Inclusion Bodies and Refolding: Turning Insoluble Protein Into the Right Shape
When a recombinant protein is expressed in the cytoplasm of Escherichia coli, it often does not stay dissolved. It aggregates into dense particles called inclusion bodies (IBs) [6]. For a long time these aggregates were treated as inactive, unfolded waste material produced by heterologous overexpression [5]. They are not waste. Compared with other expression strategies, inclusion body formation allows high product titres and makes it possible to express proteins that are toxic to the host [6].
The cost is that the protein comes out of the cell in the wrong shape. To get a soluble, correctly folded molecule you have to break the cells, collect the particles, dissolve them and let the chain find its native structure again. That sequence — solubilisation and refolding — is where yield is won or lost [5].
Why a protein ends up insoluble
Aggregation propensity varies by protein and is difficult to predict [3]. Single chain variable fragment (scFv) antibody production in E. coli remains challenging precisely because the recombinant protein tends to aggregate into inclusion bodies [1].
Sometimes insolubility is the goal rather than the problem. Heterologous protein production in E. coli often runs into proteolytic degradation, complex purification and toxicity towards the host, and producing the protein in an insoluble form inside inclusion bodies can relieve all three [3]. Because the tendency to form IBs is variable, one strategy is to fuse the target to an aggregation-prone polypeptide — an IB tag [3].
One such tag came out of a screen for signal sequences meant to do the opposite. The 39 amino acid signal sequence of E. coli TorA (ssTorA) did not direct the human proteins hEGF, Pla2 and IL-3 to the periplasm; it drove high-level expression of them in inclusion bodies instead [3]. ssTorA also forced aggregation of TrxA and MBP, two endogenous E. coli proteins that are normally highly soluble [3]. It worked fused at the C-terminus as well as the N-terminus, and in both K-12 and B strains [3]. Fusing several ssTorA copies in tandem had an additive effect, provoking almost complete aggregation of TrxA and MBP [3]. The tag was used to produce the intrinsically unstable hEGF and the toxic fusion partner SymE [3].
Inclusion bodies have quality attributes you can set upstream
Downstream of the fermenter — homogenisation of the cells, centrifugation, solubilisation of the IBs — process performance is prone to variation and often gives low extraction yields [6][8]. The working hypothesis in recent work is that this variation comes from the inclusion bodies themselves: they have measurable quality attributes, and those attributes can be controlled by upstream process conditions [6][8].
A study on two industrially relevant proteins tracked three attributes over time — titre, IB bead size and purity — and found all three depended on pH and temperature during expression [8]. A data-driven model from that work indicated that low temperatures and low pH favoured expression of both tested proteins [8]. With pH and temperature set, the authors then varied a physiological parameter: the specific substrate uptake rate for glucose (qs,Glu) [8]. Higher specific glucose uptake rates had a strong influence on the quality attributes and drastically increased titre and purity at early time points [8]. The conclusion drawn was that IB quality attributes can be modulated time-dependently in upstream processing to make downstream processing more robust [8].
Carbon source and induction regime matter too. Harsh induction with isopropyl β-d-1 thiogalactopyranoside is associated with stress reactions known as metabolic or product burden, which mainly show up as decreased growth rates and cell lysis at longer induction times [2]. Softer, tunable induction with lactose reduces that stress level [2]. Using glucose as the energy source alongside lactose as inducer causes catabolite repression of lactose uptake and, as a consequence, reduced product titre [2]. Glycerol shows no sign of repression when used together with lactose, and in the reported experiments it had beneficial effects both on cell viability and on recombinant protein productivity compared with glucose [2].
What the cell does to its own aggregates
Inclusion bodies are not inert inside the cell. During production of yeast alpha-glucosidase in E. coli, IBs began to disaggregate once protein synthesis stopped, in a process dependent on the chaperones ClpB and DnaK, and the released protein was then either degraded or reactivated [4]. The small heat-shock proteins IbpA and IbpB slowed disaggregation and degradation at higher temperatures but barely changed disaggregation kinetics at 15 °C [4]. Deleting ibpAB — harmless under physiological conditions — impaired culture growth during alpha-glucosidase production, and at higher temperatures intensified the accumulation of stress proteins including DnaK, ClpB, enolase and PNP [4]. Overexpressing ibpAB suppressed the heat-shock response and, when done alongside production at 42 °C, increased the yield of alpha-glucosidase activity during reactivation [4]. Without ibpAB, inclusion body removal was faster but the cells suffered more stress and growth impairment [4].
Solubilisation: harsh, mild, or chemical
A direct comparison on scFv shows how much the choice depends on the construct. Mild solubilisation recovered soluble, functional scFv from non-classical IBs with no fusion tag and no in vitro refolding step at all [1]. Its recovery yield beat the denaturation–refolding method in every case except when the scFv carried a maltose binding protein (MBP) tag [1]. MBP not only influenced cytoplasmic expression but greatly improved in vitro refolding yield [1]. Without the MBP tag or a refolding enhancer, protein obtained by mild solubilisation was better structured and functionally more active than protein obtained by in vitro refolding [1]. The authors' summary: MBP-tagged scFv gave better refolding and solubility yields, but on cost, time and the advantage of being tag-free, mild solubilisation is worth considering [1].
There is also a route that skips refolding entirely. Catalytically active inclusion bodies (CatIBs) are produced on demand by fusing short peptide tags or aggregation-inducing protein domains to a target protein, and where the target is an enzyme the resulting particles act as carrier-free enzyme immobilisates [5]. CatIB formation efficiency depends on construct design and expression conditions [5].
What to take from this
Inclusion bodies are a deliberate expression strategy, not a failure mode [6]. If you are specifying or auditing such a process, three things are worth asking about. First, whether IB quality attributes — titre, size, purity — are measured at all, since analytical methods for them exist and upstream parameters demonstrably move them [6][8]. Second, what pH, temperature and substrate feeding profile were chosen, and on what evidence, because these set the material the downstream train has to handle [8]. Third, whether denaturation–refolding is actually necessary: for some constructs mild solubilisation gave both a higher yield and better-structured protein [1], while a fusion tag can reverse that ranking [1]. The honest conclusion from the review literature is that upstream process parameters can be used to trigger different quality attributes, including activity, but are not yet exploited to a satisfying extent [6].
References
- Sarker A, Rathore AS, Gupta RD. Evaluation of scFv protein recovery from E. coli by in vitro refolding and mild solubilization process — Microbial cell factories, 2019-01-14.
- Kopp J, Slouka C, Ulonska S, Kager J, Fricke J, Spadiut O, Herwig C. Impact of Glycerol as Carbon Source onto Specific Sugar and Inducer Uptake Rates and Inclusion Body Productivity in E. coli BL21(DE3) — Bioengineering (Basel, Switzerland), 2017-12-21.
- Jong WS, Vikström D, Houben D, van den Berg van Saparoea HB, de Gier JW, Luirink J. Application of an E. coli signal sequence as a versatile inclusion body tag — Microbial cell factories, 2017-03-21.
- Lethanh H, Neubauer P, Hoffmann F. The small heat-shock proteins IbpA and IbpB reduce the stress load of recombinant Escherichia coli and delay degradation of inclusion bodies — Microbial cell factories, 2005-02-11.
- Jäger VD, Lamm R, Küsters K, Ölçücü G, Oldiges M, Jaeger KE, Büchs J, Krauss U. Catalytically-active inclusion bodies for biotechnology-general concepts, optimization, and application — Applied microbiology and biotechnology, 2020-07-10.
- Slouka C, Kopp J, Spadiut O, Herwig C. Perspectives of inclusion bodies for bio-based products: curse or blessing? — Applied microbiology and biotechnology, 2018-12-19.
- Potempa LA, Yao ZY, Ji SR, Filep JG, Wu Y. Solubilization and purification of recombinant modified C-reactive protein from inclusion bodies using reversible anhydride modification — Biophysics reports, 2015-07-18.
- Slouka C, Kopp J, Hutwimmer S, Strahammer M, Strohmer D, Eitenberger E, Schwaighofer A, Herwig C. Custom made inclusion bodies: impact of classical process parameters and physiological parameters on inclusion body quality attributes — Microbial cell factories, 2018-09-20.

