Tricaine, 1-phenyl-2-thio-urea (PTU), protease, and DMSO were purchased from Sigma-Aldrich (St. Louris, MO, United States). CM-DiI was purchased from Life Technologies (CA, United States). Fetal bovine serum (FBS) and Roswell Park Memorial Institute basal medium 1640 (RPMI 1640) were purchased from Basal Media Technologies (Shanghai, China).

NSCLC cell lines A549 and H1299 were obtained from the American Type Culture Collection (ATCC). A549 and H1299 cells were cultured in RPMI 1640 supplemented with 10% FBS at 37°C with 5% CO₂.

The transgenic zebrafish [Tg (fli-1: EGFP)] which expresses enhanced green fluorescent protein (EGFP) in endothelial cells was obtained from the Model Animal Research Centre of Nanjing University. Zebrafish were handled as we previously reported (Fan et al., 2021). Briefly, adult zebrafish around 5–10 months of age with a 1:2 male-to-female ratio were used for mating. The zebrafish embryo age was expressed as hours post fertilization (hpf) or days post injection (dpi). The zebrafish studies were approved by the Institutional Animal Care and Use Committee (IACUC) at Nanjing Tech University.

Zebrafish embryos at 24 hpf were dechorionated using protease (1 mg/mL, Sigma-Aldrich) and then placed in a PTU-containing E3 medium to prevent pigmentation. A549 and H1299 cells were labeled with CM-DiI according to the manufacturer’s instructions. Zebrafish embryos at 48 h post-fertilization (hpf) were anesthetized with tricaine. Approximately 100 cells of A549 or H1299 were injected into the perivitelline space (PVS) using a microinjector (IM-31, Narishige, Japan) for metastasis modeling. Additionally, 200-300 cells of A549 or H1299 were injected into the yolk sac to quantify cell proliferation. The injected zebrafish embryos were kept at room temperature for 1 h and then placed in an incubator at 32 °C until the end of the experiment. At 1 dpi, the successfully injected zebrafish embryos were collected under the microscope for subsequent experiments, as we have previously reported (Fan et al., 2021). For the zebrafish xenograft experiments, each experimental condition included more than 30 embryos and was independently repeated three times. Tumor cell proliferation in zebrafish embryos was calculated according to Formula 1 : F o l d   c h a n g e s = c e l l   n u m b e r s   o f   e m b r y o s   a t   4   d p i   c e l l   n u m b e r s   o f   e m b r y o s   a t   1   d p i (1)

Formula 1, Tumor cell proliferation of NSCLC cells.

BM and tail metastasis (TM) cells in zebrafish embryos were calculated according to Formulas 2, 3, respectively. F o l d   c h a n g e s = t h e   n u m b e r   o f   b r a i n   m e t a s t a s i s   c e l l s   a t   4   d p i t h e   n u m b e r   o f   b r a i n   m e t a s t a s i s   c e l l s   a t   1   d p i (2)

Formula 2, Tumor cell BM of NSCLC cells F o l d   c h a n g e s = t h e   n u m b e r   o f   t a i l   m e t a s t a s i s   c e l l s   a t   4   d p i t h e   n u m b e r   o f   t a i l   m e t a s t a s i s   c e l l s   a t   1   d p i (3)

Formula 3, Tumor cell TM of NSCLC cells.

Brain tissue, tail tissue, and injection site tissue of 265 the H1299 cells xenografted zebrafish embryos were isolated and collected (Fan et al., 2021). The collected zebrafish tissues were subjected to RNA extraction, purification, and library construction, followed by paired-end sequencing using Next-generation sequencing, based on the Illumina sequencing platform (Hiseq X ten platform).

Upon receiving the sequencing data, we conduct a quality assessment. Raw sequencing data achieved an average Q30 score >85% with 20 million reads per sample (30 Gb total depth). Read alignment to the reference genome using HISAT2 with default parameters, achieving an average alignment rate of 93.97%. We employed DESeq2 for statistical analysis, using default parameters for read count normalization and dispersion estimation. Genes with adjusted p-values <0.05 (Benjamini–Hochberg correction) and |log2FC| > 2 were considered differentially expressed. Then we have to normalize the Read Count of the genes by HTSeq software. The gene expression differential fold can be obtained by DESeq software, and the genes with adjusted p-values <0.05 and the expression differential fold |log2 FoldChange| > 2 are considered to be differentially expressed genes (DEGs). Differential gene volcano maps were drawn by Sangerbox 3.0 (http://vip.sangerbox.com/home.html). Then we used the R language Pheatmap package to perform cluster analysis on the screened differential genes, and the hierarchical clustering results of differential genes could be obtained. Gene ontology GO enrichment analysis using top GO software annotates differential genes to GO terms, counts the number of differential genes in each term, and calculates the P-value by the hypergeometric distribution method to filter out GO terms significantly enriched in differential genes with a P-value <0.05, thus identifying the major biological functions in which the differential genes are involved. Similarly, KEGG (http://www.kegg.jp/) enrichment analysis of differential genes was performed using DAVID (https://david.ncifcrf.gov/), and KEGG Pathway with significant enrichment of differential genes (P value <0.05 and |log2FC|> 2)was screened according to the P-value to identify the metabolic pathways and signaling pathways in which the differential genes are mainly involved. Differential genes significantly enriched for functions and pathways were mapped using ShinyGO v0.75 (http://bioinformatics.sdstate.edu/go/).

After GO enrichment and KEGG enrichment analysis, the enriched differential genes were subjected to clinical relevance studies to assess their association with the clinic. Clinical data on lung cancer patients were downloaded through the TCGA (https://www.cancer.gov/) and UCSC Xena databases (TCGA Lung Adenocarcinoma (LUAD) cohort, https://xenabrowser.net/heatmap/) and graphically analyzed using GraphPad Prism 9.0. DEGs were evaluated further with Kaplan–Meier Plotter database (https://kmplot.com/analysis/index.php?p=service&cancer=lung).

Trizol reagent (Yifeixue, Nanjing, China) was used for the NSCLC cells to extract total RNA. Then total RNA was reverse transcribed to cDNA using a 1st Strand cDNA Synthesis kit (Yifeixue, Nanjing, China). The real-time PCR assay was performed using the 7900 HT system (ABI, United States) and SYBR green PCR kit (Yifeixue, Nanjing, China) to detect the expression levels of genes. The GAPDH gene was used as an internal standard. The thermal cycle procedure was as follows: 95 °C for 30 s; then 40 cycles of 95 °C for 10 s and 60 °C for 30 s. The relative quantification of PCR results was determined by -ΔΔCT. The sequences of primers are listed in Supplementary Table S1. The experiments were performed with three biological replicates (independent samples).

A549 cells overexpression GNA13 (A549-pcDNA-GNA13 group) and the matched control cells were established using Lipo fectamine2000 carrying the pcDNA plasmid (Gene Pharma, shanghai, China). H1299 cells were transfected with siRNA (Gene Pharma, shanghai, China) to knock down GNA13 (H1299-siGNA13 group) using Lipo fectamine 2000. The three GNA13 siRNA sequences used were as follows: siRNA#1 Forward primer 5′→3′GGCGUCGAGAAUUUCAACUTT and Reverse primer 5′→3′AGUUGAAAUUCUCGACGCCTT; siRNA#2 Forward primer 5′→3′CCUGCUAUAAGAGCAUUAUTT and Reverse primer 5′→3′AUAAUGCUCUUAUAGCAGGTT; siRNA#3 Forward primer 5′→3′GGUGGUCAGAGAUCAGAAATT and Reverse primer 5′→3′UUUCUGAUCUCUGACCACCTT. Cells were harvested after 24 or 48 h of transfection, and then gene knockdown and overexpression efficiencies were quantified by qRT-PCR. “H1299-control” represents the untransfected wild-type cell control, while “H1299-NC” denotes the negative control group transfected with scrambled siRNA. Similarly, “A549-control” refers to the wild-type cells without overexpression, and “A549-NC” indicates the negative control group overexpressed with scrambled pcDNA vector.

Transfected NSCLC cells were seeded in 96-well plates at 4 × 10³ cells/well. Cell viability was determined by the CCK-8 assay (Vazyme, Nanjing, China) following the manufacturer’s instructions. The absorbance at a wavelength of 450 nm was then measured using a microplate reader (BioTek, Winooski, VT, United States) after 1 and 2 days.

For the NSCLC cells wound healing assay, 2.5 × 10⁵ cells of A549 or H1299 cell line were seeded in 6-well plates, and vertical scratches were made using a 1 mL pipet tip after overnight incubation. Cells were then cultured in a medium with 2% FBS for 24 h. Cell migration rate was measured using Formulas 4, 5. a v e r a g e   s c r a t c h   w i d t h = s c r a t h   a r e a s c r a t h   l e n g t h (4)

Formula 4, Mean value of scratch width c e l l   m o b i l i t y % = 0   h   s c r a t c h   w i d t h − 24   h   s c r a t c h   w i d t h 0   h   s c r a t c h   w i d t h × 100 % (5)

Formula 5 Cell mobility.

After 24 h of serum-free culture, 5 × 10⁴ cells were seeded in the upper chamber (pre-coated with Matrigel) with 200 μL medium with 2% FBS. The lower chamber was filled with 500 μL medium with 10% FBS. The invasion chamber was incubated for 24 h, then the upper chamber was washed, methanol fixed, stained with 0.5% crystal violet solution and photographed.

The total protein from H1299-control, H1299-NC, H1299-siRNA, A549-control, A549-NC, and A549-pcDNA was extracted using RIPA lysis buffer (Beyotime, Shanghai, China) with 1% PMSF (Beyotime, Shanghai, China). The concentration was measured using the BCA kit (Beyotime, Shanghai, China) according to the manufacturer’s instructions. According to the Western blotting protocol described, the level of protein investigated were evaluated with the following antibodies: N-cadherin Polyclonal antibody (22018-1-AP, 1: 2000), Snail-1 Polyclonal antibody (13099-1-AP, 1: 1000), Snail-2 Polyclonal antibody (12129-1-AP, 1: 1000), and ZEB-2 Polyclonal antibody (14026-1-AP, 1: 1000) were purchased from Proteintech (wuhan, China). ZEB-1 Rabbit mAb (3396, 1: 1000), GSK3β Rabbit mAb (9315, 1: 1000), and phos-GSK3β (Ser9) Rabbit mAb (9322, 1: 1000) were purchased from CST (Boston, United States). active β-catenin (05-665, 1: 1000) was purchased from Sigma-Aldrich (St. Louris, MO, United States). Anti-β-catenin (sc-7963, 1: 1000) was purchased from Santz Cruz (CA, United States). Anti-β-actin (ab8226, 1: 1000), Goat Anti-Rabbit IgG H&L (HRP) (ab6721, 1: 10000), and Rabbit Anti-Mouse IgG H&L (HRP) (ab6728, 1: 10000) were purchased Abcam (Cambridge, United Kingdom).

All images were obtained in a Zeiss fluorescence microscope (Axio vert A1, Zeiss, Germany) or an Olympus fluorescence microscope (IX71, Olympus, Tokyo, Japan). We used ImageJ to analyze and quantify the images.

All statistical analyses were expressed as mean ± SEM using GraphPad Prism 9.0. Data were analyzed using either unpaired t-tests or one-way/two-way ANOVA, as appropriate for each experimental design, with Tukey’s test employed for multiple comparisons when ANOVA was applied. Variance homogeneity was confirmed through Brown-Forsythe testing (all p > 0.05), while normality was verified for each experimental group using Shapiro-Wilk tests (all p > 0.05). Significance was considered when P values were lower than 0.05. All experiments were done in triplicates.

To explore the underlying mechanism in the BM of NSCLC cancer, we employed RNA-seq using brain tissue, tail tissue, and the rest tissues of 265 xenografted zebrafish. As shown in Figure 1A, the data FPKM normalized by the R language was plotted as a cluster analysis of gene expression in the BM, TM, and injection site group. The gene expression of the injected site group was used as a control, and the gene expression of the BM tissues and TM tissues were processed for DEGs analysis. We used the absolute values of log2FoldChange >2 as the screening criteria to identify differential genes, which were then plotted as volcanoes (Figure 1B). Compared to gene expression at the injection site tissue, BM tissue exhibited 668 DEGs, with 349 genes upregulated and 319 genes downregulated. TM tissue showed 739 DEGs, comprising 428 upregulated genes and 311 downregulated genes. Next, we plotted the Venn map to compare the up- and downregulated DEGs in brain and TM tissues (Figure 1C). In the brain and tail metastasis tissues, we identified 177 DEGs with opposite trends. Among these, 57 DEGs were upregulated in BM tissue but downregulated in TM tissue, and 120 DEGs were downregulated in BM but upregulated in TM tissue.

The 177 DEGs were analyzed by performing GO and KEGG enrichment. GO analysis classified DEGs into three main categories, biological processes (BP) (Figure 1D), cellular component (CC) (Figure 1E), and molecular function (MF) (Figure 1F). The top 15 items in the BP, CC, and MF were plotted as an enrichment analysis circle diagram (Figure 1G). The results show that the parts of the DEGs are involved in multiple GO items. The first five GO items were analyzed using a Venn diagram (Figure 1H), and the five DEGs with strong and significant involvement were obtained, including RHOA, GNA13, MAPK1, MAPK2K2, and BMPR2. KEGG analysis revealed that 177 DEGs are mainly associated with cancer-related pathways (Figure 2A). The results of the KEGG enrichment analysis were plotted as a network diagram showing the correlation between KEGG-enriched pathways (Figure 2B). The top ten pathways with the highest correlation of KEGG enrichment were plotted into a KEGG enrichment analysis circle diagram (Figure 2C), which showed that some DEGs were involved in multiple KEGG pathways. The top five pathways were selected from the KEGG circle diagram and plotted into a Venn diagram (Figure 2D), the top 16 DEGs with the highest involvement and significance in KEGG-enriched pathways were PIK3CB, RHOA, MAPK1, SOS1, SOS2, PIK3R1, GNA13, PIK3CA, RAF1, KRAS, PIK3R3, MAP2K1, MAP2K2, PIK3CD, NRAS, and HRAS. By integrating these GO and KEGG-derived gene sets and eliminating redundancies, we ultimately identified 17 DEGs (PIK3CB, RHOA, MAPK1, SOS1, SOS2, PIK3R1, GNA13, PIK3CA, RAF1, KRAS, PIK3R3, MAP2K1, BMPR2, MAP2K2, PIK3CD, NRAS, and HRAS) associated with lung cancer. They may play an important role in the progression of brain metastasis in lung cancer.

The 31 DEGs were randomly selected from 177 DEGs and then validated by RT-qPCR. As shown in Figures 3A,B, the log2 fold change of gene expression in RT-qPCR experiments was consistent with the results in RNA-seq regarding up- and downregulation trends. Then, the 17 key DEGs that had been selected by GO and KEGG analysis were also verified by RT-qPCR. The correlation analysis demonstrates excellent concordance between RT-qPCR and RNA-seq results, with highly significant Pearson correlation coefficients of R ² = 0.9216 (P < 0.0001) and R ² = 0.9323 (P < 0.0001) for the DEGs (Figures 3C,D).

The pan-cancer analysis was performed to determine whether 17 DEGs were expressed in NSCLC tissues. The results suggested that 7 DEGs expressions were downregulated in lung cancers compared with normal tissues, including BMPR2, PIK3CD, PIK3R1, PIK3R3, RAF1, RHOA, and SOS2 (Figure 4A). It was also found that 9 DEGs (GNA13, HRAS, KRAS, NRAS, MAP2K1, MAP2K2, PIK3CA, PIK3CB, and SOS1) expressions were higher in lung cancers than in normal tissues (Figure 4A). However, the expression of MAPK1 was no different in lung cancers relative to normal tissues, indicating that MAPK1 may be irrelevant to lung cancer. Based on the UCSC Xena database (M0 = 811, M1 = 30), we examined the 16 DEGs expressions in lung cancers with/without metastasis. As shown in Figure 4B, 6 DEGs expressions are significantly different in lung cancers with or without metastasis. GNA13 was upregulated (1.17-fold change, P = 0.001) in the M1 stage lung cancer than the M0 stage lung cancer. HRAS (0.92-fold, P < 0.001), MAPK1 (0.87-fold, P = 0.01), MAP2K2 (0.94-fold, P = 0.001), PIK3RCA (0.87-fold, P = 0.01), and PIK3R1 (0.77-fold, P = 0.02) expression were low in the M1 stage lung cancer related to the M0 stage lung cancer. We combined the results of the pan-cancer database and the UCSC Xena database to compare the DEGs expression. GNA13 expression was consistent between lung cancer vs. normal tissue and metastasis lung cancer vs. non-metastasis lung cancer (Table 1). Moreover, the Kaplan-Meier analysis suggested that GNA13 high expression in lung cancer is associated with a short Median Survival Time (Figure 4C). Together, these results indicate that GNA13 was identified as a key gene associated with lung cancer metastasis and poor prognosis.

Next, we explored the important role of GNA13 in NSCLC cells. We investigated whether knockdown or overexpression of GNA13 expression using siRNA and Plasmid could affect cell proliferation, migration, and invasion, respectively. GNA13 expression was evaluated in H1299and A549 cells. We observed that GNA13 expression was significantly higher (P < 0.001) in H1299 cells than in A549 cells (Figure 5A). Hence, we used H1299 cells for knockdown and A549 cells for overexpression of GNA13 in subsequent assays. The knockdown efficiency in H1299 cells and overexpression efficiency in A549 cells of GNA13 were assessed by qRT-PCR. Si RNA-GNA13 #1 showed the highest knockdown efficiency in H1299 cells and cell viability >91% (p = 0.001) (Figure 5B). pC-DNA-GNA13#3 (Lipo: plasmid = 4: 5) showed the highest overexpression efficiency in A549 cells but exhibited cytotoxicity (Figure 5C). pC-DNA-GNA13#1 (Lipid: Plasmid = 1:1) and pC-DNA-GNA13#2 (Lipid: Plasmid = 3:4) ensured 85% cell viability and pC-DNA#2 (Lipid: Plasmid = 3:4) showed a better effect on GNA13 expression in A549 cells. Then, siRNA-GNA13#1 and pC-DNA-GNA13#2 were chosen for subsequent experiments.

We examined the effect of GNA13 on cell viability using the CCK8 assay. As shown in Figure 5D, knockdown or overexpression of GNA13 affected the viability of H1299 or A549 cells when transfected for 24 h. Knockdown of GNA13 in H1299 cells resulted in a modest but statistically significant 10.1% reduction in cell viability (P = 0.005). Conversely, overexpression of GNA13 in A549 cells led to a 3.6% increase in viability (P = 0.001). After 48 h of transfection, GNA13 knockdown inhibited the proliferation of H1299 cells, while GNA13 overexpression promoted the proliferation of A549 cells. Compared to the negative control group (H1299-NC group), GNA13 knockdown reduced the migration of H1299 cells (H1299-siGNA13) in the wound healing assay (Figures 5E,F). In A549 cells, GNA13 overexpression increased migration compared to the negative control group (A549-NC group). GNA13 knockdown or overexpression showed similar results in the trans well assay. The GNA13 knockdown group (H1299-siGNA13) exhibited fewer cell invasions, while the GNA13 overexpression group (A549-pc-DNA-GNA13) showed more cell invasions than the negative control group (Figures 5G,H). Together, those results suggested that GNA13 participated in the proliferation, migration, and invasion capabilities of NSCLC.

To further investigate the role of GNA13 in tumor distance metastasis and proliferation in vivo, we use the zebrafish xenografted model. CM-DiI labeled H1299-control, H1299-NC, H1299-siGNA13, A549-control, A549-NC, and A549-pcDNA-GNA13 cells were injected into the yolk sac of zebrafish embryos and monitored at day 1 and 4 postinjection. At 4 dpi, the H1299-NC group showed proliferation results similar to those of the H1299-control group. Compared with H1299-NC cells, the GNA13 knockdown (H1299-siGNA13) reduced tumor cell numbers (Figures 6A,B). Consistently, the A549-NC and A549-control groups showed no difference in cell proliferation. Compared to the A549-NC cells, the overexpression of GNA13 enhances the proliferation of A549-control cells (Figures 6A,B). Based on our previous study, we established the metastasis zebrafish xenograft model to investigate the effects of GNA13 on metastasis. The CM-DiI labeled H1299-control, H1299-NC, H1299-siGNA13, A549-control, A549-NC, and A549-pcDNA-GNA13 cells were injected into the perivitelline space of zebrafish embryos.

At 4 dpi, we quantified tumor cell numbers in the brains and tails of zebrafish embryos separately. Figures 6C,D showed a significant distant metastasis (brain and tail) in H1299 and A549 xenografted. The cell TM potential was 2.14 in the H1299-control group, 2.05 in the H1299-NC group, and 1.77 in the H1299-siGNA13 group. The cell TM potential was 1.96 in the A549-control group, 2.04 in the A549-NC group, and 2.23 in the A549-pcDNA-GNA13 group. The results indicated no significant difference in the cell TM potential between the H1299-control and the H1299-NC group (p = 0.75). However, the cell TM potential of the H1299-siGNA13 cells was significantly lower than that of both the H1299-control cells (p = 0.008) and the H1299-NC cells (p = 0.04). In contrast, the A549-control cells showed no significant change in cell TM potential when compared to the A549-NC cells (p = 0.59). Notably, the A549-pcDNA-GNA13 cells exhibited a significant enhancement in cell TM potential compared to the A549-control (p = 0.004) and A549-NC cells (p = 0.03). As shown in Figures 6E,F, the brain metastatic potential of the H1299-control group was 1.94, while the H1299-NC group had a potential of 1.85, and the H1299-siGNA13 group exhibited a potential of 1.07. In A549 cells, the A549-control group had a potential of 1.56, the A549-NC group had 1.47, and the A549-pcDNA-GNA13 group showed a potential of 1.87. Quantitative analysis showed a modest numerical difference in BM potential between H1299-control (1.94) and H1299-NC (1.85) cells. However, statistical analysis revealed this difference was not significant (p = 0.63), indicating comparable metastatic potential between these two groups. The brain metastatic potential of H1299-siGNA13 cells was significantly decreased compared to H1299-control (p < 0.001) and H1299-NC cells (p < 0.001). The A549-control cells showed no significant change in BM potential compared to the A549-NC cells, whereas the A549-pcDNA-GNA13 cells exhibited a significantly increased BM potential compared to both the control (p = 0.004) and NC cells (p < 0.001). The results indicated that knocking down GNA13 expression reduced the proliferation, TM, and BM of NSCLC cells in zebrafish. Conversely, overexpressing GNA13 increased the proliferation, TM, and BM of NSCLC cells in zebrafish.

To identify the pathway underlying the effect of GNA13 on proliferation and metastasis, EMT-related proteins (N-cadherin, snail-1, snail-2, zeb-1, and zeb-2) were examined by Western Blotting assay (Figure 7A). The results indicated that the expression levels of N-cadherin (Figure 7B), Snail-1 (Figure 7C), Snail-2 (Figure 7D), ZEB-1 (Figure 7E), and ZEB-2 (Figure 7F) proteins did not show significant differences between the H1299-NC cells and the H1299-control cells. However, the GNA13 knockdown significantly decreased the expression levels of N-cadherin (p = 0.02), Snail-1 (p < 0.001), Snail-2 (p < 0.001), ZEB-1 (p = 0.001), and ZEB-2 (p < 0.001) proteins in H1299-siGNA13 cells compared to H1299 control cells. In comparison to the control cells from A549, the A549-NC cells showed no significant change in the expression levels of N-cadherin (Figure 7B, p = 0.87), Snail-1 (Figure 7C, p = 0.99), Snail-2 (Figure 7D, p = 0.79), ZEB-1 (Figure 7E, p > 0.99), and ZEB-2 (Figure 7F, p > 0.99) proteins. The relative expression of N-cadherin (p < 0.001; p < 0.001), Snail-1 (p < 0.001; p < 0.001), Snail-2 (p < 0.001; p < 0.001), ZEB-1 (p < 0.001; p < 0.001), and ZEB-2 (p < 0.001; p < 0.001) proteins was significantly increased in GNA13 overexpressing (A549-pcDNA-GNA13) cells compared to both A549-control cells and A549-NC cells. Those findings indicated that downregulated GNA13 expression inhibited EMT progression, while upregulated GNA13 expression promoted EMT progression in NSCLC cells.

The effects of GNA13 expression on the levels of phos-GSK3β (Ser9), GSK3β, active β-catenin, and β-catenin proteins in the Wnt/β-catenin signaling pathway were also investigated (Figure 7A). The results showed that there were no significant changes in the expression of active β-catenin (Figure 7G, p = 0.83), β-catenin (Figure 7H, p = 0.83), phos-GSK3β (Ser9) (Figure 7I, p = 0.81), and GSK3β (Figure 7J, p = 0.77) proteins in H1299-NC cells compared to H1299-control cells. The expression levels of the above proteins in the A549-control group were similar to those in the A549-NC group. The levels of phos-GSK3β (Ser9) (p = 0.002), active β-catenin (p < 0.001), and β-catenin (p = 0.001) proteins were significantly downregulated in the H1299-siGNA13 cells. However, the total protein expression of GSK3β remained unchanged. In contrast, in cells overexpressing GNA13 (A549-pcDNA-GNA13), the expression of phos-GSK3β (Ser9) (p < 0.001), active β-catenin (p < 0.001), and β-catenin (p = 0.005) proteins was significantly upregulated, with no significant change observed in the total GSK3β (p > 0.99) protein expression. In summary, when the expression of GNA13 was decreased in lung cancer cells, the level of Ser9 phosphorylation in GSK3β also decreased, while the total protein expression levels remained unchanged. Additionally, the protein levels of both β-catenin and active β-catenin were reduced, indicating that Wnt/β-catenin signaling was inhibited. Conversely, when GNA13 was overexpressed in lung cancer cells, the expression level of GSK3β Ser9 phosphorylation increased, leading to elevated protein levels of both β-catenin and active β-catenin, which activated Wnt/β-catenin signaling.