Methyltransferase SETD2 inhibits tumor growth and metastasis via STAT1–IL‐8 signaling‐mediated epithelial–mesenchymal transition in lung adenocarcinoma

Abstract Lung adenocarcinoma (LUAD) is a major subtype of non–small‐cell lung cancer, which is the leading cause of cancer death worldwide. The histone H3K36 methyltransferase SETD2 has been reported to be frequently mutated or deleted in types of human cancer. However, the functions of SETD2 in tumor growth and metastasis in LUAD has not been well illustrated. Here, we found that SETD2 was significantly downregulated in human lung cancer and greatly impaired proliferation, migration, and invasion in vitro and in vivo. Furthermore, we found that SETD2 overexpression significantly attenuated the epithelial–mesenchymal transition (EMT) of LUAD cells. RNA‐seq analysis identified differentially expressed transcripts that showed an elevated level of interleukin 8 (IL‐8) in STED2‐knockdown LUAD cells, which was further verified using qPCR, western blot, and promoter luciferase report assay. Mechanically, SETD2‐mediated H3K36me3 prevented assembly of Stat1 on the IL‐8 promoter and contributed to the inhibition of tumorigenesis in LUAD. Our findings highlight the suppressive role of SETD2/H3K36me3 in cell proliferation, migration, invasion, and EMT during LUAD carcinogenesis, via regulation of the STAT1–IL‐8 signaling pathway. Therefore, our studies on the molecular mechanism of SETD2 will advance our understanding of epigenetic dysregulation at LUAD progression.


| INTRODUC TI ON
Lung cancer remains one of the deadliest malignancies worldwide. [1][2][3] Most patients have advanced NSCLC. Histologically, it can be divided into lung adenocarcinoma (LUAD) and squamous cell carcinoma, between which LUAD is the most common subtype, accounting for more than 70% of NSCLC. 4 Despite tremendous advances in chemotherapy and radiation over the past few decades, the outlook for LUAD patients is bleak, with just over 15% surviving for 5 years after diagnosis. 5,6 Therefore, there is an urgent need to unveil underlying mechanisms and develop new therapeutic strategies to improve the treatment of LUAD.
Epigenetics is defined as heritable changes in gene expression, not as a result of changes in DNA sequence. 7 Over the past few years, it has become increasingly clear that dysfunctional epigenetic regulatory processes play a central role in the development and progression of cancer. 8,9 In contrast with DNA mutations, epigenetic modifications are reversible and therefore suitable for drug intervention. 10 Reversible histone methylation is an important process in epigenetic regulation and its role in cancer has made lysine methyltransferase and demethylase promising targets for new anticancer drugs. 11 Set domain-containing 2 (SETD2) is a major mammalian methyltransferase, responsible for catalyzing the trimethylation of histone 3 on lysine 36 (H3K36me3). 12,13 The SETD2 mutation has been found in colorectal cancer, renal clear cell carcinoma, breast cancer, glioma, acute leukemia, chronic lymphocytic leukemia, and other tumors. [13][14][15] Previous studies have identified that SETD2 and its dependent H3K36me3 both participate in both active or suppressed transcriptional regulation in a series of cellular processes reviewed previously. 16 SETD2-mediated H3K36me3 can prevent the initiation of spurious transcription by recruiting histone deacetylase complexes, and depletion of Set2 results in increased levels of this adverse transcription. 17,18 In recent decades, research on SETD2 dysfunction has been ongoing in order to explore its role in NSCLC progression. 15,[19][20][21] However, the underlying mechanism of SETD2 in the pathological progression of LUAD is largely unknown.
Here, combined with data mining of public patient data sets and human LUAD tissue arrays, we mainly utilized lung cancer cell lines to establish that SETD2 functions as a putative tumor suppressor in LUAD. Mechanistic investigation indicated that SETD2 and its mediated H3K36me3 negatively regulate IL-8 transcription in a Stat1dependent manner. Consequently, SETD2 ablation upregulates IL-8 expression to stimulate EMT progression to promote tumorigenesis in LUAD. Therefore, our results highlight that patients with SETD2 loss exhibit worse clinical outcomes with potentially therapeutic implications in LUAD.

| RNA isolation and real-time qPCR
Total RNA was extracted using TRIzol reagent according to the manufacturer's instructions. First-strand cDNA was synthesized using Superscript II (Invitrogen) and 1 μg of total RNA was used in each cDNA synthesis reaction. SYBR green Universal Master Mix reagent (Roche) and primer mixtures were used for real-time qPCR. GAPDH was used as the internal reference for mRNAs.
F I G U R E 1 SETD2 expression is downregulated in human lung cancer. (A) The mRNA expression differences for SETD2 between LUAD tissues and matched tumor-adjacent tissues were analyzed using real-time PCR. (B) Protein levels of SETD2 in LUAD tissues and matched tumor-adjacent tissues were analyzed using IHC. (C) Kaplan-Meier analyses were conducted to explore the correlation of SETD2 expression with overall survival of NSCLC patients. (D) mRNA expression differences for SETD2 between human lung cancer cell lines (A549, H1975, H1299, H1650, and PC-9) and human bronchial epithelial cell line HBE. Data are presented as means ± SEM; statistical significance: *p < 0.05; **p < 0.01

| Cell proliferation assay
Cell proliferation assay was performed using CellTiter 96 ® Non-Radioactive Cell Proliferation Assay (CCK-8) kit (Dojindo) according to the manufacturer's instructions. Cells (1 × 10 4 cells/ml) were seeded into a 96-well plate (100 µl/well) and cultured in an incubator with 5% CO 2 at 37°C. Each well was supplemented with 10 µl CCK-8 solution and then incubated at 37°C for 2 h. The spectrophotometric absorbance at 590 nm was measured for each sample. All experiments were repeated three times in triplicate.
For soft agar colony formation assays, cells were suspended in RPMI 1640 medium containing 0.35% low-melting agar (Invitrogen) and 10% FBS and seeded onto a coating of 0.8% low-melting agar in RPMI 1640 medium containing 10% FBS. Plates were incubated at 37°C and 5% CO 2 . Colonies were counted after a 3-week or 4-week culture. Triplicates were required for each experiment.

| Wound healing assay
The evaluation of cell migration ability was performed using a wound scrape assay to examine the role of SETD2 and IL-8 in the regulation of the migration ability of LUAD cell lines as described previously. 20

| Transwell invasion assay
The transwell culture system was performed to examine the invasive ability of SETD2 and IL-8 on LUAD cell lines as described previously. 20

| RNA-seq and data analysis
HiSeq RNA-seq was used to detect total RNA with or without the SETD2 deletion in A549 cells. Transcriptome reads from the RNAseq assay were mapped to the reference genome (HG19) using the bow tie tool. The gene expression level was quantified using the RSEM software package. Significance was treated by setting the pvalue threshold to 0.05. Differentially expressed genes were then analyzed using the clusterProfile software package to enrich for biological pathways.

| Co-immunoprecipitation
Co-immunoprecipitation experiments were conducted using the methods described previously. Anti-SETD2, anti-H3K36me2, and anti-p-STAT1 antibodies (10 μg) were used individually for the co-immunoprecipitation experiments with normal or SETD2knockdown A549/H1975 cells lysates. The mixture of antibodies and powders dissolved products was incubated at room temperature during a 50-min rotation and immunoprecipitation reaction of antibodies was performed by adding 50 μl protein A-Sepharose (Sigma-Aldrich) from 50% (W/V) and phosphate mud at room temperature for 30 min. The protein A-Sepharose-antibody protein complex was centrifuged at 2000 g for 5 min at 4°C in a chilled microcentrifuge, and the supernatant was discarded. Beads were washed with PBS five times. The washed beads were boiled in SDS loading buffer, separated by SDS-PAGE, and then analyzed using western blotting.

| Immunohistochemistry assays
Tumors from xenograft models were collected and then embedded in paraffin after being fixed in 4% paraformaldehyde (PFA).

| Xenograft tumor model
Here, 5-week-old BALB/c nude mice was purchased from the SLAC Animal Center (Shanghai, China) and then used for the xenograft tumor model. A549 cells were injected subcutaneously into nude mice and then tumor volumes were monitored every 5 days. Tumor volumes were estimated by length and width and calculated using the following formula: At ~1 month later, the nude mice were sacrificed and then tumors were excised, photographed, and weighed.

| Statistical analysis
All experiments were performed using three independent repeated experiments with cells. GraphPad Prism 8.0 software was applied for statistical analyses. Data in all figures are presented as the mean ± SEM. Statistical significance was determined using multiple t-test, one-way ANOVA, two-way ANOVA, Pearson correlation coefficients, or log-rank test. For all statistical tests, the 0.05 level of confidence (two-sided) was accepted for statistical significance.

| SETD2 expression is downregulated in human lung cancer
To explore the possible role of SETD2 in lung cancer, data mining using a public data set indicated that SETD2 expression was decreased in tumors compared with the normal counterparts ( Figure 1A). Moreover, the SETD2 expression pattern was also examined using human LUAD tissue array. Results showed that the expression level of SETD2 was significantly decreased in tumors compared with normal adjacent lung epithelial tissues ( Figure 1B).
The survival analysis using KM Plot (http://kmplot.com/) also demonstrated that the patients with low expression of SETD2 had a poor prognosis ( Figure 1C). Similarly, human lung cancer cell lines (A549, H1975, H1299, H1650 and PC-9) showed a significantly lower expression of SETD2 compared with human bronchial epithelial cell line HBE ( Figure 1D). Together, these findings highlighted SETD2 as a prognostic biomarker for LUAD patients and the causal role of SETD2 in tumorigenesis of lung cancer.

| SETD2 impairs proliferation, migration, and invasion ability of LUAD cells
To investigate the biological roles of SETD2 in LUAD, we first ectopically overexpressed and silenced SETD2 in lung cancer cells A549 and H1975. Real-time qPCR (Figure 2A) and western blotting ( Figure 2B) analyses were performed to confirm the overexpression of SETD2.
We found that increased expression of SETD2 significantly attenuated cell proliferation ( Figure 2C), migration ( Figure 2D), and invasion ( Figure 2E) in A549 and H1975 LUAD cells, whereas silencing of SETD2 improved these features (except invasion). Together, our results indicated that SETD2 impaired proliferation, migration, and invasion ability of LUAD cells.

| SETD2 impairs EMT of LUAD cells
EMT is a conserved cellular process in which epithelial tumor cells lack polarity and transform into a mesenchymal phenotype. To demonstrate whether the SETD2 affected EMT, we used western blotting to assay the molecular marker levels of EMT (E-cadherin, N-cadherin, and vimentin). As shown in Figure 3A Tumor volume = length * width 2 ∕2 F I G U R E 3 SETD2 inhibits EMT of A549 and H1975 cells. (A) The molecular marker levels of EMT (E-cadherin, N-cadherin, and vimentin) in SETD2-overexpressed A549 and H1975 cell were evaluated using western blot. (B) The molecular marker levels of EMT (E-cadherin, N-cadherin, and vimentin) in the SETD2-knockdown A549 and H1975 cells were evaluated using western blot. (C) The morphological change and vimentin/E-cadherin switch in SETD2-overexpressed A549 and H1975 cells were evaluated using confocal microscopy. Data are presented as means ± SEM, n = 3; statistical significance: *p < 0.05

| SETD2 negatively regulates IL-8 expression
To get an insight into the molecular basis of SETD2 impairing lung cancer progression, we performed gene expression analysis using the SETD2 overexpression and control A549 cells. As shown in the volcano plots ( Figure 4A), IL-8 was found to be significantly downregulated upon SETD2 overexpression and upregulated by SETD2 knockdown ( Figure 4B). We further found that IL-8 expression was significantly  To evaluate the relationship between SETD2 and IL-8, the expression profile of lung cancer in the Gene Expression Omnibus (GEO) database (ID: GSE40791) was used. The level of IL-8 in lung tumors was significantly higher than that in the normal lung tissues ( Figure 4G) and negatively correlated with the level of SETD2 ( Figure 4H). Taken together, these results revealed that SETD2-catalyzed H3K36me3 transcriptionally inhibited IL-8 transcription.

| SETD2-mediated H3K36me3 prevents STAT1 from activating IL-8 expression
To understand how H3K36me3 negatively regulated IL-8 expression, we first confirmed that SETD2-catalyzed H3K36me3 was associated with IL-8 expression. Western blotting analyses were performed using cell lysates from A549 and H1975 cells with or without SETD2 overexpression or knockdown. Results indicated that the H3K36me3 levels were significantly increased or reduced upon SETD2 overexpression and knockdown, respectively ( Figure 5A), as the IL-8 promoters contained binding sequences for STAT1. Moreover, activation of the JAK-STAT1 signaling pathways stimulated IL-8 transcription. Therefore, we next investigated the role of STAT1 in SETD2-inhibited IL-8 expression. Western blot analysis revealed that the phosphorylation levels of STAT1 were significantly increased or reduced upon SETD2 overexpression and knockdown, respectively ( Figure 5B). It is well established that STAT family members are phosphorylated by receptor associated kinases, and then form homodimers or heterodimers that translocate to the cell nucleus where they act as transcription activators. Therefore, translocation of STAT1 into the nucleus was analyzed using a Zeiss confocal microscope. As expected, SETD2 overexpression obviously increased the phosphorylation levels of STAT1 in the nucleus in the A549 and H1975 cells ( Figure 5C).
Moreover, co-immunoprecipitation (co-IP) experiments in A549 and H1975 cells using antibodies against SETD2 showed that SETD2 co-located with STAT1, and that this was dramatically diminished by SETD2 downregulation (Figure 5D). In addition, independent ChIP-qPCR results showed enrichment of the IL-8 promoter in anti-SETD2,

F I G U R E 6 IL-8 administration reverses the impairment of SETD2 on the proliferation, migration, and invasion ability of LUAD cells. (A)
The effects of IL-8 treatment with or without STED2 overexpression on the cell proliferation ability of A549 and H1975 cells were analyzed using a CCK-8 assay. (B) The effects of IL-8 treatment with or without STED2 overexpression on the cell migration ability of A549 and H1975 cells were analyzed using a wound healing assay. (C) The effects of IL-8 treatment with or without STED2 overexpression on the cell invasion ability of A549 and H1975 cells were analyzed using a transwell invasion assay. Data are presented as means ± SEM, n = 3; statistical significance: *p < 0.05, **p < 0.01 H3K36me3, and STAT1 ChIP assays ( Figure 5E), which were similarly decreased in the SETD2-knockdown A549 and H1975 cells ( Figure 5E).
Taken together, these results revealed that SETD2-mediated H3K36me3 prevented STAT1 from activating IL-8 expression.

| IL-8 administration reverses the impairment of SETD2 on proliferation, migration, and invasion ability of LUAD cells
To evaluated the critical role of IL-8 in the SETD2 impaired proliferation, migration, and invasion ability of LUAD cells, we next detected the effect of exogenous IL-8 treatment in SETD2-overexpressed LUAD cells, compared with vector control LUAD cells. As shown in Figure 6A, the SETD2-impaired cell viability of LUADs was obviously rescued using endogenous IL-8 treatment. Similarly, the decreased migration ( Figure 6B) and invasion ( Figure 6C

| DISCUSS ION
LUAD's tumorigenesis is thought to require multiple subsequent mutations in genes associated with cell growth, differentiation, and survival. 22,23 Some studies have shown that H3K36 methylation is associated with abnormal differentiation or proliferation. 10 It has been reported that the absence of SETD2 would impair the differentiation of ES cells, while the expression of H3.3 mutants that inhibited the methylation of H3K36 would impair the differentiation of chondrocytes and mesenchymal progenitors. 24 SETD2 and its catalyzed H3K36me3 play crucial roles in maintaining chromosome integrity and regulating gene transcription. 25,26 Mutations and deletions of the SETD2 gene have been identified in several cancers. [27][28][29][30][31] In a very recent study, SETD2 was reported to inhibit colon cancer by modulating alternative splicing. 20 Moreover, SETD2 was reported to directly mediate STAT1 methylation on lysine 525 via its methyltransferase activity, which reinforced IFN-activated STAT1 phosphorylation and antiviral cellular responses. 12 However, the roles of SETD2 loss in the progression of LUAD have not been well established. Here, we determined that SETD2 was significantly decreased in LUAD. SETD2 loss greatly upregulated IL-8 expression via its enzymatic activity in catalyzing H3K36me3, which bound STAT1 to the IL-8 promoter and activated its expression ( Figure 7D). have been used to study the role of IL-8 in carcinogenesis. 36,37 In this study, we identified that SETD2-mediated H3K36me3 could prevent transcription factor STAT1 assembly on the IL-8 promoter, which impaired the activation of IL-8. Furthermore, IL-8 treatment obviously reversed the inhibitory effect of SETD2 in the proliferation and metastasis of LUAD cells in vitro and in xenografts.

IL-8 is a pro-inflammatory
Tumor EMT is a phenotypic transformation that promotes the acquisition of fibroblast-like morphology by epithelial tumor cells, thereby enhancing the motility and aggressivity of tumor cells, enhancing metastasis tendency and resistance to chemotherapy, radiotherapy, and some small-molecule targeted therapies. 38,39 Evidence suggests that an autocrine-positive ring exists between IL-8 and the EMT transcription factor Brachyury in breast and lung cancer cell lines, which has been well documented. The addition of purified IL-8 to cancer cells has been shown to increase the percentage of aldehydepositive cells and enhance the migration and invasiveness of cancer cells in vitro. [40][41][42] In this study, we demonstrated that IL-8 is negatively regulated by SETD2-mediated H3K36me3. Using in vivo xenograft experiments, we also confirmed that IL-8 administration significantly reversed the inhibitory of EMT by SETD2. In summary, our findings highlighted the suppressive role of SETD2/H3K36me3 in cell proliferation, migration, invasion, and EMT during LUAD carcinogenesis, via regulation of the STAT1-IL-8 signaling pathway. Therefore, our studies on the molecular mechanism of SETD2 will advance our understanding of epigenetic dysregulation at LUAD development.

ACK N OWLED G M ENTS
We thank all patients involved in this study.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

CO N S E NT FO R PU B LI C ATI O N
Written consents for publication were obtained from all the patients involved in our study.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data in our study are available upon request.