NUP1/NUP136 is essential for obligatory crossover formation during meiosis in Arabidopsis

TL;DR

This study shows that the nuclear pore complex protein NUP1/NUP136 is crucial for proper chromosome pairing and crossover formation during meiosis in Arabidopsis.

Contribution

The study identifies a novel role for the plant-specific nucleoporin NUP1/NUP136 in ensuring obligatory crossover formation during meiosis.

Findings

Loss of NUP136 reduces chiasma frequency and causes chromosome interlocks during meiosis.

NUP136 is required for proper homologous chromosome pairing and crossover formation.

NUP136 affects the spatial distribution of centromeres, telomeres, and NORs during meiotic prophase I.

Abstract

The nuclear pore complex (NPC) is a major component of the nuclear envelope (NE), which mediates nucleocytoplasmic transport and is involved in a variety of transport-independent processes, including genome organization and cell division. In plants, several NPC subunits are species specific, and their roles in meiosis remain poorly understood. Here, we characterize the function of the plant-specific nuclear basket nucleoporin NUP1/NUP136 during meiosis in Arabidopsis thaliana. Loss of NUP136 leads to a marked reduction in chiasma frequency, resulting in univalents, and the persistence of chromosome interlocks at metaphase I. This phenotype is consistent with defects in early chromosome interactions and crossover (CO) formation, as evidenced by a reduced number of MLH1 foci. In the mutant, there is also an altered spatial distribution of centromeres, telomeres, and nucleolar organizer regions (NORs), pointing to changes in the dynamics of these chromosomal domains during meiotic prophase I. Meiotic defects in nup136-2 mutants are modestly aggravated by the loss of the related paralog NUP82. Our results demonstrate that NUP1/NUP136 is important for proper homologous chromosome pairing and for ensuring the formation of the obligatory CO, likely by contributing to NE organization and facilitating the chromosomal contacts that support recombination during prophase I.

Graphical AbstractFor image description, please refer to the figure legend and surrounding text.

Figures

Meiotic division in nuclear basket Nup mutants. PMCs from WT, nup82, and nup136 mutant lines. Chromosomes were visualized with DAPI. nup82-1 and nup82-2 mutants show normal meiotic progression, including full synapsis, formation of 5 bivalents at metaphase I, and accurate chromosome segregation through both meiotic divisions, comparable to WT. In contrast, nup136-1 and nup136-2 mutants exhibit univalents at metaphase I (arrowheads), chromatin bridges during telophase I, mis-segregated chromosomes (arrowheads), and micronuclei at the tetrad stage. Notably, nup136-2 also displays persistent chromosomal interlocks from pachynema to metaphase I (arrows). The NUP136:GFP line shows a WT-like phenotype. Bars = 5 μm.

Chiasma frequency and distribution in nup136-2. Chromosomes were identified by FISH using 5S (magenta; chromosomes 3, 4 and 5) and 45S (green; chromosomes 2 and 4) rDNA probes and counterstained with DAPI (gray). a) Representative metaphase I cells in WT (n = 69) and nup136-2 (n = 41). Chromosomes are numbered based on rDNA labeling. b) Examples of nup136-2 metaphase I cells showing bivalent magnifications and schematic interpretations of chromosomal interlocks. c) Chiasma frequency per cell in WT (gray, left) and nup136-2 (blue, right). P < 0.0001 (2-tailed Mann–Whitney U test). d) Proportions of univalents and bivalents in WT (gray, left) and nup136-2 (blue, right). e) Observed (solid lines) and Poisson-predicted (dashed lines) distributions of chiasma numbers per cell in WT (gray) and nup136-2 (blue). Bars = 5 μm.

Analysis of meiotic recombination proteins in WT and nup136-2 during prophase I. a) Zygotene nuclei immunolabeled with ASY1 (green) and RAD51 (magenta). b) Scatter plot of RAD51 foci in WT (n = 24) and nup136-2 (n = 20). c) Zygotene nuclei labeled with ASY1 (green) and DMC1 (magenta). d) Scatter plot of DMC1 foci in WT (n = 17) and nup136-2 (n = 32). e) Pachytene nuclei labeled with ZYP1 (green) and MLH1 (magenta). f) Scatter plot of MLH1 foci in WT (n = 22) and nup136-2 (n = 19). In scatter plots, black horizontal lines indicate mean values. *P < 0.05; P < 0.0001 (2-tailed Mann–Whitney U test). Bars = 5 μm.

nup136-2 displays altered distribution of CEN, TEL, and NORs during prophase I. FISH analysis using CEN (green; bigger signals) and TEL (magenta; smaller signals) probes, counterstained with DAPI (gray), in WT and nup136-2 nuclei. Representative nuclei from different stages of prophase I are shown. DAPI-free images are shown to enhance visualization of specific signals. The white dotted line outlines the nucleolus. In nup136-2 early leptonema and leptonema, 2 associated NORs are highlighted (blue line), and arrows indicate paired CENs, suggesting a slightly advanced association of these regions compared to WT. In contrast, TEL pairing appears delayed, as judged by the higher number of distinct TEL signals persisting into late leptonema. In nup136-2 zygonema and pachynema, magnified views and schematic interpretations of complex chromosomal interlocks are shown. Bars = 5 μm.

Association of CEN, TEL, and NORs in WT and nup136-2 during prophase I. Graphs represent the dynamics of CENs (green), TELs (magenta), and NORs (blue) from G2 to pachynema (P). The left y axis indicates the number of TEL and CEN signals (associated signals), while the right y axis shows the percentage of NORs associated. Top, WT; bottom, nup136-2. In WT, TEL signals gradually decrease during early prophase I, indicating progressive TEL clustering, while CEN and NOR associations increase more slowly. In contrast, nup136-2 nuclei show delayed TEL pairing and a precocious association of CENs and NORs, suggesting temporal uncoupling in nuclear reorganization processes.

Immunolocalization of SC, cohesion, and NE proteins in WT and nup136-2. a) Pachytene nuclei showing ASY1 (green) and ZYP1 (magenta) localization in WT and nup136-2. Synapsis is not full in the mutant. b) Zygotene nuclei showing ASY1 (green) and SYN1 (magenta) distribution in WT and nup136-2. c) Zygotene nuclei stained with DAPI (gray) showing Cter-SUN-domain proteins (magenta) in WT and nup136-2. d) Zygotene nuclei of WT as negative control and nup136-2 complemented with NUP136:GFP showing GFP signal (green) and DAPI (gray). Bars = 5 µm.

Cytological characterization of nup82-1 nup136-2 PMCs. a) Meiotic progression in the nup82-1 nup136-2 double mutant. Chromosomes were visualized with DAPI. Pachynema with an interlock (arrow); metaphase I with a pair of univalents (arrowheads) and an interlock between 2 bivalents (arrow); telophase I showing a chromatin bridge and a fragment (arrow); metaphase II with unbalanced segregation (6:4); and a tetrad with nuclei of unequal size. b) Representative metaphase I spreads from nup136-2 and nup82-1 nup136-2. Individual chromosomes or bivalents are numbered. nup82-1 images have been omitted for simplification, since they were indistinguishable from WT. c) Proportion of univalent and bivalents in nup136-2 (n = 41, blue, left) and nup82-1 nup136-2 (n = 79, orange, right). Bars = 5 µm.

Tables

Table 1.. Mean chiasma frequencies per cell, per bivalent, and per bivalent arm in WT and nup136-2.

	Chromosome									Total + SEM	n
	1	2		3		4		5
	S + La	S	L	S	L	S	L	S	L
WT	2.52	0.61	1.14	0.90	1.26	0.48	1.01	0.97	1.30	10.2 ± 0.14	69
		1.75		2.16		1.49		2.28
nup136-2	1.20	0.17	0.88	0.46	0.98	0.20	0.90	0.51	0.78	6.07 ± 0.22	41
		1.05		1.44		1.10		1.29
Decrease chiasma frequency (%)	52.61	71.95	23.31	48.43	22.62	59.20	11.05	47.25	40.16	40.48
Univalent percentage	17.07	12.20		4.88		9.76		19.51		51.22

Table 2.. Quantification of RAD51, DMC1 and MLH1 foci in WT and nup136-2.

	WT	nup136-2	P
RAD51	165.58 ± 4.74	151.65 ± 3.82	0.0286
	(107 to 204)	(123 to 184)
	n = 24	n = 20
DMC1	151.71 ± 3.96	138.66 ± 4.04	0.0231
	(123 to 176)	(92 to 192)
	n = 17	n = 32
MLH1	7.18 ± 0.38	4.58 ± 0.43	<0.0001
	(4 to 11)	(1 to 8)
	n = 22	n = 19

Table 3.. Mean chiasma frequencies per cell, per bivalent, and per bivalent arm in WT and nup82-1 nup136-2 double mutant.

	Chromosome									Total ± SEM	n
	1	2		3		4		5
	S + La	S	L	S	L	S	L	S	L
WT	2.52	0.61	1.14	0.90	1.26	0.48	1.01	0.97	1.30	10.2 ± 0.14	69
		1.75		2.16		1.49		2.28
nup136-2	1.20	0.17	0.88	0.46	0.98	0.20	0.90	0.51	0.78	6.07 ± 0.22	41
		1.05		1.44		1.10		1.29
nup82-1 nup136-2	1.18	0.12	0.80	0.34	0.78	0.32	0.76	0.54	0.93	5.63 ± 0.18	79
		0.89		1.14		0.97		1.46
Decrease in chiasma frequency (%)	53.32	79.97	29.70	62.00	38.10	33.70	25.47	44.74	28.94	44.79
Univalent percentage	26.58	24.05		18.99		24.05		12.66		74.68

Full text

Introduction

The nuclear envelope (NE) is a double lipid bilayer that separates the nucleus from the cytoplasm and is an essential component of all eukaryotic cells. The NE is composed of an inner and outer nuclear membranes (INM and ONM, respectively), which are separated by the perinuclear space, where the nuclear pore complexes (NPCs) are embedded (Fernández-Jiménez and Pradillo 2020). NPCs are versatile channels that mediate the active bidirectional transport of larger molecules between the nucleus and the cytoplasm. They have an octagonal and cylindrical symmetric structure comprising a complex network of ∼30 nucleoporins (Nups) present in various stoichiometries (Alber et al. 2007). Beyond transport, NPCs provide a binding platform for the underlying chromatin and play roles in RNA processing, gene expression regulation, DNA repair, and mitotic chromosome segregation, among other functions (Strambio-De-Castillia et al. 2010; Beck and Hurt 2017; Raices and D’Angelo 2017; Simon et al. 2024). The study of NPC functions in plants has become a rapidly expanding field, with several studies demonstrating the involvement of Nups in processes such as hormonal signaling and response to biotic and abiotic stresses, including temperature, salinity, and pathogen defense (Gu 2018; Margalha et al. 2023; Nie et al. 2023; Xie et al. 2024; Yang et al. 2024).

In addition to NPCs, other components contribute to the landscape of the NE. The linker of nucleoskeleton and cytoskeleton (LINC) complexes have been characterized in several organisms (Malone et al. 1999; Murphy et al. 2010; Zhou and Meier 2013). These protein complexes connect chromatin to cytoskeletal components and consist of the conserved eukaryotic SUN and KASH domain proteins (Starr and Han 2002; Padmakumar et al. 2005; Crisp et al. 2006; McGee et al. 2006). The LINC complex also facilitates the association of chromosome ends with the NE, enabling telomere (TEL)-led movements that are essential for promoting homologous chromosome interactions during meiosis (Murphy et al. 2014; Varas et al. 2015; Zhang et al. 2020). Compared to the well-characterized NE complexes in animals and yeast, the understanding of these proteins in plants is still limited. However, recent proteomic approaches have led to the identification of several plant NPC components, some of which are plant specific and may functionally replace vertebrate Nups (Tamura et al. 2010; Tamura and Hara-Nishimura 2013; Meier et al. 2017; Tang et al. 2020). For example, NUP1, also known as NUP136, is a plant-specific Nup that likely serves as a functional homolog of yeast Nup1 and vertebrates Nup153 (Neumann et al. 2006), both components of the nuclear basket subcomplex. In mammals, Nup153 interacts directly with SUN proteins, linking the NPC to the LINC complex. This interaction has been proposed to regulate NPC distribution along the NE (Liu et al. 2007; Lu et al. 2008; Talamas and Hetzer 2011; Jahed et al. 2016). In plants, NUP1 functions as a soluble, highly mobile protein, exhibiting dynamic behavior as it continuously assembles and disassembles from the NPC structure (Tamura et al. 2010). Loss of NUP1 results in alterations in nuclear morphology, reduced fertility, and early flowering (Tamura et al. 2010; Bao et al. 2019). Additionally, NUP82, another plant-specific Nup, interacts directly with NUP1 in the nuclear basket, and both Nups share functions in immune response and plant growth (Tamura et al. 2017).

During meiosis, NPCs have been proposed to act as “press-studs” that facilitate chromosome attachment to the NE (Engelhardt and Pusa 1972; La Cour and Wells 1974). In Arabidopsis, TELs are associated with the nucleolus from meiotic S-phase to early leptonema, when they begin to pair (Armstrong et al. 2001). The homology search then starts mediated by the attachment of TELs to the NE and their connection to the cytoskeleton. These connections lead to the formation of the bouquet, a clustered arrangement of TELs, at the onset of zygonema, which facilitates homologous chromosome pairing and synapsis (Gerton and Hawley 2005; Zickler and Kleckner 2015). Full synapsis is typically achieved by pachynema, completing the establishment of homologous chromosome interactions. At this stage, TELs are widely distributed throughout the nucleus. Studies in human spermatocytes have shown that TEL repositioning during meiotic prophase I occurs in regions of the NE distinct from those with a high NPC density, suggesting the presence of specialized functional NE domains that facilitate the complex choreography of meiotic processes (Scherthan et al. 2000). In budding yeast, nuclear basket Nups are essential for structural remodeling during meiotic progression (King et al. 2023). In fission yeast, Nup132 deficiency compromises nuclear integrity during meiosis I, disrupting SUMOylation processes (Yang et al. 2023). In Drosophila male meiosis, the absence of Nup107 leads to cytokinesis failure (Hayashi et al. 2016).

Recently, it has been reported that prophase I in Arabidopsis thaliana involves rapid prophase chromosome movements (RPMs), which are linked to significant reorganization of the NE. These movements are facilitated by interactions between chromosomes and the cytoskeleton, mediated by SUN-domain proteins at the inner NE (Cromer et al. 2024 ). While the role of NPCs in this process remains unclear, they may contribute to NE reorganization and chromosome dynamics during meiosis in plants. Supporting this idea, the scaffold Nups SAR1 (NUP160) and SAR3 (NUP96) are essential for proper meiotic progression (Fernández-Jiménez et al. 2023). To further explore the involvement of NPCs in plant meiosis, in this study, we focused on one of their less characterized subcomplexes: the nuclear basket. This structure is thought to participate in interactions with chromatin or chromatin-associated proteins, potentially influencing chromosomal dynamics during meiosis. Here, we have characterized the meiotic consequences of NUP1/NUP136 deficiency and have demonstrated that this protein is required to ensure the formation of the obligatory chiasma and, consequently, to produce balanced gametes. Our findings provide new insights into the role of the nuclear basket in plant meiosis.

Materials and methods

Plant material

Plants were grown, materials harvested, and DNA extracted as previously described by Pradillo et al. (2012). Arabidopsis thaliana accession Columbia (Col-0) was used as the wild-type (WT) control. Mutant seeds were obtained from the Nottingham Arabidopsis Stock Center (Nottingham, UK; https://www.arabidopsis.info), including T-DNA alleles from the genes NUP82 (AT5G20200) nup82-1 (SALK_001707), nup82-2 (SALK_024526), NUP136 (AT3G10650) nup136-1 (SALK_104728), and nup136-2 (SAIL_796_H02). The NUP136:GFP line was kindly provided by Dr. Chang Liu (Department of Epigenetics, University of Hohenheim) (Bi et al. 2017). To generate the double mutant, nup136-2 was crossed with nup82-1, and the F2 progeny were genotyped to identify plants homozygous for both mutations. Seeds were sown directly on soil, and plants were grown under long-day conditions (16 h light/8 h dark) at 19 °C. Homozygous plants were identified by PCR screening using primers listed in Supplementary Table 4.

Cytological analyses

Fixation, pollen mother cell (PMC) slide preparation, and FISH were performed as previously described (Sanchez-Moran et al. 2001). The DNA probes used were 45S rDNA (pTa71) (Gerlach and Bedbrook 1979), 5S rDNA (pCT4.2) (Campell et al. 1992), centromeres (CENs) (pAL1) (Martinez-Zapater et al. 1986), and TELs (pLT11) (Richards and Ausubel 1988). At least 3 plants of each genotype were used in all cytological analyses.

Immunolocalization was conducted using fresh young anthers. A chromosome spreading technique was used for the detection of homologous recombination-related proteins (Armstrong et al. 2009). The following primary antibodies were kindly provided by Prof. Chris Franklin (School of Biosciences, University of Birmingham): anti-ASY1 (rat; 1:1,000), anti-SYN1 (rabbit; 1:500), anti-ZYP1 (rat or rabbit; 1:500), anti-RAD51 (rabbit; 1:250), anti-DMC1 (rabbit; 1:250), and anti-MLH1 (rabbit; 1:250) (Mercier et al. 2003; Higgins et al. 2005; Jackson et al. 2006; Sanchez-Moran et al. 2007). Detection of NE proteins in PMCs was carried out using the squash technique as previously described (Oliver et al. 2013). Anti-SUN (rabbit; 1:200, Agrisera) was used for SUN-domain proteins, and anti-GFP (rat; 1:200, BioLegend) for detection of NUP136:GFP. Secondary antibodies used were FITC-conjugated anti-rat IgG (1:50, Agrisera) and Alexa Fluor 555-conjugated anti-rabbit IgG (1:500, Molecular Probes). We scored all images blind to genotype.

Cells were observed using an Olympus BX-61 microscope equipped with an Olympus DP71 digital camera and analySIS software (Soft Imaging System). Images were processed and analyzed with ImageJ v1.52p and Adobe Photoshop CC v19.

Statistical analysis

Standard errors are reported throughout. Fisher's exact test (2-tailed) was used for categorical data, and the Mann–Whitney U test (2-tailed) to compare 2 independent groups. Chi-squared (χ^2^) goodness-of-fit test was used to evaluate deviations from Poisson distributions. All statistical analyses were performed using GraphPad Prism v8.

Results

NUP1/NUP136 is essential for proper meiotic progression in A. thaliana

We carried out cytological analysis of PMCs from T-DNA insertion mutants in genes encoding nuclear basket Nups. Specifically, we analyzed mutants corresponding to 2 alleles of NUP82 (nup82-1 and nup82-2) and 2 alleles of NUP1/NUP136 (nup136-1 and nup136-2). While nup82 mutants did not exhibit any detectable meiotic defects and showed normal chromosome pairing, complete synapsis, 5 bivalents at metaphase I, and regular chromosome segregation through both meiotic divisions, mutations in NUP1/NUP136 led to severe and specific meiotic alterations. Both nup136 mutants exhibited univalents at metaphase I, chromosome mis-segregation during anaphase I, and the formation of unbalanced tetrads, although complete synapsis was observed in some cells (Fig. 1). Notably, the nup136-2 mutant showed particularly drastic phenotypes, including persistent interlocks between nonhomologous chromosomes during prophase I, which remained visible in 24.39% of metaphase I cells (n = 41) (Fig. 2b). To confirm the causal relationship between the observed phenotypes and NUP1/NUP136 loss of function, we performed a complementation assay using a NUP136:GFP construct (Bi et al. 2017). This transgene restored the WT meiotic phenotype (Fig. 1), confirming that NUP1/NUP136 is essential for normal meiotic progression and validating the full functionality of the line NUP136:GFP. Based on this evidence, we selected nup136-2 for further in-depth characterization.

Meiotic division in nuclear basket Nup mutants. PMCs from WT, nup82, and nup136 mutant lines. Chromosomes were visualized with DAPI. nup82-1 and nup82-2 mutants show normal meiotic progression, including full synapsis, formation of 5 bivalents at metaphase I, and accurate chromosome segregation through both meiotic divisions, comparable to WT. In contrast, nup136-1 and nup136-2 mutants exhibit univalents at metaphase I (arrowheads), chromatin bridges during telophase I, mis-segregated chromosomes (arrowheads), and micronuclei at the tetrad stage. Notably, nup136-2 also displays persistent chromosomal interlocks from pachynema to metaphase I (arrows). The NUP136:GFP line shows a WT-like phenotype. Bars = 5 μm.

*Chiasma frequency and distribution in nup136-2. Chromosomes were identified by FISH using 5S (magenta; chromosomes 3, 4 and 5) and 45S (green; chromosomes 2 and 4) rDNA probes and counterstained with DAPI (gray). a) Representative metaphase I cells in WT (n = 69) and nup136-2 (n = 41). Chromosomes are numbered based on rDNA labeling. b) Examples of nup136-2 metaphase I cells showing bivalent magnifications and schematic interpretations of chromosomal interlocks. c) Chiasma frequency per cell in WT (gray, left) and nup136-2 (blue, right). ***P < 0.0001 (2-tailed Mann–Whitney U test). d) Proportions of univalents and bivalents in WT (gray, left) and nup136-2 (blue, right). e) Observed (solid lines) and Poisson-predicted (dashed lines) distributions of chiasma numbers per cell in WT (gray) and nup136-2 (blue). Bars = 5 μm.

Chiasma formation is disrupted in nup136

In Arabidopsis thaliana, the combination of 45S and 5S rDNA probes with chromosome morphology allows the identification of each chromosome and chromosome arm. Using this approach, bivalent configurations and chiasmata could be analyzed. A rod bivalent is defined as a bivalent in which chiasmata are present in only 1 chromosome arm, whereas a ring bivalent is defined as a bivalent in which chiasmata occur in both arms (Fig. 2a). In the WT (Col-0, the genetic background of nup136-2), the mean chiasma frequency per cell was 10.20 ± 0.14 (n = 69), ranging from 8 to 13 chiasmata per cell. In contrast, nup136-2 mutants showed a significant reduction in chiasma number (6.07 ± 0.22, n = 41; P < 0.0001), with a broader range of 3 to 10 chiasmata per cell (Fig. 2c). This reduction in chiasma frequency affected all chromosomes except the long arm of chromosome 4 (Table 1). Moreover, nup136-2 exhibited a significant increase in rod bivalents compared to the WT, where ring bivalents predominate (P < 0.0001) (Fig. 2d; Supplementary Table 1). This reduction in crossover (CO) formation results in the loss of CO assurance, leading to the appearance of univalents in 53.66% of metaphase I cells (n = 41). Further analysis revealed that 12.68% of homologous chromosome pairs in nup136-2 failed to form a CO, with chromosomes 1 and 5 being the most affected (Fig. 2d; Table 1). Chiasma distribution among cells in the WT deviates significantly from a random Poisson distribution (P < 0.0001) (Fig. 2e), consistent with the existence of CO interference (Li et al. 2021). In nup136-2, however, chiasma counts do not significantly deviate from a Poisson distribution (P = 0.447), indicating that at least a subset of COs may be randomly distributed. Because class I COs are normally subject to interference, this shift toward a Poisson-like distribution suggests a reduction or loss of positive interference. As described above, we also observed a large variation in the number of chiasmata per cell and changes in their distribution among bivalents, with longer chromosomes showing a stronger reduction in chiasma number. Together, these observations are consistent with a disruption of interference in nup136-2, allowing COs to occur more randomly across the genome.

NUP1/NUP136 deficiency alters recombination protein landscape

To further investigate the potential causes underlying the reduced chiasma frequency observed in nup136-2, we performed immunolocalization of key proteins involved in different stages of the meiotic recombination pathway (Fig. 3). As an indirect measure of programmed DNA double-strand breaks (DSBs), we quantified RAD51 foci and observed a slight but statistically significant reduction in nup136-2 (151.70 ± 3.82, n = 20) compared to WT (165.60 ± 4.74, n = 24; P = 0.029) (Fig. 3a and b; Table 2) during zygotene stage. A similar trend was detected for DMC1, which is required for interhomolog recombination, with nup136-2 (138.70 ± 4.04, n = 32) showing fewer foci than the WT (151.70 ± 3.96, n = 17; P = 0.0231) (Fig. 3c and d; Table 2). Most notably, a significant decrease in MLH1 foci was observed in nup136-2 (4.58 ± 0.43, n = 19) compared to WT (7.18 ± 0.38, n = 22; P < 0.0001) (Fig. 3e and f; Table 2). MLH1 marks a subset of COs that are sensitive to positive interference (Jackson et al. 2006), and its reduction in nup136-2 (36.24% decrease relative to WT) was proportionally more pronounced than that of RAD51 (8.39%) or DMC1 (8.57%). This preferential loss of MLH1-marked COs is consistent with the decrease in chiasma number and the altered chiasma distribution in nup136-2, supporting the conclusion that interference-sensitive COs are particularly affected.

*Analysis of meiotic recombination proteins in WT and nup136-2 during prophase I. a) Zygotene nuclei immunolabeled with ASY1 (green) and RAD51 (magenta). b) Scatter plot of RAD51 foci in WT (n = 24) and nup136-2 (n = 20). c) Zygotene nuclei labeled with ASY1 (green) and DMC1 (magenta). d) Scatter plot of DMC1 foci in WT (n = 17) and nup136-2 (n = 32). e) Pachytene nuclei labeled with ZYP1 (green) and MLH1 (magenta). f) Scatter plot of MLH1 foci in WT (n = 22) and nup136-2 (n = 19). In scatter plots, black horizontal lines indicate mean values. *P < 0.05; ***P < 0.0001 (2-tailed Mann–Whitney U test). Bars = 5 μm.

nup136-2 shows alterations in the distribution of CENs, TELs, and nucleolar organizer regions during prophase I

Given that the phenotype of nup136-2 at metaphase I and later stages resembles that observed in Cter-SUN-domain protein-deficient plants (Varas et al. 2015 ), and considering that sun1 sun2 PMCs display disrupted spatial organization of CENs, TELs, and 45S rDNA loci (nucleolar organizer regions, NORs), we investigated whether nup136-2 PMCs display comparable alterations from premeiotic interphase (G2) to pachynema (Figs. 4 and 5; Supplementary Table 2). At G2, both WT and nup136-2 PMCs showed 20 individual TEL signals distributed around the periphery of the nucleolus. During leptonema, WT cells showed a gradual reduction in TEL signals as homologous chromosomes began paired via their telomeric ends. In nup136-2, although a similar reduction was observed, the process appeared to be delayed, as more than 10 TEL signals were still detected at the end of leptonema, suggesting impaired or slower TEL-mediated homolog pairing. In WT, CEN signals remained scattered throughout the nucleus, whereas NOR signals (located on the short arms of chromosomes 2 and 4) began to associate with the nucleolus by late leptonema. In contrast, nup136-2 showed an earlier reduction in the number of detectable CEN signals (from 10 to 5), and NORs became associated with the nucleolus more rapidly than in WT. At zygonema, CENs clustered completely in WT nuclei, and TELs adopted a bouquet-like arrangement in 1 nuclear hemisphere. However, nup136-2 zygotene nuclei did not display this typical bouquet configuration and often showed unresolved interlock chromosome structures. In pachynema, CEN, TEL, and NOR signals were paired and evenly distributed across the nuclear volume in both genotypes.

nup136-2 displays altered distribution of CEN, TEL, and NORs during prophase I. FISH analysis using CEN (green; bigger signals) and TEL (magenta; smaller signals) probes, counterstained with DAPI (gray), in WT and nup136-2 nuclei. Representative nuclei from different stages of prophase I are shown. DAPI-free images are shown to enhance visualization of specific signals. The white dotted line outlines the nucleolus. In nup136-2 early leptonema and leptonema, 2 associated NORs are highlighted (blue line), and arrows indicate paired CENs, suggesting a slightly advanced association of these regions compared to WT. In contrast, TEL pairing appears delayed, as judged by the higher number of distinct TEL signals persisting into late leptonema. In nup136-2 zygonema and pachynema, magnified views and schematic interpretations of complex chromosomal interlocks are shown. Bars = 5 μm.

Association of CEN, TEL, and NORs in WT and nup136-2 during prophase I. Graphs represent the dynamics of CENs (green), TELs (magenta), and NORs (blue) from G2 to pachynema (P). The left y axis indicates the number of TEL and CEN signals (associated signals), while the right y axis shows the percentage of NORs associated. Top, WT; bottom, nup136-2. In WT, TEL signals gradually decrease during early prophase I, indicating progressive TEL clustering, while CEN and NOR associations increase more slowly. In contrast, nup136-2 nuclei show delayed TEL pairing and a precocious association of CENs and NORs, suggesting temporal uncoupling in nuclear reorganization processes.

Synapsis is incomplete in nup136, but SYN1/REC8 and Cter-SUN protein localization remain unaffected

To investigate synapsis in nup136-2, we analyzed the localization of 3 key proteins in this process: ASY1, which associates with the chromosome axis (Ross et al. 1997); SYN1/REC8, a meiosis-specific cohesin (Bai et al. 1999); and ZYP1, the transverse filament component of the synaptonemal complex (SC) (Higgins et al. 2005). A substantial proportion of mutant meiocytes failed to achieve complete synapsis, as indicated by the reduced frequency of pachytene-like nuclei showing full ZYP1 signals (9.37%, n = 32) (Fig. 6a). On the other hand, we did not detect any abnormalities related to the loading of SYN1/REC8 in nup136-2 chromosomes (Fig. 6b), suggesting that sister chromatid cohesion is not compromised. Additionally, Cter-SUN-domain protein localization during meiosis was comparable to WT, showing a continuous signal around the NE from leptonema to pachynema, indicating that the global distribution of SUN proteins is not altered (Fig. 6c).

Immunolocalization of SC, cohesion, and NE proteins in WT and nup136-2. a) Pachytene nuclei showing ASY1 (green) and ZYP1 (magenta) localization in WT and nup136-2. Synapsis is not full in the mutant. b) Zygotene nuclei showing ASY1 (green) and SYN1 (magenta) distribution in WT and nup136-2. c) Zygotene nuclei stained with DAPI (gray) showing Cter-SUN-domain proteins (magenta) in WT and nup136-2. d) Zygotene nuclei of WT as negative control and nup136-2 complemented with NUP136:GFP showing GFP signal (green) and DAPI (gray). Bars = 5 µm.

Consistent with previous observations in somatic interphase cells (Bi et al. 2017), NUP136:GFP localized to the NE in meiocytes, displaying a perinuclear localization from leptotene to pachytene stages (Fig. 6d), similar to that observed for other Nups (Fernández-Jiménez et al. 2023). This localization pattern indicates that NUP1/NUP136 is integrated into the NPCs during meiosis, maintaining its association with the NE throughout prophase I.

NUP82 loss slightly enhances meiotic defects observed in nup136-2

Functional overlap is often observed among Nups belonging to the same subcomplex. In Arabidopsis, NUP82 and NUP1/NUP136 have previously been shown to be partially redundant, particularly in the context of the salicylic acid signaling pathway, since the double mutant nup82 nup136 displays a more severe phenotype than either single mutant (Tamura et al. 2017). To explore whether this functional redundancy extends to meiosis, we analyzed meiotic progression in the double mutant nup82-1 nup136-2. While nup82-1 (n = 60) and nup82-2 (n = 63) single mutants did not show any meiotic abnormalities (Fig. 1), the nup82-1 nup136-2 double mutant exhibited meiotic defects broadly similar to those of nup136-2 (Fig. 7a). A total of 19.05% of PMCs displayed unresolved chromosome interlocks at metaphase I (n = 84), a frequency comparable to that observed in the nup136-2 single mutant (24.39%, n = 41; P = 0.488). On the other hand, unlike nup136-2, the double mutant exhibited chromatin bridges accompanied by fragmentation at telophase I (51.85%, n = 27), suggesting increased severity in chromosome segregation defects. The average number of chiasmata per cell in the double mutant (5.63 ± 0.18, n = 79) was slightly lower than in nup136-2 (6.07 ± 0.22, n = 41; P = 0.0947) (Table 3), although the difference was not statistically significant. However, the proportion of homologous chromosome pairs failing to form at least 1 CO (univalents) increased significantly in the double mutant, rising to 21.27%, compared to 12.68% in nup136-2 (P = 0.010). Consequently, 74.68% of metaphase I cells contained at least 1 univalent pair and 6.78% exhibited up to 3 univalent pairs, a phenotype not observed in the single mutant nup136-2 (Fig. 7b; Table 3; Supplementary Table 3), while the frequency of ring bivalents remained comparable between nup136-2 and nup82-1 nup136-2 (P = 0.783) (Fig. 7c; Supplementary Table 3). These observations suggest that although NUP82 alone is not essential for meiosis, its absence in the mutant nup136-2 can slightly increase the severity of meiotic defects, suggesting a possible auxiliary contribution to CO assurance.

Cytological characterization of nup82-1 nup136-2 PMCs. a) Meiotic progression in the nup82-1 nup136-2 double mutant. Chromosomes were visualized with DAPI. Pachynema with an interlock (arrow); metaphase I with a pair of univalents (arrowheads) and an interlock between 2 bivalents (arrow); telophase I showing a chromatin bridge and a fragment (arrow); metaphase II with unbalanced segregation (6:4); and a tetrad with nuclei of unequal size. b) Representative metaphase I spreads from nup136-2 and nup82-1 nup136-2. Individual chromosomes or bivalents are numbered. nup82-1 images have been omitted for simplification, since they were indistinguishable from WT. c) Proportion of univalent and bivalents in nup136-2 (n = 41, blue, left) and nup82-1 nup136-2 (n = 79, orange, right). Bars = 5 µm.

Finally, we attempted to generate the sar1-4 nup136-2 double mutant to investigate the potential synergistic effects between Nups from different NPC subcomplexes. However, homozygous double mutants could not be recovered, even among the progeny of plants homozygous for 1 mutation and heterozygous for the other (n = 134). This suggests a genetic interaction between SAR1 and NUP1/NUP136 that compromises viability when both are impaired, as reported for other double mutants affected in Nups (Parry 2014).

Discussion

NPCs are the major mediators of nucleocytoplasmic exchange, and their components have been extensively studied in A. thaliana (Tamura et al. 2010; Tang et al. 2020). While the scaffold components of the NPC are generally conserved across eukaryotes, several Nups specific to plants, such as NUP1/NUP136 and NUP82, have been identified in A. thaliana (Tamura et al. 2010, 2017). However, the biological significance of these plant-specific Nups remains unknown, particularly during specialized developmental processes such as meiosis.

Previous studies have reported that the amount of NUP1/NUP136 affects nuclear morphology and size (Tamura and Hara-Nishimura 2011; Tamura et al. 2010). In somatic tissues, nup136 mutants exhibit more rounded nuclei, as well as an mRNA accumulation that could be caused by defects in nucleocytoplasmic transport of macromolecules (Lu et al. 2010; Tamura et al. 2010). In addition, nup136 mutants exhibit developmental defects, including fewer rosette leaves, early flowering, and low fertility (Lu et al. 2010; Tamura et al. 2010; Bao et al. 2019). The results presented in this study demonstrate that these fertility problems could be linked to meiotic abnormalities.

NUP1/NUP136 promotes conditions that support efficient CO formation

Our results indicate that NUP1/NUP136 contributes significantly to proper chiasma formation and subsequent accurate chromosome segregation during meiosis (Fig. 1). From the FISH analysis, we observed that nup136-2 exhibits a significant reduction in the average number of chiasmata per cell compared to the WT, representing a 40.48% decrease (Fig. 2; Table 1). This analysis, carried out at the bivalent level, revealed that this reduction affects all chromosomes but is particularly pronounced in chromosomes 1 and 5, which are the largest in the Arabidopsis complement and could have more complex spatial requirements for synapsis (Fransz et al. 1998). These chromosomes also show the highest frequency of univalents, suggesting a direct relationship between chromosomal size and the ability to pair and form COs in the mutant background. Interestingly, a greater reduction in chiasma frequency was observed in the short arms of acrocentric chromosomes (2 and 4), while no differences were detected on the long arm of chromosome 4. Accordingly, a lower frequency of univalents was observed for these chromosomes (Table 1). This suggests that the effect of nup136-2 is not uniform and may depend not only on chromosome size but also on chromosomal architecture and specific sequences. For instance, chromosomes 2 and 4 carry NORs, which remain associated with the nucleolus during early prophase I (Yang et al. 2006; López et al. 2012). The involvement of NOR in pairing was previously described in other organisms as Drosophila and certain rodent species (McKee and Karpen 1990; Stitou et al. 1997).

Despite the significant reduction in chiasma frequency in nup136-2, no chromosome fragments were detected. Therefore, the defect likely lies in the repair of DSBs using the homologous chromosome. The quantification of RAD51 and DMC1 foci (Fig. 3) suggests that DSB formation and early strand invasion steps proceed, although their slightly reduced number may reflect a subtle delay or instability in early recombination intermediates (Koszul and Kleckner 2009). More importantly, the number of MLH1 foci is reduced proportionally to the observed loss in chiasmata (Fig. 3), suggesting a defect in the formation of interfering COs. This alteration may explain the increased randomness in chiasma distribution observed in the mutant (Fig. 2). However, in this scenario, the apparent shift toward a Poisson-like distribution does not necessarily reflect a complete loss of the interference mechanism. The observed pattern may reflect a partial loss of interference, as well as a statistical effect due to the relative increase of noninterfering (class II) COs when class I COs are reduced.

These findings suggest that the recombination defect in nup136-2 reflects an indirect consequence of altered chromosome dynamics and pairing, which may compromise the stable association of homologs and the maturation of a subset of COs. Interestingly, meiotic alterations have also been reported in yeast Nup mutants. In Saccharomyces cerevisiae, the nuclear basket Nups Nup2 and Nup60 are essential for proper meiotic progression and transiently dissociate from the NPC during the first meiotic division. It has been proposed that their role involves tethering chromosomes to the nuclear periphery during prophase I (Chu et al. 2017; Komachi and Burgess 2022; King et al. 2023). These findings suggest that the role of nuclear basket components in facilitating recombination and chromosomal dynamics is broadly conserved across eukaryotes, albeit through species-specific mechanisms.

It is noteworthy that the recombination phenotype observed in nup136-2 differs markedly from that described in Arabidopsis sar1 and sar3 mutants, which also affect NPC structure (Fernández-Jiménez et al. 2023). In these mutants, meiotic defects do not involve the formation of univalents; instead, they present severe alterations in chromosome condensation and segregation, but with apparently normal homologous pairing and synapsis. These contrasting phenotypes underscore the intriguing possibility that Nups from distinct NPC subcomplexes contribute differently to key meiotic events.

NUP136 may influence chromosomal interactions to enable proper pairing and recombination

The persistence of unresolved interlocks from pachynema to metaphase I in the nup136-2 mutant (Fig. 2), together with the observation of full synapsis and normal SC protein loading in a subset of meiocytes (Fig. 6), suggests that the defect lies in early chromosome dynamics rather than in SC assembly itself. Several models have been proposed to explain interlock resolution including enzymatic mechanisms involving type II topoisomerases (eg TOPII) and physical resolution driven by TEL-led chromosome movement during the bouquet stage (Holm and Rasmussen 1980; Rasmussen and Holm 1980; Zickler and Kleckner 1999). In Arabidopsis, TOPII has been shown to participate in interlock resolution independently of chromosome movement, and mutations in both TOPII and NUP136 have an additive effect on the persistence of interlocks (Martinez-Garcia et al. 2018). In addition, the delayed TEL pairing observed in nup136-2, together with the earlier association of CENs and NORs compared to the WT (Figs. 4 and 5), is consistent with a role for NUP1/NUP136 in coordinating chromosome dynamics along the NE. In this context, similar phenotypes, characterized by an increased frequency and persistence of interlocks, have been observed in mutants defective for SUN1 and SUN2 in Arabidopsis (Varas et al. 2015) and PAM1 in maize (Golubovskaya et al. 2002). Furthermore, recent studies have demonstrated that rapid prophase chromosome movements (RPMs) in Arabidopsis are linked to essential reorganization at the NE, involving SUN-domain-containing proteins (Cromer et al. 2024). In this context, interactions between specific Nups and SUN1 have been described in Arabidopsis and are associated with CEN distribution during mitosis (Ito et al. 2024). In the nup136-2 mutant, although SUN proteins appear broadly distributed along the NE (Fig. 6c), it is possible that the absence of NUP1/NUP136 affects the fine-scale organization or dynamics of LINC complexes. This may lead to a partial impairment in TEL anchoring and movement, consistent with the variability of the meiotic phenotype observed. Moreover, previous studies in plants have also shown that NUP1/NUP136 influences nuclear shape and likely connects NPCs to the plant lamina (Tamura and Hara-Nishimura 2011, 2013). Given this molecular and functional context, it is tempting to speculate that NUP1/NUP136 may contribute to meiotic chromosome dynamics by modulating the interplay between NPCs and the LINC/lamina network during early prophase I.

Functional divergence of the NUP1/NUP136 paralog NUP82 in meiosis

NUP82 is a plant-specific paralog of NUP1/NUP136, likely the result of a gene duplication event (Tamura et al. 2017). In addition, both proteins, NUP1/NUP136 and NUP82, connect with the nuclear lamina (Mermet et al. 2023; Groves et al. 2025). In Arabidopsis, NUP1/NUP136 and NUP82 have previously been shown to act redundantly in salicylic acid signaling (Tamura et al. 2017). This prompted us to analyze meiotic progression in the nup82-1 nup136-2 double mutant. While nup82-1 single mutants do not exhibit visible meiotic defects (Fig. 1), the double mutant presents a mild enhancement of the nup136-2 phenotype including a further increase in univalent formation and occasional chromosome fragmentation at telophase I not present in the single mutants (Fig. 7; Table 3). Therefore, while NUP82 does not have an essential role in meiosis, it appears to subtly influence chromosomal interactions in the context of NUP1/NUP136 deficiency.

Conclusions

Our study uncovers a key role for the plant-specific Nup NUP1/NUP136 in chromosomal interactions and CO formation in A. thaliana. The reduced chiasma frequency and persistence of chromosomal interlocks in nup136-2 suggest that NUP1/NUP136 contributes to homolog recognition and recombination, although successful pairing and even the formation of ring bivalents can still occasionally occur. Altogether, our findings highlight that plant Nups, beyond their canonical roles in nucleocytoplasmic transport, are important for orchestrating meiotic chromosome behavior and underscore the functional diversity of NPC subcomplexes in meiosis.

Supplementary Material

iyag010_Supplementary_Data