Tripping up transport across bacterial membranes

The prevalence of bacteria resistant to multiple drugs underscores the need to discover entirely new classes of antibiotics. One factor that greatly contributes to the infectious capacity of pathogenic bacteria is their ability to regulate the transport of drugs, nutrients and other molecules across their cell wall and membrane barriers. From an antibiotic drug discovery perspective, the challenge is clear: bacteria are remarkably adept at ensuring essential nutrients get transported in and that toxic compounds get transported out.

Like Comment

The paper in Nature is here:

Among the many tools that bacteria use to control the transport of molecules to and from their environment are membrane-spanning ATP-binding cassette (ABC) transporters. These essential molecular machines use the power of ATP hydrolysis to catalyze substrate transport across a membrane. It is somewhat surprising that no antibiotics targeting ABC transporters have yet been clinically approved, despite the importance of this ubiquitous protein family to the survival of all bacteria. 

In 2015, we published in Structure a strategy for targeting a Gram-positive bacterial ABC transporter. The goal of our work was to determine whether and how an antibody fragment could access and block a Staphylococcus aureus ABC transporter that is essential for virulence. Even though the Gram-positive cell wall is a barrier to full-length antibodies, we were surprised to find that a smaller antibody fragment could access and antagonize an ABC transporter pathway. 

An antibody fragment blocking a bacterial ABC transporter.

Unfortunately, no antibody fragment can access the ABC transporters of pathogenic Gram-negative bacteria, like Escherichia coli. This is because Gram-negative bacteria have two membranes. The near-impenetrable outer membrane prevents the entry of even small molecule drug-like molecules, with the exception being polar molecules smaller than 600 Da that can traverse through water-filled pores. The permeability barrier is imparted by an asymmetric bilayer consisting of a complex glycolipid known as lipopolysaccharide, or LPS, on the outer leaflet and phospholipid on the inner leaflet. LPS is the major lipid component of the bacterial outer membrane but its synthesis begins on the inner (cytoplasmic) leaflet of the inner membrane. As the first step in its transport to the outer membrane, LPS must be flipped across the inner membrane by an ABC transporter known as MsbA.

Our recent work in Nature showcases our efforts to discover and structurally characterize inhibitors of MsbA. This project was launched with the hypothesis that inhibiting MsbA would disrupt LPS transport and kill Gram-negative bacteria. Initially, we sought to discover small molecule inhibitors through high-throughput biochemical screening of ~3 million compounds on purified, full-length MsbA. The screening of full-length protein allowed for the possibility that we would identify molecules that blocked the conformational changes necessary for transport, whereas screening only the isolated ATPase domain would have likely limited hits to those binding the cytoplasmic ATP pocket. We then biased the hit selection for phenotypic activity by immediately following this with growth inhibitory assays on Escherichia coli.

One of the most powerful secondary assays we next employed was the analysis of inhibitor-treated bacterial cells by electron microscopy. Only those compounds specifically targeting MsbA mimicked the cytosolic membrane elaboration phenotype that is a hallmark of msbA loss-of-function. So striking and telling was this phenotype that one of our most memorable team meetings was when the first microscopy results were revealed for a pathogenic strain of Escherichia coli treated with inhibitor.

Membrane elaboration upon MsbA inhibition.

The electron microscopy data, along with the subsequent isolation of resistant mutations in the msbA gene, convinced us that we had identified inhibitors with on-target phenotypic activity. We then set ourselves the aim of enabling structure-based drug design by determining structures of MsbA in complex with these inhibitors. We explored multiple crystallization strategies, including those used in earlier MsbA structural studies, but obtained well-diffracting complex crystals suitable for structure determination only when a steroid-based facial amphiphile detergent was used. 

We hope that visualization of the first inhibitor-MsbA complex will inspire additional functional studies and aid in the development of selective modulators of other ABC transporters. 

Structure of the MsbA dimer in complex with inhibitor and LPS.

See the published paper in Nature here:

Christopher M Koth

Director and Sr. Scientist, Genentech