Can you immobilise protein B on a sepharose column with cyanogen bromide then flush protein A down the column. then so long as XYZ have not bound all of the A the washed column will have quite pure ab

If the proteins interact very tightly, then they can be purified as a complex, if they are both expressed in a soluble form, by using affinity chromatography with at least one of the two proteins expressed with an affinity tag.

If the interaction is not tight enough for that, then purify them individually and then mix them together.

I'm not quite sure your goal of this purification. If you want to show the two proteins can interact directly you can try pairwise binding assay. If your aim is to see how they interact, you can try clone these two proteins into one plasmid and separate them with a linker. Then you can purify, crystalize, and analyze them.

Yuning Hou Thank you Yuning. My goal is to look at the 3D structure using X ray crystallography, that's why I want to have protein A-B. But, if you express this plasmid in bacteria, would you still get protein A-X, A-Y, A-Z, etc., in addition to A-B?

Try cell-free protein expression? Or purify A and B separately and then combine them?

If you express this plasmid in bacteria using a duet vector, both genes in the same plasmid chance of getting the complex is higher. If you are just expressing A and B, I am not sure why you will get A_X, A-Y, A-Z etc. If the protein are gone interact with protein in the expression system, it it better to purify them separately and then combine them.

Yuchen Zheng, if you express two fusion proteins with peptide linkers, you may see A-B-X or A-B-Y, etc., but I don't think you will see something like A-X or A-Y. There are plenty of articles about this topic. Another thing that needs your attention is purification optimization, you need wisely chose the tag, plasmid, as well as induction conditions.