Báo cáo khoa học: Uptake of bilirubin into HepG2 cells assayed by thermal lens spectroscopy Function of bilitranslocase doc

Uptake of bilirubin into HepG2 cells assayed by thermal
lens spectroscopy
Function of bilitranslocase
Sabina Passamonti
1
, Michela Terdoslavich
1
, Alja Margon
1,2
, Alessandra Cocolo
1
, Nevenka Medic
1
,
Fulvio Micali
1
, Giuliana Decorti
3
and Mladen Franko
2
1 Dipartimento di Biochimica, Biofisica e Chimica delle Macromolecole, Universita
`
di Trieste, Italy
2 Laboratory for Environmental Research, Nova Gorica Polytechnic, Slovenia
3 Dipartimento di Scienze Biomediche, Universita
`
di Trieste, Italy
Bilirubin-IXa is a lipophilic tetrapyrrole derived from
heme catabolism in mammals [1]. In the plasma, it is
transported as a reversible complex with serum albu-
min, characterized by a dissociation constant in the

range 10
)7
)10
)8
m [2,3]. The concentration of bilirubin
in the plasma (0.3–1 mgÆ100 mL
)1
; 5–17 lm) results
from a balance between its production from heme
(mainly from hemoglobin) and its elimination into the
bile. Hepatic disposal of bilirubin is an energy-depend-
ent process, as it is first conjugated with glucuronic
acid [4] by UDP-glucuronyl transferase 1 [5], then
actively excreted into the biliary tract. The latter is a
rate-limiting step [6], catalysed by the primary active
transporter MRP2 [7], driving the overall flux of biliru-
bin from the blood into the bile.
The transport of bilirubin from the blood to the liver
is a carrier-mediated mechanism, shared with both
bromosulfophthalein (BSP) and indocyanine green [8].
This generated the working hypothesis that the bilirubin
carrier could be identified, and possibly isolated, by
applying its property to bind and transport BSP, instead
of bilirubin. Initially, by applying an assay of BSP bind-
ing, some proteins were isolated from liver plasma mem-
brane fractions, such as bilitranslocase [9], the organic
anion binding protein [10] and the BSP ⁄ bilirubin-bind-
ing protein [11]. Later, by application of the functional
Keywords
bilirubin; bilitranslocase; HepG2 cells;

thermal lens spectrometry; uptake assay
Correspondence
S. Passamonti, Dipartimento di Biochimica
Biofisica e Chimica delle Macromolecole,
Universita
`
di Trieste, via L. Giorgeri 1,
I-34127 Trieste, Italy
Fax: +39 040 558 3691
Tel: +39 040 558 3681
E-mail:
URL:
(Received 28 July 2005, revised 18 August
2005, accepted 30 August 2005)
doi:10.1111/j.1742-4658.2005.04949.x
Bilitranslocase is a carrier protein localized at the basolateral domain of
the hepatocyte plasma membrane. It transports various organic anions,
including bromosulfophthalein and anthocyanins. Functional studies in
subcellular fractions enriched in plasma membrane revealed a high-affinity
binding site for bilirubin, associated with bilitranslocase. The aim of this
work was to test whether the liver uptake of bilirubin depends on the activ-
ity of bilitranslocase. To this purpose, an assay of bilirubin uptake into
HepG2 cell cultures was set up. The transport assay medium contained
bilirubin at a concentration of  50 nm in the absence of albumin. To ana-
lyse the relative changes in bilirubin concentration in the medium through-
out the uptake experiment, a highly sensitive thermal lens spectrometry
method was used. The mechanism of bilirubin uptake into HepG2 cells
was investigated by using inhibitors such as anti-sequence bilitranslocase
antibodies, the protein-modifying reagent phenylmethanesulfonyl fluoride
and diverse organic anions, including nicotinic acid, taurocholate and

digoxin. To validate the assay further, both bromosulfophthalein and indo-
cyanine green uptake in HepG2 cells was also characterized. The results
obtained show that bilitranslocase is a carrier with specificity for both bili-
rubin and bromosulfophthalein, but not for indocyanine green.
Abbreviations
BSP, bromosulfophthalein; ICG, indocyanine green; Oatp, organic-anion-transporting polypeptide; PMSF, phenylmethanesulfonyl fluoride.
5522 FEBS Journal 272 (2005) 5522–5535 ª 2005 FEBS
expression cloning technique, itself based on an assay of
membrane transport of BSP [12], the first organic-
anion-transporting polypeptide (Oatp) was isolated
from rat liver [13]. Further congeners of these polypep-
tides, all belonging to a single-gene superfamily, were
identified in both rat and human tissues and found to
mediate the membrane transport of diverse substrates,
including BSP, bile salts, drugs and toxins [14]. A role as
bilirubin carriers has been recently proposed for some
of them, such as the human members OATP1B1 and
OATP1B3, on the basis of various experimental obser-
vations, both direct and indirect [15–20].
It has been calculated that about 0.25 g bilirubin per
day is transported from the blood into the liver [1].
Fine regulation of this step is probably achieved
through the expression of more than one type of bili-
rubin carrier, as originally inferred from in vivo obser-
vations in the rat [8].
A major property of rat liver bilitranslocase, a BSP
carrier [21,22], is its ability to bind bilirubin, forming a
complex with dissociation constant  2nm. Interest-
ingly, this property has been reported in two previous
studies on rat liver plasma membrane vesicles, where the

protein occurs in its native environment and is assayed
as an electrogenic BSP carrier [23,24]. The high affinity
of bilitranslocase for bilirubin points to a possible inter-
action with the albumin-free fraction of plasma bili-
rubin, the concentration of which is < 10
)7
m [25].
The aim of this work was to investigate if the uptake
of bilirubin in isolated liver cells requires the activity
of bilitranslocase. An assay of bilirubin uptake by
HepG2 cell cultures was set up, using an albumin-free
transport medium containing 50 nm bilirubin. Thermal
lens spectrometry was used to analyse these very dilute
solutions of bilirubin, as the limit of detection of ana-
lytes is 100-fold lower than with conventional spectro-
photometry [26,27]. The human hepatocarcinoma cell
line HepG2 provided the experimental cell model. The
data obtained indicate that a carrier with the func-
tional properties of bilitranslocase controls the per-
meability of HepG2 cells to bilirubin.
Results and Discussion
Expression of bilitranslocase in HepG2 cells
HepG2 cells were used to study hepatic uptake of bili-
rubin as they are of human origin [28], they can be
easily handled, and they retain various cellular func-
tions typical of normal liver [29]. An early study, based
on both immunological and functional analysis, repor-
ted that bilitranslocase is expressed in HepG2 cells
[30]. In addition, HepG2 cells express other putative
bilirubin carriers, such as OATP1B1 and OATP1B3,

to a very low level, if at all [31]. These properties
simplify the interpretation of the experimental results.
To confirm the presence of bilitranslocase in this cell
model, a postmitochondrial fraction was isolated from
a HepG2 homogenate, and the resulting proteins were
separated by SDS ⁄ PAGE. Bilitranslocase was detected
by immunoblot analysis as a band with electrophoretic
mobility close to 38 kDa (Fig. 1A). The antibody used
was an anti-sequence anti-bilitranslocase, prepared as
described previously [24].
Electrogenic BSP uptake in HepG2 membrane
vesicles
The membrane fraction obtained from HepG2 cells
was also tested for bilitranslocase-specific transport
activity, assayed as electrogenic (valinomycin-induced)
BSP transport, as described previously [32,33]. This
Fig. 1. (A) Immunoblot of postmitochondrial fractions obtained from either rat liver or HepG2 cells. Samples were separated by SDS ⁄ PAGE
and transferred to a nitrocellulose membrane. The blot was developed with a bilitranslocase antibody raised in rabbit (antibody A) as the pri-
mary antibody. The secondary antibody was an anti-rabbit IgG conjugated with alkaline phosphatase. The membrane was stained by addition
of bromochloroindolyl phosphate and nitroblue tetrazolium. Lane 1, erythrocyte ghosts (negative control); lane 2, rat liver; lane 3,
HepG2 cells. Further details are described in Experimental procedures. (B) Immunogold particles visualized by scanning electron microscopy
of a sector of a HepG2 cell. HepG2 cells were preincubated with an anti-sequence anti-bilitranslocase IgG (antibody A) raised in rabbit. The
primary immunocomplexes were detected by the formation of secondary immunocomplexes, using colloidal gold (20 nm)-conjugated
anti-rabbit IgGs. Gold particles are clearly visible as bright spots.
S. Passamonti et al. Bilirubin uptake into HepG2 cells
FEBS Journal 272 (2005) 5522–5535 ª 2005 FEBS 5523
activity was found to depend on substrate concentra-
tion, in accordance with the Michaelis–Menten equa-
tion. The derived K
m

value (3.55 ± 0.26 lm BSP) was
similar to that of analogous fractions obtained from
rat liver [33]. Moreover, these data were in close agree-
ment with the value derived from experiments of BSP
uptake in intact HepG2 cells (3.6 ± 1 lm) [30]. To
check whether the electrogenic BSP uptake activity
was related to bilirubin, the kinetics of BSP uptake in
the presence of bilirubin was examined. The pigment
acted as a competitive inhibitor (K
i
¼ 116 ± 7 nm)
with respect to BSP. These results agree with those
previously obtained with rat liver plasma membrane
vesicles [34], suggesting both a strong functional simi-
larity of the rat and human homologues of bilitrans-
locase and the involvement of the BSP electrogenic
carrier in bilirubin binding and transport.
Anti-sequence antibodies as tools to establish
the role of bilitranslocase in organic anion uptake
in HepG2 cells
The uptake of polar solutes into cells is based on the
activity of membrane carriers [35]. Once dissolved in
aqueous solution (pH 7.4) up to about 50 nm, bilirubin
may occur in solvated metastable aggregates [36], obvi-
ously in equilibrium with solvated monomers. Under
these conditions, monomeric bilirubin is the only spe-
cies presumably taken up into liver cells by a carrier-
mediated mechanism. When bilitranslocase was
assayed as the carrier catalysing the electrogenic BSP
uptake in rat liver plasma membrane vesicles, it was

shown to be inhibited by an anti-sequence antibody,
targeting segment 65–75 (EDSQGQHLSSF) of its pri-
mary structure [24]. From the effect of bilirubin on the
antibody inhibition kinetics, it was concluded that this
antibody had targeted a high-affinity binding site of
the electrogenic BSP carrier (K
d
of the carrier-bilirubin
complex ¼ 2nm) [24].
Another anti-sequence antibody, targeting segment
235–246 (EFTYQLTSSPTC) of the primary structure
of bilitranslocase, was recently shown to have similar
effects on the electrogenic BSP uptake in rat liver
plasma membrane vesicles [34]. This antibody was
shown to interact with a distinct bilirubin-binding
domain, characterized by even higher affinity for
bilirubin (K
d
¼ 0.33 nm) [34].
For the sake of clarity, the two antibodies are
referred to as antibody A (anti-65–75) and antibody B
(anti-235–246).
The ability of these antibodies to form immuno-
complexes on the extracellular surface of liver cells was
checked by identifying the primary immunocomplexes
with colloidal gold-conjugated secondary antibodies,
visualized by scanning electron microscopy. Figure 1B
shows a scanning electron micrograph of the surface
of a HepG2 cell. Colloidal gold particles appear as
white spots, locating the epitope of the bilitranslocase

targeted by antibody A. Similar results were also
obtained in primary rat hepatocytes and again in both
types of cell using antibody B (not shown).
Moreover, both antibodies were shown to inhibit the
uptake into HepG2 cells of BSP [37], malvidin 3-gluco-
side [37], and other newly identified competitive inhibi-
tors of bilitranslocase (M. Terdoslavich, unpublished
data). Therefore it appears that both antibodies not
only bind to, but also partially impair, the bilitranslo-
case function when assayed in intact cells, making
them useful tools for studying the mechanism of bili-
rubin uptake into HepG2 cells.
Bilirubin analysis by thermal lens spectroscopy
To obtain a calibration curve and determine the limit
of detection of thermal lens spectrometry, serial solu-
tions of bilirubin ranging from 2 to 50 nm were pre-
pared in NaCl ⁄ P
i
. Before the measurement, the
solutions were diluted 1 : 1 (v ⁄ v) with methanol to
improve the thermo-optical properties of the samples
(higher temperature coefficient of the refractive index,
lower thermal conductivity) and thus to increase the
sensitivity of the method. To avoid substantial photo-
degradation from the excitation laser beam (476 nm,
120 mW), readings of thermal lens spectrometry sig-
nals were taken within the first minute after insertion
of the sample cell into the instrument. During this time
interval, the decrease in the signal was less than 5%,
which is a typical maximal relative error of the thermal

lens spectrometry technique. Figure 2 shows that there
is good correlation between the concentration of bili-
rubin and the detected signal. From the regression line,
it can be assumed that the limit of detection is some-
where between 1 and 2 nm. Thermal lens spectrometry
hence appears to be suitable for the detection of bili-
rubin in albumin-free physiological solutions.
To check for any possible contribution to the chan-
ges in absorbance from the plastic-ware, appropriate
blank samples were analysed. No changes in the back-
ground signal were observed with time.
Assay of bilirubin uptake into HepG2 cells
Cells were grown to confluence in 25 cm
2
flasks.
Before the assay, the cell growth medium was removed
and the monolayer rinsed three times with 5 mL
NaCl ⁄ P
i
. Then, the transport medium, consisting of
Bilirubin uptake into HepG2 cells S. Passamonti et al.
5524 FEBS Journal 272 (2005) 5522–5535 ª 2005 FEBS
7 mL NaCl ⁄ P
i
containing 50 nm bilirubin, was added
to the cell monolayer. Samples of the transport med-
ium were withdrawn at time intervals and delivered to
conical tubes containing an equal volume of methanol.
The tubes were centrifuged and the supernatants ana-
lysed by thermal lens spectroscopy within the same

day. It was expected that the uptake of bilirubin into
the cells would result in a decrease in the initial bili-
rubin concentration in the assay medium. Figure 3
shows the results of an experiment to examine the
uptake of bilirubin at three concentrations (10, 30 and
50 nm) by HepG2 cells. In the absence of the pigment,
the signal was stable, showing that the cell monolayer
did not release compounds, such as carotenoids or fla-
vins, that might interfere with the spectroscopic analy-
sis. The data show that the signal was stable even in
the presence of bilirubin.
This finding, although unexpected, ruled out the
possibility that bilirubin may bind unspecifically to
the cell surface or that it could be destroyed during the
experiment. On the other hand, it also suggested that
these cells either did not take up the pigment or, less
likely, did not retain it inside the cytoplasm. A third
possibility is that carriers for bilirubin were either
absent or in an inactive state under the experimental
conditions. On the assumption that one of the bili-
rubin carriers may be bilitranslocase, we attempted to
increase its transport activity.
In isolated rat liver plasma membrane vesicles, this
carrier has been shown to occur in a metastable equilib-
rium of two functional states, characterized by either
high (C conformer) or low (C* conformer) affinity for
the substrate BSP [33]. This equilibrium is regulated,
in vitro, by the concentration of certain substrates, such
as BSP itself and nicotinic acid. It was speculated, how-
ever, that other allosteric effectors, possibly resulting

from intracellular metabolism, may regulate the overall
activity of bilitranslocase [33]. Data from this laborat-
ory (S. Passamonti, unpublished data) clearly show that
the redox equilibrium of the nicotinamide nucleotides
modulates the allosteric equilibrium of bilitranslocase.
In particular, a low NADH ⁄ NAD
+
ratio, such as that
occurring in physiological conditions [38], favours the
low-affinity state.
An analogous allosteric equilibrium of the bilitran-
slocase homologue may occur in HepG2 cells. Thus,
increasing the relative concentration of NADH in the
cytoplasm may activate bilitranslocase. This could be
achieved by preincubating the cell monolayer in the
presence of 5 mm lactate for 1 h. In the cells, lactate
oxidizes to pyruvate, lowering the NAD
+
⁄ NADH
ratio. Under these conditions, the bilirubin concentra-
tion in the cell medium was found to decrease by
 40% within 100 s at 37 °C, but not at 0 °C (Fig. 4,
inset), which is probably an effect of a temperature-
dependent uptake into the cell monolayer. A similar
uptake could also be observed by replacing lactate
with 5 mm ethanol, another NADH-generating
Fig. 2. Calibration curve of bilirubin analysed by thermal lens spectro-
metry. A series of bilirubin solutions was prepared by appro-
priately diluting a bilirubin stock (10 l
M in dimethylsulfoxide) in

NaCl ⁄ P
i
⁄ methanol (1 : 1, v ⁄ v). Samples (1.2 mL) were added to the
spectrophotometric cuvette and analysed by thermal lens spectro-
metry, with excitation laser operating at 120 mW power and
476 nm wavelength. Data (n ¼ 3) are means ± SEM and were fit-
ted to the y ¼ y
0
+ mx equation. The following parameters were
obtained: y
0
¼ 0.6414, m ¼ 0.0939, r
2
¼ 0.9915.
Fig. 3. Thermal lens spectrometry signal of bilirubin solutions
applied to HepG2 monolayers. Monolayers of HepG2 cells grown in
25-cm
2
flasks were exposed to 7 mL of the NaCl ⁄ P
i
solution in
either the absence (d) or presence of 10 n
M (s), 30 nM (m)or
50 n
M (n) bilirubin. Samples were withdrawn at the indicated
times, processed and analysed as described in Experimental
procedures. All procedures were carried out at 37 °C. Data are
mean ± SEM (n ¼ 4) and were fitted to the y ¼ y
0
+ mx equation.

S. Passamonti et al. Bilirubin uptake into HepG2 cells
FEBS Journal 272 (2005) 5522–5535 ª 2005 FEBS 5525
substrate. Figure 4 shows the data from three separate
experiments, each carried out in quadruplicate.
Thus, the increased permeability of HepG2 cells to
bilirubin may be due to either a direct, activating effect
on bilirubin membrane carriers or an increased ability
of the cells to accumulate bilirubin. In either case, we
concluded that the conditions for assaying bilirubin
uptake by the cells must include a 1-h preincubation in
the presence of lactate (or ethanol) to decrease the
intracellular NAD
+
⁄ NADH ratio.
Bilirubin uptake into HepG2 cells: effect of
anti-sequence bilitranslocase antibodies
To examine whether the uptake of bilirubin can be
accounted for by the activity of bilitranslocase, cells
were preincubated with antibody A, added to fresh,
serum-free growth medium containing 5 mm lactate for
1 h before the transport experiment. Figure 5 shows
that cells lost the ability to take up bilirubin. It is
worth noting that essentially no free immunoglobulins
were present in the transport medium, as cells had
been extensively rinsed before the addition of bilirubin.
The effect observed is therefore due to the formation
of stable immunocomplexes on the monolayer’s sur-
face. In a separate experiment, it was confirmed that
unspecific immunoglobulins had no influence on bili-
rubin uptake (Fig. 5, inset).

When antibody B was used, no effect on the bili-
rubin uptake was observed (Fig. 5, inset). This result,
although unexpected, is consistent with the view that
the protein segment targeted by this antibody may not
be involved in the translocation of bilirubin by the car-
rier. This result also shows the high specificity of the
biological action of antibody A, together with the
absence of effects by unspecific IgGs. It can be conclu-
ded that bilirubin uptake into HepG2 cells depends on
the activity of a membrane carrier.
Bilirubin uptake into HepG2 cells: effect of a
protein-modifying reagent
We attempted to disrupt the integrity of the bilirubin
carrier by means of the protein-modifying reagent
phenylmethanesulfonyl fluoride (PMSF). This serine-
specific reagent was used because it had been shown
to inhibit electrogenic BSP uptake in rat liver plasma
membrane vesicles [23]. A crucial observation was
that bilirubin and nicotinic acid could both prevent
this inhibition. In both cases, the protection dis-
played a hyperbolic concentration dependence, with
Fig. 4. Bilirubin uptake into HepG2 monolayers: effect of reducing
substrates in the preincubation and of the assay temperature.
Monolayers of HepG2 cells grown in 25-cm
2
flasks were preincu-
bated for 1 h in the presence of either 5 m
M lactate (squares and
triangles) or 5 m
M ethanol (circles). After removal of the culture

medium, cells were washed and exposed to 7 mL NaCl ⁄ P
i
solution
containing 50 n
M bilirubin. Samples were withdrawn at the indica-
ted times, processed and analysed as described in Experimental
procedures. The uptake assay was carried out at 37 °C (circles and
squares) and at 0 °C (triangles). Data are mean ± SEM (n ¼ 4). The
inset shows the raw data obtained by thermal lens spectrometry.
Error bars are not visible if smaller than symbols.
Fig. 5. Bilirubin uptake into HepG2 monolayers: effect of antibody
A in the preincubation. Monolayers of HepG2 cells grown in
25-cm
2
flasks were preincubated for 1 h in the presence of 5 mM
lactate, either without (squares) or with (m) 0.25 lg antibody A per
mL. After removal of the culture medium, cells were washed and
exposed to 7 mL NaCl ⁄ P
i
solution containing 50 nM bilirubin. Sam-
ples were withdrawn at the indicated times, processed and ana-
lysed as described in Experimental procedures. The uptake assay
was carried out at 37 ° C. Data are mean ± SEM (n ¼ 4). The inset
shows the results obtained by preincubating HepG2 cell mono-
layers in the absence (squares) or presence of rabbit immunoglobu-
lins purified from preimmune sera (r) or antibody B (circles), both
used as 0.25 lgIgGÆmL
)1
. Error bars are not visible if smaller than
symbols.

Bilirubin uptake into HepG2 cells S. Passamonti et al.
5526 FEBS Journal 272 (2005) 5522–5535 ª 2005 FEBS
half-maximal effects at 2 and 11 nm, respectively.
Moreover, at saturating concentrations of the two
ligands, electrogenic BSP uptake was fully refractory
to PMSF [23]. Thus, it was inferred that PMSF
targeted a specific site on the BSP electrogenic car-
rier involved in the binding of both bilirubin and
nicotinic acid. At that time, it was also speculated
that bilirubin uptake into the liver may be impaired,
if not blocked, by the chemical modification of
serines of the BSP electrogenic carrier. This predic-
tion was tested experimentally in the following
experiment.
HepG2 cells were preincubated with lactate as speci-
fied above; 20 min before the end of the preincubation,
0.1 mm PMSF was added to the growth medium. Bili-
rubin uptake was then assayed as described above.
Figure 6 shows that the uptake of bilirubin was
strongly inhibited under these conditions.
Bilirubin uptake into HepG2 cells: effect of
nicotinic acid, taurocholate and digoxin
Previous work has shown that the inhibition of electro-
genic BSP uptake into rat liver plasma membrane vesi-
cles by either antibody A [24] or PMSF [23] is strongly
influenced by both bilirubin and nicotinic acid, sug-
gesting that bilirubin and nicotinic acid share a com-
mon binding site on the carrier involved. To test this,
bilirubin uptake in HepG2 cells was tested in the pres-
ence of 1 lm nicotinic acid, a saturating concentration

for bilitranslocase [23,24].
As shown in Fig. 7, bilirubin uptake was completely
blocked under these conditions, probably as a result of
competition at the level of the bilirubin carrier, rather
than due to intracellular bilirubin binding and conju-
gation, as these two steps have not been documented
in the hepatic metabolism of nicotinic acid [39].
The effects of taurocholate and digoxin were also
examined. These compounds have been shown not to
inhibit bilitranslocase transport activity in rat liver
plasma membrane vesicles (data not shown and [40]),
but are well-known substrates of OATP carriers. In
particular, it was expected that taurocholate would
inhibit both OATP carriers expressed in HepG2, i.e.
OATP1A2 (OATP-A) and OATP1B3 (OATP8) [31],
whereas digoxin would inhibit only OATP1B3 [41].
This test is important because the latter has been
shown to transport bilirubin [18].
As shown in Fig. 7, 100 lm taurocholate did not
influence bilirubin uptake, whereas 2 lm digoxin was
found to delay the onset of bilirubin uptake. We are
unclear about the biological meaning of these results,
because, had OATP1B3 been a digoxin-sensitive bili-
rubin carrier, it would have been inhibited by tauro-
cholate as well.
Fig. 6. Bilirubin uptake into HepG2 monolayers: effect of PMSF in
the preincubation. Monolayers of HepG2 cells grown in 25-cm
2
flasks were preincubated for 1 h in the presence of 5 mM lactate,
without (squares) or with (m) 0.1 m

M PMSF added 20 min before
the end of the preincubation. After removal of the culture medium,
cells were washed and exposed to 7 mL NaCl ⁄ P
i
solution contain-
ing 50 n
M bilirubin. Samples were withdrawn at the indicated
times, processed and analysed as described in Experimental proce-
dures. The uptake assay was carried out at 37 °C. Data are mean ±
SEM (n ¼ 4). Error bars are not visible if smaller than symbols.
Fig. 7. Bilirubin uptake into HepG2 monolayers: effects of the addi-
tion of nicotinic acid, taurocholate and digoxin. Monolayers of
HepG2 cells grown in 25-cm
2
flasks were preincubated for 1 h in
the presence of 5 m
M lactate. After removal of the culture med-
ium, cells were washed and exposed to 7 mL NaCl ⁄ P
i
solutions
containing 50 n
M bilirubin in the absence (squares) or presence of
either 1 l
M nicotinic acid (m) or 100 lM taurocholate (d)or2lM
digoxin (.). Samples were withdrawn at the indicated times, proc-
essed and analysed as described in Experimental procedures. The
uptake assay was carried out at 37 °C. Data are mean ± SEM (n ¼
4). Error bars are not visible if smaller than symbols.
S. Passamonti et al. Bilirubin uptake into HepG2 cells
FEBS Journal 272 (2005) 5522–5535 ª 2005 FEBS 5527

BSP uptake into HepG2 cells: effect of bilirubin,
nicotinic acid and taurocholate
The uptake of bilirubin and BSP from the blood into
the liver has long been known to occur by an appar-
ently common mechanism of transport [8,42]. To test
this concept in HepG2 cells, BSP uptake was studied
using an identical experimental approach to the bili-
rubin-uptake experiments, i.e. by measuring the time
course of disappearance of the substrate from the
extracellular medium. The latter was interpreted as
uptake of BSP into the cell monolayer and plotted
accordingly against time, as in a previous study [37].
Figure 8 shows that BSP uptake increased rapidly with
time and reached the steady state in  30 s. In the
presence of 100 lm taurocholate, both the rate and
extent of BSP uptake were unchanged, suggesting that
bile salt carriers are not involved in the event. In con-
trast, the figure shows that both 50 nm bilirubin and
1 lm nicotinic acid inhibited BSP uptake; interestingly,
both substrates inhibited only the onset of BSP uptake,
which was slightly delayed, in analogy with the effects
produced by the antibodies to bilitranslocase [37].
Thus, it can be concluded that BSP uptake into
HepG2 cells is partially accounted for by the activity
of a carrier of bilirubin and nicotinic acid, presumably
bilitranslocase.
Indocyanine green (ICG) uptake into HepG2 cells:
effects of PMSF and antibodies to bilitranslocase
Another anionic dye, ICG, is also known to be taken
up by the liver via a carrier-mediated mechanism,

shared with both BSP and bilirubin [8]. When tested as
a reversible inhibitor of electrogenic BSP uptake in rat
liver plasma membrane vesicles, it displayed an unusual
effect, consisting of a sharp increase in the initial rate
of uptake, such that it could not be reliably measured
(data not shown). This precluded the kinetic characteri-
zation of its effect. Therefore, to investigate the poss-
ible involvement of bilitranslocase in ICG transport,
the uptake of this dye into HepG2 cells was examined.
Figure 9 shows the rapid uptake of 1.5 lm ICG in
HepG2 cells and its temperature-dependence. Neither
antibody A nor PMSF had any effect on the time
course of the dye uptake. Similar results were also
obtained at a higher ICG concentration (6 lm, not
shown).
ICG uptake into HepG2 cells: effects of substrates
specific for either bilitranslocase or OATP carriers
Figure 10 shows the results obtained by adding various
organic anions to the ICG solution. Figure 10A shows
that the time course of the dye uptake in the presence
Fig. 8. BSP uptake by HepG2 monolayers: effects of the addition
of bilirubin, nicotinic acid and taurocholate. Monolayers of HepG2
cells were grown in 25-cm
2
flasks. After removal of the culture
medium, cells were washed and exposed to 3.5 mL NaCl ⁄ P
i
solu-
tion containing 24 l
M BSP in the absence (squares) or presence of

50 n
M bilirubin (s)or1lM nicotinic acid (n) or 100 l M taurocholate
(.). Samples were withdrawn at the indicated times, processed
and analysed as described in Experimental procedures. The uptake
assay was carried out at 37 °C. Data are mean ± SEM (n ¼ 4).
Error bars are not visible if smaller than symbols.
Fig. 9. ICG uptake by HepG2 monolayers: effect of either antibody
A or PMSF in the preincubation. Monolayers of HepG2 cells grown
in 25-cm
2
flasks were p reincubated for 20 min in the absence (cir-
cles and diamonds) or presence of 0.25 lg antibody AÆml
)1
(.)or
0.1 m
M PMSF (m). After removal of the cell culture medium, cells
were washed and exposed to 7 mL NaCl ⁄ P
i
solution containing
1.5 l
M ICG. Samples were withdrawn at the indicated times, proc-
essed and analysed as described in Experimental procedures. The
uptake assay was carried out at 37 °C, in all cases, except for a
control (diamond) carried out at 0 °C. Data are mean ± SEM (n ¼
4). Error bars are not visible if smaller than symbols.
Bilirubin uptake into HepG2 cells S. Passamonti et al.
5528 FEBS Journal 272 (2005) 5522–5535 ª 2005 FEBS
of 1 lm nicotinic acid was identical with that of the
control. In the same set of tests, 50 nm bilirubin had
only a small inhibitory effect on the late phase of the

uptake. In contrast, both taurocholate (100 lm) and
digoxin (2 lm) inhibited the overall time course of the
dye uptake (Fig. 10B). As digoxin is a specific substrate
of OATP1B3, this result suggests that this carrier,
although expressed to a limited extent in this cell line
[31], may also be involved in ICG uptake. However, as
the inhibition caused by this compound was only par-
tial, we cannot conclude whether it is due to nonsatu-
rating concentrations of this inhibitor or to the activity
of other unknown digoxin-insensitive ICG carriers.
Conclusions and perspectives
Bilirubin uptake in HepG2 cells is a carrier-
mediated event
The possibility of using thermal lens spectroscopy for
determining bilirubin concentration in aqueous solu-
tions, within the limits of its solubility at physiological
pH, has opened up the unprecedented opportunity to
examine its cellular uptake from a solvent-free and
albumin-free solution. Under these conditions, no
unspecific adsorption of the pigment to the cell surface
could be detected. Indeed, we observed that the pig-
ment concentration in the assay medium remained
constant: (a) in all cases (Fig. 3) except when the cells
were preincubated in the presence of reducing sub-
strates (Fig. 4); (b) when the assay was carried out on
ice, even with cells preincubated in the presence of
reducing substrates (Fig. 4); (c) when the assay med-
ium contained nicotinic acid, even with cells preincu-
bated in the presence of reducing substrates (Fig. 7);
and (d) when the cells were preincubated in the

presence of reducing substrates and the protein-modi-
fying reagent PMSF (Fig. 6).
Thus, the apparent disappearance of bilirubin from
the extracellular medium was assumed to be due to cel-
lular uptake. The latter is clearly a carrier-mediated
event, as it was not only temperature sensitive (Fig. 4),
but also greatly and specifically reduced by a reversible
inhibitor, such as nicotinic acid (Fig. 7), by a covalent
protein-modifying reagent (Fig. 6) and by a bilitrans-
locase antibody (Fig. 5).
Under the prevailing assay conditions,  90 pmol
bilirubin disappeared from the medium and were accu-
mulated in the cell monolayer. Applying the typical
ratio of 7 lL per mg protein to the 3 mg protein of
the HepG2 monolayer, the monolayer volume can be
estimated to be  20 lL. The intracellular bilirubin
concentration can thus be estimated to be 4.5 lm, i.e.
90 times higher than the extracellular concentration.
The rate of bilirubin uptake observed in this assay is
 5 pmolÆmin
)1
Æ10
)6
cells. Assuming that the initial rate
of uptake is directly proportional to the number of
cells, it can be estimated that 0.75 lmol bilirubinÆmin
)1
could be extracted by 1.5 · 10
11
cells, the estimated

number of cells in a normal human liver. This corres-
ponds to the extraction of about 1 mmol per day, i.e.
0.58 g per day. This value is slightly higher than the
physiological flux of bilirubin from the blood into the
liver (0.25 g per day [1]), which is itself rate-limited by
the canalicular excretion step.
The main bilirubin carrier in HepG2 cells has the
functional features of the electrogenic BSP carrier
in rat liver plasma membrane vesicles
Various pieces of evidence consistently indicate that
the functional features of bilirubin uptake in HepG2
Fig. 10. ICG uptake by HepG2 monolayers: effects of either nicotinic acid and bilirubin (A) or taurocholate and digoxin (B). Monolayers of
HepG2 cells were grown in 25-cm
2
flasks. After removal of the culture medium, cells were washed and exposed to 7 mL NaCl ⁄ P
i
solution
containing 1.5 l
M ICG in the absence (squares, in both panels) or presence of 50 nM bilirubin (s,A), 1 lM nicotinic acid (e, A), 100 lM tauro-
cholate (.,B)or2l
M digoxin (m, B). Samples were withdrawn at the indicated times, processed and analysed as described in Experimental
procedures. The uptake assay was carried out at 37 °C. Data are mean ± SEM (n ¼ 4). Error bars are not visible if smaller than symbols.
S. Passamonti et al. Bilirubin uptake into HepG2 cells
FEBS Journal 272 (2005) 5522–5535 ª 2005 FEBS 5529
cells, shown in this study, closely match some of the
functional features of electrogenic BSP uptake in rat
liver plasma membrane vesicles.
First, the electrogenic BSP uptake in rat liver plasma
membrane vesicles can be inhibited by the serine-speci-
fic reagent PMSF [23]. Complete protection against

such inhibition can be yielded by both bilirubin and
nicotinic acid at nanomolar concentrations. Kinetic
analysis of these effects has enabled the calculation of
the dissociation constants of the complexes of bilitrans-
locase with the two ligands (2 and 11 nm, respectively),
suggesting that the electrogenic BSP carrier is a high-
affinity binding protein for those molecules [23]. The
above results show that the bilirubin carrier in HepG2
cells is also sensitive to both PMSF and nicotinic acid.
Moreover, the almost complete blockade of bilirubin
uptake in HepG2 cells by PMSF contrasts with the
only partial (no more than 30%) inhibition of electro-
genic BSP uptake by the latter. This is in line with the
earliest prediction, that the occupation of the bilitrans-
locase bilirubin-binding site by either PMSF (by cova-
lent binding to serines) or nicotinic acid (by reversible
interaction) would totally impair bilirubin uptake in
intact cells [23].
Second, the electrogenic BSP uptake in rat liver
plasma membrane vesicles is inhibited by a polyclonal
antibody raised against an undecapeptide correspond-
ing to segment 65–75 of the primary structure of bili-
translocase [24]. As in the case of PMSF, both
bilirubin and nicotinic acid protected against such inhi-
bition, yielding quite similar K
d
values for the com-
plexes of bilitranslocase with these ligands.
Third, the electrogenic BSP uptake is regulated by
the NAD

+
⁄ NADH ratio in a physiologically meaning-
ful range (S. Passamonti, unpublished data). Crucial to
the successful observation of bilirubin uptake in
HepG2 cells was the choice of preincubating cells with
reducing substrates such as lactate or ethanol.
Collectively, these data suggest that the bilirubin
carrier in HepG2 cells is quite similar to the electro-
genic BSP carrier in rat liver plasma membrane vesi-
cles. Therefore, this assay is indeed a reliable tool for
investigating the transport of bilirubin and related
substrates in the liver.
The enigma of the primary structure of rat liver
bilitranslocase and the biological activities of
anti-sequence bilitranslocase antibodies
The amino-acid sequence of bilitranslocase was origin-
ally deduced from a clone selected from an expression
library, on the basis of its ability to express a protein
that reacted with a monoclonal antibody that inhibited
electrogenic BSP uptake in rat liver plasma membrane
vesicles [24].
Unfortunately, the clone was truncated, preventing
its further characterization. The very high homology
of the nucleotide sequence of the clone with that of the
antisense strand of ceruloplasmin suggested that it was
the product of a wrong cloning strategy. Nonetheless,
it was noted that the translated primary structure of
the clone contained a short segment (62–99) that was
highly homologous to segment 6–45 of a number of
a-chains of phycocyanins [24]. In this class of proteins,

this segment is invariant and has been shown to be
involved in accommodating the prosthetic group phyco-
cyanobilin, an open tetrapyrrole [43]. Segment 62–99
of the clone was therefore expected to provide a poten-
tial structural component for high-affinity bilirubin
binding. This observation per se suggested investiga-
tion of the occurrence of a protein encoded by the
cloned sequence.
Thus, the issue was examined experimentally using
antibody A, which targets segment 65–75, and, under
stringent conditions, reacts with: (a) a 38-kDa mem-
brane protein in rat liver; (b) a 38-kDa protein purified
using a standard protocol for the isolation of bilitrans-
locase [44]; (c) the protein expressed by the clone in
Escherichia coli [24]. Moreover, this antibody displayed
a remarkable biological activity, in that it inhibited
electrogenic BSP uptake in rat liver plasma membrane
vesicles, by targeting a high-affinity binding site for
both bilirubin and nicotinic acid [24].
In this work, the biological activity of this antibody
is fully confirmed. Not only does it identify a homo-
logue of rat liver bilitranslocase in the plasma mem-
brane of HepG2 cells, but also it shows that the
homologue in question is indeed a bilirubin carrier.
To test further the existence of a liver membrane
protein with the trait of the primary structure of bili-
translocase, a second polyclonal anti-sequence anti-
body (antibody B) was produced and found to have
similar biological properties to those of the first
[34].This antibody was, however, found to be ineffec-

tive in inhibiting bilirubin uptake in HepG2 cells.
This shows that although the segment targeted by
antibody B (EFTYQLTSSPTC) binds bilirubin with
extraordinarily high affinity [34], it does not disturb
the transport of bilirubin from its surface binding
site, which is targeted by antibody A, into the
hepatocyte. This lack of inhibition is an exception, as
the hepatocellular uptake of both BSP and malvidin
3-glucoside [37], as well as that of two recently identi-
fied competitive inhibitors of electrogenic BSP uptake
(M. Terdoslavich, unpublished data), were indeed
inhibited by antibody B. This would suggest that the
Bilirubin uptake into HepG2 cells S. Passamonti et al.
5530 FEBS Journal 272 (2005) 5522–5535 ª 2005 FEBS
multiple interactions of bilirubin with the carrier,
which conceivably underlie the substrate translocation
through the transport pore of the carrier, are some-
what different from those of the other substrates. A
similar conjecture has previously been reported, based
on certain properties of electrogenic BSP uptake in
rat liver plasma membrane vesicles [23].
Other carriers participate in organic anion uptake
in HepG2 cells
Antibody A is the most specific bilitranslocase inhib-
itor and is shown above to inhibit quite strongly
bilirubin uptake in HepG2 cells, suggesting that
bilitranslocase is the main carrier involved. Both
PMSF and nicotinic acid fully inhibited bilirubin
uptake, suggesting that, although in principle not spe-
cific just for bilitranslocase, they can impair a cellular

function exclusively performed by bilitranslocase.
The uptake of BSP into HepG2 cells was in fact
only partially inhibited by nicotinic acid, leaving intact
the activity of other putative BSP carriers. In an effort
to identify the latter carriers, taurocholate was used,
but no effect was observed. These results suggest that
bile salt carriers take no part in BSP uptake in HepG2
cells, which is consistent with the finding that sodium-
dependent bile salt uptake is absent from these cells
and sodium-independent bile salt uptake is lower than
in primary hepatocytes [45]. The residual BSP uptake
activity, however, calls for the existence of so far
unidentified BSP carriers, which are undoubtedly not
involved in bile salt transport. The activity of the
BSP ⁄ bilirubin binding protein has been documented in
HepG2 cells [46] and may be very strongly involved in
our assay of BSP uptake.
The fact that bilirubin uptake is not inhibited by
taurocholate suggests again that bile salt carriers are
inactive. This conclusion is supported by the observa-
tion that the uptake of ICG, a high-affinity substrate
of OATP1B1 [15], is not affected by taurocholate.
However, the partial inhibition of both bilirubin and
ICG uptake by digoxin suggests that OATP1B3,
although poorly expressed [31], may act as a carrier
for both bilirubin and ICG in HepG2 cells.
In conclusion, this work shows that bilitranslocase is
involved in the uptake of both bilirubin and BSP, but
not of ICG, in HepG2 cells. The activity of bilitrans-
locase in this cell line, which is very similar to that

observed in normal liver, has apparently not been lost
as a result of transformation, unlike that of bile salt
carriers. As a consequence, the occurrence of addi-
tional bilirubin carriers in normal liver cannot be ruled
out.
Experimental procedures
Cell culture
Human hepatoblastoma HepG2 cells were obtained from
the American Type Culture Collection (Rockville, MD,
USA) and maintained in Eagle’s minimum essential med-
ium supplemented with 10% (v ⁄ v) fetal bovine serum,
1mm sodium pyruvate, 100 UÆmL
)1
penicillin and
100 lgÆmL
)1
streptomycin. Cells were grown in a humid-
ified incubator at 5% CO
2
⁄ 95% air (v ⁄ v) at 37 °C and
after 6 days harvested by exposure to 0.05% trypsin and
0.02% EDTA (all these reagents were purchased from
Euroclone Ltd, Wetherby, Yorks., UK) for 5 min. Cells
(8.4 · 10
5
cells, at a cell density of 1.2 · 10
5
cellsÆmL
)1
)

were seeded in 25 cm
2
flasks. After 6 days, as cells reached
confluence, the medium was replaced, and uptake experi-
ments were performed on the following day.
Postmitochondrial fraction of HepG2
homogenates
This fraction collects the plasma membrane and micro-
somes of the cell homogenate. Cells (3.6 · 10
8
) were harves-
ted by scraping the flask bottom, collected in 8 mL of an
ice-cold solution (10 mm Hepes, 0.25 m sucrose, pH 7.4)
and homogenized in a Dounce tube using a small clearance
pestle. The homogenate was centrifuged at 700 g for
10 min. The supernatant was collected and centrifuged
again at 100 000 g for 1 h. The pellet was resuspended in
the above solution at a final protein concentration of
 10 mgÆmL
)1
and stored in aliquots at )80 °C. Bilitrans-
locase transport activity was almost entirely recovered in
this fraction.
Assay of bilitranslocase transport activity
Bilitranslocase transport activity was assayed spectrophoto-
metrically as previously described in detail [32–34]. Briefly,
3–10 lL( 10 lg protein) of the postmitochondrial fraction
was added to a stirred cuvette containing 2 mL assay med-
ium (0.1 m potassium phosphate, pH 8.0, with 18 lm BSP,
without or with 100 nm bilirubin as a reversible inhibitor) at

room temperature. This addition caused an instantaneous
fall in absorbance (recorded at k ¼ 580–514 nm). After the
attainment of a steady-state (4 s), a second fall in absorb-
ance was brought about by adding 5 lg valinomycin in
1 lL methanol, which was due to valinomycin-induced K
+
diffusion potential. This K
+
diffusion drove the substrate
into the vesicles [32]. The slope of the linear phase of this
decrease in absorbance, lasting about 1 s, is referred to as
electrogenic BSP uptake and is related to bilitranslocase
transport activity [24,47]. The pH in the assay medium was
constant throughout the test, as previously shown with an
analogous preparation from rat liver [32].
S. Passamonti et al. Bilirubin uptake into HepG2 cells
FEBS Journal 272 (2005) 5522–5535 ª 2005 FEBS 5531
Antibodies to bilitranslocase
Antibody A was raised in one rabbit (Oryctolagus cuniculus,
New Zealand White strain), immunized with a multiantigen
peptide-based system as described [24], using the peptide
EDSQGQHLSSF, corresponding to segment 65–75 of the
primary structure of bilitranslocase. Sera were purified by
affinity chromatography as described previously [24]. Anti-
body B was obtained using peptide EFTYQLTSSPTC, cor-
responding to segment 235–246 of the primary structure of
bilitranslocase. The peptide was conjugated to maleimide-
activated keyhole limpet hemocyanin (Antibody production
and purification kit; Pierce, Rockford, IL, USA) and injec-
ted into a rabbit; sera were purified by affinity chromato-

graphy as previously described [34].
Assay of bilirubin uptake in HepG2 cells
On the day of the experiment, the cell medium was
replaced, and a 1 h preincubation at 37 °C was started by
adding 7 mL serum-free fresh medium containing 5 mm lac-
tic acid, or 5 mm ethanol, or 2.5 mm glucose, as indicated
in the legends to the figures. In some experiments, anti-
sequence antibodies or immunoglobulins (both at 0.25 lg
IgGÆmL
)1
), purified from preimmune sera as described pre-
viously [34], were included in the preincubation. The pre-
incubation was stopped by removing the medium, and cells
were washed four times with NaCl ⁄ P
i
,at37°C.
The uptake assay was carried out under dim light and
started by the addition of 50 nm bilirubin (7 mL) dissolved
in NaCl ⁄ P
i
, pH 7.4, at 37 °C. In some experiments, this
transport medium contained 100 lm taurocholate, or 2 lm
digoxin, or 1 lm nicotinic acid.
The flasks were kept in a water bath at 37 °C and gently
shaken 60 times a minute. Samples (600 lL) of the extra-
cellular solution were collected at time intervals, transferred
to conical tubes containing 600 lL methanol at 0 °C, and
centrifuged at 1200 g for 5 min. The supernatants were kept
in ice until analysis by thermal lens spectroscopy on the
same day, as described below. All experiments were per-

formed in quadruplicate.
Assay of BSP and ICG uptake in HepG2 cells
On the day of the experiment, the cells were preincubated
for 20 min with 2.5 mL fresh medium, in either the absence
or presence of bilitranslocase antibodies (0.25 lg
IgGÆmL
)1
). Then, the medium was removed, and the cells
were washed four times with NaCl ⁄ P
i
,at37°C. The uptake
assay was started by the addition of transport medium,
consisting of either 3.5 mL 24 lm BSP or 7 mL 1.5 l m
ICG, in NaCl ⁄ P
i
, pH 7.4, at 37 °C. In some experiments,
the transport medium contained 100 lm taurocholate, or
2 lm digoxin, or 1 lm nicotinic acid, or 50 nm bilirubin.
The flasks were kept in a water bath at 37 °C and shaken
as above. Samples were collected at time intervals and
transferred to polycarbonate spectrophotometric cuvettes
for analysis.
Thermal lens spectrometric analysis of bilirubin
The thermal lens spectrometric analysis was performed in
batch mode on a dual-beam, mode-mismatched thermal
lens spectrometer [27] schematically shown in Fig. 11.
The excitation source (pump laser beam) was provided
by an argon ion laser (Coherent; Innova 90, Santa Paula,
CA, USA), tuned to a 476-nm line (120 mW), and modula-
ted by a mechanical chopper (Scitec Instruments, Redruth,

Cornwall, UK) at 10 Hz. The pump beam was focused on
to the sample cell by a 250-mm focal length lens. A second
lens (80 mm focal length) focused a helium–neon laser
(Uniphase 1103P, 632.8 nm, 2 mW; eFiberland, Fremont,
CA, USA) probe beam in front of the sample cell. The dis-
tance between the second lens and the sample cell, 138 mm,
was determined experimentally to achieve the maximal ther-
mal lens spectrometric signal. A dichroic mirror provided
collinear propagation of the pump beam and the probe
beam through the sample cell. An optical filter placed in
front of the photodiode removed the pump beam, allowing
the probe beam only to reach the photodiode. Changes in
the intensity on the probe beam axis, which are propor-
tional to the absorbance of the sample, were detected using
a silicon photodiode (Thorlabs Inc., Newton, NJ, USA;
model 201 ⁄ 5797227), situated 1.5 m beyond the sample cell.
The photodiode was connected to a lock-in amplifier (Stan-
ford Research Instruments, Sunnyvale, CA, USA; model
SR830), amplifying only the component of the input signal
that appears with the frequency of the reference signal from
the modulation. This is achieved by a Fourier transforma-
tion of the signal while filtering out all other frequencies
Fig. 11. Experimental set up for measurements with a dual-beam,
mode-mismatched thermal lens spectrometer in batch mode. Two
laser beams (pump and probe laser beams) pass through the sam-
ple cell. The pump beam induces the photothermal effect, consist-
ing of a change in the refractive index of the sample, proportional
to the concentration of the analyte. The refractive index change can
be quantified by the deflection of the probe beam. The instrumental
set up is described in Experimental procedures. This figure is not

to scale.
Bilirubin uptake into HepG2 cells S. Passamonti et al.
5532 FEBS Journal 272 (2005) 5522–5535 ª 2005 FEBS
with a low-pass filter. The readings of the signal values
were made directly from the lock-in amplifier.
Spectrophotometric analysis
This was carried out using a UV ⁄ VIS spectrophotometer
(Ultrospec 2100 pro; Amersham Biosciences, AB, Uppsala,
Sweden).
BSP
Aliquots of 100 lL of the extracellular solutions containing
BSP were added to cuvettes containing 900 lL 0.1 m
NaOH and analysed at k ¼ 580 nm ( e ¼ 64 000 m
)1
Æcm
)1
).
ICG
Aliquots of 800 lL of the extracellular solutions containing
ICG were analysed at k ¼ 780 nm (e ¼ 77 500 m
)1
Æcm
)1
).
SDS ⁄ PAGE and immunoblot analysis
Aliquots of the postmitochondrial fraction of HepG2 homo-
genates were resuspended in loading buffer (4 mg pro-
teinÆmL
)1
) and incubated at 100 °C for 5 min; the samples

(80 lg protein) were separated by SDS ⁄ PAGE (12% gel).
The proteins were transferred to a nitrocellulose membrane
by electroblotting for 2 h at 2 mAÆcm
)2
with the semi-dry
multiphor protein blotter system (Pharmacia Biotech,
Milano, Italy).
The membrane was kept for 2 h in blocking solution
[0.15 m NaCl, 50 mm Tris ⁄ HCl, pH 7.5, containing 0.05%
(v ⁄ v) Tween 20 and 3% (w ⁄ v) skimmed milk] and then incu-
bated overnight at 4 °C with the primary antibody (antibody
A, 1.5 lg IgGÆmL
)1
) diluted in the blocking solution. After
being washed, the blotted membrane was incubated with an
alkaline phosphatase-conjugated secondary antibody [goat
anti-(rabbit IgG) Ig; dilution 1 : 1000; Sigma, Milan, Italy]
for 1 h at room temperature. The membrane was extensively
washed and finally stained by the addition of bromochloro-
indolyl phosphate and nitroblue tetrazolium.
Scanning electron microscopy
HepG2 cells were seeded on Thermanox
TM
(Nunc, New
York, NY, USA) disks (diameter, 13 mm). The latter were
rinsed three times with NaCl ⁄ P
i
and fixed by treatment
with 2% (v ⁄ v) paraformaldehyde in 0.15 m cacodylate buf-
fer (pH 7.4) for 1 h at room temperature. After being

washed three times with the same buffer, disks were incuba-
ted in the presence of a blocking solution [NaCl ⁄ P
i
, con-
taining 1% (w ⁄ v) BSA and 20% (v ⁄ v) normal goat serum]
for 1 h at room temperature in a humidity chamber. The
blocking solution was then removed and disks were
incubated in the presence of the primary antibody solution
[NaCl ⁄ P
i
, containing 1% (w ⁄ v) BSA, 1% (v ⁄ v) normal goat
serum, 4% (v ⁄ v) fetal bovine serum, 0.1% (v ⁄ v) Tween 20
and antibody A at a concentration of 3.5 lg IgGÆmL
)1
] for
2 h at room temperature in a humidity chamber. This solu-
tion was removed and disks were washed 5 times in
NaCl ⁄ P
i
, and twice in a conditioning solution (50 mm
Tris ⁄ HCl and 0.15 m NaCl, pH 8.4). Disks were transferred
to a humidified chamber for the incubation (for 2 h at
room temperature) with the colloidal gold (20 nm)-conju-
gated anti-rabbit IgGs (British BioCell, Cardiff, UK; dilu-
ted 1 : 100), dissolved in a solution consisting of 50 mm
Tris ⁄ HCl, 0.15 m NaCl (pH 8.4), 1% (w ⁄ v) BSA, 1% (v ⁄ v)
normal goat serum, 4% (v ⁄ v) fetal bovine serum and 0.1%
(v ⁄ v) Tween 20. At the end of the incubation, disks were
washed six times with 50 mm Tris ⁄ HCl ⁄ 0.15 m NaCl
(pH 8.4) and three times with MilliQ grade water. To

enhance the detection of the colloidal gold particles, these
were coated with silver, using the Enhancer kit (British Bio-
Cell) for 5 min. The reaction was stopped by rinsing the
disks three times with water. Disks were dehydrated in gra-
ded ethanol solutions, treated with hexamethyldisilazane
(Sigma), which were finally removed by evaporation. The
disks were exposed to carbon vapours, to visualize the col-
loidal gold particles. The samples were observed under a
Leica Stereoscan 430i microscope (Leica Cambridge Ltd,
Cambridge, UK). Controls were carried out using immuno-
globulins purified from preimmune rabbit sera as described
previously [34] and displayed no colloidal gold-conjugated
secondary immunocomplexes.
Acknowledgements
Thanks are due to Professor G. L. Sottocasa and Pro-
fessor C. Tiribelli (University of Trieste), for useful dis-
cussions, and to Dr Marco Stebel (Animal Facility
Manager, CSPA, University of Trieste), for the immun-
ization and bleeding of rabbits. Financial support from
the University of Trieste (Fondi 60% and the Alpe-
Adria fellowship to A.M.), the Regione Friuli Venezia
Giulia (L.R. 3 ⁄ 98, art.16, fondo anno 2002), the Minis-
tero dell’Istruzione, Universita
`
e Ricerca (PRIN pro-
jects 2002055532 and 2004070118), the Progetto D4
(European Social Fund, Regione Friuli Venezia Giulia
and Italian Ministry of Welfare) are acknowledged.
References
1 McDonagh AF (2001) Turning green to gold. Nat

Struct Biol 8, 198–200.
2 Petersen CE, Ha CE, Harohalli K, Feix JB & Bhagavan
NV (2000) A dynamic model for bilirubin binding to
human serum albumin. J Biol Chem 275, 20985–20995.
S. Passamonti et al. Bilirubin uptake into HepG2 cells
FEBS Journal 272 (2005) 5522–5535 ª 2005 FEBS 5533
3 Weisiger RA, Ostrow JD, Koehler RK, Webster CC,
Mukerjee P, Pascolo L & Tiribelli C (2001) Affinity of
human serum albumin for bilirubin varies with albumin
concentration and buffer composition: results of a novel
ultrafiltration method. J Biol Chem 276, 29953–29960.
4 Billing BH, Cole PG & Lathe GH (1957) The excretion
of bilirubin as a diglucuronide giving the direct van den
Bergh reaction. Biochem J 65, 774–784.
5 Bosma PJ, Seppen J, Goldhoorn B, Bakker C, Oude
Elferink RP, Chowdhury JR, Chowdhury NR & Jansen
PL (1994) Bilirubin UDP-glucuronosyltransferase 1 is
the only relevant bilirubin glucuronidating isoform in
man. J Biol Chem 269, 17960–17964.
6 Goresky CA (1965) The hepatic uptake and excretion of
sulfobromophthalein and bilirubin. Can Med Assoc J
92, 851–857.
7 Jedlitschky G, Leier I, Buchholz U, Hummel-Eisenbeiss
J, Burchell B & Keppler D (1997) ATP-dependent trans-
port of bilirubin glucuronides by the multidrug resis-
tance protein MRP1 and its hepatocyte canalicular
isoform MRP2. Biochem J 327 (1), 305–310.
8 Scharschmidt BF, Waggoner JG & Berk PD (1975)
Hepatic organic anion uptake in the rat. J Clin Invest
56, 1280–1292.

9 Tiribelli C, Lunazzi G, Luciani M, Panfili E, Gazzin
B, Liut G, Sandri G & Sottocasa GL (1978) Isolation
of a sulfobromophthalein-binding protein from hepa-
tocyte plasma membrane. Biochim Biophys Acta 532,
105–112.
10 Wolkoff AW & Chung CT (1980) Identification, purifi-
cation, and partial characterization of an organic anion
binding protein from rat liver cell plasma membrane.
J Clin Invest 65, 1152–1161.
11 Stremmel W, Gerber MA, Glezerov V, Thung SN,
Kochwa S & Berk PD (1983) Physicochemical and
immunohistological studies of a sulfobromophthalein-
and bilirubin-binding protein from rat liver plasma
membranes. J Clin Invest 71, 1796–1805.
12 Jacquemin E, Hagenbuch B, Stieger B, Wolkoff AW &
Meier PJ (1991) Expression of the hepatocellular
chloride-dependent sulfobromophthalein uptake system
in Xenopus laevis oocytes. J Clin Invest 88, 2146–2149.
13 Jacquemin E, Hagenbuch B, Stieger B, Wolkoff AW &
Meier PJ (1994) Expression cloning of a rat liver Na
(+) -independent organic anion transporter. Proc Natl
Acad Sci USA 91, 133–137.
14 Hagenbuch B & Meier PJ (2003) The superfamily of
organic anion transporting polypeptides. Biochim
Biophys Acta 1609, 1–18.
15 Cui Y, Konig J, Leier I, Buchholz U & Keppler D
(2001) Hepatic uptake of bilirubin and its conjugates by
the human organic anion transporter SLC21A6. J Biol
Chem 276, 9626–9630.
16 Vavricka SR, Van Montfoort J, Ha HR, Meier PJ &

Fattinger K (2002) Interactions of rifamycin SV and
rifampicin with organic anion uptake systems of human
liver. Hepatology 36, 164–172.
17 Cui Y & Walter B (2003) Influence of albumin binding
on the substrate transport mediated by human hepato-
cyte transporters OATP2 and OATP8. J Gastroenterol
38, 60–68.
18 Briz O, Serrano MA, MacIas RI, Gonzalez-Gallego J &
Marin JJ (2003) Role of organic anion-transporting
polypeptides, OATP-A, OATP-C and OATP-8, in the
human placenta-maternal liver tandem excretory path-
way for foetal bilirubin. Biochem J 371, 897–905.
19 Huang MJ, Kua KE, Teng HC, Tang KS, Weng HW &
Huang CS (2004) Risk factors for severe hyperbilirubi-
nemia in neonates. Pediatr Res 56, 682–689.
20 Campbell SD, de Morais SM & Xu JJ (2004) Inhibition
of human organic anion transporting polypeptide OATP
1B1 as a mechanism of drug-induced hyperbilirubine-
mia. Chem Biol Interact 150, 179–187.
21 Sottocasa GL, Baldini G, Sandri G, Lunazzi G &
Tiribelli C (1982) Reconstitution in vitro of sulfobromo-
phthalein transport by bilitranslocase. Biochim Biophys
Acta 685, 123–128.
22 Miccio M, Lunazzi GC, Gazzin B & Sottocasa GL
(1990) Reconstitution of sulfobromophthalein transport
in erythrocyte membranes induced by bilitranslocase.
Biochim Biophys Acta 1023, 140–142.
23 Passamonti S, Battiston L & Sottocasa GL (1997) Aryl-
sulfonylation of bilitranslocase in plasma membranes
from rat liver enables to discriminate between natural

and artificial substrates. Biochim Biophys Acta 1323,
130–136.
24 Battiston L, Passamonti S, Macagno A & Sottocasa GL
(1998) The bilirubin-binding motif of bilitranslocase and
its relation to conserved motifs in ancient biliproteins.
Biochem Biophys Res Commun 247, 687–692.
25 Ahlfors CE (2001) Bilirubin-albumin binding and free
bilirubin. J Perinatol 21 (Suppl. 1), S40–S42; discussion
S59–62.
26 Franko M, van de Bovenkamp P & Bicanic D (1998)
Determination of trans-beta-carotene and other
carotenoids in blood plasma using high-performance
liquid chromatography and thermal lens detection.
J Chromatogr B Biomed Sci Appl 718, 47–54.
27 Franko M (2001) Recent applications of thermal lens
spectrometry in food analysis and environmental
research. Talanta 54, 1–13.
28 Aden DP, Fogel A, Plotkin S, Damjanov I & Knowles
BB (1979) Controlled synthesis of HBsAg in a differen-
tiated human liver carcinoma-derived cell line. Nature
282, 615–616.
29 Knasmuller S, Mersch-Sundermann V, Kevekordes S,
Darroudi F, Huber WW, Hoelzl C, Bichler J & Majer
BJ (2004) Use of human-derived liver cell lines for the
detection of environmental and dietary genotoxicants;
current state of knowledge. Toxicology 198, 315–328.
Bilirubin uptake into HepG2 cells S. Passamonti et al.
5534 FEBS Journal 272 (2005) 5522–5535 ª 2005 FEBS
30 Marchegiano P, Carubbi F, Tiribelli C, Amarri S, Stebel
M, Lunazzi GC, Levy D & Bellentani S (1992) Transport

of sulfobromophthalein and taurocholate in the HepG2
cell line in relation to the expression of membrane carrier
proteins. Biochem Biophys Res Commun 183, 1203–1208.
31 Marin JJ, Mangas D, Martinez-Diez MC, El-Mir MY,
Briz O & Serrano MA (2003) Sensitivity of bile acid
transport by organic anion-transporting polypeptides to
intracellular pH. Biochim Biophys Acta 1611, 249–257.
32 Baldini G, Passamonti S, Lunazzi GC, Tiribelli C &
Sottocasa GL (1986) Cellular localization of sulfo-
bromophthalein transport activity in rat liver. Biochim
Biophys Acta 856, 1–10.
33 Passamonti S, Battiston L & Sottocasa GL (1998)
Bilitranslocase can exist in two metastable forms with dif-
ferent affinities for the substrates: evidence from cysteine
and arginine modification. Eur J Biochem 253, 84–90.
34 Passamonti S, Cocolo A, Braidot E, Petrussa E, Peresson
C, Medic N, Macri F & Vianello A (2005) Characteriza-
tion of electrogenic bromosulfophthalein transport in car-
nation petal microsomes and its inhibition by antibodies
against bilitranslocase. FEBS J 272, 3282–3296.
35 Evans AM (1996) Membrane transport as a determinant
of the hepatic elimination of drugs and metabolites. Clin
Exp Pharmacol Physiol 23 , 970–974.
36 Boiadjiev SE, Watters K, Wolf S, Lai BN, Welch WH,
McDonagh AF & Lightner DA (2004) pKa and aggre-
gation of bilirubin: titrimetric and ultracentrifugation
studies on water-soluble pegylated conjugates of bili-
rubin and fatty acids. Biochemistry 43, 15617–15632.
37 Passamonti S, Vanzo A, Vrhovsek U, Terdoslavich M,
Cocolo A, Decorti G & Mattivi F (2005) Hepatic

uptake of grape anthocyanins and the role of bilitrans-
locase. Food Res Int 38, 953–966.
38 Tischler ME, Friedrichs D, Coll K & Williamson JR
(1977) Pyridine nucleotide distributions and enzyme
mass action ratios in hepatocytes from fed and starved
rats. Arch Biochem Biophys 184, 222–236.
39 Pieper JA (2003) Overview of niacin formulations: dif-
ferences in pharmacokinetics, efficacy, and safety. Am J
Health Syst Pharm 60, S9–14; quiz S25.
40 Passamonti S & Sottocasa GL (1988) The quinoid
structure is the molecular requirement for recognition of
phthaleins by the organic anion carrier at the sinusoidal
plasma membrane level in the liver. Biochim Biophys
Acta 943, 119–125.
41 Kullak-Ublick GA, Ismair MG, Stieger B, Landmann
L, Huber R, Pizzagalli F, Fattinger K, Meier PJ &
Hagenbuch B (2001) Organic anion-transporting poly-
peptide B (OATP-B) and its functional comparison with
three other OATPs of human liver. Gastroenterology
120, 525–533.
42 Serra MA, Aparisi L, Garcia F, Del Olmo JA, Gilabert
MS, Rodriguez F, Escudero A, Wassel A & Rodrigo
JM (1997) Plasma sulfobromophthalein disappearance
in Gilbert’s syndrome. Hepatogastroenterology 44,
210–214.
43 Schirmer T, Bode W & Huber R (1987) Refined
three-dimensional structures of two cyanobacterial
C-phycocyanins at 2.1 and 2.5 A
˚
resolution. A

common principle of phycobilin–protein interaction.
J Mol Biol 196, 677–695.
44 Sottocasa GL, Lunazzi GC & Tiribelli C (1989) Isola-
tion of bilitranslocase, the anion transporter from liver
plasma membrane for bilirubin and other organic
anions. Methods Enzymol 174, 50–57.
45 Kullak-Ublick GA, Beuers U & Paumgartner G (1996)
Molecular and functional characterization of bile acid
transport in human hepatoblastoma HepG2 cells.
Hepatology 23, 1053–1060.
46 Stremmel W & Diede HE (1990) Cellular uptake of
conjugated bilirubin and sulfobromophthalein (BSP) by
the human hepatoma cell line Hep G2 is mediated by a
membrane BSP ⁄ bilirubin binding protein. J Hepatol 10,
99–104.
47 Miccio M, Baldini G, Basso V, Gazzin B, Lunazzi GC,
Tiribelli C & Sottocasa GL (1989) Bilitranslocase is the
protein responsible for the electrogenic movement of
sulfobromophthalein in plasma membrane vesicles from
rat liver: immunochemical evidence using mono- and
poly-clonal antibodies. Biochim Biophys Acta 981,
115–120.
S. Passamonti et al. Bilirubin uptake into HepG2 cells
FEBS Journal 272 (2005) 5522–5535 ª 2005 FEBS 5535