18F-FBHGal for asialoglycoprotein receptor imaging in a hepatic fibrosis mouse model

Article abstract of DOI:10.1016/j.bmc.2012.12.022

Quantification of the expression of asialoglycoprotein receptor (ASGPR), which is located on the hepatocyte membrane with high-affinity for galactose residues, can help assess ASGPR-related liver diseases. A hepatic fibrosis mouse model with lower asialoglycoprotein receptor expression was established by dimethylnitrosamine (DMN) administration. This study developed and demonstrated that 4-¹⁸F-fluoro-N-(6-((3,4,5-trihydroxy-6-(hydroxymethyl) tetrahydro-2H-pyran-2-yl)oxy)hexyl)benzamide (¹⁸F-FBHGal), a new ¹⁸F-labeled monovalent galactose derivative, is an asialoglycoprotein receptor (ASGPR)-specific PET probe in a normal and a hepatic fibrosis mouse models. Immunoassay exhibited a linear correlation between the accumulation of GalH-FITC, a fluorescent surrogate of FBHGal, and the amount of ASGPR. A significant reduction in HepG2 cellular uptake (P <0.0001) was observed using confocal microscopy when co-incubated with 0.5 μM of asialofetuin, a well known ASGPR blocking agent. Animal studies showed the accumulation of ¹⁸F-FBHGal in fibrosis liver (14.84 ± 1.10 %ID/g) was appreciably decreased compared with that in normal liver (20.50 ± 1.51 %ID/g, P <0.01) at 30 min post-injection. The receptor indexes (liver/liver-plus-heart ratio at 30 min post-injection) of hepatic fibrosis mice derived from both microPET imaging and biodistribution study were significantly lower (P <0.01) than those of normal mice. The pharmacokinetic parameters (T_1/2α, T_1/2β, AUC and Cl) derived from microPET images revealed prolonged systemic circulation of ¹⁸F-FBHGal in hepatic fibrosis mice compared to that in normal mice. The findings in biological characterizations suggest that ¹⁸F-FBHGal is a feasible agent for PET imaging of hepatic fibrosis in mice and may provide new insights into ASGPR-related liver dysfunction.

Full text of DOI:10.1016/j.bmc.2012.12.022

Bioorganic & Medicinal Chemistry 21 (2013) 912–921
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
18F-FBHGal for asialoglycoprotein receptor imaging in a hepatic
fibrosis mouse model
Hao-Wen Kao a, , Chuan-Lin Chen a, , Wen-Yi Chang a, Jenn-Tzong Chen b, Wuu-Jyh Lin b,
Ren-Shyan Liu c,d, Hsin-Ell Wang a,
⇑
a Department of Biomedical Imaging and Radiological Sciences, National Yang-Ming University, No. 155, Sec. 2, Linong St., Beitou District, Taipei 112, Taiwan, ROC
b Department of Isotope Application, Institute of Nuclear Energy Research, Taoyuan, Taiwan, ROC
c School of Medicine, National Yang-Ming University, Taipei, Taiwan, ROC
d National PET/Cyclotron Center, Veterans General Hospital, Taipei, Taiwan, ROC
a r t i c l e i n f o
a b s t r a c t
Article history:
Quantification of the expression of asialoglycoprotein receptor (ASGPR), which is located on the
hepatocyte membrane with high-affinity for galactose residues, can help assess ASGPR-related liver
diseases. A hepatic fibrosis mouse model with lower asialoglycoprotein receptor expression was estab-
lished by dimethylnitrosamine (DMN) administration. This study developed and demonstrated that
4-18F-fluoro-N-(6-((3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)hexyl)benzamide
Received 28 September 2012
Revised 7 December 2012
Accepted 7 December 2012
Available online 22 December 2012
(
18F-FBHGal), a new 18F-labeled monovalent galactose derivative, is an asialoglycoprotein receptor
Keywords:
(ASGPR)-specific PET probe in a normal and a hepatic fibrosis mouse models. Immunoassay exhibited
a linear correlation between the accumulation of GalH-FITC, a fluorescent surrogate of FBHGal, and the
amount of ASGPR. A significant reduction in HepG2 cellular uptake (P <0.0001) was observed using
Liver fibrosis
Asialoglycoprotein receptor
Galactose
confocal microscopy when co-incubated with 0.5 lM of asialofetuin, a well known ASGPR blocking agent.
MicroPET
Animal studies showed the accumulation of 18F-FBHGal in fibrosis liver (14.84 1.10 %ID/g) was
appreciably decreased compared with that in normal liver (20.50 1.51 %ID/g, P <0.01) at 30 min post-
injection. The receptor indexes (liver/liver-plus-heart ratio at 30 min post-injection) of hepatic fibrosis
mice derived from both microPET imaging and biodistribution study were significantly lower (P <0.01)
than those of normal mice. The pharmacokinetic parameters (T1/2a, T1/2b, AUC and Cl) derived from
microPET images revealed prolonged systemic circulation of 18F-FBHGal in hepatic fibrosis mice
compared to that in normal mice. The findings in biological characterizations suggest that 18F-FBHGal
is a feasible agent for PET imaging of hepatic fibrosis in mice and may provide new insights into
ASGPR-related liver dysfunction.
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
hypertension, ascites, encephalopathy, synthetic dysfunction, and
impaired metabolic capacity. Liver fibrosis is reversible,2,4 whereas
The chronic liver disease and cirrhosis, not including hepatoma,
were the 8th leading cause of death for men and the 10th for wo-
men in Taiwan in 2010.1 Liver fibrosis is a dynamic and complicat-
edly regulated wound-healing response, resulting from chronic
liver damage due to accumulation of extracellular matrix proteins,
which is a characteristic of most types of chronic liver diseases.2
Hepatitis B and C are the main causes of liver fibrosis. In addition
to viral-associated hepatitis, alcohol abuse, non-alcoholic steato-
hepatitis and autoimmune disease also contribute to the pathogen-
esis of liver fibrosis.3 The fibrotic response underlies virtually all
the complications of end-stage liver disease, including portal
advanced cirrhosis may not be completely reversible, and recovery
depends on the etiology and stage of the disease. Thus, early detec-
tion of liver fibrosis in patients has become an important clinical
issue for predicting the prognosis and early treatment to prevent
disease progression.
Traditionally, liver biopsy and histological examination is con-
sidered the ‘gold standard’ for identifying the cause of liver disease
and assessing the stage of fibrosis. However, several limitations
were recognized for this invasive procedure, such as the possibility
of sampling error, intra- and inter-observer variation, and only pro-
vides a static data.5 In recent years, several imaging approaches
have been adapted to detect liver fibrosis6 including ultrasonogra-
phy (FibroScan7 and acoustic radiation force impulse8), CT and
MRI. These morphologic imaging techniques mainly provide
information about structural aspects of liver fibrosis, such as the
⇑
Corresponding author. Tel.: +886 2 28267215; fax: +886 2 28201095.
E-mail addresses: hewang@ym.edu.tw, hewang2@gmail.com (H.-E. Wang).
These authors contributed equally to this work.
0968-0896/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2012.12.022

H.-W. Kao et al. / Bioorg. Med. Chem. 21 (2013) 912–921
913
changing in liver echogenicity and nodularity, degree of liver stiff-
2.1.1. Synthesis of 6-(acetoxymethyl)tetrahydro-2H-pyran-
2,3,4,5-tetrayl tetraacetate (1) ( /b = 1.9)
A mixture of b- -galactose (2.0 g, 11.1 mmol) and acetic anhy-
ness as well as signs of portal hypertension. Molecular imaging is a
new biomedical discipline that enables the visualization, charac-
terization, and quantification of biologic processes at the cellular
and molecular levels within living subjects. Thus, efforts to diag-
nose liver fibrosis down to the molecular level using non-invasive
strategy have direct clinical implications.
a
D
dride (10.0 mL, 106.0 mmol) in anhydrous pyridine (14 mL) was
stirred at ambient temperature for 24 h. The reaction mixture
was quenched by cold water (50 mL) and extracted by ethyl ace-
tate (50 mL ꢀ 3), washed with water (20 mL ꢀ 3) and brine. The or-
ganic layer was dried and evaporated under vacuum to give
compound 1 as yellow syrup (4.1 g, 95% yield). Rf = 0.54 (silica
gel 60F254 on an aluminum sheet, ethyl acetate/hexane = 3/2 as
the developing agent). 1H NMR (400 MHz, CDCl3): d 6.35 (br, 1H,
The asialoglycoprotein receptors (ASGPR) are renowned for
existing in the mammalian liver and located on the surface of
hepatocytes which contain 100,000–500,000 binding sites per
cell.9,10 The receptors play an essential role in maintaining serum
glycoprotein homeostasis by the recognition, binding, and endocy-
tosis of glycoprotein owning galactose residues on the terminal po-
sition of saccharide chain, such as asialoglycoproteins.11,12 After
internalized via clathrin-coated pits and fused with endosomes,
the asialoglycoproteins are released in the acidic environment of
the endosome and transported to lysosomes for degradation, while
the receptor is recycled back to the cell surface. In the previous re-
ports,13,14 Sawamura et al. demonstrated that the asialoglycopro-
tein receptor activity in the cirrhotic liver was about 28% of the
control value in patients. Since 1985, 99mTc-labeled neoglycoalbu-
min (99mTc-NGA)15,16 and 99mTc-labeled galactosyl-human serum
albumin (99mTc-GSA)17–19 were developed as single photon emis-
sion computed tomography (SPECT) imaging probes for acquiring
the asialoglycoprotein receptor scintigraphy to evaluate the
liver function. Fluorine-18 labeled galactosyl-neoglycoalbumin
H-1 ), 5.68 (d, J = 8.2 Hz, 0.5H, H-1b), 5.48 (br, 1H, H-4 ), 5.40 (d,
a
a
J = 8.2 Hz, 0.5H, H-4b), 5.34–5.29 (m, 2.5H, H-2 , H-2b, H-3 ), 5.05
a
a
(dd, J = 10.5, 3.4 Hz, 0.5H, H-3b), 4.35–4.29 (m, 1.5H, H-6b, H-5 ),
a
4.17–4.03 (m, 3.0H, 2 H-6 H-6b, H-5b), 2.14 (s, 1.6H, CH3COb),
a
,
2.14 (s, 3H, CH3CO ), 2.14 (s, 3H, CH3CO ), 2.03 (s, 1.6H, CH3COb),
a
a
2.02 (s, 3.2H, 2 CH3COb), 2.02 (s, 3H, CH3CO ), 2.00 (s, 3H, CH3CO ),
a
a
1.98 (s, 3H, CH3CO ), 1.97 (s, 1.6H, CH3COb).
a
2.1.2. Synthesis of 2-(acetoxymethyl)-6-bromotetrahydro-2H-
pyran-3,4,5-triyl triacetate (2)
A mixture of b-D-galactosyl pentaacetate 1 (3.9 g, 10.0 mmol) in
an anhydrous methylene chloride (100 mL) was added acetic acid
(4 mL), acetic anhydride (4 mL) and hydrobromic acid solution
(33% in acetic acid) under 0 °C. The reaction mixture was stirred
overnight at ambient temperature, then quenched with cold water
and extracted by methylene chloride (50 mL ꢀ 3). The organic
layer was dried and evaporated under vacuum to give compound
2 as yellow syrup (4.0 g, 98% yield). Rf = 0.61 (silica gel 60F254 on
an aluminum sheet, ethyl acetate/hexane = 1/2 as the developing
(
18F-FNGA)20 was presented as a radiotracer for positron emission
tomography (PET) scans. In recent years, polymer backbone radio-
tracers were also reported as ASGPR targeting agents, such as
chitosan21 and poly(vinylbenzyl-O-b-
D-galactopyranosyl-D-gluc-
onamide) (PVLA).22 Although the exact physiological role of ASGPR
has not been elucidated, previous studies suggested that an abnor-
mal cell-surface distribution of this receptor might occur in the li-
ver dysfunction such as cirrhosis or liver cancer.12,23 Evaluation of
ASGPR expression based on 99mTc-NGA scintigraphic imaging has
been employed for distinguishing and monitoring the graft and na-
tive liver functions in patients.24
agent). H NMR (400 MHz, CDCl3): d 6.67 (d, J = 4.0 Hz, 1H, H-10),
1
5.49 (d, J = 3.2 Hz, 1H, H-40), 5.38 (dd, J = 10.8, 3.2 Hz, 1H, H-30),
5.02 (dd, J = 10.8, 4 Hz, 1H, H-20), 4.46 (dd, J = 6.4, 6.4 Hz, 1H, H-
50), 4.16 (dd, J = 11.2, 6.4 Hz, 1H, H-60), 4.08 (dd, J = 11.2, 6.4 Hz,
1H, H-60), 2.13 (s, 3H, CH3CO), 2.10 (s, 3H, CH3CO), 2.03 (s, 3H,
CH3CO), 1.99 (s, 3H, CH3CO).
Although previous studies had suggested that the in vitro
ASGPRs binding affinity of multi- and di-valent galactoside were
104 to 105-fold and 10-fold higher than the monovalent ligand,25,26
2.1.3. Synthesis of 4-fluoro-N-(6-hydroxyhexyl)-benzamide
(FHB, 3)
6-Aminohexanol (1.0 g, 8.5 mmol), dissolved in dichlorometh-
ane (17 mL) containing triethylamine (3.6 mL, 26.0 mmol) was
added p-fluorobenzoyl chloride (1.0 mL, 8.5 mmol) dropwise at
0 °C. The mixture was stirred overnight at ambient temperature.
The crude product was washed with cold pure water (30 mL ꢀ 2),
dried and evaporated under vacuum. After recrystallized in a
mixed solvent of dichloromethane and hexane, the desired com-
pound was obtained as a white crystalline solid (1.8 g, 90% yield).
1H NMR (500 MHz, DMSO-d6): d 8.44 (br, 1H, NH), 7.89 (dd,
J = 8.5 Hz, JHF = 6.0 Hz, 2H, Ar), 7.26 (dd, J = 8.5 Hz, JHF = 7.0 Hz,
2H, Ar), 3.38–3.35 (m, 2H, CH2O), 3.38–3.35 (m, 1H, OH), 3.21
(dt, J = 6.5, 6.5 Hz, 2H, NCH2), 1.51–1.48 (m, 2H, NCCH2), 1.41–
1.38 (m, 2H, CH2CO), 1.30–1.21 (m, 4H, CH2); 13C NMR (125 MHz,
DMSO-d6): 165.0 (C@O), 163.8 (d, JCF = 246.8 Hz, F-C), 131.2 (Ar),
29.7 (d, JCF = 9.1 Hz, Ar), 115.1 (d, JCF = 21.5 Hz, Ar), 60.6 (CH2OH),
39.1 (NHCH2), 32.4 (CH2CH2OH), 29.1 (NHCH2CH2), 26.4, 25.2
a
18F-labled monoantennary galactoside may still be useful for
noninvasive evaluation of ASGPR-related hepatic fibrosis owing
to the much rapid pharmacokinetics than that of multivalent
macromolecules in vivo. This study reported for the first time a
monoantennary galactose derivative 4-18F-fluoro-N-(6-((3,4,5-tri-
hydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)hexyl)benz-
amide (18F-FBHGal) as a ASGPR-specific PET probe. The biological
evaluations demonstrate that 18F-FBHGal PET can characterize
the reduction of ASGPR in liver fibrosis in a dimethylnitrosamine
(DMN)-treated mouse model and may have the potential to evalu-
ate ASGPR-related liver disease.
2. Materials and methods
2.1. Synthesis of b-D
-galactosyl precursor for 18F labeling and
the authentic standard of FBHGal
(CH2); HRESI-MS: calcd for C13H18FNNaO2 262.1214 (M+Na+),
+
found 262.1205 (M+Na+).
The synthetic scheme is shown in Figure 1. Nuclear magnetic
resonance (NMR) spectra were recorded on a Bruker Avance III
400 MHz NMR spectrometer or Varian 500 MHz NMR spectrome-
ter and referenced to CDCl3 (1H NMR: CDCl3 at 7.24 ppm and 13C
NMR: CDCl3 at 77.0 ppm), DMSO-d6 (1H NMR: DMSO-d6 at
2.49 ppm and 13C NMR: DMSO-d6 at 39.5 ppm) and methanol-d4
(1H NMR: methanol-d4 at 3.30 ppm and 13C NMR: methanol-d4 at
49.0 ppm).
2.1.4. Synthesis of 4-fluoro-N-(6-((3,4,5-trihydroxy-6-
(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)hexyl)
benzamide (FBHGal, 4a)
A mixture of compound 2 (500.0 mg, 1.2 mmol), compound 3
(438.0 mg, 1.8 mmol), silver carbonate (330.0 mg, 1.2 mmol), and
activated 4 Å molecular sieves (1.0 g) was added in anhydrous
dichloromethane (2.5 mL) at 0 °C. After stirring for 24 h under an

914
H.-W. Kao et al. / Bioorg. Med. Chem. 21 (2013) 912–921
Figure 1. Synthesis of 18F-labeled and FITC-conjugated monomeric galactose derivatives, 18F-FBHGal and GalH-FITC.
argon atmosphere at ambient temperature, the result solution was
filtered through celite, and washed with dichloromethane (2 mL).
The filtrate was washed with saturated aqueous sodium bicarbon-
ate (30 mL), dried and evaporated under vacuum. The crude prod-
crystalline solid (3.85 g, 90% yield). 1H NMR (400 MHz, CDCl3): d
7.34 (m, 5H, Ar), 5.07 (s, 2H, PhCH2), 4.73 (br, NH), 3.61 (t,
J = 6.4 Hz, 2H, CH2O), 3.16 (dt, J = 6.8, 6.8 Hz, 2H, NCH2), 1.61–
1.32 (m, 8H, CH2); 13C NMR (100 MHz, CDCl3): 156.4 (C, C@O),
136.6, 128.5, 128.1 (CH, Ar), 66.6 (CH2, PhCH2), 62.7 (CH, CH2OH),
40.9 (CH, NHCH2), 32.5 (CH2, CH2CH2OH), 29.9 (CH2, NHCH2CH2),
26.3 (CH2, CH2CH2CH2OH), 25.3 (CH2, NCH2CH2CH2).
ucts (a mixture of
a- and b-galactopyranosides) were purified by
flash chromatography (eluent, ethyl acetate/hexane = 1/1). The
crude product re-dissolved in anhydrous methanol (1.5 mL) was
treated with 0.5 M sodium methoxide in methanol (0.02 mL). After
stirring for 24 h, the reaction mixture was filtered through celite
and evaporated to give compound 4a as a white solid (122.4 mg,
2.1.6. Synthesis of 2-(acetoxymethyl)-6-((6-(((benzyloxy)
carbonyl)amino)hexyl)oxy)tetrahydro-2H-pyran-3,4,5-triyl
triacetate (6)
25% yield). 1H NMR (500 MHz, methanol-d4):
d
7.85 (dd,
J = 8.0 Hz, JH-F = 5.6 Hz, 2H, H-10, H-12), 7.17 (dd, J = 8.0 Hz, JH-F
=
Starting from 6-(N-benzyloxycarbonylamino)-hexanol, com-
pound 6 was synthesized in 26% yield using a procedure similar
to that of compound 4a. 1H NMR (400 MHz, CDCl3): d 7.32 (m,
5H, Ar), 5.36 (d, J = 3.6 Hz, 1H, H-40), 5.17 (dd, J = 10.4, 8.0 Hz, 1H,
H-20), 5.06 (s, 2H, Ar-CH2), 4.99 (dd, J = 10.6, 3.6 Hz, 1H, H-30),
4.41 (d, J = 8.0 Hz, 1H, H-10), 4.16 (dd, J = 11.2, 6.8 Hz, 1H, H-60),
4.10 (dd, J = 11.2, 6.8 Hz, 1H, H-60), 3.88–3.83 (m, 2H, H-1, H-50),
3.43 (dt, J = 9.6, 6.8 Hz, 1H, H-1), 3.16 (dt, J = 13.0, 6.8 Hz, 2H, H-
6), 2.13 (s, 3H, CH3CO), 2.06 (s, 6H, CH3CO), 1.95 (s, 3H, CH3CO),
1.54–1.28 (m, 8H, H-2, H-3, H-4, H-5); 13C NMR (100 MHz, CDCl3):
170.3, 170.2, 170.1, 169.3, 156.3 (C, C@O), 136.6 (C, C-9), 128.4,
128.0 (CH, Ar), 101.2 (CH, C-10), 70.9 (CH, C-30), 70.5 (CH, C-50),
70.0 (CH2, C-1), 68.8 (CH, C-20), 67.0 (CH, C-40), 66.5 (CH2, C-8),
61.2 (CH2, C-60), 40.8 (CH2, C-6), 29.2, 26.3, 25.4 (CH2, C-2, C-3, C-
4, C-5), 20.7, 20.6, 20.5 (CH3, CH3CO); ESI-MS: calcd for
8.0 Hz, 2H, H-9, H-13), 4.20 (d, J = 7.0 Hz, 1H, H-10), 3.89 (dt,
J = 9.5, 6.5 Hz, 1H, H-1), 3.83 (d, J = 3.0 Hz, 1H, H-40), 3.76–3.69
(m, 2H, H-60), 3.54 (dt, J = 9.5, 6.5 Hz, 1H, H-1), 3.52–3.44 (m, 3H,
H-20, H-30, H-50), 3.36 (t, J = 7.0 Hz, 2H, H-6), 1.65–1.60 (m, 4H, H-
2, H-5), 1.44–1.40 (m, 4H, H-3, H-4); 13C NMR (125 MHz, metha-
nol-d4): 168.9 (C, C@O), 166.0 (d, J = 22.0 Hz, C, C-11), 132.1 (C,
C-8), 130.8 (d, J = 9.0 Hz, C, C-9, C-13), 116.3 (d, J = 9.0 Hz, C, C-
10, C-12), 104.9 (CH, C-10), 76.5 (CH, C-50), 75.0 (CH, C-30), 72.5
(CH, C-20), 70.6 (CH2, C-1), 70.2 (CH, C-40), 62.4 (CH2, C-60), 40.9
(CH2, C-6), 30.6 (CH2, C-2), 30.3 (CH2, C-5), 27.8 (CH2, C-4), 26.7
+
(CH2, C-3); HRESI-MS: calcd for C19H28FNNaO7 424.1742
(M+Na+), found 424.1730 (M+Na+).
2.1.5. Synthesis of benzyl-(6-hydroxyhexyl)carbamate (5)
6-aminohexanol (5.0 g, 42.7 mmol) dissolved in a mixture of
tetrahydrofuran (50 mL) and water (15 mL) was added benzylchlo-
roformate (7.3 mL, 51.2 mmol) dropwise at 0 °C. Sodium hydroxide
(1 N, 12 mL) was added at the same temperature for 30 min to
keep the pH of the reaction solution at 8–9 then stirred at ambient
temperature for additional 30 min. The reaction mixture was ex-
tracted with diethyl ether (30 mL ꢀ 3), dried and evaporated under
vacuum. After recrystallized in a mixed solvent of dichloromethane
and hexane, the desired compound was obtained as a white
C
28H39NNaO12 604.2 (M+1), found 604.4 (MH+). HRESI-MS: calcd
for C28H39NNaO12 604.2364 (M+Na+), found 604.2355 (M+Na+).
+
2.1.7. Synthesis of 2-(acetoxymethyl)-6-((6-aminohexyl)oxy)-
tetrahydro-2H-pyran-3,4,5-triyl triacetate (7)
A mixture of above product 6 (200.0 mg, 0.34 mmol) and 10%
Pd/C (50 mg) was added in anhydrous methanol (1.7 mL) and stir-
red under 1 atm of hydrogen for 24 h. The reaction mixture was
purged through celite and evaporated under vacuum to give a thick

H.-W. Kao et al. / Bioorg. Med. Chem. 21 (2013) 912–921
915
oil of 7 (150.0 mg, 98% yield). 1H NMR (400 MHz, CDCl3): d 5.35 (d,
J = 3.6 Hz, 1H, H-40), 5.13 (dd, J = 10.4, 8.0 Hz, 1H, H-20), 4.99 (dd,
J = 10.6, 3.6 Hz, 1H, H-30), 4.43 (d, J = 8.0 Hz, 1H, H-10), 4.15 (dd,
J = 11.2, 6.4 Hz, 1H, H-60), 4.08 (dd, J = 11.2, 7.2 Hz, 1H, H-60),
3.90–3.81 (m, 2H, H-1, H-50), 3.44 (dt, J = 9.6, 6.8 Hz, 1H, H-1),
2.96 (t, J = 7.6 Hz, 2H, H-6), 2.11 (s, 3H, CH3CO), 2.06 (s, 3H, CH3CO),
2.04 (s, 3H, CH3CO), 1.96 (s, 3H, CH3CO), 1.61–1.12 (m, 8H, H-2, H-
3, H-4, H-5); 13C NMR (100 MHz, CDCl3): 170.4, 170.2, 170.1, 169.6
(C, C@O), 101.2 (CH, C-10), 71.0 (CH, C-20), 70.8 (CH, C-50), 70.5 (CH2,
C-1), 68.9 (CH, C-30), 66.9 (CH, C-40), 61.1 (CH2, C-60), 39.4 (CH2, C-
6), 29.3, 29.1, 26.5, 25.5 (CH2, C-2, C-3, C-4, C-5), 20.8, 20.7, 20.6
(CH3, CH3CO); ESI-MS: calcd for C20H34NO10 448.2 (M+1), found
448.3 (MH+).
min, 18F-FBHGal retention time 10.5 min). The desired fraction
was collected, evaporated to dryness, re-dissolved in normal saline
and filtered through a 0.22 lm membrane filter (Millex-OR, Milli-
pore) into a sterile vial to afford the final product of 18F-FBHGal
(4b) injection. The final product was identified by co-injection of
an aliquot of product solution with authentic FBHGal (4a) using
an analytical HPLC system (PurospherÒ STAR RP-18,
5 lm,
250 mm ꢀ 4 mm, Merck, Darmstadt, Germany; mobile phase: ace-
tonitrile/0.1% acetic acid buffer = 1/3 (vol/vol); flow rate 0.8 mL/
min). The stability of 18F-FBHGal injection was determined by
HPLC after storage at room temperature for 3 h.
2.3. ASGP receptor linked to uptake of GalH-FITC
2.1.8. Synthesis of 2-(6-hydroxy-3-oxo-3H-xanthen-9-yl)-5-(3-
(6-((-3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-
2-yl)oxy)hexyl)thioureido)benzoic acid TEA (GalH-FITC, 8)
A mixture of compound 7 (200.0 mg, 0.45 mmol) and FITC
(174.0 mg, 0.45 mmol) were dissolved in anhydrous methanol
Recombinant mouse ASGPR-1 (R&D Systems, Inc., Minneapolis)
were coated (250, 500, 750 and 1500 ng/well) in 96 well plates.
Sixteen hours later, an excess of FITC-conjugated galactose deriva-
tive (GalH-FITC, 250 ng/well) was added in the plates and incu-
bated for 1 h at 37 °C. After extensive wash with PBS,
fluorescence was counted in a TECAN Magellan microplate reader
with an excitation wavelength of 495 nm and an emission wave-
length of 521 nm.
(9 mL) containing triethylamine (63 lL, 0.45 mmol) and stirred in
the dark for 24 h at ambient temperature. The result solution
was evaporated, re-dissolved in dichloromethane and purified by
silica gel chromatography (eluent, dichloromethane/2-propa-
nol = 5/1). The crude product re-dissolved in methanol (25 mL)
2.4. Imaging of receptor-mediated endocytosis by confocal
microscopy
was treated with 0.5 M sodium methoxide in methanol (132 lL)
at room temperature for 24 h to achieve deacetylation. The reac-
tion solution was evaporated under vacuum to afford compound
8
The specific binding of galactose derivative to ASGP receptor
was examined on HepG2 hepatocellular carcinoma cells which ex-
pressed high levels of ASGPR.29 One day before the experiments,
HepG2 cells were seeded at a density of 2 ꢀ 105 cells/well into 6-
well plates containing collagen-coated glass coverslips. The cells
(GalH-Gal) as a
red solid (42.5 mg, 14% yield). 1H NMR
(400 MHz, methanol-d4): d 7.98 (s, 1H, H-15), 7.71 (d, J = 8.0 Hz,
1H, H-24), 7.19 (d, J = 8.0 Hz, 1H, H-26), 7.14–7.12 (m, 2H, H-9,
H-27), 6.62–6.59 (m, 4H, H-16, H-18, H-21, H-23), 4.21 (d,
J = 8.0 Hz, 1H, H-10), 3.91 (dt, J = 9.6, 6.8 Hz, 1H, H-1), 3.82 (d,
J = 2.4 Hz, 1H, H-40), 3.75–3.70 (m, 2H, H-60), 3.62–3.51 (m, 6H,
H-20, H-30, H-50, H-6, H-1), 3.01 (q, J = 7.6 Hz, 6H, CH2), 1.71–1.32
(m, 8H, H-2, H-3, H-4, H-5), 1.16 (t, J = 7.6 Hz, 9H, CH3); 13C NMR
(100 MHz, methanol-d4): 182.1 (C, C@S, C-7), 176.6, 173.0 (C,
C@O, C-11, C17, C22), 158.9 (C, C-8, C-10), 132.4 (CH, C-9, C2–7),
130.6 (CH, C-26), 126.3 (CH, C-15), 124.6 (CH, C-24), 121.6 (C, C-
12, C-13), 114.6 (C, C-14, C-19, C-20, C25), 105.0 (CH, C-10), 104.1
(C, C-16, C-18, C-21, C23), 76.6 (CH, C-50), 75.0 (CH, C-30), 72.6
(CH2, C-30), 70.7 (CH, C-1), 70.3 (CH, C-40), 62.5 (CH2, C-60), 47.4
(CH2), 40.6 (CH2, C-6), 30.7, 29.9, 27.8, 26.8 (CH2, C-2, C-5, C-3, C-
4), 9.2 (CH3); HRESI-MS: calcd for C33H37N2O11S 669.2113 (M+1),
found 669.2191 (MH+).
were incubated with GalH-FITC (0.1 mL, 100 lM) in serum free
MEM (0.9 mL) on ice for 1.5 h in the dark. After the binding step,
the cells were washed extensively with cold PBS. Then fresh, pre-
warmed, MEM containing 10% fetal bovine serum was adminis-
tered, and the cells were incubated at 37 °C for 30 min in a
humidified CO2 atmosphere (5%, v/v), leading to the internalization
of the receptor-bound compounds into the cells. After the internal-
ization step, the cells were washed twice with PBS and then fixed
with 4% paraformaldehyde in PBS for 10 min at room temperature.
After fixation, the coverslips were washed abundantly with PBS
and mounted upside down. For blocking experiments, cells were
incubated with asialofetuin (ASF, purchased from Sigma–Aldrich,
0.5 mg) for 30 min prior to incubation with GalH-FITC. Images of
fixed cells were acquired on an Olympus FV-1000 laser scanning
confocal system (FluoView FV1000; Olympus, Tokyo, Japan) with
objective UPLSAPO 60ꢀ O NA: 1.35. Mean fluorescent intensity
was evaluated with the ImageJ program.
2.2. Synthesis of 18F-FBHGal (4b)
Compound 7 was conjugated with N-succinimidyl-4-18F-fluor-
obenzoate (18F-SFB) to afford the title compound 18F-FBHGal. 18F-
SFB was synthesized from tert-butyl 4-N,N,N-trimethylammonium
benzoate triflate (Sigma–Aldrich) as previously reported with min-
or modifications following the literature method.27,28 The overall
radiochemical yield of 18F-SFB was about 40% (decay corrected,
EOB) and the radiochemical purity was >95% determined using
TLC method (silica gel 60F254 on an aluminum sheet, dichloro-
methane as the developing agent, Rf = 0.69). The purified 18F-SFB
was added to a acetonitrile solution containing compound 7
2.5. Establishment of hepatic fibrosis in mouse
All experimental protocols involving animals were reviewed
and approved by the institutional animal experimentation com-
mittee of the National Yang-Ming University. Male C57BL/6 mice
(weighing 20–25 g, 8-weeks-old) were purchased from the Na-
tional Laboratory Animal Center, Taiwan.
Hepatic fibrosis was induced using dimethylnitrosamine (DMN)
following the literature method.30,31 DMN (5 mg/mL in normal sal-
ine) at a dose of 10 mg/kg body weight was administered intraper-
itoneally on three consecutive days a week for 4 weeks. The sham
group was injected with an equivalent amount of 0.9% normal sal-
ine with the same frequency. To validate the hepatic fibrosis, mice
were sacrificed at 4 weeks after treated with DMN. The liver tissues
were fixed with 4% phosphate-buffered paraformaldehyde and
embedded in paraffin for Masson’s trichrome staining.
(300
ture was incubated at 50 °C for 30 min and hydrolyzed with so-
dium methoxide (1 M solution in methanol, 20 L) at 60 °C for
lg) and N,N-diisopropylethylamine (20 lL). The reaction mix-
l
15 min. After dilution with 1 mL of 0.1% acetic acid solution, the
mixture was purified by semi-preparative HPLC system (Puro-
spherÒ STAR RP-18, 5
Germany). A mixed solvent of acetonitrile and 0.1% acetic acid buf-
l
m, 250 mm ꢀ 10 mm, Merck, Darmstadt,
fer (1/3, vol/vol) was used as the mobile phase (flow rate, 3 mL/

916
H.-W. Kao et al. / Bioorg. Med. Chem. 21 (2013) 912–921
2.6. Measurement of ASGPR protein content
each group of mice (4 mice per group) for 60 min after intravenous
injection of 7.4 0.2 MBq of 18F-FBHGal. Time–activity curves of
heart and liver were generated from regions of interest (ROIs).
The receptor index was derived by dividing the radioactivity of liver
ROI by that of liver-plus-heart ROIs 30 min p.i. The pharmacokinetic
The frozen liver tissue pieces (about 100 mg) of normal and he-
patic fibrosis mice were homogenized in 0.25 M sucrose containing
protease inhibitors (Complete Mini, Roche, Indianapolis, Indiana,
USA), and the concentration of protein were determined using
the Bio-Rad Protein Assay (Bio-Rad Laboratories, Hercules, CA,
USA). An aliquot of homogenized suspension (containing 700 lg
protein) was incubated with mouse ASGPR-1 primary antibody
(polyclonal goat IgG, AF2755, R&D Systems, Minneapolis, MN,
parameters, including half-life (T1/2a and T1/2b), maximum blood
concentrations (Cmax), blood clearance (Cl), and areas under curves
(AUCs), were estimated after fitting a nonlinear regression curve to
the mean of heart radioactivity at each time point using the Win-
Nonlin software (version 5.3; Pharsight Corp., Mountain View, CA,
USA). Two-compartmental analysis (model, 7; input, iv-bolus)
was used with the log/linear trapezoidal rule.
USA) for 2 h at 4 °C, then 30
Cruz Biotechnology, Santa Cruz, CA, USA) was added. After over-
night incubation, beads were washed five times (500 L lysis buf-
fer each), and centrifuged at 2000g for 5 min. Proteins were eluted
by adding 50 L sample buffer and heating at 100 °C for 10 min.
lL of Protein A/G PLUS-Agarose (Santa
l
3. Results
l
The samples were electrophoresed in a 10% SDS–polyacrylamide
gel under reducing conditions and then transferred onto a polyvi-
nylidene fluoride membrane (Perkin–Elmer Life Sciences, Boston,
MA, USA). After blocking in 4% low-fat milk, the membrane was
incubated sequentially with the mouse ASGPR-1 primary antibody
(1:2000) followed by three times wash with PBS-T (PBS containing
0.05% Tween 20) and again incubated at room temperature for 2 h
with horseradish peroxidase-conjugated anti-goat IgG secondary
antibody (1:4000, Sigma, A5420, St. Louis, MO, USA). After incuba-
tion, the membrane was washed with PBS-T for five times and
visualized by enhanced chemiluminescence detection according
to the manufacturer’s instructions (T-Pro Biotechnology, Taipei,
Taiwan). The images were detected using LAS-4000 mini image
analyzer system (Fujifilm, Tokyo, Japan).
3.1. Chemistry and radiochemistry
Starting from D-galactose, the b-D-galactosyl precursor (7) suit-
able for 18F labeling can be obtained via a four-step synthesis with
an overall yield of 25% (Fig. 1). The authentic FBHGal (4a) was pre-
pared similarly with an overall yield of 22%. All reaction products
were characterized by 1H NMR, 13C NMR and ESI-MS analysis. After
coupling the precursor (7) with 18F-SFB, 18F-FBHGal (4b) was ob-
tained with a high radiochemical purity (>98%, Fig. 2). The specific
activity was 0.23 GBq/lmol as determined with HPLC using the
standard addition method. The radiochemical yield of 18F-FBHGal
(decay corrected, EOB) was 20–25% after about 10 runs. The
in vitro stability of 18F-FBHGal in physiological saline at ambient
temperature was good, the radiochemical purity was >98% after
3 h incubation (Fig. 2).
2.7. In vivo biodistribution studies
3.2. Establishment of hepatic fibrosis in mouse
In vivo biodistribution study was carried out in normal mice
and hepatic fibrosis mice after intravenous injection of 18F-FBHGal
After intermittent intraperitoneal injections (3 doses per week)
of a small amount of DMN to mice for 4 weeks, histological liver
examinations showed progressive deposition of collagen fiber in
the liver (Fig. 3A). The liver parenchyma was uniformly divided
into fibrotic nodules by fibrous septa. The results revealed exten-
sive fibrosis accompanied with significant hepatocyte damage in
the livers of DMN-treated mice. Although treatment-related body
weight loss was noticed in fibrosis mice (Fig. 3B), the body weight
change was less than 10% after 4-week DMN treatment.
(3.7 MBq in 100 lL normal saline). At designated time points (5
and 30 min), mice (n = 4 at each time point) were sacrificed. The
tissues and organs of interest were collected, wet weighed and as-
sayed for radioactivity with a gamma counter (Wallac 1470 WIZ-
ARD Automatic Gamma Counter, Perkin–Elmer Life Science,
Boston, MA, USA). The uptake of 18F-FBHGal in tissues was ex-
pressed in counts per minute (cpm) corrected with decay and nor-
malized to the percentage injected dose per gram of tissue (%ID/g).
The receptor index was calculated by dividing the radioactivity of
liver by that of liver plus heart 30 min post-injection (p.i.).
3.3. ASGP receptor linked to uptake of GalH-FITC
2.8. Metabolites analysis
To demonstrate the correlations between the fluorescent probe
GalH-FITC and the target receptor ASGPR, we incubated GalH-FITC
with different amounts of ASGPR and the data were calculated
using a linear regression (Fig. 4A). High linear correlations between
the amount of coated ASGPR and the retention of GalH-FITC in the
wells were observed (r2 = 0.99, P <0.01), suggested that the accu-
mulation of monomeric galactose derivatives (either GalH-FITC
or 18F-FBHGal) may provide appropriate estimation of ASGPR
expression level.
Urine samples (100
after a bolus injection of 18F-FBHGal (7.4 MBq) were passed
through a 0.45 m nylon membrane filter (Pall Gelmann Labora-
lL) collected from normal mice at 30 min
l
tory, Ann Arbor, MI, USA). The radioactive components in the fil-
trate were assayed using the above-mentioned analytical HPLC
system. The authentic compounds of possible metabolites, such
as 4-fluoro-N-(6-hydroxyhexyl)-benzamide (FHB), 4-fluorobenzoic
acid (FBA) and N-succinimidyl-4-fluorobenzoate (SFB) were pur-
chased form Aldrich (Sigma–Aldrich, Germany) or synthesized
and identified using nuclear magnetic resonance. The retention
time of authentic FBHGal, FHB, FBA and SFB was about 8.7, 14.0,
17.9 and 23.5 min, respectively.
3.4. Imaging receptor-mediated endocytosis by confocal
microscopy
HepG2 cells were incubated with GalH-FITC for 1.5 h on ice to al-
low binding to the receptors on cell membrane. In the washing step,
unbound ligand was removed, and the cells were incubated for an
additional 30 min at 37 °C to allow receptor-mediated endocytosis
of bound ligand. Significant green fluorescence emission in the cyto-
plasm revealed receptor-mediated accumulation of GalH-FITC in
HepG2 cells. Fluorescence was not observed in the nuclei (Fig. 4B).
2.9. MicroPET imaging and pharmacokinetic studies
The microPET imaging was performed on a microPET R4 scanner
(Concorde Microsystems, Inc., Knoxville, TN, USA) on Day 0, Day14
and Day 28 after treatment with DMN. The mice were anesthetized
with 1% isoflurane in oxygen. Dynamic imaging was conducted to

H.-W. Kao et al. / Bioorg. Med. Chem. 21 (2013) 912–921
917
Figure 2. HPLC chromatograms of authentic FBHGal and some possible metabolites (FBA, SFB and FHB). Assay of radioactive components in 18F-FBHGal injection (after 3-h
storage) and in urine sample (obtained at 30 min p.i.) revealed the in vitro and in vivo stability of 18F-FBHGal.
Figure 3. Effect of intraperitoneal injection of DMN (10 mg/kg of body weight) on liver histology and body weight change of the hepatic fibrosis mouse model. (A)
Histological examination of liver tissues. After 4-week DMN or saline treatment (sham group), liver tissues were dissected, fixed, stained with Masson’s trichrome (MT) and
examined. Lower photographs show higher original magnification (40ꢀ) of respective upper photographs (10ꢀ). Arrows indicate fibrosis region. (B) The body weight loss
during 12 doses of DMN and saline in C57B/L6 mice.
Asialofetuin (ASF), a natural glycoprotein having triantennary
galactose terminal sugar chains, is known as an excellent ligand
molecule selectively recognized by ASGPR. In the control group,
ASF was added to saturate the binding site of ASGPR during the up-
take period. The subcellular localization of GalH-FITC was largely
reduced when co-incubated with a blocking dose of ASF (Fig. 4B).
After incubation with GalH-FITC for 30 min, the average green fluo-
rescence emission per HepG2 cell was about 7.1-fold (48.3 vs 6.82,
P <0.0001) higher than those co-incubated with an equimolar con-
centration of ASF (Fig. 4C).
liver cirrhosis.13,14 In this study, the densitometric analysis of the
immunoblots showed that ASGPR content in the liver of hepatic
fibrosis mouse was decreased by an average of 23% compared with
that of normal mouse (Fig. 5A). The results revealed that adminis-
tration of DMN not only induce hepatic fibrosis, but also reduce the
ASGPR expression in liver.
3.6. In vivo biodistribution
The biodistribution studies of 18F-FBHGal were conducted in
both normal and DMN-induced hepatic fibrosis mouse models to
evaluate the distribution pattern in vivo (Table 1). At 30 min p.i.,
the liver uptake was 20.50 1.51 %ID/g in normal mice, while that
in the DMN-induced hepatic fibrosis mice was significantly re-
duced (14.84 1.10 %ID/g, P <0.01). The receptor index was
0.95 0.003 in normal mice, also remarkably higher than that in
hepatic fibrosis mice (0.89 0.01, P <0.001). At 30 min after
3.5. Measurement of ASGPR protein content
For human, the expression level of ASGPR was declined and the
distribution was abnormal in cirrhosis liver.13,32 Sawamura et al.
reported that the density of ASGPR in liver could be largely reduced
(maybe up to 70% reduction) in patients depends on the severity of

918
H.-W. Kao et al. / Bioorg. Med. Chem. 21 (2013) 912–921
Figure 4. The selectivity and specificity of GalH-FITC to ASGPR in vitro. (A) Fluorescence emission from bound GalH-FITC was linearly proportional to the amount of coated
ASGPR. (B) Confocal images of the HepG2 tumor cells after incubation with 10 M of GalH-FITC with/without ASF. Significant green fluorescence emission in the cytoplasm
revealed receptor-mediated accumulation of GalH-FITC in HepG2 cells. Fluorescence was not observed in the nuclei. In each picture, top is the fluorescence images and
l
bottom is the optical merged images, respectively. Bars represent 10 lm. (C) Relative fluorescence intensity of GalH-FITC per cell derived from the confocal images shown in
B. ⁄P <0.01.
injection of 18F-FBHGal in normal mice, high radioactivity collected
in urine (1171.40 269.07 %ID/g) suggested that excretion of 18F-
FBHGal was predominantly through the renal pathway. However,
partial elimination via hepatobiliary system (43.89 5.82 %ID/g
in bile) was also observed. Less rapid blood clearance of 18F-FBHGal
in the hepatic fibrosis mice was noticed. The radioactivity reten-
tion in kidney and blood of the hepatic fibrosis mice (8.13 1.19
%ID/g in kidney; 2.61 0.37 %ID/g in blood) at 30 min p.i. was sig-
nificantly higher than those of the normal mice (4.6 0.92 %ID/g in
kidney, P <0.01; 1.60 0.24 %ID/g in blood, P <0.01).
18F-FBHGal accounted for ꢁ80% of radioactivity in the urine at
30 min p.i., indicated high in vivo stability in the mouse model.
None of the above-mentioned possible radioactive metabolites,
for example, FHB, FBA, SFB, were detected in the urine.
3.8. MicroPET imaging and pharmacokinetic studies
A 1-h dynamic microPET scan of normal mouse was conducted
immediately after 18F-FBHGal injection via the tail vein. The
images revealed predominant hepatic uptake began soon after tra-
cer injection, increased rapidly, reached plateau within 10 min and
only slightly declined till 1 h p.i. The noted accumulation in kidney
and bladder revealed that renal clearance is the predominant route
of excretion. The nonspecific uptake in other organs was at very
low level in normal mice. Representative coronal and transaxial
3.7. Metabolites analysis
The radioactive metabolites in urine after administration of
18F-FBHGal in normal mouse were assayed by HPLC (Fig. 2). Intact

H.-W. Kao et al. / Bioorg. Med. Chem. 21 (2013) 912–921
919
Figure 5. 18F-FBHGal microPET imaging of normal and hepatic fibrosis mice. (A) Western blotting of ASGPR-1 in liver tissues. (B) Coronal (top) and transaxial (bottom) images
of normal and hepatic fibrosis mice at 5 and 30 min after intravenous injection of ꢁ3.7 MBq of 18F-FBHGal; L, liver; H, heart. (C) Time–activity curves of 18F-FBHGal in heart of
normal mice and hepatic fibrosis mice derived from dynamic microPET images. (D) The receptor index derived from microPET images. (n = 4 per group; bars represent
means SD). ⁄P <0.01, compared with control mice group by student paired t test.
Table 1
Biodistribution of 18F-FBHGal in normal and hepatic fibrosis mice at 5 and 30 min after intravenous injection of 3.7 MBq of 18F-FBHGal
Normal mice
Induced liver fibrosis mice
30 min
2.61 0.37*
5 min
30 min
5 min
5.83 0.50*
Blood
Heart
Lung
Stomach
SI
3.82 0.69*
1.60 0.24*
1.53 0.06*
1.83 0.40*
2.50 0.87
1.14 0.05*
0.83 0.20
2.42 0.12*
3.32 0.29*
2.83 0.29
7.86 1.03*
2.90 0.45
2.17 0.18*
4.15 0.38*
1.54 0.10*
1.32 0.62
32.67 1.46
99.33 92.10
1.73 0.11*
1.12 0.22
1.50 0.57
3.18 1.33
1.19 0.22
0.75 0.27
1.59 0.30*
1.00 0.07*
0.77 0.13
8.13 1.19*
37.14 5.92
294.80 127.68
1.08 0.32
14.84 1.10*
0.89 0.01*
0.88 0.05
2.86 0.24
0.90 0.12
5.01 0.11*
2.04 0.32
LI
Spleen
Pancreas
Muscle
Bone
Kidney
Bile
Urine
Feces
Liver
RI
1.33 0.07*
2.72 0.18*
0.95 0.10*
0.96 0.04
0.70 0.15
0.93 0.10*
0.66 0.09*
0.67 0.08
30.09 7.95
11.84 0.52
385.32 289.12
0.76 0.09
4.66 0.92*
43.89 5.82
1171.40 269.07
1.98 0.69
425.56 149.09
1.16 0.31
24.39 3.38
20.50 1.51*
0.95 0.003*
18.86 2.85
Expressed as percentage injection dose per gram of organ (%ID/g). Each value represents the mean SD (n = 4). SI, small intestine; LI, large intestine; RI, receptor index.
*
Significant difference between control mice and hepatic fibrosis group, two-tailed t test, P <0.01.
microPET images of the normal and hepatic fibrosis mice at 5–10
and 25–30 min p.i. were shown in Figure 5B. The uptake levels of
interested organs were quantitated by measuring ROIs encompass-
ing the organ in the microPET coronal images. The receptor index
was 0.74 0.05 in normal mice, again appreciably higher than that
in hepatic fibrosis mice (0.46 0.09, n = 4 in each group, P = 0.004,
Fig. 5D). The results of microPET imaging coincide with those ob-
served in biodistribution studies.
injection of 18F-FBHGal and the data fit well to a two-compartment
pharmacokinetic model. Following intravenous administration of
18F-FBHGal in normal mice, the time–radioactivity curve of blood
exhibited a rapid distribution phase with a mean distribution
half-life (T1/2a) of 0.09 min followed by a relatively slower elimina-
tion phase with a half-life (T1/2b) of 1.48 min (Fig. 5C). The distribu-
tion and elimination half-lives of 18F-FBHGal in hepatic fibrosis
mice were much longer than those in normal mice (2.1- and
55.8-fold, respectively). The AUC of 18F-FBHGal in blood of hepatic
fibrosis mice were 1434.66 min %ID/mL, which is 23.9-fold higher
than that in normal mice (60.15 min %ID/mL). The faster blood
clearance (Cl) in normal mice may explain the lower AUC. The
The time–radioactivity curves of 18F-FBHGal in normal and he-
patic fibrosis mice were shown in Figure 5C. The pharmacokinetic
parameters derived from these curves were summarized in Table 2.
The radioactivity in blood declined in a bi-exponential way after

920
H.-W. Kao et al. / Bioorg. Med. Chem. 21 (2013) 912–921
Table 2
showed slower clearance (=5 %ID/g) from several major organs,
for example, spleen, stomach, intestine and bone, of normal mice.
Liver fibrosis in experimental animals can be produced by a
variety of methods involving induction by hepatotoxic agents
(e.g., dimethylnitrosamine (DMN),30,31 carbon tetrachloride40 and
ethanol41), bile duct ligation42 or using transgenic animals.43,44 Of
the methods that were currently employed, DMN-induced liver
fibrosis in mouse is simple, relatively inexpensive, reproducible,
and displays many of the features of human cirrhosis. The biodis-
tribution of 18F-FBHGal in liver, and also in many other organs, of
healthy and hepatic fibrosis mice after intravenous injection was
significantly different (Table 1). The liver uptake was higher in
healthy than in fibrosis mice, while the remaining blood activity
was higher in fibrosis mice than in healthy ones. Liver is a vital
blood reservoir of body. The liver uptake derived from the liver
ROI of microPET images may be overestimated due to the contribu-
tion of blood in it, especially in the cases of hepatic fibrosis mice.
Receptor index has been reported as a useful diagnostic parameter
in estimating the dysfunction of hepatocytes.19,45,46 In this study,
we demonstrated that receptor index derived from the microPET
images is able to distinguish the fibrosis liver from the healthy
(Fig. 5D). The time–radioactivity curves of blood derived from
microPET images (Fig. 5C) were consistent with those observed
in biodistribution studies.
Ablation of ASGPRs on the hepatocytes reported in the process
of fibrosis13,47 was also observed in this DMN-induced hepatic
fibrosis mouse model (Fig. 5A). Steirer et al.48 reported that ASGPR
contributed to the regulation of the relative concentrations of plas-
ma glycoproteins bearing terminal galactose. Due to the decrease
of ASGPR, the increased amount of serum asialoglycoproteins as
native ASGPR ligands might compete against the radiotracer with
terminal galactose. Thus, after administration of 18F-FBHGal in
hepatic fibrosis mice, the uptake of tracer by the fibrosis liver
through receptor-mediated process would be reduced and led to
longer retention in blood by either coherent decrease of ASGPR
or competition with the native ASGPR ligands. After binding to li-
gand, internalizing into the cell, and dissociating from its ligands
within the acidic environment of the sorting compartment, ASGPR
will recycle back to the cell surface within 8 min.49 The remanent
ASGPRs in hepatic fibrosis liver still function and reveal perceiv-
ably accumulation of 18F-FBHGal. Moerover, the monovalent
galactoside 18F-FBHGal, with a relatively lower binding affinity
Estimated pharmacokinetic parameters derived from microPET images of mice after
intravenous injection of 18F-FBHGal
Parameter
Unit
Normal mice
Four-week
DMN-treated mice
T1/2a
min
min
%ID/mL
0.09
1.48
170.78
1.16
0.19
82.55
87.60
0.05
1434.66
T1/2b
Cmax
Cl
AUC
l
Ci/min ꢀ (%ID/mL)
min ꢀ %ID/mL
60.15
Values were derived by data of mean (n = 4). Calculated with the WinNonlin pro-
gram (for a two compartmental model). T1/2, half-life; Cmax, maximum radioactivity;
Cl, blood clearance; AUC, area under the curve.
results revealed prolonged systemic circulation of 18F-FBHGal in
hepatic fibrosis mice compared to that in normal mice.
4. Discussion
Measurement of hepatic function in liver fibrosis patients has
been linked to several complications, such as ascites, encephalop-
athy, and portal hypertension. Noninvasive quantification of the
ASGPR density on the hepatocytes in vivo provided unambiguous
evaluation of hepatic function. 99mTc-labeled galactosyl-neoglyco-
albumin
(
99mTc-NGA) and galactosyl-human serum albumin
(
99mTc-GSA) have been used as ASGPR-binding agents in clinical
studies. However, using albumin-based detection probe may de-
bate on systemic changes that include cardiovascular, hematolog-
ical, renal, pulmonary, and immunological effects.33 In addition,
those tracers using albumin backbone usually have high stability
in the bloodstream and a long circulation time, which is unfavor-
able for a diagnostic probe. Positron emission tomography (PET)
is one of the most sensitive molecular imaging techniques, and
fluorine-18 has the most favorable physical properties in all the
positron nuclides employed in PET. This study developed a new
18F-labeled monomeric galactoside 18F-FBHGal as an ASGPR probe
for PET imaging. Biological characterization demonstrated that 18F-
FBHGal was promising for quantitative assay of ASGPR with fast
pharmacokinetics in mouse models.
ASF, containing 12.4 galactose and 1.6 galactosamine moieties in
one molecule, exhibited high binding capacity (14-fold higher than
a GalH-FITC molecule) to ASGPR.34 When co-incubated with an
equimolar concentration of ASF, the uptake of GalH-FITC in HepG2
cells observed under confocal microscopy was almost completely
blocked. The residual contrast in the blocking experiment is
probably due to nonspecific accumulation in the extracellular
space. A competitive binding assay was conducted to evaluate the
binding affinity of nonradioactive FBHGal using 131I-labeled ASF
as a reference compound. FBHGal exhibited an IC50 value of
1.53 mM for the ASGPR on HepG2 cells (data not shown), similar
to those of monovalent galactose derivatives reported in earlier lit-
erature.35–37 The results of in vitro studies demonstrated that both
GalH-FITC and 18F-FBHGal displayed specific binding to ASGP
receptor.
than multivalent ligands, revealed
a lower hepatic uptake
(24.39 3.38 %ID/g) than 18F-FNGA (79.18 7.17 %ID/g)20 at
5 min post injection in normal mice. The differences in liver up-
takes between normal and hepatic fibrosis mice, though not
remarkable in 18F-FBHGal PET imaging, still reached statistical sig-
nificance in biodistribution study (P <0.01). Except liver, radioac-
tivity accumulation in other organs such as heart, blood, lung,
pancreas, kidney and muscle, were significant higher compared
to those of normal mice mainly due to the higher remanent radio-
activity in blood. Yang et al. have demonstrated in an inhibition
study20,39 that, most ASGP receptors on the hepatocyte membrane
were occupied when excess inhibitor was preinjected, a condition
similar to the hepatic cirrhosis in which the available receptors
were largely reduced. For the patients suffering severe liver fibrosis
or cirrhosis, administration of albumin-based ASGPR binding
radiotracers, such as 18F-FNGA and 99mTc-NGA that were currently
used, would provide only vague liver images due to significant
retention of radiotracers (in %ID/g) in bloodstream as well as in li-
ver.22,50 Compared with albumin- or polymer-based ligands,22,50
18F-FBHGal can be rapidly excreted from blood and normal organs
via the urinary system in its intact form (Fig. 2) in mice and thus
the re-uptake of radiotracer by the recycled receptor would be
limited.
Both biodistribution studies and microPET imaging revealed
high and sustained accumulation of 18F-FBHGal in the liver of nor-
mal mice (24.39–20.50 %ID/g from 5 to 30 min p.i. in biodistribu-
tion studies). Noted radioactivity in kidneys and urine indicated
that 18F-FBHGal and its radioactive metabolites excreted primarily
through the urinary system (Fig. 2). In other organs, except the kid-
neys, rapid clearance of 18F-FBHGal was observed and the uptakes
were less than 1.0 %ID/g at 30 min p.i. For the albumin-based
probes, however, 18F-labeled NGA20 exhibited rather shorter liver
retention, 79.18–13.85 %ID/g from 5 to 30 min p.i., while 99mTc-la-
beled neolactosyl human serum albumin (LSA)38 and NGA39

H.-W. Kao et al. / Bioorg. Med. Chem. 21 (2013) 912–921
921
14. Sawamura, T.; Kawasato, S.; Tsuda, M.; Naitoh, Y.; Shiozaki, Y.; Sameshima, Y.
Gastroenterol. Jpn. 1985, 20, 201.
5. Conclusion
15. Vera, D. R.; Stadalnik, R. C.; Krohn, K. A. J. Nucl. Med. 1985, 26, 1157.
16. Stadalnik, R. C.; Vera, D. R.; Woodle, E. S.; Trudeau, W. L.; Porter, B. A.; Ward, R.
E.; Krohn, K. A.; O’Grady, L. F. J. Nucl. Med. 1985, 26, 1233.
17. Matsuzaki, S.; Onda, M.; Tajiri, T.; Kim, D. Y. Hepatology 1997, 25, 828.
18. Ohno, Y.; Ishida, H.; Hayashi, A.; Kamagata, S.; Hirobe, S.; Ishii, K. J. Nucl. Med.
2002, 43, 1611.
19. Sasaki, N.; Shiomi, S.; Iwata, Y.; Nishiguchi, S.; Kuroki, T.; Kawabe, J.; Ochi, H. J.
Nucl. Med. 1999, 40, 1652.
20. Yang, W.; Mou, T.; Peng, C.; Wu, Z.; Zhang, X.; Li, F.; Ma, Y. Bioorg. Med. Chem.
2009, 17, 7510.
21. Kim, E. M.; Jeong, H. J.; Park, I. K.; Cho, C. S.; Kim, C. G.; Bom, H. S. J. Nucl. Med.
2005, 46, 141.
22. Yang, W.; Mou, T.; Shao, G.; Wang, F.; Zhang, X.; Liu, B. J. Nucl. Med. 2011, 52,
978.
23. Sawamura, T.; Kawasato, S.; Shiozaki, Y.; Sameshima, Y.; Nakada, H.; Tashiro, Y.
Gastroenterology 1981, 81, 527.
In this study, an ASGPR-specific probe conjugated with FITC or
labeled with fluorine-18 was developed and characterized with
HepG2 hepatoma cells (high ASGPR expression) and a hepatic
fibrosis mouse model. The in vitro studies demonstrated that
GalH-FITC can be specifically bound and internalized into HepG2
cells through ASGP receptor-mediated endocytosis. 18F-FBHGal,
with its highly specific ASGPR-targeting ability, good in vivo stabil-
ity, and the exhibition of significantly different pharmacokinetic
properties in normal and DMN treated mice, is demonstrated a fea-
sible PET probe for monitoring hepatic fibrosis in mice, and may
provide new insights into the ASGPR-related liver diseases.
24. Woodle, E. S.; Vera, D. R.; Stadalnik, R. C.; Ward, R. E. Surgery 1987, 102, 55.
25. Baenziger, J. U.; Maynard, Y. J. Biol. Chem. 1980, 255, 4607.
26. Kim, S. H.; Goto, M.; Akaike, T. J. Biol. Chem. 2001, 276, 35312.
27. Okarvi, S. M. Eur. J. Nucl. Med. 2001, 28, 929.
28. Chuang, C. H.; Chuang, K. H.; Wang, H. E.; Roffler, S. R.; Shiea, J. T.; Tzou, S. C.;
Cheng, T. C.; Kao, C. H.; Wu, S. Y.; Tseng, W. L.; Cheng, C. M.; Hou, M. F.; Wang, J.
M.; Cheng, T. L. Clin. Cancer Res. 2012, 18, 238.
29. Schwartz, A. L.; Fridovich, S. E.; Knowles, B. B.; Lodish, H. F. J. Biol. Chem. 1981,
256, 8878.
30. Jang, J. H.; Kang, K. J.; Kim, Y. H.; Kang, Y. N.; Lee, I. S. Transplant. Proc. 2008, 40,
2700.
31. Ogata, H.; Chinen, T.; Yoshida, T.; Kinjyo, I.; Takaesu, G.; Shiraishi, H.; Iida, M.;
Kobayashi, T.; Yoshimura, A. Oncogene 2006, 25, 2520.
32. Hilgard, P.; Schreiter, T.; Stockert, R. J.; Gerken, G.; Treichel, U. Hepatology 2004,
39, 1398.
33. Pulimood, T. B.; Park, G. R. Crit. Care 2000, 4, 151.
34. Spiro, R. G. J. Biol. Chem. 1960, 235, 2860.
Acknowledgements
This study was supported by grants from the Institute of Nuclear
Energy Research of Taiwan (982001INER071 and 992001INER081).
We thank the staff of the Molecular and Genetic Imaging
Core/National Research Program for Genomic Medicine (NSC100-
2319-B010-003) who kindly provided the radioisotope and
excellent technical assistance. The technical support in part
by the Imaging Core Facility of Nanotechnology in National
Yang-Ming University and the Division of Experiment Surgery of
the Department of Surgery in Taipei Veterans General Hospital
are also gratefully acknowledged.
35. Collins, B. E.; Paulson, J. C. Curr. Opin. Chem. Biol. 2004, 8, 617.
36. Groman, E. V.; Enriquez, P. M.; Jung, C.; Josephson, L. Bioconjugate Chem. 1994,
5, 547.
Supplementary data
37. Schwartz, A. L. CRC Crit. Rev. Biochem. 1984, 16, 207.
38. Jeong, J. M.; Hong, M. K.; Lee, J.; Son, M.; So, Y.; Lee, D. S.; Chung, J. K.; Lee, M. C.
Bioconjugate Chem. 2004, 15, 850.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmc.2012.12.022.
39. Yang, W.; Mou, T.; Zhang, X.; Wang, X. Appl. Radiat. Isot. 2010, 68, 105.
40. Louis, H.; Van Laethem, J. L.; Wu, W.; Quertinmont, E.; Degraef, C.; Van den
Berg, K.; Demols, A.; Goldman, M.; Le Moine, O.; Geerts, A.; Deviere, J.
Hepatology 1998, 28, 1607.
41. Barak, A. J.; Beckenhauer, H. C.; Kerrigan, F. J. Gut 1967, 8, 454.
42. Ebrahimkhani, M. R.; Kiani, S.; Oakley, F.; Kendall, T.; Shariftabrizi, A.;
Tavangar, S. M.; Moezi, L.; Payabvash, S.; Karoon, A.; Hoseininik, H.; Mann, D.
A.; Moore, K. P.; Mani, A. R.; Dehpour, A. R. Gut 2006, 55, 1606.
43. Paradis, V.; Dargere, D.; Vidaud, M.; De Gouville, A. C.; Huet, S.; Martinez, V.;
Gauthier, J. M.; Ba, N.; Sobesky, R.; Ratziu, V.; Bedossa, P. Hepatology 1999, 30,
968.
44. Kanzler, S.; Lohse, A. W.; Keil, A.; Henninger, J.; Dienes, H. P.; Schirmacher, P.;
Rose-John, S.; Zum Buschenfelde, K. H.; Blessing, M. Am. J. Physiol. 1999, 276,
G1059.
45. Kudo, M.; Todo, A.; Ikekubo, K.; Hino, M. Am. J. Gastroenterol. 1992, 87, 865.
46. Shiomi, S.; Sasaki, N.; Tamori, A.; Habu, D.; Takeda, T.; Nishiguchi, S.; Kuroki, T.;
Kawabe, J.; Ochi, H. J. Gastroenterol. Hepatol. 1999, 14, 566.
47. Burgess, J. B.; Baenziger, J. U.; Brown, W. R. Hepatology 1992, 15, 702.
48. Steirer, L. M.; Park, E. I.; Townsend, R. R.; Baenziger, J. U. J. Biol. Chem. 2009, 284,
3777.
References and notes
1. Department of Health, E. Y., Taiwan. 2011. http://www.doh.gov.tw/ufile/doc/
99%E5%B9%B4%E7%89%88%E5%85%AC%E5%85%B1%E8%A1%9B%E7%94%9F%E5%
B9%B4%E5%A0%B1.pdf.
2. Bataller, R.; Brenner, D. A. J. Clin. Invest. 2005, 115, 209.
3. Levrero, M. Oncogene 2006, 25, 3834.
4. Benyon, R. C.; Iredale, J. P. Gut 2000, 46, 443.
5. Martinez, S. M.; Crespo, G.; Navasa, M.; Forns, X. Hepatology 2011, 53, 325.
6. Thabut, D.; Moreau, R.; Lebrec, D. Hepatology 2011, 53, 683.
7. Castera, L.; Forns, X.; Alberti, A. J. Hepatol. 2008, 48, 835.
8. Friedrich-Rust, M.; Wunder, K.; Kriener, S.; Sotoudeh, F.; Richter, S.; Bojunga, J.;
Herrmann, E.; Poynard, T.; Dietrich, C. F.; Vermehren, J.; Zeuzem, S.; Sarrazin, C.
Radiology 2009, 252, 595.
9. Ashwell, G.; Morell, A. G. Adv. Enzymol. Relat. Areas Mol. Biol. 1974, 41, 99.
10. Stokmaier, D.; Khorev, O.; Cutting, B.; Born, R.; Ricklin, D.; Ernst, T. O.; Boni, F.;
Schwingruber, K.; Gentner, M.; Wittwer, M.; Spreafico, M.; Vedani, A.; Rabbani,
S.; Schwardt, O.; Ernst, B. Bioorg. Med. Chem. 2009, 17, 7254.
11. Ashwell, G.; Harford, J. Annu. Rev. Biochem. 1982, 51, 531.
12. Stockert, R. J. Physiol. Rev. 1995, 75, 591.
49. Schwartz, A. L.; Ciechanover, A.; Merritt, S.; Turkewitz, A. J. Biol. Chem. 1986,
261, 15225.
50. Yang, W.; Mou, T.; Guo, W.; Jing, H.; Peng, C.; Zhang, X.; Ma, Y.; Liu, B. Bioorg.
Med. Chem. Lett. 2010, 20, 4840.
13. Sawamura, T.; Nakada, H.; Hazama, H.; Shiozaki, Y.; Sameshima, Y.; Tashiro, Y.
Gastroenterology 1984, 87, 1217.