Protocol Online logo
Top : New Forum Archives (2009-): : ELISA and Immunoassay

Troubleshooting development of sandwich ELISA - Where am I going wrong? (Feb/01/2010 )

Hi,

I have recently started a PhD in biomaterials. My first project is to try and develop a sensitive human albumin ELISA. I used to work for an ELISA department for my placement in their R&D section so I guess I thought I knew most of the ways it could be done but this is making me pull my hairs out!

I started off by coating Goat anti HSA (Human Serum Albumin) onto ELISA plates, double diluting across columns but left 12 blank, in sodium carbonate/bicarbonate buffer overnight in a moistbox at 4C.
I then emptied the wells and banged them dry. I subsequently filled them with 300µl of block buffer (I've tried three different ones: PBS 3% skim milk, PBS 1% BSA, Stabilguard (commercial BSA free blocker- tried this as I was afraid components in the skim milk could cross react with the antibodies. Only tried PBS 1%BSA as it was featured in several research papers). I blocked overnight at room temperature.
I then emptied the block, banged the plates dry and added my sample (HSA in TBS) which was double diluted across the rows, leaving H blank.
I incubated at RT for 1 hr and then washed the plate (PBS 0.1%Tween 20, 300µl/well, 3 washes per step)
I added detection antibody which was Rabbit anti HSA in TBS 0.05% Tween 20 and incubated for another hour.
Another wash step
Conjugate= Goat anti Rabbit IgG HRP in TBS 0.05% Tween 20 for 1 Hr
Wash
TMB (from tablet, made up in DMSO and phosphate-citrate buffer)
Stop= 1M H2SO4
Read at 450 nm

I tried this several times with different concentrations and different blocks but the capture antibody dilutes fine and the antigen (HSA) doesn't seem to be a parameter at all. The blank always results in colour. Naturally I then did an Ouchterlony to try and look at the specificity of my antibodies. No cross-reactivity between Capture and Detection nor Conjugate was found. The Rabbit anti HSA did react with BSA so it was hypothesised that it might have reacted with the block or something. A new antibody was ordered.

The same assay was run but this time with Monoclonal mouse anti HSA as the capture antibody, Gt anti HSA as the detection antibody and Rb anti Gt IgG HRP as the conjugate. Two blocks were tried (PBS 1% BSA and Stabilguard (No BSA)). The same thing happened again?

Am I incapable of finding antibodies which aren't crossreacting? Did I swap the wrong one?
Am I doing anything else wrong? I have done this for such a long time I think I am becoming blind to a lot of things so any pointers would be very helpful.

Thank you for reading.

Nilla

-runstrog-

Hello,

Human albumin and bovine albumin are structurally very similar. To avoid problems do not block with any proteins similar to what you wish to detect. All the commercial blockers probably have some type of albumin. I suggest using SeaBlock, fish protein, (East Coast Biologics) in a buffer with glycine.

You can screen your detection abs for reaction to the blocking agents by coating wells with your various blocking solutions, adding your capture or detection abs, then conjugates. (washing in between steps) If there is binding then do not use that agent for blocking.

Another option would be to run the Ouchterlony with your capture and detection abs against the blocking agents to see if any react.

In your elisa Make sure you have 3-5 washes between steps especially before the conjugate is added. (it was unclear if you washed several times between each addition in each of your experiments.

let us know how it works out.

-sgt4boston-

sgt4boston on Feb 1 2010, 08:20 PM said:

Hello,

Human albumin and bovine albumin are structurally very similar. To avoid problems do not block with any proteins similar to what you wish to detect. All the commercial blockers probably have some type of albumin. I suggest using SeaBlock, fish protein, (East Coast Biologics) in a buffer with glycine.

You can screen your detection abs for reaction to the blocking agents by coating wells with your various blocking solutions, adding your capture or detection abs, then conjugates. (washing in between steps) If there is binding then do not use that agent for blocking.

Another option would be to run the Ouchterlony with your capture and detection abs against the blocking agents to see if any react.

In your elisa Make sure you have 3-5 washes between steps especially before the conjugate is added. (it was unclear if you washed several times between each addition in each of your experiments.

let us know how it works out.



Cheers for answering :P
I do wash 3 times in every wash step so that should be ok. Have never heard of SeaBlock so will investigate that further. Will also try the Ouchterlony again with the Abs and the blocking solutions.
In the meantime I'm going to try a method described in a research paper to see how it compares.
Will keep you informed on how I get along.

Nilla

-runstrog-

Don't overlook coating the wells with the various "blockers" and screening them with your abs for reactivity. This would give you a good indication if you will have a problem in your sandwich assay. All those blockers should give you a negative result if they work properly.

-sgt4boston-

I have now made a detailed screening of my blocks and they are all ok to use. Clearly the fault is not there. I also tried another configuration of antibodies (Ms anti HSA -Rb anti HSA -Gt anti Rb HRP) but this didn't work either. I wanted to try it as all my other assays had been utilising Gt-anti HSA in at least one of the steps.
I still had the same problem of antibody diluting out but antigen not making any impact at all. I still didn't get a blank.
I am now trying a conjugated antibody so the system is Ms-anti HSA and Sheep anti HSA- HRP.
Hopefully this might solve the problem!

Will let you know but would always welcome other suggestions of course :(

Nilla

-runstrog-

How do your abs respond when you react them with HSA on the wells and blocked with seablock/glycine? Your capture and detection abs should each work independantly and bind to the hsa in the wells. This would be the 'screening' test the ab supplier does.

Other options are to see if any of the suppliers sell paired abs?
Also if the abs you purchased have info on their sensitivities.

-sgt4boston-