SMYD3 synergises with RACK1 to promote colorectal cancer lung metastasis by recruiting SMAD3
Xiaoming Bai, Dong Han, Jie Chen, Siqi Sheng, Haimei Feng, Hongyu Wang, Ke Xu, Yadi Huang, Mengxi Huang, Xiaoyuan Chu, Yitian Chen, Zengjie Lei
TL;DR
This study identifies a protein complex involving SMYD3 and RACK1 that promotes colorectal cancer metastasis to the lungs, suggesting a new therapeutic target.
Contribution
The study reveals a novel tripartite interaction between SMYD3, RACK1, and SMAD3 that drives CRC metastasis.
Findings
SMYD3 overexpression correlates with poor CRC prognosis and increased metastasis.
RACK1 mediates SMYD3-SMAD3 complex formation to activate TSKU transcription.
Disrupting the SMYD3-RACK1-SMAD3 axis may inhibit CRC metastasis.
Abstract
Cancer metastasis is the leading cause of mortality associated with cancer, and the prognosis for patients diagnosed with colorectal cancer(CRC) largely depends on the occurrence of metastasis during the progression of the disease. A comprehensive understanding of the mechanisms underlying metastasis in CRC is essential for advancing treatment strategies. Through integrated bioinformatics analysis of mRNA expression profiles and epigenetic modifiers, we identified SMYD3 as the top differentially expressed histone modifier in CRC. Clinically, SMYD3 overexpression significantly associates with poor prognosis and enhances metastatic potential. Utilizing immunoprecipitation-mass spectrometry, we discovered RACK1 as a novel SMYD3-interacting protein. Subsequent mechanistic studies revealed a tripartite interaction network: SMYD3 recruits SMAD3 through RACK1-mediated scaffolding, facilitating…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7- —https://doi.org/10.13039/501100001809National Natural Science Foundation of China
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsChromatin Remodeling and Cancer · Epigenetics and DNA Methylation · Mechanisms of cancer metastasis
Introduction
Colorectal cancer (CRC) is the third most common malignancy and the second leading cause of cancer-related mortality globally [1]. Current first-line treatments have significantly improved the therapeutic efficacy for most advanced colorectal cancer cases.Nevertheless, the prognosis for CRC patients remains unsatisfactory due to the high occurrence of drug resistance and metastasis [2]. Therefore, it is essential to explore more effective therapeutic strategies. Epigenetic aberrations are considered hallmarks of cancer, with common changes including histone modification, abnormal DNA methylation, and dysregulated expression of non-coding RNAs [3]. Notably, epigenetic alterations associated with histone modification play a critical role in tumorigenesis and progression [4, 5]. The aberrant expression of enzymes that regulate histone modifications, such as histone methyltransferases (HMTs), histone demethyltransferases (HDMTs), and histone acetyltransferases (HATs), along with the consequent changes in epigenetic processes, are significant in the occurrence and development of various cancers [6]. Therefore, investigating the role of histone modification enzymes in the onset and progression of tumors is of paramount importance.
The SET and MYND domain (SMYD) family constitutes one of the major families of histone lysine methyltransferases (KMTs), comprising five members (SMYD1-5) [7, 8]. In colorectal cancer, SMYD3 interacts with H3K4me3 and is associated with poor prognosis, acting as an oncogene that regulates cell proliferation and metastatic spread [9–11]. The SMYD3 inhibitor BCI-121 has demonstrated efficacy in inhibiting tumor growth in various tumor types [12–14]. However, the comprehensive epigenetic regulatory impact of SMYD3 in colorectal cancer remains unclear.
RACK1 (Receptor for Activated C Kinase 1) is an intracellular scaffolding protein characterized by seven WD40 repeats, which enables it to bind a variety of proteins through these motifs [15]. RACK1 has been reported to serve as a scaffold for protein–protein interactions [16]. This scaffolding protein interacts with and regulates the activation and/or stability of multiple transcription factors [17]. For instance, RACK1 binds to and activates STAT1 and STAT3, respectively [18, 19]. As a scaffolding protein, RACK1 is involved in a variety of signaling pathways and is associated with cell growth, differentiation, movement, and migration [20, 21]. Numerous studies have demonstrated that RACK1 enhances the invasive and migratory capabilities of tumor cells in certain cancer environments [22–25]. In colorectal cancer, studies have shown that RACK1 also promotes the invasion and metastasis of tumor cells [26, 27]. However, the underlying molecular mechanisms require further investigation.
SMAD3 is a member of the SMAD protein family and is involved in the TGF-β/Smads signaling pathway [28]. The TGF-β1/Smads pathway is crucial in regulating cell growth, invasion, metastasis, and apoptosis. SMAD3 has been shown to promote cancer cell metastasis across various tumor types, and the JICD1/SMAD3 transcriptional complex has been implicated in promoting glioblastoma metastasis [29]. Additionally, POH1 (also known as Rpn11, Regulatory particle subunit number 11) enhances lung cancer cell proliferation, migration, and invasion by stabilizing SMAD3 [30]. The TGF-β1/Smad3 pathway promotes colorectal cancer metastasis by inducing the overexpression of FSTL1, which facilitates CRC cell migration and invasion through its binding to VIM [31]. However, further research is necessary to elucidate the role of SMAD3 in promoting invasive metastasis in colorectal cancer. Prompted by existing literature suggesting a potential interaction between SMYD3 and SMAD3 [32, 33], we experimentally validated this interaction in colorectal cancer (CRC) cell lines. Our results confirmed the physical interaction between these two proteins, consistent with previous reports. To further elucidate the functional consequences of this interaction, we performed comprehensive genomic analyses, including RNA-seq, Cut&Tag, and ChIP-seq. These integrated approaches enabled the identification of co-regulated downstream targets of the SMYD3-SMAD3 complex in CRC. Our findings provide novel insights into the molecular mechanisms underlying colorectal cancer progression and offer potential therapeutic targets for further investigation.
Tsukushi (TSKU) is a secreted protein initially identified as a regulator of embryonic development and metabolic homeostasis [34]. Mechanistically, TSKU acts by regulating key signalling pathways: TSKU inhibits Wnt signalling by binding to Frizzled 4 [35]. In lung cancer, TSKU promotes epithelial-mesenchymal transition (EMT), leading to upregulation of Snail and downregulation of E-cadherin, and promotes lung cancer cell proliferation [36]. Our study shows that TSKU is highly expressed in colorectal cancer and is associated with poor prognosis. Our integrated clinical and bioinformatic evidence positions TSKU as a promising dual-function biomarker for both prognostic stratification and therapeutic targeting in CRC.
In this study, we investigated the role of SMYD3 in colorectal cancer metastasis. Bioinformatics analyses revealed that SMYD3 is highly expressed in CRC, a finding confirmed by subsequent experiments. Furthermore, SMYD3 promotes CRC cell metastasis both in vitro and in vivo models. Mechanistically, co-immunoprecipitation assays demonstrated the interaction between SMYD3, RACK1, and SMAD3. RACK1 recruits SMAD3 to SMYD3, promoting the transcriptional activation of a novel SMYD3-SMAD3 downstream gene, TSKU. In summary, the SMYD3-RACK1-SMAD3 transcriptional complex identified in this study serves as a promising therapeutic target for colorectal cancer.
Methods
Cell culture and reagents
NCM460, HCT-8, HCT-15, HCT-116, SW480, RKO, Lovo, and SW620 cell lines were procured from the cell bank of the Shanghai Institute of Biochemistry and Cell Biology (Shanghai, China). NCM460, HCT-8, HCT-15, and Lovo were cultured in RPMI 1640 medium (Gibco, USA) supplemented with 10% fetal bovine serum (Excell, USA) and 1% Penicillin–Streptomycin. HCT-116 cells were maintained in McCoy's 5 A medium (Gibco, USA) with 10% fetal bovine serum (Excell, USA) and 1% Penicillin–Streptomycin. RKO cells were cultured in MEM medium (Gibco, USA) supplemented with 10% fetal bovine serum (Excell, USA) and 1% Penicillin–Streptomycin. All cell lines were incubated at 37 °C in a 5% CO2 atmosphere.
Patients and samples
All human colorectal tissues utilized in this study were collected immediately following surgical resection at Jinling Hospital, Nanjing University. Informed consent was obtained from all patients participating in this study, and the investigation received approval from the Ethical Review Committee of Jinling Hospital, Nanjing University. Paraffin-embedded colorectal cancer specimens and adjacent normal tissues were utilized for immunohistochemical (IHC) staining.
Plasmids construction, transfection
Overexpression plasmids for SMAD3, SMYD3, and RACK1 were constructed by cloning the respective full-length coding sequences (CDS) from NCBI (accession numbers: SMAD3, NM_005902; SMYD3, NM_001167740; RACK1, NM_006098) into a PLVX vector. Each construct was engineered to express a C-terminal epitope tag (6 × His, 3 × Flag, or Myc, respectively) to facilitate the detection of protein expression in cells, protein purification, and viral packaging. To generate these constructs, specific primers were designed to amplify the target gene fragments, which were subsequently digested with restriction enzymes and ligated into the PLVX vector. The recombinant plasmids were then transformed into E. coli competent cells for cloning and amplification. Following blue-white screening, individual colonies were selected and verified by sequencing. The correct clones were used to prepare the final overexpression vectors. All plasmid construction procedures were assisted by Shanghai Ji Kai Gene Medical Technology Co., Ltd.
Lentivirus infection
Cells were cultured with the supernatant containing lentivirus to transfection. Following transduction, stable cell lines were established through selection with puromycin, neomycin, or hygromycin. Short hairpin RNA (shRNA) was synthesised to specifically target the RACK1 and SMAD3 genes, and the above plasmids were produced by Shanghai Jikai Genomics Technology Co. The interfering sequences are listed in Table S6.
Western blot analysis
RIPA lysis buffer (KeyGEN Bio TECH, China, #KGB5204-100), supplemented with protease and phosphatase inhibitors (NCM Biotech, China, #P002), was applied to colorectal cancer (CRC) cells or tissues, and the insoluble precipitate was removed through shaking and centrifugation. The protein samples were subsequently separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a 0.2 µm polyvinylidene difluoride (PVDF) membrane (Millipore, #ISEQ00010, Ireland). Following sealing with 5% skimmed milk, the PVDF membrane was incubated overnight at 4 °C with primary antibodies: anti-SMYD3 (CST, #12,859, 1:1000), anti-RACK1 (CST, #5432, 1:1000), anti-SMAD3 (Abclonal, #A16913, 1:1000), anti-TSKU (Proteintech, #12,370–1-AP, 1:1000), anti-Flag (CST, #14793S), anti-His (Proteintech, #66,005–1-Ig, 1:1000), anti-p-SMAD3 (CST, #9520, 1:1000), anti-H3K4me3 (CST, #9751, 1:1000), anti-GAPDH (Proteintech, #10,494–1-AP, 1:1000), and anti-tubulin (Proteintech, #66,031–1-Ig, 1:1000). The following day, the blots were incubated with secondary antibodies for 1 h at 37 °C. After thorough washing, immunoreactive bands were visualized using an enhanced chemiluminescence (ECL) detection system.
Quantitative reverse transcription PCR (qRT-PCR)
Total RNA was isolated from colon cancer cell lines using TRIzol reagent (Invitrogen) and subsequently treated with DNase to eliminate any contaminating DNA. The RNA was reverse transcribed into complementary DNA (cDNA) using reverse transcription reagents (Accurate Biology, #AG11706), and the relative mRNA levels were quantified using a SYBR Green Pro Taq HS Premixed qPCR kit (Accurate Biology, #AG11702). The fold changes of the target genes were normalized to β-actin. Primer sequences for quantitative PCR (qPCR) are provided in Table S1. All experiments were performed in triplicate to ensure statistical validity.
Immunohistochemistry
Tissue sections were deparaffinized in xylene and subjected to a series of graded ethanol concentrations. Following treatment with EDTA antigen retrieval solution and goat serum, the slides were incubated overnight at 4 °C with primary antibodies, including SMYD3 (Abcam, #ab187149, 1:100), SMAD3 (Abclonal, #A19115, 1:200), RACK1 (CST, #5432, 1:200), and TSKU (Proteintech, #12,370–1-AP, 1:100). Subsequently, the tissue sections were incubated with the appropriate secondary antibodies and treated with DAB, followed by hematoxylin staining for subsequent microscopic observation.
Immunohistochemical Scoring: Four high-magnification fields of view were randomly selected and assessed in a double-blind manner by two experienced pathologists, based on the percentage of positive cells and the intensity of positive cell staining. The intensity of staining was categorized into four grades: 0 for no staining, 1 for weak staining, 2 for moderate staining, and 3 for strong staining. Concurrently, the percentage of positive cells was classified into five grades: 0 for no stained cells, 1 for < 25%, 2 for 26%−50%, 3 for 51%−74%, and 4 for ≥ 75%. The immunohistochemical score for each sample was calculated using the formula: score = intensity of staining × proportion of positive cells stained.
Wound healing assay and transwell assay
For the wound healing assay, 2 × 10^6 cells were inoculated into a 6-well plate until complete confluence was achieved. Subsequently, the cells were scored using a 20 µL pipette tip. After a 48-h incubation period, photographs were captured under a microscope and plotted on a graph.
For the transwell assay, 1 × 10^5 cells mixed with 200 µL of serum-free medium were placed in the upper chamber (Corning, New York, USA, #3422), while the lower chamber was supplemented with complete medium to assess cell migration. After 24 h, the chambers were washed twice with PBS, and the cells on the upper membrane were carefully removed. The chambers were then fixed in 4% paraformaldehyde for 20 min, washed twice with PBS, and subsequently immersed in 1% gentian violet solution for 30 min for staining. The invaded cells were observed in three randomly selected fields using an inverted microscope.
Animal experiments
In the lung metastasis model, 2 × 10^6 colorectal cancer (CRC) cells (HCT116 mock-Luc, HCT116-LVSMYD3-Luc, HCT116-shRACK1-Luc) were injected into the tail vein of BALB/c nude mice (male, 5 weeks old). Five weeks post-injection, in vivo imaging of the mice was conducted using the IVIS Spectrum imaging system, which is based on luciferase labeling. Six weeks after injection, the mice were euthanized in accordance with ethical guidelines, and the lungs were subjected to hematoxylin and eosin (HE) staining. For the administration of the SMYD3 inhibitor BCI-121, nude mice injected with HCT116 mock and HCT116-shRACK1 cells were selected after tail vein injection and treated with BCI-121 (Selleck, #432,529–82-3) at a dose of 5 mg/kg via peritoneal injection (i.p.) three times per week for five weeks. All animal experiments were approved by the Institutional Animal Care and Use Committee of Jinlin Hospital, Nanjing University.
Coimmunoprecipitation
Lysis buffer (50 mM Tris base, 150 mM NaCl, 2 mM EDTA, 1% NP40), supplemented with protease and phosphatase inhibitors (NCM Biotech, China, #P002), was added to CRC cells, and protein samples were obtained by centrifugation to remove insoluble precipitates. Protein A/G magnetic beads were incubated with anti-Flag (CST, #14793S), anti-SMAD3 antibody (CST, #9523S, 1:100), anti-RACK1 antibody (CST, #5432S, 1:100), anti-His antibody (Proteintech, #66,005–1-Ig, 1:1000), or IgG (Beyotime, #A7058) for 4 h at 4 °C on a rotary shaker. The immune complex solutions were mixed with the protein samples and incubated overnight at room temperature, followed by washing with washing buffer (50 mM Tris base, 150 mM NaCl, 2 mM EDTA, 0.1% NP40). Bound immune complexes were separated from the beads using lysis buffer and utilized for immunoblot analysis. For secondary antibody incubation, 5% skimmed milk was added with the appropriate secondary antibody (Abbkine, #A25012,#A25022,#A25222, 1:5000), which avoids interference between light and heavy chains.
Immunofluorescence staining
An appropriate number of cells was inoculated into confocal petri dishes. After 24 h, the cells were washed three times with phosphate-buffered saline (PBS) and fixed in 4% paraformaldehyde for 15 min. The cells were then washed three times with PBS, permeabilized with 0.5% Triton X-100 for 20 min at room temperature, and subsequently washed three times with PBS. Blocking was performed using goat serum for 30 min at room temperature, followed by incubation with primary antibodies (SMYD3,Abcam, #ab187149, 1:100; RACK1,CST, #5432S, 1:100; SMAD3 (Abclon, #A19115; TSKU (Proteintech, #12,370–1-AP) at 4 °C overnight. After three washes with PBS, diluted fluorescent secondary antibodies (Cy3 Proteintech, #SA00009-1 and Alexa Fluor® 488,Jackson ImmunoResearch Laboratories, Inc, #111–545-144) were added, and the cells were incubated for 1 h at room temperature in the dark. Subsequently, the cells were incubated with DAPI (Beyotime, #C1006) for 5 min at room temperature in the dark to stain the nuclei. Excess DAPI was removed, and the coverslips were sealed with an anti-fluorescence quencher (Beyotime, #P0126). The acquired images were observed under a fluorescence microscope.
CUT&Tag
The CUT&Tag assay was performed using the ActiveMotif CUT&Tag-IT® Assay Kit as follows: Fresh cells were collected in Eppendorf (EP) tubes at room temperature and counted. Centrifugation was performed several times to remove the solution. Cell Perforation Buffer and Binding Buffer resuspended ConA beads were added to the tubes, and the cells were gently inverted and mixed for 5 to 10 min. Antibody buffer and 0.5 to 1 mL of primary antibody were added and gently shaken; the EP tube was then incubated on a shaker at room temperature for 2 h. The secondary antibody was diluted 1:100 in Perforation Buffer, added to the tube, and gently shaken; the EP tube was incubated on a shaker at room temperature for 30 to 60 min. Following this, 0.8 to 1 mL of Perforation Buffer was added, and the tube was inverted up and down 10 times. ChiTag Transposome was diluted 1:250 in Perforation Buffer 2 and added dropwise to the cells; the mixture was incubated for 1 h at room temperature on a shaker, followed by the addition of 1 mL of Perforation Buffer 2, with the tube inverted up and down 10 times. The Tn5 transposase was activated by the addition of a magnesium-containing reaction solution, which cleaved the DNA region bound by the target protein while simultaneously ligating the carry-over library junctions to the DNA fragments. DNA was extracted, and PCR amplification was employed to construct the library. High-throughput sequencing was contracted to Shanghai Jiayin Biotechnology Co.
Chromatin immunoprecipitation (ChIP)-qPCR
ChIP was performed using the ChIP-IT Express Enzymatic Kits (Active Motif) according to the manufacturer's instructions. Cells were fixed with 1% formaldehyde for 10 min, and the reaction was terminated with glycine. The fixed cells were then placed in pre-cooled 1 × PBS. Chromatin fragments were prepared through enzymatic digestion, and the reaction was halted with EDTA. A small aliquot of chromatin was collected to assess breakage and measure concentration. The sheared chromatin was pre-conjugated to Protein G magnetic beads, and after 2 h, each mixture of chromatin and Protein G magnetic beads was incubated with 4 µg of anti-Flag antibody (CST, #14793S), anti-H3K4me3 antibody (CST, #9751S), anti-His antibody (Proteintech, #66,005–1-Ig), or normal rabbit IgG (Active Motif) overnight at 4 °C. The Protein G beads with antibody-bound DNA complexes were collected following centrifugation. Quantitative PCR (qPCR) was performed after elution, uncrosslinking, and treatment of DNA with proteinase K to detect DNA enriched by ChIP. The TSKU promoter ChIP primers were as follows: F: AGGTCTTGCGGGAGACAAAG, R: CCCACAAACGTGAGATGCAC.
Bioinformatics
For the analysis of differentially expressed genes regulating colorectal cancer development, we utilized the GEO database (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi) to obtain the raw CEL files of colorectal cancer mRNA expression profiling microarrays from three data platforms: GPL570, GPL5175, and GPL96. This included ten datasets: GSE33113, GSE18105, GSE20916, GSE21510, GSE32323, GSE4107, GSE24551, GSE29638, GSE41258, and GSE49355, encompassing a total of 1,215 colorectal cancer genes. Specifically, GSE41258 and GSE49355 provided gene expression profiling microarrays for 1,215 colorectal and paraneoplastic tissue samples, which included 1,013 tumor cases and 202 paraneoplastic cases.
For the reading and pre-processing of the raw data, the affy package of R software (http://www.bioconductor.org) was utilized to read the raw CEL files. The annotate package was employed to annotate the gene chips, allowing for the identification of genes common to the three platforms. Additionally, the sva package was used to mitigate the batch effect across different platforms, resulting in a total of 1,215 cases. Ultimately, the expression levels of 10,135 genes across these cases were obtained.
For the differential analysis of mRNA gene expression, the limma package was employed to conduct the analysis. The differential gene screening criteria were set as a fold change (tumor tissue/normal tissue) greater than 1.5 (or less than 2/3) with a P-value of less than 0.05. The results of the differential analysis were visualized using a volcano plot created with the ggplot2 package, while the heatmap package was utilized to depict the top 100 most significantly upregulated or downregulated genes in tumor tissues.
For the SMAD3 ChIP-seq analysis, the relevant data were downloaded from the database (https://guolab.wchscu.cn/hTFtarget/).
To compare the expression levels of TSKU in colorectal cancer versus paraneoplastic cancer, we retrieved the expression data for TSKU from the TNM plot database (https://tnmplot.com/analysis/).
For gene expression level correlations, we employed the TNM plot database (https://tnmplot.com/analysis/) to perform correlation analyses between SMAD3-TSKU and SMYD3-TSKU.
Statistical analysis
Statistical analysis was conducted using GraphPad Prism 9 software. The log-rank test was employed for survival analysis. Normally distributed data were expressed as the mean ± standard deviation and analyzed using Student's t-test. Correlation analysis was performed using Pearson's correlation coefficient. One-way ANOVA was utilized for comparisons among more than two patient groups. A p-value of less than 0.05 was considered indicative of a statistically significant difference.
Data availability
All data generated during this study are available from the corresponding author on reasonable request.
Results
SMYD3 is highly expressed in colorectal cancer tissues and correlates with poor prognosis
To identify key genes in colorectal cancer (CRC) development, we performed a differential analysis of 1,215 samples (1,013 tumors and 202 matched adjacent normal tissues) from three GEO platforms (GPL570, GPL5175, GPL96). By cross-referencing the differentially expressed genes with a curated list of 154 epigenetic regulators, we identified 22 dysregulated histone-modifying enzymes, with the top three being EZH2, SMYD3 and HDAC2 (Fig. 1A-B). Notably, SMYD3 has not been extensively studied in advanced colorectal cancer. Examination of eight pairs of clinical tissue samples revealed high expression of the histone methyltransferase SMYD3 in colorectal cancer tissues compared to para-carcinoma tissues (Fig. 1C-D). Additionally, data from the GEO database indicated that SMYD3 is highly expressed in colorectal cancer (Fig. S1A). Analysis of the TCGA database further confirmed elevated RNA expression levels of SMYD3 in colorectal cancer tissues (Fig. S1B-C). Within the TCGA cohort, significantly increased SMYD3 expression was observed in patients with advanced-stage colorectal cancer pathology (Fig. S1D).Fig. 1SMYD3 is highly expressed in colorectal cancer tissues and correlates with poor prognosis. A Colorectal cancer mRNA expression profiling microarray analysis of differentially expressed genes with 154 common epimerase modifying enzymes taking the intersection Venn plot. B Heatmap of the expression profiles of 22 differentially expressed epimerase modifiers. C-D Western blot analysis of SMYD3 expression in colorectal cancer tissues (T) and corresponding paracancerous tissue (P) (n = 8). E mRNA expression of SMYD3 in colorectal cancer tissues (T) and corresponding paracancerous tissue (P) (n = 8). F Representative immunohistochemical (IHC) staining of SMYD3 expression in colorectal tumors and corresponding paracancerous tissue.(n = 96). G Immunohistochemical (IHC) scoring of SMYD3 expression in colorectal tumors and corresponding paracancerous tissue (n = 96). H Kaplan–Meier curves of SMYD3 expression levels in tumour and normal tissues and patient prognosis(high: n = 62, low: n = 34). For D and F, significance was determined with the student unpaired t test.For G, significance was determined with Log–rank (Mantel–Cox) test. ns, not significant, p > 0.05; , p < 0.05; , p < 0.01;, p < 0.001; ****, p < 0.0001. Errorbars, ± SD
Furthermore, immunohistochemical analysis of 96 clinical colorectal cancer tissue samples demonstrated widespread expression of SMYD3 in both the cytoplasm and nucleus of cancer cells, with reduced expression in para-carcinoma tissues. When comparing moderately differentiated tumor tissues to poorly differentiated ones, SMYD3 expression was significantly higher in the latter (Fig. 1E-F). Based on immunohistochemical scoring criteria, colorectal cancer patients were classified into high and low SMYD3 expression groups. We investigated the correlation between SMYD3 expression and clinicopathological features in 96 pairs of colorectal cancer (CRC) tissue samples. Our analysis revealed that elevated levels of SMYD3 expression in CRC patients were significantly associated with poor differentiation (P = 0.0019) and advanced TNM staging (P < 0.0001) (Table 1). Moreover, survival analysis using Kaplan–Meier curves indicated that patients with high SMYD3 expression experienced shorter overall survival (OS), disease-free survival (DFS), and metastasis-free survival (MFS) compared to those with low SMYD3 expression, with statistically significant differences (Fig. 1G). Additionally, data from the TCGA database suggested that high expression of SMYD3 is associated with poor prognosis (Fig. S1E). Thus, SMYD3 plays a promoting role in the development of colorectal cancer and is correlated with a worse prognosis for colorectal cancer patients.Table 1. Correlation between SMYD3 expression and clinicopathological parameters in CRC (n = 96)/n (%)ParametersCategoryNoSMYD3 expressionPLow(%)High(%)Total Age9634620.5979< 605320(58.8)33(53.2)≥ 604314(41.2)29(46.8)Gender0.1249Male5222(64.7)30(48.4)Female4412(35.3)32(51.6)Tumor size(cm)0.7408< 54316(47.1)27(43.5)≥ 55318(52.9)35(56.5)T stage0.5395T1/T2229(26.5)13(21.0)T3/T47425(73.5)49(79.0)TNM stage< 0.0001I/II4125(73.5)16(25.8)III/IV559(26.5)46(74.2)Histological Grade0.0019Well-Moderately6630(88.2)36(58.1)Poorly304(11.8)26(41.9)
SMYD3 promotes colorectal cancer cell metastasis
We assessed the protein expression levels of SMYD3 in various colon cancer cell lines and immortalized colorectal epithelial cells (NCM460). Compared with normal colonic epithelial cells, SMYD3 is highly expressed in most colon cancer cells, with the exception of the RKO cell line (Figure S2A-B). Colon cancer cell lines with relatively low SMYD3 expression (HCT116, HCT15) were selected to establish stable cell lines overexpressing SMYD3, with verification conducted at both the protein and mRNA levels (Fig. S2C-D). Wound healing assays and transwell assays demonstrated that the overexpression of SMYD3 enhanced the migratory capabilities of tumor cells (Fig. 2A-C and Fig. S2E). Colon cancer cell lines with relatively high SMYD3 expression (Lovo, HCT8) were chosen to establish knockdown cell lines, with verification conducted at the protein and mRNA levels following SMYD3 knockdown (Fig. S2F-G). Wound healing assays and transwell assays confirmed that SMYD3 knockdown inhibited the migratory abilities of tumor cells (Fig. 2D-E and Fig. S2H). Collectively, these in vitro experiments indicate that SMYD3 promotes the metastatic potential of colon cancer cells.Fig. 2SMYD3 promotes colorectal cancer cell metastasis. A Wound healing assays evaluating migration in CRC transfected with control or SMYD3 overexpression lentivirus and CRC mock cells, assessing migration. Scale bar = 50 μm. B-C Transwell experiment showing the migration capabilities of CRC cells transfected with control or SMYD3 overexpression lentivirus and CRC mock cells. Scale bar = 50 μm. D Wound-healing assays evaluating migration in CRC infected with control or SMYD3 knockdown lentivirus and CRC mock cells, assessing migration. Scale bar = 50 μm. E Transwell experiment showing the migration capabilities of CRC cells infected with control or SMYD3 knockdown lentivirus and CRC mock cells. Scale bar = 50 μm. F Statistical analysis of luciferase bioluminescence intensity (n = 5). G Representative living images of mice injected with HCT116 infected by indicated lentivirus into tail vein. H Representative images of metastases in murine lung of each group; n = 5/group. The black arrow indicated the metastasis. I Statistical analysis of the number of pulmonary metastases of each group. J H&E staining of pulmonary tissue sections. For A,C,E,F and I, significance was determined with the student unpaired t test. ns, not significant, p > 0.05; *, p < 0.05; ***, p < 0.001; ****, p < 0.0001. Errorbars, ± SD
To investigate the role of SMYD3 in colon cancer in vivo, we first established a colon cancer cell line, HCT116, that stably overexpresses SMYD3, which has been labeled with luciferase (Fig. S2I). We developed a tail vein to lung metastasis model. HCT116 cells and HCT116-LV-SMYD3 cells were injected into mice via the tail vein, and real-time imaging of the mice was performed five weeks after the injection. A significant increase in fluorescence intensity of lung metastases was detected in the SMYD3 overexpression group compared to the control group (Fig. 2F-G). Significantly more lung metastases were observed in the SMYD3 overexpression group than in the control group (Fig. 2H-I). Hematoxylin and eosin (H&E) staining further indicated that the number, size, and degree of malignancy of lung metastases were significantly elevated in the SMYD3 overexpression group (Fig. 2J).
SMYD3, RACK1, and SMAD3 interact with one another, with the interaction between SMYD3 and SMAD3 being contingent upon RACK1
To gain further insight into the mechanism by which SMYD3 exerts its pro-cancer biological functions, we conducted immunoprecipitation and silver staining analyses in HCT116 cells stably overexpressing Flag-SMYD3 using anti-Flag antibodies (Fig. 3A). Additionally, we performed immunoprecipitation and mass spectrometry analyses using anti-Flag in both HCT116 and HCT15 cells stably overexpressing Flag-SMYD3 (Table S2). The intersection of these two results revealed that the scaffolding protein RACK1 exhibited higher scores (Fig. 3B-C). To further elucidate the interaction between SMYD3 and RACK1, we overexpressed Flag-SMYD3 in colon cancer cells (HCT116 and RKO) and demonstrated the interaction between SMYD3 and RACK1 following immunoprecipitation with anti-Flag and anti-RACK1 antibodies (Fig. 3D and Fig. S3A). Furthermore, we overexpressed Flag-SMYD3 in 293 T cells, confirming the interaction with RACK1 through immunoprecipitation using anti-Flag and anti-RACK1 antibodies (Fig. S3B).Fig. 3SMYD3, RACK1 and SMAD3 interact with each other and the interaction between SMYD3 and SMAD3 depends on RACK1. A Silver staining results of proteins pulled down by Flag-SMYD3 in HCT116 cells. B The intersection veen plot of the immunoprecipitation and mass spectrometry analysis in colon cancer cell lines HCT15 and HCT116. C List of the candidate target proteins identified by the mass spectrometry. D Coimmunoprecipitation (Co-IP)-Western blot (WB) analysis of the SMYD3-RACK1 interaction in HCT116 cells transfected with Flag-tagged SMYD3 (Flag-SMYD3). E Co-IP-WB analysis of the SMYD3-RACK1-SMAD3 interaction in HCT116 cells transfected with Flag-tagged SMYD3 (Flag-SMYD3). F Immunoprecipitation experiments were performed to detect the endogenous interaction of SMYD3,SMAD3, RACK1 in HCT116 cells. G Immunofluorescence assay of endogenous SMYD3 and RACK1 in HCT116 cells. DAPI was used to counterstain the nucleus. H Immunofluorescence assay of endogenous SMYD3 and SMAD3 in HCT116 cells. DAPI was used to counterstain the nucleus. I Immunofluorescence assay of endogenous RACK1 and SMAD3 in HCT116 cells. DAPI was used to counterstain the nucleus. J Effects of RACK1 knockdown on SMYD3 interacted with SMAD3 in HCT116-LV-Flag-SMYD3 and RKO-LV-Flag-SMYD3 cells. K Effects of RACK1 knockdown on SMYD3 interacted with SMAD3 in HCT116-LV-His-SMAD3 and RKO-LV-His-SMAD3 cells
RACK1 serves as a scaffold for protein–protein interactions. As a scaffolding protein, RACK1 is involved in various signaling pathways and has been implicated in cell growth, differentiation, motility, and migration. Existing literature indicates that the scaffold protein RACK1 interacts with the transcription factor SMAD3 [37, 38]. There is also evidence that SMYD3 has an interaction with SMAD3 [32, 33]. Therefore, we postulated a potential interaction among the histone methyltransferase SMYD3, the scaffold protein RACK1, and the transcription factor SMAD3. The interaction between Flag-SMYD3, RACK1, and SMAD3 was demonstrated in colon cancer cells (HCT116, RKO, 293 T) through immunoprecipitation using anti-Flag antibodies (Fig. 3E and Fig. S3C-D). Furthermore, we confirmed the endogenous interaction of SMYD3, RACK1, and SMAD3 following immunoprecipitation with SMAD3 and RACK1 in colon cancer cells (HCT116 and RKO) (Fig. 3F and Fig. S3E). Immunofluorescence co-localization experiments in the colon cancer cell line HCT116 demonstrated that endogenous RACK1, SMAD3, and SMYD3 were localized in the nuclear plasma region of the cells, indicating that all three proteins were present in the same area (Fig. 3G-I). Collectively, these experiments confirm the interactions between SMYD3, RACK1, and SMAD3.
The effects of RACK1 on its interacting proteins can be categorized into four types: altering the activity of interacting proteins, regulating intermolecular interactions, relocating its interacting proteins, and strengthening or weakening intermolecular interactions. Therefore, we hypothesized that the interaction between SMYD3 and SMAD3 is dependent on RACK1. To verify this hypothesis, we first assessed the protein expression levels of RACK1 and SMAD3 in various colon cancer cells and immortalized colorectal epithelial cells (NCM460) (Fig. S3F). We generated HCT116 and RKO cell lines stably overexpressing Flag-SMYD3 or His-SMAD3. In these cells, RACK1 was knocked down via shRNA (Fig. S3G-J). Co-immunoprecipitation (Co-IP) assays using anti-Flag or anti-His antibodies were then performed. The results demonstrated that the interactions between SMYD3 and SMAD3 were diminished following the interference of RACK1 (Fig. 3J-K). These findings indicate that the interactions between SMYD3 and SMAD3 are indeed dependent on RACK1.
SMYD3 and SMAD3 co-regulate the expression of downstream TSKU
To investigate the mechanisms underlying the interaction between SMYD3 and SMAD3 in colon cancer cells that promote colorectal cancer carcinogenesis, we analysed the downstream target genes regulated by SMYD3,SMAD3. Initially, we performed RNA-seq analysis on HCT8 cells with SMYD3 interference and on HCT116 cells with SMAD3 interference (Fig. 4A-B, and Table S3). The differentially expressed genes are illustrated in the volcano plot (Fig. S4A-B). We next performed CUT&Tag sequencing in HCT15 cells stably overexpressing Flag-SMYD3 (Table S4). This analysis revealed that SMYD3-binding peaks were primarily enriched at the transcription start site (TSS) of target genes (Fig. 4C, S4C), with a majority localized to promoter regions (Fig. 4D). These results suggest that SMYD3 plays a significant role in regulating the transcriptional expression of downstream genes.Fig. 4SMYD3 and SMAD3 co-regulate the expression of downstream TSKU. A Heatmap of putative SMYD3 direct target genes with upregulation or downregulation upon SMYD3 depletion. B Heatmap of putative SMAD3 direct target genes with upregulation or downregulation upon SMAD3 depletion. C Distribution of SMYD3 Cut-tag reads of binding signals around the 3-kb windows centeredon the transcription start site (TSS) of genes. D Pie-plot showing the distribution of SMYD3 in transcriptionally regulated regions of genes. E Venn diagram showing putative SMYD3 and SMAD3 direct target genes by combinational analyses of both RNA-seq, Cut-tag and ChIP-seq datasets. F Gene Ontology (GO) analysis of the top 10 biological processes enriched by the differentially expressed putative SMYD3 direct target genes. H RT–qPCR analysis of representative genes in HCT116 cells with SMYD3 or SMAD3 overexpression, individually. G Differential ploidy of target genes downstream of SMAD3-SMYD3. Data were shown as mean ± SD. The data were analyzed by Two-way ANOVA. ***p < 0.001
Additionally, we downloaded the ChIP-seq data for SMAD3 from the database (https://guolab.wchscu.cn/hTFtarget/) (Table S5) and overlapped it with the results from the CUT-tag analysis of SMYD3 and the two RNA-seq datasets (Fig. 4E), identifying ten downstream genes co-regulated by SMYD3 and SMAD3. Notably, Gene Ontology (GO) annotation was performed on these ten genes, revealing that most are associated with the positive regulation of cell motility and cell adhesion (Fig. 4F). The differential expression levels of six genes co-regulated by SMYD3 and SMAD3, which are linked to invasive metastasis, are shown in Fig. 4G. RT-qPCR assays validated that TSKU is a common downstream target of both SMAD3 and SMYD3 (Fig. 4H), and Western blot analysis further confirmed that TSKU is positively regulated by both SMYD3 and SMAD3 (Fig. S4D, E). To elucidate the mechanism by which the SMYD3-SMAD3 complex upregulates TSKU expression, we conducted a ChIP assay to assess the recruitment of SMYD3, H3K4me3, and SMAD3 to the TSKU site. We successfully detected the binding of both Flag-SMYD3, H3K4me3, and His-SMAD3 to the TSKU promoter (Fig. S4F-G). Thus, SMYD3 and SMAD3 upregulate TSKU expression by jointly activating its transcription.
RACK1 recruits SMAD3 to SMYD3 and facilitates the transcriptional activation of genes downstream of the SMYD3-SMAD3 complex
We sought to investigate whether RACK1 participates in the regulation of SMYD3-SMAD3 downstream TSKU expression and promotes colorectal cancer. Initially, we overexpressed SMYD3 and SMAD3 in RACK1 knockdown cells. Western blot analysis demonstrated that RACK1 knockdown suppressed the expression of downstream TSKU and inhibited SMAD3 phosphorylation, whereas the levels of H3K4me3 were not significantly altered. Furthermore, the re-expression of SMYD3 or SMAD3 individually was unable to reverse the downregulation of downstream TSKU (Fig. 5A, Fig. S5A). Additionally, RT-qPCR assays further confirmed that RACK1 knockdown inhibited TSKU expression, and this inhibition was not reversed upon the overexpression of SMYD3 or SMAD3 (Fig. 5C, Fig. S5C).Fig. 5RACK1 recruits SMAD3 to SMYD3 and promotes the transcriptional activation of the genes downstream of SMYD3-SMAD3. A Effect of RACK1 knockdown on TSKU, H3K4me3, p-SMAD3 protein expression in RKO cells. B Effect of RACK1 overexpression on TSKU, H3K4me3, p-SMAD3 protein expression in HCT8 cells. C Effect of RACK1 knockdown on TSKU mRNA expression in RKO cells. D Effect of RACK1 overexpression on TSKU mRNA expression in HCT8 cells. E ChIP-qPCR with anti-Flag antibody showed binding of SMYD3 to TSKU. F ChIP-qPCR with anti-His antibody showed binding of SMAD3 to TSKU Data were shown as mean ± SD. The data were analyzed by Two-way ANOVA. ***p < 0.001
Concurrently, we disrupted SMYD3 and SMAD3 while overexpressing RACK1. The Western blot results indicated that RACK1 overexpression upregulated the phosphorylation of SMAD3 and TSKU expression, which was reversed upon the disruption of SMYD3 and SMAD3, respectively (Fig. 5B, Fig. S5B). RT-qPCR assays also confirmed that RACK1 overexpression promoted TSKU transcription, but this promotion was negated when SMYD3 and SMAD3 were disrupted (Fig. 5D, Fig. S5D). Additionally, we validated Slug, Twist involved in EMT, MYC involved in cell proliferation, and SOX2 involved in stemness. The results indicate that Slug, Twist, MYC, and SOX2 are not activated by the SMYD3-SMAD3-RACK1 complex.These results led us to hypothesize that RACK1 recruits SMAD3 to SMYD3, thereby facilitating the transcriptional activation of their downstream targets(Fig. S5E, Fig. S5F). To test this hypothesis, we conducted ChIP experiments that demonstrated enrichment of SMYD3 and H3K4me3 in the promoter region of TSKU, which was unaffected by RACK1 interference (Fig. 5E). Simultaneously, we disrupted RACK1 while overexpressing SMAD3, and the results showed that RACK1 disruption decreased the enrichment of SMAD3 in the promoter region of TSKU (Fig. 5F). Thus, our findings confirm that RACK1 recruits SMAD3 to SMYD3, thereby promoting the transcriptional activation of SMYD3-SMAD3 downstream genes.
The SMYD3-SMAD3 complex promotes colon cancer cell metastasis both in vitro and in vivo, and this process is dependent on RACK1
Given our findings that RACK1 recruits SMAD3 to SMYD3 and promotes the transcriptional activation of SMYD3-SMAD3 downstream genes, we hypothesized that the role of the SMYD3-SMAD3 complex in promoting colorectal cancer metastasis is contingent upon RACK1. Transwell and wound healing assays demonstrated that RACK1 interference significantly inhibited the metastasis of colon cancer cells compared to the control group (Fig. 6A-B; Fig. S6A-B), and this inhibition was not reversed by the overexpression of SMYD3 or SMAD3 (Fig. 6A-B; Fig. S6A-B).Fig. 6SMYD3-SMAD3 promotes colon cancer cell metastasis in vitro and in vivo dependent on RACK1. A Wound healing assay shows the effect of RACK1 knockdown on vector-, SMYD3-overexpressing and SMAD3-overexpressing RKO and HCT116 cell metastases. B Transwell assay shows the effect of RACK1 knockdown on vector-, SMYD3-overexpressing and SMAD3-overexpressing RKO and HCT116 cell metastases. C Wound healing assay shows the effect of RACK1 overexpression on vector-, SMYD3- knockdown and SMAD3- knockdown RKO and HCT8 cell metastases. D Transwell assay shows the effect of RACK1 overexpression on vector-, SMYD3- knockdown and SMAD3- knockdown RKO and HCT8 cell metastases. E The schematic diagram of tail vein lung metastasis model construction. F Representative living images of mice injected with HCT116 transfected by indicated lentivirus into tail vein. The lentivirus was Luci-labelled and therefore stably transfected HCT116 cell lines had in vivo luciferase activity. G Statistical analysis of luciferase bioluminescence intensity (n = 5). H Statistical analysis of the number of pulmonary metastases of each group (n = 5). I Representative images of metastases in murine lung of each group and H&E staining of pulmonary tissue sections; The black arrow indicated the metastasis. For A,B,C and D, significance was determined with the Two-way ANOVA. For G and H, significance was determined with the student unpaired t test. ns, not significant, p > 0.05; *, p < 0.05; ***, p < 0.001; ****, p < 0.0001. Errorbars, ± SD
In parallel, we conducted similar experiments in RACK1 overexpression cell lines, which revealed that RACK1 overexpression significantly enhanced the metastasis of colon cancer cells compared to the control group. However, this enhancement was reversed upon interference with SMYD3 and SMAD3, respectively (Fig. 6C-D; Fig. S6C-D). Additionally, we performed in vivo experiments utilizing lung metastasis models. Initially, we established the colon cancer cell line HCT116 with stable RACK1 interference (Fig. S6E) and developed a tail vein-to-lung metastasis model (Fig. 6E). HCT116 cells and HCT116-shRACK1 cells were injected into mice via the tail vein, and the mice were imaged in real-time five weeks post-injection. During this period, the HCT116-shRACK1 group was treated with the SMYD3 inhibitor BCI-121. Notably, the intensity of lung metastasis fluorescence detected in the shRACK1 group was significantly attenuated compared to the control group. BCI-121 treatment did not further enhance the inhibitory effect of RACK1 interference on lung metastasis in mice (Fig. 6F-G). Compared to the control group, the number of lung metastases was significantly reduced in the RACK1 interference group, and this reduction was not further decreased by BCI-121 treatment (Fig. 6H-I). Together, these in vitro and in vivo data establish that RACK1 is essential for SMYD3-SMAD3-mediated promotion of CRC metastasis.
TSKU is highly expressed in colorectal cancer and associated with poor prognosis
High expression of TSKU in lung cancer cells has been shown to induce changes in molecules associated with epithelial-mesenchymal transition [34], however, the role of TSKU in colorectal cancer remains unclear. To investigate the role of TSKU in colorectal cancer, we first accessed the TNMplot database (https://tnmplot.com/analysis/) to analyze the expression levels of TSKU in colorectal cancer and para-carcinoma tissues, revealing that TSKU is significantly overexpressed in colorectal cancer (Fig. 7A). We also assessed the expression of TSKU at both the mRNA and protein levels in clinical colorectal cancer tissues and adjacent para-carcinoma tissues, confirming the elevated expression of TSKU in colorectal cancer samples (Fig. 7B-C).Fig. 7TSKU is highly expressed in colorectal cancer and associated with poor prognosis. A Box plot shows TSKU expression in Para-carcinoma and colorectal cancer tissues (Tumor) from TNMplot database. B mRNA expression of TSKU in colon cancer tissues (Tumor) and corresponding paracancerous tissue (P) (n = 8). C Western blot analysis of TSKU expression in CRC tissues (T) and corresponding paracancerous tissue (P) (n = 8). D Immunohistochemical (IHC) staining of TSKU expression levels in tumor and normal tissues.A total of 76 pairs of tumor and normal tissues were analysed. E Kaplan–Meier estimates of OS of patients with strong positive TSKU expression vs those with weak positive TSKU expression. F Representative images of immunohistochemical staining of SMYD3、SMAD3、 RACK1 and TSKU of colon cancer specimens (n = 76). G TSKU expression correlates with SMYD3 levels in tissue microarray of colorectal cancer samples. Protein levels of TSKU and SMYD3 were quantified in colon cancer specimens. H TSKU expression correlates with SMAD3 levels in tissue microarray of colorectal cancer samples. Protein levels of TSKU and SMAD3 were quantified in colon cancer specimens. I TSKU expression correlates with RACK1 levels in tissue microarray of colorectal cancer samples. Protein levels of TSKU and RACK1 were quantified in colon cancer specimens. For B, significance was determined with the student unpaired t test. For E, significance was determined with Log–rank (Mantel–Cox) test. ns, not significant, p > 0.05; *, p < 0.05; ***, p < 0.001; ****, p < 0.0001. Errorbars, ± SD
Furthermore, we examined the expression of TSKU in tissue sections from 76 clinical colorectal cancer patients. Immunohistochemical analysis indicated that TSKU was highly expressed in colorectal cancer tissues, while expression levels were notably lower in adjacent non-cancerous tissues. Additionally, TSKU expression was found to be higher in poorly differentiated tumor tissues compared to moderately differentiated tumor tissues (Fig. 7D). We conducted an investigation into the correlation between TSKU expression and clinicopathological features in 76 pairs of colorectal cancer (CRC) tissue samples. Our analysis demonstrated that elevated levels of TSKU expression in CRC patients were significantly associated with poor differentiation (P = 0.0032), advanced T stage (P = 0.0088) and advanced TNM staging (P < 0.0001) (Table 2). Kaplan–Meier survival analysis revealed that overall survival (OS) was significantly shorter in patients exhibiting high TSKU expression compared to those with low TSKU expression (Fig. 7E). Moreover, we retrieved prognostic information regarding TSKU in colorectal cancer from the Kaplan–Meier database, which further corroborated that high TSKU expression is associated with poor patient prognosis (Fig. S7A).Table 2. Correlation between TSKU expression and clinicopathological parameters in CRC (n = 76)/n (%)ParametersCategoryNoTSKU expressionPLow(%)High(%)Total Age34420.4275< 602610(29.4)16(38.1)≥ 605024(70.6)26(61.9)Gender0.7847Male4620(58.8)26(61.9)Female3014(41.2)16(38.1)Tumor size(cm)0.8578< 55022(64.7)28(66.7)≥ 52612(35.3)14(33.3)T stage0.0088T1/T22818(52.9)10(23.8)T3/T44816(47.1)32(76.2)TNM stage< 0.0001I/II2419(55.9)5(11.9)III/IV5215(44.1)37(88.1)Histological Grade0.0032Well-Moderately328(23.5)24(57.1)Poorly4426(76.5)18(42.9)
We conducted immunohistochemical staining on 76 paired colorectal cancer and para-carcinoma tissue samples, with representative images displaying SMYD3, RACK1, SMAD3, and TSKU staining shown in Fig. 7F. High expression levels of SMYD3, RACK1, SMAD3, and TSKU were observed in the majority of colorectal cancer tissues. Based on the immunohistochemical scoring results, we categorized the patients into groups based on high and low expression of SMYD3-TSKU, SMAD3-TSKU, and RACK1-TSKU, and generated Kaplan–Meier survival curves for survival analysis. The results indicated that OS for patients with high expression of SMYD3-TSKU, high expression of SMAD3-TSKU, and high expression of RACK1-TSKU was significantly shorter than that of patients with low expression of these markers (Fig. S7B-D). Additionally, we assessed the correlation between TSKU and SMYD3, SMAD3, and RACK1 based on the immunohistochemical scores, revealing a significant positive correlation among these proteins (Fig. 7G-I). We also examined the correlation between RACK1 expression and clinicopathological features in 76 pairs of colorectal cancer (CRC) tissue samples. Our analysis indicated that elevated levels of RACK1 expression in CRC patients were significantly associated with poor differentiation (P = 0.0243) and advanced TNM staging (P = 0.0221) (Table 3). Consistent with the TNMplot database analysis, IHC scores revealed significant positive correlations between TSKU and SMYD3, as well as between TSKU and SMAD3, in CRC tissues (Fig. S7E, F). Collectively, our data provide compelling clinical evidence that high TSKU expression correlates with poor prognosis in colorectal cancer patients.Table 3. Correlation between RACK1 expression and clinicopathological parameters in CRC (n = 76)/n (%)ParametersCategoryNoRACK1 expressionPLow(%)High(%)Total Age36400.8785< 602612(33.3)14(35.0)≥ 605024(66.7)26(65.0)Gender0.4003Male4620(55.6)26(65.0)Female3016(44.4)14(35.0)Tumorsize(cm)0.5240< 55025(69.4)25(62.5)≥ 52611(30.6)15(37.5)T stage0.7256T1/T22814(38.9)14(35.0)T3/T44822(61.1)26(65.0)TNM stage0.0221I/II2416(44.4)8(20.0)III/IV5220(55.6)32(80.0)Histological Grade0.0243Well-Moderately3220(55.6)12(30.0)Poorly4416(44.4)28(70.0)
Discussion
Cancer metastasis is considered the leading cause of death from cancer, with the vast majority of cancer patients succumbing to metastatic disease rather than the primary tumor [39]. Tumor metastasis consists of a series of biological events: primary tumor cells gradually acquire the ability to invade deeper tissues through mucous membranes; spread via direct infiltration through the bloodstream, lymphatic system, or adjacent structures; and ultimately colonize distant organs while regaining the ability to proliferate [40]. The five-year survival rate for colorectal cancer patients is 64%, but individual prognosis is highly dependent on whether the patient develops metastases during the course of the disease [41]. Despite the many therapeutic options available for colorectal cancer, the prognosis for affected patients remains poor, highlighting the urgent need to develop targeted therapies for its treatment.
Gene transcriptional regulation is frequently associated with dysregulation of epigenetic modifications. Epigenetic regulation, which modulates gene expression through reversible and heritable mechanisms without altering the genomic DNA sequence, primarily involves three molecular mechanisms: DNA methylation, histone post-translational modifications (including methylation and acetylation), and chromatin remodelling [42, 43]. At the molecular level, transcriptional regulation is typically achieved through the formation of transcription factor-mediated multi-protein complexes. These complexes orchestrate the transcriptional activation or repression of target genes by recruiting epigenetic modifiers and associated transcriptional co-regulators [44, 45]. Histone methylases that are aberrantly expressed in tumour cells have been shown to regulate the expression of oncogenes and cancer genes, often in conjunction with transcription factors, and to play an important role in a number of pathophysiological processes such as cell metastasis and mesenchymal transition [46–50].
SMYD3 is a structural protein containing SET (Suppressor of variegation, Enhancer of Zeste, Trithorax) and MYND (Myeloid-Nervy-DEAF-1) domains [8]. SMYD3 is overexpressed in various cancers and is typically associated with disease progression and aggressiveness [14, 51–53]. In colorectal cancer, SMYD3 induces carcinogenesis by stimulating the transcription of several important regulators [51]. Therefore, SMYD3 is a promising diagnostic and therapeutic target for colorectal cancer. Previous studies have shown that SMYD3 promotes prostate cancer metastasis and invasion through the MAPK pathway [54]; facilitates hepatocellular carcinoma metastasis via the upregulation of CDK2 and MMP2 [55]; and enhances the transcription of EZR and LOXL2, facilitating the metastasis of esophageal squamous cell carcinoma cells [56]. This study indicates that SMYD3 is linked to poor survival outcomes in colorectal cancer patients and promotes metastasis in this context.
To further investigate the role played by SMYD3 in colorectal cancer, we performed mass spectrometry combined with immuno-total precipitation in colorectal cancer cells overexpressing Flag-SMYD3, and based on the results of the RACK1, it was scored highly. Furthermore, our experiments revealed the existence of an interaction between SMYD3 and RACK1. Given these findings, it can be suggested that RACK1 may represent a novel interacting protein of SMYD3.
RACK1, as a scaffold protein, mediates multiple signaling pathways that regulate cell proliferation, migration, and transcription. RACK1 has been studied in colorectal cancer; for example, isobutyric acid promotes CRC metastasis through the activation of RACK1, and MYO10 is involved in the FAK pathway, contributing to the malignant phenotype of CRC via RACK1 [26]. It has been reported that RACK1 acts as a scaffolding protein to mediate the interaction between two proteins; for instance, RACK1 enhances the binding of P-glycoprotein to Anxa2 and Src, thereby increasing the invasive ability of drug-resistant cells [57]. RACK1 also inhibits the binding of β-catenin to PSMD2 through competitive inhibition, increasing β-catenin stability and thereby promoting breast cancer progression [58]. Numerous studies have explored the role of scaffolding proteins in mediating protein interactions in colorectal cancer; for instance, PDP1 acts as a scaffolding protein to promote the binding of BRAF and MEK1, facilitating the progression of KRAS mutant colorectal cancer [59]. Isthmin-1 promotes the interaction of EGFR and YBX-1, thereby enhancing the growth and progression of colorectal cancer [60]. Furthermore, literature suggests that RACK1 interacts with SMAD3, leading us to hypothesize that RACK1 mediates the interactions between SMAD3 and SMYD3 [37, 38]. Our experiments demonstrated that interfering with RACK1 inhibited the binding of SMYD3 to SMAD3, as well as the effect of both proteins in promoting colorectal cancer metastasis.
Through integrated analysis of RNA-seq, Cut&Tag, and ChIP-seq data, we identified ten downstream genes co-regulated by SMYD3 and SMAD3 in colorectal cancer (CRC). Gene Ontology (GO) annotation revealed that six of these genes were functionally associated with cell motility. Subsequent qPCR analysis demonstrated that TSKU exhibited the most significant differential expression among these candidates. Western blot analysis further confirmed that both SMYD3 and SMAD3 positively regulated TSKU expression at the protein level. Based on these findings, we selected TSKU for in-depth functional characterization. To evaluate the clinical relevance of TSKU, we performed immunohistochemical analysis on a cohort of CRC patient samples (n = 76). The results revealed that elevated TSKU expression was significantly correlated with advanced tumor stageand poor overall survival. These findings suggest that TSKU may serve as a novel prognostic biomarker and potential therapeutic target in CRC.
In conclusion, our findings demonstrate the complex interplay between SMYD3, RACK1, and SMAD3 in colorectal cancer, wherein the interaction between SMYD3 and SMAD3 is dependent on RACK1. Disruption of RACK1 results in decreased interaction between SMYD3 and SMAD3. From a mechanistic perspective, RACK1 facilitates the recruitment of SMAD3 to SMYD3, thereby promoting the transcriptional activation of the downstream gene TSKU (Fig. S7G). Furthermore, the pathological correlation between TSKU, SMYD3, SMAD3, and RACK1 may have significant clinical implications in colorectal cancer. Additionally, our study demonstrated that RACK1 interference inhibited colorectal cancer cell metastasis in vitro and in vivo. However, treatment with the SMYD3 inhibitor BCI-121 did not enhance the inhibition of lung metastasis in colorectal cancer induced by RACK1 interference. Therefore, RACK1 is a key player in the process of colorectal cancer metastasis, and inhibition of the interaction between SMYD3 and SMAD3 may represent a promising avenue for treatment.
To the best of our knowledge, there are currently no inhibitors for RACK1. The identification of novel RACK1 inhibitors to attenuate the interaction between SMYD3 and SMAD3 holds promise for inhibiting colorectal cancer metastasis. In summary, the SMYD3-RACK1-SMAD3 transcriptional complex identified in this study represents a promising therapeutic target for colorectal cancer.
Supplementary Information
Supplementary Material 1.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Han D, Huang M, Wang T, Li Z, Chen Y, Liu C, et al. Lysine methylation of transcription factors in cancer. Cell Death Dis. 2019;10(4):290.10.1038/s 41419-019-1524-2PMC 644109930926778 · doi ↗ · pubmed ↗
- 2Bottino C, Peserico A, Simone C, Caretti G. SMYD 3: an oncogenic driver targeting epigenetic regulation and signaling pathways. Cancers (Basel). 2020;12(1):142.10.3390/cancers 12010142 PMC 701711931935919 · doi ↗ · pubmed ↗
- 3Yoshino Y, Chiba N. Roles of RACK 1 in centrosome regulation and carcinogenesis. Cell Signal. 2022;90:110207.10.1016/j.cellsig.2021.11020734843916 · doi ↗ · pubmed ↗
- 4Fan Y, Si W, Ji W, Wang Z, Gao Z, Tian R, et al. Rack 1 mediates tyrosine phosphorylation of Anxa 2 by Src and promotes invasion and metastasis in drug-resistant breast cancer cells. Breast Cancer Res. 2019;21(1):66.10.1186/s 13058-019-1147-7PMC 653002431113450 · doi ↗ · pubmed ↗
- 5Dong ZQ, Guo ZY, Xie J. The lnc RNA EGFR-AS 1 is linked to migration, invasion and apoptosis in glioma cells by targeting mi R-133b/RACK 1. Biomed Pharmacother. 2019;118:109292.10.1016/j.biopha.2019.10929231545240 · doi ↗ · pubmed ↗
- 6Yu B, Luo F, Sun B, Liu W, Shi Q, Cheng SY, et al. KAT 6A acetylation of SMAD 3 regulates myeloid‐derived suppressor cell recruitment, metastasis, and immunotherapy in triple‐negative breast cancer. Adv Sci. 2021;8(20):e 2100014.10.1002/advs.202100014 PMC 852949434392614 · doi ↗ · pubmed ↗
- 7Kim JY, Hong N, Park S, Ham SW, Kim E-J, Kim S-O, et al. Jagged 1 intracellular domain/SMAD 3 complex transcriptionally regulates TWIST 1 to drive glioma invasion. Cell Death Dis. 2023;14(12):822.10.1038/s 41419-023-06356-0PMC 1071934438092725 · doi ↗ · pubmed ↗
- 8Yuan Y, Li Y, Wu X, Bo J, Zhang L, Zhang J, et al. POH 1 induces Smad 3 deubiquitination and promotes lung cancer metastasis. Cancer Lett. 2024;582:216526.10.1016/j.canlet.2023.21652638061486 · doi ↗ · pubmed ↗