kinase MSP1, and is required to limit sporocyte numbers. Plant Journal 54, 375387.
  • Zhou, S.R., Wang, Y., Li, W.C. et al. (2011) Pollen Semi-Sterility1 encodes a kinesin-1-like protein important for male meiosis, anther dehiscence, andfertility in rice. Plant Cell, 23, 111129.
  • Zhu, J., Chen, H., Li, H., Gao, J.F., Jiang, H., Wang, C., Guan, Y.F. and Yang, Z.N. (2008) Defective in Tapetal Development and Function 1 is essential for anther development and tapetal function for microspore maturation in Arabidopsis. Plant J. 55, 266277.

Supporting Information

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

Figure S1. Expression patterns of DTM1. (a) Transcript levels in various organs. (b) Transcript levels in floral organs. Y axis, gene expression relative to rice Ubiquitin (Os03g0234200) transcript level. Results are average of three independent experiments.

Figure S2. Callose deposition in anthers and expression of callose related genes. (a, b) Aniline blue staining of WT (a) and dtm1-1 (b) anthers at Stage 6–8a. Arrows indicate callose wall. Bars: 10 μm. (c) Transcript level of Osg1, a callase gene. (d) Transcript level of OsGSL5, a callose synthase gene. Y axis, gene expression relative to rice Ubiquitin (Os03g0234200) transcript level. Results are average of three independent experiments.

Figure S3. TEM analyses of anthers. (a–d) Cross sections of WT (a, c) and dtm1-1 (b, d) anther walls at early Stage 6 (a, b) and Stage 7 (c, d). (e–h) Magnified images of WT (e, g) and dtm1-1 (f, h) tapetal cells at early Stage 6 (e, f) and Stage 7 (g, h). (i–l) Magnified images of WT (i, k) and dtm1-1 (j, l) PMC at early Stage 6 (i, j) and Stage 7 (k, l). Arrowheads indicate undeveloped and unusual cellular organs. En, endothecial cell; Ep, epidermal cell; lb, lipid body; M, mitochondrion; Ml, middle layer; n, nucleus; p, plastid; PT, peritapetal region; T, tapetum cell. Bars: 1 μm (a–h) or 5 μm (i–l).

Figure S4. TEM analyses of 3rd leaf from WT and dtm1. (a) Mesophyll cell of WT. (b) Magnified image of solid box in a showing concentrated circular form of ER. (c) Magnified image of dashed box in a showing rough ER. (d) Mesophyll cell of dtm1. (e) Magnified image of solid box in d showing concentrated circular form of ER. (f) Magnified image of dashed box in d showing rough ER. Ch, chloroplast; Nu, nucleus. Bars: 5 μm (a, d), 1 μm (b, e), or 0.5 μm (c, f).

Figure S5. Chromosome spreads of WT male meiocytes. (a) diakinesis. (b) metaphase I. (c) anaphase I. (d) telophase I. Bars: 10 μm.

Table S1. Cross pollination test.

Table S2. Primers used for real-time PCR.

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_4864_sm_FigureS1-S5-TableS1-S2.pdf2826KSupporting info item
TPJ_4864_sm_Supporting-Information-legend.doc32KSupporting 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.