Synthesis of chemically functionalized superparamagnetic nanoparticles as delivery vectors for chemotherapeutic drugs

Article abstract of DOI:10.1016/j.bmc.2007.12.051

Superparamagnetic iron oxide nanoparticles (SPIONs) are in clinical use for disease detection by MRI. A major advancement would be to link therapeutic drugs to SPIONs in order to achieve targeted drug delivery combined with detection. In the present work,

Full text of DOI:10.1016/j.bmc.2007.12.051

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 16 (2008) 2921–2931
Synthesis of chemically functionalized superparamagnetic
nanoparticles as delivery vectors for chemotherapeutic drugs
Stephen Hanessian,a,* Justyna A. Grzyb,a,ꢀ
Feride Cengellib,ꢀ and Lucienne Juillerat-Jeanneretb,*
aUniversite´ de Montre´al, Department of Chemistry, PO Box 6128, Station, Centre-ville, Montreal, Que., Canada H3C 3J7
bCentre Hospitalier Universitaire Vaudois (CHUV) and University of Lausanne (UNIL), Bugnon 25, CH-1011 Lausanne, Switzerland
Received 9 November 2007; revised 20 December 2007; accepted 26 December 2007
Available online 31 December 2007
Abstract—Superparamagnetic iron oxide nanoparticles (SPIONs) are in clinical use for disease detection by MRI. A major advance-
ment would be to link therapeutic drugs to SPIONs in order to achieve targeted drug delivery combined with detection. In the
present work, we studied the possibility of developing a versatile synthesis protocol to hierarchically construct drug-functional-
ized-SPIONs as potential anti-cancer agents. Our model biocompatible SPIONs consisted of an iron oxide core (9–10 nm diameter)
coated with polyvinylalcohols (PVA/aminoPVA), which can be internalized by cancer cells, depending on the positive charges at
their surface. To develop drug-functionalized-aminoPVA-SPIONs as vectors for drug delivery, we first designed and synthesized
bifunctional linkers of varied length and chemical composition to which the anti-cancer drugs 5-fluorouridine or doxorubicin were
attached as biologically labile esters or peptides, respectively. These functionalized linkers were in turn coupled to aminoPVA by
amide linkages before preparing the drug-functionalized-SPIONs that were characterized and evaluated as anti-cancer agents using
human melanoma cells in culture. The 5-fluorouridine-SPIONs with an optimized ester linker were taken up by cells and proved to
be efficient anti-tumor agents. While the doxorubicin-SPIONs linked with a Gly-Phe-Leu-Gly tetrapeptide were cleaved by lyso-
somal enzymes, they exhibited poor uptake by human melanoma cells in culture.
ꢀ 2008 Elsevier Ltd. All rights reserved.
1. Introduction
magnetic iron oxide nanoparticles (SPIONs) have been
studied as potential targeted drug carriers with some
promise.8–16 A less studied area is to covalently conju-
gate17–21 an anti-tumor agent to SPIONs so that they
can be internalized by the cancer cells, and the drug sub-
sequently released through an enzyme mediated mecha-
nism to exert its expected therapeutic effects with
minimal toxicity. With this ultimate objective in mind,
we first set out to covalently attach two well known
anti-tumor agents, 5-fluorouridine (Fur) and doxorubi-
cin (DOX), to a mixed polymer of polyvinylalcohol
(PVA) and poly(vinylalcohol/vinylamine) (aminoPVA),
then to construct the drug-SPION conjugate by adding
the ferrofluid consisting of iron oxide nanoparticles. The
aim of the present study was to develop chemical meth-
ods of ligating the drugs to the aminoPVA polymer
through appropriate bifunctional linkers capable of
releasing the active entity intracellularly. In previous
studies, we have prepared and characterized various
SPIONs coated with polyvinylalcohols, which were
found to be internalized by non-phagocytic human tu-
mor cells.22–24 We report herein that a hierarchical
assembly of drug-SPION conjugates is chemically feasi-
Drug entrapping polymeric nanoparticles have been well
studied, and some are under clinical evaluation for the
treatment of cancer.1–3 The nanoparticles can act as
drug vectors in matrix systems where drugs are either
incorporated into a hydrophobic or hydrophilic core
surrounded by a polymeric membrane, or displayed at
the nanoparticle surface. The release of these drugs from
the nanoparticles may either be a passive phenome-
non,3,4 which depends on the properties of the polymers
and the cellular environment (ionic strength, pH, redox
state), or an active phenomenon,5–7 where in isolated in-
stances drugs have been covalently linked to the poly-
meric nanocarriers via enzyme-specific releasing routes.
Anti-cancer drugs non-covalently bound to superpara-
Keywords: SPIONs; Ferrofluid; Drug delivery; Anti-tumor agent.
*
Corresponding authors. Tel.: +1 514 343 6738; fax: +1 514 343 5728
(S.H.), tel.: +41 21 314 7173; fax: +41 21 314 7115 (L.J.); e-mail
addresses: stephen.hanessian@umontreal.ca; lucienne.juillerat@
chuv.ch
ꢀ
These authors contributed equally to this work.
0968-0896/$ - see front matter ꢀ 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.12.051

2922
S. Hanessian et al. / Bioorg. Med. Chem. 16 (2008) 2921–2931
ble, and that internalization by human melanoma cancer
cells affects their survival.
formation with the aminoPVA was realized using
EDC and HOBt in aqueous DMF. Purification by dial-
ysis and lyophilization gave a white amorphous solid 6
where 43% of the amines on aminoPVA were conju-
gated with 5-fluorouridine as determined by IR analysis.
The level of drug incorporation was determined by mea-
suring the ratio of the area under the azide peak at
2100 cmꢀ1 to that of a reference peak in the aminoPVA
at 1100 cmꢀ1 in the IR spectrum. Treatment of 5 in
aqueous buffer with pig liver esterase at pH 7.2 resulted
in the release of 5-fluorouridine within 24 h, while 6 re-
leased about 50% of the 5-fluorouridine under the same
conditions. The drug-linker conjugate was stable in the
pH range 5–7 for several days. Three additional linkers
were also prepared and coupled to aminoPVA to give 5-
fluorouridine-linked aminoPVAs 7, 8, and 9 containing
27%, 22%, and 19% of 5-fluorouridine, respectively
(Scheme 2 and Supplementary Information).
2. Chemistry
2.1. Synthesis of drug-functionalized-aminoPVAs
In order to evaluate the optimal ester linkage we chose
5-fluorouridine 50-esters as a prototypical anti-tumor
agent, which would be released intracellularly by the ac-
tion of esterases. We varied the length of the dicarbox-
ylic acid tethers, each containing an azide group as an
IR probe to assess the extent of conjugation (loading)
with the aminoPVA (Supplemental Table S1).
The preferred linker consisted of an end-differentiated
azido dicarboxylic acid as shown in Scheme 1. The read-
ily available acetonide 125 was converted into the bis-al-
lyl ether 2, and the latter was oxidized26 and selectively
protected as the monoester 3. Esterification with 20,30-
isopropylidene-5-fluorouridine27 gave the ester 4, which
was subsequently converted to the free acid 5. Amide
Preparation of the DOX-linker-aminoPVA conjugate is
shown in Scheme 3. We used a Gly-Phe-Leu-Gly tetra-
peptide linker known to be cleaved by the lysosomal
thiol protease cathepsin B.28 Standard peptide chemistry
O
N3
O
N3
N3
O
N3
O
F
NH
f
a,b
c,d
e
O
O
O
O
N
O
O
O
O
O
O
O
O
TMSEO
OH
TMSEO
O
4
2
3
1
O
O
N
O
N
O
F
N3
O
N3
N3
O
F
F
NH
NH
NH
O
O
O
O
O
O
O
O
N
O
h
g
O
O
O
O
O
O
O
HO
O
Fe
N
O
N
O
H
H
5
6
6S
HO OH
HO OH
HO OH
Scheme 1. Reagents and conditions: (a) cat. HCl, THF/H2O, 90 ꢁC, 2 h; (b) NaH, THF, 15 min, then allyl bromide, 24 h, rt; (c) NaIO4, RuCl3ÆH2O,
CCl4/MeCN/H2O (1:1:1.5), rt, o/n; (d) 2-(trimethylsilyl)ethanol, EDC, DMAP, CH2Cl2, rt, o/n; (e) 20,30-isopropylidene-5-fluorouridine, EDC,
DMAP, CH2Cl2, rt, 2 d; (f) TFA/H2O (11:1), rt, 35 min; (g) EDC, HOBt, aminoPVA, DMF/H2O, rt, 1 d; (h) ferrofluid.
O
N
O
N
F
F
NH
NH
O
O
O
O
O
O
a
O
O
Fe
N
H
O
O
N
H
O
N3
N3
7S
Fe
7
HO OH
HO OH
O
N
O
N
F
F
NH
NH
O
O
O
H
N
O
H
N
O
O
a
O
N
H
O
N
H
O
O
N3
O
O
N3
8
8S
HO OH
F
HO OH
O
O
F
NH
NH
O
O
O
H
H
N
N
O
N
O
a
O
N
O
Fe
N
H
O
N
H
O
O
N3
O
N3
9S
9
HO OH
HO OH
Scheme 2. Assembly of 5-fluorouridine-SPIONs 7S, 8S, and 9S. Reagents: (a) ferrofluid.

S. Hanessian et al. / Bioorg. Med. Chem. 16 (2008) 2921–2931
2923
starting with the trimethylsilylethyl ester of N-Cbz-Gly
10 led to the protected peptide 13. A 2-azido-L-Glu moi-
ety29 was appended to furnish the Fm ester 14, which
was coupled from the Gly OH end to DOX giving 15
after deprotection of the ester. Coupling with amin-
oPVA using EDC/HOBT in aqueous DMF, and purifi-
cation by multiple cycles of dialysis gave the adduct 16
as a red solid after lyophilization. Quantitative IR anal-
ysis estimated that 28% of the amines on aminoPVA
were functionalized with 15.
PVA during coating of the ferrofluid (Supplemental
Table S2). A constant ratio of polymer to iron of 10
was maintained.
The drug-SPIONs were characterized for their size and
surface charge by photon correlation spectroscopy
(PCS) (Table 1) and were consistent with their expected
compositions. All drug-SPIONs and aminoPVA-SPI-
ONs had comparable size and particle size distributions
as measured by PCS. Electrophoretic mobility (charge-
to-size ratio) and zeta potential were calculated accord-
ing to the Smoluchowski approximation.
2.2. Preparation and characterization of drug-SPION
adducts
To evaluate the stability of the drug-SPIONs at physio-
logical pH, we performed gel-filtration experiments on
the 5-fluorouridine-SPIONs-1 6S and DOX-SPIONs
16S at pH 7, and compared the elution profile of iron
and PVA for 5-fluorouridine-SPIONs (Fig. 1A) and iron
and PVA elution as well as the fluorescence of DOX for
DOX-SPIONs (Fig. 1B). The 5-fluorouridine-SPIONs
6S and DOX-SPIONs 16S eluted as a single high-molec-
ular weight fraction encompassing iron, PVA, and the
fluorescent DOX, with negligible elution of low-molecu-
lar weight fluorescent molecules.
The drug-functionalized-aminoPVA-SPIONs (drug-
SPIONs) were prepared by adding the ferrofluid to mix-
tures of PVA/aminoPVA and drug-linker-aminoPVA to
obtain homogeneous amber-brown colored solutions.
We have previously shown that SPIONs coated with
PVA/aminoPVA at a ratio of 45:1 (w/w) and a poly-
mer/iron ratio of 10:1 (w/w) are efficiently taken up by
cells,23,24 and that the efficiency and biocompatibility
of the uptake were dependent on the number of positive
charges at the surface of aminoPVA-SPIONs.23 The
number of positive charges was maintained by compen-
sating for the positive charges lost through drug substi-
tution of the amino groups on the aminoPVA. This was
accomplished by increasing the ratio of aminoPVA to
In order to stabilize the various SPIONs for storage, an
excess of PVA has to be added during the preparation22
and some trailing can be observed after the high-molec-
O
O
O
H
d,f
b,c
d,e
a
O
O
H
N
CbzHN
BocHN
CbzHN
SiMe3
N
OH
O
BocHN
Ph
BocHN
O
OTMSE
N
H
OTMSE
h,i,j
O
11
10
O
12
O
O
O
O
O
H
N
H
H
N
H
N
d,g
N
FmO
N
N
OTMSE
N
OTMSE
H
H
H
N3
O
O
O
O
13
14
Ph
Ph
O
OH O
O
OH O
HO
HO
HO
HO
O
OH O OMe
Me
O
OH O OMe
Me
H
O
H
O
O
O
O
O
O
O
k
Gly-Phe-Leu-Gly
Gly-Phe-Leu-Gly
HO
N
H
N
H
N
H
N
H
N
H
N3
OH
N3
OH
16
15
O
HO
OH O
HO
O
OH O OMe
Me
H
O
l
O
O
O
Fe
Gly-Phe-Leu-Gly
N
N
N
H
H
H
N3
OH
16S
Scheme 3. Reagents and conditions: (a) 2-(trimethylsilyl)ethanol, EDC, DMAP, CH2Cl2, rt, 24 h; (b) H2, 10% Pd/C, 40 psi, rt, 16 h; (c)
BocLeuCO2HÆH2O, EDC, HOBt, i-Pr2EtN, DMF, 0 ꢁC, 1 h, 0 ꢁC ! rt, o/n; (d) 4 N HCl/dioxane, rt, 1 h; (e) BocPheCO2H, EDC, HOBt, N-
methylmorpholine, DMF, 0 ꢁC ! rt, 1 d; (f) BocGlyCO2H, EDC, HOBt, N-methylmorpholine, DMF, 0 ꢁC ! rt, 1 d; (g) 2-azido-pentanedioic acid
5-(9H-fluoren-9-ylmethyl) ester, EDC, HOBt, N-methylmorpholine, DMF, 0 ꢁC ! rt, 1 d; (h) TFA/CH2Cl2 1:1, rt, 80 min; (i) EDC, HOBt, DMF, rt,
30 min, then doxorubicinÆHCl, NEt3, rt, 1 d; (j) 1:2 N-methylpyrrolidine/DMF, rt, 1 h; (k) EDC, HOBt, NEt3, aminoPVA, DMF/H2O, rt, 1 d; (l)
ferrofluid.

2924
S. Hanessian et al. / Bioorg. Med. Chem. 16 (2008) 2921–2931
Table 1. Physicochemical characteristics of the drug-SPIONs prepared
Coated nanoparticles Nanoparticle diameter [nm] pH
Calculated free amine [mM] Mobility [10ꢀ8msꢀ1Vꢀ1m] Zeta potential [mV]
AminoPVA-SPIONs
Fur-SPIONs-1 6S
Fur-SPIONs-2 7S
Fur-SPIONs-3 8S
Fur-SPIONs-4 9S
DOX-SPIONs 16S
56.4 4.1
67.2 1.3
74.7 6.7
52.6 6.5
113.7 16.3
21.6 2.5
7.00 1.59
6.85 1.59
6.64 1.59
6.30 1.59
6.82 1.59
6.33 1.61
1.3 0.1
1.2 0.1
1.4 0.1
1.2 0.1
1.2 0.1
1.5 0.1
17.0 0.5
14.8 1.2
18.0 0.7
14.8 0.8
15.4 0.8
19.2 0.6
The size, pH, surface charge (amino groups), and the electrophoretic mobility and zeta potential are provided. Residual free amino groups on the
aminoPVA were calculated from the known number (2.5% N, w/w) of amino groups in unfunctionalized-aminoPVA and the number of drugs per
aminoPVA.
Me300 cells, as compared with the uptake of the previ-
ously described aminoPVA-SPION.23–25 In order to bet-
ter understand the physicochemical characteristics of the
linker allowing for optimal cell uptake, we compared the
calculated length and hydrophobicity of the 5-fluorouri-
dine-linker conjugates with their actual uptake by
melanoma cells (Table 2). The rate of cell uptake of 5-
fluorouridine-SPIONs was dependent on the length of
the particular linker, but not its hydrophobicity (Table
2). The highest uptake by Me300 cells was exhibited
by 6S. A time-dependent effect on DNA synthesis and
cell survival was observed (Fig. 2C). The 5-fluorouri-
dine-SPIONs 6S acted more slowly than free 5-fluoro-
uridine (Fig. 2B), but when exposure time of cells to
6S was increased, we observed an efficient inhibition
of melanoma cell survival and DNA synthesis, demon-
strating the potential of drug-ester functionalization of
SPIONs in drug delivery. It should be noted that
although active as an anti-tumor agent, 5-fluorouridine
itself is not used in the clinic due to side-effects such as
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
A
0.1
1.5
2.9
4.3
5.7
elution [ml]
2
1.8
1.6
1.4
1.2
1
3500
B
3000
2500
2000
1500
1000
500
leucopenia,
toxicity.30
thrombopenia,
and
gastrointestinal
0.8
0.6
0.4
0.2
0
3.2. Effects of DOX-SPIONs on human melanoma cells
Uptake by human Me300 melanoma cells of DOX-SPI-
ONs 16S was evaluated by the amount of cell-bound
iron and fluorescence (Fig. 3A), as compared with that
of the previously described aminoPVA-SPIONs.22–24
The uptake of DOX-SPIONs by Me300 melanoma cells
was much lower than that of aminoPVA-SPIONs, even
if the positive cell surface charges necessary for cell up-
take were maintained at the same level in the DOX-SPI-
ONs as in the aminoPVA-SPIONs (Table 1). No
decrease in either DNA synthesis or cell survival was ob-
served with DOX-SPIONs 16S, while free DOX as a
control was highly active in these cells (Fig. 3B). In or-
der to determine whether proteases from melanoma cells
have the potential to release DOX from the DOX-SPI-
ONs 16S, we extracted lysosomal enzymes from the
melanoma cells under conditions able to maintain
thiol-cathepsin activity,31 and exposed DOX-SPIONs
16S to this extract. Release of free DOX was demon-
strated by gel-filtration, showing the potential of lyso-
somal enzymes under our cell extraction conditions to
release DOX from our DOX-SPIONs 16S (Fig. 3C),
and suggesting that the DOX-SPIONs can be suitable
substrates for these lysosomal proteases. It is possible
that the amount of DOX-SPIONs 16S taken up by
Me300 melanoma cells is too low to achieve efficacy.
0
0.2
1.9
3.6
5.3
elution [ml]
Figure 1. Gel-filtration of the drug-SPIONs. Fur-SPIONs-1 6S (A) or
DOX-SPIONs 16S (B) were gel chromatographed on Sephadex G-75
in phosphate buffer at pH 7. Iron (Ç) and PVA (ꢁ) were measured in
the eluting fractions at 690 nm by Prussian Blue and iodine reactions,
respectively. SPION-bound doxorubicin (m) was determined by its
fluorescence at kex/kem 485/580 nm.
ular weight peak of the gel-filtration. However, these
trailing fractions did not display anti-cancer activity
when evaluated on cells (results not shown).
3. Biological results
3.1. Effects of 5-fluorouridine-SPIONs on human mela-
noma cells
Uptake by human Me300 melanoma cells of 5-fluorouri-
dine-SPIONs1–4 6S, 7S, 8S, and 9S was determined by
the amount of cell-bound iron (Fig. 2A). With the
exception of 7S, all drug-SPIONs were taken up by

S. Hanessian et al. / Bioorg. Med. Chem. 16 (2008) 2921–2931
2925
·109
B120
120
3.0
2.5
2.0
1.5
1.0
0.5
0.0
C
A
100
100
80
60
40
20
0
80
60
40
20
0
1 d.
1 d.
1 d.
1.5 d.
2 d.
3 d.
4 d.
1 d.
1.5 d.
·1012
2,3,4 d.
0
1
2
3
4
5
6
7
8
9
0.0
1.0
2.0
3.0
0.0
1.0
2.0
3.0
number of particles added
μ
Fur concentration [ M]
Fur concentration [ M]
μ
Figure 2. Uptake and cellular effects of the Fur-SPIONs 6S–9S in human melanoma cells. (A) Human Me300 melanoma cells were exposed for 16 h
to increasing concentrations of the Fur-SPIONs 6S (d), 7S (m), 8S (Ç) or 9S (j), or aminoPVA-SPIONs (X) for comparison, and the cell uptake of
the Fur-SPIONs was determined by cell-bound iron content. (B) Human Me300 melanoma cells were exposed to increasing concentrations of free
Fur then either cellular metabolic activity (Alamar Blue reduction, (d) dotted lines) or DNA synthesis ([3H]thymidine incorporation, (•) solid lines)
was determined. (C) Human Me300 melanoma cells were exposed for 1–4 days (1d–4d) to Fur-SPIONs-1 6S, then either cellular metabolic activity
(Alamar Blue reduction, (d) dotted lines) or DNA synthesis ([3H]thymidine incorporation, (d) solid lines) was determined.
Table 2. Relationship between the length or hydrophobicity of the linker-Fur with uptake of the various Fur-SPIONs by human Me300 melanoma
cells
˚
Linker length [A]
Drug-SPION
% Iron uptake at 16 h
41 12
logPa
AminoPVA-SPIONs
Fur-SPIONs-1 6S
Fur-SPIONs-2 7S
Fur-SPIONs-3 8S
Fur-SPIONs-4 9S
5-Fluorouridine
—
—
37
5
4
1
1
2
18.95
13.23
20.65
22.18
—
ꢀ1.165
ꢀ1.220
ꢀ1.881
ꢀ1.067
ꢀ1.676
32
24
—
Cells were exposed for 16 h to the Fur-SPIONs 6S, 7S, 8S or 9S, then the percentage of SPIONs taken up by cells was determined as the ratio (%) of
cell-bound iron to the amount of iron added to the cells. This uptake was compared to the calculated length or the hydophobicity (logP) of the
different linkers.
a logP was calculated from www.molinspiration.com/cgi-bin/properties.
·109
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
140
120
100
80
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
A
B
C
4000
3000
2000
1000
0
0
2
4
6
8
number of particles added
60
40
20
·1012
0
0
1
2
3
4
5
6
7
8
9
0.0
0.5
1.0
1.5
2.0
2.5
0.0 1.0 2.0 3.0 4.0 5.0 6.0
eluant [ml]
number of particles added
DOX concentration [ M]
Figure 3. Uptake and cellular effects of the DOX-SPIONs 16S in human melanoma cells. (A) Human Me300 melanoma cells were exposed for 36 h
to increasing concentrations of DOX-SPIONs 16S. The uptake of DOX-SPIONs 16S was determined by measuring cell-bound iron (m) and the
fluorescence of cell-bound DOX (insert). The uptake of aminoPVA-SPIONs (X) was determined for comparison. (B) Human Me300 melanoma cells
were exposed for 36 h to increasing concentrations of either free DOX or DOX-SPIONs 16S, then either cell survival (Alamar Blue reduction, AlaB)
or DNA synthesis ([3H]thymidine incorporation, 3HT) was determined as the ratio of treated to untreated cells. Free DOX (AlaB: n, dotted line;
3HT: n, solid line); DOX-SPIONs 16S (AlaB: m, dotted line; 3HT: m, solid line). (C) Either native DOX-SPIONs 16S (m) or DOX-SPIONs 16S
exposed for 24 h to proteases extracted from human Me300 melanoma cells (n) were chromatographed on Sephadex G-75 in phosphate buffer at pH
7. DOX fluorescence of either free DOX or SPIONs-bound DOX was measured in the eluting fractions at kex/kem 485/580 nm as the relative
fluorescence to the baseline.

2926
S. Hanessian et al. / Bioorg. Med. Chem. 16 (2008) 2921–2931
4. Conclusion
mixture was stirred for 24 h. After evaporation of
DMF, the residue was dissolved in H2O and the aqueous
layer was extracted with CH2Cl2 (4·). The combined or-
ganic layers were dried (MgSO4), and concentrated.
Product was obtained following column chromatogra-
phy (2.5:97.5 EtOAc/hexane) as a yellow oil (14.3 g,
In conclusion, we have developed a versatile chemical
synthesis protocol allowing for the hierarchical con-
struction of covalent drug-SPION conjugates with
mixed carrier polymers, such as PVA and aminoPVA.
We have also shown that for doxorubicin and 5-fluoro-
uridine, the cell uptake mechanisms, the accessibility of
enzymatic activities for releasing the drugs from their
SPIONs carriers, in particular the length and chemical
structure of the linker, and the intracellular routing of
these drug-SPIONs are important parameters for even-
tual drug delivery. Further studies evaluating the sub-
cellular localization of the various drug-functionalized-
SPIONs are underway. The survival of melanoma cells
was most significantly reduced with the 5-fluorouri-
dine-SPION conjugate 6S resulting in decreased mela-
noma cell viability and DNA synthesis (IC50 < 0.5 lM)
after 48 h exposure. Further studies with other anti-tu-
mor agents conjugated to SPIONs as ester linkages are
underway in our laboratories and will be published in
due time.
1
95%); H NMR (400 MHz, CDCl3) d 5.94–5.84 (2H,
m), 5.29–5.15 (4H, m), 3.97 (2H, t, J = 1.5Hz), 3.95
(2H, t, J = 1.5 Hz), 3.32 (2H, s), 3.28 (2H, s), 3.27
(2H, s), 0.98 (3H, s); 13C NMR (100 MHz, CDCl3) d
135.0, 117.0, 73.1, 72.4, 55.7, 41.2, 18.3; IR (neat/NaCl)
2857, 2103, 1273, 1093 cmꢀ1
.
5.1.3. [2-Azidomethyl-2-methyl-3-(2-trimethylsilanyl-eth-
oxycarbonylmethoxy)-propoxy]-acetic acid (3). To a
solution of 2 (1.05 g, 4.66 mmol) in CCl4/MeCN/H2O
1:1:1.5 (56 mL) were added sequentially sodium perio-
date (8.17 g, 38.2 mmol) and ruthenium trichloride
monohydrate (0.043 g, 0.20 mmol) and the thick mix-
ture was stirred vigorously at room temperature over-
night. The reaction mixture was diluted with H2O and
CH2Cl2 until the solids dissolved and the aqueous layer
was extracted with CH2Cl2 (4·), dried (MgSO4) and
concentrated. The resulting brown oil was dissolved in
saturated NaHCO3 and washed twice with EtOAc.
The aqueous layer was acidified with 6 N HCl and ex-
tracted with CH2Cl2 (4·). The combined organic layers
were dried (MgSO4) and concentrated to give the
biscarboxylic acid as a yellow-brown solid, which was
used without further purification (0.920 g, 76%); 1H
NMR (300 MHz, CDCl3) d 10.34 (2H, s), 4.14 (4H, s),
3.45 (4H, s), 3.37 (2H, s), 0.99 (3H, s); 13C NMR
(75 MHz, CDCl3) d 175.6, 74.5, 68.3, 55.5, 41.0, 17.9;
IR (neat/NaCl) 3117, 2920, 2105, 1729, 1246,
1131 cmꢀ1; LRMS (ESI, m/z, MH+) 262. The crude
biscarboxylic acid (0.500 g, 1.91 mmol) was dissolved
in CH2Cl2 (15 mL) and to it were added 2-(trimethyl-
silyl)ethanol (274 lL, 1.91 mmol), DMAP (0.117 g,
0.96 mmol), and EDC (0.367 g, 1.91 mmol). The mix-
ture was stirred at room temperature for 16 h. The crude
reaction was diluted with H2O and extracted with
CH2Cl2 (4·). The combined organic layers were dried
(MgSO4) and concentrated. Product was obtained fol-
lowing column chromatography (0:10–1:9 MeOH/
EtOAc) as a yellow oil (0.256 g, 37%); 1H NMR
(300 MHz, CDCl3) d 10.20 (1H, s), 4.24–4.19 (2H, m),
4.06 (2H, s), 4.03 (2H, s), 3.42–3.34 (6H, m), 1.01–0.96
(5H, m), 0.02 (9H, s); 13C NMR (75 MHz, CDCl3) d
174.8, 170.8, 74.6, 74.4, 69.0, 68.7, 63.4, 55.4, 40.9,
17.9, 17.4, ꢀ1.5; IR (neat/NaCl) 3204, 2955, 2103,
1751, 1132, 838 cmꢀ1; LRMS (ESI, m/z, MH+) 362.
5. Experimental
5.1. Chemistry
5.1.1. General. All commercially available reagents were
used without further purification. All reactions were per-
formed under nitrogen atmosphere. NMR (1H, 13C)
spectra were recorded on Bruker ARX-400, AV-300,
and AV-400 RG spectrometers. Low- and high-resolu-
tion mass spectra were recorded using electrospray tech-
nique. Optical rotations were recorded on a Perkin-
Elmer Model 343 polarimeter in a 1 dm cell at ambient
temperature. Analytical thin-layer chromatography was
performed on pre-coated silica gel plates. Flash column
chromatography was performed using (40–63 lm) silica
gel. IR was measured using Perkin-Elmer Spectrum One
FT-IR spectrometer. IR spectra of organic compounds
were taken as film, while of polymer mixtures they were
taken after aqueous polymer sample was mixed with
KBr and lyophilized to give a fine powder, which was
pressed into a disk.
5.1.2. 3-(3-Allyloxy-2-azidomethyl-2-methyl-propoxy)-
propene (2). To a solution of 125 (3.26 g, 17.6 mmol) in
THF (59 mL) and H2O (12 mL) were added 40 drops
of concentrated HCl and the mixture was stirred at
90 ꢁC for 2 h. The solvent was removed in vacuo and
to remove residual water, the crude product was dis-
solved in EtOAc, dried (MgSO4), and concentrated to
give the diol25,32 as a white solid (2.55 g, quant); mp
5.1.4. {2-Azidomethyl-3-[6-(5-fluoro-2,4-dioxo-3,4-dihy-
dro-2H-pyrimidin-1-yl)-2,2-dimethyl-tetrahydro-furo[3,4-
d][1,3]dioxol-4-ylmethoxycarbonylmethoxy]-2-methyl-
propoxy}-acetic acid 2-trimethylsilanyl-ethyl ester (4). To
a solution of 3 (0.275 g, 0.76 mmol) in CH2Cl2 (5 mL)
were added DMAP (0.046 g, 0.38 mmol) and 20,30-iso-
propylidene-5-fluorouridine27 (0.253 g, 0.84 mmol).
Lastly EDC (0.160 g, 0.84 mmol) was added and reac-
tion mixture was stirred at room temperature for 2 days.
The crude reaction mixture was diluted with H2O and
extracted with CH2Cl2 (4·). The combined organic lay-
1
40–42 ꢁC; H NMR (300 MHz, CDCl3) d 3.59 (4H, s),
3.44 (2H, s), 2.81 (2H, s), 0.85 (3H, s); 13C NMR
(75 MHz, CDCl3) d 68.0, 55.6, 41.0, 17.6; IR (neat/
NaCl) 3351, 2936, 2104, 1468, 1283, 1042 cmꢀ1; LRMS
(ESI, m/z, MH+) 146. To a solution of the diol (9.73 g,
67.0 mmol) in THF (335 mL) was added sodium hydride
(60% suspension in oil, 8.04 g, 201.1 mmol). After stir-
ring the mixture at room temperature for 15 min, allyl
bromide (24.33 g, 201.1 mmol) was added and reaction

S. Hanessian et al. / Bioorg. Med. Chem. 16 (2008) 2921–2931
2927
ers were dried (MgSO4) and concentrated. Product was
obtained following column chromatography (1:1 hex-
ane/EtOAc) as a pale yellow oil (0.260 g, 54%); 1H
NMR (400 MHz, CDCl3) d 9.94 (1H, d, J = 4.7 Hz),
7.42 (1H, d, J = 5.7 Hz), 5.65 (1H, d, J = 1.8 Hz), 4.97
(1H, dd, J = 1.9 Hz, J = 6.4 Hz), 4.80 (1H, dd,
J = 3.5 Hz, J = 6.3 Hz), 4.40–4.32 (3H, m), 4.21–4.17
(2H, m), 4.10–4.00 (4H, m), 3.40–3.33 (6H, m), 1.53
(3H, s), 1.32 (3H, s), 0.99–0.95 (5H, m), 0.01 (9H, s);
13C NMR (75 MHz, CDCl3) d 170.8, 170.0, 157.0 (d,
J = 26.6 Hz), 148.8, 140.6 (d, J = 238.9 Hz), 126.5 (d,
J = 33.8 Hz), 114.9, 94.6, 85.1, 84.4, 80.8, 74.5, 74.3,
68.8, 68.6, 64.0, 63.3, 55.4, 41.1, 27.2, 25.3, 17.8, 17.4,
ꢀ1.4; IR (neat/NaCl) 3215, 2955, 2104, 1712, 1465,
1252, 1132 cmꢀ1; LRMS (ESI, m/z, MH+) 646.
lowing column chromatography (8:2 hexane/EtOAc) as
a clear oil (1.53 g, 99%); H NMR (300 MHz, CDCl3)
1
d 7.38–7.32 (5H, m), 5.35 (1H, br s), 5.13 (2H, s),
4.28–4.22 (2H, m), 3.96 (2H, d, J = 5.5 Hz), 1.04–0.98
(2H, m), 0.05 (9H, s); 13C NMR (75 MHz, CDCl3) d
170.2, 156.3, 136.3, 128.6, 128.3, 128.2, 67.1, 64.0,
43.0, 17.4, ꢀ1.4; IR (neat/NaCl) 3354, 2954, 1727,
1524, 1250, 1194 cmꢀ1; LRMS (ESI, m/z, MNa+) 332.
5.1.8. (2-tert-Butoxycarbonylamino-4-methyl-pentanoyla-
mino)-acetic acid 2-trimethylsilanyl-ethyl ester (11). To a
solution of 10 (1.476 g, 4.77 mmol) in anhydrous EtOH
(60 mL) was added 10% Pd/C (0.150 g). The reaction
mixture was stirred under 40 psi of H2 at room temper-
ature for 16 h. The mixture was filtered and concen-
trated to give crude amine which was used in the next
step without further purification (0.863 g, quant). Bo-
cLeuCO2HÆH2O (2.033 g, 8.16 mmol) was dissolved in
DMF (58 mL), to it were added HOBt (1.323 g,
9.79 mmol) and i-Pr2NEt (2.13 mL, 12.23 mmol) and
the reaction mixture was cooled in an ice bath. EDC
(1.876 g, 9.79 mmol) was added and the reaction mix-
ture was stirred at 0 ꢁC for 1 h. To this were added the
prepared crude amine (0.715 g, 4.08 mmol) and i-
Pr2EtN (2.13 mL, 12.234 mmol). The reaction mixture
was stirred at 0 ꢁC for 30 min and then at room temper-
ature overnight. Solvent was removed in vacuo, the res-
idue dissolved in saturated NaHCO3 and extracted with
CH2Cl2 (4·). The combined organic layers were dried
(MgSO4) and concentrated. Product was obtained fol-
lowing column chromatography (3:7 EtOAc/hexane) as
5.1.5. {2-Azidomethyl-3-[5-(5-fluoro-2,4-dioxo-3,4-dihy-
dro-2H-pyrimidin-1-yl)-3,4-dihydroxy-tetrahydro-furan-
2-ylmethoxycarbonylmethoxy]-2-methyl-propoxy}-acetic
acid (5). The ester 4 (0.072 g, 0.11 mmol) was dissolved
in TFA/H2O 11:1 (4.3 mL) and stirred at room temper-
ature for 35 min. Solvent was removed in vacuo and
then traces of TFA were removed by co-evaporating
with MeOH. The product was obtained after column
chromatography (95:4:1 EtOAc/MeOH/AcOH) as an
1
amorphous solid (0.032 g, 56%); H NMR (300 MHz,
CD3OD) d 7.86 (1H, d, J = 6.7 Hz), 5.81 (1H, dd,
J = 1.3 Hz, J = 3.8 Hz), 4.44 (2H, d, J = 3.6 Hz), 4.21–
4.10 (5H, m), 4.05 (2H, s), 3.44 (2H, s), 3.40 (2H, s),
3.37 (2H, s), 0.98 (3H, s); 13C NMR (100 MHz,
CD3OD) d 174.0, 171.7, 159.4 (d, J = 26.2 Hz), 150.8,
141.8 (d, J = 233.4 Hz), 126.1 (d, J = 34.7 Hz), 91.7,
82.9, 75.4, 75.1, 75.0, 71.0, 69.6, 69.3, 64.6, 56.5, 42.1,
18.1; IR (neat/NaCl) 3404, 2105, 1709, 1259,
1125 cmꢀ1; LRMS (ESI, m/z, MH+) 506; HRMS (ESI)
for C18H24N5O11F Calcd (MH+): 506.1529. Found
506.1529; Calcd (MNa+): 528.1349. Found 528.1331.
1
a clear oil (0.950 g, 60%); [a]D ꢀ21.8 (c 1, CHCl3); H
NMR (300 MHz, CDCl3) d 6.76 (1H, br s), 5.00 (1H,
d, J = 8.0 Hz), 4.26–4.14 (3H, m), 4.00–3.97 (2H, m),
1.73–1.36 (11H, m), 1.30–1.10 (1H, m), 1.02–0.91 (8H,
m), 0.03 (9H, s); 13C NMR (75 MHz, CDCl3) d 173.0,
169.9, 155.8, 80.2, 64.0, 53.0, 41.5, 28.4, 24.8, 23.1,
22.0, 17.4, ꢀ1.4; IR (neat/NaCl) 3307, 2956, 1751,
1665, 1528, 1173 cmꢀ1; LRMS (ESI, m/z, MH+) 389,
(MNH4+) 406.
5.1.6. 5-Fluorouridine-linker1-aminoPVA (6). A solution
of
5
(18.0 mg, 0.0375 mmol), HOBt (4.8 mg,
0.0375 mmol), and EDC (13.7 mg, 0.0714 mmol) in
DMF (714 lL, final concentration 0.05 M) was stirred
for 60 min. Then a solution of aminoPVA in DMF/
H2O (1:1, 2 mL, 10 mg/mL, 2.5% N, 0.0375 mmol) was
added and the pH adjusted to ꢂ6.5 with 0.1 N HCl
and/or 0.1 N NaOH as needed. The mixture was stirred
at room temperature for 1 day. After 16 h the pH was
adjusted to 6.5 if it has changed. The solution was then
dialyzed (Spectra/Porꢂ 6 dialysis membrane, MWCO
3500) against 0.2 N NaCl solution for 3 days and then
against H2O for 2 days changing the solution twice dai-
ly. Product was obtained following lyophilization as a
white solid (21.4 mg).
5.1.9. [2-(2-tert-Butoxycarbonylamino-3-phenyl-propio-
nylamino)-4-methyl-pentanoylamino]-acetic acid 2-tri-
methylsilanyl-ethyl ester (12). Compound 11 (0.720 g,
1.85 mmol) was dissolved in 4 N HCl/dioxane (7 mL)
and stirred at room temperature for 1 h. Solvent was re-
moved in vacuo to give the crude amine as an HCl
salt, which was used directly in the next step (0.602 g,
quant). The residue (0.564 g, 1.74 mmol) was
dissolved in DMF (25 mL) and to it were added HOBt
(0.469 g, 3.47 mmol), N-methylmorpholine (1.14 mL,
10.41 mmol), and BocPheCO2H (0.783 g, 2.95 mmol).
The reaction mixture was cooled in an ice bath and to
it was added EDC (0.665 g, 3.47 mmol). The mixture
was allowed to warm to room temperature and was stir-
red for 1 day. Solvent was removed in vacuo, the crude
dissolved in saturated NaHCO3 and extracted with
CH2Cl2 (4·). The combined organic layers were dried
(MgSO4) and concentrated. Product was obtained fol-
lowing column chromatography (35:65 EtOAc/hexane)
as a foamy solid (0.850 g, 91%); mp 53–56 ꢁC; [a]D
ꢀ27.3 (c 1, CHCl3); 1H NMR (300 MHz, CDCl3) d
7.26–7.13 (5H, m), 6.98 (1H, br s), 6.94 (1H, br d),
5.1.7. Benzylamino-acetic acid 2-trimethylsilanyl-ethyl
ester (10).33,34 To
a solution of Cbz-GlyCO2H
(1.046 g, 5.00 mmol) in CH2Cl2 (25 mL) were added
DMAP (0.428 g, 3.50 mmol), 2-(trimethylsilyl)ethanol
(1.07 mL, 7.50 mmol), and EDC (1.917 g, 10.00 mmol).
The reaction mixture was stirred at room temperature
for 1 day, then diluted with H2O and extracted with
CH2Cl2 (4·). The combined organic layers were dried
(MgSO4) and concentrated. Product was obtained fol-

2928
S. Hanessian et al. / Bioorg. Med. Chem. 16 (2008) 2921–2931
5.39 (1H, d, J = 7.4 Hz), 4.59–4.51 (1H, m), 4.47–4.40
(1H, m), 4.22–4.16 (2H, m), 3.90–3.87 (2H, m), 3.10–
2.95 (2H, m), 1.73–1.42 (2H, m), 1.36 (9H, s), 1.30–
1.10 (1H, m), 1.00–0.95 (2H, m), 0.87 (6H, d,
J = 5.0 Hz), 0.02 (9H, s); 13C NMR (75 MHz, CDCl3)
d 172.1, 171.7, 169.7, 155.7, 136.7, 129.4, 128.6, 126.8,
80.1, 63.7, 55.8, 51.6, 41.4, 40.9, 38.1, 28.3, 24.6, 22.9,
22.0, 17.3, ꢀ1.5; IR (neat/NaCl) 3291, 2956, 1752,
1690, 1649, 1527, 1174 cmꢀ1; LRMS (ESI, m/z, MH+)
536, (MNH4+) 553.
(MgSO4) and concentrated. Product was obtained fol-
lowing column chromatography (8:2 EtOAc/hexane) as
a foamy solid (0.200 g, 65%); mp 140 ꢁC, dec.; [a]D
ꢀ18.8 (c 1, CHCl3); 1H NMR (300 MHz, CDCl3) d
8.71 (3H, br s), 8.42 (1H, br s), 7.77 (2H, d,
J = 7.4 Hz), 7.68–7.63 (2H, m), 7.42–7.30 (4H, m),
7.21–7.10 (5H, br s), 5.26 (1H, br s), 4.91 (1H, br s),
4.52–4.38 (3H, m), 4.32–3.92 (7H, m), 3.10 (1H, br s),
2.92 (1H, br s), 2.66 (2H, br s), 2.28 (2H, br s), 1.67
(2H, br s), 1.41 (1H, br s), 1.02–0.84 (8H, m), 0.02
(9H, s); 13C NMR (75 MHz, CDCl3) d 172.5, 172.4,
170.7, 170.0, 169.3, 168.2, 143.7 (2), 141.3, 136.2,
129.8, 128.5, 127.9, 127.2, 126.8, 125.1, 120.2, 66.8,
63.8, 60.6, 54.4, 51.5, 46.8, 43.0, 41.6, 40.7, 30.6, 27.2,
24.6, 23.4, 22.0, 17.4, ꢀ1.5; IR (neat/NaCl) 3278, 2956,
2103, 1739, 1634, 1543 cmꢀ1; LRMS (ESI, m/z, MH+)
826.
5.1.10. {2-[2-(2-tert-Butoxycarbonylamino-acetylamino)-
3-phenyl-propionylamino]-4-methyl-pentanoylamino}-ace-
tic acid 2-trimethylsilanyl-ethyl ester (13). Compound 12
(0.810 g, 1.51 mmol) was dissolved in 4 N HCl/dioxane
(5.5 mL) and stirred at room temperature for 1 h. Sol-
vent was removed in vacuo to give the crude amine as
an HCl salt, which was used directly in the next step
(0.714 g, quant). The residue (0.714 g, 1.33 mmol) was
dissolved in DMF (19 mL) and to it were added HOBt
(0.360 g, 2.66 mmol), N-methylmorpholine (0.88 mL,
7.10 mmol), and BocGlyCO2H (0.397 g, 2.26 mmol).
The reaction mixture was cooled in an ice bath and to
it was added EDC (0.551 g, 2.66 mmol). The reaction
mixture was allowed to warm to room temperature
and was stirred for 1 day. Solvent was removed in va-
cuo, the crude dissolved in saturated NaHCO3 and ex-
tracted with CH2Cl2 (4·). The combined organic
layers were dried (MgSO4) and concentrated. Product
was obtained following column chromatography (6:4
EtOAc/hexane) as a foamy solid (0.750 g, 95%); mp
64–66 ꢁC; [a]D ꢀ25.8 (c 1, CHCl3); 1H NMR
(300 MHz, CDCl3) d 8.05 (1H, br d), 7.88 (1H, br t),
7.61 (1H, br d), 7.20–7.07 (5H, m), 6.03 (1H, br t),
5.03–4.96 (1H, m), 4.73–4.65 (1H, m), 4.23–4.18 (2H,
m), 4.10–3.84 (4H, m), 3.04–2.86 (2H, m), 1.73–1.55
(2H, m), 1.41 (9H, s), 1.25–1.12 (1H, m), 1.02–0.96
(2H, m), 0.90–0.85 (6H, m), 0.03 (9H, s); 13C NMR
(75 MHz, CDCl3) d 172.2, 170.7, 169.8, 169.3, 156.3,
136.4, 129.6, 128.5, 126.8, 79.7, 63.6, 54.3, 51.6, 44.0,
41.9, 41.3, 39.4, 28.4, 24.7, 22.7, 22.6, 17.4, ꢀ1.5; IR
5.1.12. 4-Azido-4-(GFLG-DOX)-butyric acid (15). Com-
pound 14 (0045 g, 0.054 mmol) was dissolved in TFA/
CH2Cl2 1:1 (1.0 mL) and stirred at room temperature
for 80 min. Solvent was evaporated and then traces of
TFA were removed by co-evaporating with MeOH.
The carboxylic acid was used in the next step without
further purification (0.039 mg, quant). The carboxylic
acid (0.039 g, 0.054 mmol) was dissolved in DMF
(6.0 mL) and to it were added HOBt (0.018 g,
0.13 mmol) and EDC (0.031 g, 0.16 mmol). The reaction
mixture was stirred at room temperature for 30 min and
then doxorubicinÆHCl (0.031 g, 0.054 mmol) and NEt3
(0.03 mL, 0.22 mmol) were added and the reaction mix-
ture was stirred in the dark for 1 day. The crude reaction
was diluted with H2O and extracted with CH2Cl2 (4·).
The organic layers were washed with brine, dried
(MgSO4), and concentrated. The product was obtained
following column chromatography (5:95 MeOH/
EtOAc) as a red solid (0.045 g, 67%); [a]D +76.0 (c
0.312, CH3OH); 1H NMR (400 MHz, DMSO-d6) d
8.36–8.33 (1H, m), 8.14 (1H, t, J = 7.0 Hz), 7.94 (1H,
t, J = 6.0 Hz), 7.87–7.82 (4H, m), 7.63 (2H, d,
J = 7.4 Hz), 7.60–7.56 (1H, m), 7.50 (1H, d,
J = 8.2 Hz), 7.40 (2H, t, J = 7.4 Hz), 7.31 (2H, t,
J = 7.4 Hz), 7.22–7.19 (5H, m), 7.16–7.12 (1H, m), 5.43
(1H, s), 5.22 (1H, s), 4.90–4.86 (2H, m), 4.80 (1H, d,
J = 6.0 Hz), 4.58 (2H, d, J = 5.9 Hz), 4.56–4.50 (1H,
m), 4.37 (2H, d, J = 6.8 Hz), 4.26–4.16 (3H, m), 4.00–
3.93 (4H, m), 3.84 (1H, t, J = 6.9 Hz), 3.76–3.58 (4H,
m), 3.41–3.38 (1H, m), 3.02–2.84 (2H, m), 2.76–2.66
(1H, m), 2.41 (1H, t, J = 7.6 Hz), 2.22–2.18 (1H, m),
2.11–2.06 (1H, m), 1.92–1.81 (2H, m), 1.56–1.50 (1H,
m), 1.48–1.40 (3H, m), 1.13 (3H, d, J = 6.4 Hz), 0.84–
0.77 (6H, m); 13C NMR (100 MHz, DMSO-d6) d
186.4, 186.3, 172.0, 171.9, 170.9, 169.2, 168.1, 167.8,
160.7, 156.1, 154.5, 143.6 (2), 140.7, 137.7, 136.2,
135.4, 134.6, 134.0, 129.2, 128.0, 127.7, 127.2, 126.2,
125.0, 120.1, 119.9, 119.6, 118.9, 110.7, 110.6, 100.4,
74.9, 70.0, 68.0, 66.6, 65.6, 63.7, 60.6, 56.5, 53.8, 51.2,
46.2, 45.1, 42.0, 41.7, 40.7, 37.5, 36.6, 32.0, 29.8, 26.6,
24.1, 22.9, 21.6, 17.0; IR (neat/NaCl) 3281, 2927, 2106,
1728, 1631, 1446, 1284 cmꢀ1; LRMS (ESI, m/z, MH+)
1252; HRMS (ESI) for C51H60N8O18 Calcd (MNa+):
1273.4700. Found 1273.4681. The red solid (0.023 g,
0.018 mmol) was then dissolved in DMF (0.4 mL) and
(neat/NaCl) 3289, 2956, 1644, 1546, 1250, 1175 cmꢀ1
;
LRMS (ESI, m/z, MH+) 593, (MNH4+) 610.
5.1.11. 4-Azido-4-{[(1-{3-methyl-1-[(2-trimethylsilanyl-
ethoxycarbonylmethyl)-carbamoyl]-butylcarbamoyl}-2-
phenyl-ethylcarbamoyl)-methyl]-carbamoyl}-butyric acid
9H-fluoren-9-ylmethyl ester (14). Compound 13
(0.220 g, 0.37 mmol) was dissolved in 4 N HCl/dioxane
(2.0 mL) and stirred at room temperature for 1 h. Sol-
vent was removed in vacuo to give the crude amine as
an HCl salt, which was used directly in the next step
(0.196 g, quant). The residue (0.196 g, 0.37 mmol) was
dissolved in DMF (6 mL) and to it were added HOBt
(0.100 g, 0.742 mmol), N-methylmorpholine (0.07 mL,
0.631 mmol), and 2-azido-pentanedioic acid 5-(9H-fluo-
ren-9-ylmethyl) ester29 (0.169 g, 0.48 mmol). The reac-
tion mixture was cooled in an ice bath and to it was
added EDC (0.142 g, 0.74 mmol). The reaction mixture
was allowed to warm to room temperature and was stir-
red for 1 day. Solvent was removed in vacuo, the crude
dissolved in saturated NaHCO3 and extracted with
CH2Cl2 (4·). The combined organic layers were dried

S. Hanessian et al. / Bioorg. Med. Chem. 16 (2008) 2921–2931
2929
to it was added N-methylpyrrolidine (0.2 mL) and the
reaction mixture was stirred in the dark at room temper-
ature for 1 h. The solvent was removed in vacuo, the
crude product dissolved in small amount of CH2Cl2
and MeOH, then Et2O was added dropwise and with
stirring to precipitate the product. The mixture was cen-
trifuged and the supernatant was discarded. After cen-
trifugation the pellet was isolated and dried. Product
was obtained following column chromatography (7:3
SPIONs were stabilized at pH 6–7, adjusted with diluted
HNO3 or NH3.
5.3. Characterization of drug-functionalized-SPIONs
Particle size distributions were measured by photon cor-
relation spectroscopy (PCS) using a Brookhaven appa-
ratus
equipped
with
a
BI-9000AT
digital
autocorrelator instrument and a He–Ne laser beam at
a wavelength of 661 nm (scattering angle of 90ꢁ). The
CONTIN method was used for data processing. Viscos-
ity and refraction index of pure water were used for size
distribution calculation. Coated-SPIONs were diluted
51-fold in 0.01 M HNO3 in ultrapure water pre-filtered
on 20 nm ceramic filters (Whatman, Anodisc 25). The
same setting equipped with platinum electrodes was
used for electrophoretic mobility measurements and zeta
potential was calculated using the Smoluchowski
approximation. The electrodes were cleaned for 5 min
in an ultrasonic bath prior to each measurement.
1
EtOAc/MeOH) as a red solid (0.014 g, 70%); H NMR
(400 MHz, DMSO-d6) d 8.52–8.30 (3H, m), 7.97–7.79
(2H, m), 7.73–7.66 (1H, m), 7.59–7.52 (1H, m), 7.41–
7.35 (1H, m), 7.26–7.11 (5H, m), 5.28–5.22 (1H, m),
4.95 (1H, br s), 4.61–4.46 (2H, m), 4.28–4.11 (2H, m),
4.02–3.81 (4H, m), 3.73–3.56 (4H, m), 3.47–3.23 (4H,
m), 3.04–2.79 (3H, m), 2.26–2.08 (3H, m), 2.01–1.79
(3H, m), 1.61–1.40 (4H, m), 1.25–1.03 (4H, m), 0.90–
0.74 (6H, m); IR (neat/NaCl) 3334, 2953, 2107, 1651,
1574, 1410, 1286 cmꢀ1; HRMS (ESI) for C51H60N8O18
Calcd (MNa+): 1095.3918. Found 1095.3888.
5.1.13. DOX-linker-aminoPVA (16). A solution of 15
(13.4 mg, 0.0125 mmol), HOBt (1.7 mg, 0.0125 mmol)
NEt3 (3.5 lL, 0.0250 mmol), and EDC (4.8 mg,
0.0250 mmol) in DMF (240 lL, final concentration
0.05 M) was stirred for 60 min. Then a solution of amin-
oPVA in DMF/H2O (1:1, 0.7 mL, 10 mg/mL, 2.5% N,
0.0125 mmol) was added and the pH adjusted to ꢂ8.3
with 0.1 N HCl and/or 0.1 N NaOH. The mixture was
stirred at room temperature for 1 day. The solution
was then dialyzed (Spectra/Porꢂ 6 dialysis membrane,
MWCO 3500) against 0.2 N NaCl solution for 3 days
and then against H2O for 2 days changing the solution
twice daily. Product was obtained following lyophiliza-
tion as a red solid (9.1 mg).
Size distribution was also determined by gel-filtration.
Samples of 15 lL of aminoPVA-SPIONs or drug-SPI-
ONs were applied to a Sephadex G-75 (Pharmacia,
Uppsala, Sweden) column (0.6 · 19 cm) equilibrated in
25 mM NaH2PO4Æ2H2O–0.15 M NaCl, pH 7.0–7.1
(PBS), pre-filtered through a 0.22 lm filter (Millexꢂ
GP, Millipore), at a flow rate of 20 mL/h and room tem-
perature. The fractions were analyzed for iron, polyvi-
nylalcohol, and fluorescence. For iron determination
50 lL of each fraction was mixed with 50 lL, 6 N HCl
for 1 h at room temperature, and 100 lL of a 5% solu-
tion of K4[Fe(CN)6Æ3H2O] was added for 30 min. Absor-
bance was read at 690 nm in a multiwell-plate reader
(Labsystems iEMS Reader MF, BioConcepts, Allschwil,
Switzerland) and iron was quantified from a standard
curve of aminoPVA-SPIONs treated in same conditions.
For PVA quantification, to 20 lL of each fraction,
110 lL of distilled water and 70 lL of an iodine solution
(0.45% KI, 0.225% I2, and 3.6% H3BO3 (w/v)) were
added and incubated at room temperature for 30 min.
Absorbance was read at 690 nm in a multiwell-plate
reader and the amount of PVA was quantified from a
standard curve of PVA treated in same conditions.
The fluorescence of DOX in 6 N HCl was measured at
kex/kem of 485/580 nm in a multiwell-plate reader (Cyto-
Fluor, Series 4000, PerSeptive Biosystems).
5.2. Preparation of drug-functionalized-SPIONs
Superparamagnetic iron oxide nanoparticles (ferrofluid)
were prepared according to previously described meth-
ods.22–24,35,36 Briefly, ferrofluid was prepared by alkaline
co-precipitation of ferric (0.086 M) and ferrous
(0.043 M) chlorides (Fluka, Buchs, Switzerland). After
washing with water, the black precipitate was refluxed
in 0.8 M nitric oxide–0.21 M Fe(NO3)3Æ9H2O (Fluka,
Buchs, Switzerland) for 1 h, cooled, and the brown pre-
cipitate was dispersed in H2O and dialyzed for 2 days
against 0.01 M HNO3.
5.4. Biological assays
To obtain aminoPVA-SPIONs the ferrofluid was mixed
with poly(vinylalcohol) (PVA 3-83, Mowiol, MW
7600 g/mol, courtesy of Clariant) and poly(vinylalco-
hol/vinylamine) (aminoPVA M12, Erkol, MW
20,000 g/mol, courtesy of Erkol) as described previ-
ously22,24 and the pH was adjusted to 7.0. For the exper-
iments described here, the ratio of polymer to iron was
10 and the ratio of PVA to aminoPVA copolymer was
45 (mass ratios).
5.4.1. Enzymatic digestion of DOX-SPIONs 16S and
evaluation of drug release. For the evaluation of the re-
lease of DOX from DOX-SPIONs 16S by cellular prote-
ases, human melanoma cells (Me300, see below) were
grown to confluence in Petri dishes and thiol-cathepsins
were extracted from the cell layer in 0.1 M Na2HPO4Æ2-
H2O, 2 mM EDTA, 8 mM DTT, pH 6.0, at 4 ꢁC by 3
cycles of freeze–thaw in liquid nitrogen. Cathepsin B
activity in the extract was controlled with the substrate
essentially as previously described.31 Then 2 volumes
of DOX-SPIONs 16S were exposed for 24 h at 37 ꢁC
to 1 volume of cell extract, centrifuged 5 min at
1000 rpm, and 20 lL of supernatant was applied to a
For the preparation of drug-functionalized-SPIONs, the
ferrofluid was mixed in ultrapure water with PVA,
aminoPVA copolymer, and drug-linker-aminoPVA at
various ratios (see Table S2). The drug-functionalized-

2930
S. Hanessian et al. / Bioorg. Med. Chem. 16 (2008) 2921–2931
MFe
MTðFeÞ
Sephadex G-75 column in phosphate buffer, pH 7. The
eluting fractions were monitored for iron, PVA, and
DOX, as described above.
N6–12 ¼ 100 ꢃ
MT(Fe): total iron (Fe) mass of all 100 nanoparticles
sample [mg];
5.4.2. Cells and cell treatment. The Me300 human mel-
anoma cell line (a kind gift of D. Rimoldi, Ludwig
Institute, Lausanne, Switzerland) was grown in RPMI
1640 medium containing 10% FCS and antibiotics (all
from Gibco, Invitrogen, Basel, Switzerland) at 37 ꢁC
and 6% CO2. Two to three days prior to experiments,
the cells were detached with trypsin–EDTA (Gibco),
centrifuged, and grown in complete medium in multi-
well-plates (Costar, Corning, NY, USA), in order to
reach 60–80% confluence on the day of experiment.
Fresh complete medium was added prior to exposure
to the drugs for the concentration and time indicated.
Fur was obtained from Sigma (Steinheim, Germany)
and 25 mM stock solutions were prepared in PBS.
DOX solution for clinical use (Adriblastin solution,
10 mg/5 mL in 0.9% NaCl) was purchased from Pfizer
AG. Experiments were performed in triplicate wells.
Thymidine incorporation to quantify DNA synthesis
and Alamar Blue reduction to determine cell viability
were performed (see below). Alternatively, the cell lay-
ers were washed twice with saline and cellular iron
content was quantified using the Prussian Blue method
and drug content by measuring drug fluorescence (see
below). Cell number was determined by counting in a
hemocytometer under a phase-contrast microscope
and protein content was determined in cell extracts
using the BCA reagent (Pierce, Socochim, Lausanne,
Switzerland).
MFe: mass of iron (Fe) in a sample (measured by Prus-
sian Blue reaction) [mg];
N6–12: amount of Fe2O3 nanoparticles in MFe mg of iron
(Fe) in a sample, calculated between 6 and 12 nm range.
5.4.5. Calculation of results. Each experiment was re-
peated in triplicate wells at least twice. Means and stan-
dard deviation were calculated.
Acknowledgments
We want to thank Mrs. Catherine Chapuis Bernasconi for
excellent technical assistance and Dr. Heinrich Hofmann,
EPFL, for helpful discussion. This works was supported
by grants from the Swiss National Scientific Research
Foundation (Grant No. 3152A0-105705) and the Swiss
League and Research against Cancer (Grant No. KLS-
01308-02-2003). We also thank NSERC (Canada) and
FQRNT (Province of Quebec) for financial assistance.
Supplementary Material
The complete syntheses of compounds 7, 8, and 9, as
well as Schemes S1, S2, and S3 and Tables S1 and S2
are provided as supplementary material. Supplementary
data associated with this article can be found, in the on-
line version, at doi:10.1016/j.bmc.2007.12.051.
5.4.3. Evaluation of cell viability and DNA synthesis. To
quantify metabolically active cells, the cells were ex-
posed to 10% Alamar Blue (Serotec, Dusseldorf, Ger-
¨
many) added to the cell culture medium and
fluorescence increase was directly measured in a multi-
well fluorescence reader (kex/kem = 530/580 nm) after
2 h at 37 ꢁC. To assess DNA synthesis 1 lCi/mL
References and notes
[3H]thymidine (Amersham Pharmacia, Dubendorf,
¨
Switzerland) was added for the last 2 h of exposure to
compounds and incorporation was quantified in a
beta-counter (Rackbeta, LKB) after precipitation with
10% trichloroacetic acid and solubilization in 0.1 N
NaOH–1% SDS.
1. Haag, R.; Kratz, F. Angew. Chem. Int. Ed. 2006, 45, 1198–
1215.
2. Brigger, I.; Dubernet, C.; Couvreur, P. Adv. Drug Deliv.
Rev. 2002, 54, 631–651.
3. Juillerat-Jeanneret, L. In Nanomaterials for Cancer Ther-
apy; Kumar, C. S. S. R., Ed.; Wiley-VCH Verlag GmbH:
Weinheim, 2006; Vol. 6, pp 199–232.
5.4.4. Total cell-bound iron and drug determination. To
quantify cellular iron content, the cell layer was dis-
solved for 1 h in 6 N HCl (125 lL/well of a 48-well
plate), then 125 lL of a 5% solution of K4[Fe(CN)6]Æ3-
H2O in H2O was added for 20 min and the absorbance
was read at 690 nm in a multiwell-plate reader. A stan-
dard curve of aminoPVA-SPIONs in 6 N HCl was trea-
ted in the same conditions to quantify the amount of
cell-bound iron. Cellular drug uptake was determined
in lysed cells by measuring DOX fluorescence intensity
in 6 N HCl at kex/kem of 485/580 nm in a multiwell fluo-
rescence plate reader. To calculate the number of SPI-
ONs added to cells or taken up by cells, a mean
diameter of 9 3 nm, corresponding to a mean iron
content of 1.3 · 10ꢀ15 mg Fe per particle, was calculated
according to:22
4. Ranney, D.; Antich, P.; Dadey, E.; Mason, R.; Kulkarni,
P.; Singh, O.; Chen, H.; Constantinescu, A.; Parkey, R. J.
Control. Release 2005, 109, 222–235.
´
5. Arruebo, M.; Fernandez-Pacheco, R.; Ibarra, M. R.;
Santamaria, J. Nano Today 2007, 2, 22–32.
6. Juillerat-Jeanneret, L.; Schmitt, F. Med. Res. Rev. 2007,
27, 574–590.
7. Schoenmakers, R. G.; van de Wetering, P.; Elbert, D. L.;
Hubbell, J. A. J. Control. Release 2004, 95, 291–300.
8. Alexiou, C.; Arnold, W.; Klein, R. J.; Parak, F. G.; Hulin,
P.; Bergemann, C.; Erhardt, W.; Wagenpfeil, S.; Lubbe,
¨
A. S. Cancer Res. 2000, 60, 6641–6648.
9. Jain, T. K.; Morales, M. A.; Sahoo, S. K.; Leslie-Pelecky,
D. L.; Labhasetwar, V. Mol. Pharm. 2005, 2, 194–205.
10. Nasongkla, N.; Bey, E.; Ren, J.; Ai, H.; Khemtong, C.;
Guthi, J. S.; Chin, S.-F.; Sherry, A. D.; Boothman, D. A.;
Gao, J. Nano Lett. 2006, 6, 2427–2430.

S. Hanessian et al. / Bioorg. Med. Chem. 16 (2008) 2921–2931
2931
11. Lubbe, A. S.; Bergemann, C.; Huhnt, W.; Fricke, T.;
¨
Riess, H.; Brock, J. W.; Huhn, D. Cancer Res. 1996, 56,
4694–4701.
12. Lubbe, A. S.; Bergemann, C.; Riess, H.; Schriever, F.;
¨
24. Cengelli, F.; Maysinger, D.; Tschudi-Monnet, F.; Mon-
tet, X.; Corot, C.; Petri-Fink, A.; Hofmann, H.;
Juillerat-Jeanneret, L. J. Pharmacol. Exp. Ther. 2006,
318, 108–116.
Reichardt, P.; Possinger, K.; Matthias, M.; Do¨rken,
B.; Herrinann, F.; Gurtler, R.; Hohenberger, P.; Haas,
N.; Sohr, R.; Sander, B.; Lemke, A.-J.; Ohlendorf, D.;
Huhnt, W.; Huhn, D. Cancer Res. 1996, 56, 4686–
4693.
25. Liu, S.-T.; Liu, C.-Y. J. Org. Chem. 1992, 57, 6079–6080.
26. Carlsen, P. H. J.; Katsuki, T.; Martin, V. S.; Sharpless, K.
B. J. Org. Chem. 1981, 46, 3936–3938.
27. Watanabe, K. A.; Matsuda, A.; Halat, M. J.; Hollenberg,
D. H.; Nisselbaum, J. S.; Fox, J. J. J. Med. Chem. 1981,
24, 893–897.
13. Rudge, S.; Peterson, C.; Vessely, C.; Koda, J.; Stevens, S.;
Catterall, L. J. Control. Release 2001, 74, 335–340.
14. Kreuter, J. J. Nanosci. Nanotechnol. 2004, 4, 484–488.
15. Alexiou, C.; Jurgons, R.; Schmid, R. J.; Bergemann, C.;
Henke, J.; Erhardt, W.; Huenges, E.; Parak, F. J. Drug
Target. 2003, 11, 139–149.
16. Cinteza, L. O.; Ohulchanskyy, T. Y.; Sahoo, Y.; Bergey,
E. J.; Pandey, R. K.; Prasad, P. N. Mol. Pharm. 2006, 3,
415–423.
17. Devineni, D.; Blanton, C. D.; Gallo, J. M. Bioconjug.
Chem. 1995, 6, 203–210.
18. Devineni, D.; Klein-Szanto, A.; Gallo, J. M. J. Neuroon-
col. 1995, 24, 143–152.
19. Kohler, N.; Sun, C.; Wang, J.; Zhang, M. Langmuir 2005,
21, 8858–8864.
20. Kohler, N.; Sun, C.; Fichtenholtz, A.; Gunn, J.; Fang, C.;
Zhang, M. Small 2006, 2, 785–792.
21. Weissleder, R.; Kelly, K.; Sun, E. Y.; Shtatland, T.;
Josephson, L. Nat. Biotechnol. 2005, 23, 1418–1423.
22. Chastellain, M.; Petri, A.; Hofmann, H. J. Colloid
Interface Sci. 2004, 278, 353–360.
23. Petri-Fink, A.; Chastellain, M.; Juillerat-Jeanneret, L.;
Ferrari, A.; Hofmann, H. Biomaterials 2005, 26, 2685–
2694.
ˇ
´
28. Lu, Z.-R.; Shiah, J.-G.; Sakuma, S.; Kopeckova, P.;
ˇ
Kopecek, J. J. Control. Release 2002, 78, 165–173.
29. Lundquist, J. T., IV; Pelletier, J. C. Org. Lett. 2002, 4,
3219–3221.
30. Ozaki, S.; Akiyama, T.; Ike, Y.; Mori, H.; Hoshi, A.
Chem. Pharm. Bull. 1989, 37, 3405–3408.
31. Aubert, J. D.; Juillerat-Jeanneret, L.; Fioroni, P.; Dayer,
P.; Plan, P. A.; Leuenberger, P. Pulmonol. Pharmacol.
Ther. 1998, 11, 263–269.
32. McLay, I. M.; Halley, F.; Souness, J. E.; McKenna, J.;
Benning, V.; Birrell, M.; Burton, B.; Belvisi, M.; Collis, A.;
Constan, A.; Foster, M.; Hele, D.; Jayyosi, Z.; Kelley, M.;
Maslen, C.; Miller, G.; Ouldelhkim, M.-C.; Page, K.;
Phipps, S.; Pollock, K.; Porter, B.; Ratcliffe, A. J.;
Redford, E. J.; Webber, S.; Slater, B.; Thybaud, V.;
Wilsher, N. Bioorg. Med. Chem. 2001, 9, 537–554.
33. Sieber, P. Helv. Chim. Acta 1977, 60, 2711–2716.
34. Burns, M. R.; Coward, J. K. Bioorg. Med. Chem. 1996, 4,
1455–1470.
35. Massart, R.; Dubois, E.; Cabuil, V.; Hasmonay, E. J.
Magn. Magn. Mater. 1995, 149, 1–5.
36. van Ewijk, G. A.; Vroege, G. J.; Philipse, A. P. J. Magn.
Magn. Mater. 1999, 201, 31–33.