Trp53 and Rb1 de ciency in chondrocytes spontaneously develop chondrosarcoma by activation of YAP signaling


 Chondrosarcoma (CHS) is a rare type of soft sarcoma with increased production of cartilage matrix arising from soft bone tissues. Currently, surgical resection is the primary clinical treatment for chondrosarcoma due to the poor response to radiotherapy and chemotherapy. However, the therapeutic effect is not satisfactory due to the higher local recurrence rate. Thus, management and elucidation of the pathological mechanism of chondrosarcoma remain an ongoing challenge, and development of effective chondrosarcoma mouse models and treatment options are urgently needed. Here, we generated a new transgenic chondrosarcoma model by double conditional deletions of Trp53 and Rb1 in chondrocyte lineage which spontaneously caused spinal chondrosarcoma and lung metastasis. Bioinformatic analysis of human soft sarcoma database showed that Trp53 and Rb1 genes had higher mutations, reaching up to approximately 33.5% and 8.7%, respectively. Additionally, Trp53 and Rb1 signatures were decreased in the human and mouse chondrosarcoma tissues. Mechanistically, we found that YAP expression and activity were significantly increased in mouse Col2-Cre;Trp53f/f/Rb1f/f chondrosarcoma tissues compared to the adjacent normal cartilage. Knockdown of YAP in primary chondrosarcoma cells significantly inhibited chondrosarcoma proliferation, invasion, and tumorsphere formation. Chondrocyte lineage ablation of YAP delayed chondrosarcoma progression and lung metastasis in Col2-Cre;Trp53f/f/Rb1f/f mice. Moreover, we found that metformin served as a YAP inhibitor, which bound to the activity area of YAP protein, and inhibited chondrosarcoma cell proliferation, migration, invasion, and progression in vitro and significantly suppressed chondrosarcoma formation in vivo. Collectively, this study identifies the inhibition of YAP may be effective therapeutic strategies for treatment of chondrosarcoma.

Chondrosarcoma is a rare type of primary bone cartilage malignancies with an 70 incident rate of about 2 new cases per a million populations per year [1,2]. It is the 71 second most common primary malignant bone tumor and has a higher local recurrence 72 rate. Although most of solid tumors have infrequent metastasis, the lung metastasis is 73 the most common in chondrosarcoma [3,4,5]. It predominantly occurs in adults aged 74 after 40 years old [6]. Currently, surgical resection is the primary clinical treatment for 75 chondrosarcoma due to the poor response to radiotherapy and chemotherapy. However, 76 the therapeutic results are not unfavorable due to the higher local recurrence and 77 mortality rates. Thus, management and elucidation of the pathological mechanism of 78 chondrosarcoma remain an ongoing challenge, and development of effective 79 chondrosarcoma mouse models and treatment options are urgently needed. 80 To discover a new drug or for drug repurposing, developing a mouse model that 81 closely mimics human chondrosarcoma initiation and progression is one of the most 82 important approaches in clinic. Currently, only few chondrosarcoma animal models 83 have been developed including allograft tumor transplanted into the hamster or rat [7, 84 8]. Despite these models are more useful for evaluating chondrosarcoma growth, there 85 are less relevant to the human disease that restricts the elucidation of the pathological 86 mechanism and development of novel therapeutic drugs. Transgenic cancer models are 87 becoming more favored, because these types of models have been characterized to be 88 accurate models in oncology, which can finely control tumor genetics and more 89 accurately study the tumor initiation and development, and delineate the potential 90 molecular drivers or inhibitors of these pathologies [9, 10]. One transgenic 91 chondrosarcoma mouse model has been developed by transgenically overexpressing c-92 Fos. However, unpredictability of tumor location, varying phenotypes (including 93 osteosarcoma), and multiple tumors formation lead to the potential problem of 94 interpreting data using this model for developing therapies [11,12]. 95 The identification of some tumor suppressors ' Fig. S1C). Moreover, Col2-Cre;Trp53 f/f /Rb1 f/f 160 mice at 5 months showed a big mass ( Fig. 2A). X-ray image showed a disruption in the 161 vertebrate bone in the thoracic spine region (Fig. 2B). The average volume of 162 chondrosarcoma was increased with age (Fig. 2C) we found the proliferation rate of primary chondrocytes in Col2-Cre;Trp53 f/f /Rb1 f/f 180 mice was significantly increased compared to that in the controls (Fig. 3A). 181 Concomitantly, the colony numbers were also remarkably increased in Trp53 and Rb1 182 deficient chondrocytes ( Cre;Trp53 f/f /Rb1 f/f mice compared to the controls (Fig. 4E (6GE3 from PDB) of YAP protein and prohibit the interaction of YAP and TEAD 220 (Fig. 5A). To further characterize the effects of metformin on chondrosarcoma 221 progression, we explored WST-1 assay to examine the effect of metformin on 222 chondrosarcoma cell proliferation. As expected, we found metformin prohibited the 223 chondrosarcoma cell growth in dose dependent manner (Fig. 5B). Moreover, the 224 activities of migration and invasion of primary chondrosarcoma cells were significantly 225 inhibited after treatment with metformin compared with the controls (Fig. 5C, D). 226 Metformin inhibited anchorage-independent chondrosarcoma cell colony formation in 227 soft agar (Fig. 5E, F) and the formation and size of tumorspheres, respectively (Fig. 5G,  228 H). Mechanistically, we found that metformin inhibited the nuclear localization and 229 transcriptional activity of YAP (Fig. 5I, J), and significantly reduced YAP target genes' 230 expression (Fig. 5K). Collectively, our findings suggested that metformin inhibits 231 chondrosarcoma progression through YAP signaling. 232

Inhibition of YAP significantly suppresses chondrosarcoma progression in Col2-233
Cre;  (Fig. 6A, B). 241 Furthermore, we also found the chondrosarcoma lung metastasis was inhibited by loss 242 of YAP (Fig. 6C). Additionally, the Kaplan-Meier survival curves plotted for the mice 243 showed a significantly longer mean survival rate in the triple cKO mice compared with 244 the double cKO mice (Fig. 6D). To test the capacity to promote tumor growth after loss 245 of YAP, we performed WST-1 and soft agar assays using chondrosarcoma cells from 246 triple cKO mice and double cKO mice. Noteworthy, the tumor growth was significantly 247 inhibited after loss of YAP (Fig. 6E, F). Meanwhile, deletion of YAP significantly 248 inhibited cell migration and invasion (Fig. 6G, H). In accordance with the reduced 249 mobility of osteosarcoma cells, loss of YAP remarkably decreased the numbers and 250 size of tumorsphere compared with that in double cKO group (Fig. 6I). To investigate 251 the role of metformin in vivo, we further used metformin to treat the Col2-252 Cre;Trp53 f/f /Rb1 f/f mice via intraperitoneal injection three times per week starting at 4 253 weeks of age, when the spine began to expand in Col2-Cre;Trp53 f/f /Rb1 f/f mice (Fig.  254  6J). The analysis of X-ray images and chondrosarcoma volume indicated that 255 metformin significantly inhibited tumor growth and improved the mobility of Col2-256 Cre;Trp53 f/f /Rb1 f/f mice at the age of 4 months (Fig. 6J, K), as evidenced by Safranin 257 O/Fast Green staining (Fig. 6L). 258

259
Chondrosarcoma is a rare type of soft sarcoma with increased production of System and the SYBR Green mixture (Bio-Rad, USA). GAPDH was served as an 363 internal control and was determined by 2 -∆∆Ct method. The primers of qRT-PCR in this 364 study were listed in Supplemental Table S1. 365

Cell functional assays 366
Cell proliferation, migration, invasion, and tumorsphere assays were carried out as 367 we performed previously [17]. 368

Luciferase reporter assay 369
For luciferase reporter assay, the chondrocytes were seeded and co-transfected with 370 luciferase reporter and the indicated plasmids in the 12-well plate. After culturing for 371 48 hr, the luciferase activities were analyzed by the Dual-Luciferase Assay Kit as we 372 performed previously [17]. 373

Radiographic procedures analysis 374
Radiographic procedures were performed in the Siemens X-ray equipment 375 (Madison, WI, USA) as we performed previously [17]. 376

Histological analysis 377
Vertebrate bones from 5-month-old Col2-Cre;Trp53 f/f /Rb1 f/f mice and controls 378 were harvested, fixed in 4% paraformaldehyde (PFA) overnight at 4°C, decalcified with 379 14% EDTA in PBS (pH 7.4) for 1 month, and then embedded in paraffin. Soft tissues 380 including the lung, kidney, spleen, brain, and liver from Col2-Cre;Trp53 f/f /Rb1 f/f mice 381 and controls were fixed and embedded in paraffin. Six-micrometer sections of the above 382 tissues were prepared, and then the stainings of Haemotoxylin and Eosin (H&E), 383 Safranin O/fast green, and Alcian blue staining were conducted as we previously 384 reported [17,18]. 385

Immunofluorescence and immunohistochemistry 386
For immunofluorescence staining, the primary chondrocytes from 5-month-old 387 Col2-Cre;Trp53 f/f /Rb1 f/f and Col2-Cre mice were seeded and cultured on coverslips. 388 After culture of 48 hr, the coverslips were fixed in 4% PFA for 5 minutes at room 389 temperature, and then permeabilized with 0.3% Triton X-100 in PBS (PBST) for 3 390 times per 5 minutes each. Next, the cells were blocked by 1% BSA for 1 hr at room 391 temperature and probed with primary antibody against rabbit anti-YAP (1:200 dilution) 392 overnight at 4°C. After washing for 3 times with PBST, the cells were incubated with 393 Alexa Fluor ® 594-conjugated second anti-rabbit antibody (1:1000 dilution) for 1 hr at 394 dark. Then, counter stain of nuclei was performed with DAPI and washed 3 times with 395 PBST, and then the cells were visualized under a fluorescence microscope as we 396 previously reported [17,18]. Immunohistochemistry was carried out as we previously 397 reported [17]. 398

Western blot 399
Briefly, the fresh chondrosarcoma tissues and adjacent normal cartilage from 400 Col2-Cre;Trp53 f/f /Rb1 f/f mice were harvested and lysed with RIPA lysis buffer and 401 protein inhibitor cocktail (Fisher Scientific™, USA), respectively. And then, equal 402 amount of the above proteins was subjected to SDS-PAGE gel (Bio-Rad, USA), 403 transferred to a PVDF membrane (GVS Life Sciences, USA), and immunoblotted with 404 primary antibodies YAP, pYAP and GAPDH (1:1000 dilution) for overnight at 4°C. 405 Following 3 times washing with 0.1% TBST (Tween-20 in TBS), the PVDF 406 membranes were incubated with HRP-conjugated anti-rabbit antibody (1:1000 dilution) 407 for 1 hr at room temperature. After washing 3 times with TBST, the membranes were 408 analyzed by ECL solution (Thermo Fisher, USA) as we previously reported [17,18]. 409

Docking and bioinformatic analysis 410
The   Cre;Trp53 f/f /Rb1 f/f mice and age-matched controls from Col2-Cre; Trp53 f/f /Rb1 f/f mice 602 and controls. After culture of 10 days, the macromass cultures as indicated were stained 603 by Alcian blue staining. Error bars were the means ± SEM from three independent 604 experiments. *P < 0.05, **P < 0.01. 605 Col2-Cre;Trp53 f/f /Rb1 f/f mice or primary chondrocytes from Col2-Cre mice were co-617 transfected with luciferase reporter and pRL-TK plasmids (internal control) as indicated, 618 respectively. After transfection of 48 hr, the luciferase activities were identified by the 619 Dual-Luciferase Assay Kit. H Soft agar analysis after silence of YAP using two 620