Recovery of proteins through desalting columns such as the Zeba or PD-10 can be very sensitive to the volume applied (too much or too little). Make sure to use a volume within the recommended range. Also, use the recommended centrifugation conditions and the recommended protein concentration. There will always be some loss, but the proportion lost increases as the protein concentration decreases.

Poor recovery may also be due to over-labeling. A high labeling stoichiometry can cause the protein to become insoluble. Keep the labeling stoichiometry down to 1 or 2.

Try using a PD-10 desalting column. I've used them many times with much success.

Recovery of proteins through desalting columns such as the Zeba or PD-10 can be very sensitive to the volume applied (too much or too little). Make sure to use a volume within the recommended range. Also, use the recommended centrifugation conditions and the recommended protein concentration. There will always be some loss, but the proportion lost increases as the protein concentration decreases.

Poor recovery may also be due to over-labeling. A high labeling stoichiometry can cause the protein to become insoluble. Keep the labeling stoichiometry down to 1 or 2.

How are you estimating antibody recovery? If it is by measuring antibody activity (by ELISA for example) then it may indicate over-labelling with biotin and reduction of the biotin: antibody molar ratio may solve the problem.

If antibody loss is actually quantity loss then this can be due to aggregation caused by over labelling with biotin making the antibody hydrophobic which again can be solved by reducing the ratio of biotin: antibody.

Finally if loss is due to the antibody sticking to the surface of container, gel or dialysis membrane then I would suggest: after completion of the coupling reagent, add 100uL of 0.1M glycine for 1Hr to quench any left reactive biotin reagent. Then add bovine serum albumin equivalent to 10mg/mL in your reaction volume before dialysis or gel filtration. The BSA will prevent antibody sticking to surfaces as well as confer stability for your conjugate

I agree with Ahmed's answer: add carrier protein and dialysis, rather than columns, will reduce loss of reagent. But, some loss will always happen.

I agree with Ahmed. Make sure the glycine (or Tris also works) is neutral pH, and determine antibody concentration before adding BSA.

NAP-5 desalting column is an alternative for the removal of biotin.

Have your tried the

EZ-Link™ Sulfo-NHS-LC-Biotinylation Kit it has columns in it to get rid of the excess biotin.

Thank you all for your recommendations and sorry for the delay in answering. At the moment I have bought another biotinylated antibodies while I try to solve this problem.

I have not tried with NAP colums (NAP-5 or NAP-10) because the elution buffer volumen is higher than 500 ul and that means that if my starting volumen is 100 ul (this is the amount of antibody that I incubate with biotin) I will dilute the antibody 5 times or more, and this is something I have no interest in. For this reason, in GE healthcare web, they recommend to use the MicroSpin G-25 columns when the final elution volumen required is 100-150 ul. I have no tried this last columns, but are so similar to the Zeba Spin columns which I tried obtaining bad results also. About the columns provided with the kit (Peter Durcan), I tried it but again I get so diluted sample and bad recovery (they are similar to NAP ones).

About the volume applied and the recommended centrifugation conditions they are ok, but maybe you; Adam B Shapiro are rigth and it is due the concentration of the starting sample (0.5 mg/mL), maybe it is to diluted...but is the concentration in which the antibody is supplied and maybe a step for getting a more concentrated antibody also implies loss of antibody. About the over-labeling, I have optimized this step with anti-HRP trying x 30, x40 and x50 molar excess of biotin and I decided to work with x50 because with this amount I got to bing 8 biotin per AB, but maybe it is to much and cause insolubility problems (even though I did not see aggregates).

Ahmed Jehanli, I am estimating the amount of AB recovered measuring the absorbance at 280 nm. I never measured the AB prior to pass it through the purification column, and it would be an interesting control in order to know if at this point I have aggregation due to the excess of biotin or if the problem is the antibody sticking to the membrane (of columns or dialysis kit). In this case I can try adding glycine and BSA as you suggested.

Thanks again

I frequently label antibodies with biotin and the

Zeba™ Spin Desalting Columns, (40K MWCO or 7K, 0.5 mL: this size allows max 130ul protein sample for loading) works beautifully if you just follow the product manual. It is very fast with high recovery (almost 100%). Attention should be given is that this column can just remove 80% free biotin residual at one time. (you can use to two or three columns consecutively if you require much higher purity. If there is an washing step following your biotinylated antibody (before streptavidin-conjugate), somehow biotin residual is totally OK.

The other way is to do dialysis. Slide-A-Lyzer MINI device (10-100 ul) is perfect for you sample size. 20K is efficient for you 150 KDa IgG. This device is easy to handle and have high recovery.

https://www.thermofisher.com/us/en/home/life-science/protein-biology/protein-purification-isolation/protein-dialysis-desalting-concentration/dialysis-products/slide-a-lyzer-mini-dialysis-devices.html

Good luck.