CYTOKINE secretion - assay or PCR?

M. Doherty M_Doherty at
Tue Feb 6 09:44:09 EST 1996

In article <alena.6.00105A9D at>, alena at wrote:

> I am kind of new in the field of immunology and have more experience in 
> cloning. Part of my new project aims the quantitation of cytokine production 
> by T cells. I was wondering what the easiest way is to do this. Is 
> Quantitative PCR the best and reliable way to do this? Or does this type of 
> PCR take a lot of time to optimize or is it just not possible to do? Or can 
> one best do cytokine assays (like the CTLL-2 assay for IL-2 determination?
> Everyone who has experience in this area could really be of help to me! So 
> please respond, thanks in advance!

I have a LOT of experience in doing cytokine measurement protocols, so
here's my take.

ELISA is more meaningful, since it measures protein as distinct from
message, and is also easier to do.  It is easier to optimise, mostly
because it is a more robust technique and therefore usually requires
little "tweaking".  In my experience (three different labs now) it has
been much easier to set up new ELISAs and make them perform reproducibly
than to do the same for PCR.

Certainly quantitative PCR is possible (although most people content
themselves with a semi-quantitative approach, since it is easier to set
up).  PCR is harder to set up and requires more effort to keep performing
at peak.  It also is more labour-intensive.  It reqires a larger initial
capital outlay (if you're not looking at setting up the procedure from
scratch, this probably doesn't matter).  It is more sensitive than ELISA
(given the caveat that what you are looking at may not reflect protein
output) and has the advantage that once you have PCR working well in the
lab, it is relatively trivial to get new primers made if you want to
examine another factor.  Finally, PCR lets you easily examine multiple
factors from small samples.  A million cells will give you enough RNA to
examine everything you could possibly want to look at, and still have
material for more.  You won't necessarily have the same capacity with
cytokines.  It is easy to store RNA or cDNA in the -70°C freezer, so that
you can re-analyse samples at a later date.

Bioassays (performed correctly) are in my opinion the gold standard, since
they measure activity, not protein or message.  However, to set them up
(depending on the assay) can require a great deal of work and they can be
very easy to mess up.  They can also be difficult to interpret, unless the
reagent measured is very well defined, since many other potential factors
in your supernatants can alter reactivity.  It all depends on the assay. 
Simple CTLL or CT4S proliferation assays are a doddle.

To sum it up - if you want to just measure a few cytokines occasionally,
ELISA is probably going to give you less headaches.  If you are going to
get into cytokine measurement seriously, I would go for PCR, but use
ELISAs and or bioassay as a backup to ensure that the cytokine message you
see actually relates to anything of biological.

Be aware that both ELISA and PCR have significant costs in consumables. 
Unless you are rolling in loot, DON'T buy commercial ELISA kits, unless
you know you only need to assay a very small number of samples.  It is
much cheaper to buy/beg/borrow/steal the antibodies you need and optimise
the assays yourself.  Once that's done, you can aliquot and freeze the
reagents and you'll have a lot of assays for a much lower cost.  There is
plenty of expertise available in setting ELISAs up, so don't be afraid to

Cheers, Mark

More information about the Immuno mailing list