Synthesis of multi-galactose-conjugated 2′-O-methyl oligoribonucleotides and their in vivo imaging with positron emission tomography

Article abstract of DOI:10.1016/j.bmc.2014.10.034

⁶⁸Ga labelled 2′-O-methyl oligoribonucleotides (anti-miR-15b) bearing one, three or seven d-galactopyranoside residues have been prepared and their distribution in healthy rats has been studied by positron emission tomography (PET). To obtain t

Full text of DOI:10.1016/j.bmc.2014.10.034

Bioorganic & Medicinal Chemistry 22 (2014) 6806–6813
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis of multi-galactose-conjugated 20-O-methyl
oligoribonucleotides and their in vivo imaging with positron
emission tomography
Jussi Mäkilä b, Satish Jadhav a, Anu Kiviniemi a, Meeri Käkelä c, Heidi Liljenbäck c,d, Päivi Poijärvi-Virta a,
Tiina Laitala-Leinonen b, Harri Lönnberg a, Anne Roivainen c,d, Pasi Virta a,
⇑
a Department of Chemistry, University of Turku, FI-20014 Turku, Finland
b Skeletal Biology Consortium, Department of Cell Biology and Anatomy, University of Turku, FI-20520 Turku, Finland
c Turku PET Centre, University of Turku and Turku University Hospital, FI-20520 Turku, Finland
d Turku Center for Disease Modelling, University of Turku, FI-20520 Turku, Finland
a r t i c l e i n f o
a b s t r a c t
68Ga labelled 20-O-methyl oligoribonucleotides (anti-miR-15b) bearing one, three or seven
D-galactopy-
Article history:
Received 2 September 2014
Revised 22 October 2014
Accepted 24 October 2014
Available online 31 October 2014
ranoside residues have been prepared and their distribution in healthy rats has been studied by positron
emission tomography (PET). To obtain the heptavalent conjugate, an appropriately protected 1,4,
7-triazacyclononane-1,4,7-triacetic acid (NOTA) precursor bearing
a
4-[4-(4,40-dimethoxytrityloxy)
butoxy]phenyl side arm was first immobilized via a base labile linker to the support and the oligonucleo-
tide was assembled on the detritylated hydroxyl function of this handle. A phosphoramidite building
block bearing two phthaloyl protected aminooxy groups and one protected hydroxyl function was intro-
duced into the 50-terminus. One acetylated galactopyranoside was coupled as a phosphoramidite to the
hydroxyl function, the phthaloyl protections were removed on-support and two trivalent galactopyran-
oside clusters were attached as aldehydes by on-support oximation. A two-step cleavage with aqueous
alkali and ammonia released the conjugate in a fully deprotected form, allowing radiolabelling with
68Ga in solution. The mono- and tri-galactose conjugates were obtained in a closely related manner. In
vivo imaging in rats with PET showed remarkable galactose-dependent liver targeting of the conjugates.
Ó 2014 Elsevier Ltd. All rights reserved.
Keywords:
Oligonucleotide-glyco-conjugates
68Ga-labelled oligonucleotides
Solid-phase synthesis
PET-imaging
Liver targeting
1. Introduction
and cancer metastasis.10–14 Affinity of monovalent carbohydrate
ligand to proteins is generally low. Biologically relevant binding
Premature elimination via renal clearance, degradation by
exonucleases, undesired accumulation in non-parenchymal cells
of liver together with poor cellular uptake are hurdles of potential
oligonucleotide-based drugs (antisense oligonucleotides and siR-
NA), which limit their systemic delivery to a target cell.1–5 These
shortcomings may at least partly be overcome by conjugation of
oligonucleotides with agents exhibiting affinity to a certain cell-
type.6–9 Targeted delivery to a desired site of action may prevent
waste accumulation, decrease passive time in the systemic circula-
tion and enhance internalization by receptor-mediated endocyto-
sis. Among the potential conjugate groups, sugars deserve a
special attention, since their interaction with proteins (lectins) is
responsible for many biologically important events such as cell–
cell communication, host–pathogen interaction, immune response
requires multiple simultaneous interactions of several identical
sugars in a correct spatial arrangement.15 To mimic these multiva-
lent interactions, numerous artificial constructs have been pre-
pared by decorating a variety of multiantennary scaffolds with
monosaccharides.16–19 Conjugation of such glycoclusters to thera-
peutic oligonucleotides may allow their targeting to lectin-like
proteins present on the surface of certain cell types, and for this
purpose, several approaches for the preparation of oligonucleo-
tides bearing multiple sugar units have been described.20–33
However, the data on the effects of carbohydrate conjugation on
oligonucleotide distribution in vivo is still meagre. The early studies
of Biessen et al.34,35 have shown that a tetraantennary lysine-based
galactose conjugate of a 20-mer 32P-labeled oligodeoxyribonucleo-
tide (ODN) is taken up in the liver of rat about four times as readily
as its unconjugated counterpart. More recent studies with mono-
valent galactosylated polyethylene glycol conjugate of 33P-ODN
have indicated that hepatocytes rather than nonparenchymal
cells were targeted.36 Similarly, N-acetylgalactosylated (GalNAc)37
⇑
Corresponding author. Tel.: +358 333 6777.
E-mail address: pamavi@utu.fi (P. Virta).
http://dx.doi.org/10.1016/j.bmc.2014.10.034
0968-0896/Ó 2014 Elsevier Ltd. All rights reserved.

J. Mäkilä et al. / Bioorg. Med. Chem. 22 (2014) 6806–6813
6807
and galactosylated38 peptide nucleic acid (PNA) has been success-
fully targeted to rat liver. GalNAc-conjugated siRNAs which target
transthyretin (TTR) mRNA have been shown to reduce TTR protein
in blood in early stage human trials.39 Very recent studies have
verified that a triantennary GalNAc enhances the silencing potency
of second-generation gapmer antisense oligonucleotides (ASOs)
6–10-fold in mouse liver.40 To learn more about the role of the
glyco-cluster effect on the whole body distribution of oligonucleo-
tides, we now report on synthesis of mono-, tri- and hepta-valent
galactose conjugates of a 20-O-methyl oligoribonucleotide (anti-
miR-15b) bearing a 30-terminal (1,4,7-triazanonane-1,4,7-triyl)tri-
acetic acid (NOTA) group and on positron emission tomography
studies of the distribution of their 68Ga complexes in healthy rats.
An oligonucleotide conjugate without backbone modification was
studied. The target miRNA (miR-15b) of the sequence is involved
in hepatocyte apoptosis.41,42
2-Cyanoethyl (methyl 2,3,4-tri-O-acetyl-
O-yl)-N,N-diisopropylphosphoramidite (3) was obtained by
a
-
D
-galactopyranoside-6-
phosphitylation of methyl 2,3,4-tri-O-acetyl-a-D-galactopyrano-
side52 with 2-cyanoethyl N,N-diisopropylphosphoramidochlori-
dite. The trivalent galactose cluster (6) was prepared according to
a previously reported protocol23 based on conjugation of 3-azido-
propyl 2,3,4,6-tetra-O-acetyl-b-D
-galactopyranoside (4)53 to a 4-
(tri-O-propargylpentaerythrityloxy)benzaldehyde (5)22 scaffold
by Cu(I) promoted click chemistry (Scheme 1). Esterified precursor
of NOTA 7 was then immobilized to LCAA-CPG, as we have previ-
ously described (Scheme 2).49 Pentane-1,5-diol was used instead
of propane-1,3-diol for the transesterification step (step i in
Scheme 2). This adjustment was noticed to improve the reproduc-
ibility of immobilization. On this support (8), a 22-mer 20-O-methyl
oligoribonucleotide (50-UGU AAA CCA UGA UGU GCU GCU A-30,
anti-miR-15b) (12) was assembled on a 1.0 lmol scale on an auto-
matic DNA/RNA-synthesizer. Non-nucleosidic phosphoramidites
1–3 were then coupled to the 50-terminus, giving supports 9–11.
Benzylthiotetrazole as an activator and coupling times of 300 s
and 600 s were used for the standard 20-O-methyl ribonucleoside
building blocks and for the non-nucleosidic phosphoramidites 1–
3, respectively. Support 9 was obtained by consecutive couplings
of 1 and 3. It should be noted that the branching unit (1) must
be capped by a phosphodiester bond in order to stabilize the
structure against base-catalyzed retro aldol condensation.50 The
phthaloyl protections of 9 and 10 were removed by a treatment
with 0.5 mol Lꢀ1 hydrazinium acetate and then subjected to oxi-
mation with galactose cluster 6 (0.17 mol Lꢀ1 in MeCN, at rt for
overnight).23 In order to release the NOTA moiety as carboxylate,
the solid supported conjugates were released using a previously
optimized two step cleavage protocol49: first hydrolysis in aq
NaOH (3 h at 55 °C) and then, after neutralization with NH4Cl,
conventional ammonolysis. The resulting conjugates 13–15 were
purified by RP HPLC and their authenticity was verified by MS
(ESI TOF) (cf. Table 1 and Fig. 5). Conjugate 16 was prepared
earlier.49 The isolated yields of conjugates 13 and 14 remained
ca. 5% due to the known problems related to solid-supported
oximation: that is, formation of formaldoximes (cf. the main
side-product: tr = 24 min with 13 in i/Fig. 5). Coupling of 3 did
not show difference to standard nucleosidic phosphoramidite
building blocks. The isolated yields (20%) of 15 and 16 were close
to those obtained for 20-mer oligonucleotides in general.
2. Results and discussion
2.1. Synthesis of the PET-labelled oligonucleotide
galactoconjugates (13–15)
Although several protocols for the preparation of oligonucleo-
tide glycoconjugates are available,20–33 incorporation of a metal-
ion-binding functionality as an additional pendant group markedly
complicates the synthesis. The most straightforward approach, viz.
post synthetic conjugation with a bifunctional chelator43 in solu-
tion, for example, suffers from insufficient difference in the chro-
matographic behavior between the radioligand-conjugate and the
large parent oligonucleotide glycoconjugate.44–48 To overcome this
problem, a solid supported NOTA precursor was recently intro-
duced.49 The support allows automated assembly of 30-radiometal-
lated oligonucleotides, leaving the 50-terminus free for further
conjugation. In the present study, this support was applied for
the preparation of 50-glycoconjugated oligonucleotides. Figure 1
shows the structures of the non-nucleosidic phosphoramidite
building blocks introduced to the 50-terminus of the support bound
oligonucleotides to allow subsequent galactose conjugation. Among
these, 2-cyanoethyl 3-(4,40-dimethoxytrityloxy)-2,2-bis{3-[(phthaliim-
idooxy)propyl]carbamoyl}propyl-N,N-diisopropylphosphoramidite
(1)50 and 2-cyanoethyl-5-(phthaliimidooxy)pent-1-yl-N,N-diiso-
propylphosphoramidite (2)51 have been synthesized previously.
O
N
O
OAc
H
AcO
AcO
O
DMTrO HN
O
O
O
O
N3
O
O
O
O
OAc
O
O
4
+
5
O
NH
O
O
N
P
N
CN
1
O
i
67%
O
O
O
N
N
P
O
OAc
N
N
N
AcO
AcO
OAc
O
2
O
O
OAc
O
N
NC
N
N
O
O
O
O
O
N
N
N
AcO
NC
H
P
O
AcO
OAc
O
N
AcO
6
O
O
O
O
AcO
AcO
AcO
OMe
AcO
OAc
3
AcO
Figure 1. Non-nucleosidic phosphoramidite building blocks used for the synthesis
of the conjugates 13–15.
Scheme 1. Synthesis of the trivalent galactose cluster. Reagents: (i) CuSO4, sodium
ascorbate, H2O, dioxane.

6808
J. Mäkilä et al. / Bioorg. Med. Chem. 22 (2014) 6806–6813
COOMe
N
N
COOMe
DMTrO
N
O
COOMe
7
8
i, ii, iii, iv
COOR
N
N
O
COOR
DMTrO
N
O
O
O
O
OMe
OAc
N
H
O
O
CN
N
O
O
O
v
O
O
O
O
NH
OAc
O
P
O
9: X =
COOR
Pr
AcO
O
N
N
O
N
O
COOR
CN
XO5'
P
O3'
O
O
O
antimiR15b
P
O
N
O
O
O
N
H
O
O
NC
O
O
9, 10
O
OH
O
OH
vi, vii
viii
10: X =
N
H
O
N
P
CN
O
O
O
O
11, 12
HO
O
O
N
N
OH
OH
N
HO
O
N
O
N
O
O
OAc
O
HO
P
CN
OH
N
N
O
11: X = AcO
O
N
N
O
OH
HO
HO
O
O
O
AcO
O
OH
MeO
OMe
H
OH
12: X = H
N
N
N
N
O
O
OH
OH
O-
P
O
HO
O
O
O
O
O
NH
OH
N
OH
COO-
N
O
N
N
O
O
O-
HO
OH
O
H
O
P
N
COO-
COO-
O
P
O
O
O5'
O3'
N
N
O
O
O
anti-miR-15b
N
O
N
HO
O
O-
N
H
N
OH
13
HO
O
O
HO
HO
OH
H
COO-
N
HO
N
O
O
P
COO-
O
O
O
-O
N
N
N
O
P
O5'
O3'
O3'
O3'
O
O
N
anti-miR-15b
N
O
O
O
O
O-
COO-
N
14
N
O
N
N
O
N
N
OH
O
HO
HO
COO-
N
O
O
O-
O5'
O
P
HO
N
COO-
O
OH
P
OH
HO
O
anti-miR-15b
N
HO
HO
O-
O
OH
COO-
O
HO
HO
15
HO
HO
MeO
HO
COO-
N
O
P
O-
N
COO-
5'
H
O
O
anti-miR-15b
N
O
COO-
16
Scheme 2. Reagents and conditions: (i) NaO(CH2)5OH, in pentane-1,5-diol, MeCN, (ii) succinic anhydride, DMAP, pyridine, (iii) LCAA-CPG, PyBOP, DMAP, DMF, (iv) Ac2O, 2,6-
lutidine, N-methylimidazol, THF, (v) standard automated oligonucleotide synthesis by the phosphoramidite strategy, (vi) hydrazine acetate, AcOH, pyridine, (vii) 0.17 mol Lꢀ1
of 6 in MeCN, overnight at rt, (viii) (1) 0.1 mol Lꢀ1 aq NaOH, 3h at 55 °C (2) concd aq NH3, overnight at 55 °C.
2.2. PET-imaging
concentrations were measured ex vivo with a gamma counter.
Conjugates 13 (seven galactoses) and 14 (three galactoses) showed
a high accumulation in liver (Fig. 2). In the ex vivo measurements,
the liver radioactivity concentration of conjugate 13 was 8-fold
higher compared to the nonglycosylated oligonucleotide 16 (P
<0.05) and that of 14 5-fold higher compared to 16 (Fig. 3). Conju-
gate 15 (one galactose) did not exhibit a significant difference in
Conjugates 13–16 were labelled as described earlier44–46,49 and
injected into the tail vein of Sprague–Dawley rats. The injection
was followed by a 60 min dynamic PET-imaging with the high res-
olution research tomography. After imaging, the tissue samples of
the euthanized animals were collected and the radioactivity

J. Mäkilä et al. / Bioorg. Med. Chem. 22 (2014) 6806–6813
6809
Table 1
MS (ESI-TOF) data of the conjugates 13–15
Entry
Oligonucleotide conjugate
Observed average molecular massa
Calculated average molecular mass
1
2
3
13
14
15
10719.5
9146.8
8097.1
10719.0
9148.6
8097.4
a
Calculated from the most intensive isotope combination at [(Mꢀ11H+K)/10]ꢀ10
.
the liver radioactivity compared to 16. It may also be supposed that
the seventh galactose unit of 13, added mainly for synthetic rea-
sons, is situated far from the six other galactose units and, hence,
does not contribute to the recognition. According to the time-activ-
ity curves of the liver, 13 achieved a steady level of radioactivity at
6 min after intravenous injection (Fig. 4), while trivalent 14 expe-
rienced a slightly slower increase of radioactivity as a function of
time after injection. The radioactivity level of the time-radioactiv-
ity curves of 15 and 16 was lower compared to 13 and 14.
Conjugate 16 showed the highest radioactivity, followed by 15,
both in the kidney ex vivo measurements (Fig. 3) and in the
time–radio-activity curves (Fig. 4). The radioactivity of 13 and 14
was much lower in kidneys. Conjugate 13, in turn, had the highest
activity in urine according to both the ex vivo and the time–radio-
activity data (Figs. 3 and 4). This appears paradoxal, because low
kidney uptake should lead to low excretion into urine. This may
be explained by metabolic cleavage of 13 that takes place in liver
and release the PET tracer into urine. Recent studies of GalNac–
ASO-conjugates have shown that only a small proportion of the
conjugate could be extracted in intact form from liver even at 1 h
timepoint.40 It was also verified that metabolism occurred once
the conjugate was internalized into the liver.
Figure 2. Maximum intensity projections of PET images. PET images are mean presentations of all time frames of 60-min acquisition. Liver (L) showed the highest
radioactivity concentration with conjugate 13 (seven galactoses), kidneys (K) showed the highest radioactivity concentration with conjugate 16 (no galactose) and the highest
radioactivity concentration in the urinary bladder (B) was observed with conjugate 13.
Figure 3. Ex vivo measured radioactivity concentration in rat liver, kidneys and urine samples.

6810
J. Mäkilä et al. / Bioorg. Med. Chem. 22 (2014) 6806–6813
Figure 4. Time radioactivity curves of liver (A), kidneys (B) and urinary bladder (C).
-13
13
[M-(1-n)H+K]n
-n
n =
13
-14
-12
RP HPLC-
purification
-15
-11
MS
-10
-16
-17
-9
-8
-7
i)
ii)
Figure 5. An example (the most complex conjugate 13) of RP HPLC chromatograms (i: crude and ii: homogenized product) and MS (ESI-TOF) spectra of the conjugates. RP
HPLC conditions: system B in General remarks.
studies, the hepta-valent galactose conjugate (13) exhibited the
highest and fastest uptake in liver, followed by the trivalent conju-
gate (14). Accordingly, the liver uptake correlated with the number
of sugar residues on the conjugate. The liver-seeking conjugate 13
had the lowest kidney uptake among the oligonucleotide conju-
gates studied, but at the same time its excretion into urine was
the highest.
Asialoglycoprotein receptor (ASGPR), abundantly expressed by
mammalian liver parenchymal cells, recognizes galactose clusters
on the surface of the glycoproteins and is important for the
3. Conclusions
68Ga-labelled oligonucleotides (anti-miR-15b) bearing multiple
galactose units on the 50-terminus were synthesized using auto-
mated phosphoramidite chemistry and on-support oximation. A
solid-supported NOTA precursor was applied for the synthesis,
which, in contrast to post-synthetic labelling methods, gave the
desired products without contamination by traces of unreacted
radioligands. Whole-body biodistribution of the conjugates
(13–16) in rats was monitored by PET. Among the compounds

J. Mäkilä et al. / Bioorg. Med. Chem. 22 (2014) 6806–6813
6811
endocytosis of these proteins.41 The marked difference between
of CuSO4ꢄ5H2O (36
l
L, 50 mmol Lꢀ1
)
and sodium ascorbate
the liver uptake of these conjugates (13 > 14 ꢁ 15 ꢂ 16) may pos-
sibly be attributed to the glycocluster effect that takes place in
binding to ASGP-R. It has been shown that ASGP-R usually clearly
prefers trivalent glycoconstructs over their di- or monovalent
counterparts.54
Kidney uptake of compound 13 was the lowest among the com-
pounds studied. The increasing number of galactose ligands clearly
reduced accumulation in kidneys. One reasonable explanation is
the competition between the liver and kidneys, that is, excretion
through hepatobiliar or renal pathways. However, compound 13
showed the highest radioactivity in urine (Fig. 4). This could be
explained by the same behavior that has recently been observed
with GalNac–ASO-conjugates:40 that is, metabolic cleavage of the
conjugate 13 in liver that release the PET-tracer into urine.
(900 l
L, 0.1 mol Lꢀ1) were added. The mixture was first heated at
50 °C for 4 h, and then stirred at room temperature overnight. To
complete the reaction, an additional heating at 50 °C for 5 h was
required. The reaction mixture was evaporated to dryness and
the residue was dissolved in EtOAc and washed twice with brine.
The organic fraction was dried with Na2SO4, evaporated to dryness
and purified by silica gel chromatography (5% MeOH in EtOAc).
Cluster 6 (0.20 g) was obtained as white foam in 67% yield. 1H
NMR (500 MHz, CDCl3): dppm 9.89 (s, 1H), 7.82 (d, 2H, J = 8.7 Hz),
7.51 (s, 3H), 6.96 (d, 2H, J = 8.7 Hz), 5.40 (br d, 3H, J = 3.2 Hz),
5.21 (dd, 3H, J = 10.4 and 8.0 Hz), 5.03 (dd, 3H, J = 10.5 and
3.4 Hz), 4.58 (s, 6H), 4.48 (d, 3H, J = 8.0 Hz), 4.44 (m, 3H), 4.35
(m, 3H), 4.19–4.09 (m, 6H), 4.03 (s, 2H), 3.94–3.87 (m, 6H), 3.61
(s, 6H), 3.49 (m, 3H), 2.21–2.08 (m, 6H), 2.16 (s, 9H), 2.09 (s, 9H),
2.04 (s, 9H), 1.99 (s, 9H); 13C NMR (125 MHz, CDCl3): dppm 190.7,
170.3, 170.2, 170.1, 169.6, 164.1, 145.0, 131.9, 130.0, 122.7,
114.8, 101.2, 70.8, 70.7, 68.8, 68.7, 67.0, 66.0, 64.9, 61.2, 46.7,
45.1, 30.3, 20.9, 20.7, 20.7, 20.6; HRMS (ESI): [M+Na]+ C72H97N9
NaO35 requires 1670.5985, found 1670.5993.
4. Experimental procedures
4.1. General remarks
The NMR spectra were recorded at 500 MHz. The chemical
shifts are given in ppm from internal TMS. The mass spectra were
recorded using a MS (ESI-TOF) spectrometer. RP HPLC analysis and
4.4. NOTA-CPG-support (8)
Dimethyl [7-(1-{4-[4-(4,40-dimethoxytrityloxy)butoxy]phenyl}-
2-methoxy-2-oxoethyl)-1,4,7-triazacyclononane-1,4-diyl]diacetate
purification of the oligonucleotides were performed using
Thermo ODS Hypersil C18 (150 ꢃ 4.6 mm, 5 m) analytical column
and a Phenomenex Oligo-RP C18 (250 ꢃ 10 mm, 5 m) semi-
a
(7, 30 mg, 37 l
mol) was synthesized as previously described49 and
l
dissolved in a mixture of 0.1 mol Lꢀ1 NaO(CH2)5OH in 1,5-pentane-
diol (0.2 mL) and MeCN (1.8 mL) (step i in Scheme 2). The mixture
was stirred for 3 h at ambient temperature, neutralized by addition
of pyridinium hydrochloride (60 mg), diluted with DCM and
washed with saturated NaHCO3 and brine. The organic layer was
dried with Na2SO4, filtered and evaporated to dryness. The residue,
l
preparative column with a gradient elution either (A) from 0% to
35% MeCN in aqueous 0.1 mol Lꢀ1 Et3NH+AcOꢀ or (B) from 0% to
35% MeCN in aqueous 50 mmol Lꢀ1 NH+4AcOꢀ (0–35 min). The flow
rate was 1.0 mL minꢀ1 (analytical) or 3.0 mL minꢀ1 (semi-prepara-
tive) and the detection wavelength 260 nm. 68Ga was obtained in
the form of [68Ga]Cl3 from a 68Ge/68Ga generator (Eckert & Ziegler,
Valencia, California, USA).
succinic anhydride (4.0 mg, 39 lmol) and a catalytic amount of
4-(N,N0-dimethylamino)pyridine (DMAP) were dissolved in dry
pyridine (0.5 mL), the mixture was stirred overnight at ambient
temperature and evaporated to dryness (step ii in Scheme 2). The
residue was dissolved in dry DMF (1.5 mL), suspended with long
chain alkylamine controlled pore glass (LCAA-CPG)-support
4.2. 2-Cyanoethyl (methyl 2,3,4-tri-O-acetyl-a-D-galacto-
pyranoside-6-O-yl)-N,N-diisopropyl phosphoramidite (3)
Triethylamine (1.32 mL, 9.5 mmol) and 2-cyanoethyl N,N-dii-
(350 mg), PyBOP (38 mg, 73
lmol), DMAP (4.0 mg, 32 lmol) and
sopropylphosphoramidochloridite (500
to a mixture of methyl 2,3,4-tri-O-acetyl-
l
L, 2.3 mmol) were added
-galactopyranoside52
DIEA (26 L) were added, and the suspension was shaken overnight
l
a
-D
at ambient temperature (step iii in Scheme 2). The crude NOTA-
loaded support 8 was filtered, washed with DMF, DCM and MeOH,
and dried under vacuum. The unreacted hydroxyl groups and
amino groups on the parent support were finally capped by acety-
lation: The support was suspended in a mixture of Ac2O, 2.6-luti-
dine and N-methylimidazol in THF (5:5:8:82, v/v/v/v, for
2 ꢃ 15 min at 25 °C), washed with DMF, DCM and MeOH and dried
under vacuum. According to DMTr-cation assay, a loading of
(0.60 g, 1.9 mmol) in dichloromethane (7 mL) under nitrogen. After
2 h, the mixture was filtered through a short dried silica gel
column (50% EtOAc in hexane, 0.1% Et3N) to yield a mixture of
RP- and SP-diastereomers of 3 (0.69 g, 70%) as colorless oil. 1H
NMR (500 MHz, CD3CN a mixture of RP and SP diastereomers): d
5.47–5.43 (m, 1H), 5.27 (dd, 1H, J = 10.8 and 3.3 Hz), 5.08 (dd,
1H, J = 10.5 and 3.5 Hz), 4.95 (d, 0.5H, J = 3.5 Hz), 4.95 (d, 0.5H,
J = 3.5 Hz), 4.20–4.15 (m, 1H), 3.88–3.70 (m, 2.5H), 3.67 (dd, 1H,
J = 7.5 and 7.0 Hz), 3.65–3.56 (m, 2.5H), 3.40 (s, 1.5H), 3.40
(s, 1.5H), 2.69–2.64 (m, 2H), 2.13 (s, 1.5H), 2.12 (s, 1.5H), 2.04 (s,
3H), 1.95 (s, 3H), 1.21–1.15 (m, 12H); 13C NMR (125 MHz, CD3CN):
d 170.2, 170.13, 170.13, 170.08, 169.87, 169.85, 68.2, 68.1, 67.9,
67.9, 67.8 (d, J = 7.1 Hz), 67.7 (d, J = 7.1 Hz), 67.54, 67.50, 61.2
(d, J = 17.0 Hz), 61.1 (d, J = 16.9 Hz), 58.5 (d, J = 18.6 Hz), 58.4 (d,
J = 18.6 Hz), 54.85, 54.83, 42.9 (d, J = 5.6 Hz), 42.8 (d, J = 5.5 Hz),
23.95, 23.92, 23.90, 23.87, 20.1, 20.0, 19.94, 19.89; 31P NMR
(200 MHz, CD3CN): d 148.64 and 148.56; HRMS (ESI): [M+H]+
C22H38N2O10P requires 521,2264, found 521.2281.
27 l
mol gꢀ1 was obtained. An automatic test synthesis with a short
oligonucleotide (T6), followed by cleavage and RP HPLC and MS
(ESI-TOF) analysis of the product, verified the quality of support 8.
4.5. Synthesis of oligonucleotide conjugates 13–16
The 22-mer 20-O-methyl oligoribonucleotide (anti-miR-15b)
was assembled on an automatic DNA/RNA-synthesizer on four par-
allel batches (each in 1.0 l
mol scale) of support 8.49 Standard RNA
coupling cycle (300 s coupling time using benzylthiotetrazole as an
activator) was used. For the synthesis of the galactose conjugates
(13–15), phthalimidooxy- (1 and 2)50,51 and galactose-derived (3)
phosphoramidites were introduced to the 50-terminus of the sup-
ported oligonucleotides (9–11 in Scheme 2). A prolonged 600 s
coupling time and slightly elevated phosphoramidite concentra-
tion (0.13 mol Lꢀ1 solution of 1, 2 or 3 in MeCN used to load the
reagent vessel) gave quantitative couplings for these non-nucleos-
idic building blocks. The support obtained by coupling branching
4.3. Trivalent galactose cluster (6)
3-Azidopropyl 2,3,4,6-tetra-O-acetyl-b-D
-galactopyranoside (4)53
(64 mg, 0.18 mmol) and 4-[3-(prop-2-yn-1-yloxy)-2,2-bis(prop-2-
yn-1-yloxymethyl)propoxy]benzaldehyde (5)23 (370 mg, 0.86 mmol)
were dissolved in 1,4-dioxane (1.3 mL) and the aqueous solutions

6812
J. Mäkilä et al. / Bioorg. Med. Chem. 22 (2014) 6806–6813
unit 1 was subjected to an additional detritylation step50 and sub-
sequent coupling of 3. After the chain assembly, supports 9 and 10
were removed from the synthesizer, treated in reaction columns
with a mixture of 0.5 mol Lꢀ1 hydrazinium acetate in AcOH–pyri-
dine (1:4, v/v, 15 min at rt), washed with pyridine and MeCN and
dried under vacuum. Triantennary galactose cluster 6 was then
conjugated to the exposed aminooxy groups by on-support oxima-
tion.50,51 The supports were transferred to microcentrifuge tubes
and suspended in a 0.17 mol Lꢀ1 solution of 6 in dry MeCN.23 The
suspensions were shaken overnight at ambient temperature, fil-
tered, washed with MeCN, and dried under vacuum. Finally, the
oligonucleotide conjugates were released from the support and
Finland). The radioactivity uptake was reported as SUV. All animal
experiments were approved by the national Animal Experiment
Board in Finland and carried out in compliance with the European
Union laws relating to the conduct of animal experimentation.
Acknowledgments
The financial support from the Academy of Finland (251539,
252097, 256214 and 258814) and from the Erasmus Mundus
Action 2, Strand 1 (EMA2) for the award of EXPERTS fellowships
are gratefully acknowledged.
deprotected by previously optimized two-step cleavage protocol49
:
(step 1/viii in Scheme 2) The supports in microcentrifuge tubes
were treated with 0.1 mol Lꢀ1 aq NaOH (1.0 mL) for 3 h at 55 °C.
The suspensions were neutralized by addition of 1.0 mol Lꢀ1 aque-
ous ammonium chloride (1.1 mL), filtered and the filtrates were
evaporated to dryness (step 2/viii in Scheme 2). The residues were
dissolved in concentrated aqueous ammonia (overnight at 55 °C)
and evaporated to dryness. The crude products 13–16 were
purified by RP HPLC and their authenticity was verified by MS
(ESI-TOF) (See Table 1 and Fig. 5, characterization data for 16, see
Ref. 49).
Supplementary data
Supplementary data (additional material for the preparation of
3, NMR spectra (1H, 13C, 31P, COSY and HSQC) of 3 and 6 and table
considering ex vivo-biodistribution of intravenously injected 68Ga
labeled compounds 13–16 in healthy rats) associated with this
article can be found, in the online version, at http://dx.doi.org/
10.1016/j.bmc.2014.10.034.
References and notes
4.6. 68Ga labelling of conjugates 13–16.
1. Burnett, J. S.; Rossi, J. J. Chem. Biol. 2012, 19, 60.
2. Kole, R.; Krainer, A. R.; Altman, S. Nat. Rev. Drug Discovery 2012, 11, 125.
3. Juliano, R. L.; Ming, X.; Nakagawa, O. Bioconjugate Chem. 2012, 23, 147.
4. Kurreck, J. Angew. Chem., Int. Ed. 2009, 48, 1378.
5. Castanotto, D.; Rossi, J. J. Nature 2009, 457, 426.
6. Juliano, R. L.; Ming, X.; Carver, K.; Laing, B. Nucleic Acid Ther. 2014, 24, 101.
7. Winkler, J. Ther. Delivery 2013, 4, 791.
8. Ming, X. Exp. Opin. Drug Delivery 2011, 8, 435.
9. Marlin, F.; Simon, P.; Saison-Behmoares, T.; Giovannangeli, C. ChemBioChem
2010, 11, 1493.
10. Ernst, B.; Magnani, J. L. Nat. Rev. Drug Discovery 2009, 8, 661.
11. Seeberger, P. H.; Werz, D. B. Nature 2007, 446, 1046.
12. Collins, B. E.; Paulson, J. C. Curr. Opin. Chem. Biol. 2004, 8, 617.
13. Bertozzi, C. R.; Kiessling, L. L. Science 2001, 291, 2357.
14. Lis, H.; Sharon, N. Chem. Rev. 1998, 98, 637.
15. Dam, T. K.; Brewer, C. F. Adv. Carbohydr. Chem. Biochem. 2010, 63, 139.
16. Blanco, J. L. J.; Mellet, C. O.; Fernández, J. M. G. Chem. Soc. Rev. 2013, 42, 4518.
17. Wittmann, V.; Pieters, R. J. Chem. Soc. Rev. 2013, 42, 4492.
18. Chabre, Y. M.; Roy, R. Chem. Soc. Rev. 2013, 42, 4657.
19. Lee, R. T.; Wang, M.-H.; Lin, W.-J.; Lee, Y. C. Bioorg. Med. Chem. 2011, 19, 2494.
20. Aviñó, A.; Ocampo, S. M.; Lucas, R.; Reina, J. J.; Morales, J. C.; Perales, J. C.; Eritja,
R. Mol. Divers. 2011, 15, 751.
68Ga was obtained in the form of [68Ga]Cl3 from a 68Ge/68Ga
generator (instrument details in general remarks) by elution with
0.1 mol Lꢀ1 HCl. Sodium acetate was then added to give
a
0.4 mol Lꢀ1 solution with regard to sodium acetate. The pH was
adjusted to 3.5 with HCl, an oligonucleotide conjugate (13–16,
5–12 nmol, as a 1 mM solution) was added and the reaction
mixture was incubated at 95 °C for 10–15 min. The radiochemical
purity of [68Ga]-chelated 13–16 was determined by a RP radio-
HPLC (see instrument details in general remarks) with a gradient
elutions either from 0.1 mol Lꢀ1 Et3NH+AcOꢀ in H2O to 0.1 mol Lꢀ1
Et3NH+AcOꢀ in 50% aq MeCN or from 50 mmol Lꢀ1 NH4+AcOꢀ in H2O
to 50 mmol Lꢀ1 NH+4AcOꢀ in 50% aq MeCN. After each run, the
column was washed with 50 mmol Lꢀ1 aqueous phosphoric acid.
4.7. Dynamic PET-imaging of the biodistribution
21. Kiviniemi, A.; Virta, P.; Lönnberg, H. Bioconjugate Chem. 2011, 22, 1249.
22. Ugarte-Uribe, B.; Pérez-Rentero, S.; Lukas, R.; Aviñó, A.; Reina, J. J.; Alkorta, I.;
Eritja, R.; Morales, J. C. Bioconjugate Chem. 2010, 21, 1280.
23. Karskela, M.; Helkearo, M.; Virta, P.; Lönnberg, H. Bioconjugate Chem. 2010, 21,
748.
24. Kiviniemi, A.; Virta, P.; Lönnberg, H. Bioconjugate Chem. 2010, 21, 1890.
25. Pourceau, G.; Meyer, A.; Chevolot, Y.; Souteyrand, E.; Vasseur, J.-J.; Morvan, F.
Bioconjugate Chem. 2010, 21, 1520.
26. Jayaprakash, K. N.; Peng, C. G.; Butler, D.; Varghese, J. P.; Maier, M. A.; Rajeev, K.
G.; Manoharan, M. Org. Lett. 2010, 12, 5410.
27. Pourceau, G.; Meyer, A.; Vasseur, J.-J.; Morvan, F. J. Org. Chem. 2009, 74, 1218.
28. Singh, Y.; Murat, P.; Defrancq, E. Chem. Soc. Rev. 2010, 39, 2054.
29. Lönnberg, H. Bioconjugate Chem. 2009, 20, 1065.
Fourteen male and two female Sprague–Dawley rats were used
(n = 2–6 per tracer; weight 358 70 g) in the study. Labelled
tracers (17.3 5.8 MBq per animal, specific radioactivity 5.08
0.35 MBq/nmol) were injected into the tail vein. The injection
was followed by a 60 min dynamic PET-imaging with the High Res-
olution Research Tomography (Siemens Medical Solutions, Knox-
ville, TN, USA). Two rats were imaged simultaneously. For PET
imaging, rats were anesthetized with isoflurane and placed on a
thermal pad. After the scan the collected imaging data was recon-
structed with the ordered-subsets expectation maximization 3D
algorithm (OSEM3D). Quantitative analysis was performed by
defining volumes of interest (VOI) on the liver, kidneys and urinary
bladder areas (Carimas 2.8; Turku PET Centre, Turku, Finland;
http://www.turkupetcentre.fi/carimas). The average radioactivity
concentration in the VOI (kBq/mL) was used for further analyses.
The uptake was reported as a standardized uptake value (SUV),
which was calculated as a ratio of radioactivity of the VOI and
injected radioactivity (kBq) divided by body weight (g). Time–
radioactivity curves, representing radioactivity concentration in
the organ of interest versus time after tracer injection, were
extracted from dynamic images accordingly. After imaging, the
animals were euthanized and tissue samples were collected,
weighed and radioactivity concentrations were measured ex vivo
with the 1480 Wizard 300 gamma counter (Perkin Elmer, Turku,
30. Karskela, M.; Virta, P.; Malinen, M.; Urtti, A.; Lönnberg, H. Bioconjugate Chem.
2008, 19, 2549.
31. Kiviniemi, A.; Virta, P.; Lönnberg, H. Bioconjugate Chem. 2008, 19, 1726.
32. Morvan, F.; Meyer, A.; Jochum, A.; Sabin, C.; Chevolot, Y.; Imberty, A.; Praly, J.-
P.; Vasseur, J.-J.; Souteyrland, E.; Vidal, S. Bioconjugate Chem. 2007, 18, 1637.
33. Duff, R. J.; Deamond, S. F.; Roby, C.; Zhou, Y.; Ts’o, P. O. P. Methods Enzymol.
1999, 313, 297.
34. Biessen, E. A. L.; Vietsch, H.; Rump, E. T.; Fluiter, K.; Bijsterbosch, M. K.; van
Berkel, T. J. C. Methods Enzymol. 2000, 314, 324.
35. Biessen, E. A. L.; Vietsch, H.; Rump, E. T.; Fluiter, K.; Kuiper, J.; Bijsterbosch, M.
K.; van Berkel, T. J. C. Biochem. J. 1999, 340, 783.
36. Zhu, L.; Ye, Z.; Chemg, K.; Miller, D. D.; Mahato, R. I. Bioconjugate Chem. 2008,
19, 290.
37. Hamzavi, R.; Dolle, F.; Tavitian, B.; Dahl, O.; Nielsen, P. E. Bioconjugate Chem.
2003, 14, 941.
38. Biessen, E. A. L.; Sliedregt-Bol, K.; Hoen, P. A. Chr. T.; Prince, P.; Van der Bilt, E.;
Valentijn, A. R. P. M.; Meeuwenoord, N. J.; Princen, H.; Bijsterbosch, M. K.; Van
der Marel, G. A.; Van Boom, J. H.; Van Berkel, T. J. C. Bioconjugate Chem. 2002,
13, 295.

J. Mäkilä et al. / Bioorg. Med. Chem. 22 (2014) 6806–6813
6813
39. Rajeev, K. G.; Zimmermann, T.; Manoharan, M.; Maier, M.; Kuchimanchi, S.;
Charisse, K. Patent WO 2013075035 A1, 2013.
40. Prakash, T. P.; Graham, M. J.; Yu, J.; Carty, R.; Low, A.; Chappell, A.; Schmidt, K.;
Zhao, C.; Aghajan, M.; Murray, H. F.; Riney, S.; Booten, S. L.; Murray, S. F.; Gaus,
H.; Crosby, J.; Lima, W. F.; Guo, S.; Monia, B. P.; Swayze, E. E.; Seth, P. P. Nucleic
Acids Res. 2014, 42, 8796.
46. Schlesinger, J.; Koezle, I.; Bergmann, R.; Tamburini, S.; Bolzati, C.; Tisato, F.;
Noll, B.; Klussmann, S.; Vonhoff, S.; Wuest, F.; Pietzsch, H.-J.; Steinbach, J.
Bioconjugate Chem. 2008, 19, 928.
47. Li, Y.; Schaffer, P.; Perrin, D. M. Bioorg. Med. Chem. Lett. 2013, 23, 6313.
48. Rockey, W. M.; Huang, L.; Kloepping, K. C.; Baumhover, N. J.; Giangrande, P. H.;
Schultz, M. K. Bioorg. Med. Chem. 2011, 19, 4080.
41. An, F.; Gong, B.; Wang, H.; Yu, D.; Zhao, G.; Lin, L.; Tang, W.; Yu, H.; Bao, S.; Xie,
Q. Apoptosis 2012, 17, 702.
42. Guo, C. J.; Pan, Q.; Li, D. G.; Sun, H.; Liu, B. W. J. Hepatol. 2009, 50, 766.
43. Liu, W.; Hao, G.; Long, M. A.; Anthony, T.; Hsieh, J.-T.; Sun, X. Angew. Chem., Int.
Ed. 2009, 48, 7346.
44. Lendvai, G.; Velikyan, I.; Bergström, M.; Estrada, S.; Laryea, D.; Välilä, M.;
Salomäki, S.; Långström, B.; Roivainen, A. Eur. J. Pharm. Sci. 2005, 26, 26.
45. Roivainen, A.; Tolvanen, T.; Salomäki, S.; Lendvai, G.; Velikyan, I.; Numminen,
P.; Välilä, M.; Sipilä, H.; Bergström, M.; Härkönen, P.; Lönnberg, H.; Långström,
B. J. Nucl. Med. 2004, 45, 347.
49. Kiviniemi, A.; Mäkelä, J.; Mäkilä, J.; Saanijoki, T.; Liljenbäck, H.; Poijärvi-Virta,
P.; Lönnberg, H.; Laitala-Leinonen, T.; Roivainen, A.; Virta, P. Bioconjugate Chem.
2012, 23, 1981.
50. Katajisto, J.; Virta, P.; Lönnberg, H. Bioconjugate Chem. 2004, 15, 890.
51. Salo, H.; Virta, P.; Hakala, H.; Prakash, T. P.; Kawasaki, A. M.; Manoharan, M.;
Lönnberg, H. Bioconjugate Chem. 1999, 10, 815.
52. Brask, J.; Jensen, K. Bioorg. Med. Chem. Lett. 2001, 11, 697.
53. Joosten, J. A. F.; Loimaranta, V.; Appeldoorn, C. C. M.; Haataja, S.; El Maate, F. A.;
Liskamp, R. M.; Finne, J.; Pieters, R. J. J. Med. Chem. 2004, 47, 6499.
54. Lee, Y. C. FASEB J. 1992, 6, 3193.