NPR3 inhibits EMT and Tumour Metastasis of Human Osteosarcoma by Activating AKT/mTOR-mediated Autophagy

Osteosarcoma is an aggressive primary malignant cancer of bone mainly occurring in adolescence with a characteristic of high metastasis and relapse rate. In our previous study, we rst identied that NPR3 was signicantly decreased in OS samples. Here, we purposed to investigate the effect and the possible mechanisms of NPR3 on the progression of human OS. The expression of NPR3 in OS patients and cells was detected by qRT-PCR, and IHC analysis. The effect of the expression of NPR3 on tumour metastasis was examined in vitro and in vivo. The molecular mechanisms of the regulation of NPR3 were evaluated in vitro and in vivo. The clinical relevance of 5-year overall survival with the expression of NPR3 was evaluated in 294 patients with OS. presented as mean ± standard deviation from three independent


Background
Osteosarcoma (OS) is an aggressive primary malignant cancer of bone mainly occurring in adolescence with the characteristic of high metastasis and relapse rate compared with other bone tumors [1]. In clinical, those patients were often diagnosed in an advanced stage, and almost 90% of patients died from pulmonary metastases [2]. Despite the use of the neoadjuvant chemotherapy, the 5-year survival of OS with metastatic disease is approximately 20% [3,4]. Currently, the overall clinical prognosis of patients with OS is poor, moreover there are no effective targeted agents approved for OS. Hence, there is an urgent need to explore the new key protein biomarkers for metastatic or relapsed OS to improve their prognosis.
The mechanisms of OS development are complex, but some progress has been made. Efforts in the past few decades have focused on the mechanism of some non-coding RNAs (ncRNAs) [5], autophagy [6] and EMT pathways [7] in the metastasis of OS, as well as in sporadic OS. It is noteworthy that some potential target genes are mainly concentrated in the PI3K/Akt/mTOR pathway, making PI3K/AKT/mTOR pathway identi ed as potentially valuable therapeutic target exploitation. Coincidentally, small molecule inhibitors for this pathway have shown encouraging results [8]. In our previous study, we identi ed a differentially low expressed target gene NPR3 in OS patients using meta-analysis for the rst time [9]. However, the clinical value of NPR3 in OS patients with metastasis remains elucidated. NPR3, located at human chromosome 16p13.3, is a member of the natriuretic peptide receptor family expressed in numerous normal tissues, including the skeletal muscle, heart, kidney, liver, and pancreas [10]. Accumulated research reported that NPR3 was play critical roles in cartilage homeostasis, preserves cardiac function [11][12][13]. The latest study mentions its role in colorectal cancer (CRC) [14] and clear cell renal cell carcinoma (ccRCC) metastasis [15]. Intriguingly, advanced studies have provided the evidence proving the relationship between NPR3 and PI3K/AKT pathway, autophagy pathway [16,17]. Thus, it is very essential to explore the regulatory mechanisms of this connection in tumorigenesis and progression of human OS.
In this paper, we rst explored the function role of NPR3 in the development of OS. Furthermore, we intend to explored whether NPR3 affects EMT and autophagy of OS through the AKT/mTOR pathway. This study is expected to provide a new idea for the treatment of OS.

Patients and clinical specimens
In this study, a total of 122 fresh primary bone tumour tissues with metastases and 172 bone tumour tissues without distance tissues metastases of OS patients who underwent surgery were collected from Yunnan Cancer Hospital, Kunming, China (December 2013 to December 2020). These tissues were stored at − 80°C immediately after surgery. All the resected tissues were identi ed by pathologic examination. RNA level of NPR3 was measured by those samples. All samples were staged according to Enneking surgical staging [18]. The median fellow-up time was 22 months. Overall survival was de ned as the interval between the date of surgery and death or the end point of follow up. Para n sections of 19 OS cases with and without lung metastases were obtained from the Department of Pathology of the Yunnan Cancer Hospital, Kunming, China (May 2019 to December 2020) including the primary bone tumour tissues, para-cancer tissues. Protein levels of NPR3 was measured by those para n sections. Written informed consent was obtained from all patients. This study was approved by the Ethics Committee of Yunnan Cancer Hospital.

Cell culture and transfection
Human normal osteoblastic cell hFOB, and immortalized OS cell lines U2OS, HOS-mnng were purchased from Shanghai Institutes for Biological Sciences, Chinese Academy of Science (Shanghai, China). And the MG63 and Saos-2 cells were obtained from the Cell Resource Centre of Kunming institute of Zoology, Chinese Academy of Science (Kunming, China). All cell lines were cultured in Dulbecco's Modi ed Eagle Medium (DMEM, Gibco, CA) with 10% fetal bovine serum (FBS) and 1% streptomycin/penicillin at 37°C, 5% CO 2 . The plasmids pcDNA3.1-EGFP-NPR3 and pcDNA-EGFP-NC were purchased from the Honor-Gene company to overexpress NPR3. The cells stably over-expressing (OE) of NPR3 was constructed in our lab. The transfection experiments were performed according to the manufacturer's instructions included with Lipofectamine 2000 TM (Invitrogen). Stable cell lines were cultured in complete medium supplemented with 600µg/ml G418.

Immunocytochemistry analysis (IHC)
A total of 57 para n-embedded OS primary bone tumour samples from 19 patients with OS were collected from the Department of Pathology, Yunnan Cancer Hospital. After that, the slides were incubated with 1:500 dilution of antibody to NPR3 (Abcam) at 4°C overnight and rinsed with phosphatebuffered saline (PBS). Then goat anti-rabbit secondary antibody labeled polymer-HRP was incubated for 1h in dark. After wishing with PBS, the color reaction was developed by 3,3'diaminobenzidinetetrachloride (DAB), followed by counterstained with hematoxylin. The negative control is the IgG isotype antibody. Samples were examined with a uorescence microscope (Nikon). Then we scored the staining intensity by two independent pathologists. The IHC scored divided into 4 grades as follows: 0 (negative, < 5%), 1 (weak, 5-25%), 2 (moderate, 25%-50%), 3 (strong, 50%-75%), 4(strong, > 75%). p values were based on the two-sided statistical analysis and p-value less than 0.05 was considered to be statistically signi cant.

qRT-PCR
Total RNA was extracted from cultured OS cell lines or specimens using the TRIzol reagent (Promega). The reverse transcription and the qRT-PCR were carried out as previously described [19]. The total RNA was reverse-transcribed using a First Strand cDNA Synthesis Kit (Promega). qRT-PCR was performed using the SYBR qPCR Master Mix (Promega). Expression data were normalized to the geometric mean of the housekeeping gene GAPDH to control the variability in expression levels and calculated using the The primers of NPR3 was used for qPCR as follows, 5-ACCCTGGACCGTATACCCTGA-3 (forward), 5-CACCACTTTCTCGTTCTGGCC-3 (reverse).

Wound healing assay
Both non-transfected and stably-transfected MG63 and Saos-2 cells were cultured in 12-well plates respectively, media was changed to 0.1% FBS DMEM before scratching. Scratch wound was generated using a 10µl micropipette tip after 12h, and then photographs were captured at indicated time points for 0, 12h. The percentage of wound healing was calculated using the Image J software. Wound healing rate (%) = (initial wound area − non-healing area)/initial wound area.

Matrigel Invasion Assay
The assay was carried out using the BD Bio-coat Matrigel Invasion Chamber with inserts of 8-µm pore size for 24-well plates (BD Biosciences) as the manufacturer's instructions. Brie y, the OS cells were suspended with medium without FBS, then 5×10 4 cells were plated in the upper chamber in 200ul serumfree medium. The bottom chamber contained complete medium with 10% FBS. Undergoing migration for 12h, the cells that had invaded through the membrane to the lower surface were xed with 4% paraformaldehyde and stained with 0.1% Crystal violet for 1 h at room temperature. The cell numbers were counted in ve random elds of each chamber under the microscope (Nikon).

Cell viability assays
Cell proliferation was measured by the Cell Counting Kit-8 (CCK-8) assay following the manufacturer's instructions. Brie y, the cells at a density of 5 × 10 3 cells / well were cultured in 96-well plates. The mixture was then treated with CCK-8 reagent. The absorbance at 450 nm was measured to determine the cell proliferation viability.

Cell apoptosis assay
To investigate the effect of NPR3 on cell apoptosis, apoptosis was evaluated by ow cytometry (FC) using the annexin V apoptosis detection FITC Kit. Data obtained from early and late apoptosis are depicted together as apoptosis in the results. The apoptotic effect on OS cells was con rmed by Annexin V-FITC/PI double staining. Statistical analysis was performed using the Mann-Whitney U test.

Xenograft tumour model
To evaluate the effect of NPR3 on primary tumour growth and spontaneous metastasis, we do the animal model. The male BALB/C nude mice (4 weeks old, purchased from Beijing Vital River Ltd., China) were injected with 2×10 6 /50µL MG63-OE-NPR3 MG63 cells or OE-NC MG63 cells in tibial medullary cavity in each group. One week later, mice were divided into different groups (n = 15) and injected with normal saline or Baf-A1 (75 nm/kg/d) every day for 3 weeks. Tumour volume and weight were measured beginning 5 weeks post injection and measured every 2 days until killing at 8 weeks. On day 56, the IVIS Lumina Imaging System (Xenogen) was used to capture the signal by taking photographs. The tumours were excised, weighed, photographed, and processed for WB, IHC assays. Lungs, livers, kidney were processed for routine HE staining. And, the number of metastatic nodules in the lung was quanti ed. All experimental procedures were approved by the animal care and use committee of Kunming Medical University.

Statistical analysis
All statistical analyses were carried out using GraphPad 7.0 software (San Diego, CA) and SPSS 25.0 (Chicago, IL, USA). All results were reported as mean ± SD from at least three separate experiments. The associations between NPR3 expression and the clinicopathological characteristics of the patients were analyzed using the Chi-squared test. Statistical analysis between different groups were performed with one-way ANOVA, between two groups were analyzed by Student's t test. The overall Survival curves were plotted using the Kaplan-Meier method and compared using the log-rank test. p values < 0.05 was considered statistically signi cant differences.

NPR3 was signi cantly down-regulated in metastatic human OS patents
Our previous work identi ed a differentially down-regulated expression protein NPR3 in OS patients using meta-analysis for the rst time [9]. To better understand the effect of NPR3 on the metastasis of OS, we examined the protein and mRNA levels of NPR3 in OS tissues with and without distance metastases. Interestingly, we observed that NPR3 was dramatically lower expression in tissues with metastases than in tissues without metastases at both the mRNA level and protein level (Fig. 1a, 1b,1d). The associated AUC in the ROC analysis was as high as 0.840 (0.7951-0.8844, p < 0.001) (Fig. 1c). Moreover, we found that NPR3 was down-regulated expression in MG63 and Saos-2 than in the two other OS cell lines and 1 normal bone cell lines by qRT-PCR assay (Fig. 1e). Therefore, we selected the MG63 and Saos-2 to carry out the subsequent experiments in this research. These results suggested that NPR3 may promote the metastasis of OS.
Down-regulated expression of NPR3 was correlated with poor prognosis of OS patients The correlation between NPR3 express and clinicopathologic characteristic was shown in Table 1. The analysis of the relationship between NPR3 expression level and clinical features of OS patients was shown in Table 2. We found that the expression level of NPR3was correlated with clinical stage (p = 0.016), tumour metastasis (p < 0.001) and tumour recurrence (p = 0.002). No difference was found for age, gender, tumour size, BMI, primary tumour site or pathological fracture. The risk factors associated with the overall survival of OS patients were shown in Table 3. Moreover, Kaplan-Meier survival analysis indicated that the clinical stage, metastasis and NPR3 expression level were the independent prognosis factor for the overall survival of OS. And the patients with low level of NPR3 had poorer overall survival than those with high expression level of NPR3 (p < 0.001) (Fig. 1f). Together, these results strongly suggested the NPR3 was signi cantly correlated with the prognosis of OS.   Overexpression NPR3 inhibited the proliferation, migration and invasion of MG63 and Saos-2 Next, we evaluated whether NPR3 regulated the proliferation, migration and invasion of OS. We generated stably-transfected cells with overexpression NPR3 expression in MG63 and Saos-2, and subsequently detected by qRT-PCR and WB after 48h (Fig. 2a, b). Then the proliferative rate, the migration and invasive ability of OS cells were tested. The results showed that NPR3 overexpression dramatically suppressed the proliferation of in OS cell lines MG63 and Saos-2 (Fig. 2c, d). Moreover, in transwell and matrigel assays, the overexpression of NPR3 signi cantly decreased the migration and invasive cell numbers compared with the control group and OE-NC group (Fig. 2e, f, g, h). These results demonstrated that overexpression of NPR3 inhibited the proliferation, cell migration and invasion of OS cells in vitro.
Over-expression of NPR3 inhibited the tumour growth and metastasis of OS in vivo We next investigated the effect of NPR3 in vivo. Stably-transfected MG63 cells were injected subcutaneously into tibial medullary cavity (left and right) of nude mice. The animals were monitored for tumour growth and distance metastases from 5 weeks. We observed that lung metastases were found in all ve mice (15/15) from the OE-NC group, whereas 66% nude mice were found having lung metastatic nodes in the OE-NPR3 group (10/15) (Fig. 3a). Compared to the control group, tumors were signi cantly smaller in volume and weight in OE-NPR3 group (Fig. 3b, c, d, e). The lung, liver and kidney tissue were collected to determine the effects of NPR3 on tumour metastasis (Fig. 3f). The results revealed that the degree of distant tissue metastasis was signi cantly higher in the OE-NPR3 group than in the OE-NC group (p < 0.001). These results con rmed the role of NPR3 in inhibiting tumour growth and distant metastasis in vivo.

Over-expression of NPR3 induced cell death and inhibited EMT in OS cells
Generally, metastasis is mainly carried out in four stages: cell survival, passing through the matrix into the circulatory system, colonizing into the distant metastatic sit, and activating. In this process, tumour cells often undergo EMT, then invade into the circulatory system [21]. Thus, we examined the effect of NPR3 on EMT and cell death. The WB assay revealed that overexpression of NPR3 in MG63 cells signi cantly the increased the expression level of epithelial cell marker E-cadherin, but decreased the expression levels of EMT-related proteins including N-cadherin, Vimentin, Snail, Twist1 and ZEB1 (Fig. 4a,  b). Moreover, overexpression of NPR3 increased the expression level of cell apoptosis proteins Bax and caspase 3 and decreased the expression of Bcl-2 (Fig. 4c, d). Cell death of Saos-2 and MG63 cells was determined by staining with Annexin V-FITC and PI. Flow cytometry data con rmed that overexpression of NPR3 induced cell apoptosis and necrosis (Fig. 4e, f). These data above suggested that overexpression of NPR3 inhibited the metastasis of OS through inducing the cell death and suppressing the EMT.

Effect of NPR3 on cell apoptosis and EMT may depends on the cell autophagy
Previous studies reported autophagy effects on cell apoptosis [22]. Therefore, we assessed the possible association between NPR3 and cell autophagy in OS cell lines. We rst investigated whether NPR3 could induce the cell autophagy. We found that the percentage of GFP-LC3 punctapositive cells was signi cantly increased in the OE-NPR3 group compared with the control group (p < 0.05) (Fig. 5a, b). As expected, Overexpression of NPR3 increased the expression of LC3 while decreased the expression of p62 (Fig. 5c). Furtherly, to further investigate the underlying molecular basis for NPR3-mediated autophagy, we found that overexpression of NPR3 reduced phosphorylation levels of AKT and mTOR while no signi cant effect on total protein levels of AKT and mTOR in MG63 and Saos-2 cells (Fig. 5d). Collectively, these results indicated that NPR3 may act as a potential autophagy promoter partially depend on the AKT/mTOR pathway.

Autophagy inhibitors Baf-A1 and CQ attenuated the inhibitory effect of NPR3 on tumour invasion in vitro, tumour growth and metastasis in vivo
Finally, loss-of-function experiments were performed to con rm the interaction between autophagy and NPR3 in OS cells. As shown in Fig. 6(a, b, c, d), wound-healing assay showed that autophagy inhibitors Baf-A1 and CQ reversed the inhibitory effect of OE-NPR3 on inhibiting tumour invasion. Stable OS cells overexpressing NPR3 were injected into nude mice. At the indicate times, pulmonary metastasis, BLI, Micro-CT assay and X-ray were used to measure the primary tumour growth and lung metastasis in nude mice deal with OE-NPR3 and OE-NPR3 + Baf-A1. The results showed that OE-NPR3 inhibited the tumour growth and distance metastasis in nude mice, while the autophagy inhibitor Baf-A1 reversed the function of NPR3 overexpression (Fig. 6e, f, g, h, i, j, k). These results con rmed that NPR3 inhibit the tumour growth and distance metastasis mainly depend on the autophagy pathway.
Autophagy inhibitors Baf-A1 attenuated the effect of NPR3 on the EMT Also, autophagy inhibitors reversed inhibited-EMT effect of OE-NPR3 in MG63 cells (Fig. 7a, 7c) and in Saos-2 cell (Fig. 7b, 7d). These results con rmed that overexpression of NPR3 inhibited the EMT partly depended on the autophagy.

Discussion
OS is an aggressive cancer with the characteristic of high metastatic rate and high rate of relapse. It is important to comprehensive analysis the potential mechanism in the development of OS. In our previous study, we identi ed a differentially low expressed target gene NPR3 in OS patients using meta-analysis for the rst time [9]. Interestingly, we identi ed that patients with downregulated of NPR3 have higher metastasis rates and poorer prognosis than those with high expression of NPR3. Then, we examined the biologic function of NPR3 on OS. We found that the proliferation and invasion abilities were inhibited in OE-NPR3 cells compared with the control. In vivo, overexpression of NPR3 decreases the incidence of metastasis. Excitingly, we elaborated that low expression of NPR3 can promote distance tissues metastasis of OS, mainly by inhibiting autophagy-induced apoptosis and EMT. And this process partly depended on the AKT/mTOR pathway. Those results indicated that NPR3 may be a potential tumour suppressor protein for OS.
The previous studies reported that NPR3 was play critical roles in cartilage homeostasis, and regulate blood pressure [23][24][25][26]. However, there are few reports on the role of NPR3 in the tumourigenesis.
Emerging studies showed NPR3 possess anti-tumour activity in some cancers. NPR3 signi cantly inhibited the proliferation and apoptosis of colorectal cancer cells [14]. Li et al. reported that NPR3 promotes Clear cell renal cell carcinoma metastasis via activating p38-MAPK signaling [15]. We investigated the anti-tumour effect of NPR3 on human OS for the rst time [9]. The results strongly indicated that low expression of NPR3 is likely to correlate with metastasize and poor survival.
Meanwhile, our data demonstrates that down-regulation of NPR3 inhibited autophagy and apoptosis, promoted EMT partly through the PI3K/AKT/mTOR pathway.
EMT-like phenotype is the characteristic of tumors with metastatic ability, which often occurs in nonepithelial tumors. Kahlert et al proved that a mesenchymal transition process involved in the onset and progression of sarcomas. More evidence showed OS cells undergo the EMT process that associate with their metastatic ability [27][28][29]. Consistently, we found NPR3 overexpression led to increase the expression level of E-cadherin and decrease the expression level of EMT-activating proteins, such as Ncadherin, vimentin, snail, twist1, ZEB1 which increased migratory and invasive potential [30]. These results suggested that NPR3 regulated the invasion of OS depend on EMT.
Autophagy, as a cancer-suppressive function, could inhibit metastasis by selectively down-regulating critical transcription factors of EMT [31]. Also, Iml1p-Npr2p-Npr3p complexes may regulate the formation of autophagosomes [32], which was strongly impaired in NPR3 defective cells or Arabidopsis thaliana [33]. Therefore, we monitored the regulation of NPR3 on cell autophagy. The data con rmed that overexpression of NPR3 could induce autophagy, including increased GFP-LC3II dot number, promote LC3I changed to LC3II, reduced p62 level. PI3K/AKT/mTOR pathway may be a key modulator of autophagy, which plays a pivotal role in tumour metastasis [34]. The activate of AKT have the ability to promote protein synthesis and cell growth through activating mTOR [35]. And the mTOR has the function to promote cellular macromolecular synthesis and suppresses autophagy [36]. Many studies suggest that aberrant activation of the AKT/mTOR pathway is associated with human OS [37][38][39]. Addition studies have demonstrated that various signaling pathways involved in the progression of OS, including Wnt, PI3K/AKT, and TSC1-TSC2 [40,41]. In consistent with this view [16], our present results also revealed that NPR3 exposure inactivated phosphorylation of mTOR/AKT and induced autophagy. We further used Baf-A1 and CQ to inhibit autophagy, and reversely verify the correlation between autophagy and NPR3. The autophagy inhibitor Baf-A1 and CQ blocked OE-NPR3-inhibited migration, invasion and EMT in OS.
These results established that down-regulation of NPR3 induces migration/invasion and EMT in OS partly through the AKT/mTOR pathway.

Conclusions
In this study, we reported a novel target gene which is down-regulated in metastatic OS and associated with poor prognosis of patients with OS. Further functional studied showed that overexpression of NPR3 inhibited migration and invasion ability of OS in vitro and tumour distant tissue metastasis in vivo.
Mechanistically, we found that NPR3 induced autophagy partly via AKT/mTOR pathway. Conclusively, NPR3 inhibited OS metastasis by suppressing EMT via the AKT/mTOR mediated autophagy. Therefore, Targeting NPR3 might represent a therapeutic and preventive target against OS.  Data are presented as the mean ± standard deviation. *p < 0.05. (c) Evaluation of differential diagnosis of tumour metastases using the expression of NPR3 by ROC curve analysis. and subsequently subjected to quantitative analysis. (g) and (h) Using the invasion assay, the effects of OS cell lines MG63 and Saos-2 were measured after transfection with vectors containing constitutively OE-NPR3 or OE-NC *p < 0.05, **p < 0.01, ***p < 0.001 compared with the Ctrl. Data are presented as the mean ± standard deviation from three independent experiments.

Figure 3
Page 24/28 Overexpression of NPR3 inhibits the tumour growth and metastasis of osteosarcoma in vivo. Stable OS cells overexpressing NPR3 were injected into nude mice. At the indicate times, (a) number of metastatic clusters per lung, (b) tumour volume, (c) primary bone tumour, (d) tumour weight were measured (mean ± standard deviation; n = 15). (e) In vivo imaging data are presented. (f) Representative metastatic foci of lungs, livers and kidney were subjected to anatomical and histological analyses. The data are shown as the mean ± standard deviation. *p < 0.05, **p < 0.01, ***p < 0.001 vs. OE-NC.

Figure 4
Effects of NPR3 overexpression on EMT and apoptosis, which were correlated to tumour metastasis. After transfections of plamid (a, b) EMT markers, including E-cadherin, N-cadherin, vimentin, Snail, Twist and ZEB1 and (c, d) apoptosis markers Bax, Bcl-2, and caspase3 were measured by WB in Saos-2 and MG63 cells. β actin was used as an internal control. The data are shown as the mean ± standard deviation. * p < 0.05 vs. ctrl. (e, f) Cell death of Saos-2 and MG63 cells was determined by staining with Annexin V-FITC and PI using ow cytometry.