ANKRD49 promotes the metastasis of NSCLC via activating JNK-ATF2/c-Jun-MMP-2/9 axis

Nine pairs of fresh NSCLC tissues and the corresponding adjacent normal tissues were collected from patients who underwent surgery at the First Hospital of Shanxi Medical University from October 2020 to December 2020. No patients in this study received other treatment before surgery.

The human bronchial epithelial cell line (HBEC) and NSCLC cell lines H1299, A549, H446, H460, Calu-3, H1703, sk-MES-1 were obtained from the Cell Culture Center of the Chinese Academy of Medical Sciences (Beijing, China). These NSCLC cell lines were cultured in Advanced RPMI 1640 medium (Seven, Beijing, China) supplemented with 10% fetal bovine serum (Every Green, Zhejiang, China) and incubated at 37 °C in 5% CO₂.

H1299 and H1703 cells were infected with lentivirus-5 ANKRD49 or lentivirus-3 ANKRD49-shRNA and corresponding vectors (Gene Pharma, Shanghai, China) according to the manufacturer’s protocols to construct the stable ANKRD49 overexpression (ANKRD49-OE) or ANKRD49 knockdown (ANKRD49-sh) groups. Meanwhile, the controls for overexpression (LV5) and knockdown (LV3) were also established. After infection of lentivirus, cells were incubated for more than 48 h and puromycin (2 µg/ml) was added to screen the stable cells. LV3 and ANKRD49-sh sequences were listed in Supplementary Table 1.

Puromycin (P816466) was purchased from MACKLIN (Shanghai, China). Anisomycin (GC11559) was purchased from GLPBIO (Shanghai, China). The p38 mitogen-activated protein kinase (MAPK) inhibitor (HY-10,256, SB203580) and JNK inhibitor (HY-12,041, SP600125) were purchased from MedChemExpress (New Jersey, USA). MMPs inhibitor (SF4180, ilomastat) was purchased from Beyotime Biotechnology (Shanghai, China). Rabbit anti-ANKRD49 primary antibody (Proteintech, Cat# 25034-1-AP, RRID:AB_2879860), rabbit anti-MMP-2 polyclonal antibody (Cat# bs-4599R, RRID:AB_11083963), rabbit anti-MMP-9 polyclonal antibody (Cat# bs-0397R, RRID:AB_10853038), rabbit anti-c-Jun polyclonal antibody (Cat# bs-0670R, RRID:AB_10857880), rabbit anti-p-ATF2 polyclonal antibody (Cat# bs-3033R, RRID:AB_10883830) and Alexa Fluor® 488-conjugated goat anti-rabbit IgG (Cat# bs-0295G-A488, RRID:AB_10893781) were obtained from Bioss Biotechnology (Beijing, China). The mouse anti-p-c-Jun monoclonal antibody (Cat# 558,036, RRID: AB_2249448) was obtained from BD Biosciences (New Jersey, USA). Rabbit anti-p38 monoclonal antibody (Cat# 8690, RRID: AB_10999090), rabbit anti-p-p38 monoclonal antibody (Cat# 4511, RRID: AB_2139682), rabbit anti-ERK monoclonal antibody (Cat# 4695, RRID: AB_390779), rabbit anti-p-ERK monoclonal antibody (Cat# 4376, RRID: AB_331772), rabbit anti-SAPK/JNK monoclonal antibody (Cat# 9258, RRID: AB_2141027), and rabbit anti-p-JNK monoclonal antibody (Cat# 4668, RRID: AB_823588) were purchased from Cell Signaling Technology (Danvers, MA, USA). Rabbit anti-ATF2 polyclonal antibody (Cat# BS1022, RRID: AB_1664115), rabbit anti-β-actin polyclonal antibody (Cat# AP0060, RRID: AB_2797445), rabbit anti-GAPDH polyclonal antibody (Cat# AP0063, RRID: AB_2651132), rabbit anti-Histone-H3 polyclonal antibody (Cat# BS7675), rabbit anti-PARP polyclonal antibody (Cat# BS7190), and rabbit anti-β-tubulin polyclonal antibody (Cat# AP0064, RRID: AB_2797447) were acquired from Bioworld (Minnesota, USA). Rabbit anti-ATF2 polyclonal antibodies (Cat# D155200) and rabbit anti-p-ATF2 polyclonal antibodies (Cat# D155010) were purchased from Sangon Biotech (Shanghai, China). Anti-rabbit IgG (H + L) (Cat# BA1050, RRID: AB_2904507) and anti-mouse IgG (H + L) (Cat# BA1054, RRID: AB_2734136) were purchased from Boster Biological Technology (Wuhan, China). Dylight 649 and goat anti-mouse IgG (Cat# 610-143-002, RRID: AB_11182582) were obtained from Rockland (Philadelphia, USA).

Total protein from cells or tissues was extracted using SDS lysis buffer containing 2% SDS, 10% glycerol, and 50 mM Tris-HCl (pH 6.8), and quantified using the Enhanced BCA Protein Assay Kit (Beyotime Biotechnology). Proteins (30 µg) were loaded onto a 12% SDS-PAGE gel, separated, and transferred onto polyvinylidene fluoride (PVDF) membranes. The membranes were blocked using 5% skimmed milk, followed by probing with the indicated primary antibodies at 4 °C overnight, and then incubated with an anti-rabbit/mouse HRP-conjugated IgG at room temperature for 1 h. Blots were developed using the ECL kit (Seven, Beijing, China) and detected using the ChemiDoc imaging system (Bio-Rad, California, USA).

Total RNA from NSCLC cells, nine pairs of human NSCLC tissues, and matched adjacent normal tissues were prepared using the Total RNA Extraction Kit (Promega, Madison, WI, USA) following the manufacturer’s instructions. RT-qPCR was performed using the RT-PCR Kit (Seven, Beijing, China) on the Quant Studio 3 (Thermo Fisher Scientific, Massachusetts, USA), according to the manufacturer’s procedures. All the primers are listed in Supplemental Table 2. β-actin was used as the control, and the relative mRNA expression was defined using the 2 ^{− ΔΔCt} method.

MMP-2 and MMP-9 activities were evaluated using gelatin zymography. 5 × 10⁵ cells were pretreated with serum-free medium for 48 h. The medium was collected, centrifuged at 1,500× g for 15 min at 4 °C, and mixed with non-reducing sample buffer for electrophoresis on a polyacrylamide gel containing 0.1% (w/v) gelatin. Proteolysis was assessed as a white zone in the dark blue fields. The 62 kDa and 92 kDa bands corresponded to active MMP-2 and MMP-9, respectively. The gelatinase activity of MMP-2 and MMP-9 was illustrated via the area of the clear zone in the dark blue gel and analyzed using ImageJ analysis software (National Institute of Health, Maryland, USA).

The CCK-8 assay was performed to evaluate cellular proliferation, according to the manufacturer’s protocols. Briefly, ANKRD49-OE or LV5 H1299 cells, ANKRD49-sh, or LV3 H1703 cells were cultured in 96-well plates at a density of 1 × 10³ cells per well in a 5% CO₂ incubator at 37 °C. Next, 100 µl serum-free medium containing 10 µl CCK-8 reagent (Seven, Beijing, China) was added to each well, followed by another 1 h incubation. The absorbance of the cells was measured at 450 nm using a microplate reader (Molecular Devices, California, USA).

ANKRD49-OE or LV5 H1299 cells, ANKRD49-sh, or LV3 H1703 cells were plated in six-well plates at a density of 5 × 10² cells/well and cultured at 37 °C for 1 week. After washing with PBS, the cells were fixed with methanol and stained with 0.1% crystal violet solution. The colonies were photographed and quantified using a light microscope (Eclipse Ts2; Nikon, Tokyo, Japan).

Cells (2 × 10⁵) were seeded into six-well plates and cultured for 24 h. Wounds in the confluent cells were created using a 200 µl pipette tip, followed by washing the cells with PBS to remove any floating cells. Next, complete medium was added and the wound margins were photographed. Subsequently, wound healing was photographed at 24 and 48 h using a Nikon photomicroscope (Eclipse Ts2, Nikon, Tokyo, Japan). Migratory ability was calculated by measuring the distance between different groups of cells simultaneously.

Cells (2 × 10⁴) were seeded into the upper chamber of 24-well inserts (8.0 μm, Corning, California, USA). The upper chamber was pre-coated with Matrigel (BD Biosciences, California, USA), and a medium containing 10% FBS was added to the lower chamber for the invasion assay. After incubation at 37 °C for 24 h, the remaining cells in the upper chamber were scraped off and the cells on the lower side of the chamber were fixed, stained, and viewed using a Nikon photomicroscope (Eclipse Ts2). Data were analyzed by counting at least five random fields.

ANKRD49-OE or LV5 H1299 cells were plated in 10 cm plates at a density of 5 × 10⁶ cells/plate and cultured at 37 °C for 2 days. The subcellular proteins in ANKRD49-OE or LV5 H1299 cells were prepared using a subcellular structure cytoplasm and nucleus extraction kit (Bioss Biotechnology). In brief, samples were centrifuged at 750 × g for 5 min at 4 °C. The supernatant was collected and centrifuged at 3400, × g, 10 min (cytoplasmic fraction). This process was repeated four times to remove nuclear contamination. The pellet was washed with fractionation buffer and centrifuged at 750 × g, 5 min to remove cytoplasmic contamination at least three times. The nuclear/membrane fraction was dissolved in RIPA buffer. The protein samples (20 µg) were subjected to Western blot with p-ATF2 or p-c-Jun antibodies.

H1299 cells were transfected with pMSCVpuro-ANKRD49 or the vector for 48 h, and 5 × 10⁴ cells were seeded onto sterile coverslips in a twelve-well plate for another 24 h. The medium was removed, and cells were fixed with 4% paraformaldehyde and then permeabilized with 0.01% Triton X-100 at room temperature for 10 min. Cells were incubated with anti-p-ATF2 antibody or anti-p-c-Jun antibody at 4 °C overnight, and then probed with Alexa Fluor® 488-conjugated goat anti-rabbit IgG or Dylight 649 goat anti-mouse IgG at room temperature for 1 h in darkness. Nuclei were stained with DAPI (Beyotime) for 5 min, and fluorescence signals were observed using a fluorescence microscope (Eclipse Ts2).

Nucleus proteins of the ANKRD49-OE or LV5 H1299 cells (1 × 10⁶) were extracted and 25 µg of them were precleared using Protein A/G Agarose (Santa Cruz, California, USA). Subsequently, 2 µl rabbit anti-ATF2 or normal rabbit IgG was added to the lysates and incubated at 4 °C for 12 h. Next, 20 µl of Protein A/G PLUS Agarose was mixed into the lysates and incubated at 4 °C overnight. Agarose was washed three times using wash buffer, and the bound proteins were separated by SDS-PAGE.

The chromatin immunoprecipitation (CHIP) assay was conducted in accordance with the manufacturer’s protocols (Beyotime Biotechnology). The binding site of ATF2/c-Jun in the MMP-2 or MMP-9 promoter region and the position of the CHIP primers are illustrated in Supplementary Fig. S1. Chromatin solutions were sonicated and incubated with anti-p-ATF2, anti-p-c-Jun, or control IgG at 4 °C for 12 h. DNA-protein cross-links were reversed, and chromatin DNA was subsequently purified and subjected to PCR analysis. Primers for amplifying the MMP-2 or MMP-9 promoter, which contains the binding site of ATF2/c-Jun, are listed in Supplementary Table 3. The immunoprecipitated DNA was analyzed by qPCR. The qPCR products were resolved on a 1.2% agarose gel and visualized by nucleic acid dye staining.

Ten female BALB/c nude mice and ten male BALB/c nude mice (5~6-week-old) were obtained from the Experimental Animal Center of Shanxi Medical University. The mice were housed under SPF conditions with a 12 h light-dark cycle and ad libitum access to food and water at 23 °C with 60% humidity. Female mice were randomly allocated to either LV5 (n = 5) or ANKRD49-OE H1299 cells (n = 5). Male mice were randomly allocated to either LV3 (n = 5) or ANKRD49-sh H1703 cells (n = 5). One week of acclimation, 4 × 10⁶ cells in 200 µl saline were injected into the lateral tail veins of the nude mice. After the injection, the health of the mice was monitored daily. After 4 weeks, the mice were sacrificed, and whole lungs were collected to count the metastatic nodules on the surface. Lung tissues were fixed in 10% paraformaldehyde for further hematoxylin-eosin (H&E) staining and IHC analysis. All animal experiments were approved by the Animal Ethics Committee of the Shanxi Medical University (approval number SYDL-2020016) and was performed according to the procedures that complied with the ARRIVE guidelines.

Five micrometer-thick sections of formalin-fixed paraffin-embedded lung tissues from nude mice were prepared for IHC staining. Slides were deparaffinized, rehydrated, and microwaved for antigen retrieval in 0.01 M, PH 6.0 sodium citrate buffer. Endogenous peroxidase activity was blocked using 3% H₂O₂. Subsequently, sections were incubated with the corresponding primary antibody (1:100) for 45 min at room temperature. Diaminobenzidine (DAB) (Boster, Wuhan, China) was used to visualize peroxidase activity. The immunostaining intensity of ANKRD49, NAPSA, NKX2-1, MMP-2, MMP-9, p63, p-JNK, p-c-Jun, and p-ATF2 was calculated according to the published H-score method: H-score value = (unstained tumor cells) % × 0 + (weakly stained tumor cells) % × 1 + (moderately stained tumor cells) % × 2 + (strongly stained tumor cells) % × 3.

Correlations between the level of ANKRD49 and the levels of MMP-2, MMP-9, p-JNK, p-ATF2, and p-c-Jun were analyzed using Pearson’s correlation coefficient. A strong correlation was defined as an r-value of ≥ 0.5. SPSS18.0 (IBM, Inc) and GraphPad Prism7.0 (GraphPad Software Inc.) software were used to analyze the data. The analyses were repeated in triplicate. Quantitative data are expressed as the mean ± standard deviation (SD) and analyzed using a Student’s t‑test or one‑way analysis of variance followed by Tukey’s post hoc test using GraphPad Prism software. A two‑sided P‑value < 0.05 was considered to indicate a statistically significant difference.

To investigate the expression pattern of ANKRD49 in fresh tissues from NSCLC, nine fresh tumor tissues and adjacent normal tissues (6 cases of LUAD and 3 cases of LUSC) were collected and the ANKRD49 mRNA was measured using RT-qPCR. The data showed that the expression of ANKRD49 mRNA in tumor tissues was significantly higher than that in adjacent normal tissues (Fig. 1A). Next, the expression of ANKRD49 mRNA in human bronchial epithelial cell line (HBEC), as well as in five LUAD lines (H1299, A549, H446, H460, Calu-3) and two LUSC lines (H1703, sk-MES-1) was assessed by RT-qPCR. The results demonstrated that the expression of ANKRD49 was obviously higher in H1299, A549, H446, H1703, and sk-MES-1 cells than in the human bronchial epithelial cell line (Fig. 1B). Accordingly, H1299 cell line was selected for ANKRD49 overexpression assay, while H1703 cell line was opted for ANKRD49 knockdown analysis.

To investigate the function of ANKRD49 in LUAD cells, we established stable ANKRD49-OE or ANKRD49-sh H1299 cells, as well as their respective control cells LV5 and LV3. RT-qPCR and Western blot assays showed that ANKRD49-OE and ANKRD49-sh H1299 cells were constructed (Fig. 2A, B). CCK-8 and colony formation assays revealed that ANKRD49 had no effect on H1299 cell proliferation (Supplementary Fig. S2). Wound healing, transwell migration, and invasion assays illustrated that ANKRD49-OE markedly enhanced the migration and invasion of H1299 cells compared to the LV5 group, and the opposite results were observed in the ANKRD49-sh cells compared to the LV3 group (Fig. 2C-F). Herein, we wanted to know whether endogenous levels of ANKRD49 correlated with differential endogenous migration/invasion potential, wound healing, transwell migration and invasion assays were performed using H1703 cells (higher expression of ANKRD49) and H1299 cells (lower expression of ANKRD49). The results exhibited that H1703 cells had the higher capability in cellular migration and invasion than those in H1299 cells (Supplementary Fig. S3). Taken together, these data demonstrated that ANKRD49 accelerated the migration and invasion of H1299 and 1703 cells.

The high expression of MMP-2 and MMP-9 in cancerous tissues has been correlated with tumor cell matrix degradation, invasion, and migration [12‐14]. To uncover the molecular basis of ANKRD49 on invasion and migration, the expression of MMP-2 and MMP-9 was tested when ANKRD49 was overexpressed or downregulated by RT-qPCR and Western blot assays. The results revealed that MMP-2 and MMP-9 levels were elevated in ANKRD49-OE and attenuated in ANKRD49-sh H1299 cells compared with the LV5 and LV3 groups, respectively (Fig. 3A, B). Next, the gelatinase activities of MMP-2 and MMP-9 were markedly boosted in ANKRD49-OE H1299 cells and declined in ANKRD49-sh H1299 cells, compared with those in the LV5 and LV3 groups, respectively (Fig. 3C, D). In order to confirm whether the effect of ANKRD49 was dependent on MMP-2 and MMP-9, ANKRD49-OE H1299 cells were pretreated with 10 µM ilomastat (MMP inhibitor) for 1 h. and the wound healing assay displayed that inhibition of MMP-2/MMP-9 attenuated the migration and invasion of H1299 cells (Fig. 3E, F).

It is well established that mitogen-activated protein kinases (MAPKs), including p38 MAPK, ERK, and JNK, are involved in tumor metastasis and invasion [15‐17]. As illustrated in Fig. 4A, the levels of p-p38 and p-JNK were obviously augmented in ANKRD49-OE and decreased in ANKRD49-sh H1299 cells compared with the LV5 and LV3 groups, respectively, while the levels of p-ERK were not altered. Next, p38 MAPK inhibitor (10µM SB203580), JNK inhibitor (10µM SP600125), or DMSO (10 µL) was used to pretreat ANKRD49-OE H1299 cells for 1 h. We found that SP600125 significantly decreased MMP-2/MMP-9 levels compared to DMSO-treated cells, whereas SB203580 had no effect on MMP-2/MMP-9 expression (Fig. 4B, C). Besides, the wound healing assay also exhibited that inhibition of JNK abated the migration and invasion of ANKRD49-OE H1299 cells, whereas inhibition of p38 had no effect on migration and invasion of ANKRD49-OE H1299 cells (Fig. 4D, E). To explore the specific role of JNK on MMP-2/MMP-9 expression, p-ATF2 and p-c-Jun, which are regulatory targets of JNK [18], were analyzed by Western blot. As shown in Fig. 4F, the levels of p-ATF2 and p-c-Jun were both boosted in ANKRD49-OE and weakened in ANKRD49-sh H1299 cells compared with the LV5 and LV3 groups, respectively. Since ATF2 also has been reported to be regulated by p38 MAPK [19], in order to clarify whether p38 MAPK was involved in the regulation of ATF2 in ANKRD49-OE settings, SP600125 or SB203580 was used and p-ATF2 was analyzed. The results illustrated that SP600125 decreased the level of p-ATF2, but SB203580 did not (Fig. 4G, H), indicating that p38 MAPK didn’t participate in the regulation of ATF2 in ANKRD49-OE H1299 cells. Overall, these data demonstrate that ANKRD49 promotes the invasion and migration of H1299 cells by enhancing MMP-2/MMP-9 expression mediated by the JNK but not p38 MAPK pathway.

ATF2 and c-Jun belong to the AP-1 family [20], translocate into the nucleus, and regulate the expression of target genes, including MMPs [21, 22]. Accordingly, the nuclear distribution of p-ATF2 and p-c-Jun was measured. The results showed that the nuclear levels of ATF2 and c-Jun as well as their phosphorylation were significantly enhanced in ANKRD49-OE H1299 cells compared to those in the LV5 group (Fig. 5A, B). It is widely acknowledged that ATF2 contains a nuclear export signal (NES) in its leucine zipper region and two nuclear localization signals (NLS) in its basic region, leading to continuous shuttling between the nucleus and the cytoplasm [23]. It is essential for ATF2’s transcriptional activation that ATF2 dimerizes with c-Jun in the nucleus which prevents the export of ATF2 [24]. Co-immunoprecipitation assay revealed that ANKRD49 promoted the interaction between p-ATF2 and p-c-Jun in the nucleus (Fig. 5C). In addition, immunofluorescence assay also showed that ANKRD49 promoted the nuclear co-localization of p-ATF2 and p-c-Jun (Fig. 5D). Furthermore, chromatin immunoprecipitation assay was conducted to analyze the binding ability of ATF2 or c-Jun to the MMP-2 or MMP-9 promoter region. Our results illustrated that ANKRD49 enhanced the binding of ATF2 or c-Jun to the MMP-2 or MMP-9 promoter region (Fig. 5E-G). Taken together, these findings suggest that ANKRD49 activates ATF2/c-Jun transcription factor as a heterodimer to regulate MMP-2/MMP-9 expression in H1299 cells.

Furthermore, the function of ANKRD49 in LUSC was investigated by establishing stable ANKRD49 knockdown (ANKRD49-sh) H1703 cells and a corresponding control LV3. RT-qPCR and Western blot assays demonstrated that ANKRD49-sh H1703 cells were successfully prepared (Fig. 6A, B). CCK-8 and colony formation assays revealed that ANKRD49 had no effect on H1703 cells’ proliferation (Supplementary Fig. S4). Wound healing, transwell migration, and invasion tests exhibited that ANKRD49-sh dramatically attenuated the migration and invasion of H1703 cells compared to the LV3 group (Fig. 6C-F). Additionally, we found that downregulation of ANKRD49 diminished both the expression and activity of MMP-2 and MMP-9 (Fig. 7A-D). Similarly, downregulation of ANKRD49 reduced the levels of p-JNK, p-p38, p-ATF2, and p-c-Jun in H1703 cells (Fig. 7E, F). To further confirm whether the effect of ANKRD49 on the MMP-2/MMP-9 levels in H1703 cells was dependent on JNK pathway just like it’s function in H1299cells, Anisomycin (20µM), an activator of JNK and p38 MAPK [25] was used to pretreat ANKRD49-sh H1703 cells for 1 h, followed by administration with SB203580, SP600125 or DMSO for another 24 h. We found that Anisomycin significantly augmented MMP-2/MMP-9 levels in SB203580-treated cells, whereas Anisomycin had no effect on MMP-2/MMP-9 expression in SP600125-treated cells (Fig. 7G, H). These data suggest that ANKRD49 promotes the invasion and migration of H1703 cells by enhancing MMP-2/MMP-9 expression mediated by the JNK rather than p38 MAPK pathway.

To further explore the effect of ANKRD49 on the migration of NSCLC cells, ANKRD49-OE or LV5 H1299 cells and ANKRD49-sh or LV3 H1703 cells were injected intravenously into the tail of nude mice and maintained for 4 weeks. The results appeared that ANKRD49-OE enhanced the incidence of lung metastasis and the number of metastatic foci (Fig. 8A, B). Downregulation of ANKRD49 substantially reduced the incidence of lung metastasis and number of metastatic foci (Fig. 9A, B). Images of the mouse lungs are shown in Supplementary Fig. S5. HE staining also illustrated that more tumor cells appeared in perivascular and peribronchial regions in lung tissues from mice injected with ANKRD49-OE H1299 cells than in control mice (Fig. 8C). Nevertheless, the distribution of tumor cells in the lung tissues of mice injected with ANKRD49-sh or LV3 showed the opposite trend (Fig. 9C).

NAPSA and NKX2-1 are markers for LUAD [26‐28], whereas p63 is a marker for LUSC [29]. Immunohistochemical staining illustrated that the distribution of ANKRD49-positive cells as well as NAPSA-positive, NKX2-1-positive, or p63-positive cells was consistent with the results of HE staining (Figs. 8D and 9D). In addition, immunohistochemical staining showed that the levels of ANKRD49, MMP-2, MMP-9, p-JNK, p-c-Jun, and p-ATF2 were elevated in lung tissues from mice injected with ANKRD49-OE H1299 cells (Fig. 8E-G), whereas their levels were decreased in lung tissues of mice injected with ANKRD49-sh H1703 cells (Fig. 9E-G), compared to control mice. Correlation analysis displayed that the metastatic rates of the tumor cells were positively correlated with the expression of ANKRD49, MMP-2, MMP-9, p-JNK, p-c-Jun or p-ATF2 (Supplementary Fig. S6). In addition, MMP-2, MMP-9, p-JNK, p-ATF2, and p-c-Jun levels were positively correlated with the levels of ANKRD49 (Supplementary Fig. S7). In summary, ANKRD49 accelerated the invasion and metastasis of NSCLC cells via JNK-mediated transcription activation of c-Jun and ATF2 which regulated the expression of MMP-2/MMP-9.