GREM1 overexpression inhibits proliferation, migration and angiogenesis of osteosarcoma

Experimental Cell Research 384 (2019) 111619

Contents lists available at ScienceDirect

Experimental Cell Research journal homepage: www.elsevier.com/locate/yexcr

GREM1 overexpression inhibits proliferation, migration and angiogenesis of osteosarcoma

T

Qingguo Gua,1, Yibin Luob,1, Cheng Chenc,1, Dongjie Jianga,∗∗∗, Quan Huanga,∗∗, Xinwei Wangb,∗ a

Department of Orthopedic Oncology, Changzheng Hospital, Second Military Medical University, 415 Fengyang Road, Shanghai, 200003, China Department of Orthopedics, Changzheng Hospital, Second Military Medical University, 415 Fengyang Road, Shanghai, 200003, China c Department of Orthopedics, Shanghai University of Medicine &health Sciences Affiliated Zhoupu Hospital, China b

A R T I C LE I N FO

A B S T R A C T

Keywords: GREM1 Migration Invasion Proliferation Osteosarcoma

Osteosarcoma is the most common malignancy of bone that occurs in young adults and children, with a five-year survival rate of 60–70%. Metastasis of osteosarcoma maintains an even poorer prognosis. GREM1 plays an important role in regulating organogenesis, body patterning, and tissue differentiation. However, there are limited studies on GREM1 in osteosarcomas. This study was carried out to characterize the expression and function of GREM1 in osteosarcoma cells, thus extending our understanding of osteosarcoma metastasis. GREM1 expression was detected in hBMSC, hFOB1.19, Saos-2, MG63 and U2OS cell lines using quantitative real-time polymerase chain reaction (qRT-PCR) and Western blot analysis. Gain- and loss-of-function approaches were used to assess the biological function of GREM1 in U2OS cells. The effects of GREM1 on U2OS cell proliferation were examined using the CCK-8 and colony formation assay. Migration and invasion ability were confirmed by the wound healing and Transwell assay, respectively. Flow cytometry was used to analyse the effect of GREM1 on the cell cycle and apoptosis. The expression of GREM1 targets was evaluated by qRT-PCR and western blotting. The expression of GREM1 was significantly downregulated in osteosarcoma. GREM1 overexpression inhibited the proliferation, migration and invasion of U2OS cells. GREM1 overexpression suppressed tumour cell-induced endothelial cell migration and invasion ability. The effect of GREM1 may be transduced through regulation of the BMP target transcription factor inhibitor of MMP-2 and -9 as well as Id1. GREM1 overexpression and knockdown regulates the tumorigenesis of osteosarcoma in vivo. In conclusion, GREM1 is downregulated in osteosarcoma cells, and overexpression of GREM1 inhibits the proliferation, migration, invasion and angiogenesis abilities of osteosarcoma cells in vitro and in vivo.

1. Introduction Osteosarcoma is the most common malignancy of bone that occurs in young adults and children, originating from bone marrow mesenchymal stem cells. The five-year survival rate of osteosarcoma continues to ranges from 60 to 70% [1]. Metastasis of osteosarcoma occurs in a small portion of patients, leading to poor prognosis with a five-year survival rate of 11–13% [2]. Surgery is the main treatment method for osteosarcoma, assisted by radiotherapy and chemotherapy. In recent years, despite improvements in surgical methods and development of new drugs, osteosarcoma metastasis occurs in approximately 40% of cases, and the five-year survival rate has not significantly improved. The investigation of new genes that play roles in osteosarcoma

would be helpful in determining the pathogenesis of metastasis and developing a targeted therapy for this disease. Bone morphogenetic protein (BMP) is named for its ability to induce the formation of bone and cartilage, it is not only involved in the formation and development of embryos, differentiation and proliferation of tissue cells, but also closely related to the occurrence, development and metastasis of some tumors [3]. It has been found that BMP4 can activate BMP signaling pathway to induce epithelial-mesenchymal transition (EMT) in stem cells and cancer cells, which results in enhanced invasion and migration of cancer cells [4,5]. High expression of BMP4 was detected in osteosarcoma, chondrosarcoma and other bone tumors [6]. Overexpression of BMP9 can inhibit the growth, migration, invasion of osteosarcoma cells and activate BMP/Smads signaling

∗

Corresponding author. Corresponding author. ∗∗∗ Corresponding author. E-mail addresses: [email protected] (D. Jiang), [email protected] (Q. Huang), [email protected] (X. Wang). 1 Qingguo Gu, Yibin Luo and Cheng Chen contributed equally to this work. ∗∗

https://doi.org/10.1016/j.yexcr.2019.111619 Received 14 April 2019; Received in revised form 31 August 2019; Accepted 6 September 2019 Available online 13 September 2019 0014-4827/ © 2019 Elsevier Inc. All rights reserved.

Experimental Cell Research 384 (2019) 111619

Q. Gu, et al.

pathway, inhibit Wnt/β-catenin signaling pathway and reduce the expression and activity of MMP [7]. BMP-2 regulates fibronectin-integrin beta-1 to make osteosarcoma cells metastasize and invade outward [8]. GREM1, as a member of the BMP (bone morphogenic protein) antagonist family, plays an important role in regulating organogenesis, body patterning, and tissue differentiation [9]. Like other extracellular BMP antagonists, GREM1 contains a cysteine knot. The role of GREM1 as a BMP antagonist has been identified in the kidney, limb development and lung branching morphogenesis [10]. In recent years, additional studies have reported the clinical significance of GREM1 in human cancers. Guan et al. first examined the expression level and function of GREM1 in glioma, demonstrating that knockdown of GREM1 inhibited cell viability, migration and invasion as well as the EMT process in glioma cells [11]. In basal cell carcinomas, GREM1 has been reported to be a marker for activated myofibroblasts in the cancer stroma or in scar tissue [12]. GREM1 has also been investigated in human cancers, including carcinomas of the stomach [13], colon [14], oesophagus [15], kidney [16], and pancreas [17]. However, few studies have examined GREM1 in osteosarcoma. This study was carried out to characterize the expression and function of GREM1 in osteosarcoma cells, extending our understanding of osteosarcoma metastasis.

Table 1 Sequence of primers used for PCR. Gene GREM1 MMP1 MMP2 MMP9 MMP13 ID1 GAPDH

Forward Primer Reverse Primer Forward Primer Reverse Primer Forward Primer Reverse Primer Forward Primer Reverse Primer Forward Primer Reverse Primer Forward Primer Reverse Primer Forward Primer Reverse Primer

Sequence(5’-3’) GGAGCCCTGCATGTGACG GAAGCGGTTGATGATGGTGC AAAATTACACGCCAGATTTGCC GGTGTGACATTACTCCAGAGTTG CCCACTGCGGTTTTCTCGAAT CAAAGGGGTATCCATCGCCAT AGACCTGGGCAGATTCCAAAC CGGCAAGTCTTCCGAGTAGT ACTGAGAGGCTCCGAGAAATG GAACCCCGCATCTTGGCTT CTGCTCTACGACATGAACGG GAAGGTCCCTGATGTAGTCGAT AAGGTGAAGGTCGGAGTCAAC GGGGTCATTGATGGCAACAATA

concentration. Ten micrograms of protein was subjected to electrophoresis in each lane on an 8% SDS-PAGE and then transferred to a polyvinylidene difluoride membrane. Blocking was performed with 5% BSA in TBST (TBS containing 0.1% Tween-20) at room temperature for 1 h. The membrane was incubated with primary antibodies at 4 °C overnight and secondary antibodies at room temperature for 1 h. The membranes were washed 3 times with washing buffer (PBS containing 0.1% Tween) for 10 min after each incubation. The LI-COR Odyssey Scanner was used to capture the images. Primary antibodies were used as follows: GREM1, MMP2, MMP9 mouse mAb (1:1000, Abcam, Cambridge, MA, USA); pSmad1, Smad1 Goat pAb (1:500, Santa Cruz, Dallas, Texas, USA) and GAPDH Rabbit mAb (1:2000, Cell Signaling Technology, Danvers, MA, USA).

2. Material and methods 2.1. Cell culture The hBMSC, hFOB1.19, Saos-2, MG63 and U2OS cell lines were selected to detect the expression level of GREM1, hBMSC, hFOB1.19 were used as the normal control cells and Saos-2, MG63 and U2OS were used as osteosarcoma cells; U2OS and Saos-2 was used for the in vitro loss/gain-of-function study. These cell lines were acquired from Key Laboratory for Endocrine and Metabolic Diseases of Chinese Health Ministry (Shanghai, China). hBMSCs were cultured in DMEM (Gibco, Carlsbad, CA, USA), MCDB (Sigma, Saint Louis, Missouri, USA), and F12 (Gibco) (2:1:1) supplemented with 10% of FBS (Gibco), 100 U/ml penicillin, and 100 μg/ml streptomycin. TPC1 were cultured in RPMI 1640 Medium HEPES (Gibco) supplemented with 10% of FBS (Gibco, Carlsbad, CA, USA), 100 U/ml penicillin, and 100 μg/ml streptomycin. Human umbilical vein endothelial cells (HUVECs) were cultured in DMEM (Gibco, Carlsbad, CA, USA) supplemented with 10% of FBS (Gibco), 100 U/ml penicillin, and 100 μg/ml streptomycin. The cell lines were incubated at 37 °C in a 5% CO2 humidified atmosphere.

2.4. Lentiviral transduction The GREM1 gene was ligated into pLVX-IRES-puro to construct the GREM1 overexpression plasmid. HEK293T viral packaging cells were infected with the pLVX-IRES-puro (negative control) and pLVX-IRESpuro-GREM1 constructs together with the psPAX2 and pMD2. G plasmids. For the overexpression experiments, the viral supernatant was collected for infection of U2OS cells 48 h after transfection. The mRNA and protein expression level of GREM1 were then analysed by qRT-PCR and Western blot analysis, respectively. 2.5. RNA interference

2.2. RNA extraction and qRT-PCR (quantitative reverse transcriptasepolymerase chain reaction)

GREM1-specific shRNA and scramble negative control shRNA were purchased from GenePharma Co., Ltd. (Shanghai, China). The sequences were as follows: GREM1-siRNA, sense, 5′-AAAUCGAUGGAUA UGCAACGA-3′, antisense, 5′-GUUGCAUAUCCAUCGAUUUGG-3’; scramble shRNA, sense, 5′-UUCUCCGAACGUGUCACGUTT-3′, antisense, 5′-ACGUGACACGUUCGGAGAATT-3’. For the knockdown experiments, U2OS cells were transfected with 100 nM GREM1-specific siRNA or scramble shRNA constructed in pLKO.1 using Lipofectamine 2000 (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer's instructions. The mRNA and protein expression level of GREM1 were then analysed by qRT-PCR and Western blot, respectively.

Total RNA was extracted from the tissues of primary osteosarcoma and its metastatic lesions in the lung, as well as cell lines (hBMSCs, hFOB1.19, Saos-2, MG63 and U2OS) using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. A Reverse Transcription Kit (Takara, Dalian, China) was used for reverse transcription of total RNA (1 μg). SYBR® Premix Ex TaqTM II Kit (Takara, Dalian, China) PCR was used for PCR according to the instructions supplied with the VIIA7 system (Applied Biosystems, California, USA). The gene primers used for PCR are listed in Table 1. GAPDH was used as the internal control for normalization of all the amplifications. The data were analysed using 2-△△Ct and expressed as fold change compared with the negative control. Three independent experiments were analysed in each test.

2.6. Colony formation assay Forty-eight hours after transfection, U2OS cells were seeded in 6well plates at a density of 1000 cells/well and then cultured at 37 °C in a 5% CO2 humidified atmosphere. The medium was changed every other day. After 7 days of culture, the medium was removed, and the cells were washed twice with PBS. Then, cells were fixed in methanol for 20 min, stained with 1% crystal violet for 30 min at room temperature, and washed again. The images were then captured.

2.3. Western blot analysis Cells were collected and split with RIPA. The supernatant containing the protein was collected after a 30-min reaction on ice and 15min centrifugation at 13000 rpm 48 h. The BCA Protein Assay Kit (Pierce, Illinois, USA) was applied for determination of the protein 2

Experimental Cell Research 384 (2019) 111619

Q. Gu, et al.

2.7. CCK-8 assay Forty-eight hours after transfection, U2OS cells were seeded into 96well plates at a density of 1 × 104 cells/well. The Cell Counting Kit-8 (CCK-8; Dojindo Molecular Technologies, Japan) was used to evaluate the cell viability at 12 h, 24 h, 36 h, 48 h and 72 h. The absorbance was detected using a TECAN infinite M200 plate reader at 450 nm. 2.8. Flow cytometry assay Forty-eight hours after transfection, U2OS cells were collected for cell cycle analysis. The cells were fixed in 75% ethanol at 4 °C overnight. The Cell Cycle Analysis Kit (Beyotime biotechnology, Jiangsu, China) was applied to stain the cells according to the manufacturer's instructions. Gallios Flow Cytometry (Beckman Coulter, USA) was used to quantify the cell cycle. All experiments were performed independently in triplicate. 2.9. Wound healing assay Forty-eight hours after transfection, U2OS cells were plated in each well of a 6-well plate and incubated until they reached 100% confluency. Subsequently, a scratch was created with a pipette tip. The medium was then replaced with serum-free medium. During the subsequent 24-h incubation, the wound closing process was observed, and images were captured. All experiments were performed independently in triplicate. 2.10. Transwell invasion assay Fig. 1. GREM1 is downregulated in osteosarcoma cells. (A) qRT-PCR analysis of the expression of GREM1 in hBMSC, hFOB1.19, Saos-2, MG63 & U2OS cells. The values were normalized to GAPDH mRNA expression. Data were expressed as the mean ± SD of three independent experiments. “**” indicates P < 0.01. (B). Western blot analysis of the expression of GREM1 in hBMSC, hFOB1.19, Saos-2, MG63 & U2OS cells. hBMSC and hFOB1.19 refer to non-cancer cells and Saos-2, MG63 & U2OS cells refer to cancer cells. GAPDH was used as a loading control. Representative images of three repeated experiments are shown. AbbreviationsqRT-PCR, quantitative real-time PCR; GREM1; gremlin 1; hBMSCs: human bone marrow mesenchymal stem cells; GAPDH: glyceraldehyde-3-phosphate dehydrogenase.

The Transwell invasion assay was performed using Transwell chambers with 8 μm pores (Costar, Corning, NY, USA). Matrigel (BD Biosciences, New Jersey, USA) was used to coat the top side of the membrane insert. The upper chamber was filled with 200 μl serum-free medium, while the lower chamber was filled with 600 μl medium with 5% FBS. Forty-eight hours after transfection, 1 × 104 U2OS cells were seeded into the upper chamber and maintained at 37 °C, 5% CO2 with a humidified atmosphere. After a 24-h incubation, the noninvaded cells on the top side of the membrane insert were removed using cotton swabs. Subsequently, the membrane inserts were fixed in methanol for 20 min and stained with 1% crystal violet for 30 min. The number of invaded cells on the bottom of the membrane was calculated under a microscope, and images were captured. All experiments were performed independently in triplicate.

3. Results 3.1. GREM1 is downregulated in osteosarcoma cells The expression level of GREM1 was analysed in the hBMSC, hFOB1.19, Saos-2, MG63 and U2OS cell lines. The qRT-PCR analysis showed that GREM1 was significantly downregulated in Saos-2 (Fold change 0.351, p value 0.00895), MG63 (Fold change 0.462, p value 0.00956) and U2OS (Fold change 0.257, p value 0.00549) cells compared with hBMSC and hFOB1.19 cells (Fig. 1A). The Western blot analysis also showed that the protein level of GREM1 was lower in Saos2, MG63 and U2OS than in hBMSC and hFOB1.19 cells (Fig. 1B). Both the qRT-PCR and Western blot results showed that U2OS maintained the lowest expression of GREM1 among the assessed cell lines.

2.11. Xenograft nude mouse model To examine the effect of GREM1 expression in tumour formation, male BALB/c nude mice (6 weeks old) were used for xenograft studies. All animal experiments were approved by the Animal Care and Use Committee of Changzheng Hospital, The Second Military Medical University. Exponetially growing U2OS cells were injected subcutaneously into the flanks of nude mice (1 × 107 cells per animal). Tumors size were measured every 1 week by caliper to determine tumour volume. All mouse were killed 3 weeks after seeding of tumour cells, and the tumour weights measured.

3.2. GREM1 overexpression suppresses the proliferation, invasion and migration of osteosarcoma cells

2.12. Statistical analysis and quantification The gain-of-function study was performed in U2OS and Saos-2 cell transfected with Lenti-GREM1 and Lenti-vector. Efficient overexpression of GREM1 in U2OS and Saos-2 cells was confirmed by qRTPCR and Western blot analysis (Fig. 2A and B and Figs. S1A and B). In the CCK8 assay, the proliferative ability of U2OS and Saos-2 cells was examined at six time points (0 h, 12 h, 24 h, 48 h, 72 h, 96 h) after transfection with Lenti-GREM1 and Lenti-vector. The cell viability was

The statistical analysis was performed using GraphPad Prism 5.0 (GraphPad Software, La Jolla, CA, USA) and Statistical Program for Social Sciences 19.0 software (SPSS, Chicago, IL, USA). Data were presented as the mean ± SD, and comparisons were calculated using the Student's t-test. Experiments were repeated in triplicate. P < 0.05 was considered to indicate a significant difference. 3

Experimental Cell Research 384 (2019) 111619

Q. Gu, et al.

Fig. 2. GREM1 overexpression suppresses the proliferation, invasion and migration of osteosarcoma cells. (A) qRT-PCR analysis of the mRNA expression levels of GREM1 in U2OS cells infected with Lenti-Vector and Lenti-GREM1. (B) Western blot analysis and quantitative analysis of GREM1 in U2OS cells infected with LentiVector and Lenti-GREM1. GAPDH was used as a loading control. Representative images of three repeated experiments are shown.(C) The CCK-8 assay was performed to evaluate the proliferation of U2OS cells with GREM1 overexpression after 12 h, 24 h, 48 h, 72 h and 96 h.(D) Flow cytometry images of cell apoptosis in U2OS cells infected with Lenti-Vector and Lenti-GREM1. (E) The colony formation assay was performed to evaluate the proliferation of U2OS cells infected with Lenti-Vector and Lenti-GREM1. The colonies were captured and counted. The colony formation assay results are presented as a histogram. (F) Flow cytometry images of the cell cycle in U2OS cells. The quantified cell cycle results are shown as a percentage of the total cells. (G) The wound healing assay was performed to determine the migration ability of U2OS cells infected with Lenti-Vector and Lenti-GREM1. Representative images obtained at 0 and 24 h from three repeated experiments are shown. (H) The Transwell assay was performed to determine the invasion ability of U2OS cells infected with Lenti-Vector and Lenti-GREM1. Representative images of invasive cells in the lower chamber stained with crystal violet are shown. Quantifications of cell invasion was performed by determining the numbers of invasive cells. All data are expressed as the mean ± SD of three independent experiments. “*” indicates P < 0.05, “**” indicates P < 0.01, “***” indicates P < 0.001.

2 cells transfected with Lenti-GREM1 and Lenti-vector. The distance of the scratch wound in the Lenti-GREM1 group was significantly larger compared with the Lenti-vector group (Fig. 2G and Fig. S1G). In the Transwell invasion assay, the number of cells that had invaded through the chamber was significantly reduced in the Lenti-GREM1 group compared with the Lenti-vector group (Fig. 2H and Fig. S1H). The above data indicate that overexpression of GREM1 suppresses the proliferation, migration and invasion of U2OS and Saos-2 cells.

determined based on a proliferation curve determined by the absorbance at 450 nm. The cell viability in the Lenti-GREM1 group was suppressed compared with the Lenti-vector group (Fig. 2C and Fig. S1C). In addition, the apoptotic ability of the U2OS and Saos-2 cells was analysed by flow cytometry. The results revealed no differences regarding the proportions of apoptotic cells between the Lenti-GREM1 and Lenti-vector groups (Fig. 2D and Fig. S1D), indicating that overexpression of GREM1 did not impact the apoptotic ability of U2OS and Saos-2 cells. A colony formation assay was also performed to evaluate the impact of GREM1 on the cell proliferation ability. The results showed fewer clones in the Lenti-GREM1 group than the Lenti-vector group (Fig. 2E and Fig. S1E). In the cell cycle analysis, the proportions of cells in G1 phase was higher in the Lenti-GREM1 group compared with the Lenti-vector group, while it was lower in S, G2/M phase in the Lenti-GREM1 group (Fig. 2F and Fig. S1F). The wound healing assay was performed to evaluate the migration ability of U2OS and Saos-

3.3. GREM1 knockdown promotes the proliferation, invasion and migration of osteosarcoma cells Additionally, the loss-of-function study was performed in U2OS and Saos-2 cells transfected with pLKO.1-GREM1 and pLKO.1-NC, which were constructed with GREM1-specific shRNA and scramble shRNA. A significant knockdown efficiency was observed in U2OS and Saos4

Experimental Cell Research 384 (2019) 111619

Q. Gu, et al.

Fig. 3. GREM1 knockdown promotes the proliferation, invasion and migration of osteosarcoma cells. (A) qRT-PCR analysis of the mRNA expression levels of GREM1 in U2OS cells transfected with pLKO.1-NC and pLKO.1-GREM1. (B) Western blot analysis of GREM1 in U2OS cells transfected with pLKO.1-NC and pLKO.1-GREM1. GAPDH was used as a loading control. Representative images of three repeated experiments are shown. (C) The CCK-8 assay was performed to evaluate the proliferation of U2OS cells with GREM1 knocked down after 12 h, 24 h, 48 h, 72 h and 96 h. (D) Flow cytometry images of cell apoptosis in U2OS cells transfected with pLKO.1-NC and pLKO.1-GREM1. (E) The colony formation assay was performed to assess the proliferation of U2OS cells transfected with pLKO.1-NC and pLKO.1-GREM1. The colonies were captured and counted. The colony formation assay results are presented as a histogram. (F) Flow cytometry images of the cell cycle in U2OS cells. The quantified cell cycle results are shown as a percentage of the total cells. (G) The wound healing assay was performed to determine the migration ability of U2OS cells transfected with pLKO.1-NC and pLKO.1-GREM1. Representative images at 0 and 24 h from three repeated experiments are shown. (H) The Transwell assay was performed to determine the invasion ability of U2OS cells transfected with pLKO.1-NC and pLKO.1-GREM1. Representative images of invasive cells in the lower chamber stained with crystal violet are shown. Quantification of cell invasion was determined based on the numbers of invasive cells. All data are expressed as the mean ± SD of three independent experiments. “*” indicates P < 0.05, “**” indicates P < 0.01.

number of clones in the pLKO.1-GREM1 than the pLKO.1-NC group (Fig. 3E and Fig. S2E). Cell cycle analysis showed that the proportion of cells in G1 phase was significantly lower in the pLKO.1-GREM1 compared with the pLKO.1-NC group, and the cell proportions were larger in S, G2/M phase in the pLKO.1-GREM1 group (Fig. 3F and Fig. S2F). In the wound healing assay, the distance spanning the scratch wound was significantly smaller in the pLKO.1-GREM1 group compared with the pLKO.1-NC group (Fig. 3G and Fig. S2G). In the Transwell invasion assay, a significantly greater number of cells had invaded through the chamber in the pLKO.1-GREM1 compared with the pLKO.1-NC group (Fig. 3H and Fig. S2H). These results indicate that GREM1 knockdown promotes the proliferation, invasion and migration of osteosarcoma cells.

2 cells transfected with pLKO.1-GREM1, as confirmed by qRT-PCR and Western blot analysis (Fig. 3A and B F and Fig. 2A and B). In the CCK8 assay, the proliferation ability of U2OS cells was examined at six time points (0 h, 12 h, 24 h, 48 h, 72 h, 96 h) after transfection with pLKO.1GREM1 and pLKO.1-NC. The cell viability was evaluated based on a proliferation curve generated by the absorbance at 450 nm. The proliferative ability of the cells in the pLKO.1-GREM1 group was promoted compared with the pLKO.1-NC group (Fig. 3C and Fig. S2C). In addition, the apoptotic ability of the U2OS and Saos-2 cells was analysed by flow cytometry. The results revealed no differences regarding the proportions of apoptotic cells between the pLKO.1-GREM1 and pLKO.1-NC groups (Fig. 3D and Fig. S2D), indicating that knockdown of GREM1 did not impact the apoptotic ability of U2OS and Saos-2 cells. A colony formation assay was also conducted to evaluate the impact of pLKO.1GREM1 on cell proliferation, and the results revealed a much greater 5

Experimental Cell Research 384 (2019) 111619

Q. Gu, et al.

Fig. 4. GREM1 overexpression and knockdown impacts the expression of MMPs and the BMP target gene Id1. qRT-PCR analysis of the mRNA expression levels of MMP1 (A), MMP2 (B), MMP9 (C), MMP13 (D) and Id1 (E) in U2OS cells with GREM1 overexpression and knockdown. The values were normalized to GAPDH mRNA expression. Data are expressed as the mean ± SD of three independent experiments. “**” indicates P < 0.01, “***” indicates P < 0.001. (F) Western blot analysis of the expression of MMP9, MMP2, p-Smad1 and Smad1 in U2OS cells with GREM1 overexpression and knockdown. Representative images of three repeated experiments are shown.

3.5. Tumour cell-derived GREM1 regulates endothelial cell function

3.4. GREM1 regulates the expression of matrix-degrading enzymes and BMP pathway activity

Subsequently, this study verified the function of tumour cell-derived GREM1 on the surrounding endothelial cells. HUVECs were incubated with conditioned medium from GREM1 overexpression or knockdown U2OS cells compared with the control. In the wound healing assay, the distance spanning the scratch wound was significantly larger in the Lenti-GREM1 group than the Lenti-vector group, while it was significantly smaller in the pLKO.1-GREM1 group compared with the pLKO.1-NC group (Fig. 5A). In the transwell invasion assay, significantly fewer cells had invaded through the chamber in the LentiGREM1 compared with the Lenti-vector group, while significantly more cells had invaded in the pLKO.1-GREM1 compared with the pLKO.1-NC group (Fig. 5B). These results indirectly indicate that GREM1 overexpression suppresses the migration and invasion ability of HUVECs. Furthermore, the expression of angiogenic factors was detected in HUVECs cultured with the supernatant from GREM1 overexpression and knockdown U2OS cells. The results showed that vascular endothelial growth factor (VEGF) was significantly decreased while endostatin, PAI-1 and thrombospondin-1 were increased in HUVECs treated with GREM1 overexpression conditioned medium. Conversely, VEGF was significantly increased while endostatin, PAI-1 and thrombospondin-1 were decreased in HUVECs treated with GREM1 knockdown conditioned medium (Fig. 5C). In the tube-formation assay,

The potential mechanism by which GREM1 decreases U2OS cell proliferation and invasion was further explored. The expression of matrix metalloproteinases (MMP1, MMP2, MMP9, MMP13) and Id1 were detected in GREM1 overexpression and knockdown cells. The mRNA expression levels of MMP2, MMP9, and Id1 were significantly downregulated in GREM1 overexpression U2OS cells (Fig. 4B,C,E), whereas that of MMP1 and MMP13 remained unaffected (Fig. 4A,D). The protein expression level of MMP2 and MMP9 were also decreased in GREM1 overexpression U2OS cells compared with the control (Fig. 4F). In contrast, when GREM1 was knocked down in U2OS cells, MMP2, MMP9 and Id1 were significantly overexpressed at the mRNA level, while MMP1 and MMP13 remained unaffected (Fig. 4A,B,C,D,E); MMP2 and MMP9 were also decreased at the protein level (Fig. 4F). Furthermore, the protein expression level of p-Smad1 was suppressed when GREM1 was overexpressed, while it was enhanced when GREM1 was knocked down (Fig. 4F). However, the protein expression level of Smad1 was not affected by GREM1 overexpression or knocked down (Fig. 4F). These data indicate that GREM1 inhibits the expression of MMP2 and MMP9 and the activity of the BMP pathway.

6

Experimental Cell Research 384 (2019) 111619

Q. Gu, et al.

Fig. 5. GREM1 overexpression and knockdown regulates the impact of U2OS cells on endothelial cell function.

3.6. GREM1 overexpression and knockdown regulates the tumorigenesis of osteosarcoma in vivo

significantly lower tube formation was in the Lenti-GREM1 group compared with the Lenti-vector group, while significantly more tube formation was in the pLKO.1-GREM1 than that in pLKO.1-NC group (Fig. 5D). These results indicate that overexpression of GREM1 inhibits endothelial cells to acquire angiogenic ability.

To further confirm the effect of GREM1 on osteosarcoma progression, xenograft experiments were performed. We seeded the U2OS cells 7

Experimental Cell Research 384 (2019) 111619

Q. Gu, et al.

Fig. 6. GREM1 overexpression and knockdown regulates the tumorigenesis of osteosarcoma in vivo. GREM1 overexpression and knockdown in U2OS cells affects the tumorigenesis of osteosarcoma in vivo (A) and the tumour volume (B) and tumour weight. All data are expressed as the mean ± SD of five samples form each group. “*” indicates P < 0.05, “**” indicates P < 0.01.

osteosarcoma as comparison to the control. Moreover, when the expression of GREM1 was artificially upregulated, the invasive and proliferative abilities of osteosarcoma cells were suppressed, indicating that the presence of GREM1 inhibited the metastatic potential in osteosarcoma. Thus, GREM1 was found to be associated with less malignant behaviour in osteosarcoma. Studies have shown that GREM1 can act as a ligand of VEGFR2, activate the signal pathway of VEGFR2 and promote angiogenesis [21,22]. However, in our study, we found that the effect of culture supernatant of osteosarcoma cells on migration after overexpression of GREM1 on the invasion and angiogenesis of HUVEC cells decreased. Although our results are in contrast with previous reports, it should be pointed out that this discrepancy may be linked to the use of different cellular models as well as to different experimental approaches. First of all we did not study the direct effect of GREM1 on HUVEC in our study, but the effect of GREM1 on HUVEC cells by affecting osteosarcoma cells. We also detected that GREM1 could affect the release of vascular endothelial growth factor in osteosarcoma cells. Secondly, the effect of GREM1 on the expression of VEGFR2 is bidirectional, that is, GREM1 can be used as an inhibitor or an agonist [23]. The mechanism by which GREM1 regulates proliferation and invasion was explored by investigating how MMPs were modulated by GREM1. MMPs are in the zinc-dependent endoproteinase family, with enzymatic activity that results in the degradation of extracellular matrix during invasion and migration. Until now, MMPs have been confirmed to modulate tumour progression by enabling tumour cells detach from each other for migration and invasion [24,25]. Kalhori et al. examined the role of MMP2 and MMP9 in S1P-induced invasion of ML-1 cells, showing that the knockdown of MMP2 and MMP9 could attenuate S1Pinduced invasion in thyroid cancer cells [26]. Li S examined the role of BMP9 in breast cancer and found that it inhibited the growth of breast cancer cells by inhibiting the PI3K/Akt signaling pathway both in vivo and in vitro [27]. In the current study, MMP2 and MMP9 were inhibited by overexpression of GREM1, even though MMP1 and MMP13 were not affected. A negative regulatory relationship was found between GREM1 and MMP2, MMP9, indicating that GREM1 inhibits cancer metastasis. Another mechanism by which GREM1 modulates metastasis has been verified by the angiogenic ability of GREM1. Angiogenesis is an important step in tumour progression, involving a vascular network for tumour cell growth as well as metastasis. In the current study,

into nude mice. At indicative time points, we measured the osteosarcoma volumes and found that GREM1 knockdown promoted the tumorigenesis of osteosarcoma and GREM1 overexpression decreased the proliferation of osteosarcoma in vivo (Fig. 6). Consistently, GREM1 knockdown led to reduced tumour weights, however, GREM1 overexpression increased the tumour weights (Fig. 6). 4. Discussion Osteosarcoma is the most common malignancy in bone, with poor prognosis when metastasis is present. Investigations of gene alterations in osteosarcoma would be helpful for understanding the mechanism of metastasis and identifying potential targets for therapy. GREM1 has been studied in some other types of cancers, while it was reported for the first time in osteosarcoma in the current study. According to our study, the expression of GREM1 was significantly downregulated in osteosarcoma cells compared with non-neoplastic cells. The in vitro experiment indicated that overexpression of GREM1 inhibited the proliferation, cell cycle, migration and invasion ability in osteosarcoma cell. Additionally, matrix-degrading enzymes and BMP pathway activity were influenced by up- or downregulation of GREM1 in osteosarcoma. The expression levels of Id1 and pSmad1 were also modulated by GREM1. Moreover, the angiogenic compacity was suppressed when GREM1 was overexpressed in osteosarcoma with changes in the expression of a series of downstream molecules (VEGF, endostatin, PAI-1 & thrombospondin). As mentioned above, GREM1 is an extracellular BMP-specific antagonist [], which plays a critical role in several biological processes including cancer biology. The roles of BMPs have been extensively studied in various cancerous diseases, whereas reports of BMPs in osteosarcoma are limited similarly to GREM1. Honma et al., [29491067] reported that GREM1-negative gastric cancer showed more advanced clinicopathological factors and had a poorer survival rate than GREM1positive gastric cancer. However, other public data indicate that GREM1 is highly overexpressed in cancerous tissues. Sato et al. reported that the high expression of GREM1 mRNA was significantly correlated with a bulky tumour, one of the most malignant features of cancer [18]. Others have also reported that GREM1 is secreted from stromal cancerassociated fibroblasts, and its expression is localized to the site of invasion [19,20]. In our study, lower GREM1 expression was detected in 8

Experimental Cell Research 384 (2019) 111619

Q. Gu, et al.

angiogenic factors, including VEGF, endostatin, PAI-1 and thrombospondin-1, were detected after GREM1 was overexpressed or knocked down. We confirmed the regulatory role of GREM1 in tumour cell angiogenesis. VEGF, as an angiogenic factor, induces the proliferation and migration of vascular endothelial cells, and it is essential for both physiological and pathological angiogenesis [28]. The relationship of BMP activity and GREM1 was also discussed in our study. BMPs have been detected in the cellular components of a series of biological processes, including mammalian reproduction [29,30], liver fibrosis [31], and cancer progression [32], among others. In the current study, GREM1 negatively regulated BMP activity by modulating the expression of Id1 and Smad1 phosphorylation.

213 (2007) 768–774. [6] H. Yoshikawa, T. Nakase, A. Myoui, T. Ueda, Bone morphogenetic proteins in bone tumors, J. Orthop. Sci. : Off. J. Jpn. Orthop. Assoc. 9 (2004) 334–340. [7] Z. Lv, D. Yang, J. Li, M. Hu, M. Luo, X. Zhan, P. Song, C. Liu, H. Bai, B. Li, Y. Yang, Y. Chen, Q. Shi, Y. Weng, Bone morphogenetic protein 9 overexpression reduces osteosarcoma cell migration and invasion, Mol. Cells 36 (2013) 119–126. [8] H. Hanamura, Y. Higuchi, M. Nakagawa, H. Iwata, H. Nogami, M.R. Urist, Solubilized Bone Morphogenetic Protein (BMP) from Mouse Osteosarcoma and Rat Demineralized Bone Matrix, Clinical orthopaedics and related research, 1980, pp. 281–290. [9] D.R. Hsu, A.N. Economides, X. Wang, P.M. Eimon, R.M. Harland, The Xenopus dorsalizing factor Gremlin identifies a novel family of secreted proteins that antagonize BMP activities, Mol. Cell 1 (1998) 673–683. [10] G.G. Wulf, B. Chapuy, L. Trumper, [Mesenchymal stem cells from bone marrow. Phenotype, aspects of biology, and clinical perspectives], Med. Klin. 101 (2006) 408–413. [11] Y. Guan, W. Cheng, C. Zou, T. Wang, Z. Cao, Gremlin1 promotes carcinogenesis of glioma in vitro, Clin. Exp. Pharmacol. Physiol. 44 (2017) 244–256. [12] H.S. Kim, M.S. Shin, M.S. Cheon, J.W. Kim, C. Lee, W.H. Kim, Y.S. Kim, B.G. Jang, GREM1 is expressed in the cancer-associated myofibroblasts of basal cell carcinomas, PLoS One 12 (2017) e0174565. [13] R. Honma, N. Sakamoto, A. Ishikawa, D. Taniyama, K. Fukada, T. Hattori, K. Sentani, N. Oue, K. Tanabe, H. Ohdan, W. Yasui, Clinicopathological and prognostic significance of epithelial Gremlin1 expression in gastric cancer, Anticancer Res. 38 (2018) 1419–1425. [14] Y. Laitman, E. Jaeger, L. Katz, I. Tomlinson, E. Friedman, GREM1 germline mutation screening in Ashkenazi Jewish patients with familial colorectal cancer, Gene. Res. 97 (2015) e11. [15] D. Hong, T. Liu, W. Huang, Y. Liao, L. Wang, Z. Zhang, H. Chen, X. Zhang, Q. Xiang, Gremlin1 delivered by mesenchymal stromal cells promoted epithelial-mesenchymal transition in human esophageal squamous cell carcinoma, cellular physiology and biochemistry, Int. J. Exp. Cell. Physiol., Biochem., and Pharmacol. 47 (2018) 1785–1799. [16] R.H. Church, I. Ali, M. Tate, D. Lavin, A. Krishnakumar, H.M. Kok, J.R. Hombrebueno, P.D. Dunne, V. Bingham, R. Goldschmeding, F. Martin, D.P. Brazil, Gremlin1 plays a key role in kidney development and renal fibrosis, Am. J. Physiol. Renal. Physiol. 312 (2017) F1141–F1157. [17] M.H. Chen, Y.C. Yeh, Y.M. Shyr, Y.H. Jan, Y. Chao, C.P. Li, S.E. Wang, C.H. Tzeng, P.M. Chang, C.Y. Liu, M.H. Chen, M. Hsiao, C.Y. Huang, Expression of gremlin 1 correlates with increased angiogenesis and progression-free survival in patients with pancreatic neuroendocrine tumors, J. Gastroenterol. 48 (2013) 101–108. [18] M. Sato, K. Kawana, A. Fujimoto, M. Yoshida, H. Nakamura, H. Nishida, T. Inoue, A. Taguchi, J. Takahashi, K. Adachi, K. Nagasaka, Y. Matsumoto, O. Wada-Hiraike, K. Oda, Y. Osuga, T. Fujii, Clinical significance of Gremlin 1 in cervical cancer and its effects on cancer stem cell maintenance, Oncol. Rep. 35 (2016) 391–397. [19] G.S. Karagiannis, N. Musrap, P. Saraon, A. Treacy, D.F. Schaeffer, R. Kirsch, R.H. Riddell, E.P. Diamandis, Bone morphogenetic protein antagonist gremlin-1 regulates colon cancer progression, Biol. Chem. 396 (2015) 163–183. [20] G.S. Karagiannis, A. Treacy, D. Messenger, A. Grin, R. Kirsch, R.H. Riddell, E.P. Diamandis, Expression patterns of bone morphogenetic protein antagonists in colorectal cancer desmoplastic invasion fronts, Mol. Oncol. 8 (2014) 1240–1252. [21] S. Mitola, C. Ravelli, E. Moroni, V. Salvi, D. Leali, K. Ballmer-Hofer, L. Zammataro, M. Presta, Gremlin is a novel agonist of the major proangiogenic receptor VEGFR2, Blood 116 (2010) 3677–3680. [22] R. Erdmann, C. Ozden, J. Weidmann, A. Schultze, Targeting the Gremlin-VEGFR2 axis - a promising strategy for multiple diseases? J. Pathol. 236 (2015) 403–406. [23] E. Grillo, C. Ravelli, M. Corsini, K. Ballmer-Hofer, L. Zammataro, P. Oreste, G. Zoppetti, C. Tobia, R. Ronca, M. Presta, S. Mitola, Monomeric gremlin is a novel vascular endothelial growth factor receptor-2 antagonist, Oncotarget 7 (2016) 35353–35368. [24] Q. Cui, B. Wang, K. Li, H. Sun, T. Hai, Y. Zhang, H. Kang, Upregulating MMP-1 in carcinoma-associated fibroblasts reduces the efficacy of Taxotere on breast cancer synergized by Collagen IV, Oncology letters 16 (2018) 3537–3544. [25] D. Zhu, M. Ye, W. Zhang, E6/E7 oncoproteins of high risk HPV-16 upregulate MT1MMP, MMP-2 and MMP-9 and promote the migration of cervical cancer cells, Int. J. Clin. Exp. Pathol. 8 (2015) 4981–4989. [26] V. Kalhori, K. Tornquist, MMP2 and MMP9 participate in S1P-induced invasion of follicular ML-1 thyroid cancer cells, Mol. Cell. Endocrinol. 404 (2015) 113–122. [27] S. Li, H. Dai, Y. He, S. Peng, T. Zhu, Y. Wu, C. Li, K. Wang, BMP9 inhibits the growth of breast cancer cells by downregulation of the PI3K/Akt signaling pathway, Oncol. Rep. 40 (2018) 1743–1751. [28] B. Ren, Protein kinase D1 signaling in angiogenic gene expression and VEGFmediated angiogenesis, Front. Cell Develop. Biol. 4 (2016) 37. [29] S. Shimasaki, R.K. Moore, F. Otsuka, G.F. Erickson, The bone morphogenetic protein system in mammalian reproduction, Endocr. Rev. 25 (2004) 72–101. [30] F. Otsuka, K. Inagaki, Unique bioactivities of bone morphogenetic proteins in regulation of reproductive endocrine functions, Reprod. Med. Biol. 10 (2011) 131–142. [31] J. Bi, S. Ge, Potential roles of BMP9 in liver fibrosis, Int. J. Mol. Sci. 15 (2014) 20656–20667. [32] P.K. Panda, P.P. Naik, P.P. Praharaj, B.R. Meher, P.K. Gupta, R.S. Verma, T.K. Maiti, M.K. Shanmugam, A. Chinnathambi, S.A. Alharbi, G. Sethi, R. Agarwal, S.K. Bhutia, Abrus agglutinin stimulates BMP-2-dependent differentiation through autophagic degradation of beta-catenin in colon cancer stem cells, Mol. Carcinog. 57 (2018) 664–677.

5. Conclusion In conclusion, GREM1 is downregulated in osteosarcoma. Overexpression of GREM1 could suppress cancer metastasis and angiogenesis by inhibiting cancer cell proliferation, migration and invasion behaviours as well as tumour-associated angiogenesis. Disclosure The authors report no conflicts of interest in this work. Acknowledgements This work was supported by grants from the National Natural Science Foundation of China (No. 81572168) and the Science and Technology Commission of Shanghai Municipality (No. 15411950700). Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.yexcr.2019.111619. GREM1 overexpression and knockdown in U2OS cells affects the migration (A) and invasion and quantitatively analysed (B) the ability of endothelial cells. HUVECs were incubated with the supernatant of U2OS cells transfected with Lenti-Vector, Lenti-GREM1, pLKO.1-NC and pLKO.1-GREM1, respectively. Then, HUVECs were subjected to the wound healing and Matrigel assays. Representative micrographs of HUVECs stimulated with supernatants of GREM1 overexpression and knockdown U2OS cells are shown. (C) The expression levels of VEGF, endostatin, PAI-1 and thrombospondin-1 in U2OS cells exposed to GREM1 overexpression and knockdown supernatant, as measured by ELISA. (D) Tube-formation assay. HUVECs were incubated with the supernatant of U2OS cells transfected with Lenti-Vector, Lenti-GREM1, pLKO.1-NC and pLKO.1-GREM1, respectively. Representative micrographs of HUVECs stimulated with supernatants of GREM1 overexpression and knockdown U2OS cells are shown. All data are expressed as the mean ± SD of three independent experiments. “**” indicates P < 0.01. References [1] N. Omer, M.C. Le Deley, S. Piperno-Neumann, P. Marec-Berard, A. Italiano, N. Corradini, C. Bellera, L. Brugieres, N. Gaspar, Phase-II trials in osteosarcoma recurrences: a systematic review of past experience, Eur. J. Cancer 75 (2017) 98–108. [2] J. Gill, M.K. Ahluwalia, D. Geller, R. Gorlick, New targets and approaches in osteosarcoma, Pharmacol. Ther. 137 (2013) 89–99. [3] P.S. Leboy, Regulating bone growth and development with bone morphogenetic proteins, Ann. N. Y. Acad. Sci. 1068 (2006) 14–18. [4] A.K. Teo, Y. Ali, K.Y. Wong, H. Chipperfield, A. Sadasivam, Y. Poobalan, E.K. Tan, S.T. Wang, S. Abraham, N. Tsuneyoshi, L.W. Stanton, N.R. Dunn, Activin and BMP4 synergistically promote formation of definitive endoderm in human embryonic stem cells, Stem cells 30 (2012) 631–642. [5] S. Hamada, K. Satoh, M. Hirota, K. Kimura, A. Kanno, A. Masamune, T. Shimosegawa, Bone morphogenetic protein 4 induces epithelial-mesenchymal transition through MSX2 induction on pancreatic cancer cell line, J. Cell. Physiol.

9