Establishment of immortalized human hepatic stellate scavenger cells to develop bioartificial livers1 : Transplantation

Journal Logo


Establishment of immortalized human hepatic stellate scavenger cells to develop bioartificial livers1

Watanabe, Takamasa2; Shibata, Norikuni3; Westerman, Karen A.4 5; Okitsu, Teru2; Allain, Jean E.6; Sakaguchi, Masakiyo7; Totsugawa, Toshinori2; Maruyama, Masanobu2; Matsumura, Toshihisa2; Noguchi, Hirofumi2; Yamamoto, Shinichiro3; Hikida, Masaki8; Ohmori, Akira8; Reth, Michael9; Weber, Anne6; Tanaka, Noriaki2; Leboulch, Philippe4 5 10; Kobayashi, Naoya2 11

Author Information
doi: 10.1097/01.TP.0000064621.50907.A6
  • Free


Acute liver failure (ALF) is often life threatening and dramatically diminishes the quality of life of patients (1). Orthotopic liver transplantation has become a successful therapy for ALF, but this procedure is costly, limited by the scarcity of donor livers, and associated with high morbidity and mortality. There is a compelling need for developing effective alternatives for patients with ALF. Considering the potential of the liver to regenerate, temporary support with bioartificial livers (BAL) is an attractive approach (2,3).

Previous attempts to develop BAL have focused on the hepatocyte biosynthetic function, ignoring the reticuloendothelial role performed by liver sinusoidal lining cells. Recently, heterotypic cell interactions between parenchymal cells and nonparenchymal neighbors have been recognized to be central to the function of many organ systems. In both the developing and mature adult liver, cell-to-cell interactions are imperative for coordinating the sophisticated liver functions (4). Hepatic stellate cells (HSC) are an important component of the tissue along the hepatic sinusoid and are responsible for microcirculation in the liver and scavenger functions. Co-culture of hepatocytes with HSC has been shown to help maintain the differentiated phenotype of hepatocytes in vitro, such as albumin and cytochrome P450-associated proteins (CYP) (4–6).

Terminally differentiated human HSC have a limited proliferating potential and eventually enter senescence, thereby limiting their utility in BAL devices. The catalytic subunit of human telomerase reverse transcriptase (hTERT) is an essential participant in the cellular immortalization process, and constitutive expression of hTERT in various primary human cell types after gene transfer often results in indefinite expansion of the cells without damage to their genome (7,8).

Here, to develop a BAL, the authors report immortalization of human HSC by means of a retroviral vector expressing hTERT and green fluorescent protein (GFP) cDNA flanked by loxP site-specific recombination targets. The clonal cell-line obtained, TWNT-1, expresses HSC-specific gene products and undergoes efficient excision of hTERT and GFP cDNAs in the presence of Cre recombinase. Protein expression of the detoxifying CYP isoenzymes 3A4 and 2C9 and urea synthesis by the human hepatocytic cell line NKNT-3 (9) was increased when co-cultured in the presence of TWNT-1 cells.


Structure of a Recombinant Retroviral Vector SSR#197

SSR#197 vector was constructed in the same manner as SSR#69, as previously described (10). The polycistronic retroviral vector, SSR#197, was derived from LXSN, and consisted of the following, from 5′ to 3′: (1) a long terminal repeat with packaging signal; (2) a 511 loxP site followed by the hTERT cDNA (kindly provided by Dr. Robert Weinberg, Whitehead Institute for Biomedical Research at Massachusetts Institute of Technology); (3) the encephalomyocarditis virus internal ribosome entry site; (4) the GFP gene; (5) a second 511 loxP site; (6) the hepatitis B posttranscriptional regulatory element; and (7) another long terminal repeat preceded by its polypurine tract. Ψ Crip cells were cultured with Dulbecco’s minimal essential medium containing 10% newborn calf serum, 100 U/mL penicillin G, and 100 μg/mL streptomycin sulfate.

Transduction of Human HSC with SSR#197

Human HSC strain LI 90, established from a mesenchymal liver tumor of a 55-year-old Japanese woman (11), was used in the present study. The cells exhibited characteristics consistent with those of HSC. Such characteristics included various connective tissue components (including collagen types I, III, IV, V, and VI), laminin, and fibronectin in addition to vitamin A storage and biosynthesis of tenascin (11). LI 90 cells at population doubling level (PDL) 5 were transduced with 5 mL of Ψ Crip cell supernatant per T75 flask in the presence of 12 μg/mL Polybrene (Sigma, St. Louis, MO) at 37°C (98.6°F) for 4 hr/day for 3 days. The cells were cultured with ASF104 (Ajinomoto Co., Tokyo, Japan) supplemented with 10% fetal bovine serum and P/S.

Flow Cytometric Analysis and Cell Line Establishment

When cells grew to confluence after SSR#197 transduction, LI 90 cells were subjected to flow cytometric analysis using FACSCalibur (Becton Dickinson Co., Mountain View, CA) to obtain GFP-positive populations. The percentage of GFP-positive cells and the mean fluorescent intensity of the positive cells were determined after compensating for autofluorescence using untransduced LI 90 as a negative control. FL-1 was used for GFP and data were analyzed using CellQuest software (Becton Dickinson). Subsequently, the GFP-expressing cells were placed on a well of 96-well plates to be cloned using a limiting dilution method. One of six clones derived from SSR#197-transduced LI 90 cells, TWNT-1, was used in the present study.

Immunofluorescent Study for α-Smooth Muscle Actin and Platelet-Derived Growth Factor β Receptor (PDGFβ-R) in TWNT-1 Cells

Indirect immunofluorescent staining for α-smooth muscle actin (SMA) and platelet-derived growth factor (PDGF)β receptor (R) was carried out. The primary antibodies for α-SMA and PDGFβ-R were mouse monoclonal immunoglobulin (Ig) G antibody against human α-SMA (Upstate Biotechnology, Lake Placid, NY) and rabbit polyclonal IgG antibody against human PDGFβ-R, respectively. Tetramethylrhodamine isothiocyanate-conjugated goat polyclonal antibodies to mouse IgG and rabbit IgG (Sigma) were used as the secondary antibody for these expression assays.

Electron-Microscopic Examination

TWNT-1 cells cultured with ASF104 medium containing retinol (Sigma) were subjected to transmission electron microscopy. Ultrathin sections of the samples were stained with uranyl and observed under Hitachi H-7100, as previously reported (12).

Gene Expression of HSC Markers and hTERT

TWNT-1 cells (5×106) harvested at two time points (PDL 50 and 150) and an equivalent number of LI 90 cells (PDL 10) were analyzed by reverse-transcriptase (RT) polymerase chain reaction (PCR) to assess gene expression of hepatocyte growth factor (HGF), flt-1, and KDR–flk-1 using 35 cycles and collagen type Iα1 and Iα2 using 30 cycles. NKNT-3 cells (simian virus 40T-immortalized human hepatocytes) were used as a negative control (9). Total RNA was isolated and used as templates, as previously described (9). The human β-actin gene served as an internal control. Primers used were as follows:

  • flt-1 (496 base pair [bp]): sense, 5′-TCTCAACAAGGATGCAGCAC-3′; antisense, 5′-GGAGAAGATTTCCCACAGCA-3′
  • KDR–flk-1 (501 bp): sense, 5′-CCCACCCCCAGAAATAAAAT-3′; antisense, 5′-ACATTTGCCGCTTGGATAAC-3′
  • HGF (493 bp): sense, 5′-CAAATGTCAGCCCTGGAGTT-3′; antisense, 5′-TTCTCCTTGACCTTGGATGC-3′
  • Collagen type Iα1 (601 bp): sense, 5′-CTGGTCCTGATGGCAAAACT-3′; antisense, 5′-ACCAGCATCACCCTTAGCAC-3′
  • Collagen type Iα2 (602 bp): sense, 5′-GGAGCAGCGGTTACTACTGG-3′; antisense, 5′-GGCGTGATGGCTTATTTGTT-3′
  • hTERT (680 bp): sense, 5′-CTGACCAGGGTCCTATTCCA-3′; antisense, 5′-TGGTTATCCCAAGCAAGAGG-3′

Protein Expression for HGF and Collagen Type I in TWNT-1 Cells

To examine HGF and collagen type I protein synthesis, total cell lysates were prepared from TWNT-1 cells (PDL 50 and PDL 150) in a cell lysis buffer, resolved on 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and transferred onto nitrocellulose membranes (Amersham International Plc., Tokyo, Japan), as previously described (13). Samples were treated with a mouse monoclonal antibody against human HGF (Genezyme, Cambridge, MA) (1:100) or against human collagen type I (Oncogene Research Products, San Diego, CA) followed by a horseradish peroxidase-conjugated anti-rabbit IgG secondary antibody (1:2,000) (MBL, Nagoya, Japan). Human actin protein served as an internal control. NKNT-3 cells were used as a negative control.

Uptake of Fluorescent DiI-Acetylated Low-Density Lipoprotein.

To examine the scavenger function of the cells, TWNT-1 cells at PDL 50 and PDL 150 were incubated with DiI-labeled acetylated low-density lipoprotein (Biomedical Technologies Inc., Stoughton, MA) for 4 hr at 37°C, washed with phosphate-buffered saline three times, and visualized under ultraviolet light. Normal human umbilical vein endothelial cells (HUVEC) at PDL 4 and NKNT-3 cells were used as a positive and negative control, respectively.

Cell Growth and Senescence-Associated β-Galactosidase Staining

To determine the growth characteristics of the cells, 5×104 of TWNT-1 cells (PDL 150) or an equal number of unmodified parental LI 90 at PDL 10 were plated in six-well plates. Cells were trypsinized and viable cells regularly counted using a trypan blue exclusion test for a period of 7 days. Senescence-associated (SA) β-galactosidase (Gal) staining was performed in TWNT-1 cells (PDL 150) and LI 90 cells at near crisis (PDL 20), as previously described (14). Normal HUVEC (PDL 45) and NKNT-3 cells were used as controls.

Telomerase Activity and Telomere Length

Telomerase activity in TWNT-1 cells was assayed by a TRAP-ease telomerase detection kit (Oncor, Gaithersburg, MD) according to the manufacturer’s protocol. Cellular extracts from LI 90 at PDL 10 and TWNT-1 cells (PDL 50 and PDL 150) were diluted to 100 ng/mL. Two microliters of each cellular extract was subjected to the TRAP assay. After the TRAP reaction, samples were loaded onto 10% polyacrylamide gel, and products were visualized by SYBR Green I staining (Molecular Probes, Eugene, OR). High-molecular-weight DNA was prepared by a standard protocol for telomere length assay. The isolated DNA (2 μg) was digested with HinfI and MspI under conditions recommended by the manufacturer (Toyobo, Osaka, Japan). Samples were migrated by electrophoresis in 0.5% agarose gel (1.3 C/cm) and transferred to nylon membranes (Schleicher & Schuell, Dassel, Germany). The oligonucleotide probe (TTAGGG)4 was synthesized and 3′-end-labeled with [α-32P] ddATP using T4 polynucleotide kinase (Amersham). Prehybridization and hybridization were performed at 46°C (114.8°F). Washes were performed in 5× saline sodium citrate (SSC) at room temperature and subsequently in 1× SSC at 60°C (140°F). Filters were autoradiographed with an intensifying screen at −80°C for 24 hr. Autoradiographs were scanned with a densitometer (Scanning Imager 300SX; Molecular Dynamics, Sunnyvale, CA) and the telomere restriction fragment (TRF) length was calculated with Excel (Microsoft Corporation, Redmond WA).

Assay for Oncogenicity

To evaluate the tumorigenicity of TWNT-1 cells, 1×107 cells were inoculated into the thighs of three different severe combined immunodeficiency (SCID) mice. As a control, 1×106 transformed human liver tumor PLC-PRF-5 cells were transplanted at the opposite side, used as a positive control. The mice were observed for at least 3 months after transplantation.

Construction and Purification of pTAT-Glutathione S-Transferase and pTAT-Cre Recombinase Expression Vectors

The authors constructed pTAT-Cre expression vectors to make a TAT-Cre fusion protein for Cre-loxP recombination. pTAT-glutathione S-transferase (GST) and pTAT-Cre vectors were constructed by inserting the PCR fragments encompassing the open-reading frames of GST from pGEX6P1 vector (Amersham) and the Cre recombinase from pANMerCreMer vector (15) into XhoI and EcoRI sites of a pTAT-HA prokaryote expression vector. The nucleotide sequences of the expression vectors were confirmed by DNA sequencing. TAT fusion proteins (TAT-Cre and TAT-GST) were purified by the method described previously (13).

Cre-loxP Recombination in TWNT-1 Cells

TWNT-1 cells (2×106) were plated in six-well plates and treated with TAT-Cre, ranging from 1 μg/mL to 10 μg/mL. Forty-eight hours after exposure to TAT-Cre, the percentage of GFP-negative cells was calculated, recovered with FACSCalibur, and subjected to RT-PCR assay for hTERT expression.

Co-culture of NKNT-3 and TWNT-1 Cells for CYP Expression and Urea Synthesis Studies

The authors randomly inoculated NKNT-3 cells with TWNT-1 cells into 100-mm culture dishes at a ratio of 10:1, respectively, and maintained then for 72 hr. Total cell lysates were prepared from NKNT-3–TWNT-1 co-culture and from single NKNT-3 culture. Samples were treated with a rabbit polyclonal antibody (DAKO, Tokyo, Japan) against human CYP 3A4 (1:100) and CYP 2C9 (1:100) followed by a horseradish peroxidase-conjugated anti-rabbit IgG secondary antibody (1:2,000) (MBL, Nagoya, Japan). Proteins obtained from single culture of TWNT-1 cells were used as a negative control. Relative percentage of CYP 3A4 and CYP 2C9 to human β-actin, used as an internal control, was calculated using National Institutes of Health Image (National Institutes of Health, Bethesda, MD). The total amounts of urea secreted into these culture media for 72 hr were measured by a dehydrogenase assay.


Cell Morphology and Expression of α-SMA and PDGFβ-R of TWNT-1 Cells

TWNT-1 cells, one of the SSR#197-immortalized LI 90 cells, were carefully characterized. Cells uniformly expressed GFP by FACS analysis (Fig. 1A) and were longitudinally spindle-shaped and formed a hills-and-valleys morphology, one of the characteristics of HSC in vitro, at confluency (Fig. 1B). TWNT-1 cells maintained in the presence of retinol showed abundant fat droplets stored in their cytoplasm (Fig. 1C). These droplets were uniformly positive for Oil Red III staining to the same extent as parental LI 90 cells (Fig. 1D). TWNT-1 cells showed expression of α-SMA and PDGFβ-R, two of the important hallmarks of HSC (Fig. 1E and F). These findings demonstrated that TWNT-1 cells possessed HSC characteristics.

Figure 1:
Morphology and expression of α-SMA and PDGFβ-R of TWNT-1 cells. TWNT-1 cells are uniformly positive for GFP by FACSCalibur (A) and cells at confluence display hills-and-valleys formations typical of HSC under phase-contrast microscopy (B) (magnification × 50). Transmission electron microscopy of TWNT-1 cells revealed abundant cytoplasmic inclusions resembling fat droplets (C) (magnification × 5,000). The fat content of these inclusions was confirmed by Oil Red staining of TWNT-1 cells (D) (magnification × 2,500). α-SMA (E) (magnification × 400) and PDGFβ-R (F) (magnification × 400) were expressed in TWNT-1 cells.

TWNT-1 Cells Expressed HSC Markers and Endocytosed Fluorescent DiI-Acetylated LDL

HSC markers and scavenger function were examined in TWNT-1 cells. The HSC-specific Flt-1, KDR–flk-1, HGF, and collagen type I mRNA were expressed for at least passage 150 without substantial difference between parental LI 90 and TWNT-1 cells (Fig. 2A). HGF expression was lower and collagen expression higher in TWNT-1 cells at PDL 150 than at PDL 50 (Fig. 2B). TWNT-1 cells endocytosed low-density lipoprotein to the same degree as LI 90 cells (Fig. 2C). NKNT-3 cells, used as a negative control, showed neither expression of such markers nor low-density lipoprotein uptake.

Figure 2:
Comparative expression of HSC markers and uptake of fluorescent DiI-acetylated low-density lipoprotein in parental LI 90 and TWNT-1 cells. (A) TWNT-1 cells expressed HFSC markers (M, size marker; 1, Flt-1; 2, KDR–Flk-1; 3, HGF; 4, collagen type Iα1; 5, collagen type Iα2; 6, human β-actin). Equal loading of PCR product was verified by detecting the human β-actin mRNA. (B) TWNT-1 showed protein expression of HGF and collagen type I. Equal loading of protein was verified by detecting the human actin protein. (C) TWNT-1 cells endocytosed fluorescent DiI-acetylated low-density lipoprotein to the same extent as parental LI 90 cells. The data are representative of three independent experiments.

Cell Growth and SA β-Gal Activity

Growth potential of TWNT-1 cells was examined. TWNT-1 cells (PDL 150) grew much more rapidly than parental LI 90 cells at PDL 10 (Fig. 3A). After 7 days in culture, TWNT-1 cells reached to PDL 153; in contrast, parental LI 90 cells were still at PDL 11. At the time of cell crisis, LI 90 cells (PDL 20) were morphologically enlarged with a flattened shape (Fig. 3B) and were stained with SA β-Gal (Fig. 3B) as strongly as HUVEC at crisis (PDL 45). In contrast, TWNT-1 cells (PDL 150) were negative for SA β-Gal (Fig. 3B). These observations clearly demonstrated that TWNT-1 cells were immortalized.

Figure 3:
Cell growth, SA β-Gal staining, and tumorigenicity. (A) TWNT-1 cells (PDL 150) grew much more rapidly than unmodified parental LI 90 cells (PDL 10). After 7 days in culture, TWNT-1 cells reached to PDL 153; in contrast, parental LI 90 cells were still at PDL 11. (B) Cell morphology and staining for SA β-Gal. Senescent LI 90 cells at PDL 20 (upper left) were stained with SA β-Gal (upper right), but TWNT-1 (lower left, PDL 150) had no detectable SA β-Gal activity (lower right). (C) PLC-PRF-5 cells formed tumors 3 weeks after inoculation; no tumor was detected after contralateral injection of TWNT-1 cells.

Oncogenetic Assay of TWNT-1 Cells

For oncogenetic study, TWNT-1 cells were transplanted into three SCID mice. No tumor developed at the injection site of TWNT-1 (1×107) (Fig. 3C) for at least 3 months after transplantation, whereas PLC-PRF-5 cells (1×106) formed tumors approximately 3 weeks after inoculation (Fig. 3C). These data suggested that TWNT-1 cells were not tumorigenic.

Telomerase Activity and Telomere Length

Telomerase activity was evaluated in both TWNT-1 and LI 90 cells. No telomerase activity was observed in LI 90 cells; in contrast, TWNT-1 cells showed telomerase activity (Fig. 4A). TWNT-1 cells expressed hTERT by RT-PCR, whereas hTERT was not detected in LI 90 cells (Fig. 4B). TRF length of LI 90 cells was shorter than that of TWNT-1 cells and, notably, TWNT-1 cells at PDL 150 had longer TRF compared with that of TWNT-1 at PDL 50 (Fig. 4C).

Figure 4:
Telomerase activity and telomere length. (A) TWNT-1 cells showed a telomerase activity similar to that found in the positive control. In contrast, parental LI90 cells expressed no telomerase activity. (B) TWNT-1 cells expressed hTERT by RT-PCR, whereas parental LI 90 cells were negative for hTERT. (C) TRF length of TWNT-1 cells tends to increase after passage. These data are representative of three independent experiments.

TAT-Cre–Mediated Excision of hTERT in TWNT-1 Cells

To evaluate Cre-loxP recombination efficacy, the authors analyzed the ratio of GFP expression in TWNT-1 cells after treatment with a TAT-Cre fusion protein (TAT-Cre). The recombination efficiency of TAT-Cre assayed by GFP-negative ratio was increased in a dose-dependent manner of TAT-Cre used, as shown in Figure 5. The retrovirally transferred GFP was removed from 60% of TWNT-1 cell population at 10 μg/mL TAT-Cre (Fig. 5B). These GFP-negative populations were subsequently collected with cell sorting and such cells were negative for hTERT expression assayed by RT-PCR. The removal of GFP from TWNT-1 cells was not observed in a TAT-GST control experiment. These observations demonstrated that TAT-Cre treatment and subsequent GFP-negative cell sorting allowed the collection of hTERT-negative reverted TWNT-1 cells.

Figure 5:
GFP expression in TWNT-1 cells after TAT-Cre treatment. GFP expression in TWNT-1 cells was analyzed by FACSCalibur 2 days after TAT-Cre treatment. (green line) Untreated TWNT-1; (white line) TAT-Cre–treated cells. (A and B) Data representative of at least three independent experiments. GFP-negative populations increased in a dose-dependent manner of TAT-Cre used, resulting in a maximum of 60% GFP-negative cells (plus 5% according to experiments) when 10 μg/mL TAT-Cre was used (B). These findings were not observed in a control experiment with TAT-GST.

Morphology, Enhanced Protein Expression of CYP3A4 and CYP2C9, and Urea Synthesis in NKNT-3–TWNT-1 Co-culture

The NKNT-3–TWNT-1 co-culture system was morphologically and functionally evaluated. Such co-culture formed hepatic lobular architecture (Fig. 6A). Although proteins derived from NKNT-3 cells in co-culture contributed less to the total protein sample than when obtained from pure NKNT-3 cell culture, the expression of CYP3A4 and CYP2C9 increased 1.9-fold and 1.5-fold in NKNT-3–TWNT-1 co-culture compared with pure culture of NKNT-3 cells, respectively (Fig. 6B). No bands for CYP 3A4 and 2C9 were detected in pure culture of TWNT-1 cells. Urea synthesis increased 1.6-fold in co-culture compared with single culture of NKNT-3. These data suggested that TWNT-1 cells could be used to improve metabolic and synthetic functions in BAL.

Figure 6:
Morphology and enhanced protein expression of CYP3A4 and CYP2C9 in NKNT-3–TWNT-1 co-culture. The morphologic appearance of NKNT-3–TWNT-1 co-culture developed a hepatic lobular architecture (A) (phase contrast micrograph; magnification × 100). The expression of CYP3A4 and for CYP2C9 increased 1.9-fold and 1.5-fold in NKNT-3–TWNT-1 co-culture compared with pure culture of NKNT-3 cells, respectively.


Although normal human livers are an ideal source of cells for BAL therapy, their availability for liver cell isolation is unfortunately limited by competition for their use in whole-organ transplantation. As an alternative, isolated pig hepatocytes or a liver cell line, C3A, clonally derived from a human hepatoblastoma has been used in a preclinical BAL trial (2,3). Concerns about porcine cells include xenozoonosis and immunologic and physiologic incompatibility with the human host, whereas human cell lines expose patients to the potential risk of releasing tumor cells or tumorigenic products from the BAL device into their circulation (16,17). To overcome these issues, the authors previously developed an immortalized human hepatocytic cell line, NKNT-3, as an alternative to primary hepatocytes (9). Extension of this procedure to other cell types present in the human liver would allow studies of cell-cell interaction and further contribute to the development of cell therapies and BAL. Among the authors’ long-term goals is the development of BAL systems that closely mimic the function of the normal liver in vivo. Pure cultures of hepatocytes recapitulate several key liver functions but fail to provide adequate levels of a few important detoxifying enzymes that include the CYP. Among the known crosstalk between hepatocytes and other liver cells, HSC are believed to play an essential role (18,19).

Here, the authors have applied a reversible immortalization procedure to HSC to yield an unlimited supply of a human cell line having the phenotype of HSC. The starting material was the human hepatic stellate cell strain LI 90 (11). LI 90 cells show differentiated characteristics of HSC but grow slowly, with a limited proliferative capacity in culture. Human cells can escape the proliferation crisis and continue to divide indefinitely at low frequency (20). The hypothesis that replicative senescence is caused by progressive telomere shortening has recently been strengthened by the demonstration that expression of hTERT was able to prevent growth arrest and immortalize human fibroblasts and epithelial cells (7,8). Hence, the authors integrated the hTERT cDNA into the chromosomes of LI 90 cells by means of the Moloney murine leukemia virus-based retroviral vector SSR#197. One of the resulting clones, TWNT-1, acquired an extensive proliferating capacity and retained its state of differentiation for at least PDL 150. Maintenance of the differentiated phenotype was evaluated by analyzing the expression of several specific mRNA and proteins (flt-1, KDR–flk-1, HGF, collagen type I). Although the gene expression profile was stable overall, a decrease in HGF expression and an increase in collagen expression were noted at high PDL 150. The authors found that the co-culture of NKNT-3 cells with TWNT-1 cells increased CYP3A4 and CYP2C9 expression by Western blotting analysis and urea synthesis. These findings support the contention that heterotypic cell interaction is of importance for enhancing the production of liver-specific enzymes by hepatocytes in vitro (4–6). Because newly developed drugs are still screened for their safety and efficacy in animal models, BAL of multiple cell composition should become an attractive platform as an alternative to animal testing. The TWNT-1 cell line the authors generated should be useful for the basic study of this specific cell type, in the design of various BAL, and for pharmacotoxicology testing. TWNT-1 cells were not tumorigenic when injected into SCID mice. However, possible tumorigenicity in compromised hosts with severe liver insufficiency cannot be ruled out; therefore, redundant safeguards should be provided in these genetically modified cells for clinical BAL treatment, especially in the event of unexpected device failure. Such safety measures include the procedure of reversible cell immortalization by retroviral transfer of an oncogene that can be subsequently excised by Cre-loxP site-specific recombination (9). The authors have successfully applied this procedure to human hepatocytes and endothelial cells (10,14).

In the authors’ previous studies, they used an adenovirus-mediated transient expression of Cre recombinase (AxCANCre) for Cre-loxP recombination, but considerable cytotoxicity associated with AxCANCre was observed. Therefore, in this work, the authors adopted protein transduction for Cre-loxP recombination in TWNT-1 cells. Protein fused to the Tyr-Gly-Arg-Lys-Lys-Arg-Arg-Gln-Arg-Arg-Arg peptide from the human immunodeficiency virus TAT protein, which has the protein transduction domain (PTD), can efficiently enter the cell membrane and be incorporated into cells (21). Transduction of TAT fusion protein occurs without binding to receptors targeting the lipid bilayer component of the cell membrane. In addition, the TAT fusion protein has an interesting feature in that it is able to move into the nucleus by the embedded nuclear localization signal in the TAT PTD. TAT-mediated transduction could provide several advantages over DNA transfection, the current standard method of intracellular protein expression: (1) most eukaryotic cell types, excluding yeast, are susceptible to transduction; (2) because transduction occurs so rapidly (15 min rather than 12 hr in serum-free media for transfection), issues of timing can be overcome; and (3) the exact intracellular concentration can be controlled precisely just by varying the amount added to the culture medium. Thus, the authors performed the excision of hTERT and GFP cDNA flanked by a pair of loxP targets by exposing TWNT-1 cells to TAT-Cre. The GFP cDNA was efficiently excised in 60% of TWNT-1 cells with the use of 10 μg/mL TAT-Cre. These recovered GFP-negative cells, reverted TWNT-1, were negative for hTERT expression by RT-PCR analysis. These observations demonstrate the feasibility of removing a segment containing hTERT and GFP cDNA intervened by loxP by TAT-Cre–mediated recombination. Subsequent cell sorting of the GFP-negative populations can make rapid preparations of reverted HSC feasible.


The authors have established a differentiated human stellate cell line, TWNT-1, and demonstrated enhancement of CYP expression and urea synthesis in the co-culture of immortalized human hepatocytes NKNT-3 and TWNT-1.


We thank T. Matsuura, M.D. (Jikei University School of Medicine, Tokyo, Japan), and K. Murakami, M.D. (Tohoku Nenkin Hospital, Sendai, Japan), for providing LI 90 cells; and Y. Kusumoto, M.D., for helpful discussion (Hiroshima University School of Medicine, Hiroshima, Japan).


1. Schiodt FV, Atillasory E, Shakil AO, et al. Etiology and outcome for 295 patients with acute liver failure in the United States. Liver Transplant Surg 1999; 5: 86.
2. Watanabe FD, Mullon CJ, Hewitt WR, et al. Clinical experience with a bioartificial liver in the treatment of severe liver failure: A phase I clinical trial. Ann Surg 1997; 225: 484.
3. Ellis AJ, Hughes RD, Wendon JA, et al. Pilot-controlled trial of the extracorporeal liver assist device in acute liver failure. Hepatology 1996; 24: 1446.
4. Bhatia SN, Balis UJ, Yarmush ML, et al. Effect of cell-cell interactions in preservation of cellular phenotype: Cocultivation of hepatocytes and nonparenchymal cells. FASEB J 1999; 13: 1883.
5. Loreal O, Levavasseur F, Fromaget C, et al. Cooperation of Ito cells and hepatocytes in the deposition of an extracellular matrix in vitro. Am J Pathol 1993; 143: 538.
6. Rojkind M, Novikoff PM, Greenwel P, et al. Characterization and functional studies of rat liver fat-storing cell line and freshly isolated hepatocyte coculture system. Am J Pathol 1995; 146: 1508.
7. Bondar AG. Extension of life-span by introduction of telomerase into normal human cells. Science 1998; 279: 349.
8. Zhu J, Wang H, Bishop JM, et al. Telomerase extends the lifespan of virus-transformed human cells without net telomere lengthening. Proc Natl Acad Sci USA 1999; 96: 3723.
9. Kobayashi N, Fujiwara T, Westerman KA, et al. Prevention of acute liver failure in rats with reversibly immortalized human hepatocytes. Science 2000; 287: 1258.
10. Westerman KA, Leboulch P. Reversible immortalization of mammalian cells mediated by retroviral transfer and site-specific recombination. Proc Natl Acad Sci 1996; 93: 8971.
11. Murakami K, Abe T, Miyazazwa M, et al. Establishment of a new human cell line, LI90, exhibiting characteristics of hepatic Ito (fat-storing) cells. Lab Invest 1995; 72: 731.
12. Murakami T. A metal impregnation method of biological specimens for scanning electron microscopy. Arch Histol Jpn 1973; 35: 323.
13. Sakaguchi M, Miyazaki M, Inoue Y, et al. Relationship between contact inhibition and intranuclear S100C of normal human fibroblasts. J Cell Biol 2000; 149: 1193.
14. Noguchi H, Kobayashi N, Westerman KA, et al. Controlled expansion of human endothelial cell populations by Cre/loxP-based reversible immortalization. Hum Gene Ther 2002; 13: 321.
15. Zhang Y, Riesterer C, Ayrall AM, et al. Inducible site-directed recombination in mouse embryonic stem cells. Nucleic Acids Res 1996; 24: 543.
16. Gartner HV, Seidl C, Luckenbach G, et al. Genetic analysis of a sarcoma accidentally transplanted from a patient to a surgeon. N Engl J Med 1996; 335: 1494.
17. Paradis K, Langford G, Long Z. Search for cross-species transmission of porcine endogenous retrovirus in patients treated with living pig tissue. Science 1999; 285: 1236.
18. Begue JM, Guguen-Guillouzo C, Pasdeloup N, et al. Prolonged maintenance of active cytochrome P-450 in adult rat hepatocytes co-cultured with another liver cell type. Hepatology 1984; 4: 839.
19. Perrot N, Chesne C, De Waziers I, et al. Effects of ethanol and clofibrate on expression of cytochrome P-450 enzymes and epoxide hydrolase in cultures and cocultures of rat hepatocytes. Eur J Biochem 1991; 200: 255.
20. Shay JW, Wright WE. Quantitation of the frequency of immortalization of normal human diploid fibroblasts by SV40T large T-antigen. Exp Cell Res 1989; 184: 109.
21. Vives E. A truncated HIV-1 Tat protein basic domain rapidly translocates through the plasma membrane and accumulates in the cell nucleus. J Biol Chem 1997; 272: 16010.
© 2003 Lippincott Williams & Wilkins, Inc.