Purpose: To allow the longitudinal investigation of molecular events associated with the progression of human hepatocellular carcinoma (HCC), we sought to develop a murine model by orthotopic implantation of tumor fragments obtained from patients diagnosed at early stage.

Experimental Design: Tumor pieces (2 × 2 mm) were implanted on the liver surface of nu/nu mice. After xenograft growing, subsequent passages were performed to achieve long-term implant viability. Isolation of tumoral hepatocytes was done to establish new cell lines. HCC characteristics, proliferation rate, apoptotic index (terminal deoxynucleotidyl transferase-mediated nick end labeling), and expression of cell-cycle regulators (cyclins E and A, p21Cip1, p27Kip1, p16INK4a, pRb, and p53) were assessed by Western Blot and immunohistochemistry, to correlate them with tumor progression.

Results: Five (50%) of the 10 primary HCCs resulted in small slow-growing liver implants. Three of them are viable after 48 months, whereas the remaining two survived for 15 and 13 months. Xenografts throughout passages exhibited a more aggressive phenotype with a poorer degree of differentiation, intense proliferation, moderate apoptosis, cell-cycle deregulation, p53 alterations, microvascular invasion, and dissemination. In one single passage, we observed critical growth delay, which was associated with significant p27kip1 overexpression. We established the anchor-free growing BCLC-9 cell line from one xenograft. This has gains of chromosomes 7, 5p, 6q, and 9q, is hepatitis B virus-DNA positive, does not secrete α-fetoprotein, and has TP53 missense mutations in codons 192 and 242.

Conclusions: The orthotopic implantation of early HCC fragments in nude mice provides a useful model to investigate the mechanisms of human HCC evolution and to establish new cell lines.

Hepatocellular carcinoma (HCC) mostly affects patients with viral cirrhosis (1, 2) and constitutes a worldwide major cause of cancer-related death (3, 4). At initial evolutionary stages, the tumor is well differentiated and the progression to more advanced stages with loss of differentiation and appearance of an invasive phenotype may require months or even years (5, 6). Interestingly, in a minority of patients, HCC can regress or even disappear because of unknown mechanisms (7). Tumor progression is achieved by the sequential accumulation of molecular alterations that bring the cells into a more transformed stage characterized by an increased proliferation. Cell proliferation is tightly regulated by the cell cycle machinery (8) that includes cyclins, cyclin-dependent kinases (CDK), and their inhibitors (CDK inhibitors) as p27Kip1, p21Cip1, and p16Ink4a, which are frequently altered in cancer (9). Enhanced vascularization, and acquired invasion and dissemination potential are also features gained during tumor progression (10). The elucidation of the molecular basis of these different evolutionary processes would require the careful longitudinal analysis of tumor tissues at different time points, and this is ethically unfeasible in humans. Accordingly, the development of animal models reproducing the evolution of human tumors is needed.

Models based on the injection of transformed cell lines may not accurately reproduce the human HCC biology. Similarly, the use of isolated cells obtained by disruption of the tumor may prompt cell death because of matrix detachment (11) and, thus, represent an artificial selection of specific cell clones. The implantation of solid tumor fragments with intact architecture may overcome some of these limitations (12). Small human tumor pieces can be placed directly on the target organ of an immunosuppressed animal and result in orthotopic implants (13, 14, 15). This approach has been successfully followed to develop experimental models for colon and pancreatic cancers (16, 17, 18). However, the sole reported attempt in liver cancer used tumor fragments from very advanced and disseminated tumors (19). Thus, it has been useful to study the mechanisms involved in a terminal tumor stage, but it offered no insight into the processes that regulate the biological evolution at earlier stages.

The present investigation assesses, first of all, the feasibility of developing an animal model derived from early human HCC and evaluates whether it reproduces the evolutionary characteristics observed in humans. Accordingly, intact specimens of human HCC at initial stages were orthotopically implanted in nude mice, and their progression was assessed pathologically and by registering proliferation activity and apoptotic rate, as well as the expression of cell cycle regulators (cyclin E, cyclin A, p21Cip1, p27Kip1, p16InK4a, pRb, and p53). Finally, on establishing successful liver implants (xenografts), we sought to use them to establish new human cell lines.

Human HCC Tissues

Ten fresh pieces of tumor tissue were obtained at the time of surgical resection or liver transplantation, in a series of nine patients diagnosed with early HCC (Table 1). All but two patients had underlying cirrhosis due to hepatitis B (HBV) or hepatitis C viruses (HCV). The two patients with a healthy liver had no serological markers of active viral infection. However, in one of them, HBV-DNA was found in the liver tissue, thus indicating occult HBV infection. In two patients (HC-3 and HC-7), small additional well and moderately differentiated tumors without evidence of vascular invasion were disclosed at pathological examination in explanted liver, probably representing synchronic tumors.

Tumor samples were obtained in sterile conditions avoiding apparently necrotic and fibrotic areas. Samples were cut into 2 × 2-mm pieces and were immediately placed in Dulbecco’s modified Eagle’s medium (Biological Industries, Kibbutz Beit Haemek, Israel) and Nutrient Mixture F-12 HAM (SIGMA Chemical Co., St. Louis, MO) supplemented with 50% fetal calf serum, 1% essential amino acids, 2 mml-glutamine, 50 units/ml penicillin, and 50 μg/ml streptomycin (all purchased from Biological Industries) and 1 mm pyruvic acid (SIGMA Chemical Co) until its implantation. Additional fragments were immediately frozen in liquid nitrogen, in which they were kept until analysis. The remaining fragments were fixed in formalin and embedded in paraffin for conventional pathological study and immunohistochemistry.

The Ethics Committee of the Hospital Clinic approved this study. Informed consent was requested from the patients before tissue collection. Animal-use and animal-care protocols were approved by the Ethics Committee of Animal Experimentation of Institut Català d’Oncologia and by the Animal Experimentation Commission of Generalitat de Catalunya according to National and European legal requirements.

Orthotopic Implantation in Nude Mice

Five-to-6-week-old male nu/nu Swiss mice, weighing 18–22 g (Iffa-Credo Animaux de Laboratoire, L’Abresle, France), were used for HCC implantation. Two pieces of each tumor were implanted in two different mice. The mice were anesthetized using i.p. injection of Avertin, 2,2,2-Tribromo-ethanol (SIGMA Chemical Co) at a dose of 0.3 mg/g. After performing a median laparotomy, the tumor fragment was anchored to the anterior hepatic lobe with a Prolene 6-0 suture. The abdominal incision was closed with staples. In the first passages, another tumor piece was placed s.c. after Matrigel soaking. Mice were inspected twice a week. Development of the successful liver implant was examined by an exploratory laparotomy under anesthesia at least every 6 weeks. Successive passages into new mice were done at least 3 months after implantation. At that time, the mice were sacrificed by cervical dislocation and were extensively explored to rule out the presence of hepatic or extrahepatic dissemination. At each passage, small fragments of the xenograft were snap-frozen or formalin fixed, and, if enough tissue was available, we used fragments to isolate tumoral hepatocytes. Blood was obtained by cardiac puncture.

Characterization of Xenografts

All of the primary HCC tumors and xenografts collected throughout successive passages were characterized in detail. The human origin of the passaged xenografts was confirmed by checking for microsatellite instability (20) within two markers, BAT26 and D12S95, which show distinct and specific band profiles between human and mouse samples.

Study of Cell Proliferation and Apoptosis.

Cell proliferation and protein expression of cell cycle regulators was assessed by Ki67 immunohistochemistry and Western blot, respectively, as published elsewhere (21). In one of the tumors (HC-2), we did not obtain enough protein yield. Fifty μg of protein, 25 μg for pRb Western, were separated electrophoretically in SDS-polyacrylamide gels and transferred onto Immobilon-P membranes (Millipore, Bedford, MA). To control protein loading, strips were incubated with 20% Ponceau S solution (SIGMA Chemical Co). Antigens were identified using polyclonal antibodies against cyclin A, p21Cip1, p16Ink4a, pRb (1:100 dilution except 1:200 for pRb; sc-751, sc-397, sc-468, and sc-050, respectively; Santa Cruz Biotechnology, Inc, Santa Cruz, CA), cyclin E (1:2500 dilution, 06-459; Upstate, Lake Placid, NY), and monoclonal antibody anti-p27Kip1 (1:200 dilution, sc-1641; Santa Cruz Biotechnology, Inc). To correct for potential contamination of mouse cells (stromal and blood cells), we assumed a 10% cell contamination. Hence, to properly assess the protein expression by Western blot, we added in each blot a lane containing 20% of mouse liver sample as a cross-reaction control. This did not show any significant band in any Western blots.

Apoptosis was estimated by measuring nuclear DNA fragmentation using the terminal deoxynucleotidyl transferase-mediated nick end labeling assay (22). p53 protein was studied by immunohistochemistry using monoclonal antibody to human p53 (1:50 dilution, BP53-12; Novocastra, Newcastle, United Kingdom). Papillary endometrial adenocarcinoma and 0.1% BSA in PBS instead of primary antibody were used for positive and negative control, respectively. In addition, Tp53 gene-mutations between exon 4 to 9 were analyzed by PCR using human intronic primers (23) and by direct sequencing.

HCV and HBV Detection.

The presence of HCV-RNA and HBV-DNA in serum and tissue was assessed by reverse transcription after PCR (24) and nested PCR (25), respectively.

Isolation and in Vitro Culture of Human Tumoral Hepatocytes

A xenograft piece was minced and then digested with 0.5 mg/ml collagenase type IV (SIGMA Chemical Co) at 37°C. Material was then placed in culture flasks containing supplemented Dulbecco’s modified Eagle’s and Nutrient Mixture F-12 HAM medium with 10% fetal calf serum in a humidified incubator at 37°C with 5% CO2 in air. Medium was changed 3 times a week. For subculture, 0.25% trypsin-EDTA (Biological industries, Kibbutz Beit Haemek Israel) was used. Cultured cells were frozen and stored in liquid nitrogen in 10% dimethylsulfoxide in fetal calf serum.

Characterization of Human HCC Cell Line

Determination of Doubling Time.

Cells (5 × 104) were seeded in 60-mm culture dishes. Viable cells determined by trypan blue exclusion were counted with a Neubauer chamber every day during 5 days after trypsin treatment. Population doubling time was calculated by plotting the mean cell number of three dishes versus days in culture during the exponential growth phase.

Cytogenetic Analysis.

The cytogenetic study was carried out on G-banded chromosomes obtained from 24-h nonstimulated cell culture after adding 10 μg/ml colchemid (Life Technologies, Inc., Grand Island, NY). Karyotype was described according to the International System for Human Cytogenetic Nomenclature (26). Comparative genomic hybridization (CGH) and fluorescence in situ hybridization (FISH) analysis were performed with VYSIS material (Downers Grove, IL) according to the manufacturer’s protocols. In CGH analysis, 1 μg of DNA from cell line and normal reference were labeled with different fluorescent dUTP by nick translation. Labeled probes were mixed and hybridized onto healthy male metaphase target slides. Repetitive sequences were blocked by COT-1 DNA. The average ratio profile was calculated based on the analysis of 16 selected metaphase images. Thresholds for gains and losses were set at 1.25 and 0.75, respectively. FISH analysis was done using WCP5, WCP9, and WCP13 probes. FISH and CGH images were captured, processed, and evaluated with a Cytovision ULTRA (Applied Imaging, Pittsburgh, PA).

Tumorigenecity Assay.

Cells (3.5 × 106) were injected intrahepatically and s.c. in the back at two separate sites in two anesthetized nude mice. Animals were inspected weekly and sacrificed 1 month after injection. Developed tumors were embedded in paraffin for pathological study.

Anchorage-Independent Growth Assay.

This was performed in three six-well dishes by suspending 5 × 104 cells in 0.3% agar in complete medium and overlaying the suspension on presolidified 0.5% agar. Cells were incubated at standard conditions for 3 weeks.

Study of Cell Proliferation and Cell Cycle Regulators.

Ki67 and p53 immunohistochemistry and Western blot of cell cycle regulators were assessed in nonsynchronic nonconfluent cells as described previously. For Western blot analysis, cells were lysed in 80 mm Tris-HCl (pH 6.8) with 2% SDS. Lysates were then sonicated twice for 10 s, and protein content was estimated by Lowry procedure.

Orthotopic Implantation in Nude Mice

Successful hepatic grafting with growth of a small tumor nodule with well-defined limits and less than 1 cm in diameter was registered in five implants (Table 2; Fig. 1, A and B). Hepatic grafting coincided with the growth of the s.c. seed. Successful implants were obtained from patients with HBV or HCV infection, and also from tumor developed in a normal liver without evidence of viral infections. Three of the xenografts (HC-3, HC-7A, and HC-9) are still viable throughout 48, 33, and 32 months involving 12, 9, and 7 passages, respectively. The other two xenografts (HC-2 and HC-5) survived for 15 (3 passages) and 13 (3 passages) months, respectively. The successful grafting of the xenografts was not related to any of the clinicopathological or molecular primary tumor characteristics depicted in Table 1.

Local and distant dissemination was observed in two perpetuated xenografts (Table 2). HC-3 and HC-7A displayed several satellite nodules with evidence of vascular invasion. A later passage of HC-3 showed lymph node metastasis that was orthoimplanted in new mice, resulting in xenograft HC-3Mx.

Characterization of Orthotopic Implanted Tumors

Microsatellite analysis confirmed the human origin of the xenografts and did not disclose instability in any sample (data not shown).

Pathological Study.

This analysis revealed that the xenografts at initial passages did not retain the architectural disposition of corresponding primary HCCs and showed a more undifferentiated cell pattern (Fig. 1,C) with some necrotic areas. These malignant hepatocytes exhibited an immature appearance. Throughout successive passages, xenografts maintained a solid and acinar cell pattern (Table 2) with increased mitotic index.

Study of Cell Proliferation and Apoptosis.

Ki67 labeling index of the primary tumors was heterogeneous ranging from 0 to 21.6% (Table 1) and relatively low because only two cases exceeded the accepted 20% cutoff used to classify tumors as highly proliferating. Proliferation activity of xenografts (Table 2) was increased with respect to their primary counterparts (Table 1) at first passages and reached almost 100% in the last passages. The increased proliferation in xenografts was associated with an altered expression of cell cycle regulators with respect to their nontumor counterparts, suggesting a cell cycle deregulation (Fig. 2). The expression pattern of positive regulators of cell cycle (cyclin E and cyclin A) in most of perpetuated xenografts increased throughout successive passages. Although the expression of negative regulators (p21Cip1, p27Kip1, p16Ink4a) was reduced in several passages, in others their expression was increased. In this sense, in HC-9 at passage 3, we detected an overexpression of p27Kip1 and an induction of p21Cip1 expression with respect to the previous passage, this coinciding with a less aggressive tumor progression; only one of the two grafted mice resulted in a very tiny viable xenograft, which had a very slow growth curve that was maintained in ulterior passages. These expressed CDK inhibitors were functional and able to inhibit CDK2, because we observed a simultaneous decrease of CDK2 total kinase activity (18-fold less) with respect to the previous passage without p27Kip1 overexpression and in the absence of p21Cip1 (data not shown).

Primary tumors showed 2–5% of terminal deoxynucleotidyl transferase-mediated nick end labeling-positive cells, whereas this increased to 7–8% in xenografts. p53 nuclear expression was always ≤5% in primary tumors and was considered negative. In contrast, in the different xenografts analyzed (early and late passages) a positive nuclear staining was detected in ≥30% of tumor cells. None of the primaries and xenografts exhibited abnormalities in the Tp53 gene except for HC-9 xenograft. This showed two previously reported mutations (27) in exon 6 (codon 192, Gln-Arg) and exon 7 (codon 242, Cys-Tyr) in both early and late passages.

HBV and HCV Detection.

Sera of mice were constantly negative for both HBV-DNA and HCV-RNA. Whereas HBV-DNA was detected in all surviving xenografts of HBV-related patients, HCV-RNA was not maintained in the xenografts.

Establishment and Characterization of Human HCC Cell Line BCLC-9

A long-term viable cell line, named BCLC-9, has been established from xenograft HC-9 and is still maintained 24 months after its obtaining. Plated cells present polygonal and epithelial-like shape under phase-contrast microscope (Fig. 3). They grow as a monolayer or as colonies. Anchorage-independent growth was confirmed when these cells, plated in soft agar, were able to form macroscopic colonies in 2 weeks. Doubling time of BCLC-9 cells at passage 17 was 27.7 h. Karyotype analysis confirmed the human origin, and the hepatocyte lineage was confirmed by cytokeratin profile. The BCLC-9 cell line is VHB-DNA positive, and does not secrete α-fetoprotein. In addition, BCLC-9 cells expressed E-cadherin, vascular endothelial growth factor, inducible-nitric oxide synthase, and showed an overexpression of Skp2 (data not shown).

Cytogenetic Analysis.

The 13 analyzed metaphases of BCLC-9 cells at passage 7 showed a complex karyotype with trisomy of chromosome 7, loss of chromosome 13, and two marker chromosomes, mar1 and mar2 (Fig. 4 A). In addition, chromosome 6 was longer in 60% of the metaphases, whereas tetrasomy of chromosome 7 was observed in only 15%.

CGH and FISH analysis were performed to characterize the two marker chromosomes. CGH showed a partial gain of chromosome 5 (p11-pter), chromosome 6 (q24-qter), and chromosome 9 (q22.3-q34), and a whole gain of chromosome 7, confirming the trisomy and tetrasomy observed previously (Fig. 4 B). FISH findings disclosed that mar1 was composed of the missing chromosome 13 and of the amplified fragment of chromosome 9, whereas mar2 was the result of the short-arm amplification of chromosome 5 (data not shown). These chromosome alterations were maintained at passage 36 after 16 months of in vitro culture.

Tumorigenecity Assay.

Tumoral nodules measuring 0.36 ± 0.11 cm developed in 3 of 4 s.c. injection sites 1 month after BCLC-9 cell injection. Pathological analysis showed a human HCC-like pattern mimicking xenograft HC-9.

Study of Cell Proliferation and Cell Cycle Regulators.

The Ki67 labeling index in BCLC-9 cells was 84.6%, indicating a high level of proliferating activity that coincided with the analysis of the cell cycle regulators by Western blot analysis. BCLC-9 cells showed a loss of p16Ink4a and an expression profile of cell cycle regulators characteristic of proliferating cells: high levels of cyclins E and A and low levels of p27Kip1 and p21Cip1 (Fig. 2). Moreover, cells express p53 protein by immunocytochemistry and presented the same Tp53 gene mutations as reported in the HC-9 xenograft from which they were derived.

This study indicates that orthotopic implantation of histologically intact tissue of early HCC obtained by surgical resection constitutes not only a useful experimental model to study events associated with tumor progression but also a help in the establishment of new cell lines. Previous investigations in other tumors have shown this potential, but the only data in HCC reported thus far have been obtained from samples derived from patients diagnosed at a very advanced tumor stage (19). The 50% initial take rate compares with the data in other neoplasms and the 30% long-term viability should be considered acceptable because of the early nature of the tumors used. Given the fact that early HCCs are usually well differentiated (5), they will keep several functions of normal hepatic cells, such as apoptotic death due to the stress associated with changes in tumor microenvironment. In that sense, several authors have shown that surgical maneuvers may sharply change the activation of several signal transduction pathways (28, 29, 30, 31). This reaction to tissue handling cannot be accurately controlled or prevented and, ultimately, may partially explain the failure to identify any specific pathological or molecular characteristics associated with initial and long-term success. Thus, successful grafting has not been related to the viral etiology of the underlying liver disease, which, in HBV patients, would be potentially mediated by the genetic changes related to the integration of HBV-DNA into the genome of the hepatocytes (1).

All of the primary tumors that we used belonged to an early evolutionary stage; this was evident because they retained hepatocytic differentiation, showed no evidence of vascular invasion, had a low-to-moderate proliferation rate, and did not exhibit p53 abnormalities, which in the Western world are mostly detected at an advanced stage (32, 33, 34). The evolution of xenografts reproduces the biological progression of human HCC at early stage. They need long periods to reach significant size, have low dissemination potential (in early passages, only one of the five initial successful grafts resulted in intrahepatic dissemination), and, in some cases, the progression might be halted by unknown mechanisms, occasionally leading to tumor regression (as in cases HC-2 and HC-5). However, it has to be noted that xenografts at first passages lose their differentiation degree, show a high level of proliferating activity, and express p53, which in one case was associated with two Tp53 gene mutations. These characteristics reflect mostly an advanced tumor stage and suggest that this model probably selects the more undifferentiated clones that have growth and survival advantages and that can exist even in tumors classified as early stage (35, 36). If not resected, these undetectable and undifferentiated clones would progressively evolve into overt cancer, finally leading to a typical advanced stage. These more aggressive cells would be able to resist extreme conditions such as loss of stimulation by cytokines and growth factors, and low oxygen concentration, this last capacity being at least partially related to p53 loss of function (10, 37).

Our data clearly show that the expression of cell cycle regulators is modified throughout successive passages. Thus, as expected, tumor growth and increased proliferation were associated clearly with an increase of positive cell cycle regulators. Accordingly, the high levels of cyclins E and A that were observed in xenografts unequivocally reflect a proliferating status, because they would not be expressed when cell cycle is arrested. The CDK inhibitor levels in xenografts did not show a consistent pattern and varied from decreases to evident elevations, which could be the result of the activation of growth-inhibitory pathways at early evolutionary stages of HCC. In this sense, the most striking event has been observed in the case HC-9 at passage 3, which presented a sharp increase of p27Kip1 expression and a lesser increase of p21Cip1. This coincided with a clear-cut reduction of its progression rate that we attributed mainly to the effects of p27Kip1. This is supported by the results of CDK2 activity, which demonstrate that this overexpressed CDK inhibitor was functional and inhibited CDK2-cyclin E and CDK2-cyclin A complexes. We and others (21, 38, 39, 40) have shown that HCC with high p27Kip1 expression have a better prognosis, and, thus, all these data strongly indicate that this protein is involved in an effective inhibitory pathway that may abrogate aggressive tumor progression at early stages. Two additional xenografts showed significant increases in p27Kip1 expression. HC-5 was not perpetuated, suggesting the possibility that p27Kip1 levels blocked xenograft viability. By contrast, HC-7A showed high p27Kip1 levels and resulted in a viable xenograft with high proliferative index. This observation might be due to the coexistence of several cell populations within the tumor (41), whose balance within the graft would determine its ultimate fate.

An intriguing observation that deserves some comment is the apparent discrepancy between high proliferation indexes and slow growth of the xenografts. These showed areas of necrosis and a slightly increased apoptotic rate as compared with primary HCC. Clearly, this should be the consequence of the lack of proper vessel formation to nourish the grafts. Early after implantation, the poor blood supply with hypoxic cell damage should prompt necrosis and activate apoptotic mechanisms, in which p53 could be involved if not mutated. This would initially shift the balance between proliferation and death toward a slow progression rate, whereas at more advanced stages proliferation would dominate and allow an increase in macroscopic tumor burden.

In addition to the confirmation of the orthotopic model as a viable option for translational research, we have been able to establish the BCLC-9 cell line from graft HC-9, and two additional lines are in the process of being established. The cytogenetic analysis of BCLC-9 cells showed a whole gain of chromosome 7 and partial gains of 5p, 6q, and 9q. These amplifications have been stable over a year in culture, and the affected chromosome regions contain an important number of genes (12 proto-oncogenes, Skp2, telomerase reverse transcriptase, hepatocyte growth factor, and c-met) that may contribute to the malignant phenotype. The observed chromosome 7 polysomy coincides with the findings in the few cell lines of European origin: HBGs and HepaRG (42, 43), indicating similarities among cell lines with the same geographic area. Furthermore, the reported gains of BCLC-9 cells have been less frequently observed, and those that have been more frequently described (44, 45) have not been detected. This represents an advantage of the new cell BCLC-9 cell line. Thus, this new cell line and, presumably, the cell lines that will be derived from the other xenografts will enlarge the currently reduced number of European human HCC cell lines, which surely represent a specific genetic background as compared with those derived from Asian patients.

In summary, we have developed an orthotopic model of human HCC in nude mice that reproduces the biological evolution of human HCC. This model constitutes a useful tool to study the molecular basis of the transition from early to advanced carcinomas and represents excellent source material from which to derive cell lines. Both tools should be valuable in the study of HCC progression as well as in the identification and testing of new therapeutic targets both in vitro using cell lines and, thereafter, in vivo using the murine model.

Grant support: Supported by grants from the Comisión Interdepartamental de Ciencia y Tecnología (SAF 98–004, SAF 00–81, SAF 99–028, and FIS 01/1264), and by a research grant from Leo-Pharma Chemicals, and by grants from the Instituto de Salud Carlos III (C03/10 and C03/09). C. Armengol was supported by research grants from the Ministerio de Ciencia y Tecnología and from the Fundació Rius i Virgili.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Notes: C. Armengol and G. Tarafa contributed equally to this study.

Requests for reprints: Jordi Bruix, BCLC Group, Liver Unit, Hospital Clinic, Villarroel 170, 08036 Barcelona, Spain. Fax: 34-932279803; E-mail: jbruix@clinic.ub.es

Fig. 1.

Successful orthotopic implant in the mouse liver (A). The xenograft is recognized as a solitary nodule protruding from the liver surface (B). Microscopic picture of HC-3 xenograft, which exhibits a lesser differentiation degree with immature cells and high mitotic activity compared with its primary tumor, H&E; original magnification, ×200 (C).

Fig. 1.

Successful orthotopic implant in the mouse liver (A). The xenograft is recognized as a solitary nodule protruding from the liver surface (B). Microscopic picture of HC-3 xenograft, which exhibits a lesser differentiation degree with immature cells and high mitotic activity compared with its primary tumor, H&E; original magnification, ×200 (C).

Close modal
Fig. 2.

Western blot of the cell cycle regulators [cyclin E (cycE) and A (cycA), p21Cip1, p27Kip1, p16Ink4a, pRb] in cases HC-3, HC-5, HC-7, and HC-9 and in the established human BCLC-9 cell line. The mouse liver sample used as a cross-reaction control did not show any significant band. Nontumoral surrounding liver (C); primary hepatocellular carcinoma (T); s.c. xenograft (sc1); xenograft at sequential hepatic passages (H1, H2, H3). For patient HC-7, two nodules used for grafting (Ta and Tb). Only Ta implant showed growth in nude mice.

Fig. 2.

Western blot of the cell cycle regulators [cyclin E (cycE) and A (cycA), p21Cip1, p27Kip1, p16Ink4a, pRb] in cases HC-3, HC-5, HC-7, and HC-9 and in the established human BCLC-9 cell line. The mouse liver sample used as a cross-reaction control did not show any significant band. Nontumoral surrounding liver (C); primary hepatocellular carcinoma (T); s.c. xenograft (sc1); xenograft at sequential hepatic passages (H1, H2, H3). For patient HC-7, two nodules used for grafting (Ta and Tb). Only Ta implant showed growth in nude mice.

Close modal
Fig. 3.

Phase-contrast microscopy image of BCLC-9 cells. ×200.

Fig. 3.

Phase-contrast microscopy image of BCLC-9 cells. ×200.

Close modal
Fig. 4.

BCLC-9 cell shows a complex G-banded karyotype with trisomy of chromosome 7, loss of chromosome 13 and two marker chromosomes, mar1 and mar2 (A). Comparative genomic hybridization findings of BCLC-9 cell line (B) show a partial gain in chromosome 5 (p11-pter), chromosome 6 (q24-qter), and chromosome 9 (q22.3-q34), and a whole gain of chromosome 7.

Fig. 4.

BCLC-9 cell shows a complex G-banded karyotype with trisomy of chromosome 7, loss of chromosome 13 and two marker chromosomes, mar1 and mar2 (A). Comparative genomic hybridization findings of BCLC-9 cell line (B) show a partial gain in chromosome 5 (p11-pter), chromosome 6 (q24-qter), and chromosome 9 (q22.3-q34), and a whole gain of chromosome 7.

Close modal
Table 1

Clinical and pathologicala characteristics of patients with hepatocellular carcinoma (HCC) from which orthotopic implants were derived

Patient codeSex/ Age (yr)Underlying diseaseTrbAFPTumor size (cm)Differ. DegreeCell patternKi67 LIc (%)No. of HCCsVascular invasion
HC-1 M/63 HCV liver cirrhosis Trabecular-acinar 8.1 No 
HC-2 M/52 Healthy liver 3.4 W to Mod Trabecular-acinar 0.2 No 
HC-3 F/57 HCV liver cirrhosis LT W to Mod Trabecular-acinar 3.3 No 
HC-4 F/65 HCV liver cirrhosis W to Mod Trabecular 11.6 No 
HC-5 F/71 HBV liver cirrhosis Mod Trabecular 21.3 No 
HC-6 F/69 HCV liver cirrhosis Trabecular 16.7 No 
HC-7d M/60 HBV liver cirrhosis LT 5 (A) Mod Trabecular-acinar No 
     2 (B) Mod Macrotrabecular 3.8 No 
HC-8 M/56 HCV liver cirrhosis LT 5.5 Mod Macrotrabecular-pseudoacinar 7.6 No 
HC-9 F/70 HBV Healthy liver Microtrabecular-pseudoacinar 21.6 No 
Patient codeSex/ Age (yr)Underlying diseaseTrbAFPTumor size (cm)Differ. DegreeCell patternKi67 LIc (%)No. of HCCsVascular invasion
HC-1 M/63 HCV liver cirrhosis Trabecular-acinar 8.1 No 
HC-2 M/52 Healthy liver 3.4 W to Mod Trabecular-acinar 0.2 No 
HC-3 F/57 HCV liver cirrhosis LT W to Mod Trabecular-acinar 3.3 No 
HC-4 F/65 HCV liver cirrhosis W to Mod Trabecular 11.6 No 
HC-5 F/71 HBV liver cirrhosis Mod Trabecular 21.3 No 
HC-6 F/69 HCV liver cirrhosis Trabecular 16.7 No 
HC-7d M/60 HBV liver cirrhosis LT 5 (A) Mod Trabecular-acinar No 
     2 (B) Mod Macrotrabecular 3.8 No 
HC-8 M/56 HCV liver cirrhosis LT 5.5 Mod Macrotrabecular-pseudoacinar 7.6 No 
HC-9 F/70 HBV Healthy liver Microtrabecular-pseudoacinar 21.6 No 
a

Defined according to pathology examination.

b

Tr, treatment; HCV, hepatitis C virus; HBV, hepatitis B virus; R, surgical resection; LT, liver transplant; AFP, α-fetoprotein; N, normal levels (≤20 ng/ml); E, elevated levels (>20 ng/mL); Differ., differentiation; W, well differentiated; Mod, moderately differentiated; LI, labeling index.

c

Number of cells stained/total cells analyzed in Ki67 immunohistochemistry.

d

Data of patient HC-7 reflect the characteristics of the two nodules (A and B) used for grafting.

Table 2

Yield of implantation, pathological characteristics, proliferating activity, and dissemination patterns of human hepatocellular carcinoma (HCC) xenografts in nude mice

Xenograft codeShort-term yieldLong-term yieldViability (mo)No. of passagesTumor sizea (cm)Differ.b DegreeaCell patternaKi67 LIa,c (%)Vascular invasionaDissemination pattern
HC-1 No          
HC-2 Yes No 15 0.40/0.36 M/ND Acinar/ND 65.3/ND No/ND  
HC-3 Yes Yes 48 12 0.45/0.47 M to P/P Acinar-solid/solid 29.5/98.6 Yes/Yes Intrahepatic, lymph node 
HC-3Mxd Yes Yes 16 0.41/0.38 M/ND Solid/ND 85.1/ND Yes/ND  
HC-4 No          
HC-5 Yes No 13 0.35/0.30 M/ND Solid/ND 48.1/ND No/ND  
HC-6 No          
HC-7A Yes Yes 33 0.31/0.36 M/M Acinar-solid/acinar-solid 96.9/94.8 No/Yes Intrahepatic 
HC-7B No          
HC-8 No          
HC-9 Yes Yes 32 0.43/0.45 M/M to P Trabecular/trabecular-solid 96.3/98.2 No/No  
Xenograft codeShort-term yieldLong-term yieldViability (mo)No. of passagesTumor sizea (cm)Differ.b DegreeaCell patternaKi67 LIa,c (%)Vascular invasionaDissemination pattern
HC-1 No          
HC-2 Yes No 15 0.40/0.36 M/ND Acinar/ND 65.3/ND No/ND  
HC-3 Yes Yes 48 12 0.45/0.47 M to P/P Acinar-solid/solid 29.5/98.6 Yes/Yes Intrahepatic, lymph node 
HC-3Mxd Yes Yes 16 0.41/0.38 M/ND Solid/ND 85.1/ND Yes/ND  
HC-4 No          
HC-5 Yes No 13 0.35/0.30 M/ND Solid/ND 48.1/ND No/ND  
HC-6 No          
HC-7A Yes Yes 33 0.31/0.36 M/M Acinar-solid/acinar-solid 96.9/94.8 No/Yes Intrahepatic 
HC-7B No          
HC-8 No          
HC-9 Yes Yes 32 0.43/0.45 M/M to P Trabecular/trabecular-solid 96.3/98.2 No/No  
a

At early passages (passage 1 except for HC-5 case, which is passage 2)/at late passages (HC-3, passage 6; HC-7A, passage 8; HC-9, passage 4).

b

Differ., differentiation; M, moderately differentiated; P, poorly differentiated; LI, labeling index; ND, not done (material not available).

c

Number of cells stained/total cells analyzed in Ki67 immunohistochemistry.

d

HC-3Mx xenograft was obtained from the orthotopic implantation of the HC-3 lymph node metastasis.

We thank Margarita Mainar and Elena Gonzalvo for their technical assistance in immunohistochemistry, and Maria Jesús Pujol and Agnès Figueras Amat for their support in assessing CDK2 kinase activity and the terminal deoxynucleotidyl transferase-mediated nick end labeling assay, respectively.

1
Chen PJ, Chen DS Hepatitis B virus infection and hepatocellular carcinoma: molecular genetics and clinical perspectives.
Semin Liver Dis
,
19
:
253
-62,  
1999
.
2
Colombo M Hepatitis C virus and hepatocellular carcinoma.
Semin Liver Dis
,
19
:
263
-9,  
1999
.
3
Bruix J, Llovet JM Prognostic prediction and treatment strategy in hepatocellular carcinoma.
Hepatology
,
35
:
519
-24,  
2002
.
4
Parkin DM, Bray F, Ferlay J, Pisani P Estimating the world cancer burden: Globocan 2000.
Int J Cancer
,
94
:
153
-6,  
2001
.
5
Kojiro M Pathological evolution of early hepatocellular carcinoma.
Oncology
,
62
:
43
-7,  
2002
.
6
Takayama T, Makuuchi M, Hirohashi S, et al Malignant transformation of adenomatous hyperplasia to hepatocellular carcinoma.
Lancet
,
336
:
1150
-3,  
1990
.
7
van Halteren HK, Salemans JM, Peters H, Vreugdenhil G, Driessen WM Spontaneous regression of hepatocellular carcinoma.
J Hepatol
,
27
:
211
-5,  
1997
.
8
Morgan DO Cyclin-dependent kinases: engines, clocks, and microprocessors.
Annu Rev Cell Dev Biol
,
13
:
261
-91,  
1997
.
9
Sherr CJ The Pezcoller lecture: cancer cell cycles revisited.
Cancer Res
,
60
:
3689
-95,  
2000
.
10
Bergers G, Benjamin EL Tumorigenesis and the angiogenic switch.
Nat Rev
,
3
:
401
-10,  
2003
.
11
Grossmann J Molecular mechanisms of “detachment-induced apoptosis–Anoikis.”.
Apoptosis
,
7
:
247
-60,  
2002
.
12
An Z, Jiang P, Wang X, Moossa AR, Hoffman RM Development of a high metastatic orthotopic model of human renal cell carcinoma in nude mice: benefits of fragment implantation compared to cell-suspension injection.
Clin Exp Metastasis
,
7
:
265
-70,  
1999
.
13
Fidler IJ Rationale and methods for the use of nude mice to study the biology and therapy of human cancer metastasis.
Cancer Metastasis Rev
,
5
:
29
-49,  
1986
.
14
Giovanella BC, Stehlin JS, Williams LJ, Lee SS, Shepard RC Heterotransplantation of human cancers into nude mice: a model system for human cancer chemotherapy.
Cancer (Phila.)
,
42
:
2269
-81,  
1978
.
15
Sharkey FE, Fogh J Considerations in the use of nude mice for cancer research.
Cancer Metastasis Rev
,
3
:
341
-60,  
1984
.
16
Fu XY, Besterman JM, Monosov A, Hoffman RM Models of human metastatic colon cancer in nude mice orthotopically constructed by using histologically intact patient specimens.
Proc Natl Acad Sci USA
,
88
:
9345
-9,  
1991
.
17
Capella G, Farre L, Villanueva A, et al Orthotopic models of human pancreatic cancer.
Ann N Y Acad Sci
,
880
:
103
-9,  
1999
.
18
Tarafa G, Villanueva A, Farre L, et al DCC and SMAD4 alterations in human colorectal and pancreatic tumor dissemination.
Oncogene
,
19
:
546
-55,  
2000
.
19
Sun FX, Tang ZY, Liu KD, et al Establishment of a metastatic model of human hepatocellular carcinoma in nude mice via orthotopic implantation of histologically intact tissues.
Int J Cancer
,
66
:
239
-43,  
1996
.
20
Gonzalez-Garcia I, Moreno V, Navarro M, et al Standardized approach for microsatellite instability detection in colorectal carcinomas.
J Natl Cancer Inst (Bethesda)
,
92
:
544
-9,  
2000
.
21
Armengol C, Boix L, Bachs O, et al p27(Kip1) is an independent predictor of recurrence after surgical resection in patients with small hepatocellular carcinoma.
J Hepatol
,
38
:
591
-7,  
2003
.
22
Gavrieli Y, Sherman Y, Ben-Sasson SA Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation.
J Cell Biol
,
119
:
493
-501,  
1992
.
23
Evans SC, Mims B, McMasters KM, et al Exclusion of a p53 germline mutation in a classic Li-Fraumeni syndrome family.
Hum Genet
,
102
:
681
-6,  
1998
.
24
Garcia-Retortillo M, Forns X, Feliu A, et al Hepatitis C virus kinetics during and immediately after liver transplantation.
Hepatology
,
35
:
680
-7,  
2002
.
25
Koike K, Kobayashi M, Gondo M, Hayashi I, Osuga T, Takada S Hepatitis B virus DNA is frequently found in liver biopsy samples from hepatitis C virus-infected chronic hepatitis patients.
J Med Virol
,
54
:
249
-55,  
1998
.
26
Karger S Mitelman F eds. .
ISCN (1995): an international system for human cytogenetic nomenclature
, Basel Karger  
1995
.
27
Olivier M, Eeles R, Hollstein M, Khan MA, Harris CC, Hainaut P The IARC TP53 database: new online mutation analysis and recommendations to users.
Hum Mutat
,
19
:
607
-14,  
2002
.
28
Webber EM, Bruix J, Pierce RH, Fausto N Tumor necrosis factor primes hepatocytes for DNA replication in the rat.
Hepatology
,
28
:
1226
-34,  
1998
.
29
Cornell RP Acute phase responses after acute liver injury by partial hepatectomy in rats as indicators of cytokine release.
Hepatology
,
11
:
923
-31,  
1990
.
30
Raz A, Levine G, Khomiak Y Acute local inflammation potentiates tumor growth in mice.
Cancer Lett
,
148
:
115
-20,  
2000
.
31
Teoh N, Dela Pena A, Farrell G Hepatic ischemic preconditioning in mice is associated with activation of NF-κB, p38 kinase, and cell cycle entry.
Hepatology
,
36
:
94
-102,  
2002
.
32
Buetow KH, Sheffield VC, Zhu M, et al Low frequency of p53 mutations observed in a diverse collection of primary hepatocellular carcinomas.
Proc Natl Acad Sci USA
,
89
:
9622
-6,  
1992
.
33
Kubicka S, Trautwein C, Schrem H, Tillmann H, Manns M Low incidence of p53 mutations in European hepatocellular carcinomas with heterogeneous mutation as a rare event.
J Hepatol
,
23
:
412
-9,  
1995
.
34
Laurent-Puig P, Flejou JF, Fabre M, et al Overexpression of p53: a rare event in a large series of white patients with hepatocellular carcinoma.
Hepatology
,
16
:
1171
-5,  
1992
.
35
Tanaka S, Toh Y, Adachi E, Matsumata T, Mori R, Sugimachi K Tumor progression in hepatocellular carcinoma may be mediated by p53 mutation.
Cancer Res
,
53
:
2884
-7,  
1993
.
36
An FQ, Matsuda M, Fujii H, et al Tumor heterogeneity in small hepatocellular carcinoma: analysis of tumor cell proliferation, expression and mutation of p53 and β-catenin.
Int J Cancer
,
93
:
468
-74,  
2001
.
37
Graeber TG, Osmanian C, Jacks T, et al Hypoxia-mediated selection of cells with diminished apoptotic potential in solid tumours.
Nature (Lond.)
,
379
:
88
-91,  
1996
.
38
Fiorentino M, Altimari A, D’Errico A, et al Acquired expression of p27 is a favorable prognostic indicator in patients with hepatocellular carcinoma.
Clin Cancer Res
,
6
:
3966
-72,  
2000
.
39
Thomas GV, Szigeti K, Murphy M, Draetta G, Pagano M, Loda M Down-regulation of p27 is associated with development of colorectal adenocarcinoma metastases.
Am J Pathol
,
153
:
681
-7,  
1998
.
40
Ito Y, Matsuura N, Sakon M, et al Expression and prognostic roles of the G1-S modulators in hepatocellular carcinoma: p27 independently predicts the recurrence.
Hepatology
,
30
:
90
-9,  
1999
.
41
Gonzalez-Garcia I, Sole RV, Costa J Metapopulation dynamics and spatial heterogeneity in cancer.
Proc Natl Acad Sci USA
,
99
:
13085
-9,  
2002
.
42
Glaise D, Ilyin GP, Loyer P, et al Cell cycle gene regulation in reversibly differentiated new human hepatoma cell lines.
Cell Growth Differ
,
9
:
165
-76,  
1998
.
43
Gripon P, Rumin S, Urban S, et al Infection of a human hepatoma cell line by hepatitis B virus.
Proc Natl Acad Sci USA
,
99
:
15655
-60,  
2002
.
44
Buendia MA Genetics of hepatocellular carcinoma.
Semin Cancer Biol
,
10
:
185
-200,  
2000
.
45
Thorgeirsson SS, Grisham JW Molecular pathogenesis of human hepatocellular carcinoma.
Nat Genet
,
31
:
339
-46,  
2002
.