With the COVID-19 pandemic continuing to dominate the headlines, people who have never once set foot in a lab have begun readily using the term “antigen-test” in conversation. Of course, what they are referring to is an immunoassay, most likely an ELISA, but seemingly overnight everyone has become an expert in this tried and trusted technique. Those who have spent a lifetime designing and validating immunoassays see things very differently, with many individuals becoming extremely passionate when discussing assay selection and best practices. Read on for some real-life examples and top tips for ensuring immunoassay success.

One size does not fit all

The design of an immunoassay depends not only on the nature and abundance of the target analyte, but also on the antibodies available, the required assay sensitivity and specificity, and the complexity of the sample material, according to Robert Hooper and Yu-Hung Huang, both senior scientists at Rockland Immunochemicals. “One common misconception about ELISAs is that they are all essentially the same assay into which any target antigen can be plugged in,” reports Hooper. “However, conditions for one analyte may not hold for another and rigorous optimization is vital. Even the type of wash buffer can affect the signal, meaning every component of the assay must be carefully considered.”

Dr. Tobias Polifke, co-founder and managing director at CANDOR Bioscience, agrees, pointing out that the vast majority of scientific publications describes using milk powder or BSA as a surface blocker, and a buffer comprising PBS/BSA/Tween as a diluent, leading to complacency when it comes to immunoassay design. “The ‘one-buffer-fits-all’ idea is a huge misconception,” he says. “For example, in a recent article describing a test for serum antibodies to COVID-19, researchers had to dilute the sample material 100-fold to overcome the high background associated with using milk powder as a surface blocker. In contrast, by carefully selecting a more suitable blocking reagent, we demonstrated a 10-fold improvement in the lower limit of detection simply by enabling a 10x dilution factor and at the same time much better assay sensitivity.”

It’s not just about the antibodies

Whenever an immunoassay goes wrong, researchers typically default to blaming the antibodies, says Dr. Michael Fiebig, vice president product portfolio and innovation at Absolute Antibody. Yet many other variables could also be in play. “Designing an immunoassay is like building a house,” he says. “You can have the best bricks in the world, but there are many more components that go into it that need to be carefully matched. Although headlines like “Blame it on the antibodies” have forced researchers to think more carefully about the antibody reagents they use, other factors I feel should move back into the limelight include the concept of controls in experiments and the basic requirements for immunoassay optimization.”

To illustrate how a well-chosen control can highlight the limitations of an immunoassay, Fiebig describes how Absolute Antibody recently evaluated some COVID-19 IgG/IgM tests intended to detect the presence of anti-virus antibodies in human serum. “When we used recombinant anti-SARS-CoV-2 spike protein antibodies in both human IgG1 and IgM formats, we were surprised to find that none of the tests worked,” he says. “In contrast, when we used our recently launched anti-SARS-CoV-2 nucleocapsid antibodies (also human IgG1 and IgM), the tests worked very well. Clearly the tests detected anti-nucleocapsid antibodies only, which is concerning for two reasons. First, research has shown that neutralizing antibodies appear to be directed against the spike protein. Second, there is far less diversity in the nucleocapsid protein than the spike protein, with many reports of anti-nucleocapsid antibodies being cross-reactive across coronaviruses. So not only does such a test leave you in the dark about the protective potential of any SARS-CoV-2 neutralizing antibodies detected in patient sample material, but it also opens up the possibility of falsely identifying SARS-CoV-2.”

Taking a step back from COVID-19, Sian Bolitho, commercial manager at Biorbyt, notes that thinking about the complex interactions that occur between different immunoglobulins is key when designing an immunoassay. “It should never be assumed that a primary antibody will bind the analyte, a secondary antibody will detect the primary, and nothing else will happen,” she explains. “Common mistakes include using a goat primary antibody on a western blot that has been blocked with milk (where the anti-goat secondary will react with milk proteins to cause high background), and not understanding that an antibody validated for IHC with FFPE tissue won’t necessarily work with frozen sample material (where freezing can distort epitopes). Instead of seeing the antibodies as the sole source of problems, more thought should be given to potential unwanted interactions, cross-binding, and indeed every other aspect of an immunoassay.”

Safeguarding antibody performance

In many instances, where antibodies are blamed for poor assay performance, the root cause is that they have been handled inappropriately. Most commonly, this involves storing them at the wrong temperature, with repeated freeze-thaw cycles being especially problematic. “If an antibody is stored at -20oC, consider making aliquots and thawing those on ice immediately prior to use to avoid repeated freeze-thawing,” suggests Huang. Conjugated antibodies also require careful handling; for instance, HRP is denatured by freezing, while a FITC-labeled antibody left on the lab bench will photo-bleach rapidly.

Batch-to-batch inconsistencies are another reason for antibodies to be accountable for skewing immunoassay data. “Researchers should never assume that each lot of an antibody will share the same or similar characteristics,” says Hooper. “Conducting bridging studies is advised to control for lot variance and can save considerable time and effort downstream by minimizing the likelihood of unexpected results.” Bolitho seconds this point, adding that although an antibody supplier will usually keep a record of which lot was sold, the onus is on researchers to record this information as they generate immunoassay data.

“To safeguard existing monoclonal antibodies against mutation, contamination, or accidental loss, we recommend antibody sequencing,” notes Fiebig. “Once you have a sequence, the very same antibody can be produced recombinantly for the entire lifespan of the immunoassay, guaranteeing long-term batch-to-batch reproducibility as well as an absolutely defined protein concentration and formulation.” Using recombinant antibodies also enables engineering to suit specific assay requirements, removing the need to use sub-optimal antibody pairings that could impact assay sensitivity and background.

Rigorous validation is key to reproducibility

Without the right validation process, it is impossible to trust immunoassay results. “A good validation strategy should be tailored to the intended use of the assay and based on a broad set of critical samples,” says Polifke. “These should include different matrices, different potentially cross-reacting substances, a number of real samples, and some spiking experiments, to name but a few. Although physical measurement techniques have come a long way since ELISA was first described in the literature, successfully designing and validating a reliable immunoassay still relies on the same basic principles: having the right binding reaction, using the right tools for biochemistry, and understanding any limitations of the assay as early in the process as possible.”

Tips and tricks

  • Consider the intended use of the immunoassay
  • Confirm the primary antibody binds the target
  • Check for cross-reactivities—this includes ensuring that secondary antibodies do not bind any other source of immunoglobulin in the assay system
  • Handle antibodies appropriately
  • Evaluate different surface blockers—try using modern alternatives
  • Consider using interference-reducing sample diluents
  • Think about using stabilizers or preservatives
  • Be aware that changing batches of ancillary reagents can affect experimental performance
  • Perform bridging studies to control for antibody lot variance
  • Switch to using recombinant monoclonals for guaranteed batch-to-batch reproducibility
  • Evaluate assay performance with a broad range of sample types
  • Include relevant controls and standard curves
  • Be consistent with timings, temperatures, volumes, and wash steps
  • Ensure reproducibility through multiple runs by multiple users
  • Keep detailed records
  • Challenge existing protocols
  • Don’t be afraid to ask for help