Tumour-derived exosomal lncRNA SNHG16 induces telocytes to promote metastasis of hepatocellular carcinoma via the miR-942-3p/MMP9 axis

Abstract

Background

Hepatocellular carcinoma (HCC) cell-derived exosomal LncRNA SNHG16 is highly expressed and associated with poor overall survival of patients. Telocytes (TCs), as novel interstitial cells, have been reported to promote HCC metastasis. Therefore, in our study, we investigated whether a molecular interaction occurred between exosomal LncSNHG16 and TCs in the tumor microenvironment.

Methods

LncSNHG16 expression in HCC tissues and cell lines was measured, and bioinformatics analysis was performed. Exosomes were isolated and purified from HCC cells with LncSNHG16 overexpression/knockdown vectors and cocultured with TCs. Then, markers of the LncSNHG16/miR-942-3p/MMP9 axis were tested in TCs. Transwell assays and cell wound healing assays were designed to examine the invasion and migration of HCC cells after coincubation with TCs. RNA immunoprecipitation (RIP) assays and dual-luciferase gene reporter assays were performed to verify the binding effect of LncSNHG16, miR-942-3p, and MMP9 mRNA. In vivo, experimental animal models were established to confirm the effect of exosomal LncSNHG16-induced MMP9 expression on HCC metastasis.

Results

Exosomal LncSNHG16 was phagocytized by TCs and downregulated miR-942-3p, which induced targeted MMP9 upregulation, and it had specific binding sites with miR-942-3p in TCs to facilitate the migration of HCC cells in vitro and in vivo. Exosomal LncSNHG16 was found to act as a competing endogenous RNA of the miR-942-3p/MMP9 axis in TCs.

Conclusion

Tumour-derived exosomal LncSNHG16 modulates MMP9 via competitively binding to miR-942-3p in TCs, thus promoting the metastasis of HCC.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13402-022-00746-w.

Introduction

Hepatocellular carcinoma (HCC) is the most common type of primary liver cancer, and it is the second leading cause of cancer-associated mortality and ranks third among all cancers in China [1]. Therefore, it has always attracted attention due to its high recurrence and metastasis to distant organs, as well as its poor overall survival (OS) for patients. Therefore, it is of pivotal significance to reveal the comprehensive and complicated molecular mechanisms of HCC metastasis.

Extracellular vesicles (EVs) are generic terms for secreted lipid bilayer encapsulated, cell-derived vesicles defined in 2011 by the International Society for EVs [2]. They are members of a family comprising microvesicles, apoptotic bodies and exosomes, which were first identified in 1981 as shedding vesicles with 5' nucleotidase activity [3]. Exosomes (30–150 nm in diameter) are characterized by their size and mechanism of generation, and they contain multiple proteins and RNAs, such as microRNAs (miRNAs), mRNAs, and noncoding RNAs [4]. Exosomes are thought to be novel contributors to cell-to-cell signalling delivery in the tissue microenvironment and circulatory system [5]. Intriguingly, tumour-derived exosomes in HCC were confirmed to participate in promoting the metastasis of cancer cells by activating surrounding cells in the tumour microenvironment (TME) [5]. Therefore, tumour-derived exosomes have become a critical component in the investigation of HCC development.

Long noncoding RNAs (lncRNAs) are classified into nonprotein-coding RNAs > 200 nucleotides in length with gene regulatory properties, which are involved in various biological processes, such as transcriptional regulation, mRNA stability, tumorigenesis and cancer development [6]. Accumulating evidence has reported that lncRNA ZFPM2-AS1 acts as a miRNA sponge and promotes cell invasion by regulating the miR-653/GOLM1 axis in HCC [7], and lncRNA CASC9 participates in HCC progression by acting as a competing endogenous (ceRNA) of miR-424-5p [8]. Moreover, cancer-derived lncRNAs influence surrounding cells or distant cells by being packaged into exosomes and secreted into the circulatory system [9–11]. To date, exosomal lncRNAs are considered common mediators that build signal bridges, linking cancer cells and other cells.

Telocytes (TCs) are a unique subset of interstitial cells and extensively exist in human organs [12]. They differ from fibroblasts in morphological, immunohistochemical, secretome, gene expression profiles and miRNAs [12]. Currently, the function of TCs in the development of cancers has received increasing attention. In our previous study, we identified that a high number of TCs was associated with a poor prognosis of HCC patients, while low expression of miR-942-3p in TCs promoted matrix metalloproteinase-9 (MMP9) secretion to accelerate the metastasis of HCC [13], but the detailed molecular mechanism that resulted in the low expression of miR-942-3p in TCs is still unknown. Whether HCC cell-derived exosomes impact TCs needs to be further elucidated. Combined with our previous high-throughput sequencing test in the blood serum exosomes of HCC patients, we speculate that exosomal lncRNA small nucleolar RNA host gene 16 (lncRNA SNHG16) may interact with TCs in the TME.

Therefore, in this study, we focused on the expression level of exosomal lnc-SNHG16 in HCC patients and analyzed its potential correlation with patient prognosis. Next, we verified whether LncSNHG16 promotes TCs to secrete MMP9 by acting as a miR-942-3p “ceRNA”. Both in vitro and in vivo experiments were designed to reveal this molecular mechanism.

Materials and methods

Clinical specimens and ethical approval

Human HCC and paracancer tissues were obtained from HCC patients who underwent surgery at Shandong Cancer Hospital and Institute between January 2020 and January 2022 (78 cases). The patients never received chemo- or radiotherapy before surgery. Before and after the operation, peripheral venous blood (10 ml) was collected to separate exosomes from the serum of HCC patients (10 cases). All pathological diagnoses were confirmed according to the guidelines of the American Joint Committee on Cancer. The study was approved by the Ethics Committee of Shandong Cancer Hospital and Institute (SHTHEC: 2,022,003,015), and informed consent was obtained from all patients and healthy volunteers.

Cell lines and co-culture

The human-derived HCCLM3 cell line, MHCC97 cell line, HL7702 cell line and 293 T cell line were purchased from Zhongqiaoxinzhou Company of China (www.zqxzbio.com) and cultured in DMEM or RPMI-1640 media at 37 °C in a humidified 5% CO₂ incubator. Primary TCs were separated from fresh paracancer tissues, incubated for 72 h, and then generated into an appropriate number of cells. Containers with dual chambers were used when TCs were incubated with the HCC cell lines, which allowed for TCs to be cultured in the upper space and cancer cells in the lower space, but the supernatant was shared between the two spaces.

Telocyte identification

The protocol of TC identification in the liver tissue can be found in our previous study (Suppl Fig. 1-A) [13].

Exosome isolation and identification

Serum and cancer cell supernatants were centrifuged at 2000 × g for 15 min and 10,000 × g for 30 min and then filtered through a 0.22 µm filter to remove debris. The pellet was ultracentrifuged at 1,200,000 × g for 70 min (BECKMAN: Optima XPN-100) and then resuspended in PBS (E607008, Sangon Biotech). CD81, HSP70 and TSG were used to identify exosomes by western blot assay, and the three dimensions of exosomes were observed by transmitted electron microscopy (TEM, HITACHI: HT7800, Shanghai, China). The volume of the exosomes was detected by nanometre particle analysis (Particle Metrix, ZetaView PMX110).

Exosome fluorescence tracking assay

After collecting exosomes from the HCC cell-line suspension, 5 µl of EvLINK505(CL12100220, TINGO, USA) dye was placed into 150 µl of the exosomes, lucifuge incubated at room temperature 30 min. Then, 15 µg/ml stained exosomes cocultured with TCs were incubated for 24 h in a 37℃ 5% CO₂ incubator. The cells were washed three times with PBS buffer, fixed with 4% paraformaldehyde for 30 min and then treated with 5 µl of CellLINK555 dye (1:100 dilution; EL012100200, TINGO, USA) for 30 min in the dark. All images were photographed by confocal laser scanning microscopy (3,869,000,207, LSM800 with Airyscan2, Zeiss, Germany).

Quantitative real-time PCR (qRT‒PCR)

Total RNA was extracted by TRIzol reagent (Invitrogen), and reverse transcription was performed with 2 µg of total RNA. The mRNA or miRNA expression levels were calculated by the Bio-Rad CFX96 system. Expression levels were normalized to GAPDH (for mRNA) or U6 (for miRNA), and data were assessed by 2^−ΔΔCt values. All primers were purchased from Jinweizhi Company of China (Suppl Table 7).

Overexpression and short hairpin RNAs

MHCC97 cells and TCs were instantaneously transfected with shRNA or scrambled shRNA negative control (sh-NC). All individual shRNAs were designed and synthesized by ABCAM company. All sequences of miR-942-3p and LncSNHG16 were constructed, combined with a pWSLV-05 carrier, and then cells were transfected according to the manufacturer’s instructions for overexpression. The cells were harvested for 48 h. MiR-942-3p or LncSNHG16 with pWSLV-05 was transfected into TCs or MHCC97 cells using Lipofectamine 2000 (Invitrogen; Suppl Fig. 1B, C, D).

Divergent lncRNA bioinformatics analysis

The distinct expression of lncRNAs in exosomes between HCC patients and healthy volunteers was detected by Novogen Company LDT of China. Comparing those different lncRNAs by the GenBank database, the functional enrichment analysis of LncRNAs was performed using the Genetic Ontology (GO) database. After screening miRNAs of interest, their targets were selected by StarBase and TargetScan databases, and JEFFERSON and Venn Diagram Web Tools were used to identify targets. The coexpression and survival analysis of LncSNHG16 and MMP9 in HCC were analyzed by StarBase v3.0.

Western blot

Cells or exosomes were dissolved in RIPA lysis buffer, and 30 µg of protein was run on SDS/PAGE gels and transferred to PVDF membranes. After blocking for 1 h and rinsing with TBST three times, the membranes were incubated with relevant primary antibodies at 4 °C overnight. Thereafter, the membranes were incubated with secondary antibodies and finally imaged by the ECL system (Thermo Fisher Scientific). All primary antibodies are listed in Suppl Table 1.

Immunohistochemistry (IHC)

Formalin-fixed paraffin-embedded primary tumour and paracancerous tissues were used for IHC analysis (Suppl Table 1). For heat-induced antigen retrieval, slides were soaked in citric acid buffer and maintained at a sub-boiling temperature for 8 min. Sections were observed using a light microscope (XSP-C204, CIC, China) and scanned using a laser scanning confocal microscope (Eclipse Ti-E, Nikon, Japan) at 40 × magnification. Datums were quantified on digital immunohistochemical slides using a Leica Aperio positive pixel count algorithm through whole slide analysis (PANNORAMIC DESK/MIDI/250/1000, 3DHISTECH, Hungary).

Cell invasion and migration assays

Transwell assays for cancer cell invasive detection were performed by using 8 μm pore size chambers in 24-well plates, and the upper chamber was filled with 100 μl of diluted Matrigel (1:20). Then, 5 × 10⁴ MHCC97 cells in 150 μl serum-free media were added to the upper chamber, while 1 × 10⁴ TCs were added to the lower chambers. After 48 h, the invaded cancer cells were fixed with methanol and stained with crystal violet or DAPI for 30 min. At least five random fields were calculated under a fluorescence microscope (Thermo. Com, USA). Each sample was run and repeated three times. The cell wound healing assay for cancer cell migration detection utilized different stimuli according to the experimental groups.

Dual-Luciferase gene reporter

The 358 bp wild-type or mutant 3′UTR of LncSNHG16 was inserted into the psiCHECK2-basic vector (Promega) and transduced into TCs with miR-942-3p mimics or MMP9 mutant type and 293 T cells using Lipofectamine 3000 (Invitrogen). Luciferase activities were measured 48 h after transfection by a Dual-Luciferase Assay (Promega).

RNA immunoprecipitation (RIP) assay

RIP was conducted using the Magna Nuclear RIP™ (Cross-Linked) Nuclear RNA-Binding Protein Immunoprecipitation Kit (Millipore) according to the manufacturer’s protocol. IgG antibodies (ab150077) were used for the RIP assay. Subsequent sequencing after RIP was performed, and Tophat (v1.4.0) was used to map the RIP-seq raw reads to the human reference genome (hg19/GRCh37). GAPDH was used as an internal reference [14].

In vivo model construction

To construct the subaxillary transplantation nude mouse model, 100 µl of PBS containing 1 × 10⁶ MHCC97-luciferase cells with distinct groups was injected into the right axilla of nude mice, and the mice were sacrificed on Day 18 according to the volume of tumors. Then, we randomly selected one mouse from the control group and resected the tumour under aseptic conditions. Then, a tumour of the same weight was placed into the liver tissues of nude mice to construct the orthotopic liver transplantation tumour model, and lung metastasis was observed (Suppl Fig. 2A). After successful establishment, the mice were evaluated by the IVIS Spectrum CT In Vivo Imaging System every 7 days. The GW4869 and MMP9 inhibitor (ab142180, ABCAM company, China) groups of mice were established by administering a 100 µl intraperitoneal injection every two days after establishing the orthotopic liver transplantation tumour models for 7 days. The mice were sacrificed by being exposed to a high level of carbon dioxide after 21 days, and their tissues were fixed with 4% paraformaldehyde. We assured that all animals received humane care.

Statistical analysis

SPSS software version 26.0 (Chicago, IL, USA) and Prism 8 (San Diego, CA, USA) were used to analyse all the data in the study, and each experiment was repeated at least 3 times. A p value < 0.05 was considered to represent significance. Student’s t test was utilized to compare microarray data and two groups of data, while the mean ± standard deviation was used for all numerical results. Kaplan‒Meier analysis and the log-rank test were used to evaluate the OS of patients, and Spearman’s rank correlation was calculated. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.000).

Results

Exosomal lnc-SNHG16 was upregulated in HCC patients and correlated with poor prognosis

Accumulating evidence has reported that cancer cell-derived exosomes, carrying distinct lncRNAs, are secreted into the blood circulation system and participate in the process of tumorigenesis and development [15–17]. Therefore, we selected five blood serum samples from healthy volunteers and thirty-five vein blood serum samples from HCC patients who had been diagnosed by tumour biopsies before surgery. From these HCC patients samples, we determined the differential expression of lncRNAs by high-throughput sequencing and found that exosomal LncSNHG16 was highly expressed in the blood serum of HCC patients (Fig. 1A, B) and contributed to multiple cellular processes, components and functions based on GO bioinformatics analysis (Fig. 1C). Subsequently, LncSNHG16 expression in HCC tissues was higher than that in normal liver tissues, and the overall survival (OS) of HCC patients with high LncSNHG16 expression indicated a poor prognosis according to the data from StarBase v2.0 (Fig. 1D, E, F; Suppl Table 2). To verify the differential expression of LncSNHG16 in HCC tissues and para-cancer tissues, we collected 38 HCC samples and measured LncSNHG16 expression by a qRT‒PCR assay, which showed that LncSNHG16 levels in HCC tissues were higher compared with para-cancer tissues (Fig. 1G; Suppl Tables 3 and 4). To verify whether high LncSNHG16 expression was associated with the prognosis of HCC patients, we analyzed 78 cases with low or high expression of LncSNHG16 by Pearson’s correlation and univariate statistical analysis, which suggested that high LncSNHG16 expression was positively correlated with HCC relapse within 2 years (p = 0.000; Fig. 1H).

Fig. 1.

Differential expression of blood serum exosomes of HCC patients and lncRNA bioinformatics enrichment analysis. (A) The volcano graph of the differential analysis of lncRNAs in the blood serum of HCC patients (n = 35) compared with healthy volunteers (n = 5). Exosomal LncSNHG16 was marked in the high expression group. (B) The top 10 highly expressed lncRNAs are listed with their p values. (C) The bioinformatics enrichment analysis from the GO database of distinct exosomal lncRNAs. (D) LncSNHG16 was expressed at a high level in HCC tissue. (E) Overall survival for LncSNHG16 in HCC patients from Starbase v2.0 by Kaplan‒Meier analysis and log-rank testing. (F) The expression of LncSNHG16 with distinct T stages. (G) LncSNHG16 expression in HCC and para-HCC tissues from 38 samples. (H) Univariate analysis and Pearson’s analysis were used to verify the correlation between lncRNA SNHG16 expression and relapse in HCC patients. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.000)

Exosomal lnc-SNHG16 could be phagocytized by telocytes

Although we confirmed that exosomal lnc-SNHG16 was upregulated in the blood circulation system of HCC patients, we also determined whether it was highly expressed in HCC cell lines. Then, we extracted exosomes from the supernatant of MHCC97, HCCLM3 and HL7702 cells and authenticated them by transmission electron microscopy (TEM, Fig. 2A), nanoparticle size testing (Fig. 2B, C) and western blot assays (Fig. 2D; Suppl Table 8). Compared with healthy donors, the blood serum exosomal LncSNHG16 level increased in the preoperative HCC patients and decreased after tumour resections (n = 10), indicating that serum LncSNHG16 was mainly produced by cancer cells (Fig. 2E). Consistent with the upregulation of LncSNHG16 in MHCC97 and HCCLM3 cells, LncSNHG16 levels were significantly higher than those in normal liver cells (Fig. 2F, G). Subsequently, to further investigate the existing pattern of extracellular LncSNHG16 in exosomes, we utilized RNase A and Triton X100 to measure its expression and found that the levels of LncSNHG16 in exosomes were unchanged upon RNase A treatment but significantly decreased upon treatment with RNase A and Triton X100 simultaneously (Fig. 2H). This result demonstrated that extracellular LncSNHG16 was mainly contained within exosomes instead of being directly released.

Fig. 2.

The qualification test of exosomes and phagocytosis of TCs. (A, B, C) Identification of exosomes by transmission electron microscopy and nanoparticle size testing. (D) Exosomes from MHCC97 and HCCLM3 cell lines with CD81, HSP70 and TSG markers by western blot assay. (E) The level of LncSNHG16 in the patient’s blood serum of pre-operation and post-operation(n = 30). (F) The level of LncSNHG16 in three different HCC cell lines. (G) The level of LncSNHG16 in exosomes from three different HCC cell lines. (H) Different expression levels of LncSNHG16 in exosomes of MHCC97 and HCCLM3 cells treated with RNase A and Triton X-100. (J) LncSNHG16 expression in TCs cocultured for 48 h with different groups. (I) TCs were stained with CellLINK555 dye (red), and HCC cell-derived exosomes were stained with EvLINK505 (green) to detect phagocytosis by TCs. The cell nucleus was stained with DAPI. Each measurement was repeated 3 times and evaluated as the mean ± standard deviation. The P value was analyzed by unpaired t test or one-way ANOVA. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.000)

In our previous study, TCs were considered a pivotal type of surrounding cell in HCC tissues and were involved in promoting the migration of cancer cells into distant organs [12]. To determine whether HCC cell-derived exosomes could be internalized by TCs, we produced fluorescence tracking exosomes (EvLINK505) from MHCC97 cells cocultured with TCs for 24 h, confirming that cancer-derived exosomal LncSNHG16 could be phagocytized by TCs (Fig. 2I). Then, we synchronously detected LncSNHG16 expression in TCs after coculture with MHCC97 cell exosomes (MHCC97-exos) and HCCLM3 cell exosomes (HCCLM3-exos) for 48 h. MHCC97-exos and HCCLM3-exos increased LncSNHG16 expression in TCs, and this process was blocked by GW4869 (exosomal inhibitor; Fig. 2J). In summary, an increased level of LncSNHG16 in TCs was derived from extrinsic exosomal LncSNHG16.

Exosomal LncSNHG16 induced telocytes to promote HCC metastasis

Given the intercellular transfer of exosomes into TCs, we found that MHCC97 cell-derived exosomes affected MMP expression in TCs, indicating that these exosomes increased the expression levels of MMP2 and MMP9, which could be abolished by GW4869, while the expression of MMP3, MMP11 and MMP14 showed no significant change (Fig. 3, B; Suppl Table 9). To elucidate the molecular mechanism by which MMP9 expression was upregulated in TCs, we designed LncSNHG16 overexpression vectors and knockdown vectors in MHCC97 cells. When TCs were incubated with LncSNHG16 vectors MHCC97-exosomes (MHCC97-exos), the MMP9 level was significantly higher than that with shSNHG16-MHCC97-exos or with GW4869 by western blot assay and qRT‒PCR assay (Fig. 3C, D). This phenomenon suggested that exosomal LncSNHG16 played a critical role in this effect. Furthermore, in our previous study, we elucidated that TCs promoted the metastasis of HCC cells by secreting MMP9. Combined with the aforementioned experimental results that exosomal LncSNHG16 increased the MMP9 expression level in TCs, we constructed a unique Transwell assay model to demonstrate the interaction between TCs and MHCC97 cells via extracellular exosomal medium (Fig. 3E). LncSNHG16-MHCC97 overexpression cells possessed powerful invasion abilities, while the invasive function of MHCC97 cells could be abrogated by shSNHG16 in cancer cells, along with the GW4869 group, simultaneously. In addition, a cell wound healing assay was conducted to examine the migration of MHCC97 cells. Exosomes from LncSNHG16 vectors or shSNHG16 MHCC97 cells were extracted and purified and then incubated with TCs for 48 h. Finally, these distinct suspensions were treated with MHCC97 cells. As shown in Fig. 3F, the migration capability of cancer cells was significantly stronger in TCs treated with exosomal LncSNHG16 than in TCs treated with shSNHG16 exosomes or GW4869, suggesting that the promoting function of TCs in HCC cells was due to the presence of exosomal LncSNHG16.

Fig. 3.

The influence of HCC cell-derived exosomal LncSNHG16 on TCs. (A, B) MMP expression in TCs cocultured with MHCC97 exosomes by western blot assay and qRT‒PCR assay. (C, D) MMP9 expression of TCs incubated with overexpressed/silenced LncSNHG16 of MHCC97 exosomes by western blot assay and qRT‒PCR assay. (E) The simulated diagram of the Transwell assay showing the effect of TCs on the invasion of MHCC97 cells when incubated with overexpressed/silenced exosomal LncSNHG16. (F) The flow diagram of the wound healing assay of MHCC97 cells with TCs that were treated with distinct types of exosomal LncSHNG16. Each measurement was repeated 3 times and evaluated as the mean ± standard deviation. The P value was analyzed by an unpaired t test. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.000)

LncSNHG16 as a “ceRNA” of miR-942-3p facilitated MMP9 expression in TCs

Since exosomal LncSNHG16 was predominantly internalized in TCs, it might function as a ceRNA to sequester miRNAs, resulting in the release of corresponding miRNA-targeted transcripts [18] To validate this hypothesis, we performed coexpression analysis of LncSNHG16 and MMP9 in HCC in the StarBase v2.0 database and found that the LncSNHG16 level was positively correlated with MMP9 expression (Fig. 4A). Accumulating evidence between miRNA-targeted MMP mRNAs and LncRNAs was integrated by Cytoscape 8.0 software, and these data revealed the LncSNHG16/miR-942-3p/MMP9 mRNA-associated axis (Fig. 4B; Suppl Tables 5 and 6). Next, we first transfected TCs with LncSNHG16 vectors or shSNHG16 to detect changes in miR-942-3p expression, and the results showed that LncSNHG16 overexpression efficiently downregulated miR-942-3p levels (Fig. 4C). Coincidently, MMP9 mRNA analysis showed that LncSNHG16 overexpression increased MMP9 expression level (Fig. 4D). Subsequent bioinformatics analysis in the TargetScan and the miRanda database revealed specific binding sites among LncSNHG16, miR-942-3p and MMP9 mRNA (Fig. 4E). Intriguingly, the Dual-Luciferase gene reporter assay confirmed the unique interactions between LncSNHG16 and miR-942–3, as well as miR-942-3p and MMP9 mRNA. Moreover, transfecting miR-942-3p mimics into TCs or 293 T cells significantly inhibited the luciferase activity of MMP9 mRNA, while the LncSNHG16 mutation type showed no response to miR-942-3p mimics, revealing that LncSNHG16 had positive binding sites with miR-942-3p and the latter had special binding sites with MMP9, which was consistent with our previous study. Next, a RIP assay was performed to further verify the endogenous binding of Lnc SNHG16 and miR-942-3p, and the results demonstrated that in the miR-942-3p mimic group, SNHG16 enrichment was strong and had statistical significance (p < 0.001; Fig. 4F). Subsequently, we determined whether LncSNHG16-mediated sequestration of miR-942-3p was responsible for the upregulation of MMP9 expression by western blot assay. The level of MMP9 increased in the LncSNHG16 vector group and was rescued by miR-942-3p silencing in TCs, whereas miR-942-3p mimics attenuated MMP9 expression directly (Fig. 4G). These results were further confirmed at both the MMP9 RNA and protein levels. Taken together, LncSNHG16 acted through miR-942-3p to promote MMP9 expression.

Fig. 4.

LncSNHG16 acts as a “ceRNA” for the miR-942-3p/MMP9 axis in TCs. (A) The coexpression analysis of LncSNHG16 and MMP9 in hepatocellular carcinoma from the ENCORI Pan-Cancer Analysis Platform. (B) The “ceRNA” network of MMPs was constructed by Cytoscape software. (C, D) miR-942-3p and MMP9 expression in TCs incubated with LncSNHG16 vectors or shSNHG16 by qRT‒PCR assay. (E) The predicted binding sites between LncSNHG16, miR-942-3p and MMP9 were screened by the JEFFERSON database and GeneCard internet. A luciferase gene reporter assay verified the binding efficiency between LncSNHG16, miR-942-3p and MMP9 by wild-type/mutated plasmids transfected into HEK-293 T cells. An empty plasmid vector with miR-942-3p was used as a parallel control. (F) RIP experiments combined with qRT‒PCR assays were performed to detect LncSNHG16 enrichment in 293 T cells and TCs. (G) The expression level of MMP9 in TCs after transfecting scrambled LncSNHG16 vectors, miR-942-3p mimics or sh-miRNA by western blot assay. Each measurement was repeated 3 times and evaluated as the mean ± standard deviation. The P value was analyzed by unpaired t test or one-way ANOVA. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.000)

Exosomal LncSNHG16-mediated TCs promoted HCC metastasis in vitro and in vivo

To evaluate the LncSNHG16/miR-942-3p/MMP9 axis-mediated interaction between HCC cells and TCs, we systematically analyzed the invasion and migration of HCC cells in vitro and in nude mouse models of MHCC97 cells in vivo. The Transwell assay results demonstrated that si-SNHG16 MHCC97 cells lost their powerful invasive ability when incubated with TCs compared to the control group, while MHCC97 cells incubated with si-miR-942-3p TCs exhibited a stronger invasion capability. Conversely, these effects were abolished by miR-942-3p mimics and MMP9 inhibitor in TCs, respectively (Fig. 5A). Additionally, results from the wound healing assay, which could evaluate the migrative capacity of MHCC97 cells, were similar to the Transwell assay results. Furthermore, we estimated the metastatic capability of HCC cells with orthotopic liver transplantation tumours in a nude mouse model as well as their lung metastasis tumours (Fig. 5B). Bioluminescence imaging indicated that the metastatic tumour number in the lung transferred from liver transplantation tumours with LncSNHG16-overexpressing MHCC97 cells was greater than that in the shSNHG16 group, MMP9 inhibitor group and GW4869 group. Moreover, the images showed the growth tendency of tumors in the liver tissues, suggesting that the larger the tumour is in the liver, the more metastatic tumors there are in the lung through statistical analysis (Fig. 5C, D). The haematoxylin–eosin (HE) staining in the lung metastatic tumour supported the preceding results (Fig. 6A). Moreover, the IHC assay results suggested that the SNHG16 vector group had high levels of MMP9 in both liver tumors and lung metastatic tumors (Fig. 6B, C).

Fig. 5.

Exosomal LncSNHG16 induced TCs to promote metastasis of HCC cells in vitro and in vivo. (A) Interactions between HCC cells and TCs with distinct treatments, respectively, by Transwell assay and wound healing assay. (B) The simulated diagram of orthotopic liver transplantation tumour model of nude mice. (C) An orthotopic liver transplantation tumour model with MHCC97-luciferase cells was constructed. Tumours in the liver and lung were tested by IVIS Spectrum after 1 day, 7 days and 21 days. (D) The gross specimens were resected from animal models. Each measurement was repeated 5 times and evaluated as the mean ± standard deviation. The P value was analyzed by unpaired t test or one-way ANOVA. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.000)

Fig. 6.

MMP9 expression in orthotopic liver tumours and lung metastatic tumours in vivo. (A) HE staining of lung metastatic tumours. (B, C) IHC staining of MMP9 in orthotopic liver tumours and lung metastatic tumours. Each measurement was repeated 5 times and evaluated as the mean ± standard deviation. The P value was analyzed by an unpaired t test. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.000)

Collectively, these results demonstrated that exosomal LncSNHG16 as an intercellular signal from cancer cells into TCs, induced TCs to promote the metastasis of HCC by downregulating miR-942-3p and unregulating MMP9, which revealed a novel and complex molecular mechanism in the TME (Fig. 7).

Fig. 7.

Diagram of the molecular mechanism involving exosomal LncSNHG16 and telocytes in the HCC-associated TME. Tumour-derived exosomal lnc-SNHG16 was phagocytized by TCs and competitively bound miR-942-3p, releasing the target gene of MMP9 mRNA. The increasing expression level of MMP9 in TCs promotes the metastasis of HCC cells

Discussion

Metastasis and recurrence of HCC are the major causes of cancer-related deaths. Hence, it is necessary to investigate the biological basis of invasion and migration in HCC development and explore the molecular mechanism of tumorigenesis between HCC cells and surrounding cells in the TME. In this study, we focused on the LncSNHG16 which was highly expressed in the blood serum exosomes of patients with HCC and functionally interacted with TCs. A series of biochemical reactions were activated after TCs phagocytized exosomal LncSNHG16, indicating that LncSNHG16 competitively bound with miR-942-3p and upregulated MMP9 expression. Thereby, the high expression of MMP9 in HCC tissue dramatically became as a trigger to enhance the metastatic behavior of HCC. Thus far, feedback interactions between cell-to-cell communications mediated by tumour-derived exosomes have been established.

Telocytes differentiate from cancer-associated fibroblasts and pericytes in morphology but possess analogical functions in contributing to the development and evolution of cancer cells [19]. Recently, Soha A reported that TCs facilitated angiogenesis by secreting VEGF to promote endothelial proliferation and migration and by releasing MMP9 to degrade the capillary basement membrane and prune vessels [20]. Furthermore, Lucio Diaz-Flores found that CD34 + /TCs play an important role in the tumour stroma formation of invasive ductal carcinoma in the breast [21]. In this study, we also verified a similar mechanism of TCs in HCC. However, the explicit functions of TCs in malignant tumours are still unclear, and more studies should be conducted in the future.

Exosomal transport is believed to be an effective means to modulate cell signalling and biological function in recipient cells [22–25]. Exosomes can not only carry various stimulants and proteinases but also transmit genetic nucleic acids, such as microRNAs, lncRNAs and circular RNAs [26, 27]. It was reported that lncRNAs packaged into exosomes accounted for 20.19% of exosomal RNA extracted from the plasma of prostate cancer patients [28]. In recent years, the modulatory role of exosomal lncRNAs in multiple types of cancer has gradually been discovered. For instance, tumour-derived exosomal lncRNA-SOX2OT was found to promote bone metastasis of non-small cell lung cancer by targeting the miRNA-194-5p/RAC1 signaling axis in osteoclasts [4]. The exosomal lncRNA PVT1/VEGFA axis was revealed to accelerate colon cancer metastasis and stemness by downregulating the tumour suppressor miR-152-3p [29]. Coincidently, Zheng H found that tumour-derived exosomal BCYRN1 activated the WNT5A/VEGF-C/VEGFR3 feedforward loop to drive lymphatic metastasis of bladder cancer [30]. These studies indicated that exosomal lncRNA was a stimulative factor in promoting cancer development and progression, which agreed with our theory. Moreover, it was widely believed that the molecular mechanism underlying exosomal lncRNA function was that of “ceRNAs” to competitively bind with a target miRNAs, resulting in the expression of related functional proteins. Therefore, the enormous proportion of exosomal lncRNAs in the serum of cancer patients should be given sufficient attention when investigating the molecular mechanisms of tumorigenesis and the development of cancers [18, 31].

Consistent with the discoveries in the literature [32, 33], LncSNHG16 exerted an oncogenic role in osteosarcoma by acting as a “ceRNA” of miR-1285-3p [32]; miR-4518 downregulation and upregulation of the target gene PRMT5 (protein arginine methyltransferase 5) conferred by LncSNHG16 induced the proliferation of glioma cells [33], and LncSNHG16 modulated the activity of the PI3K/Akt signalling pathway [34]. These studies demonstrated that LncSNHG16 directly regulated the behaviours of cancer cells and influenced the prognosis of patients LncSNHG16 also induced cancer-related surrounding cells in the TME in changing quantities of functional proteins. Interestingly, accumulating evidence provides support to confirm that exosomal lncRNAs may become a critical therapeutic target.

Conclusion

Collectively, this study revealed that the high expression of HCC cell-derived exosomal LncSNHG16 was a negative factor for the prognosis of HCC patients, and it facilitated TCs to secrete MMP9 by competitively binding with miR-942-3p, which promoted the metastasis of HCC. Unfortunately, recent technologies used in experimental nude mouse models in vivo did not satisfy the requirement of knocking down MMP9 nucleic acids only in TCs; therefore, we utilized an MMP9 inhibitor to simulate the process of abolishing MMP9 expression. In reality, exosomal LncSNHG16 is not the only cause of poor HCC evolution, and there exists a largely interactive network between HCC cells and surrounding cells composed of cytokines, chemokines, ions and extracellular vesicles. Therefore, more complex genetic mechanisms and sufficient clinical proof should be supported in future research.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

HCC

Hepatocellular carcinoma

TEM

Tumor micro-environment

miRNAs

MicroRNAs

qPCR

Quantitative polymerase chain reaction

OS

Overall survival

LncSNHG16

LncRNA SNHG16

TCs

Telocytes

RIP

RNA immunoprecipitation

ceRNA

Competing endogenous RNA

EVs

Extracellular vesicles

LncRNAs

Long non-coding RNAs

MMP9

Matrix metalloproteinase-9

TEM

Transmitted electron microscope

sh-NC

ShRNA negative control

PRMT5

Protein arginine methyltransferase 5

RIP

RNA immunoprecipitation

GO

Genetic Ontology

IHC

Immunohistochemistry

HE staining

Hematoxylin-eosin staining

Author contributions

Ying Xu supervised the study, supported the fund, wrote the manuscript, designed experiments and constructed figures. Guangchao Luan analyzed some partial data and figures. Ziming Liu was in charge of in vivo experiments. Zhongchao Li, Guangyang Qin and Yifu Chu collected the clinical samples and settled basic data of patients.

Funding

Medical and health science and Technology Development of Shandong (202104080599). Natural Science Foundation of Shandong Province (ZR2022QH066).

Data availability

The datasets used and/or analyzed during the present study are available.

Declarations

Ethical approval and consent to participate

The study was approved by the Ethics Committee of Shandong Cancer Hospital and Institute(SHTHEC: 2022003015), and informed consent was gained from all patients and healthy volunteers.

Consent for publication

All authors consent for publication.

Competing interests

There is no conflict of interest in our article.