Hyaluronic Acid Matrices for In Situ Measurement of Protein Diffusion Coefficients

In vitro measurement of protein diffusion within matrices that simulate the subcutaneous (SQ) environment is of interest, given that protein-based therapeutics formulated for SQ injection comprise the largest class of biologics. To mimic the in vivo transport of a biologic from the SQ injection site through the extracellular matrix (ECM), in vitro diffusion assays typically utilize hyaluronic acid (HA) matrices, as it is the principal component of ECM. However, broad utility has been hampered by inherent lot-to-lot variability in commercially sourced HA, wherein key properties that impact protein diffusion (for example, molecular weight distribution and viscosity) differ across lots, even when nominal molecular weights are identical, making it challenging to compare results across matrices prepared from different HA lots. To address this gap, we report a facile approach wherein binary HA blends generated from individual HA matrices derived from distinct HA lots are functionally equivalent with respect to protein diffusion, that is, the diffusion of a representative set of proteins matches that in a previously reported single HA lot-derived matrix that served as a representative reference. Taken altogether, our protocols enable preparing blended HA matrices with consistent diffusion properties, enabling the use of in vitro assays that leverage this capability.

Practical application: The measurement of in vitro diffusion of IgG-type proteins enables calculation of diffusion coefficients that could help to guide the formulation of protein-based therapeutics, administered by subcutaneous (SQ) injection, and used for treating a range of diseases, including cancer. The side-by-side comparison of these proteins over a period of time provides confirmation of consistency of properties when in vitro hyaluronic acid matrices, within which injected protein diffusion is measured, are also consistent. However, their broad utility has been hindered by the inherent variability of commercial sources of HA used to make-up matrices that simulate the SQ environment in a predictable manner. Our research addresses this gap by defining an approach (validated with rheological and diffusion measurements) that facilitates the preparation of blended matrices from different lots of HA. The resulting matrix properties enable reliable measurement of protein diffusion from one lot to the next.