Synthesis and stability of oxetane analogs of thalidomide and lenalidomide

Article abstract of DOI:10.1021/ol401705a

Oxetanes are used in drug discovery to enable physicochemical and metabolic property enhancement for the structures to which they are grafted. An imide C=O to oxetane swap on thalidomide and lenalidomide templates provides analogs with similar physicochemical and in vitro properties of the parent drugs, with an important exception: oxetane analog 2 displays a clear differentiation with respect to human plasma stability. The prospect of limiting in vivo stability/metabolism, blocking in vivo racemization, and potentially altering teratogenicity is appealing.

Full text of DOI:10.1021/ol401705a

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Synthesis and Stability of Oxetane
Analogs of Thalidomide and Lenalidomide
Johannes A. Burkhard,† Georg Wuitschik,‡ Jean-Marc Plancher,‡
Mark Rogers-Evans,*,‡ and Erick M. Carreira*,†
Laboratorium fu€r Organische Chemie, ETH Zu€rich, CH-8093 Zu€rich, Switzerland, and
F. Hoffmann-La Roche AG, pRED, Discovery Chemistry, CH-4070 Basel, Switzerland
mark.rogers-evans@roche.com; carreira@org.chem.ethz.ch
Received June 17, 2013
ABSTRACT
Oxetanes are used in drug discovery to enable physicochemical and metabolic property enhancement for the structures to which they are grafted.
An imide CdO to oxetane swap on thalidomide and lenalidomide templates provides analogs with similar physicochemical and in vitro properties
of the parent drugs, with an important exception: oxetane analog 2 displays a clear differentiation with respect to human plasma stability. The
prospect of limiting in vivo stability/metabolism, blocking in vivo racemization, and potentially altering teratogenicity is appealing.
Thalidomide, an antiemetic and sedative, was intro-
duced in the late 1950s and early 1960s to devastating
effect: more than 10000 children in 46 countries were born
with birth defects such as phocomelia, dysmelia, amelia,
bone hypoplasticity, and other congenital defects affecting
the ear, heart, or internal organs.1 As a consequence of
these tragic events, thalidomide’s therapeutic use lay dor-
mant. But, despite a ban, its pharmacology was revisited in
a landmark study by Sheskin2 in 1965 to effectively treat
erythema nodosum leprosum (ENL). Later it was deter-
mined that thalidomide leads to a decrease in TNF-R levels
by enhancing the degradation of its mRNA, and subse-
quently it was approved by the FDA in 1998 with strict
prescription controls. In 1994, studies suggested that tha-
lidomide-linked malformations were the result of the
drug’s interference with vasculogenesis and that a similar
mechanism might prevent the growth of blood vessels in
solid tumors. This led to the 2006 accelerated FDA
approval of thalidomide in combination with dexameth-
asone for the treatment of multiple myeloma. It is under
investigation for various types of cancer, as well as for
inflammatory conditions,3 although, until recently, the
molecular basis of its teratogenicity has been unclear.4
Further studies5 have revealed that thalidomide is a potent
angiogenesis inhibitor in vivo, but no effect was observed
on the proliferation of endothelial cells in culture, leading
to the postulation that a thalidomide metabolite might be
the active agent. This was later demonstrated by determin-
ing that the addition of human liver microsomes was es-
sential to observe an antiangiogenic effect of thalidomide
in exvivostudies.6 Thetwo enantiomersof thalidomideun-
dergo ready interconversion in vivo, and each elicit distinct
effects with (þ)-R as a sedative and (ꢀ)-S as a teratogen.
The metabolic modification of thalidomide and its
susceptibility to racemization led us to consider the pro-
spect of a carbonyl to oxetane switch to control both:
altering the in vivo metabolic profile and rendering the
compound configurationally stable.7 As a first step to
(4) Ito, T. Science 2010, 327, 1345.
(5) D’Amato, R. J.; Loughnan, M. S.; Flynn, E.; Folkman, J. Proc.
Natl. Acad. Sci. U.S.A. 1994, 91, 4082.
(6) (a) Bauer, K. S.; Dixon, S. C.; Figg, W. D. Biochem. Pharmacol.
1998, 55, 1827. (b) Lepper, E. R.; Smith, N. F.; Cox, M. C.; Scripture,
C. D.; Figg, W. D. Curr. Drug Metab. 2006, 7, 677. (c) Kumar, N.;
Sharma, U.; Singh, C.; Singh, B. Curr. Top. Med. Chem. 2012, 12, 1436.
(7) Oxetanes have been shown to have increased metabolic stability
relative to other cyclic ethers: Stepan, A. F.; Karki, K.; McDonald,
W. S.; Dorff, P. H.; Dutra, J. K.; DiRico, K. J.; Won, A.; Subramanyam,
C.; Efremov, I. V.; O’Donnell, C. J.; Nolan, C. E.; Becker, S. L.;
Pustilnik, L. R.; Sneed, B.; Sun, H.; Lu, Y.; Robshaw, A. E.; Riddell,
D.; O’Sullivan, T. J.; Sibley, E.; Capetta, S.; Atchison, K.; Hallgren,
A. J.; Miller, E.; Wood, A.; Obach, R. S. J. Med. Chem. 2011, 54, 7772.
† ETH Z€urich.
‡ F. Hoffmann-La Roche AG.
(1) Matthews, S. J.; McCoy, C. Clin. Ther. 2003, 25, 342.
(2) Sheskin, J. Clin. Pharmacol. Ther. 1965, 6, 303.
(3) Chen, M.; Doherty, S. D.; Hsu, S. Dermatol. Clin. 2010, 3, 577.
r
10.1021/ol401705a
XXXX American Chemical Society

address these issues, this communication describes our
preliminary investigation of both enantiomers of the ox-
etane analogues of thalidomide and its close congener
lenalidomide (Figure 1), which was approved for the
treatment of multiple myeloma in 2004 and under clinical
investigation for a variety of other cancers, together with a
comparison of theirphysicochemicalandinvitrometabolic
properties.
group. The mild conditions we have previously reported
for the conversion of primary nitro groups to the corre-
sponding oximes11 (BnBr, KOH, cat. TBAI) were success-
fully applied to the nitro group present in 8, and the oximes
9 (as a 1:1.7 stereoisomeric mixture) were isolated in 76%
yield and an overall yield of 64% from 5.
Scheme 1. Pathways toward Oximes 9
Figure 1. Thalidomide, lenalidomide, and oxetane analogues.
Subsequent heating of oximes 9 in xylenes led to the
formation of lactam 10, which was isolated as a single
diastereomer in 84% yield (Scheme 2).12 Reduction of the
oxime withRaney-Ni under an atmosphere ofH2 delivered
amine 11 in excellent yield. Amine 11 was treated in the
subsequent step with phthaloyl chloride/Et3N, followed
by DBU,13 leading to clean formation of phthalimide 12
in 82% yield.14 Finally, deprotection of the PMB-amide
with CAN in CH3CN/H2O gave thalidomide analog 2
(54% yield).
We envisioned that amine 11 would also provide access
to the oxetane analog (4) of lenalidomide (3 in Figure 1).
Consequently, 11 was treated with substituted benzyl
bromide 1315 and Et3N in hot DMF to afford the iso-
indolin-1-one 14 in 73% yield (Scheme 3).16 Deprotection
of the PMB group was effected with CAN in aqueous
CH3CN to give free amide 15 (43% yield). Reduction of
The synthesis of the oxetane analogue 2 commenced
with a Henry addition of methyl γ-nitrobutanoate8 (5) and
oxetan-3-one,9 a reaction conducted neat in the presence of
0.2 equiv of Et3N (Scheme 1). The 3-oxetanyl alcohol
formed in situ was dissolved in CH2Cl2 and treated at
low temperature with MsCl and Et3N. Slow warming of
the reaction mixture and quenching upon reaching ambi-
ent temperature was necessary for optimal yield of6. When
stirring is allowed to continue at room temperature, sig-
nificant amounts of a rearranged isoxazole compound are
otherwise formed, as previously described.10 At this point
chromatography on silica gel afforded two product frac-
tions, containing nitroalkene 6 as well as a mixture of 6
and mesylate 7. These fractions were independently pro-
cessed en route toward 2.
Treatment of a 1:4 mixture of nitroalkene 6 and mesylate
7 with Et3N and 4-methoxybenzylamine in THF afforded
3-aminooxetane 8 in 46% yield, a compound which is also
readily formed when 6 is separately treated with 4-meth-
oxybenzylamine. All attempts at forming the δ-lactam at
this stage produced extensive decomposition of starting
material. Consequently, we focused on the manipulation
of the nitro group. Although reduction to the correspond-
ing amine was possible (e.g., with Zn, aq. HCl in i-PrOH),
the outcome of these reactions was unsatisfactory. There-
fore, we decided to attempt partial reduction of the nitro
(11) (a) Czekelius, C.; Carreira, E. M. Angew. Chem., Int. Ed. 2005,
44, 612. (b) Diethelm, S.; Carreira, E. M. J. Am. Chem. Soc. 2013, 135,
8500.
(12) We were unable to determine the stereochemistry of the oxime. It
is tentatively assigned as the energetically favored (E)-isomer.
(13) (a) Myers, A. G.; Gin, D. Y.; Rogers, D. H. J. Am. Chem. Soc.
1993, 115, 2036. (b) Myers, A. G.; Gin, D. Y.; Rogers, D. H. J. Am.
Chem. Soc. 1994, 116, 4697.
(14) It is worth noting that heating amine 11 with phthalic anhydride
in toluene led to extremely sluggish product formation, presumably due
to the sterically hindered nature of the amino group in 11.
(15) Prepared from the corresponding toluene in a radical bromina-
€
tion: Rohrig, S.; Jeske, M.; Perzborn, E.; Gnoth, M. J.; Prezborn, E.
(Bayer Healthcare AG). US2010010060, 2010.
(16) This type of transformation is frequently used in the synthesis of
lenalidomide and its analogues. Selected references: (a) Muller, G. W.;
Chen, R.; Huang, S. Y.; Corral, L. G.; Wong, L. M.; Patterson, R. T.;
Chen, Y. X.; Kaplan, G.; Stirling, D. I. Bioorg. Med. Chem. Lett. 1999, 9,
1625. (b) Muller, G. W.; Stirling, D. I.; Chen, R. S.-C. (Celgene Corp.).
US 6335349, 2002. (c) Fujimoto, H.; Noguchi, T.; Kobayashi, H.; Miyachi,
H.; Hashimoto, Y. Chem. Pharm. Bull. 2006, 54, 855.
(8) Prepared according to: Simoni, D.; Rondanin, R.; Morini, M.;
Baruchello, R.; Invidiata, F. P. Tetrahedron Lett. 2000, 41, 1607.
(9) We are grateful to Dr. Thomas Fessard at SpiroChem AG,
Zurich, Switzerland for providing oxetan-3-one as a gift.
(10) Burkhard, J. A.; Tchitchanov, B. H.; Carreira, E. M. Angew.
Chem., Int. Ed. 2011, 50, 5379.
B
Org. Lett., Vol. XX, No. XX, XXXX

the NꢀH pKa’s were not significantly affected, and overall,
all properties were not significantly altered, with the
exception of reduced membrane permeation for 4, mea-
sured via PAMPA.
Scheme 2. Synthesis of Oxetano-thalidomide, 2
Table 1. Physicochemical Properties of 1ꢀ4
b
log Da
pKa
sol.c
PAMPAd
1
2
3
4
0.24
10.7/-
29
3.5
0.0
>11/-
78
1.3
ꢀ0.64
10.4/2.6
>11/2.6
100
100
0.47
0.06
ꢀ0.67
a Logarithmic n-octanol/water distribution coefficient at pH 7.4.
b Ionization constants determined at 23 °C by spectrophotometry in
water. c Intrinsic solubility measured at pH 6.5 by a Lyophilization
Solubility Assay [μg/mL]. d Passive diffusion speed measured at pH 6.5
in a Parallel Artificial Membrane Permeation Assay [10ꢀ6 cm/s]. For
details, see the Supporting Information.
the nitro group was readily accomplished with Pd/C and
H2, and oxetano-lenalidomide (4) was obtained in 99%
yield.
The synthesis of the described oxetane analogs of both
thalidomide and lenalidomide produced the targeted com-
pounds in racemic form. However, because the complex
pharmacology of thalidomide results from the differential
effects associated with the separate enantiomers and their
metabolic byproducts, we sought to gain access to enan-
tiomerically enriched forms of the oxetane analogues.
Unfortunately, separation of the target compounds 2
and 4 through the implementation of chiral HPLC was
difficult because of issues of solubility. Accordingly, other
intermediates were examined for resolution. Amine 11 was
found to be readily separable by chiral HPLC, thereby
giving gram quantities of (þ)-11in 37% yield and (ꢀ)-11 in
35% yield. Both enantiomers were subsequently taken
forward to the corresponding thalidomide and lenalido-
mide analogs (see the Supporting Information for details).
Similarly, the known low in vitro microsomal and
hepatocyte intrinsic clearances of the parent compounds,
thalidomide and lenalidomide, were predictably marginally
affected by introduction of the oxetane and remained
in a favorable range (Table 2) for the corresponding
analogs. Additionally, 2C9, 2D6, and 3A4 cytochrome
inhibition remained >50 μM for oxetane derivatives 2
and 4, thus maintaining the reduced risk for drugꢀdrug
interactions.
Table 2. Intrinsic Clearance in Microsomes and Hepatocytes
microsomal clearancea
human
hepatocyte
human
mouse
rat
clearanceb
Scheme 3. Synthesis of Oxetano-lenalidomide, 4
1
6
5
2
3
1
3
0
0
4
5
6
3
2
3
0
0
9
6
6
7
0
1
0
-
3.7
2
2.3
(þ)-2
(ꢀ)-2
3
2.6
2.5
1.5
4
0.17
0.23
0.40
(þ)-4
(ꢀ)-4
a Intrinsic clearance rates [μL/min/mg] measured in human, mouse,
and rat liver microsomes. b Intrinsic clearance rates [μL/min/106 cells]
measured in human hepatocytes. For details, see the Supporting
Information.
Screening of the derivatives in a human plasma stability
assay, which is useful for interpreting follow-up in vivo
efficacy, toxicity, and pharmacokinetic studies, proved
intriguing (Table 3). In particular, thalidomide was un-
stable in comparison to its oxetane counterpart. It is
therefore reasonable to expect that the latters’ in vivo phar-
macokinetic and toxicity profile will also be critically
differentiated (studies pending).
With the aim of profiling the oxetano-derivatives, phy-
sicochemical properties were evaluated and benchmarked
against thalidomide and lenalidomide, all in their racemic
forms (Table 1). Despite the introduction of the oxetane,
Thalidomide and lenalidomide are therapeutically use-
ful, but their patient distribution is severely limited due to
Org. Lett., Vol. XX, No. XX, XXXX
C

proach to potentially control both of these factors. In
summary, we have prepared the oxetane analogs of thali-
domide and lenalidomide and effected their resolution.
The comparative study reveals no significant differences in
their physicochemical and in vitro metabolic profiles.
However, compound 2 is strongly differentiated from
thalidomide (1) in human plasma. The experiments under-
score unexpected benefits of oxetanes in drug discovery
that surface in this case for plasma stability. Efforts are
ongoing in the design of experiments to determine the in
vivo stability/metabolic profiles of the compounds synthe-
sized as well as assaying their teratogenic potential.
Table 3. Stability in Human Plasmaa
incubation time (h)
0.25
1
5
category
1
100
100
100
100
100
100
100
100
69
40
74
72
68
63
77
68
71
unstable
stable
stable
stable
stable
stable
stable
stable
2
76
(þ)-2
(ꢀ)-2
3
74
102
93
4
112
124
117
(þ)-4
(ꢀ)-4
Acknowledgment. Mr. Daniel Zimmerli (F. Hoffmann-
La Roche) is acknowledged for the chiral separation of
amine 11, and Dr. Manfred Schneider (F. Hoffmann-La
Roche), for plasma measurements and proofreading. We
also thank the ETH for support of this research in the form
of a grant ETH-15 10-2. J.A.B. thanks Novartis and the
Roche Research Foundation for graduate fellowships. We
are also grateful to F. Hoffmann-La Roche for the gener-
ous support of our research program.
a Stability in human plasma after 0.25, 1, and 5 h incubation time,
expressed as a percentage of the initial concentration. The reported
values represent the average of three runs (N = 3). For details, see the
Supporting Information.
their inherent teratogenic properties. This negative aspect
might be ameliorated by controlling their in vivo stability/
metabolism and/or their proclivity to racemization. The
use of oxetanes as carbonyl surrogates17 offers a novel ap-
Supporting Information Available. Experimental pro-
cedures and characterization for all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(17) (a) Wuitschik, G.; Rogers-Evans, M.; Buckl, A.; Bernasconi, M.;
€
Marki, M.; Godel, T.; Fischer, H.; Wagner, B.; Parrilla, I.; Schuler, F.;
Schneider, J.; Alker, A.; Schweizer, W. B.; Muller, K.; Carreira, E. M.
€
Angew. Chem., Int. Ed. 2008, 47, 4512. (b) Wuitschik, G.; Carreira,
E. M.; Wagner, B.; Fischer, H.; Parrilla, I.; Schuler, F.; Rogers-Evans,
€
M.; Muller, K. J. Med. Chem. 2010, 53, 3227. For a review on the use of
oxetanes in synthesis and drug discovery, see: (c) Burkhard, J. A.;
€
Wuitschik, G.; Rogers-Evans, M.; Muller, K.; Carreira, E. M. Angew.
The authors declare the following competing financial interest: E.M.
C. sits on the Board of SpiroChem AG, the company that provided
oxetane-3-one as a gift (see ref 8).
Chem., Int. Ed. 2010, 49, 9052.
D
Org. Lett., Vol. XX, No. XX, XXXX