BrdU/cell cycle summary of responses

Hi everyone!

  Thanks for all those who responded to my BrdU/cell cycle inquiry!  
I've included a summary of responses.
  Also - to note, many suggested use of the Click-iT EdU kit from 
Invitrogen.  The researcher who is doing the BrdU assay attempted to use 
this in the past with their cells, with no or limited success (how much 
they did troubleshooting however, I do not know - they tried this at 
another institution).	Also - I believe they are interested in having 
both a nucleic acid stain vs BrdU plot AND a DNA histogram, so that they 
can overlay the cell cycles of control vs mutant.  I think they are 
going to end up using BrdU/PI in ethanol fix.

Thanks again, and Happy May Day. :)

Best,

Christiane

*Responses*


Try to replace the BrdU with the EdU from Invitrogen.

Attached a crude experiment with PMA/A23187 stimulation of mouse spleen cells with EdU in
the medium.

Cells are stained for surface molecules (here CD4-APC) and then like for intracellular
cytokines with anti-EdU-FITC. No harsh treatment necessary.


I did not stain with PI to analyze the cell cycle, I just wanted to see if I can stain
with anti-CD4 together with anti-EdU. You can see that not all CD4+ cells proliferate.


Hope this helps.



******************



I have recently published a paper in JBC using the BrdU assay to assess 
DNA synthesis in

polyploid megakaryocytes.


Fuhrken PG, Apostolidis PA et al 2008


My suggestion is use RNAse to improve the quality of the ploidy data 
(even though BD does

not recommend it, I have found it to be absolutely essential). It may 
also matter what

instrument you use for ploidy (as well as the flow rate of course). The 
data I show in the

paper were acquired in an Aria using PI for ploidy. PI worked even 
better in our old LSRII.

I also found I could improve the quality of my CVs using 7AAD in the 
Aria... All flow rates

were < 100 events/s.



******************


Have you tried plotting a histogram based on BrdU

instead of 7ADD? It is my understanding the cell cycle

using BrdU should be analyzed  based on BrdU signal

instead of 7ADD. I think you should be able to get a

nice picture based on BrdU histogram.



******************


For a widespread technique, the BrdU techniques does often throw up some

problems! Its like a lot of things, when you get a protocol that works,

stick with it but getting to that stage can take a while and help isn't

always forthcoming. There are a number of points to address from your mail.


First of all, yes ethanol (or alcohol) fixation is the best way to get nice

DNAS peaks. Alcohols are dehydrating fixatives and they work by coagulating

proteins - this means that the histones etc that bind to DNA are 'shrunk'

and this allows access of the DNA-binding dye to the DNA. Aldehydes are

cross-linking fixatives and these make the proteins forma sort of cage

around the DNA which means that access for the DNA dye isn't as good. This

generally manifests itself as a broad CV.


In all	 almost all BrdU protocols, the DNA must be unwound this can be by

using acid, alkali, heat or enzyme. Probably the commonest  method is to use

2N HCl to denature the DNA sufficiently for the BrdU Ab to get in. However

this technique isn't particularly compatible with other Ab staining or

fluorescent protein expression. Ion these cases, we would usually use Dnase

treatment to unwind the eh DNA. In general this is after PFA fixation, not

ethanol so immediately you start to compromise but that is the price to pay

for more information. IT is possible to improve the DNA CV by adding a

permeabilisation step using Triton (saponin is often seen used but it isn't

a very good permeabiliser of nuclear membranes).


With your cells, I would fix some in EtOH and some in various strengths of

PFA and just do a DNA (PI) stain - the ethanol will be the 'gold standard'

and you cans see how close the PFAs get to that. If you see a less intense

staining it is due to less dye being incorporated. Makes sure you use a

saturating concentration (over 10 ug/ml for mammalian cells), use RNAse and

ideally stain overnight before reading. This may also explain the

disappearance of the G2 peak - this will appear lost if there is

insufficient dye or if the CV is too broad - the CV of the G1 and G2 peaks

should be the same but as the G2 is higher up the scale it has more channels

to spread into and can artificially disappear.


You show algorithm fitted curves to your DNA profiles - why do this? All

single parameter DNA analysis suffers from inherent limitations ie the cells

in early S don't show much of a shift along the axis as they don't have that

much more DNA than a G1 cell. In an attempt to get over this several

mathematical models have been produced to (1) make the analysis more

objective and (2) attempt to model the curves so that early and late S cells

can be distinguished from G1 and G2 respectively. The other drawback of

single parameter staining is that we have no kinetic information ie are the

cells with S phase DNA really dividing? And how long does a cell take to

traverse S. This is why the BrdU method was such a boon to researchers in

cell cycle. There is really no need to show a histogram (apart from to

please the old school!). The bivariate plot of PIU v BrDU gives all the

information you need and is a more accurate measure of cells in S. This also

means that you can get away with broader CVs to a large extent.


You also use 7AAD. What about a different dye. PI is the standard but if you

have the right machine ie one with a violet laser, DAPI is also a good bet.

PI is good because it is an intercalator and has no base pair specificity.

7AAD is a G-C binder and DAPI is A-T (and some other modes which is good!).


There is also other recent good news in that Invitrogen has developed an EdU

reaction kit. This uses a 'click' reaction to label another thymidine analog

- EdU - that is incorporated into S phase cells like BrdU. The major

advantage here is that no DNA unwinding is required. IT uses a simple Fix

and Perm step and can be combined with fluorescent proteins and some other

fluorochromes (its not that good with large organic fluors like PE and APC).

There is more information on the Invitrogen site and I can say that we have

used them in my Lab with great success and I think these kits will prove

very popular.


Sorry for the essay but if you have any questions, I would be pleased to try

to answer them.



*************************


Have you tried the Click-IT (EdU) assay from Invitrogen? It works better 
than the BrdU assays (BD etc) for us in fluorescence microscopy and flow 
cytometry (note: different versions of the kit for each).


************************


I have not worked with this system, but just FYI, there is an

alternative from Invitrogen which "promises" to be faster, gentler, and

preserve antigens (which may be of value for the next step the

investigator wants.  It is called Click-iT EdU, does not require HCl,

heat, or DNase for the cell cycle stain.  Don't know about PFA with this