Comprehensive Structure-Activity Relationship of Triantennary N-Acetylgalactosamine Conjugated Antisense Oligonucleotides for Targeted Delivery to Hepatocytes

Article abstract of DOI:10.1021/acs.jmedchem.5b01948

The comprehensive structure-activity relationships of triantennary GalNAc conjugated ASOs for enhancing potency via ASGR mediated delivery to hepatocytes is reported. Seventeen GalNAc clusters were assembled from six distinct scaffolds and attached to ASOs. The resulting ASO conjugates were evaluated in ASGR binding assays, in primary hepatocytes, and in mice. Five structurally distinct GalNAc clusters were chosen for more extensive evaluation using ASOs targeting SRB-1, A1AT, FXI, TTR, and ApoC III mRNAs. GalNAc-ASO conjugates exhibited excellent potencies (ED₅₀ 0.5-2 mg/kg) for reducing the targeted mRNAs and proteins. This work culminated in the identification of a simplified tris-based GalNAc cluster (THA-GN3), which can be efficiently assembled using readily available starting materials and conjugated to ASOs using a solution phase conjugation strategy. GalNAc-ASO conjugates thus represent a viable approach for enhancing potency of ASO drugs in the clinic without adding significant complexity or cost to existing protocols for manufacturing oligonucleotide drugs.

Full text of DOI:10.1021/acs.jmedchem.5b01948

Subscriber access provided by RUTGERS UNIVERSITY
Article
Comprehensive Structure Activity Relationship of
Triantennary N-Acetylgalactosamine Conjugated Antisense
Oligonucleotides for Targeted Delivery to Hepatocytes
Thazha P. Prakash, Jinghua Yu, Michael T. Migawa, Garth A. Kinberger, W. Brad Wan, Michael E Østergaard,
Recaldo L Carty, Guillermo Vasquez, Audrey Low, Alfred Chappell, Karsten Schmidt, Mariam Aghajan, Jeff
Crosby, Heather M. Murray, Sheri L. Booten, Jill Hsiao, Armand Soriano, Todd Machemer, Patrick Cauntay,
Sebastien A. Burel, Susan F. Murray, Hans J. Gaus, Mark J. Graham, Eric E. Swayze, and Punit P Seth
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01948 • Publication Date (Web): 25 Feb 2016
Downloaded from http://pubs.acs.org on February 28, 2016
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.
Journal of Medicinal Chemistry 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 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Comprehensive Structure Activity Relationship of Triantennary
NꢀAcetylgalactosamine Conjugated Antisense Oligonucleotides
for Targeted Delivery to Hepatocytes
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
Thazha P. Prakash*, Jinghua Yu, Michael T. Migawa, Garth A. Kinberger, W. Brad Wan,
Michael E. Østergaard, Recaldo L. Carty, Guillermo Vasquez, Audrey Low, Alfred Chappell,
Karsten Schmidt, Mariam Aghajan, Jeff Crosby, Heather M. Murray, Sheri L. Booten, Jill Hsiao,
Armand Soriano, Todd Machemer, Patrick Cauntay, Sebastien A. Burel, Susan F. Murray, Hans
Gaus, Mark J. Graham, Eric E. Swayze and Punit P. Seth
IONIS Pharmaceuticals Inc., 2855 Gazelle Court, Carlsbad, CA 92010
*tprakash@ionisph.com, Tel 760-603-2590
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 61
1
2
3
4
ABSTRACT
5
6
7
8
9
The comprehensive structureꢀactivity relationships of triantennary GalNAc conjugated ASOs for
enhancing potency via ASGR mediated delivery to hepatocytes is reported. Seventeen GalNAc
clusters were assembled from six distinct scaffolds and attached to ASOs. The resulting ASO
conjugates were evaluated in ASGR binding assays, in primary hepatocytes and in mice. Five
structurally distinct GalNAc clusters were chosen for more extensive evaluation using ASOs
targeting SRBꢀ1, A1AT, FXI, TTR and ApoC III mRNAs. GalNAc ASO conjugates exhibited
excellent potencies (ED50 0.5–2 mg/kg) for reducing the targeted mRNAs and proteins. This
work culminated in the identification of a simplified trisꢀbased GalNAc cluster (THAꢀGN3),
which can be efficiently assembled using readily available starting materials and conjugated to
ASOs using a solution phase conjugation strategy. GalNAcꢀASO conjugates thus represent a
viable approach for enhancing potency of ASO drugs in the clinic without adding significant
complexity or cost to existing protocols for manufacturing oligonucleotide drugs
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

Page 3 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
INTRODUCTION
8
9
Antisense oligonucleotides bind their cognate mRNA and modulate its function to yield a
pharmacological response.1 A second generation gapmer ASO typically consists of a central
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
gap region of 8ꢀ14 DNA nucleotides adjoined on either end with 2’ꢀOꢀmethoxyethyl RNA
(MOE, Figure 1) nucleotides and phosphorothioate (PS) backbone chemistry.3 Kynamro, a
second generation ASO targeting Apolipoprotein Bꢀ100 mRNA, was recently approved by the
FDA for the treatment of homozygous familial hypercholesterolemia.4 Currently, there are more
than 35 ASOs advancing in clinic for the treatment of cardioꢀvascular, diabetes, cancer and
several rare and orphan disease indications.5,6 Significant number of these ASOs target mRNAs
expressed primarily in hepatocytes in the liver.6,7 Therefore, delivery approaches which enhance
ASO potency for suppressing gene targets of therapeutic interest expressed in hepatocytes are
beneficial.
The ASGR (asialoglycoprotein receptor)8 is abundantly expressed on hepatocytes
(~500,000 copies/cell)9,10 and it clears glycoproteins from serum by receptor mediated
endocytosis.11 The functional ASGR exhibits high affinity for Nꢀacetyl galactosamine (GalNAc)
terminated oligosaccharides and is conserved across all mammals.12 The receptor binding is
mediated by calcium13 and is sensitive to pH. The receptor ligand complex dissociates in acidic
pH within endosomes and receptor recycle back to the cell surface.14 ASGR has been exploited
for the targeted delivery of modified lipoprotein particles15, genes16, chemically modified
oligonucleotides17,18 siRNA19,20 and microRNA antagonists21 to hepatocytes in rodents.
Additionally, in an early stage human trials, siRNA conjugates targeting transthyretin (TTR)
mRNA, have exhibited good activity for reducing TTR protein in blood.22
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 61
1
2
3
4
5
6
7
8
9
We recently reported that targeted delivery of ASOs to hepatocytes using a previously
reported triꢀantennary GalNAc cluster attached to the 3’ꢀend of an ASO improved ASO activity
7ꢀ10 fold in rodents.23 Our initial studies were carried out using the TrisꢀGN3 cluster (Figure 2)
which has been successfully used for evaluation of siRNA GalNAc conjugates in rodents and in
humans and is attached to the 3’ꢀend of the oligonucleotide.20 The 3’ꢀTrisꢀGN3 oligonucleotide
conjugates were synthesized by preꢀloading an appropriately modified TrisꢀGN3 cluster onto
solidꢀsupport followed by automated oligonucleotide synthesis.19 While this approach was
efficient, it was not amenable for rapid evaluation of multiple GalNAc clusters. Furthermore, the
synthesis of the TrisꢀGN3 cluster can be cumbersome as additional steps are required for
attaching the propanediamine adaptor which links the GalNAcꢀacid moiety to the Trisꢀ
tricarboxylic acid scaffold.
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
To address these issues, we showed that attaching the TrisꢀGN3 cluster to the 5’ꢀend of
ASOs can be carried out efficiently using a solutionꢀphase conjugation reaction and that the 5’ꢀ
TrisꢀGN3 ASO was slightly more active than the corresponding 3’ꢀTrisꢀGN3 ASO conjugate in
mice.24 To determine if the structure of the TrisꢀGN3 cluster can be further simplified while
retaining ASGR binding and in vivo activity, we carried out a systematic structure activity
relationship (SAR) study to identify the optimal structural requirements of triantennary GalNAc
conjugates for enhancing ASO delivery to hepatocytes. As part of this effort, we examined the
ASGR binding, in vitro and in vivo activity of ASOs conjugated to several GalNAc clusters with
different linking scaffolds and tethers. In addition, we also examined the effect of ASO backbone
chemistry and the linker moiety on the potency of GalNAc conjugated ASOs. After initial
profiling of the GalNAc conjugates, five clusters were selected for more extensive evaluation
using multiple ASO sequences for potency and toxicity assessments. This effort led to the
ACS Paragon Plus Environment

Page 5 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
identification of several triantennary GalNAc clusters with simplified structures which are easier
to synthesize and enhance ASO potency >10ꢀfold in mouse models. Our work represents a
comprehensive study on the structure activity relationships of triantennary GalNAc conjugated
ASOs for enhancing potency by targeted delivery to hepatocytes in animals.
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
RESULTS AND DISCUSSION
Previous studies have described the SAR of multivalent galactose and GalNAc (Figure 1)
clusters derived from amino acid and dendrimer based scaffolds for ASGR binding.25ꢀ28 Affinity
of the sugar clusters for the ASGR was dependent on the nature of the sugar
(GalNAc>galactose),12 number of sugars (four=three>two>one),29ꢀ31 and on the length and
hydrophobicity of the spacer between the sugar moieties and the branching point of the
dendrites.25, 26 A galactose cluster with a 20 Å spacer exhibited 2000ꢀfold higher affinity for the
ASGR than galactose clusters lacking the spacer.26 These data suggested that the spatial
orientation of the sugars within the cluster was important for optimal ASGR binding. To
determine if these structural requirements were also valid for triantennary GalNAc clusters
attached to single stranded chemically modified ASOs, we examined GalNAc clusters assembled
on six distinct scaffolds. These scaffolds were abbreviated as Tris, Triacid, LysꢀLys, LysꢀGly,
Trebler and Hydroxyprolinol (Figure 2) based on the building blocks used to assemble the
clusters. For several of the scaffolds, we also varied the length and hydrophobicity of the linkers
used to attach the GalNAc sugars to the branching points on the scaffold.
Synthesis of GalNAc clusters. The general synthetic protocols for the preparation of the Tris,
Triacid, LysꢀLys, and hydroxyprolinol type GalNAc clusters (1ꢀ11, Scheme 1) have been
described previously and additional details are provided in the supporting information.23,26,32ꢀ34
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 61
1
2
3
4
5
6
7
8
9
Synthesis of the simplified Trisꢀbased cluster THAꢀGN3ꢀPfp 3 is shown in Scheme 2. The Nꢀ
acetylgalactosamine building block 12 (Scheme 2) was synthesized by glycosylation of known35
oxazoline 13 with benzyl (6ꢀhydroxyhexyl)carbamate 14 in presence of trimethylsilyl
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
trifluoromethanesulfonate (TMSOTf). The previously reported Tris based triantennary coreꢀunit
36
15 (Scheme 2)
was functionalized by monoꢀbenzyl protected glutaric acid 16 (Supporting
Information) under standard amino acid coupling conditions followed by treatment with
trifluoroacetic acid (TFA) to deprotect tertꢀbutyl ester yielded tris acid 17 (80%). The tris acid 17
was further reacted with pentafluorophenolꢀtrifluoracetate (PfpꢀTFA) in the presence of N,Nꢀ
diisopropylethylamine resulting in clean conversion to the corresponding Pfp ester 18 (83%).37
The tris Pfp ester 18 was reacted with GalNAc analog 12 in presence of palladium hydroxide in
THF under an atmosphere of hydrogen to yield THAꢀGN3 glutaric acid analog 19 (76%). The
THAꢀGN3 glutaric acid analog 19 was conveniently converted to its Pfp ester 3 by reacting with
PfpꢀTFA in the presence of triethylamine in high yield.
Synthesis of GalNAc-ASO conjugates. The general protocol for synthesis of the GalNAc
cluster ASO conjugates is depicted in Scheme 1.24 In brief, a solution of 10ꢀ12 mM 5’ꢀ
hexylamino ASO 20 (Scheme 1) in sodium tetraborate buffer (pH 8.5) was added to a solution of
triantennary GalNAc Pfp esters (2ꢀ3 mole equivalents, 30ꢀ40 mM) in DMSO and the resulting
solution was stirred at room temperature for 3 h. The 5’ꢀhexylamino ASO 20 was completely
consumed in 3 h as seen by LC MS analysis. The Oꢀacetyl groups from the GalNAc sugar were
removed by treating with aqueous concentrated ammonia (28ꢀ30 wt%) for 2ꢀ3 h followed by
HPLC purification to provide the triantennary GalNAc conjugated ASOs.
The synthesis of Trebler based GalNAc clusters (Figure 2, PGN3, PipꢀPGN3) was
accomplished according to reported procedures using GalNAc phosphoramidites 21 and 22
ACS Paragon Plus Environment

Page 7 of 61
Journal of Medicinal Chemistry
1
2
(Figure 3, Supporting Information).24 Hydroxyprolinol based GalNAc cluster conjugated ASOs
(Figure 2, 3’ꢀHPꢀHꢀGN3, 3’ꢀHPꢀDGN3, 3’ꢀHPꢀHAHꢀGN3) were synthesized on a DNA
synthesizer using the phosphoramidites 23ꢀ24 (Scheme 3) and 25 (Figure 3). Tetra acetylꢀ
GalNAc analogs 26 and 27 (Scheme 3, Supporting Information) were conveniently coupled to 6ꢀ
OꢀDMTꢀhydroxyprolinol23,34 28 (Scheme 3) under peptide coupling conditions to yield
compounds 29ꢀ30 (Scheme 3, Supporting Information). Compounds 29ꢀ30 were phosphitylated
to provide phosphoramidites 23ꢀ24 (Scheme 3, Supporting Information).
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
The phosphoramidites 23ꢀ25 were sequentially coupled three times on a UnyLinker™
solid support to create the hydroxyprolinolꢀbased trivalent GalNAc clusters (Figure 2).
Subsequently, DNA or MOE nucleoside phosphoramidites were coupled to the solid support to
assemble the ASO using standard oligonucleotide synthesis protocols.38 Interestingly, the
opposite configuration where the hydroxyprolinol GalNAc phosphoramidites were coupled to the
5’ꢀend of the ASO on solid support was not successful due to poor coupling efficiencies. After
completion of the synthesis, the solid support was treated with 50% triethylamine in acetonitrile
at room temperature to remove cyanoethyl group from internucleosidic linkages. Next, the solid
support bearing the ASO GalNAc cluster conjugate was treated with aqueous ammonia to release
the 3’ꢀhydroxyprolinol based GalNAc ASO conjugate from the solid support and to remove the
protecting groups from exocyclic amino groups of the nucleobases. ASO conjugates were
purified by HPLC on a strong anion exchange column and desalted by HPLC on a reverse phase
column. Purity and molecular weight of all GalNAcꢀASO conjugates were determined by ionꢀ
pair LCMS analysis (Supporting Information).
SAR of Tris Based Triantennary GalNAc Conjugated ASOs
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 61
1
2
3
4
5
6
7
8
9
We recently showed that TrisꢀGN3 (Figure 2) conjugation improved the potency of a 5ꢀ
10ꢀ5 MOE gapmer ASO 31 (Table 1) targeting scavenger receptor class B, member 1 (SRBꢀ1)
mRNA in mice.23,24 SRBꢀ1 is ubiquitously expressed in all cell types in the liver and in
extrahepatic tissues and ASOs targeting mouse SRBꢀ1was used as a model compound to study
the effect of chemical modification on ASO activity in animal models.39 We used the 5’ꢀTrisꢀ
GN3 SRBꢀ1 ASO 32 24 (Table 1) as a reference GalNAc cluster conjugate for our SAR study. To
study the effect of spacer length between the GalNAc sugar and the trisꢀscaffold on ASGPR
binding and activity in cells and in mice, we synthesized 5’ꢀTEAꢀGN3 ASO 33, 5’ꢀTBAꢀGN3
ASO 34, 5’ꢀTHAꢀGN3 ASO 35 and 5’ꢀTDAꢀGN3 ASO 36 with spacer lengths of 11.7 Å, 15.0
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
Å, 18.0 Å and 26.6 Å respectively (Table 1, Figure 2). The spacer length was defined as the
distance between the anomeric center of the sugar and the branching point on the cluster.26
ASOs 33ꢀ36 containing tris scaffold based GalNAc clusters with varying spacer lengths
were compared with standard 5’ꢀTrisꢀGN3 ASO 32 in biological assays (Table 1). First we tested
the ability of these GalNAc cluster conjugated ASOs to bind to the ASGR using a competition
binding assay.12,40 ASOs 33ꢀ35 containing 5’ꢀTEA, 5’ꢀTBA, THA GN3 exhibited similar binding
affinity (Table 1, Ki 6.1ꢀ7.0 nM) to ASGR relative to 5’ꢀTrisꢀGN3 ASO 32 (Table 1, Ki 8 nM)
while ASO 36 containing 5’ꢀTDAꢀGN3 with a longer tether (26.6 Å) bound to ASGR with a
slightly lower affinity (Ki 23 nM, Table 1) relative to 5’ꢀTrisꢀGN3 ASO 32.26
Next we ascertained the potency of ASOs 31ꢀ36 in primary mouse hepatocytes which
retain ASGR expression in cell culture.41 The unconjugated and triantennary GalNAc conjugated
ASOs 31ꢀ36 were delivered without any transfection agents.42 The triantennary GalNAc
modified ASOs 32ꢀ35 were more potent than unconjugated ASO 31 (Table 1). Consistent with
ACS Paragon Plus Environment

Page 9 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
ASGR binding data, ASO 36 containing the longer and more hydrophobic spacer was 4 fold less
potent than ASO 32.
7
8
9
We also evaluated the potency of SRBꢀ1 ASOs 32ꢀ36 in mice and compared with the
unconjugated ASO 31.23 Mice (C57BL/6, n=4/group) were injected subcutaneously with a single
dose of ASO 32ꢀ36 (0.5, 1.5, 5 and 15 mg kgꢀ1). Mice were sacrificed after 72 h, livers were
homogenized and analyzed for reduction of SRBꢀ1 mRNA. As expected, all GalNAc cluster
conjugated ASOs 32ꢀ36 showed improved potency (Table 1) relative to unconjugated ASO 31
(Table 1). ASOs 32ꢀ35 exhibited similar potency (ED50 2.4ꢀ 4.2 mg kgꢀ1, Table 1) and ASO 36
was less potent (ED50 7.2 mg kgꢀ1). ASOs used in this study were well tolerated with no
elevations in plasma transaminases or organ weights (data not shown).
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
SAR of Triacid Based Triantennary GalNAc Conjugated ASOs
The triacid scaffold (Figure 2) represents a simplified starting point for the synthesis of
triantennary GalNAc clusters using the commercially available and inexpensive
nitromethanetrispropionic acid starting intermediate.33 The triacidꢀscaffold is slightly more
hydrophobic as compared to the trisꢀscaffold as it lacks the oxygen atoms near the branching
point on the scaffold (Figure 2). 5’ꢀTAHꢀGN3 ASO 37, (spacer length 20.8 Å, Tableꢀ1, Figure
2), 5’ꢀTAPꢀGN3 ASO 38 (spacer length 19.1 Å), 5’ꢀTAꢀGN3 ASO 39, (spacer length 15.3 Å),
5’ꢀTABꢀGN3 ASO 40, (spacer length 12.7 Å) and 5’ꢀTAEꢀGN3 ASO 41 (spacer length 10.1 Å)
were synthesized using reported procedures.33
In general, ASOs containing 5’ꢀTriacid GN3 with varying spacer lengths exhibited
similar binding affinity (Ki 6.2ꢀ10 nM, Table 1) to ASGR relative to 5’ꢀTrisꢀGN3 ASO 32,
except for 5’ꢀTAB GN3 ASO 40 which showed slightly lower binding (Ki 26.0 nM, Table 1,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 61
1
2
3
4
5
6
7
8
9
Figure 2). The potency of ASOs 37-41 in mouse primary hepatocytes was determined using the
procedure described for the trisꢀconjugates.42 The GalNAc cluster modified SRBꢀ1 ASOs 37-41
were more potent (Table 1, Figure 2) than the parent ASO 31 (Figureꢀ1, Table 1). We also
evaluated the potency of SRBꢀ1 ASOs 37-41 in mice. ASOs 37-41 were injected subcutaneously
to mice (C57BL/6, n=4/group) with a single dose of 0.5, 1.5, 5 and 15 mg kgꢀ1and mice were
sacrificed after 72 h of dosing. The livers were homogenized and analyzed for reduction of SRBꢀ
1 mRNA. Animals treated with ASOs 37-41 exhibited similar potency (ED50 2.4ꢀ3.8 mg kgꢀ1,
Tableꢀ1) to 5’ꢀGN3 ASO 32. There was no elevation in plasma transaminases or organ weights
(data not shown) in the animals treated with these ASOs.
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
Peptide Scaffold Based Triantennary GalNAc Conjugated ASOs
Bifunctional amino acids such as glutamic acid, aspartic acid or lysine have been used as
branching elements for the construction of multiantennary GalNAc ligands.32 Valentijin et al.
showed that trivalent GalNAc clusters derived from bisꢀlysine scaffolds show strong binding to
the ASGR in competition binding binging assays.32 To determine if GalNAc clusters based on
amino acid scaffolds could be used for targeting ASOs to hepatocytes, we synthesized 5’ꢀLysꢀ
LysꢀHꢀGN3 ASO 42 (Table 1, Figure 2) using the solution phase conjugation strategy described
in Scheme 1. Another triantennary GalNAc cluster which binds the ASGR with high affinity was
based on 6ꢀaminohexyl GalNAc assembled on a lysineꢀglycine scaffold (Figure 2, LysꢀGlyꢀHAꢀ
GN3).43 Solution phase conjugation strategy (Scheme 1) using Pfp esters of corresponding GN3
10 and 11 (Scheme 1) was used for the synthesis of ASOs 42 and 43 (Table 1).24
As expected 5’ꢀpeptide scaffold based triantennary GalNAc ASO conjugates 42-43 with
varying spacer lengths, exhibited similar binding affinity (Ki 6.3ꢀ10.1 nM, Tableꢀ1, Figure 2) to
ASGR relative to 5’ꢀTrisꢀGN3 ASO 32 (Table 1). Consistent with the binding affinity,
ACS Paragon Plus Environment

Page 11 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
conjugation of peptide scaffold based triantennary GalNAc improved potency of SRBꢀ1 ASOs
42-43 (IC50 6ꢀ30 nM, Table 1, Figure 2) relative to unconjugated SRBꢀ1 ASO 31 (IC50 250 nM,
Table 1) in mouse primary hepatocytes.42 The ASOs 42-43 also demonstrated superior potency
(ED50 2.1ꢀ2.3 mg kgꢀ1, Table 1) in mice relative to unconjugated ASO 31 (ED50 18.3 mg kgꢀ1,
Table 1). All ASOs were well tolerated with no elevations in plasma transaminases or organ
weights (data not shown).
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
Trebler Based Triantennary GalNAc Conjugated ASOs
We also developed a strategy for solid phase assembly of GalNAc clusters using well
established phosphoramidite chemistry. We hypothesized that the synthesis of triantennary
GalNAc ASO conjugates could be simplified by using this approach.44 In addition, this approach
provides an opportunity to examine the effect of introducing a negatively charged phosphodiester
group in the tether of the GalNAc cluster on ASGR binding. In brief, this approach involved
sequential coupling of a commercially available trebler phosphoramidite to assemble the
triantennary scaffold, followed by coupling of the GalNAc phosphoramidite 21 (Figure 3) to
provide 5’ꢀPGN3 SRBꢀ1 ASO 44 (Table 1, Figure 2).
The effect of varying spacer length of triantennary GalNAc cluster has been reported.26
However, the effect of restricting the spacer flexibility of trivalent GalNAc cluster has not been
characterized. In order to understand the effect of restricting flexibility of the spacer between the
GalNAc sugar and the branching point on ASGR binding and biological activity we designed 5’ꢀ
PipꢀPGN3 ASO 45 (Table 1, Figure 2). Phosphoramidite based GalNAc cluster synthesis44 on a
solid support was used to synthesize ASO 45 using GalNAc piperidine phosphoramidite 22
(Figure 3, Supporting Information).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 61
1
2
3
4
5
6
7
8
9
ASOs 44-45 containing 5’ꢀPGN3 and 5’ꢀPipꢀPGN3 respectively exhibited similar
binding affinity (Ki 20ꢀ23 nM, Tableꢀ1) to ASGR. However, ASOs 44-45 exhibited weaker
ASGR binding relative to 5’ꢀTrisꢀGN3 ASO 32 (Ki 8 nM, Tableꢀ1). Interestingly, ASOs 44-45
showed differences in their ability to inhibit mRNA expression in hepatocytes and in mouse liver
(Table 1). Both ASOs were more potent than unconjugated ASO 31 (Table 1). The ASO 45
containing 5’ꢀPipꢀPGN3 with a constrained spacer was 3ꢀ4 fold less potent in hepatocytes (IC50
150 nM, Table 1) and in animals (ED50 9.8 mg kgꢀ1, Table 1) relative to ASO 44 (Table 1) as
well as ASO 32 containing 5’ꢀTrisꢀGN3. These ASOs were well tolerated with no elevations in
plasma transaminases or organ weights (data not shown).
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
Hydroxyprolinol Based Triantennary GalNAc Conjugated ASOs
We also explored another approach where a serial incorporation of monomer GalNAc
provides the required triantennary GalNAc structure as shown in ASOs 46ꢀ48 (Table 1, Figure
2) using nonꢀnucleoside hydroxyprolinol GalNAc phosphoramidites 23-24 (Scheme 3) and 25
(Figure 3).34,45 This approach eliminates the need for a multistep synthesis of triantennary
GalNAc or preꢀloading the solid support with the bulky triantennary GalNAc for solidꢀphase
oligonucleotide synthesis. Furthermore, this approach allows investigation of linear arrangement
of GalNAc sugars as opposed to the branched arrangement with other clusters.
The biological properties of the hydroxyprolinol GN3 containing ASOs 46ꢀ48 (Table 1)
were studied.24 The ASOs 46ꢀ48 showed differences in their binding to the ASGR.40,12
Consistent with GalNAc cluster ASOs based on tris scaffold (ASO 36, Table 1) binding affinity
of ASO 47 (Ki 40 nM, Figure 1) containing longer spacer chain between GalNAc sugar and the
originating point was weaker than ASO 46 (Ki 12 nM, Table 1). ASO 48 with a similar spacer
ACS Paragon Plus Environment

Page 13 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
chain length as the 3’ꢀHPꢀDꢀGN3 ASO 47 (Table 1, Figure 3) but with an additional amide
linkage binds tighter (Ki 9.6 nM) to ASGR, suggesting that enhancing hydrophilicity of long
aliphatic spacers improves binding of triantennary GalNAc ASOs to ASGR.
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
The potency of ASOs 46ꢀ48 were determined in mouse primary hepatocytes using the
protocol described previously. Consistent with other classes of triantennary GalNAc conjugated
ASOs, the ASOs 46ꢀ48 (IC50 30ꢀ55 nM, Table 1) showed improved potency in mouse primary
hepatocytes relative to unconjugated ASO 31 (Table 1, IC50 250 nM). Similarly, ASOs 46ꢀ48
were more potent (ED50 2.2 to 4.8 mg kgꢀ1, Table 1) in mouse liver relative to unconjugated ASO
31 (ED50 18.3 mg kgꢀ1, Table 1).
Effect of ASO Backbone Chemistry on the Biological Property of 5’-Tris-GN3, 5’-THA-
GN3, 5’-TA-GN3 and 5’-Lys-Lys-H-GN3 Conjugated ASOs
Phosphorothioate modified (PS, Figure 1) ASOs binds more avidly to proteins than
phosphodiester (PO, Figure 1) ASOs.46 PS ASOs are highly (90ꢀ98%) protein bound in plasma
which facilitates ASO delivery to peripheral tissues.47 However, nonꢀspecific protein binding has
been implicated in some of the proꢀinflammatory effects of PS ASOs.48 We determined if
reducing the PS content of GalNAcꢀconjugated ASOs, by replacing some of the PS linkages in
the MOE wings with PO linkages (termed mixedꢀbackbone MBB ASOs) could be beneficial.
MBB ASOs bind less avidly to plasma and cellꢀsurface proteins, and as a consequence, are
eliminated more readily in the urine resulting in 2ꢀ3 fold reduced activity in animals (data not
shown). We proposed that attaching trivalent GalNAc clusters to MBB ASOs could enhance
delivery to hepatocytes. Furthermore, the MBB design used was expected to be metabolically
stable, as the more labile DNAꢀgap region and both ends of the ASO were protected using PS
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 61
1
2
3
4
5
6
7
8
9
linkages. We selected TrisꢀGN3, THAꢀGN3, TAꢀGN3 and LysꢀLysꢀHꢀGN3 (Figure 2) to profile
the effect of GalNAc conjugation on the potency of MBB ASOs in primary hepatocytes and in
mice.
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 5’ꢀhexylaminoꢀmodified MBB SRBꢀ1 ASO (six PO and 13 PS linkages) was
conjugated to 5’ꢀTrisꢀGN3 49 (Table 2), 5’ꢀTHAꢀGN3 50 (Table 2), 5’ꢀTAꢀGN3 51 (Table 2)
and 5’ꢀLysꢀLysꢀHꢀGN3 52 (Table 2) clusters using the solution phase approach (Scheme 1).
Purity and mass of ASOs 49ꢀ52 were determined by ionꢀpair LCMS analysis (Supporting
Information).
We compared the potency of the MBB ASOs with the corresponding full PS ASOs 32,
35, 39 and 42 (Table 2) in mouse primary hepatocytes and in the liver. We found that the MBB
ASOs 49-52 show similar or slightly enhanced potency compared to the corresponding full PS
ASOs (ASOs 32, 35, 39 and 42 Tableꢀ2) in mouse primary hepatocytes and in mouse liver.
These data suggest that conjugation of trivalent GalNAc clusters to MBB ASOs can be beneficial
for reducing ASO PS content without compromising potency. These designs have the potential to
mitigate the nonꢀspecific protein binding properties and any resulting hybridizationꢀindependent
toxicities of full PS ASOs for clinical applications.49
5’-Tris-GN3, 5’-THA-GN3, 5’-TA-GN3 and 5’-Lys-Lys-H-GN3 Conjugation Improves
Potency of MOE gapmer ASO Targeting Mouse A1AT
We next examined the effect of conjugating selected triantennary GalNAc clusters on the
potency of a MBB ASO targeting alphaꢀ1 antitrypsin (A1AT) mRNA in mouse liver. ASOs
targeting A1AT were recently evaluated for the treatment of alphaꢀ1 antitrypsin deficiency
(AATD) which results from accumulation of mutant A1AT protein in liver.50 A MOE gapmer
ACS Paragon Plus Environment

Page 15 of 61
Journal of Medicinal Chemistry
1
2
3
4
ASO 53 (Figure 4) targeting mouse A1AT was used as the unconjugated control ASO for this
5
6
study.50 Synthesis of
5’ꢀTrisꢀGN3, 5’ꢀTHAꢀGN3, 5’ꢀTAꢀGN3 and 5’ꢀLysꢀLysꢀHꢀGN3
7
8
9
conjugated A1AT ASOs 54ꢀ57 (Figure 4) were accomplished by a solution phase coupling of
corresponding Pfp esters to 5’ꢀhexylamino ASO (Scheme 1). ASOs 54ꢀ57 were characterized by
ionꢀpair LCMS analysis (Supporting Information).
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
Mice (C57BL/6, 4/group) were injected subcutaneously with A1AT ASO 53 (5, 15 and
45 mg kgꢀ1, Figure 4) and corresponding various GalNAc cluster conjugated ASOs 54ꢀ57 (0.6, 2,
6 and 18 mg kgꢀ1, Figure 4) once a week for three weeks. Mice were sacrificed 72 h after the last
injection and liver and plasma were analyzed for reductions in expression of A1AT mRNA and
protein respectively (Figure 4). We observed a 10ꢀfold enhancement in potency for inhibiting
A1AT mRNA and protein expression in animals dosed with GN3 ASOs 54ꢀ57 (Figure 4, ED50
1.9ꢀ3.2 mg kgꢀ1) relative to unconjugated ASO 53 (ED50 25ꢀ27 mg kgꢀ1, Figure 4). The level of
plasma transaminases or blood urea nitrogen and organ or bodyꢀweights of animals treated with
these ASOs were normal (data not shown).
5’-Tris-GN3, 5’-THA-GN3, 5’-TA-GN3 and 5’-Lys-Lys-H-GN3 Conjugation Improves
Potency of MOE Targeting Mouse Factor XI
Next we determined the effect of conjugating 5’ꢀTrisꢀGN3, 5’ꢀTHAꢀGN3, 5’ꢀTAꢀGN3
and 5’ꢀLysꢀLysꢀHꢀGN3 on the potency of a 5ꢀ10ꢀ5 MOE gapmer ASO 58 (Figure 5) targeting
mouse coagulation factor XI (FXI). Here we used a 3ꢀweek dosing schedule which helps
understand the effect of drug accumulation and metabolism on ASO activity. FXI is a member of
the intrinsic coagulation pathway and is expressed specifically in hepatocytes and secreted in
blood. Inhibiting FXI activity shows benefits in animal models of thrombosis with a minimal risk
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 61
1
2
of bleeding.51ꢀ53 Trivalent GalNAc conjugated FXI ASOs 59ꢀ62 (Figure 5) were synthesized by
conjugation of corresponding trivalent GalNAc Pfp esters to 5’ꢀhexylamino conjugated ASO in
solution (Scheme 1). Purity and mass of the ASOs 59ꢀ62 were determined by ionꢀpair LCMS
analysis (Supporting Information).
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
Balbꢀc mice (n=4/group) were injected subcutaneously with 1, 3, 10 and 30 mg kgꢀ1 of
unconjugated full PS ASO 58 (Figure 5) and 0.08, 0.23, 0.7, 2 and 6 mg kgꢀ1 of 5’ꢀtrivalent
GalNAc conjugated mixed backꢀbone ASOs 59ꢀ62 (Figure 5) once a week for three weeks. FXI
mRNA expression was analyzed from the liver tissues of the treated mice. Trivalent GalNAc
conjugated ASOs 59ꢀ62 (ED50 0.4ꢀ0.6 mg kgꢀ1, Figure 5) showed >20 fold enhancement in
potency for reducing FXI mRNA levels in liver relative to unconjugated full PS ASO 58 (8.7 mg
kgꢀ1, Figure 5). Consistent with other studies all ASOs were well tolerated with no elevations in
serum transaminases or organ or body weight changes (data not shown).
SAR to Identify Optimized Linker Chemistry for Attaching Triantennary GalNAc to ASOs
We previously showed that the 3’ꢀTrisꢀGN3 cluster is detached by metabolism to liberate
the parent ASO in the liver.23 In that study we used a POꢀlinked deoxyadenosine (dA) nucleotide
as a linker moiety between the GalNAc cluster and the ASO to facilitate metabolism. Further
studies showed that the first site of metabolism of GalNAcꢀASO conjugates was the glycosidic
bond to the GalNAc sugars suggesting that the dA linker moiety is not required. To further probe
the importance of the PO dA linker moiety on potency, we evaluated ASO conjugates 63, 64 and
65 (Table 3) where the PO dA was replaced with PO dT, PO MOE A and PO MOE T
respectively. In addition, we also evaluated ASO conjugate 66 (Table 3) where the TrisꢀGN3
cluster was directly attached to the ASO via a 5’ꢀhexylamino PO linkage.
ACS Paragon Plus Environment

Page 17 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ASOs 63ꢀ66 showed similar potency for reducing SRBꢀ1 mRNA in mouse primary
hepatocytes. Interestingly, 5’ꢀTrisꢀGN3 ASO 66 (IC50 9 nM, Table 3) with only the hexylamino
PO linker between the GalNAc cluster and the ASO showed improved potency relative to the
other ASOs tested in the study. Next we tested the potency of 63-66 (Table 3) for inhibiting
SRBꢀ1 mRNA expression in mouse liver. Mice (C57BL/6, n=4/group) were injected
subcutaneously with a single dose of ASOs 63-66 (0.5, 1.5, 5 and 15 mg kgꢀ1). After 72 h of
dosing mice were sacrificed and analyzed for reduction of SRBꢀ1 mRNA expression in liver. All
the trisꢀGN3 ASO conjugates exhibited similar potency (ED50 1.8ꢀ2.8 mg kgꢀ1, Tableꢀ3)
suggesting that the PO dA or other nucleotide linker moieties between the GalNAc cluster and
the ASO are not required for optimal potency enhancement in mice and just a hexylamino PO
linkage is sufficient.
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’-Tris-GN3, 5’-THA-GN3, 5’-TA-GN3, 5’-Lys-Lys-H-GN3 and 3’-HP-H-GN3 Conjugation
Improves Potency of ASO Targeting Human TTR in Transgenic Mice.
We next studied the effect of conjugating triantennary GalNAc clusters on the potency of
a MOE gapmer ASO 67 (Figure 6) targeting human Transthyretin (TTR) in transgenic mouse
models. The unconjugated ASO 67 is currently being evaluated in humans for the treatment of
TTRꢀassociated polyneuropathy and has shown ~75% reduction in TTR protein in a phase 1 trial
at a dose of 300 mg/week (~4 mg/kg/week).54 Triantennary GalNAc conjugated ASOs 68-72
(Figure 6) were synthesized using similar procedure described in the previous sections. ASOs
68ꢀ72 were characterized by ionꢀpair LCMS analysis (Supporting Information). In ASO 68-72
we used only the hexylamino PO linker for attaching the triantennary GalNAc at the 5’ꢀend of
the ASOs.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 61
1
2
3
4
5
6
7
8
9
Human transgenic mice (TTRꢀIle84Ser, n=4/group) were injected subcutaneously with
the full PS ASO 67 (6, 20 and 60 mg kgꢀ1, Figure 6) and 5’ꢀtrivalent GalNAc conjugated mixed
backbone ASOs 68-72 (0.6, 2 and 6 mg kgꢀ1) once a week for three weeks. Mice were sacrificed
72 h after last injection and analyzed for reduction of hTTR mRNA in liver and for reduction of
hTTR protein in mouse plasma relative to saline treated control mice. All trivalent GalNAc
conjugated mixed backꢀbone ASOs 68-72 (ED50 0.6ꢀ0.8 mg kgꢀ1, Figure 6) showed >20 fold
enhancement in potency for reducing hTTR mRNA in liver and hTTR protein (ED50 1.3ꢀ1.6 mg
kgꢀ1) in mouse plasma relative to unconjugated full PS ASO 67 (Figure 6). ASOs 67ꢀ72 were
well tolerated with no elevations in serum transaminase levels or organ weight changes
(Supporting Figure S1ꢀS7).
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
To ascertain if the improved potency was a result of enhanced delivery to liver, tissue
concentration of TTR ASOs 67ꢀ72 in the livers of treated mice were quantitated.23,55,56 We
observed increased liver exposure of trivalent GalNAc conjugated ASOs 68-72 (95.6 to 125.9 µg
gꢀ1 tissue, Table 4) relative to unconjugated ASO 67 (61.6 µg gꢀ1 tissue, Table 4). The 5’ꢀ
GalNAc ASO conjugates 68-71 with PO linkers were fully metabolized to release the parent
ASO 67. These data further established that a hexylamino PO linker between the GalNAc cluster
and the ASO is sufficient to liberate the parent ASO by metabolism for 5’ꢀGalNAc ASO
conjugates. In contrast, the 3’ꢀHPꢀHꢀGN3 ASO 72 was not completely metabolized to the parent
ASO 67. Analysis of livers from mice treated with ASO 72 showed that 20ꢀ30% of the extracted
oligonucleotide had remnants of various linker moieties still attached to the ASO, thus making
these designs less interesting.
Toxicological Evaluation of 5’-Tris-GN3, 5’-THA-GN3, 5’-TA-GN3, 5’-Lys-Lys-H-GN3
and 3’-HP-H-GN3 Conjugated ASO Targeting Mouse ApoC III
ACS Paragon Plus Environment

Page 19 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Next we evaluated nonclinical safety of 5’ꢀTrisꢀGN3, 5’ꢀTHAꢀGN3, 5’ꢀTAꢀGN3, 5’ꢀLysꢀLysꢀ
HꢀGN3 and 3’ꢀHPꢀHꢀGN3 conjugated ASOs in mice. First we synthesized 5’ꢀTrisꢀGN3, 5’ꢀ
THAꢀGN3, 5’ꢀTAꢀGN3, 5’ꢀLysꢀLysꢀHꢀGN3 and 3’ꢀHPꢀHꢀGN3 conjugated ASOs 74-78 (Figure
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
7) of unconjugated ASO 73 (Figure 7) targeting mouse apolipoprotein CꢀIII (apoCꢀIII) mRNA.
We studied the potency of ASOs 74ꢀ78 to inhibit apoCꢀIII mRNA in mice and compared
with unconjugated apoCꢀIII ASO 73 (Figure 7). Mice (C57BL/6, n=4/group) were injected
subcutaneously with a single dose of ASO 73 (2, 6, 20 and 60 mg kgꢀ1) and ASOs 74-78 (0.6, 2,
6 and 20 mg kgꢀ1). Mice were sacrificed after 72 h, livers were homogenized and analyzed for
reduction of apoCꢀIII mRNA. As expected, ASOs 74-78 containing structurally different
GalNAc clusters (ED50 6.1ꢀ7.3 mg kgꢀ1, Figure 7) were 7ꢀ8 fold more potent than unconjugated
ASO 73 (ED50 50.0 mg kgꢀ1, Figure 7).
Subsequent, we assessed the nonꢀclinical safety of unconjugated ASO 73 and trivalent
GalNAc conjugated ASOs 74-78 (Figure 7) in mice. Mice (Male CDꢀ1, n = 8/group) were
injected subcutaneously with a dose of unconjugated ASO 73 100 mg kgꢀ1 and equal mole of
ASO 74-78 (130 mg kgꢀ1) week 0 loading dose day 1, 3 and 5 then once a week for 6 weeks. At
termination of the study, the various endpoints were assessed 48 h after the last dose. The ASOs
were administered at doses far in excess to relevant therapeutic doses. ASO 73 decreased liver
apoCꢀIII mRNA level down to 16% of PBS, whereas the trivalent GalNAc conjugated ASOs
decreased liver apoCꢀIII mRNA level down to 4 to 6% of PBS level (pꢀValue <0.01). While
trivalent GalNAc conjugated apoCꢀIII ASOs were more potent than unconjugated ASO 73, all
the GalNAc cluster conjugated ASOs 74ꢀ78 were as well tolerated as ASO 73 causing minimal
changes, none of them were considered adverse (Figure 8). The weight of the liver, kidneys,
spleen and heart relative to the body weight of animals treated with ASOs 74ꢀ78 were not
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 61
1
2
3
4
5
6
7
8
9
significantly changed compared to ASO 73 (Figure 8). None of the GalNAc conjugated ASOs
caused any significant change in liver function measured at 6 weeks, 48 h post dose alanine
transaminase (ALT) and aspartate aminotransferase (AST) (Figure 8). There was no significant
evidence of inflammation caused by ASOs 74ꢀ78 compared to ASO 73 as measured by changes
in white blood cell count (WBC) or the chemokine MIPꢀ1β plasma level (Figure 8).
Histopathological assessment (data not shown) of the liver and kidneys further demonstrated that
the trivalent GalNAc conjugated ASOs were all well tolerated at this dose level and clearly
shows an improvement in therapeutic index over unconjugated ASO 73.
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
CONCLUSIONS
In conclusion, we report the detailed structureꢀactivity relationships for enhancing ASO
potency via ASGRꢀmediated delivery of triantennary GalNAcꢀASO conjugates to hepatocytes.
GalNAc clusters assembled from six distinct branched or amino acid scaffolds (abbreviated Tris,
Triacid, LysꢀLys, LysꢀGly, Trebler and Hydroxprolinol) were synthesized and attached to ASOs
using simplified solutionꢀphase or phosphoramidite based methods. Within each cluster, the
length and hydrophobicity of the tether attaching the GalNAc sugar to the branching point on the
scaffold was varied to determine the optimal tether length and linker chemistry.
The GalNAc ASO conjugates were evaluated for ASGR binding and for activity in cell
culture and in mice where tethers as short as 6 atoms (10.1 Å) were tolerated but longer
hydrophobic tethers reduced ASGR affinity and biological activity by 2ꢀ4 folds. These results
were somewhat surprising and contrary to previously established rules for ASGR binding which
suggested that tether lengths of ~20 Å are optimal. We also found that the amide linkages in the
tethers of GalNAc conjugates can be replaced with negatively charged phosphodiester linkages
ACS Paragon Plus Environment

Page 21 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
without adversely affecting ASGR binding or activity of the conjugates. While conceptually
attractive, these designs generally reduced synthesis yields because of incomplete couplings (5’ꢀ
Trebler) or metabolism (3’ꢀhydroxyprolinol) and were less preferred.
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
We also evaluated if GalNAc conjugation could enhance the activity of mixedꢀbackbone
(MBB) ASOs in mice where some of the PS linkages from the wing region were replaced with
PO. MBB ASOs are generally less active in animals as reducing PS content lowers ASO avidity
for plasma and cellꢀsurface proteins which is important for promoting ASO uptake into cells and
tissues. MBB ASO designs have the potential to mitigate hybridizationꢀindependent toxicities
resulting from nonꢀspecific interactions of ASOs with certain cellular proteins.49 We found that
GalNAc conjugates of MBB ASOs show similar or enhanced activity compared to full PS
designs and thus provide chemical design alternatives for enhancing tolerability while improving
activity. Further refining of the linker moiety between the GalNAc cluster and the ASO showed
that GalNAc clusters linked to ASOs via a simple hexylamino phosphodiester linkage were
optimal.
After the initial profiling studies, five structurally distinct GalNAc clusters were chosen
for more extensive evaluation using multiple ASO sequences targeting mouse alphaꢀ1ꢀantitrypsin
(A1AT), mouse coagulation Factor XI, mouse apoCꢀIII and human TTR in wildꢀtype or
transgenic mice. Most GalNAc ASO conjugates exhibited excellent potencies (ED50 0.5–2
mg/kg) for reducing the targeted mRNA and protein using single and multiple dosing schedules.
Selected GalNAc ASO conjugates were also well tolerated at high doses (130 mg/kg/week/6
weeks) in mouse models used for assessing the potency and toxicity of ASO drug candidates.
This work culminated in the identification of a simplified trisꢀbased GalNAc cluster (THAꢀGN3,
Figure 9), which can be efficiently assembled using readily available starting materials and
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 61
1
2
3
4
5
6
7
8
9
conjugated to ASOs using a solution phase strategy. A THAꢀGN3 conjugate of an ASO targeting
apolipoprotein(a) was recently shown to be 30ꢀfold more potent in humans relative to the parent
ASO.57 GalNAcꢀASO conjugates thus represent a viable strategy for enhancing ASO potency in
the clinic without adding significant complexity or cost to existing protocols for oligonucleotide
synthesis and manufacturing.
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
EXPERIMENTAL SECTION
Synthesis of Conjugation of 5’-trivalent GalNAc conjugated ASOs in solution. To a solution
of 5’ꢀhexylamino ASO in 0.1 M sodium tetraborate buffer, pH 8.5 (2 mM) a solution of GalNAc
PFP ester 1-11 (3 mole equivalent, Scheme 1) dissolved in DMSO (40 mM) was added and the
reaction mixture was stirred at room temperature for 3h. Concentrated aqueous ammonia was
added (5 x reaction volume) and stirred at room temperature for 4 h. Reaction mixture
concentrated under reduced pressure and residue dissolved in water and purified by HPLC on a
strong anion exchange column (GE Healthcare Bioscience, Source 30Q, 30 µm, 2.54 x 8 cm, A =
100 mM ammonium acetate in 30% aqueous CH3CN, B = 1.5 M NaBr in A, 0ꢀ60% of B in 60
min, flow 14 mL minꢀ1). The residue was desalted by HPLC on reverse phase column to yield
the 5’ꢀtrivalent GalNAc conjugated ASOs in an isolated yield of 62ꢀ80%. The ASOs were
characterized by ionꢀpairꢀHPLCꢀMS analysis with Agilent 1100 MSD system.
Synthesis of 3’-HP-H-GN3 ASOs (46, 72, 78) 3’-HP-D-GN3 ASO (47) and 3’-HP-HAH-
GN3 ASO (48): Antisense oligonucleotides were synthesized at 40 ꢁmol scale using
UnyLinker™ solid support functionalized by modified nucleoside or GalNAc cluster. For the
synthesis of 3’ꢀHPꢀHꢀGN3 ASOs 46, 72, 78 0.1 M solution of GalNAc phosphoramidite 23
(Scheme 3) and for the synthesis of ASO 47 GalNAc phosphoramidite 24 (Scheme 3) and for
ASO 49 phosphoramidite 25 (Figure 4) in 40% dichloromethane in acetonitrile was used. First
ACS Paragon Plus Environment

Page 23 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
phosporamidites 23ꢀ25 sequentially coupled three times on a UnyLinker™ solid support and
subsequently required nucleoside phosphoramidites were coupled to synthesize the 3’ꢀtrivalant
GalNAc conjugated ASOs. A solution of all other phosphoramidites in acetonitrile (0.1 M), and
standard oxidizing and capping reagents were used. An extended coupling time of 5 minutes
was used for incorporation of the 2’ꢀOꢀMOE modified nucleotides. For each of the modified
analogs 4ꢀfold excess of 2’ꢀOꢀMOE modified nucleoside 3’ꢀphosphoramidite and GalNAc
phosphoramidites 23ꢀ25 were delivered with a 12 minute coupling time. The 5′ꢀend
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
dimethoxytrityl group was left on to facilitate purification.
Postꢀsynthetically, all
oligonucleotides were treated with 1:1 triethylamine: acetonitrile to remove cyanoethyl
protecting groups from the phosphorothioate linkages. Subsequently, solid support bearing
ASOs were treated with aqueous NH4OH (28ꢀ30 wt%) at 55 ºC for 4 h and then cooled and
o
added 10% (V/V) of 40% methylamine in water. Heating at 55 C was continued for additional
12ꢀ14 h to cleave trivalent GalNAc conjugated ASOs from support, remove protecting groups,
and hydrolyze the UnyLinker™ moiety. Oligonucleotides were purified by ionꢀexchange
chromatography using a gradient of NaBr across a column packed with Source 30Q resin as
described before. Pure fractions were desalted using HPLC on a reverse phase column. Purity
and mass of oligonucleotides were determined using ionꢀpair LCMS analysis (Supporting
Information).
ASGR Binding Assay. Binding affinity was measured using a competition binding assay using
the procedure reported.23 In brief α1ꢀacid glycoprotein (AGP) (4 mg, 100 nmol) was
desialylated by treatment with 50 mM sodium acetate buffer (pH 5) and neuraminidaseꢀagarose
(1 U) for 16 h at 37 oC. Then the protein was iodinated by using published procedure.58 AGP (1
mg of 1 mg/mL AGP) was mixed with 1 M glycine in 0.25 M NaOH (0.2 mL, pH 10) and added
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 61
1
2
to a solution containing 10 mM ICl (7 µL), Na125I (2.5 µL) and 1 M glycine in 0.25 M NaOH (25
µL). Mixture was incubated for 10 min at room temperature and free 125I was removed by
concentrating sample twice using a 3 kDa molecular weight cutꢀoff spin column. Protein labeling
and purity was determined using a HPLC system equipped with an Agilent SECꢀ3 column (7.8 x
300 mm) and a βꢀRAM counter. Freshly isolated mouse hepatocytes (106 cells/mL) were plated
in on 6ꢀwell plates in 2 mL of William’s medium containing 10 % fetal bovine serum (FBS), 1X
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
o
nonꢀessential amino acids and 1X sodium pyruvate. Cells were incubated for 16–20 h at 37 C
with 5 % CO2 and washed with medium without FBS prior to binding experiment. Cells were
o
incubated for 30 min at 37 C with 1 mL of competition medium containing 2 % FBS, 10 nM
125Iꢀlabeled AGP and GalNAc conjugated ASOs at concentrations ranging from 10ꢀ11 to 10ꢀ5 M.
Nonꢀspecific binding was determined in the presence of 10 mM GalNAc free acid. Cells were
washed twice with media without FBS to remove unbound 125I–labeled competitor GalNAc ASO
and then lysed using Qiagon’s RLT buffer containing 1 % mercaptoethanol. Lysates were briefly
freezed (10 min) and then thawed and transferred to round bottom assay tubes and radioactivity
measured by a γ counter. Nonꢀspecific binding was subtracted before dividing 125I γ counts by
the value of the lowest GalNAc conjugated ASO concentration counts. The inhibition curves
were fitted according to a single site competition binding equation using a nonꢀlinear regression
algorithm.
Cell Culture Study. Freshly isolated mouse hepatocytes were placed in wells with growth
medium containing 10 % FBS, antiꢀanti, HEPES, glutamine and varying amounts of ASO or
o
PBS control. Cells were maintained at 37 C and 5 % CO2 for 16 h and then washed with PBS
and lysed. RNA was extracted using Qiagen RNease kit and RNA levels determined by Taqman
qꢀrtPCR using the primers: 5’ꢀTGACAACGACACCGTGTCCTꢀ3’ (forward primer), 5’ꢀ
ACS Paragon Plus Environment

Page 25 of 61
Journal of Medicinal Chemistry
1
2
3
4
ATGCGACTTGTCAGGCTGGꢀ3’
(reverse
primer)
and
5’ꢀ
5
6
7
8
CGTGGAGAACCGCAGCCTCCATTꢀ3’ (probe). RNA was normalized to total RNA using
Ribogreen and all data were performed in triplicate.
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
Animal Treatment. Animal experiments were conducted in accordance with the American
Association for the Accreditation of Laboratory Animal Care guidelines and were approved by
the Animal Welfare Committee (Cold Spring Harbor Laboratory’s Institutional Animal Care and
Use Committee guidelines). The animals were housed in microꢀisolator cages on a constant 12ꢀ
hour lightꢀdark cycle with controlled temperature and humidity and were given access to food
and water ad libitum. Blood was collected by cardiac puncture exsanguination with K2ꢀEDTA
(Becton Dickinson Franklin Lakes, NJ) and plasma separated by centrifugation at 10,000 rcf for
4 min at 4°C. Plasma transaminases were measured using a Beckman Coulter AU480 analyzer.
Tissues were collected, weighed, flash frozen on liquid nitrogen and stored at ꢀ60°C. Reduction
of target mRNA expression was determined by real time RTꢀPCR using 7700 RTꢀPCR sequence
detector (Applied Biosystems). Briefly, RNA was extracted from about 50ꢀ100 mg tissue from
each mouse using PureLink Pro 96 Total RNA Purification Kit (LifeTechnologies, Carlsbad,
CA) and mRNA was measured by qRTꢀPCR using Express OneꢀStep SuperMix qRTꢀPCR Kit
(Life Technologies, Carlsbad, CA). Primers and probes for the PCR reactions were obtained
from Integrated DNA technologies (IDT). The assay is based on a targetꢀspecific probe labeled
with a fluorescent reporter and quencher dyes at opposite ends. The probe is hydrolyzed through
the 5′ꢀexonuclease activity of Taq DNA polymerase, leading to an increasing fluorescence
emission of the reporter dye that can be detected during the reaction. RNA transcripts were
normalized to total RNA levels using RiboGreen, RNA Quantitation Reagent (Molecular
Probes). RiboGreen is an ultrasensitive fluorescent nucleic acid stain which when bound to RNA
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 61
1
2
3
4
5
6
has a maximum excitation/emission at ~500nm/525nm. Data are mean values +/ꢀ standard
deviations.
7
8
9
SRB-1 Mouse Protocol. 6ꢀ8 week old C57BL/6 mice (Charles River Laboratories) were treated
according to the indicated treatment schedules. The sequences primers and probe used for mouse
SRB1 are 5’ꢀTGACAACGACACCGTGTCCTꢀ3’ for the forward primer, 5’ꢀ
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
ATGCGACTTGTCAGGCTGGꢀ3’
for
the
reverse
primer,
and
5’ꢀ
CGTGGAGAACCGCAGCCTCCATTꢀ3’ for the probe.
FXI Mouse Protocol. Male Balbꢀc mice (8 week old, Charles River Laboratories) were treated
according to the indicated treatment schedules. The sequences of primers and probe used for
mouse
FXI
were:
5’ꢀACATGACAGGCGCGATCTCTꢀ3’
(reverse),
(forward),
and
5’ꢀ
5’ꢀ
TCTAGGTTCACGTACACATCTTTGCꢀ3’
TTCCTTCAAGCAATGCCCTCAGCAATXꢀ3’ (probe). Sandwich enzyme immunoassay was
used for measuring mouse FXI plasma protein levels. . Briefly, assay plates were coated with
antiꢀFXI antibody (R&D systems), blocked with 2% BSA and incubated with diluted mouse
PPP. After extensive washes FXI protein was detected by incubation with biotinylated antiꢀFXI
antibody (R&D systems), followed by incubation with streptavidinꢀperoxidase conjugate and
TMB substrate (Sigma). FXI plasma protein levels were calculated using serial dilutions of
recombinant mouse FXI protein (R&D systems) or serial dilutions of normal mouse plasma.
A1AT Mouse Protocol. C57BL/6 mice (Charles River Laboratories) were treated according to
the indicated treatment schedules. The sequences of primers and probe used for mouse A1AT
were
Forward:
5’ꢀTTCTGGCAGGCCTGTGTTGꢀ3’.
Fluorescence
Reverse
primer:
probe:
5’ꢀ
5’ꢀ
ATCCTTCTGGGAGGTGTCTGTCTꢀ3’.
ACS Paragon Plus Environment

Page 27 of 61
Journal of Medicinal Chemistry
1
2
3
4
CCCCAGCTTTCTGGCTGAGGATGTTCꢀ3’. Plasma A1AT levels protein was measured with
5
6
an ELISA method according to manufacturer’s recommendations (Alpco 41ꢀA1AMSꢀE01).
7
8
9
TTR Transgenic Mice Protocol. TTRꢀIle84Ser transgenic mice were generated and described
previously.59 ASOs or vehicle PBS were injected subcutaneously with indicated dose for
indicated periods. Animals were sacrificed 72 h after last dosing. Liver mRNA levels were
analyzed as described above and plasma TTR levels and chemistry values were measured on the
AU480 Clinical Chemistry Analyzer (Beckman Coulter, CA, USA). Human TTR primers:
Forward 5’ꢀCCCTGCTGAGCCCCTACTCꢀ3’. Reverse: 5’ꢀTCCCTCATTCCTTGGGATTGꢀ3’.
Fluorescence probe: 5’ꢀATTCCACCACGGCTGTCGTCAꢀ3’.
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
ApoC-III Mouse Protocol. ApoCꢀIII ASOs 73ꢀ78 subcutaneously administrated to 8ꢀweek old
male C57BL/6 mice (Jackson Laboratories) at various doses. Mice were sacrificed 72 h after
ASO administration and liver collected and flash frozen. RNA was extracted using Invitrogen
PureLink Pro 96 Total RNA purification kit and quantified by qPCR using Invitrogen Express
OneꢀStep qRTꢀPCR kit using apoCꢀIII primers: 5’ꢀTGCAGGGCTACATGGAACAA ꢀ3’ and 5’ꢀ
CGGACTCCTGCACGCTACTT ꢀ3’. Probe with 5’ fluorescein and 3’ TAMRA: 5’ꢀ
CTCCAAGACGGTCCAGGATGCGC ꢀ3’.
ED50 Determination. ED50 values were determined with GraphPad Prism 5 software. The log
dose of ASOs were plotted against mRNA level relative to untreated controls. The curves thus
obtained was fitted using a 4ꢀparameter fit with variable slope and constraining bottom = 0 and
top =1.
Cytokine and Chemokine Assay. Plasma was collected from mice by cardiac puncture. MIPꢀ
1b levels in the plasma were analyzed using a custom, multiplex immunosorbent assay according
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 61
1
2
3
4
5
6
7
8
9
to the manufacturer’s protocol (Meso Scale Discovery, Gaithersburg, MD). Multiplex plates
preꢀcoated with capture antibodies for specific analytes were preꢀsoaked/blocked with
calibrator/sample diluent for 30 minutes prior to use. Diluted standards (4 replicates) and
samples in duplicate were added to the wells and incubated for two hours. All incubations were
performed at room temperature with vigorous shaking (300ꢀ1000 rpm). Plates were washed
three times with PBS + 0.05% Tweenꢀ20 before adding a cocktail of SulfoꢀTag® detection
antibodies to each well. After incubating with detection antibodies for two hours, plates were
washed three times. Read buffer was added to the wells and the plates immediately imaged
using the MSD Sector 2400 imaging system. Data was analyzed using MSD Discovery
Workbench® software. Concentrations of all unknown samples were backꢀcalculated using
results interpolated from the corresponding standard curve regression using a weighted, fourꢀ
parameter fit. Final sample concentrations (pg/ml) were calculated by factoring the dilution
factor used for each sample.
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
N-(5-Oxo-5-(perfluorophenoxy)pentanoyl)amino-tris-{1-[6-amidohexyl)-2-acetamido-3,4,6-
tri-O-acetyl-2-deoxy-β-D-galacosamine)carboxyethoxymethyl)}-methane (3). To a solution
of compound 19 (26.7 g, 15.5 mmol) in anhydrous DMF (150 mL) triethylamine (6.4 mL, 46.2
mmol), PfpꢀTFA (5.3 mL, 30.8 mmol) were added (color changed from yellow to
burgundy). After an h, the reaction was quenched with saturated aqueous sodium bicarbonate
solution (150 mL) and resulting solution was extracted with EtOAc (800 mL). The EtOAc layer
was washed with 1 N aqueous NaHSO4 solution (400 mL), dried (Na2SO4), filtered and
1
concentrated under reduced pressure to yield 3 (29.62 g, quantitative) as light brown foam. H
NMR (300 MHz, DMSO-d6): δ 1.18ꢀ1.42 (m, 12 H), 1.43ꢀ1.66 (m, 12 H,) 1.89ꢀ1.98 (m, 9 H),
2.00 (s, 9 H), 2.05 (s, 9 H), 2.09 (br s, 2 H), 2.15 (s, 9 H), 2.28ꢀ2.39 (m, 4 H), 2.43 (br t, J = 5.3
ACS Paragon Plus Environment

Page 29 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Hz, 8 H), 2.77 (t, J = 7.4 Hz, 2 H), 3.13ꢀ3.30 (m, 6 H), 3.47 (dt, J = 9.6, 6.5 Hz, 3 H), 3.64ꢀ3.78
(m, 10 H), 3.81ꢀ4.05 (m, 9 H), 4.06ꢀ4.23 (m, 6 H), 4.68 (d, J = 8.32 Hz, 3 H), 5.22ꢀ5.32 (m, 3 H),
5.36 (d, J = 3.3 Hz, 3 H), 6.43ꢀ6.59 (m, 3 H), 6.68ꢀ6.77 (m, 2 H);19F NMR (282 MHz, DMSO-
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
13
d6): δ ꢀ167.7ꢀ162.3 (m), ꢀ157.8ꢀ157.4 (m), ꢀ152.9ꢀ152.5 (m); C NMR (75 MHz, DMSO-d6): δ
20.7, 23.3, 25.4, 26.4, 29.1, 29.4, 36.7, 39.3, 51.5, 59.8, 61.5, 66.9, 67.6, 69.6, 70.1, 70.6, 101.1,
170.4, 170.5, 170.8, 171.5; HR MS (ESI) calcd for C84H125F5N7O36 [M + H]+ m/z = 1902.8080,
found 1902.8078.
1-[6-(N-Benzyloxycarbonyl)aminohexyl)-2-acetamido-3,4,6-tri-O-acetyl-2-deoxy-β-D-
galacosamine) (12). GalNAc oxazoline 13 (24.0 g, 72.3 mmol) and compound 14 (16.4, 65.1
mmol) and preꢀdried molecular sieves (25 g, 4 Å) were suspended in anhydrous dichloromethane
(150 mL). The mixture was stirred at room temperature for 30 min and TMSOTf (6.5 mL, 36.1
mmol) was introduced. After stirring at room temperature for 12 h the reaction mixture was
poured into ice cold sodium bicarbonate solution (300 mL) and resulting solution was extracted
with dichloromethane (3 x 300 mL). The organic layer was washed with brine (300 mL), dried
over Na2SO4, filtered and concentrated under reduced pressure. The residue obtained was
purified by silica gel chromatography and eluted with 2ꢀ5% methanol in dichloromethane to
yield compound 12 (27.4 g, 65.4%) as white foam. 1H NMR (300 MHz, DMSOꢀd6): δ 1.35ꢀ1.43
(m, 4 H), 1.43ꢀ 1.65 (m, 4 H), 1.94 (s, 3 H), 2.00 (s, 3 H), 2.05 (s, 3 H), 2.13 (s, 3H), 3.03ꢀ3.29
(m, 2 H), 3.48 (dt, J = 9.7, 6.5 Hz, 1 H), 3.78ꢀ4.04 (m, 3 H), 4.05ꢀ4.22 (m, 2 H), 4.65 (d, J = 8.3
Hz, 1 H), 4.92 (br s, 1H), 5.03ꢀ5.18 (m, 2 H), 5.27 (dd, J=11.26, 3.20 Hz, 1 H), 5.34 (d, J=2.94
Hz, 1 H), 5.96 (br d, J=8.58 Hz, 1 H), 7.28ꢀ7.41 (m, 5 H); 13C NMR (75 MHz, DMSOꢀd6): δ
20.7, 23.4, 25.2, 26.0, 29.0, 29.8, 40.6, 51.6, 61.5, 66.6, 66.9, 69.4, 70.0, 70.6, 100.8, 127.9,
ACS Paragon Plus Environment