O-GlcNAcylation Facilitates the Interaction between Keratin 18 and Isocitrate Dehydrogenases and Potentially Influencing Cholangiocarcinoma Progression

Glycosylation plays a pivotal role in the intricate landscape of human cholangiocarcinoma (CCA), actively participating in key pathophysiological processes driving tumor progression. Among the various glycosylation modifications, O-linked β-N-acetyl-glucosamine modification (O-GlcNAcylation) emerges as a dynamic regulator influencing diverse tumor-associated biological activities. In this study, we employed a state-of-the-art chemical proteomic approach to analyze intact glycopeptides, unveiling the critical role of O-GlcNAcylation in orchestrating Keratin 18 (K18) and its interplay with tricarboxylic acid (TCA) cycle enzymes, specifically isocitrate dehydrogenases (IDHs), to propel CCA progression. Our findings shed light on the mechanistic intricacies of O-GlcNAcylation, revealing that site-specific modification of K18 on Ser 30 serves as a stabilizing factor, amplifying the expression of cell cycle checkpoints. This molecular event intricately fosters cell cycle progression and augments cellular growth in CCA. Notably, the interaction between O-GlcNAcylated K18 and IDHs orchestrates metabolic reprogramming by down-regulating citrate and isocitrate levels while elevating α-ketoglutarate (α-KG). These metabolic shifts further contribute to the overall tumorigenic potential of CCA. Our study thus expands the current understanding of protein O-GlcNAcylation and introduces a new layer of complexity to post-translational control over metabolism and tumorigenesis.


Figures
Figure S1.O-GlcNAcylation and OGT are upregulated in human CCA tumor tissues.

Figure S3 .
Figure S3.OGT and OGA mRNA expression are dysregulated in CCA and correlate with overall survival (OS).

Figure S4 .
Figure S4.OGT and OGA mRNA expression are dysregulated in CCA cell lines.

Figure S5 .
Figure S5.Global O-GlcNAcylation levels in CCA cells are regulated by 5S and TMG treatment.

Figure S6 .
Figure S6.Regulation of O-GlcNAcylation affects cell proliferation in CCA cells.

Figure S9 .
Figure S9.O-GlcNAcylation affects the cell cycle progression of CCA cells.

Figure S10 .
Figure S10.Western blot analysis of the cell cycle and apoptosis marker in 5S-or TMG-treated RBE cells.

Figure S12 .
Figure S12.Global O-GlcNAcylation of the CCA cells is affected by regulating OGT and OGA levels.

Figure S17 .
Figure S17.Profiling of the intact glycosites in HuCCT1 and HIBEpiC cells.

Figure S18 .
Figure S18.Numbers of the intact glycosites with specific glycan in HuCCT1 and HIBEpiC cells.

Figure S25 .
Figure S25.Confocal fluorescence imaging of K18 filament organization in RBE cells.

Figure S31 .
Figure S31.Analysis of interacting proteins of K18 protein.

Figure S32 .
Figure S32.Analysis of enzymes in the TCA cycle interacting with K18 of RBE stable cell lines.

Figure S33 .
Figure S33.Distribution of K18 protein in CCA cells.

Figure S37 .
Figure S37.Ser202 of IDH2 contributes to its interaction with K18 in CCA cells.

Figure S38 .
Figure S38.Measurement of glycolytic and oxidative phosphorylation (OXPHOS) ATP levels in CCA cells.

Figure S39 .
Figure S39.O-GlcNAcylation of K18 contributes to the resistance of H2O2 stimulation in CCA cells.

Figure S2 .
Figure S2.O-GlcNAcylation is dysregulated in human CCA.Western blot analysis of O-GlcNAcylated proteins, OGT, and OGA levels in the 15 CCA tumor tissues (T) and adjacent normal tissues (N).Equal loading was confirmed using β-actin.

Figure S3 .
Figure S3.OGT and OGA mRNA expression are dysregulated in CCA and correlate with overall survival (OS).(a-b) The OGT (a) and OGA (b) mRNA expression in CCA tumor tissues and adjacent normal tissues based on GSE32879, GSE107943, GSE119336, GSE76297, and TCGA datasets.(c-d) Kaplan-Meier survival curve in the high and low OGT (c) or OGA (d) mRNA expression groups.Data (Figure S3a,b) were shown as the mean ± standard deviation (SD); statistical significance was determined by Student's t tests (two-tailed, *P < 0.05 and ***P < 0.001, ns, not significant).The P-value of the Kaplan-Meier survival curve (Figure S3c,d) was analyzed by log-rank (Mantel-Cox) test.

Figure S4 .
Figure S4.OGT and OGA mRNA expression are dysregulated in CCA cell lines.(a-b) The relative expression level of OGT (a) and OGA (b) mRNA in three CCA cells and HIBEpiC cells was determined by quantitative real-time PCR (qRT-PCR).Data were shown as the mean ± SD; statistical significance was determined by Student's t tests (two-tailed, *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001, ns, not significant).

Figure S5 .
Figure S5.Global O-GlcNAcylation levels in CCA cells are regulated by 5S and TMG treatment.(a-c) Time and dose-dependent analysis of the O-GlcNAcylated protein, OGT, and OGA levels in 5S-or TMG-treated HuCCT1 (a), RBE (b), and HCCC-9810 (c) cells by Western blot.Equal loading was confirmed using β-actin.

Figure S8 .
Figure S8.Cell apoptosis assay of 5S-or TMG-treated RBE cells.(a) The bivariate density plot in flow cytometry indicated the cell population of early apoptotic cells (FITC + /PE -) and late apoptotic cells (FITC + /PE + ).(b) The quantitative analysis for the cell population of early apoptotic cells (FITC + /PE -) and late apoptotic cells (FITC + /PE + ) in 5S-or TMG-treated RBE cells.Data were shown as the mean ± SD; statistical significance was determined by Student's t tests (two-tailed, *P < 0.05 and **P < 0.01).

Figure S9 .
Figure S9.O-GlcNAcylation affects the cell cycle progression of CCA cells.(a) Cell cycle distribution assay of 5S-or TMG-treated HuCCT1 cells.(b) Cell cycle distribution assay of 5S-or TMG-treated RBE cells.(c) Histogram plot in flow cytometry indicated the percentage of cell populations in the G0/G1, S, or G2/M phase of RBE cells.Data were shown as the mean ± SD; statistical significance was determined by Student's t tests (two-tailed, *P < 0.05 and **P < 0.01).

Figure S16 .
Figure S16.Profiling of protein O-GlcNAcylation in HuCCT1 and HIBEpiC cells.(a) Overlap of the identified O-GlcNAcylated proteins between HuCCT1 and HIBEpiC cells.(b) Motif analysis of O-GlcNAc sites identified in HuCCT1 and HIBEpiC cells using the pLogo.The log-odds binomial probability and adjusted P-value (adjusted by a conservative Bonferroni correction) were shown.(c) Cellular component gene ontology (GO) terms of the identified O-GlcNAcylated proteins in HuCCT1 and HIBEpiC cells using the Database for Annotation, Visualization and Integrated Discovery (DAVID).The annotation terms and enrichment P-values (adjusted by a modified Fisher's exact test) were shown.

Figure S17 .
Figure S17.Profiling of the intact glycosites in HuCCT1 and HIBEpiC cells.(a) Overlap of the identified intact glycosites in HuCCT1 or HIBEpiC cells between three replicates.(b) Overlap of the identified intact glycosites, glycosites, and glycoproteins between HuCCT1 and HIBEpiC cells.(c) Total numbers and percentages of the intact N-linked glycosites, intact mucin-type O-linked glycosites, and intact O-GlcNAc sites identified in HuCCT1 and HIBEpiC cells.(d) Motif analysis of the N-linked glycosites and mucin-type O-linked glycosites identified in HuCCT1 and HIBEpiC cells using the pLogo.The log-odds binomial probability and adjusted P-value (adjusted by a conservative Bonferroni correction) were shown.

Figure S18 .
Figure S18.Numbers of the intact glycosites with specific glycan in HuCCT1 and HIBEpiC cells.(a) Numbers of the intact N-linked glycosites with specific types (High-Mannose, Complex/Hybrid, Fucose, and Neu5Ac) of N-glycans.(b) Numbers of the intact mucin-type O-linked glycosites with specific O-glycans.

Figure S19 .
Figure S19.Glycan compositions from the intact glycosites in HuCCT1 and HIBEpiC cells.(a) The putative glycan structure of 91 N-glycans was shown.The putative glycan structure of O-GlcNAc and 45 mucin-type O-glycans was shown.The red or blue boxes were identified only in HuCCT1 or HIBEpiC cells, respectively.

Figure S27 .
Figure S27.O-GlcNAcylation of K18 promotes cell growth in RBE cells.(a) Western blot analysis of K18 in RBE stable cell lines with shK18-1~3.(b) Western blot analysis of K18 in RBE stable cell lines with small hairpin RNA K18 knockdown (shK18) and re-expression of shK18-resistant FLAG-K18 wild-type (shK18 + WT) or FLAG-K18 S30A (shK18 + S30A).Random small hairpin RNA with an empty vector (shNC + Mock) was used as a negative control.(c) CCK-8 analysis of RBE stable cell lines.Absorbance was measured for cell viability.(d) Clonogenic assay of cell proliferation in RBE stable cell lines.Colony numbers were quantitatively analyzed at the bottom.(e) Cell cycle distribution assays of RBE stable cell lines.Histogram plot in flow cytometry indicated the percentage of cell populations in the G0/G1, S, or G2/M phase.Quantitative analysis was shown in the right panel.(f) Cell cycle marker analysis of RBE stable cell lines by Western blot.Protein levels of FOXM1, cMyc (G1/S transition markers), and BUB1(G2/M transition marker) were analyzed.(g) Degradation analysis of K18 in RBE cells analyzed by Western blot.The cells were incubated with DMSO (vehicle), 200 μM 5S, or 1 μM TMG for 48 h, followed by treatment with 10 μg/mL

Figure S28 .
Figure S28.O-GlcNAcylation of K18 promotes cell growth in HCCC-9810 cells.(a) Western blot analysis of K18 in HCCC-9810 stable cell lines with shK18-1~3.(b) Western blot analysis of K18 in HCCC-9810 stable cell lines with small hairpin RNA K18 knockdown (shK18) and re-expression of shK18-resistant FLAG-K18 wild-type (shK18 + WT) or FLAG-K18 S30A (shK18 + S30A).Random small hairpin RNA with an empty vector (shNC + Mock) was used as a negative control.(c) CCK-8 analysis of HCCC-9810 stable cell lines.Absorbance was measured for cell viability.(d) Clonogenic assay of cell proliferation in HCCC-9810 stable cell lines.Colony numbers were quantitatively analyzed at the bottom.(e) Cell cycle distribution assays of HCCC-9810 stable cell lines.Histogram plot in flow cytometry indicated the percentage of cell populations in the G0/G1, S, or G2/M phase.Quantitative analysis was shown in the right panel.(f) Cell cycle marker analysis of HCCC-9810 stable cell lines by Western blot.Protein levels of FOXM1, cMyc (G1/S transition markers), and BUB1(G2/M transition marker) were analyzed.(g) Degradation analysis of K18 in HCCC-9810 cells analyzed by Western blot.The cells were incubated with DMSO (vehicle), 200 μM 5S,

Figure S30 .
Figure S30.K18 and its O-GlcNAcylation are dysregulated in CCA.Representative images of K18 and its O-GlcNAcylation levels from nine pairs of CCA tumor tissues (T) and adjacent normal tissues (N) by Western blot analysis.Equal loadings were confirmed using β-actin.

Figure S31 .
Figure S31.Analysis of interacting proteins of K18 protein.(a) Workflow of the analysis of interacting proteins with K18 in HuCCT1 shK18 cells transfected with FLAG-K18 WT or FLAG-K18 S30A by LC-MS/MS.(b) Western blot and immunoprecipitation analysis showing the FLAG-K18 protein levels in HuCCT1 shK18 cells transfected with FLAG-K18 WT or FLAG-K18 S30A .Equal loadings were confirmed using β-actin.

Figure S33 .
Figure S33.Distribution of K18 protein in CCA cells.(a-b) RBE cells cotransfected with FLAG-IDH2 and HA-K18 WT or HA-K18 S30A (a), as well as FLAG-IDH3A and HA-K18 WT or HA-K18 S30A (b), were homogenized and subjected to subcellular fractionation, followed by immunoblotting analysis for cellular distribution of HA-K18, FLAG-IDH2 or FLAG-IDH3A.COX4 and HSP70 were used as mitochondrial and cytoplasmic markers, respectively.

Figure S34 .
Figure S34.Analysis of IDH(s) interacted with K18 in CCA cells.(a-c) RBE cells cotransfected with FLAG-IDH3A and HA-K18 WT or HA-K18 S30A (a), FLAG-IDH3B and HA-K18 WT or HA-K18 S30A (b), as well as FLAG-IDH3G and HA-K18 WT or HA-K18 S30A (c), followed by lysing and anti-FLAG immunoprecipitation.Western blot and immunoprecipitation analysis showing the HA-K18 and FLAG-IDH(s) protein levels.Equal loadings were confirmed using β-actin in all Western blot analyses.

Figure S37 .
Figure S37.Ser202 of IDH2 contributes to its interaction with K18 in CCA cells.

Figure S38 .
Figure S38.Measurement of glycolytic and oxidative phosphorylation (OXPHOS) ATP levels in CCA cells.Data were shown as the mean ± SD.

Figure S39 .
Figure S39.O-GlcNAcylation of K18 contributes to the resistance of H2O2 stimulation in CCA cells.(a) Cell viability assay of HuCCT1 stable cell lines (shK18 + WT and shK18 + S30A) treated with H2O2 at indicated concentrations for 12 h.(b) Relative ROS levels of HuCCT1 stable cell lines (shK18 + WT and shK18 + S30A) after treating with or without 500 μM H2O2 for 12 h.Data were shown as the mean ± SD; statistical significance was determined by Student's t tests (two-tailed, **P < 0.01 and ****P < 0.0001).