Mechanistic studies of sanguinamide B derivatives: A unique inhibitor of eukaryotic ribosomes

Article abstract of DOI:10.1021/ol401749p

Described are mechanistic studies of two Sanguinamide B (San B) derivatives. These compounds were identified as eukaryotic ribosomal inhibitors. Two biotinylated San B derivatives were synthesized and used to capture protein targets in a pull-down assay.

Full text of DOI:10.1021/ol401749p

ORGANIC
LETTERS
2013
Vol. 15, No. 18
4638–4641
Mechanistic Studies of Sanguinamide
B Derivatives: A Unique Inhibitor
of Eukaryotic Ribosomes
Worawan Tantisantisom, Deborah M. Ramsey, and Shelli R. McAlpine*
School of Chemistry, The University of New South Wales, Kensington,
NSW 2052 Australia
s.mcalpine@unsw.edu.au
Received June 20, 2013
ABSTRACT
Described are mechanistic studies of two Sanguinamide B (San B) derivatives. These compounds were identified as eukaryotic ribosomal
inhibitors. Two biotinylated San B derivatives were synthesized and used to capture protein targets in a pull-down assay. LC/MS/MS analysis of
the San B-captured targets identified several proteins that comprise eukaryotic ribosomal subunits. The translation inhibitory effect of San B was
confirmed using an in vitro translation assay. Moreover, an evaluation of cell death mechanisms is reported.
The first total synthesis of the Sanguinamide B1 (San B; 1)
natural product was described by Singh et al.2 Singh con-
firmed the structure of the natural product (trans, trans-San
B) and identified the thermodynamically stable conformers
(trans, cis-San B and cisꢀcis-San B). Biological evaluation
of San B indicated pilicide activity against Pseudomonas
aeruginosa, which utilizes type IV pili in biofilm formation.3
The structureꢀactivity relationships of the San B analogues
were explored by synthesizing 12 compounds, whereby we
exchanged the valine, alanine, and leucine residues with
N-methyl, glycine, and L- or D-phenylalanine amino acids.4
Most of the analogues synthesized had multiple stable
conformations, and in some cases a specific conformation
exhibited greater bioactivity than the same compound in a
different conformation.4 All 12 San B analogs were tested in
cytotoxicity assays against a human colorectal carcinoma
cell line (HCT-116).5 Our data showed that two D-Phe
analogues (Figure 1; compounds 2 and 3) were cytotoxic
to HCT-116 cells and reduced cell survival by more than
50% at a 50 μM concentration, with IC50 values of 43 ( 4.8
and 38 ( 2.1 μM, respectively.5
Herein we report the synthesis of two biotinylated
analogs of compounds 2 and 3: 2-Tag and 3-Tag. We
show, via pull-down assays, that these two compounds
bind to proteins associated with the 40S (small) and 60S
(large) ribosomal subunits. The translation inhibitory
effect of San B was confirmed using an in vitro transla-
tion assay. Evaluation of cell death mechanisms in San
B-analog treated cells indicates that, in spite of structural
similarities, one compound induced necrosis and one
induced cell death in a manner resembling autophagy.
En route to producing the tagged compound we generated
a structurally related third San B derivative. This molecule
was the most potent of all the San B analogues published
to date, and it induced apoptosis.
Thus, given that none of the other 10 analogs or the
natural product were cytotoxic, we felt that a mechanism
investigation of compounds 2 and 3 would provide inter-
esting data on how the conformation and structure of these
two compounds might dictate their biological activity. We
synthesized two biotinylated derivatives: 2-Tag and 3-Tag
(Figure 1). Our structureꢀactivity relationship (SAR)
data5 indicated that the valine position in the natural
(1) Dalisay, D. S.; Rogers, W. W.; Edison, A. S.; Molinski, T. F.
J. Nat. Prod. 2009, 72, 732.
(2) Singh, E.; Ramsey, D. M.; McAlpine, S. R. Org. Lett. 2012, 14,
1198.
(3) Ramsey, D. M.; Wozniak, D. J. Mol. Microbiol. 2005, 56, 309.
(4) Wahyudi, H.; Tantisantisom, W.; Liu, X.; Ramsey, D. M.; Singh,
E. K.; McAlpine, S. R. J. Org. Chem. 2012, 77, 10596.
(5) McAlpine unpublished result.
r
10.1021/ol401749p
Published on Web 09/03/2013
2013 American Chemical Society

compounds’ protein target(s), following standardized
methods.7 Proteins were separated by electrophoresis and
visualized with Coomassie R-250 staining solution (Figure
S4a).6 Protein sequencing of each band was performed using
trypsin digestion and LC/MS/MS sequencing, followed by
peptide identification using the NCBI Eukaryotic database
and peptide fingerprinting software. The results for com-
pound 3-tag are shown in Table 1, and similar results were
observed for 2-tag, which pulled-down 60S ribosomal pro-
teins, as well as the eukaryotic elongation factor 2 (Table S2).6
Scheme 1. Synthesis of San B Biotinylated Tagged Derivatives
Figure 1. San B natural product (1),1ꢀ2 San B derivatives (2ꢀ3),4
and the biotinylated derivatives, 2-Tag and 3-Tag.
product could be substituted without impacting the mole-
cule’s potency. Thus, we incorporated the tag at this
position for both compounds. En route to both com-
pounds, we synthesized the carboxybenzyl (Cbz) protected
lysine derivatives, where compound 4 is the Cbz lysine
precursor of 2-Tag. Compound 4 was tested for its cyto-
toxicity, and we found that it exhibited an IC50 of 15.9 (
1.3 μM, which was 2.7-fold more potent than its parent
analog scaffold. No cytopathic effect was observed when
cells were treated with a 50 μM concentration of the Cbz
lysine precursor of 3-Tag.6 The synthesis of the Cbz-lysine
derivatives of compounds 2 and 3 were performed as
previously described for other analogues, with the excep-
tion that a Cbz-lysine was incorporated into the macro-
cycle.4 Biotinylation to generate 2-tag and 3-tag was
performed by cleavage of the Cbz protecting group using
33% HBr in acetic acid, yielding free lysine. The lysine was
readily biotinylated using 2.0 equiv of N-hydroxysuccinimide
estersꢀpolyethylene glycol (4 units)ꢀBiotin (NHS-PEG4-
biotin) and 16.0 equiv of N,N-Diisopropylethylamine
(DIPEA) dissolved in dichloromethane to yield biotinylated
San B derivatives (Scheme 1).7 The biotinylated com-
pounds were utilized in pull-down assays to determine the
San B tagged derivatives captured proteins that are
components of ribosomal subunits and play a crucial role
in protein synthesis (translation). Cell growth requires
protein synthesis to increase the overall biomass of the
cell, and malfunction of the translation machinery halts
cell cycle progression and leads to cell death.8 There are
numerous factors that comprise the translation machinery,
includingribosomalproteins(BandsF,I,J,LꢀN,Table1),
translation elongation factors (such as elongation factor 2,
Band C, Table 1), tRNAs (tRNAs), and energy sources.9
Ribosomal proteins are formed from the 40S (small) and
60S (large) ribosomal subunits.10 Our San B 3-tag deriva-
tives bound proteins from the 40S (S11) and the 60S (L3, 7,
12, 18, 23) subunits, whereas 2-tag appears to primarily
interact with 60S ribosomal subunits (L9, 10, 13, 17).
Table 1. Protein Sequencing Results from Pull-down Assays
with Compound 3-taga
(6) See Supporting Information for details.
(7) (a) Vasko, R. C.; Rodriguez, R. A.; Cunningham, C. N.; Ardi,
V. C.; Agard, D. A.; McAlpine, S. R. ACS Med. Chem. Lett. 2010, 1, 4.
(b) Ardi, V. C.; Alexander, L. D.; Johnson, V. A.; McAlpine, S. R. ACS
Chem. Biol. 2011, 6, 1357. (c) Kunicki, J. B.; Petersen, M. N.; Alexander,
L. D.; Ardi, V. C.; McConnell, J. R.; McAlpine, S. R. Bioorg. Med.
Chem. Lett. 2011, 21, 4716. (d) Ramsey, D., M.; McConnell, J., R.;
Alexander, L., D.; Tanaka, K., W.; Vera, C., M. ; Mcalpine, S., R.
Bioorg. Med. Chem. Lett. 2012, 22, 3287.
mass
MASCOT
score
band
protein
(kDa)
C
F
I
elongation factor 2
95.2
45.4
29.2
20.7
18.4
17.8
14.8
362
583
386
400
304
320
278
60S ribosomal proteins L3
60S ribosomal protein L7
60S ribosomal protein L18a
40S ribosomal protein S11
60S ribosomal protein L12
60S ribosomal protein L23
J
(8) (a) Fingar, D. C.; Blenis, J. Oncogene 2004, 23, 3151. (b) Naora,
H. Immunol. Cell Biol. 1999, 77, 197.
L
M
N
(9) (a) Marintchev, A.; Wagner, G. Q. Rev. Biophys. 2004, 37, 197.
(b) Agirrezabala, X.; Frank, J. Q. Rev. Biophys. 2009, 42, 159.
(10) (a) Panse, V. G.; Johnson, A. W. Trends Biochem. Sci. 2010, 35,
260. (b) Voigts-Hoffmann, F.; Klinge, S.; Ban, N. Curr. Opin. Struct.
Biol. 2012, 22, 768.
a See Supporting Information for all pull-down data and gels.
Org. Lett., Vol. 15, No. 18, 2013
4639

The similarity in the proteins isolated by both compounds
imply that the cytosolic targets of Sanguinamide B are
ribosomes.
To verify that compounds 2, 3, and 4 act as translation
inhibitors, we utilized a cell-free in vitro translation assay.
Measuring protein synthesis in the presence of the deriva-
tives, as well as comparing the results to a commercially
available translational inhibitor (G418),11 provides evi-
dence on whether our molecules inhibit protein synthesis
(Figure 2a). The concentrations of compounds used in the
assays were based on the IC50 values of the compounds.
Using HCT-116 cell lysate as the ribosomal source, we
added accessory proteins and monitored the translation
of active green fluorescent protein (GFP). The amount of
GFP was measured and quantitated by flow cytometry.
Relative fluorescence units (RFU) for each reaction were
normalized to a positive control, which contained the GFP
reporter system but no compounds or translational inhi-
bitors (Figure 2a). Comparing compounds 2, 3, and 4 to
the positive control, we observe a significant reduction in
translation of GFP compared to the positive control.
Compound 2 (50 μM) reduces translation by 4.5-fold
(22 ( 2.4% RFU; p < 0.0007), and compound 3
(50 μM) by 5.6-fold (18 ( 0.5% RFU; p < 0.0001).
Compound 4 (20 μM) reduced translation by 4.5-fold,
but with 2.5 times less compound (22 ( 3.9% RFU; p <
0.03). The reduction in GFP translation observed with San
B analogues was similar to that observed with the transla-
tion inhibitor G418 (5 μM) (14 ( 1.4% RFU; p < 0.0023).
Although the potency of San B derivatives must be opti-
mized, our results show that the D-Phe analogues of San B
inhibit in vitro translation by 80% at concentrations
similar to their IC50 values.
Protein synthesis, cell growth, and cell cycle progression
are tightly coupled events.8a Many ribosomal proteins
contribute extra ribosomal functions that can affect both
gene transcription and translation.12 Protein L7 (from the
60S subunit) is overexpressed in colorectal cancers, causes
cell cycle arrest, and induces apoptosis.8a,13 Our data show
that all three compounds (2, 3, and 4) are cytotoxic to the
colon cancer cell line and they bind to ribosomal subunits
including L7.
Thus, we anticipate that the San B derivatives would
halt DNA synthesis and induce apoptosis. To measure
the impact of the San B analogues on DNA synthesis, we
measured BrdU incorporation in HCT-116 cells when
treated with 2, 3, or 4 (Figure S3).6 Our data suggest that
compound 4 has a significant inhibitory effect on DNA
synthesis. To analyze the mechanism of growth inhibition,
we examined cell morphology, caspase 3/7 activity
(Figure S2),6 and cell permeability of HCT-116 cells
Figure 2. (a) Cell-free based in vitro translation assay. Positive
(Fully translated proteins), Negative (No protein production),
G418 (5 μM), 2 (50 μM), 3 (50 μM), 4 (20 μM). The average of
five independent experiments are shown. (b) Cell membrane
permeability of HCT-116 cells after 48 h of treatment, as
measured by 7-AAD uptake. Compound concentrations are
identical to those used in part a.
treated with San B derivatives. Treatment of cells with
40 μM of 2 induced the formation of sickle-shaped nuclei
in treated cells, a modest increase in caspase 3 activity
over the DMSO control, and 23% of 2-treated cells
were permeable as measured by 7-AAD incorporation
(Figures 2b, 3C, and S2). These results point to auto-
phagy as the primary mechanism for 2-mediated growth
inhibition.
Autophagy is a cellular defense mechanism whereby cells
degrade cytoplasmic molecules and misfolded proteins
when the proteasome degradation pathway is compro-
mised and/or the endoplasmic reticulum is undergoing
stress.14 Cellular components are degraded in large auto-
phagic vacuoles called autophagosomes, and cells can
utilize this mechanism to replenish amino acids and main-
tain homeostasis. There is a degree of overlap between the
pathways that control autophagy and those that induce
apoptosis, and caspase 3 can be detected in autophagic
cells.15 Since 2 binds to ribosomal proteins and inhibits
translation, we believe that 2 causes endoplasmic reticulum
stress and leads to the induction of autophagy within 48 h.
Treatment of cells with 40 μM of 3 induced 51% cell per-
meability over the DMSO control as measured by 7-AAD
staining (Figure 2b). Interestingly, the cell morphology
(11) Eustice, D. C.; Wilhelm, J. M. Antimicrob. Agents Chemother.
1984, 26, 53.
€
(12) Lindstrom, M. S. Biochem. Biophys. Res. Commun. 2009, 379,
167.
(14) Zheng, H.-y.; Zhang, X.-y.; Wang, X.-f.; Sun, B.-c. Differentia-
tion 2012, 4, 11.
(15) Bursch, W.; Karwan, A.; Mayer, M.; Dornetshuber, J.;
Frohwein, U.; Schulte-Hermann, R.; Fazi, B.; Di Sano, F.; Piredda,
L.; Piacentini, M.; Petrovski, G.; Fesus, L.; Gerner, C. Toxicology 2008,
254, 147.
(13) (a) von Mikecz, A.; Hemmerich, P.; Peter, H.-H.; Krawinkel, U.
Immunobiology 1994, 192, 137. (b) Kasai, H.; Nadano, D.; Hidaka, E.;
Higuchi, K.; Kawakubo, M.; Sato, T. A.; Nakayama, J. J. Histochem.
Cytochem. 2003, 51, 567. (c) Neumann, F.; Hemmerich, P.; Vonmikecz,
A.; Peter, H. H.; Krawinkel, U. Nucleic Acids Res. 1995, 23, 195.
4640
Org. Lett., Vol. 15, No. 18, 2013

Figure 3. Morphological analysis of HCT-116 cells treated for 48 h with San B derivatives or DMSO. Confocal microscopy (100ꢁ)
images show DNA staining (DAPI; A, C, E, G) or the merged image (DAPIþDCI; B, D, F, H). White arrows indicate cell rounding.
Yellow arrows indicate sickle-shaped nuclei. White bars indicate 40 μm.
data (Figure 3E and F) indicated no cell rounding or DNA
condensation at 48 h after treatment, although 51% of the
cells are permeable. These data indicate that 3 induces
necrosis in treated cells, and this mode of cell death
predominates in 3-treated cells.
mechanism of growth inhibition induced by each deriva-
tive is dependent on the position of the D-Phe. Ribosome
binding and inhibition have been seen in other thiazole-
containing macrocyclic peptides, including thiocillin and
Marthiapeptide A.17 Thus, our data are consistent with
those of others, although Sanguinamide B contains the
unique structural feature of a 4,2-bisheterocycle tandem,
with an oxazole directly linked to a thiazole. Placing the
D-Phe between the valine and thiazole togenerate2 inhibits
translation, causes endoplasmic reticulum stress, and leads
to the induction of autophagy. In contrast, placing the
D-Phe substituent between the thiazole and proline 2 induces
a Type II0 β-turn, which leads to necrosis in compound 3
treated cells. Finally, compound 4, which contains the
D-Phe between the Cbz-lysine and thiazole, induces apop-
tosis via a caspase 3 pathway. We suspect that compounds 2
and 4 bind to the same site on the ribosomal machinery, but
4 is more effective at inhibiting translation because of the
Cbz-lysine moiety. Studies are underway that will optimize
compound 4 as a lead structure for inhibiting the ribosomal
translation machinery.
We hypothesized that 3 may induce a Type II0 β-turn
that would allow the nonpolar face of the molecule (valine,
alanine, thiazole) to insert itself into the lipid bilayer of the
plasma membrane and enhance membrane permeability.
The Type II0 β-turn is a key feature of Gramicidin S, which
is an amphiphilic molecule. Gramicidin S forms an anti-
parallel β-sheet that inserts itself into lipid bilayers.16
The high level of cell permeability induced after treatment
with 3, combined with the lack of cell rounding or DNA
condensation observed in apoptotic cells, would strongly
support our hypothesis that 3 induces necrosis by a
mechanism that is similar to Gramicidin S.
Treatment of HCT-116 cells with 10 μM of 4 induced a
2-fold increase in caspase 3/7 activity over the DMSO
control (Figure S2).6 Cells treated with 4 (10 μM) or the
pro-apoptotic molecule 17-AAG (1 μM) were round in ap-
pearance, with condensed chromatin (compare Figure 3G
and H with Figure S1). In addition, 47% of 4 treated cells
werepermeable, asmeasuredby incorporationof 7-amino-
actinomycin D (7-AAD) (Figure 2b). These results, in
combination with the cell morphology data, indicate that
compound 4 induces apoptosis by a caspase 3 pathway. The
large degree of membrane permeability suggests that the
majority of 4-treated cells are undergoing secondary necrosis
at 48 h, which occurs after the induction of apoptosis.
In conclusion, our data indicate that the D-Phe
analogues of San B bind to ribosomal subunits, yet the
Acknowledgment. We thank C. Brownlee and the Bio-
logical Resources Imaging Laboratory(BRIL) and UNSW
for support of D.M.R., and S.R.M.
Supporting Information Available. Synthetic proce-
dures, characterization data, and biological assays. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(17) (a) Zhou, X.; Huang, H.; Chen, Y.; Tan, J.; Song, Y.; Zou, J.;
Tian, X.; Hua, Y.; Ju, J. J. Nat. Prod. 2012, 75, 2251. (b) Aulakh, V. S.;
Ciufolini, M. A. J. Am. Chem. Soc. 2011, 133, 5900.
(16) Prenner, E. J.; Lewis, R. N.; McElhaney, R. N. Biochim.
Biophys. Acta 1999, 1462, 201.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 18, 2013
4641