Fit-for-purpose synthesis of dual leucine zipper kinase (DLK) inhibitor GNE-834

Article abstract of DOI:10.1016/j.tetlet.2020.152430

A practical fit-for-purpose synthesis of dual leucine zipper kinase (DLK) inhibitor GNE-834 (1) was developed. The key C[sbnd]C bond was constructed via a Suzuki–Miyaura cross-coupling of iodopyrazole 2 and pyridine boronic ester 3 to afford ketone 12. Su

Full text of DOI:10.1016/j.tetlet.2020.152430

Tetrahedron Letters xxx (xxxx) xxx
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Fit-for-purpose synthesis of dual leucine zipper kinase (DLK) inhibitor
GNE-834
a
a
b
b
b
Jie Xu a, , Diane E. Carrera , Danial Beaudry , Michael Siu , Malcolm P. Huestis , Wendy Liu ,
⇑
Rémy Angelaud a, Francis Gosselin a
a Department of Small Molecule Process Chemistry, Genentech Inc., 1 DNA Way, South San Francisco, CA 94080, United States
b Department of Discovery Chemistry, Genentech Inc., 1 DNA Way, South San Francisco, CA 94080, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
A practical fit-for-purpose synthesis of dual leucine zipper kinase (DLK) inhibitor GNE-834 (1) was devel-
oped. The key CAC bond was constructed via a Suzuki–Miyaura cross-coupling of iodopyrazole 2 and pyr-
idine boronic ester 3 to afford ketone 12. Subsequent selective reductive amination of ketone 12 with
morpholine followed by a resolution via tosylate salt formation provided GNE-834 (1) in 39% overall yield
from iodide 2 with 98.9 A% HPLC purity.
Received 31 July 2020
Revised 25 August 2020
Accepted 30 August 2020
Available online xxxx
Ó 2020 Elsevier Ltd. All rights reserved.
Keywords:
Suzuki coupling
Reductive amination
Resolution
DLK inhibitor
ALS
Introduction
diastereoselective reductive amination; and the iodopyrazole 2
could be derived from further functionalization of ketone 4
Amyotrophic lateral sclerosis (ALS), also known as Lou Gehrig’s
Disease, is a rare but devastating progressive neurodegenerative
disease which results in the wasting away of muscle and loss of
movement [1]. There are an estimated 15,000 people in the United
States living with ALS [2]. Currently, there is no known effective
cure for ALS and the average survival from onset to death is two
to four years with deteriorating quality of life [3]. The cause of
ALS is currently unknown [2]. In brains of patients with ALS, an ele-
vated dual leucine zipper kinase (DLK) pathway activity was
observed and DLK was reported to regulate c-Jun N-terminal
kinases (JNK) which induces brain-cell death. It was hypothesized
that DLK inhibition could be a potential approach to treat neuronal
degeneration disease such as ALS [4].
GNE-834 (1) is a small molecule DLK inhibitor that was discov-
ered and developed for the potential treatment of ALS (Fig. 1) [5].
We describe herein a fit-for-purpose synthesis of GNE-834 (1)
which enabled the fast delivery of materials for pre-clinical studies.
Retrosynthetically, we envisioned that a Suzuki–Miyaura cross-
coupling [6] between iodopyrazole 2 and pyridine boronic ester 3
could construct the key CAC bond between the pyrazole and pyr-
idine moieties; the morpholine group could be installed via a
(Scheme 1).
Results and discussions
We commenced our synthesis with protection of 3-cyclopen-
tenol 5 with tert-butyldiphenylchlorosilane (TBDPSCl) followed
by Rh2(OAc)4 catalyzed cyclopropanation to give ester 6 [7]. Con-
version of ester 6 to ketone 4 proceeded smoothly via Weinreb
amide [8] formation followed by MeMgBr addition in 59% overall
yield from 3-cyclopentenol 5 over 5 steps. Claisen condensation
between ketone 4 and oxalic acid ethyl ester followed by pyrazole
formation with 1-isopropylhydrazine delivered the desired bicy-
clopyrazole 7 in 67% yield over 2 steps. Saponification of ester 7
and Curtius rearrangement [9] of the resulting acid gave Cbz pro-
tected pyrazole amine 8 in 70% yield over 2 steps. Cbz deprotection
and Sandmeyer [10] reaction then afforded the desired pyrazole
iodide 9 in 47% yield. Final TBDPS deprotection and Dess–Martin
periodinane oxidation [11] furnished the iodoketone as a mixture
of 2:1 exo:endo isomers [12]. Separation of the undesired endo-iso-
mer from the desired exo-isomer 2 by silica gel column chromatog-
raphy provided 2 in 60% yield (Scheme 2).
From
commercially
available
2-amino-3-trifluo-
⇑
romethylpyridine 10, an N-bromosuccinimide (NBS) bromination
afforded the desired bromopyridine 11 in 99% yield.
Corresponding author.
E-mail address: xu.jie@gene.com (J. Xu).
https://doi.org/10.1016/j.tetlet.2020.152430
0040-4039/Ó 2020 Elsevier Ltd. All rights reserved.
Please cite this article as: J. Xu, D. E. Carrera, D. Beaudry et al., Fit-for-purpose synthesis of dual leucine zipper kinase (DLK) inhibitor GNE-834, Tetrahedron
Letters, https://doi.org/10.1016/j.tetlet.2020.152430

J. Xu et al.
Tetrahedron Letters xxx (xxxx) xxx
when a strong base such as NaOt-Bu was employed in the
SuzukiÀMiyaura coupling in the presence of 5 mol% Pd(dppf)Cl2 in
dioxane/H2O (5:1) at 90 °C, the reaction gave complete conversion
but with a significant level (43 A% by HPLC analysis) of aldol con-
densation side product (Table 1, entry 1). Phosphate base resulted
in a slow reaction (Table 1, entry 2). K2CO3 afforded 93% conversion
and 70 A% HPLC of the desired product 12 (Table 1, entry 3).
Besides dppf, other ligands including A-caPhos, XPhos, t-BuXPhos,
and P(t-Bu)3 were also studied [14] (Table 1, entries 4–7). How-
ever, none of them gave better conversion and reaction profile than
Pd(dppf)Cl2. Since 1,4-dioxane is an environmentally undesirable
solvent, other solvents were explored. Compared to 1,4-dioxane
(Table 1, entry 10), THF and ethanol gave lower conversion and
inferior assay yield (Table 1, entries 8 and 11). DMF led to complete
conversion (Table 1, entry 9), but similar to NaOt-Bu, large
amounts of aldol condensation side product resulted in lower assay
yield. Using i-PrOH as reaction solvent minimized the formation of
the aldol condensation side product and afforded 100% conversion
in high assay yield of 88% (Table 1, entry 12). By increasing the
reaction temperature from 70 °C to 85 °C, the reaction time was
also reduced to 3 h from 24 h with better reaction profile (Table 1,
entry 13). The reaction also tolerated lower catalyst loading (1 mol
%) (Table 1, entry 14) and lower boronic ester 3 (110 mol%) (Table 1,
entry 15). Thus, the optimized reaction conditions employed 1 mol
% Pd(dppf)Cl2, and 110 mol% boronic ester 3 in i-PrOH/H2O at 85 °C
for 3 h, under which the reaction afforded complete conversion and
96% assay yield. Isolation of ketone 12 by direct crystallization
from i-PrOH/H2O mixture was achieved by adding H2O (10 mL/g)
directly into the reaction mixture, generating the coupling product
in 93% isolated yield and 99.6 A% purity by HPLC.
Fig. 1. GNE-834 (1), a DLK inhibitor.
We then focused our effort on the formation of GNE-834 (1) by
a reductive amination approach. We started our optimization of
the reductive amination of 12 and morpholine employing 10 equiv
of AcOH and a variety of reducing agents in toluene (Table 2). Gen-
erally, borohydride and borane reagents afforded better conversion
than other reducing agents such as LiAlH(Ot-Bu)3 (Table 2, entries
1–5). The reaction performed the best in terms of diastereoselec-
tivity when NaBH3CN was employed as the reducing agent (Table 2,
entry 6). Other aromatic solvents performed similarly to toluene
(Table 2, entries 7–9). Lewis Acids such as Ti(Oi-Pr)4 [15] and
ZrCp2Cl2 were also studied (Table 2, entries 10–11). Similar to
Scheme 1. Retrosynthetic analysis of GNE-834 (1).
Bromopyridine 11 was then converted to the corresponding boro-
nic ester with bispinacolatodiboron (B2pin2) in 72% yield
(Scheme 3) [13].
With both starting materials in hand, we started to examine the
SuzukiÀMiyaura coupling of iodopyrazole 2 and aminopyridine
boronic ester 3. The major challenge we needed to address for this
transformation was to suppress the impurity resulting from aldol
self-condensation of iodoketone 2 under basic conditions. Indeed,
3
Scheme 2. Synthesis of iodoketone 2.
2

J. Xu et al.
Tetrahedron Letters xxx (xxxx) xxx
Scheme 3. Synthesis of Boronic Ester 3.
Table 1
Suzuki reaction optimization.
entrya
base
catalyst
solvent
time
temp (°C)
conv (%)b
A% of 12b
assay yield (%)b
1
2
3
4
5
6
7
8
NaOt-Bu
K3PO4
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
Pd(dppf)Cl2
Pd(dppf)Cl2
Pd(dppf)Cl2
XPhos Pd G3
(A-caPhos)2 PdCl2
t-BuXPhos Pd G1
P(t-Bu)3Pd G2
Pd(dppf)Cl2
Pd(dppf)Cl2
Pd(dppf)Cl2
Pd(dppf)Cl2
Pd(dppf)Cl2
Pd(dppf)Cl2
dioxane/H2O
dioxane/H2O
dioxane/H2O
dioxane/H2O
dioxane/H2O
dioxane/H2O
dioxane/H2O
THF/H2O
DMF/H2O
dioxane/H2O
EtOH/H2O
i-PrOH/H2O
i-PrOH/H2O
i-PrOH/H2O
i-PrOH/H2O
3 h
90
90
90
90
90
90
90
70
70
70
70
70
85
85
85
100
83
93
63
78
86
65
79
100
89
57
57
70
38
56
50
60
67
70
76
57
82
94
94
96
–
–
–
–
–
–
–
77
61
85
56
88
89
88
93e
24 h
24 h
24 h
24 h
24 h
24 h
24 h
24 h
24 h
24 h
24 h
3 h
9
10
11
12
13
14c
15d
86
100
100
100
100
c
Pd(dppf)Cl2
Pd(dppf)Cl2
3 h
3 h
a
Reaction conditions: 2 (100 mg, 0.303 mmol), 3 (105 mg, 0.363 mmol, 120 mol%), base (200 mol%), and catalyst (5 mol%) in different organic solvent (15 mL/g) and H2O
(3 mL/g).
b
Determined by quantitative HPLC analysis of the crude reaction mixture
cPd(dppf)Cl2 (2.2 mg, 0.0030 mmol, 1 mol%)
Reaction condition: 2 (466 g, 1.41 mol), 3 (447 g, 1.55 mol, 110 mol%), K2CO3 (390 g, 2.82 mol, 200 mol%), and Pd(dppf)Cl2 (10.3 g, 0.0141 mol, 1 mol%) in i-PrOH (10 mL/g)
c
d
and H2O (2 mL/g)
e
Isolated yield
AcOH, both gave complete conversion and similar diastereoselec-
tivity. Since both diastereomers 1 and 13 have low solubility in
toluene, direct precipitation was achieved by adding water and
aqueous Na2CO3 solution into the reaction mixture to provide 1
and 13 (dr = 67:33) mixture in > 99% yield. Further purification
by silica gel column chromatography produced 1 in 58% yield.
Next, we set out to investigate the separation of the diastere-
omers by crystallization to eliminate chromatographic purification.
We first examined the solubility of the diastereomers 1 and 13 at
25ꢀC in a number of organic solvents. The initial study indicated
that both diastereomers have similarly low solubility in common
organic solvents (<30 mg/mL). Therefore, the separation of the
diastereomers by direct crystallization was deemed impractical
[16]. We then prepared the corresponding tosylate salt of both
diastereomers 1 and 13. The solubility of the tosylate salts was
examined in multiple solvents and we were delighted to see that
by using MeOH as the solvent, a significant solubility difference
(40–50 mg/mL) between the tosylate salts could be achieved at
elevated temperatures (Fig. 2). Generally, the solubility of 1ꢀTsOH
is low and relatively temperature independent. However, the solu-
bility of 13ꢀTsOH is highly dependent on temperature. This led us
to envision a crystallization by tosylate salt formation in MeOH at
high temperature so that 1ꢀTsOH could be obtained and the unde-
sired 13ꢀTsOH purged in the mother liquor after hot filtration.
Different temperatures were then evaluated for the resolution
by tosylate salt formation (Table 3). Starting from 67:33 mixture
3

J. Xu et al.
Tetrahedron Letters xxx (xxxx) xxx
Table 2
Reductive amination condition optimization.
entrya
reducing agent
acid
solvent
1 : 13b
conv (%)b
A% of 1b
1
2
3
4
5
6
7
8
NaBH4
AcOH
AcOH
AcOH
AcOH
AcOH
AcOH
AcOH
AcOH
AcOH
Ti(Oi-Pr)4
toluene
toluene
toluene
toluene
toluene
toluene
xylene
mesitylene
CF3C6H5
toluene
toluene
53:47
48:52
44:56
38:62
–
67:33
58:42
62:38
55:45
66:34
65:35
92
48
48
44
30
–
67
58
62
55
66
65
2-Picolineꢀborane
5-Ethyl-2-picolineꢀborane
NaBH(OAc)3
LiAlH(Ot-Bu)3
NaBH3CN
NaBH3CN
NaBH3CN
NaBH3CN
NaBH3CN
100
100
79
<5
100
100
100
100
100
100
9
10
11
c
NaBH3CN
ZrCp2Cl2
a
Reaction conditions: 12 (50.0 mg, 0.137 mmol), reducing agent (0.206 mmol, 150 mol%), morpholine (0.119 mL, 1.37 mmol, 1000 mol%), and acid (1.37 mmol, 1000 mol%)
in different organic solvent (15 mL/g) at 80 °C for 3 h.
b
Determined by HPLC analysis of the crude reaction mixture.
200 mol%.
c
Table 3
Crystallization of tosylate salt at different temperatures.
Fig. 2. Solubility of 1ꢀTsOH and 13ꢀTsOH in MeOH.
of 1 and 13 freebase, we were able to increase the dr of the isolated
tosylate salt to 94:6 at 50 °C (Table 3, entry 1). Additional methanol
slurry at 50 °C was able to increase the dr to 99:1. However, hot fil-
tration at higher temperatures resulted in marginally lower overall
yield (Table 3, entries 2 and 3). Conducting resolution and hot fil-
tration at lower temperature 45 °C afforded a slightly lower
diastereoselectivity (Table 3, entry 4). To balance the recovery
and purging of 13, 50 °C was pursued further. This process was first
scaled up to 3 g (Table 3, entry 5), consistent dr (99.4:0.6) and
recovery (53%) was obtained. When the reaction was scaled up
to 59 g, a lower dr (94:6) was observed after the tosylate salt for-
mation. The subsequent MeOH slurry did not upgrade the purity as
effectively as in the small scale (Table 3, entry 5), presumably due
to the presence of trapped freebase diastereomer 13 due to insuf-
ficient charge of TsOHꢀH2O. An additional MeOH slurry of crude
1ꢀTsOH in the presence of 5 mol% TsOHꢀH2O was thus performed
to further improve the diastereoselectivity and lead to comparable
dr (98.9:1.1) and yield (50%) (Table 3, entry 6).
entrya
temp (°C)
drb of crude 1ꢀTsOH
drb of 1ꢀTsOH
yield (%)c
1
2
3
50
55
60
45
50
50
94.0:6.0
96.5:3.5
93.6:6.4
92.6:7.4
97.6:2.4
94.0:6.0
99.2:0.8
99.4:0.6
99.2:0.8
98.8:1.2
99.4:0.6
98.9:1.1
55
53
52
56
53
50
4
5d
6e
a
Reactions conditions: 1 and 13 mixture (300 mg, 0.689 mmol), with TsOHÁH2O
(131 mg, 0.689 mmol, 100 mol%) in MeOH (5 mL/g) at different temperature.
b
Determined by HPLC analysis of the isolated tosylate salt.
Isolated yield from 1 and 13 mixture (dr = 67:33).
2.7 g scale.
59.4 g scale, one additional slurry with 5 mol% TsOHꢀH2O in MeOH was used.
c
d
e
4

J. Xu et al.
Tetrahedron Letters xxx (xxxx) xxx
By treating 1ꢀTsOH with a biphasic mixture of K2CO3 aqueous
Appendix A. Supplementary data
solution and EtOAc, the freebase was effectively liberated into
EtOAc layer. Subsequent azeotropic distillation of the EtOAc layer
followed by crystallization afforded 1. This process was carried
out at 25 g scale to give GNE-834 (1) in 87% yield and 99.6 A% HPLC
purity.
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.tetlet.2020.152430.
References
[1] A. Chiò, G. Logroscino, B.J. Traynor, J. Collins, J.C. Simeone, L.A. Goldstein, L.A.
White, Neuroepidemiology 41 (2013) 118–130.
[2] National Institute of Neurological Disorders and Stroke. Amyotrophic Lateral
Sclerosis (ALS) Fact Sheet. https://www.ninds.nih.gov/Disorders/Patient-
Caregiver-Education/Fact-Sheets/Amyotrophic-Lateral-Sclerosis-ALS-Fact-
Sheet (accessed 2020-06-01).
[3] (a) D. Petrov, C. Mansfield, A. Moussy, O. Hermine, Front. Aging Neurosci. 9
(2017) 68–78;
(b) A. Chio, G. Logroscino, O. Hardiman, R. Swingler, D. Mitchell, E. Beghi, B.G.
Traynor, Eurals Consortium Amyotroph. Lateral Sc. 10 (2009) 310–323;
(c) A. Eisen, J. Clin. Neurosci. 16 (2009) 505–512.
[4] (a) M. Siu, A. Sengupta Ghosh, J.W. Lewcock, J. Med. Chem. 61 (2018) 8078–
8087;
(b) C.E. Le Pichon, W.J. Meilandt, S. Dominguez, H. Solanoy, H. Lin, H. Ngu, A.
Gogineni, A. Sengupta Ghosh, Z. Jiang, S.H. Lee Maloney, Sci. Transl. Med. 9
(2017) eaag0394;
(c) S. Patel, W.J. Meilandt, R.I. Erickson, J. Chen, G. Deshmukh, A.A. Estrada, R.N.
Fuji, P. Gibbons, A. Gustafson, S.F. Harri, J. Imperio, W. Liu, X. Liu, Y. Liu, J.P.
Lyssikatos, C. Ma, J. Yin, J.W. Lewcock, M. Siu, J. Med. Chem. 60 (2017) 8083–
8102.
Conclusion
In conclusion, a fit-for-purpose synthesis of DLK inhibitor GNE-
834 (1) was developed. An efficient Suzuki–Miyaura coupling was
developed to construct the key CAC bond between the pyrazole
and the pyridine moieties in high yield. Despite the modest
diastereoselectivity (67:33) observed in the reductive amination,
GNE-834 (1) was obtained in excellent diastereoselectivity (99:1)
by resolution using tosylate salt formation without chromato-
graphic purification.
[5] Siu, M.; Estrada, A.; Liu, W.; Lyssikatos, J. P.; Patel, S.; Liang, G.; Chen, K. U.S.
Patent Application 14/576,890, Dec 19, 2014.
[6] For recent reviews on SuzukiÀMiyaura coupling, see: (a) Miyaura, N.; Suzuki,
A. Chem. Rev. 1995, 95, 2457. (b) Lennox, A. J. J.; Lloyd-Jones, G. C. Chem. Soc.
Rev. 2014, 43, 412.
[7] (a) A.J. Hubert, A.F. Noels, A.J. Anciaux, P. Teyssie, Synthesis 7 (1976) 600–602;
(b) A.J. Anciaux, A.J. Hubert, A.F. Noels, N. Petiniot, P. Teyssie, J. Org. Chem. 45
(1980) 695–702.
[8] S. Nahm, S.M. Weinreb, Tetrahedron Lett. 22 (1981) 3815–3818.
[9] (a) P.A.S. Smith, Org. React. 3 (1946) 337–449;
Declaration of Competing Interest
(b) H. Lebel, O. Leogane, Org. Lett. 8 (2006) 5717–5720;
(c) E.F.V. Scriven, K. Turnbull, Chem. Rev. 88 (1988) 297–368.
[10] Hodgson, H. H. Chem. Rev. 1947, 40, 251À277. (b) Galli, C. Chem. Rev. 1988,
88, 765À792
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.
[11] D.B. Dess, J.C. Martin, J. Am. Chem. Soc. 113 (1991) 7277–7287.
[12] The undesired endo-isomer was carried from compound 6 and purged during
the purification of compound 2.
[13] T. Ishiyama, M. Murata, N. Miyaura, J. Org. Chem. 60 (1995) 7508–7510.
[14] R. Martin, S. Buchwald, Acc. Chem. Res. 41 (2008) 1461–1473.
[15] C. Salmi, Y. Letourneux, J.M. Brunel, Lett. Org. Chem. 3 (2006) 396–401.
[16] Hydrochloride salts of diastereomers 1 and 13 were also prepared to evaluate
their crystallization. Unfortunately, the dr was only improved to 70:30 in
preliminary investigations.
Acknowledgments
The authors would like to thank Tina Nguyen (Genentech, Inc.)
for analytical support and Dr. Haiming Zhang (Genentech, Inc.) for
assisting manuscript preparation.
5