Structure of the ChbC dimer, as viewed from within the plane of the membrane. [PDB: 3QNQ]
Sugar uptake into bacteria is mediated by phosphoenolpyruvate-dependent phosphotransferase systems (PTSs) that transport saccharides across the cell membrane and phosphorylate them before their release into the cytosol. Phosphorylation of the sugar ensures its unidirectional flux and primes it for consumption through metabolic cycles that generate energy. PTSs consist of enzyme I and heat-stable phosphocarrier protein, which are general energy-coupling proteins, and the sugar-specific enzyme II (EII), a complex consisting of the cytosolic proteins EIIA and EIIB and the integral membrane protein EIIC. The phosphate group originates from phosphoenolpyruvate and is transferred sequentially by components of the PTS and eventually to the incoming sugar substrate bound to EIIC, which translocates the substrate and assists its phosphorylation. To provide insight into the substrate selectivity, phosphorylation and translocation mechanisms, Cao et al. (PSI NYCOMPS) determined the crystal structure of an EIIC, the ChbC homologue from Bacillus cereus, which transports diacetylchitobiose (GlcNaAc)2.
Heterologously expressed and purified protein forms a stable dimer and, once reconstituted into proteoliposomes, was shown to transport sugars with a preference for (GlcNaAc)2. Each ChbC protomer consists of ten transmembrane domains (TMs), two oppositely oriented re-entrant hairpin structures and two horizontal amphipathic α-helices. The structure of the dimer has the shape of a capsized canoe, with a concave surface facing the intracellular side. The extensive, largely hydrophobic dimer interface is formed primarily by transmembrane domains in the N-terminal half of ChbC, whereas the C-terminal half contains a deep, electronegative cleft on its intracellular side. This cleft is lined partly by the re-entrant hairpin loops, one of which contains a highly conserved glutamic acid residue. Interestingly, the arrangement of the hairpin loops resembles two re-entrant hairpin loops in the otherwise dissimilar structure of the glutamate transporter GltPh.
Although it was impossible to unambiguously determine the orientation of the (GlcNaAc)2 molecule in the protomer structure, the authors were able to model the phosphorylation active site comprising the conserved glutamic acid residue and a conserved histidine residue, together with the sixth-position hydroxyl group of the nonreducing sugar (C6-OH), which is known to be phosphorylated in Escherichia coli ChbC. However, the catalytic mechanism remains currently unknown. Modeling of different sugar substrates into the binding pocket suggests that it is well suited for (GlcNaAc)2, although it may be able to accommodate a trisaccharide as well.
The location of the sugar-binding pocket suggests that the structure represents the occluded state, implying the existence of at least two additional states, the outward-open state, which can bind substrate from the periplasm, and the inward-open state, which interacts with EIIB to phosphorylate and release the substrate into the cytoplasm. The loop between TM4 and TM5 seals off the cavity to the intracellular side and could potentially move away to expose the bound substrate, which makes it an ideal candidate for the intracellular gate. The similarities between the transport domain of GltPh and the C-terminal region of ChbC could provide further insight into how the occluded state might convert to an outward-open state by a rigid-body motion, facilitated by the extracellular loop between TM7 and TM8. Further structural studies will be necessary to shed light on these other transport states.