A Novel Orthotopic Liver Cancer Model for Creating a Human-like Tumor Microenvironment

概要

Introduction
Liver cancer is one of the leading causes of cancer deaths globally that accounts for approximately 810000 death annually, making it a major challenge for the global healthcare system (Fitzmaurice et a/., 2017). Hepatocellular carcinoma (HCC) is the most common form of liver cancer (Ferlay ef a/., 2010; Wang ef a/., 2014). Considering the prognosis of HCC remains very poor, therefore, more and more researchers are focusing on understanding of HCC pathogenesis.

Animal models have been long used to study cancer pathogenesis and as tool for drug screening. Because of the complicated etiology of cancer, the heterogeneity of cancer cells, and tumor microenvironment (TME) like fibrotic inflammatory liver, it is very difficult to develop animal model that can sufficiently mimic the developing process of human cancer. It is an urgent need to establish a novel, reliable, and robust HCC animal models that can be easily and rapidly produced, carry HCC with same genetic profiles as human HCC, and can mimic the normal and pathological liver TME in human.

Cancer is developed in its unique tissue environment called TME. TME is typically consists of tumor cells in the extracellular matrix, as well as a collection of cells such as fibroblast, endothelial cell (EC), mesenchymal cell (MC) and so on (Junttila and De Sauvage, 2013). These cells interact closely with each other and the surrounding tissue to create a microenvironment which has the optimal physical, molecular and cellular factors that support tumor growth.

In this study, we successfully improved our previous hiPSC-LB organoid technology and generated a functional sheet-like human HCC organoid in vitro. Then, we established a new orthotopic implantation method of a human HCC organoid which is capped by UPAL gel over the liver surface. It generated a novel human HCC mouse model within 2 weeks. By using this model, we demonstrated the roles of cellular TME within the tumor in promoting tumor growth. Furthermore, we also show our HCC mouse model are useful in testing the effect of fibrotic or NAFLD TME on HCC development. This model will be a useful tool for studying HCC pathogenesis and therapeutic studies.

Keywords:
Hepatocellular carcinoma, animal model, tumor microenvironment, fibrosis, ultra-purified alginate gel

Materials and Methods
1. Generation of the HCC organoid
The 100 |1L of cells suspension at a ratio of 10:2:2 (Huh7-luci: hiPSC-EC: hiPSC-MC) were applied into each well of the ibidi Culture-Insert 2 Well system (ibidi GmbH, Grafelfing, Germany) in the 6-well plate. The Huh7-Luci cells were seeded into each well of the ibidi Culture-Insert 2 Well system at a density of 5x 105 cells/ (unit). Three-dimensional HCC organoid was formed spontaneously after 1 days of culture.

2. HCC organoid implantation
The ultrasonic homogenizer (UH; Yamato Scientific, Tokyo, Japan) was used to disrupt the surface of the median lobe in the mouse liver. Next, the HCC organoids were implanted onto the surface of the median lobe in the mouse liver. After that, 0.5% UPAL (Mochida Pharmaceutical Co. Ltd. Tokyo, Japan) was applied over the implanted HCC organoids. At last,10% calcium chloride was added over the UPAL to promote the gelation of UPAL.

3. Inducing different tumor microenvironments
Liver fibrosis was induced in the 11-week-old NOD-SCID mice by injection of Thioacetamide (TAA) (Sigma-Aldrich, Darmstadt, Germany) into the peritoneal cavity (100 mg/kg) 3 times per week, for four consecutive weeks.
To induce NAFLD condition, 11-week-old NOD-SCID mice were fed with a Choline deficient L-amino acid-defined high fat diet (CDAHFD) (Research Diets Inc., New Brunswick, NJ, USA) for five consecutive weeks.
Results

1. Generation of HCC organoid from Huh7-Luci co-cultured with EC plus MC derived from human iPSC.
By modifying hiPSC-LB technology, we generated a functional HCC organoid that can reproduce TME including liver tumor cells expressing luciferase (Huh7-Luci), hiPSC-EC and hiPSC-MC. The ELISA assay results showed that the concentration of hALB secretion in the HCC organoid was significantly increased from day 2 to day 6. Our results showed that HCC organoid (Huh7-Luci/ hiPSC-EC/ hiPSC-MC) at a ratio of 10:2:2 has the highest luciferase signal, indicating that 10:2:2 is the best cell ratio for HCC organoid formation.

2. Implantation of HCC organoid onto the mouse liver surface.
The HCC organoid was either treated with UPAL gel capping (+UPAL) or without UPAL gel capping. Liver cancers were developed in all 6 mice in +UPAL group, but only 4 out of 6 mice in without UPAL group within 2 weeks. Overall, after HCC organoid attached to the mouse liver surface successfully, there was no significant difference in tumor size between +UPAL group and without UPAL group. H&E staining from both groups showed similar histology. Moreover, hiPSC-MC can be seen in the liver sections, however, hiPSC-EC cannot be observed.

3. The effect of cellular TME on the establishment of liver cancer mouse model
hiPSC-MC promotes the formation of HCC organoid in vitro. The addition of hiPSC-EC and hiPSC-MC promoted the formation of liver cancer in vivo, especially hiPSC-MC contributed to the higher engraftment. The tumor size in Huh7-Luci+EC+MC group are the largest in all groups.

4. The effect of fibrotic or NAFLD TME on HCC development
Tumor sizes in the mice implanted HCC organoid with induction of fibrosis by TAA injection (HO+TAA group) were significantly larger compared to those in HO group at 4-week of TAA injection. The fibrotic TME could promote tumor proliferation while liver fibrosis stayed in the same level during tumor growth. On the other hand, the tumor size in mice fed with CDAHFD (HO+CDAHFD 5w) is not significantly different compared to the mice fed with normal chow (NC)(HO+NC 5w).

5. HCC development using mice with humanized immune system
In order to mimic human cancer development, HCC organoids were implanted in the mice which have humanized immune system (humanized mice). Stereoscopical images of liver showed that HCC could be seen in NOG mice within 2w, however, HCC was hardly seen in the humanized mice for 2w. Moreover, comparing to the NOG mouse at 3w, HCC was much smaller in the humanized mouse.

6. HCC development using another liver cancer cell line (HepG2)
Using other liver cancer cell line HepG2 cells, liver cancer model was successfully developed, indicating that our new HCC development method could be generalized.

7. Generation of a highly metastatic pancreatic cancer model
By applying the method for development of HCC animal model to other tissue, we successfully generated a highly metastatic pancreatic cancer model. After orthotopic surgical implantation for 8w, pancreas cancer could be observed on the pancreas, and distant metastasis in liver lesions were also found in 2 out of 4 mice.

Discussion
The formation of the vascular endothelial network was not observed in our HCC organoid. This indicates that our cancer organoid model has a limitation for studying tumor angiogenesis in vitro. Interestingly, our control group had obvious vascular endothelial network formation in the same co-culture device. We speculate that the formation of vascular endothelial network in cancer cells might be induced by the release of exosomes or paracrine factors, rather than direct contact with cancer cells.
If UH is not used to form a wound over the liver surface, it will be difficult to see a luciferase positive signal on the mouse after HCC organoid implantation two weeks. We predict that the wound over the liver surface which made by UH could promote the exchange of oxygen and nutrients between the host wound surface and the implanted HCC organoid.
In +UPAL group, all mice showed HCC development, while the without UPAL group showed 66.7 % of mice developed HCC on their liver. The high engraftment rate in +UPAL group (100%) might because the UPAL gel helps implanted organoid to anchor to the liver surface, prevents it from moving to other places in the abdominal cavity. We did not observe significant difference in the engraftment area between without UPAL group and +UPAL group this time. Some reports have showed that UPAL gel can suppress lnteleukin-6 (IL-6) production in rabbit discectomy models (Ura et a/., 2021).IL-6 can facilitate tumorigenesis by regulating tumor cell surface markers and various signaling pathways (such as metabolism (White ef a/., 2013; Mauer, Denson and Bruning, 2015), apoptosis (Lin et al., 2001; Jourdan et al., 2003), survival (Grivennikov et al., 2009; Chang et a/., 2013), proliferation (Yeoh et a/., 2007; Wegiel et al., 2008), angiogenesis (Wei et al., 2003), invasion (Oh et a/., 2013; Zou, Zhang and Xu, 2016) and so on. We speculate that UPAL gel secretes some cytokines, such as suppression of IL-6 secretion in the TME, which inhibit the growth of cancer cells while increasing the engraftment rate of implanted cells.
When EC was added and co-cultured with Huh7-Luci in vitro, there was no significant increase in tumor cell number (Huh7-Luci). However, the tumor size after implantation the data in the vivo showed a significant difference. Interestingly, our immunofluorescent staining failed to detect hCD31,a human EC marker. This indicated that the growth of HCC tumor is not due to increase in the number of hiPSC-EC. We speculated that the hiPSC-EC might migrated from the tumor after implantation. Therefore, we further speculate that hiPSC-EC might promote tumor growth through distal paracrine secretion, instead of induce tumor cells growth through direct proximal cell contact.
Integrating the same HCC organoid on the surface of normal liver and fibrosis liver, we found a significant difference in the size of HCC, especially after TAA injection four weeks. Injecting TAA two or three weeks, however, there was no significant difference in tumor size compared with the control group without TAA injection. In our opinion, HCC lose contact with normal cells or tissues, and lose normal TME or ECM, which may further promote the development of HCC. However, if the degree of fibrosis is not particularly formed, this phenomenon of promoting the development of HCC is not obvious. Moreover, we found that incorporating HCC organoid between normal liver and NAFLD liver do not has significantly difference. After the mice ate CDAHFD food for five weeks, although the TG in the liver tissue increased, the severity of liver fibrosis was not obvious compared with TAA model. It seems that the degree of fibrosis promotes the development of HCC.
The luciferase signals were seen in all implanted mice after 2 weeks of implantation of HCC organoid in NOG group, however, luciferase positive signals were not detected in humanized mice until 3 weeks after HCC implantation. Compared to NOG mice, implanted HCC organoid in humanized mice showed delayed engraftment.
In this paper, we developed a new method of generating human tumors in vivo and examined the effects of TME and immunity on tumor formation. However, we did not show the more mechanism, such as, what kind of cytokines are released after the addition of hiPSC-EC, hiPSC- MC or fibrotic microenvironment, and which signal pathways related to liver cancer cell growth are activated by these released factors. A better understanding of the interaction between cancers and TME will may help us to find new ways to treat HCC.

この論文で使われている画像