Sandwich immunoassays, e.g. ELISA, are a mainstay of biological and clinical research and can offer accurate quantitative measurements of many different biomolecules. Several research groups and companies have adapted traditional sandwich immunoassays to make them applicable to detection and/or measurement of intact EVs. A variety of sandwich assay platforms have been demonstrated, including ELISA, electrochemiluminescence (ECL), fluorescent bead-based assays, and microscopy-based approaches.

©2017 Meso Scale Diagnostics, LLC. All rights reserved.

In a sandwich immunoassay for intact EVs, shown schematically in the figure above, a “capture” antibody is typically attached to a solid support and is used to capture EVs by binding a molecular feature present on the EV surface. One or more “detector” antibodies are used to decorate additional molecules on the EV surface and produce a signal that allows the captured EVs to be detected or quantified. Selecting various capture and detector antibodies enables configuration of many different sandwich assays for measuring EVs broadly or for measuring highly specific EV populations. A common approach is to use a capture antibody directed at a particular surface protein expected to be present on EVs of a population of interest, combined with one or more detector antibodies directed at commonly expressed EV surface proteins, such as CD63, CD81, or CD9. This configuration can also enable multiplexing when used with multiple capture antibodies displayed on a multi-spot ECL assay plate or bead-based multiplex assay platform and combined with a common detector antibody. Other configurations include the use of capture and detector antibodies directed at two distinct markers in order to detect highly specific EV populations.

When selecting capture and detector antibodies for intact EV sandwich assays, it is important to verify that the antibody is capable of binding the target molecule in its native conformation on the EV surface by using well-characterized EV samples, including positive and negative controls for the surface marker(s) of interest. Comparing each potential capture antibody with at least one negative control antibody is important in order to evaluate the ratio of specific EV capture to non-specific binding. Detector antibodies are typically titrated to evaluate the antibody affinity and optimize signal to background ratio (SBR). The EV Antibody database contains the results of these evaluations and should allow users to select capture and detection antibodies with confidence.

To generate the data in the Sandwich assay module, antibodies targeting putative EV surface markers were screened independently for their performance as capture or detector antibodies for intact-vesicle sandwich assays using MSD’s ECL-based assay platform.  Capture antibodies are biotinylated and up to 9 capture antibodies for a particular target as well as a negative control antibody are assembled on U-PLEX 96-well plates.  Intact EVs from a variety of relevant sources are assayed, including EVs from cell lines, primary cultures, and human biofluids.  Captured EVs are detected using ECL-labeled antibodies directed at the EV-associated proteins CD63, CD81, and CD9.  The assay format for comparing capture antibodies is shown schematically below.

©2017 Meso Scale Diagnostics, LLC. All rights reserved.

A comparison of ECL signal across antibodies for a given sample allows comparison of the relative overall ability of each antibody to capture EVs in this format. To facilitate comparisons between multiple capture antibodies and samples, we calculate a signal to background ratio (SBR) for each antibody/sample/detector combination, which is the ratio of the ECL signal with the sample of interest, minus the ECL signal of a blank sample (to account for non-specific binding between the detector antibodies and the capture antibody), divided by the ECL signal on the negative control antibody using the sample of interest. Antibodies with higher SBR values are most appropriate for EV capture.

Antibodies are similarly classified for their utility as detectors for intact EVs in a sandwich assay.  Here, EVs expected to display the surface target of interest are captured on ECL assay plates using a known high-quality capture antibody.  Typically streptavidin-coated plates or U-PLEX plates are used to display biotinylated capture antibodies targeting common EV-associated proteins, such as CD63, CD81, or CD9.  ECL-labeled detector antibodies targeting the marker of interest are used to label the captured EVs and produce a signal proportional to the number of copies of the surface marker present.We measure the affinity of each detector antibody by titrating the detector antibody concentration, and fitting the resulting ECL signals with a single-site binding model, as shown in below, which produces an EV-labeling Kd.

$$ECL=\frac{ECL_max*[Detector Antibody]}{Kd+[Detector Antibody]} $$

We also measure the non-specific binding of the detector antibody at each concentration by omitting the EV containing sample (blank). In general, each EV sandwich assay should be optimized by titrating the detector antibody concentration; however, the signal to background is typically maximized when the detector concentration is close to Kd, so we calculate a signal to background at Kd as a single figure to allow comparisons between detector antibodies. This is calculated by dividing the interpolated ECL signal at Kd for the EV sample, by the interpolated ECL signal at Kd for the blank.

$$𝑆𝐵𝑅([𝑑𝑒𝑡𝑒𝑐𝑡𝑜𝑟] = 𝐾! ) = \frac{𝐸𝐶𝐿(𝐸𝑉 𝑠𝑎𝑚𝑝𝑙𝑒, [𝑑𝑒𝑡𝑒𝑐𝑡𝑜𝑟] = 𝐾! )}{𝐸𝐶𝐿(𝐵𝑙𝑎𝑛𝑘 𝑆𝑎𝑚𝑝𝑙𝑒, [𝑑𝑒𝑡𝑒𝑐𝑡𝑜𝑟] = 𝐾! )}$$