Chemical synthesis enables biochemical and antibacterial evaluation of streptolydigin antibiotics

Article abstract of DOI:10.1021/ja2041964

Inhibition of bacterial transcription represents an effective and clinically validated anti-infective chemotherapeutic strategy. We describe the evolution of our approach to the streptolydigin class of antibiotics that target bacterial RNA polymerases (RN

Full text of DOI:10.1021/ja2041964

ARTICLE
pubs.acs.org/JACS
Chemical Synthesis Enables Biochemical and Antibacterial Evaluation
of Streptolydigin Antibiotics
Sergey V. Pronin,† Anthony Martinez,† Konstantin Kuznedelov,‡ Konstantin Severinov,‡,§
Howard A. Shuman,|| and Sergey A. Kozmin*,†
†Department of Chemistry, and Department of Microbiology, University of Chicago, Chicago, Illinois 60607, United States
‡Department of Molecular Biology and Biochemistry, Rutgers University, Piscataway, New Jersey 08854, United States
§Institutes of Gene Biology and Molecular Genetics, Russian Academy of Sciences, Moscow, Russia
S Supporting Information
b
ABSTRACT:
Inhibition of bacterial transcription represents an effective and clinically validated anti-infective chemotherapeutic strategy.
We describe the evolution of our approach to the streptolydigin class of antibiotics that target bacterial RNA polymerases (RNAPs).
This effort resulted in the synthesis and biological evaluation of streptolydigin, streptolydiginone, streptolic acid, and a series of new
streptolydigin-based agents. Subsequent biochemical evaluation of RNAP inhibition demonstrated that the presence of both
streptolic acid and tetramic acid subunits was required for activity of this class of antibiotics. In addition, we identified 10,11-
dihydrostreptolydigin as a new RNAP-targeting agent, which was assembled with high synthetic efficiency of 15 steps in the longest
linear sequence. Dihydrostreptolydigin inhibited three representative bacterial RNAPs and displayed in vitro antibacterial activity
against S. salivarius. The overall increase in synthetic efficiency combined with substantial antibacterial activity of this fully synthetic
antibiotic demonstrates the power of organic synthesis in enabling design and comprehensive in vitro pharmacological evaluation of
new chemical agents that target bacterial transcription.
’ INTRODUCTION
substitution of the amino-acid residues in the rifamycin binding
site of bacterial RNAP.4 Rapid onset of bacterial resistance is the
primary reason why current use of rifamycins is restricted to
combinations with other drugs, such as isoniazid, or to clinical
emergencies. Thus, there is a significant need for the develop-
ment of new antibiotics that target bacterial RNAPs by different
biochemical mechanisms and display broad-spectrum antibacter-
ial activity. Several other classes of natural products have been
shown to inhibit bacterial RNAPs by binding to alternative
regions of this multisubunit protein, which typically yields
notable antibiotic activity. Such compounds were found to be
effective against rifamycin-resistant RNAPs and strains.5
Cellular RNA polymerase (RNAP) is a multisubunit enzyme
that catalyzes the synthesis of RNA from the corresponding DNA
template. Despite overall structural and functional similarities,
bacterial RNAP sequences are substantially different from those
of the corresponding eukaryotic enzymes, which explains why
rifamycin antibiotics can be used to selectively block bacterial
RNAPs, while having no effect on eukaryotic RNAPs.1 As a result
of potent RNAP inhibition, rifamycins display broad-spectrum
antibacterial activity.2 Three semisynthetic derivatives of rifamy-
cin A, including rifampin, rifapentine, and rifabutin, are currently
in clinical use for treatment of Mycobacterial infections, including
tuberculosis and leprosy.3 Despite the high potency, low toxicity,
and broad antibacterial spectrum of rifamycins, pathogens devel-
op resistance to this class of antibiotics at a relatively high rate by
Received: May 7, 2011
Published: June 29, 2011
r
2011 American Chemical Society
12172
dx.doi.org/10.1021/ja2041964 J. Am. Chem. Soc. 2011, 133, 12172–12184
|

Journal of the American Chemical Society
ARTICLE
Figure 1. Structures of dienoyl tetramic acids antibiotics and degradation fragments. (A) Structure of streptolydigin (1), the parent member of this
antibiotic family, which was isolated from S. lydicus. (B) Structures of streptolic acid (2) and ydiginic acid (3), which were obtained by chemical
degradation of streptolydigin (1). (C) Structures of several derivatives of streptolydigin, which were biosynthetically produced in re-engineered S. lydicus
strains, and structure of tirandalydigin (7), which was isolated from S. tirandis. (D) Structures of two representative tirandamycins (8 and 9), which were
isolated from S. tirandis and marine-derived Streptolyces species, as well as Bu-2313B (10), which was isolated from oligosporic actinomycete strain, no.
E864-61.
Streptolydigin (1, Figure 1A) is a dienoyl tetramic acid
antibiotic,6 which elicits its antibacterial activity by inhibiting
initiation, elongation, and pyrophosphorylation steps of bacterial
RNAP.7 High-resolution X-ray crystallographic characterization
of the streptolydiginꢀRNAP complex revealed a unique bio-
chemical mechanism of RNAP inhibition.8 Streptolydigin (1)
traps the bridge-helix of the RNAP in a straight conformation and
induces opening of the trigger-loop of the enzyme. As a result,
streptolydigin (1) stabilizes the catalytically inactive substrate-
bound transcription intermediate and blocks structural isomer-
ization of RNAP into a fully active state, which requires con-
formational changes of both the bridge-helix and the trigger-
loop moieties.8 The streptolydigin-binding region is located
20 Å away from the rifamycins binding site.9 As a result of this
unique biochemical mechanism of RNAP inhibition and a dis-
tinct binding site, streptolydigin (1) and rifamycins exhibit only
minimal cross-resistance.8b,10
The structure of streptolydigin (1) features an epoxide-con-
taining bicyclic ketal connected by a polyene spacer to a highly
substituted, glycosylated acyl tetramic acid. Structure elucidation
of streptolydigin (1) entailed initial oxidative degradation of the
natural product into two simplified subunits, streptolic acid (2)
and ydiginic acid (3), which derived from the bicyclic ketal
fragment and the tetramic acid subunit of the natural product,
respectively (Figure 1B).11 Complete stereochemical assignment
of streptolic acid (2) was ultimately secured by X-ray crystal-
lographic analysis.12 Re-engineering of streptolydigin biosynth-
esis in S. lydicus enabled recent production of several new
antibiotics shown in Figure 1C, including streptolydiginone
(4),13 which represents a streptolydigin aglycone, as well as
streptolyidigin B (5)14 and streptolydigin LA (6).13 Following
the initial report on isolation of streptolydigin in 1956,6 several
other members of the dienoyl tetramic acid antibiotic family
have been identified, including tirandalydigin (7),15 tirandamy-
cins (i.e., 8 and 9),16 Bu-2312B (10),17 and nocamycins, which
were found to be structurally analogous to 10.18 The bicyclic
ketal subunit of tirandalydigin (7) is identical to that of strepto-
lydigin (Figure 1C). However, the tetramic acid moiety of this
metabolite lacks L-rhodinose and the amide containing side-
chain. Tirandamycins (8 and 9) and Bu-2312B (10) possess the
same unsubstituted acyl tetramic acid subunit of tirandalydigin
(7), but differ in the substitution of the ketal moiety (Figure 1D).
Despite isolation and biosynthetic production of a number of
structurally homologous dienoyl tetramic acid antibiotics over
the years, streptolydigin (1) features the most elaborate structure
and the highest antimicrobial activity recorded within this class.
Evaluation of the antibiotic activity of streptolydigin (1) against a
broad panel of microbial strains using standard broth dilution
testing revealed that this natural product elicited notable activity
against a number of Gram-positive organisms.6a Inhibition of
several representative Clostridium and Streptococcus species by 1
was particularly potent with minimum inhibitory concentrations
(MICs) as low as 0.04 μg/mL. In addition, inhibition of Myco-
bacterium strains, including M. tuberculosis, was also observed.6a
Streptolydigin (1) was shown to be generally more potent than
tirandalydigin (7) and tirandamycin A (8) in a panel of several
anaerobic and aerobic gram-positive bacteria.19 While streptoly-
diginone (4) and streptolydigin B (5) inhibited growth of
S. albus, they were found to be less active as compared to strepto-
lydigin (1) and streptolydigin LA (6).13,14
Streptolydigin (1) was also evaluated in a series of in vivo
studies conducted in mice.6c The maximum tolerated doses of 1
were 1800 mg/kg/day when administered orally and 500 mg/
kg/day when administered subcutaneously.6c Importantly, ad-
ministration of streptolydigin (1) at subcutaneous doses of
100ꢀ330 mg/kg/day protected animals against infection with
S. hemolyticus, S. pneumoniae, and P. multocida.6c Similarly, Bu-
2313B (10) was effective in protecting mice against P. fragilis and
C. perfingens when administered by both oral and subcutaneous
routes.17a
While several total syntheses of tirandamycin A (8)20 have
been developed over the years and the preparation of two
advanced fragments of streptolydigin (1) has been described,21
development of viable synthetic approaches to streptolydigin
(1) and closely related tirandalydigin (7) remained challenging
in part due to the presence of a highly labile alkenyl epoxide
12173
dx.doi.org/10.1021/ja2041964 |J. Am. Chem. Soc. 2011, 133, 12172–12184

Journal of the American Chemical Society
ARTICLE
Scheme 1. Synthetic Strategy to Streptolydiginone (4) and Projected Conversion to Streptolydigin (1)
Scheme 2. Synthesis of Diol 17
Scheme 3. Synthesis of Carboxylic Acid 18
fragment, which is absent in tirandamycin A (8). Indeed, while
a recent synthetic effort enabled assembly of a protected tiran-
dalydigin (7), the final acid-mediated deprotection step was
not successful presumably due to the instability of the epoxide
moiety.22
Following our initial communication of the synthesis of
streptolydigin (1),23 which afforded this natural product with a
longest linear sequence of 24 steps, we describe here the evolu-
tion of our synthetic approach to this important class of RNAP-
targeting antibiotics. This effort resulted in the assembly and
biological evaluation of a series of streptolydigin-based agents,
which demonstrated that the presence of both streptolic acid and
tetramic acid subunits was required for activity of streptolydigin
antibiotics. In addition, we identified a new RNAP-targeting
synthetic antibiotic, 10,11-dihydrostreptolydigin, which was as-
sembled with the high synthetic efficiency of 15 steps in the
longest linear sequence. Dihydrostreptolydigin inhibited three
representative bacterial RNAPs and displayed in vitro antibacter-
ial activity against S. salivarius. The overall increase in synthetic
efficiency combined with substantial antibacterial efficacy of this
synthetic congener of streptolydigin demonstrates the power
of organic synthesis in enabling design and comprehensive in
vitro pharmacological evaluation of new antibiotics that target
bacterial transcription.
from three simplified fragments, including aldehyde 11, diethyl-
phosphono-3-oxobutanthioate 12, and imide 13. The assembly
process would entail the HornerꢀWadsworthꢀEmmons olefi-
nation of aldehyde 11 with phosphonate 12, followed by
N-acylation of imide 13 and Dieckmann cyclization under the
protocol developed by Rinehart.24 The imide 13 would derive
from azide 19, which would be in turn assembled by the Evans
diastereoselective enolate azidation.25 The assembly of aldehyde
11 would entail late-stage installation of the epoxide moiety
starting with appropriately protected vicial diol 14 containing the
requisite bicyclic core of the streptolic acid, which would be
derived from the corresponding triol 15. To control the cis-
geometry of alkene 15, we envisioned that this acyclic product
would be prepared starting from lactone 16 by a controlled
carbonyl addition. The synthesis of lactone 16 would entail
chemoselective acylation of diol 17 with carboxylic acid 18,
followed by a ring-closing metathesis.
Preparation of diol 17 (Scheme 2) commenced with the aldol
reaction of crotonaldehyde (20) with titanium enolate derived
from benzyloxyketone 2126 to give the corresponding syn aldol
product with excellent diastereoselectivity (dr >95:5). Subse-
quent anti-selective EvansꢀTishchenko reduction27 and saponi-
fication of the resulting acetate delivered the target diol 17 in 70%
yield for three steps as a single diastereomer.
’ RESULTS AND DISCUSSION
Syntheses of Streptolic Acid, Methyl Streptolate, and
Streptolydiginone. Guided by the logic of the streptolydigin
biogenesis, which entails final-stage glycosylation of streptolydi-
ginone (4) with L-rhodinose,13 we devised our initial synthetic
strategy shown in Scheme 1. Streptolydiginone (4) would derive
The synthesis of carboxylic acid 18 is shown in Scheme 3.
The sequence began from hydroxy enoate 22, which was readily
prepared by BaylisꢀHillman reaction of benzyl acrylate with
12174
dx.doi.org/10.1021/ja2041964 |J. Am. Chem. Soc. 2011, 133, 12172–12184

Journal of the American Chemical Society
ARTICLE
acetaldehyde. Alcohol transposition in 22 was achieved via its
initial conversion into the corresponding allylic bromide, which
underwent hydrolysis in aqueous buffer to give primary alcohol
23. Subjection of 23 to Sharpless dihydroxylation,28 followed by
chemoselective protection of the corresponding triol as the
acetonide, delivered secondary alcohol 24 in 92% ee. Conversion
of alcohol 24 into the corresponding triflate, followed by
elimination with DBU and hydrolysis of the benzyl ester, gave
the requisite carboxylic acid 18 containing a terminal alkene.
Chemoselective acylation of alcohol 17 with acid 18 was
achieved in the presence of dicyclohexyl carbodiimide (DCC)
and dimethylaminopyridine (DMAP) to give ester 25 (Scheme 4).
Subjection of 25 directly to a series of Ru-based alkene metath-
esis catalysts afforded only low yields of the desired unsaturated
lactone, presumably due to considerable steric bulk presented by
the acetonide moiety. Acid-promoted removal of acetonide (4 N
aqueous HCl in THF), followed by protection of the primary
alcohol as a tri-isoprolylsilyl (TIPS) ether, afforded diene 26.
Ring-closure of 26 to lactone 16 proved to be superior to that
using the acetonide 25. The optimum efficiency was achieved
using Ru catalyst 2729 to give lactone 16 in 80% yield.
Scheme 4. Synthesis of Lactone 16
We next examined the conversion of lactone 16 to the requi-
site bicyclic ketal. While direct conversion of 16 to the corre-
sponding methyl ketone using methylithium proved unsuccess-
ful, this transformation could be efficiently achieved in a two-step
fashion involving initial formation of the Weinreb amide 28,
followed by methyllithium addition (Scheme 5). Subsequent
acid-catalyzed intramolecular acetalization afforded bicyclic ketal
14 in 72% yield for three steps. Reductive deprotection of benzyl
ether, followed by DessꢀMartin oxidation of the resulting
alcohol and Wittig olefination, furnished ester 29, a key structural
subunit for the assembly of streptolydigin antibiotics.
Subsequent elaboration of 29 required installation of the
exocyclic epoxide and conversion of the ester to the aldehyde
moiety. The optimized sequence for this functional group trans-
formation began with 1,2-reduction of unsaturated ester 29 using
DIBAL-H, followed by protection of the resulting allylic alcohol
as a bis-TIPS ether 30 (Scheme 6). Chemoselective removal of
the TIPS ether adjacent to the unprotected tertiary alcohol was
efficiently achieved by subjecting 30 to LiAlH4, presumably due
to the participation of the neighboring alkoxide in the desilyla-
tion step. The resulting diol 31 was converted to the epoxide via
intermediacy of the corresponding primary triflate. Finally,
deprotection of the silyl ether and DessꢀMartin oxidation
efficiently afforded aldehyde 11.
Scheme 5. Synthesis of Bicyclic Ester 29
We also examined the conversion of ester 29 to streptolic acid
(2). While streptolic acid represents a product of chemical
degradation of streptolydigin, the importance of this synthetic
effort was 2-fold. First, this conversion would enable us to verify
the correct structure of ester 29 by correlating analytical data of
the synthetically produced streptolic acid (2) with those present
in the literature. Second, the synthesis of compounds containing
the fully functionalized bicyclic ketal core of streptolydigin,
which was connected to the dienoyl moiety, would enable us
Scheme 6. Synthesis of Methyl Streptolate (33), Streptolic Acid (2), and Aldehyde 11
12175
dx.doi.org/10.1021/ja2041964 |J. Am. Chem. Soc. 2011, 133, 12172–12184

Journal of the American Chemical Society
ARTICLE
Scheme 7. Synthesis of Imide 13
Scheme 9. Synthesis of Streptolydiginone (4)
Scheme 8. Synthesis of 4-Methyl-2,4-hexadienoyl Tetramic
Acid (42)
Scheme 10. Synthesis of Protected L-Rhodinose (48)
to test the importance of the streptolic acid subunit for bio-
logical activity of the parent natural product (vide infra). This
transformation began with the reduction of ester 29 with DIBAL-
H, followed by the DessꢀMartin oxidation of the resulting
alcohol and Wittig olefination to give chain-extended ester 32
with excellent efficiency (Scheme 6). Fluoride-mediated desilyla-
tion of 32, followed by conversion of the resulting diol to an
epoxide via the intermediacy of the corresponding primary
triflate, gave methyl streptolate (33). Subjection of 33 to basic
hydrolysis afforded streptolic acid (2), which proved to be
identical to the corresponding degradation product of streptoly-
digin, thus securing the chemical structure of all prior synthetic
intermediates.11,21c
Having secured synthetic access to the streptolic acid moiety,
we turned our attention to the construction of the tetramic acid
fragment, which would arise from aspartimide 13 (Scheme 7). The
assembly process began with the conversion of known acid 34 to the
corresponding imide 36 by addition of lithium oxazolidinone 35 to
the mixed anhydride derived from 34. The diastereoselective azide
transfer was performed according to the Evans protocol,25 which
entailed generation of the potassium enolate, followed by treatment
with 2,4,6-tri-isopropylbenzenesulfonyl azide (TrisN3). The resulting
imide was treated with methylamine to give amide 19. Reduction
of azide 19 under a hydrogen atmosphere with simultaneous
hydrogenolysis of the benzyl ether, followed by in situ Boc-
protection of the resulting amine, furnished carbamate 37.
Subjection of amide 37 to a TEMPO-mediated oxidative cycliza-
tion afforded the corresponding aspartimide 38, which was
treated with trifluoroacetic acid (TFA) to enable Boc-deprotec-
tion to give primary amine 13.
We next examined the assembly of the simplified methyl
hexadienoyl tetramic acid 42, which corresponded to the struc-
ture of streptolydiginone (4) with a fully truncated bicyclic ketal
(Scheme 8). This sequence was designed to model the final
stages of the streptolydiginone synthesis and would also enable
us to test the importance of the tetramic acid subunit of this agent
on its biological activity (vide infra). The sequence began with
the HornerꢀWadsworthꢀEmmons olefination of unsaturated
12176
dx.doi.org/10.1021/ja2041964 |J. Am. Chem. Soc. 2011, 133, 12172–12184

Journal of the American Chemical Society
ARTICLE
Scheme 11. Revised Synthetic Approach to Streptolydigin (1)
Scheme 12. Synthesis of Phosphonate 49
increased steric bulk of the branched alkyl substituent in the
direct proximity to the reactive NꢀH bond of the tetramic acid.
Synthesis of Streptolydigin. Because of the inability to
glycosylate streptolydiginone (4), we examined an alternative
synthetic approach to streptolydigin (1), which entailed installa-
tion of the glycoside moiety prior to the formation of the final
tetramic acid (Scheme 11). The key disconnection of streptoly-
digin (1) was designed to unite two advanced fragments,
including bicyclic aldehyde 11 and phosphonate 49. The latter
would be assembled by N-glycosylation and N-acylation starting
with three fragments 12, 13, and 48, which were previously
prepared in our route to streptolydiginone (4). It is noteworthy
that phosphonates similar to 49 were prepared by both the
Boeckman and the Schlessinger laboratories.21a,b However, no
attempts to use such fragments for elaboration of streptolydigin
have been reported. The main challenge was to ensure that the
projected assembly process would be compatible with the highly
labile terminal epoxide moiety of aldehyde 11.
Following a series of exploratory studies, we devised a simple,
two-step sequence for the assembly of N-glycosyl phosphonate
49 (Scheme 12). Direct treatment of amine 13 with TES-
protected rhodinose 48 afforded the desired N-glycosylated pro-
duct as a single requisite diastereomer. Subsequent Ag-promoted
N-acylation of this glycosyl amine with thioester 1231 afforded
phosphonate 49 in 60% yield for two steps.
After significant optimization, we ultimately developed a one-
flask protocol for synthesis of streptolydigin (1), which entailed
treatment of the phosphonate 49 with 3 equiv of t-BuOK,
followed by the addition of aldehyde 11 and a rapid acidic workup
with 1 M aqueous solution of HCl at 0 °C (Figure 2A). This one-
pot process entailed several consecutive steps, including initial
deprotonation of 49 to produce enolate I (Figure 2B), Dieck-
mann cyclization to give tetramic acid II, subsequent abstraction
of two acidic protons to give III, followed by the Hornerꢀ
WadsworthꢀEmmons olefination and final neutralization with
concomitant acid-mediated TES deprotection to furnish 1.
Synthetic streptolydigin (1) proved to be identical in all respects
to the authentic sample of the natural product.32 This sequence
enabled the assembly of this natural product with a longest linear
sequence of 24 steps and a higher level of convergency as com-
pared to our initial route, which afforded streptolydiginone (4) in
26 chemical steps.
aldehyde 39 with known phosphonate 1230 to give keto thioester
40. Treatment of 40 with silver(I) trifluoroacetate in the
presence of amine 13 delivered ketoamide 41, which underwent
Dieckmann cyclization12 upon exposure to methanolic sodium
methoxide to give tetramic acid 42 in 90% yield.
The stage was now set for completing the assembly of strepto-
lyidiginone (4). To this end, we subjected aldehyde 11 to the
HornerꢀWadsworthꢀEmmons olefination with phosphonate
1230 (Scheme 9). This transformation efficiently delivered the
desired keto thioester 43. Subsequent treatment of 43 with
silver(I) trifluoroacetate in the presence of amine 13 resulted
in the formation of ketoamide 44. Under such conditions, both
steps proceeded with quantitative efficiency without any detect-
able degradation of the labile epoxide moiety. The next step
entailed the Dieckmann cyclization of 44 using a methanolic
solution of sodium methoxide, according to the Rinehart pro-
tocol,12 to deliverstreptolydiginone (4) in 88% yield and a longest
linear sequence of 26 steps. The structure of 4 was confirmed by
comparison of the 500 MHz 1H NMR and 125 MHz 13C NMR
spectral data for the sodium salt of synthetic streptolydiginone
with the corresponding NMR spectra for this aglycon, which was
isolated by the Salas group from the re-engineered S. lydicus strain
with suppressed glycosylation of streptolydigin.13
We next examined the glycosylation of streptolydiginone (4).
The synthesis of the L-rhodinose moiety of streptolydigin is
depicted in Scheme 10. Protection of the known homoallylic
alcohol 4530 as a triethylsilyl (TES) ether, followed by hydro-
boration-oxidation of the alkene, gave alcohol 46. Hydrogeno-
lysis of the benzyl ether, followed by PhI(OAc)2-mediated
oxidative cyclization of the resulting diol, afforded lactone 47.
Reduction of 47 with DIBAL-H delivered the desired lactol 48.
Our exploratory studies revealed that simple acyl tetramic
acids could be efficiently glycosylated with protected L-rhodinose
derivatives by treatment of the corresponding lactol or glycal
with tetramic acid in choloroform at 20 °C in the presence of a
catalytic amount of p-TsOH. However, glycosylation of the
streptolydiginone’s tetramic acid moiety could not be achieved
under the same reaction conditions presumably due to the
RNAP Inhibitory Activity of Streptolydigin, Streptolydigi-
none, Streptolic Acid, and Tetramic Acid. Our next objective
was to examine the inhibition of bacterial RNAP by streptolydi-
gin (1), streptolydiginone (4), as well as other compounds
prepared during our initial synthetic studies. While antibacterial
activity of several members of this antibiotic family has been
reported, the RNAP inhibitory profile has only been studied for
streptolydigin (1)7,8 and tirandamycin A (8).32 These studies
12177
dx.doi.org/10.1021/ja2041964 |J. Am. Chem. Soc. 2011, 133, 12172–12184

Journal of the American Chemical Society
ARTICLE
Figure 2. Completion of the streptolydigin synthesis. (A) Optimized one-flask protocol for conversion of phosphonate 49 and aldehyde 11 to
streptolydigin (1). (B) Mechanism of streptolydigin formation, which entails initial Dieckmann cyclization to give II, HornerꢀWadsworthꢀEmmons
olefination of aldehyde 11 with III, followed by subsequent acidification of IV with concomitant removal of the TES group to give 1.
50 μM of nucleotide triphosphates (NTPs) and two inhibitor
concentrations (25 and 250 μM, Figure 3A and B, respectively).
The assay consisted of assembling T. aquaticus RNAP complexes
with nucleic acid scaffolds that mimic the structure of nucleic
acids in the transcription elongation complex (Figure 3C). It was
previously shown that artificial transcription elongation com-
plexes assembled on nucleic acid scaffolds are functional35 and
can be used to monitor the effects of various inhibitors of
transcription elongation.36 To monitor RNAP-catalyzed elonga-
tion of the 8-nucleotide-long RNA component of the scaffold, it
was radioactively labeled at the 50 end, allowing its visualization
after denaturing gel-electrophoresis (lane 1). The addition of
NTPs led to the appearance of a runoff product, which resulted
from RNAP-catalyzed extension of the nascent transcript until
the end of the template (lane 2). The addition of synthetic
streptolydigin (1, 25 μM) inhibited the appearance of runoff
transcript (Figure 3A, lane 3), an expected result for a transcrip-
tion elongation inhibitor. A shorter transcript that has undergone
only two cycles of nucleotide addition was formed. At 250 μM of
inhibitor, the amount of the extended transcript decreased, and a
significant amount of the initial nascent transcript remained
intact (a product of a single nucleotide addition was also
observed, Figure 3B, lane 3). The same pattern of transcription
elongation inhibition was observed when an authentic sample of
1, which was isolated from S. lydicus, was used (data not shown).
We also found that streptolydiginone (4), both as a free tetramic
acid (lane 4) and as a corresponding sodium salt (lane 5), was
effective at inhibiting T. aquaticus RNAP. At the highest con-
centration tested, the observed RNAP inhibitory activity of 4 was
visibly lower than that of streptolydigin (compare lanes 3 with
lanes 4 and 5, Figure 3B). This observation is correlated with
weaker antibacterial activity of streptolydiginone (4) against S.
albus, which was reported previously.13
Figure 3. In vitro inhibition of bacterial transcription catalyzed by T.
aquaticus RNAP. Artificial transcription elongation complexes contain-
ing radioactively labeled 8-nt RNA primer were assembled on nucleic-
acid scaffolds (C) using T. aquaticus RNAP core enzyme. Where
indicated, reaction mixtures were supplemented with 25 (A) or 250
μM (B) of small-molecule inhibitors, and transcription was initiated by
the addition of NTPs. Reaction products were resolved by denaturing
polyacrylamide gel electrophoresis and revealed by autoradiography.
demonstrated that tirandamycin A was a substantially less potent
bacterial RNAP inhibitor.33 The antimicrobial data of dienoyl
tetramic acid antibiotics have been obtained using only a small
number of compounds, isolated from various microbial produ-
cers, which provided only limited structureꢀactivity relationship
data. While notable success in re-engineering streptolydigin
biosynthesis has been achieved, this approach has thus far
provided only a few compounds with modified glycoside and
tetramic acid subunits of the parent natural product.13,14
Our study was aimed at evaluating the importance of two
major structural subunits of streptolydigin (1), including strep-
tolic acid and the tetramic acid fragments. To this end, we
examined inhibition of a well-characterized Thermus aquaticus
RNAP34 by following in vitro transcription in the presence of
We were unable to observe any RNAP inhibition by either the
streptolic acid (2, lane 6) or the corresponding methyl ester (33,
lane 7). Furthermore, the truncated tetramic acid subunit of
streptolydiginone, which is represented by compound 42, was
also inactive at both concentrations tested (lane 8). These results
demonstrate that the presence of both structural subunits of
streptolydigin is essential for activity of this class of antibiotics.
12178
dx.doi.org/10.1021/ja2041964 |J. Am. Chem. Soc. 2011, 133, 12172–12184

Journal of the American Chemical Society
ARTICLE
Scheme 14. Synthesis of 10,11-Dihydrostreptolydigin (63)
Figure 4. General strategy to simplified analogues of streptolydigin.
The synthetic approach would entail a sequential condensation of four
reaction partners, including aldehyde-containing fragment A, phospho-
nate 12, imide 13, and protected L-rhodinose 48. The projected
assembly process entails N-glycosidation, N-acylation, Dieckman cycli-
zation, olefination, and final deprotection.
Scheme 13. Synthesis of Adamantane-Substituted Dienoyl
Tetramic Acid (56)
of the individual structural subunits of streptolydigin would
result in a substantial decrease of the binding efficiency and loss
of inhibitory activity.
Rationally Simplified Streptolydigin Analogues: Design,
Synthesis, and RNAP Inhibition. Our initial biochemical anal-
ysis revealed that truncation of either streptolic or tetramic acid
subunits of streptolydigin resulted in complete loss of RNAP
inhibitory activity. We envisioned that additional modifications
could be introduced into the streptolic subunit of the natural
product provided that the fully functionalized tetramic acid
fragment remains intact (Figure 4). Specifically, we planned to
remove the C(10)ꢀC(11) alkene moiety in the bicyclic ketal
core of the natural product (fragment A1, Figure 4). This
structural alteration was expected to considerably increase the
efficiency of our assembly process (vide infra). In addition, we
intended to substitute the entire bicyclic ketal moiety with an
adamantyl fragment, which was expected to preserve the overall
shape and most of the physicochemical properties of this
relatively hydrophobic subunit (fragment A2, Figure 4). Each
new compound would derive from a highly convergent sequence
that would entail a condensation of four building blocks follow-
ing our previously established synthetic sequences (Figure 4).
Because of structural simplification of fragment A, we envisioned
that such derivatives would be assembled with significantly
higher synthetic efficiency as compared to our original routes
to streptolydigin (1) and streptolydiginone (4), which required
24 and 26 steps, respectively.
This conclusion is also in agreement with previous crystal-
lographic studies of streptolydiginꢀRNAP complexes, which
established that both streptolic and tetramic fragments of the
natural product interacted with the bridge-helix and the trigger-
loop regions of the protein, respectively.8 Thus, removal of each
The synthesis of the adamantane-bearing derivative began
with asymmetric Evans alkylation37 of chiral imide derived
from the commercially available carboxylic acid 50 (Scheme 13).
This process delivered the requisite alkylation product in good
12179
dx.doi.org/10.1021/ja2041964 |J. Am. Chem. Soc. 2011, 133, 12172–12184

Journal of the American Chemical Society
ARTICLE
Figure 5. In vitro inhibition of transcription catalyzed by T. aquaticus RNAP by streptolydigin derivatives. (A) A comparison of RNAP inhibitory activity
of streptolydigin (1), 10,11-dihydrostreptolydigin (63), and adamantane derivative 56 using 10 μM NTP and 0.1 mM of each inhibitor. (B) Dose-
dependent RNAP inhibition by streptolydigin (1), streptolydiginone (4), and 10,11-dihydrostreptolydigin (63). (C) Quantitative assessment of
inhibitory effects of streptolydigin (1), streptolydiginone (4), and 10,11-dihydrostreptolydigin (63) on the rate of CMP incorporation into nascent 8-nt
starting RNA.
efficiency and excellent diastereomeric purity. Subsequent ela-
boration entailed reduction of imide 51 with LiBH4, followed by
DessꢀMartin oxidation of alcohol 52, and Wittig olefination of
the resulting aldehyde to give unsaturated ester 53. Reduction of
53 with DIBAL-H, followed by DessꢀMartin oxidation of the
allylic alcohol 54, afforded aldehyde 55. Treatment of aldehyde
55 with phosphonate 49 in the presence of excess of t-BuOK
using our previously developed one-flask protocol for sequential
olefination-cyclization-deprotection afforded requisite adaman-
tane-containing tetramic acid 56 in 59% yield (13-step longest
linear sequence from commercially available precursors).
Construction of 10,11-dihydrostreptolydigin (63) com-
menced with a Ti-mediated aldol reaction between ketone 21
and aldehyde 57, followed by an anti-reduction of the corre-
sponding hydroxyketone to afford diol, which was protected as
acetonide 58 (Scheme 14). Bromineꢀlithium exchange and
subsequent acylation gave the expected enone, which was treated
with PPTS in methanol to generate bicyclic ketal 59. Concave-
face epoxidation of alkene 59 delivered desired epoxide 60.
Hydrogenolysis of 60, followed by DessꢀMartin oxidation of
the resulting alcohol, afforded aldehyde 61. Subjection of 61 to a
Wittig olefination furnished an unsaturated ester, which was
subjected to reduction with DIBAL-H, followed by Dessꢀ
Martin oxidation to deliver aldehyde 62. Final assembly of
10,11-dihydrostreptolydigin (63) was performed under the
conditions developed for the preparation of streptolydigin.
Treatment of ketoamide 49 with excess potassium tert-butoxide,
followed by addition of aldehyde 62 and acidic workup, afforded
10,11-dihydrostreptolydigin (63). Removal of the C(10)ꢀ
C(11) alkene moiety enabled significant increase in the overall
synthetic efficiency of this route (15-step longest linear
sequence) as compared to our original approach to streptolydi-
gin, which required 24 synthetic steps.
positive control (lane 2, Figure 5A). We found that 10,11-
dihydrostreptolydigin (63) inhibited T. aquaticus RNAP (lane
3, Figure 5A) very similarly to streptolydigin (1), with transcript
elongation blocked after incorporation of two nucleotides into
the growing RNA strand. This observation suggested that
removal of the C(10)ꢀC(11) alkene preserved a substantial
amount of the parent activity from the parent natural product 1
(lane 2, Figure 5A). Interestingly, we did not observe any
inhibition of RNAP by adamantane-containing analogue 56 at
100 μM (lane 4, Figure 5A) or 1 mM (data not shown)
concentrations. This finding indicated that the adamantane core
of 56 did not provide a viable replacement for a bicyclic ketal
moiety of streptolydigin (1).
We next compared the relative efficacy of T. aquaticus RNAP
inhibition by three compounds, which were shown to elicit
activity, including streptolydigin (1), streptolydiginone (4),
and 10,11-dihydrostreptolydigin (63). The results are presented
in Figure 5B. At the two highest concentrations of streptolydigin
(1) (0.1 and 1.0 mM), the transcription was blocked either
completely or following the addition of one or two nucleotides
(Figure 5B, lanes 7 and 11). At the same concentrations of
streptolydiginone (4) and 10,11-dihydrostreptolydigin (63),
inhibition occurred almost exclusively after incorporation of
two nucleotides (Figure 5B, lanes 8, 9, 12, and 13). These results
once again confirmed that 10,11-dihydrostreptolydigin (63) and
streptolydiginone (4) were effective at inhibiting transcription in
vitro. This conclusion was further supported by an experiment
conducted in the presence of subsaturating (10 μM) concentra-
tions of inhibitors (Figure 5B, lanes 3, 4, and 5). Similar results
were obtained in experiments using a closely related T. thermo-
philus RNAP, as well as RNAPs from M. smegmatis (data not
shown). RNAPs from Gram-negative organisms such as Franci-
sella tularensis and E. coli required much higher concentrations of
inhibitors to achieve the same extent of inhibition (data not
shown), confirming earlier observations that RNAPs from Gram-
negative bacteria are significantly more resistant to streptolydigin
than their counterparts from Gram-positive organisms.
Having completed the syntheses of compounds 56 and 63, we
examined their ability to inhibit transcription catalyzed by T.
aquaticus RNAP (Figure 5) using a nucleic acid scaffold similar to
the one presented in Figure 3C but having a longer downstream
(transcribed) segment. Streptolydigin (1) was included as a
12180
dx.doi.org/10.1021/ja2041964 |J. Am. Chem. Soc. 2011, 133, 12172–12184

Journal of the American Chemical Society
ARTICLE
Figure 6. Antimicrobial activity of synthetic streptolydigin antibiotics against S. salivarius. Each compound was tested according to the general protocol
described in the Experimental Procedures. Most potent compounds, including 1, 4, and 63, were tested at several concentrations shown. All other
compounds were tested at a concentration of 100 μg/disk.
We also followed the rate of single nucleotide incorporation
catalyzed by T. aquaticus RNAP in the presence of streptolydigin
(1), streptolydiginone (4), and 10,11-dihydrostreptolydigin
(63). This study provided a quantitative basis for assessing the
difference in inhibitory activities of each of the three compounds
tested (Figure 5C). In this assay, transcription complexes con-
taining radioactively labeled nascent transcript were supplemen-
ted with CTP, which was necessary to extend the transcript by
only one nucleotide. The concentration of CTP (5 μM) and the
reaction temperature (25 °C) were kept low to slow the RNAP
catalysis rate, which enabled following the CMP addition in real
time in the presence of inibitors. In the absence of inhibitors, the
reaction was complete (full conversion of 8-nt starting RNA to
9-nt product) at the first time point sampled (15 s, data not
shown). In the presence of inhibitors, the reaction rate was
slowed considerably, with full extension achieved at 5 min for
streptolydiginone (4) and 10,11-dihydrostreptolydigin (63),
which appeared to inhibit nucleotide addition with similar
efficiency. Streptolydigin (1) inhibition was more potent, with
nascent transcript extension reaction being ∼50% complete after
20 min of incubation. The data obtained were used to derive the
apparent rates of RNAP-catalyzed single nucleotide extension,
which were found to be 0.11, 1.08, and 1.24 minꢀ1 in the
presence of streptolydigin (1), 10,11-dihydrostreptolydigin
(63), and streptolydiginone (4), respectively. Overall, the results
of in vitro transcription assays establish that removal of either the
glycoside moiety of streptolydigin or the C(10)ꢀC(11) alkene
preserved RNAP inhibitory activity of the resulting compounds,
while weakening their in vitro potency by about 10-fold.
Antimicrobial Activity against S. salivarius. We also exam-
ined antimicrobial activity of all streptolydigin-based compounds
prepared in our study. Originally, streptolydigin (1) was found to
be effective in inhibiting several Streptococcus species, including S.
faecalis, S. hemolyticus, S. lactis, S. mitis, S. viridans, and S.
pneumoniae with MICs in the range of 0.19ꢀ3.12 μg/mL.6a
Therefore, we selected S. salivarius, a representative Streptococcus
strain, which typically colonizes human oral and nasopharyngeal
epithelia and only becomes harmful to the host if the immune
status is altered or there is a loss of control of epithelial cell
sensing and discriminatory systems.38 Our main obective was to
use this model, normally nonpathogenic organism to correlate
the initially observed RNAP inhibitory activity of all compounds
with their antibacterial action. The results of this study, which
employed a standard disk-diffusion protocol, are shown in
Figure 6. As expected, streptolydigin (1) inhibited growth of S.
salivarius in a dose-dependent manner at concentrations of 5, 10,
and 20 μg per disk (Figure 6A). Streptolydiginone (4) was also
found to elicit antibacterial activity, albeit not as potent as that of
streptolydigin (1). While substantial growth inhibition was
observed at 40 μg of 4 per disk, no effect on bacterial proliferation
occurred at a lower dose of 10 μg per disk (Figure 6B).
Evaluation of 10,11-dihydrostreptolydigine (63) revealed dose-
dependent antibacterial action at concentrations of 20 and 40 μg
per disk (Figure 6C). Similar to streptolydiginone (4), no
antibacterial activity was observed with 63 at 10 μg per disk
(Figure 6C). Overall, the observed antibacterial activities of 1, 4,
and 63 were in excellent correlation with their RNAP inhibitory
profiles. In addition, streptolic acid (2) and methyl streptolate
(33) were found to be completely inactive even at a high
concentration of 100 μg/disk (Figure 6B and C). The adaman-
tane-containing derivative 56 and tetramic acid 42 elicited only
weak antibacterial action at 100 μg/disk (Figure 6A and B) and
no activity at lower concentrations tested (data not shown). The
lack of activity of 2, 33, 42, and 56 was also in excellent agreement
with the inability of such compounds to inhibit T. aquaticus
RNAP. This study demonstrated once again that the presence of
both streptolic and tetramic subunits of streptolydigin is required
for its antibiotic activity and established that RNAP inhibition of
streptolydigin-based compounds was well correlated with their
antibacterial profile.
’ CONCLUSIONS
We described the development of a general synthetic strategy
to streptolydigin antibiotics, as well as their comprehensive
biochemical and antimicrobial evaluation. Our initial studies
resulted in the total synthesis of streptolydiginone (4) and
streptolydigin (1) and enabled the assembly of a series of
simplified derivatives of the parent natural products, which
were designed to assess the importance of the individual
streptolic and tetramic acid subunits. Indeed, the presence of
both major fragments of streptolydigin was found to be
essential for its activity. This finding was in agreement with
previous X-ray crystallographic characterization of streptolydi-
ginꢀRNAP complexes. In addition, our effort enabled a 15-step
synthesis of 10,11-dihydrostreptolydigin, which inhibited three
bacterial RNAPs and elicited antimicrobial activity against S.
salivarius. The ability to access efficiently simplified analogues
of streptolydigin that elicit substantial antibacterial activity
demonstrates the power of our general synthetic approach,
which could be employed for subsequent rational design and
12181
dx.doi.org/10.1021/ja2041964 |J. Am. Chem. Soc. 2011, 133, 12172–12184

Journal of the American Chemical Society
ARTICLE
the development of new streptolydigin-based antibiotics with
enhanced potency and improved pharmacological profile.
added. Phases were separated, and the organic phase was washed
three times with water and once with brine, dried over MgSO4, and
concentrated in vacuo. Purification by flash chromatography on silica gel
(chloroform, followed by the elution with chloroform:methanol 100:3)
afforded 9.0 mg (60% yield) of streptolydigin (1). 1H NMR (500 MHz,
CDCl3) δ 0.70 (d, 3H, J = 6.9 Hz), 1.05 (d, 3H, J = 6.9 Hz), 1.11 (m,
6H), 1.23 (s, 3H), 1.77 (m, 1H), 1.90 (s, 3H), 1.94 (m, 1H), 2.07 (d, 1H,
J = 13.3 Hz), 2.50 (q, 1H, J = 11.7 Hz), 2.78 (m, 1H), 2.81 (d, 1H, J = 5
Hz), 2.89 (d, 3H, J = 4.4 Hz), 2.98 (d, 1H, J = 5 Hz), 3.05 (q, 1H, J = 6.9
Hz), 3.41 (s, 1H), 3.65 (m, 2H), 4.35 (t, 1H, J = 4.6 Hz), 4.86 (s, 1H),
5.59 (d, 1H, J = 11.2 Hz), 5.62 (d, 1H, J = 10.1 Hz), 5.85 (m, 1H), 6.25
(d, 1H, J = 9.9 Hz), 6.34 (dd, 1H, J = 10.1, 4.8 Hz), 7.16 (d, 1H, J = 15.6
Hz), 7.57 (d, 1H, J = 15.6 Hz); 13C NMR (125 MHz, CDCl3) δ 10.1,
12.2, 12.5, 17.1, 17.2, 21.2, 22.2, 26.8, 30.2, 34.1, 35.2, 42.1, 50.5, 55.0,
62.8, 66.3, 71.4, 76.1, 76.5, 78.9, 98.9, 99.7, 116.2, 130.5, 133.8, 134.0,
146.1, 150.5, 173.7, 175.0, 175.2, 193.9; HRMS (ESI) calculated for
C32H45N2O9 [M + H]+ 601.3125, found 601.3121. The optical rotation
of synthetic sample of 41 was determined to be [R]22D = ꢀ16.2 (c = 0.7,
CHCl3), which was in good agreement with that of an authentic sample
of streptolydigin obtained from ChemCon GmbH.
’ EXPERIMENTAL PROCEDURES
Streptolic Acid (2). A solution of methyl streptolate 33 (13.0 mg,
0.039 mmol) in a mixture of methanol (1 mL) and water (0.1 mL) was
treated with sodium hydroxide (16.0 mg, 0.39 mmol). The reaction
mixture was left overnight at room temperature, concentrated in vacuo,
and treated with aqueous HCl (0.4 mL of 1 N solution). The resulting
solution was extracted twice with ethyl acetate. Combined organic
extracts were washed with brine, dried over MgSO4, and concentrated
in vacuo. Preparative TLC (development with ethyl acetate:methanol
20:1) afforded 10.0 mg (77% yield) of streptolic acid (2). [R]24
=
D
+135.0 (c = 0.7, EtOH); 1H NMR (500 MHz, CDCl3) δ 0.69 (d, 3H, J =
7.0 Hz), 1.04 (d, 3H, J = 7.0 Hz), 1.23 (s, 3H), 1.80 (s, 3H), 1.94 (m,
1H), 2.73 (m, 1H), 2.81 (d, 1H, J = 5.0 Hz), 2.98 (d, 1H, J = 5.0 Hz), 3.63
(d, 1H, J = 10.5 Hz), 4.35 (t, 1H, J = 4.5 Hz), 5.62 (d, 1H, J = 10.0 Hz),
5.81 (d, 1H, J = 15.5 Hz), 6.13 (d, 1H, J = 10.0 Hz), 6.34 (dd, 1H, J =
10.0, 5.0 Hz), 7.44 (d, 1H, J = 15.5 Hz); 13C NMR (125 MHz, CDCl3) δ
12.2, 12.5, 17.1, 22.2, 33.7, 35.0, 50.5, 55.0, 71.4, 76.1, 98.8, 115.1, 130.5,
132.6, 133.8, 143.3, 152.0, 172.5; HRMS (ESI) calculated for C18H25O5
[M + H]+ 321.1702, found 321.1702.
Adamantane-Containing Tetramic Acid (56). A stirred solu-
tion of phosphonate 49 (23.0 mg, 0.039 mmol) in THF (0.3 mL) was
treated with potassium tert-butoxide (0.12 mL of 1 M solution in THF,
0.12 mmol) at 0 °C. The resulting solution was stirred 30 min and
transferred into a solution of aldehyde 55 (6.0 mg, 0.0259 mmol) in
THF (0.1 mL) at 0 °C. The reaction was warmed to room temperature
and left overnight. THF (0.4 mL) was added, and the reaction was
quenched with aqueous HCl (0.15 mL of 1 M solution, 0.15 mmol, slight
excess with respect to potassium tert-butoxide) at 0 °C. After 1 h, ethyl
acetate and water were added. Phases were separated, and the organic
phase was washed three times with water and once with brine, dried over
MgSO4, and concentrated in vacuo. Purification by flash chromatogra-
phy on silica gel (chloroform, followed by the elution with chloroform:
Streptolydiginone (4). A stirred solution of ketoamide 44 (16.0
mg, 0.033 mmol) in methanol (0.7 mL) was treated with sodium
methoxide (0.33 mL of 0.5 M solution in methanol, 0.165 mmol) at
0 °C. The reaction mixture was stirred overnight at room temperature
and concentrated in vacuo. The residue was dissolved in water (1 mL),
treated with aqueous HCl (0.17 mL of 1 N solution) at 0 °C, and
extracted twice with ethyl acetate. The combined organic extracts were
washed with water, brine, then dried over MgSO4 and concentrated in
vacuo to afford 14.0 mg (88% yield) of streptolydiginone (4). [R]25
=
D
+24.9 (c = 0.3, CHCl3); 1H NMR (500 MHz, CDCl3) δ 0.70 (m, 3H),
1.05 (m, 6H), 1.23 (s, 3H), 1.90 (s, 3H), 1.93 (m, 1H), 2.73ꢀ2.91 (m,
6H), 2.98 (d, 1H, J = 5.0 Hz), 3.65 (d, 1H, J = 10.5 Hz), 4.10 (s, 1H),
4.35 (t, 1H, J = 4.5 Hz), 5.63 (d, 1H, J = 10.1 Hz), 5.67 (s, 1H), 6.26 (d,
1H, J = 9.9 Hz), 6.34 (dd, 1H, J = 10.1, 4.7 Hz), 7.10 (d, 1H, J = 15.6 Hz),
7.58 (d, 1H, J = 15.6 Hz); 13C NMR (125 MHz, CDCl3) δ 11.4, 12.2,
12.5, 17.1, 22.2, 26.4, 34.1, 35.2, 40.9, 50.5, 55.0, 62.9, 71.4, 76.1, 98.9,
100.3, 116.2, 130.6, 133.8, 133.9, 146.0, 150.3, 175.0, 175.5, 175.7, 194.0;
HRMS (ESI) calculated for C26H35N2O7 [M + H]+ 487.2444, found
487.2442. A portion of the product was treated with dilute aqueous
NaHCO3, concentrated in vacuo, and azeotroped three times with
chloroform. The residue was extracted with chloroform, and the solution
phase was concentrated in vacuo and azeotroped with chloroform to
afford a sodium salt of streptolydiginone. 1H NMR (500 MHz,
(CD3)2SO) δ 0.65 (d, 3H, J = 6.9 Hz), 0.78 (d, 3H, J = 6.9 Hz), 0.98
(d, 3H, J = 6.8 Hz), 1.08 (s, 3H), 1.78 (s, 3H), 1.78 (m, 1H), 2.57 (d, 3H,
J = 4.2 Hz), 2.66 (q, 1H, J = 6.9 Hz), 2.75 (m, 1H), 2.89 (d, 1H, J = 5.0
Hz), 2.94 (d, 1H, J = 5.0 Hz), 3.58 (m, 2H), 4.10 (s, 1H), 4.33 (t, 1H, J =
4.6 Hz), 5.62 (d, 1H, J = 10.1 Hz), 5.76 (d, 1H, J = 9.7 Hz), 6.08 (s, 1H),
6.39 (dd, 1H, J = 10.0, 4.6 Hz), 7.01 (d, 1H, J = 15.6 Hz), 7.67 (d, 1H, J =
15.6 Hz), 7.84 (q, 1H, J = 4.3 Hz); 13C NMR (125 MHz, (CD3)2SO) δ
10.5, 12.2, 12.6, 17.3, 22.3, 25.6, 32.9, 34.7, 40.3, 49.8, 54.8, 60.4, 70.4,
75.9, 98.2, 102.0, 126.8, 130.2, 133.6, 134.0, 137.5, 141.1, 175.0, 176.4,
181.4, 193.9.
methanol 100:3) afforded 8.5 mg (59% yield) of 56. [R]25 = ꢀ96.4
D
(c = 0.8, CHCl3); 1H NMR (500 MHz, CDCl3) δ 0.90 (d, 3H, J = 7.0
Hz), 1.10 (m, 6H), 1.46 (d, 3H, J = 11.5 Hz), 1.54 (d, 3H, J = 11.5 Hz),
1.60 (d, 3H, J = 12.0 Hz), 1.67 (d, 3H, J = 12.0 Hz), 1.77 (m, 1H), 1.88
(s, 3H), 1.95 (s, 3H), 2.04 (m, 1H), 2.23 (m, 1H), 2.50 (q, 1H, J = 12.0
Hz), 2.89 (d, 3H, J = 4.5 Hz), 3.05 (q, 1H, J = 7.0 Hz), 3.40 (s, 1H), 3.64
(m, 1H), 4.86 (s, 1H), 5.57 (d, 1H, J = 11.0 Hz), 5.85 (s, 1H), 6.05
(d, 1H, J = 10.5 Hz), 7.12 (d, 1H, J = 15.5 Hz), 7.55 (d, 1H, J = 15.5 Hz);
13C NMR (125 MHz, CDCl3) δ 10.1, 12.5, 13.3, 17.2, 21.2, 26.8, 28.7,
30.2, 36.0, 37.2, 39.8, 42.1, 44.0, 62.8, 66.3, 76.5, 78.9, 99.5, 115.5, 133.6,
150.6, 151.2, 173.8, 175.1, 175.4, 193.8.
10,11-Dihydrostreptolydigin (63). A stirred solution of phos-
phonate 49 (23.6 mg, 0.04 mmol) in THF (0.3 mL) was treated with
potassium tert-butoxide (0.12 mL of 1 M solution in THF, 0.12 mmol) at
0 °C. The resulting solution was stirred 30 min and transferred into a
solution of aldehyde 62 (7.5 mg, 0.0268 mmol) in THF (0.1 mL) at
0 °C. The reaction was warmed to room temperature and left overnight.
THF (0.4 mL) was added, and the reaction was quenched with aqueous
HCl (0.15 mL of 1 M solution, 0.15 mmol, slight excess with respect to
potassium tert-butoxide) at 0 °C. After 1 h, ethyl acetate and water were
added. Phases were separated, and the organic phase was washed three
times with water and once with brine, dried over MgSO4, and concen-
trated in vacuo. Purification by flash chromatography on silica gel
(chloroform, followed by the elution with chloroform:methanol
100:3) afforded 9.0 mg (56% yield) of 10,11-dihydrostreptolydigin
(63). [R]25D = ꢀ53.3 (c = 0.9, CHCl3); 1H NMR (500 MHz, CDCl3)
δ 0.76 (d, 3H, J = 7.0 Hz), 1.07ꢀ1.12 (m, 9H), 1.22 (s, 3H), 1.73ꢀ1.82
(m, 4H), 1.90 (s, 3H), 1.93 (m, 1H), 2.04 (m, 1H), 2.14 (m, 1H), 2.51
(q, 1H, J = 12.0 Hz), 2.64 (d, 1H, J = 5.0 Hz), 2.81 (d, 1H, J = 5.0 Hz),
2.84 (m, 1H), 2.89 (d, 3H, J = 4.5 Hz), 3.05 (q, 1H, J = 7.0 Hz), 3.40
(s, 1H), 3.65 (m, 1H), 3.82 (d, 1H, J = 11.0 Hz), 3.98 (m, 1H), 4.86
Streptolydigin (1). A stirred solution of phosphonate 49 (30.0 mg,
0.05 mmol) in THF (0.4 mL) was treated with potassium tert-butoxide
(0.15 mL of 1 M solution in THF, 0.15 mmol) at 0 °C. After 40 min,
aldehyde 11 (7.0 mg, 0.0252 mmol) was added in one portion. The
reaction was warmed to room temperature and left overnight. THF
(0.5 mL) was added, and the reaction was quenched with aqueous HCl
(0.18 mL of 1 M solution, 0.18 mmol, slight excess with respect to
potassium tert-butoxide) at 0 °C. After 1 h, ethyl acetate and water were
12182
dx.doi.org/10.1021/ja2041964 |J. Am. Chem. Soc. 2011, 133, 12172–12184

Journal of the American Chemical Society
ARTICLE
(s, 1H), 5.59 (d, 1H, J = 11.0 Hz), 5.85 (d, 1H, J = 4.0 Hz), 6.24 (d, 1H,
J = 10.0 Hz), 7.15 (d, 1H, J = 15.5 Hz), 7.56 (d, 1H, J = 16.0 Hz); 13C
NMR (125 MHz, CDCl3) δ 10.1, 12.3, 12.7, 17.2, 21.0, 21.2, 22.5, 26.8,
27.2, 27.4, 30.2, 30.4, 35.1, 35.3, 42.1, 51.3, 58.5, 62.8, 66.3, 70.3, 76.5,
78.9, 98.2, 99.7, 116.1, 134.3, 146.3, 150.6, 173.7, 175.0.
Academy of Sciences (to K.S.). We also thank Dmitry Vassyliev
for helpful discussions and Rodrigo J. Carbajo for providing
copies of NMR spectra of 4.
RNAP and Scaffolds. For in vitro transcription experiments,
recombinant T. aquaticus core RNAP was purified as described earlier.34
To assemble a previously described35 nucleic acid scaffold used in
Figure 3 experiment, the following oligonucleotide strands were used:
RNA (50-GUAGCGGA-30), template DNA (30-CATCGCCTGTACA-
TTTCAGACAGGACC-50), and nontemplate DNA (50-TGTAAAGT-
CTGTCCTGG-30). To assemble a scaffold with longer transcribed part
used in Figure 5, the same RNA was combined with 30-CATCGCCTG-
TACATTTCAGACAGGACCAGGTGTTGGG-50 template DNA and
50-TGTAAAGTCTGTCCTGGTCCACAACCC-30 nontemplate DNA.
’ REFERENCES
(1) Floss, H. G.; Yu, T.-W. Chem. Rev. 2005, 105, 621–632.
(2) Wehrli, W.; Staehelin, M. Bacteriol. Rev. 1971, 35, 290–309.
(3) Shinnick, T., Ed. Current Topics in Microbiology and Immunology;
Academic Press: New York, 1996.
(4) (a) Wehrli, W. Rev. Infect. Dis. 1983, 5, S407–11. (b) David, H. L.
Appl. Microbiol. 1970, 20, 810–814. (c) Gillespie, S. H. Antimicrob.
Agents Chemother. 2002, 46, 267–274.
(5) (a) Chopra, I. Curr. Opin. Invest. Drugs 2007, 8, 600–607. (b)
Darst, S. A. Trends Biochem. Sci. 2004, 29, 159–160. (c) Villain-Guillot,
P.; Bastide, L.; Gulatieri, M.; Leonetti, J. Drug Discovery Today 2007,
12, 159–160. (d) Ho, M. X.; Hudson, B. P.; Das, K.; Arnold, E.; Ebright,
R. H. Curr. Opin. Struct. Biol. 2009, 19, 715–723.
32
To visualize transcription products, RNA was P-labeled at the 50-
end with T4 polynucleotide kinase before scaffold assembly.
Transcription Assays. Transcription elongation complexes were
reconstituted by combining 200 nM RNAP and 50 nM nucleic acids
scaffold in 10 μL of transcription buffer (25 mM Tris-HCl, 50 mM NaCl,
5 mM MgCl2, 0.5 mM DTT, pH 7.9) and incubation at 37 °C for 10 min.
Following the addition of streptolydigin derivatives and a 10 min
incubation at 37 °C, transcription was initiated by the addition of NTPs
at concentrations indicated in the figures. After a 3 min incubation at
37 °C, reactions were terminated by the addition of 1 volume of
formamide loading buffer. After being heated at 100 °C for 1 min,
samples were separated by 20% PAGE with 7 M Uurea, followed by
PhosphorImager analysis. For single-nuceotide addition experiments,
the concentration of CTP was 5 μM, and reaction incubation tempera-
ture was 25 °C. Reaction incubation times varied and are indicated in
Figure 5C. For data analysis, the results of single-nucleotide addition
experiments were fitted to a single exponent described by equation:
P = A*(1 ꢀ exp(ꢀkt)), where P is the fraction of the 9-nt product, A is
the fraction of the product at tf∞, and k and t are the rate and time of
the transcription reaction. The fitting was performed by nonlinear
regression using the program SigmaPlot (Version 8.0).
Bacterial Strains and Media. Streptococcus salivarius Andrewes
and Horder was obtained from ATCC (9758) and was grown in Bacto
Tryptic Soy Broth Medium at 37 °C.
Antibacterial Testing. The antibiotic activity of all compounds
was analyzed via disk diffusion assay. Several drops of S. salivarius culture
were spread in a Petri dish containing Difco Tryptic Soy Agar. Paper
disks of 5 mm diameter were used. Each disk was treated with a solution
of each individual compound dissolved in ethanol (2ꢀ15 μL depending
on the sample). Following evaporation of the ethanol, the disks were
placed on top of agar. Plates were incubated at 37 °C for 24 h.
(6) (a) DeBoer, C.; Dietz, A.; Silver, W. S.; Savage, G. M. Antibiot.
Annu. 1956, 886–892. (b) Eble, T. E.; Large, C. M.; DeVries, W. H.;
Crum, G. F.; Shell, J. W. Antibiot. Annu. 1956, 893–896. (c) Lewis, C.;
Wilkins, J. R.; Schwartz, D. F.; Nikitas, C. T. Antibiot. Annu. 1956,
897–902.
(7) (a) Siddhikol, C.; Erbstoeszer, J. W.; Weiblum, B. J. Bacteriol.
1969, 99, 151–155. (a) Schleif, R. Nature 1969, 223, 1068–1069. (c)
McClure, W. R. J. Biol. Chem. 1980, 255, 1610–1616.
(8) (a) Temiakov, D.; Zenkin, N.; Vassylyeva, M. N.; Perederina, A.;
Tahirov, T. H.; Kashkina, E.; Savkina, M.; Zorov, S.; Nikiforov, V.;
Igarashi, N.; Matsugaki, N.; Wakatsuki, S.; Severinov, K.; Vassylyev,
D. G. Mol. Cell 2005, 19, 655–666. (b) Tuske, S.; Sarafianos, S. G.;
Wang, X.; Hudson, B.; Sineva, E.; Mukhopadhyay, J.; Birktoft, J. J.;
Leroy, O.; Ismail, S.; Clark, A. D.; Dharia, C.; Napoli, A.; Laptenko, O.;
Lee, J.; Borukhov, S.; Ebright, R. H.; Arnold, E. Cell 2005, 122, 541–552.
(c) Vassylyev, D. G.; Vassylyeva, M. N.; Zhang, J.; Palangat, M.;
Artsimovitch, I.; Landick, R. Nature 2007, 448, 163–169. (d) Miropols-
kaya, N.; Artsimovitch, I.; Klimasauskas, S.; Nikiforov, V.; Kulbachinskiy,
A. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 18942–18947.
(9) (a) Campbell, E. A.; Korzheva, N.; Mustaev, A.; Murakami, K.;
Nair, S.; Goldfarb, A.; Darst, S. A. Cell 2001, 104, 901–912. (b)
Artsimovitch, I.; Vassylyeva, M. N.; Svetlov, D.; Svetlov, V.; Perederina,
A.; Igarashi, N.; Matsugaki, N.; Wakatsuki, S.; Tahirov, T. H.; Vassylyev,
D. G. Cell 2005, 122, 351–363.
(10) (a) Xu, M.; Zhou, Y. N.; Goldstein, B. P.; Jin, D. J. J. Bacteriol.
2005, 187, 2783–2792. (b) O’Neill, A.; Oliva, B.; Storey, C.; Hoyle, A.;
Fishwick, C.; Chopra, I. Antimicrob. Agents Chemother. 2000, 44, 3163–
3166. (c) Heisler, L. M.; Suzuki, H.; Landick, R.; Gross, C. A. J. Biol.
Chem. 1993, 268, 25369–25375. (d) Severinov, K.; Markov, D.;
Severinova, E.; Nikiforov, V.; Landrick, R.; Darst, S. A.; Goldfarb, A.
J. Biol. Chem. 1995, 270, 23926–23929. (e) Yang, X.; Price, C. W. J. Biol.
Chem. 1995, 270, 23930–23933.
’ ASSOCIATED CONTENT
(11) (a) Rinehart, K. L., Jr.; Beck, J. R.; Epstein, W. W.; Spicer, L. D.
J. Am. Chem. Soc. 1963, 85, 4035–4037. (b) Rinehart, K. L., Jr.; Borders,
D. B. J. Am. Chem. Soc. 1963, 85, 4037–4038. (c) Rinehart, K. L., Jr.;
Beck, J. R.; Borders, D. B.; Kinstle, T. H.; Krauss, D. J. Am. Chem. Soc.
1963, 85, 4038–4039.
(12) Duchamp, D. J.; Branfman, A. R.; Button, A. C.; Rinehart, K. L.,
Jr. J. Am. Chem. Soc. 1973, 95, 4077–4078.
(13) Olano Gꢀomez, C. C.; Pꢀerez, M.; Palomino, M.; Pineda-Lucena,
A.; Carbajo, R. J.; Bra~na, A. B.; Mꢀendez, C.; Salas, J. A. Chem. Biol. 2009,
16, 1031–1044.
(14) Horna, D. H.; Gꢀomez, C.; Olano, C.; Palomino-Sch€atzlein, M.;
Pineda-Lucena, A.; Carbajo, R. J.; Bra~na, A. F.; Mꢀendez, C.; Salas, J. A.
J. Bacteriol. 2011, 193, 2647–2651.
(15) Brill, G. M.; McAlpine, J. B.; Whittern, D. J. Antibiot. 1988,
41, 36–44.
S
Supporting Information. Detailed experimental proce-
b
dures as well as analytical and spectral characterization data for
all new compounds. This material is available free of charge via
the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
skozmin@uchicago.edu
’ ACKNOWLEDGMENT
This work was supported in part by National Institutes of
Health grants GM64530 (to K.S.) and AI090558 (to K.K.) and
Molecular and Cell Biology Program grant from the Russian
(16) (a) Meyer, C. E. J. Antibiot. 1971, 24, 558–560. (b) Hagenmaier,
H.; Jaschke, K. H.; Santo, L.; Scheer, M.; Zahner, H. Arch. Microbiol. 1976,
12183
dx.doi.org/10.1021/ja2041964 |J. Am. Chem. Soc. 2011, 133, 12172–12184

Journal of the American Chemical Society
ARTICLE
109, 65–74. (c) Carlson, J. C.; Li, S.; Burr, D. A.; Sherman, D. H. J. Nat. Prod.
2009, 72, 2076–2079.
(17) (a) Tsukiura, H.; Tomita, K.; Hanada, M.; Kobaru, S.; Tsunakawa,
M.; Fujisawa, K.; Kawaguchi, H. J. Antibiot. 1980, 33, 157–165.
(b) Tsunakawa, M.; Toda, S.; Okita, T.; Hanada, M.; Nakagawa, S.;
Tsukiura, H.; Naito, T.; Kawaguchi, H. J. Antibiot. 1980, 33, 166–172.
(c) Tomita, K.; Hoshino, Y.; Sasahira, T.; Hasegawa, K.; Akiyama,
M.; Tsukiura, H.; Kawaguchi, H. J. Antibiot. 1980, 33, 1491–1501.
(d) Bansal, M. B.; Dhawan, V. K.; Thadepalli, H. Chemotherapy 1982,
28, 200–203.
(18) (a) Gause, G. F.; Sveshnikova, M. A.; Ukholina, R. S.; Komar-
ova, G. N.; Bazhanov, V. S. Antibiotiki 1977, 22, 483–486. (b) Brazhni-
kova, M. G.; Konstantinova, N. V.; Potapova, N. P.; Tolstykh, I. V.;
Rubasheva, L. M.; Rozynov, B. V.; Horvath, G. Bioorg. Khim. 1981,
7, 298.
(19) Karwowski, J. P.; Jackson, M.; Theriault, R. J.; Barlow, G. J.; Coen,
L.; Hensey, D. M.; Humphrey, P. E. J. Antibiot. 1992, 45, 1125–1132.
(20) (a) Schlessinger, R. H.; Bebernitz, G. R.; Lin, P.; Poss, A. J. J. Am.
Chem. Soc. 1985, 107, 1777–1778. (b) DeShong, P.; Ramesh, S.; Elango, V.;
Perez, J. J. J. Am. Chem. Soc. 1985, 107, 5219–5224. (c) Boeckman, R. K.;
Starrett, J. E.; Nickell, D. G.; Sum, P. E. J. Am. Chem. Soc. 1986,
108, 5549–5559. (d) Neukom, C.; Richardson, D. P.; Myerson, J. H.;
Bartlett, P. A. J. Am. Chem. Soc. 1986, 108, 5559–5568. (e) Shimshock, S. J.;
Waltermire, R. E.; DeShong, P. J. Am. Chem. Soc. 1991, 113, 8791–8796.
(21) (a) Boeckman, R. K., Jr.; Potenza, J. C.; Enholm, E. J. J. Org.
Chem. 1987, 52, 469–472. (b) Schlessinger, R. H.; Graves, D. D.
Tetrahedron Lett. 1987, 28, 4385–4388. (c) Ireland, R. E.; Smith,
M. G. J. Am. Chem. Soc. 1988, 110, 854–860.
(22) Iwata, Y.; Maekawara, N.; Tanino, K.; Miyashita, M. Angew.
Chem., Int. Ed. 2005, 44, 1532–1536.
(23) Pronin, S. V.; Kozmin, S. A. J. Am. Chem. Soc. 2010,
132, 14394–14396.
(24) Cartwright, D.; Lee, V. J.; Rinehart, K. L. J. Am. Chem. Soc. 1978,
100, 4237–4239.
(25) Evans, D. A.; Britton, T. C. J. Am. Chem. Soc. 1987,
109, 6881–6883.
(26) Solsona, J. G.; Nebot, J.; Romea, P.; Urpi, F. J. Org. Chem. 2005,
70, 6533–6536.
(27) Evans, D. A.; Hoveyda, A. J. Am. Chem. Soc. 1990, 112, 6447–
6449.
(28) Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem. Rev.
1994, 94, 2483–2547.
(29) Garber, S. V.; Kingsbury, J. S.; Gray, B. L.; Hoveyda, A. H. J. Am.
Chem. Soc. 2000, 122, 8168–8179.
(30) Schlessinger, R. H.; Graves, D. Tetrahedron Lett. 1987, 28,
4381–4384.
(31) (a) Ley, S. V.; Woodward, P. R. Tetrahedron Lett. 1987,
28, 345–346. (b) Ley, S. V.; Smith, S. C.; Woodward, P. R. Tetrahedron
1992, 48, 1145–1174.
(32) This includes comparison of 500 MHz 1H NMR, 125 MHz 13
C
NMR, HRMS, and optical rotation. For the original detailed 13C NMR
analysis, see: Lee, V. J.; Rinehart, K. L. J. Antibiot. 1980, 33, 408–415.
(33) (a) Reusser, F. Infect. Immun. 1970, 2, 77–81. (b) Reusser, F.
Antimicrob. Agents Chemother. 1976, 10, 618–622.
(34) Kuznedelov, K.; Minakhin, L.; Severinov, K. Methods Enzymol.
2003, 370, 94–108.
(35) Korzheva, N.; Mustaev, A.; Kozlov, M.; Malhotra, A.; Nikiforov,
V.; Goldfarb, A.; Darst, S. A. Science 2000, 289, 619–625.
(36) Kuznedelov, K.; Semenova, E.; Knappe, T. A.; Mukhamedyarov,
D.; Srivastava, A.; Chatterjee, S.; Ebright, R. H.; Marahiel, M. A.; Severinov,
K. J. Mol. Biol. 2011, Mar 15 [Epub ahead of print].
(37) Evans, D. A.; Ennis, M. D.; Mathre, D. J. J. Am. Chem. Soc. 1982,
104, 1737–1739.
(38) Sherman, J. M.; Niven, C. F.; Smiley, K. L. J. Bacteriol. 1943,
45, 249–263.
12184
dx.doi.org/10.1021/ja2041964 |J. Am. Chem. Soc. 2011, 133, 12172–12184