Nonactin biosynthesis: Setting limits on what can be achieved with precursor-directed biosynthesis

Article abstract of DOI:10.1016/j.bmcl.2008.12.096

Nonactin, produced by Streptomyces griseus ETH A7796, is a macrotetrolide assembled from nonactic acid. It is an effective inhibitor of drug efflux in multidrug resistant erythroleukemia K562 cells at sub-toxic concentrations and has been shown to possess both antibacterial and antitumor activity. As total synthesis is impractical for the generation of nonactin analogs we have studied precursor-directed biosynthesis as an alternative as it is known that nonactic acid can serve as a nonactin precursor in vivo. To determine the scope of the approach we prepared and evaluated a furan-based nonactic acid derivative, 11. Although no new nonactin analogs were detected when 11 was administered to S. griseus fermentative cultures, a significant inhibition of nonactin biosynthesis was noted (IC₅₀ ～ 100 μM). Cell mass, nonactic acid production and the generation of other secondary metabolites in the culture were unaffected by 11 demonstrating that 11 selectively inhibited the assembly of nonactin from nonactic acid. While we were unable to generate new nonactin analogs we have discovered, however, a useful inhibitor that we can use to probe the mechanism of nonactin assembly with the ultimate goal of developing more successful precursor-directed biosynthesis transformations.

Full text of DOI:10.1016/j.bmcl.2008.12.096

Bioorganic & Medicinal Chemistry Letters 19 (2009) 1233–1235
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Nonactin biosynthesis: Setting limits on what can be achieved
with precursor-directed biosynthesis
*
Brian R. Kusche, Joshua B. Phillips, Nigel D. Priestley
Department of Chemistry, The University of Montana, 32 Campus Drive, Missoula, MT 59812, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Nonactin, produced by Streptomyces griseus ETH A7796, is a macrotetrolide assembled from nonactic acid.
It is an effective inhibitor of drug efflux in multidrug resistant erythroleukemia K562 cells at sub-toxic
concentrations and has been shown to possess both antibacterial and antitumor activity. As total synthe-
sis is impractical for the generation of nonactin analogs we have studied precursor-directed biosynthesis
as an alternative as it is known that nonactic acid can serve as a nonactin precursor in vivo. To determine
the scope of the approach we prepared and evaluated a furan-based nonactic acid derivative, 11.
Although no new nonactin analogs were detected when 11 was administered to S. griseus fermentative
Received 30 October 2008
Revised 16 December 2008
Accepted 17 December 2008
Available online 30 December 2008
Keywords:
Nonactin
Precursor-directed
Biosynthesis
cultures, a significant inhibition of nonactin biosynthesis was noted (IC50 ꢀ 100
lM). Cell mass, nonactic
acid production and the generation of other secondary metabolites in the culture were unaffected by 11
demonstrating that 11 selectively inhibited the assembly of nonactin from nonactic acid. While we were
unable to generate new nonactin analogs we have discovered, however, a useful inhibitor that we can use
to probe the mechanism of nonactin assembly with the ultimate goal of developing more successful pre-
cursor-directed biosynthesis transformations.
Ó 2008 Elsevier Ltd. All rights reserved.
Streptomyces griseus subsp. griseus ETH A7796 (DSM40695)
makes a series of ionophore antibiotics known as the macrotetro-
lides (Fig. 1).1 Nonactin, the prototypical macrotetrolide (1) is
assembled from two monomers of (À)-nonactic acid and two
monomers of (+)-nonactic acid assembled (+)-(À)-(+)-(À) in a
head-to-tail manner into a 32-membered macrocycle. Nonactin
has both antibiotic and anticancer properties2 and has been shown
to be an inhibitor of drug efflux in multiple drug resistant cancers.3
The natural macrotetrolide homologues produced by S. griseus
show a wide range potency with the minimum inhibitory concen-
tration of 1 being an order of magnitude greater than that of din-
actin (3) against Staphylococcus aureus and Mycobacterium bovis,
a difference that is related to the stability constants of their respec-
tive Na+ and K+ complexes.1,2,4
Nonactinis fartoohydrophobicandinsufficientlysolubletobe an
effective therapeutic.5 The development of therapeutics based upon
nonactin, therefore, will be dependent upon being able to make non-
natural analogs in a direct and efficient manner. The total synthesis
of nonactin has been achieved by a number of groups6–9 as has the
synthesis of nonactic acid.10,11 The total synthesis of nonactin ana-
logs is complicated as both enantiomers of nonacticacid, and its ana-
logs, are required as is the sequential construction of a linear
tetraester prior to a final macrolactonization reaction affording the
R1
O
O
O
O
O
O
O
O
R2
O
O
O
R4
O
R3
R1 R2 R3 R4
Nonactin (1)
Monactin (2)
Dinactin (3)
Trinactin (4)
Tetranactin (5)
Me Me Me Me
Me Me Et Me
Et Me Et Me
Et
Et Et Me
E
t
Et Et Et
OH
O
2
R
OH
8
O
6
R = Me
(+)-nonactic acid
* Corresponding author. Tel.: +1 406 243 6251.
Figure 1. The structures of the naturally occurring macrotetrolides and the
E-mail address: nigel.priestley@umontana.edu (N.D. Priestley).
monomer, nonactic acid.
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.12.096

1234
B. R. Kusche et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1233–1235
nonactin analog. This complexity does not lend itself to straightfor-
ward drug development.
Mycelia were recovered from each culture and an approximate
wet weight determined; addition of 11 had no effect upon biomass
production. Macrotetrolides present in the extracts were quanti-
fied by reverse-phase HPLC.20 It was determined that nonactin pro-
duction was inhibited by more than 90% in fermentative cultures
containing either 10 mM or 1 mM of 11. Nonactin production
was at the same level as in the control in a culture containing
0.01 mM of 11; a concentration of 0.1 mM of 11 lead to an interme-
diate level of nonactin production. In all cases, no new macrotetro-
lide analogs were evident in either MS or LC–MS analyses of the
extracts. In addition to showing that 11 had little effect on biomass
production, each of the fermentative cultures reliably generated
other secondary metabolites (phenazines) usually co-synthesized
with macrotetrolides when secondary metabolism is initiated in
S. griseus. Furthermore, analysis22 of the extracts demonstrated
that both enantiomers of nonactic acid, the monomer precursor
to nonactin, were generated in each culture irrespective of the con-
centration of 11 added. We have demonstrated, therefore, that
while 11 does not serve as a precursor to new macrotetrolides, it
Fermentative cultures of S. griseus can generate 6–15 g/L of
macrotetrolide mixtures (>90% 1 and 2). When nonactic acid is
added to fermentative cultures of either S. griseus, or genetically al-
tered strains of S. griseus that have been blocked in the early stages
of nonactin biosynthesis, it can be readily and efficiently incorpo-
rated into nonactin.12 These observations strongly suggest that
precursor-directed biosynthesis has the potential to generate
new nonactin analogs.13,14 As we know that a complex structure
such as nonactic acid will serve as a substrate for precursor-direc-
ted biosynthesis, our first task was to set bounds on the system by
discovering the simplest, most straightforward nonactic acid ana-
log that would work. To that end we completed the synthesis of
the substituted furan derivative 11 (Scheme 1) and evaluated it
in precursor-directed biosynthesis experiments.
Alkylation of a furan-derived anion with propylene oxide was
achieved using White’s method affording 8 in 66% yield after distil-
lation.10 The secondary alcohol of 6 was protected as an acetate by
reaction with acetic anhydride in pyridine to give 9 (79%).15 The
furan derivative 10 was obtained by a free radical addition reaction
with ethyl 2-iodopropionate, as described by Baciocchi and Mura-
glia, to give 10 in 11% yield.16,17 Saponification of 10 gave the free
acid 11 (81%).18 Although the synthesis generated a mixture of dia-
stereoisomers, the synthesis was quite short and effective and we
hoped that new analogs would more likely be formed using such a
mixture in a precursor-directed feeding experiment.
To assess the incorporation of compound 11 into new macrote-
trolide analogs, two fermentative cultures of S. griseus ETH A7796
were prepared from a single vegetative culture and grown for
48 h under standard conditions.19 At 48 h after inoculation of the
fermentative culture (50 mL), 56 mg of 11 (56 mg in 0.5 mL of eth-
anol) was added to one culture; a blank sample (0.5 mL ethanol)
was added to the equivalent control culture. The cultures were al-
lowed to grow for an additional 96 h and the macrotetrolide mix-
ture was isolated according to standard protocols.19 Analysis of the
macrotetrolide mixtures by HPLC20 and LC–MS (TOF)21 showed an
unexpected drastic reduction in macrotetrolide production in the
fed culture compared to the control. Unfortunately, no likely non-
actin analogs could be detected in the extract by both LC–MS and
MS analysis of the complex mixture.
We decided to further evaluate the inhibitory effects of 11 on
the production of nonactin to determine if adding less material
might allow for the production of higher levels of nonactin in
which a low incorporation of 11 might then be more evident. To
this end a series of fermentative cultures were evaluated, with a
range of concentrations of 11 ranging from 0.01 mM to 10 mM, to-
gether with a control culture receiving no exogenous compound.
is indeed a reasonably potent (IC50 ꢀ 100
lM) and selective inhib-
itor of nonactin biosynthesis. As nonactic acid production is not
perturbed, the furan derivative 11 is interfering with the assembly
of nonactin from its monomeric precursors, a process that requires
only the products of the nonK and nonJ type II polyketide synthase-
encoding genes and the nonL CoASH-dependent ligase-encoding
gene.23,24 The inhibition shown by 11 is distinguished from that
of our earlier acetylenic analogs which likely block the formation
of the monomer, nonactic acid.25
While we were not successful in the developing new nonactin
analogs through precursor-directed biosynthesis we have set limi-
tations on the nonactic acid analogs that may be used; it is likely
that the tetrahydrofuran ring of nonactic acid will be essential.
We have discovered, however, an inhibitor that likely will be of
use in our quest to determine how the macrocycle of nonactin is
constructed in vivo from monomeric precursors which, in turn, will
allow us to develop improved approaches for the discovery of non-
actin analogs by precursor-directed biosynthesis.
Acknowledgment
Financial support of this work from the NIH (CA77347) is grate-
fully acknowledged.
References and notes
1. Keller-Schierlein, W.; Gerlach, H. Fortschritted. Chem. Org. Naturstoffe 1968, 26,
161.
2. Meyers, E.; Pansy, F. E.; Perlman, D.; Smith, D. A.; Weisenborn, F. L. J. Antibiot.
1965, 18, 128.
3. Borrel, M. N.; Pereira, E.; Fiallo, M.; Garnier-Suillerot, A. Eur. J. Biochem. 1994,
223, 125.
4. Izatt, R. M.; Bradshaw, S. A.; Nielsen, S. A.; Lamb, J. D.; Christensen, J. J.; Sen, D.
Chem. Rev. 1985, 85, 271.
a
OH
OAc
b
O
7
O
O
5. Lipinski, C. A. J. Pharm. Toxicol. Meth. 2000, 44, 239.
8
9
6. Lee, J. Y. K.; Byeang, H. Tetrahedron Lett. 1996, 52, 571.
7. Lee, J. Y. K.; Byeang, H. Tetrahedron Lett. 1995, 36, 3364.
8. Fleming, I. G.; Sunil, K. J. Chem. Soc., Perkin Trans. 1 1998, 17, 2733.
9. Fleming, I. G.; Sunil, K. J. Chem. Soc., Chem. Commun. 1994, 19, 2287.
10. Arco, M. J.; Trammell, M. H.; White, J. D. J. Organic Chem. 1976, 43, 479.
11. Bartlett, P. A. M.; James, D.; Ottow, Eckhard J. Am. Chem. Soc. 1984, 106, 5304.
12. Woo, A. J.; Strohl, W. R.; Priestley, N. D. Antimicrob. Agents Chemother. 1997, 43,
1662.
13. Ritacco, F.; Graziani, E.; Summers, M.; Zabriskie, T.; Yu, K.; Bernan, V.; Carter,
G.; Greenstein, M. Appl. Environ. Microbiol. 2005, 71, 1971.
14. Jacobsen, J.; Hutchinson, C. R.; Cane, D. E.; Khosla, C. Science 1997, 277, 367.
15. 1-(Furan-2-yl)propan-2-yl acetate (9). Acetic anhydride (12.8 mL, 135 mmol) and
thenpyridine (18.4 mL, 225 mmol)were added toasolutionof8 (5.7 g, 45 mmol)
in THF (15 mL). The resulting solution was stirred at room temperature for 2 h
after which the reaction was quenched by the addition of water (100 mL). The
resulting mixture was extracted with ethyl acetate (3Â 100 mL) and the organic
phases were recovered, combined and washed with water (50 mL), aqueous
c
OAc
OAc
d
CO2Et
CO2Et
O
11
O
10
Scheme 1. Synthesis
a furan-based nonactic acid analog 11. Reagents and
conditions: (a) n-BuLi, À78 °C, THF then propylene oxide, 66%; (b) Ac2O, pyridine,
THF, 79%; (c) BEt3, Fe2(SO4)3, ethyl DL-2-iodopropionate, DMSO, 11%; (d) 2.5 M LiOH,
MeOH, THF, 81%.

B. R. Kusche et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1233–1235
1235
saturated copper sulfate (2Â 100 mL) and finally ammonium chloride solution
(2Â 100 mL). The extracts were dried (Na2SO4), filtered and concentrated to a
residue which was purified by distillation (96 °C, 24 mmHg) to give (9) (6.0 g,
79%). Rf = 0.50 (10% ethyl acetate/hexanes); 1H NMR (400 MHz, CDCl3) d 7.24 (s,
1H), 6.21 (s, 1H), 5.99 (s, 1H), 5.08 (sextet, J = 12.7, 6.4 Hz, 1H), 2.85 (dd, J = 15.1,
6.4 Hz, 1H), 2.75 (dd, J = 15.1, 6.4 Hz, 1H) 1.93 (s, 3H) and 1.17 (s, 3H); 13C NMR
(100 MHz, CDCl3) d 170.1, 151.4, 141.2, 110.0, 106.7, 69.0, 34.2, 20.9 and 19.3; IR
concentrated. The crude product was purified by chromatography on silica gel
(97:2.5:0.5 EtOAc/MeOH/AcOH) to give 11 (1.32 g, 81.0%) as a clear yellow oil.
Rf = 0.54 (99:1 EtOAc/TFA); 1H NMR (400 MHz, CDCl3) d 7.40 (s, 2H), 6.05 (d,
1H, J = 3.3), 5.98 (d, 1H, J = 3.3), 4.02 (sextet, J = 12.4, 6.2, 2H), 3.73 (q, 1H,
J = 7.3), 2.68 (d, J = 6.2, 2H), 1.46 (d, J = 7.3. 3H) and 1.15 (d, J = 6.2, 3H); 13C
NMR (100 MHz, CDCl3) d 177.3, 151.9, 151.6, 107.6, 106.8, 66.8, 39.1, 37.4, 22.2
and 15.3; IR (ATR) 3625–2375, 2973, 2922 and 1712 cmÀ1; ESI–TOF–HRMS m/z
calcd for C10H14O4H+: 199.0970; found 199.0971.
(ATR) 2982, 2937, 1736, 1372, and 1237 cmÀ1
.
16. Baciocchi, E.; Muraglia, E. Tetrahedron Lett. 1993, 34, 5015.
19. Ashworth, D. M.; Clark, C. A.; Robinson, J. A. J. Chem. Soc., Perkin Trans. 1 1989,
1461.
20. HPLC resolution and quantification of macrotetrolide mixtures was
17. Ethyl 2-(5-(2-acetoxypropyl)furan-2-yl)propanoate (10). Triethylborane (1.0 M,
182 mL, 182 mmol) was added to a suspension of 9 (30.6 g, 182 mmol), ethyl
DL-2-iodopropionate (41.5 g, 182 mmol) and ferric sulfate hydrate (72.8 g,
182 mmol) in DMSO (250 mL). After 45 min of vigorous stirring at room
temperature while open to the atmosphere, a second addition of triethylborane
(1.0 M, 182 mL, 182 mmol) was added to the suspension and the suspension
was allowed to stir vigorously for an additional 45 min. Brine (200 mL) was
then added and the resultant solution was extracted with Et2O (3Â 300 mL).
The combined organic phases were dried (Na2SO4), filtered and concentrated
under vacuum to give an oil. The crude oil was purified by flash
chromatography (1:4 EtOAc/hexanes) to give 10 (5.34 g, 11%) as a yellow-
orange oil. 1H NMR (400 MHz, CDCl3) d 6.01 (d, J = 3.3 Hz, 1H). 5.95 (d, J = 2.9
Hz, 1H), 5.06 (sext, J = 12.8, 6.6, 6.2 Hz, 1H), 4.10 (q, J = 14.3, 7.3, 7.0, 2H), 3.69
(q, J = 7.3 Hz, 1H), 2.73 (dd, J = 15, 6.2 Hz, 1H), 2.73 (dd, J = 15, 6.2 Hz, 1H), 1.96
(s, 3H), 1.43 (d, J = 7.3 Hz, 3H), 1.21–1.17 (m, 6H); 13C NMR (100 MHz, CDCl3) d
172.5, 170.3, 152.1, 150.7, 107.6, 106.4, 69.2, 60.8, 39.4, 34.3, 21.1, 19.4, 15.6
and 14.0; IR (ATR) 2983, 2940, and 1733 cmÀ1; ESI–TOF–HRMS m/z calcd for
C14H20O5H+: 269.1398; found 269.1389.
accomplished using
column with an isocratic mobile phase of acetonitrile–water (86:14)
containing 0.1% trifluoroacetic acid at flow rate of 1 mL/min.
a
Varian microsorb-mv 100-5 250 Â 4.6 mm C-18
a
Macrotetrolide elution was monitored by measuring the absorbance of the
eluate at 215 nm. Under these conditions the macrotetrolide retention times
were: nonactin, 7.50 min; monactin, 8.75 min; dinactin, 10.50 min; trinactin,
13.00 min and tetranactin, 16.33 min.
21. Cox, J. E.; Priestley, N. D. J. Am. Chem. Soc. 2005, 127, 7976.
22. The culture broth was acidified to pH 1–2 with 6 M HCl and then extracted
with EtOAc. The crude extracts were analyzed by TLC (95:4:1, EtOAc/MeOH/
AcOH) and stained with anisaldehyde. The crude extracts containing nonactic
and homononactic acids were also esterified by dissolving in 5% H2SO4 in
methanol and then heating at 65 °C for 1 h. The methyl nonactate and methyl
homononactate that were produced were isolated by chromatography on silica
gel (50:50 EtOAc/hexanes). Additional characterization of methyl nonactate
and methyl homononactate was achieved by gas chromatography.
23. Walczak, R. J.; Woo, A. J.; Strohl, W. R.; Priestley, N. D. FEMS Lett. 2000, 183, 171.
24. Kwon, H.-J.; Smith, W. C.; Scharon, J.; Hwang, S. H.; Kurth, M. J.; Shen, B. Science
2002, 297, 1327.
18. 2-(5-(2-Hydroxypropyl)furan-2-yl)propanoic acid (11). A solution of 10 (2.20 g,
8.2 mmol) in 40 mL of 2.5 M LiOH/MeOH/THF (1:2:3) was allowed to stir at
room temperature for 1.5 h. The reaction was acidified to pH 2 by the addition
of HCl (1.0 M), extracted with EtOAc (3Â 100 mL), dried (Na2SO4), filtered and
25. Priestley, N. D.; Earle, M. Bioorg. Med. Chem. Lett. 1997, 7, 2187.