Asymmetric ATP Hydrolysis Kinetics of ABCE1 Explained with a Markov State Model

Abstract

The eukaryotic and archaeal ATPase ABCE1 is an essential factor in ribosome recycling. As a member of the ATP Binding Cassette (ABC) superfamily, it has two highly similar homologues of a characteristic nucleotide binding domain, which are antiparallelly aligned exposing two nucleotide binding sites at their interface. Binding and hydrolysis of nucleotides drives the nucleotide binding domains to adopt either an open or a closed conformation. Kinetic measurements of two ABCE1 variants with an identical mutation in each binding site, inhibiting ATP hydrolysis at the respective binding site, resulted in a kinetically less-active ABCE1 in one case (E238Q) and, unexpectedly, in a 10-fold hyperactive ABCE1 in the other case (E485Q) compared to the wild type (WT) [Barthelme et al., 2011]. This finding suggested some kind of direct or allosteric interaction between the two binding sites, where the fast binding site hydrolyzing ATP in the E485Q mutant is assumed to be activated by an ATP bound in the opposite slow binding site. We asked if such direct interaction is actually required to explain the available experimental data. To this end, we searched for Markov state models (MSM) of the WT and the two mutants with kinetics matching the experimental ones using a Monte Carlo scheme. We found such MSMs and identified the unbinding of ADP from the slow binding site to be the rate-limiting step in the WT and the E238Q mutant. Inhibition of ATP hydrolysis at this slow binding site of the E485Q mutant enables ABCE1 to shortcut this rate-limiting step resulting in its hyperactivity. Although it remains unknown if the binding sites of ABCE1 are allosterically coupled or not, this work shows that explaining the asymmetric kinetics without any allosteric effect is possible.