Improvement of the affinity of an anti-rat P2X4 receptor antibody by introducing electrostatic interactions

Abstract

We have recently developed a mouse monoclonal antibody (12-10H) binding to the head domain region in rat P2X4 receptor (rP2X4R, which is crucial for the pathogenesis of neuropathic pain) expressed on the cell with the highest binding affinity (K_D = 20 nM). However, the 12-10H antibody failed to detect endogenously expressed P2X4Rs in microglia isolated from the spinal cord of rats whose spinal nerves were injured. Then, we prepared R5 mutant, in which five arginine residues were introduced into variable regions except for the "hot spot" in the 12-10H antibody to increase electrostatic interactions with the head domain, an anionic region, in rP2X4R. The mutation resulted in an increase of 50-fold in the affinity of the R5 mutant for the head domain with respect to the intact 12-10H antibody. As a result, detection of P2X4Rs endogenously expressed on primary cultured microglial cells originated from the neonatal rat brain and spinal cord microglia isolated from a rat model of neuropathic pain was achieved. These findings suggest a strategy to improve the affinity of a monoclonal antibody for an anionic antigen by the introduction of several arginine residues into variable regions other than the "hot spot" in the paratope.

Figure 1

Staining of rat P2X4Rs expressed on surface of living cells by fluorescent-labeled 12–10H antibody. Flow cytometry pseudocoloured scatterplots (upper panels) and histograms (lower panel) of living 1321N1 cells with or without P2X4R expression. These cells were stained with or without Alexa Fluor 488-labeled 12–10H or LKS103 antibody (40 μg/ml) for 30 min at 37 °C.

Figure 2

Homology model of the 12–10H antibody. (A) Primary sequence of the head domain (Gln111–Val167) in rP2X4R. (B) The location of mutated site in the model of 12–10H Fab based on the crystal structure of CD147 C2 domain in the complex with Fab of its monoclonal antibody 6H8 (PBD id 5X0T).

Figure 3

Fab preparation. Ion exchange of HPLC pattern of the refolded Fab from 12–10H (A) and the refolded Fab from R5 mutant (B). The column (Resource S) was eluted with a gradient of 1 M NaCl at a flow rate of 1 ml/min.

Figure 4

Binding analysis of antibodies to rHD. (A) ELISA analysis using the refolded Fab from 12–10H (colored in blue) and R5 mutant (colored in orange) as a primary antibody and alkaline phosphatase-conjugated anti-mouse Fab-specific antibody as a secondary antibody at 37 °C. B. Kinetic analysis of the interaction between 12–10H and rHD at 25 °C. (C) Kinetic analysis of the interaction between R5 mutant and rHD at 25 °C. SPR sensorgrams were fitted by global analysis (black lines) using the BIAevaluation software.

Figure 5

Detection of P2X4Rs expressed on surface of living rat microglial cells by 12–10H R5 mutant antibody. Flow cytometry pseudocoloured scatterplots (upper panels) and histograms (lower panel) of living primary cultured microglial cells. Cells were stained with or without Alexa Fluor 488-labeled 12–10H A3 or R5 mutant antibody (40 μg/ml) for 30 min at 37 °C.

Figure 6

12–10H R5 mutant antibody staining of P2X4Rs on living spinal cord microglia isolated from a rat model of neuropathic pain. (A) Flow cytometry pseudocoloured scatterplots of living microglia isolated from the spinal cord tissue of rats on day 14 after spinal nerve injury. Isolated living cells were stained with or without Alexa Fluor 488-labeled 12–10H WT, A3 or R5 mutant antibody (40 μg/ml) for 30 min at 37 °C. Numbers indicate percentage of CD11b and Alexa Fluor 488-gated cells (red square). (B) Histograms (lower panel) show overlay of P2X4R expression on spinal cord microglia, gated as in A. (C) Percentage of CD11b/Alexa488-gated spinal cord microglia (shown in A) (n = 3 rats). *P < 0.05 and **P < 0.01, one-way ANOVA with post hoc Tukey’s multiple comparison test. Data are the mean ± SEM.