Synthesis and biological activities of high affinity taxane-based fluorescent probes

Article abstract of DOI:10.1016/j.bmcl.2008.12.018

Three fluorescent probes 3a, 3b, and 4 have been synthesized through conjugation of fluorescein and difluorescein groups to the 7-OH of C-2 modified paclitaxel and cephalomannine derivatives with very high affinity to microtubules. All these probes exhibi

Full text of DOI:10.1016/j.bmcl.2008.12.018

Bioorganic & Medicinal Chemistry Letters 19 (2009) 751–754
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and biological activities of high affinity taxane-based
fluorescent probes
a,b,
Xuan Li a,b, Isabel Barasoain c, Ruth Matesanz c, J. Fernando Díaz c, , Wei-Shuo Fang
*
*
a Key Lab of Bioactive Substances and Resources Utilization of Chinese Herbal Medicine (PUMC), Ministry of Education, Beijing 100050, China
b Institute of Materia Medica, Chinese Academy of Medical Science, 1 Xian Nong Tan Street, Beijing 100050, China
c Centro de Investigaciones Biológicas, Consejo Superior de Investigaciones Científicas, Ramiro de Maeztu 9, 28040 Madrid, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Three fluorescent probes 3a, 3b, and 4 have been synthesized through conjugation of fluorescein and
difluorescein groups to the 7-OH of C-2 modified paclitaxel and cephalomannine derivatives with very
high affinity to microtubules. All these probes exhibited potent tubulin assembly promotion and tumor
cell killing activities, thus may be useful as tools for the determination of thermodynamic parameters and
exploration of ligand–microtubule interactions.
Received 8 August 2008
Revised 21 November 2008
Accepted 5 December 2008
Available online 10 December 2008
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
Paclitaxel
Cephalomannine
High affinity taxanes
Fluorescence
Fluorescein
Difluorescein
The plant-derived natural product paclitaxel (1a), exerting its
mechanism through disruption of dynamics of tubulin/microtubule
polymerization–depolymerization process,1 has been widely used
for the treatment of various solid tumors.2 In order to study the
interaction of 1a with tubulin/microtubule system and visualize
the microtubule dynamics in the presence of 1a, various spectro-
scopic methods have been applied. One of the most useful
techniques to explore such interactions is fluorescence spectros-
copy,3–5 in which various probes derived from 1a or its derivatives
with fluorophore tags at different positions of taxanes are often
used.6–12
Most of taxane-based fluorescent probes are prepared from 1a,
by the introduction of a fluorophore at C-7 position through proper
linkage. Because C-7 and other ‘‘northern hemisphere” substituents
in 1a are known not to interact with tubulin significantly, so these
probes are believed to minimize the interference of large fluores-
cent groups such as fluorescein and rhodamine in the binding pock-
et. With the fluorescein and difluorescein-tagged 1a, Flutax-1 (2a),
and -2 (2b), it is possible to determine the thermodynamic and
kinetic parameters for the taxane–tubulin binding process,13–15
analyze the ligand’s conformation in D2O/DMSO-d6,16 and locate
the centrosome and spindle pole microtubules as the main targets
for taxanes to induce cell death.11
It has been found that precise determinations of thermodynamic
parameters for ligand–tubulin binding is critical to quantify the
enthalpic and entropic contributions to the binding process, thus
predict the cytotoxicity of the ligands and help the design of high
affinity ligands (i.e. taxanes) through the binding affinity dissec-
tion.17–19 During the process of the determination of thermody-
namic parameters by the competitive displacement method,
fluorescent probes with binding affinity within three orders of
magnitudes to the targeted ligands are usually applied. However,
displacement methods cannot distinguish ligands with affinities
much higher than probes (e.g. 2a or 2b in our case) thus it is not
possible to determine the binding affinity or thermodynamic
parameters precisely.20 The thermodynamic parameter of binding
of compounds with high affinity for the taxane site have been deter-
mined by displacement of a non-fluorescent compound, Epothilone
B, with high affinity (Ka 7.5 ꢀ 108 Mꢁ1 at 35 °C) for the paclitaxel
17,19
site.
In this method, the Epothilone B concentrations in super-
* Corresponding author. Tel.: +86 10 63165237; fax: +86 10 63017757.
E-mail addresses: fer@cib.csic.es (J.F. Díaz), wfang@imm.ac.cn (W.-S. Fang).
natants and pellets required to measure the binding constant of the
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.12.018

752
X. Li et al. / Bioorg. Med. Chem. Lett. 19 (2009) 751–754
The final products 325,26 and 4,27 after chromatographic separa-
tions, are still contaminated by ca. 20–55% weight of NaOAc
(ca. 1:20–1:4 taxane–NaOAc peak ratios in the corresponding NMR
spectra). Since these impurities do not interfere with the parameter
determination, the probes are not subjected to further purifications.
Binding to the paclitaxel site in microtubules. The compounds
were tested for their ability to bind tubulin in the paclitaxel
site. All of them, upon binding to the paclitaxel site, underwent a
large increase in the anisotropy of their fluorescence as is the case
for other fluorescent derivatives of paclitaxel13 (Table 1).
competitor have to be determined employing HPLC, thus the meth-
od has a low throughput and is time consuming. Hence, a high affin-
ity fluorescent taxane is desirable for the parameter determination
by fluorescence spectroscopy during the optimization of the bind-
ing affinity of taxanes to tubulin.
The introduction of methoxy or azido group to the meta position
of C-2 benzoate phenyl in 1a and its analogs has been proved to en-
hance the tubulin polymerization promotion ability and cytotoxic-
ity significantly.22,23 It was also known that the binding affinity of
taxanes to tubulin/microtubule can be potentiated by such a mod-
ification.15,19 So, we prepared fluorescent probes 3 and 4 from
2-(m-MeO)-benzoyl paclitaxel and 2-(m-N3)-benzoyl cephaloman-
nine which are higher affinity tubulin binders than 1 and 2, and
investigated their fluorescence properties and biological activities.
Chemical syntheses. The synthesis of 3 was performed in a sim-
ilar way to that of 2-modified taxanes,19,21 except the 7-TES group
was replaced with 7-alanine (7-Ala).
As expected for the introduction of binding enhancement
groups15,19 in 1a, docetaxel (1b) does not displace them from the
crosslinked microtubules (no decrease in the anisotropy of binding
is observed after addition of 10 lM docetaxel) as in the case for
Table 1
Anisotropy of the fluorescence emission of the fluorescent taxoids in different statusa
Paclitaxel (1a) was first protected at 20-OH and then conjugated
with Boc-Ala-OH at 7-OH to afford 5. Compound 5, upon 2-deben-
zoylation and subsequent acylation, was converted to 2-(m-MeO-
Bz)-paclitaxel analog 6. Deprotection of 20-O-tert-butyldimethylsi-
lyl and N-tert-butyloxycarbonyl groups under mild conditions
furnished 7-Ala-2-modified paclitaxel analog 7, with free amino
group ready for the conjugation of fluorescent groups. Coupling
Ligands
Free
0.05 0.02
0.04 0.01
0.046 0.003
0.046 0.001
0.059 0.010
Boundb
Displacedb
2ac
2bc
3a
3b
4
0.24 0.02
0.26 0.02
0.168 0.007
0.202 0.031
0.291 0.020
0.05 0.02
0.08 0.01
0.052 0.008
0.068 0.010
0.077 0.010
a
Determined with fluorescent ligands (50 nM) in GAB buffer at 25 °C. Excitation
and emission wavelengths for 3a and 3b are 492 and 524 nm, and for 4 493 and
of
7 with 5-fluorescein succinimidyl ester (5-FAM SE) or
5-difluorescein succinimidyl ester (OG 488 SE) in sodium carbon-
ate–dicarbonate buffer (pH 9.0) furnished the desired conjugates
3a (20% yield) and 3b (60% yield) (Scheme 1).
The fluorescent probe 4 was also prepared from cephaloman-
nine similarly in 14% overall yield for seven steps.
521 nm, respectively.
b
Bound to 1
l
M of paclitaxel binding sites in stabilized microtubules by cross-
M discodermolide.
Data from Ref. 13. Bound to microtubules assembled from 20 M pure tubulin
in GAB, and further displaced by the addition of 50 M docetaxel.
linking, and further displaced by the addition of 10
l
c
l
l
Scheme 1. Reagents and conditions: (i) TBSCl, imidazole, DMF, 70 °C, 97%; (ii) Boc-Ala-OH, DCC, DMAP, toluene, 35 °C, 86%; (iii) Triton B, CH2Cl2, ꢁ20 °C, 77%; (iv) m-methoxy
benzoic acid, DCC, 4-PP, 65 °C, 73%; (v) 40% aq HF, Py, CH3CN, 35 °C, 97%; (vi) CF3COOH, PhSCH3, CH2Cl2, 0 °C, 61%; (vii) 5-FAM SE or OG 488 SE, DMF, sodium carbonate–
dicarbonate buffer (pH 9.0), rt, 20% for 3a, 60% for 3b.

X. Li et al. / Bioorg. Med. Chem. Lett. 19 (2009) 751–754
753
Figure 1. Fluorescence spectra of 4 (50 nM) black line, 4 plus microtubules (1
l
M),
Figure 2. Visualization of microtubules with 4 in A549 human lung carcinoma
dotted line and 4, microtubules plus discodermolide (10
lM), gray line in GAB
cells. A549 unfixed microtubule cytoskeletons were incubated with 1
5 min. Inset: 4 stained mitotic spindle from the same preparations. Scale bar
represents 10 m for both images.
lM 4 for
buffer.
l
Flutax-1 (2a) and Flutax-2 (2b).13 Their binding constants to micro-
tubules were also measured, and found to be 4.4 1.9 ꢀ 108,
6.8 1.7 ꢀ 108, and 12 4 ꢀ 108 Mꢁ1 for 3a, 3b, and 4, respectively.
All these facts indicate that their binding affinities are much higher
than that of docetaxel. The specificity of the binding was checked
with them and imaged by fluorescence microscopy as with other
fluorescent taxanes12 (Fig. 2). It was found that they readily and
brightly stained interphasic cellular microtubules as well as mito-
tic spindles, centrosomes and the central part of the midbody in
cytokinesis.
by adding 10 lM discodermolide, the natural paclitaxel binding site
ligand with the highest affinity known to date, to the samples con-
taining the fluorescent taxanes bound to stabilized microtubules.
Discodermolide displaces the bound taxane, resulting in a decrease
of the anisotropy of the fluorescence of the sample to the values
observed for the free compound. The same effect can be shown by
fluorescence intensity (Fig. 1). Addition of microtubules to the solu-
tion of 4 led to the change of its fluorescence spectra, with a small
increase in fluorescence intensity, while addition of discodermolide
to this solution restored the spectra, due to the displacement of fluo-
rescent probe 4 from its binding site in microtubule.
Microtubule assembly induction. The compounds were then
tested for their ability to induce microtubule assembly in condi-
tions which tubulin do not assemble.24 The compounds were able
to induce assembly of GTP-tubulin into microtubules, indicating
that the introduction of the fluorescent group does not perturb
their tubulin promotion activity (for the critical concentration data,
see Table 2). All three high affinity taxane-based fluorescent probes
were found to be tubulin assembly promoters comparable with the
more potent taxanes (including the clinically used taxanes, paclit-
axel, and docetaxel, as well as two fluorescent taxanes (2a and 2b)
previously employed12,20 (supplementary information of Ref. 19).
Among the three probes, 3b is the more active one, although the
differences are in within the experimental errors of the docetaxel
measurement.
Cytotoxicities were tested on A2780 ovarian carcinoma cells
and their P-gp overexpressing counterpart A2780AD cells. All three
newly synthesized compounds were able to kill the non-resistant
cells (Table 3). However, despite having higher affinity to microtu-
bules than paclitaxel (they cannot be displaced from crosslinked
stabilized microtubules by a 200 times excess of docetaxel), their
cytotoxicity is much lower than paclitaxel and docetaxel (Table
3) and their unlabeled partners,19 indicating that the polar fluores-
cent moiety impedes their cell permeability. The poor cytotoxicity
for 4, 2a, and 2b in P-glycoprotein overexpressing A2780AD cells
(which is not the case for their unlabeled partners19) may suggest
that the introduction of the fluorescent moiety usually makes them
better substrates for membrane transporter proteins such as P-gly-
coprotein. However, Fchitax-2 (3a) and Fchitax-3 (3b) are more ac-
tive in resistant cells (R/S index of 36 and 26, respectively), than
their parent compound Chitax-11 (R/S index 120). It is also inter-
esting to note that cytotoxicities of the difluorescein bearing
probes (2b and 3b) are less active than fluorescein bearing ones
(2a and 3a), although the former are slightly more potent tubulin
polymerization promoters.
In conclusion, all three fluorescent probes can be applied to the
determination of binding affinity or thermodynamic parameters
for taxanes with much higher affinity than 1. In addition, 3a and
Ligand binding to cellular microtubules and cytotoxicity. To test
if these fluorescent probes can bind to native cellular microtubules,
we incubated unfixed A549 human lung carcinoma cytoskeletons
Table 3
Cytotoxicity of the compounds against drug-sensitive and -resistant cells
Ligands
A2780a (nM)
A2780AD
R/Sb
Table 2
4
220 15
19.5 0.35
272 12
1.3 0.4
0.6 0.2
256 17
800 28
>20,000
>91
36.7
26.4
753
483
>78
>25
Critical concentrations of the ligand-induced assembly reactionsa
3a
3b
1a
1b
2a
2b
716 120
7200 200
980 149
290 16
>20,000
Ligands
Critical concentrations
M)
Ligands
Critical concentrations
(lM)
(l
4
3a
3b
2.45 0.41
2.26 0.40
2.17 0.39
1ab
1bb
2a
4.23 0.68
2.50 0.37
3.84 0.56
3.22 0.46
>20,000
a
IC50 values were determined in parental ovarian carcinoma A2780 cells and its
drug resistant counterpart A2780AD cells after 48 h incubation in the presence of
the drugs using an MTT cell proliferation assay.15
2b
a
Determined in the presence of a 10% excess of ligand in 10 mM phosphate, 1
mM EDTA, 4 mM MgCl2, pH 6.7 solution in the presence of 1 mM GTP.17
b
The resistance factor is calculated dividing the IC50 of the resistant cell line by
b
Data from Ref. 17.
the IC50 of the drug sensitive cell line.

754
X. Li et al. / Bioorg. Med. Chem. Lett. 19 (2009) 751–754
18. Buey, R. M.; Díaz, J. F.; Andreu, J. M.; O’Brate, A.; Ginnakakou, P.; Nicolaou, K. C.;
Sasmal, P. K.; Ritzen, A.; Namoto, K. Chem. Biol. 2004, 11, 225.
19. Matesanz, R.; Barasoain, I.; Yang, C.-G.; Lei Wang, L.; Li, X.; de Inés, C.; Coderch,
C.; Gago, F.; Jiménez-Barbero, J.; Andreu, J. M.; Fang, W.-S.; Díaz, J. F. Chem. Biol.
2008, 15, 573.
3b may also be useful in the exploration of ligand–microtubule
interactions at cellular level.
Acknowledgments
20. Diaz, J. F.; Andreu, J. M. Biochemistry 1993, 32, 2747–2755.
21. Yang, C.-G. Ph.D. Thesis, Peking Union Medical College, June 2007.
22. Chaudhary, A. G.; Gharpure, M. M.; Rimoldi, J. M.; Chordia, M. D.; Gunatilaka, A.
A. L.; Kingston, D. G. I.; Grover, S.; Lin, C. M.; Hamel, E. J. Am. Chem. Soc. 1994,
116, 4097.
23. Kingston, D. G. I.; Chaudhary, A. G.; Chordia, M. D.; Gharpure, M.; Gunatilaka, A.
A. L.; Higgs, P. I.; Rimoldi, J. M.; Samala, L.; Jagtap, P. G.; Giannakakou, P.; Jiang,
Y. Q.; Lin, C. M.; Hamel, E.; Long, B. H.; Fairchild, C. R.; Johnston, K. A. J. Med.
Chem. 1998, 41, 3175.
We thank Carne Sierra Madrid S.A. (CIF: A78074168) for provid-
ing the calf brains for tubulin purification, and Dr. Ying Zhao and
Ms. Xi Xiao for their assistance in the manuscript preparation. This
work was supported by NSFC (20572135 and 30630069), MOST
grants (2006DFA31490) to W.S.F., and BIO2007-61336 from MEC
and BIPPED-CM (S2006/BIO-0214) grants to J.F.D.
24. Díaz, J. F.; Buey, R. M. Microtubule Protocols. In Zhou, J., Ed.; Humana Press Inc.:
Totowa, NJ, 2007; pp 245–260.
References and notes
25. Spectral data for 3a. 1H NMR (500 MHz, CD3OD): d 8.41 (1H, d, J = 2.0 Hz), 7.96
(1H, dd, J = 1.5, 7.5 Hz), 7.80 (2H, m), 7.62 (1H, m), 7.59 (1H, m), 7.50–7.16
(11H, m), 6.96 (2H, dd, J = 2.0, 9.0 Hz), 6.47 (2H, dd, J = 2.0, 9.0 Hz), 6.44 (2H,
m), 6.31 (1H, s), 6.09 (1H, t, J = 9.0 Hz), 5.62 (1H, dd, J = 7.5, 11.0 Hz), 5.60 (1H,
d, J = 7.0 Hz), 5.56 (1H, d, J = 5.5 Hz), 4.96 (1H, d, J = 10.0 Hz), 4.71 (1H, d,
J = 6.0 Hz), 4.59 (1H, q, J = 7.5 Hz), 4.20 (1H, d, J = 8.5 Hz), 4.14 (1H, d,
J = 8.0 Hz), 3.87 (1H, d, J = 7.0 Hz), 3.85 (3H, s), 2.56 (1 H, ddd, J = 5.5, 8.0,
15.0 Hz), 2.30 (3H, s), 2.17 (1H, dd, J = 9.5, 15.5 Hz), 2.13 (3H, s), 1.93 (1H, dd,
J = 9.5, 15.5 Hz), 1.86 (3H, s), 1.79 (1H, overlapped), 1.76 (3H, s), 1.38 (3H, d,
J = 7.0 Hz), 1.10 (3H, s), 1.07 (3H, s). ESI-HRMS: m/z [M+H]+ 1313.42439
(found), 1313.43419 (calcd).
1. Jordan, M. A.; Wilson, L. Nat. Rev. Cancer 2004, 4, 253.
2. TAXOL Science and Applications; Suffness, M., Ed.; CRC Press Inc.: Boca Raton, FL,
1995; pp 1–448.
3. Sengupta, S.; Boge, T. C.; Liu, Y.; Hepperle, M.; Georg, G. I.; Himes, R. H.
Biochemistry 1997, 36, 5179.
4. Li, Y.; Poliks, B.; Cegelski, L.; Poliks, M.; Gryczynski, Z.; Piszczek, G.; Jagtap, P. G.;
Studelska, D. R.; Kingston, D. G. I.; Schaefer, J.; Bane, S. Biochemistry 2000, 39, 281.
5. Shanker, N.; Kingston, D. G. I.; Ganesh, T.; Yang, C.; Alcaraz, A. A.; Geballe, M. T.;
Banerjee, A.; McGee, D.; Snyder, J. P.; Bane, S. Biochemistry 2007, 46, 11514.
6. Dubois, J.; Le, G. M. T.; Gueritt-Voegelein, F.; Guenard, D.; Tollon, Y.; Wright, M.
Bioorg. Med. Chem. 1995, 3, 1357.
7. Souto, A. A.; Acuna, A. U.; Andreu, J. M.; Barasoain, I.; Abal, M.; Amat-Guerri, F.
Angew. Chem. Int. Ed. Engl. 1995, 34, 2710.
8. Guy, R.; Scott, Z.; Sloboda, R.; Nicolaou, K. C. Chem. Biol. 1996, 3, 1021.
9. Rao, C. S.; Chu, J. J.; Liu, R. S.; Lai, Y. K. Bioorg. Med. Chem. 1998, 6, 2193.
10. Baloglu, E.; Kingston, D. G. I.; Patel, P.; Chatterjee, S. K.; Bane, S. L. Bioorg. Med.
Chem. 2001, 11, 2249.
11. Abal, M.; Souto, A. A.; Amat-Guerri, F.; Acuña, A. U.; Andreu, J. M.; Barasoain, I.
Cell Motil. Cytoskeleton 2001, 49, 1.
12. Evangelio, J. A.; Abal, M.; Barasoain, I.; Souto, A. A.; Lillo, M. P.; Acuna, A. U.;
Amat-Guerri, F.; Andreu, J. M. Cell Motil. Cytoskeleton 1998, 39, 73.
13. Díaz, J. F.; Strobe, R.; Engelborghs, Y.; Souto, A. A.; Andreu, J. M. J. Biol. Chem.
2000, 275, 26265.
14. Díaz, J. F.; Barasoain, I.; Andreu, J. M. J. Biol. Chem. 2003, 278, 8407.
15. Yang, C. G.; Barasoain, I.; Li, X.; Matesanz, R.; Liu, R.; Sharom, F. J.; Yin, D.-L.;
Díaz, J. F.; Fang, W.-S. ChemMedChem 2007, 2, 691.
16. Jiménez-Barbero, J.; Souto, A. A.; Abal, M.; Barasoain, I.; Evangelio, J. A.; Acuña,
A. U.; Andreu, J. M.; Amat-Guerri, F. Bioorg. Med. Chem. 1998, 6, 1857.
17. Buey, R. M.; Barasoain, I.; Jackson, E.; Meyer, A.; Ginnakakou, P.; Paterson, I.;
Mooberry, S.; Andreu, J. M.; Díaz, J. F. Chem. Biol. 2005, 12, 1269.
26. Spectral data for 3b. 1H NMR (500 MHz, CD3OD): d 8.45 (1H, d, J = 1.5 Hz), 8.00
(1H, dd, J = 2.0, 8.0 Hz), 7.80 (2H, m), 7.63 (1H, m), 7.58 (1H, m), 7.48–7.16
(11H, m), 6.66 (2H, dd, J = 3.5, 6.5 Hz), 6.58 (2H, d, J = 7.5 Hz), 6.31 (1H, s), 6.07
(1H, t, J = 8.5 Hz), 5.61 (1H, dd, J = 7.5, 10.5 Hz), 5.59 (1H, d, J = 7.0 Hz), 5.55
(1H, d, J = 5.5 Hz), 4.96 (1H, d, J = 9.0 Hz), 4.72 (1H, d, J = 6.0 Hz), 4.61 (1H, q,
J = 7.0 Hz), 4.20 (1H, d, J = 8.0 Hz), 4.14 (1H, d, J = 8.0 Hz), 3.87 (1H, d,
J = 7.0 Hz), 3.84 (3H, s), 2.57 (1H, ddd, J = 5.5, 8.0, 15.0 Hz), 2.29 (3H, s), 2.15
(1H, dd, J = 9.5, 15.5 Hz), 2.12 (3H, s), 1.90 (1H, dd, J = 9.5, 15.5 Hz), 1.87 (3H, s),
1.79 (1H, overlapped), 1.77 (3H, s), 1.39 (3H, d, J = 7.5 Hz), 1.10 (3H, s), 1.07
(3H, s). ESI-HRMS: m/z [M+H]+ 1349.40696 (found), 1349.41535 (calcd).
27. Spectral data for 4. 1H NMR (500 MHz, CD3OD): d 8.44 (1H, s), 8.08 (1H, dd,
J = 2.0, 8.0 Hz), 7.84 (1H, d, J = 7.5 Hz), 7.73 (1H, m), 7.52 (1H, t, J = 8.0 Hz),
7.37–7.19 (7H, m), 6.79 (2H, d, J = 9.0 Hz), 6.62 (2H, d, J = 2.0 Hz), 6.53 (2H, dd,
J = 2.0, 9.0 Hz), 6.38 (1H, q, J = 7.0 Hz), 6.32 (1H, s), 6.06 (1H, t, J = 9.0 Hz), 5.62
(1H, d, J = 7.0 Hz), 5.61 (1H, dd, J = 6.5, 11.0 Hz), 5.42 (1H, d, J = 4.5 Hz), 4.96
(1H, d, J = 9.0 Hz), 4.61 (1H, d, J = 4.5 Hz), 4.58 (1H, q, J = 7.0 Hz), 4.19 (1H, d,
J = 8.0 Hz), 4.13 (1H, d, J = 8.0 Hz), 3.88 (1H, d, J = 7.0 Hz), 2.56 (1H, ddd, J = 5.5
7.5, 14.5 Hz), 2.32 (3H, s), 2.21 (1H, dd, J = 9.5, 15.5 Hz), 2.13 (3H, s), 2.00 (1H,
dd, J = 9.5, 15.5 Hz), 1.86 (3H, s), 1.79 (1H, overlapped), 1.78 (3H, s), 1.76 (3H,
s), 1.71 (3H, d, J = 7.0 Hz), 1.38 (3H, d, J = 7.0 Hz), 1.12 (3H, s), 1.08 (3H, s). ESI-
HRMS: m/z [M+H]+ 1302.44347 (found), 1302.44068 (calcd).