Heat shock protein 101 contributes to the thermotolerance of male meiosis in maize

Abstract

High temperatures interfere with meiotic recombination and the subsequent progression of meiosis in plants, but few genes involved in meiotic thermotolerance have been characterized. Here, we characterize a maize (Zea mays) classic dominant male-sterile mutant Ms42, which has defects in pairing and synapsis of homologous chromosomes and DNA double-strand break (DSB) repair. Ms42 encodes a member of the heat shock protein family, HSP101, which accumulates in pollen mother cells. Analysis of the dominant Ms42 mutant and hsp101 null mutants reveals that HSP101 functions in RADIATION SENSITIVE 51 loading, DSB repair, and subsequent meiosis. Consistent with these functions, overexpression of Hsp101 in anthers results in robust microspores with enhanced heat tolerance. These results demonstrate that HSP101 mediates thermotolerance during microsporogenesis, shedding light on the genetic basis underlying the adaptation of male meiocytes to high temperatures.

HSP101 positively mediates thermotolerance during microsporogenesis in maize.

Introduction

Meiosis, the basis of sexual reproduction, plays a critical role in genome diversity and ploidy stability in plants. Appropriate chromosome segregation in meiosis relies on a series of highly coordinated events during prophase I, including homologous chromosome pairing, synapsis, and recombination (Zickler and Kleckner, 2015). Meiotic recombination (MR) is initiated by the programmed introduction of DNA double-strand breaks (DSBs), which is catalyzed by conserved type-II topoisomerase-like enzyme sporulation protein 11 and a group of accessory proteins (Keeney et al., 1997; Grelon et al., 2001). Following that, programmed DSB ends are processed by the MRX/N (Mre11-Rad50-Xrs2/Nbs1) and Sae2/Com1/CtIP/Ctp1 protein complex into single-stranded DNA tails (Borde, 2007). Then RADIATION SENSITIVE 51 (RAD51) and DISRUPTED MEIOTIC CDNA 1 are loaded onto the single-stranded DNA to facilitate homologous searching and single-end invasion (Hunter and Kleckner, 2001). Ultimately, DSBs in meiocytes are repaired; some generate crossovers (COs), which are required for the reciprocal exchange of genetic information and accurate chromosome segregation during meiosis (Kowalczykowski, 2015; Ceccaldi et al., 2016). Mutants of meiotic proteins involved in formation, processing, or repair of DSBs usually display severe meiotic defects, including deficient MR and synaptonemal complex (SC) assembly, which ultimately result in impaired fertility in plants (Mercier et al., 2015).

Meiosis in plants is sensitive to changes in the environment, especially temperature (De Storme and Geelen, 2014; Morgan et al., 2017; Liu et al., 2019). During early meiosis I, temperature that is increased but within the fertility threshold (28°C) promotes MR via increased CO formation frequency and distribution in Arabidopsis (Arabidopsis thaliana) (Modliszewski et al., 2018). Nevertheless, in one study, CO formation was reduced at a higher temperature (32°C), probably due to effects on homolog synapsis in Arabidopsis (De Storme and Geelen, 2020). Moreover, when ambient temperatures are extremely elevated, the synapsis of homologous chromosomes tends to fail, resulting in sharp declines in MR (Loidl, 1989; Higgins et al., 2012; Zheng et al., 2014; Ning et al., 2021). In addition to their effects on MR during meiosis I, elevated temperatures interfere with chromosome segregation and cytokinesis by disturbing the microtubular cytoskeleton, which results in chromosome stickiness, laggards, micronuclei, and aneuploid gametes during meiosis II (Wang et al., 2017; Mai et al., 2019; Lei et al., 2020). Although studies have reported the impacts of high temperature on the development of pollen mother cells (PMCs) during meiosis, few of the key genes involved in microspores’ tolerance of high temperatures have been characterized in plants, and the molecular mechanism underlying how plant male meiosis responds to heat stress remains elusive.

Members of the heat shock protein (HSP) family, also known as molecular chaperones, participate in a broad range of biological processes under environmental or endogenous stress by maintaining protein homeostasis and cellular reliability, inhibiting cell death signaling, and mediating DNA repair (Al-Whaibi, 2011; Sottile and Nadin, 2018; Dubrez et al., 2020). HSPs were grouped into several classes according to their molecular weight, including HSP100, HSP90, HSP70, HSP60, and the small HSP family, which play both distinct and overlapping roles in the life cycles of plants (Parsell and Lindquist, 1993; Al-Whaibi, 2011). The first protein factor involved in the acquisition of thermotolerance to be characterized was HSP104, a 104 kD HSP, which belongs to the HSP100 family in yeast (Saccharomyces cerevisiae) (Sanchez and Lindquist, 1990). Similarly, homologs of HSP104 in plants are also required for adaptation to heat stress (Queitsch, 2000; Nieto-Sotelo et al., 2002; Lin et al., 2014).

In response to higher temperatures, plants activate heat shock transcription factors (HSFs) to regulate the expression of HSP genes (Qu et al., 2013; Ohama et al., 2017). In maize (Zea mays), Hsp101, a member of the HSP100 family, is induced by heat stress (Nieto-Sotelo et al., 2002). Transcriptome data showed that Hsp101 is highly expressed in male meiocytes under normal growth conditions (Dukowic-Schulze et al., 2014; Nelms and Walbot, 2019). Very little is known about the possible role of Hsp101 in the development of maize male gametes. In this paper, we describe the classic maize dominant male-sterile mutant Male sterile 42 (Ms42), which exhibits defective DSB repair, homologous chromosome pairing, and SC assembly, eventually triggering univalent formation and chromosome fragmentations during meiosis. We positionally cloned the Ms42 gene and found that it encodes the HSP family protein HSP101. It is interesting that hsp101 null mutants undergo normal meiotic progression under permissive temperature but have defects in RAD51 loading in PMCs under heat stress. Moreover, overexpression of Hsp101 significantly enhances heat tolerance during microsporogenesis. Taken together, our results identify a key gene involved in microspore adaptation to high temperatures in maize.

Results

Dominant Ms42 causes meiotic defects

Ms42 is a male-sterile mutant identified from pollen mutagenesis by N-nitroso-N-methyl urea almost 30 years ago (Albertsen et al., 1993). Ms42 male-sterile plants were indistinguishable from wild-type (WT) plants in terms of plant morphology and the development of reproductive organs, except for shriveled anthers that failed to produce viable pollen grains (Figure 1A). Moreover, the Ms42 plant seed set was normal when pollinated with pollen of the inbred maize line B73 and F₁ progeny had a 1:1 segregation ratio of fertile to male-sterile plants (526:474, χ² = 2.7 < χ²[0.05, 1] = 3.84), which indicates that Ms42 specifically affects male but not female reproduction and is a monofactorial, dominant mutant.

Figure 1.

Phenotypes of WT and Ms42 plants. A, Morphological comparisons of WT and Ms42 plants. The anthers of Ms42 plants merely emerged from the glumes; pollen grains stained with 1% iodine-potassium iodide solution indicate that the pollen of Ms42 plants was nonvariable. Bar = 100 µm. B, Meiotic chromosome behaviors of PMCs in WT and Ms42 plants. Bar = 10 µm.

To determine the cause of the male-sterile phenotype of Ms42 mutants, we used 4′,6-diamino-phenylindole (DAPI) staining to investigate the behaviors of meiotic chromosomes in PMCs at different stages in both WT and Ms42 mutant plants (Figure 1B). The chromosomal behavior in Ms42 mutant appeared normal as that in WT before zygotene. Nevertheless, the meiotic chromosomes of Ms42 mostly existed as single thin threads at pachytene, and most were further condensed into univalents rather than bivalents in WT at diakinesis, which suggests that the synapsis of homologous chromosomes is defective in the Ms42 mutant.

Afterward, chromosome bridges were observed in meiosis I and meiosis II in Ms42 mutants; in contrast, the homologous chromosomes segregated equally in WT. Eventually abnormal tetrads with a mass of micronuclei and chromosome fragments formed after the two rounds of cytokinesis (Figure 1B and  Supplemental Figure S1). Therefore, the abnormal meiotic progression of PMCs is responsible for the male sterility of the Ms42 mutant.

Homologous chromosome pairing is impaired in Ms42

An evolutionarily conserved chromosome arrangement in eukaryotes clusters telomeres together in a small region (“bouquet”) on the nuclear envelope (Moiseeva et al., 2017). This specific structure is supposed to promote the initiation of homologous pairing during early prophase I (Tsai and McKee, 2011). To explore whether defective Ms42 affects the formation of the telomere bouquet, we conducted fluorescence in situ hybridization (FISH) analysis in meiocytes of WT and Ms42 in the early zygotene stage using pAtT4 as a telomere-specific probe (Richards and Ausubel, 1988; Prieto et al., 2004). The FISH results revealed a typical bouquet configuration in both WT (n = 28) and Ms42 (n = 33) male meiocytes (Figure 2, A and B), which indicates that bouquet formation is normal in the Ms42 mutant.

Figure 2.

Comparison of telomere bouquet formation and homologous pairing in the WT and Ms42 by FISH. A and B, FISH detection of telomere bouquet formation using the pAtT4 probe in zygotene meiocytes of WT (A) and Ms42 (B). Bar = 10 µm. C and D, Homologous pairing revealed by FISH using the 45S rDNA probe in WT (C) and Ms42 (D) meiocytes. Bar = 10 µm. E and F, Homologous pairing revealed by FISH using the 5S rDNA probe in WT (E) and Ms42 (F) meiocytes. Bar = 10 µm.

To further confirm whether homologous chromosome pairing is defective in the mutant, we performed FISH analysis using 45S and 5S ribosomal DNA (rDNA), two probes used to assess chromosome pairing and segregation (Li and Arumuganathan, 2001; Kato et al., 2004), in WT and Ms42 meiocytes. In the WT, both 45S and 5S rDNA produced only one signal in pachytene meiocytes (n = 38, Figure 2, C and E). In contrast, both 45S and 5S probes presented as two individual signals in Ms42 PMCs (n = 41, Figure 2, D and F). This indicates that homologous chromosome pairing is disturbed in Ms42.

Homologous chromosome synapsis fails in Ms42 mutants

The SC is a meiosis-specific supramolecular structure that forms between homologous chromosomes and is required for the formation of meiotic COs. It is assembled from the zygotene to pachytene and disassembled at diplotene (Cahoon and Hawley, 2016). To quantify the formation of univalents in Ms42, we performed immunostaining with ABSENCE OF FIRST DIVISION (AFD1), ASYNAPTIC1 (ASY1), and ZIPPER1 (ZYP1) antibodies in WT and Ms42 male meiocytes. AFD1 is a cohesion protein required for sister-chromatid cohesion, homologous pairing, and axial element (AE) elongation and maturation (Golubovskaya et al., 2006). In Ms42, AFD1 was located on the entire chromosome axis as linear signals (n = 34), which was consistent with WT (n = 27; Figure 3A). ASY1, the ortholog of S. cerevisiae Hop1 and Arabidopsis ASY1, is an AE component of the SC essential for homologous chromosome pairing and SC assembly (Armstrong et al., 2002). In the WT, ASY1 loading appeared as long linear signals along chromosome axis in zygotene (n = 45). However, ASY1 signals appeared as large numbers of dots on chromosomes in Ms42 meiocytes (n = 33), which indicates that AE installation is defective in the Ms42 mutant (Figure 3B and  Supplemental Figure S2). We further investigated the distribution of a transverse filament protein, ZYP1, which is the central element of the SC (Golubovskaya et al., 2011). In the WT, as SC assembly was completed, ZYP1 signals elongated along the entire length of synapsed chromosomes (n = 27). However, ZYP1 signals could not be observed in Ms42 meiocytes (n = 34, Figure 3A and  Supplemental Figure S3A), which suggests that the SC assembly is severely disrupted by the mutation of Ms42.

Figure 3.

Immunolocalization of proteins related to the SC and DSBs on meiotic chromosomes in the WT and Ms42. A, AFD1 and ZYP1 signals in WT and Ms42 meiocytes. Bar = 10 µm. B, ASY1 and γH2AX signals in WT and Ms42 meiocytes. Bar = 10 µm. C, DAPI and RAD51 signals in WT and Ms42 meiocytes. Bar = 10 µm. D, Quantification of γH2AX and RAD51 foci per cell in WT and Ms42 meiocytes. Horizontal lines in panels represent mean values. Significant differences were analyzed by two-tailed Student’s t test (Supplemental Data Set S1).

RAD51 loading is impaired in PMCs of Ms42

In meiosis prophase I, the formation and repair of programmed DSBs is a prerequisite for correct homologous chromosome pairing, synapsis, and recombination (Inagaki et al., 2010). γH2AX is a reliable biomarker for monitoring DSB formation (Lobrich et al., 2010). In the WT, γH2AX signals appeared as dots on zygotene chromosomes. At a similar stage, the γH2AX foci in the Ms42 meiocytes were not significantly different from the WT (Figure 3, B and D), which indicates that DSB formation is normal in Ms42 meiocytes.

To determine whether homologous pairing and synaptic defects in Ms42 are associated with improper DSB repair, we immunostained WT and Ms42 meiocytes using antibodies against RAD51, a recombinase that plays an important role in DSB repair (Pawlowski et al., 2003). In the WT, a substantial number of RAD51 punctuate foci were distributed on meiotic chromosomes in zygotene meiocytes (n = 12). However, RAD51 foci were not visible in Ms42 meiocytes (n = 19) at the same stage (Figure 3, C and D and  Supplemental Figure S3B), which indicates that the recruitment/loading of RAD51 onto chromosomes was impeded and produced serious defects in subsequent meiosis in Ms42.

Ms42 encodes an HSP family protein

To isolate the mutated locus of the Ms42 mutant, we generated two large BC₁F₁ mapping populations by crossing male sterile plants with inbred B73 and Zheng58, respectively. Linkage analyses suggested that the sterility locus was located on the long arm of chromosome 6; furthermore, map-based cloning with more than 5,000 BC₁F₁ plants narrowed down the male sterility gene to a 115 kb region between the two insertion/deletion polymorphism markers M2 and M6 (Figure 4A). According to the B73 v4 reference genomic annotation from MaizeGDB (https://www.maizegdb.org/), there were six candidate genes in this region. To determine the mutation site of Ms42, we compared the transcripts of the candidate genes using RNA-seq data from meiotic tassels from the Ms42 sterile mutant and its fertile sibling. Of the six candidate genes, only one (Zm00001d038806), which encodes the HSP101, had polymorphism between the fertile and sterile plants. Zm00001d038806 of the Ms42 mutant harbored two G-to-A substitutions at nucleotide positions 1,072 and 1,094 in the first exon (Figure 4B), which resulted in two amino acid substitutions Glu³⁵⁸ to Lys and Gly³⁶⁵ to Asp in the ATP-dependent chaperone ClpB domain of HSP101, respectively (Figure 4C).

Figure 4.

Map-based cloning of the maize Ms42 gene. A, Fine-mapping of the Ms42 gene. The Ms42 locus was mapped to a 115-kb region between genetic markers M2 and M6 on the long arm of chromosome 6 that contained six annotated genes. Recombinants are indicated below the markers. R1–R6, recombinants; F, fertility; S, sterility. B, Structure of the Ms42 gene and mutation sites in Ms42. Black boxes indicate exons, connecting lines indicate introns, and arrows indicate base mutation sites. C, The structure of the MS42 protein, containing an ATP-dependent chaperone ClpB domain. The position of the E-to-K and G-to-D mutation in MS42 is indicated.

Phylogenetic analyses showed that maize has only one copy of Hsp101, and divided HSP101 homologs from monocots and dicots into two distinct branches (Supplemental Figure S4). Nevertheless, sequence alignment showed that the two substituted amino acids of Ms42 were highly conserved in maize inbred lines and homologous proteins from different species (Supplemental Figure S5).

To further explore the biological significance of the two amino acid changes in the ATP-dependent chaperone ClpB domain of HSP101, we modeled the structure of the HSP101 protein using SWISS-MODEL (Waterhouse et al., 2018) and analyzed the electrostatic potential of the molecular surfaces of different amino acid residues. The amino acid residues changed from negatively charged Glu³⁵⁸ to positively charged Lys and from not charged Gly³⁶⁵ to negatively charged Asp (Supplemental Figure S6), which might further alter the spatial structure or biochemical properties of the HSP101 protein. Therefore, we speculate that the substitutions of two conserved amino acids in Hsp101 led to the male-sterile phenotypes of Ms42.

Knocking out dominant allele of Ms42 rescues the male-sterile phenotype

To confirm that the mutation in Hsp101 is responsible for the Ms42 male-sterile phenotype, we used CRISPR/Cas9 technology to knock out the Hsp101 gene. It is interesting that the male meiosis and pollen fertility of hsp101 null mutants (Hsp101-In1/De16) are normal under field growth conditions (Supplemental Figure S7), which enabled us to knock out the dominant allele by crossing Ms42 sterile plants with T₁ transgenic plants containing the CRISPR/Cas9 vector targeting Hsp101. From the 81 F₁ progeny identified by Sanger sequencing of the Hsp101 gene, we obtained three homozygous lines, called Ms42-De2/In1/In2, containing both single-nucleotide polymorphism (SNP) mutations in Ms42 and edited target sites in the first exon (deletion 2bp/insertion 1bp/insertion 2bp), which resulted in frameshift mutations in gene editing (Figure 5A). Then we performed immunoblotting by developing a specific antibody against HSP101 protein. The HSP101 protein was specifically detected in the young tassel of the WT, but not in the three frameshift lines (Figure 5B), which indicates that Hsp101 gene editing is effective. It is interesting that unlike the Ms42 dominant mutants (F₁ plants containing only SNP mutations), which had meiotic defects, the Ms42-De2/In1/In2 lines had normal homologous chromosome pairing, synapsis, and subsequent meiosis (Figure 5C). Pollen fertility was also consistently restored in those frameshift lines (Figure 5, D and E). Taken together, these results confirm that Hsp101 is the target gene whose mutation causes the dominant male-sterile phenotype of Ms42.

Figure 5.

Knocking out Hsp101 in Ms42 rescues the mutant meiotic defects. A, The gene structure of Ms42 and sequences of gene-edited lines at the CRISPR/Cas9 editing target site. CRISPR/Cas9 edited sequences of Ms42 show frameshift mutations in their homozygous lines. The WT sequence is shown at the bottom for comparison. B, Immunoblot analysis of the HSP101 protein in meiotic tassels of the WT and three gene-edited lines (Ms42-De2/In1/In2). Actin was used as a loading control. C, Male meiotic chromosome behaviors in heterozygous plants and three homozygous knockout lines (Ms42-De2/In1/In2). Bar = 10 µm. D, Tassels and pollen staining using 1% iodine-potassium iodide of the Ms42 mutants (F₁) and three homozygous gene-edited lines (Ms42-De2/In1/In2). Bar = 100 µm. E, Fertility of the Ms42 mutants (F₁) and three homozygous gene-edited lines (Ms42-De2/In1/In2). Error bars indicate the standard deviation. Significant differences were analyzed by two-tailed Student’s t test (**P < 0.01; Supplemental Data Set S1).

HSP101 is predominantly enriched in PMCs

To explore the expression of Hsp101 in maize, we investigated the expression pattern of Hsp101 by reverse transcription (RT) quantitative PCR (RT-qPCR) using total RNA extracted from various organs of WT plants. Hsp101 was predominantly expressed in meiotic tassel, whereas it was weakly expressed in organs such as roots, stems, leaves, ears, and immature seeds (Figure 6A). Further detailed analysis suggested that Hsp101 was highly expressed in anthers at different stages of development (Figure 6A).

Figure 6.

Expression analysis of Hsp101. A, RT-qPCR analysis of the expression of Hsp101 in various tissues of the WT (Supplemental Data Set S1). Three biological replicates were used for each sample, and Zmactin1 (Zm00001d010159) was used as an internal control. B, Immunoblotting indicated that HSP101 is predominantly enriched in meiotic tassel and anthers. Anti-Actin was used as the loading control. C, Immunolocalization in the WT probed with anti-HSP101. Preimmunization serum was used as a control. Bar = 100 µm. D, Immunofluorescent localization in WT meiocytes probed with anti-HSP101. The meiocytes of Hsp101 knockout plants, Hsp101-In1 (described below), were used as a control. Bar = 25 µm. SC, somatic cells.

Then we performed immunoblotting using a specific antibody against HSP101 to detect the accumulation of HSP101 in anthers of different lengths and vegetative tissue. Consistent with the RT-qPCR results, the HSP101 protein was highly accumulated in maize anthers at different stages of development (Figure 6B). To further reveal the spatial expression of HSP101 in the anther, we performed immunolocalization with WT anther sections. From stage 7 to stage 9 (Field and Thompson, 2016), the HSP101 protein was specifically detected in microspore mother cells (Figure 6C). Moreover, we investigated subcellular localization of the HSP101 protein on WT microsporocytes by immunofluorescent localization. Distinct signals were observed in the cytoplasm and nuclei of male meiocytes, especially in the cytoplasm (Figure 6D). Likewise, a similar protein localization pattern was observed in Ms42 meiocytes (Supplemental Figure S8). This spatiotemporal specificity in HSP101 suggests that it plays a role in the development of PMCs during meiosis.

HSP101 ensures faithful progression of meiosis under heat stress

Previous studies and our data suggested that HSP101 is dispensable for whole plant development at optimal temperatures (Hong and Vierling, 2001; Nieto-Sotelo et al., 2002). Nevertheless, based on its specific expression pattern and the biochemical properties of HSP proteins, we speculate that HSP101 might play a critical role in supporting the development of microspores under heat-stress conditions.

To test this hypothesis, we isolated two homozygous Hsp101 knockout lines with frame-shift mutations called Hsp101-In1 and Hsp101-De16 (Supplemental Figure S9A). To assess the thermotolerance of male meiosis in WT and hsp101 null mutants, we exposed plants in meiosis stages to 14 h at 32°C (low heat stress [LHS]), 14 h at 35°C (moderate heat stress [MHS]), or 24 h at 35°C (high heat stress [HHS]) and then restored them to normal conditions (28°C day/22°C night). The results showed no significant difference in pollen fertility between Hsp101-In1 and WT plants after LHS treatment. Pollen viability was significantly lower in Hsp101-In1 plants than the WT after MHS treatment (Figure 7, A and B). Furthermore, after exposure to HHS treatment, only 20%–30% of the pollen in Hsp101-In1 or Hsp101-De16 plants was viable, whereas 80% of the pollen in the WT remained fertile (Figure 7, A and B and  Supplemental Figure S9, B and C). Those results indicate that the microsporogenesis of hsp101 null plants is more sensitive to heat stress than the WT.

Figure 7.

Heat stress induces abnormal meiosis in Hsp101-In1 plants. A, Pollen viability in the WT and Hsp101-In1 under different heat stress conditions. Bar = 100 µm. B, Fertility in three WT and Hsp101-In1 tassels under different heat stress conditions. Error bars indicate the standard deviation. Significant differences were analyzed by two-tailed Student’s t test (Supplemental Data Set S1). C, DAPI-stained chromosome spreads of male meiocytes under HHS treatment during meiosis. Bar = 10 µm. D, Percentages of normal and abnormal meiocytes status from diakinesis to tetrad in WT and Hsp101-In1 under HHS conditions. The 143 and 247 meiocytes were from three WT and Hsp101-In1 tassels, respectively (Supplemental Data Set S1). LHS, 14 h at 32°C; MHS, 14 h at 35°C; HHS, 24 h at 35°C.

Furthermore, we examined meiotic chromosome behavior in PMCs of both WT and Hsp101-In1 plants under HHS conditions using DAPI staining (Figure 7C). In the WT, 90% (n = 143) of chromosome spreads of meiocytes were normal, forming 10 short clear bivalents at diakinesis and regularly shaped dyads and tetrads during meiosis II (Figure 7, C and D). However, 95% (n = 247) of abnormal meiocytes containing chromosome fragments, chromosome bridges, or lagging during meiosis I and meiosis II were observed in the Hsp101-In1 plants. Ultimately, abnormal tetrads with a mass of micronuclei were generated (Figure 7, C and D). Likewise, ∼50% (n = 158) of Hsp101-In1 male meiocytes exhibited similar meiotic defects under MHS conditions (Supplemental Figure S10). Taken together, these results suggest that HSP101 is essential for basal thermotolerance in male meiosis.

HSP101 is required for proper loading of RAD51 under heat stress

To further explore the cause of defective chromosome behavior during meiosis under high temperatures, we conducted immunostaining using antibodies against AFD1, ASY1, ZYP1, γH2AX, and RAD51 in PMCs of WT and Hsp101-In1 plants under HHS treatment. Normal AFD1, ASY1, and ZYP1 linear signals (Supplemental Figure S11 and Figure 8, A and B) and similar dot-like γH2AX signals were observed at the zygotene and early zygotene stages, respectively, in the WT and Hsp101-In1 plants (Figure 8, A and D). This indicates that the axis/SC structure and the formation of DSBs are not markedly affected by short-term high temperatures in either WT or Hsp101-In1 microspores.

Figure 8.

Heat stress interferes with RAD51 recruitment/loading in Hsp101-In1 plants. A, Immunolocalization of ASY1 and γH2AX on meiotic chromosomes of the WT (n = 16) and Hsp101-In1 (n = 16) under HHS conditions. Bar = 10 µm. B, Immunolocalization of ASY1 and RAD51 on meiotic chromosomes of the WT (n = 15) and Hsp101-In1 (n = 21) under HHS conditions. Bar = 10 µm. C, Immunolocalization of ASY1 and RAD51 on meiotic chromosomes of the WT (n = 12) and Hsp101-In1 (n = 16) under NT conditions. Bar = 10 µm. D, Quantification of γH2AX and RAD51 foci per cell in WT and Hsp101-In1 meiocytes. Horizontal lines represent mean values. Significant differences were analyzed by two-tailed Student’s t tests (Supplemental Data Set S1).

Nevertheless, after HHS treatment, the number of RAD51 signals was significantly decreased at the zygotene stage in Hsp101-In1 microspores (n = 21) compared with the WT (n = 15), which resembles the Ms42 dominant mutant (Figure 8, B and D). However, there was no significant difference in RAD51 signals between Hsp101-In1 (n = 16) and WT (n = 12) microspores at the optimal temperature (Figure 8, C and D). These results show that HSP101 plays an important role in the recruitment/loading of RAD51 onto the chromosomes required for DSB repair in male meiocytes under heat stress.

Overexpression of Hsp101 enhances thermotolerance during microsporogenesis

As Hsp101 proved to be an important factor for the thermotolerance of male meiosis in maize, we wondered whether increased expression of Hsp101 could improve thermotolerance during microsporogenesis. To test this hypothesis, we generated two independent Ubi:Hsp101 homozygous transgenic lines (Hsp101-OE1/OE2) that had significantly higher expression of Hsp101 in meiotic tassels compared to the WT (Figure 9A). The Hsp101-OEs had normal vegetative growth and pollen fertility under NT conditions (Figure 9, B and C and  Supplemental Figure S12), which suggests that altered Hsp101 expression has no obvious detrimental effects on plant growth or development. It is remarkable that although the pollen viability of both the WT and Hsp101-OE plants was impaired under higher heat treatment (HHT: 16 h and 24 h at 38°C, respectively), pollen fertility was significantly higher in the Hsp101-OE lines than in the WT plants (Figure 9, B and C).

Figure 9.

Hsp101-OE lines have higher pollen fertility under heat stress. A, Detection of Hsp101 expression in meiotic tassels of WT and Hsp101-OE lines under NT conditions. Three biological replicates were used for each sample, and Zmactin1 (Zm00001d010159) was used as an internal control. (**P < 0.01; Supplemental Data Set S1). B, Pollen viability in WT and Hsp101-OE lines under NT and heat stress conditions. Bar = 100 µm. C, Fertility in WT and Hsp101-OE tassels, under NT and HHT conditions. Each experiment was repeated at least three times. Error bars indicate the standard deviation. Significant differences were analyzed by two-tailed Student’s t test (**P < 0.01; *P < 0.05; ns, not significant, Supplemental Data Set S1).

To further investigate whether HSP101 has cytoprotective functions for male meiosis under heat stress, we examined meiotic chromosome behavior in PMCs of both WT and Hsp101-OE2 plants. Cytological results showed that the meiosis of PMC was severely disturbed in both WT and Hsp101-OE2 plants under HHT conditions, including the formation of univalent, chromosome bridges, and fragments (Supplemental Figure S13A). Nevertheless, consistent with the pollen fertility phenotype, there is a higher proportion of microspores with no obvious defects during meiosis in Hsp101-OE2 plants than in WT (Supplemental Figure S13B). All these results indicate that Hsp101 positively modulates heat tolerance of microsporogenesis in maize via its cytoprotective functions.

Discussion

Improved understanding of the mechanisms underlying thermotolerance is important in developing effective strategies for breeding crop plants with high and stable production (Wheeler and von Braun, 2013; Zhao et al., 2017; Chaturvedi et al., 2021). Most studies on crop thermotolerance focus on vegetative growth, very little is known about the response to heat stress in meiosis. Our finding that HSP101 is involved in the male meiotic DSB repair process in maize under heat stress represents a successful attempt to elucidate the genetic mechanisms underlying the thermotolerance of gametogenesis in maize.

HSP101 might facilitate meiotic DSB repair

The cytoprotective functions of HSPs in meiosis have been explained by their antiapoptotic role, especially during spermatogenesis in eukaryotes (Sarge and Cullen, 1997; Ji et al., 2012; Sottile and Nadin, 2018). In the male mouse germline, HSP70-2 has a critical role in SC assembly during meiosis, possibly by inhibiting apoptosis of the spermatocyte (Dix et al., 1996; Chi et al., 2011). Moreover, HSP70-2 can interact with CDC2, a key factor in triggering the G2/M-phase transition during the mitotic and meiotic cell cycles, in the mouse testis (Zhu et al., 1997). HSP90α is required for meiotic progression, possibly acting by maintaining and activating meiotic regulators, such as the nuclear autoantigenic sperm protein required for cell cycle progression, and/or disassembling the SC that holds homologous chromosomes together in the mouse (Alekseev et al., 2003; Grad et al., 2010). In yeast, HSP90 also plays a critical role in spore maturation, possibly by interacting with Smk1, a mitogen-activated protein kinase required for spore wall assembly in yeast (Krisak et al., 1994; Zhao et al., 2005; Tapia and Morano, 2010). The regulators of HSP expression, HSFs, are expressed during mammalian spermatogenesis and could cooperate to maintain proper spermatogenesis by repackaging the DNA during spermatid differentiation (Widlak and Vydra, 2017). Similar to the pivotal role of HSPs in the reproduction of animals and yeast, our studies revealed that HSP101 has a vital function in the development of male gametes in maize.

As a molecular chaperone, HSP101 plays a critical role in plant growth under environmental stress (Queitsch, 2000; Hong and Vierling, 2001; Lin et al., 2014). Hsp101 is abundantly expressed in developing anthers under permissive temperature (Figure 6, A and B), without obvious induction under heat stress, which indicates a distinct expression pattern compared with common stress regulators (Nieto-Sotelo et al., 2002). Furthermore, we found that HSP101 predominantly accumulated in microspore mother cells during meiosis (Figure 6, C and D). These specific expression patterns likely suggest a unique role of the HSP101 in PMC development. Indeed, the consistent phenotype of disturbed RAD51 loading and DSB repair during meiosis in Ms42, as well as the similar but heat-stress-dependent phenotype of hsp101 knock out mutant (Figures 3C and 8C), points to HSP101 as a vital component of the regulatory module for meiotic development of PMC according to our cytological analysis, especially the DSB repair process (Figures 3C and 8, C and G). Furthermore, HSP101 might function as a disaggregator and elevated temperature makes recombination proteins more prone to misfolding and aggregation, which can further affect CO formation and distribution (Al-Whaibi, 2011; Bomblies et al., 2015; Morgan et al., 2017). Thus, we speculate that HSP101 may play an important role in preventing the abnormal aggregation of RAD51 and other proteins required for DSB repair during meiosis in maize. Further revealing the interacting proteins of HSP101 or exploring the stability of meiosis-related proteins through proteome may further elucidate its functional mechanism.

HSP101 is essential for thermotolerance during meiosis

Plant meiosis is a complex process jointly controlled by the coordinated effects of environmental factors and the genetic makeup of the cell (Kelliher and Walbot, 2012; Nelms and Walbot, 2019; Lee et al., 2021; Yang et al., 2022). Although HSP101 contributes to the DSB repair process during male meiosis (Figure 8C), no evident male fertility defects were observed in the hsp101 null plants at normal temperatures (NTs; Supplemental Figure S7; Nieto-Sotelo et al., 2002). In addition, the missense mutation in Ms42 triggers more severe chromosomal behavior defects during meiosis compared to hsp101 null plants under heat stress, including severely disrupted homologous pairing and SC assembly (Figures 1B and 3A). The less severe phenotype of the null mutant may be explained by the possibility that HSP101 is functionally redundant during microsporogenesis when other proteins complement the absence of HSP101 in PMCs of null plants, such as other HSPs expressed in meiosis under stress (Barnabas et al., 2008). In this scenario, the mutation in the ClpB domain of HSP101 in the dominant Ms42 may alter the spatial structure and consequently interfere with the biochemical properties of other close HSPs and further disrupt the polymerization of RAD51, Axis proteins, and SC proteins during meiosis (Supplemental Figure S6), resulting in a dominant phenotype in a gain-of-function manner even under permissive temperature (Figure 1). Alternatively, under optimal growth conditions HSP101 is stored as standby for action only in case of severe environmental stress (Nieto-Sotelo et al., 2002), in particular on RAD51 loading during male meiosis (Figure 8). As shown by the HSP101 research in Arabidopsis, a conserved Glu residue mutation in the second ATP-binding domain (hot1-1) or null hsp101 mutant allele (hot1-3) provoke a loss of function defective phenotype only under heat stress (Hong and Vierling, 2000, 2001).

These two possibilities may be both true and indistinguishable; ultimately, maize HSP101 may be an illustration of the complex mechanisms underlying environment adaption (Zinn et al., 2010). A well-described model of typical heat response is based on the fact that HSFs activate HSP gene transcription (Kim and Schoffl, 2002; Ohama et al., 2017). However, the molecular details of HSP function and the HSF network remain elusive in terms of spatial-temporal gene expression, interactions, and the large size of HSF/HSP families in plants (Kotak et al., 2007; Lin et al., 2011; Li et al., 2020). Consider HSP101 as an example: Silencing lines in Arabidopsis, rice, and maize have severely diminished thermotolerance during germination and vegetative development, but HSP101 chaperone activity is dispensable for normal development in the absence of heat stress (Supplemental Figure S7; Queitsch, 2000; Hong and Vierling, 2001; Nieto-Sotelo et al., 2002; Lin et al., 2014). Despite sustained interest in the role of Hsp101, investigations have only focused on the vegetative organs of plants. Whether HSP101 is involved in temperature-induced adaptation of reproductive development remains unclear. Our analysis of maize Hsp101-KO/OE plants confirms that HSP101 positively mediates heat tolerance during microsporogenesis (Figures 7 and 9). Moreover, our results confirm that low levels of HSP101 in roots, stems, leaves, and other vegetative tissues under normal conditions (Figure 6A), though it is expression can be induced by heat shock via HSFs (Nieto-Sotelo et al., 2002). In contrast, expression of Hsp101 in reproductive tissue is not inducible by temperature (Nieto-Sotelo et al., 2002). Such differences in inducibility by heat shock between different tissues suggest that other regulators expressed in meiosis may predominantly underlie the meiotic thermotolerance pathway compared to vegetative heat response, not just the canonical HSF–HSP network.

HSP101 homologs show highly conserved sequences in broad monocots and eudicots (Lee et al., 1994; Schirmer et al., 1994; Young et al., 2001; Supplemental Figure S5), which suggests a possible conserved function in fundamental thermotolerance in plant meiosis. More important, given that the overexpression of Hsp101 in maize anthers significantly improves pollen viability under heat stress (Figure 9), similar engineering of the expression of HSP101 during microsporogenesis may be a promising strategy for enhancing heat tolerance during reproduction in other crops, which is of great significance for breeding heat-resistant crop cultivars in the context of global warming.

Materials and methods

Plant materials and growth conditions

The maize (Z.  mays) Ms42 mutant line (502E) was obtained from the Maize Genetics Cooperation Stock Center (http://maizecoop.cropsci.uiuc.edu). All maize materials were grown at the experimental stations of the China Agricultural University in Beijing (the temperature range is 21°C–32°C when the plant is in the meiosis stage). For heat stress treatment, meiosis-staged plants were transferred to an incubator with a 14-h day/10-h night cycle and kept for 14 h at 32°C, 14 h at 35°C, or 24 h at 35°C, respectively. The meiotic tassel was fixed immediately after treatment ended.

Preparation of meiotic chromosomes

Young tassels from WT and Ms42 plants were fixed in Carnoy’s solution (ethanol: glacial acetic acid, 3:1, v/v). Meiotic anthers were crushed with forceps in 45% acetocarmine solution on glass slides and then covered with a coverslip. After staging with a light microscope, slides were frozen in liquid nitrogen, then the coverslip was removed rapidly with a razor. Chromosomes were stained with DAPI, and images were captured with an Olympus BX61 epifluorescence microscope equipped with a CCD camera.

FISH analysis

FISH was performed as described previously (Huang et al., 2016). The pTa794 clone containing 45S or 5S rDNA repeats, and the pAtT4 clone containing telomere-specific repeats were used as FISH probes. rDNA and telomere probes were labeled with a nick translation kit (Roche, Basel, Switzerland). Chromosomes were counterstained with anti-fade mounting medium with DAPI. Chromosome images were captured under an Olympus BX61 fluorescence microscope.

Antibody preparation and immunoblotting

The full-length coding sequence (CDS) of Hsp101 and a 978-bp fragment of Afd1 (encodes amino acids 137–462 of AFD1 (NCBI ID: NP_001105829.1) were cloned into the NcoI and XhoI sites of pET-28a-His (Novagen, Madison, WI, USA) using the In-Fusion HD Cloning System (Takara, Shiga, Japan). The His-tagged HSP101 and the AFD1 fusion protein were expressed in Escherichia coli strain BL21(DE3) and purified with the manufacturer’s instructions of the Beaver Beads His-tag Protein Purification kit (catalog no. 70501-K10; BEAVER, San Francisco, CA, USA). Subsequent production of HSP101 and AFD1 antibodies with rabbits was performed by Wuhan ABclonal Company according to standard protocol. The whole protein of various tissues was extracted according to the previously described protocol (Jiang et al., 2019). Immunoblot containing whole protein extracts from plants was performed with 1:8,000 dilutions of HSP101 antibody and 1:3,000 dilutions of ACTIN antibody (catalog no. BE7008; EASYBIO). The goat anti-rabbit IgG conjugated to horseradish peroxidase (1:5,000 dilution; catalog no: BE0101; EASYBIO, Kangwon-do, South Korea) was used as the secondary antibody.

Fluorescence immunolocalization assay

Immunolocalization assay was performed as described previously (Pawlowski et al., 2003). Fresh young anthers were dissected from immature tassel and fixed in 1× Buffer A containing 4% (w/v) paraformaldehyde for 30 min at room temperature (25°C), then washed in 1× buffer A at room temperature twice and stored in 1× buffer A at 4°C. Anthers at the proper stages were crushed with forceps in 1× buffer A on glass slides, then squashed and frozen in liquid nitrogen. The coverslip was rapidly removed. Subsequently, the slides were incubated with different antibodies diluted in blocking buffer in a humidity chamber at 4°C overnight and washed three times in 1× PBS. Alexa Fluor 555-conjugated goat anti-rabbit or/and FITC-conjugated goat anti-mouse antibodies (1:100 dilution in buffer A) were added to slides at 37°C for 1 h. After being washed three times in 1× PBS, the slides were counterstained with DAPI in antifade solution. Chromosome images were captured under an Olympus BX61 fluorescence microscope. Polyclonal antibodies against ASY1 were prepared by specific sequence fusion-peptide immunization in rabbit. Polyclonal antibodies against ZYP1 and γH2AX were prepared by specific sequence fusion-peptide immunization in mouse. ASY1, ZEP1, and RAD51 antibodies were gifts from Yan He (China Agricultural University, Beijing).

Mapping-based cloning of Ms42

Ms42 was crossed with inbred lines B73 and Zheng58, and the BC₁ F₁ individuals of two genetic backgrounds were used for the mapping population. We mapped the Ms42 locus to the long arm of maize chromosome 6 (6L) using insertion-deletion and simple sequence repeat markers distributed throughout the whole genome. For fine mapping, several insertion-deletion and SNP markers were designed with Oligo7 (Supplemental Table S1). Both fertile and sterile individuals of Ms42 × Zheng58 and Ms42 × B73 BC₁F₁ were used for fine mapping.

Phylogenetic analysis

Protein sequences of putative orthologs of HSP101 were obtained from Gramene (http://www.gramene.org) and then aligned with ClustalW using default parameters. Evolutionary analysis was performed with the neighbor-joining method based on the Poisson model by MEGA version 6 (http://www. megasoftware.net/) using the bootstrap method with 1,000 replications. The alignment and tree are provided as Supplemental Files S1 and S2.

RT-qPCR assays

Total RNA was isolated using TRIzol reagent (Invitrogen, Waltham, MA, USA) from maize tissues (roots, stems, leaves, tassels, seeds of 3, 6, and 9 days after pollination, and anthers of different developmental stages). Total RNA (1.0 μg) was treated with DNase I (Promega, Madison, WI, USA) to eliminate genomic DNA contamination, and RT reactions were conducted using the PrimeScript RT reagent Kit (Takara) according to the manufacturer’s instruction. The RT product (1.5 µL) was used as the template for PCR. RT-qPCR was performed using the ABI 7500 real-time PCR system with SYBR Green PCR Master Mix (Takara) in a final volume of 15 µL. ZmActin (Zm00001d010159) was used as the internal control (Dong et al., 2013). RT-qPCR was performed in three biological and technical replicates. Relative expression was calculated using the 2^−ΔΔCT method (Livak and Schmittgen, 2001). The expression level of Ms42 in the anther of less than 1.5 mm was set to 1. All primers used for RT-qPCR are listed in Supplemental Table S1.

Genetic transformation of maize

For the CRISPR/Cas9 gene-editing construct of Ms42, the target sequence for CRISPR/Cas9 editing was designed across the junction of the first exon with the NGG protospacer adjacent motif. We obtained the target sequences by annealing the primer sets listed in Supplemental Table S1 and cloning them into the BsaI sites of pBUE411 (Xing et al., 2014). All plasmids were introduced into the receptor line LH244 via Agrobacterium tumefaciens-mediated transformation, according to the standard protocol of the Center for Crop Functional Genomics and Molecular Breeding of China Agricultural University. Gene editing results were determined by Sanger sequencing. In addition, T₁ plants harboring Cas9-gRNA were crossed with the Ms42 dominant mutants. Three homozygous lines containing both the edited target sites and the SNP mutation (referred to as Ms42-De2, Ms42-In1, and Ms42-In2) were identified from the F₁ progenies. The full-length CDS (without the stop codon) of Hsp101 was amplified and cloned into the AscI and KpnI sites of the binary vector pBCXUN-MYC to generate the overexpression construct driven by the Ubiquitin promoter. The primers used for construct generation and transgenic plant identification are listed in Supplemental Table S1.

Immunohistochemical staining

WT anthers were sampled at stages S7–S9 for immunolocalization with a 1:500 dilution of HSP101 antibody. Immunolocalization was performed as described previously (Sessions, 2008), except for epitope retrieval, which involved immersing rehydrated paraffin sections in boiled 10 mM sodium citrate buffer (pH 6.0) for 10 min. Preimmunization rabbit serum was used as a control.

Accession numbers

Sequence data from this article can be found in the GenBank/EMBL data libraries under the following accession numbers: Ms42/Hsp101(Zm00001d038306) and Zmactin (Zm00001d010159). Accession numbers for the sequences used in the phylogenetic analysis are listed on the tree.

Supplemental data

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

Supplemental Figure S1. Meiotic chromosome behaviors of male meiocytes in Ms42.

Supplemental Figure S2. Immunolocalization of ASY1 on meiotic chromosomes in the WT and Ms42.

Supplemental Figure S3. Immunolocalization of proteins related to the SC and DSBs on meiotic chromosomes in the WT and Ms42.

Supplemental Figure S4. Phylogenetic analysis of MS42-related proteins.

Supplemental Figure S5. The two substituted amino acids of HSP101 are highly conserved in maize inbred lines and homologous proteins from different species.

Supplemental Figure S6. Change in the charge properties of amino acid residues in Ms42.

Supplemental Figure S7. Phenotypic analysis of hsp101 null mutants.

Supplemental Figure S8. Immunofluorescent localization of HSP101 in Ms42.

Supplemental Figure S9. Heat stress decreases pollen fertility in Hsp101-De16 plants.

Supplemental Figure S10. MHS induces abnormal meiosis in Hsp101-In1 plants.

Supplemental Figure S11. Immunolocalization analysis of AFD1 and ZYP1 in the microspores of WT and Hsp101-In1 under HHS conditions.

Supplemental Figure S12. Hsp101-OEs have normal vegetative growth under normal growth conditions.

Supplemental Figure S13. The heat tolerance of microsporogenesis has enhanced in Hsp101-OE2 plants.

Supplemental Table S1. Primers used in this study.

Supplemental Data Set S1. Data for statistical analyses.

Supplemental File S1. HSP101 sequence alignments.

Supplemental File S2. HSP101 phylogenetic tree.

 

W.J., Y.L., and W.H. designed the project. Y.L., H.S., T.W., L.P., and X.Z. performed the experiments. Y.L., Y.H., and W.R. analyzed the data and generated graphs. Y.L., W.J., W.H., and Z.D. wrote the manuscript.

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 (https://academic.oup.com/plcell) is: Weiwei Jin (weiweijin@cau.edu.cn).