Creation of a S1P Lyase bacterial surrogate for structure-based drug design

Article abstract of DOI:10.1016/j.bmcl.2016.02.084

S1P Lyase (SPL) has been described as a drug target in the treatment of autoimmune diseases. It plays an important role in maintaining intracellular levels of S1P thereby affecting T cell egress from lymphoid tissues. Several groups have already published approaches to inhibit S1P Lyase with small molecules, which in turn increase endogenous S1P concentrations resulting in immunosuppression. The use of structural biology has previously aided SPL inhibitor design. Novel construct design is at times necessary to provide a reagent for protein crystallography. Here we present a chimeric bacterial protein scaffold used for protein X-ray structures in the presence of early small molecule inhibitors. Mutations were introduced to the bacterial SPL from Symbiobacterium thermophilum which mimic the human enzyme. As a result, two mutant StSPL crystal structures resolved to 2.8 ? and 2.2 ? resolutions were solved and provide initial structural hypotheses for an isoxazole chemical series, whose optimization is discussed in the accompanying paper.

Full text of DOI:10.1016/j.bmcl.2016.02.084

Accepted Manuscript
Creation of a SIP Lyase Bacterial Surrogate for Structure-Based Drug Design
Maria A. Argiriadi, David Banach, Elzbieta Radziejewska, Susan Marchie,
Jennifer DiMauro, Jurgen Dinges, Eric Dominguez, Charles Hutchins, Russell
A. Judge, Kara Queeney, Grier Wallace, Christopher M. Harris
PII:
S0960-894X(16)30209-8
DOI:
http://dx.doi.org/10.1016/j.bmcl.2016.02.084
BMCL 23638
Reference:
To appear in:
Bioorganic & Medicinal Chemistry Letters
Received Date:
Revised Date:
Accepted Date:
13 January 2016
25 February 2016
27 February 2016
Please cite this article as: Argiriadi, M.A., Banach, D., Radziejewska, E., Marchie, S., DiMauro, J., Dinges, J.,
Dominguez, E., Hutchins, C., Judge, R.A., Queeney, K., Wallace, G., Harris, C.M., Creation of a SIP Lyase Bacterial
Surrogate for Structure-Based Drug Design, Bioorganic & Medicinal Chemistry Letters (2016), doi: http://
dx.doi.org/10.1016/j.bmcl.2016.02.084
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.

Bioorganic & Medicinal Chemistry
jo u r n a l h o m e p a g e : w w w . e ls e v ie r . c o m
Creation of a SIP Lyase Bacterial Surrogate for Structure-Based Drug Design
a*
b
b
b
b
Maria. A. Argiriadi, David Banach , Elzbieta Radziejewska , Susan Marchie , Jennifer DiMauro , Jurgen
c
b
c
c
b
a
Dinges , Eric Dominguez , Charles Hutchins , Russell A. Judge , Kara Queeney , Grier Wallace and
Christopher M. Harris
b
*
Corresponding author. Tel +1 508 849 2643. E-mail address:maria.argiriadi@abbvie.com
AbbVie Bioresearch Center, 381 Plantation Street, Worcester, MA 01605
AbbVie Bioresearch Center, 100 Research Drive, Worcester, MA 01605
a
b
c
AbbVie, 1 North Waukegan Road, North Chicago, IL 60064
ARTICLE INFO
ABSTRACT
Article history:
Received
Received in revised form
Accepted
S1P Lyase (SPL) has been described as a drug target in the treatment of autoimmune diseases. It plays
an important role in maintaining intracellular levels of S1P thereby affecting T cell egress from
lymphoid tissues. Several groups have already published approaches to inhibit S1P Lyase with small
molecules, which in turn increase endogenous S1P concentrations resulting in immunosuppression.
The use of structural biology has previously aided SPL inhibitor design. Novel construct design is at
times necessary to provide a reagent for protein crystallography. Here we present a chimeric bacterial
protein scaffold used for protein x-ray structures in the presence of early small molecule inhibitors.
Mutations were introduced to the bacterial SPL from Symbiobacterium thermophilum which mimic the
human enzyme. As a result, two mutant StSPL crystal structures resolved to 2.8 Å and 2.2 Å
resolutions were solved and provide initial structural hypotheses for an isoxazole chemical series,
whose optimization is discussed in the accompanying paper.
Available online
Keywords:
S1P Lyase
Site-directed mutagenesis
Homology modeling, crystal structure
2
015 Elsevier Ltd. All rights reserved.
Small molecule drug discovery often requires structural
biology to facilitate the design of compounds that can interact
with a given target protein. There are instances when protein
reagent generation is difficult with hurdles such as low protein
yield/solubility or protein conformational heterogeneity. This
hampers progression into wide crystallization screening and
ultimately iterative crystal structures. Protein construct design
can be used to overcome this hurdle. Here we will describe a
novel construct approach to provide structural support for the
protein target S1P Lyase.
was obtained from both a prokaryotic SPL homolog in
Symbiobacterium thermophilium (StSPL) and yeast SPL (Dpl1p),
which gave the first view of the SPL structural scaffold and key
6
features of the binding pocket. One of the conclusions from this
study was that both prokaryotic and eukaryotic homologs shared
the same quaternary fold (overall r.m.s.d 1.2 Å for dimer), with
relatively conserved features in the active site and the PLP co-
factor binding pocket. Due to our difficulty with protein
expression and generation of human SPL for crystallography and
based on the strong resemblances between prokaryotic and
eukaryotic scaffolds, we embarked on a bacterial surrogate
approach to use a more “human-like” mutant version of StSPL to
crystallize with our early SPL inhibitor leads. Here we present
the surrogate construct strategy and key crystal structures
resulting from this effort. The accompanying paper discusses
application of these structures for further inhibitor optimization.
S1P Lyase (SPL) catalyzes the irreversible breakdown of
sphingosine-1-phosphate (S1P) into trans-2-hexadecenal and
ethanolamine phosphate. This activity is critical for maintaining
the low S1P tone in secondary lymphoid organs which is
1
essential for normal lymphocyte trafficking. SPL inhibition
prevents lymphocyte egress from lymph node, thymus, and
spleen,
immunosuppression.
investigated as a potential target for a range of autoimmune
resulting
in
peripheral
lymphopenia
and
Several attempts were made to create human constructs for
SPL in order to structurally enable design chemistry efforts.
Protein generated from these constructs suffered from poor yields
and large aggregation issues. We were fortunate to read the
published work of Bourquin and colleagues which presented a
high resolution crystal structure of a prokaryotic homolog of SPL
1,2
Therefore this protein has been
3
,4
disorders such as rheumatoid arthritis and multiple sclerosis.
SPL is a member of the PLP (pyridoxal 5’-phosphate)
5
6
superfamily and is located in the endoplasmic reticulum. Initial
from Symbiobacterium thermophilium (PDB code 3MAD).
SPL structural information of the soluble PLP binding domain

6
6
6
5
2.0´10
1.5´10
1.0´10
5.0´10
Figure 1a: Model of S1P (magenta) bound to human SPL (tan)
0
0
500
1000
Substrate (mM)
Figure 2: Kinetic characterization of SPL constructs. K
1500
2000
2500
M
and
max for mutant StSPL are 310 mM and 7.1x10 cts for mutant
StSPL (■) and estimated to be >2,000 mM and 5x10 cts for
huSPL (●).
5
V
6
in the presence of PLP co-factor were grown and diffracted to
high resolution. In parallel, an HTS screen was conducted using
the wild-type human enzyme and several compounds were
identified as preliminary hits. As an example, Compound 1
(
Figure 3) (N-(2-((4-methoxy-2,5-dimethylbenzyl)amino)-1-
6
Figure 1b: Overlay of StSPL (green) (PDB code 3MAD ) with
human SPL (residues shown in tan) which shows four divergent
residues proximal to the S1P binding site
phenylethyl)-5-methylisoxazole-3-carboxamide) was identified
with modest potency with an IC50 of 1.0 mM. This compound was
used in crystal-soaking procedures in hopes of obtaining a ligand
bound StSPL mutant crystal structure. At that time, we needed
structural guidance to provide a binding mode, preferred
chirality, and design opportunities for optimization of this
isoxazole-amide chemotype.
Using this crystal structure, a homology model of human (with
S1P bound) was generated to observe the potential placement of
the natural substrate and the amino acids that surround the
7
substrate. In the model (shown in Figure 1a), the phosphate of
S1P is located in the putative phosphate binding site and makes
interactions with residues Y150, K359 and H174. F290 caps the
active site forming a fairly narrow channel for the S1P substrate
to bind. The active site opening contains an assortment of
hydrophobic residues which presumably stabilize the long alkyl
tail of the substrate. Based on this modeling hypothesis, we
superimposed the high resolution structure of StSPL (PDB code
Compound 1
Compound 2
3
MAD) and the human SPL model to observe key residue
differences proximal to S1P. While most locations in the active
site remained fairly constant (i.e., PLP binding pocket and
phosphate binding site), four positions (StSPL: Y249, L344,
F346 and L497, Human: F290, I386, A388, S542) showed
diversity between the homologs (Figure 1b). Y249 superimposes
with the key residue F290 and L344/ L479/L497 line the active
site opening to solvent. We then decided to use the StSPL protein
scaffold and introduce site-directed mutations at these locations
to create a chimeric “human-like” active site for crystallography.
Figure
dimethylbenzyl)amino)-1-phenylethyl)-5-methylisoxazole-3-
carboxamide Compound N-(1-(4-(3-hydroxyprop-1-yn-1-
yl)phenyl)-2-((4-methoxy-2,5-dimethylbenzyl)amino)ethyl)-5-
methylisoxazole-3-carboxamide
3:
Compound
1
N-(2-((4-methoxy-2,5-
2
A 2.8 Å resolution structure of Compound 1 bound to the
StSPL mutant was solved with reasonable refinement statistics
1
1
(
Figure 4). The overall structure had high similarity to the
previously reported (PDB code 3MAD) structure with an overall
r.m.s.d of 0.4 Å. Additionally, when comparing this crystal
An E. coli construct of StSPL 2-570 (Y249F, L344I, F346A,
6
L497S,) was created, which resulted in robust protein expression
8
structure to the recently reported human SPL structure (PDB
and final protein yields. Catalytic activity of this mutant
4
code 4Q6R ), we noticed a strong similarity to backbone and side
bacterial construct was assessed using a substrate analog in
which the C13 hydrocarbon tail of S1P is replaced with an ether-
chain placement in the active site. Electron density was observed
in both active sites and both Compound 1 and PLP could be
modeled in the density. The binding mode of Compound 1
assumes a “wishbone” conformation due to the flexible nature of
the molecule and reaches to several regions in the SPL active site
proximal to the PLP cofactor (Figure 4). The 3D structure
confirms an S-isomer stereochemistry where the ligand is
stabilized by several contacts with the protein. For example, there
is a clear edge-to-face interaction of Compound 1’s central
phenyl ring with nearby StSPL residue Y345. Beyond the central
phenyl is a crystallographic water which lies close to the bound
phosphate and PLP cofactor. The ligand wishbone conformation
9,10
linked umbelliferone group.
M
The K of S1P for StSPL is
approximately 10-fold lower than human SPL, and its Vmax is
lower by a similar degree (Figure 2). The decrease in Vmax may
simply reflect the added stabilization of the SPL-substrate
complex with respect to the reaction transition state, as reflected
M
by the decrease in K . At the standard assay condition of 500 mM
substrate, the two constructs have similar specific activities and
enable a comparison of IC50 values for Compound 1: 0.17 mM for
StSPL and 1.0 mM for human SPL (not shown). Protein crystals

is further stabilized by two hydrogen bonds: the amine N bound
to the backbone carbonyl of I344 and the amide carbonyl to the
backbone N of A346. The isoxazole binds in a shallow channel
outlined by residues from both monomers: F500/L504 (monomer
A) and P348 (monomer B). The methoxy benzyl projects towards
the active site entrance and is surrounded by aromatic residues
Y481 and F500. The N-terminus of the protein was disordered
and as previously hypothesized might serve as a flexible gate
identical solvent network was observed in the human structure
at this location with a succinic acid molecule in the place of
Compound 2’s alkyne alcohol (Figure 6). It was intriguing that
initial structural observations seen in the bacterial structural
surrogate were similar to those in the human SPL crystal
structure.
6
which closes the ligand-bound active site from solvent.
Figure 5: Crystal structure of mutant StSPL and Compound 2
(
orange) gives structural rationale for improved potency via
solvent network. Compound 2 is overlayed with Compound 1
light blue) to show comparable binding modes. The green sphere
Figure 4: Crystal structure of StSPL mutant (grey) bound to
Compound 1
(
illustrates the water identified in the StSPL/Compound 1
structure and replaced in the StSPL/Compound 2 structure.
This initial structure of Compound 1 bound to the surrogate
SPL drove medicinal chemistry design strategies to improve its
inhibition. Several ideas emerged: 1) rigidification of the
wishbone linkages to remove rotational freedom 2) optimization
of both the isoxazole and benzyl methoxy rings to improve
induced fit in their respective binding pockets, and 3) para
substitution off of the central phenyl ring to interact or replace
the crystallographic water near the PLP co-factor. Here, we will
focus on the third strategy.
As mentioned previously, the central phenyl ring is positioned
near a crystallographic water with ample pocket volume for
potential substitution. Superimposing the models of S1P and
Compound 1 suggests a substitution at the para position of the
phenyl ring in Compound 1 to either replace or interact with the
water. Several attempts were made to design into this region
resulting in Compound 2 (Figure 3) which introduces an alkyne
linking segment capped with an –OH to replace the
crystallographic water. This substitution proved to be successful
in moderately improving the IC50 to 0.15 mM. When solving the
structure of Compound 2 bound to StSPL, we were fortunate to
collect higher resolution data with the opportunity to model more
Figure 6: Solvent network in mutant StSPL complexed to
Compound 2 as it compares to human SPL (PDB code 4Q6R).
4
11
water molecules. Compound 2 binds identically to Compound 1
with the alkyne-hydroxyl capping group replacing the water
identified in the first crystal structure (Figure 5). Interestingly,
the hydroxyl taps into a larger solvent network which engages the
the PLP co-factor. Specifically, the group hydrogen bonds to a
nearby water which in turn binds to the phosphate of PLP and
nearby residue Y345. Therefore Compound 2 fills productive
VDW space with its alkyne linkage and improves the active site
solvent network with its hydroxyl. We were happy to see that the
mutant StSPL surrogate crystal structures provided us with a
successful hypothesis regarding the 3D solvent environment.
This in turn helped identify a key potency driver in the SPL
active site. After our structural work was completed, the human
Here we present a surrogate crystallographic approach for the
target protein S1P Lyase. In the absence of well-behaved human
protein, we chose a surrogate crystallographic approach to guide
initial medicinal chemistry design. A construct was designed to
incorporate mutations which mimicked the human active site.
Protein generated from this construct was tested for activity and
used for crystallography. As a result, two key crystal structures
were solved and provided some initial structural hypotheses
which aided the design of subsequent analogs, which is discussed
in the accompanying paper. These compounds and analogs
showed efficacious activity against human SPL providing a lead
chemical series.
4
structure of SPL (PDB code 4Q6R) was published which
.
provided additional retrospective support to our strategy. An

sample (mutant StSPL + PLP complex) was concentrated and
loaded onto a Superdex 200 column (GE Healthcare, Life
Sciences) with Buffer B: 50 mM Phosphate pH 7.2, 100 mM
NaCl, 1 mM EDTA, 1 mM DTT and 0.01 mM PLP. Two major
mutant StSPL peaks eluted corresponding to monomer and
dimeric peaks of StSPL. Dimer-pooled samples were concentrated
to 11 mg/ml and used for crystallography and assay work. Sample
purity was assessed with SDS-PAGE
Acknowledgments
All authors are employees of AbbVie. The design, study conduct,
and financial support for this research were provided by AbbVie.
AbbVie participated in the interpretation of data, review, and
approval of the publication.
9
1
.
Bedia, C.; Camacho, L.; Casas. J.; Abad, J. L.; Delgado, A.; Van
Veldhoven, P. P.; Fabriàs, G. Chem. Bio. Chem. 2009, 10, 820.
The HTS hit Compound 1 was identified by Hetal Patel.
0. Enzyme activity assays were performed by a variation of the
1
3
9
All structural figures were created in the program Pymol.
method developed by Bedia and colleagues. The following
components were added to 96 half-well plates: 20 mL compound
solution in 10% DMSO/90% water, 10 mL 2 mM substrate, and 10
mL 50 nM S1P Lyase with 50 mM pyridoxal-5’-phosphate. Human
S1P Lyase was purchased from BPS Biosciences. Substrate and
enzyme were formulated in 200 mM potassium phosphate, pH 8.0,
0.6 mM EDTA, 0.3 M NaCl, 20% glycerol, 50 mM NaF, and
0.1% Tween-20. Reactions were incubated at room temperature
for 22 hr, following which umbelliferone fluorescence was read
with excitation at 380 nm and emission at 460 nm. IC50 values
were analyzed by least-means-squares fitting to the following
equation: Percent Activity = 100%/(1+([inhibitor]/IC50)). Assay
precision was determined from a set of 30 compounds with repeat
data by calculation of a Test/Re-test Mean Significant Ratio of
Use of the IMCA-CAT beamline 17-ID (or 17-BM) at the
Advanced Photon Source was supported by the companies of the
Industrial Macromolecular Crystallography Association through
a contract with Hauptman-Woodward Medical Research Institute.
Use of the Advanced Photon Source was supported by the U.S.
Department of Energy, Office of Science, Office of Basic Energy
Sciences, under Contract No. DE-AC02-06CH11357.
Supplementary data
1
2
2
.3, according to published methods.
1
1
1. Atomic cooridinates and structure factors have been deposited to
the Protein Data Bank (www.rcsb.org). PDB codes are 5EUE and
Supplementary data (crystallographic data and additional
experimental procedures) associated with this article can be
found, in the online version, at doi:
5
EUD.
2. Beck, G.; Chen, Y.-F.; Dere, W.; Devanarayan, V.; Eastwood,
B.J.; Farmen, M.W.; Iturria, S.J.; Iversen, P.W.; Kahl, S.D.;
Moore, R.A.; Sawyer, B.D.; Weidner, J. Assay Operations for
SAR Support. 2012 May 1 [Updated 2012 Oct 1].
References and notes
Sittampalam, G.S.; Coussens, N.P.; Nelson, H.; Arkin, M.; Auld,
D.; Austin, C.; Bejcek, B.; Glicksman, M.; Inglese, J.; Iversen,
P.W.; Li, Z.; McGee, J.; McManus, O.; Minor, L.; Napper, A.;
Peltier, J.M.; Riss, T.; Trask, O.J. Jr.; Weidner, J.;. editors. Assay
Guidance Manual [Internet]. Bethesda (MD): Eli Lilly &
Company and the National Center for Advancing Translational
Sciences; 2004-. Available from:
1
2
.
.
Schwab, S. R.; Pereira, J. P.; Matloubian, M.; Xu, Y.; Huang,Y.;
Cyster, J. G. Science 2005, 309, 1735.
Vogel, P.; Donoviel, M. S.; Read, R.; Hansen, G. M.; Hazlewood,
J.; Anderson, S. J.; Sun, W.; Swaffield, J.; Oravecz, T.; PLoS One
2
009, 4(1):e4112.
3
.
Bagdanoff, J. T.; Donoviel, M. S.; Nouraldeen, A.; Tarver, J.; Fu,
Q.; Carlsen, M.; Jessop, T. C.; Zhang, H.; Hazelwood, J.; Nguyen,
H.; Baugh, S. D.; Gardyan, M.; Terranova, K. M.; Barbosa, J.;
Yan, J.; Bednarz, M.; Layek, S.; Courtney, L. F.; Taylor, J.;
Digeorge-Foushee, A. M.; Gopinathan, S.; Bruce, D.; Smith, T.;
Moran, L.; O'Neill, E.; Kramer, J.; Lai, Z.; Kimball, S. D.; Liu, Q.;
Sun, W.; Yu, S.; Swaffield, J.; Wilson, A.; Main, A.; Carson, K.
G.; Oravecz, T.; Augeri, D.J. J. Med. Chem. 2009, 52, 3941.
Weiler, S.; Braendlin, N.; Beerli, C.; Bergsdorf, C.; Schubart, A.;
Srinivas, H.; Oberhauser, B.; Billich, A. J. Med. Chem. 2014, 57,
http://www.ncbi.nlm.nih.gov/books/NBK91994/
1
3. The PyMOL Molecular Graphics System, Version 1.2r3pre,
Schrödinger, LLC
4
.
5
074.
Van Veldhoven, P. P.; Mannaerts, G.P. Adv. Lipid. Res. 1993, 26,
9.
Bourquin, F.; Riezman, H.; Capitani, G.; Grutter, M. G. Structure
010, 18, 1054.
5
6
7
.
.
.
6
2
A human SPL homology model was created using the program
InsightII (InsightII, Accelrys Inc., San Diego, USA, 1998)
and using the known crystal structure of 3MAD Symbiobacterium
6
thermophilium SPL (NCBI YP_075103) as a template.
Comparison of the human model and the bacterial crystal structure
showed 4 amino acid positions that were different in the active site
proximal to the location of the PLP binding site. These positions
were mutated in the bacterial S1P Lyase to create a “human-like”
SPL active site using the overall bacterial protein scaffold.
Quadruple mutations Y249F, L344I, F346A, L497S were
introduced to the sequence (2-507) using a pQE70 vector, T5
promotor.
8
.
Recombinant pQE70 Symbio SPL 2-507 (Y249F, L344I, F346A,
L497S)-His6 was produced from a high density expression using
XL1 Blue competent cells after 20 hours of growth induced
culture for 4 hr at 30°C with 2% ethanol and 0.1 mM IPTG. Cell
paste was stored atQuu -80°C and subsequently thawed to 4°C.
Lysis buffer contained 50 mM Phosphate pH 7.2, 2 mM BME, 0.4
mM pyridoxal phosphate (PLP), benzoase and protease inhibitors.
Soluble extract was collected after centrifugation (GSA rotor,
1
7,000 rpm, 25 min). The supernatant was applied to a Nickel
affinity column (GE Healthcare Life Sciences) and mutant StSPL
protein was eluted with 50 mM Phosphate, pH 7.2, 2 mM BME,
0
.4 mM pyridoxal (Buffer A) + 10 mM imidazole. The pooled

*Graphical Abstract (for review)
Graphical Abstract
To create your abstract, type over the instructions in the template box below.
Leave this area blank for abstract info.
Creation of a S1P Lyase Bacterial Surrogate
for Structure-Based Drug Design
a*
b
b
b
b
Maria. A. Argiriadi, David Banach , Elzbieta Radziejewska , Susan Marchie , Jennifer DiMauro ,
c
b
c
c
b
Jurgen Dinges , Eric Dominguez , Charles Hutchins , Russell A. Judge , Kara Queeney , Grier
Wallace and Christopher M. Harris
c
b
a
AbbVie Bioresearch Center, 381 Plantation Street, Worcester, MA 01605
AbbVie Bioresearch Center, 100 Research Drive, Worcester, MA 01605
b
c
AbbVie, 1 North Waukegan Road, North Chicago, IL 60064
Fonts or abstract dimensions should not be changed or altered.