Comparative studies on the interaction of proteins with a polydimethylsiloxane elastomer. I. Monolayer protein capture capacity (PCC) as a function of protein pl, buffer pH and buffer ionic strength

Polydimethylsiloxane (PEP) is widely used in medical prostheses and therefore is in contact with plasma and secretory proteins. Two pair of globular proteins, lactoferrin (Lf) and transferrin (Trf), and bovine IgG1 and IgG2a, which differ substantially between pair members in their pl, were used to study the interaction of a PEP widely used in breast implants and soluble protein. Studies were done using iodinated proteins over a concentration range that resulted in an apparent protein monolayer. Secondary incubations with dilute protein solutions were needed to form the monolayer on PEP, possibly as a consequence of micro air bubbles trapped on its highly textured surface as shown by atomic force microscopy. Immunoassay quality polystyrene microtiter wells were used as controls. Adsorption studies were routinely performed at pH 4, 7 and 10 and at ionic strengths corresponding to 0.95, 9.5 and 90.0 mS. The protein capture capacity (PCC) of PEP for Lf and Trf was optimal at physiological pH and ionic strength and comparable under these conditions to that of Immulon 2 (Imm 2) microtiter wells. While increasing the ionic strength and pH further increases the PCC of Imm 2 for Lf and Trf, this markedly lowered the PCC of PEP for these proteins suggesting that initial polar interactions may precede subsequent hydrophobic bonding to PEP. This was tested using a hydrophilic variant of PEP, which when tested in a 90.0 mS buffer, showed a >five‐fold lower PCC at neutral and alkaline pH. The greatly reduced PCC of the hydrophilic variant might also suggest that hydrophilic variants of silicone would be more biocompatible than those currently used. The PCC of PEP for the IgGs was less than that of Imm 2 but still optimal at physiological conditions. Consistent with the data on Lf/Trf, PCC progressively decreased with increasing ionic strength at alkaline pH. Differences in pl between the protein pairs had only a marginal effect on the PCC of PEP. Monolayer adsorption on both PEP and Imm 2 was slowly reversible and greater in the presence of free ligand (<2% in 16 h) suggesting that the process follows Mass Law principles. However, even in the presence of non‐ionic detergent and free ligand, 85–90% remained bound on either surface. Thus, desorption of proteins in the monolayer should not complicate subsequent immunochemical studies conducted on adsorbed monolayers. © 1997 John Wiley & Sons, Ltd.