Multivalent ions for tuning the phase behaviour of protein solutions

Abstract

Protein phase behaviour is of importance in various areas of research such as structural biology, rational drug design and delivery, medicine (in particular protein condensation diseases), biotechnology, food science and cell biology. A particularly intriguing variety of phase behaviours can be induced in negatively charged, globular proteins in the presence of multivalent salts such as lanthanide (Ln) chlorides. These behaviours include reentrant condensation, crystallisation and cluster formation as well as liquid-liquid phase separation (LLPS) into a protein-rich and a protein-poor phase. LLPS can occur upon a temperature decrease or increase, which is referred to as an upper or a lower critical solution temperature (UCST- and LCST-LLPS), respectively. In the present thesis, the complex phenomenon of LCST-LLPS in systems of bovine serum albumin (BSA) and multivalent salts is investigated from different perspectives including thermodynamic, (non-)equilibrium and spectroscopic studies in a challenging set of experiments. In the first part of this thesis, the rather unusual phenomenon of LCST-LLPS in aqueous systems consisting of BSA and yttrium chloride (YCl3) is characterised thermodynamically. Surface charge (zeta potential) and isothermal titration calorimetry (ITC) measurements show LCST-LLPS to be a hydration entropy-driven condensation. As the Y3+ cations bind to negatively charged residues on the protein surface and bridge protein molecules, highly ordered water structures around both the surface residues as well as the cations break up and water molecules are released into the bulk solution. This leads to an increase of the overall entropy of the system. Starting from the Y3+-induced LCST-LLPS described above, the second part of this thesis is concerned with the influence that the nature of the multivalent cations used has on this phase transition. The experiments focus on the three multivalent salts HoCl3, YCl3 and LaCl3. Temperature-controlled UV-Vis absorbance measurements demonstrate that the transition temperature T_trans separating homogeneous from phase-separated states of the BSA-salt systems shifts to lower values when HoCl3 is used. In contrast, using LaCl3 leads to higher T_trans values. YCl3, used as a reference system, leads to intermediate T_trans. These findings indicate that the interprotein interactions induced by HoCl3 are much stronger than those induced by LaCl3 or YCl3. Importantly, this finding is corroborated by synchrotron small-angle X-ray scattering (SAXS) data which show the reduced second virial coefficient B_2/B_2(HS) to be lowest in BSA-HoCl3 systems, again pointing towards a stronger interprotein attraction induced by this salt. Zeta potential measurements confirm that Ho3+ has a stronger affinity to BSA than Y3+ and La3+. The overall protein-protein and protein-cation interaction strengths can therefore be ranked according to the order Ho3+ > Y3+ > La3+. Taking into account their various characteristics such as radius, electron configuration and, importantly, hydration behaviour, multivalent cations are thus shown to be a sensitive tool to fine-tune protein interactions and their resulting phase behaviours in solution. Having established the influence of cation characteristics on BSA phase behaviour, the third part is concerned with the kinetics of LCST-LLPS of BSA in the presence of varying ratios of HoCl3 and LaCl3. Using synchrotron ultra-small-angle X-ray scattering (USAXS), it is found that with an increasing HoCl3 concentration --- i.e., with increasingly attractive BSA-BSA interactions --- the growth behaviour of the characteristic system length χ(t,T) is more likely to deviate from the χ~t^{1/3} growth law. A stronger interprotein attraction, moreover, leads to arrested states at lower temperatures. The results imply that both temperature and the overall cation-mediated protein-protein interaction strength can be used to obtain multidimensional control over the kinetics of LLPS in the BSA-cation systems used. In the final part, the mechanism behind LCST-LLPS is investigated spectroscopically on the molecular scale. To this end, the change in the coordination number (CN) of Y3+ cations in BSA solutions is monitored using extended X-ray absorption fine structure (EXAFS) spectroscopy. Applying this method to protein-poor and protein-rich phases, it can be shown that the CN of Y3+ is higher in the protein-poor than in the protein-rich phase. This is attributed to the fact that in the protein-poor phase more Y3+ cations are surrounded by hydration shells and not bound to BSA or forming cation bridges between BSA molecules. The results obtained using EXAFS align well with the current rationalisation of LCST-LLPS as a hydration entropy-driven phenomenon. The results obtained indicate that a careful choice of the multivalent cation used can fine-tune protein interactions and their phase behaviour in solution. In addition, EXAFS data provide atom-level insights into the mechanism of LCST-LLPS. These findings are of strong interest not only for a fundamental understanding of protein and soft matter thermodynamics, but are also potential anchoring points for the design of stimuli-responsive “smart” materials based on polymers, colloids or proteins.