AMBRA1 attenuates the proliferation of uveal melanoma cells

Abstract Uveal melanoma (UVM) is the most common primary intraocular malignancy in adults with high metastasis rates. D-type cyclins (CCNDs) are central regulators of the cell division cycle and are among the most frequently deregulated therapeutic targets in human cancer. Recently, the E3 ligase adaptor, autophagy and beclin 1 regulator 1 (AMBRA1), was reported to regulate the stability of CCNDs, including CCND1, but its role in UVM has not been demonstrated. AMBRA1 is lowly expressed in UVM cells, and the ablation of AMBRA1 promotes the proliferation of 92.1 and OMM1 cells, whereas ectopically expressing AMBRA1 attenuates the proliferation of UVM cells. Further studies found that AMBRA1 promotes the ubiquitination and degradation of CCND1, and AMBRA1 regulates the proliferation of UVM cells in a CCND1-dependent manner. Thus, this study suggests that AMBRA1 serves as an important tumor suppressor by limiting UVM cell growth.


Introduction
Uveal melanoma (UVM) is a melanoma of the eyes, which arises from melanocytes in the uvea, 90% of which involve the choroid, 6% of which are confined to the ciliary body, and 4% of which involve the iris [1,2]. Approximately, 50% of patients diagnosed with UVM will develop metastasis within 10 years from the primary intraocular tumor [3,4]. The majority of UVM is thought to occur as sporadic [5], meaning everyone is at risk of developing UVM. The causes of somatic mutations in UVM, including sun exposure, radiation, cigarette smoking, and certain chemicals, are known to increase the risk of mutations in genes that control cancers [1].
The ocular treatment usually aims to conserve the eyeball, preserving the visual performance. The treatment of primary UVM includes phototherapy, radiotherapy, and local resection and reservation of enucleation in advanced cases [1]. So far, no standard treatment exists to reduce the risk of metastasis and improve the overall survival of patients [6]. Targeted therapy is widely used in cancer treatment due to its lesser side effects compared to other strategies. Novel therapeutics targeting mutant guanine nucleotide-binding protein G(q) subunit alpha/guanine nucleotide-binding protein subunit alpha-11, regulators of G-protein signaling, inactivated BAP1, and immune-checkpoint blockade is emerging [1]. Different inhibitors specifically target particular mechanisms of UVM occurrence and progression. Thus, new targets for melanoma treatment are urgently needed.
Protein ubiquitination is a dynamic multifaceted posttranslational modification involved in nearly all aspects of eukaryotic biology, including cancer development and progression, cell survival and differentiation, innate and adaptive immunity, and individual development and sexual maturity [7][8][9]. Ubiquitination includes a sequence of three enzymatic steps, and aberrations in the pathway can lead to tumor development and progression as observed in many cancer types [10], including UVM. BAP1 is a deubiquitinating enzyme, and its somatic and germline mutations have been found in UVM, malignant mesothelioma, and clear-cell renal cell carcinoma tumors [11].
D-type cyclins are central regulators of the cell division cycle and are among the most frequently deregulated therapeutic targets in human cancer [12]. The autophagy and beclin 1 regulator 1 (AMBRA1) was identified as the main regulator of cyclin D in several earlier studies, and it mediates ubiquitination and degradation of cyclin D as a substrate receptor for the cullin 4 E3 ligase complex [13,14]. The role of AMBRA1 in UVM has not been reported. In this study, the function of AMBRA1 in UVM was extensively studied.

Cell culture and transfection
The normal human pigment epithelial cell line ARPE-19 and three human UVM cell lines 92.1, MuM2B, and OMM1 were kindly gifted from Professor Shengfang Ge (Shanghai Jiao Tong University, China). All these cells were cultured in Dulbecco's Modified Eagle Medium (Hyclone, USA), supplemented with 10% fetal bovine serum (Gibco, USA) and 100 U/mL penicillin and 100 mg/mL streptomycin (Gibco), in a 37°C humidified atmosphere of 5% CO 2 . Plasmids were transiently transfected into cells using Lipofectamine 2000 (Life Technologies, USA) according to the manufacturer's instructions, and the stably transfected cell lines were screened by puromycin (5 μg/mL) or hygromycin B (100 µg/mL). When needed, cells were treated with cycloheximide (CHX; 100 μg/mL, Selleck, China) for 6 h before harvesting and then subjected to other analyses.

Reverse transcription quantitative PCR
(RT-qPCR) The total RNAs were extracted from cells using total RNA kit (Tiangen, Beijing, China). cDNA was synthesized using ReverTra Ace RT Master Mix (Toyobo, Osaka, Japan). qPCR assay was performed to assess the relative abundances of AMBRA1 and GAPDH mRNAs using specific primers ( Table 3), stained by SYBR Green (Tiangen) on the CFX96 real-time PCR system (Bio-Rad, California, USA). The relative abundances of AMBRA1 were normalized to that of GAPDH, using the 2 ΔΔCt method [15]. All data were obtained from three independent experiments.

AMBRA1 knockout cell line
AMBRA1 knockout cell line was generated using the CRISPR-CAS9 (clustered regularly interspaced short palindromic repeats and CRISPR-associated protein 9) technique as previously described [16]. Briefly, the 92.1 cells were transfected with CRISPR-CAS9-based sgRNA (PX330-AMBRA1-sgRNA), and monoclonal were chosen and detected by immunoblotting analysis. Then, the genetic ablation of AMBRA1 was confirmed by Sanger sequencing.

Cell viability assay
Cells stably transfected with shRNAs targeting AMBRA1 or CCND1 or ectopically expressing PCDH-AMBRA1 were

Colony formation assay
One thousand cells stably transfected with PCDH or PCDH-AMBRA1 were seeded into six-well plates. Seven days later, plates were fixed with 4% paraformaldehyde (Merck, Darmstadt, Germany) at room temperature (RT) for 15 min and stained with 0.1% crystal violet (Beyotime, Jiangsu, China) for 30 min. Images were obtained by a camera (Sony, Tokyo, Japan), and the number of colonies was counted and calculated.

Expression and purification of recombinant proteins
Recombinant proteins were purified as previously described [18]. His6-tagged proteins were expressed in BL21 Escherichia coli cells. After isopropyl-β-D-thiogalactopyranoside (Sangon) induction, the cells were pelleted, lysed in PBS buffer, and incubated with Ni 2+ TA beads (Sangon) to enrich the respective proteins, followed by elution with 500 mM imidazole (Sangon) dissolved in PBS buffer and then dialysed in PBS buffer supplemented with 20% glycerol before being aliquoted and preserved at −80°C.

Statistical analysis
The data were analyzed by one-way ANOVA with Tukey's post hoc test using GraphPad Prism 5 (GraphPad Software Inc., San Diego, USA). *P < 0.05 was considered to be significantly different; **P < 0.01 was considered to be very significantly different.

AMBRA1 is lowly expressed in human UVM cells
To explore the expression profile of AMBRA1 in human UM cell lines, the lysates of three human melanoma cell lines (92.1, MuM-2B, and OMM1) and one human pigment epithelial cell line (ARPE-19) were prepared. Immunoblotting analysis indicated that the protein levels of AMBRA1 were significantly higher in ARPE-19 cells compared to that of UVM cells (Figure 1a), which was consistent with the mRNA levels of AMBRA1 detected by RT-qPCR ( Figure 1b). However, AMBRA1 is highly expressed in UVM across tumors as revealed by TCGA (The Cancer Genome Atlas) data mined using the GEPIA2 (http://gepia.cancer-pku.cn/) ( Figure A1a).

The ablation of AMBRA1 promotes the proliferation of UVM cells
To were quantified and compared to that of GAPDH. The data are expressed as mean ± standard deviation (SD) and were analyzed using oneway ANOVA with Tukey's post hoc test. **P < 0.01, very significantly different, and the experiment was repeated thrice. (b) The mRNA levels of AMBRA1 in human pigment epithelial cells and UVM cells were detected by RT-qPCR. Data are expressed as the mean ± SD and were analyzed using one-way ANOVA with Tukey's post hoc test. **P < 0.01, very significantly different, three independent experiments. cells. Immunoblotting analysis indicated that AMBRA1 significantly decreased in cells stably transfected with shRNA1 and shRNA3 (Figure 2a), and these two shRNAs were chosen for further study. The cell viability of 92.1 and OMM1 was detected by the CCK-8 assay, and a significant cell growth promotion was observed in AMBRA1 knockdown cell lines compared to scramble groups (Figure 2b).
Two AMBRA1 knockout 92.1 cell lines (KO-1 and KO-2) were generated using the CRISPR-CAS9 sgRNA-based method and validated by immunoblotting analysis and Sanger sequencing (Figure A2a and b). The viability of 92.1 cells was increased in AMBRA1 knockout cells compared to wild-type (WT) cells as revealed by the CCK-8 assay ( Figure A2c).

AMBRA1 inhibits the proliferation of UVM cells
MuM-2B and 92.1 cells were stably transfected with empty vector PCDH or PCDH-AMBRA1 (Figure 3a). CCK-8 assay indicated that the cell viability of 92.1 and MuM-2B cells that stably expressing AMBRA1 was reduced than that of control groups (Figure 3b). Decreased colony numbers were observed in AMBRA1 overexpression groups compared to that of control groups, both in 92.1 and MuM-2B cell lines (Figure 3c). These results suggest that AMBRA1 inhibits the growth of UVM cells.

AMBRA1 reduces the stability of CCND1 protein
The E3 ligase adaptor, AMBRA1, has been reported to regulate the stability of CCND1 [13,14], but its function in UVM has not been reported. Immunoblotting analysis indicated that AMBRA1 promoted the degradation of CCND1 in a dose-dependent manner both in 92.1 and MUM-2B cells (Figure 4a), and the protein levels of CCND1 were upregulated in AMBRA1 knockdown ARPE-19 cells (Figure 4b). Further study indicated that the protein levels of CCND1 were upregulated in AMBRA1 knockout 92.1 cells compared to that of WT cells ( Figure A3a). These results suggested that AMBRA1 negatively regulates the stability of AMBRA1 protein.  empty vector PCDH or PCDH-AMBRA1 were seeded into six-well plates, and the colonies were fixed and stained. The images were captured using a camera, and the colony numbers were counted and calculated. The data were expressed as mean ± SD and analyzed using one-way ANOVA with Tukey's post hoc test. **P < 0.01, very significantly different, three samples for each group.

AMBRA1 interacts with and promotes the ubiquitination of CCND1
Co-IP analysis showed that endogenous AMBRA1 formed a complex with endogenous CCND1 in ARPE-19 cells (Figure 5a), and an in vitro ubiquitination assay was carried out with recombinant proteins indicated that AMBRA1 indeed mediates the ubiquitination of CCND1 (Figure 5b). As showed in Figure A4a  MUM-2B and 92.1 cells that were transiently transfected with different amounts of pCDNA3.0-AMBRA1-Flag or empty vectors, treated with or without 100 μg/mL CHX for 6 h, and then were subjected to immunoblotting analysis using indicated antibodies. The bands of CCND1 were quantified and compared to that of GAPDH. The data are expressed as mean ± SD and were analyzed using one-way ANOVA with Tukey's post hoc test. **P < 0.01, very significantly different, and the experiment was repeated thrice. (b) The protein levels of CCND1 were upregulated in AMBRA1-knockdown cells. The protein levels of AMBRA1 and CCND1 were detected by immunoblotting analysis in ARPE-19 cells that stably express AMBRA1 shRNAs after treatment with or without CHX (100 μg/mL) for 6 h. The bands of CCND1 were quantified and compared to that of GAPDH. The data are expressed as mean ± SD and were analyzed using one-way ANOVA with Tukey's post hoc test. **P < 0.05, significantly different; **P < 0.01, very significantly different, the experiment was repeated thrice. endogenous CCND1 in ARPE-19 cells or 92.1 cells. Further study indicated that the ubiquitination of CCND1 was significantly increased in 92.1 and MUM-2B cells that stably expressing PCDH-AMBRA1 compared to that of control groups, regardless of treated with proteasome inhibitor bortezomib (BTZ) or not (Figure 5c).  Figure 6: AMBRA1 regulates the proliferation of UVM cells in a CCND1-dependent manner. (a) Test of the knockdown efficiency of shRNAs targeting CCND1 by immunoblotting analysis. The shRNAs for CCND1 were transfected into 92.1 cells, and then hygromycin B (100 µg/mL) selection was used to establish stably expressing cell lines before subjecting to immunoblotting analysis. (b) Detection of the protein levels of AMBRA1 and CCND1 in stable expression cell lines. The 92.1 cells stably expressing CCND1 were transfected with an empty vector PCDH or PCDH-AMBRA1, selected by hygromycin B selection (20 µg/mL) and puromycin (5 µg/mL), and then subjected to immunoblotting analysis. The bands of CCND1 were quantified and compared to that of GAPDH. The data are expressed as mean ± SD and were analyzed using oneway ANOVA with Tukey's post hoc test. NS, no significantly different; **P < 0.01, very significantly different, the experiment was repeated thrice. (c) AMBRA1 inhibited the proliferation of 92.1 cells in a CCND1-dependent manner. The 92.1 cells stably expressing shRNAs of CCND, and PCDH or PCDH-AMBRA1, were seeded into 96-well plates. The cell viability was detected using the CCK-8 assay at the indicated time points. The data were expressed as mean ± SD and were analyzed using one-way ANOVA with Tukey's post hoc test. NS, no significantly different; **P < 0.01, very significantly different, and the experiment was repeated thrice. compared to that of the scramble group (Figure 6b). AMBRA1 inhibited the proliferation of control group cells, but not CCND1-knockdown 92.1 cells, as revealed by the CCK-8 assay at different time points (Figure 6c). These data suggested that CCND1 is essential for AMBRA1mediated growth inhibition.

Discussion
Recently, researchers from three collaborating teams of scientists in the United States and Europe found that AMBRA1 marks CCNDs involved in helping cells divide for destruction when cell division is not needed [13,14,19]. When the gene that produces AMBRA1 is damaged and the protein is missing or does not work properly, the cell cycle loses one of its main brakes, potentially letting cell division spiral out of control [13]. The loss of the AMBRA1 protein drives the tumor formation in mice and is linked with worse outcomes in some human tumors, and they also found that a lack of AMBRA1 may make some tumors resistant to drugs called CDK4/6 inhibitors, which are a promising new class of cancer treatments [14]. Thus, AMBRA1 serves as an important tumor suppressor by limiting tumor growth.
In this study, first, we found that AMBRA1 is lowly expressed in UM cells (Figure 1), whereas AMBRA1 is highly expressed in UVM across TCGA tumors ( Figure A1a). For lack matched TCGA normal data, it is difficult to judge whether AMBRA1 is highly expressed in human pigment epithelial tissues compared to UVM tissues, which needs to be further explored. Then, we found that the ablation of AMBRA1 promoted the proliferation of 92.1 and OMM1 cells (Figure 2), which was somehow consistent with recently reported that loss of AMBRA1 promotes melanoma growth and invasion [20]. Next, we verified that ectopically expressing AMBRA1 inhibits the proliferation and colony formation of UVM cells (Figure 3). Other studies also reported that AMBRA1 controls antigen-driven activation and proliferation of naive T cells [21].
Cyclin D1 (CCND1) is an important regulator of G1 to S phase progression in many different cell types, which is important for the development and progression of several cancers, including those of the breast, esophagus, bladder, lung, and eye. [22,23]. The overexpression of CCND1 is a common event in cancer but does not occur solely as a consequence of gene amplification. Rather, increased levels of CCND1 frequently result from its defective regulation at the posttranslational level. Besides CRL4-AMBRA1 E3 ligase complex, other E3 ligases, for example, SKP2, FBXO31, HACE1, and FBXO4, were also reported that mediated the ubiquitination and degradation CCND1 [24][25][26][27]. AMBRA1 promotes the degradation of CCND1 both in 92.1 and MUM-2B cells, and AMBRA1 knockdown increases the protein levels of CCND1 in ARPE-19 cells (Figure 4a and b). These suggest that AMBRA1 acts as an important factor for CCND1 degradation in UVM. One weak point of our study is that no in vivo experiments were involved, and an ongoing effort is to test the function of AMBRA1-CCND1 axis in animal models and patient samples of UVM. Wild type (WT) and AMBRA1 knockout 92.1 cells were seeded into 96-well plates, and the cell viability was detected using the CCK-8 assay at the indicated time points. The data were expressed as mean ± SD and analyzed using one-way ANOVA with Tukey's post-hoc test. NS, no significantly different; **P < 0.01, very significantly different, the experiment was repeated three times.