Hydrophobic interaction chromatography (HIC) separates proteins based on differences in surface hydrophobicity, binding them to a hydrophobic stationary phase at high salt concentration and eluting with a descending salt gradient. In the context of antibody-drug conjugate (ADC) purification, this mechanism is uniquely suited to the problem at hand: each drug molecule conjugated to the antibody increases the hydrophobicity of the resulting species in proportion to the number of attached payloads, creating a hydrophobicity ladder that HIC resolves directly into discrete drug-to-antibody ratio (DAR) species. For bispecific antibodies, HIC provides an orthogonal selectivity mechanism to protein A and ion-exchange chromatography for removing homodimer impurities whose charge and size profiles are too similar to the target molecule for conventional polishing steps to resolve.
Quick Take
- HIC is the reference technique for preparative ADC polishing because drug conjugation increases surface hydrophobicity proportionally with DAR, enabling discrete resolution of DAR0, DAR2, DAR4, and DAR6 species under a single inverse salt gradient
- Resin ligand selection determines the window of hydrophobicity that can be resolved without denaturing the protein - less hydrophobic ligands (ether, PPG) preserve protein integrity at the cost of reduced resolution; more hydrophobic ligands (phenyl, butyl) increase resolution but risk unfolding at high salt concentrations
- Ammonium sulfate is the standard kosmotropic salt for HIC mobile phases; starting concentrations of 1.0-1.5 M are typical for mAb and ADC applications, with elution driven by descending gradients to zero or near-zero salt
- Bispecific antibody hole-hole homodimers are detectable and quantifiable by analytical HIC, and preparative HIC removes them based on differential surface hydrophobicity arising from the asymmetric CH3 engineering
- Scale-up of HIC follows the same constant linear velocity and gradient volume-in-column-volumes rules that govern ion-exchange chromatography scale-up, with the additional consideration that high-salt mobile phases increase buffer consumption significantly at commercial scale
How HIC Separates ADC DAR Species
The selectivity of HIC for ADC polishing rests on a straightforward physical principle. Cysteine-linked ADCs are produced by partial reduction of interchain disulfide bonds followed by conjugation of the drug-linker payload to the resulting free thiols. The resulting mixture contains species with zero, two, four, six, and eight drug molecules per antibody - the even DAR distribution reflects the paired nature of the interchain disulfides. Each drug-linker payload carries a lipophilic character that increases the overall surface hydrophobicity of the conjugated antibody proportionally with the number of attached drugs. DAR0 - the unconjugated antibody - is the least hydrophobic species; DAR8 is the most hydrophobic. Under an inverse ammonium sulfate gradient from approximately 1.5 M to zero, these species elute in order of increasing DAR, with the unconjugated antibody eluting first and the highest-loaded species eluting last.
The separation relies on the kosmotropic effect of ammonium sulfate: at high salt concentration, water molecules organize preferentially around the salt ions, reducing hydration of exposed hydrophobic protein surfaces and driving their interaction with the hydrophobic ligand on the resin. As salt concentration decreases, the hydration layer reforms and proteins elute in reverse order of their hydrophobic surface area. For ADCs, this mechanism produces a characteristic chromatogram in which each DAR species appears as a distinct peak, enabling both quantification of the drug load distribution and preparative collection of target DAR fractions. The separation of DAR species by inverse salt gradient is well established for cysteine-linked ADCs; a 2025 study in mAbs demonstrated that ammonium tartrate - a kosmotropic salt that is thermally decomposable and therefore mass spectrometry-compatible - delivers DAR separation equivalent to conventional ammonium sulfate, with elution order following the predicted hydrophobicity increase with each additional drug-linker unit.
Resin Selection: Matching Ligand Hydrophobicity to the Target Molecule
HIC resin selection for ADC polishing requires balancing two competing requirements: resolution between adjacent DAR species and protein integrity under adsorption conditions. More hydrophobic ligands - butyl and phenyl being the most widely used - provide stronger hydrophobic interactions and sharper resolution between DAR species, but the high ammonium sulfate concentrations required to drive binding can induce partial unfolding of the adsorbed protein if the ligand-protein interaction is too strong. Less hydrophobic ligands - ether and polypropylene glycol (PPG) - require lower salt concentrations for binding and reduce the risk of denaturation, but at the cost of resolution, particularly between closely eluting DAR species.
For most cysteine-linked IgG1 ADCs, phenyl and butyl resins achieve adequate resolution under standard ammonium sulfate gradients. The choice between them is molecule-specific: the native hydrophobicity of the parent antibody, the hydrophobicity of the drug-linker, and the isoelectric point all influence retention behavior on HIC, and a resin that resolves DAR species cleanly for one ADC may compress the distribution for another. Resin scouting at bench scale using a linear gradient from 1.5 M to 0 M ammonium sulfate in phosphate buffer at pH 6.5-7.0 is the standard starting point, with ligand hydrophobicity and gradient slope adjusted empirically to achieve the required peak-to-peak resolution between adjacent DAR species before any scale-up decisions are made.
| Ligand Type | Hydrophobicity | Typical Starting Salt | Best-Fit Application | Key Limitation |
|---|---|---|---|---|
| Ether / PPG | Low | 0.75-1.0 M (NH4)2SO4 | Aggregation-prone mAbs, sensitive ADCs | Reduced resolution between DAR species |
| Phenyl | Medium | 1.0-1.5 M (NH4)2SO4 | Standard IgG1 ADC polishing | Molecule-specific optimization required |
| Butyl | Medium-high | 1.0-1.5 M (NH4)2SO4 | High-resolution DAR separation | Risk of unfolding at high salt |
| Octyl | High | 0.5-1.0 M (NH4)2SO4 | Analytical applications | Denaturation risk; rarely used preparatively |
HIC for Bispecific Antibody Homodimer Removal
Knob-into-hole bispecific antibodies are produced by co-expressing two different heavy chains engineered with complementary steric modifications to promote heterodimerization. Despite this design, homodimerization - particularly of the hole-hole heavy chain pair - occurs as a minor but functionally significant by-product. Hole-hole homodimers are a critical quality attribute in bispecific antibody manufacturing because they lack one of the two target-binding arms and cannot be distinguished from the bispecific product by size alone.
The surface hydrophobicity of hole-hole homodimers differs from the target bispecific molecule due to the asymmetric CH3 domain engineering that defines the knob-into-hole architecture. This differential hydrophobicity is small enough that HIC analytical columns are typically used to quantify homodimer content rather than to achieve preparative separation, but preparative HIC has been applied as a polishing step where the hydrophobicity window is sufficient. A 2025 study in the journal Membranes examining HIC aggregate and homodimer removal from bispecific antibody preparations demonstrated that protein-resin interactions modulated by ammonium sulfate concentration gradients can achieve selective retention of aggregate and homodimer species while the target bispecific passes through, under optimized flow-through conditions. The operating mode - flow-through versus bind-and-elute - is determined by whether the target molecule binds the resin under the load conditions; in flow-through HIC, the load salt concentration is set so the target does not bind while more hydrophobic impurities are retained, exactly as mixed-mode chromatography flow-through polishing operates for aggregate removal.
Method Development: Salt Gradient Design and pH Effects
HIC method development for ADC polishing begins with salt gradient scouting to establish the ammonium sulfate concentration at which the target ADC species bind the chosen resin and the gradient slope required to resolve adjacent DAR peaks. The standard approach uses a linear inverse gradient from the starting salt concentration to zero over 20-30 column volumes, matching the gradient volume rule used for IEX: number of column volumes, not absolute buffer volume, is the scaling invariant. Gradient slope affects both resolution and run time - a shallower gradient improves peak spacing but increases buffer consumption and cycle time at commercial scale, where buffer management is a significant operational constraint.
pH has a secondary but meaningful effect on HIC selectivity. At low pH, proteins carry more positive surface charge, which can modulate the effective hydrophobicity of the resin interaction; at high pH, surface charge effects are reduced. For most IgG-based ADCs, a phosphate buffer pH of 6.5 to 7.0 provides a good starting point, but pH scouting in combination with salt concentration scouting is warranted for molecules that show unusual retention behavior at standard conditions. Temperature also affects HIC retention - hydrophobic interactions generally strengthen with increasing temperature, meaning that separations developed at ambient temperature may shift at the higher temperatures that can occur during a long production run if column temperature is not controlled.
For preparative ADC polishing in a GMP context, the target DAR fraction is typically collected by UV fraction cutting rather than by time-based collection, with the cut points defined by the resolution achieved between DAR species. The acceptable DAR distribution - typically centered around DAR4 for cysteine-linked IgG1 ADCs, with specification limits on unconjugated antibody (DAR0) and high-drug-load species (DAR6 and above) - defines the required resolution that the HIC step must achieve and therefore the resin, salt, and gradient conditions that the method must deliver.
HIC in the ADC and Bispecific Downstream Purification Sequence
HIC for ADC polishing typically follows the Protein A capture step and virus inactivation, occupying the polishing position where the bulk of DAR distribution adjustment occurs. The high salt concentration required for HIC binding means the eluate conductivity is low - the inverse gradient ends near zero ammonium sulfate - and the eluate typically requires buffer exchange or dilution before the next ion-exchange polishing step. This conductivity transition is an important process design consideration at commercial scale: the buffer exchange step adds capital and time to the process, and selecting a downstream IEX resin that tolerates direct HIC eluate loading can eliminate it.
For bispecific antibodies, HIC is more often deployed analytically for in-process monitoring of homodimer content than as a preparative polishing step, with mixed-mode chromatography or nanobody-based affinity resins providing higher selectivity for specific impurity classes at production scale. Where preparative HIC is used for bispecific polishing, it sits within the same advanced chromatography downstream sequence that governs all polishing step design, and the column sizing, gradient volume, and scale-up approach described in the context of downstream purification apply directly.
This article was produced under Separation Science's AI Editorial Guidelines.




