MicroRNA-302a enhances 5-fluorouracil sensitivity in HepG2 cells by increasing AKT/ULK1-dependent autophagy-mediated apoptosis

Introduction

Liver cancer (LC) is ranked as the third leading cause of cancer-related mortality, contributing to an estimated annual fatality count of approximately 830,000 individuals [1, 2]. Hepatectomy and adjuvant chemotherapy are the primary treatments for LC [3]. Unfortunately, the overall survival rate has failed to meet expectations because of the development of chemoresistance and the high rate of recurrence [4]. The molecular basis of chemoresistance is still unclear and continues to be a big challenge that significantly limits the efficacy of chemotherapy. Therefore, understanding the underlying molecular mechanisms associated with chemoresistance is essential to improve the clinical outcome of LC patients.

MicroRNAs (miRNAs) are involved in regulating the biological behaviors of tumor cells, such as proliferation, invasion, apoptosis, autophagy, and chemoresistance [5], [6], [7], [8], [9]. Therefore, identification of the miRNA machinery that regulates LC chemoresistance will not only facilitate the understanding of LC progression but also aid in the development of potential therapeutic strategies for LC chemoresistance [10, 11]. MicroRNA-302a (miR-302a), a member of the miR-302 family (miR-302/367 cluster), inhibits the progression of various cancers, including gastric cancer, LC, and breast cancer (BC) [12], [13], [14]. Additionally, miR-302a over-expression enhances the sensitivity of multiple tumor cells to chemotherapy agents [9, 15], [16], [17]. Although miR-302a plays a regulatory role in various tumors, its effect on 5-fluorouracil (5-FU) sensitivity in LC cells remains unclear.

Autophagy is also implicated in the modulation of chemoresistance [18], [19], [20], [21], [22], [23]. However, autophagy plays a dual role in the regulation of chemotherapy. On one hand, autophagy serves as a mechanism to enhance the sensitivity of cancer cells to chemotherapy, acting as a chemosensitizing agent. For instance, autophagic cell death can significantly increase the responsiveness of non-small cell lung cancer and cervical cancer cells toward chemotherapeutic agents [18, 19]. On the other hand, autophagy can have an antagonistic effect on chemotherapy-induced cell death and can reduce sensitivity toward chemotherapy [20], [21], [22], [23]. These differences may be related to cell types, cell states, autophagy threshold, etc. Nevertheless, regulation of autophagy may be a novel strategy to overcome chemoresistance. Autophagy is regulated by multiple genes, miRNAs, and kinases, among which UNC-51 like kinase 1 (ULK1) acts as a switch for autophagy [24]. ULK1 is the substrate of the mammalian target of rapamycin (mTOR) complex 1 (mTORC1), which is accommodated by the upstream protein kinase B (PKB/Akt)/mTOR signaling [25, 26]. Interestingly, miR-302a has been documented to suppress Akt signaling while inducing cancer cell death and enhancing sensitivity toward chemotherapeutic agents [16, 27]. In the study, we identified Akt as a potential target gene for miR-302a using target gene prediction software. These findings suggest an intrinsic relationship between miR-302a, Akt, autophagy, and 5-FU sensitivity in LC cells. Therefore, our study aimed to explore the correlation between miR-302a, Akt, autophagy, and 5-FU sensitivity in LC cells and to provide evidence on the regulatory role of miR-302a on 5-FU chemoresistance in LC cells via the Akt/ULK1-dependent autophagy and apoptosis.

Materials and methods

Results

MiR-302a levels are correlated with 5-FU-resistance in HCC

To examine the role of miR-302a in HCC, we initially used ISH to observe the expression of miR-302a in HCC tissues. ISH images showed that miR-302a was down-regulated in HCC tissues compared to the adjacent non-tumor tissues (Figure 1A and B, p<0.01), which was further confirmed by real-time PCR analysis (Figure 1C, p<0.01). Consistently, miR-302a levels were low in cultured LC cell lines (Figure 1D, p<0.01), especially in HepG2 cells, wherein it was only 0.4-fold of that in the cultured normal liver cells. Chemoresistance is one of the major contributors to poor prognosis in HCC patients [6]. Therefore, we collected HCC tissues that had been surgically resected before the receipt of FP chemotherapy. The HCC tissues were classified into chemosensitive and chemoresistant tissues, according to the response of patients to FP chemotherapy, and their miR-302a levels were detected using ISH and real-time PCR. The results revealed that compared to the chemosensitive HCC tissues, miR-302a levels in the chemoresistant HCC tissues were significantly decreased (Figure 1E–G, p<0.01). Similarly, we found that miR-302a levels were significantly decreased in the 5-FU-resistant HepG2/R cells, with approximately 0.5-fold expression level of that of the parent HepG2 cells (Figure 1H, p<0.01). These findings suggest that 5-FU chemoresistance is associated with the down-regulation of miR-302a expression in HCC cells.

Figure 1:

Decreased miR-302a levels correlate with 5-FU chemoresistance in HCC. (A–C) The levels of miR-302a in HCC tissues (n=10) and adjacent non-tumor tissues (n=10) were determined by ISH (A, B) and real-time PCR (C). ^**p<0.01 vs. non-tumor. (D) Real-time PCR analysis of miR-302a levels in THLE-3, primary hepatocytes, and 4 LC cell lines. ^**p<0.01 vs. THLE-3 cells, n=6. (E–G) ISH (E, F) and real-time PCR (G) analysis of miR-302a levels in chemosensitive (n=20) or chemoresistant (n=17) HCC tissues. ^**p<0.01 vs. chemoresistant. (H) Real-time PCR analysis of miR-302a levels in 5-FU-resistant HepG2/R cells and parent HepG2 cells. ^**p<0.01 vs. HepG2 cells, n=6.

MiR-302a inhibition decreases 5-FU sensitivity in HepG2 cells

To elucidate the effects of miR-302a on 5-FU chemosensitivity, we transfected HepG2 cells with miR-302a-m, miR-302a-i (Figure S2) in response to 5-FU treatment. As shown in Figure 2A and B, HepG2 cell viability was not significantly altered after miR-302a over-expression or inhibition under basal conditions (without 5-FU treatment). However, 5-FU decreased the viability of HepG2 cells, which was enhanced by miR-302a up-regulation but diminished by miR-302 inhibition (Figure 2A and B, p<0.01). These results were further confirmed by TUNEL staining, which revealed that the number of TUNEL⁺ HepG2 cells was significantly increased in the miR-302a-m+5-FU group but significantly decreased in the miR-302a-i+5-FU group, indicating an increase in 5-FU-induced apoptosis in the miR-302a-m-transfected cells (Figure 2C and D, p<0.01). Additionally, the over-expression of miR-302a augmented 5-FU susceptibility in other HCC cells (Figure S3). To elucidate the regulatory role of miR-302a in 5-FU-associated HepG2 cell death, we determined the level of caspase cleavage. The results revealed that 5-FU elevated the levels of caspase-3 and caspase-9 cleavage (Figure 3E–H, p<0.01). Moreover, the levels of caspase-3 and caspase-9 cleavage were not significantly altered by miR-302a up-regulation and down-regulation under basal conditions but were promoted by miR-302a over-expression and reduced by miR-302a inhibition in response to 5-FU (Figure 3E–H, p<0.01). These results indicate that miR-302a over-expression enhanced the 5-FU sensitivity in HepG2 cells.

Figure 2:

MiR-302a increases 5-FU sensitivity in HepG2 cells. (A, B) Statistical analysis on the effects of miR-302a-m (A) or miR-302a-i (B) transfection on the cell viability of 5-FU-treated (100 μM) HepG2 cells. ^**p<0.01 vs. control, ^##p<0.01 vs. 5-FU, n=6. (C, D) TUNEL examination and statistical analysis on the effects of miR-302a-m (C) or miR-302a-i (D) transfection on the apoptosis of 5-FU-treated (100 μM) HepG2 cells. ^**p<0.01 vs. 5-FU, n=6. (E–H) Western blot analysis and statistical analysis on the levels of caspase-3 (E, G) and caspase-9 (F, H) cleavage in 5-FU-treated (100 μM) HepG2 cells. ^**p<0.01 vs. control, ^##p<0.01 vs. 5-FU, n=6.

Figure 3:

MiR-302a increases 5-FU sensitivity in HepG2 via autophagy. (A, B) GFP-LC3 fluorescent probe detection (A) and statistical analysis (B) on the effects of 5-FU treatment on the formation of LC3-positive puncta in HepG2 cells. LC3-positive puncta are shown from at least 40 cells from five random fields. **p<0.01 vs. control (0 μM), n=6. (C) Cell viability analysis of 3-MA-pretreated (5 μM) and 5-FU-treated (100 μM) HepG2 cells, ^**p<0.01 vs. untreated control, ^##p<0.01 vs. 5-FU, n=6. (D, E) Levels of LC3B-I/II (D) and miR-302a (E) in 5-FU-treated (10, 50, or 100 μM) HepG2 and HepG2/R cells. ^*p<0.05, ^**p<0.01 vs. untreated HepG2 cells, ^#p<0.05 vs. untreated HepG2/R cells, n=6. (F) Correlation between miR-302a levels and LC3B-II expression. (G, H) GFP-LC3 fluorescent probe detection (G) and statistical analysis (H) on the effects of miR-302a-m or miR-302a-i on the formation of LC3-positive puncta in 5-FU-treated (100 μM) HepG2 cells. ^**p<0.01 vs. 5-FU, n=6. (I–K) Examination of LC3B-I/II expression (I), p62 expression (J), and cell viability (K) of miR-302a-m-transfected HepG2 cells pretreated with or without 3-MA for 30 min, and treated with 5-FU (100 μM). ^**p<0.01 vs. untreated HepG2 cells, ^##p<0.01 vs. 5-FU+NC-m, ^&&p<0.01 vs. 5-FU+miR-302a-m, n=6.

MiR-302a targets Akt to promote autophagy-associated HepG2 cell death

To determine how miR-302a mediates autophagy in HepG2 cells, we used the TargetScan software 8.0 (https://www.targetscan.org/vert_80/) to predict the molecular targets of miR-302a. The results revealed that miR-302a might directly target Akt, a key upstream regulator of ULK1 that is involved in autophagy initiation [24, 25, 28]. Moreover, the results revealed that the miR-302a-binding site in the Akt 3′UTR (7-bp fragment) was conserved in humans and mice (Figure 4A), indicating a good conservative character. Furthermore, the results of luciferase reporter analysis revealed the luciferase activity of Akt 3′UTR decreased significantly in the miR-302a-overexpressing HepG2 cells and decreased in the miR-302a-inhibited cells (p<0.01); however, the luciferase activity of the mutant Akt 3′UTR did not alter with miR-302a over-expression or down-regulation (Figure 4B). Consistently, Akt protein expression in HepG2 cells decreased with miR-302a up-regulation and increased with miR-302a down-regulation (Figure 4C and D, p<0.01). However, ULK1 expression showed an opposing trend in HepG2 cells (Figure 4C and D, p<0.01), indicating that Akt is a negative regulator of ULK1 [23, 24]. Considering that ULK1 has been reported to be involved in autophagy-dependent processes [23, 29, 30], we investigated whether miR-302a promotes cell death in HepG2 cells via ULK1-mediated autophagy. As shown in Figure 4E, compared with the 5-FU+NC-m group, miR-302a over-expression significantly enhanced the 5-FU-induced increase in LC3B-II levels by approximately 4.5-folds (p<0.01), which was attenuated by the ULK1 inhibitor SBI-0206965 (p<0.01). In addition, SBI-0206965 also reversed the miR-302a overexpression-induced reduction in p62 expression under 5-FU stimulation (Figure 4F, p<0.01). Moreover, SBI-0206965 significantly inhibited the enhanced activity of miR-302a on 5-FU-induced HepG2 cell death, as evidenced by the CCK-8 assay and TUNEL staining (Figure 4G–I, p<0.01). The effect of Akt1 inhibitor GSK69063 on cell viability was consistent with those of miR-302a over-expression (Figure 4G, p<0.01). Additionally, the sensitivity of miR-302a-over-expressing HepG2/R cells to 5-FU was found to be comparable to that of the parental HepG2 cells (Figure S1). Altogether, these results indicate that miR-302a enhanced 5-FU chemosensitivity by promoting Akt/ULK1-dependent autophagy in LC cells.

Figure 4:

MiR-302a negatively regulates Akt expression and promotes autophagic death. (A) The 3′UTR of Akt was predicted as the target of miR-302a. The complementary sequences are highlighted in red and green, while the mutated sequences are highlighted in blue. (B) HepG2 cells were co-transfected with wild-type or mutant Akt 3′UTR luciferase reporter and miR-302a-m or miR-302a-i and the luciferase reporter activity was assessed. ^**p<0.01 vs. control, n=6. (C, D) Western blot analysis and statistical analysis of Akt (C) and ULK1 (D) protein levels in HepG2 cells transfected with increasing concentrations of miR-302a-m (C) or miR-302a-i (D) for 48 h. ^*p<0.05, ^**p<0.01 vs. control, n=6. (E–G) Examination of LC3B-I/II levels (E), p62 levels (F), and cell viability (G) of miR-302a-m-transfected HepG2 cells pretreated with SBI-0206965 (10 μM) or GSK690693 (1 μM) for 30 min and treated with or without 5-FU for 24 h. ^**p<0.01 vs. 5-FU+NC-m, ^##p<0.01 vs. 5-FU+miR-302a-m, n=6. (H, I) TUNEL staining of miR-302a-m-transfected HepG2 cells pretreated with SBI-0206965 (10 μM) or GSK690693 (1 μM) for 30 min and treated with or without 5-FU for 24 h (H) and TUNEL⁺ cell quantitation (I). ^**p<0.01 vs. 5-FU, ^##p<0.01 vs. 5-FU+miR-302a-m, n=6.

Discussion

5-FU is the most commonly used anti-cancer chemotherapeutic agents for various solid tumors, including HCC [31]. Several patients with solid tumors are initially sensitive to 5-FU but ultimately acquire chemoresistance through continuous drug administration [32], which may be attributed to changes in gene expressions and/or signaling pathways by continuous 5-FU administration. In this study, we showed that miR-302a may be an important regulator of 5-FU resistance and Akt/ULK1-dependent autophagy in LC cells. MiR-302a levels were significantly decreased in chemoresistant HCC tissues as well as 5-FU-resistant LC cells. Additionally, overexpression of miR-302a augmented the sensitivity of HepG2 cells to 5-FU treatment, while inhibition of miR-302a diminished this sensitivity. Furthermore, inhibition of miR-302a enhanced 5-FU resistance by suppressing the Akt/ULK1-mediated autophagy and subsequently inhibiting HepG2 cell apoptosis. These findings suggest that targeting miR-302a/Akt/ULK1 axis is a potential strategy for treating 5-FU resistance in HCC patients. Increasing evidence suggests that miRNAs mediate cancer cell survival and are potential regulators of chemoresistance [33]. The roles of multiple microRNAs, including miR-375, miR-205-5p, and miR-141, have been well-documented in 5-FU chemoresistance in HCC [8, 34]. For example, miR-375 levels were found to be markedly higher in HCC tissues compared to the adjacent tissue, and its down-regulation was associated with the inhibition of 5-FU-induced cell apoptosis [8]. MiR-141 can serve as a potential 5-FU chemoresistance-reversal agent in HCC by inhibiting the antioxidant pathway [34, 35]. MiR-302a belongs to the miR-302/367 cluster, which overlaps with the La-related protein (LARP) genes, involved in RNA processing and folding [36]. This cluster has been well-characterized as a key regulator of cell proliferation and apoptosis [37]. With regards to tumor cells, the miR-302/367 cluster is suggested to be associated with increased chemosensitivity via the regulation of different target genes and signaling pathways [9, 15], [16], [17]. As for BC, miR-302a/b/c/d up-regulation targeted MAP/ERK kinase 1 (MEKK1) and BC resistance protein (BCRP) and increased chemosensitivity of adriamycin and mitoxantrone [15, 38]. Herein, we demonstrated that miR-302a expression was decreased in HepG2 cells and sensitized cells to 5-FU treatment, suggesting that miR-302a may function as a cancer suppressor gene in LC, consistent with some previous reports [9, 15], [16], [17, 38, 39].

Experimental and clinical investigations have shown that cell survival is critically associated with autophagic activity [18], [19], [20]; however, the mechanisms underlying this interplay are still poorly understood. Our results showed that autophagy was significantly induced by 5-FU in HepG2 cells but not in 5-FU-resistant cells. Previous studies found that the administration of 5-FU induces autophagic responses in various cancer cells, including LC cells, gastric cancer cells, and cervical cancer cells [19], [20], [21], [22, 39]. It is worth noting that the role of autophagy in mediating cancer cell fate is inconsistent, which likely functions as a pro-survival or pro-apoptotic regulator [40]. Autophagy-mediated cell death is the predominant mechanism in response to chemotherapy [18, 19, 39]. On the other hand, autophagy may enhance tumor adaptation to stress, resulting in resistance to cell death and chemotherapy [20], [21], [22], [23]. Therefore, maintaining a precise level of autophagic activity in specific cell types may represent a more efficacious strategy for mitigating the development of chemoresistance. Our data revealed that 5-FU-induced autophagy functions as a pro-apoptotic mechanism since the inhibition of autophagy inhibited the effects of 5-FU on HepG2 cell viability, consistent with the results of previous reports [18, 19, 39]. Moreover, the 5-FU-induced autophagy of LC cells was promoted by miR-302a over-expression and attenuated by miR-302a inhibition. These alterations provide a possible explanation for why miR-302a increases the chemosensitivity of 5-FU in LC cells.

We subsequently explored the underlying mechanisms associated with miR-302a-induced increase in LC cell autophagy using the TargetScan software, which revealed that miR-302a can directly target Akt, which is involved in autophagy initiation. Akt 3′UTR luciferase activity and protein expression were inhibited by miR-302a over-expression and increased by miR-302a inhibition. Growing evidence indicated that mTOR, a major downstream effector of Akt, exerts a critical role in regulating autophagy [23, 24, 41]. Inhibition of mTOR followed by increased ULK1 activity is critical for facilitating autophagy development [41]. Therefore, the down-regulation of Akt expression has been widely acknowledged as a reliable indicator for autophagy initiation. In this study, the miR-302a-induced inhibition of Akt was found to be accompanied by an up-regulation in ULK1 expression, providing further evidence for the involvement of miR-302a-mediated regulation of Akt levels in autophagy. Although the antineoplastic role of miR-302a may be associated with its inhibitory effects on Akt-dependent cell proliferation, our results demonstrated that ULK1 inhibitor could reverse the effects of miR-302a, suggesting that miR-302a increases chemosensitivity, at least partially, through the activation of Akt/ULK1-dependent autophagy.

Conclusions

We identified certain limitations in our study. For instance, it may be possible that miR-302a is involved in strengthening the 5-FU sensitivity of HCC cells via other targets or signaling pathways. Additionally, it remains unclear whether miR-302a also influences the sensitivity of other chemotherapy agents. Therefore, further research needs to be conducted on these aspects in the future. Our current findings strongly suggest that the inhibition of miR-302a in chemoresistant HCC tissues or cells may contribute to the development of chemoresistance. In contrast, miR-302a over-expression promoted 5-FU-induced autophagy and apoptosis of LC cells via the down-regulation of Akt. Therefore, the findings of our study unveiled the role of miR-302a in increasing the chemosensitivity of LC cells toward 5-FU treatment and provided a novel therapeutic strategy for LC treatment.