(387cq) Optimizing ligand density and properties of charged membranes for concentrating siRNA by ultrafiltration

Small interfering RNA (siRNA) can inactivate mRNA associated with a wide range of diseases via site-specific binding and complex formation, a process known as RNA Interference (RNAi) that provides exciting opportunities in gene therapy applications. siRNA therapeutics have been successfully developed for treatment of neurologic disorders, atherosclerosis, and liver disease among others. However, siRNA typically requires large therapeutic doses and must be delivered at very high concentrations to meet the limits of subcutaneous injections. This creates significant challenges for the design and operation of the ultrafiltration process used for final formulation. Recent work in our lab has demonstrated that ultrafiltration with negatively charged membranes could concentrate siRNA to 180 mg/mL, more than 3 times the maximum concentration achieved using commercial (non-modified) cellulose membranes. This dramatic improvement in performance was due to the reduction in membrane fouling and concentration polarization arising from the electrostatic repulsion between the negatively charged membranes and the negatively charged siRNA.

The present study is focused on optimizing reaction conditions and ligand properties to explore the effect on the ultrafiltration performance during siRNA concentration. A series of charge modified cellulose membranes were produced by chemical modification with sulfonic acid functionalities over different reaction times (8, 24 and 48 h), ligand concentrations (1, 2 and 4 M), and spacer arm lengths (varying the number of carbons between the sulfonic acid group and the surface between 2, 3 and 4 carbon atoms). The membranes were characterized using X-ray Photoelectron Spectroscopy (to evaluate the Sulfur content) and using zeta potential measurements (to evaluate the effective membrane charge). Ultrafiltration experiments were performed over a range of feed siRNA concentrations in stirred cells at constant transmembrane pressure (TMP). The ligand charge density was higher at higher ligand concentrations and longer reaction times; however, the highest final siRNA concentration achieved during ultrafiltration was obtained using intermediate charging conditions (2 M ligand for 24 h). Membranes with higher degrees of modification showed better siRNA retention but suffered from lower initial fluxes, likely due to lower membrane permeability and pore-crowding steric effects. Spacer arm length was only optimized for the 2 M, 24 h condition. Increasing the spacer arm length from 2 to 4 carbons reduced the measured surface zeta potential from -13 to -10 mV, although the membrane with the longer spacer arm showed better performance during siRNA ultrafiltration. The effect of ligand chemistry on the membrane performance was further explored by replacing the sulfonate ligand with a carboxylic acid group. The resulting membranes, however, were less effective at siRNA ultrafiltration than those produced using the more strongly acidic sulfonic acid group. These results provide important insights into the factors controlling the performance and optimization of these novel charged ultrafiltration membranes for processing siRNA therapeutics.

Research Interests