Potent inhibitors of pro-inflammatory cytokine production produced by a marine-derived bacterium

Article abstract of DOI:10.1021/jm801110j

Cytokines produced through the antigen presenting cell (APC)-T-cell interaction play a key role in the activation of the allergic asthmatic response. Evaluating small molecules that inhibit the production of these pro-inflammatory proteins is therefore im

Full text of DOI:10.1021/jm801110j

J. Med. Chem. 2009, 52, 2317–2327
2317
Potent Inhibitors of Pro-Inflammatory Cytokine Production Produced by a Marine-Derived
Bacterium
Wendy K. Strangman,† Hak Cheol Kwon,† David Broide,‡ Paul R. Jensen,† and William Fenical*,†,§
Center for Marine Biotechnology and Biomedicine, Scripps Institution of Oceanography, UniVersity of California, San Diego, La Jolla,
California 92093-0204, Skaggs School of Pharmacy and Pharmaceutical Sciences, UniVersity of California, San Diego, La Jolla,
California 92093-0204, and Department of Medicine, UniVersity of California, San Diego, La Jolla, California 92093-0635
ReceiVed September 6, 2008
Cytokines produced through the antigen presenting cell (APC)-T-cell interaction play a key role in the
activation of the allergic asthmatic response. Evaluating small molecules that inhibit the production of these
pro-inflammatory proteins is therefore important for the discovery of novel chemical structures with potential
antiasthma activity. We adapted a mouse splenocyte cytokine assay to screen a library of 2,500 marine
microbial extracts for their ability to inhibit TH2 cytokine release and identified potent activity in a marine-
derived strain CNQ431, identified as a Streptomyces species. Bioactivity guided fractionation of the organic
extract of this strain led to the isolation of ten new 9-membered bis-lactones, splenocins A-J (1-10). The
new compounds display potent biological activities, comparable to that of the corticosteroid dexamethasone,
with IC50 values from 2 to 50 nM in the splenocyte cytokine assay. This study provides the foundation for
the optimization of these potent anti-inflammatory compounds for development in the treatment of asthma.
Introduction
there have been very few novel drug classes developed. In fact,
antileukotriene agents and an anti-IgE monoclonal antibody are
the only new asthma drug classes to be released in the past 20
years.6 As our knowledge of asthma-related molecular mech-
anisms increases, novel pathways can be targeted for drug
development. Assessing natural product extracts presents an
excellent way to probe biological systems with a wide variety
of molecules in a relatively high-throughput manner.7 By
focusing even further on previously unexplored, marine micro-
bial secondary metabolites, the chances of discovering a novel
small molecule inhibitor are further increased.8 In the current
study, 2,500 marine-derived microbial extracts were screened
in a mouse splenocyte assay, which models the APC-TH2 cell
population critically involved in the pathogenesis of allergic
inflammation related to asthma.
The whole culture extract of a marine-derived actinomycete
strain, culture CNQ431, identified as a Streptomyces sp.,
displayed significant biological activity in the mouse splenocyte
assay (IC80 ) 21 µg/mL for the crude extract) and was therefore
targeted for further investigation. This bacterial strain was
isolated from oceanic bottom sediments at 30 m depth roughly
one mile off the coast of the Scripps Institution of Oceanography
in La Jolla, California. Bioactivity guided fractionation and final
purification of the cytokine inhibitors from this strain led to
the discovery of a new set of bioactive molecules, the splenocins
A-J (1-10, Figure 1). These molecules display potent sup-
pression of cytokine production in ranges from low micromolar
to low nanomolar and exhibit minimal mammalian cell cyto-
toxicity, which is only evident at levels in the low molar to
high micromolar range. These data indicate a therapeutic ratio
in this assay of over 200 with some metabolites, and over 500
in the case of compound 2. Further biological investigation
revealed that all of these molecules inhibited not only the
production of TH2 cytokines IL-5 and IL-13 but also the
production of the dendritic cell-associated cytokines IL-1 and
TNF-R, indicating immunosuppressive effects on both the APCs
(i.e., dendritic cells) and the TH2 cells.
Asthma is a disease affecting over 15 million people in the
United States. The disease is characterized by airway obstruc-
tion, bronchial inflammation, and bronchial smooth muscle
hyper-reactivity.1 Symptoms include characteristic wheezing,
shortness of breath or difficulty with breathing, and in the USA,
approximately 5,000 patients die from asthma each year. Chronic
allergic asthma is a complex disease mediated by a number of
cellular processes. In chronic asthma, one of the early cellular
events in the pathogenesis of the disease is the stimulation of
TH2 lymphocytes by antigen presenting cells (APCs) such as
dendritic cells.2 Upon activation, TH2 cellsa secrete a number
of cytokines such as interleukins-4, -5, -9, and -13 that activate
cellular and molecular cascades, eventually leading to the
asthmatic phenotype.3 Specifically, the TH2 cytokine interleu-
kin-5 (IL-5) has been shown to be an activator of eosinophils,
bone marrow-derived leukocytes responsible in part for persis-
tent inflammation and airway remodeling seen in patients with
chronic asthma.3 Similarly, another TH2 cytokine, interleukin
13 (IL-13), has been shown to play a critical role in murine
models of asthma through its actions on airway smooth muscle
and epithelial cells.4 While inhibition of a single TH2 cytokine
may provide little benefit in the treatment of asthma, compounds
that suppress the production of multiple cytokines could possess
significantly greater therapeutic potential.5
Despite advances in understanding of the cellular and
molecular mechanisms involved in the pathogenesis of asthma,
* Corresponding author. E-mail: wfenical@ucsd.edu. Tel: (858) 534-
2133. Fax: (858) 534-1318.
†
Center for Marine Biotechnology and Biomedicine, Scripps Institution
of Oceanography.
‡
Department of Medicine.
Skaggs School of Pharmacy and Pharmaceutical Sciences.
§
a Abbreviations: TH2 cells, helper T lymphocytes; IgE, immunoglobulin
E; TNF-R, tumor necrosis factor alpha; COSY, two-dimensional correlation
spectroscopy; HMBC, heteronuclear multiple bond correlation (spectros-
copy); HRMS, high resolution mass spectroscopy; HR ESI-TOF MS, high
resolution electrospray time-of-flight mass spectroscopy; NOE, nuclear
Overhauser enhancement; DMAP, dimethylaminopyridine; CD, circular
dichroism; OVA, ova albumin (chicken egg white); MTPA-Cl, methox-
y(trifluoromethyl)phenylacetyl chloride.
Compounds 1-10 are characterized by a central nine-
membered cyclic bis-lactone core. An N-formylamino salicylic
10.1021/jm801110j CCC: $40.75
2009 American Chemical Society
Published on Web 03/26/2009

2318 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8
Strangman et al.
Figure 1. Structures of 1-10 and known antimycins 11-17.
acid moiety is positioned at C-3′ via an amide bond in all
members. This fundamental structure, together with the cyclic
bis-lactone core, is also seen in molecules of the antimycin
class.9-11 The splenocins are further characterized and dif-
ferentiated from the antimycins by the presence of benzyl groups
positioned at C-7 (1-3, 10) and benzoyl groups at C-8 (4-8)
of the bis-lactone ring. Compound 9 contains aromatic func-
tionalities at both the C-7 and C-8 positions. In addition to the
benzyl group at C-7, compound 10 contains a hydroxyl group
at C-8 instead of the ester functionality seen in splenocins A-I.
This hydroxyl functionality can also be seen in other previously
described 8-hydroxy antimycins which contain alkyl chains at
the C-7 position and differ from each other by the number of
carbons in the chain.12 Representative members of this group,
which include deisovalerylblastmycin (C4 alkyl chain)13 (11),
the urauchimycins (C5)14 (12), and kitamycins (C6)15 (13), were
also isolated from the culture broth of Streptomyces strain
CNQ431.
Compounds 1-9 displayed low nanomolar activity in the
suppression of cytokine production by OVA stimulated sple-
nocytes, similar to that we have observed with the corticosteroid
drug dexamethasone. Currently, corticosteroids are the most
potent anti-inflammatory drugs on the market. Compound 10
exhibited low micromolar activity in the splenocyte assay. In
this paper, we describe the isolation, structure elucidation, and
structure activity relationships of the splenocins in the in Vitro
mouse splenocyte assay.
Figure 2. Substructures with key HMBC correlations used to establish
the planar structure of compound 1.
NMR correlations to both H-4 (δH 5.60) and a secondary amide
proton (δH 6.98). Proton H-4 showed an additional correlation
to a methyl doublet (δH 1.16). Similarly, a second spin system
was assembled starting with oxygenated methine proton H-9
(δH 5.01), which showed 1H-1H COSY NMR correlations to a
methyl doublet (δH 1.32) and H-8 (δH 5.19). Proton H-8 showed
a further correlation to H-7 (δH 2.90), which showed correlations
to two methylene protons H1-1′′a and H1′′b (δH 2.97 and 2.71)
in the 1H-1H COSY NMR spectrum. HMBC NMR correlations
from H-4 and H-7 to carbonyl carbon C-6 (δC 172.8) and protons
H-3 and H-9 to carbon C-2 (δC 170.0) allowed the fragments
to be assembled into a nine-membered bis-lactone. The orienta-
tions of the ester bonds within the bis-lactone ring were
determined by 1H and 13C chemical shift analysis, and in
particular the proton shifts of H-4 and H-9.
Results
Structure Assignments for the Splenocins. Compound 1
was isolated as an optically active amorphous white powder,
which displayed strong IR bands at 3381 and 1748 cm-1
,
indicating the presence of hydroxyl and carbonyl functional
groups. A molecular formula of C26H28N2O9 was assigned based
on interpretation of HR ESI-TOF MS data (obsd [M + Na]+ at
m/z 535.1683). Together with the molecular formula and
information derived from interpretation of spectral data, several
substructures were assembled based on analyses of 1D and 2D
NMR experiments (Figure 2; Table 1).
A second substructure (Figure 2B) was assembled starting
with a downfield resonance (δH 8.48) that correlated to the
carbon at δC 159.1. This shift was explained by a 1H-1H COSY
The first substructure (Figure 2A) was assembled starting with
the methine proton H-3 (δH 5.24), which showed 1H-1H COSY

Splenocins A-J, New Anti-Inflammatory Compounds
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8 2319
Table 1. NMR Spectral Data for 1-3 in CDCl
3
compound 1
compound 2
compound 3
a
b
a
b
a
b
position
δC
δH (J in Hz)
δC
δH (J in Hz)
δC
δH (J in Hz)
2
3
4
6
7
8
9
170.0
53.8
70.6
172.8
51.6
75.6
170.1
53.4
70.9
171.9
51.9
75.0
170.0
53.5
70.9
172.4
52.0
75.1
5.24 dd (8.0, 8.0)
5.60 dq (8.0, 6.5)
5.26 dd (7.5, 7.5)
5.60 dq (7.5, 7.0)
5.25 dd (7.5, 7.5)
5.59 dq (7.5, 7.0)
2.90 ddd (10.5, 10.5, 3.5)
5.19 t (10.5)
5.01 m
2.87 ddd (11.5, 10.5, 3.0)
5.20 t (10.5)
5.01 m
2.87 ddd (10.5, 10.5, 3.5)
5.21 t (10.5)
5.00 m
75.0
74.7
74.8
4-Me
9-Me
15.0
18.0
1.16 d (6.5)
1.32 d (6.5)
14.7
17.8
1.15 d (7.0)
1.31 d (6.5)
14.7
17.9
1.14 br d (7.0)
1.31 d (6.5)
1′
2′
3′
4′
5′
6′
112.5
150.6
127.7
124.9
119.0
120.2
169.5
112.6
150.7
127.4
124.8
118.8
120.3
169.3
112.5
150.6
127.4
124.8
119.0
120.0
169.3
8.52 d (7.5)
6.89 t (7.5)
7.20 d (7.5)
8.52 d (7.5)
6.89 t (7.5)
7.18 d (7.5)
8.51 d (8.0)
6.88 t (7.5)
7.18 d (8.0)
1′-CONH
1′-CONH
2′-OH
3′-NHCHO
3′-NHCHO
1′′
6.98 d (8.0)
12.59 br s
7.87 br s
8.48 s
2.97 dd (13.5, 10.5)
2.71 dd (13.5, 3.5)
7.03 d (7.5)
12.62 br s
7.94 s
8.47 s
2.97 dd (13.5, 11.5)
2.68 dd (13.5, 3.5)
6.98 d (8.0)
12.60 br s
7.89 br s
8.49 s
2.97 dd (13.0, 10.5)
2.70 dd (13.5, 3.5)
159.5
34.7
159.1
34.5
158.9
34.5
2′′
3′′
4′′
5′′
6′′
7′′
1′′′
2′′′
3′′′
138.5
128.8
128.6
126.6
128.6
128.8
170.2
20.9
137.9
128.7
128.5
126.6
128.5
128.7
175.7
34.1
137.9
128.7
128.6
126.6
128.6
128.7
175.3
41.3
7.12 d (8.0)
7.24 t (8.0)
7.18 t (8.0)
7.24 t (8.0)
7.12 d (8.0)
7.11 d (8.0)
7.24 t (8.0)
7.17 t (8.0)
7.24 t (8.0)
7.11 d (8.0)
7.11 d (8.0)
7.24 t (8.0)
7.18 t (8.0)
7.24 t (8.0)
7.11 d (8.0)
2.04 s
2.59 m
1.22 d (7.0)
2.43 m
1.75 m
18.9
26.5
1.53 m
4′′′
5′′′
11.8
16.7
0.95 t (7.5)
1.20 d (7.5)
b
a Assignment by 13C, HSQC, and HMBC NMR methods at 125 MHz. Assignment by H,gHSQC, and gHMBC NMR methods at 500 MHz.
1
NMR correlation from the formamide proton to an amide proton
(δH 7.87), which indicated the presence of an N-formyl amino
group. Next, 1H-1H COSY NMR correlations between aromatic
protons H-4′ (δH 8.52), H-5′ (δH 6.89), and H-6′ (δH 7.20) and
HMBC correlations from these aromatic protons to quaternary
carbons C-1′ (δC 112.5) and C-3′ (δC 127.7) and hydroxyl-
substituted C-2′ (δC 150.6) allowed us to assemble a trisubsti-
tuted aromatic ring. An HMBC NMR correlation from the amide
proton of the N-formyl amino group to the aromatic quaternary
carbon C-3′ allowed the two pieces to be connected, thereby
forming the N-formyl-amino salicylic acid moiety as a second
substructure. Finally, interpretation of 1H-1H COSY correlations
allowed a monosubstituted benzene ring to be defined (Figure
2C). In addition, the methyl singlet, H3-2′′′ (δH 2.04, δC 20.9),
showed an HMBC correlation to carbonyl carbon C-1′′′ (δC
170.2) defining a single acetate unit (Figure 2D).
The connectivity between the substructures was achieved by
interpretation of HMBC NMR correlation data (Figure 2E). The
aromatic proton H-6′ showed a C-H long-range correlation to
a carbonyl carbon at δC 169.5. Both H-3 and the secondary
amino proton adjacent to H-3 in the cyclic bis-lactone also
showed C-H long-range correlations to this carbonyl, thereby
allowing the N-formyl amino salicylic acid substructure to be
connected to the bis-lactone ring via an amide bridge between
C-1′ and C-3. This bis-lactone core substructure is the same as
that seen in the antimycins, which are characterized by the
presence of a C-7 alkyl and C-8 acyl unsaturated hydrocarbon
chains9-11 and in the 8-hydroxy antimycins, deisovalerylblast-
mycin,13 urauchimycins,14 and kitamycins.15 Additional C-H
long-range correlations allowed the benzyl substituent to be
connected to C-7 of the bis-lactone. Further, an HMBC NMR
correlation from H-8 to the C-1′′′ carbonyl carbon allowed the
acetate unit to be connected to C-8, thereby completing the
planar structure for 1 (Figure 2E).
The structures of 2 and 3 are very similar to that of 1 in that
they contain a benzyl group positioned at C-7 (Table 1).
However, they differ from 1 in the length of the acyl chain at
C-8. The length and branching of the side-chains in 2 and 3
were defined by a combination of HRMS and proton NMR
analysis.
Similar to compounds 1-3, compounds 4-8 (Table 2) also
possess ester functionalities positioned at C-8. However, for this
subset of compounds, HMBC analysis showed correlations from
the C-8 ester carbonyl to protons on a second aromatic ring
H-3′′′ (δH 8.04) and H-7′′′ (δH 8.04), suggesting that these
molecules contained a benzoyl group at C-8 rather than an acyl
chain (Table 2). In further contrast to 1-3, compounds 4-8
possess alkyl chains originating at C-7 rather than the corre-
sponding benzyl moieties of 1-3. The C-7 alkyl chains of 4-8
were assembled based on analysis of HRMS and proton NMR
spectroscopic data.
Compound 9 is unique in that it contains aromatic groups at
both C-7 and C-8 (Table 3). Similar to 1-3, the benzyl group
at C-7 in 9 was connected to the bis-lactone by the observation
of COSY NMR correlations from H-7 to H-1′′a and H-1′′b, and
by analysis of HMBC NMR correlations from H-7 to the
quaternary aromatic carbon C-2′′ and from H-3′′ to C-1′′. As
was observed in compounds 4-8, HMBC NMR correlations
were observed from both H-8 and H-3′′′ to the C-1′′′ ester
carbon (δC 165.3).
Compound 10 contains the same benzyl functionality at C-7
as in 1 - 3, and 9 (Table 3). However, unlike any of the other

2320 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8
Strangman et al.
Table 2. NMR Spectral Data for 4-8 in CDCl3
compound 4
compound 5
compound 6
a
b
a
b
a
b
position
δC
δH (J in Hz)
δC
δH (J in Hz)
δC
δH (J in Hz)
2
3
4
6
7
8
9
170.5
53.9
71.2
172.6
52.0
76.4
75.0
15.2
18.0
170.1
53.7
71.0
172.9
50.2
76.3
75.0
15.0
17.9
170.1
53.9
71.3
172.9
50.2
76.5
75.2
15.0
17.9
5.32 dd (8.0, 7.5)
5.76 dq (7.5, 7.0)
5.30 dd (8.0, 7.5)
5.76 dq (7.5, 6.5)
5.30 dd (8.0, 7.0)
5.75 dq (7.5, 7.0)
2.62 ddd (10.0, 10.0, 3.0)
5.33 t (10.0)
2.68 ddd (10.0, 10.0, 3.0)
5.32 t (10.0)
2.65 ddd (10.0, 10.0, 3.0)
5.32 t (10.0)
5.15 m
5.14 m
5.14 m
4-Me
9-Me
1.32d d (7.0)
1.33 dd (6.5)
1.33d d (7.0)
1.33d d (6.5)
1.32 dd (6.5)
1.33d d (7.0)
1′
2′
3′
4′
5′
6′
112.7
151.0
127.0
124.8
119.2
120.3
169.0
112.5
150.6
127.4
124.8
120.1
120.8
169.4
112.5
150.6
127.4
124.8
119.4
120.4
169.4
8.54 d (8.0)
6.91 t (8.0)
7.23 mc
8.54 d (8.0)
6.91 t (8.0)
7.23 mc
8.53 d (8.0)
6.90 t (8.0)
7.23 mc
1′-CONH
1′-CONH
2′-OH
3′-NHCHO
3′-NHCHO
1′′a
7.08 d (8.0)
12.60 s
7.91 br s
8.49 s
7.07 d (8.0)
12.60 s
7.92 s
8.49 s
1.76 m
707 d (8.0)
12.60 s
7.92 s
159.0
21.8
158.9
28.2
158.9
28.2
8.49 s
1.75 m
1.76 m
1.40 m
1.08 m
1′′b
2′′
1.50 mc
0.86 t (8.0)
1.40 m
1.15 m
11.5
29.3
36.5
1.24 m
3′′
4′′
5′′
22.3
13.8
1.23 m
0.80 t (7.0)
26.5
22.3
22.3
1.45 m
0.77d d (7.0)
0.78d d (7.0)
6′′
7′′
1′′′
2′′′
3′′′
4′′′
5′′′
6′′′
7′′′
165.5
129.0
130.0
128.7
133.8
128.7
130.0
165.2
129.0
129.9
128.7
133.7
128.7
129.9
1
165.2
129.0
129.9
128.7
133.9
128.7
129.9
8.04 d (8.0)
7.47 t (8.0)
7.61 t (8.0)
7.47 t (8.0)
8.04 d (8.0)
b
8.05 d (8.0)
7.47 t (8.0)
7.61 t (8.0)
7.47 t (8.0)
8.05 d (8.0)
8.04 d (7.5)
7.47 t (7.5)
7.63 t (7.5)
7.47 t (7.5)
8.0 d (7.5)
a Assignment by 1D 13C methods at 125 MHz. Assignment by H, gHSQC, and gHMBC NMR methods at 500 MHz. c The multiplicity can not be
assigned due to solvent peak overlap. d Signals are interchangeable.
splenocins, the proton at C-8 (δH 3.73/ δC 77.3) in 10 showed
1H-1H COSY NMR correlation to a hydroxyl proton,
secondary amide proton connected to C-3 and a very weak
correlation to H-3, while H-3 and H-4 showed strong NOESY
correlations to each other. Additionally, there was no apparent
NOE correlation between H-4 and the secondary amide proton.
Taken together with the coupling constant data, this indicated
that H-3 and H-4 were in the syn configuration although not
quite eclipsed. (Figure 3).
On the opposite side of the ring, the coupling constants
between H-7 and H-8, and between H-8 and H-9, are between
9 and 10.5 Hz, which is strong evidence for the assignment of
an anti configuration. These coupling constants are also
consistent with those observed in the antimycin bis-lactone ring
(Table 4). NOE correlations were analyzed to further confirm
this configuration. The methyl protons at C-9 showed strong
NOE NMR correlations to H-8. Proton H-9 showed a strong
NOE correlation to H-7, and H-7 showed a strong NOE
correlation to a proton in the benzyl ring at C-7, indicating that
they are in close proximity on the bottom face of the molecule.
All splenocins showed similar NOE NMR correlations between
protons in the bis-lactone ring and displayed similar coupling
constants, thereby indicating that they share identical relative
configurations (Table 4).
a
indicating the presence of a secondary alcohol at C-8 rather
than an ester. This C-8 hydroxyl group is also shared by the
known compounds 11,13 12,14 and 13,15 which contain C4, C5,
and C6 alkyl chains, respectively, at C-7 of the bis-lactone ring.
Additionally, we isolated compound 14, a new member of the
8-hydroxy antimycin subset of molecules. Analysis of 1D and
2D NMR spectral data, combined with mass spectral fragmenta-
tion information, showed that 14 contains a C-4 branched C7
alkyl chain originating from C-7 (Supporting Information).
Also isolated from the extract of Streptomyces sp., strain
CNQ431, were three known derivatives, antimycins A6a, A4a
and A7a (14, 15, and 16). As antimycin structures differ from
each other by the length and branching of their alkyl and acyl
chains, their structures were determined through interpretation
9-11
1
of HRMS and by comparison to reported H NMR data.
Relative Configurations of the Splenocins. The relative
configurations of the splenocins were assigned by interpretation
of proton NMR coupling constants and by analysis of NOESY
NMR data (Table 4 and Figure 3, respectively). On the side of
the bis-lactone containing the N-formylamino salicylic acid
group, protons H-3 and H-4 were coupled with coupling
constants of 7.0 and 7.5 Hz (Table 4). While this value is
suggestive of a syn configuration, the coupling constant is not
small enough to be definitive.10 To further support this assign-
ment, NOE NMR correlations were also analyzed. The methyl
protons at C-4 showed strong NOE correlations with the
For those compounds possessing a benzyl group at C-7 (1-3,
9, 10), a strong NOE NMR correlation was observed between
H-8 and H-1′′a and a weaker NOE correlation between H-8
and H-1′′b, which, when combined with the NOE correlation
between H-7 and the benzyl at C-7, indicated the spatial
orientation for the benzyl functionality (Figure 3). In compound

Splenocins A-J, New Anti-Inflammatory Compounds
Table 3. NMR Spectral Data for 9 and 10 in CDCl3
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8 2321
ring. Additional NMR-based evidence for this connectivity for
10, and also for the other benzyl-containing compounds 1-3
and 9, comes from proton chemical shift analysis of the methyl
group at C-4. In the splenocins that contain a benzyl group at
C-7, the methyl protons at C-4 are shifted upfield roughly 0.2
ppm (from 1.32 ppm to 1.15 ppm) relative to those of the
nonbenzylated compounds 4-8 (Tables 1, 2, and 3). This upfield
shift, combined with the NOE correlation seen in 10, suggests
that the methyl protons at C-4 are being shielded by the adjacent
benzyl group at C-7.
compound 9
compound 10
a
b
a
b
position
δC
δH (J in Hz)
δC
δH (J in Hz)
2
3
4
6
7
8
9
170.1
170.1
53.6 5.29 dd (8.0, 7.5)
71.0 5.64 dq (7.5, 7.0)
172.0
52.0 3.06 dt (10.0, 10.0) 54.0 2.69 ddd (10.5, 10.0, 3.5)
76.1 5.45 t (10.0)
74.9 5.12 m
53.5 5.21 dd (7.5, 7.0)
70.7 5.58 dq (7.5, 7.0)
173.0
77.3 3.73 dt (10.0, 10.0)
76.9 4.87 dq (10.0, 10.0)
1.62 br s
14.7 1.15 d (7.0)
18.4 1.47 d (6.0)
112.5
8-OH
4-Me
9-Me
1′
14.8 1.18 d (6.5)
18.0 1.37 d (6.5)
112.5
Absolute Configurations of the Splenocins. The availability
of the secondary alcohol at C-8 made compound 10 an
appropriate candidate for absolute configuration determination.
The absolute configuration of C-8 in 10 was thus defined by
application of the modified Mosher NMR method.16 Compound
10 was treated with (R)-(-)-R-methoxy-R-(trifluoromethyl)phe-
nylacetyl chloride ((R)-MTPA-Cl) and (S)-(+)-MTPA-Cl in
separate experiments. Initially, and quite unexpectedly, the
reaction yielded mono-S-Mosher and mono-R-Mosher amides,
respectively, at the 3′-aminophenyl position. The creation of
this Mosher amide can be explained through an acyl transfer
mechanism in which the formyl group is first displaced by
DMAP forming an aniline (Scheme 1.). This group then attacks
the Mosher-DMAP conjugate resulting in the Mosher-amide.
While this reaction is unprecedented under these conditions,
similar deformylation reactions of aromatic N-formyl groups
have been shown to occur upon heating, or under strongly acidic
or strongly basic conditions.17-19 It would be interesting to
investigate this mechanism in greater detail to determine whether
it is specific to the splenocins or whether it can be applied more
generally to other types of N-formyl containing compounds.
2′
150.6
150.6
3′
127.4
127.3
4′
5′
6′
124.7 8.52 d (8.0)
119.0 6.89 t (8.0)
120.0 7.19 t (8.0)
124.8 8.51d (8.0
119.0 6.88 t (8.0)
120.1 7.18 d (8.0)
169.3
1′-CONH 169.4
1′-CONH
2′-OH
7.01 br d (8.0)
12.6 br s
7.01 d (7.5)
12.6 s
3′-NHCHO
7.88 br s
7.87 br s
3′-NHCHO 158.9 8.48 s
159.0 8.47 s
1′′a
1′′b
2′′
34.7 3.06 dd (20.0, 10.0) 35.1 3.19 dd (13.5, 3.5)
2.78 dd (20.0, 10.0) 2.98 dd (13.5, 10.5)
137.8
138.4
3′′
4′′
5′′
6′′
128.7 7.09 d (7.5)
128.5 7.19 t (7.5)
126.6 7.13 t (7.5)
128.5 7.19 t (7.5)
128.7 7.09 d (7.5)
165.3
126.6 7.17 d (7.5)
128.6 7.26 t (7.5)
128.8 7.16 t (7.5)
128.6 7.26 t (7.5)
126.6 7.17 d (7.5)
7′′
1′′′
2′′′
3′′′
4′′′
5′′′
6′′′
7′′′
129.0
129.9 8.05 d (8.0)
128.7 7.48 t (8.0)
133.9 7.62 t (8.0)
128.7 7.48 t (8.0)
129.9 8.05 d (8.0)
Allowing the reaction to proceed longer produced the di-S-
Mosher product 10a and di-R-Mosher product 10b, respectively,
with the second Mosher ester addition at the aromatic hydroxyl
position of the N-formyl amino salicylic acid moiety (Figure
4A). Continuation of the reaction finally produced the tri-S-
Mosher product 10c and the tri-R-Mosher product 10d with the
third ester product at the desired C-8 alcohol position (Figure
4B). However, initial analysis of the tri-Mosher products gave
inconclusive stereochemical results. Specifically, subtraction of
NMR shifts of the tri-R-Mosher product from the tri-S-Mosher
product following the modified Mosher method protocol gave
a negative value (-0.01) for 9-Me and a positive value (+0.01)
for H-9 on the same side of the Mosher ester at C-8, as well as
positive values for H-7 (+0.01) and H-1′′a (+0.04) (Figure 4B).
Suspecting that the ambiguous data were due to anisotropic
effects of the phenyl groups from the Mosher products at the
3′-amine and the 2′ positions, it was reasonable to subtract the
1H NMR chemical shift differences of the di-Mosher products
10a and 10b (∆δdiS-diR) from the differences of the tri-Mosher
products 10c and 10d (∆δtriS-triR, Figure 4C). Analysis of these
chemical shift differences (∆δS(tridi)-R(tridi)) gave negative values
for the methyl protons at C-9 (-0.02) and H-9 (-0.01), and
positive values for protons at H-7 (+0.01) and H-1′′a (+0.04),
thus suggesting that the absolute configuration at C-8 is R.
a Assignment by 13C, HSQC, and HMBC NMR methods at 125 MHz.
b Assignment by 1H, gCOSY, gHSQC, and gHMBC NMR methods at 500
MHz.
Table 4. Coupling Constant Values (Hz) for Key Protons on the
bis-Lactone Ring
3JH3-H4
3JH7-H8
3JH8-H9
compounds 1-3
compounds 4-8
compound 9
compound 10
compound 17
7.5
7.5
7.5
7.5
7.5
10.5
10.0
10.0
9.0
10.5
10.0
10.0
9.0
10.0
10.0
10, a further NOE correlation was observed between a proton
on the benzyl ring at C-7 and the methyl attached to C-4 on the
opposite side of the bis-lactone ring, thereby offering evidence
relating the configurations of the two halves of the bis-lactone
Since not all splenocins displayed NOE NMR correlations
between the left and right sides of the bis-lactone ring as was
seen for compound 10, it was important to determine the
absolute configurations of C-3 and C-4 of the embedded
threonine molecule of the bis-lactone ring in order to assign
the absolute configuration for the entire molecule. Given the
syn relative configuration between H-3 and H-4 within the
constraints of the bis-lactone ring system, the D and L isomers
of threonine would give opposite configurations within the
Figure 3. Key NOESY correlations for establishing the relative
configuration of compound 10.

2322 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8
Strangman et al.
Scheme 1. Proposed Mechanism for Deformylation and Subsequent Acylation of 10
molecule (Figure 5). To determine whether the threonine moiety
was D or L, we performed an acid-catalyzed hydrolysis of 10
followed by derivatization with Marfey’s reagent and subsequent
Marfey’s analysis.20,21 Based on the retention time of the
derivatized splenocin-derived threonine, relative to derivatized
standards, we were able to determine that the threonine
component of the bis-lactone ring had the L configuration,
thereby confirming the complete absolute configuration for 10.
Similar experiments were performed with compound 6 to
confirm that both benzyl and benzoyl containing splenocins
contained L-threonine in the 3- and 4-positions of the bis-lactone
ring.
Since only compound 10 was a feasible candidate for
Mosher’s analysis, we used a series of circular dichroism (CD)
spectral comparisons to confirm the absolute configurations of
the related compounds 1-9 (Supporting Information). Com-
parisons of splenocin CD spectra allowed us to group the
splenocins into two different CD spectral categories (Figure 6).
Compounds 1-3 and 10 displayed a weak absorption at λ
220 nm. Compounds 4-9 comprise the second group,
possessing a negative Cotton effect at λ 240 nm and
corresponding positive Cotton effect at λ 225 nm (Figure 6
and Supporting Information). Based upon the strong similari-
ties in their CD spectra, it can be assumed that 4-9 all have
the same absolute configurations.
As the negative Cotton effect seen in compounds 4-9 is
observable only in compounds containing the C-8 benzoyl
group, it is likely that this Cotton effect is due to exciton
coupling of the 8-O-benzoyl group with the 3′-N-formyl amino
salicylic acid moiety (Figure 7A). In applying the rules for
exciton coupling of two chromophores, we analyzed three-
dimensional representations of 4-9 to estimate angular orienta-
tions of the electronic transition dipole moment vectors for both
the 8-O-benzoate and the 3′-N-formyl amino salicylamide
(Figure 7).22 The resulting vectors were in accordance with the
observed CD spectral data supporting our absolute configuration
assignment.
To further confirm the absolute configurations of 4-9, we
performed benzoylation reactions on compounds 9 and 10,
whose absolute configurations had already been established
chemically. These reactions created identical tribenzoylated
products (9b and 10g, Scheme 2). The CD spectra for these
two compounds proved to be identical, displaying a negative
cotton effect at 235 nm and a corresponding positive signal at
213 nm. Since the absolute configuration of 10 had already been
determined through Mosher’s and Marfey’s analyses, it can now
be unambiguously assumed that 9 also shares this configuration.
Further, as the initial set of CD experiments showed that 4-9
all have nearly identical Cotton effects, it is reasonable to assume
that they all have the same absolute configuration as compound
10.
A similar set of benzoylation reactions was also performed
with compounds 1-3 in an attempt to add an additional
chromophore on the salicylic acid (Figure 7) and create exciton
coupling with the C-7 benzyl group. These products were
obtained, but they yielded no observable change in the Cotton
effects of their CD spectra and so a different technique was
needed. Due to their close structural similarity to compound
10, we measured their optical rotation values, [R]D, to determine
if they share identical absolute configurations. Compounds 2
and 10 both showed positive optical rotations ([R]D +68.0 and
[R]D +69.6 respectively). These data, combined with the
similarities of the CD spectra, provide strong evidence that 1-3
have the same absolute configurations as compounds 4-10. The
results of these experiments confirm that compounds 1-10 all
share the same absolute configuration, 3S, 4R, 7R, 8R, and 9S,
Figure 4. The structures of derivatives of 10 from Mosher acylation
and the differentiation of H NMR chemical shift values between S
1
and R Mosher esters (∆S-R). (A) Compound 10 2′,3′-di-Mosher
derivatives and ∆S-R values. (B) Compound 10 2′,3′,7-tri-Mosher
derivatives and ∆S-R values. (C) Tri-Mosher ∆S-R - di-Mosher ∆S-R
.
Figure 5. Orientations of (A) L-threonine and (B) D-threonine as would
be seen with cis-relative configurations within the bis-lactone ring.

Splenocins A-J, New Anti-Inflammatory Compounds
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8 2323
Figure 6. Representative CD spectra (in MeCN, 22 °C) of splenocin structural groups indicating configuational similarity. Compounds 3 and 10
lack Cotton effects, whereas splenocins E and I show more complex CD spectra.
showed low nanomolar suppression of IL-5 production, while
compound 10 showed inhibition in the low micromolar range
(Table 5). The inhibitory activities of 1-10 are comparable to
that of the corticosteroid drug dexamethasone at similar nano-
molar concentrations (Figure 8). Currently, corticosteroids are
the most potent suppressors of cytokine production in asthma.
Both dexamethasone and compound 2 inhibit antigen induced
IL-5 production levels by greater than 80% of the levels of OVA
stimulated control splenocytes. The inhibitory levels of dex-
amethasone- and 2-treated cells are comparable to the production
levels measured in unstimulated splenocytes.
To determine whether the splenocins inhibited other TH2
cytokines in addition to IL-5, levels of the TH2 cytokine IL-13
were also analyzed in the splenocyte cultures (Table 5). These
data indicated that compounds 1-10 inhibit multiple TH2
cytokines including IL-5 and IL-13 at low ng/mL concentrations
in Vitro.
Since the splenocyte cell culture population is a heterogeneous
mixture of cell types, we were interested in determining whether
splenocins were acting solely on TH2 cells or if they also
inhibited the cytokine production capabilities of APCs. To
investigate this, we quantified cytokine levels of the APC-
derived cytokines IL-1 and TNF-R. These studies demonstrated
that the splenocins are capable of inhibiting both TH2 and
APC derived cytokines (Figure 9). Compounds 2 and 7
inhibited IL-1 by approximately 85% and TNF-R by ap-
proximately 75% relative to the OVA stimulated control cells.
Compounds 2 and 7 are as effective as dexamethasone in
inhibiting TH2 cytokine production and are only slightly less
effective than dexamethasone in suppressing APC cytokine
production of IL-1 and TNF-R.
Figure 7. Three dimensional structure (A) and vector diagram (B)
for compounds 4-9 and 10g illustrating the conformation which
supports the observed CD Cotton effect.
which is identical with that of the stereocenters reported for
the bis-lactone ring of the antimycins.23
In Vitro Biological Evaluation of Splenocins. The splenocins
show potent biological activities in an in Vitro mouse model of
allergen induced TH2 splenocyte cytokine production charac-
teristic of allergic asthma. As mentioned above, the parent
Streptomyces strain CNQ431 was selected for further investiga-
tion due to the low µg/mL noncytotoxic inhibition of OVA
stimulated splenocyte cytokine interleukin 5 (IL-5) production
seen from the crude extract. In pure form, compounds 1-9

2324 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8
Strangman et al.
Scheme 2. Reaction Scheme for the Benzoylation of Compounds 9 and 10 To Yield Identical Tribenzoylated Products (9b, 10g) and
Corresponding CD Spectra (in MeCN, 22 °C)
Table 5. Compounds 1-10 and dexamethasone (IC50 Values) for
Inhibition of TH2 Cytokine IL-5 and IL-13 Production and Cytotoxicity
(LD50 Values)
reports that antimycins induce cellular apoptosis in hepatocytes
by binding to the hydrophobic groove of Bcl-2/Bcl-xL proteins
on the surface of mitochondria.24 However, these studies also
report cell death in the micromolar ranges, and similar experi-
ments with lymphocytes or splenocytes have not been reported.
Recently reported studies with antimycins have also shown that,
at concentrations lower than the cytotoxic threshold, immuno-
suppressive anticancer activities are observed. Specifically, in
ViVo cancer models explored by Nose and co-workers have
shown nontoxic inhibition of angiogenesis and HIF-1R inhibition
by antimycins at concentrations similar to those used in our
investigations.25 Another recent in ViVo study investigated the
nontoxic cardioprotective effects of antimycin in combating
ischemia, while in Vitro studies have examined the nontoxic
effects of antimycins in mouse hepatocytes.26 To our knowledge,
the present study reflects the first investigations of the novel
splenocins and the structurally related antimycins in a model
of allergic inflammation.
IC50 (nM)
IL-5
IL-13
cytotoxicity, LD50 (nM)
dexamethasone
compound 1
compound 2
compound 3
compound 4
compound 5
compound 6
compound 7
compound 8
compound 9
5 ( 0.1
3.1 ( 1.2
1.8 ( 0.2
6.7 ( 0.2
47.9 ( 2.9
16.6 ( 1.8
9.4 ( 2.8
5 ( 0.4
5 ( 0.1
>7100
>1800
>1700
>560
>550
>570
>560
>550
>1600
>540
>6000
naa
1.6 ( 0.02
7.3 ( 4.2
43.7 ( 3.5
15.9 ( 1.1
6.8 ( 0.3
5.2 ( 0.1
5.1 ( 0.1
15.2 ( 1.3
4.3 ( 0.5
15.8 ( 1.0
compound 10 1022.7 ( 52.3 826.3 ( 187.6
a Not available.
The splenocyte assay we have utilized provides a novel assay
to screen natural product libraries for compounds with the ability
to inhibit pro-inflammatory cytokines produced by the antigen-
Figure 8. Interleukin 5 (IL-5) levels of cultured splenocytes. (A)
Unstimulated control. (B) Ova-stimulated control. (C) Ova-stimulated
+ compound 2 (5 nM). (D) Ova-stimulated + dexamethasone (5 nM).
Due to the structural similarity of the splenocins to the known
antimycins, compound 17 (Sigma) was also tested in this assay.
Antimycin A2, which has a hexyl alkyl chain at C-7 and a propyl
acyl chain at C-8, displayed a similar activity profile to that
seen from the splenocins (data not shown). A recent paper
Figure 9. Splenocyte cytokine inhibition: Levels of TH2 (IL-5, IL-
13) and APC related (IL-1, TNF-R) cytokines of Ova-stimulated
splenocytes incubated with either compound 2 or 7 at 10 nM or
dexamethasone at 10 nM.

Splenocins A-J, New Anti-Inflammatory Compounds
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8 2325
mixtures of n-hexanes-EtOAc (10:1), n-hexanes-EtOAc (5:1),
n-hexanes-EtOAc (2.5:1), n-hexanes-EtOAc (1:1), EtOAc,
EtOAc-MeOH (10:1), and 100% MeOH, successively. The
EtOAc-MeOH (2.5:1) and EtOAc-MeOH (1:1) eluting fractions
(labeled Q431-3 and -4) showed the most potent IL-5 inhibition
levels in the mouse splenocyte assay, and by LCMS, contained
1-9 as well as several antimycins. The 100% EtOAc (CNQ431-5)
fraction contained 10 as well as several 8-hydroxy antimycins.
Fractions CNQ-431-3, -4, and -5 were refractionated by C-18
reversed-phase HPLC with an MeCN-H2O step gradient from 40%
MeCN to 100% MeCN over 110 min to obtain nearly pure 1-10.
Final purification was performed by C-8 and C-18 reversed-phase
column chromatography with isocratic solvent systems ranging
between 50:50 and 90:10 MeCN:H2O.
presenting cell-T-cell interaction. By using corticosteroids as
our control therapy, we are able to compare the anti-inflam-
matory potency of novel compounds to the most potent anti-
inflammatory molecules in clinical use today. Interestingly, in
this study we have identified a series of new compounds, 1-9,
that display potency levels equal to those of the corticosteroid
dexamethasone. As corticosteroids are currently the most
effective therapy in asthma, identifying new compounds that
have potent anti-inflammatory activities, but do not have the
same negative side effects associated with corticosteroid therapy,
would be an important therapeutic advance.
While further immunological experiments are in preparation
to further investigate this activity using the mouse allergic
asthma model in ViVo, our initial data suggests a potential for
therapeutic development of a splenocin-derived drug for inflam-
matory diseases such as asthma, as well as their potential uses
as anti-inflammatory agents in other diseases in which suppres-
sion of cytokine production by APCs and T-cells may be
beneficial.
Compound 1. White amorphous powder: 2 mg. CD (MeCN):
∆ε220 +2.5, ∆ε190 -4.8. UV (EtOH): 230 (ε ) 23,000), 360 (ε )
5,750) nm. IR (film): νmax 3381, 1748, 1527, 1368, and 736 cm-1
.
HR ESI-TOF MS: Obsd m/z 535.1683 [M + Na]+ (C26H28N2O9Na
requires 535.1687). See Table 1 for NMR data.
Compound 2. White amorphous powder: 6 mg; [R]D ) +68
(0.1, CH3OH). CD (MeCN): ∆ε213 +4.2, ∆ε190 -10.8. UV (EtOH):
230 (ε 18,000), 320 (ε ) 4,300) nm. IR (film): νmax 3361, 1746,
1527, 1369, and 737 cm-1. HR ESI-TOF MS: Obsd m/z 541.2167
[M + H]+ (C28H33N2O9 requires 541.2180). See Table 1 for NMR
data.
Experimental Section
General Experimental Procedures. 1H, 13C, and 2D NMR
spectral data were obtained on Varian Inova 300 and 500 MHz
NMR spectrometers. UV spectra were recorded on a Varian Cary
50 Conc UV/visible spectrophotometer with a path length of 1 cm.
IR spectra were recorded in a Nicolet IR100 FT-IR spectrometer.
Optical rotations were measured using a Rudolph Research Autopol
III automatic polarimeter with a 10 cm cell. High resolution
MALDI-FTMS data were collected on an IonSpec Ultima mass
spectrometer at the Scripps Research Institute, La Jolla. Low
resolution LC/MS data were obtained on a Hewlett-Packard series
1100 LC/MS system with a reversed phase C18 column (Agilent,
4.6 mm × 100 mm, 5 µm) with a 0.7 mL/min flow rate. CD spectra
were obtained on a Jasco 810 spectropolarimeter with a 1 cm path
length. CD spectra were also recorded using an AVIV Instruments
Inc. model 215 circular dichroism Spectrometer. OD measurements
of ELISA experiments were recoded at 450 nm on a Biorad Model
680 microplate reader. The purity of all new natural products
(1-10) was assessed by analytical HPLC using either silica gel
(normal phase) or C-18 reversed phase columns. HPLC traces,
indicating high purity (ca. 95% or greater) for 1-10, are found in
the Supporting Information.
Compound 3. White amorphous powder: 2 mg. CD (MeCN):
∆ε220 +4.4, ∆ε190 -7.0. UV (EtOH): 230 (ε ) 19,400), 320 (ε )
4,400) nm. IR (film): νmax 3381, 1747, 1527, 1355, and 744 cm-1
.
HR ESI-TOF MS: Obsd m/z 555.2322 [M + H]+ (C29H35N2O9
requires 555.2337). See Table 1 for NMR data.
Compound 4. White amorphous powder: 2 mg. CD (MeCN):
∆ε238 -2.2, ∆ε220 +2.3. UV (EtOH): 230 (ε ) 16,400), 320 (ε )
3,000) nm. IR (film): νmax 3361, 1740, 1537, 1350, and 702 cm-1
.
HR ESI-TOF MS: Obsd m/z 511.1716 [M - H]- (C26H27N2O9
requires 511.1722). See Table 2 for NMR data.
Compound 5. White amorphous powder: 2 mg. CD (MeCN):
∆ε237 -2.9, ∆ε221 +4.0. UV (EtOH): 230 (ε ) 18,400), 320 (ε )
3,300) nm. IR (film): νmax 3381, 1753, 1534, 1345, and 737 cm-1
.
HR ESI-TOF MS: Obsd m/z 541.2168 [M + H]+ (C28H31N2O9
requires 541.2180). See Table 2 for NMR data.
Compound 6. White amorphous powder: 2 mg. CD (MeCN):
∆ε241 -2.2, ∆ε220 +4.5. UV (EtOH): 230 (ε ) 21,100), 320 (ε )
4,400) nm. IR (film): νmax 3347, 1760, 1540, 1369, and 744 cm-1
.
HR ESI-TOF MS: Obsd m/z 555.2326 [M + H]+ (C29H35N2O9
requires 555.2337). See Table 2 for NMR data.
Compound 7. White amorphous powder: 2 mg. CD (MeCN):
∆ε239 -2.1, ∆ε221 +2.8. UV (EtOH): 230 (ε ) 22,700), 320 (ε )
5,700) nm. IR (film): νmax 3347, 1740, and 764 cm-1. HR ESI-
TOF MS: Obsd m/z 569.2481 [M + H]+ (C30H37N2O9 requires
569.2493). See Table 2 for NMR data.
Isolation of CNQ431 Strain, Identification, Cultivation,
and Extraction. Marine-derived Streptomyces strain CNQ431 was
isolated using growth medium A1+C (10 g of starch, 4 g of
peptone, 2 g of yeast extract, 1 g of calcium carbonate, 18 g of
agar, and 1 L of seawater) from a marine sediment collected from
a depth of 30 m 1 mile Northwest of the Scripps Institution of
Oceanography (La Jolla, CA). Strain CNQ431 was classified as a
Streptomyces sp. based on 16S rDNA analysis which showed 100%
sequence identity between strain CNQ431 and 5 strains including
Streptomyces fungicidus, strain AY636155; two strains of an
unidentified Streptomyces, strains AF429400 and AM21770.1; and
two identified strains of bacteria in the family Streptomycetaceae,
AY944261 and AY944255. The strain was cultured in 60 × 1 L
volumes of media A1 while shaking at 230 rpm for 3 days. After
72 h, 20 g/L of XAD-7 adsorbent resin was added to each flask
and the flasks were shaken for an additional 24 h. At this time, an
additional 20 g/L of XAD-7 adsorbent resin was added. The cultures
were shaken for an additional 24 h, at which time the resin was
collected by filtration through cheesecloth, washed with deionized
water and eluted twice with acetone. Evaporation of the acetone in
Vacuo left a wet residue, which was partitioned with EtOAc,
providing 130 mg of dry organic extract per 1 L of culture after
solvent removal.
Compound 8. White amorphous powder: 7 mg. CD (MeCN):
∆ε240 -10.3, ∆ε221 +14.9. UV (EtOH): 230 (ε ) 23,000), 320
(ε ) 6,000) nm. IR (film): νmax 3315, 1725, 1534, 1275, 785 cm-1
.
HR ESI-TOF MS: Obsd m/z 583.2664 [M + H]+ (C31H39N2O9
requires 583.2577). See Table 2 for NMR data.
Compound 9. White amorphous powder: 5 mg. CD (MeCN):
∆ε238 -4.5, ∆ε217 +3.9. UV (EtOH): 230 (ε ) 23,000), 350 (ε )
5,750) nm. IR (film): νmax 3374, 1744, 1535, 1370, and 734 cm-1
.
HR ESI-TOF MS: Obsd m/z 575.2000 [M + H]+ (C31H31N2O9
requires 575.2024). See Table 3 for NMR data.
Compound 10. White amorphous powder: 6 mg. [R]D ) +70
(c 0.18, CH3OH). UV (EtOH): 230 (ε ) 29,600), 337 (ε ) 5,600)
nm. IR (film): νmax 3375, 1740, 1540, 1360, and 730 cm-1. HR
ESI-TOF MS: Obsd m/z 493.1573 [M + Na]+ (C24H26N2O8Na
requires 493.1581). See Table 3 for NMR data.
Mosher MTPA Esters 10a/10b, 10c/10d, 10e/10f. Compound
10 (1.0 mg) was divided into two equal portions, and each was
dissolved in 500 µL of pyridine in separate reaction vials. To each,
5 mg of DMAP and either 10 µL of (R)-MTPA-Cl or 10 µL of
(S)-MTPA-Cl was added. After 5 h, LC-MS analysis indicated
that equal amounts of di- and tri-Mosher ester products were formed.
Isolation and Purification of Compounds 1-10. The EtOAc
extract of a 60 × 1 L fermentation (7.0 g) was subjected to silica
gel column chromatography purification eluting with solvent

2326 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8
Strangman et al.
The reaction was terminated and reaction products were purified
by C-18 reversed phase HPLC. The H NMR spectra for the (R)-
MTPA-Cl and (S)-MTPA-Cl di- and tri-Mosher esters were
recorded (see Supporting Information for NMR data).
Grant CA44848. The authors express thanks to C. Kauffman
for culturing this strain, A. Prieto-Davo´ for isolating and
identifying this Streptomyces strain, K. and S. McElwain and
R. Serrano for mouse immunizations and Dr. M. Miller and
Dr. J-.Y. Cho for helpful discussions. We are particularly
grateful to Ted Molinski (UCSD) for advice and discussion on
applying CD spectral data to the splenocins.
1
Benzoylation of Compounds 9 and 10 To Yield 9b and
10g. Compounds 9 and 10 (0.5 mg each) were dissolved in separate
1.0 mL volumes of pyridine. To each sample, DMAP (5 mg) and
benzoyl chloride (1 mg) were added, and the reaction was allowed
to proceed for 12 h. The reaction mixture was filtered and
concentrated under reduced pressure. The resulting benzoylated
products 9b and 10g were purified by C-18 HPLC and analyzed
by LC-MS. LRMS (ESI) 9b, 10b: m/z 777.24 [M + Na], m/z
777.24 [M + Na] (see Supporting Information for CD spectra).
Marfey’s Analysis of the Compound 6- and 10-Derived
Threonine. Approximately 0.2 mg each of compounds 6 and 10
were separately hydrolyzed with 6 N HCl (0.8 mL) overnight in a
hot oil bath at 105 °C. The hydrolysates were then evaporated to
dryness and resuspended in 0.2 mL of H2O. To each aqueous
hydrolysate, 0.5 mg of 1-fluoro-2,4-dinitrophenyl-5-L-leucinamide
(L-FDLA) suspended in 0.1 mL of acetone and then 20 µL of 1 N
NaHCO3 were added and the mixtures heated at 40 °C for 5 min.
The mixtures were next cooled to room temperature, neutralized
with 2 N HCl and evaporated to dryness. The residues were then
resuspended in 500 µL of MeCN and subjected to LC-MS analysis
by comparison to standard D-threonine and L-threonine L-FDLA
derivatives that had been prepared in the same fashion. The
derivatized threonine residues, derived from compounds 6 and 10,
were found to have identical retention times as the L-threonine
derivatized standard (L-threonine ) 19.7 min; D-threonine ) 22.7
min).
Supporting Information Available: Characterization data for
compounds 1-10, including 1D and 2D NMR spectra, UV
spectra, HRMS data, CD spectra and HPLC traces, as well as
1H NMR chemical shift data for di- and tri-Mosher products.
This material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) Ikeda, R.; Nayar, J.; Cho, J. Y.; Miller, M.; Rodriguez, M.; Raz, E.;
Broide, D. H. Resolution of airway inflammation following ovalbumin
challenge. Am. J. Respir. Cell Mol. Biol. 2003, 28, 655–663.
(2) Itano, A.; Jenkins, M. Antigen presentation to naive CD4 T cells in
the lymph node. Nat. Immunol. 2003, 4, 733–739.
(3) Greenfeder, S.; Umland, S. P.; Cuss, F. M.; Chapman, R. W.; Egan,
R. W. Th2 cytokines and asthma: the role of interleukin-5 in allergic
eosinophilic disease. Respir. Res. 2001, 2, 71–79.
(4) Wills-Karp, M.; Chiaramonte, M. Interleukin-13 in asthma. Curr. Opin.
Pulm. Med. 2003, 9, 21–27.
(5) Barnes, P. J. New drugs for asthma. Nat. ReV. Drug DiscoVery 2004,
3, 831–844.
(6) Salvi, S. S.; Krishna, M. T.; Sampson, A. P.; Holgate, S. T. The anti-
inflammatory effects of leukotriene modifying drugs and their use in
asthma. Chest 2001, 119, 1533–1546.
Mouse Immunization, Cell Culturing and Cell Viability. In
order to create a mouse model of allergen induced TH2 cytokine
production suitable for screening, female Balb-C mice age 6-8
weeks were immunized subcutaneously with 25 µg of ovalbumin
(OVA) allergen adsorbed to 1 mg of alum adjuvant in 200 µL of
phosphate-buffered saline (PBS) weekly for four weeks according
to a protocol previously described by our laboratory.27 At the
conclusion of the four-week immunization period, the mice were
sacrificed and their spleens surgically removed. Splenocytes were
separated through a cell strainer. The splenocytes were then
suspended in RPMI-1640 media, centrifuged, and resuspended in
(7) Koehn, F.; Carter, G. T. The evolving role of natural products in drug
discovery. Nat. ReV. Drug DiscoVery 2005, 4, 206–220.
(8) Fenical, W.; Jensen, P. R. Developing a new resource for drug
discovery: marine actinomycete bacteria. Nat. Chem. Biol. 2006, 2,
666–673.
(9) Barrow, C. J.; Oleynek, J. J.; Marinelli, V.; Sun, H. H.; Kaplita, P.;
Sedlock, D. M.; Gillum, A. M.; Chadwisk, C. C.; Cooper, R.
Antimycins, inhibitors of ATP-citrate lyase, from a Streptomyces sp.
J. Antibiotics 1997, 50, 729–733.
(10) Hosotani, N.; Kumagai, K.; Nakagawa, H.; Shimatani, T.; Saji, I.
Antimycins A10-A16, seven new antimycin antibiotics produced by
Streptomyces spp. SPA10191 and SPA-8893. J. Antibiotics 2005, 58,
460–467.
(11) Shiomi, K.; Hatae, K.; Hatano, H.; Matsomoto, A.; Takahashi, Y.;
Jiang, C.-L.; Tomoda, H.; Kobayashi, S.; Tanaka, H.; Omura, S. A
new antibiotic, antimycin A9, produced by Streptomyces sp. K01-0031.
J. Antibiotics 2005, 58, 74–78.
(12) Fondja Yao, C. B.; Schiebel, M.; Helmke, E.; Anke, H.; Laatsch, H.
Prefluostatin and new urauchimycin derivatives produced by Strep-
tomycete isolates. Z. Naturforsch. 2006, 61b, 320–325.
(13) Ishiyama, T.; Endo, T.; Otake, N.; Yonehara, H. Deisovalerylblast-
mycin produced by Streptomyces sp. J. Antibiot. 1976, 29, 804–808.
(14) Imamura, N.; Nishijima, M.; Adachi, K.; Sano, H. Novel antimycin
antibiotics, urauchimycins A and B, produced by a marine actino-
mycete. J. Antibiot. 1992, 46, 241–246.
(15) Hayashi, K.-I.; Nozaki, H. Kitamycins, new antimycin antibiotics
produced by Streptomyces sp. J. Antibiotics 1999, 52, 325–328.
(16) Ohtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. High-field FT
NMR application of mosher’s method. The absolute configurations
of marine terpenoids. J. Am. Chem. Soc. 1991, 113, 4092–4096.
(17) Ohkawa, S.; Fukatsu, K.; Miki, S.; Hashimoto, T.; Sakamoto, J.; Doi,
T.; Nagai, Y.; Aono, T. 5-Aminocoumarans: dual inhibitors of lipid
peroxidation and dopamine release with protective effects against
central nervous system trauma and ischemia. J. Med. Chem. 1997,
40, 559–573.
(18) Kikuchi, H.; Saito, Y.; Sekiya, J.; Okano, Y.; Saito, M.; Nakahata,
N.; Kubohara, Y.; Oshima, Y. Isolation and synthesis of a new
aromatic compound, brefelamide, from dictyostelium cellular slime
molds and its inhibitory effect on the proliferation of astrocytoma cells.
J. Org. Chem. 2005, 70, 8854–8858.
6
RPMI-1640 media to a concentration of 1.3 × 10 cells/mL and
then aliquoted into 96 well plates (100 µL/well). To each test well,
containing splenocytes suspended in media, an aliquot of either
crude extract, semipurified extract fractions, or pure compounds
suspended in DMSO were added, in addition to an aliquot of
stimulating OVA allergen. In addition, replicate wells containing
either DMSO or DMSO and OVA allergen were plated as
controls. The plates were then incubated for 48 h, during which
time the OVA allergen was taken up by the splenocyte antigen
presenting cells, and allergen fragments were displayed on the
APC surface to stimulate OVA allergen specific TH2 cells
induced to be present in the spleen by the in ViVo OVA
immunization. Control OVA-stimulated splenocytes secrete a
variety of cytokines into the supernatant at much higher levels
than non-OVA-stimulated splenocytes. Adherent splenocytes
were stained with trypan blue and counted visually on a
hemocytometer to determine cell viability.
Cytokine Quantification. At the end of the splenocyte incuba-
tion period, the culture supernatants were removed and cytokine
production was quantified using colorimetric-based sandwich
enzyme linked immunosorbant assays (ELISAs). Mouse IL-1, -5,
-13, and TNF-R (R&D Systems) ELISAs were carried out according
to the manufacturer’s specifications. Cytokine standards were
included in each replicate assay in order to create accurate standard
curves.
(19) Campbell, T. W.; McCoy, V. E.; Kauer, J. C.; Foldi, V. S. Preparation
of some 9,10-difunctional derivatives of 9,10-dihydro-9,10-ethanoan-
thracene. J. Org. Chem. 1961, 26, 1422–1426.
(20) Fujii, K.; Ikai, Y.; Mayumi, T.; Oka, H.; Suzuki, M.; Harada, K. A
nonempirical method using LC/MS for determination of the absolute
configuration of constituent amino acids in a peptide: elucidation of
limitations of Marfey’s method and of its separation mechanism. Anal.
Chem. 1997, 69, 3346–3352.
Acknowledgment. This work was a result of financial
support from the National Institutes of Health (NIGMS,
GMO67557 as part of a T-32 training program in marine
biotechnology) and in part by the National Cancer Institute under

Splenocins A-J, New Anti-Inflammatory Compounds
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8 2327
(21) Fujii, K.; Ikai, Y.; Oka, H.; Suzuki, M.; Harada, K. A nonempirical
method using LC/MS for determination of the absolute configuration
of constituent amino acids in a peptide: combination of Marfey’s
method and with mass spectrometry and its practical application. Anal.
Chem. 1997, 69, 5146–5151.
(22) Berova, N.; Nakanishi, K.; Woody, R. W. Circular Dichroism
Principles and Practice, 2nd ed.; Wiley-VCH, Inc.: New York,
2000.
(23) Kinoshita, M.; Aburaki, S.; Umezawa, S. Absolute configurations
of antimycin lactones and antimycin A. J. Antibiot. 1972, 15, 373–
376.
(24) Tzung, S.-P.; Kim, K. M.; Basanez, G.; Giedt, C. D.; Simon, J.;
Zimmerberg, J.; Zhang, K. Y. J.; Hockenbery, D. M. Antimycin A
mimics a cell-death-inducing Bcl-2 homolgy domain 3. Nat. Cell Biol.
2001, 3, 183–191.
(25) Maeda, M.; Hasebe, Y.; Egawa, K.; Shibanuma, M.; Nose, K.
Inhibition of angiogenesis and HIF-1R activity by antimycin A1. Biol.
Pharm. Bull. 2006, 29, 1344–1348.
(26) Kabir, A. M. N.; Cao, X.; Gorog, D. A.; Bassi, R.; Bellahcene, M.;
Quinlan, R. A.; Davis, R. J.; Flavell, R. A.; Shattock, M. J.; Marber,
M. S. Antimycin A induced cardioprotection is dependant on pre-
ischemic p38 MAPK activation but independant of MKK3. J. Mol.
Cell. Cardiol. 2005, 39, 709–717.
(27) Broide, D.; Schwarze, J.; Tighe, H.; Gifford, T.; Nguyen, M.-D.;
Malek, S.; Van Uden, J. H.; Martin-Orozco, E.; Gelfand, E. W.; Raz,
E. Immunostimulatory DNA sequences inhibit IL-5, eosinophilic
inflammation, and airway responsiveness in mice. J. Immunol. 1998,
161, 7054–7062.
JM801110J