1. Introduction
In December 2019, coronavirus disease 2019 (COVID-19) emerged as a global health crisis. COVID-19 is caused by a novel coronavirus and spreads through respiratory droplets, indirect contact, and aerosols [
1], with an average incubation period of 5.2 days [
2]. As the disease progresses, approximately 15% of COVID-19 cases become severe COVID-19 cases [
3]. The transition of COVID-19 infection from mild to severe is often due to the interweaving of multiple factors, including atelectasis of the lung, inadequate nasal ventilation that increases the risk of aspiration pneumonia, edema of the mediastinum and upper respiratory tract caused by a persistently high heart rate, impaired sputum excretion leading to exacerbated pulmonary infections, damage to renal function, and the deterioration of basic health. Within a week of symptom onset, patients with severe COVID-19 frequently experience dyspnea and hypoxemia, which can rapidly escalate into life-threatening conditions such as acute respiratory distress syndrome (ARDS), septic shock, metabolic acidosis, coagulation dysfunction, and multiorgan failure [
4]. The pathological mechanism underlying the transition from mild to severe COVID-19 mainly involves the overactivation of the immune system and the occurrence of cytokine storms, leading to damage and dysfunction in the lungs and other vital organs. The virus-host cell interaction triggers the excessive release of proinflammatory cytokines (e.g., interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α)), leading to pulmonary alveolar injury and extrapulmonary dysfunction [
5]. Immune dysregulation, which is characterized by the activation of immune cells and the release of inflammatory cytokines, plays a crucial role in COVID-19 pathogenesis. Modulating the immune system through immunoregulatory agents [
6] or traditional Chinese medicines [
7] can eliminate excessive inflammation and reduce tissue damage. Targeting cytokine storms, a characteristic feature of severe COVID-19, by inhibiting the release of inflammatory cytokines or blocking their signaling pathways can arrest disease progression and decrease disease severity. Furthermore, underlying conditions such as cancer or cardiovascular diseases may be exacerbated by COVID-19. Addressing these comorbidities is essential to prevent mild cases from progressing to a severe disease [
8]. Considering the high mortality rate among patients with severe COVID-19, an urgent need exists for effective therapeutic strategies to impede disease progression and alleviate the associated pathological changes.
Fucosylation, a key form of protein glycosylation, is intricately involved in the progression of COVID-19, spanning from mild to severe pneumonia. Fucosylation involves the attachment of fucose residues to
N-glycans,
O-glycans, and glycolipid molecules, profoundly influencing their functional properties [
9]. The synthesis of fucosylated glycans is orchestrated by fucosyltransferases (FUTs), a diverse family of hexosyltransferases that catalyze the addition of
L-fucose residues. To date, the human genome is known to encodes 13
FUT genes, which can be categorized into four distinct groups based on the specific type of fucose linkage they form. FUT1 and FUT2 are pivotal in the biosynthesis of H blood group antigen and sialic acid Lewis Y antigen [
10]; FUT3-7 and FUT9-11 are responsible for the synthesis of sialic acid Lewis X antigen [
11]; FUT8 is an important enzyme responsible for core fucose modification of
N-glycan chains [
12]; and protein O-FUT1 (POFUT1) and POFUT2 are enzymes specifically responsible for the fucosylation of proteins containing epidermal growth factor-like repeat sequences and thrombospondin type 1 repeat sequences [
13]. Fucosylation plays a multifaceted role in various human biological processes, including protein folding, trafficking, localization, cell differentiation and development, cell recognition, signal transduction, and the immune response. In the context of COVID-19, fucosylation may serve as a crucial regulatory factor in disease progression. It potentially modulates antibody-dependent cellular cytotoxicity (ADCC) and influences the activation of inflammatory pathways, thereby contributing to the transition from mild to severe manifestations of COVID-19.
Glycosylation is considered a potential therapeutic target for multiple lung diseases. Studies have indicated that the targeted modulation of protein glycosylation can regulate respiratory inflammation and serve as a new therapeutic target for chronic obstructive pulmonary disease (COPD) [
14]. Additionally, research has revealed that targeting the
O-glycosyltransferase GALNT2 is a key therapeutic approach for overcoming radiation resistance in lung cancer [
15]. Furthermore, monoclonal antibodies that specifically target glycosylation pathways have become ubiquitous in the therapeutic landscape for numerous diseases. Obinutuzumab, a glycoengineered therapeutic anti-CD20 antibody, is widely used in combination therapy for patients with chronic lymphocytic leukemia (CLL) and follicular lymphoma. Clinical trials have demonstrated that integrating these monoclonal antibodies with chemotherapy regimens significantly prolongs the overall survival and progression-free survival of CLL patients while increasing the complete response rate [
16].
In this study, we explored alterations in fucosylation patterns among patients with severe COVID-19. We subsequently directed our efforts toward targeting fucosylation to identify potential therapeutic agents for the treatment of this severe disease. Our findings contribute significantly to the development of a novel strategic framework to overcome this formidable global health challenge.
2. Methods
2.1. Collection of samples from patients with COVID-19 and detection of fragment crystallizable of immunoglobulin G (IgG-Fc) glycopeptide levels
A case-control study was conducted at the Second Affiliated Hospital of Guangzhou University of Chinese Medicine, China and approved by the Ethics Committee of the hospital (approval No. ZE2023-163-01), in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants prior to their enrollment in the study. The participants were provided with detailed information about the study’s purpose, procedures, potential risks, and benefits, and their consent was documented. According to the World Health Organization (WHO) COVID-19 case definition, COVID-19 is defined as a positive result from a nucleic acid amplification test (NAAT), primarily quantitative reverse transcription polymerase chain reaction (qRT-PCR), or a rapid diagnostic test (RDT) for the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) antigen (SARS-CoV-2 antigen-RDT) within 48 h, depending on the availability of resources and the clinical context. qRT-PCR was used in 42.37% of the cases, whereas rapid antigen tests were used in 57.63% of the cases. The inclusion criteria were as follows: ① aged ≥ 18 years and ② met the COVID-19 case definition. The exclusion criteria were as follows: ① patients with autoimmune diseases, advanced malignancies, or severe liver or kidney dysfunction; ② patients with gastrointestinal ulcers or bleeding requiring fasting; ③ patients who had used corticosteroids or other medications with a clear impact on immune function within the past month; ④ patients with a life expectancy of less than 7 d due to underlying medical conditions. The control group was rigorously selected based on the following criteria: negative results for SARS-CoV-2 infection through NAAT or rapid diagnostic testing; the absence of clinical symptoms associated with COVID-19 (fever, cough, sore throat, or nasal congestion); no evidence of infection with other pathogens, as determined through comprehensive evaluation of vital signs and background medical information; and no history of SARS-CoV-2 infection, confirmed through a medical history review and antibody testing.
Sixty COVID-19 patients and 30 individuals who served as controls were included. Demographic data, underlying health conditions, and other pertinent information were extracted from the electronic medical records. The severity of COVID-19 in patients was assessed according to the WHO ordinal scale [
17] and the WHO Clinical Progression Scale [
18]. Venous blood samples (4 mL each) were obtained between 6:00 and 8:00 AM and centrifuged at 3000 r∙min
−1 for 10 min to isolate the serum, which was then aliquoted into Eppendorf (EP) tubes, sealed, and stored at −80 °C until further testing.
Using the methods described in a previous article [
19], we purified protein G from human serum and then prepared IgG-Fc glycopeptides through trypsin digestion. The quantitation of the IgG-Fc glycopeptides was performed using an Agilent 1290 Infinity ultra high-performance liquid chromatography (UHPLC) system (Agilent Technologies, USA) coupled to an Agilent 6495 QQQ mass spectrometry (MS) spectrometer (Agilent Technologies). Chromatography was performed on an Agilent ZORBAX Eclipse Plus C18 column (2.1 mm × 100 mm, 1.8 μm; Agilent Technologies). The mobile phase consisted of 0.1% formic acid in Milli-Q water (A) and 0.1% formic acid in acetonitrile (B). The flow rate was 0.3 mL∙min
−1, and the linear gradient was optimized as follows: 0-8 min, 2% to 6% B; 8-16 min, 6% to 10% B; 16-20 min, 10% to 35% B; 20-25 min, 35% to 50% B; 25-30 min, 50% to 100% B; 30-32 min, 100% B; 32.1-40 min, followed by equilibration with 2% B for 8 min. The injection volume was 10 μL and the column temperature was maintained at 40 °C. The Agilent Jet Stream electrospray ionization (AJS-ESI) was performed in positive mode and the source parameters were as follows: The dry gas temperature and flow rate were 200 °C and 16 L∙min
−1, nebulizer pressure was 35 psi (1 psi = 6.89 kPa), sheath gas temperature and flow rate were 300 °C and 12 L∙min
−1, capillary voltage was 1.8 kV, nozzle voltage was 1200 V, and the radio frequency (RF) voltages of the high-pressure and low-pressure ion funnels were 200 and 150 V, respectively. The dwell time was set to 10 ms. The intraday and interday repeatability of the multiple reaction monitoring (MRM) method was evaluated by injecting the IgG-Fc glycopeptide sample three times within a day and on three consecutive days.
2.2. Transcriptomic analysis of COVID-19 patients
COVID-19-related datasets were retrieved from the Gene Expression Omnibus (GEO) database. Among the 15 databases employing whole blood samples, we carefully selected appropriate datasets for analysis, considering the sequencing platforms and diagnostic criteria. Consequently, GSE161731, GSE163151, and GSE157103 were chosen for the comprehensive analysis of transcriptome data from COVID-19 patients, with GSE172114 serving as a validation dataset for our findings. Within these selected datasets, we had a comprehensive representation of both healthy controls and COVID-19 patients. Specifically, GSE161731 included data from 19 healthy controls and 77 COVID-19 patients, GSE163151 included data from 20 healthy controls and 7 COVID-19 patients, and GSE157103 included data from 26 healthy controls and 100 COVID-19 patients.
Our initial step involved integrating and analyzing the data from these three GEO datasets. We delved into the differential expression patterns among COVID-19 patients with varying disease manifestations, generating insightful heatmap, volcano plots, and differential violin plots. Furthermore, we conducted a coexpression network analysis of the integrated gene set to visualize their intricate relationships through a coexpression network diagram.
We compared the targets from COVID-19 patients with a dataset of glycosyltransferases and glycosidases to identify the glycosidase-related genes that were targeted in COVID-19 patients. Differential gene expression was validated in the GSE172114 dataset, and receiver operating characteristic (ROC) curves were plotted and area under the curve (AUC) coefficients were calculated. Finally, using the R language ssGSEA package, we calculated the expression levels of differentially expressed genes and presented them in a heatmap, providing a comprehensive overview of their involvement in the immune response to COVID-19.
2.3. Single-cell transcriptomic analysis of COVID-19 patients
We employed two single-cell transcriptome datasets, GSE165080 and GSE192391, to validate the expression of glycosyltransferases in COVID-19 patients. Leveraging these databases, we generated t-distributed stochastic neighbor embedding (t-SNE) and uniform manifold approximation and projection (UMAP) maps to visualize the distinct subpopulations of cells. Subsequently, we scrutinized the expression patterns of glycosyltransferases, particularly FUT1-11 and α-L-fucosidase 1/2 (FUCA1/2), across these cell subpopulations and presented our findings in scatter plots, providing a comprehensive view of their cellular distribution.
2.4. Network pharmacology study of Fuzheng Jiedu Decoction (FZJDD) in the treatment of COVID-19
FZJDD is composed of Danfu tablets, dried ginger, licorice, honeysuckle, soapberry thorn, five finger peach, patchouli, and tangerine peel. FZJDD was recommended for use in patients with severe COVID-19 in the tenth edition of the Diagnosis and Treatment Protocol for Novel Coronavirus Infection issued by the National Health Commission of the People’s Republic of China. However, the mechanism through which FZJDD treats severe COVID-19 is still unclear. This study aimed to explore the mechanism through which FZJDD, treats severe COVID-19 and its effects on fucosylation.
The active ingredients of the FZJDD were obtained using the traditional Chinese medicine systems pharmacology database and analysis platform (TCMSP), symptom mapping (SYMMAP), encyclopedia of traditional Chinese medicine (ETCM), and high-throughput experiment and reference guided database of traditional Chinese medicine (HERB) databases and the literature. After removing duplicates, active ingredients were screened based on the oral bioavailability (OB) and drug likeness (DL) values. The potential action targets of FZJDD were compared with the genes of patients with COVID-19 to obtain possible COVID-19-related targets of the compounds. The FZJDD herb-component-target network was constructed. A protein-protein interaction (PPI) network was generated to map the interaction network between the FZJDD action targets and COVID-19 proteins. In addition, molecular docking of the above targets with the active components of FZJDD was performed to identify the active components that may play a role in treating COVID-19.
2.5. Effects of FZJDD on septic in mice
The mouse septic model represents a valuable tool for simulating the intense systemic inflammatory response observed in COVID-19 patients, particularly the “cytokine storm.” COVID-19 infection not only affects the respiratory system, but can also damage multiple organs, including the heart and kidneys. The mouse model of septic also has certain advantages in simulating multiorgan dysfunction syndrome, thereby assisting researchers in gaining a deeper understanding of the mechanisms underlying multiorgan damage caused by COVID-19 infection. Therefore, we chose mice with septic as the animal model.
Specific-pathogen-free (SPF) grade 6-8 week-old male C57BL/6 mice weighing 20-23 g were obtained from Guangdong Yaokang Biotechnology Co., Ltd. (China). The certificate number was 44824700027077. The mice were housed under controlled conditions with a temperature of (22 ± 1) °C and humidity of 40%-50%. The mice were divided into a control group, a model group, a positive drug group treated with dexamethasone (Dex), a group gavaged with 0.32 g∙mL
−1 FZJDD (FZJDD-L) for 7 d, a group gavaged with 0.64 g∙mL
−1 FZJDD (FZJDD-H) for 7 d, and a group gavaged with 0.64 g∙mL
−1 FZJDD for 4 h (FZJDD-H-4h). Dex and FZJDD were administered to the respective drug groups, adhering to the quality control measures outlined previously [
20]. All the mice, except those in the control group, subsequently received an intraperitoneal injection of lipopolysaccharide (LPS) to induce septic. The survival rate of the animals was monitored and recorded at 24 h post-injection. Lung tissues and serum samples were harvested from the mice for further analysis.
2.6. FZJDD intervention in cells transfected with FUT8/FUCA1 overexpression plasmids
qRT-PCR and Western blot (WB) analyses were performed to determine the messenger RNA (mRNA) and protein levels of FUT8 and FUCA1 in HEK-293 cells transfected with FUT8/FUCA1 overexpression plasmids following treatment with FZJDD-containing serum to assess the regulatory effects of FZJDD on glycosyltransferase gene expression.
2.7. Statistics
The data are presented as the mean ± standard errors of the means (SEMs). One-way analysis of variance (ANOVA) was used to compare multiple groups. The Wilcox-test was used to compare the COVID-19 patient group with the healthy control group. Pearson’s χ2 test or Fisher’s exact test was used to compare categorical variables in baseline characteristics. Pearson correlation analysis was performed to evaluate the relationship between clinical parameters and the abundance of IgG N-glycopeptides. A multivariable logistic regression analysis was performed to investigate the degree of association of IgG-Fc glycopeptides with severe COVID-19. Statistical significance was set at P < 0.05. Analyses and graphing were performed using GraphPad Prism 9.0 software (USA) and R 4.3.1 software.
3. Results
3.1. Correlation between the progression of COVID-19 and fucosylation levels
The basic information of the patients is shown in
Table 1. The results indicate no significant differences in demographic characteristics between the two groups, suggesting comparability. In terms of disease-specific characteristics, both body temperature and the presence of heart failure comorbidities were significantly different between the groups, reflecting the active features of the disease. Initially, we quantified galactosylation, fucosylation, and sialylation levels in COVID-19 patients, revealing a significant decrease in fucosylation levels compared with those of healthy controls (
Fig. 1(a)). An analysis of the IgG subclass expression patterns revealed an increase in IgG1 levels and a decrease in IgG2 levels (
Fig. 1(b)). Notably, the changes in FUT activity were observed specifically in COVID-19 patients with an IgG2 subclass. The regression analysis underscored the strong association between IgG2 and fucosylation levels (AUC value = 0.89,
Fig. 1(c)). A subsequent correlation analysis between IgG2 fucosylation levels and patient baseline characteristics indicated a negative correlation with body temperature; as the body temperature increased, IgG2 fucosylation levels decreased (
Fig. 1(d)). In summary, the reduction in fucosylation levels in COVID-19 patients is primarily associated with the IgG2 subclass and is linked to body temperature. In addition, no significant differences in the fucosylation levels of IgG1, IgG2, and IgG3/4 were observed in patients with different durations of symptom onset (Fig. S1 in Appendix A).
In the correlation analysis between the severity of COVID-19 and IgG-Fc glycopeptides levels, a significant association was observed between the levels of multiple glycopeptides and indicators of disease severity (
Fig. 1(e)). The differences in IgG fucosylation levels between COVID-19 patients classified based on the disease severity are illustrated in Fig. S2 in Appendix A. The variations in the abundance of IgG-Fc glycopeptides are presented in Fig. S3 in Appendix A. We constructed a multivariable logistic regression prediction model using four fucosylation-related glycopeptides: IgG2 3_5_0_0, IgG2 3_5_1_0, IgG2 5_4_0_1, and IgG2 5_4_1_1. The model demonstrated an AUC value of 0.74 (95% confidence interval (95% CI): 0.61-0.87) for the correlation between IgG2-Fc glycopeptides levels and COVID-19 severity, with a sensitivity of 0.90 (95% CI: 0.79-1.00,
Table 2). Notably, IgG2 3_5_1_0 and IgG2 5_4_1_1 emerged as independent risk factors of severe COVID-19 (
Figs. 1(f) and (g)).
3.2. Differentially expressed genes associated with glycosyltransferases and glycosidases in patients with severe COVID-19
We conducted a thorough transcriptomic analysis to compare COVID-19 patients with healthy controls, and subsequently validated our findings in a dataset of patients with severe COVID-19. The transcriptome datasets of GSE161731, GSE163151, and GSE157103 were shown in Fig. S4 in Appendix A. A total of 482 genes presented decreased expression and 1448 genes presented increased expression (
Fig. 2(a)). We constructed a hierarchical clustering tree based on their correlation coefficients of these genes to further explore the relationships among these genes. This tree distinctly segregated genes into distinct modules, each represented by a unique color (
Fig. 2(b)). Using weighted correlation analysis, we grouped genes with similar expression patterns into coherent modules. Notably, the red and blue modules were strongly correlated with COVID-19 patients, whereas the yellow and brown modules were more closely associated with healthy individuals. Among the 856 genes related to COVID-19, 290 genes involved in glycosyltransferase and glycosidase activities were identified. We obtained the shared genes by constructing a Venn diagram, which revealed a total of 13 shared genes, including
FUT7, chitotriosidase 1 (
CHIT1), UDP glycosyltransferase family 3 member A2 (
UGT3A2), hyaluronidase 3 (
HYAL3), lysozyme G2 (
LYG2), β-4-galactosyltransferase 2 (
B4GALT2), mannosyl (β-1,4-)-glycoprotein β-1,4-
N-acetylglucosaminyltransferase (
MGAT3), chondroitin polymerizing factor (
CHPF), sperm acrosome associated 3 (
SPACA3), α1,3-galactosyltransferase 2 (
A3GALT2), globoside α-1,3-
N-acetylgalactosaminyltransferase 1 (
GBGT1), β-1,4-
N-acetyl-galactosaminyltransferase 2 (
B4GALNT2), and polypeptide
N-acetylgalactosaminyltransferase 14 (
GALNT14) (
Fig. 2(c)). These differentially expressed genes were predominantly expressed in type 17 T helper cells, plasmacytoid dendritic cells, and immature dendritic cells (
Fig. 2(d)). Next, we examined the expression of eight of these genes in the GSE172114 dataset to validate the role of fucosylation in severe COVID-19. ROC curves were generated (
Fig. 2(e)), and AUC values were calculated. Notably, the
FUT7 and
GALNT14 genes had AUC values of 0.84 and 0.88, respectively, indicating their potential significance in the development of COVID-19.
3.3. Detection of the FUT family in cell subpopulations using single-cell datasets of COVID-19 patients
Using the single-cell datasets GSE165080 and GSE192391, we generated
t-SNE and uniform manifold approximation and projection (UMAP) plots to visualize the data (
Figs. 3(a)-(d)). Through scatter plots and violin plots, we detected and displayed the expression of FUT family members across various cell subpopulations. Our analysis revealed that FUT8 was predominantly expressed in plasma cells, whereas FUCA2 was expressed in plasma cells, dendritic cells, and monocytes (
Figs. 3(e) and
(f)). Furthermore, analysis of the GSE192391 single-cell dataset revealed that FUT8 was expressed primarily in platelets and B cells (
Figs. 3(g) and
(h)). These results indicate that the FUT family is expressed in the plasma cells and B cells of COVID-19 patients.
3.4. Network pharmacology analysis of the effect of FZJDD on severe COVID-19
Building upon the discoveries of fucosylation in patients with severe COVID-19, we aimed to validate whether fucosylation can serve as a therapeutic target for this disease. Therefore, we selected FZJDD and explored whether its therapeutic mechanisms are mediated by fucosylation.
By utilizing network pharmacology, we investigated the potential mechanisms by which FZJDD treats patients with severe COVID-19. We sourced the ingredients of FZJDD from multiple websites, including TCMSP, SYMMAP, ETCM, and HERB, to compile a comprehensive list. Each database revealed varying numbers of ingredients: TCMSP revealed 916, SYMMAP revealed 2057, ETCM revealed 418, and HERB revealed 1561 active ingredients. Additionally, our literature review revealed 17 unique components. By meticulously eliminating duplicates, we amassed a total of 1964 distinct pharmaceutical ingredients. Next, we leveraged the TCMSP database to screen these ingredients based on their OB and DL values, ultimately yielding 159 active ingredients. Using the Swiss Target Prediction, we identified 1019 potential drug targets associated with these ingredients (Table S1 in Appendix A). We subsequently constructed a component-target network diagram (
Fig. 4(a)), which visually illustrates the intricate relationships between the components of FZJDD and their respective targets. We further explored the relevance of FZJDD in severe COVID-19 by comparing its targets with known severe COVID-19-related genes, identifying 69 common genes (
Fig. 4(b)). Moreover, we generated a protein-protein interaction (PPI) network diagram, focusing on the intersection of proteins between FZJDD targets and severe COVID-19-related genes (
Fig. 4(c)). These findings underscore the potential of FZJDD as a multicomponent, multichannel, and multitarget therapy for severe COVID-19. We conducted molecular docking studies between multiple fucosylated proteins and the active components of FZJDD to validate our hypotheses. Encouragingly, our results demonstrated robust binding abilities between the active components and proteins such as FUT7, FUT8, FUCA1, and FUCA2 (
Fig. 4(d)). These findings suggest that FZJDD may modulate the functions of these proteins through fucosylation, contributing to its therapeutic effects on severe COVID-19.
3.5. Effects of FZJDD on mice with LPS-induced septic
LPS was intraperitoneally injected to establish a mouse septic model. Our findings revealed that during the administration of FZJDD from day 1 to day 6, no notable alteration in the body weight of the mice was observed. However, a notable decrease in body weight was observed on the seventh day after the LPS injection (
Fig. 5(a)). The survival rate curve was plotted based on mouse mortality (
Fig. 5(b)). Under these conditions, the mortality rate in the model group was 30%, whereas the positive control drug (Dex) group presented a slightly higher mortality rate of 40%. However, pretreatment with different doses of FZJDD one week in advance, the mortality rates were 20% and 40%, respectively. The survival rate curve of FZJDD-H group animals overlaps with that of Dex group. These finding suggest that FZJDD has a protective effect when it is administered sufficiently in advance of LPS challenge. In contrast, when FZJDD was administered only 4 h before LPS exposure, the mortality rate in the FZJDD group increased to 70%. These findings indicate that the 4 h pretreatment window was insufficient to confer protective effects, further emphasizing the importance of timing in the therapeutic efficacy of FZJDD. The qRT-PCR analysis of mouse spleen tissues also showed that FZJDD can significantly reduce the LPS-induced expression of the mRNAs encoding inflammatory factors, including
IL-1β,
IL-6,
IL-4, and
S100A8 (
Fig. 5(c)). The primers for the genes were shown in Table S2 in Appendix A. We also confirmed that FZJDD exerted significant protective effects in a murine model of LPS-induced acute lung injury (Fig. S5 in Appendix A).
The pathological examination of the lung tissues of septic mice revealed that the mice in the model group presented increased infiltration of inflammatory cells, alveolar collapse, and many red blood cells in the lung tissues, indicating the presence of pulmonary hemorrhage. In contrast, animals treated with FZJDD-L and FZJDD-H presented relatively normal lung tissue structures, with significantly reduced inflammatory cell infiltration and no red blood cell infiltration (
Fig. 5(d)). The infiltration of inflammatory cells and red blood cells in the lung tissues of FZJDD-H treated animals at 4 h was still relatively more significant. These pathological results suggest that FZJDD administered in advance could significantly alleviate LPS-induced septic-induced lung tissue injury. We obtained similar results for mice with LPS-induced acute lung injury, further validating the anti-inflammatory effect of FZJDD (Fig. S4).
Next, we examined the plasma of septic mice treated with FZJDD to investigate the potential changes in fucosylation levels analogous to those observed in COVID-19 patients. Our results revealed a significant reduction in the total plasma fucosylation level in septic mice (
Fig. 5(e)). Notably, a significant difference in the total fucosylation level was not observed between the model group and the positive drug control group or the FZJDD group. Furthermore, we investigated the fucosylation levels of various isoforms. Intriguingly, IgG2 fucosylation was increased in the model group, consistent with the increased IgG1 fucosylation levels observed in COVID-19 patients (
Figs. 5(f) and
(g)). Based on previous research [
21,
22], a certain correspondence exists between the binding affinity and properties of the mouse IgG1 subtype and the human IgG2 subtype. Therefore, in our study, the fucosylation level of the IgG1* subtype in septic mice and the IgG2 subtype in the serum of patients with severe COVID-19 decreased.
3.6. Detection of key glycosyltransferases of the IgG2 subtype in patients with severe COVID-19
We analyzed the IgG2 glycosylation level in serum from patients with COVID-19. The results revealed that the level of serum IgG2 3_5_0_0 glycosylation was increased during the disease, whereas the level of serum IgG2 3_5_1_0 glycosylation was decreased during the disease. Similar trends were observed for serum IgG2 5_4_0_1 and IgG2 5_4_1_1 levels (
Figs. 6(a) and
(b)). These findings confirm a decrease in serum glycosylation levels as the disease progresses. Moreover, research has demonstrated that FUT8 and FUCA1 are pivotal enzymes involved in modulating serum glycosylation patterns within the context of severe COVID-19 (
Fig. 6(c)). We subsequently performed a Mendelian randomization analysis of fucosylation-related genes and COVID-19. The results suggested that the expression of
FUT8 was negatively correlated with COVID-19 progression disease (
Fig. 6(d)). Additionally, we employed an immunohistochemistry (IHC) assay to assess FUCA1 protein expression in the lung tissues of septic mice. Notably, our findings revealed a decrease in FUCA1 expression in the model group, whereas FZJDD-H treatment led to an increase in FUCA1 expression in the lung tissue (
Fig. 6(e)). We detected the mRNA and protein expression of FUT in the plasma of COVID-19 patients and the lung tissues of mice with septic, as shown in Fig. S6 in Appendix A. Treatment with FZJDD-containing serum effectively suppressed
FUT8 or
FUCA1 overexpression induced by plasmid transfection
in vitro, indicating its regulatory effect on glycosylation-related enzymes, as shown in
Figs. 6(f) and
(g). The semi quantitative images of WB protein bands were shown in Fig. S7 in Appendix A. In summary, our study highlights the primary role of the FUT8 and FUCA1 glycosyltransferases in regulating IgG2 fucosylation in patients with severe COVID-19, providing valuable insights into the underlying mechanisms of this disease.
4. Discussion
Glycosylation, as a non-template dependent post-translational modification, is highly dependent on the specific expression and activity regulation of glycosyltransferases (such as
N-acetylglucosamine (GlcNAc) TIII encoded by MGAT3), which catalyze the transfer of double-branched GlcNAc groups to construct characteristic double-branched sugar chain structures, dynamically affecting protein function and cell fate. Epigenetic regulation (especially DNA methylation) directly regulates the expression of glycosylation related genes such as
FUT7,
MGAT3, and
B4GALNT2, and drives changes in IgG glycosylation patterns (such as double branched GlcNAc enrichment) under the interaction of environmental factors (such as smoking), participating in the pathological processes of lung cancer, ovarian cancer, colorectal cancer (CRC), inflammatory bowel disease (IBD), and immune disorders [
23,
24]. As a dynamic “epigenetic language,” sugar chains respond to environmental signals through chemical modifications (sialylation and fucosylation) [
25]. Their non-template driven synthesis mechanism is determined by the expression of glycosyltransferases and epigenetic networks, forming an interactive regulatory axis of genetic epigenetic environmental factors. This provides a theoretical basis for the diagnosis of complex diseases and intervention strategies targeting glycosyltransferases.
Fucosylation modifies the glycan ends of cell surface proteins and lipids. Previous studies have shown that fucosylation levels are associated with the occurrence of a variety of lung diseases [
26,
27] and are associated with COVID-19 severity [
28]. Our study revealed that the extent of fucosylation in COVID-19 patients is correlated with the disease progression, confirming that fucosylation is a potential indicator of COVID-19 progression.
Currently, extensive research has been conducted on the relationship between fucosylation and T cells [
29,
30], but less attention has been given to B cells. IgG antibodies are secreted by B cells, and the resulting cell signaling is an important component of the humoral immune response [
31]. Structural variation in the IgG-Fc domain between humans was found to be driven primarily by differences in the N-linked glycosylation of the IgG subclasses and CH2 domains [
32]. A mass spectrometry analysis revealed that IgG-B cell receptor (BCR) molecules are glycoproteins rich in core fucosyl groups [
33]. Glycosylation is crucial for maintaining the water solubility and three-dimensional conformation of BCR [
34]. Inadequate glycosylation can compromise the rigidity of the BCR peptide chain, whereas excessive glycosylation may obstruct the Ag binding site of BCR, disrupting its interaction with Ag. The core fucosyl modification plays an important regulatory role in the biological activity of IgG-BCR [
33]. Our study has revealed that fucosylation predominantly occurs in effector B cells (plasma cells) and dendritic cells in patients with severe COVID-19. Importantly, this study is the first to confirm alterations in fucosylation levels within the B cell population of patients with severe COVID-19.
The glycosylation of total IgG, especially the dynamic changes in fucosylation levels, play crucial roles in the immunopathological mechanisms triggered by viral infections [
35]. The absence of fucose at the N-glycosylation site of the IgG-Fc segment significantly increases its affinity for low affinity immunoglobulin gamma Fc region receptor III-A (FcγRIIIa), thereby activating the immune response through enhanced ADCC [
36]. However, this process may also exacerbate cytokine storms. In patients with severe COVID-19, the fucosylation level of anti-viral specific IgG1 is significantly reduced (accounting for approximately 6%) [
37,
38]. These nonfucosylated IgGs preferentially form immune complexes with enveloped viruses (such as SARS-CoV-2) and promote the release of proinflammatory cytokines and acute-phase reactions by overactivating the FcγRIIIa receptor pathway, directly participating in the pathological process of organ damage [
39]. Although virus-specific IgG levels were not measured separately, the fucosylation profile of total IgG can still reflect the virus-specific immune status due to its antigen selectivity. Studies have shown that IgG1 fucosylation levels between different antigens (such as human immunodeficiency virus (HIV) and cytomegalovirus) are not correlated, while the fucosylation level of anti-hepatitis B surface antigen (HBsAg) IgG in individuals vaccinated with the hepatitis B vaccine is significantly higher than that in naturally infected individuals [
40]. Therefore, glycoengineering to modulate IgG fucosylation levels has become an important therapeutic strategy. For example, the development of low-fucosylated monoclonal antibodies can increase ADCC and has shown potential in antiviral and antitumor therapies.
Investigations have revealed that in patients with severe COVID-19, the levels of nonfucosylated IgG antibodies are markedly elevated [
37]. By increasing the activation of Fc receptors, these antibodies exacerbate cytokine storms and contribute to severe immunopathological outcomes; thus they are strongly associated with disease severity [
41]. In severe cases, the core fucosylation level of the IgG1-Fc segment is significantly decreased. Non-fucosylated IgG1 antibodies are more likely to form immune complexes, which in turn overactivate the FcγRIIIa receptor, promoting the release of proinflammatory cytokines such as IL-6 and TNF-α, and thereby intensifying inflammatory responses [
42]. Additionally, while the levels of IgG3 are slightly yet significantly increased in patients with severe COVID-19, the observed reduction in IgG2 levels may be indicative of underlying immune regulatory mechanisms [
43]. Collectively, these changes reflect the intricate immune responses elicited by the body in response to viral infections. Alterations in fucosylation levels have emerged as potential biomarkers for severe COVID-19, suggesting new avenues for disease diagnosis and therapeutic intervention. IgG antibodies can be classified into four subtypes: namely, IgG1, IgG2, IgG3, and IgG4, each characterized by variations in disulfide bond placement and number, yet they share remarkable similarities in their spatial structures. Among them, IgG2 primarily neutralizes antigens or block the binding of receptor ligands [
44], and its complement-dependent cytotoxicity (CDC) and ADCC effects are weak. However, IgG2 is the only subtype that can bind FcγRIIa (CD32a) [
45]. Despite limited research on IgG2 glycosylation in the context of COVID-19, studies have suggested a potential association between IgG1 fucosylation and this disease [
28]. Larsen et al. [
37] reported that IgG1-Fc glycosylation may serve as an early marker of severe COVID-19. The results of this study show that the level of fucosylation of IgG2 is significantly reduced in the blood of COVID-19 patients. This finding may suggest that the fucosylation of IgG2 may be more meaningful than that of the IgG1 type in patients with COVID-19. In addition, previous studies have shown that when afucosylated anti-S IgG appears, plasma IL-6 and C-reactive protein (CRP) concentrations increase [
37], indicating that afucosylated anti-S IgG in COVID-19 patients may lead to the excessive release of proinflammatory cytokines and subsequent systemic inflammation.
Our data shows a significant negative correlation between IgG2-Fc fucosylation levels and body temperature (
Fig. 1(d)), indicating that the glycosylation status of IgG2 regulates inflammatory responses. We propose the following mechanism explanation. The reduction of IgG2 fucosylation enhances its binding affinity to Fc γ receptors (such as FcγRIIa) on activated innate immune effector cells (such as monocytes, macrophages, natural killer (NK) cells). This stronger binding amplifies effector functions, including antibody dependent cellular phagocytosis (ADCP) and ADCC, and crucially increases the production of pro-inflammatory cytokines, particularly potent endogenous pyrogens IL-1β, IL-6, and TNF-α [
46]. These cytokines act in the hypothalamus, raising the body temperature set point and causing fever. On the contrary, higher levels of fucosylation may weaken FcγR binding [
47] and downstream signaling, leading to reduced cytokine release and thus lowering body temperature. Further research on the specific FcγRs involved and direct measurement of the cytokine pathway regulated by IgG2 fucosylation is crucial for fully validating this mechanism.
FUT8 is the sole known enzyme responsible for adding core fucosylation. In
FUT8 knockout (KO) mice, the absence of this enzyme completely abolishes core fucosylation, leading to impaired transforming growth factor-β1 (TGF-β1) receptor function and the subsequent abnormal upregulation of the enzymes matrix metalloprotein-9 (MMP-9), MMP-12, and MMP-13. These lung abnormalities are attributed to the dysfunction of cell surface receptors due to the lack of core fucosylation [
48]. Smoking causes a decrease in core fucosylation in mouse lung tissue, and smoking-induced emphysema develops earlier in hybrid
FUT8 KO mice than in wild-type mice [
49]. Interestingly, studies have reported a decrease in FUT8 glycosyltransferase levels following the administration of COVID-19 mRNA vaccines, but this phenomenon is confined to individuals without prior vaccination or antigen exposure [
50]. Research has shown that changes in α-
L-fucosidase activity in COVID-19 patients precede those in IgM and serum amyloid A (SAA), and are preferentially associated with glucose metabolic disorders. This finding suggests the timeliness of the response of fucosidase to disease states [
51].
In our study, we found for the first time that the occurrence of IgG2 fucosylation in patients with severe COVID-19 is closely related to the regulation of FUT8 and FUCA1. Both enzymes are integral to fucose metabolism, yet they have opposing functions: FUT8 catalyzes fucose addition, whereas FUCA1 facilitates fucose removal. Consequently, these enzymes may collaborate to maintain a balanced glycosylation state and optimal glycosylation structure within cells. Moreover, FUT8 and FUCA1 may be related to the occurrence and development of the disease by affecting the glycosylation state of IgG molecules. However, this study has several limitations. Although no experiments have proven the direct interaction between FUT8 and FUCA1, they may play complementary roles in regulating glycosylation kinetics. FUT8 core fucosylated glycoprotein and FUCA1 remove terminal fucose, which may affect cancer glycosylation. Future studies will use kinins, functional assays and metabonomic analysis to confirm their interactions and joint effects on glycosylation profiles. While we identified a correlation between the levels of IgG-Fc glycopeptides and the severity of COVID-19, these findings are based on data collected at a single time point. The absence of longitudinal data prevents us from establishing causal relationships. Further investigation is warranted to elucidate their precise roles and interrelationships in severe COVID-19.
5. Conclusions
In summary, this study revealed a notable decrease in fucosylation levels during the progression of COVID-19, particularly in the IgG2 subclass and plasma cells. The correlation analysis further revealed that variations in IgG2 fucosylation levels were associated with the severity of the disease and body temperature among patients. Through comprehensive transcriptomic and single-cell sequencing analyses, we validated significant alterations in fucosylation patterns among patients with severe COVID-19, with FUT8 and FUCA1 being identified as potential key FUTs. Furthermore, leveraging the antiseptic mechanism of FZJDD, a recommended medication for severe COVID-19 treatment. We demonstrated that defucosylation may be a crucial strategy to prevent the progression from mild to severe COVID-19.
CRediT authorship contribution statement
Caiping Zhao: Writing - review & editing, Writing - original draft, Validation, Methodology, Formal analysis, Conceptualization. Jingrong Wang: Software, Methodology, Data curation. Yuan Liu: Visualization, Software, Investigation, Data curation. Baoling Shang: Visualization, Funding acquisition. Danna Lin: Software, Methodology, Data curation. Yao Xiao: Methodology. Hong Ren: Validation, Methodology. Yue Li: Methodology. Wen Rui: Methodology, Data curation. Xu Zou: Supervision. Hudan Pan: Supervision. Liang Liu: Supervision.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This work was financially supported by the National Key Research and Development Project of China (2022YFC0867500), the Science and Technology Projects in Guangzhou (2024A03J0040), and the National Funded Postdoctoral Researcher Program (GZC20230626).
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.eng.2025.08.004.
PDF
(6482KB)
