Synthesis of SB 222618. A potential PDE IV inhibitor

Article abstract of DOI:10.1016/S0040-4039(99)02198-X

PDE IV inhibitor SB 222618 was prepared by regioselective S(N)2' addition of 9 to bromoallene 6a followed by a stereoselective borane reduction of 4 to afford 2 which by a palladium mediated coupling with 3, delivered SB 222618 (1) in good yield. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(99)02198-X

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 811–814
Synthesis of SB 222618. A potential PDE IV inhibitor†
Jose J. Conde ∗ and Wilford Mendelson
Department of Synthetic Chemistry, SmithKline Beecham Pharmaceuticals, 709 Swedeland Road, PO Box 1539,
King of Prussia, PA 19406, USA
Received 1 November 1999; revised 23 November 1999; accepted 24 November 1999
Abstract
PDE IV inhibitor SB 222618 was prepared by regioselective SN20 addition of 9 to bromoallene 6a followed by
a stereoselective borane reduction of 4 to afford 2 which by a palladium mediated coupling with 3, delivered SB
222618 (1) in good yield. © 2000 Elsevier Science Ltd. All rights reserved.
SB 222618 (1) has been a target of synthetic chemists at SmithKline Beecham owing to its potential
PDE IV1 inhibitor activity against inflammatory diseases such as asthma. Apart from its biological
activity, this cyclohexanol provided the opportunity to explore the non-obvious construction of the
quaternary C4 of 1 whose retrosynthesis is shown below (Scheme 1).
As depicted in Scheme 1, SB 222618 (1) could be obtained via the Sonogashira2 coupling protocol
between heterocycle 3 and alkyne 2. The substituted cyclohexanol 2 could then be made from ketone
4 by selectively reducing the keto group delivering the desired stereochemistry at C1. To prepare 4, we
envisioned the possibility of performing a SN20 addition of cuprate 9 to bromoallene 6a which would be
obtained from the corresponding carbinol 7 coming from treating ketone 8 with lithium acetylide.
The most critical steps in the synthesis of SB 222618 are the preparation of intermediate 4 and further
selective reduction of the keto group to deliver 2. The preparation of bromoallenes is well precedented in
5
the literature. Reagents such as HBr,3 PBr3,4 SOBr2 were extensively used to synthesize bromoallenes
from propargylic alcohols. Other conditions such as addition of cuprates6 to propargylic mesylates or
tosylates are also known.
Propargylic alcohol 7 was prepared without incident from commercially available cyclohexanedione
mono-ethyleneketal 8 by reaction with excess of lithium acetylide7 (Scheme 2). For the preparation of
bromoallene 6a, we first decided to investigate the SN20 addition of milder reagents such as LiCuBr28 to
the corresponding mesylate. This reaction afforded a mixture of regioisomers 6a and 6b (2:1).
This poor regioselectivity was later improved by using SOBr2 in the presence of Et3N as acid scavenger
obtaining bromoallene 6a as the sole product in 90% yield.
∗
†
Corresponding author.
This work is dedicated to the memory of Dr. Lendon Pridgen
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(99)02198-X
tetl 16126

812
Scheme 1.
Scheme 2. Reaction conditions. (a) lithium acetylide·EDA/dioxane/rt/80%; (b) nBuLi/MsCl/THF −60°C; (c) LiCuBr2/−60°C
to rt (33% of 6a); (d) SOBr2/Et3N/TBME/0°C/90%
The preparation of cuprate 9 (Scheme 3) would require having bromophenol 11 in place which
then would be alkylated to afford 10. Starting with readily available guaiacol, preliminary attempts to
regioselectively brominate the 5 position were unsuccessful and gave, instead, mixtures of 4-bromo-2-
methoxyphenol (major), some desired product (11) and dibrominated by-products.
Scheme 3.
To overcome this problem we masked the hydroxy group in situ as its trifluoroacetate intermediate
12 (Scheme 4) which without isolation cleanly afforded bromide 13 upon reaction with NBS. Further
conversion into 11 after aqueous work-up followed by alkylation with cyclopentyl bromide gave 10 in
excellent yield.
With 10 in hand, we prepared and evaluated the addition of several organocuprates (M=CuLi and
M=CuCN) to bromoallene 6a. The majority of these reagents produced mixtures of regioisomers among
other impurities. We then found a report9 in the literature regarding the fact that Vermeer-type10

813
Scheme 4. Reaction conditions. (a) (CF3CO)2O/cat. KtOBu/CH3CN/30 min; (b) NBS/rt/36 h; (c) NaOH (3N) then HCl
(3N)/85%; (d) cyclopentyl bromide/K2CO3/DMF/100°C/90%; (e) Mg/THF/rt/CuLiBr2/−70°C to rt/80%
organocopper species of formula RCuMg2Br3·LiBr were far more regioselective by adding preferentially
in a SN20 fashion instead of direct bromide displacement.
Hence, 9 (M=CuMg2Br3.LiBr) was prepared (Scheme 4) and reacted with 6a to yield the desired
product (5) which was then hydrolyzed to give ketone 4 (Scheme 5) that was submitted to a variety of
reducing agents such as DIBALH, NaBH4, LiBH4 and BH3·THF under various conditions. The addition
of less hindered hydride donors to cyclohexanones tends to give predominantly the equatorial alcohol.
One explanation of the preference for formation of the equatorial isomer involves the torsional strain that
develops in formation of the axial alcohol.11 We found that selectivity decreased in the order: BH3·THF
(96% d.e.)>LiBH4 (93% d.e.)>NaBH4 (85% d.e.)»>DIBALH. Hence carbonyl reduction with BH3 at
−40°C afforded alcohol 212 in 80% yield. SB 222618 (1) was finally obtained by Pd(0)/Cu(I) catalyzed
coupling with aminopyrimidine 3.
Scheme 5. Reaction conditions. (f) 6a/THF/−70°C to rt/60%; (g) AcOH (50%)/90°C/90%; (h) BH3·THF/−40°C/80%; (i)
3/Pd(PPh3)4/CuI/Et2NH/DMSO/80%
In summary, we report herein a very efficient synthesis of PDE IV inhibitor SB 222618 (1) by using as
key step the SN20 addition of organocuprate 9 to bromoallene 6a followed by BH3 reduction of 4 to give

814
2 in 96% d.e. We also developed a one-pot regioselective method of bromination of 2-methoxyphenol to
afford 11. Application of this methodology to different substituted phenols is currently in progress.
Acknowledgements
We thank Dr. Andrew Allen for 400 MHz 1H NMR measurements.
References
1. 30,50-Cyclic monophosphate (cAMP)-specific cyclic nucleotide phosphodiesterases (PDEs).
2. Sonogashira, K.; Tohda, V.; Hagihara, N. Tetrahedron Lett. 1975, 4467.
3. Misahiro, M.; Masaaki, Y.; Sejji, S.; Keisuke, T.; Takashi, T. Bull. Chem. Soc. Jpn. 1988, 61, 2067.
4. Shoji, E.; Tomoyuki, I.; Yasuyuki, K.; Tadashi, S. J. Chem. Soc. Chem. Commun. 1985, 958.
5. Corey, E.; Boaz, N. Tetrahedron Lett. 1984, 25, 3055.
6. Caporusso, A.; Consoloni, C.; Lardicci, L. Gazz. Chim. Ital., 1988, 118, 857; Elsevier, C.; Vermeer, P. J. Org. Chem. 1984,
49, 1649; Montury, M.; Gore, J. Synth. Commun. 1980, 10, 873.
7. Huffman, J.; Arapakos, P. J. Org. Chem. 1965, 30, 1604.
8. LiCuBr2 was prepared from stoichiometric LiBr and CuBr in THF.
9. Caporruso, A.; Polizzi, C.; Lardicci, L. J. Org. Chem. 1987, 52, 3920.
10. Westmijze, H.; Vermeer, P. Synthesis 1979, 390.
11. Cherest, M.; Felkin, H.; Prudent, N. Tetrahedron Lett. 1968, 2205.
12. The molecular structure of SB 222618 was determined by 3-dimentional X-ray crystallography.