Analysis of HIF-1 inhibition by manassantin A and analogues with modified tetrahydrofuran configurations

Article abstract of DOI:10.1016/j.bmcl.2009.04.071

We have shown that manassantin A downregulated the HIF-1α expression and inhibited the secretion of VEGF. We have also demonstrated that the 2,3-cis-3,4-trans-4,5-cis-configuration of the tetrahydrofuran is critical to the HIF-1 inhibition of manassantin

Full text of DOI:10.1016/j.bmcl.2009.04.071

Bioorganic & Medicinal Chemistry Letters 19 (2009) 3783–3786
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Analysis of HIF-1 inhibition by manassantin A and analogues with modified
tetrahydrofuran configurations
Amanda C. Kasper a, Eui Jung Moon b, Xiangqian Hu a, Yongho Park a, Ceshea M. Wooten a, Hyoungsu Kim a,
Weitao Yang a, Mark W. Dewhirst b,c, Jiyong Hong a,
*
a Department of Chemistry, Duke University, Durham, NC 27708, United States
b Department of Pathology, Duke University Medical Center, Durham, NC 27708, United States
c Department of Radiation Oncology, Duke University Medical Center, Durham, NC 27708, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
We have shown that manassantin A downregulated the HIF-1a expression and inhibited the secretion of
VEGF. We have also demonstrated that the 2,3-cis-3,4-trans-4,5-cis-configuration of the tetrahydrofuran
is critical to the HIF-1 inhibition of manassantin A.
Received 2 March 2009
Revised 14 April 2009
Accepted 17 April 2009
Available online 22 April 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Manassantin A
Hypoxia
Hypoxia inducible factor-1
Tetrahydrofuran analogue
Tetrahydrofuran configuration
Vascular endothelial growth factor
2,3-cis-3,4-trans-4,5-cis-Tetrahydrofuran
2,3-cis-3,4-trans-4,5-trans-Tetrahydrofuran
2,3-trans-3,4-trans-4,5-trans-
Tetrahydrofuran
Conformation–activity relationship
Molecular oxygen (O2) is required for aerobic metabolism to
maintain intracellular bioenergetics and to serve as an electron
acceptor in many organic and inorganic reactions.1 Hypoxia, usu-
ally defined as 62% of O2, occurs in a variety of pathological condi-
tions, including stroke, tissue ischemia, inflammation, and tumor
growth.2 Mammalian tissues have developed a number of essential
mechanisms to cope with the stress of hypoxia. Among these cop-
ing mechanisms is the response mediated by the hypoxia-induc-
ible transcription factor 1 (HIF-1). It is a basic helix-loop-helix
(bHLH)-PER-ARNT-SIM (PAS) family protein that forms a heterodi-
Through HIF-1, tumors adapt to hypoxia by increasing angio-
genesis and metastatic potential, altering apoptosis, and regulating
metabolism.6 These adaptations make tumors more aggressive and
treatment-resistant resulting in poor patient prognosis.7 Immuno-
histochemical analyses have revealed that HIF-1 is overexpressed
in many human cancers.8 HIF-1 overexpression is not only in-
volved in tumor progression but also associated with resistance
to radiation9 and chemotherapy.10 It has been shown that inhibi-
tion of HIF-1 activity suppresses tumor growth, destroys blood ves-
sels, enhances tumor apoptosis, and increases radiosensitivity,11
making HIF-1 a good potential target for anti-cancer treatment.
Due to the importance of HIF-1 in tumor development and pro-
gression, a considerable amount of effort has been made to identify
HIF-1 inhibitors for treatment of cancer.12 A variety of anticancer
drugs, most of which were not developed as HIF-1 inhibitors, have
been reported to inhibit HIF-1. However, these compounds possess
relatively low HIF-1 inhibitory activity (Pmicromolar range). In
addition, most of them lack the desired selectivity for the HIF-1
signaling pathway or toxicity profiles required for a useful thera-
peutic agent. The dineolignans manassantin A (1), manassantin B
(2), manassantin B1 (3), and 4-O-demethylmanassantin B (4)
mer with its
There are two additional HIF-1
and HIF-3a 4
Like HIF-1 , they also bind to HIF-1b (ARNT) for
a
and b subunits and acts as a transcription factor.3
-related bHLH-PAS proteins: HIF-
a
2a
.
a
activation. HIF-1 is a main regulator of hypoxia since it activates
more than 60 genes involved in angiogenesis (VEGF), glucose
transport (GLUT1), glycolytic pathways (LDHA), ROS signals (iNOS),
and erythropoiesis (EPO), as well as a number of other processes.5
* Corresponding author. Tel.: +1 919 660 1545; fax: +1 919 660 1605.
E-mail address: jiyong.hong@duke.edu (J. Hong).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.04.071

3784
A. C. Kasper et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3783–3786
(Fig. 1), isolated from the aquatic plant Saururus cernuus L. (Saurur-
aceae), have been shown to be potent inhibitors of HIF-1.13 How-
ever, their molecular mechanisms of action have yet to be
established. Since manassantins may sensitize cancer cells to che-
mo- and/or radiotherapy by HIF-1 inhibition,11c,14 use of manas-
santins in combination with other cytotoxic drugs and/or
radiation has great potential for therapeutic applications.
In broad connection with our interest in the stereoselective syn-
thesis of substituted tetrahydrofurans,15,16 we recently completed
the synthesis of 1 and 2 via a direct nucleophilic addition of an
organozinc reagent to a 2-benzenesulfonyl cyclic ether to synthe-
size the 2,3-cis-3,4-trans-4,5-cis-tetrahydrofuran moiety of 1 and
2.17 In addition, we also showed that the (R)-configuration at C-7
and C-7000 is not critical for HIF-1 inhibition and that the hydroxyl
group at C-7 and C-7000can be replaced with carbonyl group without
significant loss of activity.
Herein, we present initial biological data demonstrating the
significant potential of 1 as a potent HIF-1 inhibitor with little
cytotoxicity. In addition, we describe the synthesis and conforma-
tion–activity relationship study of tetrahydrofuran core analogues
of 1 to characterize the effect of tetrahydrofuran conformation on
HIF-1 inhibitory activity.
Previously, we reported that 1 exhibited a highly potent level of
HIF-1 inhibitory activity in a luciferase-reporter based assay
(IC50 = 1–10 nM).17 Based on the luciferase assay data, we used
Western blots to further confirm the HIF-1 inhibitory activity of
1. 4T1 cells, a mouse mammary carcinoma, were grown under hyp-
oxic conditions (0.5% O2) for 24 h with various concentrations (0, 1,
10, 100 nM, and 1 lM) of 1. Western blots were performed on nu-
clear extracts as reported.11b A dose–response study revealed that
exposure of 4T1 cells to 1 at concentrations higher than 10 nM for
24 h significantly inhibited hypoxia-induced expression of the HIF-
1a
protein (Fig. 2a).
To determine if 1 also inhibits chemically-induced HIF-1
a
expression,18 we treated 4T1 cells with 240
l
M of CoCl2 for 24 h
and carried out Western blots with nuclear extracts. HIF-1
expression induced by CoCl2 was inhibited by (100 nM)
(Fig. 2b) indicating that 1 inhibits chemically induced HIF-1
expression as well as hypoxia-induced HIF-1 expression. It should
be noted that 1 was reported to have no significant effect on iron
a
1
a
a
chelator-induced HIF-1 activation in T47D cells (10 lM 1,10-phe-
nanthroline, 16 h).13c To examine if the HIF-1 inhibition by 1 is
cell-type specific, we carried out the same experiment with
MDA-MB-231, a human breast cancer cell line, and observed the
same inhibition effect, which showed that the HIF-1 inhibition ef-
fect by 1 is not cell-type specific (Fig. 2b).
As stated earlier, more than 60 genes have been identified as
targets of HIF-1. Vascular endothelial growth factor (VEGF) is a
gene that is highly involved in tumor progression as a pro-angio-
genic factor. The effects of 1 on HIF-1 regulated VEGF secretion
were examined in 4T1 cells using ELISA. Cells were incubated un-
der hypoxia (0.5% O2 for 24 h) with various concentrations (0, 1, 10,
100 nM, and 1 lM) of 1. Cell culture supernates were collected, and
VEGF levels in supernates were measured by a commercially avail-
able kit (R&D systems, Minneapolis, MN). As we observed from
HIF-1 expression, VEGF induced by hypoxia was significantly
inhibited by 1 at concentrations higher than 10 nM (Fig. 3).
Cytotoxicity of 1 was examined using the MTS assay, a standard
colorimetric cytotoxicity assay (see Supplementary data for de-
tails). 4T1 cells were seeded in a 96-well plate and incubated with
serially diluted 1 (0–10
tion examined (10
that 1 completely inhibited the expression of HIF-1 at the concen-
tration 6100 nM, 1 possesses a significant therapeutic window
(IC50 (cytotoxicity)/IC50 (HIF-1 inhibition) P 100).
lM) for 24 h. Up to the highest concentra-
M), cells had ꢀ70% survival rate. Considering
l
The configuration of 1 is largely determined by the 2,3-cis-3,4-
trans-4,5-cis-configuration of the tetrahydrofuran core. We
hypothesized that the overall conformation should be an impor-
tant determinant for the binding mode and affinity toward molec-
Figure 2. Inhibition of HIF-1
hypoxia (0.5% O2, 24 h) with and without 1, HIF-1
Western blots. By comparison with loading control (histone H1), nuclear expression
of HIF-1 was significantly inhibited by 1 at concentrations higher than 10 nM. (b)
a
expression by 1. (a) After treating 4T1 cells under
a
expression was evaluated using
a
In 4T1 and MDA-MB-231 cells, HIF-1a expression induced by CoCl2 (240 lM, 24 h)
Figure 1. Structure of dineolignans from Saururus cernuus.
was also inhibited by 1.

A. C. Kasper et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3783–3786
3785
we used 4T1-ODD-Luc cells21 stably transfected with the oxygen-
dependent-degradation (ODD) domain of HIF-1 and a firefly lucif-
a
erase reporter. This ODD-Luc reporter contains a CMV promoter,
which is constitutively active. Since its ODD domain is identical
to that of HIF-1, it enables us to directly detect the stability of
HIF-1. Cells were seeded in the 24-well plate at a density of
105 cells/well. After 16-h incubation, cells were treated with
240 lM of CoCl2 and serially diluted compounds for 24 h. Since
luciferase requires O2 for its activity but the ODD-Luc is highly sen-
sitive to reoxygenation, we induced the HIF-1 expression by CoCl2,
not by hypoxia to accurately determine the effect of the com-
pounds on HIF-1 stability. Luciferase signals were detected and
quantified as relative light units (RLUs). The ODD-Luc assay to as-
sess HIF-1 inhibitory activity of 11 and 13 revealed that 11 was
Figure 3. Inhibition of VEGF secretion by 1. Inhibition of VEGF was determined in
4T1 cells after hypoxia treatment (0.5% O2, 24 h). After treatment with and without
1, cell culture supernates were collected, and VEGF secretion was measured using
ELISA. Compound 1 significantly inhibited VEGF at concentrations higher than
10 nM (p < 0.001).
nearly inactive and 13 was less active than
1 by 10-fold
(IC50 = 47 nM) (Fig. 4). Based on these results, the 2,3-cis-3,4-
trans-4,5-cis-configuration of the tetrahydrofuran core is critical
for HIF-1 inhibition.
ular target(s), potency, and HIF-1 signaling specificity of 1. To test
the hypothesis, we prepared and evaluated manassantin A ana-
logues with modifications in tetrahydrofuran configuration. These
analogues were easily prepared through the procedures previously
reported by our group (Scheme 1).15,17
The synthesis of manassantin A analogues with 2,3-cis-3,4-
trans-4,5-trans- and 2,3-trans-3,4-trans-4,5-trans-tetrahydrofuran
cores (11 and 13) was accomplished as described in Scheme 1.
Briefly, the 2,3-cis-3,4-trans-4,5-trans-tetrahydrofuran 7a was pre-
pared via BF3ꢁOEt2-promoted deoxygenation of cyclic hemiketal
615 followed by stereoselective reduction of the oxocarbenium
ion intermediate. Deprotection of the Bn and TBS groups, BEMP-
mediated coupling, and polymer-supported BH4-reduction com-
pleted the synthesis of 11.19 Compound 13 was prepared via
BF3ꢁOEt2-promoted epimerization/reductive deoxygenation fol-
lowed by deprotection, BEMP-mediated coupling, and polymer-
supported BH4-reduction.20
HIF-1 Stability (ODD-luciferase)
compound 11
compound 13
140
120
100
80
60
40
20
0
0
1
10
100
500
1,000
Concentration (nM)
To determine HIF-1 inhibitory activity of 11 and 13, we used a
luciferase-reporter based assay as a primary screen. For this assay,
Figure 4. Inhibition of HIF-1 by 11 and 13.
Scheme 1. Reagent and conditions: (a) Ar2Li, THF, ꢂ78 °C, 40 min, 70%; (b) BF3ꢁOEt2, NaBH3CN, CH2Cl2, ꢂ78 °C, 30 min, 99%; (c) BF3ꢁOEt2, CH2Cl2, ꢂ78 to ꢂ20 °C, 2 h; then,
NaBH3CN, ꢂ78 °C, 30 min, 96%; (d) H2, Pd/C, EtOAc/EtOH (3:1), 25 °C, 2 h; (e) TBAF, THF, 25 °C, 1 h, 88% for 2 steps; (f) 10, BEMP, CH2Cl2, 25 °C, 18 h, 66%; (g)
(polystyrylmethyl)trimethylammonium borohydride, MeOH, 25 °C 48 h, 75%; (h) 10, BEMP, CH2Cl2, 25 °C, 20 h, 92%; (i) (polystyrylmethyl)trimethylammonium borohydride,
MeOH, 25 °C, 30 h, 86%.

3786
A. C. Kasper et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3783–3786
Figure 5. Optimized conformations of truncated structures of 1, 11, and 13 (a) truncated structures (14–16). (b) Overlay I. (c) Overlay II: front view. (d) Overlay II: side view
(14 in black, 15 in red, 16 in blue, and tetrahydrofurans in green).
8. Semenza, G. L. Drug Discovery Today 2007, 12, 853.
9. Aebersold, D. M.; Burri, P.; Beer, K. T.; Laissue, J.; Djonov, V.; Greiner, R. H.;
Semenza, G. L. Cancer Res. 2001, 61, 2911.
10. Koukourakis, M. I.; Giatromanolaki, A.; Sivridis, E.; Simopoulos, C.; Turley, H.;
Talks, K.; Gatter, K. C.; Harris, A. L. Int. J. Radiat. Oncol. Biol. Phys. 2002, 53, 1192.
11. (a) Kung, A. L.; Wang, S.; Klco, J. M.; Kaelin, W. G.; Livingston, D. M. Nat. Med.
To further characterize the effect of tetrahydrofuran conforma-
tion of 1, 11, and 13 on the HIF-1 inhibition, we optimized the con-
formations of truncated structures22 using density functional
theory (B3LYP)23 at the 6-31G level (GAUSSIAN 03, D.02 version24).
*
As shown in Figure 5, compound 14 adopted a nearly linear confor-
2002, 6, 1335; (b) Zhang, X.; Kon, T.; Wang, H.; Li, F.; Huang, Q.; Rabbani, Z. N.;
mation, but compound 15 adopted a bent-shaped conformation
remarkably different from that of 14. However, in one of the two
possible orientations for 16 (Overlay II), the conformation was rel-
atively close to that of 14 (Fig. 5c and d)25 indicating that the lin-
ear-shaped conformation resulting from the 2,5-trans-
configuration is critical to the HIF-1 inhibition. Thus, designing a li-
gand that mimics the overall conformation of 1 may improve the
potency and selectivity toward the hypoxia signaling pathway.
Kirkpatrick, J. P.; Vujaskovic, Z.; Dewhirst, M. W.; Li, C. Y. Cancer Res. 2004, 64,
8139; (c) Moeller, B. J.; Cao, Y.; Li, C. Y.; Dewhirst, M. W. Cancer Cell 2004, 5,
429.
12. (a) Giaccia, A.; Siim, B. G.; Johnson, R. S. Nat. Rev. Drug Disc. 2003, 10, 803; (b)
Lin, X.; David, C. A.; Donnelly, J. B.; Michaelides, M.; Chandel, N. S.; Huang, X.;
Warrior, U.; Weinberg, F.; Tormos, K. V.; Fesik, S. W.; Shen, Y. Proc. Natl. Acad.
Sci. U.S.A. 2008, 105, 174; (c) Zhang, H.; Qian, D. Z.; Tan, Y. S.; Lee, K.; Gao, P.;
Ren, Y. R.; Rey, S.; Hammers, H.; Chang, D.; Pili, R.; Dang, C. V.; Liu, J. O.;
Semenza, G. L. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 19579.
13. (a) Rao, K. V.; Alvarez, F. M. Tetrahedron Lett. 1983, 24, 4947; (b) Hodges, T. W.;
Hossain, C. F.; Kim, Y.-P.; Zhou, Y.-D.; Nagle, D. G. J. Nat. Prod. 2004, 67, 767; (c)
Hossain, C. F.; Kim, Y.-P.; Baerson, S. R.; Zhang, L.; Bruick, R. K.; Mohammed, K.
A.; Agarwal, A. K.; Nagle, D. G.; Zhou, Y.-D. Biochem. Biophys. Res. Commun.
2005, 333, 1026.
14. (a) Moeller, B. J.; Dreher, M. R.; Rabbani, Z. N.; Schroeder, T.; Cao, Y.; Li, C. Y.;
Dewhirst, M. W. Cancer Cell 2005, 8, 99; (b) Moeller, B. J.; Batinic-Haberle, I.;
Spasojevic, I.; Rabbani, Z. N.; Anscher, M. S.; Vujaskovic, Z.; Dewhirst, M. W. Int.
J. Radiat. Oncol. Biol. Phys. 2005, 63, 545; (c) Viola, R. J.; Provenzale, J. M.; Li, F.;
Li, C. Y.; Yuan, H.; Tashjian, J.; Dewhirst, M. W. Am. J. Roentgenol. 2008, 191,
1779; (d) Tarnawski, A. S.; Jones, M. K. J. Mol. Med. 2003, 81, 627; (e)
Macpherson, G. R.; Figg, W. D. Cancer Biol. Ther. 2004, 3, 503; (f) Dang, D. T.;
Chen, F.; Gardner, L. B.; Cummins, J. M.; Rago, C.; Bunz, F.; Kantsevoy, S. V.;
Dang, L. H. Cancer Res. 2006, 66, 1684.
In summary, we have shown that 1 downregulated the HIF-1
a
expression and inhibited the secretion of VEGF. We have also dem-
onstrated that the 2,3-cis-3,4-trans-4,5-cis-configuration of the tet-
rahydrofuran is critical to the HIF-1 inhibition of 1. This
conformation–activity relationship study may help to identify
structural motifs required for HIF-1 inhibition and allow structural
modifications to increase specificity and decrease off-target effects.
Acknowledgments
15. Kim, H.; Wooten, C. M.; Park, Y.; Hong, J. Org. Lett. 2007, 9, 3965.
16. Kim, H.; Baker, J. B.; Lee, S.-U.; Park, Y.; Bolduc, K. L.; Park, H.-B.; Dickens, M. G.;
Lee, D.-S.; Kim, Y.-C.; Kim, S. H.; Hong, J. J. Am. Chem. Soc. 2009, 131, 3192.
17. Kim, H.; Kasper, A. C.; Moon, E. J.; Park, Y.; Wooten, C. M.; Dewhirst, M. W.;
Hong, J. Org. Lett. 2009, 11, 89.
We thank Dr. Chuan-Yuan Li (Department of Radiation Oncol-
ogy, University of Colorado Health Sciences Center) for the 4T1-
ODD-Luc. This work was supported by grants from Duke University
(J.H.), Duke Chemistry Undergraduate Summer Research Program
(Y.P.), and National Institutes of Health (NIH PO1 CA42745 and
NIH/NCI CA40355 to M.W.D; NIH R01GM61870-09 to W.Y.). H.K.
gratefully acknowledges the Korea Research Foundation Grant
funded by the Korean Government (MOEHRD) (KRF-2006-352-
E00028) for a postdoctoral fellowship.
18. Wang, G. L.; Semenza, G. L. Blood 1993, 82, 3610.
19. For 11: 1H NMR (400 MHz, CDCl3) d 6.80–7.09 (m, 12 H), 5.13 (d, J = 8.4 Hz, 1
H), 4.63 (d, J = 8.0 Hz, 1 H), 4.62 (d, J = 8.0 Hz, 1 H), 4.42 (d, J = 9.2 Hz, 1 H),
4.06–4.14 (m, 2 H), 4.09 (br s, 2 H), 3.85–3.92 (m, 18 H), 2.23–2.30 (m, 1 H),
1.76–1.84 (m, 1 H), 1.16 (d, J = 6.0 Hz, 3 H), 1.15 (d, J = 6.0 Hz, 3 H), 1.08 (d,
J = 6.4 Hz, 3 H), 0.67 (d, J = 7.2 Hz, 3 H); LRMS (FAB) found 732.4 [calcd for
C42H52O11 (M)+ 732.4].
20. For 13: 1H NMR (400 MHz, CDCl3) d 6.96–7.01 (m, 4 H), 6.91 (s, 2 H), 6.90 (dd,
J = 8.4, 1.2 Hz, 4 H), 6.81 (d, J = 8.0 Hz, 2 H), 4.65 (d, J = 9.2 Hz, 2 H), 4.62 (d,
J = 8.4 Hz, 2 H), 4.05–4.13 (m, 4 H), 3.92 (s, 6 H), 3.86 (s, 6 H), 3.85 (s, 6 H), 1.78–
1.81 (m, 2 H), 1.14 (d, J = 6.4 Hz, 6 H), 1.06 (d, J = 5.6 Hz, 6 H); HRMS (FAB)
found 732.3492 [calcd for C42H52O11 (M)+ 732.3510].
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2009.04.071.
21. Li, F.; Sonveaux, P.; Rabbani, Z. N.; Liu, S.; Yan, B.; Huang, Q.; Vujaskovic, Z.;
Dewhirst, M. W.; Li, C. Y. Mol. Cell 2007, 26, 63.
22. To avoid any unnecessary complications and/or exaggeration by the flexible
side chains, truncated structures of 1, 11, and 13 were used instead of the full
structures. Initial geometries were determined by conformational search based
on molecular mechanics (MMFF).
References and notes
1. Semenza, G. L.. Nat. Rev. Cancer 2003, 56, 721.
23. (a) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785; (b) Vosko, S.; Wilk, L.;
Nusair, M. Can. J. Phys. 1980, 58, 1200; (c) Becke, A. D. Phys. Rev. A 1988, 38,
3098; (d) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
2. Bertout, J. A.; Patel, S. A.; Simon, M. C. Nat. Rev. Cancer 2008, 12, 967.
3. Wang, G. L.; Jiang, B. H.; Rue, E. A.; Semenza, G. L. Proc. Natl. Acad. Sci. U.S.A.
1995, 92, 5510.
24. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.,
et al GAUSSIAN 03, Revision D.02; Gaussian: Wallingford, CT, 2004.
4. Flamme, I.; Frolich, T.; Risau, W. Mol. Pharmacol. 2006, 70, 1469.
5. Lisy, K.; Peet, D. J. Cell Death Differ. 2008, 15, 642.
6. Vaupel, P. Semin. Radiat. Oncol. 2004, 14, 198.
25. Optimization of the full structures of 1, 11, and 13 using density functional
7. (a) Brizel, D. M.; Sibley, G. S.; Prosnitz, L. R.; Scher, R. L.; Dewhirst, M. W. Int. J.
Radiat. Oncol. Biol. Phys. 1997, 38, 285; (b) Hockel, M.; Schlenger, K.; Aral, B.;
Mitze, M.; Schaffer, U.; Vaupel, P. Cancer Res. 1996, 56, 4509.
*
theory (B3LYP) at the 6-31G level provided the same conclusion (see
Supplementary data for details).