NEDD4-Mediated Endothelial-Mesenchymal Transition Participates in Radiation-Induced Lung Injury Through the ATM Signaling Pathway

Abstract

Objective

To elucidate the role of NEDD4 in ionizing radiation (IR)-induced endothelial-mesenchymal transition (EndMT) and its molecular mechanism in radiation-induced lung injury (RILI), given the unclear regulatory pathways of EndMT in RILI pathogenesis.

Methods

IR-induced EndMT was observed during RILI in vivo and in vitro by immunohistochemical staining and Western blot analysis. Proteomics identified NEDD4 as a candidate, validated by RNA sequencing (RNA-seq) and quantitative real-time polymerase chain reaction (qRT‒PCR). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis linked NEDD4 to PI3K-AKT signaling. Co-immunoprecipitation (Co-IP) confirmed NEDD4-ATM interaction.

Results

IR upregulated NEDD4 in endothelial cells, correlating with EndMT progression. NEDD4 overexpression enhanced ATM pathway activation, modulating genes upstream/downstream of ATM. Co-IP verified physical NEDD4-ATM binding, suggesting NEDD4 stabilizes ATM to promote EndMT.

Conclusion

Overall, our study shows that NEDD4 mediates EndMT to participate in RILI through the ATM signaling pathway, which may break new ground for understanding the occurrence and development of RILI.

Introduction

Radiation-induced lung injury (RILI), the most common type of dose-limiting toxicity due to radiotherapy for thoracic malignancies, clinically manifests as early ionizing radiation (IR)-induced pneumonitis and late radiation-induced lung fibrosis (RILF).^1,2 The degree of RILI is generally correlated with IR dose, the irradiated area, basal lung function, age and smoking. ³ In recent decades, numerous studies on the mechanism of RILI have emerged; these studies have identified mechanisms that include free radicals, cytokines, and cell senescence.^4,5 However, the exact roles of the inflammatory cytokines and free radicals involved in RILI and whether they interact remain unclear and may vary from tissue to tissue. ⁶

The injury of vascular endothelial cells, which are important target cells of RILI, and the corresponding outcome are closely related to the course of RILI. Recently, Diana Klein et al found that mesenchymal stem cells can restore expression of superoxide dismutase 1 (SOD1), an antioxidant enzyme, and significantly inhibit IR-induced endothelial cell damage and the development of long-term lung fibrosis. ⁷ This study has provided a new direction for the precise prevention and treatment of RILI. Nevertheless, the specific molecular mechanism underlying vascular endothelial cell injury during the progression of RILI has not been fully elucidated.

Endothelial-mesenchymal transition (EndMT) refers to the process by which endothelial cells change from endothelial cells to mesenchymal cells or fibroblasts upon stimulation, which exhibits upregulation of Vimentin and alpha-smooth muscle actin (α-SMA) and downregulation of endothelial cell adhesion molecule 1 (CD31) and Vascular Endothelial Cadherin (VE-cadherin).^8,9 Originally discovered as a key mechanism for cardiac development, EndMT is also involved in embryonic development as well as tissue healing and organ fibrosis. ¹⁰ Ionizing radiation induces vascular damage and inflammation. Persistent inflammation leads to vascular endothelial cell damage and contributes to capillary permeability and pulmonary edema. Prolonged vascular damage leads to EndMT and eventually culminates in fibrotic changes. ² Zheng’s study has shown that alleviating oxidative damage and EMT/EndMT exert significant protective effects against lung injury, inflammatory responses, and fibrosis. ¹¹ Recently, numerous studies have reported that EndMT participates in lung fibrosis, possibly by activating the transforming growth factor-β (TGF-β) or mitogen-activated protein kinase (MAPK) pathway.^12-14 Despite these findings, the relationship between EndMT and RILI as well as the possible role and specific mechanism have not been widely reported.

Human umbilical vein vascular endothelial cells (HUVECs) are commonly used as a cell model for endothelial cells in the event of a radiologic catastrophe.^15-17 In addition, the rat is an excellent animal model to investigate radiation-induced lung injury.^18,19 On the basis of these findings, in this study, HUVECs and rats were used to identify the role of EndMT in RILI in vitro and in vivo. Then, we further clarified the molecular mechanism underlying EndMT, providing potential therapeutic strategies against RILI.

Materials and Methods

Reagents and Materials

DAPI was purchased from Sigma Aldrich (St. Louis, MO). Lipofectamine 3000 was obtained from Invitrogen (Carlsbad, CA, USA). Phalloidin and Nucspot were purchased from Thermo Fisher Scientific (Waltham, MA, USA). Antibodies against GAPDH, β-actin, and α-tubulin; HRP - conjugated anti-mouse; and anti-rabbit immunoglobulin G (IgG) were purchased from Beyotime (Nantong, China). Antibody against VE-cadherin (#YT5611) was purchased from Immunoway (Newark, DE, USA). Antibody against α-SMA (#CY5295) was obtained from Abways (Shanghai, China). Antibody against CD31 (#3528) was purchased from Cell Signaling Technology (Beverly, MA, USA). Antibodies against Vimentin (#ab92547), S100A4 (#ab197896), NCL (#ab129200), ATM (#ab32420), NBS1 (#ab23996), RAD50 (#ab124682), MRE11 (#ab109623), CHK2 (#ab109413), p-CHK2 (#ab85743) and p53 (#ab32389) were purchased from Abcam (Cambridge, MA, USA). Antibodies against NEDD4 (#21698-1-AP) and IL-8 (#27095-1-AP) were purchased from Proteintech (Chicago, IL, USA). The human NEDD4-coding region (GenBank accession no. NM_006154.4) was amplified by PCR using primer pairs specific for NEDD4. The plasmid was then sequenced by GeneChem (Shanghai, China) for confirmation.

Animals and Irradiation

All animal experiments in this study were approved by the Animal Ethics Committee of Soochow University (Suzhou, China) and met the animal ethics review requirements. A total of 30 male SD rats (250-300 g) were obtained from Shanghai SLAC Laboratory Animal Co, Ltd (Shanghai, China). An RILI model was established in the rats with unilateral lung irradiation as reported previously. ¹⁸ The right lung tissues of the rats were irradiated with 6 MeV X-rays (Varian 23EX linear accelerator, Palo Alto CA) at a single dose of 0 (n = 5) or 20 Gy (n = 25) and a dose rate of 5 Gy/min. Lung tissues were harvested at 3 (n = 5), 7 (n = 5), 12 (n = 5), 18 (n = 5), and 26 (n = 5) weeks after IR.

Hematoxylin and Eosin (H&E) Staining

Lung tissues were fixed and embedded in paraffin. After deparaffinization, 3 μm-thick slices were treated with citrate buffer (pH = 6.0) for 7 min. The lung tissue was stained with H&E. ²⁰ Simply, tissue sections were stained with Harris hematoxylin for 8 min. After differentiation in 1% acid alcohol and bluing in 0.2% ammonia water, cytoplasmic counterstaining was performed with eosin Y for 2 min. Sections were dehydrated, cleared in xylene, and mounted.

Masson’s Trichrome Staining

The 3 μm-thick slices were deparaffinized with xylene (2 × 5  min) and a descending alcohol series (100%, 90%, 80%, 70% and 50%) for deparaffinization. Next, a Masson’s Trichrome Stain Kit (Solarbio, Beijing, China) was used for staining as described previously. ²⁰ Simply, tissue sections were stained with Weigert’s hematoxylin for 10 min. After differentiation (1% acid alcohol), sections were treated sequentially with Biebrich scarlet-acid fuchsin, phosphomolybdic acid, and aniline blue. Dehydration, xylene clearing, and mounting completed the process.

Cell Culture and Irradiation

HUVECs were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA). Cells were cultured in ECM (Procell Life Science & Technology, Wuhan, China) and incubated at 37°C in a humidified atmosphere with 5% CO₂. Cells were exposed to IR (0, 2, 5, or 10 Gy) using an X-ray linear accelerator (RadSource, Suwanee, GA, USA) at a fixed dose rate of 1.15 Gy/min.

Endothelial Tube Formation Assay

A tube formation assay for HUVECs was performed according to the instructions. ²¹ Briefly, Matrigel was polymerized in a 96-well plate for 30 min at 37°C. Nonirradiated or irradiated HUVECs were incubated in 100 μL if ECM before image acquisition.

Immunofluorescence Assay

Cells were fixed in 4% paraformaldehyde and blocked 3% bovine serum albumin (BSA) for 1 h at room temperature. The cells were incubated with primary antibody at 4°C overnight. The next day, cells were incubated with secondary antibody for 1 h at room temperature. DAPI was used for nuclear staining, and images were observed by an FV1200 confocal microscope (Olympus, Tokyo, Japan).

Western Blot Analysis

HUVECs and lung tissues were collected and lysed in RIPA buffer on ice. After centrifugation, the supernatant was collected for Western blot. Protein was separated by 10% SDS-PAGE, transferred to polyvinylidene difluoride membranes (EMD Millipore, Bedford, MA, USA), and incubated with primary antibody at 4°C overnight. After washing with PBST, membranes were then incubated with the corresponding HRP-conjugated secondary antibodies for 1 h at room temperature. Finally, membranes were exposed to the ECL chemiluminescence chromogenic reagent (Beyotime, Nantong, China) and detected by a FluorChem^TM M System (Protein Simple, San Jose, CA, USA).

Enrichment Separation of Vascular Endothelial Cells of the Rat Lung

Anesthetized rats were fixed, and the abdominal skin was cut open and bluntly separated into the mouth. The muscle was completely separated along the direction of the trachea, a “T” mouth was cut, and the tracheal tube was inserted. The thoracic cavity of the rats was opened, the aorta and pulmonary arteries were fully exposed and arterial cannulation was performed. The vessels were flushed with PBS, the left atrial appendage was cut open to release blood, and air was immediately injected into the tracheal tube to open the alveoli and ensure clean blood. Biotin was infused from the arterial cannula for 20 min at room temperature. And then glycine was infused via an arterial cannula followed by flushing with PBS. Lung tissues were harvested, embedded in OCT and placed in 50 mL centrifuge tubes. Lung tissue blocks were homogenized and sonicated on ice. After 10% SDS was added, the supernatant was centrifuged for 24 h at 4°C. After dialysis, the homogenate was transferred to a 50 mL centrifuge tube and centrifuged 20 000 rpm for 1 h at 4°C and the supernatant was collected.

Proteomics

The homogenates from rats 3 weeks after 20 Gy irradiation and unirradiated rats were fully lysed on ice and denatured at 95°C for 5 min. The SDS-PAGE gel electrophoresis was performed and the colloids were removed and cut for labeling. Then Coomassie Blue G250 staining and Coomassie blue decoloring were then performed. The protein points analyzed from the gel were cut down for enzymatic digestion and ZipTip desalination. The eluted samples were used for matrix-assisted laser desorption ionization time of flight mass spectrometry. The protein sequence database was constructed with Pepsea administrator 2.2, and the mass spectrometry data from MALDI-TOF-MS were retrieved by MASCOT. The maps were searched by peptide mass fingerprinting.

RNA Sequencing

HUVECs were transfected with NEDD4 overexpression plasmid or empty plasmid, followed by 10 Gy irradiation. 72 h later, TRIzol reagent was used to extract total RNA from cells according to the manufacturer’s protocols. Aglient 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA) was used to access RNA integrity. RNA sequencing, Gene Ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were performed by the OE Biotech Company (Shanghai, China).

Quantitative Real-Time Polymerase Chain Reaction (qRT‒PCR)

Total RNA was extracted from HUVECs with TRIzol reagent (Invitrogen, Carlsbad, CA, USA). Extracted RNA was reverse transcribed to cDNA using the PrimeScript RT kit (Takara, Shiga, Japan), and qRT‒PCR was applied using the SYBR Green Master Mix Kit (Takara, Shiga, Japan) on the ABI ViiA 7 Real-Time PCR system (Applied Biosystems, Foster City, CA, USA). qRT‒PCR was carried out as follows: predenaturation, 95°C, 15 s, cycle number 1; PCR: denaturation, 95°C, 10 s; annealing/extension, 61°C, 32 s, cycle number 40. The obtained data were analyzed using the 2^−ΔΔCT method. The primer sequences were as follows:

NEDD4 Forward: 5′-GTG​CAG​ACT​CAC​CTT​GCA​GA-3′

NEDD4 Reverse: 5′-TTT​TTC​TTC​CCA​ACC​TGG​TG-3′

GAPDH Forward: 5′-GAC​ATG​CCG​CCT​GGA​GAA​AC-3′

GAPDH Reverse: 5′-AGC​CCA​GGA​TGC​CCT​TTA​GT-3′.

Coimmunoprecipitation (Co-IP)

HUVEC lysates were suspended in IP lysis buffer for incubation with antibodies against NEDD4 or ATM or negative control IgG at 4°C overnight. The next day, beads were mixed with the lysate-antibody complex for 2 h and then washed with IP lysis buffer before the immunoprecipitated proteins were analyzed by Western blot.

Statistical Analysis

All data are expressed as the mean ± SEM of at least 3 independent experiments. The data statistic was performed using Prism 8 software (GraphPad Software, La Jolla, CA, USA). Paired or unpaired t test (2 tails) was used to compare experiments with only 2 groups. One-way ANOVA with multiple comparisons was performed for experiments with more than 2 groups. Differences for which P < 0.05 were considered statistically significant.

Results

IR-Induced EndMT Accompanies the Development of RILI in vivo

To investigate the role of EndMT during RILI, a rat model of RILI was established via a single dose of 20 Gy to the unilateral thorax, and lung tissues were collected at 3, 7, 12, 18 and 26 weeks, as described previously. ¹⁸ As shown via H&E staining, pathological damage, including thickened interstitial edema, alveolar septa and infiltrated inflammatory cells, appeared in the irradiated groups (Figure 1A). Additionally, the release of IL-8 was markedly increased at 3 weeks after IR (Figure 1B). A pronounced increase in collagen fiber deposition in the lung tissue was observed in the irradiated rats compared to the nonirradiated rats (Figure 1C). Furthermore, EndMT-associated protein expression in rat lung tissues was examined by immunohistochemical staining. As shown in Figure 1D, CD31 expression was downregulated, α-SMA and Vimentin expression was upregulated in the lung tissue at 3 weeks after IR. In addition, Western blot analysis confirmed the IR-induced variations in the expression of EndMT-related markers (Figure 1E and F). These data confirmed the occurrence of EndMT during RILI in vivo.

Figure 1.

IR-Induced EndMT is Involved in the Development of RILI in vivo. (A) Representative Images of H & E-Stained Lung Tissues From a Rat Model of RILI. Scale bar = 100 μm. (B) The Expression of IL-8 in Rat Lung Tissues was Determined by Immunohistochemistry. Scale bar = 50 μm. (C) The Right Lung Tissues of the Rats Were Not Irradiated (Control Group) or Irradiated With 20 Gy X-Ray. Three Weeks after Irradiation, Lung Tissues Were Collected. Representative Images of Masson’s Trichrome-Stained Rat Lung Tissues. Scale Bar = 100 μm. (D) Representative Images Following Immunohistochemistry for CD31, Vimentin and α-SMA Protein Levels are Shown. (E) Western Blot Assay of VE-Cadherin and α-SMA Protein Levels. (F) Quantitative Analyses of VE-Cadherin and α-SMA Protein Levels (n = 5; Mean ± SEM; *P < 0.05 or **P < 0.01 Compared With the Control Group; One-Way ANOVA Test of Variance)

IR-Induced EndMT is Involved in the Development of RILI in vitro

Additionally, the role of EndMT during RILI was confirmed in vitro. HUVECs were exposed to 0 or 10 Gy X-rays. The number of cell tubes that formed was obviously reduced after IR exposure compared to that of the unirradiated cells, suggesting that the tube-forming capacity was obviously damaged (Figure 2A). The cell morphology was observed at 0, 24, 48 and 72 h after IR. With increasing IR dose and prolonged irradiation time, the cell morphology gradually changed from the stone-like form typical of endothelial cells to the long spindle-like form characteristic of stromal cells (Figure 2B and C). The same variations were demonstrated by immunofluorescence staining of phalloidin (Figure 2D and E). Correspondingly, CD31 expression was significantly decreased, and Vimentin and α-SMA expression was increased in a dose-dependent manner, as observed by Western blot analysis (Figure 2F and G). In addition, time-dependent decreases in CD31 expression and increases in Vimentin and α-SMA expression were identified in HUVECs (Figure 2H and I). Immunofluorescence also verified the decreased expression of CD31 after IR exposure (Figure 2J and K). These data showed that EndMT accompanies the development of RILI in vitro.

Figure 2.

IR-Induced Lung EndMT Accompanies the Development of RILI in vitro. (A) HUVECs Were Exposed to 0 or 10 Gy X-rays. The Tube-formation Capacity of HUVECs was Investigated at 2 h after Irradiation. Scale Bar = 500 μm. (B) HUVECs Were Exposed to 0, 2, 5 or 10 Gy X-rays. Representative Images Showing the Cell Morphology are provided at 72 h after Irradiation. Scale Bar = 200 μm. (C) HUVECs Were Exposed to 0 or 10 Gy X-rays. The Cell Morphology was Observed at 24, 48 and 72 h After Irradiation. Scale bar = 200 μm. (D) HUVECs Were Exposed to 0, 2, 5 or 10 Gy X-rays. The Distribution of Phalloidin in HUVECs was Detected by Immunofluorescence at 72 h after Irradiation. Scale bar = 20 μm. (E) HUVECs Were Exposed to 0 or 10 Gy X-rays. The Distribution of Phalloidin in HUVECs was Detected at 0.5, 6, 24, 48 and 72 h after Irradiation. Scale Bar = 20 μm. (F) HUVECs Were Exposed to 0, 2, 5 or 10 Gy X-Rays. The Protein Expression of Vimentin, α-SMA and CD31 in HUVECs was Examined by Western Blot at 72 h after Irradiation. (G) Quantitative Analyses of CD31, Vimentin and α-SMA Protein Levels (n = 3; Mean ± SEM; *P < 0.05 or **P < 0.01 Compared With the Control Group; One-Way ANOVA Test of Variance). (H) HUVECs Were Exposed to 0 or 10 Gy X-rays. The Protein Expression of EndMT-Related Markers in HUVECs was Detected at 0.5, 6, 24, 48 and 72 h after Irradiation. (I) Quantitative Analyses of CD31, Vimentin and α-SMA Protein Levels (n = 3; Mean ± SEM; *P < 0.05 or **P < 0.01 Compared With the Control Group; One-Way ANOVA Test of Variance). (J) HUVECs Were Exposed to 0, 2, 5 or 10 Gy X-rays. The Expression of CD31 was Examined by Immunofluorescence at 72 h After Irradiation. Scale Bar = 20 μm. (K) HUVECs Were Exposed to 0 or 10 Gy X-rays. The Expression of CD31 was Examined at 0.5, 6, 24, 48 and 72 h After Irradiation. Scale Bar = 20 μm

NEDD4 Mediates EndMT to Participate in the Regulation of RILI

Further clarifying the molecular mechanism by which EndMT participates in RILI, in the rat model of RILI prepared as mentioned above, lung tissue homogenates were collected from irradiated and nonirradiated rats. The vascular endothelial cell protein CD31 was pulled down from and enriched in the homogenate (Figure 3A). Biotin labeling was mainly concentrated in the endovascular lumen in rat lung tissues. Immunoprotein blotting showed that compared with that in the input sample, CD31 expression was obviously increased in the homogenate obtained through pulmonary circulatory perfusion. That is, enriched vascular endothelial cell proteins were successfully obtained after pulmonary circulatory perfusion. ²² Eighty-five proteins expressed in only the irradiated rats were successfully selected by proteomics (Supplemental Table 1). The key differentially expressed proteins, including S100 calcium-binding protein A4 (S100A4), nucleoli (NCL) and NEDD4 E3 ubiquitin protein ligase (NEDD4), were identified (Table 1). Furthermore, expression of the 3 candidate proteins in rat lung tissues were evaluated by Western blot. The expression of NEDD4, but not that of S100A4 or NCL, was significantly increased in the irradiated lung tissues at 3 weeks after IR (Figure 3B and C). Western blot analysis showed that NEDD4 expression in the HUVECs increased after IR exposure in a dose- and time-dependent manner (P < 0.01) (Figure 3D–G), suggesting that NEDD4 plays a vital role in the early stage of RILI. These results collectively suggested that the expression of endothelial cell-derived NEDD4 could be induced by IR both in vitro and in vivo.

Figure 3.

NEDD4 Mediates EndMT to Participate in the Regulation of RILI. (A) Biotin Labeling and Homogenate Pull - Down Enrichment. (B) Western Blot was Used to Evaluate S100A4, NCL and NEDD4 Expression. (C) Quantitative Analyses of S100A4, NCL and NEDD4 Protein Levels (n = 3; Mean ± SEM; **P < 0.01 Compared With the Control Group; t test of Variance). (D and E) HUVECs Were Exposed to 0, 2, 5 or 10 X-rays. NEDD4 Protein Expression was Measured at 72 h after Irradiation. (F and G) HUVECs Were Exposed to 0 or 10 X-rays. NEDD4 Protein Expression was Measured by Western Blot Analysis and Quantified at 0.5, 6, 24, 48 and 72 h after Irradiation (n = 3; Mean ± SEM; **P < 0.01 Compared With the Control Group; One-Way ANOVA Test of Variance)

Table 1.

The Key Dysregulated Proteins of Proteomics Analysis

No.	Gene name	Accession	Description	Peptides (95%)
1	NCL	P13383	Nucleolin	6
2	NEDD4	Q62940	E3 ubiquitin-protein ligase NEDD4	3
3	BASI	P26453	Basigin	3
4	APMAP	Q7TP48	Adipocyte plasma membrane-associated protein	3
5	S100A4	P05942	S100 Calcium Binding Protein A4	2

To explore whether NEDD4 is involved in IR-induced EndMT, a HUVEC model with NEDD4 overexpression was constructed, and NEDD4 overexpression was confirmed by Western blot (Figure 4A and B). Subsequently, a tube-formation assay showed that the tube-formation capacity of HUVECs decreased after IR exposure; however, NEDD4 overexpression further reduced the tube-forming capacity (Figure 4C). As shown in Figure 4D, NEDD4-overexpressing cells exhibited a swollen and elongated morphology after 10 Gy IR exposure. Furthermore, HUVECs that were transfected with NEDD4 overexpression plasmid displayed decreased CD31 expression and increased Vimentin expression with or without IR exposure (Figure 4E–H). The above results collectively suggest that NEDD4 can promote EndMT in endothelial cells.

Figure 4.

NEDD4 Participates in EndMT in Endothelial Cells. (A) NEDD4 Protein Levels in NEDD4 - Overexpressing HUVECs. (B) Quantitative Analyses of NEDD4 Protein Level (n = 3; Mean SEM; **P < 0.01 Compared With the Control Group; t test of Variance). (C) NEDD4 Overexpression Reduced the Tube-formation Capacity of HUVECs. Scale Bar = 500 μm. (D) Phase-Contrast Micrographs of Control and NEDD4-Overexpressing HUVECs. Scale Bar = 200 μm. (E–H) CD31 and Vimentin Protein Expression in Control and NEDD4-Overexpressing HUVECs (n = 3; Mean ± SEM; *P < 0.05 or **P < 0.01 Compared With the Control Group; t test of Variance)

NEDD4 Mediates EndMT to Participate in RILI Through the ATM Signaling Pathway

To delineate how NEDD4 mediates IR-induced EndMT during RILI, HUVECs were transfected with NEDD4 overexpression plasmid or empty plasmid and then exposed to 10 Gy irradiation, and RNA was collected for RNA sequencing 72 h later. A total of 208 upregulated genes and 45 downregulated genes were successfully identified (Figure 5A and B). To further analyze the differentially expressed genes, they were divided into functional groups based on the results of Gene Ontology (GO) enrichment analysis. Among the GO terms that were obtained, the most dominant among biological process terms included chemokine-mediated signaling pathways, jaw development, endochondral ossification, the cellular response to tumor necrosis factor, and skeletal system development (Figure 5C). In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis demonstrated that the differentially expressed genes are related to PI3K-Akt signaling, focal adhesions and the ECM-receptor interaction (Figure 5D). PI3K-Akt signaling showed the greatest enrichment (Figure 5D). The transcriptomic results suggested that ATM expression increased with the overexpression of NEDD4. As an essential member of the PI3K-associated protein kinase family, ATM in the PI3K-Akt signaling pathway may play an important role in NEDD4-mediated IR-induced EndMT. ²³ Then, qRT‒PCR was used to verify the sequencing results (Figure 5E). These results illustrated that overexpression of NEDD4 promoted IR-induced EndMT and increased the expression of ATM. We next examined the potential interaction between NEDD4 and ATM. NEDD4-overexpressing HUVECs were exposed to 10 Gy X-rays, and Co-IP was performed 72 h later. As shown in Figure 5F, NEDD4 could specifically bind the ATM protein, and ATM could specifically bind the NEDD4 protein, suggesting an interaction between NEDD4 and ATM.

Figure 5.

NEDD4 Stimulates EndMT to Participate in RILI Through the ATM Signaling Pathway. HUVECs Were Treated With 10 Gy X-rays Alone or in Combination With NEDD4 Overexpression Plasmid. (A) Volcano Plot Showing the Gene Profiles of the NEDD4-Overexpressing and 10 Gy Irradiated Control Groups. (B) Differentially Expressed Gene Distribution in HUVECs After NEDD4 Overexpression. Red and Blue Bars Represent Upregulated and Downregulated Differentially Expressed Genes, Respectively. (C) GO Term Enrichment in the Differentially Expressed Genes From HUVECs is Shown. (D) KEGG Enrichment Analysis of the Significantly Differentially Expressed Genes Between the 2 Groups is Shown. (E) ATM Expression was Verified by qRT‒PCR (n = 3; Mean ± SEM; **P < 0.01 Compared With the Control Group; t test of Variance). (F) HUVECs Were Subjected to a Co - IP Assay Using IgG, NEDD4 or ATM Antibodies, Followed by Western Blot Analysis to Detect the NEDD4-ATM Interaction. (G, H) Protein Expression of Genes Upstream and Downstream of ATM in Nonirradiated or Irradiated HUVECs. (I) Schematic Representation of the NEDD4/ATM in IR-Induced EndMT

Then, Western blot was used to measure the effect of NEDD4 overexpression on the expression of genes upstream and downstream of ATM in HUVECs. NEDD4 overexpression decreased the expression of members of the MRN (MRE11/RAD50/NBS1) complex, which are genes upstream of ATM, whereas expression in CHK2, p-CHK2 and p53, which are genes downstream of ATM, showed almost no variation (Figure 5G). However, in response to IR exposure, NEDD4 overexpression increased the expression of genes upstream and downstream of ATM (Figure 5H). Taken together, these results confirmed that NEDD4 and ATM interact and that NEDD4 mediates EndMT to participate in RILI through the ATM signaling pathway.

Discussion

RILI, which is characterized by acute lung inflammation and irreversible lung fibrosis, has been identified as a major dose-limiting factor for radiotherapy of chest malignancies.^1,2,24-26 To date, no therapeutic approach for RILI is available, and the underlying mechanism of RILI remains unclear. EndMT, a specialized type of EMT that was initially observed during cardiac development, plays a key regulatory role in fibrotic diseases. ²⁷ Recently, HIF1α-related EndMT was observed in mouse and human RIPF tissues. ²⁸ Furthermore, EndMT has been reported to be related to RILI, and F-actin stress fibers have been identified as a phenotypic marker of IR-induced EndMT. ²⁹ However, Song et al failed to reveal the in-depth mechanism underlying the involvement of EndMT in RILI. ²⁹ Therefore, further investigation of EndMT will help to determine whether it can modulate RILI to reduce late tissue damage after IR exposure.

In the present study, we have illustrated the role of EndMT in RILI. We found that IR induced EndMT in RILI in vivo and in vitro, accompanied by a decrease in the expression of endothelial markers and an increase in the expression of interstitial markers (Figures 1 and 2). Considering the importance of EndMT in RILI, we further clarified the molecular mechanism of EndMT by proteomics. NEDD4 was then selected as a candidate target due to the dose- and time-dependent increase in its expression in the early stage of RILI in vitro and in vivo. The NEDD4 family contains 9 members: NEDD4-1 (RPF1), NEDD4L (NEDD4-2), ITCH/atropine-1 interaction protein 4 (AIP4), WW domain-containing E3 ubiquitin protein ligase 1 (WWP1), WWP2/Atropine-1-interacting protein 2 (AIP2), NEDL1 (HECW1), NEDL2 (HECW2), SMAD-specific E3 ubiquitin protein ligase 1 (SMURF1) and SMURF2. ³⁰ NEDD4 has been shown to be involved in the ubiquitination of a variety of proteins involved in a variety of processes, including in anti-inflammatory processes, wound repair, the regulation of heart failure, and EMT. ³¹ Furthermore, downregulated NEDD4 expression was found to mediate TGF-β-induced EMT in lung cancer cells, ³² and in TGF-β1-induced alveolar epithelial cells, and silencing NEDD4L promoted EMT and fibrosis in alveolar epithelial cells. ³³ In nasopharyngeal carcinoma cells, the upregulation of NEDD4 was found to be related to EMT. ³⁴ Additionally, NEDD4 has been reported to participate in diverse cellular processes related to epithelial homeostasis and has been implicated in the development of chronic lung disease and fibrosis when dysregulated. ³⁵ In our study, NEDD4 promoted EndMT in endothelial cells, consistent with the findings of Feng et al in nasopharyngeal carcinoma cells. ³⁴ NEDD4 has been reported to participate in diverse cellular processes. For instance, NEDD4 was shown to play a key role in myocardial reperfusion injury and cardiac vascular calcification in mice, effectively promoting the regeneration and repair of the heart. ³⁶ In addition, NEDD4 was implicated in the regulation of epithelial Na⁺ channels, which are critical for airway surface hydration and mucus clearance, and TGF-β signaling regulation, which promotes fibrotic remodeling. ³⁵ Nevertheless, the biological function of NEDD4 in the pathogenesis of RILI remains to be elucidated.

Furthermore, the detailed mechanism by which NEDD4 mediates IR-induced EndMT during RILI was delineated. KEGG enrichment analysis demonstrated that NEDD4-mediated EndMT during RILI may be related to the PI3K-AKT signaling pathway, a key sensor of genome integrity that promotes DNA double-strand repair. ³⁷ As an essential member of the PI3K-associated protein kinase family, the ATM protein is vital for the DNA damage response, can regulate the DNA repair signaling network, and participates in cell cycle arrest and even apoptosis. ³⁸ Importantly, NEDD4 is required for the ubiquitination and nuclear trafficking of ATM activators during the DNA damage response. ²³ It is worth noting that ATM participates in the occurrence and development of EMT.^39,40 In addition, Bouten’s study found that ATM and EndMT-associated genes increase after radiation in endothelial cells. This study revealed a time course of integrated gene expression and protein activation leading from early DNA damage response and cell cycle arrest to senescence, pro-inflammatory gene expression, and EndMT. ⁴¹ Therefore, we hypothesized that ATM is involved in the regulation of IR-induced EndMT in vascular endothelial cells. We found by RNA-seq and qRT‒PCR that ATM expression in irradiated vascular endothelial cells was increased with the upregulation of NEDD4 and confirmed the interaction between NEDD4 and ATM through Co-IP. Notably, the apparent contradiction—where NEDD4, an E3 ubiquitin ligase, fails to promote ATM degradation despite potential ubiquitination—can be explained by the following mechanisms: (1) NEDD4 may mediate non-degradative ubiquitination (e.g., K63-linked chains) of ATM, which is associated with signaling modulation rather than proteasomal targeting. Such modifications could stabilize ATM by enhancing its interaction with DNA damage repair complexes (e.g., BRAT1) or promoting its nuclear retention, as seen in Low’s study. ²³ (2) NEDD4 might compete with other E3 ligases (e.g., MDM2) for ATM binding. If NEDD4 occupies ATM ubiquitination sites without adding K48-linked chains (required for degradation), it could block degradation pathways while amplifying ATM’s activation. (3) NEDD4 might compete with other E3 ligases (e.g., MDM2) for ATM binding. If NEDD4 occupies ATM ubiquitination sites without adding K48-linked chains (required for degradation), it could block degradation pathways while amplifying ATM’s activation. As with Low’s study, ²³ we guess that NEDD4 might require adaptors like Ndfip1 to recognize ATM. Ndfip1 could recruit NEDD4 to ATM, enabling ubiquitination. However, whether NEDD4 can directly ubiquitinate ATM needs to be further verified. In the present work, the upregulation of NEDD4 significantly increased the expression of genes upstream and downstream of ATM in response to IR, suggesting that after DNA damage, DNA double-strand breaks are recognized by the MRN complex, which recruits ATM to promote the phosphorylation of downstream CHK2 and the activation of p53.

Conclusion and Limitation

In summary, we have demonstrated that vascular endothelium-derived NEDD4 may play a biological role in the occurrence and development of IR-induced EndMT during RILI through the ATM signaling pathway (Figure 5I). These observations reveal a novel, critical role of NEDD4-mediated EndMT and provide novel strategies to ameliorate RILI. Nevertheless, further in-depth analyses to reveal the complete NEDD4-associated regulatory network in RILI are indispensable. For example, as NEDD4 is a ubiquitination-regulatory molecule, whether the molecular interaction between NEDD4 and ATM is related to the regulation of ubiquitination and how this interaction regulates IR-induced EndMT in endothelial cells deserve further exploration.

Supplemental Material

Supplemental Material - NEDD4-Mediated Endothelial-Mesenchymal Transition Participates in Radiation-Induced Lung Injury Through the ATM Signaling Pathway

Supplemental Material for NEDD4-Mediated Endothelial-Mesenchymal Transition Participates in Radiation-Induced Lung Injury Through the ATM Signaling Pathway by Yang Feng, Lirong Zhang, Youbin Zhang, Ying Xu, Kaixiao Zhou, Zhao Yang, Wei Zhu, Qi Zhang, Jianping Cao, Lili Wang and Yang Jiao in Dose-Response.