The TCGA-lung adenocarcinoma (LUAD) (Firehose Legacy, 586 samples) and TCGA-lung squamous cell carcinoma (LUSC) (Firehose Legacy, 511 samples) reverse-phase protein array (RPPA) datasets were sourced from cBioPortal (https://www.cbioportal.org/) for Cancer Genomics. Samples with higher expression of TNIK protein as indicated by z-scores in RPPA were collected: 14 LUAD samples with high expression of TNIK, and 140 LUSC samples with high expression of TNIK. The gene expression profiles of LUAD and LUSC RNAseq (Firehose Legacy) were downloaded from xenabrowser.net (https://xenabrowser.net/datapages/). Gene set enrichment analysis (GSEA) (https://www.gsea-msigdb.org/gsea/index.jsp) was used to compare the mRNA profile of tumor tissues with high expression of TNIK in the RPPA to that of 59 adjacent normal samples.

Two tissue chips were purchased from Shanghai OUTDO Biotech: HLugA150CS03 containing 75 LUAD samples, and HLugS120CS01 containing 60 LUSC samples. Immunohistochemical (IHC) staining for TNIK (HPA0121128, 1:100 dilution, Sigma, St. Louis, MO, USA) was performed on the tissue chips, with Image-J software (version 1.48; National Institute of Health, Bethesda, MD, USA) used to measure the signal density. Integrated optical density (IOD) value/cell was calculated to quantify the IHC score. Shanghai OUTDO Biotech committees approved this study on clinical samples, and all applicable ethical regulations were followed (protocol code SHYJS-CP-1901006, approved on 1st January 2019 for HLugS120CS01; protocol code SHYJS-CP-1901005, approved on 1st January 2019 for HLugA150CS03).

A549 cells were cultured in DMEM-F12 (SH30023.01, Hyclone, Waltham, MA, USA) and PC-9 cells in DMEM (SH30243, Hyclone) containing 10% heat-inactivated fetal bovine serum (FSP500, ExCell, Uruguay) and penicillin (100 IU/mL)/streptomycin (100 mf/mL) (SV30010, Hyclone). Cells were grown at 37 °C in a water-saturated atmosphere containing 5% CO₂. Both cell lines were validated by short tandem repeat (STR) profiling and tested negative for mycoplasma.

Two shRNA sequences of the TNIK gene (shT-1#: 5^′-CCATCTCATATTCAGGGCAA-3^′; shT-2#: 5^′-GCCTCAAAGAACAACTTCTAT-3^′) and a shRNA control sequence (shCtrl: 5^′-TTCTCCGAACGTGTCACGT-3^′) were designed and constructed into a plko.1-puro lentivirus vector (#8453, Addgene, Watertown, MA, USA). As per the manufacturer’s instructions, 60%~70% of 293T cells in a 10 cm plate were transfected with either the control vector or TNIK-shRNA vector, along with Neofect™ DNA transfection reagent (TF20121201, NEOFECT, Beijing, China) to generate recombinant lentiviruses. Fresh culture medium was added to the cells 24 h after transfection. The cell supernatant was collected 48 h after transfection, concentrated using polyethylene glycol precipitation, filtered through 0.45 µm cellulose acetate filters (SLHV033RB, Merck Millipore, San Jose, CA, USA), and centrifuged for 40 minutes at 4 °C and 3500 rpm. For infection with lentivirus, A549 and PC-9 cells were first seeded at a density of 3 $\times$ 10⁵ cells per well in 6-well plates and grown for 24 h. The culture fluid was then replaced with 1900 µL of fresh culture medium together with 100 µL of lentivirus particles (control or TNIK-shRNA). After 48 h, the supernatant was replaced with fresh medium containing 1 µg/mL of puromycin (A1113806, Thermo Fisher Scientific, Waltham, MA, USA) for six days to enable stable cell line selection.

Two siRNA sequences for LIMK1 (siL-1#: 5^′-CCATGGGTGCTCTGAGCAAAT-3^′; siL-2#: 5^′-GATCGTCCTGTGCGAGATCAT-3^′) and two for ROCK2 (siR-1#: 5^′-GCACAGTTTGAGAAGCAGCTA-3^′; siR-2#: 5^′-GCCTTGCATATTGGTCTGGAT-3^′) were designed (Shanghai GenePharma, Shanghai, China). The siRNAs were transfected into the indicated cells using Lipofectamine RNAiMAX reagent (13778150, Invitrogen, Waltham, MA, USA) according to the manufacturer’s instructions. Cells were collected 48–72 h after transfection and subjected to analysis.

For the clone formation assay, cells were resuspended at a density of 200 cells/mL in a single-cell suspension and 400 cells per well were then seeded into a 6-well plate. The cells were cultured for 14 days before being fixed in absolute ethanol for 10 min, followed by staining with 0.1% crystal violet (A600331, Sangon, Shanghai, China) for 30 minutes. Colonies with more than 50 cells were counted, and the relative number of colonies was calculated for each group.

For the cell migration trajectory assay, A549 and PC-9 cells were seeded (2 $\times$ 10⁵ cells/well) in a 6-well plate. After 12 h, the cells were monitored using HoloMonitor M4 (Phase Holographic Imaging PHI AB, Skiffervägen, Lund, Sweden) and photographs were taken every 5 minutes for 20 minutes. The cell migration trajectory was analyzed with HoloMonitor software (Version 2.7.1, Phase Holographic Imaging PHI AB, Skiffervägen, Lund, Sweden).

A total of 3 $\times$ 10⁵ cells per well were seeded into a 24-well plate. After 24 h, the cells were labeled with 5-ethynyl-2^′-deoxyuridine (EDU) using the Cell-Light EdU Apollo488 In Vitro Kit in accordance with the manufacturer’s instructions (C10310-3, Ribobio, Guangzhou, China). Alternately, the cells were treated with chemotherapy drugs for 48 h prior to the EDU labeling. Finally, the cells were mounted with medium containing 4^′, 6-diamidino-2-phenylindole (DAPI, D9564, 1:1000 dilution, Sigma) and viewed with a Leica confocal microscope (Leica Microsystems GmbH, Wetzlar, Germany).

Following stable TNIK knockdown, cells were photographed using a Leica light microscope (Leica SP8 Confocal Microscope, Wetzlar, Germany) at different locations on the coverslips (40$\times$ magnification). The single cell area was analyzed using Image-J software.

Ten male BALB/c nude Mice (4–6 weeks old, 18–22 g) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). The mice were bred and housed in a specific pathogen-free (SPF) facility under controlled conditions: constant temperature (21 $\pm$ 0.5 °C), 12-hour light/dark cycle and 50% humidity. They had free access to autoclaved food and water, and were allowed a 7-day acclimation period before the start of experiments. Animal ethics approval for all experiments was obtained from the Chongqing Medical University Institutional Animal Care and Treatment Committee (Ethic Approval Number: IACUC-CQMU-2024-0593).

For tumor xenograft experiments, 6 $\times$ 10⁶ A549 cells infected with lentivirus (control shRNA or TNIK-shRNA) were subcutaneously implanted into the right axillary region of male nude mice (4 mice per group). At 35 days after injection, the mice were anesthetized with 2% isoflurane and euthanized by cervical dislocation. Tumors were harvested, and their weight and volume were measured. The tumor size (width and length) was measured using a caliper, and the volume was computed as (width² $\times$ length)/2 (expressed in mm³) and weight was recorded in milligrams (mg). Harvested tumors were immediately fixed in 10% neutral-buffered formalin for subsequent analyses. Formalin-fixed tumors were embedded in paraffin, and 4-micrometer sections were analyzed for proliferating cell nuclear antigen (PCNA) by IHC (13110, 1:500 dilution, CST, Danvers, MA, USA).

Cells with stable TNIK knockdown and control cells underwent RNA-seq analysis. Total RNA was extracted and sequencing was performed by BGI (Shenzhen, China). Differentially expressed genes (DEGs) were screened using the criteria of $\mid$log₂fold change$\mid$ $>$1, and adjusted p-value 0.05. Altered biological processes were identified using gene set enrichment analysis (GSEA).

Total RNA was isolated from lung cancer cells using trizol (15596026, Invitrogen). Next, reverse transcription was performed with 1 µg of RNA using the PrimeScript™ RT Master Mix (RR036A, Takara, Kusatsu, Shiga, Japan) in a 20-microliter reaction volume. Quantitative PCR was carried out using the SYBR Green qPCR Master Mix (B21202, Bimake, Houston, TX, USA). Gene expression was normalized to GAPDH and analyzed using the 2⁻^Δ⁢Δ⁢CT method. Primers were synthesized by BGI-Write (Beijing, China) and the primer sequences utilized in this study are presented in Supplementary Table 1.

For immunoblot experiments, the cells were lysed in RIPA buffer containing a protease inhibitor cocktail (B14001 and B15001, Bimake). The protein concentration in each lysate was measured, followed by denaturation at 100 °C for 5 min. Equal volumes of the lysates (30 µg of protein) were separated by 8% or 10% standard SDS-PAGE, and then electro-transferred onto polyvinylidene difluoride membranes (IPVH00010, Merck Millipore). The membranes were blocked with 5% non-fat dry milk in TBST at room temperature for 2 h, probed with the indicated primary antibodies at 4 °C overnight, followed by incubation with matching secondary antibodies at 37 °C for 2 h. The antibodies used for immunoblot were as follows: TNIK (HPA0121128, 1:1000 dilution, Sigma), RHOA (2117, 1:1000 dilution, CST), RHOB (2098, 1:1000 dilution, CST), ROCK1 (28999, 1:1000 dilution, CST), ROCK2 (47012, 1:1000 dilution, CST), LIMK1 (3842, 1:1000 dilution, CST), GAPDH (AB-P-R001, 1:1000 dilution, Xianzhi Bio, Hangzhou, China), vinculin (V9264, 1:2000 dilution, Sigma), p-FAK³⁹⁷ (8556, 1:1000 dilution, CST) and FAK (3285, 1:1000 dilution, CST). The membranes were washed three times with Tris Buffered Saline with Tween (TBST) (10 min per wash) between antibody incubations. For secondary antibody incubation, horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (H+L) (A0208, 1:1000 dilution, Beyotime, Shanghai, China) or goat anti-mouse IgG (H+L) (A0216, Beyotime, 1:1000 dilution) was used, depending on the species origin of the primary antibodies. The membranes were incubated with secondary antibodies at 37 °C for 2 h, followed by three additional TBST washes (10 min each). Protein bands were visualized using an enhanced chemiluminescence (ECL) detection system.

For immunofluorescence, cells were fixed and then incubated with phalloidin–TRITC (P1951, 1:1600 dilution, Sigma), vinculin (V9264, 1:200 dilution, Sigma) or tubulin (AT819, 1:200 dilution, Beyotime). They were subsequently incubated with Alexa 488/568-conjugated secondary antibodies (A32766/A32790, 1:1000 dilution, Invitrogen) and then mounted with medium containing 4^′,6-diamidino-2-phenylindole (DAPI, D9564, 1:1000 dilution, Sigma). A Leica confocal microscope was used to view the preparations.

Cells with stable knockdown of TNIK and control cells were incubated with the chemotherapy agents paclitaxel (PTX) (S1150, Selleck, Houston, TX, USA), cisplatin (also known as cis-Diaminodichloroplatinum, DDP) (S1166, Selleck) and adriamycin (ADR) (S1280, Selleck) at the dosage and duration shown in the figure legend. Cell viability was then quantified by cell counting kit-8 (CCK-8) assay (B34304, Bimake, Houston, TX, USA). The optical density (OD) at 490 nm was measured using a microplate reader (51119100, Thermo, Waltham, MA, USA), and the 50% inhibition of cell proliferation (IC50) value was calculated with GraphPad Prism software (version 9; GraphPad Software, Boston, MA, USA).

Briefly, cells with stable knockdown of TNIK and control cells were incubated with chemotherapy agents for 48 h, stained with Annexin V-FITC and propidium iodide (PI) (MA0220, meilunbio, Dalian, China) for 30 minutes in the dark, and then analyzed by FACS.

GraphPad Prism software was used for statistical analysis. Two-tailed Student’s t test was used for the analysis and comparison of data from two groups. Statistical data were presented as the mean $\pm$ SEM. A p-value of 0.05 represents statistical significance. * indicates p 0.05, ** indicates p 0.01, *** indicates p 0.001, and **** indicates p 0.0001 (p $>$ 0.05 refers to no significant difference).

To evaluate TNIK protein expression in lung cancer, we analyzed TNIK protein expression data obtained using reverse-phase protein array (RPPA) and available from The Cancer Genome Atlas (https://www.cbioportal.org). TNIK protein expression was found to be elevated in 2.3% of LUAD cases and 27.4% of LUSC samples (Fig. 1A). Next, we used IHC to evaluate TNIK protein expression in clinical samples of LUAD and LUSC. Tissue chip analysis revealed that TNIK protein expression was upregulated in both LUAD and LUSC samples compared to adjacent normal tissues (p 0.0001, Fig. 1B,C). Interestingly, TNIK protein in LUAD was mainly localized in the cytoplasm, whereas TNIK in the adjacent normal tissues and in LUSC was mainly localized in the nucleus (Fig. 1B,D). Collectively, these results imply that TNIK is a positive regulator for LUAD progression and may have distinct effects on the development and progression of LUAD and LUSC. Previous research reported that upregulation of TNIK strengthens and maintains cell viability in LUSC [20], although the role of TNIK in LUAD has so far remained unclear.

Fig. 1.

Upregulation of TNIK protein is associated with the progression of LUAD. (A) The ratio of TNIK expression in the TCGA RPPA datasets for LUAD and LUSC. (B) Representative IHC staining of TNIK protein showing subcellular localization in adjacent normal tissue (left), LUAD (upper right) and LUSC (lower right). Scale bar, 50 µm. (C,D) Statistical analysis of TNIK expression and subcellular distribution in adjacent normal and tumor tissue of 75 LUAD and 60 LUSC cases. Cytosol: TNIK is mainly localized to the cytosol; nucleoplasm: TNIK is mainly localized to the nucleoplasm; others: no TNIK is expressed in either the cytosol or nucleoplasm. **** p 0.0001. TNIK, Traf2- and Nck-interacting kinase; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; TCGA, The Cancer Genome Atlas; RPPA, reverse-phase protein arrays; IHC, immunohistochemical; A, adjacent normal tissue.

To evaluate the role of TNIK in LUAD progression, we generated cells with stable TNIK knockdown using the two LUAD cell lines A549 and PC-9. The knockdown effects were verified by qRT-PCR and immunoblotting analysis (p 0.001, Fig. 2A,B). Next, we examined whether TNIK affected LUAD cell proliferation by conducting colony formation and EDU labeling assays. The results showed that silencing of TNIK significantly inhibited the proliferation of A549 and PC-9 cells (p 0.01, Fig. 2C–F). Furthermore, cell morphology analysis revealed that TNIK knockdown increased the area of cell spread (p 0.0001, Fig. 2G,H). We then assessed the significance of TNIK on cancer cell movement. Cell migration assays showed that depletion of TNIK slowed the movement of A549 and PC-9 cells (p 0.01, Fig. 2I,J). These findings indicate that in vitro proliferation and movement of LUAD cells are enhanced by TNIK expression. To determine whether TNIK can also regulate the growth of LUAD cells in vivo, A549 cells with a stable control vector or TNIK knockdown were subcutaneously injected into nude mice to enable tumor formation. As shown in Fig. 2K–M (p 0.001), TNIK silencing greatly slowed the growth of xenograft tumors in mice compared to the control. IHC staining also confirmed that TNIK silencing inhibited PCNA expression (p 0.01, Fig. 2N,O). Together, these findings show that TNIK is necessary to maintain the LUAD cell malignant phenotype, both in vitro and in vivo.

Fig. 2.

TNIK knockdown suppresses the malignant phenotype of lung adenocarcinoma cells. (A) qRT-PCR analysis of TNIK knockdown efficacy. (B) Immunoblot analysis of the indicated proteins in control and TNIK knockdown A549 and PC-9 cells. The loading control was GAPDH. The TNIK expression level relative to GAPDH was standardized to control cells (n = 3 per group). (C,D) Representative images of control and TNIK knockdown A549 and PC-9 colonies and the relative colony number (n = 3 per group). (E,F) Representative images of control and TNIK knockdown A549 and PC-9 cells with EDU labeling, and the positive EDU count. Merge panels include DAPI (blue, nuclei). Scale bar, 20 µm. (G) The morphology of control and TNIK knockdown A549 and PC-9 cells. Scale bar, 100 µm. (H) The cell area was measured with Image J software, with 50 cells counted per condition. (I,J) Representative images of the migration trajectories of control and TNIK knockdown cells, and the calculated velocity and distance of cell migration. (K) Tumor xenografts from control and TNIK knockdown A549 cells. (L,M) The growth curve and weight of xenografts from control and TNIK knockdown A549 cells. (N,O) IHC staining of PCNA in control and TNIK knockdown A549 tissues, and the PCNA-positive cell count. Scale bar, 50 µm. ** p 0.01, *** p 0.001, **** p 0.0001. shCtrl, a shRNA control; shT-1#, a shRNA of TNIK-1#; shT-2#, a shRNA of TNIK-2#; qRT-PCR, quantitative reverse transcription polymerase chain reaction; EDU, 5-ethynyl-2^′-deoxyuridine; DAPI, 4^′,6-diamidino-2-phenylindole; PCNA, proliferating cell nuclear antigen.

IHC analysis showed that TNIK was localized mainly in the cytoplasm of LUAD cells, but also in the nucleus. This indicates a critical function for cytoplasmic TNIK in the progression of LUAD. To study the molecular mechanism underlying the TNIK-associated malignant phenotype in LUAD, RNAseq analysis was performed to ascertain changes in the gene expression profile at the whole genome level in A549 cells with stable TNIK knockdown. A total of 2543 genes were dysregulated in TNIK knockdown cells compared to control cells. GSEA revealed that TNIK depletion significantly inhibited the mitosis-related biological process (Fig. 3A). KEGG pathway analysis showed that dysregulated genes were involved in the cell cycle and in several cancer-related pathways, including p53 and the HIF-1 signaling pathway (Fig. 3B). Moreover, GSEA analysis showed that high TNIK expression detected by RPPA in TCGA LUAD samples promoted the mitosis-associated biological process (Fig. 3C). In addition, light microscopy revealed increased cell spreading in the TNIK knockdown group compared to the control group (Fig. 2G). Previous research has suggested that TNIK overexpression regulates actin assembly, thereby inducing cell spreading and causing adherent cells to round up and lose adhesion to the culture plate [13]. Cell spreading is dependent on actin filament dynamics and force generation by the FA contacts on the surface [21]. We therefore focused on the cytoskeleton-related biological process. As shown in Fig. 3A–C, the three biological processes highlighted in the red box were associated with the cytoskeleton: microtubule organization involved in mitosis, cytoskeleton-dependent cytokinesis and FA. Furthermore, we analyzed the top-ranked GSEA gene sets identified in the TNIK knockdown transcriptome of A549 cells and in LUAD samples from the TCGA database with high expression of TNIK. Other processes related to cytoskeleton organization were found, including microtubule organization and stress fiber assembly (Fig. 3D,E). The control of cell polarity and shape during cell movement and division depends on crosstalk between actin and microtubules [4], of which the RHO family of small GTPases (RHO, Rac, Cdc42) and their effectors are central players [4]. Analysis of the transcriptome profile revealed significantly lower levels of RHOA/B, ROCK 1/2 and LIMK1 in cells with TNIK silencing (Fig. 3F). In contrast, there was no difference in the level of two other small GTPases, Rac and Cdc42, between cells with TNIK knockdown and control cells (Fig. 3G). Further analysis of the RHO family and their effectors by qRT-PCR confirmed that TNIK knockdown decreased the mRNA expression of RHO/ROCK/LIMK1 (Fig. 3H,I). Moreover, Western blot analysis confirmed that RHOA/B, ROCK2 and LIMK1 protein expression were reduced upon TNIK silencing (Fig. 3J). These data strongly suggest that TNIK controls F-actin and microtubule dynamics via the RHO/ROCK2/LIMK1 signaling pathway.

Fig. 3.

TNIK regulates F-actin and microtubule cytoskeletal organization via the RHO/ROCK/LIMK1 signaling pathway. (A) The top 20 enriched gene sets identified by GSEA analysis based on the differentially expressed genes between A549 cells with stable TNIK knockdown and the control. The biological processes highlighted in the red box were associated with the cytoskeleton: microtubule organization invovled in mitosis. (B) KEGG pathway analysis based on the differentially expressed genes between A549 cells with stable TNIK knockdown and the control. The biological processes highlighted in the red box were associated with the cytoskeleton: focal adhesion. (C) Top 20 enriched gene sets identified by GSEA analysis based on the differentially expressed genes between 14 LUAD samples with high expression of TNIK detected by RPPA and 59 adjacent normal samples from the TCGA database. The biological processes highlighted in the red box were associated with the cytoskeleton: cytoskeleton dependent cytokinesis. (D) GSEA plot showing RNA-seq data from cells with stable TNIK knockdown relative to control cells. (E) GSEA plot showing TCGA data from 14 LUAD samples with high TNIK expression detected by RPPA relative to 59 adjacent normal samples from the TCGA database. (F,G) Heatmap displaying the transcript levels of indicated genes. (H,I) qRT-PCR verification of the expression of small GTPase RHO-related genes in control (shCtrl) and stable TNIK knockdown cells. Data represent the mean $\pm$ SEM. (J) Immunoblot analysis of the indicated proteins in A549 and PC-9 cells with stable TNIK knockdown or the control. “ns” indiactes p $>$ 0.05 and is used to indicate no significant difference. ** p 0.01, *** p 0.001, **** p 0.0001. GSEA, gene set enrichment analysis; KEGG, Kyoto Encyclopedia of Genes and Genomes; RHO, Ras homolog gene family; RAC, Ras-related C3 botulinum toxin substrate; ROCK, RHO-associated kinase; LIMK, LIM motif-containing protein kinase; PAK, p21-activated kinase; ARPC, actin related protein 2/3 complex.

Microtubules and actins are essential for cell division and movement [4]. Chromosome segregation during cell division relies on properly arranged microtubules and cytokinesis during the last stages of cell division and is driven by actin filaments. However, microtubules and actin filaments also provide forces to drive changes in cell shape and migration. During these processes, FAs act as a bridge between the cytoskeleton and the ECM to direct mechano-transduction, cell migration and division [8]. The assembly and disassembly of FAs depend on microtubules and actins [6]. Active pFAK-Y397 present in FAs controls several biological processes, including cell survival, proliferation, FA turnover and migration [22]. Vinculin regulates and promotes FA assembly, but FA disassembly requires the loss or inactivation of vinculin [23].

We examined the role of TNIK in the FA dynamics of LUAD cells. Phalloidin–TRITC staining was used to identify F-actin, while vinculin immunostaining was used to mark the FAs. Stable knockdown of TNIK significantly increased the number and area of FAs (p 0.01, Fig. 4A–C). Immunoblot analysis was used to further examine the expression of the FA-associated key proteins vinculin and p-FAK397. Stable knockdown of TNIK led to increased expression of vinculin and decreased expression of p-FAK397 (Fig. 4D). Additionally, TNIK knockdown was found to disturb the microtubule organization (Fig. 4E) and induce formation of syncytium (p 0.01, Fig. 4F).

Fig. 4.

TNIK regulates focal adhesion turnover and mitosis. (A) Representative images of vinculin (green) and F-actin (red) in A549 and PC-9 control and TNIK knockdown cells. Merge panels include DAPI (blue, nuclei). Scale bar, 10 µm. (B,C) Quantification of the number and average area of FAs in A549 and PC-9 control and TNIK knockdown cells. (D) Immunoblot analysis of the indicated proteins in A549 and PC-9 control and TNIK knockdown cells. (E) Immunofluorescence staining of tubulin in A549 and PC-9 control and TNIK knockdown cells. Merge panels include DAPI (blue, nuclei). Scale bar, 10 µm. (F) Statistical analysis of the syncytium ratio in the results from (E). More than 50 cells were counted per condition in each repeat. Data represent the mean $\pm$ SEM (n = 3 per group). (G) Immunoblot analysis of the indicated proteins in the A549 and PC-9 control and ROCK2 knockdown cells. (H) Immunoblot analysis of the indicated proteins in the A549 and PC-9 control and LIMK1 knockdown cells. * p 0.05, ** p 0.01, *** p 0.001, **** p 0.0001. FA, focal adhesion; FAK, focal adhesion kinase; siCtrl, a siRNA control; siR, a siRNA of ROCK; siL, a siRNA of LIMK1.

Next, we hypothesized that FA dynamics would be disrupted by inhibition of the RHO/ROCK/LIMK1 signaling pathway. TNIK knockdown was found to induce a low level of ROCK2 protein expression, but not ROCK1 (Fig. 3J). To test our hypothesis, we assessed the levels of p-FAK397 and vinculin protein in LUAD cells treated with siRNA targeting ROCK2 or LIMK1, respectively. Immunoblot analysis revealed that ROCK2 expression was silenced by siRNA, while ROCK2 knockdown enhanced the expression of vinculin and decreased the level of p-FAK397 (Fig. 4G and Supplementary Fig. 1A). Consistent with these findings, LIMK1 silencing resulted in low expression of p-FAKY397 and high expression of vinculin (Fig. 4H and Supplementary Fig. 1B). Additionally, LIMK1 knockdown was observed to increase the number and area of FAs (Supplementary Fig. 2). Taken together, these findings indicate that TNIK may control FA turnover and mitosis by affecting the dynamics of F-actin and microtubules via the RHO/ROCK2/LIMK1 signaling pathway.

Paclitaxel (PTX), cisplatin (DDP) and adriamycin (ADR) are common and conventional anti-cancer drugs. However, drug resistance is still a major reason for chemotherapy failure in cancer [24]. Cell adhesion to the ECM is one of the critical micro-environmental elements that drives cancer cell resistance. Multifunctional and multiprotein FA complexes are essential mechanical components that functionally and structurally regulate the cell shape and cytoplasmic signaling for cell survival, proliferation, differentiation and motility [7]. The effectiveness of any specific molecular therapy is challenging due to the considerable redundancy, intricacy and interaction between downstream signaling and FA-related receptors. In the present study, we found that TNIK silencing affected FA dynamics and cytoskeletal organization, including microtubules and F-actin. The targeting of TNIK in combination with chemotherapeutic agents may overcome drug resistance. CCK-8 assays revealed that TNIK knockdown increased the sensitivity of A549 and PC-9 cells to PTX, DDP and ADR (Fig. 5A–C). PTX had IC50 values of 3.044 µM and 1.4 µM/1.108 µM in control and TNIK knockdown A549 cells, respectively (Fig. 5A), and 2.025 nM and 1.058 nM/0.805 nM in control and TNIK knockdown PC-9 cells, respectively (Fig. 5A). DDP had IC50 values of 2.801 µM and 1.102 µM/1.381 µM in control and TNIK knockdown A549 cells, respectively (Fig. 5B), and 1.188 µM and 0.893 µM/0.767 µM in control and TNIK knockdown PC-9 cells, respectively (Fig. 5B). ADR had IC50 values of 0.785 µM and 0.347 µM/0.284 µM in control and TNIK knockdown A549 cells, respectively (Fig. 5C), and 0.515 µM and 0.209 µM/0.140 µM in control and TNIK knockdown PC-9 cells, respectively (Fig. 5C). To determine whether the cytotoxicity of these drugs against A549 and PC-9 cells with TNIK knockdown was related to apoptosis, we utilized flow cytometry to measure the proportion of apoptotic cells. TNIK knockdown in A549 and PC-9 cells markedly increased the rate of apoptosis in comparison to the control group (p 0.01, Fig. 5D and Supplementary Fig. 3A). Consistent with this finding, TNIK knockdown resulted in fewer EDU-positive cells (p 0.001, Fig. 5E and Supplementary Fig. 3B). Together, these results indicate that TNIK knockdown improves the tumor-suppressive efficacy of chemotherapeutic drugs.

Fig. 5.

TNIK knockdown improves the tumor-suppressive efficacy of chemotherapeutic drugs in vitro. (A–C) CCK-8 assays were used to evaluate cell viability in A549 and PC-9 cells with either stable TNIK knockdown or control vector following treatment with PTX (A), DDP (B) or ADR (C) for 48 h (n = 3 per group). (D) Cell apoptosis was analyzed by Annexin V-FITC/PI double staining of A549 and PC-9 control and stable TNIK knockdown cells after treatment with PTX (1.108 µM for A549, 0.8045 nm for PC-9), DDP (1.102 µM for A549, 0.7667 µM for PC-9), or ADR (0.284 µM for A549, 0.14 µM for PC-9) for 48 h. Results are representative of three independent experiments. (E) Cell proliferation was assessed by EDU labeling of A549 and PC-9 control and stable TNIK knockdown cells after treatment with PTX (1.108 µM for A549, 0.8045 nm for PC-9), DDP (1.102 µM for A549, 0.7667 µM for PC-9), or ADR (0.284 µM for A549, 0.14 µM for PC-9) for 48 h. ** p 0.01, *** p 0.001, **** p 0.0001. PTX, paclitaxel; DDP, cisplatin; ADR, adriamycin; CCK-8, cell counting kit-8; PI, propidium iodide.