Tuftelin1 drives experimental pulmonary fibrosis progression by facilitating stress fiber assembly

Recombinant human TGF-β1 was obtained from R&D systems (Minneapolis, MN, USA). Bleomycin sulfate was obtained from Hisun Pharm (Taizhou, China). Sodium orthovanadate (SOV) was obtained from Sigma-Aldrich (Shanghai, China). Wiskostatin (Wis) was obtained from Abcam (Shanghai, China). N-WASP was obtained from Proteintech (Wuhan, China). Goat anti-rabbit IgG of β-actin, Alpha-smooth muscle actin (α-SMA), p^Y256N-WASP and GAPDH were purchased from Affinity Biosciences (Changzhou, China). Goat anti-mouse IgG of Vimentin and E-cadherin were purchased from Cell Signaling Technology (Shanghai, China), Goat anti-rabbit IgG of N-cadherin, Fibronectin and Collagen I were purchased from Cell Signaling Technology (Shanghai, China), TUFT1 was purchased from Thermo Fisher (Suzhou, China).

Human lung tissues were obtained by surgical lung biopsy (SLB) from patients in Xinxiang central hospital, IPF was diagnosed according to the international guidelines [25]. The human lung tissue study was approved by the Xinxiang central hospital Medical Research Ethics Committee (No.2019-01-12) and was conducted in accordance with the Declaration of Helsinki. All patients who donated tissues have provided informed consent.

The human lung adenocarcinoma cell line A549 (ATCC^® CCL-185™, ATCC, Manassas, VA), negatively tested for mycoplasma, was cultured in DMEM/F12 medium supplied with 10% fetal bovine serum and antibiotics (100 units/mL penicillin and 100 μg/mL streptomycin). The human embryonic lung fibroblast cell line (MRC-5) was purchased from Procell Life Science &Technology Co., Ltd. (Wuhan, China) and cultured in MEM medium (containing NEAA) (Shanghai Zhong Qiao Xin Zhou Biotechnology Co., Ltd., ZQ-300) supplied with 10% fetal bovine serum and antibiotics (100 units/ml penicillin and 100 μg/ml streptomycin) in an atmosphere of 5% CO₂ and humidified at 37 °C. For TGF-β1 stimulation, cells were serum-starved overnight before 5 ng/mL TGF-β1 was added for 24 h.

TUFT1/siRNA-1(5′-GGAGTCCCATGATGGACAT-3′), TUFT1/siRNA-2(5′-GCAGAGGCUGUGUGACAAATT-3′), TUFT1/siRNA-3(5′-GAGGAACTTCGGAGCAACA-3′) were purchased from RIBOBIO (Guangzhou, China). To transfect siRNA into A549 and MRC-5 cells, the cells were cultured on 6-well plates at 50–70% coverage before transfection. Lipofectamine RNAi MAX (RIBOBIO, Guangzhou, China), Opti-MEM (Thermo Fisher) and Individual siRNA (50 nM) were mixed completely and incubated for 10 min at room temperature. The efficiency of transfection was testified by western blot in 48 h after transfection.

Mice from each group were anesthetized by isoflurane and fixed in the supine position. The pulmonary CT images of each mouse were photographed by a CT imaging system (SkyScan 1276, Bruker, Kontich, Belgium) at the 14th day. And the parameters were set as follows: 65 kV, 200μA, over a total angle of 360◦ for a total scan time of approximately 15 min, and the micro-CT image acquired with a respiratory-gated technique. Micro-CT images of pulmonary were reconstructed by using insta-Recon software (Bruker, Kontich, Belgium).

A hydroxyproline assay kit (MAK008, Sigma, St. Louis, MO, USA) was used to detect the content of hydroxyproline in mouse right lung [26]. Briefly, the fresh lung tissue was homogenized in water (100 µL water of every 10 mg lung tissue), followed by hydrolyzing the samples after adding equivalent 12 N HCl for 3 h at 120 °C. Next, the samples were centrifuged (12,000 rpm, 4 °C, 3 min; Beckman Microfuge 20R) and extracted 10 µL supernatant of every sample, the samples were transferred to a 96-well plate and evaporated at 56 ℃. The samples in the plate were incubated for 5 min at 20 ℃ after added 100 µL chloramine T reagent into each well. Afterward, the samples were incubated for 90 min at 56 °C after mixed with 100 µL p-dimethylaminobenzaldehyde reagent. Absorbance of samples was measured by BioTek ELx800 plate reader at λ = 560 nm (Winooski, VT, USA). The hydroxyproline content of samples was calculated by a standard sample and presented as µg hydroxyproline per right lung.

The lungs of mice were collected and fixed with 4% paraformaldehyde. The lung tissues were embedded in paraffin and sliced into 3 μm sections. The sections were stained with Masson’s trichrome (Beyotime, Guangzhou, China) and H&E (Beyotime, Guangzhou, China) to assess the collagen content and evaluate the morphometric changes. The acquisition of panoramic images was performed using a BioTek Cytation C10 Confocal Imaging Reader. Images were analyzed by Image Pro Plus software (version 6.0, Media Cybernetics). In order to devise an adapted Ashcroft score, five fields from each lung section were randomly chosen under 100× magnification. The mean score for each field was determined independently by two pathologists in a blinded fashion [27].

The lung slides were prepared following previously described protocols [28]. The lung sections were incubated with the primary antibodies overnight at 4 °C, then the secondary antibody biotinylated anti-rabbit IgG (1:100, Beyotime, Guangzhou, China) was incubated for 60 min at 37 °C. Followed by incubating with SABC (1:100, Beyotime, Guangzhou, China), and then visualized by DAB stain (Beyotime, Guangzhou, China). Cell nuclei were stained with hematoxylin.

SiRNA control or TUFT1 siRNA was transfected into A549 and MRC-5 cells. After 24 h, the cells were placed in a 6-well plate using a culture medium containing 1% FBS. Upon reaching confluence, a pipette tip was utilized to create a direct scratch across the cell monolayer. Pictures were taken at the predetermined time intervals to document the progression of wound healing. Wound healing was calculated as follows: [1−(Width of the wound at a given time/width of the wound at t = 0)] × 100%.

Cell migration assay was performed using a transwell chamber (Corning Incorporated, Corning, NY, USA, cat:3422). A549 cells transfected with SiRNA control or TUFT1 siRNA were seeded into transwell plates with serum-free medium, then medium containing 10% FBS was added to the bottom of the chamber. After 24 h, Migration cells were fixed and stained with 1% crystal violet and counted in three random fields under an optical microscope.

Collagen gel contraction assay was performed using a cell contraction assay kit (CBA-201, Cell Biolabs, Inc., San Diego, CA, USA). MRC-5 cells were suspended in bovine type 1 collagen solution (chilled on ice) at the concentration of 1.85 mg/mL and seeded at a density of 1 × 10⁶ cells/well in a Costar low attachment surface polystyrene 24-well plates (Corning Incorporated, Corning, NY, USA). The plates were then placed in a 37 °C incubator for 1 h to allow for collagen gel polymerization, and 1 mL culture medium was added on top of the gel lattice, and cells were incubated overnight. Then cells were pre-incubated with sodium orthovanadate (20 µM) for 24 h or Wiskostatin (10 µM) for 1 h. Once the contraction agonists were added, the gels were meticulously detached from the sidewalls of each well by a sterile scalpel. Photographs were captured after 24 h, and the analysis of gel disc area was conducted using Image J software for quantification.

The 4% paraformaldehyde was used to fix the lung sections and cell slides and 0.3% Triton X-100 was used to elevate the permeable ability of cells. Then, the sections were block by 10% goat serum for 0.5 h at 37 °C and incubated with primary antibodies overnight at 4 °C. Corresponding secondary antibodies were incubated for 1 h at 37 °C. F-actin was stained with Phalloidin-iFluor 594 Reagent (ab176757, Abcam, USA) and nuclei were stained with DAPI. The images were captured by a fluorescence microscope (Zeiss LSM780, Carl Zeiss).

The steps of Western blot analyses were described as previously [29]. Initially, lung homogenates or cell lysates were prepared by RIPA lysis buffer (AR0102, Boster, Wuhan, China). The protein concentrations were detected by BCA kit (Solarbio, Beijing, China). The protein was loaded onto SDS-PAGE followed by electrotransferred onto PVDF membranes completely. The membranes were incubated with primary antibodies overnight at 4 °C and incubated with corresponding secondary antibodies for 1 h at room temperature. The images were captured by Odyssey Imaging System (LI-COR Biosciences, Lincoln, NE, USA). The densitometry band quantification was measured by the Image Studio software (LI-COR Biosciences, NE, USA).

IHC showed that TUFT1 was increased significantly in IPF lung tissues, and TUFT1 was mainly located in myofibroblasts in the fibroblast foci of the IPF lung (Fig. 1a). Meanwhile, we detected the expression of Tuft1 in bleomycin induced fibrotic mouse lung tissue. Consistent with IPF lung, Tuft1 was increased significantly in BLM-induced fibrotic mouse lungs (Fig. 1b–d). To further establish the association between TUFT1 and pulmonary fibrosis in an in vitro context, A549 and MRC-5 cells were exposed to TGF-β1 (5 ng/mL) for a duration of 24 h. This treatment resulted in a marked reduction in the epithelial marker E-cadherin, coupled with a pronounced elevation in the mesenchymal markers N-cadherin and Vimentin. Furthermore, the protein levels of TUFT1 were notably upregulated in TGF-β1-induced A549 cells (Fig. 1e). Similarly, the ECM protein collagen I and myofibroblasts marker α-SMA increased in TGF-β1-treated human lung fibroblasts (MRC-5 cells), and the protein levels of TUFT1 were increased in TGF-β1-induced MRC-5 cells (Fig. 1f, g). These results sustain the idea that augmented TUFT1 may have clinically relevant implications in pulmonary fibrosis.

To investigate the role of TUFT1 in pulmonary fibrosis, we examined fibrotic responses after the silenced TUFT1 expression by using Adeno-associated virus knockdown of TUFT1 (Tuft1/short hairpin RNA (shRNA)) in bleomycin-induced mouse pulmonary fibrosis model. The schematic diagram illustrated the time and course of the experiments in Fig. 2a. Micro-CT images taken on day 14 showed the bleomycin treatment caused an obvious density increase in mouse lungs, and the Tuft1/shRNA groups showed a marked reduction in the lung density indicated by increased parenchymal opacity (Fig. 2b). Moreover, mice treated with Tuft1/shRNA adenovirus showed normal lung tissues structure compared with the bleomycin group as detected by H&E staining (Fig. 2c). Meanwhile, the Masson’s trichrome staining showed that the collagen deposition increased obviously in bleomycin groups, whereas this phenomenon could be rectified by disruption of TUFT1 (Fig. 2c), furthermore, silencing the Tuft1 reduced the expression of α-SMA as indicated by immunohistochemistry staining (Fig. 2c). Ashcroft histopathological grading score was calculated according to hologram of the Masson’s trichrome staining (Additional file 1: Fig. S1), bleomycin treated mice showed increased Ashcroft histopathological grading score compared to saline group, while silencing the Tuft1 decreased the Ashcroft score significantly (Fig. 2d). As expected, the bleomycin treatment enhanced the level of Hydroxyproline sharply, and the Tuft1/shRNA alleviated this phenomenon (Fig. 2e). Meanwhile, the levels of Fibronectin and collagen I was decreased by Tuft1/shRNA in pulmonary fibrosis process (Fig. 2f–i). Interestingly, the Phalloidin staining revealed a marked abundance of F-actin within the fibrotic tissues of the bleomycin-induced lung fibrosis model, whereas the Tuft1/shRNA led to a reduction in F-actin formation in bleomycin-challenge lung tissues (Additional file 1: Fig. S2). Altogether, these data indicate that inhibition of TUFT1 alleviated bleomycin-induced experimental models of lung fibrosis.

Next, we intend to explore the influence of TUFT1 on the fibrotic phenotype of epithelial cells and fibroblasts in vitro. Knockdown efficiency of silencing TUFT1 in A549 cells was evaluated by Western Bolt (Fig. 3a). Phalloidin staining revealed that the silencing of TUFT1 resulted in distinct cell morphology changes in TGF-β1 treated A549 cells (Fig. 3b). Moreover, silencing TUFT1 impaired the migration ability of TGF-β1 treated A549 cells (Fig. 3c, d). Meanwhile, the knockdown of TUFT1 also reduced the invasion of TGF-β1 treated A549 cells detected by transwell (Fig. 3e, f). In parallel, we also found that the silencing of TUFT1 influences the microfilaments morphology in TGF-β1 treated MRC-5 cells detected by phalloidin staining (Fig. 3g). Moreover, the knockdown of TUFT1 could decrease the expression of Fibronectin, collagen I, and α-SMA in TGF-β1 treated MRC-5 cells (Fig. 3h). Also, silencing TUFT1 impaired the cell migration of the MRC-5 cells post the induction of TGF-β1 (Fig. 3i, j). Finally, we found that knocking down TUFT1 decreased the contraction ability of TGF-β1 treated MRC-5 cells (Fig. 3k, l). These data implicated that disruption of TUFT1 suppresses the activation of both epithelial cells and fibroblasts in vitro.

As silencing of the TUFT1 distorted the shape of microfilaments, in order to illustrate the mechanism of TUFT1 influence the microfilaments, we examined the expression level of N-WASP and p^Y256N-WASP, which influenced the assembly of cell microfilaments. The p^Y256N-WASP was decreased significantly by silencing TUFT1 in TGF-β1 treated A549 cells (Fig. 4a–c). Interestingly, N-WASP and p^Y256N-WASP were augmented in IPF lungs compared with the non-IPF donor as indicated by immunohistochemistry staining (Fig. 4d, e). Consistently, N-WASP and p^Y256N-WASP were increased significantly in the bleomycin-induced fibrosis mice model compared to the control (Fig. 4f, g).

Next, we aimed to ascertain whether the p^Y256N-WASP was involved in the TUFT1-mediated pro-fibrotic phenotype. Sodium orthovanadate (20 µM), a phosphorylase inhibitor of Tyrosine, was used to interfere with the dephosphorylation of p^Y256N-WASP after the silenced of TUFT1 for 48 h. The result indicated that sodium orthovanadate reversed the impaired contract ability of TGF-β1 treated MRC-5 cells transfected with si-TUFT1 (Fig. 4h, i). Moreover, silencing TUFT1 decreased the level of p^Y256N-WASP in TGF-β1 treated MRC-5 cells, as well as the myofibroblast marker α-SMA, and the sodium orthovanadate reversed the decrease of both p^Y256N-WASP and α-SMA (Fig. 4j–m).

To explore the relationship between TUFT1 and N-WASP in epithelial cells, we constructed a vector transiently expressing N-WASP. Over-expressed N-WASP led to a substantial increase in TUFT1 fluorescence intensity in MRC5 cells, N-WASP and TUFT1 were co-localized in the cytoplasm (Fig. 5a). Furthermore, exogenously expressed N-WASP and TUFT1 were co-immunoprecipitated suggesting a physical interaction between them (Fig. 5b, c). Likewise, over-expressed TUFT1 also induced the accumulation of phosphorylated N-WASP in MRC5 cells (Fig. 5d). Interestingly, while TGF-β1-induced p^Y256N-WASP accumulated in the cell nucleus, silencing TUFT1 led to a dispersed distribution of p^Y256N-WASP in the cytoplasm, and this phenomenon coincidence with the microfilament disruption induced by silencing TUFT1 in MRC-5 cells (Fig. 5e, f). A similar scenario was observed in A549 cells (Additional file 1: Fig. S3). Collectively, these above results showed that TUFT1 interacts with N-WASP and influenced the level and distribution of p^Y256N-WASP.

N-WASP was known to play a critical role in regulating the formation of α-SMA-containing cytoplasmic filaments during myofibroblast differentiation and contractility through active the ARP2/3 complex [24]. We found phosphorylation of N-Wasp at Y256 was induced in BLM-challenged fibrotic lung, whereas silencing the Tuft1 could decrease the level of phosphorylated N-Wasp and collagen I in BLM-challenged fibrotic lung (Fig. 6a, b). Meanwhile, we found silencing the Tuft1 could disperse the p^Y256N-Wasp in the fibrotic areas (Additional file 1: Fig. S4).

TUFT1 was engaged in the formation of microfilaments and silence of the TUFT1 could decrease the percentage of cells with actin cores in A549 cells, which is similar to the effect of wiskostatin, a selective N-WASP inhibitor which can bind the GTPase binding domain stabilizing the autoinhibited conformation (Fig. 6c, d). Over-expressed TUFT1 led to nucleus-translocation of p^Y256N-WASP in MRC-5 cells, whereas the Wiskostatin could reverse this phenomenon completely (Additional file 1: Fig. S5). Meanwhile, TUFT1 enhanced the contractility of MRC-5 cells, whereas Wiskostatin inhibited the enhanced contractility induced by TUFT1 (Fig. 6e, f). Interestingly in MRC-5 cells, over-expressed TUFT1 induced a high level of p^Y256N-WASP as well as FN, collagen I and α-SMA, while wiskostatin inhibited TUFT1 induced FN, collagen I, and α-SMA (Fig. 6g–k). Taken together, the above data illustrated that TUFT1 facilitates profibrotic responses in a p^Y256N-WASP-dependent manner.