CTGF Mediates Tumor-Stroma Interactions Between Hepatoma Cells and Hepatic

Home , CTGF

Author Manuscript Published OnlineFirst on July 2, 2018; DOI: 10.1158/0008-5472.CAN-17-3844 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

CTGF mediates tumor-stroma interactions between hepatoma cells and hepatic

stellate cells to accelerate HCC progression.

Yuki Makino1; Hayato Hikita1; Takahiro Kodama1; Minoru Shigekawa1; Ryoko

Yamada1; Ryotaro Sakamori1; Hidetoshi Eguchi2; Eiichi Morii3; Hideki Yokoi4; Masashi

Mukoyama4, 5; Suemizu Hiroshi6; Tomohide Tatsumi1; Tetsuo Takehara1

Departments of 1Gastroenterology and Hepatology, 2Gastroenterological Surgery, and

3Pathology, Osaka University Graduate School of Medicine, Osaka, Japan

4Department of Nephrology, Kyoto University Graduate School of Medicine, Kyoto,

Japan

5Department of Nephrology, Kumamoto University Graduate School of Medical

Sciences, Kumamoto, Japan

6Central Institute for Experimental Animals, Kanagawa, Japan

*Correspondence:

Tetsuo Takehara, MD, PhD

Department of Gastroenterology and Hepatology

Osaka University Graduate School of Medicine

2-2 Yamadaoka, Suita, Osaka, 565-0871, Japan

Tel.: +81-6-6879-3621 Fax: +81-6-6879-3629

E-mail: [email protected]

Running title: CTGF accelerates HCC growth via tumor-stroma interactions.

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 2018 American Association for Cancer Research. Author Manuscript Published OnlineFirst on July 2, 2018; DOI: 10.1158/0008-5472.CAN-17-3844 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Word Count: 4566

Number of Figures: 6

Number of Tables: 1

Financial Support

This work was partially supported by a Grant-in-Aid for Scientific Research (to T.

Takehara. JP26670381 and JP26253947) and a Grant-in-Aid for Young Scientists (to Y.

Makino. JP17K15943) from the Ministry of Education, Culture, Sports, Science, and

Technology, Japan, a research grant from Bristol-Myers Squibb (to T. Takehara), and a

Grant-in-Aid for Research from the Japan Agency for Medical Research and

Development (JP18fk0210021, JP18fk0310108, JP18fk0210018 and JP18fk0210026) .

Conflict of interest: The authors declare no potential conflicts of interest.

Abstract

Connective tissue growth factor (CTGF) is a matricellular protein related to hepatic

fibrosis. This study aims to clarify the roles of CTGF in hepatocellular carcinoma

(HCC), which usually develops from fibrotic liver. CTGF was overexpressed in 93

human HCC compared with non-tumorous tissues, primarily in tumor cells. Increased

CTGF expression was associated with clinicopathological malignancy of HCC. CTGF

was upregulated in hepatoma cells in hepatocyte-specific Kras-mutated mice (Alb-Cre

KrasLSL-G12D/+). Hepatocyte-specific knockout of CTGF in these mice (Alb-Cre

KrasLSL-G12D/+ CTGFfl/fl) decreased liver tumor number and size. Hepatic stellate cells

(HSC) were present in both human and murine liver tumors, and α-SMA expression, a

marker of HSC activation, positively correlated with CTGF expression. Forced

expression of CTGF did not affect growth of PLC/PRF/5 cells, a hepatoma cell line

with little CTGF expression, but facilitated their growth in the presence of LX-2 cells, a

hepatic stellate cell line. The growth of HepG2 cells, which express high levels of

CTGF, was promoted by co-culture with LX-2 cells compared with monoculture.

Growth promotion by LX-2 cells was negated by an anti-CTGF antibody in both culture

and xenografts. Co-culturing LX-2 cells with HepG2 cells drove LX-2-derived

production of IL-6, which led to STAT-3 activation and proliferation of HepG2 cells. An

anti-CTGF antibody reduced IL-6 production in LX-2 cells and suppressed STAT-3

activation in HepG2 cells. In conclusion, our data identify tumor cell-derived CTGF

as a keystone in the HCC microenvironment, activating nearby HSC which transmit

pro-growth signals to HCC cells, this interaction is susceptible to inhibition by an

anti-CTGF antibody.

(Abstract: 254 words)

Significance

Pro-tumor crosstalk between cancer cells and hepatic stellate cells presents an

opportunity for therapeutic intervention aganst HCC.

Introduction

Connective tissue growth factor (CTGF), also known as CCN2, is a member of the

CCN (CCN1-6) family proteins [1]. CTGF is a secreted matricellular protein that

interacts with various molecules in the extracellular matrix (ECM). CTGF contains an

N-terminal secretory peptide followed by four conserved domains with sequence

homologies to insulin-like growth factor-binding proteins, the von Willebrand factor C

(VWC) domain, a motif related to thrombospondin, and a C-terminal domain that

contains a cysteine-knot motif [1-3]. CTGF has various biological functions, including

cell adhesion, migration, proliferation, differentiation and ECM production, and

participates in the development of many organs under normal physiological conditions

[2]. CTGF is pathologically viewed as a central mediator of tissue remodeling and

fibrosis of various organs, including the lung, heart, liver and kidney [1-4]. In addition

to fibrotic diseases, CTGF has also been reported to be associated with the progression

of various malignant diseases, such as breast, pancreatic, and gastric cancers [5-8].

Hepatocellular carcinoma (HCC) is a major cancer type and is the third most common

cause of cancer-related mortality worldwide [9]. One study has reported that this is not

the case, but in general, it has been reported that CTGF is highly expressed in HCC

tissues [10-12]. CTGF is expressed in several cell types in the liver, including

hepatocytes, cholangiocytes, hepatic stellate cells (HSCs), fibroblasts, and sinusoidal

endothelial cells [4, 8, 10, 13-15]. The types of cells among HCC tissues that express

CTGF have not been clarified. CTGF increases the proliferation of cancer-associated

fibroblasts [10] and the migration of macrophages [14]. In addition, it has been reported

that CTGF acts directly on hepatoma cells and increases DNA synthesis [11], cell cycle

progression [13], invasion and migration abilities [13] and resistance to doxorubicin and

TRAIL-induced apoptosis [11]. However, these reported actions were mostly based on

in vitro experiments, and the significance of CTGF in vivo is completely unknown.

Here, we clarified that CTGF was upregulated in HCC compared with surrounding

non-tumorous tissues and that CTGF is expressed primarily in cancer cells. Using a

genetically engineered mouse model, a hepatocyte-specific knockout of CTGF

suppressed the activation of HSCs and the progression of liver cancer. Through the

specific inhibition of CTGF function with an anti-CTGF-neutralizing antibody, which is

currently being tested in clinical trials for other diseases [3, 16, 17], we further

demonstrated that the growth-promoting effect of CTGF is mediated by tumor-stroma

interactions between cancer cells and HSCs. This report provides the first demonstration

that CTGF produced in cancer cells activates HSCs in the tumor microenvironment and

accelerates the progression of liver cancer.

Materials and Methods

Human samples

Liver samples were collected from both tumorous and non-tumorous liver tissues of

HCC patients who underwent hepatectomy. Written informed consent was obtained

from all patients. The study protocol conformed to the ethical guidelines of the

Declaration of Helsinki. Approval for the use of resected samples was obtained from the

Institutional Review Board (IRB) Committees at Osaka University Hospital (IRB No.

13556 and 15267).

Mice

C57BL/6/129 background mice carrying a lox P-stop-lox P (LSL) termination sequence

with the KrasG12D point mutation allele (KrasLSL-G12D/+) were kindly provided by the

National Cancer Institute (Bethesda, MD, USA). Hepatocyte-specific Kras-mutated

mice (KrasLSL-G12D/+Alb-Cre; KrasG12D mice) were generated by mating KrasLSL-G12D/+

mice and heterozygous Alb-Cre transgenic mice that expressed the Cre recombinase

gene under the promoter of the albumin gene [4]. C57BL/6/J background mice that

carried two floxed CTGF alleles (CTGFfl/fl) have been previously described [18]. The

CTGFfl/fl mice were genotyped using the following primers for the ctgf allele:

5’-ACAATGACATCTTTGAGTCC-3’ and 5’-AGTCTAATGAGTTCGTGTCC-3’. We

generated hepatocyte-specific CTGF-knockout Kras-mutated mice by mating KrasG12D

mice and CTGF-floxed mice and littermates generated were used to compare the

phenotypes among experimental groups. NOD/Shi-scid/IL-2Rγ (null) (NOG) mice have

been previously described [19]. Only male mice were used in the experiments. The

institute of experimental animal sciences of Osaka University Graduate School of

Medicine specifically approved these studies and the mice were maintained in a specific

pathogen-free facility and were treated humanely.

Cell culture

The human hepatoma cell lines HepG2, Huh7, PLC/PRF/5, and HLF were obtained

from the JCRB/HSRRB cell bank (Osaka, Japan) in 2012, 2015, 2015, and 2011,

respectively. The human HSC line LX-2 was kindly provided by Prof. Eiji Miyoshi

(Department of Molecular Biochemistry and Clinical Investigation, Osaka University

Graduate School of Medicine, Osaka, Japan) in 2015 and the cell authentication was

performed by the JCRB/HSRRB cell bank (Osaka, Japan). Mycoplasma testing was

performed for all cell lines using MycoAlert™ Mycoplasma Detection Kit (Basel,

Switzerland) according to the manufacturer’s recommended protocols. The latest date

the cells were tested is December 5, 2017. The cells were cultured at 37°C with 5% CO2

in Dulbecco's Modified Eagle’s Medium (DMEM) (Sigma-Aldrich, St. Louis, MO,

USA) supplemented with 10% fetal calf serum and antibiotics unless otherwise

indicated. All experiments were performed using cells with the passage number of less

than 30. The Transwell insert system (Corning, Corning, NY, USA) was used in the

co-culture experiments.

Cell proliferation was analyzed via the WST-8 assay (Nacalai Tesque, Kyoto, Japan).

LY294002, a phosphoinositide 3-kinase (PI3K) inhibitor, FR180204, an extracellular

signal-regulated kinase (Erk) inhibitor, and U0126, a mitogen-activated protein

kinase/Erk kinase (Mek) inhibitor, were purchased from Sigma-Aldrich. Recombinant

human epidermal growth factor (EGF) protein was purchased from Thermo Fisher

Scientific (Waltham, MA, USA). Recombinant human CTGF protein was obtained from

FibroGen (San Francisco, CA, USA). Anti-IL-6 neutralizing antibody was purchased

from Thermo Fisher Scientific. Kras siRNA, IL-6 siRNA, CTGF siRNA, STAT-3

siRNA and control siRNA (Thermo Fisher Scientific) were transfected into the cells

using Lipofectamine RNAiMAX (Thermo Fisher Scientific) according to the reverse

transfection protocol. The CTGF cDNA plasmid vector or the control empty plasmid

vector (OriGene, Rockville, MD, USA) was transfected using Lipofectamine 3000

(Thermo Fisher Scientific) according to the manufacturer’s recommended protocols. To

establish a cell line that stably overexpresses CTGF, transfected cells were selected over

four weeks with G418 (Nacalai Tesque) treatment, and several single colonies were

isolated after selection. The expression levels of CTGF were compared among these

colonies, and the one with the highest CTGF expression was used in the subsequent

experiments. CTGF shRNA plasmid (Santa Cruz Biotechnology, Dallas, TX, USA) or

control shRNA plasmid (Santa Cruz Biotechnology) were transfected according to the

manufacturer’s protocol. Puromycin (Thermo Fisher Scientific) treatment was

employed to select the stably transfected cells.

Anti-CTGF neutralizing antibody

The fully recombinant human monoclonal IgG1 anti-CTGF antibody FG-3019 and a

non-specific human IgG control antibody (hIgG) were kindly provided by FibroGen.

Experimental protocol for xenograft tumor models

Under general anesthesia, 7- to 9-week-old male NOG mice were subcutaneously

inoculated in the bilateral flank portions with 1×107 hepatoma cells with or without

1×107 LX-2 cells suspended in Matrigel (Corning). To evaluate the effect of CTGF

inhibition, 40 mg/kg FG-3019 or hIgG was intraperitoneally administered twice per

week from the day of inoculation. The length and width of the xenograft tumors were

measured twice per week, and the tumor volume was calculated according to the

following formula: tumor volume = 1/2 × (major axis) × (minor axis)2 [20].

Histological analyses

Liver sections were routinely stained with hematoxylin and eosin.

Immunohistochemical analyses were performed using paraffin-embedded liver sections

and a CTGF antibody (Santa Cruz Biotechnology), an α-SMA antibody (Abcam,

Cambridge, UK), and a PCNA antibody (Cell Signaling Technology, Danvers, MA,

USA). PCNA-positive cells were counted in high-power fields (HPFs) at 400-fold

magnification. Positive areas of α-SMA staining within the liver sections were

quantified as the α-SMA area divided by total tissue area under a HPF at 400-fold

magnification using the image-analysis software WinROOF (Mitani Corporation, Fukui,

Japan), as previously described [21-23]. Tumor area / total tissue area ratio in

macroscopically non-tumorous liver tissues was quantified at 40-fold magnification

using WinROOF (Mitani Corporation). Double immunofluorescence staining was

performed using frozen liver sections and a CTGF antibody conjugated to Alexa Fluor

488 (Santa Cruz Biotechnology) and an α-SMA antibody conjugated to Alexa Fluor 594

(Santa Cruz Biotechnology). We used 4',6-diamidino-2-phenylindole (DAPI) to label

the nuclei.

RNA isolation and quantitative real-time reverse-transcription PCR (qRT-PCR)

Total RNA was extracted from cell lines or liver tissues using the RNeasy Mini Kit

(QIAGEN, Hilden, Germany) according to the manufacturer’s recommended protocols,

and cDNA was produced by reverse transcription as previously described [21]. The

mRNA expression of specific genes was quantified by real-time RT-PCR using TaqMan

Gene Expression Assays (Applied Biosystems, Foster City, CA, USA) as follows:

mouse-ctgf (Mm01192933_g1), mouse-afp (Mm00431715_m1), mouse-glypican-3

(Mm00516722_m1), mouse-acta2 (Mm00725412_s1), mouse-β-actin

(Mm02619580_g1), human-ctgf (Hs00170014_m1), human-kras (Hs00364284_g1),

human-β-actin (Hs01060665_g1), human-il 6 (Hs00985639_m1), and human-acta2

(Hs00426835_g1). All expression levels were normalized to those of β-actin.

Western blot analysis

Liver tissue was lysed in lysis buffer (1% Nonidet P-40, 0.5% sodium deoxycholate,

0.1% sodium dodecyl sulfate (SDS), protease inhibitor cocktail (Nacalai Tesque),

phosphatase inhibitor cocktail (Nacalai Tesque), and PBS pH 7.4). Equal amounts of

protein were electrophoretically separated using SDS polyacrylamide gels and

transferred onto a polyvinylidene fluoride membrane. The following antibodies were

used for immunodetection: anti-Ras, anti-phospho-Erk, anti-phospho-Akt,

anti-phospho-STAT-3, anti-β-actin (Cell Signaling Technology), anti-α-SMA (Abcam)

and anti-CTGF (Santa Cruz Biotechnology). ImageJ software (NIH, Bethesda, MD,

USA) was used to quantify the bands and the expression levels were normalized to

those of β-actin.

Enzyme-linked immunosorbent assay (ELISA)

Commercial ELISA kits for CTGF (PeproTech, Rocky Hill, NJ, USA) and IL-6 (R&D

systems) were used to quantify their concentrations in cell culture supernatants

according to the manufacturer’s recommended protocols.

Statistical analysis

Differences between unpaired groups with normal or non-normal distributions were

compared using Student’s t-test or the Mann-Whitney U test, respectively. The

Wilcoxon signed-rank test was used to compare two paired samples. Correlations were

assessed using the Pearson product-moment correlation coefficient. The chi-square test

was used to analyze categorical data. Parametric or non-parametric multiple

comparisons were performed using one-way analysis of variance (ANOVA) followed by

the Tukey-Kramer post-hoc test or Kruskal-Wallis test followed by the Steel-Dwass test,

respectively. P-values less than 0.05 were considered to indicate statistical significance.

Results

CTGF is highly expressed in human HCC tissues and is related to the malignant

characteristics of HCC

The CTGF gene expression levels in tumorous tissues and surrounding non-tumorous

tissues were compared among 93 human liver samples from patients who underwent

surgical hepatectomy for HCC (Supplementary Table 1). CTGF gene expression was

significantly upregulated in tumorous tissues compared with non-tumorous tissues (Fig.

1A). A western blot analysis also showed upregulation of CTGF protein in tumorous

tissues (Fig. 1B). Positive immunohistochemical staining for CTGF was observed

primarily in the cytoplasm of hepatoma cells (Fig. 1C). Immunohistochemistry using

serial-sections revealed AFP expression in CTGF-positive cells in tumor tissues

(Supplementary Fig. 1A). In tumor tissues, α-SMA-positive HSCs were detected by

immunohistochemistry, and α-SMA mRNA was upregulated in these tissues compared

with non-tumorous liver tissues (Fig. 1D). Alpha-SMA expression in tumorous tissues

were positively correlated with that of CTGF (r = 0.21, p<0.05). In double

immunofluorescence staining, α-SMA-positive cells expressed CTGF, but comprised

only a small fraction of the total CTGF expression in the tumor (Supplementary Fig.

1B). We further investigated the association between CTGF gene expression levels and

tumor characteristics. According to the median value of the ratio of CTGF mRNA

expression in tumorous tissues to non-tumorous tissues, we divided 93 patients into a

CTGF high-expression group (CTGF mRNA expression ratio > 1.3, N = 46) and a

low-expression group (CTGF mRNA expression ratio ≤ 1.3, N = 47) and compared the

tumor characteristics between the two groups. The CTGF high-expression group

showed higher positive rates of des-γ-carboxy prothrombin, tumor multiplicity, portal

invasion, and malignant macroscopic tumor classification, which suggests more

malignant characteristics of HCC (Table 1). In addition, the CTGF high-expression

group exhibited higher α-SMA expression in tumorous tissues (Fig. 1E).

CTGF is upregulated in liver tumors of KrasG12D mice

We examined the relationship between liver tumors and CTGF using genetically

modified mice that develop liver tumors. KrasG12D mice developed liver tumors at high

rates after 4 months of age, and the number and diameter of the tumors increased with

age (Fig. 2A and 2B). In the liver tissues of KrasG12D mice, Ras protein accumulation

was confirmed, and the Ras signaling pathway was activated, as assessed by increased

Erk and Akt phosphorylation (Fig. 2C). The expression of p-Erk was more prominent in

tumors than in non-tumorous tissues from the KrasG12D mice. A gene expression analysis

revealed that CTGF and tumor markers for HCC, such as AFP and glypican-3, were

significantly upregulated in tumors compared with non-tumorous tissues from control

mice and KrasG12D mice (Fig. 2D). A western blot analysis showed that CTGF was

overexpressed in tumors from KrasG12D mice compared with non-tumorous tissues from

KrasG12D mice and control mice (Fig. 2C). Immunohistochemical staining for CTGF

demonstrated expression primarily in hepatoma cells (Fig. 2E). An analysis of

immunohistochemical staining in serial sections showed a positive reaction for AFP in

CTGF-positive cells in tumor tissues (Supplementary Fig. 1C). Immunohistochemistry

also revealed α-SMA-positive HSCs in liver tumors (Fig. 2F). Alpha-SMA was

upregulated in tumorous tissues compared with non-tumorous liver tissues from control

mice and KrasG12D mice (Fig. 2C and 2F). In tumorous tissues, α-SMA gene expression

was positively correlated with that of CTGF (Fig. 2F). Double immunofluorescence

staining α-SMA and CTGF suggested that although α-SMA-positive cells were also

positive for CTGF, CTGF derived from α-SMA-positive cells accounted for only a

small fraction of the total CTGF expressed in the tumor (Supplementary Fig. 1D).

The Ras/Mek/Erk pathway regulates CTGF expression in hepatoma cells

We examined the relationship between the Ras signaling pathway and CTGF expression

through in vitro and in silico analyses. The stimulation of Huh7 cells, which are Kras

wild-type hepatoma cells, with EGF activated Erk and Akt downstream of Ras and

increased gene expression and secretion of CTGF (Supplementary Fig. 2A). In HepG2

cells, which contain mutant Kras, CTGF was downregulated by siRNA-mediated Kras

knockdown (Supplementary Fig. 2B). PI3K inhibition and Mek/Erk inhibition

specifically down-regulated p-Akt and p-Erk, respectively (Supplementary Fig.2C).

CTGF expression in HepG2 cells was also decreased by Mek and Erk inhibitors but not

by a PI3K inhibitor (Supplementary Fig. 2D) 6 hours after their addition, whereas the

cell viability was not affected (Supplementary Fig.2D). Furthermore, to investigate an

association between CTGF gene expression levels and the activity of the Ras/Mek/Erk

pathway in HCC patients, we performed a single-sample gene set enrichment analysis

(ssGSEA) using a publicly available microarray dataset of HCC patients. The ssGSEA

revealed a positive correlation between CTGF expression and Ras/Mek/Erk pathway

activity in the liver (Supplementary Fig. 2E). These results suggest that Ras/Mek/Erk

pathway activation is associated with CTGF upregulation in human liver tissue.

Hepatocyte-specific CTGF deficiency inhibits tumor development and progression

in KrasG12D mice

To analyze the functions of CTGF in HCC, we generated hepatocyte-specific

CTGF-deficient KrasG12D mice and compared the phenotypes of three groups at 8

months: 1) KrasG12D CTGF+/+ mice (CTGF+/+ KrasLSL-G12D/+Alb-Cre), 2) KrasG12D

CTGF+/- mice (CTGFfl/+ KrasLSL-G12D/+ Alb-Cre), and 3) KrasG12D CTGF-/- mice (CTGFfl/fl

KrasLSL-G12D/+ Alb-Cre). The incidence rate of macroscopic liver tumors was lower in

KrasG12D CTGF-/- mice than in KrasG12D CTGF+/+ mice, although this difference was not

statistically significant (Fig. 3A and 3B). Importantly, KrasG12D CTGF-/- mice showed a

significant reduction in the number of macroscopic tumors, tumor size, liver/body

weight ratio, and histological tumor/non-tumor area ratio compared with KrasG12D

CTGF+/+ mice (Fig. 3B and 3C). The number of PCNA-positive cancer cells in these

liver tumors was lower in KrasG12D CTGF-/- mice than in KrasG12D CTGF+/+ mice, which

suggests the attenuation of tumor proliferation in KrasG12D CTGF-/- mice (Fig. 3D). Gene

and protein expression of CTGF in these liver tumors was decreased in KrasG12D

CTGF-/- mice compared with KrasG12D CTGF+/+ mice (Fig. 3E). The number of positive

areas of immunohistochemical staining for α-SMA was decreased in liver tumors in

KrasG12D CTGF-/- mice compared with KrasG12D CTGF+/+ mice (Fig. 3F). KrasG12D

CTGF+/- mice exhibited intermediate phenotypes (Fig. 3). These results indicate that

CTGF produced by hepatoma cells is involved in the development and progression of

liver tumors, as well as the activation of HSCs in tumorous tissues.

Forced CTGF expression does not affect hepatoma cell growth but increases their

growth in the presence of HSCs

To examine the influence of CTGF on the proliferation of hepatoma cells, we

established CTGF-overexpressing PLC/PRF/5 cells, which originally showed the lowest

endogenous levels of CTGF expression and secretion among several hepatoma cell lines

(Huh7, HLF, HepG2, PLC/PRF/5 and Hep3B) (Supplementary Fig. 3A and 3B). The in

vitro growth of CTGF-overexpressing PLC/PRF/5 cells was not different from that of

the parental or mock-transfected PLC/PRF/5 cells (Fig. 4A). After PLC/PRF/5 cells

were xenografted into NOG mice, the growth was not found to be different between

tumors derived from CTGF-overexpressing PLC/PRF/5 cells and those derived from

mock-transfected PLC/PRF/5 cells, which is similar to the results obtained from the in

vitro experiments (Fig. 4B). We subsequently investigated the influence of CTGF on the

proliferation of hepatoma cells in the presence of HSCs. During the co-culture of a

human HSC cell line with LX-2 cells in a Transwell system, CTGF-overexpressing

PLC/PRF/5 cells proliferated faster than mock-transfected PLC/PRF/5 cells (Fig. 4C).

In a xenograft model, the tumor volumes of CTGF-overexpressing PLC/PRF/5

cell-derived tumors were significantly larger than those of mock-transfected PLC/PRF/5

cell-derived tumors when the cells were co-injected with LX-2 cells (Fig. 4D). While

α-SMA positive cells were present along with the fibrous bands in xenograft tumors

mixed with LX-2 cells (Fig. 4E), predominant cell population in tumors was composed

of PLC/PRF/5 cells. It is therefore suggested that CTGF contributes to hepatoma cell

growth in the presence of HSCs both in vitro and in xenograft models.

The growth-promoting effect of HSCs on hepatoma cells with high CTGF

expression is abolished by inhibition of CTGF

In HepG2 cells, which showed the highest endogenous expression and secretion of

CTGF (Supplementary Fig. 3A), the application of an anti-CTGF-neutralizing antibody

FG-3019 did not affect cell growth (Fig. 5A). HepG2 cells grew faster when co-cultured

with LX-2 cells than when cultured alone (Fig. 5B). The proliferation of HepG2 cells

was facilitated by co-culture with LX-2 cells when the cells were exposed to control

hIgG, but the growth promoting effect of the co-culture with LX-2 cells was abolished

after exposure to an anti-CTGF neutralizing antibody (Fig. 5C). In a xenograft model,

the volumes of tumors derived from HepG2 cells co-injected with LX-2 cells were

larger than those derived from HepG2 cells injected alone in the hIgG group (Fig. 5D).

In contrast, in the anti-CTGF-neutralizing antibody group, the volumes of tumors

derived from HepG2 cells did not increase even when these cells were co-injected with

LX-2 cells (Fig. 5D). Similar to the experiments with an anti-CTGF neutralizing

antibody, hepatoma cell-specific CTGF inhibition by siRNA also abolished enhanced

hepatoma cell growth by the co-culture with HSCs in vitro (Supplementary Fig. 4A). In

xenograft models, shRNA-mediated CTGF knockdown in hepatoma cells abolished the

growth-promoting effect of HSCs (Supplementary Fig. 4B and 4C). Enhanced tumor

growth induced by the co-existence of hepatoma cells and HSCs is therefore suggested

to be CTGF-dependent.

CTGF induces IL-6 secretion from HSCs, which promotes the growth of hepatoma

cells

To investigate the mechanisms underlying the growth-facilitating effect of HSCs

mediated by CTGF, we focused on IL-6, which has been reported to be produced by

HSCs in the HCC microenvironment, where it facilitates tumor progression [24]. IL-6

was secreted not from HepG2 cells but from LX-2 cells in monoculture (Fig. 6A). The

secretion levels of IL-6 from LX-2 cells were increased by the recombinant CTGF

supplementation (Fig. 6B and Supplementary Table 2). Moreover, the concentration of

IL-6 in the supernatant was elevated by co-culture with HepG2 cells compared with

LX-2 cells alone (Fig. 6A). The growth-facilitating effect of LX-2 cells on HepG2 cells

was abolished by the siRNA-mediated knockdown of IL-6 in LX-2 cells, and this effect

was accompanied by the downregulation of p-STAT-3 in HepG2 cells (Fig. 6C).

Anti-IL-6 neutralizing antibody treatment also decreased the growth of HepG2 cells

co-cultured with LX-2 cells and down-regulated p-STAT-3 in HepG2 cells

(Supplementary Fig. 5A). The growth-facilitating effect of LX-2 cells was canceled by

STAT-3-knockdown in HepG2 cells (Supplementary Fig. 5B). When HepG2 cells and

LX-2 cells were cultured together, the application of an anti-CTGF-neutralizing

antibody also decreased the concentration of IL-6 in the supernatant and downregulated

p-STAT-3 in HepG2 cells (Fig. 6D), which resulted in the inhibition of enhanced cell

growth due to co-culture with LX-2 cells (Fig. 5B). Consistent with these results,

p-STAT-3 expression in liver tumors in KrasG12D CTGF-/- mice was lower than that in

KrasG12D CTGF+/+ mice (Supplementary Fig. 6).

Discussion

In the present study, using a genetically modified murine model, we demonstrated that

CTGF is highly expressed in human HCC tissues and is involved in liver tumor

formation. In addition, we showed evidence supporting the hypothesis that CTGF

promotes tumor growth via its interaction with HSCs within the tumor

microenvironment. HCC is a vasculature-rich cancer that arises predominantly from

fibrotic liver, and it is known that VEGF, Ang-2 and PDGF produced by hepatoma cells

or other cell types causes endothelial cell proliferation [25, 26]. It is also known that

HGF, FGF, and TGFβ produced by stromal cells induce the proliferation of hepatoma

cells [27, 28]. The present study provided the first demonstration of a previously

unknown role of CTGF in the association of hepatoma cells and HSCs.

Our ssGSEA revealed that CTGF expression was correlated with activation of the

Ras/Mek/Erk pathway in human HCC tissues (Supplementary Fig. 2E). EGF activated

the Ras signaling pathway in Huh7 cells, which led to CTGF upregulation

(Supplementary Fig. 2A). HepG2 cells, which harbor mutant Ras, produced high levels

of CTGF, whereas inhibition of the Ras signaling pathway decreased CTGF production

(Supplementary Fig. 2B and 2D). The activation of Ras might not be sufficient for high

expression of CTGF because the CTGF expression levels in non-tumorous liver tissues

of KrasG12D mice were not significantly different from those of control mice (Fig. 2D).

However, because the expression of CTGF was clearly increased in liver tumors that

developed in these mice, activation of the Ras signaling pathway would induce CTGF

expression in the tumor cells.

In the present study, we applied KrasG12D mice as a model of liver tumors. One of the

limitations of this model is the uncommon driver gene. While various gene mutations

were reported to be observed in human HCC [29], the frequency of Kras mutations is

not high and has been reported to be approximately 5-7% [30]. However, the Ras

signaling pathway is frequently activated in human HCC. For example, Ito Y et al.

demonstrated activation of MAPK/Erk in 58% of surgically resected human HCC [31].

Calvisi DF et al. reported increased phosphorylation of Raf, Mek, Erk and Akt

compared with surrounding non-tumorous liver tissues in all 35 cases of human HCC

examined [32]. Indeed, Raf kinase downstream of Ras is an important molecular target

of HCC, and sorafenib, a Raf kinase inhibitor, is approved by the FDA as a molecular

targeting agent for HCC. The Erk/Akt pathway was constitutively activated in the liver

tissues of KrasG12D mice, which is consistent with activation of the Ras signaling

pathway in human HCC. In agreement with the observations in human HCC, tumors in

these mice produced high levels of CTGF. Importantly, the hepatocyte-specific

knockout of CTGF significantly reduced the number and size of liver tumors in these

mice. Thus, CTGF produced by hepatocytes and hepatoma cells contributes to tumor

development and progression.

Hepatic fibrosis is not observed in KrasG12D mice, despite the fact that most HCCs

develop from fibrotic livers. Hepatic fibrosis is characterized by the excessive

accumulation of ECM in the liver [33]. ECM is mainly produced by HSCs, which

account for 5%-8% of the cells in the liver and are key regulators of liver fibrosis

[34-38]. Although the causative relationship between liver fibrosis and

hepatocarcinogenesis has been incompletely understood, there are abundant evidence

that liver fibrosis and HSCs contribute to the initiation of HCC [24, 33, 34]. Owing to

the absence of background hepatic fibrosis, KrasG12D mice might not be appropriate for

focusing on cancer initiation. Nevertheless, we consider this model is still useful for

analyzing the roles of CTGF in cancer cell-HSC interactions in HCC microenvironment,

since CTGF was up-regulated in cancer cells and activated HSCs were present in

tumors, similar to human HCC.

CTGF has been reported to promote DNA synthesis and cell cycle progression in

hepatoma cells [11, 13]. In our experiments, CTGF did not directly affect hepatoma cell

growth but accelerated it in the presence of HSCs in both culture and xenograft models.

HSCs also infiltrate tumorous tissues and exist as α-SMA-positive cells in

perisinusoidal areas of HCC tissues [36, 37]. Activated HSCs secrete several humoral

factors and accelerate tumor proliferation, invasion, and angiogenesis [24, 34-38].

Coulouarn et al. reported that the cross-talk between hepatoma cells and activated HSCs

promoted the production of VEGF and MMP9 from HSCs, which leads to vascular

angiogenesis and tissue remodeling [39]. In the present study, we revealed that CTGF

produced by tumor cells participated in HSC activation and HCC progression.

Furthermore, CTGF-mediated cross-talk between hepatoma cells and activated HSCs

induced IL-6 production from HSCs, which was involved in hepatoma cell growth.

Paracrine IL-6 production by inflammatory cells was reported to be deeply involved in

early hepatocarcinogenesis [40]. IL-6 has been reported to be produced by HSCs in the

HCC microenvironment, where it facilitates the progression of HCC [24].

CTGF-induced IL-6 secretion from HSCs might be one of the causes of accelerated

hepatoma cell growth. Meanwhile, our secretome analysis revealed that LX-2 cells

secreted several humoral factors other than IL-6 after the stimulation with recombinant

CTGF protein (Supplementary Table 2). Although our results indicated that IL-6 was

related to the CTGF-mediated interaction between cancer cell and HSCs, other

molecules might also be involved in this interaction.

The application of an anti-CTGF-neutralizing antibody alone did not affect the growth

of hepatoma cells (Fig. 5A). In contrast, an anti-CTGF-neutralizing antibody decreased

hepatoma cell growth in the presence of HSCs both in vitro and in xenograft models

(Fig. 5C and 5D). An anti-CTGF-neutralizing antibody was thus shown to prevent

CTGF-mediated tumor-stroma interactions between hepatoma cells and HSCs and to

successfully inhibit tumor growth. Several pharmacological inhibitors of CTGF that

block the biological action of CTGF have been explored [41]. An

anti-CTGF-neutralizing antibody (FG-3019) is capable of specifically binding to CTGF

with reasonable affinity and can exhibit an inhibitory effect [3]. An

anti-CTGF-neutralizing antibody has been demonstrated to be effective in animal

models of several malignant diseases, such as pancreatic cancer, ovarian cancer,

leukemia and melanoma, and a phase 2 clinical trial is currently being conducted in

pancreatic cancer patients [3, 6, 41-45]. The effectiveness of an anti-CTGF-neutralizing

antibody on HCC was previously unknown. Due to the tumor-facilitating actions of

HSCs, HSCs are currently considered a novel therapeutic target in HCC [24]. Our

results suggest that an anti-CTGF-neutralizing antibody might be a novel therapeutic

agent that can be used to inhibit the actions of HSCs in HCC.

In conclusion, CTGF produced by hepatoma cells exerts tumor-promoting effects

through tumor-stroma interactions between tumor cells and HSCs. Approaches using

CTGF-targeted biologics might be able to attenuate this effect.

Acknowledgments

This work was partially supported by a Grant-in-Aid for Scientific Research (to T.

Takehara. JP26670381 and JP26253947) and a Grant-in-Aid for Young Scientists (to Y.

Makino. JP17K15943) from the Ministry of Education, Culture, Sports, Science, and

Technology, Japan, a research grant from Bristol-Myers Squibb (to T. Takehara), and a

Grant-in-Aid for Research from the Japan Agency for Medical Research and

Development (JP18fk0210021, JP18fk0310108, JP18fk0210018 and JP18fk0210026).

The authors thank FibroGen for kindly providing the anti-CTGF-neutralizing antibody

FG-3019.

(4566 words from Introduction to Acknowledgments)

References

1. Jun JI, Lau LF. Taking aim at the extracellular matrix: CCN proteins as emerging

therapeutic targets. Nat Rev Drug Discov. 2011;10:945-63.

2. Hall-Glenn F, Lyons KM. Roles for CCN2 in normal physiological processes. Cell

Mol Life Sci. 2011;68:3209-17.

3. Lipson KE, Wong C, Teng Y, Spong S. CTGF is a central mediator of tissue

remodeling and fibrosis and its inhibition can reverse the process of fibrosis.

Fibrogenesis Tissue Repair. 2012;5(Suppl 1):S24.

4. Kodama T, Takehara T, Hikita H, Shimizu S, Shigekawa M, Tsunematsu H, et al.

Increases in p53 expression induce CTGF synthesis by mouse and human

hepatocytes and result in liver fibrosis in mice. J Clin Invest. 2011;121:3343-56.

5. Lai D, Ho KC, Hao Y, Yang X. Taxol resistance in breast cancer cells is mediated by

the hippo pathway component TAZ and its downstream transcriptional targets Cyr61

and CTGF. Cancer Res. 2011;71:2728-38.

6. Bennewith KL, Huang X, Ham CM, Graves EE, Erler JT, Kambham N, et al. The

role of tumor cell-derived connective tissue growth factor (CTGF/CCN2) in

pancreatic tumor growth. Cancer Res. 2009;69:775-84.

7. Jiang CG, Lv L, Liu FR, Wang ZN, Liu FN, Li YS, et al. Downregulation of

connective tissue growth factor inhibits the growth and invasion of gastric cancer

cells and attenuates peritoneal dissemination. Mol Cancer. 2011;10:122.

8. Wells JE, Howlett M, Cole CH, Kees UR. Deregulated expression of connective

tissue growth factor (CTGF/CCN2) is linked to poor outcome in human cancer. Int J

Cancer. 2015;137:504-11.

9. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA Cancer J Clin.

2016;66:7-30.

10. Mazzocca A, Fransvea E, Dituri F, Lupo L, Antonaci S, et al. Down-regulation of

connective tissue growth factor by inhibition of transforming growth factor beta

blocks the tumor-stroma cross-talk and tumor progression in hepatocellular

carcinoma. Hepatology. 2010;51:523-34.

11. Urtasun R, Latasa MU, Demartis MI, Balzani S, Goñi S, Garcia-Irigoyen O, et al.

Connective tissue growth factor autocriny in human hepatocellular carcinoma:

oncogenic role and regulation by epidermal growth factor receptor/yes-associated

protein-mediated activation. Hepatology. 2011;54:2149-58.

12. Zhang H, Li W, Huang P, Lin L, Ye H, Lin D, et al. Expression of CCN family

members correlates with the clinical features of hepatocellular carcinoma. Oncol

Rep. 2015;33:1481-92.

13. Xiu M, Liu YH, Brigstock DR, He FH, Zhang RJ, Gao RP. Connective tissue

growth factor is overexpressed in human hepatocellular carcinoma and promotes

cell invasion and growth. World J Gastroenterol. 2012;18:7070-8.

14. Akahoshi K, Tanaka S, Mogushi K, Shimada S, Matsumura S, Akiyama Y, et al.

Expression of connective tissue growth factor in the livers of non-viral

hepatocellular carcinoma patients with metabolic risk factors. J Gastroenterol.

2016;51:910-22.

15. Gressner OA, Gressner AM. Connective tissue growth factor: a fibrogenic master

switch in fibrotic liver diseases. Liver Int. 2008;28:1065-79.

16. Adler SG, Schwartz S, Williams ME, Arauz-Pacheco C, Bolton WK, Lee T, et al.

Phase 1 study of anti-CTGF monoclonal antibody in patients with diabetes and

microalbuminuria. Clin J Am Soc Nephrol. 2010;5:1420-8.

17. Raghu G, Scholand MB, de Andrade J, Lancaster L, Mageto Y, Goldin J, et al.

FG-3019 anti-connective tissue growth factor monoclonal antibody: results of an

open-label clinical trial in idiopathic pulmonary fibrosis. Eur Respir J.

2016;47:1481-91.

18. Toda N, Mori K, Kasahara M, Ishii A, Koga K, Ohno S, et al. Crucial Role of

Mesangial Cell-derived Connective Tissue Growth Factor in a Mouse Model of

Anti-Glomerular Basement Membrane Glomerulonephritis. Sci Rep. 2017;7:42114.

19. Ito M, Hiramatsu H, Kobayashi K, Suzue K, Kawahata M, Hioki K, et al.

NOD/SCID/gamma(c)(null) mouse: an excellent recipient mouse model for

engraftment of human cells. Blood. 2002;100:3175-82.

20. Tomayko MM, Reynolds CP. Determination of subcutaneous tumor size in athymic

(nude) mice. Cancer Chemother Pharmacol. 1989;24:148-54.

21. Kodama T, Takehara T, Hikita H, Shimizu S, Li W, Miyagi T, et al.

Thrombocytopenia exacerbates cholestasis-induced liver fibrosis in mice.

Gastroenterology. 2010;138:2487-98, 2498.e1-7.

22. Kawano S, Kojima M, Higuchi Y, Sugimoto M, Ikeda K, Sakuyama N, et al.

Assessment of elasticity of colorectal cancer tissue, clinical utility, pathological and

phenotypical relevance. Cancer Sci. 2015;106:1232-9.

23. Varma S, Stéphenne X, Komuta M, Bouzin C, Ambroise J, Smets F, et al. The

histological quantification of alpha-smooth muscle actin predicts future graft

fibrosis in pediatric liver transplant recipients. Pediatr Transplant. 2017;21. doi:

10.1111/petr.12834.

24. Coulouarn C, Clément B. Stellate cells and the development of liver cancer:

therapeutic potential of targeting the stroma. J Hepatol. 2014;60:1306-9.

25. Yoshiji H, Kuriyama S, Noguchi R, Yoshii J, Ikenaka Y, Yanase K, et al.

Angiopoietin 2 displays a vascular endothelial growth factor dependent synergistic

effect in hepatocellular carcinoma development in mice. Gut. 2005;54:1768-75.

26. Abou-Shady M, Baer HU, Friess H, Berberat P, Zimmermann A, et al. Transforming

growth factor betas and their signaling receptors in human hepatocellular carcinoma.

Am J Surg. 1999;177:209-15.

27. Zhu AX, Duda DG, Sahani DV, Jain RK. HCC and angiogenesis: possible targets

and future directions. Nat Rev Clin Oncol. 2011;8:292-301.

28. Neaud V, Faouzi S, Guirouilh J, Le Bail B, Balabaud C, Bioulac-Sage P, et al.

Human hepatic myofibroblasts increase invasiveness of hepatocellular carcinoma

cells: evidence for a role of hepatocyte growth factor. Hepatology. 1997;26:1458-66.

29. Fujimoto A, Furuta M, Totoki Y, Tsunoda T, Kato M, Shiraishi Y, et al.

Whole-genome mutational landscape and characterization of noncoding and

structural mutations in liver cancer. Nat Genet. 2016;48:500-9.

30. Pylayeva-Gupta Y, Grabocka E, Bar-Sagi D. RAS oncogenes: weaving a

tumorigenic web. Nat Rev Cancer. 2011;11:761-74.

31. Ito Y, Sasaki Y, Horimoto M, Wada S, Tanaka Y, Kasahara A, et al. Activation of

mitogen-activated protein kinases/extracellular signal-regulated kinases in human

hepatocellular carcinoma. Hepatology. 1998;27:951-8.

32. Calvisi DF, Ladu S, Gorden A, Farina M, Conner EA, Lee JS, et al. Ubiquitous

activation of Ras and Jak/Stat pathways in human HCC. Gastroenterology.

2006;130:1117-28.

33. Affo S, Yu LX, Schwabe RF. The role of cancer-associated fibroblasts and fibrosis

in liver cancer. Annu Rev Pathol. 2017;12:153-186.

34. Yin C, Evason KJ, Asahina K, Stainier DY. Hepatic stellate cells in liver

development, regeneration, and cancer. J Clin Invest. 2013;123:1902-10.

35. Friedman SL. Hepatic stellate cells: protean, multifunctional, and enigmatic cells of

the liver. Physiol Rev. 2008;88:125-72.

36. Enzan H, Himeno H, Iwamura S, Onishi S, Saibara T, Yamamoto Y, et al.

Alpha-smooth muscle actin-positive perisinusoidal stromal cells in human

hepatocellular carcinoma. Hepatology. 1994;19:895-903.

37. Faouzi S, Le Bail B, Neaud V, Boussarie L, Saric J, Bioulac-Sage P, et al.

Myofibroblasts are responsible for collagen synthesis in the stroma of human

hepatocellular carcinoma: an in vivo and in vitro study. J Hepatol. 1999;30:275-84.

38. Hernandez-Gea V, Toffanin S, Friedman SL, Llovet JM. Role of the

microenvironment in the pathogenesis and treatment of hepatocellular carcinoma.

Gastroenterology. 2013;144:512-27.

39. Coulouarn C, Corlu A, Glaise D, Guénon I, Thorgeirsson SS, Clément B.

Hepatocyte-stellate cell cross-talk in the liver engenders a permissive inflammatory

microenvironment that drives progression in hepatocellular carcinoma. Cancer Res.

2012;72:2533-42.

40. Naugler WE, Sakurai T, Kim S, Maeda S, Kim K, Elsharkawy AM, et al. Gender

disparity in liver cancer due to sex differences in MyD88-dependent IL-6 production.

Science. 2007;317:121-4.

41. Brigstock DR. Strategies for blocking the fibrogenic actions of connective tissue

growth factor (CCN2): From pharmacological inhibition in vitro to targeted siRNA

therapy in vivo. J Cell Commun Signal. 2009;3:5-18.

42. Lu H, Kojima K, Battula VL, Korchin B, Shi Y, Chen Y, et al. Targeting connective

tissue growth factor (CTGF) in acute lymphoblastic leukemia preclinical models:

anti-CTGF monoclonal antibody attenuates leukemia growth. Ann Hematol.

2014;93:485-92.

43. Moran-Jones K, Gloss BS, Murali R, Chang DK, Colvin EK, Jones MD, et al.

Connective tissue growth factor as a novel therapeutic target in high grade serous

ovarian cancer. Oncotarget. 2015;6:44551-62.

44. Dornhöfer N, Spong S, Bennewith K, Salim A, Klaus S, Kambham N, et al.

Connective tissue growth factor-specific monoclonal antibody therapy inhibits

pancreatic tumor growth and metastasis. Cancer Res. 2006;66:5816-27.

45. Neesse A, Frese KK, Bapiro TE, Nakagawa T, Sternlicht MD, Seeley TW, et al.

CTGF antagonism with mAb FG-3019 enhances chemotherapy response without

increasing drug delivery in murine ductal pancreas cancer. Proc Natl Acad Sci U S A.

2013;110:12325-30.

Table 1. Tumor characteristics of the CTGF high-expression and low-expression

groups.

CTGF CTGF

Characteristics Low-expression High-expression P-value

group group

AFP (negative/positive) 24/23 15/31 0.074

PIVKA-II (negative/positive) 16/31 7/39 0.024

Tumor number (1-5/6+) 43/4 34/12 0.025

Maximum tumor diameter (cm) 3.8 (0.8-20.0) 4.0 (1.0-18.0) 0.48

Macroscopic tumor classification 1/20/15/5/6 0/12/12/11/11 0.021 (SN-IM/SN/SN-EG/CMN/IF)

Portal invasion (negative/positive) 40/7 31/15 0.044

Intrahepatic metastasis 35/12 26/20 0.069 (negative/positive)

TNM stage (I/II/III/IVa/IVb) 8/24/7/8/0 4/18/14/9/1 0.082

AFP, α-fetoprotein; DCP, des-γ-carboxy prothrombin; SN-IM, small nodular type with

indistinct margin; SN, simple nodular type; SN-EG, simple nodular type with

extranodular growth; CMN, confluent multinodular type; IF, infiltrative type.

The cut-off values for AFP and DCP are 20 ng/mL and 40 mAU/mL, respectively.

Maximum tumor diameter are presented as the medians (ranges).

Figure Legends

Fig. 1. CTGF is upregulated in human HCC, and CTGF mRNA expression levels

are associated with those of α-SMA, a marker of HSC activation.

Liver tissues were obtained from HCC patients who underwent surgical hepatectomy.

(A) CTGF mRNA expression levels in tumorous tissues (T) and non-tumorous tissues

(NT) (N = 93). (B) Western blot analysis for the expression of CTGF in the livers of

representative five HCC patients and CTGF protein in the tumorous (T) and

non-tumorous tissues (NT) (N = 25). (C) Immunohistochemistry for CTGF expression

in HCC. A representative image from 10 HCC samples is presented. (D)

Immunohistochemistry for α-SMA in HCC and α-SMA mRNA expression levels in

tumorous tissues (T) and non-tumorous liver tissues (NT) (N = 93). A representative

immunohistochemical image from four HCC samples is shown. (E) Alpha-SMA mRNA

expression levels in tumorous tissues in the CTGF low-expression group and in the

high-expression group. CTGF-High; T/NT CTGF expression ratio > 1.3 (N = 46),

CTGF-Low; T/NT CTGF expression ratio ≤ 1.3 (N = 47). The Wilcoxon signed-rank

test was used to compare the expression levels between tumorous and non-tumorous

tissues in each patient (A, B, D). The horizontal bars denote medians. *p < 0.05.

Fig. 2. Hepatocyte-specific Kras-mutated mice (KrasG12D mice) develop liver

tumors in which CTGF is upregulated similarly to human HCC.

KrasG12D mice and their littermate Alb-Cre transgenic control mice were sacrificed at

the indicated age in months. (A) Representative images of the livers of 9-month-old

mice. (B) Macroscopic tumor incidence rate, tumor number, and maximum tumor

diameter according to age in months (N = 7-22 for each group). (C) Western blot

analysis for the expression of Ras, phospho-Erk, phospho-Akt, CTGF, α-SMA, and

β-actin in the livers of 9 to 11-month-old mice. (D) Gene expression levels of

glypican-3, afp and ctgf in tumorous liver tissues (T) and non-tumorous liver tissues

(NT) from KrasG12D mice and control mice from 9 to 11 months of age (N = 6-12 per

group). (E) Immunohistochemistry for CTGF in the liver tissues of 9-month-old mice. A

representative image from five samples is presented. (F) Immunohistochemistry for

α-SMA in liver tumors, α-SMA mRNA expression levels in tumorous liver tissues (T)

and non-tumorous liver tissues (NT) in KrasG12D mice and control mice (N = 7-10 per

group), and the correlation between α-SMA mRNA expression levels and CTGF mRNA

expression levels in tumorous tissues from KrasG12D mice (N = 20). Representative

immunohistochemical images from five samples are shown. In a box plot, box denotes

25th and 75th percentiles and top or bottom whiskers indicate the values higher than

75th percentiles or lower than 25th percentiles, respectively. In bar graphs, bars and top

whiskers represent mean values and standard deviations, respectively. *p < 0.05.

Fig. 3. Hepatocyte-specific CTGF deficiency suppresses the development and

progression of liver tumors in KrasG12D mice.

KrasG12D CTGF+/+, KrasG12D CTGF+/- and KrasG12D CTGF-/- littermates were sacrificed at

8 months of age, and the tumor characteristics were compared among these three

groups.

(A) Representative images of the livers. (B) Macroscopic tumorigenesis rate, tumor

number and maximum tumor diameter (N = 11-15 per group). (C) Representative

images of hematoxylin and eosin staining (×40) of the macroscopically non-tumorous

liver tissues and quantified tumor area / total tissue area ratio (N = 7-8 per group).

Arrows indicate tumors. (D) Representative immunohistochemistry images for PCNA

(×400) and the number of PCNA-positive cells per high-power field (HPF) in liver

tumors (N = 8-10 per group). (E) Gene and protein expression levels of CTGF in

tumorous liver tissues (N = 8-10 per group). (F) Representative immunohistochemistry

images for α-SMA (×400) and α-SMA-positive areas per HPF in liver tumors (N =

10-13 per group). In a box plot, box represents 25th and 75th percentiles and top or

bottom whiskers indicate the values higher than 75th percentiles or lower than 25th

percentiles, respectively. In bar graphs, bars and top whiskers indicate mean values and

standard deviations, respectively. *p < 0.05.

Fig. 4. Forced expression of CTGF does not affect the growth of PLC/PRF/5 cells

alone but increases their growth in the presence of LX-2 cells.

The growth of PLC/PRF/5 cells was evaluated in vitro using a WST-8 assay or

xenograft tumor models at the indicated time points. For xenograft models,

mock-transfected (PLC/PRF/5-mock) or CTGF-overexpressing PLC/PRF/5 cells

(PLC/PRF/5-CTGF), with or without LX-2 cells were subcutaneously injected into the

left and right flanks of NOG mice. The mice were sacrificed, and xenograft tumors were

enucleated 36 days after inoculation. Sequential tumor volume and representative

images of the xenograft tumors are presented. (A) Growth of parental, mock-transfected

(PLC/PRF/5-mock) or CTGF-overexpressing PLC/PRF/5 cells (PLC/PRF/5-CTGF).

(B) Xenograft model of mock-transfected or CTGF-overexpressing PLC/PRF/5 cells.

cells in co-culture with LX-2 cells. Mock-transfected PLC/PRF/5 cells or

CTGF-overexpressing PLC/PRF/5 cells were co-cultured with the same number of

LX-2 cells in a Transwell system. (D) Xenograft model of mock-transfected or

CTGF-overexpressing PLC/PRF/5 cells mixed with the same number of LX-2 cells. (E)

Immunohistochemistry for α-SMA in xenograft tumors composed of mock-transfected

or CTGF-overexpressing PLC/PRF/5 cells and LX-2 cells. Arrowheads indicate α-SMA

positive cells. N = 4 (A, C) or 9 (B, D) per group. In line graphs, plots with whiskers

indicate mean values ± standard deviations. *p < 0.05 vs the control.

Fig. 5. The growth of HepG2 cells is accelerated by co-culture with LX-2 cells,

which is abolished by inhibition of CTGF.

The growth of HepG2 cells was evaluated in vitro using a WST-8 assay or xenograft

tumor models at the indicated time points. For the co-culture experiments, HepG2 cells

were incubated in Transwell plates and co-cultured with the same number of LX-2 cells.

(A) Growth of HepG2 cells exposed to 100 ng/mL FG-3019 or hIgG in monoculture.

(B) Growth of HepG2 cells in monoculture or in co-culture with LX-2 cells. (C) Growth

of HepG2 cells in monoculture or in co-culture with LX-2 cells after exposure to 100

ng/mL FG-3019 or hIgG. (D) Xenograft model of HepG2 cells alone or HepG2 cells

with LX-2 cells. HepG2 cells alone and HepG2 cells mixed with the same number of

LX-2 cells were injected into the left and right flanks of mice. Inoculated mice were

divided into the hIgG and the anti-CTGF-neutralizing antibody administration groups.

From the day of inoculation, 40 mg/kg hIgG or FG-3019 was intraperitoneally

administered twice per week. The mice were sacrificed, and the xenograft tumors were

enucleated 31 days after inoculation. Representative images of enucleated tumors and

sequential tumor volumes are presented. N = 4 (A-C) or 5-6 (D) per group. In line

graphs, plots with whiskers represent mean values ± standard deviations. *p < 0.05 vs

the control.

Fig. 6. CTGF induces IL-6 production in HSCs to promote hepatoma cell growth.

The concentration of IL-6 in the culture supernatant was measured by ELISA after a

24-hour incubation. Cell growth was evaluated in vitro via a WST-8 assay. (A)

Concentration of IL-6 in the supernatant of HepG2 cells alone, LX-2 cells alone, and the

co-culture of HepG2 cells and LX-2 cells. The total volume of supernatant was

normalized among the three groups. (B) IL-6 concentration in the supernatant of LX-2

cells incubated with or without 5 nM recombinant CTGF protein (rhCTGF). (C) HepG2

cells were incubated in monoculture or in Transwell co-culture with LX-2 cells

transfected with IL-6 siRNA or control siRNA. Seventy-two hours before the start of

the co-culture, LX-2 cells were transfected with IL-6 siRNA or control siRNA. The cell

viability, IL-6 concentration in the supernatant, and protein expression in HepG2 cells

were evaluated 24 hours after the start of the co-culture. (D) HepG2 cells were

incubated in monoculture or in Transwell co-culture with LX-2 cells and exposed to 100

ng/mL FG-3019 or hIgG. The IL-6 concentration in the supernatant and protein

expression in HepG2 cells were evaluated after a 24-hour incubation. N = 4 per group.

In bar graphs, bars and top whiskers denote mean values and standard deviations,

respectively. *p < 0.05 vs the control.

CTGF mediates tumor-stroma interactions between hepatoma cells and hepatic stellate cells to accelerate HCC progression.

Yuki Makino, Hayato Hikita, Takahiro Kodama, et al.

Cancer Res Published OnlineFirst July 2, 2018.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-17-3844

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2018/06/30/0008-5472.CAN-17-3844.DC1

Author Author manuscripts have been peer reviewed and accepted for publication but have not yet Manuscript been edited.

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/early/2018/06/30/0008-5472.CAN-17-3844. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.