Antibodies can bind to cells in a specific manner – where the FAB portion of the antibody binds to a high-affinity specific target or the FC portion of the antibody binds to the FcR on the surface of some cells.
They can also bind to cells in a nonspecific manner, where the FAB portion binds to a low affinity, non-specific target. Further, as cells die, and the membrane integrity is compromised, antibodies can non-specifically bind to intracellular targets.
So, the question is, how can you identify and control for this observed nonspecific antibody binding?
To answer this question, many research groups started using a control known as the isotype control.
The concept of this control is that an antibody targeting a protein not on the surface of the target cells has the same isotype (both heavy and light chain) as the antibody of interest. When used to label cells, those that showed binding to the isotype would be excluded as they represented the non-specific binding of the cells.
Why Isotype Controls Often Fall Short
Isotype controls were once the most popular negative control for flow cytometry experiments.
They are still very often included by some labs, almost abandoned by others, and a subject of confusion for many beginners. What are they, why and when do I need them? Are they of any use at all, or just a waste of money?
Most importantly, why do reviewers keep asking for them when they review papers containing flow data?
Isotype controls were classically meant to show what level of nonspecific binding you might have in your experiment. The idea is that there are several ways that an antibody might react in undesirable ways with the surface of the cell.
Not all of these can be directly addressed by this control (such as cross-reactivity to a similar epitope on a different antigen, or even to a different epitope on the same antigen). What it does do is give you an estimate of non-specific (non-epitope-driven) binding. This can be Fc mediated binding, or completely nonspecific “sticky” cell adhesion.
In order to be useful, the isotype control should ideally be the same isotype, both in terms of species, heavy chain (IgA, IgG, IgD, IgE, or IgM) and light chain (kappa or lambda) class, the same fluorochrome (PE, APC, etc.), and have the same F:P ratio. F:P is a measurement of how many fluorescent molecules are present on each antibody.
This, unfortunately, makes the manufacture of ideal isotype controls highly impractical.
There is even a case to be made that differences in the amino acid sequence of the variable regions of both the light and heavy chains might result in variable levels of undesirable adherence in isotypes versus your antibody of interest.
Moving Beyond Isotype Controls
Many flow cytometry researchers are no longer using isotype controls, with some suggesting they be left out of almost all experiments.
If you spend any time browsing the Purdue Cytometry list, you’ll see these same arguments presented in threads about isotype controls.
A report in Cytometry A presents options for controls in several categories, the options available, and pros and cons of each option. The report's section on isotype controls summarizes the problems with the use of isotype controls very clearly.
A second report in Cytometry B presents options for controls in several categories, the options available, and pros and cons of each option.
The section of the above paper focusing on isotype controls summarizes the problems with their use very clearly.
The article also illustrates difference in undesirable binding at different levels using the same clone from different manufacturers.
For example, the figure below shows how even the same isotype control clone can result in highly variable levels of undesirable staining.
If you do use isotype controls in your experiment, they must match as many of the following characteristics as possible for your specific antibody — species, isotype, fluorochrome, F:P ratio, and concentration.
Here are 5 cases against using isotype controls alone...
1. Isotype controls are not needed for bimodal experiments.
You don’t need isotype controls for experiments that are clearly bimodal. For example, if you are looking for T cells and B cells in peripheral blood, the negative cells also in the circulation will provide gating confidence.
As seen in the second figure below, it is extremely easy to pick out CD4 and CD8 positive cells in the sample of lysed mouse blood.
2. Isotype controls are not sufficient for post-cultured cells.
If you are using post-cultured cells, the isotype control might give you some information about the inherent “stickiness” of your cells.
However, this measurement is not a value you can subtract from your specific antibody sample to determine fluorescence intensity or percent positive.
Instead, the measurement is simply a qualitative measure of “stickiness” and the effectiveness of Fc-blocking in your protocol.
3. Isotype controls should not be used as gating controls.
If you are using multiple dyes in your search, and your concern is positivity by spectral overlap, you will be better served by using a fluorescence-minus-one control (FMO), in which all antibodies are included except the one you suspect is most prone to error from spectral overlap.
4. Isotype controls should not be used to determine positivity.
You should absolutely not be using isotype controls to determine positive versus negative cells — or, as mentioned in #3 above, as a gating control.
5. Isotype controls are not always sufficient for determining non-specific antibody adherence.
Isotype controls cannot always determine non-specific antibody adherence versus, for example, free fluorochrome adherence. For this, you need to use isoclonic controls. If you add massive amounts of non-fluorochrome conjugated monoclonal antibody to your staining reaction, your fluorescence should drop. If it does not, your issue is not due to nonspecific antibody binding, but to free fluorochrome binding.