Telomeres, the nuclear lamina, and membrane remodeling: Orchestrating meiotic chromosome movements

As chromosomes evolved from an ancestral circular structure to a linear configuration during the prokaryote-to-eukaryote evolutionary juncture (Lue, 2018), the nucleoprotein structures at chromosomal ends, the telomeres, emerged as indispensable elements for protecting genetic information as well as for ensuring faithful sexual reproduction. The telomeric nucleoprotein is composed of a tandem array of short repetitive DNA sequences and their associated sequence-specific DNA-binding proteins known as the shelterin complex (de Lange, 2018). Telomeres fulfill three evolutionarily conserved functions.

First, telomeres maintain their repetitive DNA sequences, thus counteracting the inherent limitations of canonical DNA polymerase, which cannot fully replicate the 5′ end of lagging strand DNA (Olovnikov, 1973). Without this maintenance, telomeric DNA would gradually shorten with each cell cycle, eventually exposing non-telomeric regions and risking genetic instability. To prevent this, telomeres activate DNA-lengthening pathways, including telomerase-mediated reverse transcription and homologous recombination–based alternative lengthening mechanisms (Cesare and Reddel, 2010; Shay and Wright, 2019). These pathways protect against replicative senescence and maintain genome integrity in highly proliferative cells.

Second, telomeres prevent aberrant activation of DNA damage response (DDR) pathways (de Lange, 2009). The natural ends of linear DNA mimic DNA double-strand breaks and thus can mistakenly trigger ATM/ATR-dependent DDRs. Telomeres prevent such responses by sequestering DDR proteins, forming protective T-loop structures, and condensing telomeric chromatin (Bandaria et al., 2016; Doksani et al., 2013; Okamoto et al., 2013). DDRs are also activated as a consequence of critical telomere shortening in senescent cells; thus, the first and second roles are interconnected (d’Adda di Fagagna et al., 2003).

Third, in a function specific to germ cells, telomeres facilitate chromosome movements during meiotic prophase I, promoting the pairing and synapsis of homologous chromosomes—an essential process for ensuring faithful sexual reproduction (Zickler and Kleckner, 2016). This telomeric function is highly conserved throughout eukaryotes; however, in rare exceptions, it is substituted by other chromosomal regions, such as centromeres in Drosophila melanogaster (Rubin et al., 2022; Takeo et al., 2011) or repetitive DNA sequences near telomeres in Caenorhabditis elegans (discussed later in this review) (Rog and Dernburg, 2013).

While the first two functions and their implications for human health have been extensively reviewed elsewhere (Arndt and MacKenzie, 2016; de Lange, 2018; Maciejowski and de Lange, 2017; Palm and de Lange, 2008), this review will focus on the third, less explored function of telomeres in germ cells, which has seen rapid advancements in recent years.

Eukaryotic genomes are partitioned into linear chromosomes, with the chromosome number varying by organism such as 2n = 6 in fission yeast (Schizosaccharomyces pombe), 32 in budding yeast (Saccharomyces cerevisiae), 46 in humans, 40 in house mice, and 12 in the nematode C. elegans. The chromosomes comprising these diploid genomes occupy distinct regions in the nucleus known as the chromosomal territories, and the pairs of homologous chromosomes are spatially separated (Cremer and Cremer, 2010) (Fig. 1 A). Specifically during meiotic prophase I, homologous chromosomes move rapidly within the nucleoplasm to search for their correct homologous partners. Upon locating their partners, the chromosomes pair, synapse, and recombine, a process that is prerequisite for the subsequent reductional segregation of homologous chromosomes in meiosis I (Fig. 1 B). As early as 1921, the distinctive arrangement of meiotic telomeres was reported during flatworm meiosis, a process in which all telomeres are attached to the nuclear envelope (NE) and cluster on one side of the nucleus near the centrosome (Gelei, 1921). This meiosis-specific configuration of the chromosomes is referred to as the chromosomal bouquet (Fig. 1 B), and it is conserved within a wide range of eukaryotes from unicellular organisms like fungi to higher eukaryotes, including plants, fish, mice, and humans (Harper et al., 2004; Scherthan, 2001). The bundling of chromosomal ends likely facilitates the homology search by bringing homologous DNA strands into close proximity, aligning them along the entire length of the chromosomes from one telomere to the opposite telomere. This, in turn, promotes the initial recombination-independent homolog pairing through a complex set of mechanisms, including DNA sequence–dependent base paring, noncoding RNA-dependent RNA-DNA hybrid formation, and protein-dependent recognition processes (Ding et al., 2012; Gladyshev and Kleckner, 2014; Ishiguro et al., 2014; Rhoades et al., 2021). In addition to this classical point of view, the telomere-driven chromosome movements likely have noncanonical roles, such as the elimination of inappropriate pairing or entanglements between nonhomologous chromosomes (Koszul and Kleckner, 2009; Zickler and Kleckner, 2015). Considering that telomere-driven chromosome movements persist until pachynema, at which point homologous chromosomes are fully paired and synapsed (Lee et al., 2015; Shibuya et al., 2014b), there is likely to be some undescribed postsynapsis function as well, for example, promoting the disassembly of the synaptonemal complex or the resolution of recombination intermediates by applying mechanical forces to the chromosomal axes. Indeed, the chromosome movements nearly cease in late diplonema, coinciding with the completion of meiotic recombination and the desynapsis of homologous chromosomes (Shibuya et al., 2014b).

The key regulatory proteins behind the meiosis-specific telomere functions were first discovered in fission yeasts and budding yeasts during the golden age of yeast genetics research, spanning from the 1990s to the 2000s. In fission yeast, telomeres in the vegetative cells are steadily attached to the inner nuclear membrane (INM) via the interaction between the shelterin component Rap1 and the INM protein complex Bqt3/4 (Chikashige et al., 2009) (Fig. 2 A). Upon entry into meiosis, the meiosis-specific telomere-binding proteins, Bqt1/2, are recruited to the telomere-NE interface via interactions between Rap1 and Bqt2 (Chikashige et al., 2006). Bqt1/2 recruit the linker of nucleoskeleton and cytoskeleton (LINC) complex comprising the SUN-domain protein Sad1 and the two KASH-domain proteins Kms1/2. Sad1 and Kms1/2 localize at the INM and the outer nuclear membrane (ONM), respectively, and they interact with each other in the NE lumen (Fig. 2 A). This transmembrane linkage connects telomeres in the nucleoplasm to the microtubule cytoskeletons and drives the movements of telomeres via cytoplasmic dynein and kinesin-dependent forces (Chikashige et al., 2006; Yoshida et al., 2013). In budding yeast, the Ndj1 protein, a functional homolog of Bqt1/2, facilitates the linkage of meiotic telomeres to the LINC complex comprising the SUN-domain protein, Mps3, and the two KASH-like proteins, Mps2/Csm4 (Conrad et al., 1997, 2008; Lee et al., 2020; Trelles-Sticken et al., 2000) (Fig. 2 B). The Mps3 and Mps2/Csm4 complex transverses the INM and ONM, links the NE-anchored telomeres to the actin cytoskeletons, and drives the movements of telomeres via actin-based myosin motor proteins, such as Myo2 (Conrad et al., 2008; Kosaka et al., 2008; Lee et al., 2020; Wanat et al., 2008). Yeast cells with mutations in these genes all show defects in rapid chromosome movements and homologous recombination and thus have reduced spore viability (Chikashige et al., 2006; Conrad et al., 2007; Kosaka et al., 2008; Wu and Burgess, 2006).

Following the golden age of yeast genetics, research on mammalian meiosis flourished. The meiosis-specific telomere-binding proteins TERB1 and TERB2, along with the meiosis-specific INM-binding protein MAJIN, were discovered in mice in 2014 and 2015 (Shibuya et al., 2014a, 2015). The TERB1–TERB2–MAJIN ternary complex plays analogous roles as the fission yeast Bqt1/2/3/4 and budding yeast Ndj1 and ensures telomere attachment to the INM. This attachment is mediated by the interaction of TERB1 with the shelterin protein TRF1, the INM-binding activity of the transmembrane protein MAJIN, and the TERB2-dependent molecular linkage between TERB1 and MAJIN (Shibuya et al., 2014a, 2015) (Fig. 2 C and Fig. 3 A). In this review, the term “attachment” in the context of mammalian meiosis refers to this proteinaceous connection of telomeres to the lipid bilayer, mediated by TERB1–TERB2–MAJIN and characterized by the formation of the attachment plate structure observed by electron microscopy. Clarifying this definition is crucial because, under certain conditions where attached telomeres lose their mobility—such as in Cdk2 and Sun1 knockout (KO) meiocytes (details to be discussed later) or wild-type meiocytes treated with nocodazole—an incipient attachment mediated by TERB1–TERB2–MAJIN still forms at the nuclear periphery accompanied by intact attachment plate structures, but the remaining telomeres stay within the internal nuclear region (Link et al., 2014; Shibuya et al., 2015; Viera et al., 2015). This phenotype is primarily attributed to defects in telomere movement, which plays a role in facilitating the attachment of internally located unattached telomeres by shuffling the nucleoplasm. To date, no mutants other than those in Terb1, Terb2, and Majin have been reported to completely lack an attachment plate structure (Shibuya et al., 2015).

After the establishment of the incipient attachment, the TERB1–TERB2–MAJIN ternary complex recruits the LINC complex, consisting of SUN1 at the INM and KASH5 at the ONM (Ding et al., 2007; Morimoto et al., 2012) (Fig. 2 C). The LINC complex ensures the transmembrane linkage between telomeres and the cytoskeleton and drives the cytoplasmic microtubule–dependent chromosome movement, which further facilitates the attachment of the remaining internal telomeres. Both TERB1 and MAJIN directly bind to the SUN1 N-terminal domain, which protrudes to the nucleoplasm (Shibuya et al., 2014a; Wang et al., 2020) (Fig. 3 A). KASH5 directly binds to the dynein light-intermediate chain via its cytoplasmic N-terminal region, which is composed of a pair of EF-hands and a coiled-coil domain (Agrawal et al., 2022; Garner et al., 2023) (Fig. 3 A), in a manner conserved among dynein-activating adapter proteins that converts dynein and dynactin into a processive motor (Olenick and Holzbaur, 2019). The KASH5–light-intermediate chain interaction is indispensable for the recruitment of dynein complexes to meiotic telomeres or to the ONM in the case where KASH5 is ectopically expressed in somatic cells (Agrawal et al., 2022; Garner et al., 2023). KASH5 activates dynein motility in vitro, and its effect is enhanced by the addition of LIS1, a dynein regulatory factor, to achieve full velocity (Agrawal et al., 2022). Thus, KASH5 not only functions as a transmembrane linker for the dynein complex, but also functions as an activating dynein adapter.

The KO mice for all involved genes, namely Terb1, Terb2, Majin, Sun1, and Kash5, show severe defects in homolog pairing and recombination, which leads to both male and female infertility (Ding et al., 2007; Horn et al., 2013; Shibuya et al., 2014a, 2015). Mutations in human homologs of these genes are associated with human infertility, including nonobstructive azoospermia and ovarian insufficiencies, implying their conserved roles in human meiosis (Fakhro et al., 2018; Meng et al., 2023; Salas-Huetos et al., 2021; Sharifi et al., 2025; Wu et al., 2022; Yalcin et al., 2024). Fission yeast Bqt1/2, budding yeast Ndj1, and mammalian TERB1–TERB2–MAJIN are highly divergent proteins with no detectable sequence homology, suggesting rapid evolution and functional convergence. In contrast, the TERB1–TERB2–MAJIN module is remarkably well conserved within the metazoan clade, having emerged early in the metazoan era, as evidenced by homologs found in basal metazoans, such as Cnidaria, Placozoa, and Porifera (da Cruz et al., 2020).

The domain conformation of TERB1 is highly conserved from basal metazoans to humans, and it consists of an N-terminal armadillo repeat domain, a C-terminal TRF1-binding (TRFB) domain, and a MYB-like DNA-binding (MYB) domain at the C-terminal end (da Cruz et al., 2020; Shibuya et al., 2014a) (Fig. 3 A). The armadillo repeat binds to SUN1 in yeast two-hybrid analyses and pull-down assays (Shibuya et al., 2014a; Wang et al., 2020), but the in vivo significance of this interaction has yet to be determined. The TRFB domain is necessary and sufficient for the telomere localization of TERB1 through binding to TRF1 (Zhang et al., 2017) (Fig. 3 A). Protein crystal structures show that two TRFB peptides of TERB1 bind to a TRF1 homodimer, forming a 2:2 stoichiometric TERB1–TRF1 complex in a manner analogous to the interaction between TRF1 and the shelterin component TIN2 (Long et al., 2017; Pendlebury et al., 2017). Indeed, TERB1 and TIN2 use similar motifs, ILLTP in TERB1 and FNLAP in TIN2 (where the bold letters match the FxLxP TRF1-binding consensus, barring the presence of a phenylalanine to isoleucine substitution in TERB1), for binding to the TRF1-homology domain of TRF1, and they bind to TRF1 in a mutually exclusive manner (Chen et al., 2008; Pendlebury et al., 2017). These findings suggest that there exist two mutually exclusive complexes seeded by distinct populations of TRF1 at meiotic telomeres, namely the canonical shelterin complex consisting of TRF1–TIN2–TRF2–RAP1–TPP1–POT1 and the meiosis-specific TRF1–TERB1–TERB2–MAJIN complex. CDK-dependent phosphorylation of TERB1 T648, which is located within the TRF1-binding consensus motif (ILLTP) (Fig. 3 A), significantly weakens the TERB1–TRF1 interaction by inducing electrostatic repulsion with a glutamate in the TRF1-homology domain of TRF1 (E106 and E93 in humans and mice, respectively) (Long et al., 2017; Pendlebury et al., 2017; Shibuya et al., 2014a). In vivo, phosphorylation of TERB1 at T648 in telomeres gradually accumulates throughout prophase I, reaching its peak toward late-pachynema in a CDK activity-dependent manner (Shibuya et al., 2015). While TERB1–TERB2–MAJIN and TRF1 all colocalize at telomeres by forming the tetrameric complexes in early prophase I, the TERB1–TERB2–MAJIN ternary complex dissociates from TRF1, as well as the rest of the shelterin components, at the early to late-pachytene transition and forms a small compartment surrounded by the broader ring-like shelterin subdomains (Shibuya et al., 2015). This cytologically defined phenomenon, termed the telomere cap exchange, is likely triggered by the weakening of the TRF1–TERB1 interaction due to the gradual accumulation of TREB1 T648 phosphorylation. Indeed, the artificial enforcement of the TRF1–TERB1 interaction by introducing the TRF1 E93K substitution in mice inhibits the cap exchange (Pendlebury et al., 2017).

The generation of Terb1 knock-in mice harboring TRF1 binding-deficient mutations, namely the substitution of ILLTP to ILAEA in the Terb1 gene, showed that this mutation did not completely abrogate the telomere localization of TERB1 (Long et al., 2017). This is due to the remaining TRF1 interaction mediated by the rest of the TRFB domain in TERB1 aside from this short motif, as suggested by in vitro biochemistry experiments (Pendlebury et al., 2017). This Terb1^AEA knock-in mouse showed moderate pairing defects in males but not females meiosis, suggesting that the attenuated, if not completely abrogated, TERB1–TRF1 interaction leads to hypomorphic defects specifically in male meiosis (Long et al., 2017).

The TRFB domain of TERB1 contains a short motif called the TERB2-binding motif, which directly binds to TERB2 (Zhang et al., 2017) and is highly conserved within metazoans (da Cruz et al., 2020) (Fig. 3 A). Two TERB2-binding motifs within the 2:2 stoichiometric TRF1–TERB1 heterotetrametric complex bind to a 2:2 stoichiometric TERB2–MAJIN heterotetrametric complex to attach the telomere to the NE (Dunce et al., 2018). Of note, there are patches of basic amino acids within TERB2 and MAJIN (Fig. 3 A: BP) that directly bind to double-stranded and single-stranded DNA in a sequence-independent manner (Dunce et al., 2018; Shibuya et al., 2015). Mutations in a basic patch of the MAJIN protein attenuate the attachment of telomeres to the NE (Shibuya et al., 2015). These findings suggest that the TERB2–MAJIN complex likely takes over telomeric DNA from TRF1 after telomeric DNA is brought into close proximity to the INM through the interaction between DNA-bound TRF1 and membrane-bound TERB1 to establish the more stable attachment involving the direct tethering of telomeric DNAs to the membrane-bound TERB2–MAJIN complex (Fig. 3, B–D). This transition from initial attachment to stable attachment might correspond to the cytologically observed telomere cap exchange, which occurs at the early to late-pachytene transition. Indeed, fluorescent in situ hybridization signals of telomeric DNAs colocalize with TERB1–TERB2–MAJIN rather than the surrounding shelterin ring-like structure after the cap exchange reaction (Dunce et al., 2018; Shibuya et al., 2015).

The MYB domain in TERB1 shares significant sequence homology with that of TRF1 and TRF2, which directly bind telomeric double-stranded DNA, but lacks conserved DNA-contacting residues, resulting in the loss of DNA-binding capacity (Zhang et al., 2022). The TERB1 MYB domain is dispensable for the telomeric localization of TERB1 and most TERB1 functions, including the recruitment of TERB2-MAJIN and SUN1-KASH5, the establishment of NE attachment, and even its role in fertility. The role of the MYB domain was initially suggested by the expression of TERB1 lacking the MYB domain (TERB1^ΔMYB) in Terb1 KO testes (Shibuya et al., 2014a), and this was later confirmed by the analysis of Terb1 point mutant mice expressing TERB1^ΔMYB protein from the endogenous Terb1 locus (Zhang et al., 2022). In both cases, the TERB1^ΔMYB protein rescued the telomere attachment defects but not the telomere structural defects seen in Terb1 KO testes, such as splits and bridges/fusions of telomeric DNAs between nonhomologous chromosomes. Cohesin mislocalization is likely a direct cause of the telomere structural defects seen in Terb1 KO and Terb1^ΔMYB mice, and this is supported by the phenotypic similarity seen in Smc1β KO spermatocytes, all of which show mislocalization of the cohesin axial core specifically at telomeres, which causes similar telomere structure defects (Adelfalk et al., 2009; Biswas et al., 2018). Indeed, the TERB1 C terminus region, comprising the TRFB domain and MYB domain, directly binds to the meiotic cohesin subunit STAG3 in yeast two-hybrid analysis (Shibuya et al., 2014a), but this interaction requires further validation using alternative methods, such as biochemical approaches. In vivo, the TERB1^ΔMYB protein co-immunoprecipitates less cohesin and axial element proteins compared with the full-length protein from testis extracts, suggesting that the MYB domain is indispensable for the stable interaction with the cohesin axial core (Zhang et al., 2022) (Fig. 3 A).

The telomere structural defects caused by the loss of the TERB1 MYB domain are cytologically reminiscent of the fragile telomere phenotypes reported in TRF1-depleted mitotic telomeres (Sfeir et al., 2009); however, it does not induce the DDR unlike the TRF1-depleted mitotic cells (Zhang et al., 2022). This suggests that meiotic telomere defects caused by deletion of the TERB1 MYB domain are fundamentally distinct biological phenomena from fragile telomeres caused by the replication stress seen in TRF1-depleted meiotic cells, despite their apparent structural similarities. The molecular mechanisms underlying this enigmatic phenotype are yet to be explored. Even though the TERB1 MYB domain is dispensable for the normal fertility in laboratory-housed mice, the TERB1 MYB domain is highly conserved across a wide array of metazoan species from basal metazoans to humans (da Cruz et al., 2020), suggesting that the MYB domain–dependent suppression of meiotic telomere splitting, bridge/fusion, and ultimately DNA erosion might protect telomeric DNA from genomic instability over long evolutionary timescales and could potentially help maintain species-specific karyotypes by inhibiting chromosome end-to-end fusion, which would otherwise be inherited by the next generation.

In addition to the TERB1–TERB2–MAJIN complex, the CDK2 kinase and its noncanonical activator SPDYA (also known as RingoA) stably localize to meiotic telomeres throughout prophase I (Ashley et al., 2001; Mikolcevic et al., 2016) (Fig. 2 C). Both Cdk2 and Spdya KO mice show defects in the nuclear peripheral distribution of telomeres, leading to both male and female infertility (Mikolcevic et al., 2016; Ortega et al., 2003; Viera et al., 2015). In Spdya KO mice, the kinase activity of CDK2 in testis extracts significantly decreases, and the telomeric CDK2 signals disappear completely, suggesting that SPDYA functions as the activator of CDK2 kinase and regulates its localization to meiotic telomeres (Mikolcevic et al., 2016; Tu et al., 2017). Notably, TERB1 still localizes to telomeres, and thus the attachment plates are formed at the nuclear periphery while SUN1 localization is completely lost in Spdya KO spermatocytes, suggesting that CDK2-SPDYA functions downstream of TERB1–TERB2–MAJIN-dependent incipient telomere attachment and is indispensable for the recruitment of the SUN1-KASH5 movement apparatus (Mikolcevic et al., 2016) (Fig. 3, B and C). Conversely, the telomeric localization of CDK2 depends on SUN1 (Liu et al., 2014), suggesting that SUN1 and CDK2-SPDYA localize to meiotic telomeres in an interdependent manner. In support of this idea, the N-terminal domain of SUN1, which protrudes into the nucleoplasm, directly binds to SPDYA and is phosphorylated by CDK2-SPDYA in vitro (Mikolcevic et al., 2016; Wang et al., 2020) (Fig. 3 A). The MAJIN-binding site resides adjacent to the SPDYA-binding sites within the SUN1 N terminus (Fig. 3 A), and inhibition of CDK2 activity attenuates MAJIN–SUN1 interactions in vitro (Wang et al., 2020).

The introduction of the SPDYA binding-deficient mutation W151R in the murine Sun1 gene drastically impairs the SUN1-KASH5 telomeric localization and this Sun1^W151R knock-in mouse phenocopies Sun1 and Spdya KO mice (Chen et al., 2021). The localization of SPDYA is also drastically reduced in Sun1^W151R spermatocytes, consistent with the idea that the telomeric localizations of CDK2-SPDYA and SUN1 are interdependent.

Collectively, these findings lead to a simple working model where CDK2-SPDYA binds to and phosphorylates the SUN1 N terminus, which then triggers the relocalization of SUN1 from the entire INM to the telomeric puncta by enhancing SUN1–MAJIN interactions (Fig. 3, C and D). In vitro kinase assay using the recombinant CDK2–SPDYA protein complex and SUN1 N-terminal fragments identified serine 48 of SUN1 as a major CDK2-SPDYA–-dependent phosphorylation site (Mikolcevic et al., 2016), but its biological significance in vivo remains unverified. Consequently, the specific SPDYA-CDK2–dependent phosphorylation sites on SUN1 that impact protein interactions or in vivo functions have yet to be determined.

C. elegans is an exceptional model system among metazoans for the study of meiosis-specific chromosome movements, as it is not the telomeres but clusters of 11–12 base pair DNA motifs, known as pairing centers (PCs), near one end of each chromosome that mediate chromosome attachment to the NE (MacQueen et al., 2005; Phillips and Dernburg, 2006; Phillips et al., 2005, 2009) (Fig. 2 D). Despite this unique arrangement, the regulation of the PCs closely parallels that of meiotic telomeres in other metazoans, offering significant insights into evolutionary conservation and diversity. Each PC recruits one of four meiosis-specific zinc finger proteins, ZIM-1/2/3 and HIM-8, through unique repetitive DNA sequence motifs (Li et al., 2024; Phillips et al., 2009) (Fig. 2 D). Subsequently, these PC proteins recruit the LINC complex consisting of SUN-1 at the INM and ZYG-12 at the ONM (Penkner et al., 2007; Sato et al., 2009) to connect PCs to cytoplasmic microtubules and associated dynein motors, driving the chromosome movements (Wynne et al., 2012). MJL-1, a functional homolog of MAJIN in C. elegans, mediates the interaction between PC proteins and SUN-1, though their direct protein interaction remains untested experimentally (Kim et al., 2023) (Fig. 2 D). It is noteworthy that C. elegans lacks TERB1 and TERB2 homologs, suggesting that the TERB1–TERB2 complex, which is specialized for meiotic telomere functions in other metazoans, has been functionally replaced by PC proteins in C. elegans through convergent evolution. Similarly, although MAJIN and MJL-1 share no sequence homology except for the presence of a single transmembrane domain at their C terminus, they play a common role as dedicated meiosis-specific binding partners of SUN domain proteins, regardless of whether they localize to telomeres or PCs.

Analogous to the involvement of CDK2-SPDYA in mammalian meiotic telomere regulation, the polo-like kinases PLK-1/2 and the checkpoint kinase CHK-2 localize at the PCs in a PC protein-dependent manner, promoting SUN-1 accumulation at PCs and the subsequent movement of PCs along the NE in C. elegans (Harper et al., 2011; Kim et al., 2015; Labella et al., 2011) (Fig. 2 D). PLK-2 and CHK-2 phosphorylate the N terminus of SUN-1 at multiple sites (Harper et al., 2011; Penkner et al., 2009). However, non-phosphorylatable mutations in all identified in vivo SUN-1 phospho-residues do not cause severe meiotic defects, unlike the depletion of SUN-1, PLK-2, or CHK-2, suggesting that the key physiologically relevant phospho-residue remains unidentified (Woglar et al., 2013). Thus, phospho-regulation of the N-terminal nuclear extrusion of SUN-domain proteins, catalyzed by CDK2-SPDYA in mammals and by PLK-2 and CHK-2 in C. elegans, appears to be conserved across phyla, and this facilitates the mobility of the LINC complex and enhances its direct molecular interaction with MAJIN in mammals and probably with MJL-1 in C. elegans.

The nuclear lamina is a polymeric protein network associated with the nucleoplasmic surface of the INM and functions as the structural scaffold of the INM (Karoutas and Akhtar, 2021). Various lamin proteins and their isoforms are expressed in mammalian mitotic cells, including lamin B1 (encoded by Lmnb1), lamin B2 (encoded by Lmnb2), and two A-type lamin isoforms, lamins A and C (encoded by Lmna) (Höger et al., 1988; Lin and Worman, 1993; Röber et al., 1989; Zewe et al., 1991). Compared with somatic cells, mammalian spermatogenic cells express distinct combinations of lamin proteins. Among the somatic lamins, only lamin B1 is expressed in spermatogenic cells (Vester et al., 1993). Furthermore, spermatogenic cells express two germline-specific short lamin-splicing variants, namely lamin C2 and B3, which arise from the Lmna and the Lmnb2 genes, respectively (Nakajima and Abe, 1995; Wise, 1975). Whereas the expression of lamin B3 is restricted to postmeiotic stages (i.e., spermiogenesis), lamin C2 is exclusively expressed in spermatocytes, implying its role in meiotic chromosome movements (Alsheimer and Benavente, 1996; Furukawa et al., 1994; Schütz et al., 2005; Smith and Benavente, 1992).

The ectopic expression of meiosis-specific lamin C2 in somatic cells results in higher mobility compared with its somatic counterpart lamin C (Jahn et al., 2010). Furthermore, the telomere-NE interface in rat spermatocytes locally enriches lamin C2 (Alsheimer et al., 1999). Direct functional insight was gained from the analysis of a lamin C2 isoform-specific KO mouse model, which exhibited defects in homologous synapsis and male-specific infertility (Link et al., 2013). Neither the telomere attachments nor the recruitment of SUN1 are defective in the mutant mice, but there is an accumulation of bouquet stage spermatocytes, which are otherwise rare and only transiently observed between the leptotene and zygotene stages in WT spermatocytes, suggesting that the telomere-driven chromosome movements are attenuated in the absence of meiotic lamin C2 (Link et al., 2013). These finding suggest that incorporation of meiosis-specific lamin C2 fluidizes the nuclear lamina to allow for the efficient movements of NE-attached telomeres.

C. elegans has only one B-type lamin homolog, LMN-1, which is encoded by the lmn-1 gene (Liu et al., 2000). LMN-1 is phosphorylated by CHK-2 and PLK-2 at the PC-NE–attachment sites, which triggers the mobilization of the nuclear lamina for the efficient movement of PCs along the NE (Link et al., 2018). LMN-1 phosphorylation by PLK-2 in C. elegans also plays a checkpoint function by triggering the mechanosensitive cation channel, PEZO-1, which eliminates synapsis-defective oocytes by activating apoptosis (Baghdadi et al., 2024). Thus, in C. elegans, LMN-1 is not only involved in meiosis-specific chromosome movements, but also in the quality control mechanism carried out by the PC-bound PLK-2 subpopulation. It is unlikely that an analogous checkpoint mechanism is mediated by the meiotic telomere-localizing kinase CDK2-SPDYA in other metazoan systems because CDK2-SPDYA remains constitutively localized at meiotic telomeres throughout prophase I. This is in contrast to PLK-2 in C. elegans, which relocalizes from PCs to the synaptonemal complex upon the completion of homologous synapsis and thus inactivate the checkpoint (Brandt et al., 2020; Harper et al., 2011).

In addition to the reorganization of the nuclear lamina, the composition of fatty acids comprising the NE seems to be under testis-specific regulation in mammals to fluidize the NE. The fluidity of the membrane is primarily promoted by the elongation of fatty acid carbon chains through the activity of the fatty acid elongases as well as the incorporation of more double bonds through the activity of fatty acid desaturases (de Mendoza and Pilon, 2019; Pilon and Ruiz, 2023). Mammalian testis are enriched in phospholipids and sphingolipids comprising the very long-chain polyunsaturated fatty acids (VLC-PUFAs), which contain carbon chains with 22 or more carbon atoms and three to six double bonds. The genetic depletion of fatty acid metabolism enzymes, including the elongase ELOVL2, the ceramide synthase CerS3, and the desaturase FADS2, causes complete male infertility, with depletion of FADS2 also resulting in female infertility (Rabionet et al., 2008, 2015; Stoffel et al., 2020; Zadravec et al., 2011). The analysis of the genetic deletion of an evolutionarily conserved membrane fluidity sensor, AdipoR2 (Pilon, 2021), showed that AdipoR2 upregulates the expression of ELOVL2 at both the mRNA and protein level to ensure the synthesis of VLC-PUFAs (Zhang et al., 2024). This parallels the role of AdipoR1, a close paralog of AdipoR2, in promoting VLC-PUFA synthesis in retinal cells for photoreceptor renewal through transcriptional activation of Elovl2 (Osada et al., 2021; Rice et al., 2015). Thus, the AdipoR1/2-ELOVL2 functional axis appears to be the master regulator of VLC-PUFA synthesis and is likely to be conserved beyond the specific tissue contexts. AdipoR2 KO male germ cells have abnormally stiffened cellular membranes, and this leads to defects in intercellular bridge stability, defects in meiotic homolog pairing/recombination, and male infertility. In these cells, even though TERB1-TERB2-MAJIN-SUN1-KASH5 all localize at meiotic telomeres and establish normal attachment plates, some telomeres localize at internal nuclear regions due to the invagination of the NE-attached telomeres toward the inside of the nucleoplasm (Zhang et al., 2024). These defects suggest that dynein-dependent movement forces applied to telomeres embedded in the stiffened NE-lacking VLC-PUFA lead to the NE curvature rather than the movement of telomeres within the NE. AdipoR2 KO female mice are fertile (Zhang et al., 2024), similar to the lamin C2-deficient female mice (Link et al., 2013), suggesting the sexual dimorphic requirement for the fluidization of the nuclear membrane and nuclear lamina for the faithful progression of meiosis. Alternatively, distinct factors from those in the testis might regulate the fluidity of the nuclear lamina and nuclear membrane in the ovary. Emerging evidence suggest that Bqt4 in fission yeast also plays a direct role in the regulation of membrane homeostasis by directly binding to and recruiting phosphatidic acid to the NE (Hirano et al., 2024) as well as interacting with lipid synthesis enzymes (Hirano et al., 2023). Although not yet directly tested, the role of Bqt4 in regulating the composition of the NE may be crucial for modifying the meiotic lipidome, fluidizing the NE, and facilitating rapid chromosome movements in a similar manner as AdipoR2-ELOVL2 in mammals. It is further possible that MAJIN, the functional homolog of Bqt4 in mammals, also plays a similar direct role in the modification of membrane homeostasis in meiosis in concert with the action of AdipoR2-ELOVL2.

The molecular regulation of meiosis-specific rapid chromosome movements, crucial for sexual reproduction and heredity, has been progressively elucidated by integrating genetics, cytology, biochemistry, protein crystallography, and lipidomics studies in various model organisms. Nevertheless, this field remains a key area of active research, with many critical questions yet to be answered. For instance, how do chromosomes identify the correct homologous partners once physical juxtaposition is established? What is the role of persistent rapid chromosome movements after homolog pairing and synapsis are completed? How are telomere lengthening and protection regulated in germ cells, along with meiosis-specific functions? How do distinct cellular compartments, such as chromosomes, the nuclear lamina, the lipid bilayer, and the cytoskeleton, interact and coordinate to accomplish a single molecular task? Recent advances hold great promise for rapidly advancing our understanding of these and other important questions.

Some schematics in Figs. 1 and 2 were created using https://BioRender.com. The images of fission and budding yeast in Fig. 2 are Copyrighted 2007 by Emi Kosano.

This work was supported by institutional funding from RIKEN.

Author contributions: H. Shibuya: conceptualization, funding acquisition, visualization, and writing—original draft, review, and editing.