Engineering proteins with novel mechanical properties by recombination of protein fragments.

The mechanical properties of proteins are crucial in living cells as proteins serve as basic units in cells to constantly sense, generate, and bear mechanical forces. Outside the cell, nature exploits the mechanical features of proteins and produces a variety of materials with superb mechanical properties (e.g. spider dragline silk) which often outperform man-made materials. Studies of the mechanics of proteins are not only important in understanding fundamental biophysical principles underlying various biological processes, but they also underscore the great potential in engineering protein-based advanced materials and using proteins as building blocks for the bottom-up construction of functional nanomechanical devices. Despite the great significance of mechanical proteins, the investigation of the mechanical design of these proteins at the single-molecule level was not possible until the development of single-molecule force spectroscopy techniques a few years ago. Since then, single-molecule atomic force microscopy (AFM) has become the workhorse in the field of protein nanomechanics. 12–15] Initial single-molecule AFM studies focused on naturally occurring mechanical proteins that are placed under stretching force under physiological conditions, such as the immunoglobulin (Ig) domains of the giant muscle protein titin and the fibronectin type III domains from fibronectin and tenascin. 10,11,16–19] Recently, non-mechanical proteins have also been studied by single-molecule force spectroscopy in the search for mechanically stable protein folds and to expand the toolbox of mechanical proteins. With the increasing understanding of the relationship between protein structure and protein mechanical stability, scientists have attempted to engineer or modify the mechanical properties of proteins. Such efforts are limited to sitedirected mutagenesis. With the advent of nanobiotechnology, it is desired to engineer novel multifunctional mechanical proteins that combine desired mechanical properties with other functional properties, such as enzymatic activity. Site-directed mutagenesis may not suffice for such challenges. Here, we demonstrate the feasibility and the first attempt towards engineering novel mechanical proteins through DNA-shuffling-based recombination. Recombination is an important mechanism for proteins to acquire novel functions. Recombination offers the advantage of combining beneficial mutations from multiple parents into a single offspring and has been exploited extensively by nature during evolution in improving protein traits such as enzymatic activity. This method has also been used extensively in the directed-evolution of proteins in the laboratory and has become one of the most important strategies in engineering proteins with novel functions. 30] Recombination is based on DNA-shuffling of proteins sharing high sequence homology and identity. Recent developments have extended this method to proteins that are distantly related and share low sequence homology. Here, as a proof of principle, we demonstrate the feasibility of using this powerful method to engineer proteins with novel mechanical stability. This will serve as the first step towards engineering multifunctional proteins in which mechanical stability is combined with desired additional functionality. We used the 27th and 32nd immunoglobulin domains (I27 and I32, following the nomenclature by Labeit) from human cardiac titin as the parent model proteins to construct hybrid proteins by recombining different fragments from the two parents. I27 and I32 are good model systems for mechanical proteins and have been studied extensively by single-molecule AFM. It was shown that I27 unfolds at a force of 200 pN while I32 is mechanically more stable and unfolds at a force of 300 pN. The big difference in their mechanical stability may have interesting ramifications for the mechanical properties of hybrid recombinant proteins. Extensive single-molecule atomic force microscopy studies, 26–28,34,35] in conjunction with steered molecular dynamics simulations, revealed the critical importance of the A’ and G b strands (referred to as the A’G patch) to the mechanical stability of the I27 domain. When I27 is stretched from its N and C termini, a shear force is applied to the hydrogen-bonding network in the A’G patch, which forms the force-bearing parts of I27 and constitutes the strongest mechanical resistance to unfolding. This shear topology of the hydrogen-bonding network seems to be a common feature among stable mechanical proteins, and suggests that mechanical stability may be a rather local phenomenon. Hence, it seems possible to modulate the mechanical properties of I27-like proteins by shuffling the A’G patch among I27 homologous proteins. Here, we show that, by interchanging the force-bearing A’ and G strands or the presumably nonforce-bearing C, D, and E strands between I27 and I32, we can successfully generate mechanically stable, hybrid proteins. To the best of our knowledge, this is the first successful recombination approach to impart novel mechanical properties to proteins. I27 and I32 share a high sequence homology (identity 42%, similarity 57%, according to the program CLUSTALW, Figure 1A). Although the three-dimensional structure of I32 [*] Dr. D. Sharma, Y. Cao, Prof. Dr. H. Li Department of Chemistry The University of British Columbia 2036 Main Mall Vancouver BC, V6T1Z1 (Canada) Fax: (+1)604-822-2847 E-mail: [email protected]