Tuning Rate of the Bergman Cyclization of Benzannelated Enediynes with Ortho Substituents

Article abstract of DOI:10.1021/ol0255054

matrix presented The Bergman cyclization of benzannelated enediynes is highly sensitive to ortho substitution. This finding opens possibilities for the rational design of conformer-specific and pH-dependent DNA-cleaving agents.

Full text of DOI:10.1021/ol0255054

ORGANIC
LETTERS
2002
Vol. 4, No. 7
1119-1122
Tuning Rate of the Bergman Cyclization
of Benzannelated Enediynes with Ortho
Substituents
Igor V. Alabugin,* Mariappan Manoharan, and Serguei V. Kovalenko
Department of Chemistry and Biochemistry, Florida State UniVersity,
Tallahassee, Florida 32306-4390
alabugin@chem.fsu.edu
Received January 2, 2002
ABSTRACT
The Bergman cyclization of benzannelated enediynes is highly sensitive to ortho substitution. This finding opens possibilities for the rational
design of conformer-specific and pH-dependent DNA-cleaving agents.
The thermal cyclization of (Z)-enediynes (the Bergman
cyclization)1 has a number of practical applications in the
design of DNA-cleaving agents,2 the development of poly-
meric materials with enhanced thermal properties,3and
synthesis of polycyclic compounds.4 Introduction of func-
tional groups in benzannelated enediynes can provide a
straightforward and synthetically viable way to control both
the cyclization rate and molecular recognition of the enediyne
moiety by DNA. Although the reactivity of heterocyclic
enediynes5 shows considerable variation, only slight changes
in the cyclization barrier were found for para-substituted
benzene analogues.6,7 The developing radical centers are
orthogonal to the aromatic π-system, and therefore, the para
substituents influence reactivity predominantly via the field
effect.7 Electron-acceptor substituents are generally found
to accelerate the cyclization as suggested by Koga and
Morokuma8 because electron repulsion between the occupied
in-plane alkyne π-orbitals should increase the cyclization
barrier. A number of experimental results are consistent with
this suggestion.9
Scheme 1. Bergman Cyclization of Benzannelated Enediynes
(1) Jones, R. R.; Bergman, R. G. J. Am. Chem. Soc. 1972, 94, 660.
Bergman, R. G. Acc. Chem. Res. 1973, 6, 25.
(2) (a) Nicolaou, K. C.; Smith, A. L. Acc. Chem. Res. 1992, 25, 497. (b)
Maier, M. E.; Bosse, Folkert, Niestroj, A. J. Eur. J. Org. Chem. 1999, 1, 1.
(c) Grissom, J. W.; Gunawardena, G. U.; Klingberg, D.; Huang, D.
Tetrahedron 1996, 19, 6453. (d) Fallis, A. G. Can. J. Chem. 1999, 7, 159.
(e) Caddick, S.; Delisser, V. M.; Doyle, V. E.; Khan, S., Avent, A. G.;
Vile, S. Tetrahedron 1999, 55, 2737. (f) Wang, K. K. Chem. ReV. 1996,
96, 207. (g) Enediyne Antibiotics as Antitumor Agents; Borders, D. B.,
Doyle, T. W., Eds.; Marcel Dekker: New York, 1995. (h) Neocarzinosta-
tin: The Past, Present, and Future of an Anticancer Drug; Maeda, H.,
Edo, K., Ishida, N., Eds.; Springer: New York, 1997.
(3) Chen, X.; Tolbert, L. M.; Hess, D. W.; Henderson, C. Macromolecules
2001, 34, 4104. Shah, H. V.; Babb, D. A.; Smith, D. W., Jr. Polymer, 2000,
41, 4415. John, J. A.; Tour, J. M. J. Am. Chem. Soc. 1994, 116, 5011.
(4) Bowles, D. M.; Anthony, J. E. Org. Lett. 2000, 2, 85.
The present study was prompted by our experimental
finding of an accelerating effect by an electron-donating
OMe group at the ortho position.10 Although the effect is
subtles2,3-diethynylanisole is only two times more reactive
(6) (a) Boger, D. L.; Zhou, J. J. Org. Chem. 1993, 58, 3018. (b) Grissom,
J. W.; Calkins, T. L.; McMillen, H. A.; Jiang Y. J. Org. Chem. 1994, 59,
5833. (d) Note, however, that benzannelation might alter the rate-limiting
step in enediyne cycloaromatization. Kaneko, T.; Takahashi, M.; Hirama,
M. Tetrahedron Lett. 1999, 40, 2015. Fujimura. Y.; Hirama, M. J. Phys.
Chem. A 1999, 103, 7672.
(5) (a) Kim, C.-S.; Russell, K. C. J. Org. Chem. 1998, 63, 8229. (b)
Choy, N.; Russell, K. C. Heterocycles 1999, 51, 13. (c) Kim, C.-S.; Russell,
K. C. Tetrahedron Lett. 1999, 40, 3835. (d) Kim, C.-S.; Diez, C.; Russell,
K. C. Chem. Eur. J. 2000, 6, 1555.
10.1021/ol0255054 CCC: $22.00 © 2002 American Chemical Society
Published on Web 03/03/2002

Table 1. Activation Barriers and Reaction Energies (kcal mol-1) for Bergman Cyclizations of ortho- and para-Substituted
1,2-Diethylnylbenzenes Calculated at the B3LYP/6-31G** Level; Data for para-Substituted Cases Given in Parentheses
a Planar and nonplanar conformers of NH2, syn and anti conformers of OH, CHO, and OMe, and staggered and eclipsed conformations of OMe group
were considered. b δE ) Eortho - Epara, difference in the absolute energies between ortho and para isomers. δER is the difference for the enediynes, and δETS
is for transition states. c The ZPE correction is in the range of 0.2-0.8 kcal mol-1
.
than 1,2-diethynylbenzene at 170 °C11 (Table 2, Supporting
Information)sthe direction of the change is unexpected. The
accelerating influence of an electron-donor substituent clearly
contradicts with the simple premise of the Koga-Morokuma
hypothesis.
Intrigued by this result we undertook this theoretical
analysis of the “proximity effect”12 of ortho substituents in
the Bergman cyclization. Below, we report our finding that
the cyclization step is highly sensitive to the nature of ortho
substitution. We also evaluate specific mechanisms account-
ing for this phenomenon.
Scheme 2a
(7) Choy, N.; Kim, C.-S.; Ballestero, C.; Argitas, L., Diez, C.; Licht-
enberg, F.; Shapiro, J.; Russell, K. C. Tetrahedron Lett. 2000, 41, 6995.
Correlation of the reaction rate with Hammett σm value revealed a low
sensitivity to substituents (F ) 0.654). The Swain-Lupton model yielded
a larger field parameter (0.662) than resonance parameter (0.227) indicating
that conjugative effects in the out-of-plane π-system are of lesser importance
than field effects.
a
Reagents and conditions: (a) (1) NaOH, Hg(OAc)2, (2) KI,
I2; (b) (1) MeI, (2) N2H4, FeCl3, (3) NaNO2, KCl, KI; (c) Pd(PPh3)2,
(1) Cu(I), HCCSiMe3, (2) K2CO3, MeOH.
(8) Koga, N.; Morokuma, K. J. Am. Chem. Soc. 1991, 113, 1907.
(9) Schmittel, M.; Kiau, S. Chem. Lett. 1995, 953. Mayer, M. E.; Greiner,
B. Liebigs Ann. Chem. 1992, 855. Semmelhack, M. F.; Neu, T.; Foubelo,
F. J. Org. Chem. 1994, 59, 5038. Nicolau, K. C.; Zuccarello, G.; Riemer,
C.; Estevez, V. A.; Dai, W.-M. J. Am. Chem. Soc. 1992, 114, 7360.
(10) Synthesis of the enediyne is outlined in Figure 1. The key step
involved Sonogashira coupling of TMS-acetylene with 3-OMe-1,2-diodo-
benzene in 78% yield. (See Supporting Information)
All geometries were computed13,14 using either the re-
stricted (reactant) or broken-spin unrestricted (TS and
diradical product)15 B3LYP/6-31G** method which has been
shown to provide reasonable accuracy for the Bergman
cyclization.16,17 The computational results are summarized
in Table 1.
(11) We used the conditions of Grissom et al.:6b 5.4 mg (0.034 mmol)
of 1.2-diethynyl-3-methoxybenzene, 0.27 g (3.4 mmol) of 1,4-cyclohexa-
diene, and 4.7 mg (0.011 mmol) of tetraphenylnaphthalene internal standard
were dissolved in 10 mL of chlorobenzene. Next, 50-µL aliquots were sealed
in thin capillaries (at the height where no dead volume was observed) after
freeze-pump-thaw degassing and placed in an oil bath, the temperature
being maintained at 170 °C with a “Temp-O-Trol” temperature controller
((0.2 °C). The range of reaction times corresponded to 2-3 half-lives.
The reaction mixtures were diluted with hexane and analyzed by analytical
HPLC. The cyclization of 1,2-diethynyl benzene was studied under the same
conditions (see Supporting Information). Importantly, the presence of an
ortho substituent did not decrease the efficiency of the Bergman cyclization
In agreement with the experimental results of Russell and
co-workers,7 the influence of para substituents on the
cyclization is small. The activation energies for the neutral
as indicated by the yield of OMe-naphthalene (67%). The yield of
naphthalene in the cyclization of 1,2-diethynylbenzene was 73%. Products
formed by recombination of naphthyl and cyclohexadienyl radicals constitute
the rest of the reaction mixtures.
(12) Charton, M. Prog. Phys. Org. Chem. 1971, 8, 235.
1120
Org. Lett., Vol. 4, No. 7, 2002

Figure 1. Correlation between the calculated activation energy of the Bergman cyclization and the product of natural charges at the
terminal acetylenic atoms of benzannelated enediynes. Only para substituents obey the correlation.
substituents lie within a range of only 0.6 kcal/mol (Table
1). A larger decrease in the activation energy for a positively
charged substituent (NH3 ) demonstrates the predominant
products of natural atomic charges18,19 at the acetylenic
carbons with the activation energy (Figure 1).
+
However, the influence of ortho substituents on the
activation energy is not limited to the field effect, as indicated
by large deviations from the above correlation (Table 1 and
Figure 1). As a result the accessible range of activation
energies is much larger, from 27.7 to 32.3 kcal/mol. At 37
°C, all other factors being equal, it corresponds to an almost
2000-fold change in the reaction rate!18 This difference is
especially remarkable since the electronic effects of ortho
and para substituents are often considered similar.12
Further insight into the role of ortho substituents is
provided from the difference in absolute energies between
ortho- and para-substituted enediynes (δER) as well as their
cyclized TSs (δETS) given in Table 1. Positive values of δER
and δETS are indicative of steric repulsion and lower
stabilities of the ortho compounds compared to those of the
corresponding para compounds. Negative values of δER and
δETS indicate that ortho isomer is more stable as a result of
stabilizing interactions of the substituent with the enediyne
moiety. It is important to note here that only difference
between δER and δETS and not the sign of the δEs determines
the effect of the substituent on the cyclization rate.
role of the field effect.
The field effect of electronegative substituents decreases
electron density at the terminal acetylenic atoms thus
alleviating the electron repulsion. We found that this simple
rationale is supported by an excellent correlation of the
(13) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick,
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.;
Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi,
I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M.
W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,
M.; Replogle, E. S.; Pople, J. A. Gaussian 98; Gaussian, Inc.: Pittsburgh,
PA, 1998.
(14) The size of the systems precluded the use of accurate but
computationally expensive multiconfigurational methods. (a) Nicolaides,
A.; Borden, W. T. J. Am. Chem. Soc. 1993, 115, 11951. (b) Lindh, R.;
Lee, T. J.; Bernhardsson, A.; Persson, B. J.; Karlstro¨m, G. J. Am. Chem.
Soc. 1995, 117, 7186. (c) Kraka, E.; Cremer, D.; Bucher, G.; Wandel, H.;
Sander, W. Chem. Phys. Lett. 1997, 268, 313. (d) McMahon, R. J.; Halter,
R. J.; Fimmen, R. L.; Wilson, R. J.; Peebles, S. A.; Kuczkowski, R. L.;
Stanton, J. F. J. Am. Chem. Soc. 2000, 122, 939. (e) Cramer, C. J.; Nash,
J. J.; Squires, R. R. Chem. Phys. Lett. 1997, 277, 311. Cramer, C. J.; Squires,
R. R. J. Phys. Chem. A 1997, 101, 9191.
(15) UB3LYP/6-31G** slightly overestimates the reaction barrier and
underestimates the endothermicity (see Supporting Information). However,
this accuracy is acceptable since we are considering relatiVe trends in
reactivity.
(16) (a) Kraka, E.; Cremer, D. J. Am. Chem. Soc. 2000, 122, 8245. (b)
Gra¨fentein, J.; Hjerpe, A. M.; Kraka, E.; Cremer, D. J. Phys. Chem. 2000,
122, 8245. (c) Jones, G. B.; Warner, P. M. J. Am. Chem. Soc. 2001, 123,
2134. (d) Feldgus, S.; Shields, G. C. Chem. Phys. Lett. 2001, 347, 505.
(17) BLYP is another DFT method commonly used for theoretical studies
on the Bergman cyclization: Prall, M.; Wittkopp, A.; Fokin, A. A.;
Schreiner, P. R. J. Comput. Chem. 2001, 22, 1605. Prall, M.; Wittkopp,
A.; Schreiner, P. R. J. Phys. Chem. 2001, 105, 9265. (See Supporting
Information.)
Sterically compact ortho substituents such as Me, OH, anti-
OMe, F, Cl, and CN destabilize the ground and transition
states to a similar degree. As a result, changes in the
activation energy are minor. The net effect of these substit-
uents on the cyclization rate is similar to that of the para
substituents, and the corresponding computational data fit
well into the correlation in Figure 1.
(18) The product of atomic charges is proportional to the electrostatic
repulsion between the atoms (with constant interatomic distance).
(19) Weinhold, F.; Natural Bond Orbital Methods. In Encyclopedia of
Computational Chemistry; Schleyer, P.v. R., Ed.; Wiley: New York, 1998;
Vol. 3, p 1792.
Org. Lett., Vol. 4, No. 7, 2002
1121

+
On the other hand, the presence of ortho-NO2, NH2, NH3 ,
interaction disappears since the overlap of σ*X-H orbital and
the in-plane π-bond decreases as a result of inward bending
of the acetylene moiety away from the X-H group.
CHO, CF3, and syn-OMe groups results in large changes in
the activation energy. Depending on the substituent, three
different factors account for these changes: steric assistance20
(decrease in steric destabilization in TS), extra stabilization
of the TS, and decrease in TS stabilization. The first two
factors decrease the activation energy.
syn-CHO, NO2, CF3, and syn-OMe facilitate the cyclization
through steric assistance. The ground state is destabilized
by steric factors as reflected in the decreased C1-C6 distances
for the ortho isomers and higher energies of the latter (more
than 5 kcal/mol for syn-OMe and NO2). The steric effects
are alleviated in the TS where the acetylene moiety is bent
away from the ortho substituent.
An interesting feature of sterically demanding substituents
is the dependence of the activation energy on the conforma-
tion. This conformational dependence of the Bergman
cyclization may provide an additional way to control
reactivity of enediynes in vivo when binding to the minor
groove of DNA is conformer specific.
However, Curtin-Hammett analysis shows that syn-
conformers are unlikely to contribute significantly to the rate
of the Bergman cyclization of 2,3-diethynylanisole because
these conformers are destabilized significantly compared with
the lowest energy (a,s)-conformer (3.9 kcal mol-1 (s,s) and
6.5 kcal mol-1 (s,e)). Although the 2-fold difference in the
reaction rates for cyclization of 1,2-diethynylbenzene and
2,3-diethynylanisole is consistent with the small difference
(0.2 kcal mol-1) in the computed barrier, theory predicts the
anisole to be less reactive. Thus, at this stage, theory does
not provide a definitive explanation for the subtle increase
in the reactivity of the ortho-OMe enediyne. An intriguing
possibility is that the diradical is trapped intramolecularly
by the ortho-methoxy group.21 This trapping might prevent
the retro-Bergman ring opening and make the cyclization
step irreversible. We plan to seek further insight into the
details of the hydrogen atom abstraction using isotope
labeling and to carry out additional kinetic studies to
determine the rate-limiting step in the cycloaromatization
cascade.
In contrast, in the case of a positively charged functional
group, NH3 , the strength of hydrogen bonding increases at
+
the beginning of the cyclization step, thus providing extra
stabilization to the TS. This accelerating effect can be
attributed to a larger electrostatic component in the H-
bonding in the case of a positively charged group and to a
concomitant strong through-space electron transfer from the
adjacent in-plane π-bond of acetylene moiety to the am-
monium group. As a result, population of this π bond is
markedly decreased (Table 3, Supporting Information), thus
decreasing electron repulsion in the transition state.
The large decrease in the activation energy upon proto-
nation is important because cancer cells are more acidic (pH
5.523) than normal cells. At this pH, anilines are protonated
noticeably.24 Hence, the fact that protonation of the ortho-
amino group decreases the activation energy by 3.9 kcal/
mol (and thus can speed up the reaction by the factor of 600
at 37 °C) can be used in the design of tumor-specific DNA
cleaving agents.16, 25
In conclusion, we have found that the Bergman cyclization
of aromatic enediynes is highly sensitive to ortho substitution
as a result of a combination of electronic, steric, and
electrostatic effects. This finding opens possibilities for the
rational design of and conformer-specific and pH-dependent
DNA-cleaving agents.
Acknowledgment. A.I. thanks Florida State University
for First Year Assistant Professor Award and Ms. Carol Joud
for technical assistance.
Supporting Information Available: Optimized geom-
etries and energies for all compounds, synthetic procedures,
and spectral and kinetic data for 2,3-diethynylanisole and
1,2-diethynylbenzene. This material is available free of
charge via the Internet at http://pubs.acs.org.
CH3, NH2, and syn-OH increase the activation energy
because they stabilize the starting material by a hydrogen
bond between the X-H moiety and in-plane acetylenic
π-orbital. The strength of the hydrogen bond increases with
the electronegativity of X.22 This interaction results in an
increase of the C1-C6 distance and outward bending of the
acetylene group. As the reaction proceeds, the stabilizing
OL0255054
(22) NBO analysis18 finds the following energies for the hydrogen bonds
(in kcal/mol). 1.3 (NH2), 2.7 (syn-OH), 5.2 (NH3+). Interestingly, direct
orbital interaction of substituents at the vinylic position with the developing
radical center can be either stabilizing or destabilizing.16c
(23) Osinsky, S.; Bubnovskaya, L. Arch. Geschwulstforsch. 1984, 54,
463. Tannock, I. F,: Rotin, D. Cancer Res. 1989, 49, 4373.
(24) Basicity of ortho-amino-enediynes is unknown, but it is reasonable
to suggest that the protonation will occur to a noticeable extent and can be
further controlled by substitution. Aniline (pKb ) 9.40) is 7% protonated
and p-anisidine (pKb ) 8.70) is 29% protonated at pH 5.5.
(20) Eliel, E. L.; Wilen, S. H.; Doyle, M. P. Basic Organic Stereochem-
istry; Wiley-Interscience: New York, 2001; p 459.
(21) The activation barrier for the intramolecular hydrogen abstraction
is 6.7 kcal/mol, and the reaction is 9.2 kcal/mol exothermic (UB3LYP/6-
31G**).
(25) Hoffner, J.; Schottelius, J.; Feichtinger, D.; Chen, P. J. Am. Chem.
Soc. 1998, 120, 376-385.
1122
Org. Lett., Vol. 4, No. 7, 2002