Substrate and inhibitor specificity of kynurenine monooxygenase from Cytophaga hutchinsonii

Article abstract of DOI:10.1016/j.bmcl.2017.02.080

Kynurenine monooxygenase (KMO) is a potential drug target for treatment of neurodegenerative disorders such as Huntington's and Alzheimer's diseases. We have evaluated substituted kynurenines as substrates or inhibitors of KMO from Cytophaga hutchinsonii.

Full text of DOI:10.1016/j.bmcl.2017.02.080

Accepted Manuscript
Substrate and Inhibitor Specificity of Kynurenine Monooxygenase from Cyto-
phaga hutchinsonii
Robert S. Phillips, Andrew D. Anderson, Harvey G. Gentry, Osman F. Güner,
J. Phillip Bowen
PII:
S0960-894X(17)30223-8
DOI:
http://dx.doi.org/10.1016/j.bmcl.2017.02.080
BMCL 24752
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
7 January 2017
27 February 2017
28 February 2017
Please cite this article as: Phillips, R.S., Anderson, A.D., Gentry, H.G., Güner, O.F., Bowen, J.P., Substrate and
Inhibitor Specificity of Kynurenine Monooxygenase from Cytophaga hutchinsonii, Bioorganic & Medicinal
Chemistry Letters (2017), doi: http://dx.doi.org/10.1016/j.bmcl.2017.02.080
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Substrate and Inhibitor Specificity of Kynurenine Monooxygenase from Cytophaga
hutchinsonii
Robert S. Phillips*a,b, Andrew D. Andersonb, Harvey G. Gentryb, Osman F. Günerc,1, and J. Phillip
Bowenc
aDepartment of Chemistry, University of Georgia, Athens, Georgia 30602
bDepartment of Biochemistry and Molecular Biology, University of Georgia, Athens, Georgia 30602
cCenter for Drug Design, Department of Pharmaceutical Sciences, College of Pharmacy, Mercer
University, Atlanta, GA 30341, USA.
1Present address: Department of Chemistry & Physics, Santa Rosa Junior College, Santa Rosa, CA
95401, USA
*To whom correspondence should be addressed at Department of Chemistry, University of Georgia,
Athens, Georgia 30602. E-mail: plp@uga.edu

ABSTRACT
Kynurenine monooxygenase (KMO) is a potential drug target for treatment of neurodegenerative
disorders such as Huntington's and Alzheimer's diseases. We have evaluated substituted kynurenines as
substrates or inhibitors of KMO from Cytophaga hutchinsonii. Kynurenines substituted with a halogen
at the 5-position are excellent substrates, with values of kcat and kcat/Km comparable to or higher than
kynurenine. However, kynurenines substituted in the 3-position are competitive inhibitors, with KI
values lower than the Km for kynurenine. Bromination also enhances inhibition, and 3,5-
dibromokynurenine is a potent competitive inhibitor with a KI value of 1.5 µM. A pharmacophore
model of KMO was developed, and predicted that 3,4-dichlorohippuric acid would be an inhibitor. The
KI for this compound was found to be 34 µM, thus validating the pharmacophore model. We are using
these results and our model to design more potent inhibitors of KMO.

The kynurenine pathway is the major pathway for tryptophan catabolism in many organisms,
including man1, as shown in Scheme 1. Tryptophan undergoes cleavage of the 2,3-double bond
catalyzed by a hemoprotein, either tryptophan dioxygenase (TDO) in the liver or indoleamine
dioxygenase (IDO) in extrahepatic tissues, to produce N-formyl-L-kynurenine.
Subsequent
deformylation by kynurenine formamidase gives kynurenine, which then is hydroxylated by a
mitochondrial flavoprotein, kynurenine monooxygenase (KMO), to give 3-hydroxy-L-kynurenine.
Subsequently, kynureninase, a pyridoxal-5'-phosphate dependent enzyme, catalyzes the hydrolytic
cleavage of 3-hydroxy-L-kynurenine to 3-hydroxyanthranilate and L-alanine. 3-Hydroxyanthranilate
dioxygenase (3HAO) then produces 2-amino-3-carboxymuconate semialdehyde, which can either
cyclize spontaneously to quinolinate2 or undergo enzymatic decarboxyation by 2-amino-3-
carboxymuconate semialdehyde decarboxylase (ACSD) to 2-aminomuconate semialdehyde. The
former reaction ultimately leads to NAD(P)+ formation, and the latter reaction leads to complete
catabolism via the 2-ketoadipate pathway. Several of the intermediates in the kynurenine pathway are
cytotoxic, including 3-hydroxykynurenine and 3-hydroxyanthranilate, which are apoptotic3, and
quinolinate, which is a neurotoxin due to NMDA receptor agonism4. Overexpression of the kynurenine
pathway has been linked to neurodegenerative diseases such as Huntington's5, Parkinson's6 and
Alzheimer's diseases7. Thus, enzymes in the kynurenine pathway are drug targets, and KMO has
emerged as the target of choice, since all downstream metabolites are potentially cytotoxic. As proof of
concept, a KMO inhibitor was shown to delay the onset of disease in an Alzheimer's mouse model8.
Kynurenine analogues were some of the first KMO inhibitors to be examined, and m-
nitrobenzoylalanine was shown to be an effective KMO inhibitor in the brains of mice9. It was
recognized that binding to KMO required a substituted aromatic ring and a carboxylate separated by 3
spacer atoms. Later, N-(4-phenylthiazol-2-yl)benzenesulfonamides were found to be potent inhibitors,
with KI values in the nM range10. Since the crystal structure of the Saccharomyces cerevisiae enzyme

Scheme 1. The kynurenine pathway for tryptophan metabolism.was published,11 KMO inhibitors have
been designed computationally12. However, a systematic study of the effects of substituents on
kynurenine on the KMO substrate activity and inhibition has not been performed. Recently, we
prepared a series of substituted kynurenines and evaluated their activity as substrates for human and
Pseudomonas fluorescens kynureninase13. In the present study, we have evaluated these substituted
kynurenines as substrates and inhibitors for KMO from Cytophaga hutchinsonii, and we have designed
and validated a pharmacophore model for KMO which predicted new inhibitors. The active site
residues which contact substrates and inhibitors in the crystal structure of KMO from Saccharomyces
cerevisiae11 are conserved in KMO from both C. hutchinsonii and humans, so the enzyme from C.
hutchinsonii is a suitable model to develop lead compounds.

Table 1
Substrates for C. hutchinsonii KMO
Substrate
Km (µM)
kcat (s-1)
2.0 ± 0.2
0.6 ± 0.1
0.8 ± 0.1
kcat/Km(M-1 s-1)
(5.3 ± 1) x104
(1.3 ± 0.3) x105
(4.0 ± 1) x105
L-Kynurenine (1)
5-Bromo-L-Kyn (2)
5-Chloro-L-Kyn (3)
67±15
4.5±3.0
2.0±0.7
We tested 5-halo-L-kynurenines as potential inhibitors for KMO, but discovered high blank
rates in assays containing the compounds without kynurenine. In addition to NADPH consumption in
the presence of 5-bromo-L-kynurenine (2), we also observed increasing absorbance in the 370-400 nm
region during the reaction, as expected since hydroxylation of kynurenine causes a red shift in the
absorption spectrum. Furthermore, HPLC analysis of reaction mixtures of 2 incubated with KMO
Figure 1. HPLC analysis of 5-bromo-L-kynurenine reaction with KMO. Black line: Reaction mixture
without addition of KMO. Red line: Reaction mixture after addition of KMO and incubation at 37 °C
for 30 minutes.

shows a new peak which elutes earlier (Figure 1, red line) than kynurenine (Figure 1, black line), and
exhibits λmax red-shifted from 380 to 388 nm. ESI-LC/MS analysis of this reaction mixture exhibited
new peaks at m/z = 303 and 305 for (M+1)+, as expected for the product, 5-bromo-3-
hydroxykynurenine. Thus, these compounds are not acting as uncouplers, consuming NADPH without
hydroxylation, although we found previously that β-benzoylalanine is an uncoupler of KMO from
Pseudomonas fluorescens14. 5-Chloro-L-kynurenine (3) is also a substrate based on NADPH
consumption (Table 1). Both kynurenines halogenated in the 5-position are excellent substrates for C.
hutchinsonii KMO (Table 1), with kcat/Km values greater than kynurenine itself. The 2-3-fold decrease
in kcat for these compounds is more than compensated by the lower values of Km. It is interesting that
we also found previously that these 5-halogenated kynurenines are excellent substrates for both human
and P. fluorescens kynureninase13.
In contrast, kynurenines with substitutents in the 3-position were found to be competitive
inhibitors of NADPH consumption in the spectrophotometric assay (Table 2). These compounds could
be considered product analogues of KMO. 3-Chloro (5), and 3-methylkynurenine (7) have KI values
lower than the Km of kynurenine. However, it is not clear if the value of Km ≈ Kd for kynurenine. As
these 3-substituted kynurenines are racemic, the KI value for the L-isomers should be half of the values
given in Table 2. Since 5-bromination lowers the Km for substrates (Table 1), we were interested to see
if this also applied to inhibitors. Bromination of 3-methyl- and 3-chloro-DL-kynurenine with Br2 in
acetic acid gave the corresponding 5-bromo-3-substituted kynurenines (see supplemental data). As
anticipated, these compounds have comparable or lower KI values than the 3-substituted kynurenines
(Table 2). The most potent inhibitor tested is 3,5-dibromo-L-kynurenine (4), with a KI value of 1.5 µM
(Table 2). The 3-halokynurenines do not appear to be efficient uncouplers of hydroxylation, since the
blank rates in the presence of the compounds were not significantly higher than normal.

Table 2
Inhibition of KMO by 3-substituted kynurenines
Compound
KI (µM)
3, 5-diBr-L-Kyn (4)
3-Cl-DL-Kyn (5)
5-Br-3-Cl-DL-Kyn (6)
3-Me-DL-Kyn (7)
5-Br-3-Me-DL-Kyn (8)
1.5 ± 0.6
13.8 ± 8.4
11.0 ± 3.6
8.6 ± 5.8
10.2 ± 4.7
Figure 2. Hydrogen bonding interaction with Arg83. The x-ray crystal structure complex of UPF648
bound to yeast KMO (pdb code: 4J36). The important hydrogen bonding interactions with Arg83 are
shown at the right, along with other interactions.
Based on these results and the published crystal structure of Saccharomyces cerevisiae KMO,
we developed a pharmacophore model. Pharmacophore modeling has been an extremely successful

approach for lead identification and modification15-17. There are two broadly defined approaches for
pharmacophore modeling: (i), ligand-based pharmacophore development, where there is no drug-target
structural information and small ligands with different binding affinities are used, and (ii), structure-
based pharmacophore development where specific binding conformations of ligands are used. For the
work reported herein, we have adopted primarily the latter approach.
Figure 3. Receptor-based pharmacophoric models of 3,4-dimethoxyhippuric acid aligned with the
conformation of UPF648 as bound to yeast KMO (pdb code: 4J36). The green spheres are
hydrophobic centers, the red spheres are acidic centers, the pink spheres with arrows are H-bond
acceptors, and the orange rings are aromatic planes.
Our pharmacophore modeling concentrated on lead identification. Multiple models were
generated using the Schrödinger software package18, and each was scored in search of novel
biologically active chemotypes. We carried out pharmacophore modeling of the bioactive conformation
of UPF648 bound yeast KMO19, 20
.
The approach we used is structure-based modeling using the bound conformation of known
active compounds in KMO X-ray crystal structures. A good understanding of the SAR of KMO is

desirable for the success of the design and development of novel inhibitors for KMO. If we have a
reliable description of the inhibitor-specific structural changes in the KMO catalytic site, we can
efficiently enrich compound libraries from which we can identify hits with minimized false positives
and negatives. Our proof-of-concept work was performed on a single bound conformation of a small
molecule KMO inhibitor, UPF648, bound to yeast KMO19, 20 (Fig. 2). On the periphery, UPF648 has a
pair of hydrogen bonding interactions with Arg83 while the rest of the UPF648 is placed in the larger
lipophilic pocket formed by Leu221, Ile232 and Leu234. The pharmacophore centers based on the
bound conformation of UPF648 with KMO were generated (Fig. 3) and a number of compounds were
retrieved through virtual screening of a database of commercially available chemicals. Two compounds
with a high fit value to the pharmacophore model, 3,4-dichlorohippuric acid and 3,4-
dimethoxyhippuric acid (Chart 1), have been identified as high scoring hits from this initial screening.
Table 3
Inhibition of KMO by substituted hippuric acids
Compound
KI (µM)
3,4-Dimethoxyhippuric acid (9)
3, 4-Dichlorohippuric acid (10)
(1.0 ± 0.6) x103
34 ± 20
3,4-Dimethoxyhippuric acid (9) and 3,4-dichlorohippuric acid(10) were then tested, with the
results shown in Table 3. Although 9 is a weak inhibitor, with a KI value of about 1 mM, 10 is much
better, with a KI value of 34 µM, thus validating the pharmacophore model in Figure 3. We were
surprised that the 3,4-dimethoxy substitution is not tolerated for binding of hippuric acids, because it is
found in one of the N-(4-phenylthiazol-2-yl)benzenesulfonamides with an IC50 value of 37 nM10. This
suggests different binding modes for these two compounds. 3,4-Dichloro substitution was also found
to be optimal for 1112, which can be considered as a cyclized analog of 10. It should be noted that the
3-position in these hippuric acids is equivalent to the 5-position in the kynurenines (see Chart 1), so the
effect of the halogen is similar in increasing affinity of both substrates and inhibitors of KMO. A recent

paper used molecular modeling to screen a series of arylpyrimidine carboxylic acids related to 11 as
potential inhibitors of KMO21. A similar compound, GSK-180 (12) was found to be a potent inhibitor
of KMO and is active against a rodent model of acute pancreatitis22. KMO inhibitors were the subject
of a recent comprehensive review23.
Chart 1. Compounds examined in this study and that were identified by pharmacophore methods using
the bound conformation of UPF 648, and other active compounds, 11 and 12.

In conclusion, kynurenines substituted at the 3-position are competitive inhibitors of KMO, and
halogen substitution at C-5 increases both substrate and inhibitor activity. The α-amino group is not
necessary for inhibitor binding, since halogen-substituted hippuric acids have KI values comparable to
kynurenines. The pharmacophore modeling results provide a preliminary validation of our approach
and set the stage for future structure-based modeling. We have begun to explore in greater detail the
binding modes of several key inhibitors within the KMO catalytic site to design more potent inhibitors
of KMO.
Supplementary data. The supplementary data contains the details of enzyme purification and assay,
and the synthesis and characterization of kynurenine derivatives.
Acknowledgement. The expression system for KMO from Cytophaga hutchinsonii was generously
provided by Professor Tadgh Begley of Texas A&M University.
Funding sources. This research did not receive any specific grant from funding agencies in the public,
commercial, or not-for-profit sectors.
References and notes.
1. Badawy, A. A.-B. Targeting the Broadly Pathogenic Kynurenine Pathway. Springer
International Publishing, 2015, 11.
2. Colabroy, K. L.; Begley, T. P. J. Am. Chem. Soc. 2005, 127, 840.
3. Fallarino, F.; Grohmann, U.; Vacca, C.; Bianchi, R.; Orabona, C.; Spreca, A.; Fioretti, M.
C.; Puccetti, P. Cell Death Diff. 2002, 9, 1069.
4. Kim, J. P.; Choi, D. W. Neurosci. 1987 23, 423.
5. Campesan, S.; Green, E. W.; Breda, C.; Sathyasaikumar, K. V.; Muchowski, P. J.;
Schwarcz, R.; Kyriacou, C. P.; Giorgini, F. Curr. Biol. 2011, 21, 961.
6. Szabó, N.; Kincses, Z. T.; Toldi J.; Vécsei, L. J. Neurolog. Sci. 2011, 310, 256.
7. Guillemin, G. J.; Brew, B. J. Redox Report 2002,7, 199.

8. Zwilling, D.; Huang, S.Y.; Sathyasaikumar, K.V.; Notarangelo, F. M.; Guidetti, P.; Wu, H.
Q.; Lee, J.; Truong, J.; Andrews-Zwilling, Y.; Hsieh, E.W.; Louie, J. Y.; Cell 2011, 145, 863.
9. Chiarugi, A.; Carpenedo, R.; Molina, M. T.; Mattoli, L.; Pellicciari, R.; Moroni, F. J.
Neurochem. 1995, 65, 1176.
10. Röver, S.; Cesura, A. M.; Huguenin, P.; Kettler, R.; Szente, A. J. Med. Chem.1997, 40,
4378.
11. Amaral, M.; Levy, C.; Heyes, D. J.; Lafite, P.; Outeiro, T. F.; Giorgini, F.; Leys, D.;
Scrutton, N. S. Nature 2013, 496, 382.
12. Toledo-Sherman, L. M.; Prime, M. E.; Mrzljak, L.; Beconi, M. G.; Beresford, A.;
Brookfield, F. A.; Brown, C. J.; Cardaun, I.; Courtney, S. M.; Dijkman, U.; Hamelin-Flegg,
E. J. Med. Chem. 2015, 58, 1159.
13. Maitrani, C.; Phillips, R. S. Bioorg. Med. Chem. 2013, 21, 4670.
14. Crozier-Reabe, K. R.; Phillips, R. S.; Moran, G. R. Biochemistry 2008, 47, 12420.
15. Bowen, J. P. in Computational chemistry and computer-assisted drug design. Block J, Beale
J.M., Jr, editors. Wilson and Gisvold’s textbook of organic medicinal and pharmaceutical
chemistry, 11th. Philadelphia: Lippincott, Williams, and Wilkins; 2004. pp. 919-947.
16. Pharmacophore Perception, Development, and Use in Drug Design; Güner, O. F., Ed.;
International University Line: La Jolla, California, 2000.
17. Zhong, H.; Wees, M. A.;. Faure, T. D.; Carrillo, C.; Arbiser, J.; Bowen, J. P. ACS Med.
Chem. Letts. 2011, 2, 455-460.
18. Small-Molecule Drug Discovery Suite 2013-3: QikProp, version 3.8, Schrödinger, LLC,
New York, NY, 2013.
19. Guner, O. F. Curr. Top. in Med. Chem., 2002, 2, 1321-1332.
20. Guner, O.F. IDrugs 2005, 8, 567-572.
21. Amin, S. A.; Adhikari, N.; Jha, T.; Gayen, S. Bioorg. Med. Chem. Letts. 2016, 26, 5712-

5718.
22. Mole, D. J.; Webster, S. P.; Uings, I.; Zheng, X.; Binnie, M.; Wilson, K.; Hutchinson, J. P.;
Mirguet, O.; Walker, A.; Beaufils, B.; Ancellin, N.; Trottet, L.; Beneton, V.; Mowat, C. G.;
Wilkinson, M.; Rowland, P.; Haslam, C.; McBride, A.; Homer, N. Z. M.; Baily, J. E.; Sharp,
M. G. F.; Garden, O. J.; Hughes, J.; Howie, S. E. M.; Holmes, D. S.; Liddle, J.; Iredale, J. P.
Nat. Med., 2016, 22, 202-209.
23. Smith, J. R.; Jamie, J. F.; Guillemin, G. J. Drug Discovery Today, 2016, 21, 315-324.

•
•
•
5-Halokynurenines are good substrates for kynurenine monooxygenase
3-Substituted kynurenines are competitive inhibitors of kynurenine monooxygenase.
A pharmacophore model was prepared and predicted 3,4-dichlorohippuric acid as an
inhibitor.