I've been reading about cohesin lately, and I'm confused about the head subunit interactions. I've read a few papers, and also found this nice figure from wiki that demonstrates the crux of my question (note: SMC here corresponds to Cohesin, or am I mistaken? I'm pretty sure this is correct, or pretty much correct):
From reading around, I'm aware generally the head subunit is an ATPase, yet from this diagram (and from the typical putative "use" of Cohesin in TAD maintenance), I'm not sure what exactly is the ATPase motif so useful for, it seems that the ATP is sort of "glueing" the the two head domains together, but that seems opposite of what ATPase would typically do, ie catalyze ATP -> ADP. This makes me think I'm just misunderstanding something, and wondering what is the function of the Walker A/B ATPase motif.
Also, I'm wondering if there are any nice PDB files that show a cohesin loop, or elsewise are generally the most instructive about Cohesin. I've searched through a few, but would like more informed advice about some good structures, and maybe why they are nice.