Transcriptome and Proteome Analysis Reveals Corosolic Acid Inhibiting Bladder Cancer via Targeting Cell Cycle and Inducing Mitophagy in Vitro and in Vivo

Existing chemotherapy and radiotherapy methods have drawbacks such as high toxicity, high side effects, poor ecacy, and chemoresistance. Therefore, there is an urgent need to develop new anti-cancer drugs with low toxicity and high eciency for cancer therapy and the prevention of recurrence. Corosolic acid (CA) is a medicine and food homologue and has many biological activities of health care. However, the anti-tumor effects and mechanism of CA in bladder cancer remain unexplored. To study the anticancer effect of CA on bladder cancer cells, CCK8, EdU, high-content living cell imaging experiments were performed. And the mice model was used to verify the anticancer effect and toxicity for mice. To investigate the molecular mechanism of CA inhibition on bladder cancer cells, transcriptomics and proteomics were employed. To further verify the pharmacological mechanism, ow cytometry, RT-qPCR, western blotting and immunouorescence staining experiments were performed, and the CA targeting molecules were analyzed in combination with our experimental and public clinical data. We found that CA inhibited bladder cancer cell proliferation in a concentration- and time-dependent manner. Xenograft experiments showed that CA inhibited the growth of subcutaneous tumors, and had no toxicity in mice. Integration of omics analysis revealed that low concentrations of CA inhibited bladder cancer cells proliferation via attenuating DNA synthesis by downregulating TOP2A and LIG1, and via diminishing mitosis by downregulating CCNA2, CCNB1, CDC20, and RRM2. High concentrations of CA induced cell death not through the apoptotic pathway but through triggering mitophagy pathway via upregulating SQSTM1/P62, NBR1, and UBB. proliferation. EdU (5-ethynyl-2′-deoxyuridine) is a thymidine analog which can be substituted for thymidine in DNA synthesis. We used the EdU-594 cell proliferation detection kit (C0078L, Beyotime, China) to examine the synthesis of DNA according to manufacturer’s instruction. The results of the EdU staining were photographed by CellInsight CX7 High-Content Screening (HCS) platform (Thermo Scientic, US).

crape myrtle, loquat, and other plants. It has been discovered that CA has a variety of biological activities, including anti-oxidation, regulation of blood sugar levels, and anti-fungal and anti-tumor properties [14].
Owing to these bene ts, CA has been widely used as a health food supplement [15,16]. More importantly, due to its signi cant role in diabetes treatment, CA is known as "plant insulin" [17]. CA has been reported to inhibit certain protein tyrosine phosphatases to enhance insulin receptor β phosphorylation and stimulate glucose metabolism, which in turn decreases blood sugar levels [18].
In addition, recent progress has been made in anti-tumor studies involving CA. For instance, CA inhibits colorectal cancer cell growth by suppressing the PI3K/Akt/PKA signaling pathway through binding to the outer domain of HER3 and forming stable hydrogen bonds with Gly515, Arg444, Ser412, and Pro512 [19]. Moreover, Wang et al. reported that CA dose-dependently suppressed Y-79 retinoblastoma cell growth by blocking the cell cycle and inducing apoptosis by targeting maternal embryonic leucine zipper kinase (MELK) and forkhead box M1 (FoxM1) [20]. The anti-cancer function of CA in urinary system tumors has also been studied. CA has been found to inhibit TRAMP-C1 prostate cancer cell growth by decreasing the methylation level of nuclear factor erythroid 2-related factor 2 (Nrf2) promoter CpG sites, resulting in the upregulation of both mRNA and protein levels of Nrf2. In addition, CA can induce the expression of heme oxygenase-1 (HO-1) and NADH quinone oxidoreductase 1 (NQO1), suppressing the transformation of TRAMP-C1 cells [21]. However, the inhibitory effect of CA on bladder cancer and its underlying molecular mechanism remains unknown. In the present study, using cell proliferation, colony formation, and DNA synthesis assays, we found that CA signi cantly inhibited bladder cancer cell growth and veri ed the anticancer e cacy of CA. Using transcriptome and proteome analysis, we revealed a panoramic view of the mechanism by which CA inhibits bladder cancer: CA does not induce apoptosis but represses DNA replication and mitosis via signi cant downregulation of cell multiplication-related molecules TOP2A, LIG1, CCNA2, CCNB1, and CDC20. Interestingly, high concentrations of CA can lead to cell death in bladder cancer cells through induction of mitophagy by upregulating NBR1, SQSTM1/P62, UBB, and LC3. This study provides a comprehensive insight into the anti-cancer effects, toxicology, and pharmacological mechanisms of CA, which would provide compelling support for the development of anti-tumor drugs based on CA.
Dual uorescence staining of living/dead cells Calcein AM is able to stain living cells showing green uorescence; propidium iodide (PI) is able to stain dead cells showing red uorescence. The two uorescence probes (Calcein AM/PI) were employed to detect the activity of cell lactase and the integrity of cell membrane respectively, thus re ecting the cell activity and cytotoxicity. This kit was purchased from Beyotime (China, Beijing, C2015S).

Transcriptome and proteomic tests
In transcriptome sequencing, three control sampls and three CA-treated samples (5×10 6 cells each sample) were cleaved with Trizol, and then library building and mRNA sequencing were performed in Novogene (Beijing, China). For proteome assay, three control and three CA-treated (1.2×10 7 cells each sample) were lyzed with BD lysis buffer (containing 8M, 0.1M NaHCO3, 0.2% SDS), and then determined by 6-label TMT mass spectrometry at Norogena (Norogena, Beijing, China).

RNA extraction and quantitative RT-PCR
Total RNA was extracted from cells with Trizol reagent (Roche), and reverse transcribed to cDNA using high capacity reverse transcription kit (TIANGEN Biotech, China). Quantitative real-time PCR was performed using SuperReal PreMix Color (SYBR Green) (TIANGEN, China). Relative mRNA levels were normalized to GAPDH expression levels. All primers used for the PCR are listed in Table 1.

Flow cytometry for cell cycle and apoptosis test
To examine the effects of the CA on the cell cycle, ow cytometry was employed. The cells were harvested by trypsin digestion after treatment, and xed in 70% cold ethanol (in PBS) overnight at 4°C.

Immuno uorescencent Staining
Cells growing in 24-well plates were xed in 4% paraformaldehyde, labeled with primary antibodies 2 h at 37°C, and then incubated with species-appropriate secondary antibodies at room temperature for 1 h.
Nuclei were stained with DAPI, and images were collected using CellInsight CX7 High-Content Screening

Immunohistochemical Staining
Immunohistochemistry was employed to detect biomarkers in the xenografted tumor of nude mice. The 10% formalin xed blocks were cut into 4-µM slices and dehydrated. The endogenous peroxidase was inactivated by H 2 O 2 . The antigenic was repaired with citric acid solution, heating for 20 minutes. After being incubated with primary and secondary antibodies, the samples were developed with DAB solution (Solarbio, DA1015). After being re-dyed by hematoxylin, the slides were dehydrated and sealed. The antibodies used in immuno uorescencent staining were listed as following: Ki67 (1:300, Abcam, ab15580), CDC20 (1:150, Proteintech, 10252-1-AP).

Mitochondrial membrane potential detection
Mitophagy is accompanied by the attenuation of mitochondrial membrane potential. The effect of CA on the mitochondrial membrane potential of bladder cancer cells was detected by Mito-Tracker Red CMXRos kit (Beyotime, China, C1049B) according to the manufacturer's instructions.
Xenograft in nude mice 10 six-week-old BALB/C-nude mice were injected subcutaneous with 10 7 bladder cancer SW780 cells.
When the tumor grew to 100mm 3 , the mice were randomly divided into two groups with 5 mice in each group. The CA-treatment group was intraperitoneally injected once every 2 days at a dose of 8 mg/kg (CA was rst dissolved in DMSO and then dissolved in corn oil), while the control group was given DMSO and corn oil at the same volume at the same IP injection frequency. After administration, the tumor volume was measured every 2 days and the mice were weighed every 3 days. After 10 doses injection, the mice were sacri ced, and the subcutaneous tumors were taken, photographed and weighed, then xed with 10% formalin, dehydrated and made into para n blocks for subsequent studies. At the time of sacri cing mice, blood was collected for routine blood measurement.

Statistical analysis
The data were analyzed using one-way ANOVA and analysis was performed using Microsoft Excel and/or GraphPad Prism. P < 0.05 was considered signi cant.

Corosolic acid inhibited bladder cancer cell proliferation
Corosolic acid, also known as 2α-hydroxy ursolic acid, is known as a plant-resourced parainsulin due to its hypoglycemic function [18]. However, there are few investigations on the anti-cancer function of CA. In the present study, we provided multiple evidences suggesting that CA inhibited bladder cancer, through cell and animal experiments and multi-omics analysis. CCK8 kit was used to detect the inhibitory effects of CA against bladder cancer cells. As shown in Fig. 1B and 1C, bladder cancer cell growth was suppressed by CA in a dose-dependent manner. IC50 for SW780 and UM-UC-3 were 6.31 µg/ml and 7.25 µg/ml, respectively. Following treatment with CA, the morphological characteristics of the bladder cancer cells were changed: rst, the cell intensity was decreased; second, the cells were wrinkled, and attachment became unstable ( Fig. 1D and 1E). Because the proliferation marker Ki67 represents the percentage of highly proliferative tumor cells correlating with the S-phase fraction and mitosis, we detected Ki67 in CAtreated cells using a high-content cell imaging system. Figure 1F shows the representative image of Ki67 staining, and Fig. 1G shows the quantitative evaluation of the staining intensity analyzed by the cell imaging system. As shown in Fig. 1F and 1G, CA reduced the nuclear staining ratio of Ki67, indicating that CA inhibited bladder cancer cell proliferation. To study the time-dependent anti-cancer effect of CA, we incubated the bladder cancer cells SW780 and UM-UC-3 with CA for 24h, 48h, and 72h and recorded the cell activity and morphological alterations. As seen in Fig. 1H and 1I, 7.0 µg/ml of CA suppressed the growth of SW780 and UM-UC-3 cells in a time-dependent manner. Cell morphologies were gradually altered by prolonged treatment with CA ( Fig. 1J and 1K): while 48h treatment caused changes in cell shape, 72h treatment caused the bladder cancer cells to lose their attachment and extension features.
The above results indicate that CA can inhibit the proliferation of bladder cancer cells in a time-and dosedependent manner, suggesting that CA may be a potential natural source compound for the treatment of bladder cancer.

Corosolic acid inhibited transplanted tumor in vivo
To investigate the anti-tumor effect of CA against bladder cancer as well as its toxicity and side effects, we employed a xenograft nude mouse model. As shown in Fig. 1A, subcutaneous tumor growth slowed down with 8 mg/kg of CA treatment. After the mice were administered with CA, the tumor volume and tumor weight signi cantly decreased ( Fig. 2B and 2C); however, the mice's body weight did not change signi cantly (Fig. 2C). To study the effect of CA on tumor pathology, the tumor sections were analyzed by H&E staining and IHC staining for Ki67 ( Fig. 2D and 2E). As shown in Fig. 2E, the proliferation marker Ki67 was signi cantly decreased in the CA-treated group, indicating that CA treatment inhibited the proliferation of transplanted bladder cancer cells in nude mice. The side effects of CA were estimated through routine blood examination and kidney, heart, and liver dissection. As shown in Fig. 4F, CA did not cause a change the mouse blood index; only the average erythrocyte hemoglobin content (MCH) was increased. As shown by the H&E staining data in Fig. 4G-4I, CA did not affect the structure and H&E staining characteristics of mouse kidney, heart, and liver. The above data suggest that CA is an effective anti-bladder cancer compound with low toxicity to mice.

Corosolic acid inhibited bladder cancer via suppression of DNA replication and mitosis-related pathways
Pharmacological analysis based on multi-omics measurement has become a powerful method for revealing pharmacological mechanisms. In order to explore the anti-cancer mechanisms of CA, we performed transcriptome and proteome analyses of CA-treated bladder cancer cells along with control cells. As shown in the volcano map in Fig. 3A, using "p < 0.05 and absolute" and "value of Log2FoldChange greater than 0" as the screening threshold, CA-treated cells exhibited 3304 upregulated and 2917 downregulated transcripts. GO enrichment analysis of the differential transcripts revealed that the altered transcripts were enriched in DNA synthesis and cell mitotic division-related pathways (Fig. 3B). As shown in Fig. 3C, the heat map clearly showed that a number of genes related to DNA replication, including TOP2A, LIG1, and TYMS, as well as genes related to cell mitosis, including AURKB, RRM1, RRM2, and CDC20, were all signi cantly downregulated by CA treatment. To validate the results of highthroughput sequencing, we designed speci c primers and performed RT-qPCR. Figure 3D shows that CCNA2, CCNB1, TOP2A, and LIG1 were signi cantly downregulated by CA treatment, which was consistent with the high-throughput sequencing results.

Corosolic acid at low concentration inhibited bladder cancer cells via suppression of DNA replication
EdU (5-ethynyl-2′-deoxyuridine) is a substance that can only be incorporated into newly synthesized strands of DNA and is a powerful measure of the DNA synthesis rate. The inhibitory effect of CA on DNA synthesis in bladder cancer cells was evaluated using the EdU method. As shown in Fig. 4A-4D, CA dosedependently suppressed DNA replication in SW780 and UM-UC-3 bladder cancer cells, which was consistent with the enrichment results of the high-throughput transcriptome sequencing. To further con rm the anti-cancer mechanism of CA, we performed a proteomic assay on CA-treated bladder cancer cells. Then, a combined analysis of the transcriptome and proteome was performed (Fig. 4E). Through the transcriptome-proteome combined analysis, we found that many genes related to DNA synthesis and mitosis, including CCNA2, CDC20, TOP2A, AURKB, TYMS, RRM1, and RRM2 were downregulated by CA at both mRNA and protein levels. To double con rm the effect of CA on these molecules, immunoblotting and immuno uorescence staining were performed. The effect of CA on CCNA2 protein was veri ed by cellular immuno uorescence (Fig. 4G). Moreover, the protein alteration caused by CA treatment was analyzed by Western blotting. Figure 4H shows that CCNB1, CCNA2, and RRM2 levels were decreased by CA treatment. The inhibitory effect of CA on cell mitosis-related critical proteins was also veri ed in vivo in mice. Figure 4I and 4J show that CDC20 and RRM2 in subcutaneous tumors of mice treated with CA were signi cantly reduced compared with those in the control group. The above results strongly indicate that CA inhibits bladder cancer proliferation via the suppression of DNA replication and mitosis.
Corosolic acid at high concentration induced non-apoptotic cell death in bladder cancer cells During the analysis of morphological changes induced by CA treatment, we found that low concentrations of CA could decrease the growth rate of bladder cancer cells, while high concentrations could cause cell death. Figure 5A shows that the colony formation of bladder cancer cells was signi cantly inhibited. Notably, high concentration of CA resulted in the failure of cell clone formation and loss of the cells' stretched state, which could be a sign that the cells were predisposed to die. Using ow cytometry, we found that CA signi cantly reduced the percentage of cell population in the S and G2 phases (Fig. 5B), which veri ed that the pathways found enriched by multi-omics were rational and correctly identi ed. It is worth noting that although CA at a concentration of less than 7.0 µg/ml did not induce a signi cant increase in the subG1 cell population, but a high concentration of CA dramatically increased this population. This result indicates that high concentrations of CA might induce bladder cancer cell death. The calcein-AM/PI double staining results con rmed the above indication that 8.0 µg/ml of CA treatment led to bladder cancer cell death (Fig. 5C). Next, we performed annexin-V/PI double staining to test for apoptosis. As shown in Fig. 5D, CA did not induce apoptosis in bladder cancer cells, suggesting that CA could cause a non-apoptotic cell death mode. Although induction of apoptosis has long been the most common mechanism of anti-cancer agents [22], the discovery of various cell death modes in recent years enables a broader view for screening novel anti-cancer agents [23]. Here, we discovered that CA suppressed bladder cancer cell proliferation at low concentrations and induced nonapoptotic cell death at high concentrations.

Corosolic acid at high concentration induced mitochondrial autophagy
In this study, the mitochondrial autophagy (mitophagy) pathway was found to be enriched by differential protein enrichment analysis (Fig. S1). The heat map in Fig. 6A shows that the mitophagy markers LC3, SQSTM1/p62, NBR1, TAX1BP1, and UBB were upregulated upon CA treatment. We then tested the mitochondrial potential to evaluate mitochondrial damage caused by CA. As shown in Fig. 6B, mitochondrial potential decreased signi cantly with the increasing doses of CA, which is consistent with the enrichment analysis results. Since SQSTM1/p62 was reported to be an important mitophagy inducer because of its ability to promote mitochondrial ubiquitination independent of parkin [24], we examined SQSTM1/p62 by immuno uorescence staining. Figure 6C shows that treatment of the cells with high concentrations of CA resulted in the upregulation of SQSTM1/p62, which is an important marker of CAinduced mitophagy. To further verify the induction of mitophagy by CA, LC3 was detected by immuno uorescence staining. As shown in Fig. 6D, with increasing CA concentration, LC3 positive staining was gradually enhanced. Based on the proteomics data and immuno uorescence staining results, we summarized that CA upregulated the autophagy receptor NBR1 and TAX1BP1 which were able to act as an autophagy inducer. Moreover, CA augmented UBB protein level and increased SQSTM1/p62 (a mitophagy marker), in turn leading to LC3 accumulation which marked the onset of autophagy (Fig. 6E).

Discussion
As a natural compound with medicinal value and drug/food homology, CA has recently attracted widespread attentions due to its anticancer effect. Some scholars pointed out that CA inhibited cell proliferation through suppression of STAT3 and NF-κB [25],and some reported that CA promoted cancer cell death through induction of apoptosis via activating caspase − 3, 8 and 9 [26]. But there is a lack of systematic understanding of the anticancer mechanism of CA. Based on multi-omics and multiperspective analysis, we elucidated a comprehensive understanding on the pharmacological mechanism of CA inhibiting bladder cancer. CA inhibited the proliferation of bladder cancer cells when the concentration of CA was above 5.0µg/ml, and caused cell death when the concentration higher than 7.0µg/ml. 5.0µg/ml ~ 7.0µg/ml CA repressed bladder cancer proliferation was not just achieved by affecting a single protein, but through mitigating many DNA replication and mitosis related proteins (e.g.CCNA2, CCNB1, TOP2A, LIG1 and CDC20). These proteins have been demonstrated as therapeutic targets for many different tumors. CCNA2, which is short for cyclin A2, is a regulator of cyclin-dependent kinases that play an important role in cell cycle promotion [27]. The mRNA expression level of CCNA2 in bladder cancer patients was signi cantly higher than that in the normal population (Fig. 3E), suggesting that CCNA2 may be involved in bladder tumorigenesis. Recently, CCNA2 was considered as a bladder cancer therapy target by Li et al. [28]. The mRNA expression level of CCNB1 in bladder cancer patients was also signi cantly higher than that in the normal population (Fig. 3E), and CCNB1 has also been reported as a potential target for bladder cancer [29]. In the present study, we found that CA treatment resulted in the downregulation of CCNA2 and CCNB1, which suggests that inhibition of the CCNA2/B1dependent cell cycle is one of the molecular mechanisms by which CA inhibits bladder cancer. TOP2A encodes a DNA topoisomerase, an enzyme that controls and alters the topological states of DNA during transcription [30]. TOP2A is highly expressed in many tumors, including bladder cancer (Fig. 3E), gastric cancer [31], lung cancer [32]and liver cancer [33], and is widely considered as an anti-cancer target molecule [34]. In the present study, TOP2A expression was found to be suppressed by CA treatment, suggesting that repression of DNA replication is a mechanism of CA-inhibition of bladder cancer. LIG1 also plays an important role in DNA replication, as well as the base excision repair process. Although it has not been widely considered as an anti-cancer target molecule, many studies have reported that LIG1 abnormalities are associated with multiple tumorigenesis [35]. In this study, we found that CA signi cantly downregulated LIG1, which may lead to a disruption in DNA synthesis in bladder cancer cells. CDC20 (cell division cycle protein 20 homolog) is required for two microtubule-dependent processes: nuclear movement prior to anaphase and chromosome separation and chromosome segregation and mitosis [36]. CDC20 is well accepted as a novel therapeutic target for cancer [37]. Figure 3E shows that CDC20 mRNA expression levels in bladder cancer patients were signi cantly higher than those in the normal population, suggesting that CA might be a potential drug candidate to inhibit bladder cancer tumors by targeting CDC20. RRM2, short for ribonucleotide reductase M2, catalyzes the formation of deoxyribonucleotides from ribonucleotides and is a tumor biomarker [38]. Figure 3E shows that the RRM2 mRNA expression level in bladder cancer was signi cantly higher than that in the normal population. Recently, it was reported that inhibiting RRM2 could enhance the anti-cancer activity of chemotherapy [38], suggesting that CA could be a potent combination drug for chemotherapy, targeting RRM2.
Furthermore, we elucidated that high concentration of CA inhibited the bladder cancer mainly through the up-regulation of TAX1BP1, NBR1, UBB, SQSTM1/P62 and LC3, thus leading to mitochondrial autophagy. TAX1BP1 has been reported to be an autophagy receptor which is recruited and required for the clearance of stress-induced aggregations [39]. NBR1 functions as a speci c autophagy receptor for the selective autophagic degradation of peroxisomes, which in turn promotes mitophagy [40]. Besides, NBR1 is involved in cleaning autophagosomes, independently from SQSTM1/p62 [41]. In our study, we found that CA treatment increased TAX1BP1 and NBR1 expression, which would account for the induction of mitophagy. Ubiquitin B (UBB) is one of the major elements for the guidance of cellular protein degradation by the 26S proteasome [42]. Abnormally low expression of UBB has been found in many cancers. The transcriptional inhibition of UBB was found in a large number of women with genital tract cancer [43], and UBB expression is suppressed in approximately 30% of patients with high-grade serous ovarian cancer [43].
Autophagy is a kind of programmed death of eukaryotic cells. Although autophagy was previously thought to play a role in tumor development, Dr. Karlseder demonstrated that the clinical use of autophagy inhibitors did not have a positive effect on cancer patients, but had a tumorigenesis effect. Karlseder pointed out that abnormal DNA replication occurred when cells were stressed or in danger, which is called a "replicative crisis" [44]. Interestingly, cells in replicative crisis tended to die dominantly through autophagy pathway. Karlseder and Nassour discovered that when autophagy was blocked, cells in replicative crisis went crazy and went into a phase of in nite replication. This suggests that autophagy would be an important cancer inhibition mechanism [44]. Consistent to Karlseder's view, our ndings indicate that elevation of autophagy by CA could strongly inhibit bladder cancer cells, which might be a new anticancer strategy.
The graphical abstract in Fig. 7 describes the main mechanisms of CA-inhibition of bladder cancer but does not deny the existence of other mechanisms. For example, as shown in Fig. 3A, CA signi cantly upregulated the expression of FBXO32 mRNA. FBXO32 has been shown to be a target gene of TGFβ/Smad4 and plays the role of a tumor suppressor in a variety of tumors [45]. Moreover, the expression of FBXO32 in bladder cancer patients is signi cantly lower than that in the normal population. FBXO32 is a member of the ubiquitin-protein ligase complex and plays an important role in protein ubiquitination and degradation [45], and it might be responsible for the mitophagy of CA-induced bladder cancer cells.
Although the health value of CA has been widely recognized and health food has been developed due to its ability to lower blood sugar and anti-in ammatory properties [46,47], there is still a lack of studies on its anti-tumor activities. In particular, there is a need for studies revealing the comprehensive molecular mechanism of the CA biological activities based on multi-omics analysis.

Conclusions
Our study discovered the inhibitory effect of CA on bladder cancer using cell and animal model. Through combined transcriptome and proteomics study, we analyzed the underlying anti-cancer mechanisms of CA from a comprehensive perspective and revealed that low concentration of CA inhibited proliferation of cancer cells by repressing DNA replication and restricting cell mitosis through the suppression of LIG1, TOP2A, CCNA2, CCNB1, CDC20, RRM2 and so on, while high concentrations of CA induced mitophagy of bladder cancer cells through upregulation of SQSTM1/P62, NBR1, and UBB (Fig. 7). This study provides comprehensive insights into the pharmacological mechanisms of CA in inhibiting bladder cancer, which would be helpful in developing new anti-tumor drugs based on CA.

Declarations
Ethics approval and consent to participate The whole protocols were approved by the Ethics Committee of Jining Medical University, China.

Consent for publication
All the authors consent for this manuscript publication.
Availability of data and materials       CA induced bladder cancer cell death in a non-apoptotic manner. (A) Bladder cancer cells were coincubated with different concentration of CA, and the morphological alteration was recorded by a microscope. Note: as the red arrows showed, with the CA concentration lower than 6 μg/ml, bladder cancer cells growth speed decreased; while the CA concentration higher than 6 μg/ml, the cell morphology changed obviously (marked by red arrows). (B) The CA treated bladder cancer cells were stained by PI, and ow cytometry was employed to detect the cell cycle. The comparison of each stage was shown in the lower gure (column gure). The red arrow indicated impaired cells. (C) The DMSO or CA treated cells were triple stained by calcein AM / PI / H33342 to examine the cell death. H33342 stained cells were shown in blue, calceim AM stained cells were shown in green, and PI stained cells were shown in Red. (D) The cells were double stained by Annix-V-FITC / PI, and the ow cytometry was used for testing the apoptosis. Note: as shown in Fig.5D, the CA induced bladder cancer cell death by a nonapoptotic manner.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.