Ph.D Thesis Colloquium : Aditya Upasani

The effect of membrane composition on the activity of membrane disrupting agents.

Membrane targeting agents such as pore-forming toxins (PFTs) and antimicrobial peptides (AMPs) specifically disrupt target cell membranes leveraging its lipid constituents, organization and biophysical properties. Cells react to these attacks by altering the composition and properties of their membranes, aiming to mitigate the disruption. Throughout the co-evolution of hosts and pathogens, both have developed the ability to adapt to these membrane alterations. In this study, I present two contrasting systems within the host-pathogen spectrum where modifications to the membrane facilitated by host cell signalling can either aid or hinder the function of proteins or peptides targeting the membrane.

First, I discuss how drug resistance to colistin, an antimicrobial peptide that disrupts bacterial membranes, in gram-negative bacteria emerges via membrane modifications. Many colistin resistant clinical isolates with possible links to membrane modifications have been reported though the mechanisms that lead to the emergence of colistin resistance is not well understood. Using single-cell imaging and mass spectrometry, I show that a broad heterogeneous level of pre-existing tolerance against colistin enables the E. coli bacteria to progressively adapt to high colistin concentrations by modifying their lipid A to reduce colistin binding and activity. These tolerant bacteria are detectable early during the first exposure to colistin in isogenic naive populations suggesting phenotypic adaptation as the precursor to emergence of resistance. A second class of tolerant bacteria that displays similar reduced colistin binding activity but via gene duplications serves as the precursors of pmrB gene point mutation variants that eventually allow fixing this membrane modification as dominant resistance. These colistin tolerant and resistant bacteria display characteristic lipid A modifications including phosphoethanolamine addition leading to lowering of negative charge of the membrane albeit at the cost of membrane integrity due to increased fraction of four-acyl chained lipid A. I take advantage of this reduced membrane integrity (potential collateral sensitivity), to demonstrate that the rate of resistance emergence against colistin, as well as other AMPs such as human AMP LL-37, can be reduced by combining AMP treatment with small membrane-acting bioactive molecules, thymol and terpineol. Using single-molecule tracking and super-resolution imaging, I show that AMPs displays membrane insertion and clustering with time on supported lipid bilayers and live bacteria. AMPs aids the membrane permeation due to bioactive molecules, thymol and terpineol. I propose that the increase in membrane fluidity due to thymol and terpineol enhances the membrane permeation activity of the AMPs, which results in overall increased membrane susceptibility to membrane acting peptides like colistin and LL-37.

In the second half, I examine how sterol-oxysterol balance changes in the mammalian cell membranes affect the activity of pore forming toxins. Mammalian cells respond to pathogen attacks by disrupting the cholesterol synthesis which is known to be essential for action of bacterial pore-forming toxins. Additionally, interferon mediated activation of CH25H gene by bacterial components like LPS leads to oxidation of cholesterol. Oxidised sterols like 25-hydroxycholesterol (25-HC) tend to act as broad-spectrum antimicrobials and inhibit cholesterol dependent cytolysins though their action against other pore-forming toxins is poorly known. I show that Cytolysin A (ClyA), a traditional alpha-toxin and a cholesterol-dependent pore-forming toxin (PFT), has evolved the ability to avoid inhibition by 25-HC and has even adapted to substitute cholesterol with 25-HC to enhance its activity. Single-molecule imaging on synthetic lipid membrane systems shows that this enhancement results from the increased membrane insertion of ClyA and enhanced ClyA assembly on 25-HC membranes. Additionally, to augment the understanding of ClyA assembly dynamics measured on heterogeneous membranes, I developed a ClyA assembly dynamics model on phase separated membranes using stochastic spatial simulations. These simulations show that ClyA membrane insertion and assembly depends on cholesterol content and size of cholesterol-rich membrane domains. These results support the experimentally observed increased assembly rates for ClyA in membranes containing cholesterol and sphingomyelin and how it can be perturbed by sterol composition and availability.

Overall, this study underscores the importance of understanding the interplay between membrane-disrupting proteins and peptides and their target cell membranes to understand the effectiveness of  membrane composition modifications as a defence mechanism against these agents.