Jonathan, you may not need to use isotype controls at all.

During both the Royal Microscopy Society's advanced flow cytometry course in York 2010 and the advanced data analysis course at the CYTO2013 the consensus was that isotype controls in flow cytometry experiments are of no benefit (but just add to the cost).

The recommended controls for multicolour flow cytometry were:

- unstained control

- compensation controls (preferrably beads)

- fluorescence minus one (FMO) controls

However, your question is regarding the need for the use of the same fluor. I believe this is covered in Mario Roederer's post on the Purdue cytometry list - find section copied below as well as link.

"...Furthermore, there is the problem that even at the same F/P ratio, the

conjugates could be significantly different. It is quite possible that in one

antibody there is a fluor at a critical "background" binding site; on another

antibody, this site is unconjugated. This could significantly change the

"stickiness" of an antibody.

For example, when you conjugate antibodies to Texas Red, you can get hugely

different "background" binding characteristics depending on the reactive form of

TR that you use--even after getting exactly the same F/P ratios. Clearly, the

sites on an antibody that are conjugated are different by these different

reactive forms of TR, and those differences translate into different

"background" binding characteristics."

https://lists.purdue.edu/pipermail/cytometry/1998-April/009860.html

Good luck!

The fluorochromes do not affect binding. The isotype control is used to determine non specific binding of your antibody. Your monoclonal antibody (mAb) will recognize an antigen by its Fab fragment that is highly specific, and the fluorochrome associated with this mAb will show you the % of cells expressing this antigen and the level of expression (MFI). However, all mAb can also bind cells in a non specific manner (not by Fab specific recognition) thus creating a background of non-specific fluorescence. This non specific binding is more or less specific to the Ig you are using (isotype). The isotype control is then a Ig from the same isotype that your specific mAb but whose Fab does not recognize nothing on your cell of interest (usually specific of another species). This control can then only bind in a non-specific manner to your cells. As this control will define the background fluorescence for a given channel, it must be of the same fluorescence than your specific mAb or in the worse case a close relative (like AF488 for FITC). A control Ig labelled in PE cannot in any case show you the background fluorescence in FITC for example.

I hope this helped !

S.L.

Thanks for the answer Sebastien. I am curious as to why you can't use different fluorochromes though. If I find that my PE labeled isotype has the same fluorescence as background (unstained cells), then I know that my cells are not binding this IgG non-specifically. Why then would I need to do it again with a FITC labeled isotype?

If you want to use isotype controls, they have to be fluorochrome matched. Ideally, the antibody/fluorochrome ratio has to be the same, too.

Regards,

Bastian

Your cells may have an auto-fluorescence which may change due to culture conditions, different stimulation/activation etc. Although this mainly affects FL1 channel, it may well flow over to FL2. So, you may have different background fluorescence levels in those two channels, independent from unspecific antibody staining. Additionally the compensation settings would make it difficult to translate the antibody based unspecific signal from one channel to another.

short answer is YES.

you should also consider using FMOs (fluorescence minus one), if you stain for more than one fluorochorme. An FMO control contains all the flurochromes in a panel, except for the one that is being measured. Basically you can use this to set the position of your markers/gates accurately.

see http://flowjo.typepad.com/the_daily_dongle/2011/09/fmo-vs-isotype-controls.html for more information.

happy flow

Mara

Jonathan,

Bastian Schilling is right. The fluorochromes should be matched. In adition you want to have the same antibody/fluorochrome ratio. To get this most optimal it is recommended to order the antibody aswell as the isotype control from the same companie. They use the same protocol to label the antibody and their recommended/related isotype.

Thanks for the answers but I am trying to understand WHY they need to be fluorochrome matched. It is my understanding that isotypes are to make sure that your cells are not binding your antibody of interest non-specifically. If I can show using one label (e.g. PE) that my cells are not binding non-specifically to my isotype, then why would they all of a sudden start binding one that is FITC labeled? I realize that cells have different autofluorescence in different channels, but assuming that one found that the PE-labeled isotype is not binding cells, this will be the same as an unstained sample (that is how I set my background sensitivity).

It's not just a matter of showing your cells don't mind, it's a matter of being able to set your gates properly with the same fluorophore.

Teresa, what is wrong with setting your gates on unstained cells (or FMO)? If the isotype isn't binding the cells, the cells will have identical fluorescence to unstained cells. When using multiple colors the background needs to be set for each channel, so why not use unstained cells (if you have already shown that the isotype isn't binding and changing the fluorescence?

Of course this is not true of the isotype is binding to your cells, in that case I totally understand why you would have to use a fluorochrome matched isotype to set the background for each channel.

Note: I do use fluorophore matched isotype controls, but I am trying to understand the rationale behind it. For example, if I am doing FLOW with 4 mouse IgG1 clones (PE, FITC, APC, AF700 for example), and I show that the mouse IgG1-PE isotype does not bind the cells at all (fluorescence of isotype is identical to unstained), then what is the point of doing an IgG1-FITC, APC, and AF700 isotype control?

Many fluorochromes bind to surface receptors (this has been well documented for PE and cyanins) (see examples below). While doing isotype control, you better match the fluorochrome as well just in case.

Jahrsdörfer, B., Blackwell, S.E. & Weiner, G.J., 2005. Phosphorothyoate oligodeoxynucleotides block nonspecific binding of Cy5 conjugates to monocytes,

Shapiro, H.M., 2004. Practical Flow Cytometry 4th Edt,

Takizawa, F., Kinet, J.P. & Adamczewski, M., 1993. Binding of phycoerythrin and its conjugates to murine low affinity receptors for immunoglobulin G. Journal of Immunological Methods, 162(2), pp.269–272.

van Vugt, M.J., van den Herik-Oudijk, I.E. & van de Winkle, J.G., 1996. Binding of PE-CY5 conjugates to the human high-affinity receptor for IgG (CD64). Blood, 88(6), pp.2358–2361.

I hear your frustration! One thing to add is that non-specific binding is not only receptor mediated. Fluorophores will affect the charge of the antibody molecule and this charge may mediate non-specific binding. This may be particularly relevant if cell viability is compromised. The fluorophore needs to be matched in order to set your cytometer up correctly as discussed above. The importance of this will vary depending on the application. Much flow cytometry is undertaken poorly with incorrectly set up cytometers and inadequate controls: even published work. Often this does not matter. Sometimes it is critically important. However, should we not always strive to carry out our experiments as correctly as possible?

Yes, flow staining can be a pain sometimes. I agree that specific fluorochromes can bind, emit some fluorescence, and interfere with gating and compensation. So it's always a good idea to use an isotype of the same fluorochrome. With surface or intracellular staining you sometimes get low levels of non-specific binding with isotypes and using the same fluorochrome ensures you compensate this false positive (versus just simple cell autofluorescence) while being able to utilize a larger panel of fluorochromes. So if you are staining a target with PE and use a FITC isotope you may get a few positive events and thus you won't be able to normalize to the isotype false positives or compensate out the FITC signal.

And another thought. The different fluorochromes are very different molecules. PE is a large molecule, FITC very small. The addition of fluorochromes will affect antibody binding so the PE conjugate may bind differently, both specifically and non-specifically, as compared to exactly the same antibody conjugated with FITC. Even the amount of a given fluorochrome added will make a difference. The better companies recognise this, as above, and will use identical protocols for their 'active' antibody conjugates and isotope control in order to ensure as much as possible that the amount of fluorochrome conjugated is similar. This is important for signal intensity but also to replicate non-specific binding. This is particularly relevant to biotinylation as biotin binds to certain amino-acid residues and may modify binding to the specific target but also to FcR. The level of biotinylation will have a dramatic effect on specific and non-specific binding. The same is true, but less dramatically so, for other conjugates.

It seems that not only you should use isotype conjugate to the same fluorochrome but also from the same company. Apparently, depending on the company, the experimental procedure for antibody labelling may differ resulting in different coupling efficiencies and, understandably, in different non-specific binding. It's quite a good commercial argument 😉 for companies but also makes some sense from a scientifc point of view. Whether this could result in significant changes in term of staining, this I cannot tell.

Does level of conjugation make a difference? With biotin it certainly does. I routinely FITC conjugated and biotinylated my own antibodies for flow cytometry. The same antibody biotinylated to different degrees would behave very differently. It is even possible to completely abrogate specific binding of some antibodies by over-biotinylation, depending largely on the number of amines presented by the amino acids in the variable region. This is discussed at length elsewhere on RG. This was less important with FITC, but differences were seen. With practice it becomes easy to control the level of conjugation for these molecules and it can of course be measured. I have no first hand experience of PE conjugation but as it is a large molecule I would be surprised if similar considerations did not apply.

And this is directly relevant to the use of a 'different' control antibody. The PE conjugate MAY behave very differently to e.g. the FITC conjugate for these reasons. Sometimes you will 'get away with it', but why take the risk. If looking at rare events this is crucially important. If I am reviewing data dependent on rare events my first port of call is the controls and instrument set-up for exactly these reasons. I only believe conclusions based on rare events if these things are perfect.

Jonathan, you may not need to use isotype controls at all.

During both the Royal Microscopy Society's advanced flow cytometry course in York 2010 and the advanced data analysis course at the CYTO2013 the consensus was that isotype controls in flow cytometry experiments are of no benefit (but just add to the cost).

The recommended controls for multicolour flow cytometry were:

- unstained control

- compensation controls (preferrably beads)

- fluorescence minus one (FMO) controls

However, your question is regarding the need for the use of the same fluor. I believe this is covered in Mario Roederer's post on the Purdue cytometry list - find section copied below as well as link.

"...Furthermore, there is the problem that even at the same F/P ratio, the

conjugates could be significantly different. It is quite possible that in one

antibody there is a fluor at a critical "background" binding site; on another

antibody, this site is unconjugated. This could significantly change the

"stickiness" of an antibody.

For example, when you conjugate antibodies to Texas Red, you can get hugely

different "background" binding characteristics depending on the reactive form of

TR that you use--even after getting exactly the same F/P ratios. Clearly, the

sites on an antibody that are conjugated are different by these different

reactive forms of TR, and those differences translate into different

"background" binding characteristics."

https://lists.purdue.edu/pipermail/cytometry/1998-April/009860.html

Good luck!

Isotypes are a waste of time and money. Usually you compare 2 or more different groups with the same coloration, and you are interested in the difference, not the absolute value, so you don't care about any unspecific binding, which usually never happens if your cells are alive and the antibody is of good quality. If you use them anyway, you do that because some closed minded reviewer will want you to. And, for many of the above mentioned reasons and others you can never be sure how it binds and it is not comparable sometimes to the specific antibody, even if they are both from the same manufacturer. So they are not good as controls (or anything else for that matter - maybe except to make sales for the biotec company)

Surely it is the detail that matters. As discussed at length above the wrong isotype control can be more than useless. But the details of the experimental setup matter. For many cells non-specific binding is non-contributory. However, some cells can be very tricky. Macrophages are the ones I have most experience with. Auto fluorescence levels are high and change depending on activation status. They also have abundant FcR. The background signal can be high and I would suggest that in this setting a carefully controlled experiment is essential? Your call re your setup.

Just a thought with regard to abrogation of binding. As antibodies are large molecules and conjugates are incorporated randomly, not all potential fluorophore binding sites will be occupied. So although in some molecules binding will be affected, in others it will not. This of course depends critically on the conjugation ratios, but the commercial products get this right. Complete abrogation of binding is rare unless over-conjugated.

Thanks everyone for all of the great answers. I know a lot more about isotype controls than I did before posting the question (maybe more than I really want to know? ;)) and it is much more clear now. Given that the fluorochrome can affect binding of the isotype, it makes much more sense to me now why isotypes should be matched as closely as possible.

Thanks again!

J

Teresa,

No, no and no !!! Iso-ctrl shouldn't be used to set gates !!! For all the reasons mentioned post above.

Ina is absolutely right. Have a look on the post from Mario in the Perdue mailing list about : "we should all stop using isotype controls to set gates".

As Mario mention in his post : "Gating is an art, one which requires considerable experience and knowledge of the system". And Iso-ctrl doesn't give you any informations helpful on "gate setting". I reiterate what Ray Hicks (and Mario) said: "Isotype controls can be a valuable tool for rooting out problems. However, it is a rare problem that will be solved with isotype controls".

It's interesting returning to this discussion! I have read the document Ina provided the link to with interest. I agree! It makes the points about choosing isotype controls made above. Surely the answer is not that we should not use isotype controls, but that the wrong isotype control is worse than no isotype control. This answers the original question definitively. If you are going to use them they MUST be the same fluorochrome. And all the considerations about conjugation ratios etc apply, as I argued above. But surely we just need to think carefully about our experiments and understand what we are doing; the systems we are working with. I agree that in many settings isotype controls are unnecessary and potentially harmful. Occasionally they are essential. If they are essential, it IS possible to get the right molecule; it just takes a bit of time and effort.

I just read this conversation and agree with Richard , but I am really wondering that why some of the people in science community like to see only white and black colours. Isotype controls may not be necessary but in some cases they are helpful and in those circumstances, they have to be from the same subclass and fluorochrome. As simple as this.

I totally agree with Ina, Miguel and Richard on their comments for use of isotype controls. It may not be a total waste if isotype controls are combined with FMO controls, especially in multicolor experiments. However, the type-matching and titration are tricky and time-consuming, may not worth to do it.