The core of glycoprotein structural characterization does not lie in confirming a single chemical entity, but in resolving a multilayered information system jointly defined by the protein backbone, glycosylation sites, site occupancy, glycan composition, branching patterns, terminal modifications, charge heterogeneity, population state, and higher-order structure. Because glycosylation is not a directly template-encoded process, a given glycoprotein usually exists as a heterogeneous population of multiple glycoforms. Accordingly, its quality attribute analysis cannot rely on a single technique, but instead requires a systematic framework built on orthogonal methods for structural confirmation, heterogeneity analysis, and structure-function attribution.
Keywords: glycoprotein; glycosylation; quality attributes; critical quality attributes; released glycan analysis; glycopeptide analysis; intact mass spectrometry; higher-order structure; multi-attribute method; comparability study
I. Methodological Positioning of Glycoprotein Structural Characterization
1.1 Core analytical targets in glycoprotein analysis
(1) Protein backbone information
This level mainly includes amino acid sequence, terminal structures, disulfide bond connectivity, and backbone-associated chemical modifications. The analytical objective is to confirm molecular identity and exclude truncation, mismatch, oxidation, deamidation, and other abnormalities at the backbone level.
(2) Glycan structural information
This level mainly includes N-glycan and O-glycan types, the locations of glycosylation sites, site occupancy, glycan composition and branching patterns, as well as terminal or subterminal features such as sialylation, galactosylation, fucosylation, and high-mannose content. The analytical objective is to define the glycoform space of the glycoprotein and the distribution of site-specific heterogeneity.
(3) Conformational and population-level information
This level mainly includes charge variants, aggregates, fragments, higher-order structural stability, and local dynamic changes. The analytical objective is to determine whether glycan differences are further translated into differences in population behavior and conformational consequences.
1.2 Basic requirements for glycoprotein quality attribute analysis
(1) Analytical results must support hierarchical attribution
Glycoprotein analysis cannot stop at merely “detecting a difference,” but must further determine whether that difference occurs at the backbone level, global glycan profile level, site-specific level, conformational level, or functional level. Only when hierarchical attribution is achieved does quality attribute evaluation acquire methodological significance.
(2) Analytical results must be linked to functional consequences
Differences in glycoforms do not necessarily constitute critical quality attributes. Only when glycan changes affect receptor binding, biological activity, stability, pharmacokinetic behavior, or immune-related risk do such differences enter the scope of critical quality attribute discussion.
1.3 Orthogonal analysis principles in glycoprotein characterization
(1) No single technique can cover all structural information
Released glycan analysis is well suited for global glycan distribution, glycopeptide analysis is optimal for site-specific resolution, intact mass spectrometry is effective for overall heterogeneity profiling, higher-order structural analysis is useful for conformational and dynamic information, and functional assays establish biological relevance. These techniques are complementary rather than interchangeable.
(2) Method combinations should serve the specific analytical question
If the goal is comparison of batch-to-batch glycan trends, released glycan analysis and charge variant analysis are often more efficient. If the goal is site-specific attribution, glycopeptide analysis is central. If the goal is overall profile comparison before and after a manufacturing change, intact molecular and subunit analysis is more efficient. The analytical system must be designed around the question rather than assembled mechanically.
II. Hierarchical Framework of Critical Quality Attributes in Glycoproteins
2.1 Primary structure and backbone-related attributes
(1) Sequence and termini
These are used to confirm backbone identity, processing completeness, and potential truncation, and they constitute the foundational layer of glycoprotein characterization.
(2) Disulfide bonds and backbone modifications
These are used to confirm the structural basis of molecular folding and backbone-associated chemical variants, thereby avoiding the misassignment of backbone abnormalities as glycan-related effects.
2.2 Glycan-related attributes
(1) Global glycan profile
This includes the proportion of high-mannose glycans, the proportion of complex glycans, the degree of galactosylation, the level of sialylation, the level of core fucosylation, and the overall glycan profile distribution.
(2) Site-specific glycosylation
This includes glycosylation site localization, site occupancy, and site-specific glycoform distribution. This level is one of the most critical analytical targets for structure-function attribution.
2.3 Population heterogeneity and conformational attributes
(1) Charge variants and size variants
These are used to evaluate the impact of glycan modifications, charge modifications, aggregation, and fragmentation on sample homogeneity.
(2) Higher-order structure and stability
These are used to evaluate whether glycan differences alter local conformation, surface exposure, flexibility, thermal stability, and receptor-binding interface states.
Table 1. Critical Quality Attributes of Glycoproteins and Their Main Analytical Levels
Structural Level | Key Attributes | Main Analytical Questions | Representative Methods |
Primary structure | Amino acid sequence, termini, disulfide bonds | Is the backbone correct, and are mismatch or truncation present? | Peptide mapping, LC-MS, reducing and non-reducing analysis |
Global glycan profile | Glycan composition, branching, terminal modification | How is the overall glycan distribution, and is glycan drift present? | Released glycan analysis, HILIC-FLD, CE-LIF, LC-MS |
Site-specific glycosylation | Site location, occupancy, site-specific glycoforms | Which sites are glycosylated, and has site-specific heterogeneity changed? | Glycopeptide LC-MS/MS |
Population heterogeneity | Charge variants, aggregates, fragments | Are variant populations present that affect consistency? | cIEF, IEX, SEC, CE-SDS |
Higher-order structure | Conformational stability, local dynamics | Do glycan differences affect folding and stability? | CD, DSC, DSF, HDX-MS |
Functional attributes | Binding activity, biological activity, effector function | Are structural differences translated into functional differences? | Receptor binding assays, cell-based activity assays |
III. Sample Pretreatment and Analytical Strategy Design
3.1 Basic principles of pretreatment
(1) Preserve the integrity of glycan and backbone information
Pretreatment of glycoproteins should not solely pursue sufficient solubilization and denaturation, but must simultaneously control for sialic acid loss, glycan release, disulfide rearrangement, increased deamidation, and nonspecific degradation.
(2) Maintain compatibility with downstream analytical platforms
Released glycan analysis, glycopeptide analysis, intact mass spectrometry, and higher-order structural analysis differ in their tolerance to buffers, salt concentrations, surfactants, and reducing agents. Pretreatment must therefore be matched to the downstream platform.
3.2 Pretreatment priorities for different methods
(1) Released glycan analysis
The emphasis is on fully exposing glycosylation sites and ensuring efficient enzymatic release of glycans. The degree of denaturation, reduction state, and completeness of enzymatic digestion must be carefully controlled.
(2) Glycopeptide analysis
The emphasis is on achieving controlled digestion, preserving glycopeptide integrity, reducing artificial modification background, and improving both site coverage and glycopeptide enrichment efficiency.
(3) Intact molecular and higher-order structural analysis
The emphasis is on preserving the overall conformation and population distribution as much as possible while minimizing strong denaturation and the introduction of complex nonvolatile components.
IV. Core Methodological Systems for Glycoprotein Structural Characterization
4.1 Released glycan analysis
(1) Methodological positioning
Released glycan analysis is used to obtain information on the overall glycan distribution and is suitable for monitoring global changes such as high-mannose content, galactosylation, fucosylation, and sialylation.
(2) Analytical value
This method offers relatively high separation efficiency and good throughput, making it suitable for batch comparison, process development, stability trend monitoring, and overall glycan profile analysis.
(3) Methodological boundaries
Released glycan analysis does not retain site information and therefore cannot independently support site-level attribution. For glycoproteins with multiple glycosylation sites, the results are more suitable as global profile information rather than final attribution evidence.
4.2 Glycopeptide analysis
(1) Methodological positioning
Glycopeptide analysis preserves both peptide sequence information and site-retained glycan information, making it the core method for resolving site-specific glycosylation.
(2) Analytical value
This method is well suited for deep characterization during development, manufacturing change studies, comparability analysis, and structure-function attribution studies. Its interpretive power for changes in glycoforms and occupancy at key sites is significantly stronger than that of released glycan analysis.
(3) Methodological boundaries
Glycopeptide analysis is highly dependent on digestion efficiency, glycopeptide enrichment strategy, fragmentation mode, and data interpretation algorithms. If site coverage is insufficient or low-abundance glycopeptide response is unstable, local observations should not be overgeneralized.
4.3 Intact molecular and subunit mass spectrometry analysis
(1) Methodological positioning
Intact mass spectrometry is used for rapid observation of the overall mass distribution and macroscopic heterogeneity of glycoproteins, whereas subunit analysis improves resolution while preserving structural context.
(2) Analytical value
These methods are suitable for overall profile comparison, rapid screening before and after process changes, and assessment of subpopulation distributions in antibodies and fusion proteins.
(3) Methodological boundaries
Intact molecular profiles are suitable for comparison but do not equal site-level resolution. If a shift in overall mass distribution is observed, released glycan or glycopeptide analysis is still required for structural attribution.
4.4 Peptide mapping and multi-attribute methods
(1) Methodological positioning
Peptide mapping is used to confirm backbone sequence, local modifications, and partial information related to glycosylated peptides. Multi-attribute methods are based on LC-MS peptide maps and allow simultaneous monitoring of multiple quality attributes within a single workflow.
(2) Analytical value
These approaches are suitable for primary structure confirmation, backbone modification monitoring, and the integration of analytical attributes from development into routine control.
(3) Methodological boundaries
The core of multi-attribute methods lies in quantitative robustness and consistency of assignment, not in the number of peaks generated. If assignment, thresholds, and reproducibility have not been established, the method should not directly support judgments regarding critical quality attributes.
4.5 Charge variant analysis
(1) Methodological positioning
cIEF, iCIEF, and IEX are mainly used to evaluate charge heterogeneity and are especially suitable for monitoring distribution changes caused by sialylation, deamidation, and other charge-related modifications.
(2) Analytical value
These methods are suitable for evaluating batch consistency, performing stability studies, analyzing sialylation-related trends, and supporting routine quality monitoring.
(3) Methodological boundaries
Changes in charge variants do not necessarily equal changes in glycan structure. If abnormal charge peaks are observed, released glycan analysis, glycopeptide analysis, and backbone modification analysis should be used together for interpretation.
4.6 Size variant analysis
(1) Methodological positioning
SEC and CE-SDS are mainly used to analyze aggregates, fragments, and incompletely assembled species, thereby evaluating population homogeneity and stability-related risk.
(2) Analytical value
These methods are suitable for process development, stability studies, stress testing, and consistency evaluation.
(3) Methodological boundaries
Size variant methods reflect population state rather than glycan structure itself. If aggregation or fragmentation increases, further attribution at the backbone, glycan, and conformational levels remains necessary.
4.7 Higher-order structural analysis
(1) Methodological positioning
CD, DSC, DSF, and HDX-MS are used to evaluate whether glycan changes cause local conformational perturbation, altered thermal stability, or changes in surface dynamics.
(2) Analytical value
These methods are suitable for structure-function attribution, monitoring conformational drift, and assessing the impact of manufacturing changes on molecular stability.
(3) Methodological boundaries
Higher-order structural methods do not directly provide glycan compositional information, but rather evaluate the conformational consequences of glycan differences. They are therefore usually interpreted in combination with glycan-level methods.
4.8 Functional analysis
(1) Methodological positioning
Functional analysis is used to determine whether glycan differences are translated into altered receptor binding, shifted biological activity, differences in effector function, or stability-related functional consequences.
(2) Analytical value
These methods are suitable for confirmation of critical quality attributes, comparability studies, and mechanistic validation.
(3) Methodological boundaries
Functional analysis represents the endpoint of the attribution chain rather than the sole line of evidence. Without structural data, the source of functional differences remains difficult to define.
Table 2. Analytical Positioning of Major Methods for Glycoprotein Structural Characterization
Method Category | Main Analytical Level | Main Question Addressed | Main Advantages | Main Limitations |
Released glycan analysis | Global glycan profile | Has the overall glycan profile changed, such as high-mannose content, galactosylation, sialylation, or fucosylation? | High sensitivity, good throughput, suitable for batch comparison and trend analysis | Does not retain site information and cannot independently complete site-level attribution |
Glycopeptide analysis | Site-specific glycosylation | Which site is glycosylated, what is the occupancy, and has the site-specific glycoform changed? | Provides both peptide sequence and site-retained glycan information; strong attribution capability | Complex sample preparation, high data analysis threshold, and strong dependence on coverage and algorithms |
Intact mass spectrometry | Overall heterogeneity | Has the overall mass distribution changed, and are glycoform clusters or macroscopic heterogeneity shifted? | Allows rapid observation of the overall profile and is suitable for quick comparison before and after changes | Limited structural resolution and difficult to directly provide fine site-level information |
Subunit mass spectrometry | Subunit-level heterogeneity | Are there differences in glycoforms and mass distribution at the domain or subunit level? | Balances overall context and resolution; suitable for complex glycoproteins | Still requires combination with glycopeptide or released glycan analysis for fine attribution |
Peptide mapping | Primary structure and local modifications | Is the backbone sequence correct, and are oxidation, deamidation, truncation, or other backbone-associated variants present? | Foundational method for backbone confirmation and multi-attribute monitoring | Limited coverage of overall glycan profile information |
Charge variant analysis | Population heterogeneity | Has the charge distribution changed, and are acidic or basic variant shifts present? | Suitable for monitoring sialylation-related changes and overall consistency | Charge variation has multiple sources and cannot be directly equated with glycan changes |
Size variant analysis | Population state | Are aggregates, fragments, and incompletely assembled species present? | Suitable for evaluation of stability and homogeneity | Does not directly provide glycan structural information |
Higher-order structural analysis | Conformation and stability | Do glycan changes cause local conformational perturbation, altered thermal stability, or changes in surface dynamics? | Suitable for linking glycan differences to conformational consequences | Does not directly answer questions of glycan composition or site localization |
Functional analysis | Functional attributes | Are structural differences translated into receptor binding, biological activity, or effector function differences? | Important endpoint evidence for confirmation of critical quality attributes | Cannot independently localize the structural cause |
Multi-attribute methods | Integrated attribute monitoring | Can glycosylation, backbone modifications, and other key attributes be monitored simultaneously in a unified workflow? | Facilitates connection between development and routine monitoring | High requirements for robustness, assignment consistency, and data processing |
V. Key Reagents for Glycoprotein Structural Characterization and Quality Attribute Analysis
Name | CAS No. | Applicable Method Category | Main Use | Methodological Significance |
PNGase F | N-glycan release analysis, released glycan profiling, global glycoform analysis | Specifically cleaves most N-linked glycans | Initiating enzymatic reagent for overall N-glycan analysis | |
Sialidase | Terminal glycan modification validation, released glycan confirmation, glycan function studies | Removes terminal sialic acid residues | Used for attribution of sialylation and confirmation of terminal capping | |
Trypsin | Glycopeptide analysis, peptide mapping, multi-attribute methods | Generates peptides and glycopeptides suitable for LC-MS analysis | Fundamental enzyme for site coverage and peptide map reproducibility | |
Lys-C | Glycopeptide analysis, peptide mapping | Assists or substitutes for trypsin to generate peptides of different lengths | Improves site coverage in complex proteins | |
Glu-C | Glycopeptide analysis, peptide mapping | Generates peptides with specific cleavage patterns | Used to supplement regions inadequately covered by trypsin | |
Urea | Released glycan pretreatment, glycopeptide pretreatment, peptide mapping | Denatures proteins and increases site exposure and digestion efficiency | Influences glycan release efficiency and digestion accessibility | |
Guanidine Hydrochloride | Released glycan pretreatment, glycopeptide pretreatment | Strong denaturing treatment to improve unfolding of complex glycoproteins | Suitable for difficult sample pretreatment | |
SDS | Released glycan pretreatment, protein denaturation | Enhances protein unfolding and improves glycan release efficiency | Strong denaturing capacity, but limited downstream compatibility | |
Triton X-100 | Released glycan pretreatment, digestion-compatible treatment | Mitigates inhibitory effects of SDS on certain enzymatic activities | Common auxiliary reagent in released glycan pretreatment systems | |
DTT | Glycopeptide analysis, peptide mapping, subunit analysis | Reduces disulfide bonds and improves digestion and subunit separation | Directly affects the quality of disulfide bond processing | |
TCEP-HCl | Glycopeptide analysis, intact molecular pretreatment, subunit analysis | Stably reduces disulfide bonds | Suitable for MS-compatible workflows | |
Iodoacetamide (IAA) | Glycopeptide analysis, peptide mapping | Alkylates free sulfhydryl groups to prevent disulfide reformation | Influences control of artificial modification background | |
Chloroacetamide (CAA) | Glycopeptide analysis, peptide mapping | Alternative alkylating reagent to IAA | Used to optimize alkylation-related side reaction background | |
N-Ethylmaleimide (NEM) | Disulfide bond analysis, nonreducing peptide mapping | Blocks free sulfhydryl groups and preserves native disulfide information | Suitable for disulfide linkage studies | |
2-Aminobenzamide (2-AB) | Released glycan fluorescent labeling, HILIC-FLD | Enhances fluorescence detection sensitivity for released glycans | Commonly used in routine released glycan profiling | |
APTS | CE-LIF released glycan analysis | Suitable for capillary electrophoresis fluorescence detection | Suitable for high-resolution migration analysis | |
Procainamide Hydrochloride | Released glycan fluorescent labeling, released glycan LC-MS | Simultaneously enhances fluorescence signal and part of the MS response | Suitable for platforms integrating FLD and MS | |
Sodium Cyanoborohydride | Released glycan reductive amination labeling | Used for labeling reactions with 2-AB, 2-AA, APTS, procainamide, and related reagents | Determines labeling efficiency and reproducibility | |
Ammonium Bicarbonate | Glycopeptide digestion, peptide mapping | Provides a volatile buffer system | Common LC-MS-compatible digestion buffer | |
Tris | Pretreatment, enzymatic digestion, structural studies | Common buffering system | Suitable for certain non-direct MS analytical workflows | |
Formic Acid | LC-MS mobile phase, sample acidification | Adjusts acidity and improves peak shape and ionization efficiency | Common volatile acidic additive | |
Trifluoroacetic Acid (TFA) | Released glycan separation, monosaccharide composition analysis, certain chromatographic methods | Adjusts acidity and retention behavior | Commonly used in methods not primarily intended for MS | |
Ammonium Acetate | HILIC-MS, glycopeptide LC-MS, native MS | Volatile salt system that maintains separation and MS compatibility | Influences retention and ionization behavior | |
Ammonium Formate | HILIC-MS, glycopeptide LC-MS | Volatile buffer salt | Used for fine adjustment of separation conditions | |
Acetonitrile | HILIC separation, LC-MS mobile phase | Adjusts organic phase proportion and improves retention of glycans and glycopeptides | Key mobile phase component in HILIC separation | |
D2O | HDX-MS, higher-order structural analysis | Used for hydrogen-deuterium exchange to characterize local conformational dynamics | Core reagent for acquisition of higher-order structural dynamic information |
VI. Data Integration and Method Validation in Quality Attribute Analysis
6.1 Logic of data integration
(1) Integration of global and site-specific data
When released glycan analysis indicates a glycoform shift, glycopeptide analysis is usually needed to identify the originating site and quantify the extent of the change.
(2) Integration of structural and functional data
If a site-specific glycoform change further induces changes in conformation, charge distribution, or binding activity, that difference is more likely to qualify as a critical quality attribute.
6.2 Key points of method validation
(1) Accuracy and reproducibility
Released glycan, glycopeptide, and peptide mapping methods all require evaluation of recovery, peak area reproducibility, site coverage, and batch-to-batch consistency.
(2) Separation capability and interpretive capability
Special attention should be paid to isomer separation, stability of peak assignment, completeness of fragmentation information, and consistency of data interpretation.
(3) Method applicability boundaries
Deep characterization methods used during development are not necessarily all suitable for transfer into routine control. Platform complexity, sample throughput, and feasibility for long-term execution must also be evaluated.
VII. Common Evaluation Metrics and Application Scenarios in Research
7.1 Common metrics at the structural level
(1) Glycan and site-related metrics
Common metrics include the number of glycosylation sites, site occupancy, abundance of major glycoforms, high-mannose proportion, degree of galactosylation, level of fucosylation, and proportion of sialylation.
(2) Population heterogeneity metrics
Common metrics include main peak purity, proportion of charge variants, proportion of aggregates, proportion of fragments, and intact molecular mass distribution.
7.2 Common metrics at the functional level
(1) Binding and activity metrics
Common metrics include receptor-binding activity, biological activity, enzymatic activity, cell-based effector function, and effector recruitment capacity.
(2) Stability metrics
Common metrics include thermal transition temperature, stress-induced degradation rate, aggregation tendency, and conformational retention capacity.
7.3 Typical research application scenarios
(1) Structural confirmation during development
The emphasis is on identifying glycoform space, confirming key sites, and establishing the integrated framework of backbone-glycan-conformation relationships.
(2) Process change and comparability studies
The emphasis is on comparing whether global glycoforms, site-specific glycoforms, charge variants, higher-order structure, and functional readouts remain consistent before and after the change.
(3) Mechanistic studies
The emphasis is on explaining how glycan differences affect conformation, receptor binding, pharmacological activity, and stability.
Methodological studies on glycoprotein structural characterization and quality attribute analysis should establish a hierarchically clear and attribution-focused analytical framework around the backbone, glycans, conformation, population heterogeneity, and functional consequences. For glycoproteins, the value of an analytical system is determined not by the number of technologies included, but by whether it can robustly support identification of critical quality attributes and judgment of structure-function relationships.
For more related articles, please see below:
[1] Glycoprotein Hormone Family: Molecular Features, Signaling Bias, and In Vivo Pharmacokinetics