(707f) Redesign of E. coli Water Channel Protein, Ompf, Using Iterative Protein Redesign and Optimization (IPRO) Suite

Chowdhury, R., The Pennsylvania State University
Kumar, M., The Pennsylvania State University
Grisewood, M., The Pennsylvania State University
Maranas, C., The Pennsylvania State University
The water footprint for food production, processing and energy production is increasing everyday with the world population and freshwater is becoming an increasingly limited resource. Ultra-permeable membranes are rapidly emerging as a widely implemented technology to utilize and efficiently recycle marginal water sources. Existing membranes are mainly dense solution or diffusion-based or channel- based. Aquaporins (AQPs) are the most popular biological channel-based membranes due to their short pore lengths and high water permeabilities of billion molecules per second. Another particular member of the â??porinâ? class of beta-barrel proteins is Outer Membrane Protein F (OmpF). Unlike AQPs, OmpF is exceptionally stable with the ability to survive contact with solvents such as 100% ethanol. While the stability and mutation tolerance of OmpF makes it a suitable candidate for computational design and subsequent experimental validation for performing separations at the angstrom scale, it can also be easily assembled into stable block-copolymer membrane sheets. To this end, we employed the Iterative Protein Redesign & Optimization (IPRO) suite (1) of programs to perturb the OmpF protein backbone, mutate and recursively repack the sidechains in order to obtain the selective internal structure of AQP inside the stable beta-scaffold of OmpF.

We started with the wild-type E.coli OmpF (elliptical pore constriction with major and minor axes of ~1.1 nm and 0.7 nm respectively). Pore sizes between 0.3 nm and 0.5 nm are key to several industrial and environmental separations such as methane/CO2 and salt/water, which is accessible to the less stable AQPs. We first performed ensemble analysis on molecular dynamics (MD) simulations of water permeation through E. coli AQP1 using k-means clustering technique. Next, we fixed the relative coordinates of water molecules in a water-wire from each ensemble and put them inside the OmpF pore. Subsequently, we identified the OmpF pore-constricting residues and allowed them to be mutated to only hydrophobic residuesto eliminate any interactions between the water-wire and the pore wall. In addition, we extended the IPRO suite with the ability to select designs based on a geometric criterion in which only designs that have both the pore-constriction axes smaller than 0.35 nm, were accepted. Interestingly, we identified three main types of mutants from our results. The first type comprises mutants where the design positions were only allowed to mutate to tryptophan (longest hydrophobic side chain). This was performed to test out the physical limit up to which the OmpF pore area can be reduced. We have identified a mutant with 25 tryptophan mutations having pore axes of 0.224 nm and 0.18 nm, respectively. The other two types of mutants comprise one with typical hour-glass shaped constriction (pore axes lengths: 0.319 nm and 0.2552 nm respectively) and another which appears as two stacks of ellipses, one arranged on top of the other with the major axes of the lower stack nearly at a right angle to that of the ellipses in the stack above, thereby making the pore very small (axes lengths: 0.236 nm and 0.194 nm respectively). We will perform subsequent experimental validation in establishing a platform for designing precisely tuned membrane transporters. These engineered membrane proteins will serve as important industrial workhorses in energy-efficient water treatment.

1. Pantazes RJ, Grisewood MJ, Li T, Gifford NP, Maranas CD. The Iterative Protein Redesign and Optimization (IPRO) suite of programs. J Comput Chem. 2015;36(4):251-63.