Question 9 of 20 Fundamentals Practical
How does a sandwich ELISA work, and what makes it specific?
Short answer
A capture antibody on the plate grabs the analyte, and a second, labeled detection antibody binds a different epitope on it. An enzyme then turns a substrate into a measurable signal. Specificity comes from needing two antibodies to bind the same molecule at once.
Why
The plate is coated with capture antibody, then blocked with protein so nothing else sticks to the plastic. Sample is added and the target binds. After washing, a detection antibody binds a second epitope, either carrying an enzyme or a tag such as biotin that recruits one.
The enzyme converts a substrate into a colored, fluorescent or luminescent product read on a plate reader. Concentration comes from a standard curve of known amounts, usually fitted with a four- or five-parameter logistic model. Thermo Fisher's overview stresses that the two antibodies must recognize non-overlapping epitopes, a matched pair.
Using the same antibody for both limits sensitivity. Requiring two binding events rejects many proteins that might stick to one antibody, which is why the sandwich format is the most widely used. Specificity is not guaranteed, though.
Heterophilic antibodies or rheumatoid factor in human serum can bridge the two antibodies with no analyte present. Very high analyte can saturate both antibodies and give a falsely low signal, the hook effect. Spike recovery and dilution linearity in the real matrix are how a lab shows the assay measures what it claims.
A strong answer covers
- Capture, block, sample, detect, substrate, read
- Matched pair binding non-overlapping epitopes
- Standard curve with a logistic fit
- Heterophilic antibodies and the hook effect
- Spike recovery and dilution linearity in the matrix
Common mistakes
- Assuming two antibodies make interference impossible in human serum
- Reading samples above the top standard instead of diluting them
- Fitting the standard curve with a straight line across its full range
The interviewer might ask next
How would you show that a sample matrix such as plasma is not interfering with your ELISA?