HPMA-Based Nanocarriers for Effective Immune System Stimulation

Article abstract of DOI:10.1002/mabi.201800481

The selective activation of the immune system using nanoparticles as a drug delivery system is a promising field in cancer therapy. Block copolymers from HPMA and laurylmethacrylate-co-hymecromone-methacrylate allow the preparation of multifunctionalized

Full text of DOI:10.1002/mabi.201800481

Full PaPer
Nanocarriers
www.mbs-journal.de
HPMA-Based Nanocarriers for Effective Immune
System Stimulation
Stefan Kramer, Jens Langhanki, Matthias Krumb, Till Opatz, Matthias Bros,
and Rudolf Zentel*
clinical trials, like Doxil, Abraxane or
Narekt-102, Paclical, and others.[3–5] These
formulations take advantage of the so-
called enhanced permeability and reten-
tion (EPR) effect, which is a characteristic
of many solid tumors.[6,7] Besides tumors,
also other types of tissue like the liver and
the spleen have an open vasculature and
may accumulate nanoparticle (NP). Due
to the presence of antigen presenting cells
(APC) in the spleen, this organ constitutes
an attractive target for nanoparticles.[8]
Concerning antitumor immunotherapies,
nanocarriers that codeliver an antigen and
an adjuvant are intended to target APC
which in turn elicit an adaptive antitumor
T cell response.[9–11]
The selective activation of the immune system using nanoparticles as a drug
delivery system is a promising field in cancer therapy. Block copolymers from
HPMA and laurylmethacrylate-co-hymecromone-methacrylate allow the prep-
aration of multifunctionalized core-crosslinked micelles of variable size. To
activate dendritic cells (DCs) as antigen presenting cells, the carbohydrates
mannose and trimannose are introduced into the hydrophilic corona as DC
targeting units. To activate DCs, a lipophilic adjuvant (L18-MDP) is incorpo-
rated into the core of the micelles. To elicit an immune response, a model
antigen peptide (SIINFEKL) is attached to the polymeric nanoparticle—in
addition—via a click reaction with the terminal azide. Thereafter, the differ-
ently functionalized micelles are chemically and biologically characterized.
While the core-crosslinked micelles without carbohydrate units are hardly
bound by DCs, mannose and trimannose functionalization lead to a strong
binding. Flow cytometric analysis and blocking studies employing mannan
suggest the requirement of the mannose receptor and DC-SIGN for effective
micelle binding. It could be suppressed by blocking with mannan. Adjuvant-
loaded micelles functionalized with mannose and trimannose activate DCs,
and DCs preincubated with antigen-conjugated micelles induce proliferation
of antigen-specific CD8+ T cells.
Generally,
the
immune
system
responds to pathogens like viruses or
bacteria. Moreover, the immune system
can—under certain circumstances—
also eliminate cancer cells and it has the
potential to combat metastases, which is
a major challenge in cancer therapy.[12,13]
To efficiently induce an antigen-specific
adaptive immune response, the nano-
1. Introduction
particle needs to codeliver an antigen and an APC-stimulating
agent (adjuvant) into an APC like a dendritic cell (DC).[14,15]
The activated DCs then mediate the activation/differentiation
of antigen-specific cytotoxic T cells (CD8+ T cells) and T helper
cells (CD4+ T cells), and thereby induces a tumor antigen
specific immune response.[16] Especially, activated DCs are
able to elicit primary immune responses. To target DCs, a
specific targeting vector which binds to a distinct surface pro-
tein is required. This ligand can be an antibody or carbohy-
drates like mannose or trimannose.[17] It is known that cells
of the immune system, especially DCs, express the mannose
receptor (CD206)[18,19] and DC-SIGN (CD209).[20] Both recep-
tors preferably bind mannose and trimannose, respectively. As
a consequence, if the nanoparticle carries these carbohydrates
in its corona, it may be specifically endocytosed by receptor-
expressing immune cells.[21,22] In order to activate DCs, toll
like receptor (TLR) or nucleotide-binding domain like receptor
(NLR) ligands are often used.[23,24] Such small molecules can
be incorporated into the core of micellar nanoparticles easily.
As outlined above, an antigen is necessary to elicit an antigen-
specific immune response. To this end, a model antigen
can be attached to the drug delivery system. Accordingly, a
Nanomedicine has found growing interest concerning cancer
treatment during the last decades. Advantages of nanofor-
mulated drug delivery systems are that they can encapsulate
poorly soluble drugs and improve their circulation time in
blood.[1,2] In the last few decades several nanoparticulate drug
delivery systems have entered the clinic or are evaluated in
Dr. S. Kramer, Dr. J. Langhanki, M. Krumb, Prof. T. Opatz, Prof. R. Zentel
Institute of Organic Chemistry
Johannes Gutenberg-University Mainz
Duesbergweg 10–14, 55128, Mainz, Germany
E-mail: zentel@uni-mainz.de
Dr. M. Bros
Department of Dermatology
University Medical Center
Johannes Gutenberg-University Mainz
Obere Zahlbacher Straße 63, 55131, Mainz, Germany
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/mabi.201800481.
DOI: 10.1002/mabi.201800481
©
1800481 (1 of 10)
Macromol. Biosci. 2019, 1800481
2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.mbs-journal.de
trifunctional nanoparticle, bearing an APC targeting function,
an adjuvant, and an antigen, is essential for an effective activa-
tion of the immune system.
following components: pump PU 1580, auto sampler AS
1555, UV-detector UV 1575, RI-detector RI 1530 from Jasco.
Applied columns were purchased from MZ-Analysentechnik:
MZ-Gel SDplus 102 Å, MZ-Gel SDplus 104 Å, MZ-Gel SDplus
106 Å. HPMA polymers were analyzed using HFIP as sol-
vent containing 3 g L−1 potassium trifluoroacetate. For HFIP-
GPC, the following components were used: pump PU 2080+,
autosampler AS1555, and RI detector RI2080+ from Jasco. As
standard polystyrene (PS) respectively polymethylmethacrylate
(PMMA) for THF- and HFIP-GPC was used. For evaluation,
the software WinGPC Uni Chrom was used. The flow rate was
1 mL min−1 at 25 °C (THF) or 40 °C (HFIP).
To accomplish this sophisticated task, HPMA-based amphi-
philic block copolymers are a promising platform. For the
present study, a polymerization strategy via reversible addi-
tion fragmentation chain transfer (RAFT) polymerization and
the reactive ester approach was chosen.[25] To this end, a poly-
pentafluorophenyl methacrylate (p(PFPMA)) was synthesized
using an azide-functionalized chain transfer agent (CTA). This
homopolymer was further utilized as a macro-chain transfer
agent (macro-CTA). Thereafter a hydrophobic block containing
lauryl methacrylate (LMA) and a cross linkable hymecromone
methacrylate (HCMA) were attached as recently described.[26,27]
In the last step of polymer synthesis, the so-called polymer ana-
logues reaction, different functionalities can be attached to the
polymer backbone yielding the final p(HPMA)-b-p(LMA)-ran-
p(HCMA) containing carbohydrate moieties and dye labels.
This polymerization strategy yields well-defined and function-
alized block copolymers. Thereafter, the size of the micellar
aggregates can be adjusted[27] during preparation to fulfill the
requirements of a DC-focused immunotherapeutic approach.
UV–vis spectroscopy was carried out in a Jasco UV–vis spec-
trometer type V-360 using a 10 mm quartz cell. The spectra
were analyzed using Spectra Manager (Version 2.04) software.
2.3. Synthesis of Pentafluorophenyl 4-cyano-4-
(phenylcarbonothioylthio)pentanote
In
a
reaction flask, 2.05
g
(7.16 mmol) 4-cyano-4-
(phenylcarbonothioylthio)pentanoic acid (COOH-CTA) was
dissolved in abs. THF. The solution was kept under argon and
triethylamine (1.99 mL, 14.32 mmol) was added under stirring.
Then, pentafluoro phenyl trifluoroacetate was added in THF.
Subsequently, the pink solution was stirred for 4 h at room tem-
perature. The solvent was then removed by evaporation and the
residue dissolved in abs. dichloromethane (DCM). The solution
was extracted with water and dried over magnesium sulfate.
The pink solid was purified twice by column chromatography
(ethyl acetate/petroleum ether, 15:1, DCM/petroleum ether,
1:5), yielding a dark red solid. Yield: 1.33 g (2.98 mmol, 42%)
1H-NMR: (CDCl3, 400 MHz): δ [ppm] = 7.96 (dt, J = 8.6,
2. Experimental Section
2.1. Materials
Solvents of technical grade were distilled before usage. Tetrahy-
drofuran (THF), hexane and 1,4-dioxane were distilled from
sodium, using benzophenone as an indicator, chloroform and
dichlormethane from CaH2. All solvents of p.a. (pro analysis)
grade were purchased from Acros (Geel, Belgium), Sigma-
Aldrich (Darmstadt, Germany), Roth (Karlsruhe, Germany), or
Fluka (Schwerte, Germany) and utilized without further puri-
fication. Water-free DMSO (dimethyl sulfoxide) was stored in
a sealed bottle over molecular sieves (3 Å). The chain transfer

1.6 Hz, 2H, Ar H); 7.61 (ddt, J = 8.7, 7.1, 1.2 Hz, 1H,


Ar H); 7.44 (m, 2H, Ar H); 3.16–3.02 (dt, J = 9.4,6.0 Hz, 2H,




CH2 ); 2.83–2.75 (m, 1H, CH2 ); 2.62–2.55 (ddd, J = 14.3,
9.6, 6.4 Hz, 1H, CH2); 2.02 (s, 3H, CH3).
agent
(CTA)
4-cyano-4-phenylcarbonothioylthio)pentanoic
19F-NMR: (CDCl3, 400 MHz): δ [ppm] = −153.66 (d, 2F, o);
−158.40 (t, 1F, p) −163.01 (dd, 2F, m).
acid was purchased from Sigma-Aldrich. The Initiator V-70
2,2´-azobis(4-methoxy-2,4-dimethylvaleronitrile) was obtained
from Wako Chemicals (Neuss, Germany). Lauryl methacrylate
(LMA) was purchased from Sigma-Aldrich and distilled prior
to use. Oregon Green 488 cadaverine was purchased from
InvitroGen (Darmstadt, Germany). The functionalized pep-
tide SIINFEKL-DBCO was purchased from Bachem (Weil am
Rhein, Germany). The adjuvant L18-MDP was obtained from
InvivoGen (Toulouse, France).
2.4. Synthesis of 1-azido-24-cyano-21-oxo-2,5,8,11,14,17-hex-
aoxa-20-azapentacosan-24-yl Benzodithioate
A total of 70 mg (0.157 mmol) of pentafluorophenyl 4-cyano-
4-(phenylcarbonothioylthio)pentanote (PFP-CTA) was dissolved
in ≈10 mL abs. THF. The solution was kept under argon atmos-
phere and 18 mg trimethylamine (0.181 mmol) was added
under stirring. H2N-(PEG)6-N3 was dissolved in abs THF and
was slowly added to the red solution. The reaction mixture was
stirred for 4 h at room temperature. Afterward, the solvent was
removed in vacuo. The red solid was dissolved in abs. DCM
and was washed with saline and dried in vacuo. The solid was
purified by column chromatography (ethylacetate/isopropanol,
2.2. Characterization
1H- and 19F-NMR spectra were recorded at 400 MHz on a
Bruker Avance II 400 equipped with a 5 mm BBFO probe.
All spectra were measured at room temperature and analyzed
using MestReNova software.
Polymers were dried at 40 °C overnight and afterward
were analyzed using gel permeation chromatography (GPC).
GPC of precursor polymers was accomplished using THF
or hexafluoroisopropanol (HFIP) as eluent and using the
4:1), yielding a dark red solid. Yield: 49 mg (0.082 mmol, 52%)
1

H-NMR (CDCl3, 400 MHz): δ [ppm] = 7.93 (m, 2H, Ar H);


7.58 (tt, J = 7.1, 1,2, 1H, Ar H); 7.42 (m, 2H, Ar H); 6.53 (s,
1H, NH); 3.69–3.67 (m, 20 H, O CH2 CH2); 3.60–3.58 (t,
J = 5.3 Hz, 2H, CH2 ); 3.49–3.48 (m, 2H, CH2 ); 3.42–3.39






©
1800481 (2 of 10)
Macromol. Biosci. 2019, 1800481
2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.mbs-journal.de
19F-NMR (400 MHz, CDCl3): δ [ppm] −151.5 to −153.1 (br,
2F), −157.9 to −158.2 (br, 1F), −162.9 to −163.4 (br, 2F).


 

(t, J = 5.3, 2H, NH CH2 ); 2.69–2.63 (m, 1H, C O CH2 );



2.62–2.56 (m, 2H, CH2 ); 2.48–2.41 (m, 1H, CH2–); 1.97 (s,



3H, CH3); 1.30–1.27 (m, 1H, N3 CH2 ).
2.9. Synthesis and Functionalization of p(HPMA)-b-p(LMA)-ran-
2.5. Synthesis of Pentafluorophenyl Methacrylate Monomer
p(HCMA) Copolymer
The pentafluorophenyl methacrylate monomer (PFPMA)
monomer was synthesized according to literature.[28]
In a typical reaction, 100 mg (0.45 µmol) of reactive ester polymer
was dissolved in 2 mL abs. dioxane. Then, 1.7 mg (0.5 µmol) of
the dye Oregon Green cadaverin in 1 mL DMSO and 0.12 mL of
triethylamine (TEA) as a base were added. After stirring 6 h at
40 °C, 5 equiv. of the specific targeting moiety and 0.12 mL TEA
were added. This solution was further stirred overnight at 40 °C.
Complete attachment can be verified by thin layer chromatog-
raphy as described in ref. .[29] Subsequently, hydroxypropylamine
(HPA) (47 mg, 0.6 mmol) was added to the solution and stirred
for 24 h under argon atmosphere. Complete reaction was proved
by the disappearance of the fluorine signals of the reactive ester
(see Figure S2, Supporting Information). For purification, the mix-
ture was diluted with Milli-Q water and transferred into a dialysis
bag (molecular weight cut-off (MWCO) 3.5 kDa) and dialyzed for
3 days against water changing water three times a day. The orange
solution was then lyophilized to give an orange fluffy powder.
1H-NMR (400 MHz, DMSO-d6): δ [ppm] = 7.6 (br, 1H),
2.6. Synthesis of Hymecromone Methacrylate Monomer
The hymecromone methacrylate monomer (HCMA) was
synthesized as recently described.[27]
2.7. Synthesis of Poly-Pentafluorophenyl Methacrylate
Homopolymer p(PFPMA)
In a typical reaction, PFPMA (4.08 g, 16.19 mmol) and
4-cyano-4-phenylcarbonothioylthio)pentanoic
acid
(CTA)
(52 mg, 0.186 mmol) and V-70 (5.7 mg, 0.0186 mmol) were
dissolved in 10 mL of abs. dioxane. The pink solution was
subsequently degassed by three freeze pump thaw cycles and
polymerized at 40 °C for 17 h. Afterward, conversion was

7.4–7.3 (br, 1H, NH), 7.1 (br, 2H, Ar–H), 6.3 (br, 1H, Ar–H),
 

4.7 (br, 1H, C OH), 4.4 (br, 1H, COOCH2–), 3.7 (br, 1H,
1




determined by H-NMR the polymer was precipitated three
CH HPA), 3.1–2.6 (br, 2H, NH CH2 HPA), 1.44–0.76 (br,
polymer backbone).
times in cold n-hexane. After removing the solvent in vacuo,
the homopolymer was yielded as a pink powder. Typical yield:
1.92 g (48 wt%)
1H-NMR (300 MHz, CDCl3): δ [ppm] 2.1–2.5 (br, 2H), 1.3–
1.6 (m, 3H).
2.10. Synthesis of Mannose–Amine
19F-NMR (400 MHz, CDCl3): δ [ppm] −151.5 to −153.1 (br,
2F), −157.9 to −158.2 (br, 1F), −162.9 to −163.4 (br, 2F).
All synthetic details of the synthesis of 11-amino-3,6,9-trioxa-
undecyl-α-d-mannopyranoside (M) are described by Mohr
et al.[21]
2.8. Synthesis of p(PFPMA)-b-(LMA)-ran-p(HCMA) Precursor
Block Copolymer
2.11. Synthesis of Trimannose–Amine
For the synthesis of the block copolymer the p(PFPMA) homo
polymer was used as a macro-CTA. Hence, p(PFPMA) (1 g,
0.05 mmol), hymecromone methacrylate (HCMA) (146 mg,
0.6 mmol), lauryl methacrylate LMA (144 mg, 0.6 mmol), and
V-70 (1.7 mg, 0.005 mmol) were dissolved in abs. dioxane. After
three pump–thaw cycles, the solution was polymerized for 72 h
at 40 °C. The block copolymer was precipitated three times in
cold ethanol to ensure removal of excess monomer. Remaining
solvent was removed in vacuo and the polymer was character-
ized by NMR and GPC.
Removal of the dithiobenzoate endgroups was performed
by adding 25-fold excess of V-70 to a block copolymer solution
in abs. dioxane. The reaction mixture was stirred for 16 h at
40 °C. Afterward, the colorless solution was precipitated three
times in cold ethanol and dried in vacuo, yielding the final
precursor polymer as a colorless powder. Typical yield: 970 mg
For details of the synthesis of (1-(2-(2-(2-(2-aminoethoxy)
ethoxy)ethoxy)ethyl)-1H-1,2,3-triazol-4-yl)methoxy)-3,6-di-O-
α-d-mannopyranosyl-α-d-mannopyranose (T), see Supporting
Information.
2.12. Micelle Preparation
Three milligrams of the distinct p(HPMA)-b-p(LMA)-ran-
p(HCMA) containing no targeting unit, mannose or triman-
nose was dissolved in anhydrous DMSO. The clear solution
was transferred into a dialysis bag (MWCO 3.5 kDa) and gradu-
ally dialyzed against Milli-Q water. For micelles consisting of
trimannose and mannose polymers, respectively, 1.5 mg of
both polymers were applied.
For encapsulation 30 µg of L18-MDP were dissolved in
DMSO and added to the polymer solution prior to dialysis. To
verify the incorporation of L18-MDP into the block copolymer
micelles, we proved that the dialyzed aqueous solution was free
of L18-MDP by bioassays assessing its activating properties.
(75 wt%).
1

H-NMR (300 MHz, CDCl3): δ [ppm] 7.6 (br, 1H, Ar H), 7.1



(br, 2H, Ar H), 6.3 (br, 1H, Ar H), 3.9 (br, 2H, CH2), 2.4–0.9
(m, 5H, polymer backbone).
©
1800481 (3 of 10)
Macromol. Biosci. 2019, 1800481
2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.mbs-journal.de
After complete solvent exchange, the aqueous solution was
transferred to a plastic vessel and crosslinked under UV light
for 10 min.
FACS analysis was performed on an Attune Nxt flow cytometer
(ThermoFisher). Data was analyzed using Attune Nxt software.
For subsequent CD8+ T cell proliferation assays, BMDC
were incubated overnight with the indicated type of micelles
(each 30 mg mL−1). On the next day, BMDC were harvested,
thoroughly washed, and cocultured at serially titrated numbers
(starting concentration: 2 × 104 per well) in triplicates with
SIINFEKL-specific OT-I CD8+ T cells (each 5 × 104 per well) in
96 well cluster plates for 3 days in a volume of 0.2 mL culture
medium. T cell proliferation was assayed as genomic incorpo-
2.13. SIINFEKL Click Reaction
To generate peptide containing particles, 1 mL of the particle
solution (1 mg mL−1) was transferred to a 2 mL Eppendorf
vial equipped with a stirring bar. The peptide (0.16 mg) was
dissolved in DMSO and added to the particle solution. The
solution was stirred at room temperature for 24 h. Afterward,
the particles were spin-filtered ≈20 times (MWCO 30 kDa).
Thereby, no detectable free SIINFEKL (protein) was found in
the low molar mass fraction.
3
ration of H-thymidine (0.25 µCi per well) added on day 2 of
coculture for 16 h. Cells were harvested onto glass fiber filters,
and retained radioactivity was measured in a ß counter (1205
Betaplate, LKB Wallac, Turcu, Finland).
3. Results and Discussion
2.14. Characterization of the Polymeric Micelles
3.1. Polymer Synthesis
DLS measurements were executed on a multiangle setup, con-
sisting of an ALV-SP125 goniometer equipped with a single-
photon detector SO-SIPD, an ALV-5000 Multiple-Tau digital
correlator and a Spectra Physics 2060 argon ion laser (500 mW
output at 514.5 nm wavelength). The scattering intensity was
divided in half by a beam splitter and each half was detected
by a photomultiplier. The signals were cross-correlated for
elimination of nonrandom electronic noise. All samples were
measured from 30° to 150° in steps of 15°. Data analysis was
implemented according to literature.[30,31] Nanoparticle samples
were quantified at 0.1 g L−1 in PBS. Samples were prepared in
a dust-free cylindrical quartz cuvette (20 mm diameter, Hellma,
Mühlheim, Germany) and filtered through Millex GHP filters,
0.2 µm pore size (Millipore).
The goal of this study was to develop a polymer platform for
nanoparticle formulations to introduce the three important
components required for the induction of selective immune
response: targeting, activation, antigen presenting. The poly-
mers used for this purpose should allow it, in addition, to
prepare core-crosslinked micelles of variable size. In order to
establish such a system, RAFT polymerization was used. In a
first step, an azide functionalized chain-transfer agent (CTA
2) was synthesized via a reactive-ester CTA (1-R). (Scheme 1a)
Then CTAs 1 or 2 were used for homopolymerization, using the
low-temperature initiator V-70, yielding a non-functionalized
(P1-H) or an azide functionalized reactive ester homo polymer
N3-p(PFPMA) P2-H (Scheme 1b). For the purpose of synthe-
sizing amphiphilic block copolymers P1 or P2, the homopoly-
mers were used as a macro-CTA and copolymerized with the
two hydrophobic monomers LMA and HCMA resulting in the
precursor block copolymers P1-R or N3-p(PFPMA)-b-p(LMA)-
ran-p(HCMA) P2-R (see Table 1). GPC result of P1-R is pre-
sented in Supporting Information. The hymecromone unit is
important to obtain core-crosslinked micelles. Since it is a cou-
marin derivative, it can undergo a [2 + 2]-cycloaddition under
irradiation, which leads to crosslinking in the hydrophobic core
to stabilize the final drug delivery system.[27,33]
Standard dynamic light scattering measurements were per-
formed at 37 °C using a Malvern Zetasizer NanoZS with a
633 nm He/Ne Laser at a fixed scattering angle of 173°. Nano-
particles were measured at 0.1 g L−1 in water and filtered with
GHP filters 0.2 µm pore size.
2.15. Biomedical Experiments
Bone marrow-derived dendritic cells (BMDC) were differenti-
ated from C57BL/6 bone marrow progenitor cells in 12 well
cell cluster plates (2 × 105 per well) in the presence of GM-CSF
(10 ng mL−1; R&D Systems, Wiesbaden, Germany) for 1 week.
Media was replenished on days 3 and 6 of culture. CD8+ T cells
were immunomagnetically sorted (Miltenyi Biotec, Bergisch-
Gladbach, Germany) from spleens of transgenic OT-I mice.[32]
Expression of BMDC surface markers was assessed using
fluorescence-labeled receptor-specific rat antimouse antibodies
(CD11c: BV421 or PE, CD86: eFl450, CD206: BV421, CD209:
eFl660, MHCII: APC) obtained from ThermoFisher (Waltham,
MA) or Biolegend (San Diego, CA). Receptor expression was
assessed for CD11c+ BMDCs. Data denote either frequencies
of receptor-positive BMDC or the mean fluorescence intensity
(MFI) of receptor expression. In blocking studies, BMDC were
preincubated with mannan (2 mg mL−1; Saccharomyces cerevi-
siae; Sigma-Aldrich) for 30 min prior to application of micelles.
The molecular weight of the reactive block-copolymers P1-R
and P2-R was determined by GPC (see Figure S1, Supporting
Information); their composition was determined by NMR. All
properties are compiled in Table 1.
The final conversion into the amphiphilic HPMA-based
block copolymers P1 and P2 was realized by polymer analogues
reactions using 2-hydroxy-propylamine (HPA) at the end. In
the case of functionalized polymers 1 mol% of Oregon Green
cadaverine bearing an amine functionality was added at first,
allowing 2 h for full reaction with reactive ester moieties. To
prepare the polymers containing the targeting moieties, 5 mol%
of amine-functionalized mannose (M) or trimannose T (see
Supporting Information for their synthesis) was added and the
solution was stirred for an additional 6 h (Figure 1). Previously,
we had shown that such a reaction of a reactive ester polymer
and amino-functionalized carbohydrates runs quantitatively.[29]
©
1800481 (4 of 10)
Macromol. Biosci. 2019, 1800481
2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.mbs-journal.de
Scheme 1. Rational of polymer synthesis a) 2: azide functionalized chain transfer agent, b) P1-H/P2-H reactive ester homo polymer and macro-CTA
bearing endgroup Re1 or Re2, c) precursor block copolymer, d) final HPMA block copolymer carrying different side functionalizations Rs (compare
Figure 1).
In the final step, the remaining PFP-groups were quenched
with an excess of hydroxy-propylamine.
mannose units (T) in the hydrophilic block. It should be noted
that P1, P1-M, and P1-T (as well as P2, P2-M, and P2-T) possess
the same degree of polymerization as they are made from the
same precursor polymer P1-R or P2-R.
As has been shown previously,[34] the conversion of such
reactive ester polymers with primary amine groups works
quantitatively without side reactions, if proper conditions are
applied. 19F-NMR was used to confirm complete conversion
of reactive ester units (see 2, Supporting Information). The
amount of targeting units M or T was determined by NMR (see
Figures S32 and S33, Supporting Information). It corresponds
to the molar amount added. This protocol leads to polymers
compiled in Table 2. They contain 70–80 hydrophilic units and
about 10 hydrophobic units per block copolymer. They are—on
average—functionalized with four mannose (M) or four tri-
3.2. Nanoparticle Preparation
To compare the efficacy of different targeting moieties like
mannose or trimannose, we prepared various particles using
the described amphiphilic block copolymers (Table 2). First, a
“naked” particle with no targeting function; second, a mannose-
functionalized particle; third, a particle bearing trimannose and
©
1800481 (5 of 10)
Macromol. Biosci. 2019, 1800481
2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.mbs-journal.de
Table 1. Precursor polymers P1-R and P2-R bearing different end groups (see Supporting Information for GPC).
b
c
n [g mol−1
22 200
]
Label
P1-R
P2-R
Polymer
Hydrophobic blocka (LMA and HCMA)
Ð
Pn
Pn
M
p(PFPMA)-b-p(LMA)-ran-p(HCMA)
N3-p(PFPMA)-b-p(LMA)-ran-p(HCMA)
10 mol%
9.5 mol%
1.18
1.36
18
21
70
83
26 200
a)As determined by NMR; b)Degree of polymerization of the hydrophobic block; c)Degree of polymerization of the hydrophilic block.
Figure 1. Structure of mannose-NH2 (M) and trimannose-NH2 (T) used for side chain functionalization by reaction with the reactive ester units.
fourth a particle containing both targeting units mannose and
trimannose were prepared. (Figure 2) These micellar nanoparti-
cles were prepared using slow dialysis[27] (see Experimental Sec-
tion). Thus, 3 mg of the desired polymer were dissolved in 1 mL
abs. DMSO, transferred into a dialysis bag (MWCO 3.5 kDa)
and dialyzed against water for 3 days changing water three
times per day. In order to generate nanoparticles comprising
both targeting moieties, 1.5 mg of each polymer were used. The
resulting polymeric micelles possess therefore the same total
amount of targeting units (either mannose or trimannose) as
the other particles but carry only half of the amount of either
mannose or trimannose per particle compared to the particles
decorated with pure carbohydrate units. For incorporation of
the adjuvant, the polymer was dissolved in DMSO and after 1 h
the drug L18-MDP (see Supporting Information for molecular
structure) was added in DMSO giving the solution 1 h to equili-
brate. Afterward, the solution was dialyzed against water as
described in the Experimental Section. Based on the composi-
tion, 1 mg of nanoparticles was loaded with 10 µg of L18-MDP.
After preparation, the particles were crosslinked for 10 min
under UV light, as described recently,[27] to prepare stable nano-
particles. Composition and size of all nanoparticles, which are
core-crosslinked micelles are collected in Table 3.
In order to attach the antigen, preformed—and eventually
loaded—photo crosslinked nanoparticles were transferred into a
1.5 mL Eppendorf vial. One equivalent of SIINFEKL-DBCO (see
Supporting Information for molecular structure) was added in
DMSO under stirring. After stirring the solution at room tem-
perature for 24 h, it was purified by spin filtration (≈20 times,
MWCO 30 kDa) yielding NP7–NP11 (Table 3). This process leads
to a loading of about 160 µg of antigen per 1 mg of nanoparticles.
The size of all nanoparticles (NP1–NP11) was determined by
dynamic light scattering (see Figure S34, Supporting Informa-
tion). The data are summarized in Table 3. It can be seen that
all core-crosslinked nanoparticles NP1–NP6 have a comparable
diameter, which is between 20 and 26 nm, irrespective of the
nature of the sugar moiety and independent
Table 2. Amphiphilic block copolymers P1 and P2 bearing mannose (M) or trimannose (T) as
of the loading with L18-MDP as adjuvant.
Functionalizing the nanoparticles NP7–NP11
with an antigen (SIINFEKL) leads, however,
to an increase of the particle size. How-
ever, all particles possess a suitable size for
biomedical applications, which is between
20 nm (above the threshold for renal elimina-
tion) and 65 nm, which is small enough for
an efficient tissue penetration.[5,35]
targeting moieties as well as the amount of targeting units.
Targeting moietiesa
—
Mnb [g mol−1
14 600
]
Label
P1
Polymer
p(HPMA)-b-p(LMA)-ran-p(HCMA)
Man-p(HPMA)-b-p(LMA)-ran-p(HCMA)
P1-M
5 mol%
15 800
221.5 µmol g−1
5 mol%
P1-T
Triman-p(HPMA)-b-p(LMA)-ran-p(HCMA)
17 400
201.2 µmol g−1
—
P2
N3-p(HPMA)-b-p(LMA)-ran-p(HCMA)
17 200
18 400
3.3. Biological Evaluation
P2-M
N3-Man-p(HPMA)-b-p(LMA)-ran-p(HCMA)
5 mol%
217.4 µmol g−1
5 mol%
3.3.1. Binding
P2-T
N3-Triman-p(HPMA)-b-p(LMA)-ran-p(HCMA)
20 000
200.0 µmol g−1
A selective uptake of nanoparticles by the
desired immune cell population is impor-
tant, especially for systemic administra-
tion. To analyze the binding properties of
a)Targeting moieties in mol% per hydrophilic block and as µmol g−1 (of polymer); b)Molecular weight
calculated based on the molecular structure (including carbohydrates) assuming full conversion as
described in refs.[29,34]
©
1800481 (6 of 10)
Macromol. Biosci. 2019, 1800481
2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.mbs-journal.de
Figure 2. Core-crosslinked micelles with different targeting moieties.
the core-crosslinked micelles to DCs, flow cytometry was per-
formed. To this end, murine bone marrow derived dendritic
cells (BMDCs) were incubated in parallel assays with NP1–
NP4 overnight. On the next day, the frequencies of NP-positive
BMDC were assessed. To assess the binding of the differentially
functionalized micelles (non-decorated, mannose, trimannose,
mannose + trimannose functionalized micelles) to the intended
target surface receptor expressing DC subpopulations, that
is, of mannose-conjugated micelles to the mannose receptor
(CD206) and of trimannosylated micelles to DC-SIGN (CD209),
expression of these surface markers was detected in addition.
In Figure 3, the results are presented for different nanopar-
ticle concentrations (3–15 µg mL−1). As the molecular weight
of P1-M and P1-T is rather similar, equal amounts of nanopar-
ticles correspond to “nearly” equal molar amounts of carbohy-
drate targeting units. In this way, the concentration of nanopar-
ticles in Figure 3 transforms into 3, 5, or 15 times 221 pmol
mL−1 (for mannose) or 201 pmol mL−1 (for trimannose). Thus,
the molar concentration of trimannose is about 10% smaller,
but trimannose contains three closely packed monomer units
(Figure 1), and it might also bind to the mannose receptor.
As shown in Figure 3, CD206+ and CD209+ presenting
BMDC subpopulations displayed similar NP binding patterns.
The non-functionalized NP1 showed only very low binding at
lower concentrations. It increased considerably to the highest
NP concentration tested (15 µg mL−1).[36] In contrast, already at
the lowest concentration, nanoparticles surface-functionalized
with mannose (NP2), trimannose (NP3), or both carbohydrates
(NP4) were engaged at almost maximal extent by CD206+ and
CD209+ BMDC, respectively.
Next, to evaluate the functional relevance of CD206 and DC-
SIGN to bind the differentially surface-functionalized micelles,
BMDC were preincubated for 30 min with mannan as a high
affinity ligand—especially for CD206—to block the binding
of subsequently applied ligands.[37] Then, both untreated and
mannan-pretreated BMDC were incubated with the differen-
tially functionalized micelles NP1–NP4 for 3 h. As shown in
Figure 4, the minor unspecific binding of non-functionalized
Table 3. Core-crosslinked micelles for in vitro tests.
Particle name
NP1
Targeting unit
none
Polymer
P1
Diameter [nm]
Adjuvants
—
Antigen
—
23
20
26
22
25
23
61
45
51
64
65
NP2
Mannose
P1-M
—
—
NP3
Trimannose
Man + Triman
Man
P1-T
—
—
NP4
P1-M/P1-T
P1-M
—
—
NP5
L18-MDP
L18-MDP
—
NP6
P1-M/P1-T
P2
—
Man + Triman
none
NP7
SIINFEKL
SIINFEKL
SIINFEKL
SIINFEKL
SIINFEKL
NP8
Mannose
P2-M
—
—
NP9
Trimannose
Man + Triman
Man + Triman
P2-T
NP10
NP11
P2-M/P2-T
P2-M/P2-T
—
L18-MDP
©
1800481 (7 of 10)
Macromol. Biosci. 2019, 1800481
2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.mbs-journal.de
Figure 3. Binding of mannose- and trimannose-functionalized micelles NP1–NP4 (µg nanoparticle per mL) to BMDC. BMDC were incubated
with various amounts of NP1–NP4 (see Table 3). On the next day, BMDC surface expressions of the mannose receptor 206 and DC-SIGN (CD209)
and of NP-specific OG488 were monitored by flow cytometry. Data denote the frequencies of OG488+ BMDC that express CD206 (left panel) and
CD209 (right panel). Data show the mean SEM of 3–4 independent experiments each. For FCS measurements, see also Figure S36, Supporting
Information.
micelles NP1 to BMDC was not affected by mannan. In con-
trast, pretreatment of BMDC with mannan strongly reduced the
specific binding of mannosylated or trimannosylated micelles
(NP2–NP4). The reduction was strongest for mannosylated
particles (NP2) for which binding was attenuated to a similar
extent as observed for the unmodified particle NP1. This dem-
onstrates that the binding is due to specific interactions of car-
bohydrate-functionalized NP with carbohydrate-binding recep-
tors expressed by BMDC and not due to interactions with other
types of receptors.
These results are in agreement with previous studies, which
show that in particular CD206, but also CD-SIGN, can bind
to mannan. CD206 is blocked probably more easily than DC-
SIGN, which requires three correctly placed mannose units for
optimal binding.[38] It also demonstrates that trimannose (T) is
probably a more robust targeting unit, since its receptor is less
easily blocked with simple carbohydrates.
3.3.2. Activation
Based on the improved BMDC binding properties of micelles
conjugated with mannose plus trimannose (NP4), we next
assessed the potential of NP4 micelles to serve as a platform
for nanovaccines intended to target APCs like DCs to code-
liver antigen and adjuvant to induce antigen-specific immune
responses. First, the suitability of such micelles to deliver an
adjuvant to BMDCs to mediate their activation was assessed.
So far, most nanosized drug delivery systems are equipped
with hydrophilic immunostimulatory molecules that engage
intracellular receptors after internalization, for example, CpG-
rich oligonucleotides that bind TLR9 (toll-like receptor 9).[39]
However, a growing number of lipophilic adjuvants has been
introduced as well which may be encapsulated by the type of
micelles presented here. As a proof of concept, we opted for a
derivative of muramyl dipeptide (MDP) which constitutes the
minimal peptidoglycan motif occurring in all bacteria. MDP
specifically engages the cytoplasmic danger receptor NOD2
(nucleotide-binding oligomerization domain-containing pro-
tein 2). A 6-O-acyl derivative with a stearic acid (L18-MDP, see
Figure S38, Supporting Information) was reported to display an
improved membrane permeability and thereby a strong stimu-
latory potential on APC.[40]
Figure 4. Competitive binding of mannan and differentially functional-
ized micelles to BMDC. BMDC were preincubated with mannan at high
concentration (2 mg mL−1; 30 min) prior to the application of NP1-NP4
(each 15 µg mL−1). On the next day, frequencies of OG488+ BMDC were
assessed by flow cytometry. Data denote the binding efficiency of NP
applied to BMDC preincubated with mannan given as fold of binding to
non-preincubated BMDC (see also Figure S39, Supporting Information).
©
1800481 (8 of 10)
Macromol. Biosci. 2019, 1800481
2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.mbs-journal.de
Figure 6. T cell proliferation induced by BMDC preincubated with dif-
ferentially functionalized micelles. BMDC were incubated in parallel with
different NP formulations (each 15 µg mL−1; see Table 3). On the next day,
harvested BMDC were titrated and cocultured with OVA peptide-specific
CD8+ T cells in triplicates for 3 d. T cell proliferation was assessed by incor-
poration of 3H thymidine. Data denote the mean SEM of triplicates. The
data are representative of two independent experiments. Statistical differ-
ences were assessed pairwise between the different groups (t-test). T cell
proliferation of SIINFEKL-containing micelles (NP7, NP10, NP11) was
significantly higher as compared to non-functionalized NP4 (not shown
for reasons of clarity). Statistically significant differences versus NP7 (*)
and NP10 (+) are indicated (*p < 0.05, **p < 0.01, +++p < 0.001).
Figure 5. Activation of BMDC by adjuvant-loaded micelles. BMDC were
incubated in parallel with different NP formulations (each 15 µg mL−1;
see Table 3) and free L18-MDP or L18-MDP in nanoparticle MP6 (both
0, 15 µg mL−1). On the next day, expression of surface activation markers
was assessed by flow cytometry. Data denote the mean fluorescence
intensity SD of two independent experiments each and are given as
fold of untreated BMDC (see also Figure S37, Supporting Information).
We incubated BMDCs in vitro with NP1 (as reference), free
L18-MDP and NP6 loaded with the same amount of L18-MDP
and characterized their activation states by detecting surface
expression of the antigen presenting receptor MHCII and the
main costimulatory receptor CD86. The results are presented
in Figure 5. At first, non-loaded micelles (NP1) did not activate
BMDCs. The addition of free L18-MDP to activate BMDCs led
to an upregulation of MHCII and CD86 expression (Figure 5).
However, the addition of DC-targeting NP6 micelles with
encapsulated L18-MDP mediated even stronger DC activation.
Therefore, micelles engineered to cotarget CD206 and CD209
are suitable for the delivery of lipophilic adjuvants to stimulate
APC.
3
assessed by incorporation of H-thymidine applied on day 2
of coculture. As expected, BMDC preincubated with targeting
micelles that lacked antigen and adjuvant (NP4) induced no
T cell proliferation (Figure 6). BMDC that were preincubated
with micelles decorated with antigen, but lacking both targeting
moieties and adjuvant (NP7), facilitated moderate CD8+ T cell
proliferation. Antigen-decorated micelles that actively targeted
BMDC in addition (NP10), mediated stronger T cell prolifera-
tion. However, BMDC preincubated with trifunctional micelles
that contained L18-MDP as well (NP11), induced CD8+ T cell
proliferation to the highest extent.
Altogether, these results indicate that trifunctionalized
micelles that target APC via CD206 and CD209, contain an
antigenic peptide, and codeliver an encapsulated (lipophilic)
APC-stimulating adjuvant may be suitable to induce antigen-
specific T cell responses as required to eliminate, for example,
virus-infected as well as tumor cells.
3.3.3. Proliferation
So far, we have demonstrated that APC-targeting micelles were
efficiently engaged by BMDC, and that the encapsulated adju-
vant L18-MDP facilitated their activation. Next, we asked for the
capability of accordingly functionalized micelles to transfer an
antigen to BMDC. After cellular uptake of the micelle and intra-
cellular release of the antigen from the micelle, the antigen
may be presented on the cell surface in the context of MHC
receptors to T cells. A T cell whose T cell receptor binds the
MHC/antigen complex at sufficient affinity is activated if the
DC provides sufficient costimulatory signals conferred by
receptors like CD86. To this end, micelles were decorated with
a model peptide antigen (SIINFEKL) derived from ovalbu-
mine. This peptide is specifically recognized by all CD8+ T cells
derived from transgenic OT-I mice. BMDC were incubated with
differentially functionalized micelles overnight. After washing,
BMDC were serially diluted and cocultured with OT-I CD8+
T cells. The extent of antigen-specific T cell proliferation was
4. Conclusions
In this work, we introduced a versatile HPMA-based block-
copolymer system for the activation of the immune system.
These amphiphilic block copolymers self-assemble into poly-
meric micelles which can be stabilized by photocrosslinking.
Very importantly, these core-crosslinked micelles show very
little unspecific binding if they are not functionalized with man-
nose or trimannose as targeting units. They also do not activate
dendritic cells. The resulting core-crosslinked micelles could
be loaded with the lypophilic adjuvant L18-MDP. The drug
delivery systems showed a carbohydrate-dependent binding and
induced activation of DCs. Additionally, DCs incubated with
©
1800481 (9 of 10)
Macromol. Biosci. 2019, 1800481
2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.mbs-journal.de
[11] C. C. Preston, E. L. Goode, L. C. Hartmann, K. R. Kalli,
K. L. Knutson, Immunotherapy 2011, 3, 539.
[12] C. Liang, L. Xu, G. Song, Z. Liu, Chem. Soc. Rev. 2016, 45, 6250.
[13] Q. Zhou, L. Zhang, H. Wu, Nanotechnol. Rev. 2017, 6, 473.
[14] B. L. Narendra, K. E. Reddy, S. Shantikumar, S. Ramakrishna,
Inflammation Res. 2013, 62, 823.
micelle-conjugated peptide antigen subsequently induced a
CD8+ T cell response. The extent of T cell activation was highest
when using targeting micelles that codelivered antigen and
adjuvant. Our results show that this versatile polymer system is
a promising platform for new polymer-based vaccines.
[15] P. Kalinski, Curr. Opin. Investig. Drugs 2009, 10, 526.
[16] J. M. den Haan, S. M. Lehar, M. J. Bevan, J. Exp. Med. 2000, 192,
1685.
[17] K. Palucka, J. Banchereau, Nat. Rev. Cancer 2012, 12, 265.
[18] J. Rauen, C. Kreer, A. Paillard, S. Van Duikeren, W. E. Benckhuijsen,
M. G. Camps, A. R. P. M. Valentijn, F. Ossendorp, J. W. Drijfhout,
R. Arens, S. Burgdorf, PLoS One 2014, 9, e103755.
[19] S. Burgdorf, V. Lukacs-Kornek, C. Kurts, J. Immunol. 2006, 176,
6770.
Supporting Information
Supporting Information is available from the Wiley Online Library or
from the author.
Acknowledgements
[20] J. den Dunnen, S. I. Gringhuis, T. B. H. Geijtenbeek, Cancer
Immunol. Immunother. 2009, 58, 1149.
The authors acknowledge funding by the DFG (SFB1066).
[21] N. Mohr, C. Kappel, S. Kramer, M. Bros, S. Grabbe, R. Zentel,
Nanomedicine 2016, 11, 2679.
[22] A. Holla, A. Skerra, Protein Eng., Des. Sel. 2011, 24, 659.
[23] T. Satoh, S. Akira, Microbiol. Spectr. 2016, 4, 1.
[24] T. A. Kufer, G. Nigro, P. J. Sansonetti, Microbiol. Spectr. 2016, 4, 1.
[25] L. Nuhn, M. Barz, R. Zentel, Macromol. Biosci. 2014, 14, 607.
[26] L. Nuhn, L. Kaps, M. Diken, D. Schuppan, R. Zentel, Macromol.
Rapid Commun. 2016, 37, 924.
[27] S. Kramer, K. O. Kim, R. Zentel, Macromol. Chem. Phys. 2017, 218,
1700113.
[28] M. Eberhardt, R. Mruk, R. Zentel, P. Théato, Eur. Polym. J. 2005, 41,
1569.
[29] K. E. Moog, M. Barz, M. Bartneck, F. Beceren-braun, N. Mohr,
Z. Wu, L. Braun, J. Dernedde, E. A. Liehn, F. Tacke, T. Lammers,
H. Kunz, R. Zentel, Angew. Chem., Int. Ed. 2017, 56, 1416.
[30] M. Schmidt, in Dynamic Light Scattering, The Method and Some
Applications (Ed: W. Brown), Clarendon Press, Oxford 1993.
[31] K. Fischer, M. Schmidt, Biomaterials 2016, 98, 79.
[32] K. A. Hogquist, S. C. Jameson, W. R. Heath, J. L. Howard, M.
J. Bevan, F. R. Carbone, Cell 1994, 76, 17.
[33] M. Talelli, M. Barz, C. J. F. Rijcken, F. Kiessling, W. E. Hennink,
T. Lammers, Nano Today 2015, 10, 93.
[34] N. Mohr, M. Barz, R. Forst, R. Zentel, Macromol. Rapid Commun.
2014, 35, 1522.
Conflict of Interest
The authors declare no conflict of interest.
Keywords
HPMA block copolymers, micelles, targeting, trimannose, vaccines
Received: December 21, 2018
Revised: March 22, 2019
Published online:
[1] R. Duncan, Nat. Rev. Drug Discovery 2003, 2, 347.
[2] H. Ringsdorf, J. Polym. Sci. 1975, 153, 135.
[3] Y. Matsumura, K. Kataoka, Cancer Sci. 2009, 100, 572.
[4] A. Kakkar, G. Traverso, O. C. Farokhzad, R. Weissleder, R. Langer,
Nat. Publ. Gr. 2017, 1, 1.
[5] T. Sun, Y. S. Zhang, B. Pang, D. C. Hyun, M. Yang, Y. Xia, Angew.
Chem. Int. Ed. 2014, 53, 12320.
[6] Y. Matsumura, H. Maeda, Cancer Res. 1986, 46, 6387.
[7] H. Maeda, H. Nakamura, J. Fang, Adv. Drug Delivery Rev. 2013, 65,
71.
[35] M. Longmire, P. L. Choyke, H. Kobayashi, Nanomedicine 2008, 3, 703.
[36] M. Weilbächer, M. Allmeroth, M. Hemmelmann, S. Ritz,
V. Mailänder, T. Bopp, M. Barz, R. Zentel, C. Becker, J. Biomed.
Nanotechnol. 2013, 9, 1.
[8] S. Grabbe, K. Landfester, D. Schuppan, M. Barz, R. Zentel,
Nanomedicine 2016, 11, 2621.
[37] B. Sedaghat, R. Stephenson, I. Toth, Curr. Med. Chem. 2014, 21,
3405.
[9] X. Zang, X. Zhao, H. Hu, M. Qiao, Y. Deng, D. Chen, Eur. J. Pharm.
Biopharm. 2017, 115, 243.
[38] M. Colmenares, A. Puig-Kröger, O. M. Pello, A. L. Corbí, L. Rivas,
J. Biol. Chem. 2002, 277, 36766.
[10] D. M. Smith, J. K. Simon, J. R. Baker, Nat. Rev. Immunol. 2013, 13,
592.
[39] S. Bauer, H. Wagner, Curr. Top. Microbiol. Immunol. 2002, 270, 145.
[40] T. E. Kristensen, Beilstein J. Org. Chem. 2015, 11, 446.
©
1800481 (10 of 10)
Macromol. Biosci. 2019, 1800481
2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim