Background

Non-small cell lung cancer (NSCLC) represents a grave concern for public health, leading the spectrum of malignancies in both incidence and mortality [1, 2]. Due to the significant advancements in targeted therapy and immunotherapy in recent years, the survival of NSCLC patients has improved [3]. Nevertheless, a substantial subset of patients is diagnosed at an advanced stage, with metastases already present [4]. The metastasis of lung cancer impacts the quality of life of patients, and results in the leading cause of death [5, 6]. Consequently, an in-depth exploration of the molecular mechanisms underlying NSCLC metastasis is imperative. This research is pivotal for identifying novel therapeutic targets and developing innovative pharmacological strategies, thereby enhancing treatment outcomes for patients with NSCLC.

Cytoskeletal remodeling is a pivotal process in cancer cell metastasis, with small Rho GTPases acting as key regulators. Among these, Rac, cell division cycle protein 42 (CDC42), and Rho are extensively characterized for their distinct roles in cytoskeletal dynamics. Rac activation is associated with the generation of lamellipodia and membrane ruffles, CDC42 activation promotes filopodia assembly, and Rho activation is linked to the formation of stress fibers [7, 8]. These small Rho GTPases have garnered attention as critical targets in anticancer therapy due to their central roles in modulating cytoskeletal dynamics, which is fundamental to cancer metastasis. They facilitate metastatic spread by not only driving the migratory and invasive capabilities of cancer cells [9,10,11] but also by augmenting the migration of endothelial cells, a process essential for angiogenesis [12,13,14].

Rho GTPase proteins exhibit activity when localized at the plasma membrane or endomembrane (e.g., endosomes, Golgi) [15]. The intracellular trafficking of these proteins plays a crucial role in their cellular functions by influencing their subcellular distribution. For instance, Arf6-mediated membrane trafficking can modulate the localization and activation of CDC42, thereby impacting CDC42-driven cell migration [16]. Additionally, Rab5, a key regulator of endocytosis, has been shown to activate Rac and promote lamellipodia formation by facilitating the trafficking of Rac to the cell membrane [17]. Furthermore, Rab5 has been implicated in integrin trafficking [18], cytoskeleton remodeling [19], and cell metastasis [20,21,22]. Notably, the downregulation of Rab5 has been observed to inhibit the activity of small Rho GTPases, leading to a suppression of filopodia and lamellipodia formation [23]. Currently, these findings revealed membrane trafficking regulatory proteins modulate small Rho GTPase activity via intercellular trafficking pathways. Generally, small Rho GTPase activity can be directly regulated by GTPase Activating Protein (GAP) and Guanine Exchange Factor (GEF), it is unknown whether membrane trafficking regulatory proteins can directly interact with and regulate small Rho GTPase activity.

Rabenosyn-5 (Rbsn) serves as an effector for the small Rab GTPases Rab4 and Rab5, orchestrating their function as elucidated in prior research [24,25,26]. The role of Rbsn in the endocytic pathway is well-established, where it governs the trafficking of a plethora of internalized receptors, including but not limited to the transferrin receptor, major histocompatibility complex class I, and the epidermal growth factor receptor (EGFR) [26,27,28]. Beyond receptor trafficking, Rbsn is implicated in the preservation of microtubule polarity and modulating the subcellular distribution of intracellular RNA [29]. In Drosophila, Rbsn is indispensable for the biogenesis of early endosomes in the context of vesicular fusion, as evidenced by one investigation [30]. Disruption of Rbsn function leads to compromised epithelial polarity and aberrant tissue proliferation, culminating in tumorigenesis [30]. Despite these insights, the involvement of Rbsn in tumor metastasis and its capacity to modulate small Rho GTPase activity remain enigmatic. There is a pressing need for further inquiry to unravel the role of Rbsn in these critical processes and to ascertain its potential as a modulatory agent of small Rho GTPase-mediated signaling.

In this study, we observed a downregulation of Rbsn in NSCLC compared to normal lung tissue and demonstrated its ability to suppress filopodia formation and lung cancer metastasis by inhibiting the GTPase activity of CDC42. Interestingly, we found that this process is regulated by the phosphorylation of the Rbsn Thr253 site, mediated by Akt. Therefore, our findings reveal a novel mechanism by which Rbsn regulates cell migration and invasion through function as a GAP for small Rho GTPases, thereby impacting lung cancer metastasis.

Methods

Plasmid construction and antibodies

The retrovirus-based Flag-Rbsn expressing plasmid was generated by flanking PCR fragment of human Rbsn cDNA and synthesizing 3xFlag sequence, then inserted into a pRetroQ-AcGFP-C1 vector (Clontech, San Jose, CA, USA) using BamHI and EcoRI, and NheI and BglII respectively. The retrovirus-based SF (StrepII-Flag)-Rbsn expressing plasmid for tandem affinity purification (TAP) experiment was generated by flanking PCR fragment of human Rbsn cDNA, then inserted into pMXs-SF (StrepII-Flag) vector using EcoRI and NotI. Akt1 was cloned from 293 T cell cDNA via PCR and inserted into the pLVx-HA vector. To simulate the kinase active state (DA), the S473D mutation (Akt1 S473D forward: 5′-ccgctggccgagtagtcgaactgggggaagtg-3′; reverse: 5′-cacttcccccagttcgactactcggccagcgg-3′), and the kinase-dead mutant K179M (Akt1 K179M forward: 5′-cttccttcttgaggatcatcatggcgtagtagcgg-3′; reverse: 5′-ccgctactacgccatgatgatcctcaagaaggaag-3′) were introduced. The primers corresponding to these mutations were used to amplify pLVx-HA-Akt1. The pLVx-HA-Akt1 plasmid was then digested with DpnI enzyme to remove the template plasmid. Subsequently, the constructs were transformed into Bl21-competent cells for plasmid extraction.

The following antibodies were employed: anti-Flag (clone M2, Sigma, St. Louis, Missouri, USA), anti-Rbsn (sc-82729, Santa Cruz Biotechnology, Dallas, Texas, USA), anti-β-actin (clone AC-74, Sigma), anti-CDC42 (ACD03, Cytoskeleton, Denver, CO, USA), anti-Rac1 (ARC03, Cytoskeleton), anti-Rho (ARH04, Cytoskeleton), anti-Akt (CST#9272, Cell Signaling Technology, Denver, CO, USA), anti-phospho-Akt (Ser473) (CST #4060, Cell Signaling Technology), and anti-phospho Akt substrate (RXXS*/T*) (CST #9614, Cell Signaling Technology).

Cell culture and establish stable cell lines by retrovirus infection

A549, H1299, and 293T cells used in this study were obtained from the American Type Culture Collection (ATCC). BEAS-2B cells used in this study were obtained from the National Collection of Authenticated Cell Cultures (Shanghai, China). A549 cell line was cultured in Roswell Park Memorial Institute (RPMI) 1640 (RPMI 1640, Gibco, Carlsbad, CA, USA). H1299, BEAS-2B and 293T cell lines were cultured in Dulbecco’s Modified Eagle Medium (DMEM, Gibco) supplemented with 10% fetal bovine serum (FBS, Sigma-Aldrich) and 1% penicillin-streptomycin (Gibco) in a humidified atmosphere containing 5% CO2 at 37 °C.

The retrovirus plasmids (24 μg) and 4 μg of vesicular stomatitis virus G protein (pVSV-G) vector (Clontech Laboratories, Palo Alto, CA, USA) were introduced into GP2-293 packaging cells (Clontech Laboratories). After 24 h, the medium was replaced with 5 mL of fresh DMEM supplemented with 10% FBS. The supernatants containing the viral particles were collected 48 h post-transfection and subsequently used to infect the target cells. The infected cells were then subjected to selection with 2 μg/mL puromycin.

For the transient transfection of 293T cells, the cells were seeded into 6 cm dishes and allowed to adhere and grow until they reached the logarithmic phase of growth. The plasmids were mixed with the transfection reagent Lipo8000 (C0533, Beyotime Biotechnology, Shanghai, China) at a mass-to-volume ratio of 1:2 in 200 μl of basic DMEM medium, and this mixture was then added dropwise to the cell culture medium. Protein was extracted 48 h later.

Co-immunoprecipitation and Western blotting analysis

Cells were lysed in immunoprecipitation lysis buffer (20 mM Tris-HCl, 120 mM NaCl, 0.8% Triton X100, 1 mM EDTA, pH 7.4) supplemented with protease inhibitors Cocktail (Selleck, Houston, TX, USA). The supernatants were then incubated with 1 μg of primary antibody or control IgG at 4 °C overnight. Following this, 20 μL of protein A/G sepharose beads (Thermo Fisher Scientific, Waltham, MA, USA) were added and incubated at 4 °C for 3 h. The beads were subsequently washed with immunoprecipitation lysis buffer, and the protein complex was eluted using 100 μL of SDS sample buffer.

To investigate the interaction between Rbsn and CDC42, cells were lysed with GTPase immunoprecipitation buffer (20 mM Tris-HCl, 120 mM NaCl, 1 mM EDTA, 0.8% Triton X-100, 5 mM MgCl2, pH 7.4). All of the Mg2+ in each 500 μL of cell lysate were chelated with 7.5 μL of 0.5 M EDTA (pH 8.0). Subsequently, 5 μL of 0.5 M Guanosine-5’-(γ-thio)-triphosphate (GTPγS, a non-hydrolysable GTP analog which can create constitutively GTP bound GTPases) or GDP (final concentration, 0.5 mM) was added to the lysate and incubated for 15 min at room temperature to load GTPγS or GDP. Following this, 30 μL of 1 M MgCl2 was added to the lysate to stabilize the binding of the nucleotide to the dynamin GTPase. The primary antibody was then added and incubated for 3 h, followed by the addition of protein A/G sepharose beads for another 3 h. The protein complex was eluted using 100 μL of SDS sample buffer.

Western blotting was performed as previously described [31]. Briefly, the protein sample was prepared and separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). After the separation, the protein was transferred from the gel to polyvinylidene difluoride (PVDF) membranes (GE Healthcare, Marlborough, MA, USA). The membrane was then blocked with 5% dry milk in phosphate-buffered saline (PBS) for 1 h at room temperature and was then incubated with the indicated primary antibody at 4 °C overnight. After 3 times washing, the membrane was incubated with horseradish peroxidase-conjugated secondary antibodies (Jackson ImmunoResearch, West Grove, PA, USA) at room temperature for 1 h, followed by 3 times PBS washing. After washing, the ECL reagent (GE Healthcare) was added to the membrane and the images were taken.

Small Rho GTPase activity assay

The activation of small Rho GTPase activity was performed with the RhoA/Rac1/CDC42 activation assay combo biochem kit (Cytoskeleton) according to the manufacturer’s instructions. Briefly, 40 µL of the specific affinity beads targeting RhoA, Rac1, and CDC42 were added to the 1 mL of cell lysate in assay lysis buffer and were incubated at 4 °C for 1 h. The active form of small Rho GTPase was pulled down in the bead pellet by centrifugation for 10 s at 14,000 x g. The supernatant was discarded, and the bead pellet was washed 3 times with 0.5 mL of assay buffer. After the last wash, the bead pellet was mixed with 40 µL of SDS-PAGE sample buffer and was boiled for 5 min. After the centrifugation, the supernatant was used to perform Western blotting with the specific antibody targeting RhoA, Rac1, or CDC42.

To assess the GTPase activity of CDC42, a malachite green-based colorimetric assay was conducted following previously described methods [32]. Briefly, the purified proteins were mixed as the indicated molar ratio in a diluting buffer for 30 min. Then the GTPase assay buffer (12.5 mM NaCl, 2.5 mM MgCl2, 0.0625% Tween-80, 1.25 μg/mL of leupeptin, 125 μM PMSF, 375 μM GTP, 15 mM Tris–HCl, pH 7.4) was added to 96-well plates, followed with the addition of the protein mixture. The samples were incubated at 30 °C with continuous shaking at a rate of 300 rpm, and the reaction was terminated by adding 0.5 M EDTA at the indicated time. The samples were treated with 150 μL of malachite green reagent and developed the color. The absorption spectra at 650 nm were measured.

Tandem affinity purification (TAP) of protein complex

To purify Rbsn-interacting proteins, H1299 cells stably expressing SF (StrepII-Flag dual tag) -Flag or SF-Rbsn were established after the infection with indicated plasmid and were selected with 2 μg/mL puromycin. To extract the proteins, 5x15 cm dishes of each cell were lysed using immunoprecipitation lysis buffer (20 mM Tris-HCl, 120 mM NaCl, 1 mM EDTA, 0.8% Triton X-100, 5 mM MgCl2, pH 7.4) and then incubated with 200 μl of Strep-Tactin superflow resin from IBA LifeSciences (Louvain-La-Neuve, Belgium) for 6 h at 4 °C. After washing with lysis buffer 3 times, the protein complex was eluted by employing 1 mL of elution buffer#1 (Tris-buffered saline (TBS) buffer containing 2 mM Desthiobiotin from IBA). Following this, the eluate was subjected to the second round of purification by incubating with 100 μl anti-Flag resin overnight at 4 °C. After washing with lysis buffer 3 times, the protein complex was subsequently eluted using elution buffer#2 (TBS buffer containing 200 μg/mL of Flag peptide from Sigma). Then 20 μl samples were loaded for SDS-PAGE and visualized by silver staining using The Pierce Silver Stain Kit (#24612, Thermo Fisher Scientific).

Identify Rbsn-interacting proteins by mass spectrometry analysis

For the identification of proteins included in the tandem affinity purification eluate, the whole eluates were digested with Trypsin Gold (Promega, Madison, WI) for 16 h at 37 °C after reduction, alkylation, demineralization and concentration, followed by analysis on the Orbitrap Fusion mass spectrometer (Thermo Fisher Scientific). The protein–protein interaction network was generated by STRING (https://cn.string-db.org/).

CCK8 assay

Cells were seeded in 96-well plates at a density of 2000 cells per well and allowed to adhere overnight. The medium was then removed, and 20 µl of Cell Counting Kit-8 (CCK-8) solution along with 200 µl of medium was added to each well. The plates were incubated for an additional 1 h at 37 °C. Absorbance at 450 nm was measured using a microplate reader to determine cell viability. CCK-8 measurements were taken every 24 h. Each treatment was performed in triplicate, and data were normalized to control wells containing medium without the test compound. Results were expressed as mean ± SD. Statistical significance was assessed using Student’s t-test.

Wound healing assay

A549 or H1299 cells were seeded in 6-well plates at a density of 5×105cells per well and allowed to adhere for 24 h to achieve a confluent monolayer. A wound was introduced to the monolayer using a 200 µl pipette tip, ensuring consistent wound width. Following scratch application, wells were gently washed twice with PBS to remove cellular debris and floating cells. The cells were then incubated in basic RPMI 1640 or DMEM to minimize proliferation. Images of the same field were taken immediately after wounding and again after 24 h using a microscope. The wound closure ratio was calculated using the formula: (Wound area at 0 h–Wound area at 24 h)/Wound area at 0 h * 100%. The experiment was repeated three times and data were presented as mean ± SD. Statistical significance was assessed using Student’s t test.

Transwell assay

For invasion assays, 8.0 µm pore size Transwell inserts (Corning, NY, USA) were coated with Matrigel (diluted to 1 mg/mL, incubated overnight at 4 °C) to mimic the extracellular matrix; no coating was used for migration assays. Cells were seeded at a density of 5 × 104 cells/mL in the basic medium into the upper chamber, with a medium containing 20% FBS in the lower chamber. After incubating for 24 h, cells that migrated or invaded the lower side of the membrane were fixed with 4% formaldehyde for 20 min and stained with 0.1% crystal violet for 30 min. Images of the stained cells were captured using a Leica microscope, and the cells were quantified using ImageJ software. The experiment was repeated three times and data were presented as mean ± SD. Statistical significance was assessed using Student’s t-test.

Immunofluorescence staining

The cells were washed with PBS 3 times and then were fixed with 4% paraformaldehyde for 15 min at room temperature. The cells were incubated at 0.1% Triton X-100 in PBS at room temperature for 10 min. After washing with PBS 3 times, the cells were incubated with the indicated primary antibody overnight at 4 °C, followed by washing with PBS 3 times. Then the cells were incubated with the appropriate fluorochrome-conjugated secondary antibody (Invitrogen, Carlsbad, CA, USA) for 1 h at room temperature. After washing with PBS, the cells were incubated with DAPI for 10 min at room temperature and were mounted with Fluoroshield (sigma). Images were captured using a Leica SP8-DMIL confocal microscope with a 63 x oil objective lens. For each group, 100 cells were randomly selected and assessed for the presence of prominent filopodia. Each experiment was repeated three times. Statistical analysis was performed using the Student’s t-test.

Immunohistochemistry

Tissues were fixed in 4% paraformaldehyde, dehydrated, and embedded in paraffin according to the standard procedure. The tissues were cut into 5-μm-thick sections and were blocked with 5% goat serum for 1 h at room temperature. The sections were then treated with the indicated primary antibody in 5% goat serum overnight at 4 °C and were washed in PBS 3 times. Then sections were subsequently incubated with the second antibody for 1 h at room temperature. After being washed in PBS, the slides were developed with 3, 3′-diaminobenzi dine and images were taken.

Animal experiment

2 × 106 H1299 cells in 100 μL of PBS were injected into the tail veins of 6-8 weeks female BALB/c nude mice (Liaoning Changsheng Biotechnology, Benxi, Liaoning, China), with 6 mice used per group, and the groups were completely randomized. No blinding was done. Eight weeks after injection, the mice were sacrificed and the tumor burden was checked in the lung through hematoxylin-eosin staining (HE) staining. In this study, all animal experiments were conducted in strict adherence to the ethical principles outlined in the Declaration of Helsinki. We ensured that all experimental procedures were aimed at minimizing pain and discomfort to the animals and were carried out in strict compliance with relevant national and international guidelines and regulations. The experimental protocols were reviewed and approved by the Laboratory Animal Ethics Committee of Xiangya Hospital (approval #202110138).

Clinical samples

Non-small cell lung cancer samples were collected from the Department of Pathology, Xiangya Hospital. A total of 95 NSCLC patients, excluded receive any treatment before surgery, were selected. The expression of Rbsn in the lung cancer sample and adjacent para-cancerous sample were examined. This study was approved by the Clinical Research Ethics Committee of Xiangya Hospital (approval #2021101149). Patient details are shown in Table 1. Researchers blinded the clinical information of patients before analyzing the data. Statistical significance was assessed using the Chi-square test. A p-value of less than 0.05 was considered statistically significant.

Table 1 The Rbsn expression in NSCLC patients.
Full size table

Statistical analysis

Data are presented as mean ± standard error (SE). Statistical significance between two groups was assessed using a two-tailed Student’s t test. For comparisons involving more than two groups, one-way ANOVA was employed. In instances where the data involved two independent factors, two-way ANOVA was utilized to evaluate the main effects and the interactions between these factors. Overall survival rates were estimated using the Kaplan–Meier method, with statistical comparisons conducted using the Log-rank test. A p-value of less than 0.05 was considered statistically significant.

Results

Low expression of Rbsn is correlated with poor prognosis of NSCLC patients

To discern the clinical relevance of Rabenosyn-5 (Rbsn) in non-small cell lung cancer (NSCLC), we embarked on an initial exploration of NSCLC patient datasets. This preliminary analysis disclosed a diminution of Rbsn expression in NSCLC specimens relative to non-cancerous lung tissue (Fig. 1A, S1A, and S1B). Notably, this downregulation correlated with a decline in overall patient survival (Fig. 1B). To corroborate these observations, we assessed Rbsn levels across a spectrum of lung cancer cell lines and compared them with the human normal lung epithelial cell line BEAS-2B (Fig. 1C). Consistently, Rbsn expression was found to be lower in the most lung cancer cell lines than that in the BEAS-2B cells (Fig. 1C). Subsequent immunohistochemical analyses of NSCLC tissues, sourced from patients undergoing surgical intervention at Xiangya Hospital, further substantiated the reduced expression of Rbsn in tumor cells (Fig. 1D, E). This diminished expression was again linked to poorer overall survival outcomes (Fig. 1F). Moreover, a significant relationship was observed between Rbsn expression levels, the incidence of lymph node metastasis, and the TNM staging in NSCLC patients (Table 1). The expression of Rbsn in patients was found to be independent of age, gender, and tumor size. Besides, patients with high Rbsn expression exhibited a lower rate of lymph node metastasis and were more likely to be in the early stages of tumor development. In aggregate, these data intimate that Rbsn may serve as an inhibitory factor in the progression and metastatic behavior of NSCLC.

Fig. 1: Rbsn is lowly expressed in NSCLC.
figure 1

A The expression of Rbsn in NSCLC patients was analyzed in the LUNG CANCER EXPLORER platform (TCGA_LUAD_2016 dataset). The numbers in parentheses represent the sample size. A two-tailed Student’s t-test was used for statistical analyses. B Kaplan–Meier analysis of prognosis about Rbsn expression in NSCLC patients from an R2 clinic patient database (TCGA-515-rsem-tcgars dataset). Rbsn expression levels were divided into high and low groups based on the median expression level of Rbsn. Log-rank test was used for statistical analyses. C The expression of Rbsn was assessed in different cell lines using Western blotting. D, E Immunohistochemical staining of Rbsn was performed in human NSCLC tissues. The right panels in (D) show a higher magnification of the regions within the black boxes. The quantification results of Rbsn staining are presented in (E). A two-tailed Student’s t test was used for statistical analyses. Data are presented as means ± SE (n = 95), ***p < 0.001. F Overall survival analysis was conducted to evaluate the impact of Rbsn expression on NSCLC patients. Rbsn expression levels were divided into high and low groups based on the median expression level of Rbsn. Log-rank test was used for statistical analyses.

Full size image

Rbsn inhibits lung cancer cell migration and invasion

To delineate the contribution of Rbsn to NSCLC pathobiology, we engineered A549 and H1299 cell lines to stably overexpress Rbsn. Intriguingly, Rbsn upregulation did not markedly impact the proliferation of these NSCLC cell lines, as indicated by Cell Counting Kit-8 (CCK8) assay outcomes (Fig. 2A). Similarly, the capacity for anchorage-independent growth remained unchanged following Rbsn enhancement (Fig. 2B, C). Conversely, a pronounced attenuation of migratory and invasive capabilities was observed in cells overexpressing Rbsn, as evidenced by wound healing (Fig. 2D, E) and transwell assays (Fig. 2F, G). These data collectively imply that Rbsn principally impedes NSCLC progression by thwarting the migratory and invasive potential of lung cancer cells.

Fig. 2: Rbsn inhibits lung cancer cell migration and invasion.
figure 2

A The effect of Rbsn on the proliferation of lung cancer cells was examined using CCK8 analysis (n = 3, technical replicates) and the experiments were conducted with at least three independent repetitions. B, C The colony formation of lung cancer cells in soft agar was assessed, with or without Rbsn overexpression. The quantification data are shown in (C) (n = 3, technical replicates) and the experiments were conducted with at least three independent repetitions. D, E Lung cancer cell migration was evaluated using a wound healing assay, with or without Rbsn overexpression. The quantification data are shown in (E) (n = 3, technical replicates) and the experiments were conducted with at least three independent repetitions. F, G The invasion of lung cancer cells was examined using a transwell assay, with or without Rbsn overexpression. The quantification data are shown in (G) (n = 3, technical replicates) and the experiments were conducted with at least three independent repetitions. In the results of (A), (C), (E), and (G), a two-tailed Student’s t test was used for statistical analyses. Data are presented as means ± SE, ** p < 0.01. EV indicates empty vector; n.s. indicates no significant difference.

Full size image

Rbsn inhibits CDC42 activity

To elucidate the mechanistic basis for Rbsn-mediated suppression of cancer cell migration and invasion, we examined the small Rho GTPase signaling pathway, a pivotal modulator of cytoskeletal organization. Employing assays to measure the activity of Rac, CDC42, and Rho, we discovered a significant reduction in active CDC42 in the Flag-Rbsn-expressing cells relative to the empty vector (EV) controls, while levels of active Rac1 and Rho remained unaltered between the two groups (Fig. 3A–D, S2A, and S2B). This suggests that Rbsn overexpression selectively downregulates CDC42 activity without affecting Rac and Rho pathways. Complementing these findings, we noted a decrease in filopodia formation, a cellular process governed by CDC42, in both A549 (Fig. 3E, G) and H1299 cells (Fig. 3F, H) overexpressing Rbsn. These findings suggest that Rbsn impedes cancer cell motility and invasiveness by specifically targeting and inhibiting CDC42 activity.

Fig. 3: Rbsn inhibits CDC42 activity.
figure 3

A–D The GTPase activity of CDC42, Rac1, and Rho in lung cancer cells, both with and without Rbsn overexpression, was assessed through repeated experiments using a small Rho GTPase activity assay, along with the statistical analysis related to A549 (C) and H1299 (D). E–H Cells were infected with indicated plasmids, and immunofluorescence was performed 72 h later to visualize filopodia formation in lung cancer cells with or without Rbsn overexpression. Closed boxes are magnified in the insets. The percentage of cells with filopodia in each group (100 cells from three independent experiments) is quantified in (G) and (H), respectively. Two-tailed Student’s t test was used for statistical analyses. Data are presented as means ± SE, * p < 0.05, ***p < 0.001, ****p < 0.0001. EV indicates empty vector; TCL indicates total cell lysate.

Full size image

Rbsn is a GTPase-activating protein (GAP) for CDC42

To delineate the regulatory influence of Rbsn on CDC42 activity, we employed tandem affinity purification (TAP) to isolate Rbsn-associated proteins from H1299 cell lysates. Extracts from cells expressing either EV or SF-tagged Rbsn underwent TAP, with the resultant protein complexes visualized by silver staining and subjected to mass spectrometric identification (Fig. 4A). Remarkably, CDC42 emerged as a prominent member of the Rbsn interaction network (Fig. S3). Co-immunoprecipitation assays further corroborated the Rbsn-CDC42 interaction in A549 and H1299 cells (Fig. 4B, C). Additionally, Rbsn demonstrated a binding preference for the GTP-bound form of CDC42 when in the presence of non-hydrolyzable GTPγS, but not with GDP (Fig. 4D–G). This specificity suggests a direct interaction between Rbsn and the active state of CDC42. Subsequent colorimetric GTPase assays, varying both the temporal and molar parameters of Rbsn and CDC42, indicated that Rbsn potentiates the GTPase activity of CDC42 in a time- and dose-dependent fashion (Fig. 4H). Moreover, knocking down Rbsn in A549 and H1299 cells was observed to enhance the activity of CDC42 (Fig. S4A–B). These findings strongly implicate Rbsn as a GAP for CDC42, enhancing the conversion of GTP to GDP and thereby attenuating CDC42 activity.

Fig. 4: Rbsn is a GAP for CDC42.
figure 4

A Tandem affinity purification was performed to isolate the Rbsn interacting proteins in H1299 cells, followed by visualization of the protein complex using silver staining. B, C The interaction between endogenous Rbsn and CDC42 was examined by co-immunoprecipitation in A549 (B) and H1299 (C). D–G The interaction between endogenous Rbsn and CDC42 was examined by co-immunoprecipitation in the presence of either GTPγS or GDP, along with the statistical analysis related to A549 (F) and H1299 (G), a two-way ANOVA was used for statistical analyses. H CDC42 GTPase activity assays were conducted to assess the impact of the Rbsn on CDC42 GTPase activity. The results demonstrated that the Rbsn increased CDC42 GTPase activity in a time- and dose-dependent manner. The numbers represent the molar ratio of the Rbsn protein to the CDC42 protein utilized in this assay. A one-way ANOVA test was used for statistical analyses. Data are presented as means ± SE (n = 6, technical replicates) and the experiments were conducted with at least three independent repetitions, *p < 0.05, **p < 0.01, ***p < 0.001. IP indicates immunoprecipitation.

Full size image

Akt-induced Rbsn Thr253 phosphorylation regulates GAP activity of Rbsn

To investigate the regulatory mechanisms of Rbsn GAP activity, we scrutinized the amino acid sequence of Rbsn, noting that a threonine residue at position 253 resides within conserved Akt phosphorylation motifs (R-x-R-x-x-S/T) (Fig. 5A). Moreover, we found that the interaction between Rbsn and Akt occurs in A549 and H1299 cells respectively, as demonstrated by co-immunoprecipitation assays (Fig. 5B, C). To determine whether Akt phosphorylates Rbsn at Thr253, 293 T cells were co-transfected with vectors expressing either constitutively active Akt (Akt DA) or kinase-dead Akt (Akt KD) along with a Flag-Rbsn vector. Immunoprecipitation using an anti-Flag antibody, followed by detection with an anti-phospho Akt substrate (RXXS*/T*) antibody, revealed that Akt DA facilitated phosphorylation of wild-type Rbsn, whereas Akt KD did not (Fig. 5D). Mutation of Rbsn at Thr253 to alanine (T253A) abrogated its phosphorylation, pinpointing Thr253 as the specific Akt phosphorylation site (Fig. 5D). Epidermal growth factor (EGF) stimulation, which activates Akt, induced phosphorylation of wild-type Rbsn in 293 T cells but not the T253A mutant, further validating Thr253 as the target (Fig. 5E). Simultaneously, EGF also stimulated the activation of Rbsn in A549 cells (Fig. 5F). Functional assays revealed that overexpression of wild-type Rbsn attenuated CDC42 activity in A549 and H1299 cells, as assessed by pull-down assays using CDC42 affinity beads. This inhibitory effect was not observed with the Rbsn T253A mutant (Fig. 5G, H). Moreover, the Rbsn T253A mutation negated the effects on filopodia formation (Fig. 6A, B), cell migration (Fig. 6C, D), and invasion (Fig. 6E, F). These results collectively underscore the pivotal role of Akt-mediated phosphorylation of Rbsn at Thr253 in modulating its GAP activity toward CDC42, with downstream consequences on cancer cell morphology, motility, and invasive behavior.

Fig. 5: Akt-induced Rbsn phosphorylation is essential for CDC42 activity regulation.
figure 5

A Amino acid sequences of Rbsn from various species are presented, with the conserved threonine at position 253 (highlighted in red) conforming to the Akt substrate motif. B, C The endogenous interaction between Rbsn and Akt was checked by co-immunoprecipitation in lung cancer cells. D Indicated plasmids were transfected into 293T cells, and the immunoprecipitation was performed with anti-Flag antibody or control IgG, followed by p-Akt substrate to check the phosphorylation of Rbsn, along with statistical analysis of three independent experiments. A One-way ANOVA was used for statistical analyses. E, F The indicated plasmids were transfected into 293 T cells (E), or A549 Flag-Rbsn cell line (F) stimulated with or without 100 ng/ml EGF for 30 min. Subsequently, the immunoprecipitation was performed with anti-Flag antibody or control IgG. The phosphorylated Rbsn was examined with a p-Akt substrate, along with statistical analysis of three independent experiments. A One-way ANOVA and two-tailed Student’s t test was used for statistical analyses. G, H CDC42 activity assay to examine CDC42 activity in lung cancer cells stably expressing with indicated plasmids. The active CDC42 was purified by a pulldown experiment with the affinity beads, along with statistical analysis of three independent experiments. A One-way ANOVA was used for statistical analyses. WT indicates wild type; IP indicates immunoprecipitation; TCL indicates total cell lysate. Data are presented as means ± SE, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Full size image
Fig. 6: Rbsn Thr253 phosphorylation is essential for lung cancer cell migration and invasion.
figure 6

A, B Cells were infected with indicated plasmids, and immunofluorescence was performed 72 h later to visualize filopodia formation in lung cancer cells expressing the indicated plasmids. Closed boxes in (A) are magnified in the insets. The percentage of cells with filopodia in each group (100 cells from three independent experiments) is quantified in (B), and a One-way ANOVA was used for statistical analyses. C, D Wound healing assay to examine cell migration in lung cancer cells stably expressing with indicated plasmids. The quantification data from three independent experiments are shown in (D) (n = 3, technical replicates) and the experiments were conducted with at least three independent repetitions. E, F Cell invasion in lung cancer cells expressing the indicated plasmids was examined using a transwell assay. Quantification data for cell invasion are presented in (F) (n = 3, technical replicates) and the experiments were conducted with at least three independent repetitions. In the results of (B), (D), and (F), a One-way ANOVA was used for statistical analyses. Data are presented as means ± SE, *p < 0.05, **p < 0.01, ***p < 0.001. EV indicates an empty vector; WT indicates wild type.

Full size image

Rbsn inhibits lung cancer metastasis in vivo

To validate the impact of Rbsn on lung cancer metastasis in an in vivo setting, H1299 cells were intravenously injected into immunocompromised nude mice (Fig. 7A). After 8 weeks, the mice were sacrificed, and their lungs were collected for assessment of tumor burden. Notably, the overexpression of wild-type Rbsn led to a significant reduction in lung metastasis compared to the EV control cells. Conversely, the overexpression of the Rbsn T253A mutant showed minimal impact on metastasis, similar to that observed in the EV control group (Fig. 7B, C). These findings provide further evidence supporting the inhibitory role of Rbsn in lung cancer metastasis.

Fig. 7: Rbsn inhibits lung cancer metastasis in vivo.
figure 7

A Schematic of the mice models to check lung cancer metastasis. B, C Representative images depicting the metastatic tumor foci in the lungs are presented. Quantification data for the number of metastatic tumor foci are shown in (C). A One-way ANOVA was used for statistical analyses. Data are presented as means ± SE (6 mouses per group), **p < 0.01. EV indicates an empty vector; WT indicates wild type.

Full size image

Discussion

Metastatic dissemination is the leading cause of mortality in cancer patients. Proteins involved in intracellular trafficking have emerged as pivotal modulators of cytoskeletal dynamics. In this study, we have elucidated that the expression level of Rbsn is markedly reduced in NSCLC, correlating with a poorer patient prognosis. Our data further demonstrate that Rbsn functions as a GAP for the small GTPase CDC42, thereby attenuating CDC42-mediated signaling pathways that drive filopodia assembly and, consequently, tumor cell metastasis. Moreover, we have identified a critical post-translational modification; specifically, phosphorylation of Rbsn at threonine 253 by Akt kinase is requisite for its GAP activity, underscoring a novel regulatory mechanism of Rbsn function in cancer progression.

CDC42, a pivotal member of the Rho GTPase family, toggles between an active GTP-bound state and an inactive GDP-bound state. The regulation of small Rho GTPase activity is principally orchestrated by GEFs and GAPs. GEFs facilitate the transition to the active GTP-bound conformation, thereby stimulating small Rho GTPase activity, whereas GAPs expedite GTP hydrolysis, effectively terminating the signaling [33]. Specific mutations in CDC42, such as G12V and Q61L, have been shown to compromise GTPase hydrolytic function, resulting in aberrant cellular morphology [34]. In contrast, the F28L mutation enhances the dynamic interconversion between GDP- and GTP-bound states [35], culminating in filopodia formation [36]. Dysregulation of small Rho GTPase signaling, as evidenced in various neoplasms, underscores the importance of modulating GTPase activity in the context of tumorigenesis [37].

The functional diversity of CDC42 GAPs in key biological processes is well-documented. SrGAPs, a subset of CDC42 GAPs, have been implicated in the regulation of Slit-mediated neuronal migration [38, 39]. In Saccharomyces cerevisiae, the CDC42 GAP Rga1 influences the establishment of polarity axis orientation, asymmetric cell division, and longevity, with Rga1-deficient yeast displaying a reduced lifespan due to compromised polarity establishment [40, 41]. Conversely, Rga3 appears to be non-essential for cell polarization during mitotic growth in yeast [42]. Rga4 has been identified as a determinant of growth patterns across successive generations [43]. In Schizosaccharomyces pombe, Rga6 regulates the septin complex’s cortical localization, thereby controlling monopolar growth during spore germination; loss of Rga6 results in aberrant bipolar growth due to impaired septin localization [44]. The involvement of CDC42 GAPs in oncogenesis has also been explored. CdGAP, a CDC42-specific GAP, is upregulated following TGF-β stimulation and participates in TGF-β-mediated migration and invasion of breast cancer cells [45]. Furthermore, CdGAP modulates gene expression associated with epithelial-mesenchymal transition (EMT) [46], influences cancer cell migration and invasion [47], and is implicated in VEGF-dependent angiogenesis [48]. Despite the critical functions of CDC42 GAPs elucidated in various studies, the repertoire of known GAPs remains comparatively small against the backdrop of identified GEFs.

Previous studies have primarily centered on its role as an effector of the small Rab GTPases Rab4 and Rab5, orchestrating endocytic trafficking pathways with implications for diseases such as congenital neutrophil defect syndrome and intractable seizures [49, 50]. Furthermore, Rab5A has been implicated in promoting Rbsn-dependent endo/exocytic cycles, which are critical for protein degradation and the invasive capacity of tumors, potentially enhancing metastatic progression [51]. In addition, previous studies revealed intracellular trafficking proteins modulate small Rho GTPases activity via regulating their intracellular trafficking [52, 53]. However, our study has revealed for the first time that Rbsn functions as a GAP of small Rho GTPase CDC42 and can directly inhibit the activity of CDC42 through its GAP function. This discovery sheds light on the crosstalk between small Rab GTPases and small Rho GTPases, suggesting that intracellular trafficking may regulate the functions of small Rho GTPases.

Our study also has some limitations. A notable limitation is the absence of a specific antibody targeting the Thr253 phosphorylation site on Rbsn, which precluded the analysis of phosphorylated Rbsn within tumor specimens. Additionally, the possibility of phosphorylation at alternative sites on Rbsn and the consequent effects on its regulatory activity warrant further exploration.

Conclusions

Our work demonstrates that Rbsn suppresses lung cancer metastasis by reducing CDC42 activity.