A differential cell capture assay for evaluating antibody interactions with cell surface targets

Abstract

Many biological and biomedical laboratory assays require the use of antibodies and antibody fragments that strongly bind to their cell surface targets. Conventional binding assays, such as the enzyme-linked immunosorbent assay (ELISA) and flow cytometry, have many challenges, including capital equipment requirements, labor intensiveness, and large reagent and sample consumption. Although these techniques are successful in mainstream biology, there is an unmet need for a tool to quickly ascertain the relative binding capabilities of antibodies/antibody fragments to cell surface targets on the benchtop at low cost. We describe a novel cell capture assay that enables several candidate antibodies to be evaluated quickly as to their relative binding efficacies to their cell surface targets. We used chimeric rituximab and murine anti-CD20 monoclonal antibodies as cell capture agents on a functionalized microscope slide surface to assess their relative binding affinities based on how well they capture CD20-expressing mammalian cells. We found that these antibodies' concentration-dependent cell capture profiles correlate with their relative binding affinities. A key observation of this assay involved understanding how differences in capture surfaces affect the assay results. This approach can find utility when an antibody or antibody fragment against a known cell line needs to be selected for targeting studies.

Fig. 1

Schematic diagram of a cell capture assay used to measure the relative binding affinity of antibodies. A grid is first stamped onto a dense streptavidin substrate slide, after which varying concentrations of biotinylated antibodies are added to different grid elements. Spots are blocked with a blocking buffer containing biotin in order to saturate the streptavidin binding sites. The substrate slide is subsequently incubated with a suspension of cells expressing a particular antigen, after which the cells are fixed and counted using a microscope. Cell capture by different immobilized antibodies is quantitated and used as a measure of relative binding affinities of the capture agents.

Fig. 2

(A) SDS-PAGE and (B) Western blot results. The four antibodies analyzed were biotinylated rituximab (1), rituximab (2), commercially biotinylated murine anti-CD20 mAb (3), and murine anti-CD20 mAb (4). Antibodies were detected in the Western blot with avidin-AP.

Fig. 3

MALDI-TOF results with rituximab, biotinylated rituximab, murine anti-CD20 mAb, and biotinylated murine anti-CD20 mAb (commercially and in-lab biotinylated). The full mass spectrum of rituximab is shown, along with highlights of the molecular ion peaks from the other spectra. The average extent of biotinylation for each antibody was calculated based on the molecular weights from the corresponding mass spectra. The average of the molecular ion peak, the doubly ionized mass peak, and the triply ionized mass peak was calculated to determine the mass of the antibody.

Fig. 4

Flow cytometry scatter plots of biotinylated rituximab and commercially biotinylated murine anti-CD20 mAb binding to CD20-positive Daudi and CD20-negative Jurkat cells. Fluorescence was detected with Streptavidin-PE. (A) Cells were incubated with no primary antibody and just Streptavidin-PE (negative control). (B) and (C) represent the binding of biotinylated rituximab and biotinylated murine anti-CD20 mAb, respectively. FL2-H, the vertical shift, measures the fluorescence intensity of PE.

Fig. 5

Antibody binding curve of biotinylated rituximab and commercially biotinylated murine anti-CD20 mAb to Daudi cells. The Y-axis represents the average of the mean fluorescence intensity (MFI) as measured by flow cytometry. Binding of biotinylated murine anti-CD20 mAb was detected with FITC-conjugated goat anti-mouse IgG AffiniPure F(ab’)₂ fragment (Fc_γ-specific), and binding of biotinylated rituximab was detected with FITC-conjugated goat anti-human IgG AffiniPure IgG (Fc_γ-specific).

Fig. 6

Microscope images of Daudi cells captured at different concentrations of (A) biotinylated rituximab and (B) murine anti-CD20 mAb (commercially biotinylated). The top (black background) image at each concentration point is a fluorescence microscope image based on a DAPI stain. The bottom is a bright field microscope image. (C) shows two bright field microscope images of Jurkat cell capture with biotinylated rituximab and murine anti-CD20 mAb (commercially biotinylated) at the highest capture concentration (54 × 10⁻³ mM) of antibody. All images are at 4× magnification (1 pixel = 2.56 μm²).

Fig. 7

Cell capture assay results for biotinylated rituximab and murine anti-CD20 mAb (commercially biotinylated) (A) and a comparison of the cell capture profiles of the two biotinylated murine anti-CD20 mAbs (in-lab and commercially biotinylated) (B). Standard deviations in (A) represent the average of four data points from a single experiment. The experiment was repeated three times for validity. The data in (B) was normalized to the biotinylated rituximab curve to demonstrate the difference between the two biotinylated murine anti-CD20 mAbs. Only Daudi cell capture was used in (B). The “Cell Density” parameter represents a measure of the total number of cells captured divided by the average pixel area of the cell capture spots in a digital image of the capture surface.

Fig. 8

An intact human IgG1 (Protein Data Bank accession code 1HZH) antibody showing solvent surface (A). A surface-rendering of this IgG with all domains visible and the Fc domain rendered for clarity (B). There are 88 lysines in 1HZH, 19 in the CH2/CH3 area that are exposed and possibly biotinylated (see insert) (C). An intact murine IgG2a (Protein Data Bank accession code 1IGT) zoomed in on the Fc domain (D). A demonstration of the 18 lysines in the CH2/CH3 area that are exposed and possibly biotinylated (E). An overlay showing that although there is some structural similarity and a fairly similar number of exposed lysines, there are localized areas (circled) where the ratio of potential biotins may tend to cause stronger off-axis orientations for the human IgG than the mouse IgG (F).