Date: Tue Feb 01 2005 - 17:21:33 EST
Dear Dagna, You have touched on a subject that has many answers, so let me summarise my feelings on this Isotype controls should be (and in the main are) matched to give a similar protein concentration as the antibody you are using, as well , obviously , as the fluorochrome Occasionally you will find Isotype controls with much higher protein concentrations than those of the antibodies you are using in which case an adjustment has to be made to achieve the same protein concentration Isotype controls can never have the identical tertiary structure of the test antibody you are trying to control (or else it would indeed be the antibody) . The non specific binding , and therefore the degree of fluorescence added to your negative cells by non specific / FC binding may not be identical as when you add the antibody of your test to the cells and monitor the negative population, sometimes the Isotype control does it's job well , other times it does not The best negative control ,I feel, is the negative population within your positive sample, but this is only useful if you have a clear demarcation of negative to positive for whatever antibodies you are working with , here you see exactly the non specific/FC binding attributable to your antibody on the negative population(which will be made up of primarily Instrument noise ,autofluorescence ,non specific binding)I would place this negative population within the first log decade, or now with extra sensitivity and different lasers, perhaps monitoring the lowest CV for this population and placing the negative population determined by this factor may be a more reproducible way of setting up . Negative cells, I take it you refer to cells that have not bee in contact with any antisera , can lead to users using slightly elevated PMT voltages, which may lead to increased compensation values, also this population will not take into account any non specific binding that occurs with your test antibody and hence this is the reason why when setting up with negative cells , you put your antibody test through and your negatives are a little further up the log scale than you wish them to be(proportional to degree on non specific binding of your test antibody on your negative cells), it is debatable , on a test to test basis , whether this will adversely affect your results To Squish or not to Squish-That is the question? The squish scenario will only exist if you use unstained cells and not ,either negative cells within the positive population, or good isotypic controls, if you have to do a concerted squish (this is a big one ) then your negative cells do not represent the negative cells within your test. In terms of publication , I feel that all events are relative(in terms of scatter and fluorescence) on the dot plot /histograms therefore placing the negatives in the first log decade is not written in stone , what I think is more important is that the data should be real ie not going off the end of the histograms / dot plots and not being built up in the zero channel too much, this is where the Bi exponential displays by the Stanford university group , I believe , will make looking at flow data either in publications or just in the lab much easier . Ian (good word squish, I think it is a relation of squidge) Ian Dimmick Flow Cytometry Applications Specialist / BD BD Biosciences 21, Between Towns Road, Cowley, Oxford OX4 3LY UK tel: 0044 1865 781688 cell: 0044 78 9966 2787 fax: 0044 1865 781 578 E-mail: Ian_Dimmick@europe.bd.com Website: www.bd.com Dagna Sheerar <dsheerar@robarts.ca> 01/02/2005 16:44 To: Cytometry Mailing List <cytometry@flowcyt.cyto.purdue.edu> cc: (bcc: Ian DIMMICK/Europe) Subject: Isotype Titration? Hello All - I have some questions regarding isotypes (just to clear up some "discussions" I've had with some users): The first is whether it is "correct" to titrate an isotype and how you would determine the difference bona fide background staining and just too much isotype. The second is whether or not it is acceptable to "squish" an isotype down into the first decade of the log scale. For instance; you have an unstained control that you set to sit nicely in the first decade as your negative control. Then, you run your isotype and see that it is fluorescing well into the second decade. Should you push this isotype down into the first decade? I say no. I say the unstained control is your true negative, the isotype shows the background level of staining and then you have the fluorescence level of your markers. And, finally, a statement was made in my lab that if you want to publish flow cytometry data, your isotype controls MUST be set within the first decade. True or False? All insight, opinions, rants and raves welcomed. thank you all, Dagna __________________ Dägna Solveig Sheerar Manager London Regional Flow Cytometry Facility Robarts Research Institute P.O. Box 5015, 100 Perth Drive London, Ontario N6A 5K8 (519) 663 5777 x 34042 dsheerar@robarts.ca
