Knockdown of CDCA8 inhibits the proliferation and enhances the apoptosis of bladder cancer cells

Expression profile data of bladder cancer

The processed expression datasets of bladder cancer were obtained from the GEO database (https://www.ncbi.nlm.nih.gov/geo/). The details of the four bladder cancer datasets are shown in Table 1, and the clinical information of GSE13507 dataset is shown in Table 2. The above sample data are obtained from the public database and hence do not require approval by the ethics committee.

After obtaining the GSE13507, GSE7476, GSE37815 and GSE65635 datasets, we used Perl software (version: 5.28) to transform the gene probe ID of the dataset expression profile into the gene symbol of the platform file and obtained four gene expression profile files containing standard gene symbol. We then extracted the expression value of CDCA8 in all samples from each dataset for differential analysis. Univariate and multifactorial COX regression analysis was performed by survival package in R language. In addition, we downloaded the bladder cancer gene expression data and overall survival data from TCGA database (https://portal.gdc.cancer.gov/) and extracted the expression data and prognosis data of CDCA8 in 19 normal samples and 414 bladder cancer samples for analysis.

Immunohistochemical analysis

We collected seven normal bladder tissues and 35 bladder cancer tissues from Haikou Municipal People’s Hospital from October 2018 to December 2019. All sample collections were approved by the Haikou Municipal People’s Hospital Ethics Committee under approval number SC20180238. We placed the tissue slices into a 63 °C oven for 1 h and performed dewaxing process involving 15 min in xylene and 7 min in anhydrous ethanol. We then used the Dako automatic immunohistochemistry pretreatment system to repair antigens. After repairing the antigen, the tissues were rinsed with PBST buffer 3 times for 1 min each. In accordance with the instructions, we diluted the first antibody with the antibody diluent. After adding an antibody, the samples were incubated at room temperature for 30 min. The slices were rinsed with PBST buffer 3 times for 1 min each and then added with the second antibody. The slice was incubated with diluted DAB for 10 min and then washed for 5 min. The slice was added with haematoxylin (SIGMA) for 1 min, immersed in 0.25% hydrochloric acid alcohol (400 mL of 70% alcohol + 1 mL of concentrated hydrochloric acid) for not less than 2 s, rinsed with tap water for more than 2 min, dried at room temperature and sealed. Finally, the dyeing results were read by a pathologist. The criteria for judging dyeing intensity and percentage were as follows: intensity: negative, score 0; weak, score 1; moderate, score 2; strong, score 3; percentage: 0%, score 0; 1–10%, score 1; 11–50%, score 2; 51–100%, score 3. The two scores were multiplied to obtain the final score (Hashimoto et al., 2020; She et al., 2019).

Cell culture

Human T24 and 5637 bladder cancer cell lines were purchased from the Shanghai Institutes for Biological Sciences. These cells were cultured at 37 °C in a 5% CO₂ atmosphere with DMEM containing 10% 10–15 mL of foetal bovine serum. The medium was changed every 3 days. The cells were washed 3 times with PBS and digested with 0.25% trypsin while the cells were in the logarithmic growth phase. The suspension of cells was separately inoculated into 6-well plates for the preparation for subsequent experiments.

Cell transfection

The lentivirus was transfected when the cell density in the 6-well plate reached approximately 15–30%. The lentiviral vector was synthesised by Shanghai Jikai Company. The target sequence of CDCA8 gene selected in this experiment was (si-CDCA8): 5′-TTGACTCAAGGGTCTTCAA-3′ with a GC content of 42.11%. The RNAi negative control sequence was (si-NC): 5′-TCTCCGAACGTTTCACGT-3′ with a GC content of 52.63%. The cells were transfected with MOI = 10, the amount of virus in the si-CDCA8 group and the si-NC group was 2 μL, and the virus titre was 5 × 10⁸ TU/mL. After the transfection was completed, the cells were further cultured in a 37 °C and 5% CO₂ incubator for approximately 72 h. The expression of the green fluorescent protein (GFP) in the cells was observed under a fluorescence microscope. GFP expression is used to judge transfection efficiency. In addition, we used quantitative real-time polymerase chain reaction (qRT-PCR) and Western blot (WB) to verify the effect of siRNA transfection. The transfected cells were then prepared for subsequent steps.

qRT-PCR assay

After 72 h of transfection, total RNA was extracted using 1 mL of Trizol. cDNA reverse transcription was then performed using an M-MLV kit (Promega, Shanghai, China). Approximately 2.0 μg of total RNA and 1 μL of Oligo dT (0.5 μg/μL) were added to the PCR system and then supplemented with RNase-free H₂O to 10 μL. The primers were: CDCA8 forward, 5′-GCAGGAGAGCGGATTTACAAC-3′ and reverse, 5′-CTGGGCAATACTGTGCCTCTG-3′; and GAPDH forward, 5′-TGACTTCAACAGCGACACCCA-3′ and reverse, 5′-CACCCTGTTGCTGTAGCCAAA-3′. The cycle condition was 95 °C for 5 min, followed by 50 cycles at 95 °C for 30 s and 60 °C for 45 s. The last step was 72 °C for 30 min. ∆Ct = Ct value of target gene is the Ct value of internal reference gene; −∆∆Ct = mean value of ∆Ct of NC group is the ∆Ct value of each sample; the expression level of target gene in each sample relative to NC group is represented by 2^−∆∆Ct (Ci et al., 2019; Livak & Schmittgen, 2001).

WB assay

After 72 h of transfection, the cells were washed with PBS 3 times, cleaved with radio immunoprecipitation assay lysis buffer, and then broken by 200 W ultrasound on the ice. The procedure was repeated 4 times, 5 s per time with an interval of 2 s. The supernatant was centrifuged for 15 min at 4 °C and 12,000×g. Protein concentration was determined by BCA reagent. Total protein was separated by SDS-PAGE, which consisted of 10% polyacrylamide sequencing gel, and then transferred to the PVDF membrane sealed with 5% skimmed milk powder TBST sealing solution for 1 h. CDCA8 antibody (Abeam, Shanghai, China) was diluted with PBS, incubated with the blocked PVDF membrane for 2 h or at 4 °C overnight at room temperature, and washed 4 times with TBST for 8 min/time. The secondary antibody was diluted with PBS, incubated with PVDF membrane for 1.5 h at room temperature, and then washed 4 times with TBST for 8 min/time. Finally, the PVDF membrane was placed in a cassette and added with A and B mixed solutions.

Cell proliferation assay

After 72 h of transfection, the cells were digested and inoculated into 96-well plates at a density of 4,000 cells/well. Cell counting kit-8 (CCK-8) (Sigma, Shanghai, China) was used to detect cell viability. In accordance with the operating instructions, the cell viability of T24 and 5637 was detected every 24 h (24, 48, 72, 96 and 120 h). Approximately 10 μL of CCK-8 reagent was added to the wells 2 h before the end of culture without changing the solution. The CCK-8 reagent was then added to the enzyme detector for 4 h to detect the OD value of 450 nm. A growth curve was then constructed. Each sample was tested 3 times.

Cell clone formation assay

The infected cells were digested with trypsin, resuspended and then counted in the complete medium. According to the number of 500 cells/holes, the cells were re-inoculated into the 6-well plate culture plate. Three multiple holes were set up in each group, the culture medium was changed every 2–3 days, and the cell state was observed. After 10 days of culture, cell growth was observed under a fluorescence microscope before the end of the experiment. The cells were then washed once with PBS, fixed with 1 ml of 4% paraformaldehyde for 30–60 min and then washed with PBS. Approximately 1,000 μL of clean crystal purple dye solution was added to each hole and allowed to stain the cells for 10–20 min. The cells were washed with ddH₂O several times, dried and photographed with a digital camera. Three repeated experiments were set up in each group.

Wound healing assay

The cells were digested with trypsin, suspended in complete medium and then counted. The cell density of the 96-well plate was adjusted to 50,000 cells/well, and the volume was 100 μL/well. Three multiple wells were set up in each group, and culture medium containing 1% serum was replaced the next day. The lower centre of the 96-well plate was aligned with the scraping device and gently pushed upward to form a scratch. The cells were washed 2–3 times with serum-free complete medium, cultured with medium containing 0.5% FBS and photographed at 0 h. Cells were cultured in a 37 °C incubator with 5% CO₂ and scanned with Celigo at 0, 8 and 24 h after scraping. Migration area was analysed by Celigo: migration area = Xh cell area − 0 h cell area. Three repeated experiments were set up for each sample.

Transwell assay

The chamber was placed in a new 24-well plate. Approximately 100 μL of serum-free medium was added to the upper chamber, and the plate was placed in a 37 °C incubator for 1 h. The cells in logarithmic growth phase were digested and prepared into a cell suspension with serum-free medium. The cells were counted and inoculated in a 24-well plate. The number of inoculated cells was controlled at 10⁵/well. The medium in the upper chamber was carefully replaced with 100 μL of the cell suspension. Approximately 600 μL of 30% FBS medium was then added to the lower chamber. The cells were cultured in a 37 °C incubator for 40 h. The plate was then inverted onto an absorbent paper to remove the culture medium. The nontransferred cells in the chamber were slowly removed with cotton swabs, and the cells were fixed in 4% paraformaldehyde solution for 0.5 h. After fixation, the fixative on the surface of the chamber was absorbed with paper. The cells were then stained with 1–2 drops of staining solution under the membrane so that the transferred cells could be stained for 1–3 min. The chamber was then soaked and rinsed several times and allowed to dry naturally. Four 100× photos and nine 200× photos were randomly captured from the Transwell’s visual field. The number of metastatic cells per field was counted using 200× photos. Three repeats was performed in each group.

Cell cycle analysis

After 72 h of transfection, the cells in the logarithmic growth phase were collected, and the supernatant was discarded after centrifugation. The cells were washed with D-Hanks (pH = 7.2–7.4). After centrifugation, the cells were fixed with 75% ethanol precooled at 4 °C for at least 1 h. The cells were suspended with 0.6–1 mL of cell staining solution. The formulation of the staining solution was as follows: 40 × PI mother solution (2 mg/mL): 100 × RNase mother solution (10 mg/mL): 1 × D-Hanks = 25:10:1,000. Cell cycle was detected by flow cytometry, and the results were analysed using the ModFit software. Three repeats were applied in each group.

Apoptosis assay

At logarithmic growth stage, cells were digested and washed with D-Hanks (pH = 7.2–7.4) at 4 °C after centrifugation. The cells were washed again with 1 × binding buffer. After centrifugation, 200 μL of 1 × binding buffer was added to obtain a cell suspension, and then 10 μL of Annexin V-APC was added. The reaction was carried out in an opaque environment at room temperature for 10–15 min. Flow cytometry was used to detect apoptosis. Results were analysed using the guava InCyte software. Each group was repeated 3 times.

Statistical analysis

Statistical analysis was carried out using SPSS 23.0 software, and the counting data were tested for normality and homogeneity of variance. Normality analysis was carried out by Kolmogorov–Smirnov test. Independent sample t-test was used to compare differences between the two groups. For the counting data which did not meet the t-test conditions of independent samples, the Mann–Whitney U nonparametric test was used to compare the differences between the two groups, and the Kruskal–Wallis test was used to compare data from more than three groups. Prognostic analysis was performed using the Kaplan–Meier and Log-rank tests. The difference was statistically significant at P < 0.05.

CDCA8 is upregulated in bladder cancer tissues

We extracted the expression values of CDCA8 from normal tissues and bladder cancer tissues in each dataset. The difference of CDCA8 expression between the two groups is shown in Fig. 1. Compared with normal bladder tissues, CDCA8 expression in bladder cancer tissues in the GSE7476 dataset was significantly higher than that in normal bladder tissues (Fig. 1A; P < 0.01). CDCA8 expression in bladder cancer was significantly increased in the GSE13507 dataset (Fig. 1B; P < 0.001) and in the GSE37815 dataset (Fig. 1C; P < 0.01). The results of GSE65635 dataset analysis also showed that CDCA8 expression in bladder cancer was significantly higher than that in normal tissues (Fig. 1D; P < 0.01). In the TCGA database, we obtained the same results. CDCA8 expression in bladder cancer was significantly higher than that in normal tissues (Fig. 1E; P < 0.001).

We analysed the correlation between CDCA8 expression and the prognosis of patients with bladder cancer in the GSE13507 dataset. Depending on the median expression of CDCA8 in bladder cancer tissues, the patients were divided into high- and low-expression patients. Cancer-specific survival analysis and overall survival analysis were carried out. The correlation between cancer-specific survival rate and CDCA8 expression is shown in Fig. 1F. The prognosis of patients with high CDCA8 expression was poor (P < 0.00028). The correlation between overall survival rate and CDCA8 expression showed the same results, and the prognosis of patients with high CDCA8 expression was poor (Fig. 1G; P < 0.0006). However, in the TCGA database, no correlation was observed between CDCA8 expression and the prognosis of patients with bladder cancer (Fig. 1H; P = 0.6).

Correlation between CDCA8 expression and the clinical characteristics of patients with bladder cancer

To further explore the effect of CDCA8 expression on the progression of bladder cancer, we obtained the CDCA8 expression value of each sample from the GSE13507 dataset and divided the samples into groups according to the clinical information of GSE13507. The difference of CDCA8 expression between groups in the same category was compared. As shown in Fig. 2, CDCA8 expression is higher in individuals with high stage and grade tumour compared with individuals with low stage and grade tumour (P < 0.05). We also found that CDCA8 expression was higher in advanced bladder cancer than in non-advanced bladder cancer (P < 0.05).

To understand whether the correlation between CDCA8 expression in GSE13507 dataset and the prognosis of patients with bladder cancer is independent of other clinical factors, we performed univariate COX regression and multivariate COX regression analysis on CDCA8 expression, gender, age, TNM stage and tumour grade and progression. In the results of multivariate COX regression analysis based on overall survival, no significant correlation was found between CDCA8 expression and the prognosis of patients with bladder cancer (Table 3; P = 0.114). In the results of multivariate COX regression analysis based on cancer-specific survival, CDCA8 expression was significantly associated with the prognosis of patients with bladder cancer (Table 4; P = 0.036).

Verification of CDCA8 expression in samples by IHC

We verified CDCA8 expression in seven normal bladder tissues and 35 bladder cancer tissues by IHC. As shown in Fig. 3, the expression score of CDCA8 in tumour tissues was significantly higher than that in normal tissues. In addition, we analysed CDCA8 expression in different clinical characteristics (gender, age, grade, lymph node invasion and T stage). The analysis results are shown in Table 5. CDCA8 expression in patients with high grade was significantly higher than that in patients with low grade (P = 0.023). CDCA8 expression in patients with high T stage was significantly higher than that in patients with low T stage (P = 0.031).

CDCA8 knockdown in T24 and 5637 cells

We transfected T24 and 5637 cells (si-CDCA8) with lentiviruses carrying siRNA vectors and used nontargeted siRNA sequences as controls (si-NC). The total RNA of the cells was extracted after transfection, and CDCA8 mRNA expression was detected by qRT-PCR. We also analysed CDCA8 protein expression by WB. Results showed that the expression rate of GFP was more than 80% (Figs. 4A–4D), indicating that the lentivirus was successfully transfected into the cells. CDCA8 expression was significantly reduced in T24 and 5637 cells, and the knock-down efficiency of mRNA was more than 70% (Figs. 4E–4G).

CDCA8 knockdown inhibits the proliferation of bladder cancer cells

After knocking down CDCA8 expression in T24 and 5637 cells, the absorptivity of wavelength at 450 nm in si-CDCA8 group and si-NC group was detected for 5 days. The value of OD450 could indirectly reflect the proliferation of cells. In T24 cells, the OD value after 5 days was significantly lower than that in si-NC group (Fig. 5A; P < 0.001). The same results were observed in 5637 cells (Fig. 5B; P < 0.001). In addition, the clone formation ability of T24 and 5637 cell lines was significantly inhibited in si-CDCA8 group compared with si-NC group (Figs. 5C–5G; P < 0.001). These results suggest that CDCA8 expression knockdown can inhibit the proliferation of bladder cancer cells.

CDCA8 knockdown inhibits the migration and invasion of bladder cancer cells

We then investigated the effect of CDCA8 knockdown on the migration of T24 and 5637 cells through wound healing experiments. As shown in Figs. 6A–6J, the ability of cell migration in si-CDCA8 group was significantly inhibited compared with that in the si-NC group (P < 0.001).

In addition, we conducted a Transwell assay to examine the effects of CDCA8 knockdown on the invasive ability of T24 and 5637 cells. The results showed that the number of invasive cells in the si-CDCA8 group was significantly lower than that in the si-NC group (Figs. 6K–6P; P < 0.001).

Effects of knocking down CDCA8 on the growth cycle of bladder cancer cells

To determine whether CDCA8 is involved in the regulation of T24 and 5637 cell growth cycles, we used flow cytometry to examine the effects of CDCA8 knockdown on the growth cycle of T24 and 5637 cells. The results showed that in T24 cells, the proportion of G0/G1 phase cells was decreased (P < 0.001), and the proportion of cells in S phase was increased (P < 0.01) after CDCA8 knockdown in si-CDCA8 group (Figs. 7A–7G). In 5637 cells, the proportion of G0/G1 phase cells in si-CDCA8 was decreased (P<0.001), whereas the proportion of cells in S phase and G2/M phase increased (P < 0.05), indicating that CDCA8 gene expression knockdown could result in 5637 cells S and G2/M phase growth arrest (Figs. 7H–7N).

CDCA8 knockdown promotes the apoptosis of bladder cancer cells

After T24 and 5637 cells were successfully infected with lentivirus for 5 days, the apoptosis of T24 and 5637 cells was detected by flow cytometry. As shown in Fig. 8, the apoptosis rates of T24 and 5637 cells in si-CDCA8 group were significantly higher than those in si-NC group (P < 0.001), indicating that inhibiting CDCA8 expression could promote the apoptosis of bladder cancer cells.

Analysis of the regulatory mechanism of CDCA8 in bladder cancer

In order to understand the regulatory mechanism of CDCA8 in bladder cancer, we calculated the CDCA8 co-expressed genes based on the transcriptome data of 414 bladder cancer samples in the TCGA database. The screening conditions were |Pearson correlation coefficient| > 0.4 and P < 0.01. A total of 978 genes were screened, including 171 negatively related genes and 807 positively related genes, and the top 20 positively related and top 20 negatively related genes were selected for expression heat map (Fig. 9A). We used the DAVID database (latest version 6.8, https://david.ncifcrf.gov) for GO functional annotation and KEGG pathway enrichment analysis. The GO function annotation includes three aspects: biological process (BP), molecular function (MF) and cell component (CC). The results showed that CDCA8 may have functions of regulating cell DNA binding, DNA repair and cell division in bladder cancer (Fig. 9B). KEGG pathway enrichment results showed that CDCA8 may affect the Cell cycle and P53 signalling pathways (Fig. 9C).

We divided the samples into two subgroups (high and low CDCA8 expression level subgroups) based on the CDCA8 gene expression values in these 414 bladder cancer samples and used GSEA analysis to predict the pathways that CDCA8 may affect. We used GSEA 3.0 software to calculate pathway enrichment scores and R language software to draw enrichment maps. We found several important pathways in the GSEA analysis results (Table 6; Fig. 9D), which are similar to those obtained using the co-expression analysis method.

In addition, we also used the GSE48075 dataset to analyse the correlation between CDCA8 expression and common mutant genes in bladder cancer. The results showed that CDCA8 overexpression was significantly associated with P53 mutation and RB1 mutation (Figs. 9E–9H).