An improved process for the manufacture of 5′-O-(4,4′-dimethoxytrityl)-N2-isobutyryl-2′-O-(2-methoxyethyl)guanosine

Article abstract of DOI:10.1021/acs.oprd.0c00261

A revised, optimized process for the manufacture of 5′-O-(4,4′-dimethoxytrityl)-N2-isobutyryl-2′-O-(2- methoxyethyl)guanosine (MOE G PNS) that controls critical impurities to less than 0.2% was developed. The 2′-O-alkylation of 2,6-diaminopurine riboside

Full text of DOI:10.1021/acs.oprd.0c00261

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE
Full Paper
An Improved Process for the Manufacture of 5#-O-(4,4#-
2
Dimethoxytrityl)-N-isobutyryl-2#-O-(2-methoxyethyl)guanosine
Andrew McPherson, Daniel Capaldi, Lijian Chen, and Philip Olsen
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.0c00261 • Publication Date (Web): 23 Sep 2020
Downloaded from pubs.acs.org on September 23, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
An Improved Process for the Manufacture of 5′-O-
(4,4′-Dimethoxytrityl)-N2-isobutyryl-2′-O-(2-
methoxyethyl)guanosine
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
Andrew K. McPherson*, Daniel Capaldi, Lijian Chen, Philip Olsen
Ionis Pharmaceuticals, Inc., 2855 Gazelle Court, Carlsbad, CA 92010
Keywords: MOE G PNS, oligonucleotide precursor, DAPR alkylation, guanosine
protection, design of experiments (DOE)
Abstract
A revised, optimized process for the manufacture of 5′-O-(4,4′-dimethoxytrityl)-N2-
isobutyryl-2′-O-(2-methoxyethyl)guanosine (MOE G PNS) that controls critical impurities
to less than 0.2% was developed. The 2′-O-alkylation of 2,6-diaminopurine riboside
(DAPR) with 1-bromo-2-methoxyethane (MOE-Br) in DMSO was examined using a
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 38
1
2
3
4
5
6
7
8
9
Design of Experiments (DOE) approach, which lead to the selection of LiOH as the
optimal base choice. Equivalents of base and MOE-Br were optimized to control levels of
the critical 3′-O-(2-methoxyethyl) impurity and residual DAPR. DMSO was removed with
a solid-phase extraction using SP-207 resin, affording the desired 2′-O-(2-methoxyethyl)-
2,6-diaminopurine riboside (MOE DAPR) in an aqueous solution in 53% yield. The
methylbromide impurity commonly found in bulk sourced MOE-Br was controlled using a
co-distillation with MEK, thereby controlling the resulting critical 2′-O-methyl nucleoside
impurity. The aqueous solution was telescoped into the enzymatic conversion of MOE-
DAPR to 2′-O-(2-methoxyethyl)guanosine (MOE G) using adenosine deaminase. MOE G
purity is enhanced at this step due to enzymatic selectivity and crystallization from the
reaction mixture, which together significantly reduce levels of the 3′ MOE G isomer.
Installation of the isobutyryl and 4,4-dimethoxytrityl protecting groups proceeded similarly
to published literature methods, but the work-up and isolation conditions were optimized.
A series of extractions after isobutyrlation was used to control critical and non-critical
impurities, and the final product was crystallized to upgrade purity. The improved process
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
2

Page 3 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
was carried out at production-scale to afford 59 kg of MOE G PNS with no critical impurity
over 0.2% in an overall yield of 26%.
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
ABBREVIATIONS
MOE, 2-methoxyethyl; DAPR, 2,6-diaminopurine riboside; PNS, protected nucleoside;
DCI, 4,5-dicyanoimidazole.
Introduction
Oligonucleotides are a distinct and growing therapeutic platform with the ability to target
genes linked to diseases that are considered undruggable by classic small molecule
approaches.1,2,3 Several nucleic-acid based drugs have been approved, and many more
are in the late stages of clinical development. Nucleic-acid therapeutics are manufactured
by solid-phase synthesis, where phosphoramidite monomers are coupled sequentially on
a solid support, which is typically functionalized with a universal linker such as
UnyLinker™ to allow easy removal of the desired sequence from the solid support.4
Synthesis is typically carried out using a four step cycle comprising acid treatment to
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 38
1
2
3
4
5
6
7
8
9
remove a 5′-O-4,4′-dimethoxytrityl protecting group, attachment of a phosphoramidite to
the support-bound oligonucleotide, conversion of the newly formed phosphite triester to
either its phosphorothioate triester with a sulfurizing reagent or a phosphate triester with
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
an oxidizing reagent, and finally acetylation of unreacted 5′-hydroxyls.5 In this manner the
6,7
desired sequence is constructed from the 3′ to 5′ terminus.
Several commercially
approved drugs and many late-stage clinical candidates contain 2′-O-(2-methoxyethyl)
modified (MOE) nucleosides. Historically, the most expensive and challenging MOE
phosphoramidite to manufacture has been 2′-O-(2-methoxyethyl) guanosine
phosphoramidite (MOE G phosphoramidite), 2, which is produced by phosphitylating the
corresponding 2′-O-(2-methoxyethyl) guanosine protected nucleoside (MOE G PNS) 1
using 2-cyanoethyl-N,N,N′,N′-tetraisopropylphosphorodiamidite (Phos reagent8) in the
presence of weakly acidic activators such as tetrazole and DCI (Scheme 1). The
synthesis of high purity phosphoramidites relies on the production of high purity protected
nucleosides.9
ACS Paragon Plus Environment
4

Page 5 of 38
Organic Process Research & Development
1
2
3
4
5
Scheme 1. Phosphitylation of 2′-O-(2-Methoxyethyl) Guanosine Protected Nucleoside
6
7
8
9
O
N
CN
NH
O
O
N
N
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
N
O
O
P
H
N
N
NH
O
DMTrO
(iPr)2N
N(iPr)2
N
O
O
O
O
P
H
CN
activator
DMTrO
N
O
O
O
HO
1
2
MOE G PNS
MOE G phosphoramidite
Previously, we reported a process for the manufacture of MOE G PNS, 1, where
alkylation of the 2′ oxygen of 2,6-diaminopurine ribose was carried out using 1-bromo-2-
methoxyethane and KOH in DMSO.10 While the process was successfully used in a pilot
plant setting, further scale-up resulted in uncontrolled and variable levels (0.11% - 1.22%)
of an impurity containing a 2′-O-methyl in place of the desired 2′-O-(2-methoxyethyl),
which we thought formed by reaction of methyl bromide originating as both a trace
impurity in 1-bromo-2-methoxyethane (MOE-Br), and from degradation of MOE-Br during
the alkylation reaction. After failed attempts to bring the 2′-O-methyl impurity under control
at production scale, the need for material to support clinical studies drove us to set aside
the MOE-Br/DMSO conditions in favor of a lower yielding aqueous alkylation process,
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 38
1
2
3
4
5
6
7
8
9
which used methoxyethyl mesylate and did not result in detectable methylation. Over the
past several years we have carried out process development on the MOE-Br/DMSO
process to enable control of the 2′-O-methyl impurity without chromatography. Herein we
report an improved process for MOE G PNS using MOE-Br in DMSO, which has been
successfully demonstrated at ca 60 kg batch scale. The overall route is the same as
previously published (Scheme 2), however many changes have been made to the
reaction conditions and workups.
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. Preparation of 2′-O-(2-Methoxyethyl) Guanosine Protected Nucleoside (MOE
G PNS)
Br
O
N
NH2
N
NH2
N
O
N
N
N
N
N
N
4
NH
NH2
ADA enzyme
N
NH2
N
NH2
O
O
O
LiOH, DMSO
sodium phosphate buffer(aq)
HO
HO
HO
O
O
OH
O
O
HO
HO
HO
3
5
6
DAPR
MOE DAPR
MOE G
O
O
N
N
NH
O
NH
O
N
DMTr-Cl, 2,6-lutidine,
DCM
N
N
N
TMSCl, pyridine;
iBuCOCl
N
N
O
O
H
H
DMTrO
HO
O
O
O
HO
O
HO
7
1
MOE G-ibu
MOE G PNS
ACS Paragon Plus Environment
6

Page 7 of 38
Organic Process Research & Development
1
2
3
4
5
Results and Discussion
6
7
8
9
Attempts have been made to find a cost-effective route to MOE G PNS starting from
guanosine; however, direct alkylation of guanosine leads to preferential O6-alkylation.11,12
Conversion of guanosine to 2,6-diaminopurine riboside (DAPR) prior to alkylation protects
the nucleobase against alkylation. Once converted to DAPR, another difficulty is
encountered while attempting to alkylate the 2′-hydroxyl selectively. That is, while the 2′-
OH vs. 3′-OH selectivity favors the desired product, a significant amount of the 3′-MOE
regioisomer impurity 8 is also formed. Note that the 3′-MOE regioisomer is a critical
impurity because, if carried through to phosphoramidite, the 3′-MOE phosphoramidite
impurity will compete with the parent phosphoramidite and result in a 3′-MOE
oligonucleotide impurity that cannot be purified away from the desired parent
oligonucleotide. Additionally, the reaction is complicated by over alkylation products. The
2′-hydroxyl proton of DAPR is more acidic than the 3′-hydroxyl (predicted pKa value for
2′-OH is 12.9 and 3′-OH is 14.2), but once the 2′-hydroxyl is alkylated, the pKa of the 3′-
hydroxyl becomes closer to that of the 2′-hydroxyl of DAPR (predicted to be 13.4).
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
7

Organic Process Research & Development
Page 8 of 38
1
2
3
4
5
6
7
8
9
Therefore, the 3′-hydroxyl of the desired 2′-MOE DAPR 5 competes in the alkylation
reaction, resulting in the bisalkylated impurity 10 (Scheme 3). Note that impurity 10 is non-
critical because if carried through to phosphoramidite, it would not have a free alcohol to
be phosphitylated and therefore would not incorporate into the oligonucleotide. To
complicate matters further, the base needed to carry out the reaction slowly hydrolyzes
MOE-Br (4) to 2-methoxyethanol (9). Thus, insufficient MOE-Br leaves residual DAPR
starting material 3, which, if carried through to phosphoramidite results in an unreactive,
non-critical, cyclic phosphite triester impurity due to the cis-diol functionality reacting twice
with the Phos reagent during phosphitylation.
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
While the alkylation reaction is complex, preparing MOE G PNS (1) in this manner is
the most efficient and cost-effective route we have found. Attempts at avoiding the
selectivity complications using 3′,5′ protection strategies such as TIPDS or di-t-butylsilyl
led to increased cost due to the two additional steps required and the cost of the silyl
reagents with no discernable benefit in throughput or critical impurity control.12–14
ACS Paragon Plus Environment
8

Page 9 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
We developed the MOE-Br/DMSO process and the control strategy with the goal of
producing MOE G PNS with tightly controlled levels of critical impurities. Rather than hold
each reaction step to a criterion of high purity, we developed the process to reduce or
purge critical impurities throughout to ensure material of high quality at the MOE G PNS
stage.
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 3. Alkylation Products with Critical Modifications Relative to Parent in Red
NH2
N
N
NH2
N
N
N
N
Br
NH2
N
O
N
NH2
O
+
N
N
N
NH2
4
O
HO
O
OH
HO
O
N
NH2
O
LiOH, DMSO
O
O
HO
HO
8
5
OH
HO
3
DAPR
NH2
OH
O
N
N
N
9
N
NH2
O
HO
O
O
O
O
10
To determine which reaction parameters affected the yield and levels of critical impurity
8, the alkylation reaction conditions to produce intermediate 5 were reexamined using a
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 38
1
2
3
4
5
6
7
8
9
two level, five-factor, 16 run 25-1 screening DOE. Equivalents of base and alkylbromide 4
were assumed to be important and were fixed in the DOE, while the volume of DMSO,
temperature, base type, base addition rate, and alkylbromide addition rate were varied.
(Table 1) Reaction progress was followed over time and the reaction profile assessed for
maximum product formation, time to maximum product, time product remained within
2.5% of the maximum value (reaction plateau), and impurity levels at the time of maximum
product formation (Figure 1). Analysis of the results indicated that the maximum product
was formed when the reaction was carried out under more dilute conditions (15 vol
DMSO) using LiOH, with reaction volume being slightly more impactful than the type of
base. No other main effects or two factor interactions were observed. The time the
reaction took to reach maximum product was governed by a complicated set of factors
comprising type of base, temperature, base addition rate, and an aliased two-factor
interaction, which is either reaction volume × type of base or temperature × bromide
addition rate. Of these, base type was the predominant factor, with LiOH being favored.
Analysis of the length of the reaction plateau showed that LiOH produced a longer, flatter
reaction maximum, with a lower reaction temperature of 15 °C also contributing to a small
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
10

Page 11 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
extent. The amount of residual DAPR starting material 3 at the point when the reaction
reached its maximum level of product was unaffected by any of the parameters varied in
the DOE. However, the amount of non-critical bis-alkylated impurity 10 was inversely
affected by the reaction volume, and the critical 3′-MOE impurity 8 was affected by both
the reaction volume and the type of base. When the reaction was concentrated, the level
of 3′-MOE impurity 8 was lower, while the bis-alkylated impurity 10 was higher.
Presumably, the more concentrated reaction conditions favor the continued conversion
of 3′-MOE into the bis-alkylated species. The use of LiOH also favored lower levels of 3′-
MOE impurity 8 over those formed with KOH. While higher volumes of solvent made a
slight positive impact on maximum yield, the lower volume of solvent was ultimately
selected to improve control of critical 3′-MOE impurity 8 and throughput. An optimization
model built upon the DOE data for 5 volumes of DMSO suggested that the optimal
balance of maximum yield, lowest 3′-MOE impurity 8, and long plateau would be obtained
using LiOH at 24 °C with the base and alkylbromide charged as single additions.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. DOE Variables
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 38
1
2
3
4
5
6
7
8
9
Variable
Type
Low
5
High
15
Vol DMSO
Numeric
Numeric
Categoric
Temperature (°C)
base type
15
25
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
LiOH
KOH
base addition rate
Categoric All at once Portions
Categoric All at once Portions
alkylbromide addition
rate
Equiv. base
Constant
Constant
2.6
2.7
Equiv. 4
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0
10
20
Hours
2'-MOE DAPR
DAPR
3'-MOE DAPR
Bis-alkylated
ACS Paragon Plus Environment
12

Page 13 of 38
Organic Process Research & Development
1
2
3
4
5
Figure 1. Example reaction profile from the DoE using 5 vol DMSO, LiOH, portion-wise
6
7
8
addition of base, portion-wise addition of alkylbromide, 25 °C
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
70%
60%
50%
40%
30%
20%
10%
0%
0
5
10 15 20 25 30 35 40 45 50
Hours
2'-MOE DAPR
DAPR
3'-MOE DAPR
Bis-alkylated
Figure 2. Reaction profile using 5 vol DMSO, 1.3 equiv. LiOH added all at once, 1.4 equiv.
alkylbromide 4 added all at once, 25 °C
Armed with the information from the DOE, we turned our attention to the base and
alkylbromide equivalents. Reducing the equivalents of LiOH to 1.3 and alkylbromide 4 to
1.4 did not change the reaction profile (10 hour time point Figure 1 vs. 30 hour time point
Figure 2); however, the reaction progression slowed considerably. Under these conditions
the reaction has reached just above 60% product formation (not yet the maximum) after
48 hours with approximately 6% residual DAPR starting material 3, 10% 3′-MOE impurity
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 38
1
2
3
4
5
6
7
8
9
8, and 19% bis-alkylated 10. While reducing the equivalents of reactants is generally
desirable, these conditions are not ideal for several reasons. To ensure rugged control in
the downstream steps, we wanted critical 3′-MOE impurity 8 levels to be as low as
possible at the crude stage of the alkylation. Further, non-critical DAPR starting material
3 continues to react throughout the process and the resulting species are not easily
removed; therefore, it would be beneficial to control 3 to low levels in the alkylation crude.
Both impurities can be reduced by running the reaction past the product maximum and
trading some yield for impurity control. Running the reaction to the point at which not more
than 2.0% 3 remains typically produces levels of 3′-MOE impurity 8 and bis-alkylated
impurity 10 around 5% and 33% respectively. To reach this point using reduced
equivalents of base and alkylbromide would take more than 48 hours, which was deemed
unacceptable in terms of throughput. After screening several pairings of base and
alkylbromide equivalents, 2.15 equivalents of both LiOH and alkylbromide 4 were chosen
as a reasonable balance between reagent usage and throughput, with the reaction
reaching 2.0% 3 in approximately 18 hours while maintaining a long product plateau.
Despite the yield loss (ca. 5%) associated with running the reaction past the point of
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
14

Page 15 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
maximum product, the new conditions provide a doubling of the yield over the former
aqueous alkylation process (~54% vs. 28%).
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
As mentioned previously, the quality of 1-bromo-2-methoxyethane (4) is critical. Bulk
supplies of 4 can contain small amounts of alkylbromide impurities that can compete
during the alkylation reaction (Figure 3). Most of these impurities can be controlled by an
appropriate specification on 1-bromo-2-methoxyethane; however, bromomethane is a
decomposition by-product of the reagent, which complicates control of the resulting 2′-O-
Me nucleoside impurity. A simple co-distillation of alkylbromide 4 and MTBE in DMSO
was sufficient to reduce 2′-O-Me impurity formation to very low levels (<0.05%). The co-
distillation procedure was implemented at production scale successfully. However, when
the same process was used in a subsequent campaign with a different nucleoside
(adenosine), unexpectedly high levels of the corresponding 2′-O-Me impurity were
generated. Because the co-distillation was expected to behave similarly regardless of
nucleoside, the result brought into question control of the 2′-O-Me impurity by co-
distillation. We therefore reexamined the co-distillation conditions and determined that
ACS Paragon Plus Environment
15

Organic Process Research & Development
Page 16 of 38
1
2
3
4
5
6
7
8
9
carrying out the alkylbromide/MTBE distillation in the absence of DMSO resulted in
reproducible control of the 2′-O-Me impurity. The new conditions were demonstrated
during a subsequent campaign to alkylate adenosine (data to be disclosed in a future
publication). While the new co-distillation conditions have not formally been demonstrated
for MOE G PNS, control of the 2′-O-Me impurity in the adenosine series indicates the
procedure will be similarly successful in the synthesis of 1.
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
Br
Br
Br
bromomethane
bromoethane
1-bromobutane
Br
O
Br
O
O
1-bromo-2-ethoxyethane
1-bromo-2-
(2-methoxyethoxy)ethane
Figure 3. Critical impurities that may be present in bulk 1-bromo-2-methoxyethane
The next step in the process is enzymatic deamination of MOE DAPR 5 to MOE G 6
using adenosine deaminase. The enzyme can tolerate up to ca. 20% DMSO; however,
even a small amount of DMSO impairs crystallization of 6 from the reaction mixture, which
is a key component of the impurity control strategy. Therefore, a method to remove the
DMSO prior to the enzymatic reaction was required. Attempts at biphasic aqueous
ACS Paragon Plus Environment
16

Page 17 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
extractive removal of DMSO using brine and THF or n-butanol showed an unfavorable
partitioning of MOE DAPR 5 into the aqueous phase. Organic solvent nanofiltration and
counter current extractions both showed promise, but required substantial capital
investment in specialized equipment, and so an alternative solid-phase extraction was
developed. The solid phase chosen was brominated polystyrene matrix SP-207, which
was placed in a column with a height to diameter ratio of 2:1. The 2:1 ratio provides
enough depth of matrix to hold the product and remove DMSO, even without uniform
loading. After capturing the crude mixture on the solid-phase, the resin is washed with
10% NaOAc(aq), which is critical in disrupting the DMSO binding. Once the DMSO is
removed, the resin is rinsed with water to remove sodium acetate (a necessity for the
following enzymatic reaction), and the product released from the resin using 80%-100%
methanol(aq) with a recovery of about 94%. Methanol is removed under reduced pressure
to afford the crude alkylation mixture in water, which is used directly in the next step.
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 enzymatic deamination of DAPR intermediate is carried out using adenosine
deaminase (ADA) in the presence of aqueous phosphate buffer. Reaction conditions
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 38
1
2
3
4
5
6
7
8
9
were optimized to balance conversion speed with amount of enzyme. We settled on an
enzyme charge of 176 units of ADA per gram 2′-MOE DAPR, which afforded MOE G 6
in 9 hours in 84-90% yield. During the reaction, not only is 2′-MOE DAPR 5 deaminated
to form the desired 2′-MOE G 6, but other components in the crude alkylation mixture
are also deaminated (Scheme 4). Of these, only 3′-MOE G 12 is critical. The other
critical impurity observed in the crude reaction mixture is residual 2′-MOE DAPR 5.
Interestingly there are two characteristics of the reaction that we took advantage of in
the control strategy of critical impurities 5 and 12. First, the ADA enzyme selectively
deaminates 2′-MOE DAPR 5 over 3′-MOE DAPR 8. This is evinced by the enrichment
of the 2′-MOE G isomer over the course of the reaction (ratio of 2′ to 3′ MOE DAPR at
~7:1 in the alkylation reaction mixture vs. the ratio of 2′ to 3′ MOE G at ~18:1 in the
deamination crude). Secondly, 2′-MOE G 6 crystallizes selectively from the sodium
phosphate buffer. Provided the reaction is run until not more than 1.0% of 5 remains,
the crystallization allows for a simple workup by filtration, as the product precipitates
from the reaction mixture to afford material with a ~180:1 ratio of 6 to 12 and residual 5
reduced to ~0.15% in ~81% (purity corrected) yield.
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
18

Page 19 of 38
Organic Process Research & Development
1
2
3
4
5
Scheme 4. Mixture of Compounds in the Deamination Reaction with Critical Impurities
6
7
8
Demarcated in Red
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
NH2
N
NH
NH2
N
N
ADA enzyme
N
N
N
sodium phosphate buffer(aq)
O
N
NH2
O
HO
HO
OR2
OR2
R1O
R1O
R1 = H, R2 = CH2CH2OCH3
R1 = H, R2 = H
R1 = CH2CH2OCH3, R2 = H
R1 = H, R2 = CH2CH2OCH3
R1 = H, R2 = H
R1 = CH2CH2OCH3, R2 = H
R1 = CH2CH2OCH3, R2 = CH2CH2OCH3
5
6
3
11
12
13
8
R1 = CH2CH2OCH3, R2 = CH2CH2OCH3
10
The next step in the process is to install the isobutyryl protecting group. The reaction is
carried out with transient protection of the 3′ hydroxyl as the trimethylsilyl ether, followed
by N-acylation using isobutyryl chloride.15 The reaction is quenched with methanol and
adjusted to neutral pH with ammonium hydroxide.
Under the previous aqueous alkylation conditions, the alkylation reaction was stopped
near its maximum product formation to maximize throughput in a poor yielding process.
To obtain the desired product 2′-MOE Gibu 7 in an organic phase while keeping more
polar impurities such as triol 14 in the aqueous phase, an extensive continuous
extraction followed by a complicated series of back extractions were required. However,
ACS Paragon Plus Environment
19

Organic Process Research & Development
Page 20 of 38
1
2
3
4
5
6
7
8
9
because the alkylation in DMSO can be pushed further along the reaction pathway with
minor yield loss (vide supra), there is much less impurity 14 (carried forward from
unreacted 1). This allows for the continuous extraction to be replaced with a simpler
batch extraction workup. The acylation reaction is quenched with methanol and
precipitated pyridine hydrochloride removed via filtration. The filtrate is solvent swapped
to water, and the aqueous solution extracted with 3% methanol in DCM to extract the
product and leave most of impurity 14 in the aqueous layer. Attempts to replace DCM
with other solvents failed to identify a suitable alternative. The combined organic extract
is washed with brine to reduce impurity 14 to undetectable levels. The organic layer is
azeotropically dried, concentrated to 4 vol, and swapped to acetone to crystallize the
desired product 7 in ~82% yield, further reducing the level of critical impurity 15 and
removing trace amounts of the critical isobutyrylated products of 2′-MOE DAPR 5
carried through from the previous step (17, 18, 19).
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
20

Page 21 of 38
Organic Process Research & Development
1
2
3
4
5
Scheme 5. Mixture of Compounds in the Isobutyryl Protection Reaction
6
7
8
9
NHR4
O
N
O
N
N
N
N
N
N
N
N
N
NH
NH2
NH
O
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
NHR5
N
O
O
O
H
HO
HO
R3O
OR2
OR2
O
R1O
R1O
O
HO
1. TMSCl, pyridine;
i-BuCOCl;
R1 = H, R2 = CH2CH2OCH3
R1 = H, R2 = H
R1 = CH2CH2OCH3, R2 = H
R1, R3 = H, R2 = CH2CH2OCH3
R1, R3 = H, R2 = H
R1 = CH2CH2OCH3, R2, R3 = H
R1 = H, R2 = CH2CH2OCH3, R3 = i-Bu
R4 = H, R5 = COCH(CH3)2
R4 = COCH(CH3)2, R5 = H
R4, R5 = COCH(CH3)2
6
7
17
18
19
11
12
14
15
16
2. NH4OH(aq)
NH2
N
N
N
N
NH2
O
HO
O
O
HO
5
The final step in the process is to install the dimethoxytrityl protecting group. The
reaction and workup were modified slightly from the previously mentioned published
procedure. At this point in the synthesis, it becomes difficult to upgrade purity, so
emphasis is placed on clean conversion to 1. Based on previous work, 2,6-lutidine was
preferred over pyridine, since it increased the regioselectivity of the tritylation (5′ over 3′
hydroxyl). Running the trityl protection with 2,6-lutidine proceeded smoothly over 30
minutes to give 1 in excellent yield and purity, and so no further optimization of the
reaction conditions was performed. However, the workup was extensively modified. It
ACS Paragon Plus Environment
21

Organic Process Research & Development
Page 22 of 38
1
2
3
4
5
6
7
8
9
was previously reported that the reaction was quenched with methanol, vacuum distilled
to a residue, dissolved and treated with a series of extractions, purified by silica gel
chromatography followed by another series of extractions to remove residual lutidine,
before again evaporating to a residue, and triturating with hexane to induce formation of
an amorphous solid, which could be isolated by filtration. While this process was viable
at low kg scales, we wanted a rugged process capable of being more easily scaled.
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 methanol quench was maintained, but no concentration to a residue was
employed. Rather, sodium bicarbonate washes were carried out on the DCM/MeOH
solution directly (reducing residual levels of 7 to less than 0.1%). The column
chromatography step was removed and replaced by precipitation and recrystallization
from DCM/cyclohexane to afford crystalline MOE G PNS 1 in ~73% (w/w) yield. To our
knowledge, this is the first report of crystalline MOE G PNS. Attempts to replace DCM in
the crystallization with other solvents failed to provide a suitable alternative, so to
simplify the process, DCM was chosen as the reaction solvent.
Scheme 6. Mixture of Compounds in the DMTr Protection Reaction
ACS Paragon Plus Environment
22

Page 23 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
O
O
N
N
NH
O
N
N
NH
O
DMTr-Cl, 2,6-lutidine,
DCM
N
N
N
O
N
H
O
H
DMTrO
HO
OR2
R1O
OR2
R1O
73% (w/w) yield
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
R1 = H, R2 = CH2CH2OCH3
R1 = CH2CH2OCH3, R2 = H
R1 = H, R2 = CH2CH2OCH3
R1 = CH2CH2OCH3, R2 = H
7
1
15
17
The improved process was demonstrated at production scale starting with 91.2 kg 2,6-
DAPR to afford 58.8 kg MOE G PNS (assay corrected) with all critical impurities below
0.15%. This represents an overall yield of ~26%, which is a 153% increase over
material produced by the former unpublished aqueous alkylation process.
Experimental Section
Materials and Instrumentation. Except for 1-bromo-2-methoxyethane, commercially
available starting materials were used as received without further purification. 1H and
13C NMR spectra were recorded on a Bruker 300 MHz spectrometer at ambient
temperature (298 K) using tetramethylsilane as an internal standard. The multiplicities of
1H NMR signals are indicated as follows: s = singlet, d = doublet, t = triplet, m =
ACS Paragon Plus Environment
23

Organic Process Research & Development
Page 24 of 38
1
2
3
4
5
6
7
8
multiplet, br = broad, or combinations of thereof. Coupling constants (J) are reported to
the nearest 0.1 Hz.
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
5′-O-(4,4′-dimethoxytrityl)-N2-isobutyryl-2′-O-(2-methoxyethyl)guanosine (MOE G
PNS) (1). Into a 2000 L reactor at 20-35 °C under nitrogen was charged DCM (714 L),
47.6 kg 7 (a/a% purity corrected: 47.1 kg, 115 mol), 2,6-lutidine (49.1 kg, 458 mol), and
DMTr-Cl (42.8 kg, 126 mol). The reaction mixture was stirred at 20-35 °C for 30 minutes
at which point unreacted 7 was undetectable by HPLC. The reaction was cooled to 10-
15 °C, quenched with methanol (3.1 L, 97 mol) at 10-15 °C, and the organic layer
washed three times with 10% sodium bicarbonate(aq) (NaHCO3: 9.52 kg, water: 95 L)
at 20-35 °C and once with sat. NaCl(aq) (NaCl: 28.6 kg, water: 95 L). The organic layer
was dried with sodium sulfate (47.6 kg, 335 mol), filtered, and the filtered sodium
sulphate rinsed with DCM (95L). To the combined filtrate and rinse was charged
cyclohexane (1428 L) and the material was precipitated by removing the DCM by
vacuum distillation (<45 °C). The mixture was cooled to 20-35 °C and filtered. The
resulting crude solid was charged into a reactor containing DCM (476 L) and stirred for
ACS Paragon Plus Environment
24

Page 25 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
15 minutes to dissolve. To the solution was added cyclohexane (762 L) and the mixture
was stirred at 20-35 °C until solid formation was observed (ca. 75 minutes). Upon
formation of solid, the slurry was maintained at 20-35 °C for 6 hr, then filtered, and the
filtered solid rinsed with 38% DCM in cyclohexane (DCM: 90 L, cyclohexane: 143 L).
The solid was dried under vacuum at 45-50 °C until the water content was less than
1.00% (ca. 16 hr.) to afford MOE G PNS, 1 (62.2 kg, 87 mol), with an assay of 95.5%
and a purity of 98.96% a/a% in a 72.7% yield (assay corrected). Impurity profile (a/a%):
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
1
2′-O-methyl G PNS (0.06%), 3′-MOE G PNS 17 (0.05%) H NMR (300 MHz, CDCl3) δ
11.99 (s, 1H), 7.90 (s, 1H), 7.84 (s, 1H), 7.58 – 7.48 (m, 2H), 7.45 – 7.34 (m, 4H), 7.33 –
7.16 (m, 4H), 6.87 – 6.73 (m, 4H), 5.88 (d, J = 7.0 Hz, 1H), 4.91 (dd, J = 7.1, 4.9 Hz,
1H), 4.49 (dt, J = 4.8, 2.4 Hz, 1H), 4.26 – 4.19 (m, 1H), 3.88 (ddd, J = 11.3, 4.2, 2.4 Hz,
1H), 3.83 (d, J = 2.6 Hz, 1H), 3.78 (s, 3H), 3.76 (s, 3H), 3.69 (ddd, J = 11.2, 8.0, 2.5 Hz,
1H), 3.63 – 3.49 (m, 2H), 3.46 (ddd, J = 10.6, 4.2, 2.5 Hz, 1H), 3.35 (s, 3H), 3.13 (dd, J
= 10.7, 3.3 Hz, 1H), 1.67 (hept, J = 6.8 Hz, 1H), 0.92 (d, J = 6.9 Hz, 3H), 0.73 (d, J = 6.9
Hz, 3H). 13C NMR (75 MHz, CDCl3) δ 178.38, 158.76, 155.51, 148.26, 147.17, 145.03,
ACS Paragon Plus Environment
25

Organic Process Research & Development
Page 26 of 38
1
2
3
4
5
6
7
8
139.02, 136.10, 135.71, 130.02, 128.06, 127.15, 122.43, 113.28, 86.43, 86.27, 84.51,
81.47, 77.24, 71.78, 70.03, 69.44, 63.78, 59.03, 55.28, 36.07, 18.57, 18.49.
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
Co-distillation of 1-bromo-2-methoxyethane (4) to remove MeBr
Methyl tert-butyl ether (32.4 Kg) and 2-methoxyethyl bromide (24 Kg) were charged to a
reactor at 20-30 °C. The mixture was stirred for 15 minutes and then MTBE and any
MeBr impurity present was distilled under vacuum (300-350 mm Hg) at 20-30 °C to 0.9-
1.2 vol remaining in reactor. Note that temperatures above 30 °C at 500 mm Hg will
evaporate 2-methoxyethyl bromide and should be avoided.
2′-O-(2-methoxyethyl)-2,6-diaminopurine riboside (5). DMSO (295 L), 2,6-diaminopurine
riboside 3 (91.2 kg, 323 mol), and lithium hydroxide monohydrate (29.1 kg, 694 mol)
were charged to a 2000 L reactor at 20-35 °C under nitrogen and stirred for 30 min. The
mixture was cooled to 15-20 °C and maintained at that temperature while 330 kg of a
solution of 2-methoxyethyl bromide 4 (29.47 w/w%) in DMSO was charged over 75
minutes. The temperature was raised to 20-35 °C and maintained at that temperature
until residual 3 < 2.0% a/a% by HPLC (1.58% in approximately 19 hr). The reaction was
ACS Paragon Plus Environment
26

Page 27 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
quenched and diluted with chilled water (1368 L, 15-20 °C) over 190 min. The pH of the
mixture was adjusted to 7.1 with 5N HCl(aq) (30 L). Sodium acetate (20.5 kg) was
added to the crude mixture and the pH was again adjusted to 7.1 with 5N HCl(aq)
(0.912 L). Crude reaction profile (a/a%): alkylated product 5 (56.6%), residual 3 (1.6%),
3′-MOE impurity 8 (5.3%), and bis-alkylated impurity 10 (33.2%).
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
Solid-phase extraction of DMSO from crude 2′-O-(2-methoxyethyl)-2,6-diaminopurine
riboside (5). SP-207 resin from Mitsubishi Chemical Corporation manufactured by
Resindion Srl (14 vol relative to 3, 1272 L) was charged to a 1500 L column (2:1 m
depth:diameter). At this scale, the flow rate of all equilibration and sodium acetate
containing solutions through the resin should be not more than 20 L/min (linear velocity
~25 m/min). The resin was conditioned with 5% methanol(aq) (70 L methanol, 1330 L
water, 1.1 CV), water (1780 L, 1.4 CV), and 10% sodium acetate(aq) (254.4 kg NaOAc,
2544 L water, 2 CV). Half of the crude solution of 5 (929 kg) was charged to the column.
The loaded column was washed with 10% sodium acetate(aq) (254.4 kg NaOAc, 2544
L water, 2 CV). Note that most of the DMSO was contained in the first CV (~16% assay)
ACS Paragon Plus Environment
27

Organic Process Research & Development
Page 28 of 38
1
2
3
4
5
6
7
8
9
with the remainder in the second CV (~2.5% assay). The residual sodium acetate was
washed from the resin with water (2544 L, 2 CV). Once the sodium acetate is removed,
the flow rate of the following washes can be increased to a linear velocity of ~50 m/min
or 40 L/min at this scale. The alkylated product 5 was liberated from the resin with 80%
MeOH(aq) (2036 L MeOH, 508 L water, 2 CV). The resin is rinsed with MeOH (2544 L,
2 CV) to remove any residual material. The equilibration, capture, and release process
was repeated for the second half of the crude reaction mixture. Note that after the final
MeOH wash, the resin can be stored in MeOH and reused (demonstrated up to 25
times in the lab with a similar resin). The washes containing 80% MeOH and 5 from the
two capture release operations were combined and the volume was reduced to ~20% of
the original volume via vacuum distillation keeping the solution below 50 °C. The
resulting aqueous solution of 2′-MOE-DAPR 5 (438.3 kg, 13.4% w/w, 58.8 kg 5, 173
mol, 53.4% assay yield) was used directly in the next step.
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′-O-(2-methoxyethyl)guanosine (6). The ca 13%w/w aqueous solution of 2′-MOE-
DAPR 5 (58.8 kg assayed, 173 mol) was charged to a 1000 L reaction vessel and
ACS Paragon Plus Environment
28

Page 29 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
stirred at 25-30 °C. Sodium phosphate (9.47 kg, 58 mol) was charged into the solution
and the pH was adjusted to 7.3 using a solution of sodium dihydrogen phosphate (8.17
kg, 68 mol) in water (59 L). In a second vessel, sodium phosphate (4.53 kg, 28 mol)
was dissolved in water (118 L), and the pH was adjusted to 7.3 using a solution of
sodium dihydrogen phosphate (3.12 kg, 26 mol) in water (22.8 L). To this second
buffered solution was charged adenosine deaminase ADA-101 from Codexis (470 g, 22
units/mg activity, 176 units enzyme/g 2′-MOE-DAPR 5). The enzyme solution was
cooled and maintained at 5-10 °C for 90 min, and then charged into the solution of 5,
which was held at 25-30 °C. The reaction mass was maintained at 25-30 °C and a pH of
7.3, adjusting the pH as necessary using a solution of sodium dihydrogen phosphate
(25 kg) in water (37 L). After 8 hours of maintenance, seed crystals of 6 (94 g) were
added to the reaction, and the reaction was maintained at 25-30 °C and a pH of 7.3 until
the residual level of 5 was not more than 1.0% (another 12.25 hr, 20.25 hr in total). The
reaction mass was cooled to 15-20 °C and maintained for 35 min, before collecting the
2′-O-(2-methoxyethyl)guanosine product by filtration. The filter cake was washed with
water (118 L), slurried in acetone (118 L), filtered, and transferred into a hot air dryer
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
29