A nanopatterning technique, termed femtosecond laser nanomachining, was developed in order to study the effect of clustering, placement, and overall cell adhesive ligand density on cell function. The general patterning strategy was to render a quartz cell culture substrate non-fouling through the grafting of a pEG polymer, followed by femtosecond laser ablation of the polymer to generate regions into which cell adhesive peptides or proteins could be adsorbed. Specifically, a high density ultrathin pEG brush layer was synthesized directly from the surface using surface initiated atom transfer polymerization, (SI-ATRP) a living radical polymerization with a constant growth rate that provides very low polydispersities. As the film thickness varied linearly with time, demonstrated with in situ growth monitoring in a quartz crystal microbalance with dissipation (QCM-D), thickness was easily tunable, and the resulting surfaces had very low roughness (less than 1 nm RMS) as verified by AFM. Surface chemistry was verified with x-ray photoelectron spectroscopy, (XPS) with a notable peak growth for energies associated with carbon-oxygen bonding, expected for a ethylene glycol rich polymer. Exposure of pEG modified quartz oscillators to various peptide solutions resulted in minimal frequency and dissipation changes, reaffirming that our film robustly inhibits protein adsorption.
Our pEG thin film was then ablated with 400 nm wavelength femtosecond laser pulses focused with various far field objectives to elucidate the ablation thresholds of both our film and the underlying quartz substrate. We identified an ideal processing window between fluences of 0.7 J/cm2 (polymer threshold) and 1.5 J/cm2 (quartz threshold), which results in clean film removal without damaging the underlying substrate. The effect of laser fluence on feature diameter for various film thicknesses and objective magnifications was characterized to map processing parameters for patterned sample generation. Protein adsorption into our ablated features was verified by fluorescence microscopy and AFM scanning samples prior to and post protein modification. Cell adhesion on a variety of avidin derivative proteins and biotinylated peptides were quantified using a fluorescence based assay to assess an optimal ligand presentation system from our surface. Neutravidin with biotin-bsp-RGD(15) was identified as an ideal adhesive ligand presentation system, due to maintenance of a high degree of cell adsorption with minimum background adhesion to the neutravidin itself. We used laser ablation to create a variety of substrates that are not easily accessible to other nanopatterning techniques, including one dimensional and two dimensional gradients of feature pitch, patterns designed to isolate single cells in different morphologies, and line patterns to control cell alignment and aspect ration.
To highlight the link between adhesion island size, pitch, and overall surface ligand density to the morphology and cytoskeletal development within the cell, gradient patterns and isometric pitch samples were created to isolate these variables. Gradient patterns of both adhesive island diameter and pitch were created in order to define a critical ligand density for cell adhesion, determined to lie in the range of 0.15-0.3 pmol/cm2 for the human mesenchymal stem cells on these patterned substrates. Patterns consisting of isometric pitch and diameter constrained within a non-fouling border were also created. For these patterns, rat mesenchymal stem cells were found to adhere down to densities of 0.03 pmol/cm2. Nuclear distension was quantified with the nuclear shape index metric (maximum cross sectional area over the high of the nucleus) and a statistically significant difference in NSI (p<0.05) was found for cells adhered surfaces on designed to project ligand densities of 0.03 pmol/cm2 and 1 pmol/cm2, suggesting the underlying surface density modulates intracellular tension and nuclear distension.