Static and dynamic properties of concentrated monoclonal antibody solutions

NAISS 2024/5-292


NAISS Medium Compute

Principal Investigator:

Peter Schurtenberger


Lunds universitet

Start Date:


End Date:


Primary Classification:

10402: Physical Chemistry

Secondary Classification:

10603: Biophysics




Monoclonal antibodies are highly effective for the treatment of immunological diseases as well as for cancer treatments. Unfortunately, they often present high viscosities in concentrated solutions, dramatically altering their flow properties. Therefore, preparing stable solutions suitable for subcutaneous self-administration that would be advantageous for patients and healthcare systems remains challenging. A coarse-graining modeling inspired by polymer and colloid physics can bring remarkable advantages for the investigation of the collective behavior of antibodies using computer simulations. Here we build on the previous project, where we have successfully established a coarse graining strategy and multiscale simulation protocol. Starting from the all-atom description of an individual antibody, a bead-based model is created, that takes into account the inherent anisotropic shape of the antibody, its excluded volume and an effective charge. We now plan to use the same approach to investigate other mAbs with more complex solution behavior, and also extend the simulations to dynamic properties. We aim at calculating static and dynamic properties of an ensemble of coarse-grained antibodies using molecular dynamics (MD) simulations. The work is integrated in an international consortium within the LINXS Institute of advanced Neutron and X-ray Science, which also involves the participation of the US National Institute of Standards and Technology and industrial pharmaceutical partners such as Novartis. After having appropriately mapped the features of the individual antibodies onto a coarse-grain bead model, we will first run equilibrium MD simulations at different salt and concentration conditions. We will compare two different methods for modeling charged domains, both implicitly, by means of effective screened Coulomb interactions, and explicitly, by also including counterions in the simulation box. Despite being computationally more expensive, a correct treatment of charged domains have been reported to be fundamental for other charged colloidal systems. The latter approach in particular will allow us to go beyond state-of-the-art as compared to current simulation methodologies for such systems. With this sets of simulations, we will be able to extract static structure factors, information on eventual clustering, diffusion coefficients and stress correlation functions, the latter being connect to the viscosity of the system via a Green-Kubo relation. Secondly, we will run non-equilibrium MD simulations under shear with the goal of directly extracting the viscosity of our samples. To the best of our knowledge, none of the previous work on the topic relied on a direct estimate of the viscosity with such simulations, but they have been always based on the use of phenomenological theories for an estimate of the viscosity. Our participation in the LINXS mAb research program will further allow to directly compare our findings with experimental results, thus providing a microscopic justification for the observed behavior. We expect our research to provide valuable insights on the reported increase in viscosity for concentrated antibodies samples, offering solutions for the clever design of antibodies and for guiding the formulation of therapeutics with desired physical response.