Parkinson’s

Disease–Associated Remodeling of Synaptic N‑Glycans in the Human Prefrontal Cortex

Authors: Moyinoluwa Adeniyi, Mojibola Fowowe, Odunayo Oluokun, Sarah Sahioun, Vishal Sandilya, Oluwatosin Daramola, Andrew I. Bennett, Yehia Mechref

Abstract

Parkinson’s disease (PD) is characterized by progressive motor and cognitive dysfunction and is associated with synaptic pathology and impaired neurotransmission, including dysfunction of synaptic vesicles (SVs) and presynaptic terminals. However, the role of protein N-glycosylation within synaptic subcellular fractions remains understudied. Here, N-glycomics analysis was performed on synaptosomes and SVs enriched from the prefrontal cortex (PFC) of post-mortem PD and control brains using high-resolution LC-MS/MS. A total of 66 N-glycans were identified in synaptosomes and 68 in SVs, with PD-associated glycomics alterations showing clear fraction-specific patterns. PD synaptosomes exhibited reduced sialofucosylation alongside increased fucosylated and neutral glycans, including altered sialyl Lewis X (sLeX)-bearing structures, suggesting potential involvement in neuroinflammation and aberrant cell signaling. In contrast, PD SVs showed elevated high-mannose and neutral glycans. Isomer-resolved N-glycomics revealed distinct remodeling of glycan isomers in PD, characterized by altered branching, fucosylation, and sialylation in both fractions. Notably, PD synaptosomes displayed shifts toward highly branched fucosylated and sialylated isomers, while PD SVs exhibited changes consistent with impaired glycan maturation. Together, these findings demonstrate that PD induces distinct N-glycan alterations in PFC synaptic fractions, providing new insights into synaptic dysfunction associated with cognitive decline in PD.

Introduction

Parkinson’s disease (PD) is a progressive neurodegenerative disorder that primarily impairs motor function and, over time, significantly affects cognition. , The pathological hallmark of PD is the degeneration of dopaminergic neurons in the substantia nigra, a brain region essential for regulating voluntary movement. −

This neuronal loss gives rise to the characteristic motor symptoms of PD, including tremor, rigidity, bradykinesia, and postural instability. In addition to motor deficits, PD has been recognized as a multisystem disorder involving nonmotor symptoms such as cognitive impairment, mood disturbances, and autonomic dysfunction. A defining cellular feature of PD is the intracellular accumulation of misfolded α-synuclein protein in the form of Lewy bodies and Lewy neurites.

While the degeneration of dopaminergic neurons and formation of Lewy bodies are central features of PD, mounting evidence suggests that early synaptic dysfunction may precede overt neurodegeneration and play a pivotal role in disease onset and progression. , Specifically, presynaptic terminals and their associated molecular machinery are among the earliest structures to be affected, with disruptions in neurotransmission observed before measurable neuronal loss. , Synaptosomes, metabolically active, isolated presynaptic terminals, and synaptic vesicles (SVs), which store and release neurotransmitters, are valuable systems for investigating the molecular mechanisms underlying synaptic pathology in PD. −

SV dysfunction, driven by α-synuclein aggregation and impaired vesicle cycling, has been shown to disrupt synaptic signaling and homeostasis. , While intact synaptosomes are often used to study functional aspects of neurotransmission, synaptosomes also provide a biochemically enriched source of presynaptic components, including SVs, mitochondria, and membrane-associated glycoconjugates, making them highly suitable for glycomics and glycoproteomics analysis. This approach enables a deeper understanding of how α-synuclein pathology perturbs synaptic integrity through alterations in glycosylation patterns and glycan-mediated signaling.

Glycosylation is an essential post-translational modification that governs various aspects of protein folding, trafficking, receptor function, and intercellular signaling. In neurons, proper glycosylation is essential for maintaining synaptic integrity. N-linked glycosylation modulates the structure, trafficking, and function of key synaptic proteins, including voltage-gated ion channels, synaptic vesicle components, and neurotransmitter receptors. −

These glycan modifications are crucial for membrane localization and activity, influencing processes such as axon firing, synaptic vesicle release, and neurotransmitter signaling. Additionally, glycosylation supports astrocytic glutamate transport and unfolded protein response pathways, both of which are vital for preserving synaptic homeostasis and preventing neurodegeneration. , Disruptions in glycosylation machinery in PD have been associated with lysosomal dysfunction, impaired synaptic vesicle recycling, and altered localization of key synaptic glycoproteins such as synaptophysin and synaptic vesicle glycoprotein 2A (SV2A), which are known to carry complex fucosylated N-glycans critical for neurotransmitter release and synaptic plasticity. , Furthermore, changes in the expression of glycosyltransferases and sialyltransferases in PD have been linked to neuroinflammatory responses and sphingolipid metabolism dysregulation. , To investigate these glycosylation alterations in detail, advanced analytical platforms are essential.

High-resolution mass spectrometry (HRMS) systems have enabled detailed compositional and structural analyses of glycans, providing the sensitivity and accuracy necessary for comprehensive glycomics analyses of complex biological samples, including those derived from neurodegenerative diseases such as Alzheimer’s disease (AD) and PD. −

To complement HRMS and further improve structural resolution, the mesoporous graphitic carbon (MGC) columns have proven particularly effective for isomeric glycan separation, further enhancing the structural resolution of glycomic data sets. −

Studies on PD brain tissue, particularly within subcortical regions such as the substantia nigra and striatum, suggest that subtle alterations in glycosylation, including sialylation and fucosylation patterns, can significantly impact protein function, cellular communication, and neuroinflammatory processes. ,, However, to our knowledge, no study has yet examined N-glycan alterations specifically within the synaptic compartments of the prefrontal cortex (PFC) in PD.

The PFC is a critical region for executive function, decision-making, and working memory, which are cognitive domains frequently impaired in PD. Notably, the PFC undergoes dopaminergic disruption and α-synuclein pathology in PD, suggesting its direct involvement in disease progression and cognitive decline. The PFC was selected to investigate cortical involvement in PD at a stage when PD-related pathology is present but overt cognitive impairment and advanced neurodegeneration are limited. Although PFC involvement typically occurs at later Braak stages, subtle molecular and transcriptional alterations in cortical regions are thought to precede widespread neocortical Lewy body deposition and clinically apparent cognitive decline. , PD cases in this study exhibited Lewy body pathology consistent with cortical involvement, while Alzheimer’s disease-type pathology and clinical dementia were minimal, enabling investigation of PD-related cortical synaptic alterations rather than mixed or end-stage neurodegeneration. Despite this, the molecular underpinnings of PD in the PFC have yet to be fully elucidated. Given the crucial role of glycosylation in synaptic physiology, , characterizing glycan alterations in this region may provide novel insights into the mechanisms driving cortical synaptic dysfunction and cognitive impairment in PD.

In this study, we used high-resolution LC–MS/MS to profile glycans in synaptosomes and SVs isolated from the PFC of post-mortem PD brains. We comprehensively characterized both compositional and isomeric N-glycan features within these synaptic fractions to identify glycosylation changes associated with PD pathology. We hypothesized that PD-associated cognitive decline involves alterations in N-glycosylation specific to synaptic subfractions of the PFC. To our knowledge, this is the first glycomics characterization of synaptosomes and SVs from the PFC in PD.

Materials and Methods

Sample Information

In this study, post-mortem brain tissue samples from the prefrontal cortex (Brodmann area 9) were obtained from two sources. The Washington University Movement Disorders Center provided samples from 10 patients with PD. Control samples from individuals without PD were obtained from the University of California, San Diego Alzheimer’s Disease Research Center. This study was conducted in strict accordance with the ethical principles outlined in the Declaration of Helsinki and Texas Tech University guidelines for research involving human samples with IRB 504702. A summary of the sample demographics is presented in Table . PD cases were clinically diagnosed during life and confirmed at autopsy to exhibit Lewy body diseases, spanning Lewy body Braak stages 3–6, consistent with cortical involvement. Despite this, Alzheimer’s disease-type pathology was minimal and clinical dementia was rare, supporting analysis of PD-related cortical changes without significant AD confounding.

Chemicals and Reagents

4-(2-Hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), sodium hydroxide (NaOH), potassium hydroxide (KOH), ammonium bicarbonate (ABC), borane-ammonia, acetic acid, formic acid (FA), sucrose, dimethyl sulfoxide (DMSO), iodomethane, NaOH beads, dithiothreitol (DTT), iodoacetamide (IAA), and mesoporous graphitic carbon (MGC) material were procured from Sigma-Aldrich (St. Louis, Missouri, USA). Peptide: N-glycosidase F (PNGase F) enzyme was sourced from New England Biolabs (Ipswich, Massachusetts, USA), while difluoroacetic acid (DFA) was obtained from Acros Organics (New Jersey, USA). Halt Protease and Phosphatase Inhibitor Cocktail (100×), high-performance liquid chromatography (HPLC)-grade isopropyl alcohol (IPA), acetonitrile (ACN), methanol (MeOH), and water were acquired from Fisher Scientific (Fair Lawn, New Jersey, USA). Microcolumns were purchased from Harvard Apparatus (Holliston, Massachusetts, USA), and Isolute C18 (EC) cartridges were secured from Biotage (Charlotte, North Carolina, USA). The fused silica capillary was sourced from Polymicro Technologies (Phoenix, Arizona, USA), and the Kasil frit kit was obtained from Next Advance (Troy, New York, USA).

Synaptosome Enrichment

Synaptosome and SV enrichment from brain tissue was performed through a series of steps, including sucrose density gradient separation and ultracentrifugation, as illustrated in the workflow in Figure . The process began with the enrichment of synaptosomes from frozen brain tissue. First, 350 mg of frozen human brain tissue was excised and rinsed three times with homogenization buffer containing 320 mM sucrose and 5 mM HEPES, pH 7.4, adjusted with sodium hydroxide (NaOH). The tissue was then homogenized in 6 mL of homogenization buffer with 60 μL of a phosphatase-protease inhibitor using a tight 7 mL Teflon homogenizer with 15–20 strokes. The homogenate underwent an initial centrifugation at 900g for 10 min (min), after which the supernatant containing synaptosomes was carefully collected while the pellet was discarded. This supernatant was further centrifuged at 15,000g for 15 min, yielding a second supernatant fraction and a pellet fraction. The pellet contained synaptosomes and was carefully collected while avoiding the dark-brown center enriched in mitochondria and peroxisomes. Finally, the pellet was resuspended in 2 mL of ice-cold double-distilled water, and a 150 μL aliquot of the synaptosome suspension was collected, centrifuged at 10,000g for 10 min, resuspended in PBS buffer, and stored for subsequent synaptosome analysis.

Synaptic Vesicle Enrichment

The enrichment of synaptic vesicles (SVs) from the synaptosomal fraction was performed using differential centrifugation and sucrose gradient separation for subsequent analyses. First, the remaining synaptosomal fraction was adjusted to 6 mL with ice-cold double-distilled water (ddH2O), after which 30 μL of 1 M HEPES (pH 7.4, adjusted with NaOH) and 60 μL of a protease-phosphatase inhibitor were added and homogenized using a 7 mL Teflon homogenizer with seven to eight strokes. The lysate was then centrifuged at 48,000g for 25 min to separate the SV-containing supernatant from the pellet, which was discarded. To create a sucrose gradient, 4 mL of 700 mM sucrose solution was added to a centrifuge tube as the bottom layer, followed by a careful overlay of 1 mL of 300 mM sucrose solution. The 5 mL SV-containing supernatant was gently layered on top. The sample was then centrifuged at 132,000g for 1 h. After centrifugation, the top 6 mL was removed, leaving 4 mL containing SVs. These vesicles were pelleted by centrifugation at 300,000g for 2 h using a 50.2 Ti rotor. Finally, the supernatant was discarded, and the pellet containing SVs was resuspended in phosphate-buffered saline (PBS) and stored for subsequent processing.

Characterization of Synaptosome and Synaptic Vesicle by Transmission

Electron Microscopy (TEM)

A synaptosome sample was prepared for electron microscopy through a series of fixation, washing, dehydration, embedding, and cutting steps. The synaptosome sample was first fixed in 2.5% glutaraldehyde prepared in 0.05 M cacodylate buffer. Briefly, the synaptosome sample was transferred into 1.5 mL microcentrifuge tubes and centrifuged to form a pellet. After removing the supernatant, the tube was filled with the glutaraldehyde fixative, and the pellet was gently resuspended to ensure thorough fixation. The sample was fixed for 1 h at room temperature. Prior to embedding, the sample was washed three times in 0.05 M cacodylate buffer, 10 min each, to remove residual fixative. Postfixation was then performed using 1% osmium tetroxide for 40 min to 1 h at room temperature. Following osmication, the sample underwent three additional 10 min washes in cacodylate buffer. Dehydration was performed using a graded ethanol series (25, 50, 75, 85, 95, and 100%), with each step lasting 10–15 min, followed by two 15 min incubations in 100% acetone. Infiltration with EPON resin was performed using increasing resin-to-acetone ratios of 4, 1, and 1, with each step lasting 1 h. Finally, samples were incubated in 100% resin twice for 2 h each, embedded in fresh EPON, and polymerized in an oven. After polymerization, the embedded blocks were trimmed and sectioned using an ultramicrotome. Ultrathin sections were cut using a diamond knife and collected onto copper grids. The sections were then stained with uranyl acetate and lead citrate to enhance contrast and were imaged using a Hitachi H-7650 transmission electron microscope (Hitachi High-Tech America, Inc., USA) at an accelerating voltage of 100 kV.

For the SV, a suspension was applied dropwise onto carbon-coated grids and allowed to sit for 1 min. The excess sample was gently removed using lint-free filter paper, and the grids were then negatively stained with a freshly prepared 1% uranyl acetate solution for 30 s. After blotting off the excess stain, the grids were allowed to air-dry thoroughly. The dried grids were imaged with a Hitachi H-7650 transmission electron microscope.

Vesicle Lysis and Protein Quantitation

To prepare synaptosome and SV lysates, 400 μm zirconium beads were added to 2 mL microcentrifuge screw cap tubes. The synaptosome or SV samples were transferred into tubes containing the beads and mixed with 5% sodium deoxycholate (SDC) at a 1 sample-to-SDC ratio. The mixture was homogenized using a Beadbug Microtube Homogenizer at 4 °C, set to 400 rpm, with alternating cycles of 30 s of beating and 30 s of rest for six cycles. Following homogenization, samples were subjected to continuous sonication on ice slurry using a Branson 2510 sonicator bath for 1 h. The homogenate was kept immersed in an ice slurry bath throughout the procedure, and ice was replenished as needed to maintain a low temperature and prevent protein denaturation. After sonication, the lysate was centrifuged at 14,800 rpm for 10 min, and the supernatant containing the synaptosome or SV lysate was carefully collected. Finally, the protein content of the lysates was determined using the micro-BCA protein assay kit according to the manufacturer’s instructions.

Characterization of Synaptosome and Synaptic Vesicle by LC-MS/MS

Proteomics Analysis

Tryptic Digestion of Proteins

A bottom-up LC-MS/MS proteomics approach was employed for this characterization, involving the tryptic digestion of 50 μg of protein from synaptosome and SV lysate. The process began with denaturation in a 90 °C water bath for 15 min. Next, 200 mM dithiothreitol (DTT) (1/40 of the sample volume) was added, followed by incubation in a 60 °C water bath for 45 min. The addition of DTT reduces disulfide bonds between cysteine residues, thereby unfolding the protein and exposing cleavage sites for trypsin. Subsequently, 200 mM iodoacetamide (IAA) (four times the DTT volume) was added, and the sample was incubated in a 37 °C water bath for 45 min. IAA alkylates free thiol groups on cysteine residues after reduction by DTT, irreversibly modifying them to prevent the reformation of disulfide bonds. An additional 200 mM DTT (1/40 of the sample volume) was introduced, followed by incubation at 37 °C for 30 min. The pH was checked to ensure an optimal value of 8 for trypsin activity. Trypsin was added to the sample at a trypsin-to-sample ratio of 25 (w/w). The sample was then incubated at 37 °C for 18 h.

C18 Desalting of Tryptic-Digested Peptides

The tryptic-digested peptides were purified using a TopTip C18 column, prefilled with C18 material occupying one-third of its volume. Briefly, the column was washed three times with 50 μL of desalting solution B (60% ACN, 0.1% FA) and desalting solution A (water, 0.1% FA), each wash followed by centrifugation at 1000g for 1 min. The sample was then loaded, centrifuged at 500g for 40 s, and incubated for 2 min. The flow-through was reapplied and incubated again, and this process was repeated three times. An additional three-wash steps were carried out with desalting solution A, and the sample was eluted by adding desalting solution B and centrifuging at 2000g for 1 min, repeated three times. This was followed by a final elution step using desalting solution C (100% ACN, 0.1% FA), also centrifuged at 2000g for 1 min, and repeated three times. Finally, the collected eluates were dried using a SpeedVac vacuum concentrator, completing the sample preparation for LC-MS/MS analysis.

LC-MS/MS Analysis of Tryptic-Digested Peptides

The dried sample was reconstituted to a final concentration of 1 μg/μL in 2% ACN with 0.1% FA, and 2 μg of protein was injected for LC-MS/MS analysis. Peptide separation was conducted using a reversed-phase C18 Acclaim PepMap 100 Å capillary column (150 mm × 75 μm). The column oven temperature was maintained at 29.5 °C. A multistep gradient elution was applied over a total run time of 120 min at a flow rate of 0.350 μL/min. The MS acquisition parameters were configured on the Fusion Lumos Tribrid Orbitrap MS. The nanoESI source was set to positive ion mode with an applied voltage of 1600 V. The ion transfer tube temperature was maintained at 275 °C. Data-dependent acquisition (DDA) mode was employed with a full MS resolution of 120,000, a scan range of 400–2000 m/z, and the detector type set to Orbitrap. The Automatic Gain Control (AGC) target was set at 1 × 10,^6^ with an isolation window of 1.6 m/z. For the MS2 scan, the resolution was set to 30,000 using the Orbitrap detector, with higher-energy collision dissociation (HCD) as the activation type and a stepped collision energy of 15, 30, and 35. The AGC target for MS2 was set at 1 × 10,^5^ with an isolation window of 1.6 m/z.

LC-MS/MS Proteomics and N-Glycopeptide Data

Processing

The raw data files obtained from proteomics LC-MS/MS analysis were analyzed using Thermo Scientific’s Proteome Discoverer 3.2 software to detect and measure protein levels in synaptosome and synaptic vesicle (SV) fractions. The Sequest HT search engine was configured with the following cleavage sites at arginine (R) and lysine (K) on the C-terminal side, fully specific tryptic digestion, a maximum of two missed cleavages, and mass tolerances of 10 ppm for precursors and 0.02 Da for fragments. The maximum retention time shift was set to 10 min. Modifications include fixed carbamidomethylation on cysteine and variable modifications, including methionine oxidation and acetylation at the protein N-terminus. N-Glycopeptides were identified by analyzing the proteomics raw files using Byonic (Protein Metrics, v4.1.10). Searches were performed using full trypsin specificity, with enzymatic cleavage occurring C-terminal to lysine (K) and arginine (R), and permitting up to two missed cleavage events. Mass accuracy thresholds were set to 10 ppm for precursor ions and 0.02 Da for fragment ions. Carbamidomethylation of cysteine was defined as a static modification, whereas oxidation of methionine residues and N-terminal protein acetylation were considered as variable modifications. Spectral searches were conducted against the UniProtKB/Swiss-Prot human reference proteome, supplemented with a curated database of mammalian N-glycan compositions. Identified glycopeptides were consolidated across searches, and redundancies were removed prior to downstream analysis. Confidence filtering was applied using multiple criteria, including a two-dimensional false discovery rate below 1%, an absolute log probability greater than 1, precursor charge states between 2 and 6, and a minimum Byonic score of 300. All retained N-glycopeptide assignments were manually reviewed to confirm accurate glycan composition and site localization. The identified N-glycopeptides were quantified using Skyline software (version 24.1.0.199).

N-Glycan Release and Purification

A total of 100 μg of protein from each sample of enriched synaptosome and SV fractions was transferred to individual Eppendorf tubes and diluted to a final volume of 100 μL with 50 mM ABC buffer. Proteins were denatured by heating the samples at 90 °C for 15 min. After cooling, 3000 units of PNGase F were added to each tube to enzymatically release N-glycans. The reaction mixtures were then incubated at 37 °C for 18 h to ensure complete deglycosylation. Following N-glycan release, the samples were dried using a vacuum concentrator. The dried residues were reconstituted in 300 μL of 5% acetic acid (v/v) for cleanup and removal of deglycosylated proteins.

Solid-phase extraction (SPE) cleanup was performed using C18 cartridges. The cartridges were conditioned by rinsing three times with 1 mL of methanol, followed by three rinses with 1 mL of 5% acetic acid. The reconstituted samples were then loaded onto the prepared cartridges and washed three times with 300 μL of 5% acetic acid. The flow-through, containing the released N-glycans, was collected into 1.5 mL tubes and dried again using a vacuum concentrator in preparation for downstream analysis.

N-Glycan Reduction and Permethylation

Once released and purified, the N-glycans were reduced using a protocol adapted from a previously published method. A 10 mg/mL solution of borane–ammonia complex was prepared using HPLC-grade water. To each sample, 10 μL of freshly prepared borane–ammonia complex solution was added, followed by incubation at 60 °C for 1 h. Subsequently, excess borane–ammonia was removed from the samples by repeatedly adding 1000 μL of methanol until the resulting methyl borate was completely evaporated using a vacuum concentrator. After reduction, the N-glycans underwent solid-phase permethylation according to an established protocol. , This procedure entailed reconstituting the reduced N-glycan samples in 30 μL of DMSO, followed by the addition of 1.2 μL of water and 20 μL of iodomethane. Microspin columns, prepacked with NaOH beads suspended in DMSO, were centrifuged at 1800 rpm for 2 min and then rinsed with 200 μL of DMSO and centrifuged again at the same speed. The sample solution was loaded into the columns and incubated in the dark at room temperature for 25 min. An additional 20 μL of iodomethane was introduced, and the column was incubated for a further 15 min. After incubation, the columns were centrifuged at 1800 rpm for 2 min to collect the eluent, followed by the addition of 30 μL of ACN for a second elution. The permethylated N-glycans were collected by centrifugation, dried, and reconstituted in an aqueous solution of 20% acetonitrile and 0.1% formic acid for LC-MS analysis.

LC-MS/MS Analysis of Permethylated N-Glycans

C18-LC-MS/MS Parameters

Reduced and permethylated N-glycans, derived from an equivalent of 25 μg of starting protein, were injected into the LC-MS/MS system. The analysis was conducted using an UltiMate 3000 nano UHPLC system (Thermo Scientific, San Jose, CA, USA) coupled to an Orbitrap Fusion Lumos mass spectrometer (Thermo Scientific, San Jose, California, USA) operating in positive ion mode. Sample loading and online purification were performed using an Acclaim PepMap 100 C18 trapping column (75 μm × 2 cm, 3 μm particle size, 100 Å pore size, Thermo Scientific) with mobile phase A (MPA) delivered at a flow rate of 3 mL/min for 10 min. Subsequent chromatographic separation was achieved on a 15 cm reversed-phase C18 Acclaim PepMap column. The mobile phases consisted of 98% HPLC-grade water, 2% acetonitrile, and 0.1% formic acid (FA) for MPA, while mobile phase B (MPB) contained 100% ACN, with 0.1% FA. Separation was carried out at 55 °C with a flow rate of 0.35 mL/min, using a gradient elution that started at 20% MPB for 10 min, increased to 55% over 35 min, and then rose to 90% within 5 min. The gradient was then returned to 20% and maintained for 5 min to re-equilibrate the column. Following chromatographic separation, permethylated glycan samples were introduced into the Orbitrap Fusion Lumos mass spectrometer. Ionization was performed with a spray voltage of 2 kV, while the capillary temperature was maintained at 305 °C. Full-scan MS acquisition was carried out at a resolution of 120,000, with an AGC target set to standard, an expected peak width of 30 s, and a mass scan range of 400–2000 m/z. MS/MS spectra were acquired in data-dependent mode, selecting the top 20 most intense ions for fragmentation. A quadrupole was used for precursor isolation with a 2 m/z isolation window, and fragmentation was achieved using collision-induced dissociation (CID) at a fixed collision energy of 35. The MS/MS scans were acquired at a resolution of 30,000, with the AGC target set to standard and maximum injection time set to auto.

MGC-LC-MS/MS Parameters for Isomeric Separation

The N-glycan isomers associated with PD were characterized using an in-house-prepared 10 mm MGC column, which has been shown to separate isomers of permethylated glycans and glycopeptides effectively. −

This analysis utilized an UltiMate 3000 nano UHPLC system (Thermo Scientific, San Jose, California, USA) connected to a Q-Exactive HF mass spectrometer (Thermo Scientific, San Jose, California, USA), operating in positive ion mode. Briefly, all samples underwent a 90 min multistep gradient on the MGC column to effectively separate glycans. Mobile phase A (MPA) consisted of 98% water, 2% acetonitrile (ACN), and 0.1% difluoroacetic acid (DFA), whereas mobile phase B (MPB) was made up of 50% ACN, 50% isopropanol (IPA), and 0.1% DFA. The column was maintained at a temperature of 75 °C with a flow rate of 0.3 μL/min. The gradient began at 20% MPB for 10 min, and then rose to 60% over 20 min, and subsequently increased to 95% over 30 min. The gradient was kept constant at 95% for 20 min. It then decreased to 20% over 8 min and was maintained at that level for 2 min to re-equilibrate the column. Following LC separation, N-glycans were analyzed using a mass spectrometer fitted with a nanoelectrospray ionization (ESI) source operating in positive ion mode. The spray voltage was set to 1.6 kV, and the transfer tube temperature was maintained at 305 °C. Full MS spectra were acquired with an Orbitrap mass analyzer, covering a mass range of 400 to 2000 m/z. The Orbitrap was set to a resolution of 120,000 and an accuracy of 5 ppm. The maximum injection time was 50 ms, with an AGC target of 1e6. The tandem MS/MS Orbitrap scan used a data-dependent acquisition mode to select the 20 most abundant precursor ions for MS/MS scanning at a normalized collision energy (NCE) of 23%. The isolation window was configured at 2 m/z and an AGC target of 1e5. The mass analyzer resolution was adjusted to 30,000, with a maximum injection time of 100 ms and a loop count of 20.

Data Analysis

N-Glycan identification was performed using a combined approach involving MaxQuant and R scripts. First, LC-MS raw files were processed in MaxQuant without specifying any Fasta files. The MS2 output was saved as a text file, and precursor mass and charge data were extracted for each MS2 spectrum. Then, R (version 4.4.1) and the tidyverse package suite were utilized to match precursor m/z and charge values to glycan structures from an in-house database. The results were filtered to include only glycans detected with two or more adducts at the same retention time. Each identified glycan structure was verified and quantified using Skyline software (version 24.1.0.199).

The absolute proteomics and N-glycopeptides and N-glycan abundances were transformed using log2 transformation and normalized by quantile normalization. Statistical analysis was performed using the Reproducibility-Optimized Test Statistic (ROTS) method in R. The ROTS is a data-driven method that selects the most reproducible form of a t-like test statistic by optimizing its parameters through bootstrapping. It maximizes the reproducibility of the top-ranked features across resampled data sets, resulting in a robust and stable ranking of differential features. , Principal component analysis (PCA), heatmaps, effect size analysis, and boxplots were generated using R. Extracted ion chromatogram (EIC) and MS spectrum were generated using Xcalibur software (Version 4.2., Thermo Scientific) and Pyteomics package in Python (version 3.13.2).

Results

Characterization of Synaptosomes and Synaptic Vesicles by Transmission

Electron Microscopy (TEM) and LC-MS/MS Proteomics Analysis

Synaptosome and SV were enriched from the PFC of post-mortem human brain tissues from PD patients and matched controls. The enrichment process utilized sequential differential centrifugation and sucrose density gradient ultracentrifugation (Figure ). Transmission electron microscopy (TEM) was used to validate vesicle enrichment and assess vesicle morphology, size, and purity (Figure ). Positive staining of synaptosomes revealed with high-contrast, membrane-bound structures containing mitochondria and synaptic vesicles (Figure a), while negative staining of SVs revealed uniformly spherical vesicles with clear membrane outlines (Figure b).

To further characterize the molecular content of each synaptic fraction, LC-MS/MS-based proteomics analysis was carried out. A total of 1281 proteins were identified in synaptosomes and 907 in SVs using the Proteome Discoverer software (3.2). Enrichment analysis using g:Profiler revealed that synaptosomes were enriched in diverse cellular components, including mitochondrial outer membrane translocase complexes and other mitochondrial complexes (Figure c). SVs showed strong enrichment for vesicle-specific components such as V-type ATPase complexes, the synaptobrevin 2-SNAP-25-syntaxin complex, and other synaptic vesicle membrane complexes (Figure d).

The Venn diagram (Figure e) showed that 742 proteins were common to both vesicle types, while 539 proteins were unique to synaptosomes and 165 were unique to SVs. Synaptosome-specific proteins include mitochondrial markers such as ATP synthase FO complex subunit B1, cytochrome c1, and the mitochondrial calcium uniporter. SV-specific proteins included synaptosomal-associated protein 23 and others involved in vesicle docking and fusion. Shared proteins included synaptotagmin-1, synapsin-1 and -2, synaptic vesicle glycoproteins 2A/2B (synaptosomal-associated protein 25) SNAP-25, and V-type proton ATPase catalytic subunit A.

N-Glycan Profile of Synaptosome Synaptic Vesicle

from PD and Control Brains

N-Glycan structures were annotated using a four-digit nomenclature, with each digit representing the number of monosaccharide units corresponding to specific sugar types in the N-acetylglucosamine (GlcNAc), hexose (mannose/galactose), fucose (Fuc), and N-acetylneuraminic acid (Neu5Ac). For example, the structure “5–5–1–1” denotes a glycan containing five units of GlcNAc, five Hexoses, one Fucose, and one Neu5Ac. Isomeric variants were denoted using suffixes, such as “5–5–1–1_Iso2”, indicating isomer 2 of the “5–5–1–1” glycan. The corresponding glycan symbols used in this study are illustrated in the caption of Figure .

A total of 66 N-glycans were identified in the synaptosome fraction and 68 in the SV. Of these, 56 were common to both, while 10 were unique to the synaptosome and 12 were unique to SV (Figure S1). To illustrate the glycan profiles, the extracted ion chromatogram (EIC) of 14 representative common N-glycans in both synaptic fractions is shown in Figure a,b for synaptosome and SV, respectively. Additionally, Figure c shows the spectral distribution of N-glycans across an average retention time of 25–45 min. Tables summarizing the normalized abundance for each group are shown in Tables S1–S4. The normalized abundance data were used for subsequent analyses and statistical comparisons. The most abundant N-glycan found in both the synaptosome, and SV samples was the fucosylated structure 5–3–1–0. Figure S2 presents the chromatographic traces of the N-glycan 5–3–1–0 in control and PD synaptosome fractions, control and PD SV fractions, and the MS/MS spectrum with annotated fragment ions for this N-glycan.

The N-glycans in synaptosome (Figure S3a) and SV fractions (Figure S3b) were categorized based on glycan type, branching, fucosylation, and sialylation profiles. In terms of glycan type, synaptosomes were predominantly enriched in sialofucosylated structures (43.94%), followed by fucosylated (30.30%), sialylated (10.61%), high mannose (9.09%), and other glycans (6.06%). Similarly, SVs also exhibited a high proportion of sialofucosylated glycans (38.24%), followed by fucosylated forms (36.76%), while sialylated and high mannose glycans were present at 8.82% each, and other types accounted for 7.35%. In terms of branching, synaptosomes showed a predominance of biantennary (28.79%) and triantennary (27.27%) glycans, whereas SVs exhibited higher levels of biantennary (26.47%) and tetra-antennary (26.47%) structures, followed by triantennary (23.53%) glycans. The fucosylation profile showed that monofucosylated glycans were highly represented in both fractions, 37.88% in synaptosomes and 45.59% in SVs, along with moderate levels of difucosylated and nonfucosylated species. The sialylation profile indicated that the highest proportion of glycans were nonsialylated, followed by monosialylated and disialylated glycans in both synaptosome and SV fractions. These profiles provide a comprehensive snapshot of the glycan landscape within each synaptic compartment (Figure S3a,b). Further classification of all fucosylated glycans into core-, antenna-, and mixed-fucosylated species (core and antenna) revealed a similar distribution in synaptosomes and SVs. Classification was based on diagnostic MS/MS fragmentation patterns indicative of fucose position (Figure S4). In synaptosomes, mixed-fucosylated glycans constituted the largest fraction (53.1%), followed by core-fucosylated species (44.9%), whereas purely antenna-fucosylated glycans represented only a minor fraction (2.0%) (Figure S5a). Synaptic vesicles (SVs) exhibited a comparable pattern, with mixed and core-fucosylated populations accounting for 51.0 and 45.1%, respectively, alongside a slightly increased antenna-fucosylated fraction (3.9%) (Figure S5b). Cohen’s d effect size analysis highlighted a distinct PD-specific signature characterized by a divergence between antenna and core modifications. In synaptosomes, PD was associated with a robust increase in antenna fucosylation and a moderate rise in mixed species, contrasted by a significant reduction in core-only fucosylation (Figure S5c). This pattern was largely mirrored in the SV fraction, which exhibited a positive effect size for antenna fucosylation and a marked negative effect size for core fucosylation, while mixed species showed negligible change (Figure S5d). These indicate a PD-associated shift in fucosylation processing toward antenna modifications at the expense of core fucosylation.

Unsupervised Principal Component Analysis

An unsupervised principal component analysis (PCA) was performed using the normalized abundance data from control and PD N-glycans from the synaptosome and SV. PCA was used to explore intrinsic differences in glycan composition between the control and PD groups. The PCA reduces dimensionality and reveals distinct clustering patterns that reflect underlying molecular variation in high-throughput omics data. The principal components (PCs) represent the major sources of variance in the data set. In synaptosomes, PC1 and PC2 account for 25.19 and 19.24% of the total variance, respectively (Figure a). In SVs, PC1 explains 23.29% of the variance, while PC2 accounts for 19.77% (Figure b).

Compositional Analysis of N-Glycans in Synaptosomes

and Synaptic Vesicles

To evaluate N-glycan compositional differences in synaptosomes and SVs between control and PD groups, N-glycans were classified into five categories: sialofucosylated, neutral, and high mannose, and Cohen’s d effect size analysis was conducted on the mean normalized abundance of each glycan class. In the synaptosome (Figure c), the analysis revealed a strong negative effect size (−0.63) for sialylated N-glycans, indicating a pronounced decrease in sialylation in PD synaptosomes relative to controls. Sialofucosylated glycans also exhibited a moderate negative effect size (−0.23), suggesting reduced levels in the PD group. Conversely, both fucosylated and neutral glycan categories showed positive effect sizes of approximately 0.3, reflecting an increase in these glycan types in PD. High mannose glycans had a negligible effect size, indicating minimal differences between the two groups. The effect size analysis using Cohen’s d was also applied to assess differences in N-glycan types between control and PD groups in SVs (Figure d). The results showed a strong positive effect size (0.89) for the neutral and high mannose glycan categories, indicating substantial increases in PD SVs. In contrast, fucosylated glycans displayed a moderate negative effect size (−0.31), indicative of a reduction in PD, while sialylated and sialofucosylated glycans showed negligible or near-zero effect sizes, suggesting little to no difference between the groups for these types in the SVs of PD brains.

Differential Expression Analysis of N-Glycans

Differential expression of N-glycans between PD and control groups in synaptosome and SV fractions were assessed using ROTS analysis, with N-glycans showing a false discovery rate (FDR) <0.05 considered statistically significant. In synaptosomes, five N-glycans were found to be statistically significant, with three being upregulated (3–5–2–1, 4–6–2–0, and 5–8–1–2) and two downregulated (4–5–0–1 and 5–5–1–1) in PD samples compared to controls (Figure a). In SVs, nine N-glycans were statistically significant, with four upregulated (2–3–1–0, 2–5–0–0, 5–3–0–0, and 6–4–2–0) and five downregulated (4–5–1–0, 5–4–1–0, 5–4–1–1, 5–5–1–0, and 6–5–1–1) in PD (Figure b). Hierarchical clustering heatmaps (Figure S6) showed clear group-specific clustering, indicating distinct glycan expression profiles between PD and control samples in synaptosome and SV. The box plots in Figure illustrate the normalized abundance of each differentially expressed glycan, emphasizing both the direction and degree of expression changes in synaptosome and SV fractions.

Lewis-Type Structures Identified among Differentially Expressed N-Glycans

Linkage-specific and positional analysis of differentially expressed N-glycans, based on diagnostic fragment ions in MS/MS data, revealed the presence of Lewis-type and sialyl Lewis structures. Specifically, the N-glycans 3–5–2–1 and 5–8–1–2, which were upregulated in synaptosomes (Figure a­(i and ii)), exhibited fragment patterns consistent with the sialyl Lewis X (sLeX) motif. These included fragments corresponding to the characteristic tetrasaccharide Neu5Ac­(α2–3)­Gal­(β1–4)­[Fuc­(α1–3)]­GlcNAc. The corresponding MS/MS spectra, including key diagnostic fragments supporting this annotation, are shown in Figure . The sLeX fragment was not found in the MS/MS spectra of other significant N-glycans. The MS/MS spectra of selected differentially expressed N-glycans from the two synaptic fractions, not having sLeX, are shown in Figure S4. This includes the N-glycan 5–5–1–1 from the synaptosome fraction and the N-glycans 6–4–2–0 and 5–4–1–1 from the SV fraction.

Differential Expression Analysis of Isomeric N-Glycans in PD Synaptosome and Synaptic Vesicle

Among the 90 N-glycan isomers identified in the synaptosome, two representative examples of isomeric separation with significant alterations are highlighted in Figure . These include the sialofucosylated N-glycans 4–5–1–1 and 5–5–1–1. The extracted ion chromatogram and box plots of the isomers of N-glycan 4–5–1–1, with two distinguishable isomers identified in the synaptosome and SV, are shown in Figure a and Figure b, respectively. Only 4–5–1–1_Iso2, which is expected to be the α2,6-linked sialic acid isomer, was statistically significant in the synaptosome fraction, while neither isomer was significant in the SV fraction. The extracted ion chromatogram and box plots of the isomeric N-glycan 5–5–1–1, with four distinguishable isomers identified in the synaptosome and SV, are shown in Figure c and Figure d, respectively. In this case, none of the isomers were statistically significant in the synaptosome, but isomers 2, 3, and 4 (5–5–1–1_Iso2, 5–5–1–1_Iso3, and 5–5–1–1_Iso4, respectively) were statistically significant in the SV fraction.

Building on this, a broader quantitative comparison revealed multiple differentially expressed N-glycan isomers. Specifically, eight N-glycan isomers in the synaptosome showed significant alterations with an FDR less than 0.05 (Figure a). Of these, three were upregulated (6–3–1–0_Iso1, 6–4–0–1_Iso1, and 7–3–1–0_Iso1), while five were downregulated (4–4–1–0_Iso1, 4–4–1–0_Iso2, 4–5–0–1_Iso1, 4–5–1–1_Iso2, and 5–3–1–0_Iso2) in the PD group compared to the control. In the SV group, 9 out of 88 identified N-glycan isomers exhibited significant differences (Figure b), with three upregulated (6–4–2–0_Iso1, 6–5–1–2_Iso2, and 7–5–2–1_Iso2) and six downregulated (4–5–2–1_Iso1, 5–4–1–1_Iso2, 5–4–1–1_Iso3, 5–4–1–1_Iso4, 5–4–2–1_Iso2, and 5–5–2–0_Iso1) in PD compared to control. Hierarchical clustering was visualized using heatmaps to reveal expression patterns of significant N-glycan isomers across synaptosome and SV samples, as shown in Figure S7. The boxplots in Figure a and Figure b display the distribution of each differentially expressed N-glycan isomer in synaptosome and SV, respectively.

Site-Specific, Protein-Normalized N-Glycoproteomics

Analysis Reveals Glycan Remodeling Independent of Protein Abundance Changes

Site-specific N-glycopeptide analysis was carried out, and glycopeptide abundances were normalized to their corresponding protein abundance, enabling discrimination between protein-driven effects and true glycan remodeling. In synaptosome, multiple glycopeptides, including those derived from dynamin-1-like protein, and choline transporter-like protein 1, displayed significantly altered glycopeptide abundances after protein normalization, indicating true glycan remodeling rather than changes driven by protein expression (Table S5).

Similarly, in SV, protein-normalized glycopeptide analysis revealed widespread glycan remodeling. Notably, multiple glycopeptides from SNAP-25 exhibited significant glycopeptide abundances after protein normalization, further supporting glycosylation-specific regulation (Table S6). These findings demonstrate that synaptic compartments undergo active, selective glycan remodeling in PD independent of protein changes.

Discussion

Synaptic transmission deficits are a hallmark of PD. , The role of glycomics in synaptic transmission in PD pathogenesis has gained increasing attention, with aberrant glycosylation patterns observed on key synaptic proteins such as α-synuclein and synaptic vesicle glycoprotein 2A (SV2A), which can impact key processes, including vesicle trafficking and neurotransmitter release. Synaptosomes and SVs have emerged as invaluable tools for studying these synaptic abnormalities , The PFC is a critical region of cognitive and executive functions and has been implicated in PD. In this study, we performed LC-MS/MS glycomics analysis on enriched synaptosome and SV N-glycans from the PFC of PD brains. Our findings reveal distinct N-glycomic alterations that may contribute to synaptic dysfunction and cognitive decline in PD.

The differences in shared and unique N-glycans between synaptosomes and SVs, as shown in Figure S1, indicate specific glycosylation patterns in the synaptic fractions that reflect their distinct protein compositions and functional roles in neurotransmission. Consistent with previous reports, our characterization of the human brain N-glycome (Figure S3) aligns with regio-specific mapping. High-mannose glycans are low in abundance in the human cortex, in contrast to mouse models, where they dominate the central nervous system. In contrast, human cortical synapses are enriched in mature, multiantennary sialofucosylated N-glycans. This dominance in the structural diversity of sialofucosylated glycans in both synaptosomes and SVs, as shown in Figure S2, is consistent with the well-established role of complex N-glycans in neural cell–cell recognition, synaptic stability, and plasticity. , The relatively high proportion of high mannose glycans, although less abundant in number compared to complex forms, is typical for neural tissues and may reflect the presence of immature glycoproteins or those involved in rapid trafficking. Branching analysis indicates a greater number of branched N-glycans in both synaptosomes and SVs. Such branching is associated with increased functional diversity and enhanced interactions with glycan-binding proteins such as lectins, which can modulate SV trafficking and protein sorting.

Compositional and effect size analysis of the N-glycans from synaptosome and SV fractions suggests significant differences in N-glycan profiles induced by PD-related changes. In synaptosomes (Figure c), a marked reduction in sialylated and sialofucosylated N-glycans in PD may reflect altered terminal glycosylation, potentially due to impaired sialyltransferase activity, dysregulation of the glycosylation machinery, or increased activity of sialidases. Sialylation is critical for synaptic stability, neuronal signaling, and protection against immune-mediated damage; its reduction may contribute to synaptic dysfunction and neuroinflammation observed in PD. The observed increase in fucosylated and neutral N-glycan types could signify a compensatory shift in glycan biosynthesis or remodeling in response to PD-associated stress. In contrast, SV glycan profiles (Figure d) exhibited elevated levels of high mannose and neutral glycan types in PD samples, suggesting alterations in glycan processing within the secretory pathway or impaired maturation of glycoproteins in the disease state. The limited changes in sialylated and sialofucosylated glycans observed in SV compared to synaptosomes reflect selective glycomic alterations within synaptic subfractions in PD. The slight decrease in fucosylated glycans may reflect subtle regulatory changes in fucosylation within SVs. Our findings in the PFC synaptic compartments both align with and diverge from existing tissue-level studies in other brain regions affected in PD. For instance, the increase in high-mannose SV glycans mirrors alterations observed in the substantia nigra, suggesting a disease-general maturation defect. However, the reduction in synaptosomal sialofucosylation mirrors PFC and hippocampus changes seen in Alzheimer’s disease but contrasts with the global increases reported in subcortical PD tissues such as substantia nigra and striatum, potentially representing a region-specific synaptic signature of cortical cognitive decline. This pattern may therefore represent a region-specific synaptic glycan signature associated with cortical involvement and cognitive decline.

The differentially expressed glycans in synaptosome suggest a shift toward hyper-fucosylation and increased glycan complexity on synaptosomal proteins in the PFC of PD brain. These changes may reflect upregulated fucosyltransferase activity and altered sialylation pathways in the diseased brain. Notably, elevated fucosylation, both core and antenna fucose, has been reported as a general feature of the PD brain glycome, along with a modest overall increase in sialylation. SV glycoproteins in the PD PFC show increased expression of less mature N-Glycans, and a decrease in expression of galactosylated and sialylated glycans, suggesting a deficit or alteration in glycan processing. Also, there is an increase in hyper-fucosylated glycans, containing two fucose residues, while standard monofucosylated complex glycans are decreased. This indicates altered fucosylation patterns in PD, with enhanced core and outer arm fucosylation, and typical maturation steps such as galactosylation and sialylation are disrupted. These changes align with broader glycomic trends in PD, which include elevated fucosylation and accumulation of oligomannose-type glycans. The reduction in sialylated structures in SVs may reflect localized deficits in sialylation capacity or glycan stability on vesicle proteins. These findings indicate that PD alters N-glycosylation in both synaptic membranes and vesicles within the prefrontal cortex, with distinct shifts in glycan branching, fucosylation, and sialylation. The synaptosomal glycome exhibits increased complexity and terminal modifications, while the SV glycome shifts toward less mature and more variably fucosylated forms. These changes in synaptic compartments may influence synaptic stability, vesicle trafficking, and neuron–glia communication in PD, potentially contributing to the cognitive decline associated with the disorder.

Linkage-specific and positional analysis of differentially expressed N-glycans revealed the presence of sialyl Lewis X (sLeX) structures in synaptosomes (upregulation of N-glycans 3–5–2–1 and 5–8–1–2). The sLeX epitope, composed of the tetra saccharide Neu5Acα2–3Galβ1–4­(Fucα1–3)­GlcNAc, is known for mediating cell–cell adhesion, immune cell trafficking, and inflammatory signaling through interactions with selectins and other lectins. , A recent study has demonstrated increased expression of sLeX in the cortex and striatum of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-treated mouse models of PD, linked to the upregulation of fucosyltransferase 7 (FUT7), the enzyme responsible for α1–3 fucosylation that generates Lewis X-type structures. Immunohistochemical data localized sLeX expression to both neurons and activated microglia, implicating it in neuroinflammatory signaling and potentially maladaptive neuron–glia interactions. The identification of sLeX-bearing glycans in human prefrontal cortex synaptosomes aligns with this model and suggests similar glycomics remodeling at the synapse.

Although compositional glycan profiling reveals changes in abundance and structure, it lacks the resolution to distinguish isomeric glycans with different linkages or branching, differences that critically influence glycoprotein function and interactions. Selective alteration of the 4–5–1–1_Iso2 species in synaptosomes, likely corresponding to an α2,6-sialylated form, suggests localized dysregulation of sialylation at the synaptic membrane, while the absence of changes in SVs underscores compartment specificity. Conversely, multiple 5–4–1–1 isomers were altered exclusively in the SV fraction, pointing to vesicle-specific modifications in branching or fucosylation that may influence vesicle stability and neurotransmitter handling. Among the differentially expressed N-glycan isomers in synaptosome (Figure a), 6–3–1–0_Iso1, 6–4–0–1_Iso1, and 7–3–1–0_Iso1 were upregulated, suggesting enhanced biosynthesis of complex, highly branched glycans with fucosylation and/or sialylation. These changes may reflect increased enzymatic activity of branching glycosyltransferases such as MGAT4 or MGAT5, or core and terminal fucosyltransferases like FUT8 and FUT9. , In contrast, five isomers, 4–4–1–0_Iso1, 4–4–1–0_Iso2, 4–5–0–1_Iso1, 4–5–1–1_Iso2, and 5–3–1–0_Iso2, were downregulated, which correspond to bi- or tri-antennary structures. The differential regulation of isomers from the same glycan composition (i.e., 4–4–1–0_Iso1 and 4–4–1–0_Iso2) indicates disease-specific altered positional or linkage-specific processing, which may reflect sub-Golgi enzyme dysfunction or disrupted glycoprotein trafficking in PD. In the SV fraction (Figure b), the upregulated isomers represent highly branched, multifucosylated, and sialylated structures, potentially indicative of altered glycoprotein sorting or stabilization within vesicles. Conversely, the downregulated isomers point to a selective loss of structurally mature complex glycans. These shifts may indicate impaired terminal glycosylation or early truncation, possibly due to deficits in galactosylation or sialylation, which have been reported in PD. Importantly, these isomer-specific changes that synaptic glycoproteins in PD may bear structurally distinct glycan signatures that influence their stability, signaling capacity, or immune recognition. , As such, N-glycan isomeric profiles may reflect synaptic pathology-related molecular changes in PD and its associated cognitive decline.

Age, sex, and PMI are potential confounders in post-mortem brain omics analyses. In this study, the PD and control cohorts were well matched for sex, with no statistically significant difference between groups, reducing the likelihood that the observed glycomics differences reflect sex-related bias. Although mean age differed modestly between groups, the age distributions substantially overlapped, and both cohorts fell within an advanced elderly range. Importantly, key age-associated neuropathological measures were comparable between groups, arguing against a meaningful contribution of age-related pathology to the observed differences. PMI distributions also overlapped extensively between PD and control samples and did not differ significantly between groups. Moreover, post-mortem effects are expected to influence proteins and glycans in a largely non–disease-specific manner, whereas the observed alterations were selective, synaptic compartment–specific, and biologically coherent, supporting a disease-related rather than technical or demographic origin. Nonetheless, PMI-related variability remains an inherent limitation of post-mortem human brain studies. Future studies with larger cohorts and tighter matching will enable analyses with greater statistical power and further refine the interpretation of synaptic glycomic remodeling in PD.

Conclusions

This study provides a detailed N-glycomics profile of synaptosomes and SVs enriched from the PFC of PD and control brains, revealing distinct glycosylation patterns and disease-related changes. High-resolution TEM and LC-MS/MS proteomics confirmed the integrity and enrichment of each synaptic fraction. Both synaptic compartments are mainly composed of complex sialofucosylated structures, which are key mediators of synaptic communication and stability. PD synaptosome is associated with reduced sialylated and sialofucosylated glycans, along with an increase in fucosylated and neutral types, suggesting altered terminal glycosylation that may impair synaptic adhesion and signaling. In SVs, PD was characterized by increased levels of high mannose and neutral glycans, indicating disrupted glycoprotein maturation. Differential expression analysis identified multiple N-glycans in PD. Notably, some upregulated synaptosomal glycans displayed sialyl Lewis X (sLeX) epitopes (3–5–2–1 and 5–8–1–2), implicating them in neuroinflammatory processes and altered neuron–glia interactions. SV glycan changes indicated a shift toward less mature, hypogalactosylated, and hyper-fucosylated forms. Isomer-resolved profiling revealed linkage- and position-specific changes in glycan structure, highlighting complex remodeling not visible in standard compositional analysis. Overall, these findings demonstrate that PD induces specific alterations in the synaptic glycome of synaptosome and synaptic vesicle fractions of the PFC. These alterations may contribute to synaptic dysfunction, neuroinflammation, and cognitive decline associated with PD. Importantly, differentially expressed glycan structures with sLeX motifs provide promising avenues for further research aimed at restoring synaptic glycosylation balance in PD. However, the relatively small sample size in this study may limit statistical power, emphasizing the need for validation in larger, independent cohorts.

Supplementary Material