Start by optimizing five key variables: host strain, promoter system, induction conditions, growth temperature, and codon usage. Choose a strain matched to your protein's requirements -- BL21(DE3) for non-toxic cytoplasmic proteins, C41/C43(DE3) for membrane or toxic proteins, SHuffle for disulfide-bond-containing proteins, or Origami for multiple disulfide bonds. Use codon-optimized genes for heterologous expression, but avoid over-optimization that creates mRNA secondary structures. Systematically test induction OD (0.4-0.8), IPTG concentration (0.1-1.0 mM), post-induction temperature (16-37°C), and duration (4 h to overnight). Our optimizer tool automates these recommendations based on your protein's molecular weight, origin, and structural features.

For most T7-based expression systems, 0.1-0.5 mM IPTG is optimal, though many researchers default to 1 mM unnecessarily. IPTG is not metabolized by E. coli, so it persists throughout the culture, and lower concentrations often yield more soluble protein. For soluble expression of difficult proteins, start with 0.1 mM IPTG at 16-18°C overnight. For high-yield expression of robust proteins, 0.4-1.0 mM at 37°C for 3-4 hours works well. Auto-induction media (Studier's ZYM-5052) is an excellent alternative that eliminates IPTG entirely, using lactose for gradual induction after glucose depletion. This typically gives higher cell densities and more total protein per culture volume. Always run a small-scale IPTG titration (0.05, 0.1, 0.25, 0.5, 1.0 mM) before committing to large-scale expression.

Lower the expression rate to give proteins more time to fold correctly. The most effective single intervention is reducing post-induction temperature to 16-20°C, which slows translation and promotes proper folding. Additionally, reduce IPTG concentration (0.05-0.2 mM) to decrease transcription rate, or switch to a weaker promoter (araBAD, tac) instead of T7. Co-expression of molecular chaperones (GroEL/GroES, DnaK/DnaJ/GrpE, trigger factor) can rescue misfolded proteins. Fusion tags such as MBP, SUMO, or Trx dramatically improve solubility for many proteins. Using osmotic stress (0.5 M sorbitol + 2.5 mM betaine) or adding ethanol (2-3% v/v) can also induce chaperone expression. If inclusion bodies persist, they can be solubilized and refolded, though this adds significant process complexity.

There is no single best strain -- the optimal choice depends on your protein's properties. BL21(DE3) is the default workhorse for cytoplasmic expression of non-toxic proteins, as it lacks the Lon and OmpT proteases. For toxic proteins, C41(DE3) and C43(DE3) carry mutations that reduce basal T7 expression. For proteins requiring disulfide bonds, SHuffle T7 (cytoplasmic DsbC, trxB/gor deletions) or Origami 2 are preferred. For proteins with rare codons, Rosetta 2(DE3) supplies tRNAs for AGG, AGA, AUA, CUA, CCC, and GGA codons. BL21-AI uses the araBAD promoter to control T7 RNAP, giving tighter regulation. For periplasmic expression (signal peptide-directed), standard BL21 strains work, but Lemo21(DE3) allows fine-tuning of expression levels through titrable lysozyme expression.

Lower temperature (16-25°C) is generally better for soluble, correctly folded protein, while 37°C maximizes total expression but often produces inclusion bodies. At 37°C, the translation rate is highest, which can overwhelm the cellular folding machinery. Reducing temperature to 16-18°C slows translation 3-5 fold, reduces hydrophobic interactions that drive aggregation, and allows more time for chaperone-assisted folding. The trade-off is longer induction times (16-24 hours vs 3-4 hours). A practical approach is to grow cells at 37°C to mid-log phase (OD 0.4-0.8), then shift to the lower induction temperature 30 minutes before adding IPTG. For enzymes where activity matters more than yield, 25°C is often a good compromise. For structural biology or biophysical studies requiring maximum soluble protein, 16°C overnight is standard practice.

First, verify that your gene is being transcribed and translated by checking total cell lysate on SDS-PAGE (not just soluble fraction). If you see no band at all, check plasmid integrity by sequencing, verify antibiotic selection is maintained, confirm the reading frame is correct, and ensure your promoter is being induced. If you see a band in the insoluble fraction only, the protein is expressed but misfolding (see inclusion body strategies above). For truly low expression, consider codon optimization -- rare codons like AGG, AGA, CUA, and AUA can cause ribosome stalling and premature termination. Check for mRNA secondary structure near the ribosome binding site using tools like RNAfold. Toxic proteins may be selected against; use tighter promoter control (pLysS, pLysE) or BL21-AI. Sometimes simply changing the N-terminal sequence (first 5-7 codons) dramatically improves expression by altering mRNA structure around the start codon.