Diastereoselective addition of chlorotitanium enolates of N-acyl thiazolidinethione to activated imines - A novel synthesis of β-lactams

Article abstract of DOI:10.1139/V06-179

We report a novel methodology for preparing enantiomerically pure β-lactams, starting from nitriles in diastereomeric ratios up to 10:1. The power of the methodology was demonstrated by the efficient synthesis of the cholesterol absorption inhibitor SCH 4

Full text of DOI:10.1139/V06-179

1696
COMMUNICATION / COMMUNICATION
Diastereoselective addition of chlorotitanium
enolates of N-acyl thiazolidinethione to activated
imines — A novel synthesis of ␤-lactams1
Marike Herold-Dublin and Dennis C. Liotta
Abstract: We report a novel methodology for preparing enantiomerically pure β-lactams, starting from nitriles in
diastereomeric ratios up to 10:1. The power of the methodology was demonstrated by the efficient synthesis of the
cholesterol absorption inhibitor SCH 48462.
Key words: β-lactam, N-metalloaldimine, titanium Lewis acids, cholesterol absorption inhibitor.
Résumé : Nous rapportons une nouvelle méthode de préparation des β-lactames énantiomériquement pures à partir de
nitriles avec des rapports diastereomériques allant jusqu’à 10 : 1. L’efficacité de la méthode a été démontrée par la
synthèse de l’inhibiteur d’absorption du cholestérol SCH 48462.
Mots clés : β-lactame, N-metalloaldimine, acides de Lewis à base de titane, inhibiteur d’absorption du cholestérol.
[Traduit par la Rédaction] Herold-Dublin and Liotta 1699
Scheme 1. Diastereoselective synthesis of azetine derivatives,
starting from nitriles.
Introduction
S
The biological importance of β-lactams has been clearly
demonstrated over the last five decades, particularly regard-
ing the central role they play as antibiotics (1). The search
continues for new β-lactams that exhibit a broader spectrum
of antibacterial activity and a better resistance profile. Fur-
thermore, the relevance of β-lactams in many important non-
antibiotic uses, such as cholesterol absorption inhibitors (2)
or a variety of enzyme inhibitors (3–6), continues to expand
at a surprising rate.
O
N
S
3
TiCl4,
(–)-Sparteine
S
Ph
N
O
S
N
Ph-CN
S
TiCl4
Ph
N
+
+
Cp2ZrHCl
N
75%
Recently, our research group reported a diastereoselective,
one-pot procedure for preparing novel azetine derivatives (1)
from readily accessible nitriles (Scheme 1) (7). The diastereo-
selective additions of chlorotitanium enolates of N-propionyl-
thiazolidine-2-thione (3) to various metalloaldimines, derived
from the hydrometallation of the corresponding nitriles, resulted
in the preferred formation of azetine 1 over tetrahydropyri-
midinone derivatives (2) for both enolizable and non-enolizable
substrates. Here, we describe an efficient, diastereoselective
one-pot procedure for preparing β-lactams starting from nitriles.
1:2 = 17:1
1
2
Results and discussion
In the synthesis of azetine derivative 1, a β-lactam side-
product could be detected in a low percent. The formation of
a β-lactam can mechanistically be explained as follows
(Scheme 2).
Received 8 December 2005. Accepted 6 October 2006. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on
5 January 2007.
M. Herold-Dublin, D.C. Liotta.2 Department of Chemistry, Emory University, 1521 Dickey Drive, Atlanta, GA 30322, USA.
1This article is part of a Special Issue dedicated to Dr. Alfred Bader.
2Corresponding author (e-mail: dliotta@emory.edu).
Can. J. Chem. 84: 1696–1699 (2006)
doi:10.1139/V06-179
© 2006 NRC Canada

Herold-Dublin and Liotta
1697
Table 1. Formation of β-lactam in the presence of titanium Lewis acids.
O
S
N
S
3
TiCl4,
O
O
(–)-Sparteine
HN
8a
Lewis acid/
TiCl3Cp*
R
14 h, 24 oC
R
C N
+
Cp2ZrHCl
CH2Cl2
+
2 h, 24 oC
1 h, 0 o
C
HN
8b
R
Entry
Nitrile (R)
Lewis acid
Yield (%)a
anti:syn Ratiob
70
89
1
2
Ph
Ph
TiCl3(OiPr)
TiCl2(OiPr)2
4:1
2:1
3
4
5
6
7
8
9
Ph
Ph
Ph
TiCl(OiPr)3
80
1:2
7:1
7:1
7:1
6:1
1:3
8:1
TiCl3Cp*
<50b
TiCl3Cp*/TiCl3(OiPr)
TiCl3Cp*/TiCl2(OiPr)2
TiCl3Cp*/TiCl2(OiPr)2
TiCl3Cp*/TiCl2(OiPr)2
TiCl3Cp*/TiCl3(OiPr)
81
p-MeO–Ph
p-Me–Ph
p-CF3–Ph
2-Thienyl
76
70
74
65
aIsolated material after silica gel chromatography.
1
bDetermined by H NMR analysis of the crude reaction mixture; ~5% error of the stated values.
zwitterion (5) undergoes ring closure by nitrogen attack at
the carbonyl group (8).3 When the oxygen in intermediate
(6) coordinates with a second titanium Lewis acid, it be-
comes the preferred leaving group, leading to the formation
of the azetine 1 (path A). Alternatively, Lewis acid coordi-
nation with the chiral auxiliary converts it to a reasonable
leaving group, resulting in the production of the β-lactam (8)
(path B). In principle, redirecting the reaction to favor the
lactam over the azetine could be achieved by employing a
Lewis acid that is more thiaphilic than oxaphilic (i.e., one
that would preferably coordinate to the thiocarbonyl group
of the chiral auxiliary). In this way, the thiazolidine-2-thione
could be activated as a leaving group, thereby leading to
lactam formation.
Scheme 2. Proposed mechanism for the formation of lactam 8
and azetine 1.
TiCl3
S
M
N
O
S
Ph
N
N
S
Ph
N
S
1
6
Path A
Cl3
Ti
Cl3
Ti
M
M
H
O
S
S
N
O
N
+
N
Ph
N
S
Ph
S
4
In our initial investigation, we chose the thiaphilic Lewis
acid, titanium trichloroisopropoxide, and were gratified to
find that a mixture of syn and anti β-lactams was formed in a
70% yield (Table 1, entry 1). Each of these isomers was pro-
duced in an enantiomerically pure form as confirmed by
chiral HPLC studies. The absolute stereochemistry of the β-
lactams was elucidated using X-ray crystallography. The
yields (but not the diastereomeric ratios) could be improved
by increasing the isopropoxide content of the Lewis acid
(Table 1, entries 2 and 3). To increase the diastereo-
selectivity of the lactam formation, a screening of other
titanium Lewis acids was conducted. Although the use of
pentamethylcyclopentadienyl titanium trichloride increased
the ratio of the lactam diastereoselectivity (Table 1, entry 4),
M = metal
5
Path B
Cl3Ti
TiClx(OiPr)y
H
N
M
N
O
S
S
Ph
O
Ph
N
8
7
The nitrile is converted in situ to the N-metalloaldimine
species (4), thereby activating it for attack by the titanium
enolate. After the C–C bond formation, the resulting
3 Consistent with literature results from both Crimmins laboratory (see ref. 8) and our own, we have never observed any stereoinduction re-
sulting from the use of sparteine in titanium addition reactions. We use it primarily because we obtain better chemical yields relative to
other tertiaty amines, e.g. triethylamine and Huenig’s base.
© 2006 NRC Canada

1698
Can. J. Chem. Vol. 84, 2006
Scheme 3. Proposed transition-states for the formation of
(a) azetine and (b) lactam.
Scheme 4. Synthesis of the cholesterol inhibitor SCH 48462.
O
S
S
S
N
Ph
S
(b)
(a)
O
9
N
HN
Ph
N
TiCl4,
(–)-Sparteine
H
Ph
8a
N
1
MeO
O
TiCl3Cp*/
TiCl2(OiPr)2
MeO
CN
Clx
Ti
+
ClxTi
Ph
1 h, 0 oC
14 h, 24 o
C
2 h, 24 oC,
CH2Cl2
Cp2ZrHCl
10
O
O
H
Ph
71%
S
Ph
Me
Me
N
N
N
H
S
anti:syn = 10:1
N
ClxTi
(OiPr)y
ClxTi
S
OMe
S
TiClx(OiPr)y
MeO
I
N
O
MeO
K3PO4, Cat. Cul,
dihydropyrimidinone 2 was also produced in significant
quantities (8:2 ratio was 2:1). Because we observed high
lactam yields and only traces of dihydropyrimidinone
(Table 1, entries 1–3), we speculated that the 6-membered
ring formation might be suppressed by adding titanium
trichlorodiisopropoxide to the reaction mixture. Indeed, a
combination of the Lewis acids, pentamethylcyclopenta-
dienyl titanium trichloride and titanium trichloride
isopropoxide, resulted in an efficient formation of the mix-
ture of lactams in a good diastereomeric ratio with nearly
complete suppression of dihydropyrimidinone (Table 1, en-
try 5). The process works well with other non-enolizable
nitriles (Table 1, entries 6–9) but resulted in poor yields (0–
20%) when attempted with several enolizable nitriles. The
diastereoselectivity is highly dependent on the electronic
properties of the nitriles, favoring syn-lactam fromation for
nitriles with electron withdrawing substituents (Table 1, en-
try 8).
Interestingly, the absolute stereochemistries observed in
the formation of anti-azetine derivatives are opposite to
those observed in the anti-lactams (Scheme 3). To explain
the difference in the stereochemical outcome, we propose
the competing transition-states shown in Scheme 3 as our
models. In the azetine formation, we postulate an open tran-
sition-state that minimizes dipole interactions (9) between
the (Z)-titanium enolate (10–12) and the titanium imine.4 As
a consequence, the thiazolidinethione blocks the re side of
the enolate, and the imine must approach from the si side.
Diastereomeric transition-states presumably create larger di-
pole and steric interactions. In case of the lactam formation;
however, the coordination of an additional thiaphilic Lewis
acid to the thiocarbonyl group of the chiral auxiliary elimi-
nates chelation, thereby allowing it to rotate 180° to mini-
mize dipole effects of the thiocarbonyl group and the
carbonyl group of the enolate. As a consequence, the
thiazolidinethione blocks the si side of the enolate and the
imine has to attack from the re side. The syn-lactam is pre-
sumably formed by attack on the other face of the imine.
To probe the scope of this newly developed methodology,
we applied it to the synthesis of the cholesterol absorption
inhibitor SCH 48462 (Scheme 4) (2, 13–15). Using the com-
Cat. diamine
dioxane, 23 h, 100 oC
Ph
SCH 48462
89%
bination of titanium dichlorodiisopropoxide and pentamethyl
cyclopentadienyl titanium trichloride, the reaction proceeded
in 71% yield and gave improved diastereoselectivity (an anti
to syn ratio of 10:1). In the second step of the synthesis, the
β-lactam 10 was readily N-arylated (89% yield) by using
copper catalysis as described by Klapars et al. (16) (63%
overall yield for the two-pot process).
Conclusions
In summary, we have developed a novel methodology for
the preparation of enantiomerically pure β-lactams, starting
from nitriles in diastereomeric ratios up to 10:1. The power
of the methodology was demonstrated by the efficient syn-
thesis of the cholesterol absorption inhibitor SCH 48462.
Experimental section
General procedure for ␤-lactam formation
The nitrile (1.72 mmol) and Cp2ZrHCl (445.00 mg,
1.72 mmol) were dissolved in dry dichloromethane (2 mL)
in a 50-mL flask and stirred for 2 h under argon at room
temperature. The respective titanium Lewis acid (3.45 mmol)
was added, and the reaction mixture was stirred for 1 h at
0 °C. During this time, titanium enolate of compound 3 or 9
was generated by dissolving 3 or 9 (1.15 mmol) in dry di-
chloromethane (5 mL) at 0 °C under argon, then dropwise
addition of titanium tetrachloride (0.13 mL, 1.15 mmol),
stirring for 5 min, and the addition of (–)-sparteine
(0.66 mL, 2.87 mmol). The brown solution was stirred
40 min at 0 °C and then added to the first reaction mixture
via a cannula. The reaction mixture was stirred for 14 h at
room temperature. It was then quenched with 20 mL satu-
rated aq. ammonium chloride solution with continuous stir-
ring, and dichloromethane (30 mL) and distilled water
4 We considered and rejected Zimmerman–Traxler-like transition-states primarily because the absolute stereochemistries of the newly formed
stereogenic centers would have been derived from an unfavorable trans diaxial substituent pattern.
© 2006 NRC Canada

Herold-Dublin and Liotta
1699
5. A.D. Borthwick, G. Weingarten, T.M. Haley, M. Tomaszewski,
W. Wang, Z. Hu, J. Bedard, H. Jin, L. Yuen, and T.S. Mansour.
Bioorg. Med. Chem. Lett. 8, 365 (1998).
6. W.W. Ogilvie, C. Yoakim, F. Do, B. Hache, L. Lagace, J.
Naud, J.A. O’Meara, and R. Deziel. Bioorg. Med. Chem. 7,
1521 (1999).
(30 mL) were added. The organic phase was separated, and
the aq. phase extracted with dichloromethane (2 × 30 mL).
The combined organic phases were washed with saturated
aq. sodium bicarbonate solution, then with brine, dried with
anhyd. magnesium sulfate, and finally concentrated under
vacuum. The product mixture was purified by silica gel flash
chromatography (EtOAc:hexanes was 1:10 – 1:3).
7. N.B. Ambhaikar, M. Herold, and D.C. Liotta. Heterocycles,
62, 217 (2004).
For spectroscopic and physical data of 8a, 8b, 9, 10, and
8. (a) M.T. Crimmins and K. Chaudhary. Org. Lett. 2, 775
(2000); (b) M.T. Crimmins, B.W. King, E.A. Tabet, and K.
Chaudhary. J. Org. Chem. 66, 894 (2001); (c) M.T. Crimmins,
B.W. King, and E.A. Tabet. J. Am. Chem. Soc. 119, 7883
(1997).)
SCH 48462, refer to the supplementary data.5
Acknowledgments
We would like to thank Dr. Kenneth Hardcastle (Depart-
ment of Chemistry, Emory University) for obtaining the X-
ray structures.
9. H. Danda, M.M. Hansen, and C.H. Heathcock. J. Org. Chem.
55, 173 (1990).
10. D.A. Evans, J. Bartroli, and T.L. Shih. J. Am. Chem. Soc. 103,
2127 (1981).
11. D.A. Evans, F. Urpi, T.C. Somers, J.S. Clark, and M.T.
Bilodeau. J. Am. Chem. Soc. 112, 8215 (1990).
12. W.A. Kleschick, C.T. Buse, and C.H. Heathcock. J. Am.
Chem. Soc. 99, 247 (1977).
13. D.A. Burnett. Tetrahedron Lett. 35, 7339 (1994).
14. D.A. Burnett, M.A. Caplen, H.R. Davis, R.E. Burrier, and J.W.
Clader. J. Med. Chem. 37, 1733 (1994).
15.J.W. Clader, D.A. Burnett, M.A. Caplen, M.S. Domalski, S.
Dugar, W. Vaccaro, R.R. Sher, M.E. Browne, H.R. Zhao, R.E.
Burrier, B. Salisbury, and H.R. Davis. J. Med. Chem. 39, 3684
(1996).
16. A. Klapars, X.H. Huang, and S.L. Buchwald. J. Am. Chem.
Soc. 124, 7421 (2002).
References
1. G. Burton, N.F. Osborne, M.J. Pearson, and R. Southgate. In
Burger’s medicinal chemistry and drug discovery. Vol. 4.
Edited by M.E. Wolff. Wiley & Sons, New York. 1995. p. 277.
2. D.A. Burnett. Curr. Med. Chem. 11, 1873 (2004).
3. R.M. Adlington, J.E. Baldwin, B.N. Chen, S.L. Cooper, W.
McCoul, G.J. Pritchard, T.J. Howe, G.W. Becker, R.B.
Hermann, A.M. McNulty, and B.L. Neubauer. Bioorg. Med.
Chem. Lett. 7, 1689 (1997).
4. W.T. Han, A.K. Trehan, J.J.K. Wright, M.E. Federici, S.M.
Seiler, and N.A. Meanwell. Bioorg. Med. Chem. 3, 1123
(1995).
5 Supplementary data for this article are available on the journal Web site or may be purchased from the Depository of Unpublished Data,
Document Delivery, CISTI, National Research Council Canada, Ottawa, ON K1A 0S2, Canada. DUD 5116. For more information on ob-
taining material, refer to http://cisti-icist.nrc-cnrc.gc.ca/irm/unpub_e.shtml. CCDC 628790 and 628791 contain the crystallographic data for
this manuscript. These data can be obtained, free of charge, via http://www.ccdc.cam.ac.uk/conts/retrieving.html (or from the Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax +44 1223 336033; or deposit@ccdc.cam.ac.uk).
© 2006 NRC Canada