In bacteria, adenosine 5′-triphosphate (ATP)–binding cassette (ABC) importers are essential for the uptake of nutrients including the nonreducing disaccharide trehalose, a metabolite that is crucial for the survival and virulence of several human pathogens including Mycobacterium tuberculosis. SugABC is an ABC transporter that translocates trehalose from the periplasmic lipoprotein LpqY into the cytoplasm of mycobacteria. Here, we report four high-resolution cryo–electron microscopy structures of the mycobacterial LpqY-SugABC complex to reveal how it binds and passes trehalose through the membrane to the cytoplasm. A unique feature observed in this system is the initial mode of capture of the trehalose at the LpqY interface. Uptake is achieved by a pivotal rotation of LpqY relative to SugABC, moving from an open and accessible conformation to a clamped conformation upon trehalose binding. These findings enrich our understanding as to how ABC transporters facilitate substrate transport across the membrane in Gram-positive bacteria.
Trehalose (α-d-glucopyranosyl-α-d-glucopyranoside) is an essential nonreducing disaccharide used in bacteria, fungi, parasitic nematodes, insects, and plants, but it is not used in vertebrates (1–5). It has several functions including as a source of carbon and energy, as a signaling molecule to control metabolic pathways, and also to protect cellular membranes and proteins from denaturation or inactivation caused by a variety of stress conditions including desiccation, heat, dehydration, cold, and oxidation (3, 5, 6). Trehalose is crucial for the survival and virulence of several pathogens including Clostridium difficile (C. difficile) and Mycobacterium tuberculosis (Mtb) (7, 8).
Mycobacteria cause many diseases in humans, including tuberculosis (TB) (9). A metabolic difference between mycobacteria and humans is the requirement of trehalose, an integral component of various glycolipids that are important cell wall structures and thus vital for the survival and virulence of these bacteria (5, 10–12). In these glycolipids, trehalose monomycolate (TMM) and trehalose dimycolate (TDM; cord factor) are the best known and are of vital importance for these pathogens (12–15). TMM is the precursor of TDM that is the chief immunostimulatory component of the Mtb cell wall that includes peptidoglycan, arabinogalactan, mycolic acids, and glycolipids (13, 16). TMM also serves as a transport form for mycolic acids, a key component of the outer membrane for Mtb (10, 17). The outer membrane of Mtb has some differences compared to that found in Gram-negative bacteria. These include the presence of arabinogalactan-bound mycolic acids and noncovalently linked solvent-extractable lipids that intercalate into the mycolate layer including TMM, TDM, diacyltrehalose, polyacyltrehalose, phthiocerol dimycocerosate, and sulfoglycolipid (18). Notably, during the process of mycolic acid and TDM synthesis, the trehalose moiety is released through mycolyltransferase antigen 85 complex catalysis at the periplasmic space (19–21). Important questions are whether there is a recycling system that mediates the retrograde transport of released trehalose for effective use and whether this varies substantially for the survival of different pathogens, especially in a harsh environment and nutritionally deficient conditions. The trehalose specific importer LpqY-SugABC has been identified in mycobacteria including Mtb, and it is believed to be responsible for recycling trehalose from the periplasm to cytoplasm (7). The transporter LpqY-SugABC and mycolic acid transporter MmpL3, a drug target for the development of anti-TB drugs, form a complete trehalose transport circuit in mycobacteria (7, 10, 17, 22). Notably, disruption of trehalose recycling strongly impairs virulence of Mtb (7). In addition, this system is also essential for Mtb growth in human macrophages as well as in mice (23, 24). Therefore, the trehalose importer LpqY-SugABC is a potential therapeutic drug target. However, the molecular basis as to how import is achieved is yet to be resolved.
The trehalose importer LpqY-SugABC, containing five subunits, belongs to the adenosine 5′-triphosphate (ATP)–binding cassette (ABC) transporter superfamily (7, 25). In these transporters, functionality is driven by ATP hydrolysis to allow substrates to transport across the membrane (22). It is therefore suggested that LpqY-SugABC could have a similar mechanism. In LpqY-SugABC, it has been proposed that the two SugC subunits form the nucleotide-binding domains (NBDs) that bind and hydrolyze ATP during transport, while SugA and SugB have helical transmembrane domains that form a translocation pathway for trehalose. Furthermore, it has been proposed that LpqY, which is the counterpart to maltose-binding protein (MBP) in the Escherichia coli MBP-MalFGK2, functions to deliver trehalose to the transporter (26). It has also been suggested that LpqY may control and trigger conformational changes required for trehalose transport (7, 27). Unlike MBP, LpqY is a lipoprotein with an N-terminal lipid modification (7). It is proposed that this feature may help it anchor to the membrane during trehalose transport. Moreover, substrate-binding proteins (SBPs) with lipid modifications are ubiquitous in Gram-positive bacteria (28), which implies that they have multiple and varied functions.
Herein, we have used single-particle cryo–electron microscopy (cryo-EM) to determine high-resolution structures of the trehalose-specific importer LpqY-SugABC from Mycobacterium smegmatis (M. smegmatis), which shares 76% sequence identity with its Mtb counterpart, in four different states. These structures provide the first example of how mycobacteria can transport substrates across the membrane by using a type I ABC importer. On the basis of these structures, we are able to delineate the structural basis for uptake and release by this importer. The structures presented here greatly enhance our understanding of the mechanism of ABC transporters in general and facilitate our understanding of substrate transport across membranes in Gram-positive bacteria. Furthermore, these findings will greatly inform the design and development of trehalose analogs as potential antibacterial drugs.
SugABC and LpqY-SugABC (E164Q mutant, which is inactive) were overexpressed in M. smegmatis (mc2155) and then purified to homogeneity (see Materials and Methods). They were both purified in N-dodecyl-β-d-maltoside (DDM) and lauryl maltose neopentyl glycol (LMNG), respectively. Before cryo-EM studies, the two samples were exchanged into A8-35 and glyco-diosgenin (GDN) detergent, respectively (fig. S1, A and B). SugABC has adenosine triphosphatase (ATPase) activity in A8-35, while it is inactive in GDN, which indicates that detergents strongly influence ATPase activity (fig. S1C). This phenomenon is also observed in other ABC transporters such as ABCA1 (29), LptB2FGC (30), and cystic fibrosis transmembrane conductance regulator (31). To obtain the structures of the four different complexes, we prepared samples of (i) SugABC in the absence of LpqY, (ii) the LpqY-SugABC (E164Q) in complex with no bound ligands, (iii) LpqY-SugABC (E164Q) with trehalose alone, and (iv) LpqY-SugABC (E164Q) with trehalose, Mg2+, and ATP. The cryo-EM maps for the four complexes have resolutions in the 3.3- to 3.8-Å range (fig. S2 and table S1). These maps revealed densities for the side chains in the transmembrane domains, most side-chain and secondary structural elements in NBDs, as well as the locations of trehalose, ATP, and the Mg2+ ions (fig. S3).
Overall structure of SugABC, preresting state
The structure of SugABC was determined at 3.4-Å resolution (table S1). The four subunits of SugABC assemble as observed in a typical type I ABC importer adopting an inward-facing orientation with the TM trehalose–binding site exposed to the cytoplasmic side of the membrane (Fig. 1A). SugA and SugB have a combined total of 12 transmembrane helices to form a substrate translocation pathway (Fig. 1A). The TM1 helix of SugA (or SugB) crosses over the dimer interface and packs against the TM2 to TM6 helical bundle of SugB (or SugA) (Fig. 1, A and B). The observed helix arrangement of SugAB is analogous to the maltose transporter MalFGK2 (32) and molybdate/tungstate importer ModBC (33) but is different to the side-by-side arrangement observed in MetNI (fig. S4A) (34). Although the two polypeptide chains of SugA and SugB share only a low sequence identity (18%), they have similar topologies. Pseudo-twofold symmetry is observed at the center of the interface of SugA and SugB, with SugA related to SugB along an axis perpendicular to the membrane bilayer [root mean square deviation (RMSD) of 2.33 Å after superimposition of 217 Cα atoms] (Fig. 1, B to D). SugC contains two subdomains: an NBD and a regulatory domain (Fig. 1A). In the SugABC complex, the SugC dimer is in an open configuration and is bound to the transmembrane region primarily through contacts with EAA (EAA-X-(3)-G) loops (26, 35) of SugA and SugB, both of which contain two short helices (Fig. 1, A and D). One of the two helices, identified as the coupling helix (Fig. 1A) (36), inserts into the surface cleft of each SugC NBD and is stabilized by numerous hydrophilic and hydrophobic interactions (Fig. 1A and fig. S7A).
SugABC in complex with ligand-free LpqY: A unique resting state structure
Next, we determined the structure of SugABC (E164Q) in complex with LpqY, but in the absence of trehalose (Fig. 2A). SugABC in the LpqY-SugABC complex is similar to its structure in the absence of LpqY (RMSD of 1.55 Å for 1273 Cα atoms) (fig. S6A). The largest differences are observed in transmembrane domains SugA and SugB, which is most likely due to the binding of LpqY. Compared with the differences in the transmembrane region, the ATP catalytic domains are similar in the two structures (RMSD of 0.64 Å for 655 Cα atoms). Thus, we conclude that the current resolution is insufficient to reveal the molecular basis for the effects of different detergents on ATPase activity, such as A8-35 and GDN.
The notable feature of the LpqY-SugABC complex is that the SBP LpqY is tilted/rotated by ~26° relative to the transmembrane region of SugABC, and its N terminus likely inserts into the detergent micelle (Fig. 2, A and B). However, because of the strong influence of these micelles, it is difficult to identify the N-terminal lipid of LpqY in our structure. Nonetheless, mass spectrometry confirmed that N-terminal cysteine (Cys23) is modified by lipid (fig. S1D). By comparing possible lipid modification forms that have been reported in mycobacteria and other bacteria (37–40), the lipid modification of LpqY appears to be a diacylglycerol moiety (fig. S1, D and E).
As observed in other SBPs (41), two lobes of LpqY (N-lobe and C-lobe) are connected by a hinge form a cleft with the trehalose-binding site located at the interface (fig. S5, A and B). The cleft is exposed to the solvent to facilitate substrate capture. The interactions of LpqY with the TM region are mainly mediated by N-lobe helix 2 (65 to 90 amino acids), the C-terminal region of LpqY (459 to 465 amino acids), SugA P1 and P3 loops, and SugB P2 and P3 loops (Fig. 2C). An interesting question is whether the N-terminal lipid moiety of LpqY involves the interactions between LpqY and transmembrane region, in addition to serving to anchor LpqY to the cell membrane. Since the lipid moiety is not observed in our structure, this important issue needs further studies.
The C-terminal region of LpqY at the membrane surface is located between the SugA P1 and P3 loops and is mainly stabilized by hydrophobic interactions (Fig. 2D). Another critical contact is made by helix 2 of LpqY, which interacts with the SugA P3 loop and SugB P1 and P3 loops (Fig. 2C). There are several residues (e.g., Lys68 and Asp85) that anchor helix 2 of LpqY to the TM interface (Fig. 2, E and F).
Structure of LpqY-SugABCTRE with a trehalose-bound, pretranslocation state
Our third structure is of LpqY-SugABC in complex with trehalose, which represents a pretranslocation state. It makes a tight assembly with dimensions 145 Å by 80 Å by 60 Å (Fig. 3A). The cryo-EM map identifies a trehalose located in the substrate-binding pocket of LpqY (Fig. 3, A and D). An obvious feature is that the C-lobe of LpqY rotates by ~26° (compared to the resting state) and docks onto the TM region of SugABC (Fig. 3B). In this structure, there are many more interactions between LpqY and the TM region than observed in the resting state (fig. S6C). Superimposition of the apo (LpqY-SugABC) structure and the complex structure (LpqY-SugABCTRE) gives an RMSD of 1.56 Å for 1385 aligned Cαs with the largest differences located in the LpqY subunit (fig. S6B).
In this state, LpqY is in a closed conformation and is contracted to hold the trehalose in place (Fig. 3A and fig. S5C). That is achieved by the movement of helices 8, 10, 11, 12, 19, and 20 of LpqY toward the cleft, and as a result, the entrance pathway of trehalose is blocked (Fig. 3E and fig. S5A), a process akin to the mechanism of a Venus flytrap (42). Notably, the salt bridge between Lys68 (LpqY) and Glu239 (SugB) that stabilizes the conformation of LpqY when in the resting state is broken because of the conformational changes of the P3 loop of SugB, and instead, Glu239 interacts with Lys286 of LpqY (Fig. 2F). Considering the vital importance of the interactions of P3 loop (SugB) and helix 2 (LpqY) in the resting state, we suggest that these changes may trigger the conformation changes of LpqY upon the trehalose binding. In the trehalose-binding pocket, there are extensive interactions between trehalose and LpqY with the side chains of Asn39, Glu40, Asp94, Asn148, Glu255, and Arg418 all forming hydrogen bonds with trehalose (Fig. 3C). An additional 12 residues are involved in stabilizing the trehalose by hydrophobic interactions.
Structure of a catalytic intermediate state of trehalose importer LpqY-SugABCTRE_ATP
Our fourth structure is for LpqY-SugABC (E164Q) in complex with ATP, Mg2+, and trehalose. It has an outward-facing conformation (Fig. 4A). At the cytoplasmic side, the two SugC subunits form a closed dimer rotated by ~12° compared with that in the resting and pretranslocation states (Fig. 5A). The two ATP molecules are bound at the interfaces between the Walker A (residues 41 to 48) and LSGGQ motifs (residues 139 to 143) (Fig. 4A and fig. S7B). In the transmembrane region, because of ATP binding, the TM region switches from an inward- to outward-facing conformation. As a result, the substrate-binding site is exposed to the periplasmic side of the membrane (Fig. 4A). These changes to the TM region are also reflected in the separation of the coupling helices. In the catalytic intermediate state, they are much closer together than in the resting and pretranslocation states (Fig. 5B). To understand the conformation changes that occur in the TM region when transitioning from the pretranslocation to catalytic intermediate state, a superimposition of TM regions was performed (Fig. 5C). The RMSD for 455 aligned Cαs is 2.78 Å with the largest differences observed in SugB. It is apparent that upon ATP binding, the TM2 to TM5 helix bundle of SugB undergoes a rigid conformational change and moves toward the center of the TM region at the cytoplasmic side (Fig. 5C). The largest conformational changes occur in the TM5 helix of SugB, with a rotation angle of about 20° resulting in a translation distance of up to 11.6 Å (Fig. 5C). Notably, because of the conformational changes of the TM region on the cytoplasmic side, there is not enough space to accommodate the C terminus of SugB (residues 266 to 278), resulting in the structural rearrangements of this part from the helix to the loop (Figs. 3A and 4A). The C-terminal segment of SugB is inserted partway into the SugC dimer (Fig. 4, A and F). Residues 271 to 278 of SugB are running along the Q loop (residues 88 to 94) of one SugC subunit, with the terminal carboxyl group of SugB making a hydrogen bond with the side chain of Tyr89 on the Q loop in the adjacent protomer (Fig. 4F). In addition, the main chain of Gly274 and Lys277 of SugB interacts with the main chain of Ala90 and side chain of Tyr89 from SugC, respectively (Fig. 4F).
At the periplasmic side, the TM3 to TM5 helical bundles of SugA and SugB move away from the center of the TM region, which opens the periplasmic gate and, at the same time, leads to LpqY binding in the open form (Figs. 4A and 5D), and as a result, trehalose is released from LpqY into the substrate-binding pocket of the TM region. In this binding pocket, a trehalose is bound at the interface of SugA and SugB and is surrounded by 12 residues, seven of which (Glu173, Lys176, Thr177, Arg230, Asp233, and Asn240 in SugA and His118 in SugB) form hydrogen bonds with trehalose (Fig. 4, B and C).
In this catalytic intermediate state, LpqY in a ligand-free open form can dock onto the transmembrane domain of SugA and SugB where it makes more extensive interactions with the TM subunits compared with that in the resting and pretranslocation states (fig. S6C). A notable feature at the LpqY-TM region interface is that the periplasmic loop P3 of SugB (residues 216 to 240) identified as the “scoop loop,” connecting TM5 and TM6 helices, inserts into the LpqY sugar-binding cleft and occupies the trehalose-binding site (Fig. 4, D and E), which suggests that this loop plays a vitally important role in the release of trehalose from LpqY. A similar scoop loop is also found in the E. coil maltose transporter (fig. S4B) and has been demonstrated to be essential for substrate release to the transmembrane region in maltose transporter (26, 27, 43, 44). The scoop loop in this structure is stabilized by hydrogen bonding networks with two lobes of LpqY. Asn231 (main chain), Thr233, Gln237, Phe238 (main chain), and Glu239 of SugB from the scoop loop interact with Asn39 (main chain), Glu40, Lys68, Glu255, and Asn256 of LpqY, respectively (Fig. 4D). Furthermore, a salt bridge between SugB Glu239 and LpqY Arg259 and the stacking interactions of SugB Phe238 with Trp273 and Try257 of LpqY contribute to stabilizing the scoop loop (Fig. 4D). The side chains of Gln237 and Phe238 of SugB undergo a major conformation change in the catalytic intermediate state compared to the pretranslocation state. On one hand, the two residues participate in the stabilization of the conformation of scoop loop (Fig. 4D). On the other hand, their side chains are oriented toward and occupy the trehalose-binding pocket (Fig. 4E). Therefore, it is speculated that the two residues may play vitally important roles in substrate transport.
Mechanism of trehalose translocation
On the basis of these structural data, we propose an alternating access mechanism for trehalose transport across the membrane (Fig. 6). Initially, the transporter SugABC is assembled on the cell membrane in an inward-facing conformation. The lipoprotein, LpqY (SBP), is then secreted into the periplasm, where its N-terminal signal peptide is cleaved. This protein is then anchored to the membrane by lipid modification, and it leans against the TM region of SugABC when in the open form (i.e., without substrate bound). Free trehalose is then captured by LpqY with the C-lobe of LpqY rotating toward the membrane and docking onto the TM region of SugABC, while SugABC adopts an inward-facing conformation. After substrate capture, ATP binding causes SugC to assume a closed conformation, and this change is coupled to the conformation changes of the TM region and LpqY, which leads to the opening of the trehalose transport pathway from the periplasmic side to TM region. As a result, trehalose transfers into a translocation cavity in the TM region. Last, ATP hydrolysis with release of adenosine 5′-diphosphate and Pi opens the cytoplasmic exit gate, and the substrate is released into the cytoplasm. At the same time, the transporter LpqY-SugABC reverts back to the resting state.
ABC transporters are ubiquitous in prokaryotes and eukaryotes and play vital roles in transporting diverse substrates across the cell membrane (45, 46). In bacteria, ABC importers are essential for the uptake of nutrients and micronutrients, and they usually have an SBP to bind the specific substrate at the periplasm side (46–49). The mechanism of ABC importers in Gram-negative bacteria is understood on the basis of multiple crystal structures showing snapshots of the E. coli MBP-MalFGK2 complex (26, 27, 43, 50). However, the types of SBPs are different in Gram-negative and Gram-positive bacteria (49, 51). Generally, the SBPs can shuttle between the inner and outer membranes in Gram-negative bacteria, while in Gram-positive bacteria, the SBPs need to be anchored on the membrane or the transporter (52). These differences suggest varied substrate capture mechanisms in these two bacteria. To date, few structures of an ABC transporter are reported in Gram-positive bacteria. As a result, the precise details as to how these transporters work have remained unclear in these bacteria.
Here, the mechanism of a type I ABC importer is revealed in Gram-positive bacteria. Notably, it is also the first time that a full complex of lipid-anchored SBP (lipoprotein) and ABC transporter has been determined. Interestingly, in the resting state, the orientation of LpqY (SBP) relative to SugABC is distinct from these of equivalent SBPs in the E. coil (a Gram-negative bacteria) maltose transporter MalFGK2 (32) and Methanosarcina acetivorans molybdate/tungstate importer ModBC (fig. S4A) (33). On the basis of these differences, we suggest that LpqY binds to the transporter before substrate binding and is held there throughout the entire process of import. To our knowledge, this type of association in the resting state of ABC transporters has not previously been observed. Besides that, previous studies of E. coil maltose transporter MBP-MalFGK2 in a pretranslocation state have shown that the MalFGK2 structure is located between inward- and outward-facing states (50), while SugABC in the complex adopts an inward-facing conformation in this state. This difference may be caused by the maltose bound in the transmembrane domain of MBP-MalFGK2.
Among the various types of transporters, ABC importers are widely present in bacteria and can be used as potential antibacterial targets (46, 53, 54). On the basis of the three-dimensional (3D) structure of the transporter in complex with the substrate, substrate analogs can be designed as potential antibacterial drugs. These analogs can directly inhibit transporters or uptake of metabolic nutrients into cells. Mycobacteria cause a variety of diseases, of which TB caused by Mtb is the most widespread (9). It is estimated that there are nearly 10 million new Mtb infections per year, with around 1.3 million deaths occurring annually (55). Thus, despite serious efforts to try to curtail this disease in recent years, there remains an urgent need to develop drugs to target TB.
In mycobacteria, trehalose metabolism is vital for the survival and virulence of these bacteria (12). Notably, trehalose analogs have already shown great potential as antibacterial agents and can inhibit the growth and the formation of biofilms in mycobacteria (19, 22, 56). To date, LpqY-SugABC has been identified as the only trehalose importer in these bacteria and is required for these trehalose analogs to be transported across the membrane (7, 22). Therefore, the structural data presented here will greatly facilitate the design and development of trehalose analogs as potential antibacterial drugs. Furthermore, the outbreaks of hypervirulent strains of C. difficile are associated with trehalose metabolism (8, 57). These strains can metabolize low concentrations of trehalose, which contributes to hypervirulence of the pathogen and increases the risk of disease (8, 57). Thus, our structures also provide insights for finding drug leads that can prevent disease caused by other pathogenic bacteria such as C. difficile.
MATERIALS AND METHODS
Expression and purification of SugABC
The genes encoding the ABC transporter SugABC were cloned from M. smegmatis strain mc2155 genomic DNA and overexpressed in M. smegmatis mc2155. MSMEG_5058, MSMEG_5059, and MSMEG_5060 were amplified by polymerase chain reaction (PCR) and inserted into a pM261 vector, with a C-terminal 6× His-tag attached to SugC. The recombinant plasmid was transformed into M. smegmatis mc2155 cells by electroporation. The cells were cultivated at 37°C in LB liquid media supplemented with kanamycin (50 μg ml−1), carbenicillin (20 μg ml−1), and Tween 80 (1 g ml−1) until the optical density at 600 nm reached 1.0. Overexpression of protein was induced by 0.2% (w/v) acetamide at 16°C. After 4 days, cells were harvested by centrifugation at 4000 rpm for 20 min. Cell pellets were resuspended in buffer containing 20 mM Hepes, (pH 7.5) and 300 mM NaCl and then lysed by passing through a high-pressure homogenizer at 1200 bar. Cell debris was removed by centrifugation at 12,000 rpm for 10 min at 4°C. The supernatant was collected and ultracentrifuged at 37,000 rpm for 1.5 hours to isolate the membrane fractions. The membrane fraction was solubilized in buffer with the addition of 1% (w/v) DDM (Anatrace) at room temperature for 1 hour. The suspension was ultracentrifuged and the supernatant supplemented with 20 mM imidazole was loaded onto a nickel–nitrilotriacetic acid (Ni-NTA) agarose beads (Qiagen) affinity column. The beads were rinsed in 20 mM Hepes, (pH 7.5), 150 mM NaCl, 0.02% (w/v) DDM, and 50 mM imidazole, and the protein was eluted from the beads with 20 mM Hepes, (pH 7.5), 150 mM NaCl, 0.02% (w/v) DDM, 10 mM dithiothreitol (DTT), and 500 mM imidazole. The eluted sample was concentrated and applied to a size exclusion chromatography column (Superose 6 Increase, GE Healthcare) preequilibrated with 20 mM Hepes, (pH 7.5), 150 mM NaCl, 2 mM DTT, and 0.02% (w/v) DDM. SugABC in DDM was mixed with amphipols at 1:3 (w/w) with gentle agitation at room temperature for 15 min. Detergent was removed by rocking incubation with Bio-Beads SM-2 (Bio-Rad) at room temperature for 3 hours. Beads and insoluble substances were removed by centrifugation, and supernatant was applied to a size exclusion chromatography column (Superose 6 Increase, GE Healthcare) preequilibrated with 20 mM Hepes, (pH 7.5), and 150 mM NaCl. The peak fractions were collected and concentrated to 0.8 mg ml−1 for cryo-EM sample preparation.
Expression and purification of LpqY-SugABC-E164Q
The LpqY-SugABC operon was coexpressed in M. smegmatis mc2155. To prevent ATP hydrolysis and trap ATP in the nucleotide-binding site, the catalytic glutamate in the consensus site [E164 in MSMEG_5058 (SugC)] was mutated to glutamine. A coexpression vector was constructed by amplification of MSMEG_5058 (E164Q), MSMEG_5059, MSMEG_5060, and MSMEG_5061 from M. smegmatis strain mc2155 genomic DNA by PCR and insertion into a pM261 expression vector with a C-terminal 6× His-tag. LpqY-SugABC was expressed and purified following a similar protocol to SugABC. The membrane fraction was resuspended and solubilized in buffer with the addition of 1% (w/v) LMNG (Anatrace) at room temperature for 1 hour. The suspension was centrifuged at 18,500 rpm for 40 min at 4°C, and the supernatant supplemented with 20 mM imidazole was loaded onto a Ni-NTA agarose beads (Qiagen) affinity column. The beads were washed in 50 mM imidazole, followed by incubation with a buffer containing 0.04% (w/v) GDN for 10 min. The protein was eluted from the beads with 20 mM Hepes, (pH 7.5), 150 mM NaCl, 0.04% (w/v) GDN, 10 mM DTT, and 500 mM imidazole. The eluted sample was concentrated and further purified by a size exclusion chromatography column (Superose 6 Increase, GE Healthcare) in a buffer containing 20 mM Hepes, (pH 7.5), 150 mM NaCl, 2 mM DTT, and 0.04% (w/v) GDN. The peak fractions were collected and concentrated for cryo-EM sample preparation. The protein sample (4 mg ml−1) was incubated with 1 mM trehalose alone or 8 mM ATP, MgCl2, and 1 mM trehalose on ice for 30 min before freezing the grids.
All ATPase assays were performed using the ATPase/GTPase Activity Assay Kit (Sigma-Aldrich, catalog number MAK113-1KT). Purified SugABC (0.72 μg) or 5.46-μg LpqY-SugABC-E164Q in A8-35 was incubated in a 20-μl reaction volume containing 10 mM Hepes, (pH 7.5), 75 mM NaCl, 1 mM ATP, and 2.5 mM MgCl2 (Sangon Biotech) for 30 min at 37°C. The reaction was stopped by adding 100 μl of reagent (catalog number MAK113A) and incubated for an additional 30 min at room temperature to generate the colorimetric product. Absorbance at 620 nm was measured at room temperature using a SpectraMax iD3 multifunction reader (Molecular Devices). ATPase activity was represented as phosphate (nanomoles) produced by 1 mg of protein per minute. Experiments were performed in triplicate.
LpqY was separated by SDS–polyacrylamide gel electrophoresis from purified LpqY-SugABC (E164Q). The bands excised from gel were destained with 50% (v/v) methanol and dried by vacuum centrifugation. The dried gel pieces were rehydrated with 2 μl of trypsin (10 ng ml−1). The resulting digests were analyzed by mass spectrometry after chloroform/methanol extraction. Mass spectrometry was performed using a matrix-assisted laser desorption/ionization TOF/TOF 5800 analyzer (AB Sciex, Framingham, MA, USA) in a positive reflectron mode. Saturated ɑ-cyano-4-hydroxycinnamic acid (CHCA) solution in a chloroform/methanol (2:1, v/v) solvent was used as matrix. A thin layer of CHCA matrix was prepared, and the samples were deposited on the matrix.
EM sample preparation and data collection
Aliquots (3 μl) of freshly purified SugABC or LpqY-SugABC (E164Q) complexes were applied to glow-discharged holey carbon grids (Quantifoil Cu R1.2/1.3). The glow discharge was followed by a standard receipt of a mixture of H2 and O2. Grids were blotted for 2.5 s and flash-frozen in liquid ethane cooled by liquid nitrogen using a FEI Mark IV Vitrobot operated at 8°C and 100% humidity. Cryo-EM data were collected on a FEI Titan Krios electron microscope operated at 300 keV with a Gatan K3 camera at ×29,000 nominal magnification in superresolution mode and binned to a pixel size of 0.82 Å. Automated single-particle data acquisition was performed with SerialEM (58).
EM data analysis
For cryo-EM image processing, all steps were performed using cryoSPARC (59). For the SugABC complex, 4126 original images were aligned, and local motion was corrected using patch motion correction with dose weighting. Micrographs that exhibited defects in the Thon rings due to excessive drift, ice contamination, or astigmatism were discarded. Particles (1,825,045) were automatically selected and submitted for several rounds of reference-free 2D classification to discard bad particles. After particle cleaning, 423,850 were used for ab initio reconstruction to generate 3D models as references to perform heterogeneous refinement. After heterogeneous refinement, 197,191 particles were refined using non-uniform (NU) refinement to generate the final cryo-EM map with an estimated average resolution of 3.41 Å according to the gold-standard Fourier shell correlation cutoff of 0.143 (60). Local resolution ranges were also analyzed within cryoSPARC. All four datasets were processed in the same way.
Model building and refinement
Model building of the SugABC complex was based on the 3.4-Å cryo-EM map and was performed using Coot (61). Aromatic residues were used as landmarks to identify specific regions of structure. The models for the LpqY-SugABC (E164Q), LpqY-SugABC (E164Q)_TRE–bound (trehalose, TRE), and LpqY-SugABC (E164Q)_TRE_ATP–bound complexes were generated on the basis of their cryo-EM maps and using the model of SugABC as a starting point. Trehalose and ATP were fitted as rigid bodies into the cryo-EM map using Coot (61). The structures were refined in real space using PHENIX with secondary structure and geometry restraints to prevent overfitting (62). The final atomic model was evaluated using MolProbity (63). Cryo-EM data collection and model refinement statistics are shown in table S1.
All the figures were generated using UCSF (University of California, San Francisco) ChimeraX (64), UCSF Chimera (65), or PyMOL (www.pymol.org).
Acknowledgments: We thank the staff from the Electron Microscopy Facility of ShanghaiTech University, for assistance during cryo-EM data collection. We also thank C. Peng and X. Tian of the Mass Spectrometry System at the National Facility for Protein Science in Shanghai (NFPS), Zhangjiang Lab, Shanghai Advanced Research Institute, Chinese Academy of Science, China for data collection and analysis. We are thankful to the analytical chemistry platform of Shanghai Institute for Advanced Immunochemical studies (SIAIS) for assistance in mass spectrometry analysis. Funding: This work was supported by grants from the National Key Research and Development Program of China (grant no. 2017YFC0840300) and the National Natural Science Foundation of China (grant no. 81520108019) to Z.R. Author contributions: Z.R. initiated and supervised the project. Z.R., B.Z., and F.L. designed experiments. F.L. and J.L. made all constructs and purified the proteins. F.L. and Y.G. collected and processed cryo-EM data. B.Z., F.L., and Y.G. built and refined the structure models. F.L., J.L., B.Z., and Y.G. performed functional experiments with help from X.Y. and T.H. F.L., J.L., B.Z., Y.G., X.Y., T.H., H.Y., W.X., and Z.R. analyzed and discussed the results. The manuscript was written by B.Z., F.L., J.L., Y.G., L.W.G., and Z.R. with the help of all other authors. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. The accession numbers for the 3D cryo-EM density maps reported in this paper are EMD-30327, EMD-30328, EMD-30329, and EMD-30330. Atomic coordinates and structure factors for the SugABC, LpqY-SugABC (E164Q), LpqY-SugABC (E164Q)_TRE, and LpqY-SugABC (E164Q)_TRE_ATP structures have been deposited in the Protein Data Bank with identification codes 7CAD, 7CAE, 7CAF, and 7CAG. Additional data related to this paper may be requested from the authors.
- Copyright © 2020 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. Distributed under a Creative Commons Attribution NonCommercial License 4.0 (CC BY-NC).