[Immunology] Re: ELISA problems ('dead wells')

GS not at home.com
Thu Feb 23 15:37:19 EST 2006

m.bristow at apu.ac.uk wrote:
> Hi,
> I'm experiencing a new (to me) problem with ELISA.  I have an double
> sandwich human IgA ELISA using HRP (and TMB as the substrate) using
> Greiner Medium bind plates.    The standard curve is good.  I have had
> some problems with poor CV% on some runs and we are now seeing 'dead
> well' - wells that just don't develop any colour (9 in one 96 well
> plate today) and some wells with much more colour than others, giving
> us poor CV% in wells.
> I can't see how these dead wells could be caused by the overnight
> coating of the capture antibody (STAR92) in the fridge or by the
> incubation/ shaking stages (can plate shaking cause problems?).  I'm
> also confident that this isn't an operator problem.
> I'm left with the feeling it is the plate washer (Dynex Ultrawash),
> though I can't  think of any way it could be causing it.
> I'd appreciate any ideas/ thoughts/guidance on what could be causing
> these 'dead wells' and/ or if anybody has had similiar problems to
> this.
> Thanks,
> Matt

Here's some things worth looking at -

The first thing to do is identify what has changed. Are you using a new 
batch of plates? New batch of capture antibody? New batch of pipette 
tips? Is there any consistency in the position of the "dead" wells?

Are you using a multichannel pipette or a repeating dispenser? When did 
you last - (a) calibrate the pipette/dispenser? (b) Check for channel 
leakage? check for contamination of the pipette? (d) clean it?

Just because you're confident it's not an operator problem doesn't mean 
it isn't. If more than one person is doing the assay and they're all 
seeing the same issue, it's probably not an operator thing. If there's 
only one assayist, get an independent observer (with ELISA experience) 
to watch the operator. Sometimes it's the only way to spot a problem.

I don't think that shaking would necessarily cause a problem, if the 
plate is sealed during the operations. If it's not sealed, there is the 
possibility of splashes on to the top of the plate or into other wells.

You don't want to do this until you've solved the problem, but have you 
looked at what effect shaking actually has? I suspect you'll find that 
even with static incubation, something like 90% of the binding will 
happen in the the first 15 minutes at 37 degrees C.

Good luck


More information about the Immuno mailing list