M-CSF Receptor Antagonists Inhibit the Initiation and Progression of Hepatocellular Carcinoma in Mice

Background/Aim: The aim of this study was to investigate the effects of the macrophage colony-stimulating factor (M-CSF) receptor antagonist on hepatic carcinogenesis in mice. Materials and Methods: Mice were injected with diethylnitrosamine (DEN) and treated with M-CSF receptor antagonist GW2580 (GW) or a saline vehicle just after (early treated group) or 2 weeks after (late treated group) DEN injection. Animals were sacrificed after 28 weeks and incidence of tumor was assessed. Isolated Kupffer cells were co-cultured with M-CSF in the presence or absence of GW, and the concentration of VEGF was measured. Results: The incidence of tumors was significantly blunted both in the early- and the late-treated groups. In addition, angiogenesis within the tumor was also suppressed in both groups. The concentration of VEGF increased in Kupffer cells treated with M-CSF compared to those cultured without M-CSF. This increase was blunted by GW. Conclusion: M-CSF and its receptor could be novel molecular targets for hepatocellular carcinoma.

Immunohistochemistry for CD31, F4/80, and CD206. Briefly, formalin-fixed, paraffin-embedded tissue specimens were cut into 4μm sections. Each section was mounted on a silane-coated glass slide, deparaffinized, and incubated in antigen retrieval solution for 40 min at 95˚C using Dako REAL Target Retrieval Solution (Dako, Carpinteria, CA, USA). Endogenous peroxidases were quenched by incubation at room temperature in 0.3% H 2 O 2 , followed by rinsing with PBS. Endogenous biotin was quenched using Dako Biotin Blocking System (Dako). Sections were blocked using 5% normal blocking serum for 20 min. Rabbit polyclonal antibody to CD31 (1:50, Abcam, Cambridge, UK) was applied for 2 h at room temperature. Anti-F4/80 primary antibody (1:100, Serotec, Oxford, UK) was applied for 30 min at room temperature. CD206 (Anti-Mannose Receptor antibody) (1:1,000 Abcam) was applied overnight at 4˚C. Immunoperoxidase staining was completed using a Vectastain ABC elite kit (Vector Laboratories, Burlingame, CA, USA) and diaminobentizine (DAB) as a chromogen. The number of positive cells against each antibody was assessed under microscopy in five random (400×) fields for calculation of the proliferation index. Histological samples were evaluated by one of the authors and by an outside expert in rodent liver pathology. Quantitative analysis of the CD31 positive area was calculated from five different fields and indicated as a percentage of the total area of the field using Image J software.
Statistical analysis. Data are expressed as mean±SEM. The Student's t-test was used for the determination of significance as appropriate. p<0.05 was considered significant.

Results
In animals administered with DEN, HCC was observed at 28 weeks after administration in the GW2580 non-treated group. On the other hand, tumor incidence was significantly blunted in both the early GW2580 administered and the late administered groups compared to the non-treated group  ( Figure 2A, B and C). Furthermore, the number and the maximum diameter of tumor were significantly blunted in both the early and the late GW2580 administered groups compared to the non-treated group ( Figure 2D and E). There were no significant differences between the early group and the late administered group.

Analysis of angiogenesis in intratumor and peritumor tissues after the administration of DEN.
To evaluate angiogenesis in the peritumor and intratumor tissues, immunohistochemical staining for CD31 was performed. The number of CD31-positive cells in the peritumor and intratumor tissues was significantly lower in both the early and late GW2580 administered groups than in the nontreated group. There were no significant differences between early administered group and late administered groups ( Figure 3A and B).
Analysis of the phenotype of macrophages in intratumor and peritumor liver tissues after the administration of DEN. The ratio of M2-type macrophages (CD206)/M1-type macrophages (F4/80) in the peritumor tissues was also significantly lower in both the early and late GW2580 administrated groups than in the non-treated group ( Figure 4A  Analysis of the population of hepatic macrophages after the administration of DEN. Double immunofluorescence staining indicated that the ratio of M2-type macrophages (CD163) / all macrophages (CD68) in the peritumor tissue was also significantly lower in both the early and the late GW2580 administered groups than in the non-treated group ( Figure  5A and B). There were no significant differences between the early and the late administered groups. Figure 6A). The concentration of VEGF increased significantly in cells cultured in the presence of M-CSF. This increase was significantly blunted in the presence of GW2580.

Effects of M-CSF on the proliferation of MH 134 cells in vitro.
There were no significant differences in the number of MH134 among the groups studied, suggesting that proliferation of MH134 was not affected by treatment with M-CSF in vitro ( Figure 6B).

Discussion
Angiogenesis is strongly associated with oncogenesis and tumor growth (12,(15)(16)(17). A previous study from our laboratory showed that macrophages activated by M-CSF   were involved in tumor initiation and progression (18,19). M2-type macrophages include Kupffer cell expressing the M-CSF receptor. It has been reported that M2-type macrophages activated by M-CSF induce neovascularization by increasing production of VEGF (6), which is an activator of angiogenesis. Importantly, VEGF production by isolated Kupffer cells was inhibited by GW2580 ( Figure 6A). These results indicated that M-CSF is involved in tumor progression by inducing production of VEGF by macrophages including Kupffer cells and monocytes.
In this study, occurrence of chemically induced HCC was significantly suppressed by treatment GW2580 in vivo, suggesting that the M-CSF receptor antagonists was effective in suppressing induction of hepatocarcinogenesis ( Figure  2E). The maximum diameter of the hepatic tumors was also significantly smaller in the GW2580 administered group ( Figure 2D), suggesting that GW2580 also inhibited tumor progression. After DEN administration, oxidative stress derived from hepatic macrophages has been shown to be involved in carcinogenesis in the liver (9,20,21). Thus, administration of GW2580 may lead to inhibition of activation of macrophages and reduction of oxidative stress.
GW2580 was not found to inhibit HCC cell proliferation, suggesting that it indirectly inhibits tumor growth ( Figure  6B). It has previously been reported that Kupffer cells stimulated by M-CSF promote proliferation of vascular endothelial cells by producing VEGF (6,13). In this study, VEGF production by isolated Kupffer cells was suppressed by GW2580 in vitro. After DEN administration, the ratio of M2-type/M1-type macrophages and the area of angiogenesis were reduced in both the early and the late GW2580 administered groups ( Figure 3A and B). Thus, M-CSF most likely acts on macrophages to promote production of VEGF, leading to angiogenesis and tumor growth.
The mechanism of induction of HCC involves chronic inflammation caused by viral hepatitis, alcohol and nonalcoholic steatohepatitis, and chemicals, suggesting that prevention of inflammation could be effective in inhibiting occurrence of HCC (22). In these inflammatory conditions, activated Kupffer cells play a critical role. In this study, using a chemically-induced HCC model, oral medication by M-CSF receptor antagonist inhibited activation of Kupffer cells and occurrence or progression of tumor. Since oral medication is appropriate and easy for long-term therapy than intravenous administration, it may be possible to treat HCC using this compound long-term.
In conclusion, M-CSF and/or its receptor may be novel molecular targets for the therapy of HCC.

Conflicts of Interest
The Authors declare that they have no conflict of interest. No Authors disclosed any financial conflicts for this manuscript, and there were no personal relationships with other people or organizations that may have potentially and inappropriately influenced our work or conclusions.

Authors' Contributions
Yoshihiro Akazawa contributed to the data collection and writing of the paper; Hiroshi Kono contributed to the conception and design of the study; Michio Hara contributed to critical revision of the article; Shinji Furuya contributed to critical revision of the article; Yuuki Nakata contributed to the analysis and interpretation of the study; Hiroyuki Wakana contributed to data collection; Hisataka Fukushima contributed to data collection; Chao Sun contributed to the data collection and writing of the paper; Hideki Fujii contributed to the conception and design of this study, and obtained funding; Daisuke Ichikawa contributed to the conception and design of this study, and obtained funding.