Conformational control in activation of an enediyne

Article abstract of DOI:10.1021/ja036763e

In the bicyclo[7.3.1]tridec-4-ene-2,6-diyne framework characteristic of calicheamicin, DFT calculations predict that the chair conformer should be much more reactive toward cycloaromatization compared to the boat form. A functionalized derivative of this

Full text of DOI:10.1021/ja036763e

Published on Web 08/07/2003
Conformational Control in Activation of an Enediyne
M. F. Semmelhack,* Lingyun Wu, Robert A. Pascal, Jr., and Douglas M. Ho
Department of Chemistry, Princeton UniVersity, Princeton, New Jersey 08544
Received June 18, 2003; E-mail: mfshack@princeton.edu
Scheme 1. Preparation of Key Intermediates and Attempted
The enediyne toxins have attracted attention due to their novel
Cyclizationa
structures, high toxicity, and unique mechanism of action.1 Their
activity is based on a sequence of thermal reactions: (a) a chemical
triggering event which converts a stable form into a reactive form,
(b) rapid cycloaromatization to an arene 1,4-diradical, and (c) DNA
cleavage. There has been interest in designing novel frameworks
that might allow activation under a variety of conditions and could
be tailored for selective activation.2 Efforts have been made to for-
mulate simple structure/reactivity relationships in order to predict
the reactivity of a new enediyne before and after triggering.3-5 How-
ever, there is no general analysis that is effective in complex sys-
tems, with theoretical prediction and experimental verification. Here
we explore, through computational analysis of the transition states
and through synthesis, a new basis for triggering the cyclic ene-
diynes.
It was suggested on the basis of molecular mechanics analysis
in an early model study that the calicheamicin/esperamicin frame-
work, represented by 1, could exist in either a chair (1c) or boat
(1b) conformation and that the chair form would be more reactive
toward cycloaromatization.6 We have evaluated the reactivity of
1c compared to 1b using DFT at the BLYP/6-31G(d) level7 and
found a relative transition state energy difference of 4.1 kcal/mol.
Molecules of general structure 1 show experimental half lifetimes
for cycloaromatization of ca. 10 min at 21 °C, presumably through
the chair conformer.8 If the conformation were restricted to the boat
form, the framework would presumably be quite stable until the
restriction is removed. The boat-to-chair conversion could then serve
as a triggering event for cycloaromatization.
a (a) i. H2O, Me2CO, 23 °C, 24 h; ii. I2/KI, NaHCO3, H2O, 23 °C, 2 h;
(b) i. ClC(O)OEt, NEt3, THF, -78 °C; ii. NaBH4, H2O/THF, 0 °C, 2 h; iii.
TBSCl, imidazole, DMF, 23 °C, 12 h; (c) for X, Y ) Cl, lauroyl peroxide,
heptane/PhCl, 90 °C, 12 h; (d) for X ) Ph, Y ) H, (tBuO)2, PhCl, 130 °C,
12 h; (e) X, Y ) Cl, tBuLi, THF, -78 °C, 3 h; (f) for X ) Ph, YdH: i.
O3, CH2Cl2, -78 °C; ii. Me2S, -78 °C; (g) N2CHP(O)(OMe)2, tBuOK,
THF, -78 to 0 °C, 10 h. (h) ClHCdCHCtCTΜS, Pd(PPh3)4, CuI, BuNH2,
C6H6, 23 °C, 24 h; (i) i. NIS, AgNO3, DMF, 23 °C, 5 h; ii. HF, MeCN/
H2O, 0 °C, 0.5 h; ii. ClC(O)C(O)Cl, DMSO, NEt3, CH2Cl2, -78 to 0 °C;
(j) CrCl2, THF, 23 °C, 24 h.
lactonization of 6 to give 7 which was then converted to the TBS
ether 8.9 Several tactics were tested for replacement of iodide by a
carbon unit with inversion of configuration.10 Two versions of a
radical addition/elimination process using vinyl sulfones 9 were
remarkably stereoselective, giving 10 in > 9:1 ratio over the epi-
mer.11 Reaction of 10 (X,Y ) Cl) with tBuLi12 gave 11 directly in
28% yield, while 10 (X ) H, Y ) Ph) was converted to 11 by
ozonolysis to aldehyde 12 followed by Gilbert’s reagent (49% yield
overall).13
Various strategies for completion of the enediyne ring were
considered. The common protocol,14 involving chain extension to
13, introduction of the aldehyde as in 14, and then intramolecular
alkyne nucleophile addition to the aldehyde to give 15, was not
successful. For example, using Cr(II) activation15 of the alkynyl
iodide 14 (X ) I) under conditions of high dilution, the only product
characterized (17% yield) was the symmetrical dimeric ring
structure 16.16 An obvious problem is the requirement that the
favored chair conformation with equatorial side chains (14c) must
flip to the boat (14b) in order to allow cyclization to 15.
An alternate strategy is based on the Cr(II)-promoted addition
of an allylic bromide to an aldehyde,17 with an expectation of
coordination of the allyl-Cr(III) unit to the aldehyde carbonyl in a
large ring (Scheme 2, structure A) in order to favor the desired
conformer for cyclization.18 The sequence began with alkyne
addition (Scheme 2, step a) to aldehyde 17 (obtained from 11 by
desilylation and Swern oxidation; 74% yield) to give a mixture of
We considered the 1,3-bridged frameworks 2 and 4 and the ring-
opened analogues 3 and 5. The barrier for cycloaromatization
calculated for 2 is 24.6 kcal/mol while the ring-opened version 3
has a much lower barrier, 17.0 kcal/mol. With a three-atom bridge
(4), the framework has a barrier of 23.0 kcal/mol while the ring-
opened version (5) is lower, at 17.3 kcal/mol. These activation
energies predict long lifetimes at 25 °C for 2 and 4 but rapid
rearrangement for analogues 3 and 5.
1
epimers at the new hydroxyl group (ratio 2:1 by H NMR). After
chromatography, the epimers were protected as 18a and 18b
(yields: 18a 45%; 18b 26%) and carried forward separately; only
The synthesis (Scheme 1) of a test case begins with the iodo-
9
10496
J. AM. CHEM. SOC. 2003, 125, 10496-10497
10.1021/ja036763e CCC: $25.00 © 2003 American Chemical Society

C O M M U N I C A T I O N S
Scheme 2. Successful Synthesis and Rearrangement Kineticsa
diisobutylaluminum hydride was studied. Reproducible data were
obtained by following the disappearance of the UV absorption for
the enediyne unit (λmax 296 nm) upon reaction with DIBALH in
THF.20 The half lifetime for the intermediate 26 was 43.5 min at
24.5 °C (average of three runs). From a multi-milligram run in THF
containing excess CHD, the cycloaromatized product 27b (spon-
taneous loss of the silyl protecting group) was isolated in 20% yield.
The observed half lifetimes for 22 and intermediate 26 correspond
to free energies of activation for the cyclizations of approximately
29 and 22 kcal/mol, respectively. Thus the BLYP/6-31G(d) calcu-
lations underestimate the barriers but correctly predict the difference
in the activation energies for the framework and ring-opened com-
pounds. These results open the way for the design of selective trig-
gers for cycloaromatization on the basis of conformational control.
Acknowledgment. This work was supported by grants from
the Public Health Service [NIH CA54819, to M.F.S.] and the
National Science Foundation [CHE-0077990, to R.A.P.].
Supporting Information Available: Full characterization data on
all new compounds including selected NMR spectra; a crystallographic
information file (CIF) containing the full X-ray structural data for 10
(X,Y ) Cl), 16 and 21; an ASCII text file containing the atomic
coordinates of 1b, 1c, 2-5, and the corresponding transition states;
and general conditions for rate studies. This material is available free
of charge via the Internet at http://pubs.acs.org.
a (a) i. LiCtCCH2OTHP, CeCl3, THF, -78 °C, 5 h; ii. TBSCl,
imidazole, DMF, 23 °C; (b) cis-ICHdCHCH2OH, Pd(PPh3)4, CuI, NEt3,
THF, 23 °C, 12 h; (c) i. MsCl, NEt3, CH2Cl2, -50 °C, 1 h; ii. NaBr, CH2Cl2,
23 °C, 3 h; iii. PPTS, IPA, 48 h, 23 °C; iv. ClC(O)C(O)Cl, DMSO, NEt3,
CH2Cl2, -78 to 0 °C; (d) CrCl2, THF, 23 °C, 12 h; (e) i. MsCl, NEt2,
CH2Cl2, -20 °C, 2 h; ii. DBU, CH2Cl2, 23 °C, 4 h; (f) 74 °C, C6D6; (g) 1
M NaOMe, MeOH, 1,4-CHD, 23 °C, 4 h; (h) (iBu)2AlH, 1,4-CHD, THF,
23 °C.
References
(1) Thorson, J. S.; Sievers, E. L.; Ahlert, J.; Shepard, E.; Whitwam, R. E.;
Onwueme, K. C.; Ruppen, M. Curr. Pharm. Des. 2000, 6, 1841-1879.
(2) Jones, G. B.; Fouad, F. S. Curr. Pharm. Des. 2002, 8, 2415-2440.
(3) The first suggestion, based on the distance between the ends of the alkyne
units (“cd” distance), is useful only for simple monocyclic systems:
Nicolaou, K. C.; Zuccarello, G.; Ogawa, Y.; Schweiger, E. J.; Kumazawa,
T. J. Am. Chem. Soc. 1988, 110, 4866-4868.
(4) Strain energy (MM) analysis has been used in more complex frameworks
related to the natural products: Magnus, P.; Fortt, S.; Pitterna, T.; Snyder,
J. P. J. Am. Chem. Soc. 1990, 112, 4986-4987.
(5) DFT calculations have been reported recently for enediyne cyclizations:
(a) Jones, G. B.; Wright, J. M.; Hynd, G.; Wyatt, J. K.; Warner, P. M.;
Huber, R. S.; Li, A.; Kilgore, M. W.; Sticca, R. P.; Pollenz, R. S. J. Org.
Chem. 2002, 67, 5727-5732. (b) Ahlstrom, B.; Kraka, E.; Cremer, D.
Chem. Phys. Lett. 2002, 361, 129-135. (c) Prall, M.; Wittkopp, A.;
Schreiner, P. R. J. Phys. Chem. A 2001, 105, 9265-9274.
(6) Magnus, P.; Carter, P. A. J. Am. Chem. Soc. 1988, 110, 1626-1628.
(7) For discussion, see Schreiner, P. R. J. Am. Chem. Soc. 1998, 120, 4184-
4190. Density functional calculations were performed using GAUSSIAN
98; the built-in default thresholds for wave function and gradient
convergence were employed; transition state structures were located by
using the QST3 option and verified by frequency calculations.
(8) Sarpong, R. Ph.D. Dissertation, Princeton University, 2001.
(9) Based on reports by the following: (a) Belleau, B.; Dvornik, D. Can J.
Chem. 1965, 43, 2545-2550. (b) Kobayashi, S.; Eguchi, Y.; Shimada,
M.; Ohno, M. Chem. Pharm. Bull. 1990, 38, 1479-1484.
(10) Direct SN2 displacement with several carbon nucleophiles was not
successful.
the major epimer was successful in the key cyclization step. The
yields reported in Scheme 2 are for series a.
Both epimers were converted through Sonogashira coupling with
3-hydroxy-1-iodo-(Z)-prop-1-ene into alcohols 19.19 The hydroxyl
group was activated as the mesylate and converted to the bromide.
Then cleavage of the THP ether allowed oxidation to the aldehydes,
20. The critical cyclization step was carried out on 20a with Cr(II)
activation and gave a fairly complex mixture from which the cyclic
diyne 21 was isolated in 15-30% yield. An X-ray determination
established the structure of 21 and then elimination of water was
provoked under mild conditions to give the cyclic enediyne 22.
Attempted cyclization of 20b under the same conditions gave a
complex mixture from which only dimeric products could be
isolated.
(11) The configurations of the substitution products 10 were established for
10 (X, Y ) Cl) via X-ray analysis and conversion to common intermediate
11.
(12) Bertrand, F.; Quiclet-Sire, B.; Zard, S. Z. Angew. Chem., Int. Ed. 1999,
38, 1943-1946.
(13) Brown, D. G.; Velthuisen, E. J.; Commerford, J. R.; Brisbois, R. G.; Hoye,
T. R. J. Org. Chem. 1996, 61, 2540-2541.
Consistent with the calculations, 22 is indefinitely stable at 23
°C and decomposes with a half lifetime of 110 h at 74 °C (but
gives none of the typical cycloaromatization products from the
expected intermediate diradical, 23).
(14) Kende, A. S.; Smith, C. A. Tetrahedron Lett. 1988, 29, 4217-4220.
(15) Crevisy, C.; Beau, J.-M. Tetrahedron Lett. 1991, 32, 3171-3174.
(16) The structure of 16 was confirmed by an X-ray diffraction study.
(17) Wessjohann, L. A.; Scheid, G. Synthesis 1999, 1-36.
(18) Wender, P. A.; McKinney, J. A.; Mukai, C. J. Am. Chem. Soc. 1990,
112, 5369-5370.
To evaluate the conformational effect on cycloaromatization
reactivity, 22 was stirred at 23 °C in 1.0 M NaOMe/MeOH solution
containing a ca. 20-fold excess of 1,4-cyclohexadiene (1,4-CHD).
Attempts to isolate the ring-opened ester 24 through rapid workup
failed, and the only discrete product characterized was 25 from
cycloaromatization, in ca. 20% yield. From this experiment, the
half lifetime of 24 is less than about 2 h at 23 °C. In an effort to
open the ring rapidly at low temperature and follow the cycloaro-
matization process by spectroscopy, the reaction of 22 with
(19) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett. 1975, 16,
4467-4470.
(20) Comparable rate data were obtained by monitoring the vinyl H at δ 6.38
ppm in the 1H NMR spectrum, but serious line broadening left greater
uncertainty in the values.
JA036763E
9
J. AM. CHEM. SOC. VOL. 125, NO. 35, 2003 10497