Structurally Defined Chitosan Oligosaccharides (COS) Suppress Macrophage Inflammatory Signaling and Enzymes in Atherosclerosis Relevant Pathways

1. Introduction

The process of atherosclerosis is an ongoing, lipid-based inflammatory process of the arterial walls which leads to approximately twenty million cardiovascular related deaths annually throughout the world [1]. The focus on the role of inflammation in atherosclerosis was initiated with the recognition that it is not simply a passive process of lipid deposition into arterial walls, but rather a chronic inflammatory condition cause d by endothelial dysfunction, oxidative stress and prolonged macrophage activity [2,3,4].

Hyperlipidemia and reactive oxygen species stimulate the endothelial surface causing further penetration and oxidation of the low-density lipoproteins (LDL), creating modified forms of LDL (ox-LDL), which promotes further attraction of monocytes and their subsequent transformation into foam cells through the action of scavenger receptors; this is characteristic of early lesions within the arterial wall [5,6,7,8]. Due to the fact that oxidative and inflammatory signals from macrophages contribute to continued growth and expansion of the lesions, macrophages have become the primary target for the development of novel therapeutic interventions aimed at preventing or treating atherosclerotic disease.

Chitosan oligosaccharides (COS), low-molecular-weight β-(1→4)-linked D-glucosamine derivatives of chitosan, have attracted considerable attention as marine-derived bioactive compounds because of their antioxidant, anti-inflammatory, and immunomodulatory properties, as well as their ability to attenuate atherosclerotic lesion development in experimental models [9,10,11]. However, the reported biological activities of COS vary considerably among studies because many preparations differ in molecular weight, degree of acetylation, and oligomer composition, while their structural characteristics are often incompletely defined. These physicochemical differences substantially influence the biological activity of COS, making it difficult to establish clear structure–activity relationships [9,12,13].

In our previous study, we developed a simple and reproducible hydrochloric acid hydrolysis method to produce a structurally characterized low-molecular-weight COS fraction. This preparation demonstrated improved physicochemical properties that are expected to enhance cellular uptake and bioavailability. Moreover, in 3T3-L1 adipocytes, COS reduced intracellular lipid accumulation and triglyceride content while promoting lipolysis and insulin-stimulated glucose uptake, suggesting beneficial effects on metabolic dysfunction [14]. Despite these promising findings, the molecular mechanisms by which structurally defined COS regulate macrophage inflammation remain poorly understood. Most previous studies have focused on reductions in nitric oxide and pro-inflammatory cytokines as biological endpoints without identifying the molecular targets responsible for these effects or examining global transcriptional responses associated with macrophage activation, foam cell formation, and atherosclerosis [15,16,17]. Therefore, integrating transcriptomic profiling with molecular docking provides an opportunity to identify the signaling pathways and candidate target proteins underlying the anti-inflammatory and anti-atherosclerotic activities of COS.

Herein we describe a method to prepare a structurally defined COS with a low average molecular weight by treatment of Scylla olivacea shell waste with HCl. We used NMR spectroscopy and mass spectrometry to confirm that our preparation was indeed a low-average molecular weight chitosan oligosaccharide (COS) [14,18,19]. Our COS preparation was then assessed using an integrated approach that included determining its antioxidant capabilities, assessing cytotoxicity and its ability to suppress NO in LPS stimulated RAW 264.7 cells, performing molecular docking against eleven receptors associated with both inflammation and atherosclerosis and comparing them against the known NSAID diclofenac, and conducting genome wide RNAseq analysis to assess global gene expression with subsequent enrichment analysis of the results into relevant pathways. Our goal was to integrate the predictions made by molecular docking regarding where COS would bind into macrophage cells with the observed effects of those bindings on gene expression to demonstrate that COS is capable of modulating multiple aspects of macrophage inflammation versus being a singular pathway inhibitor and therefore could be considered as a candidate for preventing atherosclerosis.

2. Results

2.1. Structural Characterization of HCL-Derived COS

The physicochemical characteristics of COS used in this study were previously reported by Songkoomkrong et al., 2026. Briefly, MALDI-TOF mass spectrometry demonstrated that COS consisted predominantly of low-molecular-weight oligomers (approximately 300-2000 Da) with degrees of polymerization ranging from 2 to 9 (Figure S1). Structural confirmation by ¹³C NMR spectroscopy indicated that the COS was mainly composed of glucosamine units with a high degree of deacetylation (92.35%) and minor residual N-acetylglucosamine residues (Figure S2). Detailed characterization data, including MALDI-TOF MS spectra and NMR analysis, was described (Figure S1) [14].

2.2. Antioxidant Activity of COS in the DPPH and ABTS Assay

In this study, antioxidant capacity was initially evaluated using the DPPH radical-scavenging assay, in which COS exhibited an IC₅₀ of 2.86 mg/mL. To further characterize its antioxidant properties, the ABTS radical scavenging assay was performed and showed an IC₅₀ value of 2.49 mg/mL. The lower IC₅₀ obtained in the DPPH assay compared with the ABTS assay indicates that antioxidant constituents present in COS may possess stronger scavenging efficiency or higher affinity toward DPPH radicals than toward ABTS radicals. The DPPH assay confirms that COS possesses measurable free radical-scavenging capacity and its potential to reduce oxidative stress in biological systems.

2.3. Effects of COS on Survival Rate of RAW 264.7 cells

The cytotoxicity of HCl-derived COS isoform on RAW 264.7 macrophages was assessed via an MTT assay following 24-hour and 48-hour treatments across concentrations between 20-1280 µg/mL. The result demonstrates that COS displayed no observable cytotoxicity at any concentration tested, with cell viability consistently ranging from 95% to 110% compared to the untreated control. This demonstrates that COS is biocompatible and well-tolerated by RAW cells, even at elevated doses. Furthermore, COS treatment did not induce abnormal cell proliferation, as viability values remained within normal physiological variations and did not exceed control levels in a dose-dependent manner. These results demonstrate that COS is non-toxic in RAW 264.7 macrophages, supporting its suitability for downstream anti-inflammatory assays without confounding cytotoxic effects (Figure 1).

2.4. Inhibition Assay of Cellular Nitric Oxide Production

The ability of COS to inhibit LPS-induced inflammatory activation was assessed by quantifying nitrite accumulation in RAW 264.7 macrophages via the Griess assay. Stimulation with LPS (1 µg/mL) markedly elevated NO production compared with the untreated control, confirming successful induction of the inflammatory response. COS treatment attenuated this LPS-induced increase in a dose-dependent manner. While 40 µg/mL COS produced a modest reduction in NO levels, significant inhibition was observed at 80 and 160 µg/mL, where nitrite production decreased to approximately 80% and 70% of the LPS group, respectively. These findings indicate that COS effectively suppresses NO synthesis in activated macrophages, supporting its potential anti-inflammatory activity (Figure 2).

2.5. Structural analysis of macrophage-relevant inflammatory receptors

Atomic coordinates and mature amino-acid sequences for inflammation-associated receptors related in macrophage activation and foam-cell formation, including Aldo-Keto Reductase Family 1 Member B1, Aldose Reductase (AKR1B1), Cyclooxygenase-2 (COX-2), Epidermal Growth Factor Receptor (EGFR0, Fibroblast Growth Factor 1 (FGF1), Fibroblast Growth Factor 2 (FGF2), Galectin-3, Heat Shock Protein 90 Alpha Family Class A Member 1 (HSP90AA1), Heparanase (HPSE), Nitric Oxide Synthase (NOS), Transient Receptor Potential Vanilloid 1 (TRPV1), Tumor Necrosis Factor Alpha (TNF-α), Vascular Endothelial Growth Factor A (VEGFA), were retrieved from the RCSB Protein Data Bank and used as receptor targets for docking of chitosan-oligosaccharide (COS). These proteins collectively integrate redox signaling (AKR1B1, NOS), pro-inflammatory eicosanoid and cytokine axes (COX-2, TNF-α), growth-factor/trophic signal (FGF1/2, VEGFA), ion-channel-dependent activation (TRPV1), extracellular-matrix remodeling (HPSE), and pattern-recognition lectin signaling (galectin-3) to induce the atherogenic macrophage program.

The proteins exhibit domain organizations that adhere to their canonical folds, featuring well-defined ligand-accessible cavities. Enzymatic receptors (AKR1B1, COX-2, NOS) exhibited profound, hydrophobic clefts adjacent to catalytic residues, while lectin and growth-factor interfaces (galectin-3 CRD; FGF/VEGF heparin-binding grooves) revealed solvent-exposed channels reinforced by mixed a/b scaffolds. HSP90AA1 displayed the established Bergerat NTP-binding pocket, but TRPV1 provided a substantial cytosolic vestibule. Binding-site volumes were obtained from the docking grids and ranged over two orders of magnitude (e.g., compact cavities for FGF1/2 at approximately 64–70 ų compared to vast vestibules for TRPV1 and COX-2 at 25,745 and 4425 ų). These aligned with the structural diversity of the panel and allowed for an evaluation of ligand accommodation based on cavity layout.

2.6. Molecular docking of COS and reference ligands

The binding of COS was evaluated across all receptors and compared to the pharmacological positive control, diclofenac, a therapeutically utilized non-steroidal anti-inflammatory medication (NSAID). Docking scores (kcal/mol; AutoDock Vina) and cavity dimensions are presented in Table 1. The binding affinities of COS with AKR1B1, FGF1, FGF2, HPSE, HSP90AA1, TRPV1, VEGFA, COX-2, TNF-α, NOS, and galectin-3 were assessed using CB-Dock [20], with diclofenac utilized as a positive control for its inhibition of inflammatory pathways, thus offering advantageous effects in atherosclerosis-related conditions. COS demonstrated a clear binding preference within the target set, exhibiting its highest affinity for COX-2 (−7.7 kcal/mol), followed by galectin-3 (−7.4 kcal/mol), iNOS (−7.1 kcal/mol), TNF-α (−6.7 kcal/mol), and HPSE (−6.6 kcal/mol). Intermediate affinities were observed for TRPV1 (−6.5 kcal/mol), HSP90AA1 (−6.4 kcal/mol), and FGF1 (−6.3 kcal/mol), while the weakest interactions were seen for VEGFA (−6.1 kcal/mol), AKR1B1 (−5.7 kcal/mol), and FGF2 (−5.1 kcal/mol). COS had a greater affinity than diclofenac for various inflammatory mediators, including HPSE (−6.6 vs. −6.4 kcal/mol), VEGFA (−6.1 vs. −5.8 kcal/mol), and FGF1 (−6.3 vs. −6.0 kcal/mol). COS bound more strongly than diclofenac at three targets (FGF1, HPSE, and VEGFA) and showed comparable affinity at several others, although diclofenac exhibited higher predicted affinity at the majority of receptors.

The influence of cavity size on binding posture stability was minimal, since COS exhibited robust binding in both large pockets (COX-2, TRPV1) and medium-sized pockets (galectin-3). This finding suggests that electrostatic complementarity and hydrogen-bond networks, rather than pocket volume, are the principal factors influencing COS recognition. This pattern aligns with the polycationic and hydrogen-bond-dense characteristics of COS. These data suggest that COS interacts with several inflammatory targets, potentially facilitating a reduction in macrophage inflammation.

2.7. Effects of COS Treatment on the Transcriptome of RAW 264.7 cells

2.7.1. Differential Expressed Genes and KEGG Pathway Analysis

To investigate the molecular mechanisms underlying the protective effects of COS against inflammatory stimulation, transcriptomic profiling was performed in COS-pretreated RAW macrophages followed by LPS induction compared with LPS-treated control cells.

The volcano plot demonstrated extensive transcriptomic alterations after COS pretreatment (Figure 3A). A total of 15,590 differentially expressed genes (DEGs) were identified (Table S1), including 11,399 upregulated transcripts (73.1%) and 4,191 downregulated transcripts (26.9%), indicating that COS induced widespread transcriptional remodeling in macrophages under inflammatory conditions. The predominance of upregulated transcripts suggests that COS may activate multiple protective and adaptive cellular responses rather than exclusively suppressing gene expression. Hierarchical clustering analysis of DEGs further demonstrated clear separation between the COS pretreatment and LPS-treated groups (Figure 3B), indicating distinct transcriptomic signatures and confirming that COS substantially altered the LPS-induced cellular response. To understand the biological relevance of these transcriptomic changes, KEGG enrichment analysis was performed using significantly regulated transcripts (Figure 3C) (Table S2). Several pathways closely related to inflammation, macrophage activation, foam cell formation, and atherosclerosis development were significantly enriched.

Among the enriched pathways, the TNF signaling pathway and NF-κB signaling pathway were particularly notable because these pathways are major regulators of macrophage inflammatory responses and are activated during LPS stimulation. Their enrichment suggests that COS pretreatment modulates inflammatory signaling networks involved in cytokine production and immune activation. Importantly, enrichment of the lipid and atherosclerosis pathway indicates that COS may regulate processes involved in macrophage lipid handling and vascular inflammation. Since excessive lipid uptake and sustained inflammatory activation promote the conversion of macrophages into foam cells, modulation of this pathway suggests a potential role of COS in reducing mechanisms associated with foam cell formation and subsequent atherosclerotic progression. Additional enrichment of pathways, including cellular senescence, protein processing in the endoplasmic reticulum, ubiquitin-mediated proteolysis, and cell cycle regulation, suggests that COS may further support cellular adaptation and restoration of macrophage homeostasis under inflammatory stress.

2.7.2. Transcriptomic Identification of Key Differentially Expressed Genes Involved in Inflammatory Regulation

2.7.2.1. Inflammatory relevance

Transcriptomic profiling of COS-treated macrophages revealed coordinated differential expression across the principal innate-immune and atherosclerosis-relevant pathways, including Toll-like receptor, NOD-like receptor, NF-κB, MAPK, TNF, HIF-1, and lipid-and-atherosclerosis signaling (Table 2). The most pronounced change was the strong downregulation of the inducible nitric oxide synthase Nos2 (iNOS; log₂FC −7.92, q = 1.05 × 10⁻²⁶), providing a transcriptional correlate for the reduced nitrite measured by the Griess assay, since iNOS is the enzymatic source of macrophage-derived NO. This effector-level suppression was accompanied by concordant downregulation of the prostaglandin-synthesizing enzyme Ptgs2 (COX-2; −0.84), the pro-inflammatory cytokines Il1b (−2.94), Il18 (−1.94), and Il6 (−1.80), the chemokine Ccl5 (−5.42), and the stress-activated kinases Mapk14 (p38α; −5.74) and Mapk3 (ERK1; −1.59), all at stringent FDR thresholds. In contrast, the upstream pattern-recognition and signal-transduction apparatus was transcriptionally induced, including the Toll-like receptors Tlr2, Tlr4, and Tlr6, the inflammasome sensor Nlrp3, the adaptors Irak4 and Traf6, the IκB kinase subunits Ikbkb (IKKβ) and Ikbkg (NEMO) and the NF-κB subunit Rela (p65), with reciprocal downregulation of the inhibitor Nfkbia (IκBα). The antioxidant regulator Nfe2l2 (NRF2; +1.59) was upregulated, consistent with the DPPH result, whereas the endothelial synthase Nos3 (eNOS; −2.56) was downregulated, opposite in physiological consequence to inducible NOS. Taken together, COS most strongly attenuated the terminal effector arm of macrophage inflammation, in agreement with the measured reduction in NO, while the upstream TLR–NF-κB–inflammasome machinery did not track the effector outcome and cannot, alone, be interpreted as pathway inhibition.

2.7.2.2. Foam Cell Formation

Within the lipid-and-atherosclerosis pathway, COS treatment produced coordinated differential expression of the nuclear-receptor and calcium-dependent transcriptional regulators that govern macrophage lipid handling (Table 3). The PPARγ–RXRα heterodimer partners Rxra (RXRα; log₂FC +9.29, q = 9.67 × 10⁻⁵⁸) and Pparg (PPARγ; +5.24) were among the most strongly upregulated transcripts, together with the PPARγ target gene Cd36 (CD36/FAT; +2.21), a scavenger receptor for oxidized and modified lipoproteins. The calcineurin–NFAT arm was coordinately induced, comprising the calcineurin catalytic subunit Ppp3cb (calcineurin Aβ; +3.97) and all three NFAT isoforms Nfatc1(+6.00), Nfatc2 (+7.94), and Nfatc3 (+2.47), indicative of activated Ca²⁺/calcineurin-dependent transcription, all at stringent FDR thresholds. In parallel, genes controlling lipoprotein uptake and reverse cholesterol transport were downregulated, including the lectin-like oxidized-LDL receptor Olr1 (LOX-1; −4.92, q = 1.05 × 10⁻⁴), the low-density lipoprotein receptor Ldlr (−1.41), and the efflux transporters Abca1 (ABCA1; −0.89) and Abcg1 (ABCG1; −1.11). The pronounced suppression of Olr1/LOX-1, the largest-magnitude change among the lipid-uptake receptors, is consistent with attenuated scavenging of oxidized LDL, a primary driver of foam-cell initiation, while reduced Ldlr further limits receptor-mediated lipoprotein internalization. Taken together, these data indicate that COS engages the principal PPARγ–RXRα and calcineurin–NFAT programs governing macrophage cholesterol metabolism and identifies lipoprotein uptake and cholesterol efflux as candidate pathways for targeted functional validation, including cholesterol-efflux and foam-cell-formation assays, before the net effect on lipid handling can be assigned.

2.7.2.3. Atherosclerosis-Associated Pathways

Within the lipid-and-atherosclerosis pathway, COS treatment produced coordinated differential expression of Rho-family GTPase and actin-cytoskeletal signaling, which governs macrophage adhesion, polarization, and directional migration (Table 4). Upregulated transcripts were concentrated in the RhoA–ROCK2 contractility axis, comprising Rhoa (log₂FC +3.42, q = 2.31 × 10⁻²⁸) and Rock2 (+2.24), the small GTPase Cdc42 (+1.81), and their upstream guanine-nucleotide exchange factors Vav2 (+2.33) and Vav3 (+2.31), indicative of enhanced activation of the GEF–GTPase circuitry that drives cytoskeletal remodeling and cell motility. Concurrent induction of the non-receptor tyrosine kinase Src (+1.02), the class IA PI3K catalytic subunit Pik3ca (+2.66), the Notch-ligand E3 ubiquitin ligase Mib2 (+7.68), and the transcription factor Pou2f1 (+6.89) is consistent with broad activation of receptor-proximal kinase signaling coupled to cytoskeletal output, all at stringent FDR thresholds. In contrast, the downregulated set comprised effectors of the inflammatory and extracellular-matrix-remodeling program, including the immune-restricted PI3Kδ and Akt2 nodes Pik3cd (−7.33) and Akt2 (−9.73), focal-adhesion kinase Ptk2 (FAK; −6.42), the RhoGEF Arhgef1 (−2.58), the type I interferon master regulator Irf7 (−4.07), the monocyte chemoattractant Ccl12 (−3.77), the costimulatory receptor Cd40 (−3.59), and the matrix metalloproteinase Mmp3 (stromelysin-1; −4.33). Downregulation of Ccl12, Cd40, Mmp3, and Irf7 is consistent with attenuated chemotactic recruitment, costimulatory signaling, and ECM degradation, processes that promote monocyte infiltration and foam-cell formation. The endothelial synthase Nos3 (eNOS; −2.56) was also downregulated, a change opposite in physiological consequence to inducible Nos2 and warranting separate interpretation. Taken together, these data indicate that COS exerts a pathway-selective effect on Rho-GTPase–dependent motility and inflammatory effector signaling in macrophages rather than a global suppression of the program, and the cytoskeletal and migratory arm of this response warrants direct functional validation before any atheroprotective interpretation is assigned.

2.8. Integration of Molecular Docking and Transcriptomic Profiling Identifies the Potential Anti-Inflammatory Targets of COS

To identify potential molecular targets underlying the anti-inflammatory activity of COS, docking results were integrated with transcriptomic profiling of COS-treated LPS-induced RAW 264.7 cells. Among the predicted receptors, COS exhibited the strongest binding affinity toward COX-2 (−7.7 kcal/mol), while showing moderate interaction with iNOS (−7.1 kcal/mol). Transcriptomic analysis supported these observations by revealing significant downregulation of Nos2 (iNOS; log₂FC = −7.92) and Ptgs2 (COX-2; log₂FC = −0.84). Reduced Nos2 expression was consistent with decreased nitrite production measured by the Griess assay, whereas lower Ptgs2 expression suggested attenuation of prostaglandin-associated inflammatory signaling. Thus, the overlap between docking targets and differentially expressed genes identified iNOS and COX-2 as candidate anti-inflammatory targets of COS, supporting subsequent analysis of their molecular binding interactions.

The docking complexes of diclofenac and COS with COX-2 and iNOS are presented together with their corresponding two-dimensional interaction maps. In COX-2, diclofenac exhibited a stronger binding affinity than COS, despite COS showed a notable strength in forming a broader and more diversified interaction network, characterized by multiple conventional hydrogen bonds and van der Waals interactions with residues distributed throughout the pocket (Figure 4A-B) COS interacted with Gln178 and Tyr341, which are key residues involved in inhibitor recognition and binding within the COX-2 active site [21]. This pattern suggests that COS may stabilize receptor binding through extensive polar contacts and flexible accommodation within the active site rather than relying predominantly on hydrophobic interactions.

Similarly, for iNOS, diclofenac showed greater binding affinity than COS. However, COS maintained stable binding through multiple hydrogen-bond interactions and contacts with residues located within the catalytic region, including Glu377, Tyr373, Tyr347, Asp382, Arg388, and Gln263, which have been reported to contribute to inhibitor recognition and selective iNOS binding (Figure 4C-D) [22]. These interactions suggest favorable receptor accommodation despite the lower predicted binding affinity. The binding profile of COS indicates its potential to engage inflammatory targets through a distinct interaction mode supported by extensive polar contacts, which may contribute to stable receptor association and modulation of inflammatory signaling.

Integration with transcriptomic analysis further demonstrated significant downregulation of Nos2 (iNOS) and Ptgs2 (COX-2) following COS treatment, supporting these receptors as convergent anti-inflammatory targets. Together, these findings suggest that COS may suppress inflammatory signaling and reduce macrophage inflammatory responses through stable and multifaceted interactions with COX-2 and iNOS.

3. Discussion

This study investigated the precisely defined low molecular weight distribution of the hydrolyzed COS prepared by hydrochloric acid that possessed antioxidant and anti-inflammatory properties. This COS obtained from mud crab shell waste could sustainably promote local and industrial economy as the daily waste was turned into a high-value biomedical material. The current study differs from many previous studies that used large, diverse and poorly characterized COS mixes. Advantages of HCl-derived COS offer a rapid, cost-effective, and highly reproducible approach compared to enzymatic digestion, oxidative degradation or chemical depolymerization [18,19]. Subsequently, the structural identity and integrity of HCl-derived COS confirmed through ¹³C NMR spectroscopy and MALDI-TOF mass spectrometry ensured reproducibility for downstream biological and mechanistic investigations [23,24]. Previous studies from Guo et al (2018) reported that 1H NMR spectra of two COS samples, finding major signals at ~2.0 ppm (CH3 of N-acetyl group) and ~3.1 ppm (H-2 of GlcN) and calculated degrees of deacetylation (DD) of 88.4% and 87.4% respectively [9]. In this study, ¹³C NMR and MALDI-TOF were used to verify glycosidic integrity and linkage uniformity that COS used in this study contained molecular sizes ranging between 318 to 498 Da, with a 344 Da as a principal constituent identified to be D-glucosamine dimer (DP2).This guided us to predict the chain length, charge and hydrogen-bonding networks of COS molecules obtained from this study. Since low-molecular-weight oligosaccharides exhibit better solubility in aqueous environments, enhanced membrane permeability, and superior cellular uptake in comparison to their high-molecular-weight counterparts [25,26], the predominance of dimeric DP in this study was therefore significant to be expected for more effective intracellular bioactivity. In contrast to larger polymeric chitosan which is typically restricted to extracellular or membrane-localized effects [27,28], the precisely defined low molecular weight COS might be readily transported through cell membranes and entered intracellular signaling pathways [29,30]. Therefore, selectively modulate excessive inflammatory signaling without inducing nonspecific macrophage suppression.

This study differs from previous reports relying on structurally heterogeneous COS mixtures or chitosan derivatives with high-molecular-weight where variability in degree of polymerization and acetylation complicates mechanistic interpretation. Specifically, earlier studies demonstrated impact of COS in suppressing LPS-induced nitric oxide, iNOS, and COX-2 expression in RAW 264.7 macrophages [16,17], and reduce atherosclerotic lesion burden in ApoE⁻/⁻ mice [10] without precise structural definition. In contrast, the HCl-derived COS used in this study exhibits a narrowly defined low-molecular-weight profile with dimeric D-glucosamine, allowing accurate correlation between molecular structure and anti-inflammatory bioactivity. With low-molecular-weight COS reported to display superior solubility and intracellular accessibility compared with larger oligomers or parent chitosan polymers [9,11], the structurally verified COS applied in this study offers a clear mechanistic framework for researching atherosclerosis-relevant macrophage-driven inflammation.

The COS prepared in this study was proven for its antioxidant and anti-inflammatory properties. Since the basic mechanism of anti-inflammatory action of COS remains poorly defined [3,31] and existing works rarely identify the molecular receptors through which COS exerts these effects and few studies have contextualized COS activity within macrophage-driven atherogenesis [12,16]. In this study, we speculated that COS played its intracellular actions by infiltrating into macrophage cells due to its small size molecular properties. This was also reported in study of COS in cellular uptake, where COS effectively gets into the cells to play its actions [29,30,32]. COS displayed non-cytotoxic and non-proliferating effects, confirming high biocompatibility of its molecules and this property is confirmed by other COS studies [28,32]. As in case of atherosclerosis, excess ROS causing LDL oxidation and oxidative damage could trigger inflammatory cytokines could activate macrophage to form the foam cells [2,4]. Therefore, treatment of COS in RAW 264.7 cell culture could be suitable model for studying the therapeutic effects of COS on arteriosclerosis.

Inflammatory cytokines that activate macrophages involve multiple pathways, i.e., NF-kB, MAPK, and nitric oxide synthase pathways [3]. To ensure that COS effectively attenuate these multiple pathways in atherosclerosis, the proof that COS firmly binds with key molecules is essential to this hypothesis. In this study, we elucidated putative molecular mechanisms underlying ROS-induced macrophage activation and we employed molecular docking of COS against eleven inflammation-related target proteins selected based on their established roles in macrophage activation, oxidative stress, angiogenesis, extracellular matrix remodeling, and foam-cell formation. Together, these targets – AKR1B1, FGF1, FGF2, HPSE, HSP90AA1, TRPV1, VEGFA, COX-2, TNF-α, inducible nitric-oxide synthase, and galectin-3 – represented the main metabolic axis of vascular inflammation mediated by macrophages [3,4,31]. Therefore, rather than focusing on a single target mechanism, the selection of this receptor panel was intended to highlight the multi-pathway character of atherogenesis.

Docking was performed as a predictive strategy, as it enabled rapid, structure-based hypothesis generation regarding ligand–target interactions. Diclofenac is a clinically proven NSAID that primarily acts by inhibiting COX and suppressing prostaglandin formation [33,34], so it was chosen as a positive control. Diclofenac is useful for treating acute inflammation, but it has negative effects on the gastrointestinal tract, kidneys, and heart [33], which highlights the need for safer substitutes. In this study, COS showed broad affinity to multiple receptor molecules for macrophage activation in atherogenic process. Strongly predicted interaction of COS with COX-2 provided structural support for its observed suppression of NO production and inflammatory signaling in RAW macrophages. Similarly, COS binding to galectin-3 and VEGFA points to additional ways to prevent pathological angiogenesis, macrophage migration, and foam-cell formation in atherosclerotic plaques. Additionally, binding to HPSE suggests a possible function in preventing plaque destabilization and extracellular matrix degradation. While various COS preparations reported suppression of LPS-induced NO [16,17], these studies largely interpreted NO reduction as an endpoint and left out molecular-level interactions between COS and inflammatory regulators. However, in this study, docking analyses were used to support observed attenuation of nitrite accumulation and demonstrated stable COS binding to multiple key inflammation targets connected to atherosclerosis, including COX-2, galectin-3, VEGFA, FGF1, and HPSE. By linking a functional inflammatory readout with predicted receptor-level engagement this approach extends beyond prior phenomenological observations of NO suppression alone.

To move beyond predicted binding and toward expressed mechanism, transcriptomic profiling of COS-treated macrophages was cross-referenced against the docking panel. Of the receptors selected for docking, four were represented among the differentially expressed transcripts (COX-2/Ptgs2, inducible and endothelial nitric oxide synthase/Nos2 and Nos3, and TNF-α/Tnf), permitting a direct comparison between predicted engagement and transcript-level response; the remaining targets were not differentially expressed in this dataset and could not be evaluated in this manner. The most coherent convergence was observed for the inducible nitric oxide synthase axis. COS suppressed nitrite accumulation in the Griess assay, and Nos2 (iNOS), the enzymatic source of macrophage-derived NO [35], was the most strongly downregulated transcript in the dataset (log₂FC −7.92, q = 1.05 × 10⁻²⁶). Together with the inclusion of NOS in the docking panel, this provides converging evidence at three levels, functional (reduced nitrite), transcriptional (reduced Nos2), and structure-based (predicted COS–NOS engagement), that the observed NO suppression reflects downregulation of inducible NO synthesis rather than a non-specific endpoint. A comparable convergence was found for COX-2, for which COS exhibited its highest predicted affinity, alongside concordant downregulation of Ptgs2 (log₂FC −0.84). The agreement between predicted catalytic-pocket occupancy and reduced transcript is consistent with a dual mode of action in which COS may both bind and limit expression of COX-2, supporting attenuation of prostaglandin-driven signaling [36]. As transcript abundance does not establish a corresponding change in protein or catalytic activity, immunoblot confirmation of iNOS and COX-2 remains necessary before structural and transcriptional concordance can be interpreted as functional inhibition.

The endothelial isoform Nos3 (eNOS) was downregulated (log₂FC −2.56), and because eNOS-derived NO is vasoprotective [37], its suppression is not a favorable outcome in the atherosclerotic context. This distinction is mechanistically important and indicates that the inducible and endothelial synthases must be treated as functionally separate targets rather than a single NOS entity, where the protective endothelial enzyme and the inflammatory inducible enzyme should not be conflated. In addition, although TNF-α was included in the docking panel and COS displayed moderate predicted affinity for it (−6.7 kcal·mol⁻¹), the Tnf gene was modestly upregulated (log₂FC +0.45) rather than suppressed, suggesting that any influence of COS on TNF-α is more plausibly exerted through direct binding or post-transcriptional mechanisms than through transcriptional repression [38]. A similar caution applies to the broader pattern of the transcriptomic data: while the terminal effectors of macrophage inflammation (iNOS, COX-2, and the IL-1β/IL-6/IL-18 cytokine axis) were coordinately downregulated in agreement with the NO result, the upstream Toll-like receptor, NF-κB, and inflammasome machinery was transcriptionally induced [39], and within the lipid-handling program the simultaneous upregulation of the PPARγ–RXRα–CD36 axis [40] and downregulation of the ABCA1/ABCG1 efflux transporters [41], together with elevated RhoA/ROCK2 signaling [42], did not uniformly track an atheroprotective direction. The clearest favorable signals within these arms were the strong suppression of the oxidized-LDL receptor Olr1 (LOX-1) [43] and of the chemotactic and costimulatory mediators Ccl12, Ccl5, Cd40, and Mmp3 [44].

Considered together, the integration of docking predictions with transcript-level data strengthens the mechanistic interpretation of COS activity specifically for the COX-2 and inducible-NOS axes, where structure-based, transcriptional, and functional readouts align, while delimiting the broader claim. The eNOS, TNF-α, and lipid-handling results indicate that predicted binding and transcript abundance do not uniformly translate into an anti-inflammatory or atheroprotective outcome, and that the net effect of COS is target-selective. This convergent-and-divergent pattern is consistent with the interpretation of COS as a multi-target modulator rather than a uniform single-pathway inhibitor, and it identifies COX-2 and iNOS as the most defensible nodes for subsequent protein-level and functional validation (iNOS/COX-2 immunoblot, phospho-NF-κB p65 and MAPK, and cholesterol-efflux and foam-cell assays), before any plaque-level protective effect can be assigned.

In summary, this study presented the precisely defined COS molecules obtained by HCl hydrolysis, in performing antioxidant and anti-inflammatory actions. The molecular docking of COS by strong interactions with other candidates involving macrophage activation in atherogenic processes enhanced the possibility of COS in having preventive and therapeutic functions against atherosclerosis. From a translational perspective, COS emerges as a promising candidate for development as a nutraceutical, preventive supplement, or lead compound for anti-atherosclerotic drug design. Future research should focus on candidate gene expression profiling of COS-treated macrophages including the in vivo atherosclerosis models to confirm plaque-level protection.

4. Materials and Methods

4.1. Extraction and Hydrolysis of Chitosan Oligosaccharides (COS)

Shells of mud crab, S. olivacea were collected from Coastal Aquaculture Research and Development Regional Center 2, Samut Sakhon, Thailand. The waste crab shells were carefully rinsed and dried in an oven at 60 °C for 6 hours. Subsequently, dried crab shell was homogenized in a blender into small even pieces (<20 mesh) and maintained frozen until used. The demineralization step was carried out by integrating 10% HCl 1:20 (w/v) with crab shells. The reaction proceeded at room temperature under agitation at 250 rpm for 6 h. Afterwards, the demineralized shells were filtered and washed with distilled water until neutralized. They were bleached in ethanol for 10min and dried in an oven at 70 °C. Deproteinization step was performed by integrating 1 M NaOH with the dried demineralized shells at a solid/liquid ratio of 1:20 (w/v). The reaction was carried out under agitation at 70 °C or 75 °C for 24 h. The solid was filtered and washed with distilled water until it reached neutral pH. It was then immersed in ethanol for 10 min for further bleaching, and resulting chitin was dried in an oven at 60 °C. Chitin was deacetylated by reacting it with 12.5 M NaOH at a solid/liquid ratio of 1:20 (w/v) under agitation at 70 °C for 6 or 9 h and washed with distilled water until it neutralized. Obtained Chitosan was dried in the oven at 60 °C for 4 h.

Chitosan (1g) was then dissolved in 100 mL of 2 M HCl solution and heated at 70 °C for 2 hours. The resulting solutions were then precipitated by adding 95% Ethanol in a 1:2 (v/v) ratio of COS to Ethanol, followed by incubation overnight at 4 °C to complete the precipitation. The mixture was then centrifuged at 4000 xg for 5 minutes, and precipitation was collected to be washed with 70% ethanol for 5 times to remove residual acids and impurities. After that, purified products were then ultimately dried at 50 °C to obtain COS powder.

4.2. Characterization and Identification of COS

The molecular weight distribution of COS was determined by MALDI-TOF mass spectrometry using a SpiralTOF MALDI Imaging-TOF/TOF instrument (JEOL, Japan) operated in positive ion mode with 2,5-dihydroxybenzoic acid (DHB) as the matrix. Mass spectra were collected over the selected m/z ranges, and ion signals were detected mainly as protonated and sodium-associated species. The structural characteristics of COS were further confirmed by solid-state 13C CP/MAS NMR spectroscopy using a 400 MHz FT-NMR spectrometer (Bruker, USA). Spectral analysis was performed to identify characteristic carbon resonances corresponding to the glucosamine backbone and residual N-acetyl groups. Chitosan oligosaccharide lactate was used as a reference standard, described previously [14].

4.3. Determination of Antioxidant property of COS

4.3.1. DPPH Free Radical Scavenging Assay

The antioxidant activity of COS was determined using the 2,2-diphenyl-1-picryl hydrazyl (DPPH) radical scavenging method. A DPPH working solution (0.2 mM) was freshly prepared in ethanol according to previously reported procedures. For the assay, 100 µL of COS at different concentrations (12.5, 6.25, 3.12, 1.60, and 0.80 mg/mL) was dispensed into a 96-well microplate and mixed with 100 µL of DPPH solution. Each treatment was conducted in triplicate. Following incubation at room temperature for 30 min under protected conditions, absorbance was recorded at 517 nm using a Multiskan SkyHigh Plate Reader (Thermo Fisher Scientific, Waltham, MA, USA). Ascorbic acid (vitamin C) and Trolox were included as reference antioxidants. For the blank group, an equivalent amount of Type I water served as the substitute for the sample. Each group underwent two parallel measurements. The hydroxyl radical scavenging rate was then calculated using the formula below:

$$\mathit{Hydroxyl}\mathit{radical}\mathit{scavenging}(\%)={\displaystyle \frac{A0-A1}{A0}}\text{×}100$$

4.3.2. ABTS Radical Scavenging Assay

ABTS radical cations were prepared by reacting 7 mM ABTS solution with 70 mM potassium persulfate at a ratio of 100:1 (v/v) and incubating the mixture in the dark at room temperature for 12–16 h. Then, the stock solution was diluted with distilled water to obtain an absorbance of approximately 0.700 at 745 nm. For the assay, 10 µL of COS at the same concentrations as described above was mixed with 100 µL of ABTS working solution in a 96-well plate and incubated in the dark for 5 min. Absorbance was then measured at 734 nm using a microplate reader. The initial control was the same as described before. The percentage of radical scavenging activity (%) was calculated using the same equation as described for the DPPH assay, and the IC₅₀ value was determined from the concentration–response curve.

4.4. Cell Experiment

4.4.1. Cell Culture

RAW 264.7 murine macrophage cells (TIB-71; ATCC, VA, USA) were obtained from the American Type Culture Collection. Cells were cultured in complete growth medium (GM) consisting of Dulbecco’s Modified Eagle Medium with high glucose (DMEM-HG; Cytiva, South Logan, Utah) supplemented with 10% fetal bovine serum (FBS; ATCC, VA, USA) and 100 units/mL penicillin plus 100 µg/mL streptomycin (P/S; Gibco, Thermo Fisher Scientific, CA, USA). Cultures were maintained at 37 °C in a humidified atmosphere of 5% CO₂ and 95% air. The growth medium was replaced every 2-3 days, and cells were then subcultured by gentle scraping prior to reaching confluence.

4.4.2. Cytotoxicity Test Using MTT Assay

For the MTT (3-(4,5-dimethylthiazolyl-2)-2,5-diphenyltetrazolium bromide) assay, RAW 264.7 cells were seeded into 96-well plates at a density of 2 × 10³ cells per well in triplicate and allowed to adhere for 24 hours. After incubation, the cells were treated with COS at a range of concentrations (20, 40, 80, 160, 320, 640, and 1280 µg/mL) for 24 hours and 48 hours. Following COS exposure, the culture medium was removed and replaced with 100 µL of 0.5 mg/mL MTT reagent (Sigma-Aldrich). Cells were then incubated for 2 hours at 37 °C under 5% CO₂ to allow intracellular reduction of MTT to insoluble formazan crystals. After incubation, the MTT solution was carefully aspirated, and the formazan precipitate was dissolved with 100 µL dimethyl sulfoxide (DMSO, Merck). Absorbance was measured at 570 nm with a 630 nm reference wavelength using a microplate reader (Varioskan Flash, Thermo Scientific, Finland). All experiments were performed with at least five replicates. Cell viability (%) was calculated relative to vehicle control using the formula:

$$\%\mathit{Cell}\mathit{viability}={\displaystyle \frac{{\mathit{OD}}_{\mathit{COS}\mathrm{treatment}}-{\mathit{OD}}_{\mathit{COS}\mathit{control}}}{{\mathit{OD}}_{\mathit{cell}\mathit{control}}-{\mathit{OD}}_{\mathit{medium}\mathit{control}}}}\text{×}100$$

4.4.3. Determination of NO Content

Nitric oxide production in RAW 264.7 cells was quantified using the Griess reagent method (N-(1-naphthyl) ethylenediamine dihydrochloride (Component A), Sulfanilic acid (Component B) G7921). Following a 2-hour treatment, cells were incubated with LPS (µg/mL) for 24 hours, after which the culture supernatant was collected. For the microplate assay, 130 µL of deionized water, 20 µL of Griess reagent, and 150 µL of each sample or sodium nitrite standard were added to individual wells. The plate was incubated at room temperature for 15 minutes, and absorbance was measured at 548 nm. Nitric oxide levels were determined by interpolating sample absorbance values from a sodium nitrite standard curve and expressed as nitrite concentration.

4.5. Structural Analysis and Selection of Inflammatory Target Proteins

The Swiss Target Prediction database (http://www.swisstargetprediction.ch/, accessed on 21 Oct 2025) was used to predict target disease anti-inflammation specifically for COS. COS was selected for molecular docking verification against the 7 core inflammatory targets: AKR1B1 from Aldose reductase, TRPV1 from Transient receptor potential cation channel subfamily V member 1 (PDB ID: 7LPE), HSP90AA1 from Cytosolic (PDB ID: 1UY6), HPSE from endo-b-glucuronidase (PDB ID: 5E8M), FGF1 from acidic FGF (PDB ID: 1RG8), FGF2 from basic FGF (PDB ID: 1BFG), and VEGFA from Vascular endothelial growth factor A (PDB ID: 1VPF). COX-2 from Cyclooxygenase-2 (PDB ID: 3LN1), TNF-α from Tumor necrosis factor alpha (PDB ID: 2AZ5), human inducible nitric oxide synthase (iNOS) (PDB ID: 4NOS), and Galectin-3 were obtained from a previous research. These receptors represent key regulators of macrophage activation, which drive foam cell formation. The 3D structures were retrieved in PDB format from the RCSB Protein Data Bank (https://www.rcsb.org/, accessed on 21 Oct 2025) and were virtualized by using Discovery Studio (version 2024). COS ligand structure was downloaded in SDF format from PubChem (https://pubchem.ncbi.nlm.nih.gov/, accessed on 21 Oct 2025).

4.6. Molecular Docking of Chitosan-Oligosaccharide (COS) with Inflammation-Related Target Receptors

All three-dimensional structures of the target proteins and the selected ligands used and diclofenac as positive ligand, with their SDF structures obtained from PubChem (https://pubchem.ncbi.nlm.nih.gov/, accessed on 21 Oct 2025). Molecular docking was performed using CB-Dock2 (https://cadd.labshare.cn/cb-dock2/ index.php, accessed on 21 Oct 2025), which provided binding cavity parameters (A) [20] and binding docking score (kcal/mol). The resulting protein–ligand complexes were visualized and analyzed with Discovery Studio (version 2024).

4.7. Gene Expression of RAW 264.7 Cells by Transcriptomic Analysis

To evaluate the effect of COS on gene expression, RAW 264.7 cells were pretreated with or without COS (160 μg/mL) prior to LPS-induced inflammation, as described above. Total RNA was extracted from three independent biological replicates using TriPure™ Isolation reagent (Roche, Germany). Equal amounts of RNA from each replicate were pooled and submitted to BGI Genomics (Hong Kong) for RNA quality assessment and mRNA sequencing using the DNBseq platform (150 bp read length). Raw sequencing data were filtered to remove adapters and low-quality reads prior to alignment to the mouse reference genome using HISAT2 (version 2.0.4) and Bowtie2 (version 2.2.5). Gene expression levels were quantified as FPKM and TPM values. Differentially expressed genes (DEGs) were identified using DESeq2 with thresholds of Q value ≤ 0.05 and log₂FC ≥ 0.2. Functional enrichment analyses, including KEGG enrichment pathways, were performed using a hypergeometric test with significance defined at Q value ≤ 0.05 [45]. Transcriptomic analyses were conducted using the Dr. Tom platform.