Endothelial cells do not arise from tumor-initiating cells in human hepatocellular carcinoma
© Ghanekar et al.; licensee BioMed Central Ltd. 2013
Received: 26 July 2013
Accepted: 15 October 2013
Published: 20 October 2013
Conventional models of carcinogenesis suggest that tumors recruit blood vessel formation from normal host tissues. This concept has recently been challenged by prominent studies of glioblastoma, which suggest that intratumoral endothelial cells (ECs) may arise from cancer stem cells/tumor-initiating cells (TICs). Hepatocellular carcinoma (HCC) is a common, highly vascularized tumor with few effective therapies, against which anti-angiogenic strategies are being actively explored. TICs are felt to play a role in HCC pathobiology, but their contributions to tumor vasculature have not been studied.
We examined human HCCs in settings that selected for tumor formation from functionally defined TICs, and in which the origin of intratumoral ECs from TICs as opposed to host tissues could be clearly distinguished. We generated HCC nodules in the livers of immunodeficient mice by intrasplenic injection of HCC cells from cell lines and patient specimens and studied the tumor ECs by immunohistochemistry for mouse and human markers. We then used immunohistochemistry for EC markers in combination with fluorescence in situ hybridization (FISH) for X and Y chromosomes to study the endothelium of recurrent HCC specimens resected from sex-mismatched liver allografts of patients who had undergone liver transplantation for HCC.
We observed that all ECs in intrahepatic human HCC xenografts expressed mouse rather than human CD31. FISH analysis of recurrent HCCs resected from patients with sex-mismatched liver allografts revealed that all CD31+ and CD34+ intratumoral ECs originated from the donor allograft rather than the tumor.
These observations suggest that the vasculature of human HCC arises from normal host tissues rather than from TICs, supporting ongoing efforts to target angiogenesis in HCC as it is currently understood, and suggesting that the contribution of TICs to the vasculature of other cancers is disease-specific.
KeywordsHepatocellular carcinoma Endothelial cell Angiogenesis Cancer stem cell Tumor-initiating cell CD31 CD34 Xenograft Liver transplantation Recurrence
Conventional models of carcinogenesis posit that tumor vasculature develops in response to factors produced by neoplastic tissues (such as vascular endothelial growth factor) that elicit the sprouting of new blood vessels from surrounding pre-existing “normal” vasculature while also promoting the recruitment and differentiation of bone marrow-derived vascular progenitor cells . This concept has recently been challenged by the cancer stem cell (CSC) model, which suggests that tumors are sustained by a distinct population of cancer cells with stem cell-like properties that self-renew and are also capable of generating the full repertoire of cell types found within tumors, including those contributing to tumor vasculature . Most notably, two recent reports focused on glioblastoma have provided compelling evidence that intratumoral vascular endothelium arises from CSCs [3, 4].
Hepatocellular carcinoma (HCC) is one of the five most common cancers worldwide, and the third most common cause of cancer death . Early-stage HCC confined to the liver is amenable to potentially curative therapies such as ablation, partial hepatic resection, or liver transplantation, though disease recurrence is common . Patients with more advanced disease have a median survival less than one year due to the inefficacy of chemotherapy and radiation . HCC is a highly vascularized tumor, and this characteristic is exploited as a key diagnostic criterion on radiological imaging . Pathways and mediators regulating angiogenesis and neovascularization are thus recognized to play an important role in the development and progression of HCC and have emerged as attractive targets in the development of new therapies against this lethal disease .
Several studies suggest that CSCs, functionally assayed by their operational role as tumor-initiating cells (TICs), play a role in the pathogenesis of human HCC. Many different subpopulations of cells with tumor-initiating capacity in xenograft models have been isolated from HCC cell lines and patient tissues in recent years but their relative contributions to various aspects of HCC pathobiology remain to be clearly defined or understood [10–13]. Despite important implications for the correct targeting of new therapies against tumor vessel formation in HCC, the specific question of whether HCC vasculature arises from TICs or from surrounding tissues through “conventional” models of angiogenesis remains unexplored.
In this study, we sought to determine whether intratumoral endothelial cells in human HCC arise from TICs or from the surrounding non-cancerous tissue. We studied human HCC in biologically relevant in vivo models chosen for their ability to propagate tumors from single cells defined by their tumor-initiating function rather than a specific phenotypic cell surface marker. We first studied HCC tumor nodules generated in the livers of immunodeficient mice following intrasplenic injection of human HCC cells from established cell lines as well as patient tumors. We then examined human HCC tumors that developed in sex-mismatched liver allografts of patients who had undergone liver transplantation for treatment of HCC. We demonstrate that in both settings, endothelial cells (ECs) in human HCC do not arise from TICs.
Ethical approvals were obtained from the University Health Network Research Ethics Board (Protocol #08-0697-TE) and Animal Care Committee (Animal Use Protocol #1595). Human tissues were obtained with written consent from patients.
Human HCC cell lines
HepG2 cells were obtained from ATCC (catalog no. HB-8065) and grown in culture according to the supplier’s instructions. Instead of fetal bovine serum, growth medium was supplemented with 15% KnockOut SR XenoFree CTS serum replacement (Invitrogen). Huh7 cells were a kind gift from Dr. Paolo Parini (Karolinska Institute, Sweden) and were grown in DMEM/F12 media (Invitrogen) supplemented with 1X GlutaMAX (Invitrogen) and 15% KnockOut SR XenoFree CTS serum replacement (Invitrogen). Cells were harvested for xenografting with Accutase (StemCell Technologies) according to the manufacturer’s instructions.
Patient samples and preparation of tumor cell suspensions
HCC samples were obtained from patients undergoing surgery. Clinical data were obtained by chart review. Fresh tumor samples of approximately 1 cm3 were taken from the viable periphery of tumors within 30 minutes of resection and transferred on ice to the laboratory. Samples were sectioned for fixation in formalin, preservation in RNAlater (Invitrogen), or immediate digestion. Bulk tumor cell suspensions were prepared from patient samples by digestion of minced tumor fragments with 200 units/mL of type IV Collagenase (Sigma) and 0.8 units/mL of dispase II (Stem Cell Technologies) for 30-60 minutes at 37°C in an equal mixture of Dulbecco’s Modified Eagle Medium (DMEM) and F12 medium (both from Invitrogen), passage through a 70 μm cell strainer (BD Biosciences), centrifugation, and lysis of red blood cells using RBC lysis buffer (eBioscience). Human CD45+ cells and dead cells were excluded by fluorescence activated cell sorting on a FACSAria II cell sorter (BD Biosciences) after staining with anti-human CD45 PE-Cy7 (BD Biosciences) and Live/Dead Fixable Violet Dead Cell Stain Kit (Invitrogen) as per the manufacturer’s instructions; the viable CD45- fraction was then used for xenografting.
6-8 week old female NSG mice (non-obese diabetic severe combined immunodeficiency interleukin 2 receptor gamma chain null) (Jackson Laboratory) were anesthetized with inhalational isoflurane. Using aseptic surgical technique, a left subcostal incision was made and the spleen exposed. 106 HCC cells in 30 μL of cold Matrigel (BD Biosciences) were injected into the lower pole of the spleen using a 29-gauge needle. Upon removal of the needle the lower pole of the spleen was ligated with a silk ligature and the abdominal incision then closed with sutures. After recovery from surgery, animals were observed for 90 days and then sacrificed. Tissues were fixed in formalin or preserved in RNAlater. The xenografts analyzed in this study were generated from two different human HCC cell lines (HepG2 and Huh7) as well as from primary cells isolated from three different HCC specimens originating from three distinct patients. Xenografts were generated in duplicate in different mice from each cell line or patient sample and analyzed independently as described below.
Reverse transcription polymerase chain reaction (RT-PCR)
Ribonucleic acid (RNA) was isolated using TRIzol (Invitrogen) from tissues stored in RNAlater and then analyzed as previously described .
Immunohistochemistry and in situ hybridization
Formalin-fixed, paraffin-embedded tissues were utilized. Hematoxylin and eosin (H & E) stains were performed using standard techniques. For immunohistochemistry on human HCC xenografts and mouse liver tissue, dewaxed sections were blocked with 3% hydrogen peroxide, avidin/biotin blocking kit (Vector Labs), and 10% normal serum from the secondary antibody species, then incubated at room temperature with primary antibody overnight as follows: monoclonal mouse anti-human CD31 (Dako), polyclonal PECAM-1 antibody which recognizes both mouse and human CD31 (Santa Cruz), or monoclonal mouse anti-mouse H-2K (BD Biosciences). This was followed by biotin labeled secondary antibody (Vector Labs) for 30 minutes and horseradish peroxidase-conjugated ultrastreptavidin labeling reagent (ID Labs) for 30 minutes. Color was developed with DAB solution (Dako). Sections were counterstained with Mayer’s hematoxylin, dehydrated and mounted in Permount (Fisher). For analysis of human HCC specimens resected from sex-mismatched liver allografts, fluorescent immunohistochemistry was performed first using monoclonal mouse anti-human CD34 (Dako) or polyclonal goat anti-CD31 (Santa Cruz), in combination with monoclonal mouse anti-human CD45 (Dako) followed by Cy5 conjugated goat anti-rabbit or goat-anti-mouse IgG (Jackson ImmunoResearch Laboratories) or by Cy3 conjugated goat anti-mouse CD45 antibody (Jackson) according to manufacturer’s protocols. Counterstaining was performed with DAPI (Vector Labs). Images were captured with a CV-M4 + CL progressive scan monochrome camera (JAI) on a Zeiss Imager M1 using Metasystems workstation and ISIS software programs. Coordinates for all images were recorded. Slides were then washed in PBS, immersed in 100% ethanol and air dried. Fluorescence in situ hybridization (FISH) for X and Y chromosomes was performed using CEPX(DXZ1) SpO and CEPY(DYZ1) SpG Alpha Satellite DNA probes (Abbott Labs) as per the manufacturer’s recommendations. Dual color probes and target DNA were co-denaturated at 75°C for 5 minutes on HyBRITE followed by overnight hybridization at 37°C. Slides were washed, air dried and counterstained with DAPI. Using recorded coordinates, fields matching immunohistochemistry images were recaptured to evaluate staining of X and Y chromosomes in cells of interest by counting signals in at least 100 non-overlapping intact cells.
Results and discussion
Because of the likelihood that cell surface marker phenotypes of TICs vary between cell lines and among different patient tumors, we utilized the approach of intrasplenic injection of bulk tumor cells in order to select for HCC cells with tumor-initiating function in an unbiased manner rather than purify specific populations from the wide variety of putative HCC TICs described in the literature. By injecting cells in the spleen rather than directly into the mouse liver, we also hoped to avoid the potentially confounding contributions to xenograft vasculature from human ECs isolated along with bulk tumor cells from enzymatically digested HCC specimens. Furthermore, we believe this indirect approach yields intrahepatic tumor nodules derived from single HCC cells rather than from cell aggregates that might yield tumors from bulk tumor cell suspensions implanted directly into the mouse liver.
To validate this observation directly in primary human HCC tissues, we sought to study patients in whom the origin of intratumoral ECs could be clearly distinguished from surrounding malignant epithelial tissue without complex genetic analyses. Because orthotopic liver transplantation is routinely performed at our institution as definitive therapy for patients with HCC confined to the liver , we were able to identify three highly unique patients who had received liver allografts from sex-mismatched organ donors, subsequently developed a new HCC within the transplanted liver, and then had this recurrent HCC resected from the allograft. All of the patients identified were male and, following complete removal of their native livers, had received liver allografts from deceased organ donors who were female. The first patient underwent liver transplantation at age 56 for hepatitis C cirrhosis and a 5 cm moderately differentiated HCC; three years later, he underwent resection of a 9 cm moderately differentiated HCC in the right lobe of the liver allograft. The second patient underwent liver transplantation at age 66 for hepatitis C cirrhosis complicated by moderately differentiated multifocal HCC; he required resection of an 18 cm moderately differentiated HCC in the right lobe of the liver three years after transplantation. The third patient underwent liver transplantation at age 56 for hepatitis B cirrhosis and moderately differentiated multifocal HCC; he underwent resection of a 3 cm moderately differentiated HCC in the left lobe of the liver allograft three years after transplantation.
Limitations of this study include the ability of patient-derived xenograft models to accurately recapitulate all aspects of human cancer pathobiology, which have been criticized due to the immunodeficient state of the murine host and incompatibilities between murine and human factors in the tumor microenvironment . In order to prevent liver allograft rejection, the patients with HCC recurrences that we studied to validate our observations in xenografts were also receiving immunosuppressive therapy, which is known to facilitate cancer development and progression . In addition, all of the resected HCC recurrences that we studied by FISH were moderately differentiated tumors; due to the extremely unique characteristics of the patients from whom these samples originated, we were unable to study specimens that were either well differentiated or poorly differentiated. Although it is possible that endothelial cell origins are different in such tumors, we feel that the observations presented in this study provide valuable insights into the pathobiology of human HCC.
Taken together, our observations from two unique, biologically relevant, high-fidelity in vivo models of functional tumor-initiation strongly suggest that human HCC conforms to conventional models of tumor angiogenesis in which the tumor vasculature is recruited from surrounding “normal” tissues rather than generated through differentiation of TICs. This supports the notion that the biology of TICs may vary between different cancers, and that the observations prominently reported concerning glioblastoma may be specific to that disease. Reports validating a TIC origin of ECs in other cancer types have yet to emerge in the literature. While there is increasing attention to the potential contributions of CSCs/TICs to the pathobiology of human HCC and other tumors, our observations demonstrate that the roles of these cells likely vary between different cancers. The data presented herein firmly support the importance and relevance of many ongoing research efforts targeting pathways and mechanisms of tumor angiogenesis and neovascularization as they are currently understood  in order to develop novel therapeutic strategies for this frequently lethal cancer.
Cancer stem cell
Fluorescence in situ hybridization
- H & E:
Hematoxylin and eosin
Non-obese diabetic severe combined immunodeficiency interleukin 2 receptor gamma chain null
Reverse transcription polymerase chain reaction
Tumor initiating cell.
This research was supported by funding from The Cancer Research Society (AG) and Canadian Institutes of Health Research (AG). The authors wish to thank Olga Ludkovski and Melanie Peralta for technical assistance with tissue staining protocols, Dr. Paul Greig for assistance in identifying liver transplant patients with resected HCC recurrences, and Dr. John Dick for his support of this project.
- Hanahan D, Weinberg RA: Hallmarks of cancer: the next generation. Cell. 2011, 144 (5): 646-674. 10.1016/j.cell.2011.02.013.View ArticlePubMedGoogle Scholar
- Ping YF, Bian XW: Concise review: contribution of cancer stem cells to neovascularization. Stem Cells. 2011, 29 (6): 888-894. 10.1002/stem.650.View ArticlePubMedGoogle Scholar
- Ricci-Vitiani L, Pallini R, Biffoni M, Todaro M, Invernici G, Cenci T, Maira G, Parati EA, Stassi G, Larocca LM, et al: Tumour vascularization via endothelial differentiation of glioblastoma stem-like cells. Nature. 2010, 468 (7325): 824-828. 10.1038/nature09557.View ArticlePubMedGoogle Scholar
- Wang R, Chadalavada K, Wilshire J, Kowalik U, Hovinga KE, Geber A, Fligelman B, Leversha M, Brennan C, Tabar V: Glioblastoma stem-like cells give rise to tumour endothelium. Nature. 2010, 468 (7325): 829-833. 10.1038/nature09624.View ArticlePubMedGoogle Scholar
- El-Serag HB, Rudolph KL: Hepatocellular carcinoma: epidemiology and molecular carcinogenesis. Gastroenterology. 2007, 132 (7): 2557-2576. 10.1053/j.gastro.2007.04.061.View ArticlePubMedGoogle Scholar
- Bruix J, Sherman M: Management of hepatocellular carcinoma. Hepatology. 2005, 42 (5): 1208-1236. 10.1002/hep.20933.View ArticlePubMedGoogle Scholar
- Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, Blanc JF, de Oliveira AC, Santoro A, Raoul JL, Forner A, et al: Sorafenib in advanced hepatocellular carcinoma. N Engl J Med. 2008, 359 (4): 378-390. 10.1056/NEJMoa0708857.View ArticlePubMedGoogle Scholar
- Tang A, Cruite I, Sirlin CB: Toward a standardized system for hepatocellular carcinoma diagnosis using computed tomography and MRI. Expert Rev Gastroenterol Hepatol. 2013, 7 (3): 269-279. 10.1586/egh.13.3.View ArticlePubMedGoogle Scholar
- Pang R, Poon RT: Angiogenesis and antiangiogenic therapy in hepatocellular carcinoma. Cancer Lett. 2006, 242 (2): 151-167. 10.1016/j.canlet.2006.01.008.View ArticlePubMedGoogle Scholar
- Ma S, Chan KW, Hu L, Lee TK, Wo JY, Ng IO, Zheng BJ, Guan XY: Identification and characterization of tumorigenic liver cancer stem/progenitor cells. Gastroenterology. 2007, 132 (7): 2542-2556. 10.1053/j.gastro.2007.04.025.View ArticlePubMedGoogle Scholar
- Yamashita T, Ji J, Budhu A, Forgues M, Yang W, Wang HY, Jia H, Ye Q, Qin LX, Wauthier E, et al: EpCAM-positive hepatocellular carcinoma cells are tumor-initiating cells with stem/progenitor cell features. Gastroenterology. 2009, 136 (3): 1012-1024. 10.1053/j.gastro.2008.12.004.View ArticlePubMedGoogle Scholar
- Yang ZF, Ho DW, Ng MN, Lau CK, Yu WC, Ngai P, Chu PW, Lam CT, Poon RT, Fan ST: Significance of CD90+ cancer stem cells in human liver cancer. Cancer Cell. 2008, 13 (2): 153-166. 10.1016/j.ccr.2008.01.013.View ArticlePubMedGoogle Scholar
- Lee TK, Castilho A, Cheung VC, Tang KH, Ma S, Ng IO: CD24(+) liver tumor-initiating cells drive self-renewal and tumor initiation through STAT3-mediated NANOG regulation. Cell Stem Cell. 2011, 9 (1): 50-63. 10.1016/j.stem.2011.06.005.View ArticlePubMedGoogle Scholar
- Chen K, Ahmed S, Adeyi O, Dick JE, Ghanekar A: Human solid tumor xenografts in immunodeficient mice are vulnerable to lymphomagenesis associated with Epstein-Barr virus. PLoS One. 2012, 7 (6): e39294-10.1371/journal.pone.0039294.View ArticlePubMedPubMed CentralGoogle Scholar
- Lafreniere R, Rosenberg SA: A novel approach to the generation and identification of experimental hepatic metastases in a murine model. J Natl Cancer Inst. 1986, 76 (2): 309-322.PubMedGoogle Scholar
- Ahmed S, Zair M, Chen K, Iu M, He F, Adeyi O, Cleary S, Ghanekar A: Generation of subcutaneous and intrahepatic human hepatocellular carcinoma xenografts in immunodeficient mice. J Vis Exp. 2013, 79: e50544-PubMedGoogle Scholar
- DuBay D, Sandroussi C, Sandhu L, Cleary S, Guba M, Cattral MS, McGilvray I, Ghanekar A, Selzner M, Greig PD, et al: Liver transplantation for advanced hepatocellular carcinoma using poor tumor differentiation on biopsy as an exclusion criterion. Ann Surg. 2011, 253 (1): 166-172. 10.1097/SLA.0b013e31820508f1.View ArticlePubMedGoogle Scholar
- Cui S, Hano H, Sakata A, Harada T, Liu T, Takai S, Ushigome S: Enhanced CD34 expression of sinusoid-like vascular endothelial cells in hepatocellular carcinoma. Pathol Int. 1996, 46 (10): 751-756. 10.1111/j.1440-1827.1996.tb03544.x.View ArticlePubMedGoogle Scholar
- Kelly PN, Dakic A, Adams JM, Nutt SL, Strasser A: Tumor growth need not be driven by rare cancer stem cells. Science. 2007, 317 (5836): 337-10.1126/science.1142596.View ArticlePubMedGoogle Scholar
- Penn I: Post-transplant malignancy: the role of immunosuppression. Drug Saf. 2000, 23 (2): 101-113. 10.2165/00002018-200023020-00002.View ArticlePubMedGoogle Scholar
- Albini A, Tosetti F, Li VW, Noonan DM, Li WW: Cancer prevention by targeting angiogenesis. Nat Rev Clin Oncol. 2012, 9 (9): 498-509. 10.1038/nrclinonc.2012.120.View ArticlePubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2407/13/485/prepub
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.