Synthesis and biological evaluation of truncated derivatives of abyssomicin C as antibacterial agents

Article abstract of DOI:10.3762/bjoc.15.147

The synthesis and antibacterial activity of two new highly truncated derivatives of the natural product abyssomicin C are reported. This work outlines the limits of structural truncation of the natural product and consequently provides insights for furthe

Full text of DOI:10.3762/bjoc.15.147

Synthesis and biological evaluation of truncated derivatives
of abyssomicin C as antibacterial agents
Leticia Monjas1,2, Peter Fodran1,2, Johanna Kollback1, Carlo Cassani1,2,3,
Thomas Olsson1, Maja Genheden2,4, D. G. Joakim Larsson2,4
and Carl-Johan Wallentin*1,2
Letter
Open Access
Address:
Beilstein J. Org. Chem. 2019, 15, 1468–1474.
1Department of Chemistry and Molecular Biology, University of
Gothenburg, Kemigården 4, 412 96, Gothenburg, Sweden, 2Centre
for Antibiotic Resistance Research (CARe), University of Gothenburg,
Gothenburg, Sweden, 3current affiliation: Hit Discovery, Discovery
Sciences, R&D BioPharmaceuticals, AstraZeneca, Gothenburg,
Sweden and 4Department of Infectious Diseases, Institute of
Biomedicine, The Sahlgrenska Academy at the University of
Gothenburg, Guldhedsgatan 10, 413 46, Gothenburg, Sweden
doi:10.3762/bjoc.15.147
Received: 05 March 2019
Accepted: 18 June 2019
Published: 02 July 2019
The article is dedicated to Dr. Fredrik Pettersson who unfortunately
passed away recently.
Email:
Associate Editor: A. Kirschning
Carl-Johan Wallentin* - carl.wallentin@chem.gu.se
© 2019 Monjas et al.; licensee Beilstein-Institut.
License and terms: see end of document.
* Corresponding author
Keywords:
abyssomicin C; antibiotic; antibiotic resistance; MRSA; truncated
natural products
Abstract
The synthesis and antibacterial activity of two new highly truncated derivatives of the natural product abyssomicin C are reported.
This work outlines the limits of structural truncation of the natural product and consequently provides insights for further
structure–activity relationship studies towards novel antibiotics targeting 4-amino-4-deoxychorismate (ADC) synthase. Specifi-
cally, it is demonstrated that the synthetically challenging bicyclic motif is essential for activity towards methicillin-resistant
Staphylococcus aureus (MRSA).
Introduction
Antibiotic resistance is a major threat to global health, and nu- appropriate physicochemical properties that NPs typically show
merous actions are presently taken in both academia and [3]. Abyssomicin C (AbC) is an NP with antibacterial activity
industry to combat this major societal challenge. There are that was isolated from the marine actinomycete strain Verruco-
many different approaches to develop new antibiotics, but the sispora AB-18-032 in 2004 [4,5]. It shows antibacterial activity
reevaluation of known natural products (NP) and the identifica- against Gram-positive bacteria, including resistant pathogens
tion of new NPs and derivatives thereof is considered an espe- such as methicillin- and vancomycin-resistant Staphylococcus
cially important pathway in the search for more robust antibiot- aureus (MRSA and VRSA) with minimum inhibitory concen-
ics [1,2]. The main reasons are the inherent high selectivity and tration (MIC) values of 4 and 13 µg/mL, respectively [4]. AbC
1468

Beilstein J. Org. Chem. 2019, 15, 1468–1474.
belongs to a small family of NPs and two congeners of this Because of its intriguing structure and antibacterial activity,
family lacking the enone functionality, abyssomicins B and D, several synthetic studies [11-14], formal [15] and total
were discovered at the same time, and these compounds did not [7-9,16,17] syntheses of AbC and its slightly more active atro-
show any antibacterial activity (Figure 1). Consequently, the pisomer (atrop-AbC) have been reported in the last ten years.
enone moiety of AbC was proposed to play an essential role in Due to its complex structure, these total syntheses are long and
the antibiotic activity. Indeed, in a subsequent study it was thus not compatible with further drug development studies.
shown that AbC has a unique mechanism of action amongst
NPs: it is a covalent inhibitor of 4-amino-4-deoxychorismate An advantageous approach to strategically truncate complex
(ADC) synthase, which is the enzyme that catalyzes the conver- molecules and thereby simplify their synthesis is referred to as
sion of chorismate and glutamine into ADC and glutamate, the function-oriented synthesis (FOS) [18]. This strategy allows for
first step in the biosynthesis of p-aminobenzoic acid (PABA) in a more rapid drug development process, and it has previously
bacteria [6]. Specifically, AbC binds via a Michael addition be- been applied to AbC. Accordingly, structure–activity relation-
tween a cysteine residue in the immediate proximity to the ship (SAR) studies have been performed, which indicate that
active site and the enone moiety of AbC. The PABA pathway is benzylation of the hydroxy group and the complete demethyla-
essential in bacteria but absent in humans, making AbC a prom- tion of the core structure of AbC retain the antibiotic activity
ising compound for further development towards an antibiotic towards MRSA while simultaneously decrease cytotoxicity
drug candidate.
towards various human cell lines [9,10].
Figure 1: Structure of abyssomicins C, B, D, atrop-abyssomicin C, atrop-O-benzyl-abyssomicin C and atrop-O-benzyl-desmethyl-abyssomicin C with
their biological activities, and the truncated derivatives 1 and 2. In red: part of abyssomicin C that is maintained in the truncated derivatives [4,7-10].
1469

Beilstein J. Org. Chem. 2019, 15, 1468–1474.
Based on these observations, and in line with the principles of modeling. Accordingly, the crystal structure of ADC synthase
FOS, we hypothesized that further simplification could be com- (PDB ID: 1K0E) [19] was used for docking of known and pro-
patible with retained activity and consequently also a more posed ligands. The crystal structure contains a tryptophan mole-
rapid and exhaustive SAR exploration. Specifically, we identi- cule in the active site. Restrained molecular dynamics [20,21]
fied truncated derivative 1 as the most simplified structure that was employed to position the active site cysteine (Cys-263) in a
still includes all structural features that have been previously position that would allow covalent binding of the ligands in the
found to be essential for activity. This truncated derivative still active site. The resulting protein conformation was allowed to
has the tetronate functionality and the enone-equipped 11-mem- relax to ensure that the found conformation was an energy
bered ring system but lacks the oxabicyclo[2.2.2]octane unit. In minimum. This protein structure was then used for docking of
addition to a simplified core structure, 1 only has one stereo- ligands using Glide [22-24]. Glide XP docking positioned both
center as compared to seven in AbC. Furthermore, the benzyl- AbC and the analog atrop-O-benzyl-desmethyl-abyssomicin C
ated derivative 2 was also envisioned to mimic the favorable with low energy conformations and with expected interaction
toxicity profile observed for atrop-O-benzyl-desmethyl- points in the binding site, providing suitable docking scores (see
abyssomicin C (Figure 1) [10]. Here we present the synthesis Supporting Information File 1). The key interactions include a
and biological evaluation of 1 and 2, the hitherto most trun- hydrogen bond to the backbone of Arg-45 and lipophilic inter-
cated derivatives of AbC.
actions in the deep pocket defined by Phe-241 and Leu-34
(Figure 2). Our compound 2 docks in a similar way, filling the
same subpocket, maintaining the key interactions and posi-
Results and Discussion
An initial evaluation of the truncation strategy was made by tioning the Cys-263 near the reacting double bond (Figure 2A).
assessing the potential efficiency of binding to the target by The additional ring system of atrop-O-benzyl-desmethyl-
Figure 2: A) Docking poses for atrop-O-benzyl-desmethyl-abyssomicin C (orange) and compound 2 (green) showing key interactions and reasonable
positioning for a reaction with Cys263. B) Ligand interaction diagrams for atrop-O-benzyl-desmethyl-abyssomicin C (left) and compound 2 (right),
showing similar interactions and similar shape.
1470

Beilstein J. Org. Chem. 2019, 15, 1468–1474.
abyssomicin C does not seem to make any significant interac- We synthesized the common building block 3 in six steps
tion with the protein, which suggests a suitable level of struc- (Scheme 2), starting with the monoprotection of 1,5-pentane-
tural truncation of compound 2.
diol (6), followed by Swern oxidation, which afforded 7 in an
overall yield of 57% [26]. The so obtained aldehyde 7 was
Further docking studies, using covalent docking, also showed treated with vinylmagnesium bromide in the presence of cerium
that both atrop-O-benzyl-desmethyl-abyssomicin C and 2 can chloride to afford 8 in 89% yield. An overall translocation of
bind to the active site cysteine via a Michael addition to the α,β- the TBDMS group from the primary to the secondary alcohol
unsaturated ketone, still maintaining the benzyl group in the was achieved by a protection/deprotection sequence, which
favorable subpocket position (Figure S1 in Supporting Informa- smoothly provided alcohol 10 in 71% yield over two steps. A
tion File 1).
final Swern oxidation gave the building block 3 in 91% yield
(33% overall yield from 1,5-pentanediol 6).
Synthesis
The syntheses of 1 and 2 were inspired, in part, by previous The other key intermediates, building blocks 4 and 5 (4, R = Me
strategies employed in the total synthesis of AbC [7-9,16,17]. and 5, R = CH2CH2Ph) were synthesized starting with allyl-
Hence, the construction of the 11-membered ring was based on dioxazaborolidine 11, an allyl-transfer reagent that was pre-
a carbonyl addition reaction between aldehyde 3 and the pared as previously reported (Scheme 3) [27]. Allylation of
tetronate derivatives 4 and 5, followed by a ring-closing metath- methyl pyruvate (12) or 13 (synthesized from dimethyl oxalate
esis (Scheme 1). Since both enantiomers of AbC have similar and phenethylmagnesium bromide, see Supporting Information
activity [25], we pursued racemic synthesis of the targeted com- File 1) using 11 in the presence of trifluoroacetic acid afforded
pounds.
the corresponding homoallylic alcohol 14 and 15 in 84% and
92% yield, respectively. The next steps involved acylation of
alcohol 14 and 15 using bromoacetyl bromide in refluxing tolu-
ene, which afforded the corresponding α-bromo esters 16 and
17, in 68% and 72% yield, respectively. Finally, intramolecular
Wittig reaction of 16 or 17 in the presence of N,N-diisopropyl-
ethylamine and triphenylphosphine provided building blocks 4
and 5 in 75% and 77% yield (43% and 51% overall yields from
allyldioxazaborolidine 11), respectively.
The tetronate building blocks 4 and 5 were further converted to
the corresponding enolates using lithium diisopropylamide and
allowed to react with the common key intermediate 3 to afford
the hydroxyalkylated products 18 and 19, in 78% and 59%
yield, respectively (Scheme 4). Treatment of 18 and 19 with
Scheme 1: Retrosynthetic analysis of the truncated derivatives 1 and
2 of abyssomicin C.
Scheme 2: Synthesis of the building block 3. a) TBDMSCl (0.5 equiv), NaH (1.0 equiv), THF, 0 °C to rt, 16 h, 82%. b) i: (COCl)2 (1.2 equiv), DMSO
(2.4 equiv), CH2Cl2, −78 °C, 1.5 h; ii: Et3N (5.0 equiv), −78 °C to rt, 1 h, 70%. c) vinylmagnesium bromide (1.4 equiv), CeCl3 (1.4 equiv), THF,
−78 °C, 2 h, then rt, 30 min, 89%. d) TBDMSCl (2.0 equiv), imidazole (4.0 equiv), DMF, rt, 2 h, 95%. e) HF·pyridine, pyridine, THF, rt, 45 h, 75%.
f) i: (COCl)2 (1.2 equiv), DMSO (2.4 equiv), CH2Cl2, −78 °C, 1.5 h; ii: Et3N (5.0 equiv), −78 °C to rt, 1 h, 91%.
1471

Beilstein J. Org. Chem. 2019, 15, 1468–1474.
Hoveyda–Grubbs second generation catalyst in refluxing 1,2-
dichloroethane afforded the ring-closing metathesis products 20
and 21 in 62% and 45% yield, respectively. Deprotection of the
TBDMS group using tetrabutylammonium fluoride followed by
final Dess–Martin oxidation gave the targeted compounds 1 and
2 in 17% and 30% yield over the two steps, respectively. For
the final step, several oxidation agents were evaluated, and the
highest yields were achieved with Dess–Martin periodinane,
which converted 22 and 23 to 1 and 2 in 24% and 41% yield,
respectively.
Biological evaluation
The antibacterial activity of the truncated AbC derivatives 1 and
2, as well as the tetronate building blocks 4 and 5, was evalu-
ated against S. aureus and two different strains of MRSA using
vancomycin as a positive control (Table 1).
Scheme 3: Synthesis of the building blocks 4 and 5. a) TFA
(1.1 equiv), CH2Cl2, rt, 24 h, 84% (14: R = Me) and 92% (15: R =
CH2CH2Ph). b) bromoacetyl bromide (1.5–2.4 equiv), toluene, reflux,
20–24 h, 68% (16: R = Me) and 72% (17: R = CH2CH2Ph). c) PPh3
(1.5 equiv), DIPEA (1.2 equiv), THF, 70 °C, 16 h, 75% (4: R = Me) and
77% (5: R = CH2CH2Ph).
Table 1: Minimum inhibitory concentrations (MIC) of compounds 1, 2,
4 and 5 against S. aureus and two strains of MRSA.
Compound
MIC (µg/mL)
S. aureus
MRSA
MRSA
CCUG15915 CCUG58065 CCUG38266
vancomycin
1
>200
150
>200
>200
1
>200
150
>200
>200
0.5
>200
100
>200
>200
1
2
4
5
It has been previously shown that structural prerequisites for
truncated derivatives of AbC to harbor activity against Gram-
positive bacteria are that the 11-membered ring is intact and
that this macrocycle is equipped with an enone functionality
[4,5]. In accordance with these previous findings, compounds 4
and 5 were found not to inhibit the growth of any of the bacteri-
al strains investigated. Likewise, and to our dismay, the targeted
AbC derivative 1 did not show any inhibition either. Only
compound 2 showed a low inhibition of bacterial growth
(100–150 μg/mL), which, taken together, suggests that the
oxabicyclo[2.2.2]octane moiety from the original structure of
AbC is crucial for antibacterial activity. The activity difference
between compounds 1 and 2 could be explained by the pre-
dicted more efficient binding of 2 with the active site. However,
such a conclusion would be highly speculative based solely on
these results. Due to the low activity of compound 2, no further
investigations related to the conserved mode of action or toxici-
ty were performed.
Scheme 4: Reaction of the building block 3 with 4 or 5 for the synthe-
sis of compounds 1 and 2. a) 4 or 5 with LDA (1.5 equiv), THF,
−78 °C, 30 min; then 3, −78 °C, 1.5 h; 78% (18: R = Me) and 59%
(19: R = CH2CH2Ph). b) Hoveyda–Grubbs second generation catalyst
(5 mol %), 1,2-dichloroethane (0.002 M), reflux, 1 h, 62% (20: R = Me)
and 45% (21: R = CH2CH2Ph). c) TBAF (10 equiv), THF, rt, 16 h, 73%
(22: R = Me) and 74% (23: R = CH2CH2Ph). d) Dess–Martin periodi-
nane (2.5 equiv), CH2Cl2, rt, 2 h, 24% (1: R = Me) and 41% (2: R =
CH2CH2Ph).
Conclusion
We have described the synthesis and antibacterial activity of
two new truncated derivatives of AbC. Previous work indicated
1472

Beilstein J. Org. Chem. 2019, 15, 1468–1474.
2. Rossiter, S. E.; Fletcher, M. H.; Wuest, W. M. Chem. Rev. 2017, 117,
12415–12474. doi:10.1021/acs.chemrev.7b00283
3. Wright, G. D. Nat. Prod. Rep. 2017, 34, 694–701.
doi:10.1039/c7np00019g
4. Riedlinger, J.; Reicke, A.; Zähner, H.; Krismer, B.; Bull, A. T.;
Maldonado, L. A.; Ward, A. C.; Goodfellow, M.; Bister, B.; Bischoff, D.;
Süssmuth, R. D.; Fiedler, H.-P. J. Antibiot. 2004, 57, 271–279.
doi:10.7164/antibiotics.57.271
5. Bister, B.; Bischoff, D.; Ströbele, M.; Riedlinger, J.; Reicke, A.;
Wolter, F.; Bull, A. T.; Zähner, H.; Fiedler, H.-P.; Süssmuth, R. D.
Angew. Chem., Int. Ed. 2004, 43, 2574–2576.
doi:10.1002/anie.200353160
6. Keller, S.; Schadt, H. S.; Ortel, I.; Süssmuth, R. D.
Angew. Chem., Int. Ed. 2007, 46, 8284–8286.
doi:10.1002/anie.200701836
7. Nicolaou, K. C.; Harrison, S. T. Angew. Chem., Int. Ed. 2006, 45,
3256–3260. doi:10.1002/anie.200601116
8. Nicolaou, K. C.; Harrison, S. T. J. Am. Chem. Soc. 2007, 129,
429–440. doi:10.1021/ja067083p
9. Bihelovic, F.; Karadzic, I.; Matovic, R.; Saicic, R. N.
Org. Biomol. Chem. 2013, 11, 5413–5424. doi:10.1039/c3ob40692j
10.Matovic, R.; Bihelovic, F.; Gruden-Pavlovic, M.; Saicic, R. N.
Org. Biomol. Chem. 2014, 12, 7682–7685. doi:10.1039/c4ob01436g
11.Snider, B. B.; Zou, Y. Org. Lett. 2005, 7, 4939–4941.
doi:10.1021/ol0518941
12.Rath, J.-P.; Eipert, M.; Kinast, S.; Maier, M. E. Synlett 2005, 314–318.
doi:10.1055/s-2004-837208
13.Rath, J.-P.; Kinast, S.; Maier, M. E. Org. Lett. 2005, 7, 3089–3092.
doi:10.1021/ol0511068
that the three methyl groups of AbC were not required for activ-
ity [10]. In this work, we went one step further with the design
of two new truncated AbC derivatives, in which besides
removing the three methyl groups, the oxabicyclo[2.2.2]octane
structure was also subjected to truncation. We found that the
antibacterial activity is either decreased by one order of magni-
tude or even completely lost, given that the MIC values of the
two new derivatives are ≥100 µg/mL. These unexpectedly low
activities could possibly be rationalized from the initial
modeling studies, which predicted a longer distance from the
reacting cysteine to the electrophilic carbon associated with the
enone functionality for compound 2 than for the active atrop-O-
benzyl-desmethyl-abyssomicin C (Figure 2A). This difference
is not reflected in the docking scores, however, it is still indica-
tive of a less favorable binding for 2 when compared to atrop-
O-benzyl-desmethyl-abyssomicin C. Even though these results
are somewhat discouraging, together with previous studies they
clearly outline the upper level of truncation that is tolerated for
further SAR studies of AbC derivatives. As such, this work
demonstrates that continued efforts towards drug development
based on AbC should focus on derivatives that include struc-
tural motifs close to the oxabicyclo[2.2.2]octane with substitu-
ents that can more systematically explore interactions with the
deep channel of the ADC synthase.
14.Zografos, A. L.; Yiotakis, A.; Georgiadis, D. Org. Lett. 2005, 7,
4515–4518. doi:10.1021/ol051872e
15.Couladouros, E. A.; Bouzas, E. A.; Magos, A. D. Tetrahedron 2006, 62,
5272–5279. doi:10.1016/j.tet.2006.01.106
16.Zapf, C. W.; Harrison, B. A.; Drahl, C.; Sorensen, E. J.
Angew. Chem., Int. Ed. 2005, 44, 6533–6537.
doi:10.1002/anie.200502119
17.Bihelovic, F.; Saicic, R. N. Angew. Chem., Int. Ed. 2012, 51,
5687–5691. doi:10.1002/anie.201108223
Supporting Information
Supporting Information File 1
Experimental part (modelling and docking, synthesis and
biological evaluation), and copies of NMR spectra.
[https://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-15-147-S1.pdf]
18.Wender, P. A.; Verma, V. A.; Paxton, T. J.; Pillow, T. H.
Acc. Chem. Res. 2008, 41, 40–49. doi:10.1021/ar700155p
19.Parsons, J. F.; Jensen, P. Y.; Pachikara, A. S.; Howard, A. J.;
Eisenstein, E.; Ladner, J. E. Biochemistry 2002, 41, 2198–2208.
doi:10.1021/bi015791b
20.Pronk, S.; Páll, S.; Schulz, R.; Larsson, P.; Bjelkmar, P.; Apostolov, R.;
Shirts, M. R.; Smith, J. C.; Kasson, P. M.; van der Spoel, D.; Hess, B.;
Lindahl, E. Bioinformatics 2013, 29, 845–854.
Acknowledgements
The Olle Engkvist Byggmästare foundation, the Centre for
Antibiotic Resistance Research at the University of Gothen-
burg (CARe) and the Swedish Research Council are gratefully
acknowledged for financial support. The Carl-Tryggers founda-
tion is acknowledged for a scholarship for P.F. The European
Commission is acknowledged for a Marie Skłodowska-Curie
fellowship (H2020-MSCA-IF-2014n:660668) for C.C.
doi:10.1093/bioinformatics/btt055
21.Abraham, M. J.; Murtola, T.; Schulz, R.; Páll, S.; Smith, J. C.; Hess, B.;
Lindahl, E. SoftwareX 2015, 1–2, 19–25.
doi:10.1016/j.softx.2015.06.001
22.Schrödinger, Release 2017-2; Schrödinger, LLC: New York, NY, 2017.
23.Friesner, R. A.; Murphy, R. B.; Repasky, M. P.; Frye, L. L.;
Greenwood, J. R.; Halgren, T. A.; Sanschagrin, P. C.; Mainz, D. T.
J. Med. Chem. 2006, 49, 6177–6196. doi:10.1021/jm051256o
24.Zhu, K.; Borrelli, K. W.; Greenwood, J. R.; Day, T.; Abel, R.;
Farid, R. S.; Harder, E. J. Chem. Inf. Model. 2014, 54, 1932–1940.
doi:10.1021/ci500118s
25.Freundlich, J. S.; Lalgondar, M.; Wei, J.-R.; Swanson, S.;
Sorensen, E. J.; Rubin, E. J.; Sacchettini, J. C. Tuberculosis 2010, 90,
298–300. doi:10.1016/j.tube.2010.08.002
ORCID® iDs
Leticia Monjas - https://orcid.org/0000-0003-0624-9934
Peter Fodran - https://orcid.org/0000-0003-0348-8331
Carl-Johan Wallentin - https://orcid.org/0000-0003-1983-9378
References
1. Moloney, M. G. Trends Pharmacol. Sci. 2016, 37, 689–701.
doi:10.1016/j.tips.2016.05.001
1473

Beilstein J. Org. Chem. 2019, 15, 1468–1474.
26.Frankowski, K. J.; Golden, J. E.; Zeng, Y.; Lei, Y.; Aubé, J.
J. Am. Chem. Soc. 2008, 130, 6018–6024. doi:10.1021/ja800574m
27.Reilly, M. K.; Rychnovsky, S. D. Org. Lett. 2010, 12, 4892–4895.
doi:10.1021/ol1020515
License and Terms
This is an Open Access article under the terms of the
Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0). Please note
that the reuse, redistribution and reproduction in particular
requires that the authors and source are credited.
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
(https://www.beilstein-journals.org/bjoc)
The definitive version of this article is the electronic one
which can be found at:
doi:10.3762/bjoc.15.147
1474