α-[Amino(4-aminophenyl)thio]methylene-2-(trifluoromethyl)benzeneacetonitrile; Configurational equilibria in solution

Article abstract of DOI:10.1016/j.bioorg.2021.104955

Inconsistent results have been reported for the effects of the mitogen-activating extracellular kinase (MEK) inhibitor α-[amino(4-aminophenyl)thio]methylene-2-(trifluoromethyl)benzeneacetonitrile (SL 327) on ethanol-induced conditioned place preference (EtOH-CPP). Since such inconsistencies may be due to the configurational composition of administered SL 327, the interconvertibility of the geometric isomers of this class of compounds has been investigated. This study provides conditions for determination of configurational composition of this class of compounds by HPLC and by ¹H NMR and reports details of configurational equilibria as a function of medium and time in solution along with solubility data for SL 327 in aqueous DMSO. The results suggest that the apparently inconsistent results reported for CPP-EtOH may be due to the administration of suspension vs. solutions, as well as to different configurational compositions of SL 327.

Full text of DOI:10.1016/j.bioorg.2021.104955

Bioorganic Chemistry 113 (2021) 104955
Contents lists available at ScienceDirect
Bioorganic Chemistry
journal homepage: www.elsevier.com/locate/bioorg
α
-[Amino(4-aminophenyl)thio]methylene-2-(trifluoromethyl)
benzeneacetonitrile; Configurational equilibria in solution
Charles J. McElhinnya, Anita H. Lewina *, Larry Brieaddya, Scott Fixa, Gregory H. Imlerb,
,
Jeffrey Deschampsb, S. Wayne Mascarellaa, Herbert H. Seltzmana, P. Anantha Reddya,
F. Ivy Carrolla
a Research Triangle Institute, P.O. Box 12194, Research Triangle Park, NC 27709, United States
b United States Naval Research Laboratory (NRL), Laboratory for Biosensors and Biomaterials, Center for Biomolecular Science and Engineering, Washington, DC
20375, United States
A R T I C L E I N F O
A B S T R A C T
Keywords:
Inconsistent results have been reported for the effects of the mitogen-activating extracellular kinase (MEK) in-
SL 327
hibitor
α-[amino(4-aminophenyl)thio]methylene-2-(trifluoromethyl)benzeneacetonitrile (SL 327) on ethanol-
MEK inhibitor
E/Z configuration
Solubility
induced conditioned place preference (EtOH-CPP). Since such inconsistencies may be due to the configura-
tional composition of administered SL 327, the interconvertibility of the geometric isomers of this class of
compounds has been investigated. This study provides conditions for determination of configurational compo-
sition of this class of compounds by HPLC and by 1H NMR and reports details of configurational equilibria as a
function of medium and time in solution along with solubility data for SL 327 in aqueous DMSO. The results
suggest that the apparently inconsistent results reported for CPP-EtOH may be due to the administration of
suspension vs. solutions, as well as to different configurational compositions of SL 327.
Formulation
1. Introduction
study of the configurational preferences of SL 327 (1a), and of the
analog 1b, in physiologically relevant solutions are described below.
The potent and selective mitogen-activating extracellular kinase
(MEK) inhibitor SL 327 ( -[amino(4-aminophenyl)thio]methylene-2-
α
2. Materials and methods
(trifluoromethyl)benzeneacetonitrile, 1a) has been reported to impair
acquisition of cocaine-, Δ9-tetrahydrocannabinol (THC)-, and 3,4-meth-
ylenedioxymethamphetamine (MDMA)-induced conditioned place
preference (CPP) in mice.[1,2,3] Somewhat inconsistent results have
been reported for the effects of SL 327 on ethanol-induced CPP (EtOH-
CPP). Thus, SL 327 was reported to neither impair acquisition of EtOH-
CPP, nor to interfere with extinction of EtOH-CPP,[4] but more recent
studies have claimed that SL 327 significantly blocks the acquisition of
EtOH-CPP, but not the acquisition of conditioned place aversion (CPA).
[5] Since SL 327, the only systemically administered MEK inhibitor,[4]
is capable of existing in both the E and the Z configurations, these in-
consistencies may be due to differences in the configuration associated
with each administered formulation. We therefore undertook to study
the configurational preferences of SL 327 (1a) and of a simplified
2.1. Materials
With exception of DMSO and THF all solvents were from Fisher
Scientific; DMSO, THF, Et3N, CuI, NH4OAc, propanedinitrile, 4-amino-
benzenethiol, and 4-methylbenzenethiol were from Sigma Aldrich. 1-
Iodo-2-trifluoromethylbenzene was from Alfa Aesar, K2CO3 was from
EM Scientific, CDCl3 was from Cambridge isotope laboratories Inc, and
DMSO‑d6 and CD3OD were from Acros Organics. The deionized water
used was obtained from a Millipore Milli-Q Plus Ultra-Pure Water
System.
2.2. Instrumentation
comparator molecule,
α-[amino(4-methylphenyl)thio]methylene-2-
Melting points were determined on a Thomas-Hoover capillary tube
apparatus. Thin layer chromatography was carried out using Merck
Silica Gel 60F254 TLC plates; visualization was under UV. High pressure
(trifluoromethyl)benzeneacetonitrile (1b), in which the aromatic amino
substituent was replaced by a methyl group (Fig. 1). The results of our
* Corresponding author.
https://doi.org/10.1016/j.bioorg.2021.104955
Received 4 February 2021; Received in revised form 21 April 2021; Accepted 27 April 2021
Available online 4 May 2021
0045-2068/© 2021 Elsevier Inc. All rights reserved.

C.J. McElhinny et al.
Bioorganic Chemistry 113 (2021) 104955
2.3.1. (
α
-[amino(4-aminophenyl)thio]methylene-2-(trifluoromethyl)
benzeneacetonitrile (1a)
2.3.1.1. 2-[2-[(trifluoromethyl)phenyl]propanedinitrile (4). A mixture of
1-iodo-2-trifluoromethylbenzene (2) (9.77 g, 0.036 mol), propane-
dinitrile (3) (4.95 g, 0.075 mol), K2CO3 (20.73 g, 0.150 mol), and CuI
(0.714 g, 0.004 mol) in DMSO (100 mL) was stirred at 120 ◦C for 21 h,
then cooled to RT and water (500 mL) and EtOAc (250 mL) was added.
The organic phase was separated, treated with water (500 mL), and 6 N
HCl was added until acidic to litmus paper. The organic phase was
separated, dried over sodium sulfate, filtered, and concentrated in vacuo
to afford a dark red oil. The crude oil was purified by column chroma-
tography (SiO2; hexane/EtOAc 70:30) giving 4 as a yellow solid (3.46 g,
46% yield). 1H NMR (300 MHz, CDCl3); δ 5.31 (s, 1H, ArCH), 7.67 (t, J
= 7.7 Hz, 1H, ArH), 7.80 (m, 2H, ArH), 7.94 (d, J = 7.8 Hz, 1H, ArH).
Fig. 1. Isomers of SL 327 (1a) and its analog 1b.
liquid chromatography was carried out on a system comprised of two
Rainin Dynamax Pumps Model SD-300 equipped with a Rheodyne
Model 7725i Injector, and a Varian ProStar 330 Detector; the system was
controlled using Varian Star Version 5.5.
2.3.1.2. (
α
-[amino(4-aminophenyl)thio]methylene-2-(trifluoromethyl)
stirring solution of 2- [2- [(tri-
Chromatographic purification was performed on a Teledyne ISCO
CombiFlash Rf. Nuclear magnetic resonance (NMR) spectra were
recorded on a Bruker Avance 300 MHz or a Varian Unity Inova 500 MHz
NMR spectrometer. Microanalyses were performed by Atlantic Micro-
lab, Inc. X-ray diffraction analysis was carried out by the Naval Research
Laboratory.
benzeneacetonitrile, (1a). To
a
fluoromethyl) phenyl] propanedinitrile (4) (5.49 g, 0.026 mol) and 4-
aminobenzenethiol (5a) (3.23 g, 0.026 mol) in THF (45 mL) at ꢀ 78 ◦C
was added Et3N (2.61 g, 0.026 mol). After slowly warming to RT the
solution was stirred for 21 h. The solvent was concentrated in vacuo to
afford a viscous yellow oil that was purified by column chromatography
(SiO2; hexane/Me2CO 1:1). The product-containing fractions were
concentrated in vacuo to an oil that solidified upon standing. Trituration
of the solid with Et2O gave 1a as an off-white solid (2.08 g, 22% yield):
MP 145–146 ◦C; TLC (SiO2, hexane/Me2CO (1:1), UV single spot Rf 0.52.
Anal. calcd for C16H12 F3N3S: C 57.31, H 3.61, N 12.53. Found C 57.48,
H 3.51, N 12.39. 1H NMR (300 MHz, CDCl3); δ 3.98 (bs, 2H, NH2), 4.78
(bs, 2H, ArNH2), 6.67 (m, 2H, SArH), 7.26 (m, 2H, SArH), 7.51 (m, 3H,
ArH), 7.76 (d, J = 7.8 Hz, 1H, ArH). X-ray analysis confirmed the E
configuration,
2.3. Synthesis
The compounds 1a and 1b were synthesized following the literature
protocol described for SL 327 (1a) (Scheme 1).[6] Since the synthetic
intermediate 4 and the product 1b have not been previously charac-
terized both were rigorously characterized.
Scheme 1. Synthesis of SL 327 (1a) and its analog 1b.
2

C.J. McElhinny et al.
Bioorganic Chemistry 113 (2021) 104955
2.3.2.
α
-{Amino[(4-methylphenyl)-thio]methylene}-2-(trifluoromethyl)
final R1 value of 3.56% on the observed data and wR2 of 9.17% on all
data (Fig. 2). Complete information on data collection and refinement is
available in the supporting information.
benzeneacetonitrile (1b)
To a stirring solution of 2- [2- [(trifluoromethyl) phenyl] propane-
dinitrile (4) (2.00 g, 0.009 mol) and 4-methylbenzenethiol (5b) (1.17 g,
◦
0.009 mol) in THF (20 mL) at ꢀ 78 C was added Et3N (0.95 g, 0.009
mol). The solution was allowed to slowly warm to RT and stirring was
continued for 21 h. The solvent was concentrated in vacuo affording a
heterogeneous white solid in a yellow oil. The mixture was sonicated in
Et2O (15 mL) and the resulting white solid was collected and dried (1.73
g). The mother liquor was concentrated down to ca 15 mL depositing a
second crop of white crystals (0.37 g) after overnight at RT, for an
overall yield of 67%; MP 150–151 ◦C; TLC (SiO2, hexane/Me2CO (1:1),
UV) single spot Rf 0.72. Anal. calcd for C17 H13F3N2S: C 61.07, H 3.92, N
8.38. Found C 60.99, H 3.97, N 8.48. 1H NMR (300 MHz, CDCl3); δ 2.39
(s, 3H, CH3), 4.07 (bs, 2H, NH2), 7.25 (d, overlapped by CDCl3, J = 7.5
Hz, 2H, SArH), 7.48 (m, 4H, SArH + ArH), 7.61 (t, J = 7.6 Hz, 1H, ArH),
7.77 (d, J = 7.9 Hz, 1H, ArH). X-ray analysis confirmed the Z
configuration,
2.4.2. α-{Amino[(4-methylphenyl)-thio]methylene}-2-(trifluoromethyl)
benzeneacetonitrile (1b)
Single-crystal X-ray diffraction data on 1b were collected using
MoKα radiation (λ = 0.71073 Å) and a Bruker APEX-2 CCD area de-
tector. The crystal was prepared for data collection by coating with high
viscosity microscope oil. The oil-coated crystal was mounted on a micro-
mesh mount (Mitergen, Inc.) and transferred to the cold stream on the
diffractometer; diffraction data were collected at 150 K. The 0.210 ×
0.176 × 0.067 mm3 crystal was monoclinic in space group P21, with unit
cell dimensions a = 7.6833(17), b = 8.761(2), c = 23.197(5) Å, and β =
90.029(3)◦. The data were 97.7% complete to 29.22◦ θ (~0.73 Å) with
an average redundancy of 1.87. The structure was solved by direct
methods and refined by full-matrix least squares on F2 values using the
programs found in the SHELXTL suite (Bruker, SHELXTL v6.10, 2000,
Bruker AXS Inc., Madison, WI). Corrections were applied for Lorentz,
polarization, and absorption effects. Parameters refined included atomic
coordinates and anisotropic thermal parameters for all non-hydrogen
atoms. Hydrogen atoms on carbons were included using a riding
2.4. X-ray crystallography
2.4.1. (α-[amino(4-aminophenyl)thio]methylene-2-(trifluoromethyl)
–
model [coordinate shifts of C applied to H atoms] with C H distance set
benzeneacetonitrile, (1a)
at 0.96 Å (Fig. 3). The final anisotropic full matrix least-squares
refinement on F2 with 406 variables converged to a final R1 value of
11.10% for the observed data and wR2 of 34.21% on all data. The crystal
was determined to be a two-component pseudo-merohedral twin using
the program Platon. The ratio of the major to the minor component was
approximately 80:20. All anisotropic displacement parameters were
restrained with enhanced rigid-bond restraints (esds = 0.001) and to be
approximately isotropic (esds = 0.01). Bond lengths in the benzene rings
C1-C6 and C1′-C6′ were restrained to 1.39 Å (esds = 0.01). Complete
information on data collection and refinement is available in the sup-
porting material.
The off-white solid 1a was recrystallized from (CH2Cl)2 by slow
evaporation of the solvent. Single-crystal X-ray diffraction data on 1a
were collected at 173 K using MoKα radiation and a Bruker APEX 2 CCD
area detector. A 0.26 × 0.15 × 0.09 mm3 crystal was prepared for data
collection coating with high viscosity microscope oil (Paratone-N,
Hampton Research). The oil-coated crystal was mounted on a micro-
mesh mount (MiTeGen Inc., Ithaca, NY) and transferred immediately
to the cold stream on the diffractometer. The crystal was monoclinic in
space group P21/c with unit cell dimensions a = 13.6596(4) Å, b =
5.48190(10) Å, c = 21.4135(6) Å, and β = 107.682(1)◦. Corrections
were applied for Lorentz, polarization, and absorption effects. Data were
99.4% complete to 28.28◦ θ (approximately 0.74 Å) with an average
redundancy of 2.5. The structure was solved by direct methods and
refined by full-matrix least squares on F2 values using the programs
found in the SHELXTL suite (Bruker, SHELXTL v6.10, 2000, Bruker AXS
Inc., Madison, WI). Parameters refined included atomic coordinates and
anisotropic thermal parameters for all non-hydrogen atoms. Hydrogen
atoms on carbons were included using a riding model [coordinate shifts
2.5. High pressure liquid chromatography
Solutions of 1 were analyzed using a 20 min gradient of 80% → 30%
5 mM ammonium acetate in water/MeCN (9:1) to 5 mM ammonium
acetate in water/MeCN (1:9), followed by a 20 min hold, at 1.0 mL/
min., on a Luna-C18 (250 × 4.6 mm, 5 μm) column, detecting at λ 270
–
of C applied to H atoms] with C H distance set at 0.96 Å (Fig. 2). The
nm. Retention times were: 20.955 min for E-1a; 24.088 min for Z-1a;
27.688 min for E-1b; 34.680 min for Z-1b.
asymmetric unit contains a single molecule. The structure refined to a
Fig. 2. The structure of SL 327 (1a) as determined by X-ray diffraction. Displacement ellipsoids are shown at the 30% probability level.
3

C.J. McElhinny et al.
Bioorganic Chemistry 113 (2021) 104955
Fig. 3. The structure of 1b as determined by X-ray diffraction. Displacement ellipsoids are shown at the 30% probability level. The asymmetric unit contained two
molecules of 1b. For clarity only one molecule has been shown.
3. Results and discussion
aqueous acid (DCl/D2O) to the solution of SL 327 in CD3OD led to
downfield shifts and appeared to have only small effects on the area
ratio (compare Supp. Fig. 2a and 2d). Dissolution of 1a (5 mg) in CD3OD
Our preparation of the potent and selective mitogen-activating
extracellular kinase (MEK) inhibitor SL 327 (1a), (Scheme 1) was car-
ried out by reaction of 2-[2-(trifluoromethyl)phenyl]propanedinitrile
(4) with 4-aminothiophenol (5a), following the patent literature,[6]
with the exception that it was carried out on a larger (3-fold) scale; the
reaction yielded a two-component mixture (TLC, silica gel, acetone/
hexane), in agreement with the literature.[6] However, whereas, the
literature reported column chromatography (silica gel, ethyl acetate/
hexane) to afford two products, (6% yield of the faster eluting and 41%
yield of the slower eluting)[6] the fractions collected from column
chromatography (silica gel, acetone/hexane) of our product contained
only one component (22% yield). The literature[6] identified the faster
eluting compound to be the E-isomer; the 1H NMR spectra of the E- and
Z-isomers were reported to be essentially identical except for the
chemical shifts of the thiophenyl protons which were at 6.66 and 7.24
ppm for the E-isomer and at 6.68 and 7.40 ppm for the Z-isomer.
Comparison of the 1H NMR spectrum of our product, as recorded in
deuterochloroform (CDCl3), suggested that our product was the faster
eluting E-isomer, consistent with the reported[6] spectrum of E- 1a,
exhibiting the expected[6] resonances for the thiophenyl protons at 6.68
and 7.26 ppm (Supp. Fig. 1).
(650
μL)/D2O (150
μ
L) gave a 1H NMR spectrum with a 1:1 area ratio of
the upfield doublets (Table 1). These observations were difficult to
interpret, particularly since 1a possesses two amino groups subject to
protonation. Thus, the simple interpretation that the configurational
composition of 1a is changed by altering the pH of the medium could be
wrong if the chemical shifts of the configurational isomers were pH
dependent. To facilitate interpretation of the results it was decided to
synthesize the comparator molecule
α-[amino(4-methylphenyl)thio]
methylene-2-(trifluoromethyl)benzeneacetonitrile (1b), in which the
aromatic amino substituent has been replaced by a methyl group.
The synthesis of 1b was carried out by the condensation of 2-[2-[(tri-
fluoromethyl)phenyl]propanedinitrile (4) with 4-methylbenzenethiol
(5b) following the procedure utilized to prepare 1a (Scheme 1). Pure
1b, which crystallized upon sonication in ether, was analyzed by X-ray
crystallography to determine its configuration. Comparison of the
observed X-ray crystallographic intensities versus the calculated in-
tensities of 1b showed evidence of unresolved twinning. Attempts to
find a crystal without twinning were unsuccessful. The crystal was
determined to be a two-component pseudo-merohedral twin (51:49) and
the program PLATON[7] was employed to determine a twin law (-1
To unambiguously determine the configuration of the product iso-
lated by us (1a), it was decided to analyze the sample by X-ray crys-
tallography. Since crystals of the synthetic product 1a as isolated were
found to be unsuitable for diffraction, 1a was recrystallized from 1,2-
dichloroethane. Analysis of the recrystallized solid by X-ray crystallog-
0 0 0 1 0 0 0–1). The structure of 1b (Fig. 3) was found to be Z-α-[amino
(4-methylphenyl)thio]methylene-2-(trifluoromethyl)benzeneacetoni-
trile (Z-1b).
Analysis of the CDCl3 solution prepared from crystalline 1b by 1H
NMR showed the expected doublets for the thiophenyl protons at 7.24
pm and 7.49 ppm, consistent with the Z-configuration (Supp Fig 3).
Likewise, the 1H NMR spectrum recorded in CD3OD (Supp. Fig 4a)
showed predominantly the Z-1b configuration(93.5%), as did a solution
raphy confirmed the product to be E-α-[amino(4-aminophenyl)thio]
methylene-2-(trifluoromethyl)benzeneacetonitrile (E-1a) (Fig. 2),
consistent with the literature.[6,7]
Although 1H NMR analysis showed the configuration of SL 327 in
CHCl3 solution to be the same as in the crystalline solid, i.e. the E-
configuration, it was important to determine what the preferred
configuration of SL 327, which is insoluble in neutral water, would be in
a physiologically relevant medium. Examination of the 1H NMR spec-
trum of SL 327 in perdeuterated methanol (CD3OD) showed the area of
the resonance at 6.59 ppm, which is associated with E-1a, to correspond
to 71% of the area expected for 2 hydrogens and the area of a resonance,
at 6.62 ppm, (associated with Z-1a) to correspond to 29% of the area
expected for 2 hydrogens (Supp. Fig. 2a). Addition of one drop of water
(D2O) resulted in downfield shifts and appeared to alter the area ratio of
the upfield doublets from 3:1 to ca 2:3 (compare Supp. Fig. 2a and 2b).
Addition of 1 drop of deuterated aqueous sodium hydroxide (NaOD/
D2O) to the solution of SL 327 in CD3OD had no apparent effect on the
chemical shifts but changed the area ratio of the upfield doublets to
approximately 1:3 (compare Supp. Fig. 2a and 2c). Addition of a drop
of 1b in CD3OD/D2O (650:100 μL) (Supp Fig 4b). Lowering the pH of the
aqueous methanol solution of 1b by the addition of deuterated aqueous
hydrochloric acid (DCl/D2O) had no effect on the chemical shifts and
only slightly increased the percentage of Z-1b (Supp. Fig 4d) suggesting
that protonation was not taking place. Raising the pH of the aqueous
methanol solution of 1b by the addition of deuterated aqueous sodium
hydroxide (NaOD/D2O, pH 12) also had no effect on the chemical shifts,
as expected from the DCl results, but, surprisingly, increased the per-
centage of E-1b from 6.5% to 28% (Supp. Fig 4c). Because this unex-
pected result was observed after the sample had been in solution
overnight, the effect of time in solution on the configurational compo-
sition of both 1a and 1b was investigated.
Monitoring of the area ratio of the thiophenyl resonances for solu-
tions of crystalline 1a and 1b in CDCl3, CD3OD and deuterated dimethyl
sulfoxide (DMSO‑d6) over a 48-hr period, at room temperature, showed
that in CDCl3 the configuration of both 1a and 1b remained unchanged
4

C.J. McElhinny et al.
Bioorganic Chemistry 113 (2021) 104955
Table 1
1H NMR Determination of % Z-1 as a Function of Medium and Time.
Time (min.)
% Z-1a
% Z-1b
CDCl3
CD3OD
DMSO‑d6
CD3OD/D2O (6.5:1)
CDCl3
CD3OD
DMSO‑d6
CD3OD/D2O (6.5:1)
15*
0
0
0
0
33
69
69
69
62
71
71
70
50
100
100
100
100
96
73
73
73
74
74
74
74
70
360
ND
ND
ND
ND
70
1440
2880
ND
*
minimum time required to prepare the solution and record the spectrum
from that of those of crystalline solid. In CD3OD and DMSO‑d6 E-1a
rapidly isomerized to afford solutions containing 33% and 62% Z-1a,
respectively, and afforded solutions containing ca 70% Z-1a after being
in solution at room temperature for 6 h. For 1b, rapid isomerization, to
afford the equilibrium value of 74% Z-1b, occurred in DMSO‑d6; in
CD3OD the equilibrium value of 73% Z-1b was reached in less than 6 h
(Table 1).
results for the analog 1b using the normal phase HPLC system showed
that for a MeOH solution prepared from crystalline 1b some isomeri-
zation took place within 5 min while virtually no isomerization occurred
for a CHCl3 solution prepared from crystalline 1b (Table 2). To improve
the peak shape and to enable analysis of DMSO solutions, use of a
reversed-phase column (Luna C18) was examined. It was found that
using a gradient of aqueous MeCN buffered at pH 7 to analyze an “aged”
These observations highlighted the need for a simple chromato-
graphic method that would allow researchers to determine the config-
urational composition of the formulations used in their experiments. To
ensure that isomerization would not occur during HPLC analysis, the
configurational stability of 1a and 1b in several solvents was examined
by thin layer chromatography (TLC). As expected, analysis of a CHCl3
solution of crystalline 1a on silica-gel plates, eluting with acetone
(Me2CO)/hexane gave a single spot. Solutions of crystalline 1a in Me2CO
also gave a single spot, indicating that 1a was not undergoing isomeri-
zation in Me2CO solution. By contrast, analogous analysis of solutions of
crystalline 1a in MeOH, or in acetonitrile (MeCN), gave two separate
spots, consistent with isomerization of 1a, as had been seen for meth-
anolic solutions by 1H NMR. TLC analysis of a CHCl3 solution of the
residue obtained by evaporating the solvent from a MeOH solution of
crystalline 1a, gave the same two spots observed in the TLC analysis of
the MeOH solution of crystalline 1a. Similar results were observed for
solutions of 1b. These results show that whereas crystalline 1 adopts a
single configuration, the configurational composition of solid (non-
crystalline) 1 depends on the solvent from which it was obtained.
Since 1a and 1b were found to maintain their crystalline configu-
rations in CHCl3, Me2CO and hexanes, it was attempted to develop a
normal-phase HPLC system to analyze and monitor the configurational
composition of 1a and 1b. While a system employing a silica gel (Zorbax
(5 h) solution of 1b in MeOD gave two well separated peaks (RT =
27.576 and 34.549 min) representing 73.4% Z- and 26.6% E-1b (Supp
Fig 5b) in agreement with the results obtained by 1H NMR analysis
(Table 1), whereas a solution freshly prepared from crystalline 1b in
MeOH gave 98.1% Z-1b (Supp Fig 5a), confirming that no significant
isomerization was taking place during analysis, despite the fact that
solutions of 1b in MeCN were found to isomerize (Table 2). Using this
HPLC system, which enabled the determination of the configurational
composition immediately after dissolution, a time course was deter-
mined for the interconversion of the E- and Z forms of the model com-
pound 1b. The results showed that in both MeOH (Supp. Fig 5a) and
DMSO (Supp. Fig 6a) the configurational preference of the solid (Z-1b)
was largely retained for at least 5 min and an equilibrium mixture,
containing 73.4% Z-1b in MeOH and 76.6% Z-1b in DMSO, was reached
in 3 to 5 h (Table 2; Supp. Figs 5b and 6b). Due to solubility issues,
solutions of 1b in CHCl3 were only analyzed using normal phase con-
ditions; a very small amount (3.3%) of isomerization to E-1b was seen to
occur after 24 h. A time course study of DMSO solutions prepared from
crystalline 1a showed that immediately after dissolution the sample
consisted of predominantly E-1a (as in the solid, Supp. Fig 7a; RT
=
23.933 min) but isomerized within 1 h to a mixture containing 67.8% Z-
1a (RT = 20.867 min) and reached equilibrium (70% Z-1a) in 2 h (Supp.
Fig 7b). Slower isomerization was seen in MeOH: isomerization to
37.9% Z-1a was seen within one hour and equilibrium (67.9% Z-1a) was
approached after 6–7 h. As further validation of the suitability of these
HPLC conditions a solution prepared from crystalline 1a in MeCN/buffer
SIL, 5
μ 4.6x250 mm) column and Me2CO/hexane as eluent gave
reasonably good separation of the geometric isomers of 1a and 1b, the
peaks were broad and misshapen, hampering integration. Preliminary
Table 2
HPLC Determination of the Time Course for Equilibration of 1.*
Time (min)
%Z-1b
MeOH
Zorbax
%Z-1a
DMSO
Luna
CHCl3
MeOH
Luna
DMSO
Luna
MeCN/Buffer (1:1)
Luna
Luna
Zorbax
5
97.6
89.7
81.5
78.5
74.6
72.0
ND
98.1
ND
97.0
ND
100
100
100
100
100
100
ND
ND
ND
ND
ND
96.7
ND
ND
ND
ND
7.7
12.3
ND
9.7
ND
25.0
ND
ND
ND
ND
ND
ND
26.8
ND
ND
ND
ND
43.1
ND
30
ND
60
86.7
ND
77.6
ND
37.9
ND
67.8
ND
90
120
150
180
240
300
360
420
1440
1500
5760
14,400
20,160
79.9
ND
76.8
ND
53.6
ND
70.3
ND
75.9
74.5
73.4
73.5
ND
76.6
76.6
ND
60.9
64.9
66.7
67.7
67.9
ND
70.3
70.4
ND
ND
ND
ND
ND
ND
ND
ND
ND
72.0
ND
ND
ND
ND
ND
ND
68.2
ND
ND
ND
ND
76.5
ND
ND
ND
ND
ND
69.9
ND
ND
ND
ND
69.0
*
Zorbax SIL 5
μ
4.6x250 mm; Me2CO/hexane; Luna C18
5μ
4.6x250 mm; A: 5 mM NH4OAc/MeCN (9:1), B: 5 mM NH4OAc/MeCN (1:9), 20% B->70% B over 20 min
5

C.J. McElhinny et al.
Bioorganic Chemistry 113 (2021) 104955
(1:1) was found to isomerize very slowly, consistent with the absence of
significant isomerization during HPLC analysis (Table 2). Interestingly,
although the Z/E ratio of 1a in MeOH and DMSO was ca 7:3, the % Z-1a
in the solution prepared from crystalline 1a using aqueous MeCN buff-
ered to pH 7 to exceed 50% even after 10 days at room temperature
implying that the equilibrium ratio of Z/E in this medium may be around
1:1, in contrast to the Z/E ratio observed in MeOH and DMSO.
the supernatant indicated that approximately 9% of the total SL 327 (1a)
was in solution (0.14 mg/mL) and that the configurational composition
of the dissolved 1a was 63.6% Z-1a after one hour, changing to 78.6% Z-
1a after two hours. Assuming that animals were dosed with a solution,
and not with a suspension, the administered 1a may have been 9% of the
intended dose (if the sample were injected through a syringe filter)
which could account for no effects having been seen. On the other hand,
if the animals were, in fact, dosed with a suspension, solid SL 327 in the
suspension would, most likely, possess the E-configuration (E-1a), as
was found by X-ray crystallographic analysis of crystalline SL 327. This
configuration would likely persist as the suspended SL 327 interacted
with the lipophilic membrane surface, since we had found that config-
urational equilibration does not take place in a lipophilic environment
(e.g. CHCl3). If that were the case, the observation that SL 327 had no
effect in blocking the acquisition of EtOH-CPP[4] would suggest that,
whereas Z-1a is active in blocking the acquisition of EtOH-CPP, as re-
ported by Rosas,[5] E-1a is not.
Having established the time dependence of the configurational
composition of 1, the 1H NMR spectra of solutions prepared from crys-
talline SL 327 (1a) in CD3OD (Supp. Figure 8a), CD3OD/D2O (650:100
μ
L) (Supp. Figure 8b), CD3OD/D2O (650:100
μ
L) + NaOD (5
μL) (Supp.
Fig 8c) and CD3OD/D2O (650:150
μ
L) + DCl (5 L) (Supp. Figure 8d)
μ
recorded after allowing for sufficient equilibration time were compared.
This re-investigation revealed that whereas addition of 13% water to a
methanolic solution of 1b had no effect on the configurational compo-
sition Supp Fig. 4a, 4b), the addition of water dramatically altered the
configurational composition of 1a (Supp. Fig 8a, 8b). The1H NMR
spectrum of 1a in aqueous CD3OD did not change between pH 3 and pH
11, i.e. no changes were observed for either the chemical shifts or the
areas of the aromatic protons (Supp Fig 8b-8d), as had been observed for
the analog 1b (Supp. Fig 4b-4d). Although of no physiological signifi-
cance, it was noted that further lowering of the pH induced significant
changes in the chemical shifts, suggesting that protonation of the aniline
was occurring, and, by contrast to the previous finding where the effect
of time was not controlled for, significantly lowered the % Z-1a (from
70% Z-1a at neutral pH to 40% Z-1a at pH 1).
The observed insolubility of SL 327 in aqueous DMSO calls into
question results of experiments using aqueous DMSO formulations of SL
327. For example, since our study demonstrated that when the water
content of an aqueous DMSO solution of SL 327 reaches 60%, the use of
a formulation of SL 327 in 25% DMSO[1] for dosing animals in a study of
the involvement of the extracellular signal-regulated kinase cascade for
cocaine-rewarding properties suggests that the dosing consisted of a
suspension, or, if the dosing were performed using a syringe filter, less
than the calculated dose would have been delivered. This could also be
the case for the study that reported that extracellular signal-regulated
kinase (ERK) inhibition does not prevent the development or expres-
sion of tolerance to, and dependence on, morphine in mice, where SL
327 was administered at 5 mg/mL in 50% DMSO/0.9% saline[17], since
saline would decrease the solubility of SL 327 (salt effect).
The sum total of these observations may have a profound impact on
the interpretation of results of studies carried out with the MEK inhibitor
SL 327 (1a) and chemically related analogs. Since this class of com-
pounds is totally insoluble in water, 1a, which is the only systemically
administered MEK inhibitor,[4] has been reported to be dissolved in
various media including DMSO,[3,8] aqueous DMSO,[1,2] absolute
ethanol diluted in 40% polyethylene glycol,[9] Tween-20,[10] aCSF
containing 1% DMSO,[11] 10% Cremaphor (Sigma-Aldrich) and 90%
saline,[12] or to be suspended in 45% β-cyclodextrin.[13] For prepa-
ration of an injectable formulation aqueous propylene glycol, made
isotonic with sodium chloride, has been recommended[6] and for oral
administration a suspension in sodium carboxymethylcellulose, sodium
benzoate, sorbitol and vanillin has been described.[6]
Since we had noted that deionized water was, at times, slightly acidic
(pH 5) a small amount (3%) of concentrated hydrochloric acid was
added to the 15% DMSO mixture containing the precipitated solid,
which led to complete dissolution of the solid. Analysis of this acidified
solution showed the configurational composition of 1a•HCl to be 81%
Z. If the deionized water used to dilute the DMSO solutions of SL 327
were acidic (as has been noted in our labs) the desired dose of 1a,
consisting of ca 80% Z-1a would have been administered.
Our observation that solutions prepared by dissolving E-
α
-[amino(4-
While the observations that although replacement of the aromatic
amino group of 1a, which had crystallized in the E-configuration, by a
methyl group resulted in an analog (1b) that crystallized in the Z
configuration, DMSO and MeOH solutions of this analog contain
approximately the same Z/E ratio (3:7) observed for 1a are striking, this
is probably a coincidence, and should not be interpreted as suggesting a
general preference for the Z configuration for this class of compounds in
DMSO and MeOH solutions.
aminophenyl)thio]methylene-2-(trifluoromethyl)benzeneacetonitrile
(E-1a) in DMSO initially (time = 0) contain 15% of Z-1a, changing over
time to 70% Z-1a, indicates that an abundance of caution must be used
in comparing and extrapolating experimental results. For example, since
stereochemistry has been shown to play a role in pharmacological ef-
fects,[14,15] as well as in physiologic distribution,[16] the observed
effects of solvent and time in solution on the configurational composi-
tion of 1a could serve to explain the apparently contradictory results
reported for the effect of 1a on the acquisition of EtOH-CPP. In a study
that reported no effect of SL 327 (1a) on the acquisition of EtOH-CPP,[4]
aqueous solutions containing 15% and 50% DMSO that were prepared
fresh daily by dissolving 1a in 100% DMSO followed by dilution with
deionized H2O, were used, while a more recent study,[5] which reported
significant inhibition by 1a on acquisition of EtOH-CPP, used a vehicle
made of DMSO, Cremaphor and isotonic saline 30/30/40% (v/v). In
neither case was the time between preparation of the formulation and
administration of the sample reported. Having developed the required
HPLC methodology to rapidly determine the configurational composi-
tion of SL 327 in DMSO, we undertook to mimic the formulations uti-
lized in the study of the effects of SL 327 on the acquisition, and
extinction, of EtOH-CPP.[4] To our disappointment (and surprise) we
were unable to prepare a solution of SL 327 at 1.5 mg/mL in 15% DMSO.
Thus, slow addition of deionized water to a DMSO solution of SL 327 (10
mg/mL) resulted in a milky suspension, at 40% DMSO, which persisted
throughout the further addition of water. With continued addition of
water to reach 15% DMSO, a white precipitate settled out. Analysis of
4. Conclusions
The results of the above reported study have shown that an abun-
dance of caution must be applied in using formulations of the enamines
exemplified by SL 327 (1a) and its analog 1b and in interpretation of the
results obtained by using such formulations. We had found that
(a) although the MEK inhibitor
α-[amino(4-aminophenyl)thio]
methylene-2-(trifluoromethyl)benzeneacetonitrile (SL 327, 1a)
has been reported to have been administered in aqueous DMSO,
SL 327, which is completely insoluble in neutral water, has
minimal (~9%) solubility in aqueous medium containing 40%
(and less) DMSO
(b) the solubility of SL 327 in 15% DMSO/85% neutral water is 0.14
mg/mL
(c) the configurational composition of a solution of SL 327 (1a)
immediately after its preparation mimics the configurational
composition of the solid
6

C.J. McElhinny et al.
Bioorganic Chemistry 113 (2021) 104955
[2] E. Valjent, C. Pages, M. Rogard, M.J. Besson, R. Maldonado, J. Caboche, Delta 9-
tetrahydrocannabinol-induced MAPK/ERK and Elk-1 activation in vivo depends on
dopaminergic transmission, Eur J Neurosci 14 (2) (2001) 342–352.
[3] J. Salzmann, C. Marie-Claire, S. Le Guen, B.P. Roques, F. Noble, Importance of ERK
activation in behavioral and biochemical effects induced by MDMA in mice, Br J
Pharmacol 140 (5) (2003) 831–838.
(d) the configurational composition of solid SL 327 depends on the
solvent from which it was obtained; crystallized SL 327 (1a) is the
E- isomer
(e) SL 327 (1a) in DMSO solution isomerizes over 2–3 h to 70% Z-1a
(f) SL 327 (1a) in 15% DMSO and 85% neutral water contains 79%
[4] P.A. Groblewski, F.H. Franken, C.L. Cunningham, Inhibition of extracellular signal-
regulated kinase (ERK) activity with SL327 does not prevent acquisition,
expression, and extinction of ethanol-seeking behavior in mice, Behav Brain Res
217 (2) (2011) 399–407.
Z-1a
(g) SL 327 (1a) in 15% DMSO and 85% acidic water contains 81% Z-
1a
[5] M. Rosas, S. Porru, R. Longoni, L. Spina, A.T. Peana, M. Collu, E. Acquas,
Differential effects of the MEK inhibitor SL327 on the acquisition and expression of
ethanol-elicited conditioned place preference and aversion in mice,
J Psychopharmacol 31 (1) (2017) 105–114.
Because the configurational composition of SL 327 in the various
formulations used for administration may play a significant role,
particularly in efficacy, we recommend that solutions of SL 327 be
prepared ahead of time and kept at room temperature for at least three
hours to ensure complete dissolution and configurational equilibrium.
Analysis by HPLC to establish the configurational composition of the
specific formulation is highly desirable.
[6] Hobbs, F. W. Amino-Thio-Acrylonitriles as MEK Inhibitors. U.S. Patent 6,703,420
B1, 2004.
[7] A.L. Spek, Structure validation in chemical crystallography, Acta Crystallogr D Biol
Crystallogr 65 (2) (2009) 148–155.
[8] J.C. Selcher, C.M. Atkins, J.M. Trzaskos, R. Paylor, J.D. Sweatt, A necessity for
MAP kinase activation in mammalian spatial learning, Learn Mem 6 (5) (1999)
478–490.
[9] P.A. Scherle, W.-g. Ma, H. Lim, S.K. Dey, J.M. Trzaskos, Regulation of
cyclooxygenase-2 induction in the mouse uterus during decidualization. An event
of early pregnancy, J Biol Chem 275 (47) (2000) 37086–37092.
[10] N.M. Urs, T.L. Daigle, M.G. Caron, A dopamine D1 receptor-dependent beta-
arrestin signaling complex potentially regulates morphine-induced psychomotor
activation but not reward in mice, Neuropsychopharmacology 36 (3) (2011)
551–558.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
[11] D.J. Knapp, B.A. Whitman, T.A. Wills, R.A. Angel, D.H. Overstreet, H.E. Criswell,
Z. Ming, G.R. Breese, Cytokine involvement in stress may depend on corticotrophin
releasing factor to sensitize ethanol withdrawal anxiety, Brain Behav Immun 25
(Suppl 1) (2011) S146–S154.
Acknowledgements
[12] J.D. Raybuck, T.J. Gould, Extracellular signal-regulated kinase 1/2 involvement in
the enhancement of contextual fear conditioning by nicotine, Behav Neurosci 121
(5) (2007) 1119–1124.
The research was supported by National Institute on Drug Abuse
Contract No. HHSN2712013000007C. The X-ray crystallographic work
was supported by NIDA through Interagency Agreement (IAA)
ADA17001 with the Naval Research Laboratory (NRL).
[13] A.E. Agoglia, A.C. Sharko, K.E. Psilos, S.E. Holstein, G.T. Reid, C.W. Hodge,
Alcohol alters the activation of ERK1/2, a functional regulator of binge alcohol
drinking in adult C57BL/6J mice, Alcohol Clin Exp Res 39 (3) (2015) 463–475.
[14] D. Schoepp, B. Johnson, R. True, J. Monn, Comparison of (1S,3R)-1-
aminocyclopentane-1,3-dicarboxylic acid (1S,3R-ACPD)- and 1R,3S-ACPD-stimu-
lated brain phosphoinositide hydrolysis, Eur J Pharmacol. 207 (4) (1991) 351–353.
[15] Schoepp, D.; Johnson BG; Salhoff CR; McDonald JW; MV., J., In vitro and in vivo
pharmacology of trans- and cis-(+-)-1-amino-1,3-cyclopentanedicarboxylic acid:
dissociation of metabotropic and ionotropic excitatory amino acid receptor effects.
J Neurochem. 1991, 56 (5), 1789-96.
Appendix A. Supplementary material
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.bioorg.2021.104955.
´
´
[16] Jindra Valentova, Katarína Bauerova, Leila Farah, Ferdinand Devínsky, Does
stereochemistry influence transdermal permeation of flurbiprofen through the rat
skin? Arch Dermatol Res 302 (8) (2010) 635–638.
References
´
[17] L. Mouledous, M.F. Díaz, G. Hb, Extracellular signal-regulated kinase (ERK)
inhibition does not prevent the development or expression of tolerance to and
dependence on morphine in the mouse, Pharmacol Biochem Behav. 88(1) (2007)
39-46.
`
[1] E. Valjent, J.-C. Corvol, C. Pages, M.-J. Besson, R. Maldonado, J. Caboche,
Involvement of the extracellular signal-regulated kinase cascade for cocaine-
rewarding properties, J Neurosci 20 (23) (2000) 8701–8709.
7