close

Вход

Забыли?

вход по аккаунту

?

Mutasynthesis of Aureonitrile An Aureothin Derivative with Significantly Improved Cytostatic Effect.

код для вставкиСкачать
Communications
Antitumor Agents
Mutasynthesis of Aureonitrile: An Aureothin
Derivative with Significantly Improved Cytostatic
Effect**
Martina Ziehl, Jing He, Hans-Martin Dahse, and
Christian Hertweck*
Polyketides constitute a highly diverse group of natural
products that play a leading role in drug discovery.[1] All
polyketides are assembled from an acyl starter unit by
repetitive Claisen condensations of extender units derived
from malonyl coenzyme A (CoA) in a manner that closely
parallels fatty acid biosynthesis.[2] While the vast majority of
polyketides are primed with acetate, a number of polyketide
synthases (PKSs) utilize alternative starter units that enlarge
polyketide diversity and provide the molecule with important
structural and biological features.[3] Consequently, the variation of starter units can be an efficient approach to yield
complex natural product analogues with potentially improved
activity, selectivity, availability, and fewer unwanted side
effects. A strategy to achieve this goal is the addition of nonnatural primers to a mutant, in which particular genes that
code for the biosynthesis of the starter unit have been deleted.
This technique, which suppresses the formation of the natural
metabolites, is also referred to as “mutasynthesis”.[4–7] In the
PKS area this approach has been successful generating
analogues of erythromycin,[3, 8, 9] rifamycin,[10] and enterocin.[11]
The most impressive example for mutational synthesis of a
polyketide derivative to date is the engineered biosynthesis of
the anthelmintic agent doramectin by exchanging the isobutyrate starter of avermectin for cyclohexanoate.[12–14]
A highly unusual polyketide starter unit, p-nitro benzoate
(PNBA), is employed in the biosynthesis of aureothin (1), a
polyketide-shikimate hybrid metabolite from the soil bacterium Streptomyces thioluteus.[15, 16] Aureothin exhibits a vari-
[*] M. Ziehl, J. He, Dr. H.-M. Dahse, Dr. C. Hertweck
Hans-Knoell-Institute for Natural Products Research
Beutenbergstrasse 11a
07745 Jena (Germany)
Fax: (+ 49) 3641-656-705
E-mail: christian.hertweck@hki-jena.de
[**] Financial support by the DFG priority program SPP1152: “Evolution
of metabolic diversity” (HE3469/2) is gratefully acknowledged. We
thank Dr. U. Mllmann and Dr. A. Hrtl for biological testing, and
K.-D. Menzel and M. Steinacker for assistance with fermentation.
We are also grateful to Prof. Xiufen Zhou of Shanghai Jiaotong
University, China, and Prof. Jose Salas and Dr. Carmen Mendez,
University of Oviedo, Spain, for the gifts of S. lividans ZX1 and
S. albus, respectively.
1202
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ety of pharmacological properties, which include weak
cytotoxic, antifungal, and antiviral activities.[17] As a basis
for engineering aureothin derivatives with altered bioactivity
profiles, we recently cloned and analyzed the entire aureothin
(aur) biosynthesis gene cluster.[18] In addition to employing
the rare PNBA starter unit, the multifunctional PKS enzyme
complex has some peculiar features. First, the aureothin PKS
does not have an obvious loading domain like that found in
conventional bacterial type I PKS,[3] and thus the priming
mechanism is obscure. Possibly the primer is loaded directly
onto the KS1 domain after activation by the putative
acyl CoA ligase AurE. Second, contrary to the principle of
colinearity, the first module (AurA) is used iteratively.[18]
Consequently, the KS of AurA has a double editing role, as
it needs to tolerate the alternative PKS starter moiety before
and after the first elongation cycle (Figure 1). Thus, using the
aureothin PKS in precursor-directed biosynthesis would be
intriguing.
Analyses of the aur genes suggested that the biosynthesis
of PNBA involves a p-aminobenzoate (PABA) synthase,
AurG, which converts chorismate into PABA, and a novel Noxygenase, AurF, for the N-oxygenation of PABA.[16] Thus, for
the mutasynthesis of aureothin derivatives, two genes, aurF
and aurG, were targeted for the construction of suitable null
mutants (Figure 1). The PABA synthase gene aurG was
successfully excised from the aur gene cluster on expression
plasmid pHJ48 using the l Red method.[19, 20] The resulting
plasmid, pHJ97 (Figure 2), was introduced into the heterologous expression hosts S. lividans ZX1, by polyethyleneglycol
(PEG) induced protoplast transformation, and S. albus, by
intergeneric conjugation.[21] The recombinant strains were
fermented and their metabolic profiles monitored by thin
layer chromatography (TLC) and HPLC/MS. Remarkably,
the transconjugant S. albus::pHJ97 was still capable of
aureothin biosynthesis, albeit in strongly reduced amounts.
This result is in accord with the finding that not only one, but
several PABA synthase genes may be present in a Streptomyces genome, including the essential PABA synthase genes
required for folic acid synthesis.[22, 23] In contrast to the results
obtained with S. albus, aureothin biosynthesis was completely
abolished in S. lividans ZX1::pHJ97, and only upon administering synthetic PABA to the mutant, was aureothin production restored. It thus appears that the titer of PABA is straindependent and that PABA involved in folate biosynthesis
does not interfere with the aureothin pathway. Another null
mutant lacking the N-oxygenase gene was generated analogously to the aurG null mutant (Figure 2). Transfer of the
manipulated gene cluster into S. lividans ZX1 resulted in a
strain (S. lividans ZX1::pHJ79) that is incapable of producing
aureothin in the absence of synthetic PNBA.[16] This result
underlines our earlier finding that PNBA, and not PABA,
serves as starter unit and suggests that the aureothin synthase
has a rather high primer specificity.
Expression of the mutated gene clusters in the S. lividans
host appeared suitable for mutasynthesis, since the aur
pathway could only be restored with exogenously supplied
PABA or PNBA. To explore the substrate specificity of the
aur PKS, first p-, m-, and o-PNBA were administered to a
growing culture of S. lividans ZX1::pHJ79. Not surprisingly,
DOI: 10.1002/anie.200461990
Angew. Chem. Int. Ed. 2005, 44, 1202 –1205
Angewandte
Chemie
gates were tested, either as free acids or as the
corresponding N-acetyl cysteamine (NAC) thioesters, which were synthesized using the dicyclohexyl
carbodiimide/4-dimethylaminopyridine (DCC/DMAP) method.[24, 25] The NAC
adducts serve as activated acyl CoA mimics[26]
that may diffuse into the bacterial cells and
bypass a potential bottleneck, the putative
acyl CoA ligase AurE. Of the various p-substituted benzoic acids, p-iodo, p-bromo, p-chloro,
and p-fluoro benzoate, as well as p-N-acetamido
anthranilic acid and p-dimethylamino benzoate
were probed on a 100-mL fermentation scale.
Unfortunately, novel aureothin derivatives
could not be detected in the crude extracts of
these feeding experiments. Also, toluic acid and
terephthalic acid monomethyl ester failed to
incorporate. Strikingly though, addition of pcyano benzoic acid to a culture of the aurF null
mutant yielded a novel metabolite, which was
detected by ESI-MS in the positive mode
(m/z 378). Since only relatively low quantities
of the new compound were produced, we tested
if p-cyano benzoyl-SNAC would provide higher
yields, as it does not need to be activated as an
acyl CoA adduct by the ligase (Figure 3). HowFigure 1. Model of aureothin biosynthesis starting from chorismate and the strategy
ever, the yield did not exceed that obtained with
for priming the type I PKS with alternative activated benzoates (R’: NAC, CoA). Key
the free acid, which reveals that in this case acid
enzymes (AurF, AurG) involved in the biosynthesis of the starter unit (PNBA) are taractivation by the ligase AurE is not a bottleneck.
gets for mutasynthesis and highlighted with arrows.
To obtain sufficient quantities of the new
compound for structure elucidation and biological testing, a medium-scale fermentation (50 L)
was performed. Fermentation broth and mycelium of the
mutant supplemented with cyano benzoate were exhaustively
extracted with ethyl acetate, and the crude extracts were
subjected to sequential column chromatography on silica and
sephadex LH-20. Final purification of the new compound was
by preparative HPLC.
HR-MS and 13C NMR spectroscopy corresponded to a
Figure 2. Organization of a) the entire aur gene cluster, as in pHJ48,
b) aurF null mutant (pHJ79), and c) aurG null mutant (pHJ97); aurmolecular formula of C23H23NO4. The presence of a nitrile
ABC PKS genes, aurD regulator gene, aurE acyl-CoA ligase gene,
moiety was confirmed by the 13C NMR spectroscopy signal at
aurF N-oxygenase gene, aurG PABA synthase gene, aurH P-450 monod = 142.14 ppm, as well as the very characteristic IR band
oxygenase gene, aurI O-methyl transferase gene.
(nCN) at 2223 cm 1. Except for the signals of the aromatic
protons, the 1H NMR spectroscopic data of 1 and the new
compound were very similar, and HH-COSY, HMQC, and
only the p-substituted acid was used as substrate, while the
HMBC data revealed the same connectivities. All spectroregioisomers were not incorporated. In a series of further
metric (MSn) and spectroscopic (IR, NMR) data support the
feeding experiments a variety of p-substituted PNBA surro-
Figure 3. Mutasynthesis of aureonitrile with PCBA and PCBA-SNAC.
Angew. Chem. Int. Ed. 2005, 44, 1202 –1205
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1203
Communications
expected structure of an aureothin derivative substituted with
a nitrile moiety instead of the nitro group. The new compound
was thus named aureonitrile (2). In addition, NOE measurements revealed the same configuration for 2 as for aureothin.
It is surprising that out of the number of p-substituted
benzoates only p-cyano benzoate was tolerated by the PKS,
but not related compounds with substituents that induce
comparable I- and M effects, such as methyl carboxylate
and N-acetamido. This observation may be rationalized by a
similar charge distribution of nitrile and nitro groups.
Aureonitrile (2) was subjected to broad biological testing,
with focus on antimicrobial and cytotoxic assays. While 2
shows only weak antibacterial activities, it has a reasonably
high potency against fungal pathogens, for example Candida
albicans (IC50 of 12.5 mm). The most remarkable results,
however, were obtained in cytotoxicity and antiproliferative
assays.
Most importantly, we observed a flat dose-response curve
of the pyrones, which indicates a cytostatic effect (Figure 4).
Figure 4. Cytostatic effect of aureothin (1; ^) and aureonitrile (2; &)
on adherent human cervix carcinoma HeLa cells. For control (&) and
for dilutions of 1 and 2, the same number of cells was seeded per
microplate well. After incubation (72 h) the relative amount of treated
cells was monitored versus the control.
Compared to 1, compound 2 shows a significantly enhanced
activity (> 200 %) against various tumor cell lines, such as
HeLa and K-562, at a wide range of concentration. In contrast
to classical anticancer drugs, which often exert unwanted
cytotoxic effects in addition to cytostatic properties, the new
generation of antitumor agents should show only cytostatic
effects in a broad therapeutic window. Such drugs stop cell
proliferation at the checkpoints of mitosis or interfere with
DNA synthesis.[27] Owing to the intriguing properties of 2,
in vivo assays and the elucidation of its molecular target are
now underway.
In conclusion, we have successfully explored the potential
of the iterative aureothin PKS for precursor-directed biosynthesis. For this purpose, suitable mutants were engineered,
which also supported the biosynthetic model of PNBA
generation via PABA involving a specific PABA synthase.
For complementation of the mutants, a number of PNBA
surrogates were tested as free acids, as well as the corresponding synthetic N-acetyl cysteamine thioesters. The iterative aur polyketide synthase, which lacks a loading domain,
exhibits a rather high substrate specificity, besides PNBA it
1204
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
only allows p-cyano benzoate to be further processed.
However, the resulting fully characterized nitrile analogue
of aureothin, shows significantly enhanced cytostatic effects
against HeLa and K-562 tumor cells compared to the native
compound in the range 50 mg L 1 to 10 ng mL 1. This result
underlines the power of mutasynthesis towards polyketide
derivatives with improved properties. Strain optimization and
complementary synthetic approaches are currently underway
to obtain larger quantities of aureonitrile.
Experimental Section
Construction of aurG null mutant: Deletion of aurG from the aur
gene cluster was performed in analogy to the described method for
inactivation of aurF,[16] using the l Red method with primers aurGF
(5’-AGG ACG ACC CCG CTC GAA CCG ACA GGG AGG CTC
CCG ATG ATT CCG GGG ATC CGT CGA CC-3’) and aurGR (5’CTT GAA GCC CGT CGA ACG ACG GGC GGA ACG GCC
GTA CTA TGT AGG CTG GAG CTG CTT C-3’). The polymerase
chain reaction (PCR) product was introduced into E. coli BW25113/
pIJ790 containing cosmid pHJ48, which includes the entire aureothin
biosynthesis gene cluster, with concomitant substitution of the aurG
gene by the extended spectinomycin resistance cassette. The inserted
cassette was removed through expression of the FLP-recombinase in
E. coli, to yield a 81 base pair (bp) “scar” in the preferred reading
frame. The resulting plasmid, pHJ97, was then introduced into
S. albus by intergeneric conjugation using E. coli ET12567 containing
the RP4 derivative pUZ8002 or into S. lividans ZX1 by PEGmediated protoplast transformation.
Synthesis of p-nitrobenzoyl-SNAC: Dimethylaminopyridine
(0.82 mmol, 98 mg) N-acetyl cysteamine (6.1 mmol, 0.7 mL), ,and
dicyclohexylcarbodiimide (4.43 mmol, 4.4 mL, 1m solution in
dichloromethane) were sequentially added to a solution of (4 mmol,
593 mg) p-cyano benzoic acid in dry dichloromethane (24 mL) and
dry THF (8 mL) at 0 8C. The mixture was stirred for 10 min at 0 8C and
3 days at room temperature. After filtration through celite the
solution was passed through a short column of copper sulfate
impregnated silica. PNBA-SNAC was purified by precipitation with
petroleum ether. Yield: 594.4 mg (59 %). 1H NMR (300 MHz,
CDCl3): d = 1.95 (s, 3 H, CH3); 3.24 (t, J = 6,5 Hz, 2 H, S-CH2), 3.51
(q, J = 6.3 Hz, 2 H, CH2-N), 5.94 (br s, 1 H, NH-CO), 7.73 (d, J =
8.47 Hz, 2 H), 8.01 ppm (d, J = 8.39 Hz, 2 H); 13C NMR (75 MHz,
CDCl3): d = 23.14 (CH3), 29.03 (S-CH2), 39.25 (CH2-N), 116.87 (CCN), 117.65 (CN), 127.69 (2C, Ar), 132.52 (2C, Ar), 139.75 (C-CO),
170.32 (N-CO), 190.75 ppm (CO-S); IR (film): ñ = 2231 (nCN), 1642,
1553 (nOCNH, nC=C), 919 cm 1 (gCH, Ar); HR-MS calcd for
C12H13SN2O2 : 249.0692; observed: 249.0696.
Fermentation and isolation of aureonitrile: All fermentations were
performed in liquid M10 medium supplemented with apramycin
(50 mg mL 1) as described elsewhere.[18] Precursors were added as
sterile solutions in H2O/DMSO (v/v = 75:25, c = 1 mg/100 mL). A
S. lividans::pHJ79 seed culture was used to inoculate 50 L M10 (apra)
medium in a stainless steel fermenter. The culture was pulse-fed with
p-cyano benzoate (1.325 g) in two portions. After 5 days with constant
stirring at 28 8C mycelium and filtrate were separated and exhaustively extracted with ethyl acetate. The crude extracts were separated
on silica by using a chloroform/methanol gradient as eluent and on
sephadex by using methanol for elution. All fractions containing the
main product were combined and further purified by RP-HPLC on a
Phenomenex Luna C18 column (10 m, 250 21.2 mm) using a 80:20–
100:0 acetonitrile/water gradient for elution over 10 min at a flow rate
of 20 mL min 1. Yield: 5.9 mg of a dark yellow solid. 1H NMR
(300 MHz, CDCl3): d = 1.84 (s, 3 H, 17a-CH3), 2.0 (s, 3 H, 9a-CH3),
2.01 (s, 3 H, 15a-CH3), 2.98 (m, 2 H, 13-CH2), 3.92 (s, 3 H, O-CH3), 4.71
(d, J = 14.2 Hz, 1 H, 11a-CH2), 4.84 (d, J = 14.2 Hz, 1 H, 11a-CH2),
5.12 (t, J = 6.9 Hz, 1 H, 12-CH), 6.16 (s, 1 H, 10-CH), 6.30 (s, 1 H, 8-
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 1202 –1205
Angewandte
Chemie
CH), 7.32 (d, J = 8.2 Hz, 2 H, Ar H3,H5), 7.61 ppm (d, J = 8.3 Hz, 2 H,
Ar H2, H6); 13C NMR (75 MHz, CDCl3): d = 6.89 (17a-CH3), 9.42
(15a-CH3), 17.65 (9a-CH3), 38.27 (13-CH2), 55.26 (O-CH3), 70.10
(11a-CH2), 73.33 (12-CH), 100.11 (17-C), 110.10 (1-C), 118.89 (4-C),
120.18 (15-C), 126.00 (10-CH), 128.72 (8-CH), 129.53 (Ar C3,C5),
132.00 (Ar C2, C6), 137.89 (9-C), 140.22 (11-C), 142.15 (CN), 154.72
(14-C), 162.09 (18-C), 180.60 ppm (16-C); IR (film): ñ = 2223 (nCN),
1718 (nC=O), 1662 (nC=O), 1591, 1499, w (nC=C), 915 cm 1 (nC-O-C); HRMS: calcd for C23H24NO4 : 378.1700; observed: 378.1705.
Antiproliferative and cytotoxic assay: Cells of K-562 (DSM
ACC 10), L-929 (DSM ACC 2), and HeLa (DSM ACC 57) were
cultured in RPMI 1640 medium (GIBCO BRL), supplemented with
gentamicin sulfate (25 mg mL 1; Cambrex), 10 % heat inactivated fetal
bovine serum (GIBCO BRL), and GlutaMAX-1 (GIBCO BRL) at
37 8C in high density polyethylene flasks (NUNC). Aureothin (1) and
aureonitrile (2) were dissolved in ethanol (10 mg mL 1) before being
diluted in RPMI 1640. The adherent cells of L-929 and HeLa were
harvested at the logarithmic growth phase after soft trypsinization,
using 0.25 % trypsin in PBS (phosphate buffer saline) containing
0.02 % EDTA (Biochrom KG). For cytotoxicity assays, approximately 10 000 HeLa cells per well of the 96-well microplates cells were
seeded with RPMI 1640 (0.2 mL) and pre-incubated for 48 h without 1
and 2. After formation of a subconfluent monolayer, cells were
treated with dilutions of 1 and 2 and incubated under physiological
conditions (72 h at 37 8C in a humidified atmosphere and 5 % CO2).
For the antiproliferative assay, 1 and 2 were tested against L-929 and
K-562. For each experiment approximately 10 000 cells were seeded
with RPMI 1640 (0.1 mL) per well of the 96-well microplates and
subsequently exposed to dilutions of 1 and 2 and incubated under
physiological conditions for 72 h. The measurement of cytotoxicity
and cell growth inhibition was based on registration of living cells
after incubation with 1 and 2 versus untreated control. Non-adherent
K-562 cells were analyzed by an electronic cell analyzer system
CASY 1 (Schrfe) as described elsewhere.[28] Adherent L-929 and
HeLa cells were fixed by glutaraldehyde (Merck) and stained with
methylene blue (SERVA) for 15 min. After gentle washing the stain
was eluted with HCl (0.2 mL, 0.33 n) in the wells. The optical densities
were measured at 660 nm in a SUNRISE microplate reader
(TECAN).
[13] H. A. I. McArthur, A Novel Avermectin, Doramectin—A Successful Application of Mutasynthesis, Society for Industrial
Microbiology, Fairfax, VA, 1998.
[14] T. A. Cropp, D. J. Wilson, K. A. Reynolds, Nat. Biotechnol. 2000,
18, 980.
[15] Y. Hirata, H. Nakata, K. Yamada, K. Okuhara, T. Naito,
Tetrahedron 1961, 252.
[16] J. He, C. Hertweck, J. Am. Chem. Soc. 2004, 126, 3694.
[17] J. L. Schwartz, M. Tishler, B. H. Arison, H. M. Shafer, S. Omura,
J. Antibiot. 1976, 29, 236.
[18] J. He, C. Hertweck, Chem. Biol. 2003, 10, 1225.
[19] K. A. Datsenko, B. L. Wanner, Proc. Natl. Acad. Sci. USA 2000,
97, 6640.
[20] K. C. Murphy, K. G. Campellone, A. R. Poteete, Gene 2000, 246,
321.
[21] T. Kieser, M. J. Bibb, M. J. Buttner, K. F. Chater, D. A. Hopwood, Practical Streptomyces Genetics, The John Innes Foundation, Norwich, UK, 2000.
[22] M. P. Brown, K. A. Aidoo, L. C. Vining, Microbiology 1996, 142,
1345.
[23] J. M. Green, B. P. Nichols, J. Biol. Chem. 1991, 266, 12 972.
[24] B. Neises, W. Steglich, Angew. Chem. 1978, 90, 556.
[25] C. J. Wilkinson, E. J. Frost, J. Staunton, P. F. Leadlay, Chem. Biol.
2001, 8, 1197.
[26] J. Staunton, A. C. Sutkowski, J. Chem. Soc. Chem. Commun.
1991, 1110.
[27] S. N. Gardner, M. Fernandes, J. Exp. Ther. Oncol. 2004, 4, 9.
[28] H.-M. Dahse, B. Schlegel, U. Grfe, Pharmazie 2001, 56, 489.
Received: September 14, 2004
Published online: January 11, 2005
.
Keywords: antitumor agents · mutasynthesis · natural products ·
polyketides
[1] D. OHagan, The Polyketide Metabolites, Ellis Horwood, Chichester, 1991.
[2] J. Staunton, K. J. Weissman, Nat. Prod. Rep. 2001, 18, 380.
[3] B. S. Moore, C. Hertweck, Nat. Prod. Rep. 2002, 19, 70.
[4] K. L. Rinehart, L. S. Shield, Fortschr. Chem. Org. Naturst. 1976,
33, 231.
[5] K. L. Rinehart, Pure Appl. Chem. 1977, 49, 1361.
[6] S. J. Daum, J. R. Lemke, Annu. Rev. Microbiol. 1979, 33, 241.
[7] W. Wohlleben, S. Pelzer, Chem. Biol. 2003, 10, 1163.
[8] K. Konishita, P. G. Williard, C. Khosla, D. E. Cane, J. Am. Chem.
Soc. 2001, 123, 2495.
[9] R. P. Desai, T. Leaf, Z. Hu, C. R. Hutchinson, A. Hong, G. Byng,
Biotechnol. Prog. 2004, 20, 38.
[10] D. Hunziker, T.-W. Yu, C. R. Hutchinson, H. G. Floss, C. Khosla,
J. Am. Chem. Soc. 1998, 120, 1092.
[11] J. A. Kalaitzis, M. Izumikawa, L. Xiang, C. Hertweck, B. S.
Moore, J. Am. Chem. Soc. 2003, 125, 9290.
[12] C. J. Dutton, S. P. Gibson, A. C. Goudie, K. S. Holdom, M. S.
Pacey, J. C. Ruddock, J. D. BuLock, M. K. Richards, J. Antibiot.
1991, 44, 357.
Angew. Chem. Int. Ed. 2005, 44, 1202 –1205
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1205
Документ
Категория
Без категории
Просмотров
1
Размер файла
121 Кб
Теги
significantly, aureothin, effect, aureonitrile, mutasynthesis, improve, cytostatic, derivatives
1/--страниц
Пожаловаться на содержимое документа