High-risk histological subtype-related FAM83A hijacked FOXM1 transcriptional regulation to promote malignant progression in lung adenocarcinoma

Patients and tissue samples

In order to conduct additional research, LUAD and nearby normal tissues were taken from 12 patients at the Xuzhou Central Hospital and preserved at −80 °C. Pathologists from our facility verified all cancers and the corresponding normal tissues. None of the participants in this study had any other cancers, nor had they undergone any radiation or chemotherapy prior to surgery. Each patient’s clinical and pathological characteristics were subsequently gathered following their surgical procedure. All patients gave written informed permission for the study, and it complied with Xuzhou Central Hospital Ethics Committee standards and the Ethics Committee of Jiangsu Cancer Hospital. The Ethics Committee of Jiangsu Cancer Hospital approved the study.

Cell culture

Human LUAD cell lines (A549, PC9, H1975, A427, H358 and DV90) and normal lung epithelial cells (HBE) were purchased from the Chinese Academy of Sciences (Shanghai, China). The aforementioned cells were grown in DMEM media (Invitrogen, Carlsbad, CA, USA), 10% FBS, and 1% penicillin/streptomycin supplements, all from the same source, at 37 °C and 5% CO₂. The media was then replaced every 2 to 3 days while cell growth was periodically monitored. For the following studies, cells in the logarithmic growth phase were chosen. According to earlier investigations (Goel et al., 2017), palbociclib (HY-50767, MCE) was added to the cell media at a concentration of 1 nM, incubated for 48-h.

Cell transfection

Target FAM83A’s sgRNA and FOXM1’s siRNA were designed by GenScript GenCRISPR gRNA Design Tool (https://www.genscript.com/tools/gRNA-design-tool) and BLOCK-iT^™ RNAi Designer, respectively. Both were synthesized by GenScript (Nanjing, China). The backbone of sgFAM83A used pSpCas9 BB-2A-Puro (Ran et al., 2013).Transient transfection of sgRNA-Cas9 plasmid and FAM83A overexpression plasmid were performed via Lipofectamine 3000 (Thermo Fisher Scientific, Waltham, MA, USA). The infected cells were subject to puromycin (1 μg/ml) selection for 5 days after transfection. Individual puromycin resistant colonies were picked up manually and then expanded in 12-well plates. Colonies carrying a deletion allele were checked by Western blots. Lipofectamine iMAX (Thermo Fisher Scientific, Waltham, MA, USA) was performed for siRNA transfection. FOXM1 cDNA, which was synthesized by Genscript (the synthetic sequences were referenced to Addgene #73237) and cloned into the pcDNA3.1 expression vector. According to the instructions, Lipofectamine 3000 (Thermo Fisher Scientific, Waltham, MA, USA) was transfected. Oligonucleotide sequences are listed in Table S1.

Experimental animals

The Institutional Animal Care and Use Committee of Xuzhou Medical University provided full approval for this research (IACUC-2202049). BALB/c nude mice were acquired from GemPharmatech, Inc. (Nanjing, China). BALB/c nude mice were obtained from GemPharmatech, Inc. (Nanjing, China). The participants were randomly divided into two groups (n = 5, with no blinding performed in either group). Cells were injected subcutaneously into the axilla of the nude mice, with each mouse receiving a dosage of 1 × 10⁷ cells. The maximum diameter of the tumor was measured every 2 days after implantation. After 12 days, the mice were euthanized, and the tumor was removed and weighed. AB line zebrafish were acquired from China Zebrafish Resource Center (Wuhan, China). They were randomly divided into two groups (n = 16, no blinding was performed, respectively). After A549 cells incubated with CFSE (Thermo Fisher, Waltham, MA, USA), xenograft perivitelline injection were conducted in 2 day-post-fertilization zebrafish via a Picoliter Microinjector (PLI-100A; Warner Instruments, Hollister, MA, USA) with a glass capillary needle (Sutter, Q100-50-10) made on a laser-based needle puller (P-2000; Sutter, Sacramento, CA, USA). Observed cell extravasation in the caudal vasculature at 3 days post-transplant and imaging on a Zeiss AxioZoom V16 fluorescence microscope.

Dual-luciferase reporter assay

The FOXM1 promoter truncature plasmid were inserted into the pGL3 basic vector (Promega Corporation, Madison, WI, USA). All were co-transfected with a pRL-TK plasmid into cells by using Lipofectamine 3000 (#L3000075; Thermo Fisher Scientific, Waltham, MA, USA) in triplicate. Luciferase activity was measured using the Dual-Glo Luciferase Assay System (#DD1205-01; Vazyme, Nanjing, China) according to the manufacturer’s guidelines. Results represent the mean ± standard deviation (SD) of three experiments.

Immunohistochemistry (IHC)

The LUAD tissues were embedded in paraffin after being treated in 10% formalin. The tissues were then divided into 5 μm-thick sections, and the sections were treated with the primary antibodies anti-FAM83A (#20618-1-AP), anti-Ki-67 (#27309-1-AP), and anti-FOXM1 (#13147-1-AP) (Proteintech, China) for an overnight period. The sections were then stained with a 3,3-diaminobenzidine solution and treated with an HRP-polymer-conjugated secondary antibody (CST, San Antonio, TX, USA) for 1 h at 37 °C.

Immunofluorescence microscopy

Cells were permeabilized with PBS containing 0.05% Triton X-100 after being fixed with 4% paraformaldehyde for 30 min. Cells were washed three times with PBS before being incubated with FOXM1 primary antibody (#ab207298, dilution: 1/200; Abcam, Boston, MA, USA) for a whole night at room temperature. Following three PBS washes, the cells were treated for 1 h with a secondary antibody that was FITC-conjugated. The ZEISS confocal system took pictures of fluorescence (Lsm710, Zeiss, Baden-Württemberg, Germany).

Total RNA isolation, reverse transcription, and qRT PCR

Using the Total RNA Isolation Kit (TIANGEN, Sichuan, China), total RNA was isolated from issues, and cell lines and kept in a freezer at −80 °C. All complementary DNA (cDNAs) were created using the GoldenstarTM RT6 cDNA synthesis kit (TSINGKE, Sichuan, China) in accordance with the user’s manual and kept in a freezer at −80 °C. ABI 7,500 equipment (Applied Biosystems, Foster City, CA, USA) and ChamQ Universal SYBR qPCR Master Mix (Vazyme, Shuangbai Alley, China) were used to conduct qRT-PCR experiments. Table S1 displays the sequence of primers.

Western blot

The media was discarded, and RIPA lysate was added to the cells for total protein extraction. The total protein concentration was measured using the BCA technique (Thermo, Shanghai, China). Protein samples were separated through SDS-PAGE and transferred onto a PVDF membrane. After blocking with 5% skim milk at room temperature for 1 h, the membrane was washed three times with TBST and then incubated overnight at 4 °C with the primary antibody at the recommended dilution ratio. Subsequently, the membrane was washed again with TBST and incubated at room temperature for 1 h with the DyLight 800-labeled secondary antibody (ROCKLAND, USA). The membrane underwent three additional TBST washes before imaging using a fluorescence scanning instrument (Odyssey, Danbury, CT, USA). Primary antibodies against FOXM1 (ab207298, dilution: 1/1,000), FAM83A (ab128245, dilution: 1/1,000), CDC20 (ab183479, dilution: 1/2,000), CDC6 (ab109315, dilution: 1/4,000), cyclin B (ab32053, dilution: 1/8,000), H3 (ab1791, dilution: 1/5,000), and GAPDH (ab9485, dilution: 1/2,500) were bought from Abcam for this work.

Nucleus–cytoplasm fractionation

First, two washes of precooled PBS were performed on 1 * 106 LUAD cells. After scraping the cell layer into 500 l of PBS, it was centrifuged at 500 g for 5 min at 4 °C. The supernatant was then eliminated, and the washing procedure was repeated twice. In the end, LUAD cells in culture were used to isolate nuclear and cytoplasmic protein using NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Fisher Scientific, Waltham, MA, USA) in accordance with the manufacturer’s instructions.

Cell cycle analysis with flow cytometry

After trypsin digestion, cells were collected, twice-washed with PBS, and overnight fixed in ice-cold 70% ethanol. Fixed cells were rinsed twice with PBS before being stained for 30 min at 37 °C in PBS containing propidium iodide (PI, 50 g/mL) and RNase (100 g/mL). On a Beckman Coulter Epics XL-MCL flow cytometer, cell cycle analysis was carried out using System II (version 3.0) software (Beckman Coulter, Brea, CA, USA).

Transwell and matrigel invasion assay

Transwell assay was used to test the LUAD cells’ capacity to invade 24-well Boyden chambers with precoated Matrigel (Corning Incorporated, Corning, NY, USA). Briefly, 1× 105 cells in 200 μL of serum-free media were plated into the top chamber, and 600 μL of RPMI 1,640 medium with 10% FBS was added to the bottom chamber. Cells on the upper surfaces of the transwell chambers were removed after 24 h of incubation at 37 °C with 5% CO₂, and cells that had crossed to the bottom surface were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet. A fluorescence microscope (Zeiss, Baden-Württemberg, Germany) was used to collect the invading cells, which were then counted from five randomly selected areas.

5-Ethynyl-2′-deoxyuridine (EdU) assay

After being planted onto 96-well templates, A549 and PC9 cells were collected 48 h after transfection. After that, cells were treated with 50 mM EdU (#KGA331; KeyGEN, Nanjing, China) for 2 h, fixed with 4% paraformaldehyde, and labeled with Apollo Dye Solution. DAPI was used to counterstain the cell nuclei. A fluorescence microscope (Zeiss, Baden-Württemberg, Germany) was used to spot growing cells with green signals.

Colony formation assay

A total of 300 cells with FAM83A knockdown or NC were seeded in 6-well plates over the course of one night, treated with DMSO (NC), or palbociclib, and then cultivated for an additional 14 days. The cells were then fixed in 4% formaldehyde in PBS (vol/vol) for 10–15 min, washed twice with PBS to remove floating cells, and stained for 20 min in 0.1% crystal violet/10% ethanol. Colonies were observed using an Odyssey Imaging System (LI-COR) after the staining solution had been aspirated, three water washes, air drying, and visualization.

Cell counting kit-8 (CCK8)

Cells with FAM83A knockdown or NC incubated in 96-well plates were treated as indicated and cell proliferation was assessed by CCK-8 assay (SAB Biotech, College Park, MD, USA) at 0, 24, 48 and 72 h post treatment following the manufacturer’s instruction. Optical density (OD) was recorded at 450 nm.

Chromatin immunoprecipitation (ChIP) assay

ChIP was carried out using the Magna ChIP A kit (#17-610; MilliporeSigma, Burlington, MA, USA) in accordance with the manufacturer’s instructions. Anti-Flag antibodies(#14793S) were purchased from CST Inc. The qPCR analysis was carried out as previously mentioned.

Statistical analysis and in silico bioinformatics analysis

R (version 4.1.2) and GraphPad Prism 8 were used to conduct the statistical analysis. The difference between the several groups was compared, and the data are given as the means ± standard deviation (SD). Analysis of differentially expressed genes (DEGs) was performed using the “DESeq2” R package. GSEA assay were conducted via “clusterProfiler” R package. Copy number assay were performed by “maftool” R package. All published data used in this study reveal in Data Availability. The t test or one-way ANOVA were used to examine differences between the various groups. All assay was performed three independent times. At p < 0.05, a difference was deemed significant.

FAM83A was screened as a histological subtype-related transcription factor in LUAD

H&E-stained whole slide images of 541 patients from the TCGA-LUAD database were obtained, and the histological subtypes and proportions of tumor tissues were evaluated, including nine patients with predominance of low-risk pathological subtypes (>50% of lepidic lung adenocarcinoma), 107 patients with predominance of intermediate-risk pathological subtypes (>50% of acinar and papillary lung adenocarcinoma), and 73 patients with predominance of high-risk pathological subtypes (>50% of micropapillary and solid lung adenocarcinoma). Transcription factors play a pivotal role in shaping the molecular evolution patterns and transcriptional profiles of malignant tumors (Goossens et al., 2017; Yu, Pardoll & Jove, 2009). Meanwhile, a variety of transcription factors has been identified as the key determinants of molecular subtypes in malignant tumors. To further explore the differences in transcription factor expression profiles among histological subtypes of lung adenocarcinoma, we analyzed the differences in transcription factor RNA expression profiles among three risk groups, which indicated that 26 transcription factors displayed specifically highly expression in high-risk pathological subtypes, whilst 24 and 23 transcription factors were specifically highly expressed in low-risk and intermediate-risk pathological subtypes, respectively (Figs. 1A, S1A and Table S2). Data retrieved from TCGA-LUAD database and GTEx database (Tang et al., 2017) revealed that transcription factor FAM83A displayed the most significant increase in expression levels within LUAD tissues (log₂(FC) = 4.715) (Fig. 1B). Meanwhile, FAM83A was significantly increased in a variety of malignant tumors including LUAD, ESCA, LUSC, and UCEC, suggesting that it could serve as a malignant phenotypic transcription factor on a pan-cancer scale (Fig. 1C). Subsequently, survival analysis in the TCGA-LUAD database and multiple large-scale LUAD high-throughput cohorts (Bild et al., 2006; Okayama et al., 2012; Rousseaux et al., 2013; Xie et al., 2011) revealed that patients exhibiting high FAM83A expression had significantly poorer overall survival rates and progression-free rates (Figs. 1D–1G). Copy number analysis conducted on tumor tissues and adjacent tissues extracted from LUAD patients demonstrated significant copy number amplification at the FAM83A locus in lung adenocarcinoma tissues (Fig. 1H), which could potentially explain the observed high expression of FAM83A in tumor tissues. We continued to explore the distribution of FAM83A expression in patients with lung adenocarcinoma. It was observed that FAM83A exhibited significantly higher expression levels in patients diagnosed with advanced stage lung adenocarcinoma, as opposed to those diagnosed at earlier stages (Figs. S1B–S1D). We also detected the mRNA expression of FAM83A in 48 pairs of lung adenocarcinoma and adjacent tissues, which showed that FAM83A was significantly overexpressed in lung adenocarcinoma tissues (Fig. S1E), consistent with previous analysis results. Taken together, these data revealed that the expression of transcriptional factor FAM83A was specially increased in the high-risk group among histological subtypes, and also associated with poor outcomes and advanced stages in LUAD.

FAM83A promotes LUAD tumor progression in vitro and in vivo

We evaluated FAM83A mRNA expression levels in multiple LUAD cell lines and normal human alveolar epithelial cells via qRT-PCR, which uncovered that FAM83A expression levels were higher in all six LUAD cell lines than in HBE cell line (Fig. 2A). For functional experiments, A549 cells featuring the lowest expression of FAM83A, and PC-9 cells exhibiting the highest expression of FAM83A were selected. We confirmed that the expression level of FAM83A was significantly decreased and increased when PC9 cells transfected with sgRNA and A549 cells transfected with overexpression, respectively (Fig. S1F). Subsequently, we performed colony formation and CCK8 assays in A549 and PC9 cells to evaluate the proliferative function of FAM83A, which showed that knockdown of FAM83A suppressed cell proliferation in both A549 and PC9 cells, while overexpression promoted cell growth in those cells (Figs. 2B, 2C, and S2A). Furtherly, we conducted EdU assays in A549 cells, and the knockdown of FAM83A inhibited A549 cell proliferation while the overexpression of FAM83A increase the proliferation (Figs. 2D and S2B). We also examined the effect of FAM83A on cell invasion and migration in normal Transwell or coating with Matrigel. As expected, FAM83A upregulation increased cell invasion and migration capability, whereas reduced FAM83A expression led to the opposite results both in A549 and PC9 cells (Figs. 2E, 2F, S2C and S2D). Collectively, these data suggest that FAM83A could promote the malignant phenotype of LUAD in vitro.

Furthermore, zebrafish and BALB/c nude mice tumor xenotransplantation models were used to evaluate the effect of FAM83A on LUAD progression in vivo (Fig. 3A). We administered A549-sgNC and A549-sgFAM83A cells via perivitelline injection in the zebrafish 3 days’ post-injection, decreased expression of FAM83A resulted in a significant reduction in the number of metastasis cells in zebrafish tail (Fig. 3B). At 2 weeks after injection of A549-sgNC and A549-sgFAM83A cells into the axilla of nude mice, downregulation of FAM83A significantly inhibited LUAD tumor growth, and tumor volume was smaller than that of the control group at the end of the experiment (Figs. 3C and 3D). The application of immunohistochemistry (IHC) staining revealed a marked downregulation of FAM83A and Ki-67 expressions in tumor tissues of the sgFAM83A group as opposed to that of the control group (Fig. 3E). Together, these results indicate that FAM83A promotes LUAD tumor progression in vivo.

FAM83A regulates the key genes involved in cell cycle at transcriptional level

To explore the possible transcriptional regulation of FAM83A in LUAD, we analyzed the difference in bulk RNA-seq data between patients exhibiting high FAM83A expression (TOP15%) and low FAM83A expression (Bottom15%) in TCGA-LUAD database. A total of 98 genes with high expression and 50 genes with low expression in top 15% were screened out (Fig. 4A and Table S3). Pathway enrichment analysis showed that patients with high FAM83A expression were significantly enriched in cell cycle-related pathways (Fig. 4B). These results suggested that FAM83A may be involved in the transcriptional activation of cell cycle key genes. The above results were also validated in GSE30219 (Figs. 4C, 4D, and Table S4). Subsequently, we identified key genes that are known to play a role in cell cycle regulation from the selected pool of highly expressed genes within the top 15% (CDC6, CDC20, FOXM1, CDKN3), and assessed the extent of the correlation between FAM83A and these genes within PC-9 cells, which showed that FAM83A could significantly positively regulate the transcription of the aforementioned genes (Fig. 4E). We also verified that certain proteins were related to cell cycle pathway (FOXM1, cyclin B, CDC20 and CDC6), and obtained corresponding results (Fig. 4F). Moreover, the FACS results also indicated that reduced FAM83A expression increased cell G1/G0 cell arrest, whereas FAM83A upregulation led to the opposite results (Fig. 4G). These results confirmed that FAM83A could increase the key genes in transcriptional level to regulate the cell cycle.

FAM83A hijacks the transcriptional regulation of FOXM1 to control the cell cycle

We proceeded to investigate the molecular mechanism by which FAM83A promoted transcription of key cell cycle genes. We firstly explored the correlation between the expression of FAM83A and the aforementioned cell cycle related genes, including FOXM1, CDC6, CDC20, and CDKN3, which revealed that FAM83A exhibited a more significant positive correlation with FOXM1 (P = 6.9e−05, R = 0.18) than others in TCGA-LUAD (Fig. 5A). FOXM1 is a forkhead box transcription factor that is involved in regulating cell proliferation, differentiation and tumorigenesis (Kalathil, John & Nair, 2021; Katzenellenbogen, Guillen & Katzenellenbogen, 2023; Liao et al., 2018). FOXM1 promotes tumor cell proliferation by activating genes involved in cell cycle progression and DNA synthesis. It drives cancer cells to enter and pass through the G1/S and G2/M cell cycle checkpoints (Alvarez-Fernández & Medema, 2013; Chen et al., 2013). To examine the ability of FAM83A as a transcription factor binding specific chromatin region, we designed a Flag tag fused plasmid containing FAM83A open reading frame (ORF) (Fig. 5B, upper), and designed four pairs of 500 bp primers at the promoter of FOXM1. ChIP-PCR results showed that Flag antibody significantly bound to the region 500 bp upstream of FOXM1 transcription start site (Fig. 5B, bottom). The truncated luciferase reporter plasmids with intervals of 500 bp according to the FOXM1 promoter were designed (Fig. 5C, left), which confirmed that the first 500 bp region upstream the transcription start site (TSS) was the core region of FOXM1 transcriptional regulation (Fig. 5C, right). FOXM1 is a transcription factor that serves as a crucial regulator of cell cycle progression, thereby controlling cell proliferation, DNA replication, and mitosis by modulating the expression of genes that are vital for these processes (Tan, Raychaudhuri & Costa, 2007). Nucleocytoplasmic separation analysis and immunofluorescence localization experiments revealed that increased FAM83A expression significantly enhanced the nuclear expression of FOXM1 (Figs. 5D and 5E). In conclusion, these results suggested that FAM83A hijacked transcription of FOXM1, the key gene of cell cycle, by binding to upstream 500 bp region of the promoter.

FAM83A regulated the cell cycle dependent on FOXM1

Furthermore, we further explored whether FAM83A promoted malignant progression in LUAD cells dependent on FOXM1. Colony formation and FACS cell cycle assays revealed that cell proliferation and cell cycle suppressed by FAM83A knockdown were reversed by co-transfection with FOXM1 overexpression plasmid. Conversely, the overexpression of FAM83A was found to promote LUAD cell proliferation and mitosis, while co-transfection with FOXM1 siRNA-pool resulted in a dampening effect on these outcome (Figs. 6A, 6B, S3A and S3B). Palbociclib is a selective inhibitor of the cyclin-dependent kinases CDK4 and CDK6 (Rocca et al., 2014). By inhibiting CDK4/6, Palbociclib ensures that the cyclin D-CDK4/6 complex cannot aid in phosphorylating Rb, which prevents the cell from G1 phase, and in turn from proceeding through the cell cycle (Xu et al., 2017). Therefore, we continued to explore whether Palbociclib could effectively restrict the malignant progression of cancer cells with high FAM83A expression. The Colony formation showed that Palbociclib could only slightly inhibit cell proliferation and cell cycle in LUAD cells A549 with low FAM83A expression. However, Palbociclib significantly inhibited cell proliferation and cell cycle in A549 cells overexpressing FAM83A (Figs. 6C, 6D, and S3C). This suggested that tumors exhibiting high FAM83A expression regulated cell cycle dependent on FOXM1, which may benefit from Palbociclib treatment.

FAM83A/FOXM1 axis positively correlates with malignant progression in LUAD patients

We measured the expression of FAM83A in 12 LUAD patients and divided them into two groups according to their respective FAM83A expression levels (Figs. 7A and 7B). The result demonstrated that patients presenting with high levels of FAM83A expression also displayed increased expressions of Ki-67 and FOXM1, as compared to patients exhibiting low FAM83A expression (Fig. 7C). Moreover, it was observed that patients with high levels of FAM83A/FOXM1 expression exhibited poorer patient outcomes, when compared to their low-expression counterparts (Fig. 7D). Additionally, the expression of CDK2 and CDC20, the keys genes involved in cycle cell, were found to be heightened in patients with high FAM83A/FOXM1 expression as opposed to those displaying low expression levels (Fig. 7E). Collectively, our data indicated that FAM83A/FOXM1 axis expression indicated malignant progression in LUAD patients.