Detection of plasma proteins in multicomponent protein films by atomic force microscopy

Introduction: The success of long-term blood-contacting implanted devices greatly depends upon the interaction of the blood components with the device material. The ability to study these interactions has been hindered by a lack of methods to measure single-molecule interactions with these materials particularly in light of the highly complex multi-protein layers that form following blood contact. Recent advances in atomic force microscopy allow detection of proteins on model surfaces using antibody-coupled probes that measure antibody-antigen interactions.1 In this study, we continue to develop techniques for molecular identification and seek to extend these tools to rough polymeric materials, utilizing poly (dimethylsiloxane) (PDMS) as a model biomedical polymer, with the ultimate goal of developing techniques for single molecule detection in complex multi-component protein films on complex polymers used in medical applications. Materials and Methods: Triangular silicon nitride cantilevers with integral probe (k=0.06N/m²) were modified with polyclonal antibodies against human fibrinogen as described previously. Protein patterns were prepared using PDMS stamps consisting of an array of pillars having heights of 0.7μm and diameter of 0.7μm. PDMS stamps were incubated with bovine serum albumin (BSA, 250μg/mL) for 1 hr, and BSA was patterned onto either hydrophilic muscovite mica or a plasmacleaned PDMS by microcontact printing. The patterned sample was subsequently incubated in 100μg/mL fibrinogen solution for 1 hour to fill in the patterned areas. All measurements were made using a Nanoscope IIIa Multimode AFM (Digital Instruments, CA) operating under aqueous buffer. Sulfo-N-Hydroxy-Succinimido Nanogold (Nanoprobes) was conjugated to polyclonal rabbit antifibrinogen using the manufacturer supplied protocol. Patterned samples were prepared as described previously and then incubated with gold-labeled antibodies for 30 minutes. The distribution of the conjugated nanogold particles was visualized by phase imaging. Results / Discussion: Hydrophilic muscovite mica was chosen as a model surface. It is molecularly smooth and is easily patterned by microcontact printing. Fig 1a and 2a show patterned samples imaged by adhesion force and phase imaging, respectively. Fig 1b illustrates an adhesion force map between the polyclonal anti-Fb AFM probe and the surface. A statistical algorithm was used to differentiate between specific and non-specific interactions, yielding a binary recognition map ( fig 1c) with specific interactions denoted by white pixels. The pattern is superimposed on the map. Separately, 2 component protein films were prepared and blocked by the Ab. This resulted in > 80% reduction of recognition events, demonstrating that contrast arises from the specific interactions. Subsequent experiments have utilized purified platelet integrin receptors as probes to demonstrate fibrinogen functionality rather then just location (manuscript in review). Fig 2 shows AFM phase images of a patterned sample on mica (fig 2a), the same sample after filling with fibrinogen (2b), and the sample after incubation with the gold bead conjugated antibodies (2c). The pattern is clearly visible after labeling. This method offers two advantages over the adhesion force technique, it can be acquired at much higher data resolution (512 × 512 pixels vs 32 × 32 pixels) and much faster (8 minutes vs 60 minutes). However, it is an indirect measure as opposed to the direct interaction measurements and effects of non-specific adsorption are not clear. Similar experiments were performed using air plasma-treated PDMS substrates. The patterned BSA substrate is seen (Fig 3a) the sample was backfilled with fibrinogen (3b) and adhesion maps taken with modified probes. Interactions were detected on the sample but patterns not easily visualized, primarily due to increased background forces (Fig 3c). Nano-gold particles conjugated with anti-fibrinogen were also used with PDMS. BSA patterns were seen (fig 4a) and fibrinogen was adsorbed (4b). However, phase imaging did not show the protein patterns. Current efforts are focused on application and optimization of these techniques to the polymer biomaterials. Summary: Molecular detection of individual proteins in multi-protein films on polymers is important for understanding blood material interactions. The AFM offers unique advantages for detecting single protein types in multi-component protein films through a variety of detection techniques, and even offers opportunity to understand function of proteins at the nanometer scale.