(603c) Effect of Protein Charge Distribution on Transport through Semipermeable Ultrafiltration Membranes

Several recent studies have clearly demonstrated that protein transport through semipermeable ultrafiltration membranes is strongly affected by electrostatic interactions between the charged membrane and the charged protein. All previous analyses of this phenomenon have implicitly assumed that the key factor is the protein surface charge density (i.e., net charge per unit area), with no effect of the detailed charge distribution over the protein surface. This is in sharp contrast to the behavior seen in ion exchange chromatography where the presence of charge "patches" can have a significant affect on the adsorption behavior. The objective of this study was to examine the possible effects of the protein charge distribution on the magnitude of the electrostatic interactions that occur during protein ultrafiltration.

Experimental studies were performed using cytochrome c and a peracetylated derivative formed by reaction of cytochrome c with acetic anhydride to block all free lysine groups. The peracetylated derivative had almost identical mass as the native protein but with an isoelectric point of pH 4 compared to pH 10 for the native cytochrome c due to the elimination of the positively-charged amino groups on the 19 lysine residues. The net charge of cytochrome c and the peracetylated derivative were determined theoretically, with the results in good agreement with data obtained by capillary electrophoresis. Ultrafiltration experiments were performed using 30 kD Ultracel membranes, both as received and surface-modified by chemical attachment of a quaternary ammonium functionality to generate a positively-charged membrane. The membrane charge was evaluated as a function of pH for both the native and modified membranes using streaming potential measurements.

The transmission of the native cytochrome c was maximum near its isoelectric point when using positively-charged membrane, but was a monotonically decreasing function of increasing pH with the unmodified (slightly negatively-charged) membrane. In contrast, the transmission of the peracetylated derivative increased with increasing pH for the positively-charged membrane but decreased with increasing pH for the unmodified membrane. These differences were a direct result of the different charge-pH profiles for the native and peracetylated proteins, giving rise to different electrostatic interactions. There were also distinct differences in the protein sieving coefficients for the native and peracetylated proteins having the same net electrical charge, which appear to be related to differences in the distribution of charge groups over the protein surface. These results provide important insights into the effects of electrostatic interactions on protein transport through ultrafiltration membranes.