The facile synthesis of multifunctional PAMAM dendrimer conjugates through copper-free click chemistry

Article abstract of DOI:10.1016/j.bmcl.2012.03.052

The facile conjugation of three azido modified functionalities, namely a therapeutic drug (methotrexate), a targeting moiety (folic acid), and an imaging agent (fluorescein) with a G5 PAMAM dendrimer scaffold with cyclooctyne molecules at the surface thro

Full text of DOI:10.1016/j.bmcl.2012.03.052

Bioorganic & Medicinal Chemistry Letters 22 (2012) 3152–3156
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The facile synthesis of multifunctional PAMAM dendrimer conjugates
through copper-free click chemistry
⇑
Baohua Huang , Jolanta F. Kukowska-Latallo, Shengzhuang Tang, Hong Zong, Kali B. Johnson,
⇑
Ankur Desai, Chris L. Gordon, Pascale R. Leroueil, James R. Baker Jr.
Michigan Nanotechnology Institute for Medicine and Biological Sciences, University of Michigan, 9220 MSRB III, Box 0648, Ann Arbor, MI 48109, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The facile conjugation of three azido modified functionalities, namely a therapeutic drug (methotrexate),
a targeting moiety (folic acid), and an imaging agent (fluorescein) with a G5 PAMAM dendrimer scaffold
with cyclooctyne molecules at the surface through copper-free click chemistry is reported. Mono-, di-,
and tri-functional PAMAM dendrimer conjugates can be obtained via combinatorial mixing of different
azido modified functionalities simultaneously or sequentially with the dendrimer platform. Preliminary
flow cytometry results indicate that the folic acid targeted nanoparticles are efficiently binding with KB
cells.
Received 31 January 2012
Revised 8 March 2012
Accepted 13 March 2012
Available online 21 March 2012
Keywords:
PAMAM dendrimer
Copper-free click conjugation
Drug delivery
Ó 2012 Elsevier Ltd. All rights reserved.
For several decades, polyamidoamine (PAMAM) dendrimers
have been used as a platform to deliver therapeutics and imaging
agents, both in vitro1 and in vivo.2 A sequential synthetic strategy
based on amide and ester chemistry is often used to conjugate
small molecules, such as therapeutic drugs, targeting groups and
imaging agents, to generate the functional forms of the dendri-
mer.3 While sufficient for small-scale needs, the extensive purifica-
tion & characterization required between each small molecule
addition, as well as batch-to-batch reproducibility issues, made it
ill-suited for large-scale needs. Cu(I)-catalyzed alkyne azide
1,3-dipolar cycloaddition (CuAAC), widely used for bioconjugation
due to its high yields, tolerance toward functional groups, and lack
of concurrent side-reactions, was initially an attractive alternative
to our amide and ester conjugation chemistry.4 However, the cop-
per catalyst proved difficult to remove from the dendrimers5 and
the cytotoxicity6 associated with the residual copper prevented
CuAAC from being a viable option. Several metal-free bioorthogo-
nal labeling reactions such as strain-promoted cycloaddition of
azides,7 which relies on the unusually high ring strain associated
with cyclooctyne to react rapidly and selectively with azide
groups,8 have since been developed.9
O
OH
O
N3
NH2
N
NH
O
NH
N
N
N
O
N
OH
O
NH2
γ
-N3-MTX (2)
CO2H
O
Ac92
O
NH
G5
N3
NH NH
O
20
N3-FITC (3)
G5 dendrimer platform (1)
Figure 1. Structures of PAMAM dendrimer platform,
azido modified fluorescein (numbers of functionalities are rounded to the nearest
c
-azido modified MTX, and
integers).
amine groups on the dendrimer surface were neutralized by acet-
ylation ( Fig. 1, compound 1). Variable equivalents (5, 10, and
18 equiv) of c-azido modified methotrexate (c-N3-MTX) were re-
acted with the dendrimer platform efficiently with >90% of the
modified drug conjugated. Here we report the facile conjugation
of multiple functional groups with the PAMAM dendrimer plat-
form via copper-free click chemistry. Three types of azido modified
Recently, we synthesized a G5 PAMAM platform with cyclooc-
tyne ligand on the surface for copper-free click conjugation with
azido modified functionalities.10 Briefly, the G5 PAMAM dendrimer
was first partially (70%) acetylated and then coupled with 20 cyc-
looctyne ligands through amide bonds. The remaining primary
functionalities, namely a therapeutic drug (
c-N3-MTX, 2), a target-
ing group ( -N3-folic acid, 7), and an imaging agent (N3-FITC, 3)
were reacted with the G5 dendrimer platform simultaneously or
sequentially to achieve multifunctional nanodevices.
c
⇑
Corresponding authors.
E-mail addresses: baohua@med.umich.edu, baohua@umich.edu (B. Huang),
jbakerjr@med.umich.edu (J.R. Baker).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.03.052

B. Huang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 3152–3156
3153
c
-N3-MTX (Fig. 1, compound 2) was synthesized using our pub-
lished method. Briefly, 4-amino-4-deoxy-N10-methylpteroic acid
was first reacted with Glu- -OtBu using BOP reagent to give
OtBu-MTX. The -carboxylic acid group was then coupled with
3-azido-propylamine in the presence of HATU. After exposure of
the carboxylic acid functional group using trifluoroacetic acid,
-N3-MTX was obtained in excellent yield.10 Azido modified fluo-
they would have similar reactivity. Therefore, the di-functional
conjugate was made by mixing these two compounds simulta-
neously with the platform. The conjugates containing the structur-
ally dissimilar fluorescein were synthesized in a slightly different
manner to ensure that the average number of fluoresceins per con-
jugate was identical, a critical requirement for the in vitro cell
binding studies. In that case, we mixed the N3-FITC with the den-
drimer platform and allowed the click reaction to go to completion.
The reaction mixture was then separated into four equal aliquots;
one aliquot was purified to give compound 11. Subsequent addi-
a
a-
c
a
c
rescein (Fig. 1, compound 3) was synthesized in a straightforward
manner using fluorescein isothiocyanate and 3-azido-propylamine
(see Supplementary data).
A number of cancer cell lines exhibit high levels of the cell sur-
face folate receptor (FR) while its expression is highly restricted
in most normal tissues. The degree of overexpression of the FR on
human epithelial cancer cells ranges from 30-fold (ovarian carci-
noma) to 1400-fold (endometrial cancer).4b Hence, the FR is a very
attractive molecular target for receptor-based tumor imaging and
therapy.11 The folic acid structure offers two positions for bioconju-
tion of c-N3-FA, c-N3-MTX, and both of these two compounds to
three of the aliquots gave conjugates 12, 13, and 14, each contain-
ing the same average number of fluorescein molecules (Scheme 2).
All of the click reactions were stirred for 24 h at room temperature
shielded from light. After removal of the organic solvent, the resi-
dues were dissolved in phosphate buffer saline (1Â PBS buffer,
pH = 7.4) and purified by 10 K centrifugal filters (PBS Â 3, DI
water  6) and then lyophilized to give compounds 8–14 with
good yields. More than 90% of the azido modified functionalities
were conjugated through click reaction with the dendrimer
platform.
gation, the
a- and c-carboxylic acid. Whether or not the binding
affinity and in vivo pharmacological behavior is dependent on the
site of conjugation is still debated in the literature.12 Here we
synthesized a
synthesized a
c
-azido modified folic acid (Scheme 1). Mindt et al.
c-azido modified folic acid starting from double-pro-
Using the integration of the acetamide proton of the dendrimer
platform as an internal reference, the average number of each type
of small molecule conjugated was determined using 1H NMR. The
aromatic proton in the pteridine ring of MTX and folic acid had a
chemical shift at 8.61 and 8.64 ppm.10,13 After the conjugation,
the pteridine proton signal of folic acid shifted down field to
8.71–8.73 ppm.13 By comparing the integration of these two peaks
to the integration of the acetamide proton peak we determined the
numbers of MTX and folic acid molecules conjugated to the dendri-
mer platform. The 1H NMR spectra of compounds 8, 9, and 10 are
shown in Figure 2. In these cases, the integrations of the internal
reference peaks were set to 276 which represented 96 acetamide
groups.10 The results indicated that the average numbers of MTX
and folic acid in compound 10 were 5.95 and 3.06. Compound 8
had 5.93 MTX molecules and compound 9 had 2.79 folic acid mol-
ecules per dendrimer molecule (see inserts in Fig. 2), respectively.
The 1H NMR spectra of compounds 11–14 are shown in Figure
3. The number of fluorescein molecules conjugated was first calcu-
lated based on the relative proton integrations of the eight aro-
matic protons in the fluorescein structure, located between 6.40
and 7.81 ppm, and the internal reference. The result indicated
there were 3 dye molecules per dendrimer molecule averagely
(Fig. 3c). Because compounds 12, 13, and 14 were synthesized by
subsequent additions of azido modified MTX and/or folic acid,
the average number of fluorescein molecules per conjugate was
identical. Using the same method described above, we determined
tected pteroic acid.13 Starting from commercially available N10-(tri-
fluoroacetyl) mono-protected pteroic acid, we synthesized
folic acid (5) in five steps with good yield. Originally, we tried to use
BocGluOMe as a precursor,13 but the final removal of
-methyl es-
c-azido
a
ter with NaOH yielded 8–15% racemization (data not shown). In or-
der to avoid the basic conditions that were responsible for the
racemization, we replaced BocGluOMe with FmocGluOtBu. Briefly,
FmocGluOtBu was coupled with 3-azido-propylamine to obtain
compound 4, followed by the exposure of amino group with piper-
idine to give compound 5. N10-(trifluoroacetyl) pteroic acid was
then coupled with 5 to give a folic acid intermediate with a-tBu es-
ter and N10-(trifluoroacetyl) protecting groups (6). The removal of
these two groups sequentially with ammonium hydroxide and
TFA gave compound 7. Compound 7 was obtained by precipitation
with diethyl ether and collected by centrifugation (see Supplemen-
tary data). We later found that the protection of the primary amino
group at the pteridine ring was not necessary.
The copper-free click conjugation of the dendrimer platform
with azido modified functionalities was then carried out (Scheme
2). Briefly, the methanol solution of G5 PAMAM dendrimer with
cyclooctyne ligand molecules (1) was mixed simultaneously or
sequentially with one, two or three of the previously mentioned
targeting moiety (c-N3-FA, 7), therapeutic drug (c-N3-MTX, 2),
and imaging agent (N3-FITC, 3). Because of the structural similarity
of the azido modified folic acid and methotrexate, we assumed that
O
O
O
O
O
H
Piperidine/DCM
Fmoc-Glu-OtBu
H
N3
H2N
N
N3
O
N
H
N
N3
HATU
85%
H2N
O
DMF/DIPEA
O
O
98%
4
5
O
OH
O
O
O
H
N
H
N3
N3
N
O
O
N10
-
(trifluoroacetyl)
N
N
H
H
N
N
N
pteroic acid
O
O
HN
1. NH4OH
2. TFA/DCM
70%
N
HN
H2N
N
H
HATU
DMF/DIPEA
52%
H2N
N
N
O
CF3
N
7
6
Scheme 1. The synthesis of a c-azido modified folic acid.

3154
B. Huang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 3152–3156
(MTX)6
(FITC)3
(MTX)6
12
G5
G5
G5
8
γ
-
N
TX
3
-
M
M
-
T
X
A
N 3
-
γ
(M TX)6
O
γ
O
-N3-FA
N3-FITC
γ
-N3-MTX
-N3-FA
N
H
G5
G5
G5
(FA)3
G5
γ
(FA)3
20
(FITC)3
(FITC)3
X
13
9
T
1
11
M
-
3
N
-
γ
F
-
γ
3
N
-
N3
-
γ
-
F
A
(F A) 3
(FITC)3
(MTX)6
10
G5
14
(FA)3
Scheme 2. The copper-free click conjugations of the dendrimer platform with azido modified functionalities.
Figure 2. 1H NMR of compounds 10(a), 8(b), and 9(c). Inserts are the enlargements of the spectra (8.40–8.90 ppm) showing the pteridine ring proton signals of folic acid (FA)
and methotrexate (MTX).

B. Huang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 3152–3156
3155
Figure 3. 1H NMR of compounds 11(a), 13(b), 12(c), and 14(d).
the average numbers of MTX and folic acid molecules conjugated
to the platform. Results indicated that there were 3.53 molecules
of folic acid and 6.10 molecules of MTX in compound 13
(Fig. 3b), 6.06 molecules of MTX in compound 12 (Fig. 3c), and
3.37 molecules of folic acid in compound 14 (Fig. 3d), respectively.
The KB human tumor cell line, which overexpresses the folate
receptor14 was purchased from the American Type Tissue Collec-
tion (Manassas, VA) and maintained in vitro at 37 °C, 5% CO2 in fo-
late-deficient RPMI 1640 supplemented with penicillin (100 units/
mL), streptomycin (100 lG/mL), and 10% heat-inactivated fetal bo-
vine serum. The cells were harvested with trypsin–EDTA solution,
washed and resuspended in PBS. The cell suspension (106 cells in
0.1 mL) was incubated for 30 min on ice with 0.3 lM of targeted
(G5-FITC-FA) or non-targeted (G5-FITC) conjugate. After incuba-
tion the cells were washed once with phosphate-buffered saline
(PBS) and re-suspended in PBS for analysis. Flow cytometry analy-
sis was done using Accuri C6 Flow Cytometer (BDBiosciences) and
CFlow Plus software. Data were reported as the mean channel fluo-
Figure 4. Flow cytometry analysis of targeted (G5-FITC-FA), non-targeted (G5-
FITC), and unstained control (PBS).
rescence of the cell population stained with 0.3
or G5-FITC. Flow cytometry results indicated that the targeted
l
M of G5-FITC-FA
nanoparticle binds specifically to KB cells. The non-targeted nano-
particle did not bind to KB cells (Fig. 4).

3156
B. Huang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 3152–3156
2. Myc, A.; Kukowska-Latallo, J.; Cao, P.; Swanson, B.; Battista, J.; Dunham, T.;
In conclusion, we have successfully synthesized mono-, di-, and
Baker, J. R. Anticancer Drugs 2010, 21, 186.
tri-functionalized G5 PAMAM dendrimer conjugates with copper-
free click conjugation method. An azido modified targeting moiety,
a therapeutic drug, and an imaging reagent were mixed with a G5
PAMAM dendrimer nanoplatform simultaneously or sequentially
to give mono-, di-, and tri-functional conjugates. The reactions
were very efficient and reproducible. Preliminary flow cytometry
results indicate that the FA targeted nanoparticles are efficiently
binding with KB cells. The in vitro toxicity studies of these materi-
als toward different cancer cell lines are under investigation.
3. (a) Thomas, T. P.; Shukla, R.; Kotlyar, A.; Liang, B.; Ye, J. Y.; Norris, T. B.; Baker, J.
R. Biomacromolecules 2008, 9, 603; (b) Shukla, R.; Thomas, T. P.; Peters, J.;
Kotlyar, A.; Myc, A.; Baker, J. R. Chem. Commun. 2005, 5739.
4. (a) Meldal, M.; Tornoe, C. W. Chem. Rev. 2008, 108, 2952; (b) Mullen, D. G.;
McNerny, D. Q.; Desai, A.; Cheng, X.-m.; DiMaggio, S. C.; Kotlyar, A.; Zhong, Y.;
Qin, S.; Kelly, C. V.; Thomas, T. P.; Majoros, I.; Orr, B. G.; Baker, J. R., Jr.; Holl, M.
M. B. Bioconjug. Chem. 2011, 22, 679.
5. van Dijk, M.; Rijkers, D. T. S.; Liskamp, R. M. J.; van Nostrum, C. F.; Hennink, W.
E. Bioconjug. Chem. 2001, 2009, 20.
6. (a) Link, A. J.; Vink, M. K. S.; Tirrell, D. A. J. Am. Chem. Soc. 2004, 126, 10598; (b)
Gierlich, J.; Burley, G. A.; Gramlich, P. M. E.; Hammond, D. M.; Carell, T. Org. Lett.
2006, 8, 3639.
7. Sletten, E. M.; Nakamura, H.; Jewett, J. C.; Bertozzi, C. R. J. Am. Chem. Soc. 2010,
132, 11799.
Acknowledgements
8. Ornelas, C.; Broichhagen, J.; Weck, M. J. Am. Chem. Soc. 2010, 132, 3923.
9. Becer, C. R.; Hoogenboom, R.; Schubert, U. S. Angew. Chem., Int. Ed. 2009, 48,
4900.
10. Huang, B. H.; Desai, A.; Zong, H.; Tang, S. Z.; Leroueil, P.; Baker, J. R. Tetrahedron
Lett. 2011, 52, 1411.
This project has been funded in whole or in part with Federal
funds from the National Cancer Institute, National Institutes of
Health, under award 1 R01 CA119409.
11. (a) Kukowska-Latallo, J. F.; Candido, K. A.; Cao, Z. Y.; Nigavekar, S. S.; Majoros, I.
J.; Thomas, T. P.; Balogh, L. P.; Khan, M. K.; Baker, J. R. Cancer Res. 2005, 65,
5317; (b) Majoros, I. J.; Myc, A.; Thomas, T.; Mehta, C. B.; Baker, J. R.
Biomacromolecules 2006, 7, 572.
Supplementary data
12. (a) Ross, T. L.; Honer, M.; Lam, P. Y. H.; Mindt, T. L.; Groehn, V.; Schibli, R.;
Schubiger, P. A.; Ametamey, S. M. Bioconjug. Chem. 2008, 19, 2462; (b) Leamon,
C. P.; DePrince, R. B.; Hendren, R. W. J. Drug Target. 1999, 7, 157; (c) Ke, C. Y.;
Mathias, C. J.; Green, M. A. Nucl. Med. Biol. 2003, 30, 811.
13. Mindt, T. L.; Muller, C.; Melis, M.; de Jong, M.; Schibli, R. Bioconjug. Chem. 2008,
19, 1689.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2012.03.052.
References and notes
14. Turek, J. J.; Leamon, C. P.; Low, P. S. J. Cell Sci. 1993, 106, 423.
1. Thomas, T. P.; Majoros, I. J.; Kotlyar, A.; Kukowska-Latallo, J. F.; Bielinska, A.;
Myc, A.; Baker, J. R. J. Med. Chem. 2005, 48, 3729.