E. coli is still the default prokaryotic factory for recombinant proteins: cheap, fast, genetically tractable, and scalable from 5 mL to 100 L. But "clone into pET, induce with IPTG, lyse, purify" is a cartoon. Real expression work is a negotiation between the host, the vector, the protein's folding energy landscape, and the downstream assay's tolerance for imperfection.
The central switch: T7 RNA polymerase / lac repression
pET vectors place your gene under T7 promoter; the host (BL21(DE3), Rosetta, Tuner, Lemo21) carries T7 RNA polymerase behind lacUV5. IPTG displaces LacI and turns on T7 pol. Problem: basal T7 transcription leaks even without IPTG, and toxic proteins kill the cells before induction. That is why auto-induction media, tighter repressors (T7 lysozyme in BL21(DE3)pLysS), or strains with tuned polymerase help.
The solubility dilemma
Eukaryotic proteins, multidomain enzymes, disulfide-rich proteins, and anything intrinsically disordered often fail to fold in the bacterial cytosol. Overexpression dumps thousands of copies per cell; chaperones saturate; the protein precipitates into inclusion bodies. Inclusion bodies are not garbage—they're dense, pure-ish, protease-protected aggregates. You can denature them in urea/guanidine, refold by dialysis/gradient, and sometimes recover active protein. But refolding is black magic: redox buffers, arginine as a folding osmolyte, slow dialysis, chaperone co-expression, tandem tags. Some proteins refold beautifully; others never do.
Expression conditions are tuning knobs
- Inducer: IPTG vs lactose vs arabinose; IPTG is strong but toxic at high dose.
- Induction OD: too early = metabolic burden; too late = less biomass.
- Temperature: 37 °C maximizes yield but favors aggregates; 16–25 °C slows translation, helps folding, often improves solubility.
- Induction duration: 3 h vs overnight changes yield vs degradation.
- Media: LB is rich and inconsistent; TB/M9 auto-induction gives reproducible biomass and often better soluble fractions.
- Codon usage: human proteins in E. coli may stall on Arg/Leu/Ile codons; Rosetta(DE3) supplies rare tRNAs, but rare-tRNA saturation can create misincorporation under strong induction.
Fusion tags are strategic
His₆ is for purification, not solubility. MBP (maltose-binding protein) and NusA are solubility enhancers. GST helps pull-downs and sometimes solubility. SUMO is small and often proteolytically removable with native fold preserved. Thioredoxin (TrxA) helps disulfide handling in the cytosol when paired with origami strains (trxB/gor mutants). Periplasmic targeting (PelB/OmpA signal peptides) moves the protein toward an oxidizing compartment better for disulfides—but yields drop and processing is messy.
Strain selection is underrated
BL21(DE3) lacks Lon and OmpT proteases, stabilizing recombinant proteins. BL21(DE3)pLysS adds T7 lysozyme to silence leakiness. Rosetta adds rare tRNAs. SHuffle strains engineer disulfide isomerization in the cytoplasm. C41/C43 tolerate toxic membrane proteins. ClearColi reduces endotoxin for mammalian/cell-based assays. You do not "pick a strain"; you match a strain to a failure mode.
Detection before scale-up is mandatory
Small 2 mL cultures: induce side-by-side at 16/25/37 °C, lyse by sonication/French press, split soluble vs pellet by centrifugation, run SDS-PAGE, and stain with Coomassie. If the protein is only in the pellet, do not scale up. Change tag, strain, temperature, inducer concentration, or truncate the problematic domain. A membrane protein needs detergents; a secreted protein needs periplasmic prep; a toxic protein needs tightly repressed systems.
What a mature expression campaign answers
Where is the protein (soluble / insoluble / supernatant / inclusion body / periplasm)? Is the band the right size (fusion + linker + target)? Is the host burden acceptable (growth curve not collapsing)? Is the product full-length (degradation ladder on gel)? Is it antigenically/functionally useful even if not "native"? Many people want native eukaryotic protein; E. coli often gives a workable reagent, not a perfect mimic. Knowing the difference saves years.
Yield is not success
The bigger lesson: yield is not success. A 200 mg/L inclusion-body precipitate you cannot refold is worse than 5 mg/L soluble active protein. Expression work rewards patience: one variable at a time, solubility screens, tag swaps, protease-site planning, and small-scale triage before any 1 L flask. The bacteria do not care about your Western blot. They care about thermodynamics, codon pressure, and whether you gave them a promoter they can survive.
References
- Rosano GL, Ceccarelli EA. Recombinant protein expression in Escherichia coli: advances and challenges. Front Microbiol. 2014;5:172.
- Terpe K. Overview of bacterial expression systems for heterologous protein production. Appl Microbiol Biotechnol. 2006;72(2):211–222.
- Esposito D, Chatterjee DK. Enhancement of soluble protein expression through the use of fusion tags. Curr Opin Biotechnol. 2006;17(4):353–358.
- Geertsma ER, Poolman B. Production of membrane proteins in Escherichia coli and Lactococcus lactis. Methods Mol Biol. 2008;474:217–228.
- Studier FW. Protein production by auto-induction in high-density shaking cultures. Protein Expr Purif. 2005;41(1):207–234.
- Baneyx F, Mujacic M. Recombinant protein folding and misfolding in Escherichia coli. Nat Biotechnol. 2004;22(11):1399–1408.
- Lobstein J, Emrich CA, Jeans C, et al. SHuffle, a novel Escherichia coli protein expression strain capable of correctly folding disulfide bonded proteins in its cytoplasm. Microb Cell Fact. 2012;11:56.