Modification of STIM2 by m6A RNA methylation inhibits metastasis of cholangiocarcinoma.

Background: N6-methyladenosine (m6A) is the most frequent internal methylation of eukaryotic RNA (ribonucleic acid) transcripts and plays an important function in RNA processing. The current research aimed to investigate the role of m6A-STIM2 axis in cholangiocarcinoma (CCA) progression.
Methods: The expression of STIM2 (Stromal Interaction Molecule 2) in CCA was measured using quantitative polymerase chain reaction (PCR) and immunohistochemistry (IHC). STIM2 was examined in vivo for its effects on the malignant phenotypes of CCA cells. The m6A modification of STIM2 was assessed through MeRIP (methylated RNA Immunoprecipitation)-PCR.
Results: Based on the GEPIA (Gene Expression Profiling Interactive Analysis) 2 database findings, a low STIM2 mRNA (messenger RNA) level was related to a poor prognosis in individuals with CCA. Quantitative PCR and IHC assays indicated decreased protein satin in CCA tissues and were associated with extrahepatic metastasis. Vianude mice tail vein injection model indicated that increased STIM2 levels suppressed CCA cell metastasis in vivo, while KRT8 (keratin 8) was detected as the direct downstream target of STIM2-mediated CCA cell metastasis in vivo. Meanwhile, based on SRAMP database and MeRIP assays indicated that m6A alteration resulted in abnormal STIM2 expression in CCA via METTL14 and YTHDC2. Conclusions: Our findings revealed the epi-transcriptomic dysregulation in CCA and metastasis by proposing a complicated STIM2-KRT8 regulatory paradigm based on m6A alteration. 2022 Annals of Translational Medicine. All rights reserved.

Introduction

Cholangiocarcinoma (CCA) is the second most aggressive form of primary liver cancer, with an increasing incidence worldwide (1,2). CCA has gradual signs, a high degree of malignancy, and a poor prognosis, as several parts of its clinical features are identical to those of hepatocellular carcinoma (HCCA) (3). Radiation therapy, chemotherapy, intervention, surgery, immune targeting, CCA ablation treatment, and other multidisciplinary approaches are all options for treating CCA (4-7). CCA is generally unresponsive to radiotherapy and chemotherapy, and consequently has inferior clinical outcomes.

CCA is a highly fatal malignant tumor originating from the bile duct epithelium, which is characterized by late clinical manifestations and lack of effective treatment. Chronic inflammation, including primary sclerosing cholangitis, Fasciola hepatica infection and hepatolithiasis, is listed as a risk factor, but the cause is unclear for most cases of CCA. The latest progress in molecular pathogenesis emphasizes the apparent legacy. The importance of genetic changes, including promoter hypermethylation and histone deacetylation, and genetic changes in cholangiocarcinoma (8). N6-methyladenosine (m6A) is a common RNA (ribonucleic acid) alteration that predictably affects carcinogenesis and tumor growth (9). The alteration of m6A RNA is a reversible process that is coordinated by methyltransferase (m6A “writers” and “readers”) and demethylase (m6A “erasers”) proteins (10-12). Recent research indicates that m6A modification may play a critical role in various malignant cancers, including CCA (13-16). However, the particular chemical mechanism by which m6A induces CCA development remains largely unknown.

Orai1 are essential primary proteins that are clustered and activated by STIM1 (Stromal Interaction Molecule 1) and STIM2 (Stromal Interaction Molecule 2). STIM1 is the direct activator of Orai1, and loss of the protein eliminates Orai1 activation, thereby regulating downstream Ca2+ (calcium)-dependent cell functions (17). STIM1 clustering of Orai1 in endoplasmic reticulum (ER)-plasma membrane (PM) junctions reinforces the coupling of Orai1-mediated Ca2+ entry with nuclear factor of activated T cells 1 (NFAT1) initiation (18). In contrast, the activation of diffusely localized channels in the membrane by the cytosolic STIM1 C terminus or SOAR (the STIM1 Orai activating region) fragments has been demonstrated in some research to cause nuclear translocation of NFAT1 and gene expression (17-20). STIM2 is also considered to be a relevant participant in pathological conditions associated with aging, Alzheimer’s and Huntington’s diseases, autoimmune diseases and cancer (21). STIM2 is considered to play a role in tumor development control; however, research on this is limited. However, the role of STIM2 in CCA regulation remains unclear. The purpose of this study was to investigate the clinical significance and biological roles of STIM2 in CCA.

We present the following article in accordance with the ARRIVE reporting checklist (available at https://atm.amegroups.com/article/view/10.21037/atm-21-6485/rc).

Methods

Samples and cell lines from human origin

The Eastern Hepatobiliary Surgery Hospital (EHBH) used CCA tissues and neighboring normal tissues from 10 CCA patients undergoing curative surgical resections. All procedures performed in this study involving human participants were in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by Ethics Review Committee of Eastern Hepatobiliary Surgery Hospital and all patients provided written informed consent. Human CCA cell lines HuCCAT1 (ATCC, Manassas, USA)were cultured in Roswell Park Memorial Institute (RPMI)-1640 medium with 100 g/mL streptomycin, 100 U/mL penicillin, and 10% fetal calf serum (GE Healthcare, Life Sciences, USA). The third-party biology services used short tandem repeat (STR) analysis to characterize all cell lines (Feiouer Biology Co., Ltd., Chengdu, China).

RT and qPCR (reverse transcription-reaction and quantitative polymerase chain reaction)

The PrimeScript RT reagent kit (Takara Bio Inc., China) produced the first-strand cDNA (complementary DNA). To synthesize microRNA (miRNA), the first-strand cDNA Poly-A Tailing Kit (Sangon Bio Inc., Shanghai, China) was used to reverse-transcribe miRNAs. In a quantitative RT-PCR, SYBR Premix Ex TaqII (Takara Bio Inc., Dalian, China) was utilized (qRT-PCR). The endogenous controls for Messenger RNA (mRNA) and miRNAs were β-actin and U6, respectively. The relative fold-change of target expression was calculated using the comparative 2-ΔΔCt technique. Sangon designed and manufactured all of the primers (Shanghai, China).

Immunohistochemical (IHC) staining

IHC was performed according to the previously described method (22). The evaluation fields were set at 200× magnification power for H-score calculations (23). In malignant cells, staining concentration was sorted as 0, 1, 2, or 3, referring to the existence of negative, weak, intermediary, or solid brown staining, respectively. All of the cells were counted and stained at different intensities to obtain an overall cell count. To calculate the average percentage of positive cells, we used the formula shown below.

H-score = (% of cells stained at intensity level 1×1) + (% of cells stained at intensity level 2×2) + (% of cells stained at intensity level 3×3). H-scores of 0–3 were obtained, with 0 indicating that all cells stained negatively and 3 indicating that all cells stained positively.

Establishment of stable cell lines overexpressing STIM2 and Keratin 8 (KRT8)

Hanbio provided the lentivirus vectors expressing STIM2 or KRT8 (LV-STIM2 or LV-KRT8) and its control vector (LV-NC) (Shanghai, China). CCA cells were infected with the lentivirus vector and then treated with puromycin (at a 2 µg/mL concentration) for 1 week.

In vivo experimentation through metastasis assays

A protocol was prepared before the study without registration. Animal experiments were approved by the Second Military Medical University’s Institutional Animal Care and Use Committee (Shanghai, China), in compliance with the Second Military Medical University guidelines for the care and use of animals. Four-week-old male BALB/c mice were purchased from the Second Military Medical University Animal Center and maintained in pathogen-free conditions with a regular pellet diet and water. Metastatic lung and liver models were used as previously described (23).

Methylated RNA immunoprecipitation (MeRIP) PCR

MeRIP-PCR analysis was performed according to a previously published technique (24).

Statistical analysis

The mean± standard error of the mean (SEM) were used to present the data. All data were obtained from three independent experiments. GraphPad Prism v.8 (San Diego, USA) and SPSS v.25.0 (Chicago, USA) software were used for all statistical analyses. Survival curves were calculated using the Kaplan-Meier method, and the differences were determined using the log-rank test. Univariate and multivariate Cox proportional hazards regression models were used to examine the independent components. P values less than *0.05, **0.01, and ***0.001 were deemed statistically significant.

Results

STIM2 deficiency in CCA tissues is related to a poor prognosis and extrahepatic metastases

According to the GEPIA(Gene Expression Profiling Interactive Analysis) 2 database, patients with CCA who had lower levels of STIM2 had a worse overall prognosis and were more likely to die from the condition (). STIM2 mRNA expression was significantly lower in 10 matched CCA tissues compared to adjacent normal tissues by quantitative PCR (qPCR). Notably, the CCA that metastasized extra hepatically had the lowest STIM2 mRNA expression (). Meanwhile, IHC revealed that STIM2 protein levels were commensurate with the trend in mRNA levels (). Collectively, these findings suggested that STIM2 may be involved in the evolution of CCA and may serve as a possible prognostic indicator for patients with CCA.

Figure 1

Decreased STIM2 in CCA tissues is associated with poor prognosis and extrahepatic metastasis. (A) STIM2 expression and overall survival or disease-free survival in CCA patients were compared using Kaplan-Meier survival analysis. A median expression level from the GEPIA 2 database served as the cutoff point. (B) The mRNA expression levels of STIM2 were examined in 10 paired CCAs (with or without extrahepatic metastases) and neighboring normal tissues. (C) IHC pictures showing STIM2 staining in the CCA tumor or surrounding tissues (magnification ×200). (D) H-score from (C). ± SD. *, P<0.05 indicated statistical significance. CCA, cholangiocarcinoma; STIM2, Stromal Interaction Molecule 2; GEPIA 2, Gene Expression Profiling Interactive Analysis; mRNA, messenger RNA; IHC, immunohistochemical; SD, standard deviation.

Inhibition of tumor metastasis by STIM2 in vivo

The mRNA level of STIM2 was then studied in three CCA cell lines, with RBE exhibiting relatively lower STIM2 expression, while HUCCAT1 and Huh28 exhibited markedly higher STIM2 expression (). To investigate the role of STIM2 in preventing CCA cell metastasis, RBE cells were infected with LV-STIM2 or LV-NC (), and the effects of STIM2 on tumor metastasis were examined in vivo. To create the liver orthotopic-implanted models, cells were infected with LV-STIM2 or LV-NC and transplanted into the livers of nude mice. Six weeks after transplantation, the fluorescence signal intensities of metastatic liver nodules in the LV-STIM2 group were significantly lower than those in the LV-NC group (). There were significantly fewer metastatic foci found in liver tissue slices from patients who had been treated with LV-STIM2 (), indicating that STIM2 inhibited the ability of CCA cells to spread intra-hepatically. Using nude mice as a model for lung metastasis, we injected cells labeled with firefly luciferase into their tail veins. Compared to the LV-NC group, the LV-STIM2 group had significantly lower bioluminescence signal intensities in mice, substantially lower fluorescence signal intensities of metastatic lung nodules (), and fewer metastatic foci in lung tissue sections (), indicating that STIM2 can inhibit hepatoma cells’ metastatic lung potential.

Figure 2

STIM2 inhibits tumor metastasis in vivo. (A) qRT-PCR was used to verify the mRNA levels of STIM2 in three CCA cells. (B) RBE cells overexpressed STIM2. qRT-PCR was used to verify the efficacy of the overexpression. (C) Photographs of the intrahepatic metastases taken in bioluminescence. (D) Sample microscopic pictures of metastatic lung foci stained with HE to show the indicated CCA cells in liver tissue from which to extrapolate, magnification ×50. (E) Quantitative findings from (D). (F) The IVIS@ Lumina II imaging system (Caliper Life Sciences, Hopkinton, MA, USA) captured bioluminescence images of the mouse tail vein and injection metastasis lung models at specific periods. (G) A microscopic picture of pulmonary metastatic foci identified CCA cells in pulmonary tissue sections stained with HE, magnification ×50. (H) Quantitative statistics from (G). *, P<0.05. qRT-PCR, quantitative reverse transcription-polymerase chain reaction; CCA, cholangiocarcinoma; STIM2, Stromal Interaction Molecule 2; HE, hematoxylin-eosin.

STIM2 inhibits tumor metastasis by decreasing KRT8 levels in vivo

Next, we explored the targeted gene of STIM2 inhibition of tumor metastasis in vivo. The UALCAN database (https://ualcan.path.uab.edu/) showed that the top five genes (KRT18, KRT8, MPST, MRPS17, and C20orf24) were negatively correlated with STIM2 in CCA (). We observed that LV-STIM2 lowered KRT8 mRNA expression in CCA cells, indicating that STIM2 may regulate KRT8 expression in CCA cells (). As a result, we are curious as to whether STIM2 inhibits tumor metastasis via decreased KRT8 levels. SK-Hep1 cells transfected with LV-STIM2 or LV-NC were infected in vitro with LV-KRT8 or LV-N, respectively (). LV-STIM2 decreased the capacity of CCA to metastasize in orthotopic implanted and lung metastasis models. Simultaneously, LV-KRT8 restored the inhibitory effect of STIM2 on CCA cell metastasis in vivo (), demonstrating that STIM2 suppresses tumor metastasis in vivo by lowering KRT8 levels.

Figure 3

STIM2 inhibits tumor metastasis by decreasing KRT8 levels in vivo. (A) Five top genes (KRT18, KRT8, MPST, MRPS17, and C20orf24) negatively correlated with STIM2 in CCA based on the UALCAN database (https://ualcan.path.uab.edu/). (B) The mRNA level of KRT8 in CCA with LV-NC or LC-STIM2 was validated using qRT-PCR. (C) KRT8 was overexpressed in CCA cells with LV-NC or LV-STIM2. The overexpression efficiency was validated using qRT-PCR. (D) Representative bioluminescence photographs of the intrahepatic metastases. (E) Microscopic representations of pulmonary metastatic foci in liver tissue slices stained with HE identified CCA cells, magnification ×50. (F) Statistics derived from (E). (G) The IVIS@ Lumina II imaging system captured bioluminescence images of the mouse tail vein and injection lung metastasis models at specific times. (H) Representative microscopic pictures of pulmonary metastatic foci from indicated CCA cells were obtained through HE staining, magnification ×50. (I) Quantitative statistics from (H). *, P<0.05. CCA, cholangiocarcinoma; STIM2, Stromal Interaction Molecule 2.

Abnormal STIM2 levels in CCA are responsible for m6A methylation modification

Through GEPIA 2 analysis, we observed the relationship between STIM2 and the expression of methylase METTL3, METTL14, WTAP, and KIAA1429, as well as demethylase ALKBH5 and FTO. In the CCA, we detected that STIM2 was related toMETTL14 and YTHDC2 (). Meanwhile, shows that STIM2 mRNA had three very confident m6A sites, indicating that m6A modification may be related to aberrant STIM2 levels in CCA, as demonstrated by the SRAMP (sequence-based RNA adenosine methylation site predictor) database (https://www.cuilab.cn/sramp). Further research using a RIP (RNA immunoprecipitation) assay confirmed the presence of the m6A-3-site in STIM2 mRNA (). In the clinic, MeRIP-qPCR results showed that in 10 paired CCA patients, the m6A CPEB1 level was lower in the tumors than the surrounding normal tissues (). According to the above findings, aberrant STIM2 levels in CCA are caused by m6A alteration.

Figure 4

Abnormal STIM2 levels in CCA are responsible for m6A methylation modification. (A) Expression correlation with methylase (METTL3, METTL14, WTAP, and KIAA1429), demethylase (ALKBH5 and FTO), and readers (YTHDC1-2 and YTHDF1-3) in CCA based on the GEPIA database. (B) The m6A site of STIM2 mRNA was predicted by the SRAMP database. (C) m6A-site level STIM2 in CCA cells were obtained by MeRIP-qPCR. (D) m6A STIM2 levels in 10 paired primary CCA were obtained by MeRIP-qPCR. *, P<0.05. CCA, cholangiocarcinoma; STIM2, Stromal Interaction Molecule 2.

Discussion

CCA, which develops from epithelial cells confronting the biliary tree lumen, is the second most common primary hepatic tumor worldwide after HCCA (25). As a result of the lack of early recognized clinical symptoms and the unavailability of precise tumor biomarkers, CCA is typically detected at an advanced, incurable stage (26). Despite recent advances in the development of molecularly targeted medicines, the outlook for this lethal malignancy remains bleak. As a result, there is a high priority in unraveling the molecular mechanisms and pathways underlying this disease to improve the clinical outcomes of CCA patients. STIM2 is the primary protein involved in Orai1 clustering and activation (27).

There is little known regarding the involvement of STIM2 in the advancement of CCA. Based on the GEPIA 2 database, this study discovered that low STIM2 mRNA levels were related to poor survival in CCA patients. Quantitative PCR and IHC studies revealed a decrease in the protein expression in CCA tissues was linked to extrahepatic metastasis. Increased STIM2 levels inhibited CCA cell metastasis in vivo, confirming the inhibitory effect of STIM2 on the tumorigenesis and metastasis of CCA.

Our mechanistic investigation revealed that the STIM2-KRT8 axis is critical in CCA metastasis. KRT8 (keratin 8) is an epithelial marker and one of the most important keratin proteins (28). KRT8 and its filament partner, KRT18, have been found to control the cellular response to stress stimuli and contribute to cell resistance to apoptosis, in addition to maintaining cell mechanical integrity (29). Recent investigations showed that KRT8 levels were higher in the aqueous humor of patients with age-related macular degeneration (AMD) (30). In vitro studies have also indicated that high KRT8 levels could protect RPE cells from deterioration under oxidative stress (31). KRT8 has been demonstrated to be important in a variety of malignant tumors (32-34). However, the molecular mechanism of KRT8 in the pathogenesis of CCA remains unknown. In this study, the GEPIA database and qRT-PCR were used to determine the changed expression of the KRT8 mRNA levels controlled by STIM2. According to functional studies, STIM2 suppresses tumor metastasis in vivo by reducing KRT8 levels. In future research, we will explore the unique regulatory mechanism of m6A-induced STIM2 abnormalities in CCA.

A recent study revealed that m6A mRNA modification is critical for RNA fate, including splicing, localization, translation, transport, and mRNA stability (35). A multicomponent complex composed of METTL3, METTL14, and WTAP mediates adenosine N6 methylation in mammals (36). To date, FTO and ALKBH5 have been shown to influence m6A demethylation (37). m6A has been shown to interact with the targeted mRNA or miRNA and contribute to the progression of various types of cancer (9). In recent years, emerging studies have suggested that m6A-related genes have crucial roles in the initiation and progression of cancers, including acute myeloid leukaemia, renal cell carcinoma, and hepatocellular carcinoma etc. (24). Several m6A regulator abnormal expressed is closely related to the poor prognosis of tumor patients including CCA (38). Recently study identified PD-L1 mRNA as a target of ALKBH5 and reveals a role for ALKBH5 in regulating the tumor immune microenvironment and immunotherapy efficacy in CCA (39). However, research on the m6A alteration of STIM2 in CCA is lacking. In the present study, the SRAMP database and MeRIP-seq (sequencing) were used to determine the altered expression of STIM2 mRNA transcripts controlled by m6A alteration. In this study, we found that the m6A/IGF2BP2 axis functioned in enhancing CCA metastasis, suggesting that it might be a new target for CCA diagnosis and treatment. Finally, as previously stated, this was the first time that STIM2 was identified as a direct target of m6A-mediated transformation. The particular regulatory mechanism through which m6A induces STIM2 abnormality in CCA remains an open question for further research.

Conclusions

Finally, our findings have provided new insight into the critical involvement of the m6A-STIM2 axis in the progression of CCA metastasis. The current results highlighted the importance of the molecular process of m6A epi-transcriptomic alteration in cancer research, which was previously undiscovered.

The article’s supplementary files as