Synthetic approaches to a chiral 4-amino-3-hydroxy piperidine with pharmaceutical relevance

Article abstract of DOI:10.1039/c2ob25411e

Four synthetic strategies were evaluated towards the preparation of (-)-(3R,4R)-1-benzyl-4-(benzylamino)piperidin-3-ol (1), which was constructed with control over the relative and absolute stereochemistry of the 4,3-amino alcohol moiety. The first strategy employed a novel Rh^I catalyzed asymmetric hydrogenation, while two other strategies exploited the existing stereochemistry in 2-deoxy-d-ribose, and the fourth explored both biocatalytic and classical resolution techniques as a means to impart enantioenrichment to racemic intermediates en route to targeted structure (-)-1. The Royal Society of Chemistry 2012.

Full text of DOI:10.1039/c2ob25411e

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Organic &
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 5253
www.rsc.org/obc
PAPER
Synthetic approaches to a chiral 4-amino-3-hydroxy piperidine with
pharmaceutical relevance†
Adrian Ortiz,* Ian S. Young, James R. Sawyer, Yi Hsiao, Amarjit Singh, Masano Sugiyama,
R. Michael Corbett, Melissa Chau, Zhongping Shi and David A. Conlon
Received 24th February 2012, Accepted 10th May 2012
DOI: 10.1039/c2ob25411e
Four synthetic strategies were evaluated towards the preparation of (−)-(3R,4R)-1-benzyl-4-(benzylamino)-
piperidin-3-ol (1), which was constructed with control over the relative and absolute stereochemistry of
the 4,3-amino alcohol moiety. The first strategy employed a novel RhI catalyzed asymmetric
hydrogenation, while two other strategies exploited the existing stereochemistry in 2-deoxy-D-ribose, and
the fourth explored both biocatalytic and classical resolution techniques as a means to impart
enantioenrichment to racemic intermediates en route to targeted structure (−)-1.
Introduction
The piperidine 11 is an advanced key intermediate utilized in the
synthesis of the investigational new drug candidate,
BMS-690514 (2, see Fig. 1), developed for treatment of non-
small cell lung cancer.1 An efficient synthesis of 1 from readily
available starting materials was required in order to furnish
larger quantities of BMS-690514 to support ongoing clinical
studies. The initial in-house route used to prepare 1 took advant-
age of a well-precedented chiral pool based strategy starting
from (R)-pyroglutamic acid (see Fig. 2).2 While this strategy was
successful in preparing 1 to support initial clinical development,
it was lengthy (requiring seven multi-operational steps and three
isolations) and was projected to contribute significantly to the
overall cost of BMS-690514 (2). This complexity led us to
pursue an alternative synthesis of 1. In this article we report on
the development and comparison of four alternative strategies
towards this valuable intermediate.
Fig. 1 Retrosynthetic analysis of drug substance 2, and key piperidine 1.
Results and discussion
Our first strategy sought to employ an asymmetric catalytic
method for the installation of the key 4-amino-3-hydroxy
moiety. Hydrogenation of keto esters is a powerful tool, pro-
viding access to the corresponding hydroxy esters.3 In our
system a trans-selective asymmetric hydrogenation would be
required from commercially available keto ester
4 (see
Scheme 1).4 The hydrogenation of 2-hydroxycyclohex-1-ene-
carboxylic acid, has been reported using Ru-BINAP catalyst
with high diastereoselectivity (95 : 5) and enantioselectivity
(90% ee),5 however, this catalyst was not effective for the hydro-
genation of compound 4, giving both low diastereoselectivity
(55 : 45) and enantioselectivity (70% ee).
Chemical Development, Bristol-Myers Squibb Company, One Squibb
Drive, New Brunswick, New Jersey08903, U.S.A.
After extensive catalyst screening, [RuCl(cymene)(R-C3-
tunephos)]Cl 6 (1 mol%) was identified as a suitable catalyst for
this transformation. Our optimized conditions for this reaction
utilized a hydrogen atmosphere (50 bar, 17 h) in a mixture of
EtOAc–MeOH (9 : 1) to yield the desired hydroxy ester 5 as a
E-mail: Adrian.Ortiz@bms.com; Tel: +732-227-5816
†Electronic supplementary information (ESI) available: Experimental
procedures and characterization data (1H NMR, 13C NMR, HRMS and
IR) and copies of 1H NMR and 13C NMR for all new compounds.
CCDC 878581 for HCl salt of 24. For ESI and crystallographic data in
CIF or other electronic format see DOI: 10.1039/c2ob25411e
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 5253–5257 | 5253

View Online
Fig. 2 Retrosynthetic analysis of 1 from initial pyroglutamic route.
Scheme 2 Synthesis of carbamate derivative 9.
Scheme 1 Preparation of 5 by asymmetric hydrogenation.
94 : 6 trans : cis mixture and in 94% ee. The crude product 5 was
saponified and isolated as its sodium salt (7, NaOTMS, THF,
87% over the two steps, see Scheme 2). Curtius rearrangement
(DPPA, t-BuOH, 80 °C) proceeded smoothly with in situ
trapping of the intermediate isocyanate, to afford cyclic carba-
mate 9 (78%). An interesting advantage in the formation of 9 is
that the oxazolidinone simplifies the protecting group strategy
for the eventual coupling with fragment 3. Despite the success
of the asymmetric reduction approach, we unfortunately encoun-
tered difficulties in finding commercial sources to provide ligand
6 and therefore continued to pursue alternative strategies.
The second strategy investigated was based on a literature pre-
cedent for the preparation of chiral 3,4-substituted piperidines.6
The ring expansion of a 2-deoxy-D-ribose derivative was exam-
ined as an avenue to access 1 (Scheme 3). Formation of the
methyl glycoside 10 (1.25 : 1 mixture of anomers) was con-
ducted at 0 °C (AcCl, MeOH) to minimize competing formation
of the hexose sugar (<8%).7 Subsequent transformation to its
Scheme 3 Synthesis of 1 in enantiomerically pure form from 2-deoxy-
D-ribose.
bis-tosylate derivative (11, 83%) yielded a substrate suitable for
glycoside exchange, via treatment with an excess of benzyl
alcohol in the presence of catalytic TsOH, thus providing 12 as a
5254 | Org. Biomol. Chem., 2012, 10, 5253–5257
This journal is © The Royal Society of Chemistry 2012

View Online
Scheme 5 Synthesis of racemic ( )-1 from tetrahydropyridine 20.
cis-diol 18 (84%). Subsequent treatment with Moffatt’s reagent
and basic hydrolysis (K2CO3, MeOH), resulted in ring closure,
and delivered epoxide (−)-19 in 26% yield and >99.9% ee.
Enantiomerically pure epoxide (−)-19 required only regio-
selective opening with benzylamine to complete the synthesis
of 1 (see Scheme 5).10 Despite the perceived advantages of
using 2-deoxy-D-ribose to prepare 1, this route suffered from
unattractive low yields.
Scheme 4 Synthesis of epoxide (−)-19 from 2-deoxy-D-ribose.
1.3 : 1 mixture of anomers. The excess benzyl alcohol was
difficult to remove from the product at this stage. Therefore, the
crude bis-tosylate 12 was directly subjected to displacement with
BnNH2, providing chromatographically separable cascade
precursors 13a/b in 52% combined yield from 11.8 Interestingly,
the rate of product formation for each anomer (13a/b) was differ-
ent under hydrogenative conditions (H2, Pd/C); the more polar
anomer had a diminished rate for –O-Bn removal, resulting in a
reduced yield of (−)-1 (27%). Extended reaction times (4.5 h),
significantly increased –N-Bn removal. However, the less polar
anomer reached greater than 90% conversion in 2 h, to produce
compound (−)-1 in 48% yield, and >99.9% enantiomeric excess.
This reaction has not been optimized with respect to hydrogen
pressure, catalyst, or acid, and the major by-products are the
–N-benzyl cleaved compounds and the methyl glycosides of the
starting material. Although high quality 1 was prepared using
this strategy, the differential reactivity of 13a/b reduced the
attractiveness of this approach.
A third strategy utilizing 2-deoxy-D-ribose as a chiral building
block was also explored (Scheme 4). It has previously been
reported6 that the N-Boc derivative of 18 could be converted to
the corresponding epoxide by treatment with Moffatt’s reagent.9
In our case, the N-benzyl substrate had yet to be utilized in such
a transformation, and as will be observed (vide infra11), an
N-alkyl substituent is required to obtain a high degree of regio-
selectivity in the epoxide opening with benzylamine. To this
end, 2-deoxy-D-ribose was protected as its isopropylidene deriva-
tive 15 (2-methoxypropene, PPTS) in 67% yield. Reduction of
15 (LiAlH4) afforded the corresponding diol, which after conver-
sion to the bis-mesylate 16 (82%, 2 steps), was displaced with
benzylamine to give piperidine 17 (93%). The isopropylidene
moiety was cleaved under acidic conditions (aq. HCl) to furnish
The final strategy towards 1 focused on the development of a
simple racemic synthesis to epoxide 19 as a key intermediate.
We hoped to take advantage of a recent report detailing the
highly regioselective opening of 3,4-epoxy piperidines by nitro-
gen nucleophiles in the presence of lithium salts.11 This report
thus rendered regioselectivity a non-issue, and left enantioenrich-
ment via resolution as the key step still requiring development.
The synthesis of 1 began with epoxidation of N-benzyl 1,2,5,6-
tetrahydropyridine (20, see Scheme 5) utilizing a two-step pro-
cedure (TFA–H2O then NCS),12 to circumvent potential N-oxide
formation. Thus, the regioisomeric chlorohydrins were treated
with base (K2CO3, MeOH) to provide racemic epoxide 19. The
3,4-epoxy piperidine 19 was isolated as its fumaric acid salt 21
in high yield, 73% (from 20). A salt break (aq. KOH) followed
by Li+-mediated, regioselective (>20 : 1) epoxide opening with
BnNH2 (LiCl, CH3CN)11b afforded ( )-1 as a crystalline solid
(85%). It has been previously shown11b that lithium chelation
between both the oxygen and the basic piperidine nitrogen (i.e.
N-Bn rather than N-Boc) is required to achieve the desired
regioselectivity.
With a simple synthetic strategy to racemic 1 in place, we
explored potential avenues for its enantioenrichment. Initially,
we focused on the resolution of racemic epoxide ( )-19, as resol-
ution earlier in the synthetic sequence had the potential to
increase the efficiency of our route. We were encouraged by the
work of Grishna and coworkers on the biotransformations of
N-benzyl 1,2,5,6-tetrahydropyridines to their corresponding
trans-diol intermediates (i.e. 20 → 22).13 We therefore, investi-
gated the potential to perform a microbial resolution of racemic
epoxide ( )-19. A significant screening of microbes was
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 5253–5257 | 5255

View Online
Fig. 3 X-ray derived ORTEP of bis-HCl salt 24.
Scheme 6 Biocatalytic/kinetic resolution of racemic epoxide (+)-19.
enantiomerically pure form. Each route is characterized by
apparent advantages and limitations, and exploits differing strat-
egies for achieving enantioenrichment. A preliminary evaluation
of these four routes was conducted in terms of safety, cost, ease
of intermediate isolation, availability of starting material, and
overall yield, with the final route, employing a classical resol-
ution, being selected for further optimization. The implemen-
tation of this route on a kilogram scale is currently under
investigation in our laboratories and will be reported in due
course.
Acknowledgements
We would like to thank Dr Qi Gao for providing X-ray crystallo-
graphic analysis of (+)-24. We would also like to thank Dr David
Kronenthal, Dr Martin Eastgate, and Dr Michael Cassidy for
careful review of the manuscript.
Scheme 7 Classical resolution of ( )-1.
conducted (see Scheme 6)14 which interestingly, resulted in only
the undesired enantiomer [i.e. selective hydrolysis of (−)-19 over
(+)-19]. The optimized conditions, [yielding (+)-19] utilized
Aspergillus niger SC 16295 to provide selective hydrolysis of
(−)-19 while leaving epoxide (+)-19 largely intact (49.6% yield,
97.1% ee).15
An alternative enantioenrichment strategy focused on a classi-
cal resolution of racemic intermediate ( )-1. It was discovered,
after a high throughput crystallization screen, that treatment of
( )-1 with R-O-acetyl mandelic acid (1.0 eq.)16 in EtOH (200
proof) generated the desired diastereomeric salt [(+)-23] in
41.7% yield and 93.8% ee (Scheme 7). Unambiguous confir-
mation for the absolute stereochemistry present in 23 was pro-
vided by single crystal X-ray crystallographic analysis (see
Fig. 3).17 This final route provided the target structure (1) in
26% overall yield, and required only two chemical transfor-
mations and a classical resolution.
Notes and references
1 For Bristol-Myers Squibb patents covering
2
see:
(a) C. Wei,
D. J. Norris, Crystalline forms of (3R,4R)-4-amino-1-[[4-[(3-methoxyphenyl)-
amino]pyrrolo[2,1-F][1,2,4]triazin-5-yl]methyl]piperidine-3-ol. US Patent
0191375, August 16, 2007; (b) Bristol-Myers Squibb Company, Pyrrolotria-
zine Compounds as Kinase Inhibitors, WO Patent 066176, July 21, 2005;
(c) Bristol-Myers Squibb Company, Pyrrolotriazine Compounds as Kinase
Inhibitors, US Patent 7,141,571 B2.
2 N. Langlois and O. Clavez, Synth. Commun., 1998, 28, 4471;
J. K. Thottathil, J. L. Monoit, R. H. Mueller, M. K. Y. Wong and
T. P. Kissick, J. Org. Chem., 1986, 51, 3140.
3 General reviews: R. Noyori, Homogeneous Asymmetric Hydrogenation,
in Asymmetric Catalysis in Organic Synthesis, John Wiley & Sons,
New York, 1994, ch. 2, p. 16. For more recent report; X. Sun, W. Li,
G. Hou, L. Zhou and X. Zhang, Adv. Synth. Catal., 2009, 351, 2553;
M. Eggenstein, A. Thomas, J. Theuerkauf, G. Franciò and W. Leitner,
Adv. Synth. Catal., 2009, 351, 725.
4 Z. Guo, B. P. Patel, R. M. Corbett, A. Goswami and R. N. Patel, Tetra-
hedron: Asymmetry, 2006, 17, 2015.
5 R. Noyori, T. Ikeda, T. Ohkuma, M. Widhalm, M. Kitamura, H. Takaya,
S. Akutagawa, N. Sayo and T. Saito, J. Am. Chem. Soc., 1989, 111, 9134;
M. Kitamura, T. Ohkuma, M. Tokunaga and R. Noyori, Tetrahedron:
Asymmetry, 1990, 1, 1.
6 D. Wang and W. A. Nugent, J. Org. Chem., 2007, 72, 7307.
7 O. R. Ludek and V. E. Marquez, Tetrahedron, 2009, 65, 8461.
Conclusion
In summary, four strategies have been demonstrated on lab scale
with the potential to prepare benzyl-protected piperidine 1 in
5256 | Org. Biomol. Chem., 2012, 10, 5253–5257
This journal is © The Royal Society of Chemistry 2012

View Online
8 The importance of the benzyl glycoside to initiate the rearrangement
under the influence of hydrogen was evident at this stage, as no product
was observed if the corresponding methyl glycoside derivative was used
(i.e. 11→1).
9 S. Greenberg and J. G. Moffatt, J. Am. Chem. Soc., 1973, 95, 4016.
10 The conversion of 19 to 1 has been established both in this article (see
Scheme 5) and previously in the literature (ref. 11b).
11 For the regioselective opening of 3,4-epoxy piperidines see:
(a) I. S. Veselov, I. V. Trushkov, N. S. Zefirov and G. V. Grishina,
Russ. J. Org. Chem., 2009, 45, 1050; (b) O. Tokuda, T. Aikawa,
T. Ikemoto and I. Kurimoto, Tetrahedron Lett., 2010, 51, 2832.
12 M. Surkyn, S. F. A. Van Brandt, M. G. C. Verdonck, L. Moens,
J. Cuypers and J. R. M. A. Bosmans, Tetrahedron, 2008, 64, 2456.
13 P. B. Terentiev, T. M. Zilberstein, A. A. Borisenko, V. A. Shmorgunov,
N. F. Piskunkova and G. V. Grishna, Chem. Heterocycl. Compd., 2003,
39, 885.
14 See ESI† for a condensed list of the microbes examined.
15 The absolute stereochemistry of (+)-19 was determined by its conver-
sion to (+)-1 (LiCl, BnNH2) and then comparison of both chiral
HPLC and optical rotation data to that of the known compound.
See ESI.†.
16 A high throughput crystallization screen of epoxide ( )-19 yielded
≤30% ee.
17 Bis-HCl salt 24 (MeOH co-solvate) was serendipitously prepared from
(+)-23 while attempting its recrystallization (from MeOH–CH2Cl2) to
generate crystals appropriate for single crystal X-ray analysis.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 5253–5257 | 5257