Skip to main content
NCBI home page
Springer logoLink to Springer
. 2025 Aug 10;58(3):789–802. doi: 10.1007/s11255-025-04708-7

High-density lectin microarray uncovers stage-specific plasma glycoprofiling signatures for non-muscle-invasive and muscle-invasive bladder cancer

Xiansheng Xiang 1,#, Xue Mei 2,#, Xiaojing Song 2, Bing Pan 1, Kun Ma 1, Chunlin Tan 3, Lin Liu 4, Shengmao Li 2, Qiang Dong 5,✉, Tielong Tang 3,✉
PMCID: PMC12935817  PMID: 40783897

Abstract

Objectives

The critical need for minimally invasive biomarkers to stratify non-muscle-invasive (NMIBC) and muscle-invasive bladder cancer (MIBC) prompted our investigation into tumor-specific glycan signatures. We hypothesized that lectin-based glycoprofiling could stratify MIBC and NMIBC via stage-specific glycan remodeling. Leveraging the emerging role of lectin microarrays in glycoprofiling, the first comprehensive plasma-based analysis comparing NMIBC and MIBC glycosylation patterns using a high-density 95-lectin platform was conducted.

Methods

A pilot cohort of 48 plasma samples (22 NMIBC, 10 MIBC, 16 NL) was analyzed using RayBiotech’s high-density 95-lectin microarray platform (GA-Lectin-95). Advanced bioinformatics pipelines (Limma package, R/Bioconductor) identified differentially expressed lectins (LogFC > 1.2, adj.p < 0.05) with stringent validation, including ROC analysis (Receiver Operating Characteristic, pROC package) and cross-validated AUC (Area Under the Curve) calculations.

Results

Stage-specific glycoarchitectural remodeling revealed profound pathophysiological divergence: NMIBC plasma exhibited global glycan depletion marked by SSA/ASA/SHA downregulation (diagnostic AUC 0.79–1.00), whereas MIBC demonstrated aggressive sialylation/fucosylation signatures dominated by VVA/WFA overexpression (AUC 0.91–0.95). A decalectin panel in DSA/ASA/STL/ABL/MNM-M/PWA/PTL-1/SSA/CAA/GAL3C-S achieved perfect stratification of NMIBC from controls (AUC = 1.00), while four invasion-associated lectin complexes GAL1-S/RCA60/MPL/LSL-N showed MIBC-specific amplification (AUC > 0.80). Notably, CAA/DSA/SSA precisely discriminated MIBC from NMIBC (AUC = 1.00). GAL2 emerged as a novel metastasis checkpoint through its dramatic suppression in invasive tumors, suggesting its role as a glycosylation gatekeeper restraining muscle invasion.

Conclusions

These findings support plasma glycoprofiling as a paradigm-shifting tool for bladder cancer subtyping. The results of this study provide actionable biomarkers for clinical stratification and diagnostics applications in bladder cancer management.

Keywords: Biomarkers, Lectin, MIBC, NMIBC, Plasma glycopatterns

Introduction

Bladder cancer is the global malignant tumor with poor prognosis, accounting for over 570,000 new cases worldwide in 2020, a number projected to double by 2040 [1]. The incidence and aggressiveness of bladder cancer are dramatically different between male and female [2]; it ranks as the sixth common malignant cancer in males but only the seventeenth in females [3]. Despite lower incidence, female patients often experience more advanced disease and poor prognosis [4, 5], with lower survival rates compared to male patients, particularly during the first 2 years after diagnosis [6]. Multiple efforts have been made in the past few years to explore the role of race in bladder cancer biology, and further advances are expected in this field [7]. Bladder cancer can be categorized as non-muscle-invasive bladder cancer (NMIBC; stages Tis, Ta and T1) and muscle-invasive bladder cancer (MIBC; stages T2-T4) [8]. Before muscle invasion, bladder cancer is typically treatable with a good prognosis, while once the disease becomes muscle-invasive, it has the potential to metastasize to lymph nodes, bones, lungs, and liver, often leading to poor survival outcomes (~ 15 months) [9]. Within the NMIBC category, Ta tumors often exhibit a benign disease course in comparison to T1 and Tis, and their treatment differs accordingly [10]. These different stages are closely associated with different genetic features, which can be used as biomarkers for disease progression. The critical role of these markers in disease management further increases as molecular pathology becomes more predominant in diagnosis, treatment selection, and follow-up visiting[11]. The most informative molecular markers of genetic variants include TP53, ERCC1, and FGFR3, which are important for understanding disease progression, chemotherapy sensitivity, and the selection of small molecule therapeutics [12].

Glycosylation is the most common co- and post-translational modification that affects more than half of all known proteins [13] and also influences various cellular pathways [14]. Glycans, which are covalently bonded to biomolecules, modulate cellular functions by both direct interactions—such as the recognition of glycan structures by binding partners—and indirect mechanisms that affect protein conformation and stability [15]. This type of protein modification is known as post-translational modification (PTM), which contains over 200 PTMs characterized to date.Glycosylation is unique among these modifications because it is a template-free process that results in the addition of heterogeneous structures, leading to a large mass shift [15]. The attachment between carbohydrate moieties and amino acid residues during the glycosylation process often occurs on asparagine, serine, threonine, tryptophan, hydroxyproline, hydroxylysine, and rarely tyrosine [16, 17]. Glycobiology is a newly appeared hot topic in malignant tumor research because almost all secreted proteins and cellular membranes carry glycans, which are involved in multiple biological processes such as molecular signaling and cell communications. Altered glycosylation is an important hallmark of most malignant tumors [18]. Tumor-associated carbohydrate antigens, including Tn antigen (N-acetylgalactosamine, GalNAc-O-Ser/Thr) in breast, lung, colon, prostate, and pancreatic cancer, harbor critical roles in multiple important events during cancer progression and modulation of the anti-tumor immune responses [19–21]. There are two main types of protein glycosylation, which depend on the addition of an oligosaccharide to either the nitrogen atom (N-linked glcosylation) or the oxygen atom (O-linked glycosylation) of the polypeptide chain [22]. Cancer-associated changes in glycosylation, including fucosylation, sialylation, N-glycan truncation, and N-/O-linked glycan branching, have been reported in the development and progression of multiple tumors [20]. Lectins are glycan-binding proteins isolated from trees, legumes, and other sources that are highly specific for glycans [23]. Therefore, lectin microarray is an effective tool for analyzing glycans [24]. The screening of lectin levels in tumors can be valuable for the diagnosis of malignant tumors. This technology has unveiled glycoprofiling biomarkers in breast cancer, colon cancer, cervical cancer, lung cancer, and liver cancer, but its application in bladder cancer is nascent.

In this study, we conducted the first plasma-based lectin microarray analysis (95-lectin panel) to identify stage-specific glycopatterns in 48 human plasma samples, including 22 NMIBC, 10 MIBC, and 16 NL. Using a lectin microarray containing 95 lectins, we explored the relevance of changes in plasma glycoproteins associated with different bladder types. The alteration levels in glycoproteins were also compared between NMIBC, MIBC, and NL groups, with ROC and combination analysis performed on candidate lectins to evaluate the potential of plasma glycoproteins as effective biomarkers for the diagnosis of bladder cancer. By integrating advanced bioinformatics and clinical validation, our goal was to address the critical unmet need for non-invasive biomarkers in bladder cancer subtyping.

Materials and methods

Collection of plasma samples

This study was approved by the Ethics Committee on Biomedical Research of Nanbu County People's Hospital (Approval No. 2025–88). A total of 48 human plasma samples were enrolled, including 22 NMIBC patients, 10 MIBC patients, and 16 healthy controls (NL). Patients were diagnosed via cystoscopy and histopathology, with exclusion criteria of diabetes, chronic inflammatory diseases, or prior chemotherapy. All recruited patients had diagnostic pathological information and clinical characteristics (Table 1). All plasma samples were collected, processed, and stored at −80 °C until analysis.

Table 1.

Clinical information of samples

Variable NMBC
(n = 22)		 MIBC
(n = 10)		 NL
(n = 16)
Age, year 59.95 ± 13.42 61.50 ± 13.41 60.50 ± 8.70
Gender n (%)
 Male 16(72.73%) 5(50%) 8(50%)
 Female 6(27.27%) 5(50%) 8(50%)
Smoking status n (%)
 Smoking 12(54.55%) 4(40%) 4(25%)
 No smoking 10(45.45%) 6(60%) 12(75%)
Treatment history n (%)
 Total bladder resection + ileal bladder 2(9.09%) 4(40%)
 Total palliative bladder resection 1(4.55%) 1(10%)
Clinical staging, n (%)

T1

T2

T3/4

 22(100%)

8(80%

2(20%)
Pathological grade n (%)
High 7(31.82%) 9(90%)
Low 15(68.18%) 1(10%)
Open in a new tab

Lectin microarray profiling

The lectin microarray conducted in this study is GA-Lectin-95 purchased from Ray Biotech Company (Peachtree Corners, Georgia, USA). This kit was firstly dialyzed and all plasma samples were biotin-labeled. The microarray slides which contain the lectin chips were removed and equilibrated at room temperature (20–30 min); then, 100 μL of 1 × closure solution was added to each chip well and closed at room temperature. After discarding the closure solution, 100 μL of the sample was added in each well. All biotin-labeled plasma samples were diluted 50-fold according to the pre-experimental closure solution. After removing all samples and washing two times (150 μL) of the slides with wash buffer, Cy3 equivalent dye-streptavidin (80 μL) was added to the chip (80 μL/well) by adding 1.4 mL of closure solution and incubating for 1 h in a dark box. Decant the samples from each well with 150 μL 1 × wash buffer I at room temperature with gentle shaking and completely remove wash buffer. Fluorescence detection was then conducted after putting the glass slide into Slide Dryer by centrifuging at 1,000 rpm for 3 min without a cap; the signals were visualized using a laser scanner equipped with a Cy3 wavelength such as Axon GenePix.

Normalization of array data

To normalize signal intensity data, one sub-array is defined as the “reference”, to which all other arrays are normalized. The selection of this reference array is arbitrary. For example, in our Analysis Tool Software, the array represented by the data entered in the left-most column each worksheet is the default “reference array.” The normalized values can calculated using the following formula: X(Ny) = X(y) * P1/P(y) Where: P1 is the mean signal intensity of POS spots on reference array. P(y) is themean signal intensity of POS spots on array “y”. X(y) is the mean signal intensity for spot “X” on array “y”. X(Ny) is the normalized signal intensity for spot “X” on Array “y”.

Analysis of differential proteins

The Limma data package was conducted for analyzing the normalized raw data, with the analyze method based on moderated t-statistics sourced from R/Bioconductor. Differential proteins were screened using adjusted p-value (corrected with the BH method) and LogFC (l fold change in expression, with a base of 2). The selection conditions were as follows: LogFC > log2 (1.2), with a difference threshold of 1.2, and a corrected p-value of less than 0.05.

Volcano visualization for differential proteins

The scatter plot plotting was generated using the ggplot2 function, with the ggfortify data package from R. In the volcano map, the horizontal axes represent the average expression level (AveExp) of each group (the logarithm of each sample value), while the vertical axis displays the negative logarithm of Val or p-value—log10 (adj.P.Val or p-value). In the plot, red points indicate upregulated proteins, blue points represent downregulated proteins, and gray points indicate proteins with no significant difference. The volcano plot is then performed to visualize the differentially expressed proteins, with logFC (log-fold change) on the x-axis and the adjusted p-value on the y-axis, highlighting the significance of differential protein expression between the two samples.

Principal component analysis

Principal component analysis (PCA) aims to use the idea of dimensionality reduction to transform multiple indicators into a few comprehensive indicators. After performing PCA, the first two principal components are selected, and the distribution of the samples on the two-dimensional plane is plotted based on the scores of the principal components. The sample can be classified intuitively through the graph, and outliers far away from the majority of sample points can be identified through the graph. The prcomp function from R is used for this analysis, with the autoplot function for visualization, and the ggfortify data package from R/Github is employed for plotting.

Cluster heatmap

Hierarchical clustering is the process of treating all objects as the content of a cluster, continuously decomposing them into smaller but increasingly numerous clusters, until all objects form a cluster on their own or meet certain termination conditions. The method adopted is the heatmap.2 function, and the data packet is gplots from R/Bioconductor. The distance between two samples is calculated using the Euclidean distance. The distance between two clusters is calculated using the farthest neighbor method (complete linkage), and the inter-cluster distance is defined as the maximum distance between the data points of two clusters.

ROC curve analysis

ROC (Receiver Operating Characteristic) curve analysis is conducted to target lectins using the pROC package from R4.0.2 software to assess the sensitivity and specificity of the target lectins. The AUC (Area Under the Curve) is performed to estimate the diagnostic value. Logistic regression analysis is constructed to evaluate the lectin combination. To reduce the risk of overfitting and enhance model robustness given the relatively small sample size (N = 48; NMIBC = 22, MIBC = 10,NL= 16), we applied a repeated five-fold cross-validation procedure across the entire cohort. The dataset was randomly divided into five approximately equal-sized subsets, preserving the proportional distribution of NMIBC, MIBC, and healthy control subjects in each fold. In each iteration, one fold was reserved as the validation set, while the remaining four served as the training set. This five-fold procedure was repeated five times with different random splits, yielding a total of 25 cross-validation runs. The repeated design allowed us to mitigate sampling variability and obtain more stable and generalizable performance estimates. Evaluation metrics such as the area under the ROC curve (AUC) were calculated for each run and averaged across all iterations.

Results

Lectin microarray analysis reveals differential glycoprotein levels in bladder cancer subtypes and NL

All 48 participants in this study were divided into three groups: the NMIBC group (22 cases), the MIBC group (10 cases) and NL (16 cases). All plasma samples were detected by the lectin microarray (Fig. 1A), which is currently the largest lectin array containing 95 lectins.The details of this chip are shown in Table 1. Raybiotech’s Excel-based analysis software was conducted on original data from chip scanning to remove and normalize the chip background. The different normalized fluorescence intensities of lectins can reflect the different glycoproteins among the samples because specific lectin binding recognizes different carbohydrates. In our study, all plasma samples were analyzed in the heatmap with 90 samples as columns and 95 lectins as rows, which revealed the different glycoprotein levels in different groups (Fig. 1B). In detail, the total lectin levels significantly declined in the NMIBC group in comparison to the healthy group, while part of the lectins increased dramatically in the MIBC group (Fig. 1B). Moreover, the top 10 variation proteins are shown in Fig. 1C, which are SSA, ASA, PTL-1, PWA, VVA-M, ASL, MNA-M, CAA, DSA, and STL (Fig. 1C). Overall, significant variations in lectin levels were detected both in MIBC and NMIBC groups in comparison to NL.

Fig. 1.

Fig. 1

Open in a new tab

Detection of glycopatterns in MIBC, NMIBC, and NL using the lectin chip. A Flow chart of the detection procedures. All samples were labeled with biotin, blocked, and incubated with Cy3 equivalent dye-streptavidin. The signals were scanned with a laser scanner (InnoScan 300), with a wavelength of 532 nm and a resolution of 10 μm. B The total scanning results of all 48 samples. Columns were plasma samples and rows were 95 lectins. The color and intensity of each square represented the level of expression relative to the rest of the data in the row. Red stands for up-regulation, blue stands for down-regulation, and white stands for no change. C The top 10 variation lectins with the most variation among MIBC, NMIBC, and NL groups. Red stands for up-regulation, blue stands for down-regulation, and white stands for no change

Comparison analysis between MIBC/NMIBC and NL

We firstly analyzed the correlation between bladder cancer (BC) group (MIBC/NMIBC) and NL of candidate lectins (Fig. 2A). Blue represents positive correlation, red represents negative correlation, with darker shades indicating stronger correlations, while an 'X' represents no significant correlation. Our analysis revealed that among the top 10 candidate lectins, the correlation of ABA and SHA was greater than 0.8 in the comparison of BC and NL. The correlation coefficients between VVA-M and PWA, ABA and PWA, SSA and PWA were all greater than 0.7, indicating the significant correlation among these lectins (Fig. 2A). The lectin signal patterns were divided into three categories to evaluate the significant variation in plasma glycopatterns: (1) significant upregulation (fold change > 1.2, p value < 0.05), (2) significant downgregulation (fold change < 0.83, p value < 0.05) and (3) no significant difference (fold change 0.83–1.2, p value > 0.05). Glycopatterns were altered in BC compared to NL, with the majority of altered lectins being downregulated in BC patient samples, such as SSA, ASA, SHA, PWA, ABA, and NPA. However, 11 lectins, including VVA, WFA, PSL1A, UDA, SNA-B, GRFT, WGA, SBA, ACL, F17AG, and GAL7-S were overexpressed (Fig. 2B). The differential lectins were visualized by volcano plot, including upregulated (WFA, SNA-II, VVA and F17AG) and downregulated (SHA, ASA, PWA, ABA and F17AG) (Fig. 2C). The analysis between BC and NL also revealed 10 differential lectin clusters across all participant samples (Fig. 2D). ROC curves were drawn to evaluate the role of each differentially expressed lectin in the diagnosis of BC patients, with AUC values as follows: ABA (0.90), ASA (0.78), F17AG (0.87), GAL1 (0.79), PWA (0.76), SHA (0.81), SSA (0.79), VVA (0.95), VVA-M (0.75) and WFA (0.91) (Fig. 2E). Significant variations were also confirmed by scatter plot analysis (Fig. 2F), which indicated the role of lectins in the diagnosis of BC.

Fig. 2.

Fig. 2

Open in a new tab

The variation of glycopatterns level in bladder cancer (MIBC + NMIBC) in comparison to NL (NL). A Top 10 lectins identified in correlation analysis between tumor samples and NL. The red represents positive correlation, blue represents negative correlation, and the “X” represents no significant correlation. B Total lectins in correlation analysis between tumor samples and NL. The red represents positive correlation, blue represents negative correlation, and the “X” represents no significant correlation. C Volcano plot of the differences of glycopattern between tumor group and NL. Red indicates up-regulated lectins, blue indicates down-regulated lectins, and gray indicates insignificant differences. D The heatmap analysis between tumor group and NL. Rows represent samples, and columns represent candidate lectins. The color and intensity of each square represent the expression level relative to the rest of the data in the row. Red represents up-regulation, blue represents down-regulation, and white represents insignificance. E ROC analysis of candidate lectins in tumor group in comparison to NL. F Scatter analysis of candidate lectins with significantly differential signals between tumor patients and NL. **p < 0.01, ***p < 0.001, ****p < 0.0001, n.s., no significance

Alteration in plasma glycopatterns between MIBC and NL

Correlation analysis was performed to examine alterations in lectins between MIBC and NL, and the top 10 lectins, as well as all co-related lectins, were listed in Fig. 3A and B. It was found that GAL2 showed a negative correlation with almost all candidate lectins (Fig. 3B). On the other hand, the most significant positive correlations involved lectins such as GALS-1, which was positively correlated with RCA60 (0.83), MPL (0.80), LSL-N (0.76), ACL (0.69), WFA (0.67) and Jacalin (0.61); RCA60, which was positively correlated with GAL1-S (0.83), LSL-N (0.81), MPL (0.76), ACL (0.75), PSL1A (0.67) and WFA (0.58); ACL, which was positively correlated with GAL1-S (0.73), LSL-N (0.83), Jacalin (0.67), MPL (0.76), PSL1A (0.66) and RCA60 (0.74) (Fig. 3A). Volcanoplot was performed to analyze the differentially expressed lectins between MIBC and NL, and the results indicated that only GAL2 was down-regulated in MIBC, while multiple lectins were increased dramatically in the MIBC group in comparison to the healthy control group, including VVA, UDA, ACL, GAL1-S, MPA, and WFA (Fig. 3C).The analysis also revealed two different lectin clusters: one where all NL and three MIBC patients (MIBC185, MIBC 270 and MIBC 245) harbored similar lectin expression levels, while the remaining 7 MIBC patients exhibited a different lectin pattern (Fig. 3D). In ROC analysis of the top 10 variation lectins, including ACL, GAL1-S, Jacalin, LSL-N, MPL, PSL1A, RCA60, UDA, VVA, and WFA (Fig. 3E and F), which were all dramatically up-regulated in the MIBC group (all p < 0.05). In ROC analysis, the AUC values of LSL-N, GAL1-S, RCA60, and MPL were more than 0.80, which indicated the strong diagnostic potential for MIBC. Additionally, the AUC values of PSL1A, VVA, Jacalin, WFA, ACL, and UDA were more than 50%, underscoring their important role in diagnosis (Fig. 3E). Therefore, plasma lectin levels in MIBC could be used as effective biomarkers for MIBC diagnosis.

Fig. 3.

Fig. 3

Open in a new tab

Variation of glycopattern levels in MIBC in comparison to NL. A Top 10 lectins in correlation analysis between MIBC samples and NL. The red represented positive correlation, the blue represented negative correlation, and the X represented no significant correlation. B Total lectins in correlation analysis between MIBC and NL. The red represented positive correlation, the blue represented negative correlation, and the X represented no significant correlation. C Volcano plot of the differences of glycopatterns between MIBC samples and NL. Red indicated up-regulated lectins, blue indicated down-regulated lectins, and gray indicated insignificant differences. D The heatmap analysis between MIBC samples and NL. The samples were rows, and the candidate lectins were columns. The color and intensity of each square represented the expression level relative to the rest of the data in the row. Red represented up-regulation, blue represented down-regulation, and white represented insignificance. E ROC analysis of candidate lectins of MIBC patients in comparison to NL. F Lectin factors with significantly differential signals of candidate lectins in MIBC samples in comparison to NL. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

The differential analysis of lectin levels in NMIBC compared to NL

Firstly, we conducted correlation analysis between NMIBC and NL (Fig. 4A and B) and the results showed that WFA, VVA, and F17AG were negatively correlated with other candidate lectins, while the majority of variation lectins were positively correlated. The top 10 lectins, including CAA, GAL1, ABA, ASA, PWA, SHA, SSA, VVA-M, MAA, and PTL-1, showed strong correlation with each other (Fig. 4A). Next, differential expression lectins between NMIBC and NL were analyzed by volcanoplot.The results revealed 8 enhanced expression lectins in NMIBC (WFA, VVA, PSL1A, SNA-B, GRFI, WFA, SBA and F17AG). The majority of differential lectins, such as SSA, ASA, SHA, PWA, VVA-M, PTL-1, GNA, ABA, MAA, et al., were down-regulated in NMIBC samples (Fig. 4C). The heatmap results indicated two different clusters; the NMIBC and NL could be completely separated based on differential lectin levels (Fig. 4D). ROC analysis was conducted on the top 10 differential lectins, including CAA, GAL1, ABA, ASA, PWA, SHA, SSA, VVA-M, MAA, and PTL-1. Except for GAL1 (0.98) and ASA (0.96), the other eight lectins (CAA, ABA, PWA, SHA, SSA, VVA-M, MAA and PTL-1) showed an ROC value of 1.00 (Fig. 4E). Finally, we performed scatter analysis to evaluate the diagnostic value of these lectins and found significant variations (all p < 0.001) between NMIBC and NL groups (Fig. 4F). Together, our study demonstrated that plasma lectin levels in NMIBC could act as diagnostic biomarkers for NMIBC patients.

Fig. 4.

Fig. 4

Open in a new tab

Variation of glycopattern levels in NMIBC in comparison to NL. A Top 10 lectins in the correlation analysis between NMIBC samples and NL. Red represents a positive correlation, blue represents a negative correlation, and "X" represents no significant correlation. B Total lectins in the correlation analysis between NMIBC samples and NL. Red indicates up-regulated lectins, blue indicates down-regulated lectins, and gray indicates insignificant differences. Red represents a positive correlation, blue represents a negative correlation, and "X" represents no significant correlation. C Volcano plot of the differences of glycopatterns between NMIBC samples and NL. Red indicates up-regulated lectins, blue indicates down-regulated lectins, and gray indicates no significant differences. D Heatmap analysis between NMIBC samples and NL. Rows represent samples, and columns represent candidate lectins. The color and intensity of each square represent the expression level relative to the rest of the data in the row. Red represents up-regulation, blue represents down-regulation, and white represents no significant change. E ROC analysis of candidate lectins of NMIBC samples in comparison to NL. F Lectin factors with significantly differential signals of candidate lectins in NMIBC samples in comparison to NL. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

The differential analysis between MIBC and NMIBC patients

As MIBC and NMIBC were two important invasive statuses of bladder cancer, it was critical to select valuable lectins which could be used to predict the occurrence of MIBC. In correlation analysis of differential lectins between MIBC and NMIBC, we found that F17AG, LPA, EEL, and MOA had no correlation with other candidate lectins (Fig. 5A and B). The GAL2 had a negative correlation with the majority of differential lectins (Fig. 5B), while all other differential lectins were positively correlated. The top 10 lectins were DSA, ASA, STL, ABL, MNM-M, PWA, PTL-1, SSA, CAA, and GAL3C-S, which harbored significant correlation between each other (correlation co-efficiency > 0.8) except for ASA, which showed 0.6–0.7 co-efficiency with other lectins (Fig. 5A). The volcanoplot results revealed two down-regulated lectins, F17AG and GAL2 (Fig. 5C), while the majority of variation lectins between MIBC and NMIBC were up-regulated, including CAA, DSA, SSA, PWA, CGL2, VVA-M, et al. (Fig. 5C). This indicates the importance of these lectins in distinguishing NMIBC and MIBC patients. The heatmap analysis revealed four clusters, with the differential lectins completely separating the NMIBC and MIBC groups (Fig. 5D). In ROC analysis, the top 10 lectins (DSA, ASA, STL, ABL, MNM-M, PWA, PTL-1, SSA, CAA and GAL3C-S) showed a ROC value of 1.00 (Fig. 5E). The scatter analysis also indicated the significant differences in lectin levels (all p < 0.0001) between NMIBC and MIBC patients (Fig. 5F). Taken together, our study uncovered critical lectins which could be applied to predict the invasive potential in different stages of bladder cancer.

Fig. 5.

Fig. 5

Open in a new tab

Variation of glycopatterns level in MIBC in comparison to NMIBC samples. A Top 10 lectins in the correlation analysis of MIBC and NMIBC samples. The red represents positive correlation, the blue represents negative correlation, and the "X" represents no significant correlation. B Correlation analysis of all lectins between MIBC and NMIBC samples. Red indicates up-regulated lectins, blue indicates down-regulated lectins, and gray indicates no significant differences. C Volcano plot of the differences of glycopattern between MIBC and NMIBC samples. Red indicates up-regulated lectins, blue indicates down-regulated lectins, and gray indicates no significant differences. D Heatmap analysis between MIBC and NMIBC samples. Rows represent samples, and columns represent candidate lectins. The color and intensity of each square represent the expression level relative to the rest of the data in the row. Red represents up-regulation, blue represents down-regulation, and white represents no significant change. E ROC analysis of candidate lectins of MIBC samples in comparison to NMIBC samples. F Lectin factors with significantly differential signals of candidate lectins in MIBC samples in comparison to NMIBC samples. ****p < 0.0001

Discussion

Glycobiology has gained increased importance in cancer research and provides multiple targets for diagnostic application and tumor therapy [25–28]. Glycosylation plays a critical regulatory mechanism controlling several physiopathological processes [29]. Glycosylation variation associates closely with oncogenic transformation, and tumor cells exhibit a wide range of glycosylation alterations compared with non-transformer cells [19]. Glycans have been found to participate in numerous essential biological processes in malignant tumors, including cell–cell adhesion, immune surveillance, angiogenesis, cell–matrix interaction, and cellular metabolism [19]. In a gastric cancer study, for instance, the overexpression of GnT-V induces E-cadherin cellular mislocalization from the cell membrane into the cytoplasm, leading to incorrect assembly and non-functional adherens junctions. This, in turn, affects downstream signaling pathways and finally contributes to tumor invasion and metastasis [30]. The extracellular matrix (ECM) is composed of a dynamic and complex array of glycoproteins, collagens, and proteoglycans, which provide mechanical structural support for signaling events and offer direct implications in tumor development, as well as maintenance of stem cell niches and cancer invasion [31]. The shift from oxidative phosphorylation to aerobic glycolysis is a key feature of tumor metabolism [32]. In breast cancer, the level of O-GlcNAc transferase (OGT) dramatically increases, and knockdown of OGT reduces cancer hyper-O-GlcNAcylation, inhibits tumor development, invasion, and metastasis [33] O-GlcNAc modifications are implicated in many key events in malignant tumors, including tumor proliferation, survival of cancer cells, angiogenesis, and upregulation of VEGFA and matrix metalloproteinases (MMPs), all of which contribute to tumor invasion [34]. Furthermore, alterations in the sialylation and fucosylation of N- and O-glycans play important roles in brain cancer [15].

Glycans is an important source for the development of non-invasion biomarkers for cancer diagnosis. Multiple clinically utilized circulating biomarkers for cancer diagnosis are glycoproteins, including prostate-specific antigen (PSA) [35] in prostate cancer, carcinoma antigen 125 (CA125) in ovarian cancer [36], carcinoma antigen 129 (CA129) and carcinoembryonic antigen (CEA) in colon cancer [37], aberrantly glycosylated MUC1 in breast cancer and CA19-9 in pancreatic cancer [38]. Lectin screening for glycopatterns in tumors is now applied for selecting effective markers for cancer diagnosis. In lung cancer, a study using lectin microarray with 16 lung adenocarcinoma (LUAD) samples, 16 lung squamous cell carcinoma (LUSC) samples and 16 NL identified 7 candidate lectins that offered potential diagnostic utility in LUAD and LUSC, with F17AG exhibiting an important role in LUSC patients (AUC > 0.80)[39]. In another lung cancer study examining altered serum glycans, the Tn antigen (precursor of O-linked glycans) was found to be significantly increased in lung adenocarcinoma serum in comparison to NL [40]. In colorectal cancer, serum samples from 633 patients and 478 age-/sex-matched controls revealed obvious differences in N-glycosomes levels at all stages of CRC [23]. In bladder cancer, α2-6 sialylation, β1-4 branching, β1-6 branching and outer-arm fucosylation could be used as biomarkers for diagnosis [41]. Therefore, glycosylation alterations in glycoconjugates on the cell membrane or secreted by cancer cells are commonly used in biomarker selection in tumor diagnosis [42]. Recent advancements in lectin microarray technology provide a useful tool for high-throughput screening of glycans in plasma or serum from malignant tumors [43].

In this study, our objective was to compare different types of bladder cancer (MIBC and NMIBC) with NL to select candidate lectin biomarkers for bladder cancer diagnosis. Lectin chip technology was performed on all participant plasma samples, including 10 MIBC, 22 NMIBC, and 16 NL. The chip contained 95 lectin factors, which covered most of the glycan structures. We found that a total of 11 lectin factors increased dramatically in MIBC samples, including VVA, WFA, PSL1A, UDA, SNA-B, GRFT, WGA, SBA, ACL, F17AG, and GAL7-S. In NMIBC samples, six lectin factors were found down-regulated, including SSA, ASA, SHA, PWA, ABA, and NPA. Finally, combination analysis revealed excellent diagnostic potential in both NMIBC and MIBC, with an ROC value of 1.00 in comparison to NL. Lectins play significant roles in the pathogenesis of bladder cancer, particularly in tumor invasion, immune evasion, and metastasis. The differential expression of these lectins between NMIBC and MIBC suggests their involvement in critical cancer-related processes. It has been reported that SSA can significantly sensitize cancer cells to anoikis-induced death both in vitro and in vivo [44]. This suggests that SSA may similarly enhance the susceptibility of tumor cells to anoikis, potentially overcoming the anoikis resistance mechanisms commonly associated with cancer progression. In this study, SSA was identified as being downregulated in NMIBC samples, which may be linked to reduced tumor cell adhesion, a critical factor in the transition of cancer cells from benign to invasive forms. On the other hand, DSA was found to bind to specific glycoproteins expressed in tumor tissues, such as canine visceral hemangiosarcoma. DSA-bound glycoproteins showed varying intensities across different tissue components (hemangiosarcoma cells, arterial endothelium, and venous endothelium) [45]. This suggests that DSA may interact with tumor-associated glycoproteins that contribute to tumor progression. Similarly, PWA's selective accumulation in tumors indicates its potential role in tumor cell migration and metastasis. PWA has also been reported to selectively accumulate in tumors, suggesting a similar role in tumor cell migration and metastasis. By binding to glycan structures on tumor cells, PWA may affect the interactions between tumor cells and the extracellular matrix, thereby influencing cell adhesion and migration processes that are critical for metastasis [46]. Here, DSA and PWA, which were upregulated in MIBC, could potentially facilitate tumor invasion by promoting glycan-mediated signaling pathways that enhance cell migration and metastasis.

Moreover, GAL1, another important lectin that shows varying expression between the stages of bladder cancer, has been implicated in immune evasion [47]. It is reported that GAL1 contributes to immune evasion by preventing T cell migration into the tumor microenvironment. It acts by reprogramming the tumor endothelium, which is the layer of cells that line blood vessels within tumors, to upregulate immune checkpoint molecules such as PD-L1 and galectin-9. These molecules promote immune suppression, thus inhibiting the effective immune response against the tumor [48].The upregulation of CAA and VVA-M in MIBC may similarly play a role in immune evasion, potentially affecting immune cell recognition and the subsequent immune response [49, 50].

The differential glycosylation patterns observed in bladder cancer highlight the critical role of glycan-mediated interactions in cancer cell biology. Lectins like F17AG and GAL2, which were downregulated in MIBC, may also influence metastasis by regulating the extracellular matrix and the cellular microenvironment. For instance, GAL2 is known to be involved in cellular communication and the maintenance of tissue integrity, and its suppression in invasive tumors could facilitate the detachment of cancer cells from the primary tumor, promoting metastasis [51, 52].

These findings underscore the importance of lectins in bladder cancer biology and suggest that they could serve as potential biomarkers for tumor progression, immune evasion, and metastasis. Further investigation into the specific mechanisms by which these lectins contribute to tumorigenesis is warranted to fully understand their role in bladder cancer progression.

However, this study still has some limitations. First, the small sample size of the cohort may limit the statistical power of our findings, and consequently, the generalizability of the results should be interpreted with caution. And although we observed AUC values of 1.00 for some markers, indicating potential overfitting, future studies with larger, independent cohorts are needed for external validation to confirm the diagnostic potential of these markers and evaluate their broader applicability in clinical settings. Second, the components in plasma are complex and exhibit individual variation related to factors such as patient age, physical condition, and the presence of comorbidities. Third, we mapped lectin binding to glycan structures, but we did not identify the specific glycoproteins involved. Future work should incorporate lectin pulldown coupled with mass spectrometry (MS) to enable precise glycoprotein identification. Additionally, complementary approaches, such as lectin affinity chromatography, biotinylated lectin-based Western blotting, and hydrazide chemistry-based glycoprotein enrichment, may further expand the glycoprotein landscape and provide deeper insights into the glycosylation patterns distinguishing NMIBC from MIBC.

Conclusion

This study investigated the altered plasma glycopatterns and systematically compared bladder cancer (NMIBC and MIBC) with NL. The glycan patterns were significantly altered in both NMIBC and MIBC samples. In MIBC, multiple factors (including CAA, DSA, SSA, PWA and STL et. al) were up-regulated compared with healthy control, while SSA, ASA, SHA, PWA, and VVA-M were down-regulated in NMIBC samples. These alterations could serve as effective biomarkers in future diagnosis of bladder cancer.

Acknowledgements

Thanks to all staff of the Department of Urology, West China Hospital, Sichuan University, for their support and suggestions.

Author contributions

The authors confirm their contribution to the paper as follows: Tielong Tang, Qiang Dong, and Liu Lin conceived and designed the study; Xiansheng Xiang prepared the samples; Bing Pan and Kun Ma analyzed the data. Xiansheng Xiang, Xue Mei, Xiaojing Song, Chunlin Tan, and Shengmao Li wrote the manuscript. All authors reviewed the manuscript. All authors read and approved the final manuscript.

Funding

Funding from the Regional Collaborative Innovation Project of North Sichuan Medical College and the Affiliated Nanbu County People’s Hospital to Tielong Tang, the Scientific Startup Foundation for Doctors of North Sichuan Medical College to Xue Mei (CBY23-QDA11), and the Nanchong Science and Technology Bureau awarded to Lin Liu (No. 22SXQT0162).

Data availability

No datasets were generated or analyzed during the current study.

Declarations

Conflict of interest

The authors declare that there are no competing interests.

Ethical approval

Ethical approval for this study (No. 2025–88) was provided by the Ethics Committee on Biomedical Research, Nanbu County People’s Hospital.

Informed consent

Not applicable.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Xiansheng Xiang and Xue Mei contributed equally to this work.

Contributor Information

Qiang Dong, Email: dqiang@gmail.com.

Tielong Tang, Email: cdzt2004@163.com.

References

  • 1.Dyrskjøt L, Hansel DE, Efstathiou JA, Knowles MA, Galsky MD, Teoh J et al (2023) Bladder cancer. Nat Rev Dis Primers 9(1):58. 10.1038/s41572-023-00468-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Theodorescu D, Li Z, Li X (2022) Sex differences in bladder cancer: emerging data and call to action. Nat Rev Urol 19(8):447–449. 10.1038/s41585-022-00591-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Internationa W. Bladder Cancer Statistics 2022. Available from: https://www.wcrf.org/cancer-trends/bladder-cancerstatistics.
  • 4.Dobruch J, Daneshmand S, Fisch M, Lotan Y, Noon AP, Resnick MJ et al (2016) Gender and bladder cancer: a collaborative review of etiology, biology, and outcomes. Eur Urol 69(2):300–310. 10.1016/j.eururo.2015.08.037 [DOI] [PubMed] [Google Scholar]
  • 5.Hyldgaard JM, Jensen JB (2021) The inequality of females in bladder cancer. APMIS 129(12):694–699. 10.1111/apm.13183 [DOI] [PubMed] [Google Scholar]
  • 6.Radkiewicz C, Edgren G, Johansson ALV, Jahnson S, Häggström C, Akre O et al (2020) Sex differences in urothelial bladder cancer survival. Clin Genitourin Cancer 18(1):26–34. 10.1016/j.clgc.2019.10.020 [DOI] [PubMed] [Google Scholar]
  • 7.You S, Kim M, Widen S, Yu A, Galvan GC, Choi-Kuaea Y et al (2022) Characterizing molecular subtypes of high-risk non-muscle-invasive bladder cancer in African American patients. Urol Oncol 40(9):410. 10.1016/j.urolonc.2022.04.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Partin AW PC, Kavoussi LR, Dmochowski RR, Wein AJ Campbel-Wash-Wein Urology Twelfth Edition Review Elesevier Health Sciences, 2020.
  • 9.Freedman ND, Silverman DT, Hollenbeck AR, Schatzkin A, Abnet CC (2011) Association between smoking and risk of bladder cancer among men and women. JAMA 306(7):737–745. 10.1001/jama.2011.1142.PubMedPMID:21846855;PubMedCentralPMCID:PMCPMC3441175 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Wilhelm-Benartzi CS, Christensen BC, Koestler DC, Houseman EA, Schned AR, Karagas MR et al (2011) Association of secondhand smoke exposures with DNA methylation in bladder carcinomas. Cancer Causes Control 22(8):1205–1213. 10.1007/s10552-011-9788-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Bellamri M, Walmsley SJ, Brown C, Brandt K, Konorev D, Day A et al (2022) DNA damage and oxidative stress of tobacco smoke condensate in human bladder epithelial cells. Chem Res Toxicol 35(10):1863–1880. 10.1021/acs.chemrestox.2c00153 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Mertens LS, Claps F, Mayr R, Bostrom PJ, Shariat SF, Zwarthoff EC et al (2022) Prognostic markers in invasive bladder cancer: FGFR3 mutation status versus P53 and KI-67 expression: a multi-center, multi-laboratory analysis in 1058 radical cystectomy patients. Urol Oncol 40(3):110. 10.1016/j.urolonc.2021.10.010 [DOI] [PubMed] [Google Scholar]
  • 13.Apweiler R, Hermjakob H, Sharon N (1999) On the frequency of protein glycosylation, as deduced from analysis of the SWISS-PROT database. Biochim Biophys Acta 1473(1):4–8. 10.1016/s0304-4165(99)00165-8. (PubMed PMID: 10580125) [DOI] [PubMed] [Google Scholar]
  • 14.Taniguchi N, Kizuka Y (2015) Glycans and cancer: role of N-glycans in cancer biomarker, progression and metastasis, and therapeutics. Adv Cancer Res 126:11–51. 10.1016/bs.acr.2014.11.001 [DOI] [PubMed] [Google Scholar]
  • 15.Veillon L, Fakih C, Abou-El-Hassan H, Kobeissy F, Mechref Y (2017) Glycosylation changes in brain cancer. ACS Chem Neurosci 9(1):51–72. 10.1021/acschemneuro.7b00271 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Ihara YIY, Ikezaki M, Matsui IS, Manabe S, Ito Y et al (2015) Glycoscience: biology and medicine. Springer, Japan Tokyo, pp 1091–1099 [Google Scholar]
  • 17.Song E, Mechref Y (2013) LC–MS/MS identification of the O-glycosylation and hydroxylation of amino acid residues of collagen α-1 (II) chain from bovine cartilage. J Proteome Res 12(8):3599–3609. 10.1021/pr400101t [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Freire T, Bay S, Vichier-Guerre S, Lo-Man R, Leclerc C (2006) Carbohydrate antigens: synthesis aspects and immunological applications in cancer. Mini Rev Med Chem 6(12):1357–1373. 10.2174/138955706778992996 [DOI] [PubMed] [Google Scholar]
  • 19.Pinho SS, Reis CA (2015) Glycosylation in cancer: mechanisms and clinical implications. Nat Rev Cancer 15(9):540–555. 10.1038/nrc3982 [DOI] [PubMed] [Google Scholar]
  • 20.Mereiter S, Balmaña M, Campos D, Gomes J, Reis CA (2019) Glycosylation in the era of cancer-targeted therapy: where are we heading? Cancer Cell 36(1):6–16. 10.1016/j.ccell.2019.06.006 [DOI] [PubMed] [Google Scholar]
  • 21.Fu C, Zhao H, Wang Y, Cai H, Xiao Y, Zeng Y et al (2016) Tumor-associated antigens: Tn antigen, sTn antigen, and T antigen. Hla 88(6):275–286. 10.1111/tan.12900 [DOI] [PubMed] [Google Scholar]
  • 22.da Costa V, van Vliet SJ, Carasi P, Frigerio S, García PA, Croci DO et al (2021) The Tn antigen promotes lung tumor growth by fostering immunosuppression and angiogenesis via interaction with macrophage galactose-type lectin 2 (MGL2). Cancer Lett 518:72–81. 10.1016/j.canlet.2021.06.012 [DOI] [PubMed] [Google Scholar]
  • 23.Doherty M, Theodoratou E, Walsh I, Adamczyk B, Stöckmann H, Agakov F et al (2018) Plasma N-glycans in colorectal cancer risk. Sci Rep. 10.1038/s41598-018-26805-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Dan X, Liu W, Ng TB (2016) Development and applications of lectins as biological tools in biomedical research. Med Res Rev 36(2):221–247. 10.1002/med.21363 [DOI] [PubMed] [Google Scholar]
  • 25.Fuster MM, Esko JD (2005) The sweet and sour of cancer: glycans as novel therapeutic targets. Nat Rev Cancer 5(7):526–542. 10.1038/nrc1649. (PubMed PMID: 16069816) [DOI] [PubMed] [Google Scholar]
  • 26.Reis CA, Osorio H, Silva L, Gomes C, David L (2010) Alterations in glycosylation as biomarkers for cancer detection. J Clin Pathol 63(4):322–329. 10.1136/jcp.2009.071035. (PubMed PMID: 20354203) [DOI] [PubMed] [Google Scholar]
  • 27.Freeze HH (2013) Understanding human glycosylation disorders: biochemistry leads the charge. J Biol Chem 288(10):6936–6945. 10.1074/jbc.R112.429274 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.S H. (2002) Glycosylation defining cancer malignancynew wine in an old bottle. Proc Natl Acad Sci 99:10231–10233 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Ohtsubo K, Marth JD (2006) Glycosylation in cellular mechanisms of health and disease. Cell 126(5):855–867. 10.1016/j.cell.2006.08.019 [DOI] [PubMed] [Google Scholar]
  • 30.Pinho SS, Reis CA, Paredes J, Magalhães AM, Ferreira AC, Figueiredo J et al (2009) The role of N-acetylglucosaminyltransferase III and V in the post-transcriptional modifications of E-cadherin. Hum Mol Genet 18(14):2599–2608. 10.1093/hmg/ddp194 [DOI] [PubMed] [Google Scholar]
  • 31.Kim S-H, Turnbull J, Guimond S (2011) Extracellular matrix and cell signalling: the dynamic cooperation of integrin, proteoglycan and growth factor receptor. J Endocrinol 209(2):139–151. 10.1530/joe-10-0377 [DOI] [PubMed] [Google Scholar]
  • 32.Slawson C, Copeland RJ, Hart GW (2010) O-GlcNAc signaling: a metabolic link between diabetes and cancer? Trends Biochem Sci 35(10):547–555. 10.1016/j.tibs.2010.04.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Ma Z, Vosseller K (2014) Cancer metabolism and elevated O-GlcNAc in oncogenic signaling. J Biol Chem 289(50):34457–34465. 10.1074/jbc.R114.577718 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Zhu WLB, Andrews DW (2001) Cytoplasmic O-glycosylation prevents cell surface transport of E-cadherin during apoptosis. EMBO J 20:5999–6007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Gilgunn S, Conroy PJ, Saldova R, Rudd PM, O’Kennedy RJ (2013) Aberrant PSA glycosylation–a sweet predictor of prostate cancer. Nat Rev Urol 10(2):99–107. 10.1038/nrurol.2012.258 [DOI] [PubMed] [Google Scholar]
  • 36.Zurawski VR Jr., Orjaseter H, Andersen A, Jellum E (1988) Elevated serum CA 125 levels prior to diagnosis of ovarian neoplasia: relevance for early detection of ovarian cancer. Int J Cancer 42(5):677–680. 10.1002/ijc.2910420507 [DOI] [PubMed] [Google Scholar]
  • 37.Goldstein MJ, Mitchell EP (2005) Carcinoembryonic antigen in the staging and follow-up of patients with colorectal cancer. Cancer Invest 23(4):338–351. 10.1081/cnv-58878 [DOI] [PubMed] [Google Scholar]
  • 38.Safi F, Schlosser W, Kolb G, Beger HG (1997) Diagnostic value of CA 19–9 in patients with pancreatic cancer and nonspecific gastrointestinal symptoms. J Gastrointest Surg 1(2):106–112. 10.1016/s1091-255x(97)80097-2 [DOI] [PubMed] [Google Scholar]
  • 39.Zeng LYXJ, Chen HY, Mao SQ, Liu L, Zhang L (2023) Glycoprofiling of early non-small cell lung cancer using lectin microarray technology. Oncologie 25:469–480 [Google Scholar]
  • 40.Liang Y, Han P, Wang T, Ren H, Gao L, Shi P et al (2019) Stage-associated differences in the serum N- and O-glycan profiles of patients with non-small cell lung cancer. Clin Proteomics. 10.1186/s12014-019-9240-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Ahmad A, Ferdosi S, Ho TH, Castle EP, Stanton ML, Borges CR (2018) Behavior of blood plasma glycan features in bladder cancer. PLoS ONE. 10.1371/journal.pone.0201208 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.RodrÍguez E, Schetters STT, van Kooyk Y (2018) The tumour glyco-code as a novel immune checkpoint for immunotherapy. Nat Rev Immunol 18(3):204–211. 10.1038/nri.2018.3 [DOI] [PubMed] [Google Scholar]
  • 43.Dang K, Zhang W, Jiang S, Lin X, Qian A (2020) Application of lectin microarrays for biomarker discovery. ChemistryOpen 9(3):285–300. 10.1002/open.201900326 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Song KJ, Jeon SK, Moon SB, Park JS, Kim JS, Kim J et al (2017) Lectin from Sambucus sieboldiana abrogates the anoikis resistance of colon cancer cells conferred by N-acetylglucosaminyltransferase V during hematogenous metastasis. Oncotarget 8(26):42238–42251. 10.18632/oncotarget.15034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Oungsakul P, Choi E, Shah AK, Mohamed A, O’Leary C, Duffy D et al (2021) Candidate Glycoprotein Biomarkers for Canine Visceral Hemangiosarcoma and Validation Using Semi-Quantitative Lectin/Immunohistochemical Assays. Vet Sci. 10.3390/vetsci8030038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Kojima S, Jay M (1987) An experimental study on differential diagnosis of tumor from inflammation by using 125I labeled Pisum sativum agglutinin. Eur J Nucl Med 13(9):474–477. 10.1007/bf00281863 [DOI] [PubMed] [Google Scholar]
  • 47.Cooper D, Ilarregui JM, Pesoa SA, Croci DO, Perretti M, Rabinovich GA (2010) Multiple functional targets of the immunoregulatory activity of galectin-1: Control of immune cell trafficking, dendritic cell physiology, and T-cell fate. Methods Enzymol 480:199–244. 10.1016/s0076-6879(10)80011-4 [DOI] [PubMed] [Google Scholar]
  • 48.Nambiar DK, Aguilera T, Cao H, Kwok S, Kong C, Bloomstein J et al (2019) Galectin-1-driven T cell exclusion in the tumor endothelium promotes immunotherapy resistance. J Clin Invest 129(12):5553–5567. 10.1172/jci129025 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Tomana M, Matousovic K, Julian BA, Radl J, Konecny K, Mestecky J (1997) Galactose-deficient IgA1 in sera of IgA nephropathy patients is present in complexes with IgG. Kidney Int 52(2):509–516. 10.1038/ki.1997.361 [DOI] [PubMed] [Google Scholar]
  • 50.Schoenbeck S, Hammen MJ, Kagnoff MF (1989) Vicia villosa agglutinin separates freshly isolated Peyer’s Patch T cells into interleukin 5- or interleukin 2-producing subsets. J Exp Med 169(4):1491–1496. 10.1084/jem.169.4.1491 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Negedu MN, Duckworth CA, Yu LG (2022) Galectin-2 in health and diseases. Int J Mol Sci. 10.3390/ijms24010341 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Sasaki T, Saito R, Oyama M, Takeuchi T, Tanaka T, Natsume H et al (2020) Galectin-2 has bactericidal effects against Helicobacter pylori in a β-galactoside-dependent manner. Int J Mol Sci. 10.3390/ijms21082697 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

No datasets were generated or analyzed during the current study.

Articles from International Urology and Nephrology are provided here courtesy of Springer

RESOURCES