Glycosylated Reversible Addition-Fragmentation Chain Transfer Polymers with Varying Polyethylene Glycol Linkers Produce Different Short Interfering RNA Uptake, Gene Silencing, and Toxicity Profiles

Article abstract of DOI:10.1021/acs.biomac.7b01168

Achieving efficient and targeted delivery of short interfering (siRNA) is an important research challenge to overcome to render highly promising siRNA therapies clinically successful. Challenges exist in designing synthetic carriers for these RNAi constructs that provide protection against serum degradation, extended blood retention times, effective cellular uptake through a variety of uptake mechanisms, endosomal escape, and efficient cargo release. These challenges have resulted in a significant body of research and led to many important findings about the chemical composition and structural layout of the delivery vector for optimal gene silencing. The challenge of targeted delivery vectors remains, and strategies to take advantage of nature's self-selective cellular uptake mechanisms for specific organ cells, such as the liver, have enabled researchers to step closer to achieving this goal. In this work, we report the design, synthesis, and biological evaluation of a novel polymeric delivery vector incorporating galactose moieties to target hepatic cells through clathrin-mediated endocytosis at asialoglycoprotein receptors. An investigation into the density of carbohydrate functionality and its distance from the polymer backbone is conducted using reversible addition-fragmentation chain transfer polymerization and postpolymerization modification.

Full text of DOI:10.1021/acs.biomac.7b01168

Subscriber access provided by University of Florida | Smathers Libraries
Article
Glycosylated RAFT polymers with varying PEG linkers produce
different siRNA uptake, gene silencing and toxicity profiles
Elizabeth G.L. Williams, Oliver Earl Hutt, Tracey M. Hinton, Sophie C Larnaudie, Tam
Le, James M. MacDonald, Pathiraja Gunatillake, San H. Thang, and Peter J Duggan
Biomacromolecules, Just Accepted Manuscript • DOI: 10.1021/acs.biomac.7b01168 • Publication Date (Web): 23 Oct 2017
Downloaded from http://pubs.acs.org on October 23, 2017
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 free 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 accessible to all
readers and 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.
Biomacromolecules 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 41
Biomacromolecules
1
2
3
4
5
6
7
8
9
Glycosylated RAFT polymers with varying PEG
linkers produce different siRNA uptake, gene
silencing and toxicity profiles
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
Elizabeth G. L. Williams,a‡ Oliver E. Hutt,*a‡ Tracey M. Hinton,b Sophie C. Larnaudie,a‡ Tam
Le,a James M. MacDonald,a Pathiraja Gunatillake,a San H. Thang,a Peter J. Duggana,c
a CSIRO Manufacturing, Private Bag 10, Clayton South, VIC, 3169, Australia.
b CSIRO Health and Biosecurity, Port Arlington Rd, East Geelong, VIC 3219, Australia.
c School of Chemical and Physical Sciences, Flinders University, Adelaide, South Australia,
5042, Australia.
Supporting Information
The following files are available free of charge.
1H and 13C NMR data for sugarꢀPEGꢀazide click precursors 18, 19, 20, 30 and 31, ABA
triblock 11, and clicked polymers 32, 33, 34, 35 and 36 (PDF)).
AUTHOR INFORMATION
* Corresponding author: Oliver Hutt, Email: oliver.hutt@csiro.au
Author Contributions The manuscript was written through contributions of all authors. All
authors have given approval of the final version of the manuscript. ‡These authors
contributed equally.
ABSTRACT
Achieving efficient and targeted delivery of short interfering (siRNA) is an important
research challenge to overcome to render highly promising siRNA therapies clinically
successful. Challenges exist in designing synthetic carriers for these RNAi constructs that
ACS Paragon Plus Environment

Biomacromolecules
Page 2 of 41
1
2
3
4
5
6
7
8
9
provide protection against serum degradation, extended blood retention times, effective
cellular uptake through a variety of uptake mechanisms, endosomal escape and efficient
cargo release. These challenges have resulted in a significant body of research, and led to
many important findings about the chemical composition and structural layout of the delivery
vector for optimal gene silencing. The challenge of targeted delivery vectors remains, and
strategies to take advantage of nature’s selfꢀselective cellular uptake mechanisms for specific
organ cells, such as the liver, have enabled researchers to step closer to achieving this goal. In
this work we report the design, synthesis and biological evaluation of a novel polymeric
delivery vector, incorporating galactose moieties to target hepatic cells through clathrinꢀ
mediated endocytosis at asialoglycoprotein receptors. An investigation into the density of
carbohydrate functionality and their distance from the polymer backbone is conducted using
Reversible Addition Fragmentation chain Transfer (RAFT) polymerisation and postꢀ
polymerisation modification.
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
INTRODUCTION
In recent years, many disease interventions have focussed on the naturally occurring RNA
interference mechanism, which consists of specific gene suppression by the action of small
ACS Paragon Plus Environment

Page 3 of 41
Biomacromolecules
1
2
3
4
5
6
7
8
interfering RNA (siRNA) chains.1 The basis of this approach is the delivery of artificial
exogenous siRNA to cells to knockꢀdown targeted genes.2 As such, siRNA could be used in
the treatment of genetic disorders3, in cancer therapy,4 or as antivirals.5 However, the strong
9
6
negative charge and instability of naked RNA in vivo requires a suitable carrier for these
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
siRNAs. Viral carriers are easy to develop, but present problems such as immunogenicity or
oncogenicity, low nucleic acid loading capacity and limited control over the type of cells that
is transduced.7ꢀ8 Therefore, siRNA delivery has become a very active field of research.9
The liver is an important target for RNAiꢀbased therapies due to its role in the production and
secretion of serum proteins into the blood, potentially enabling a permanent therapeutic
response.10 The present study focuses on interventions for viral liver diseases (such as
hepatitis A, B, C and E) and seeks to couple a polymeric siRNA delivery vehicle with a
targeting strategy to take advantage of the presence of asialoglycoprotein receptors (ASGPꢀ
R) on mammalian liver cells.11
The asialoglycoprotein receptor (ASGPR) is a member of the large family of Ca2+ꢀdependent
lectins (Cꢀtype lectins) present in hepatocytes that remove desialylated proteins from the
blood serum. The ASGPRs recognise and clear glycoproteins with either terminal galactose
(Gal) or Nꢀacetyl galactosamine (GalNAc) through activation of clathrinꢀmediated
endocytosis. Upon encapsulation, the decrease in pH in the endosomal compartment is
sufficient to break the carbohydrateꢀreceptor interaction and the ASCGR is ultimately
returned to the surface within 15 minutes (Fig. 1).
ACS Paragon Plus Environment

Biomacromolecules
Page 4 of 41
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: Clathrinꢀmediated endocytosis, triggered by binding of Gal or GalNAc ligands at
the ASGP receptor.
The high level of expression of ASGPR in the liver and it’s specificity for Gal and GalNAc
hasled to the successful exploitation of this naturally occurring mechanism to target therapies,
including smallꢀmolecule cytotoxics for hepatic cancer. For example, the GalNAcꢀ
DoxorubicinꢀHPMA copolymers reported by Kopeček et al are retained in the human liver
for more than 24 hours compared with controls (Fig. 2A).12 Incorporation of enzymatically
cleavable tetrapeptide units enabled the targeted delivery of doxorubicin.
GalNAcꢀsiRNA conjugates developed by Alnylam also effectively target the liver and are
effective in targeting the genes involved in range of diseases including transthyretinꢀrelated
amlyoidosis, hemophilia, and porhyria (Fig. 2B).13 These triꢀantennary GalNAc recognition
elements were specifically designed to be 15ꢀ20 Å apart, within a tripod, to mimic the
triangular spatial arrangement of hepatic ASGP receptors, thereby maximising ligand
recognition and cellular uptake.14 Similarly, GalNAcꢀsiRNA dynamic polyconjugates, which
rely on a reversible succinimideꢀamide linkage, have shown efficacy against hepatitis B
(Figure 2C)15. More recently, a tetradentate GalNAc targeting group was also reported for
ASGPR (Figure 2D)16.
ACS Paragon Plus Environment

Page 5 of 41
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 2. Asiaglycoprotein receptor targeting therapeutic delivery systems.
Whilst triꢀantennary based GalNAc vectors have shown a much stronger affinity to ASGPR
compared to monovalent vectors, based on IC50 values, they often require a complex
synthesis.17ꢀ18 With the aim of developing synthetically simple delivery vectors, the Reineke
group have reported an AB diblock copolymer comprising a GalNAcꢀfunctionalised 2ꢀ
hydroxyethyl methacrylamide (HEMAA) and 2ꢀaminoethyl methacrylamide (AEMA)
monomers (Fig. 3) for the delivery of nanomedicine formulations, including plasmid DNA, to
the liver19. This delivery system incorporates a cationic block (AEMA), allowing the RNAiꢀ
based therapeutics to associate with the carrier polymer via an electrostatic interaction, rather
than the covalent delivery vectors designed by Alnylam. By using the Reversible Additionꢀ
Fragmentation Chain Transfer (RAFT) – a controlled free radical polymerisation process,8
Reineke and colleagues were able to interrogate optimum cationic carrier block length, whilst
ACS Paragon Plus Environment

Biomacromolecules
Page 6 of 41
1
2
3
4
5
6
keeping the targeting GalNAc block constant (Fig. 3A). A glucose control (Fig. 3B) was also
synthesised to evaluate targeting effects of the GalNAc polymers.
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 3: GalNAc and Glucose containing RAFT polymers prepared by the Reineke group.
ASGPꢀR are carbohydrateꢀbinding proteins (lectins) that bind βꢀgalactose, and galactosylated
polymers have been reported to show good hepatocyte specificity and higher transfection
efficiency than non galactoseꢀcontaining polymers.20
The multivalent GalNAc copolymers were found to provide stronger interactions with
ASGPRs at 5 orders of magnitude greater than monomeric GalNAc. The polymers were
assayed against HepG2 cells, derived from human hepatocellular carcinoma, and HeLa cells,
derived from human cervical adenocarcinoma. Lower HepG2 cell viability was observed for
the GalNAcꢀcontaining polymers at N:P (nitrogen (Polymer): phosphate (siRNA backbone))
ratios of 2.5 and 5. Transgene expression using gWizꢀluc plasmid polyplexes in HepG2 and
HeLa cells showed that GalNAc containing polymers promoted higher cellular uptake and
therefore transgene expression in HepG2 cells than in HeLa cells due to the presence of the
ASGPꢀreceptors on HepG2 cells. However, MAGꢀcontaining control polymers show similar
levels of gene expression in HeLa cells compared to HepG2 cells, and efficient gene delivery
ACS Paragon Plus Environment

Page 7 of 41
Biomacromolecules
1
2
3
4
5
6
7
8
9
is seen even at low N:P ratios. This result suggests that the cationic copolymer block or
excess free polymer may promote nonspecific cell surface interactions and nonꢀligand
dependent uptake mechanisms, in addition to the ASGPRꢀmediated uptake promoted by the
GalNAc block. Following tailꢀvein injection of fluorescently labelled p(MAGalNAc62ꢀbꢀ
AEMA33) and p(MAG46ꢀbꢀAEMA13) pDNA polyplexes, total fluorescence in all harvested
tissues (harvested 30 min postꢀinjection) was measured to determine the biodistribution of the
two vectors. p(MAGalNAc62ꢀbꢀAEMA33) was found mainly in the liver, accounting for 80%
of total injected fluorescence, whilst p(MAG46ꢀbꢀAEMA13) displayed 64% of total
fluorescence in the liver, with 20% of fluorescence accounted for by polymer in the kidneys.
Collectively, the in vitro and in vivo results show that the MAGalNAc polymer does provide
some targeting to the liver above the glucose control, however, the majority of the retention is
due to other MAGalNAcꢀindependent cellular uptake mechanisms.
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
Narain and coꢀworkers also used a RAFT diꢀblock copolymer approach incorporating a
cationic AEMA segment and an asialoglycoprotein targeting 2ꢀlactobionoamidoethyl
methacrylamide (LAEMA) segment (Fig. 4)21. Statistical copolymers were also synthesised
to understand toxicity profiles of concentrated cationic segments compared to the random
spread of these functionalities. These polymer vectors were used to condense DNA (βꢀ
galactosidase plasmid) rather than siRNA, with a focus on transfection to HepG2 cells for the
treatment of hepatocellular carcinoma.
ACS Paragon Plus Environment

Biomacromolecules
Page 8 of 41
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 4: Lactobionicꢀcontaining block and statistical copolymers for targeting DNA
plasmid delivery to HepG2 and Huh7.5 cells via endocytosis through the ASGPꢀR.
Their studies found that toxicity decreased (obtained by MTT assay in HepG2 cells) with
increasing carbohydrate content, which is expected due to lower toxic cationic content.
However, block copolymers were found to be more toxic than their statistical counterparts,
even when a higher carbohydrate content is incorporated into the diblock copolymer. This
was attributed to the inability of the block copolymer to mask the toxic cationic components.
When complexed with βꢀgalactosidase plasmid, the resulting polyplexes were stable at N/P
ratios of between 48ꢀ65. Zeta potentials of the block copolymers was found to be close to
zero, proving that the condensed DNA is mostly held within the core of the polyplex, whilst
the statistical copolymer polyplexes had net positive zeta potentials. Block copolymer
polyplexes also exhibited smaller sized complexes by dynamic light scattering (DLS) and
therefore able to condense DNA more efficiently into the polyplex core, compared to their
statistical counterparts. When HepG2 and Huh 7.5 cells were treated with these polyplexes,
higher molecular weight polymers provided improved transfection, however these were also
more toxic, proving that a careful balance of cationic charge and carbohydrate ratio, and the
ACS Paragon Plus Environment

Page 9 of 41
Biomacromolecules
1
2
3
4
5
6
7
8
9
appropriate molecular weight must be struck for efficient transfection. DNA transfection was
higher in HepG2 cells than in Huh7.5 cells, particularly when comparing p(AEMA58ꢀbꢀ
LAEMA56). When evaluated in hepatic cells lines that do not express ASGPR’s on their
surface (HeLa and SK hep 1), these polyplexes exhibit limited transfection: 25 U/mg in SK
hep 1 compared to 80 U/mg in HepG2 for p(AEMA81ꢀstꢀLAEMA58), proving uptake via
ASGPRꢀmediated endocytosis. To confirm this finding, a competitive uptake study using free
asialofetuin was conducted. Asialofetuin did not prevent transfection completely, which was
attributed to higher multiples of galactose molecules on the surface of the polyplexes of the
block copolymer complexes, enabling higher binding affinity, despite competition from
asialofetuin. Confocal studies of Cy3ꢀlabeled plasmid to identify uptake in the presence and
absence of asialofetuin. The high transfection efficiency of the nearly neutral p(AEMA58ꢀbꢀ
LAEMA56) polyplex suggests that the distances between the galactose residues are optimal in
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
this complex, consistent with the ‘cluster effect’ identified by Lee et al17, 22
.
Narain and coꢀworkers recently reported a novel polyplex formulation that exhibited
improved complexation with siRNA for the down regulation of the epidermal growth factor
receptor (EGFR) proteins in cervical cancer cells23. By combining an oxazoboroleꢀ
functionalised monomers with Nꢀisopropyl acrylamide (NIPAM) copolymer as a stimuliꢀ
responsive protective system; and a glycopolymer containing cationic residues for siRNA
complexation, crossꢀlinking between the oxazoborole and diols of the glycopolymer enable a
stable polyplex to be formed. This system exhibited gene silencing of up to 60% in HeLa
(cervical) cancer cells with low toxicity. This approach is particularly novel given the use of
two complimentary polymers to provide specific roles within the siRNA delivery and release
process. The targeting lactobionic functionalities were incorporated in a statistical copolymer,
and therefore the effect of spacing and density was not investigated. Nonetheless, the
ACS Paragon Plus Environment

Biomacromolecules
Page 10 of 41
1
2
3
4
5
6
polyplex has shown impressive results and is a promising avenue of further investigation for
oxazoboroleꢀcontaining polymers.
7
8
9
Whilst distance between targeting groups is important, another factor to take into
consideration while establishing the synthetic strategy is the density of targeting moieties on
the polymer chains. Some results have indeed shown that density matters for efficient
recognition by ASGPꢀR and cell internalization:24 if the density of targeting groups is too low
they might not be able to bind the receptors and if it is too high, steric hindrance problems
might be encountered. For similar reasons, the length of the spacers between the polymer
backbone and the targeting group also has to be considered and optimized.
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
In light of these results, it is reasonable to consider incorporating the triangular spatial
arrangement of Alnylam’s hepatic ASGP receptors into a polymer structure to obtain
improved performance in a polymer context. We postulated that having a more accessible
targeting moiety, further removed from the polymer backbone may further promote ASGPRꢀ
mediated uptake.
We have decided to focus primarily on polymeric materials. From a biological point of view,
the structure of the chosen polymeric carrier has to be very precisely controlled.25 This
control is possible by polymerization processes such as Reversible AdditionꢀFragmentation
chain Transfer (RAFT) (Figure 4, which enable the fabrication of controlled molecular
weight and polymers with very narrow polydispersity.26ꢀ28 The RAFT process is a living free
radical polymerization process which uses a chain transfer agent (CTA), typically a
thiocarbonylthio compound, to perform the radical exchange between propagating and
dormant species.
To be an efficient siRNA delivery vector, the chosen carrier has to meet a number of
requirements. The vector has to be hydrophilic and nonꢀcytotoxic, it should bind the siRNA
ACS Paragon Plus Environment

Page 11 of 41
Biomacromolecules
1
2
3
4
5
6
7
8
9
strongly enough to protect it from degradation, but release it, and enable gene silencing upon
cellular uptake.8 Our team has recently synthesised an ABA triblock polymer constituted of a
cationic core of 110 units of N,Nꢀdimethylaminoethyl methacrylate (DMAEMA) and
hydrophilic 25ꢀunits outer blocks made of polyethyleneglycol methacrylate (PEGMA)
(Figure 4). This delivery vector has shown 75% gene silencing on the tested cell lines and
also good serum stability and toxicity profiles.29 Given this success we were interested in
installing ASGPR targeting groups into the outer block in order to target siRNA delivery to
the liver.
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
In order to achieve and understand the effect of installing ASGPR targeting moieties into a
polymeric ABAꢀtriblock delivery vehicle, three main parameters must be correlated. The
nature of the targeting group, the length of the linker between the polymer backbone and the
target group, and the brush density (i.e., the number) of the targeting groups per polymer
chain.24 However, the above requirements are difficult to combine with the inherent batchꢀtoꢀ
batch variation of any given polymerization process. Although it can be argued that the
variation is relatively minor, it is still sufficient to engender doubt into the interpretation of
biological results when comparing multiple polymers across a series.
ACS Paragon Plus Environment

Biomacromolecules
Page 12 of 41
1
2
3
4
5
6
7
8
Figure 5: a. (OEGMA 475)aꢀ(DMAEMA)b ABAꢀtriblock copolymer previously reported to
achieve 75% knockdown. b. New (PEGMAꢀHPMA)aꢀ(DMAEMA)b ABAꢀtriblock copolymer
reported here.
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
In this light, postꢀpolymerisation modification of a single polymer scaffold has emerged as a
strategy to address the batchꢀtoꢀbatch polymerisation variation. Using this approach a library
of polymers can be accessed from a single batch of wellꢀdefined polymer. Here we report the
synthesis a biodegradable bisꢀRAFT agent and its subsequent use to synthesise a polymeric
ABAꢀtriblock coꢀpolymer 2 made of an Nꢀ(2ꢀhydroxypropyl)methacrylamide (HPMA),
propargyl methacrylate (PgMA) and DMAEMA (Fig.5b). We also describe the synthesis of
azideꢀbearing polyethylene glycol (PEG) linked galactose targeting agents and corresponding
glucose controls and the grafting of these groups onto the polymer backbone with the copperꢀ
catalysed azideꢀalkyne cycloaddition (CuAAC, “click” reaction)30ꢀ33. Finally, we report our
biological results, including effect of PEG linker length, and the efficacy of the PEGꢀlinked
galactose moiety as targeting unit for the ASGPR receptor.
EXPERIMENTAL SECTION
General Experimental: N,Nꢀdimethylaminoethyl methacrylate (DMAEMA) and Nꢀ(2ꢀ
Hydroxypropyl) methacrylamide (HPMA) monomers were purchased from Aldrich and
purified by stirring in the presence of inhibitorꢀremover for hydroquinone or hydroquinone
monomethyl ether (Aldrich, catalogue number 311332) for 30 min prior to use. 1,10ꢀ
Azobis(cyclohexanecarbonitrile) (VAZOꢀ88) and azobisisobutyronitrile (AIBN) initiator
(DuPont) were used as received. Dichloromethane (DCM), nꢀheptane, diisopropyl ether,
diethyl ether, pentane, propargyl bromide, methyl iodide, methanol, and other chemical
substances were commercial reagents and used without further purification. 1H and 13C
NMR (400 MHz) spectra were recorded using a Bruker Av400 spectrometer at 25°C.
ACS Paragon Plus Environment

Page 13 of 41
Biomacromolecules
1
2
3
4
5
6
7
8
9
Chemical shifts are reported in ppm downfield from tetramethylsilane (TMS, 0 ppm). The
following abbreviations were used to explain the multiplicities: s = singlet, d = doublet, t =
triplet, dd = doublet of doublets, dq = doublet of quadruplets, bs = broad singlet, m =
multiplet. For carbon NMR, all peaks correspond to one carbon, unless specified. M(th)
refers to the theoretical molecular weight based on monomer conversion, M(NMR) to the
calculated molecular weight based on relative integration of signals corresponding to the
different units. Mass spectrometric analyses were performed on a Thermo Scientific Q
Exactive mass spectrometer fitted with a HESI ion source. Positive and negative ions were
recorded in an appropriate mass range at 140,000 mass resolution. The probe was used with
0.6ml/min flow of solvent (usually methanol), and a solution of Reserpine was also
introduced into the probe during the experiments to serve as a lock mass in both positive and
negative ion modes. The nitrogen nebulizing/desolvation gas used for vaporization was
heated to 100°C in these experiments. The sheath gas flow rate was set to 25 and the
auxiliary gas flow rate to 7 (both arbitrary units). The spray voltage was 3.8 kV and the
capillary temperature was 300°C. FTIR spectra were obtained using a Thermo Nicolet 6700
spectrometer using a SmartATR (attenuated total reflectance) attachment fitted with a
diamond window. Gel permeation chromatography (GPC) measurements were performed on
two different sets of equipment. The first used a Shimadzu system equipped with a CMBꢀ
20A controller system, a SILꢀ20A HT autosampler, a LCꢀ20AT tandem pump system, a
DGUꢀ20A degasser unit, a CTOꢀ20AC column oven, a RDIꢀ10A refractive index detector
and 4 × Waters Styragel columns (HT2, HT3, HT4, HT5 each 300 mm × 7.8 mm providing
an effective molar mass range of 100 – 4 × 106). This GPC system used N,Nꢀ
dimethylacetamide (DMAc) (with 2.1 g L–1 of lithium chloride (LiCl)) as eluent with a flow
rate of 1 mL/min at 80°C. The molecular weights reported are relative to linear poly(methyl
methacrylate) (PMMA) standards. The second GPC apparatus was an Agilent SECcurity
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
ACS Paragon Plus Environment

Biomacromolecules
Page 14 of 41
1
2
3
4
5
6
7
8
9
GPC system (1200 series), comprising a quaternary pump (G1311B), a solvent degasser
(G1379B), an autosampler (G1329B), a refractive index detector (G1362) at 35 °C and a
temperature controlling unit (G1316A). The temperature controller unit housed three
Eprogen, Inc. CATSEC columns (250 × 4.6 mm) (CCS201ꢀ25 5µ 100 Å, CCS203ꢀ25 5µ 300
Å, and CCS210ꢀ25 7µ 1000 Å) and an Eprogen, Inc. CATSEC guard column (50 × 4.6 mm)
(FCCS210ꢀ5 7µ 1000 Å) at 25 °C. This system uses an eluent of 1 wt % acetic acid/0.10 M
Na2SO4 (aqueous) at a flow rate of 0.3 mL min–1. The molecular weights reports are relative
to standards of linear polyethyleneglycol (PEG) or poly(N,Nꢀdimethylaminoethyl
methacrylate) homopolymers quaternized with methyl iodide (quat. pDMAEMA).
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 lactobiono-O-(2-amidoethyl)-O’-(2-azidoethyl)-heptaethylene glycol 18.
Method 1. To a solution of lactobionic acid (419 mg, 1.17 mmol) in dry methanol (5 mL)
were added Oꢀ(2ꢀaminoethyl)ꢀO’ꢀ(2ꢀazidoethyl)ꢀheptaethylene glycol (3) (1 eq., 1.17 mmol,
513 mg) and triethylamine (2 eq., 2.34 mmol, 320 µL). The mixture was stirred at 50°C
overnight. The solvent was evaporated, the product rinsed with toluene and condensed.
Without further purification, the product was dissolved in pyridine (10 mL) and an excess of
acetic anhydride was added (4 mL, 42 mmol). The mixture was stirred for 15 min at 0°C and
a small amount of 4ꢀdimethylaminopyridine (DMAP) was added. The flask was then placed
in an oil bath at 50°C and stirred for one hour. The solvent was removed under reduced
pressure, the residue dissolved in ethyl acetate and washed successively with H2O, 2% citric
acid, saturated sodium bicarbonate and brine. The aqueous phases were reꢀextracted twice
with ethyl acetate. The organic phases were combined, dried on anhydrous magnesium
sulfate and condensed. The crude mixture was purified by column chromatography on a silica
column first using 1% methanol/dichloromethane, then 5% methanol/dichloromethane,
leading to the amide 7 in 60% yield over the two steps. The alcohols were then deprotected
using sodium methoxide (0.1 eq.) in anhydrous methanol. The mixture was stirred at room
ACS Paragon Plus Environment

Page 15 of 41
Biomacromolecules
1
2
3
4
5
6
temperature for one hour. Acid resin was used to remove sodium methoxide and the solvent
was evaporated, leading to the clickable unit 10 in 36% yield over the three steps.
7
8
9
Following the previous protocol, lactobionic acid and the other PEG aminoꢀazide compounds
13 and 14 were typically reacted together. The clickable units 19 and 20 were obtained in
36% and 45% yield respectively.
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
Method 2. Lactobionic acid was first partially converted to the corresponding
lactobionolactone. This was achieved by dissolving lactobionic acid (25g, 70 mmol) and ca. 1
mL of trifluoroacetic acid in anhydrous methanol (150 mL) at 50°C, followed by vacuum
distillation. The procedure was repeated three times until equilibrium was reached, and no
further conversion could be observed. After solvent evaporation, the product was dried in the
vacuum oven for 24 hours. In a second step, to a solution of this mixture (408 mg, 1.14
mmol) in dry methanol (5 mL) were added the PEGꢀ7 aminoꢀazide 12 (1 eq, 1.14 mmol, 500
mg) and triethylamine (2 eq, 2.28 mmol, 310 µL). The mixture was stirred at 50°C overnight
and the solvent was evaporated. The mixture was passed through a short column of acid resin
to remove the unreacted 12, and 18 was obtained in 81% yield.
18. Rf = 0.69 (CH3Cl3/MeOH/H2O 15/10/1). 〖[α]〗_D^19 +15.9 (c 1.02 in MeOH). 1H
NMR (CD3OD, δ, ppm): 4.49 (d, J 7.6, 1H), 4.36 (d, J 2.5, 1H), 4.23 (dd, J 2.5, 3.9, 1H), 3.93
(dd, J 4, 6.6, 1H), 3.85ꢀ3.90 (m, 1H), 3.75ꢀ3.84 (m, 4H), 3.62ꢀ3.76 (m, 31H), 3.53ꢀ3.61 (m,
4H), 3.35ꢀ3.52 (m, 6H). 13C NMR (CD3OD, δ, ppm): 175.2, 105.8, 83.2, 77.2, 74.8, 73.9,
73.2, 72.8, 72.4, 71.6, 71.5 (8C), 71.3, 71.1, 70.6, 70.4, 63.8, 62.7, 51.8, 40.0. IR: 1649
(C=O), 2106 (N3). m/z (ESI+) 691 (M+H, 100%). m/z (HRꢀEI+) found: 691.3242;
C26H51O17N4 691.3244.
19. Rf = 0.77 (CH3Cl3/MeOH/H2O 15/10/1). 1H NMR (CD3OD, δ, ppm): 4.48 (d, J 7.5, 1H),
4.36 (d, J 2.6, 1H), 4.21 (dd, J 2.6, 3.9, 1H), 3.93 (dd, J 3.9, 6.6, 1H), 3.85ꢀ3.90 (m, 1H),
ACS Paragon Plus Environment

Biomacromolecules
Page 16 of 41
1
2
3
4
5
6
7
8
9
3.74ꢀ3.84 (m, 5H), 3.61ꢀ3.73 (m, 38H), 3.53ꢀ3.60 (m, 4H), 3.36ꢀ3.52 (m, 6H). 13C NMR
(CD3OD, δ, ppm): 175.1, 105.7, 83.2, 77.2, 74.8, 73.9, 73.2, 72.8, 72.4, 71.6, 71.5 (12C),
71.3, 71.1, 70.5, 70.3, 63.8, 62.7, 51.8, 40.0. IR: 1646 (C=O), 2105 (N3). m/z (ESI+) 779
(M+H, 65%). m/z (HRꢀEI+) found: 779.3766; C30H59O19N4 779.3768.
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
20. Rf = 0.86 (CH3Cl3/MeOH/H2O 15/10/1). 〖[α]〗_D^21 +12.4 (c 1.02 in MeOH). 1H
NMR (CD3OD, δ, ppm): 4.48 (d, J 7.5, 1H), 4.36 (d, J 2.5, 1H), 4.22 (dd, J 2.5, 4, 1H), 3.93
(dd, J 3.9, 6.6, 1H), 3.87 (m, 1H), 3.74ꢀ3.84 (m, 4H), 3.61ꢀ3.72 (m, 47H), 3.53ꢀ3.60 (m, 4H),
3.34ꢀ3.52 (m, 6H). 13C NMR (CD3OD, δ, ppm): 175.2, 105.8, 83.2, 77.2, 74.8, 73.9, 73.2,
72.8, 72.4, 71.6, 71.5 (16C), 71.3, 71.1, 70.6, 70.4, 63.8, 62.7, 51.8, 40.0. IR: 1647 (C=O),
2105 (N3). m/z (ESI+) 867 (M+H, 30%). m/z (HRꢀEI+) found: 867.4292; C34H67O21N4
867.4292.
Synthesis of macroRAFT p(DMAEMA): 1,1'ꢀAzobiscyclohexanecarbonitrile (vazo 88, 5.65
mg, 2.31x10ꢀ5 mol), bisꢀRAFT agent 13 (20 eq, 428 mg, 4.63x10ꢀ4 mol), DMAEMA (3140
eq, 11.418 g, 7.26x10ꢀ2 mol) and DMF (42.36 mL, 40 g) were mixed together in a vial. After
30 min of stirring, the mixture was transferred in a 100 mL Schlenk flask and submitted to
four freezeꢀpumpꢀthaw cycles to remove oxygen. The flask was placed at 90°C for 17 hours;
the reaction was then quenched by cooling and exposure to air. After 1H NMR determination
of monomer conversion (see Scheme 4), the solvent was evaporated. The crude oil was
diluted with dichloromethane (DCM) and the solution was precipitated by dropwise addition
in heptane/diisopropyl ether 2/1. The solvent was decanted and the procedure repeated twice.
The polymer was dissolved in DCM and the solvent evaporated overnight. 4.95 g of yellow
solid foam were obtained. 1H NMR (CDCl3, δ, ppm): 4.00ꢀ4.12 (bs, 2H, CH2O), 2.51ꢀ2.63
(bs, 2H, CH2N), 2.24ꢀ2.33 (bs, 6H, N(CH3)2), 1.71ꢀ2.06 (m, 2H, CH2 backbone), 0.79ꢀ1.17
ACS Paragon Plus Environment

Page 17 of 41
Biomacromolecules
1
2
3
4
5
6
(m, 3H, CH3 backbone). GPC in DMAc against PMMA standards: Mn = 23.5 kDa, PDI =
1.15.
7
8
9
Synthesis of TMS-protected PgMA: Silver chloride (1.46 g, 9.91x10ꢀ3 mol) was suspended in
145 mL of dry dichloromethane. Propargyl methacrylate (13 g, 1.05x10ꢀ1 mol) and DBU (20
mL, 1.34x10ꢀ2 mol) were added to the suspension, which became clear. The reaction mixture
was heated to 40°C and trimethylsilane chloride (TMSCl, 19.8 mL, 1.49x10ꢀ1 mol) added
dropwise via a syringe pump. The mixture was then stirred for 24 hours. The obtained dark
red solution was diluted with 300 mL heptane and washed successively with water, 1% HCl,
saturated NaHCO3 (200 mL each). The organic phase was dried on MgSO4, filtered and
concentrated under reduced pressure. The crude product was purified by silica column
chromatography in a heptane/diethyl ether 25/1 mixture to afford PgMA in 74% yield. 1H
NMR (CDCl3, δ, ppm): 6.17 (m, 1H, C=CH), 5.61 (m, 1H, C=CH’), 4.76 (s, 2H, CH2), 1.96
(m, 3H, CH3), 0.18 (s, 9H, TMS). 13C NMR (CDCl3, δ, ppm): 166.8 (C=O), 135.9 (CH2=C),
126.5 (CH2=C), 99.4 (CH2ꢀC≡C), 92.1 (C≡CꢀTMS), 53.5 (CH2O), 18.4 (CH3), ꢀ0.1
(Si(CH3)3). IR: 2200 (weak, C≡C stretch), 1723 (strong, C=O stretch), 1639 (medium, C=C
stretch).
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 the ABA triblock 10: AIBN (1.03 mg, 6.2x10ꢀ6 mol), pDMAEMA (590.2 mg,
2.51x10ꢀ5 mol), TMSꢀPgMA (54.8 mg, 2.79x10ꢀ4 mol) and HPMA (323.79 mg, 2.26x10ꢀ3
mol) were mixed together in a vial with DMAc (total monomer concentration 0.7 mol/L). The
ratio HPMA to TMSꢀPgMA was 9/1 and the total molar ratio of monomer to pDMAEMA
(macro RAFT agent) to initiator was 100/1/0.25. The mixture was transferred in a Schlenck
flask, submitted to four freezeꢀpumpꢀthaw cycles and placed in a 60 °C oil bath. After 24
hours, the reaction was quenched by cooling and exposure to air and NMR analysis was used
to determine monomer conversions: PgMA conversion = 80% (8 units per chain) and HPMA
conversion = 47% (42 units per chain). M(th) = 25.7 kDa. The solvent was evaporated, the
ACS Paragon Plus Environment

Biomacromolecules
Page 18 of 41
1
2
3
4
5
6
7
8
9
polymer was dissolved in methanol and precipitated in a mixture of diethyl ether and pentane
1/1.5 4 times. 1H NMR (CD3OD, δ, ppm): 7.35ꢀ7.60 (bs, 1H, NH (HPMA) partially
exchanged), 4.51ꢀ4.73 (bs, 2H, CH2O (PgMA)), 4.03ꢀ4.21 (bs, 2H, CH2O (DMAEMA)),
3.79ꢀ3.96 bs, 1H, CH (HPMA)), 2.89ꢀ3.26 (m, 2H, CH2 (HPMA)), 2.60ꢀ2.74 (bs, 2H, CH2N
(DMAEMA)), 2.27ꢀ2.44 (m, 6H, N(CH2)3 (DMAEMA)), 0.80ꢀ2.23 (m, CH2 and CH3
backbone + CH3 (HPMA)), 0.15ꢀ0.27 (bs, 9H, TMS (PgMA)). PgMA/HPMA/DMAEMA =
1/3.8/9.6, M(NMR) = 26.6 kDa. GPC in DMAc against PMMA standards: Mn = 36.4 kDa,
PDI = 1.18.
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
For the second batch, all the quantities were multiplied by 5 and the same procedure was
followed. NMR analysis was used to determine monomer conversions: PgMA conversion =
80% (8 units per chain) and HPMA conversion = 33% (30 units per chain). M(th) = 24 kDa.
The solvent was evaporated and the polymer was precipitated 4 times in a cold diethyl
ether/pentane mixture 1/1.5, paying particular attention to the dilutions in minimal amounts
of methanol. The NMR spectrum of the precipitated polymer showed the same peaks as the
previous batch. PgMA/HPMA/DMAEMA = 1/4.7/15.7, M(NMR) = 24.3 kDa. GPC in
DMAc against PMMA standards: Mn = 33 kDa, PDI = 1.33.
Deprotection to ABA triblock 11: 225 mg of the ABA triblock 16 (M = 26.4 kDa, DP(PgMA)
= 10, 8.52x10ꢀ5 mol TMS groups) were dissolved in 15 mL of THF. One drop of acetic acid
was added, the mixture was bubbled with nitrogen for 10 min and cooled down to ꢀ20 °C.
TBAF in solution in THF (0.2 M, 0.5 mL, 1.1 eq., 9.4x10ꢀ5 mol) was slowly added via
syringe. The mixture was stirred at this temperature for 30 min, then warmed to ambient
temperature and stirred for 3 more hours. The reaction solution was passed through a short
silica pad in order to remove the excess of TBAF and the pad was subsequently washed with
additional THF. The resulting solution was then concentrated under reduced pressure. The
crude polymer was dissolved in methanol and precipitated in a cold diethyl ether/pentane
ACS Paragon Plus Environment

Page 19 of 41
Biomacromolecules
1
2
3
4
5
6
7
8
9
1/1.5 mixture. 1H NMR (CD3OD, δ, ppm): 7.35ꢀ7.60 (bs, 1H, NH (HPMA) partially
exchanged), 4.51ꢀ4.73 (bs, 2H, CH2O (PgMA)), 4.03ꢀ4.21 (bs, 2H, CH2O (DMAEMA)),
3.79ꢀ3.96 bs, 1H, CH (HPMA)), 2.89ꢀ3.26 (m, 2H, CH2 (HPMA)), 2.60ꢀ2.74 (bs, 2H, CH2N
(DMAEMA)), 2.27ꢀ2.44 (m, 6H, N(CH2)3 (DMAEMA)), 0.80ꢀ2.23 (m, CH2 and CH3
backbone + CH3 (HPMA)). GPC in DMAc against PMMA standards: Mn = 36 kDa, PDI =
1.24.
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
For the second batch of ABA triblock, the same molar ratios and procedure were used for the
deprotection of 2.1 g of the ABA triblock (M = 24.3 kDa, DP(PgMA) = 8, 6.91x10ꢀ4 mol
TMS groups). The NMR spectrum of the precipitated polymer (20) showed the same peaks as
the previous batch. GPC in DMAc against PMMA standards: Mn = 39.8 kDa, PDI = 1.25.
Grafting of the targeting units 18, 19, 20, 30 and 31 on polymer 11 by click reaction:34 100
mg of the deprotected ABA triblock 11 (M = 23.7 kDa, DP(PgMA) = 8, 1.6x10ꢀ5 mol alkyne
groups), the clickable unit (1.5 eq., 5.04x10ꢀ5 mol, 34.8 mg , 39.2 mg, 43.7 mg, 38.5 mg and
42.9 mg for 18, 19, 20, 29 and 30, respectively) and PMDETA (1.9 µL, 0.3 eq, 9.24x10ꢀ6
mol) were solubilized in 4 mL DMF. The solution was bubbled with nitrogen for 15 min and
copper bromide was added (1.3 mg, 0.3 eq., 9.24x10ꢀ6 mol). The solution was bubbled 10 min
more and stirred at room temperature for 24 hours. Quadrasil resin (ca. 100 mg) was added
and the mixture was gently stirred for 2 days, filtered and condensed under reduced pressure
at 50°C. The crude products 32ꢀ36 were purified by size exclusion chromatography in a LHꢀ
20 column in DMF.
32. 1H NMR (CD3OD, δ, ppm): 8.10ꢀ8.20 (bs, 1H (PgMA), triazole), 4.61ꢀ4.70 (bs, 2H
(PgMA), CH2O), 4.02ꢀ4.20 (bs, 2H (DMAEMA), CH2O), 3.77ꢀ4.00 (m, 1H (HPMA), CH +
7H (18)), 3.66ꢀ3.73 (m, 34H (18)), 2.92ꢀ3.24 (m, 2H (HPMA), CH2), 2.60ꢀ2.73 (bs, 2H
(DMAEMA), CH2N), 2.28ꢀ2.41 (m, 6H (DMAEMA), N(CH2)3), 0.79ꢀ2.21 (m, 5H (PgMA) +
ACS Paragon Plus Environment

Biomacromolecules
Page 20 of 41
1
2
3
4
5
6
8H (HPMA) + 5H (DMAEMA), CH2 and CH3 backbone + CH3 (HPMA)). GPC in DMAc
against PMMA standards: Mn = 55.3 kDa, PDI = 2.05.
7
8
9
33. 1H NMR (CD3OD, δ, ppm): 8.08ꢀ8.24 (bs, 1H (PgMA), triazole), 4.61ꢀ4.68 (bs, 2H
(PgMA), CH2O), 4.48ꢀ4.54 (bs, 1H (19)), 4.36ꢀ4.39 (bs, 1H (19)), 4.22ꢀ4.26 (bs, 1H (19)),
4.03ꢀ4.19 (bs, 2H (DMAEMA), CH2O), 3.77ꢀ3.99 (m, 1H (HPMA), CH + 7H (19)), 3.56ꢀ
3.70 (m, 48H (19)), 2.93ꢀ3.24 (m, 2H (HPMA), CH2), 2.60ꢀ2.73 (bs, 2H (DMAEMA),
CH2N), 2.28ꢀ2.41 (m, 6H (DMAEMA), N(CH2)3), 0.73ꢀ2.20 (m, 5H (PgMA) + 8H (HPMA)
+ 5H (DMAEMA), CH2 and CH3 backbone + CH3 (HPMA)). GPC in DMAc against PMMA
standards: Mn = 57.5 kDa, PDI = 1.49.
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
34. 1H NMR (CD3OD, δ, ppm): 8.07ꢀ8.25 (bs, 1H (PgMA), triazole), 4.61ꢀ4.68 (bs, 2H
(PgMA), CH2O), 4.49ꢀ4.54 (bs, 1H (20)), 4.37ꢀ4.39 (bs, 1H (20)), 4.22ꢀ4.26 (bs, 1H (20)),
4.05ꢀ4.17 (bs, 2H (DMAEMA), CH2O), 3.78ꢀ3.98 (m, 1H (HPMA), CH + 6H (20)), 3.56ꢀ
3.69 (m, 57H (20)), 2.92ꢀ3.25 (m, 2H (HPMA), CH2), 2.60ꢀ2.73 (bs, 2H (DMAEMA),
CH2N), 2.28ꢀ2.41 (m, 6H (DMAEMA), N(CH2)3), 0.80ꢀ2.20 (m, 5H (PgMA) + 8H (HPMA)
+ 5H (DMAEMA), CH2 and CH3 backbone + CH3 (HPMA)). GPC in DMAc against
PMMA standards: Mn = 48.5 kDa, PDI = 1.71.
35. 1H NMR (CD3OD, δ, ppm): 8.08ꢀ8.25 (bs, 1H (PgMA), triazole), 5.02ꢀ5.21 (bs, 1H,
(30)), 4.74ꢀ5.00 (bs, 2H (PgMA), CH2O), 4.56ꢀ4.71 (bs, 1H (30)), 4.30ꢀ4.46 (bdd, 1H (30)),
3.97ꢀ4.20 (bs, 2H (DMAEMA), CH2O), 3.79ꢀ3.96 (m, 1H (HPMA), CH + 7H (30)), 3.47ꢀ
3.72 (m, 57H (30)), 2.89ꢀ3.44 (m, 2H (HPMA), CH2), 2.54ꢀ2.72 (bs, 2H (DMAEMA),
CH2N), 2.23ꢀ2.38 (m, 6H (DMAEMA), N(CH2)3), 0.74ꢀ2.10 (m, 5H (PgMA) + 8H (HPMA)
+ 5H (DMAEMA), CH2 and CH3 backbone + CH3 (HPMA)). GPC in DMAc against PMMA
standards: Mn = 41.4 kDa, PDI = 1.41.
ACS Paragon Plus Environment

Page 21 of 41
Biomacromolecules
1
2
3
4
5
6
7
8
9
36. 1H NMR (CD3OD, δ, ppm): 8.07ꢀ8.20 (bs, 1H (PgMA), triazole), 5.01ꢀ5.24 (bs, 1H,
(31)), 4.76ꢀ4.96 (bs, 2H (PgMA), CH2O), 4.55ꢀ4.69 (bs, 1H (31)), 4.29ꢀ4.46 (bdd, 1H (31)),
3.96ꢀ4.20 (bs, 2H (DMAEMA), CH2O), 3.77ꢀ3.95 (m, 1H (HPMA), CH + 7H (31)), 3.47ꢀ
3.76 (m, 57H (31)), 2.87ꢀ3.42 (m, 2H (HPMA), CH2), 2.53ꢀ2.73 (bs, 2H (DMAEMA),
CH2N), 2.19ꢀ2.40 (m, 6H (DMAEMA), N(CH2)3), 0.72ꢀ2.11 (m, 5H (PgMA) + 8H (HPMA)
+ 5H (DMAEMA), CH2 and CH3 backbone + CH3 (HPMA)). GPC in DMAc against PMMA
standards: Mn = 41.4 kDa, PDI = 1.50.
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
Cells:
Chinese Hamster Ovary cells constitutively expressing Green Fluorescent protein (CHOꢀ
GFP) (kindly received from K. Wark; CSIRO CMHT Australia) were grown in MEMα
modification supplemented with 10% foetal bovine serum, 10 mM Hepes, 0.01% penicillin
and 0.01% streptomycin at 37 ºC with 5% CO2 and subcultured twice weekly. Human
adenocarcinomic human alveolar basal epithelial cells (A549) and human hepatocarcinoma
(huh7) cells were grown in DMEM supplemented with 10% foetal bovine serum, 10 mM
Hepes, 2 mM glutamine, 0.01% penicillin and 0.01% streptomycin at 37 ºC with 5% CO2 and
subcultured twice weekly.
Toxicity Assay:
CHOꢀGFP, A549 and Huh7 cells were seeded at 2x104 cells in 96ꢀwell tissue culture plates in
triplicate and grown overnight at 37 ºC with 5% CO2.
The serially diluted polymer materials or polymer/diRNA complexes were added to 3 wells
in the 96 well culture plates for each sample and incubated for 72 h at 37 ºC in 200 µl
standard media. Toxicity was measured using the Alamar Blue reagent (Invitrogen USA)
according to manufacturer’s instructions. Briefly, media was removed, cells were washed
ACS Paragon Plus Environment

Biomacromolecules
Page 22 of 41
1
2
3
4
5
6
7
8
9
with PBS and replaced with 100 µl of standard media containing 10% Alamar Blue reagent,
cells were then incubated for 4 h at 37 ºC with 5% CO2. The assay was read on an EL808
Absorbance microplate reader (BIOTEK, USA) at 540 nm and 620 nm. Cell viability was
determined by subtracting the 620 nm measurement from the 540 nm measurement. Obtained
data was analysed in Microsoft Excel. Results are presented as a percentage of untreated cells
and the presented data are representative of three separate experiments in triplicate.
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
siRNA:
The antiꢀGFP and negative control siRNAs were obtained from QIAGEN (USA).
The antiꢀGFP siRNA sequence is sense 5’ gcaagcugacccugaaguucau 3’ and antisense
5’gaacuucagggucagcuugccg 3’. [6 FAM] labelled si22 was purchased from Sigma Aldrich
(USA) with the FAM label on the sense strand.
The nonꢀsilencing control is the same sequence as the siRNA as a DNA duplex. This will not
be recognised by RISC, therefore should nit silence GFP.
The siDeath siRNA mix was obtained from QIAGEN.
The antiꢀCOPA siRNA sense sequences are
siCOPAꢀ1 ACUCAGAUCUGGUGUAAUA[dT][dT]
siCOPAꢀ2 GCAAUAUGCUACACUAUGU[dT][dT]
siCOPAꢀ3 GAUCAGACCAUCCGAGUGU[dT][dT]
siCOPAꢀ4 GAGUUGAUCCUCAGCAAUU[dT][dT]
Formation of polymer/siRNA complexes:
ACS Paragon Plus Environment

Page 23 of 41
Biomacromolecules
1
2
3
4
5
6
7
8
9
Molar ratios of polymer to 50 pmole siRNA or siDNA were calculated. Complexes were
formed by the addition of OPTIMEM media (Invitrogen, USA) to eppendorf tubes. The
required amount of polymer resuspended in water was added to the tubes and the mixture
vortexed. 50nM of si22 or di22 was then added to the tubes and the sample vortexed.
Complexation was allowed to continue for 1 h or overnight at room temperature.
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
Agarose gel:
Samples containing 50 pmole of siRNA were electrophoresed on a 2% agarose gel in TBE at
100V for 40 min. siRNA was visualised by gel red (Jomar Bioscience) on a UV
transilluminator with camera, the image was recorded by the GeneSnap program (Syngene,
USA).
Silencing Assay:
CHOꢀGFP cells were seeded at 1x104 cells in 96ꢀwell tissue culture plates in triplicate and
grown overnight at 37 ºC with 5% CO2. For positive and negative controls siRNAs were
transfected into cells using Lipofectamine 2000 (Invitrogen, USA) as per manufacturer’s
instructions. Briefly, 50 picomole of the relevant siRNA were mixed with 1 µl of
Lipofectamine 2000 both diluted in 50 µl OPTIꢀMEM (Invitrogen, USA) and incubated at
room temperature for 20 mins. The siNA: lipofectamine mix was added to cells and
incubated for 4 h. Cell media was replaced and incubated for 72 h.
For polymer/siRNA complexes prepared, cell media was removed and replaced with 200 µl
OPTIꢀMEM. The siRNA: polymer complexes in a volume of 10µl was added to 3 wells of
cells per sample and incubated for 5 h. Cell media was replaced and cells incubated for a
further 72 h.
ACS Paragon Plus Environment

Biomacromolecules
Page 24 of 41
1
2
3
4
5
6
7
8
Cells were washed twice with PBS, and read on a Fluoroskan Ascent FL (Thermo Scientific,
USA) and EGFP silencing was analysed as a percentage of the siDNA or polymer/siDNA
complexes mean EGFP (ex 488nm em 516nm) fluorescence.
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
For A549 and Huh7 cells; Cell death was determined by the Alamar Blue reagent (Invitrogen
USA) according to manufacturer’s instructions. Briefly, media was removed, cells were
washed with PBS and replaced with 100 µl of standard media containing 10% Alamar Blue
reagent, cells were then incubated for 4 h at 37 ºC with 5% CO2. The assay was read on an
EL808 Absorbance microplate reader (BIOTEK, USA) at 540 nm and 620 nm. Cell viability
was determined by subtracting the 620 nm measurement from the 540 nm measurement.
Obtained data was analysed in Microsoft Excel. Results are presented as a percentage of di22
control treated cells and the presented data are representative of three separate experiments in
triplicate.
Uptake of [6’FAM] si22
Polymer/si22ꢀFAM complexes (50 pmol) were assembled as described above using the
[6FAM] labelled si22.
Flow cytometry; CHOꢀWT cells were seeded at 1x105 cells in 96ꢀwell tissue culture plates in
triplicate (Nunc, USA) and grown overnight at 37ºC with 5% CO2. For positive controls
[6FAM] labelled si22 was transfected into cells using Lipofectamine 2000 as described
above. Polymer and labelled siRNA complexes were produced as described above and added
to the cells for 5 or 24 hrs. Cells were washed with PBS, trypsinised and washed twice with
FACS wash (PBS with 1% FBS). Cells were subjected to flow cytometry and the FITC
fluorescence at emission of 516 nm was analysed in Microsoft Excel based on the uptake
index which is the percentage of cell transfected multiplied by the mean fluorescence of the
transfected cell population. Results are presented relative to [6FAM] labelled si22 alone
ACS Paragon Plus Environment

Page 25 of 41
Biomacromolecules
1
2
3
4
5
6
treated cells and the presented data are representative of three separate experiments in
triplicate.
7
8
9
RESULTS AND DISCUSSION
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 alkyne-bearing polymers
The polymer synthesis started with the synthesis of a biodegradable disulfideꢀcontaining bisꢀ
RAFT agent 5, which was derived through standard coupling of the trithiocarbonate 3 and
alcohol 4 (Scheme 1). Our previous studies have demonstrated the optimal central
pDMAEMA block of the polymer is 110 units to be an effective siRNA carrier, with low
toxicity, which was synthesised employing standard RAFT protocols with DMAEMA 6 [14].
Thus, the resulting pDMAEMA macro RAFT agent 7 was purified by precipitation and
characterized by 1H NMR and gel permeation chromatography (GPC). The calculated Mn of
the pDMAEMA was 18.2 kDa (based on the NMR determination of the monomer conversion
(70%)), and the molecular weight determined by GPC (23.4 kDa). As expected, the
polydispersity of this polymer was narrow, 1.15 (Table 1).
ACS Paragon Plus Environment

Biomacromolecules
Page 26 of 41
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
Scheme 1: Synthesis of alkyneꢀbearing RAFT ABAꢀtriblock copolymers
To obtain a discrete distribution of alkyne functionality randomly dispersed in the outer
blocks of the pDMAEMA 7, the hydrophilic monomer HPMA 9 was copolymerized with
trimethylsilylꢀpropargyl methacrylate (TMSꢀPgMA) 8. The TMS protecting group is the
established method to prevent the alkyne groups participating in radical reactions during the
polymerisation process31. Indeed, our early attempts with PgMA gave polymers with much
broader polydispersity. This may be attributed to the involvement of the alkyne groups in the
radical addition process, probably leading to crossꢀlinking and branching phenomena.35 The
TMS protected PgMA 8 was conveniently synthesized by treating PgMA with TMSꢀCl in the
presence of AgCl.36 Thus, TMSꢀPgMA 8 was copolymerized with HPMA 9 in DMAc using
standard RAFT protocols to give the ABAꢀtriblock copolymer 10. 1H NMR analysis showed
ACS Paragon Plus Environment

Page 27 of 41
Biomacromolecules
1
2
3
4
5
6
7
8
9
that the obtained ABAꢀtriblock copolymer 10 contained 33 HPMA units and 7 PgMA units
per polymer chain (for a calculated molecular weight of 24.3 kDa). The GPC against
polymethylmethacrylate (PMMA) standards found a polydispersity of 1.33 and molecular
weight 33 kDa. The ABAꢀtriblock copolymer 10 was then easily deprotected using
tetrabutylammonium fluoride (TBAF) in THF. The NMR analysis showed complete
disappearance of the TMS peak and GPC analysis confirmed a slight decrease in MW from
36.4 kDa to 35.9 kDa.
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 azide-bearing ASGPR targeting groups
The synthesis of the galactose targeting group took advantage of commercially available
lactobionic acid which is ideally configured, in that it bears a βꢀgalactose group and a
convenient spacer between the targeting group and the PEG linker, allowing conformational
flexibility during ligand binding. Lactobionic acid was condensed directly with three different
amines (Compounds 12-14) containing azide end groups and polyethylene glycol chains
(PEG) with increasing EG repeats from 5 to 9 units. The reaction presumably proceeded
through the in situ formation of the lactone 21, and after acetate protection, delivers the
amides 15-17 in 50 to 60% yield over the two steps. Acetate protection was used to aid with
purification of the products. With pure compounds in hand, the acetates 15, 16, and 17 were
deprotected, leading to clickable targeting agents 18, 19 and 20 in 35 to 45% yield over the
three steps, respectively (Scheme 2).
An alternate approach was also investigated, which consisted of converting the lactobionic
acid to lactobionolactone 21 prior to reaction with the PEG amine compounds 12, 13 and
1437. Despite incomplete conversion of lactobionic acid to the lactone 21, this route led to
amides 18, 19 and 20 in excellent yields of 81% yield. Key to this conversion was the
removal of any excess amine 12, 13 or 14 with Amberlyst 15 resin, which delivered material
ACS Paragon Plus Environment

Biomacromolecules
Page 28 of 41
1
2
3
4
5
6
comparable to that synthesised through the protection method and required no further
purification (Scheme 2).
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
Scheme 2: The synthesis of azideꢀbearing galactose ASPGR targeting groups.
Synthesis of the glucose controls was performed via a similar protocol, employing
commercially available cellobiose 22, to mimic the galactose linker strategy. An initial
approach was to oxidise cellobiose into cellobionolactone for the facile introduction of the
azideꢀbearing PEG linker (Scheme 3) however, the lactone appeared unreactive towards
amide formation under a range of conditions.38 A second option involved oxidation and
amide formation in one step, which was attempted with a model amine, ethylamine.39 Upon
acetyl protection of the resultant amide, only a 5% yield of the product was gained. It was
therefore deemed necessary to attach the azidoꢀPEG groups through a glycosyl link by
initially activating the anomeric alcohol and displacing it with a hydroxyl terminated azideꢀ
bearing PEG linker. For this approach, cellobiose was fully protected as acetyl cellobiose 23
before deprotection of the anomeric hydroxyl using hydrazine in DMF at 50 °C. The
anomeric alcohol 24 was converted into the trichloroacetimidate 25 for the addition of PEGꢀ7
and PEGꢀ9 linkers, 26 and 27, respectively, employing TMSOTf and molecular sieves. The
ACS Paragon Plus Environment

Page 29 of 41
Biomacromolecules
1
2
3
4
5
6
7
8
9
resultant 1:1 α/β anomeric mixtures 28 and 29 were then deprotected with NaOMe in
methanol to provide the glucoseꢀcontaining PEGꢀ7 and PEGꢀ9 glycosides 30 and 31,
respectively (Scheme 3). Whilst it was decided to carry the 1:1 α/β anomeric mixture forward
without separation, it is worthy to note that the terminal glucose residue (like the galactose
analogue) exists exclusively as its βꢀanomer.
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
OAc
OH
OAc
H2NNH2 AcOH,
OAc
OH
AcO
O
HO
O
AcO
O
OH
DMF, 50
C
Ac2O
O
AcO
O
HO
O
AcO
AcO
AcO
HO
HO
AcO
AcO
O
O
O
OAc
OH
OAc
HClO4, 0 C to rt
95 %
OAc
OH
OAc
66 %
22 - Cellobiose
24
23
Cl3C
Cs2CO3, DCM, rt
N
69 %
NH
O
N3
HO
O
n
OAc
26: n = 7
27: n = 9
O
CCl3
OAc
AcO
O
O
O
n
OAc
O
AcO
O
O
N3
AcO
AcO
O
AcO
AcO
AcO
O
O
OAc
TMSOTf, 4A mol.sieve,
DCM, -30 C to 0
OAc
OAc
= 1:0
NaOMe
C
OAc
MeOH, 0
C
25
:
28: n = 7; 81 % yield,
29: n = 9; 77 % yield,
:
:
= 1:1
= 1:1
OH
O
O
n
OH
O
O
N3
O
HO
O
HO
HO
OH
OH
30: n = 7; 92 % yield,
31: n = 9; 91 % yield,
:
:
= 1:1
= 1:1
Scheme 3: Synthesis of azideꢀbearing glucose control moieties
Grafting of the ASGPR targeting and glucose control units.
The targeting units 18, 19, 20, 30 and 31 were grafted in parallel onto the polymer 11, in the
presence of copper bromide as the catalyst and N,N,N’,N”,N”ꢀpentamethyldiethylenetriamine
(PMDETA) as the ligand,33 leading to the polymers 32, 33, 34, 35 and 36. GPC analysis
showed that the polymeric products contained azide starting material. Early attempts to purify
by dialysis were unsuccessful (data not shown). Purification by precipitation was similarly
unsuccessful because of the similarly in solubility of the azides (18-20, 30, 31) and the
resulting polymers. However, purification of the obtained polymers by size exclusion
chromatography (Sephadex LHꢀ20) allowed completed removal of the excess clickable units
ACS Paragon Plus Environment