Motualevic acids and analogs: Synthesis and antimicrobial structure-activity relationships

Article abstract of DOI:10.1016/j.bmcl.2010.05.073

Synthesis of the marine natural products motualevic acids A, E, and analogs in which modifications have been made to the ω-brominated lipid (E)-14,14-dibromotetra-deca-2,13-dienoic acid or amino acid unit are reported, together with antimicrobial activities against Staphylococcus aureus, methicillin-resistant S. aureus, Enterococcus faecium, and vancomycin-resistant Enterococcus.

Full text of DOI:10.1016/j.bmcl.2010.05.073

Bioorganic & Medicinal Chemistry Letters 20 (2010) 4108–4111
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Motualevic acids and analogs: Synthesis and antimicrobial
structure–activity relationships
*
Pradeep Cheruku, Jessica L. Keffer, Cajetan Dogo-Isonagie, Carole A. Bewley
Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD 20892, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Synthesis of the marine natural products motualevic acids A, E, and analogs in which modifications have
Received 8 April 2010
Revised 15 May 2010
Accepted 18 May 2010
Available online 9 June 2010
been made to the
x-brominated lipid (E)-14,14-dibromotetra-deca-2,13-dienoic acid or amino acid unit
are reported, together with antimicrobial activities against Staphylococcus aureus, methicillin-resistant S.
aureus, Enterococcus faecium, and vancomycin-resistant Enterococcus.
Published by Elsevier Ltd.
Keywords:
Antibacterial
Gram-positive bacteria
Marine natural product
Dibromo olefin
Starting with Fleming’s discovery of penicillin in the 1920’s, the
discovery and development of new antibiotics was marked by un-
matched success in the area of drug discovery and remained at the
fore for decades.1 Numerous classes of antibacterial agents were
discovered from Nature, and many of those are in clinical use to-
day. The expansion of antibiotic research reached its peak by the
middle of the last century, then paused for four decades before ap-
proval of oxazolidinone linezolid in 20002 and the lipopeptide dap-
tomycin in 2003.3 Nonetheless, a variety of infectious diseases
persist as true global health burdens,4 and infections associated
with drug resistant organisms5 such as methicillin-resistant Staph-
ylococcus aureus (MRSA) and vancomycin-resistant Enterococcus
spp. (VRE) continue to rise. To help alleviate this problem, develop-
ment of new antibiotics that function with unique mechanisms of
action and can escape the resistance mechanisms used by drug
resistant bacteria are needed. Combined studies involving discov-
ery and synthesis of new antimicrobial natural products can be
valuable in this area.6a,b
against SA only at much higher concentrations (50 lg/disk).
Interestingly, the naturally occurring amides 1b and 1c and the Z
isomer of 1a also exhibited greatly diminished antimicrobial activity
providing some insight into the structure–activity relationships for
this series. To obtain a more general picture of the chemical features
necessary for antimicrobial activity, we have synthesized motual-
evic acids A and E together with a number of analogs and tested their
effects on growth of several bacterial strains, the results of which are
reported here.
Synthesis of motualevic acid A (1a) began with monoprotection
of commercially available and inexpensive 1,9-nonanediol with
TBDPSCl (Scheme 1). Conversion of the remaining hydroxyl group
to a,b-unsaturated ester 5 was accomplished by an oxidation-Wit-
tig reaction sequence.8 Successive reductions of the double bond
O
O
Br
Recently we reported the isolation, structure elucidation and
antibacterial properties of motualevic acids A–E (e.g., 1a–c and 8)
from the marine sponge Siliquariaspongia sp.7 These compounds
N
CO2R
N
H
CO2H
9
H
n
Br
2a n = 7
2b n = 1
motualevic acid A: 1a
motualevic acid C: 1b
motualevic acid D: 1c
R = H
R = NH2
contain an unusual
x-brominated lipid (E)-14,14-dibromotetra-
R = NMe2
deca-2,13-dienoic acid, where motualevic acids A–D are its glycyl
conjugates and motualevic acid E the free acid (Fig. 1). Antimicrobial
assays revealed that motualevic acid A (1a) inhibited the growth of S.
aureus (SA) and MRSA at low micromolar concentrations, while
motualevic acid E (8) showed greatly diminished activity, effective
O
O
Br
Br
N
H
CO H
x
2
n
OH
9
Br
Br
3
n = 7, x = 1
motualevic acid E: 8
4a n = 2, x = 1
4b n = 2, x = 7
* Corresponding author. Tel.: +1 301 594 5187.
E-mail address: caroleb@mail.nih.gov (C.A. Bewley).
Figure 1. Examples of motualevic acids and synthetic analogs.
0960-894X/$ - see front matter Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2010.05.073

P. Cheruku et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4108–4111
4109
Scheme 1
i-iii
iv,ii,iii
MeO2C
MeO2C
HO
OH
OTBDPS
OTBDPS
7
89%
3 steps
57%
3 steps
7
9
5
6
O
vi,ii,vii
viii
ix,viii
MeO2C
Br
HO2C
Br
Br
Br
Br
HO2C
N
H
quant.
87%
3 steps
73%
2 steps
9
9
9
Br
Br
Br
7
8
1a
3
Scheme 2
O
MeO2C
Br
HO2C
Br
Br
vi,ii,vii
89%
viii
ix,viii
5
7
7
HO2C
N
93%
2 steps
93%
7
Br
H
3 steps
9
10
Scheme 3
viii,ix,viii
viii,x,viii
78%
2 steps
O
4a
4b
MeO2C
Br
Br
v,iii
ii,vii
89%
2 steps
MeO2C
OH
2
2
51%
O
2 steps
11
12
Schemes 1–3. Reagents and conditions: (i) TBDPSCl, imidazole, THF, 0 °C, 5 h; (ii) Dess–Martin periodinane, DCM, rt, 0.5–1 h; (iii) (C6H5)3P@CHCO2CH3, toluene, 80 °C, 4 h;
(iv) pd/C. H2, rt, 0.5 h; (v) DIBAL-H, toluene, À70 °C, 4 h; (iv) TBAF, THF, rt, O/N; (vii) PPh3, CBr4, CH2Cl2, 0 °C to 15 °C, 1 h; (viii) 1 M NaOH, THF, rt, on; (ix) glycine methylester
hydrochloride, EDC, HOBt, DMF, 0 °C to rt, 4–6 h; (x) caprylic acid methyl ester, EDC, HOBt, DMF, 0 °C to rt, 4–6 h.
and ester by catalytic hydrogenation and treatment with DIBAL-H,
followed by a second two-carbon Wittig olefination afforded ,b-
ey–Fuchs reaction conditions (Scheme 2). Saponification and cou-
pling to glycine provided analog 3, whose fatty acid unit is two
carbons shorter than natural product 1. As shown in Scheme 3,
the C7 analog (E)-7,7-dibromohex-2,6-dienoic acid present in ana-
a
unsaturated ester 6. Fluoride-induced cleavage of the silyl-protect-
ing group, followed by oxidation and a Corey–Fuchs gem-dibromo
olefination9 yielded the vinyl gem-dibromo olefin 7.10 Saponfication
of 7 provided motualevic acid E (8), that was in turn coupled to gly-
cine methyl ester using standard coupling reagents (EDC, HOBT);
and saponification of the methylester completed the total synthesis
of motualevic acid A (1) in 31% overall yield (1 mmol scale).11
Analytical data for synthetic motualevic acids A and E7 were identi-
cal to those of the natural products.
logs 4a–b was prepared starting with commercially available
butyrolactone. Reduction with DIBAL-H provided the hemiacetal,
and a Wittig reaction afforded the
,b-unsaturated ester 11.14 Oxi-
c-
a
dation of 11 to its aldehyde followed by Corey–Fuchs gem-dibromo
olefination afforded 12, and subsequent saponification, coupling to
glycyl or caprylic acid methyl ester, and saponification gave ana-
logs 4a–b.
With motualevic acid E (8) in hand, we synthesized a range of
additional conjugates 1d–n, where amino acids or diamines were
chosen to explore the effects of charge, hydrophobicity, and flexi-
bility in this region of the molecule. Acid 8 was activated with
HOBt and coupled to respective amino acid methyl esters, followed
by hydrolysis to give 1d–m; or coupled to 2-amino-N,N,N-trimeth-
ylethaneammonium chloride to give quaternary ammonium 1n
(Table 1).
Antimicrobial susceptibility testing was carried out using agar
disk diffusion and microbroth dilution assays using Clinical Labora-
tory Standards Institute guidelines as described in our previous
work.7 Initial screening was performed using the solid agar assays,
and for compounds exhibiting zones of inhibition of at least 8–
10 mm at 25
performed.
lg/disk, microbroth dilution assays were also
Prior to synthesizing the suite of amino acid conjugates, com-
pounds designed to investigate the antimicrobial effects of fatty
acid chain length and composition were tested first. As shown in
Table 1, the composition of the fatty acid was critical to the antimi-
crobial activity, and any changes that were made to the dienoic
acid in this study diminished or abolished activity. Thus glycyl con-
jugates 2a and 2b, both of which lack the terminal dibromoolefin
and possess shorter alkyl chains (C12 and C6, respectively) were
inactive. Surprisingly, activity was diminished by at least fivefold
for analog 3 that is almost identical to motualevic acid A differing
only by deletion of an ethylene unit. Compounds 4a and 4b were
constructed both to test whether a truncated dienoic acid contain-
ing a terminal dibromoolefin could substitute for the natural C14
dienoic acid, and whether total chain length was sufficient for anti-
microbial activity and, in the case of 4b, whether position of the
amide mattered. Truncated glycyl conjugate 4a was inactive. Mod-
est activity was observed for the more conservative analog 4b, but
zones of inhibition were hazy. Thus, modification of motualevic
acid A within the fatty acid portion of the molecule is detrimental
to antibacterial activity and led us to construct all other conjugates
from motualevic acid E (8).
To explore the effects of fatty acid chain length and composition
on antimicrobial activity, we synthesized glycyl conjugates of the
a,b-unsaturated C6, C7 and C12 acids, with and without the termi-
nal dibromovinyl olefin. Knowing that the Z configuration at C2–C3
diminished activity, (E)-didec-2-enoic acid was synthesized start-
ing from decanal using the same homologation approach employed
for synthesis of 1 according to published procedures.12,13 Coupling
of the synthetic (E)-didec-2-enoic acid or commercially available
(E)-hex-2-enoic acid to glycine provided alkenes 2a–b. Similarly,
the protected C11
a,b-unsaturated ester 5 was converted to the
C12 gem-dibromoolefin 9 after deprotection, oxidation and Cor-
Table 1
In vitro antimicrobial activities of compounds 1–4 and 8a
Compound(s)
SA
MRSA
1a
10
na
25
50
50
10
na
25
50
na
1b–c, 2a–b, 4a
4b
3
8
Subsequent to our original report describing antimicrobial
activities of motualevic acids A–E toward SA and MRSA, these com-
pounds were further screened for effects on EF and VRE. Similar to
a
Determined by agar disk diffusion assays; values are in
l
g/disk and represent
zones of inhibition of 8–11 mm; na (not active at 50
l
g/disk).

4110
P. Cheruku et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4108–4111
the anti-staphylococcal activities, motualevic acid A inhibited the
growth of EF and VRE with similar potency (Table 2). Thus, amino
acid/amine conjugates 1d–n were tested against these four bacte-
rial strains.
Slight differences in potency were observed in disk diffusion as-
says for analogs 1d and 1e, conjugates of b-alanine and 3-amino-
propanoic acid, respectively; but in liquid culture both analogs
inhibited the growth of SA and MRSA with MIC values around
bial profile for 1i (L-proline) was almost identical to motualevic
acid A; and somewhat surprisingly, its enantiomer 1j showed re-
duced activity for all strains but VRE. The hydroxyproline conju-
gate showed only modest activity in disk diffusion assays and
was not tested further.
The structure–activity profile for natural motualevic acids
showed that the presence of a free carboxylic acid was essential
for antimicrobial activity. This prompted us to synthesize aspartic
acid- and diacetic acid conjugates 1l and 1m, analogs with a charge
of À2. In disk diffusion and microbroth dilution assays both were
found to be inactive against all strains in both formats. Finally,
mindful of the properties of quaternary ammonium compounds
as general antiseptics, we prepared the 2-amino-N,N,N-trimethyle-
thanaminium conjugate 1n. Similar to the leucine analog 1h, 1n
showed improved potency toward all strains relative to 1a, with
3
lg/mL, an improvement over 1a; and inhibited EF and VRE with
MIC values of 8.5–12 g/mL, similar to 1a. Two aromatic amino
l
acid conjugates 1f and 1g were prepared by coupling 8 to phenyl-
glycine or phenylalanine, respectively. Surprisingly, these analogs
showed very different potencies where MICs for 1f averaged
20
2.9 to 6.3
conjugate 1h proved to be the most potent of all analogs, and
inhibited the growth of SA and MRSA with MICs less than 3 g/
lg/mL over the four strains, while values for 1g ranged from
lg/mL and showed good activity toward VRE. The leucine
MIC50 values of 2.4–5.2 lg/mL.
l
In summary, we have synthesized the antibacterial marine nat-
ural products motualevic acids A and E, together with analogs de-
signed for structure-function relationships within the fatty acid
mL. To test the effects of rigidity of the amino acid unit on activity,
several proline-containing analogs were prepared. The antimicro-
Table 2
Antimicrobial data for conjugates 1a and 1d–n
O
Br
coupling
8
R
9
Amine (R),
HOBt, EDC
Br
R
(Disk diffusiona) microbroth dilutionb
SA
MRSA
EF
VRE
N
H
CO2H
CO2H
CO2H
1a
1d
1e
(10) 11
(10) 3.6
(25) 2.5
(10) 9.3
(25) 9.3
(10) 12
2
3
N
H
(10) 3.3
(25) 2.4
(25) 11
(10) 11
(10) 12
(10) 8.5
N
H
Ph
Bn
CO2H
NH
CO2H
NH
1f
1g
1h
1i
(25) 17
(na) 5.0
(25) 2.3
(25) 10
(25) 18
(25) 21
(50) 4.4
(25) 2.9
(10) 9.0
(25) 19
(10) 25
(5) 6.3
(10) 22
(5) 2.9
(5) 3.7
(5) 12
CO2H
(10) 6.4
(25) 10
(25) 15
NH
CO2H
N
CO2H
1j
(10) 11
N
HO
CO2H
1k
(na) nt
(25) nt
(25) nt
(25) nt
N
CO2H
NH
HO2C
N
1l
(na) na
(50) na
(10) 2.4
(na) na
(50) na
(10) 4.2
(na) na
(na) na
(25) 3.7
(na) na
(na) na
(25) 5.2
CO2H
CO2H
1m
1n
N
N
H
Cl
a
Results for disk diffusion assays are shown in parentheses in
l
g/disk and correspond to values giving 8–10 mm zones of inhibition.
b
MIC50 values reported as
l
g/mL. Standard deviations averaged 25%; na = not active at 50 g/disk or higher; nt = not tested. Bacterial strains included Staphylococcus
l
aureus (SA), methicillin-resistant S. aureus (MRSA), Enterococcus faecium (EF), and vancomycin-resistant Enterococcus (VRE).

P. Cheruku et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4108–4111
4111
and amino acid units. This work has shown that the
x-brominated
References and notes
lipid (E)-14,14-dibromotetra-deca-2,13-dienoic acid present in all
of the natural products is required for low micromolar antimicro-
bial activity, replacement of glycine by more hydrophobic amino
acids (with the exception of phenylglycine) increases potency,
and introduction of polar amino acids or more than one negative
charge abrogates activity.
Compounds 1a, 1g, 1h, and 1n were tested for cytotoxicity to-
ward a control mammalian cell line (BSC1); each exhibited mild
toxicity only when treated at 100 lg/mL or higher. Interestingly,
survival assays showed that the proline analogs 1i and 1j along
with motualevic acid E are bacteriostatic to SA and MRSA, while
analogs 1d–h, 1n and motualevic acid A are bactericidal, each at
2–5Â their MICs. Together these results indicate that motualevic
acid A analogs with improved antibacterial potency can be con-
structed, and that their mode of antibacterial activity can be al-
tered with composition. These analogs provide starting points for
synthesis of motualevic acid probes that may be useful in studying
their mechanism/s of action in more detail.
1. Sheehan, J. C. The Enchanted Ring: The Untold Story of Penicillin; The MIT Press:
Boston, 1984.
2. (a) Barbachyn, M. R.; Ford, C. W. Angew. Chem., Int. Ed. 2003, 42, 2010;
Reviewed by (b) Brickner, S. J.; Barbachyn, M. R.; Hutchinson, D. K.; Manninen,
P. R. J. Med. Chem. 2008, 51, 1981.
3. (a) Raja, A.; LaBonte, J.; Lebbos, J.; Kirkpatrick, P. Nat. Rev. Drug Disc. 2003, 2,
943; (b) Alder, J. D. Drugs Today 2005, 41, 81.
4. (a) World Health Organization (WHO) Geneva, World Health Report-2002,
2002.; (b) Nathan, C. Nature 2004, 431, 899; (c) Pinner, R. W.; Teutsch, S. M.;
Simonsen, L. J. Am. Med. Assoc. 1996, 275, 189.
5. (a) Payne, D. J.; Gwynn, M. N.; Holmes, D. J.; Pompliano, D. L. Nat. Rev. Drug Disc.
2007, 6, 29; (b) Levy, S. B.; Marshall, B. Nat. Med. 2004, 10, 122; (c) Walsh, C.;
Wright, G. Chem. Rev. 2005, 105, 391.
6. (a) Singh, S. B.; Barrett, J. F. Biochem. Pharmacol. 2006, 71, 1006; (b) von
Nussbaum, F.; Brands, M.; Hinzen, B.; Weigand, S.; Bich, D. H. Angew. Chem., Int.
Ed. 2006, 45, 5072.
7. Keffer, J. L.; Plaza, A.; Bewley, C. A. Org. Lett. 2009, 11, 1087.
8. Lu, K.; Huang, M.; Xiang, Z.; Liu, Y.; Chen, J.; Yang, Z. Org. Lett. 2006, 8, 1193.
9. Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 13, 3769.
10. López, S.; Fernández-Trillo, F.; Midón, P.; Castedo, L.; Saá, L. J. Org. Chem. 2006,
71, 2802.
11. Barrett, A. G. M.; Head, J.; Smith, M. L.; Stock, N. S.; White, A. J. P.; Williams, D. J.
J. Org. Chem. 1999, 64, 6005.
12. While preparing this manuscript a paper reporting the synthesis of several
motualevic acids was published where Sudhakar and co-workers followed a
similar strategy: Sudhakar, G.; Kadam, V. D.; Reddy, V. V. N. Tetrahedron Lett.
2010, 51, 1124.
13. Ryzhkov, L. R. J. Org. Chem. 1996, 61, 2801.
14. Kigoshi, H.; Suzukib, Y.; Aokib, K.; Uemura, D. Tetradedron Lett. 2000, 41, 3927.
Acknowledgments
We thank J. R. Lloyd and N. Whittaker for mass spectral data.
This work was supported by the NIH Intramural Research Program
(NIDDK) and the Intramural AIDS Targeted Antiviral Program, Of-
fice of the Director, NIH (C.A.B.).