Pharmacological assessment of sepiapterin reductase inhibition on tactile response in the rat

Article abstract of DOI:10.1124/jpet.119.257105

There is an unmet medical need for nonopioid pain therapies in human populations; several pathways are under investigation for possible therapeutic intervention. Tetrahydrobiopterin (BH4) has received attention recently as a mediator of neuropathic pain. Recent reports have implicated sepiapterin reductase (SPR) in this pain pathway as a regulator of BH4 production. To evaluate the role of SPR inhibition on BH4 reduction, we developed analytical methods to monitor the relationship between the plasma concentration of test article and endogenous pterins and applied these in the rat spinal nerve ligation pain model. Sepiapterin is an endogenous substrate, which accumulates upon inhibition of SPR. In response to a potent inhibitor of SPR, plasma concentrations of sepiapterin increased proportionally with exposure. An indirect-effect pharmacokinetic/pharmacodynamic model was developed to describe the relationship between the plasma pharmacokinetics of test article and plasma sepiapterin levels in the rat, which was used to determine an in vivo SPR IC₅₀ value. SPR inhibition and mechanical allodynia were assessed coordinately with pterin biomarkers in plasma and at the site of neuronal injury (i.e., dorsal root ganglion). Upon daily oral administration for 3 consecutive days, unbound plasma concentrations of test article exceeded the unbound in vivo rat SPR IC₉₀ throughout the dose intervals, leading to a 60% reduction in BH4 in the dorsal root ganglion. Despite evidence for pharmacological modulation of the BH4 pathway, there was no significant effect on the tactile paw withdrawal threshold relative to vehicle-treated controls.

Full text of DOI:10.1124/jpet.119.257105

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
Manuscript title
Pharmacological assessment of sepiapterin reductase inhibition on tactile response in the rat
Authors and affiliations
James T. Meyer, Brian A. Sparling, William J. McCarty, Maosheng Zhang, Marcus Soto, Stephen Schneider,
Hao Chen, Jonathan Roberts, Helming Tan, Thomas Kornecook, Paul Andrews, Charles G. Knutson.
Amgen Research, One Amgen Center Drive, Thousand Oaks, CA 91320, United States (JTM, MZ, MS, HT,
TK)
Amgen Research, 360 Binney Street, Cambridge, MA 02142, United States. (BAS, WJM, SS, HC, JR, PA,
CGK)
Public Information
1

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
Running Title:
PK/PD relationship with sepiapterin reductase inhibition
Corresponding Author:
Charles G. Knutson
Amgen Research
Department of Pharmacokinetics and Drug Metabolism
Amgen Inc.
360 Binney Street
AMA1/4-G-10
Cambridge, MA 02142
Tel. 617.444.5470
E-mail: charlie.knutson@amgen.com
Numbers:
Text pages: 34
Number of tables: 3
Number of figures: 4
Number of references: 27
Number of words in Abstract: 241
Number of words in Introduction: 652
Number of words in Discussion: 1262
Public Information
2

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
Nonstandard abbreviations:
AA, Ascorbic Acid
ALS, autosampler
BH4, Tetrahydrobiopterin
BIO, 6-Biopterin
CAD, collision activated dissociation
CE, collision energy
CI, confidence interval
CUR, curtain gas
CXP, collision cell exit potential
DMSO, Dimethyl Sulfoxide
DP, declustering potential
DRG, dorsal root ganglion
FA, Formic Acid
GCH1, GTP cyclohydrolase 1
HCl, Hydrochloric Acid
HPLC, high performance liquid chromatography
I, Iodine
IS, Internal Standard
Public Information
3

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
i.v., intravenous
KI, Potassium Iodide
MSA, methyl sulfonic acid
2
K EDTA, di-potassium ethylenediaminetertra-acetic acid
LC-MS/MS, liquid chromatography tandem mass spectrometry
NH OH, Ammonium Hydroxide
4
RHAM, Rhamnopterin [6-(L-1,2,3-Trihydroxybutyl)-pterin]
RP, reverse phase
SEP, L-Sepiapterin
SNL, spinal nerve ligation
SPR, Sepiapterin Reductase
TOL, Tolbutamide
VER, Verapamil-HCl
Recommended section assignment:
Drug Discovery and Translational Medicine
Public Information
4

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
Abstract
There is an unmet medical need for non-opioid pain therapies in human populations; several pathways are
under investigation for possible therapeutic intervention. Tetrahydrobiopterin (BH4) has received attention
recently as a mediator of neuropathic pain. Recent reports have implicated sepiapterin reductase (SPR) in this
pain pathway as a regulator of BH4 production. To evaluate the role of SPR inhibition on BH4 reduction, we
developed analytical methods to monitor the relationship between the plasma concentration of test article and
endogenous pterins and applied these in the rat spinal nerve ligation pain model. Sepiapterin is an
endogenous substrate, which accumulates upon inhibition of SPR. In response to a potent inhibitor of SPR,
plasma concentrations of sepiapterin increased proportionally with exposure. An indirect effect PK/PD model
was developed to describe the relationship between the plasma PK of test article and plasma sepiapterin
levels in the rat, which was used to determine an in vivo SPR IC50 value. Evaluation of SPR inhibition and
mechanical allodynia was assessed coordinately with pterin biomarkers in plasma and at the site of neuronal
injury (i.e., dorsal root ganglion). Upon QD p.o. administration for 3 consecutive days, unbound plasma
concentrations of test article exceeded the unbound in vivo rat SPR IC90 throughout the dose intervals leading
to a 60% reduction in BH4 in the dorsal root ganglion. Despite evidence for pharmacological modulation of the
BH4 pathway, there was no significant effect on the tactile paw withdrawal threshold relative to vehicle-treated
controls.
Public Information
5

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
Introduction
Chronic pain affects close to 300 million people in the U.S. annually. Due to the lack of therapeutic options,
one common pharmacological intervention is to prescribe opioids. The mortality rate from opioid overdose
exceeds 90 Americans per day (Rudd et al., 2016). Over-prescription of opioids (such as fentanyl) leads to
misuse and addiction that has an increasing burden on societal and health care systems(Volkow et al., 2014).
In effort to reduce the production and use of opioid analgesics, the pharmaceutical industry is investigating
alternative biological pathways of pain reduction, which has led to the rapid prosecution and validation of new
pain targets. Accordingly, several putative targets for analgesia have emerged in recent years; of particular
interest are enzymes that regulate tetrahydrobiopterin (BH4) synthesis. BH4 is reported to enhance pain
sensitivity in the dorsal root ganglia (DRG) in preclinical models with associated links to human
pain(Latremoliere and Costigan, 2011; Latremoliere et al., 2015). Subsequently, drug discovery and
development efforts have emerged to reduce BH4 by developing inhibitors of biosynthetic enzymes involved in
BH4 biosynthesis (Tebbe et al., 2017).
BH4 is an essential cofactor for several enzymes (nitric oxide synthase and aromatic amino acid hydrolases)
and its production occurs through de novo synthesis and enzymatic salvage pathways in vivo (Latremoliere
and Costigan, 2011). GTP cyclohydrolase 1 (GCH1) is elevated during chronic pain and performs the rate-
limiting step in de novo synthesis of BH4 by converting GTP to 7,8-dihydroneopterin-triphosphate
(
Hatakeyama et al., 1991; Tegeder et al., 2006). Sepiapterin reductase (SPR), which is also elevated during
chronic pain, is involved in both de novo biosynthesis and scavenging pathways(Tegeder et al., 2006;
Latremoliere and Costigan, 2011) (Figure 1). Since GCH1 is a master regulator of BH4 production in vivo,
inhibition of GCH1 is likely to significantly reduce systemic BH4 levels, which may lead to undesirable
biological consequences. Therefore, the downstream enzyme, SPR, has emerged as a more desirable target
to reduce BH4 levels because several SPR-catalyzed steps in BH4 production are also catalyzed by other
enzymes (Latremoliere and Costigan, 2011; Latremoliere et al., 2015). Recent efforts have focused on
development of SPR inhibitors that would lead to reduction of BH4 but not its complete elimination. As due
diligence to assess new targets in the discovery pipeline for neuropathic pain, we recently evaluated candidate
Public Information
6

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
inhibitors of SPR. To evaluate our conviction in the target it was necessary to benchmark internal efforts to
those reported in the literature. Therefore, we developed bioanalytical methods to quantify components of the
BH4 biosynthesis pathway to guide our interpretation of pharmacodynamic response related to SPR inhibition
in preclinical behavior models of tactile allodynia.
Using available literature reports, our objective was to develop low-volume, high-throughput analytical methods
for the quantitation of BH4, biopterin (BIO), sepiapterin (SEP), and test article in both plasma and DRG tissues
to quickly make decisions regarding the advancement of molecules. Multiple bioanalysis methods have been
reported for quantitation of BH4, BIO and SEP (Stea et al., 1980; Schmidt et al., 2006; Fismen et al., 2012; Kim
et al., 2012; Arning and Bottiglieri, 2016). The stability of BIO and SEP allow these molecules to be analyzed
by routine protein precipitation methods. However, BH4 is susceptible to autoxidation, which complicates its
direct analysis. Our bioanalytical approach developed two separate analysis arms to account for: 1, direct
measurement analytes (BIO, SEP, test article); 2, indirect measurement of BH4 via oxidative stabilization to
BIO. A total of 6 assays were developed to measure biomarkers and test article in both plasma and DRG
matrixes (Figure 2). These efforts were designed to support early discovery portfolio with efficient use of
resources; thus, the analytical assays were developed as fit-for-purpose without the requirement of analytical
method validation (Cummings et al., 2008; Administration, 2013; Leonard et al., 2014; Cowan et al., 2016).
These biomarker assays for BH4, BIO and SEP were successfully employed for the quantitative measurement
of target engagement (SPR inhibition) in preclinical PK and PK/PD studies.
Public Information
7

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
Methods
Chemicals
Dimethyl Sulfoxide (DMSO), Water for HPLC, Methanol for HPLC, Isopropanol, Acetonitrile for HPLC, Formic
Acid (88-91%), Hydrochloric Acid (37%), Iodine (99.8%), Potassium Iodide USO 99%, Ascorbic Acid,
Ammonium Hydroxide (28-30%), BH4, 6-Biopterin, SEP, Verapamil-HCl (VER), Tolbutamide (TOL),
gabapentin and methyl sulfonic acid were obtained from Sigma-Aldrich (Saint Louis, MO, U.S.). BH4 was
obtained from Combi Blocks (San Diego, CA, U.S.). Rhamnopterin (RHAM), was obtained from Carbosynth Ltd
(
(
Berkshire, UK). Tween 80 was obtained from Croda (East Yorkshire, UK). Hydroxy propyl methyl cellulose
HPMC), 50 cps grade, was obtained from Spectrum Chemicals (New Brunswick, NJ, U.S.). Sprague-Dawley
rat K EDTA plasma and DRG were obtained from BioreclamationIVT (Westbury, NY, U.S.).
2
Chemical Synthesis of SPRi3 and Q-1195
SPRi3 was synthesized in 2 steps from 5-methoxy-2-methyltryptamine via stepwise O-demethylation and N-
acylation with 2-methoxyacetic acid. The synthesis is further described in the Supplemental Methods.
Q-1195 was identified from U.S. Patent 2017/0096435 and a synthetic route was designed (Tebbe et al.,
2017). Briefly, selective hydrolysis of 2,4-dichloropyrrolotriazine followed by nucleophilic displacement with
hydrazine afforded the intermediate hydrazinylpyrrolotriazinone. The condensation product of methyl
acetoacetate with dimethylformamide dimethyl acetal was then coupled with the hydrazinylpyrrolotriazinone to
afford the methyl pyrazolecarboxylate intermediate, which was saponified to the corresponding carboxylic acid
in preparation for amide formation. The amine coupling partner was prepared via reductive amination of Boc-
protected diazabicyclo[3.1.1]heptane with 1,1,-ethoxy(trimetylsilyl)cyclopropane, followed by Boc deprotection.
Amide formation of the acid with the amine afforded Q-1195. The synthesis is further described in the
Supplemental Methods.
Instrumentation and LC-MS Conditions
Public Information
8

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
BIO and SEP LC-MS/MS. The HPLC system contained the following modules: CBM20A controller; solvent
degassing unit; 2 LC20ADXR pumps; SIL30ACMP temperature-controlled autosampler; CTO30A column oven
(
(
Shimadzu, Kyoto, Japan). Chromatographic separation was achieved on a Phenomenex Synergi 4µ Polar RP
80A, 50 x 2 mm) column maintained at 40°C. Mobile phase A consisted of water with 0.002% FA and mobile
phase B consisted of methanol with 0.002% FA. The gradient profile was as follows: 0.0 to 0.2 min at 0% B;
.2 to 1.2 min linear increase to 10% B; 1.2 to 1.3 min linear increase to 100% B; 1.3 to 1.95 min hold at 100%
B; 1.95 to 2.0 min return to 0% B; 2.0 to 2.5 min hold at 0% B. The total run time was 2.5 min at a flow rate of
.5 ml/min. Samples were kept at 15°C in the autosampler. Twenty µl of sample extract was injected with the
0
0
column effluent plumbed to an AB Sciex 4000 triple quadrupole mass spectrometer (AB Sciex, Framingham,
MA, U.S.). The mass spectrometer was operated in negative ion mode turbo spray with the following
parameters: CUR at 20, GS1 and GS2 at 30, gas temperature at 500°C, CAD gas at 12, ion spray voltage
at -4500 V and entrance potential at -10 V. Tuning parameter settings were optimized by post-column infusion
of 1 µg/ml analyte at 1 to 5 µl/ml into a flow of 0.1 ml/min 50% B. All biomarker analytes and internal standards
were scanned in multiple reaction monitoring (MRM) mode using the parameter settings shown in
Supplemental Table 1. Analyst version 1.6.1 software (AB Sciex, Framingham, MA, U.S.) was used for HPLC
control, data acquisition and initial data processing.
Q-1195 LC-MS/MS conditions. Chromatographic separation was achieved on a Phenomenex Synergi 4µ Polar
RP (80A, 50 x 2 mm) column maintained at 40°C. Mobile phase A consisted of water with 0.002% FA and
mobile phase B consisted of methanol with 0.002% FA. The gradient profile was as follows: isocratic hold of
5
% B from 0.0 to 0.15 min; linear increase to 100% B from 0.15 to 0.75 min; isocratic hold at 100% B from 0.75
to 1.25 min; linear decrease to 5% B from 1.25 to 1.3 min; isocratic hold (re-equilibration) at 5% B from 1.3 to
.7 min. The total run time was 1.7 min at a flow rate of 0.6 ml/min. Samples were kept at 15°C in the
1
autosampler over the course of analysis. Two µl of sample extract was injected with the column effluent in line
with an AB Sciex 4000 triple quadrupole mass spectrometer. The mass spectrometer was operated in positive
ion mode turbo spray with the following parameters: CUR at 20, GS1 and GS2 at 30, gas temperature at
500°C, CAD gas at 12, ion spray voltage at 4500 V and entrance potential at 10 V. Tuning parameter settings
Public Information
9

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
were optimized by post-column infusion of 1 µg/ml analyte at 1 to 5 µl/ml into a flow of 0.1 ml/min 50% B.
Analyte and internal standard were scanned in MRM mode using the parameter settings shown in
Supplemental Table 2.
Compound Stock Solutions
BH4, BIO, SEP, RHAM, Q-1195, VER and TOL were weighed in duplicate and dissolved in DMSO:methanol
(
1:1) at 1.0 mg/ml as free base. RHAM required addition of 20 mM HCl to solubilize. Prior to use, the duplicate
.0 mg/ml stock solutions were analyzed with multiple injections (n = 6); the mean area was determined for
1
each stock solution and, if within 20%, then aliquoted at 50 µl per tube and stored at -70°C for up to 4 months.
Standard and Sample Preparation and Analysis
Homogenization of DRG. Due to the challenges in handling the small DRG tissues (1 to 2 mg per DRG), DRG
L4, L5 and L6 from the ipsilateral side were pooled; separately, DRG L4, L5 and L6 from the contralateral side
were pooled from each subject on study. The pooled DRGs were weighed and 9-times their weight in water
was added, representing a 10-fold dilution of the matrix and providing sufficient volume for subsequent sample
preparation steps. Homogenization of the DRGs was accomplished using two 3 mm tungsten carbide beads
per sample in snap cap tubes and processing in the TissueLyzer (Qiagen, Hilden, Germany) with a frequency
of 15 cycles/s for 5 min.
Selection of stock solution solvent. There are varied literature reports of solvents used to prepare stock
solutions of BH4, BIO and SEP. Stock solutions (1.0 mg/ml) were prepared by weighing in duplicate in the
following solvents: water, 25 mM HCl in water, DMSO and DMSO:Methanol (1:1) (Fukushima and Nixon,
1
980a; Zhao et al., 2009; Fismen et al., 2012). DMSO:Methanol (1:1) showed a near 10-fold increase in MS
signal. Subsequently, all stock solutions were prepared in DMSO:Methanol (1:1). Stock solution aliquots (50
L) in individual tubes were capped and stored at -70°C. Stock solution stability was established out to 4

months under these conditions.
Rat plasma. Endogenous levels of BIO and SEP were evaluated in 6 individual lots of blank K EDTA rat
2
plasma using this method. BIO levels were BQL (< 2.5 ng/ml) in 4 of 6 lots; the remaining 2 lots contained 11
Public Information
1
0

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
and 55 ng/ml of BIO. The lots with less than 2.5 ng/ml were reserved for use as blank matrix. SEP levels were
BQL (< 1 ng/ml) in all 6 individual lots of K EDTA rat plasma. These results are similar to previous literature
2
reports (Fukushima and Nixon, 1980a; Schmidt et al., 2006; Latremoliere et al., 2015).
Preparation of standard curves. To facilitate high-throughput analysis of Q-1195, BIO and SEP, both standards
and samples were prepared in 96-well deep block microtiter plates (Microliter Analytical, Suwanee, GA).
Standard curve concentrations at 1.0, 2.5, 5, 10, 25, 50, 100 and 250 ng/ml were prepared in control Sprague-
Dawley rat K
DMSO:Methanol (1:1) into wells containing 20 µl of control plasma. Standard curve concentrations at 10, 15,
5, 50, 100, 250, 500 and 1000 ng/g in DRG homogenate (1:10 diluted in water prior to homogenization) were
2
EDTA plasma. The Tecan D300e was used to dispense analyte from 100 µg/ml stock solution in
2
prepared using the Tecan D300e to dispense analyte from 100 µg/ml stock solution in DMSO:Methanol (1:1)
into 20 µl of control DRG homogenate.
Bioanalysis of test article, BIO and SEP. For determination of Q-1195, BIO and SEP in both plasma and DRG
matrixes, standards and samples were processed as follows: to 20 µl sample, 150 µl of IS was added (VER
and TOL at 25 ng/ml in ACN), vortexed for 30 s and then centrifuged for 5 min at 3200 x g in a Beckman-
Coulter Allegra 25R centrifuge (Danvers, MA, U.S.). A Tomtec Quadra 4 liquid handling system was used to
transfer 125 µl of supernatant to a fresh 96-well deep block microtiter plate. The supernatant was taken to
dryness under 15 psi N gas at 60°C. The dry down procedure required 50 min in an UltraVap (Porvair
2
Sciences, Wrexham, UK). Standards and samples were reconstituted in 100 µl of mobile phase A, sealed with
a pre-slit 96-round-well sealing mat (Phenomenex, Torrance, CA), vortexed for 1 min on a plate shaker (VWR
VX-2500, Manasquan, NJ) and placed in the autosampler.
Bioanalysis of BH4. Autoxidation of BH4 to BIO can complicate direct assessment of BH4 in biological
matrixes. Stabilization of BH4 in its reduced form immediately upon sample collection has been
reported(Fukushima and Nixon, 1980a; Zhao et al., 2009), but due to limited sample volumes and sizes,
sample processing in this manner was determined to be impractical for routine analysis. Other well-established
reports oxidize BH4 all the way to BIO and indirectly determine BH4 levels by quantitation of BIO (Schmidt et
al., 2006); this method was adopted in our analysis. Correspondingly, a separate analysis arm that
Public Information
1
1

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
implemented solid-phase extraction (SPE) was incorporated into our procedure to remove these oxidizing
reagents from downstream workflow so as not to interfere with the stability of other analytes. Plasma and DRG
standards and samples were processed as follows: 20 µl sample was combined with 100 µl of RHAM (25
ng/ml in 0.1 M HCl) then vortexed for 30 s. Fifty µl of oxidizing reagent (10 mg/ml I in 20 mg/ml KI in water)
was added to this mixture, vortexed for 30 s and incubated for 30 min at room temperature in the dark. To stop
the reaction, 100 µl of ascorbic acid at 50 mg/ml in water was added and the plate(s) were vortexed for 30 s.
Note that on addition of ascorbic acid the standards and samples went colorless. Both the ascorbic acid and
iodine solutions were made fresh immediately before use.
SPE method for BH4 bioanalysis. Oasis MCX 96-well 30 µm/30 mg plates (Waters, Wexford, Ireland) were
prepared as follows: 1.0 ml of methanol was added and washed through each well with positive pressure
(
~ 1.0 psi). This was followed by 1.0 ml of water that was washed through each well being careful to ensure
that a small amount of water remained to cover the column head. The oxidized BH4 standards and samples
were transferred to the plate and loaded under ~ 1.0 psi positive pressure. The plates were washed with 1.0 ml
of 0.1 M HCl followed by 1.0 ml of methanol. BIO, from oxidized BH4, was eluted with two 250 µl additions of
4
8% NH OH in 25% methanol in water. The eluates were collected into a fresh 96-well plate and taken to
dryness under N
2
gas (15 psi) at 60°C for 1 h. Standards and samples were reconstituted in 100 µl of mobile
phase A, sealed with a pre-slit 96-round-well sealing mat (Phenomenex, Torrance, CA), vortexed for 1 min on
a plate shaker (VWR VX-2500, Manasquan, NJ,) and placed in the autosampler. Standard curve
concentrations at 10, 15, 25, 50, 100, 250, 500 and 1000 ng/ml in Sprague-Dawley rat K EDTA plasma were
2
prepared using the Tecan D300e dispenser and dispensed BH4 from 100 ug/ml stock solution in
DMSO:Methanol (1:1) into wells containing 20 µl of blank plasma. Standard curve concentrations at 10, 15, 25,
50, 100, 250, 500 and 1000 ng/g in DRG homogenate (1:10 diluted in water prior to homogenization) were
prepared using the Tecan D300e dispenser and dispensed BH4 from 100 µg/ml stock solution in
DMSO:Methanol (1:1) into 20 µl of control homogenate. Standard curves were oxidized and applied to the SPE
for wash and elution as described above for the sample analysis.
Public Information
1
2

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
Standard curve parameters. Standard curves were acquired at the beginning and end of each analytical run.
2
Standard curve linear regression fitting was performed in Watson version 7.4.2 using 1/X weighting. The
inter-assay precision [(S.D./mean)*100] and accuracy [(measured/nominal)*100] of the standard calibrators
across multiple analytical runs are reported in Supplemental Table 3. Supplemental Table 4 summarizes the
standard curve performance in both plasma and DRG matrices.
Fractional Binding
Pooled plasma and brain homogenates from male Sprague-Dawley rats were obtained from BioIVT (Hicksville,
NY). Plasma was treated with lithium heparin and each homogenate was prepared as 1 g of tissue per 4 ml of
phosphate-buffered saline. Plasma or homogenate (1 ml) was spiked with test article to a final concentration of
5
µM. Triplicate 200 µl aliquots were transferred to 220 µl ultracentrifuge tubes and spun at 37°C and 190,000 x
g for 4.5 h in a Type 42.2 rotor (Beckman Coulter). Twenty-five µl aliquots of spiked matrix were transferred to a
sample plate and kept on ice to represent total drug. The remainder of spiked plasma was incubated at 37°C for
the duration of the centrifugation period and used as a stability control. After centrifugation, 25 µl aliquots of the
unbound fractions were transferred to the sample plate. Twenty-five µl aliquots of the stability control plasma
were also transferred to the sample plate. All samples were matched with an equal volume of blank matrix.
Samples were extracted by adding 3 volumes of acetonitrile containing internal standard (1 µM TOL) and 0.1%
formic acid. Samples were vortexed, centrifuged at 4°C for 15 min at 3400 x g and analyzed by LC-MS using a
Thermo Q-Exactive in positive ion mode.
Fraction unbound (f
u
) was calculated from the ratio of test article detected in the water layer after centrifugation
relative to the total concentration in the original matrix. The f
u
in tissue homogenates was corrected for dilution
according to the equation described by Kalvass(Kalvass et al., 2007):
1
ꢀ
)
퐷
f
u2 = (
ꢆ
1
1
ꢁ
ꢂ
ꢃ−ꢄꢅ+ꢀ )
푓푢1
퐷
where fu2 is the adjusted fraction unbound in tissues, fu1 is the estimated fraction unbound assuming no dilution,
and D is the fold-dilution of the tissue.
Public Information
1
3

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
Formulation
Q-1195 was formulated as a solution of 2 mg/ml in dimethyl sulfoxide (DMSO, Sigma-Aldrich, St. Louis, MO)
for intravenous (i.v.) bolus administration in rats. A platform p.o. formulation vehicle of 2% (w/v)
hydroxypropylmethylcellulose (HPMC, 50 cps grade, Spectrum Chemicals) and 1% (w/v) Tween 80 (Croda,
East Yorkshire, UK) in water was used, and Q-1195 was formulated as an in-situ salt in the vehicle by
adjusting to pH 2 with methyl sulfonic acid (Sigma-Aldrich, St. Louis, MO) for oral administration in rats.
In Vivo Experiments
Animals were cared for in accordance to the Guide for the Care and Use of Laboratory Animals, 8th Edition
(
National Research Council (U.S.). Committee for the Update of the Guide for the Care and Use of Laboratory
Animals. et al., 2011). Animals were individually or group-housed at an Association for Assessment and
Accreditation of Laboratory Animal Care, an internationally accredited facility, in static or ventilated caging on
corn cob bedding. All research protocols were approved by the Institutional Animal Care and Use Committee.
Animals had ad libitum access to pelleted feed (Harlan 2020X, IN, USA) and reverse osmosis-purified water
via water bottles or automatic watering system. Animals were maintained on a 12:12 hour light:dark cycle in
rooms at 18 to 26°C and a 30 to 70% humidity range and had access to enrichment opportunities. All animals
were determined free of specific pathogens (http://www.criver.com/files/pdfs/rms/hmsummary.aspx). Animals
were allowed at least 1 week acclimation to the facility prior to any procedures.
Pharmacokinetic In Vivo Experiments
A total of 18 surgically-modified male Sprague-Dawley rats (8 to 12 weeks old) were obtained from Charles
River Laboratories (CA, USA). Rats that were to receive i.v. administration had a femoral vein catheter for
dosing and a jugular vein catheter for blood sampling. Rats that were to receive p.o. administration via gavage
had only a jugular vein catheter for blood sampling. Rats were arbitrarily assigned to treatment groups (n = 3
rats/group) and received either a single 1 mg/kg i.v. bolus dose of Q-1195 administered at a dose volume of
0.5 ml/kg in 100% DMSO or a single 5, 15 or 45 mg/kg p.o. dose of Q-1195 administered at a dose volume of
10 ml/kg in 2% HPMC, 1% Tween 80, pH 2 adjusted with MSA. Q-1195 was also administered orally once
Public Information
1
4

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
daily at 15 mg/kg over 3 subsequent days to a group of rats. For single-dose studies, blood was collected at
.083 (i.v. treatment group only), 0.25, 0.5, 1, 2, 4, 6, 8 and 24 h post-dose. For multiple-dose studies, blood
was collected on Day 1 and Day 3 at 0.25, 0.5, 1, 2, 3, 4, 6, 8 and 24 h post-dose. Blood specimens were
mixed in SARSTEDT Microvette® K EDTA plasma separation tubes and centrifuged at 4°C at 14,000 x g for 5
0
3
min to generate plasma. In multiple-dose studies, lumbar DRG (L4, L5 and L6) were harvested from 2 sub-
groups of rats at 3 h post-dose on Day 1 and Day 3 immediately following carbon dioxide asphyxiation. DRG
L4, L5 and L6 from the ipsilateral side were pooled and DRG L4, L5 and L6 from the contralateral side were
pooled, weighed, and frozen. All plasma and DRG specimens were maintained at -70°C (± 10°C) prior to
quantitative analysis. Non-compartment analysis was applied to the plasma concentration versus time results
to determine pharmacokinetic parameters.
PK/PD Modeling of Q-1195 Exposure to SPR Inhibition
Phoenix WinNonlin/NLME (Certara USA, Princeton, NJ) software was used for the PK and PK/PD modeling.
Parameter imprecision was estimated using 95% confidence intervals (CI) around the mean estimates. The
plasma PK after single administration i.v. (1 mg/kg) and p.o. (5, 15 and 45 mg/kg) dosing was consistent with a
linear extravascular one-compartmental model with first-order elimination. Because lower than dose-
proportional plasma PK were observed at the 45 mg/kg dose, bioavailabilities were estimated for each dose
and the doses were adjusted for bioavailability when fitting the PK/PD model. Sequential PK/PD model fitting
was performed for animals with plasma SEP data. A standard indirect effect model (inhibition of dissipation)
(
Mager et al., 2003) was used to describe the effect of Q-1195 on the time-course of SEP in plasma with Imax
fixed at 1. This model is consistent with the anticipated mechanism of Q-1195 inhibiting the conversion of SEP
to BH4 by SPR, with the plasma levels of SEP as the PD biomarker. The PK/PD model and best-fit parameters
to the 5, 15 and 45 mg/kg data sets were used to project the plasma concentrations of Q-1195 and SEP
following 60 mg/kg p.o. dose administered during the open-field activity and SNL studies (described below).
Rat Open-field Activity
Public Information
1
5

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
Male Sprague-Dawley rats were habituated in a reversed light cycle room in home cages for at least 1 week
and acclimated to the testing room for 1 h prior to dosing. Animals were counterbalanced to n = 8 per
treatment groups. Q-1195 (15 and 60 mg/kg) or vehicle (2% HPMC, 1% Tween 80, pH 2.2, with MSA) was
administered 3 h before placing an animal in the open-field apparatus. Open-field activity was measured using
a system that counts interruptions of a set of photobeams for the course of 30 min (Kinder Scientific, Poway,
CA). To begin a session, animals were removed from the home cage and placed individually into an
independent Plexiglas box (41 L x 41 W x 38 cm H) surrounded by a frame consisting of 32 photocells (16Y
and 16X) that track the movement of the animal. Photobeam breaks were used as an indication of activity and
were measured as the following parameters per min: basic movements (counts) and total rearing (counts).
Following completion of behavioral measurements, animals were euthanized with carbon dioxide followed by
immediate blood collection via cardiac puncture for pharmacokinetic analysis (approximately 3.5 h post-dose).
Blood collections were mixed in SARSTEDT Microvette® K
3
EDTA plasma separation tubes and centrifuged at
°C at 14,000 x g for 5 min to generate plasma. All plasma and DRG specimens were maintained at -70°C (±
0°C) prior to LC-MS/MS analysis. All behavioral data was scored by an automated device.
4
1
SNL-induced Tactile Allodynia
Adult male Sprague-Dawley rats weighing 120 to 150 g were obtained from Harlan (CA, USA) and kept in the
reversed light cycle for the entire study. SNL surgeries were performed under a stereomicroscope using
standard aseptic surgical techniques as described previously (Kim and Chung, 1992). Spinal nerve injury was
caused by ligating the left L5 and L6 (ipsilateral) spinal nerves with special care taken to avoid damage to the
L4 spinal nerve. Briefly, under gaseous anesthesia with a mixture of O
2
and isoflurane (4% for induction and
2% for maintenance), skin was excised and longissimus lumborum muscle and the fascia above L6 spinal
nerves were carefully removed. These procedures provided a clean and spacious working area to enable
complete resection of the L6 transverse process and to separate the L5 from L4 spinal nerve. The L5 and L6
spinal nerves were each tightly ligated with 6-0 silk suture. The entire surgical procedure beginning with skin
incision and ending with wound clipping of the skin lasted 7 min or less.
SNL Behavioral Testing
Public Information
1
6

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
Two weeks post-operation, mechanical sensitivity was measured by determining the median 50% ipsilateral
hind paw withdrawal threshold for von Frey filaments using the up-down method. Rats were placed under a
plastic cover (9 x 9 x 20 cm) on a metal mesh floor (Chaplan et al., 1994). von Frey filaments (Semmes-
Weinstein monofilaments, Stoelting, Wood Dale, IL) were applied to the middle glabrous area between the
footpads of the plantar surface of the injured hind paw. This plantar area was touched with a series of nine
calibrated von Frey filaments with approximately exponentially incremental bending forces (von Frey filament
numbers: 3.61, 3.8, 4.0, 4.2, 4.41, 4.6, 4.8, 5.0 and 5.2; equivalent to 0.41, 0.63, 1.0, 1.58, 2.51, 4.07, 6.31, 10
and 15.8 g). The von Frey filament was presented perpendicular to the plantar surface with sufficient force to
cause slight bending and held for approximately 3 to 4 s. To avoid possible reflex responses, only abrupt
withdrawal of the foot accompanied by pain-indicative behaviors (namely paw flinching, shaking or licking for
more than 2 s) was recorded as a response. Any post-surgery rat that displayed a mechanical threshold of
more than 3.16 g or less than 0.7 g was eliminated from the study.
After measuring basal threshold, animals were randomized into 4 groups (a minimum of 10 animals per group)
and treated orally with either vehicle (single-dosed), Q-1195 (60 mg/kg, single-dosed or multi-dosed once per
day for 3 days) or gabapentin (100 mg/kg, single-dosed). Measurement of the tactile threshold was assessed
at 3 h post-dose for Q-1195-treated and vehicle groups, and at 1 and 2 h post-dose for gabapentin-dosed
animals. Following completion of behavioral measurements, animals were euthanized with carbon dioxide
followed by immediate blood collection via cardiac puncture for pharmacokinetic analysis (approximately 3.5 h
post-dose). Lumbar DRGs (L4, L5 and L6) from the ipsilateral side of individual animals were collected and
pooled in 1.5 ml microcentrifuge tubes; separately, DRG L4, L5 and L6 from the contralateral side were pooled
from each subject into microcentrifuge tubes. Blood collections were mixed in SARSTEDT Microvette®
3
K EDTA plasma separation tubes and centrifuged at 4°C at 14,000 x g for 5 min to generate plasma. All
plasma and DRG specimens were maintained at -70°C (± 10°C) prior to LC-MS/MS analysis.
Human and Rat SPR Protein Production
SPR cloning and expression. Human (NM_003124.5) and rat (NM_019181.1) sepiapterin reductase genes
were synthesized by Integrated DNA Technologies (IDT) and cloned into pET28 vector (Millipore Sigma)
Public Information
1
7

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
containing an N-terminal FLAG tag followed by thrombin cleavage site. The sequence confirmed expression
constructs were transformed into BL21(DE3) and subsequently a single colony of each construct was
inoculated into 60 ml Terrific Broth (TB) medium supplemented with 50 µg/ml kanamycin. The overnight
cultures were used to inoculate 6 l of TB/kanamycin at a 1:100 v/v ratio. The cultures were grown at 37°C by
shaking at 250 rpm until OD600 of ~ 1.0; protein expression was then induced by the addition of 0.5 mM IPTG
(
(
isopropyl-β-D-1-thiogalactopyranoside) for 3 h. Cells were pelleted by centrifugation, washed once with TBS
20 mM Tris pH 8, 150 mM NaCl) and frozen at -80°C until protein purification.
SPR protein purification. Frozen cells were resuspended in TBS with protease inhibitor cocktail (Roche) and
lysed by a microfluidizer. Cell debris was removed by centrifugation at 40,000 rpm at 4°C using a Beckman
Coulter Ti75 rotor. The cleared cell lysates were incubated with 3 ml of anti-FLAG M2 resin (Sigma-Aldrich)
overnight at 4°C with gentle nutation. Resin was packed into a gravity column and washed with 20 column
volumes of TBS, and proteins were then eluted with 20 ml TBS supplemented with 0.15 mg/ml 3X FLAG®
peptide (Sigma-Aldrich). SPR proteins were further purified by size-exclusion chromatography using a HiLoad
16/60 Superdex-200 column equilibrated in Buffer B (20 mM MES pH 6.5, 100 mM NaCl). SPR-containing
fractions were pooled and concentrated to ~ 0.5 mg/ml using Amicon Ultra-15 filters with 10 kDa MWCO.
Biochemical Assessment of SPR Binding
A low volume high throughput NADP/NADPH-GloTM Assay (Promega, Madison, WI) was developed to
measure the NADPH dependent SPR reduction of SEP to 7,8-dihydrobiopterin. The conversion of NADPH to
NADP is proportional to the conversion of SEP to 7,8 dihydrobiopterin by SPR. NADP/NADPH-GloTM
technology generates a luminescent signal upon reaction with reductase reducing the proluciferin substrate to
luciferin. Luciferin is then detected with Ultra-GloTM rLuciferase (Promega, Madison, WI) where the light
generated is proportional to the NADPH in the sample. Inhibition of SPR is assessed by the inhibition of
luciferin signal in response to increasing concentrations of inhibitor.
Assay plates were prepared by dispensing either human FLAG-SPR or rat FLAG-SPR to a final concentration
of 10 nM and 3 nM, respectively, in assay buffer (25 mM HEPES, pH 7.5, 150 mM NaCl, 0.01% Tween-20)
Public Information
1
8

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
containing 10 nM SEP into a Corning 384-well white, low volume plate with a Thermo Multidrop Combi.
Inhibitors were dissolved in 100% DMSO to a top concentration of 5 mM and then serially diluted (1:2
stepwise) with 100% DMSO to generate a 22-point curve in a standard 384-well polypropylene plate. Assay
buffer supplemented with 10 µM NADPH was added to the inhibitor titration plate and the reaction was initiated
by transferring from the inhibitor titration plate to the assay plate using a Thermo Vprep liquid dispenser. The
final top concentration of inhibitor was 50 µM in the assay. The reaction was incubated for 40 min at room
temperature and stopped by the addition of sulfasalazine (to a final concentration of 10 µM). The detection
buffer NADP/NADPHGloTM was prepared and added to the stopped reaction as recommended by the
manufacturer (omitting the use of NADP cycling enzyme). After 45 min of incubation the plate was read on an
Envision 2104 Multimode Plate Reader (PerkinElmer) using the ultra-sensitive luminescence detection method.
Statistical Analyses
Since the von Frey filament set was calibrated on a logarithmic scale by the vendor (Stoelting), our selection of
9
filaments for the up-down method was also based on nearly equal logarithmic intervals, and because it is our
experience that variability noticeably increases with threshold value (Dixon, 1980), data were analyzed
following logarithmic transformation prior to statistical analysis. Actual gram values were plotted on a gram
scale Y-axis of the figures for convenience. Behavioral data are expressed as S.E.M. Results were analyzed
using one-way analysis of variance (ANOVA) with Dunnett’s multiple comparisons post hoc test for significance
relative to vehicle. Statistical calculations and graphs were made using GraphPad Prism 7.01 (GraphPad
Software Inc., San Diego, CA, USA).
Public Information
1
9

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
Results
Chemical Synthesis and Biochemical Inhibition of SPR Inhibitors
Literature evaluation of chemical matter revealed several molecules that have demonstrated biochemical
inhibition of SPR as well as reduction in BH4 in vivo. Synthesis of these molecules was initiated; concurrently
we developed a biochemical assay to assess in vitro potency of inhibition. The SPR NADPH-Glo assay IC50
determinations were consistent with reported values in Table 1 (Zhao et al., 2009; Latremoliere et al., 2015;
Tebbe et al., 2017). Q-1195 was determined to be the most potent inhibitor of those tested. Thus, Q-1195 was
moved forward into in vivo characterization to determine its suitability as a tool molecule to probe for behavioral
and biochemical changes associated with SPR inhibition in vivo.
Bioanalytical Method
Accurate and reliable bioanalytical methods were required to monitor the pterin biomarkers associated with
SPR inhibition. A coordinated series of method development steps were undertaken in a fit-for-purpose matter
during bioanalytical method development. Several bioanalytical methods provided the basis for internal
development efforts(Fukushima and Nixon, 1980b; Schmidt et al., 2006). Details of these investigations may
be found in the Methods section. Briefly, stock solutions were prepared in 1:1 DMSO:Methanol, which resulted
in a near 10-fold increase in MS response compared to aqueous and DMSO-alone solutions. A shallow LC
gradient with methanol and water (containing 0.002% formic acid) was selected with separations performed on
a Phenomenex Polar RP column. Stabilization of BH4 to BIO was achieved by oxidation of matrix (plasma or
DRG) using I/KI in HCl solutions. Biological samples (plasma and DRG) were split and processed by two
routes: 1, protein precipitation with Q-1195 detection in positive ion mode followed by immediate reanalysis of
the samples for BIO and SEP in negative ion mode; 2, oxidation (I/KI in HCl) followed by SPE trapping for
detection of BH4 measured as BIO (Figure 2).
Protein Binding Assessment
Fractional binding to plasma proteins and neural tissues was assessed by ultracentrifugation in effort to
determine the unbound fraction (f
u
) in separate tissues. Brain homogenates were used as a surrogate for
Public Information
2
0

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
DRG (Liu et al., 2018). Attempts to determine fu,DRG were made, but these efforts proved challenging because
the DRG homogenate would seize, making it unusable to liquid transfers. Fractional binding studies produced
f
u,plasma = 0.20 and fu,brain = 0.32.
PK Results
Q-1195 was administered to male Sprague-Dawley rats by i.v. and p.o. routes of administration and results are
summarized in Table 2. Following i.v. administration (1.0 mg/kg), Q-1195 underwent first-order elimination,
displayed low clearance (0.063 l/h/kg) and a long half-life (4.5 h). A low Vss (0.4 l/kg) indicated that Q-1195 was
primarily restricted to plasma water. P.o. administration (5 mg/kg) demonstrated good bioavailability (86.7%)
with a Cmax of 26.3 µM at 2 h post-dose. Based on these initial findings, Q-1195 was deemed suitable for
subsequent p.o. dose-escalation studies. P.o. dose escalation at 15 mg/kg and 45 mg/kg indicated less than
dose-proportional increases in AUC (Table 2). Additionally, repeat dosing of Q-1195 at 15 mg/kg for 3 days
demonstrated minimal dose accumulation (Day 1 AUCinf = 434 µM*h; Day 2 AUCinf = 509 µM*h). A PK model
was developed using Phoenix WinNonlin/NLME and adequately described the PK of Q-1195 (Figure 3);
2
predicted concentrations correlated well with measured values (R = 0.91). The plasma PK was 80% of dose-
proportional between 5 and 15 mg/kg, and 46% of dose-proportional between 5 and 45 mg/kg, suggesting
decreased bioavailability. The model estimations of bioavailability were consistent with this trend, fitting to
100% (CI: 80 to 120%) at 5 and 15 mg/kg and 58% (CI: 53 to 63%) at 45 mg/kg. PK parameter values and CI
estimated with the model are shown in (Table 3).
PK/PD Results
Concomitant with the assessment of plasma concentrations of Q-1195, SEP was measured in plasma for each
dose group (i.v. and p.o.). The appearance of SEP in plasma increased in response to increased p.o. exposure
of Q-1195, indicating Q-1195 was engaged with its target (SPR) and that SPR inhibition was measurable in the
plasma compartment (Figure 3). An indirect PK/PD model adequately described the relationship between
plasma PK of Q-1195 and the plasma SEP levels (Figure 3). Predicted SEP levels correlated with measured
2
values (R = 0.86). The PK/PD derived in vivo SPR IC50 for inhibition of the conversion of SEP to BH4 by SPR
Public Information
2
1

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
was 2.2 µM (Table 4), with the corresponding unbound in vivo SPR IC50 = 0.44 µM. The PD response was 80%
dose-proportional between 5 and 15 mg/kg and 74% dose-proportional between 5 and 45 mg/kg,
demonstrating that increased Q-1195 exposure led to an increase in SEP concentration. We then used the
indirect PK/PD model to project a dose that would cover the in vivo SPR IC90 at Cmin for use in subsequent pain
studies. Results from the model projection indicated that a Q-1195 dose of 60 mg/kg dose yields plasma
concentrations at Cmin exceeding 2.2 µM.
To determine the extent of SPR inhibition at the presumed site of action (DRG), SEP was measured in excised
DRG from intact rats. After p.o. dosing at 15 mg/kg, SEP levels in DRG increased from undetectable
(
< 15 ng/g) in sham-dosed animals to 26.0 ng/g on Day 1 and 50.9 ng/g on Day 3. The corresponding plasma
concentrations of SEP were 20.8 and 19.1 ng/ml on Day 1 and Day 3, respectively. Q-1195 was measured in
the same DRG samples and found to be 9.1 nmol/g (9.1 µM) on Day 1 and 11.8 nmol/g (11.8 µM) on Day 3.
The corresponding plasma concentrations of Q-1195 were 29.6 and 30.7 µM on Day 1 and Day 3,
respectively.
Q-1195 Did Not Affect Open-field Activity
Potential analgesic effects of a new chemical entity can be confounded by behavioral changes that produce
sedative or motor side effects and may lead to a false positive response. Therefore, prior to testing in a pain
behavioral model, we evaluated Q-1195 in an open-field assessment to observe general behavior changes
after exposure to test article. Following 15 and 60 mg/kg p.o. administration of Q-1195, the mean plasma
concentrations at 3.5 h post-dose were 79.8 and 193 µM, respectively, while the SEP plasma concentration
was 41.1 and 59.8 ng/ml, respectively. The Cu,plasma was 38.6 µM at the 60 mg/kg dose indicating that an
8.6-fold coverage of the unbound in vivo SPR IC90 was achieved near Cmax. The 60 mg/kg projection from the
PK/PD model matched reasonably well to the data with both the Q-1195 and SEP falling within 2-fold of the
prediction (Figure 3). The measured Q-1195 plasma concentrations were 1.5-fold higher than projected
(
130 µM), and the measured plasma SEP concentrations were 1.6-fold higher than projected (37.9 µM). At 3 h
post-dose, basic movement counts (5532 ± 165.7) were not significantly different relative to the vehicle-treated
group (5406 ± 386.2), and total rearing counts (184.4 ± 14.8) were not significantly different relative to the
Public Information
2
2

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
vehicle-treated group (173.3 ± 15.1) (Figure 4A & 4B). Similarly, at the lower dosing group (15 mg/kg) no
significant effect on basic movements or total rearing counts were observed. Results from the open-field
assessment indicate that any behavioral change in the SNL assessment would likely be attributable to a
pharmacological effect rather than an overt behavioral change.
Q-1195 Did Not Reverse SNL-induced Tactile Allodynia
Following no observable effect in the open-field assessment, Q-1195 was assessed for its ability to reverse
tactile allodynia in the SNL rat pain model. Q-1195 was administered at 60 mg/kg p.o. for 3 consecutive days
to SNL rats (n = 10); testing was assessed on Day 1 and reassessed on Day 3 to evaluate the
pharmacological response from acute and repeat administration, respectively. The typical baseline, pre-
surgery threshold for tactile response was 13.5 g. Two weeks after SNL surgery, sham-treated rats exhibited a
tactile threshold of 2.53 ± 1.45 g (3 h post-single-dosing of vehicle) and 2.30 ± 1.71 g (3 h post-multi-dosing of
vehicle on Day 3) indicating SNL surgery produced the expected endpoint of reduced pain threshold.
Gabapentin (100 mg/kg p.o.) was administered as a positive control to separate cohorts of SNL rats on Day 1
or Day 3 of the experiment. Previous open-field assessment of gabapentin at this dose level was determined to
have no impact on basic movements or rearing counts (negative in the open-field assessment). Gabapentin
significantly reversed SNL-induced tactile allodynia with an increase in the tactile threshold to 8.94 ± 3.95 g
(
Day 1) and 10.12 ± 4.45 g (Day 3). Q-1195 at 60 mg/kg p.o. did not increase the tactile threshold relative to
vehicle on either Day 1 or Day 3 of this multi-dose study (p > 0.05 by Dunnett’s multiple comparisons test) with
observed mean tactile thresholds of 2.61 ± 1.45 g and 2.94 ± 1.71 g on Day 1 and Day 3, respectively (Figure
4C).
Biochemical Assessment of SPR Inhibition in the SNL Rat
In addition to the von Frey behavioral response assessment of tactile threshold, biochemical endpoints of SPR
inhibition (BH4 and SEP) were also evaluated. Plasma samples were collected after behavioral assessment
(
3.5 h post-dose) on Day 1 and Day 3 of the study and DRG were collected only at the termination of the study
on Day 3. Vehicle-dosed animals were run as the negative control. Analysis of plasma samples (3.5 h post-
Public Information
2
3

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
dose) from Day 1 and Day 3 demonstrated Q-1195 concentrations of 241 µM and 231 µM, respectively (Figure
D). These plasma levels were consistent with (and slightly higher than) the measured plasma concentrations
4
from the 60 mg/kg p.o. dose in the open-field activity assessment. The unbound in vivo IC90 coverage was
determined to be 11-fold on both Day 1 and Day 3. Correspondingly, an increase in plasma SEP levels was
observed in Q-1195-treated rats relative to the vehicle-dosed controls. SEP plasma concentrations increased
from undetectable (LLOQ 1 ng/ml) in vehicle-treated rats to mean plasma levels of 46.1 ng/ml and 55.6 ng/ml
on Day 1 and Day 3, respectively, following administration of Q-1195 (Figure 4E).
The anticipated site of pharmacological effect for Q-1195 is the SPR expressed in the DRG. The SNL
procedure was performed on the left-side of the rat spine and was designated as the ipsilateral side, and the
non-surgical (right-side) was labeled as the contralateral. Because SNL surgery was only applied to the
ipsilateral side, the contralateral side served as an internal control to monitor changes in BH4. Quantitation of
BH4 in the ipsilateral DRG of vehicle-dosed animals demonstrated a 6.7-fold increase relative to the
contralateral DRG (from 18.5 to 124.7 ng/g) (Figure 4H), which indicated SNL surgery led to elevated BH4
levels as previously reported (Latremoliere et al., 2015). This is also in agreement with the decrease in tactile
threshold observed during von Frey testing (described above). Analysis of the ipsilateral DRG from Q-1195-
dosed animals demonstrated a 59% reduction in BH4 levels relative to sham-dosed controls (from 124.7 to
51.5 ng/g). Despite this decrease in BH4 in Q-1195-exposed rats, there was no significant increase in tactile
threshold observed during von Frey testing. In good agreement with the decrease in BH4 in the ipsilateral DRG
following exposure to Q-1195, levels of SEP were significantly increased over sham-dosed DRGs from BQL to
156.8 ng/g in ipsilateral DRG and 57.6 ng/g in contralateral DRG (Figure 4G). Analysis of DRG on Day 3
revealed the presence of Q-1195 at levels of 26.8 µM (Figure 4F). The Cu,DRG for Q-1195 was 8.6 µM 3.5 h
post-dose, which was in excess of the unbound in vivo SPR IC90 by 2-fold.
Discussion
To support the early pipeline and evaluate novel non-opioid targets for analgesia, we developed a
fit-for-purpose analysis suite to quantify both test article and relevant target-specific biomarkers in a
pharmacological model of pain sensitivity. The requirement to generate data for multiple analytes from
Public Information
2
4

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
low-volume rodent samples presented several analytical challenges. The methods reported here
accommodated the analysis of BIO and SEP in plasma using the same negative ion acquisition method. Q-
1195, along with other test articles, in plasma was successfully analyzed by immediate re-injection of the
sample plates employing a positive ion acquisition method. BIO, SEP and Q-1195 were quantitated in DRG
matrix using the same methods used in the analysis of plasma samples. The oxidation reaction implemented to
convert BH4 to BIO required a second sample preparation event in both plasma and DRG samples for
quantitation of BH4. The methods reported here allowed for the rapid turnaround of bioanalytical results for
BH4, BIO, SEP and Q-1195 in both plasma and DRG samples. The turnaround time for the complete analysis
of a behavioral (SNL) study including sample preparation, data analysis and report generation was
approximately 4 days.
The DRG matrix presented unique sample handling challenges. Homogenates were prepared for both matrix
spiking of analytical standard curves as well as for the analysis of in vivo samples. Homogenates of DRG were
typically prepared as 2:1 mixtures of PBS:tissue. Gradually, these mixtures would seize and become
unworkable to any subsequent liquid transfers, and it became impossible to determine fractional binding in
DRG by ultracentrifugation because a total recovery sample could not be reproducibly transferred due to these
severe matrix effects. Homogenates diluted 10:1 in PBS:tissue produced similar results. However, when total
recovery was desired, precipitation and extraction with organic solvent (acetonitrile) was successful as
evidenced by the linearity of spiked standard curves used for bioanalysis. Liu et al. recently reported strong
correlation in fractional binding among neural tissues (brain, spinal cord, sciatic nerve and DRG) by using rapid
equilibrium dialysis (Liu et al., 2018). Dialysis was not attempted in our analysis, but it’s possible that
congealing events observed in our studies may be overcome using dialysis, provided the solidified DRG matrix
permitted small molecule equilibrium to be achieved between dialysis solution reservoirs.
Application of these bioanalytical methods to preclinical studies enabled reliable determinations of test article
and pterin metabolites in both plasma and DRG. Collectively, this provided a quantitative basis to derive
PK/PD relationships and assess pharmacological inhibition of SPR in a rat model of neuropathic pain. Using
the PK analysis and PK/PD projections, the unbound plasma concentrations from a 60 mg/kg p.o. dose in the
Public Information
2
5

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
SNL experiment was designed to exceed the in vivo SPR IC90 at Cmin over the dose interval. Plasma
quantitation of Q-1195 in the SNL experiment demonstrated the unbound plasma concentrations exceeded the
unbound in vivo SPR IC90 by 11-fold at 3.5 h post-dose (the approximate Cmax). The PK model projection at
3.5 h post-dose projected 6.8-fold coverage of the in vivo SPR IC90. Thus, we inferred that unbound plasma
concentrations exceeded the in vivo unbound IC90 at Cmin as well and was consistent with the projections.
The opportunity to assess target engagement (SPR inhibition) through SEP appearance in a typical p.o. PK
experiment was very attractive. The PK/PD model produced a reliable prediction of SPR inhibition and likely
would have been a useful tool in a drug discovery campaign, provided a relationship between PD and efficacy
was established. This approach would have minimized the reliance on SNL studies, which are time and
resource intensive. Prioritizing compounds based off both PK and PD early in the discovery flow scheme likely
would have provided an efficient use of resources, where by the SNL studies would have been reserved for
true lead candidates, and greater confidence in the outcomes would be expected.
There was a large difference in potency observed between the calculated in vitro rat SPR IC50 (0.004 µM) and
the in vivo SPR IC50 (2.2 µM and unbound in vivo SPR IC50 = 0.44 µM) of approximately 110-fold. Prior studies
demonstrated a strong right-shift in potency from biochemical to cell for human SPR with SPRi3 of
approximately 70-fold (not adjusted for fractional binding in the cell-based assay) (Latremoliere et al., 2015).
Large shifts in potency from biochemical to cell-based assays are common; further, shifts from cell-based
potency to in vivo potency are frequently right-shifted. The biochemical, recombinant protein-based assay is
largely a measure of potency and may not fully account for mode of action and intracellular compartment
effects that would be increasingly prevalent from cell culture to in vivo. In the absence of a reliable cell-based
potency assessment, we used the calculated in vivo assessment of SPR inhibition (measured by the increase
in SEP in plasma) to benchmark the target coverages for the determination of efficacy in the SNL investigation.
Despite biochemical evidence for inhibition of SPR in both plasma and DRG, there was no observable increase
in tactile paw withdrawal to SNL rats. The reason for this lack of behavioral change is unclear, but several
explanations are possible. A possible disconnect between the unbound plasma coverage and behavior effect
may result from unanticipated differences in unbound concentrations of test article in the DRG. SEP, BH4 and
Public Information
2
6

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
Q-1195 were all quantified in DRG; increased levels of SEP were observed as well as a decrease in BH4,
which suggested that SPR inhibition was achieved in DRG. Total concentrations of Q-1195 were measured in
DRG, but the unbound fraction estimates were determined from brain matrix studies (fu,brain). Even though there
appears to be a strong relationship between fractional binding in brain relative to DRG (Liu et al., 2018), it’s
possible that some chemotypes will deviate from this relationship. If the actual fu,DRG was 10-fold lower than
f
u,plasma then it’s conceivable that in DRG only the in vivo SPR IC50 was exceeded at Cmin. However, fractional
binding was also evaluated in spinal cord and nerve homogenates with Q-1195 and these values were within
-fold of the fu,brain value (data not shown), which is consistent with prior reports (Liu et al., 2018). Penetration
2
into the tissue could also impact the local concentration; however, in vitro permeability assessment of Q-1195
in parental MDCK cell monolayers indicated favorable passive permeability (Papp = 15 µcm/s, data not shown),
which should not be limiting for tissue distribution. BH4 levels in healthy DRG were generally in the 20 to 30
ng/g range. SNL surgery caused an increase in BH4 that was greater than 100 ng/g. There was often a tight
S.D. around the measured levels but a large range of values within a given experiment. This heterogeneity in
response may also have impacted our ability to effectively reduce BH4. A 60% reduction in BH4 was observed
following exposure to Q-1195 and it’s possible that a larger magnitude of BH4 reduction is required to return
animals to their tactile threshold baseline. It’s worth noting that even for the positive control, gabapentin, there
was a large variation in tactile threshold response (ranging from 2.32 to 15.0 g). A large sample number (n =
10 rats) for all SNL groups was used to gain a better estimate of the sample mean. Though not evaluated in
this study, the selection of rodent strains and vendors can impact physiological and behavioral endpoints in
preclinical studies. Additionally, several neuropathic pain models are used in preclinical research; it’s possible
that varied results could be observed with SNL, spared nerve injury or chronic constriction injury models.
Behavioral models often exhibit large variability in response (pain models included) that present additional
obstacles in the drug discovery pipeline.
Public Information
2
7

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
Acknowledgements
Special thanks to Sarah O’Connor for careful editing of the final manuscript. Thanks to Dean Hickman and
Chris James for timely discussions and scientific review of the research findings. For assistance with the open-
field activity assessment we thank Beth Youngblood and Dave Matson.
Public Information
2
8

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
Authorship contributions
Participated in research design: Meyer, Sparling, McCarty, Zhang, Soto, Schneider, Tan, Kornecook, Andrews,
Knutson
Conducted experiments: Meyer, Sparling, McCarty, Zhang, Soto, Schneider, Chen, Roberts, Tan
Contributed new reagents or analytical tools: Meyer, Sparling, McCarty, Schneider, Chen
Performed data analysis: Meyer, Sparling, McCarty, Zhang, Schneider, Roberts, Tan, Kornecook, Andrews,
Knutson
Wrote or contributed to the writing of the manuscript: Meyer, Sparling, McCarty, Zhang, Soto, Schneider, Chen,
Roberts, Tan, Kornecook, Andrews, Knutson
Public Information
2
9

JPET Fast Forward. Published on May 20, 2019 as DOI: 10.1124/jpet.119.257105
This article has not been copyedited and formatted. The final version may differ from this version.
JPET 257105
References
Administration USFaD (2013) Guidance for Industry: Bioanalytical Method Validation.
Arning E and Bottiglieri T (2016) LC-MS/MS Analysis of Cerebrospinal Fluid Metabolites in the Pterin Biosynthetic
Pathway. JIMD reports 29:1-9.
Chaplan SR, Bach FW, Pogrel JW, Chung JM, and Yaksh TL (1994) Quantitative assessment of tactile allodynia in the rat
paw. Journal of neuroscience methods 53:55-63.
Cowan K, Gao X, Parab V, Nguy T, Wu L, Arron JR, Townsend M, Wallin J, Cheu M, Morimoto A, and Wakshull E (2016)
Fit-for-purpose biomarker immunoassay qualification and validation: three case studies. Bioanalysis.
Cummings J, Ward TH, Greystoke A, Ranson M, and Dive C (2008) Biomarker method validation in anticancer drug
development. British journal of pharmacology 153:646-656.
Dixon WJ (1980) Efficient analysis of experimental observations. Annual review of pharmacology and toxicology 20:441-
4
62.
Fismen L, Eide T, Djurhuus R, and Svardal AM (2012) Simultaneous quantification of tetrahydrobiopterin,
dihydrobiopterin, and biopterin by liquid chromatography coupled electrospray tandem mass spectrometry.
Anal Biochem 430:163-170.
Fukushima T and Nixon JC (1980a) Analysis of reduced forms of biopterin in biological tissues and fluids. Anal Biochem
1
02:176-188.
Fukushima T and Nixon JC (1980b) Chromatographic analysis of pteridines. Methods Enzymol 66:429-436.
Hatakeyama K, Inoue Y, Harada T, and Kagamiyama H (1991) Cloning and sequencing of cDNA encoding rat GTP
cyclohydrolase I. The first enzyme of the tetrahydrobiopterin biosynthetic pathway. The Journal of biological
chemistry 266:765-769.
Kalvass JC, Maurer TS, and Pollack GM (2007) Use of plasma and brain unbound fractions to assess the extent of brain
distribution of 34 drugs: comparison of unbound concentration ratios to in vivo p-glycoprotein efflux ratios.
Drug metabolism and disposition: the biological fate of chemicals 35:660-666.
Public Information
3
0