Affinity biosensor interface engineering for real-time biohazard monitoring

Abstract

Our environments abound with biological materials which can impact upon our health. The consequence of our exposure can range from mild irritation to acute and life threatening infection. To defend effectively against exposure to this spectrum of biohazards detection systems must precisely interrogate complex heterogeneous mixtures of biological entities sampled from the environment. Such a system must respond instantaneously and alert those at risk of exposure so that appropriate action may be taken. In principle, affinity biosensor technologies (e.g. surface plasmon resonance sensors) may be used within a detection system to continuously interrogate an aqueous sample (e.g. continuously delivered by an air sampler). The specific benefit of affinity biosensors is that, at the point of analysis, they are reagentless. This permits simple and automated operation without need of complex fluidics or mandrolic processing and, most importantly of all, on-line and near real-time analysis. All that is required to exploit these technologies is an affinity sensing interface which binds only the materials of concern and rejects any and all other materials. The concept is simple, but this belies the technical complexity of its realisation. Recognition elements (e.g. antibodies) are broadly available as binding ligands for application in affinity sensors. However, their integration and performance within such technologies, facilitated by other necessary enabling components (e.g. hydrated polymers), presents a long standing and multidisciplinary technical challenge. Current technologies are prone to false indications making them unreliable. For example; spontaneous non-specific biological adsorption compromises binding specificity yielding false positive results, and variable integration of recognition elements (e.g. random orientation) may compromise sensitivity yielding false negative results. Theoretical models suggest the necessary functionality required of affinity biosensor interfaces is a credible possibility (e.g. those built upon the polymer brush model described by de Gennes in 1980 [1]). However, even the most recent iterations of such models are not predictive of interfacial biological processes as the molecular behaviour of the solvent (i.e. water) is excluded (principally as it is both ill-defined and computationally challenging). There is, therefore, a need for further empirical exploration of such biological interfaces to understand how materials interact within such systems and how to fabricate them. The work herein has systematically, and at the molecular scale, engineered different components of the affinity biosensor interface. The aim has been to develop an interfacial system to enable real-time environmental affinity biosensing to become a practical reality, i.e. to resist all non-specific biological adsorption, and reproducibly integrate biological recognition elements. Surface plasmon resonance has been used throughout as a representative affinity biosensor technology. This class of affinity biosensors are sensitive and the transduction mechanism does not physically or chemically perturb the sensing interface (e.g. by introduction of mechanical or electrochemical processes) and is, therefore, unlikely to change the behaviour of the interfacial system which is being interrogated. This should enable facile data interpretation in pursuit of new design rules for biosensor interface engineering. Specific objectives were as follows; To engineer recombinant antibodies to simultaneously enable both site-specific chemical modification of the product, and facile manufacture within an industrial microbial expression system; To engineer thin-film polymer coatings to passivate a SPR interface so that it may resist all detectable non-specific adsorption from complex and heterogeneous biological mixtures; To demonstrate specific biological detection within complex and heterogeneous biological mixtures through the immobilisation of site-specifically engineered recombinant antibodies upon a non-specific adsorption resistant SPR interface. These objectives have been partially met. The principle outcomes are as follows; A novel protein engineering tool has been developed and, where E. coli periplasmic expression is used to manufacture recombinant antibodies, a new novel tag may be universally applied which does not reduce soluble yield. Once purified the intrinsic, labile chemistry within the tag may be selectively activated via the reduction of an intrinsic disulphide bond to yield two free thiols for subsequent manipulation. A novel affinity biosensor interface has been developed. All covalent interfacial construction methods assessed (e.g. ATRP or SAM deposition methods of zwitterionic or oligoethylene glycol functionalities) yielded thin-films which adsorbed protein non-specifically. These were discarded and a non-covalent, self-organising interface was developed. Adsorption from undiluted animal serum was found to be undetectable on this novel affinity interface. The properties of the self-organising molecular components are; they all present pendant oligoethylene glycol of between 7 and 10 repeats, as surfactants they reduce the surface tension of water to 30 mN.m-1 or below, they have hydrophobic functionalities which match the underlying hydrophobic interface, or they are more hydrophobic than the interface (though the surface must also be hydrophobic). The covalent integration of recognition elements was successfully demonstrated by the insertion of polyethylene ether tethers within the non-covalent self-organising coating. However, once again, these surfaces were found to foul non-selectively. The recurrence of non-selective adsorption is thought to be attributable to the physical behaviour of the tether where localised chain entropy is reduced. As far as the author is aware, this is the first observation of protein fouling that may be directly attributed to hydrated polyethylene ether which has not been driven by exogenous mechanical or chemical forces. The tools and understanding demonstrated through SPR sensing are anticipated to find broad utility across affinity sensors. This work presents a tangible start point to study the behaviour of complex soft matter assemblies, and to build more functional self-organising interfacial molecular systems. Prospective future work: The biosensor interface must be engineered as a concerted soft matter system. The integration of recognition elements within the non-covalent, self-organising coating must be undertaken in a way that sustains the entropy of the molecular tethers or, alternatively, completely shrouds them to sterically preclude fouling. However, considering the need to covalently immobilise the recognition element, and therefore reduce entropy at least at this specific point, it is possible that non-specific adsorption is simply a fundamental thermodynamic inevitability. This may therefore defeat the simple concept of real-time affinity biosensing, at least through mass sensing modalities. As a broad enabler of affinity biosensor platforms, the self-organising building blocks defined herein should be amenable to systemic molecular engineering to further integrate properties to suit varied transduction principles, e.g. redox activity. The opportunity to augment optical sensors with electrochemical functionality should enable confirmatory analytical tests to be undertaken upon the same sensor interface. The consequence of any residual non-specific adsorption may then be negated. The self-organising system may also be amenable to phase partitioning to spatially direct specific molecular species in the system to predefined locations, e.g. upon a hydrophobically patterned surface, or solvated components such as nanoparticles. In so doing it may be possible to further develop the self-organising concept to explore interfacial assemblies and molecular electronics which may obviate the need of the optical sensing modality (leading further to miniaturisation etc.). Beyond environmental biosensing, the interface may find application in basic research studying biological interactions in complex and congested molecular environments. This may be applied to generate insights into complex molecular processes such as; interaction dynamics of intrinsically disordered proteins, behaviour of DNA-protein interactions within the nucleoplasm, or amyloid formation in varied physiological environments. Quantitative insight into such biological soft matter systems may be a broad enabler of computational modelling to benefit bioengineering disciplines such as synthetic biology and medial therapeutics.