Jason Bubier wrote:
> What kind of ELISA (direct, indirect, sandwich)? SANDWICH
> -What cytokine is of interest? IL-15 FAMILY
> -Are you using a commercially available antibody, or home-brewed?
> COMMERCIAL R&D Systems
> -What are you using as a blocking solution? PBS + 1% BSA
> -Are you putting your antibodies in the blocking solution? CAPTURE is in
> PBS, detection is diluted in blocking solution
>> ---
Now you've ruled out all the simple explanations, darn it!
I'm not familiar with the IL-15 family of cytokines, but perhaps I've
got something that will work for you.
You could try doing the ELISA as an indirect ELISA. Use some plates that
have high binding ability (The exact name escapes me right now, but I
believe I used to use an Immulon 4HBX plate) and put your sample of
interest on the bottom of the well. Let them bind overnight, maybe at
4degree C, maybe at room temp.
Next day wash, and add the capture antibody and incubate for however
long you please (I'm thinking for a reasonably long time, a few hours
perhaps). Then add a secondary antibody that is anti-primary. You decide
what reporter you want on the secondary.
Why would I suggest this when the sandwich ELISA is usually superior? If
there is something in the serum that binds the site recognized by
detection antibody that might explain the lack of detectable levels,
which I think is in line with what your colleagues have suggested. If
it's the capture antibody that has it's target blocked you'll still have
the same problem though.
Have you tried a plasma sample as opposed to a serum sample? It'd be a
beauty if it were that simple.
There's probably some separation techniques you could also use to get
different fractions of the serum, coupled with a good mass spec you
might be able to nip it in the bud quickly, but that's probably more
complicated than an ELISA (I can't speak from experience on mass spec
though). Let me know how it goes!
CMN
