Bypassing the Fundamental Trade-off Between Structural Order and Flexibility

Bypassing the Fundamental Trade-off Between Structural Order and Flexibility

Our paper titled "Hyperexpandable, self-healing macromolecular crystals with integrated polymer networks" can be found here:

Crystals are solid materials that have periodic spacing of their components at the atomic or molecular level. Salt, sugar, and snowflakes are all examples of crystalline materials. While the high level of order within a crystal lattice is often advantageous, especially for physical properties that rely on periodicity, most crystals are quite brittle and lack the flexibility for large-scale conformational changes. Amorphous materials, such as textiles and plastics, are much more malleable since they lack the precise arrangements found in crystalline materials. The lack of a highly ordered bonding network allows these materials to undergo significant changes both in size and shape. In the Tezcan lab, we frequently work with crystalline protein-based materials and always wanted to bridge the gap between flexible polymers and brittle crystals. Being chemists, we believed that if a protein lattice could be “chemically” bonded to a polymer network, we could obtain a singular hybrid material that would possess the best of both worlds, namely the structural order crystals and the flexibility of polymers. We thought this could be accomplished by taking a preformed protein crystal, filling the pores in it with polymer precursors, and then instigating these precursors to form a flexible mold around the protein molecules in the lattice.

One important aspect of constructing this hybrid material was choosing a suitable protein. We wanted a well-characterized protein crystal that contained large pores and had discrete crystal contacts that could be toggled on and off. With this in mind, we choose the cage-like iron storage protein, human heavy-chain ferritin, since the protein forms a highly symmetric crystal lattice with nanometer-sized pores. Unlike many proteins, ferritin is quite easy to crystallize. The protein forms micrometer-sized crystals in hours upon the addition of calcium. By removing the bound calcium, these assemblies can be converted back to free protein cages. For our amorphous component, we selected a hydrogel based on sodium acrylate and acrylamide. These components have been previously used to form hydrogels that can uptake water and swell to 100 times their original size. The monomers are much smaller than the pores of the ferritin crystal lattice and can readily diffuse into the preformed crystal. Once inside the lattice, we used a chemical reaction to polymerize these monomers, forming a hydrogel throughout the crystal lattice, and creating a crystal-hydrogel hybrid.

Our first test was to take a crystal-hydrogel hybrid and transfer it into deionized water. To our delight, the crystal started expanding slowly over the course of minutes! The crystal seemed to expand uniformly in all directions and contracted back to its original size after adding salt. This behavior is very similar to a polyacrylate gel. Contraction of the crystal-hydrogel hybrid occasionally lead to cracks, likely due to a temporary spatial gradient of salt throughout the lattice. Remarkably, these cracks spontaneously self-healed as the salt solution equilibrated throughout the material. This behavior is shown in this short video below:

Encouraged by these results, we wanted to monitor changes in the lattice periodicity at the atomic level using X-ray diffraction. We conducted a series of experiments to monitor the crystal-hydrogel hybrids at different stages throughout their expansion and contraction. Consistent with our visual observations, the unit cell of the crystal lattice expands in water and contracts with the addition of salt. The X-ray data led us to a more detailed understanding of our system. The atomic order of the lattice was lost immediately during expansion, but the long-range periodicity of the ferritin cages persisted significantly longer. After all of the crystalline order disappeared, the long-range periodicity could be restored by contracting the hybrid with the addition of a monovalent salt (i.e. sodium chloride). To our surprise, the atomic order was completely recovered after the addition of calcium or other divalent salts.   

After collecting single crystal X-ray diffraction data on these hybrids, we were surprised to see that many of the contracted crystals were better ordered than the native ferritin crystals we had previously characterized. We had unexpectedly obtained the highest resolution crystal structure of ferritin! We are very excited about the prospect of applying our methodology to improve traditional crystallographic techniques and enhance the resolution of the crystal structures of other proteins. We are also exploring different ways to exploit the unique properties of these hybrid materials.