isotype controls anad PFC

Uriel TK utk1 at 013.net
Thu Jan 18 16:04:42 EST 2007


Wayne:
I must confess I had trouble following your train of thought, but I think I understood your general question. 

My 2c: 
I think that a good way to approach the problem is to think what is what you're trying to achieve. 
If you're doing multicolor flow and your objective is to detect the presence of the % of a given population of cells, then what are the isotypes contributing? When you do your gate, when you see your results, what do you compare to decide "here, these are my cells"? It is clear to me that you should be using FMOs to decide that. 
You can add a 2nd tube which would be your FMO+the isotype control of your marker of interest. But as you mention yourself, with tandems and custom reagents, is the isotype control really a control or is it too different to be called a control? Maybe it could be called an "indicator" instead of control, which would be interpreted like this: 
Since I am 1)using a control antibody which is very similar to my test antibody, and 2)using the same probe (albeit at a different ratio/composition); Then, if 1)the background I get is low, and specially if it is lower than the threshold I set with my FMO, it significantly strengthens the probability that the events I call positive are really positive and not due to unspecific interactions. On the other side, if the background is high and specially if it crosses the threshold I set with my FMO, I should seriously consider the specificity of my positive events and confirm the results with other assays or with other panels.

So, the isotype "indicator" would be giving useful information, but the point here is that it is not a golden standard or something like that. It is a further element of verification that you add to improve the validity of your results. You should ask yourself whether that verification is needed or not, and for which color/marker. Multicolor experiments can get complex and big, so the point of "where do I add complexity and samples and controls?" is not trivial. If it is a panel you run commonly and have already validated it for a given experimental setup, then the in my opinion the isotype control certainly looses much of its appeal, and becomes a tradition, in the sense of "Traditions are solutions for which we have forgotten the problems" (Peterson's principle).

Another problem is that there is some arbitrariness to their use. As you scale your colors, the number of combinations grows , I.e. adding a 6th color adds more possible combinations than adding a 5th, and adding a 5th adds more than adding a 4th. Therefore if you want to go "by the book" and use many or all the possible controls (FMOs, isotypes, positive and negative controls, etc) you are adding even more samples and you find yourself rapidly with a huge, complicated experiment with so many variables to watch for and a big risk of technical mistakes and errors. What is standard with 3 colors might be possible with 4, a hassle with 5 and plain impossible with 6. That is one of the biggest challenges I think, to balance the scientific objective with the technical feasibility. That is inherent to polychromatic flow but regarding isotypes you are adding complexity and it is not always clear what is best or needed, and there are many possibilities. So, when you have to choose which controls you use and which ones you don't, and how, etc. what you call criterion or common sense, a reviewer/critic could call arbitrary or nonsense (sometimes justified, sometimes not...). 

Finally I think that there is a big conceptual difference (and accordingly, technical approach) between measuring expression levels of a given marker and population studies. Realizing that difference when approaching an experiment with multiple colors was very important for me, and in that context, isotypes can (or cannot) serve different roles depending on your objective, your specific problem, etc.

Best of luck,

Uriel.
Uriel Trahtemberg, M.Sc.
MD/PhD student
The Laboratory for Cellular and Molecular Immunology
The Hebrew University - Hadassah Medical Organization
Jerusalem - ISRAEL

 
  ----- Original Message ----- 
  From: Wayne Harris 
  To: Cytometry Mailing List 
  Sent: 17 January, 2007 1:26 AM
  Subject: isotype controls anad PFC


  Hello everyone,



  I just wanted to throw out a question to see what people's opinions are. i know that several times before the subject of the utility of isotype controls in analysis of multicolor cytometry has come up. The basic perspectives that seem to repeat are there are those that use isotypes because they always do, or it's better to use isotypes than not to. And still others favor the use of FMO controls. But somehow it all seems incompletely accepted what is best. 


  My question is really if you are using a six or seven color experiment (perhaps more) with tandem conjugates on a digital system can you really use isotype controls to aid in your analysis at all since they will be even more different from the actual reagents used in comparison to the case for our "normal" reagents (FITC,PE, PerCP, APC)  where the argument against the utility of isotypes focussed on differences in Ig concentrations etc. And furthermore, if the tandems can vary so significantly from lot to lot,	to what degree might we see shifts in the staining patterns as the levels of background or just simply staining characteristics of the reagents vary for the different antibodies in our panels.... 


  ie. even though we create antibody specific compensations in the matrix, do the differences in the panels because of the different antibodies cause significant shifts in background etc that your threshold for signal determination may itself shift from panel to panel thus making isotypes virtually useless as a control other than to say whether or not there is significant background.


  Simply stated, does polychromatic flow cytometry further significantly increase the problems in analyzing data using isotype controls because you may have greater problems with backgrounds etc  for different panels even for samples acquired on the same machine with the same settings... Are FMO controls the only option then for each reagent in each panel you use?


  just wanted to get an idea how people are dealing with this question.


  thanks








  Wayne A. C. Harris
  Bone Marrow Transplant Laboratory
  Immunology and Flow Cytometry Assays Laboratory
  Winship Cancer Institute, Emory University
  1701 Uppergate Drive
  WCI, building C, Rm#4032
  Atlanta GA  30322
  404-727-3086





-------------- next part --------------
HTML attachment scrubbed and removed


More information about the Cytometry mailing list