Nevertheless, Mcd1p presence from S phase through mid M is dependent upon its binding to both Smc1p and Smc3p as Mcd1p is rapidly degraded if either Smc1p-AID or Smc3p-AID are depleted (?amdere et al., 2015; Guacci et al., 2015). that in vivo all the regulators of cohesin promote DNA binding of cohesin by mechanisms independent of opening this gate. Furthermore, we show that this interface has a gate-independent activity essential for cohesin to bind chromosomes. We propose that this interface regulates DNA entrapment by controlling the opening and closing of one or more distal interfaces formed by cohesin subunits, likely by inducing a conformation change in cohesin. Furthermore, cohesin regulators modulate the interface to control both DNA entrapment and IV-23 cohesin functions after DNA binding. function (Chan et al., 2012; Beckou?t et al., 2016). These experiments led to an exit gate model whereby DNA can escape cohesin entrapment by Wpl1p-dependent opening the Smc3p Mcd1p interface. Smc3p acetylation would inhibit Wpl1p, thereby keeping the interface closed and DNA topologically bound (Chan et al., 2012; Beckou?t et al., 2016). Thus, the Smc3p Mcd1p interface has been postulated to serve as a regulated DNA exit gate and as well as a regulated entrance gate. While compelling in its simplicity, IV-23 viewing the function of the Smc3p Mcd1p interface and cohesin regulators solely through the lens of a putative DNA gate has been challenged by several observations. First, cohesin is stably bound to DNA when Smc3p K112 K113 acetylation is prevented by mutating the critical lysines ((Unal et al., 2008; Rowland et al., 2009). Thus, acetylation of these residues is not needed to stabilize topological entrapment of DNA in vivo. However, both and mutants exhibit a dramatic defect in sister chromatid cohesion (Rowland et al., 2009; Sutani et al., 2009; Chan et al., 2012; Guacci and Koshland, 2012). These results suggest that acetylation promotes cohesion by an additional step beyond preventing DNA release through an exit gate. Finally, recent experiments suggest that the function of cohesin and other Smc complexes requires a conformational change of the coiled coil, likely driven by the head ATPase activity (Soh et al., 2015). The binding of Mcd1p to both the head domain of Smc1 and the coiled coil of Smc3 provides a potential mechanism to transduce the ATP state of the head domain to initiate a conformation change of the coiled coils. Here, we use a fusion protein of Smc3p and Mcd1p to permanently shut the putative DNA gate as a tool to evaluate gate-independent functions of this interface. We examined whether the fusion could suppress the need for the cohesin loader Scc2p, the Scc3p cohesin subunit, Pds5p and Smc3p K112 K113 acetylation. Our results show that in vivo all these regulators of cohesin promote DNA binding or cohesion by mechanisms independent of opening the Smc3p Mcd1p interface. Furthermore, mutations altering the interface reveal it has a gate-independent activity SIR2L4 essential for cohesin to bind DNA in vivo. These observations suggest new models for the Smc3p Mcd1p interface in topological entrapment and as a target of cohesin regulators. Results We wanted to assess whether cohesin regulators control cohesin function by modulating the transient opening of the Smc3p Mcd1p interface. To do so, we used a previously characterized gene fusion in which the open-reading frame (ORF)?of was placed in frame at the end of the ORF?for (Gruber et al., 2006). The product of this gene fuses the Mcd1p N-terminus to the Smc3p C-terminus so this putative DNA gate cannot open. It supports viability as sole source of both Smc3p and Mcd1p. Consequently, regulators that act solely by stabilizing the exit gate should no longer be needed for cohesin function in fusion bearing strains. We IV-23 used the Smc3p Mcd1p fusion to test gate-dependent functions of three proteins associated with the trimer cohesin, Scc2p and Pds5p regulators as well as the the Scc3p cohesin subunit. Scc3p is required for cohesin binding to DNA, cohesion and condensation (Tth et al., 1999; Roig et al., 2014; Orgil et al., 2015). Since Scc2p, Scc3p and Pds5p are subunits of complexes that bind near the interface (Rowland et al., 2009; Orgil et al., 2015), they are in a position to modulate the Smc3p Mcd1p interface to control its dissociation to.