RA190, a Proteasome Subunit ADRM1 Inhibitor, Suppresses Intrahepatic Cholangiocarcinoma by Inducing NF-KB-Mediated Cell Apoptosis

Intrahepatic cholangiocarcinoma (ICC) is the second most common primary liver cancer, with a 5-year survival rate less than 10% [1, 2]. The morbidity and mortality associated with ICC have increased worldwide in the past decade [2-4]. Since most ICC patients are diagnosed at an advanced stage and surgical resection remains the optimum curative option for ICC [5, 6], standardized treatment for these patients is currently inadequate [3, 5, 6]. Moreover, although gemcitabine and cisplatin are commonly used for inoperable ICC patients, they respond poorly to these chemotherapies [3, 7]. Therefore, it is essential to identify novel anti-tumor targets for ICC.

The 26S proteasome is a major component of the ubiquitin proteasome pathway, which is responsible for more than 80% of protein degradation in mammalian cells [8]. Since cancer cells require the ubiquitin proteasome pathway to permanently activate pro-tumor signal cascades, which promote cell cycle progression and prevents cell death resulting from aberrant stress [9], the use of inhibitors targeting the ubiquitin proteasome pathway is an attractive strategy for treating different cancers [10, 11]. Among these inhibitors, bortezomib is the first agent approved by the US Food and Drug Administration (FDA) for treating multiple myeloma [12].

Recently, a ubiquitin receptor ADRM1 [13] was found to be overexpressed in several malignancies [14-18]. However, the role of ADRM1 in these tumors has not been explored fully. Moreover, knockdown of ADRM1 suppresses tumor growth in different malignancies [15-18]. RA190, a novel specific inhibitor of ADRM1, was found to have a significant suppressive effect on multiple myeloma [14, 19]. However, the expression of ADRM1 in ICC and the effect and mechanism of action of targeting ADRM1 in ICC remain unclear.

In this study, we found that ADRM1 was elevated in ICC tissues compared with normal bile duct tissues; this was further validated by Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) datasets. In addition, ADRM1 overexpression predicted poor prognosis in ICC patients. Furthermore, targeting ADRM1 showed a significant therapeutic effect in ICC via inducing G₂-M cell cycle arrest and apoptosis both in vitro and in vivo. Together, we have provided proof-of-concept evidence that ADRM1 is a promising anti-tumor target in ICC. More importantly, targeting ADRM1 with RA190 may improve patient outcome in ICC.

Human ICC cell lines QBC939 (donated by Professor SG Wang at the Third Military Medical University, China) and RBE (purchased from the cell bank of Typical Culture preservation commission of the Chinese Academy of Sciences, Shanghai, China) were cultured in Dulbecco’s modified Eagle’s medium (HyClone, Logan, UT, USA), containing 10% fetal bovine serum (Gibco, Grand Island, New York, USA) and 1% penicillin-streptomycin solution, at 37 °C with 5% CO₂. RA190 was purchased from Xcess Biosciences (San Diego, CA, USA). For in vitro studies, RA190 stock solution (50 mM) was prepared in dimethyl sulfoxide (DMSO) and stored at -20 °C as small aliquots until needed. For in vivo studies, RA190 was dissolved in 15% 2-hydroxypropyl-β-cyclodextrin (HPBCD), and the solution was prepared freshly every week and stored in the dark at 4°C before use.

The primary ICC cells (ICC-1 and ICC-2) were established using freshly resected human ICC samples as previously described [20]. The primary cells were validated by their unique DNA short tandem repeat “fingerprints” matching that of the patient’s tumor tissue.

In this study we used paraffin-embedded tissue samples from 102 consecutive ICC patients who underwent primary and potentially curative liver resections at the authors' institution from 2009 to 2014. Patients had not received any preoperative radiotherapy or chemotherapy. The clinicopathological and baseline demographic characteristics of the patients, including age, gender, tumor size, and tumor site, were retrospectively collected. Overall survival (OS) was calculated from the date of surgery to either the date of death or the last follow-up. Recurrence-free survival (RFS) was defined as the interval between the date of surgery and the first recurrence, or from the date of surgery to the date of last follow-up for patients without recurrence. Follow-up was terminated in December 2016. Ethical approval was obtained from the Research Ethics Committee, Fudan University, Shanghai, China and written informed consent regarding the use of tissue and data for scientific purposes was obtained from all participating patients.

Human ICC tumor tissue arrays were immunohistochemically stained with anti-ADRM1 antibody. The tissue array sections (5 µm) were dehydrated and subjected to peroxidase blocking. Primary antibodies were added and incubated at room temperature for 30 min using the DAKO AutoStainer and DakoCytomation EnVision+ System-HRP detection kit (Dakocytomation, Carpinteria, CA, USA). Slides were counterstained with hematoxylin. The stained slides were observed under microscopy, and images were acquired. Based on staining intensity, we classified the samples into four groups with increasing intensity from negative (Group 1) to strong (Group 4) for analysis [21].

Data are publicly available from the GEO (accession number: GSE26566, https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE26566) [22] and TCGA (https://cancergenome.nih.gov/) databases.

Lentiviral infection was performed as described [23]. The shRNA sequences were: ADRM1 shRNA1: 5′-CCGGCCCTGACGACTGTGAGTTCAACTCGAGTTGAACTCACAGTCGTCAGGGTTTTTG-3′, shRNA2: 5′-CCGGCCGCGGATGAGATCCAGAATACTCGAGTATTCTGGATCTCATCCGCGGTTTTTG-3′; IκBα: shRNA: 5′-CCGGCTCCGAGACTTTCGAGGAAATCTCGAGATTTCCTCGAAAGTCTCGGAGTTTTTG-3′.

RBE, QBC939, ICC-1, and ICC-2 cells were each seeded in 96-well plates with 1 × 10³ cells per well in triplicate, and then treated with RA190 or DMSO for various time periods. Cell viability was detected by Cell Counting Kit-8 (Dojindo, Kumamoto, Japan) following the manufacturer’s instructions.

RBE and QBC939 cells (each 300/well) and ICC-1 and ICC-2 cells were each seeded in 6-well plates each 1000/well per well in triplicate and cultured for 12 days. The colonies were fixed with 4% paraformaldehyde and stained with crystal violet. Colonies with more than 30 cells were counted. Representative results of three independent experiments with similar trends are presented.

Total RNA was isolated using Trizol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions and treated with RNase-free DNase. The reverse transcription reaction was performed using 2 µg total RNA per sample and the PrimerScript reverse transcription reagent kit (Takara, Shiga, Japan) according to the manufacturer’s protocol. After reverse transcription, real-time polymerase chain reaction (PCR) was performed using Power SYBR Green PCR MasterMix (Applied Biosystems, Foster City, CA, USA) and the ABI 7900 thermocycler (Applied Biosystems) following the manufacturer’s manual. The primer sequences are as follows: Human β-actin: forward 5′-TGACGTGGACATCCGCAAAG-3′, reverse 5′-CTGGAAGGTGGACAGCGAGG-3′; Human ADRM1: Forward 5′-CTGGCTGTGGCTCATGTTT-3′, reverse 5′-ACCCAAGACAGACCAGGATG-3′.

Cells treated with RA190 or DMSO for various time periods were harvested and fixed in 70% ethanol at -20°C overnight, and stained with propidium iodide (PI, 36 µg/mL, Sigma, St. Louis, MO) containing RNase (10 µg/mL, Sigma) at 37°C for 30 min, then analyzed for apoptosis and cell cycle profile with CyAnTM ADP (Beckman Coulter, Fullerton, CA). Apoptosis was measured as the percentage of cells in the sub-G₁ population. Data were analyzed with ModFit LT software [24].

Cells were treated with DMSO or RA190 for the indicated times. Then, cells were collected and stained with annexin V-fluorescein isothiocyanate (FITC) and PI using an annexin V-FITC apoptosis detection kit (BD Biosciences, San Jose, CA) followed by flow cytometric analysis. Data were analyzed with FlowJo 7.6 software [25].

Cell lysates were prepared with cell lysis buffer (Beyotime, Haimen, China) and the protein concentration in the lysates was quantified using an enhanced BCA protein assay kit (Beyotime). Protein samples (50µg) were loaded for immunoblotting using antibodies against γ-H2AX, cleaved-PARP, PARP, cleaved-CASPAS3, PUMA (Cell Signaling Technology, Inc., Danvers, MA, USA), p-IκBα (S32) (Sangon Biotech, Shanghai, China), WEE1, p21, p27, ADRM1 and GAPDH (Abcam, Cambridge, MA, USA).

Five-week-old male athymic nude mice were purchased from Shanghai Experimental Animal Center (Shanghai, China). QBC939 cells were trypsinized, resuspended in phosphate-buffered saline, and subcutaneously implanted into the groin (5 × 10⁶ cells per injection). The PDX mouse model was generated according to the previously reported protocol [26]. After 1 week, tumor- bearing mice were randomly divided into two groups (eight mice/group) and treated with RA190 or vehicle alone, twice a week for 4 weeks. Tumor size was measured weekly using vernier calipers. At the end of the study, the mice were killed and tumor tissues were collected, photographed, and weighed. Paraffin-embedded tissues were sectioned for immunohistochemistry of Ki-67 (Abcam), TUNEL assay (Promega, Madison, WI, USA), WEE1 (Abcam) and p-IκBα (Sangon Biotech). All procedures were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Data are presented as mean ± standard deviation. Student’s t test was used for the comparison of parameters between two groups or two-way analysis of variance when comparing more than two groups. Survival was analyzed using the Kaplan–Meier method. A Cox proportional hazards model was used to calculate hazard ratio (HR) and the corresponding 95% confidence interval (CI) with adjustment for potential confounders. Statistical Program for Social Sciences software 17.0 (SPSS, Chicago, IL, USA) was used for statistical analyses.

First, we examined the expression of ADRM1 in 12 paired ICC tissues and normal bile duct tissues and found that the protein level of ADRM1 was increased in tumor tissues compared with paired normal bile duct tissues (Fig. 1A and 1B). Then the mRNA levels of ADRM1 were tested in 167 cases of ICC tissues and 12 cases of normal bile duct tissues from our institution. The results showed that ADRM1 mRNA levels were increased in ICC tissues compared with normal bile duct tissues (Fig. 1C). The up-regulation of ADRM1 in ICC was further validated in reported GEO (GSE26566) [22] and TCGA datasets (Fig. 1D and1E). These data indicate that ADRM1 is overexpressed at both mRNA and protein levels in human ICC tissues.

To investigate the prognostic value of ADRM1 for ICC, we further analyzed the pathological clinical characteristics of 102 ICC patients. These patients were divided into high-ADRM1 (strong or moderate intensity) or low-ADRM1 (weak or negative intensity) groups according to protein level (Fig. 2A and 2B). We found that ADRM1 levels in tumor tissues were significantly correlated with tumor size (Fig. 2C and Table 1) and that patients in the high-ADRM1 group had shorter OS and reduced RFS compared with those in the low-ADRM1 group (Fig. 2D). High expression of ADRM1 was associated with poor OS and RFS of ICC patients in univariable (OS: p = 0.003, HR = 2.383, 95% CI = 1.357 to 4.188; RFS: p = 0.033, HR = 1.710, 95% CI = 1.045 to 2.796) and multivariable (OS: p = 0.006, HR = 2.246, 95% CI = 1.259 to 4.006; RFS: p = 0.031, HR = 1.760, 95% CI = 1.054 to 2.939) survival analyses (Table 2). In the analysis of TCGA data, we found that ADRM1 overexpression also predicted poor prognosis of liver cancer (p = 0.00177), lung cancer (p = 0.0214), and renal cancer (p < 0.001) and glioblastoma (p = 0.0248) (for all online suppl. material, see www. karger.com/doi/10.1159/000490210, Suppl. Fig. 1). These results demonstrate that ADRM1 is overexpressed in human ICC and associated with reduced OS and RFS.

ADRM1 overexpression in ICC suggests that it may be a useful therapeutic target. To validate this hypothesis, we knocked down ADRM1 in RBE and QBC939 cells using two shRNAs targeting two well-defined regions of ADRM1 (shA1 and shA2, respectively). The knockdown efficiency was validated by immunoblotting (Fig. 3A). ADRM1 knockdown significantly inhibited the proliferation of RBE and QBC939 cells (Fig. 3B). ADRM1 silencing also notably suppressed colony formation of these two cell lines (Fig. 3C). In addition, with regard to in vitro tumor formation ability, tumor growth in the shADRM1 group was significantly suppressed compared to the control group (Fig. 3D and 3E). These data indicate that ADRM1 is required for tumor growth of ICC.

Recent studies have demonstrated that RA190 is an inhibitor of ADRM1. RA190 can covalently bind to ADRM1 in the 19S regulatory particle, resulting in inhibition of proteasome function and accumulation of higher-molecular weight polyubiquitylated proteins [14, 19]. Since ADRM1 also mediates the degradation of IκBα [27], we hypothesized that RA190 could impede proteasome function and induce the accumulation of higher-molecular weight polyubiquitylated proteins and phosphorylated IκBα (p-IκBα) in ICC cells. Our results showed that RA190 could induce higher-molecular weight polyubiquitylated protein accumulation in ICC cells (Fig. 4A). We also found that RA190 could induce p-IκBα accumulation in a dose-dependent manner (Fig. 4B). To further determine the specificity of RA190 against ADRM1, we treated QBC939 shA1 and wild-type cells with RA190. Treatment with RA190 triggered a significant decrease in cell viability of wild-type cells (Fig. 4C). These data show that RA190-induced activity is dependent on ADRM1 in ICC cells.

To evaluate the efficacy of RA190 in ICC, we isolated two primary ICC cells (ICC-1 and ICC-2) from two fresh ICC specimens. These primary ICC cells and two ICC cell lines RBE and QBC939 were treated with RA190 for the indicated time. RA190 showed a significant suppressive effect on cell proliferation (Fig. 5A) and clone formation (Fig. 5B) of established ICC cell lines in a dose-dependent manner. By using two primary ICC cells, RA190 also exhibited similar suppressive effect on cell proliferation (Fig. 5A) and clone formation (Fig. 5B), in accordance with results from established ICC cell lines. These data demonstrate that inhibition of ADRM1 by RA190 suppresses the proliferation of ICC cells.

Previous studies demonstrated that ADRM1 was required for proper cell cycle progression [28]. We explored whether RA190 has an effect on the cell cycle of ICC cells. We observed a prominent increase of the G₂-M phase cell population after treatment with RA190 for 24 h (Fig. 6A). After 48 h of treatment with RA190, the sub-G₁ population, representing the apoptosis subgroup [29], was increased significantly (Fig. 6A). When RA190 treatment was performed after cell cycle synchronization by thymidine, the G₂-M phase cell population accumulated significantly in RA190-treated groups (see online suppl. material, Suppl. Fig. 2A). To investigate how RA190 induces G₂-M phase cell cycle arrest, we analyzed cell cycle regulators, including p21 [30], p27 [30, 31] and WEE1 [30, 32], which have been reported to be degraded by the proteasome after ubiquitination [33]. After RA190 treatment, both p21 and WEE1 accumulated significantly, but p27 levels were not affected (see online suppl. material, Suppl. Fig. 2B), indicating that RA190 induces G₂-M phase cell cycle arrest of ICC cells by accumulation of p21 and WEE1.

To determine whether RA190 indeed induces cell apoptosis, we examined the expression of apoptosis markers, cleaved-CASPAS3 and cleaved-PARP [30, 33]. Our results showed that cleaved-CASPAS3 and cleaved-PARP were increased when cells were treated with RA190 (Fig. 6B). Moreover, PUMA, which has a pro-apoptotic role [34], was increased significantly, and γ-H2AX, a DNA damage marker indicating cell apoptosis [30, 33], also accumulated significantly (Fig. 6B). RA190 induction of ICC cell apoptosis was further validated by annexin V/PI double staining assay (Fig. 6C). These data indicate that RA190 induces apoptosis of ICC cells. Taken together, our findings suggest that RA190-induced cell cycle arrest occurs before cell apoptosis in ICC cells.

RA190 can inactivate proteasome function by inhibiting ADRM1. In addition, ADRM1 functions as an adaptor protein regulating degradation of many functional substrates, such as the well-defined NF-κB inhibitor IκBα, [27]. Thus, we hypothesized that the substrates of ADRM1 contributed to RA190-induced cell apoptosis. As shown in Fig. 4B and see online suppl. material, Suppl. Fig 3, p-IκBα accumulated after RA190 treatment in ICC cells.

To determine whether p-IκBα accumulation contributed to RA190-induced apoptosis, we constructed an IκBα knockdown plasmid, and the knockdown efficiency was validated by immunoblotting (Fig. 7A). Results indicated that IκBα knockdown decreased the expression of PUMA upon RA190 treatment and attenuated the expression of the apoptosis markers, cleaved-CASPAS3 and cleaved-PARP, compared with the control group (Fig. 7A). Moreover, an apoptosis assay using annexin V/PI double staining further showed that IκBα knockdown attenuated RA190-induced cell apoptosis of ICC cells (Fig. 7B). These data suggest that p-IκBα accumulation is necessary but not sufficient for RA190-induced cell apoptosis.

To further explore the role of RA190 on ICC growth, we investigated the anti-tumor effect of RA190 in vivo. QBC939 cells were injected subcutaneously into nude mice. At the end of RA190 treatment (4 weeks), tumor size and weight were measured (Fig. 8A and 8B). As compared with the control group, tumor size and weight of the RA190-treated group were significantly reduced (Fig. 8A and 8B). No obvious treatment-related adverse effects on body weight, liver function, or kidney function during treatment were observed (see online suppl. material, Suppl. Fig. 4A). These data show that RA190 has promising anti-ICC effects and is well-tolerated in vivo.

To further validate the ICC inhibitory mechanism of RA190 in vivo, immunohistochemical analyses of proliferation and apoptosis were performed. As expected, increased TUNEL staining was observed in the RA190-treated group compared with the control group (Fig. 8C). Ki-67 staining decreased significantly in RA190-treated group compared with the control group, indicating that proliferation was suppressed by RA190 (Fig. 8C). Enhanced p-IκBα and WEE1 staining were observed in RA190-treated tumors, consistent with the in vitro expression (Fig. 8C). Finally, we randomly extracted proteins from treated and control tumors to measure the markers of apoptosis, p-IκBα, WEE1 and p21. Consistently, RA190-induced elevated expressions of cleaved-PAPR, cleaved-CASPAS3, and p-IκBα, as well as accumulation of WEE1 and P21, were observed in the RA190-treated group (Fig. 8D). Taken together our data indicate that RA190 has a similar anti-tumor effect and mechanism of action in vitro and in vivo.

We further validated the suppressive effect of RA190 in two PDX mouse models. Consistent with findings from the in vitro and subcutaneous mouse model experiments, RA190 showed a powerful suppressive effect on tumor growth in the PDX mouse. Tumor size and weight were reduced by RA190 treatment (Fig. 8A and 8B). Moreover, RA190 treatment was well tolerated (see online suppl. material, Suppl. Fig. 4A).

In the PDX mouse, TUNEL staining was also increased in the RA190-treated group and Ki-67 staining was decreased significantly (Suppl. Fig. 4B). Enhanced staining of p-IκBα and WEE1 were observed in RA190-treated tumors (see online suppl. material, Suppl. Fig. 4B). These data demonstrate that RA190 has a significant anti-tumor effect in the ICC PDX model, consistent with the findings of in vitro and subcutaneous mouse model studies.

Overexpression of ADRM1 was found in some malignancies [14-18], and high expression of ADRM1 indicated poor prognosis in ovarian cancer [17] and gastric carcinoma [18]. However, to our knowledge, its expression and roles in ICC have not been studied previously. Our results showed that ADRM1 was upregulated at the mRNA and protein levels in ICC tumor tissues. This up-regulation of ADRM1 was correlated with ICC tumor size and indicated a shorter RFS and OS in ICC patients. ADRM1 overexpression also predicted a poor prognosis in liver, lung cancer, and renal cancers and glioblastoma according to TCGA dataset. These findings suggest that ADRM1 has a pathogenetic role in cancer and may serve as a potential anti-cancer target in ICC.

Targeting ADRM1 by RNA interference significantly suppressed ICC cell proliferation both in vivo and in vitro. These data are consistent with some early studies showing that ADRM1 knockdown suppressed cell proliferation in hepatocellular carcinoma [15] and ovarian cancer [16]. These results imply that targeting ADRM1 is a promising strategy to treat cancers. However, there is no drug targeting ADRM1 in clinical use.

Recently, a small molecule targeting ADRM1 was developed. RA190, which can covalently bind to cysteine 88 of ADRM1 in the 19S regulatory particle of proteasome, inhibits proteasome function, resulting in higher-molecular weight polyubiquitylated protein accumulation [14, 19]. In HCT116 cells, RA190 can also bind with UCH37, a deubiquitinating enzyme interacting with ADRM1, to suppress proteasome function [35]. Meanwhile, it was reported that KDT-11 could also suppress multiple myeloma by targeting ADRM1 [36]. In ICC cells the specificity of RA190 on ADRM1 inhibition was confirmed by: (1) the accumulation of higher-molecular weight polyubiquitylated proteins and p-IκBα in RA190-treated ICC cells; and (2) reduced RA190 sensitivity following knockdown of ADRM1 in ICC cells [14].

Previous studies showed that RA190 can suppress cancer cell growth and induce cell apoptosis in multiple myeloma [14, 19] and ovarian carcinoma [19], and can overcome bortezomib resistance in multiple myeloma [14, 19]. In the present study, we found that RA190 significantly suppressed cell proliferation and colony formation in both established ICC cell lines and primary ICC cells, and targeting ADRM1 by RA190 in these ICC cells could induce G₂-M phase cell cycle arrest and the accumulation of WEE1 and p21, which was similar to previous findings [28].

Cancer cells have a tendency to escape cell apoptosis by inactivating apoptosis signal pathways through down-regulating pro-apoptotic proteins or up-regulating anti-apoptotic proteins. Thus, induction of cell apoptosis was an exportable anti-tumor strategy in cancer therapy [37, 38]. We noted that targeting ADRM1 by RA190 induced cell apoptosis of multiple myeloma [14, 19], but the mechanism remains unclear. It was also reported that RA190 induced cell apoptosis by triggering endoplasmic reticulum stress and resulted in an unresolved unfolded protein response [19]. In the present study, we found that targeting ADRM1 induced ICC cell apoptosis through increasing PUMA expression by inducing p-IκBα accumulation. IκBα is a substrate of ADRM1, and a key inhibitor of the NF-κB signaling pathway [27]. The NF-κB pathway is constitutively active in a variety of human cancers including ICC [39]. Abolishing NF-κB activation by curcumin or caffeic acid phenethyl ester suppressed ICC cell growth by inducing apoptosis [39, 40]. It was reported also that AKT/NF-κB activation suppressed PUMA expression by up-regulating Slung upon fragile histidine triad loss in non-small cell lung cancer [41]. Moreover, PUMA, a downstream gene of p53[42], was directly regulated by TNF-α through activating NF-κB [43]. We found that the NF-κB signaling pathway was suppressed by p-IκBα accumulation upon RA190 treatment and PUMA was up-regulated resulting in cell apoptosis. However, levels of p53 and p-p53 (s15) were not affected by RA190 (data not shown). Nevertheless, the mechanism of action was not considered in this study. Also, IκBα knockdown only partially rescued RA190-induced cell apoptosis and PUMA accumulation. It is possible that there are other signaling pathways involved in RA190-induced cell apoptosis, and further investigations are needed to understand the relationship between NF-κB signaling pathway and PUMA upon targeting ADRM1 and the signaling pathways involved in targeting ADRM1-induced cell apoptosis.

Recently, the PDX mouse model has been used to identify personalized therapy for cancer patients [44]. In our study, a subcutaneously transplanted tumor model of ICC and two PDX mouse models demonstrated the anti-tumor effects of RA190 in vivo. Consistent with its effects in vitro, RA190 suppressed transplanted tumor growth significantly. Our analysis revealed that RA190 induced apoptosis in treated tumors, suggesting that the mechanisms of action of RA190 are similar in vitro and in vivo. Notably, RA190-treated mice showed good tolerance during the entire treatment period. In fact, another proteasome inhibitor, bortezomib, was approved by the FDA. However, since bortezomib suppresses the degradation of all substrates in the proteasome system, it is highly toxic [45]. RA190 only blocks the degradation of specific substrates regulated by ADRM1. This selectively might provide a high degree of safety of RA190 for clinical application [14, 19, 28].

In summary, this is the first report that ADRM1 was overexpressed in ICC both at the mRNA and protein levels, and overexpression of ADRM1 was associated with tumor size and predicted poor prognosis in ICC. Based on both in vitro and in vivo investigations, including in established cell lines, primary ICC cells, and subcutaneously transplanted tumor and PDX models, we showed that RA190 suppressed ICC cell growth by inducing G₂-M phase cell cycle arrest and NF-κB-regulated cell apoptosis (Fig. 8E). These findings indicate that ADRM1 may provide a promising target for ICC treatment and support the clinical investigation of ADRM1 inhibitors (e.g., RA190 and KDT-11) for ICC treatment in the future.

This research was supported by the “973” State Key Basic Research Program of China (2013CB910500, L.X.Q), China National Natural Science Foundation (81672820, L.X.Q), China National Key Projects for Infectious Disease (2012ZX10002-012, L.X.Q), and the National Key Research and Development Program of China (2017 YFC 1308604, Q.Z.D).

The authors declare that they have no competing interests.

G.-Y. Yu, X. Wang and S.-S. Zheng contributed equally to this work.