Mechanisms of suppression of pistil primordia in male flowers and of stamen primordia in female flowers differ in diclinous plants. In this study, we investigated how cell death and cell cycle arrest are related to flower organ formation in Silene latifolia. Using in situ hybridization and a TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labeling) assay, we detected both cell cycle arrest and cell death in suppressed stamens of female flowers and suppressed pistils of male flowers in S. latifolia. In female flowers infected with Microbotryum lychnidis-dioicae, developmental suppression of stamens is released, and cell cycle arrest and cell death do not occur. Smut spores are formed in S. latifolia anthers infected with M. lychnidis-dioicae, followed by cell death in the endothelium, middle layer, tapetal cells and pollen mother cells. Cell death is difficult to detect using a fluorescein isothiocyanate-labeled TUNEL assay due to strong autofluorescence in the anther. We therefore combined a TUNEL assay in an infrared region with transmission electron microscopy to detect cell death in anthers. We show that following infection by M. lychnidis-dioicae, a TUNEL signal was not detected in the endothelium, middle layer or pollen mother cells, and cell death with outflow of cell contents, including the nucleoplast, was observed in tapetal cells.
Introduction
Hermaphroditic flowers, which have both pistils and stamens, blossom in most flowering plants. However, unisexual flowers blossom only in approximately 10% of flowering plants (Ainsworth 2000). Half of these are diclinous plants, as they have male and female flowers in the same plant. The other half are dioecious plants, as they have male and female flowers in individual plants (Charlesworth and Guttman 1999). Most of the diclinous and dioecious plants have stamen and pistil primordia during their early developmental stages.
In later stages, the anther is suppressed until maturity in the female flower, whereas the pistil is suppressed until maturity in the male flower (Ainsworth 2000, Mitchell and Diggle 2005). The timing of stamen and pistil suppression differs among species. When stamen and pistil primordia are formed in the diclinous plants Zea mays and Cucumis sativus, developmental suppression occurs in the pistils of male flowers and in the stamens of female flowers (Calderson-Urrea and Dellaporta 1999, Bai et al. 2004). On the other hand, when stamens form pollen in the female flowers and pistils form an embryo sac in the male flowers in Opuntia stenopetala diclinous plants, developmental suppression occurs in the pistils of the male flowers and in the stamens of the female flowers (Strittmatter et al. 2006, Flores-Rentería et al. 2013).
It is thought that suppression related to male and female flower formation involves two mechanisms. One mechanism is cell cycle arrest in specific organs, and the other is cell death, also in specific organs (Diggle et al. 2011). In situ hybridization (ISH) using histone and cyclin probes is effective for detecting the activity of cell division in suppression organs (Matsunaga et al. 2004, Kim et al. 2007, Daher et al. 2010). Cells that express the cyclin and histone genes undergo cell division. Histone genes are expressed in the DNA synthetic period and during endoreduplication, and cyclin genes are expressed in the DNA synthetic and Gap 2 periods. Organ suppression is related to cell cycle arrest, because histone H4 is not expressed in stamens in female flowers or in pistils in male flowers in Phoenix dactylifera (Daher et al. 2010).
Developmental suppression is caused by cell cycle arrest, because cyclin B is not expressed in the stamen primordia in the female flowers in Z. mays. The terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) signal, which detects cell death, is observed in the pistil primordia of male flowers of Z. mays (Kim et al. 2007). Specific TUNEL signals and chromatin condensation are detected in the stamen primordia of the female flowers at Stage 7 for C. sativus (Hao et al. 2003). However, TUNEL signals and chromatin condensation are not detected in the pistils of the male flowers of C. sativus (Bai et al. 2004). Cell death is detected only in the pistil primordia of the male flowers of Z. mays using a TUNEL assay (Kim et al. 2007). Cell death is detected only in the stamen primordia in C. sativus female flowers (Hao et al. 2003). The suppression mechanisms of pistil primordia in male flowers and in the stamen primordia in female flowers differ in diclinous plants.
Cell death in tissues of dioecious plants is detected by the TUNEL assay, but practical examples of the TUNEL assay in dioecious plants do not exist. However, it is well known that programmed cell death occurs in tapetal cells in plants. The TUNEL assay is often used on tissue sections to detect cell death in tapetal cells (Li et al. 2006). The programmed cell death of tapetal cells is characterized by a gradual decrease in cytoplasm structure. Cytoplasmic reduction, fragmentation of DNA, vacuolar explosion and expansion of the endoplasmic reticulum have been observed as tapetal cells in Lobiva raushii and Tillandsia albida collapse (Papini et al. 1999). The programmed cell death of the tapetal cell occurs with the post-meiotic development process (Sanders et al. 1999).
The bisexual form was observed in the male and the female flowers of Silene latifolia in Stages 1–6 (Grant et al. 1994; Supplementary Figs. S1, S2). The pistil developed in the female flower after Stage 7, and the stamen was suppressed (Supplementary Fig. S1). Then, the stamen developed and the pistil was suppressed in the male flower (Supplementary Fig. S2). Microbotryum violaceum has recently been differentiated into three species due to host specificity, one being renamed M. lychnidis-dioicae (Denchev et al. 2009). Male flowers infected with M. lychnidis-dioicae formed a smut spore instead of pollen in the anthers. The female flowers infected with M. lychnidis-dioicae formed stamens, which are typically not formed in female flowers, and the stamens formed smut spores instead of pollen in the anthers (Uchida et al. 2003).
Schema of areas predicted to exhibit cell death and cell cycle arrest in the flower organs of a wild-type (WT) female, an infected female, a WT male and an infected male. We hypothesized that cell death and/or cell cycle arrest can be detected in the areas shown in red. The black area indicates that the flower was infected with smut fungus.
Results
Cell division at organ extension and mature stages
In situ hybridization using cyclin B and histone H4 probes in Stages 3–4, 7–8 and 10–11 of WT males, WT females, infected males and infected females. Signals were detected in Stages 3–4 (a–d) and 7–8 (e–h) of each flower bud meristem (a–h). Signals were detected in the suppressed stamen in Stages 10–11 of WT females (i, m). Signals were not detected in the suppressed pistil in Stages 10–11 of WT males or infected males (k, l, o, p). However, signals were detected in the developed filament in Stages 10–11 of the infected females, but signals were not detected in the anther (j, n). (a) Female flowers at Stages 3–4, (b) infected female flowers at Stages 3–4, (c) male flowers at Stages 3–4, (d) infected male flowers at Stages 3–4, (e) female flowers at Stages 7–8, (f) infected female flowers at Stages 7–8, (g) male flowers at Stages 7–8, (h) infected male flowers at Stages 7–8, (i, m) female flowers at Stages 10–11, (j, n) infected female flowers at Stages 10–11, (k, o) male flowers at Stages 10–11, (l, p) infected male flowers at Stages 10–11. Squares indicate high magnification areas. ds, developed stamen; dp, developed pistil; o, ovule; ow, ovary wall; se, sepal; ss, suppressed stamen; sp, suppressed pistil; and p, petal. Scale bar = 100 µm for a–p.
Both signals were detected in the stamen and pistil primordia of the WT female flowers in early developmental stages (Fig. 2a). Both signals were detected in the developing pistils of the WT female flowers in the organ extension stages (Stages 7–8; Fig. 2e). Many signals were detected in the developing pistils, especially the ovaries and ovary walls, in the WT female flowers at the mature stages (Stages 10–11; Fig. 2i, m). Neither signal was detected in the suppressed stamens of the WT female flowers at the mature stages (Stages 10–11; Fig. 2i, m). Therefore, it was revealed that cell division was arrested in the suppressed stamens of the WT female flowers at the mature stages (Stages 10–11; Fig. 2i, m).
Differences were not observed in these stages in the male flowers infected with M. lychnidis-dioicae (Fig. 2d, h, l, p). Signals were detected in the developmental stamens of female flowers infected with M. lychnidis-dioicae in the organ extension stages (Stages 7–8; Fig. 2f) as well as the WT male flowers (Fig. 4g) and male flowers infected with M. lychnidis-dioicae (Fig. 2h).
Signals were detected in the developmental stamens in female flowers infected with M. lychnidis-dioicae during the mature stages (Stages 10–11), but there were fewer signals than in the WT male flowers (Fig. 2j, k, n, o). The cell division activity in the stamens of the female flowers infected with M. lychnidis-dioicae was lower than that in the WT male flowers.
Developmental suppression of organ and cell death
Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay in Stages 3–4, 7–8 and 10–11, and anther locules of WT males, WT females, infected males and infected females. Signals were not detected in Stages 3–4 (a–d) or 7–8 (e–h) of each flower bud meristem. Nuclei are indicated by red fluorescence, while TUNEL-positive nuclei are indicated by green fluorescence (a–h). Signals were detected in the suppressed stamen in Stages 10–11 of WT females (i, m). Signals were detected in the suppressed pistil in Stages 10–11 of WT males and the infected males (k, l, o, p). However, signals were not detected in the developed stamens in Stages 10–11 of infected females. (a) a female flower at Stages 3–4, (b) an infected female flower at Stages 3–4, (c) a male flower at Stages 3–4, (d) an infected male flower at Stages 3–4, (e) a female flower at Stages 7–8, (f) an infected female flower at Stages 7–8, (g) a male flower at Stages 7–8, (h) an infected male flower at Stages 7–8, (i, m) a female flower at Stages 10–11, (j, n) an infected female flower at Stages 10–11, (k, o) a male flower at Stages 10–11, (l, p) an infected male flower at Stages 10–11, (q) bright field images of a WT male at Stages 10–11, (r) fluorescent image of a WT male at Stages 10–11, (s) high magnification of anthers from (r), (t)anther locules. Squares indicate high magnification areas. ds, developed stamen; dp, developed pistil; pg, pollen grain; ss, suppressed stamen; sp, suppressed pistil; se, sepal; and ta, tapetal cell. Scale bar = 100 µm for a–t. Double bar = 50 µm for m–p.
TUNEL signals were not detected in the WT female or male flowers during the early developmental (Stages 3–4) or organ extension (Stages 7–8; Fig. 3a, h) stages. TUNEL signals were detected in the suppressed stamens of the WT female flowers during the mature stages (Stages 10–11; Fig. 3i, m). Cell death did not occur in the pistils or stamens of female flowers infected with M. lychnidis-dioicae, as TUNEL signals were not detected (Fig. 3j, n). However, TUNEL signals were detected in the suppressed pistils in the WT male flowers (Fig. 3k, o). TUNEL signals were observed over the base from the center of the suppressed pistils, but these signals were not confirmed because they were weak with low magnification. TUNEL signals were detected in the males infected with M. lychnidis-dioicae, as well as in the suppressed pistils of the WT male flowers (Fig. 3l, p). These results showed that cell death occurred in the suppressed pistils of WT males and in the male flowers infected with M. lychnidis-dioicae at maturity, as well as in the suppressed stamens of WT females at maturity (Stages 10–11).
Cell death in the anther locule and M. lychnidis-dioicae
High-resolution TUNEL assay in WT anthers. The anthers of the five developmental stages in WT males were compared for nuclear DNA fragmentation using a TUNEL assay. Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI), as indicated by blue, whereas TUNEL-positive nuclei are indicated by magenta and appear white when the blue and magenta areas are merged. (a-1–a-6) WT males at Stage 5, (b-1–b-6) WT males at Stage 6, (c-1–c-6) WT males at Stage 7, (d-1–d-6) WT males at Stage 8, (e-1–e-6) WT males at Stage 9. Scale bars = 50 µm for a-1 to e-6.
High-resolution TUNEL assay in male anthers infected with M. lychnidis-dioicae. The anthers of the five developmental stages in male anthers infected with M. lychnidis-dioicae were compared for nuclear DNA fragmentation using the TUNEL assay. Nuclei were stained with DAPI, as indicated by blue, whereas TUNEL-positive nuclei were indicated by magenta and appear white when the blue and magenta areas are merged. (a-1–a-6) An infected male at Stage 5, (b-1–b-6) an infected male at Stage 6, (c-1–c-6) an infected male at Stage 7, (d-1–d-6) an infected male at Stage 8 and (e-1–e-6) an infected male at Stage 9. The arrowhead indicates a high electron density karyoplasm, probably the same structure indicated in Fig. 6. Scale bars = 50 µm for a-1 to e-6.
Transmission electron micrographs of tapetal cells in infected males. Tapetal cells lead to abnormal disintegration caused by M. lychnidis-dioicae infection. (a) Tapetal cells in an infected male at Stage 6. (b) Disintegrated tapetal cells in an infected male at Stage 7. Arrows indicate a disrupted cell wall. THe arrowhead indicated a high electron density karyoplasm, Mc, middle layer cell; Mb, multivesicular body; N, nucleus, S, spore; Tc, tapetal cells; Tc*, disintegrated tapetal cells. Scale bar = 2 µm for a and b.
We used high magnification to observe the pollen mother and tapetal cells in the anthers. We could not detect programmed cell death in tapetal cells using an FITC-labeled TUNEL assay because tapetal cells strongly autofluoresce in the 520 nm range (Fig. 3r–t). Therefore, we used Alexa 647 to observe anthers in an infrared region with the least autofluorescence. The nucleus was stained with DAPI, as indicated by blue, whereas TUNEL-positive nuclei were indicated by magenta and white merged with blue and magenta.
TUNEL signals were not detected in the WT male flowers at Stage V (Fig. 5a-1–a-6). At Stage VI, the TUNEL signal was detected in the tapetal and pollen mother cells during meiosis (Fig. 4b-1–b-6). TUNEL signals were detected with tapetal cells only in Stages VII and VIII (Fig. 4c-1–c-6, d-1–d-6). TUNEL signals were detected in parts of just a few tapetal cells during Stage IX (Fig. 4e-1–e-6).
TUNEL signals were detected with tapetal cells only in the male flowers infected with M. lychnidis-dioicae at Stage V (Fig. 5a-1–a-6). During Stages VI–VIII, the pollen mother cell also collapsed, but the liquid-state-formed TUNEL signals was detected only in tapetal cells (Fig. 5b-1–b-6, c-1–c-6, d-1–d-6). The TUNEL signals were only detected with the tapetal cells at Stage IX (Fig. 5e-1–e-6). The TUNEL signals were detected only in the tapetal cells in anther locules of male flowers infected with M. lychnidis-dioicae. The collapse of the pollen mother cells in anther locules of male flowers infected with M. lychnidis-dioicae was due to the necrosis-like cell death without the TUNEL signal.
Observation of the collapse of tapetal cells using transmission electron microscopy
We observed anther locules in Stages VI and VII using transmission electron microscopy (TEM). The nebula-formed TUNEL signals were observed in the tapetal cells of the male flowers infected with M. lychnidis-dioicae but not in the tapetal cells in WT male flowers (Fig. 5). As a result, we discovered that the collapse of the membranes and cell walls of the tapetal cells in the male flowers infected with M. lychnidis-dioicae was caused by M. lychnidis-dioicae infection, and the contents of the cells were eluted (Fig. 6a, b). The tapetal cells, in which we detected TUNEL signals in the male flowers infected with M. lychnidis-dioicae, were impaired at the nuclear membranes and formed numerous vacuoles at Stage VI (Fig. 6a). The cell then completely collapsed in a high electron density karyoplasm, and the vacuole was observed (Fig. 6b). The high electron density karyoplasm was in the same area as the tapetal cells and the area of the liquid-state-formed TUNEL signal (Figs. 5,6). We also only detected the TUNEL signal in tapetal cells in the anther locules in the male flowers infected with M. lychnidis-dioicae, as TUNEL signals were not detected in the collapse of pollen mother cells caused by M. lychnidis-dioicae.
Discussion
Usefulness of the TUNEL assay for observing the infrared region in the plant
The TUNEL assay is effective for detecting cell death in animals (Grasl-Kraupp et al. 1995). There are fewer examples of applying the TUNEL assay to plants compared with animals. This difference is attributable to strong autofluorescence in plants. We performed the TUNEL assay in the infrared region, which has the weakest autofluorescence in tapetal tissues. The autofluorescence of tapetal tissue was very strong near 520 nm (Figs. 4,5). The autofluorescence of Chl consisted of red (near 685–690 nm) and infrared (near 730–740 nm) regions (Buschmann 2007). However, chloroplasts do not exist in stamens and pistils. Therefore, we could avoid autofluorescence by using the infrared region (near 730–740 nm).
Drs. Hao and Bai performed the TUNEL assay in a light field (Hao et al. 2003, Bai et al. 2004), and Dr. Kim performed the TUNEL assay as fluorescent labeling with FITC; this approach detected strong autofluorescence near 520 nm (Kim et al. 2007). However, it was difficult to detect a clear signal. We were able to get a clear signal in the tapetal cells, as the TUNEL assay was effective in the infrared region in flower organs (Figs. 4,5).
The TUNEL signal was detected in pollen mother and tapetal cells during meiosis in Stage 6 (Fig. 5c-4, c-5, c-6). Double-strand breaks, indicating recombination during meiosis, were detected. We were able to detect double-stranded breaks, such as recombination during meiosis, using high resolution, as autofluorescence was eliminated and observation in the infrared region was facilitated.
Cell death and cell cycle arrest in the flower organ
Stamens were completely suppressed in the mature female flowers of S. latifolia (Supplementary Fig. S1n). Pistils of the mature male flowers remained as filament-formed rudimentary organs (Supplementary Fig. S2k, l; Grant et al. 1994). Cell cycle arrest in the pistils of male flowers was less severe than cell cycle arrest in the stamens of female flowers (Fig. 2m–p). TUNEL signals were detected only in a few cells of the pistils (Fig. 3m–p). Therefore, we think that the pistils of the male flowers remain as rudimentary organs due to the weak developmental suppression of their pistils.
Development of the female flower stamens was suppressed by cell death, and the pistils of the male flowers were suppressed by arresting the cell division in Z. mays (Kim et al. 2007). Cell death was not observed in the pistils of the male flowers in C. sativus. Cell death, accompanied by the TUNEL signal, was observed in the stamens of female flowers (Hao et al. 2003, Bai et al. 2004). Developmental suppression of C. sativus and Z. mays was related to either cell cycle arrest or death. We identified cell cycle arrest by ISH using cyclin B and histone H4 probes, similar to the technique of Matsunaga et al. (2004) (Fig. 2). We determined cell death using the TUNEL assay (Fig. 3). When the pistil of the male flower and the stamen of the female flower were suppressed in S. latifolia, cell death and cell cycle arrest were detected. On the other hand, when the pistil of the male flower and the stamen of the female flower were developed in S. latifolia, cell division was detected (Figs. 3i, m, l, p; 4i, m, l, p). Therefore, we suggest that the function of suppressing flower development may not be specialized between the male and female in S. latifoli a because of differences between dioecious plants, such as S. latifolia, and diclinous plants, such as C. sativus and Z. mays, or because of the sexual differentiation of S. latifolia. Therefore, it was concluded that both cell cycle arrest and death, reflecting developmental suppression, were present in male and female flowers in S. latifoia.
Lytic cell death in the tapetal cell
Tapetal cells exist in the innermost four layers that form anthers, and these comprise the epidermis, endothecium, middle layer and tapetum. The tapetal cells directly touch the gametophyte, which become pollen and play an important role in development from microspore to pollen (Pacini et al. 1985). The tapetal cell, as a secretion cell layer, provides nourishment for a microspore released from a pollen tetrad until pollen development (Goldberg et al. 1993). The cytoplasm of the tapetal cell collapses in the late stage of pollen development. It is thought that this process of the collapse in the tapetal cell is related to programmed cell death in many plants (Papini et al. 1999, Wu and Cheun 2000). Programmed cell death of the tapetal cells is characterized by a graded decrease in cytoplasm structure. Cytoplasmic reductions, fragmentation of DNA, a vacuolar explosion and expansion of the endoplasmic reticulum have been observed during collapse in the tapetal cells in Lobiva raushii and Tillandsia albida (Papini et al. 1999). Tapetal cell collapse occurs at the same time as the post-meiotic developmental process of pollen mother cells (Sanders et al. 1999). Malfunction of the tapetal cells triggers male sterility (Sorensen et al. 2002, Yang et al. 2003).
Tapetal cells include amoeba and secretion types, and most of the angiosperm are of the secretion type of tapetal cells. The thickness of the secretion tapetal cell walls in A. thaliana is reduced before pollen mother cells perform meiosis, and the pollen mother cells simultaneously thin (Matsuo et al. 2013).
Lytic cell death was observed when the cell walls of males of S. latifolia infected with M. lychnidis-dioicae thinned before meiosis (Fig. 6a, b). Cell death seen in the tapetal cells in the males infected with M. lychnidis-dioicae caused programmed cell death, with the fragmentation of DNA as expected. However, it is thought that the nucleus and its contents is eluted from tapetal cells, because collapse of the cell walls and membranes is caused by M. lychnidis-dioicae.
Microbtryum lychnidis-dioicae and cell death
The TUNEL signal is detected in the tapetal cells with pollen development (Fig. 4). Tapetal cells collapse to supply nutrition and sporopollenin, which are components of the pollen wall (Goldberg et al. 1993). Male sterility, which occurs when pollen is not formed, is caused by collapsed tapetal cells. Pollen mother cells are killed and eliminated in the anther of males infected with M. lychnidis-dioicae (Uchida et al. 2003). However, cell death with TUNEL signals was not observed with the tapetal cells in the pollen mother cell (Fig. 5). These TUNEL signals were in a liquid state, unlike normal TUNEL signals. We hypothesized that the fragmented nucleus with the TUNEL signal was detected in anther locules, caused by tapetal cell collapse, and eluted to anther locules due to the cell death of the tapetal cells, which had occurred earlier as a result of M. lychnidis-dioicae. In other words, M. lychnidis-dioicae caused cell death, but it could not inhibit cell death from the anthers. Microbtryum lychnidis-dioicae does not inhibit cell death and affects cell cycle and expression of B or C function genes, because it promotes formation of stamens in female flowers (Figs. 2b, f, j, n, 3b, f, j, n), and M. lychnidis-dioicae did not control cell death (Fig. 6). Microbtryum lychnidis-dioicae changed the expression of upper genes, but it could not directly control cell death due to expression of SLM2, which is a homolog of PISTILLATA of the B function gene in A. thaliana, which is induced when M. lychnidis-dioicae infects male flowers (Kazama et al. 2005).
Materials and Methods
Plant materials and plant growth conditions
Silene latifolia seeds of inbred lines (K-line) were stored in our laboratory. The K-line was propagated for 17 generations of inbreeding to obtain a genetically homogeneous population. We also used asexual mutants, which were obtained from crossing an asexual mutant (ESS1) and an inbred line (K-line). Plants were grown from vernalized seeds in pots in a regulated chamber at 23°C, with 16 h light/8 h dark cycles.
Microbtryum lychnidis-dioicae inoculation
A1 and A2 sporidia were cultured on potato dextrose agar (BD Difco) at 23°C for 5 d and suspended at 2 × 106 cells ml–1 in distilled water. Equal concentrations of A1 and A2 sporidial mixtures were used throughout the inoculations. The inoculation treatments were performed on 10-day-old seedlings of S. latifolia on 0.8% agar plates. The base of each 10-day-old seedling was injected with 2 µl of the mixture. Inoculation was repeated after 3 d. Three weeks after the inoculation, we transferred seedlings to soil in pots and grew them in a regulated chamber at 23°C, with a 16 h light/8 h dark cycle.
In situ hybridization
Total RNA (100 ng) was reverse-transcribed into cDNA using a first-stand cDNA synthesis kit (GE Healthcare Biosciences). The probes used were SlCycA1 and SlH4, with the SlCycA1-specific primers (SlCycA1F, 5′-GGAACGCTACTTTGCGGCAT-3′ and SlCycA1R, 5′-CTACAGTCTCAGTACATGGC-3′) and SlH4-specific primers (SlH4F2, 5′-AGGAAAAAAGAAGACCAAAC-3′ and SlH4R2, 5′-AAACCCGAAACCAAAACGAA-3′). The amplified insert was used to produce digoxigenin (DIG)-labeled sense and antisense RNA probes, using a DIG RNA Labeling Kit SP6/T7 (Roche Applied Science). Flower buds were immediately fixed in FAA solution [3.7% (v/v) formaldehyde, 50% (v/v) ethanol and 5% (v/v) acetic acid] at 4°C. The fixed buds were dehydrated in an ascending ethanol series (25, 50, 75 and 100%; each step for 20 min at 4°C) and stored overnight in 100% ethanol. The samples were embedded in Histosec (Merck). The 8 µm sections were cut with a microtome and mounted on slides at 37°C overnight. ISH was performed as described by Kazama et al. (2005).
TdT-mediated dUTP nick end labeling (TUNEL) assay
Flower buds were immediately fixed in FAA solution [3.7% (v/v) formaldehyde, 50% (v/v) ethanol and 5% (v/v) acetic acid] at 4°C. Anthers were immediately double-fixed overnight in 4% glutaraldehyde in 0.1 M phosphate buffer (pH 7.2) at 4°C and post-fixed for 4 h in 2% osmium tetraoxide in distilled water. After being washed in 0.1 M phosphate buffer (pH 7.2), the fixed flowers were dehydrated in an ethanol series (30, 50, 70, 80, 90, 95 and 100%, each step for 15 min at room temperature) and stored overnight in 100% ethanol at 4°C. The ethanol was replaced with xylene and embedded in paraffin. Flowers embedded in paraffin were cut into 10 µm sections using a microtome (RV-240, Yamato Kohki). Cut sections were deparaffinized in xylene and rehydrated in an ethanol series (100, 95, 90, 80, 70, 50 and 30%, each step for 10 min at room temperature).
In situ nick end labeling of nuclear DNA fragmens was performed in a humid chamber for 1 h in the dark at 37°C with an In situ Cell Death Detection Kit (Roche) in flower buds and with a Click-iT TUNEL Alexa Fluor 647 Imaging Assay for microscopy and HCS in anthers. Samples were analyzed under a fluorescence microscope (LeicaCTR6000, Leica Microsystems). The fluorescent filter was set to view the green fluorescence of FITC at 527 ± 30 nm, the infrared fluorescence of Alexa Fluor 647 at 700 ± 70 nm and the blue fluorescence of DAPI at 470 ± 40 nm.
Transmission electron microscopy
Flower buds were dissected with fine forceps, placed in 1-hexadecene and frozen in a high-pressure freezing machine (HPM010, BAL-TEC) that was cooled with liquid nitrogen (−196°C). The samples were immediately transferred to 2% OsO4 in dry acetone at −0°C and incubated at −80°C for 100 h. The samples were then gradually warmed from −80 to 0°C over 5 h, held for 1 h at 0°C, warmed again from 0 to 23°C over 1 h, and incubated for 1 h at 23°C (Leica EM AFS, Leica Microsystems). The samples were washed three times with dry acetone at room temperature, infiltrated with increasing concentrations of Spurr’s resin in dry acetone, and finally infiltrated with Spurr’s resin. Ultra-thin sections (50 nm) were cut with a diamond knife (DIATOME Ltd.) and mounted onto Formvar-coated copper grids. The sections were stained with 3% uranyl acetate for 2 h at room temperature and examined using an electron microscope (H-7600, Hitachi Co.) at 100 kV.
Supplementary data
Supplementary data are available at PCP online.
Funding
This work was supported by the Japan Society for the Promotion of Science [a Grant-in-Aid for Challenging Exploratory Research (24657046)].
Abbreviations
- DAPI
4′,6-diamidino-2-phenylindole
- FITC
fluorescein isothiocyanate
- ISH
in situ hybridization
- TEM
transmission electron microscopy
- TUNEL
terminal deoxynucleotidyl transferase dUTP nick end labeling
- WT
wild type
Acknowledgments
We thank Michael E. Hood for his generous gift of M. lychnidis-dioicae.
Disclosures
The authors have no conflicts of interest to declare.
