Specific Smc kleisin interactions have hitherto been detected in
Specific Smc/kleisin interactions have hitherto been detected in vivo using a bi-functional thiol-specific cross-linking reagent, BMOE, to induce rapid cross-linking between cysteine residue pairs inserted within individual ring interfaces (Gligoris et al., 2014). The results of these experiments imply that about 25% of cohesin rings are cross-linked simultaneously at all three Smc1/3 hinge, Smc3/Scc1, and Smc1/Scc1 interfaces (Gligoris et al., 2014). Chemical closure in this manner can then be exploited to detect DNA entrapment. Thus, entrapment of individual circular DNAs by chemically circularized cohesin rings leads to a modest retardation in their migration during gel electrophoresis even when all proteins have been denatured by heating in the presence of SDS (Haering et al., 2008). Likewise, co-entrapment of monomeric sister DNAs within chemically circularized cohesin causes them to migrate as dimers instead of monomers. Because they are catenated exclusively by cohesin rings, these sister DNA pairs are known as catenated dimers (CDs) (Gligoris et al., 2014). Analysis of numerous mutants has revealed a perfect correlation between the incidence of CDs and whether cells had established sister chromatid cohesion (Srinivasan et al., 2018). Thus, co-entrapment of sister DNAs within individual cohesin rings provides a mechanistic explanation for cohesion and for how cleavage of Scc1 by separase triggers sister chromatid disjunction at Tacrolimus (Uhlmann et al., 2000).
These studies have not hitherto taken into account the possibility that DNAs are entrapped within the ring’s sub-compartments, namely S or K compartments associated with E or J heads. Indeed, it has been proposed on numerous occasions that cohesion is in fact conferred by entrapment within E-S compartments and that the interconnection of E heads by kleisin merely reinforces this entrapment (Elbatsh et al., 2016, Huber et al., 2016, Li et al., 2017, Murayama et al., 2018, Murayama and Uhlmann, 2015, Stigler et al., 2016, Uhlmann, 2009, Uhlmann, 2016). Support for this notion stems from the observation that abolition of Smc3 de-acetylation by inactivation of the HOS1 de-acetylase delays sister chromatid disjunction during anaphase despite efficient Scc1 cleavage (Li et al., 2017). If DNAs were in fact entrapped within E-S compartments, then cleavage of their coiled coils by separase should suppress the delayed disjunction, which is precisely what was found. Entrapment within the S or K compartments of complexes whose heads are engaged is likewise consistent with the claim that cohesion can be established by viable Smc1D1164E mutations that are supposedly incapable of hydrolysing ATP (Çamdere et al., 2015, Çamdere et al., 2018, Elbatsh et al., 2016).
Discussion Sister chromatid cohesion is a feature of chromosome segregation that is universal among eukaryotes and a property that distinguishes them from bacteria. An important clue regarding the mechanism was the finding that the Scc1, Smc1, and Smc3 subunits of the cohesin complex responsible bind each other in a pairwise manner to create a huge tripartite ring whose cleavage by separase triggers the dissolution of cohesion at anaphase. This raised the possibility that cohesin holds sister DNAs together using a topological principle, namely co-entrapment of sister DNAs inside the tripartite SK ring formed by the binding of Scc1’s N- and C-terminal domains to the necks and heads of Smc3 and Smc1 that are themselves associated via their hinges (Haering et al., 2002, Haering et al., 2008). Though previous thiol-specific cross-linking studies have confirmed the entrapment of sister minichromosome DNAs inside such rings (Gligoris et al., 2014), they have not hitherto taken into account another key feature of Smc/kleisin complexes, namely that the ATPase heads at the vertices of V-shaped Smc1/3 heterodimers can themselves interact. Such interactions divide the ring into two compartments, an S compartment created by association of Smc1 hinges and heads with equivalent domains within Smc3 and a K compartment created by juxtaposition of Smc1 and Smc3 heads that are also associated with the C- and N-terminal domains of Scc1 (Arumugam et al., 2003). Cross-linking studies using a series of novel cysteine pairs described here show that cohesin’s Smc1 and Smc3 ATPase heads in fact associate in two very different modes, the canonical one that sandwiches a pair of ATP molecules between engaged heads (E) and another in which the heads are rotated and translocated in a fashion that juxtaposes their signature motifs in the absence of ATP (J). The transition between these two states driven by the binding and hydrolysis of ATP, first discovered in bacteria (Diebold-Durand et al., 2017), may be a universal feature of Smc/kleisin complexes (Bürmann et al., 2019). There are accordingly two types of S and K compartments, those associated with J or E heads.