A practical guide to E. coli inclusion bodies, protein refolding strategies, and their role in recombinant protein manufacturing.
Author: Agnes Zimmer, PhD, Senior Scientist, NorthX Biologics | August 2026
Abstract
E. coli inclusion bodies are commonly viewed as unwanted aggregates that form when recombinant protein expression exceeds the host cell’s folding capacity. However, their role in biomanufacturing is more nuanced. Although inclusion body production introduces additional operations (including isolation, solubilisation, refolding and structural characterisation), it can also deliver high titres, facilitate early product enrichment and protect protease-sensitive or toxic proteins. This article explores the molecular and process-related factors that drive inclusion body formation and reviews modern strategies for recovering biologically active proteins through milder solubilisation, controlled refolding and redox management. It also examines commercially successful products manufactured using this approach, including insulin, filgrastim, interferons, BMP-2, aldesleukin and moxetumomab pasudotox. For developers evaluating recombinant protein production, the central question is whether the potential manufacturing advantages outweigh the downstream complexity for a specific molecule and process.
E. coli inclusion bodies in recombinant protein production
Approximately 30% of approved recombinant therapeutic proteins are produced in E. coli, exploiting its rapid growth, genetic tractability, cost-efficient cultivation, and scalability [1]. To date, more than 250 recombinant therapeutic proteins produced in E. coli have received FDA and/or EMA approval [2]. Despite the advantages, E. coli has a well-recognized limitation: the frequent formation of inclusion bodies (IBs), which are often described as dense, insoluble aggregates of misfolded or partially folded recombinant proteins that accumulate in the bacterial cytoplasm, periplasm, or nucleus [3]. Under certain conditions, recombinant proteins expressed in E. coli are frequently reported to accumulate as IBs, particularly when expression levels exceed the folding capacity of the host cell [3]. IB formation can also occur in yeast, insect, and mammalian expression systems [4]. For protein classes with complex tertiary structures, multiple disulfide bonds, or eukaryotic post-translational modifications, IB formation in E. coli is especially common because bacterial cytoplasmic expression does not naturally support all folding and modification requirements [5]. Thus, for many recombinant proteins, inclusion body formation is the rule rather than the exception.
What drives the formation of inclusion bodies?
Each protein has a specific functional three-dimensional structure encoded by its primary sequence [6], and folding is thermodynamically driven toward the most energetically favorable conformation [7]. However, within the energy landscape of protein folding, native folding and aggregation are competing pathways. Intramolecular folding toward the native state can therefore compete with misfolding and intermolecular association, ultimately leading to the formation of inclusion bodies [3].
During synthesis and folding, nascent polypeptide chains pass through partially folded intermediate states in which hydrophobic residues and flexible regions may be exposed. If these regions are not rapidly buried within the native fold or transiently shielded by chaperones, they can promote intermolecular interactions and become incorporated into growing aggregates [3,8]. Inclusion body formation is therefore often favored when protein synthesis exceeds the capacity of the cellular folding machinery. Additionally, inefficient or incorrect disulfide bond formation due to the reducing environment of the bacterial cytoplasm, and the absence of eukaryotic post-translational modifications can contribute to misfolding and aggregation [5]. While such protein-intrinsic factors may be difficult to modify, process-controllable parameters—including inducer concentration and induction strategy, host strain, promoter strength, growth medium, pH, and temperature—can be adjusted to reduce inclusion body formation [9].
Challenges and advantages of inclusion bodies
The foe perspective on inclusion bodies
The main drawback of inclusion bodies is that the target protein accumulates in a non-native, misfolded, and largely inactive form [3]. Unlike soluble expression, where the protein can proceed directly to purification, IB-derived proteins must first be unfolded and then refolded into their functional state. This adds several unit operations to the workflow, including IB isolation and washing, solubilization in strong denaturants, refolding, aggregate removal, and often also disulfide bond formation [10]. Each step requires case-specific optimization, increasing development time, costs, and material consumption. Refolding is particularly sensitive to protein concentration, pH, temperature, additive selection, and the rate of denaturant removal, often requiring extensive screening without any guarantee of success or acceptable yield [11]. Formation of correct disulfide bonds can further complicate the process, especially for proteins containing multiple disulfide bonds and/or free cysteines [5]. Refolding commonly requires substantial dilution of the solubilized protein, which creates large process volumes during scale-up [11].
After refolding, the protein must also be characterized structurally and functionally to confirm correct folding and distinguish active product from misfolded but soluble species [10]. This adds further analytical effort to an already more demanding workflow, increased costs, complexity, and scale-up risk.
Turning the tide, changing perspectives
However, these challenges are not insurmountable. Over the last decades, inclusion body processing has evolved from an empirical rescue strategy into a controlled and optimizable downstream workflow. Solubilization, refolding and disulfide bond formation remain critical bottlenecks, but each can be addressed through rational process design.
For solubilization, the traditional approach relies on strong chaotropic agents such as urea and guanidine hydrochloride to fully unfold aggregated proteins [4]. While effective, harsh solubilization erases even residual native-like structure. Therefore, attention has been given to milder solubilization strategies, including reduced denaturant concentrations and alternative solubilizing agents that aim to preserve refolding-competent structure [12]. The goal is not simply to dissolve the inclusion body material, but to recover the target protein in a state that is as refolding-competent as possible. This is particularly relevant for non-classical inclusion bodies, where part of the protein may already retain native-like structure [13].
Refolding is mainly competing with aggregation, which can be mitigated by maintaining low protein concentrations, carefully controlling the rate of denaturant removal, and screening additives that stabilize productive folding intermediates [11]. Common solutions include dilution refolding, diafiltration, fed-batch refolding, and on-column refolding, while using additives that reduce intermolecular interactions [11]. As refolding is highly protein-specific, systematic screening is needed, and design-of-experiments approaches can help to find robust refolding conditions.
During refolding of proteins that contain disulfide bonds, correct oxidative conditions are needed. This is commonly achieved by using defined redox buffers to balance disulfide formation and reshuffling of non-native pairings [10]. The optimal redox potential depends on the number and accessibility of cysteines, the folding pathway, and the tendency of the protein to form intermolecular disulfide-linked aggregates. In some cases, protein engineering can simplify the challenge by removing non-essential free cysteines, introducing stabilizing mutations, or modifying expression constructs to favor correct folding [14]. Host strain engineering, periplasmic expression, or co-expression of foldases and disulfide isomerases can also reduce the downstream refolding burden [5].
The friend perspective on inclusion bodies
Despite the challenges and extended workflow, inclusion bodies can provide important compensating advantages. For an increasing number of proteins, they are therefore not merely an accepted production route but a deliberately selected manufacturing strategy [15,16].
Because inclusion bodies sequester the target protein in an inert form, they reduce the cellular burden associated with folding and allow resources to be directed toward protein production. As a result, target proteins can accumulate to 30–50% of total cell protein [17], supporting high space-time yields, a key economic driver. Their high density also enables straightforward separation from soluble host cell proteins by low-speed centrifugation. Following detergent or urea washes, the target protein typically represents approximately 60% of the washed-pellet protein [18], providing partial purification before downstream processing. For individual proteins, IB purity of up to 95% has been reported [19].
For some proteins, the IB pathway can therefore be an attractive option for high-yield manufacturing. IB formation may protect protease-sensitive proteins that would otherwise be degraded as soluble cytoplasmic species [20]. Similarly, proteins that are toxic to the host can be rendered inactive through aggregation into IBs, enabling higher biomass formation and improved product yields [15,17].
Case studies: commercially successful refolded proteins
The feasibility of inclusion body-based manufacturing is best illustrated by the number of approved recombinant proteins that are produced through this route. Despite the need for solubilization and refolding, several IB-derived proteins have been successfully developed into high-volume and clinically important biopharmaceuticals [16]. The following examples highlight both the challenges and the industrial relevance of refolded proteins.
Insulin
Insulin, including products such as Humulin® (Eli Lilly), was the first approved recombinant biopharmaceutical in 1982 and remains one of the most prominent examples of large-scale E. coli inclusion body-based manufacturing [1].
- Refolding challenge: Insulin contains three disulfide bonds, and incorrect pairing can generate inactive or potentially immunogenic variants [11].
- Scale: Annual demand reaches thousands of kilograms worldwide.
- Relevance: This example demonstrates that IB-based processes can support some of the world’s most high-volume biopharmaceutical products.
Granulocyte colony-stimulating factor (G-CSF)
Filgrastim (Neupogen®, Amgen) is a non-glycosylated recombinant human Granulocyte colony-stimulating factor (G-CSF) produced in E. coli and used for the treatment of chemotherapy-induced neutropenia. Several biosimilars are available, and G-CSF is widely described as product commonly recovered from inclusion bodies [13].
- Refolding challenge: The presence of two disulfide bonds and one free cysteine can promote disulfide scrambling and formation of inactive variants [11].
- Recent innovations: G-CSF has been shown to accumulate into non-classical inclusion bodies (ncIBs) that retain native-like secondary structure and can be solubilized under mild, non-denaturing conditions, reducing reliance on conventional harsh solubilization and refolding steps [13].
Interferons
Like insulin, interferons were among the earliest recombinant proteins commercially produced from E. coli inclusion bodies [1].
Interferon alpha-2b (Intron A®, Schering-Plough), used to treat hepatitis B/C and certain cancers, contains two disulfide bonds. Interferon beta-1b (Betaseron®, Bayer; Extavia®, Novartis), used for the treatment of multiple sclerosis, was engineered with a serine-for-cysteine substitution to reduce incorrect disulfide bond formation during refolding [21,14]. Interferon gamma-1b (Actimmune®), a homodimeric protein, is used to reduce serious infections in chronic granulomatous disease and to delay progression of severe malignant osteopetrosis.
- Key learning: Protein engineering, such as removal of free cysteines, can improve refolding efficiency and product quality.
Bone Morphogenetic Protein-2 (BMP-2)
Bone Morphogenetic Protein-2 (BMP-2), (Infuse® Bone Graft, Medtronic) is a homodimeric growth factor used in orthopedic surgery to promote bone regeneration and represents a particularly demanding refolding case.
- Refolding challenge: Its cystine-knot motif contains multiple intrachain disulfide bonds and an interchain disulfide bond linking the homodimer, requiring tightly controlled redox conditions [22].
- Key learning: The clinical success of Infuse® shows that even highly complex refolding processes can be translated to commercial-scale manufacturing.
Interleukin-2 (IL-2)
Aldesleukin (Proleukin®, Prometheus/Clinigen) is a recombinant human Interleukin-2 (IL-2) product produced in a genetically engineered E. coli strain and used for the treatment of metastatic renal cell carcinoma and melanoma [23,24].
- Refolding: Compared with multi-disulfide proteins, refolding of IL-2 is relatively straightforward. It contains one disulfide bridge, and substitution of a third cysteine with serine helps to prevent intermolecular disulfide formation and aggregation [14].
Immunotoxin
Moxetumomab pasudotox (Lumoxiti®, AstraZeneca/MedImmune) is an anti-CD22 recombinant immunotoxin approved for hairy cell leukemia and represents one of the most complex proteins commercially refolded from inclusion bodies [25].
- Structure and refolding: The molecule consists of VH and VL domains from an anti-CD22 antibody linked by a disulfide bond and fused to a truncated Pseudomonas exotoxin (PE38). Correct folding requires both chains to fold properly and associate, and commercial development included scalable fed-batch dilution refolding of the two-chain immunotoxin [25].
Inclusion bodies: not inherently good or bad, but process-dependent
Inclusion bodies are neither wholly friend nor wholly foe — they are what the process developer makes of them. For a small, soluble, disulfide-free protein, they remain an obstacle best avoided. But for a high-value, protease-sensitive, toxic, or high-titer target, the very properties that once earned them a poor reputation — density, robustness, and enrichment of the product — become decisive advantages. The deciding factor is, and always has been, process understanding: the better we characterize and control solubilization and refolding, the more firmly the balance tips from foe toward friend.
References
[1] Huang, C.-J., Lin, H., & Yang, X. (2012). Industrial production of recombinant therapeutics in Escherichia coli and its recent advancements. Journal of Industrial Microbiology & Biotechnology, 39, 383–399.
[2] Niazi, S. K., & Magoola, M. (2023). Advances in Escherichia coli-based therapeutic protein expression: mammalian conversion, continuous manufacturing, and cell-free production. Biologics, 3(4), 380–401.
[3] Bhatwa, A., Wang, W., Hassan, Y. I., Abraham, N., Li, X.-Z., & Zhou, T. (2021). Challenges associated with the formation of recombinant protein inclusion bodies in Escherichia coli and strategies to address them for industrial applications. Frontiers in Bioengineering and Biotechnology, 9, 630551.
[4] Palmer, I., & Wingfield, P. T. (2004). Preparation and extraction of insoluble inclusion-body proteins from Escherichia coli. Current Protocols in Protein Science, Chapter 6, Unit 6.3.
[5] de Marco, A. (2009). Strategies for successful recombinant expression of disulfide bond-dependent proteins in Escherichia coli. Microbial Cell Factories, 8, 26.
[6] Anfinsen, C. B. (1973). Principles that govern the folding of protein chains. Science, 181(4096), 223–230.
[7] Jaenicke, R. (1996). Protein folding and association: In vitro studies for self-organization and targeting in the cell. Current Topics in Cellular Regulation, 34, 209–314.
[8] Mitraki, A., & King, J. (1989). Protein folding intermediates and inclusion body formation. Bio/Technology, 7, 690–697.
[9] Singh, A., Upadhyay, V., Upadhyay, A. K., Singh, S. M., & Panda, A. K. (2015). Protein recovery from inclusion bodies of Escherichia coli using mild solubilization process. Microbial Cell Factories, 14, 41.
[10] Tsumoto, K., Ejima, D., Kumagai, I., & Arakawa, T. (2003). Practical considerations in refolding proteins from inclusion bodies. Protein Expression and Purification, 28(1), 1–8.
[11] Yamaguchi, H., & Miyazaki, M. (2014). Refolding techniques for recovering biologically active recombinant proteins from inclusion bodies. Biomolecules, 4(1), 235–251.
[12] Klausser, R., Kopp, J., Brichtova, E. P., Gisperg, F., Elshazly, M., & Spadiut, O. (2023). State-of-the-art and novel approaches to mild solubilization of inclusion bodies. Frontiers in Bioengineering and Biotechnology, 11, 1249196.
[13] Vo, T. M. T., Tran, L. T., & Thao, D. T. P. (2021). Production of recombinant human G-CSF from non-classical inclusion bodies in Escherichia coli. Brazilian Journal of Microbiology, 52, 331–338.
[14] Wang, A., Lu, S. D., & Mark, D. F. (1984). Site-specific mutagenesis of the human interleukin-2 gene: structure-function analysis of the cysteine residues. Science, 224(4656), 1431–1433.
[15] Ramón, A., Señorale-Pose, M., & Marín, M. (2014). Inclusion bodies: not that bad. Frontiers in Microbiology, 5, 56.
[16] Kopp, J., & Spadiut, O. (2023). Inclusion bodies: from unwanted aggregates to functional biomaterials and process intermediates. Applied Microbiology and Biotechnology, 107, 5431–5446.
[17] Rinas, U., Garcia-Fruitós, E., Corchero, J. L., Vázquez, E., Seras-Franzoso, J., & Villaverde, A. (2017). Bacterial inclusion bodies: Discovering their better half. Trends in Biotechnology, 35(2), 93–106.
[18] Palmer, I., & Wingfield, P. T. (2012). Preparation and extraction of insoluble inclusion-body proteins from Escherichia coli. Current Protocols in Protein Science, Chapter 6, Unit 6.3.
[19] Singh, S. M., & Panda, A. K. (2005). Solubilization and refolding of bacterial inclusion body proteins. Journal of Bioscience and Bioengineering, 99(4), 303–310.
[20] Komolov, A. S., Sannikova, E. P., Gorbunov, A. A., Gubaidullin, I. I., Plokhikh, K. S., Konstantinova, G. E., Bulushova, N. V., Kuchin, S. V., & Kozlov, D. G. (2024). Synthesis of biologically active proteins as L6KD-SUMO fusions forming inclusion bodies in Escherichia coli. Biotechnology and Bioengineering, 121(2), 535–550.
[21] Nekoufar, S., Fazeli, A., & Fazeli, M. R. (2020). Solubilization of human interferon β-1b inclusion body proteins by organic solvents. Advanced Pharmaceutical Bulletin, 10(2), 233–238.
[22] Rane, A. M., Jonnalagadda, S., & Li, Z. (2013). On-column refolding of bone morphogenetic protein-2 using cation exchange resin. Protein Expression and Purification, 90(2), 135–140.
[23] U.S. Food and Drug Administration. (2023). Proleukin® (aldesleukin) for injection, prescribing information. FDA.
[24] U.S. Food and Drug Administration. (2012). Proleukin® (aldesleukin) injection label. FDA.
[25] Linke, T., Aspelund, M. T., Thompson, C., Xi, G., Fulton, A., Wendeler, M., Pabst, T. M., Wang, X., Wang, W. K., Ram, K., & Hunter, A. K. (2014). Development and scale-up of a commercial fed batch refolding process for an anti-CD22 two-chain immunotoxin. Biotechnology Progress, 30(6), 1380–1389.