Computationally guided organometallic chemistry: Preparation of the heptacyclic pyrazine core of ritterazine N

Article abstract of DOI:10.1021/jo052656m

Diels-Alder cycloaddition of 10 followed by Wittig homologation and intramolecular diene cyclozirconation of the resulting triene under equilibrating conditions led to the tricyclic 6-6-5 ketone 5 with high diastereocontrol. The derived α-azido ketone 16

Full text of DOI:10.1021/jo052656m

Computationally Guided Organometallic Chemistry: Preparation of
the Heptacyclic Pyrazine Core of Ritterazine N
Douglass F. Taber* and Karen V. Taluskie
Department of Chemistry and Biochemistry, UniVersity of Delaware, Newark, Delaware 19716
taberdf@udel.edu
ReceiVed December 27, 2005
Diels-Alder cycloaddition of 10 followed by Wittig homologation and intramolecular diene cyclozir-
conation of the resulting triene under equilibrating conditions led to the tricyclic 6-6-5 ketone 5 with
high diastereocontrol. The derived R-azido ketone 16 cyclized efficiently to the heptacyclic pyrazine
core of ritterazine N.
Introduction
The cephalostatins, represented by cephalostatin 1 (1),1 and
ritterazines, represented by ritterazine N (2), comprise a family
of 45 structurally unprecedented marine products with selective
cytotoxicity against human tumors. The cephalostatins, from
the Indian Ocean hemichordate Cephalodiscus gilchristi, and
the ritterazines, from the Japanese marine tunicate Ritterella
tokioka, are clearly related (ritterazine K, for example, contains2
the “left half” of cephalostatin 7), although identical alkaloids
have not yet been found in both species. Isolated from C.
gilchristi, cephalostatin 1 exhibits remarkable cytotoxic ac-
tivity3 against P388 murine leukemia cells with IC50 values of
10-4-10-6 ng/mL. Ritterazine B treated HL-60 cells become
multinucleated, leading at 20 nM to apoptosis by a novel
mechanism. Ritterazine N shows the same profile, but at near-
micromolar concentrations. It is particularly exciting that this
family of alkaloids induces apoptosis in apoptosis-resistant
malignant cell lines.
These compounds are isolated in only microgram quantities
from natural sources. If their potent physiological activity is to
be pursued, they must be prepared by synthesis. Fuchs and co-
workers have completed much elegant work toward the synthesis
of the cephalostatins and ritterazines from steroid precursors.4
(1) For the isolation and activity of the ritterazines and cephalostatins,
see: (a) Pettit, G. R.; Tan, R.; Xu, J.; Ichihara, Y.; Williams, M. D.; Boyd,
M. R. J. Nat. Prod. 1998, 61, 955. (b) Fukuzawa, S.; Matsunaga, S.;
Fusetani, N. J. Org. Chem. 1997, 62, 4484.
(2) Ganesan, A. Angew. Chem., Int. Ed. Engl. 1996, 35, 611.
(3) For the physiological activity of the cephalostatins and ritterazines,
see: (a) Komiya, T.; Fusetani, N.; Matsunaga, S.; Kubo, A.; Kaye, F. J.;
Kelley, M. J.; Tamura, K.; Yoshida, M.; Fukuoka, M.; Nakagawa, K. Cancer
Chemother. Pharm. 2003, 51, 202. (b) Dirsch, V. M.; Mueller, I. M.;
Eichhorst, S. T.; Pettit, G. R.; Kamano, Y.; Inoue, M.; Xu, J. P.; Ichihara,
Y.; Wanner, G.; Vollmar, A. M. Cancer Res. 2003, 63, 8869. (c) Mueller,
I. M.; Dirsch, V. M.; Rudy, A.; Lopez-Anton, N.; Pettit, G. R.; Vollmar,
A. M. Mol. Pharm. 2005, 67, 1684.
(4) For leading references to the synthesis of the fused 6-5 cephalostatins
and ritterazines, see: (a) Li, W.; Fuchs, P. L. Org. Lett. 2003, 5, 2849. (b)
Flessner, T.; Jautelat, R.; Scholz, U.; Winterfeldt, E. Prog. Chem. Org.
Nat. Prod. 2004, 87, 1. (c) Tietze, L. F.; Krahnert, W. R.; Chem. Eur. J.
2002, 8, 2116. No work has yet been reported toward the spiro 5-5
ritterazines.
10.1021/jo052656m CCC: $33.50 © 2006 American Chemical Society
Published on Web 03/09/2006
J. Org. Chem. 2006, 71, 2797-2801
2797

Taber and Taluskie
SCHEME 1
SCHEME 2
We report the preparation of the symmetrical heptacyclic core
of the spiro 6-6-5-5 ritterazines.
conditions of the reaction. The cyclization is readily reversible,
even though the equilibrium lies toward the cyclized product.
To predict diastereoselectivity, it was important to evaluate the
relative thermodynamic stability of the diastereomeric zircona-
cycles. ZINDO calculations indicated that zirconacycle 11a
(which would give tricyclic ketone 5, Scheme 2) was more stable
than its nearest competitor by 2.24 kcal/mol.
Results and Discussion
Retrosynthetic Analysis. We envisioned (Scheme 1) that the
symmetrical pyrazine 2 might be prepared by reductive dimer-
ization of the azido ketone 3. The ketone would be available
by convergent coupling of the sulfonate 4 with the tricyclic
ketone 5. It was particularly encouraging that computational
studies on the four diastereomeric zirconacycles derived from
6 led to the prediction that the equilibrating cyclozirconation
should proceed with substantial stereocontrol. We report the
preparation of the ketone 5, by cyclozirconation5,6 of the triene
6, and the dimerization of 5 to the heptacyclic pyrazine core of
ritterazine N 2.
SCHEME 3
Preparation of Triene 6. Diels-Alder cycloaddition (Scheme
2) of the aldehyde derived from 87 to butadiene delivered the
aldehyde 10. This set the relative configuration of the A-B
ring system. Another advantage of this approach is that it in-
stalls the ∆-2 alkene (steroid numbering) needed for pyra-
zine construction. Wittig methylenation of 10 then provided the
triene 6.
Oxygenation of the Zirconacycle. We initially ran the
cyclozirconation of 6 (Scheme 4) at room temperature. After 3
h, the intermediate zirconacycles were converted to the diols
by exposure to molecular oxygen. We recovered a dominant
kinetic product 12c from this reaction, corresponding to
intermediate zirconacycle 11c (Scheme 3). The structure of diol
12c was confirmed by X-ray analysis. A minor diol 12b was
also isolated. On reacting the triene 6 with zirconocene
dichloride and 2 equiv of butyllithium, heating for 4 h at 70
°C, and then oxygenating, we recovered a new dominant diol,
derived from the equilibrated zirconacycle 11a. This was
assigned the structure 12a.
Time Course of the Cyclozirconation. An HPLC assay was
developed for the benzoates of 11a-c. To determine the
optimum time and temperature for equilibration of the zircona-
cycles, a time and temperature study was performed (Fig-
ure 1). After 1 h at room temperature, the 13a/13b/13c ratio
was 10:25:65. After an additional 1 h at 55 °C, the ratio was
Computational Analysis. The four diastereomers of the
zirconacycle (Scheme 3) will be equilibrating5,6 under the
(5) For the development of intramolecular diene cyclozirconation,
including computational analysis, see: (a) Nugent, W. A.; Taber, D. F. J.
Am. Chem. Soc. 1989, 111, 6435. (b) Taber, D. F.; Louey, J. P.; Wang, Y.;
Nugent, W. A.; Dixon, D. A.; Harlow, R. L. J. Am. Chem. Soc. 1994, 116,
9457.
(6) For the development of Zr-based intramolecular diene cyclocarbo-
nylation, including computational analysis, see: (a) Negishi, E.; Miller, S.
R. J. Org. Chem. 1989, 54, 6014. (b) Taber, D. F.; Wang, Y. J. Am. Chem.
Soc. 1997, 119, 22. (c) Taber, D. F.; Zhang, W.; Campbell, C. L.; Rheingold,
A. R.; Incarvito, C. D. J. Am. Chem. Soc. 2000, 122, 4813. (d) Both ZINDO
and molecular mechanics were used as implemented on a Tektronix CAChe
workstation.
2798 J. Org. Chem., Vol. 71, No. 7, 2006

Computationally Guided Organometallic Chemistry
FIGURE 1. Time course of the cyclozirconation.
SCHEME 5
45:10:45. After a third hour at 70 °C, 13a was the single
dominant product.
These results indicated that the reaction is complete after 3
h total and that there was no need to heat it past 2 h at 70 °C
in order for the zirconacycles to fully equilibrate. This helped
to improve our yields, since significant degradation was
observed on prolonged heating. Using the results from the time
and temperature study, the cyclozirconation of 6 was maintained
at 70 °C for 1.5 h to equilibrate the zirconacycles. The resulting
solution was exposed6 to carbon monoxide, and a single
dominant ketone 5 was formed.
SCHEME 4
In the past, such azido ketones have been reduced to the
dimeric pyrazines using phosphines.10 These reactions did not
work well with our substrate. We were pleased to observe,
however, that reduced Te11 efficiently dimerized 16 to give 17.
This is the first preparation of the heptacyclic pyrazine core of
the 6-6-5-5 ritterazines.
Conclusion
We expect that the key Diels-Alder/Wittig/diene cyclozir-
conation approach illustrated here, supported by ZINDO
calculations, will be a general strategy for the stereocontrolled
construction of polycarbacyclic target structures. The tricyclic
ketone 5 may become a useful “truncated steroid” platform for
drug discovery.
Preparation of the Heptacyclic Pyrazine. Before proceeding
with the preparation of 4 (Scheme 1), we thought it wise to
explore the prospective azido ketone dimerization (Scheme 5).
As expected from the steroid precedent,8 epoxidation of 5
proceeded to give 14 as a single major diastereomer. Opening
of the epoxide9 with NaN3 also proceeded to give a single
product 14 from the preferred diaxial opening of the epoxide.
Oxidation then gave ketone 15.
(9) Campbell, M. M.; Craig, R. C.; Boyd, A. C.; Gilbert, I. M.; Logan,
R. T.; Redpath, J.; Roy, R. G.; Savage, D. S.; Sleigh, T. J. Chem. Soc.,
Perkin Trans. 1 1979, 2235.
(7) (a) Cahiez, C.; Alexakis, A.; Normant, J. F. Synthesis 1978, 528. (b)
Johnston, B. D.; Oehlschlager, A. C. Can. J. Chem. 1984, 62, 2148.
(8) Forcellese, M. L.; Calvitti, S.; Camerini, E.; Martucci, I.; Mincione,
E. J. Org. Chem. 1985, 50, 2191.
(10) Heathcock, C. H.; Smith, S. C. J. Org. Chem. 1994, 59, 6828.
(11) (a) Suzuki, H.; Kawaguchi, T.; Takaoka, K. Bull. Chem. Soc. Jpn.
1986, 59, 665. (b) Jeong, J. U.; Sutton, S. C.; Kim, S.; Fuchs, P. L. J. Am.
Chem. Soc. 1995, 117, 10157.
J. Org. Chem, Vol. 71, No. 7, 2006 2799

Taber and Taluskie
mixture was partitioned between EtOAc and a 1:1 mixture of 5%
aqueous H2SO4 and saturated aqueous Na2SO4. The combined
organic extract was dried (Na2SO4) and concentrated. The residue
was chromatographed to afford the ketone 5 (337 mg, 59% yield)
Experimental Section
Cyclic Aldehyde 10. To a stirred suspension of activated
manganese dioxide (123.0 g, 1.418 mol) and dry CH2Cl2 (284 mL)
was added allylic alcohol 8 neat (20.0 g, 142 mmol) dropwise over
30 min. The residual allylic alcohol was rinsed into the reaction
mixture with 5 mL of CH2Cl2. After 48 h, the mixture was filtered
through Celite using an additional 3 × 300 mL of CH2Cl2. The
organic extract was concentrated, and the aldehyde was used crude
in the next reaction: TLC Rf (PE/MTBE ) 9/1) ) 0.47; 1H NMR
(400 MHz, CDCl3) δ 9.40 (1H, s), 6.48-6.52 (1H, t, J ) 7.25),
5.77-5.87 (1H, m), 5.01-5.10 (2H, m), 2.44-2.49 (2H, m), 2.24-
2.30 (2H, m), 1.75 (3H, s); 13C NMR δ u 28.3, 32.4, 115.9, 139.7;
d 9.4, 137.1, 153.8, 195.3; IR (film) 3350, 3079, 2979, 2925, 2818,
2762, 2711, 1688, 1642, 1442, 1404, 1379, 1359, 1320, 1251, 1188,
1068, 992, 915, 859, 824, 788, 641; LRMS m/z (rel intensity) 123
(6), 109 (41), 106 (13), 95 (71), 67 (54), 55 (100), 41 (44); HRMS
calcd for C8H12O 125.096640, obsd 125.096525.
The aldehyde (7.16 g, 51.1 mmol) was dissolved in dry CH2Cl2,
and methylene blue (10 mg) was added. The reaction flask was
then cooled to -30 °C. Butadiene (21.7 mL) was condensed using
a coldfinger condensor, taken up in cold CH2Cl2 (118 mL), and
added to the reaction mixture. Triflic acid in CH2Cl2 (12 mL,
6.0 mmol) was added dropwise at -10 °C over 30 min. The reac-
tion mixture was warmed to room temperature and stirred for
an additional 5 h. The reaction mixture was partitioned between
CH2Cl2 and saturated aqueous NaHCO3. The combined organic
extract was dried (Na2SO4) and concentrated. The residue was
chromatographed to afford the aldehyde 10 (6.18 g, 68% yield) as
a colorless oil: TLC Rf (PE/MTBE ) 9/1) ) 0.63; 1H NMR (400
MHz, CDCl3) δ 9.44 (1H, s), 5.63-5.81 (3H, m), 4.95-5.04 (2H,
m), 2.17-2.34 (3H, m), 1.89-2.01 (2H, m), 1.66-1.71 (2H, m),
1.24-1.32 (2H, m), 0.99(3H, s); 13C NMR δ u 27.2, 30.2, 31.8,
32.0, 48.4, 115.2; d 14.0, 35.4, 123.6, 126.0, 138.4, 206.4; IR (film)
3076, 3027, 2976, 2921, 2835, 2689, 1728, 1640, 1454, 1433, 1189,
994, 912, 752; LRMS m/z (rel intensity) 178 (3), 163 (16), 149
(54), 145 (17), 135 (40), 121 (30), 107 (73), 93 (100), 79 (90), 67
(63), 55 (39), 41 (39); HRMS calcd for C12H18O 178.135765, obsd
178.136210.
1
as a colorless oil: TLC Rf (PE/MTBE ) 8/2) ) 0.53; H NMR
(400 MHz, CDCl3) δ 5.60-5.65 (2H, m), 2.36-2.42 (1H, dd, J )
6 Hz, J ) 17 Hz), 2.17-2.23 (1H, dd, J ) 6 Hz, J ) 17 Hz),
1.93-2.03 (4H, m), 1.75-1.82 (4H, m), 1.47-1.59 (3H, m), 1.22-
1.36 (2H, m), 0.81 (3H, s); 13C NMR δ u 28.9, 29.8, 32.1, 34.1,
39.4, 40.0, 46.3, 218.4; d 11.0, 37.5, 41.3, 54.6, 125.3, 126.3; IR
(film) 3467, 3019, 2961, 2914, 2852, 1744, 1651, 1445, 1431, 1406,
1384, 1366, 1201, 1153, 989, 912; LRMS m/z (rel intensity) 204
(61), 189 (7), 176 (6), 162 (13), 150 (41), 122 (100), 107 (27), 93
(41), 79 (30), 67 (10), 53 (9), 41 (12); HRMS calcd for C14H20O
204.151415, obsd 204.151768.
Epoxide 14. To a stirred solution of ketone 5 (37 mg, 0.18
mmol) and CH2Cl2 (1.2 mL) was added m-CPBA (43 mg, 0.25
mmol). After 30 min, the reaction mixture was partitioned between
CH2Cl2 and saturated aqueous NaHCO3. The combined organic
extract was dried (Na2SO4) and concentrated. The residue was
chromatographed to afford the epoxide 14 (34 mg, 87% yield) as
1
a colorless oil: TLC Rf (CH2Cl2/acetone ) 9.5/0.5) ) 0.43; H
NMR (400 MHz, CDCl3) δ 3.23 (1H, s), 3.14-3.16 (1H, m), 2.36-
2.39 (1H, m), 2.17-2.24 (1H, dd, J ) 7 Hz, J ) 17 Hz), 1.90-
2.02 (3H, m), 1.77-1.83 (3H, m), 1.48-1.59 (5H, m), 1.18-1.31
(2H, m), 0.83 (3H, s); 13C NMR δ u 28.6 (2), 31.9, 33.3, 38.8,
39.2, 46.1, 217.9; d 12.5, 36.2, 37.5, 50.6, 52.5, 54.2; IR (film)
3053, 2921, 2850, 1741, 1604, 1448, 1406, 1380, 1320, 1265, 1169,
1144, 984, 898, 801, 737; LRMS m/z (rel intensity) 220 (41), 205
(37), 192 (100), 187 (13), 174 (14), 162 (27), 145 (31), 122 (91),
107 (63), 93 (80), 79 (61), 67 (33), 55 (30), 41 (38); HRMS calcd
for C14H20O2 220.146330, obsd 220.145760.
Azido Alcohol 15. To a stirred solution of epoxide 14 (146.4
mg, 0.67 mmol) in an 8:1 methanol/water solution (3.35 mL)
was added NaN3 (130 mg, 2.0 mmol). After 48 h at 80 °C (bath),
the reaction mixture was partitioned between CH2Cl2 and satu-
rated aqueous NaHCO3. The combined organic extract was dried
(Na2SO4) and concentrated. The residue was chromatographed to
afford the azido alcohol 15 (111 mg, 63% yield) as a colorless oil:
1
Triene 6. To methyltriphenylphosphonium bromide (17.98 g,
50.33 mmol) and potassium tert-butoxide (5.28 g, 47.19 mmol)
was added THF (125 mL). The reaction mixture was stirred at room
temperature for 30 min, and then aldehyde 10 (5.60 g, 31.46 mmol)
in THF (10 mL) was added dropwise over 15 min. After an
additional 30 min, the reaction mixture was partitioned between
diethyl ether and, sequentially, saturated aqueous NH4Cl and brine.
The combined organic extract was dried (Na2SO4) and concentrated.
The residue was chromatographed to afford the triene 6 (5.09 g,
92% yield) as a colorless oil: TLC Rf (PE/ CH2Cl2 ) 9.5/0.5) )
TLC Rf (CH2Cl2/acetone ) 9.5/0.5) ) 0.20; H NMR (400 MHz,
CDCl3) δ 3.96 (1H, s), 3.81-3.83 (1H, m), 2.36-2.42 (1H, dd, J
) 7 Hz, J ) 17 Hz), 2.21-2.27 (1H, dd, J ) 7 Hz, J ) 17 Hz),
1.90-2.06 (3H, m), 1.79-1.89 (4H, m), 1.67-1.72 (2H, m), 1.44-
1.51 (1H, m), 1.40-1.42 (3H, m), 1.28-1.37 (1H, m), 1.07 (3H,
s); 13C NMR δ u 28.1, 31.4, 32.1, 35.2, 36.9, 39.4, 46.1, 218.4; d
12.0, 36.9, 38.5, 55.1, 61.3, 68.5; IR (film) 3429, 2921, 2102, 1738,
1447, 1254, 1169, 1025, 902.
Azido Ketone 16. To a stirred solution of the azido alcohol 15
(71 mg, 0.27 mmol) in CH2Cl2 (1.8 mL) at room temperature was
added Dess-Martin periodinane (228 mg, 0.54 mmol). After an
additional 20 min, the reaction mixture was partitioned between
CH2Cl2 and a 1:1 mixture of saturated aqueous Na2S2O3 and
saturated aqueous NaHCO3. The organic extract was dried
(Na2SO4) and concentrated. The residue was adsorbed onto silica
gel and chromatographed to afford the azido ketone 16 (44 mg,
63% yield) as a colorless oil: TLC Rf (CH2Cl2/acetone ) 9.5/0.5)
) 0.62; 1H NMR (400 MHz, CDCl3) δ 4.08-4.11 (1H, m), 2.41-
2.55 (3H, m), 2.24-2.30 (1H, m), 1.81-2.11 (6H, m), 1.45-1.62
(3H, m), 1.20-1.29 (1H, m) 1.16 (1H, m), 1.04 (3H, s); 13C NMR
δ u 28.6, 31.4, 35.1, 39.5, 41.6, 43.3, 206.2, 216.8; d 12.8, 37.0,
43.5, 45.7, 54.7, 63.7; IR (film) 2922, 2102, 1740, 1444, 1404,
1261, 1187, 1147, 1101, 1015, 746; HRMS calcd for C14H20O2N3
(M + H) 262.155552, obsd 262.156570.
1
0.83; H NMR (400 MHz, CDCl3) δ 5.66-5.86 (4H, m), 4.95-
5.06 (4H, m), 2.08-2.28 (3H, m), 1.92-1.98 (1H, m), 1.67-1.76
(2H, m), 1.59-1.63 (1H, m), 1.46-1.50 (1H, m), 1.06-1.12 (1H,
m), 0.96 (3H, s); 13C NMR δ u 28.4, 30.0, 32.2, 38.7, 39.1, 111.3,
114.5; d 16.6, 40.1, 125.4, 125.9, 139.3, 149.0; IR (film) 3079,
3023, 2974, 2918, 2880, 2833, 1823, 1639, 1432, 1413, 1373, 1350,
1327, 1225, 1188, 1001, 909, 875, 783; LRMS m/z (rel intensity)
176 (<1), 161 (6), 147 (6), 134 (13), 122 (21), 107 (33), 93 (43),
81 (100), 77 (17), 67 (16), 55 (10), 41 (14); HRMS calcd for C13H20
176.156501, obsd 176.155848.
Ketone 5. Zirconocene dichloride (992 mg, 3.4 mmol) was added
to a 50 mL round-bottomed flask. Triene 6 (500 mg, 2.8 mmol) in
toluene (8 mL) was added to the flask. The reaction mixture was
cooled to -78 °C, and n-BuLi was added (2.17 M in hexanes, 3.10
mL) via syringe. The reaction was warmed to room temperature
over 15 min and then was heated to 70 °C for 1.5 h. The reaction
was cooled to room temperature, and then 8 mL of THF was added.
CO was bubbled through the reaction mixture for 45 min. The
reaction was stirred under a CO atmosphere for 24 h. Glacial acetic
acid (1.18 mL) was added to the reaction mixture. The reaction
Pyrazine 17. Sodium hydrogen telluride was prepared by heating
powdered tellurium (75 mg, 0.59 mmol) and NaBH4 (52 mg, 1.38
mmol) to 75 °C in ethanol (2.2 mL) for 1 h. To the resulting dark
red solution was added azido ketone 16 (59.4 mg, 0.23 mmol) in
ethanol (0.6 mL) with stirring. The color instantly turned black.
The mixture was stirred open to the air overnight. The ethanol was
2800 J. Org. Chem., Vol. 71, No. 7, 2006

Computationally Guided Organometallic Chemistry
removed under high vacuum, and then the reaction mixture was
adsorbed directly onto silica gel and chromatographed to afford
the heptacyclic pyrazine core 17 (41 mg, 84% yield) as a pale
yellow solid: mp 185 °C dec; TLC Rf (CH2Cl2/acetone ) 9.5/0.5)
) 0.34; 1H NMR (400 MHz, CDCl3) δ 2.83-2.84 (2H, m)
2.61-2.75 (6H, m), 2.40-2.46 (2H, dd, J ) 7 Hz, J ) 17 Hz),
2.28-2.34 (2H, dd, J ) 7 Hz, J ) 17 Hz), 1.78-2.11 (12 H, m),
1.67-1.70 (2H, m), 1.44-1.50 (2H, m), 1.21-1.35 (2H, m), 0.84
(6H, s); 13C NMR δ u 28.8, 32.1, 35.3 (2), 39.8, 46.4 (2), 148.7,
149.0, 218.0; d 11.6, 37.5, 41.9, 54.4; IR (film) 2920, 2854, 1741,
1446, 1398, 1158, 1016, 914, 733; HRMS calcd for C28H37N2O2
433.285504, obsd 433.284413.
Acknowledgment. We thank the NIH (GM42056) for
financial support of this work and J. V. Mitten for preliminary
experiments. This study is dedicated to E. I. Negishi and his
students, pioneers of reductive cyclozirconation.
Supporting Information Available: General experimental
details, preparation of 8 and 12a-c, HPLC details, and spectra for
all new compounds. This material is available free of charge via
the Internet at http://pubs.acs.org.
JO052656M
J. Org. Chem, Vol. 71, No. 7, 2006 2801