A new synthesis of cytoxazone and its diastereomers provides key initial SAR information

Article abstract of DOI:10.1016/S0960-894X(03)00131-8

A short, enantioselective, and diastereoselective synthesis of cytoxazone, a Th2-selective immunomodulator from Streptomyces, is described. The route was readily adapted to the synthesis of the three other stereoisomers of natural cytoxazone. Evaluation of these compounds revealed that the stereochemical configuration of the oxazolidinone ring did not influence their biological activity.

Full text of DOI:10.1016/S0960-894X(03)00131-8

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1237–1239
A New Synthesis of Cytoxazone and Its Diastereomers
Provides Key Initial SAR Information
Percy H. Carter,* Jacob R. LaPorte, Peggy A. Scherle and Carl P. Decicco
Bristol-Myers Squibb Company, Pharmaceutical Research Institute, Princeton, NJ 08543-4000, USA
Received 8 November 2002; revised 27 January 2003; accepted 3 February 2003
Abstract—A short, enantioselective, and diastereoselective synthesis of cytoxazone, a Th2-selective immunomodulator from
Streptomyces, is described. The route was readily adapted to the synthesis of the three other stereoisomers of natural cytoxazone.
Evaluation of these compounds revealed that the stereochemical configuration of the oxazolidinone ring did not influence their
biological activity.
# 2003 Elsevier Science Ltd. All rights reserved.
In 1998, Osada and co-workers reported the isolation of
cytoxazone (1, Fig. 1) from a soil sample of Strepto-
myces sp,1 and identified it as a selective modulator of
TH2 cytokine secretion.1 The structure of 1 was defini-
tively determined using X-ray crystallographic analysis,2
and the absolute stereochemistry was assigned via com-
parison of the circular dichroism spectrum of 1 with the
Figure 1. Retrosynthesis of cytoxazone 1.
spectra of (S)- and (R)-4-phenyl-oxazolidin-2-ones. The
structure was later confirmed by the Sharpless AD-
mediated synthesis of 1 from 4-methoxycinnamate by T.
Nakata and co-workers.3 Since TH2 cells play a role in
mediating the immune response to allergens,4 we rea-
soned that cytoxazone might be a useful lead compound
for the development of therapeutic agents for atopic
dermatitis and asthma, and we designed a new synthesis
of 1 that utilized an aldol/Curtius sequence (Fig. 1).5
The brevity and convergency of our synthesis make it
ideal for analogue generation.
The syntheses of cytoxazone 1 and epi-cytoxazone 9 are
shown in Scheme 1. Acylation of the commercially
available (R)-4-phenylmethyloxazolidin-2-one with
4-methoxyphenylacetic acid provided the known imide
2.6 Subsequent reaction of the dibutylboron enolate of 2
with the commercially available benzyloxyacetaldehyde
according to the pioneering work of Evans7 provided
the aldol 3 in good syn-diastereoselectivity (>95:5, H
NMR). Removal of the chiral auxiliary from 3 using
Scheme 1. Synthesis of cytoxazone 1 and epi-cytoxazone 9. Notes:
1
PMP=para-methoxyphenyl;
Xq=(R)-4-benzyloxazolidin-2-one.
Reagents and conditions: (a) Bu2BOTf, i-Pr2EtN, À78 ꢀC, 20min;
BnOCH2CHO, À78 ꢀC to 0 ꢀC, 1.5 h; (b) 4:1 THF:H2O, H2O2, LiOH,
0 ꢀC, 1 h; NaHSO3; (c) (PhO)2PON3, PhCH3, 23 ꢀC, 40min; 110 ꢀC, 3
h; (d) 1 atm H2, Pd(OH)2, MeOH, 23 ꢀC, 24 h; (e) Bu2BOTf, i-Pr2EtN,
0 ꢀC, 30min; add BnOCH 2CHO precomplexed w/0.5 equiv SnCl4,
À78 ꢀC, 3 h; (f) (PhO)2PON3, CH2Cl2, 23 ꢀC, 40min; 45 ꢀC, 12 h.
*Corresponding author. Tel.: +1-609-252-4144; fax: +1-609-252-
7410; e-mail: percy.carter@bms.com
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00131-8

1238
P. H. Carter et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1237–1239
lithium hydroperoxide provided the acid 4 in good
yield; notably, 4 underwent retro-aldol reaction upon
silica gel chromatography, and so it was purified with a
carefully-controlled acid/base extractive workup. Acid 4
was transformed into the oxazolidinone 5 in a one-pot,
three-step (acyl azide formation, Curtius rearrangement,
isocyanate trapping) procedure that was initiated using
diphenylphosphoryl azide (toluene, 23 ꢀC) and then
completed at elevated temperature (110 ꢀC, 3 h). Ether 5
was not purified, but rather directly debenzylated using
Pearlman’s catalyst to provide cytoxazone 1 as a white
powder after silica gel chromatography. The spectral
characteristics (1H and 13C NMR, IR, HRMS, and
optical rotation) of synthetic 1 were the same as those
reported in the literature for 1.1À3 The enantiomer of 1
was prepared in identical fashion to cytoxazone itself
(Scheme 1), beginning with (S)-4-phenylmethyl-oxazoli-
din-2-one.8
Osada.1 We also examined the effects of 1 on stimulated
human peripheral blood mononuclear cells, and found
that 30 mM concentrations of 1 selectively inhibited the
production of the prototypical TH2 cytokines IL-4
(73%), IL-5 (66%) and IL-10(94%) relative to the
secretion of the TH1 cytokines IL-2 (19%) and inter-
feron-g (7%).12 To our surprise, ent-1, 9, ent-9, and 5 all
showed equivalent effects in this system (Fig. 3B). Like-
wise, ent-1, 9, ent-9, and 5 were equipotent to 1 for the
inhibition of cytokine secretion of PWM-stimulated
murine splenocytes (data not shown). To confirm that
these effects were not due to some type of toxicity, we
also examined 1, ent-1, 9, ent-9, and 5 for cytotoxicity in
the human Ramos cell line, but none of these com-
pounds were cytotoxic at concentrations up to 50 mM in
a standard 24 h assay.13 Furthermore, these compounds
were not potent anti-proliferative agents in PWM-sti-
mulated splenocytes (<15% inhibition of 3H-thymidine
incorporation at 10 mM; see Figure 3A for data
obtained with 1).11b
The synthesis of 4-epi-cytoxazone 9 required the use of
an anti-selective aldol protocol (Scheme 1). Addition of
a precomplexed solution of benzyloxyacetaldehyde and
0.5 equiv of SnCl4 to the dibutylboryl enolate of 2
according to the procedure of Heathcock9 proceeded
smoothly to give the desired aldol 6 in reasonable anti-
In summary, our work has provided a new synthetic
approach to the natural product cytoxazone and its
stereoisomers. These syntheses represent the shortest
published routes to cytoxazone14À18 and epi-cytox-
azone.16,18,19 In addition, they are the first routes that
offer a fully convergent approach to the cytoxazone
core.14b The biological data described above also
extend our understanding of 1 in several key ways: (1)
its TH2-selective inhibition of cytokine secretion is
effective on human T-cells; (2) it is neither an anti-
proliferative nor cytotoxic agent; and (3) its biological
1
diastereoselectivity (75:25, H NMR); chromatography
provided diastereomerically-pure 6 in 64% yield. Nota-
bly, no detectable product was obtained when imide 2
was used with the alternative Et2AlCl- or Mg(II)-based
anti-selective protocols.9,10 Hydrolysis of 6 (H2O2,
LiOH; NaHSO3) provided the carboxylic acid 7, which
was then transformed into oxazolidinone 8 with diphe-
nylphosphoryl azide. Due to the poor solubility of 7 in
toluene, the rearrangement was conducted in methylene
chloride, which highlights the low temperature require-
ments of this particular Curtius rearrangement (com-
plete conversion at 45 ꢀC after 12 h). Hydrogenation of
the benzyl ether 8 as before provided 4-epi-cytoxazone
9, which had spectral characteristics (1H and 13C NMR,
IR, HRMS, and optical rotation) consistent with those
reported in the literature.3 The enantiomer of 9 (5-epi-
cytoxazone) was synthesized in identical fashion, begin-
ning with (S)-4-phenylmethyloxazolidin-2-one. The
anti-disposition of the oxazolidinone ring of ent-9 was
shown via comparison to 1 in one-dimensional nuclear
Overhauser enhancement (NOE) experiments (Fig. 2).
We first confirmed the reported biological selectivity of
1 by demonstrating that synthetic 1 preferentially
blocked the secretion of IL-10relative to IL-2 in poke-
weed mitogen (PWM)-stimulated murine splenocytes
(Fig. 3A).11 The IC50 for this inhibition was approxi-
mately 10 mM, which was consistent with the work of
Figure 3. (A) The dose-dependent inhibition of IL-10production by 1
from PWM-stimulated murine splenocytes is compared with its effects
on IL-2 production and 3H-thymidine uptake (see ref 11). (B) Shown
is the selective inhibition of TH2 cytokine release from stimulated
human PBMCs by cytoxazone and its stereoisomers (see ref 12).
Figure 2. Stereochemical proof of the synthetic sequence via NOE
studies. Note that a weak NOE between the two oxazolidinone ring
protons was also observed in the trans diastereomer ent-9.

P. H. Carter et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1237–1239
1239
activity is shared by its three stereoisomers and the
O-benzylated derivative 5. These initial SAR data sug-
gest that the oxazolidinone ring of 1 — which is a rare
component of bacterial natural products1 — does not
function simply to hold the aryl and hydroxymethyl
groups in rigid, stereo-defined arrangement. They may
also indicate that the hydroxymethyl group of 1 can be
replaced with more lipophilic moieties. Our studies on
other analogues of 1 and on our attempts to define its
biological target are in progress and will be reported in
due course.
9. Walker, M. A.; Heathcock, C. H. J. Org. Chem. 1991, 56,
5747.
10. Evans, D. A.; Tedrow, J. S.; Shaw, J. T.; Downey, C. W.
J. Am. Chem. Soc. 2002, 124, 392.
11. (a) Splenocytes were isolated from 6–8 week old female
Balb/c mice (Charles River). Following lysis of red blood cells
with NH4Cl, lymphocytes were plated in 96-well plates
(Costar) at 5Â105 cells/well in RPMI (Gibco BRL) +10%
fetal calf serum (HyClone) +50 mM 2-ME. Compounds were
dissolved in DMSO and serially diluted in media to give a final
concentration of 0.3% DMSO. Compounds were added to
cells 5 min prior to stimulating the cells with pokeweed mito-
gen (5 mg/mL). Cells were cultured in a final volume of 200 mL
at 37 ꢀC for 48 h. Supernatants were then harvested and cyto-
kine levels measured by ELISA according to the manufac-
turer’s instructions (R&D Systems). (b) Alternatively, cells
were pulsed for 18 h with 3H-thymidine (1 mCi/well) and then
harvested and counted.
12. Human PBMCs were isolated by centrifugation over
Ficoll-Hypaque (Pharmacia). Cells were plated in AIM V
media (Gibco BRL) in 96-well plates (Costar) at 2Â105 cells/
well. Compounds were dissolved in DMSO and serially diluted
in media to give a final concentration of 0.3% DMSO. Com-
pounds were added to cells 5 min prior to stimulating the cells
with anti-CD3 (R&D Systems) at 1 mg/mL and anti-CD28
(Caltag Laboratories) at 200 mg/mL. Cells were cultured in a
final volume of 200 mL at 37 ꢀC for 24 h. Supernatants were
then harvested and cytokine levels measured by ELISA
according to the manufacturer’s instructions (Pharmingen).
13. Ramos cells were seeded at 3Â104 cells/well in serum free
AIM V media (Gibco BRL) in 96-well plates with various
concentrations of compound. After incubation for 24 h at
37 ꢀC, MTS (Promega) was added to each well at a final con-
centration of 333 mg/mL. Cell viability was determined after
an additional 3 h incubation at 37 ꢀC by reading the plates at
490nM absorbance. Inhibition was determined relative to cells
cultured with 1% DMSO.
Acknowledgements
We acknowledge Dr. Robert C. Newton for bringing
reference 1 to our attention, and thank Drs. Robert
Cherney and Matthew Voss for helpful discussions.
References and Notes
1. Kakeya, H.; Morishita, M.; Kobinata, K.; Osono, M.; Ish-
izuka, M.; Osada, H. J. Antibiotics 1998, 51, 1126.
2. Kakeya, H.; Morishita, M.; Koshino, H.; Morita, T.;
Kobayashi, K.; Osada, H. J. Org. Chem. 1999, 64, 1052.
3. Sakamoto, Y.; Shiraishi, A.; Seonhee, J.; Nakata, T. Tet-
rahedron Lett. 1999, 40, 4203.
4. (a) For reviews of the subject, see: Stirling, R. G.; Chung,
K. F. Eur. Respir. J. 2000, 16, 1158. (b) Renauld, J. C. J. Clin.
Pathol. 2001, 54, 577.
5. Takacs has described a conceptually identical approach as
a generalized synthesis of 4-substituted oxazolidin-2-ones. See:
Takacs, J. M.; Jaber, M. R.; Vellekoop, A. S. J. Org. Chem.
1998, 63, 2742.
6. (a) Prashad, M.; Kim, H.-Y.; Har, D.; Repic, O.; Black-
lock, T. J. Tetrahedron Lett. 1998, 39, 9369. (b) The PivCl/
LiCl/NEt3 method was also effective. See: Ho, G.-J.; Mathre,
D. J. J. Org. Chem. 1995, 60, 2271.
7. (a) Gage, J. R.; Evans, D. A. Org. Syn. 1990, 68, 77. (b)
Note that the corresponding enantioselective addition of enol-
silanes to benzyloxyacetaldehyde using chiral Cu(II) com-
plexes does not work well for phenylacetate derivatives.
See: Evans, D. A.; Kozlowski, M. C.; Murry, J. A.; Burgey,
C. S.; Campos, K. R.; Connell, B. T.; Staples, R. J. J. Am.
Chem. Soc. 1999, 121, 669.
14. (a) Carda, M.; Gonzalez, F.; Sanchez, R.; Marco, J. A.
Tetrahedron: Asymmetry 2002, 13, 1005. (b) After the com-
pletion of our work, Carda, et al. reported a variation of the
aldol/Curtius strategy towards cytoxazone 1 (but not 9) that
utilizes an unusual chiral ketone derived in five steps from
l-erthrulose.
15. Seki, M.; Mori, K. Eur. J. Org. Chem. 1999, 2965.
16. Miyata, O.; Asai, H.; Naito, T. Synlett. 1999, 1915.
17. Madhan, A.; Kumar, A. R.; Rao, B. V. Tetrahedron:
Asymmetry 2001, 12, 2009.
8. Optical rotation data confirmed the enantiomeric relation-
ship of 1 and ent-1: [a]D=À72.4ꢀ for 1 (c, 0.96, MeOH, 25 ꢀC),
[a]D=+72.2ꢀ for ent-1 (c, 0.16, MeOH, 25 ꢀC); literature
[a]D=À71ꢀ for 1 (c=0.1, MeOH, 23 ꢀC; see ref 1).
18. Hamersak, Z.; Ljubovic, E.; Mercep, M.; Mesic, M.;
Sunjic, V. Synthesis 2001, 1989.
19. Park, J. N.; Ko, S. Y.; Koh, H. Y. Tetrahedron Lett. 2000,
41, 5553.