Cryo-EM structure and resistance landscape of M. tuberculosis MmpL3: An emergent therapeutic target

Abstract

Tuberculosis (TB) is the leading cause of death from a single infectious agent and in 2019 an estimated 10 million people worldwide contracted the disease. Although treatments for TB exist, continual emergence of drug-resistant variants necessitates urgent development of novel antituberculars. An important new target is the lipid transporter MmpL3, which is required for construction of the unique cell envelope that shields Mycobacterium tuberculosis (Mtb) from the immune system. However, a structural understanding of the mutations in Mtb MmpL3 that confer resistance to the many preclinical leads is lacking, hampering efforts to circumvent resistance mechanisms. Here, we present the cryoelectron microscopy structure of Mtb MmpL3 and use it to comprehensively analyze the mutational landscape of drug resistance. Our data provide a rational explanation for resistance variants local to the central drug binding site, and also highlight a potential alternative route to resistance operating within the periplasmic domain.

Keywords: LMNG; MmpL; MmpL3; RND transporter; antitubercular; cryo-EM; drug resistance; membrane protein; mycolic acid; tuberculosis.

Graphical abstract

Figure 1

Cryo-EM map and structural analysis of Mtb MmpL3 (A) Cryo-EM density of Mtb MmpL3_1-753 as viewed from the membrane plane (front), or perpendicular to it from the periplasm (top). Proteinaceous density is blue (contour level 0.427, sharpened map), the surrounding detergent micelle gray, and a bound LMNG molecule gold (contour level 0.150, unsharpened map). Annotations highlight the transmembrane domain (TMD), and the N-terminal (PN) and C-terminal (PC) lobes of the periplasmic domain (PD). (B) Cartoon representation of the Mtb MmpL3_1-753 atomic model, colored blue (N terminus) to red (C terminus). LMNG is shown as sticks (gold). (C) Licorice representations of Mtb MmpL3_1-753 (gray) accompanied by surface depictions emphasizing its PD cavity network (wheat, calculated using Caver Analyst 2.0 [Jurcik et al., 2018]) and TMD drug binding pocket (teal). The TMs (1–4 and 7–10) lining cavity exits into the outer leaflet are indicated. (D) Electrostatic surface renderings of the MmpL3_1-753 PD, highlighting the hydrophilic residues gating the labeled PD_F, PD_B, and PD_T cavity openings. See also Figures S1 and S2, and Table S1.

Figure 2

Comparison of MmpL3 periplasmic ligands (A) Close-up of the Mtb MmpL3_1-753 PD LMNG. Residues within 4.5 Å, plus the detergent itself (gold), are shown as sticks; bar glycines shown as balls. Side chains are rainbow colored, as in Figure 1B. Dashed lines denote hydrogen bonds. Asterisks highlight residues conferring inhibitor resistance when mutated. (B) Side-by-side comparison of the 6DDTre (green), DDM (salmon), and PE (blue) molecules co-crystallized within the Msmg MmpL3 PD. Ligands are shown as spheres and proteins (gray) in licorice representation. PDB codes are in brackets. (C) Overlay of the Mtb and Msmg PD ligands, each depicted as sticks and colored as in (A) and (B). Common alkyl chain sites, and the positions of the PD apertures, are annotated.

Figure 3

Mapping MmpL3 drug-resistance mutations (A) Eighty resistance positions (RPs) plotted onto Mtb MmpL3_1-753 as balls, color-coded by distance from the universal drug pocket (teal). The PD cavity is also shown (wheat surface). (B) Residues of the 22 RPs causing cross-resistance to a minimum of three classes of MmpL3 inhibitor. Black asterisks denote cross-RPs only observed in M. abscessus MmpL3. See also Figures S3 and S4, and Table S2.

Figure 4

Wider non-synonymous landscape of MmpL3 (A) Structural mapping of MmpL3 variants mined from over 45,000 Mtb genomes, represented by balls and color-coded by mutational frequency. The most abundant substitutions, and/or those with apparent Mtb lineage association (colored asterisks) are labeled. (B and C) (B) Incidence of non-synonymous mutations in the TM6-7 loop, and (C) C-terminal domain. See also Figure S4 and Table S3.