Heat stress interferes with chromosome segregation and cytokinesis during male meiosis in Arabidopsis thaliana

ABSTRACT

In higher plants, male meiosis is a key process of microsporogenesis and is crucial for plant fertility. Male meiosis programs are prone to be influenced by altered temperature conditions. Studies have reported that an increased temperature (28°C) within a fertile threshold can affect the frequency of meiotic recombination in Arabidopsis. However, not much has been known how male meiosis responses to an extremely high temperature beyond the fertile threshold. To understand the impact of extremely high temperature on male meiosis in Arabidopsis, we treated flowering Arabidopsis plants with 36-38°C and found that the high-temperature condition significantly reduced pollen shed and plant fertility, and led to formation of pollen grains with varied sizes. The heat stress-induced unbalanced tetrads, polyad and meiotic restitution, suggesting that male meiosis was interfered. Fluorescence in situ hybridization (FISH) assay confirmed that both homologous chromosome separation and sister chromatids cohesion were influenced. Aniline blue staining of tetrad-stage pollen mother cells (PMCs) revealed that meiotic cytokinesis was severely disrupted by the heat stress. Supportively, immunolocalization of ɑ-tubulin showed that the construction of spindle and phragmoplast at both meiosis I and II were interfered. Overall, our findings demonstrate that an extremely high-temperature stress over the fertile threshold affects both chromosome segregation and cytokinesis during male meiosis by disturbing microtubular cytoskeleton in Arabidopsis.

KEYWORDS:

• Male meiosis
• Heat stress
• Chromosome segregation
• Meiotic cytokinesis
• Microtubule

Introduction

Meiosis is a specialized type of cell division during which DNA duplicates once while chromosomes separate twice.¹ At prophase of meiosis I, meiotic recombination occurs which ensures the reciprocal exchange of genetic information between parental chromosomes and allows the novel combination of gene alleles. Initiation of meiotic recombination is launched by the formation of double-strands breaks (DSBs) that is catalyzed by a highly conserved topoisomerase VI protein SPO11.² In Arabidopsis, both SPO11-1 and SPO11-2 are required for meiotic recombination.^3–5 Thereafter, recombinases RAD51 and DMC1 process the DSBs for recombination, with RAD51 acting as an accessory factor of DMC1.^6–9 After the completion of meiotic recombination, homologous chromosomes separate at anaphase of meiosis I. At early stages of meiosis II, the centromeric cohesions between sister chromatids are still retained, and are released at anaphase II, leading to the disjoining of sister chromatids at the end of meiosis with a resultant four haploid chromosome sets, reviewed by .¹⁰ In Arabidopsis, many key cohesion factors, including AtSMC1, AtSMC3, SYN1 (AtRec8), AtSCC3 and AtSGO1, have been identified and characterized.^11–16 Meiotic cytokinesis takes place afterward of chromosome segregation, while dicot plants and monocot plants have distinct meiotic cytokinesis patterns. For dicot plants, i.e., Arabidopsis, meiotic cell walls are formed after the separation of sister chromatids at the end meiosis II, and this pattern is named ‘simultaneous cytokinesis.’¹⁷ In contrast, monocot plants, e.g., rice and maize, undergo successive-type of cytokinesis. For this kind of meiotic cytokinesis, each round of chromosome segregation is followed by a construction of cell wall formation, which leads to a dyad formation after meiosis I and a tetrad formation after meiosis II.¹⁷ In Arabidopsis, male meiotic cytokinesis is regulated by a PAMK signaling cascade AtNACK2/TES/STUD-AtANP3-AtMKK6-AtMPK4.^18–22

Male meiosis in plants is sensitive to either increase or decrease of environment temperatures.^23–25 In both dicot and monocot plants, cold stress has been found to interfere with cytoskeleton organization and cell wall formation during male meiosis and consequently lead to the production of unreduced gametes.^26–28 In addition, studies have reported that a minor decrease of temperature (e.g., 19°C) can influences the frequency of crossover (CO) in Arabidopsis.^29,30 Compared with cold stress, high-temperature stress seems to affect more aspects of meiosis. Male meiotic progression is interfered in hexaploid wheat (Triticum aestivum L.) incubated shortly under high temperatures (30 and/or 35°C).³¹ An increased but within fertile threshold temperature (28°C) causes a higher meiotic recombination rate in A. thaliana, which was shown to act through the class I CO pathway.^29,32,33 Higher temperatures beyond fertile range have been found to influence chromatin configuration at early meiotic stages, and induces polyploid and/or aneuploid spore formation.^34,35 Nevertheless, the cellular mechanism underpinning how plant male meiosis responses to an extremely increased temperature that impairs plant fertility, is not understood yet. To this end, we studied the impact of a high-temperature condition (36–38°C) on male meiosis in Arabidopsis. We found that the heat stress severely impairs pollen shed and plant fertility. Cytological analysis of developing PMCs revealed that the heat stress interferes with both chromosome segregation and cytokinesis, probably by impacting the construction of tubulin-dependent spindle and phragmoplast. This study contributes to the understanding how plant male meiosis responses to high-temperature stress.

Results

Heat stress reduces plant fertility and induces pollen grains with varied sizes

To examine the impact of extremely high temperature on Arabidopsis, flowering wild-type Col-0 plants were incubated at 36–38°C, which has been shown to severely influence the seed set in plants, ^36–38 for 24 h and were transferred back to 20°C for recovery. After 10 d post the heat stress, we checked the silique development and counted the number of seeds in the influenced siliques. Compared with control plants, which displayed normal silique development (Figure 1a), heat-stressed plants produced shorter siliques on both main shoots and lateral shoots (Figure 1a, see red arrow). The number of seeds yielded in the heat-affected siliques was significantly reduced (Figure 1b). Alexander staining was then performed to check the pollen development at 5 d later upon the heat treatment. In contrast to control anthers, the 24 h-heat stress severely reduced pollen shed (Figure 1d,e). The 12 h heat treatment influenced pollen shed slightly, and the 6 h treatment did not show an obvious impact (Figure 1f,g). Remarkably, pollen grains with different sizes were observed in all the three treatment conditions (Figure 1c,e–g, see blue boxes). Germination assay was then performed to verify the bioactivity of pollen grains with different sizes. At 24 h later after setting on germination medium, pollen grains from control plants germinated and produced pollen tubes (Figure 1h). However, differently sized-pollen grains from 24 h heat-stressed plants did not germinate (Figure 1i–k). Overall, these findings indicated that the heat stress significantly affects male sterility and the seed set of Arabidopsis. Additionally, since differently sized-pollen grains could have different chromosome numbers, ³⁹ earlier processes during microsporogenesis were likely to be impacted by the heat stress.

Figure 1. Heat stress reduces pollen viability and plant fertility. (a), Siliques from main shoots and lateral shoots in control plants and heat-stressed plants. Red arrows indicate short siliques induced by the heat stress. Scale bar = 5 cm. (b), Histogram showing the average number of seeds yielded by the siliques of control plants and the heat-stressed plants. For both control and heat stress conditions, three individuals were used for quantification. Error bars represent standard deviation. (c), Histogram showing the diameter of 10 pollen grains produced by control, 24 h-stressed, 12 h-stressed and 6 h-stressed plants. (d–g), Alexander staining of anthers in control plants (d), and in the plants treated by 24 h (e), 12 h (f) and 6 h (g) heat stresses. Blue boxes indicate the pollen grains with different sizes. Scale bar = 100 μm. (h–k), Germinated pollen grain in control plant (h), and non-germinated pollen grains with varied sizes from the 24 h heat-stressed plants (i–k). n indicates the number of observed pollen grains with corresponding phenotypes. Scale bar = 20 μm.

Heat stress induces aberrant tetrad formation

To elucidate whether heat-induced pollen grains with varied sizes were caused by alterations in male meiosis, we analyzed tetrad stage PMCs by performing orcein staining. Flowering Col-0 plants were treated by 36–38°C for 24 h, and were transferred back to 20°C for 6 h recovery. In control plants, PMCs produced normal tetrads with four equal-sized spores (Figure 2a,b). However, in the heat-stressed plants, about 56% irregular meiotic products were observed (Figure 2a). Unbalanced-triads, which harbor two or more nuclei in one spore, occupied the highest proportion among the abnormal products (Figure 2a, around 34%; F and G). Unbalanced-dyads occurred at 9.21% (Figure 2a,d). Balanced-dyads (Figure 2h), unbalanced-tetrads that contain four spores but with more than one nucleus in each spore (Figure 2c), and polyads (Figure 2e), were consistently visualized, either, and occurred at a same frequency (Figure 2a, 3.95%). Group of multiple nuclei with varied sizes in one spore (Figure 2c,f,g) and occurrence of polyad (Figure 2e) suggested that chromosome segregation is unbalanced. Dyad formation (Figure 2d,h) indicated that meiotic cytokinesis is affected. These figures thus suggest that the heat stress-induced pollen grains with different sizes originated from these abnormal meiotic products, and the chromosome dynamics and/or cytokinesis during male meiosis were influenced.

Figure 2. Heat stress induces abnormal meiotic products. (a), Pie chart showing the frequency of meiotic products from control plants and heat-stressed plants. For control 50 meiotic products were counted, and for heat stress treatment 76 meiotic products were quantified. (b–h), Orcein staining of tetrad in control plant (b), and unbalanced-tetrads with multiple nuclei in one spore (c), unbalanced-dyad (d), polyad (e), unbalanced-triad with multiple nuclei in one spore (f,g), and balanced-dyad (h) in the heat-stressed plants. Scale bar = 10 μm.

Heat stress induces unbalanced chromosome segregation

To elucidate the impact of heat stress on male meiosis, we analyzed the chromosome dynamics by performing fluorescence in situ hybridization (FISH) on meiotic spread using a centromere-specific DNA probe. At diakinesis and metaphase Ι, control plants displayed five pairs of bivalents with paired centromeric foci that represent a full synapsis of homologous chromosomes and a regular formation of CO (Figure 3a,b). In contrast, the heat-stressed plants exhibited 10 univalents together with 10 centromeric foci, which suggested that either homology pairing and/or synapsis, or meiotic recombination are compromised by the heat stress (Figure 3g,h). The formation of univalents caused unbalanced separation of homologous chromosomes after meiosis Ι (Figure 3i,j); moreover, centromeric foci with a number over 10 at anaphase Ι and interkinesis were occasionally observed, which suggested an occurrence of premature loss of sister chromatid cohesion (Figure 3i,j). Finally, at metaphase ΙΙ and telophase ΙΙ, unbalanced chromosome sets were formed in the heat-stressed plants (Figure 3k,l). These data demonstrate that the heat stress interferes with chromosome dynamics of male meiosis and leads to unbalanced chromosome segregation.

Figure 3. Centromeric FISH analysis of heat-stressed PMCs. A-F, Diakinesis- (a), metaphase Ι- (b), anaphase Ι- (c), interkinesis- (d), metaphase ΙΙ- (e) and telophase ΙΙ-stage (f) PMCs in control plants. G-L, diakinesis- (g), metaphase Ι- (h), anaphase Ι- (i), interkinesis- (j) metaphase ΙΙ- (k) and telophase ΙΙ-stage (l) PMCs in heat-stressed plants. Chromosomes are stained by DAPI, and centromeres are labeled by centromere-specific cen180_oligo2/5/6 probe, and displayed by green. Bifluorescence are shown by merged cells. Numbers displayed in the Centromere columns represent the number of chromosomes. n in the Merged columns indicates the number of cells observed with the corresponding phenotypes. Scale bars = 10 μm.

Heat stress interferes with male meiotic cytokinesis

Next, we analyzed meiotic cell wall formation by performing aniline blue staining that marks callosic cell walls to examine the impact of heat stress on male meiotic cytokinesis. Tetrads in control plants displayed ‘cross’-shaped meiotic cell walls between the four separated nuclei (Figure 4a); in contrast, the PMCs from heat-stressed plants produced abnormal meiotic products with unbalanced cell walls (Figure 4b–h). The smaller and larger cavities represented spores containing different chromosome numbers, which were supportive for the chromosome segregation defects. Moreover, we found that the callosic cell walls in the heat-induced aberrant products were thicker than that in the control tetrads (Figure 4b–h). These findings revealed that the high-temperature stress interferes with male meiotic cytokinesis.

Figure 4. Heat stress causes abnormal meiotic cytokinesis. A-H, Aniline blue staining of tetrad-stage PMCs in control plant (a) and heat-stressed plants (b-h). Scale bar = 10 μm.

Heat stress influences the construction of microtubule cytoskeleton

The formation of meiotic cell walls relies on the construction of cytoskeleton between the separated nuclei.²⁶ In order to clarify the cause of abnormal meiotic cytokinesis upon the heat stress, we performed immunolocalization of ɑ-tubulin in both control and heat-stressed PMCs. In control plants, homologous chromosomes are fully paired at pachytene (Figure 5a), and COs are formed at diakinesis, when five bivalents are visible and microtubular network is constructed surrounding the chromosomes (Figure 5b). At metaphase Ι, a spindle structure is shaped at two sides of aligned chromosomes (Figure 5c), which pulls the homologous chromosomes to the opposite polar of the cell at anaphase Ι (Figure 5d). Phragmoplast is temporarily formed between the separated homologies after meiosis Ι (Figure 5e). At metaphase ΙΙ, two pairs of spindles are constructed (figure 5f) and conduct the disjoining of sister chromatids, which consequently contributes to tetrad formation, when mini-phragmoplast; i.e., radial microtubule arrays are generated surrounding the haploid nuclei (Figure 5g). Balanced chromosome segregation and complete meiotic cytokinesis during male meiosis lead to haploid spore production (Figure 5h).

Figure 5. Microtubular cytoskeleton in PMCs of control plants. A-H, PMCs at pachytene (a), diakinesis (b), metaphase Ι (c), anaphase Ι (d), interkinesis (e), metaphase ΙΙ (f), tetrad (g) and unicellular microspore (h) stages in control plants. White: DAPI signal; green: ɑ-tubulin signal. Scale bar = 10 μm.

On the contrary, heat-stressed plants displayed alterations in the microtubular cytoskeleton formation. At early stages of male meiosis, the heat stress seemed to have no obvious impact on the construction of microtubule around the chromosomes (Figure 6a,b). However, at metaphase Ι, irregular spindle with reduced microtubule abundance and dotted fibers was observed (Figure 6c); and occasionally the typical bipolar configuration of spindle was not formed with the chromosomes linked by linear microtubules (Figure 6d,e). This suggests that the heat stress severely impaired the formation and/or stability of spindle at meiosis Ι. The mis-segregation of chromosomes was thereafter induced with the chromosomes unevenly grouped at anaphase Ι (Figure 6f,g). These aneuploid chromosome sets are likely to develop into the mini-nuclei at later stages (Figure 6h,i,l, see red arrows). ‘Extra’ phragmoplast was found to construct between the aneuploid chromosome sets at anaphase Ι (Figure 6f,i, see yellow arrows). These ‘extra’ phragmoplast would define the location of cell walls and therefore support the observation by aniline blue staining (Figure 4c,e,g). Meanwhile, significantly reduced microtubule fibers occurred suggesting that the phragmoplast stability was influenced, either (Figure 6g). Mini-chromosome groups underwent nucleation at interkinesis (Figure 6h,i, see red arrows). At metaphase ΙΙ, spindle-like structures were visualized but were incomplete and the direction was abnormal (Figure 6j). Unbalanced chromosome sets occurred either at telophase ΙΙ and irregular phragmoplasts were formed between the dispersive chromosome groups (Figure 6k). At tetrad stage, both adjacent nuclei and multiple nuclei were found, and for the former case a complete omission of RMAs occurred between the unseparated nuclei which indicates for meiotic restitution (Figure 6l and M, see yellow arrows). Consequently, aneuploid microspores were yielded (Figure 6n). These findings demonstrate that the high-temperature stress has a significant impact on microtubular cytoskeleton during male meiosis, and suggests that the meiotic cytokinesis defects result from the misstructured microtubules.

Figure 6. Microtubular cytoskeleton in PMCs of heat-stressed plants. A-N, PMCs at pachytene (a), diakinesis (b), metaphase Ι (c–e), anaphase Ι (f,g), interkinesis (h,i), metaphase ΙΙ (j), telophase ΙΙ (k), tetrad (l,m) and unicellular microspore (n) stages in heat-stressed plants. White: DAPI signal; green: ɑ-tubulin signal. Red arrows indicate mini-nucleus (h,i,l), and yellow arrows indicate extra phragmoplast or omission of RMAs (f,i,l). Scale bar = 10 μm.

Discussion

In flowering plants, male meiosis ensures the stability of genome ploidy of life cycle by halving the chromosome number to generate haploid gametes. Alterations occurring during meiosis either cause unbalanced chromosome segregation that leads to aneuploid gametes and impaired plant fertility, or evoke meiotic restitution with resultant unreduced gametes.^14,40 As sessile organisms, plants are prone to be influenced by environmental conditions, and the male reproductive development is especially sensitive to temperature changes.^23,24 In this study, we found that a 36–38°C heat stress condition that severely reduces plant fertility interferes with both chromosome dynamics and cytokinesis in Arabidopsis. This report demonstrates that an extreme increase of environmental temperature condition would not only affect fertility but also disturb genomic ploidy consistency of plants by influencing with male meiosis. As in Populus, ⁴¹ the heat stress induced about 55% meiotic defects, but caused a severely impacted male sterility and seed set in Arabidopsis (Figure 1), which suggests that high-temperature stress induces male sterility of plants not merely by affecting male meiosis, but also by damaging post-meiosis developmental programs; e.g., tapetum development and microspore maturation.⁴²

Heat stress affects chromosome dynamics in both meiosis Ι and ΙΙ. At diakinesis, PMCs with 10 univalents (Figure 3g) suggested that homology synapsis and/or recombination was omitted by the heat. Since the pairing and synapsis of homologous chromosomes depend on proper construction of chromosome axis and synaptonemal complex (SC), the univalents could be induced by heat-interfered chromosome structures.^43–46 Examination of the key axis element ASY1 and the SC central element ZYP1 would contribute to the understanding of univalent formation. On the other hand, abnormalities of machineries that catalyze meiotic recombination and/or CO formation also cause univalents.^9,46-48 In wheat, high temperature has been shown to induce a significantly increased number of univalent and irregularities in chromosome alignment at metaphase Ι, and DMC1 was proposed to mediate the heat tolerance.⁴⁹ It is possible that heat stress predominantly interferes with DSB repair and meiotic recombination during male meiosis. To this end, a detection of recombinases DMC1 and RAD51 should be applied in future study. At the same time, we observed a premature separation of sister chromatids prior to anaphase ΙΙ in heat-stressed plants (Figure 3i,j), which suggests an impacted protection of sister centromeric cohesion under the high-temperature stress. This could be caused by alterations occurring in loading or stability of cohesion factors; e.g., AtREC8, AtSGO1 or other cohesion complex members on the chromosomes. In Arabidopsis, meiotic recombination rate increases under a milder heat stress (28°C), suggesting that a limited elevation of temperature has a positive effect on CO formation.^29,32 The fully impaired bivalent formation visualized in our study points to a notion that meiotic recombination machineries of plants can only endure a relatively small range of temperature increase. Monitoring of multiple temperature conditions at a wider range may favor the understanding of how plant meiotic recombination responses to temperature alterations. In Populus and persimmon, similar meiotic defects and unreduced gamete formation were recorded.^41,50 This suggests that the response of male meiosis to high-temperature stress may be conserved among plant species.

Our tubulin immunostaining assay and the report by Wang et al. suggested that heat-induced cytokinesis irregularities are caused by omission and/or deconstruction of phragmoplast, or RMAs (Figure 5).⁴¹ It was evidenced that high-temperature stress interferes with microtubular organization independently of affecting protein synthesis.⁵¹ It is thus plausible that during meiotic cytokinesis in both Arabidopsis and Populus, the stability of microtubules decreases upon temperature increasing, and the restoration of phragmoplast after incubation under normal temperature is due to recovered tubulin polymerization. The generation of RMAs is mediated by the interaction of microtubule organizing center (MTOC)-originated actin fibers and microtubule, and MTOCs localize at the surface of newly formed nuclei at telophase ΙΙ.^52–54 It was proposed that the positioning of separated chromosome sets and the later nucleation determines the spatial sites where RMAs are set.^55–57 Therefore, the cytokinesis defects caused by the heat-interfered RMA structuring could be owing to chromosome segregation irregularities and the consequent position of unbalanced and adjacent nuclei. When the formed nuclei get too close to each other, RMAs are not able to form which then causes meiotic restitution (Figure 6l, see yellow arrows). This possibility suggests that the spindle alteration may be the primary effect of heat stress on male meiosis in plants. A real-time monitoring of GFP-tagged tubulin dynamics would contribute to the understanding how tubulin-dependent cytoskeleton during male meiosis responses to temperature increase.^58,59

Cold stress can induce alteration in the landscape of male meiotic recombination.²⁴ In addition, cold stress seems to specifically influence meiotic cytokinesis in both Arabidopsis and crops.^26–28 It thus is likely that the formation of meiotic cell walls is more sensitive to cold than other meiotic programs, e.g., chromosome condensation and segregation. Moreover, compared with cold stress that does not influence spindle formation, ^26,27 heat stress impacts multiple meiotic processes, which suggests that the response of male meiosis to low and high temperature is mediated by different signalings. Transcriptome study and examination on the expression of meiosis genes in both cold- and heat-stressed meiosis-staged anthers may contribute to the understanding how male meiosis responses to different temperature alterations.

Material and methods

Plant materials and growth conditions

Arabidopsis (Arabidopsis thaliana) Columbia-0 plants were obtained from the Nottingham Arabidopsis Stock Center. Seeds were germinated in soil for 6–8 d and seedlings were transferred to soil and cultivated in growth chambers with a 16 h day/8 h night, 20°C, and 50% humidity condition.

Heat stress treatment

Young flowering plants were transferred to a humid chamber with a 16 h day/8 h night, 36–38°C stress condition and cultivated for 24 h. For fertility analysis, siliques and anthers were examined after 10 and 5 d of heat treatment, respectively. For tetrad analysis, heat-treated plants were transferred to control condition for 6 h for recovery and tetrad stage flower buds were examined. For immunolocalization experiment, samples were fixed right away upon the finish of heat treatment.

Cytology and immunolocalization

Pollen viability was examined by Alexander staining.⁶⁰ Tetrad analysis by orcein and aniline blue staining was performed as described previously.⁶¹ Buds producing significant numbers of mature meiotic products were used for quantification. FISH experiment was performed by using the protocol reported previously.⁶² Immunolocalization analysis of ɑ-tubulin was performed as described previously.²⁷ ɑ-tubulin monoclonal antibody (YL1/2) (Invitrogen, MA1-80017) was used with a dilution of 1:200, and goat anti-rat IgG (H + L) secondary antibody conjugated by FITC (Invitrogen, 31629) was diluted by 1:400.

Pollen germination assay

Mature pollen grains from control plants or the 24 h heat-stressed plants with a 5 d recovery at normal temperature conditions, are soiled on pollen growth medium (5 mM CaCl₂, 1 mM MgSO₄, 5 mM KCl, 0.01 mM H₃BO₃, 18% Suc, 1.5% agarose, pH at 7.5 modified by 0.1 M NaOH), and the pollen tubes are observed after incubation in a humid chamber under dark for 24 h.

Microscopy

Bright-field images was pictured using M-Shot ML31 microscope equipped with a MS60 camera. Fluorescence microscopy was performed using an Olympus IX83 inverted fluorescence microscope equipped with an X-Cite lamp and a Prime BSI camera. Bifluorescent images and Z-stacks were processed using Image J. Brightness and contrast settings were adjusted using PowerPoint.

Disclosure statement

All authors claim that there is no interest of conflict in this work.

Additional information

Funding

The work is supported by the Fundamental Research Funds for the Central Universities, South-Central University for Nationalities [CZY20001] (to B.L.); and the Fundamental Research Funds for the Central Universities, South-Central University for Nationalities [304/YZZ18007] (to B.L.).