The vulnerability of glycomics is structural isomerism: two sugars with the same mass, but different linkages, anomers or branching can have dramatically different biological effects, sometimes opposing actions. Chromatographic, ion-mobility and chemical strategies, orthogonal methods that separate molecules based on shape, charge or reactivity prior to MS confirmation can untangle these analytes and turn ambiguity into regulatory suitable linkage knowledge.
Isomer separation is one of the fundamental issues facing glycan analysis. Glycans exist on multiple axes of structural variation including linkage position, anomeric stereochemistry, and branching order creating many more degrees of isomerism than their linear polymer counterparts DNA and protein. Sequence in DNA and protein means that each molecule has a unique linear arrangement of monomeric units. Glycans however can be constructed from identical monosaccharide building blocks into many different three-dimensional structures. Isomers that share the same mass will often have vastly different receptor binding profiles, clearance mechanisms or immunogenicities. Analysis of glycans that cannot distinguish between structures will be compositionally accurate but biologically meaningless.
Constant and variable regions of N-glycan structures.1,5
Structural complexity in glycans is not limited to elemental composition alone, as glycans have four forms of isomerism. As carbohydrates are stereoisomers, glycans cannot be distinguished based on mass alone. Also, unique linkages can form between monosaccharides when they connect through different hydroxyl groups. This is known as regioisomerism and results in different molecular shapes. Alpha versus beta anomeric pairs also exist at every glycosidic linkage, further multiplying the potential structural diversity. Branching patterns can also differ among glycans with identical compositions, leading to isomers with different antennarity. Combined, these features allow for two monosaccharides to form several isomeric disaccharides and complex N-glycans to form into several hundred isomeric structures that all share the same mass.
Three-dimensional isomers of glycans are often functionally relevant since many carbohydrate recognition systems display specificity that prohibits binding of certain glycan structures. α-2,3 and α-2,6 sialic acid linkage position isomers will preferentially bind to different viral hemagglutinins or selectins that mediate leukocyte trafficking; these positional isomers therefore affect viral infection and inflammation. α/β anomers can display different stabilities towards glycosidases, changing glycans half lives in circulation. Finally, branching isomers with different antennary types can affect avidity towards multivalent lectin targets and have been shown to have differing biological activity; for instance, biantennary and triantennary N-glycans can exhibit dramatically different immunological profiles. Because of this structural dependency on function, knowledge of composition alone is not enough to identify glycans because isomers of glycans can have drastically different functions but would be reported as the same structure if isomeric resolution is not achieved.
Table 1 Biological Significance of Glycan Isomerism
| Isomeric Feature | Biological System | Functional Impact | Clinical Relevance |
| Sialic acid linkage (α2-3 vs α2-6) | Viral binding and immune recognition | Determines host tropism and cell trafficking | Influenza infection susceptibility |
| Core fucosylation position | Antibody effector function | Modulates Fc receptor binding affinity | Therapeutic antibody efficacy |
| Branching architecture | Lectin-mediated clearance | Affects circulatory persistence | Pharmacokinetic optimization |
| Anomeric configuration | Enzymatic degradation | Determines metabolic stability | Drug half-life prediction |
Composition analysis techniques, such as intact mass spectrometry or elemental analysis, are inherently limited since they cannot determine glycan isomers. The inability to distinguish glycan isomers is perhaps carbohydrate chemistry's biggest weakness. Composition analysis does not distinguish how monosaccharides are linked or which of the possible glycosidic linkages are present, or branch ordering. This inherent limitation means that any observed composition could have dozens of possible isomeric forms that may vary widely in their biological properties, making the information provided by composition alone biologically irrelevant. Also since combinations of monosaccharides can yield the same mass (isobaric), such as a fucose+hexose combination vs. two pentoses, mistaken identity is possible when only the mass over charge ratio is measured. Structural analysis is required to overcome this limitation.
Glycans show structural isomerism in at least four different ways. For example, saccharides can be connected at different linkage positions. Also, monosaccharides can occupy different positions within a glycan chain. Branching patterns can vary within glycans (branching topology isomers) as can the stereochemical configuration of each linkage. Each type of isomer presents its own separation challenge. In addition, multiple types of isomerism may coexist in any biological glycan, making complete glycan characterization quite difficult.
Linkage isomerism refers to linkage of monosaccharides through different hydroxyl groups on the acceptor sugar. Linkage isomers are molecules that share the same composition but differ in molecular structure (regioisomers). Alpha-2,3/alpha-2,6 and alpha-1,3/alpha-1,4 linkage patterns on hexoses (simple sugars containing six carbon atoms) are common examples that create distinct structural isomers. One example involving sialic acids can involve alpha-2,3 linkage vs. alpha-2,6 linkage when binding to galactose. This difference can greatly affect the recognition of the sugar by a receptor protein such as a virus or dendritic cell. Glycan fragmentation is required for linkage determination by mass spectrometry.
Terminal monosaccharides or modifications at different antennae are examples of positional isomerism. The antennae positioning of a terminal residue on branched glycans produces topological isomers that have the same composition and branching order, but differ in 3D topology. For example, a terminal fucose can be linked to the reducing-terminal N-acetylglucosamine (core fucosylation) or to antennae (antennary fucosylation). These topological isomers can have profoundly different biological effects, while only differing slightly in mass. The addition of sialic acid residues to multiple antennae can also produce positional isomers that alter the net charge of glycans and recognition by proteins. Topological isomers often co-elute during glycan profiling and generate similar fragment ions, therefore techniques like specific lectin binding or chromatography methods are needed to resolve topological differences.
Topology or branching isomerism results in different arrangements of the same monosaccharide composition as they are partitioned into varying antenna arrangements. Well-known topological categories include biantennary, triantennary, and tetraantennary glycans. Additional topological differences can include details such as which antenna contains fucose or sialic acid, or whether specific oligosaccharide chains reside on the alpha or beta branch. The topology of glycans plays an important role in their function, influencing the avidity of lectin binding and spatial arrangement of functional groups. The number of possible topologies increases exponentially with increasing antennarity, resulting in complexity that requires two or more separation dimensions to resolve, often coupled with chromatography, ion mobility, and mass spectrometry (MS).
Another layer of complexity is added when considering anomeric isomers (α and β) that exist at every sugar ring. These stereoisomers have the same connectivity, but differ in the configuration at the anomeric center (carbon bonded to oxygen in two places). The α-anomer is generally more sterically crowded (equatorial), while the β-anomer generally less so (axial). This structural difference is important because enzymes recognize each sugar stereoisomer (glucose vs galactose vs mannose for example) and anomeric form (α vs β) differently. Since mass spectrometry does not directly detect stereochemistry, analysis by itself cannot differentiate between anomeric pairs or stereoisomers. Detection of isomeric forms requires chiral derivatization methods or separate enzymatic recognition using sugar specific glycosidases.
Typically, separation methods for glycan isomers remain difficult because all isomers have the same mass and very similar physicochemical properties. Therefore, mass spectrometry alone or retention times during chromatography are not able to differentiate between isomers. Instead, separation of isomers typically requires two or more dimensions that can take advantage of small differences between structures by using multiple separation techniques, derivatization, or fragmentation.
LODES and CID spectra of triantennary N-glycans.2,5
Because glycan isomers have identical elemental composition and molecular weight, mass spec-based techniques cannot distinguish between regioisomers, anomers, or linkage variants based on mass alone. Sialic acid linked α-2,3 vs α-2,6 to galactose will generate identical m/z's and very similar fragmentations during standard tandem MS experiments. Isomers also possess identical polarity, hydrophobicity, and charge so they cannot be distinguished by bulk physicochemical properties.
Front-end chromatographic separation is the most common approach to resolve isomers. Retention mechanisms of glycan isomers can be very similar. Although glycans can be separated using hydrophilic interaction chromatography (HILIC) based on size and charge, linkage isomers containing sialic acids can often co-elute or only be partially resolved using HILIC. Specialized methods such as linkage specific derivitization, using high column temperatures and capillary electrophoresis can be used to resolve these linkage isomers. Porous graphitized carbon chromatography can provide excellent resolution of glycans based on subtle shape differences in the isomers. However, even with these types of columns, resolution of some isomers such as larger biantennary structures may not be possible due to lack of selectivity between the small conformational differences.
Limitations of MS interpretation of glycans arise from intrinsic ambiguity of the MS data itself. Isomers give rise to ions of equal mass and so will not be distinguished from each other as precursor ions. The fragmentation spectra are dominated by glycosidic bond cleavages which are often similar for isomers regardless of linkage position or anomeric configuration. Although collision induced dissociation provides B and Y ions that encode sequence information, regioisomers may form virtually identical series of fragment ions, with only slight differences in relative abundances. These slight differences are also lost when isomers co-elute, resulting in a complex summed spectrum that cannot be deconvoluted. Peaks due to in-source fragmentation may also complicate spectra and lead to identification of these fragments as separate glycans.
Resolution of glycan isomers during chromatography is non-trivial. Since glycan isomers generally differ only slightly in their physicochemical properties, different chromatographic properties must be utilized to obtain full resolution of isomers on chromatography columns. Often times one dimension of chromatography is not enough to separate all isomers. Specific stationary phases, two-dimensional chromatography, and appropriate standards can aid in separation of glycan isomers. As always, trade-offs between highest resolution and robust/reproducible methodologies must be considered.
Orthogonal to mass spectrometry, HILIC separations are great for providing an overview of glycans by their size and charge but separation of structural isomers is inherently difficult. α-2,3 vs α-2,6 sialic acid linkage isomers often co-elute, or show only partial separation by standard HILIC methodologies, making biological interpretation unclear. Positional isomers, such as terminal modifications on different antennae also cannot be separated by HILIC alone. Separation of anomers is also limited. While derivatization strategies that report linkage information or MRM transitions that selectively report on one class of isomers can improve selectivity for these types of glycans, HILIC does not typically allow for complete linkage assignments without orthogonal confirmation or another separation mode.
The strongest method available for separation of glycan isomers is porous graphitized carbon chromatography (PGCC), which can separate linkage isomers, regioisomers, and anomers based on distinct shape-selective retention mechanisms provided by interaction with the planar graphitic surface. Extensive separation of linkage isomers (resolution of sialic acid linkage isomers not achieved by any other chromatography method), regioisomers of fucosylation and sulfation, and anomers can be obtained with PGCC. Challenges to using PGCC routinely include tailing/elusive peak shapes caused by unknown surface chemistry of PGCC, harsh conditioning protocols required to preserve column stability and performance, and overall lack of robustness. Improved retention time reproducibility has been recently achieved by normalizing retention with a dextran ladder and reporting system independent GU (Graphitized Unit) values. Despite advances in the methodology, PGCC use remains limited due to its complexity relative to other normalization techniques.
Pairs of orthogonal separation techniques can be joined together for multidimensional liquid chromatography with complete resolution of isomers. Commonly, two-dimensional liquid chromatography approaches first use anion exchange to separate glycans based on number of sialic acids and overall size. A second separation using HILIC can then separate out structures for structural analysis, or PGC for isomeric resolution. By separating out one class of glycans before further analysis, it may be possible to simplify samples before high resolution separations. On the other hand, while these approaches may offer the most complete separation available, they are technically challenging and require at minimum two liquid chromatography instruments in series or Offline fractionation.
Retention time libraries based on GU metrics provide an instrument-independent standardization for glycan identification in HILIC. When retention times are normalized to an internal standard such as dextran ladder, glycan retention times from different laboratories can be compared. GU metrics are specific for HILIC separations and can not be used with other separation modes such as PGC since the retention mechanisms are inherently different between separation modes. An independent GU value has been created for PGC separations by normalizing retention times internally to a dextran standard using post-acquisition normalization processing; however libraries for PGC separations are not as complete as those available for HILIC separations. GU metrics can not be used to predict retention times of untested glycans, including novel structures or glycans that do not behave predictably. Retention times will also vary between instruments running the same chromatographic method due to subtle differences in column and instrument performance.
Mass spectrometry (MS) has fragmentation capabilities that enable structural elucidation of glycan isomers. Glycan isomers can be separated physically by chromatographic methods. Tandem MS can differentiate glycans chemically by producing fragment ions that are diagnostic for differences in their linkage and configuration. Fragmentation patterns indicate how a linkage or configuration affect dissociation of the molecule, leaving a spectral fingerprint that can be used to annotate structures. Regular fragmentation however struggles to differentiate certain isomeric configurations, requiring further resolution.
Information from diagnostic fragment ions formed during tandem mass spectrometry is used to deduce linkage of glycan isomers. Fragment ions created from cleavages across ring bonds (commonly referred to as A- and X-type ions) are informative of linkage positions because they reveal what ring bonds were broken during fragmentation. Knowing where fragment ions broke can reveal where monosaccharides are attached to each other. Glycosidic bond cleavages result in B- and Y-type ions that verify sequence and branch connectivity but often do not reveal linkage ambiguity. Loss of neutral sugars or combinations that are common to a specific linkage isomer result in diagnostic fragmentation patterns.
Disadvantages with fragmentation-only approaches are however apparent in instances where isomers cannot be distinguished from each other confidently based on their fragment ion set al.ne. Mixture of isomeric species which co-elute during chromatography will produce summed spectra from more than one precursor molecule and thereby complicate analysis, especially when done manually. Linkage positional isomers and anomeric forms may often give rise to identical fragment ions upon CID fragmentation, leaving one unable to differentiate between glycan structures. For larger glycans that fragment at multiple sites this also becomes challenging. The low abundance of certain glycans can lead to poor signal to noise ratio.
Greater discrimination can be achieved using other fragmentation techniques that access different dissociation energies than typical collision-induced dissociation. These include electron-based dissociation techniques such as electron transfer dissociation or electron excitation dissociation which cause fragmentation via radical mechanisms and can produce characteristic cross-ring cleavages that complement those produced via collision-induced dissociation. Ultraviolet-induced photodissociation or infrared multiphoton dissociation introduces energy via photons and can cause distinct fragmentations that help differentiate between isomers. Stepwise or incrementally increasing collision energies can be used to maximize structural data. Ultraviolet photodissociation tends to produce more cross-ring cleavages.
Ion mobility mass spectrometry (IM-MS) has become an invaluable technique for the separation of glycan isomers. Ion mobility spectrometry separates molecules according to their gas-phase shape and size. Therefore it provides another orthogonal separation technique to chromatography and standard MS because it can separate species based on their three-dimensional structure, despite having the same molecular weight. The drift time it takes ions to travel through a buffer gas in an electric field can be used to calculate collision cross-sections (CCS), which are intrinsic physicochemical properties useful for identification.
Ion mobility separation physically accelerates ions through a drift region filled with neutral gas. They drift towards a detector under an electric field. The more frequent their collisions with gas molecules, the slower they will reach the detector. CCS describes the surface area that is available to collide with buffer gas molecules. Highly compact molecules have smaller CCS values and will reach the detector faster than extended conformers. Drift tube ion mobility measures drift times which can be translated into CCS using the Mason-Schamp equation. Ions are moved through the drift region by using static electric fields. Traveling wave ion mobility moves ions by applying a voltage wave. The wave essentially moves the ions through the drift region. Arrival times can be translated into CCS after calibrating with known mobility standards.
Collision cross section differences can be used to distinguish between isomeric glycans based on different conformations they may adopt in the gas phase. Position isomers, for example glycan structures that differ by the position of a linkage (alpha2,3 linked versus alpha2,6 linked sialic acid), may have unique CCS values due to different geometries of the glycan chain and the flexibility afforded by the glycosidic linkage. Alpha and beta anomers can be distinguished by their CCS due to the equatorial glycosidic linkage of beta anomers promoting a more extended glycan conformation relative to the axial glycosidic linkage of alpha anomers. Differences in branching (biantennary vs. triantennary) can also affect the CCS because of differences in overall shape. Differences in CCS can help unambiguously assign glycan structures when used in combination with retention time and fragmentation data.
Ion mobility measurements are currently capable of routinely resolving anomeric differences as well as many linkage position isomers; however, it is unable to completely resolve all linkage isomers due to very similar conformations. An example of this limitation is incomplete separation of some larger biantennary forms due to flexibility causing wide distributions of arrival times which often overlap isomers. Transmission loss in the mobility region also decreases sensitivity towards low-level glycans. While CCS predictions for complex glycans are improving, they are not yet accurate enough to unequivocally identify unknown glycans without comparison to a standard. Careful calibration and temperature stabilization are required to obtain reproducible CCS values, and coupling to chromatography introduces additional complexity that could affect analysis throughput.
Chemical approaches may also be taken to help distinguish structural isomers, such as employing enzymes (glycosidases) that hydrolyze specific monosaccharide linkages. Glycosidases have defined specificities for a given monosaccharide linkage (types of linkage), stereochemistry (α- or β-anomer), and position on the sugar chain. Glycosidases sequentially remove monosaccharides from the non-reducing end of oligosaccharides, allowing sequential removal of sugars to determine structure by identifying what is removed and what remains on the glycan. Exoglycosidase arrays or chains of exoglycosidases can be used to convert structural determination into a biochemical assay in addition to chemical and physical separation techniques and MS.
Exoglycosidase arrays are another useful tool for obtaining linkage information. Glycans are incubated with sets of linkage specific enzymes either in parallel or in a combinatorial manner, and resulting mass differences or RT differences are used to determine terminal sugars, and linkages. For example, incubating glycans with individual beta-galactosidases (one specific for beta-1,4 and one specific for beta-1,3 linkages) will result in a mass shift if the glycan has a terminal beta-linked galactose. Since an array analyzes glycans with many enzymes at once, this allows simultaneous testing of many hypotheses about sugar linkage and structure. Quickly comparing which enzymes digested a glycan and which failed to digest it allows identification of linkages such as sialic acid linkages, fucose attachment, and terminal galactoses that make up biological epitopes.
Stepwise degradation methods: Enzymatic digestion methods progressively treat glycans with exoglycosidases. After each step, terminal residues are cleaved off allowing for the determination of glycan structure based on the sum of mass differences and chromatographic changes. Each enzymatic digestion results in a puzzle where the identity of the terminal residue can be deduced. Continuing the digestion steps will allow progressively inner parts of the glycan structure to be elucidated. Glycan structure can be determined by matching changes in retention time with each digestion step to a predicted list of possibilities. The order of enzyme use is important to ensure that any residues that would block enzymes from cleaving other residues are removed before that digestion step is performed.
There are several pitfalls associated with enzyme mediated approaches. The specificity of enzymes used is not always absolute and many glycosidases can act on more than one linkage type (cross-peak reassignment). Branching can cause enzymes to be unable to cleave certain sugars on a glycan (false negatives). Some exoglycosidases have broader specificity than previously thought and are able to cleave different types of linkages (ambiguous structures). Resistance to cleavage by an enzyme can result from modification of the glycan or steric issues rather than expected linkages on the glycan. Finally, there are glycan classes that are not susceptible to any enzyme treatments. These include many O-glycans which do not have a uniform core (regions of glycans that cannot be probed biochemically). These pitfalls can be overcome by selecting the appropriate enzymes and controls, as well as corroborating with orthogonal techniques.
Table 2 Enzymatic Approach Challenges and Solutions
| Challenge Category | Specific Issue | Impact on Analysis | Mitigation Strategy |
| Specificity overlap | Enzymes recognize multiple linkages | Ambiguous structural assignment | Use of highly specific enzymes with validation |
| Incomplete digestion | Steric hindrance from branching | Underestimation of susceptible linkages | Extended incubation or alternative enzymes |
| Modified substrate resistance | Substitutions block enzyme access | False negative results | Chemical modification analysis or alternative methods |
The only way to ensure accurate assignment of glycan isomers is to combine orthogonal techniques so that evidence for a glycan's structure can be verified in multiple, independent ways. Liquid chromatography, mass spectrometry, ion mobility spectrometry, and exoglycosidase digestion all provide different pieces of information that can be used together to unequivocally identify a glycan. If the same structure is deduced by each of two or more techniques, then it is very likely the correct structure. However, if one or more techniques provide evidence for a different structure, there may be ambiguity.
Liquid chromatography/tandem mass spectrometry/ion mobility spectrometry/enzymatic digestion provide orthogonal structural information. By separating (1) chromatographic retention, (2) tandem fragmentation, (3) ion mobility conformation, and (4) enzyme susceptibilities into complementary pieces of information, LC/MS(/MS)/IMS/EDI narrows the range of possible structural interpretations more than any one technique alone. Chromatographic retention separates isomers and provides relative glucose unit values that correspond to size and charge. Tandem mass spectrometric fragmentation provides sequence and linkage information based on diagnostic fragment ions. Ion mobility separates gas-phase conformers based on their CCS. Enzymatic digests identify the presence of known epitopes based on known mass shifts in the fragments. When each piece of information supports the same structural interpretation, confidence in that assignment is increased.
Another approach would be to implement a confidence scoring system, which ranks structural assignments to one of four levels that clearly indicate how much information is known about that structure. At level 1, the assignment would only specify composition (e.g. elemental formula). At level 2 the glycan class is confidently assigned, and some structural features can be inferred from biosynthetic considerations. Level 3 assignments require supporting evidence for linkages positions/anomericity (e.g. based on fragmentation pathways/chromatographic properties). Finally, level 4 represents unambiguous assignment of stereochemistry, confirmed by orthogonal evidence such as NMR or X-ray. Adoption of scoring systems like this would ensure users of glycan databases and literature are aware of the supporting evidence for each annotation.
Annotation artefacts may arise if sparse spectral evidence is used to make provisional assignments that are then reported as confidently identified. This would allow inaccurate structural information to become propagated in databases and publications. In practice over-assignment may happen if assignment is made based solely on retention time correlation, or if fragment ions are "bent" to fit proposed structures, or if predicted masses are accepted without experimental corroboration. Annotation can be improved by correctly applying confidence levels to data, honestly reporting ambiguity when known, avoiding confirmation bias from spectral database searching and recognizing when only compositional information can be reported. The importance of under- rather than over-posting has been recognized by the MS community.
The issue of glycan isomerism is critical with regards to both quantitation and interpretation because most techniques cannot differentiate between different isomers of glycans. For example, if one wished to quantify a glycan using mass spectrometry or chromatography and the instrument does not resolve all of the isomers from one another, the quantitative result will be a sum of all of the isomers. Each isomer may have different biological activity so if one intends on using glycan profile to predict biological response, monitor drug activity, or define specifications the fact that glycans exist in isomeric forms must be considered.
Ideally quantitative analysis should be isomer specific, meaning there is no bias towards detecting one structural isomer over another. Deviations from this perfect situation are referred to as quantitative bias. Certain linkage isomers of sialic acid like alpha-2,3 vs. alpha-2,6 may show different ionization efficiencies during MS analysis or differences in chromatographic recovery. If chromatographic or mass signal separation of the linkage isomers is incomplete then the area under the combined peak will be affected by quantitative bias. It is then more difficult to compare relative abundances of isomers between different samples especially if they have been analyzed on different instruments or in different labs. Two glycans may appear to be the same but contain different isomers in differing amounts that could have different biological significance.
Two glycans with the same elemental composition can have vastly different properties if they are isomers of each other. The position of sialic acid affects hemagglutinin binding affinities to determine whether an influenza virus will preferentially infect a human or animal host. Fucosylation of the core of antibody glycans can greatly affect binding to Fc receptors and resulting cellular responses. Glycan branching can affect plasma half-life by altering affinity for hepatocyte asialoglycoprotein receptors. Because of differences like these, summed quantity measurements that do not differentiate isomers may overlook important functional information or result in established quality metrics that do not correlate with clinical outcomes.
Determining glycans isomers is of importance for biopharmaceutical quality control purposes. Specifications for quality control tests have to guarantee that the manufacturing process will ensure drugs with the same therapeutic effect from batch to batch. Since glycan biomarkers measured by quality control assays often lack the specificity to measure the contributing isomers associated with effector function or clearance, linking quality attributes to these activities may result in control of a surrogate that may not assure control of the actual functional attribute. When developing biosimilars it is necessary to show that the isomeric structure of glycans are the same as the originator/reference product and as such requires analytics beyond what may be used for quality control purposes to demonstrate structural comparability. Isomeric distributions can also change if process conditions alter the enzymatic glycosylation activities, even if the overall glycan distribution remains unchanged. This can lead to a quality gap if the methods approved for quality control do not measure these changes.
Isomeric resolution need must be weighed against application requirements, resources, and throughput limitations. In some glycan profiling workflows, the application may not require true isomeric resolution. An example might be the simple compositional tracking typically needed for quality control purposes. However, studies requiring mechanistic detail or glycan-based biosimilar comparability exercises will require complete structural characterization including linkage placement and anomericity. Workflow solutions can be envisioned to selectively apply the necessary levels of information depending upon the need.
When deciding whether linkage-specific resolution of isomers is required for a given application consider if detailed structural information is needed or if compositional information is adequate. For example analytical coverage for routine release testing of existing therapeutics might only need coverage by methods such as HILIC profiling that report on overall distribution of glycoform populations without resolving linkage isomers. Biosimilar characterization, profiling of lead candidates for development, or discovery of disease related biomarkers require linkage specific isomeric resolution by PGC chromatography, ion mobility or multi-dimensional methods to determine quality attributes or disease modifying structures. This approach allows optimal allocation of analytical effort based on need and avoids overburdening simple applications.
Methods for analysis of glycans obtained from glycoproteins typically allow some customization based upon desired throughput, complexity of sample etc. Applications which are discovery-based and may target unknown glycans/disease pathways are generally aimed at highest structural fidelity possible, even at the cost of throughput and ease of use in order to differentiate isomers. Validated methods with a goal of controlling attributes of interest (critical quality attributes) for Biopharmaceutical production need to strike a balance between structural fidelity and providing evidence of controlled reproducibility. Quality control methods aimed at regular batch release need high throughput and ease of use, with some tradeoffs in structural fidelity. Regulatory requirements may also limit methods to those that have been well validated and shown to be reproducible between labs.
Trade-offs among analytical resolution, cost and throughput are inherent in choosing an approach for resolving isomers. Techniques such as PGC chromatography, ion mobility separations, and two-dimensional separations allow complete isomeric resolution, but require expensive instrumentation, significant expertise, and long run times precluding high throughput analysis. Standard HILIC separation trades resolution of isomers for speed and inexpensive, robust analysis that can be used for high throughput sample screening. Instrument and expertise costs must be weighed against the benefits of isomeric resolution when choosing how to approach analyzing separable isomers. In some cases it may make sense to use higher resolution techniques for critical decision making steps, or failure troubleshooting while relying on more rapid techniques for routine analysis.
There are several challenges currently limiting glycan analysis that directly affect structural determination of isomers and thus our ability to completely understand their biology. Separation techniques available today are often unable to fully resolve complex mixtures of isomeric glycans found in biological samples. Standardization remains an issue when comparing results from different laboratories and instruments. Lastly, we are often unable to infer function from detailed structural isomer information.
Complex matrices such as biological samples are difficult to fully resolve into their individual isomers because of the huge combinatorial space of possible glycans and trace components that can still be biologically active. No chromatography technique will baseline resolve all possible isomers in a mixture, even multidimensional separations can elute structural isomers. MS will often fail to differentiate isomers with similar fragmentation, and ion mobility may not fully resolve flexible structures. We may think these features are resolved but could be missing subtle glycan species that alter biology and could play a role in disease or pharmaceutical responses.
One of the major obstacles towards standardization for glycans isomer measurements lies in retention time, CCS value, and fragmentation patterns variations that are strongly method-dependent between instruments and laboratories. Indices such as glucose unit index can standardize HILIC separation but are not directly transferable to other chromatography modes or ion mobility devices. Further complicating quality assurance and validation is the limited availability of certified reference materials with well-characterized isomeric distributions. Variability in column batches, mobile phase composition, and instrument tuning parameters contribute to systematic discrepancies that hinder direct comparisons and regulatory alignment, requiring significant inter-laboratory study efforts.
Isomeric resolution does not necessarily help predict biological roles even if analytical approaches can distinguish between the fine structure of glycans with high resolution. For example, if one can distinguish between alpha-2,3 and alpha-2,6 linked sialic acids using an analytical approach, it does not mean that this information can predict how these linkages will affect a disease state or be exploited for therapeutic purposes. Many structure-function studies are correlative and glycan structures can exist in various combinations making it difficult to predict biological properties based on analytical results. Glycosylation is also cell condition dependent so knowing the relative isomeric structure in vitro may not be useful for determining biological relevance.
Table 3 Bridging Analytical and Biological Understanding
| Gap Category | Analytical Capability | Biological Understanding Need | Bridging Strategy |
| Static vs dynamic analysis | Snapshot of isomeric composition | Temporal and contextual functional information | Longitudinal studies with functional assays |
| Structure-function correlation | Detailed structural assignment | Mechanistic role of specific isomers | Integration with receptor binding and cell-based assays |
| Predictive capability | Comprehensive isomeric profiling | Prediction of biological outcomes | Machine learning and systems biology approaches |
The hallmark of glycans that sets them apart from other classes of polymers is structural isomerism. Linkage position, anomeric state, and branch order can lead to multiple structures with the same composition. Resolution of isomers therefore requires thoughtful application of advanced analytical tools that can provide structural information beyond monosaccharide composition. Orthogonal techniques such as chromatography, mass spectrometry fragmentation, ion mobility spectrometry, and enzymatic assays can be employed together to resolve biologically significant isomers. While complete structural resolution is currently impossible, appropriate pairing of the depth of analysis with the goal of the study—be it research or biopharmaceutical—ensures that glycan analyses will provide useful information. Understanding that compositional data is functionally incomplete motivates the glycoscience community to work towards methods that resolve structural isomers.
Structural isomerism remains one of the most challenging aspects of glycan analysis. Glycans with identical compositions can differ in linkage, branching, or stereochemistry—differences that may significantly impact biological function, therapeutic efficacy, or regulatory assessment. Resolving these isomers requires more than standard mass measurement; it demands carefully integrated analytical strategies and expert structural interpretation. Our advanced glycan structural analysis services are designed to address isomer-level complexity using orthogonal technologies and validated workflows tailored to research and biopharmaceutical applications.
No single analytical platform can fully resolve all glycan isomers. We implement integrated workflows that combine complementary techniques to increase structural confidence, including:
By aligning separation mode, fragmentation method, and validation approach with the specific analytical objective, we reduce ambiguity in isomer assignment and improve reproducibility across studies. These workflows are particularly valuable in:
Even with advanced instrumentation, confident structural annotation requires expert interpretation. Glycan isomers may generate similar MS/MS spectra, overlapping chromatographic peaks, or ambiguous fragmentation patterns. Our structural interpretation services include:
We emphasize transparency in structural annotation, helping clients understand both the strengths and limitations of the analytical data. For biopharmaceutical applications, this level of rigor supports glycosylation characterization as a critical quality attribute (CQA) and strengthens regulatory documentation.
If your project requires advanced glycan isomer resolution, structural confirmation, or expert interpretation of complex glycan data, our team provides customized analytical solutions tailored to your research or biopharmaceutical objectives. Contact us to discuss your sample type, required structural depth, and timeline for comprehensive glycan structural analysis.
References:
