6
J.L. Newell et al. / Neurochemistry International xxx (2014) xxx–xxx
was not anticipated, as these compounds emerged during the
course of structure–activity-relationship (SAR) studies aimed at
the optimization of rationally designed competitive inhibitors.
Thus, the isoxazole scaffold was selected for analogue development
because of the previously characterized inhibitory activity of a
number of closely related compounds, including quisqualate, ibot-
enate, and bromo-homoibotenate (Bridges et al., 2012b). Employ-
end’’ complex (i.e., an inactive transporter) and that is does so
in a manner that does not alter the binding of the substrate (e.g.,
L
-Glu). Among the three isoxazoles examined, only 5-Benzyl-
4-bis-TFM-HMICA exhibited a substantial difference between the
two replots methods, yielding a Ki of about 20 M from the slope
replot and 60 M from the 1/V intercept replot. While there are dif-
l
l
ferent types of mixed inhibition involving multiple sites of interac-
tion that could produce such results, the most straightforward
interpretation suggests that 5-Benzyl-4-bis-TFM-HMICA is non-
ing amino-3-carboxy-5-methylisoxazole (ACPA) as
a starting
point, it was found that the inhibitory activity increased as aryl
groups were systematically introduced at either the 5 position on
the isoxazole ring (replacing the methyl group of ACPA) or the 4
competitively inhibiting SxÀc with a Ki of ꢀ20
lM, but that when
it is bound the analogue also reduces the affinity of the transporter
for its substrate (Segel, 1993). Such an interpretation is also sup-
ported by the fact that the other closely related di-substituted isox-
azoles are both more potent and act as ‘‘pure’’ noncompetitive
inhibitors.
position of the isoxazole via a hydrazone linkage (replacing a-ami-
no acid moiety) (Matti et al., 2013; Patel et al., 2010). Of these
derivatives, S-2-naphthyl-ethyl-ACPA emerged as one of the more
potent SxÀc inhibitors. Detailed kinetics analysis similar to those
employed in the present study confirmed that it competitively
The switch from a competitive to a noncompetitive mechanism
observed with the di-substituted isoxazoles raises intriguing ques-
tions as to the molecular relationships between the potential sites
of action on SxÀc . Based upon the competitive action of the mono-
substituted isoxazoles, such as 4-bis-TFM-HMICA, it was hypothe-
inhibited the SxÀc -mediated uptake of 3H-
with a Ki of about 50
L
-Glu into SNB-19 cells
l
M (Patel et al., 2010). Given the structural
similarities between the analogues, it was assumed at that time
that 4-bis-TFM-HMICA, in which the aryl substitution was made
at the 4 position of the isoxazole ring was also acting as a compet-
itive manner. That this was indeed the case is confirmed in the
present report, where Michaelis–Menten and LWB analyses dem-
sized that the isoxazole portion of the molecule was acting as an L-
Glu mimic and interacting with substrate binding domains, while
the trifluoromethyl-substituted benzyl group was interacting with
an adjacent lipophilic (aryl-binding) domain. As a consequence of
occupying some portion of the substrate site, it competitively
onstrated it competitively inhibited the uptake of 3H-
SNB-19 cells with a Ki of about 60
L
-Glu into
l
M. The results also confirm
the utility and potency of 4-substituted aryl isoxazole as inhibitors.
These SAR data were particularly valuable, as the results suggest
that there are lipophilic (or aryl-binding) pockets adjacent to the
substrate binding site on SxcÀ and that the presence of these do-
mains can be exploited, much in the same manner as has been
done with the EAATs (Bridges and Patel, 2009), to develop more
potent and specific inhibitors.
blocks the binding of L-Glu. The 5-monosubstituted isoxazole
would be hypothesized to bind in an analogous manner; only it
would be interacting with a different lipophilic domain, the pres-
ence of which is supported by the inhibitory action of the di-
substituted ‘‘hybrid’’ isoxazoles. If this is occurring, the present
demonstration of the noncompetitive action of the di-substituted
analogues may reflect the optimal binding of the two aryl groups
to their respective lipophilic (aryl-binding) domains in a manner
that still inhibits uptake, but also repositions the isoxazole ring
SAR-based comparisons between the 4- and 5-aryl substituted
isoxazole raised intriguing questions as to whether the aryl groups
were interacting with the same domains on the transporter (neces-
sitating a change in the manner in which the isoxazole ring was
accommodated) or that there are two distinct lipophilic domains
present on SxÀc . This issue was directly addressed through the syn-
thesis and testing of comparable isoxazoles that were modified at
both positions. Although markedly less potent than the mono-
substituted isoxazoles, the ability of these first ‘‘hybrid’’ analogues
such that it is no longer directly precludes the binding of L-Glu.
This, in turn, would lead to the hypothesis that the substrate bind-
ing and lipophilic domains are located in close proximity to one an-
other. The competitive action of the other isoxazoles would also be
consistent with this model. Further, lipophilic (aryl-binding) do-
mains have been identified in a number of other transport systems
that are in close proximity to the substrate binding domains,
including the EAATs and the serotonin transporter (SERT) (Bridges
and Patel, 2009; Leary et al., 2011; Zhou et al., 2009b). However,
the possibility that the aryl-substituted isoxazoles are acting at a
site well removed from the substrate-binding site cannot be ex-
cluded. In such an instance, analogue binding would have to pro-
duce conformational changes that inhibit transport activity in a
to also block the SxÀc -mediated uptake of 3H-
L-Glu into SNB-19
cells supported the conclusion that there were at least two distinct
lipophilic (aryl-binding) domains on the transporter, although the
mechanism of inhibition remained to be elucidated (Patel et al.,
2010). These results prompted the optimization of the di-substi-
tuted isoxazoles and the preparation of the three analogues charac-
terized in the present study. As initial screening assays suggested
that these new isoxazoles were among the best inhibitors yet
developed for SxÀc (Table 1), kinetic studies were carried out to
determine Ki values. Surprisingly, the Michaelis–Menten and
LWB analyses of the concentration dependence with which the
di-substituted isoxazoles inhibited the SxÀc -mediated uptake of
manner that may or may not also block L-Glu binding, reflecting
competitive and noncompetitive mechanisms, respectively.
Whichever mechanisms are ultimately resolved, the marked in-
crease in the potency of the present inhibitors, likely also reflect
the presence of the trifluoromethyl groups on the aryl substituents.
The role of fluorine in enhancing the binding affinity of the ligands
likely arises from either the proper filling of apolar pockets, multi-
3H-
L-Glu into SNB-19 cells revealed a noncompetitive mechanism
rather than a competitive one. Similarly, when the uptake rates
were analyzed using Eadie–Hofstee plots as a second approach,
the resulting pattern of lines was again indicative of a noncompet-
itive inhibitor. The Lineweaver–Burk plots were further analyzed
by repotting both the slope and 1/Vmax,apparent (1/Vintercept) vs. [I].
If a compound is acting as ‘‘pure’’ noncompetitive inhibitor, then
these two replots should both be linear and yield the same value
for Ki (i.e., ÀX intercept). This was the case for 5-4-TFM-Benzyl-
polar C–FÁÁÁH–N, C–FÁÁÁC@O, and C–FÁÁÁH–C
a interactions or polar
interactions with electropositive side chains (Bissantz et al.,
2010; Muller et al., 2007; Zhou et al., 2009a; Zurcher and Diede-
rich, 2008). It is entirely plausible that with three trifluoromethyl
groups present on the most potent inhibitor, 5-4-TFM-Benzyl-4-
bis-TFM-HMICA, that both types of interactions contribute to the
enhanced binding affinity. Future work within this evolving library
of compounds will focus on the continued optimization of these
aryl group interactions. The protein domains with which these li-
gands interact have been postulated to represent either intermedi-
ate binding sites guiding substrate permeation (e.g., ‘‘vestibule
sites’’) or potential allosteric regulatory sites (e.g., ‘‘halogen
4-bis-TFM-HMICA, where a Ki value of about 5 lM places it among
the most potent SxÀc inhibitors yet identified (Bridges et al., 2012b).
Mechanistically, the data suggest this inhibitor can bind to either
the empty or substrate-occupied transporter to produce a ‘‘dead
Please cite this article in press as: Newell, J.L., et al. Novel di-aryl-substituted isoxazoles act as noncompetitive inhibitors of the system xÀc cystine/gluta-
Newell
