Protein binding and non binding surfaces on biochips lab on chip 2012
- 1. Protein Binding and Non-Binding Surfaces on Biochips Jennins D, Baumgartner T, Abernethy N, Chung E, Fontanelle BT, Gao Y, Koudijs MM, Cooper S, Yang L, Ling T, Vukovic P, Savvina O, Wong A, Maeji,NJ Anteo Diagnostics Ltd, Brisbane, Australia; www.anteodx.com Introduction Results, Cont. Results, Cont. Design requirements for miniaturized and/or multiplexed assay systems require stringent device fabrication where materials such as silicon wafers, splutter coated metal oxides, ceramics and/or highperformance thermoplastics are used. Regardless whether the biochip is array- or microfluidics-based, its performance is critically dependent on certain regions having protein binding capability while other regions require the opposite. Many surface modification approaches have been developed many of which involve multi-step processing and are applicable to only a subset of materials and surfaces. Mix&Go converts a PS surface that passively binds proteins into a surface that binds proteins by chelation. Like with nano- and micron-sized particles, the consequence of this change is faster and more uniform protein binding that requires fewer capAb and has tighter uniformity between replicates. Synthetics Binding to Synthetics For example, IL6 sandwich assays on Greiner Low Binding PS plates that normally have inter-well CVs of >10% using passive binding gave %CV of 3.95% using Mix&Go. These %CVs are comparable to passive binding on the industry gold standard Nunc Maxisorp plates (%CV: 4.91%). The Greiner plate costs $0.55 while Maxisorp plates cost $4.92 (9x more expensive). 1.20 96-Well Plate 1.00 OD (450nm - 620nm) To simplify surface fabrication of such devices, we developed a family of metal polymer solutions called Mix&GoTM. Metal ions in polymeric form are able to interact by chelation with any material with electron donating potential. In isolation, each metal chelation point is not strong enough to anchor a biomolecule but the avidity benefits of the polymeric construct change almost any surface into one that gently but strongly binds proteins. 0.80 0.60 0.40 0.20 0.00 0 500 1000 1500 2000 2500 Fig. 2. IL-6 Sandwich assay on 96- (top) vs. 384- (bottom) well plates indicate that relative sensitivity performance of both Mix&Go formulations (compared to passive binding) improves with decreasing surface area (384-well). To show that Mix&Go can be used to produce synthetic coatings on top of other synthetic materials, gold colloids (50nm) were added to Mix&Go activated glass slides to form a thin gold film. Similarly, other biological or synthetic materials with electron donating groups can be fixed onto Mix&Go surfaces. Silica Oxides And Other Metal Oxide Surfaces Silica oxide, titanium oxide, aluminium oxide and nearly all other metal oxides have residual hydroxyl groups on their surfaces which allows direct chelation of Mix&Go to form an activated surface for protein binding. No chemical modification such as aminopropyl silanization is required e.g. glass slides are dipped into Mix&Go solution for as little as 5 mins, washed, and dried to give a Mix&Go activated glass microarray slides whose protein binding reactivity is maintained for at least 6 months (tested so far). Using Mix&Go coating Schott Glass B were comparable to the industry gold standard Schott Epoxy slides. Further optimisation studies are underway. Antigen (pg/mL) Mix&Go C10 1.20 1.200 0.40 1.000 0.20 200 100 0.00 0 500 1000 1500 2000 2500 Antigen (pg/mL) The ability to bind proteins a variety of surfaces also means the ability to bind other materials onto surfaces to manipulate its surface characteristics. For example, polymers having sufficient electron donating groups can be used to form thin films having completely different surface properties to the underlying substrate. If the electron donating potential is not high enough, small molecule tags that specifically bind Mix&Go sites can be coupled onto such polymers. By these approaches, non-binding surfaces can also be generated on all sorts of surfaces. Previously, we have shown on a number of different nano- and micron-sized particles that Mix&Go has given the following advantages compared to gold standard benchmark methods: Rapid and stable surface activation Significant savings in Abs and particle use for equivalent or better performance Increased sensitivity and dynamic range Excellent reproducibility and scalability. Using two types of Mix&Go with different polymeric structures (designated N10 and C10), the objective of our work was to demonstrate coating of proteins and other materials onto surfaces commonly used in the fabrication of biochips. Methods To activate surfaces for protein binding, the substrates in this study were immersed in Mix&Go C10 or N10 solutions from 10 min up to 1hr at room temperature. The substrates were then washed with deionised water and are ready for immediate use. Alternatively, activated Mix&Go surfaces can be stored after drying. Prior studies have shown that Mix&Go activated microparticles are stable for over 2 years. Preliminary studies on Mix&Go glass slides and Mix&Go PS plates indicate that these coated substrates are have comparable stability. Fig. 3. Using IL-6 Antigen at 5,000 pg/ml, cAb was titrated on Mix&Go Greiner Low Binding plates in comparison to Nunc Maxisorp using passive binding. Coupling time of capAb was 60 min. Not shown: above 1 ug/ml, both plates plateau out under the experimental conditions. 50 200 100 50 ug/ml ug/ml 0.800 0.600 Mix&Go N10 0.400 Passive (Maxi) As another example to prove its versatility, Mix&Go was coated onto 10-15nm iron oxide nanoparticles to form very small magnetic particles by an one step water based reaction. Similarly, any metal oxides whether as planer sheets, prefabricated shapes or as nanoparticles, can be easily activated to form a stable protein binding surface. 0.200 Millions 3 0.000 0 0.2 0.4 0.6 CAb (ug/mL) 0.8 1 R² = 0.98 1.2 2.5 2 Binding and Non-Binding Polymers and Plastics 1.5 1 0.5 Mix&Go can also be used to bind proteins to other synthetic substrates e.g., polyvinyl alcohols, polyhydroxyethylmethacrylates, sepharose and other carbohydrates commonly used as membranes. However, Mix&Go does not bind to polypropylene (PP), polycarbonates, cyclic olefin copolymers (COC), and other plastics with low levels of electron donating potential. To passively bind proteins, such plastics can be plasma- or -radiation treated to generate oxygen containing functionalities. Mix&Go also binds to such surface modified plastics and can present a more controlled protein binding surface by metal chelation. Protein Binding to Proteins The metal polymers in Mix&Go form a thin reactive surface (<1nm) to bind a monolayer of proteins. However, even the mildest binding conditions cannot maintain protein functionality if the remaining regions of the protein are exposed. To improve the microenvironment of proteins co-coupling with other proteins is used to improve distribution and functionality. Alternatively, a 3D scaffold created from proteins or synthetic polymers, e.g., nitrocellulose, can help maintaining antibody function. To show that Mix&Go can be used to bind proteins on proteins, it was used to bind antibodies to a protein surface (BSA) immobilised on beads by Mix&Go. The resulting outcome was near 50% increase in antibody loading compared to direct antibody binding to Mix&Go beads. Thousands Fig 4. BSA was coupled to Mix&Go activated M270 Dynabeads, and Mix&Go activation was repeated onto these BSA beads. Compared to antibody directly bound to Mix&Go beads, total antibody loading on antibody-BSA-beads was approx 50% higher by anti-species Ab binding. 300 Results 250 Mix&Go on Polystyrene (PS) Surfaces 200 MFI Mix&Go binds proteins gently but strongly, thus maximizing the number of functional proteins per unit surface area. If there is a sufficiently large surface area, e.g. in the wells of a 96-well microtiter plate, even if the majority of immobilized capture antibodies (capAbs) are damaged sensitivity of the assay is not significantly impacted. However, with decreasing surface area - such as on biochips or in the wells of 384-well plates - passive binding of antibodies to the PS surfaces can become a critical sensitivity limiting step. Fig 5. Mix&Go (A) activated Schott Glass B slides (plain glass) were spotted with antibodies at 200, 100 and 50 ug/ml. Spotting volume was 2 nL. These slides were compared to Epoxy slides (B) spotted under recommended conditions of manufacturer. cAb Titration on Mix&Go vs. Passive 0.60 RLU M M 0.80 150 100 50 0 Beads+M&G+BSA Beads+M&G+ FITC+M&G+ mouse IgG mouse IgG Bare beads Beads+M&G+ BSA FITC 0 0 500 1000 -0.5 1500 2000 2500 3000 3500 Fig 6. Chemiluminescense sandwich assay using anti-TnI antibodies bound to Mix&Go iron oxide (1015 nm) nanoparticles. LoD was 160 pg/ml in this first unoptimized experiment and can be improved upon. A loading assay showed no antibody binding to the bare iron oxide particles. pg/ml TnI Blocking Residual Mix&Go Sites BSA and other common protein blockers can be used to block Mix&Go sites remaining after capAb binding. Alternatively, non-protein blockers for Mix&Go are available. Polyethylene glycol (PEG) polymers of 2/5 kD were functionalized at one end with a modified amino acid fragment that was selected for strong binding to Mix&Go. PEG and other hydrophilic polymers with Mix&Go binding motifs can be used to generate low NSB regions around protein immobilized surfaces. Thousands 160 140 120 100 80 60 40 20 0 RLU M B 384-Well Plate OD (450-620nm) M M A Passive 1.00 OD (450nm-620nm) M M Fig.1. By avidity, the polymeric metal ions (Mix&Go) chelate to available electron donating groups on the synthetic substrate surface. Since not all the chelation potential is used to bind Mix&Go to the substrate, the remaining chelation points of Mix&Go are available for protein binding. Therefore, Mix&Go can be considered a molecular glue. Mix&Go N10 PA2.5 PA2.25 PA11 30ug BSA/mg beads No Fig 7. Shows different Mix&Go specific blockers. Blocking efficiency of BSA is compared to three different PEG blockers (called PAs) designed to specifically bind Mix&Go. These blockers were tested for their ability to block antibody binding (at 1 mg/ml) in comparison to a BSA control. 10ug Ab/mg beads Conclusion We have used an aqueous metal polymer solution called Mix&Go to create protein binding and nonbinding regions on all materials routinely used in the manufacture of biochips. Coating surfaces with Mix&Go is an easy and fast process which is identical for all surfaces evaluated. Non-protein blockers make it possible to render defined areas non-binding, an important feature for biochip and medical device manufacture. Mix&Go facilitates layer-by-layer deposition and the creation of new thin films, e.g. gold. Mix&Go will prove to be a easy-to-use, affordable and universal solution in biochip development and manufacture.