br Materials and methods br Results br

Materials and methods
Results
Discussion Discovery and characterization of cancer-associated markers effective in cancer diagnosis and treatment is a main focus in modern medicine. Several targets have been described [13,14]. Furthermore, we recently showed that the malarial VAR2CSA protein targets a specific type of placental-like chondroitin sulfate found in the placenta and on most cancer p2y receptor [27]. The field of medical science increasingly recognizes that a thorough molecular analysis of the biomarker recognition event is needed for full evaluation of a target for diagnostic and treatment purposes [11]. Such thorough analysis requires an in depth understanding of the molecular mechanism underlying the specific interaction as well as its kinetic properties. For this purpose scientists are utilizing a wide range of techniques including nuclear magnetic resonance (NMR) [12,26], mass-spectrometry (MS) [12], Enzyme-linked-immunosorbent-assay (ELISA), isothermal titration calorimetry (ITC) [15], and biosensor technology [1,5,19,22,24,31]. Common for these techniques is a requirement for the isolation and purification or recombinant expression of the target molecule. While these methods are valuable in describing a given biomarker and how to target it effectively, biomarker extraction is laborious and may interrupt key features only present in the biomarker\'s, or target\'s, native environment. We macro-dissected out a small circular piece of paraffin-embedded tissue and put it on COP-1 Attana QCM chip surfaces. We then removed the paraffin and performed antigen retrieval. One problem was that the tissue did not stick sufficiently to non-treated COP-1 surfaces. A coating with poly-l-lysine was needed for proper immobilization. The antigen retrieval needed was optimized by IHC staining using the analytes in question. Subsequently, the tissue-coated chips were treated in an identical way, allowing for direct comparison. Since no light can go through the gold COP-1 surface, we visualized the immobilized tissue by a DNA stain and a fluorescent microscope before and after the experiments to ensure the integrity of the tissue. No change in morphology was seen, showing that the proposed method maintains proper tissue integrity for subsequent analysis. The exact epitope of recognition of the anti-AR (N-20) antibody and its molecular nature is not known. It is likely that antibody specific optimization on tissue treatments is needed to accommodate analysis of other antibodies, such as antibodies targeting conformational epitopes. Having optimized the immobilization of the FFPE tissue, we performed binding experiments using rVAR2 and Anti-AR (N-20) monoclonal antibodies on tissues verified for the presence of the target antigen in question. All binding experiments were verified against appropriate cell lines, and in the case of rVAR2 also purified pl-CS. rVAR2 showed similar binding kinetics to purified placental-like CS, cell lines, and tissue. This confirms our strategy and supports the use of primary tissue for QCM experiments. It is common to evaluate specific binding in biosensor experiments by subtracting binding to a blank reference [25]. However, in working with immobilized cells and tissues we found that binding of an analyte to a blank chip is not representative of the background seen in binding to a chip coated with cells or tissues. When investigating the interaction between an analyte and a purified receptor in classic SPR, the surface area available for non-specific interactions on the chip can be considered comparable. This is not the case when investigating a chip covered in tissue or cells. Subtracting the signal from the reference chip therefore introduced a negative bias on the ‘true’ interaction, visualized as a negative spike in the real-time observation of the on-rate (Data not shown). Instead we selected cells and tissues negative for the ligand in question as proper negative controls, and presented them as separate figures to demonstrate the background independently. In this, rVAR2 did not interact with normal tissue (tonsil), confirming our previous observations that rVAR2 staining is placenta and cancer specific [27]. Furthermore, rVAR2 did not interact with the CS negative CHO-A475a cells and binding could be inhibited with soluble CSA, confirming the CS specificity. The analysis of the AR interaction showed the same correspondence with AR expression in both cell line and tissue setups. It is notable that such analysis was possible using a nuclear target.