Formation and resolution of meiotic chromosome entanglements and interlocks Free

Meiosis is a specialized cell division that is required to produce gametes containing only one copy of each parental (homologous) chromosome. The generation of haploid gametes relies on a single round of DNA replication followed by two successive rounds of cell division (meiosis I and meiosis II); meiosis I separates homologous chromosomes, whereas meiosis II separates sister chromatids, similar to mitosis (reviewed in Zickler and Kleckner, 2023). To enable this adaptation, meiosis relies on tightly regulated interactions between the homologous chromosomes. At the chromosomal level, transient, localized associations (‘pairing’) give rise to intimate end-to-end juxtaposition and ultimately to physical linkages, called chiasmata, that resist the pulling forces of the spindle (Kurdzo and Dawson, 2015). Chiasmata reflect exchanges of genetic information between the homologous chromosomes, which are the products of recombination-mediated repair of programmed DNA double-strand breaks (DSBs; reviewed in Lam and Keeney, 2014). This repair process initially generates reversible associations (repair intermediates) between the homologous chromosomes that, in some species, mediate pairing (reviewed in Zickler and Kleckner, 2015). Eventually, a subset of repair intermediates matures to become exchanges between homologous chromosomes (crossovers) that promote genetic diversity and enable proper chromosome segregation (reviewed in Hunter, 2015; Zickler and Kleckner, 2023).

Interactions between homologous chromosomes are enabled and regulated by a dedicated and evolutionarily conserved chromosome organization (Fig. 1). Early in meiosis, chromatin is organized as loops that are anchored along a structural rod called the axial element, which also serves as a platform for DSB formation and repair (Kumar et al., 2010; Panizza et al., 2011; Stanzione et al., 2016). Subsequently, a tripartite proteinaceous structure called the synaptonemal complex (SC) assembles between the axial elements of homologous chromosomes, aligning them and regulating the formation of crossovers (Colaiácovo et al., 2003; de Vries et al., 2005; Gowen, 1933; Higgins et al., 2005; Page and Hawley, 2001; Sym and Roeder, 1994). The SC is composed of two axial elements (called lateral elements following their incorporation into the SC) that run along the length of each homologous chromosome, stacked transverse filaments that connect the lateral elements, and a central element that seems to stabilize the transverse filaments (Bolcun-Filas et al., 2007, 2009; Collins et al., 2014; Gómez-H et al., 2016; Hamer et al., 2008; Humphryes et al., 2013; Page et al., 2008; Schramm et al., 2011). For simplicity, we will refer to axial and lateral elements as the chromosome axis. SC formation integrates local and long-range sequence information to align homologous chromosomes and sequences. In the budding yeast Saccharomyces cerevisiae (Agarwal and Roeder, 2000), mammals (Romanienko and Camerini-Otero, 2000), fish (Blokhina et al., 2019) and some plants (Grelon et al., 2001; Stacey et al., 2006), DSBs and their repair play a crucial role in assessing local sequence homology, whereas in other species, including Caenorhabditis elegans (Dernburg et al., 1998) and Drosophila females (McKim et al., 1998), homology is assessed independently of DSBs. Regardless of the mechanisms that establish local sequence homology, SC assembly initiates at sites of localized pairing, and this is followed by processive extension to span the length of the chromosomes (synapsis) (reviewed in Zickler and Kleckner, 2015).

Pairing and synapsis occur as chromosomes undergo motor-driven movements (reviewed in Koszul and Kleckner, 2009). These movements are facilitated by attachment of chromosomal sites (commonly the telomeres) to the nuclear envelope (NE). Within the NE, the conserved linker of nuclear cytoskeleton (LINC) complex serves as the bridge between telomeres and the cytoplasmic cytoskeleton that anchors the motor proteins responsible for chromosome movements (Agrawal et al., 2022; Horn et al., 2013; Kim et al., 2023; Lee et al., 2020; Spindler et al., 2019). A second conserved tethering complex, the TERB1–TERB2–MAJIN complex, contributes to anchoring of chromosomes in the NE in mammals (Dunce et al., 2018; Pendlebury et al., 2017; Shibuya et al., 2014, 2015; Wang et al., 2019; Zhang et al., 2017). Chromosome movements tend to result in the clustering of telomeres towards one side of the nucleus, thereby forming the telomere bouquet (Carlton et al., 2003; Chikashige et al., 1994; Conrad et al., 2007, 2008; Kosaka et al., 2008; Koszul et al., 2008; Scherthan and Adelfalk, 2011; Scherthan et al., 1996). Although tethering of chromosomes could potentially limit their mobility and ability to interact, it could also help promote interactions. Tethering telomeres (or pairing centers in C. elegans) to the NE limits the homology search volume and promotes chromosome pairing, an effect that could be further facilitated by active chromosome movements (Baudrimont et al., 2010; Chua and Roeder, 1997; Conrad et al., 1997, 2008; Koszul et al., 2008; Lee et al., 2012; Parvinen and Söderström, 1976; Penkner et al., 2009; Sato et al., 2009; Trelles-Sticken et al., 2000; Wanat et al., 2008; Wynne et al., 2012). Additionally, chromosome movements promote SC assembly (Penkner et al., 2009; Rog and Dernburg, 2015; Sato et al., 2009) and have been proposed to eliminate unwanted connections between nonhomologous chromosomes (reviewed in Koszul and Kleckner, 2009).

As the tethered chromosome ends move along the NE and chromosomes undergo pairing, DSB repair and synapsis, inappropriate topological configurations between unrelated chromosomes can occur, namely entanglements, entrapments and interlocks (Fig. 2). Here, we define entanglements as chromosomes, either paired or unpaired, that are twisted around one another but not topologically linked (Fig. 2A). Entanglements can involve two or more unrelated chromosomes. Chromosome entrapments occur when physical impediments, such as flanking SC or recombination intermediates, topologically lock unrelated chromosomes together. Interlocks constitute a subset of entrapments where the SC of one chromosome pair (‘bivalent’) flanks either side of an entrapped chromosome or bivalent (reviewed in Zickler and Kleckner, 2023). Interlocks can exist in two configurations: bivalent and chromosomal. A bivalent interlock arises when both chromosomes of an individual bivalent are entrapped within another partially synapsed bivalent (Fig. 2B). A chromosomal interlock occurs when only one chromosome of an asynapsed or partially synapsed bivalent is entrapped within a partially synapsed bivalent (Fig. 2C). A bivalent interlock prevents one bivalent from fully synapsing, whereas a chromosomal interlock prevents both bivalents from completely synapsing. Interlocks involving more than two bivalents are referred to as complex interlocks and can be an amalgamation of bivalent and/or chromosomal interlocks (Wang et al., 2009). Interlocks persisting into metaphase can lead to mis-segregation of the chromosomes and aneuploid gametes (called chromosome nondisjunction; Buss and Henderson, 1971). Additionally, interlocks prevent complete SC formation, which can impact the formation of crossovers (Libuda et al., 2013; Sym et al., 1993; Voelkel-Meiman et al., 2019). Such outcomes could compromise meiotic fidelity directly and might also impact fertility through the activation of meiotic checkpoints and apoptosis (reviewed in Subramanian and Hochwagen, 2014).

In this Review, we explore the nature of these topological chromosome configurations and their surprisingly common occurrence throughout different species (reviewed in von Wettstein et al., 1984). Notably, entanglements and entrapments are rarely observed in later stages of meiosis I, suggesting that they are efficiently resolved. We highlight the current body of research addressing mechanisms that minimize and resolve these inappropriate chromosome configurations and the topological considerations that underlie such mechanisms. We also provide a focus on zebrafish as a model to study topological chromosome configurations and C. elegans as a model for analyzing chromosome structure and dynamics.

Entanglements occur when one chromosome or pair of chromosomes twists around another. Since entanglements are, in principle, topologically open, they could both be formed and resolved via the cytoskeletal-led telomere movements described above, through Brownian (diffusive) motion, or as an indirect effect of nuclear motion (Conrad et al., 2008; Koszul et al., 2008; Newman et al., 2022; Nozaki et al., 2021). In silico modeling suggests that telomere movements within the NE, which significantly increase the rate of pairing (Marshall and Fung, 2016), could also untwist two unrelated chromosomes without altering chromosome architecture.

Several processes could serve to limit the prevalence of entanglements. Partial pairing of homologous chromosomes prior to meiotic entry, which has been observed in mouse (Boateng et al., 2013; Ishiguro et al., 2014; Solé et al., 2022), budding yeast (Burgess et al., 1999) and Drosophila (Rubin et al., 2022), could help to establish a less mixed configuration before chromosomes enter the more dynamic stages of meiotic prophase. In addition, the telomere bouquet could help limit entanglements prior to pairing and synapsis by reducing the range of possible configurations that chromosomes can adopt (Scherthan et al., 1994; Trelles-Sticken et al., 1999). The limited frequency of entanglements can also be facilitated by other non-random chromosome configurations, such as the homology-independent pairwise association of centromeres (centromere coupling) (Obeso and Dawson, 2010) or chromosome associations with components of the nuclear pore complex (Komachi and Burgess, 2022). The process of homolog pairing itself is potentially hazardous with respect to the formation of interlocks since it can cement topological configurations by promoting synapsis or by forming irreversible recombination intermediates (see below). However, if stabilization of chromosome interactions occurs in a stepwise fashion, then this could reduce interactions with unrelated chromosomes by gradually sampling the ‘pairing landscape’ to avoid getting trapped in local minima (Storlazzi et al., 2010). The topoisomerase Top3, an enzyme that catalyzes a single-stranded DNA break to relieve DNA topological constraints (reviewed in McKie et al., 2021), and its accessory factor Rmi1 also aid in resolving recombination-dependent entanglements (Hartung et al., 2008; Kaur et al., 2015; Tang et al., 2015).

Repulsive forces between the brush-like chromatin of meiotic chromosomes, as well as increased persistence length of chromosomes as a consequence of condensation and axis assembly (i.e. reduced bendability and increased stiffness), would bias entanglement resolution (Marko and Siggia, 1997). Interestingly, entanglements appear to also be limited in non-meiotic interphase cells (Tavares-Cadete et al., 2020), suggesting that general properties of chromosome organization disfavor entanglements, and by extension, that entanglements might pose a risk to somatic cells. One such property is the individualization of chromosomes into territories in interphase cells (Cremer and Cremer, 2001). Distinct chromosomal territories, however, do not prevent entanglements between sister chromatids – events that could lead to the formation of chromatin tethers linking chromosomes that have segregated to the daughter cells (anaphase bridges; Finardi et al., 2020). Finally, chromosome condensation, which helps to resolve sister chromatid entanglements in mitosis, limits chromosome segregation errors in meiosis (Chan et al., 2004).

Entanglements by themselves are unlikely to severely impact meiotic processes or subsequent chromosome segregation since the different chromosomes remain only loosely associated. Although simple entanglements may be resolved by condensation, more topologically complex configurations (such as those involving more than two chromosomes) could potentially impede separation at anaphase I.

The stabilization of transient pairing interactions through synapsis and formation of stable repair intermediates introduces topological concerns. This Review focuses on topologically entrapped chromosomes as they occur within the context of zygotene, the sub-stage of meiotic prophase at which the SC forms. Depending on the organism, SC assembly initiates at one end of a bivalent [in C. elegans (MacQueen et al., 2005; Rog and Dernburg, 2015)], at both ends of a bivalent [in zebrafish (Blokhina et al., 2019; Saito et al., 2014), human males (Brown et al., 2005) and barley (Higgins et al., 2012)] or at multiple locations along a bivalent [in yeast (Pyatnitskaya et al., 2022; Tsubouchi et al., 2008), mice (Boateng et al., 2013) and some plants (Hobolth, 1981; Holm, 1977; Moens, 1968)]. Since synapsis can initiate simultaneously at multiple locations on the chromosome, one or both axes of a bivalent can become topologically trapped between the two axes of another bivalent. This will result in an entrapment, and, upon further synapsis, in an interlock. Importantly, the relative position of telomeres or other anchors at the NE prior to SC assembly and the formation of recombination intermediates will precondition chromosomes for inevitable entrapment (Fig. 2A).

Interlocks were first observed in zygotene meiocytes of the flatworm Dendrocoelum lacteum over 100 years ago using light microscopy (Gelei, 1921). Interlocks have since been detected cytologically in many species including maize (Zea mays; Gillies, 1981), the silkworm moth Bombyx mori (Rasmussen, 1976), the basidiomycete Coprinus cinereus (also known as Coprinopsis cinera; Holm et al., 1981), the filamentous fungus Sordaria macrospora (Storlazzi et al., 2010), zebrafish (Danio rerio; Blokhina et al., 2019), white sturgeon (Acipenser transmontanus; Van Eenennaam et al., 1998) and humans (Bojko, 1983; Rasmussen and Holm, 1978). Interlocks can also be seen in mutant backgrounds of other species such as C. elegans (Link et al., 2018) and Arabidopsis thaliana (Martinez-Garcia et al., 2018) or by heat treatment in the locust Locusta migratoria (Buss and Henderson, 1971). The two-dimensional flattening that occurs during the preparation of a chromosome spread, a common technique to analyze meiotic chromosomes, can generate artifactual overlapping chromosomes that are difficult to differentiate from interlocks. However, three-dimensional reconstruction of serially or optically sectioned nuclei has confirmed that interlocks are bona fide topological entrapments (Hobolth, 1981; Holm et al., 1981; Rasmussen, 1976; Rasmussen and Holm, 1978; Wang et al., 2009; Zickler and Sage, 1981).

Interlocks were initially considered to be a rare event since they are uncommon in pachytene, a later sub-stage of meiotic prophase when synapsis is complete (reviewed in von Wettstein et al., 1984). How could interlock rarity in pachytene be explained? A potential explanation is the activity of surveillance mechanisms that monitor the progression of DSB repair and synapsis. Defects in either process activate checkpoints that slow meiotic progression and could eventually lead to apoptosis (or cell cycle arrest in yeast) (Bhalla and Dernburg, 2005; Bishop et al., 1992; Börner et al., 2004; Di Giacomo et al., 2005; Li and Schimenti, 2007; MacQueen and Villeneuve, 2001; Mitra and Roeder, 2007; Odorisio et al., 1998; San-Segundo and Roeder, 1999; Woltering et al., 2000; Wu and Burgess, 2006). Since interlocks impede the completion of synapsis, they could trigger a synapsis checkpoint or a DNA repair checkpoint as a consequence of a synapsis defect. However, a lack of interlocks in pachytene cannot be fully attributed to the elimination of nuclei with interlocks; in the fungus S. macrospora, more than 99% of meiocytes form viable spores despite the high prevalence of interlocks in zygotene, indicating that essentially all interlocks are efficiently resolved (Zickler and Espagne, 2016).

Following pachytene, the SC disassembles at the diplotene stage of meiotic prophase and homologs remain physically linked via chiasmata – the consequence of crossovers and sister chromatid cohesion – which are resolved only upon release of cohesion in anaphase (Anderson et al., 1999; Buonomo et al., 2000; Fu and Sears, 1973; Kanda and Kato, 1980; Martinez-Perez et al., 2008; Moens, 1978; Moens and Spyropoulos, 1995; Ogushi et al., 2021; Siomos et al., 2001). SC disassembly relieves some of the topological constraints, allowing some interlocks and entrapments to be resolved; however, interlocks that occur between two chiasmata would remain entrapped and could perturb chromosome segregation at meiosis I. Interlocking of two or more chromosomes at metaphase I has been reported in locusts, newts and plants (Buss and Henderson, 1971; Callan and Pearce, 1979; Gairdner and Darlington, 1931; Heslop-Harrison and Bennett, 1985; Mancino et al., 1979; Sax and Anderson, 1933; Yacobi et al., 1982). It is not clear how problematic these events are for proper chromosome separation at meiosis I, since the loss of sister chromatid cohesion at anaphase I essentially disrupts the entrapping chiasmata structures as chromosomes are pulled apart. Interlocks could potentially affect centromere orientation and consequently normal segregation (Heslop-Harrison and Bennett, 1985), though to our knowledge this has not been explored.

Physical features of meiotic nuclei and meiotic chromosomes could help to minimize entrapments. Despite harboring the same diploid genome, meiotic prophase nuclei have a larger volume than somatic nuclei in several plant species, mouse, grasshopper (Beasley, 1938) and S. macrospora (Zickler, 1977). Computational modeling predicts that fewer interlocks would form in larger nuclei (Navarro et al., 2022). Organisms with smaller and/or fewer chromosomes, such as fungi, appear to experience fewer interlocks in zygotene than those with longer chromosomes (Bojko, 1983; Holm and Rasmussen, 1980; Holm et al., 1981; Rasmussen and Holm, 1978, 1982; Zickler and Sage, 1981). As shown in the locust, long chromosomes form interlocks at higher frequency than shorter chromosomes (Buss and Henderson, 1971).

In silico modeling of chromosome pairing and interlock resolution has also shown that having only one end of a chromosome attached to the NE results in a considerable reduction in the occurrence of interlocked chromosomes relative to that observed in situations with two attached ends (Navarro et al., 2022). Similarly, staggered synapsis initiation, or synapsis initiation rates that are much slower than the extension rate, would decrease one source of topological constraints and, as a result, could decrease the transition from entanglements to interlocks. Interestingly, in C. elegans, synapsis initiates on a single site on each chromosome (MacQueen et al., 2005; Rog and Dernburg, 2015) and only one end of each chromosome is attached to the NE, leaving the other chromosome end free (Goldstein and Slaton, 1982). These features likely account for the near absence of interlocks in this species (Box 1). Nonetheless, perturbing the nuclear lamina structure in C. elegans leads to increased interlocks (Link et al., 2018).

Three general mechanisms have been proposed to resolve interlocks (Fig. 3) (Rasmussen and Holm, 1978). The first mechanism involves breaking and rejoining the proteinaceous structure of the axis and the DNA of both sister chromatids of the entrapping chromosome with the aid of topoisomerase II. The second involves desynapsis of the entrapping chromosomes to allow the entrapped chromosome to be released, and the third involves pulling the entrapped chromosome through the asynapsed ‘eye’ of the entrapping chromosome. Below we examine these three mechanisms in more detail.

It has been proposed that the breaking and rejoining of chromosomes to resolve interlocks could involve the activity of type II topoisomerases (Holm and Rasmussen, 1980). Type II topoisomerases generate a transient DSB in one DNA molecule to allow passage of a second DNA molecule before the break is resealed (Liu et al., 1980). Rampant topoisomerase II activity allows passage of tangled DNA molecules through one another, rendering DNA essentially oblivious to topological constraints. However, in prophase I, each chromosome is composed of two sister chromatids associated with a proteinaceous axis that cannot be accommodated by topoisomerase II. Therefore, interlock resolution involving topoisomerase II would need to occur via a multi-step process: sequential (and potentially coordinated) activity of topoisomerase II to pass up to four DNA molecules through another set of two DNA molecules coupled to local disassembly and reassembly of the axis (reviewed Zickler and Kleckner, 1999).

Topoisomerase II proteins localize to meiotic prophase chromosomes of budding yeast (Heldrich et al., 2020; Klein et al., 1992), the rooster Gallus gallus domesticus, (Moens and Earnshaw, 1989), mouse (Li et al., 2013), and the brassica plants A. thaliana and Brassica oleracea (Martinez-Garcia et al., 2018). Topoisomerase II is enriched at entangled regions of zygotene chromosomes of A. thaliana, and a hypomorphic topoisomerase II mutant (topII, also known as top2) has an increased incidence of chromosome interlocks in meiosis prophase I (57% compared to 18% in wild type); interlocks are not detected in wild-type nuclei at metaphase I, whereas interlocks are detected in 20% of nuclei from the hypomorphic topII mutant (Martinez-Garcia et al., 2018). Inhibitors of topoisomerase II in mouse spermatocytes lead to defects in chromosome condensation and segregation (Kallio and Lähdetie, 1996; Russell et al., 2004). Similarly, perturbing topoisomerase II in Schizosaccharomyces pombe (fission yeast) results in defects in chromosome separation at meiosis I, although the underlying reason for these defects remains unclear (Hartsuiker et al., 1998). We note that the role of topoisomerase II in various nuclear processes, including transcription, chromosome condensation and meiotic recombination, complicates the interpretation of these results (Heldrich et al., 2020; Joshi et al., 2012; Li et al., 2013; Zhang et al., 2014). No interlocks or abnormalities are seen in toposiomerase II mutants of budding yeast, though exit from pachytene is delayed (Rose and Holm, 1993).

Although it is unknown how the axes might be locally disassembled to accommodate the transit of the entrapped chromosome, discontinuities in the axes have been observed near interlocks and could represent transient axis disassembly (Blokhina et al., 2019; Bojko, 1983; Gillies, 1985; Hobolth, 1981; Holm and Rasmussen, 1980; Holm et al., 1981; Rasmussen, 1986; Rasmussen and Holm, 1978). Local removal of axis components could be achieved through the activity of the cohesin-accessory protein WAPL, which removes the axis component cohesin from chromosomes (Castellano-Pozo et al., 2023; Challa et al., 2019; De et al., 2014), or the AAA+ ATPase PCH2 (also known as TRIP13), which remodels HORMA domain-containing axis proteins (Börner et al., 2008; Chen et al., 2014; Lambing et al., 2015; Russo et al., 2023; Wojtasz et al., 2009); however, their involvement in interlock resolution has not yet been documented. Another possibility is that the chromosome axis is much less continuous than its appearance in fixed images and electron micrographs has led us to believe. In extensively spread meiotic chromosomes from C. elegans, axis proteins appear as linear arrays of foci (Woglar et al., 2020), suggesting that periodic gaps in the axis could allow for the passage of the entrapped chromosome. The recent discovery that cohesins are molecular motors capable of loop extrusion also supports the idea that the contiguity of the axis is more metastable than previously appreciated (Davidson et al., 2019; Kim et al., 2019). Cohesins, in addition to their role in sister chromatid cohesion, extrude chromatin loops, which could stack and align the base of the loops emanating from the axis. In worms, perturbing the loop-extruding activity of cohesins after the axis has already formed leads to axis disintegration (Castellano-Pozo et al., 2023). Such disintegration, if local and transient, could allow passage of one axis through another. Future studies to examine the dynamics of axis proteins in a variety of organisms, as well as efforts to reconstitute axis formation in vitro, are likely to shed light on physical properties of the axis and, in turn, on the challenges it poses for passage of another axis through it.

A second mechanism to resolve interlocks is the disassembly and reassembly of the SC (Rasmussen and Holm, 1978). There is ample evidence that the SC is labile; the integrity of the central region of the SC in worms, budding yeast and flies relies on weak hydrophobic interactions (Rog et al., 2017), and live imaging in C. elegans has shown the liquid-like properties of the central region of the SC, including the internal mobility of its subunits and its ability to undergo reassembly after reversible dissolution using hexanediol (Nadarajan et al., 2017; Pattabiraman et al., 2017; Rog et al., 2017) (Box 1). The SC can undergo post-assembly rearrangements to minimize topological complexities such as junctions, bubbles and asynapsed axes, through a process known as synaptic adjustment, as has been observed in mouse (Moses and Poorman, 1981), humans (Guichaoua et al., 1986), the mealmoth Ephestia kuehniella (Weith and Traut, 1986), C. elegans (Henzel et al., 2011), the lizards Sceloporus graciosus and Sceloporus undulatus (Reed et al., 1990), and the frog Leptodactylus pentadactylus (Noronha et al., 2020). Synaptic adjustment might not be universal, however, as it has not been observed in maize (Anderson et al., 1988; Maguire, 1981) or Drosophila (Gong et al., 2005).

Crossovers locally deform the SC (Libuda et al., 2013; Pattabiraman et al., 2017; Woglar and Villeneuve, 2018), which might reflect a general response of the SC to spatial hindrances, such as crosslinked homologs or the large aggregates of pro-crossover proteins known as recombination nodules (Carpenter, 1975). Such a response could promote SC disassembly specifically near interlocks. In C. elegans, the SC locally disassembles following exogenous DNA damage (Couteau and Zetka, 2011). In zebrafish (Box 2) and B. mori, individual pairs of chromosomes often show extensive or complete asynapsis, sometimes with another bivalent entrapped in the asynapsed region (Rasmussen, 1986) (Fig. 2). If these are indeed transient interlock resolution intermediates, then fully synapsed chromosomes found in pachytene nuclei would be created through a second round of synapsis.

The dynamic nature of the SC can also be inferred from live-cell imaging studies. Measuring kinetics of SC assembly in budding yeast has identified instances of ‘abortive’ SC disassembly (Pollard et al., 2023); the transient nature of these events would allow for the resolution of an interlock (Navarro et al., 2022). Live-cell imaging of yeast has also shown that in late prophase, when synapsis is complete, fluorescently tagged homologous loci undergo multiple cycles of separation and colocalization. Foci are found to be spaced 750 nm apart on average but are occasionally separated by up to 2.0 µm (which is approximately the diameter of the yeast nucleus; Newman et al., 2022). Considering a loop size of ∼0.5 µm, based on electron microscopy images of yeast chromosome spreads (Møens and Pearlman, 1988), and the width of the SC as ∼100 nm (Schmekel, 2000), distances between foci would be expected to reach a maximum separation of ∼1.1 µm. A separation of 2.0 µm requires axis separation due to transient desynapsis or wholesale disruptions in integrity of the axis. Brief separation of terminal chromosomal ends has been observed in time-lapse imaging of C. elegans meiocytes (Wynne et al., 2012), although the synapsis status of these chromosomes was not addressed. Terminal separation of axes in C. elegans has also been observed in fixed samples (Mlynarczyk-Evans et al., 2013) and in pachytene chromosomes of zebrafish (Box 2), mouse (Zwettler et al., 2020), B. mori spermatocytes (Rasmussen, 1986) and the lamprey Lampetra fluviatilis (Matveevsky et al., 2023).

Transient SC disassembly can only help to remove interlocks if there are no downstream inter-homolog repair intermediates that crosslink the two homologs, which would impede interlock migration. One mechanism to minimize crosslinks is limiting the number of crossover precursors. This can occur via crossover interference, a phenomenon where a crossover reduces the likelihood of an adjacent crossover forming on the same bivalent by channeling nascent inter-homolog repair intermediates towards non-crossover outcomes (Pazhayam et al., 2021). Correspondingly, weaker crossover interference entails more crossovers and more crosslinks that could impede interlock resolution. Work in S. macrospora and rice has shown that absence of the endonuclease Mlh1 leads to an increase in interlocks (Storlazzi et al., 2010; Xin et al., 2021). This might be due to excess crossovers, which Mlh1 usually disassembles into non-crossover products (Storlazzi et al., 2010). Additionally, the Bloom syndrome helicase (BLM) might facilitate interlock resolution through its anti-crossover activity (Holloway et al., 2010; Rockmill et al., 2003). In mice, BLM localizes near the SC during late prophase, hinting at a potential role in interlock resolution (Walpita et al., 1999). However, whether absence of BLM leads to increased interlocks has not been reported.

The third mechanism of interlock resolution is specific for bivalent interlocks. In this mechanism, one end of the interlocked bivalent is detached from the NE and passed through the asynapsed segment of the interlocking bivalent (Rasmussen and Holm, 1978). Detachment could be coordinated with transient SC disassembly (see above) to first nudge the interlock towards the telomeric end of the chromosomes and then resolve via brief telomere detachment (Navarro et al., 2022). Evidence for this mechanism comes from the lily (Lilium longiflorum), where several telomeres are detached from the NE (Holm, 1977). However, the feasibility of resolving interlocks in physiological timescales is unclear. Telomere detachments are rarely observed in fixed nuclei from several species (Bass et al., 1997; Elkouby and Mullins, 2017; Moens, 1969; Rasmussen, 1976; Rasmussen and Holm, 1978; Wang et al., 2019), suggesting that they are quite infrequent and/or transient. Live-cell imaging will be an important tool to shed light on the prevalence of such a mechanism in vivo.

Importantly, motor-driven chromosome movements are likely to be essential to allow all three mechanisms to resolve interlocks in physiologically relevant timescales (Koszul and Kleckner, 2009). Indeed, a reduction in motor-driven movements in maize (Golubovskaya et al., 2002), C. elegans (Link et al., 2018) and A. thaliana (Martinez-Garcia et al., 2018) is linked to the accumulation of interlocks.

The cell may also employ ‘last resort’ enzymes to process unresolved DNA links that persist to anaphase. For example, the mammalian nuclease ANKLE1 (LEM-3 in C. elegans) cleaves branched DNA species to remove chromatin bridges at late stages of meiosis and mitosis (Brachner et al., 2012; Dittrich et al., 2012; Hong et al., 2018a,b; Song et al., 2020).

Although interlocks have been widely documented across species, mechanisms to remove interlocks remain poorly understood. Recent progress in microscopy promises to provide unprecedented insight into meiotic structures (Blokhina et al., 2019; Cahoon et al., 2017; Ding et al., 2021; Köhler et al., 2017; Kubalová et al., 2023; Prakash et al., 2015; Wang et al., 2009; Xu et al., 2019; Yoon et al., 2018; Zwettler et al., 2020). Advanced fluorescence microscopy can be used to compare SC and axis structures before and after a desynapsis event, identify proteins localizing to interlock boundaries, and visualize potentially labile structures that allow transient detachment of telomeres from the inner NE or create transient gaps in the axis to transit one chromosome through another. Advances in live imaging and in vitro reconstitution of SC components will shed light on the dynamic features of meiotic prophase that are likely to play important roles in interlock resolution, including kinetic analysis of SC depolymerization and repolymerization (Pollard et al., 2023; Rog et al., 2017). Finally, advances in computational modeling can uncover propensities for DNA molecules to become entangled and/or form interlocks (Brahmachari and Marko, 2019; Navarro et al., 2022; Pouokam et al., 2019).

Although the basic premise of meiosis – to segregate chromosomes into haploid gametes – is universal, thematic motifs, including DSB formation, chromosome pairing, the formation and disassembly of the SC, crossing over, and cytoskeleton-mediated chromosome movements, occur in different orders and might have different functions among eukaryotes. Model genetic organisms are particularly important for mechanistic studies by offering a rich genetic toolkit. However, it is important to remember that one species might rely on one mechanism more than another, which might influence interpretation of the effects of genetic and other perturbations.

In considering interlocks, particularly important are the seemingly conflicting roles played by the SC. At first glance it might appear that the SC causes interlocks by assembling between transiently paired chromosomes and by promoting crossover formation – two processes that lock in topologically enclosed configurations. However, the SC might also help to avoid and resolve entanglements and interlocks through modulation of chromosome stiffness, progressive assembly and negative regulation of crossovers. Future studies that consider these potential effects of perturbing the SC could shed light on these important functional contributions to meiosis.

We thank scientific illustrator Maria Diaz de la Loza for graphical work.