The screening of hybridoma supernatants for antibodies directed against cell surface antigens is important in the discovery and development of therapeutic antibodies. Here we report on the use of the ImageXpress® Velos System with a homogeneous assay format to quantify cell surface bound antibody for discovery and primary screening applications. The proprietary optics allows very sensitive detection of cell-surface bound fluorescence with excellent background rejection. The system acquires data over the whole-well and can scan a 384-well plate in under 5 minutes. Acquisition and analysis can be done in the presence of fluorescence backgrounds allowing true homogeneous cell surface binding assay for the detection of targetspecific human antibodies. In addition, we compared the use of a 488nm blue with a 640nm red laser with the same assay and appropriate fluorophore-labeled secondary antibodies.
Homogeneous Cell Surface Binding Assay. A flow diagram of the 384-well homogeneous cell surface binding assay is shown in Figure 1. The target specific human IgG, parental and target expressing CHO-S cells were provided by Medarex (Milpitas, CA). The cells were provided as frozen aliquots (10 x 106 cells/ mL) stored at -80o C. The vial of cells was thawed and the cells were immediately diluted into 5 mL of PBS containing 5% FBS. After 10 min room temperature incubation the cells were centrifuged at 800 x g for 2 min. The supernatant was discarded, the cell pellet was resuspended in 5 mL of cell dissociation buffer (enzyme free, Invitrogen, Cat # 13151-014) and a portion removed for cell counting. The cells were adjusted to 4 x 105 cells/ mL in cell dissociation buffer. A dilution series of the target-specific human IgG antibody was prepared from a 10 μg/ml stock solution in PBS containing 5% FBS. The cell binding assay was assembled by adding 20 μL/well of the dilution series of the target specific antibody in a 384-well black-walled plate (medium bind surface, Greiner Bio-One, Cat # 781096). Next, the parental or transfected CHO cells plus the anti-human IgG labeled with Alexa-Fluor 488 (AF488 conjugated anti-human IgG, H+L-chain specific, Invitrogen, Cat # A-11013) or with Cy5 (Cy5 conjugated anti-human IgG, Fcγ fragment specific, Jackson Immunoresearch Laboratories, Inc., Cat # 109-175-098) was added (40 μL/well). The cell binding reaction of 8000 cells/well was incubated at room temperature with the plate protected from light exposure.
Detection of Antibody Binding to Cells. The ImageXpress Velos System DL laser scanning platform used for this demonstration was configured with two excitation sources: a 20mW 488nm laser and a 50mW 640nm laser. The 488nm laser was used for AF488 detection using channels 1 and 2 (Ch1 & Ch2) with 510- 540nm (green) band pass filters and 560nm LP dichroic filter. The 640nm laser was used for Cy5 detection using channels 3 and 4 (Ch3 & Ch4) with 660-680nm (red) band pass filters. The image acquisition was done at 5 x 5 micron sampling and an entire 384- well plate was scanned. Images from the two channels were averaged together before processing as described below. The total scan and analysis time under these conditions is < 5 minutes.
Results & Discussion
Homogeneous Cell Binding Assays
A schematic of the ImageXpress Velos System optical setup that provides highly sensitive detection in homogeneous assay formats is shown in Figure 2. The intersection of the excitation laser with the collection optical path creates a confined detection region which limits the collection of the fluorescence to near the bottom of a well. This reduces the background fluorescence and improves the measurement of cell bound fluorescence signals.
Two symmetric collection heads provide a binocular detection setup and double the amount of signal collected. This improves the assay by increasing the S/N of the measurement and reducing well shadowing effects.
Image Processing and Analysis
The images of each well were processed to determine the total, or integrated, cellular fluorescence intensity. A region of interest (ROI) was defined in the image, an image flattening algorithm was applied, and a threshold was automatically determined based on background intensity levels in the ROI. The gray scale values of all pixels above threshold were summed and background corrected. This value was automatically generated after image acquisition by an integrated image analysis program. The plate results were reported in well-by-well basis in a text file and imported into Excel (Microsoft) using a macro developed by BBI for further analysis and data processing. Representative images taken with both the 488nm and 640nm lasers are shown in Figure 3.
The total fluorescence response measured as a function of primary antibody concentration at various incubation times is also shown in Figure 3. A time and antibody concentration dependent increase in fluorescence intensity is observed with both the AF488 and Cy5 assays. As commonly observed in homogeneous binding assays, the fluorescence intensity associated with cells decreases at high primary antibody concentrations. This is attributed to saturation of the cell binding sites and a correspondingly higher concentration of primary antibody in solution. If the 2nd antibody concentration is limited (as is the case in these studies) then as the primary concentration increases a larger percentage will bind to primary in solution and the cell fluorescence will decrease.
Differences in the kinetics of the binding for the AF488 and Cy5 assay systems were observed. For example, significant fluorescence intensity at the 30 min incubation time was only observed with the Cy5 based assay. The AF488 assay signal was not observed until 60 minutes. After overnight incubation (16 hr), both the AF488 and Cy5 assays had excellent signals with similar total intensities. However, the AF488 assay saturated at lower primary concentration and had a wider plateau region.
The ImageXpress Velos System enables a simple, fast and high throughput homogeneous assay format to quantify cell surface bound antibody for discovery and primary screening applications. As demonstrated in this report, image acquisition and analysis can be done in the presence of fluorescence backgrounds allowing a true homogeneous cell surface binding assay for the detection of target-specific human antibodies. In addition, we compared the use of a 488nm blue or a 640nm red laser with the same assay and appropriate fluorophore-labeled secondary antibodies. The results showed a similar pattern of antibody concentration dependent increase in fluorescence intensity after 60 and 90 min, and 16 hr incubation. This assay format has only two liquid handling steps and performance features suggesting that it is ready for use in 1536-well plates. The unique optics and scanning engine of this platform enables simple “plug and play” applications to meet the needs of life science researchers in both academia and industry. The assay robustness and fast read time also make it ideal for bioprocess control in development and manufacturing of therapeutic antibodies.