Phage display can be used to prepare combinatorial protein libraries with up to 10¹¹ unique members. The tremendous sequence variability present in these libraries makes possible the identification of polypeptides that bind with high affinity (K_D < 10^-8 M) to virtually any target molecule. We have been developing a modular biosensing platform in which peptides on the surface of the M13 bacteriophage perform the recognition and binding functions. M13 virus particles are covalently attached to a self-assembled monolayer (SAM) covalently bonded to a gold surface (Fig. 1) to form a "covalent virus surface" (henceforth: "CVS"). Also present in the CVS are molecules of bovine serium albumin (BSA) intended to fill residual pin-holes in the SAM, and block the surface from nonspecific adsorption. Here, we report an in-depth examination of the properties of immobilized M13 and the CVS for the mass-based detection of a 148.2 kD antibody with binding specificity for the virus.

How much phage is on these surfaces? To answer this question, phage were detached from the CVS by photooxidizing the thiol SAM using UV light. Mass loss from a CVS immersed in PBF buffer commences promptly with UV light illumination and the ensuing loss mass transient is always characterized by a sharp drop in mass after which the mass stabilizes at a new value. The total phage coverage measured in this way was 3.0 microg/cm corresponding to 1.1 x 10¹¹ particles cm^-2 or the equivalent of 6.5 close-packed monolayers.

A calibration plot for the binding of p-Ab to the CVS was constructed by recording the mass increases associated with exposure of the surface to injections of p-Ab at concentrations varying from 6.6 nM to 200 nM (Fig. 2a). After each 180 microl injection, approximately 2400 s was allowed to elapse after which the bound p-Ab was removed by washing with 180 microl of 0.5 M HCl thereby regenerating the CVS. Each mass transient seen in Fig. 2a shows the same pattern: After each injection in the data set of Fig. 2a, the mass of the CVS increased for an average of 760 + 350s and then stabilized at a value proportional to the concentration of p-Ab (Fig. 2c). As indicated by its standard deviation of 350 s, this "rise-time" (to 90% maximum mass) was subject to considerable experimental uncertainty, but the mean value approximates the calculated 660s contact time of the p-Ab bolus with the CVS-modified gold QCM electrode. The CVS was then regenerated at after 2400s by injecting 180 microl of 0.5 M HCl using a higher flow rate of 100 microl/min. We also investigated the response of the CVS to a control antibody, n-Ab, which showed no measurable affinity for K07 phage in ELISA binding measurements (ELISA with n-Ab data not shown). The CVS showed no measurable mass increase upon exposures to n-Ab at concentrations up to 200 nM (Fig. 2b).

So it appears that this CVS surface may be a reasonably good affinity medium for biosensing. Now the issue is whether the tranduction of binding can be carried out using a more compact techique that, down the road, will be adaptable to nanowires.

---

Copyright 2008 R.M. Penner