A highly stereoselective approach to novel 2,2,4-trisubstituted pyrrolidines by halocylization: Key intermediates towards syntheses of nitrogen analogs of Noxafil

Article abstract of DOI:10.1016/j.tetlet.2003.08.114

A convenient synthesis of novel 2,2,4-trisubstituted pyrrolidines via stereoselective 5-exo iodocyclization of the 2,2-disubstituted olefins 8 and 9 is described. Single crystal X-ray analysis of the oxazolidinone 15 confirmed the relative stereochemistry in this cyclization. As observed earlier in the formation of 2,2,4-trisubstituted tetrahydrofurans, the presence of a bulky aryl substituent on the olefin appears to direct the stereochemical outcome of these halocyclizations.

Full text of DOI:10.1016/j.tetlet.2003.08.114

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 7997–8000
A highly stereoselective approach to novel 2,2,4-trisubstituted
pyrrolidines by halocylization: key intermediates towards
syntheses of nitrogen analogs of Noxafil®
Jinping L. McCormick,a Rebecca Osterman,a Tze-Ming Chan,a Pradip R. Das,a
Birendra N. Pramanik,a Ashit K. Ganguly,a Viyyoor M. Girijavallabhan,a Andrew T. McPhailb and
Anil K. Saksena*
aSchering-Plough Research Institute, 2015 Galloping Hill Road, Kenilworth, NJ 07033, USA
bPaul M. Gross Chemical Laboratories, Duke University, Durham, NC 27706, USA
Received 26 June 2003; revised 26 August 2003; accepted 27 August 2003
Abstract—A convenient synthesis of novel 2,2,4-trisubstituted pyrrolidines via stereoselective 5-exo iodocyclization of the
2,2-disubstituted olefins 8 and 9 is described. Single crystal X-ray analysis of the oxazolidinone 15 confirmed the relative
stereochemistry in this cyclization. As observed earlier in the formation of 2,2,4-trisubstituted tetrahydrofurans, the presence of
a bulky aryl substituent on the olefin appears to direct the stereochemical outcome of these halocyclizations.
© 2003 Elsevier Ltd. All rights reserved.
Noxafil® is a novel 2,2,4-trisubstituted tetrahydrofuran
based antifungal agent currently undergoing late-stage
phase III clinical trials. It has shown improved thera-
peutic potential over existing drugs against a variety of
invasive fungal infections in normal and immunocom-
promised patients refractory to or intolerant of stan-
dard therapy.1 Halocyclization provides an effective
methodology for stereoselective construction of substi-
tuted tetrahydrofurans. Recently we described the
enantioselective synthesis of (−)-2R-cis-tosylate 1,
which acted as a central intermediate towards Noxafil®
and a variety of highly active analogs.2 Synthesis of 1
was accomplished in a highly stereoselective manner via
a novel 2,4-diastereoselective halocyclization of an opti-
cally active 2,2-disubstituted d-olefin.2c
ment of the 2,2,4-trisubstituted tetrahydrofuran core
with a pyrrolidine moiety. Examples dealing with
stereoselective electrophilic cyclization of unsaturated
nitrogen compounds from acyclic precursors are rela-
tively few compared to the analogous oxygen cases.3,4
Mercury or selenium induced intramolecular cycliza-
tions of amido or amino precursors have been reported
to provide mainly 2,5-disubstituted pyrrolidines with
good stereocontrol.5 An elegant stereoselective
approach to 2,5-trans or 2,5-cis pyrrolidines, via overall
5-endo-trig iodocyclization of homoallylic tosylamides,
was recently described.6 A careful search of the litera-
ture led us to conclude that no example existed of
stereoselective halocyclizations leading to 2,2,4-trisub-
stituted pyrrolidines such as 10–13.7,8
A logical extension would be to prepare the nitrogen
analog of this important therapeutic agent by replace-
The olefinic homochiral diol 2, readily available from a
previous study,2c was converted to the benzylidine
* Corresponding author. Fax: +1-908-740-7152; e-mail: anil.saksena@spcorp.com
0040-4039/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2003.08.114

7998
J. L. McCormick et al. / Tetrahedron Letters 44 (2003) 7997–8000
acetal 3. Reduction of 3 with diisobutylaluminiun
hydride (DIBAL) provided the monobenzyl alcohol 4 in
94% yield over two steps. Attempted direct displacement
of the hydroxyl group in 4 with diphenylphosphorylazide
(DPPA) failed to provide the azide 6. However, conver-
sion of the alcohol 3 to the mesylate 5 followed by
displacement with sodium azide provided the desired
azide 6 in 50% yield over two steps. Lithium aluminium
hydride (LAH) reduction of 6 and treatment of the
resulting amine 7 with di-tert-butyl dicarbonate or p-tol-
uene sulfonyl chloride provided the N-Boc and N-tosyl
protected amines 89 and 910 in 89 and 92% yield,
respectively (Scheme 1).
cis:trans ratio of the cyclized product from 8 was superior
(10:11, >95:5; 88% yield) compared to the cyclized
products from 9 (12:13, 83:17; 72% yield). This bias in
favor of the cis isomers did not change in the absence of
base (Scheme 2).
Since 10/11 existed as rotamer mixture, removal of the
N-Boc group with trifluoroacetic acid was employed to
give the amine salt 14 from which the cis:trans isomer
ratio could be accurately determined by NMR. NOE
experiments on 1413 showed independently that the
iodomethyl and benzyloxymethyl groups are cis to each
other. In contrast to the corresponding tetra-
hydrofurans,2 not surprisingly direct displacement of
iodine even by the highly nucleophilic triazolyl anion
with either N-Boc or N-tosyl protected pyrrolidines
proved challenging. Thus treatment of 10 with Na-1,2,4-
triazole at elevated temperature, afforded the crystalline
oxazolidone 15,14 mp 78–80°C (Scheme 3). Treatment of
14 with aqueous NaHCO3 in a one-pot sequence pro-
vided the aziridine 16 as a yellow oil (ꢀ90% yield).
Iodocyclization of the protected amino alcohols 8 and 9,
under irreversible conditions proceeded in excellent
yields and 2,2,4-diastereoselectivity to provide 10/1111
and 12/1312 as cis:trans mixtures. The major isomers in
both cases were cis with respect to the iodomethyl and
benzyloxymethyl functionalities. Presumably due to the
greater steric bulk imposed by the t-butyl group, the
Scheme 1. Reagents and conditions: (i) PhCH(OMe)2, PPTS, DMF, rt; (ii) DIBAL, benzene–hexane, rt; (iii) MsCl, Et3N, CH2Cl2,
0–5°C; (iv) NaN3, DMF, 80°C; (v) LAH, EtOEt, rt; (vi) Boc2O, CH2Cl2, rt; (vii) TsCl, pyridine, CH2Cl2, 0–5°C.
Scheme 2.
Scheme 3. Reagents and conditions: (i) Na-triazole, DMF, 60°C, 24 h; (ii) TFA, CH2Cl2, rt; aq. NaHCO3.

J. L. McCormick et al. / Tetrahedron Letters 44 (2003) 7997–8000
7999
The crystal structure of 15 was solved by direct meth-
References
ods (MULTAN 11/82)15 (Fig. 1). The asymmetric unit
consists of two molecules. Approximate coordinates for
the non-hydrogen atoms were derived in part from an
E-map and from a series of weighted Fo and difference
Fourier syntheses phased successively by an increasing
number of atoms. Positional and thermal parameters of
these atoms (first isotopic and then anisotropic) were
adjusted by means of several rounds of full-matrix
least-squares calculations. Hydrogen atoms were incor-
porated at their calculated positions, and an extinction
correction was included as a variable during the later
cycles. A final difference Fourier synthesis contained no
unusual features.
1. Everett, N. Antifungal Drug Discovery and Development—
Fifth International Summit, 28–29 March 2001, Princeton,
NJ, USA, IDrugs 2001, 4, 652–655.
2. (a) Saksena, A. K.; Girijavallabhan, V. M.; Lovey, R. G.;
Bennett, F.; Pike, R. E.; Wang, H.; Pinto, P.; Liu, Y. T.;
Patel, N.; Ganguly, A. K. 35th Interscience Conference on
Antimicrobial Agents and Chemotherapy ( ICAAC), San
Francisco, 17–20 September1995 (abstract nos. F 61 and
F 83); (b) Saksena, A. K.; Girijavallabhan, V. M.; Lovey,
R. G.; Pike, R. E.; Wang, H.; Liu, Y.-T.; Pinto, P.;
Bennett, F.; Jao, E.; Patel, N.; Desai, J. A.; Rane, D. F.;
Cooper, A. B.; Ganguly, A. K. Anti-infectives, Recent
Advances in Chemistry and Structure–Activity Relation-
ships; The Royal Society of Chemistry, Special Publica-
tion No. 198, 1997; (c) Saksena, A. K.; Girijavallabhan,
V. M.; Lovey, R. G.; Bennett, F.; Pike, R. E.; Wang, H.;
Ganguly, A. K.; Morgan, B.; Zaks, A.; Puar, M. Tetra-
hedron Lett. 1995, 36, 1787–1790.
The observed 2,4-distereoselectivity in the present
examples is consistent with the rationale that we pro-
posed ealier for the stereoselective formation of 2,2,4-
trisubstituted tetrahydrofurans.2c In
a
chairlike
transition state first proposed by Harding and Burks16
for the amidomercuration of olefinic amides, equatorial
disposition of the 4-aminomethyl substituent as
depicted in [A] may account for the major 10 and 12
type products. A minor pathway such as [B] will have
unfavorable steric repulsion between the 2-aryl and the
axially disposed 4-hydroxymethyl substituents.
3. For reviews, see: (a) Bartlett, P. A. Asymmetric Synthesis;
Academic Press: New York, 1984; Vol. 3, p. 411; (b)
Cardillo, G.; Orena, M. Tetrahedron 1990, 46, 3321.
4. Rychnovsky, S. D.; Bartlett, P. A. J. Am. Chem. Soc.
1981, 103, 3963.
5. (a) Tokuda, M.; Yamada, Y.; Suginome, H. Chem. Lett.
1988, 1289–1290; (b) Tamaru, Y.; Kawamura, S.; Bando,
T.; Tanaka, K.; Hojo, M.; Yoshida, Z. J. Org. Chem.
1988, 53, 5491–5501.
In conclusion we have demonstrated a simple and
highly stereoselective synthesis of 2,2,4-trisubstituted
pyrrolidines 10 and 12. These and derived key interme-
diates 15 and 16 offer further opportunities for a vari-
ety of selective transformations possibly leading up to
analogs of Noxafil® and other biologically interesting
compounds.
6. Jones, A. D.; Knight, D. W. Chem. Commun. 1996, 915.
7. Procedure for the synthesis of 10: An acetonitrile solution
of 8 was cooled to 0°C and treated with pyridine (42 mL,
2.1 equiv.) followed by iodine (126 mg, 2.0 equiv.). The
solution was shielded from light and stirred at rt
overnight (16 h) TLC indicated that a small amount
starting material was left. The reaction mixture was
treated with Na2S2O3 (aq.) solution until the purple color
faded. The mixture was diluted with EtOAc and washed
with water and brine. The crude product was purified on
silica gel column using 10% Ethyl acetate/hexane as
eluent to afford the cyclized product 10 (118 mg, 88%,
clear oil). The NMR experiments showed that 10 existed
in rotamers. Removal of the Boc group resulted in pre-
dominantly the cis product from which the ratio of
cis/trans isomers was determined.
1
8. All new compounds were characterized by H, 13C, NMR
and high resolution mass spectra. When necessary 1D
NOE and 2D NOESY NMR spectra were obtained to
confirm relative stereochemistry. Elemental analysis were
obtained for crystalline compounds only. Yields refer to
isolated products and have not been optimized. Selective
spectral data is given here.
Figure 1.

8000
J. L. McCormick et al. / Tetrahedron Letters 44 (2003) 7997–8000
9. 8: FAB-MS m/z 418 (M+H), 362 (M−Bu+H), 318 (M−
(d, J=10.7 Hz, 1H), 4.38 (d, J=10.7 Hz, 1H), 4.50 (s,
2H), 6.57 ddd, J=2.6, 8.5, 10.7 Hz, 1H), 6.90 (ddt,
J=1.0, 2.7, 8.7 Hz, 1H), 7.19 (d, J=8.1 Hz, 2H), 7.31 (m,
1H), 7.45 (d, J=8.3 Hz, 2H), 7.75 (dt, J=6.2, 9.1 Hz,
1H).
1
Boc, 100%). H NMR [300 MHz, CDCl3] l 1.42 (s, 9H),
1.79 (m, 1H), 2.50 (m, 2H), 3.12 (m, 1H), 3.25 (m, 1H),
3.37 (dd, J=4.2, 9.2 Hz, 1H), 3.42 (dd, J=4.5, 9.4 Hz,
1H), 4.44 (s, 2H), 4.95 (bs, 1H), 5.18 (s, 1H), 5.22 (s, 1H),
4.79 (m, 2H), 7.30 (m, 6H).
13. 14: FAB-MS m/z 444 (M+H), 316 (M−I, 100%). 1H
NMR [300 MHz, CDCl3] l 2.41 (dd, 1H), 2.76 (m, 1H),
2.95 (dd, 1H), 3.55 (m, 4H), 3.90 (AB, 2H), 4.55 (AB,
2H), 6.98 (m, 1H), 7.35 (m, 6H).
10. 9: FAB-MS m/z 472 (M+H, 100%), 380 (M−CH2Ph, 366
1
(M−OCH2Ph), 316 (M−Ts). H NMR [300 MHz, CDCl3]
l 1.75 (m, 1H), 2.42 (s, 3H), 2.44 (d, J=11 Hz, 2H), 2.90
(m, 1H), 3.05 (m, 1H), 3.28 (dd, J=6.0, 8.3 Hz, 1H), 3.41
(dd, J=3.8, 9.3 Hz, 1H), 4.39 (s, 2H), 5.14 (s, 2H), 5.17
(s, 1H), 5.24 (t, J=6.0 Hz, 1H), 6.78 (m, 2H), 7.10 (m,
1H), 7.30 (m, 7H), 7.65 (d, J=9.0 Hz, 2H).
14. 15: FAB-MS m/z 360 (M+H, 100%). 1H NMR [400
MHz, CDCl3] l 1.89 (dt, J=1.8, 12.1 Hz, 1H), 2.44 (m,
2H), 3.44 (d, J=5.8 Hz, 2H), 3.50 (dd, J=8.5, 11.7 Hz,
1H), 3.64 (dd, J=6.9, 11.7 Hz, 1H), 4.27 (d, J=9.2 Hz,
1H), 4.49 (s, 2H), 4.68 (dd, J=2.3, 9.1 Hz, 1H), 6.88 (m,
2H), 7.30 (m, 6H).
11. 10: CI-MS m/z 544 (M+H), 488 (M−Bu+H), 444 (M−
1
Boc, 100%), 360 (M−I). H NMR [300 MHz, CDCl3] l
15. Atomic parameters, bond lengths, bond angles and tor-
sion angles for 15 have been deposited at the Cambridge
Crystallographic Data Center, Cambridge, CB2 1EZ,
UK. Any request should be accompanied by full litera-
ture citation for this communication.
1.37 (s), 1.40 (s), 2.24 (m), 2.65 (m), 3.08 (m), 3.25 (m),
3.47 (m), 3.80 (m), 4.07 (d, J=10 Hz), 4.22 (m), 4.50 (s),
4.54 (d, J=10 Hz), 6.78 (m), 7.10 (m), 7.30 (m).
12. 12: CI-MS 598 (M+H), 490 (M−BnOH), 470 (M−I), 456
(M−CH2I), 442 (M−Ts), 380 (M−Bn−I), 315 (M−Ts−I).
cis-Isomer: 1H NMR [400 MHz, CDCl3] l 2.4 (s, 3H),
2.65 (m, 1H), 3.45 (m, 2H), 3.94 (t, J=8.7 Hz, 1H), 4.22
16. Harding, K. E.; Burks, S. R. J. Org. Chem. 1981, 46,
3920–3922.