CAF-derived miR-642a-3p supports migration, invasion, and EMT of hepatocellular carcinoma cells by targeting SERPINE1

Cell culture

Huh7 cells (a human hepatoma cell line) and human hepatocellular CAFs were obtained from Jiangsu KeyGEN Biotechnology Co., Ltd. (KeyGEN BioTECH, Nanjing, China) and Shanghai Fusheng Industrial Co., Ltd. (Shanghai, China), respectively. The cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) (KeyGEN BioTECH) supplemented with 10% fetal bovine serum (FBS) (Gibco, USA) and 1% penicillin-streptomycin (P/S) (KeyGEN BioTECH) in a humidified incubator at 37 °C with 5% CO₂.

Huh7 cells and CAFs were co-cultured in a transwell system. CAFs were seeded in the upper chamber of the transwell, while Huh7 cells were seeded in the lower chamber. A transwell without CAFs served as the control. Co-culture was maintained for 48 h.

MiRNAs bioinformatics prediction

The miRNAs binding to SERPINE1 3′ untranslated region (UTR) were predicted using the starBase (https://rnasysu.com/encori/), miRwalk (http://mirwalk.umm.uni-heidelberg.de/), TargetScan (https://www.targetscan.org/vert_80/), and miRDB (https://mirdb.org/) databases. The predicted miRNAs were intersected by Venn diagram.

Cell transfection

Huh7 cells were transfected with miR-642a-3p mimics, miR-642a-3p inhibitor, or SERPINE1 siRNA (si-SERPINE1) (General Biology, Anhui, China) for 48 h using Lipofectamine™ 3000 transfection reagent (Invitrogen, USA) according to the manufacturer’s protocol. Briefly, 125 µL Opti-MEM medium containing 100 pmol SERPINE1 siRNA (20 µM) or miR-642a-3p mimics/inhibitor (20 µM) and 5 µL P3000™ reagent was gently mixed. Similarly, 125 µL Opti-MEM medium and 3.75 µL Lipofectamine™ 3000 reagent were combined. The diluted Lipofectamine™ 3000 reagent was then added to the diluted siRNA mixture, gently mixed, and incubated at room temperature for 10-15 min. Once the cell confluence reached 70–80% in the 6-well plate, 250 µL of the transfection mixture was added, and the cells were cultured at 37 °C for 48 h. Negative control groups included cells transfected with miR-642a-3p mimics negative control (mimics NC), miR-642a-3p negative inhibitor (inhibitor NC), or siRNA negative control (si-NC). Each experimental group was performed in triplicate. The sequences of the miR-381-3p mimics, inhibitor, SERPINE1 siRNA, and respective negative controls are provided in Table 1.

CCK-8 assay

Huh7 cell proliferation was quantified using the CCK-8 Cell Proliferation Detection Kit (KeyGEN BioTECH). Briefly, Huh7 cells transfected with either si-NC or si-SERPINE1 were cultured in 96-well plates for 48 h. Subsequently, 10 µL of CCK-8 reagent was added to each well, followed by incubation at 37 °C for 2 h. Absorbance values were measured at 450 nm using an ELx800 Microplate Reader (BioTek, Winooski, VT, USA).

Wound healing assay

Huh7 cells were seeded into six-well cell culture plates at a density of 1 × 10⁵ cells/mL and incubated overnight. A sterile pipette tip was employed to create a scratch wound in each well. Unattached cells were removed by washing with 1× PBS, followed by the replacement of the culture medium with fresh medium. The cells were then concurrently transfected. After a 48-hour incubation period, images were captured at 100× magnification using an IX51 microscope (Olympus, Tokyo, Japan). Wound width was measured at 0 h (a) and 48 h (b), and the wound healing ratio [(a − b)/a × 100%] was calculated to assess migratory capacity.

Transwell assay

To assess Huh7 cell invasion, a 24-well transwell chamber (Corning Incorporated, USA) coated with Matrigel (BD, Franklin Lakes, NJ, USA) was employed. The cells were seeded at a density of 1 × 10⁵ cells/mL within the transwell chamber. The lower chamber was filled with 500 µL of DMEM supplemented with 10% FBS. Following a 48-hour incubation period, the cells on the upper surface of the membrane were removed using cotton swabs. The cells on the lower surface of the membrane were stained with 0.1% crystal violet (Sigma) for 30 min at 37 °C, washed twice with 1× PBS, imaged using an IX51 microscope (Olympus, Japan), and quantified.

Total RNA extraction and RT-qPCR

Total RNA was isolated using TRIzol Reagent (Invitrogen, USA) according to the manufacturer’s protocol. RNA integrity was assessed via agarose gel electrophoresis, and concentration and purity were determined using a Nano100 spectrophotometer (Hangzhou Allsheng Instruments Co., Ltd., Hangzhou, China). First-strand cDNA synthesis was performed using the PrimeScript™ RT reagent Kit (Takara, Shiga, Japan) with total RNA as a template. For miRNA expression analysis, reverse transcription was conducted with a Bulge-Loop™ miRNA RT-PCR Starter Kit (RiboBio). GAPDH or U6 served as the endogenous control gene. Quantitative PCR was performed using SYBR Green PCR Mix (Takara) on a StepOnePlus Real-Time PCR System (ABI, USA). Relative gene mRNA and miRNA expression levels were calculated using the 2^−ΔΔCt method based on three biological replicates with three technical replicates each. The primers sequences for miR-642a-3p, miR-3135a, miR-449b-5p, miR-642a-3p, miR-3135a, SERPINE1, GAPDH, and U6 were synthesized by General Biosystems (Anhui) Co., Ltd. (Anhui, China) and are listed in Table 2.

Western blotting

Total protein extraction and quantification were performed using a total protein extraction kit (KeyGEN BioTECH) and a BCA protein content detection kit (KeyGEN BioTECH), respectively, according to the manufacturer’s protocols. As described by previous study (Chehade et al., 2021), the protein samples were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred to polyvinylidene fluoride (PVDF) membranes. Immunoblotting was conducted using primary antibodies against SERPINE1 (ab222754, Abcam), E-cadherin (ab76055, Abcam), N-cadherin (66219-1-Ig, Proteintech), vimentin (bsm-33170m, Bioss), and GAPDH (ab9485, Abcam) at dilutions of 1:1000, 1:1000, 1:2000, 1:1000, and 1:2000, respectively. A secondary anti-rabbit IgG H&L antibody (ab6721, Abcam) was used at a dilution of 1:5000. Protein bands were visualized using an enhanced chemiluminescence (ECL) detection kit (KeyGEN BioTECH) and a ChemiDoc Touch 1708370 imaging system (Bio-Rad, USA). Densitometric analysis was performed using ImageJ software.

Dual-luciferase reporter assay

The 293T cells were seeded into 12-well plates. Upon reaching approximately 50% confluency, pmirGLO-SERPINE1-WT or pmirGLO-SERPINE1-MUT recombinant plasmid was co-transfected with miR-642a-3p mimics or negative control into 293T cells using Lipofectamine™ 3000 transfection reagent (Invitrogen, USA). After a 48-hour transfection period, 50 µL of cell lysate was transferred to each well of a 96-well black plate. Subsequently, Dual-Glo® Luciferase Reagent (Promega, USA) was added, and firefly luciferase luminescence was quantified using a Tecan Spark microplate reader (Tecan, Männedorf, Switzerland). Finally, 100 µL of Dual-Glo® Stop & Glo® Reagent (Promega, Madison, WI, USA) was added to each well to measure Renilla luciferase luminescence. Relative luciferase activity was normalized to Renilla luciferase activity.

Animal model

Four- to six-week-old male BALB/c nude mice obtained from Shanghai Lingchang Biotechnology Co., Ltd. were housed in single cages. The study protocol was approved by the Institutional Animal Care and Use Committee of Nanjing Ramda Pharmaceutical Co., Ltd. (IACUC-20230505) and carried out in accordance with the guidelines of the Animal Care Committee. Mice were acclimatized for one week under standard specific pathogen-free (SPF) conditions with a temperature of 20–26 °C, relative humidity of 40–70%, and a 12-hour light/12-hour dark cycle. Ten mice were randomly assigned to two groups: NC and shmiR-642a-3p (n = 5). Under abdominal anesthesia, nude mice were positioned supine, and the surgical site was disinfected. A midline abdominal incision was made to expose the liver, and the liver lobe exterior to the incision was gently removed with a cotton swab. Tumor cells were injected into the liver parenchyma approximately three mm deep using a needle, delivering 100 µL of Huh7-luc cells (Shanghai Zhong Qiao Xin Zhou Biotechnology Co., Ltd., China) at a density of 2 × 10⁸ cells/mL. The liver was repositioned, and the abdomen was closed layer by layer. One week post-inoculation, mice in the NC and shmiR-642a-3p groups received intraperitoneal injections of 2 × 10¹¹ vg of AAV-vector or AAV-shmiR-642a-3p, respectively. The animals (n = 5 per group) were monitored twice weekly for behavioral changes, food consumption, and weight. At eight weeks, one mouse from each group was anesthetized with carbon dioxide and subjected to liver tumor nodule examination (Fig. S1). The remaining mice were euthanized with carbon dioxide, and their livers were excised and photographed. Liver tissues were divided: one half was fixed in 4% paraformaldehyde, while the other half was snap-frozen in liquid nitrogen and stored at −80 °C.

Hematoxylin-eosin (H & E) staining

Liver tissues were fixed in a 4% paraformaldehyde solution and subsequently embedded in paraffin. Tissue sections with a thickness of 4 µm were prepared and stained with H&E. Histomorphological analysis of each liver was conducted using a SLIDEVIEW VS200 research slide scanner (Olympus).

Immunohistochemistry (IHC) assay

Livers were immediately immersed in 4% paraformaldehyde solution and fixed overnight. Tissue blocks were embedded in paraffin and sectioned to a thickness of 4 µm. Immunohistochemical staining for Ki67 expression in liver tissue was performed using the EnVision two-step method. Rabbit anti-Ki67 (ab16667; Abcam) at a dilution of 1:50 served as the primary antibody. Subsequently, sections were incubated with a specified HRP-conjugated secondary antibody (MXB Biotechnologies, Fuzhou, China). Diaminobenzidine (DAB) solution (MXB Biotechnologies) and hematoxylin (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) were used for color development. Ki67 expression in liver tissue was visualized using a SLIDEVIEW VS200 research slide scanner (Olympus).

Statistical analysis

Data were analyzed using SPSS version 21.0 and presented as mean ±  standard error (SE) with at least three times. Independent-sample t-tests were employed to compare differences between two groups, while one-way analysis of variance (ANOVA) was used to assess differences among multiple groups. Post hoc comparisons were conducted using the Tukey method. All experiments were repeated at least three times. A P-value less than 0.05 was considered statistically significant.

CAF-induced SERPINE1 underexpression promoted proliferation and invasion of Huh7 cells

CAFs are pivotal contributors to tumor progression, exhibiting a diverse range of functions encompassing collagen deposition and immunosuppression. Our prior investigation revealed that CAFs stimulated the proliferation and migration of Huh7 cells. To elucidate whether CAFs regulate Huh7 cells via SERPINE1, we initially quantified SERPINE1 mRNA expression in Huh7 cells co-cultured with CAFs. The results indicated a significant decrease in SERPINE1 mRNA expression within the co-culture group relative to the control group (Fig. 1A). Subsequently, SERPINE1 gene knockdown markedly enhanced Huh7 cell proliferation and invasion (P < 0.001) (Figs. 1B–1D), suggesting that CAFs may promote Huh7 cell proliferation and invasion through the suppression of SERPINE1 expression.

CAFs inhibited SERPINE1 expression in Huh7 cells by secreting miR-642a-3p

MicroRNAs function as intercellular messengers, transmitting information between cells, tissues, and organs. Within the TME, miRNAs contribute to tumor initiation and progression by regulating aberrant gene expression. To investigate whether CAFs influence SERPINE1 expression in Huh7 cells through miRNA secretion, we identified 18 potential miRNAs targeting the SERPINE1 3′ UTR through multiple miRNA online databases (Fig. 2A). Subsequent screening revealed five miRNAs of interest, among which miR-642a-3p, miR-3135a, and miR-449b-5p exhibited significantly elevated expression in the co-culture group (P < 0.05) (Figs. 2B–2F). Moreover, miR-642a-3p and miR-3135a expression was markedly increased in Huh7 cells with SERPINE1 knockdown (P < 0.01), with miR-642a-3p demonstrating a more pronounced difference between groups (Fig. 2G). These findings collectively suggest that CAFs suppress SERPINE1 expression in Huh7 cells by secreting miR-642a-3p.

CAFs-derived miR-642a-3p promoted migration, invasion, and EMT of Huh7 cells by inhibiting SERPINE1

To investigate whether CAF-derived miR-642a-3p promotes Huh7 cell migration, invasion, and EMT by regulating SERPINE1, miR-642a-3p mimics or inhibitors were transfected into Huh7 cells. Results demonstrated that miR-642a-3p mimics significantly upregulated miR-642a-3p expression while downregulating SERPINE1 mRNA expression (P < 0.001), whereas the miR-642a-3p inhibitor exhibited the opposite effect (Figs. 3A and 3B). Moreover, miR-642a-3p mimics significantly enhanced Huh7 cell migration and invasion, while the miR-642a-3p inhibitor significantly suppressed these processes (Figs. 3C and 3D). Notably, miR-642a-3p mimics markedly increased N-cadherin and vimentin protein expression and decreased E-cadherin protein expression in Huh7 cells (P < 0.001) (Fig. 3E). Conversely, the miR-642a-3p inhibitor significantly upregulated E-cadherin protein expression and downregulated N-cadherin and vimentin protein expression (P < 0.001) (Fig. 3E), suggesting that CAF-derived miR-642a-3p promotes EMT in Huh7 cells.

To investigate the binding interaction between miR-642a-3p and the 3′ UTR of SERPINE1, recombinant plasmids pmirGLO-SERPINE1-WT and -MUT were constructed based on predicted binding sites from a miRNA database. A dual-luciferase reporter assay demonstrated a significant decrease in fluorescence activity in the WT group upon miR-642a-3p mimic overexpression (P < 0.05), whereas no effect was observed in the MUT group (P > 0.05) (Fig. 4A). Rescue experiments revealed that SERPINE1 knockdown attenuated the inhibitory effects of miR-642a-3p knockdown on cell migration, invasion, and EMT in Huh7 cells (Figs. 4B–4D), suggesting that miR-642a-3p promotes migration, invasion, and EMT in Huh7 cells by targeting SERPINE1.

miR-642a-3p knockdown inhibited tumor growth and EMT in vivo

To assess the in vivo impact of miR-642a-3p on tumor infiltration, an orthotopic tumor model was established by injecting Huh7 cells into the livers of nude mice. As depicted in Fig. 5A, tumors were observed in the NC group but absent in the shmiR-642a-3p group. While no significant weight difference was observed between groups, a slight increase in weight was noted for the shmiR-642a-3p group (Fig. 5B). Histopathological examination of the NC group revealed visible tumor lesions invading the hepatic parenchyma with a larger invasion area compared to the reduced invasion area observed in the shmiR-642a-3p group (Figs. 5C and 5D). Moreover, miR-642a-3p knockdown significantly suppressed miR-642a-3p expression while enhancing SERPINE1 expression (Fig. 5E). Notably, miR-642a-3p knockdown upregulated SERPINE1 and E-cadherin protein levels while downregulating N-cadherin and vimentin protein levels (Fig. 5F), indicating an inhibitory effect on tumor infiltration and dissemination.