Proteases are a large family of enzymes involved in many important biological processes. Quantitative detection of the activity profile of specific target proteases is in high demand for the diagnosis and treatment monitoring of diseases such as cancers. The overexpression and enhanced activity of proteases were found to correlate well with cancer development. Protease inhibitors are also a class of major anti-cancer drug candidates. However, due to the presence of a large number (over 600) of proteases in human body, their complex interactions with each other, and large influence by the surrounding factors (such as buffer composition, pH value and temperature), it is challenging to accurately detect the activity of specific proteases in complex media. It is even more difficult to measure the activity profiles of a large set of proteases that are relevant to cancers. Here we summarize our studies on the development of a multiplex electrode array biosensor platform toward cancer diagnosis based on rapid profiling of multiple protease activities. Both nano- and micro-electrode arrays have been demonstrated for electrochemical detection of protease activities. The individually addressed 3x3 gold thin-film microelectrode array is particularly attractive for rapid profiling of multiple protease activities.

The general sensor scheme involves attaching short peptides with specific sequences of 4 to 8 amino acids on the electrode surface of the electrochemical sensor chip, which serve as the substrates for selective proteolysis by the cognate proteases. The far end is covalently attached with a ferrocene (Fc) group as the redox tag for electrochemical measurements. The quantity of the intact peptides on the electrode surface is measured with an AC voltammetry (ACV) method, which is proportional to the peak current at the specific electrode potential when Fc oxidation is oxidized into ferrocenium (Fc⁺). As the peptide is cleaved by the cognate protease, the peak current decreases continuously in the continuously repeated ACV curves. The recorded kinetic proteolytic curve can be quantitatively described by the heterogeneous Michaelis-Menten model following an exponential decay. The inverse of exponential decay time constant, i.e., 1/τ, is found to represent the protease activity which equals to [E](k_cat/K_M), where [E] is the protease concentration and k_cat/K_M is the specificity constant defined by the kinetic proteolytic reaction constant k_cat and the Michaelis equilibrium constant K_M. The k_cat/K_M values have been investigated versus the substrate peptide sequence and length, temperature, and buffer composition to optimize the detection for activity of cathepsin B, a potential cancer biomarker. The limit of detection (LOD) for activated cathepsin B can reach down to 0.57 pM, sufficient to directly detect cathepsin B in diluted human serum.

This electrochemical method has been compared and validated with the commonly used affinity assay, i.e., enzyme-linked immunosorbent assay (ELISA). The ELISA measurements were found to be more sensitive to the concentration of proenzyme (zymogen), while the electrochemical method detects the activity of the proteases which reflects the intrinsic biological function. This provides more critical information than the concentration derived from ELISA.

Currently, two sets of peptides have been synthesized which are selective for detection of cathepsin B and ADAM-17, respectively. They are functionalized to the 3x3 gold thin-film microelectrode array and used to simultaneously detect these two types of proteases in serum samples from breast cancer patients at different stages. The activity profiles will be correlated to the cancer development.