Mechanistic implications of the interaction of the soluble substrate-binding protein with a type II ABC importer

Abstract

ATP-Binding Cassette (ABC) Transporters employ homologous ATPase domains to drive transmembrane transport of diverse substrates ranging from small molecules to large polymers. Bacterial ABC importers require an extramembranous substrate binding protein (SBP) to deliver the transport substrate to the extracellular side of the transporter complex. Previous studies suggest significant differences in the transport mechanisms of type I vs. type II bacterial ABC importers, which contain unrelated transmembrane domains. We herein use ensemble fluorescence resonance energy transfer (FRET) experiments to characterize the kinetics of SBP interaction in the E. coli BtuCD-F complex, a canonical type II ABC importer that transports vitamin B12 . We demonstrate that, in the absence of B12 , BtuF (the SBP) forms a 'locked' (kinetically hyper-stable) complex with nanodisc-reconstituted BtuCD that can only be dissociated by ATP hydrolysis, which represents a futile reaction cycle. Notably, no type I importer has been observed to form an equivalent locked complex. We also show that either ATP or vitamin B12 binding substantially slows formation of the locked BtuCD-F complex, which will limit the occurrence of futile hydrolysis under physiological conditions. Mutagenesis experiments demonstrate that efficient locking requires concerted interaction of BtuCD with residues on both sides of the B12 binding pocket in BtuF. Combined with the kinetic inhibition of locking by ATP binding, these observations imply that the transition state for the locking reaction involves a global alteration in the conformation of BtuCD that extends from its BtuF binding site in the periplasm to its ATP-binding sites on the opposite side of the membrane in the cytoplasm. These observations suggest that locking, which seals the extracellular B12 entry site of the transporter, may help push B12 through the transporter and directly contribute to the transport mechanism in type II ABC importers.

Realtime observation of ATP-driven single B12molecule translocation through BtuCD-F

“…At saturation, binding of B12 quenches BtuF* intensity by ~54% in ensemble FRET-Q experiments in the absence of BtuCD (Fig. 1a), and fitting of fluorescence quenching curves 34 (e.g., as shown in Fig. 1a) yields a dissociation constant (Kd) of ~7.9 nM for the BtuF*•B12 binding reaction, matching previous reports 19,34,43 .…”

“…The corresponding 2 nM Kd (given by the ratio of krelease /kbinding ) represents a reasonable match 44,45 44,45 to the value of ~8 µM given by ensemble experiments conducting using several methods 19,34,43 (Fig. 1a).…”

“…S3a). Based on the close match of these values to those in our ensemble FRET experiments 34 (Fig. 1a), we interpreted the ~0.46 state to be the B12-bound state of BtuF*(BtuF*•B12), the ~1.0 state to be the B12-free state (apo BtuF*), and the transitions between them to reflect spontaneous binding and release of B12 by BtuF*.…”

“…S3b). Our ensemble FRET-Q experiments 34 demonstrate that the B12 binding site in BtuF* is inaccessible in the locked BtuCD-F* complex that can only be dissociated by Mg•ATP hydrolysis. Therefore, we infer that formation of the H state involves movement of B12 away from a cavity in the TMD in BtuCD concomitant with sealing of BtuF* on the periplasmic surface of BtuCD to form the locked BtuCD-F* complex.…”

“…We hypothesized that these that these fluctuations, which occur at a much higher rate in the presence of Mg•ATP than in the absence of nucleotide or the presence of non-hydrolyzable analogs (Fig. 1e FRET-Q and anisotropy 34 , the L state represents a dynamic equilibrium between free BtuF*•B12 and the pre-transport BtuF*•B12•BtuCD complex prior to movement of B12 into BtuCD (states 1-2 on the left in Fig. 4 below).…”