Phosphokinase antibody arrays on dendron-coated surface

Abstract

Monitoring protein phosphorylation at the cellular level is important to understand the intracellular signaling. Among the phosphoproteomics methods, phosphokinase antibody arrays have emerged as preferred tools to measure well-characterized phosphorylation in the intracellular signaling. Here, we present a dendron-coated phosphokinase antibody array (DPA) in which the antibodies are immobilized on a dendron-coated glass slide. Self-assembly of conically shaped dendrons well-controlled in size and structure resulted in precisely controlled lateral spacing between the immobilized phosphosite-specific antibodies, leading to minimized steric hindrance and improved antigen-antibody binding kinetics. These features increased sensitivity, selectivity, and reproducibility in measured amounts of protein phosphorylation. To demonstrate the utility of the DPA, we generated the phosphorylation profiles of brain tissue samples obtained from Alzheimer's disease (AD) model mice. The analysis of the profiles revealed signaling pathways deregulated during the course of AD progression.

Figure 1. A dendron-coated surface.

The 27-acid dendrons are coated on the surface. The chemical structure of the dendron is shown. After the immobilization and deprotection, the NH₂ group at the apex is activated with DSC to generate NHS group. Through the deprotection, the NH group denoted in red becomes the primary amine. Phosphosite-specific antibodies are immobilized at the apex of dendron molecules through the amine reaction with the NHS group. A proper lateral spacing (6∼7 nm) of the immobilized antibodies is obtained. The dendron density on the surface was estimated to be 0.03 dendrons per nm².

Figure 2. Signaling pathways represented by the proteins including 22 phosphosites.

The components involved in the 15 pathways labeled in bold were obtained from KEGG and BioCarta . The 22 phosphosites were denoted as the circles while the signaling components in the 15 pathways were denoted as the labels. The number after the residue represents the residue position in the human protein containing the phosphosite (Table S2). The arrows represent activation while the inhibition symbols represent inactivation.

Figure 3. Evaluation of sensitivity and selectivity.

A. Scatter plots between sample masses and measured intensities of four phosphosites in individual samples. The 11 sample masses hybridized onto the DPA in the ranges between 1.00×10² and 2.00×10⁵ ng are shown (X-axis). Signal intensities of four technical replicates for the four phosphosites were plotted with mean ± standard deviation (Y-axis). For each selected antibody, a linear relationship was observed (R²>0.93). B. Scatter plots between amounts of the antibodies and measured intensities. For a fixed amount of the analyte (Apo-I), the amounts of the antibodies were varied (X-axis). The signal intensity (Y-axis) detected by Ab-A specific for Apo-I was apparent while the intensity detected by Ab-B was residual. The image corresponding to the signal intensities is also shown (left).

Figure 4. Application of DPA to AD brain tissues.

A. Experimental scheme for phosphorylation profiling from AD brain tissues obtained from normal (Control) and AD mouse model (AD) at the ages of 2 and 6 months. B. Boxplots of the four DPSs whose alterations in AD, compared to control, were confirmed by Western blotting. ***P<0.001 from ANOVA followed by post-hoc tests with Bonferroni correction. C. Results of Western blotting of the four DPSs. Up-regulation of STAT3(Y705) and RelA(S534) at 6 months, and down-regulation of PKCδ/θ(S643/676) at 2 months and Akt1(S473) at 6 months were confirmed in independent samples (n = 3). Data are normalized to the β-actin abundance and presented as percentage changes from the control group. Data are shown as means ± SEM. *P<0.05 from ANOVA followed by Tukey's leat significant difference post-hoc tests.

Figure 5. A network model describing perturbed signaling pathways in the AD brain.

The nodes are arranged based on the signaling pathways in which the nine DPSs are involved (see text for the nine DPSs used for the network modeling). The node (center) and boundary colors represent the increase (red) and decrease (green) in phosphorylation measured by the DPA in AD samples at two and six months, respectively, compared to Control. The nodes corresponding to the four DPSs whose differential phosphorylation was confirmed by western blotting (STAT3, RelA, PKCδ/θ and Akt1) were denoted by large nodes. The arrows represent activation while the inhibition symbols represent inactivation. The dashed lines indicate interactions obtained from Alzpathway, and the solid lines indicate protein-protein interactions collected from BIND, HPRD, and BioGrid databases.