How to Separate Insulin from Host Cell Proteins

Property Comparison

Insulin (Target)

Molecular Weight5,808 Da

TypePeptide Hormone

Charge (pH 4)+2 to +3

pI5.3

SolubilitySoluble at pH <4 or >7

Structure2 chains, 3 disulfide bonds

HydrophobicityModerate (GRAVY −0.1)

vs

Host Cell Proteins (Impurity)

Molecular Weight10–300 kDa (diverse)

TypeE. coli Proteome

Charge (pH 4)Variable (+/−)

pI4–11 (broad range)

SolubilityVariable

Count~4,300 unique proteins

HydrophobicityVaries widely

Why This Separation Works

Insulin is a small, well-characterized peptide with defined pI (5.3) and hydrophobicity. HCPs are a diverse mixture of thousands of proteins with varying properties. Multi-dimensional chromatography exploits different selectivity at each step to progressively remove HCP subpopulations:

Property	Insulin	Most HCPs	Separation Step
Charge at pH 4	+2 to +3 (below pI)	Variable, many neutral or negative	Cation exchange (CEX)
Size	5.8 kDa (very small)	10–300 kDa (larger)	Size exclusion (SEC)
Hydrophobicity	Defined C18 retention	Diverse retention times	Reverse-phase HPLC

No single step achieves pharmaceutical purity. The combination of orthogonal selectivities (charge + hydrophobicity + size) is required to reduce HCPs to <100 ppm.

Recommended Process Route

1

Cell Lysis & Inclusion Body Isolation

Harvest E. coli cells by centrifugation. Lyse by high-pressure homogenization (800–1000 bar). Wash inclusion bodies (IBs) with 1% Triton X-100 and 1 M urea to remove membrane fragments, DNA, and loosely bound HCPs. IBs are >50% proinsulin at this stage.

Primary recovery

2

Solubilization & Refolding

Dissolve IBs in 6–8 M guanidine HCl with 10 mM DTT. Dilute 1:20 into refolding buffer (pH 10.5, 1 mM cysteine/cystine redox pair) to form the three correct disulfide bonds (A6–A11, A7–B7, A20–B19). Refolding yield: 60–80%.

Renaturation

3

Cation Exchange Chromatography

Adjust to pH 4.0 (below insulin pI 5.3). Insulin binds to SP Sepharose as a cation; most HCPs with pI <4 do not bind. Elute with NaCl gradient (0–0.5 M). This step removes >90% of HCPs and misfolded variants. Capacity: 15–25 mg insulin per mL resin.

Capture step

4

Enzymatic Conversion (Proinsulin Route)

Cleave C-peptide from proinsulin using trypsin and carboxypeptidase B at pH 7.5, 30°C, 4–8 hours. Conversion >95%. This step generates insulin, C-peptide, and partially cleaved intermediates that must be separated in the polishing step.

Enzymatic processing

5

Reverse-Phase HPLC Polishing

Final purification on C8 or C18 RP-HPLC column with acetonitrile/water gradient containing 0.1% TFA. Resolves insulin from des-amido variants (Asp→isoAsp), C-peptide, and remaining HCPs. Achieves >98% purity by HPLC. This is the critical step for pharmaceutical compliance.

Final polishing

Expected Results

30–50%

Overall Insulin Yield

>98%

Insulin Purity (HPLC)

5 steps

Total Process Length

HCP levels must be reduced to <100 ppm for pharmaceutical insulin. Endotoxin removal (<5 EU/dose) requires additional polishing or endotoxin-specific affinity steps.

Alternative Techniques

Technique	Feasibility	Notes
Affinity Chromatography (anti-insulin)	Good	Immunoaffinity or Zn²+-IMAC capture gives excellent selectivity. High cost per cycle; mainly used for analytical-scale or high-value biosimilars.
Hydrophobic Interaction (HIC)	Good	Complements CEX as orthogonal step. Insulin binds at high ammonium sulfate and elutes in decreasing salt gradient. Used in some commercial processes.
Size Exclusion (SEC)	Moderate	Removes high-MW aggregates and large HCPs but low throughput. Best as a final polishing step, not primary capture. Typical column: Superdex 75.
Membrane Chromatography	Moderate	Anion exchange membrane adsorbers (e.g., Sartobind Q) in flow-through mode remove DNA and endotoxin. Useful as a polishing step but cannot replace primary capture.

Frequently Asked Questions

Why is insulin produced in inclusion bodies rather than soluble form?

E. coli overexpression of insulin (or proinsulin) at high levels (>20% total cell protein) causes the protein to aggregate into insoluble inclusion bodies. While this requires solubilization and refolding steps, it has advantages: IBs are easily isolated by centrifugation, are naturally resistant to proteolysis, and the initial purity (>50% target protein) is higher than soluble expression.

Why is reverse-phase HPLC needed despite ion exchange giving >90% purity?

Pharmaceutical insulin requires >98% purity with specific impurity limits: desamido insulin <1%, other related substances <1% each. CEX cannot resolve insulin from its deamidation variants (same charge, same size). RP-HPLC exploits subtle hydrophobicity differences between insulin and its degradation products, providing the resolution needed for pharmacopeial compliance.

What is the overall yield of recombinant insulin production?

Typical overall yields are 30–50% from inclusion bodies to final product. Major losses occur at refolding (20–40% misfolded), enzymatic conversion (5–10% incomplete cleavage), and chromatography steps (10–20% in tails/side fractions). Modern optimized processes can achieve the higher end of this range.

Related Separation Guides

Lysozyme from Ovalbumin Whey Protein from Lactose Lactic Acid from Glucose Penicillin from Broth Insulin Properties Ion Exchange Guide