- © 2015 American Society of Plant Biologists. All Rights Reserved.
Abstract
In plants, the copy number of the mitochondrial DNA (mtDNA) can be much lower than the number of mitochondria. The biological significance and regulatory mechanisms of this phenomenon remain poorly understood. Here, using the pollen vegetative cell, we examined the role of the Arabidopsis (Arabidopsis thaliana) mtDNA-binding protein WHIRLY2 (AtWHY2). AtWHY2 decreases during pollen development, in parallel with the rapid degradation of mtDNA; to examine the importance of this decrease, we used the pollen vegetative cell-specific promoter Lat52 to express AtWHY2. The transgenic plants (LWHY2) had very high mtDNA levels in pollen, more than 10 times more than in the wild type (ecotype Columbia-0). LWHY2 plants were fertile, morphologically normal, and set seeds; however, reciprocal crosses with heterozygous plants showed reduced transmission of LWHY2-1 through the male and slower growth of LWHY2-1 pollen tubes. We found that LWHY2-1 pollen had significantly more reactive oxygen species and less ATP compared with the wild type, indicating an effect on mitochondrial respiration. These findings reveal that AtWHY2 affects mtDNA copy number in pollen and suggest that low mtDNA copy numbers might be the normal means by which plant cells maintain mitochondrial genetic information.
Reflecting their endosymbiotic origin, mitochondria contain DNA genomes (mtDNA) encoding several key proteins for oxidative phosphorylation. As most genes identified in the mitochondrial genome are indispensable for mitochondrial function, it is generally believed that each mitochondrion must possess at least one full copy of the genome. Indeed, this seems to be the case in animals. For example, although the number of mitochondria per cell varies in human, mouse, rabbit, and rat cell lines, the mtDNA copy number per mitochondrion remains constant at 2.6 ± 0.3 (Robin and Wong, 1988). Also, in mouse egg cells, each mitochondrion contains an estimated one to two copies of the mtDNA (Pikó and Matsumoto, 1976).
Plant cells, however, have very few copies of the mtDNA compared with the number of mitochondria. For example, in the Cucurbitaceae, cells containing 110 to 140 copies of the mtDNA have 360 to 1,100 mitochondria (Bendich and Gauriloff, 1984). In Arabidopsis (Arabidopsis thaliana), leaf cells each contain approximately 670 mitochondria (Sheahan et al., 2005) and approximately 50 copies of the mtDNA (Draper and Hays, 2000). Thus, in plant cells, each mitochondrion does not possess one complete copy of the mtDNA, a phenomenon that occurs commonly in somatic cells of plants (Preuten et al., 2010). In addition, work in Arabidopsis, barley (Hordeum vulgare), and tobacco (Nicotiana tabacum) showed that cells in leaves, stems, and roots contain few copies of the mtDNA (40–160), whereas cells in root tips contain more copies (300–450; Preuten et al., 2010). This is consistent with the mitochondrial nucleoid diminishment previously observed in developing root and shoot tips (Fujie et al., 1993, 1994), which suggests that the low copy numbers in plant cells result from a decrease in the mtDNA copy number in nondividing cells during development.
One question raised by these findings is whether some mitochondria have complete mtDNAs while others have no mtDNA or whether mitochondria have partial mtDNAs. Using techniques for the direct visualization of small amounts of DNA, our group revealed that up to two-thirds of mitochondria in Arabidopsis mesophyll cells totally lack mtDNA and the remaining one-third of mitochondria possess mtDNA of about 100 kb on average (Wang et al., 2010). This agrees well with a previously reported value for mtDNA copy number (about 50 copies per cell; Draper and Hays, 2000) and is consistent with the idea that plant mitochondrial genomes exist as submolecules smaller than the total genomic sizes (Satoh et al., 1993; Kubo and Newton, 2008). Among plant cells possessing low mtDNA copy numbers, the vegetative cell in the pollen grains is an extreme case; a mature pollen grain of Antirrhinum majus, containing many more mitochondria than a somatic cell, possesses only 16 copies of the mtDNA (Wang et al., 2010). Similar to the changes observed in somatic cells, this extremely low level of mtDNA in pollen vegetative cells results from a rapid decrease in mtDNA copy number during pollen development (Sodmergen et al., 1991; Nagata et al., 1999). In A. majus, the vegetative cell in its initial developmental stage has 482.7 copies of the mtDNA per cell, indicating a 30-fold decrease (482.7/16) during development (Wang et al., 2010). These results from both somatic and reproductive cells led to the intriguing idea that the mtDNA copy number in plants decreases in parallel with cell differentiation, to a very low value, and thus that several mitochondria must share the genetic information carried on a single copy of the mtDNA. Plant cell mitochondria undergo frequent and coupled fusions and divisions, which may explain how mitochondria share this information (Arimura et al., 2004). However, the biological significance of why plant cells lose their mtDNA, and how this benefits these cells, remains unknown. Given that pollen germination, pollen tube elongation, and sperm cell delivery all require energy conversion, the extremely low mtDNA copy numbers, such as in pollen vegetative cells, must not compromise mitochondrial function.
The mtDNA copy numbers remain constant in various tissues, however, indicating that cellular mechanisms accurately regulate the levels of mtDNA in relation to cell type (Robin and Wong, 1988; Preuten et al., 2010). In yeast and animals, this regulation involves the core enzymes of mtDNA replication, such as DNA polymerase-γ (Sharief et al., 1999), RNA polymerase (Wanrooij et al., 2008), and mitochondrial helicase (Liu et al., 2009), as well as a group of DNA-binding proteins such as ARS-binding factor2 protein in yeast (Saccharomyces cerevisiae; Newman et al., 1996), MITOCHONDRIAL TRANSCRIPTION FACTOR A (TFAM) in human (Alam et al., 2003), and mitochondrial single-stranded DNA binding protein in Drosophila spp. (Maier et al., 2001). Overexpression of TFAM causes an increase in the mtDNA copy number, and RNA interference of TFAM decreases the mtDNA copy number (Ekstrand et al., 2004; Kanki et al., 2004). Also, the homozygous knockout of TFAM in mouse results in embryos that lack mtDNA and thus fail to survive (Larsson et al., 1998). Clearly, protein factors within mitochondrial nucleoids play a crucial role in the regulation of mtDNA copy number.
Recent investigation in Arabidopsis revealed that, similar to the case in yeast and animal cells, DNA polymerase, the core enzyme of mtDNA replication, functions to maintain mtDNA levels in plants. Mutation of Arabidopsis PolIA or PolIB (homologs of bacterial DNA polymerase I) causes a reduction in mtDNA copy number, and double mutation of these proteins is lethal (Parent et al., 2011). Also, an Mg2+-dependent exonuclease, DEFECTIVE IN POLLEN ORGANELLE DNA DEGRADATION1 (DPD1), degrades organelle DNA, helping to produce the proper amounts of mtDNA in pollen cells (Matsushima et al., 2011; Tang et al., 2012). These results provide insights into the molecular control of mtDNA levels in plants, via both mtDNA replication and mtDNA degradation. Except for these enzymes, however, other protein factors (such as TFAM in animals) have not been identified in plants. The DNA-binding proteins, such as MutS Homolog1 (MSH1), Organellar Single-Strand DNA Binding Protein1 (OSB1), Recombinase A1 (RecA1), RecA3, and WHIRLY2 (WHY2), identified so far in plant mitochondria likely participate in genomic maintenance by affecting substoichiometric shifting (Abdelnoor et al., 2003), stoichiometric transmission (Zaegel et al., 2006), genomic stability (Shedge et al., 2007; Odahara et al., 2009), and DNA repair (Cappadocia et al., 2010). None of these plant nucleoid factors (DNA-binding proteins) has been implicated in the control of mtDNA copy number; thus, the mechanisms by which nonenzyme protein factors regulate mtDNA copy number in plants remain obscure.
To test whether nucleoid DNA-binding proteins can affect mtDNA copy number, we examined the effect of producing Arabidopsis WHY2, a single-stranded DNA-binding protein (Cappadocia et al., 2010), in the pollen vegetative cell, which generally does not express WHY2 (Honys and Twell, 2004). We found that expression of WHY2 resulted in a 10-fold increase in mtDNA copy number in the pollen vegetative cell. This increase affected mitochondrial respiration, mitochondrial size, and pollen tube growth. Thus, our results uncover a novel function for WHY2, a member of the plant Whirly protein family, in regulating mtDNA amounts and indicate that, in plants, low mtDNA copy number does not compromise mitochondrial function but rather promotes proper mitochondrial function.
RESULTS
Arabidopsis Pollen Vegetative Cells Have Few Copies of the mtDNA
In flowering plants, the pollen vegetative cell generally has an extremely low mtDNA copy number, and the amount of mtDNA can be directly observed by fluorescence microscopy of cells stained for DNA. In Arabidopsis, we found that the fluorescent signals for the mtDNA were undetectable in the vegetative cell of the mature pollen grain (Fig. 1, A and B), showing that this cell has very little mtDNA. We did observe fluorescent signals for mtDNAs in the BCP (Fig. 1, A and B) and other, earlier developmental stages (Supplemental Fig. S1), indicating a decrease of mtDNA levels during pollen development, as indicated previously (Nagata et al., 1999; Matsushima et al., 2011; Tang et al., 2012). Using a competitive PCR technique developed for determining the copy number of mtDNA in single cells (Wang et al., 2010), we measured 133.3 ± 11.5 copies of the mtDNA per BCP and 10.3 ± 2.1 per mature pollen grain (Fig. 1C; Supplemental Fig. S2), indicating a more than 10-fold decrease during pollen development. We also counted 486 ± 107 mitochondria per BCP and 1,708 ± 100 mitochondria per mature pollen grain (Fig. 1D). Thus, each mitochondrion in the vegetative cell at the BCP stage contains an average of 0.274 (133.3/486) copies of the mtDNA, but each mitochondrion at the mature pollen stage contains only 0.006 (10.3/1,708) copies, equivalent to about 170 mitochondria per mtDNA. Since the vegetative cell does not divide during pollen development, the decrease of mtDNA likely results from mtDNA degradation.
The pollen vegetative cells of Arabidopsis have a very low mtDNA copy number. A, Flowers and pollen of ecotype Columbia-0 (Col-0) Arabidopsis at bicellular pollen (BCP) and mature pollen grain (MPG) stages. The pollen grains were stained with 4′,6-diamino-phenylindole (DAPI). Arrows indicate dotted signals of cytoplasmic DNA. GN, Generative nucleus; SN, sperm nucleus; VN, vegetative nuclei. Bars = 10 μm. B, Pollen vegetative cells with transgenic expression of mitochondrial GFP (Lat52:Dips-GFP). The pollen grains were stained with DAPI. Arrows indicate mitochondria having mtDNA signals. Double arrowheads indicate the signals of plastid DNA. Bars = 1 μm. C, Copy number of mtDNA per pollen. The values were acquired from three individual quantifications (Supplemental Fig. S2). Error bars indicate sd. D, Number of mitochondria per pollen. The values were acquired from 10 individual pollen grains of each stage. Error bars indicate sd.
AtWHY2 Levels Decrease with the mtDNA Copy Number
A previous microarray analysis of pollen showed that AtWHY2 transcription decreases during pollen development and is absent in mature pollen grains (Honys and Twell, 2004). Our reverse transcription (RT)-PCR assays verified the array data (Supplemental Fig. S3A). To examine the levels of AtWHY2 in pollen, we used immunoblots with anti-AtWHY2 antibodies against total proteins from pollen at different developmental stages. Our results showed a decrease in AtWHY2 during pollen development and an undetectable level of AtWHY2 in mature pollen (Supplemental Fig. S3B). This is consistent with the transcriptional profiling and with the expression from the AtWHY2 promoter analyzed by fusion with the GUS reporter gene (Supplemental Fig. S3C), indicating that transcriptional regulation causes the decrease in AtWHY2 in wild-type vegetative cells from pollen mitosis I. Thus, the decrease of AtWHY2 occurs at the same time as the degradation of mtDNA during pollen development, consistent with our hypothesis that WHY2 may affect mtDNA levels.
To further examine the role of AtWHY2 in regulating mtDNA levels, we next tested whether AtWHY2 in cells could cause an increase in mtDNA copy numbers. We tested this by expressing AtWHY2 in vegetative cells under the control of the Lat52 promoter, which drives consistent, high-level expression in pollen vegetative cells (Twell et al., 1989); we termed the resulting transgene LWHY2 (Fig. 2A). We also constructed a transgene expressing an AtWHY2-GFP fusion under the control of the Lat52 promoter (LWHY2-GFP) and a Lat52-GFP control transgene (LGFP). We succeeded with this strategy (Fig. 2B) and obtained homozygous LWHY2 transgenic lines showing stable transcription of AtWHY2 (Fig. 2C), accumulation of AtWHY2 (Fig. 2D), and, most interestingly, mtDNA signals in the vegetative cells of mature pollen (Fig. 2E). The signals appeared in independent transgenic lines for LWHY2 and LWHY2-GFP (Fig. 2E; Supplemental Fig.S4A) but not in the LGFP lines (Fig. 2E), implying that transgenic expression of AtWHY2 causes a higher level of mtDNA in the pollen vegetative cells. Quantification of mtDNA levels showed 124 ± 11.1, 135.7 ± 16.6, and 128.7 ± 18.4 copies of the mtDNA per mature pollen grain in LWHY2-1, LWHY2-2, and LWHY2-3 plants, respectively, in contrast to 9 ± 2.6 copies per pollen in LGFP control plants (Fig. 2F; Supplemental Fig. S4, B and C). Thus, the pollen vegetative cells expressing AtWHY2 have more than 10-fold more copies of the mtDNA, indicating a role for AtWHY2 in the elevation of mtDNA levels. The transgenic plants of all lines showed no abnormal phenotypes in plant growth, flowering, or seed set (Supplemental Fig. S5; see below).
The mtDNA copy number is highly elevated in mature pollen grains that express AtWHY2. A, Construction of the transgenic vectors. B, WHY2-GFP signals in mature pollen grains of LWHY-GFP plants and absence of signal in WWHY2-GFP, indicating the successful driving of AtWHY2 expression in mature pollen vegetative cells by Lat52. Bar = 10 μm. C, RT-PCR showing the increased transcript levels of AtWHY2 in pollen of the transgenic plant. MSP, Microspore; TCP, tricellular pollen. D, Immunoblot showing the accumulation of AtWHY2 in mature pollen grains of a transgenic plant in comparison with the undetectable level in pollen of Col-0. HSP is the loading control. E, Appearance of dotted DAPI (mtDNA) signals (arrows) in mature pollen grains of LWHY2-1 plants and absence of the signal in LGFP control plants. SN, Sperm cell nucleus; VN, vegetative nuclei. Bar = 10 μm. F, Copy number of mtDNA per mature pollen grain determined with three independent transgenic lines of LWHY2 (LWHY2-1, LWHY2-2, and LWHY2-3). Error bars indicate sd.
AtWHY2 Localizes to the Mitochondrial Nucleoid
To understand how AtWHY2 affects the level of mtDNA in cells, we next examined its subcellular localization. Previous work using transient expression systems showed that AtWHY2 localizes to the mitochondria in potato (Solanum tuberosum) and onion (Allium cepa) cells (Krause et al., 2005). We examined the mitochondrial localization of AtWHY2 in pollen vegetative cells of Arabidopsis using a homozygous transgenic plant expressing AtWHY2-GFP and mitochondria-localized RFP (for red fluorescent protein; Fig. 3, A and B), from a cross between the transgenic lines LWHY2-GFP-1 and LDips-RFP (a construct containing Lat52:Dips-RFP, where Lat52 is the promoter and Dips is the mitochondrial localization signal; Matsushima et al., 2008). Fluorescence microscopy showed that the AtWHY2-GFP signals appeared dimorphic, with strong and weak signals (Fig. 3A). With normal and controlled (reduced) exposures for GFP, we show that both the strong and weak signals colocalize with the RFP signal (Fig. 3, A and B). The mtDNA signals, however, colocalized with the strong AtWHY2-GFP signals (Fig. 3C). These results indicate that AtWHY2 incorporates unequally into mitochondria, with the mitochondria-possessing mtDNA having significantly more AtWHY2. In the mitochondria of LWHY2-1 plants, electron microscopy revealed electron-dense nucleoid-like structures (Fig. 3D; Supplemental Fig. S6A), and immunogold electron microscopy showed that gold signals of mtDNA and AtWHY2 colocalize within the nucleoid-like structure (Fig. 3E). These observations indicate that AtWHY2 forms a DNA-protein complex in the nucleoid of mitochondria. The mitochondria in the LWHY2-1 pollen vegetative cells were also larger than those in the cells of Col-0 (Fig. 3D; Supplemental Fig. S6, A and B). Given that we observed fewer mitochondria per pollen grain in LWHY2-1 plants (Supplemental Fig. S6C), AtWHY2 may compromise mitochondrial division during pollen development via an unknown mechanism.
AtWHY2 localizes in mitochondria and aggregates with mtDNA. A, Signals of AtWHY2-GFP and mitochondrial RFP in mature pollen grains in normal exposure. Large and small arrows indicate strong and weak signals of AtWHY2-GFP, respectively. The blocked area was enlarged (bottom) to show localization of the weak signals in mitochondria. Bar = 3 μm. B, Signals of AtWHY2-GFP and mitochondrial RFP in mature pollen grains in reduced exposure. The blocked area was enlarged (bottom) to show localization of the strong AtWHY2-GFP signals in mitochondria. Bar = 3 μm. C, Signals of mtDNA (DAPI staining) merged with the strong signals of AtWHY2-GFP. Bar = 3 μm. D, Electron microscopy of mitochondria in mature pollen vegetative cells of Col-0 and LWHY2-1 plants. Arrows indicate the electron-dense nucleoid-like structure appearing in mitochondria of LWHY2-1 cells. Bars = 200 nm. E, Immunogold electron microscopy of mitochondria in mature pollen vegetative cells of Col-0 and LWHY2-1. The signals of mtDNA (5-nm gold) and AtWHY2 (10-nm gold) localize, and they colocalize into the electron-dense nucleoid-like structure. Bars = 200 nm.
The WHY2 DNA-Binding Motif Is Required to Affect mtDNA Levels
Structural study revealed a KGKAAL motif in Whirly proteins, and this motif is required for Whirly proteins to bind to single-stranded DNA (Desveaux et al., 2002). The motif in AtWHY2 appears at amino acids 62 to 67 from the N terminus. To see if the elevation of mtDNA levels requires this putative DNA-binding motif, we mutated this motif in AtWHY2 by replacing KGKAAL with QGQGGV (Fig. 4A). We prepared a transgenic construct consisting of the Lat52 promoter and the mutated AtWHY2 (AtWHY2M62-67) and GFP (Lat52:AtWHY2M62-67-GFP), obtained a homozygous transgenic plant, and crossed it with LDips-RFP. In pollen cells expressing AtWHY2M62-67-GFP and mitochondrial RFP, AtWHY2M62-67-GFP localized to mitochondria, as expected, but we observed no mtDNA signals in the mature pollen vegetative cell (Fig. 4B) and no nucleoid-like structures in the mitochondria (Fig. 4C). Quantification revealed 15.3 ± 3.5 copies of mtDNA per pollen of LWHY2M62-67-GFP, similar to the value detected for Col-0 plants (10.3 ± 2.1 copies per pollen) and much smaller than the value for LWHY2-1 plants (124 ± 11.1 copies per pollen). We thus conclude that AtWHY2 elevation of mtDNA levels requires the DNA-binding motif. Interestingly, similar to the LWHY2-1 plants, mitochondria in the LWHY2M62-67-GFP cells also appeared larger than Col-0 and LGFP mitochondria (Fig. 4C; Supplemental Fig. S6B), and LWHY2M62-67-GFP cells also had fewer mitochondria per cell (Supplemental Fig. S6C). This suggests that AtWHY2 may inhibit mitochondrial division, but this inhibition does not require the DNA-binding motif.
The DNA-binding motif in AtWHY2 is necessary for the regulation of mtDNA levels. A, Mutation of the motif KGKAAL to QGQGGV. B, Mature pollen grains carrying AtWHY2M62-67-GFP (GFP) and mitochondrial RFP (mtRFP). The pollen grains were stained with DAPI. There was no detectable signal for mtDNA in the pollen vegetative cell. SN, Sperm cell nucleus; VN, vegetative nuclei. Bar = 10 μm. C, Electron microscopy of mitochondria in the vegetative cells of LWHY2M62-67 and LGFP. Bar = 200 nm.
AtWHY2 Decreases Transmission through the Male Germ Line
The homozygous transgenic lines set seeds that appeared normal in morphology and number (Fig. 5A). This indicates that expression of WHY2 in pollen vegetative cells and the resulting high levels of mtDNA do not affect the function of pollen in sexual reproduction. We observed normal arrival of the pollen tube at the embryo sac (Fig. 5B), in agreement with this idea. However, segregation analysis of self-crosses using three heterozygous transgenic lines of LWHY2 showed skewed ratios of segregation, 2.17:1, 2.4:1, and 2.28:1 kanamycin-resistant (KanR) to kanamycin-sensitive (KanS) progeny (Table I), all deviating from the expected ratio of 3:1 and indicating unequal transmission of the transgene. As significant differences were observed in two independent transgenic lines, LWHY2-1 and LWHY2-3, we conclude that the change in segregation did not result from an effect of the insertion site of the transgene. We performed reciprocal crosses between heterozygous Col-0 and LWHY2-1 to quantify the male and female transmission efficiencies and found a significant reduction in transmission through the male lineage but no difference in transmission through the female (Table II).
Seed and pollen phenotypes of plants expressing AtWHY2. A, Developing seeds of homozygous LWHY2-1 plants compared with Col-0. Bar = 1 mm. B, Arrival of pollen tubes to the embryo sac 8 h after fertilization observed with Aniline Blue staining of pollen tubes. Arrows indicate the top of the pollen tube. Bars = 50 μm. C, Mature pollen grains of Col-0 and LWHY2-1 stained with Alexander solution. Bars = 50 μm. D, In vitro germination of pollen. Bars = 30 μm. E, In vivo growth of pollen tubes 4 and 8 h after pollination observed with Aniline Blue staining. Arrows indicate the positions of leading pollen tubes. Bar = 500 μm. F, Length of the leading pollen tubes 2, 4, 8, and 10 h after pollination. For both the genotypes, four leading pollen tubes were measured at each time point. Error bars indicate sd (Student’s t test: *, P < 0.05).
KanR and KanS were observed with T1 seeds. TE (transmission efficiency of gametes) = number of KanR/number of KanS × 100%. Double asterisks indicate highly significant difference from the expected TE (100%; χ2, P < 0.01).
To understand how transmission through the male parent is affected in the transgenic plants, we first examined the viability of mature pollen grains freshly collected from dehiscing anthers of LWHY2-1 plants by Alexander staining and in vitro germination. Our results revealed 98.3% (n = 580) and 98.7% (n = 628) positive staining of pollen grains in LWHY2-1 and Col-0, respectively, and 43% (n = 500) and 38.6% (n = 500) germination (after 4 h in germination medium) of pollen in LWHY2-1 and Col-0, respectively. This indicates that pollen of LWHY2-1 and Col-0 shows no significant difference in viability or germination (Fig. 5, C and D). Next, we inspected the growth of the pollen tube by Aniline Blue staining using Col-0 flowers pollinated with LWHY2-1 and Col-0 pollen. We found that the leading pollen tubes of LWHY2-1 were shorter than those of Col-0 (Fig. 5E). This indicates that pollen tubes of LWHY2-1 may grow more slowly than those of Col-0. Averaging the length of four leading pollen tubes within an ovary showed significant differences between the pollen tubes at 4 h (1,433 ± 57 µm for Col-0 and 1,196 ± 42 µm for LWHY2-1; P = 0.049) and 8 h (2,243 ± 100 µm for Col-0 and 1,893 ± 115 µm for LWHY2-1; P = 0.038) after pollination but showed similar lengths at 2 h (553 ± 116 µm for Col-0 and 540 ± 125 µm for LWHY2-1; P = 0.926) and 10 h (2,263 ± 75 µm for Col-0 and 2,213 ± 49 µm for LWHY2-1; P = 0.185) after pollination (Fig. 5F). These results imply that pollen grains of LWHY2-1 show reduced pollen tube growth from 500 to 1,000 µm in length. The similar length of pollen tubes at 2 h after pollination indicates that LWHY2-1 and Col-0 pollen grains show little or no difference in germination and initial growth of the pollen tube. Also, the pollen tubes exhibit a similar length at 10 h after pollination, having reached the end of the ovary.
The reduced growth of LWHY2 pollen tubes may reduce the chance for the pollen tube to reach unfertilized ovules located in the lower half of the ovary when competing with pollen tubes of Col-0. This may explain the abnormal segregation in self-crosses and the lower male transmission in reciprocal crosses (Tables I and II). To test this, we performed limited pollinations and observed that the segregation ratio of KanR to KanS approached 1:1 for pollination of Col-0 with pollen from heterozygous LWHY2 plants when we limited the number of pollen grains to no more than the number of ovules per ovary. Thus, pollination with fewer than 40 pollen grains per stigma yielded a segregation ratio of 1.02 ± 0.03, significantly different (P = 0.014) from the ratio of 0.75 ± 0.03 that we observed in pollinations with more than 200 pollen grains per stigma (Fig. 6A). We also predicted that the ratio of KanR to total would vary in seeds harvested from the upper or lower position of the silique in crosses with excess heterozygous LWHY2 pollen. Indeed, when we divided the silique into three equal portions (Fig. 6B), the seeds showed decreasing ratios of KanR to total, with 48.7% ± 0.7% KanR in the top, 41% ± 0.7% in the middle, and 34.7% ± 1% in the bottom portions (Fig. 6C). The percentage of KanR differed significantly between each two portions (P = 0.003 between the top and the middle, P = 0.009 between the middle and the bottom, and P = 0.002 between the top and the bottom), indicating that the growth of most LWHY2 pollen tubes slows when they reach approximately 700 µm in length, one-third of the distance from the top of the ovary. Thus, AtWHY2 causes a disadvantage in male competence owing to reduced pollen tube growth.
Decreased transmission by pollen expressing AtWHY2. A, Segregation KanR-KanS ratios in F1 progeny harvested with pollination of heterozygous LWHY2-1 pollen grains onto Col-0 pistils. EnPG, Excess pollen; LnPG, limited number of pollen grains; n, number of seedlings inspected. Error bars indicate sd (Student’s t test: *, P < 0.05). B, Equal division of a mature fertilized silique into top (Top), middle (Mid.), and bottom (Bot.) portions. C, The ratios of KanR to total in F1 progeny harvested from each of the silique portions with pollination with excess heterozygous LWHY2-1 pollen grains onto Col-0 pistil. n, Number of seedlings inspected. Error bars indicate sd (Student’s t test: **, P < 0.01).
AtWHY2 appears to inhibit mitochondrial division and increase the copy number of the mtDNA. To examine which of these cellular events, or both, causes the reduced strength of pollen tube growth, we examined segregation ratios in the heterozygous transgenic lines of LWHY2M62-67, as these derive from homozygous lines that show inhibition of mitochondrial division without increased mtDNA copy number (Fig. 4, B and C; Supplemental Fig. S6, B and C). Our results for three independent transgenic lines showed KanR:KanS ratios of 2.91:1, 3.03:1, and 2.97:1, which fits the expected ratio of 3:1 (Table III). This indicates equal transmission of gametes in a self-cross of heterozygous LWHY2M62-67, revealing that the transgenic alterations of mitochondrial size and number do not affect male competence (pollen tube growth). Therefore, we conclude that the high copy number of mtDNA in pollen vegetative cells causes the reduced pollen tube growth.
High mtDNA Copy Number Causes Respiratory Abnormalities
Pollen germination and pollen tube growth require rapid and ample mitochondrial respiration (Rounds et al., 2011). Given that neither pollen viability nor morphology is aberrant in LWHY2 plants (Fig. 5C), the reduced pollen tube growth may indicate defects in energy production or consumption in LWHY2 pollen. We measured the ATP level in mature anthers of Col-0 and LWHY2-1 plants and found 19.98 ± 1.58 mmol ATP g−1 (fresh weight) in LWHY2-1 and 30.84 ± 3.61 mmol g−1 in Col-0 anthers (Fig. 7A). Thus, pollen grains of LWHY2-1 harbor less ATP, 65% of that in Col-0 (19.98/30.84) as measured in the entire anther. For a better understanding of the mitochondrial condition that causes this reduction in cellular ATP, we further measured ADP, NADH, and NAD+, finding 236.8 ± 40.1 nmol ADP g−1 fresh weight in LWHY2-1 and 285.6 ± 68.2 nmol g−1 in Col-0 anthers (Fig. 7A). This reveals an imbalance in the ATP-ADP ratio, which was 84.4 (19.98/0.2368) for LWHY2-1 and 107.9 (30.84/0.2856) for Col-0 (Fig. 7A), indicating possible mitochondrial dysfunction. NADH, at 89.5 ± 13.5 nmol g−1 for LWHY2-1 anthers, exhibited a 246% increase compared with 36.4 ± 2.4 nmol g−1 for Col-0. NAD+, at 351.1 ± 18.9 nmol g−1 for LWHY2-1 anthers compared with 410.3 ± 23.1 nmol g−1 for Col-0, and the imbalanced ratio of NADH to NAD+, at 0.255 (89.5/18.9) for LWHY2-1 compared with 0.089 (36.4/410.3) for Col-0 (Fig. 7B), support the idea that defective respiration occurs in the pollen grains of LWHY2 possessing more copies of the mtDNA.
Respiratory abnormality in mature pollen grains of LWHY2-1. A, Amounts of ATP and ADP quantified in mature LWHY2-1 anthers in comparison with Col-0. Error bars indicate sd (Student’s t test: *, P < 0.05). B, Amounts of NADH and NAD+ quantified in mature LWHY2-1 anthers in comparison with Col-0. Error bars indicate sd (Student’s t test: *, P < 0.05; and **, P < 0.01). FW, Fresh weight.
Mitochondrial dysfunction generally also involves the generation of excess reactive oxygen species (ROS; Alandijany et al., 2013). We detected the accumulation of extra ROS in the pollen grains of LWHY2-1 (Fig. 8, A and B), corroborating the abnormality of LWHY2-1 mitochondria. To test if this increased ROS level might cause oxidative stress in pollen cells, we examined the expression of ALTERNATIVE OXIDASE (AOX) and ALTERNATIVE INTERNAL NAD(P)H DEHYDROGENASE (NDA), which encode proteins that participate in the alternative pathway of plant mitochondria known to reduce the formation of ROS (Maxwell et al., 1999; Svensson and Rasmusson, 2001). We also examined AOX1 protein and observed increased AOX1 levels (Fig. 8C). Our results showed increased AOX1a, AOX1d, NDA1, and NDA2 transcript levels (Fig. 8D) in the pollen grains of LWHY2-1, indicating that genes in this pathway express at higher levels. In contrast, we observed no increase in proteins in the ordinary pathway of mitochondrial respiration in the LWHY2-1 pollen (Supplemental Fig. S7). The enhanced levels of ROS might cause cellular stress, thus activating the alternative pathway, the cellular mechanism involved in reducing the level of ROS, to eliminate the stress. Our further inspection showed significant and rapid loss of pollen viability during storage (Fig. 8, E and F), supporting the idea that pollen grains of LWHY2-1 likely experience a certain degree of cellular stress.
ROS accumulation and reduced pollen viability over time in LWHY2-1. A, Mature pollen grains of Col-0 and heterozygous and homozygous LWHY2-1 incubated with 2′,7′-dichlorodihydrofluorescein diacetate. Green fluorescence indicates the level of ROS. Bar = 40 μm. B, Relative levels of ROS measured by quantification of the fluorescence from each pollen grain. The values are averages of five measurements. Error bars indicate sd. C, Relative amount of AOX1 shown by immunoblot. HSP is the loading control. D, Relative transcript levels of AOX1a, AOX1d, NDA1, and NDA2, which encode proteins in the alternative pathway. The values were normalized to the nucleus-encoded UBIQUITIN5 and averaged from three independent experiments, with the transcript abundance of Col-0 adjusted as 1. Error bars indicate sd. E, Viability staining of pollen after 4 and 48 h of storage. Red fluorescence indicates pollen grains that lost viability. Bar = 60 μm. F, Percentage of living pollen grains after various times of storage. The values were acquired by averaging three individual tests with 300 pollen grains per test. Error bars indicate sd.
DISCUSSION
Low mtDNA Copy Numbers May Be Sufficient for Plant Cell Viability
Direct observation revealed that approximately two-thirds of mitochondria in Arabidopsis mesophyll cells lack mtDNA (Wang et al., 2010). This insufficiency of mtDNA occurs not only in Arabidopsis but also in different tissues of various plants (Preuten et al., 2010). Thus, although the universality of this phenomenon remains to be examined in many plant clades, we can conclude that plants can have fewer copies of the mtDNA than the number of mitochondria in their cells. This is a special property of plant cells, because animal cells have more than one copy of the mtDNA per mitochondrion (Pikó and Matsumoto, 1976; Robin and Wong, 1988) and reduction in the mtDNA copy number in animal cells causes mitochondrial defects and disease (for review, see Montier et al., 2009). Given that the insufficiency of mtDNA in the somatic or the pollen cells of plants did not compromise cell respiration (Preuten et al., 2010; Wang et al., 2010), we predict that plant and animal cells may have evolved different mechanisms to maintain and use the information in the mitochondrial genome. A previous study observed frequent and coupled fusions and divisions of mitochondria in cultured tobacco cells (Arimura et al., 2004), implying that plant mitochondria may have a dynamic mechanism to share the insufficient copies of the mtDNA within the cell.
To test whether the low mtDNA copy number is appropriate for plant cells to maintain and use mitochondrial genetic information, here we used a direct experimental strategy to elevate the amount of mtDNA per cell and examined the result. No homolog of TFAM, the well-known regulator of mtDNA levels in animals (Larsson et al., 1998; Alam et al., 2003; Ekstrand et al., 2004; Kanki et al., 2004), has been identified in plants; therefore, we used a plant-specific mtDNA-binding protein, WHY2, which may similarly affect mtDNA levels in plants. Our analysis targeted the pollen vegetative cell because it has a very low mtDNA copy number (about 10 copies per pollen; Wang et al., 2010) and WHY2 does not express in pollen (Honys and Twell, 2004), making the pollen vegetative cell a natural knockout cell line of WHY2. As shown above, our results provide experimental evidence for the prediction that, in plants, having more mtDNA than the insufficient level does not benefit mitochondrial function; rather, it compromises mitochondrial function. The low copy number of mtDNA, therefore, is conducive to proper mitochondrial function and growth of plant cells. This indicates that the mtDNA copy number also matters in plant cells, as is observed in animal cells (Moraes, 2001; Montier et al., 2009).
The mechanism by which mtDNA copy number affects mitochondrial function, however, remains unknown. With the transgenic LWHY2-1 plants, we detected abnormal activation of mitochondrial transcription (Supplemental Fig. S8). This was abnormal because the level of mitochondrial proteins remained unvaried (Supplemental Fig. S7). We propose that this abnormality may occur as part of the cellular responses to the abnormally high mtDNA copy number and may require future research.
AtWHY2 Alters mtDNA Copy Numbers
Another purpose of this study was to test whether AtWHY2, an Arabidopsis mtDNA-binding factor, could affect the copy number of mtDNA. As one of the ubiquitous plant-specific Whirly family proteins, WHY2 localizes in plant mitochondria, unlike the other family members, WHY1 and WHY3, which localize in plastids (Krause et al., 2005). These Whirly family proteins share a typical DNA-binding structure (Desveaux et al., 2002) and participate in the maintenance of organellar genomes. Recent work demonstrated that a double knockout of AtWHY1 and AtWHY3 results in chloroplast defects due to illegitimate recombination in the plastid genome (Maréchal et al., 2009; Lepage et al., 2013), and AtWHY1 and AtWHY3 appear to stabilize the plastid genome by repressing short-range rearrangements (Zampini et al., 2015). Structural analysis also indicated that AtWHY2 binding may stabilize mtDNA to favor accurate repair of DNA double-strand breaks (Cappadocia et al., 2010). Taken together, the evidence acquired so far shows the great importance of WHY proteins in maintaining the stability, rather than affecting the quantity, of organelle genomes. Here, we show a significant increment of mtDNA copy number (10-fold or greater) with accumulation of AtWHY2 in the pollen vegetative cells, indicating a remarkable effect of AtWHY2 upon mtDNA quantity. Given that high copy numbers of mtDNA occurred in independent transgenic lines and that transgenic expression of GFP with the equivalent vector (LGFP) did not affect the copy number of mtDNA, we conclude that AtWHY2 affects the mtDNA levels in the pollen vegetative cell.
Thus, our result reveals a potential novel function for this Whirly family protein in regulating mtDNA levels. In addition, our investigation reports, to our knowledge, the first nonenzyme protein factor that affects mtDNA levels in plants. In fact, except for the core enzymes of mtDNA replication, PolIA and PolIB, which increase the amount of mtDNA (Parent et al., 2011), and an exonuclease, DPD1, which degrades mtDNA (Matsushima et al., 2011; Tang et al., 2012), few plant factors have been experimentally shown to affect mtDNA quantity. The DNA-binding factors so far identified in plants, such as OSB, RecA, and MSH, likely affect mtDNA via its recombination and repair rather than regulate mtDNA levels (for review, see Maréchal and Brisson, 2010).
Intriguingly, however, our result in pollen appears to contradict the results obtained in somatic cells. In Arabidopsis, transgenic overexpression of AtWHY2 using the 35S promoter does not cause an increase in mtDNA or mitochondrial transcript levels but rather a decrease in their levels (Maréchal et al., 2008). To verify this, we repeated the previous transgenic approach and obtained identical results (Supplemental Fig. S9). The exact reason for such a conflict is unclear. However, given that the complete knockout of AtWHY2 results in no phenotypes in the plant or in mitochondria, this suggests that functional homologs of AtWHY2, or other compensating mechanisms, may exist in somatic mitochondria (Maréchal et al., 2008). If this were the case, the result of this report may imply that the homologs or mechanisms probably do not function in the pollen vegetative cell.
AtWHY2 May Impede mtDNA Degradation
The regulation of mtDNA copy number is of particular interest in animal cells because imbalance in this regulation, causing either increases or decreases in copy number, has been implicated in human disease (Moraes, 2001; Montier et al., 2009). Work in animal cells demonstrated that continuous, well-balanced replication and turnover (degradation) maintain the levels of mtDNA in cells (Berk and Clayton, 1974; Kai et al., 2006). Accordingly, factors affecting either degradation or replication likely function as key regulators of mtDNA levels. TFAM, as one of the key regulators of mtDNA levels in animals, is proposed to regulate the copy number of mtDNA by limiting the rate of mtDNA turnover (Ekstrand et al., 2004; Matsushima et al., 2004). TFAM associates with the full mtDNA and packages it into protein-DNA aggregates (nucleoids; Alam et al., 2003; Garrido et al., 2003), thus protecting the mtDNA from degradation. This depends on TFAM’s tight but nonspecific DNA binding and high abundance in mitochondria (Fisher et al., 1992; Ekstrand et al., 2004). The other regulatory factors found in animals, TWINKLE helicase and mtSSB, facilitate mtDNA replication by helix destabilization (Korhonen et al., 2003, 2004; Tyynismaa et al., 2004), a different mechanism from that of TFAM.
The mechanism by which AtWHY2 affects mtDNA copy number is not yet clear. However, as a nonenzyme DNA-binding factor without sequence specificity (Maréchal et al., 2008; Cappadocia et al., 2010), AtWHY2 may affect mtDNA copy number in a way similar to TFAM, by impeding mtDNA degradation by pollen mitochondrial DNases such as DPD1 (Matsushima et al., 2011; Tang et al., 2012). This agrees with a previous study showing that AtWHY2 protects DNA against nuclease digestion in vitro (Cappadocia et al., 2010). Our results showing the appearance of nucleoid-like structures containing mtDNA (Fig. 3) and that knockout of the DNA-binding ability of AtWHY2 results in the loss of its effect on mtDNA copy numbers (Fig. 4) support this idea.
Another important observation indicating that AtWHY2 may act by protecting mtDNA from degradation, rather than by promoting mtDNA replication, is that AtWHY2 expression results in an equivalent amount of mtDNA to that detected in BCPs. Our results show that, relative to the mature pollen of Col-0, which contain approximately 10 copies of mtDNA per pollen cell, the LWHY2 transgenic lines contain approximately 120 to 130 copies per pollen grain. This is equivalent to that detected in the BCPs (approximately 130 copies; Fig. 1). The elevation of mtDNA by AtWHY2 may occur through a mechanism maintaining the mtDNA copy number during pollen development. Given that the degradation of mtDNA in pollen vegetative cells starts at the early BCP stage (Nagata et al., 1999) and that the Lat52 promoter drives gene expression in pollen vegetative cells beginning at the microspore stage (Twell et al., 1989, 1990; Eady et al., 1995), earlier accumulation of AtWHY2 in the vegetative cell may prevent the degradation of mtDNA in the transgenic lines. To solidify this inference, we expressed AtWHY2:GFP in the generative cell of Arabidopsis using the Histone Three Related10 promoter, which drives gene expression beginning at the late BCP stage (Ingouff et al., 2007). Our results showed the correct import of AtWHY2 into the generative mitochondria but undetectable signals of mtDNA in the sperm cells of TCP (Supplemental Fig. S10). This observation indicates that AtWHY2 fails to maintain mtDNA levels when imported to mitochondria after the initiation of mtDNA degradation, agreeing with the idea that AtWHY2 affects the copy number of mtDNA by protecting the mtDNA from degradation.
MATERIALS AND METHODS
Plant Materials and Growth Conditions
Arabidopsis (Arabidopsis thaliana) ecotype Col-0 was used as wild-type plant material and for transgenic screening. Seeds were surface sterilized and sown on 0.7% (w/v) agarose plates containing one-half-strength Murashige-Skoog medium (Sigma-Aldrich) supplemented with 3% (w/v) Suc. For transgenic screening, the medium above was supplemented with kanamycin to 30 mg L−1. Arabidopsis seedlings were grown at 23°C under 16 h of light/8 h of darkness and transferred to sterilized nutrient soil 14 d after imbibition.
Constructs and Plant Transformation
All plasmids constructed in this study were modified from pGreen-Lat52:Dips-GFP (Matsushima et al., 2008). A 717-bp region of AtWHY2 was amplified from wild-type leaf complementary DNA (cDNA) using the primers WHY2-SalF and WHY2-PstR. The product was cloned into pGreen-Lat52:Dips-GFP to replace the Dips coding sequence and generate pGreen-Lat52:AtWHY2-GFP (LWHY2-GFP), which was then verified by sequencing. AtWHY2 amplified by primers WHY2-SalF and WHY2-XbaR was cloned into pGreen-Lat52:Dips-GFP to replace Dips-GFP and generate pGreen-Lat52:AtWHY2 (LWHY2), which was then verified by sequencing. Subsequently, GFP was amplified with the primers GFP-SalF and GFP-XbaR and cloned into pGreen-Lat52:Dips-GFP to replace Dips-GFP and create pGreen-Lat52:GFP, which was used as a control. To create pGreen-WHY2pro:GUS and pGreen-WHY2pro:WHY2-GFP (WWHY2-GFP), which were used to control the expression of GUS and WHY2-GFP, a 2,010-bp fragment of the AtWHY2 promoter was amplified from wild-type genomic DNA using the primers proWHY2-KpnF and proWHY2-SalR and cloned into the pGreen vector.
The generative cell-specific promoter HTR10 was amplified using the primers proHTR10-KpnF and proHTR10-SalI and cloned to pGreen-Lat52:AtWHY2-GFP to replace Lat52 and create pGreen-HTR10:AtWHY2-GFP. To generate AtWHY2M62-67, two separate mutant fragments were amplified from pGreen-Lat52:AtWHY2 using the primers WHY2-SalI and WHY2M62-67-R and the primers WHY2M62-67-F and WHY2-PstI. Then, the target AtWHY2M62-67 was amplified using the previous two products as templates and the primers WHY2-SalF and WHY2-PstR via overlapping PCR. The resulting product was cloned into pGreen-Lat52:AtWHY2-GFP to replace the native AtWHY2 and generate pGreen-Lat52:AtWHY2M62-67-GFP (LWHY2M62-67-GFP). All of the primers used are listed in Supplemental Table S1. Plants were stably transformed with Agrobacterium tumefaciens (GV3101) via the floral dip method as described previously (Zhang et al., 2006).
Determination of mtDNA Copy Number per Pollen Grain
A competitive PCR method that quantifies the copy number of the mtDNA in single plant cells, including single pollen grains (Wang et al., 2010), was used in this study to measure mtDNA copy number per pollen grain. All procedures of the quantification, including the pretreatment of pollen to release mtDNA from the cell, the preparation of competitor template, and the settings for PCR, were performed exactly as described by Wang et al. (2010). Briefly, five pollen grains for each quantification were gently crushed and pretreated by rapid freezing, thawing, and proteinase K digestion. The resulting mixture was divided equally into five PCR tubes. The mtDNA copy number per pollen grain was obtained after two rounds of competitive PCR with mitochondrial matR as target. The efficiency coefficient for target/competitor amplification (0.8 for AtMatR±; Wang et al., 2010) was used in the elaboration of the value.
RNA and DNA Extraction and RT-Quantitative PCR
Total RNA was extracted from different Arabidopsis tissues using TRIzol reagent (Invitrogen). The concentrations of total RNA were quantified spectrophotometrically by the Nanodrop ND-2000c system (Thermo Fisher Scientific). A total of 5 μg of RNA was reverse transcribed using the PrimeScript First-Strand cDNA Synthesis Kit (TaKaRa) according to the manufacturer’s instructions. The quantitative PCR amplification containing 5 μL of 2×SYBR premix ExTaq (TaKaRa), 0.2 μm of each primer, and 80 ng of cDNA in a 10-μL mixture was performed on the LightCycler II Real-Time PCR System (Roche). The primers used for RT-quantitative PCR are listed in Supplemental Table S3. The PCR conditions consisted of an initial denaturation at 95°C for 10 min followed by 45 cycles of 95°C for 15 s and 60°C for 45 s. Data were analyzed using LightCycler3 software (Roche). The expression levels of target transcripts were normalized to the level of the internal standard UBIQUITIN5 using the ΔCT method (2−ΔCT = relative amount of transcripts; ΔCT = CTTarget – CTInternal standard).
The quantification of mtDNA levels in leaves used quantitative PCR as described (Preuten et al., 2010). Total DNA was extracted from fresh leaves using a cetyl-trimethyl-ammonium bromide protocol (Murray and Thompson, 1980). The mtDNA levels in total DNA were measured by amplification of mitochondrial matR, which was measured by quantitative real-time PCR. Quantitative real-time PCR was performed in the LightCycler II Real-Time PCR System (Roche) using the SYBR premix ExTaq kit (TaKaRa). Real-time PCR was performed in a 10-μL reaction mixture containing 5 μL of 2×SYBR premix ExTaq, 0.2 μm of each primer, and 40 ng of genomic DNA. The primers used for quantitative analysis of mtDNA are listed in Supplemental Table S1. The following amplification was used: 95°C for 10 min and 45 cycles of 95°C for 20 s and 60°C for 45 s. Data were analyzed using LightCycler3 software (Roche). All mtDNA levels were normalized to the internal standard nucleus-encoded single-copy gene phage-type RNA polymerase as described previously (Preuten et al., 2010) using the ΔCT method (see above).
Polyclonal Antibody Production and Immunoblot Analysis
To produce polyclonal antibodies against WHY2, ATP1, ATP3, apocytochrome b, and succinate dehydrogenase1-1, the synthetic polypeptides of these proteins were used to immunize rabbits, and the resulting polyclonal antibodies were purified from antisera using antigen-antibody affinity chromatography (Abgent). The sequences of polypeptides used for immunization are listed in Supplemental Table S2.
The mature pollen grains from newly opened flowers at stage 13, according to flower development in Arabidopsis (Smyth et al., 1990), were collected and ground in ice-cold extraction buffer (100 mm Tris-HCl, 15 mm EDTA, and 200 mm NaCl, pH 7.5, supplemented with 0.5% [v/v] Triton X-100, 1 mm phenylmethylsulfonyl fluoride, and 0.1% [v/v] β-mercaptoethanol), and then the homogenized slurry was centrifuged at 10,000g and 4°C for 15 min. The supernatant was mixed with SDS-PAGE loading buffer, boiled, and separated by gel electrophoresis. Then, the gel was transferred to a polyvinylidene difluoride membrane (Millipore) in transfer buffer (25 mm Tris, 190 mm Gly, and 20% [v/v] methanol) for 1.5 h at 18 V. The polyvinylidene difluoride membrane was blocked for 30 min at room temperature in 1× Tris-buffered saline (TBS; 0.2 m Tris-HCl, pH 7.5, and 5 m NaCl) containing 5% (w/v) skim milk and probed with the corresponding polyclonal antibodies (1:1,000 in 1× TBS containing 0.1% [v/v] Tween 20) for 1 h at room temperature. After three washes in 1× TBS containing 0.1% (v/v) Tween 20, the membranes were incubated with the secondary antibody, donkey anti-rabbit IgGIRDye (dilution, 1:10,000; ODYSSEY; Li-Cor), for 1 h at room temperature. The immunoblots were detected with the ODYSSEY (Li-Cor) according to the manufacturer’s instructions.
Fluorescence Microscopy
For direct observation of mtDNA signals from the pollen vegetative cell, fresh pollen grains placed onto a glass slide were immersed in TAN buffer (Nemoto et al., 1988) supplemented with 3% (v/v) glutaraldehyde and 10 μg mL−1 DAPI (Invitrogen). After covering with a coverslip, the samples were examined with an epifluorescence microscope (DMI 6000B; Leica). For mature pollen grains, to remove the interference of the pollen walls, the coverslip was pressed down, bursting the pollen grains and releasing the vegetative cytoplasm, thus allowing clear detection of mtDNA signals. Photomicrographs of squashed pollen grains were captured with a CCD camera (DFC480; Leica) attached to the microscope.
Electron Microscopy
For transmission electron microscopy, fresh pollen grains from open flowers were fixed with 5% (v/v) glutaraldehyde in 0.1 m sodium cacodylate buffer (pH 7.4) at room temperature for 4 h and then postfixed overnight in 1% (w/v) osmium tetroxide at 4°C. The samples were dehydrated through a series of alcohol solutions and embedded in Spurr’s resin. Ultrathin sections were incubated with 1% (w/v) uranyl acetate and lead citrate and examined with an electron microscope (TECNAI G2 20; FEI) at an accelerating voltage of 120 kV.
Immunoelectron microscopy for detecting mtDNA and AtWHY2 was performed according to the standard method (Johnson and Rosenbaum, 1990). Briefly, fixation of pollen grains was performed as described above, except that the postfixation step was omitted. The fixed samples were embedded in LR White resin (Sigma-Aldrich). For localization of mtDNA, the sections were incubated first with a mouse anti-DNA antibody (1:20 dilution; Sigma-Aldrich) and then with goat anti-mouse IgM conjugated to 5-nm colloidal gold (1:60 dilution; British BioCell International). As a negative control, sections were pretreated with DNase and processed as above. For the localization of AtWHY2, the sections were incubated first with a rabbit anti-AtWHY2 antibody prepared as described above (1:50 dilution) and then with goat anti-rabbit IgG conjugated to 10-nm colloidal gold (1:60 dilution; British BioCell International). The colocalization of mtDNA and AtWHY2 was performed with sections treated with the above procedures in series. All sections were examined with an electron microscope (TECNAI G2 20; FEI) at the accelerating voltage of 120 kV after a brief staining with 1% (w/v) uranyl acetate.
Phenotypic Analyses of Pollen and Pollen Tubes
For in vitro germination, mature pollen grains from newly opened flowers were spread on culture plates prepoured with germination medium (1.5% [w/v] agar, 0.01% [w/v] H3BO3, 5 mm KCl, 5 mm CaCl2, 1 mm MgSO4, and 18% [w/v] Suc, pH 7.5; Boavida and McCormick, 2007). The plates were sealed with Parafilm and incubated at 23°C. To assay in vivo pollen tube growth, wild-type pistils were pollinated with wild-type and LWHY2-1 pollen. After specific periods of tube growth, the pistils were fixed in acetic acid:ethanol (1:3) solution, cleared in 8 m NaOH, and stained with Aniline Blue as described previously (Mori et al., 2006). The images of germinated pollen and pollen tubes stained by Aniline Blue in the pistils were taken on a microscope (DMI 6000B; Leica) equipped with Leica Application Suite software, and the length of pollen tubes was measured with ImageJ software (National Institutes of Health; http://rsb.info.nih.gov/ij).
The viability of mature pollen grains was investigated after propidium iodide (50 μg mL−1; Sigma-Aldrich) staining for 3 min, and pollen was examined with a fluorescence microscope (DMI 6000B; Leica; Hashida et al., 2013). The mature pollen grains were preprocessed by storage for various periods (4, 12, 24, 36, 48, 60, and 72 h) in dry conditions at 23°C.
The ROS level of mature pollen grains was indicated by staining with 2′,7′-dichlorodihydrofluorescein diacetate (10 μm; Sigma-Aldrich) in 1× PBS (pH 7.4) for 5 min using a fluorescence microscope (DMI 6000B; Leica) with absorption and emission wavelengths at 488 and 530 nm, respectively (Matsushima et al., 2008). Image capture and quantification of the pollen fluorescence intensity were performed by the Leica software. Each detection was repeated three times.
ATP, ADP, NAD+, and NADH Measurements
To assay ATP and ADP concentrations in mature pollen, pollen from four newly opened flowers was harvested and ground in 200 μL of distilled, deionized water, the mixture was centrifuged at 10,000g at 4°C for 5 min, and the supernatant was divided equally into four microfuge tubes. The ATP levels were measured by the luciferin-luciferase method according to the instructions of the ENLITEN ATP Assay System (Promega). The ADP levels were tested using the ADP-Glo Kinase Assay Kit (Promega) for converting ADP to ATP, and total ATP was quantified by luminescence in the Multifunctional Microplate Reader (FlexStation3; Molecular Devices).
For the measurement of NAD+ and NADH, mature pollen from eight newly opened flowers was collected and ground in 400 μL of NAD+/NADH Extraction buffer using the NAD+/NADH Quantification Kit (Sigma-Aldrich), and then the homogenized slurry was centrifuged at 10,000g at 4°C for 3 min and the supernatant was divided equally into eight microfuge tubes. The amounts of NAD+ and NADH were measured for the A450 using the Multifunctional Microplate Reader (FlexStation3; Molecular Devices) according to the manufacturer’s instructions.
Sequence data from this article can be found in the Arabidopsis Genome Initiative or GenBank/EMBL databases under the following accession numbers: AtWHY2, At1g71260; AtHTR10, At1g19890; mitochondrial maturase of Arabidopsis, Y08501.
Supplemental Data
The following supplemental materials are available.
Supplemental Figure S1. Pollen mtDNA decreases to low levels during pollen development.
Supplemental Figure S2. Quantification of mtDNA per BCP and mature pollen grain of Col-0 using competitive PCR.
Supplemental Figure S3. Pattern of AtWHY2 expression in Col-0 Arabidopsis.
Supplemental Figure S4. Pollen mtDNA levels in different transgenic lines.
Supplemental Figure S5. Plant and flower phenotypes of LWHY2-1.
Supplemental Figure S6. Characteristics of mitochondria from Col-0 and transgenic lines.
Supplemental Figure S7. Proteins crucial for the ordinary pathway of mitochondrial respiration in pollen do not differ significantly between Col-0 and LWHY2-1.
Supplemental Figure S8. Mitochondrial transcription was significantly enhanced in the pollen of LWHY2-1 plants.
Supplemental Figure S9. Overexpression of AtWHY2 driven by the 35S promoter did not substantially affect the copy number of mtDNA in somatic cells.
Supplemental Figure S10. AtWHY2 localization in the mitochondria of generative cells when expressed under the control of the AtHTR10 promoter.
Supplemental Table S1. Primers used in this study.
Supplemental Table S2. Sequences of polypeptides used for polyclonal antibody production.
Acknowledgments
We thank Wataru Sakamoto for generously providing the transgenic Arabidopsis seeds (Lat52:Dips-GFP and Lat52:Dips-RFP) and Dr. Ying-Chun Hu for technical assistance in electron microscopy.
Footnotes
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.plantphysiol.org) is: Sodmergen (sodmergn{at}pku.edu.cn).
Q.C. and L.G. performed the experiments and analyzed the data; Z.-R.S. contributed to the data analysis; D.-Y.W. contributed technical assistance to the single pollen quantification of mtDNA; Q.C. prepared the article; Q.Z. supervised and complemented the writing; Q.Z. conducted the experiments; S. conceived the project and revised the article.
↵1 This work was supported by the National Basic Research Program of China (grant no. 2013CB126900).
↵2 These authors contributed equally to the article.
Glossary
- mtDNA
- mitochondrial DNA
- BCP
- bicellular pollen
- RT
- reverse transcription
- Col-0
- Columbia-0
- KanR
- kanamycin-resistant
- KanS
- kanamycin-sensitive
- ROS
- reactive oxygen species
- cDNA
- complementary DNA
- TBS
- Tris-buffered saline
- DAPI
- 4′,6-diamino-phenylindole
- Received March 24, 2015.
- Accepted July 15, 2015.
- Published July 20, 2015.
REFERENCES
