ELISA calls for help

Deitiker, Philip R via methods%40net.bio.net (by pdeitik from bcm.edu)
Fri Apr 15 11:34:57 EST 2011

Methyl orange is close to yellow above ph4.0 you can play with the pH to optimize its adsorption spectrum. I believe that Methyl orange can be chemically activated with NO3 and sulfanilic acid and therefore theoretically can link to proteins.   

If I suspect pipetting errors the first step is to check the pipetters. 
I weigh water in a sensitive balance. Generally 10 pipette amounts. I will check a p10 at 3ul @ 10 = 30 mg, 10 ul @10 = 100 mg. For P20 I will check at 2ul and 20 ul each at 10. For P100 I check 10ul and 100ul at 10. For P1000 I will check 200ul and 1000ul at 10. 

I check both in forward and reverse mode. If there is a variance of more than 0.2% then I know its time to replace the seals, seals can be bought from rainin instruments, I always keep a few seals extra. Pipetter malfunction is a most common source of error that people generally don't notice until after the error has creeped into their work. Before big projects test the pipettors. 

For testing binding assays you want a negative control and a positive control. As the previous poster has stated make sure you block. If you are having radically different replicates results from expected there is generally two causes.

1. Your antibody has precipitated (centrifugation can remove  precipitates). Precipitation can form after freezing. Always freeze antibodies in a stable freezer, don't freeze in frost-free freezers as this leads to sublimation and temperature changes can cause loss of liquid from sealed eppendorf tubes. Precipitation can also occur in seras of very high titers (in the old days we used to do repeated immunizations with CFA) or in patients with autoimmune disorders where excess antibody production has occurred if not properly stored (sera can be diluted with BSA and stored). Freezing and rethawing can cause precipitation (sera should be immediately aliquoted). Complete compliment fixation occurs at 50'C, but 30 minutes at 37'C is sufficient to separate sera from clotted blood in mice. The use of anticoagulants such as citric acid prevents clotting so complete sera formation is prevented. Improperly prepared sera will have excess intracellular protein and can favor precipitation.    

2. Blocking is not adequate, not only block, but block the wells completely. I have seen instances were one well in every 4th replicate was off. Asked the person to show me the equipment and find that tips were not seated properly or the multi-pipette used for blocking was not working properly. If you are plating 50 ul of antigen and 50 ul of Antibody(s) then you need at least 150 ul of block to make sure there is no residual surface. I simply fill to the top of the well, if its not to the top of the well, then find out why. I used 0.5% micropulverized casein in PBS w/thimersol (prepared by diluting the caseine in PBS at 70'C) for mice. For human sera, backgrounds are going to be much higher, some people use goat serum to block, others use 2% BSA, etc. 
Whatever the source of block, you don't want it to react with your 2ndary antibody or developing reagent (e.g. 125-I protein A). Individually, I have studied patients with AID, and these patients tend to have high IgG and IgM levels, for each patient Caseine or BSA may work better. I am a good example, I am gluten sensitive, people with GSE tend to have high antibodies to dietary proteins, one of the most common dietary antigens is to milk proteins, casein is a milk protein and is the major antigen. Thus when I test my own sera I have high background reactivity to milk. Theoretically these antibodies should be IgE and IgA, however there is adequate leakage in several diseases to allow systemic infiltration of dietary allergens and IgG production. In fact some people have the IgA-less phenotype, in these individuals you see increased IgG production against antigens that would normally have IgA production. So be prepared to block and carry with different agents. One size does not fit all.  

The best way to test adhesion to tips is 125-I protein A solutions. This material is very sticky, and if you use forward mode you will generally find a 5 to 10% drop off between the first and second pipettings. The protein also drops when solution is transferred from old to new plastic tubes as a lot of protein sticks to the tubes. 
Dye-labeled proteins can accomplish the same task. One can measure the level of Dye in a solution transferred by pipette 1, 2, 3, 4, ... times. If you see a drop in the Dye, then you know protein is sticking. 

There are companies that claim to make less sticky tips. Sigmacote can be used on certain materials to prevent sticking, etc. If a problem is detected one can purchase or treat accordingly. Increasing the carrier protein concentration also helps. I typically have used 0.2% casein in PBS or 0.5% BSA in PBS. If you fail to use carrier antibody will end up on the side of test tubes. This is also true with 125-I protein A. 

With 125-I protein A, using clarified Ab and well calibrated Ab, I have gotten deviations in triplicates repeated less that 0.3% when CPMs are above 20,000. When one is seeing deviations of more than 2% one needs to start looking for sources of errors. At lower CPMs one has to expect wider deviations due to decay and randomnicity (BPD issues). 

-----Original Message-----
From: methods-bounces from oat.bio.indiana.edu [mailto:methods-bounces from oat.bio.indiana.edu] On Behalf Of WS
Sent: Wednesday, April 13, 2011 3:13 PM
To: methods from magpie.bio.indiana.edu
Subject: Re: ELISA calls for help

Dear Philip,

which dye(s) would you recommend for scrutinizing pipetting? Anything
that is yellow (for OD450) or do you have a personal favorite for
aqueous solutions?
Actually, I don't want to cook TMB with peroxidase and add acid, just
to be as close as possible to the original. The purpose I want to use
it for is calibrate a homemade pipetting robot for microplates. It's
for TMB based ELISAs


Methods mailing list
Methods from net.bio.net

More information about the Methods mailing list