hsa_circ_0004018 suppresses the progression of liver fibrosis through regulating the hsa-miR-660-3p/TEP1 axis

The low expression of hsa_circ_0004018 especially in HSCs was associated with the poor progression of liver fibrosis in mouse model

To examine the expression of hsa_circ_0004018 in the progression of liver fibrosis, we obtained the liver samples from the mice injected with CCl₄ for 0, 30 and 45 days and detected the relative RNA levels of hsa_circ_0004018 by real-time PCR. We found that the severity of liver fibrosis and the serum level of IL-6 both increased over time upon CCl₄ injection (Figure 1A–1C), while the relative expression of hsa_circ_0004018 was downregulated (Figure 1D). And we further examined the expression of hsa_circ_0004018 in IL-6 treated HSCs, and found that the relative RNA levels of hsa_circ_0004018 were significantly downregulated as the HSCs were treated with IL-6 at 0, 20 and 40 ng/L for 24 hours (p<0.01, Figure 1E)

It has been reported that salvianolic acid B (Sal B) can inhibit the progression of liver fibrosis through regulating the activation of HSCs [21–23]. We further examined the expression of hsa_circ_0004018 in Sal B treated HSCs, and found that the relative RNA levels of hsa_circ_0004018 were significantly upregulated over time as the HSCs were treated with 100 ug/L of Sal B for 0, 3 and 6 days (p<0.01, Figure 1F). In addition, the relative RNA levels of hsa_circ_0004018 significantly enhanced with the increase of Sal B concentration (p<0.01, Figure 1G), suggesting that the low expression of hsa_circ_0004018, especially in HSCs, was associated with the poor progression of liver fibrosis.

Overexpression of hsa_circ_0004018 suppressed the proliferation and activation of HSCs in vitro

To further study the role of hsa_circ_0004018 in HSCs, we overexpressed hsa_circ_0004018 in the primary HSCs (Figure 2A) and examined the proliferation of HSCs through the EDU incorporation experiment (Figure 2B). We found that the EDU incorporation levels of hsa_circ_0004018 overexpressed HSCs were lower than the vector control (Figure 2B, 2C). We further analyzed the effect of hsa_circ_0004018 overexpression on cell cycle, and found that the upregulation of hsa_circ_0004018 resulted in higher G0/1 phase ratio compared with the vector control (p<0.01, Figure 2D–2F), indicating that overexpression of hsa_circ_0004018 could suppress the proliferation of HSCs through promoting the cell cycle arrest.

In addition, we examined the expression of α-SMA and COL1A1, which reflect the activation of HSCs, in the hsa_circ_0004018 overexpressed and the control HSCs. We found that both RNA levels and protein levels of α-SMA and COL1A1 were downregulated upon overexpression of hsa_circ_0004018 comparing with the control cells (Figure 2G–2K), suggesting that the overexpression of hsa_circ_0004018 inhibited the activation of the primary HSCs in vitro.

The hsa_circ_0004018 acts as a sponge of hsa-miR-660-3p

The role of circular RNA as a sponge in regulation of microRNA has been well known [16]. In order to find out the potential target of hsa_circ_0004018, we used the Arraystar’s home-made microRNA target prediction software to scan globally the matching microRNA. We found that hsa_circ_0004018 shares microRNA response elements (MREs) of hsa-miR-660-3p (Figure 3E), which may interact with hsa_circ_0004018 in HSCs.

To confirm this hypothesis, we designed the biotin labeled hsa_circ_0004018 probe to detect the interaction in the primary HSCs. As expected, the hsa_circ_0004018 probe enriched more hsa_circ_0004018 (p<0.05, Figure 3A), as well as the hsa-miR-660-3p (p<0.01, Figure 3B), than the negative control probe. Consistently, the biotin labeled hsa-miR-660-3p enriched more hsa_circ_0004018 than the negative control probe (Figure 3C, 3D). To further explore the function of this interaction, the luciferase plasmids containing the wild type 3’ terminal (WT) or the MRE deleting 3’ terminal (MT) of hsa_circ_0004018 were constructed (Figure 3E) and simultaneously co-transfected with hsa-miR-660-3p mimic, negative control or inhibitor into the HSCs. After that, the luciferase activity of each combination was detected. It showed that the hsa-miR-660-3p mimic significantly suppressed the luciferase activity of WT (p<0.05, Figure 3F, Column 2) but not MT (Figure 3F, Column 5). Conversely, the hsa-miR-660-3p inhibitor significantly increased the luciferase activity of WT (p<0.01, Figure 3F, Column 3) but not MT (Figure 3F, Column 6), indicating that hsa_circ_0004018 functions as a sponge of hsa-miR-660-3p in HSCs.

The hsa-miR-660-3p can promote the proliferation and activation of HSCs in vitro, and counteract with the function of hsa_circ_0004018

To explore the role of hsa-miR-660-3p in the HSCs and the progression of liver fibrosis, we firstly examined the expression of hsa-miR-660-3p in the primary HSCs isolated from the CCl₄-induced liver fibrosis mice. It showed that the expression levels of hsa-miR-660-3p significantly increased over time upon CCl₄ injection (p<0.01 at Day 60 and p<0.001 at Day 90, Figure 4A). The transfection of hsa-miR-660-3p mimic significantly promoted the proliferation of HSCs (p<0.01, Figure 4B, 4D), while the transfection of inhibitor showed the inhibitory effect (p<0.01, Figure 4C, 4E). In addition, the expression of α-SMA and COL1A1 significantly increased after transfection with mimic in HSCs (Figure 4F), whereas decreased after transfection with inhibitor (Figure 4G).

We then co-transfected the mimic with hsa_circ_ 0004018 in the HSCs and examined the proliferation and activation effect. As control, the negative control or the vector was transfected in combination as indicated (Figure 4H, 4I). The HSCs co-transfected with mimic and the vector control exhibited the fastest proliferation rate (black line, Figure 4H). In compare, the proliferation of HSCs co-transfected with mimic and hsa_circ_ 0004018 was significantly slower than the mimic/vector combination (red line vs. black line, Figure 4H), but faster than the negative control/ hsa_circ_0004018 combination (red line vs. yellow line, Figure 4H). We also examine the activation of the co-transfection HSCs, and found that the expression levels of α-SMA and COL1A1 were significantly lower in the mimic/ hsa_circ_0004018 combination than the mimic/vector combination (Lane 1 vs. Lane 4, Figure 4I), but higher than the negative control/hsa_circ_0004018 combination (Lane 1 vs. Lane 2, Figure 4I), indicating that hsa_ circ_0004018 counteract with the function of hsa-miR-660-3p.

TEP1 was validated as a bona fide target of hsa-miR-660-3p

To find out the downstream target of hsa-miR- 660-3p, we used the miRBase for screening (http://www.mirbase.org/). As standard, the potential gene target should harbor a similar MRE with the seed sequence of hsa-miR-660-3p and hsa_circ_0004018. Accordingly, we finally focused on TEP1. To confirm this, we transfected the HSCs with hsa-miR-660-3p mimic or inhibitor, and examined the expression of TEP1. It showed that the transfection of mimic resulted in an obvious decrease in both RNA and protein levels of TEP1 (Figure 5A, 5C). In contrast, the transfection of inhibitor resulted in an obvious increase in both RNA and protein levels of TEP1 (Figure 5B, 5D).

In further, we constructed the luciferase plasmids containing the wild type 3’ UTR (WT-3’UTR) or the MRE deleting 3’ UTR mutant (MT-3’UTR) of TEP1 (Figure 5E). Either of these two plasmids was co-transfected with hsa-miR-660-3p mimic, negative control or inhibitor in the HSCs, and then the luciferase activity was examined. It showed that the hsa-miR-660-3p mimic significantly suppressed the luciferase activity of WT-3’UTR (p<0.01, Figure 5F, Column 2) but not MT-3’UTR (Figure 5F, Column 5). In contrast, the hsa-miR-660-3p inhibitor significantly increased the luciferase activity of WT-3’UTR (p<0.01, Figure 5F, Column 3) but not MT-3’UTR (Figure 5F, Column 6), indicating that TEP1 is a bona fide target of hsa-miR-660-3p in HSCs.

Overexpression of hsa_circ_0004018 increased the expression of TEP1, which can significantly suppress the proliferation and activation of HSCs

To explore the role of TEP1, we overexpressed TEP1 in the primary HSCs (Figure 6A). We found that the upregulation of TEP1 significantly inhibited the proliferation of HSCs (p<0.01, Figure 6B), and resulted in an obvious decrease in the expression of α-SMA and COL1A1 (Figure 6C).

In addition, both the RNA and protein levels of TEP1 significantly increased after overexpression of hsa_ circ_0004018 in the HSCs (Figure 6D, 6E). We further co-transfected the mimic with hsa_circ_0004018 in the HSCs and then examined the expression of TEP1. As control, the negative control or the vector was transfected in combination as indicated (Figure 6F, 6G). As expected, both the RNA and protein levels of TEP1 of the mimic/vector combination were significantly lower than the negative control/vector combination (p<0.01, Figure 6F, Column 1 vs. Column 2; Figure 6G, Lane 2 vs. Lane 1). In parallel, the RNA and protein levels of TEP1 of the mimic/hsa_circ_0004018 combination were significantly lower than the negative control/ hsa_circ_0004018 combination (p<0.01, Figure 6F, Column 3 vs. Column 4; Figure 6G, Lane 3 vs. Lane 4), indicating that hsa_circ_0004018 and hsa-miR-660-3p counteract with each other in the regulation of TEP1.

Overexpression of hsa_circ_0004018 in vivo alleviated the progression of CCl₄-induced liver fibrosis in mouse model

To explore the therapeutic potential of hsa_circ_ 0004018 overexpression, we further transduced the hsa_circ_0004018 expressing lentivirus into the livers of CCl₄-induced liver fibrosis mice. It showed that liver fibrosis was alleviated in the hsa_circ_0004018 lentivirus transduced mice comparing with the control mice (Figure 7A, 7B). In accordance with the results of the in-vitro experiments, the relative levels of hsa-miR-660-3p were significantly lower in the hsa_circ_0004018 lentivirus transduced livers than the control livers (Figure 7C), while both the RNA and protein levels of TEP1 were significantly higher in the hsa_circ_0004018 lentivirus transduced livers than the control livers (Figure 7D, 7E). Meanwhile, the expression levels of α-SMA and COL1A1 obviously decreased after transduction with hsa_circ_0004018 lentivirus (Figure 7F–7H), suggesting that the upregulation of hsa_circ_ 0004018 suppressed the activation of HSCs in vivo.

Plasmids

The sequence of hsa_circ_0004018 was obtained from CircBase (http://circbase.org, genome position: chr17: 1703150-1704318). The cDNA of hsa_circ_0004018 was cloned into the pcDNA3.1(+) circRNA mini vector with EcoRI and SacI restriction sites:

 (forward primer: 5’-GAATTCACACTCCAATTCTC TGCCTAC-3’;

 reverse primer: 5’-GAGCTCCAAGACCAGTCTGG GCAAC-3’),

and into the pLO-ciR vector with EcoRI and NdeI resctriction sites:

 (forward primer: 5’-GAATTCACACTCCAATTCTC TGCCTAC-3’;

 reverse primer: 5’-CATATGCAAGACCAGTCTGG GCAAC-3’).

The sequence of TEP1 was obtained from NCBI (NM_001319035). The cDNA was cloned into pcDNA3.1(+) vector with NotI and XbaI restriction sites:

 (forward primer: 5’-GCGGCCGCATGGAAAAACT CCATGGGC-3’;

 reverse primer: 5’-TCTAGATCATTCCCAATTCAG AAAGTAC-3’).

All the plasmids were sequenced before use.

Cell culture

The isolation of HSCs was referred to the previous protocol [26, 27]. Firstly, we carefully isolated the livers from the normal mice or the liver fibrosis mice. Then we digested the livers with pronase E (0.4mg/ml, Sigma Aldrich, Catalog# 1074330001)/collagenase P (1.5mg/ml, Sigma Aldrich, Catalog#11213857001) solution to made disperse cells. After being sorted by FACS, the HSCs population was collected and seeded at a density of 2×10⁴ cells/cm² in the culture plates with HSC culture medium (DMEM plus 10% FBS, 4mM L-glutamine and 1% penicillin-streptomycin solution) (DMEM: GIBCO, Catalog#11965; FBS: Biological Industry, Catalog#04-001-1A; L-glutamine: Sigma Aldrich, Catalog#G3126), and incubated in the incubator with 5% CO₂ at 37 °C overnight. After attachment, the cells were carefully replaced with new prewarmed medium. Since the expression of α-SMA is a classical marker to positively identify HSCs upon culturing, we examined the expression of α-SMA of the harvested cells by immunofluorescence with the α-SMA antibody (CST, Catalog#19245, 1:500). In addition, to distinguish the HSCs from the hepatic myofibroblasts which also express α-SMA, we also detected the expression of desmin and glial fibrillary acidic protein (GFAP), which have been reported to be almost absent in myofibroblasts [27–29], by immunofluorescence with desmin antibody (Abcam, Catalog#ab32362, 1:500) and GFAP antibody (Abcam, Catalog#ab7260, 1:500). We observed the expression of those markers under the confocal microscopy (Leica TCS SP5) and found that almost all (>90%) of the harvested cells co-expressed the markers of α-SMA, desmin and GFAP. The HSCs used in the following experiments were all cultured for 3 days from cell attachment.

For IL-6 treatment experiments, the primary HSCs (1×10⁶ cells/plate in 6-cm dish) from the normal mouse livers were respected treated with 0, 20 and 40 ng/L of recombinant murine IL-6 (Peprotech, Catalog#216-16) for 24 hours. As for salvianolic acid B (Sal B) treatment experiments, the primary HSCs from the fibrotic mouse livers were treated with 100 ug/L of Sal B (Sigma, Catalog#49724) for 0, 3 and 6 days, followed by cell collection and RNA isolation. In the Sal B gradient experiment, the primary HSCs from the fibrotic mouse livers were respectively treated with 0, 80 and 160 ug/L for 3 days, followed by cell collection and RNA isolation. All the experiments were performed in triplicate.

Immunofluorescence microscopy

For the EDU incorporation experiment, iClick™ EdU Andy Fluor™ 647 Imaging Kit (GeneCopoeia, Catalog#A006) was used. The experiment procedure was performed according to the manual of the kit. Firstly, the HSCs (transfected with hsa_circ_0004018 overexpressing or control plasmids) were seeded with the density of 2×10⁴ cells/cm² in the 12-well culture plates with poly-D-lysine-coated glass coverslips on the bottom. After cell attachment, the medium of those for EDU incorporation experiment was replaced with EDU containing (10μM of the final concentration) new HSC culture medium. After incubation for 24 hours, the coverslips were collected and fixed with 4% paraformaldehyde, followed by DAPI staining and washing steps.

As for the detection of α-SMA and COL1A1 expression, the coverslips seeded with the primary HSCs were harvested and fixed in 4% paraformaldehyde, followed by the IgG (CST, Catalog#2729) blocking step. Then, those IgG-blocked coverslips were stained with α-SMA antibody (CST, Catalog#19245) or COL1A1 antibody (CST, Catalog# 39952) with the dilution of 1: 500, followed by secondary antibody incubation, DAPI staining and washing steps.

All the coverslips mentioned above were inverted onto the glass slides with mounting media and carefully observed and photographed under confocal microscopy (Leica TCS SP5).

Biotin labeled probe pull down assay

The pull assay was referred to the previous study [30]. The biotin labeled hsa_circ_0004018 probe and hsa-miR-660-3p probe were synthesized by GeneChem. The sequence of hsa_circ_0004018 probe is as follows: 5’-CAGACTGGTCTTGACACTCCAATTCTCTGCCTA-3’-biotin.

The sequence of hsa-miR-660-3p probe is as follows: 5’-ACCUCCUGUGUGCAUGGAUUA-3’-biotin.

For purification, about 1×10⁷ cells were collected and lysed in lysis buffer. After that, 3μg biotin labeled probe was added to the buffer and incubated at room temperature for 4 hours. To pull down the cirRNA-microRNA complex, strepavidin magnetic beads (Thermo Fisher, Catalog#88816) were added to the buffer and slowly rotated for another 4 hours, followed by washing step for 4 times. Finally, the beads were collected by centrifuge, and the binding RNA was extracted with TRIzol reagent (Thermo Fisher, Catalog#15596026) for further real-time PCR examination.

Cell cycle analysis

We collected about 3×10⁵ HSCs of each group (hsa_circ_0004018 overexpression vs. vector control in triplicate) for the staining of propidium iodide according to the manufacture protocol of the cell cycle analysis kits (Beyotime, Catalog#C1052). The prepared cell suspension was analyzed by FACS (BD LSRFortessa). All the FACS data were re-analyzed by FlowJo software.

Real-time PCR

The real-time PCR experiments were performed according the manufacture protocol (GoTaq qPCR System, Promega, Catalog#A6001) on the ABI 7500 Real-Time PCR System. Each sample were examined in triplicate from three independent experiments. The PCR primers used in this study are as follows:

Western blotting

We collected the same number of cells from each group. Cells were extracted with 1% SDS cell lysis/loading buffer. All the protein samples were subjected to electrophoresis by SDS-PAGE method and then transferred to PVDF membrane for further immunoblot. The primary antibodies used in this study include: α-SMA antibody (CST, Catalog#19245, 1:1000), COL1A1 antibody (CST, Catalog# 39952, 1:1000), TEP1 antibody (Abcam, Catalog#ab64189, 1:1000), β-actin antibody (CST, Catalog#3700T, 1:3000). The blots were exposed using chemiluminescence and photographed by Tannon 3500 Imager. Photos were analyzed by ImageJ software.

Cell proliferation assay (CCK8 assay)

For this assay, we cultured 1×10³ cells per wells on 96-well plates and examined the cell numbers in each well using the Cell Counting Kit (CCK8, Dojindo, Catalog#CK04) at 24, 48 and 72 hours. Each group of cells was examined in triplicate. The OD value was determined by Microplate Absorbance Reader (Bio-rad, iMark^TM).

Luciferase reporter assay

This assay was referred to the protocol of the previous study [31]. A 200bp fragment of hsa_circ_0004018 3’ terminal or TEP1 3’UTR was inserted into the 3’ terminal of the luciferase gene in pGL3 vector (pGL3-3’) and sequenced before use. About 1×10⁵ cells were seeded in 48-well plate before transfection. After attachment, the cells of each well were respectively transfected with 320 ng of pGL3-3’ and 30 ng of pRL-TK containing Renilla luciferase plus hsa-miR-660-3p mimic (original sequence: 5’-ACCUCCUGUGUGCA UGGAUUA-3’), mimic negative control or inhibitor (GenePharma) using Lipfatamin 2000 (Invitrogen, Catalog#11668019). 48 hours after transfection, the relative luciferase absorbance value of each group was examined by Dual-Luciferase Reporter Assay System (Promega, Catalog#E1910) and normalized to Renilla luciferase absorbance value. Each group was performed in triplicate from three independent experiments.

Lentivirus in-vivo administration

The lentivirus expressing hsa_circ_0004018 or negative control was purchased from GeneChem. Lentivirus solution was injected via the tail vein (about 5×10⁸ units of each mouse) twice respectively at the first and the second week of CCl₄ injection (once a week for 4 weeks). All groups of mice were sacrificed at the end of 4 weeks for further pathological examination.

Immunochemistry

The liver tissue samples collected from the model and the control mice were fixed in 4% paraformaldehyde overnight at room temperature. Following with the steps of dehydration, embedding in paraffin, the samples were sliced into 5-8μm thickness and transferred onto glass slides. The sections were then immunostained with α-SMA and COL1A1 primary antibody at 4 °C overnight and then incubated with biotinylated secondary antibody. After incubation with Sav-HRP conjugates, the sections were applied with DAB substrate for color development and observed under microscopy.

CCl₄-induced liver fibrosis mouse model

The liver fibrosis mouse model was established referring to the previous report [32]. C57BL/6 male mice of 6-8 weeks old and 20±2 g in weight were supplied by the Experimental Animal Center of Hubei University of Medicine. Each mouse was treated with 2 ml of CCl₄ (Sinopharm Chemical Reagent Co., Catalog# 10006480)/olive oil (1:1, v/v) per kg body weight by intraperitoneal injection twice for 6 weeks. Mice injected with equal volume of olive oil were performed as control.

All the mice used in this study were under human care and provided with enough food and water. When it came to the end of each experimental point, those mice were euthanatized with CO₂. All the experiments were approved and supervised by the Animal Welfare and Ethics Committee of Renmin Hospital, Hubei University of Medicine.

Masson’s trichrome staining and the morphometry

The liver samples were isolated from the liver fibrosis mice sacrificed on day 0, 30 and 45 upon injection with CCl₄ and then fixed in 4% paraformaldehyde and embedded in paraffin, followed by frozen section. For the Masson’s trichrome staining of the liver sections, Trichrome Stain Kit (Abcam, Catalog#ab150686) was used. Briefly, the sections were firstly deparaffinized and rehydrated in distilled water. Then, the sections were incubated successively with preheated Bouin's Fluid, Weigert's Iron Hematoxylin, Biebrich Scarlet/Acid Fuchsin solution, phosphomolybdic/phosphotungstic acid solution, Aniline Blue solution and acetic acid solution, with rinse steps at the interval of every two incubation steps. The duration of each incubation step was controlled according to the recommendatory procedures of the manual. The stained sections were then carefully observed and photographed under microscope (Olympus). The proportion of each staining area was analyzed by ImageJ software. The blue staining area represents the collagen-enriched tissues, which was considered as the fibrosis area of the livers.

Statistical analysis

All the data were analyzed using SPSS software. Student’s t test was used to compare the difference of data between groups. P<0.05 represents significant difference. The asterisks *, ** and *** stand for p<0.05, p<0.01 and p<0.001, respectively.

Data availability statement

The datasets used in the current study are available from the corresponding author on reasonable request.