What Is CE-SDS? Capillary Electrophoresis for Proteins

CE-SDS, or capillary electrophoresis–sodium dodecyl sulfate, is an analytical technique that separates proteins by size inside a narrow glass tube (a capillary) rather than on the flat gel slabs traditionally used in biochemistry labs. It works by coating proteins with the detergent SDS so they all carry a uniform negative charge, then pulling them through a polymer-filled capillary with an electric field. Smaller proteins navigate the polymer mesh faster, larger ones slower, producing a readout that tells you a protein’s approximate molecular weight and how pure your sample is. The method has become a workhorse in the pharmaceutical industry, especially for checking the quality of antibody-based drugs, and it shows up routinely in regulatory filings worldwide.

How the Separation Actually Works

In everyday lab shorthand, CE-SDS is the capillary-based cousin of SDS-PAGE, the gel electrophoresis technique that has been a staple of protein biochemistry for decades. Both rely on the same core trick: SDS is a detergent that wraps around protein chains and gives them a consistent charge-to-mass ratio. Once that ratio is roughly the same for every protein in the mixture, the only thing that determines how fast each one moves through a sieving medium is its size.

The difference is the format. Instead of casting a slab of polyacrylamide gel and running proteins through it in a chamber, CE-SDS fills a thin fused-silica capillary with a liquid polymer solution that acts as a sieving matrix. Early work showed that a replaceable polymeric solution could separate SDS-denatured proteins in a size-dependent way, with a linear relationship between the logarithm of a protein’s molecular weight and its mobility across a range of roughly 14,000 to 205,000 daltons.¹ That linear relationship is what lets you estimate the molecular weight of an unknown protein by comparing its migration time to a set of standards.

Because the sieving matrix is a liquid polymer rather than a rigid gel, it can be flushed out and replaced between runs. This is a practical advantage: each sample gets a fresh separation medium, which keeps results consistent and reduces carryover. Various polymer chemistries have been explored over the years, including polyethylene oxide and polyvinyl alcohol formulations, but most commercial CE-SDS kits use proprietary blends optimized for the protein size ranges that biopharmaceutical labs care about most.

Why Capillary Coatings Matter

One of the trickier engineering details in CE-SDS involves the capillary wall itself. Bare fused silica has charged groups on its inner surface that interact with the running buffer and create a bulk flow of liquid called electroosmotic flow. That flow can distort separations and ruin reproducibility. To suppress it, capillaries are coated with polymers that shield those surface charges. Some coatings simply adsorb onto the wall, but these can be unstable under the harsh conditions inside the capillary. More robust approaches use covalent bonding. One strategy synthesizes a modified polymer that first adsorbs loosely, then forms permanent chemical bonds to the silica surface when heated, locking the coating in place.² Getting this coating right is one of the reasons commercial CE-SDS systems cost more to set up than a basic gel rig, but it is also why they can deliver highly reproducible data over hundreds of runs.

Reduced and Non-Reduced Modes

CE-SDS is typically run in two configurations, and the choice depends on what you want to learn about your protein.

In non-reduced CE-SDS, proteins are treated with SDS but their disulfide bonds are left intact. Disulfide bonds are the chemical crosslinks that hold multi-chain proteins together. For an antibody, those bonds keep the heavy chains and light chains assembled into the familiar Y-shaped molecule. So in non-reduced mode, the intact antibody migrates as a single large species, and you can see whether any of it has broken into fragments or clumped into aggregates. A validated non-reduced method tested across 26 approved monoclonal antibodies and two antibody-drug conjugates showed it could accurately quantify size variants in about 40 minutes per sample, making it practical for routine batch-release testing.³

In reduced CE-SDS, a reducing agent is added to break those disulfide bonds before the run. The antibody falls apart into its individual heavy and light chains, and you can assess the purity of each chain separately. Reduced analysis is especially useful for spotting degradation products and modifications on individual chains. It is also the mode where certain artifacts tend to pop up. In the analysis of the antibody pertuzumab, for example, unexpected peaks appeared between the light-chain and heavy-chain signals. These turned out to be non-covalent light-chain dimers, artifacts formed during sample preparation rather than real impurities. The problem was solved by raising the capillary temperature to 60°C and lowering the separation voltage, conditions that disrupted the artificial association.⁴

Fragments from cleavage events inside the antibody’s CH2 domain present another interpretive wrinkle. These fragments are linked by intrachain disulfide bonds, so in non-reduced mode they can co-migrate with the intact antibody peak and go undetected. Reduced CE-SDS, by breaking those bonds, can resolve them as separate species.⁵ Running both reduced and non-reduced modes gives a more complete picture than either alone.

How CE-SDS Compares to Traditional Gel Electrophoresis

The most natural question for anyone who has worked with SDS-PAGE is whether CE-SDS actually gives the same answers. A direct head-to-head comparison found that CE-SDS-based methods are similar to SDS-PAGE in precision and repeatability, including the apparent molecular-mass shifts you see with glycosylated proteins.⁶ In a separate study on pertussis vaccine antigens, CE-SDS demonstrated better resolution than SDS-PAGE for separating the subunit peaks of pertussis toxin and isoforms of filamentous hemagglutinin, while delivering comparable results for purity and molecular weight estimates overall.⁷

Where CE-SDS pulls ahead is in quantification and throughput. SDS-PAGE gels are stained and then imaged, and the staining step introduces variability. CE-SDS detects proteins as they pass a detector at the end of the capillary, generating a quantitative trace (an electropherogram) directly. You get numbers, not band-intensity estimates. The technique is also automated: modern instruments load samples from 96-well plates, rinse and refill the capillary between injections, and run unattended overnight. For labs that need to analyze dozens or hundreds of samples under standardized conditions, this automation is a significant practical advantage over manually casting and running gels.

Detection Options and Sensitivity

Most CE-SDS instruments detect proteins by ultraviolet (UV) absorbance at 220 nm, which picks up the peptide bonds present in every protein. UV detection works well for concentrated samples and is the standard in quality-control labs. But some applications need much higher sensitivity, and that is where laser-induced fluorescence (LIF) detection comes in.

For LIF, proteins are labeled with a fluorescent dye before analysis. One optimized labeling scheme achieved detection limits as low as 10 ng/mL for monoclonal antibodies.⁸ A comparison of CE-SDS with UV detection versus CE-SDS with LIF for adeno-associated virus (AAV) capsid proteins found that the LIF method offered superior resolution and a significant increase in signal intensity.⁹ The extra sensitivity matters when you are working with low-abundance proteins or need to quantify trace impurities.

Applications Beyond Antibodies

Monoclonal antibodies dominate the CE-SDS literature because they dominate the biopharmaceutical market, but the technique extends well beyond them. Gene therapy products built on adeno-associated virus vectors are a growing area. AAV capsids are assembled from three viral proteins (VP1, VP2, and VP3) in specific ratios, and confirming that ratio is a critical quality check. CE-SDS with LIF is now considered the reference method for separating these capsid proteins under generic conditions.¹⁰ Sample-stacking techniques have been developed to further boost sensitivity, providing roughly a thousand-fold improvement that makes the method a practical replacement for SDS-PAGE in AAV product release testing across different serotypes.¹¹

Vaccine development is another area where CE-SDS has found a niche. A validated method for acellular pertussis vaccine antigens showed it could serve as a stability-indicating tool, detecting degradation over time in a way that supports quality-control decisions.⁷ And in biosimilar development, where a manufacturer must demonstrate that a follow-on product matches the originator, CE-SDS is used alongside other techniques to track degradation profiles under stress conditions. Studies comparing biosimilar and originator antibodies under thermal stress have used non-reduced and reduced CE-SDS to monitor the time- and temperature-dependent growth of fragments and the decline of intact protein.¹²

Sample Preparation Pitfalls

CE-SDS is sensitive enough that artifacts from sample preparation can masquerade as real impurities, and this is something that catches labs off guard. The most widely documented pitfall involves iodoacetamide (IAM), an alkylating agent routinely added during non-reduced sample preparation to cap free thiol groups and prevent disulfide bond reshuffling. For most antibodies this works fine, but for at least one recombinant protein, IAM was found to alkylate lysine and serine residues as well, creating a heavier modified form that showed up as an extra peak in the electropherogram. The amount of this artifact depended heavily on preparation conditions, including whether SDS was present during the alkylation step.¹³ The takeaway is that sample-preparation conditions cannot simply be transferred from one protein to the next without verification. A new protein might react differently with standard reagents, and the only way to know is to check.

Baseline Issues and Data Integrity

Anyone who has run CE-SDS in a regulated environment knows that the electropherograms are not always as clean as a chromatographer would like. Baseline disturbances, drifts, and ripples are common enough that they have been studied systematically. Investigation has traced these disturbances to thermal disruptions during sample runs and to the optical properties of the background electrolytes themselves. Practical recommendations include running shorter injection sequences with fewer samples per sequence and closely monitoring baseline drift to ensure accurate peak integration.¹⁴

Peak integration itself has become a regulatory flashpoint. Data-integrity audits in manufacturing labs increasingly scrutinize how peaks are integrated and whether manual adjustments to baselines are justified. The challenge is that most data-processing software was designed for chromatographic peaks, which tend to sit on flat baselines, not for electropherograms, which often do not. This mismatch means analysts spend time manually correcting integrations, and regulators want documentation that those corrections are scientifically defensible rather than arbitrary.

High-Throughput and Microchip Formats

Conventional CE-SDS is already faster than SDS-PAGE, but for screening applications where thousands of samples need to be evaluated, even 40 to 50 minutes per sample adds up. Microchip-based CE-SDS platforms compress that timeline dramatically. In one evaluation, microchip CE-SDS separated antibody samples in about 41 seconds per injection, roughly 70 times faster than the conventional capillary format, while still providing sufficient resolution and sensitivity for product-quality screening.¹⁵ These microchip platforms are widely used in early-stage antibody development, where labs may need to rank hundreds of cell-culture clones by the quality of the antibody they produce. The trade-off is that microchip results are not always directly interchangeable with conventional CE-SDS data, so companies typically use microchips for screening and reserve conventional CE-SDS for formal release testing.

Coupling CE-SDS with Mass Spectrometry

One of the technique’s longstanding limitations is that it tells you a protein fragment’s apparent size but not its identity. You see a peak, and you know it migrates like a 25 kDa species, but you do not know what that species actually is without additional work. Linking CE-SDS to mass spectrometry (MS) would solve this, but SDS itself is a notorious problem for MS instruments, suppressing ionization and fouling source components.

Researchers have developed workarounds. One approach injects positively charged surfactants and organic solvent into the capillary after separation, creating conditions that strip SDS from the proteins before they reach the mass spectrometer.¹⁶ A more elaborate platform uses two-dimensional capillary electrophoresis: the first dimension separates proteins by CE-SDS as usual, then individual peaks of interest are transferred via a switching valve into a second capillary running SDS-free capillary zone electrophoresis, which feeds into the mass spectrometer for identification.¹⁷ These hyphenated setups remain specialized tools rather than routine methods, but they represent a clear direction for the field: combining the separation power of CE-SDS with the identification power of MS in a single workflow.

Where CE-SDS Fits in a Broader Testing Panel

No single analytical method gives a complete picture of a therapeutic protein. CE-SDS excels at detecting fragments and confirming the molecular weight of intact species, but it has blind spots. Size-exclusion chromatography (SEC) is better suited for detecting soluble aggregates under native (non-denaturing) conditions, because CE-SDS denatures everything before separation and may break apart weak aggregates in the process. A comparison of SEC and CE-SDS for monitoring hinge fragmentation in an IgG1 antibody concluded that combining both techniques provides the most comprehensive understanding of a product’s fragment profile.¹⁸ In practice, regulatory submissions for antibody drugs routinely include data from CE-SDS, SEC, ion-exchange chromatography, and peptide mapping, each probing a different aspect of the molecule’s quality.

Regulatory Validation

For a method to be used in drug manufacturing and release, it needs to be formally validated, meaning a lab has to demonstrate that it performs consistently and accurately for its intended purpose. An interlaboratory validation study involving multiple sites assessed CE-SDS’s precision, accuracy, linearity, range, and limit of quantitation for antibody purity testing. The method passed all pre-set criteria, establishing that it is fit for use as a size-impurity and size-heterogeneity assay for therapeutic antibodies.¹⁹ This kind of multi-site validation is what gives regulators confidence that different manufacturing facilities will get comparable results when testing the same product, and it is a major reason CE-SDS has become embedded in quality-control workflows across the industry.

Forced-degradation studies, where a drug product is deliberately stressed by heat, light, or chemical exposure to see how it breaks down, also lean on CE-SDS as a primary readout. Comparative forced-degradation work on biosimilar versus originator antibodies uses the technique to quantify product-related impurities such as aggregates and fragments under identical stress conditions.²⁰ When a biosimilar developer can show that its product degrades in the same pattern and at the same rate as the originator under stress, that is a strong piece of evidence for similarity, and CE-SDS provides the quantitative backbone for that comparison.