Cell Transplantation, Vol. 17, pp. 899–909, 2008
Printed in the USA. All rights reserved.
Copyright 2008 Cognizant Comm. Corp.
0963-6897/08 $90.00 + .00
Use of Bioluminescent Imaging to Assay the Transplantation
of Immortalized Human Fetal Hepatocytes Into Mice
Moon Seok Choi,*†1Andreea M. Catana,*1Jian Wu,* Young Seok Kim,* Sang Jeong Yoon,*
Alexander D. Borowsky,‡ Sanjiv S. Gambhir,§ Sanjeev Gupta,¶ and Mark A. Zern*
*Transplant Research Institute, UC Davis Medical Center, Sacramento, CA, USA
†Department of Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, 135-710, Korea
‡Department of Pathology and Laboratory Medicine, UC Davis, Davis, CA, USA
§Molecular Imaging Program at Stanford, Departments of Radiology and Bioengineering, Bio-X Program,
Stanford University, Stanford, CA, USA
¶Marion Bessin Liver Research Center, Department of Medicine, Albert Einstein College of Medicine, Bronx, NY, USA
Noninvasive serial monitoring of the fate of transplanted cells would be invaluable to evaluate the potential
therapeutic use of human hepatocyte transplantation. Therefore, we assessed the feasibility of bioluminescent
imaging using double or triple fusion lentiviral vectors in a NOD-SCID mouse model transplanted with
immortalized human fetal hepatocytes. Lentiviral vectors driven by the CMV promoter were constructed
carrying reporter genes: firefly luciferase and green fluorescence protein with or without herpes simplex
virus type 1 thymidine kinase. Human fetal hepatocytes immortalized by telomerase reconstitution (FH-
hTERT) were successfully transduced with either of these fusion vectors. Two million stably transduced
cells selected by fluorescence-activated cell sorting were injected into the spleens of NOD-SCID mice pre-
treated with methylcholanthrene and monocrotaline. The transplanted mice were serially imaged with a
bioluminescence charged-coupled device camera after D-luciferin injection. Bioluminescence signal intensity
was highest on day 3 (6.10 ± 2.02 × 105p/s/cm2/sr, mean ± SEM), but decreased to 2.26 ± 1.54 × 105and
7.47 ± 3.09 × 104p/s/cm2/sr on day 7 and 10, respectively (p = 0.001). ELISA for human albumin in mice
sera showed that levels were similar to those of control mice on day 2 (3.25 ± 0.92 vs. 2.84 ± 0.59 ng/ml,
mean ± SEM), peaked at 18.04 ± 3.11 ng/ml on day 7, and decreased to 8.93 ± 1.40 and 3.54 ± 0.87 ng/ml
on day 14 and 21, respectively (p = 0.02). Real-time quantitative RT-PCR showed gene expression levels of
human albumin, α1-antitrypsin, and transferrin in mouse liver were 60.7 ± 6.5%, 26.0 ± 1.4%, and 156.8 ±
62.4% of those of primary human adult hepatocytes, respectively, and immunohistochemistry revealed cells
with human albumin and α1-antitrypsin expression in the mouse liver. In conclusion, our study demonstrated
that bioluminescent imaging appears to be a sensitive, noninvasive modality for serial monitoring of trans-
planted hepatic stem cells.
Key words: Hepatocytes; Stem cells; Transplantation; Luciferase; Imaging techniques
Primary human cells are a preferred source of hepato-
cyte function for liver cell transplantation. Because the
availability of primary human hepatocytes is severely
limited, various efforts have been made to generate
functional human hepatocytes from diverse sources, in-
cluding embryonic stem cells (ESC) (7), bone marrow
stem cells (20), and liver stem cells (17). However, pre-
vious studies to direct these cells to differentiate into
hepatocytes have shown only limited success. Another
potential source is our telomerase-immortalized fetal he-
patocytes. We previously reported that human fetal he-
Liver transplantation is the mainstay of therapy for
end-stage liver diseases, but its use is restricted by the
shortage of donor organs as well as the considerable mor-
bidity and mortality associated with the process. Liver-
directed cell therapies, including hepatocyte transplanta-
tion and extracorporeal bioartificial liver support devices,
have been suggested as alternatives (1,6,23). Obtaining
sufficient numbers of functional hepatocytes for these
cell-based therapies remains an unsolved problem (11,14).
Received June 11, 2007; final acceptance January 25, 2008.
1These authors contributed equally to the work.
Address correspondence to Mark A. Zern, M.D., Transplant Research Institute, UC Davis Medical Center, 4635 Second Avenue, Suite 1001,
Sacramento, CA 95817, USA. Tel: (916) 734-8063; Fax: (916) 734-8097; E-mail: firstname.lastname@example.org
900CHOI ET AL.
patocytes that were transduced with the catalytic
subunit of telomerase reverse transcriptase were im-
mortalized, capable of expressing liver-specific gene
products in vitro and in vivo, and were nontumorigenic
To evaluate the potential therapeutic use of hepato-
cyte transplantation, it is essential to evaluate the effec-
tiveness of transplanted cells to engraft and repopulate
the recipient liver. However, conventional transplanta-
tion methods require many animals to be sacrificed in
order to assess their fate in vivo, and they provide only
one point of observation per animal. Hence, it would
be invaluable to develop a noninvasive method to mon-
itor the fate of these cells on a longitudinal basis. Mo-
lecular imaging is a rapidly emerging research field
that may be useful for that purpose (18). Employing
specific molecular probes as the source of imaging
allows concomitant visual and analytical biological
phenotyping of animals. In addition, this methodology
eliminates the need to sacrifice each mouse for charac-
terization, and the consequent repetitive imaging makes
it possible to investigate biologic activities that are oth-
erwise difficult to interpret with single time point data
The triple fusion lentiviral vector, developed by one
of us, is an important achievement in this field that
allows the use of several imaging modalities to evaluate
the same biological system (19). This vector contains
three reporter genes: the firefly luciferase (fl) or Renilla
luciferase (rl) gene for bioluminescent imaging, the
green fluorescence protein (gfp) or red fluorescence pro-
tein (rfp) gene for fluorescence microscopy, and a mu-
tant herpes simplex virus type 1 sr39 thymidine kinase
(ttk) gene for positron emission tomography (PET). The
feasibility of molecular imaging using this vector has
been evaluated in various in vitro and in vivo studies.
To track the fate of transplanted cells, the viral vectors
have been injected directly into animals (29), or into a
variety of cells [neuroblastoma cells (5), rat glioma cells
(2,24), melanoma cells (19), human embryonic stem
cells (3,8), pancreatic islet cells (10), and 293T cells
(24)] that were then transplanted into animals, mainly
by implantation into subcutaneous tissue (2,5,19,24) or
by direct injection into a target organ (3,10,28). How-
ever, this vector system and novel imaging modality
have not previously been employed to assess the en-
graftment of human liver stem cells in mice.
Therefore, we assessed the usefulness of biolumines-
cent imaging employing this triple fusion vector system
in a NOD-SCID mouse model transplanted with immor-
talized fetal hepatocytes. Our results indicate that this
imaging system appears to be a promising approach for
repeatedly and noninvasively monitoring transplanted
MATERIALS AND METHODS
Culture of Immortalized Fetal
Our experiments were approved by the Institutional
Review Board at the University of California, Davis and
were performed in accordance with their guidelines. Hu-
man fetal hepatocytes were procured by Prof. S. Gupta
of Albert Einstein College of Medicine, Bronx, New
York with the approval of the Institutional Committee
of Clinical Investigations. Telemerase reconstitution was
done in our lab by ectopic expression of telomerase re-
verse transcriptase using a retrovirus vector as described
previously (26). These immortalized fetal hepatocytes
(FH-hTERT) were cultured in Dulbecco’s modified Ea-
gle’s medium (Invitrogen) supplemented with 10% inac-
tivated fetal bovine serum (Invitrogen), 5 µg/ml insulin
(Sigma-Aldrich, St. Louis, MO), 2.4 µg/ml hydrocorti-
sone (Sigma-Aldrich), 4 µmol/ml L-glutamine (Invitro-
gen), and standard antibiotics in a humidified 5% CO2
Construction of Double/Triple Fusion
A self-inactivated lentivirus was prepared by tran-
sient transfection of 293T cells as described previously
by De et al. (5). Briefly, the firefly luciferase gene and
the green fluorescence protein gene, with or without a
mutant herpes simplex virus type 1 sr39 thymidine ki-
nase gene, driven by the cytomegalovirus promoter
(pCMV) in the lentiviral backbone (CS) (CS-pCMV-fl-gfp
or CS-pCMV-fl-gfp-ttk), were cotransfected into 293T
cells with the HIV-1 packaging vector and vesicular sto-
matitis virus G glycoprotein-pseudotyped envelop vector
(pVSVG). Fu-Gene6 (Roche Molecular Biochemicals,
Indianapolis, IN) was used for cotransfection according
to the manufacturer. Lentivirus supernatant was harvested
48 and 72 h after transfection, centrifuged at a low speed
(1500 rpm for 5 min), and filter purified by passing
through a 0.45-µm filter. To make a high-titer concen-
trated stock of virus particles, the virus-containing me-
dium was centrifuged for 3 h at 50,000 × g using a
SW28 (Beckman-Coulter Inc., Fullerton, CA) rotor.
After centrifugation, the virus particles were dissolved
in 300 µl serum-free media and stored at −70°C in ali-
quots. The viral titers of double/triple fusion vectors
were assessed using the HIV-1 p24 Antigen EIA kit
(Beckman Coulter Inc.). Viral titers of the double/triple
fusion vectors as high as 1010transducing units/ml were
Transduction of FH-hTERT With Double/Triple Fusion
FH-hTERT were transduced with either double or tri-
ple fusion lentiviral vectors at a multiplicity of infection
BIOLUMINESCENT IMAGES OF HEPATOCYTE TRANSPLANTATION 901
(MOI) 50 to 100 in the presence of 8 µg/ml hexadime-
thrine bromide (Sigma-Aldrich). Successfully transduced
cells (FH-hTERT-DF/TF) were selected by fluorescence-
activated cell sorting (FACS) using a Cytomation MoFlo
Cell Sorter (Becton Dickinson, San Jose, CA) 3 days
µl D-luciferin potassium salt (30 mg/ml in PBS; Xeno-
gen, Hopkinton, MA), bioluminescent imaging was per-
formed with the in vivo Imaging System (IVIS, Xeno-
gen) (2). Bioluminescence was quantified in units of
maximum photons per second per centimeter squared
per steradian (p/s/cm2/sr) (27).
To evaluate the feasibility of bioluminescent imaging
in mice transplanted with FH-hTERT-TF, 2 × 106cells
sorted by FACS for GFP were injected into the spleens
of pretreated NOD-SCID mice. One week following the
cell transplantation, the mice were imaged with the CCD
camera after D-luciferin potassium salt (375 mg/kg body
weight) injection under isoflurane anesthesia. The mice
were sacrificed and tissue evaluated for luciferase activity.
Mice were sacrificed 3 weeks after cell transplantation
and liver tissue was removed to evaluate the engraftment
of transplanted cells. The expression level of human glyc-
eraldehyde phosphate dehydrogenase (hGAPDH) was
evaluated in mouse liver tissue and serially diluted cul-
tured FH-hTERT by real-time quantitative RT-PCR us-
ing mouse β-actin (mBAC) as a reference. Samples were
amplified on the same plate with standards consisting of
mouse liver tissue with 102to 106spiked human cells
(FH). All amplifications were performed in duplicate
and repeated twice to control for Poisson errors. Follow-
ing amplification, the threshold of specific amplification
was set within the log-linear region of product genera-
tion to determine the fractional cycle number at which
the reaction passes the threshold of positive amplifica-
tion (CT). hGAPDH amplification was normalized to
mBAC by subtraction: ∆CT= CT(hGAPDH) − CT(mBAC).
A standard curve was established with the ∆CTvalues
of standards and the number of the engrafted cells was
calculated from 30 mg of initial liver tissue. An engraft-
ment rate (%, number of FH-hTERT-TF after 3 weeks/
number of transplanted FH-hTERT-TFtimx100) was
calculated to estimate cell survival after 3 weeks (25).
RNA extracts from liver tissue of transplanted mice
were used for real-time quantitative RT-PCR assessment
of intrahepatic maturation of engrafted FH cells. Hu-
man-specific amplification was performed for albumin.
Amplification curves were compared to the human spe-
cific endogenous control (hGAPDH). RNA extracts
from FH cells in vitro were used as calibrators to deter-
mine relative expression levels.
FH-hTERT-DF/TF Cell Transplantation
The protocols for animal experiments performed in
the present study were approved by the UC Davis Insti-
tutional Animal Care and Use Committee, according to
Guideline for the Care and Use of Laboratory Animals
and Principals for the Utilization and Care of Verte-
brate Animals by the National Institutes of Health. FH-
hTERT-DF or FH-hTERT-TF were transplanted into
groups of 7–8-week-old NOD-CB17-prkdc-scid mice
(Jackson Labs, Bar Harbor, ME). These mice have a
severe combined immunodeficient syndrome; they ex-
hibit no T- or B-cell activity and demonstrate impaired
natural killer cell and macrophage function. The mice
were maintained in a rodent barrier facility providing a
sterile environment and fed autoclaved chow and water
ad libitum. Five days before transplantation, all animals
were treated with methylcholanthrene (Sigma-Aldrich)
50 mg/kg for 4 consecutive days. Methylcholanthrene
was dissolved in corn oil and each mouse received a
volume of 200 µl. These animals received monocrota-
line (200 mg/kg, Sigma-Aldrich) IP, 24 h before cell
transplantation. Five hundred milligrams of monocrota-
line was dissolved in HCl, pH 2.5, followed by neutral-
ization by NaOH 0.1 N to a final pH of 7.4. Normal
saline was added to a final volume of 10 ml and a final
concentration of 50 mg/ml. A dose of 200 mg/kg was
injected—the total volume depending on mouse body
weight being less than 150 µl (mouse body weight was
between 24 and 27 g). Two million FH-hTERT-DF/TF
were transplanted via intrasplenic injection into NOD-
SCID mice (9). Before midline laparotomy, general an-
esthesia was induced by IP injection of 2,2,2-tribromo-
ethanol (250 mg /kg body weight, Sigma-Aldrich). The
spleen was exposed and the tip of the spleen was ligated.
Cell suspension (200 µl with 1.0 × 107cells/ml) was in-
jected into the spleen with a 27-gauge needle. Before
transplantation, cell clumps were dispersed with the use
of a 70-µm cell strainer (26).
Preliminary Evaluation for Feasibility
of Bioluminescent Imaging
To assess the detection threshold for luciferase, trans-
duced cells (from 106to 102per well) were evaluated
using a bioluminescence optical charged-coupled device
(CCD) camera. Serial cell dilutions of FH-hTERT-TF
were done in triplicate and plated on 96-well plates for
bioluminescent imaging. After adding reporter probe 1
Assessment of Bioluminescent Imaging for Mice
Transplanted With FH-hTERT-DF/TF
NOD-SCID mice transplanted with 2 × 106FH-hT-
ERT-DF/TF cells were serially imaged as described above.
The first imaging was performed 3 days after transplan-
tation (day 3), then on day 7 and day 10 (Fig. 3). After
various intervals, animals were sacrificed by exsangui-
nation under general anesthesia. Blood was collected
902CHOI ET AL.
from the inferior vena cava for serological analysis and
liver samples were obtained for further analysis. Freshly
isolated primary human adult hepatocytes, cultured FH-
hTERT, or liver tissue samples from control mice (pre-
treated mice not undergoing transplantation) served as
controls (26). Human albumin secretion of the trans-
planted cells was evaluated by ELISA and liver sections
were stained with antibodies specific for human cells
and used for RNA extraction to assess the human liver-
specific gene expression of these cells.
Quantitative Gene Expression Analysis
by Real-Time RT-PCR
RNA extracts were obtained from 30 mg of mouse
liver tissue. Total RNA (1 µg) was extracted with
RNeasy Mini Kit (Qiagen). Following digestion with
DNase I, complementary DNA (cDNA) was generated
employing Thermoscript RT-PCR Systems and oligo
(dT)20 to prime first-strand synthesis. Digestion with
DNase I was performed before RT-PCR because not all
the primer–probe sets were intron sparing and the exis-
tence of pseudogenes was not excluded for all target se-
quences. The generated cDNA was diluted to 10 or 20
ng RNA-derived cDNA/µl, and aliquots of 5 µl were
used for various PCR amplifications (21,26,30).
Relative human gene expression analysis was per-
formed by real-time quantitative RT-PCR with the ABI
Prism 7700 Sequence Detection System (Applied Bio-
systems, Foster City, CA) and TaqMan PCR Master Mix
(Applied Biosystems). The concentration of all primers
and probes, as listed in Table 1, was optimized for spe-
cific annealing and primer–probe sets were validated to
ensure equal PCR efficiencies over a two-log range (26).
Expression levels of hepatocyte-specific genes were nor-
malized to the housekeeping control (hGAPDH or
mBAC) and compared with those of primary human
adult hepatocytes or cultured FH-hTERT.
Measurement of Luciferase Activity
Luciferase activity was measured using the Dual-
Luciferase Reporter Assay kit (Promega, Madison,
WI). In brief, 30 mg of liver tissue or each well of cul-
tured FH-hTERT-DF or FH-hTERT-TF cultured in six-
well plates was dispersed in 500 µl of 1× passive lysis
buffer for 15 min at room temperature. A mixture of 100
µl LAR II and 20 µl of lysate was used to measure the
firefly luciferase activity using the Lumat LB 9507
(Berthold Technologies GmbH & Co. KG, Germany).
The protein content of the cell lysates was determined
and the luminescence results were described as relative
light units per milligram of protein (RLU/mg).
PCR Amplification of GFP Sequences in Liver DNAs
Genomic DNA was extracted from liver of NOD-
SCID mouse transplanted with FH-hTERT-DF cells us-
ing the DNeasy Tissue Kit (Qiagen, Valencia, CA).
PCR proliferation of the GFP sequence was done as fol-
lows. Cultured FH-hTERT transduced with lentiviral
vectors carrying the GFP gene (FH-hTERT-GFP) was
used as a positive control and the PCR mix as a negative
control. The sequence of the forward primer was AGA
ACGGCATCAAGGTGAAC and that of the reverse
primer was TGCTCAGGTAGTGGTTGTCG. PCR was
done in a final volume of 50 µl containing 45 µl of
Platinum PCR Supermix (Invitrogen), 1.5 µl of each
primer (15 mM), and 200 ng genomic DNA. The condi-
tions were 94°C for 5 min, followed by 40 cycles at
94°C for 30 s, 55°C for 30 s, 72°C for 30 s, and final
elongation at 72°C for 7 min. PCR products were elec-
trophoresed in 1.5% agarose gels containing ethidium
bromide. A PCR product of 125 bp size was expected.
Immunohistology of Mouse Liver Tissue
To detect expression of human liver specific proteins,
tissue was fixed for 2 h with 4% paraformaldehyde
(PFA) in PBS at room temperature, for 1–2 h in 4%
PFA-5% sucrose in PBS, and finally overnight in 20%
sucrose in PBS. The fixed tissue was embedded in Tis-
sue-Tek Cryostat and OCT compound (Sakura Finetek
USA, Torrance, CA) and quickly frozen in methylbutane
at −70°C and stored at −70°C. Sections (5–7 µm thick)
were cut with Cryostat CM3050 (Leica, Germany) and
collected on microscope slides. Then slides were fixed
with 1% PFA for 10 min at room temperature and post-
fixed with ethyl alcohol/acetic acid (2:1) for 5 min at
−20°C. The fixed tissue was incubated sequentially
overnight with primary goat antibodies against human
albumin and α1-antitrypsin (α1-AT, Sigma-Aldrich) di-
luted 1:1000 and with secondary rabbit anti-goat IgG-
fluorescein isothiocyanate conjugates (Santa Cruz Bio-
technology, Santa Cruz, CA) diluted 1:1000 at room
temperature for 1 h to visualize human albumin and α1-
AT under a fluorescent microscope (21).
ELISA for Human Albumin
Sera from recipient mice were analyzed for human
albumin by ELISA using the Human Albumin ELISA
Quantitiation Kit (Bethyl Laboratories Inc., Montgom-
ery, TX). Affinity-purified goat anti-human albumin
antibody was used as coating antibody and goat anti-
human albumin-HRP conjugate as the HRP detection
antibody. Human reference serum was used as a cali-
Data are presented as means ± SEM. The Friedman
test and the Kruskal-Wallis test with Dunn’s multiple
BIOLUMINESCENT IMAGES OF HEPATOCYTE TRANSPLANTATION 903
Table 1. Primer Pairs and Hybridization Probes for Quantitative RT-PCR
Gene Primer–Probe Sequence 5′–3′
Human albumin F: AGTTTGCAGAAGTTTCCAAGTTAGTG
Human α1-antitrypsin F: TCGCTACAGCCTTTGCAATG
Human transferrin F: GTGTATCAGCAGAGACCACCGA
F, forward primer; R, reverse primer; T, hybridization probe; FAM, 6-carboxyfluorescein.
comparison test were used for statistical analysis, and a
value of p < 0.05 was considered significant.
pared to FH-hTERT P34 was graphed on a log scale.
Replicate RNA extracts (n = 3) were analyzed for each
transplanted mouse. Three weeks after cell transplanta-
tion, expression of the human albumin gene increased
by more than 5 log levels compared to cultured FH-
Transduction of Immortalized Fetal Hepatocytes
With Double/Triple Fusion Lentiviral Vectors
FH-hTERT were cultured and successfully trans-
duced with these double or triple fusion lentivirus vec-
tors showing a transduction efficiency of 40–50%, and
stably transduced cells were selected by FACS (Fig. 1).
Assessment of Bioluminescent Imaging for Mice
Transplanted With FH-hTERT-DF/TF
Two million sorted transduced cells were injected
into the spleens of another set of nine NOD-SCID mice
pretreated with methycholanthrene and monocrotaline.
Two mice died early after transplantation. Three days, 7
days, and 10 days after the cell transplantation, CCD
camera imaging was obtained. Bioluminescence signal
intensity was highest on day 3 (6.10 ± 2.02 × 105p/s/
cm2/sr, mean ± SEM), but showed a decreasing ten-
dency thereafter (2.26 ± 1.54 × 105p/s/cm2/sr on day 7
and 7.47 ± 3.09 × 104p/s/cm2/sr on day 10, p = 0.001)
(Fig. 3). A control mouse showed only background sig-
nal (data not shown).
Preliminary Evaluation of Suitability
of Bioluminescent Imaging
The detection threshold for luciferase in vitro was as-
sessed from 96-well plates containing 102to 106FH-
hTERT-TF cells by bioluminescent imaging. As few as
102cells could be detected under CCD camera (Fig. 2).
Two million sorted FH-hTERT-TF cells were in-
jected into the spleens of NOD-SCID mice that were
treated with monocrotaline and methylcholanthrene. One
week after cell transplantation, CCD camera imaging
showed the signal in both mice. Luciferase activity of
the tissue was high (2.7 × 106RLU/mg protein) confirm-
ing the CCD camera images. Three weeks after cell
transplantation, the number of surviving cells was esti-
mated from the expression level of hGAPDH as deter-
mined by quantitative RT-PCR. An engraftment rate of
7.19% was observed in the mouse with the best levels
of repopulation and engraftment.
Liver tissue extracts from transplanted NOD-SCID
mice were used for real-time quantitative RT-PCR as-
sessment of the intrahepatic maturation of engrafted FH
cells. Human-specific amplification was performed for
albumin. Amplification curves were compared to the hu-
man specific endogenous control (hGAPDH). RNA ex-
tracts from passage 34 FH-hTERT cells (FH-hTERT
P34) in vitro were used as calibrators to determine rela-
tive expression levels. The increase in expression com-
PCR Amplification of GFP Sequences in Liver DNAs
GFP sequences were amplified by PCR in genomic
DNA extracted from livers of another six transplanted
mice that were sacrificed 3, 7, and 10 days after trans-
plantation. Electrophoresis of the PCR products showed
GFP bands of 125 bp size in all six mice (Fig. 4).
ELISA for Human Albumin
Human albumin concentration was measured by
ELISA in sera of mice transplanted with FH-hTERT-DF
on day 2, 7, 14, and 21 and nontransplanted control
mice. Whereas serum levels of human albumin on day
2 were similar to those of control mice (3.25 ± 0.92 vs.
2.84 ± 0.59 ng/ml, mean ± SEM), levels increased to
reach a peak concentration of 18.04 ± 3.11 ng/ml on day
7 (p = 0.02). Then human albumin concentrations de-
904 CHOI ET AL.
Figure 1. GFP imaging of transduced FH-hTERT before and after FACS. FH-hTERT were transduced with the double or triple
fusion lentiviral vectors. While 40–50% of cells showed green fluorescence before FACS (A, B), the fluorescence was observed
in nearly all cells 2 days after sorting (C, D) (original magnification 100×).
creased to 8.93 ± 1.40 ng/ml on day 14 and to as low as
those of control mice on day 21 (Fig. 5).
Quantitative Gene Expression Analysis
We analyzed mRNA levels of human hepatocyte-spe-
cific genes from the liver tissue of mice transplanted
with the transduced cells by real-time quantitative RT-
PCR using hGAPDH as a housekeeping gene control.
Primary human adult hepatocytes were used as a refer-
ence. A total of six mice were used for quantitative gene
expression analysis. Three mice were sacrificed on day 7
and another three mice were sacrificed on day 14. Gene
expression levels of human albumin, α1-AT, and trans-
ferrin of transplanted FH-hTERT-DF in mouse liver
were 60.7 ± 6.5%, 26.0 ± 1.4%, and 156.8 ± 62.4%
(mean ± SEM) of those of primary human adult hepato-
cytes, respectively. There was no difference according
to the sacrificed time point.
Figure 2. Detection threshold for luciferase activity in vitro.
Serial cell dilutions of FH-hTERT-TF (from 106to 102per
well) were done in triplicate and plated on 96-well plates. Bio-
luminescent imaging showed the detection threshold for lucif-
erase of 100 cells per well.
Immunohistology of Mouse Liver Tissue
Immunohistochemistry using specific antibodies to
human albumin and α1-AT showed cells with positive
signals in the mouse liver specimen (Fig. 6).
BIOLUMINESCENT IMAGES OF HEPATOCYTE TRANSPLANTATION 905
Figure 3. Bioluminescent imaging and bioluminescence signal intensity over time in mice transplanted with FH-hTERT-TF. (A)
Under cooled CCD camera, a bioluminescence signal was observed from a location compatible with the liver sites in transplanted
mouse #1. The signal was detected until day 10, although signal intensity was highest on day 3 and decreased thereafter. This
mouse is representative of all mice that were transplanted. (B) Bioluminescence signal intensity reached peak level of 6.10 × 105±
2.02 × 105p/s/cm2/sr, mean ± SEM) on day 3, but showed a decreasing tendency thereafter (2.26 × 105± 1.54 × 105p/s/cm2/sr on
day 7 and 7.47 × 104± 3.09 × 104p/s/cm2/sr on day 10, p = 0.001).
Figure 4. PCR amplification of GFP sequences in DNAs from livers of transplanted mice. Mice
were sacrificed 3 days (#1, #2), 7 days (#3, #4), and 10 days (#5, #6) after transplantation. GFP
sequences were amplified by PCR in genomic DNA extracted from livers of transplanted mice
using cultured FH-hTERT transduced with lentiviral vectors carrying GFP gene (FH-hTERT-GFP)
as a positive control and the PCR mix as a negative control. Electrophoresis of the PCR products
showed GFP bands of 125 bp size in all six mice.
906CHOI ET AL.
PET and bioluminescent imaging in living mice im-
planted with the N2a cells was feasible and showed a
high correlation (R2= 0.86) between them, with optical
imaging being more sensitive (5). Ray et al. reported the
use of a triple fusion lentiviral vector containing the rl
gene, the rfp gene, and the ttk gene for bioluminescence,
fluorescence, and PET imaging in several cell lines. For
example, metastases of a human melanoma cell line
(A375M) stably expressing the triple fusion were im-
aged by microPET and optical technologies over 40–50
days in mice (19).
In most previous bioluminescent imaging studies for
cell transplantation, a tumor cell line was implanted sub-
cutaneously (2,5,19,24) or directly injected into a target
organ (3,10,28); for example, C6 rat glioma cells ex-
pressing rl were implanted subcutaneously in mice (2),
N2a murine neuroblastoma cells transfected with ttk and
fl or rl were implanted subcutaneously in nude mice
(24), or embryonic stem cells expressing the triple fu-
sion vector were injected into the myocardium (3). How-
ever, bioluminescence imaging employing this vector
system has not yet been studied in living animals trans-
planted with human hepatic cells. Moreover, it is diffi-
cult to obtain CCD camera imaging in highly vascular
organs such as liver and spleen because the efficiency
of light transmission is low due to absorption of light by
oxyhemoglobin and deoxyhemoglobin (4).
To our knowledge, our study is the first one to show
the feasibility of this novel imaging system to monitor
the engraftment of transplanted human hepatic cells into
mice. Bioluminescence imaging using double or triple
fusion lentiviral vectors was evaluated in a NOD-SCID
mouse model transplanted with FH-hTERT. FH-hTERT
were transduced with the double or triple fusion lentivi-
rus with a high transduction efficiency (40–50%) and
the transduced cells were successfully transplanted into
the spleen of mice. After FH-hTERT were injected into
the spleens of the mice, CCD camera imaging demon-
strated the bioluminescence signal in their livers, con-
sistent with high luciferase activity of the tissue, and
successful engraftment and maturation of transplanted
FH-hTERT was demonstrated by evaluation of human-
specific gene expression. The detection of the human
cell engraftment in transplanted mice was confirmed by
PCR amplification of GFP sequences in liver DNAs,
ELISA for human albumin in sera, quantitative RT-PCR
for human hepatocyte-specific genes from the liver tis-
sue, and immunohistochemistry of the mouse liver spec-
Two of our observations are worth additional atten-
tion. First, fl gene expression and bioluminescence sig-
nal intensity was highest at the earliest assessment; how-
ever, albumin synthesis (shown by ELISA) was not
Figure 5. ELISA for human albumin in sera from transplanted
mice. Sera from recipient mice (n = 2 for day 2, n = 4 for day
7, day 14, and day 21, respectively) and control mice (n = 2)
were analyzed for human albumin by ELISA. Serum levels of
human albumin on day 2 were not different form those of con-
trol mice (3.25 ± 0.92 vs. 2.84 ± 0.59 ng/ml, mean ± SEM),
but levels increased to reach a peak (18.04 ± 3.11 ng/ml, p =
0.02) on day 7. Then the concentrations decreased to 8.93 ±
1.40 ng/ml on day 14 and to as low as those of control mice
on day 21 (p = 0.02).
To assess the effectiveness of human hepatocyte
transplantation, noninvasive serial monitoring of the fate
of transplanted cells would be extremely useful. Hence,
we evaluated the feasibility of bioluminescent imaging
using double or triple fusion lentiviral vectors in a NOD-
SCID mouse model transplanted with immortalized fetal
hepatocytes. This represents the first time that human
hepatic progenitor cells have been imaged in a longitudi-
nal in vivo model of cell transplantation in rodents.
Bioluminescent imaging using fl reporter gene ex-
pression was first shown by Wu et al. (29). A cooled
CCD camera was shown to provide consistent and re-
producible results within ±8% standard deviation from
mean values, and a detection sensitivity of 1 × 106
plaque-forming units of E1-deleted adenovirus express-
ing fl driven by a CMV promoter. After injecting this
vector into the skeletal muscle of living mice, it was
determined that a cooled CCD camera can sensitively
and noninvasively track the location, intensity, and per-
sistence of fl gene expression (29). Lentiviral vectors,
which are self-inactivated and biologically safe for hu-
mans, can deliver genetic material into cells regardless
of their proliferation status (12). De et al. constructed
a lentiviral vector carrying two reporter genes: ttk for
microPET and fl for CCD camera imaging. Neuroblast-
oma (N2a) cells were stably transfected by this virus
showing high correlation (R2= 0.91) between the ex-
pression of the two reporter genes in cell culture. Micro-
BIOLUMINESCENT IMAGES OF HEPATOCYTE TRANSPLANTATION 907
Figure 6. Immunohistochemistry of mouse liver specimen. Immunohistochemistry showed expres-
sion of human liver specific proteins in liver tissue from transplanted mice. Slides prepared from
liver tissue were stained with primary goat antibodies against human albumin (ALB) and α1-
antitrypsin (α1-AT) and with secondary rabbit anti-goat IgG-fluorescein isothiocyanate conjugates.
Fluorescence microscopy showed some cells with positive signals for human albumin (A, B) and
α1-AT (C, D) in the mouse liver specimen (original magnification 200×).
active at that time. This discrepancy between fl gene
expression and liver-specific protein expression may be
explained by the speculation that fl gene expression may
be driven by the CMV promoter even at an immature
stage of the immortalized fetal hepatocytes. On the other
hand, albumin synthesis may not be fully activated until
these transplanted fetal hepatocytes mature into more
functional hepatocytes in mouse liver. This is consistent
with our previous report showing maturation of trans-
planted FH-hTERT and subsequent increase in albumin
gene expression in the mouse liver with time (26), and
the fact that albumin mRNA expression increased by 5
logs in transplanted cells in comparison to their expres-
sion in the cells in culture. Epithelial-mesenchymal tran-
sition might be a possible explanation for this observation
(22). Second, liver-specific gene and protein expression
as well as bioluminescence signal intensity decreased in-
stead of increasing over time, even though we induced
injury in the endogenous mouse organ. This result re-
flects the general finding that transplanted human cells
seldom proliferate in the mouse liver (13). This result
occurred despite our effort to facilitate cell transplanta-
tion and endogenous cell proliferation by pretreating
mice with monocrotaline, which promotes transplanted
cell engraftment and advances liver repopulation by
causing endothelial and hepatocyte injury (9), and meth-
ycholanthrene, a hepatic P450 enzyme inducer that en-
hances monocrotaline hepatic toxicity. Thus, our results
highlight a major issue in human hepatic cell transplan-
tation studies in rodent models: the difficulty in demon-
strating exogenous human cell transplantation when
human cells appear to have a selective proliferative dis-
advantage in rodent livers. This critical problem may
lend itself to a series of alternative approaches in the
future: better injury models in these fragile immunosup-
pressed animals, the use of non-human primates whose
livers may well be more welcoming to human cells, and
the use of human growth factors or other factors, such
as activation of growth factor signaling [e.g., through c-
Met agonistic antibody as shown by Kay and colleagues
to enhance survival and proliferation of human hepato-
cytes growth in mice (15,16)]. Despite these peripheral
908 CHOI ET AL.
Sakaguchi, M.; Uemura, T.; Noguchi, H.; Kondo, A.;
Tanaka, N.; Namba, M. Transplantation of highly differ-
entiated immortalized human hepatocytes to treat acute
liver failure. Transplantation 69(2):202–207; 2000.
12. Lewis, P. F.; Emerman, M. Passage through mitosis is re-
quired for oncoretroviruses but not for the human immu-
nodeficiency virus. J. Virol. 68(1):510–516; 1994.
13. Malhi, H.; Gorla, G. R.; Irani, A. N.; Annamaneni, P.;
Gupta, S. Cell transplantation after oxidative hepatic pre-
conditioning with radiation and ischemia-reperfusion leads
to extensive liver repopulation. Proc. Natl. Acad. Sci.
USA 99(20):13114–13119; 2002.
14. Nakamura, J.; Okamoto, T.; Schumacher, I. K.; Tabei, I.;
Chowdhury, N. R.; Chowdhury, J. R.; Fox, I. J. Treatment
of surgically induced acute liver failure by transplantation
of conditionally immortalized hepatocytes. Transplanta-
tion 63(11):1541–1547; 1997.
15. Ohashi, K.; Marion, P. L.; Nakai, H.; Meuse, L.; Cullen,
J. M.; Bordier, B. B.; Schwall, R.; Greenberg, H. B.;
Glenn, J. S.; Kay, M. A. Sustained survival of human he-
patocytes in mice: A model for in vivo infection with hu-
man hepatitis B and hepatitis delta viruses. Nat. Med.
16. Ohashi, K.; Meuse, L.; Schwall, R.; Kay, M. A. cMet acti-
vation allows persistent engraftment of ectopically trans-
planted xenogenic human hepatocytes in mice. Transplant.
Proc. 33(1–2):587–588; 2001.
17. Overturf, K.; Al-Dhalimy, M.; Ou, C. N.; Finegold, M.;
Grompe, M. Serial transplantation reveals the stem-cell-
like regenerative potential of adult mouse hepatocytes.
Am. J. Pathol. 151(5):1273–1280; 1997.
18. Ray, P.; Bauer, E.; Iyer, M.; Barrio, J. R.; Satyamurthy,
N.; Phelps, M. E.; Herschman, H. R.; Gambhir, S. S.
Monitoring gene therapy with reporter gene imaging.
Semin. Nucl. Med. 31(4):312–320; 2001.
19. Ray, P.; De, A.; Min, J. J.; Tsien, R. Y.; Gambhir, S. S.
Imaging tri-fusion multimodality reporter gene expression
in living subjects. Cancer Res. 64(4):1323–1330; 2004.
20. Schwartz, R. E.; Reyes, M.; Koodie, L.; Jiang, Y.; Black-
stad, M.; Lund, T.; Lenvik, T.; Johnson, S.; Hu, W. S.;
Verfaillie, C. M. Multipotent adult progenitor cells from
bone marrow differentiate into functional hepatocyte-like
cells. J. Clin. Invest. 109(10):1291–1302; 2002.
21. Shirahashi, H.; Wu, J.; Yamamoto, N.; Catana, A.; Wege,
H.; Wager, B.; Okita, K.; Zern, M. A. Differentiation of
human and mouse embryonic stem cells along a hepato-
cyte lineage. Cell Transplant. 13(3):197–211; 2004.
22. Sicklick, J. K.; Choi, S. S.; Bustamante, M.; McCall,
S. J.; Perez, E. H.; Huang, J.; Li, Y. X.; Rojkind, M.;
Diehl, A. M. Evidence for epithelial-mesenchymal transi-
tions in adult liver cells. Am. J. Physiol. 291(4):G575–
23. Strom, S. C.; Fisher, R. A.; Thompson, M. T.; Sanyal,
A. J.; Cole, P. E.; Ham, J. M.; Posner, M. P. Hepatocyte
transplantation as a bridge to orthotopic liver transplanta-
tion in terminal liver failure. Transplantation 63(4):559–
24. Wang, Y.; Iyer, M.; Annala, A. J.; Chappell, S.; Mauro,
V.; Gambhir, S. S. Noninvasive monitoring of target gene
expression by imaging reporter gene expression in living
animals using improved bicistronic vectors. J. Nucl. Med.
25. Wege, H.; Chui, M. S.; Le, H. T.; Strom, S. C.; Zern,
M. A. In vitro expansion of human hepatocytes is re-
issues, our study demonstrated the feasibility of biolumi-
nescent imaging using double or triple fusion lentiviral
vectors as a sensitive, noninvasive, and serial monitoring
modality for transplanted hepatocytes. The imaging re-
sults were confirmed by a series of biochemical and his-
tologic modalities. This should provide investigators in-
terested in hepatic cell transplantation with a new research
tool to evaluate transplanted cell engraftment and repop-
ulation in the liver.
ACKNOWLEDGMENTS: This work was supported in part by
the GlaxoSmithKline Research Fund of the Korean Association
for the Study of the Liver (M.S.C.), NIH grants AA014173,
DK073901, DK075415 and California Institute Regenerative
Medicine (CIRM) grant RC1-00359 (M.A.Z.), NIH grant
DK46952 (S.G.), and NIH grants PAR-04-069, HL078632, and
1. Allen, J. W.; Hassanein, T.; Bhatia, S. N. Advances in
bioartificial liver devices. Hepatology 34(3):447–455; 2001.
2. Bhaumik, S.; Gambhir, S. S. Optical imaging of Renilla
luciferase reporter gene expression in living mice. Proc.
Natl. Acad. Sci. USA 99(1):377–382; 2002.
3. Cao, F.; Lin, S.; Xie, X.; Ray, P.; Patel, M.; Zhang, X.;
Drukker, M.; Dylla, S. J.; Connolly, A. J.; Chen, X.;
Weissman, I. L.; Gambhir, S. S.; Wu, J. C. In vivo visual-
ization of embryonic stem cell survival, proliferation, and
migration after cardiac delivery. Circulation 113(7):1005–
4. Chance, B.; Luo, Q.; Nioka, S.; Alsop, D. C.; Detre, J. A.
Optical investigations of physiology: a study of intrinsic
and extrinsic biomedical contrast. Philos. Trans. R Soc.
Lond. 352(1354):707–716; 1997.
5. De, A.; Lewis, X. Z.; Gambhir, S. S. Noninvasive imaging
of lentiviral-mediated reporter gene expression in living
mice. Mol. Ther. 7(5 Pt. 1):681–691; 2003.
6. Fox, I. J.; Chowdhury, J. R.; Kaufman, S. S.; Goertzen,
T. C.; Chowdhury, N. R.; Warkentin, P. I.; Dorko, K.;
Sauter, B. V.; Strom, S. C. Treatment of the Crigler-Najjar
syndrome type I with hepatocyte transplantation. N. Engl.
J. Med. 338(20):1422–1426; 1998.
7. Hamazaki, T.; Iiboshi, Y.; Oka, M.; Papst, P. J.; Mea-
cham, A. M.; Zon, L. I.; Terada, N. Hepatic maturation in
differentiating embryonic stem cells in vitro. FEBS Lett.
8. Heo, J.; Factor, V. M.; Uren, T.; Takahama, Y.; Lee, J. S.;
Major, M.; Feinstone, S. M.; Thorgeirsson, S. S. Hepatic
precursors derived from murine embryonic stem cells con-
tribute to regeneration of injured liver. Hepatology 44(6):
9. Joseph, B.; Kumaran, V.; Berishvili, E.; Bhargava, K. K.;
Palestro, C. J.; Gupta, S. Monocrotaline promotes trans-
planted cell engraftment and advances liver repopulation
in rats via liver conditioning. Hepatology 44(6):1411–
10. Kim, S. J.; Doudet, D. J.; Studenov, A. R.; Nian, C.; Ruth,
T. J.; Gambhir, S. S.; McIntosh, C. H. Quantitative micro
positron emission tomography (PET) imaging for the in
vivo determination of pancreatic islet graft survival. Nat.
Med. 12(12):1423–1428; 2006.
11. Kobayashi, N.; Miyazaki, M.; Fukaya, K.; Inoue, Y.;
BIOLUMINESCENT IMAGES OF HEPATOCYTE TRANSPLANTATION 909
stricted by telomere-dependent replicative aging. Cell
Transplant. 12(8):897–906; 2003.
26. Wege, H.; Le, H. T.; Chui, M. S.; Liu, L.; Wu, J.; Giri,
R.; Malhi, H.; Sappal, B. S.; Kumaran, V.; Gupta, S.;
Zern, M. A. Telomerase reconstitution immortalizes hu-
man fetal hepatocytes without disrupting their differentia-
tion potential. Gastroenterology 124(2):432–444; 2003.
27. Wu, J. C.; Chen, I. Y.; Sundaresan, G.; Min, J. J.; De, A.;
Qiao, J. H.; Fishbein, M. C.; Gambhir, S. S. Molecular
imaging of cardiac cell transplantation in living animals
using optical bioluminescence and positron emission to-
mography. Circulation 108(11):1302–1305; 2003.
28. Wu, J. C.; Chen, I. Y.; Wang, Y.; Tseng, J. R.; Chhabra,
A.; Salek, M.; Min, J. J.; Fishbein, M. C.; Crystal, R.;
Gambhir, S. S. Molecular imaging of the kinetics of vas-
cular endothelial growth factor gene expression in ische-
mic myocardium. Circulation 110(6):685–691; 2004.
29. Wu, J. C.; Sundaresan, G.; Iyer, M.; Gambhir, S. S. Non-
invasive optical imaging of firefly luciferase reporter gene
expression in skeletal muscles of living mice. Mol. Ther.
30. Yamamoto, N.; Wu, J.; Zhang, Y.; Catana, A. M.; Cai,
H.; Strom, S.; Novikoff, P. M.; Zern, M. A. An optimal
culture condition maintains human hepatocyte phenotype
after long-term culture. Hepatol. Res. 35(3):169–177;