How to Concentrate a Protein: Semi-permeable Membranes, Protein Precipitation & Chromatography

Concentrating a protein sample is one of those tasks that sounds simple, but choosing the wrong technique could ruin your precious sample. Whether you’re prepping for a downstream assay, trying to hit a working concentration, or just trying to shrink an unwieldy volume, you have options.

In this guide, we’ll walk through three of the most common approaches to protein concentration, each suited to different sample volumes, protein properties, and downstream needs. First, we’ll cover semi-permeable membranes which is the most popular approach for everyday use. Next, we’ll look at protein precipitation, focusing on the two methods you’ll reach for most often: ammonium sulfate and trichloroacetic acid (TCA). Finally, we’ll explore a method you might not have considered for concentration at all — chromatography.

By the end, you’ll have a clear sense of which technique fits your sample and how to get the most protein out the other side.

Structure of the membrane

Semi-permeable membranes are most often made of cellulose, and although they are diagramed as a solid sheet with little holes in it, electron micrographs reveal that they look more like a sponge. The size of the pores in the sponge-like structure determine the size of the molecules that can pass through the membrane. For simplicities sake, the manufactures translate this to a protein size, assuming that all proteins fold to a sphere-like shape.

This means there are a couple of things that you should bear in mind when using these membranes.

First, the size and shape of the pores is irregular, so you should be careful when using a membrane to ensure that there is a significant difference between the size of your protein of interest and the rating of the membrane. Dialyzing your 11 kilodalton protein in a membrane rated at 10 kilodaltons will likely result in the slow disappearance of your protein over time.

Second, your protein or molecule of interest may not be shaped like a sphere, allowing it to pass through the pores with some measurable efficiency. Assuming that your favorite protein is reasonably shaped (most are), and that you are able to purchase a product with a suitable pore size (several are available), then these membranes can be used to concentrate your protein.

Dialysis bags or cassettes

While dialysis is a great way to exchange the solution conditions of proteins or DNA, there is a property of the membranes that allows some control over the volume of the protein sample as well, and manipulating the volume of the protein solution is what protein concentration is all about. This special property? Water moves quickly and easily across the membrane, even faster than ions, which are ‘bulked-up’ with hydration layers. So when there is a strong difference in the solute concentration on opposite sides of a membrane, water will rush across the membrane to dilute the salty side more quickly than the salts can make the opposite journey.

The result is that the volume of the original high-solute solution grows, and the original low-solute solution shrinks. This works against you if you are dialyzing a high-salt column fraction into a low-salt storage buffer, but you can make it work for you as well.

The principle that I described above for salt also works for glycerol. Many protein storage buffers contain glycerol to help stabilize the protein, but if you push the concentration up to the 40-50% range (as in many commercial restriction enzyme preparations), you can reduce the volume of your dialyzed protein fraction to as little as 20% of the original volume. This trick will even work if your original sample contains a much higher salt concentration than the dialysis buffer, although the volume reduction will be reduced to 30-50% of the original sample volume, depending on the salt concentrations.

This same principle is also exploited in many commercial dialysis-based concentration solutions. However, in this case large polymers are used rather than small molecules. These large polymers are, on average, too large to pass through the membrane pores themselves, which largely prevents the dialysis from coming to equilibrium. This means that the longer the dialysis bag/cassette sits in one of these solutions, the more water is extracted, and the smaller the volume of the protein solution.

Consequently, much larger volume reductions are possible compared to glycerol solutions, but failure to properly monitor the dialysis could result in volume reduction beyond the critical point where the protein precipitates out of solution (or, if you’re lucky, forms a beautiful crystal).

If you don’t want to buy the proprietary mystery solution to concentrate your protein, you can do this yourself by using very high molecular weight preparations of polyethylene glycol (PEG) and a membrane with a very small average pore size. The PEG can be dissolved to form a very high concentration solution, or even used straight out of the bottle by laying a sample in a dialysis bag directly onto a bed of PEG chips, which will draw the water out of the sample.

One of the things to bear in mind when using either the commercial solution or the homemade variety is that the polymer solution will contain a mixed population of sizes, and that while most polymers adopt a spherical, globular shape in solution, they are not folded as a protein is and more readily adopt shapes that allow them to pass through the pores in the membrane and contaminate your protein sample. Therefore, this method probably shouldn’t be used unless the protein is going to undergo further purification steps that will remove possible polymer contaminants.

Protein concentrators

Commercial protein concentrators also use semi-permeable membranes, but substitute another force for the osmotic pressure that the dialysis methods use. The most common of these use centrifugation to force the protein solution through a membrane, but others use a pressure differential, either by pumping nitrogen into a sealed vessel above the protein solution or by creating a vacuum on the opposite side of the membrane from the protein solution. As with the polymer methods above, these concentrators have to be closely monitored to ensure that the protein sample doesn’t precipitate or completely run dry.

One common problem with these concentrators is that the rate at which the solution is passing through the membrane changes over the course of time. In other words, the first half of the solution passes through quickly, but then the volume appears to stabilize. Sometimes this is a sign of a disaster – your favorite protein is now embedded in the membrane, clogging it – but sometimes this is simply an effect of poor mixing of the sample.

Before giving up on the sample or moving it to a new concentrator, try gently pipetting the remaining solution over the membrane several times (without touching it) and continue concentrating. To prevent this problem, many large-scale concentrators contain stirring mechanisms incorporated into the design. In addition, some centrifugation-based concentrators have also been specifically designed to avoid this situation, but you can negate this design feature if you use the wrong type of centrifuge rotor (fixed-angle versus swing-bucket), so make sure you read the instructions.

The Amazing Disappearing Protein

The biggest complaint about semi-permeable membranes is protein loss. This occurs for some proteins during dialysis, and for even more during force-induced protein concentration. It is an unfortunate fact of life that some proteins are just ‘sticky’ and love cellulose membranes. If you encounter this, and you’re certain your protein didn’t simply pass through the membrane, there are a couple of things you can try.

Change membranes or manufacturers. Many of the common concentrators use regenerated cellulose membranes, but even though two different manufacturers both use this material to make their membranes, there will still be subtle chemical differences between them, and that might make all the difference in the world to your protein. Even within a particular manufacturer, there will be other types of membranes offered including derivatives of cellulose or polyethersulfone that you can try. Don’t be afraid to ask for free samples to test conditions.

Pre-treat the membrane. Another direction to try is to pre-bind likely binding sites on the membrane or apparatus with a (hopefully) inert protein like bovine serum albumin (BSA). I have personally had great success with this, but it comes at a price – you will have BSA contaminating your protein sample after dialysis or concentration, which could complicate determining the concentration of your protein of interest after the procedure. This also requires some forethought on controls, since you need to demonstrate that the contaminating BSA isn’t responsible for whatever effect you are attributing to your protein of interest.

Sometimes, despite all the tricks in the book, semi-permeable membranes just aren’t going to work for some proteins or for some applications. In parts two and three of this series, I’ll describe some other approaches to this problem.

Protein Precipitation

Protein precipitation occurs primarily via hydrophobic aggregation, either by subtly disrupting the folded structure of the protein and exposing more of its hydrophobic interior to the solution, or by dehydrating the shells of water molecules that form around hydrophobic patches on the surface of properly folded proteins. Once the proteins start aggregating into larger structures, the amount of water per protein drops, enhancing the density differences between the proteins and the solute. Once these aggregates grow large enough, and if there are enough of them, they disrupt the path of light through the solution, giving it a cloudy appearance. In addition, the density differences become great enough for the aggregates to be readily pelleted in the centrifuge. Because these methods depend on the protein molecules “finding” one another in solution and forming these aggregates, the efficiency of this method depends on the concentration of the protein being precipitated – the lower the concentration, the harder it is to form aggregates. There are two practical reagents used to precipitate proteins: ammonium sulfate and trichloroacetic acid.

Ammonium sulfate

How it works.  Salting proteins out of solution was actually discovered over 120 years ago by Franz Hofmeister when he noticed that the addition of different salts caused precipitate to form in solutions of egg whites. He ordered the anions and cations by their ability to precipitate proteins in what is known as the Hofmeister series. Biochemistry has spent the 120 years since using this trick to purify proteins while also arguing about how it works. The current, mostly-accepted mechanism says that salting proteins out of solution occurs when the water molecules are titrated away from the solvent shells around the protein to the solvent shells around the ions that make up the salt. Salts high in the Hofmeister series are the most efficient at protein precipitation because of the large, stable solvent shells they maintain. This increases the surface tension of the solution, which effectively increases the hydrophobic effect, which stabilizes the protein structure while also encouraging the hydrophobic regions on the surfaces of different molecules to interact, affecting aggregation. Because of this, proteins with a larger amount of hydrophobic surface character precipitate at lower salt concentrations than one with little hydrophobic surface character, and these protein to protein differences are exploited during protein purification procedures.

In practical terms.  The go-to salt for purification or protein concentration is ammonium sulfate, since both the anion and cation are both high in the Hofmeister series. This salt has a high solubility at around 4M, with most proteins precipitating by the time the salt concentration reaches 3.2M. These concentrations can be obtained either by adding the salt directly to the protein solution or by the addition of a calculated volume of a saturated solution of the salt. Adding the salt directly to the protein solution helps keep the total volume of the solution under control, but can be problematic if the volume of the initial sample is low. Once the salt is dissolved or the solutions are mixed, the protein is allowed to precipitate for 30-60 minutes on ice (longer if the protein concentration is low) before pelleting the protein in a centrifuge. Because salting out stabilizes the protein structure, the pelleted protein will almost always readily re-dissolve into a buffer (lacking ammonium sulfate) with enzymes maintaining their specific activity (some measure of activity/mass of protein). However, even though salting out occurs via a phase transition mechanism, some quantity of salt will come down with the protein, leaving you with an undefined solution once you re-dissolve the protein. This requires dialysis or a de-salting (size exclusion) column to move the protein into a defined solution, which could result in some dilution of your now-concentrated protein.

Trichloroacetic acid

How it works. Proteins can be efficiently precipitated with trichloroacetic acid (TCA), acetone, or even ethanol, although the concentrations at which these (mostly) miscible organic solvents function can vary greatly. As with ammonium sulfate, the mechanism of precipitation is hydrophobic aggregation. However, in addition to disruption of the solvation layers of the proteins, these compounds also partially denature the proteins, exposing even more hydrophobic surface to the solvent.

In practical terms.  TCA is often the compound of choice amongst the miscible organics because it is effective at lower concentrations than the others – ~15% for TCA, ~75% for acetone, and ~90% for ethanol – which means the sample volume doesn’t increase dramatically. Because of this, the protein concentration remains higher during a TCA precipitation, increasing the efficiency of the precipitation. However, since TCA (as well as the others) partially denature the protein, there is a good chance that your protein of interest won’t dissolve back into an aqueous buffer in the absence of a detergent (SDS); and even if it does, it may have a markedly lower specific activity. For this reason, these methods are normally only employed with samples that don’t require functional enzymes, such as SDS-PAGE or mass spectrometry analysis. Because TCA is an acid, the protein pellet is either washed with 75% acetone to remove the TCA, or base is added after the pellet is resuspended in SDS-PAGE sample buffer (until the bromophenol blue turns from yellow back to blue) to neutralize the pH. However, if you happen to be lucky and have a protein that does tolerate this treatment, this method has the advantage of precipitating the protein with a minimum of salt in the pellet, possibly eliminating follow-up desalting procedures, depending on your application.

As always with proteins, your results may vary; but for many proteins precipitation has proven to be a good concentration method. As always, if you have any tips, hints, or suggestions to share, please speak up in the comments section!  Next up: protein concentration via chromatography.

Chromatography

While chromatography resins are an obvious choice for protein purification, they can also be used to concentrate a dilute protein with a couple of modifications to the approach.

Unlike the other methods detailed in this series, this approach requires more knowledge of the protein of interest than just its molecular weight. Since we are often trying to concentrate a protein that we have already purified, however, that knowledge – what chromatography resins the protein will bind to and under what conditions it does so – is at our disposal.  But, the protocol used with a particular column during purification of the protein may not be the best protocol for concentrating the already pure protein. Rather than trying to maximize the resolution power of these resins, the key to using them for protein concentration is to recover the protein from the column in the smallest volume possible. This means using the smallest column that will bind all of the protein in the sample and/or eluting all of the protein at once rather than allowing the protein to trail off a column. This mission is best accomplished with two types of chromatography resins – affinity and ion-exchange.

Affinity Resins

If your protein of interest was expressed with an affinity tag, most likely this tag was utilized in the first step of a multi-column purification because of their ability to remove the vast majority of the contaminating proteins in one step. However, if the tag isn’t removed in the course of the purification, then it can be quite handy for concentrating the protein at the end of the prep. Maltose binding protein (MBP), glutathione-s-transferase (GST), or histidine tags are the most useful for this purpose. The only other prerequisite for this method to work is that the protein has been bound to, and extensively washed on, one to two other columns in order to remove the elution molecules (maltose, glutathione, or imidazole, respectively) from the affinity tag.

The next key to this approach is to use the smallest column (volume of resin) possible to bind all the protein in your sample. Oversizing an affinity column generally has little effect on the purification power of the resin, with the exception of immobilized metal resins used for histidine-tagged proteins. However, larger columns lead to larger eluates and lower protein concentrations. Since you are now working with a purified protein, you can calculate the total amount of protein that you have, and better estimate how much resin you will need to bind it from the technical information that came with the resin. Due to the high binding capacity of these resins, you will likely be using less than a milliliter of resin, allowing you to collect very small fractions from the column, with most of the protein eluting in a single column volume. Alternatively, after washing you can load the resin into a disposable, empty spin column and elute the resin by centrifugation after a 5-10 minute incubation with the elution buffer.

Ion-exchange Resins

Ion-exchange resins bind proteins via the specific ion moieties displayed on the surface of the matrix, and include Q (NH₃⁺), S (SO₄^2-), and Bio-Rex70 (COO^–), amongst others. Unlike affinity resins, it is actually quite important to oversize an ion-exchange column during protein purification. This is because ion-exchange resins often don’t only resolve the protein of interest from contaminants during the binding phase, but also during the elution phase of the procedure. Consequently, you may or may not have access to small ion-exchange columns. If you are using a column in a FPLC, this may not matter, since these systems are designed to elute proteins in sharp, concentrated peaks. However, if you are using loose resin in a gravity column, then you will want to minimize the size of the column as discussed with the affinity resins.

When using an ion-exchange column for purification, the column is normally eluted with a salt gradient, and the ‘slower’ the gradient (ie – the rate that one moves from a low salt to a high salt concentration buffer) the better the resolution of the proteins, within reason. Sometimes under these conditions the protein of interest will elute in a nice sharp band, but some proteins aren’t as cooperative. When using an ion-exchange column for protein concentration, you can best ensure eluting your protein of interest in the smallest volume possible by not running a gradient at all, but rather to “bump” the protein off the column with a high salt buffer. The exact concentration of salt that you need is empirically determined from the purification profile, and I like to add about 50 mM to that just to be sure. For example, if the protein of interest eluted at 200 to 220 mM KCl during the prep, then I would bump the protein off the same resin with 270mM KCl when using the column to concentrate the protein.

One aspect of this approach that may be a problem for some proteins is that the protein of interest is now sitting in a buffer that is high in salt or that contains glutathione or maltose, requiring dialysis or a desalting (size exclusion) column to remove. However, this isn’t always a problem, as some proteins are best stored (ie – more stable) in high salt buffers, or the elution buffer may not affect the behavior of the protein.

Final Thoughts

Hopefully, I’ve given you some alternative approaches to consider the next time you need to concentrate your protein. If you have a favorite method that I’ve neglected to mention, please tell us about it!

Jode Plank

Jody gained a PhD in Biochemistry from Duke University, which was followed by a postdoc at the University of California at Davis. He is now the Manager – Product & Analytics Group at American Chemical Society.

Share this article: