Rcl1 suppresses tumor progression of hepatocellular carcinoma: a comprehensive analysis of bioinformatics and in vitro experiments

RNA 3’-terminal phosphate cyclase-like protein (Rcl1) is involved in pre-rRNA processing, but its implication in cancers remains unclear. RCL1 expressions in 21 malignancies was examinated through GEPIA website portal. Clinical implication data related to RCL1 level in Hepatocellular Carcinoma (HCC) samples were downloaded through TCGA, ICGC, GEO databases. Survival analysis and gene function enrichment analyses were performed through R software. The correlation between RCL1 expression and tumor immune infiltration was assessed via the TIMER2.0 database. The effects of Rcl1 overexpression or knockdown on cell growth and metastasis was evaluated by CCK8, transwell, and cell cycle assays. RCL1 expression is commonly down-regulated in HCC. The lower expression of RCL1 is associated with higher tumor stage, higher AFP level, vascular invasion, and poor prognosis. RCL1 expression has a significant correlation with immune cells infiltration in HCC, especially myeloid-derived suppressor cell (MDSC). Moreover, it was further identified that Rcl1 expression was reduced in HCC cell lines and negatively correlated with invasion of HCC cell lines. Immunofluorescence (IF) analysis revealed that the level of Rcl1 expression in the cytoplasm of HCC cells is significantly lower than that in the cytoplasm of L-02 cell. Moreover, both gain- and loss-of-function studies demonstrated that Rcl1 inhibited the growth and metastasis of HCC cells and regulated cell cycle progression in vitro. Rcl1 may serve as a novel tumor suppressor in HCC, and its biological effect needs further study. Rcl1 mRNA expression is down-regulated within HCC tissues and associated with poor prognosis and disease progression. Anti-cancer effects of Rcl1 on HCC were confirmed in vitro. Rcl1 may be a potential tumor suppressor in HCC.


Introduction
Primary liver cancer, the 6th most prevalent malignancy in the world, is the third leading cause of cancer-associated death worldwide [1]. Hepatocellular carcinoma (HCC) is the most common subtype of primary liver cancer (75-85%) [2]. Despite the large numbers of effective treatments of mid-advanced HCC exist, the long-term prognosis still remains poor. Therefore, lucubrating on the molecular mechanism of HCC development could contribute to the identification of new therapeutic targets.
Ribosome had initially been considered as a completely homogenous cellular organelle that simply synthesize the protein. However, ribosome heterogeneity was suggested having as evidence the findings that the ribosome proteome, ribosome gene transcriptome, and ribosome biogenesis factors differ between cells and tissues [3,4]. Besides, the different underlying causes of ribosomopathies and their tissue-specific phenotypes also pointed out the variation in the ribosome composition and function [5,6]. Furthermore, many studies have showed that the widespread deficiency in ribosome function and regulation of ribosome activity by oncogenes both could promote cancer development and progression [7]. Furthermore, it was reported that ribosome composition, maturation, and function could promote in the cancer chemo-resistance [8,9]. In particular, many agents targeted into ribosome could be sensitive to several cancers that had not responded to chemotherapy [10,11]. Therefore, studying the role of ribosome in cancer development becomes a necessity.
RCL1 encodes the RNA 3′-terminal phosphate cyclase-like protein, a number of RNA cyclase families but without cyclase activity [12]. In a yeast study, it was suggested that Rcl1p could serve as an endonuclease that affects the cleavage steps in the 5'-external transcribed spacer and internal transcribed spacer-1 regions of the ribosomal RNA precursor [13,14]. However, the endonuclease activity of Rcl1 in human cells is controversial [15]. Minguez et al. [16] reported that Rcl1 mRNA expression was associated with vascular invasion of HCC through transcription sequencing. Until now, the role and the exact mechanisms of Rcl1 in HCC development still remains unclear.
In the present study, we explored the mRNA expression and the clinical implications of Rcl1 in HCC patients by using several HCC cohorts. Moreover, the effects and mechanisms of Rcl1 in HCC cell line was further studied through in vitro experiments.

Data resource
RCL1 expression in various cancer and its relationship with tumor progression were analyzed via employing the GEPIA2 web portal (http:// gepia2. cancer-pku. cn/) and TISIDB web portal (http:// cis. hku. hk/ TISIDB). Public HCC gene expression matrix was download from Gene Expression Omnibus (GEO) database, Liver Hepatocellular Carcinoma Project of The Cancer Genome Atlas (TCGA-LIHC), and Liver Cancer-RIKEN, JP Project from International Cancer Genome Consortium (ICGC-LIRI-JP). The correlation of RCL1 expression with the abundance of immune infiltrates was obtained from the TIMER 2.0 (http:// timer. cistr ome. org/). The gene sets were downloaded from the Molecular Signatures Database (MSigDB) from the Gene Set Enrichment Analysis (GSEA) website (http:// www. broad insti tute. org/ gsea/ msigdb/).

Cell proliferation assay
Cell viability was monitored by Cell Count Kit 8 (Absin, China) according to the manufacturer's protocol. Briefly, HCC cells were counted and plated onto 96-well cell culture plates at a density of 2 × 10 3 /well. Proliferation rates were measured by absorbance of 450 nm at 0, 1-, 2-, 3-, and 4-days post-transfection. Experiments were repeated more than three times with similar results.

Transwell assay
For migration assay, HCC cells were seeded (7 × 10 4 ) onto the upper chamber wells with serum-free medium and then incubating for 48 h. The penetrated cells were fixed with 4% paraformaldehyde and stained with crystal violet. For the Matrigel invasion assay, the 24-well transwell chamber was coated with an extracellular matrix on the upper surface (Corning, USA). Then, HCC cells were seeded (1.5 × 10 5 ) onto the upper chamber wells with serum-free medium. After incubating for 48 h, the penetrated cells were fixed with 4% paraformaldehyde and stained with crystal violet. Then, migrating or invading cells were photographed at 100 × and counted in five random fields.

Cell cycle analysis
The

Statistical analysis
Statistical analyses were performed using R software (version 4.0.0, USA) or GraphPad Prism (La Jolla, USA). The limma package was used to evaluate Rcl1 expression between HCC tumor tissues and adjacent tissues using R software. The survival data of HCC patients were analyzed via log-rank test and Cox proportional hazard regression. According to the relevance of Rcl1 expression, genes in TCGA-LIHC were divided into Rcl1 positive and Rcl1 negative subgroups. Gene set enrichment analysis (GSEA) was performed using R software. Continuous data were expressed as the mean ± standard deviation (SD). Comparisons between groups were performed using Student's t test or ANOVA test. A value of p < 0.05 was defined as statistically significant.

RCL1 is abnormally expressed in various tumor tissues, and associated with prognosis and tumor progression
We first explored the general expression of Rcl1 in multiple human cancers using the GEPIA2 website portal (Fig. 1). The analyses of the RNA-seq data of 23 malignancies in TCGA suggested that the expression of RCL1 was significantly lower in Cholangiocarcinoma and LIHC compared to the adjacent normal tissues. However, RCL1 expression of Colon and Rectum adenocarcinomas were significantly higher than the one in normal tissues. The TISIDB web portal was used (Additional file 1: Fig.  S1A) to further evaluate the correlation of RCL1 expression and survival prognosis. Notably, RCL1 expression had a significant impact in the prognosis of 8 cancers, including brain lower-grade glioma (LGG), glioblastoma multiforme, kidney renal clear cell carcinoma (KIRC), LIHC, ovarian serous cystadenocarcinoma (OV), uterine corpus endometrial carcinoma (UCEC), uterine carcinosarcoma (UCS), and uveal melanoma (UVM). Low RCL1 expression was remarkably associated with poor prognosis in all these cancer types except for UCEC and UCS.
Moreover, the association between RCL1 expression and tumor progression across human cancers was identified. It was revealed that RCL1 expression was positively correlated to tumor stage in KIRC, LIHC, and Stomach adenocarcinoma, as well as in UVM (Additional file 1: Fig. S1B). Similarly, the expression of RCL1 in KIRC and LIHC was also positively associated with histological grade (Additional file 1: Fig. S1C). However, the RCL1 downregulation was notably correlated with higher grade of Cervical squamous cell carcinoma and endocervical adenocarcinoma and UCEC (Additional file 1: Fig. S1C).
In summary, these results confirmed that RCL1 could be a potential tumor-associated gene in several malignancies. Notably, it was in LIHC that RCL1 expression was not only significantly down-regulated but also associated with prognosis, tumor progression across many human cancers.

Low RCL1 expression is correlated with poor clinicopathological outcomes in HCC patients
Eleven HCC datasets were downloaded and analyzed to further verify RCL1 expression in HCC tumor tissue. For 9 HCC datasets including GSE22058 ( Fig  Univariate and multiple survival analyses were performed using R programming environment on the TCGA-LIHC dataset to promote understanding of the association between the RCL1 expression and the prognosis of HCC. Univariate analysis indicated that patients with the high RCL1 expression was associated with better overall survival (OS, HR = 0.607 (0.416-0.886)) and progression-free survival (PFS, HR = 0.661 (0.476-0.917)) ( Fig. 3A, B). Univariate and multivariate analyses further confirmed that RCL1 expression is an independent factor for OS (HR = 0.616 (0.420-0.905)) and PFS (HR = 0.701 (0.502-0.98)) of HCC patients (Fig. 3C, D, Table 1).
In addition, we studied the connection between RCL1 expression and the clinicopathological characteristics. It was found that low RCL1 expression was significantly correlated with female, advanced primary tumor (T classification) and TNM stage, higher AFP level, as well as vascular invasion in TCGA-LIHC cohort (Fig. 4A-E). Besides, it was found that a decrease in the RCL1 expression was associated with increasing T classification, HBV infection, portal vein and hepatic vein invasion in ICJC-LIRI-JP cohort (Fig. 4F-I). Meanwhile, a remarkable connection between RCL1 expression was lower in the patients with BCLC C stage, proliferation class, high AFP level, vascular invasion, as well as phosphorylation level of Akt, RPS6, and IGFR1 in GSE9843 cohort ( Fig. 4J-P). No significant relationship between the RCL1 mRNA expression and age (Additional file 2: Fig. S2A, C, F), cirrhosis (Additional file 2: Fig. S2B, E), as well as gender (Additional file 2: Fig. S2D, G) was found in HCC cohorts.

RCL1 expression significantly correlates with infiltrating levels of various immune cells in HCC
The correlation of RCL1 expression in HCC samples with immune infiltration levels were investigated using TIMER2.0 website portal. The results showed that RCL1 expression significantly correlated with the infiltrating levels of myeloid-derived suppressor cell (MDSC, r = − 0.395, p = 2.32e−14, Fig. 5G), endothelial cell (r = 0.336, p = 1.57e−10, Fig. 5H), hematopoietic stem cell (r = 0.296, p = 2.21e−08, Fig. 5I  Moreover, RCL1 expression was found to be significantly different between molecular and immune subtypes by exploring the TISIDB web portal (p < 0.001, Fig. 6A,  B). Furthermore, the expression levels of RCL1 in the patients with TP53 and IDH1 mutation were lower than the ones in the patients with TP53 and IDH1 wild-type (Fig. 6C, D). The expression levels of RCL1 in the patients with CTNNB1 mutation were higher than the ones in the patients with CTNNB1 wild-type (Fig. 6E) and no significant difference was observed between the expression levels of RCL1 in TERT mutation and wild-type patients (Fig. 6F).

Rcl1 is down expression in HCC cells and suppresses HCC cell growth and metastasis in vitro
The endogenous Rcl1 expression levels were detected in a collection of liver cancer cell lines and L-02 cells. Both mRNA and protein expression levels of Rcl1 were generally lower in all liver cancer lines in comparison with L-02 cells (Fig. 7A, B). The Rcl1 expression of highinvasive HCC cell lines were substantially lower than the one in the low-invasive cell lines. These findings were further supported by immunofluorescence staining. Interestingly, it was also revealed that there was a marked difference in the distribution of Rcl1 protein between the liver cells and cancer cells. In particular, Rcl1 was mostly located in the nucleus in the HCC cell lines, while it was uniformly distributed in nucleus and cytoplasm in the liver cell lines (Fig. 7C).
Then a recombinant plasmid vector encoding Rcl1 (FLAG-Rcl1) was conducted and an empty vector was used as control (Ctrl). The overexpression of Rcl1 in Huh-7 cells was validated with RT-PCR and western blot analyses (Fig. 8A, C). Forced Rcl1 expression could markedly inhibit cell growth as supported by cell viability assay in Huh-7 cell (Fig. 8E). Moreover, transwell assays indicated that the overexpression of Rcl1 significantly impaired Huh-7 cell's ability to migrate and invade (Fig. 8G). Furthermore, Rcl1 was knockdown by  (Fig. 8B, D). Consistently, knockdown of Rcl1 in Hep-3B cell strikingly enhanced cell viability, migration, and invasion (Fig. 8F,  H).

Rcl1 could potentially participate in regulating cell cycle and metabolism-associated signal pathways
Notably, the Rcl1 positive group of genes were enriched in multiple cellular metabolic processes such as xenobiotic, fatty acid, bile acid, adipogenesis, and oxidative phosphorylation, while the Rcl1 negative subgroup Fig. 4 Association between RCL1 expression and clinicopathologic characteristics in the HCC cohorts. RCL1 expression was lower in A female, B higher TNM stage, C higher T classification, D higher AFP levels, E vascular invasion in TCGA cohort, F higher T classification, G HBV, H portal vein invasion, I hepatic vein invasion in ICGC cohort, and J BCLC C stage, K proliferation class, L high AFP level, M vascular invasion, and the positive expression of N phosphorylated Akt, O phosphorylated RPS6, P phosphorylated IGFR1 in GSE9843 cohort of genes were enriched in cell cycle regulation including G2M checkpoint, E2F targets and mitotic spindle (Fig. 9A, C). Moreover, gene ontology (GO) analysis revealed that Rcl1 could potentially promote the activation of protein-binding and transmembrane transport, while simultaneously inhibiting microtubule and protein kinase activity (Fig. 9B). Mitochondria and chromosomes were the main cellular organelles of the Rcl1 positive and Rcl1 negative groups, respectively (Fig. 9D).
The effects of Rcl1 on cell cycle distribution was evaluated using flow cytometry analysis. We found that the overexpression of Rcl1 resulted in a significant increase of cells in G2M phase (35.10% vs 22.06%, p = 0.002) and a concomitantly significant decrease of cells in S phase (25.62% vs 39.53%, p = 0.0002) in Huh-7 cell (Fig. 9E). As expected, upon Rcl1 knockdown, the ratio of cells in G2M phase was decreased (18.89% vs 26.58%, p = 0.003), the ratio in S phase was increased (32.33% vs 17.59%, p = 0.0001), and the ratio in the G0/1 phase was decreased (50.78% vs 55.84%, p = 0.040) (Fig. 9F).

Discussion
Rcl1 could interact with Bms1 to endonucleolytic cleave the pre-rRNA at the A2 site, which is required for the eukaryotic pre-ribosome assembly implicated in the 18 s rRNA biogenesis [17][18][19]. However, the mechanism that allows Rcl1 to carry out the catalytic activity in human species still remains unclear since the mutation of the proposed pre-rRNA substrate binding site of human Rcl1 does not affect 2a cleavage or 18S production [15]. In the present study, we found that the expression levels of RCL1 was primarily reduced in HCC tissues and cell lines. Moreover, its expression significantly correlated with the survival prognosis, tumor progression, immune infiltration level, as well as molecular classification of HCC. Furthermore, both gain-and loss-of-function experiments demonstrated that Rcl1 has the inhibitory effects on the proliferation, migration, and invasion of HCC cells in vitro. Bioinformatic analyses revealed that Rcl1 could be significantly associated with cell cycle transition, cellular metabolic processes. Mechanistically, it was identified that Rcl1 overexpression or knockdown could inhibit or promote cell cycle transition.
RCL1 was found to be up-regulated in colorectal cancers, but downregulated in liver and bile duct cancers. Venkateswaran et al. found that non-coding variants of RCL1 could regulate granulocyte-macrophage colony-stimulating factor signaling in inflammatory bowel disease patients [20]. It is likely that single nucleotide polymorphism of RCL1 influences the gut microbiome. Gut microbiome has a key role in carcinogenesis and anticancer immune response of HCC and colorectal cancer [21,22]. And low RCL1 expression was associated with poor prognosis and tumor stage in multiple cancers. It suggested that Rcl1 may play a dualistic role by acting both as an oncogene and tumor suppressor, which is similar to other ribosome biogenesis factors [23].
Many studies have elucidated that many ribosome assembly factors could promote cell growth and metastasis in HCC by up-regulating the rate of ribosome biogenesis [9,[24][25][26][27][28]. But we found that RCL1 expression was commonly downregulated in HCC tissues and cell lines. Besides, bioinformatic analyses indicated that low RCL1 expression was the risk factor of poor survival prognosis and tumor progression, including: advanced TNM classification, high AFP level, vascular invasion in many HCC cohorts. And RCL1 expression was relatively lowest in iCluster 1 and proliferation molecular subtypes of HCC which exhibited the high frequency of macrovascular invasion and a significantly worse prognosis [29]. Recently, Zhu et al. reported that Rcl1 is essential for the 18S rRNA maturation at A1-site and for digestive organogenesis in zebrafish [30]. And its deficiency may upregulate the expression of genes responsible for ribosome biogenesis. Meanwhile, we also found that Rcl1 protein was distributed in both nucleus and cytoplasm of HCC cell and hepatocyte lines by immunofluorescence, which is consistent with a recent study [31]. Notably, the cytoplasmic Rcl1 protein level of HCC cell lines was dramatically reduced compared to hepatocyte cell. And it has been documented that the localization of BOP1, one of ribosome biogenesis factors, from nucleus to cytoplasm correlated with advanced disease and decreased survival in prostate cancer patients. Further research is needed to be proven whether Rcl1 could participate in the biological processes in the cytoplasm.
Moreover, functional tests indicated Rcl1 may be a potential tumor suppressor in HCC in vitro. Mechanistically, GSEA also suggested RCL1 was involved in cell cycle control and multiple cellular metabolic processes. It was further identified that Rcl1 overexpression could induce HCC cell cycle arrest by flow cytometer analysis. Consistently, Rcl1 knockdown obviously promote cell cycle progression of HCC cell. Large number of studies have demonstrated that aberrant pre-rRNA intermediates could lead to cell cycle arrest [32][33][34][35]. The most widely accepted model of cell-cycle arrest, upon pre-rRNA processing defect, was that the PeBoW complex could increase the extra-ribosome RPL5-RPL11-5S rRNA complex to bind and block the human homolog of mouse double minute-mediated ubiquitination and degradation of p53 [34,[36][37][38]. But it was documented that exogenous expression of Rcl1 in senescent cells could not increase the accumulation of p53, p21, and p16 [39]. Combining the bioinformatic analyses and flow cytometer analysis, we found that Rcl1 mainly control cell cycle progression by interfering the S phase.
Immune-related mechanisms play notable roles in the incidence and recurrence of HCC, and a combination of molecular and immune therapies could remarkably increase objective response in advanced HCC by 30-40% [40]. Ribosome biogenesis modulate immunosurveillance and innate immune response [41][42][43]. We also examined the correlation between the RCL1 expression and the infiltration levels of several immune cells in the TCGA-LIHC cohort. There exists a moderate negative relationship between the Rcl1 expression and the infiltration level of MDSC, CD4 + T cell, and Tregs, and significantly positive correlations with the infiltration level of endothelial cell, hematopoietic stem cell, and CD8 + T cells. MDSC contributes to the immunosuppressive network through multiple mechanisms and mediates the tumor growth, angiogenesis, and metastasis of HCC [44] while endothelial cell necroptosis induced by tumor-cells could reversely promote cancer metastasis [45]. Moreover, we also found a prominent association between the expression levels of RCL1 and the immune-subtypes in HCC patients. The RCL1 expression was relatively lower in the C1 subtype (wound healing). Its characteristics include a high proliferation rate, the upregulated expression of angiogenic genes, a Th2 cell-dominated immune  [46]. It is generally accepted that the interaction between Rcl1 and Bms1, a GTPase-activating protein, is involved in pre-rRNA processing across species [17,18]. Further observational studies are required to confirm whether the Rcl1-Bms1 complex could module the innate stimulation by activating RIG-I.
Our study has several limitations. First, this was a retrospective analysis based on public databases (TCGA and GEO). Although Rcl1 expression in human HCC cell lines has been analyzed, it should still be determined in clinical specimens. Second, despite in vitro experiments suggested that Rcl1 may be a tumor suppressor in HCC, additional animal tests should be performed. In addition, the exact molecular mechanism regarding Rcl1on cell cycle regulation should be further explored.

Conclusions
In summary, our study demonstrates that Rcl1 could serve as a favorable prognostic factor for HCC. Moreover, intracellular molecular metabolism and cell cycle control might be the primary biological processes regulated by Rcl1. Furthermore, the results of cell experiment indicated that Rcl1 plays a pivotal anti-cancer role by inhibiting of both growth and metastasis of HCC. The significant reduction of cytoplasmic Rcl1 protein   in HCC imply the additional biological function. Our study revealed that Rcl1 may act as a potential prognostic marker and tumor suppressor in HCC.