Practical synthesis of biaryl colchicinoids containing 3′,4′- catechol ether-based A-rings via Suzuki cross-coupling with ligandless palladium in water

Article abstract of DOI:10.1016/j.tetlet.2003.11.016

Eight new biaryl colchicinoids containing 3′,4′-methylene or benzodioxy ether bridges were synthesized. The key synthetic step employed a ligandless, aqueous Suzuki cross-coupling reaction catalyzed by Pd(OAc) ₂ with tetrabutylammonium bromide (TBAB) and potassium carbonate (K₂CO₃). The biaryl Suzuki products were typically formed in 5-30min and always in less than 1h.

Full text of DOI:10.1016/j.tetlet.2003.11.016

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 803–807
Practical synthesis of biaryl colchicinoids containing 30,40-catechol
ether-based A-rings via Suzuki cross-coupling with ligandless
palladium in waterq
Amy Morin Deveaua,* and Timothy L. Macdonaldb
aChemistry and Physics Department, University of New England, 11 Hills Beach Road, Biddeford, ME 04005, USA
bChemistry Department, University of Virginia, McCormick Road, PO Box 400319, Charlottesville, VA 22904, USA
Received 17 September 2003; revised 5 November 2003; accepted 7 November 2003
Abstract—Eight new biaryl colchicinoids containing 30,40-methylene or benzodioxy ether bridges were synthesized. The key synthetic
step employed a ligandless, aqueous Suzuki cross-coupling reaction catalyzed by Pd(OAc)2 with tetrabutylammonium bromide
(TBAB) and potassium carbonate (K2CO3). The biaryl Suzuki products were typically formed in 5–30 min and always in less than
1 h.
ꢀ 2003 Elsevier Ltd. All rights reserved.
1. Introduction
completed by our laboratory3 that examined the binding
of the combretastatins to the colchicine site on tubulin
has reignited our interest in the efficient syntheses of
tropolone-containing biaryls. Colchicine (Fig. 1, top),
a potent anti-mitotic agent with three unique ring sys-
tems (A, B, and C), is biologically interesting but not
Biaryl heterocycles containing functionalized catechols
often decorate the structures of anti-proliferative, clini-
cally significant natural products like the combretasta-
tins1;2 (Fig. 1, top). A recent molecular modeling study
Figure 1. The structures of compounds that possess anti-mitotic properties.
Keywords: Biaryl; Suzuki coupling; A-ring; Colchicinoid; Colchicine; Combretastatin.
q Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.tetlet.2003.11.016
* Corresponding author. Tel.: +1-207-283-0171x2813; fax: +1-207-294-5956; e-mail: adeveau@une.edu
0040-4039/$ - see front matter ꢀ 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2003.11.016

804
A. Morin Deveau, T. L. Macdonald / Tetrahedron Letters 45 (2004) 803–807
therapeutically useful for cancer chemotherapy because
of its low therapeutic index4 and high in vivo toxicity.5;6
However, colchicine may be abridged to 4.2, its biaryl
cousin containing only the A and C rings, while
retaining most of colchicineÕs anti-mitotic properties.7
Our laboratory remains interested in the rational design
and synthesis of synthetically accessible AC-colchici-
noids to generate new lead compounds for cancer che-
motherapy.
2.1. 20-benzyl alcohol derivatives: methylene- and
benzodioxy functionalized A-rings
Starting from 2,3-dihydroxybenzaldehyde, the catechol
moiety was masked as either the methylenedioxy- (78%
yield) or benzodioxy- (96% yield) aryl ether using the
appropriate dihaloalkane with Cs2CO3 in N,N-dimeth-
ylformamide (DMF) (Scheme 1). Since catechols are
prone to oxidation, we introduced the cyclic aryl ether
early in the sequence to avoid a penultimate protection/
deprotection step with our biaryl colchicinoid products.
Sodium borohydride reduction of the resulting func-
tionalized aldehydes afforded the intermediate benzyl
alcohols 4.32a (90%) and b (88%). Next, these benzyl
alcohols were regiospecifically brominated in the 6-
position using N-bromosuccinimide (NBS) with THF as
the solvent to yield 4.33a (50%) and 4.33b (71%).
Because direct bromination of aldehydes 4.32c–d was
unsuccessful16 (conditions: KBr/Bromine or Br2), the
reduction of the aldehyde was necessary to selectively
brominate at position 6. The subsequent methylation of
benzyl alcohols 4.32a–b with methyl iodide (MeI) yiel-
ded ethers 4.10a (71%), 4.11a (89%), 4.34a (71%), and
4.35a (43%). Alternatively, protection of 4.32a–b as
either a methoxymethyl ether17 (MOMCl, neat DIPEA)
or silyl ether (TBSCl, TEA, DMAP, CH2Cl2) afforded
final A-rings 4.10–4.11a–c and 4.34–4.35a–b, respec-
tively, in good yields (Scheme 1).
Although a full account of our studies with biaryl col-
chicinoids will be enumerated in due course, described
herein is our synthesis of eight novel biaryl colchicinoids
containing A-rings with catechol–ether motifs. Central
to our synthesis was the efficient application of BadoneÕs
aqueous Suzuki cross-coupling8 to form the key C–C
bond between the tropolone core and the A-ring syn-
thon. We found that solvent degassing was of para-
mount importance to the efficacy of the arylation
reaction, and cleanly generated the desired tropolone-
based biaryls usually in 5 min and always in less than
1 h. We believe this expedient Suzuki coupling could be
further optimized for widespread incorporation into
parallel or combinatorial synthetic schemes.9;10
2. Convergent synthesis of AC-based colchicinoid
targets
The tropolone and A-ring halves were each constructed
separately and then married in a final Suzuki coupling.
Following a literature sequence of dibromocyclopropa-
nation,11 dihydroxyllation,12 oxidation,13 base-pro-
moted rearrangement, and methylation, 5-bromo-2-
methoxy-tropolone 4.49b was regioselectively synthe-
sized from 1,4-cyclohexadiene (20% overall yield).14;15
2.2. Arylation of tropolone core: Suzuki coupling
methodology
We implemented the Suzuki methodology to join our
individual A-rings with the tropolone core during syn-
thesis of the desired biaryl colchicinoids. A-rings 4.10–
4.11a–c were first converted to their boronic acid
Scheme 1. Synthesis of 20-substituted benzyl alcohol A-rings containing methylene- and benzodioxy functionalities.

A. Morin Deveau, T. L. Macdonald / Tetrahedron Letters 45 (2004) 803–807
Table 1. Results from Suzuki arylation sequence
805
Product
Yield (%)a
4.66a
4.66b
4.66c
4.67a
4.67b
4.67c
58
33
65
38
70
95
Conditions––(1) aryl halide (1 equiv), n-BuLi (1.1 equiv), THF, )78 ꢁC;
(2) B(OiPr)3 (3 equiv), then NH4Cl workup; (3) ArB(OH)3 (1.01 equiv),
4.49b (1 equiv), Pd(OAc)2 (0.2 equiv), K2CO3 (2.5 equiv) TBAB
(1 equiv), water (distilled, degassed, 0.90 M in boronic acid), 70 ꢁC.
a Yields are unoptimized over the three step sequence of lithiation,
boronic acid formation, and arylation.
Scheme 2. Synthesis of target colchicinoids by arylation of 4.49b.
4.12a–f (Scheme 2). We immediately experienced success
with BadoneÕs ligandless palladium protocol, with our
best yield being 95% (Table 1). To liberate the desired
benzyl alcohols 4.66d (87%) and 4.67d (64%), final col-
chicinoids 4.66–4.67c were treated with tetrabutylam-
monium fluoride (TBAF) in THF (Scheme 2). General
procedures for boronic acid formation and Suzuki
coupling are provided below. All eight colchicinoids
synthesized and their A-ring and tropolone precursors
counterparts via lithium halogen exchange using
n-BuLi,18 and were subsequently quenched at )78 ꢁC
with triisopropyl borate (Scheme 2). After stirring at
room temperature for 12–14 h and acidification to
pH ¼ 5, the crude boronic acid of choice was in hand
and was used immediately without further purification.
Although we experienced much success in metalating the
aforementioned A-rings, it is important to note that we
were unable to lithiate ortho to any methylene- or ben-
zodioxy functionality in examples 4.34 or 4.35a,b. After
exhaustive attempts, we discovered a recent literature
account that summarized and supported our troubles.
Mattson et al. recognized that Ô1,2-methylenedioxy and
-ethylenedioxy groups are poor ortho-directed metala-
tion groupsÕ.19
1
were fully characterized by both H and 13C NMR, as
well as low and/or high resolution mass spec.21
Table 1 summarizes yields achieved for 4.66–4.67a–c
during the three-step sequence of lithiation, boronic acid
formation, and arylation. Because the amount of prod-
uct formed in the Suzuki reaction largely depends on the
quality of the boronic acid used, yields could be
improved by further optimization of the sequence pro-
moting boronic acid formation.22 Additionally, we dis-
covered that both concentration and solvent degassing
were crucial factors for success of the reaction. Success
with the cross-coupling was achieved in 5–10 min when
the solvent was vigorously degassed overnight. The
reaction was also run fairly concentrated (0.90 M); if the
reaction slurry was diluted below ꢀ0.45 M, the coupling
proceeded sluggishly and rarely to completion. Like
most Suzuki couplings, we postulate that the reaction
proceeds through a Pdð0Þ or more reactive PdðIIÞ species.9
Although the precise mechanism involving TBAB
additive is not known, our observations are in accord
with BadoneÕs8 suggestions. Furthermore, our work
represents the first application of ligandless, aqueous
Suzuki conditions to the synthesis of tropolone-con-
taining biaryls. Clearly this method is a logistical
improvement over the Stille23a coupling or other
Suzuki23b;c conditions that were used previously to
generate AC-colchicinoid derivatives.
2.3. Suzuki coupling: catalyst selection and
reaction optimization
Standard Suzuki protocol [Pd(Ph3)4, K2CO3, in DME
with heating for 12–16 h at 70–80 ꢁC] was first imple-
mented to drive the arylation reaction. Although excel-
lent yields were achieved (approximately 80–90%), we
were frustrated with the long reaction times and had
difficulty removing the residual triphenylphosphine-
based by-products. Faced with these challenges, we
discovered BadoneÕs account exploring the use of the
use of catalytic Pd(OAc)2 with tetrabutylammonium
bromide (NBu4Br, TBAB) and K2CO3 in water. Badone
observed dramatic rate enhancement due to presence of
TBAB, from which other groups20 have since benefited.
The advantages to BadoneÕs conditions include: (1)
triphenylphosphine was not a ligand and thus would not
perpetuate purification problems, (2) reaction times
were dramatically decreased from 12+ h to less than 1 h,
(3) high yields were still achieved, and (4) no organic
solvent required: nontoxic, inexpensive water is used as
a solvent, hence greener chemistry is achieved.
3. Aryl boronic acid synthesis
To a solution of aryl halide (1 equiv) in dry THF (1.0 M
in aryl halide) at )78 ꢁC, n-BuLi (1.1 equiv) was added
dropwise over a 20 min period. Care was taken to
Toward the synthesis of 4.66a–c and 4.67a–c, we cou-
pled bromotropolone 4.49b with aryl boronic acids

806
A. Morin Deveau, T. L. Macdonald / Tetrahedron Letters 45 (2004) 803–807
ensure vigorous stirring during BuLi addition. After the
addition of BuLi, the reaction was stirred for 15 min at
)78 ꢁC, then triisopropyl borate (3 equiv) was added in
one portion. The reaction was allowed to warm to room
temperature over 6 h and then stirred for an additional
12 h. The reaction was cooled to 0 ꢁC, then carefully
acidified to pH ¼ 5–6 with ammonium chloride (satd)
and/or HCl (5%). The THF was then removed in vacuo
and the aqueous residue diluted with CH2Cl2. The acidic
aqueous layer was subsequently extracted three times
with CH2Cl2. The combined organics from this extrac-
tion were then dried over MgSO4, filtered, and concen-
trated. The crude boronic acid was used immediately
without further purification.
Supplementary data, including synthetic protocols and
spectral data for 4.66–4.67a–c, is available online with
the paper in ScienceDirect.
References and notes
1. For the anti-mitotic properties of the combretastatins see:
(a) Lin, C. M.; Ho, H. H.; Pettit, G. R.; Hamel, E.
Biochemistry 1989, 28, 6984–6991; (b) Sackett, D. L.
Pharmacol. Therap. 1993, 59, 163–228.
2. For clinical significance of the combretastatins in cancers
see: (a) Hill, S. A.; Tozer, G. M.; Pettit, G. R.; Chaplin, D.
J. Anticancer Res. 2002, 22, 1453–1458; (b) Adams, J.;
Elliott, P. J. Oncogene 2000, 19, 6687–6692; (c) Prise, V. E.;
Honess, D. J.; Stratford, M. R. L.; Wilson, J.; Tozer, G.
M. Int. J. Oncol. 2002, 21, 717–726.
4. Arylation of 5-bromo-2-methoxytropolone via
ligandless Suzuki cross-coupling
3. Brown, M. L.; Rieger, J. M.; Macdonald, T. L. Bioorg.
Med. Chem. 2000, 8, 1433–1441.
4. Malawista, S. E. Arthritis Rheum. 1968, 11, 191–197.
5. Folpini, A.; Furfori, P. J. Toxicol., Clin. Toxicol. 1995, 33,
71–77.
6. For leading references see: Dustin, P. Microtubules;
Springer: New York, 1984.
7. (a) Fitzgerald, T. J. Biochem. Pharmacol. 1976, 25,
1383–1387; (b) Morin Deveau, A. Doctoral Disserta-
tion, University of Virginia, Charlottesville, VA, May
2002.
8. Badone, D.; Baroni, M.; Cardamone, R.; Ielmini, A.;
Guzzi, U. J. Org. Chem. 1997, 62, 7170–7173.
9. For leading Suzuki coupling references see: (a) Miyaura,
N.; Suzuki, A. Chem. Rev. 1995, 95, 2457–2483; (b)
Suzuki, A. J. Organomet. Chem. 1999, 576, 147–168; (c)
Kotha, S.; Lahiri, K.; Kashinath, D. Tetrahedron 2002, 58,
9633–9695.
10. For leading references on solid-phase Suzuki coupling see:
(a) Bork, J. T.; Lee, J. W.; Chang, Y. T. Tetrahedron Lett.
2003, 44, 6141–6144; (b) Brase, S.; Kirchhoff, J. H.;
Kobberling, J. Tetrahedron 2003, 59, 885–939.
11. Last, L. A.; Fretz, E. R.; Coates, R. M. J. Org. Chem.
1982, 47, 3211–3219.
To a 5 mL round bottom flask equipped with a reflux
condenser, 5-bromotropolone (1 equiv), boronic acid
(1.01 equiv), Pd(OAc)2 (0.2 equiv), K2CO3 (2.5 equiv),
NBu4Br (1 equiv), and water (doubly distilled, degassed
for 24 h by bubbling argon gas through, 0.90 M in
boronic acid) were added. The reaction was heated to
70 ꢁC and allowed to stir until TLC analysis indicated
that all tropolone was consumed (usually 5–30 min,
always less than 1 h). The dark brown/black reaction
mixture was cooled to room temperature, diluted with
water, and extracted with CH2Cl2 (3·). The combined
organics were washed with brine and dried over MgSO4
before being filtered through a pad of Celite and con-
centrated in vacuo. Unless otherwise noted, the crude
solid was purified by preparative TLC, and then crys-
tallized from hexanes/EtOAc to afford the desired
functionalized biaryl.
12. VanRheenen, V.; Kelly, R. C.; Cha, D. Y. Tetrahedron
Lett. 1976, 23, 1973–1976.
5. Conclusion
13. Amon, C. M.; Banwell, M. G.; Gravatt, G. L. J. Org.
Chem. 1987, 52, 4851–4855.
14. Banwell, M. G.; Onrust, R. Tetrahedron Lett. 1985, 26,
4543–4546.
15. Banwell, M. G.; Lambert, J. N.; Reum, M. E.; Onrust, R.
Org. Prep. Proced. Int. 1988, 20, 393–399.
16. A complex mixture of starting material and brominated
products was obtained.
17. Stork, G.; Takahashi, T. J. Am. Chem. Soc. 1977, 99,
1275–1276.
18. Alo, B. I.; Kandil, A.; Patil, P. A.; Sharp, M. J.; Sidiqui,
M. A.; Snieckus, V.; Josephy, P. D. J. Org. Chem. 1991,
56, 3763–3768.
19. Mattson, R. J.; Sloan, C. P.; Lockhart, C. C.; Catt, J. D.;
Gao, Q.; Huang, S. J. Org. Chem. 1999, 64, 8004–8007.
20. (a) Bussolari, J. C.; Rehborn, D. C. Org. Lett. 1999, 1,
965–967; (b) Hesse, S.; Kirsch, G. Synthesis 2001, 755–
758.
We contend that ACÕs anti-mitotic properties coupled
with its structural similarity to the combretastatins and
its synthetically accessible core make biaryl colchici-
noids justifiable targets that may generate new leads in
cancer chemotherapy. Overall, we successfully applied
BadoneÕs ligandless Suzuki cross-coupling (Pd(OAc)2,
TBAB, K2CO3, water) to join 5-bromo-2-methoxy-
tropolone and a series of aryl boronic acids functional-
ized with 30,40-methylene- or benzodioxy-ethers (Scheme
2). In this work, eight new AC colchicinoids containing
30,40-catechol ethers were ultimately realized (4.66–
4.67a–d) (Scheme 2 and Table 1). With control of con-
centration and extensive solvent degassing, the time
required for the Suzuki reaction could be lowered to
5 min, but in all cases less than 1 h. We think that both
the use of water as a solvent and the abbreviated reac-
tion time make these Suzuki cross-coupling conditions
attractive for future synthetic applications, such as high-
throughput production of colchicinoids and other
functionalized biaryl heterocycles.
21. Supplementary spectral data accompanies this article
online (ScienceDirect).
22. Importantly, unwanted cross-coupled by-products derived
from residual aryl bromides did not contaminate our final
colchicinoids. If any biaryl by-products were indeed

A. Morin Deveau, T. L. Macdonald / Tetrahedron Letters 45 (2004) 803–807
807
formed, there was a large enough Rf difference such that
they were separable from the tropolone-based biaryls by
chromatography.
eron, J. M.; Collis, M. P.; Crisp, G. T.; Gable, R. W.;
Hamel, E.; Lambert, J. N.; Mackay, M. F.; Reum, M. E.;
Scoble, J. A. Aust. J. Chem. 1991, 44, 705–728; (c)
Banwell, M. G.; Cameron, J. M.; Corbett, M.; Dupuche,
J. R.; Hamel, E.; Lambert, J. N.; Lin, C. M.; Mackay, M.
F. Aust. J. Chem. 1992, 45, 1967–1982.
23. (a) Banwell, M. G.; Collis, M. P.; Crisp, G. T.; Lambert, J.
N.; Reum, M. E.; Scoble, J. A. J. Chem. Soc., Chem.
Commun. 1989, 10, 616–617; (b) Banwell, M. G.; Cam-