Size Exclusion Chromatography for Protein Purification: Principles and Protocol

Size exclusion chromatography (SEC), also known as gel filtration chromatography, is a liquid chromatography method that separates proteins and other macromolecules based on their hydrodynamic volume (effective molecular size) rather than chemical interaction with the stationary phase. The column is packed with porous beads; molecules too large to enter the pores elute first in the void volume, while smaller molecules that penetrate the pores are retarded and elute later. SEC is uniquely suited for desalting, buffer exchange, removal of small-molecule contaminants, and separation of proteins that differ substantially in molecular weight (typically by a factor of ≥2). It is a gentle, nondenaturing method that preserves native protein structure and biological activity, making it an ideal final polishing step or a first step for removing aggregates and low-molecular-weight impurities from a crude protein mixture. Unlike affinity or ion exchange chromatography, SEC does not rely on binding interactions, so it can be applied to virtually any protein under conditions that maintain solubility and stability.

At a Glance

Scientific Principle: Separation by Hydrodynamic Volume

SEC separates molecules based on their ability to access the pores within a packed bed of porous, cross-linked polymer beads (typically agarose, dextran, or polyacrylamide). The stationary phase consists of beads with a defined pore size distribution. Molecules in solution diffuse into and out of these pores. A molecule that is larger than the largest pore diameter cannot enter any pore and is excluded from the internal volume of the beads. It travels only through the interstitial space (the volume between beads) and elutes first at the void volume (V₀). A molecule small enough to enter all pores will access the entire internal volume of the beads and elute last at the total column volume (Vₜ). Molecules of intermediate size partition between the interstitial and internal volumes according to their hydrodynamic radius and elute between V₀ and Vₜ.

The partition coefficient (Kₐᵥ) describes the fraction of the internal pore volume accessible to a given molecule:

Kₐᵥ = (Vₑ – V₀) / (Vₜ – V₀)

where Vₑ is the elution volume of the molecule. For a given column, Kₐᵥ ranges from 0 (completely excluded) to 1 (completely included). The relationship between log molecular weight and Kₐᵥ is approximately linear over the fractionation range of the column, which is specified by the manufacturer for globular proteins. This linear relationship allows estimation of the molecular weight of an unknown protein by comparing its elution volume to those of known standards.

The separation mechanism is purely entropic: larger molecules have fewer accessible pore configurations and therefore elute earlier. Because there is no binding or desorption step, SEC is a gentle method that does not denature proteins or require gradient elution. The mobile phase composition remains constant (isocratic elution), and the buffer can be chosen to maintain protein stability, solubility, and activity.

Column Selection: Matching Pore Size and Bed Dimensions

Choosing the correct SEC column is the most critical decision for a successful separation. Two parameters dominate: the fractionation range of the resin and the column dimensions (length and diameter).

Fractionation Range and Resin Type

The fractionation range defines the molecular weight window within which proteins are separated. For example, a resin with a fractionation range of 10–150 kDa will separate a 20 kDa protein from a 100 kDa protein but will not resolve a 100 kDa protein from a 150 kDa protein. Resins with narrower fractionation ranges provide better resolution within that window. Common resin types include:

• Superdex 200 (10–600 kDa): Broad range suitable for most soluble proteins and protein complexes
• Superdex 75 (3–70 kDa): Ideal for small proteins, peptides, and separation of monomers from dimers
• Sephacryl S-300 (10–1,500 kDa): For very large proteins and protein complexes
• Sephadex G-25 (1–5 kDa): Designed for desalting and buffer exchange; excludes proteins >5 kDa while retaining salts and small molecules

For desalting or buffer exchange, a resin with a fractionation range that excludes the target protein (i.e., the protein elutes in the void volume) while including the small-molecule contaminants is optimal. Sephadex G-25 and similar desalting resins are the standard choice.

Column Dimensions and Resolution

Resolution in SEC is proportional to the square root of column length. Longer columns provide more theoretical plates and better separation, but also increase run time and back pressure. For high-resolution separations, columns of 60–100 cm length with internal diameters of 1.0–1.6 cm are typical. For desalting, short columns (5–20 cm) with larger diameters (1.6–5.0 cm) are used to maximize sample throughput at the expense of resolution.

The sample volume must be small relative to the column bed volume for high-resolution separations. A general rule is to load no more than 5% of the bed volume for analytical separations. For desalting, sample volumes up to 30% of the bed volume are acceptable because the goal is simply to separate the excluded protein from included small molecules.

Prepacked vs. Self-Packed Columns

Prepacked columns from manufacturers (e.g., Cytiva, Bio-Rad, Tosoh) offer reproducibility, documented performance, and validated packing quality. They are recommended for most applications, especially when regulatory compliance or method transfer is required. Self-packing is possible but requires experience, appropriate slurry packing equipment, and careful quality control (e.g., checking plate count and asymmetry). For routine laboratory work, prepacked columns save time and reduce variability.

Sample Preparation and Loading

Proper sample preparation is essential for SEC success. The sample must be free of particulate matter that could clog the column frit or bed. Centrifuge the sample at 10,000–15,000 × g for 10 minutes at 4°C, or filter through a 0.22 μm or 0.45 μm syringe filter. For viscous samples (e.g., cell lysates), dilution or addition of a nuclease (e.g., Benzonase) may be necessary to reduce viscosity and prevent column fouling.

The sample buffer should match the column equilibration buffer to avoid buffer mismatch artifacts (e.g., protein precipitation, pH shifts). If the sample is in a different buffer, dialyze or dilute it into the column buffer before loading. The sample concentration should be high enough to allow detection after dilution during chromatography, but not so high that protein aggregation occurs. Typical protein loads range from 0.1–10 mg per run for analytical columns, depending on column size and detection sensitivity.

Sample volume is critical. For high-resolution separations, load ≤5% of the column bed volume. For a 120 mL column, this means a maximum sample volume of 6 mL. For desalting, load up to 30% of the bed volume. Exceeding these limits causes peak broadening and loss of resolution.

Column Equilibration and Operation

Equilibrate the column with at least 2–3 column volumes of the chosen mobile phase before loading the sample. The mobile phase should be degassed (by vacuum filtration or sonication) to prevent bubble formation in the column. Common buffers include phosphate-buffered saline (PBS), Tris-HCl, or HEPES, with or without added salts (e.g., 150 mM NaCl) to maintain ionic strength and prevent nonspecific interactions. The pH should be within the protein’s stability range, typically pH 6.0–8.0 for most soluble proteins.

Flow rate affects resolution. Lower flow rates improve resolution because they allow more time for diffusion into and out of the pores. For high-resolution columns, flow rates of 0.5–1.0 mL/min are typical for a 1.6 cm diameter column. For desalting columns, higher flow rates (2–5 mL/min) are acceptable because resolution is not the primary goal. Always stay within the manufacturer’s recommended pressure limit to avoid compressing the bed and damaging the column.

Fraction Collection and Detection

Collect fractions of constant volume, typically 0.5–2.0 mL for analytical columns or 1–5 mL for preparative columns. The fraction volume should be small enough to capture the peak of interest without excessive dilution. A good rule is to collect fractions that are 1–2% of the column bed volume.

Detection is usually by UV absorbance at 280 nm (for proteins containing aromatic amino acids) or at 214 nm (for peptide bonds). For proteins lacking tryptophan and tyrosine, alternative detection methods (e.g., fluorescence, refractive index, or post-column derivatization) may be needed. Collect fractions across the entire elution profile, not just the main peak, because unexpected peaks may contain aggregates, degradation products, or binding partners.

Quality Checks and Column Calibration

Before using a new column or after a change in application, calibrate the column with a set of known molecular weight standards. This calibration serves two purposes: it verifies that the column is properly packed and performing within specifications, and it establishes the relationship between elution volume and molecular weight for the specific column and buffer conditions.

Calibration Procedure

1. Prepare a mixture of at least 3–5 globular protein standards spanning the fractionation range of the column. Common standards include thyroglobulin (670 kDa), γ-globulin (158 kDa), ovalbumin (44 kDa), myoglobin (17 kDa), and vitamin B12 (1.35 kDa).
2. Dissolve standards in the column buffer at 1–5 mg/mL each.
3. Inject a small volume (1–5% of bed volume) and elute under standard conditions.
4. Record the elution volume (Vₑ) for each standard.
5. Plot log molecular weight vs. Vₑ or Kₐᵥ. The plot should be linear over the fractionation range.
6. Calculate the void volume (V₀) using a completely excluded molecule (e.g., blue dextran, 2,000 kDa) and the total column volume (Vₜ) using a completely included molecule (e.g., vitamin B12 or acetone).

A well-packed column should yield a linear calibration curve with R² > 0.98. Deviations from linearity may indicate column degradation, channeling, or improper packing.

Column Performance Checks

Periodically check column performance by measuring the plate number (N) and asymmetry factor (Aₛ) for a test compound (e.g., acetone). For a 60 cm column, N should be >10,000 plates/m, and Aₛ should be between 0.8 and 1.2. A decrease in plate number or increase in asymmetry over time indicates column fouling or bed compression.

Result Interpretation

The elution profile (chromatogram) shows peaks corresponding to different molecular size populations. The first peak (at V₀) contains molecules larger than the exclusion limit of the resin—typically protein aggregates, large complexes, or very high molecular weight species. The last peak (at Vₜ) contains small molecules such as salts, nucleotides, and small peptides. Between these extremes, peaks represent proteins of intermediate size.

For desalting or buffer exchange, the target protein elutes in the void volume (first peak), while salts and small molecules elute in the included volume (later peak). Collect the void volume peak and discard the rest. The success of desalting can be confirmed by measuring conductivity of the collected fractions; the protein peak should have conductivity equal to the column buffer, while the salt peak will show elevated conductivity.

For molecular weight estimation, compare the elution volume of the unknown protein to the calibration curve. Note that SEC estimates hydrodynamic volume, not true molecular weight. Nonglobular proteins (e.g., extended or rod-shaped proteins) will elute earlier than globular proteins of the same molecular weight because they have a larger hydrodynamic radius. Glycosylated proteins also behave anomalously due to the high hydration of carbohydrate chains. For accurate molecular weight determination, use SEC in combination with other methods such as mass spectrometry or analytical ultracentrifugation.

Troubleshooting

Limitations

SEC has several inherent limitations that users must understand. First, resolution is modest compared to affinity or ion exchange chromatography. Proteins that differ by less than a factor of two in molecular weight are difficult to separate completely. Second, the sample volume is severely constrained for high-resolution work, limiting the amount of protein that can be processed per run. Third, SEC dilutes the sample because the protein elutes in a volume larger than the loaded volume. For a well-packed column, the dilution factor is typically 1.5–3.0. Fourth, SEC cannot separate proteins of similar size, such as isoforms or post-translationally modified variants, unless the modification substantially alters hydrodynamic volume. Fifth, the method is relatively slow compared to binding-based methods, especially for long high-resolution columns.

For desalting and buffer exchange, SEC is effective but less scalable than dialysis or tangential flow filtration. For processing large volumes (>100 mL), dialysis or ultrafiltration may be more practical.

Documentation and Record Keeping

For reproducible and auditable results, document the following for each SEC run:

• Column type, serial number, and packing date
• Resin lot number and fractionation range
• Equilibration buffer composition, pH, and preparation date
• Flow rate and pressure during the run
• Sample identity, concentration, volume, and buffer composition
• Injection method (manual loop, automated injector)
• Fraction collection parameters (fraction volume, number of fractions)
• UV detection wavelength and sensitivity
• Calibration data (standards used, elution volumes, calibration curve)
• Column performance metrics (plate number, asymmetry) and date of last check
• Any deviations from the standard protocol

Maintain a column logbook that records every run, including any unusual observations (e.g., pressure spikes, baseline drift). This documentation is essential for troubleshooting, method transfer, and regulatory compliance.

Biosafety Considerations

SEC is performed under BSL-1 conditions when working with nonpathogenic proteins and recombinant proteins that do not require higher containment. Follow institutional biosafety guidelines as outlined in the CDC/NIH Biosafety in Microbiological and Biomedical Laboratories (BMBL) [6] and the NIH Guidelines for Research Involving Recombinant or Synthetic Nucleic Acid Molecules [7]. For work with recombinant proteins, ensure that the institutional biosafety committee has approved the project and that appropriate containment levels are used.

Standard laboratory practices include:

• Wear lab coat, gloves, and eye protection when handling samples and column buffers
• Decontaminate work surfaces before and after use with 70% ethanol or appropriate disinfectant
• Dispose of sample waste and column eluate according to institutional hazardous waste guidelines
• If the sample contains recombinant organisms, inactivate them (e.g., by filtration or chemical treatment) before loading onto the column
• Clean and sanitize the chromatography system according to manufacturer recommendations between different protein samples to prevent cross-contamination

For work with human-derived samples (e.g., plasma, serum), treat all materials as potentially infectious and follow BSL-2 practices unless the samples have been tested and confirmed negative for bloodborne pathogens. The use of SEC for isolating extracellular vesicles from human plasma or serum, as described in recent studies [2][5], requires appropriate biosafety precautions for handling human blood products.

Frequently Asked Questions

1. Can I use SEC to separate proteins of similar molecular weight, such as a 50 kDa protein from a 55 kDa protein? No. SEC cannot reliably separate proteins that differ by less than approximately 2-fold in molecular weight. For a 50 kDa and 55 kDa protein, the difference is only 10%, which is well below the resolution limit of SEC. Use ion exchange chromatography, hydrophobic interaction chromatography, or affinity chromatography for such separations.

2. Why does my protein elute earlier than expected based on its molecular weight? Several factors can cause early elution. The protein may be nonglobular (elongated or rod-shaped), which increases its hydrodynamic volume. It may be aggregated, forming dimers, trimers, or higher-order oligomers. It may be heavily glycosylated, as carbohydrate chains are highly hydrated and increase the effective size. Alternatively, the column calibration may be inaccurate if the standards are not appropriate for the protein’s shape or if the calibration has not been performed recently.

3. How do I choose between a desalting column and a high-resolution SEC column? Use a desalting column (e.g., Sephadex G-25) when the goal is simply to remove salts, small molecules, or exchange buffers. Desalting columns are fast, tolerate large sample volumes (up to 30% of bed volume), and require minimal equipment. Use a high-resolution SEC column when you need to separate proteins by size, remove aggregates from a monomeric protein, or estimate molecular weight. High-resolution columns require smaller sample volumes and longer run times but provide much better separation.

4. Can I reuse SEC columns, and how do I store them? Yes, SEC columns can be reused many times if properly maintained. After each run, wash the column with 1–2 column volumes of the running buffer. For long-term storage, equilibrate the column in a storage buffer containing 0.02% sodium azide (or 20% ethanol) to prevent microbial growth. Store at 4°C, and never let the column dry out. Replace the column if performance degrades (decreased plate number, increased asymmetry, or high back pressure).

References and Further Reading

• Engineering of Glioblastoma-Derived Biomimetic Vesicles and Their Structural and Molecular Features — Describes SEC purification of biomimetic vesicles from unbound proteins, demonstrating SEC as a polishing step after vesicle fabrication.
• From platelets to particles: Optimizing extracellular vesicle isolation in a National Blood Centre — Compares SEC with other isolation methods for platelet-derived extracellular vesicles, showing SEC yields high particle numbers with reduced protein contaminants.
• Selective Lipoprotein Removal Enables High-Purity EV Isolation from Plasma via Aptamer-Based Mesh Filtration — Integrates SEC with lipoprotein depletion for ultrapure extracellular vesicle isolation from plasma.
• Recent Advances on Extracellular Vesicles: A Natural Nanomaterial for Biomedical Application — Reviews SEC as a mainstream isolation method for extracellular vesicles, comparing its advantages and limitations relative to ultracentrifugation and microfluidic technologies.
• Improvement of small extracellular vesicle isolation from mouse model blood — Demonstrates that SEC yields higher vesicle numbers and fewer contaminating proteins compared to ultracentrifugation-density gradient methods for small-volume blood samples.
• Biosafety in Microbiological and Biomedical Laboratories (BMBL), 6th Edition — Authoritative principles for risk assessment, containment, decontamination, and microbiological laboratory practice.
• NIH Guidelines for Research Involving Recombinant or Synthetic Nucleic Acid Molecules — Institutional and biosafety framework for recombinant and synthetic nucleic acid research.
• NCBI Bookshelf: Molecular Biology and Laboratory Methods — Searchable collection of authoritative biomedical books and methods references, including detailed SEC protocols.

Related Articles

• Affinity Chromatography for Protein Purification: Principles and Column Selection
• Protein Immunoprecipitation: Principles and Protocol
• Immunofluorescence Assay: Principles and Protocol for Protein Localization
• How to Interpret a Protein Purification Table: Yield, Purity, and Fold Purification
• Lowry Assay Protocol: Protein Quantification Method
• Kinase Activity Assay: Methods for Measuring Protein Kinase Activity