Targeting the active site of the placental isozyme of alkaline phosphatase by phage-displayed scFv antibodies selected by a specific uncompetitive inhibitor

Abstract

Background: The isozymes of alkaline phosphatase, the tissue non-specific, intestinal and placental, have similar properties and a high degree of identity. The placental isozyme (PLAP) is an oncofetal antigen expressed in several malignancies including choriocarcinoma, seminoma and ovarian carcinoma. We had earlier attempted to isolate PLAP-specific scFv from a synthetic human immunoglobulin library but were unable to do so, presumably because of the similarity between the isozymes. In this work, we have employed a PLAP-specific uncompetitive inhibitor, L-Phe-Gly-Gly, to select isozyme specific scFvs. An uncompetitive inhibitor binds to the enzyme in the presence of substrate and stabilizes the enzyme-substrate complex. Several uncompetitive inhibitors have varying degrees of isozyme specificity for human alkaline phosphatase isozymes. A specific uncompetitive inhibitor would be able to unmask conformational differences between the otherwise very similar molecules. Also, such inhibitors would be directed to regions at/close to the active site of the enzyme. In this work, the library was first incubated with PLAP and the bound clones then eluted by incubation with L-Phe-Gly-Gly along with the substrate, para-nitro phenyl phosphate (pNPP). The scFvs were then studied with regard to the biochemical modulation of their binding, isozyme specificity and effect on enzyme activity.

Results: Of 13 clones studied initially, the binding of 9 was inhibited by L-Phe-Gly-Gly (with pNPP) and 2 clones were inhibited by pNPP alone. Two clones had absolute and 2 clones had partial specificity to PLAP. Two clones were cross-reactive with only one other isozyme. Three scFv clones, having an accessible His6-tag, were purified and studied for their modulation of enzyme activity. All the three scFvs inhibited PLAP activity with the kinetics of competitive inhibition. Cell ELISA could demonstrate binding of the specific scFvs to the cell surface expressed PLAP.

Conclusion: The results demonstrate the biochemical modulation of scFv binding. Also, the scFvs bound to the active site and denied the access to the substrate. The selection strategy could generate specific anti-enzyme antibodies to PLAP that can potentially be used for targeting, for modulating enzyme activity in in vitro and in vivo and as probes for the active site. This strategy also has a general application in selecting antibodies from combinatorial libraries to closely related molecules and conformations.

Figure 1

Flow Chart of the selection strategy.

Figure 2

Polyclonal MBPhE. 10¹¹ pfu of each round phage eluted with L-Phe-Gly-Gly + pNPP were checked for binding affinity to PLAP-conjugated magnetic beads (solid bars) against the background binding of phage to the unconjugated beads (white bars). The black bars on top indicate SD.

Figure 3

PLAP binding activity of the 30 phage clones (of the 100 tested) that showed positive binding in the Magnetic Bead Phage ELISA. The solid bars represent binding of the each phage clone to the PLAP-conjugated magnetic beads. The empty bars represent binding activity of phage clones to the unconjugated magnetic beads (negative control).

Figure 4

Binding features of Type I, Type II and Type III antibodies on MBPhE. 4A. Representative binding of Type I antibodies on MBPhE. The binding of these antibodies was inhibited by L-Phe-Gly-Gly + pNPP but not by L-Gly-Gly-Gly + pNPP. The unconjugated magnetic beads served as a negative control. The black bars on top indicate SD. 4B. Representative binding of Type II antibodies on MBPhE. The binding of these antibodies was inhibited both by L-Phe-Gly-Gly + pNPP and L-Gly-Gly-Gly + pNPP and by substrate alone. The unconjugated magnetic beads served as a negative control. The black bars on top indicate SD. 4C. Representative binding of Type III antibodies on MBPhE. The binding of these antibodies was not inhibited by L-Phe-Gly-Gly + pNPP or L-Gly-Gly-Gly + pNPP. The unconjugated magnetic beads served as a negative control. The black bars on top indicate SD.

Figure 5

Characteristics of PLAP-scFv binding in the presence of competing isozymes. 5A. A fixed amount of biotinylated-PLAP was incubated with immobilized soluble scFv of the clone GLE4. The extent to which this could be inhibited by varying concentrations of competing isozymes was measured. Binding in the absence of any competing isozyme was taken as 100%. The binding to biotinylated-PLAP is inhibited only by unbiotinylated PLAP and not the other two isozymes of alkaline phosphatase. The labels on x-axis represent the molar concentration of the various isozymes included in the experiment. 5B. Competition of PLAP-GLD6 binding by IAP and BAP. Normalisation of the data was done as in Figure 6A. The binding to biotinylated-PLAP is inhibited by the two isozymes (BAP and PLAP) but not the third. The labels on x-axis represent the molar concentration of the various isozymes included in the experiment. 5C. Competition of PLAP-GL2D binding by IAP and BAP. Normalisation of the data was done as in Figure 6A. Binding to biotinylated-PLAP binding was inhibited by all the isozymes of alkaline phosphatase. At the maximum concentration of competing isozyme studied, the inhibition was the same. However, there was a slight difference in the relative affinity when 50 % inhibition was calculated. The labels on x-axis represent the molar concentration of the various isozymes included in the experiment.

Figure 6

Lineweaver-Burk Plot of inhibition of PLAP activity by purified GLE4. The enzyme activity of PLAP is inhibited in a competitive manner. Km of the enzyme: 5.5 mM; Km of the enzyme with GLE4: 40 mM. Inset shows % inhibition of enzyme activity by varying amount of purified scFv GLE4 (This clone also shows a higher binding [by ELISA] to PLAP as compared to BAP or IAP).

Figure 7

Lineweaver-Burk Plot of inhibition of PLAP activity by purified GL2D. The enzyme activity of PLAP is inhibited in a competitive manner. Km of the enzyme: 5.5 mM; Km of the enzyme with GL2D: 25 mM. Inset shows % inhibition of enzyme activity by varying amount of purified scFv GL2D (This clone also shows a higher binding [by ELISA] to PLAP as compared to BAP or IAP).

Figure 8

Lineweaver-Burk Plot of inhibition of PLAP activity by purified GLD6. The enzyme activity of PLAP is inhibited in a competitive manner. Km of the enzyme: 5.5 mM; Km of the enzyme with GLD6: 15 mM. Inset shows % inhibition of enzyme activity by varying amount of purified scFv GLD6 (This clone has a higher binding [by ELISA] to PLAP and BAP as compared to IAP).

Figure 9

Cell ELISA to study the binding of the isozyme specific scFvs to the cell surface expressed PLAP. The two PLAP-specific purified scFvs, clone E4 and clone GL2D showed significant binding to the HeLa cells that express PLAP (solid bars). SaOs-2 cells, that do not express PLAP, served as control (empty bars). The scFv VB1 did not show binding to either of the cell types and served as a negative scFv control. The black bars on top indicate SD.