Mannosylated Poly(ethylene imine) Copolymers Enhance saRNA Uptake and Expression in Human Skin Explants

Article abstract of DOI:10.1021/acs.biomac.0c00445

Messenger RNA (mRNA) is a promising platform for both vaccines and therapeutics, and self-amplifying RNA (saRNA) is particularly advantageous, as it enables higher protein expression and dose minimization. Here, we present a delivery platform for targeted delivery of saRNA using mannosylated poly(ethylene imine) (PEI) enabled by the host-guest interaction between cyclodextrin and adamantane. We show that the host-guest complexation does not interfere with the electrostatic interaction with saRNA and observed that increasing the degree of mannosylation inhibited transfection efficiency in vitro, but enhanced the number of cells expressing GFP by 8-fold in human skin explants. Besides, increasing the ratio of glycopolymer to saRNA also enhanced the percentage of transfected cells ex vivo. We identified that these mannosylated PEIs specifically increased protein expression in the epithelial cells resident in human skin in a mannose-dependent manner. This platform is promising for further study of glycosylation of PEI and targeted saRNA delivery.

Full text of DOI:10.1021/acs.biomac.0c00445

Subscriber access provided by Bibliothèque de l'Université Paris-Sud
Article
Mannosylated poly(ethylene imine) copolymers enhance
saRNA uptake and expression in human skin explants
Anna K. Blakney, Yamin Abdouni, Gokhan Yilmaz, Renjie Liu, Paul
F. McKay, Clément R Bouton, Robin J Shattock, and C. Remzi Becer
Biomacromolecules, Just Accepted Manuscript • DOI: 10.1021/acs.biomac.0c00445 • Publication Date (Web): 06 Apr 2020
Downloaded from pubs.acs.org on April 9, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 32
Biomacromolecules
1
2
3
4
5
6
7
8
9
Mannosylated poly(ethylene imine) copolymers enhance
saRNA uptake and expression in human skin explants
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Anna K. Blakney1,§, Yamin Abdouni2,§, Gokhan Yilmaz3, Renjie Liu2, Paul F. McKay1, Clément
R. Bouton1, Robin J. Shattock1*, C. Remzi Becer2,4*
1 Division of Infectious Diseases, Imperial College London, Norfolk Place, London W21PG,
United Kingdom
2 School of Engineering and Materials Science, Queen Mary University of London, London,
E1 4NS, United Kingdom
3 School of Pharmacy, University of Nottingham, Nottingham, NG7 2RD, UK
4 Department of Chemistry, University of Warwick, Coventry, CV4 7AL, United Kingdom
§ These authors contributed equally to this work
* Correspondence: remzi.becer@warwick.ac.uk and r.shattock@imperial.ac.uk
KEYWORDS
Polyethylene imine, mannose, self-amplifying RNA, gene delivery, skin explants
1
ACS Paragon Plus Environment

Biomacromolecules
Page 2 of 32
1
2
3
4
5
6
7
ABSTRACT
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Messenger RNA (mRNA) is a promising platform for both vaccines and therapeutics, and self-
amplifying RNA (saRNA) is particularly advantageous as it enables higher protein expression
and dose minimization. Here, we present a delivery platform for targeted delivery of saRNA
using mannosylated poly(ethylene imine) (PEI) enabled by the host-guest interaction between
cyclodextrin and adamantane. We show that the host-guest complexation does not interfere
with the electrostatic interaction with saRNA, and observed that increasing the degree of
mannosylation inhibited transfection efficiency in vitro but enhanced the number of cells
expressing GFP by 8-fold in human skin explants. Besides, increasing the ratio of
glycopolymer to saRNA also enhanced the percentage of transfected cells ex vivo. We
identified that these mannosylated PEIs specifically increased protein expression in the
epithelial cells resident in human skin in a mannose-dependent manner. This platform is
promising for further study of glycosylation of PEI and targeted saRNA delivery.
2
ACS Paragon Plus Environment

Page 3 of 32
Biomacromolecules
1
2
3
4
1. INTRODUCTION
5
6
7
8
9
Recent advances and investment in RNA technology has enabled messenger RNA
(mRNA) to become a clinically viable platform for both vaccines and protein replacement
therapeutics. Self-amplifying mRNA (saRNA) has emerged as a next-generation approach, and
has several advantages compared to both mRNA and plasmid DNA (pDNA). Because saRNA
vectors are derived from the alphavirus genome,1 they are able to self-replicate in the
cytoplasm, resulting in amplification of the delivered dose of RNA and a higher magnitude and
duration of protein expression than mRNA.2-4 Compared to pDNA, saRNA is a minimal genetic
vector and does not pose the risk of integration or insertional mutagenesis.5 While a number of
mRNA vaccines and therapeutics are currently being tested in the clinic,6 there have not yet
been any non-viral RNA replicons tested in human clinical trials.7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
saRNA has previously been formulated with a variety of delivery platforms, including
lipid nanoparticles (LNPs),8, 9 a cationic nanoemulsion,3 cationic polymers10, 11 and ionizable
dendrimers.12 These formulations are not tailored for targeting of certain cell populations, but
rather increased overall cellular uptake and expression of the saRNA. Liang et al. previously
observed that while neutrophils, monocytes and dendritic cells infiltrate the injection site and
take up the RNA, it was mainly monocytes and dendritic cells that translated mRNA formulated
in LNPs.13 Both siRNA and mRNA has previously been targeted to leukocytes using the
Anchored Secondary scFv Enabling Targeting (ASSET) platform, in which LNPs are coated
in monoclonal antibodies to target specific leukocyte subsets.14 Furthermore, siRNA has been
directly conjugated to a synthetic triantennary N-acetylgalactosamine (GalNAc)-based ligand
that directly targets hepatocytes in vivo.15 In this study, we sought a delivery platform that
enabled tailoring of glycosylation without the use of expensive monoclonal antibodies or direct
3
ACS Paragon Plus Environment

Biomacromolecules
Page 4 of 32
1
2
3
4
5
6
conjugation to saRNA, which is much larger in size than siRNA and unlikely to be taken up
by cells without complexation.
7
8
9
Host-guest interactions between cyclodextrin (CD) and adamantane (Ad) have been
previously used as a gene delivery platform for intravenous delivery of pDNA, wherein
polyethylene glycol (PEG) was conjugated to adamantane in order to reduce toxicity of a
poly(ethylene imine) (PEI) formulation.16 CD and Ad are known to form a specific and stable
complex in aqueous environments through the interaction between adamantane and the
hydrophobic cavity of CD.17-19 Glycosylation of cyclodextrins has been performed previously,
and allows for a facile approach for attaching a variety of glycan groups.19, 20 Given the ease of
chemistry and biocompatibility of CD-Ad complexes, we chose this host-guest pair as a
platform for glycosylation of PEI as a targeted delivery vehicle for saRNA.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Herein, we have developed a mannosylated PEI complex enabled by the host-guest
interaction between CD and Ad. We designed and synthesized a library of PEI polymers with
varying degrees of mannosylation. We then characterized the polymers and the polyplexes
formed when complexed with saRNA for size, charge and transfection efficiency in vitro. After
identifying the optimal ratio of PEI to saRNA, we then tested these formulations ex vivo in a
clinically relevant human skin explant model to characterize the transfection efficiency. Finally,
we observed how the degree of mannosylation and ratio of polymer to saRNA affected cellular
expression and identify of which cellular subsets are targeted.
2. MATERIALS AND METHODS
2.1.
Materials
PEI MAX (Transfection grade linear polyethylenimine hydrochloride, MW 40,000) was
purchased from Polysciences, Inc. Dry triethylamine (TEA) (≥99.5%) equipped with septum,
4
ACS Paragon Plus Environment

Page 5 of 32
Biomacromolecules
1
2
3
4
5
6
7
8
9
1-adamantane carbonyl chloride and CuBr2 were purchased from Sigma Aldrich and used as
received. Tris(2-(dimethylamino)ethyl)amine (Me6TREN) was synthesized according to
literature procedures and stored at 4°C prior to use. Cyclodextrin initiator, mannose
glycomonomer and heptamannose β-cyclodextrin (CD-Man7) were synthesized as previously
reported and stored at -20 °C prior to use.21 22 All other reagents and solvents were purchased
from Sigma Aldrich at the highest purity available and used without further purification unless
stated otherwise.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2.2.
Instrument and Analysis
Nuclear Magnetic Resonance (NMR)
Proton (1H) Nuclear Magnetic Resonance (1H-NMR) spectra were recorded on a 400
MHz Bruker Avance III spectrometer using DMSO-d6, CDCl3, MeOD-d4 or D2O as the solvent
at 300K. 2D Nuclear Overhauser Effect SpectroscopY (NOESY) NMR experiments were
performed on a 600 MHz Bruker Avance NEO spectrometer in D2O at a temperature of 303 K
using the states-TPPI method with a 5ms Z-gradient spoil pulse in the mixing time and zero-
quantum suppression using the method of M.J. Thrippleton & J.Keeler.23 Mixing time was set
to 0.3 s, spectra were recorded using 20 scans per t1 increment and the spectral width was set
to 8 x 8 ppm.
Dynamic Light Scattering (DLS)
The hydrodynamic diameters (Dh, the volume weight diameter of the distribution)
evolution were determined characterized by a Malvern Zetasizer Nano ZS instrument equipped
with a He−Ne laser at 633 nm. DLS measurements were performed by dissolving polymer
samples at 1 mg/ml in deionized water and all the samples were passed through a 0.22 µm
nylon filter before measurement. For complex samples, polymers were dissolved separately in
deionized water and mixed together at different molar ratios. Then the samples were stirred
5
ACS Paragon Plus Environment

Biomacromolecules
Page 6 of 32
1
2
3
4
5
6
overnight at room temperature and filtered using a 0.22 nylon filter before analysis. All the
samples were measured three times at 25 °C.
7
8
9
In vitro transfection and luciferase assay
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
HEK 293T.17 cells (ATCC, USA) were plated at a density of 50,000 cells/well 48 hours
prior to transfection. The polyplexes were added to each well in a total volume of 100 μL with
a total dose of 100 ng of RNA in 20 mM HEPES with 5% glucose with n=3. The cells were
then incubated with the polyplexes for 4 hours, and then the media was replaced with 100 μL
of complete Dulbecco’s Modified Eagle’s Medium (cDMEM) (with 10% fetal bovine serum
(FBS), 5 mg/mL L-glutamine, 5 mg/mL penicillin streptomycin (ThermoFisher, UK)). After
24 hours, 50 uL of media was removed and 50 uL of ONE-Glo™ luciferase substrace
(Promega, UK) was added, and the total 100 μL was transferred to a white 96-well plate and
analyzed on a FLUOstar Omega plate reader (BMG LABTECH, UK) with a gain of 4000. The
average of the media only wells were subtracted from each sample measurement.
Human skin explant culture and digestion
Surgically resected specimens of human skin tissue were collected at Charing Cross
Hospital, Imperial College London, UK. All tissues were collected after receiving signed,
informed consent from all patients under protocols approved by the Local Research Ethics
Committee. The tissue was obtained from patients undergoing elective abdominoplasty, breast
reduction or mastectomy surgeries. Tissue was refrigerated until arrival in the laboratory. The
subcutaneous layer of fat was completely removed, and the remaining skin layers were trimmed
into ~1 cm2 sections. Explants were cultured in 10 mL of cDMEM in a petri dish at 37 °C and
5% CO2, and the media was refreshed daily.
Explants were injected with 2 μg of saRNA in a volume of 50 μL intradermally (ID)
using a Micro-Fine Demi 0.3 mL syringe (Becton Dickinson, UK). After 3 days, skin was
6
ACS Paragon Plus Environment

Page 7 of 32
Biomacromolecules
1
2
3
4
5
6
7
8
9
digested as previously described.9 Briefly, explants were minced well with scissors and
incubated in 2 mL DMEM supplemented with 1 mg/mL collagenase P (Sigma, UK) and 5
mg/mL dispase II (Sigma, UK) for 4 hours at 37°C on a rotational shaker. Digests were then
filtered through a 70 μm cell strainer and centrifuged for 5 min at 1750 RPM. Cells were
resuspended in 100 μL of FACS buffer (PBS + 2.5% FBS) and stained with 100 μL of Aqua
Live/Dead Stain (ThermoFisher, UK) diluted 1:400 in FACS buffer for 20 min on ice. Cells
were then washed with 1 mL of FACS buffer, centrifuged at 1750 RPM for 5 min and stained
with a panel of antibodies (Supplementary Table 1) to identify cellular phenotypes for 30
minutes. Cells were then washed with 1 mL of FACS buffer, centrifuged at 1750 RPM for 5
minutes and resuspended in 250 μL of PBS. Cells were fixed with 250 μL of 3%
paraformaldehyde for a total concentration of 1.5% and refrigerated until flow cytometry
analysis.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Flow cytometry analysis
Single cell suspensions were analyzed on a LSRFortessa™ (BD Biosciences, UK)
using FACSDiva software (BD Biosciences, UK) with 100,000 acquired events. Gating was
performed as previously described.9 GFP+ cells and phenotypes were quantified using FlowJo
Version 10 (FlowJo LLC, Oregon, USA).
Statistical analysis
Graphs and statistical analysis of in vitro and ex vivo data were prepared in GraphPad
Prism, version 8.0. Statistical analysis was performed using a two-tailed t test or a one way
ANOVA adjusted for multiple comparisons with α=0.05 used to indicate significance.
2.3.
Methods
General procedure for adaPEI synthesis
7
ACS Paragon Plus Environment

Biomacromolecules
Page 8 of 32
1
2
3
4
5
6
7
8
9
Linear polyethylene imine hydrochloride 40 kDa (PEI, 100 mg, 2.5 x 10-6 mol) was
suspended in 40 mL dry CHCl3 in a 100 mL RBF under Ar, equipped with stirring bar and
sonicated for 30 min. Subsequently the suspension was stirred and 1 mL of dry TEA was added.
Afterwards, the suspension was sonicated for 30 min until a fine suspension was achieved. A
solution of adamantane carbonyl chloride (50 mg, 2.515 x 10-4 mol, 0.2 eq. per repeating unit)
in 10 mL dry CHCl3 was prepared and subsequently added to the suspension. The mixture was
allowed to stir overnight at ambient temperature. After the reaction the suspension was filtered
over the MilliPore and an NMR sample was taken in D2O. Subsequently the filtered residue
was dissolved in 10 mL H2O to which 1 mL of a 32% HCl solution in water was added. The
solution was subsequently precipitated in acetone and dried under vacuum. An NMR sample
was taken in D2O and average amount of adamantanes per chain were calculated by comparing
the CH2- peak to the amount of adamantane protons. Quantities for synthesis of other adaPEIs
can be found in Table S1 and S2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
SET-LRP polymerization of CD-p(Man8)7
A Schlenk tube was charged with CD7-initiator (10 mg, 2696.88 g/mol, 3.71 µmol),
mannose glycomonomer (291 mg, 373.36 g/mol, 779 µmol, 30 eq. per initiating site),
Me6TREN (1.32 µL, 4.93 µmol, 0.19 eq. per initiating site), CuBr2 (232 µg, 1.04 µmol, 1.04
eq. per initiating site) in DMSO (2 mL), sealed with a rubber septum and subsequently degassed
by gentle bubbling of Ar gas for 15 min. The polymerization was then started by addition of
pre-activated Cu(0) wire (5 cm) wrapped around a stirring bar under a positive Ar pressure and
quickly sealed again and the reaction mixture was allowed to polymerize for 1 h at 25˚C.
Sampling was carried out using a degassed syringe to check the conversion of mannose
glycomonomer. NMR samples was dissolved in DMSO-d6 and conversion was determined as
27.5% (8.3 monomers per arm) by comparing the triazole peak to the vinyl protons. After
polymerization the glycopolymer CD-p(Man8)7 was dialysed against water to remove excess
8
ACS Paragon Plus Environment

Page 9 of 32
Biomacromolecules
1
2
3
4
glycomonomer and further impurities. Molecular weight of the polymer was then determined
5
6
via 1H NMR and was revealed to be 24.5 kDa on average.
7
8
9
Synthesis of glycoPEI: Complexation of adaPEI with CD-Man7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A solution of adaPEI2 (30.2 mg, 41088 g/mol, 0.734 µmol) and CD-Man7 (17.9 mg,
2836.53 g/mol, 6.26 µmol, 8.58 eq. per polymer chain) was prepared in 10 mL H2O amounting
to a 1 / 1 ratio of cyclodextrin derivative per adamantane and sonicated until the solution
became clear. Subsequently the solution was transferred to a 20 mL glass vial and freeze-dried.
An NMR sample was made by dissolving 10 mg in D2O for Nuclear Overhauser Effect
Spectroscopy (NOESY). Quantities for synthesis of other adaPEIs can be found in Table S1.
Synthesis of glycoPEI: Complexation of adaPEI with CD-(pMan8)7
A solution of adaPEI1 (30.7 mg, 41390 g/mol, 0.741 µmol) and CD-(pMan8)7 (86 mg,
24499 g/mol, 3.51 µmol, 4.75 eq. per polymer chain) was prepared in 10 mL H2O amounting
to a 1 / 1 ratio of gycopolymer CD-(pMan8)7 per adamantane and sonicated until the solution
became clear. Subsequently the solution was transferred to a 20 mL glass vial and freeze-dried.
saRNA synthesis and purification
Self-amplifying RNA encoding the non-structural proteins (NSPs) from the
Venezuelan Equine Encephalitis Virus (VEEV) and either firefly luciferase (fLuc) or enhanced
green fluorescent protein (eGFP) was prepared using in vitro transcription. pDNA was
transformed into Eschericia coli and cultured in 50 mL of LB broth with 1 mg/mL carbenicillin
(Sigma Aldrich, UK) and isolated using a Plasmid Plus Maxiprep kit (QIAGEN, UK). pDNA
concentration and purifity were quantified on a NanoDrop One (ThermoFisher, UK) and then
linearized using MluI for 3 h at 37 °C. Co-transcriptionally capped saRNA, used for in vitro
experiments, was synthesized using 1 μg of linearized DNA template in a mMessage
mMachine reaction (Ambion, UK) according to the manufacturer’s protocol, and purified using
9
ACS Paragon Plus Environment

Biomacromolecules
Page 10 of 32
1
2
3
4
5
6
7
8
9
a MEGAclear column (Ambion, UK) according to the manufacturer’s protocol. For ex vivo
experiments, uncapped in vitro RNA transcripts were synthesized using 1 μg of linearized DNA
template in a MEGAScript reaction (Ambion, UK) according to the manufacturer’s protocol.
Transcripts were then purified by overnight LiCl precipitation at -20 °C, pelleted by
centrifugation at 14,000 RPM for 20 min at 4°C, washed with 70% EtOH, centrifuged at 14,000
RPM for 5 min at 4°C, and then resuspended in UltraPure H2O. Purified transcripts were then
capped using the ScriptCap™ and m7G Capping System (CellScript, Madison, WI, USA) and
ScriptCap™ 2’-O-Methyltransferase Kit (CellScript, Madison, WI, USA) simultaneously
according to the manufacturer’s protocol. Capped transcripts were then purified again by LiCl
precipitation and stored at -80 °C in a buffer of 10 mM HEPES with 100 mg/mL trehalose until
use.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Particle complexation and characterization
Stock solutions of glycopolymers were prepared at a concentration of 5 mg/mL in
ultrapure H2O and purified using a 0.22 μm syringe filter (Millipore, Sigma, UK). saRNA
complexes were prepared by mixing the RNA and polymer in 20 mM HEPES buffer (pH 7.4)
with 5% glucose, with a ratio of polymer to RNA of 5:1 (w/w) unless otherwise specified. The
solution was immediately vortexed for 30 seconds and then allowed to rest for 10 minutes prior
to use.
Polyplexes were prepared in a volume of 800 μL of 20 mM HEPES with 5% glucose
for particle size and charge analysis, and characterized on a Zetasizer NanoZS (Malvern
Instruments, UK) with Zetasizer 7.1 software (Malvern Instruments, UK) in a clear disposable
1 mL cuvette. The polyplexes were analyzed using the following settings: a material refractive
index of 1.529, absorbance of 0.010, dispersant viscosity of 0.8820 cP, refractive index of 1.330
10
ACS Paragon Plus Environment

Page 11 of 32
Biomacromolecules
1
2
3
4
5
6
and dielectric constant of 79. Each sample was analyzed three times for up to 100 runs or until
the measurement stabilized.
7
8
9
Synthesis of per-(6-deoxy-6-bromine)-β-cyclodextrin (β-CD-(Br)7)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Triphenylphosphine (Ph3P, 36.72 g, 140 mmol) was dissolved in anhydrous DMF (150
mL) under stirring and cooled down to 0 °C (Scheme S1). N-bromosuccinimide (NBS, 24.92
g, 140 mmol) was dissolved in anhydrous DMF (40 mL) and the solution was added dropwise
to the Ph3P solution under Ar atmosphere and then stirred at ambient temperature for 30 min.
β-Cyclodextrin (β-CD, 11,35 g, 10 mmol) (previously recrystalized three times from water and
dried in vacuum oven at 50 °C for two days) was dissolved in anhydrous DMF (150 mL). The
obtained Ph3P/NBS solution was then added dropwise to the β-cyclodextrin solution at ambient
temperature after which the solution temperature was increased to 80 °C. The mixed brown
solution was stirred under Ar atmosphere overnight at 80 °C. Afterwards MeOH (40 mL) was
added at ambient temperature and stirring was continued for 30 min. The reaction mixture was
then cooled to 0 °C and the pH was adjusted to 9 by adding sodium methoxide, while further
stirring for 1 h. The reaction mixture was then poured into stirred ice-water (4 L) resulting in a
fine precipitate which was filtered and washed with MeOH. Heptakis (6-deoxy-6-bromo)-β-
cyclodextrin was obtained as beige solids and dried under vacuum for 1 day. Yield: 11.32 g,
70%).
1H NMR (400 MHz, DMSO-d6, 298 K, ppm): δ =6.02 (d, 7 H, 6.7 Hz), 5.89 (d, 7 H, 1.9 Hz),
4.98 (d, 7 H, 3.4 Hz), 4.00 (d, 7 H, 9.8 Hz), 3.82 (t, 7 H, 9.3 Hz), 3.65 (m, 14 H), 3.38 (m, 14
H, overlap with H2O).
MALDI-TOF MS m/z: calculated for C42H63Br7O28K+: 1614.73; found, 1614.74.
11
ACS Paragon Plus Environment

Biomacromolecules
Page 12 of 32
1
2
3
4
Synthesis of per-(6-deoxy-6-azido)-β-cyclodextrin (β-CD-(N3)7)
5
6
7
8
9
Heptakis (6-deoxy-6-bromo)-β-cyclodextrin (10 g, 6.3 mmol) was dissolved in
anhydrous DMF (80 mL) and NaN3 (5.78 g, 88.8 mmol) (Scheme S2). The resulting
suspension was stirred at 70 °C under Ar for 36 h. The suspension was then allowed to cool
down and precipitated in 2 L of stirred ice-water. The precipitate was filtered, washed with
water and redissolved in DMF (20 mL) and precipitated in 1L of stirred ice-water. The
precipitate was filtered and washed with water and with little acetone. The resulting product
was a white solid (yield: 7.2 g, 86.5 %) and was dried under vacuum overnight.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1H NMR (400 MHz, DMSO-d6, 298 K, ppm): δ =5.90 (d, 7 H, 6.8 Hz), 5.75 (d, 7 H, 2 Hz),
4.91(d, 7 H, 3,4 Hz), 3.74 (m, 14 H), 3.59 (m, 14 H), 3.36 (m, 14 H, overlap with H2O).
MALDI-TOF MS m/z: calculated for C42H63N21O28K+: 1348.37; found, 1348.27.
Synthesis of per-6-thio-β-cyclodextrin (β-CD-(SH)7)
β-CD-(Br)7 (5 g, 3.17 mmol) and thiourea (2.5 g, 33.3 mmol) were dissolved in DMF
(50 mL) and the mixture was heated to 70 °C under argon atmosphere (Scheme S3). After 24
h, DMF was removed under reduced pressure and the obtained brown oil was dissolved in
water (200 mL). Sodium hydroxide (2.22 g, 55.5 mmol) was then added and the reaction
mixture was heated to a gentle reflux under nitrogen atmosphere. After 1 h, the resulting
suspension was acidified with aqueous KHSO4 forming a white precipitate which was then
filtered and washed thoroughly with water and dried under vacuum. Compound β-CD-SH was
recovered as white powder (yield: 3.2 g, 81%).
1H NMR (400 MHz, DMSO-d6, 298 K, ppm): δ =5.91 (d, 7 H, 6.8 Hz), 5.81 (d, 7 H, 2 Hz),
4.93(d, 7 H, 3.3 Hz), 3.68 (t, 7 H, 8.5 Hz), 3.61 (t, 7 H, 9.2 Hz), 3.29-3.40 (m, 14 H, overlap
with H2O), 3.19 (m, 7H), 2.75 (m, 7 H), 2.13 (t, 7 H, 8.3 Hz).
12
ACS Paragon Plus Environment

Page 13 of 32
Biomacromolecules
1
2
3
4
Synthesis of allyl 2-bromoisobutyrate
5
6
7
8
9
Allyl alcohol (16.2 mL, 16.42 g, 282 mmol) and triethylamine (47.3 mL, 34.33 g, 339
mmol) were dissolved in diethyl ether (150 mL) and cooled down to 0 °C in an ice-water bath
(Scheme S4). A solution of α-bromoisobutyryl bromide (BIBB) (27 mL, 50 g, 217 mmol) in
20 mL diethyl ether was added dropwise over a period of 20 min. The mixture was allowed to
stir for 1 h at 0 °C after which it was allowed to reach room temperature and stirring was
continued overnight. The solution was washed 3 x 50 mL 10% HCl solution, 3 x 50 mL 5%
NaOH solution, 3 x 50 mL water and subsequently dried over MgSO4. After evaporating the
solvent via rotary evaporation, the product was purified by flash chromatography on silica gel
using chloroform as an eluent affording a colourless oil. (yield: 76%, 23 g).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1H NMR (400 MHz, CDCl3, 298 K, ppm): δ = 5.9 (ddt, 1 H, 5.5 Hz, 10.6 Hz, 17.2 Hz), 5.35
(dq, 1 H, 1.5 Hz, 17.2 Hz), 5.24 (dq, 1 H, 1.3 Hz, 10.6 Hz), 4.63 (dt, 2 H, 1.4 Hz, 5.6 Hz), 1.91
(s, 6 H).
Synthesis of per-6-deoxy-6-(thiopropyl-2-bromo-2-methylpropanoate)-β-cyclodextrin
Per-6-thio-β-cyclodextrin (2.5 g, 2 mmol), and dithiothreitol (DTT, 618 mg, 4 mmol)
were dissolved in 40 mL anhydrous DMF under Ar and heated to 60 °C (Scheme S5). After
60 h de reaction mixture was allowed to cool down to room temperature and allyl 2-
bromoisobutyrate (14.53 g, 70 mmol), 2,2-dimethoxy-2-phenylacetophenone (DMPA, 179 mg,
7 mmol) were added to the reaction mixture and stirring was continued for 5 h under UV
irradiation (365 nm).
The solution was precipitated in 500 mL of methyl tert-butyl ether (MTBE) in ten 50
mL centrifuge tubes and centrifuged at 8000 rpm for 5 min. The solvent was decanted and all
precipitated fractions collected in two 50 mL centrifuge tubes and fresh MTBE was added,
mixed and centrifuged again. This procedure was repeated 4 times in order to remove DMF
13
ACS Paragon Plus Environment

Biomacromolecules
Page 14 of 32
1
2
3
4
5
6
and allyl 2-bromoisobutyrate. Subsequently the product was dried under vacuum, yielding a
fine beige solid. (3.7 g, yield: 68%).
7
8
9
1H NMR (400 MHz, DMSO-d6, 298 K, ppm): δ = 5.90 (d, 7 H, 5.6 Hz), 5.8 (m, 7 H), 4.85 (m,
7 H), 4.22 (t, 14 H, 5.2 Hz), 3.85 (m, 7 H), 3.57 (m, 7 H), 3.33 (m, 14 H), 3.09 (d, 7 H, 10.6
Hz), 2.82 (m, 7 H), 2.69 (m, 14 H), 1.90 (s, 56 H).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
MALDI-TOF MS m/z: calculated for C91H147Br7O42S7K+:2733.12; found: 2733.36.
Synthesis of 3-azido-propan-1-ol
3-bromopropan-1-ol (7 g, 50.35 mmol) was dissolved in a solution of acetone (250 mL)
and water (50 mL) along with sodium azide (1.6 eq., 5.56 g, 86.61 mmol) and refluxed
overnight at temperature of 70 °C (Scheme S6). The organic solvent was removed by rotary
evaporation. 50 mL water was added to the remaining water phase and was then extracted with
diethyl ether (3X50 mL). The resulting ether phase was then back extracted with water (50 mL)
and dried over magnesium sulphate. The organic solvent was removed by rotary evaporation.
The product was recovered as a colourless liquid and used directly (yield: 64%).
1H NMR (400 MHz, D2O, 298 K, ppm): δ =3.66 (t, 2 H, 6.3 Hz), 4.41 (t, 2 H, 6.8 Hz), 1.81
(quin, 2 H, 6.5 Hz).
13C NMR (100 MHz, D2O, 298 K, ppm): δ = 58.78 (O-CH2), 47.93 (CH2-N3), 30.49 (C-CH2-
C).
ESI-MS m/z: calcd for C6H9N3O2 (2M+H+), 311.1; found, 311.1.
Notice: Organic azide is very sensitive compound and it should be handled with great care.
After synthesis this intermediate was directly used for the next step reaction without further
purification. Long period storage even in the fridge is not recommended.
14
ACS Paragon Plus Environment

Page 15 of 32
Biomacromolecules
1
2
3
4
Synthesis of 3-azidopropyl acrylate
5
6
7
8
9
A solution of 3-azido-propan-1-ol (6.10 g, 60.3 mmol), TEA (8.5 mL, 84.5 mmol),
hydroquinone (30 mg) and anhydrous diethyl ether (200 mL) was cooled in an ice water bath
(Scheme S7). Acryloyl chloride (5.88 mL, 72.4 mmol) in 20 mL diethyl ether was added
dropwise into the solution. The mixture was stirred in the ice bath for 1 h and then at ambient
temperature overnight. The ammonium salts were removed by filtration and the residue was
extracted sequentially with aqueous solution of hydrochloric acid (10 v%, 3X50 mL), water
(2Х50 mL), 5 wt% aqueous NaOH (3Х50 mL) and water (2Х50 mL) and dried over
magnesium sulphate. The organic solvent was removed by rotary evaporation. The product
was recovered as a yellow liquid and used directly (yield: 45%).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1H NMR (400 MHz, CDCl3, 298 K, ppm): δ =6.42 (dd, 1 H, 1.4, 17.3 Hz), 6.12 (dd, 1 H,10.4,
17.3 Hz), 5.85 (dd, 1 H, 1.4, 10.4 Hz), 4.25 (t, 2 H, 6.2 Hz), 3.41 (t, 2 H, 6.7 Hz), 1.96 (quin, 2
H, 6.4 Hz).
Synthesis of 1-(2′-Propargyl) D-Mannose
1-(2′-Propargyl) D-Mannose was prepared according to the procedure reported by
Mukhopadhyay et al.54 A suspension solution of D-Mannose (12 g, 66.6 mmol), propargyl
alcohol (19.4 mL, 333 mmol) and H2SO4-silica (333 mg) was stirred at 65 °C overnight
(Scheme S8). After cooling to ambient temperature, the reaction mixture was transferred to a
silica gel column and eluted with CHCl3-MeOH (8:1) to remove the excess propargyl alcohol.
1-(2′-Propargyl) D-Mannose was obtained as a white solid after drying under vacuum (8 g,
yield: 55 %). 1-(2′-propargyl) D-Mannose was found as an anomeric mixture in a ratio of 10:1
(α/β).
15
ACS Paragon Plus Environment

Biomacromolecules
Page 16 of 32
1
2
3
4
5
6
7
8
9
1H NMR (400 MHz, CD3OD, 298 K, ppm): δ: 4.96 (d, 1H, 1.6 Hz), 4.27 (d, 2H, 2.5
Hz), 3.84 (dd, 1H, 2.3, 11.8 Hz), 3.79 (dd, 1H, 1.8, 3.1 Hz), 3.66 (m, 3H), 3.51 (m, 1H), 2.85
(t, 1 H, 2.4 Hz)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of D-Mannose glycomonomer
1-(2′-propargyl) D-mannose (2.46 g, 12.6 mmol) and 3-azidopropyl acrylate (2.85 g,
11.8 mmol) were dissolved in MeOH/H2O (2:1 vol/vol, 60 mL), aqueous solution of
CuSO4·5H2O (246 mg, 0.9 mmol) and (+)-sodium L-ascorbate (284 mg, 1.2 mmol) were added
into the reaction solution (). The reaction mixture was stirred at ambient temperature for 24 h
and then the methanol was removed under vacuum and residue mixture was freeze dried to
remove water. The purification of the obtained product was done by silica gel column
chromatography using dichloromethane-MeOH (8:1) as eluent. After the removing of solvent,
the product was obtained as white (1.62 g, yield: 58.2%).
1H NMR (D2O, 298 K, 400 MHz): δ =8.07, 8.06 (s, overlaped, 1 H, NCH=C), 6.37 (dd, J=1.8,
15.5 Hz), 6.36 (dd, J=1.6, 15.7 Hz) (anomeric 1 H, CH2=C), 6.14 (dd, J=10.4, 6.9 Hz), 6.13(dd,
J=10.4, 7.0 Hz) (anomeric, 1 H, CH2=CHC=O), 5.89 (dd, 1 H, J=1.5, 8.9 Hz, CH2=C), 4.70-
5.05 (m, CH2-OH, H-1 of mannose , overlap with H2O), 4.64 (d, 1 H, J=12.3 Hz, CH2-OH),
4.55 (t, 2 H, J=6.9 Hz, CH2-N), 4.19 (t, 2 H, J=6.0 Hz, C=O-O-CH2), 3.40-3.92 (m, H residues
of mannose), 2.30 (m, 2H, CH2-CH2-CH2) ppm.
13C NMR (D2O, 298 K, 400 MHz): δ =146.4 (C=O), 145.4 (N-CH=C), 131.9 (CH2=C), 129.2
(CH2=C), 125.6 (N-CH=C), 100.8 (β anomeric, C1 of mannose), 100.7 (α anomeric, C 1 of
mannose), 78.4,75.2, 75.0, 72.5, 72.3, 72.0, 68.6, 68.4 (carbons of anomeric mannose),
63.0(CH2-OH), 62.6 (C=O-O-CH2), 60.7 (C-CH2-O), 48.5 (CH2-CH2-N), 28.5(CH2-CH2-CH2)
ppm.
ESI-MS m/z: calcd for C15H23N3O8 (M+Na+), 396.1; found, 396.1.
16
ACS Paragon Plus Environment

Page 17 of 32
Biomacromolecules
1
2
3
4
Synthesis of heptamannose β-cyclodextrin (CD-(Man)7).
5
6
7
8
9
β-CD-(N3)7 (1.96 g, 1.5 mmol), 1-(2‘-propargyl)-D-Mannose (2.61 g, 12 mmol) were
dissolved in DMSO (20 mL) in a Schlenk tube (Scheme S10). Bipyridine (0.37 g, 0.0024
mmol) and CuBr (0.17 g, 0.0012 mmol) were added. The resulting mixture was evacuated and
filled with argon and 3 freeze-pump-thaw cycles were performed to eliminate oxygen from the
reaction mixture. The mixture was then allowed to stir at 50 °C for 24 h. After the reaction,
water was added to the reaction medium and the resulting mixture was dialysed against water.
After dialysis, the resulting clear solution was freeze dried.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1H NMR (DMSO-d6, 298 K, 400 MHz): δ = 7.95, 7.92 (s, overlapped, 7 H, NCH=C), 5.80-
6.10 (m, 14 H, OH-2, OH-3 of CD), 5.10 (s, 7 H, H-1), 3.00-5.00 (m, CD & mannose residues,
overlap with H2O) ppm.
3. RESULTS AND DISCUSSION
3.1 Preparation and Characterization of Mannosylated PEI Polymers
A library of different PEI based polymers were successfully modified to contain
appending adamantane units along the backbone. For this, commercially available PEI was
first deprotonated using triethylamine and subsequently reacted with 1-adamantane carbonyl
chloride via a simple substitution reaction (Figure 1A). The resulting adaPEI polymers were
1
protonated and furthermore characterized using H-NMR (Figure 1B) in order to determine
the average adamantane content per polymer chain together with the change in molecular
weight (Table 1). Protonation shifts PEI peak (-CH2-CH2-) from around 3.0 ppm to around 3.5
ppm. Percentages are defined as the percentage of monomer units along the polymer backbone.
1H- NMR analysis revealed that there is a huge discrepancy between the amount of adamantyl
groups added to the reaction and adamantyl actually found along the polymer chain.
17
ACS Paragon Plus Environment

Biomacromolecules
Page 18 of 32
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Chemical reaction scheme for the synthesis of adaPEI (4) and glycoPEI (6) (A); 1H
NMR characterization of the synthesized adaPEI1 (B).
Apart from a monodisperse heptamannose β-cyclodextrin (CD-Man7) previously
synthesized within the group, a cyclodextrin based star-shaped mannose polymer was also
prepared. Mannose monomer was synthesised as previously reported (analysis provided in ESI)
and subsequently used for polymerisation starting from a heptavalent β-cyclodextrin based
18
ACS Paragon Plus Environment

Page 19 of 32
Biomacromolecules
1
2
3
4
5
6
initiator.21, 22 The polymerisation proceeded via Single Electron Transfer-Living Radical
Polymerisation (SET-LRP) in DMSO for 1h and was followed via 1H-NMR.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Characteristics of the synthesized glycoPEI polymers before supramolecular
interaction and after the supramolecular interaction.
glycoPEI
% Adamantanea
% Adamantaneb
# Adamantanes
per chainc
Mn, Theo
Mn, Theo
adaPEId (Da)
glycoPEIe (Da)
PEI1
PEI2
PEI3
PEI4
PEI5
PEI6
PEI7
PEI
40
20
10
5
2.18
1.71
0.94
0.54
0.37
0.14
0.94
0
10.96
8.58
4.76
2.72
1.83
0.71
4.76
0
41390
41080
40600
40340
40230
40090
40600
40000
72560
65480
54090
48060
45440
42110
157110
-
2.5
1.25
10
0
aTheoretical amount of adamantane before the reaction; bThe amount of adamantane according to 1H NMR after the reaction;
cThe number of adamantane per chain according to 1H NMR; dThe molecular weight of the polymers according to 1H NMR
after the reaction; eThe molecular weight of the polymers according to 1H NMR after the complexation with CD.
The resulting adaPEIs were subsequently mannosylated in a supramolecular manner by
combining the adaPEIs with CD-Man7 or CD-(pMan8)7 in water. The resulting glycoPEIs were
then characterized via 1H-NMR, 2D NOESY NMR and DLS which confirmed the anticipated
host-guest complexation between the adamantyl groups and CD-Man7 as NOESY experiments
revealed cross-peaks between the signals at 4 – 4.3 ppm assigned to the inner protons of the
CD-Man7 cavity and the signals at 1.9 – 2.4 ppm assigned to the adamantane not present when
taken 2D NOESY from the respective pure products (Figure 2A). Additionally, as seen in
Figure 2B, the host-guest interaction between PEI1 and CD-Man7 was further confirmed by
DLS, which revealed that size and size distribution increased due to the attachment of
19
ACS Paragon Plus Environment

Biomacromolecules
Page 20 of 32
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. 2D NOESY-NMR spectrum of the adaPEI1 and CD-Man7, clearly showing cross
peaks between the signals at 4.0 – 4.3 ppm assigned to the inner protons of the CD-Man7 cavity
and the signals at 1.9 – 2.4 ppm assigned to the adamantane (A); DLS measurements of the
adaPEI1 and CD-Man7 before and after supramolecular interaction (B).
CD-Man7 on the backbone of PEI1. The mean hydrodynamic size of PEI1 increased
dramatically from 125 nm to 218 nm while distribution index increased from 0.34 to 0.58 after
the host-guest complexation. Besides, large aggregates of CD-Man7 disappeared after the
complexation, which also proved that the interaction of CD-Man7 with adamantane units on
the polymer backbone could eliminate the formation of interchain assemblies of aggregations.
20
ACS Paragon Plus Environment

Page 21 of 32
Biomacromolecules
1
2
3
4
5
6
The resulting solutions were furthermore freeze-dried, resulting in a fine powder which is
easily dissolvable for use in RNA transfection.
7
8
9
3.2 Preparation and Characterization of saRNA/ManPEI Polyplexes In Vitro
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Particle size and charge was characterized after complexation with saRNA (Figure 3).
All particles were found to be between 50-200 nm in size, with a slight trend of increasing size
with increasing degree of mannosylation. All particles were positively charged after saRNA
complexation, indicating that the host-guest interaction does not interfere with the cationic
charges or further condensation of saRNA molecules.
Figure 3. Particle size and zeta potential of PEI-Ad-CD-Man7/saRNA polyplexes as
determined by DLS. Z-average particle diameter (the volume weight diameter of the
distribution) (A) and zeta potential (B) of complexes prepared at a ratio of 5:1 polymer to RNA
(w/w). Bars represent mean ± standard deviation for n=3.
In order to investigate the effects of the ratio of PEI to saRNA and degree of
mannosylation on transfection efficiency in vitro, we prepared polyplexes with saRNA and PEI
polymers with and without mannosylation. We observed how the degree of mannosylation
affected transfection efficiency, using either unmodified PEI, PEI with varying amounts of
Man7 (PEI1-6) or PEI with polyMan7 (PEI7) (Figure 4A). Because changing the mass ratio of
21
ACS Paragon Plus Environment

Biomacromolecules
Page 22 of 32
1
2
3
4
5
6
7
8
9
polymer to saRNA changes the amount of available positively charges amines, we used a fixed
ratio of PEI to saRNA of 5:1 (w/w) and then normalized the molar amount of PEI in each
formulation for PEI1-7. We observed that PEI had the highest transfection efficiency, ~106 RLU,
whereas all of the mannosylated PEIs were lower, between 1-5x105 RLU. In addition,
increasing the degree of mannosylation decreased the transfection efficiency in vitro, as PEI4-
6 had the highest transfection efficiency of the mannosylated PEIs, and PEI1-3 had the lowest.
We hypothesize that this is due to steric hinderance cause by increasing degree of
mannosylation which may limit the access that saRNA has to the amine groups on PEI.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. In vitro transfection efficiency of PEI-Ad-CD-Man7 complexes with fLuc saRNA in
HEK293T.17 cells after 24 hours. Transfection efficiency with formulations normalized to the
molar amount of PEI in the complex (A); Transfection efficiency with formulations prepare by
varying ratio of polymer to RNA. Bars represent mean ± standard deviation for n=3 (B).
Furthermore, while HEK cells can be induced to express the mannose receptor,24 they
do not naturally express it, thus these experiments exhibit how the structure of the polyplexes
affects non-mannose mediated uptake. Furthermore, we observed similar transfection
efficiency between the 5:1 and 20:1 ratios of polymer to saRNA (w/w) in vitro (Figure 4B).
This is most likely due to glycopolymers being saturated with saRNA even at lower ratios of
polymer to RNA, which is supported by the Zetasizer data (Figure 3) wherein even at a ratio
of 5:1, the particles exhibit a positive charge. Thus, adding more polymer does not increase the
22
ACS Paragon Plus Environment

Page 23 of 32
Biomacromolecules
1
2
3
4
transfection efficiency. Overall, we observed that increasing the degree of mannosylation, but
5
6
not the ratio of mannosylated PEI to saRNA, decreased the transfection efficiency in vitro.
7
8
9
3.3 Transfection Efficiency of saRNA/ManPEI Polyplexes in Human Skin Explants.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Because in vitro transfection efficiency does not generally correlate well with in vivo
efficacy,25 we sought to test these glycopolymers in a clinically translational human skin
explant model. Human skin explants have previously been shown to be a viable model for
optimization of nucleic acid formulations,9 and contain many cell types with the mannose
receptor including dendritic cells, fibroblasts, and macrophages.26-30 We first prepared
formulations with either RNA alone, PEI or mannosylated PEI (PEI1-6) (Figure 5A) at a ratio
of 5:1 polymer to saRNA (w/w). We were surprised to observe that the polyplexes did not
enhance the percentage of eGFP+ in skin explants, even with unmodified PEI (Figure S6A).
We observed a similar effect for PEI with polyMan7- only ~1% of cells expressed GFP and
there was no observed benefit to naked RNA alone (Figure S6B). We then tested whether
increasing the ratio of PEI1 to saRNA had any effect on the percentage of cells expressing GFP
(Figure 5B). We observed that increasing the ratio of PEI1 to saRNA to 10:1 and 20:1 (w/w)
did indeed increased the number of GFP+ cells to 5% and 8%, with p=0.018 and 0.00038,
respectively. This enhancement is superior to previously studied LNP formulations.9
23
ACS Paragon Plus Environment

Biomacromolecules
Page 24 of 32
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Ex vivo transfection efficiency in human skin explants. Percentage of eGFP+
epithelial cells in human skin explants after ID injection of saRNA/ PEI-Ad-CD-Man7
complexes prepared at a ratio of 20:1 (w/w) after 72 hours in culture. Bars represent mean ±
standard deviation for n=3. * indicates significance of p<0.05 (A); Percentage of eGFP+ cells
in human skin explants after treatment with saRNA/PEI-Ad-CD-Man7 complexes after 72
hours in culture. The ratio of complexes to RNA was varied from 1:1 to 20:1 (w/w) of PEI1.
Bars represent the mean ± standard deviation for n=3. * indicates significance of p<0.05
compared to RNA only(B).
3.4 Impact of Ratio of ManPEI to saRNA on Ex Vivo Phenotypic Protein Expression
We then characterized which cells were expressing the saRNA using a flow cytometry
panel capable of identifying epithelial cells (CD45-), fibroblasts (CD90+), NK cells (CD56+),
leukocytes (CD45+), Langerhans cells (CD1a+), monocytes (CD14+) dendritic cells (CD11c+),
T cells (CD3+) and B cells (CD19+). As previously observed, the majority of cells that make
up the skin are epithelial cells, fibroblasts and dendritic cells, and leukocytes, Langerhans cells,
B cells, monocytes and T cells to a lesser extent (Figure 6A). While epithelial cells make up
~64% of the total cells, they make up only 16% of the cells expressing the saRNA alone.
However, when the saRNA was complexed with mannosylated PEI1 at a ratio of 20:1 (w/w),
24
ACS Paragon Plus Environment

Page 25 of 32
Biomacromolecules
1
2
3
4
5
6
7
8
9
it was expressed in ~33% of epithelial cells (Figure 6B). GFP was expressed in either a similar
or lesser percentage of the other cell types. Overall, we observed that increasing the ratio of
mannosylated PEI to saRNA increased the number of cells expressing saRNA in human skin
explants, and we identified that epithelial cells were specifically targeted by these polyplexes.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Phenotypic identity of eGFP+ cells in human skin explants alone (A) and after ID
injection of GFP saRNA/PEI-Ad-CD-Man7 complexes at varying ratios of PEI1 to saRNA (B).
Cells were identified using the following antibodies: epithelial cells (CD45-), fibroblasts
(CD90+), NK cells (CD56+), leukocytes (CD45+), Langerhans cells (CD1a+), monocytes
(CD14+), dendritic cells (CD11c+), T cells (CD3+) and B cells (CD19+).
3.5 Impact of Degree of Mannosylation on Phenotypic Expression of saRNA Ex Vivo
25
ACS Paragon Plus Environment

Biomacromolecules
Page 26 of 32
1
2
3
4
5
6
7
8
9
Given our observation that increasing the ratio of mannosylated PEI to saRNA enhanced the
number of cells expressing saRNA, we then studied whether the degree of mannosylation
affected cellular uptake and expression ex vivo (Figure 7). We prepared polyplexes at a fixed
ratio of 20:1 (w/w) of PEI to saRNA and again evaluated which cells were expressing GFP.
We observed that at a ratio of 20:1, the PEI formulations (both unmodified and mannosylated)
increased the percentage of GFP+ cells to ~8%. Increasing the degree of mannosylation had a
trend of increasing the percentage of epithelial cells expressing GFP (Figure 7), although only
the PEI1,2,4,6 groups were found to be statistically significantly higher. While the mannose
receptor is primarily known to be expressed by macrophages, dendritic cells, fibroblasts and
keratinocytes, it has previously been shown to be expressed by vaginal epithelial cells.31 It is
possible that human skin epithelial cells also express the mannose receptor, leading to increased
polyplex uptake and saRNA expression in these cells. However, in these studies we quantified
the percentage of cells expressing the saRNA, not the percentage of cellular uptake, so it is
possible that there is increased uptake into cells that are known to express the mannose receptor.
In the context of RNA vaccines, it has yet to be defined as to which cells are desired to express
the protein; we hypothesize that an increased protein expression will result in increased
immunogenicity. Overall, we show that increasing the degree of mannosylation increases
protein expression, specifically in epithelial cells of human skin explants.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
26
ACS Paragon Plus Environment

Page 27 of 32
Biomacromolecules
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. Phenotypic identity of eGFP+ cells in human skin explants after ID injection of
saRNA/PEI-Ad-CD-Man7 complexes prepared at a ratio of 20:1 (w/w) with PEI1-7 after 72
hours in culture. Cells were identified using the following antibodies: epithelial cells (CD45-),
fibroblasts (CD90+), NK cells (CD56+), leukocytes (CD45+), Langerhans cells (CD1a+),
monocytes (CD14+), dendritic cells (CD11c+), T cells (CD3+) and B cells (CD19+).
27
ACS Paragon Plus Environment

Biomacromolecules
Page 28 of 32
1
2
3
4
4. CONCLUSION
5
6
7
8
9
A library of mannosylated PEI polymers enabled by the host-guest interaction between
cyclodextrin and adamantane for targeted saRNA delivery was investigated. We show that
while increasing the degree of mannosylation stifles in vitro transfection efficiency, it enhances
the percentage of cells expressing the saRNA in human skin explants. Furthermore, it was
investigated that increasing the ratio of polymer to saRNA also enhanced the protein expression
ex vivo, which was specifically due to an increase in epithelial cell expression. Meanwhile,
increasing the degree of mannosylation also increased expression specifically in epithelial
cells. We believe that this platform, which enables glycosylation of PEI through host-guest
chemistry, is a highly clinically translational delivery vehicle and is dually useful for targeting
specific cell types for saRNA delivery and expression.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ASSOCIATED CONTENT
Supporting Information.
Experimental details for the synthesis of all compounds are provided in the ESI.
AUTHOR INFORMATION
Corresponding Author
Remzi.Becer@warwick.ac.uk and R.Shattock@imperial.ac.uk
Author Contributions
The manuscript was written through contributions of all authors. All authors have given
approval to the final version of the manuscript. ‡These authors contributed equally.
28
ACS Paragon Plus Environment

Page 29 of 32
Biomacromolecules
1
2
3
4
Acknowledgements
5
6
7
8
9
We gratefully acknowledge the surgeons (Elizabeth A. Dex and Judith E. Hunter) and their
nursing team at Charing Cross Hospital. We also acknowledge Dormeur Investment Services
Ltd for providing funds to purchase equipment used in these studies. AKB is supported by a
Whitaker Post-Doctoral Fellowship and a Marie Skłodowska Curie Individual Fellowship
funded by the European Commission H2020 (No. 794059). PFM, CRB, and RJS are funded
by the U.K. Department of Health and Social Care through the Future Vaccine Manufacturing
Hub through the Engineering and Physical Sciences Research Council (EPSRC, grant number:
EP/R013764/1). This project has received funding from the European Union’s Horizon 2020
research and innovation programme under the Marie Skłodowska-Curie grant agreement No
642083. RL acknowledge the China Scholarship Council (CSC) and Queen Mary, University
of London for funding this project.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5. REFERENCES
1.
Perri, S.; Greer, C. E.; Thudium, K.; Doe, B.; Legg, H.; Liu, H.; Romero, R. E.; Tang, Z.; Bin,
Q.; Dubensky, T. W.; Vajdy, M.; Otten, G. R.; Polo, J. M., An Alphavirus Replicon Particle Chimera
Derived from Venezuelan Equine Encephalitis and Sindbis Viruses Is a Potent Gene-Based Vaccine
Delivery Vector. J. Virol. 2003, 77, (19), 10394-10403.
2.
Chahal, J. S.; Khan, O. F.; Cooper, C. L.; McPartlan, J. S.; Tsosie, J. K.; Tilley, L. D.; Sidik, S.
M.; Lourido, S.; Langer, R.; Bavari, S.; Ploegh, H. L.; Anderson, D. G., Dendrimer-RNA nanoparticles
generate protective immunity against lethal Ebola, H1N1 influenza, and Toxoplasma gondii challenges
with a single dose. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, (29), E4133-E4142.
3.
Geall, A. J.; Verma, A.; Otten, G. R.; Shaw, C. A.; Hekele, A.; Banerjee, K.; Cu, Y.; Beard, C.
W.; Brito, L. A.; Krucker, T.; O’Hagan, D. T.; Singh, M.; Mason, P. W.; Valiante, M. N.; Dormitzer,
P. R.; Barnett, S. W.; Rappuoli, R.; Ulmer, J. B.; Mandl, C. W., Nonviral delivery of self-amplifying
RNA vaccines. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, (36), 14604-14609.
4.
Bogers, W. M.; Oostermeijer, H.; Mooij, P.; Koopman, G.; Verschoor, E. J.; Davis, D.; Ulmer,
J. B.; Brito, L. A.; Cu, Y.; Banerjee, K.; Otten, G. R.; Burke, B.; Dey, A.; Heeney, J. L.; Shen, X.;
Tomaras, G. D.; Labranche, C.; Montefiori, D. C.; Liao, H. X.; Haynes, B.; Geall, A. J.; Barnett, S. W.,
29
ACS Paragon Plus Environment