Synthesis of Akt inhibitor ipatasertib. part 1. Route scouting and early process development of a challenging cyclopentylpyrimidine intermediate

Article abstract of DOI:10.1021/op500271w

Herein, the route scouting and early process development of a key cyclopentylpyrimidine ketone intermediate toward the synthesis of Akt inhibitor Ipatasertib are described. Initial supplies of the intermediate were prepared through a method that commenced with the natural product (R)-(+)-pulegone and relied on the early construction of a methyl-substituted cyclopentyl ring system. The first process chemistry route, detailed herein, enabled the synthesis of the ketone on a hundred-gram scale, but it was not feasible for the requisite production of multikilogram quantities of this compound and necessitated the exploration of alternative strategies. Several new synthetic approaches were investigated towards the preparation of the cyclopentylpyrimidine ketone, in either racemic or chiral form, which resulted in the discovery of a more practical route that hinged on the initial preparation of a highly substituted dihydroxypyrimidine compound. The cyclopentane ring in the target was then constructed through a key carbonylative esterification and subsequent tandem Dieckmann cyclization-decarboxylation sequence that was demonstrated in a racemic synthesis. This proof-of-concept was later developed into an asymmetric synthesis of the cyclopentylpyrimidine ketone, which will be described in a subsequent paper, along with the synthesis of Ipatasertib.

Full text of DOI:10.1021/op500271w

Subscriber access provided by Maastricht University Library
Full Paper
Synthesis of Akt Inhibitor GDC-0068 (Ipatasertib). Part
I: Route Scouting and Early Process Development of
a Challenging Cyclopentylpyrimidine Intermediate
Jonathan Lane, Keith L. Spencer, Sagar Raj Shakya,
Nicholas C. Kallan, Peter J. Stengel, and Travis Remarchuk
Org. Process Res. Dev., Just Accepted Manuscript • Publication Date (Web): 26 Oct 2014
Downloaded from http://pubs.acs.org on October 27, 2014
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.
Organic Process Research & Development 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
Synthesis of Akt Inhibitor Ipatasertib. Part I: Route
Scouting and Early Process Development of a
Challenging Cyclopentylpyrimidine Intermediate
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
Jonathan W. Lane,*,† Keith L. Spencer,†,§ Sagar R. Shakya,†,║ Nicholas C. Kallan,†,€ Peter J.
Stengel†,∞ and Travis Remarchuk*,‡
†Array BioPharma Inc., 3200 Walnut Street, Boulder, Colorado 80301, United States
‡Genentech Inc., 1 DNA Way, South San Francisco, California 94080ꢀ4990, United States
ABSTRACT: In Part I of this series, the route scouting and early process development of a key
cyclopentylpyrimidine ketone intermediate toward the synthesis of Akt inhibitor Ipatasertib are
described. Initial supplies of the intermediate were prepared through a method that commenced
with the natural product (R)ꢀ(+)ꢀpulegone and relied on the early construction of a methylꢀ
substituted cyclopentyl ring system. The first process chemistry route, detailed herein, enabled
the synthesis of the ketone on hundred gram scale, but was not feasible for the requisite
production of multikilogram quantities of this compound and necessitated the exploration of
alternative strategies. Several new synthetic approaches were investigated towards the
preparation of the cyclopentylpyrimidine ketone, in either racemic or chiral form, which resulted
in the discovery of a more practical route that hinged on the initial preparation of a highlyꢀ
substituted dihydroxypyrimidine (DHP) compound. The cyclopentane ring in the target was then
1
ACS Paragon Plus Environment

Organic Process Research & Development
Page 2 of 38
1
2
3
4
5
6
7
8
9
constructed through a key carbonylative esterification and subsequent tandem Dieckmannꢀ
decarboxylation sequence that was demonstrated in a racemic synthesis. This proofꢀofꢀconcept
was later developed into an asymmetric synthesis of the cyclopentylpyrimidine ketone, which is
described in Part II of this series in addition to the synthesis of Ipatasertib.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
INTRODUCTION
Akt/Protein Kinase B (PKB) is a member of the AGC subfamily and regulates diverse
cellular activities including cell growth, proliferation, apoptosis and metabolism.1 In response to
the observation that Akt plays a key role in the survival and progression of various types of
cancer cells,2 the concept of Akt inhibition has generated significant interest in the oncology
community.3 Through an extensive collaboration between the discovery chemistry groups at
Array and Genentech (Roche), numerous molecular scaffolds were successfully devised and
investigated both in vitro and in vivo as potential selective inhibitors of Akt, which ultimately
resulted in the nomination of Ipatasertib (Figure 1) for clinical development.4
Retrosynthetically, Ipatasertib is assembled in a convergent manner from two key
components where cyclopentylpyrimidine core transꢀ1, which possesses a piperazine linker, is
coupled to β2ꢀamino acid (S)ꢀ2 (Figure 1). In the medicinal chemistry synthesis of Ipatasertib, the
first coupling partner (S)ꢀ2 was prepared from commercially available 4ꢀchlorophenylacetic acid,
and two subsequent asymmetric syntheses of this chiral compound have recently been
communicated.5 The second coupling partner transꢀ1 arose from an asymmetric reduction of
ketone (R)ꢀ3, which was originally derived from the natural product (R)ꢀ(+)ꢀpulegone. Since (R)ꢀ
(+)ꢀpulegone was a starting material from the chiral pool, it was initially attractive from a
process pointꢀofꢀview as it provided a reliable source of chirality for the methyl group in (R)ꢀ3,
2
ACS Paragon Plus Environment

Page 3 of 38
Organic Process Research & Development
1
2
and was readily available on kilogram scale.6 Furthermore, (R)ꢀ(+)ꢀpulegone has been widely
used for the preparation of several natural products containing a methyl stereocenter in a
cyclopentyl ring system that was accessed from the wellꢀprecedented Favorskii rearrangement.7
The discovery route to ketone (R)ꢀ3 provided a viable basis for further development after
addressing several concerns including (1) the ozonolysis reaction8 that required cryogenic
temperatures as well as a facility with an ozone generator, (2) high reaction temperatures (150 °C
for 24 h) required to forge the pyrimidine ring system, and (3) the use of undesirable solvents
and reagents such as Et2O (Favorskii rearrangement), CHCl3 (mꢀCPBA oxidation) and PDC
(benzylic alcohol oxidation), which were not attractive for scaleꢀup.4a Another drawback to the
discovery route was the requirement for installation of the benzylic ketone group in (R)ꢀ3, which
is eventually reduced to provide transꢀ1. Albeit cumbersome, installation of the ketone group
was successfully accomplished by a Boekelheideꢀtype oxidationꢀrearrangement reaction9 and
subsequent Swern oxidation that enabled synthesis of the first hundredꢀgram quantities of ketone
(R)ꢀ3.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Retrosynthesis of Ipatasertib.
3
ACS Paragon Plus Environment

Organic Process Research & Development
Page 4 of 38
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In response to the fact that the (R)ꢀ(+)ꢀpulegone route was lengthy, lowꢀyielding and
employed extensive oxidation chemistry, and concomitant with our initial scaleꢀup efforts,4a two
new distinct strategies were examined and explored for the preparation of ketone (R)ꢀ3 that
would potentially better facilitate multikilogram scale production (Figure 2). In the first approach
(S1), pyrimidine ring construction was envisioned to arise from the condensation of a
formamidine equivalent onto a preꢀfunctionalized cyclopentanedione compound such as (R)ꢀ4.10
In contrast, a second approach (S2) would entail sequentially building the pyrimidine ring
followed by the cyclopentane ring in (R)ꢀ3 from a triester compound (R)ꢀ5. This second strategy
was advantageous with respect to enabling the assembly of the key chiral unit in (R)ꢀ3 through
mild intramolecular annulation approaches with the pyrimidine component already in place.
Moreover, the latter approach (S2) was the most attractive, because we could employ simple
acyclic ester derivatives that would potentially allow for access to chiral products through
enzymatic resolution methods. The scaleꢀup of the (R)ꢀ(+)ꢀpulegone route to (R)ꢀ3 and
4
ACS Paragon Plus Environment

Page 5 of 38
Organic Process Research & Development
1
2
3
4
5
6
alternative strategy investigations (S1–S2) that resulted in the development and racemic
demonstration of a new lead route to (R)ꢀ3 will be discussed herein.
7
8
9
Figure 2. Route scouting strategies for the preparation of ketone (R)ꢀ3.
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
1. Process scaleꢀup of the (+)ꢀpulegone route to ketone (R)ꢀ3. At the time of this work, the
medicinal chemistry group had reported the conversion (R)ꢀ(+)ꢀpulegone to the Favorskii
rearrangement product (R)ꢀ6 in 64% isolated yield on 600 g scale (Scheme 1).11 This procedure
was reproducible; however, we sought to eliminate the use of Et2O for the bromination step prior
to further scaleꢀup, and DCM was found to be a suitable and safer alternative. Therefore, starting
with a DCM solution of (R)ꢀ(+)ꢀpulegone12 and sodium bicarbonate, slow addition of 1.02 equiv
Br2 at –10 °C resulted in nearꢀcomplete conversion to the corresponding dibromide intermediate
(monitored by TLC). Following filtration to remove the inorganic solids, the Favorskii
rearrangement was initiated with the addition of NaOEt in EtOH at –40 °C and subsequent
warming to ambient temperature. The reaction mixture was then acidified, and the DCM solution
was concentrated to a give the crude product (R)ꢀ6 following an aqueous workꢀup. The crude
product mixture contained ketone impurities, such as unreacted (R)ꢀ(+)ꢀpulegone and likely other
unidentified reaction intermediates and/or byproducts, which required removal. These impurities
5
ACS Paragon Plus Environment

Organic Process Research & Development
Page 6 of 38
1
2
3
4
5
6
7
8
9
were successfully purged by initial treatment of crude (R)ꢀ6 with aqueous semicarbazide to form
the corresponding waterꢀsoluble semicarbazone13 derivatives [I]. The cyclopentyl ester (R)ꢀ6
could then be selectively extracted into EtOAc and isolated as an oil, which was of acceptable
quality for further processing. On 5 kg scale, (R)ꢀ(+)ꢀpulegone was converted to cyclopentyl
ester (R)ꢀ6 (3.9 kg, 60% yield with ~85 A% GC purity) as a 2:3 mixture of cis/trans isomers.14
The next step required cleavage of the exocyclic alkene in (R)ꢀ6 in order to unmask the βꢀ
ketoester functionality in (R)ꢀ7. Although alternative approaches were examined to replace the
existing ozonolysis, none were as efficient.15 Two modifications were made to the medicinal
chemistry procedure where (1) EtOAc was replaced with DCM, and (2) an alternative quench of
the ozonide was employed. The use of DCM was driven by the improvement realized in
determining the ozone saturation endpoint, which was accomplished by the visual observation of
a characteristic blue color. As well, several methods are commonly used for quenching ozonide
intermediates including Ph3P, Zn/AcOH and dimethyl sulfide (DMS), but in our experience
DMS was preferred for scaleꢀup and led to a reduction in the exotherm associated with the
quench in addition to a more straightforward charging protocol at cryogenic temperatures (–78
°C). In a typical experiment, starting with 500 g cyclopentyl ester (R)ꢀ6, the ozonolysis reaction
was complete in ≤24 h. Subsequent quench of the deep blue solution with DMS followed by
gradual warming to room temperature and aqueous extraction required an additional ~20 h of
processing time before the crude βꢀketoester intermediate (R)ꢀ7 was isolated as a yellow liquid in
~75% yield. A total of 10 kg (R)ꢀ(+)ꢀpulegone was processed over the two steps to provide ~6 kg
of βꢀketoester (R)ꢀ7 in multiple batches and variable purity (77–94 A%) by GC. The inconsistent
purity was believed to be due to the sensitivity of the ozonolysis workꢀup on increased scale;
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
6
ACS Paragon Plus Environment

Page 7 of 38
Organic Process Research & Development
1
2
3
4
however, the purity of (R)ꢀ7 was not critical to the function of the subsequent step to construct
5
6
the thiopyrimidine.
7
8
9
Scheme 1. Scaleꢀup of the Favorskii rearrangement and ozonolysis reactions.
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 medicinal chemistry group had previously employed two principle methods to
construct the pyrimidine ring associated with Ipatasertib from (R)ꢀ7 (Scheme 2). The first
method entailed a stepꢀwise condensation whereby βꢀketoester (R)ꢀ7 was initially treated with
NH4OAc to produce the corresponding enamine derivative, which was then condensed with
ammonium formate/formamide at ≤150 °C to give pyrimidine (R)ꢀ9 in ~60% yield over the two
steps.4a In the second method, the more nucleophilic thiourea was initially reacted with βꢀ
ketoester (R)ꢀ7 under milder reaction conditions (≤85 °C) to produce thiopyrimidine (R)ꢀ8 in
modest yields (45–55%), after which a subsequent desulfurization step was required to generate
pyrimidine (R)ꢀ9.16 Although the first process afforded improved yields over the thiourea
method, the higher reaction temperatures associated with this procedure were unsafe for scaleꢀup
as ammonium formate could sublime and plug the reflux condenser.17 Therefore, the thiourea
7
ACS Paragon Plus Environment

Organic Process Research & Development
Page 8 of 38
1
2
3
4
5
6
7
8
9
approach was ultimately selected for scaleꢀup based on these safety concerns. On scaleꢀup of this
process, condensation of the βꢀketoester (R)ꢀ7 with thiourea was promoted by the slow addition
of DBU to the reaction mixture, which resulted in complete consumption of starting material
after ~20 h at reflux. The product was precipitated by the addition of aqueous HCl and was then
recrystallized from iꢀPrOH to afford thiopyrimidine (R)ꢀ8 in ~50% yield with >99 A% HPLC
purity on kilogram scale. Subsequent reductive desulfurization of (R)ꢀ8 with Raney® Nickel in
ammonium hydroxide afforded hydroxypyrimidine (R)ꢀ9 in yields ranging from 83–94%.18
Installation of the piperazine linker in (R)ꢀ3 commenced with the chlorination of (R)ꢀ9 using neat
POCl3 and heating to 100 °C.19 On scaleꢀup this resulted in clean formation of chloropyrimidine
(R)ꢀ10 in 91% isolated yield and >99 A% purity.20 Nucleophilic aromatic substitution of
chloropyrimidine (R)ꢀ10 with NꢀBocꢀpiperazine then provided (R)ꢀ11 (79%, >99 A% by HPLC)
following a silica gel plug filtration and crystallization from EtOAc/heptane.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In order to install the requisite benzylic ketone group in (R)ꢀ3, a Boekelheideꢀtype
oxidationꢀrearrangement sequence was performed. The twoꢀstep transformation entailed the
initial generation of pyrimidine Nꢀoxide (R)ꢀ12 through treatment of (R)ꢀ11 with mꢀCPBA,
followed by subsequent heating of (R)ꢀ12 in acetic anhydride to afford rearrangement product
(R)ꢀ13. In the original discovery synthesis, CHCl3 was employed for the mꢀCPBAꢀmediated
oxidation; however, DCM was found to be a suitable replacement for this stage of
development.21 On scaleꢀup of this reaction, up to 4 equiv of mꢀCPBA was required in order to
achieve complete conversion and gave 74% isolated yield of pyrimidine Nꢀoxide (R)ꢀ12 after
passage of the crude mixture of a plug of silica gel.22 The resulting brown oil was then heated in
neat acetic anhydride to facilitate the desired oxidative transposition, producing the acetate
derivative (R)ꢀ13 (~ 1:1 dr) as a black tarꢀlike residue. Cleavage of the benzylic acetate group in
8
ACS Paragon Plus Environment

Page 9 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
(R)ꢀ13 with aqueous LiOH in THF and purification of the isolated product via a silica gel plug
followed by crystallization from ether/heptane afforded the benzylic alcohol (R)ꢀ14 in ~99 A%
purity and modest yields (41–54%) for the threeꢀstep sequence. A Swern oxidation protocol had
been previously employed in prior efforts to convert the hydroxyl group in (R)ꢀ14 to ketone (R)ꢀ
3. However, it was later determined that residual DMSO and dimethylsulfide byproducts
associated with (R)ꢀ3 produced in this manner were responsible for poisoning the ruthenium
catalyst in the subsequent asymmetric ketone reduction to (R,R)ꢀ1 (Figure 1). From the
examination of alternative oxidation reagents, pyridinium dichromate (PDC) afforded the highest
yield of product (R)ꢀ3 and circumvented the introduction of any potential impurities that could
adversely affect the subsequent asymmetric reduction.23 In the process, PDC was added
portionwise to a solution of (R)ꢀ14, resulting in complete consumption of the alcohol starting
material after stirring overnight at ambient temperature. Purification of the crude product over a
plug of silica gel, which was required in order to remove several unknown impurities arising
from the intrinsic instability of (R)ꢀ3 under the reaction conditions, afforded the target ketone in
modest 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
9
ACS Paragon Plus Environment

Organic Process Research & Development
Page 10 of 38
1
2
3
4
Scheme 2. First generation process route for the preparation of cyclopentyl ketone (R)ꢀ3.
5
6
7
8
9
i) thiourea (1.5 equiv)
DBU (1.3 equiv)
OH
OH
i) neat POCl3
100 °C
CH3
CH3
CO2Et
CH3
Raney-Ni
O
N
N
NH4OH
reflux
(83-94%)
ii) concentrate
iii) Aq. work-up
(91%)
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
ACN, reflux
ii) recryst. i-PrOH
HS
N
N
(R)-7
(50%)
(R)-8
(R)-9
Boc
N
Boc
N
Boc
N
Cl
CH3
N
CH3
N
H
N
1) m-CPBA, CH2Cl2
CH3
N
N
N
2) SiO2 (plug filtration)
(74%)
DIPEA, n-BuOH
N
N
N
(79%)
(R)-10
O
(R)-12
(R)-11
Boc
Boc
N
Boc
N
N
PDC [O]
DCM, rt
Ac2O
95 °C
LiOH, THF/H2O
N
N
N
CH3
O
CH3
CH3
OAc
(41-54%
3 steps)
N
N
N
(36-43%)
N
N
N
OH
R-(3)
(R)-14
(~ 1:1 dr)
(R)-13
(~ 1:1 dr)
2. (R)ꢀ(+)ꢀPulegone route liabilities. Although scaleꢀup of the discovery synthesis successfully
facilitated the critical initial delivery of 0.2 kg of ketone (R)ꢀ3 in ~2% overall yield from a 5 kg
input of (R)ꢀ(+)ꢀpulegone, this route was not practical for large scale production. The ozonolysis
reaction posed significant concerns due to long reaction times, modest yields and the scarce
availability of vendors possessing the capabilities necessary to conduct the transformation on
multikilogram scale. A suitable replacement to the ozonolysis reaction for cleavage of the
exocyclic alkene in (R)ꢀ6 was not available. As well, the lengthy and cumbersome benzylic
oxidation sequence (Boekelheide reaction) late in the route represented perhaps the most
significant challenge for two reasons. First, the initial oxidation to produce Nꢀoxide (R)ꢀ12
10
ACS Paragon Plus Environment

Page 11 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
required the use of mꢀCPBA, which presented safety concerns on scaleꢀup. Second,
rearrangement of (R)ꢀ12 to (R)ꢀ13 was not stereoselective, which necessitated a subsequent lowꢀ
yielding oxidation to ketone (R)ꢀ3 en route to an asymmetric reduction to access transꢀ1 (Figure
1). Indeed, the shortcomings associated with the (R)ꢀ(+)ꢀpulegone route to (R)ꢀ3 highlighted the
importance of establishing a new path to this key intermediate. Therefore, several alternative
strategies (S1ꢀS2, Figure 2) were explored toward the preparation of ketone (R)ꢀ3, initially with
racemic inputs in an effort to rapidly assess the feasibility of the key bondꢀforming steps.
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
3. Route scouting approach S1: Pyrimidine construction. The first approach (S1, Figure 2)
was an extension of the original (R)ꢀ(+)ꢀpulegone route and was initially chosen for investigation
in part to avoid the ozonolysis step associated with the preparation of βꢀketoester (R)ꢀ7 (Scheme
2). We envisioned that use of the alternative cyclopentadione compound racꢀ410 could also
obviate the need for the Boekelheide oxidation/rearrangement used in the scaleꢀup route to install
the hydroxyl functionality in (R)ꢀ14 that was subsequently oxidized to ketone (R)ꢀ3. Therefore,
we prepared gram quantities of cyclopentadione racꢀ4 as well as the hydroxylꢀsubstituted
derivative (racꢀ16)24 according to literature methods and tested these precursors in the key step
to forge the corresponding pyrimidine (or thiopyrimidine) ring system (Scheme 3). We adopted
reaction conditions similar to those employed in the previous route that entailed treatment of the
substrate in various solvents with either thiourea or formamidine in combination with an
alkoxide base (entries 1ꢀ3, Table 1). Unfortunately, all attempts to react either the
cyclopentadione compound racꢀ4 or the hydroxylꢀsubstituted derivative racꢀ16 under these
conditions failed to provide the desired products, ketone racꢀ15 or alcohol racꢀ17, respectively,
in any appreciable quantity as indicated by LCMS analysis. Specifically, at lower reaction
temperatures (≤50 °C) only the monoꢀaddition enamine intermediate [II] and unreacted starting
11
ACS Paragon Plus Environment

Organic Process Research & Development
Page 12 of 38
1
2
3
4
5
6
7
8
9
material were observed, suggesting that higher temperatures were required to promote the
cyclization. Although trace product was observed in some cases by LCMS, the more forcing
reaction conditions typically resulted in a complex mixture of products. From these initial
results, we concluded that we must develop a route to ketone 3 that either commenced with the
pyrimidine ring system in place or, alternatively, employed a more robust substrate that could
tolerate the harsh reactions conditions required for pyrimidine ring formation. Our next approach
would address this theory.
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. Cyclization attempts for oxygenꢀsubstituted cyclopentane compounds (S1).
OH
CH3
CH3
Entry: 1 or 2
EtO2C
R1O
N
X
R2
N
(Entry 2)
O
O
rac-15
(R2 = H, SH)
rac-4
(R1 = H, Bn)
O
CH3
EtO
HN
~ 7.8 ppm
(d6-DMSO)
X
H
NH
II
OH
CH3
CH3
EtO2C
BnO
N
X
Entry: 3
N
OR3
OR3
rac-16
(R3 = H, Bn)
rac-17
(R3 = H, Bn)
~3:2 cis/trans
Table 1. Reaction conditions to construct thiopyrimidine and pyrimidine derivatives.
Entrya Experimental Details
1
2
Thiourea, solvent (MeOH, EtOH, toluene), >100 ˚C.
HN=CHNH2 (AcOH or HCl salt), NaOMe/MeOH or NaOEt/EtOH, 80 ˚C.
HN=CHNH2 (AcOH salt), base (none, NaOMe, NaOEt, NaOtBu), solvent (EtOH, DMF,
DMSO), 50 ˚C.
3
aAll experiments were performed in sealed tubes and analyzed by LCMS (λ = 254 nm) after 8–12 h.
12
ACS Paragon Plus Environment

Page 13 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
4. Route scouting approach S2: Dual ring construction. Based on our observations from
investigations pertaining to S1,25 we proceeded to a second approach (S2, Figure 2), which relied
on acyclic triester substrates for the construction of both rings comprising the bicyclic core in
ketone (R)ꢀ3. Pyrimidine diester racꢀ19 was a logical target that could arise from ketoꢀtriester
racꢀ18 and ultimately provide the desired cyclopentyl ketone ring system in 3 via racꢀ20 through
a Dieckmann cyclization and decarboxylation sequence (Scheme 4).26 To access racꢀ19, we
prepared the αꢀoxalyl methylglutarate substrate racꢀ18 using known literature methods.27
Unfortunately, our attempts to cyclize racꢀ18 in the presence of formamidine acetate and
NaOEt/EtOH suffered a similar issue to that which we observed in prior investigations with the
cyclopentaneꢀbased substrates (Scheme 3). Only the initial condensation product enamine [III],
in addition to unreacted starting material, were observed at ambient reaction temperatures (≤23
°C). Further heating the reaction mixture resulted in complex product mixtures, and diester racꢀ
19 was not observed.
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 4. Attempted pyrimidine formation with oxalyl methylglutarate racꢀ18 (S2).
13
ACS Paragon Plus Environment

Organic Process Research & Development
Page 14 of 38
1
2
3
4
5
6
7
8
9
Therefore, efforts were refocused on the cyclization of an alternative triester substrate racꢀ21 that
was easily accessed from ethyl crotonate and diethylmalonate through reported methods.28 To
our delight, treatment of racꢀ21 with formamidine generated dihydroxypyrimidine racꢀ22 in an
unoptimized 30% yield (Scheme 5). Evaluation of the key Dieckmann cyclization strategy
toward establishing the cyclopentane unit in (R)ꢀ3 next required conversion of racꢀ22 to the
corresponding dichloropyrimidine intermediate through treatment with POCl3, which was then
reacted with NꢀBoc piperazine using SNAr conditions similar to those employed in the (R)ꢀ(+)ꢀ
pulegone route (Scheme 2) in order to arrive at Dieckmann substrate precursor chloropyrimidine
racꢀ23.
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 5. Successful preparation of chloropyrimidine racꢀ23 (S2).
Boc
N
NH2
OH CH3
HOAc
1. POCl3, ACN,
85 °C
CH3
CO2Et
H
NH
N
CH3
Cl
N
EtO2C
CO2Et
CO2Et
N
NaOEt, EtOH
23 °C–reflux
(30%)
N
OH
CO2Et
2.DIEA, EtOH,
50 °C
rac-22
rac-21
N
NBoc
HN
rac-23
(57%, 2 steps)
Having established a path to racꢀ23, the next step was to install the requisite aryl ester
group in racꢀ20 that was not directly accessible through racꢀ18. The palladiumꢀcatalyzed
carbonylative esterification of aryl chlorides, including 4ꢀchloropyrimidines, is wellꢀ
precedented,29 and a brief survey of the reaction conditions reported by AlbanezeꢀWalker and coꢀ
workers29a expeditiously revealed that the use of 1,3ꢀbis(diphenylphosphino)propane (dppp) with
Pd(OAc)2 was the most effective catalyst system for the conversion of racꢀ23 to ethyl ester
14
ACS Paragon Plus Environment

Page 15 of 38
Organic Process Research & Development
1
2
derivative racꢀ20 (Scheme 6).30 In attempts to rapidly validate the Dieckmann approach to (R)ꢀ3,
formation of the cyclopentane ring was subsequently effected under mild conditions, generating
ketoester racꢀ24 with complete consumption of starting material and clean reaction profile by
LCMS analysis.31 Unfortunately, removal of the unwanted ethyl ester group in racꢀ24 proved
very challenging. Exposure of racꢀ24 to a variety of systems typically used to promote this
transformation including base (LiOH/THFꢀH2O, KOH/MeOH), acid (TsOH/MeOHꢀH2O,
HCO2H) and neutral (Krapchoꢀtype: LiCl, DMSO)32 conditions, all resulted in either recovered
starting material or degradation products after heating the reaction mixture. From these studies it
was reasoned that the instability associated with racꢀ24 as well as the product racꢀ3,33
necessitated an alternative ester group that was amenable to decarboalkoxylation under milder
reaction conditions.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 6. Cyclopentane ring formation through a Dieckmann condensation approach.
15
ACS Paragon Plus Environment

Organic Process Research & Development
Page 16 of 38
1
2
3
4
5
6
7
8
9
To this end, racꢀ27, the corresponding benzyl ester sideꢀchain derivative of racꢀ20, was a
logical substrate to investigate in the subsequent Dieckmannꢀdecarboalkoxylation sequence.34 In
practice, we found it more convenient to install the requisite benzyl ester sideꢀchain in racꢀ27
through the initial hydrolysis of the ethyl ester group of racꢀ23 followed by the reꢀesterification
of the resulting carboxylic acid intermediate (racꢀ25), as opposed to accessing the modified sideꢀ
chain from the corresponding benzyl triester derivative of racꢀ21. Using this method, racꢀ26 was
produced in 86% yield over two steps and employed in the subsequent Pdꢀmediated
carbonylative esterification reaction. Despite the potential for ester group exchange with EtOH,
we initially chose to use substantially similar reaction conditions to those employed with the
previous substrate racꢀ23 in order to test our hypothesis, and the carbonylation proceeded to
provide mixed diester racꢀ27 in 27% yield after purification.35 The Dieckmann cyclization of
racꢀ27 to racꢀ28 was then successfully promoted through the agency of KOtBu in THF, resulting
again in complete conversion for this transformation.36 Following a simple neutralization with
aqueous HCl, benzyl ketoester racꢀ28 was immediately subjected in situ to hydrogenolysis
conditions (H2, Pd/carbon, THF), effectively enabling the desired decarboalkoxylation to racꢀ3
in excellent conversion (>95% by LCMS) and securing proofꢀofꢀconcept for the new strategy on
milligram scale.37 However, in this initial demonstration, the overꢀreduced racꢀ29 (>10:1
cis/trans) was also observed as a byproduct and suggested that additional optimization would be
required in attempts to suppress this undesired pathway.
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. Optimized lead route to ketone racꢀ3. Following the initial successful establishment of the
Dieckmannꢀbased strategy toward ketone racꢀ3, additional optimization was conducted, and
scaleꢀup was demonstrated in order to rapidly establish the viability of the new approach. The
synthesis commenced from trimethyl ester racꢀ30 instead of the ethyl ester derivative racꢀ21,
16
ACS Paragon Plus Environment

Page 17 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
and gave dihydroxypyrimidine (DHP) racꢀ31 in 87% yield (Scheme 7). Improved performance
in this reaction was realized with the NaOMe/MeOH based cyclization system due to the
enhanced solubility in the reaction mixture, which maintained a homogenous composition. The
waterꢀsoluble DHP racꢀ31 product was isolated through filtration following pH adjustment with
HCl (g) and solvent exchange to toluene. One notable observation associated with this procedure
was the contamination of racꢀ31 with significant quantities of sodium chloride that had
essentially coꢀprecipitated during the isolation. However, the presence of sodium chloride in the
product did not adversely impact the subsequent chlorination as the salts were purged in the
workup of this transformation.38
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 conversion of (DHP) racꢀ31 to dichloropyrimidine intermediate racꢀ32 through
treatment with phosphorous oxychloride (POCl3) initially posed a notable safety concern with
respect to the further scaleꢀup of this operation. The source of concern arose from a significant
exotherm that was consistently observed upon heating the mixture following the introduction of
all reaction components. Fortunately, the exotherm was successfully managed through the use of
2,6ꢀlutidine as a promoter in the context of a controlled charge of POCl3 at elevated temperature
(70 ˚C) where the accumulation of reactive intermediates could be avoided.39 Subsequent
introduction of the NꢀBocꢀprotected piperidine linker through nucleophilic aromatic substitution
of racꢀ32 proceeded in MeOH, circumventing the likelihood of ester group exchange to give
intermediate racꢀ33. Consistent with prior observations, the corresponding byproduct arising
1
from the addition of two equivalents of NꢀBoc piperidine to racꢀ32 was not observed by H
NMR spectroscopy. Hydrolysis of the methyl ester group in racꢀ33 was accomplished under
mild conditions to provide carboxylic acid intermediate racꢀ25. In the threeꢀstep sequence (racꢀ
17
ACS Paragon Plus Environment

Organic Process Research & Development
Page 18 of 38
1
2
3
4
5
6
31 to racꢀ25), intermediates racꢀ32 and racꢀ33 were processed without any formal purification,
and compound racꢀ25 was isolated by filtration in 77% overall yield.
7
8
9
Scheme 7. Secondꢀgeneration process chemistry route to cyclopentylpyrimidine ketone racꢀ3.
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
Esterification of carboxylic acid racꢀ25 was accomplished using conditions similar to
those employed in our initial studies (Scheme 6), and benzyl ester racꢀ26 was produced in
quantitative yield and used without further purification. Conversely, the subsequent
carbonylative esterification of benzyl ester racꢀ26 had been lowꢀyielding in the context of the
reaction conditions employed in our initial demonstration, and conversion as well as exchange of
the benzyl ester group with the reaction solvent (EtOH) were identified as key issues. A
preliminary survey of ligands, Pdꢀsources and bases confirmed that our original catalyst system
recruited to promote this transformation (dppp, Pd(OAc)2, K2CO3) was optimal, and we
proceeded to examine alternative solvents. From a screen of various alcohol solvents40 and coꢀ
18
ACS Paragon Plus Environment

Page 19 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
solvents, a 3:1 mixture of iPrOH and THF was identified as an optimal solvent system whereby
the reaction would proceed to completion with nearꢀcomplete suppression of benzyl ester group
exchange,41 and the resulting aryl isopropyl ester group was competent in the subsequent
Dieckmann cyclization. In our final demonstration (50 g scale), the carbonylation proceeded
smoothly to give the mixed ester racꢀ34 in 71% yield after purification over a plug of silica gel
and subsequent trituration (heptane/MTBE).42
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For the key Dieckmann cyclization and decarboxylation sequence, the choice of Pd
catalyst was found to be critical. Our initial proofꢀofꢀconcept demonstration for the
hydrogenolysis of intermediate racꢀ28 (Scheme 6) had employed 10 wt% Pd/carbon as catalyst
in the presence of H2 (1 atm), resulting in complete conversion (>95% by LCMS) to the ketone
racꢀ3 in addition to the formation of racꢀ29 from further reduction of the ketone group. In order
to suppress the observed overꢀreduction in this step, we found that the use of Pd/alumina (5
wt.%) as catalyst promoted the removal of the benzyl ester group and resulted in <1% of
compound racꢀ29 in our final 25 g demonstration. Using this procedure, ketone racꢀ3 was
produced from racꢀ34 and then isolated by precipitation (THF/heptane) in 91% yield (>99 A%)
wherein alcohol racꢀ29 was effectively rejected into the filtrate.
CONCLUSION
A new lead route was developed and demonstrated in racemic form toward preparation of the
cyclopentylpyrimidine ketone (R)ꢀ3 in 8 linear steps and 39% overall yield from triester racꢀ30.
The new strategy replaced5a the less practical firstꢀgeneration scaleꢀup route to (R)ꢀ3 that
commenced with (R)ꢀ(+)ꢀpulegone and only enabled the production of hundred gram quantities
of ketone (R)ꢀ3 in ~2% overall yield. Two key route scouting approaches were investigated, and
19
ACS Paragon Plus Environment

Organic Process Research & Development
Page 20 of 38
1
2
3
4
5
6
7
8
9
the method based on dualꢀring construction (S2) proved to be the most successful. In this
approach, construction of the pyrimidine ring of DHP racꢀ31 was accomplished in one step from
triester racꢀ30 in 87% yield. The DHP racꢀ31 was then readily converted into the diester racꢀ34
through five additional steps that culminated in a key Pdꢀcatalyzed carbonylation reaction, which
was used to install the aryl ester group necessary for the key intramolecular cyclization. In the
main event, a telescoped Dieckmann cyclization / decarboxylation sequence of diester racꢀ34
was used to forge the cyclopentanone ring system of (R)ꢀ3 in 91% yield. A salient feature of this
route is the ability to introduce chirality through an enzymatic resolution of the triester racꢀ30,
and this approach is demonstrated in the kilogram synthesis of Ipatasertib in Part II of this series.
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
1
Reagents and solvents were purchased from commercial sources and were used as received. H
and 13C NMR spectra are reported in parts per million (δ) at 400 MHz and 100 MHz,
1
respectively. In cases of H NMR spectra where mixtures of compounds are present, the desired
product is reported. In cases where weight percent assays were employed to determine product
purity, corrected yields were reported. High resolution mass spectra (HRMS) were recorded by
Dr. John Greaves at the University of California, Irvine.
Trimethyl 2ꢀmethylpropaneꢀ1,1,3ꢀtricarboxylate, racꢀ30. A 12 L 4ꢀneck roundꢀbottom flask
was charged with THF (3.5 L) followed by 25 wt% NaOMe/MeOH (128 mL, 562 mmol, 0.25
equiv). Dimethylmalonate (259 mL, 2.27 mol, 1.01 equiv) was then added to the cloudy agitated
solution at 23 °C over 10 min, which was followed by the addition of transꢀmethyl crotonate
(239 mL, 2.25 mol, 1.00 equiv) over 10 min. The resulting pale yellow reaction mixture was
heated (45ꢀ50 °C) and agitated for 4 h. The reaction was determined to be complete by inꢀ
1
process analysis using H NMR spectroscopy (<2% transꢀmethyl crotonate). The reaction was
20
ACS Paragon Plus Environment

Page 21 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
cooled to ~2 °C (ice bath) over 1 h, after which a solution of AcOH (34.7 mL, 607 mmol, 0.27
equiv) in H2O (347 mL) was charged over 15 min affording a cloudy, colorless solution. After
agitation for 30 min, the reaction was then concentrated via rotovap, and the resulting residue
was partitioned between EtOAc (2.0 L) and 0.5M aq. NaHCO3 (1.2 L). The layers were
separated, and the aqueous phase was extracted with EtOAc (400 mL). The combined organic
layers were washed with saturated aq. NaCl (500 mL), dried (MgSO4), filtered and concentrated
via rotary evaporation to give triester racꢀ30 as a colorless oil (0.50 kg, 2.17 mol, 96% yield). 1H
NMR (400 MHz, CDCl3) 3.74 (s, 3H), 3.74 (s, 3H), 3.69 (s, 3H), 3.46 (d, J = 7.2 Hz, 1H), 2.75
(m, 1H), 2.55 (1/2 ABX, JAB = 16.0 Hz, 1H), 2.33 (1/2 ABX, JAB = 15.8 Hz, 1H), 1.07 (d, J =
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
6.8 Hz, 3H); C NMR (100 MHz, CDCl3) 172.6, 169.0, 168.9, 56.0, 52.5 (2), 51.7, 38.5, 30.3,
17.7; HRMS (ESI+) calculated for C10H16ONa [M + Na]+ (m/z) 255.0845; found (m/z) 255.0844.
Methyl 3ꢀ(4,6ꢀdihydroxypyrimidinꢀ5ꢀyl)butanoate (DHP), racꢀ31. A 12 L 4ꢀneck roundꢀ
bottom flask was charged with MeOH (2.8 L), and the reaction headspace was purged with
nitrogen using two evacuation/backfill cycles. The reaction was then charged with NaOMe (25
wt% in MeOH, 1.97 L, 8.61 mol, 4.00 equiv) at 23 °C using MeOH (250 mL) to complete the
transfer, after which formamidine acetate (246.6 g, 2.37 mol, 1.10 equiv) was added in one
portion. The resulting clear, homogeneous mixture was stirred for 20 min. A solution of racꢀ30
(500.0 g, 2.15 mol, 1.00 equiv) in MeOH (500 mL) was then added over 5 min, using MeOH
(200 mL) to complete the transfer. The intense yellow reaction was stirred at 25 ˚C for 18 h. At
1
this time, the reaction was determined to be complete by inꢀprocess analysis using H NMR
spectroscopy (<2% racꢀ30) and was subsequently cooled to ~2 °C (ice bath), after which HCl (g)
was bubbled through the mixture for approximately 1 h at a temperature not exceeding 10 °C. At
this time pH 2.5 (litmus) had been achieved, and the reaction was warmed to 25 ˚C. The reaction
21
ACS Paragon Plus Environment

Organic Process Research & Development
Page 22 of 38
1
2
3
4
5
6
7
8
9
mixture was concentrated via distillation over a period of 7 h (35 °C at 200 Torr) until ~3.3 L of
distillate had been collected. The resulting viscous tan slurry was suspended in toluene (3.75 L)
and agitated for 30 min at 23 °C, after which the product was collected using a tableꢀtop filter
device equipped with polypropylene cloth. Toluene (2 × 500 mL) was used to wash and transfer
material, and the filter cake was dried under vacuum overnight and subsequently homogenized
using a spatula. DHP racꢀ31 (906.0 g, 43.8 w/w% via HPLC wt% assay, 87% yield) was isolated
as a beige solid contaminated with inorganic salts (NaCl). Analysis was conducted on a 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
1
of DHP racꢀ31 purified through recrystallization in water. H NMR (400 MHz, DMSOꢀd6) 7.87
(s, 1H), 3.52 (s, 3H), 3.36 (m, 1H), 2.62 (ABX, JAB = 15.0 Hz, 2H), 1.09 (d, J = 7.0 Hz, 3H); 13C
NMR (100 MHz, DMSOꢀd6) 172.8, 163.8, 147.1, 105.1, 51.1, 37.8, 25.5, 17.6; HRMS (ESI+)
calculated for C9H12N2O4Na [M + Na]+ (m/z) 235.0688; found (m/z) 235.0695.
Methyl 3ꢀ(4,6ꢀdichloropyrimidinꢀ5ꢀyl)butanoate, racꢀ32. A 3 L 4ꢀneck roundꢀbottom flask
was charged with crude DHP racꢀ31 (220.6 g, 44.3 w/w%, 471 mmol, 1.0 equiv) followed by
CH3CN (600 mL), resulting in a viscous slurry, and the reaction headspace was purged with
nitrogen using two evacuation/backfill cycles. The slurry was agitated for ~10 min, after which
2,6ꢀlutidine (54.9 mL, 471 mmol, 1.00 equiv) was added in one portion at 23 °C, leading to an
immediate reduction in viscosity. The reaction was then heated to 70 °C, and POCl3 (97.1 mL,
1.06 mol, 2.25 equiv) was added over 12 min. The reaction was further heated to 79 °C over 1 h
and was agitated for an additional 18 h, after which the reaction was determined to be complete
1
by inꢀprocess analysis using H NMR spectroscopy (<2% racꢀ31, DMSOꢀd6) and was cooled to
25 ˚C over 3 h. At this time, 6N NaOH (78.0 mL, 428 mmol, 1.1 equiv) was added to the mixture
at a rate of 1ꢀ2 mL/min such that the temperature did not exceed 35 °C. The mixture was
concentrated via distillation (30 °C at 85ꢀ100 Torr) until ~475 mL of distillate had been collected
22
ACS Paragon Plus Environment

Page 23 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
and the rate of distillation had decreased significantly (1 drop/10 s). The concentration process
required ~1.5 h and afforded a viscous solution, which continued to stir efficiently. The reaction
was subsequently diluted with iPrOAc (1.0 L) followed by H2O (750 mL), resulting in the
dissolution of all solids. The phases were separated, and the organic phase was washed 0.5 M aq.
NaHCO3 (250 mL) followed by H2O (100 mL). The organic phase was concentrated via rotovap
affording dichloropyrimidine racꢀ32 as a light orange oil, which was observed to slowly solidify
upon standing. Crude racꢀ32 was employed in the subsequent SNAr step without further
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
purification, and quantitative yield (117.4 g, 471 mmol) was assumed. H NMR (400 MHz,
CDCl3) 8.63 (s, 1H), 4.15 (m, 1H), 3.65 (s, 3H), 3.07 (1/2 ABX, JAB = 16.6 Hz, 1H), 2.86 (1/2
ABX, JAB = 16.6 Hz, 1H), 1.45 (d, J = 7.2 Hz, 3H); 13C NMR (100 MHz, CDCl3): δ 172.0,
155.7 (2), 135.1, 52.0, 37.7, 31.8, 17.2.
Tertꢀbutyl
4ꢀ(6ꢀchloroꢀ5ꢀ(4ꢀmethoxyꢀ4ꢀoxobutanꢀ2ꢀyl)pyrimidinꢀ4ꢀyl)piperazineꢀ1ꢀ
A 3 L 4ꢀneck roundꢀbottom flask was charged with crude
carboxylate, racꢀ33
.
dichloropyrimidine racꢀ32 (117.4 g, 471 mmol, 1.00 equiv) in MeOH (900 mL), followed by Nꢀ
Bocꢀpiperazine (96.5 g, 518 mmol, 1.10 equiv) and DIEA (90.3 mL, 518 mmol, 1.10 equiv), and
the reaction headspace was subsequently purged with nitrogen using two evacuation/backfill
cycles. The homogeneous orange mixture was then heated to 60 °C over 45 min and was agitated
for an additional 12 h. At this time, the reaction was determined to be complete by inꢀprocess
1
analysis using H NMR spectroscopy (<2% racꢀ32) and was cooled to 30 °C over ~1 h. The
mixture was concentrated via distillation (~25–30 °C at 80–100 Torr) until the rate of distillation
1
had decreased significantly (1 drop/10 s), and H NMR analysis indicated ~5–6 equiv. MeOH
remaining in the reaction. The concentration process required ~2 h and afforded a viscous slurry,
which was subsequently diluted with EtOAc (1.2 L) followed by 3M aq. NH4Cl (250 mL),
23
ACS Paragon Plus Environment

Organic Process Research & Development
Page 24 of 38
1
2
3
4
5
6
7
8
9
resulting in the dissolution of all solids after vigorous stirring. The phases were separated, and
the organic phase was washed with 3M NH4Cl (250 mL). The organic phase was then
concentrated via rotovap affording SNAr product racꢀ33 as a light orange oil, which was
observed to slowly solidify upon standing. Crude racꢀ33 was employed in the subsequent
hydrolysis step without further purification, and quantitative yield (188.0 g, 471 mmol) was
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
assumed. H NMR (400 MHz, CDCl3) 8.42 (s, 1H), 3.70ꢀ3.55 (m, 5H), 3.66 (s, 3H), 3.28 (m,
4H), 3.04 (1/2 ABX, JAB = 16.2 Hz, 1H), 2.71 (1/2 ABX, JAB = 16.2 Hz, 1H), 1.48 (s, 9H), 1.45
13
(d, J = 7.2 Hz, 3H); C NMR (100 MHz, CDCl3): δ 172.7, 168.4, 160.0, 155.2, 154.9, 125.1,
80.3, 52.0, 50.4, 44.0, 38.7, 29.7, 28.6, 17.9; HRMS (ESI+) calculated for C18H27ClN4O4Na [M +
Na]+ (m/z) 421.1619; found (m/z) 421.1614.
3ꢀ(4ꢀ(4ꢀ(Tertꢀbutoxycarbonyl)piperazinꢀ1ꢀyl)ꢀ6ꢀchloropyrimidinꢀ5ꢀyl)butanoic acid, racꢀ
25. A 3 L 4ꢀneck roundꢀbottom flask was charged with crude methyl ester racꢀ33 (188.0 g, 471
mmol, 1.00 equiv) in THF (750 mL), followed by H2O (250 mL) and LiOHꢀH2O (59.3 g, 1.41
mol, 3 equiv). The slightly cloudy, light yellow reaction was stirred for 12 h, after which H2O
(50 mL) was added, and the reaction was agitated for an additional 24 h. At this time, the
1
reaction was determined to be complete by inꢀprocess analysis using H NMR spectroscopy
(<2% racꢀ33), after which the mixture was concentrated via distillation (~30 °C at 185 Torr)
until the rate of distillation had decreased significantly (1 drop/10 s). The concentration process
required ~3 h and afforded a cloudy solution that was subsequently diluted with H2O (700 mL).
The mixture was then cooled to ~2 °C (ice bath), after which 2N aq. HCl (700 mL, 1.4 mol, 3
equiv) was added to the mixture over 3 h. The resulting white slurry (pH 3.5 by litmus) was
subsequently warmed to 25 ˚C, after which the product was collected using a Buchner funnel.
Additional H2O (2 × 250 mL) was used to wash and transfer material, and the filter cake was
24
ACS Paragon Plus Environment

Page 25 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
dried (0.1 Torr, 45 °C) for 60 h during which time the material was homogenized several times
using a spatula. Carboxylic acid racꢀ25 (146.3 g, 95.0 w/w% via HPLC wt% assay, 77% yield
from DHP racꢀ31) was isolated as a white solid. 1H NMR (400 MHz, CDCl3) 8.44 (s, 1H), 3.72ꢀ
3.51 (m, 5H), 3.28 (m, 4H), 3.11 (1/2 ABX, JAB = 16.6 Hz, 1H), 2.78 (1/2 ABX, JAB = 16.6 Hz,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13
1H), 1.49 (s, 12H); C NMR (100 MHz, CDCl3) 176.6, 168.4, 159.9, 155.1, 155.0, 125.1, 80.6,
50.3, 44.1, 38.7, 29.6, 28.6, 17.9; HRMS (ESI+) calculated for C17H25ClN4O4Na [M + Na]+ (m/z)
407.1462; found (m/z) 407.1455.
Tertꢀbutyl
4ꢀ(5ꢀ(4ꢀ(benzyloxy)ꢀ4ꢀoxobutanꢀ2ꢀyl)ꢀ6ꢀchloropyrimidinꢀ4ꢀyl)piperazineꢀ1ꢀ
carboxylate, racꢀ26. To a solution of carboxylic acid racꢀ25 (100.0 g, 246.8 mmol, 1.00 equiv)
in DMF (600 mL) at 23 °C under a nitrogen atmosphere was added Cs2CO3 (84.45 g, 259.2
mmol, 1.05 equiv) followed by benzyl bromide (30.1 mL, 253 mmol, 1.02 equiv). The reaction
was agitated for 18 h at 23 °C and was determined to be complete by inꢀprocess analysis using
1H NMR spectroscopy (<2% racꢀ25). The reaction mixture was then filtered through a plug of
celite (washed with EtOAc (200 mL)) to afford a tan solution. The filtrate was subsequently
concentrated via rotovap to afford a viscous tan residue containing some precipitated salts.
[CAUTION: Benzyl bromide is a strong lachrymator and is very irritating to the skin and
mucous membranes (see MSDS)]. The residue was partitioned between EtOAc (600 mL) and
water (500 mL), after which the aq. layer was extracted with EtOAc (100 mL), and the combined
organic layers were washed with H2O (3 × 100 mL). The combined organic layers were dried
(MgSO4), filtered, and concentrated via rotovap to afford a tan oil, which was further dried under
vacuum (0.1 Torr) for 18 h to give benzyl ester racꢀ26 (119.8 g, 246.8 mmol, quant). The
1
product was observed to slowly solidify upon standing. H NMR (400 MHz, CDCl3) 8.40 (s,
1H), 7.43ꢀ7.24 (m, 5H), 5.09 (s, 2H), 3.66 (m, 1H), 3.54 (m, 4H), 3.22 (m, 4H), 3.06 (1/2 ABX,
25
ACS Paragon Plus Environment

Organic Process Research & Development
Page 26 of 38
1
2
3
4
5
6
7
8
9
JAB = 16.0 Hz, 1H), 2.78 (1/2 ABX, JAB = 16.0 Hz, 1H), 1.49 (s, 9H), 1.45 (d, J = 7.2 Hz, 3H);
13C NMR (100 MHz, CDCl3) 172.0, 168.4, 160.1, 155.2, 154.9, 135.8, 128.8, 128.6, 128.5,
125.1, 80.3, 66.8, 50.4, 44.0, 39.0, 29.8, 28.6, 18.0; HRMS (ESI+) calculated for
C24H31ClN4O4Na [M + Na]+ (m/z) 497.1931; found (m/z) 497.1919.
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
Isopropyl
5ꢀ(4ꢀ(benzyloxy)ꢀ4ꢀoxobutanꢀ2ꢀyl)ꢀ6ꢀ(4ꢀ(tertꢀbutoxycarbonyl)piperazinꢀ1ꢀ
yl)pyrimidineꢀ4ꢀcarboxylate, racꢀ34. A 2 L 3ꢀneck roundꢀbottom flask was charged with
benzyl ester racꢀ26 (50.0 g, 105 mmol, 1.00 equiv), Pd(OAc)2 (2.36 g, 10.5 mmol, 0.10 equiv),
dppp (4.78 g, 11.6 mmol, 0.11 equiv) and K2CO3 (325 mesh, 8.00 g, 57.9 mmol, 0.55 equiv).
The flask was purged with nitrogen using three evacuation/backfill cycles and was then charged
with (3:1) iPrOHꢀTHF (600 mL) that had been previously degassed by subꢀsurface sparging with
nitrogen for 1 h. The reaction was subsequently stirred for 20 min at 23 °C, which facilitated the
dissolution of all solids, except for the K2CO3, resulting in a heterogeneous orange solution. The
flask was then equipped with a CO source consisting of a doubleꢀwall thickness balloon (balloon
inside of a balloon, Betallatex™, 11 in. inflated, 0.35 psi CO measured), and the reaction was
purged with CO using three evacuation/backfill cycles. [CAUTION: Carbon monoxide is very
toxic if inhaled, and a CO detector must be used when handling this gas. It also has a very low
flash point (–191 °C) and is considered highly flammable! Refer to the MSDS]. The resulting
mixture was stirred at 25 ˚C for 1 h, gradually becoming red in color, after which the reaction
was heated to 55 °C and agitated for an additional 40 h. At this time, the reaction was determined
1
to be complete by inꢀprocess analysis using H NMR spectroscopy (<2% racꢀ26), and the
temperature was reduced to 45 °C. In a sequence to destroy the catalyst, the CO source was
initially removed, and hexanes (600 mL) was charged to the reaction. The resulting mixture was
stirred under air atmosphere for 7 h, gradually becoming black in color, after which the reaction
26
ACS Paragon Plus Environment

Page 27 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
was stirred while sparging with air for an additional 14 h. At this time, the reaction was cooled to
25 ˚C, and additional hexanes (1.2 L) was charged in order to compensate for solvent lost due to
evaporation during the sparging process. The reaction was then filtered through a purification
plug constructed in a 4 in. diameter fritted glass funnel using (1:1) hexanesꢀEtOAc (~50 mL) to
wash and transfer material from the flask. The plug employed an initial layer of SiO2 (4.0 in.,
preꢀeluted with hexanes), which was followed by a layer of sand (0.5 in.) and a layer of celite
(1.0 in.). Following the initial charge of solution, the plug was subsequently washed with (1:1)
hexanesꢀEtOAc (4 × 400 mL). The dark yellow filtrate was partially concentrated to ~ 300 mL
total volume, after which the solution was filtered using a Buchner funnel equipped with
Whatman® GF/F paper in order to remove the remaining fine suspension of Pdꢀblack. The
filtrate was concentrated via rotovap, affording a crude beige solid (45.3 g, 82% crude 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
A 1 L roundꢀbottom flask was charged with a portion of the crude solid (35.0 g) followed by
(15:1) heptane/MTBE (320 mL) at 23 °C under air atmosphere. The suspension was agitated
vigorously for 16 h during which time some of the larger product chunks were necessarily
homogenized using a spatula. Upon completion of the trituration, the product was collected via
Buchner funnel using heptane (150 mL) to wash and transfer material. The product was dried
under vacuum (0.1 Torr) for 45 min to give mixed diester racꢀ34 as a gray solid (30.3 g, 57.5
mmol, 71% yield). 1H NMR (400 MHz, CDCl3) 8.69 (s, 1H), 7.39ꢀ7.27 (m, 5H), 5.31 (sp, J = 6.2
Hz, 1H), 5.10 (s, 2H), 3.63 (m, 1H), 3.54 (m, 4H), 3.22 (m, 4H), 2.91 (1/2 ABX, JAB = 16.4 Hz,
1H), 2.74 (1/2 ABX, JAB = 16.2 Hz, 1H), 1.50 (s, 9H), 1.43 (d, J = 6.2 Hz, 3H), 1.41 (d, J = 6.4
Hz, 3H), 1.34 (d, J = 7.0 Hz, 3H); 13C NMR (100 MHz, CDCl3) 172.0, 167.8, 166.7, 158.2,
155.7, 154.9, 135.9, 128.8, 128.5, 128.4, 125.4, 80.2, 70.7, 66.6, 50.4, 44.0, 40.8, 29.1, 28.6, 21.8
27
ACS Paragon Plus Environment

Organic Process Research & Development
Page 28 of 38
1
2
(2), 19.8; HRMS (ESI+) calculated for C28H38N4O6Na [M + Na]+ (m/z) 549.2689; found (m/z)
549.2690.
3
4
5
6
7
8
9
Tertꢀbutyl 4ꢀ(5ꢀmethylꢀ7ꢀoxoꢀ6,7ꢀdihydroꢀ5Hꢀcyclopenta[d]pyrimidinꢀ4ꢀyl)piperazineꢀ1ꢀ
carboxylate, racꢀ3. A 1 L roundꢀbottom flask was charged with racꢀ34 (25.0 g, 47.5 mmol, 1.00
equiv), purged with nitrogen using three evacuation/backfill cycles, and then charged with THF
(250 mL). The mixture was stirred at 25 ˚C until all solids were dissolved, and the solution was
cooled to ~2 ºC (ice bath), after which KOtꢀBu (1M in THF, 52.2 mL, 52.2 mmol, 1.10 equiv)
was added over 5 min, affording a yellow mixture. The reaction was stirred 5 min at ~2 ºC, and
was then subsequently warmed to 25 ˚C and stirred for an additional 10 min. At this time,
conversion to ketoꢀester intermediate racꢀ28 was determined to be complete by inꢀprocess
analysis using HPLC (<1 A% racꢀ34).37 The pH of the reaction was then adjusted via the
addition of 3M aq. HCl solution (16.9 mL, 50.6 mmol, 1.06 equiv), resulting in pH 5.5 (litmus).
The reaction was then charged with Pd/alumina (5 wt.%, powder, Aldrich) (1.01 g, 0.475 mmol,
0.01 equiv), after which the mixture was purged with nitrogen using three evacuation/backfill
cycles. The reaction flask was then equipped with a hydrogen (H2) source consisting of a doubleꢀ
wall thickness balloon (balloon inside of a balloon, Betallatex™,11 in. inflated, 0.35 psi H2
measured). The reaction was then purged with H2 using three evacuation/backfill cycles and was
agitated vigorously for 3 h at 25 ˚C. At this time, the decarboxylation was determined to be
complete by inꢀprocess analysis using HPLC (<1 A% racꢀ28) and the reaction was subsequently
sparged with nitrogen (subꢀsurface) for 10 min in order to eliminate the hydrogen gas present.
The mixture was subsequently filtered through a plug of celite (dry, tightly packed) to afford a
dark yellow filtrate. Following the initial charge of solution, the plug was washed with THF (300
mL total). The filtrate was concentrated via rotovap, after which the residue was reꢀdissolved in
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
28
ACS Paragon Plus Environment

Page 29 of 38
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
THF (75 mL) at 25 ˚C, and heptane (225 mL) was charged to the solution over 15 min resulting
in precipitation of the product. The slurry was agitated for 18 h, after which the product was
isolated via filtration. The product was washed with heptane (100 mL) and allowed to airꢀdry on
the filter. Ketone racꢀ3 was isolated as a tan solid (14.4 g, 43.2 mmol, 91% yield), with >99 A%
purity via HPLC analysis. 1H NMR (400 MHz, CDCl3) 8.73 (s, 1H), 3.88 (m, 2H), 3.72ꢀ3.62 (m,
5H), 3.53 (m, 2H), 2.96 (1/2 ABX, JAB = 19.6 Hz, 1H), 2.34 (1/2 ABX, JAB = 19.6 Hz, 1H), 1.50
(s, 9H), 1.32 (d, J = 6.8 Hz, 3H); 13C NMR (100 MHz, CDCl3) 205.3, 161.4, 158.4, 157.5, 153.9,
136.9, 79.2, 45.3, 42.9, 30.7, 28.0, 19.9; HRMS (ESI+) calculated for C17H24N4O3Na [M + Na]+
(m/z) 355.1746; found (m/z) 355.1752.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ASSOCIATED CONTENT
Supporting Information
Additional experimental procedures for the scaleꢀup of the (R)ꢀ(+)ꢀpulegone route to ketone (R)ꢀ
3 are provided. This material is available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Authors
*Eꢀmail: jlane@arraybiopharma.com
*Eꢀmail: travisr@gene.com
Present Addresses
§OPX Biotechnologies Inc., 2425 55th Street, Suite 100, Boulder, Colorado 80301, United States.
29
ACS Paragon Plus Environment