Concentration of T cells is important (at least 1x10.7/ml) and the timing (not to long) about 3 min. is enough. I labeled with 1 uM and this gives a very bright staining. Then immediately wash in prewarmed FCS and wash 3X afterwards before culture.

So decreasing the concentration of CFSE and removing the excess CFSE after labeling is crucial.

Alberta's suggestions are correct. Cfse is toxic. Do dose testing Also leave cells in warm medium for 30 minutes after quecnching labeling reaction. Good luck

Hi Alberta and James,

I used cell at 1x10^6/ ml and 10 uM CFSE for 15' as described in the invirogen datasheet ... I will try as Alberta suggest,

1x10^7/ml and 3 min instead of 15'.

@ James are you suggesting that instead of wash right after quenching I should try to leave the cells for 30 min at 37° and then wash away the excess of CFSE ???

Yes. That was a step in the original protocol that is needed to full remove unbound CFSE. Check out the Vybrant CFSE protocol that I have. We were able to get very clean proliferation studies done using mouse or human CD4 T cells.

This works best for me,

1) Resuspend 10 million cells in 1ml warm PBS in 15 ml tube.

2)Add one ul of 5uM CFSE in 0.1%BSA and incubate at 37 degree celcius for 7 minutes with intermittant shaking.

3) Fill the tube with cold PBS and spin down at 1200rpm for 5 min.

4) Wash with cold PBS 2 more times and add pre warmed medium.

This has worked for me every single time.

Alberta is correct. CFSE is toxic so titrate the CFSE to determine the concentration that yields the maximum proliferation. I labeled 10 million cells with 2.5uM CFSE for 15min in the dark with gentle shaking in a 37C water bath. I quench with autologous plasma & wash the cells x2 to remove any excess dye.

Dario, one question, when do you loose your lymphocytes? Just after staining, or some time into the culture? Of course I do agree with all the above notions - first, CFSE is toxic (as permanently binding cellular proteins) and must be titrated; we went down to 0.2 uM for 15 minutes at 37 degs, but prepurified CD4s might be more sensitive. Regarding my initial question - if you stimulate CFSE-stained PBMCs, then after 72 hours you will always see 2-3-4 divisions, but your cell count will be low, sometimes lower by half than the original count, due to both CFSE toxicity and AICD (activation-induced cell death). You may want to look up my chapter in CPC: Witkowski JM. Advanced application of CFSE for cellular tracking. Curr Protoc

Cytom. 2008 Apr;Chapter 9:Unit9.25. doi: 10.1002/0471142956.cy0925s44. PubMed

PMID: 18770652.

Hi All,

I may be able to help you with your problem. Chris Parish and I have published several papers on CFSE labelling, one of the most recent explores the use of CTV and CPD as well (JIM 2012, 379(1-2)pg1-14). Believe it or not, we actually would recommend not to follow the manufacturer instructions on labelling in serum (protein free medium), but actually label in high amine content medium (I use 10% FCS in RPMI for eg). This dramatically reduces toxicity. Although this may seem counter intuitive, what we think is going on is that this prevents excessive substitution of surface protein (which causes function loss and toxicity), while still allowing labelling of intracellular proteins which are so abundant that toxicity is limited (so essentially concentrating labelling to intracellular proteins). Anyway, if you need more help please read the paper and/or email me.

Hope this is helpful,

Ben Quah

Ben, that was inspiring. And I like your recent (Journal of Immunological Methods 387 (2013) 181–190) paper! But then, your interpretation of high amine medium benefit is puzzling, at least for bi-ester dyes requireing INTRACELLULAR esterases to become first 'activated' for protein binding and then converted to fluorescent derivatives. How would this reconcile with the idea of binding to the surface (i.e., EXTRACELLULAR) proteins? Unless of course there would be the nonspecific surface esterase present, converting free extracellular (say) CFSE to monoester binding the proteins. Is there any? Otherwise I think it is the low dye concentration which is crucial for low toxicity (simply: low concentration of the dye means low proportion of bound proteins). It is a compromise between the need to stain with enough dye to pick up the halving signals form dividing cells (depending somehow on the sensitivity of your FACS system) and the toxic effect of dye bound to cellular proteins.

Hi Jacek,

This is a common interpretation of the chemistry of the dye and what I initial thought to!, however, it is not quite correct.

The esterase is important for removing the two acetate that make the dye membrane permeant, thus concentrating the dye in the cell, this also results in the dye becoming fluorescent. However, the succinimydyl groups are active in both forms (ie with and without the acetates), this can be seen in the molecular structure of both CFDA, SE and CFSE (eg Quah and Parish 2010, JoVE) so you still get reactivity of amines on the cell surface regardless of esterase activity but the dye is not fluorescent until acetate removal (which is what the "activation" refers to). By having protein in the labelling buffer this surface reaction would be buffered against.

With this in place we can stain with these dyes at 120uM of total dye with no toxicity issues (as outlined in the above mentioned JIM paper, you can also see this in the extreme in our most recent papers on Fluorescent Target Array assays in Cytometry A (2012) and JIM (2013))

Hope this helps! :)

Sabita Islam · is correct. Pl try her protocol to avoid the toxicity of CFSE in your assay. lalitha kabilan

Ben, I do agree with your argument; however, do I understand correctly, that he non-fluorescent form of the dye would still bind surface proteins ( and 'poison' them) without eventually becoming fluorescent (due to lack of esterases on the cell surface)? In this case you are right that protein-rich staining buffer would help; yet, would it not take a substantial part of th diester out, thus reducing the proportion of the ester entering the cell and eventually becoming fluorescent?

Hi Jacek,

Yes to both your questions. It takes more initial dye to get the same levels of fluorescence in protein containing buffer than without. However, the benefit is that you can now label with ALOT more dye than you could otherwise, and not get any apreciable toxicity. This mean you can now push the fluorescence out to get many more divisions being detected before dilution to auto fluorescence. For eg labelling with 80uM initial CFSE in "protein" buffer ( 10% FCS in RPMI for eg) for 5min at RT (followed by 3 washes), you can get levels of fluorescence intensity capable of detecting 12 individual "peaks"....normally 8 would be good.... and not get any detectable toxicity! (the initial fluorescence is so bright initially however, that it would be best to analyse your samples after day 2).

Anyway I have rambled on enough!

CU,

Ben

Thanks for the interesting point of view. I tried to switch to CPD from ebio as should be less toxic. I used 5uM of dye, 10M/ml T-cell, quench for 5' on pure FCS and 1 wash in complete medium and I am still loosing 40% of CD4 T-cell. Next time I will try to follow Ben suggestion and stain the T-cell in 10% FCS. I did a tritation of the dye and actually you cab go down to 2uM without loosing the staining.

@ Jacek , I am counting the CD4 T-cell ( trypan blue) before and after the staining. I will let you know.

Thanks a lot

I had the same problem. CFSE is not a problem itself as I used it 0.5 micromolar and still have 50% of cells. Using small volumes for staining and washing (1ml cell suspension 10 mills cells, and 3x1ml washing) I have about 70% recover.

Paolo, as I stated in my earlier remark above, it is not only the CFSE toxicity, but also the AICD, by which you would loose up to 50% of initial inoculate after 72 hours into culture with anti-CD3 or ConA, even at optimal concentrations of tghe stimulant, cell density and other culture conditions. Of course manipulation of cells after staining may also compound the problem...

We use CFSE at lower concentrations (usually 1uM) with less problem. Another possibility is to change CFSE for another dye, such as of the PKH family (we use PKH26 with good results).

Hope it helps.

Good luck, Dewton

Dear all thank you for all information. I would like to share a problem that I am  facing with detection of CFSE proliferating cells at day 5 of culture (loss of fluorescence of non stimulated cells and also of the other stimulated conditions).

 We used to stain at least 10M PBMC with 5 microM of CFSE in warm (37oC) RPMI serum free, then we wash twice with RPMI containing 10% of FCS and ressuspend in RPMI10%FCS.We count the cells with trypan blue, to check the viability, we have loss of cells during the staining, but that is not a problem in our experiment. After we check the fluorescence in the” Fortessa LSR II”. In the staining day, the fluorescence of CFSE stained PBMC is around 10^5. After 5 days of culture, the fluorescence of the non stimulated cells diminish to 10^4 or 10^3. In the initial experiments everything worked well, I could see the number of divisions, the cultured cells at day 5 still presenting FL1 fluorescence about 10^5, this protocol worked for months. Now when we stain the cells with the same protocol, at the day of CFSE staining, fluorescence of stained cells is around 10^5, and 5 days later the fluorescence of non stimulated condition is about 10^3 in FL1, and I cannot see the number of cell divisions. The plate is always incubated in a 37oC / 5% of CO2, in the dark, in some experiments we covered the plate. Anyone now what could be happening with my CFSE stained cells? 

Have you changed the CFSE reagent recently? Your problem may be the toxicity of CFSE. An easy test: if after staining and washing in PBS (you need colorless medium for that) the spun pellet of your cells is visibly yellow, they are overstained and toxic effects are very likely.  THe pellet after such staining should be very palely cream colored. I would suggest titrating your CFSE concentration down to 1 microM, and staining for 10-15 minutes.