The NRP1 gene regulates proliferation, apoptosis, migration, and invasion in T24 and 1 5637 bladder cancer cells 2

Background Bladder urothelial carcinoma (BC) is a fatal invasive malignancy and the most 35 common malignancy of the urinary system. In the current study, we investigate the function and 36 mechanisms of Neuropilin-1 (NRP1), the co-receptor for vascular endothelial growth factor, in BC 37 pathogenesis and progression. 38 Methods The expression of NRP1 was assessed in several BC cell lines. Additionally, the 39 biological function of NRP1 in proliferation, apoptosis, angiogenesis, migration, and invasion of 40 BC were validated in vitro by silencing NRP1. Moreover, gene expression profiling chip analysis 41 was conducted, and the related signalling pathways were confirmed by Western blot to reveal the 42 potential molecular mechanisms by which NRP1 promotes the malignant progression of BC. 43 Results Overexpression of NRP1 was observed in several human BC cell lines. NRP1 knockdown 44 inhibited cell proliferation, promoted apoptosis, and decreased angiogenesis, migration, and 45 invasion in T24 and 5637 human BC cells. Microarray analysis results indicated that the 46 expression of NRP1 was correlated with the levels of cyclin dependent kinase (CDK) 4, 47 baculoviral IAP repeat containing 3, Cyclin E 2, CDK2, and AP-1 transcription factor subunit in 48 BC. We also demonstrated that the biological function of NRP1was associated with activation of 49 the mitogen-activated protein kinase (MAPK) signalling pathway. 50 Conclusions Our findings provide evidence that NRP1, as a potential tumour promoter, 51 contributes to the metastasis and invasion of BC, which is associated with the activation of the 52 MAPK pathway. Targeting NRP1 has the potential to become a new therapeutic strategy to 53 benefit more patients with BC or other cancers.

8 apoptosis determination kit (Ebioscience, USA), we added 100 μL of 1× binding buffer cautiously 166 to each tube. Then, 5 μL of propidium iodide (PI) (Sigma) and 5 μL of Annexin-V-APC were 167 added to the tubes. The tubes were then incubated at room temperature for 15 min, protected from 168 light, before placing on ice. Within 1 hour, apoptosis was assessed using the BD FACSCalibur 169 flow cytometer (BD Biosciences). 170 171

Flow cytometry cell cycle analysis 172
Cells were digested with 0.25% trypsin, washed with PBS, and centrifuged at 1000 rpm for 5 173 min. The cell pellet was washed twice with PBS, after which the cells were resuspended in 0.5 mL 174 of PBS. The tubes were oscillated on a low-speed oscillator, and 70% ice-cold ethanol was added 175 to fix the cells overnight at 4°C. The fixed cells were subsequently centrifuged at 1000 rpm for 5 176 min. The supernatant was discarded, and the pellet was washed with PBS and resuspended. 177 Bovine pancreatic RNase (Fermentas, Lithuania) was added at a final concentration of 2 mg/mL 178 and the tubes were incubated in a 37℃ water bath for 30 min. PI was added at a final 179 concentration of 65 µg/mL, followed by incubation in an ice bath for 30 min protected from light. 180 Finally, cell cycle detection and data analysis were performed using a BD FACSCalibur Flow 181 Cytometer filtration and FLOWJO Software (Tree Star, Inc, Ashland, OR, USA). 182 183

Transwell cell migration assay 184
Cells in the logarithmic growth stage were digested and centrifuged and then resuspended in 185 serum-free medium. A volume of 750 µL culture medium with serum was added to the bottom of 186 a 24-well plate, and migration chambers were put in the wells. We added 600 µL of 30% serum-187 9 free medium to each chamber and added 100 µL of cell suspension at a density of 1 × 10 5 188 cells/mL. After incubation at 37°C for 24 h, the medium was removed from the chambers, and the 189 wells were washed twice with PBS. Migrated cells were fixed by formaldehyde for 30 min before

Transwell cell invasion assay 196
Matrigel was diluted using serum-free medium and mixed well by pipet. A volume of 100 µL 197 prepared Matrigel was added to Transwell chambers in a 24-well plate and incubated at 37°C 198 overnight for gelling. Cells in the logarithmic growth stage were digested, centrifuged, and 199 resuspended in serum-free medium. A volume of 500 µL cell suspension at a density of 1 × 10 5 200 cells/mL was placed in the chamber. We subsequently added 750 µL culture medium with serum 201 in the bottom of the wells of a 24-well plate and placed the Transwell chambers into the wells. 202 After incubation at 37°C for 12 h, the medium was removed from the chambers, and the wells 203 were washed twice with PBS. The invasive cells were fixed by formaldehyde at room temperature 204 for 30 min, followed by a 15-min staining with Giemsa stain, and then washed twice with PBS. 205 The non-invasive cells on the bottom of the chamber were scraped off with cotton swabs. Invasive 206 cells were counted in three random fields of view using a light microscope (200×), and images 207 were captured. 208 209

Affymetrix gene expression profile chip detection 210
We extracted total RNA from normal control cells and NRP1-knockdown cells with TRIzol 211 reagent as described above and quantified RNA using the NanoDrop ND-2000 (Thermo 212 Scientific, USA). RNA integrity was further analysed using the Agilent Bioanalyzer 2100 (Agilent 213 Technologies, USA). cDNA libraries were constructed after confirming RNA purity (A260/A280: 214 1.7-2.2) and RNA integrity (RNA integrity number ≥7.0). Total RNA was transcribed to double-215 stranded cDNA and synthesized to cRNA. In this process, 2 nd -cycle cDNAs were generated and 216 further hybridized onto the microarray after fragmentation and biotin labelling. Microarrays were 217 washed and stained on the GeneChip Fluidics Station 450, and subsequent scanning was 218 performed using the GeneChip Scanner 3000 (Affymetrix, USA). The genes with fold change 219 ≥2.0 and p＜0.05 were considered significantly differentially expressed genes (SDEGs). The 220 potential pathways and Gene Ontology (GO) terms related to SDEGs were revealed by KEGG and 221 GO analysis. In addition, disease/gene-function analysis and interaction network analysis were 222 performed to explore the potential predominant diseases and genes affected by knockdown of 223 NRP1 and their associations. 224 225

Statistical analysis 226
All statistical analyses were conducted using SAS 9.43 statistical software (SAS Institute Inc., 227 Cary, NC, USA). One-way ANOVA was carried out to perform significance tests on the data 228 groups. Significant differences in continuous data (mean ± standard deviation) were evaluated 229 using the Student's t test. A p < 0.05 was considered to be statistically significant.

NRP1 is up-regulated in BC 234
Analysis of the expression of NRP1 in published profiles [8] from patients with BC showed a 235 frequent up-regulation of NRP1 in BC samples (13 cases) when compared to normal bladder 236 tissues (9 cases) (p < 0.01, Fig. 1a). A high level of NRP1 protein was also observed in BC 237 pathological sections using the Human Protein Atlas database. This staining was primarily present 238 in the cytoplasm and membrane of cancer cells (Fig. 1b). A total of six cultured BC cell lines, 239 including T24, 5637, UM-UC-3, J82, SW780, and SCaBER cells, were employed for further 240 experiments. We performed qRT-PCR to assess the expression of NRP1 in these cell lines. NRP1 241 was up-regulated in all BC cell lines but was particularly prominent in T24 and 5637 cells, which 242 have strong invasive ability (Fig. 1c). Western blotting presented similar results in these cell lines 243 (Fig. 1d). These results strongly demonstrated that NRP1 is up-regulated in bladder cancer. 244

NRP1 modulates BC cell proliferation and angiogenesis 246
To explore the role of NRP1 in the tumorigenesis and development of BC, we constructed stable 247 BC cell lines expressing one of three different shRNAs against NRP1 or a negative control 248 shRNA. shNRP1-1 generated the most consistent and significant down-regulation of NRP1 in T24 249 and 5637 cells and was therefore used in subsequent functional studies (Fig. 2a). In colony 250 formation assays, NRP1 knockdown caused a significant reduction in colony number in both T24 251 and 5637 BC cells (p < 0.05 for both) (Fig. 2b). Additionally, MTT assays indicated that NRP1 252 knockdown significantly inhibited the growth of T24 and 5637 cells, and compared to control 253 12 cells, the growth rate decreased by almost 2.0-fold after 5 days (Fig. 2c). Further, conditioned 254 medium from shNRP1 T24 or 5637 cells was able to significantly suppress the ability of tubule 255 formation by HUVECs (p < 0.05 for both) (Fig. 2d). These results demonstrated that NRP1 may 256 play a role in promoting proliferation and angiogenesis in BC. 257 258

Silencing NRP1 promotes BC cell apoptosis and cell cycle arrest 259
In order to explore the possible mechanism of the proliferation-promoting function of NRP1, 260 apoptosis was assayed in NRP1-knockdown cells. As shown in Figure 3a, silencing NRP1 261 increased the proportion of apoptotic cells compared to control cells. Cell cycle arrest is one of the 262 main ways to induce apoptosis. Flow cytometry analysis showed that NRP1 knockdown caused a 263 significant decrease in the percentage of cells in the G0/G1 peak and an increase in the percentage 264 of cells in the G2/M peak, but no statistically significant change was observed in the S peak ( Fig.  265 3b), indicating that NRP1 may promote proliferation in BC cells by reducing apoptosis through 266 mediating the G0/G1 and G2/M phase transitions. 267

NRP1 modulates the migration and invasion of BC 269
To evaluate whether NRP1 affects the process of migration and invasion in BC, we performed 270 Transwell assays in T24 and 5637 cells following NRP1 knockdown. As shown in Figure 3c and 271 3d, NRP1 knockdown significantly weakened the migration and invasion abilities in T24 and 5637 272 cells. Migration and invasion in T24 cells decreased by 51% (p < 0.05) and 72% (p < 0.05) after 273 NRP1 knockdown, respectively, and they decreased in 5637 cells by 61% (p < 0.05) and 65% (p < 274 0.05), respectively. Our results indicated that silencing NRP1 inhibited the migration and invasion 275 13 ability of BC cells. 276 277

Microarray analysis of dynamic gene expression after NRP1 knockdown in BC cells 278
To better understand the potential molecular mechanisms underlying BC malignant progression 279 associated with NRP1, we further conducted Affymetrix Gene Chip hybridization analysis in 5637 280 cells following stable NRP1 knockdown. After subsequent bioinformatic and normalization 281 analyses, we were able to distinguish the two groups clearly by hierarchical cluster and principal 282 component analyses. According to the microarray expression profiling data, 599 up-regulated and 283 880 down-regulated genes had at least a 2-fold change in expression (p < 0.05 for all) following 284 NRP1 knockdown. A heatmap of the significantly affected genes is presented in Figure 4a. Among 285 the significantly activated pathways (Fig. 4b), the cancer pathway was chosen to examine the 286 potential role of NRP1 in BC. We constructed a gene network map in this pathway to discover 287 potential NRP1-regulated genes (Fig. 4c), and SDEGs, including CDK4, BIRC3,CCNE2,CDK2,288 and FOS, emerged as the dysregulated genes associated with NRP1 knockdown. Western blot was 289 performed to verify changes in these genes with NRP1 knockdown. BIRC3 and CDK6 were up-290 regulated with NRP1 knockdown, and CDK4, CCNE2, FOS, and CDK2 were down-regulated ( Fig.  291 4d-e). 292

NRP1 is associated with the mitogen-activated protein kinase (MAPK) signalling pathway 294
By performing signalling enrichment analysis of the altered gene sets following NRP1 295 knockdown, we found that the differentially expressed genes were significantly associated with 296 the activation of p53 signalling, ERK/MAPK signalling, cAMP-mediated signalling, nuclear 297 14 factor kappa B (NF-κB) signalling, and G2/M checkpoint regulation, among others (Figure 5a). 298 Western blot analysis confirmed that NRP1 function was closely associated with the ERK/MAPK 299 and mitogen-activated protein kinase 8 (JNK)/MAPK signalling pathways. As shown in Figure 5b, 300 Ras, phospho-Raf (p-Raf), p-ERK1/2, and matrix metallopeptidase 9 (MMP9) were all decreased 301 in NRP1-knockdown cells, indicating that ERK/MAPK pathway activation is modulated by 302 NRP1. Further, the expression of JNK/MAPK signalling-related factors, such as p-JNK, p-c-jun, 303 and cyclin B1, were significantly lower in NRP1-knockdown cells (Fig. 5c), but the expression of 304 BCL2-associated X protein (Bax)/ BCL2 apoptosis regulator (Bcl2) and caspase 3 were higher, 305 which was consistent with the bioinformatics signalling enrichment assays and indicated changes 306 in the G2/M checkpoint regulation pathway. These results suggest NRP1 as a novel regulatory 307 mechanism of MAPK signalling that contributes to cell cycle modulation and drives tumorigenesis 308 in BC (Fig. 5d). 309 310 311

312
Over the past few decades, although encouraging progress has been made in the understanding of 313 the mechanisms of BC development [9] , metastatic BC remains incurable, and many patients have 314 tried the few therapeutic strategies unsuccessfully [10] . Therefore, identifying therapeutic targets 315 and discovering better treatment options for BC is vital. Angiogenesis has the ability to promote 316 growth, invasion, and metastasis in multiple cancers [11] . Through unremitting efforts, numerous 317 therapeutic agents have been developed to target angiogenesis, a pathway that largely influences 318 the clinical activity of bladder cancer. One such targeted agent is bevacizumab, a monoclonal 319 15 antibody targeting VEGF [12] . NRP1 is considered to be a co-receptor for VEGF and is 320 overexpressed in many human cancers. Recently, the overexpression of NRP1 has been reported 321 to be associated with tumour progression and poor prognosis in patients with BC, but the 322 underlying molecular mechanisms remain poorly understood [7] . Therefore, identifying the 323 mechanisms by which NRP1 modulates the progression of BC has significance for exploring and 324 optimizing the therapeutic strategy for urological malignancies. 325

326
In this study, we confirmed increased expression of NRP1 in BC cells and showed that 327 suppressing NRP1 inhibits cell proliferation, promotes apoptosis, and regulates migration, 328 invasion, and angiogenesis in human BC cells. NRP1 also regulates MAPK pathway activation. 329 Our results suggest that NRP1 plays a crucial role in the tumorigenesis and progression of BC. 330 They also provide evidence for the eligibility of NRP1 as a novel therapeutic target for BC. and three extracellular domains (a1a2, b1b2, and c) [13] . The membrane domain directly binds to type 335 III semaphorins and VEGF and can initiate downstream signalling. There are two major NRP 336 subtypes, NRP1 and NRP2. NRP1-and NRP2-knockout mice have hypoplasia and deficiency in 337 the neural system, emphasizing their roles in neural development [14] . NRP1 has been shown to be 338 overexpressed in numerous human tumour tissues, including breast, lung, colorectal, and 339 hepatocellular cancer [5,15] . Further, the expression of NRP1 is positively associated with prostate-340 specific antigen and Gleason score in prostatic cancer [16] , and overexpression may contribute to 341 16 autocrine-paracrine interactions in pancreatic cancer [17] . However, the role of NRP1 in the 342 pathogenesis of malignant diseases has not been deeply studied until now. Shi et al. found that NRP1 343 was highly expressed in oesophageal squamous cell carcinoma (ESCC), and inhibiting the 344 expression of NRP1 could suppress the proliferation of ESCC cells and the growth of xenografts [18] . 345 Cheng et al. demonstrated that NRP1 overexpression is an independent and novel prognostic factor 346 for BC patients, which can help clinicians identify high-risk patients for close follow-up and 347 intensive treatment [7] . 348

349
To explore whether NRP1 qualifies as a therapeutic target for BC, we assessed the levels of NRP1 350 in BC cell lines and found that it was high in all of the cell lines assayed. Additionally, silencing 351 NRP1 promoted apoptosis and reduced proliferation, angiogenesis, migration, and invasion in two 352 aggressive BC cell lines. These results clearly identify NRP1 as a tumour promoter in BC and 353 suggest that NRP1 has the potential to be an attractive target for BC treatment. To better 354 understand the role of NRP1 in the growth, invasiveness, and migration of BC, we further 355 performed global gene expression profiling using microarray technology. By comparing the gene 356 expression profiles between T24 cells with control shRNA and shNRP1, we observed that the 357 genes altered most significantly were mostly associated with tumour development. Among the 358 differentially expressed genes associated with the cancer pathway, 48 genes were differentially 359 expressed more than 2-fold. Among these genes, CDK6 was the most significantly up-regulated 360 gene, and CDK2 was the most significantly down-regulated gene. CDK6 plays an important role 361 in the cell cycle. To drive the progression of the cell cycle, CDK6 binds to and is activated by 362 cyclin D to enhance the transition through the G1 phase [19] . Wang et al. confirmed that the 363