Re: How long do I centrifuge serum to remove denatured proteins?

Dear Sharon,

Your research sounds very interesting! I wouldn’t have thought to use serum proteins as medical lubricants, but that is absolutely brilliant. I can appreciate that you are yourself a scientist with a different background than my own, so I will try to give you a detailed yet high level picture of the processes governing protein folding and aggregation from a biophysical perspective.

As you can likely imagine, aggregation is a macroscopic property which can have many meanings and mechanisms; we can clearly differentiate the sort of ordered fibrillization process associated with amyloidogenic proteins (where the aggregate has an ordered, regular structure) from the sort of stochastic mess often associated with aggregation from an unfolded state, of the kind you are likely dealing with. We should also use this polymorphism to think about reversible protein folding and folding landscapes.

I answered a previous question from the MadSci network regarding prion formation that is relevant to your question; specifically, the idea that a protein should really be described as an ensemble of states. The concept of an unfolded protein existing in a large number of states with roughly equal probability should be a familiar one to you. The concept that a native protein is also dynamic is also an important consideration. This is still an area of research that is quite active (some of the major players include S. Walter Englander at U. Penn, Ken Dill at UCSF, Ernesto Freire at JHU, and Chris Dobson in the UK, just to name a few), the common model being derived from a variety of experimental and theoretical approaches. Essentially, every protein has a finite probability of adopting any conformation between the fully folded state and the unfolded state(s). In fact, we can use the value of a proteins folding free energy to quantitate this probability; assuming a dG ~ 5 kcal/mol, at room temperature the equilibrium constant between N and U is roughly Keq ~ 5000. At any time, one in five thousand protein molecules has unfolded, and will generally quickly refold. For monomeric unfolding, this is true independent of the concentration; life gets only a little more complicated for multimeric proteins, but as you’re quickly learning reversible unfolding is not the only pathway available to most proteins.

In the real, non-ideal world, any number of competing pathways may be at play that in general will result in less than perfect reversibility of protein folding/unfolding pathways. Here is a Nature review from Chris Dobson you should read; in one of the figures, he shows a schematic that illustrates how aggregation pathways compete with the simple equilibrium between folded and unfolded states of the protein. It is commonly believed that the tendency of folded proteins to occasionally sample the unfolded state that makes them prone to these off-pathway aggregation pathways; predictably, aggregation gets worse with increasing concentrations of protein. Aggregation is also initially a kinetically slow process (generally); you may observe that if you centrifuge a sample and remove the aggregates that they return after some period of time (short intervals for more concentrated samples, or over several days in more dilute situations). In a clarified system, you may be able to run experiments immediately without observing evidence for aggregation; however, once “nucleated” aggregation can proceed rapidly.

It is generally accepted that aggregation occurs through hydrophobic interactions and/or disulfide (mis)bonding. The hydrophobic effect in regards to aggregation phenomena is generally relevant under “native conditions”, that is those solution conditions that favor more compact rather than expanded structures. This is clearly seen in the effect of denaturants (Urea, guanidine) in promoting reversibility and minimizing aggregation; temperature induced unfolding almost always leads to aggregation since the basic hydrodynamic properties of the system still favor a compact protein state, though the high temperature tends to make the native state less favorable than an unfolded state. Practically speaking, you can routinely include reductants (DTT, TCEP), or selectively block non-bonded cysteins to minimize the scavenging effect of free thiol groups, though this may not be practical in your applications. This simple model holds for studies of single proteins, but the rules are quite different in considering crowded systems; hypothetically, you might imagine glyocoprotein A may tend to aggregate at a certain pH or temperature, but albumin does not and so the net effect of having high albumin concentrations is to limit the tendency of glycoprotein to aggregate.

Aggregation tends to be non-specific (obvious exceptions include amyloidogenic fibril formation); in a heterogeneous system like serum, any one protein may be prone to aggregation under certain conditions, but there are ways to minimize this effect. You will find published examples describing the effect of “additives” that minimize protein aggregation; naturally occurring osmolytes tend to do just that (D. Wayne Bolen has extensively studied the thermodynamic effects of osmolytes on protein folding energetics). Certainly the intracellular environment is highly crowded and tends to minimize effects of low protein stability or aggregation through a combination of a local dilution effect (two unfolded proteins not being able to see one another) and an effect on water activity.

I realize this is a more lengthy “answer” than you were probably looking for. The short answer is that a normal desktop centrifuge can be used to spin out protein aggregates; generally ~ 13-14000 rpm for 5-10 minutes; from my discussion above, it should be clear to you that spinning out aggregates may or may not eliminate the problem (over a day or so, you may continue to loose protein to aggregation). My advice to you is to take the time to characterize which protein(s) is causing you problems; it may be that if you ran a gel on the aggregate there may be one or two major contaminants that tend to aggregate. You may wish to use a concentrated BSA solution in lieu of serum, or immunodeplete serum of the primary aggregating offender.

I would encourage you to scan the literature for reviews on protein folding and protein aggregation to pad your baseline understanding of the current models favored by active research in these fields. As I mentioned, the Nature review from Chris Dobson is very much worth reading, though his discussion of aggregation phenomena is limited to amyloidogenesis (that being one of his primary research interests). There is an article in the journal Biotechnology Advances from Harmen H.J. de Jongh of the Netherlands that may be relevant to your current applications, describing the disulfide bonding affects aggregation (I will admit to a lack of familiarity with this article, but the abstract seemed particularly relevant to your work). Your own literature searches will likely be the best place to begin working through your problems.

Regards,
Dr. James Kranz