Synthesis and evaluation of locostatin-based chemical probes towards PEBP-proteins

Article abstract of DOI:10.1016/j.tetlet.2016.04.071

Phosphatidyl ethanolamine-binding proteins (PEBPs) are implicated in various critical physiological processes in all eukaryotes. Among them is Flowering Locus T (FT), the protein recently discovered as the vital flowering hormone in plants. Small molecule

Full text of DOI:10.1016/j.tetlet.2016.04.071

Tetrahedron Letters 57 (2016) 2406–2409
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Synthesis and evaluation of locostatin-based chemical probes
towards PEBP-proteins
Jorin Hoogenboom a, Martijn Fiers b, Richard Immink b, Han Zuilhof a,c, Tom Wennekes a,d,
⇑
a Laboratory of Organic Chemistry, Wageningen University, Dreijenplein 8, Wageningen 6703 HB, The Netherlands
b Plant Research International, Bioscience, Droevendaalsesteeg 1, Wageningen 6708 PB, The Netherlands
c Department of Chemical and Materials Engineering, King Abdulaziz University, Jeddah, Saudi Arabia
d Department of Chemical Biology and Drug Discovery, Utrecht Institute for Pharmaceutical Sciences and Bijvoet Center for Biomolecular Research, Utrecht University, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
Phosphatidyl ethanolamine-binding proteins (PEBPs) are implicated in various critical physiological pro-
cesses in all eukaryotes. Among them is Flowering Locus T (FT), the protein recently discovered as the
vital flowering hormone in plants. Small molecule inhibitors and activators of FT could provide control
over plant flowering and are therefore an interesting target for industrial agriculture. No small molecule
inhibitors or activators are known for FT, but for a structurally similar PEBP, RKIP, an inhibitor called loco-
statin has been reported to covalently bind in the RKIP ligand binding pocket. Herein, we report the syn-
thesis of novel locostatin-based chemical PEBP probes and evaluate their ability and selectivity towards
the binding of FT and RKIP.
Received 18 March 2016
Revised 15 April 2016
Accepted 19 April 2016
Available online 20 April 2016
Keywords:
Inhibitor
Chemical probes
Locostatin
PEBP
Ó 2016 Elsevier Ltd. All rights reserved.
RKIP
Flowering Locus T
The phosphatidyl ethanolamine-binding protein (PEBP) family
compromises over 400 members from a variety of organisms,
including plants, bacteria, yeast and mammals.1 The PEBP family
is thus a highly conserved group of proteins, that fulfil a plethora
of different functions. Recently, a new functional addition to the
PEBP family was made due to the search towards ‘florigen’. Flori-
gen is a crucial, long sought substance that relays a signal from
the leaves to the growing tip of a plant, known as the shoot apical
meristem, to induce flowering. The extensive hunt for florigen has
led to the discovery of Flowering Locus T (FT) as the flowering
inducing agent belonging to the PEBP family of proteins in the
model plant Arabidopsis thaliana.2,3 As a PEBP, FT also possesses
the two notable features common to all PEBPs, namely (1) a com-
pact globular structure, which provides ample surface area for
interaction with other proteins and (2) a ligand binding pocket.1
The crystal structures of PEBPs from rat, human, bovine and plant
material reveal the evolutionary highly conserved nature of the
PEBP ligand binding pocket.4 For example, the very similar spatial
arrangement of the amino acid residues in the ligand binding
pocket of the FT-like protein Centrodialis (CEN) from Antirrhinum
and bovine PEBP (bPEBP) is depicted in Figure 1.
The ligand binding pocket in FT is interesting as a potential tar-
get for the development of small molecules that can activate or
inhibit the functioning of FT. In view of the importance of FT as
the key flowering-inducing protein, the applications for such com-
pounds in plant science and industry are obvious.
Currently, no small molecule inhibitor is known that can selec-
tively bind FT. In contrast, several small molecule inhibitors have
been identified for the Raf Kinase Inhibitory Protein (RKIP), a struc-
turally similar mammalian PEBP. RKIP has been well studied as a
potential therapeutic target involved in pathophysiological pro-
cesses including diabetic nephropathy,6 Alzheimer’s disease7–9
and immunotherapy against human cancer.10–13 The extensive
study of RKIP has also led to the discovery of locostatin (Fig. 1),
named after its ability to inhibit cellular locomotion in multiple
systems.10 Besides claims about its potential therapeutic value,10–
12,14–16
locostatin’s mode of action has been reported to involve
covalent and irreversible binding of the RKIP ligand binding
pocket.10,14,17 The pocket contains a highly conserved histidine
residue (His86) that was reported to act as a nucleophile on the
Michael acceptor in locostatin17 and which is also present in
bPEBP, CEN and FT (Fig. 1). Because of this conservation and the
overall structural similarity of the binding pocket, we hypothesized
that FT might also be covalently bound by locostatin in a similar
fashion and block the mode of action of FT, and hence influence
flowering. We therefore set out to investigate if locostatin affects
⇑
Corresponding author. Tel.: +31 (0)302534262.
E-mail address: t.wennekes@uu.nl (T. Wennekes).
http://dx.doi.org/10.1016/j.tetlet.2016.04.071
0040-4039/Ó 2016 Elsevier Ltd. All rights reserved.

J. Hoogenboom et al. / Tetrahedron Letters 57 (2016) 2406–2409
2407
a
Biotin
NH2
8
1
N
H
8
2
O
O
O
O
O
b
c
NH
N
N
O
O
O
Biotin
8
N
H
3
4
5
Ph
Ph
Ph
Scheme 1. Synthesis of locostatin-probe 5. Reagents and conditions: (a) EDC, HOBt,
D
-Biotin, DMF, rt, 1 d, 85%; (b) n-BuLi, acryloyl chloride, THF, À78 °C, 1 h, 17%; (c) 2,
HG-II, AcOH, 40 °C, 1 d, 26%.
the E-isomer. Not surprisingly, this compound turned out to pos-
sess only limited water-solubility so we also set out to develop
more water-soluble versions. To this end we aimed to synthesize
locostatin analogue 9 (Scheme 2) and PEGylated compound 13
(Scheme 3), which could then be coupled to a PEGylated Biotin
group through a Cu(I)-catalyzed alkyne–azide click reaction.19
Compound 9 could be obtained from (S)-4-benzyloxazolidin-2-
one 3, which was acrylated in good yield with 2-bromoacetyl bro-
mide and then converted to phosphonate 7 through an Arbuzov
reaction in almost quantitative yield. The resulting intermediate
was reacted through a Horner–Wadsworth–Emmons reaction with
aldehyde 8, providing the E-isomer as the major isomer that could
be purified to provide compound 9 in 74% yield.
Figure 1. Top: conserved residues of the PEBP ligand binding pocket in the crystal
structure of bPEBP (A) with bound phosphoryl group (PE) and FT-like protein CEN
(B).5 Bottom: Putative binding mechanism of locostatin in the ligand binding pocket
of the PEBP, RKIP.10,17
The other required building block, compound 13, could be
obtained from triethylene glycol, which was first alkylated to
provide intermediate 11 (Scheme 3). Intermediate 11 was then
oxidized via a Swern protocol to aldehyde 12. This aldehyde was
the flowering-inducing properties of FT and by developing a set of
novel locostatin-based chemical probes, if it selectively binds FT in
its ligand binding pocket.
We started by studying the effect of locostatin on the flowering
time of Arabidopsis thaliana Col0. We initially tested different con-
centrations of locostatin and noticed that this compound is toxic
for Arabidopsis at higher concentrations (ESI, Fig. S1). Based on this
reacted with intermediate 7 through another Horner–Wads-
worth–Emmons reaction to give compound 13.
With both compound 13 and compound 9 in hand we set out to
couple them to a PEGylated biotin 14 through a CuAAC click reac-
tion, which was achieved in reasonable yield (Scheme 4). With
probes 5, 15 and 16 in hand, we set out to validate if these loco-
statin-based chemical probes could covalently bind FT. For this
purpose FT was expressed in vitro and the protein mixture contain-
ing this FT was incubated with locostatin-probe 5 for 24 h at 37 °C
(ESI, Fig. S4), similar to the conditions employed earlier by Beshir
and co-workers17 for locostatin and RKIP. However, instead of
selective labelling of FT we observed that the probe seemed to
label the majority of the proteins in the cell lysate. Similarly, when
we used locostatin-probes 15 and 16 under milder conditions: 2 h
at 4 °C or room temperature, we again did not see any selectivity
for FT even at lower concentrations (Figs. 2A and ESI, S4). Never-
theless, it was unclear if this was due to poor selectivity or
due to low expression levels of FT in our system. To that end,
we expressed FT, fused to Glutathione S-Transferase (GST), in
pilot experiment a 25 lM locostatin solution in H2O was used in a
flowering time experiment. Although here we found some very
early flowering plants, no significant flowering stimulating or
repressing activity was observed at population level (ESI, Fig. S2).
The only significant effect was that locostatin negatively affected
root development of the plants (ESI, Fig. S1). It is improbable that
locostatin is affecting the roots due to interaction with FT, since
the FT gene is not expressed in roots.18 A saturated version of loco-
statin did not have a significant effect on the root development
(ESI, Fig. S3), indicating the crucial role of the reactive Michael
acceptor of locostatin for its activity.
Despite the fact that no FT-related phenotypes could be
observed in the plants treated with locostatin, we decided to inves-
tigate whether FT can be targeted by locostatin, based on the con-
servation of the PEBP ligand binding pocket and previous binding
studies for locostatin and RKIP.10,14,17 In order to visualize the
covalent modification of FT by locostatin, novel chemical probe
analogues of locostatin were designed. We aimed to modify the
Michael acceptor tail of locostatin, since it has been previously
reported for RKIP that this part of locostatin remains covalently
bound to the PEBP protein, while the oxazolidinone-moiety may
dissociate over time through hydrolysis.17 In addition, a struc-
ture–activity relationship study of locostatin with RKIP has shown
that structural modification of this part is tolerated.10
O
O
O
Br
NH
Ph
a
b
O
N
O
3
6
Ph
We aimed to synthesize target locostatin-based chemical probe
5 by starting with commercially available 1-amino-10-undecene 1,
which was coupled to biotin in good yield (Scheme 1). The result-
ing alkene 2 was then reacted with known compound 4 via a cross-
metathesis reaction. Initial attempts to carry out this reaction with
the 2nd generation Grubbs catalyst were unsuccessful, but we
were able to obtain compound 5 using the more stable 2nd gener-
ation Hoveyda–Grubbs catalyst in AcOH, which exclusively gave
O
O
O
P
OEt
O
O
O
8
c
OEt
N
N
O
O
9
7
Ph
Ph
Scheme 2. Synthesis of compound 9. Reagents and conditions: (a) bromoacetyl
bromide, n-BuLi, THF, À78 °C, 1 h, 70%; (b) P(OEt)3, neat, 50 °C, 16 h, 100%; (c) 8,
NaH, THF, rt, 1 h, 74%.

2408
OH
J. Hoogenboom et al. / Tetrahedron Letters 57 (2016) 2406–2409
O
O
OH
O
10
11
12
a
b
OH
O
3
3
O
O
O
N
3
O
c
13
O
H
Figure 2. (A) Binding of probe 15 and 16 at room temperature for 2 h with in vitro
expressed FT; (B) binding of 40 M probe 15 and 16 with purified GST-FT and
residual GST; binding of 200 M probe 15 with GST. Arrow indicates proximate
position where FT is expected.
Scheme 3. Synthesis of compound 13. Reagents and conditions: (a) NaH, 5-chloro-
1-pentyne, THF, 60 °C, 16 h, 60%; (b) oxalyl chloride, DMSO, DCM, À78 °C, Et3N, 1 h,
47%; (c) NaH, 7, THF, DCM, rt, 1 h, 26%.
l
l
O
O
H
N
O
a
N3
N
n
O
O
Biotin
+
4
9: n=0
: n=3
14
13
N
N
O
N
O
H
N
O
Biotin
O
N
n
4
O
15: n=0
16
: n=3
Scheme 4. Synthesis of locostatin-probes 15 and 16. Reagents and conditions: (a)
CuSO4, sodium ascorbate, DMF, H2O, 64% (for 15) or 68% (for 16).
Figure 3. (A) Evaluation of binding of 2.38
for 6 h at 37 °C with or without 2 mM DTT; (B) evaluation of selectivity of locostatin
lM RKIP by 200 lM locostatin probe 15
probes 15 and 16 towards 5 reference proteins (2.38 lM) with or without 2.38 lM
RKIP for 1 h at rt.
Escherichia coli and purified this fusion protein using affinity
chromatography.
The N-terminal GST fusion was used as an affinity tag. We then
incubated GST-FT with locostatin probes 15 and 16. Upon incubat-
ing GST-FT with probes 15 and 16 we found that both probes were
able to successfully bind GST-FT (Fig. 2B). However, we observed
that the probes also bound residual cleaved GST to a similar extent.
This proved that locostatin-based probes 15 and 16 were not selec-
tively binding FT. It also seemed to indicate that the Michael accep-
tor motif in locostatin is capable of reacting with the various
nucleophiles present on any given protein.
In the past it has been reported that a tritium-labelled loco-
statin analogue was able to selectively label RKIP in MDCK cell
extracts.10 We therefore next investigated if (1) our locostatin-
based chemical probes could also bind RKIP, and (2) if they display
selectivity towards RKIP. As such, we incubated recombinant RKIP
with locostatin probe 15 with or without 2 mM DTT. Although DTT
has been used by others in incubation steps of locostatin with RKIP,
we expected that the nucleophilic thiol of DTT might deactivate
our locostatin-based chemical probes.10,17 As expected we indeed
observed that binding of our locostatin-based probe 15 proceeded
much better in the absence of DTT (Fig. 3A). Similarly, locostatin
probe 5 was also able to bind RKIP, but not quantitatively, probably
due to a lower effective concentration of the probe as a result of its
lower water solubility (ESI, Fig. S6).
in line with a recent study of locostatin towards a bacterial PEBP
present in Mycobacterium tuberculosis, where the authors speculate
that locostatin probably inhibits additional targets.20
Conclusions
We successfully synthesized novel locostatin-based chemical
probes. These probes enabled us to study the locostatin binding
selectivity in cell lysates and protein mixtures, which provided sur-
prising and important new insights into the selectivity of locostatin.
While our locostatin-based probes can covalently bind both FT and
RKIP, we have shown that these probes are not selective and can
also bind a multitude of non-PEBP proteins. In addition, the appar-
ent toxicity of locostatin towards plants and lack of selectivity of
locostatin-based chemical probes towards FT indicate that loco-
statin is not a suitable lead compound towards novel small mole-
cule binders of FT and flowering time-modulating compounds.
Locostatin is currently sold commercially by many vendors as a
small molecule inhibitor of cell migration through its supposed
selective inhibition of RKIP. In addition, potential pharmacological
applications10–12,14–16 have been attributed to locostatin. However,
our findings here highlight that future research into locostatin and
its analogues and their use as inhibitors of RKIP and other PEBP tar-
gets should take into account that locostatin has limited selectivity
towards these proteins.
Finally, we tried to visualize the selectivity of our locostatin
probes towards RKIP by introducing equimolar amounts of com-
monly used, non-PEBP, reference proteins to a buffered solution
with or without RKIP. We observed that locostatin-based chemical
probes 15 and 16 seem to bind RKIP and the other negative control
proteins more or less to the same extent (Figs. 3B and ESI, S7). In
addition, the selectivity did not improve by performing the incuba-
tion for a shorter time at reduced temperatures. These results are
Acknowledgements
The authors thank Prof. Claudia Oecking for supplying the GST-
FT vector and the Netherlands Foundation for Scientific Research

J. Hoogenboom et al. / Tetrahedron Letters 57 (2016) 2406–2409
2409
8. Yuasa, H.; Ojika, K.; Mitake, S.; Katada, E.; Matsukawa, N.; Otsuka, Y.; Fujimori,
O.; Hirano, A. Dev. Brain Res. 2001, 127, 1.
9. Al-Mulla, F.; Bitar, M. S.; Taqi, Z.; Yeung, K. C. J. Cell. Physiol. 2013, 228, 1688.
10. Zhu, S.; Mc Henry, K. T.; Lane, W. S.; Fenteany, G. Chem. Biol. 2005, 12, 981.
11. Rudnitskaya, A. N.; Eddy, N. A.; Fenteany, G.; Gascon, J. A. J. Phys. Chem. B 2012,
10176.
(NWO) for funding this research via ChemThem: Chemical Biology
grant (728.011.105) and VENI grant (722.011.006) to T.W.
Supplementary data
12. Ménoret, A.; McAleer, J. P.; Ngoi, S.-M.; Ray, S.; Eddy, N. A.; Fenteany, G.; Lee, S.-
J.; Rossi, R. J.; Mukherji, B.; Allen, D. L.; Chakraborty, N. G.; Vella, A. T. J.
Immunol. 2009, 183, 7489.
13. Mc Henry, K. T.; Ankala, S. V.; Ghosh, A. K.; Fenteany, G. ChemBioChem 2002, 3,
1105.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.tetlet.2016.04.
071.
14. Shemon, A. N.; Eves, E. M.; Clark, M. C.; Heil, G.; Granovsky, A.; Zeng, L.;
Imamoto, A.; Koide, S.; Rosner, M. R. PLoS One 2009, 4, e6028.
15. Ciarmela, P.; Marzioni, D.; Islam, M. S.; Gray, P. C.; Terracciano, L.; Lorenzi, T.;
Todros, T.; Petraglia, F.; Castellucci, M. J. Cell. Physiol. 2012, 227, 1821.
16. Best, O. G.; Pyke, T.; Crassini, K.; Stevenson, W. S.; Mulligan, S. P. Blood 2014,
124, 3326.
17. Beshir, A. B.; Argueta, C. E.; Menikarachchi, L. C.; Gascón, J. A.; Fenteany, G. For.
Immunopathol. Dis. Therap. 2011, 2, 47.
18. Farrona, S.; Thorpe, F. L.; Engelhorn, J.; Adrian, J.; Dong, X.; Sarid-Krebs, L.;
Goodrich, J.; Turck, F. Plant Cell 2011, 23, 3204.
References and notes
1. Granovsky, A. E.; Rosner, M. R. Cell Res. 2008, 18, 452.
2. Lifschitz, E.; Eviatar, T.; Rozman, A.; Shalit, A.; Goldshmidt, A.; Amsellem, Z.;
Alvarez, J. P.; Eshed, Y. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 6398.
3. Corbesier, L.; Vincent, C.; Jang, S.; Fornara, F.; Fan, Q.; Searle, I.; Giakountis, A.;
Farrona, S.; Gissot, L.; Turnbull, C.; Coupland, G. Science 2007, 316, 1030.
4. Banfield, M. J.; Brady, R. L. J. Mol. Biol. 2000, 297, 1159.
5. Serre, L.; Vallée, B.; Bureaud, N.; Schoentgen, F.; Zelwer, C. Structure 1998, 6,
1255.
6. Keller, E. T.; Fu, Z.; Brennan, M. Biochem. Pharmacol. 2004, 68, 1049.
7. Sultana, R.; Perluigi, M.; Butterfield, D. A. Acta Neuropathol. 2009, 118, 131.
19. Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew. Chem., Int. Ed.
2002, 41, 2596.
20. Eulenburg, G.; Higman, V. A.; Diehl, A.; Wilmanns, M.; Holton, S. J. FEBS Lett.
2013, 587, 2936.