[Cytometry] Compensation bead caution

Michael Waring mwaring708 at gmail.com
Fri Apr 15 13:51:11 EDT 2011


Anush, you are correct that it is important to use the same tandem
dye, but you want to be sure you are only correcting for the
contribution of the fluorochrome to signals in your other channels.
When you compare positive beads and unstained cells, any difference in
autofluorescence between the particles themselves will incorrectly be
interpreted as spillover and will lead to improper compensation.  If
you are using stained beads, you need to compare it to matched
unstained beads, and for stained cells you need to compare it to
unstained cells of the same type.

Regardless of what particle your reagent is bound to, for a given set
of voltages the spillover should be the same, so it is NOT correct to
recompensate your samples depending on whether you are looking at
lymphocytes or monocytes and using them as the unstained control.  The
unstained control should match whatever particle your reagent is bound
to.

My comment on monocytes relates to the fact that a group uses stained
cells for their compensation for a sort on an Aria (so running Diva
and can only define a single "unstained tube")--their other controls
stain lymphocytes so I can gate on that region in the unstained
sample, but for the monocyte marker it would not be an accurate
comparison.  What I end up doing is appending some of the unstained
into the single stain control data file for the monocyte marker, and
gate on the monocytes by scatter and use the internal negative
population for that data file as the unstained control.  But it would
be easier if all the markers were staining beads, with a single
matched negative bead for the 'unstained"...
Mike

On Thu, Apr 14, 2011 at 10:26 PM, Anush Karapetian
<karapetian1 at gmail.com> wrote:
> For an unstained sample I usually use a whole blood sample after lysing RBC.
> To analyze the lymphocyte population I gate unstained sample on lymphocytes.
> For monocytes I will regenerate a compensation matrix with an unstained
> sample gated on monocytes and so on.
>        FlowJo uses the mean of fluorescence intensity (FIpos) of positive
> stained beads and the mean of autofluorescence intensity(FIneg) of unstained
> sample of specifically gated population and fits FIpos and FIneg to the
> linear equation for each channel and for each spillover. It then uses those
> functions to generate the compensation matrix.
> So the negative peak for each positive bead is replaced with the
> autofluorescence of real cells.
> The essential is to stain beads with the same tandem dye that will be used
> to stain cells.
> Please correct me if I am wrong,
>
> Anush.
>
> On Thu, Apr 14, 2011 at 3:08 PM, Michael Waring <mwaring708 at gmail.com>
> wrote:
>>
>> Mario beat me to it-- Anush, you have to match the positive and
>> negative particle when compensating, comparing the stained beads to
>> the unstained cells will have the same result as comparing unmatched
>> beads--either under or overcompensating due to differences in base
>> fluorescence.  I've even had issues with a broad gate for unstained
>> PBMCs and compensating a monocyte-specific marker--the unstained is
>> the average signal for everything, but the positive is the
>> autofluorescence of the monocytes only, which tends to be higher than
>> lymphs especially in FITC or PE.  (I've told this group to use beads
>> but they wont listen ;) ).
>> We had another incident of bead mismatch today, the user had one comp
>> tube of rat binding bead expiring in 2012 and the rest were a mouse
>> binding bead expiring in 2013--the latter had the low level of pac
>> blue fluorescence again, and they used the same negative bead for all
>> of the tubes. Be sure to match your lots!
>> Mike
>>
>> On Thu, Apr 14, 2011 at 1:59 PM, Anush Karapetian <karapetian1 at gmail.com>
>> wrote:
>> > When using beads for compensation I  always use  unstained cells for
>> > negative control then generate compensation matrix by using FlowJo.
>> > I usually work with blood samples and gate unstained sample on
>> > particular
>> > population I am going to analyse.
>> >
>> > Sincerely,
>> >
>> > Anush Karapetyan,PhD
>> > Senior Lab. Technician
>> > Internal Medicine,
>> > University of Kentucky
>> >
>> >
>> >
>> > On Thu, Apr 14, 2011 at 7:15 AM, Mario Roederer <roederer at drmr.com>
>> > wrote:
>> >>
>> >> Great warning!  We too have seen issues like this.  So we abandoned the
>> >> "nonbinding" bead altogether, and just use, as you suggest, unstained
>> >> beads
>> >> as our negative control.  We usually just run them separately, rather
>> >> than
>> >> mix together with the positive.
>> >>
>> >> If you do mix together with the positive control (after washing the
>> >> positive...) -- don't worry if the "negative" now picks up some
>> >> residual
>> >> antibody fluorescence.  For proper compensation, you do NOT need to
>> >> have a
>> >> completely unstained control -- the negative only needs to be far
>> >> dimmer
>> >> than the positive, it doesn't have to be "truly negative".  I would say
>> >> that
>> >> as long as the negative is no more than 1/10th the signal of the
>> >> positive,
>> >> you will get good compensation values.
>> >>
>> >> mr
>> >>
>> >>
>> >> On Apr 13, 2011, at 10:26 PM, Michael Waring wrote:
>> >>
>> >> > Hello flow community-
>> >> >
>> >> > A compensation bead issue that first occurred in my lab 2 years ago
>> >> > has resurfaced, so I think it worth broadcasting.  I’m sure this
>> >> > applies to beads from other companies, but we specifically use the BD
>> >> > Compbeads.
>> >> >
>> >> > The BD Compbead kits come with two vials - binding and non-binding
>> >> > beads.  Adding antibody to the binding bead gives you a bright single
>> >> > stained control.  The non-binding, or negative beads, can be added to
>> >> > the same tube after washing to give you an appropriate negative
>> >> > population for calculating spillover, or collected as a separate tube
>> >> > as a compensation “unstained control.”
>> >> >
>> >> > Sometimes binding and non-binding beads from different lots are
>> >> > paired
>> >> > up, either leftover vials from older kits, grabbing the wrong vial if
>> >> > all the beads are kept together, or using a single unstained tube
>> >> > with
>> >> > mixed rat and mouse binding bead controls, etc.
>> >> >
>> >> > The autofluorescence of these polystyrene beads can vary by lot.  We
>> >> > first noticed a dramatic issue in our machines’ Pacific Blue channels
>> >> > (405nm laser, 450/50 filter).  I then compared two lots in more
>> >> > channels and found all signals with blue and violet excitation are
>> >> > higher in the newer batch; red excitation does not show a dramatic
>> >> > increase.  MFIs for the 2 lots are in the first table at the bottom
>> >> > of
>> >> > this email.
>> >> >
>> >> > The table at the very bottom notes bead expiration dates and
>> >> > polystyrene bead batch numbers (provided by BD) along with “high” and
>> >> > “low” Pac Blue signal on our LSRII.   Beads from batch Z01 have half
>> >> > a
>> >> > log higher fluorescence than earlier batches.  This causes incorrect
>> >> > compensation when lots are mixed.  (If the binding bead has a higher
>> >> > autofluorescence, then the experiment would be overcompensated; if
>> >> > the
>> >> > negative bead were brighter then it would be undercompensated.)
>> >> >
>> >> > As is often said on this list, it is critical to match your positive
>> >> > and negative particles when doing compensation.  Be sure that your
>> >> > unstained Compbead is from the same lot as your binding bead!  Since
>> >> > the polystyrene batch # is not provided on the labels, make sure you
>> >> > keep your kits together.  Use the expiration dates to ensure you’re
>> >> > likely using the same batch.  If you have to use a combination of
>> >> > anti
>> >> > rat and anti mouse binding beads in your panel, make sure the
>> >> > negative
>> >> > population matches each control.  You could also use a drop of the
>> >> > binding bead without antibody as the unstained control to be certain
>> >> > it is the same exact bead as the single stained controls.
>> >> >
>> >> > Hopefully this will help other labs avoid wasting a lot of time
>> >> > trying
>> >> > to find one cause of improperly compensated data!
>> >> >
>> >> > Mike Waring
>> >> > Ragon Institute Imaging Core
>> >> > Massachusetts General Hospital
>> >> > Charlestown, MA 02129
>> >> >
>> >> > Channel       2010 exp        2012 exp
>> >> > FITC                  71              103
>> >> > PE                    67              99
>> >> > PE-TexR               86              131
>> >> > PE-Cy5                97              147
>> >> > PE-Cy5.5      73              109
>> >> > PE-Cy7                44              66
>> >> > APC           47               49
>> >> > Alexa 700     34              39
>> >> > APC Cy7       18              23
>> >> > Pac Blue             105               215
>> >> > Amcyan                73              146
>> >> > Alexa 430     29              57
>> >> >
>> >> >
>> >> > item     Exp date     PB signal       PS batch #
>> >> > Neg        4/30/2010    low           X01
>> >> > Rat+      10/31/2010     low          X02
>> >> > Neg        12/31/10     low           Y01
>> >> > Neg         3/30/11     Low?  Y01
>> >> > Mouse+  3/30/11         Low?  Y01
>> >> > Mouse+  5/30/11         high          Z01
>> >> > Neg         5/30/11     high  Z01
>> >> > Neg         9/30/11     high  Z01
>> >> > Mouse+  9/30/11         high  Z01
>> >> >
>> >> > _______________________________________________
>> >> > Cytometry mailing list
>> >> > Cytometry at lists.purdue.edu
>> >> > https://lists.purdue.edu/mailman/listinfo/cytometry
>> >> > Search the list archive at  http://tinyurl.com/cytometry
>> >>
>> >>
>> >> _______________________________________________
>> >> Cytometry mailing list
>> >> Cytometry at lists.purdue.edu
>> >> https://lists.purdue.edu/mailman/listinfo/cytometry
>> >> Search the list archive at  http://tinyurl.com/cytometry
>> >
>> >
>>
>> _______________________________________________
>> Cytometry mailing list
>> Cytometry at lists.purdue.edu
>> https://lists.purdue.edu/mailman/listinfo/cytometry
>> Search the list archive at  http://tinyurl.com/cytometry
>
>



More information about the Cytometry mailing list