Fragment-Based Discovery of Novel Potent Sepiapterin Reductase Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.9b00218

Genome-wide-association studies in chronic low back pain patients identified sepiapterin reductase as a high interest target for developing new analgesics. Here we used ¹⁹F NMR fragment screening for the discovery of novel, ligand-efficient SPR

Full text of DOI:10.1021/acs.jmedchem.9b00218

Brief Article
pubs.acs.org/jmc
Cite This: J. Med. Chem. XXXX, XXX, XXX−XXX
Fragment-Based Discovery of Novel Potent Sepiapterin Reductase
Inhibitors
Jo Alen,* Markus Schade,* Markus Wagener, Frank Christian, Sonja Nordhoff, Beatrix Merla,
Torsten R. Dunkern, Gregor Bahrenberg, and Paul Ratcliffe
Grunenthal GmbH, Zieglerstraße 6, 52078 Aachen, Germany
̈
S
* Supporting Information
ABSTRACT: Genome-wide-association studies in chronic
low back pain patients identified sepiapterin reductase as a
high interest target for developing new analgesics. Here we
used 19F NMR fragment screening for the discovery of novel,
ligand-efficient SPR inhibitors. We report the crystal
structures of six chemically diverse inhibitors complexed
with SPR, identifying relevant interactions and binding modes
in the sepiapterin pocket. Exploration of our initial fragment
screening hit led to double-digit nanomolar inhibitors of SPR
with excellent ligand efficiency.
INTRODUCTION
■
Non-opioid or new mode-of-action medicines for chronic pain
are highly sought after, even more so since the opioid crisis has
emerged in the U.S.1 Patient genetics data have been reported
as a promising strategy to unravel novel molecular targets for
analgesics.2 Single-nucleotide-polymorphisms (SNPs) associa-
tion studies of 12 independent cohorts of patients with chronic
low back pain revealed that SNPs within and close to the gene
GCH1 are associated with reduced clinical and experimental
pain sensitivity. GCH1 encodes GTP cyclohydroxylase 1
(GTPCH1) which is the first and rate-limiting enzyme in
the biosynthetic cascade for (6R)-L-erythro-5,6,7,8-tetrahydro-
biopterin (BH4), an essential cofactor for aromatic amino acid
hydroxylases, alkylglycerol monooxygenase, and nitric oxide
synthases. Reducing excessive pathological BH4 production as
found in those low back pain patients is free of CNS side
effects.3
However, care should be taken not to reduce BH4 levels too
dramatically, as it is indispensable for the synthesis of essential
neurotransmitters and for glycerol ether metabolism.4 Patients
with Segawa’s disease who carry a dominant-negative mutation
in GCH1 that reduces BH4 production have mild motor
dysfunctions, which can be reverted by L-dopa treatment.5
Complete blockage of BH4 synthesis causes severe cognitive
delays, dystonia, tremors, and hyperphenylalaninemia.6,7
Inhibition of pathological BH4 production without CNS
side effects is speculated to be achieved through blockage of
sepiapterin reductase (SPR), the last enzyme in the BH4
biosynthetic cascade, because two salvage pathways can ensure
basal BH4 production in the absence of SPR. Patients with
autosomal SPR deficiency mutations show only mild motor
impairments which can be reverted by L-dopa therapy. This is
in line with preclinical animal models, whereby systemic
administration of the SPR inhibitor SPRi3 (Figure 1) to mice
Figure 1. Known SPR inhibitors.
twice daily for 3 days reduced hypersensitivity to neuropathic
and inflammatory pain but did not show motor, exploratory, or
depression-like adverse effects.8 As a consequence, we
considered SPR an attractive target for the development of
small molecule analgesics. An 19F NMR fragment screen was
carried out to identify novel chemical matter. Supported by
newly generated crystal structures of six ligands bound to SPR,
one fragment was further explored and optimized to a double-
digit nanomolar SPR inhibitor suitable for entering a lead
finding program.
RESULTS AND DISCUSSION
■
NMR Fragment Screen and Crystallography. Various
small molecules, such as sulfapyridine, N-acetylserotonin
(NAS), and xanthurenic acid (Figure 1), inhibit SPR with
low micromolar and even nanomolar IC50 values.9 X-ray
crystallography data show that they all bind to the BH4 site of
SPR in the presence of the cosubstrate NADP. We therefore
aimed at a focused fragment screen of the BH4 substrate
pocket rather than an unbiased screen of the entire protein
Received: February 1, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.jmedchem.9b00218
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Brief Article
surface, using an 19F-labeled reporter ligand in a competitive
binding assay.10 The 19F NMR detection tolerates the presence
of high micromolar to millimolar NADP/NADPH concen-
trations and further essential buffer components. We
considered NAS with an IC50 of ∼5 μM as a suitable starting
point for the design of a 19F-labeled reporter ligand as its acetyl
moiety can be readily replaced by trifluoroacetyl (Figure 2A).
Figure 3. Overlay of the crystal structures of six screening hits bound
to SPR; 1 (magenta, PDB code 6I6F); 2 (pink, PDB code 6I6C); 3
(orange, PDB code 6I6P); 4 (green, PDB code 6I79); 5 (blue, PDB
code 6I6T); 6 (cyan, PDB code 6I6V).
Figure 2. 19F NMR fragment screen: (A) selection of the reporter
ligand TNAS; (B) competitive binding. The 19F NMR signal of 25
μM TNAS in the presence of 2 μM SPR (red) is reverted back to the
intensity of 25 μM TNAS in the absence of protein (blue) when 25
μM inhibitor SPRi3 (green) is added, indicating 100% competition.
Indeed, 19F-NMR-detected competitive binding studies of N-
trifluoroacetylserotonin (TNAS) with the nanomolar SPR
inhibitor xanthurenic acid (Figure 1) showed partial displace-
ment, while SPRi3 exhibited full displacement, with a large
difference in 19F signal intensity (Figure 2B).
For the screen, we selected a chemically diverse library of
4750 fragments composed of diverse planar, sp3-enriched and
polar fragments, most of which were rule-of-3 compliant.11 We
categorized the 19F NMR screening hits from 1 to 4 according
to the displacement of the TNAS reporter, with full inhibition
labeled category 4 and no inhibition labeled category 0. We
obtained a total of 20 category 4 hits, five category 3 hits, and
one category 2 hit. Enzymatic SPR inhibition was assessed for
these 26 primary screening hits,12 resulting in IC50 values
between 0.6 and 75 μM and five showing less than half-
maximal inhibition at 200 μM (Supporting Information, Table
S1). Ligand efficiencies ranged from 0.55 to 0.26 kcal/(number
of heavy atoms), which compares very well to the ligand
efficiency of SPRi3 (0.48).8 Overall, 21 out of 26 NMR
screening hits were confirmed in the enzymatic assay, with
excellent ligand efficiencies, comparable to those of published
SPR inhibitors.
Figure 4. Hydrogen bonding to Ser157·Oγ (S), Tyr170·OH (Y) and
Cys159·NH (C) of SPR for the six crystallized SPR inhibitors, as
compared with sepiapterin.13
Next, we aimed to obtain protein crystallography data for
the 20 category 4 screening hits by soaking into NADP loaded
SPR crystals, as previously reported.9 Six compounds yielded
atomic resolution electron densities, all of which were located
in the BH4 site of SPR (Figure 3), as expected from the NMR
screen. Though the chemical structures of these six SPR
inhibitors (Figure 4) are diverse and clearly distinct from
known inhibitors,13 they all form a hydrogen bond with the
side chain hydroxy group of Tyr170 or Ser157 or both. Tyr170
and Ser157 are key residues for catalysis because they complex
the carbonyl of the endogenous substrate sepiapterin that is
reduced by SPR to hydroxymethylene.
Initially, we focused our attention on the most potent
inhibitor 3 (Figure 5; full descriptions of the other crystal
structures can be found in the Supporting Information), which
forms hydrogen bonds to these two residues via its hydroxy
group, while its vicinal carbonyl acceptor is twisted away with a
Figure 5. Crystal structure of 3 bound to SPR (PDB code 6I6P).
distance of 3.7 Å between its carbonyl oxygen and the γ-oxygen
of Ser157. This H-bonding mode between an aromatic
hydroxyl and Tyr170 and Ser157 was previously detected in
the crystal structure of xanthurenic acid bound to SPR.9
Less intuitive is the filling of the subpocket above the
nicotinamide moiety of the cosubstrate NADP, which is where
the hydride is transferred from NADPH to sepiapterin.
Inhibitor 3 packs a sp2-hybridized ring face-to-face onto the
nicotinamide (Supporting Information, Figure S8) as observed
in the crystal structure of sulfasalazine (SSZ). Additionally, 3
forms van der Waals interactions with the side chains of
B
DOI: 10.1021/acs.jmedchem.9b00218
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Brief Article
Table 1. Influence of Aromatic Ring Substituents on the SPR Inhibitory Potency of a Carboxamide Series
Phe164, Met218, and Leu222 on the opposite side of the
substrate entry channel.
good kinetic solubility (70 μM) and an acceptable log D7.4
(3.2). Initially, we explored the SAR of the dichlorosalicyla-
mide in the nicotinamide subpocket. The crystal structure
reveals several noteworthy features that could aid the hit
optimization process. First of all, it showed that the main
interaction stems from the phenolate, via hydrogen bonding
with Tyr170 and Ser157. Second, the carbonyl group of the
Hit Exploration. Given the potent IC50 of 580 nM
(compared to 310 nM for the known inhibitor/positive control
SPRi3 in this assay), the good ligand efficiency of 0.46, and the
attractive synthetic vectors of 3, this hit was selected for
chemical optimization. Furthermore, this compound exhibited
C
DOI: 10.1021/acs.jmedchem.9b00218
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Brief Article
amide does not seem to be conjugated with the aromatic ring,
as a torsion angle of 79° is observed. This could be attributed
to steric interactions between the bulky secondary amide
moiety and the ortho substituents of the aromatic ring.
Modifications were therefore aimed at dissecting the influence
of each of these factors (Table 1).
namely, a fluoro substituent in 3-position, resulted in a slightly
reduced (3-fold) SPR affinity.
As the initial aim was the reduction of the pKa of the
phenolate, 2,4-difluoro analog 26 (pKa = 6.56) was
synthesized, which gratifyingly exhibited a 10-fold increase in
potency (compared to 3), resulting in a double-digit
nanomolar compound. Encouraged by this positive result, a
limited number of further fluoro analogs was analyzed to
establish preliminary SAR around this improved hit.
Substituting the 4-F with H led to a 5-fold reduction in
potency. Introducing a methyl group in this position abrogated
once again the potency (IC50 > 10 μM), whereas the 4-Cl-
analog 17 showed low micromolar potency. On the other
hand, substituting the 2-F with H led to complete loss of
potency (IC50 > 10 μM) that was partially restored to
submicromolar potency by introducing a chloro substituent in
this position. The 4-F,2-pyridyl analog 30 proved to have
similar potency (roughly 300 nM) as its 4-Cl analog 20, which
spurred our interest in this heteroaromatic subseries.
Removing the F-substituent led to a moderate 2-fold decrease
in potency, but its three regioisomers 32−34 were all devoid of
potency. Interestingly, the 2,4-pyrimidyl analog 35 showed
modest, low micromolar potency. It is worth mentioning that
the log D values for these last four compounds 32−35 were
markedly decreased to below 1 (and even approaching 0 in the
case of the 3-pyridyl analog 33).
Summarizing, it became clear that minor variations in the
substitution pattern could lead to dramatic effects and even
real potency cliffs. Nevertheless, we were able to find a sweet
spot that led to a 10-fold increase in potency while retaining or
even slightly improving the good physicochemical properties of
the starting hit.
Guided by the cocrystal structure of SSZ with SPR, in which
a sulfonamide forms a bidentate H-bond with Ser157·Oγ and
Cys159·NH,9 we next exchanged the amide of the series by a
sulfonamide (Table 2). Sulfonamide analogs were synthesized
for the initial hit 3, its highly potent difluoro analog 26, and the
two monofluoro derivatives. Sulfonamide 36 was markedly
more potent (7-fold) than its carboxamide analog but
Initially, we focused our attention on assessing the
importance of the phenol, which could act as a double H
bond acceptor or a mixed H bond donor/acceptor. As
expected, changing the hydroxyl into a methoxy group resulted
in complete loss of affinity (IC50 > 10 μM). We then explored
how the pKa of the phenol correlates with potency: if the
phenolate acts as a double hydrogen bond acceptor similar to
the carbonyl oxygen in fragments 1 and 2 (Figure 4), acidifying
the phenolate should increase binding affinity and potency.
To investigate this without affecting the conjugation of the
phenyl with the amide moiety (or lack thereof), only the 4-Cl
substituent was varied initially. Switching from Cl (3, pKa =
6.63) to a more electronegative F-substituent (8, pKa = 6.60)
did not have any effect on the potency, whereas the H-analog
(pKa = 7.6) 9 was found to be 2-fold less potent. Introducing
an electron donating group (and therefore destabilizing the
phenolate) led to a significant reduction (4-fold) in potency
for a dimethylamino substitution (pKa = 7.69) and even
complete loss in activity (IC50 > 10 μM) for a methyl (pKa =
7.70) and methoxy (pKa = 7.21) group, respectively. However,
the difference in potency between a methyl and dimethylamino
group with very similar pKa values cannot be explained by an
electronic effect, so it is clear that other factors are playing a
role as well. This is further exemplified by the complete lack of
potency (IC50 > 10 μM) for analogs with strong electron
withdrawing substituents, such as nitrile (pKa = 6.21),
trifluoromethyl (pKa = 6.59), and primary amide (pKa =
6.89). Whereas an increase in potency (vs 3) was expected,
based on pKa values, the opposite was observed. Finally, we
introduced a ring N-atom in 4-position (pKa = 6.78), which
also proved to be detrimental for the potency (IC50 > 10 μM).
In the crystal structure, the 4-Cl atom of 3 is located 3.3 Å
away from the carbonyl O atom of Met205, which could lead
to an attractive π−δ interaction. Hence the 4-Cl substituent is
favorable beyond its pKa lowering effect and only 4-F yields
similar potency in this position.
Table 2. Influence of Aromatic Ring Substituents on the
SPR Inhibitory Potency of a Sulfonamide Subseries
To further establish a structure−activity relationship
regarding modifications on the phenyl ring, the 2-position
was investigated, whereby modifications could indistinguish-
ably influence the electron donating effect of the phenolate, the
torsion angle of the carbonyl group, or both. Significant
reduction in potency was observed for the F and the H analogs
(roughly 13-fold and 7-fold, respectively), and complete loss of
affinity (IC50 > 10 μM) was seen when a methyl group was
present in this position. On the other hand, the pyridine analog
20 exhibited a slightly (2-fold) improved potency.
We then explored the single-point regioisomers of our initial
hit 3 by moving a single chloro substituent around in the ring
(21−24). Disappointingly, only 21 retained some potency,
though significantly reduced (13-fold) compared to 3.
Therefore, no further efforts were undertaken to evaluate
substituents at these positions. It is noteworthy that none of
the modifications had a detrimental effect on the good
solubility of the starting hit except for the 3,4-dichloro
substitution pattern in derivative 23, which exhibited a roughly
10-fold lower solubility. Introducing a third halogen atom,
D
DOI: 10.1021/acs.jmedchem.9b00218
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Brief Article
b
Table 3. Influence of the Amine Moiety on the SPR Inhibitory Potency of a Carboxamide Series
a
b
Compounds tested as racemic mixtures. ND: not determined.
simultaneously poorly soluble (10-fold drop compared to its
carboxamide analog 3). The same trend in potency, though
less pronounced, was observed for the difluoro derivative 37
(2-fold increase in potency), resulting in a satisfying potency of
30 nM. In the case of monofluoro compounds, dramatic
differences in potency were observed. Though 4-F carbox-
amide 29 was more or less inactive (IC50 > 10 μM), its
sulfonamide analog 38 exhibited excellent potency in the
double-digit nanomolar range, meaning an impressive potency
gain of >300-fold. For the regioisomeric bioisosters, the shift in
potency was the other way around, with 27 having
submicromolar potency and its sulfonamide analog 39 being
inactive (IC50 > 10 μM). In contrast to the dichloro-
sulfonamide, monofluoro- and the difluorosulfonamides
showed excellent aqueous solubility.
to the amine part of the amide (Table 3). The 6-
azaspiro[3.4]octane tertiary amide of the initial hit 3 was
initially modified through ring expansions and contractions,
resulting in spirocycles 41−44, all showing slightly reduced
potency in the low micromolar range except for ring-expanded
analog 41 which was equipotent. Subsequently, we introduced
heteroatoms in the original spirocycle, resulting in a micro-
molar oxetane 45 and an inactive (IC50 > 10 μM) aziridine
analog 46, respectively. In the next round, simplification of the
spirocycle to its monocyclic analog was carried out. 3-
Dimethylpyrrolidine derivative 47 and racemic 3-fluoro-3-
methylpyrrolidine derivatives 48 showed a 9-fold drop in
potency. The potency was partly restored by removing the
substituents, with the unsubstituted pyrrolidine 49 being
roughly 3-fold less potent than the initial hit. In order to
improve potency and introduce potential exit vectors for
further modifications along the way, monosubstituted 3-
Having tackled the first two sites of modification, i.e., ring
substitution patterns and linker moiety, the focus was switched
E
DOI: 10.1021/acs.jmedchem.9b00218
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Brief Article
pyrrolidinyl derivatives were evaluated. Introducing a lipophilic
trifluoromethyl group had a minor effect on potency. Attaching
a phenyl ring proved to be very beneficial, bringing the potency
back in the range of the initial hit. All three pyridinyl
regioisomers 52−54 exhibited submicromolar potency, with
meta-pyridinyl being the preferred heteroaromatic moiety. N-
methylated 3-pyrazolyl was tolerated as well, exhibiting similar
potency as a para-pyridinyl. Following up on these promising
compounds, a methylene spacer was incorporated to move the
aromatic ring further away. While this affected the potency in a
slightly negative way for the phenyl (51 vs 56) and 3-pyridyl
(53 vs 57) analog, the 4-pyridyl analog 58 retained potency (vs
54). Exchanging the methylene linker with an O-, NH, or
ethylene linker seemed to have a beneficial effect (about 2- to
3-fold increase in potency in all three cases).
0.78 in a Caco-2 assay), good log D7.4 (2.5), and promising in
vitro metabolic stability (hLM Clint = 6.6 μL min−1 mg−1).
CHEMISTRY
■
Salicylamides 3, 7−35, 40−69 were synthesized from their
salicylic acid precursors in one step through coupling with the
selected secondary amine or from the corresponding methyl
ethers, requiring an additional demethylation step (Scheme 1).
a
Scheme 1. Synthesis of Salicylamides
To complete this exploration of combining phenyl with
pyrrolidine, fusing the phenyl ring with the heterocycle was
evaluated. A significant potency difference was observed
between the two isomers, with the isoindoline 62 being poorly
active (as well as poorly soluble) and the indoline 63 being the
most active amine counterpart discovered so far. Therefore,
further analogs were evaluated. 2-Methyl substitution had a
detrimental (20-fold) effect on potency. Ring expansion to
tetrahydroisoquinoline 65 or -quinoline 66 was better
tolerated, though both isomers were less potent than 63
(about 2- and 6-fold, respectively). Related derivatives as the
nonfused 4-phenylpiperidine analog 67 and its corresponding
benzylic and sulfone derivate were prepared. Whereas the latter
(69) was 3-fold less potent, the former (68) more or less
retained potency.
a
Reagents and conditions: (a) HNRR′, HATU, NEt3, THF, 0 °C →
rt, 16 h; (b) BBr3, DCM, 0 °C → rt, 1 h.
Sulfonamides 36−39 were obtained in two steps (Scheme
2) by coupling the secondary amine under basic conditions
with the relevant sulfonyl chloride and subsequent depro-
tection of the phenolic methyl ether.
a
Scheme 2. Synthesis of Sulfonamides
In summary, quite a lot of variations on the amine part were
tolerated (much more shallow SAR), facilitating the
introduction of exit vectors that could potentially be used for
the optimization of pharmacokinetic parameters.
a
Reagents and conditions: (a) NEt3, DCM, 0 °C → rt, 16 h; (b) BBr3,
DCM, 0 °C → rt, 30 min.
Exemplarily, selected physicochemical properties and in
vitro ADME parameters were evaluated for the potent amide
representative 26 (Table S2). The compound seemed well
suited as a starting point for further optimization, with high
aqueous solubility and log D7.4 in the desired range, excellent
ligand efficiency (0.53), and high in vitro permeability with no
significant efflux (ER < 1). The in vitro stability in human liver
microsomes (hLM) also seems to be acceptable, with a
predicted clearance of 32% of hepatic blood flow.
EXPERIMENTAL SECTION
All compounds assessed had purities that were >95%, determined by
HPLC, UPLC, or LC/MS, based on UV detection.
■
Synthesis of (6-Aza-spiro[3.4]oct-6-yl)-(2,4-difluoro-6-
hydroxyphenyl)methanone (26). To a mixture of 6-aza-
spiro[3.4]octane hydrochloride (175 mg, 1.19 mmol, 1.0 equiv) and
2,4-difluoro-6-hydroxybenzoic acid (207 mg, 1.19 mmol, 1.0 equiv) in
THF (12 mL) were added DIPEA (1.03 mL, 5.95 mmol, 5.0 equiv)
and HATU (542 mg, 1.43 mmol, 1.2 equiv) at 0 °C, and the reaction
mixture was stirred at rt for 16 h. After extraction with ethyl acetate
(150 mL), the combined organic fractions were consecutively washed
with water (50 mL), saturated sodium bicarbonate solution (50 mL),
and brine (50 mL). The organic layer was then dried over sodium
sulfate and concentrated under reduced pressure. The crude residue
was purified by column chromatography (silica gel; 15% acetone/
hexane) followed by preparative HPLC purification to yield (6-aza-
spiro[3.4]oct-6-yl)-(2,4-difluoro-6-hydroxyphenyl)methanone (45
CONCLUSION
■
Through NMR and X-ray supported fragment screening, we
identified novel chemical matter that could potentially lead to
new series of SPR inhibitors and revealed relevant interactions
and binding modes in the sepiapterin pocket of SPR. Out of
this diverse set of starting points, fragment 3 was selected to
embark on a hit optimization program, based on its on-target
submicromolar potency (580 nM), high ligand efficiency
(0.46), and potential for attractive exit vectors. Through
variations on three different sites of modification, a chemical
series was established with generally good aqueous solubility
and log D values in a desirable range. Within this series, the
improved hit 26 was obtained, with double-digit nanomolar
SPR potency (57 nM) and excellent ligand efficiency (0.53).
Furthermore, this compound could be qualified as candidate
for follow-up hit-to-lead optimization, as it exhibited good
physicochemical and in vitro ADME properties, such as high
aqueous solubility (98 μM) and permeability (Papp = 30 × 10−6
1
mg, 0.16 mmol, 14%) as an off-white solid. H NMR (600 MHz,
DMSO-d6, trans) δ 6.69 (qd, 1H), 6.53 (tt, 1H), 3.45−3.39 (m, 2H),
3.13 (s, 2H), 1.94−1.72 (m, 8H); 13C NMR (150 MHz, DMSO-d6)
δC 162.9, 161.5, 159.6, 156.5, 111.5, 99.7, 95.1, 58.1, 44.8, 44.2, 36.3,
31.1, 30.6, 15.9.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jmed-
chem.9b00218.
■
S
cm/s in a Caco-2 assay) with no apparent efflux (ERB‑A/A‑B
=
F
DOI: 10.1021/acs.jmedchem.9b00218
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Brief Article
(9) Haruki, H.; Hovius, R.; Pedersen, M. G.; Johnsson, K.
Tetrahydrobiopterin Biosynthesis as a potential target of the
kynurenine path-way metabolite xanthurenic acid. J. Biol. Chem.
2016, 291, 652−657.
NMR binding categories, IC50 values and ligand
efficiencies for the fragment screening hits, crystal
structures of all six fragments, details on the enzymatic
assay, synthesis procedures and characterization data for
all compounds (PDF)
(10) Dalvit, C.; Vulpetti, A. Ligand-based Fluorine NMR Screening:
Principles and Applications in Drug Discovery Projects. J. Med. Chem.
2019, 62, 2218−2244.
Molecular formula strings and some data (CSV)
(11) Congreve, M.; Carr, R.; Murray, C.; Jhoti, H. A “Rule of three”
for fragment-based lead discovery. Drug Discovery Today 2003, 8,
876−877.
(12) Katoh, S. Sepiapterin reductase from horse liver: purification
and properties of the enzyme. Arch. Biochem. Biophys. 1971, 146 (1),
202−214.
(13) Auerbach, G.; Herrmann, A.; Gutlich, M.; Fischer, M.; Jacob,
U.; Bacher, A.; Huber, R. The 1.25 Å crystal structure of sepiapterin
reductase reveals its binding mode to pterins and brain neuro-
transmitters. EMBO J. 1997, 16, 7219−7230.
Accession Codes
Atomic coordinates for the X-ray structures of compounds 1
(PDB code 6I6F), 2 (PDB code 6I6C), 3 (PDB code 6I6P), 4
(PDB code 6I79), 5 (PDB code 6I6T), and 6 (PDB code
6I6V) are available from the RCSB Protein Data Bank (www.
rcsb.org). Authors will release the atomic coordinates and
experimental data upon article publication.
AUTHOR INFORMATION
Corresponding Authors
*J. A.: e-mail, Jo.Alen@grunenthal.com.
*M. S.: e-mail, Markus.Schade@grunenthal.com.
■
ORCID
Jo Alen: 0000-0002-2215-812X
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors thank Stefanie Peters for tremendously helping
with the fragment screening and maintaining of the instru-
ments. The team of Proteros Biostructures GmbH, Bunsen-
strasse 7a, D-82152 Martinsried, Germany, and in particular
Dr. Martin Augustin and Dr. Michael Blaesse are acknowl-
edged for X-ray crystal structure determination. Utz-Peter
Jagusch and Katja Ohlen are acknowledged for log D and
solubility determinations. This work was funded by Gru
GmbH.
̈
nenthal
ABBREVIATIONS USED
■
SPR, sepiapterin reductase; GTPCH1, GTP cyclohydroxylase
1; BH4, tetrahydrobiopterin; NAS, N-acetylserotonin; TNAS,
N-trifluoroacetylserotonin; SSZ, sulfasalazine
REFERENCES
■
(1) Olson, M. E.; Eubanks, L. M.; Janda, K. D. Chemical
interventions for the opioid crisis: key advances and remaining
challenges. J. Am. Chem. Soc. 2019, 141, 1798−1806.
(2) Latremoliere, A.; Costigan, M. GCH1, BH4 and pain. Curr.
Pharm. Biotechnol. 2011, 12, 1728−1741.
(3) Latremoliere, A.; Costigan, M. Combining human and rodent
genetics to identify new analgesics. Neurosci. Bull. 2018, 34, 143−155.
̈
(4) Werner, E. R.; Blau, N.; Thony, B. Tetrahydrobiopterin:
biochemistry and pathophysiology. Biochem. J. 2011, 438, 397−414.
(5) Segawa, M.; Nomura, Y.; Nishiyama, N. Autosomal dominant
guanosine triphosphate cyclohydrolase I deficiency (Segawa disease).
Ann. Neurol. 2003, 54 (Suppl. 6), 32−45.
(6) Longo, N. Disorders of biopterin metabolism. J. Inherited Metab.
Dis. 2009, 32, 333−342.
(7) Opladen, T.; Hoffmann, G. F.; Blau, N. An international survey
of patients with tetrahydrobiopterin deficiencies presenting with
hyperphenylalaninaemia. J. Inherited Metab. Dis. 2012, 35, 963−973.
(8) Latremoliere, A.; Latini, A.; Andrews, N.; Cronin, S. J.; Fujita,
M.; Gorska, K.; Hovius, R.; Romero, C.; Chuaiphichai, S.; Painter, M.;
Miracca, G.; Babaniyi, O.; Remor, A. P.; Duong, K.; Riva, P.; Barrett,
L. B.; Ferreiros, N.; Naylor, A.; Penninger, J. M.; Tegeder, I.; Zhong,
J.; Blagg, J.; Channon, K. M.; Johnsson, K.; Costigan, M.; Woolf, C. J.
Reduction of neuropathic and inflammatory pain through inhibition
of the tetrahydrobiopterin pathway. Neuron 2015, 86, 1393−1406.
G
DOI: 10.1021/acs.jmedchem.9b00218
J. Med. Chem. XXXX, XXX, XXX−XXX