Assessing mAb specificity prior to IND submissions


Ensuring that monoclonal antibodies are specific to a single, primary target receptor is increasingly being recognised as a critical step in the development of these essential biological therapies. Tissue cross reactivity (TCR) studies have traditionally been used to uncover potential off-target events that warrant further investigation. TCR shows binding that can occur in certain tissues, which, in turn, may indicate an off-target issue. While providing useful information as to the whereabouts of a potential cross-reactivity, this approach does not provide information on the identities (or even the number) of off-targets that may need to be considered.

Specificity screening using cell microarray technology identifies relevant interactions between an antibody and over 6,200 human proteins that are expressed in human cells. This highly specific platform reveals the number and, crucially, the identity of any secondary targets discovered. The incredibly low rate of false positives provides confidence that a screen will not return erroneous results and has led to the widespread acceptance of the data by both sponsors and regulators.


Antibody off-target screening – what data are produced?

The example presented here outlines a typical pre-IND off-target screen that was conducted on a human monoclonal antibody (mAb) provided by Simcere. The antibody has been developed for a solid tumour indication and is one of Simcere’s many candidates that are currently in preclinical or clinical development. In addition to its extensive pipeline, Simcere’s marketed pharmaceuticals span diverse therapeutic areas, including oncology, central nervous system disorders, and autoimmune diseases.

The primary target of Simcere’s test antibody was disclosed to Retrogenix prior to the study; however, only the anonymised data is shown below.

The Retrogenix platform

Retrogenix’s vast library of cDNA clones currently represents >6,200 human proteins. These include plasma membrane monomers, heterodimers (formed as a result of co-expression of the separate subunits) and secreted proteins (expressed with an inert cell membrane tether). Each cDNA is spotted in duplicate onto specialised slides and overlaid with HEK293 cells. These become reverse-transfected, resulting in clusters of cells each overexpressing a different, individual protein (or heterodimeric complex). Most screening studies performed using this cDNA expression platform follow a three-stage process.

  • Pre-screen: Levels of background binding to untransfected HEK293 cells are assessed using various concentrations of the test molecule to select the optimal concentration for full screening.
  • Library screening: The test molecule is screened for binding against the full library of over-expressed proteins. Typically, duplicate slides with duplicate cDNA spots are included, providing four opportunities for interactions to be detected using the platform.
  • Confirmation screening: Each “hit” from the library screens is re-expressed in duplicate, along with control receptors, and re-tested with the test molecule or the control treatments. This is often performed on live (no cell fixation) and fixed cells, and if any secondary target interactions are identified, these are often further investigated by flow cytometry as standard.

In the case of Simcere’s test antibody: full screening was performed at a concentration of 20 mg/mL. An isotype match control antibody, which was supplied by Simcere, was also included in the confirmation screening.

Specificity to primary target confirmed

Library screening revealed 15 initial hits, two of which were represented by more than one expression clone, resulting in 17 hits altogether. All these hits were re-expressed for a confirmation screen (results shown in Figure 1).


Figure 1. Confirmation screens of the Simcere test mAb and controls show binding of the test mAb to its known primary target (two clones) with no additional specific (off-target) interactions detected. An AlexaFluor647 anti-hIgG antibody was used for detection.   

After removing non-specific and non-significant hits, the test antibody showed a specific interaction with its known primary target only. This was seen in two distinct positions on the arrays, since the primary target (isoform 1) was cloned into two different vector backbones. The interaction between the test antibody and its primary target were detected in both fixed and live cell screens.

These data indicate high specificity of Simcere’s mAb for its primary target, providing data that can be used in critical decision making as well as included in IND submissions to regulatory agencies.


Specificity is a key consideration in the safety assessment of novel antibody therapeutics. Off-target screening using the Retrogenix platform can provide confidence in the specificity of an antibody; or it can provide crucial information on the number and identities of relevant off-targets.  Retrogenix screening data are widely accepted by regulatory agencies globally, and the reliability of the platform has led to increasing reliance on cell microarray results for cross-reactivity assessment, either alone or combination with other methods such as TCR.

We thank Simcere for their permission to reproduce the data presented here.

To discuss a project with a member of our science team please get in touch.