Nk isolation with negative selection Miltenyi magnetic beads - with negative selection Miltenyi magnetic beads (Dec/05/2008 )

Has anyone worked with human NK cells isolated with miltenyi magnetic beads?

I tried for more than 10 occasions to isolate NK cells using negative selection and then stimulating these NK cells with 100U/ml of IL-15 overnight.

Subsequently I will activate them with a few stimulants such as PMA+ionomycin，　which are highly lymphocytes activators that can work at extrememly low levels. The problem I faced is, when I try to detect for intracellular IFN-gamma production using flow cytometry (I add golgistop to my cells), I either go NO response or very low response. And it is certainly not the problem of IFN-gamma antobodies because another person had managed to detect high IFN-gamma response with the same bottle of antibodies as well as same amounts of it.

So I suspect its could be something i did during the isolation steps that made NK cells unresponsive, I suppose NK cells are very fastidious. Has anyone done NK isolation before, can you share with me what are the dos and don't that can cause huge disparity in NK response?

My Nk cells do not look healthy after isolation. My colleague described healthy NK cells as perfect spheres, but mine looked slightly "distorted".

Please leave some comments if you have any experience in NK isolation and NK maintainance!!! Thank you so much!!!

Hello,

Seems like you may have at least two separate issues to deal with (isolation and stimulation).

You didn't specifiy the amount of starting material (ie. how many mLs of blood?). As you may know, NKs comprise a very small percentage of total PBMC, and therefore your isolation stategy must take this into consideration. It is probably better to use manual MACS MS columns, as opposed to an AutoMACS. In my experience, I simply POSITIVELY select with CD56-microbeads after I have depleted other cell subsets such as monos, granulocytes, B cells, and CD4s. I DO NOT DEPLETE with any CD8-beads, as there are a significant proportion of human NK cells that express CD8 (note that CD8 positive T cells do not express CD56 at any point in their life). I can usually recover a substantial number of CD56 positive cells from ~40mL of PB. As far as I know, CD56-microbead isolation does not activate human NKs, but someone else may be able to comment on this.

Now on to your stimulation.

First, note that at physiological levels in vivo, IL-15 functions in the development and maintaintance of NK cells by regulating expression of anti-apoptotic molecules.

Ex vivo, human NK cells should produce a significant amount of IFNg (detectable by ELISA after 10hrs, and ICS by 4hrs) in the presence of excess amounts of IL-15. These same cells will produce even more IFNg when IL-15 synergizes with IL-12 and/or IL-18. I forget how IL-15 signaling induces IFNg off the top of my head, but IL-12 and IL-18 signaling activates two different transcription factors (STAT4 and AP-1) that bind at different spots at the IFNg locus resulting in a ton of IFNg to be expressed. Therefore, you could try to stimulate your cells with various combos of the above cytokines.

With that said, you don't mention if you have a positive control (ie. a cell type, CD8 T cells with anti-CD3 stim, that you know will make IFNg in your assay). This will tell you if your ICS worked.

Your idea to use PMA/ionomycin seems like a good one, but take note that the signaling pathways that are activated are going to be completely different than activation via the above cytokines. I can't recall if anyone in my lab (or department) has stim'ed NKs with P/I, but give it a shot.

A note on NK cell morphology and health

I'm not exactly sure what slightly distorted means. But upon activation of various human (and mouse) cells, many of them adopt a 'tennis racket-like' morphology. This is especially true of cells containing granules (CD8s, NKs, and mast cells). The exact reason for this is not known, but has been attributed to both blasting cells and formation of immunological synapse. Depending on the expected number of cells recovered after seaparation, you really should count (with Trypan blue) and analyze purity by flow cytometry. This will allow you to roughly assess the overall health of the cells.

Finally, expansion of NK cells in vitro can be done to some degree, either by generating clones or expanding LAKs with exogenous IL-2.


I'm sorry if that was too much info, but I hope that it helps.

Thanks for sharing the details of your isolation/expt.

I can't find any glaring problems with your isolation procedure at first glance. Your PBMC counts are within range for the amount of PB you are beginning with. With 90-98% purity, which is on the high side for MACS separation especially via negative selection, you should be obtaining a large number of NK cells. I know that it is a pain to do purity checks, but given that you are consistently not able to get this expt to work I would highly recommend that you do this for every separation until you begin to acheive some success. It is perfectly understandable that you want to leave the cells as 'untouched' as possible.

Not to sound offensive, but perhaps it is something in your technique. Sometimes even simple differences in pipetting technique can have drastic effects later on in the experiment. Let the person who trained you (or have your boss) stand over your shoulder while you perform the isolation procedure.

In terms of using IL15 (or IL2)

I know that there is discrepancy in the conentration between different soruces of recombinant human cytokines (ie. BD uses a different 'unit' than the 'unit' that R&D uses). I will assume that somone has performed a bioassay to establish the specific activity of your cytokines. The "optimal concentrations" will likely vary from lab to lab. It is difficult to comment on your concentrations, but I would venture to guess that 100u/mL IL15 is more than sufficient to keep your cells alive O/N. This is certainly the case with IL2.

Experiment

So when I referred to using a positive control, I was indeed suggesting that you stimulate purified T cells in parallel with your NK cells. Or any other cell type that will give you a positive result in your assay. This will ensure that your experiment worked and that there are no problems with instrumentation, reagents, etc.

You're right, it is puzzling that you can detect NKT and CD8 T cell derived IFNg in your P/I stim but not from your NKs. Clearly, the 'contaminating cells' provide a somewhat internal control. If you really want to know if your NKs are functional do a cytotoxicity assay with a permissive cell line.

Once you begin doing your experiment with infected DCs, bear in mind that you may see no response from your NKs. There are numerous pathogens that can successfully evade immune detection by NKs.

It could be possible that your NKs are dying, in which case you could check with AnnexinV staining.

It is extremely frustrating when somone in the lab can get a particular experiment to work that you are having trouble with. In these cases, it is well worth your trouble to scrutinze every step and systematically check each to ensure that you doing things correctly. Often times, these sort of technical issues seem to work themselves out. Keep trying and you will get it to work.

Good luck! And I hope the information helped.