Efficient synthesis of cis-2,6-di-(2-quinolylpiperidine)

Article abstract of DOI:10.1016/j.tetlet.2013.07.067

An efficient synthesis of cis-2,6-di-(2-quinolylpiperidine) has been developed. The key steps involve Wittig reaction of N-Cbz-protected cis-piperidine-2,6-dicarboxaldehyde (3) with 2-(triphenylphosphinyl-methyl) quinoline bromide (4) and sequential removal of the N-Cbz group and double bond reduction. This synthetic procedure provides an efficient preparation for this useful norlobelane analogue.

Full text of DOI:10.1016/j.tetlet.2013.07.067

Tetrahedron Letters 54 (2013) 5211–5213
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Efficient synthesis of cis-2,6-di-(2-quinolylpiperidine)
Derong Ding a, Linda P. Dwoskin a, Peter A. Crooks b,
⇑
a Department of Pharmaceutical Science, College of Pharmacy, University of Kentucky, Lexington, KY 40536, USA
b Department of Pharmaceutical Science, College of Pharmacy, University of Arkansas for Medical Sciences, Little Rock, AR 72205-7199, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
An efficient synthesis of cis-2,6-di-(2-quinolylpiperidine) has been developed. The key steps involve Wit-
tig reaction of N-Cbz-protected cis-piperidine-2,6-dicarboxaldehyde (3) with 2-(triphenylphosphinyl-
methyl)quinoline bromide (4) and sequential removal of the N-Cbz group and double bond reduction.
This synthetic procedure provides an efficient preparation for this useful norlobelane analogue.
Ó 2013 Elsevier Ltd. All rights reserved.
Received 28 June 2013
Accepted 9 July 2013
Available online 17 July 2013
Keywords:
Lobelane
cis-2,6-Di-(2-quinolylpiperidine)
Norlobelane analogues
Wittig reaction
Swern oxidation
Quinlobelane
Introduction
obtain more drug-like 2-quinolyl analogues of both norlobelane
and lobelane for studying both structure–activity and structure–
Lobelane (N-methyl-cis-2,6-diphenethylpiperidine, Fig. 1), a
minor piperidine alkaloid of Lobelia inflata,1 has been shown to
exhibit good affinity and selectivity for the tetrabenazine (TBZ)
binding site on the vesicular monoamine transporter (VMAT2),
and is also a potent inhibitor of vesicular dopamine uptake.2 Struc-
ture–activity relationship (SAR) studies indicate that lobelane
analogues lacking the N-methyl substituent (i.e., norlobelane ana-
logues) retain their affinity for the TBZ binding site on VMAT2,
suggesting that the presence of the N-methyl group is not critical
for interaction with VMAT2. Furthermore, other studies have
shown that norlobelane is a potent inhibitor of [3H]DA uptake into
striatal vesicles (Ki = 43 nM).2 SAR studies have also shown that the
phenyl moieties in lobelane can be replaced with 1-naphthyl or
2-naphthyl groups, resulting in analogues that are potent and
selective inhibitors of VMAT2.2a However, such analogues have
poor water-solubility and poor drug-like properties. Recent studies
have shown that lobelane analogues in which the phenyl moieties
have been replaced with heterocyclic rings, such as pyridyl, quino-
lyl, or indolyl, have improved water-solubility (Fig. 1). Only the
quinolyl analogues retained potent VMAT2 inhibitory properties,3
with quinlobelane (Fig. 1) exhibiting potent inhibition of vesicular
[3HDA] uptake (Ki = 51 nM). However, the synthetic procedures
utilized in this study were not amenable to the synthesis of 2-qui-
nolyl analogues of norlobelane [e.g., cis-2,6-di-(2-quinolylpiperi-
dine); compound 1, Scheme 1] and its analogues. In order to
property relationships, a new and efficient synthesis of cis-2,6-di-
(2-quinolylpiperidine) is now reported, which may be useful for
the general synthesis of a wide range of compounds of this type.
Results and discussion
We now report a versatile and efficient method for the prepara-
tion of 2-quinolylnorlobelane. Our retrosynthetic approach is
outlined in Scheme 1, and is centered around a Wittig reaction
for the construction of two double bonds sequentially. The requi-
site precursor 3 (Scheme 1) can be synthesized from commercial
pyridine-2,6-dicarboxylic acid.
The synthesis of target molecule 1 was executed as shown in
Scheme 2. Pyridine-2,6-dicarboxylic acid was heated under reflux
in methanol containing a few drops of concentrated H2SO4 to form
⇑
Corresponding author. Tel.: +1 501 686 6495; fax: +1 501 686 6057.
E-mail address: pacrooks@uams.edu (P.A. Crooks).
Figure 1. Structures of lobelane, heterocyclic analogs of lobelane and quinlobelane.
0040-4039/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2013.07.067

5212
D. Ding et al. / Tetrahedron Letters 54 (2013) 5211–5213
Scheme 1. Retrosynthetic analysis of cis-2,6-di-(2-quinolyl piperidine) (1).
Scheme 2. Synthetic route to compound 1. Reagents and conditions: (a) Concd H2SO4, MeOH, reflux, 71%; (b) 10% Pd/C, H2O, rt, 91%; (c) CbzCl, DIPEA, THF, rt, quant; (d) LiBH4,
THF, 0 °C?rt, 81%; (e) Swern oxidation, À78 °C; (f) 4, tert-BuOK, THF, rt, 51% (two steps); (g) 6 N HCl, reflux, quant; (h) 10% Pd/C, rt, 75%.
Scheme 3. Synthetic route to compound 4. Reagents and conditions: (a) SeO2, solvent-free, 170 °C, 81%; (b) NaBH4, EtOH, rt, 92%; (c) 33% HBr/AcOH, reflux, 98%; (d) PPh3,
toluene, reflux, 95%.
the sulfate salt 5. Reduction of 5 under 50 psi H2 pressure followed
by crystallization of the crude product from hexane provided the
pure cis-isomer 6 in 91% yield,4 which was then protected as its
N-Cbz derivative 7. In our original plan, we utilized DIBAL-H for
the reduction of 7, in the hope of obtaining aldehyde 3 in one step.
Unfortunately, an unidentified mixture was obtained after work-
up of the reaction. To obtain the key intermediate 2, we reduced
ester 7 with LiBH4 to afford alcohol 8. Then, we explored Dess–
Martin periodinane and PCC reactions in our attempts to oxidize
compound 8 to the key intermediate 3. However, only complex
mixtures were obtained, which may have been due to facile
decomposition of aldehyde 3 during purification by column chro-
matography. We subsequently found that alcohol 8 could be oxi-
dized efficiently to 3 under Swern conditions.5 In the Wittig

D. Ding et al. / Tetrahedron Letters 54 (2013) 5211–5213
5213
reaction of 3 with compound 4 (Scheme 3), which is the pivotal
step in the synthesis of 1, we found that it was advantageous to
utilize the crude aldehyde 3 directly without further purification
to afford the optimal yield of 2 (it is noteworthy that all attempts
to obtain a pure sample of compound 3 failed). In our initial at-
tempts to synthesize 2, THF was used as solvent and n-BuLi was
utilized as base, and the product was isolated in 6% yield in the
two-step procedure. In order to optimize the reaction conditions,
other bases were evaluated. The yields obtained by substituting
n-BuLi with either LiHMDS, NaHMDS, or NaOEt were 8%, 10%,
and 11%, respectively, for the two-step synthetic procedure. To
our satisfaction, when tert-BuOK was used as base, the yield of 2
improved significantly to 51% for the two step reaction.
phosphine (PPh3) in toluene to obtain the desired compound 4 in
95% yield.
Characterization data (1H NMR, 13C NMR and high resolution
mass spectrometry) for compounds 1, 2 and 9 are provided in
the References and Notes section8.
In conclusion, an efficient method for the preparation of cis-2,6-
di-(2-quinolylpiperidine) (1) has been developed. The key step in
the synthetic scheme is the introduction of the two 2-quinolyl
moieties via Wittig reaction with N-Cbz-protected cis-piperidine-
2,6-dicarboxaldehyde. Other quinolyl analogues of norlobelane
and lobeline are currently being prepared utilizing this synthetic
procedure, and their biological evaluation is underway.
With the key compound 2 in hand, our initial strategy was to
synthesize compound 1 from 2 in one step by removal of the N-
Cbz group followed by double bond hydrogenation over 20%
Pd(OH)2. However, when this procedure was followed, TLC analysis
indicated a complex mixture, which proved difficult to purify by
column chromatography.
A similar outcome was observed when 10% Pd/C was used. We
speculated that these problems might be due to hydrogenolytic
ring opening of the piperidine ring under the reduction condi-
tions utilized. In order to circumvent this problem, a strategy
involving two separate steps was employed. First, we attempted
to reduce the double bonds utilizing Wilkinson’s catalysis6 prior
to removal of the N-Cbz group; however, no reaction occurred,
and the starting material was recovered. Subsequently, we turned
to a second strategy, and attempted the removal of the N-Cbz
group of 2 followed by double bond reduction. We found that
6 N HCl at reflux could be used to deprotect the N-Cbz group
affording compound 9 in quantitative yield. Hydrogenation of 9
to the desired compound 1 was achieved utilizing 10% Pd/C as
catalyst in 75% yield.
Scheme 3 provides the synthetic route to phosphonium salt 4,
which was utilized in the Wittig reaction of compound 3 to 2
(Scheme 2). 2-Methylquinoline was oxidized by SeO2 at a high
temperature (170 °C) under solvent-free conditions to afford alde-
hyde 10 in good yield. It should be noted that when this oxidation
reaction was performed in high boiling points solvents such as 1,4-
dioxane, the yield was very low, even after prolonged reaction
times. Aldehyde 10 could be reduced by NaBH4 in EtOH to afford
carbinol 11 in 92% yield, followed by bromination of 11 in 33%
HBr/AcOH to afford bromide 12 in high yield7 (we found that when
compound 10 was brominated with 48% aq HBr, the yield of 12 was
only 72% and the reaction required a long time for completion).
Compound 12 was then heated under reflux with triphenyl
Acknowledgments
The authors gratefully acknowledge support from NIH grant
U01 DA13519. The University of Kentucky holds patents on lobe-
line and the analogues described in the current work. A potential
royalty stream to LPD and PAC may occur consistent with Univer-
sity of Kentucky policy.
References and notes
1. Wieland, H.; Schopf, C.; Hermsen, W. Liebigs Ann. Chem. 1925, 444, 40–68.
2. (a) Zheng, G.; Dwoskin, L. P.; Deaciuc, A. G.; Norrholm, S. D.; Norrholm, S. D. J.
Med. Chem. 2005, 48, 5551–5560; (b) Zheng, G.; Dwoskin, L. P.; Deaciuc, A. G.;
Norrholm, S. D.; Norrholm, S. D. Bioorg. Med. Chem. Lett. 2008, 18, 6509–6512.
3. Vartak, A. P.; Deaciuc, A. G.; Dwoskin, L. P.; Crooks, P. A. Bioorg. Med. Chem. Lett.
2010, 20, 3584–3587.
4. Robert, C.; Michael, D. J. Org. Chem. 1996, 61, 3332–3341.
5. Harry, H. W.; Karen, R.; Roman, K. Tetrahedron Lett. 1989, 30, 6077–6080.
6. Baldwin, J. E.; Adlington, R. M.; Jones, R. H.; Schofield, C. J.; Zaracostas, C.
Tetrahedron 1986, 42, 4879–4888.
7. Markus, S.; Gebhard, H. Eur. J. Org. Chem. 2009, 26, 4458–4467.
8. The spectroscopic data of all known compounds are identical to the reported data.
All new compounds were characterized by 1H NMR, 13C NMR and high
resolution mass spectrometry (HRMS). Selected data (NMR spectra were
recorded in CDCl3, 1H at 300 MHz and
C at 75 MHz): (i) Compound 2:
13
colorless oil; 1H NMR: d 7.91 (d, 2H, J = 8.4 Hz), 7.78 (d, 2H, J = 7.5), 7.68–7.60
(m, 4H), 7.48–7.25 (m, 9H), 6.86 (dd, J = 16.2 Hz, 2H), 6.82 (dd, J = 16.2 Hz, 2H),
5.24 (s, 2H), 5.21 (s, 2H), 2.13–1.84 (m, 5H), 1.68–1.63 (m, 1H) ppm; 13C NMR: d
155.9, 155.8, 148.1, 137.6, 136.8, 136.2, 131.5, 129.6, 129.4, 128.7, 128.3, 128.2,
127.6, 127.4, 126.2, 119.3, 67.8, 52.2, 28.5, 15.6 ppm; HRMS (EI) Calcd for
C
35H31N3O2, 525.2416. Found 525.2419 (ii) Compound 9: viscous oil; 1H NMR: d
8.10–8.01 (m, 4H), 7.77–7.65 (m, 4H), 7.58–7.45 (m, 4H), 7.26–6.86 (m, 4H),
3.60–3.55 (m, 2H), 2.04–1.86 (m, 3H), 1.63–1.43 (m, 3H); 13C NMR: d 156.0,
148.2, 139.3, 136.5, 130.8, 129.8, 129.4, 127.6, 127.5, 126.3, 119.0, 59.2, 32.1,
24.7; HRMS(EI) calcd for C27H25N3 391.2048; Found 391.2045 (iii) Compound 1:
viscous oil; 1H NMR: d 8.08–7.99 (m, 4H), 7.76–7.79 (m, 2H), 7.62–7.68 (m, 2H),
7.50–7.45 (m, 2H), 7.32–7.26 (m, 2H), 3.11–3.06 (m, 4H), 2.61–2.57 (m, 2H),
2.02–1.94 (m, 4H), 1.78–1.71 (m, 3H), 1.27–1.19 (m, 3H); 13C NMR: d 162.5,
148.0, 136.6, 129.6, 128.9, 127.7, 126.9, 125.9, 121.6, 56.8, 37.0, 35.9, 32.4, 24.8;
HRMS(EI) Calcd for C27H29N3 395.2361. Found 395.2367.