How do you raise an antibody against a protein you cannot purify?
Why recombinant antigen fails for these targets
A seven-transmembrane receptor has very little surface outside the membrane, and what there is only adopts its real conformation while the rest of the protein is held in a lipid bilayer. Detergent-solubilised or truncated constructs are easier to work with and produce antibodies that recognise something — often something that does not exist on a live cell.
The failure is usually discovered late. The campaign produces clones with good ELISA numbers against the recombinant construct, and then nothing binds cells. At that point the immunogen decision, made months earlier, is what needs to change.
How we do it
Immunisation uses peptides and/or cells rather than a purified protein. Panning is done directly on native or transfected cells, with rounds of subtraction against cells that do not express the target so that clones binding other surface proteins are competed away during selection. Surviving clones are confirmed by flow cytometry.
The consequence is that the clones we deliver have already demonstrated the property that matters — binding the target on a cell — rather than only binding a recombinant fragment. Performance on native tissue, where antigen density is lower than on transfected cells, still has to be screened separately.
- Cell-binding-confirmed clones, with sequences
- Flow cytometry data on target-positive and target-negative cells
- Recombinant material, recombinant and therefore reproducible batch to batch
Platform example: a rabbit monoclonal against the CB1 GPCR obtained by native-antigen panning. More detail on the method is on the native antigen panning page.
Where this also applies
- Conformational epitopes that only exist in the assembled, membrane-embedded state.
- Post-translationally modified surface targets where the modification is added by the cell — our published work on sulfated CCR5 used exactly this, panning on cells co-expressing the receptor and the relevant sulfotransferase.
- Rare-cell capture, where the antibody has to pick a population out of a mixture rather than bind a purified molecule.
- Targets with no commercial antibody at all — see our PLXDC2 work, where the antibody had to exist before the biology could be tested.
Have a target that will not express or purify?
Tell us the target, whether you have a cell line, and what the antibody has to do. A PhD scientist reads it and replies within two business days. No NDA needed to start.
Talk to a scientist →Frequently asked questions
Do I need to supply purified antigen?
No, and that is the point of this service. If the target cannot be purified or expressed as a well-behaved recombinant protein, we immunise with peptides and cells and pan directly on cells displaying the native target.
How do you avoid clones that bind the cell rather than the target?
Negative subtraction. Selection includes rounds against cells that do not express the target, so clones binding anything else on the cell surface are removed during panning rather than filtered out afterwards. Clones are then confirmed by flow cytometry on positive versus negative cells.
Can you work on GPCRs and ion channels?
Yes. Those are the typical cases for this route, because the conformational epitope only exists in the membrane. Our platform includes a rabbit monoclonal against the CB1 GPCR raised this way.
What do I need to provide to start?
The target identity, whether you have a cell line expressing it, and what the antibody has to do downstream. If no expressing line exists we can discuss transfected cell options at scoping.
Will the antibody work on native tissue, not just transfected cells?
It has to be screened for that. Overexpression on a transfected line gives far more antigen per cell than endogenous tissue, so a clone that looks strong on transfectants can be too weak on real samples. If native tissue is the endpoint, that goes into the screening cascade.