Lactone pathway to statins utilizing the wittig reaction. the synthesis of rosuvastatin

Article abstract of DOI:10.1021/jo101050z

The first entry to statins via lactonized side chain is reported, exemplified by the synthesis of rosuvastatin. The key step is Wittig coupling of (2S,4R)-4-(tert-butyldimethylsilyloxy)-6-oxotetrahydro-2H-pyran-2-carbaldehyde and phosphonium salt of an appropriately functionalized pyrimidine heterocy'le. One-pot deprotection and hydrolysis of the resulting 4-O-TBS rosuvastatin lactone provided rosuvastatin in high yield.

Full text of DOI:10.1021/jo101050z

pubs.acs.org/joc
superstatins are currently marketed including fluvastatin,7
Lactone Pathway to Statins Utilizing the Wittig
Reaction. The Synthesis of Rosuvastatin
pitavastatin,8 atorvastatin,9 and rosuvastatin.10 The value of
rosuvastatin is growing with new approved indications in phase
III studies, which raises it high on the list of synthetic targets.11
,†
Zdenko Casar,* Miha Steinbucher, and Janez Kosmrlj*
‡
,‡
ꢀ
€
ꢀ
†Lek Pharmaceuticals, d.d., Sandoz Development Center
Slovenia, API Development, Organic Synthesis Department,
‡
Kolodvorska 27, 1234 Menges, Slovenia, and Faculty of
ꢀ
Chemistry and Chemical Technology, University of Ljubljana,
SI-1000 Ljubljana, Slovenia
zdenko.casar@sandoz.com; janez.kosmrlj@fkkt.uni-lj.si
Received May 30, 2010
FIGURE 1. Left: general formula of superstatins (Het = heterocycle).
Right: rosuvastatin calcium.
Over the last three decades, diverse strategies to the synthetic
statins have been developed including those utilizing the Wittig
reaction.5 Unfortunately, many of these are long, multistep
linear reaction sequences, often involving late and unselective
functional group transformations and protection/deprotection
steps with difficult purification procedures. Tedious prepara-
tion of chiral precursors, several synthetic steps, and a need of
cryo-chemistry render these approaches less attractive (see
Supporting Information for detailed discussion).7,8,10,12-17
Interestingly, the most obvious formyl functionalized
lactonized side chain precursor 1 (Scheme 1) has not yet
(8) For selected literature, see: (a) Suzuki, M.; Iwasaki, H.; Fujikawa, Y.;
Kitahara, M.; Sakashita, M.; Sakoda, R. Bioorg. Med. Chem. 2001, 9, 2727–
2743. (b) Hayashi, T.; Yokote, K.; Saito, Y.; Iguchi, A. Expert Opin.
Pharmacother. 2007, 8, 2315–2327.
(9) For selected literature, see: Roth, B. D.; Blankley, C. J.; Chucholowski,
A. W.; Ferguson, E.; Hoefle, M. L.; Ortwine, D. F.; Newton, R. S.; Sekerke,
C. S.; Sliskovic, D. R.; Stratton, C. D.; Wilson, M. J. Med. Chem. 1991, 34,
357–366.
The first entry to statins via lactonized side chain is reported,
exemplified by the synthesis of rosuvastatin. The key step is
Wittig coupling of (2S,4R)-4-(tert-butyldimethylsilyloxy)-
6-oxotetrahydro-2H-pyran-2-carbaldehyde and phosphonium
salt of an appropriately functionalized pyrimidine heterocycle.
One-pot deprotection and hydrolysis of the resulting 4-O-
TBS rosuvastatin lactone provided rosuvastatin in high yield.
(10) For selected literature, see: Watanabe, M.; Koike, H.; Ishiba, T.;
Okada, T.; Sea, S.; Hirai, K. Bioorg. Med. Chem. 1997, 5, 437–444.
(11) Forrecentmethods for the synthesis of rosuvastatin, see: (a) Balanov, A.;
Shenkar, N.; Niddam-Hildesheim, V. U.S. Patent Appl. US 2007167625 A1,
2007; Chem. Abstr. 2007, 147, 166109. (b) Radl, S.; Stach, J.; Klvana, R.; Jirman,
J. PCT Int. Appl. WO 2007000121 A1, 2007; Chem. Abstr. 2007, 146, 121983.
ꢀ
(c) Casar, Z. PCT Int. Appl. WO 2007039287 A1 2007; Chem. Abstr. 2007, 146,
421841. (d) Patel, D. J.; Kumar, R.; Dwivedi, S. P. D. WO 2007099561 A1, 2007;
Chem. Abstr. 2007, 147, 322770.
Statins are among the most commonly prescribed drugs
worldwide for the treatment of lipid disorders.1,2 Initially,
they were discovered as fungal metabolites and were com-
posed of a chiral decaline core attached to a β-hydroxy
lactone moiety.3,4 Chemical evolution to structurally refined
molecules led to semisynthetic analogues4 and later on to fully
synthetic derivatives with the decaline core being replaced by
heteroaromatic motifs (superstatins, Figure 1).5 In these mole-
cules, the β-hydroxy lactone moiety generally remained unmo-
dified as it is essential for the biological activity.6 Several
(12) (a) Jendralla, H.; Baader, E.; Bartmann, W.; Beck, G.; Bergmann,
A.; Granzer, E.; Von Kerekjarto, B.; Kesseler, K.; Kraue, R.; Schubert, W.;
Wess, G. J. Med. Chem. 1990, 33, 61–70. (b) Sliskovic, D. R.; Picard, J. A.;
Roark, W. H.; Roth, B. D.; Ferguson, E.; Krause, B. R.; Newton, R. S.;
Sekerke, C.; Shaw, M. K. J. Med. Chem. 1991, 34, 367–373. (c) Tempkin, O.;
ꢀ
Abel, S.; Chen, C.-P.; Underwood, R.; Prasad, K.; Chen, K.-M.; Repic, O.;
Blacklock, T. J. Tetrahedron 1997, 53, 10659–10670.
ꢀ
(13) Zlicar, M. PCT Int. Appl. WO 2007017117, 2007; Chem. Abstr. 2007,
ꢀ
146, 251655.
(14) Beck, G.; Kesseler, K.; Baader, E.; Bartmann, W.; Bergmann, A.;
Granzer, E.; Jendralla, H.; Von Kerekjarto, B.; Krause, R.; Paulus, E.;
Schubert, W.; Wess, G. J. Med. Chem. 1990, 33, 52–60.
(15) (a) Wess, G.; Kesseler, K.; Baader, E.; Bartmann, W.; Beck, G.;
Bergmann, A.; Jendralla, H.; Bock, K.; Holzstein, G.; Kleine., H.; Schnierer,
M. Tetrahedron Lett. 1990, 31, 2545–2548. (b) Hiyama, T.; Minami, T.;
Takahashi, K. Bull. Chem. Soc. Jpn. 1995, 68, 364–372. (c) Koike, H.;
Kabaki, M.; Taylor, N. P.; Diorazio, L. J. PCT Int. Appl. WO 2000049014
A1, 2000; Chem. Abstr. 2000, 133, 193161. (d) Bratton, L. D.; Auerbach, B.;
Choi, C.; Dillon, L.; Hanselman, J. C.; Larsen, S. D.; Lu, G.; Olsen, K.;
Pfefferkorn, J. A.; Robertson, A.; Sekerke, C.; Trivedi, B. K.; Unangst, P. C.
Bioorg. Med. Chem. 2007, 15, 5576–5589. (e) Cai, Z.; Zhou, W.; Sun, L.
Bioorg. Med. Chem. 2007, 15, 7809–7829. (f) Pfefferkorn, J. A.; Bowles,
D. M.; Kissel, W.; Boyles, D. C.; Choi, C.; Larsen, S. D.; Song, Y.; Sun,
K.-L.; Miller, S. R.; Trivedi, B. K. Tetrahedron 2007, 63, 8124–8134.
(16) (a) Yang, Y. L.; Falck, J. R. Tetrahedron Lett. 1982, 23, 4305–4308.
(b) Paquette, L. A.; Oplinger, J. A. J. Org. Chem. 1988, 53, 2953–2959.
(1) (a) Tobert, J. A. Nat. Rev. Drug Discovery 2003, 2, 517–526. (b) Istvan,
E. S.; Deisenhofer, J. Science 2001, 292, 1160–1164.
(2) Thiebaud, P.; Patel, B. V.; Nichol, M. B. Health Econ. 2008, 17, 83–97.
(3) Endo, A.; Kuroda, M.; Tsujita, Y. J. Antibiot. 1976, 29, 1346–1348.
(4) Endo, A.; Hasumi, K. Nat. Prod. Rep. 1993, 10, 541–550.
ꢀ
(5) Casar, Z. Curr. Org. Chem. 2010, 14, 816–845.
(6) Sit, S. Y.; Parker, R. A.; Motoc, I.; Han, W.; Balasubramanian, N.;
Catt, J. D.; Brown, P. J.; Harte, W. E.; Thompson, M. D.; Wright, J. J.
J. Med. Chem. 1990, 33, 2982–2999.
(7) For selected literature, see: Fuenfschilling, P. C.; Pascale, H.; Mutz,
J.-P. Org. Process Res. Dev. 2007, 11, 13-18 and references therein.
DOI: 10.1021/jo101050z
r
Published on Web 09/03/2010
J. Org. Chem. 2010, 75, 6681–6684 6681
2010 American Chemical Society

ꢀ
Casar et al.
JOCNote
SCHEME 1. Lactone Pathway to Rosuvastatin Calcium
SCHEME 3. Synthesis of TBS-Protected Rosuvastatin Lactone E-6
SCHEME 2. Wittig Experiments with 2 and Model Aldehyde 4
To identify optimal reaction conditions for the ylide 3
formation, we decided to trap this intermediate for analytical
purposes in a form of a Wittig reaction product. To minimize
the complexity of the trapping reaction, which could arise
at this stage from the fragile δ-valerolactone part of 1, we
decided to utilize a non-epimerizable, non-lactone/ester
model aldehyde, (()-2-ethylhexanal (4, Scheme 2). After
careful addition of NaHDMS (1.0-1.1 equiv) into a toluene
solution of 2 (1 mmol) at -35 °C, followed by the introduc-
tion of aldehyde 4 (1 equiv) at -45 °C, only a sluggish fade of
the orange-red color of ylide 3 was observed, indicating low
reactivity of the Wittig components (Supporting Informa-
tion, Table S1, entries 1-3). In subsequent experiments, this
led us to gradually increase the temperatures of both steps,
the ylide formation and the olefination. Optimal reaction
conditions were identified by conducting both steps at room
temperature with olefin E-5 formed exclusively and isolated
in yields of up to 80% (Table S1, entries 4 and 5). No isomeric
Z-5 could be detected by NMR of the crude reaction
product. Similarly, n-BuLi, LiHDMS, and NaH were exam-
ined for the ylide 3 formation (Table S1, entries 6-10). In
all instances, E-5 isomer was formed exclusively albeit in
slightly lower yields than by using NaHDMS. n-BuLi caused
a partial removal of N-methylsulfonyl group from E-5,
resulting in byproduct E-5a (Table S1, entries 7 and 8).
With these results, we next performed the study with
aldehyde 1 (Scheme 3; for “Het”, see Figure 1). As reported
previously,20 the optimal way for its preparation and storage
is in the form of hydrate 10, which after being dissolved in an
appropriate solvent, equilibrates to aldehyde 1 liberating
an equivalent of water. On the basis of the relatively low
reactivity of ylide 3, we surmised that an equimolar amount
of water, liberated from 10, might not be detrimental to the
Wittig reaction. Hence, a toluene solution of 10 was initially
used in the reactions with equimolar ylide 3. From the results
shown in Table 1, entry 1, one can see that the formation of
the desired E-6 was accompanied by the isomeric Z-6,
phosphonium salt 2, and methylpyrimidine 7.24 If 2 equiv
of 3 was employed, only phosphonium salt 2 and pyrimidine
7 were detected in the reaction mixture (entry 2). The
appearance of both 2 and 7 can be ascribed to the reaction
of 3 with either water or even hydrate 10, which could
potentially remain in the solution.25 This was confirmed by
been applied for the construction of statins. This is because
all previous attempts in the preparation of 1 failed. It was
even suggested that 1 is not synthesizable because of the
β-substituted δ-lactone ring instability.18 Followed by our
recent success in the preparation of 1,19,20 now accessible via
an efficient, green, and highly stereoselective enzymatic
approach,21 here we report its application to the synthesis
of rosuvastatin. A highly convergent approach is presented,
taking advantage of the Wittig reaction between a methyl-
phosphonium-substituted heterocycle 2 and aldehyde 1 and
subsequent one-pot deprotection/hydrolysis step (Scheme 1).
Our study began by the selection of phosphonium salt 2
and optimization of its ylide formation. Although substitu-
ents R at the phosphorus atom and the counterion X- may
have influence on the course of the Wittig reaction,22,23 these
also greatly affect the solubility of the salt, which is an
important factor in a potential industrial application. Out
of several combinations of R and X-, tributyl phosphonium
- 13
trifluoromethanesulfonate 2 (R = n-Bu, X- = CF3CO2
)
was chosen on the basis of its relatively high solubility in a
solvent as apolar as toluene.
ꢀ
(17) (a) Lee, G. T.; Linder, J.; Chen, K.-M.; Prasad, K.; Repic, O.;
Hardtmann, G. E. Synlett 1990, 508–508. (b) Prasad, K.; Chen, K.-M.;
ꢀ
Repic, O.; Hardtmann, G. E. Tetrahedron: Asymmetry 1990, 1, 307–310.
(c) Minami, T.; Takahashi, K.; Hiyama, T. Tetrahedron Lett. 1993, 34, 513–
516. (d) Beck, G.; Jendralla, H.; Kesseler, K. Synthesis 1995, 1014–1018.
(e) Sun, F.; Xu, G.; Wu, J.; Yang, L. Tetrahedron: Asymmetry 2007, 18, 2454–
2461. (f) Choi, H.; Shin, H. Synlett 2008, 1523–1525.
(18) (a) Cardani, S.; Scolastico, C.; Villa, R. Tetrahedron 1990, 46, 7283–
€
7288. (b) Tararov, V. I.; Andrushko, N.; Andrushko, V.; Konig, G.;
Spannenberg, A.; Borner, A. Eur. J. Org. Chem. 2006, 5543–5550.
€
ꢀ
(19) Casar, Z. Synlett 2008, 2036–2040.
ꢀ
(20) Casar, Z.; Kosmrlj, J. Synlett 2009, 1144–1148.
ꢀ
ꢀ
ꢀ
(21) Cluzeau, J.; Casar, Z.; Mrak, P.; Oslaj, M.; Kopitar, G. PCT Int.
Appl. WO 2009092702 A2, 2009; Chem. Abstr. 2009, 151, 218948.
(22) Hon, Y.-S.; Lee, C.-F. Tetrahedron 2000, 56, 7893–7902.
(23) Fu, Y.; Wang, H.-J.; Chong, S.-S.; Guo, Q.-X.; Liu, L. J. Org. Chem.
2009, 74, 810–819.
(24) Crystal structure of 7: Zhou, X.-L.; Li, J.-S.; Wang, W.; Zhang,
Y.-Y.; Fan, X.-P. Acta Crystallogr. 2007, E63, o1293–o1294. Preparation,
spectral and analytical data of 7 have previously not been reported.
(25) Bestmann, H. J. Angew. Chem., Int. Ed. Engl. 1965, 4, 583–587.
6682 J. Org. Chem. Vol. 75, No. 19, 2010

ꢀ
Casar et al.
JOCNote
TABLE 1. Wittig Reaction of In Situ Generated Ylide 3 with Aldehyde 1a
SCHEME 5. TBS Deprotection of E-6
yield (%)c
entry
T,b °C
ylide 3(equiv)
E-6
33
0
39
25
49
57
84 (62)e
Z-6
2
7
8
1d
2d
3
4
5
20
20
20
0
40
65
110
1
2
1
1
1
1
1
4
0
5
10
7
52
47
33
2
10
19
4
11
53
10
28
12
8
0
0
13
35
22
10
0
6
7
6
7 (5)e
5
TABLE 2. Selected Results at Desilylation of E-6
aAverage data from several consecutive runs are given. Dry toluene
solution of 1 was used (see text), unless otherwise noted. bTemperature
for the olefination step. For details, see the Experimental Section.
cDetermined as ratio by 1H NMR analysis of the crude products.
dToluene solution of 10 was used. eIsolated pure product.
entry
reagent (equiv)
solvent
T, °C
time, h
producta
1
2
3
CEC (0.05)
AcCl (0.15)
MeOH
MeOH
THF
25
25
25
0.8
48
24
E-10 (71)b
E-10 (100)
E-9 (99)
n-Bu4NF (3.4)c
aPercent yield, determined by 1H NMR analysis of the crude reaction
mixtures with 1,3,5-trimethoxybenzene as an internal standard for
SCHEME 4. Ylide 3 Mediated TBSOH Elimination from E-6
c
integration. bStarting E-6 (7%) was also detected. In the presence of
AcOH (5.5 equiv).
110 °C are shown in Table 1, entries 4-7. At elevated temp-
eratures, the formation of E-8 was greatly suppressed. This
suggests different temperature profiles of the two compet-
ing reactions of ylide 3: as an olefination partner with
aldehyde 1, and as a base in TBSOH elimination at E-6. Optimal
results were seen when toluene solution of aldehyde 1 was added
to ylide 3 (1 equiv) at 110 °C. In the crude reaction product, the
desired E-6and the isomeric Z-6were detected by 1HNMRin84
and 7% yield, respectively, and after chromatographic purifica-
tion isolated in 62 and 5% yield, respectively (entry 7). No E-8
could be detected in the reaction mixture.
an independent experiment in which ylide 3, generated from
2 and 3-fold excess of NaHDMS, was treated with a 10-fold
excess of water. Products 2 (44%) and 7 (56%) were con-
firmed by 1H NMR analysis of the crude reaction mixture.
Adopting the above findings, we carried out experiments
with a dry toluene solution of aldehyde 1, (see “Preparation
of Aldehyde 1” in the Experimental Section), and a decrease
in the formation of byproduct 2, Z-6, and 7 was observed. In
this case, however, TBSOH elimination product E-8 (13%)
appeared (Scheme 3, Table 1, entry 3).
Concomitant formation of 7 in nearly the same amounts
as 8 suggested that ylide 3 as a base23 could cause TBSOH
elimination from E-6. This was confirmed by an independent
experiment in which E-6 was treated with equimolar ylide 3 at
room temperature for 15 min (Scheme 4). Complete consump-
tion of the starting compounds was observed, and on the basis of
1H NMR analysis of the reaction mixture, compounds 7 (40%)
and E-8 (20%) were formed along with several unidentified
byproducts. We did not investigate further whether the ylide 3
mediated TBSOH elimination also occurs at aldehyde 1 to give
(S)-6-oxo-3,6-dihydro-2H-pyran-2-carbaldehyde, which would
after Wittig reaction afford E-8, as well. It is known that R,β-
dehydro-δ-formyl-δ-valerolactones undergo Wittig reaction.26
Interestingly, base (NaH, DBU)-promoted TBSOH elimination
has been reported to occur at β-TBSO-substituted cyclo-
hexanones,27 whereas at the valerolactones, it has only been
achieved by the action of p-toluenesulfonic acid.28
Several methods were tested for desilylation of E-6. By
employing HCl generated in situ either from 1-chloroethyl
chloroformate (CEC)/MeOH29 or from acetyl chloride and
MeOH,30 TLC analyses of the reaction mixtures indicated
very clean transformations, but instead of the desired alcohol
E-9, these afforded methyl ether E-10 (Scheme 5, Table 2).
NMR spectral analysis of E-10 revealed two sets of signals in
the ratio of 4:6, indicating partial epimerization at C-5. The
mechanism of this ester ether formation may proceed by
acid-catalyzed transesterification at the lactone ring of either
E-6 or E-9 with methanol, followed by SN1 substitution of
the liberated C-5 hydroxyl group with methanol via allylic
carbocation intermediate. Similar lactone ring opening has
been documented at a δ-vinyl-δ-valerolactone,31 and acid-
catalyzed allylic alcohol etherification has been reviewed.32
In our case, we have provided the evidence of the above sug-
gested mechanism by the experiment in which rosuvastatin
methyl ester10 was in the presence of catalytic amounts of
HCl incubated in methanol for 19 h. A mixture of diaster-
eomeric ethers E-10 (authentic to the above-mentioned) was
produced in quantitative yield. Attempts at desilylation of
35
Different reaction conditions for the Wittig olefination
of 1 were examined to minimize the formation of E-8. Experi-
ments conducted at different temperatures ranging from 0 to
E-6 mediated by TMSCl/KF 2H2O,33 FeCl3,34 or CuCl2
3
gave the desired compound E-9, but in unsatisfactory yields
(Table S2). On the other hand, nearly quantitative conversion
(29) Yeom, C.-E.; Kim, Y. J.; Lee, S. Y.; Shin, Y. J.; Kim, B. M.
Tetrahedron 2005, 61, 12227–12237.
(26) (a) Hansen, T. V. Tetrahedron: Asymmetry 2002, 13, 547–550.
(b) Boger, D. L.; Ichikawa, S.; Zhong, W. J. Am. Chem. Soc. 2001, 123,
4161–4167. (c) Tsubuki, M.; Kanai, K.; Honda, T. Heterocycles 1993, 35,
281–288. (d) O’Connor, B.; Just, G. Tetrahedron Lett. 1986, 27, 5201–5202.
(27) (a) Hareau, G. P.-J.; Koiwa, M.; Hikichi, S.; Sato, F. J. Am. Chem.
Soc. 1999, 121, 3640–3650. (b) Jeffery, D. W.; Perkins, M. V. Tetrahedron
Lett. 2004, 45, 8667–8671.
(30) Khan, A. T.; Mondal, E. Synlett 2003, 694–698.
(31) Behr, A.; Juszak, K. D. Z. Chem. 1985, 25, 220–221.
(32) DeWolfe, R. H.; Young, W. G. Chem. Rev. 1956, 56, 753–901.
(33) Peng, Y.; Li, W.-D. Z. Synlett 2006, 1165–1168.
(34) Yang, Y.-Q.; Cui, J.-R.; Zhu, L.-G.; Sun, Y.-P.; Wu, Y. Synlett 2006,
1260–1262.
(28) Sato, M.; Nakashima, H.; Hanada, K.; Hayashi, M.; Honzumi, M.;
Taniguchi, T.; Ogasawara, K. Tetrahedron Lett. 2001, 42, 2833–2837.
(35) Tan, Z. P.; Wang, L.; Wang, J. B. Chin. Chem. Lett. 2000, 11, 753–
756.
J. Org. Chem. Vol. 75, No. 19, 2010 6683

ꢀ
Casar et al.
JOCNote
SCHEME 6. Synthesis of Rosuvastatin Calcium from E-6
(EtOAc/hexanes 3:5) to give E-6 (350 mg, 62%) and Z-6 (30 mg,
5%) as white glassy solids after removal of volatiles.
(E-6): 1H NMR (300 MHz, CDCl3) δ 0.07 (s, 3H, CH3Si), 0.09
(s, 3H, CH3Si), 0.89 (s, 9H, (CH3)3C), 1.27 (d, J = 6.7 Hz, 6H,
(CH3)2C), 1.51-1.83 (m, 2H, H-5), 2.53-2.68 (m, 2H, H-3),
3.34 (septet, J = 6.7 Hz, 1H, CH3CH), 3.51 (s, 3H, CH3SO2),
3.58 (s, 3H, CH3N), 4.23-4.33 (m, 1H, H-4), 5.12-5.25 (m, 1H,
H-6), 5.49 (dd, J = 16.1, 5.8 Hz, 1H, dCH-(C-6)), 6.70 (dd, J =
16.1, 1.3 Hz, 1H, dCH-(C-50)), 7.03-7.17 (m, 2H, H-300, H-500),
7.55-7.70 (m, 2H, H-200, H-600); 13C NMR (75 MHz, CDCl3)
δ = -4.90, -4.82 (2 ꢀ CH3Si), 18.0 ((CH3)3C), 21.6, 21.7 (2 ꢀ
CH3C), 25.7 ((CH3)3C), 32.3 (CH3CH), 33.1 (CH3N), 36.4
(C-5), 39.4 (C-3), 42.5 (CH3SO2), 63.3 (C-4), 75.4 (C-6), 115.1
(d, J = 21.7 Hz, C-300, C-500), 120.6 (C-50), 125.5 (dCH-(C-500)0),
132.1 (d, J = 8.4 Hz, C-200, C-600), 134.3 (d, J = 3.3 Hz, C-1 ),
134.8 (dCH-(C-6)), 157.6 (C-20), 163.3 (d, J = 250 Hz, C-400),
163.7 (C-40), 169.6 (C-2), 175.0 (C-60). Anal. Calcd for
C28H40FN3O5SSi: C, 58.21; H, 6.98; N, 7.27. Found: C, 58.25;
H, 7.17; N, 7.24.
into E-9 was achieved by using n-Bu4NF/AcOH/THF36
(Table 2, entry 3).
Having established this highly efficient protocol for TBS
deprotection and expecting that lactone hydrolysis with an
appropriate base should proceed without difficulties, we
decided to probe the preparation of rosuvastatin calcium
in a one-pot reaction starting from E-6 (Scheme 6). Thus,
after the TBS deprotection into E-9 was complete, the
resulting reaction mixture was made alkaline with aqueous
NaOH to give a solution of rosuvastatin sodium. From this
solution, pure amorphous rosuvastatin calcium10 was pre-
cipitated with Ca(OAc)2 in 88% overall yield from E-6.
In conclusion, we have demonstrated a practical synthesis
of rosuvastatin. The key step of our approach is highly
selective Wittig olefination between the appropriate hetero-
cyclic ylide and β-TBSO functionalized δ-formyl-δ-valero-
lactone into 4-O-TBS-protected rosuvastatin lactone.
Deprotection, lactone hydrolysis, and cation exchange reaction
into rosuvastatin calcium has been achieved in one-pot reaction
in excellent overall yield and without any detectable epimeriza-
tion. Owing to the recent advances in remarkably efficient
synthesis of β-TBSO-δ-formyl-δ-valerolactone19,20 and highly
stereoselective enzymatic preparation of its acetyloxy lactone
precursor,21 this convergent and cryo-chemistry-free approach
is superior to other methods5 for the preparation of rosuvasta-
tin. Moreover, it seems unlikely that this chemistry is restricted
to rosuvastatin, selected herein. Instead, we believe it opens new
perspective in efficient, economically and ecologically sustain-
able statin synthesis in general.
Synthesis of Rosuvastatin Calcium. To a stirred solution of
n-Bu4NF 3H2O (3.66 g, 11.6 mmol) and AcOH (1.05 mL) in dry
3
THF (60 mL) was added E-6 (3.80 g, 6.6 mmol). The reaction
mixture was stirred at 25 °C for 48 h. Then the mixture was
concentrated under reduced pressure to give an oily residue,
which was diluted with EtOAc (50 mL). The obtained solution
was washed successively with water (25 mL), saturated NaH-
CO3 solution (25 mL), brine (25 mL), and water (25 mL),
followed by removal of EtOAc under reduced pressure to give
an oily residue. The residue was dissolved in a mixture of THF/
H2O (4:1, 60 mL), warmed to 30 °C, and NaOH (8.0 M, 0.92 mL,
0.73 mmol) was added. The reaction mixture was stirred at 30 °C
for 2 h. Then THF was evaporated completely under reduced
pressure, and water was added to give 30 mL of clear rosuvas-
tatin sodium solution. The resulting solution was washed with
EtOAc (2 ꢀ 12 mL), and water layer was distilled under reduced
pressure to remove dissolved EtOAc. The resulting clear solu-
tion of rosuvastatinate sodium was diluted with H2O to 30 mL
of total volume, warmed to 40 °C, and Ca(OAc)2 H2O (0.88 g,
3
5.0 mmol in 6 mL of H2O) was added under vigorous stirring at
40 °C over 5 min. The reaction mixture was stirred for 30 min
at 40 °C. The precipitate was filtered off, washed with water
(5 mL), and resuspended in H2O (20 mL), followed by vigorous
stirring for 1 h at room temperature. The product was collected
by filtration and dried in vacuo to give pure rosuvastatin
calcium (2.88 g, 88%) with specific rotation and spectroscopic
properties in accordance with the literature data.10
Experimental Section
Synthesis of E-6. (a). Preparation of Aldehyde 1: A solution of
1020 (271 mg, 0.98 mmol) in dry CH2Cl2 (30 mL) was stirred at
ambient temperature for 23 h. Volatiles were removed by atmo-
spheric evaporation at 40 °C, and the residue (1, 253 mg, 0.98
mmol) was dissolved in dry toluene (30 mL).
Acknowledgment. The authors gratefully acknowledge
ꢀ
Mr. P. Drnovsek for support of this project, Dr. Bogdan
Kralj and Dr. Dusan Zigon (Mass Spectrometry Center,
(b). Wittig Reaction: A toluene solution of NaHDMS (1.85
mL of 0.6 M, 1.11 mmol) was added at room temperature to a
stirred suspension of 213 (638 mg, 0.98 mmol) in dry toluene
(29 mL). The reaction mixture was within 15 min warmed to
110 °C, and the above prepared toluene solution of aldehyde 1
(0.98 mmol), kept at room temperature, was added under
vigorous stirring. The reaction mixture was stirred for 5 min
and chilled to room temperature. Volatiles were removed in
vacuo, and the oily residue was subjected to chromatography
ꢀ
ꢀ
ꢀ
Jozef Stefan Institute, Ljubljana, Slovenia) for mass spectro-
metry measurements, Mr. S. Borisek for acquiring some of
ꢀ
ꢀ
the NMR spectra, Mrs. D. Jereb, Mrs. J. Anzel (Dipl. Ing.),
and Miss A. Veskovic for technical assistance.
ꢁ
Supporting Information Available: Detailed discussion on
synthetic strategies to statins, experimental procedures, Tables
S1 and S2, and copies of NMR spectra. This material is available
free of charge via the Internet at http://pubs.acs.org.
(36) Hoffman, W. F.; Alberts, A. W.; Anderson, P. S.; Chen, J. S.; Smith,
R. L.; Willard, A. K. J. Med. Chem. 1986, 29, 849–852.
6684 J. Org. Chem. Vol. 75, No. 19, 2010