Microbial transformation of trichostatin A to 2,3-dihydrotrichostatin A

Article abstract of DOI:10.1021/np1006718

A new reduced hydroxamate, 2,3-dihydrotrichostatin A, was created from trichostatin A by employing a recombinant strain of Streptomyces venezuelae as a microbial catalyst. Compared with trichostatin A, 2,3-dihydrotrichostatin A showed similar antifungal activity against Saccharomyces cerevisiae, but, interestingly, approximately twice the cytostatic activity against human small-cell lung cancer cells. The production of 2,3-dihydrotrichostatin A via microbial biotransformation demonstrates that the regiospecific and substrate-flexible hydrogenation by S. venezuelae provides a new approach for creating natural product analogues with improved bioactive properties. (Chemical Equation Presented).

Full text of DOI:10.1021/np1006718

NOTE
pubs.acs.org/jnp
Microbial Transformation of Trichostatin A to 2,3-Dihydrotrichostatin A
Je Won Park,† Sung Ryeol Park,‡ Ah Reum Han,§ Yeon-Hee Ban,‡ Young Ji Yoo,‡ Eun Ji Kim,‡ Eunji Kim,‡
and Yeo Joon Yoon*,‡
†Department of Pharmaceutical Engineering, Institute of Biomolecule Reconstruction, Sun Moon University, Chungnam 336-708,
Republic of Korea
‡Department of Chemistry and Nano Science, Ewha Womans University, Seoul 120-750, Republic of Korea
§Interdisciplinary Program of Biochemical Engineering and Biotechnology, Seoul National University, Seoul 151-742, Republic of Korea
S Supporting Information
b
ABSTRACT: A new reduced hydroxamate, 2,3-dihydrotrichos-
tatin A, was created from trichostatin A by employing a
recombinant strain of Streptomyces venezuelae as a microbial
catalyst. Compared with trichostatin A, 2,3-dihydrotrichostatin
A showed similar antifungal activity against Saccharomyces
cerevisiae, but, interestingly, approximately twice the cytostatic
activity against human small-cell lung cancer cells. The produc-
tion of 2,3-dihydrotrichostatin A via microbial biotransforma-
tion demonstrates that the regiospecific and substrate-flexible
hydrogenation by S. venezuelae provides a new approach for creating natural product analogues with improved bioactive properties.
richostatin A, isolated from the soil bacterium Strpeto-
myces hygroscopicus, is a naturally occurring hydroxamate
T
(Figure 1).1 It was originally discovered as an antifungal antibiotic
but is also a histone deacetylase (HDAC) inhibitor with a broad
spectrum of epigenetic activities, making it a potential anticancer
drug.2ꢀ4 In many cancerous cell lines, overactivation of HDAC
results in histone hypoacetylation, and HDAC inhibitors (HDACi)
specifically target the HDACs, rendering them functionally inactive.
Several HDACi have undergone clinical trials for the treatment of
leukemia and solid tumors: vorinostat is an example of a synthetic
mimic of trichostatin A approved for the treatment of cutaneous
T-cell lymphoma (Figure 1).5,6 Trichostatin A has previously shown
cytostatic activities against various cancer cell lines, such as glioma
cells and bladder cancer cells.7ꢀ9 In recent work, trichostatin A
treatment of small-cell lung cancer (SCLC) cell lines resulted in
decreased cell growth and proliferation.10 The potential of trichos-
tatin A in cancer therapy has generated interest in producing
analogues. In order to modify structurally complex natural products
such as trichostatin A and to create improved bioactive agents,
biological syntheses can be efficient alternatives to chemical ones.
However biotransformed analogues of trichostatin A have not yet
been reported.
Figure 1. Trichostatin A, its bioconverted analogue 2,3-dihydrotrichos-
tatin A, and its synthetic mimic vorinostat.
above-described structural elements are well conserved. Inter-
estingly, the expected reduced analogue (2,3-dihydrotrichostatin
A) resembles the marketed trichostatin A mimic, vorinostat,
more closely than it does the natural original, in that the
polyketide chain spanning from C1 to C6 is more saturated
(Figure 1). The regioselective hydrogenation of this conjugated
double bond using traditional synthesis methods is challenging.
However, we produced and isolated a previously uncharacterized
2,3-dihydrotrichostatin A from S. venezuelae culture supplemen-
ted with trichostatin A. Furthermore, its structure was elucidated
using HPLC-ESIMS/MS and NMR analyses, and its bioactivity
was evaluated against Saccharomyces cerevisiae and human SCLC
DMS53 cells, respectively.
We recently reported a unique and regiospecific hydrogena-
tion of unsaturated macrolide rings by Streptomyces venezuelae,
and its application to heterologous macrolides was found to
synthesize reduced macrolide antibiotics.11 The microbial hydro-
genation system could recognize the specific neighboring struc-
tural elements at each end of the target double bond: a carbonyl
and a methyl group (Figure 1). Reported here is an attempt to
expand the applicability of this unique biocatalyst to use with a
non-macrolide linear polyketide, trichostatin A, in which the
Received: September 20, 2010
Published: April 19, 2011
r
Copyright
2011 American Chemical Society and
1272
dx.doi.org/10.1021/np1006718 J. Nat. Prod. 2011, 74, 1272–1274
American Society of Pharmacognosy
|

Journal of Natural Products
NOTE
Table 1. NMR Spectroscopic Data (500 MHz, CDCl3) of 2,3-
Dihydrotrichostatin A and Trichostatin A
Table 2. Comparison of MIC and IC50 dData of Trichostatin
A and Its Analogue 2,3-Dihydrotrichostatin A
trichostatin A
2,3-dihydrotrichostatin A
2,3-dihydrotrichostatin A
trichostatin A
δC δH
MIC (μg/mL)a
IC50 (nM)a
50.0
25.0
50.0
12.5
position
δC, mult
δH, mult
a Antifungal index (MIC) was determined against S. cerevisiae, whereas
cytostatic index (IC50) was tested against human SCLC DMS53 cell line.
1
2
172.2, C
166.1
117.5
29.8, CH2
2.26, ddd
2.22, ddd
2.20, m
5.27
its chemical shifts with those of trichostatin A (Table 1 and
Supporting Information, Figures S3ꢀS5). The most obvious
difference in the 1H NMR spectra of trichostatin A and its reduced
analogue was the absence of signals at 5.27 and 7.41 ppm, typical
of olefinic protons (H-2,3) in trichostatin A. The upfield shift of
the C-2 and C-3 signals of the parent compound (from 117.5 and
145.0 ppm to 29.8 and 36.2 ppm, respectively) also corroborates
the distinction found in the above 1H NMR data, confirming that
trichostatin A was reduced at the C-2,3 double bond.
3
36.2, CH2
145.0
7.41
2.17, ddd
4
134.1, C
132.8
140.1
40.8
5
133.6, CH
40.9, CH
199.6, C
5.47, d
5.68
3.84
6
3.83, dq
7
199.4
123.8
130.8
110.8
153.6
12.6
8
123.7, C
9, 13
10, 12
11
14
15
16
130.8, CH
110.8, CH
153.6, C
7.78, d
6.82, d
7.78
6.83
The antifungal and cytostatic activities of trichostatin A as a
positive control and 2,3-dihydrotrichostain A were evaluated
against Saccharomyces cerevisiae and the human SCLC cell line,
respectively. The MICs (minimal inhibitory concentrations) of
both compounds against the yeast were 50 μg/mL. However,
2,3-dihydrotrichostatin A (IC50 ∼12.5 nM) was approximately
twice as active as trichostatin A (IC50 ∼25.0 nM) against human
SCLC DMS53 cells. This shows that the C2,3-double bond in
trichostatin A had a minimal effect on antifungal activity, but
enhanced the cytostatic effect on the human SCLC cell line
(Table 2 and Supporting Information, Figure S6). A recent study
reported that significant suppression of human SCLC cell
proliferation was noticed at doses as low as 25 nM of trichostatin
A,10 in agreement with the data of this study. The improved in
vitro cytostatic activity of 2,3-dihydrotrichostatin A caused by the
removal of the C2,3-double bond prompted the investigation of a
three-dimensional docking model of trichostatin A and its
analogue using human HDAC7 enzyme, for which crystallo-
graphic data are available (PDB ID: 3C10).13 The root mean
squared deviation (rmsd) estimated from computational docking
of both ligands into the HDAC7 catalytic site suggests that binding
of 2,3-dihydrotrichostatin A to HDAC7 is relatively more stable
(rmsd ∼0.713 Å) when compared with its parent, trichostatin A
(rmsd ∼0.737 Å) (Supporting Information, Figure S7).
16.0, CH3
17.7, CH3
39.9, CH3
1.82, s
0.94, d
3.05, s
2.21
0.93
3.05
17.8
39.9
A recombinant strain of S. venezuelae YJ028,11 in which the
pikromycin polyketide synthase-encoding gene and desosamine
biosynthetic genes had been deleted, was incubated with trichos-
tatin A. Supplementation of this non-native hydroxamic acid into
the recombinant strain cultivated in 50 mL of SCM media in
baffled Erlenmeyer flasks at 30 °C for 2 days, followed by further
cultivation for 3 days, led to approximately 18% conversion of
trichostatin A into the corresponding reduced trichostatin ana-
logue as the sole metabolite (Supporting Information, Figure S1).
In previous in vivo studies using pikromycin as substrate, the
conversion rate into its dihydro form was approximately 85%.11
In an attempt to increase trichostatin A conversion yield, we
extended the cultivation time after feeding trichostatin A up to a
week. However, no further increased conversion was observed
after day 3, as in the case of pikromycin. We also investigated
using a cell-free extract made from the YJ028 strain using the
same procedure described previously,11 but there was no sig-
nificant increase in the conversion yields (<20%). These results
suggest that the biocatalytic activity of S. venezuelae toward the
non-macrolide linear polyketide is limited compared with its
natural substrate. The organic extracts obtained from several
batch cultures of YJ028 strain fed with trichostatin A were
pooled. The bioconverted trichostatin analogue was purified by
chromatographic isolation using preparative reversed-phase
HPLC with subsequent ESIMS analyses of each fraction
(5 mL/fraction) collected. The MS/MS spectrum of trichostatin
A at m/z 303, representing its protonated molecular ion, showed
fragment ions at m/z 148, 177, and 270, identical to those
obtained from a previous report.12 The MS/MS spectrum of the
trichostatin analogue demonstrated two fragment ions at m/z
148 and 177 common to trichostatin A. However, instead of the
m/z 270 product ion from trichostatin A, there was a character-
istic product ion at m/z 272 (Supporting Information, Figure S2),
suggesting the reduction of one of the double bonds existing at
either the C2,3 or the C4,5 positions of trichostatin A.
’ EXPERIMENTAL SECTION
General Experimental Procedures. Analyses of trichostatin A
(Sigma, St. Louis, MO) and its bioconverted analogue were performed
using a Waters/Micromass Quattro micro/MS interface comprising an
analytical revered-phased Nova-Pak C18 column (Waters, Milford, MA;
150 ꢁ 3.9 mm, 4.0 μm) connected directly to a Micromass Quattro
micro MS. Chromatographic purification of the analogue of interest
was conducted using a preparative HPLC on a Watchers 120 ODS-BP
(250 ꢁ 10.0 mm, 5.0 μm).11 NMR samples were prepared by dissolving
authentic trichostatin A and purified analogue 2,3-dihydrotrichostain A
in 200 μL of CDCl3 and placing the solution in a 5 mm Shigemi
1
advanced NMR microtube (Sigma) matched to the solvent. The H
and 13C NMR spectra were acquired using a Varian INOVA 500
spectrometer at 298 K, with chemical shifts reported in ppm using
TMS as internal reference. All NMR data processing was performed
using Mnova Suite 5.3.2 software.
The chemical structure of the biotransformed analogue was
Biotransformation of Trichostatin A. The mutant strain of S.
further confirmed by 1H and13C NMR spectroscopy by comparing
venezuelae YJ028 has been described previously.11 The organism was
1273
dx.doi.org/10.1021/np1006718 |J. Nat. Prod. 2011, 74, 1272–1274

Journal of Natural Products
NOTE
initially streaked for isolation on SPA agar plates (0.1% yeast extract,
0.1% beef extract, 0.2% tryptose, 1.0% glucose, 1.5% agar, and trace
amount of FeSO4) and then incubated at 30 °C for 2 days. Isolated
colonies were then used to inoculate SCM media (1.5% soluble starch,
2.0% soytone, 0.01% CaCl2, 0.15% yeast extract, and 1.0% MOPS) in
baffled Erlenmeyer flasks. Following 2 days of growth at 30 °C on a
rotary shaker, the cultures were supplemented with trichostatin A
(dissolved in MeOH) at a final concentration of 5 μg/mL and then
incubated for an additional 3 days of cultivation. All in vivo experiments
were independently carried out in triplicate.
’ AUTHOR INFORMATION
Corresponding Author
*Tel: þ82-2-3277-4446. Fax: þ82-2-3277-3419. E-mail:
joonyoon@ewha.ac.kr.
’ ACKNOWLEDGMENT
This work was supported by National Research Laboratory
(NRL) Program (R0A-2008-000-20030-0), Basic Science Research
Program (2009-0073043), and the 21C Frontier Microbial Geno-
mics and Applications Center Program (2010K000890) through the
National Research Foundation of Korea (NRF) funded by the
Ministry of Education, Science & Technology. This research was also
supported by the Technology Development Program (20100623)
for Agriculture and Forestry of the Ministry for Food, Agriculture,
Forestry and Fisheries, Republic of Korea.
Extraction and Isolation of Bioconverted 2,3-Dihydrotri-
chostatin A. Whole cultures were extracted and partitioned twice using
equal volumes of EtOAc in a 250 mL separatory funnel, after which the
organic extracts were then combined and concentrated under vacuum. The
dried residues were immediately dissolved in 200 μL of MeOH, and a
portion of this solution was subjected to HPLC-ESIMS analysis.11
Bioactivity Assays. S. cerevisiae ATCC 9763 and the human SCLC
DSM53 cell line, which were used respectively as test organisms for
antifungal and cytostatic effects, were obtained from American Type
Culture Collection (Manassas, VA). The yeast strain was cultured and
maintained in Antibiotic 19 medium (Difco, BD Biosciences, San Jose,
CA) at 30 °C, whereas human SCLC DMS53 cells were cultured in
Waymouth’s MB752/1 medium (Invitrogen, San Diego, CA), supple-
mented with 10% fetal bovine serum (Sigma), penicillin, and strepto-
mycin (Invitrogen).10 The cell line was incubated in a humidified
atmosphere of 5% CO2 in air at 37 °C. The antifungal activity of the
test compounds was determined using the microdilution method of the
Clinical and Laboratory Standard Institute (CLSI, formerly NCCLS).16
Serial 2-fold dilutions were performed using DMSO to give final
concentrations between 12.5 and 200 μg/mL, with an aliquot of DMSO
(final 0.1%) used as a negative control. The growth of S. cerevisiae was
monitored at 600 nm using a Labsystems Bioscreen C reader, with MIC
determined as the lowest concentration diluted in broth medium that
inhibited the growth of the test strain. To evaluate the cytostatic effects,
proliferation of DMS53 cells following treatment with the two test
compounds was measured using a 3,4-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide (MTT) assay, as previously described.10
In brief, DMS53 cells were trypsinized and plated in 24-well plates and
allowed to adhere overnight. Serial 2-fold dilutions were carried out to
give final concentrations between 3.1 and 200 nM, with an aliquot of
DMSO used as a control. After treatment for 2 days, media were
removed and replaced with the medium containing MTT and incubated
for a further 2 h, after which the absorbance was measured at 540 nm.
IC50 was defined as the lowest concentration at which 50% of the growth
of DMS53 cells was inhibited compared with a control cell line grown in
the absence of test compound. All assays were performed in at least
triplicate.
’ REFERENCES
(1) Tsuji, N; Kobayashi, M.; Nagashima, K.; Wakisaka, Y.; Koizumi,
K. J. Antibiot. 1976, 29, 1–6.
(2) Finnin, M. S.; Donigian, J. R.; Cohen, A.; Richon, V. M.; Rifkind,
R. A.; Marks, P. A.; Breslow, R.; Pavletich, N. P. Nature 1999,
401, 188–193.
(3) Marks, P. A.; Richon, V. M.; Breslow, R.; Rifkind, R. A. Curr. Opin.
Oncol. 2001, 13, 477–483.
(4) Drummond, D. C.; Noble, C. O.; Kirpotin, D. B.; Guo, Z.; Scott,
G. K.; Benz, C. C. Annu. Rev. Pharmacol. Toxicol. 2005, 45, 495–528.
(5) Marks, P. A.; Breslow, R. Nat. Biotechnol. 2007, 25, 84–90.
(6) Codd, R.; Braich, N.; Liu, J.; Soe, C. Z.; Pakchung, A. A. Int.
J. Biochem. Cell Biol. 2009, 41, 736–739.
(7) Kim, M. S.; Blake, M.; Baek, J. H.; Kohlhagen, G.; Pommier, Y.;
Carrier, F. Cancer Res. 2003, 63, 7291–7300.
(8) Wetzel, M.; Premkumar, D. R.; Arnold, B.; Pollack, I. F.
J. Neurosurg. 2005, 103, 549–556.
(9) Li, G. C.; Zhang, X.; Pan, T. J.; Chen, Z.; Ye, Z. Q. Int. J. Urol.
2006, 13, 581–586.
(10) Platta, C. S.; Greenblatt, D. Y.; Kunnimalaiyaan, M.; Chen, H.
J. Surg. Res. 2007, 142, 219–226.
(11) Park, J. W.; Oh, H. S.; Jung, W. S.; Park, S. R.; Han, A. R.; Ban,
Y. H.; Kim, E. J.; Kang, H. Y.; Yoon, Y. J. Chem. Commun. 2008,
44, 5782–5784.
(12) Ueki, M.; Teruya, T.; Nie, L.; Usami, R.; Yoshida, M.; Osada, H.
J. Antibiot. 2001, 54, 1093–1095.
(13) Schuetz, A.; Min, J.; Allali-Hassani, A.; Schapira, M.; Shuen, M.;
Loppnau, P.; Mazitschek, R.; Kwiatkowski, N. P.; Lewis, T. A.;
Maglathin, R. L.; McLean, T. H.; Bochkarev, A.; Plotnikov, A. N.;
Vedadi, M.; Arrowsmith, C. H. J. Biol. Chem. 2008, 283, 11355–11363.
(14) Ommeslaeghe, K. V.; Elaut, G.; Breex, V.; Papeleu, P.; Iterbeke,
K.; Geerlings, P.; Tourwꢀe, D.; Rogiers, V. Bioorg. Med. Chem. Lett. 2003,
13, 1861–1864.
(15) Oger, F.; Lecorgne, A.; Sala, E.; Nardese, V.; Demay, F.;
Chevance, S.; Desravines, D. C.; Aleksandrova, N.; Le Guꢀevel, R.;
Lorenzi, S.; Beccari, A. R.; Barath, P.; Hart, D. J.; Bondon, A.; Carettoni,
D.; Simonneaux, G.; Salbert, G. J. Med. Chem. 2010, 53, 1937–1950.
(16) NCCLS (National Committee for Clinical Laboratory
Standards). Methods for Dilution Antimicrobial Susceptibility Tests for
Bacteria That Grow Aerobically; Approved Standard M7-A5; NCCLS:
Wayne, PA, 2000.
ProteinꢀLigand Docking Model Simulations. Recently pub-
lished crystal complex HDAC7ꢀtrichostatin A was used as a template.13 In
silico molecular modeling analyses to predict the interaction of trichostatin
ligands with the HDAC catalytic site were performed using AutoDock Vina
(version 1.1.1., The Scripps Research Institute, La Jolla, CA).17
’ ASSOCIATED CONTENT
S
Supporting Information. HPLC-ESIMS chromatograms
b
of the biotransformations, ESIMS/MS spectra of trichostatin A and
its converted 2,3-dihydrotrichostatin A, 1H and 13C NMR spectra of
2,3-dihydrotrichostatin A, their antifungal and cytostatic activities
against S. cerevisiae strain and human SCLC cell line, respectively,
and three-dimensional docking modeling of both trichostatins with
the HDAC7 catalytical site. This information is available free of
charge via the Internet at http://pubs.acs.org.
(17) Trott, A.; Olson, A. J. J. Comput. Chem. 2010, 31, 455–461.
1274
dx.doi.org/10.1021/np1006718 |J. Nat. Prod. 2011, 74, 1272–1274