Structural requirements of 2-oxoglutaric acid analogues to mimic its signaling function

Article abstract of DOI:10.1021/ol401914z

A 2-oxoglutaric acid (2-OG) probe bearing a methylene group introduced at the C4 position and a vinyl group to replace the carbonyl group at the C2 position elicited characteristic affinity for NtcA, the 2-OG receptor, while maintaining the signaling function of the parent natural metabolite 2-OG. This discovery opens new perspectives in the design, synthesis, and implementation of specific 2-OG analogues as molecular probes for investigating the complex 2-OG signaling pathways.

Full text of DOI:10.1021/ol401914z

ORGANIC
LETTERS
2013
Vol. 15, No. 18
4662–4665
Structural Requirements of 2‑Oxoglutaric
Acid Analogues To Mimic Its Signaling
Function
Xinjun Liu,†,‡ Yang Wang,†,‡ Erik Laurini,§, Paola Posocco,§, Han Chen,‡
Fabio Ziarelli,^ Annick Janicki,# Fanqi Qu,‡ Maurizio Fermeglia,§, Sabrina Pricl,§,
Cheng-Cai Zhang,# and Ling Peng*,†
ꢀ
Aix-Marseille Universite, CINaM CNRS UMR 7325, 13288 Marseille, France, College
of Chemistry and Molecular Sciences, Wuhan University, Wuhan 430072, P.R. China,
Molecular Simulation Engineering Laboratory, University of Trieste, 34127 Trieste,
Italy, National Interuniversity Consortium for Material Science and Technology,
ꢀ
Research Unit MOSE-DEA, Trieste University, Italy, Spectropole, Faculte de
Saint-Jerome, Marseille, France, and Aix-Marseille Universite, LCB CNRS UMR
ꢀ ^
ꢀ
7283, 13402 Marseille, France
ling.peng@univ-amu.fr
Received July 8, 2013
ABSTRACT
A 2-oxoglutaric acid (2-OG) probe bearing a methylene group introduced at the C4 position and a vinyl group to replace the carbonyl group at the
C2 position elicited characteristic affinity for NtcA, the 2-OG receptor, while maintaining the signaling function of the parent natural metabolite
2-OG. This discovery opens new perspectives in the design, synthesis, and implementation of specific 2-OG analogues as molecular probes for
investigating the complex 2-OG signaling pathways.
Derived from the Krebs cycle, the metabolite 2-oxoglu-
taric acid (2-OG, Scheme 1) is a strategically important
precursor for biomolecular synthesis and cellular energy
production. In addition to these traditional roles, it re-
cently regained considerable attention as a key signaling
molecule1À4 in many other biological processes in various
organisms.5À9 Studies with cyanobacteria Anabaena sp.
PCC7120 (hereafter referred to as Anabaena) consitute an
exiquisite example: in Anabaena, the accumulation of
2-OG constitutes a nitrogen starvation signal which trig-
gers a cascade of cellular responses ultimately leading
to the formation of heterocysts.3,10 Studying the signaling
roles of 2-OG in vivo is a grand challenge since this
compound is rapidly metabolized. We inventively used
† CINaM CNRS UMR 7325.
‡ Wuhan University.
§ University of Trieste.
INSTM, Italy.
^ Spectropole, Marseille.
# LCB, CNRS UPR 9043.
(5) Ninfa, A.; Jiang, P. Curr. Opin. Microbiol. 2005, 8, 168.
ꢀ
(6) Galvez, S.; Lancien, M.; Hodges, M. Trends Plant Sci. 1999, 4,
484.
(1) Zhang, S.; Bryant, D. A. Science 2011, 334, 1551.
(2) He, W.; Miao, F. J.-P.; Lin, D. C.-H.; Schwandner, R. T.; Wang,
Z.; Gao, J.; Chen, J.-L.; Tian, H.; Ling, L. Nature 2004, 429, 188.
(7) Hebert, S. C. Nature 2004, 429, 143.
(8) Loenarz, C.; Schofield, C. J. Nat. Chem. Biol. 2008, 4, 152.
ꢀ
(3) Laurent, S.; Chen, H.; Bedu, S.; Ziarelli, F.; Peng, L.; Zhang,
C.-C. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 9907.
(4) Bourrellier, A. B. F.; Valot, B.; Guillot, A.; Ambard-Bretteville,
(9) Cairns, R. A.; Mak, T. W. Cancer Discov. 2013, 3, 730.
ꢀ
(10) Li, J.-H.; Laurent, S.; Konde, V.; Bedu, S.; Zhang, C.-C.
F.; Vidal, J.; Hodges, M. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 502.
Microbiology 2003, 149, 3257.
r
10.1021/ol401914z
Published on Web 08/29/2013
2013 American Chemical Society

the fluorinated nonmetabolisable 2-OG analogue DFPA
(Scheme 1) to mimic 2-OG, thusdemonstratingfor the first
time the signaling role of 2-OG in vivo in Anabaena.3 We
then showed that another 2-OG analogue, featuring a
planar vinyl group to mimic the carbonyl group in 2-OG
(2-MPA, Scheme 1), was also able to play a similar role in
Anabaena. These findings led to us conclude that it is the
planar ketone form rather than the tetrahedral ketal form
of 2-OG that is essential for the signaling activity of this
molecule.11 Recently, our group synthesized DFMPA
(Scheme 1), a hybrid probe based on DFPA and 2-MPA
carrying a vinyl and a gem-difluoromethylene group at the
C2 and C4 positions, respectively. Surprisingly, DFMPA
could also mimic the signaling function of 2-OG and
successfully trigger the formation of heterocysts in cya-
nobacteria.12 Although DFMPA shares with 2-OG some
structural features such as two carboxylic acid terminals,
it also presents essential chemical differences, particularly
at the C4 position. The surprising yet interesting results
obtained with DFMPA suggest that structural alter-
ations at C4 in 2-OG analogues can be tolerated, leading
to 2-OG mimics able to bind their receptor and, in so
doing, exercise a regulatory function on the metabolic
pathways in Anabaena.
Scheme 1. 2-OG and Its Structrural Analogues
The synthesis of 1À5 was successfully achieved accord-
ing to the strategies presented in Scheme 2. Probe 1 was
obtained using an established protocol described in
literature,13 whereas 2À4 were prepared starting from the
precursor7 which, in turn, was obtained by coupling6 with
poly(ethyl glyoxylate).14 After tosylation of the hydroxyl
group in 7 and subsequent nucleophilic substitution with
CsF, the resulting product was hydrolyzed under alkali
Probing the chemical modifications tolerated by 2-OG is
interesting in studies on the biological role of 2-OG itself,
and at the same time, such an investigation could also be of
great help in clarifying the interaction of this metabolite
withpotentialreceptors, inprovidinguseful information to
design competent affinity probes for identifying new 2-OG
receptors, and, ultimately, in dissecting the complex signal-
ing pathways of 2-OG. We therefore further explored the
structural requirement for 2-OG analogues to mimic the
signaling role of 2-OG by specifically varying the substi-
tuents at its C4 position. Our previous work proved that
replacing the carbonyl function at the C2 position of 2-OG
with a vinyl group led to a metabolite derivative with
unaltered signaling activity.11 Therefore, we designed
new compounds 1À5 in which the vinyl functionality at
the C2 position was maintained while the position C4
was modified with a series of substituents illustrated in
Scheme 1. The choice of chemical groups at C4 encom-
passes a diversity of chemical structures with varying
nature, size, and shape which, in turn, allow us to explore
the effects of structural variation and the requirement at
the C4 position of mimicking the biological behavior of the
original metabolite. The results stemming from this study
will be instrumental in the design and synthesis of new
2-OG analogues. At the same time, and perhaps more
importantly, we expect to use these analogues in the quest
for unknown 2-OG receptors and in shedding light on the
intriguing signaling pathways of 2-OG in cyanobacteria
and other organisms.
Scheme 2. Synthesis of Probe 1À5
conditions to afford 2.15 Probe 3 was obtained via direct
benzylation of 7 followed by subsequent ester hydrolysis.
To synthesize 4, the cyclopropyl group was imported via
carbene insertion onto the vinyl group in 7; the hydroxyl
group was then oxidized to a carbonyl group, which was
further transformed into a vinyl group via Wittig reaction.
(13) Basavaiah, D.; Sharada, D. S.; Kumaragurubaran, N.; Reddy,
R. M. J. Org. Chem. 2002, 67, 7135.
(14) Einhorn, C.; Luche, J.-L. J. Organomet. Chem. 1987, 322, 177.
(15) The poor solubility of CsF in organic solvent could be the main
reason for the low yield of the nucleophilic substitution reaction.
ꢀ
(11) Chen, H.; Laurent, S.; Bedu, S.; Ziarelli, F.; Chen, H.-L.; Cheng,
Y.; Zhang, C.-C.; Peng, L. Chem. Biol. 2006, 13, 849.
(12) Liu, X.; Chen, H.; Laurini, E.; Wang, Y.; Dal Col, V.; Posocco,
P.; Ziarelli, F.; Fermeglia, M.; Zhang, C.-C.; Pricl, S.; Peng, L. Org. Lett.
2011, 13, 2924.
Org. Lett., Vol. 15, No. 18, 2013
4663

Subsequent hydrolysis led to4 ingoodyield. Finally, probe
5 was obtained by trimerization of 8 followed by simulta-
neous hydrolysis of all three ester groups into carboxylic
acid functionalities.16
Angle Spinning (HRMAS) NMR, an excellent nondes-
tructive method for in vivo analysis of the metabolic
profiles of whole cells.20 Compared to DFMPA as the
positive control, both 1 and 4 were efficiently taken up by
the KGTP strain (Figure 2). Conversely, only a limited
amount of 5 was internalized by cells, while no notable
uptake was observed for either 2 or 3. These findings
allowed us to conclude that the poor/no cellular uptake
of 2, 3, and 5 could be directly related to the failure of these
compounds in inducing heterocyst formation.
With probes 1À5 at hand, we next studied their ability to
mimic 2-OG function in Anabaena, an excellent model to
investigate the signaling role of 2-OG in nitrogen metabo-
lism since it produces morphologically distinct heterocysts
in response to combined-nitrogen depletion.17 The so-
formed heterocysts are able to fix N2 from air, allowing
Anabaena to survive even in the absence of combined
nitrogen in the nutrition medium.18 Importantly, since
heterocysts differ morphologically from vegetative cells,
they can be easily distinguished under a microscope.
Furthermore, heterocyst differentiation can be repressed
at early stages, when a combined nitrogen source such as
ammonium or nitrate is supplied to the growth medium.
Hence, we first examined the ability of 1À5 to induce
heterocyst formation in the presence of high ammonium
concentration (5 mM) which, as mentioned above, repre-
sents a strongly repressive condition to heterocyst forma-
tion. Note that, like the natural metabolite 2-OG, probes
1À5 are highly negatively charged; therefore, their passage
across the cellular membrane should, at least in principle,
be disfavored.10 To address this issue, we constructed a
recombinant strain of Anabaena expressing a heterolo-
gous 2-OG permease KgtP from E. coli (herein referred
to as the KGTP strain) which can efficiently take up 2-OG
and its analogues.3,10À12,19 Notably, among all five probes,
only 1 could trigger heterocyst formation (Figure 1) in a
manner similar to that observed for DFPA, 2-MPA, and
DFMPA.3,11,12
Figure 2. 1H HRMAS NMR study on the cellular uptake of 1À5
in KGTP strain.
To unravel the molecular basis behind the singular
behavior of 1 with respect to 4 in mimicking the signaling
role of 2-OG and inducing heterocyst differentiation in
Anabaena, we carried out an in silico study (see SI for
details) on the binding of 1 and 4 to NtcA, a 2-OG receptor
able to regulate nitrogen metabolism and responsible for
heterocyst formation in cyanobacteria.3,11,12,21 The key
elements for a favorable binding of 2-OG toward NtcA
involve a network of four hydrogen bonds between 2-OG
and Phe75, Gly76, Val77, Leu78 of NtcA, respectively,
together with the two carboxylic groups of 2-OG interact-
ing with Arg88 and Arg129 in NtcA through permanent
salt bridges (Figure S1). Interestingly, our results revealed
that 1 adopts a binding pose within the NtcA binding
pocket very similar to that of the natural ligand 2-OG
(Figure 3A and C): indeed, three out of four hydrogen
bonds are preserved and so are the two salt bridges
involving the residues Arg88 and Arg129 although the
corresponding H-bond interaction with the NH group of
the peptide bond between Phe75 and Gly76 is no longer
feasible due to the replacement of the carbonyl group of
2-OG by a vinyl moiety in 1. These stabilizing interactions
Figure 1. Induction of heterocyst differentiation under repres-
sive conditions in the KGTP stain of Anabaena incubated (A)
without any probe, and in the presence of (B) DFMPA and (C)
1, respectively. Filaments were stained with alcian blue. Hetero-
cysts are indicated by red arrows.
result in an elevated affinity of 1 toward NtcA (ΔGbind
À14.0 ( 0.8 kJ/mol), similar to that of 2-OG (ΔGbind
=
=
To understand the reason for the distinctive behavior of
probe 1 in inducing heterocyst formation, we next exam-
inedthe cellular uptake efficiencyof all 5probes. Uptakeof
1À5 by the KGTP strain was monitored by inspecting the
groups around 5À6 ppm using 1H High Resolution Magic
À15.5 ( 0.9 kJ/mol) (Table S1). The situation is quite
different for compound 4 (Figure 3B and D): the two salt
bridges between the two carboxylic groups of 4 and the
guanidinium groups of Arg88 and Arg129 in NtcA are the
only stabilizing interactions. Also, the presence of the
cyclopropyl ring in 4 is found to induce an overall,
1
characteristic H NMR signals associated with the vinyl
(16) Amri, H.; Rambaud, M.; Villieras, J. Tetrahedron Lett. 1989, 30,
7381.
ꢀ
(17) Zhang, C.-C.; Laurent, S.; Sakr, S.; Peng, L.; Bedu, S. Mol.
Microbiol. 2006, 59, 367.
(20) Griffin, J.; Pole, J.; Nicholson, J.; Carmichael, P. Biochim.
Biophysi. Acta 2003, 1619, 151.
(18) Meeks, J. C.; Elhai, J. Microbiol. Mol. Biol. Rev. 2002, 66, 94.
(19) Vazquez-Bermudez, M. F.; Herrero, A.; Flores, E. J. Bacteriol.
2000, 182, 211.
(21) Zhao, M.-X.; Jiang, Y.-L.; He, Y.-X.; Chen, Y.-F.; Teng, Y.-B.;
Chen, Y.; Zhang, C.-C.; Zhou, C.-Z. Proc. Natl. Acad. Sci. U.S.A. 2010,
107, 12487.
4664
Org. Lett., Vol. 15, No. 18, 2013

Figure 4. DNA biding affinity of NtcA at various concentra-
tions of 2-OG, DFMPA, and probe 1. Co: DNA fragment
without NtcA.
Figure 3. Equilibrated MD snapshot of 1 and 4 in complex with
NtcA. Zoomed view of 1 (A) and 4 (B) within the NtcA binding
pocket. Compounds 1 and 4 are shown as atom-colored balls-
and-sticks (O, red; C, gray). The H-atoms of the methylene
group(s) are highlighted in white. Black lines denote hydrogen
bond and salt bridge interactions. NtcA residues directly in-
volved in binding 1 and 4 are shown as labeled, colored sticks:
Phe75 (light blue), Gly76 (yellow), Val77 (green), Leu78 (dark
green), Arg88 and Arg129 (both in blue). (C) Superposition of
NtcA/2-OG (red) and NtcA/1 (blue) binding modes. (D) Super-
position of NtcA/2-OG (red) and NtcA/4 (green) binding
modes. In all panels, water molecules, ions, and counterions
are omitted for clarity.
role of 2-OG in cells. Our results revealed that only
compound 1 is endowed with cell internalization capacity,
NtcA target protein affinity, and heterocyst formation
activity similar to the natural Krebs cycle metabolite
2-OG. Molecular modeling results demonstrated that 1
and 2-OG share similar binding modes within the NtcA
binding pocket; thus, 1 is able to exert the same signaling
function as 2-OG for heterocyst formation in cyanobacte-
ria Anabaena. Collectively, our results allow us to conclude
that the C4 position of 2-OG is of great importance and
that either the vinyl or the gem-difluoromethylene group
can be introduced at this position in order to retain the
signaling function of 2-OG. These findings represent a
further contribution toward a better understanding of the
nature and mechanisms of 2-OG binding to the NctA
receptor and will assist us in the design of new 2-OG
probes. We also encourage and foresee the use of such
analogues inthe identificationofnew 2-OGreceptors, with
the ultimate, ambitious task of expanding our knowledge
of the various signaling pathways of 2-OG such as the
balance of carbon to nitrogen in available nutrients,5,6 the
role of 2-OG-dependent oxygenases in epigenetic regu-
lation,8 and the altered metabolism of 2-OG to generate
the “oncometabolite” 2-hydroxyglutarate, the reduced
form of 2-OG, in various cancers.9 Currently, we are ac-
tively pursuing research in these directions.
considerable distortion of the whole binding site; particu-
larly, the side chain of Phe75 is drastically rotated while
Val77, Leu78, and Arg129 side chains all undergo a
significant translational move. All these conformational
readjustments result in a wider protein binding pocket and,
ultimately, in less effective ligand/protein stabilizing inter-
actions (ΔGbind = À8.9 ( 1.0 kJ/mol). Thus, according to
the computational results, even if compound 4 is able to
cross the cellular membrane, its low affinity for the NtcA
receptor reflects negligible biological activity.
To support the above in silico results, we further assessed
1 for its ability to promote DNA binding of NtcA. NtcA is
the 2-OG receptor, which is also a transcription factor
regulating nitrogen metabolism and initiating heter-
ocyst formation in cyanobacteria. 2-OG enhances the
binding affinity of NtcA toward its DNA promoter, an
event which can be easily revealed by performing DNA
mobility shift assays. Similarly to 2-OG and DFMPA, we
observed that 1 could effectively promote DNA/NtcA
binding (Figure 4). In contrast, 4 could not stably elicit
DNA/NtcA binding (Figure S2). Taken together, our
results provide a reasonable molecular rationale for the
finding that only 1 among all the 2-OG analogues 1À5
designedand synthesizedin thisworkcan act asa regulator
of NtcA and effectively mimic the signaling role of 2-OG.
In summary, we designed and synthesized novel 2-OG
probes 1À5 bearing different moieties at the C4 position
while retaining the vinyl group at the C2 position with the
aim of understanding the structural requirements of 2-OG
analogues at the C4 position in mimicking the signaling
Acknowledgment. Financial support from ANR PCV
ꢀ
program, Region PACA, Wuhan University and Aix-
Marseille Universite. X.L. and Y.W. are supported by
ꢀ
the oversee PhD fellowship from the China Scholarship
Council. X.L. and Y.W. contributed equally to this
work.
Supporting Information Available. Chemical synthesis,
analytical data, NMR spectra, biological study, compu-
ter modeling as well as Figures S1 and S2, Scheme S1,
Table S1. This material is available free of charge via the
Internet at http://pubs.acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 18, 2013
4665