CLK3 regulates tumor angiogenesis by modulating HIF1α ubiquitination Authors: Jing Shen, Wen Zhang, Zhiqiang Chu, Lei Zhu, Yan Feng, Xuling Sun Published: 2026 ABSTRACT Tumor angiogenesis is a critical driver of cancer progression and is regulated by a complex network of signaling pathways. However, limited therapeutic efficacy and the emergence of drug resistance remain significant obstacles, necessitating the exploration of novel angiogenesis inhibitors, alternative pathways, and reliable biomarkers. In this study, we investigate the role of CDC-like kinase 3 (CLK3) in tumor angiogenesis and its underlying molecular mechanisms. Bioinformatic analysis of clinical datasets revealed that CLK3 expression positively correlated with angiogenesis-related pathways in colorectal cancer. In vitro and in vivo experiments confirmed that CLK3 overexpression promoted tumor angiogenesis, whereas CLK3 knockout inhibited this process. Mechanistically, we demonstrated that CLK3 directly interacted with and stabilizes hypoxia-inducible factor 1-alpha (HIF1α) by inhibiting its ubiquitin-proteasome-mediated degradation. Additionally, CLK3 promoted the phosphorylation of HIF1α at tyrosine residues. Furthermore, our functional assays confirmed that CLK3-mediated tumor angiogenesis was dependent on HIF1α. These findings highlight the importance of CLK3 in regulating tumor angiogenesis, and provide compelling evidence for targeting CLK3 as a potential therapeutic strategy for colorectal cancer treatment. Introduction Tumor angiogenesis is a complex biological process characterized by the de novo formation of new blood vessels within a tumor mass. This phenomenon is pivotal for cancer growth, progression, and metastasis, as it provides the tumor with essential nutrients and oxygen, supporting its continued expansion. Angiogenesis involves the activation of various signaling pathways, including IL-6/STAT3 signaling, HIF1 signaling and PI3K/AKT signaling, and the release of angiogenic factors that stimulate the proliferation and migration of endothelial cells, ultimately establishing a functional vascular network within the tumor microenvironment [1–5]. While targeting tumor angiogenesis is a cornerstone of cancer therapy, challenges such as inadequate treatment efficacy and acquired drug resistance persist [67]. These limitations necessitate the development of novel angiogenesis inhibitors, the exploration of additional regulatory pathways, and the identification of reliable biomarkers. Identifying novel signaling axes that govern the compensatory angiogenic pathways under hypoxic conditions is therefore essential for overcoming the limitations of current therapies. CLK3, also known as CDC-like kinase 3, is a nuclear dual-specificity kinase that phosphorylates substrates containing both serine/threonine and tyrosine residues [8]. CLK3 participates in regulating RNA splicing and translation by phosphorylating RNA processing factors, such as SRSF1 and SRSF3 [9]. Recently, CLK3 has been implicated in the development of various tumors, for example, CLK3 promotes cholangiocarcinoma progression by reprogramming nucleotide metabolism; deletion of CLK3 abrogates MFAP2-induced colorectal cancer invasion; CLK3, a direct target of miR-144, contributes to the aggressive progression of hepatocellular carcinoma; IL-6/STAT3 signaling can be activated by CLK3 [10–13]. Besides, our previous work also demonstrated that CLK3 activated the MYC signaling pathway to promote tumor proliferation [14]. Despite these advances, the functional involvement of CLK3 in the tumor microenvironment, particularly its role in orchestrating tumor angiogenesis, has not yet been explored. In this study, we demonstrated that CLK3 regulated angiogenesis in colorectal cancer through the analysis of multiple clinical datasets. Subsequent in vitro and in vivo experiments confirmed that CLK3 positively regulated tumor angiogenesis. Mechanistically, we revealed that CLK3 directly bound to HIF1α, inhibiting its degradation through the ubiquitin-proteasome pathway. Additionally, we found that CLK3 could phosphorylate HIF1α and demonstrated that CLK3-induced tumor angiogenesis depended on HIF1α. These findings further elucidated the role of CLK3 in oncogenesis, expanded the known downstream regulatory mechanisms of CLK3, and provided a theoretical basis for targeting CLK3 as a potentially effective therapeutic strategy for colorectal cancer. Materials and methods Antibodies and agents The detailed antibody information was listed in Table 1. CLK-IN-3 (HY-115470) and Ipivivint (HY-137443) were bought from MedChemExpress company.Table 1The information of antibodies.AntibodiesCatalog numbersManufacturersCLK1ab74044AbcamCLK2ab86147AbcamCLK33256CSTα-tubulin2144CSTVEGFA50661CSTHIF1α36169CSTHIF1β5537CSTPan Phospho-TyrosineAP0973ABclonalPan Phospho-Serine/ThreonineAP1475ABclonalK48-linkage Specific UbiquitinA3606ABclonalK63-linkage Specific PolyubiquitinA18164ABclonalPan Phospho-TyrosineAP0973ABclonalFlag tag14793CSTHis tag2365CST Table 1The information of antibodies.AntibodiesCatalog numbersManufacturersCLK1ab74044AbcamCLK2ab86147AbcamCLK33256CSTα-tubulin2144CSTVEGFA50661CSTHIF1α36169CSTHIF1β5537CSTPan Phospho-TyrosineAP0973ABclonalPan Phospho-Serine/ThreonineAP1475ABclonalK48-linkage Specific UbiquitinA3606ABclonalK63-linkage Specific PolyubiquitinA18164ABclonalPan Phospho-TyrosineAP0973ABclonalFlag tag14793CSTHis tag2365CST The information of antibodies. Cell lines Cells for LoVo (RRID: CVCL_0399), HCT116 (RRID: CVCL_0291), and HUVEC (RRID: CVCL_2959) were obtained from American Type Culture Collection (https://www.atcc.org/). All cells were cultured using Dulbecco’s Modified Eagle Medium plus 10% (v/v) fetal bovine serum at 37 °C under an environment with 5% carbon dioxide concentration. For hypoxic conditions, cell culture was conducted in a hypoxic incubator (Heracell Vios 160i CR, Thermo Scientific) with 1% oxygen concentration. All human cell lines used in this study were authenticated by short tandem repeat (STR) profilinge. All cell lines were routinely tested and confirmed to be free of mycoplasma contamination. Regarding the generation of CLK3 knockout (KO) cells, we utilized the CRISPR/Cas9 system by cloning the specific sgRNA targeting human CLK3 into the pSpCas9(BB)-2A-Puro (PX459) V2.0 plasmid, which was then transfected into colorectal cancer cells, followed by puromycin (2 μg/mL) selection for one week and subsequent single-clone cultivation. The successful establishment of CLK3 KO clones was validated through sequencing and Western blot analysis. The sgRNAs for CLK3 were listed as followed: sgRNA#1 (GACTGCGATACCCGTCCCGA); sgRNA#2 (GCCGTGACAGCGATACATAC); sgRNA#3 (CTCGTCATCGTCGGCGATCG). Cell counting Kit-8 (CCK8) assays For the cell proliferation assay, 3000 LoVo and HCT116 cells were seeded into 96-well plates in 100 μL medium per well and cultured overnight. The CCK-8 Assay Kits (RM02823) were purchased from ABclonal Technology (Wuhan, China). At the indicated time points, 10 μL of CCK-8 reagent was added to each well. The absorbance was measured at 450 nm using a microplate reader. All assays were performed according to the manufacturer’s instructions in at least three independent experiments. Brdu assays 5000 LoVo and HCT116 cells were seeded into 96-well plates and incubated with 10 µM BrdU to label newly synthesized DNA. After incubation, cells were fixed with 4% paraformaldehyde. Cells were blocked with 5% BSA and incubated with an anti-BrdU primary antibody, followed by a fluorescence-conjugated secondary antibody. Nuclei were counterstained with DAPI, and slides were mounted with an anti-fade medium. BrdU incorporation was visualized under a fluorescence microscope, with BrdU-positive cells indicating active DNA synthesis. Quantification was performed by comparing BrdU-positive cells to total DAPI-stained nuclei. Enzyme-linked immunosorbent assays (ELISA) For the quantification of secreted VEGFA, 1×106 LoVo and HCT116 cells were seeded into 6-well plates and cultured until they reached ~80% confluency. The culture medium was then replaced with 1.5 mL of serum-free medium and incubated for an additional 24 h. Subsequently, the cell culture supernatants were collected and centrifuged at 1000 × g for 10 min to remove cell debris. The Human VEGFA ELISA Kits (RK00023, ABclonal Technology, Wuhan, China) were used to measure VEGFA concentrations according to the manufacturer’s instructions. The absorbance was measured at 450 nm using a microplate reader. The final concentrations were normalized to the cell number. HRE-luciferase reporter assay 1 × 105 LoVo and HCT116 cells were seeded in 24-well plates and transfected with the HRE-luciferase reporter plasmid (P11208, MIAOLING) together with a Renilla luciferase plasmid (D2760, Beyotime) for normalization using HighGene (RM09014, ABclonal). For hypoxia treatment, cells were incubated at 1% O2 for 12–16 h. Luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer’s instructions. Firefly luciferase activity was normalized to Renilla luciferase, and relative luciferase activity was calculated. Real-time PCR Total RNA was isolated using TRIzol reagent (Invitrogen) according to the manufacturer’s instructions. cDNA was synthesized using a reverse transcription kit (RK20429, ABclonal). Quantitative real-time PCR was performed with SYBR Green Master Mix (RK21203, ABclonal). A-tubulin was used as an internal control. Relative mRNA expression was calculated using the 2−ΔΔCt method. Primer sequences were listed in Table 2.Table 2Primer sequences used for RT-PCR.GeneForward Primer (5' - 3')Reverse Primer (5' - 3')HIF1AGAACGTCGAAAAGAAAAGTCTCGCCTTATCAAGATGCGAACTCACAGLUT1GGCCAAGAGTGTGCTAAAGAAACAGCGTTGATGCCAGACAGPGK1TGGACGTTAAAGGGAAGCGGGCTCATAAGGACTACCGACTTGGMMP2TACAGGATCATTGGCTACACACCGGTCACATCGCTCCAGACTCA9GGATCTACCTACTGTTGAGGCTCATAGCGCCAATGACTCTGGTα-tubulinACCTTAACCGCCTTATTAGCCACACCACGGTACAACAGGCA Table 2Primer sequences used for RT-PCR.GeneForward Primer (5' - 3')Reverse Primer (5' - 3')HIF1AGAACGTCGAAAAGAAAAGTCTCGCCTTATCAAGATGCGAACTCACAGLUT1GGCCAAGAGTGTGCTAAAGAAACAGCGTTGATGCCAGACAGPGK1TGGACGTTAAAGGGAAGCGGGCTCATAAGGACTACCGACTTGGMMP2TACAGGATCATTGGCTACACACCGGTCACATCGCTCCAGACTCA9GGATCTACCTACTGTTGAGGCTCATAGCGCCAATGACTCTGGTα-tubulinACCTTAACCGCCTTATTAGCCACACCACGGTACAACAGGCA Primer sequences used for RT-PCR. Tube formation assay 96-well plates were pre-coated with 50 μL of growth factor-reduced Matrigel (Corning, NY, USA) per well and incubated at 37 °C for 30 minutes to allow polymerization. HUVECs (1.5 × 104 cells per well) were suspended in 50 μL of conditioned medium and seeded onto the Matrigel. The cells were incubated at 37 °C in a 5% CO2 atmosphere for 8 h to allow the formation of capillary-like networks. Images were captured using a phase-contrast microscope. The degree of angiogenesis was quantified by measuring the number of branching points using ImageJ software. Immunoblot (IB) and immunoprecipitation (IP) The methods had also been described in our earlier papers [15]. Cell or tissue samples were collected, and cell membranes or cell walls are lysed using NP40 extraction buffer to release proteins. The extracted protein concentration was measured using methods such as the BCA. Protein samples were loaded onto a polyacrylamide gel, and SDS-PAGE electrophoresis was performed to separate proteins based on size. Separated proteins were transferred onto a polyvinylidene fluoride (PVDF). The membrane was incubated with a specific primary antibody, allowing it to bind specifically to the target protein, followed by incubating with a secondary antibody. Chemiluminescence detection (ECL) or other methods were used to observe and record the intensity of protein bands. For immunoprecipitation, Specific antibodies were added to facilitate the formation of antigen-antibody complexes with the target protein. The protein-antibody complexes were combined with Protein A/G agarose to precipitate the complexes from the cell lysate. Trans-well assays 4–10,000 HUVEC cells with 200 μl serum-free culture medium were added into the upper chamber and 300 μl conditional culture medium was added into the lower chamber. After 12–16 h, the cells were fixed using 4% paraformaldehyde. Subsequently, cells were stained with crystal violet dye. Removed the cells from the upper chamber, and allowed the chamber to air-dry at room temperature. Capture images using an inverted microscope (Olympus IX73, 20× objective lens). Image analysis was performed using ImageJ software. Immunohistochemistry (IHC) The studies involving human participants underwent review and approval by the Ethical Committee of The First Affiliated Hospital of Shihezi University (KJ2020-018-01) and the study methodologies conformed to the standards set by the Declaration of Helsinki. All clinical samples were obtained from The First Affiliated Hospital of Shihezi University (From December 2020 to December 2022). Participants provided written informed consent to participate. For immunohistochemistry assays, tissue samples were collected and fixed with 4% paraformaldehyde. Paraffin-embedded blocks were prepared through dehydration and paraffin embedding. Thin sections, typically 4–5 micrometers in thickness, were cut from the paraffin-embedded blocks using a microtome, followed by deparaffinization, rehydration, and antigen retrieval. Samples were stained using enzyme-based methods with visualization through substrate reactions. Images were captured using a microscope (Olympus IX73, 10× objective lens). The quantitative analysis of immunohistochemical staining was independently performed by two experienced senior pathologists who were blinded to the clinical data and experimental groupings. Quantitative analysis of images was assessed using the Immunoreactive Score (IRS). Intensity scoring involves scoring the staining intensity, typically ranging from 0 (no staining) to 3 (strong staining). Percentage of positive cells scoring assesses the percentage of positive cells: 0 (0%), 1 (1–25%), 2 (26–50%), 3 (51–75%), 4 (76–100%). IRS was calculated as the product of the intensity score and the percentage of positive cells score. In vitro phosphorylation assay Flag-CLK3 was transiently expressed in HEK293T cells and immunoprecipitated using anti-Flag beads (HY-K0207, MCE). His-tagged HIF1α was expressed in E. coli and purified with Ni-NTA Magnetic Agarose Beads (HY-K0241, MCE). For the kinase reaction, Flag-CLK3–bound beads were incubated with purified His-HIF1α in kinase buffer (25 mM Tris-HCl, pH 7.5, 10 mM MgCl2, 1 mM DTT, 100 µM ATP) at 30 °C for 30–60 min. Reactions were terminated by boiling in SDS buffer, and HIF1α phosphorylation was detected by Western blot using a pan-phospho-tyrosine antibody. Animal experiments Approval for the animal study was obtained from the Institutional Animal Care and Use Committee at The First Affiliated Hospital of Shihezi University (A2020-040-01). All mice were housed in a specific pathogen-free (SPF) facility, with controlled temperature (22 ± 2 °C), humidity (50 ± 10%), and a 12-h light/dark cycle. Mice were provided with sterilized pellet diet and autoclaved water ad libitum. Animals were acclimatized for 7 days before experiments. Mice were housed in groups of 4–5 per cage in individually ventilated cages, with environmental enrichment provided. A triple-blinded procedure was applied. Specifically, the investigators responsible for animal handling, outcome assessment, and data analysis were blinded to group allocation to minimize potential bias. The sample size (n = 5 per group) was determined based on our previous experience with the subcutaneous xenograft model and prior literature to provide sufficient statistical power to detect significant differences in tumor growth. The 6-week-old nude female mice (BALB/cNJ-Foxn1^nu^/Gpt) were purchased from GemPharmatech LLC. After undergoing a health check, they were housed in a Specific Pathogen-Free (SPF) experimental environment. Mice were randomly assigned to three group. Each mouse received only one type of cell injection. Three cell lines (5 × 105) were injected subcutaneously into the dorsal area of the mice. Tumor dimensions, including the long and short diameters, were measured every 3 days. On the 16th day, the mice were euthanized by cervical dislocation, and the tumors were excised, weighed, and fixed. No samples or animals were excluded from the analysis. All animals that reached the experimental endpoint were included in the final data analysis. Tumor volume was calculated using the formula: 0.5 * long diameter * short diameter * short diameter. For survival analysis, mice were injected three cell lines subcutaneously, and were monitored until they reached the ethically defined humane endpoint. Statistical analysis Database for The Cancer Genome Atlas Program (TCGA, https://www.cancer.gov/ccg/research/genome-sequencing/tcga), Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/), GENEMANIA (https://genemania.org/), and The Human Protein Atlas (https://www.proteinatlas.org/) were used. GSEA v4.3.3 software was used for GSEA analysis. Data were shown as the mean ± standard deviation (SD). SPSS v26.0 and GraphPad Prism v9.0 software were used for statistical analysis The normality of data distribution was assessed using the Shapiro–Wilk test, and the homogeneity of variance was evaluated using Levene’s test. Student’s t test was used for normally distributed data with equal variance, while the Mann–Whitney U test was used for non-normally distributed data. ANOVA was used for comparisons among multiple groups. Pearson’s correlation test or Log-rank test was applied where appropriate. A significance level of P < 0.05 was considered meaningful. Results CLK3 functioned as a pro-angiogenesis factor in colorectal cancer The process of tumor angiogenesis is pivotal for the development, advancement, and spread of cancer. In order to elucidate the role of CLK3 in tumor angiogenesis, we initially analyzed the TCGA and GEO databases. GSEA analysis revealed that tumors with higher CLK3 expression exhibited activation of the tumor angiogenesis signaling pathway (Fig. 1A and Supplementary Fig. 1A). Multiple dataset analyses indicated a significant positive correlation between the expression levels of VEGFA and CLK3 (Fig. 1B and Supplementary Fig. 1B). Analysis of TCGA dataset indicated a significant elevation of CLK3 in tumor tissues compared to adjacent normal tissues (Supplementary Fig. 1C). We extracted proteins from 7 pairs of colorectal adenocarcinoma and adjacent tissues, and immunoblots indicated a significantly elevated expression level of CLK3 in tumor tissues compared to the corresponding normal tissues (Fig. 1C and Supplementary Fig. 1D). Subsequently, we conducted IHC experiments on 28 pairs of colorectal adenocarcinoma and adjacent tissues, and our findings revealed a significant increase in CLK3 and VEGFA expression levels in tumor tissues compared to the corresponding normal tissues (Fig. 1D, E). We further analyzed the expression levels of CLK3 and VEGFA, along with micro-vessel density (CD31-labeled), in 71 colorectal adenocarcinoma samples using IHC experiments. Our data indicated that tumors from patients with higher T stages and advanced clinical stages exhibited elevated levels of CLK3 expression (Fig. 1F, G). Moreover, CLK3 expression levels showed a significant positive correlation with both VEGFA expression and micro-vessel density (Fig. 1H). Patients with higher CLK3 expression were associated with poorer survival outcomes (Fig. 1I, J), and similar results were found in THE HUMAN PROTEIN ATLAS (Supplementary Fig. 1E). Multivariable Cox regression analysis demonstrated that high CLK3 expression was independently associated with poorer overall survival (Fig. 1K). These data confirmed that CLK3 was a prognostic biomarker in colorectal cancer, as its overexpression significantly aligned with both reduced patient survival and enhanced tumor angiogenesis.Fig. 1CLK3 may have acted as a promoter of tumor angiogenesis.A Gene Set Enrichment Analysis (GSEA) revealing the relationship between CLK3 and tumor angiogenesis using The Cancer Genome Atlas (TCGA) database (TCGA-COAD, TCGA-READ). B Analysis of the correlation between the expression levels of CLK3 and vascular endothelial growth factor A (VEGFA) using TCGA database (n for READ = 95; n for COAD = 288); Pearson’s correlation test. C Expression level of CLK3 in colorectal cancer and matched normal tissues detected by immunoblots using 7 pairs of primary colorectal tumors and matched adjacent normal tissues. D 28 paired colorectal tumor and matched normal tissue samples from 28 individual patients were used for IHC analysis; Representative images of immunohistochemistry (IHC) experiments; scale bar = 50 μm. E Analysis of CLK3 and VEGFA expression levels in colorectal cancer and matched normal tissues based on IHC from (D) (n = 28); Student’s t test and Mann–Whitney U test. F 71 colorectal tumor samples from 71 different patients were used for IHC analysis; Representative images of IHC experiments; scale bar = 50 μm; Student’s t test. G Analysis of CLK3 expression in colorectal cancer patients grouped by tumor stage (T) or clinical stage based on IHC from (F) (n = 71). H Correlation analysis of CLK3 expression, VEGFA expression, and micro-vessel density (quantified by CD31 staining) in colorectal cancer tissues (n = 71); Pearson’s correlation test. I Overall survival analysis of colorectal cancer patients grouped by CLK3 expression level using TCGA database (Low = 393, high = 250); Log-rank test. J Overall survival analysis of 71 colorectal cancer patients grouped by CLK3 expression level (low = 37, high = 34); Log-rank test. K Multivariable Cox regression analysis (incorporating age, sex, tumor stage, and CLK3 IHC score) showing that CLK3 is an independent prognostic factor in CRC. CLK3 may have acted as a promoter of tumor angiogenesis. A Gene Set Enrichment Analysis (GSEA) revealing the relationship between CLK3 and tumor angiogenesis using The Cancer Genome Atlas (TCGA) database (TCGA-COAD, TCGA-READ). B Analysis of the correlation between the expression levels of CLK3 and vascular endothelial growth factor A (VEGFA) using TCGA database (n for READ = 95; n for COAD = 288); Pearson’s correlation test. C Expression level of CLK3 in colorectal cancer and matched normal tissues detected by immunoblots using 7 pairs of primary colorectal tumors and matched adjacent normal tissues. D 28 paired colorectal tumor and matched normal tissue samples from 28 individual patients were used for IHC analysis; Representative images of immunohistochemistry (IHC) experiments; scale bar = 50 μm. E Analysis of CLK3 and VEGFA expression levels in colorectal cancer and matched normal tissues based on IHC from (D) (n = 28); Student’s t test and Mann–Whitney U test. F 71 colorectal tumor samples from 71 different patients were used for IHC analysis; Representative images of IHC experiments; scale bar = 50 μm; Student’s t test. G Analysis of CLK3 expression in colorectal cancer patients grouped by tumor stage (T) or clinical stage based on IHC from (F) (n = 71). H Correlation analysis of CLK3 expression, VEGFA expression, and micro-vessel density (quantified by CD31 staining) in colorectal cancer tissues (n = 71); Pearson’s correlation test. I Overall survival analysis of colorectal cancer patients grouped by CLK3 expression level using TCGA database (Low = 393, high = 250); Log-rank test. J Overall survival analysis of 71 colorectal cancer patients grouped by CLK3 expression level (low = 37, high = 34); Log-rank test. K Multivariable Cox regression analysis (incorporating age, sex, tumor stage, and CLK3 IHC score) showing that CLK3 is an independent prognostic factor in CRC. CLK3 positively regulated tumor angiogenesis in vitro To further investigate the role of CLK3 in tumor angiogenesis, we utilized lentivirus to establish stable high-expression CLK3 cell lines in LoVo and HCT116 cells. Immunoblots confirmed the successful construction of the cell lines (Fig. 2A). ELISA assays indicated that cultures with higher CLK3 expression exhibited higher levels of VEGFA (Fig. 2B). To determine whether these secreted factors directly influence endothelial cell behavior, HUVEC cells were cultured in conditioned media (CMs) derived from vector and CLK3 cells. Consistent with the ELISA results, The CCK8 and Brdu experiments revealed that the CMs from CLK3 cells was more effective in promoting the proliferation ability of HUVEC cells (Fig. 2C–E). Tube formation assays showed that cells cultured in CMs from CLK3 cells exhibited enhanced tube formation ability (Fig. 2F, G). Also, transwell experiments revealed that the migration capacity of HUVEC cells was significantly increased when cultured in CMs from CLK3 cells (Fig. 2H, I).Fig. 2CLK3 overexpression promotes colorectal cancer angiogenesis in vitro.A Detection of the expression of identified proteins in Vector control or CLK3-overexpressing colorectal cancer cells by immunoblotting. B Measurement of VEGFA expression levels in Vector or CLK3 colorectal cancer cells using ELISA assay (n = 3); Student’s t test. C–E HUVECs were cultured with different conditioned medias; CCK-8 (C) and BrdU (D, E) assays were performed to assess the proliferation ability of HUVECs. (Red: BrdU; Blue: DAPI) (n = 3); scale bar = 30 μm; Student’s t test. F Representative images of tube formation assays; scale bar = 1000 μm. G Statistical analysis corresponding to (F) (n = 3); Student’s t test. H Representative images of transwell migration assays; scale bar = 100 μm. I Statistical analysis corresponding to Fig. 2H (n = 3); Student’s t test. CLK3 overexpression promotes colorectal cancer angiogenesis in vitro. A Detection of the expression of identified proteins in Vector control or CLK3-overexpressing colorectal cancer cells by immunoblotting. B Measurement of VEGFA expression levels in Vector or CLK3 colorectal cancer cells using ELISA assay (n = 3); Student’s t test. C–E HUVECs were cultured with different conditioned medias; CCK-8 (C) and BrdU (D, E) assays were performed to assess the proliferation ability of HUVECs. (Red: BrdU; Blue: DAPI) (n = 3); scale bar = 30 μm; Student’s t test. F Representative images of tube formation assays; scale bar = 1000 μm. G Statistical analysis corresponding to (F) (n = 3); Student’s t test. H Representative images of transwell migration assays; scale bar = 100 μm. I Statistical analysis corresponding to Fig. 2H (n = 3); Student’s t test. Next, we employ CRISPR-Cas9 technology to knock out CLK3 in LoVo and HCT116 cells (Fig. 3A). We collected CMs from cells expressing sgNC and sgCLK3. ELISA experiments indicated that the CMs from sgCLK3 cells exhibited lower levels of VEGFA (Fig. 3B). Similarly, we cultured HUVEC cells in CMs, and confirmed that cells cultured in CMs from sgCLK3 cells exhibited lower proliferation (Fig. 3C–E), tube formation (Fig. 3F, G), and migration abilities (Fig. 3H, I). In summary, these results suggested that CLK3 had the ability to positively regulate tumor angiogenesis in vitro.Fig. 3Loss of CLK3 inhibits colorectal cancer angiogenesis in vitro.A Detection of the expression of identified proteins in sgnc (non-targeting control) or sgCLK3 (two kinds of CLK3 knockout cells, #1, #2) colorectal cancer cells by immunoblotting. B Measurement of VEGFA expression levels in sgNC or sgCLK3 colorectal cancer cells using ELISA (n = 3); One-way ANOVA test. HUVECs were cultured with different conditioned medias; CCK-8 (C) and BrdU (D, E) assays were performed to assess the proliferation ability of HUVECs (n = 3); scale bar = 30 μm; One-way ANOVA test. F Representative images of tube formation assays; scale bar = 100 μm. G Statistical analysis for (F) (n = 3); One-way ANOVA test. H Representative images of transwell migration assays; scale bar = 100 μm. I Statistical analysis for (H) (n = 3); One-way ANOVA test. Loss of CLK3 inhibits colorectal cancer angiogenesis in vitro. A Detection of the expression of identified proteins in sgnc (non-targeting control) or sgCLK3 (two kinds of CLK3 knockout cells, #1, #2) colorectal cancer cells by immunoblotting. B Measurement of VEGFA expression levels in sgNC or sgCLK3 colorectal cancer cells using ELISA (n = 3); One-way ANOVA test. HUVECs were cultured with different conditioned medias; CCK-8 (C) and BrdU (D, E) assays were performed to assess the proliferation ability of HUVECs (n = 3); scale bar = 30 μm; One-way ANOVA test. F Representative images of tube formation assays; scale bar = 100 μm. G Statistical analysis for (F) (n = 3); One-way ANOVA test. H Representative images of transwell migration assays; scale bar = 100 μm. I Statistical analysis for (H) (n = 3); One-way ANOVA test. CLK3 positively regulated the HIF-1 signaling pathway To further explore the underlying mechanism by which CLK3 regulated tumor angiogenesis, we initially conducted GSEA using multiple datasets, revealing that CLK3 could positively regulate the hypoxia signaling pathway (Fig. 4A and Supplementary Fig. 2A). Considering that HIF-1α was easily degraded under normoxic conditions, we treated the tumor cells with 1% oxygen concentration to create a hypoxic environment. Subsequently, immunoblots experiments proved that overexpression of CLK3 lead to the upregulation of HIF1α expression, while it did not show a significant impact on HIF1β in both normoxic condition and hypoxic condition (Fig. 4B and Supplementary Fig. 2B). CLK3 overexpression markedly enhanced HIF-1 transcriptional activity in HRE-luciferase reporter activity assays (Supplementary Fig. 2C). Consistently, qPCR analysis showed that overexpression of CLK3 significantly upregulated multiple classical HIF-1 downstream target genes (Supplementary Fig. 2D). To investigate whether the regulation of HIFα expression by CLK3 was dependent on its enzymatic activity, we separately transfected wild-type CLK3 and an enzymatically inactive mutant (K186M) into colorectal cancer cells. Immunoblots indicated that K186M mutant could not regulate HIF1α expression, suggesting that CLK3 enzymatic activity was crucial for regulating HIF1α expression (Fig. 4C). We also found that knocking out CLK3 significantly inhibited the expression of HIF1α (Fig. 4D). HRE-luciferase reporter activity assays and qPCR analysis obviously inhibited HIF-1 transcriptional activity (Supplementary Fig. 2E, F). Re-expression of CLK3 WT, but not CLK3 K186M, restored HIF1α protein levels in CLK3-knockout LoVo cells (Supplementary Fig. 2G). knockout of CLK3 did not affect HIF-2α expression (Supplementary Fig. 2H). We further used RT-PCR experiments to assess the mRNA expression levels of HIF1α in tumor cells. The results indicated that CLK3 overexpression, CLK3 knockout, and CLK3 inhibitor (CLK-IN-T3) application did not regulate the mRNA expression levels of HIF1α, suggesting that CLK3 regulates the expression levels of HIF1α through post-translational regulation (Supplementary Fig. 3A–D). The ubiquitin-proteasome pathway and the autophagy-lysosome pathway were two major protein degradation pathways [16]. We used chloroquine (CQ) to inhibit the autophagy-lysosome pathway and MG132 to inhibit the ubiquitin-proteasome pathway. It was worth emphasizing that the use of MG132 could eliminate the upregulation of HIF1α mediated by CLK3 overexpression, implying that CLK3 might regulate the poly-ubiquitination level of HIF1α (Fig. 4E). Also, MG132, but not CQ, application could reverse the downregulation of HIF1A expression caused by CLK3 knockdown (Supplementary Fig. 3E). Then, we transfected His-tagged ubiquitin plasmid into LoVo cells and used MG132 to inhibit the degradation of endogenous proteins. Immunoprecipitation and immunoblot experiments indicated that CLK3 overexpression obviously inhibited the poly-ubiquitination levels of HIF1α (Fig. 4F). Similarly, knocking out CLK3 significantly upregulated the poly-ubiquitination levels of HIF1α (Fig. 4G). Also, loss of CLK3 significantly accelerated the degradation of HIF1α (Fig. 4H, I). Furthermore, CLK3 overexpression reduced K48-linked, but not K63-linked, polyubiquitination of HIF1α (Fig. 4J and Supplementary Fig. 3F). Collectively, these findings revealed that CLK3 positively regulated the HIF1 signaling pathway by stabilizing the HIF1α protein.Fig. 4CLK3 positively regulates HIF1 signaling by stabilizing HIF1α.A GSEA revealing the relationship between CLK3 and hypoxia signaling using TCGA database (TCGA-COAD). B, C Detection of the expression levels of hypoxia signaling-related proteins in Vector, CLK3 WT, or CLK3 K186M mutant cells by immunoblotting. D Detection of hypoxia signaling-related proteins in sgNC or sgCLK3 cells using IB. E Detection of HIF1α expression in Vector or CLK3 cells under chloroquine (CQ) or MG132 treatment. F MG132 is used to inhibit the degradation of endogenous proteins. Detection of poly-ubiquitination levels of HIF1α in Vector, CLK3 WT, and CLK3 K186M cells. G MG132 is used to inhibit the degradation of endogenous proteins. Detection of poly-ubiquitination levels of HIF1α in sgNC or sgCLK3 cells. H, I Application of CHX to inhibit endogenous protein synthesis; IB assays to detect the remaining CLK3 protein levels over time; Student’s t test. J MG132 is used to inhibit the degradation of endogenous proteins. Immunoblotting assays to detect the K48-linked poly-ubiquitination level of HIF-1α using K48-linked ubiquitin–specific antibody in Vector and CLK3 cells. CLK3 positively regulates HIF1 signaling by stabilizing HIF1α. A GSEA revealing the relationship between CLK3 and hypoxia signaling using TCGA database (TCGA-COAD). B, C Detection of the expression levels of hypoxia signaling-related proteins in Vector, CLK3 WT, or CLK3 K186M mutant cells by immunoblotting. D Detection of hypoxia signaling-related proteins in sgNC or sgCLK3 cells using IB. E Detection of HIF1α expression in Vector or CLK3 cells under chloroquine (CQ) or MG132 treatment. F MG132 is used to inhibit the degradation of endogenous proteins. Detection of poly-ubiquitination levels of HIF1α in Vector, CLK3 WT, and CLK3 K186M cells. G MG132 is used to inhibit the degradation of endogenous proteins. Detection of poly-ubiquitination levels of HIF1α in sgNC or sgCLK3 cells. H, I Application of CHX to inhibit endogenous protein synthesis; IB assays to detect the remaining CLK3 protein levels over time; Student’s t test. J MG132 is used to inhibit the degradation of endogenous proteins. Immunoblotting assays to detect the K48-linked poly-ubiquitination level of HIF-1α using K48-linked ubiquitin–specific antibody in Vector and CLK3 cells. CLK3 phosphorylated the tyrosine residues of HIF1α After browsing GENEMANIA database, we found that there was potential binding between CLK3 and HIF1α (Supplementary Fig. 4A). Immunoblots further validated that CLK3 interacted with HIF1α, but not HIF1β, and HIF1α bound with CLK3, but not CLK1 and CLK2 (Fig. 5A and Supplementary Fig. 4B). Besides, His-pull down assays further proved the binding of CLK3 and HIF1α (Supplementary Fig. 4C). Also, CLK3 K186M mutant did not interacted with HIF1α (Supplementary Fig. 4D). CLK3 had serine/threonine and tyrosine kinase activity. We speculated that CLK3 might regulate the phosphorylation levels of HIF1α. Immunoprecipitation and immunoblot assays confirmed that CLK3 overexpression upregulated the tyrosine phosphorylation levels of HIF1α, while it did not have a significant impact on its serine and threonine phosphorylation levels (Fig. 5B and Supplementary Fig. 4E). The CLK3 K186M enzyme-deficient mutant was incapable of regulating the tyrosine phosphorylation levels of HIF1α (Fig. 5C). The knockout of CLK3 significantly inhibited the tyrosine phosphorylation levels of HIF1α protein (Fig. 5D). Consistently, CLK3 inhibitors also effectively suppressed the tyrosine phosphorylation levels of HIF1α protein (Supplementary Fig. 4F). Interestingly, we demonstrated that hypoxia could further increase the tyrosine phosphorylation levels of HIF1α protein mediated by CLK3 (Fig. 5E). In vitro kinase assay demonstrated that CLK3 directly phosphorylated HIF1α in vitro (Fig. 5F). Also, we revealed that hypoxia could enhance the binding affinity between HIF1α and CLK3 (Fig. 5G). Taken together, our results proved that CLK3 exhibited the ability to phosphorylate the tyrosine residues of HIF1α.Fig. 5The tyrosine residues of HIF1α are phosphorylated by CLK3.A Co-immunoprecipitation and immunoblotting assays to confirm the interaction between HIF1α and CLK3. B, C IP and IB assays to detect the tyrosine, serine/threonine phosphorylation level of HIF1α in Vector, CLK3 WT, or CLK3 K186M cells. D IP and IB assays to detect the tyrosine phosphorylation level of HIF1α in sgnc or sgCLK3 cells. E IP and IB assays to detect HIF1α phosphorylation in Vector or CLK3 WT cells under normoxia or hypoxia treatment. F In vitro kinase assay using Flag-CLK3 and His-HIF1α. G IP and IB assays to assess the interaction between HIF1α and CLK3 under normoxia or hypoxia treatment. The tyrosine residues of HIF1α are phosphorylated by CLK3. A Co-immunoprecipitation and immunoblotting assays to confirm the interaction between HIF1α and CLK3. B, C IP and IB assays to detect the tyrosine, serine/threonine phosphorylation level of HIF1α in Vector, CLK3 WT, or CLK3 K186M cells. D IP and IB assays to detect the tyrosine phosphorylation level of HIF1α in sgnc or sgCLK3 cells. E IP and IB assays to detect HIF1α phosphorylation in Vector or CLK3 WT cells under normoxia or hypoxia treatment. F In vitro kinase assay using Flag-CLK3 and His-HIF1α. G IP and IB assays to assess the interaction between HIF1α and CLK3 under normoxia or hypoxia treatment. CLK3-mediated tumor angiogenesis relied on HIF1α in vitro To further elucidate the relationship of the CLK3 signaling pathway and the HIF1 signaling pathway in regulating tumor angiogenesis, we constructed four kinds of colorectal cell lines stably expressing Vector + shnc, CLK3 + shnc, CLK3 + shHIF1α#1, and CLK3 + shHIF1α#2. Immunoblots confirmed the successful construction in hypoxic condition (Fig. 6A). ELISA assays were conducted to evaluate the VEGFA level in CMs and revealed that the silencing of HIF1α reversed the CLK3-mediated increase in VEGFA expression in hypoxic condition (Fig. 6B). CCK8 experiments indicated that the upregulation of HUVEC cell proliferation induced by CMs from CLK3 + shnc cells could be reversed by silencing HIF1α in hypoxic condition (Fig. 6C). Also, transwell assays revealed that the increased migration ability of HUVEC cells induced by CMs from CLK3 + shnc cells could be reversed by silencing HIF1α in hypoxic condition (Fig. 6D, E). Next, immunoblots proved CLK3 overexpression mediated the upregulation of VEGFA expression, which could be abolished by silencing HIF1α in normoxia (Fig. 6F, G). Similarly, the increased proliferation and migration abilities of HUVEC cells induced by CMs from CLK3 + shnc cells could be counteracted by silencing HIF1α in normoxia (Fig. 6H–J). Overall, our findings revealed that HIF1α was an essential factor for CLK3-mediated tumor angiogenesis.Fig. 6HIF1α is an essential factor for CLK3-mediated tumor angiogenesis.A Immunoblotting is performed to detect the expression levels of relevant proteins in colorectal cancer cells transduced with Vector + shnc, CLK3 + shnc, CLK3 + shHIF1α#1, or CLK3 + shHIF1α#2 under hypoxic conditions. B VEGFA levels are measured in the four groups of cells using ELISA under hypoxia (n = 3); Two-way ANOVA test. C HUVECs are cultured with different conditioned media, and CCK-8 assays are used to assess their proliferation under hypoxia (n = 3); Two-way ANOVA test. D Representative images of transwell migration assays under hypoxia; scale bar = 100 μm. E Statistical analysis corresponding to (D) (n = 3); Two-way ANOVA test. F IB is performed to detect the expression levels of relevant proteins in four groups of cells under normoxic conditions. G VEGFA expression was measured in the four groups using ELISA under normoxia (n = 3); Two-way ANOVA test. H HUVECs were cultured with different conditioned media, and CCK-8 assays were performed to assess proliferation under normoxia (n = 3); Two-way ANOVA test. I Representative images of transwell assays under normoxia; scale bar = 100 μm. J Statistical analysis corresponding to (I) (n = 3); Two-way ANOVA test. HIF1α is an essential factor for CLK3-mediated tumor angiogenesis. A Immunoblotting is performed to detect the expression levels of relevant proteins in colorectal cancer cells transduced with Vector + shnc, CLK3 + shnc, CLK3 + shHIF1α#1, or CLK3 + shHIF1α#2 under hypoxic conditions. B VEGFA levels are measured in the four groups of cells using ELISA under hypoxia (n = 3); Two-way ANOVA test. C HUVECs are cultured with different conditioned media, and CCK-8 assays are used to assess their proliferation under hypoxia (n = 3); Two-way ANOVA test. D Representative images of transwell migration assays under hypoxia; scale bar = 100 μm. E Statistical analysis corresponding to (D) (n = 3); Two-way ANOVA test. F IB is performed to detect the expression levels of relevant proteins in four groups of cells under normoxic conditions. G VEGFA expression was measured in the four groups using ELISA under normoxia (n = 3); Two-way ANOVA test. H HUVECs were cultured with different conditioned media, and CCK-8 assays were performed to assess proliferation under normoxia (n = 3); Two-way ANOVA test. I Representative images of transwell assays under normoxia; scale bar = 100 μm. J Statistical analysis corresponding to (I) (n = 3); Two-way ANOVA test. Depletion of CLK3 hindered tumor angiogenesis in vivo To further verify the biological function of CLK3 in vivo, we carried out animal experiments. We subcutaneously injected sgnc, sgCLK3#1, and sgCLK3#2 cell lines into the backs of mice and measured tumor size every three days. After 16 days, all mice were sacrificed, and tumors were isolated. Representative tumor image was shown (Fig. 7A). Deletion of CLK3 significantly inhibited the growth rate of tumor (Fig. 7B, C). Additionally, the survival curve indicated that knocking out CLK3 significantly prolonged the survival time of mice (Fig. 7D). IHC experiments showed that deletion of CLK3 markedly reduced the expression of HIF1α and VEGFA, and diminished the formation of intra-tumoral neo-vasculature in vivo (Fig. 7E, F). Immunoblot experiments similarly demonstrated that knocking out CLK3 inhibited the expression of HIF1α and VEGFA (Fig. 7G, H). In conclusion, these data demonstrated that the knockout of CLK3 significantly inhibited the intra-tumoral formation of neo-vasculature in vivo.Fig. 7Loss of CLK3 inhibits tumor angiogenesis in vivo.A Representative tumor images from different groups; scale bar = 1 cm. B Tumor growth curves (n = 5); One-way ANOVA test. C Statistical analysis of tumor weights. D Kaplan–Meier survival curves for the three groups of mice (n = 8). E Representative immunohistochemistry images; scale bar = 100 μm; Log-rank test. F Quantification of intra-tumoral micro-vessel density based on IHC staining (n = 15); One-way ANOVA test. G Immunoblotting is performed to detect the expression levels of related proteins in tumor tissues from each group. H Quantification of IB results shown in (G) (n = 4); One-way ANOVA test. I Schematic working model: hypoxia induces the binding of CLK3 to HIF1α, thereby inhibiting HIF1α degradation via the ubiquitin–proteasome pathway. Loss of CLK3 inhibits tumor angiogenesis in vivo. A Representative tumor images from different groups; scale bar = 1 cm. B Tumor growth curves (n = 5); One-way ANOVA test. C Statistical analysis of tumor weights. D Kaplan–Meier survival curves for the three groups of mice (n = 8). E Representative immunohistochemistry images; scale bar = 100 μm; Log-rank test. F Quantification of intra-tumoral micro-vessel density based on IHC staining (n = 15); One-way ANOVA test. G Immunoblotting is performed to detect the expression levels of related proteins in tumor tissues from each group. H Quantification of IB results shown in (G) (n = 4); One-way ANOVA test. I Schematic working model: hypoxia induces the binding of CLK3 to HIF1α, thereby inhibiting HIF1α degradation via the ubiquitin–proteasome pathway. Discussion Tumor angiogenesis, the formation of new blood vessels within a tumor mass, is a critical driver of cancer progression and metastasis [4]. Targeting this process has emerged as a promising strategy in cancer therapy. Multiple clinical trials have demonstrated that VEGFA monoclonal antibodies and VEGFR monoclonal antibodies can inhibit tumor angiogenesis [17–19]. These drugs have shown significant efficacy in the treatment of various malignant tumors. However, the emergence of treatment resistance underscores the existence of alternative angiogenic mechanisms that require further exploration. In this study, we investigated the potential role of CLK3 in tumor angiogenesis, using colorectal cancer as a model system. Through analysis of clinical datasets, we revealed a positive correlation between CLK3 expression and the activation of the tumor angiogenesis pathway. Immunoblotting and immunohistochemistry assays confirmed elevated CLK3 expression in colorectal tumor tissues compared to adjacent normal mucosa, further supporting its functional role in angiogenesis regulation. Additionally, elevated CLK3 expression levels correlated with advanced tumor stages and diminished patient survival, highlighting its clinical significance in colorectal cancer progression. In vitro experiments revealed that CLK3 positively regulated tumor angiogenesis. In vivo experiments demonstrated that knockout of CLK3 inhibited tumor growth and prolonged survival by suppressing intra-tumoral neovascularization. However, it was worth noting that not all tumors with high CLK3 expression showed a significant increase in tumor vascular density, as this was verified in our clinical specimens (Fig. 1). This suggested that CLK3 was not the only target gene regulating tumor angiogenesis, and patients with high CLK3 expression and high tumor vascular density might represent the optimal target population for CLK3-targeted interventions. HIF-1 plays a central role in promoting angiogenesis by upregulating the expression of pro-angiogenic factors such as VEGF, PDGF, and ANGPT2. Under normoxic conditions, HIF1α is hydroxylated by prolyl hydroxylase domain proteins (PHDs), marking it for ubiquitination and proteasomal degradation [20–22]. However, in hypoxic conditions, this process is inhibited, preventing HIF1α degradation. Consequently, stabilized HIF1α translocates into the nucleus, where it forms a complex with HIF1β. This heterodimer binds to hypoxia response elements (HREs) in the promoters of target genes, leading to the transcriptional activation of genes critical for the hypoxic response [2023–25]. Previous studies have established that HIF1α is regulated by various post-translational modifications such as methylation, phosphorylation, acetylation, and ubiquitination [202627]. Here, we uncovered the critical involvement of the HIF-1 signaling pathway in CLK3-mediated tumor angiogenesis. CLK3 was found to directly bind to HIF1α, stabilizing it by inhibiting its ubiquitin-proteasome-mediated degradation. Notably, we demonstrated the crucial role of CLK3’s protein kinase activity in regulating the ubiquitination of HIF1α, and further confirmed that CLK3 could phosphorylate tyrosine residues of HIF1α. However, the precise mechanism by which CLK3-mediated phosphorylation influenced the ubiquitination of HIF1α remain to be fully elucidated. These findings suggested that CLK3 regulated the HIF-1 signaling pathway by enhancing HIF1α expression and activity, thereby promoting tumor angiogenesis. Of particular importance, tumor angiogenesis inhibitors are often ineffective due to the hypoxic tumor microenvironment that activates cancer stem cells (CSCs) [28]. In low-oxygen conditions, CSCs upregulate HIFs to promote survival and stemness, which enhances their resistance to therapies and leads to tumor recurrence. The cycle of hypoxia and angiogenesis further complicates therapeutic strategies by creating niches that favor CSCs [1529]. To improve outcomes, it is crucial to address the interplay between hypoxia and CSC behavior; combining angiogenesis inhibitors with agents targeting CSCs or the hypoxic microenvironment may enhance treatment efficacy. Furthermore, the identified CLK3/HIF1α/VEGFA axis provided potential insights into the limitations of current anti-angiogenic therapies. While we currently lacked direct clinical data correlating CLK3 levels with patient responses to anti-VEGF agents (such as Bevacizumab), its role in stabilizing HIF1α suggested that high CLK3 expression might contribute to a diminished response to VEGF-targeted treatments. Addressing this in future prospective clinical cohorts or patient-derived models would be essential to determine if CLK3 ccould serve as a predictive biomarker for anti-angiogenic therapy outcomes. In summary, our study provided compelling evidence that CLK3 positively regulated tumor angiogenesis in colorectal cancer by modulating the HIF-1 axis (Fig. 7I). These findings not only elucidated the mechanistic underpinnings of CLK3-mediated tumor angiogenesis but also positioned CLK3 as a potential therapeutic target for colorectal cancer treatment. Further research into CLK3 inhibition might hold significant promise for improving cancer therapy outcomes. Supplementary information Suppelmentary figure legend Figure S1 Figure S2 Figure S3 Figure S4 Suppelmentary figure legend Figure S1 Figure S2 Figure S3 Figure S4