Substituted quinolines as inhibitors of l-glutamate transport into synaptic vesicles

Abstract

This study investigated the structure–activity relationships and kinetic properties of a library of kynurenate analogues as inhibitors of ³H-l-glutamate transport into rat forebrain synaptic vesicles. The lack of inhibitory activity observed with the majority of the monocyclic pyridine derivatives suggested that the second aromatic ring of the quinoline-based compounds played a significant role in binding to the transporter. A total of two kynurenate derivatives, xanthurenate and 7-chloro-kynurenate, differing only in the carbocyclic ring substituents, were identified as potent competitive inhibitors, exhibiting K_i values of 0.19 and 0.59 mM, respectively. The K_m value for l-glutamate was found to be 2.46 mM. Parallel experiments demonstrated that while none of the kynurenate analogues tested effectively inhibited the synaptosomal transport of ³H-d-aspartate, some cross-reactivity was observed with the EAA ionotropic receptors. Molecular modeling studies were carried out with the identified inhibitors and glutamate in an attempt to preliminarily define the pharmacophore of the vesicular transporter. It is hypothesized that the ability of the kynurenate analogues to bind to the transporter may be tied to the capacity of the quinoline carbocyclic ring to mimic the negative charge of the γ-carboxylate of glutamate. A total of two low energy solution conformers of glutamate were identified that exhibited marked functional group overlap with the most potent inhibitor, xanthurenate. These results help to further refine the pharmacological specificity of the glutamate binding site on the vesicular transporter and identify a series of inhibitors with which to investigate transporter function.

Introduction

As the primary excitatory neurotransmitter in the mammalian CNS, l-glutamate participates in both fast synaptic communication and the higher order signal processing required in learning, memory, development, and plasticity (for review see Cotman et al., 1995). Consistent with classical mechanisms of neurotransmission, l-glutamate has been shown to be sequestered in synaptic vesicles (Storm-Mathisen et al., 1983) and upon depolarization, to be released from neurons in a calcium-dependent manner (Cotman et al., 1981, Nicholls and Sihra, 1986). l-Glutamate is concentrated in these synaptic vesicles in an ATP-dependent manner by a specific uptake system (Naito and Ueda, 1985). This transporter has also been shown to correlate anatomically and temporally with the presence of glutamatergic synapses (Ueda, 1986, Fischer-Bovenkerk et al., 1988, Kish and Ueda, 1989).

Pharmacological and kinetic studies of the vesicular uptake system demonstrate that it is distinct from the well characterized cellular glutamate transporters present on neurons and glia (e.g. EAAC-1, GLT-1, and GLAST; human counterparts, EAAT 1–3 (Arriza et al., 1994, Gegelashvili and Schousboe, 1997)). In contrast to the sodium-dependency exhibited by the cellular systems, vesicular uptake is sodium-independent, coupled to an electrochemical proton gradient generated by a vacuolar type Mg⁺⁺-ATPase, and requires physiologically relevant concentrations of chloride ion (e.g. 1–4 mM) (Naito and Ueda, 1985, Maycox et al., 1988). The potential regulatory role of a chloride binding site was suggested by studies demonstrating that the anion channel blocker, 4,4′diisothiocyanatostilbene-2,2′-disulfonic acid (DIDS) inhibits vesicular glutamate transport and that this inhibition could be prevented by excess levels of chloride (Hartinger and Jahn, 1993). Kinetic studies with isolated vesicles also indicate significant differences in the K_m values determined for glutamate uptake: 1–2 mM (Naito and Ueda 1985), versus 5–50 μM for the high-affinity cellular systems (Bridges et al., 1994, Garlin et al., 1995). The specificities of the respective systems are also distinct, as a number of well known substrates and inhibitors of cellular transporters exhibit little or no activity at the vesicular system (e.g. d- and l-aspartate, l-cysteate) (Dunlop et al., 1991, Fyske and Fonnum, 1991, Tabb and Ueda, 1991). While large numbers of excitatory amino acid (EAA) analogues have been tested as potential blockers of vesicular glutamate uptake, only a few potent competitive inhibitors have been identified; e.g. trans-1-aminocyclopentane-1,3-dicarboxylic acid (trans-ACPD) (Winter and Ueda, 1993), bromocriptine (Carlson et al., 1989), certain naphthalene disulfonic acids (Roseth et al., 1995), and kynurenate (Fyske et al., 1992). This last compound is of particular interest because it is not only present in the CNS (for review see Stone, 1993) but also because it exhibits activity at EAA ionotropic receptors. Specifically, kynurenate acts as a competitive inhibitor at the glutamate sites on the NMDA, KA and AMPA receptors (Ganong et al., 1983) and at the glycine site on the NMDA receptor (Watson et al., 1988, Kessler et al., 1989).

In the present work we analyze the structure–activity relationships and kinetic properties of a library of kynurenate analogues as inhibitors of the vesicular glutamate transporter and identify two closely related derivatives, 7-chloro-kynurenate and xanthurenate, that are markedly more potent as uptake inhibitors than the parent compound. Parallel studies assessed the activity of these inhibitors at the sodium-dependent synaptosomal transporter and at the EAA ionotropic receptors. Computer-based molecular modeling was used to analyze the structure–activity data in an attempt to explain the inhibitory activities of these compounds and to postulate an active conformation of glutamate that binds to the vesicular transporter. The findings will help to further define the pharmacological specificity of the glutamate binding site on the vesicular uptake system and identify novel pharmacological probes with which to investigate its function.