Smoking-associated upregulation of CBX3 suppresses ARHGAP24 expression to activate Rac1 signaling and promote tumor progression in lung adenocarcinoma

Although tobacco smoking is a risk factor for lung adenocarcinoma (LUAD), the mechanisms by which tobacco smoking induces LUAD development remain elusive. Histone methylation levels in human bronchial epithelial cells have been reported to increase after exposure to cigarettes. In this study, we explored the mechanisms regulating histone methylation in LUAD in response to smoking. We found that the histone H3K9 methylation reader CBX3 was upregulated in current smokers with LUAD, and that CBX3 overexpression promoted LUAD progression. Functional enrichment analyses revealed that CBX3 regulated the activation of Rho GTPases in LUAD. We also found that by forming a complex with TRIM28, TRIM24, and RBBP4, CBX3 repressed the expression of ARHGAP24 and increased the amount of active Rac1 in LUAD cells. Collectively, these results suggest that smoking associated upregulation of CBX3 promotes LUAD progression by activating the ARHGAP24/Rac1 pathway. Hence, the CBX3/ARHGAP24/Rac1 axis may represent a promising therapeutic target in smoking-induced LUAD.

peak of RBBP4, TRIM24, TRIM28, CBX3 and H3K9me3 as indicated.b, the ChIP-qPCR of TRIM28 on the promoter region of ARHGAP24 in H1299 cells. Data presented as Mean ± SD with three replicates (n = 3). Ns, not significant; **, P < 0.01. c, the ChIP-qPCR of TRIM24 on the promoter region of ARHGAP24 in H1299 cells.
Data presented as Mean ± SD with three replicates (n = 3). Ns, not significant; **, P < 0.01; ***, P < 0.001. d, the ChIP-qPCR of TRIM24 on the promoter region of ARHGAP24 in H1299 cells. Data presented as Mean ± SD with three replicates (n = 3). Ns, not significant; ***, P < 0.001. e, A549 cells were transfected with indicated plasmids for 24 h. Cells were collected for the ChIP-qPCR of CBX3 on the promoter region of ARHGAP24 in A549 cells. Data presented as Mean ± SD with three replicates (n = 3). Ns, not significant; **, P < 0.01; ***, P < 0.001. f and g, A549 cells were infected with shControl or shCBX3 for 48 h. Then, cells were transfected with indicated plasmids for another 24 h. Cells were collected for RT-qPCR analysis (f) and Western blotting analysis (g). Data presented as Mean ± SD with three replicates (n = 3). Ns, not significant; *, P < 0.05; **, P < 0.01. h-j, A549 cells were infected with indicated shRNAs for 72 h. Cells were harvested for RT-qPCR analysis.

Supplementary Material and Methods
Public datasets for data mining and bioinformatics analysis 1)TCGA-LUAD：Transcriptome data and clinical information of LUAD patients were obtained from the GDC data portal (https://portal.gdc.cancer.gov/). Data from 522 LUAD patients and 60488 genes were acquired. Of these patients, 59 samples had matched normal tissues.

Identification of key genes involved in the progression of LUAD
Univariate cox regression analysis was first applied to identify genes associated with the RFS of LUAD patients in TCGA-LUAD dataset. Then, genes with P < 0.05 went through lasso-cox regression analysis with 10-fold cross-validation to further select genes associated with the RFS of LUAD. The lasso selection process was replicated by 1000 times.

Identification of smoking-related genes
Since nonsmokers are underrepresented in lung cancer, smoking-related genes were mainly identified between current smokers and former smokers in this study. We also required that the smoking-related genes were differentially expressed in both the TCGA-LUAD and TCGA-LUSC datasets.

Survival analysis
LUAD patients were divided into two groups according to the median expression level of the key gene. The differences in RFS and OS between the high and low expression groups were evaluated by the Kaplan-Meier method, followed by a log-rank test.

GSEA for the key gene
LUAD patients were first divided into two groups according to the median expression level of the key gene. Then, differential expression analysis was applied between the high and low expression groups. Input genes for GSEA were sorted by their logFC values. Signaling pathways activated or suppressed by the key gene were decided by the NES value derived from GSEA.

Downstream targets of the key gene
Downstream targets of the key gene were determined by ChIP-seq and correlation analysis. Binding site in the targeted gene promoter of the key protein or methylation were obtained from ChIP-Atlas database (https://chip-atlas.org/). Correlation analysis between key gene/protein and downstream targeted genes was applied to further validate the ChIP-seq results (|r| ≥ 0.30 & P <0.05).

PPI network between key protein and correlated proteins
Correlation analysis was applied to identify proteins correlated with the key protein (|r| ≥ 0.30 & P <0.05). PPI network among key protein and correlated proteins was constructed using STRING v11.0 (https://string-db.org/).

Statistical analysis and visualization
Microsoft R Open v4.0.2 was used for data mining, bioinformatics analysis and visualization in transcriptomics and proteomics data. IGV v2.9.0 was used for analysis and visualization of ChIP-seq data. STRING v11.0 was used for analysis and visualization of PPI network.

Cell proliferation assay
Cell proliferation was determined using an MTS assay. Lung adenocarcinoma cells Finally, the number of colonies was calculated.

Cell invasion assay
The in vitro cell invasion assay was applied by using a Bio-Coat Matrigel invasion chamber (BD Biosciences). Each transwell chamber was coated with 50 μL matrigel BALB/c-nude mice (4-5 weeks old, 18-20 g) were obtained from Vitalriver (Beijing, China). A459 cells were transduced with different lentiviral particles. After puromycin selection for 72 h, cells (1×10 7 per mouse) were subcutaneously injected into the back of mice. The procedure of xenografts assay was described previously [1].
At the study endpoint, the volume and mass of xenografts were measured. Nicotine

Preparation of CSE
Commercial cigarettes (Hongta, China, 1.0 mg nicotine per cigarette; 10 mg tar per cigarette) were smoked using a vacuum pump for 5 min each, and smoke per 10 mL of serum-free cell growth medium was used to generate 100% CSE-PBS solution. The CSE was prepared 30 min before use, and the pH was regulated to 7.4. The solution was sterilized by filtration with 0.22 µm of filter membrane [2].

Proximity ligation assay (PLA)
The A549 cells were fixed by the blocking solution following the manufacture's protocol (Duolink in situ-fluorescence, sigma). Then, the primary antibodies CBX3 Medium with DAPI was added to cells to take photos under confocal microscope.

ALT, AST, CRE and BUN detection assay
The blood serum of nude mice was collected. The ALT, AST, CRE or BUN were tested following the manufacture's protocol of Alanine aminotransferase Assay Kit    shCBX3-2