Exploiting protein symmetry to design light-controllable enzyme inhibitors

Article abstract of DOI:10.1002/anie.201307207

The activity of the metabolic branch-point enzyme PriA from Mycobacterium tuberculosis (mtPriA) can be controlled reversibly by light. Two-pronged inhibitors based on the dithienylethene scaffold were designed utilizing mtPriA's natural rotational symmetr

Full text of DOI:10.1002/anie.201307207

Angewandte
Chemie
DOI: 10.1002/anie.201307207
Light-Regulated Enzyme Activity
Exploiting Protein Symmetry To Design Light-Controllable Enzyme
Inhibitors**
Bernd Reisinger, Natascha Kuzmanovic, Patrick Lçffler, Rainer Merkl, Burkhard Kçnig,* and
Reinhard Sterner*
Abstract: The activity of the metabolic branch-point enzyme
PriA from Mycobacterium tuberculosis (mtPriA) can be
controlled reversibly by light. Two-pronged inhibitors based
on the dithienylethene scaffold were designed utilizing
mtPriAꢀs natural rotational symmetry. Switching from the
flexible, ring-open to the rigid, ring-closed isomer reduces
inhibition activity by one order of magnitude.
enzymeꢀs active site. This task can be fulfilled either by the
covalent incorporation of a molecular photoswitch near the
catalytic center[4,7] or by the design of a noncovalently bound,
light-controlled inhibitor.[4,8]
We aimed to design a light-controlled inhibitor for
phosphoribosyl isomerase A from Mycobacterium tuberculo-
sis (mtPriA). mtPriA is a branch-point enzyme in amino acid
biosynthesis as it catalyzes two chemically equivalent sugar
isomerization reactions in tryptophan and histidine biosyn-
thesis.[9] In the latter, the aminoaldose N’-[(5’-phosphoribo-
T
he artificial control of biological processes by light is
a rapidly emerging area of protein design.[1] Three basic
strategies for the light regulation of biomolecules have been
reported: key positions have been functionalized with photo-
labile protecting groups,[2] naturally
syl)formimino]-5-aminoimidazole-4-carboxamide
ribo-
nucleotide (ProFAR) is converted to the corresponding
occurring photoreceptors have
been
reprogrammed,[3]
and
designed molecules that can be
reversibly switched by light (photo-
switches) have been used to direct
protein or cellular function.[4] With
respect to the latter, substantial
progress has been made in the
regulation of neuronal activity by
designing light-inducible ligands for
ion channels and receptors.[5] As the
molecular recognition of specific
ligand parts leads to a nonlinear
signal response in neural systems,
even small changes in the binding
efficacy upon light irradiation sig- Figure 1. Reaction and structure of mtPriA. a) mtPriA catalyzes the conversion of ProFAR to PRFAR in
the histidine biosynthesis. b) Ribbon representation of the (ba)8-barrel structure of mtPriA with
nificantly influence the cellular
bound product PRFAR (PDB ID: 3ZS4[12a]). The view is along the twofold symmetry axis of the
output.[6] In contrast, for the rever-
protein. PRFAR is anchored by two opposite phosphate binding sites, which are enlarged in the
insets. Hydrogen bonds are indicated by dashed lines.
sible control of enzymatic activities,
the switching of a photoresponsive
group must substantially affect the
aminoketose N’-[(5’-phosphoribulosyl)formimino]-5-amino-
imidazole-4-carboxamide ribonucleotide (PRFAR) (Fig-
ure 1a). Since humans can synthesize neither histidine nor
[*] B. Reisinger,[+] P. Lçffler, Prof. Dr. R. Merkl, Prof. Dr. R. Sterner
Institut fꢀr Biophysik und physikalische Biochemie
Universitꢁt Regensburg, 93040 Regensburg (Germany)
E-mail: Reinhard.Sterner@ur.de
N. Kuzmanovic,[+] Prof. Dr. B. Kçnig
Institut fꢀr Organische Chemie
Universitꢁt Regensburg, 93040 Regensburg (Germany)
E-mail: Burkhard.Koenig@ur.de
tryptophan, mtPriA is a potential target for anti-tuberculosis
drugs.[10] Structurally, mtPriA belongs to the class of (ba)8-
barrels, which is a frequently encountered and highly versatile
fold among enzymes.[11] The protein exhibits a clear twofold
symmetry (Figure 1b),[12] which indicates its evolution from
a (ba)4-half-barrel precursor.[13] Consequently, two phosphate
[+] These authors contributed equally to this work.
binding sites are found opposite each other to fix the substrate
[**] B.R. was supported by a PhD fellowship from the Cusanuswerk.
ProFAR and the product PRFAR (Figure 1b). We thus
reasoned that a C2-symmetric photoswitch with terminal
phosphate anchors would be an excellent foundation for
building a light-controllable inhibitor of mtPriA.
Financial support by the Deutsche Forschungsgemeinschaft (GRK
1910) is gratefully acknowledged.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201307207.
Angew. Chem. Int. Ed. 2014, 53, 595 –598
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
595

.
Angewandte
Communications
Two types of organic photochromic systems
possess the desired twofold rotational symme-
try: stilbene[14] and azobenzene switches[15] and
the diarylethene scaffold.[16] Although azoben-
zene derivatives have been widely used in
biological systems, they suffer from incomplete
photoconversion and thermal reversibility.[1] In
contrast, photoresponsive compounds based on
1,2-dithienylethene (DTE) generally feature
photoconversions of over 90% with both photo-
isomers being thermally stable.[8b,16] Hence, we
opted for DTE as a core and provided it with
different phosphate and phosphonate anchors
(Scheme 1).
Starting
with
the
bis-
chlorodithienylethene 1, Suzuki coupling
yielded either the aromatic hydroxides 2–4 or
the aromatic bromides 8 and 9. The former were
subsequently converted to ortho-, meta- and
para-phosphates 5–7, while the latter were used
to synthesize the phosphonic acid esters 10 and
11, which were finally hydrolyzed to afford meta-
and para-phosphonates 12 and 13.
The DTE molecular structure can be toggled
reversibly between a ring-open and a ring-closed
photoisomer (Table 1), which significantly alters
its overall conformational flexibility.[8b] Energy
minimizations of the open and closed forms of
all potential inhibitors confirmed distances
between the phosphorus atoms of 15.8 to
19.6 ꢁ (Table 1), which is in good accordance
with the 16.9 ꢁ observed for the corresponding
atoms in the mtPriA structure with co-crystal-
lized PRFAR (PDB ID: 3ZS4[12a]; see the
Supporting Information for details). Only the
open and closed isomers of ortho-phosphate 5
exhibit rather short P–P distances of 12.3 ꢁ and
11.3 ꢁ, respectively, in their energetically most
Scheme 1. Synthesis of DTE phosphates and DTE phosphonates.
favorable geometries; in addition, less populated, more
extended conformers can be observed.
Table 1: Photochemical switching and corresponding calculated P–P
distances in DTE phosphates and DTE phosphonates 5–7, 12, and 13.
When ring-open forms of compounds 5–7, 12, and 13 are
irradiated with UV light (312 nm), the absorption band at
280 nm immediately decreases. Simultaneously, new absorp-
tion maxima at 350 nm and 525 nm formed, turning the
initially colorless solutions pink (Figure S1). In each case, the
spectral changes are complete after 30 s of irradiation and the
corresponding photostationary states consist of between 93%
and 97% of the closed isomers, as judged by HPLC analyses
(Figure S2). The open forms can be recovered by irradiation
with visible light (> 420 nm) and all switches are robust over
several ring-closing/ring-opening cycles (Figure S3).
Compound
d(P–P) [ꢂ]
closed
open
5
6
7
12
13
12.3
16.7
19.6
16.9
18.9
11.3
16.2
18.4
15.8
18.1
The activity of mtPriA can be measured spectrophoto-
metrically at 300 nm in
a
coupled enzyme assay
(Scheme S1).[17] As all synthesized compounds were stable
under the assay conditions, their inhibitory effect could be
investigated in steady-state enzyme kinetics. For this purpose,
substrate saturation curves were monitored in the presence of
different concentrations of compounds 5–7, 12, and 13 in their
open and closed forms (curves are shown for compound 6 in
Figure S4). Indeed, all investigated DTE phosphates and
DTE phosphonates are able to inhibit the mtPriA reaction in
both isomeric forms, thus proving the viability of the design
concept. As expected for competitive inhibition, the turnover
numbers kcat were identical in the presence and absence of
596
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 595 –598

Angewandte
Chemie
inhibitor (Table S1). The observed increase of the Michaelis
constants caused by the inhibitors (Table S1) was used to
calculate the inhibition constants Ki (Equation S1), which are
given in Table 2.
Table 2: Inhibition constants Ki for compounds 5–7, 12, and 13 in their
ring-open and ring-closed forms.
Inhibitor
Ki [mm][a]
ring-closed
ring-open
5
6
7
12
13
8.1Æ2.1
0.55Æ0.12
3.5Æ0.3
1.6Æ0.1
6.8Æ0.6
7.0Æ1.2
4.4Æ0.3
3.7Æ0.4
4.9Æ0.3
22.7Æ4.7
Figure 2. Change in mtPriA activity upon ring-closure of compound 6.
The turnover of 5 mm ProFAR was followed photometrically at 300 nm
in the presence of 4 mm 6 in its open form under typical assay
conditions (258C, 50 mm Tris/acetate pH 8.5, 100 mm ammonium
acetate, 0.18 mm HisF, and 0.15 mm mtPriA). After having reached its
maximum rate, the reaction mixture was either left in the spectropho-
tometer (gray) or removed and irradiated with 312 nm light for 10 s
(black). A reference solution without enzymes was used to correct the
baseline shift resulting from different absorption values of the open
and closed isomer at 300 nm (see Figure S1).
[a] The values of Ki were obtained from the data shown in Table S1 using
Equation S1.
The determined inhibition constants are in the low
micromolar range (Table 2) and therefore comparable or
even better than the enzymeꢀs KM value for the natural
substrate ProFAR (KMProFAR = 8.6 mm; Table S1). In agree-
ment with the appropriate distance between its phosphorus
atoms (Table 1), the open isomer of phosphate 6 exhibits the
highest binding affinity (Ki = 0.55 mm). However, when 6 is
switched to its rigid, closed form, the inhibition activity is
lowered by roughly one order of magnitude (Ki = 4.4 mm). A
similar trend is observed for the corresponding phosphonate
12, whose binding affinity in the open form (Ki = 1.6 mm)
decreases about threefold upon ring-closure (Ki = 4.9 mm). In
contrast, the inhibitory effects of phosphates 5 and 7 are
nearly identical in their ring-open and ring-closed forms
(Table 2). Interestingly, the introduction of a phosphonate
moiety in para position (compound 13) causes a threefold
difference in the inhibition activity of the open (Ki = 6.8 mm)
and closed photoisomer (Ki = 22.7 mm). The removal of the
oxygen bridge between the terminal anchor and the switch-
able core further reduces the overall flexibility, which seems
to predominantly affect the already rigid, closed isomer.
The performance of mtPriA can also be controlled by
irradiation with light during catalysis. When the reaction was
started with compound 6 in its strongly inhibiting, open form,
the reaction rate could be enhanced by about threefold upon
switching to the less active, ring-closed isomer (Figure 2).
As pointed out before, the mechanistic principle behind
the different binding affinities of the photoisomers is based on
a change in conformational flexibility.[8b] Due to free rotation
À
around the C C bonds joining the thiophene heterocycles to
the central cyclopentene ring and the terminal phenyl groups,
DTE phosphate 6 is able to adopt various geometries in its
ring-open form. On the other hand, the closed isomer is
completely conjugated and thus far more restricted in its
mobility. To gain insight into the binding modes of the
inhibitors, we performed molecular dynamics (MD) simula-
tions of the open and closed isomers of compound 6 bound to
mtPriA. Although both forms are clearly fixed at the
phosphate binding sites (Figure 3a,c), obvious differences
can be observed in their structural cores. While the open
isomer converges to similar, C2-symmetric conformers in
Figure 3. MD simulations of mtPriA and bound meta-phosphate 6. For
each isomer, three independent calculations were performed and
representative enzyme structures for the open (a) and closed (c) form
are shown. A superposition of the energetically most favorable
conformer of each simulation is depicted in for the open conformation
(b) and the closed conformation (d).
Angew. Chem. Int. Ed. 2014, 53, 595 –598
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
597

.
Angewandte
Communications
[3] a) U. Krauss, T. Drepper, K. E. Jaeger, Chem. Eur. J. 2011, 17,
2552 – 2560; b) O. Yizhar, L. E. Fenno, T. J. Davidson, M. Mogri,
K. Deisseroth, Neuron 2011, 71, 9 – 34; c) F. Zhang, J. Vierock, O.
Yizhar, L. E. Fenno, S. Tsunoda, A. Kianianmomeni, M. Prigge,
A. Berndt, J. Cushman, J. Polle et al., Cell 2011, 147, 1446 – 1457.
[4] W. Szymanski, J. M. Beierle, H. A. Kistemaker, W. A. Velema,
B. L. Feringa, Chem. Rev. 2013, 113, 6114 – 6178.
[5] a) T. Fehrentz, M. Schçnberger, D. Trauner, Angew. Chem. 2011,
123, 12362 – 12390; Angew. Chem. Int. Ed. 2011, 50, 12156 –
12182; b) I. Tochitsky, M. R. Banghart, A. Mourot, J. Z. Yao,
B. Gaub, R. H. Kramer, D. Trauner, Nat. Chem. 2012, 4, 105 –
111.
[6] M. R. Banghart, A. Mourot, D. L. Fortin, J. Z. Yao, R. H.
Kramer, D. Trauner, Angew. Chem. 2009, 121, 9261 – 9265;
Angew. Chem. Int. Ed. 2009, 48, 9097 – 9101.
[7] S. Muramatsu, K. Kinbara, H. Taguchi, N. Ishii, T. Aida, J. Am.
Chem. Soc. 2006, 128, 3764 – 3769.
[8] a) S. Herre, T. Schadendorf, I. Ivanov, C. Herrberger, W. Steinle,
K. Rꢂck-Braun, R. Preissner, H. Kuhn, ChemBioChem 2006, 7,
1089 – 1095; b) D. Vomasta, C. Hçgner, N. R. Branda, B. Kçnig,
Angew. Chem. 2008, 120, 7756 – 7759; Angew. Chem. Int. Ed.
2008, 47, 7644 – 7647; c) D. Vomasta, A. Innocenti, B. Kçnig,
C. T. Supuran, Bioorg. Med. Chem. Lett. 2009, 19, 1283 – 1286.
[9] F. Barona-Gomez, D. A. Hodgson, EMBO Rep. 2003, 4, 296 –
300.
[10] a) K. Mdluli, M. Spigelman, Curr. Opin. Pharmacol. 2006, 6,
459 – 467; b) H. Shen, F. Wang, Y. Zhang, Q. Huang, S. Xu, H.
Hu, J. Yue, H. Wang, FEBS J. 2009, 276, 144 – 154.
[11] a) B. Hçcker, C. Jꢂrgens, M. Wilmanns, R. Sterner, Curr. Opin.
Biotechnol. 2001, 12, 376 – 381; b) R. K. Wierenga, FEBS Lett.
2001, 492, 193 – 198.
[12] a) A. V. Due, J. Kuper, A. Geerlof, J. P. von Kries, M. Wilmanns,
Proc. Natl. Acad. Sci. USA 2011, 108, 3554 – 3559; b) J. Kuper, C.
Doenges, M. Wilmanns, EMBO Rep. 2005, 6, 134 – 139.
[13] F. List, R. Sterner, M. Wilmanns, ChemBioChem 2011, 12, 1487 –
1494.
[14] D. H. Waldeck, Chem. Rev. 1991, 91, 415 – 436.
[15] A. A. Beharry, G. A. Woolley, Chem. Soc. Rev. 2011, 40, 4422 –
4437.
three independent calculations (Figure 3b), more diverse
binding modes are found for the ring-closed isomer (Fig-
ure 3d). Here, one terminal phenyl ring is twisted and adopts
various geometries to facilitate proper coordination of the
terminal phosphate group. The measured difference in
inhibition activity is also reflected in the binding energies
determined during the simulations, which consistently show
that interaction with the open form is energetically more
favorable (Table S2). Taken together, the higher flexibility of
the open isomer allows for better adaptation to the enzymeꢀs
active site and apparently overcompensates for the loss in
entropy upon binding.
In summary, we have demonstrated that natural protein
symmetry can be advantageously utilized to design light-
controllable enzyme inhibitors. The two-pronged DTE
switches can be toggled reversibly between a high- and low-
affinity form, where both photoisomers are nearly quantita-
tively formed and thermally stable. Hence, the enzymeꢀs
performance can alternately be enhanced and reduced by
irradiation with UV and visible light, respectively. The
viability of a dual-anchored DTE inhibitor has been shown
before[8b] and the design concept can in principle be trans-
ferred to functionally quite different enzyme systems. Phos-
phate is a frequently encountered element of metabolic
substrates, and various other (ba)8-barrel enzymes such as
pyridoxine 5’-phosphate synthase[18] and aldolases[19] possess
two phosphate binding sites. Thus, for these enzymes,
inhibitors may be designed by a similar approach, which
would allow for the photocontrol of several metabolic
processes independently in a spatiotemporal fashion.
Received: August 16, 2013
Published online: November 25, 2013
Keywords: biosynthesis · enzyme catalysis · enzyme inhibitors ·
[16] M. Irie, Chem. Rev. 2000, 100, 1685 – 1716.
.
[17] T. J. Klem, V. J. Davisson, Biochemistry 1993, 32, 5177 – 5186.
[18] M. G. Franco, B. Laber, R. Huber, T. Clausen, Structure 2001, 9,
245 – 253.
[19] a) E. Lorentzen, E. Pohl, P. Zwart, A. Stark, R. B. Russell, T.
Knura, R. Hensel, B. Siebers, J. Biol. Chem. 2003, 278, 47253 –
47260; b) T. Wagner, I. A. Shumilin, R. Bauerle, R. H. Kret-
singer, J. Mol. Biol. 2000, 301, 389 – 399.
molecular switches · photochromism
[1] C. Brieke, F. Rohrbach, A. Gottschalk, G. Mayer, A. Heckel,
Angew. Chem. 2012, 124, 8572 – 8604; Angew. Chem. Int. Ed.
2012, 51, 8446 – 8476.
[2] A. Deiters, ChemBioChem 2010, 11, 47 – 53.
598
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 595 –598