mAb Titration-Summary

From: Yasmin Marikar (ycaldwel@com1.med.usf.edu)
Date: Thu Dec 07 2000 - 00:26:39 EST


Friends:

Thankyou very much for the suggestions and input. As many of you were
interested in the mAb titration subject I have included all the input I
received below. Since I had about 30 mAbs to titrate I was very happy to
see Mario's reponse that 4-color titration was doable. Anyway I tried
this out combining it with the other suggestions to calculate signal to
noise ratios using the arithmetic mean values on individual histograms
at each mAb dilution. Each panel of 4 mAbs were mixed at the same conc.,
6 dilutions in all. I used unlabelled cells as background, since my
mouse PBLs and splenocytes give high backgrounds with isotype controls
(in spite of Fc block). Acquistion was done keeping the
Detector/Amplification values constant for a panel, but compensating at
each subsequent dilution. The Detector/PMT values were set to get the
best resolution at maximum mAb concentration (1ug/million cells). Most
of the mAbs showed an increse in signal/noise ratio with conc. then a
plateau over 2-3 concs. and sometimes, another increase to the highest
conc. I selected as optimal the lowest conc. required for the plateau
value. The logic being that signal could be increased by changing
detector settings and compensation would be less of a nightmare. I will
know if these selected concs. work in the 4-color mix in a day or two.
All comments and criticisms are welcome!

Yasmin

a)"Dennis Broud 301-594-5879 FAX 301-594-3037" <BROUDD@cder.fda.gov>

titrate them individually first.  I generally use 10 ul of the AB
dilutions to 100 ul of whole blood, if I'm lysing.   Or if I have
prepared cells, 10 ul of ab to 100ul of cells (about 1,000,000 cells). I
generally start at 1:20 and do two fold dilutions up to 1:640 as a
starting off place for commercial antibodies, unless the manufacturer
has already specified a dilution. Once I've titered each antibody this
way, I've found that I can generally add all 3 or 4 antibodies together
into a pool.  That is use 30 or 40 ul of AB mixture to 100ul of whole
blood or to 1 million cells.  I let all my staining go for at least 30
minutes before I lyse wash, etc. If cells have been stimulated in any
way (especially mouse cells) I block with 40 ul of 1:40 dilution of FC
block (Pharmingen)to the cells  for 15 minutes at room temperature and
then add the antibody cocktail. Occasionally you'll get a wierd antibody
that will act differently in a cocktail than it does by itself, then you
have to do all the combinations to
figure out what's going on, but fortunately that's not very often.

b)  "Dennis J. Young" <djyoung@ucsd.edu>

I'd titrate individually, with the assumption that stearic hindrance is
minimal. Start at 200 - 300 micrograms per ml of Ab and use 6 serial 1:3

dilutions plus an unlabeled control. So you'd need to run only 25 tubes
(6*4 + 1) Measure signal to noise, not the "brightest" positives. Then I
would verify
the 4-Ab cocktail. See Chapter 4.1 in Current Protocols in Cytometry
(C.C. & S.J. Stewart)

c)  Mario Roederer <Roederer@DrMR.com>

We often combine and titre as you suggest--as long as the antibodies
don't cross block (i.e., don't titre FITC CD8 with PE CD8, etc.), and
as long as you set proper compensation. Then it really doesn't matter,
and you can titre all four effectively
on the same cells.

d) Maciej Simm <simmmmer@yahoo.com>
I've only titrated my AB's 1 at a time and mixed them at the lowest
titration that gave me the same signal as stock.















This archive was generated by hypermail 2b29 : Sat Mar 10 2001 - 19:31:40 EST