to BioTechniques free email alert service to receive content updates.
Protein purification: Inclusion bodies
Kristie Nybo, Ph.D.
BioTechniques, Vol. 54, No. 2, February 2013, pp. 75–76
Full Text (PDF)

This month's question from the Molecular Biology Forums (online at comes from the “Protein Methods” section. Entries have been edited for concision and clarity. Mentions of specific products and manufacturers have been retained from the original posts, but do not represent endorsements by, or the opinions of, BioTechniques.

Molecular Biology Techniques Q&A

Is it possible to recover correctly folded proteins from inclusion bodies? (Thread 19840)

Q I work with a protein expressed in inclusion bodies in E. coli and recently heard that proteins with native secondary structure can occasionally be recovered from inclusion bodies using low concentrations of urea. Since protein refolding has been problematic in this case, I would like to try this approach, but I haven't been able to locate a reference describing the method. Can someone recommend a good procedure?

A The majority of the expressed protein is in the inclusion bodies, so most researchers prefer refolding over purifying a small amount of native protein. There are many approaches for refolding your protein, so you should start there.

Q I have spent a significant amount of time unsuccessfully trying to refold this protein, and I now would like to explore another avenue. This particular protein is very difficult to refold due to a lack of glycosylation, five disulfide bonds, and the fact that it is a dimer. Every expression system I have tried results in inclusion bodies. I am trying a yeast expression system now, but the progress is slow. While I am waiting for this, I would like to try recovering the protein from the inclusion bodies.

A There are several ways to coax a protein into the soluble fraction. You probably have already tried temperature, varying IPTG concentrations, etc. But there are other approaches that you might have overlooked such as using different strains of your cells, clipping or adding amino acids to either end of the protein, using another vector with a different promoter, or using a different tag. You might also try using a small percentage of detergent to remove it from the inclusion bodies.

A What have you tried so far to refold the protein? During the last six months, I have expressed twelve proteins in E.coli using the SUMO expression system from Invitrogen. My standard approach has been to isolate the inclusion bodies, solubilize in 8 M urea or 6 M GuHCl, purify with Ni-NTA resin under denaturing conditions, refold against 3 M urea, 20 mM Tris, 1 mM DTT, 150 mM NaCl, pH 8.0 using a normal dialysis chamber, and then remove the SUMO tag with a SUMO protease.

Six of these proteins were able to refold using 2 M urea, but the others required 3 M. While I wasn't too concerned about the protein of interest refolding since I'm using it for ELISA,the SUMO protein folded, which enabled me to cleave it from my protein of interest. I think that other proteins should refold under these conditions as well.

A I have used sarkosyl to purify or increase the yields of proteins in inclusion bodies, although in one case the protein I worked with wasn't active following sarkosyl purification. A good method can be found in Analytical Biochemistry (1993) 210: 179-187.

Q Previously, I tried using different constructs, including mutations or different N- and C-terminal tags, different vectors, and different E.coli host strains, all of which resulted in inclusion bodies, including the construct with an N-terminal signal peptide that directed the protein to the periplasm. For refolding, I tried solubilizing inclusion bodies with 6M GdmCl, IMAC purification, reduction of cysteines with DTT, gel filtration purification, refolding by rapid dilution, and SDS-PAGE, native-PAGE, and activity assays to monitor refolding. Even after extensive optimization, I obtained yields of less than 5%.

A I recently expressed 16 mutant proteins, all of which were in the insoluble fraction. I used the pQE30 vector from Qiagen, so the yield was high, but initially, I had trouble refolding the proteins because they have 16 disulfide bonds and are 28KDa. I was finally successful with the Refolding Kit from Novagen, which is basically a 96-well plate with different buffers for refolding in each well. I found two combinations of buffers that refolded my protein of interest.

For your protein, can you do circular dichroism (CD)? Is the crystal structure known? You might also try running an anion or cation exchange column. If you have a mixture of folded and misfolded proteins, then you will have multiple peaks rather than a single sharp peak from a solution containing 100% correctly folded protein.

Q Did you check the disulfides by MSMS or ABD-F staining?

I check for refolding using an activity assay. I don't have access to a CD instrument and the crystal structure is unknown. I tried several methods, including chromatography, precipitation, and prep-native-PAGE, for separating the correctly folded protein from the misfolded forms. Ion exchange attempts did not work with any column tested or using a wide range of pH. Gel filtration revealed multiple species including monomer, dimer, and multiple aggregation states. Unfortunately, gel filtration did not give satisfactory separation between the monomer and dimer. I haven't tried the refolding kit, but I have made similar 96-well plate screens and used this to optimize the buffers.

A To check the refolding of my protein, I use different techniques including reactivity with monoclonal antibodies, CD, anion exchange and affinity chromatography (heparin binding), functional assays comparing with Pichia expressed protein, and MALDI using mercury. If there is a free cysteine, it will react with the mercury, causing an almost 200 mass unit increase. It is difficult to figure out, but some combination of all these approaches should give you insights.

A Are you using one of the origami-based E. coli strains from Novagen? I had to purify a 13kDa protein that had 16 cysteines and was able to do it only when I used Rosettagami cells.

A In BioTechniques 2008 May; 44(6):787-96, Thapa et al. describe expression of foreign proteins in soluble form by fusion to E. coli thermostable protein (Trigger Factor) through a ubiquitin bridge. The foreign protein can be excised with the de-ubiquitylating enzyme Usp2-cc.

A You usually get the highest recoveries using denaturing conditions, but protein misfolding will reduce those yields. It is best to isolate proteins under native conditions when the success of an experiment is dependent on proper refolding. We have successfully extracted proteins under native conditions from tissues with Covaris Adaptive Focused Acoustics (AFA), which is a method for isothermal sonication.

A Since you're using E. coli, you might consider a maltose binding protein tag to increase the solubility of your protein. The fact that the proteins are going into inclusion bodies suggests that your induction level is too high for the folding machinery in E. coli. You should troubleshoot and optimize for the best IPTG concentration. Instead of gel filtration, you might try dialyzing slowly overnight using a peristaltic pump connected to your IMAC column to decrease aggregation.

A Does your protein have a known binding partner? Sometimes insoluble proteins become soluble if co-expressed with their binding partners.