Written and maintained by CASRAI Editorial Board
Last updated
Protein refolding is the step that returns a denatured protein, usually one recovered from inclusion bodies, to its native, active fold by removing the denaturant under conditions that favour folding over aggregation. The deciding variable is protein concentration: aggregation speeds up faster than folding as concentration rises, so most refolding methods are different ways of keeping the protein dilute while the denaturant is removed.
Short answer
Refolding means dissolving aggregated recombinant protein in a strong denaturant, reducing its disulfide bonds if it has them, and then lowering the denaturant so the chain can fold. Dilution, dialysis and gradual (fed-batch) addition are the main ways to remove the denaturant. Each trades buffer volume, time and aggregation risk differently, and reported recoveries range widely between laboratory and industrial reports. A soluble product is not proof of a correct fold, so a refolded protein needs separate checks for aggregation and for activity.
Definition and scope
Recombinant proteins made in bacteria often accumulate as insoluble inclusion bodies. Yamaguchi and Miyazaki (2014) describe these aggregates as being “denatured and dissolved with a high concentration of denaturant”, with urea, guanidine hydrochloride (GdnHCl) or ionic detergents such as N-lauroylsarcosine named as the solubilizing agents. When the protein contains disulfide bonds, a reducing agent such as dithiothreitol (DTT) or 2-mercaptoethanol is added so the bonds can re-form correctly during folding.
Refolding is therefore the second half of a recovery sequence: solubilize, then remove the denaturant. It sits downstream of expression and upstream of purification. If the protein is not forming inclusion bodies, or if expression conditions can be changed so that it stays soluble, refolding is not needed at all. For that earlier decision, see the guide to protein expression in E. coli, and for what comes after, protein purification strategy and chromatography sequence.
How refolding works in practice
Why aggregation competes with folding
Pauk and colleagues (2021) frame the problem kinetically: folding is a first-order reaction, while aggregation is order two or higher. In their words, “the reaction favors aggregation with higher protein concentration, thus decreasing the refolding yield.” Yamaguchi and Miyazaki make the same point from the experimental side: aggregation “is known to be a function of the protein concentration”, and it “predominantly occurs at medium concentrations of denaturant”, the range a protein passes through while the denaturant is being removed.
The practical consequence is that every common method is a strategy for limiting how much unfolded protein is present at once, or for passing through the middle of the denaturant range with as little aggregation as possible.
Dilution, dialysis and fed-batch compared
The table below keeps each figure attached to the paper it comes from, because the two sources describe different settings (a laboratory review and a process-development review) and their numbers should not be read as one range.
| Method | What it does | Reported strengths and limits | Source |
|---|---|---|---|
| One-step dialysis | Denaturant drops rapidly across a membrane | Simple, but the rapid drop raises the rate of misfolding and aggregation. Listed at recovery up to 40%, protein at 1-10 mg/mL, 1-2 days at 4 °C. | Yamaguchi and Miyazaki 2014 |
| Step-wise dialysis | Denaturant lowered in stages | Reduces the aggregation of one-step dialysis but is a time-consuming procedure, multi-day. | Yamaguchi and Miyazaki 2014 |
| Batch dilution | Denatured protein added once to denaturant-free buffer | Described there as 100- to 1000-fold dilution; requires a large volume of buffer; denaturant diffuses faster than the protein, so aggregation can occur as in one-step dialysis. Their table lists recovery of 80% or more for dilution, citing an earlier study. | Yamaguchi and Miyazaki 2014 |
| Batch and fed-batch dilution (process scale) | Solubilized inclusion bodies added in one step (batch) or slowly (fed-batch) | Typically 10- to 50-fold dilution of denaturant, final protein 1-100 µg/mL. Fed-batch gives the chain time to fold and is described as more effective; batch carries higher aggregation risk. Industrial dilution refolding is described as giving “non-competitive yields of 15-25% of the total protein”, and very slow dilution can approach about 100% but is uneconomical. | Pauk et al. 2021 |
| Microfluidic gradual removal | Denaturant removed gradually on a chip | Refolded citrate synthase more efficiently than dilution in the reported experiment. | Yamaguchi and Miyazaki 2014 |
Two readings follow from the table. First, the 80% or more recovery cited for lab-scale dilution and the 15-25% described for industrial dilution are not contradictory; they come from different proteins, scales and definitions of yield, so neither should be quoted as what a new protein will give. Second, the methods differ most in how much buffer and time they consume, which is why a method that works at the bench may not carry over to a larger batch.
Additives and why they are used
Yamaguchi and Miyazaki review additives that shift the balance away from aggregation:
- Arginine and arginine hydrochloride improve solubility and inhibit aggregation, attributed to moderate binding to proteins. Arginineamide and glycineamide are sometimes better.
- Proline has been proposed to trap folding intermediates.
- Glycerol, polyethylene glycol and sugars act as stabilizers, and are usually used together with an aggregation inhibitor such as arginine.
- Cyclodextrins are used as inhibitors, and as stripping agents in artificial-chaperone refolding.
Checking that refolding worked
Solubility after denaturant removal says the protein did not precipitate; it does not say the protein adopted its native structure. The monitoring methods reviewed by Pauk and colleagues are a useful checklist, each with its stated limits:
- Size-exclusion HPLC is described as the standard process for measuring aggregation. See size exclusion chromatography.
- Dynamic light scattering is described as primarily a qualitative method. See how to read a DLS report.
- Circular dichroism needs about 0.2-1.0 mg/mL of protein and is slow.
- Fluorescence can give feedback within seconds; RP-HPLC and SEC-HPLC give feedback in several minutes; inline ATR-FTIR can follow folding in real time; NMR is not suited to routine real-time monitoring.
For a bench check of the starting material and the final preparation, SDS-PAGE shows size and purity but cannot tell a native fold from a misfolded one, so an activity or binding assay for the specific protein remains the only direct evidence of function.
Where it matters for labs
Refolding decisions come up when a target protein is needed in quantity and expresses only as inclusion bodies. Points that bear on that decision:
- Protein concentration during denaturant removal is the lever most directly tied to yield, so the working concentration is a design choice, not a detail.
- Dilution methods trade yield for buffer volume; the lab review describes a large buffer requirement, and the process review describes very slow dilution as uneconomical.
- Dialysis is the simpler route but, per the lab review, has the lowest listed recovery and a multi-day step-wise variant.
- Whatever the method, the output needs an aggregation measurement and a function test before it is used downstream.
Common misunderstandings
“If it is soluble, it is refolded.”
Not necessarily. Soluble aggregates and misfolded monomers can both stay in solution. Aggregation is measured (for example by size-exclusion HPLC) and function is tested separately.
“There is one correct dilution factor.”
The two reviews cited here give different figures (100- to 1000-fold in one, 10- to 50-fold in the other) because they describe different settings. The right value is protein-specific.
“Higher protein concentration saves time with no cost.”
Because aggregation rises faster with concentration than folding does, concentrating the protein tends to lower the refolding yield.
FAQ
What is protein refolding?
It is the process of returning a denatured protein, commonly solubilized from inclusion bodies, to its native active structure by removing the denaturant under conditions that limit aggregation.
Which denaturants are used to dissolve inclusion bodies?
Yamaguchi and Miyazaki name urea, guanidine hydrochloride and ionic detergents such as N-lauroylsarcosine, with DTT or 2-mercaptoethanol added to reduce disulfide bonds.
Why does a refolded protein aggregate?
Aggregation is order two or higher while folding is first order, so a higher protein concentration favours aggregation and lowers refolding yield (Pauk et al. 2021).
How do I know the protein refolded correctly?
Measure aggregation (size-exclusion HPLC is described as the standard), consider structural methods such as circular dichroism, and test activity or binding for the specific protein.
Sources
- Yamaguchi H, Miyazaki M. Refolding Techniques for Recovering Biologically Active Recombinant Proteins from Inclusion Bodies. Biomolecules 2014;4(1):235-251. https://pmc.ncbi.nlm.nih.gov/articles/PMC4030991/. Checked 2026-10-29.
- Pauk JN et al. Advances in monitoring and control of refolding kinetics combining PAT and modeling. Applied Microbiology and Biotechnology 2021;105(6):2243-2260. doi:10.1007/s00253-021-11151-y. https://pmc.ncbi.nlm.nih.gov/articles/PMC7954745/. Checked 2026-10-29.
