One-pot orthogonal copper-catalyzed synthesis and self-assembly of l-lysine-decorated polymeric dendrimers

Article abstract of DOI:10.1021/acs.macromol.5b00195

Synthetic peptides, including cyclic peptides and peptidomimetics, provide stability, protection, and long circulation times compared to free-circulating peptides. Dendritic structures with amino acids or peptides attached to the peripheral layer represent one form of peptidomimetics (i.e., a hybrid peptide/dendrimer construct) that has found use in biological applications. Constructing such dendritic structures from linear polymeric building blocks provides a further advantage of generating a highly ordered and defined structure in the nanoparticle size range. However, the rapid synthesis of such well-defined structures is still a challenge. In this work, we demonstrate that through modulating the copper activity concomitantly of the nitroxide radical coupling (NRC) and the azide-alkyne cycloaddition (CuAAC) reactions, polymeric dendrimers decorated with l-lysine on the periphery could be made rapidly in one pot at 25 °C. Three polymeric dendrimers were constructed with high purity (>94%) and with varying l-lysine density coated on the peripheral generation layer. The self-assembly of these dendrimers in water gave similar sizes to that found in organic solvents, suggesting that the aggregation number of dendritic structures in water was very low and possibly consisting of unimolecular micelles. The findings support the conclusion that the self-assembly of a dendritic architecture in water produces nanoparticles with predictable and well-controlled sizes. This synthetic methodology and the self-assembly properties represent an important step toward synthesizing peptide-decorated dendrimers targeted toward therapeutic applications.

Full text of DOI:10.1021/acs.macromol.5b00195

Article
pubs.acs.org/Macromolecules
One-Pot Orthogonal Copper-Catalyzed Synthesis and Self-Assembly
of L‑Lysine-Decorated Polymeric Dendrimers
Derong Lu, Md. D. Hossain, Zhongfan Jia, and Michael J. Monteiro*
Australian Institute for Bioengineering and Nanotechnology, The University of Queensland, Brisbane QLD 4072, Australia
S
* Supporting Information
ABSTRACT: Synthetic peptides, including cyclic peptides and peptidomimetics, provide stability, protection, and long
circulation times compared to free-circulating peptides. Dendritic structures with amino acids or peptides attached to the
peripheral layer represent one form of peptidomimetics (i.e., a hybrid peptide/dendrimer construct) that has found use in
biological applications. Constructing such dendritic structures from linear polymeric building blocks provides a further advantage
of generating a highly ordered and defined structure in the nanoparticle size range. However, the rapid synthesis of such well-
defined structures is still a challenge. In this work, we demonstrate that through modulating the copper activity concomitantly of
the nitroxide radical coupling (NRC) and the azide−alkyne cycloaddition (CuAAC) reactions, polymeric dendrimers decorated
with L-lysine on the periphery could be made rapidly in one pot at 25 °C. Three polymeric dendrimers were constructed with
high purity (>94%) and with varying L-lysine density coated on the peripheral generation layer. The self-assembly of these
dendrimers in water gave similar sizes to that found in organic solvents, suggesting that the aggregation number of dendritic
structures in water was very low and possibly consisting of unimolecular micelles. The findings support the conclusion that the
self-assembly of a dendritic architecture in water produces nanoparticles with predictable and well-controlled sizes. This synthetic
methodology and the self-assembly properties represent an important step toward synthesizing peptide-decorated dendrimers
targeted toward therapeutic applications.
applications to where larger sizes (>15 nm) were required,
including a self-adjuvanting vaccine16 and a cancer vaccine.17
Polymeric dendrimers combine the attributes of the highly
branched and symmetrical dendrimer structure, consisting of
well-defined functionality within the generational layers and
periphery, with the advantages of size and shape associated with
nanoparticles.13,18
Considerable work has been carried out to produce a wide
range of polymeric dendrimers either through divergent or
convergent methods.8,9,11,14,19−21 One of the first strategies to
combine “living” radical polymerization (LRP) and a chain-end
coupling process was denoted as TERMINI (terminator
multifunctional initiator), in which polymers were divergently
grown via LRP, terminated with a difunctional molecule that
INTRODUCTION
■
Recently, there has been a rapid increase in the number of
synthetic peptides with therapeutic efficacy covering a diverse
range of bioapplications.1 The synthesis of cyclic peptides and
peptidomimetics helped to overcome many of the drawbacks
limiting the use of peptides as therapeutics in the past. These
drawbacks included low stability in plasma, rapid degradation
by proteases, and rapid clearance from circulation.2 Synthetic
peptides in the form of dendrimers (i.e., peptidomimetics) were
found to be much more stable when incubated with human
plasma and serum.3 Dendrimers consisting of a fourth-
generation poly(L-lysine) with multivalent ligands on the
peripheral generational layer are in clinical trials as an antiviral
topical ointment.4,5 Although dendrimers built from small
molecules in each generation (with sizes <15 nm) have been
widely researched for many biological applications,6,7 den-
drimers synthesized from polymeric building blocks8−15
(denoted here as polymeric dendrimers) have extended the
Received: January 28, 2015
Revised: February 24, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
a
Scheme 1. Synthetic Route for Lysine Decorated Polymeric Third- and Fourth-Generational Layered Dendrimers
a
(i) One-pot, parallel CuAAC and NRC reaction. Polymer reactants with CuBr/PMDETA in a mixture of toluene/DMSO (50:50) for 30 min at 25
°C. (ii) Deprotection of Boc groups with TFA.
after deprotection acted as an initiator for further polymer
growtha process repeated until the desired generation was
achieved.8 Other strategies included the divergent coupling of
already synthesized linear polymers.11 More recently, with the
combination of LRP and “click” reactions, polymers were
synthesized with precise control over their architecture,
including stars,12,22 dendrimers,12,14,23−25 hyperbranched poly-
mers,26−28 multiblock polymers,29 and bioconjugates.30,31
There are now many examples of using “click” reactions to
couple linear polymers together; these include the copper(I)-
catalyzed azide−alkyne cycloaddition (CuAAC) reaction,
strain-promoted azide−alkyne coupling (SPAAC),32 Diels−
Alder,33 and thiol−ene reactions.34−37
Most dendrimers made from linear polymer building blocks
are synthesized divergently in a sequential and iterative manner.
This synthetic process requires multiple reaction steps,
including protection and deprotection of terminal end-groups,
and time-consuming purification before growth of the next
generational layer. The use of orthogonal “click” coupling
reactions avoids many of these issues and increases the coupling
efficiency with fewer reaction steps,37−39 but in many cases a
change in experimental conditions are required for the next
“click” reaction.
Here, we elaborate on a new process40 of using two
orthogonal “click”-type reactions (i.e., nitroxide radical coupling
(NRC) and CuAAC) modulated by a copper catalyst to
produce third- or fourth-generation layered polymer den-
drimers that are densely coated with L-lysine groups in one pot
at 25 °C (Scheme 1). This represents an advance on the
previous techniques which relied on changing experimental
conditions and purification at each generation step. By
controlling the experimental conditions through selection of
solvent and ligand, the synthesis of polymeric dendrimers in
one pot could proceed via a divergent, convergent, or parallel
process. The relative rates of the CuAAC and NRC reactions
could be controlled to be comparable or with one significantly
faster than the other. Utilizing the tool kit of building blocks
shown in Scheme 1, the lysine peripheral density could be well
B
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Absolute Molecular Weight Determination by Triple Detection
SEC (TD-SEC). Absolute molecular weights of polymers were
determined using a Polymer Laboratories GPC50 Plus equipped
with dual angle laser light scattering detector, viscometer, and
differential refractive index detector. HPLC grade N,N-dimethylace-
tamide (DMAc, containing 0.03 wt % LiCl) was used as the eluent at a
flow rate of 1.0 mL min−1. Separations were achieved using two PLGel
Mixed B (7.8 × 300 mm) SEC columns connected in series and held
at a constant temperature of 50 °C. The triple detection system was
calibrated using a 2 mg mL−1 PSTY standard (Polymer Laboratories:
Mw = 110K, dn/dc = 0.16 mL g−1, and IV = 0.5809). Samples of
known concentration were freshly prepared in DMAc + 0.03 wt %
LiCl and passed through a 0.45 μm PTFE syringe filter prior to
injection. The absolute molecular weights and dn/dc values were
determined using Polymer Laboratories Multi Cirrus software based
on the quantitative mass recovery technique.
controlled. The purity of the foundational building block, 5, is
critical to the success of the polymer dendrimer synthesis. In
general, the synthesis of 4-arm stars produces higher order star
architectures through star−star radical coupling.41 Limiting the
amount of star−star coupling allowed high-purity 4-arm star
polymer to be produced, enabling the high peripheral density of
L-lysine on the dendrimer. This method of producing well-
defined polymeric dendrimers with densely decorated amino
acids (or amino acid dendrons) on the periphery represents an
important step toward synthetic peptides designed as
therapeutics. To further validate the potential use for these
dendrimers as peptidomimetic nanoparticles, the self-assembly
of these structures in water was compared to that in organic
solvent.
Preparative Size Exclusion Chromatography (Prep-SEC). Crude
polymers were fractionated (i.e., purified) using a Varian Pro-Star
preparative SEC system equipped with a manual injector, differential
refractive index detector, and single wavelength ultraviolet visible
detector. The flow rate was maintained at 10 mL min−1, and HPLC
grade THF was used as the eluent. Separations were achieved using a
PL Gel 10 μm 10 × 103 Å, 300 × 25 mm preparative SEC column at
25 °C. The dried crude polymer was dissolved in THF at 100 mg/mL
and filtered through a 0.45 μm PTFE syringe filter prior to injection.
Fractions were collected manually, and the composition of each was
determined using the Polymer Laboratories GPC50 Plus equipped
with triple detection as described above.
Nuclear Magnetic Resonance (NMR). All NMR spectra were
recorded on either a Bruker DRX 400 or 500 MHz spectrometer using
an external lock (CDCl3), and all spectra were referenced to the
residual nondeuterated solvent (CHCl3). Spectra of functional
molecules and polymers containing nitroxide radicals were measured
in the presence of phenylhydrazine under an inert atmosphere to
reduce the nitroxides to their incipient hydroxylamines.
METHODS SECTION
■
Materials. Inhibitor was removed from styrene (STY: Aldrich,
>99%) before use by passing through a basic alumina column. The
following chemicals were used as received: alumina, activated basic
(Aldrich, Brockmann I, standard grade, ∼150 mesh, 58 Å), magnesium
sulfate (MgSO4: anhydrous, Scharlau, extra pure), sodium chloride
(NaCl: Univar, 99.9%), sodium iodide (NaI: Aldrich, 99.5%), sodium
azide (NaN3: Aldrich, 99.5%), 1,1,1-triisopropylsilyl chloride (TIPS-
Cl: Aldrich, 99%), tetrabutylammonium fluoride (TBAF: Aldrich, 1.0
M in THF), ethylmagnesium bromide solution (Aldrich, 3.0 M in
diethyl ether), triethylamine (TEA: Fluka, 98%), TLC plates (silica gel
60 F254), silica gel 60 (230−400 mesh ATM (SDS)), potassium
carbonate (K2CO3, AnalaR, 99.9%), 2-bromoisobutyryl bromide (BIB,
Aldrich, 98%), lithium aluminum hydride (LiAlH4, Aldrich, 98%),
diphenylphosphoryl azide (DPPA, Aldrich, 97%), 1,8-
diazabicylco[5.4.0]undec-7-ene (DBU, Aldrich, 98%), tetrabutylam-
monium fluoride hydrate (TBAF, Aldrich, 1.0 M in THF), 18-crown-6
ether (18-C-6, Aldrich, 99%), 2-bromoethanol (Aldrich, 98%),
imidazole (Aldrich, 99%) N,N′-dicyclohexylcarbodiimide (DCC,
Aldrich, 99%), and N-hydroxysuccinimide (NHS, Aldrich, 98%).
The following solvents were used as received: acetone (Chem-
Supply, AR), chloroform (CHCl3: Labscan, AR grade), dimethyl
sulfoxide (DMSO: Labscan, AR grade), dichloromethane (DCM:
Labscan, AR grade), diethyl ether (Merck, GR grade), ethyl acetate
(EtOAc: ChemSupply, AR grade), methanol (MeOH: anhydrous,
Lichrosolv, 99.9%, HPLC grade), N,N-dimethylformamide (DMF:
Labscan, AR grade), N,N-dimethylacetamide (DMAc: Aldrich, HPLC
grade), petroleum spirit (BR 40−60 °C, Univar, AR grade),
tetrahydrofuran (THF: Lichrosolv, HPLC grade), and toluene
(TOL, Univar, AR grade).
The following initiators, ligands, and metals for the various
polymerizations are given below and used as received unless otherwise
stated: N,N,N′,N″,N″-pentamethyldiethylenetriamine (PMDETA: Al-
drich, 99%), copper(II) bromide (Cu(II)Br2: Aldrich, 99%), copper(I)
bromide, and Cu(II)Br2/PMDETA complex were synthesized in our
group.
Analytical Methodologies. Size Exclusion Chromatography (RI-
SEC). All polymer samples were dried prior to analysis in a vacuum
oven for 2 days at 25 °C. The dried polymer was dissolved in
tetrahydrofuran (THF) to a concentration of 1 mg mL−1 and then
filtered through a 0.45 μm PTFE syringe filter. The molecular weight
distribution of the polymers was determined through separation on a
Waters 2695 separations module, fitted with a Waters 410 refractive
index (RI) detector maintained at 35 °C, a Waters 996 photodiode
array detector, and two Ultrastyragel linear columns (7.8 × 300 mm)
arranged in series. These columns were maintained at 40 °C for all
analyses and are capable of separating polymers in the molecular
weight range of 500 to 4 million g mol−1 with high resolution. All
samples were eluted at a flow rate of 1.0 mL min−1. Calibration was
Diffusion-Ordered Spectroscopy (DOSY) NMR. 1D DOSY experi-
ments were run to suppress the solvent and organic peroxide (from
1
THF) impurities, which appeared in H NMR spectra of all polymers
(including starting building blocks and dendrimers). A gradient
strength (gpz6) of 85% (for starting building blocks) and 90% (for
dendrimers), gradient pulse length (p30, little delta, δ = p30 × 2) 2 ms
(for starting building blocks) and 2.5 ms (for dendrimers), and
relaxation delay (d1) 5 s (≥5T1) with 256−512 scans were the
parameters used in the 1D DOSY experiment.
2D DOSY experiments were carried out to determine the diffusion
coefficients (D) for dendrimers (14, 15, and 16) in CDCl3. All 2D
DOSY experiments were conducted at 298 K at a dendrimer
concentration of 10 mg mL−1 in CDCl3 using a Bruker Avance
DRX 500 spectrometer operating at 500.13 MHz for protons and
equipped with a 5 mm triple-resonance (1H, 13C, 15N) z-gradient
probe equipped with actively shielded gradients. The z-gradient was
calibrated at 298 K with a HDO sample containing 0.1 mg mL−1
GdCl3. The maximum z-gradient amplitude was 50 G cm−1. A 90°
pulse calibration was performed for each new sample. A bipolar pulse
longitudinal eddy current delay (BPPLED) pulse sequence was used.
The pulse sequences included a 5 ms delay to allow residual eddy
currents to decay. Sine-shaped gradient pulses were utilized to further
minimize eddy currents. The gradient pulse length (p30, little delta, δ
= p30 × 2) 2.5 ms was chosen for diffusion time in order to obtain the
minimum residual signal for each component at maximum gradient
strength. The diffusion delay (Δ) was set to 200 ms. The pulse
gradients were incremented from 2 to 95% of the maximum gradient
strength in a linear ramp (16 steps). A spectral window of 6000 Hz
was accumulated in an acquisition time of 1.38 s. The relaxation time
T1 was determined by inversion recovery method. A relaxation delay of
5T1 of the slowest relaxing signal was used (5 s). The FIDs were
collected into 16K data points; 128 scans and 4 dummy scans were
acquired on each sample. Following acquisition the FIDs were Fourier
transformed applying zero-filling to 16K data points and an
exponential window function with line broadening factor 1−5 Hz.
performed using narrow molecular weight PSTY standards (PDIRI
≤
1.1) ranging from 500 to 2 million g mol−1. Data acquisition was
performed using Empower software, and molecular weights were
calculated relative to polystyrene standards.
C
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
a
Scheme 2. Synthetic Route for Building Block 5 (4-Arm PSTY-(NO•)2)
a
(i): THF, TEA, 0 °C−RT, 6 h. (ii) CuBr, PMDETA, Cu(II)Br2/PMDETA, anisole, 80 °C, 9 h. (iii) DMF, NaN3, 25 °C, 24 h. (iv) CuBr,
PMDETA, toluene, 25 °C, 1 h.
Data were processed using Bruker XWIN NMR software. The signal
decay due to gradients was fitted using
were mixed in an Eppendorf tube, vortexed, and centrifuged. 1 μL of
solution was placed on the target plate spot, the solvent evaporated at
ambient condition, and the measurement run. For lysine dendron 13,
Na(CF3COO) (1 mg mL−1 in THF) was used as cation source.
Synthesis of 4-Arm ATRP Initiator, 1. A solution containing 2-
bromoisobutyryl bromide (25.1 g, 0.109 mol) dissolved in 80 mL of
dry THF was added dropwise to another solution containing
pentaerythritol (3.0 g, 2.2 × 10−2 mol) and triethylamine (11.1 g,
0.109 mol) dissolved in 220 mL of dry THF at 0 °C. The reaction was
allowed to stir for 6 h and filtered to remove the salts, and then the
filtrate concentrated by rotary evaporation. The product was dissolved
in 300 mL of ether and sequentially washed with 10 wt % HCl,
saturated NaHCO3 solution, and brine. The organic layer was
collected, dried over anhydrous MgSO4, and filtered. The filtrate
was concentrated by rotary evaporation. The concentrate was purified
by silica gel column chromatography using ethyl acetate/petroleum
spirit (1/6, v/v) as the eluent. Product 1 was obtained as white crystals
I = I0 exp(−Dγ2g2δ2(Δ − δ/3)
(1)
where I is the resonance intensity measured for a given gradient
strength, g, I0 is the signal intensity with no gradient applied, γ is the
gyromagnetic ratio, δ is the gradient duration (p30 × 2), and Δ is the
diffusion delay. The resulting diffusion coefficients (D) of the polymer
signals and the solvent are the result of the fitting procedure (see
Supporting Information).
The hydrodynamic diameter (Dh,NMR) was determined using the
Stokes−Einstein equation
kT
3πηD
Dh = 2Rh =
(2)
where k is the Boltzmann constant (1.380 × 10−23 J K−1), T is the
temperature in kelvin (298 K), η is the viscosity of the solvent in pascal
seconds (5.3 × 10−4 Pa s for CDCl3), and D is the diffusion coefficient
obtained from 2D DOSY experiment.44
Attenuated Total Reflectance−Fourier Transform Spectroscopy
(ATR-FTIR). ATR-FTIR spectra were obtained using a horizontal,
single bounce, diamond ATR accessory on a Nicolet Nexus 870 FT-IR.
Spectra were recorded between 4000 and 500 cm−1 for 32 scans at 4
cm−1 resolution with an OPD velocity of 0.6289 cm s−1. Solids were
pressed directly onto the diamond internal reflection element of the
ATR without further sample preparation.
Matrix-Assisted Laser Desorption Ionization−Time-of-Flight
(MALDI-ToF) Mass Spectrometry. MALDI-ToF MS spectra were
obtained using a Bruker MALDI-ToF autoflex III smart beam
equipped with a nitrogen laser (337 nm, 200 Hz maximum firing
rate) with a mass range of 600−400 000 Da. Spectra were recorded in
either reflectron mode (1500−4500 Da) or linear mode (4000−
20 000 Da). trans-2-[3-(4-tert-Butylphenyl)-2-methylpropenylidene]-
malononitrile (DCTB; 20 mg mL−1 in THF) was used as the matrix
and Ag(CF3COO) (1 mg mL−1 in THF) as the cation source for all
the polystyrene samples. 20 μL of polymer solution (1 mg mL−1 in
THF), 20 μL of DCTB solution, and 2 μL of Ag(CF3COO) solution
1
(10.65 g, 66.0%). Rf (1/6 EtOAc/petroleum spirit) 0.61. H NMR
(CDCl3, 298 K, 400 MHz): δ 1.92 (s, 24H, CH3−), 4.30 (s, 8H,
−OCH2C−). 13C NMR (CDCl3, 298 K, 400 MHz): 30.62, 43.64,
55.18, 62.88, 170.89.
Synthesis of Building Block 2 (4-Arm PSTY-Br). Freshly
purified styrene (9.02 g, 8.66 × 10−2 mol), PMDETA (0.078 mL, 3.76
× 10−4 mol), 1 (0.254 g, 3.76 × 10−4 mol), anisole (10 mL), and
Cu(II)Br2/PMDETA (0.037 g, 9.41 × 10−5 mol) were added to a 20
mL Schlenk tube equipped with a magnetic stirrer and purged with
argon for 20 min. Cu(I)Br (0.054 g, 3.77 × 10−4 mol) was carefully
added under positive argon flow, and the reaction mixture was purged
with argon for a further 5 min. The flask was placed in a temperature-
controlled oil bath at 80 °C for 9 h. The reaction was stopped by
quenching in ice and exposure to air. The polymerization mixture was
diluted with DCM, and the copper salts were removed by passage
through an activated basic alumina column. The solution was
concentrated by rotary evaporation, and the polymer was recovered
by precipitation into methanol, filtered, and dried for 48 h under high
vacuum at 25 °C. The resultant polymer was precipitated from DCM
into MeOH, filtered, and dried under high vacuum for 48 h at 25 °C to
D
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
a
Scheme 3. Synthetic Route for Building Block 8
a
(i) NaI, NaN3, DMSO/H2O, 80 °C, 2 days. (ii) TEA, THF, 0 °C−RT, 6 h. (iii) CuBr, PMDETA, Cu(II)Br2/PMDETA, bulk, 80 °C, 7.5 h.
a
Scheme 4. Synthetic Route for Building Block 10
a
(i) Imidazole, THF, 0 °C−RT, 12 h. (ii) 18-C-6, K2CO3, acetone, reflux, 48 h. (iii) LiAlH4, THF, 0 °C−RT, 16 h. (iv) DPPA, DBU, toluene, 0 °C−
RT, dark, 16 h. (v) TBAF, THF, argon, RT, 12 h. (vi) TEA, THF, 0 °C−RT, 12 h. (vii) CuBr, PMDETA, Cu(II)Br2/PMDETA, bulk, 80 °C, 220
min.
give a white polymer, which was further purified by preparative SEC.
The resultant polymer was reprecipitated from THF into MeOH,
filtered, dried under high vacuum for 48 h at 25 °C, and obtained as
added to the solution under a positive flow of argon, sealed, and stirred
for 1 h at 25 °C. The polymer was diluted with DCM, and the copper
salts were removed by passage through an activated basic alumina
column. The solution was concentrated by rotary evaporation, and the
polymer was recovered by two precipitations (once from toluene and
once from DCM) into methanol, collected by vacuum filtration, and
washed extensively with methanol. The polymer was dried under high
vacuum for 48 h at 25 °C to give a peach-colored polymer. The
resultant polymer was reprecipitated from THF into MeOH, filtered,
and dried under high vacuum for 48 h at 25 °C and obtained as peach
powder (Mn,RI = 10 200, Mp,RI = 10 320, PDIRI = 1.04, Mn,TD = 11 600,
Mp,TD = 11 760, PDITD = 1.01, Mn,NMR = 12 640).
peach powder (Mn,RI = 8670, Mp,RI = 8710, PDIRI = 1.05, Mn,TD
=
9700, Mp,TD = 9840, PDITD = 1.02, Mn,NMR =10 820, monomer
conversion 39.0% as determined by gravimetric method).
Synthesis of Building Block 3 (4-Arm PSTY-N3). NaN3 (0.34 g,
5.25 × 10−3 mol) was added to a stirred solution of 2 (1.16 g, 1.31 ×
10−4 mol) in 6 mL of DMF. The reaction mixture was stirred for 24 h
at room temperature. The polymer was twice precipitated into
methanol (once from DMF, once from DCM), recovered by vacuum
filtration and washed extensively with water and methanol. The
polymer was dried under high vacuum for 48 h at 25 °C obtained as
white powder (Mn,RI = 8730, Mp,RI = 8810, PDIRI = 1.05, Mn,TD = 9630,
Mp,TD = 9790, PDITD = 1.02, Mn,NMR = 10 670).
Synthesis of Building Block N3-PSTY-Br 8. The ATRP initiator
7 (Scheme 3) was synthesized according to the literature procedure.42
The synthetic strategies to produce 7 and 8 are given in Supporting
Synthesis of 5 (4-Arm PSTY-(NO•)2). The trifunctional linker, 4,
was synthesized according to the literature procedure.40 In a 20 mL
Schlenk tube, 4 (0.096 g, 1.82 × 10−4 mol) was added to a stirred
solution of 3 (0.35 g, 4.04 × 10−5 mol) and PMDETA (7.0 × 10−3 g,
4.04 × 10−5 mol) in 6 mL of toluene. The reaction mixture was purged
with argon for 15 min. Cu(I)Br (5.7 × 10−3 g, 4.04 × 10−5 mol) was
Information (8, Mn,RI = 2370, Mp,RI = 2490, PDI = 1.08, Mn,NMR
=
2870, monomer conversion 47.6% as determined by gravimetry).
Synthesis of Building Block 10. The ATRP initiator 9 was
synthesized according to the literature procedure.19 The synthetic
strategies to produce 9 and 10 are given in Supporting Information
E
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
a
Scheme 5. Synthesis of Alk-Lysine-Boc (12) and Alk-Lsyine-Boc Dendron (13)
a
(i) NaOH, H2O, 1,4-dioxane, 0 °C−RT, 7 h. (ii) DCC, DMAP, DCM, 0 °C−RT, 24 h. (iii) TFA, DCM, RT, 2 h. (iv) DCM, DCC, NHS, 0 °C−
RT, 2 h. (v) DMF, TEA, RT, 12 h.
(10, Mn,RI = 2600, Mp,RI = 2690, PDI = 1.07, Mn,NMR = 2830, monomer
conversion 46.9% as determined by gravimetry).
56.9%). Rf (100% EtOAc) 0.51. MALDI-ToF MS: [M + Na+] Calcd =
862.52, Found = 862.89. [M + K+] Calcd = 878.50, Found = 878.72.
Synthesis of Dendrimer 14 in a One-Pot Reaction. In a 10 mL
Schlenk tube, 5 (90.0 mg, 7.76 × 10−6 mol), 8 (156 mg, 6.51 × 10−5
mol), 12 (48.0 mg, 1.24 × 10−4 mol), and PMDETA (20 μL, 9.31 ×
10−5 mol) were dissolved in 2 mL of toluene and 2 mL of DMSO. The
reaction mixture was then purged with argon for 30 min. Cu(I)Br
(13.5 mg, 9.31 × 10−5 mol) was added to the solution under a positive
flow of argon, sealed, and stirred for 30 min at 25 °C in a temperature-
controlled oil bath. The polymer was diluted with DCM, and the
copper salts were removed by passage through an activated basic
alumina column. The solution was concentrated by rotary evaporation
and the polymer recovered by precipitation into methanol, collected
by vacuum filtration, and washed extensively with methanol. The
polymer was dried under high vacuum for 48 h at 25 °C to give a white
polymer and further purified by preparative SEC. The resultant
polymer was reprecipitated from DCM into MeOH, filtered, and dried
under high vacuum for 48 h at 25 °C and obtained as white powder
Synthesis of Building Block 12 (Alkyne-Lysine-Boc). The Boc-
Lysine-OH, 11, was synthesized according to ref 43 (Scheme 5). To a
500 mL flask, 11 (6.0 g, 0.172 mol), propargylamine (1.90 g, 0.0345
mol), and DMAP (0.316 g, 2.59 × 10−3 mol) were dissolved in 120
mL of dry DCM and cooled to 0 °C in an ice bath. A mixture of DCC
and 50 mL of DCM was added dropwise into the solution over 30
min. The mixture was allowed to react for 36 h at room temperature.
The solid content was removed by filtration, and the filtrate was
washed by saturated brine (2 × 50 mL). The organic layer was
collected, dried over anhydrous MgSO4, and the solvent removed in
vacuo followed by column chromatography using ethyl acetate/
petroleum spirit (3/1, v/v) as the eluent. Product 12 was obtained as a
white solid (2.86 g, yield = 41.7%). Rf (3/1 EtOAc/petroleum spirit)
0.65. 1H NMR (CDCl3, 298 K, 500 MHz): δ 6.91 (s, 1H,
CHCCH2NHCO−), 5.32 (s, 1H, −CHNHCO−), 4.69 (s, 1H,
−CH2NHCO), 3.97−4.07 (b, 3H, CH2CCH2NH− and
−CH2CHCO−), 3.08 (t, 2H, J = 6.6 Hz, −CH2CH2NH−), 2.20 (t,
1H, J = 2.5 Hz, HCCCH2−), 1.2−2.0 (m, 24H, CH2-Lys and CH3-
Boc). 13C NMR (CDCl3, 298 K, 500 MHz): 22.69, 28.44, 28.53, 29.09,
29.70, 32.23, 40.07, 54.24, 71.60, 79.16, 79.53, 80.10, 155.97, 156.26,
172.22.
Synthesis of Alkyne-Lysine-Boc Dendron (13). 11 (2.25 g, 6.51
× 10−3 mol), NHS (1.5 g, 0.013 mol), and DCC (2.63 g, 0.013 mol)
were dissolved in 50 mL of DCM, which was cooled to 0 °C in an ice
bath. The reaction mixture was heated to room temperature and
stirred for 2 h; the solid was filtered, the filtrate was evaporated, and
the concentrate, 11a, was dissolved in 20 mL of DMF and used
without further purification. 12 (1.12 g, 2.93 × 10−3 mol) was
dissolved in 6 mL of TFA/DCM mixture (1:1, v/v), stirred for 2 h,
and the solvent evaporated. The concentrate, 12a, was dissolved in 20
mL of DMF and then added dropwise into the 20 mL DMF/11a
mixture above over 20 min. The mixture was stirred for another 12 h.
The reaction mixture was then concentrated. The concentrate was
dissolved in 200 mL of DCM and washed sequentially with 10 wt %
HCl solution, saturated NaHCO3 solution, and brine. The organic
layer was collected, dried over anhydrous MgSO4, and filtered. The
filtrate was concentrated by rotary evaporation. The concentrate was
purified by silica gel column chromatography using 100% EtOAc as
the eluent. Product 13 was obtained as white solid (1.4 g, yield =
(Mn,RI = 26 590, Mp, RI = 27 650, PDI = 1.08; Mn,TD = 35 120, Mp,TD
35 990, PDI = 1.05, Mn,NMR = 39 340).
=
Synthesis of Dendrimer 15 in a One-Pot Reaction. In a 10 mL
Schlenk tube, 5 (90.0 mg, 7.76 × 10−6 mol), 10 (171 mg, 6.51 × 10−5
mol), 12 (71.9 mg, 1.86 × 10−4 mol), and PMDETA (26 μL, 1.24 ×
10−4 mol) were dissolved in 1 mL of toluene. The reaction mixture
was then purged with argon for 15 min. Cu(I)Br (18 mg, 1.24 × 10−4
mol) was then added to the solution under a positive flow of argon,
sealed, and stirred for 30 min at 25 °C in a temperature-controlled oil
bath. The polymer was diluted with DCM, and the copper salts were
removed by passage through an activated basic alumina column. The
solution was concentrated by rotary evaporation, and the polymer was
recovered by precipitation into methanol, collected by vacuum
filtration, and washed extensively with methanol. The polymer was
dried under high vacuum for 48 h at 25 °C to give a white polymer and
further purified by preparative SEC. The resultant polymer was then
reprecipitated from DCM into MeOH, filtered, dried under high
vacuum for 48 h at 25 °C, and obtained as white powder (Mn,RI
=
29 140, Mp,RI = 30 520, PDI = 1.08; Mn,TD = 37 540, Mp,TD = 38 590,
PDI = 1.05, Mn,NMR = 43 990).
Synthesis of Dendrimer 16 in a One-Pot Reaction. In a 10 mL
Schlenk tube, 5 (90.0 mg, 7.76 × 10−6 mol), 10 (171 mg, 6.51 × 10−5
mol), 13 (160 mg, 1.86 × 10−4 mol), and PMDETA (26 μL, 1.24 ×
10−4 mol) were dissolved in 1 mL of toluene. The reaction mixture
F
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
was then purged with argon for 15 min. Cu(I)Br (18 mg, 1.24 × 10−4
mol) was then added to the solution under a positive flow of argon,
sealed, and stirred for 30 min at 25 °C in a temperature-controlled oil
bath. The polymer was diluted with DCM, and the copper salts were
removed by passage through an activated basic alumina column. The
solution was concentrated by rotary evaporation, and the polymer was
recovered by precipitation into methanol, collected by vacuum
filtration, and washed extensively with methanol. The polymer was
dried under high vacuum for 48 h at 25 °C to give a white solid and
further purified by preparative SEC. The resultant polymer was then
reprecipitated from DCM into MeOH, filtered, and dried under high
vacuum for 48 h at 25 °C obtained as a white powder (Mn,RI = 31 200,
Mp,RI = 33 360, PDI = 1.08; Mn,TD = 45 640, Mp,TD = 46 330, PDI =
1.03, Mn,NMR = 53 540).
Deprotection of Boc Groups from Boc-Lysine Units of
Dendrimer. Typically, dendrimer (30 mg) was dissolved in a mixture
of DCM (1 mL) and TFA (1 mL). The solution was kept stirring for
12 h at room temperature, the solvent was removed by evaporation,
and the residual content was dried in high vacuum for 24 h. The
resulting product was used for further micellization in water directly.
Micellization of Dendrimers. Typically, 5 mg of dendrimer (after
deprotection of the Boc groups) was dissolved in 3 mL of DMF (a
common solvent for both PSTY and Lysine units) followed by the
gradual addition (at a rate of 0.013 mL min−1) of 3 mL of water (a
nonsolvent for the hydrophobic PSTY blocks). The resulting mixture
of DMF and water was transferred to presoaked and rinsed dialysis bag
(Pierce Snakeskin, MWCO 3.5K) and dialyzed against a large volume
of Milli-Q water for 2 days to remove the organic solvent. After
dialysis, additional water was then added to the water solution to make
up 10 mL. This ensured that the final concentration of the dendrimers
was 0.5 mg mL−1. It should be noted that a small amount of
precipitant was observed after the self-assembly of 17, thereby slightly
reducing its concentration.
Dynamic Light Scattering (DLS). Dynamic light scattering
measurements were performed using a Malvern Zetasizer Nano Series
running DTS software and operating a 4 mW He−Ne laser at 633 nm.
Analysis was performed at an angle of 173° and a constant
temperature of 25 °C. The hydrodynamic diameter of dendrimer
before the removal of Boc groups from Boc-Lysine units was measured
in CDCl3 (5 mg mL−1). The hydrodynamic diameter and zeta
potential of dendrimer after the removal of Boc groups was measured
in water (0.5 mg mL−1, micelle solution). All the samples were filtered
by 0.45 μm filter before measurements. The number-average
hydrodynamic particle size, polydispersity index, and zeta potential
are reported. The polydispersity index (PDI) was used to describe the
width of the particle size distribution. It was calculated from a
Cumulants analysis of the DLS measured intensity autocorrelation
function and is related to the standard deviation of the hypothetical
Gaussian distribution (i.e., PDIPSD = σ2/ZD2, where σ is the standard
deviation and ZD is the Z average mean size).
σ2 = ln(PDI)
(5)
where eq 1 is the Gaussian distribution function of w(M) (the weight
distribution of the SEC trace), Mn is the number-average molecular
weight, Mw is the weight-average molecular weight, and the
polydispersity PDI = Mw/Mn.
RESULTS AND DISCUSSION
■
Synthesis of Building Blocks. The synthesis of a 4-arm
polystyrene star (2) by ATRP using a tetrafunctional bromine
initiator, 1 (see synthesis and characterization in Supporting
Information and 1H NMR in Figure S1), is given in Scheme 2.
The produced 4-arm star exhibited a narrow molecular weight
distribution (MWD) with an Mn,RI of 9140 and polydispersity
index (PDI) of 1.09. The SEC trace (i.e., weight distribution) of
2 (crude, curve a) given in Figure 1 showed a high molecular
Figure 1. Molecular weight distributions (MWDs) for starting
polymer (2 and 3) and product (5) obtained from SEC-RI based
on a polystyrene calibration curve: (a) 4-arm PSTY-Br (2 crude), (b)
4-arm PSTY-Br (2 after prep), (c) 4-arm PSTY-N3 (3), (d) 4-arm
PSTY-(NO.)2 (5). Dotted line represents the LND fit to 5. All SEC
traces were normalized to weight.
weight tail most likely the result of star−star coupling, which
was removed by preparative SEC (i.e., 2 after prep, curve b).
The purity of the 4-arm star (2 crude) relative to the higher
molecular weight species formed through star−star radical
coupling was determined by simulating the weight distributions
from SEC based on a log-normal distribution (model) using a
Gaussian function (Figure S2 and Table S1 in Supporting
Information). The purity of the crude 4-arm star was 87.3%,
which increased to 96.1% after purification by preparative SEC
(Table 1). In addition, the chain-end functionality of 2 after
preparative SEC was 98% determined by 1H NMR (Figure 2A),
and from the MALDI-ToF (Figure 3A) the only species
detected was the chain-end fragmentation with added Ag+ (e.g.,
m/z found was 9266.20, which was close to that calculated
9266.37). These results suggested that this 4-arm star would
provide a suitable core building block to produce the final
dendrimers in high yields.
Azidation of the bromine groups on 2 (i.e., after purification)
to produce 3 showed a distinct shift of the proton adjacent to
the azide from ∼4.5 to 4 ppm (Figure 2), a slight increase in
the Mn,RI value (Table 1), and the near-quantitative conversion
further confirmed by analysis of the MALDI (Figure 3B)
spectrum. The SEC trace of 3 (i.e., 2 after azidation) was nearly
identical to that of 2 (after prep) as shown in Figure 1. The
trifunctional linker, 4, consisted of an alkyne and two nitroxide
moiteies (see Scheme 2), with its characterization given in the
Supporting Information (Figures S4−S6). Coupling 3 and 4
using the CuAAC reaction resulted in a slight shift of the MWD
Transmission Electron Microscopy (TEM). The samples for
TEM analysis were prepared by placing a drop of the micelle solution
(0.5 mg mL−1) of dendrimer onto a Formvar-precoated copper TEM
support grid and allowed to air-dry before measurement. The micelles
were characterized on a Jeol-1010 instrument utilizing an accelerating
voltage of 80 kV at ambient temperature.
LND Simulations. We used a log-normal distribution (LND)
model based on a Gaussian function to fit the experimental MWD.
One can simulate the molecular weight distributions, and in particular
the weight distribution,45 with a log-normal distribution (see ref 46 for
more details) using the following equations:
exp(−(ln M − ln M)2/2σ2
̅
w(M) =
M(2πσ2)0.5
(3)
where
M = (M M )0.5
̅
(4)
n
w
and
G
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Table 1. SEC and NMR Data for All Building Blocks
a
a
b
b
f
c
d
i
polymer
Mn,RI
Mp,RI
PDI
Mn,TD
Mp,TD
PDI
Mn,theo
ΔHDV
Mn,NMR
chain-end functionality (%)
purity
2
9140
8670
8730
10200
2500
2370
2600
8730
8710
8810
10310
2570
2490
2690
1.09
1.05
1.05
1.04
1.11
1.08
1.07
87.3
96.1
94.9
96.0
95.4
99.0
97.9
e
2
9700
9630
9840
9790
1.02
1.02
1.01
0.89
0.90
0.88
10820
10670
12640
98.0
98.0
96.0
g
3
5
8
9550
g
11600
11760
11660
e
h
h
8
2480
2820
2870
2830
99.0
10
96.0
a
b
c
SEC (RI detector) was based on a PSTY calibration curve. MWD determined from DMAc triple detection SEC. Hydrodynamic volume change
d
e
(ΔHDV = Mp,RI/Mp,TD). Chain-end functionality calculated from 1H NMR. Polymers fractionated (purified) by preparative SEC.
f
g
Underestimation of true PDI value as light scattering has less sensitivity to low molecular weights. Mn,theo determined from Mn,TD of 2 (after
h
prep) plus the Mw of functional groups. Mn,theo determined by Mn,theo = ([M]/[I])·Mm·con% + Mi, where [M], [I], Mm, con%, and Mi are the
monomer concentration, initiator concentration, molecular weight of monomer, monomer conversion, and molecular weight of initiator,
respectively. Determined from log-normal distributions(LND) simulation.
i
to the expected molecular weight to produce 5 (Scheme 1).
The 4-arm star 5 consisted of eight free nitroxides on its
periphery. The MALDI-ToF given in Figure 3C showed that
the peaks corresponded to the expected nitroxide end-
functionalized 5 (e.g., m/z found was 12 593.38, which was
close to 12 593.20 calculated with Ag+), in agreement with the
MALDI analysis of the previously synthesized polymers with
NRC and CuAAC produced third- or fourth-generational layer
dendrimers coated on the periphery with L-lysine in one pot at
25 °C (Scheme 1). The NRC process occurs through the
reaction of a nitroxide and a free radical to form an
alkoxyamine.47,48 A method to produce radicals rapidly in situ
employed the abstraction of the halide from the polymer end
group with either Cu(I) or Cu(0).47 It was found that in
DMSO Cu(I)Br/Me6TREN was highly reactive, rapidly
producing radicals that were subsequently trapped by nitroxides
to form the resultant alkoxyamines with little or no radical−
radical coupling product.48 As a solvent, DMSO has also been
demonstrated to induce disproportionation of Cu(I)Br/
Me6TREN to the highly active Cu(0) and Cu(II) species.49
Recent results showed that in DMSO, the NRC reaction was
rapid and complete within less than 2 min.19,40 By changing the
ligand to PMDETA and the solvent to toluene, the NRC
reaction slowed significantly with coupling times greater than 2
h. Copper also catalyzes the CuAAC reaction. The coupling
reaction using Cu(I)Br/PMDETA in toluene, in contrast to the
NRC reaction, was rapid with complete coupling in less than 2
min. Conversely, using Cu(I)Br/Me6TREN in DMSO slowed
the CuAAC reaction with complete coupling in more than 30
min. However, using Cu(I)Br/PMDETA in a mixture of
toluene and DMSO (50:50 v/v %) allowed the NRC and
CuAAC reactions to proceed with similar rates of coupling.
This latter method should produce the dendrimer much faster
than in the other two solvent/ligand conditions.50
Polymeric dendrimer 14 formed through the orthogonal
NRC and CuAAC “click” reactions of 5, 8, and 12 in a molar
ratio of 1:8.4:16 (Scheme 1), respectively, utilizing an excess of
reactants 8 and 12 to the core 5. The parallel process using
CuBr/PMDETA in a solvent mixture of toluene and DMSO
(50:50 v/v %) produced 14 (crude) in 30 min at 25 °C. The
SEC trace in Figure 5A showed that all core 5 (blue dotted
line) was consumed and that the MWD contained the
remaining excess of 8 (green dotted line) and 12 (orange
dotted line). We determined a purity of 79 wt % from a fit of
the theoretical MWD of 14, denoted as “LND (14, pure)” in
Figure 5A by the black dotted line calculated from the addition
of molecular weight of all arms to the core using a change in
hydrodynamic volume (ΔHDV) of 0.77 in the LND method.
Polymeric dendrimer 14 was then purified by preparative SEC
to remove all starting reagents and other low molecular weight
coupled products. After preparative SEC (see Figure 5A), the
purity of 14 (after prep) determined by the LND method was
94% based on the fit of the theoretical MWD of 14. The 6% of
nitroxide chain-end functionality.19 The H NMR of the free
1
nitroxides on 5 led to line broadening and poor resolution,
whereas converting these nitroxides to hydroxylamines allowed
1
us to obtain accurate NMR spectra. The H NMR of 5, in
which the nitroxides have been converted to hydroxylamines,
showed near complete loss of the proton associated with the
azide and presence of peaks at 5.1, 5.3, 7.8, and 8.2 ppm
associated with peaks d, e, f, and g from the coupling reaction
to produce 5 (Figure 2C). It was also found that the absolute
Mn,TD (11 600) was close to that calculated (11 660) from the
addition of all the end-groups to the 4-arm star (Table 1).
The ATRP of initiator 7 with styrene (Scheme 3) gave
telechelic polymer 8 (crude) with approximately 4.6 wt % of
double molecular weight polymer formed through radical−
radical coupling (see Figure S10 and Table S2 in Supporting
Information). After purification by preparative SEC, its purity
increased to 99.0% (Figure S10 and Table S2) and chain-end
1
functionality was close to 99% as determined by H NMR
(Figure 4A) from the near 4:1 ratio of proton e to that of a and
c. Telechelic polymer 10 (Scheme 4), with a diazide
functionality on one chain-end and a bromine on the other,
was formed again using ATRP. As determined by the LND
simulations, it showed greater than 97.9% purity (Figure S13
and Table S3) and approximately 96% chain-end functionality
1
by H NMR (Figure 4B) as given in Table 1, and thus 10 was
used without further purification by preparative SEC. The high
purity of both 8 and 10 was further supported by the MALDI
analysis in Figures S11 and S14 in the Supporting Information.
The other two building blocks 12 and 13 (see Scheme 1)
consisted of Boc protected L-lysine groups. The synthetic
strategy to make 12 and 13 is shown in Scheme 5. The
amidation of 11 and propagylamine afforded 12 in a good yield
of 41.7%. The Boc protected L-lysine dendron 13 formed
through the reaction of 11a and 12a to give a yield of 56.9%.
Characterization of 12 and 13 is given in Supporting
Information (Figures S16−S22).
One-Pot Synthesis of Boc Protected L-Lysine Deco-
rated Polymeric Dendrimers. Modulating the catalytic
activity of copper for the orthogonal “click”-type reactions of
H
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
1
Figure 2. Comparison of H 1D DOSY NMR spectra of (A) 4-arm PSTY-Br (2), (B) 4-arm PSTY-N3 (3), and (C) 4-arm PSTY-(NO•)2 (5).
Recorded in CDCl3, 298 K, 500 MHz, gradient strength (gpz6) 85%, gradient pulse length (p30) 2.0 ms. *Residual phenylhydrazine.
byproducts most probably results from the nonfully coupled L-
lysine−alkyne to the polymeric dendrimer. The Mn,RI of 14
(prep) was 26 590 with a PDI of 1.08 using the refractive index
(RI) detector, and the Mn,TD was 35 120 with a PDI of 1.05
using triple detection (i.e., absolute molecular weight
determination) as given in Table 2. The value of Mn,TD
determined by triple detection was close to the theoretical
value (33 640) of the dendrimer calculated from the addition of
the absolute molecular weights of the building blocks (5, 8, and
12). It should be noted that the PDI from triple detection SEC
was always lower due to a lower sensitivity of the light
scattering at lower molecular weights. The change in ΔHDV
from Mp,RI to Mp,TD of 0.77 was the same as that used in the
LND method above, supporting the calculated high purity of
94%. The lower apparent molecular weight (i.e., ΔHDV of
0.77) suggested that the polymeric dendrimer was much more
compact than its corresponding linear analogue as a result of
the many tethering links in each generational layer. Each link
will reduce the amount of solvent required to swell the polymer
arms.51 Further support for the high purity comes from analysis
of the 1H NMR (see Figure 6A and Figure S23) from
integration of peaks corresponding to the three building blocks.
Table 2 shows that the ratio of the number of arms in each
generational layer was 4:8.2:8 and close to theory (4:8:8).
The next polymeric dendrimer synthesized was 15 with
double the number of L-lysine on the periphery (Scheme 1).
I
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Figure 3. MALDI-TOF mass spectra of (A) 4-arm PSTY-Br (2), (B) 4-arm PSTY-N3 (3), and (C) 4-arm PSTY-(NO•)2 (5). The spectra were
recorded in linear mode using DCTB as the matrix and Ag(CF3COO) as the cation source: (i) full spectra and (ii) expanded spectra.
The molar ratio of starting reagents 5, 10, and 12 was 1:8.4:24
and using the same reaction conditions as those to produce 14.
The SEC trace showed that 15 was produced (Figure 5B) with
near complete loss of starting polymer 5 and remaining excess
J
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Figure 4. 1H 1D DOSY NMR spectra of (A) N3-PSTY-Br (8) and (B)
(N3)2-PSTY-Br (10) recorded in CDCl3 at 298 K, 500 MHz, gradient
strength (gpz6) 85%, gradient pulse length (p30) 2.0 ms; the sample 8
was purified by preparative-SEC.
Figure 5. SEC traces of starting building blocks (5, 8, 10, 12, 13;
dotted lines) to produce the respective dendrimers. Dendrimers before
(crude, dark blue solid line) and after purification by preparative SEC
(after prep, red solid line). All SEC traces were determined by THF
SEC (RI). (A) Dendrimer 14 and LND simulation of pure
(theoretical, dotted lines) 14. (B) Dendrimer 15 and LND simulation
of pure (theoretical, dotted lines) 15. (C) Dendrimer 16 and LND
simulation of pure (theoretical, dotted lines) 16.
of reactants 10 and 12. The SEC trace showed products
corresponding to the coupling of 10 and 12 through the
CuAAC reaction. The purity of 15 (crude) was 74%
determined by the LND method using the theoretical
molecular weight of 15 and a ΔHDV of 0.76 (Figure 5B).
After fractionation of 15 (crude) by preparative SEC, the purity
increased to 96%. The Mn,RI and PDI were 29 140 and 1.07, and
the Mn,TD and PDI by triple detection was 37 540 and 1.05,
respectively (Table 2). The value of Mn,TD (38 590) was close
to the theoretical value (38 530) of the dendrimer calculated
from the absolute molecular weight of the building blocks (5,
10, and 12). In addition, the ΔHDV determined from the ratio
of Mp,RI to Mp,TD was 0.79, which was close to that used in the
produce a fourth-generation layered (G4) polymeric dendrimer
16. The molar ratio of starting reagents 5, 10, and 13 was
1:8.4:24, and using the same reaction conditions as for the
previously made 14 and 15. The MWD after the reaction
showed a loss of reactants 5 and 10, with new MWD peaks
appearing that corresponded to the coupling of 10 and 13 and
the final polymeric dendrimer 16. There was still a large peak
corresponding to the starting dendron 13, which was used in
large excess. The purity of the crude product 16 was lower
(69%) compared to both 14 and 15 using a ΔHDV of 0.70 in
the LND simulation (Figure 5C). Purification using preparative
SEC gave an Mn,RI of 31 200 (PDI = 1.08) and Mn,TD of 45 640
(PDI = 1.03), in which the value of Mn,TD (45 640) was close to
the theoretical one (46 420) of the dendrimer calculated from
the absolute molecular weight of the building blocks (5, 10, and
13). The purity after preparative SEC increased to 97%. In
addition, the ΔHDV calculated from the ratio of Mp,RI to Mp,TD
was 0.72, which was close to the value used in the LND
1
LND method above for 15 (crude). Analysis by H NMR
(Figure 6B and Figure S24) gave the number of arms in each
generational layer as equal to 4:8.1:16, which was very close to
expected for a polymeric dendrimer to form 15.
One of the main goals of this work was to create a dendrimer
with a densely coated L-lysine periphery (i.e., with 32 lysine
groups on the periphery and 16 lysine groups as the
penultimate generational layer) as shown in Scheme 1. To do
this, we produced a small molecule L-lysine dendron 13 with an
alkyne end-functionality that was then coupled to 10 and 5 to
K
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
a
Table 2. SEC, LND, and NMR Data for Dendrimers
h
ΔHDV
number of arms
b
b
c
c
d
e
f
g
i
dendrimer
reactants
Mn,RI
Mp,RI
PDI
Mn,TD
Mp,TD
PDI
Mn,theo
SEC
LND
Mn,NMR G1
G2
G3
purity (%)
14
15
16
5, 8, 12
26590
29140
31200
27650
30520
33360
1.08
1.07
1.08
35120
37540
45640
35990
38590
46330
1.05
1.05
33640
38530
46420
0.77
0.79
0.72
0.77
0.76
39340
43990
53540
4
4
4
8.20
8.12
8.14
8
94
96
97
5, 10, 12
5, 10, 13
16
j
1.03
0.70
16
a
b
c
All the dendrimer products fractionated (purified) using preparative SEC. MWD from SEC (RI). MWD determined by DMAc triple detection
d
e
SEC. Underestimation of true PDI value as light scattering has less sensitivity to low molecular weights. Mn,theo calculated from Mn of starting
f
g
materials (Table 1). Hydrodynamic volume change (ΔHDV) determined from peak molecular weight (ΔHDV = Mp,RI/Mp,TD). ΔHDV used in
h
i
log-normal distributions(LND) simulation. Determined from integration of each generation by 1H NMR. Determined from log-normal
j
distributions(LND) simulation. Represents 16 units of dendron 13, which includes 16 and 32 L-lysine groups in the third- and fourth-generation
layers, respectively.
Figure 6. 1H 1D DOSY NMR spectra of dendrimers (A) 14, (B) 15, and (C) 16, recorded in CDCl3 at 298 K, 500 MHz; gradient strength (gpz6)
90%, gradient pulse length (p30) 2.5 ms.
1
simulation (Table 2). This lower ΔHDV value suggested that
16 was more compact due to the higher peripheral lysine
branching compared to either 14 or 15. Analysis by H NMR
further showed that the ratio of the number of arms in each
L
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
a
Table 3. Self-Assembly of Dendrimers in Water and Organic Solvents
THF
CDCl3
water
b
c
d
e
dendrimer Dh,SEC (nm) diffusion coefficient (m2 s−1
)
Dh,NMR (nm) Dh,DLS (nm) PDIDLS Dh,TEM (nm) Dh,DLS (nm) PDIDLS zeta potential (mV)
14
15
16
17
18
19
8.41
8.86
9.21
1.38 × 10−10
1.03 × 10−10
9.06 × 10−11
5.97
8.00
9.06
5.80
8.44
0.168
0.164
0.228
10.44
10.8
13.0
13.5
8.38
9.83
0.374
0.447
0.673
27.5
34.1
40.4
11.07
a
b
The diameters were measured from three methods: (i) SEC, (ii) DOSY NMR, and (iii) DLS. Determined from eq 6 using the Mark−Houwink
c
parameters (K = 0.0141 cm3 g−1, a = 0.7 in a good solvent). Diffusion coefficient (D) determined by 2D DOSY NMR in CDCl3 at 298 K.
d
e
Determined from eq 2. Determined by averaging the size of over 50 particles in the TEM micrograph.
generational layer (i.e., 4:8.1:16:32) was close to theory (Figure
6C and Figure S25). The results taken together strongly
support that all three polymeric dendrimers were produced in
excellent purity. We could not obtain MALDI-ToF of the
dendrimers, presumably due to the high molecular weights of
the dendrimers.
Self-Assembly of Dendrimers in Organic and Aque-
ous Media. We used three different methods to determine the
hydrodynamic diameter (Dh) in organic solvents. The first
method used SEC to determine Dh in THF based on
polystyrene standards.46,52 Using the Mark−Houwink relation-
ship between molecular weight and intrinsic viscosity, one can
determine the hydrodynamic radius (Rh) from the relation-
ship52
All samples were filtered before analysis. The Dh,DLS found in
water increased from 8.38 to 11.07 nm for the three dendrimers
17 to 19 (Table 3). A similar trend was found from TEM (see
Figure S32 in Supporting Information), although the TEM
diameters (Dh,TEM) were slightly larger by ∼2.5 nm. The zeta
potential increased from +27.5 to +40.4 mV, as expected, with
the increased number of cationic L-lysine groups on the
periphery. The small size of the dendrimer micelles in water
could be assumed to represent crew-cut unimolecular micelles
or micelles with quite a low aggregation number. This postulate
is in agreement with the low aggregation number found from
the self-assembly of amphiphilic 4-arm block stars (consisting
of an anionic outer block and a hydrophobic core block)53,54
and, more particularly, with the low aggregation number (Z) of
9 found for polymeric dendrimers decorated with anionic
blocks.18 The reason for such low aggregation numbers
compared to linear diblock copolymers (where Z ranges from
150 to 300)53 is ascribed to junction points55 and loops.54 The
PSTY core of our dendrimers consists of a 4-arm star tethered
to a second-generational layer of linear PSTY with a cationic
peripheral layer (i.e., L-lysine groups). The many junction
points between PSTY building blocks results in stretching of
the PSTY chains51 in the core, which further stretches due to
the cationic peripheral groups. This stretching together with the
formation of loops leads to an increase in the free energy, which
can only be reduced by significantly lowering the aggregation
number. Therefore, the data for our dendrimers suggests that
the aggregation number is very low and consisting of
unimolecular micelles (i.e., where Z =1) and that the L-lysine
groups on one side of the dendrimer must be as far from the L-
lysine groups on the other side of the dendrimer to reduce loop
formation and thus entropy, thus limiting the diameter to the
length of a single dendrimer in the core. The dendrimer
architecture also directed self-assembly toward spherical
micelles compared to the lamella structure normally found
from crew-cut diblock copolymer self-assembly.
3KMa+1
3
Rh
=
10πN
(6)
A
where K = 0.0141 cm3 g−1, a = 0.7 (in a good solvent), and NA
is Avogadro’s number. It can be seen from Table 3 that the
Dh,SEC increased from 8.41 to 9.21 nm for dendrimers 14 to 16
with an increase in the number of L-lysines on the perihpery. It
should be noted that the value of K is based on linear PSTY
chains and would be expected to change for dendritic
structures. The second and more precise method allowed
determination of the diffusion coefficient (D) from a
concentration-dependent diffusion-ordered NMR spectroscopy
(DOSY).44 The value of Dh,NMR can be calculated from the
Stokes−Einstein equation (see eq 2). The diameter (i.e.,
Dh,NMR) in CDCl3 increased from 5.97 to 9.06 nm respectively
for 14 to 16, which was supported from the very close values
found by DLS (the third method). These sizes are expected,
since the solvent is good for both PSTY and Boc-protected L-
lysine to represent the unimolecular diameter (i.e., without
aggregation or self-assembly of the dendrimers). In addition,
the excellent agreement between the Dh,DLS and Dh,NMR
suggests that the size determined by DLS is accurate and
should provide some insight into the self-assembly of these
dendrimers in water.
The self-assembly of L-lysine-decorated dendrimers into
peptidomimetic nanoparticles represents the next step toward
biological efficacy. The three Boc-protected L-lysine dendrimers
were deprotected to the free L-lysine periphery by addition of
TFA (Scheme 1). Micelles of three L-lysine dendrimers (17−
19) were formed through the slow addition of water to a
solution of dendrimer dissolved in DMF over a 6.5 h period
and then further dialyzed with Milli-Q water for 2 days to
remove organic solvent. It should be noted that there was a
small amount of precipitant found after the self-assembly of 17.
CONCLUSION
■
In summary, we have demonstrated that by using copper to
catalyze two orthogonal reactions (i.e., NRC and CuAAC)
polymeric dendrimers with 3 and 4 generational layers could be
constructed rapidly and with high purity. The purity of the 4-
arm star core, 5, was essential to the success of creating
polymeric dendrimers coated with L-lysine on the periphery.
The ATRP of the 4-arm star produced 2, the precursor to 5,
with a purity of 87.3% with the other 16.7% corresponding to
higher molecular weight star−star coupling products. After
purification by preparative SEC, the purity increased to 96.1%.
Coupling various building blocks to 5, in a process where the
M
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
(12) Whittaker, M. R.; Urbani, C. N.; Monteiro, M. J. J. Am. Chem.
Soc. 2006, 128 (35), 11360−11361.
rates of reaction for the NRC and the CuAAC were similar (i.e.,
the parallel process), allowed us to produce three polymeric
dendrimers with an increase in L-lysine peripheral density. The
purity for 14, 15, and 16 was found to be 94, 96, and 97%,
respectively, as determined using the LND method. Our work
showed that the copper-catalyzed CuAAC and NRC reactions
represent a powerful synthetic method to produce dendrimers
in one pot at 25 °C. The versatility of this synthetic approach
will have utility in the synthesis of polymeric dendrimers coated
with a wide range of biomolecules. The self-assembly of these
dendrimers in water demonstrates that due to the junction
points within the dendrimer, the particle sizes were slightly
larger than that found in organic solvents. We show here that
utilizing amphiphilic dendrimers leads to control over the
particles size due to its very low aggregation numbers. This
represents an important step to implementing these peptidomi-
metic nanoparticles to increase the biological efficacy. We will
carry out in vitro and in vivo studies in the future to confirm our
hypothesis.
(13) Urbani, C. N.; Bell, C. A.; Lonsdale, D.; Whittaker, M. R.;
Monteiro, M. J. Macromolecules 2008, 41 (1), 76−86.
(14) Urbani, C. N.; Bell, C. A.; Whittaker, M. R.; Monteiro, M. J.
Macromolecules 2008, 41 (4), 1057−1060.
(15) Hossain, M. D.; Jia, Z. F.; Monteiro, M. J. Macromolecules 2014,
47 (15), 4955−4970.
(16) Skwarczynski, M.; Zaman, M.; Urbani, C. N.; Lin, I. C.; Jia, Z.
F.; Batzloff, M. R.; Good, M. F.; Monteiro, M. F.; Toth, I. Angew.
Chem., Int. Ed. 2010, 49 (33), 5742−5745.
(17) Liu, T. Y.; Hussein, W. M.; Jia, Z. F.; Ziora, Z. M.; McMillan, N.
A. J.; Monteiro, M. J.; Toth, I.; Skwarczynski, M. Biomacromolecules
2013, 14 (8), 2798−2806.
(18) Lonsdale, D. E.; Whittaker, M. R.; Monteiro, M. J. J. Polym. Sci.,
Part A: Polym. Chem. 2009, 47 (22), 6292−6303.
(19) Bell, C. A.; Jia, Z. F.; Kulis, J.; Monteiro, M. J. Macromolecules
2011, 44 (12), 4814−4827.
(20) Trollsas, M.; Claesson, H.; Atthoff, B.; Hedrick, J. L. Angew.
Chem., Int. Ed. 1998, 37 (22), 3132−3136.
(21) Trollsas, M.; Atthof, B.; Wursch, A.; Hedrick, J. L.; Pople, J. A.;
Gast, A. P. Macromolecules 2000, 33 (17), 6423−6438.
(22) Lammens, M.; Fournier, D.; Fijten, M. W. M.; Hoogenboom,
R.; Du Prez, F. Macromol. Rapid Commun. 2009, 30 (23), 2049−2055.
(23) Urbani, C. N.; Lonsdale, D. E.; Bell, C. A.; Whittaker, M. R.;
Monteiro, M. J. J. Polym. Sci., Part A: Polym. Chem. 2008, 46 (5),
1533−1547.
ASSOCIATED CONTENT
* Supporting Information
Synthesis and characterizations of all polymers, including H
■
S
1
NMR spectra, SEC traces, and MALDI-ToF spectra. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(24) Urbani, C. N.; Bell, C. A.; Lonsdale, D. E.; Whittaker, M. R.;
Monteiro, M. J. Macromolecules 2007, 40 (19), 7056−7059.
(25) Hutchings, L. R.; Roberts-Bleming, S. J. Macromolecules 2006,
39 (6), 2144−2152.
(26) Konkolewicz, D.; Monteiro, M. J.; Petrie, S. Macromolecules
2011, 44 (18), 7067−7087.
(27) Kimani, S. M.; Hutchings, L. R. Macromol. Rapid Commun.
2008, 29 (8), 633−637.
(28) Hutchings, L. R. Soft Matter 2008, 4 (11), 2150−2159.
(29) Hu, D.; Zheng, S. X. Eur. Polym. J. 2009, 45 (12), 3326−3338.
(30) Pan, P. J.; Fujita, M.; Ooi, W. Y.; Sudesh, K.; Takarada, T.;
Goto, A.; Maeda, M. Polymer 2011, 52 (4), 895−900.
(31) Lutz, J. F.; Borner, H. G.; Weichenhan, K. Macromolecules 2006,
39 (19), 6376−6383.
AUTHOR INFORMATION
Corresponding Author
*E-mail: m.monteiro@uq.edu.au (M.J.M.).
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
M.J.M. acknowledges financial support from the ARC
Discovery grant (DP140103497).
(32) Sumerlin, B. S.; Vogt, A. P. Macromolecules 2010, 43 (1), 1−13.
(33) Inglis, A. J.; Sinnwell, S.; Stenzel, M. H.; Barner-Kowollik, C.
Angew. Chem., Int. Ed. 2009, 48 (13), 2411−2414.
(34) Kade, M. J.; Burke, D. J.; Hawker, C. J. J. Polym. Sci., Part A:
Polym. Chem. 2010, 48 (4), 743−750.
(35) Campos, L. M.; Killops, K. L.; Sakai, R.; Paulusse, J. M. J.;
Damiron, D.; Drockenmuller, E.; Messmore, B. W.; Hawker, C. J.
Macromolecules 2008, 41 (19), 7063−7070.
(36) Campos, L. M.; Meinel, I.; Guino, R. G.; Schierhorn, M.; Gupta,
N.; Stucky, G. D.; Hawker, C. J. Adv. Mater. 2008, 20 (19), 3728−+.
(37) Killops, K. L.; Campos, L. M.; Hawker, C. J. J. Am. Chem. Soc.
2008, 130 (15), 5062−+.
(38) Antoni, P.; Robb, M. J.; Campos, L.; Montanez, M.; Hult, A.;
Malmstrom, E.; Malkoch, M.; Hawker, C. J. Macromolecules 2010, 43
(16), 6625−6631.
REFERENCES
■
(1) Kaspar, A. A.; Reichert, J. M. Drug Discovery Today 2013, 18
(17−18), 807−817.
(2) Vlieghe, P.; Lisowski, V.; Martinez, J.; Khrestchatisky, M. Drug
Discovery Today 2010, 15 (1−2), 40−56.
(3) Bracci, L.; Falciani, C.; Lelli, B.; Lozzi, L.; Runci, Y.; Pini, A.; De
Montis, M. G.; Tagliamonte, A.; Neri, P. J. Biol. Chem. 2003, 278 (47),
46590−46595.
(4) Jiang, Y. H.; Emau, P.; Cairns, J. S.; Flanary, L.; Morton, W. R.;
McCarthy, T. D.; Tsai, C. C. Aids Res. Hum. Retroviruses 2005, 21 (3),
207−213.
(5) Bourne, N.; Stanberry, L. R.; Kern, E. R.; Holan, G.; Matthews,
B.; Bernstein, D. I. Antimicrob. Agents Chemother. 2000, 44 (9), 2471−
2474.
(6) Lee, C. C.; MacKay, J. A.; Frechet, J. M. J.; Szoka, F. C. Nat.
Biotechnol. 2005, 23 (12), 1517−1526.
(39) Ma, X. P.; Tang, J. B.; Shen, Y. Q.; Fan, M. H.; Tang, H. D.;
Radosz, M. J. Am. Chem. Soc. 2009, 131 (41), 14795−14803.
(40) Jia, Z. F.; Bell, C. A.; Monteiro, M. J. Chem. Commun. 2011, 47
(14), 4165−4167.
(41) Matyjaszewski, K.; Davis, T. P. Handbook of Radical Polymer-
ization; John Wiley and Sons: New York, 2002.
(7) Gajbhiye, V.; Palanirajan, V. K.; Tekade, R. K.; Jain, N. K. J.
Pharm. Pharmacol. 2009, 61 (8), 989−1003.
(8) Percec, V.; Barboiu, B.; Grigoras, C.; Bera, T. K. J. Am. Chem. Soc.
2003, 125 (21), 6503−6516.
(9) Trollsas, M.; Hedrick, J. L. J. Am. Chem. Soc. 1998, 120 (19),
4644−4651.
(42) Mantovani, G.; Ladmiral, V.; Tao, L.; Haddleton, D. M. Chem.
Commun. 2005, No. 16, 2089−2091.
(10) Hedrick, J. L.; Trollsas, M.; Hawker, C. J.; Atthoff, B.; Claesson,
H.; Heise, A.; Miller, R. D.; Mecerreyes, D.; Jerome, R.; Dubois, P.
Macromolecules 1998, 31 (25), 8691−8705.
(43) Agrawal, P.; Gupta, U.; Jain, N. K. Biomaterials 2007, 28 (22),
3349−3359.
(44) Altintas, O.; Krolla-Sidenstein, P.; Gliemann, H.; Barner-
Kowollik, C. Macromolecules 2014, 47 (17), 5877−5888.
(11) Matsuo, A.; Watanabe, T.; Hirao, A. Macromolecules 2004, 37
(17), 6283−6290.
N
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
(45) Gavrilov, M.; Monteiro, M. J. Eur. Polym. J. 2015,
DOI: 10.1016/j.eurpolymj.2014.11.018.
(46) Monteiro, M. J. Eur. Polym. J. 2015, DOI: 10.1016/
j.eurpolymj.2015.01.009.
(47) Kulis, J.; Bell, C. A.; Micallef, A. S.; Jia, Z. F.; Monteiro, M. J.
Macromolecules 2009, 42 (21), 8218−8227.
(48) Kulis, J.; Bell, C. A.; Micallef, A. S.; Monteiro, M. J. J. Polym. Sci.,
Part A: Polym. Chem. 2010, 48 (10), 2214−2223.
(49) Percec, V.; Guliashvili, T.; Ladislaw, J. S.; Wistrand, A.;
Stjerndahl, A.; Sienkowska, M. J.; Monteiro, M. J.; Sahoo, S. J. Am.
Chem. Soc. 2006, 128 (43), 14156−14165.
(50) Kulis, J.; Jia, Z. F.; Monteiro, M. J. Macromolecules 2012, 45
(15), 5956−5966.
(51) Daoud, M.; Cotton, J. P. J. Phys. (Paris) 1982, 43 (3), 531−538.
(52) Hossain, M. D.; Lu, D. R.; Jia, Z. F.; Monteiro, M. J. ACS Macro
Lett. 2014, 3 (12), 1254−1257.
(53) Whittaker, M. R.; Monteiro, M. J. Langmuir 2006, 22 (23),
9746−9752.
(54) Iatrou, H.; Willner, L.; Hadjichristidis, N.; Halperin, A.; Richter,
D. Macromolecules 1996, 29 (2), 581−591.
(55) Kim, K. H.; Kim, S. H.; Huh, J.; Jo, W. H. J. Chem. Phys. 2003,
119 (11), 5705−5710.
O
DOI: 10.1021/acs.macromol.5b00195
Macromolecules XXXX, XXX, XXX−XXX