Research ArticleResearch Article

Spatiotemporal Production of Reactive Oxygen Species by NADPH Oxidase Is Critical for Tapetal Programmed Cell Death and Pollen Development in Arabidopsis

Hong-Tao Xie, Zhi-Yuan Wan, Sha Li, Yan Zhang

Published May 2014. DOI: https://doi.org/10.1105/tpc.114.125427

Abstract

Male sterility in angiosperms has wide applications in agriculture, particularly in hybrid crop breeding and gene flow control. Microspores develop adjacent to the tapetum, a layer of cells that provides nutrients for pollen development and materials for pollen wall formation. Proper pollen development requires programmed cell death (PCD) of the tapetum, which requires transcriptional cascades and proteolytic enzymes. Reactive oxygen species (ROS) also affect tapetal PCD, and failures in ROS scavenging cause male sterility. However, many aspects of tapetal PCD remain unclear, including what sources generate ROS, whether ROS production has a temporal pattern, and how the ROS-producing system interacts with the tapetal transcriptional network. We report here that stage-specific expression of NADPH oxidases in the Arabidopsis thaliana tapetum contributes to a temporal peak of ROS production. Genetic interference with the temporal ROS pattern, by manipulating RESPIRATORY-BURST OXIDASE HOMOLOG (RBOH) genes, affected the timing of tapetal PCD and resulted in aborted male gametophytes. We further show that the tapetal transcriptional network regulates RBOH expression, indicating that the temporal pattern of ROS production intimately connects to other signaling pathways regulated by the tapetal transcriptional network to ensure the proper timing of tapetal PCD.

INTRODUCTION

The control of male fertility in angiosperms has attracted wide attention in the past decades due to its vital importance for crop breeding, hybrid production, and gene flow control for transgenes (Wilson and Zhang, 2009). Extensive studies have shown that the finely tuned process of male gametophytic development requires contributions by and collaborations from the surrounding sporophytic cells (Wilson and Zhang, 2009; Parish and Li, 2010; Chang et al., 2011). Microsporocytes in anther locules are surrounded by four somatic cell layers, the epidermis, endothecium, middle layer, and tapetum, during stages 1 to 5 in Arabidopsis thaliana (Sanders et al., 1999) or stages 1 to 6 in rice (Oryza sativa) (Zhang et al., 2011). Tapetal cells develop adjacent to microspores. Following the completion of meiosis, tapetal cells play a crucial role in microsporogenesis by supplying proteins, lipids, and pigments, first through secretion and later through degradation (Parish and Li, 2010). In Arabidopsis, visible tapetal degeneration starts at stage 10 and lasts to stage 11, when pollen mitotic division occurs (Sanders et al., 1999; Parish and Li, 2010). This differs slightly from rice, in which tapetal degeneration was detected as early as the tetrad stage (Zhang et al., 2011).

The production of viable pollen requires the proper timing of tapetal degradation, which occurs via developmentally regulated programmed cell death (PCD). Tapetal PCD is tightly controlled by evolutionarily conserved transcriptional cascades (Wilson and Zhang, 2009; Parish and Li, 2010; Ma et al., 2012). In Arabidopsis, spatiotemporal expression of genes encoding transcription factors controls tapetal function (Wilson and Zhang, 2009; Parish and Li, 2010; Chang et al., 2011; Zhu et al., 2011). Mutations of these genes, such as DYSFUNCTIONAL TAPETUM1 (DYT1) (Zhang et al., 2006), DEFECTIVE IN TAPETAL DEVELOPMENT AND FUNCTION1 (TDF1/MYB35) (Feng et al., 2012), MYB33 and MYB65 (Millar and Gubler, 2005), ABORTED MICROSPORES (AMS) (Xu et al., 2010), MYB80 (also known as MYB103) (Higginson et al., 2003; Zhang et al., 2007; Phan et al., 2011), and MALE STERILITY1 (MS1) (Wilson et al., 2001; Vizcay-Barrena and Wilson, 2006; Ito et al., 2007; Yang et al., 2007), all led to pollen abortion and male sterility. Elegant transcriptomic (Xu et al., 2010; Feng et al., 2012; Ma et al., 2012) and genetic (Zhu et al., 2011) studies integrated these key transcription factors into a transcriptional network regulating tapetal PCD, in which linear as well as complex interactions among these transcription factors may be delicately intertwined to ensure the proper timing of tapetal PCD (Zhu et al., 2011; Ma et al., 2012). Most components in the Arabidopsis tapetal transcriptional network have functional counterparts in rice (Li et al., 2006; Zhang et al., 2008; Li et al., 2011; Phan et al., 2012; Niu et al., 2013), suggesting its evolutionary conservation.

Genes encoding proteolytic enzymes are often targets of the tapetal transcriptional network (Li et al., 2006; Phan et al., 2011; Niu et al., 2013). Mutations of rice TAPETUM DEGENERATION RETARDATION (TDR), a counterpart of Arabidopsis AMS (Li et al., 2006; Zhang et al., 2008), resulted in delayed tapetal PCD and pollen abortion (Li et al., 2006; Zhang et al., 2008). Interestingly, genes encoding a Cys protease and a protease inhibitor are likely targeted by TDR (Li et al., 2006). Rice ETERNAL TAPETUM1, operating downstream of TDR, was recently found to regulate the expression of two aspartic proteases, both of which promoted PCD (Niu et al., 2013). In Arabidopsis, genome-wide transcriptomic analyses identified several Cys protease–encoding genes as potential targets of MS1 (Yang et al., 2007). Indeed, a direct target of MYB80, UNDEAD, encodes an A1 aspartic protease whose functional loss resulted in premature tapetal PCD (Phan et al., 2011). These data support the existence of a protease-mediated pathway for the execution of plant developmental PCD (Lam, 2004).

Another universal theme in developmental PCD is the involvement of reactive oxygen species (ROS) (Beers and McDowell, 2001; Lam, 2004; Gechev et al., 2006; De Pinto et al., 2012). ROS, such as superoxide and hydrogen peroxide, are critical signaling molecules affecting a large number of proteins through posttranslational modifications, transcriptional changes, or activity changes (Neill et al., 2002; Lam, 2004; Mittler et al., 2004, 2011; Torres and Dangl, 2005; Suzuki et al., 2011). It was shown that rice anthers display ROS levels that increase until stages 8 and 9 and then decrease to complete absence at stage 11 (Hu et al., 2011). This temporal change of ROS was abolished in mads3 mutant anthers, in which tapetal PCD occurs prematurely (Hu et al., 2011). One target of MADS3 is MT-1-4b, encoding a protein with ROS-scavenging activity in vitro (Hu et al., 2011). Also, the determining factor of rice wild abortive cytoplasmic male sterility encodes a tapetal mitochondrial protein, WA352, whose interaction with the nucleus-encoded mitochondrial protein COX11 inhibits its ROS-scavenging activity, resulting in premature tapetal PCD and pollen abortion (Luo et al., 2013).

Although these data hinted at the involvement of ROS in tapetal PCD (Hu et al., 2011; Luo et al., 2013), it remains unclear how ROS are generated, whether ROS production undergoes temporal changes in the tapetum, and how the ROS-producing system interacts with the tapetal transcriptional network. We report here that tapetal PCD requires temporal control of ROS levels by the spatiotemporal expression of RESPIRATORY-BURST OXIDASE HOMOLOG (RBOH) genes. RBOHs encode NADPH oxidases that localize at the plasma membrane to produce O2−, which then mostly converts to apoplastic H2O2 (Keller et al., 1998; Torres and Dangl, 2005; Marino et al., 2012). RBOHs play crucial roles in cell growth, development, and abiotic and biotic stress responses (Suzuki et al., 2011; Marino et al., 2012; Lee et al., 2013). Our work indicates that the temporal ROS pattern produced by NADPH oxidases, together with other signaling pathways, is under complex control of the tapetal transcriptional network to ensure the proper timing of tapetal PCD.

RESULTS

Spatiotemporal Expression of RBOHE

The Arabidopsis genome encodes 10 RBOHs. We identified RBOHE (At1g19230) as an anther-preferential or tapetum-enriched gene by mining several microarray studies of the tapetal transcriptome (Wijeratne et al., 2007; Feng et al., 2012; Ma et al., 2012). To verify the tapetum-enriched expression of RBOHE, we generated more than 20 individual ProRBOHE:GUS (for β‑glucuronidase) reporter lines and analyzed GUS expression by histochemical analysis. The ProRBOHE:GUS transgenic anthers showed GUS signals in the tapetal layer during anther developmental stages 6 to 11 (Figure 1A), a spatiotemporal expression pattern confirmed by RNA in situ analysis (Figures 1B to 1D). These lines of evidence demonstrated the temporal expression of RBOHE in tapetum, implying its involvement in late tapetal function.

Figure 1.
Figure 1.

Expression of RBOHE in the Tapetum during Anther Stages 6 to 11.

(A) Histochemical staining of representative anther sections from ProRBOHE:GUS transgenic plants at different developmental stages.

(B) to (D) RNA in situ analyses of RBOHE expression at different developmental stages using an antisense probe either in the wild type (B) or in rbohe-1 (D). Wild-type anthers at corresponding stages were incubated with a sense probe as controls (C). Arrowheads point to the tapetal layer.

Bars = 20 µm.

Functional Loss of RBOHE Resulted in Sporophytic Male Defects

The spatiotemporal expression of RBOHE in tapetum suggested its involvement in pollen development. To test this, we characterized two T-DNA insertion mutants of RBOHE, rbohe-1 and rbohe-2. No transcript was detectable by RT-PCR in rbohe-1 and rbohe-2 mutants using gene-specific primers flanking the insertion sites (Figures 2A and 2B) or by RNA in situ hybridization in rbohe-1 (Figure 1D), supporting the idea that these are null alleles for RBOHE. F1 plants generated from crosses between rbohe-1 and rbohe-2 showed similar pollen developmental defects to either parent (Supplemental Figure 1), indicating that they are allelic. However, rbohe-1 contains another insertion that caused a long-petiole phenotype. Thus, rbohe-2 was analyzed in detail and used to generate double mutants in the following studies, unless noted otherwise.

Figure 2.
Figure 2.

Functional Loss of RBOHE Resulted in Pollen Developmental Defects.

(A) Schematic illustration of T-DNA insertions within the RBOHE genomic region. Arrows indicate the binding sites of primers used for RT-PCR.

(B) Transcript analysis by RT-PCR showing the loss of RBOHE expression in both mutant alleles. ACTIN2 (ACT2) was used as the internal control. gDNA, genomic DNA.

(C) to (E) Scanning electron micrographs of pollen grains from the wild type (C), rbohe-1 (D), and rbohe-2 (E).

(F) to (H) Scanning electron micrographs of a single pollen grain from the wild type (F), rbohe-1 (G), and rbohe-2 (H).

(I) to (L) Transmission electron micrographs of pollen from stage 11 anthers of the wild type ([I] and [K]) or rbohe-2 ([J] and [L]). dPG, degenerated pollen grain; In, intine; Ne, nexine; PC, pollen coat; PG, pollen grains; Se, sexine.

Bars in (C) to (E) = 20 µm; bars in (F) to (H) = 5 µm; bars in (I) and (J) = 500 nm; bars in (K) and (L) = 2 µm.

A substantial portion of pollen grains from both mutants aborted, as judged by scanning electron microscopy (Figures 2C to 2H) and Alexander dye staining (Supplemental Figure 1). Scanning electron microscopy of mature pollen (Figures 2C to 2H) showed that, unlike the well-organized reticular exine structure of wild-type pollen (Figures 2C and 2F), mutant pollen showed irregular exine patterns and a deformed shape (Figures 2D, 2E, 2G, and 2H). Further examination by transmission electron microscopy (Figures 2I to 2L) confirmed the pollen developmental defects in the RBOHE loss-of-function mutants. Wild-type pollen showed a layered structure with intine, nexine, sexine, and pollen coat from inner to outer layers (Figures 2I and 2K). A substantial portion of mutant pollen grains contained intact intine but hardly any nexine structure (Figures 2J and 2L). Both pollen coat and sexine structure were nearly absent in the mutant pollen (Figures 2J and 2L). As a result of the significantly increased pollen abortion rates, RBOHE loss of function resulted in partial sterility (Supplemental Figure 1).

To find out whether the defective pollen development was sporophytic or gametophytic, we first analyzed pollen from heterozygous mutants of RBOHE and found that they have a normal pollen coat (Supplemental Figure 1). Pollen germination of RBOHE heterozygous mutants was comparable to that of the wild type but significantly reduced in its homozygous mutants (Supplemental Figure 1), confirming that RBOHE loss of function resulted in sporophytic rather than gametophytic male defects.

Functional Loss of RBOHE Resulted in Delayed Tapetal Degeneration

Because the spatiotemporal expression of RBOHE occurred at the same time as tapetal degeneration (Phan et al., 2011) and functional loss of RBOHE resulted in pollen abortion (Figure 2), we wondered whether the sporophytic defects in RBOHE loss-of-function mutants were due to defective tapetal degeneration.

To test this hypothesis, we first aimed at identifying the developmental defects of rbohe-2 by histological analyses. We compared transverse semithin sections of developing wild-type and rbohe-2 anthers from stages 6 to 12. The tapetum of rbohe-2 (Figure 3F) was morphologically comparable to that of the wild type at stages 6 and 7 (Figure 3A). However, rbohe-2 tapetum remained highly densely stained (Figure 3G) at stage 8, when wild-type tapetal cells began to shrink (Figure 3B). The middle layer persisted in rbohe-2 (Figure 3G), but it largely disappeared in the wild type (Figure 3B). Wild-type tapetum gradually degenerated into a thin and broken layer at stage 11 (Figures 3C and 3D) and disappeared completely at stage 12 (Figure 3E), accompanied by septum breakage (Figure 3E). By contrast, rbohe-2 tapetum remained intact and became slightly hypertrophic until stage 11 (Figures 3H and 3I). At stage 12, a thin and broken layer of tapetum was still visible in rbohe-2 (Figure 3J). These histological analyses indicated a delayed degeneration of tapetal cells in rbohe-2.

Figure 3.
Figure 3.

Functional Loss of RBOHE Resulted in Delayed Tapetal Degeneration.

(A) to (E) Transverse semithin sections of wild-type anthers at developmental stages 6 and 7 (A), stages 8 and 9 (B), stage 10 (C), stage 11 (D), and stage 12 (E).

(F) to (J) Transverse semithin sections of rbohe-2 anthers at developmental stages 6 and 7 (F), stages 8 and 9 (G), stage 10 (H), stage 11 (I), and stage 12 (J).

Arrowheads point to the tapetal layer.

(K) to (M) Transmission electron micrographs of wild-type anthers at anther stages 8 and 9 (K) or stage 11 ([L] and [M]).

(N) to (P) Transmission electron micrographs of rbohe-2 anthers at anther stages 8 and 9 (N) or stage 11 ([O] and [P]).

Arrows point to the middle layer, which was not observed in stage 11 wild type ([L] and [M]) but persisted in stage 11 rbohe-2 anthers ([O] and [P]). Arrowheads point to the tapetal layer. E, elaioplast; Msp, microspore; T, tapetosome; V, vacuole.

Bars in (A) to (J) = 50 µm; bars in (K), (M), (N), and (P) = 1 µm; bars in (L) and (O) = 2 µm.

[See online article for color version of this figure.]

To find out more cytosolic details of the effects of RBOHE loss of function on the tapetal cells, we performed transmission electron microscopy on mutant anthers at late developmental stages. Consistent with the results of histological analyses, rbohe-2 mutants showed delayed tapetal degeneration. Microspores form an exine wall in the wild type at stage 9 (Figure 3K). Tapetal cells of the wild type at this stage accumulate numerous tapetosomes and elaioplasts (Hsieh and Huang, 2007), which eventually deposit their contents to developing microspores (Figure 3K). By contrast, rbohe-2 at the same stage contained a thicker tapetal layer (Figure 3N). At stage 11, wild-type tapetal cells degenerated, with disintegrated cellular organization (Figures 3L and 3M). By contrast, tapetal cells of rbohe-2 remained intact and contained numerous tapetosomes and elaioplasts, which failed to release into anther locules (Figures 3O and 3P). The development of microspores was significantly retarded in rbohe-2 (Figure 3O) as compared with that in the wild type (Figure 3L), confirming delayed tapetal degradation in rbohe-2.

Functional Redundancy in Pollen Development by RHD2

Despite the clear effect of RBOHE loss of function in tapetal degeneration, we wondered whether another RBOH functioned redundantly with RBOHE, as reported for other processes (Torres et al., 2002; Kwak et al., 2003). Close examination of anther transcriptomic data (Wijeratne et al., 2007) indicated that RHD2, also named RBOHC, which belongs to the same RBOH family (Torres and Dangl, 2005), was possibly also a tapetal gene, despite its involvement in root hair growth (Foreman et al., 2003).

To find out whether RHD2 was expressed in the tapetum, we performed RNA in situ hybridization in anthers of various developmental stages (Supplemental Figure 2). Positive signals were detected in tapetum from stages 6 to 10 (Supplemental Figure 2). We also generated promoter:GUS reporter lines for RHD2. Histochemical analyses of ProRHD2:GUS transgenic anthers confirmed the spatiotemporal expression of RHD2 by RNA in situ hybridization (Supplemental Figure 2). We note, however, that under the same analytical conditions, the expression level of RHD2 was much lower than that of RBOHE in the tapetum, suggesting a minor role of RHD2.

To test whether RHD2 was redundant with RBOHE in tapetum, we generated the rbohe-2 rhd2-1 double mutant by crossing rbohe-2 with rhd2-1, described previously as a null mutant for RHD2 (Foreman et al., 2003). Compared with the wild type (Figures 4A, 4D, and 4G), functional loss of RHD2 did not affect pollen development significantly (Figures 4B, 4E, and 4H; Supplemental Figure 1). In rbohe-2 rhd2-1, however, pollen development was severely compromised (Figures 4C, 4F, and 4I). More of the pollen appeared to be deformed in the double mutants, leading to significantly reduced fertility (Figure 4F; Supplemental Figure 1). The tapetum in the double mutant was comparably normal before stage 7 (Figures 4J and 4K). However, densely stained tapetal cells persisted in the double mutant until stage 12 (Figures 4L to 4P). Only at stage 13, when septum and stomium degenerated to break the anther wall, did the tapetal layer disappear in rbohe-2 rhd2-1 mutants (Figure 4Q). The middle layer disappeared at stage 8 in the wild type (Figure 3C) but was still visible in the double mutant even at stage 13 (Figure 4Q), likely a secondary effect of defective tapetal degeneration. These results showed that RHD2 plays a redundant, albeit minor, role in tapetal degeneration and pollen development compared with RBOHE.

Figure 4.
Figure 4.

RHD2/RBOHC Confers Functional Redundancy to RBOHE in Pollen Development.

(A) to (C) Alexander dye staining of mature anthers from the wild type (A), rhd2-1 (B), and rbohe-2 rhd2-1 (C).

(D) to (F) Scanning electron micrographs of dehiscing anthers from the wild type (D), rhd2-1 (E), and rbohe-2 rhd2-1 (F).

(G) to (I) Scanning electron micrographs of pollen grains from the wild type (G), rhd2-1 (H), and rbohe-2 rhd2-1 (I).

(J) to (Q) Transverse semithin sections of rbohe-2 rhd2-1 anthers at stages 5 and 6 (J), stage 7 (K), stage 8 (L), stage 9 (M), stage 10 (N), stage 11 (O), stage 12 (P), and stage 13 (Q). The septum breakage was delayed in the double mutant at stage 12 (P), likely a side effect due to incomplete degeneration of other anther cells. Arrows indicate the persistent middle layer even at stage 12 (P).

Bars in (A) to (C) = 100 µm; bars in (D) to (F) = 20 µm; bars in (G) to (I) = 5 μm; and bars in (J) to (Q) = 50 μm.

Overexpression of RBOHE Resulted in Precocious Tapetal Degeneration

To explore the possibility that the spatiotemporal RBOHE expression was crucial for tapetal degeneration and pollen development, we attempted to manipulate RBOHE levels by overexpressing RBOHE. RBOHE has two annotated splicing variants, and we cloned a novel RBOHE splicing form (RBOHE.3) by using anther cDNA libraries as a template for PCR (Supplemental Figure 3). RBOHE.3 potentially encodes an RBOHE with intact enzymatic sites, similar to other variants (Supplemental Figure 3). To prove that RBOHE.3 has NADPH oxidase function, we used a tobacco (Nicotiana benthamiana) infiltration system to examine its in vitro activity (Supplemental Figure 3). Nitrotetrazolium blue chloride (NBT) staining of tobacco leaves expressing RBOHE.3 showed enhanced ROS production, whereas leaves expressing a C-terminal truncated RBOHE.3 that lacks its enzymatic domains and leaves expressing the empty vector did not show enhanced ROS production (Supplemental Figure 3), indicating that RBOHE.3 encodes a functional NADPH oxidase.

To manipulate the expression level of RBOHE, we used the rice promoter ProOsg6B (Tsuchiya et al., 1994; Kawanabe et al., 2006), which is active in the tapetum from stages 6 to 10 (Figure 5A). We generated 35 transgenic lines expressing RBOHE.3 driven by ProOsg6B, among which 27 individual transgenic lines (OX-S1 to OX-S27) showed complete pollen abortion and complete sterility, while eight individual transgenic lines (OX-M1 to OX-M8) showed significantly increased pollen abortion (Supplemental Figure 4). Two representative transgenic lines, OX-M1 for medium overexpression and OX-S1 for strong overexpression, were chosen for further studies based on transcript analyses that correlate RBOHE expression with pollen abortion rates (Supplemental Figure 4).

Figure 5.
Figure 5.

Overexpression of RBOHE Resulted in Tapetal Dysfunction.

(A) Histochemical staining analysis of anthers (top) or anther sections (bottom) from the ProOsg6B:GUS transgenic plants at different developmental stages. Arrowheads point to the tapetal layer. In total, 15 individual ProOsg6B:GUS transgenic lines were analyzed, and representative images are shown.

(B) Transverse semithin sections of ProOsg6B:RBOHE anthers at different developmental stages. Arrowheads point to the tapetal layer. Arrows point to the middle layer that persists until late stages in the ProOsg6B:RBOHE anthers. Three individual ProOsg6B:RBOHE lines were analyzed, similar results were obtained, and representative images are presented.

(C) Transmission electron micrographs of ProOsg6B:RBOHE anthers at stages 8, 9, and 10. dMsp, degenerated microspore; ML, middle layer; Msp, microspore; Ta, tapetum.

Bars in (A) (anthers) = 100 µm; bars in (A) (anther sections) and (B) = 20 µm; bars in (C) = 2 µm.

Histological analyses of tapetal cells in developing anthers showed that ProOsg6B:RBOHE anthers were indistinguishable from those of the wild type before anther stage 6 (Figure 5B). However, starting from anther stage 7 after tetrads were released into anther locules (Figure 5B), tapetal cells were densely stained in ProOsg6B:RBOHE anthers (Figure 5B). In addition to lightly stained microspores, anther locules were filled with cellular debris (Figure 5B), suggesting cellular disintegration. At stage 9, the tapetal layer of ProOsg6B:RBOHE anthers was hardly recognizable, with more cellular debris filling anther locules (Figure 5B). At stage 10, microspores were highly vacuolated and mostly degenerated (Figure 5B). ProOsg6B:RBOHE anther locules were finally empty, with only degenerating microspores, at stage 11 (Figure 5B). By transmission electron microscopy, we found that, unlike the tapetal cells of the wild type, which showed a clear cell wall structure at stages 8 and 9 (Figure 3K) and partially degenerated tapetal cells at stage 11 (Figure 3M), no cell wall structure was detected in the tapetal layer of ProOsg6B:RBOHE anthers (Figure 5C). In addition, overexpression of RBOHE caused early degeneration of tapetal cells such that the tapetal layer was not present at stage 10 (Figure 5C). In addition, microspores degenerated as early as at stage 9 in ProOsg6B:RBOHE anthers (Figure 5C), likely due to dysfunction of the tapetum. These results suggested that overexpressing RBOHE induced precocious tapetal degeneration (Figure 5C).

RBOHE Loss or Gain of Function Interfered with the Timing of Tapetal PCD

To find out whether the delayed or precocious tapetal degeneration reflected the different timing of tapetal PCD, we performed terminal deoxynucleotidyl transferase–mediated dUTP nick-end labeling (TUNEL) assays on RBOHE loss- or gain-of-function anthers, in which positive signals indicate cells undergoing massive DNA fragmentation.

As reported previously (Vizcay-Barrena and Wilson, 2006; Phan et al., 2011), no TUNEL-positive signal could be detected in the wild type before stage 10 (Figure 6A). At stage 10, tapetal cells underwent PCD, leading to TUNEL-positive signals (Figure 6A). At stage 11, not only tapetal cells but also septum cells degenerated to produce TUNEL-positive signals (Figure 6A). By contrast, no TUNEL-positive signals could be detected at stage 10 or before in rbohe-2 (Figure 6B). Only at stage 11, when vascular and septum cells degenerated, did TUNEL-positive signals appear in the tapetal layer in rbohe-2 (Figure 6B), indicating delayed tapetal PCD in rbohe-2. In contrast with RBOHE loss of function, RBOHE gain of function resulted in precocious tapetal PCD, as indicated by the appearance of TUNEL-positive signals as early as at stage 9 (Figure 6C). In addition to tapetal cells, TUNEL-positive signals were also visible inside anther locules from degenerating microspores (Figure 6C). These results indicated that RBOHE is crucial for the proper timing of tapetal PCD.

Figure 6.
Figure 6.

RBOHE Loss or Gain of Function Interfered with the Timing of Tapetal PCD.

Fluorescence microscopy of DNA fragmentation detected using the TUNEL assays in sections of wild-type anthers (A), rbohe-2 anthers (B), and ProOsg6B:RBOHE anthers (C) is shown at different stages. Three individual ProOsg6B:RBOHE lines were analyzed, similar results were obtained, and representative images are presented. Green fluorescence indicates TUNEL-positive signals, while red fluorescence indicates propidium iodide staining. Corresponding bright-field images are placed together with the fluorescence images. Arrowheads point to the tapetal layer. Arrows in (C) indicate TUNEL-positive signals from degenerated microspores that start from stage 9 in ProOsg6B:RBOHE anthers. Bars = 20 μm.

RBOHE Loss or Gain of Function Affected Tapetal ROS Levels

The delayed or precocious tapetal PCD by RBOHE loss or gain of function suggested the importance of the spatiotemporal production of ROS by NADPH oxidases. To find out whether this was the case, we analyzed anther ROS levels by NBT staining of superoxide anion and 2′,7′-dichlorodihydrofluorescein diacetate (H2DCF-DA) staining of free radicals. We classified anthers into six groups based on their sizes to categorize anthers at different developmental stages (Supplemental Table 1).

Correlating with the temporal expression of RBOHE, developing Arabidopsis anthers showed a temporal ROS pattern similar to what has been reported in rice anthers (Hu et al., 2011). The ROS level gradually increased starting from stages 5 and 6, reaching the highest level at approximately stages 8 and 9, and then decreasing until stage 11 (Figure 7; Supplemental Figure 5). By contrast, ROS signals did not show a temporal rise in either rbohe-2 or rbohe-2 rhd2-1 (Figure 7A). Both rbohe-2 and rbohe-2 rhd2-1 displayed significantly reduced ROS levels during stages 6 to 11 compared with the wild type (Figure 7B). Between rbohe-2 and rbohe-2 rhd2-1, the only significant difference was detected at anther stages 8 and 9 (Figure 7B), suggesting that RHD2 contributes to ROS production mainly during the ROS peak.

Figure 7.
Figure 7.

RBOHE Loss or Gain of Function Affected Tapetal ROS Levels.

(A) NBT staining analysis of ROS in anthers at various developmental stages from the wild type, rbohe-2, rbohe-2 rhd2-1, or the ProOsg6B:RBOHE transgenic plant OX-S1. Classification of anther groups (1 to 6) is based on anther sizes as described in Supplemental Table 1. Bars = 100 µm.

(B) Fluorescence quantification of anther ROS levels by H2DCF-DA staining of anthers at corresponding stages. ROS signals were hardly detectable in group 6 by H2DCF-DA staining; thus, they were not quantified. Corresponding anther stages are listed below the group number. Different letters on top of the columns within each group represent significant difference (Student’s t test, P < 0.01).

In comparison with its loss of function, overexpression of RBOHE increased ROS levels significantly around stage 6 (Figure 7A; Supplemental Figure 5) and interfered with the temporal ROS pattern seen in the wild type (Figure 7B). Instead of the ROS peak around stages 8 and 9 in the wild type, RBOHE overexpression resulted in relatively uniform ROS levels from stages 6 to 10 (Figure 7B). Considering the significantly increased pollen abortion rates of overexpression lines (Supplemental Figure 4), this result suggests the importance of the spatiotemporal regulation of ROS levels for pollen development.

AMS and MYB80 Regulate RBOHE Expression

Anther transcriptomic studies indicated that RBOHE was downregulated in dyt1 and ams anthers (Feng et al., 2012) but upregulated in ms188-1 (an MYB80 mutant) anthers (Phan et al., 2011), suggesting a complex regulatory network by the tapetal transcription factors (Wilson et al., 2001; Parish and Li, 2010; Zhu et al., 2011).

To provide clues to the transcriptional control of the NADPH oxidase–coding genes in the tapetum, we performed in silico analysis on the promoter sequences of RBOHE and RHD2. AMS and MYB80 are epistatic to DYT1 and TDF1 (Zhu et al., 2011). Both have been extensively studied for their cis-elements (Xu et al., 2010; Phan et al., 2011). The promoter sequences of RBOHE contain four AMS core binding motifs and two MYB80 core binding motifs and the RHD2 promoter contains two AMS motifs (Supplemental Figure 6), suggesting potential regulation of the NADPH oxidase–coding genes by AMS and MYB80.

Next, we analyzed RBOHE expression by histochemical analyses of the ProRBOHE:GUS reporter lines in the wild type, in ams, or in ms188-1 to examine their potential regulation of the spatiotemporal expression of RBOHE. Consistent with the anther transcriptomic data (Phan et al., 2011; Feng et al., 2012), RBOHE was reduced in ams but enhanced in ms188-1 (Figure 8A). The transcriptional changes of RBOHE were further verified by RNA in situ hybridization using stages 6 and 7 anthers from the wild type, ams, or ms188-1 (Figure 8B). We further used quantitative real-time PCR (qPCR) to measure RBOHE expression in wild-type, ams, or ms188-1 anthers at stages 5 to 8 (Figure 8C). We also verified that RBOHE was reduced in dyt1 and tdf1 by ProRBOHE:GUS reporter analysis and by RNA in situ hybridization (Supplemental Figure 7). All results demonstrated the different effects of mutations at AMS and MYB80 on RBOHE expression (i.e., downregulated in ams and upregulated in ms188-1), indicating that AMS and MYB80 regulate the transcription of RBOHE, directly or indirectly.

Figure 8.
Figure 8.

Expression of RBOHE Depends on the Tapetal Transcription Factors AMS and MYB80.

(A) Representative histochemical staining of ProRBOHE:GUS transgenic inflorescences in the wild type, ams, and ms188-1. The same ProRBOHE:GUS transgenes were introduced into different genetic backgrounds by crosses for comparison. To highlight differences and avoid signal saturation, a staining duration of 12 or 8 h was used for ProRBOHE:GUS;ams or ProRBOHE:GUS;ms188-1, respectively, to compare with ProRBOHE:GUS anthers of the same staining protocol. Bars = 500 µm.

(B) In situ analysis of RBOHE expression at stages 6 and 7 using an antisense probe in the wild type, ams, and ms188-1. Anther sections of the wild type, ams, and ms188-1 at the same stage were incubated with a sense probe as a control. Arrowheads point to the tapetal layer. Bars = 50 μm.

(C) qPCR analysis of RBOHE expression in wild-type, ams, and ms188-1 anthers at stages 5 to 8. Error bars represent se. Different letters on top of the columns represent significantly different groups (Student’s t test, P < 0.001).

Genetic Interaction between MYB80 and RBOHE

The enhanced expression of RBOHE in ms188-1 suggested a negative regulation of RBOHE by MYB80. Interestingly, we found that ms188-1 showed elevated ROS levels at early stages of anther development (Supplemental Figure 8), correlating with an enhanced expression of RBOHE in ms188-1 (Figure 8). This result suggested the genetic interaction between MYB80 and RBOHE.

To test this hypothesis, we generated the ms188-1 rbohe-2 double mutant. Analyses of ROS by either NBT or H2DCF-DA staining showed that introducing rbohe-2 into ms188-1 significantly reduced the ROS level (Supplemental Figure 8), suggesting an epistatic interaction between RBOHE and MYB80 in ROS production.

To gain more insights into their genetic interactions, we performed histological analyses on the ms188-1 rbohe-2 double mutant. Transverse semithin sections of developing ms188-1 anthers showed defective tetrad separation and early degeneration of tapetum (Figures 9A and 9C to 9E), as reported (Higginson et al., 2003; Zhang et al., 2007; Phan et al., 2011). Tapetum and microspores degenerated precociously in ms188-1 (Figures 9A and 9C to 9E). In comparison, microspores degenerated as early in ms188-1 rbohe-2 (Figure 9F) as in ms188-1 (Figure 9C). At stages 10 and 11, large amounts of microspores degenerated and anther locules were filled with cell debris, possibly from degenerated microspores in ms188-1 rbohe-2 (Figures 9G and 9H). However, unlike in ms188-1, where tapetal cells fully degenerated at stage 10 (Phan et al., 2011), tapetal cells in ms188-1 rbohe-2 were still visible at stages 10 and 11 (Figures 9G and 9H), suggesting delayed tapetal degeneration caused by RBOHE loss of function.

Figure 9.
Figure 9.

Genetic Interaction between MYB80 and RBOHE.

(A) and (B) Transverse semithin sections of ms188-1 (A) and ms188-1 rbohe-2 (B) anthers at different developmental stages.

(C) to (E) Transmission electron micrographs of ms188-1 anthers at stages 8 and 9 (C) and stages 9 and 10 ([D] and [E]). (E) shows a closeup at the degenerated tapetum.

(F) to (H) Transmission electron micrographs of ms188-1 rbohe-2 anthers at stages 8 and 9 (F), stages 9 and 10 (G), and stages 10 and 11 (H).

dMsp, degenerated microspore; ML, middle layer; Ta, tapetum. Bars in (A) and (B) = 20 µm; bars in (C) and (D) = 1 µm; bars in (E) = 500 nm; bars in (F) to (H) = 2 µm.

[See online article for color version of this figure.]

To find out whether the delayed tapetal degeneration reflected a delayed tapetal PCD process, we performed TUNEL assays on the ms188-1 single mutant and the ms188-1 rbohe-2 double mutant. Indeed, unlike the precocious tapetal PCD in ms188-1 (Figure 10A), TUNEL-positive signals were significantly delayed and reduced in ms188-1 rbohe-2 (Figure 10B). TUNEL signals were hardly detectable even at stage 11, when other anther tissues started to degenerate (Figure 10B). These results suggested that RBOHE was partially epistatic to MYB80. In addition, introducing rbohe-2 into ms188-1 resulted in a novel phenotype (i.e., much delayed release of microspores from tetrads) (Figures 9B, 9F, and 9G), suggesting a complex rather than a linear interaction between RBOHE and MYB80.

Figure 10.
Figure 10.

RBOHE Is Epistatic to MYB80 in the Tapetal PCD.

Fluorescence microscopy of DNA fragmentation detected using the TUNEL assays in sections of ms188-1 (A) and ms188-1 rbohe-2 (B) anthers is shown at different developmental stages. Corresponding bright-field images are placed on top of the fluorescence images. Green fluorescence indicates TUNEL-positive signals, and red fluorescence indicates propidium iodide staining. Bars = 20 µm.

DISCUSSION

Tapetal NADPH Oxidases Control the Temporal ROS Pattern

ROS can be produced from various subcellular locations, such as chloroplasts, mitochondria, peroxisomes, and apoplast. ROS produced by mitochondrial respiration was especially noted for its roles in metazoan PCD (D’Autréaux and Toledano, 2007). Indeed, failure of ROS scavenging in mitochondria causes wild abortive cytoplasmic male sterility in rice (Luo et al., 2013). However, our results suggest that plasma membrane-localized NADPH oxidases are crucial for the temporal ROS pattern in tapetum.

By using qualitative and quantitative approaches, we show that Arabidopsis anthers display a spatiotemporal ROS pattern that reaches its highest level approximately at stages 8 and 9 (Figure 7), during which tapetal degradation commences, to promote microsporogenesis (Sanders et al., 1999; Parish and Li, 2010). Afterward, ROS rapidly decreases to a basal level at stage 11 (Figure 7), when tapetal cells mostly disintegrate (Sanders et al., 1999; Parish and Li, 2010). Similar observations were made in developing anthers in rice (Hu et al., 2011), suggesting the evolutionary conservation of temporal ROS patterns.

Genetic interference with the temporal ROS pattern compromised tapetal degeneration and resulted in pollen abortion (Figures 4 and 5). Specifically, functional loss of RBOHE and RHD2 abolished the ROS peak at stage 8 (Figure 7). However, residual ROS production in the absence of the tapetal NADPH oxidases suggests other tapetal ROS sources, such as mitochondria. Indeed, failure of scavenging mitochondria ROS leads to tapetal dysfunction and pollen abortion in rice (Hu et al., 2011; Luo et al., 2013).

In contrast to the overall reduction of ROS in RBOHE loss of function, spatiotemporal overexpression of RBOHE significantly enhanced ROS production only at early developmental stages (Figure 7) but not after, even though ProOsg6B:RBOHE transgenic anthers did show enhanced RBOHE expression (Supplemental Figure 4). The discrepancy between the expression of RBOHE and ROS levels might be due to posttranslational modifications such as phosphorylation (Marino et al., 2012), S-nitrosylation (Yun et al., 2011), Ca2+ concentration (Foreman et al., 2003; Ogasawara et al., 2008), binding to phosphatidic acid (Zhang et al., 2009), or protein–protein interaction (Wong et al., 2007), all of which affect NADPH oxidase activities.

Another question regarding NADPH oxidase–mediated spatiotemporal ROS patterns is what results in the transient ROS peak. The differential transcript abundance of NADPH oxidase–coding genes at different anther developmental stages might be one possibility (Figure 1; Supplemental Figure 2). A nonmutually exclusive possibility is temporal expression of ROS scavengers. ROXY1 and ROXY2 encode glutaredoxins whose activities at early anther stages may dampen the overall ROS level before stage 8 (Xing and Zachgo, 2008). Alternatively, late expression of ROS scavengers by the tapetal transcription factors, such as MYB80 (Phan et al., 2011), would reduce the overall ROS levels after stage 8, rendering the ROS peak transient.

ROS Generated by NADPH Oxidases May Mediate the Synchronization between Tapetal PCD and Pollen Development

Tapetal cells support pollen development by providing building materials for the pollen wall through PCD at late stages (Wilson and Zhang, 2009; Chang et al., 2011). Although RBOHE was also detected in developing microspores at early stages in addition to tapetum during stages 6 to 11 (Figure 1), its gametophytic function, if any, would be negligible because the pollen defect in RBOHE loss of function is sporophytic (Supplemental Figure 1). RBOHE loss of function resulted in disrupted pollen exine formation (Figure 3) and delayed tapetal PCD (Figure 6), while its gain of function resulted in abnormal deposition of cellular contents in anther locules (Figure 5) and precocious tapetal PCD (Figure 6). Both resulted in significant pollen abortion (Supplemental Figures 1 and 4), providing additional support to the notion that the proper timing of tapetal PCD is critical for pollen development (Higginson et al., 2003; Kawanabe et al., 2006; Li et al., 2011; Phan et al., 2011; Niu et al., 2013).

It has long been accepted that male gametophytes require synchronized development of supporting sporophytic cells. However, the identity of such intercellular signals has remained obscure. ROS produced by NADPH oxidases may serve as signaling molecules. The existence of multilevel control of RBOHs, transcriptionally and posttranslationally, makes them good candidates for fine-tuning ROS production in response to extracellular cues, which are essential prerequisites for ROS signal specificity (Mittler et al., 2004, 2011). Apoplastic ROS generated by NADPH oxidases may enter the neighboring microspores to initiate signaling events, leading to developmental synchronization.

ROS are generally not primary agonists but are superimposed cosignals that may allow the integration of cellular activity pathways in accordance with the metabolic state of the cell (D’Autréaux and Toledano, 2007). Well-characterized targets of ROS are redox-sensitive kinases (Neill et al., 2002). ROS also initiate global changes in gene expression through regulating a subset of transcription factors, including MYB family transcription factors (Gechev et al., 2006), to which most components of the transcriptional network belong (Wilson and Zhang, 2009; Parish and Li, 2010; Ma et al., 2012). Indeed, plant PCD processes often involve redox-sensitive transcriptional changes (Swidzinski et al., 2002).

RBOHE and the Tapetal Transcriptional Network

Extensive transcriptomic studies and examination of tapetal mutants have allowed the systematic identification of transcriptional events regulating tapetal PCD, in which AMS and MYB80 act downstream of the transcriptional hierarchy (Wijeratne et al., 2007; Yang et al., 2007; Xu et al., 2010; Phan et al., 2011; Feng et al., 2012; Ma et al., 2012). RBOHE was among the genes whose expression was significantly downregulated in dyt1 and ams anthers (Feng et al., 2012) but upregulated in ms188-1 anthers (Phan et al., 2011), suggesting its transcriptional control by the tapetal transcriptional network (Zhu et al., 2008, 2011), most likely through AMS and MYB80. Indeed, multiple cis-elements of AMS and MYB80 are present in the promoter of RBOHE (Supplemental Figure 6). By promoter:GUS reporter analysis, RNA in situ hybridization, and qPCR analysis, we confirmed such transcriptional changes of RBOHE by mutations at AMS and MYB80 (Figure 8).

We tried to analyze the genetic epistasis between AMS and RBOHE by introducing ProAMS:RBOHE into ams, but no transgenic lines could be recovered with exogenous RBOHE expression in ams (H.-T. Xie and Y. Zhang, unpublished data). By contrast, introducing rbohe-2 into ms188-1 partially rescued premature tapetal PCD (Figures 9 and 10) and suppressed the high ROS level in ms188-1 (Supplemental Figure 8), suggesting that RBOHE is at least partially epistatic to MYB80. However, microspore degeneration still occurred in ms188-1 rbohe-2 at stages comparable to ms188-1 (Figures 9 and 10), suggesting the presence of other downstream components of MYB80 independent of RBOHE, possibly genes directly regulated by MYB80 such as UNDEAD (Phan et al., 2011).

Both AMS and MYB80 were proposed to function as transcriptional activators (Li et al., 2007; Xu et al., 2010). However, we cannot exclude the possibility that they function as transcriptional repressors of certain genes (Phan et al., 2011). Indeed, MYB80 was proposed to be a bifunctional transcription factor due to the fact that it increases UNDEAD expression but decreases GLOX1 expression, likely through the binding of different cofactors (Phan et al., 2011). The ams ms188-1 mutants showed similar developmental defects to either dyt1 or tdf1 (Zhu et al., 2011), indicating a collaborative action of AMS and MYB80. However, their antagonistic action on the expression of RBOHE, directly or indirectly, suggests a complex rather than additive regulatory mechanism that awaits further investigation.

METHODS

Plant Materials and Growth Conditions

Arabidopsis thaliana plants were grown and transformed as described (Zhou et al., 2013). Growth conditions for Nicotiana benthamiana as well as Agrobacterium tumefaciens–mediated RBOHE expression by infiltration were conducted as described (Wong et al., 2007). The T-DNA insertion lines, SALK_146126C (rbohe-1) and SALK_150096C (rbohe-2), were obtained from the ABRC (http://www.arabidopsis.org). Columbia-0 was used as the wild type. Transgenic Arabidopsis plants were selected on half-strength Murashige and Skoog medium supplemented with either 25 µg/mL hygromycin B or 30 µg/mL glufosinate-ammonium PESTANAL (Sigma-Aldrich). Mutants including rbohe-1, rbohe-2, and ams (SALK_152147) were analyzed by genotyping PCR using the following primers: ZP953/ZP956 for RBOHE; ZP2/ZP956 for rbohe-1/2; ZP1420/ZP1421 for AMS; and ZP1420/ZP2 for ams. The single base pair mutation of rhd2-1 (CAG→CAA) (Foreman et al., 2003) was analyzed using the primers ZP1799/ZP1800 such that PCR products from the wild-type copy but not those from rhd2-1 can be digested by PstI into 40- and 200-bp fragments. The single base pair mutation of ms188-1 (CAA→TAA) was characterized by sequencing analyses as described (Zhang et al., 2007). Primers are listed in Supplemental Table 2.

RNA Extraction, RT-PCR, and qPCR

For RT-PCR of RBOHE loss-of-function mutants, total RNAs were extracted from inflorescences using the RNeasy plant miniprep kit according to the manufacturer’s instructions (Qiagen). For RT-PCR of RBOHE gain-of-function lines, total RNAs were extracted from stage 5 to 8 anthers based on the size of floral buds and the position of a given floral bud in an inflorescence as described (Peirson et al., 1996; Phan et al., 2011). Reverse transcription was performed using SuperScript III reverse transcriptase with on-column DNase-I digestion (Invitrogen). Primers used in the RT-PCR were ZP953/ZP955 for the endogenous RBOHE and ZP1628/ZP955 for the exogenous RBOHE. Arabidopsis ACTIN2 was used as the internal control for RT-PCR as described (Wang et al., 2013).

For qPCR, total RNAs were extracted from stage 5 to 8 anthers from the wild type, ams, or ms188-1. The qPCR analysis was performed with the Bio-Rad CFX96 real-time system using SYBR Green real-time PCR master mix (Toyobo) as described (Zhou et al., 2013). Primers ZP2196/ZP2197 were used for RBOHE. GAPDH and TUBULIN2 were used as internal controls as described (Zhou et al., 2013). The qPCR was performed using three biological replicates for each genotype, and three technical replicates were conducted for each sample. The data presented in Figure 8 are means of three biological replicates for which three technical replicates were averaged. All primers used in this study are listed in Supplemental Table 2.

Plasmid Construction

All constructs were generated using Gateway technology (Invitrogen). The pENTR/D/TOPO vector (Invitrogen) was used to generate all entry vectors. For GUS reporter constructs, a 2171-bp sequence upstream of the start codon of RBOHE was amplified from Arabidopsis genomic DNA using the primer pair ZP951/ZP952 to generate the entry vector for ProRBOHE. A 1634-bp sequence upstream of the start codon of Osg6B (Tsuchiya et al., 1994) was amplified from rice (Oryza sativa subsp japonica) genomic DNA using the primer pair ZP780/ZP781 to generate the entry vector for ProOsg6B. A 1968-bp sequence upstream of the start codon of RHD2 was amplified from Arabidopsis genomic DNA using the primer pair ZP1041/ZP1042 to generate the entry vector for ProRHD2 (Foreman et al., 2003). Destination vectors for the GUS reporter constructs (i.e., ProRBOHE:GUS, ProRHD2:GUS, and ProOsg6B:GUS) were generated by LR reactions with the corresponding entry vectors and pMD163 (Curtis and Grossniklaus, 2003).

For RBOHE expression vectors, the whole genomic fragment of RBOHE and the RBOHE.3 coding sequence were cloned either from genomic DNA or from the anther cDNA library (stages 6 to 10) with the primer pairs ZP953/ZP954 and ZP953/ZP1353, respectively. The gene fragment encoding RBOHEN1-440 was amplified from the anther cDNA library (stages 6 to 10) with the primer pair ZP953/ZP1225. Destination vectors for RBOHE overexpression in tapetum were constructed from pH2WG7 (Karimi et al., 2002) by replacing its Pro35S with ProOsg6B (ZP780/ZP781) and ProAMS (ZP1284/ZP1285) using the SacI/SpeI restriction sites. The corresponding expression vectors, including ProOsg6B:RBOHE and ProAMS:RBOHE, were generated by LR reactions with the RBOHE entry vector and corresponding destination vectors using LR Clonase II (Invitrogen). The expression vectors for tobacco (Nicotiana tabacum) injection, including Pro35S:RBOHE and Pro35S:RBOHEN1-440, were generated by LR reactions with the corresponding entry vectors and pB7WFG2.0 (Karimi et al., 2002). Primers are listed in Supplemental Table 2.

GUS Histochemistry

Histochemical analysis of GUS activity was performed as described (Zhou et al., 2013). To show the outlines of cells, tissue sections were stained with 100 mg/L ruthenium red (Sigma-Aldrich) for 1 to 2 min before visualization. Images were captured with an Olympus BX51 microscope equipped with a charge-coupled device camera. The histochemical analyses of GUS activity in ProRBOHE:GUS transgenic plants of different genetic backgrounds (i.e., in ams, ms188-1, dyt1, or tdf1) were performed by directly infiltrating anthers with the GUS staining solution under vacuum for 15 min followed by 37°C incubation for 8 or 12 h. Treated samples were subject to decoloration in 95% (v/v) ethanol for 24 to 48 h before being visualized. To ensure comparability, the same ProRBOHE:GUS transgenes were introduced in different genetic backgrounds and analyzed when the transgene was homozygous. Reproducible results were obtained from three independent experiments.

RNA in Situ Hybridization

RNA in situ hybridization was performed as described (Zhou et al., 2013). Briefly, inflorescences of wild-type or rbohe-1 plants were fixed in formaldehyde solution (formalin:acetic acid:ethanol:water, 1:2:10:7) overnight at 4°C, embedded in Paraplast (Sigma-Aldrich) after dehydration, and sectioned at 8 µm thickness. The 324-bp fragment specific for RBOHE was amplified from the coding region of RBOHE with the primer pair ZP1017/ZP1018. A 400-bp fragment specific for RHD2 was amplified from the coding region of RHD2 with the primer pair ZP1171/ZP1172. Sense and antisense probes were transcribed in vitro with digoxigenin-UTP by SP6 and T7 RNA polymerases (Roche), respectively. Tissue sections were hybridized with 1.0 ng/μL probes at 42°C overnight in a hybridization solution that contained 50% formamide. Hybridization signals were detected using anti-digoxigenin antibody (Anti-Digoxigenin-Ap Fab fragments; Roche). The samples were observed using an Olympus BX51 microscope.

Pollen Analysis and Anther Staging

Pollen staining with the Alexander dye, scanning electron microscopy of pollen or dehiscing anthers, and pollen in vitro germination were performed as described (Zhou et al., 2013). For pollen abortion rate, pollen grains that have collapsed, lost the normal rugby shape, or have no or disordered apertures are counted as aborted pollen grains. Pollen grains only defective in exine patterns are counted as normal.

For anther staging, because tapetal defects often affect other cell layers, such as delaying microsporogenesis, it is often difficult to determine anther stages based on either microspore developmental stage or tapetal status (Sanders et al., 1999). Therefore, we determined anther stages based on the size and the position of a given floral bud in an inflorescence as described previously (Peirson et al., 1996). In addition, we also classified anthers of different developmental stages based on anther sizes (width and overall area) as described in Supplemental Table 1.

TUNEL Assay

Whole inflorescences from the wild type, rbohe-1, rbohe-2, ms188-1, ms188-1 rbohe-2, and RBOHE overexpression lines were processed as described (Phan et al., 2011). In situ nick-end labeling of nuclear DNA fragmentation was performed with the Dead End Fluorometric TUNEL system according to the supplier’s instructions (Promega). Samples were analyzed with a confocal laser scanning microscope (Zeiss) using a 488-nm/510-nm excitation/emission spectrum for fluorescein and a 530-nm/640-nm excitation/emission spectrum for propidium iodide as described (Phan et al., 2011).

Histology of Anthers and Histochemical Assays for ROS

Floral buds of different development stages were fixed, embedded, and processed for transverse semithin sections and transmission electron microscopy as described (Wang et al., 2013).

Anthers of various developmental stages were either incubated in 10 mM potassium-citrate buffer (pH 6.0) containing 0.5 mM NBT for superoxide anion detection (Hu et al., 2011) or stained using the fluorescent dye H2DCF-DA (Sigma-Aldrich) for ROS as described (Huang et al., 2013). For H2DCF-DA staining, anthers were vacuum infiltrated with 5 µM H2DCF-DA staining solution for 15 min and then incubated at 25°C for 2 h. Fluorescence imaging of H2DCF-DA–stained anthers was performed with an Axio Observer D1 microscope equipped with a CCD camera (Zeiss) at the same parameters without reaching signal saturation in any sample for comparative quantification. Anthers were classified into six groups based on their sizes as described in Supplemental Table 1. Three independent experiments were conducted for all genetic backgrounds, and each experiment involved 20 to 25 anthers of a given group. Fluorescence intensity of whole anthers was quantified using ImageJ software (http://rsbweb.nih.gov/ij/).

For NBT staining of tobacco leaves, infiltrated plants were kept in a growth chamber for 2 to 3 d before being examined. Detached leaves were first infiltrated with 0.5 mM NBT solution by vacuum for 15 min followed by 6 h of incubation in 0.5 mM NBT solution. NBT-treated leaves were decolorized in 95% (v/v) ethanol for 48 h before being visualized.

Accession Numbers

Sequence data were archived in TAIR (www.arabidopsis.org) and the National Center for Biotechnology Information with the following accession numbers: RBOHE (At1g19230), RHD2/RBOHC (At5g51060), AMS (At2g16910), MYB80/MYB103 (At5g56110), DYT1 (At4g21330), TDF1 (At3g28470), and Osg6B (Os11g0582400).

Supplemental Data

The following materials are available in the online version of this article.

Acknowledgments

We thank Zhong-Nan Yang for the ams, dyt1, tdf1, and ms188-1 seeds and Lei Ge for the rhd2-1 seeds. This work was supported by the Major Research Plan from the Ministry of Science and Technology of China (Grant 2013CB945102), by Shandong Provincial Funds for Outstanding Young Scientists (to Y.Z.), and by the Tai-Shan Scholar program from the Shandong Provincial Government (to Y.Z.).

AUTHOR CONTRIBUTIONS

H.-T.X. and Z.-Y.W. performed the experiments. Y.Z., H.-T.X., and S.L. designed the experiments and wrote the article.

Footnotes

  • www.plantcell.org/cgi/doi/10.1105/tpc.114.125427

  • The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantcell.org) is: Yan Zhang (yzhang{at}sdau.edu.cn).

  • ↵[C] Some figures in this article are displayed in color online but in black and white in the print edition.

  • ↵[W] Online version contains Web-only data.

Glossary

PCD
programmed cell death
ROS
reactive oxygen species
NBT
nitrotetrazolium blue chloride
TUNEL
terminal deoxynucleotidyl transferase–mediated dUTP nick-end labeling
H2DCF-DA
2′,7′-dichlorodihydrofluorescein diacetate
qPCR
quantitative real-time PCR
  • Received March 17, 2014.
  • Revised April 10, 2014.
  • Accepted April 16, 2014.
  • Published May 7, 2014.

References

  1. ↵
    1. Beers E.P.,
    2. McDowell J.M.
    (2001). Regulation and execution of programmed cell death in response to pathogens, stress and developmental cues. Curr. Opin. Plant Biol. 4: 561–567.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Chang F.,
    2. Wang Y.,
    3. Wang S.,
    4. Ma H.
    (2011). Molecular control of microsporogenesis in Arabidopsis. Curr. Opin. Plant Biol. 14: 66–73.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Curtis M.D.,
    2. Grossniklaus U.
    (2003). A gateway cloning vector set for high-throughput functional analysis of genes in planta. Plant Physiol. 133: 462–469.
    OpenUrlAbstract/FREE Full Text
  4. ↵
    1. D’Autréaux B.,
    2. Toledano M.B.
    (2007). ROS as signalling molecules: Mechanisms that generate specificity in ROS homeostasis. Nat. Rev. Mol. Cell Biol. 8: 813–824.
    OpenUrlCrossRefPubMed
  5. ↵
    1. De Pinto M.C.,
    2. Locato V.,
    3. De Gara L.
    (2012). Redox regulation in plant programmed cell death. Plant Cell Environ. 35: 234–244.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Feng B.,
    2. Lu D.,
    3. Ma X.,
    4. Peng Y.,
    5. Sun Y.,
    6. Ning G.,
    7. Ma H.
    (2012). Regulation of the Arabidopsis anther transcriptome by DYT1 for pollen development. Plant J. 72: 612–624.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Foreman J.,
    2. Demidchik V.,
    3. Bothwell J.H.,
    4. Mylona P.,
    5. Miedema H.,
    6. Torres M.A.,
    7. Linstead P.,
    8. Costa S.,
    9. Brownlee C.,
    10. Jones J.D.,
    11. Davies J.M.,
    12. Dolan L.
    (2003). Reactive oxygen species produced by NADPH oxidase regulate plant cell growth. Nature 422: 442–446.
    OpenUrlCrossRefPubMed
  8. ↵
    1. Gechev T.S.,
    2. Van Breusegem F.,
    3. Stone J.M.,
    4. Denev I.,
    5. Laloi C.
    (2006). Reactive oxygen species as signals that modulate plant stress responses and programmed cell death. Bioessays 28: 1091–1101.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Higginson T.,
    2. Li S.F.,
    3. Parish R.W.
    (2003). AtMYB103 regulates tapetum and trichome development in Arabidopsis thaliana. Plant J. 35: 177–192.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Hsieh K.,
    2. Huang A.H.
    (2007). Tapetosomes in Brassica tapetum accumulate endoplasmic reticulum–derived flavonoids and alkanes for delivery to the pollen surface. Plant Cell 19: 582–596.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Hu L.,
    2. Liang W.,
    3. Yin C.,
    4. Cui X.,
    5. Zong J.,
    6. Wang X.,
    7. Hu J.,
    8. Zhang D.
    (2011). Rice MADS3 regulates ROS homeostasis during late anther development. Plant Cell 23: 515–533.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    1. Huang G.-Q.,
    2. Li E.,
    3. Ge F.-R.,
    4. Li S.,
    5. Wang Q.,
    6. Zhang C.-Q.,
    7. Zhang Y.
    (2013). Arabidopsis RopGEF4 and RopGEF10 are important for FERONIA-mediated developmental but not environmental regulation of root hair growth. New Phytol. 200: 1089–1101.
    OpenUrlCrossRefPubMed
  13. ↵
    1. Ito T.,
    2. Nagata N.,
    3. Yoshiba Y.,
    4. Ohme-Takagi M.,
    5. Ma H.,
    6. Shinozaki K.
    (2007). Arabidopsis MALE STERILITY1 encodes a PHD-type transcription factor and regulates pollen and tapetum development. Plant Cell 19: 3549–3562.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    1. Karimi M.,
    2. Inzé D.,
    3. Depicker A.
    (2002). GATEWAY vectors for Agrobacterium-mediated plant transformation. Trends Plant Sci. 7: 193–195.
    OpenUrlCrossRefPubMed
  15. ↵
    1. Kawanabe T.,
    2. Ariizumi T.,
    3. Kawai-Yamada M.,
    4. Uchimiya H.,
    5. Toriyama K.
    (2006). Abolition of the tapetum suicide program ruins microsporogenesis. Plant Cell Physiol. 47: 784–787.
    OpenUrlAbstract/FREE Full Text
  16. ↵
    1. Keller T.,
    2. Damude H.G.,
    3. Werner D.,
    4. Doerner P.,
    5. Dixon R.A.,
    6. Lamb C.
    (1998). A plant homolog of the neutrophil NADPH oxidase gp91phox subunit gene encodes a plasma membrane protein with Ca2+ binding motifs. Plant Cell 10: 255–266.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Kwak J.M.,
    2. Mori I.C.,
    3. Pei Z.M.,
    4. Leonhardt N.,
    5. Torres M.A.,
    6. Dangl J.L.,
    7. Bloom R.E.,
    8. Bodde S.,
    9. Jones J.D.,
    10. Schroeder J.I.
    (2003). NADPH oxidase AtrbohD and AtrbohF genes function in ROS-dependent ABA signaling in Arabidopsis. EMBO J. 22: 2623–2633.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Lam E.
    (2004). Controlled cell death, plant survival and development. Nat. Rev. Mol. Cell Biol. 5: 305–315.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Lee Y.,
    2. Rubio M.C.,
    3. Alassimone J.,
    4. Geldner N.
    (2013). A mechanism for localized lignin deposition in the endodermis. Cell 153: 402–412.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Li H.,
    2. Yuan Z.,
    3. Vizcay-Barrena G.,
    4. Yang C.,
    5. Liang W.,
    6. Zong J.,
    7. Wilson Z.A.,
    8. Zhang D.
    (2011). PERSISTENT TAPETAL CELL1 encodes a PHD-finger protein that is required for tapetal cell death and pollen development in rice. Plant Physiol. 156: 615–630.
    OpenUrlAbstract/FREE Full Text
  21. ↵
    1. Li N.,
    2. et al
    . (2006). The rice Tapetum Degeneration Retardation gene is required for tapetum degradation and anther development. Plant Cell 18: 2999–3014.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Li S.F.,
    2. Iacuone S.,
    3. Parish R.W.
    (2007). Suppression and restoration of male fertility using a transcription factor. Plant Biotechnol. J. 5: 297–312.
    OpenUrlCrossRefPubMed
  23. ↵
    1. Luo D.,
    2. et al
    . (2013). A detrimental mitochondrial-nuclear interaction causes cytoplasmic male sterility in rice. Nat. Genet. 45: 573–577.
    OpenUrlCrossRefPubMed
  24. ↵
    1. Ma X.,
    2. Feng B.,
    3. Ma H.
    (2012). AMS-dependent and independent regulation of anther transcriptome and comparison with those affected by other Arabidopsis anther genes. BMC Plant Biol. 12: 23.
    OpenUrlCrossRefPubMed
  25. ↵
    1. Marino D.,
    2. Dunand C.,
    3. Puppo A.,
    4. Pauly N.
    (2012). A burst of plant NADPH oxidases. Trends Plant Sci. 17: 9–15.
    OpenUrlCrossRefPubMed
  26. ↵
    1. Millar A.A.,
    2. Gubler F.
    (2005). The Arabidopsis GAMYB-like genes, MYB33 and MYB65, are microRNA-regulated genes that redundantly facilitate anther development. Plant Cell 17: 705–721.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    1. Mittler R.,
    2. Vanderauwera S.,
    3. Gollery M.,
    4. Van Breusegem F.
    (2004). Reactive oxygen gene network of plants. Trends Plant Sci. 9: 490–498.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Mittler R.,
    2. Vanderauwera S.,
    3. Suzuki N.,
    4. Miller G.,
    5. Tognetti V.B.,
    6. Vandepoele K.,
    7. Gollery M.,
    8. Shulaev V.,
    9. Van Breusegem F.
    (2011). ROS signaling: The new wave? Trends Plant Sci. 16: 300–309.
    OpenUrlCrossRefPubMed
  29. ↵
    1. Neill S.,
    2. Desikan R.,
    3. Hancock J.
    (2002). Hydrogen peroxide signalling. Curr. Opin. Plant Biol. 5: 388–395.
    OpenUrlCrossRefPubMed
  30. ↵
    1. Niu N.,
    2. Liang W.,
    3. Yang X.,
    4. Jin W.,
    5. Wilson Z.A.,
    6. Hu J.,
    7. Zhang D.
    (2013). EAT1 promotes tapetal cell death by regulating aspartic proteases during male reproductive development in rice. Nat. Commun. 4: 1445.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Ogasawara Y.,
    2. et al
    . (2008). Synergistic activation of the Arabidopsis NADPH oxidase AtrbohD by Ca2+ and phosphorylation. J. Biol. Chem. 283: 8885–8892.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Parish R.W.,
    2. Li S.F.
    (2010). Death of a tapetum: A programme of developmental altruism. Plant Sci. 178: 73–89.
    OpenUrlCrossRef
  33. ↵
    1. Peirson B.,
    2. Owen H.,
    3. Feldmann K.,
    4. Makaroff C.
    (1996). Characterization of three male-sterile mutants of Arabidopsis thaliana exhibiting alterations in meiosis. Sex. Plant Reprod. 9: 1–16.
    OpenUrl
  34. ↵
    1. Phan H.A.,
    2. Iacuone S.,
    3. Li S.F.,
    4. Parish R.W.
    (2011). The MYB80 transcription factor is required for pollen development and the regulation of tapetal programmed cell death in Arabidopsis thaliana. Plant Cell 23: 2209–2224.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Phan H.A.,
    2. Li S.F.,
    3. Parish R.W.
    (2012). MYB80, a regulator of tapetal and pollen development, is functionally conserved in crops. Plant Mol. Biol. 78: 171–183.
    OpenUrlCrossRefPubMed
  36. ↵
    1. Sanders P.M.,
    2. Bui A.Q.,
    3. Weterings K.,
    4. McIntire K.N.,
    5. Hsu Y.-C.,
    6. Lee P.Y.,
    7. Truong M.T.,
    8. Beals T.P.,
    9. Goldberg R.B.
    (1999). Anther developmental defects in Arabidopsis thaliana male-sterile mutants. Sex. Plant Reprod. 11: 297–322.
    OpenUrlCrossRef
  37. ↵
    1. Suzuki N.,
    2. Miller G.,
    3. Morales J.,
    4. Shulaev V.,
    5. Torres M.A.,
    6. Mittler R.
    (2011). Respiratory burst oxidases: The engines of ROS signaling. Curr. Opin. Plant Biol. 14: 691–699.
    OpenUrlCrossRefPubMed
  38. ↵
    1. Swidzinski J.A.,
    2. Sweetlove L.J.,
    3. Leaver C.J.
    (2002). A custom microarray analysis of gene expression during programmed cell death in Arabidopsis thaliana. Plant J. 30: 431–446.
    OpenUrlCrossRefPubMed
  39. ↵
    1. Torres M.A.,
    2. Dangl J.L.
    (2005). Functions of the respiratory burst oxidase in biotic interactions, abiotic stress and development. Curr. Opin. Plant Biol. 8: 397–403.
    OpenUrlCrossRefPubMed
  40. ↵
    1. Torres M.A.,
    2. Dangl J.L.,
    3. Jones J.D.
    (2002). Arabidopsis gp91phox homologues AtrbohD and AtrbohF are required for accumulation of reactive oxygen intermediates in the plant defense response. Proc. Natl. Acad. Sci. USA 99: 517–522.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    1. Tsuchiya T.,
    2. Toriyama K.,
    3. Ejiri S.,
    4. Hinata K.
    (1994). Molecular characterization of rice genes specifically expressed in the anther tapetum. Plant Mol. Biol. 26: 1737–1746.
    OpenUrlCrossRefPubMed
  42. ↵
    1. Vizcay-Barrena G.,
    2. Wilson Z.A.
    (2006). Altered tapetal PCD and pollen wall development in the Arabidopsis ms1 mutant. J. Exp. Bot. 57: 2709–2717.
    OpenUrlAbstract/FREE Full Text
  43. ↵
    1. Wang J.G.,
    2. Li S.,
    3. Zhao X.Y.,
    4. Zhou L.Z.,
    5. Huang G.Q.,
    6. Feng C.,
    7. Zhang Y.
    (2013). HAPLESS13, the Arabidopsis μ1 adaptin, is essential for protein sorting at the trans-Golgi network/early endosome. Plant Physiol. 162: 1897–1910.
    OpenUrlAbstract/FREE Full Text
  44. ↵
    1. Wijeratne A.J.,
    2. Zhang W.,
    3. Sun Y.,
    4. Liu W.,
    5. Albert R.,
    6. Zheng Z.,
    7. Oppenheimer D.G.,
    8. Zhao D.,
    9. Ma H.
    (2007). Differential gene expression in Arabidopsis wild-type and mutant anthers: Insights into anther cell differentiation and regulatory networks. Plant J. 52: 14–29.
    OpenUrlCrossRefPubMed
  45. ↵
    1. Wilson Z.A.,
    2. Zhang D.B.
    (2009). From Arabidopsis to rice: Pathways in pollen development. J. Exp. Bot. 60: 1479–1492.
    OpenUrlAbstract/FREE Full Text
  46. ↵
    1. Wilson Z.A.,
    2. Morroll S.M.,
    3. Dawson J.,
    4. Swarup R.,
    5. Tighe P.J.
    (2001). The Arabidopsis MALE STERILITY1 (MS1) gene is a transcriptional regulator of male gametogenesis, with homology to the PHD-finger family of transcription factors. Plant J. 28: 27–39.
    OpenUrlCrossRefPubMed
  47. ↵
    1. Wong H.L.,
    2. Pinontoan R.,
    3. Hayashi K.,
    4. Tabata R.,
    5. Yaeno T.,
    6. Hasegawa K.,
    7. Kojima C.,
    8. Yoshioka H.,
    9. Iba K.,
    10. Kawasaki T.,
    11. Shimamoto K.
    (2007). Regulation of rice NADPH oxidase by binding of Rac GTPase to its N-terminal extension. Plant Cell 19: 4022–4034.
    OpenUrlAbstract/FREE Full Text
  48. ↵
    1. Xing S.,
    2. Zachgo S.
    (2008). ROXY1 and ROXY2, two Arabidopsis glutaredoxin genes, are required for anther development. Plant J. 53: 790–801.
    OpenUrlCrossRefPubMed
  49. ↵
    1. Xu J.,
    2. Yang C.,
    3. Yuan Z.,
    4. Zhang D.,
    5. Gondwe M.Y.,
    6. Ding Z.,
    7. Liang W.,
    8. Zhang D.,
    9. Wilson Z.A.
    (2010). The ABORTED MICROSPORES regulatory network is required for postmeiotic male reproductive development in Arabidopsis thaliana. Plant Cell 22: 91–107.
    OpenUrlAbstract/FREE Full Text
  50. ↵
    1. Yang C.,
    2. Vizcay-Barrena G.,
    3. Conner K.,
    4. Wilson Z.A.
    (2007). MALE STERILITY1 is required for tapetal development and pollen wall biosynthesis. Plant Cell 19: 3530–3548.
    OpenUrlAbstract/FREE Full Text
  51. ↵
    1. Yun B.W.,
    2. Feechan A.,
    3. Yin M.,
    4. Saidi N.B.,
    5. Le Bihan T.,
    6. Yu M.,
    7. Moore J.W.,
    8. Kang J.G.,
    9. Kwon E.,
    10. Spoel S.H.,
    11. Pallas J.A.,
    12. Loake G.J.
    (2011). S-Nitrosylation of NADPH oxidase regulates cell death in plant immunity. Nature 478: 264–268.
    OpenUrlCrossRefPubMed
  52. ↵
    1. Zhang D.,
    2. Luo X.,
    3. Zhu L.
    (2011). Cytological analysis and genetic control of rice anther development. J. Genet. Genomics 38: 379–390.
    OpenUrlCrossRefPubMed
  53. ↵
    1. Zhang D.S.,
    2. Liang W.Q.,
    3. Yuan Z.,
    4. Li N.,
    5. Shi J.,
    6. Wang J.,
    7. Liu Y.M.,
    8. Yu W.J.,
    9. Zhang D.B.
    (2008). Tapetum degeneration retardation is critical for aliphatic metabolism and gene regulation during rice pollen development. Mol. Plant 1: 599–610.
    OpenUrlCrossRefPubMed
  54. ↵
    1. Zhang W.,
    2. Sun Y.,
    3. Timofejeva L.,
    4. Chen C.,
    5. Grossniklaus U.,
    6. Ma H.
    (2006). Regulation of Arabidopsis tapetum development and function by DYSFUNCTIONAL TAPETUM1 (DYT1) encoding a putative bHLH transcription factor. Development 133: 3085–3095.
    OpenUrlAbstract/FREE Full Text
  55. ↵
    1. Zhang Y.,
    2. Zhu H.,
    3. Zhang Q.,
    4. Li M.,
    5. Yan M.,
    6. Wang R.,
    7. Wang L.,
    8. Welti R.,
    9. Zhang W.,
    10. Wang X.
    (2009). Phospholipase Dα1 and phosphatidic acid regulate NADPH oxidase activity and production of reactive oxygen species in ABA-mediated stomatal closure in Arabidopsis. Plant Cell 21: 2357–2377.
    OpenUrlAbstract/FREE Full Text
  56. ↵
    1. Zhang Z.B.,
    2. et al
    . (2007). Transcription factor AtMYB103 is required for anther development by regulating tapetum development, callose dissolution and exine formation in Arabidopsis. Plant J. 52: 528–538.
    OpenUrlCrossRefPubMed
  57. ↵
    1. Zhou L.Z.,
    2. Li S.,
    3. Feng Q.N.,
    4. Zhang Y.L.,
    5. Zhao X.,
    6. Zeng Y.L.,
    7. Wang H.,
    8. Jiang L.,
    9. Zhang Y.
    (2013). PROTEIN S-ACYL TRANSFERASE10 is critical for development and salt tolerance in Arabidopsis. Plant Cell 25: 1093–1107.
    OpenUrlAbstract/FREE Full Text
  58. ↵
    1. Zhu J.,
    2. Chen H.,
    3. Li H.,
    4. Gao J.F.,
    5. Jiang H.,
    6. Wang C.,
    7. Guan Y.F.,
    8. 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: 266–277.
    OpenUrlCrossRefPubMed
  59. ↵
    1. Zhu J.,
    2. Lou Y.,
    3. Xu X.,
    4. Yang Z.N.
    (2011). A genetic pathway for tapetum development and function in Arabidopsis. J. Integr. Plant Biol. 53: 892–900.
    OpenUrlCrossRefPubMed
Back to top

Table of Contents

Print
Download PDF
Email Article
Citation Tools
Request Permissions
Spatiotemporal Production of Reactive Oxygen Species by NADPH Oxidase Is Critical for Tapetal Programmed Cell Death and Pollen Development in Arabidopsis
Hong-Tao Xie, Zhi-Yuan Wan, Sha Li, Yan Zhang
The Plant Cell May 2014, 26 (5) 2007-2023; DOI: 10.1105/tpc.114.125427
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

In this issue

View this article with LENS