(687a) Nisin Integration Into PEO-PPO-PEO Triblock Copolymer (Pluronic®) Brush Layers

Biomaterials that resist protein adsorption and bacterial adhesion offer improved biocompatibility and safety. Brush layers of polyethylene oxide (PEO) are generally considered to be cell- and protein-repellent, as a result of steric hindrance, thermodynamic unfavorability, and masking of electrostatic and van der Waals interactions. Protein adsorption can initiate clotting and immune responses at the fouled surface, with potentially life-threatening effects. Bacterial contamination of medical devices such as central venous catheters has also been correlated with protein adsorption.

An anti-fouling layer can be produced through self-assembly of PEO-PPO-PEO Pluronic^® triblock copolymers on hydrophobic surfaces. These polymers consist of a central polypropylene oxide (PPO) block between two PEO chains. The adsorbed Pluronic^® is anchored by the hydrophobic association between the surface and the PPO center block, while the pendant hydrophilic PEO side-chains form the anti-fouling layer. Pluronic^® F108 (PEO₁₄₁-PPO₄₄-PEO₁₄₁) is a commercially relevant protein-repellent surface coating, which can be applied to a material by immersing it in an aqueous polymer solution. The terminal ?OH groups of the PEO chains can also be functionalized, yielding an end-group-activated Pluronic^® (EGAP), to which bioactive molecules such as nitrilotriacetic acid (NTA) or heparin can be covalently attached.

Despite the well-established exclusion of large proteins by these Pluronic-based PEO layers, recent experimental results and theoretical studies indicate that this anti-fouling effect depends greatly upon the size of the proteins, and the brush layer spacing. Small peptides, such as the 3.5 kDa lantibiotic nisin, have been shown to penetrate and remain stably entrapped within a self-assembled F108 PEO brush layer. These findings assumed that there was no exchange of the surfactant with adsorbed nisin. In this work, we set out to demonstrate that nisin can be stably integrated into covalently-immobilized F108 and EGAP-NTA brush layers.

F108 and EGAP brushes have proven to be very effective at repelling bacteria and proteins in model systems, but can be displaced by surfactants and blood plasma proteins. To ensure the stability of the PEO brush layers on substrates, F108 or EGAP was covalently immobilized on hydrophobic silica microspheres by the method of McPherson, et al. (J. Biomed. Mater. Res. 1997, 38(4), 289-302). Non-porous 1 µm silica microspheres and silicon wafers were functionalized by a 5% solution of either allyldimethylchlorosilane (ADCS) or trichlorovinylsilane (TCVS) in dry chloroform. While TCVS can undergo polymerization and forms relatively thick films, ADCS cannot polymerize and is expected to form monolayers. The allyl/vinyl-modified silica was then cured overnight at 110°C to stabilize the coating.

The hydrophobic vinyl/allyl silica was incubated overnight in a solution of F108 or EGAP (0.5% in PBS) to produce a self-assembled brush layer. The silica with adsorbed triblocks was then subjected to γ-irradiation under water or F108/EGAP solution. Radiation-initiated radical reactions of the double-bonds presumably covalently link the PPO centerblocks of the Pluronics^® to the modified silicon surface, while the hydrophilic PEO side-chains extend into the bulk aqueous medium. The wafers were then incubated for 4 h with nisin (5 mg/mL in phosphate buffered saline (PBS), pH 7.4), and then rinsed with protein-free buffer. To determine whether the positively charged nisin molecules could facilitate adsorption and/or retention of negatively-charged fibrinogen on implant surfaces, we used the same procedure to challenge nisin-contacted samples with human fibrinogen (2.5 mg/mL in buffer). Any loosely bound protein was then removed from the microsphere surfaces by incubation with PBS buffer containing SDS. Zeta potential of the microspheres was used to characterize the degree of nisin integration into the brush layer, as well as the presence of fibrinogen on the surface.

We also attempted to determine if fibrinogen was associated with the nisin or brush surface by a modified ELISA method. Flat silicon wafers were coated with γ-immobilized F108 by the methods described above, contacted with nisin, and then challenged with fibrinogen. HRP-labeled anti-fibrinogen antibodies were allowed to bind to fibrinogen on the brush surface, and residual fibrinogen was detected with a colorimetric HRP substrate (OPD).

Although TCVS and ADCS both produced very hydrophobic surfaces, zeta potential changes indicate that TCVS produces denser, more stable F108/EGAP brushes than ADCS. For this reason, all protein-binding experiments used TCVS-silica exclusively. Similar zeta potential before and after SDS washes show that the brush layer was stabilized by γ-irradiation. Integration of nisin into the brushes was observed as a large increase in zeta potential due to the net positive charge of nisin. Introduction of fibrinogen caused a decrease in the zeta potential, consistent with its net negative charge. This decrease was substantially more pronounced for TCVS-modified silica in the absence of F108/EGAP coatings, suggesting an enhanced resistance to nisin elution due to its inclusion into the PEO brush layer.

The ability of nisin and, presumably, other small peptides to integrate into PEO brush layers offers a simple and effective mechanism to produce cell- and protein-repellent coatings that also provide specific functionalities, in this case a biocidal function. Ongoing research in our lab aims to determine the exact mechanism of the observed integration of nisin into PEO brush layers, as well as experimental parameters that affect the ability of peptides to integrate into brushes. We are also currently working on studies to quantify the activity and stability of small peptides within the brush layer.