This is one of those "all depends" questions.

The amount and volume you can load depends on the hetergeneity of the proteins in the sample, the dynamic range of the amounts of the proteins, the tendency of your protein to aggregate if it becomes to concentrated, the amount of salt in the buffer, the shape of your protein (i.e. globular, cigar-shaped) and size difference in molecular weight of your tagged protein vs. the contaminants.

First step would be to run a non-reducing gel on the eluted protein after the nickel affintiy column to determine the yield of the tagged protein, how much of it there is compared to contaminants, and how close in size the other proteins are.

Rule of thumb is for comparable amounts of similarly-shaped protein, SEC can ideally separate to baseline proteins that differ in MW by a factor of 2. You can get partial separation (a non-baseline separation) if the proteins differ in MW by at least 20%. However if you have 5-10xfold more of one or the other, the peaks will overlap, and separation may not be possible.

I would take 200 ug of your eluted protein in 100 ul, and load on the column as a test run at 0.5 ml/min. Then I would run a gel on the peaks to determine exactly where the protein is eluting, and how pure the protein peak is. This is especially important if you see a large void volume peak which might indicate aggregation of the protein.

Start with 250 ug and after getting results (chromatogram and sds-page) decide if you can increase this amount.

Hi Alex, I agree with Mary, if you have 2-4 proteins that are well separated in MW then you could get a good separation if you choose a gel filtration resin that maximizes the spread.

But all bets are off if there are protein-protein interactions or aggregates in your sample.

Generally speaking, gel filtration is not a good final step in a purification procedure, unless you have ~2-3 proteins that don't interact and can be well separated on a particular resin.

In theory, protein separation on gel filtration depends on the sample volume applied and not the protein concentration.

But quoting Mary it: "all depends".

In SEC the most important factor is VOLUME. You should apply your sample in a volume of 2 to 5% of the BED volume, to have an optimum number of theoretical plates in your column. Your column also has to be long, even the best resins display reduced separation performance at lengths below 300 mm, and lengths of 600 to 1200 mm are usually employed. This depends on the resin quality (particle average size and homogeneity).

An example, Let us assume your column is 900 X 16 mm (length X internal diameter):

-Your internal section is ~2.0 cm2 and the total volume of bed is ~181.0 ml, when completely filled (empty space in the column spoils its performance, it should not be any).

-Your sample volume should be between 3.6 and 9 ml. Now regarding the protein amount, you can concentrate your sample and apply as much protein as possible, but as Mary Cristine mentioned, factors such as the sample viscosity and protein aggregation come into focus.

Viscosity: If your sample has high viscosity, the liquid will tend to move as a separate phase in between the gel beads, this will limit the diffusion of proteins through the bead pores and into the bead interior volume and separation will not be good.

Insoluble material: If a protein or any sample component aggregates, the insoluble particles will clog the column, will reduce flow rates and might damage the resin.

Flow Rates: DO NOT EXCEED flow rates recommended by the maker of the resin/column. Flow rate is usually given in cm/min or mm/s, and represents the velocity of a virtual water circular front moving along your column. to convert this effective volume per unit of time (e.g. ml/min) multiply the flow rate by the section of your column (in the example above 2.0 cm2). Too fast flow does not give enough time to particles for filtering in and out through the resin bead pores.

Best wishes,

Rogelio

For loading, your limit is the solubility (make sure you centrifuge or filter possible precipitates). The resolution will depend on the volume injected, that should be no more than 2% of the column volume.

Regarding how separable the proteins will be, it depends on how they may interact at high concentrations. Proteins interact non-specifically, it is only a matter on how high the concentration is and the particularities of the mixture. This will be empirical, unless you know the behavior of the proteins in the mixture. These interactions are often electrostatic, so you can run test expreiments increasing NaCl concentration. In addition it depends how close in size the impurities are from your protein.

I would not risk a precious sample by risking taking the sample to very high concentrations just to have a single injection in a preparative SEC. You are better off runing 5 different SECs with the optimal volume (2% approx).

If you have a fairly pure protein and you know it does not stick to the others, your limit is the solubility. Once again, do several runs to be on the safe side, you may even have a power cut in the middle of that single run.

When you are using gel filtration the important thing to take into consideration is the bed volume which depends on the the height and the radius of the column you are using, also you should know the capacity of the gel you are using. I recommend that you have a look on one of earlier volumes of method of enzymology where there are chapters deal with this topics .Also there is a book under the title of protein purification by Scope I think will be very useful for you to have a look at ,both have information about this. with Good Luck