Chiral separation of arginine based on tailor-made FhuA β-barrel protein

Anand, Deepak; Schwaneberg, Ulrich (Thesis advisor); Böker, Alexander (Thesis advisor)

Aachen : RWTH Aachen University (2020, 2021)
Dissertation / PhD Thesis

Dissertation, RWTH Aachen University, 2020


Chirality of chemical compound is ubiquitous in nature performing central function in metabolism of many nutrients and pharmaceuticals. Even-though enantiomers of a compound have similar chemical and physical properties it can have completely different biological activity. Thus chiral molecules are of large economic value in chemical, pharmaceutical, and food industries. Many applications in these industries require the isolation and use of single chiral isomers (enantiomers) of chiral compounds. As a result, there is an ever increasing necessity of optically pure compounds but it is a challenging task to obtain it. Since the first optical resolution of tartaric acid, performed by Louis Pasteur in early 1848, various techniques have been developed for chiral resolutions of enantiomeric compounds. Methods such as chromatographic or enzymatic techniques are commonly used. However, they are limited by difficulties for large-scale productions. Crystallization can be used in large-scales but often several rounds of crystallization and recrystallization are required to obtain enantiopure compounds due to entrapments of the unwanted enantiomer during crystal growth, which often leads to significant reduction in yields (up to 50 %). Each and every one of these techniques has their own advantages and disadvantages but the membrane based approach is of particular interest because of its use in continuous operation and its ease of scale-up. But the problem with the membrane based technology developed till now is that of non-uniform pore size and limited number of pores. Inspired by nature, membrane-based approaches for chiral separation using a barrel protein can be a suitable alternative. Chiral protein-polymer membranes would be an attractive, cost-effective, and scalable method. However, the main challenges lie in the design of "filter regions" and the generation of "screening systems" to identify chiral channel proteins. In the present thesis, ferric hydroxamate uptake component A (FhuA) was engineered to generate chiral channel for separation of arginine enantiomers which can be used as a scaffold for the generation of protein-polymer membrane. FhuA is a large monomeric transmembrane protein of Escherichia coli which folds into a barrel consisting of 22 antiparallel β-strands and a barrel-plugging "corkdomain". FhuA was chosen as a functional nanopore because of its high tolerant towards organic solvents, thermal resistant, and has a high robustness towards reengineering. Structurally, FhuA contains a water channel wherein two flexible loops in the cork domain(loop1; residue 35-40 and loop2; residues 135-145) were shortened to generate two selectivity filter regions (filter1 and filter2). Additionally, cork domain was stabilized inside the barrel by substituting three amino acids (Q62D, R81W, and N117L) resulting in generation of FhuAΔLvariant having higher interaction between barrel and cork domain. In order to generate a chiral selective variant, a directed evolution protocol was developed to generate a chiral FhuAΔLvariant. A novel whole cell calorimetric screening system based on amino acid utilizing enzyme (argininedeiminase) was developed in order to identify the enantioselective variant from mutant libraries of FhuAΔL. Screening of mutant libraries led to the identification of FhuAF4 variant (amino acid substitutions: G134S, G146T) showing approximately two times higher transport of L-arginine compared to parent FhuAΔL with E-value=1.92; ee %=23.91 at 52.39 % conversion. Steered molecular dynamics (SMD) simulations were carried out for molecular understanding of observed altered enantiopreference of FhuAF4 variant towards L-arginine compared to the FhuAΔL variant. In FhuAΔL variant, less interactions of both D- and L-arginine with the filter region 2 were observed and both enantiomers pass freely. Interestingly, in FhuAF4 variant, transport of D-arginine is hindered and steered transport is slowed down, indicated by a longer residence time close to the selectivity filter region 2. The obtained results provide first proof of principle of engineering of a membrane protein towards chiral separation of amino acids and provide insight into the mechanism of chiral separation within the FhuA channel. It is likely that with the identified filter region and OmniChange libraries further improvements are achievable for other amino acids and a broader range of enantiomers. The chiral FhuA channel proteins would be an excellent scaffold for generation of chiral membrane based on protein polymer conjugates with a high potential for novel and scalable downstream processes in pharmaceutical and food industries.