Size-exclusion chromatography for peptides: secondary interactions, mobile phase design, and calibration limits

Size-exclusion chromatography (SEC) is the standard separation for resolving soluble aggregates from monomer in protein analysis, and it is often proposed as a companion method for peptides. The extension is less direct than it appears. A protein of 150 kDa and a peptide of 4 kDa differ by nearly two orders of magnitude in hydrodynamic size, and the packing materials, pore distributions, and mobile phases that work well for the former can behave quite differently for the latter. For peptides, the nominal size-based separation competes with ionic and hydrophobic interactions between analyte and stationary phase, and those secondary interactions frequently dominate retention.

This post describes how SEC separates species, why small peptides are a demanding application, how mobile phase composition is typically tuned, and what the calibration curve can and cannot be taken to show.

The separation principle and where peptides sit in it

In an ideal SEC experiment, no analyte adsorbs to the packing. Molecules diffuse into pores whose volume is accessible according to molecular size: large species are excluded from most of the pore volume and elute early, near the void volume, while small species sample nearly all of the pore volume and elute near the total permeation volume. Retention is described by a distribution coefficient, K_av, defined from the elution volume, the void volume, and the total column volume. Over the useful fractionation range of a given pore size, the logarithm of molecular mass is approximately linear in elution volume.

Peptides in the range of roughly 1 to 10 kDa occupy the low-mass end of the fractionation range of the columns commonly sold for proteins. Packings with pore diameters near 125 to 150 angstroms are generally selected for this regime, and columns with narrower pore distributions are sometimes preferred. Even so, the selective range is short. A monomer of 4 kDa and a dimer of 8 kDa differ by a factor of two in mass, which corresponds to a modest difference in hydrodynamic radius, so resolution between them depends on column length, particle size, and flow rate rather than on the wide separation window available to large proteins. Higher-order oligomers are resolved more readily, but they are also the species most affected by dilution and by the sample preparation step.

Hydrodynamic size, rather than mass, governs retention, and peptides are rarely compact. An extended or partially disordered peptide occupies a larger effective volume than a globular species of the same mass and elutes earlier than a mass-based calibration predicts. Conformational changes that accompany a change in solvent, such as helix induction by organic modifier or by ionic strength, can shift retention without any change in association state.

Secondary interactions

Silica-based SEC packings carry residual silanol groups, and polymer-based packings carry their own ionizable or hydrophobic surface features. Peptides are rich in basic residues, and positively charged side chains interact with deprotonated silanols at neutral pH. The effect is a delay in elution, peak tailing, and in extreme cases irreversible loss of analyte on the column. Hydrophobic residues, particularly aromatic and long aliphatic side chains, can interact with the bonded diol layer or the polymer matrix and produce the same pattern through a different mechanism.

These effects are not merely cosmetic. Because adsorption delays elution, an adsorbing monomer may appear to elute later than its size predicts, while an oligomer of the same peptide may appear closer to the monomer peak than expected, which compresses the apparent resolution. Conversely, ionic repulsion between a negatively charged analyte and a negatively charged surface at low ionic strength can cause early elution, an effect known as ion exclusion. A peptide with a net negative charge at the mobile phase pH may therefore elute earlier than a size-matched neutral species. Method development for peptide SEC is largely the work of suppressing both effects so that the retention that remains is size-based.

Recovery is a related concern. Mass balance between the injected amount and the integrated eluted area is the most direct check on whether material is being retained on the column. A method that returns a clean chromatogram but recovers a low or variable fraction of the injected peptide is not reporting on the population that was injected. Comparison against a known-amount injection made without the column, or against an orthogonal quantitation, is commonly used to assess this.

Mobile phase design

The mobile phase in a peptide SEC method has to do several things at once: keep the analyte soluble, preserve the association state being measured, and screen secondary interactions. Neutral-pH phosphate buffers with added salt are the usual starting point for proteins. Ionic strength in the range of 100 to 300 mM, supplied by sodium chloride or sodium sulfate, screens electrostatic interactions with the packing. Too little salt leaves ionic effects in place, while very high salt can strengthen hydrophobic interactions, so the working range is bounded on both sides.

For peptides that adsorb through hydrophobic contacts, a modest proportion of organic solvent, typically acetonitrile in the range of 10 to 30 percent, or an additive such as arginine, is frequently evaluated. Acidic mobile phases containing trifluoroacetic acid or formic acid are another common choice, particularly where compatibility with mass spectrometric detection is wanted. These conditions suppress silanol ionization and protonate acidic residues, but they also change the peptide’s charge and can dissociate non-covalent assemblies. A method that measures aggregate content under conditions that disrupt weak association is measuring something different from a method run in a near-physiological buffer, and the two results should not be treated as interchangeable.

This tension is central to interpretation. Organic modifiers and low pH improve peak shape but can dissolve reversible oligomers on the column, so a result of no detectable aggregate may reflect the mobile phase rather than the sample. Where reversible self-association is of interest, studies often compare the same sample across several mobile phase compositions and report how apparent oligomer content changes. Sample dilution on the column is a further factor: injection volume and concentration set the concentration at which the sample actually travels, and a concentration-dependent equilibrium will shift during the run.

Detection also bears on mobile phase choice. UV absorbance at 214 nm monitors the peptide bond and is sensitive but is affected by absorbing additives; absorbance at 280 nm requires tryptophan or tyrosine and may be insensitive for peptides lacking them. Light scattering and refractive index detectors, used in-line, allow molar mass to be estimated independently of elution position, although low-mass peptides scatter weakly and the signal-to-noise ratio limits their use below a few kilodaltons.

Calibration and its limits

A calibration curve built from standards of known mass is the conventional way to assign apparent molecular mass. For peptide work, standard mixtures of small proteins and peptides, sometimes supplemented with amino acids or small molecules marking the permeation limit, are used to span the range. The resulting relationship is approximate, and several caveats apply.

First, the standards must share the shape and interaction behavior of the analyte for the calibration to transfer. A globular protein calibrant and a flexible, charged peptide will not follow the same line. Second, the usable linear region is narrow at the low-mass end of the column, and extrapolation beyond the lowest calibrant is unreliable. Third, because secondary interactions shift retention, an apparent mass derived from elution position may be off by a considerable margin, and the assignment of a peak as a dimer on the basis of calibration alone is a hypothesis rather than an identification.

For this reason, peak identity in SEC is usually established by corroboration. Collecting fractions for mass spectrometry, comparing against orthogonal techniques such as analytical ultracentrifugation or dynamic light scattering, and examining the effect of changing the mobile phase all contribute to the assignment. Method qualification typically addresses specificity, linearity of the aggregate response, recovery, and precision, along with system suitability criteria such as resolution between monomer and a defined dimer marker, tailing factor, and plate count.

Synthesis

SEC occupies a particular niche in peptide characterization. It is gentle compared with reversed-phase methods, which can disrupt association, and it reports on size directly in principle. In practice, the low mass of most research peptides pushes the method to the edge of its resolving range and amplifies the influence of secondary interactions, so that mobile phase selection, recovery assessment, and orthogonal confirmation carry much of the interpretive weight. Reported oligomer content from an SEC method is best read together with a description of the column, mobile phase, injected concentration, and recovery, since each of those conditions shapes what the chromatogram reflects. Used with those qualifications, SEC remains a useful complement to reversed-phase purity analysis, which resolves chemical impurities that SEC generally cannot, and to the other tools used to characterize association state in solution.