A versatile synthesis of "tafuramycin A": A potent anticancer and parasite attenuating agent

Article abstract of DOI:10.1039/c4ob00842a

An improved and versatile synthesis of tafuramycin A, a potent anticancer and parasite-attenuating agent, is reported. The three major improvements that optimized yield, simplified purification and allowed the synthesis of more versatile duocarmycin analogues are: a first-time reported regioselective bromination using DMAP as catalyst; the control of the aryl radical alkene cyclization step to prevent the dechlorination side reaction; and the design of a new protection/deprotection method to avoid furan double bond reduction during the classical O-benzyl deprotection in the final step. This alternative protection/deprotection strategy provides ready access to duocarmycin seco-analogues that carry labile functionalities under reducing reaction conditions. Tafuramycin A (3) was prepared in either 8 steps from intermediate 6 or 7 steps from intermediate 17 in 52% or 37% yield respectively. Our strategy provides a significant improvement on the original procedure (11% overall yield) and greater versatility for analogue development. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4ob00842a

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
A versatile synthesis of “tafuramycin A”: a potent
anticancer and parasite attenuating agent†
Ibrahim M. El-Deeb,*a Faith J. Rose,a Peter C. Healyb and Mark von Itzstein*a
Cite this: Org. Biomol. Chem., 2014,
12, 4260
An improved and versatile synthesis of tafuramycin A, a potent anticancer and parasite-attenuating agent,
is reported. The three major improvements that optimized yield, simplified purification and allowed the
synthesis of more versatile duocarmycin analogues are: a first-time reported regioselective bromination
using DMAP as catalyst; the control of the aryl radical alkene cyclization step to prevent the dechlorination
side reaction; and the design of a new protection/deprotection method to avoid furan double bond
reduction during the classical O-benzyl deprotection in the final step. This alternative protection/
deprotection strategy provides ready access to duocarmycin seco-analogues that carry labile functional-
ities under reducing reaction conditions. Tafuramycin A (3) was prepared in either 8 steps from intermedi-
ate 6 or 7 steps from intermediate 17 in 52% or 37% yield respectively. Our strategy provides a significant
improvement on the original procedure (11% overall yield) and greater versatility for analogue
development.
Received 17th March 2014,
Accepted 6th May 2014
DOI: 10.1039/c4ob00842a
www.rsc.org/obc
Introduction
Duocarmycins, exemplified by duocarmycin SA (DSA, 1), are a
group of highly potent anticancer alkaloids separated from
Streptomyces sp. (Fig. 1).1–3 They exert their cytotoxic effect
through binding at the minor groove of AT-rich sequences in
the DNA and by covalently alkylating adenine-N3.2–4 In spite of
their high potency, all natural duocarmycins failed to reach
the market as anticancer drugs because of their severe toxicity,
particularly to bone marrow and liver.5–7 Over the last two
decades, numerous synthetic analogues of duocarmycins
have been developed that introduce functionalization at
both its DNA-alkylating unit (part A) and DNA-binding unit
(part B).8–11 These analogues have enabled the development of
a SAR profile of this compound class to facilitate the discovery
of more selective and less toxic derivatives that may have
clinical application (for detailed review see ref. 1–3).
Fig. 1 Structures of duocarmycin SA (1), centanamycin (2), tafuramycin
A (3) and CFI-TMI (4).
A number of these reported modifications were successful
in simplifying the structure of duocarmycins while maintain-
ing high potency, for example centanamycin (2).9,12 In
addition to its highly potent anticancer activity, centanamycin
also displayed a remarkable antiparasitic activity both in vitro
and in vivo against a number of protozoan parasites, such
as different Plasmodium species.13,14 However, the potential
in vivo toxicities of such compounds remain a limiting
factor for their clinical use, either as an anticancer or as an
antiparasitic agent. Nevertheless, a recent study on centana-
mycin opened the door to a potential use of the compound,
aInstitute for Glycomics, Griffith University, Gold Coast Campus, Queensland, 4222,
Australia. E-mail: i.el-deeb@griffith.edu.au, m.vonitzstein@griffith.edu.au
bQueensland Micro and Nanotechnology Centre, Griffith University, Nathan Campus,
Queensland 4111, Australia
†Electronic supplementary information (ESI) available: Complete experimental
procedures, characterization data, 1H and 13C NMR spectra of new intermediates
and final compounds and CIF with details of X-ray data collection and refine-
ment for compound 14. CCDC 992249. For ESI and crystallographic data in CIF
or other electronic format see DOI: 10.1039/c4ob00842a
4260 | Org. Biomol. Chem., 2014, 12, 4260–4264
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
regardless of its in vivo toxicity, where it was successfully used provide 8 in excellent yields at a reaction temperature of 25 °C
as a chemical attenuating agent for the production of a candi- (81%) or 0 °C (92%), in relatively short reaction time (1 h).
date malaria vaccine.13 It has been shown that centanamycin
The allylation of intermediate 8 with (E/Z)-1,3-dichloropro-
is effective in attenuating Plasmodium berghei both in vitro pene, following standard reaction conditions,15 provided the
and in vivo, and mice vaccinated with in vitro treated key intermediate 9 in an excellent yield (96% of a mixture of
sporozoites were protected against infection with wild type Z : E isomer ≡ 6 : 4 cf. 80% previously reported15). This
sporozoites.13
improved yield is most likely due to the greater purity of the
Following the discovery of centanamycin and its successful starting material as a consequence of our optimized bromina-
use as a parasite attenuating agent in vaccine development, tion conditions that yield selectively and exclusively the mono-
tafuramycin A (3)15 displayed superior effect compared with bromo product 8.
centanamycin. This discovery is now the subject of a patent in
Cyclization of 9 in toluene, in the presence of AIBN and
which 3 is claimed as a parasite attenuating agent in the pro- Bu3SnH, has been reported to yield a mixture of two com-
duction of a whole parasite malaria vaccine.16 A single dose of pounds; the required product 10 from a 5-exo-trig cyclization
tafuramycin A-attenuated parasite was found to be sufficient to (obtained in 40% yield) and a seco-cyclopropyl-tetrahydro-
protect the animals against subsequent malaria infection by furano[2,3-f]quinoline (seco-CFQ, Fig. 2) that resulted from
the same isolate, strain or species of Plasmodium used in the 6-endo-trig cyclization (obtained in 48% yield).15,28 A limited
immunogenic composition, or by one or more heterologous number of reports on a similar cyclization reaction that pro-
isolates, strains or species of Plasmodium.17
vides the thermodynamically less favored 6-membered quino-
Tafuramycin A (3)15 is a seco-prodrug that dehydrochlori- line have been published.29–32 In these reports, the formation
nates inside the cell to generate the active cyclopropane- of the less favored 6-membered quinoline structure is the
containing drug that subsequently alkylates DNA. It’s also an result of rearrangement of the 5-membered indoline as a con-
isoform of the previously reported18,19 furanoindoline ana- sequence of nucleophilic attack by anions such as Cl−, NC−
logue of duocarmycin SA (CFI-TMI, 4). Herein, we describe a and MeO−.31
versatile synthesis of tafuramycin A that overcomes a number
Accordingly, the second challenge in this synthesis was to
of difficulties observed in the original strategy. Furthermore, control the conditions of the cyclization step so as to exclu-
our approach allows for a high yielding, large-scale production sively obtain the required product 10, or at least to maximize
of 3 and more versatility that enables the introduction of new its yield. In our hands; the typical reaction conditions reported
functionality.
for this cyclization,15 led to a mixture of two products. Charac-
terization of each of the mixture’s components by spectro-
scopic analysis confirmed one of the products as the intended
chloromethyl furanoindole derivative 10, while the other was
determined to be its dechlorination product (methyl furanoin-
dole derivative, 13).
Results and discussion
Our approach towards the synthesis of tafuramycin
A
Surprisingly, no 6-membered furanoquinoline isomer was
(Scheme 1) utilized readily synthesized intermediates 5–7,15 observed in several trials, under various conditions. Further
with 7 further purified by crystallization from hexanes. The support for the structure of the dechloro-derivative 13 was pro-
first synthetic challenge was encountered during the regio- vided by X-ray crystal structure determination of its sub-
selective bromination of intermediate 7 to provide the mono- sequent coupled product 14 as shown in Fig. 3 (CCDC
bromo derivative 8. The reported15 reaction conditions for 992249).
bromination, i.e. NBS in anhydrous THF–MeOH and p-toluene-
The dechlorination product 13 was presumably the conse-
sulfonic acid (TsOH) at −78 °C, provides 8 in only 55% yield. quence of the excess (2.7 eq.) Bu3SnH used in the reaction. In
Despite several attempts, no improvement in yield was the presence of only 1.0 eq. of Bu3SnH we found that 10 was
observed by following similar conditions. Furthermore, the iso- the only reaction product observed. The product yield was
lated product was made up of an inseparable mixture of the however unsatisfactory (∼20%), with the majority of starting
mono- and dibromo derivatives.
material 9 unreacted. This is most likely why 2.7 eq. of Bu3SnH
Similar reported regioselective brominations have employed has been used in published procedures15 as it forces the reac-
various reaction conditions at extremely low reaction tempera- tion to completion, despite the formation of the unwanted
tures, ranging from −60 °C20 to −78 °C,21–24 in the absence of side product 13.
light and all used acid catalysis.20–24 Alternatively, the more
In our approach to 10 we sought to minimize the formation
expensive iodinating reagent NIS has also been used in place of 13 and to this end we explored a variety of conditions
of NBS.25–27 Even with NIS, acid catalysis was required, and the including varying the amount of AIBN and total solvent
reaction temperatures ranged typically between −20 °C27 and volume. Increasing the amount of AIBN had little influence on
−40 °C.25 Based on these reports, a number of different reac- reaction rate and product yield, while remarkable effects were
tion conditions, including varied solvent composition, catalyst observed by varying the solvent volume. Thus, reduction of the
and reaction temperature employed were evaluated. Both acid solvent volume led to an increase in reaction rate and
and base catalysts were tested and DMAP was shown to improved product yield. Further optimization through solvent
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 4260–4264 | 4261

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 1 Improved synthesis of tafuramycin A (3) and its methyl furanoindole analogue (15).
Fig. 2 Structure of the 6-membered furanoquinoline derivative (seco-
CFQ).
(toluene) volume adjustment to give 9 in a concentration
Fig. 3 X-ray crystal structure of 14.
0.2 mmol mL−1, in the presence of Bu3SnH (1 eq.), resulted in
formation of the desired product 10 in 89% isolated yield in
1 h. Interestingly, and as shown in Scheme 1, refluxing a
similar molar concentration of 9 in toluene with excess
Bu3SnH (3.0 eq.) for 1 h yielded the dechloro-product 13
(91%).
Intermediate 10 was then smoothly converted to 3 using
published conditions.15 Thus, deprotection of 10 and coupling
with 5,6,7-trimethoxy-1H-indole-2-carboxylic acid (TMI),33
yielded the O-benzyl protected amide 11. Finally 11 was
4262 | Org. Biomol. Chem., 2014, 12, 4260–4264
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 2 Alternative synthesis of tafuramycin A (3).
deprotected by 10% Pd/C catalysed reduction to yield the hydrolysis and Curtius rearrangement provided the Boc-
target compound 3. Unfortunately, under these catalytic protected amine 18 in 75% yield. The intermediates 19 and 20
reduction conditions, overreduction of 3 was observed and were prepared by analogous reaction conditions (Scheme 1)
afforded the dihydro compound 12 as a byproduct. Numerous employed for the synthesis of compounds 8 and 9 respectively.
significant variations of reduction conditions (e.g. changing Employing our optimized cyclization conditions (Scheme 1)
the catalyst, solvent, reaction time or temperature) failed to the key intermediate 21 was isolated (∼35%), leaving the
improve this outcome. Furthermore, the ratio of the two majority of starting material unreacted. Increasing AIBN
products (3 and 12) was variable and uncontrollable), and the equivalents from 0.1 to 1.0 improved the reaction yield (86%)
resultant mixture could be only resolved using HPLC (see within 2 h. Finally, removal of the Boc and PMB groups by
Experimental section). Similarly, Boc-deprotection of 13, coup- treatment with HCl in THF afforded an intermediate with
ling of the product with TMI to yield 14 and benzyl deprotec- free hydroxyl and amino groups that was coupled with TMI to
tion by catalytic reduction yielded a HPLC-separable mixture afford tafuramycin A (3) in 69% yield.
of 15 and its dihydro-derivative 16 (Scheme 1).
In order to avoid overreduction of 3 during benzyl deprotec-
tion, an alternative protection/deprotection strategy was inves-
tigated. The use of acetate or other ester-protecting group was
Conclusions
not possible, because of the basic reaction conditions (NaH)
employed in one of the intermediate steps (preparation of 9).
Consequently, the use of a p-methoxybenzyl (PMB) ether-pro-
tecting group was explored (Scheme 2). This alternative PMB-
ether protection strategy was only possible after the new
bromination conditions, described for the synthesis of intermedi-
ate 8, were adopted. The previously employed acid-catalysed
bromination conditions20–24 prevent the use of PMB as a pro-
tecting group as these ethers are known to be labile to oxi-
dative (NBS) and acidic (TsOH or H2SO4) conditions.34
Adoption of our optimized bromination conditions allows the
reaction to proceed under mild basic (DMAP) conditions
rather than acidic conditions. Moreover, our conditions
shorten the bromination time to only one hour, leaving the
PMB-ether installed to provide the desired bromo PMB-ether
in good yield.
In summary, an improved synthesis of the potent anticancer
and parasite attenuating agent (tafuramycin A, 3) is described.
The first improvement described is the regioselective bromina-
tion, employing DMAP as a catalyst. This bromination step is a
common key step in the synthesis of several duocarmycin ana-
logues, and therefore, the modified regioselective bromination
conditions will be applicable in many other related synthetic
steps. The second improvement is the control of the aryl
radical alkene cyclization step to yield the intended furano-
indole intermediate 10 exclusively in an excellent yield.
The structure of the side product 13 produced in the cycli-
zation step upon use of excess Bu3SnH was also identified.
X-ray crystal structure determination (CCDC 992249) of the
subsequent coupling product 14 confirmed it as the dechloro-
product of 10.
These two major improvements (Scheme 1) provided the
final product 3 starting from intermediate 6 in an overall yield
of 52%. This represents a significant improvement over the
original procedure (11% overall yield). The final improvement
The new synthetic pathway started with the treatment of 5
with PMB-chloride in the presence of potassium carbonate
and TBAI to yield intermediate 17 (84%). Subsequent ester
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 4260–4264 | 4263

View Article Online
Paper
Organic & Biomolecular Chemistry
that employed an alternative protection/deprotection strategy 11 C. M. Gauss, A. Hamasaki, J. P. Parrish, K. S. MacMillan,
(Scheme 2) overcame the problem of overreduction in the final
step. The success of the new protection method is ascribed to
T. J. Rayl, I. Hwang and D. L. Boger, Tetrahedron, 2009, 65,
6591.
the newly adopted DMAP-catalysed bromination conditions 12 K. Kiakos, A. Sato, T. Asao, P. J. McHugh, M. Lee and
that overcome the limitations of the classic acid-catalysed J. A. Hartley, Mol. Cancer Ther., 2007, 6, 2708.
conditions preventing use of a PMB ether-protecting group. 13 L. A. Purcell, S. K. Yanow, M. Lee, T. W. Spithill and
This strategy gave exclusively the desired product 3 in an A. Rodriguez, Infect. Immun., 2008, 67, 1193.
overall yield of 37% from intermediate 17. Importantly, this 14 S. K. Yanow, L. A. Purcell, G. Pradel, A. Sato, A. Rodriguez,
protection/deprotection strategy allows a more versatile syn- M. Lee and T. W. Spithill, J. Infect. Dis., 2008, 197, 527.
thesis of duocarmycin seco-analogues that carry labile groups 15 T. T. Howard, B. M. Lingerfelt, B. L. Purnell, A. E. Scott,
under reducing reaction conditions. Such analogues were pre-
viously inaccessible by previously published methods.
C. A. Price, H. M. Townes, L. McNulty, H. L. Handl,
K. Summerville, S. J. Hudson, J. P. Bowen, K. Kiakos,
J. A. Hartley and M. Lee, Bioorg. Med. Chem., 2002, 10,
2941.
Notes
16 M. Lee, M. Good and T. Spithill, PCT. Int. Appl,
WO12162731, 2012.
The authors declare no competing financial interest.
17 M. F. Good, J. M. Reiman, I. B. Rodriguez, K. Ito,
S. K. Yanow, I. M. El-Deeb, M. R. Batzlof, D. I. Stanisic,
C. Engwerda, T. Spithill, S. L. Hoffman, M. Lee and
V. McPhun, J. Clin. Invest., 2013, 123, 3353.
18 H. Muratake, A. Hayakawa and M. Natsume, Tetrahedron
Lett., 1997, 38, 7577.
19 H. Muratake, A. Hayakawa and M. Natsume, Chem. Pharm.
Bull., 2000, 48, 1558.
20 V. F. Patel, S. L. Andis, J. K. Enkema, D. A. Johnson,
J. H. Kennedy, F. Mohamadi, R. M. Schultz, D. J. Soose and
M. M. Spees, J. Org. Chem., 1997, 62, 8868.
Acknowledgements
The authors would like to acknowledge Dr Renee Winzar and
Dr Robin Thomson for their research assistance and invalu-
able discussions respectively. We gratefully acknowledge the
award of a Griffith University Postdoctoral Award (IMED),
funding from the National Health and Medical Research
Council and the support of the Central Analytical Research
Facility, Queensland University of Technology.
21 D. L. Boger and M. S. S. Palanki, J. Am. Chem. Soc., 1992,
114, 9318.
22 D. L. Boger, J. A. McKie, H. Cai, B. Cacciari and
P. G. Baraldi, J. Org. Chem., 1996, 61, 1710.
Notes and references
1 M. Searcey, Curr. Pharm. Des., 2002, 8, 1375.
2 L. Brulikova, J. Hlavac and P. Hradil, Curr. Med. Chem.,
2012, 19, 364.
3 C. Suckling, Future Med. Chem., 2012, 4, 971.
4 D. L. Boger, D. S. Johnson and W. Yun, J. Am. Chem. Soc.,
1994, 116, 1635.
23 D. L. Boger, T. V. Hughes and M. P. Hedrick, J. Org. Chem.,
2001, 66, 2207.
24 C. M. Gauss, A. Hamasaki, J. P. Parrish, K. S. MacMillan,
T. J. Rayl, I. Hwang and D. L. Boger, Tetrahedron, 2009, 65,
6591.
25 G. Jia and J. W. Lown, Bioorg. Med. Chem., 2000, 8, 1607.
5 J. P. McGovern, G. L. Clarke, E. A. Pratte and T. F. DeKoning, 26 D. L. Boger and C. W. Boyce, J. Org. Chem., 2000, 65,
J. Antibiot., 1983, 37, 63. 4088.
6 M. Ichimura, T. Ogawa, K. Takahashi, A. Mihara, 27 J. P. Parrish, D. B. Kastrinsky, I. Hwang and D. L. Boger,
I. Takahashi and H. Nakano, Oncol. Res., 1993, 5, 165. J. Org. Chem., 2003, 68, 8984.
7 K. Gomi, E. Kobayashi, K. Miyoshi, T. Ashizawa, A. Okamoto, 28 H. Pati, T. Howard, H. Townes, B. Lingerfelt, L. McNulty
T. Ogawa, S. Katsumata, A. Mihara, M. Okabe and T. Hirata,
Jpn. J. Cancer Res., 1992, 83, 113.
8 J. P. Parrish, D. B. Kastrinsky, F. Stauffer, M. P. Hedrick,
and M. Lee, Molecules, 2004, 9, 125.
29 D. L. Boger and P. Mesini, J. Am. Chem. Soc., 1994, 116,
11335.
I. Hwang and D. L. Boger, Bioorg. Med. Chem., 2003, 11, 30 N. H. Al-Said, K. Q. Shawakfeh and W. N. Abdullah, Mole-
3815. cules, 2005, 10, 1446.
9 A. Sato, L. McNulty, K. Cox, S. Kim, A. Scott, K. Daniell, 31 N. H. Al-Said, K. Q. Shawakfeh and B. A. Hammad, J. Het-
K. Summerville, C. Price, S. Hudson, K. Kiakos, J. A. Hartley,
erocycl. Chem., 2008, 45, 1333.
T. Asao and M. Lee, J. Med. Chem., 2005, 48, 3903.
32 J. P. Lajiness and D. L. Boger, J. Org. Chem., 2011, 76, 583.
10 M. S. Tichenor, K. S. MacMillan, J. S. Stover, 33 F. Zhang, Y. Zhao, L. Sun, L. Ding, Y. Gu and P. Gong,
S. E. Wolkenberg, M. G. Pavani, L. Zanella, A. N. Zaid, Eur. J. Med. Chem., 2011, 46, 3149.
G. Spalluto, T. J. Rayl, I. Hwang, P. G. Baraldi and 34 B. Classon, P. J. Garegg and B. Samuelsson, Acta Chem.
D. L. Boger, J. Am. Chem. Soc., 2007, 129, 14092.
4264 | Org. Biomol. Chem., 2014, 12, 4260–4264
Scand. Ser. B, 1984, 38, 419.
This journal is © The Royal Society of Chemistry 2014