Abstract The actions were examined of 17 frequently used glycol ether compounds on the glutamate receptor-mediated ion currents. The receptors were expressed in Xenopus oocytes by injection of rat brain mRNA. Most of the 17 glycol ethers exerted no effects on the glutamate subreceptors activated by kainate and N-methyl-D-aspartate (NMDA), whereas 2-phenoxyethanol (ethylene glycol monophenyl ether) caused a considerable reduction of NMDA-induced membrane currents in a reversible and concentration-dependent manner. The threshold concentration of the ethylene glycol monophenyl ether effect was <10 umol/l. The concentration for a 50% inhibition (IC50) was ~360 umol/l. The results indicate a neurotoxic potential for 2-phenoxyethanol.

Key words N-methyl-D-aspartate Glutamate receptor Glycol ether 2-Phenoxyethanol Ethylene glycol monophenyl ether

Introduction

Glycol ethers (GE) are extensively used as solvents, detergents and emulsifiers alone or as components in industrial and consumer applications. For example, the domestic production of ethylene glycol monoethyl ether in 1986 was estimated to be 122 million pounds (NIOSH 1991). As a result, a high potential for occupational, consumer and environmental exposure to GE exists during manufacturing and use. Inhalation and dermal uptake are the most likely routes of human exposure to GE which have been shown to be toxic by either route (Hardin 1983; Hobson et al. 1986; Piacitelli et al. 1990).

Several animal studies indicate that GE may cause embryotoxic, teratogenic and haematotoxic effects as well as atrophy of reproductive and lymphoid organs (Hardin 1983; Grafton and Hansen 1987; Moss et al. 1985; Schuler et al. 1984). Effects on the central nervous system additionally have to be expected, since GE are also capable of rapidly crossing the blood-brain barrier (Ahmed et al. 1983). In fact, disturbances of brain functions, for example headache and persistent cognitive impairment, have been found in humans exposed to GE (Morton 1990; Ohi and Wegman 1978; Parsons and Parsons 1938; Rowe and Wolf 1982).

Little is known about the neuronal mechanisms involved in acute and chronic GE neurotoxicity. In general, it may be possible that voltage-operated ion channels, which underlies the generation of action potentials, or ligand-operated ion channels, which are responsible for the synaptic transmission, or both are involved. The aforementioned clinical data, especially as far as the memory impairment is concerned, can be explained by disturbance of synaptic transmission. In order to gain more insight in this respect, we examined possible effects of GE on the function of the glutamate receptor. Thus, the interaction between 17 frequently used glycol ether compounds and ligand-operated ion channels activated by the glutamate subreceptor agonists kainate (KA) and N-methyl-D-aspartate (NMDA) were investigated. Since studies on the detailed effects of substances on these receptors are often hampered in intact brain tissue by the complexity of the CNS, we chose to use the Xenopus oocyte expression system. Xenopus oocytes are capable of expressing foreign mRNA (Dascal 1987) and the expressed glutamate and GABA receptors exhibit all of the features of the naturally occurring receptors (Sigel 1990). Thus, the oocyte expression system permits a direct assessment of drug effects on these receptors (Binding et al. 1996; Mul3hoff et al. 1995). For these reasons, ligand-operated channels were expressed by injection of rat brain RNA and the currents through the receptor channels were investigated in the presence of the different glycol ether compounds.

Materials and methods

Fig 1 Effects of 17 glycol ethers (A) on kainate (KA)-induced membrane currents (B) and on N-methyl-D-aspartate (NMDA)-induced membrane currents (C). Original recordings from a Xenopus oocyte 3 days after injection of poly(A)+ RNA from rat brain. Holding potential: —70 mV. Administration of 50 umol/1 KA and 100 pmol/l NMDA (white bars) and 100 umol/l of each glycol ether (black bars)

Results

After injection of rat brain mRNA, Xenopus oocytes acquired responsiveness to KA, and NMDA (Figs. 1, 2).

The receptor agonists evoked membrane currents in the nanoampere range, which were completely blocked by co-application with their specific antagonists (KA by 100 umol/l 6-cyano-7-nitroquinoxaline, NMDA by 100 umol/1 aminophosphonovalerate; n = 3; data not shown). The agonist responses remained stable and reproducible in amplitude and time course upon successive or prolonged administrations in a given oocyte. The administration of the various GE derivatives alone in concentrations up to 100 umol/l evoked no detectable membrane currents in the oocytes (n = 5).

In a first series of experiments, the interaction of the 17 GE derivatives (Fig. 1A) with KA- and NMDA-activated glutamate receptors was tested by the simultaneous application of the agonist with GE. At the beginning of the experiment the receptors were activated by their specific agonists until the resulting membrane current remained stable. During the prolonged application of the receptor agonists the different GE compounds (100 umol/l each) were successively applied for 30 s. As shown in Fig. lB. none of the GE tested exerted any significant effect on the KA-induced membrane currents (n = 6). When NMDA was used as a receptor agonist solution number 5 (ethylene glycol monophenyl ether = 2-phenoxyethanol) caused a considerable reduction of the agonist response (Fig. IC; n = 6). None of the other GEs exerted any significant effects.

For further clarification, the effects of 2-phenoxy-ethanol on the glutamate subreceptors were tested with respect to its dependence on the concentration (Fig. 2). The KA-induced membrane currents were not altered by 2-phenoxyethanol applied in concentrations up to 500 umol/1 (Fig. 2A). In contrast, the blocking effect of 2-phenoxyethanol on NMDA responses increased with higher concentrations of the GE derivative (Fig. 2B). The dose-response curve in Fig. 3 shows that the threshold concentration of the 2-phenoxyethanol effect was <10 umol/l. Application of 1 umol/l exerted no effect, whereas 10 umol/l inhibited significantly the NMDA-induced membrane current to 0.92 ± 0.03 (n = 7). With administration of 100 and 250 umol/l of the GE compound, the NMDA response was reduced to 0.75 ± 0.04 and 0.6 ± 0.04, respectively. The concentration for 50% inhibition (IC₅₀), obtained by fitting the mean values to a simplified Langmuir equation, was 362 umol/l. The blockade of the NMDA currents by ethylene glycol monophenyl ether was completely reversible within 2 min after washing.

Discussion

Several glycol ethers were tested in the oocyte expression assay for their ability to interact with glutamate subreceptors. The findings reported in this study demonstrate that ethylene glycol monophenyl ether (2-phenoxyethanol), caused a considerable reduction of membrane currents elicited by the glutamate subreceptor agonist NMDA. None of the other 16 glycol ether compounds tested exert an effect on the function of this glutamate subreceptor. This result is somewhat surprising considering the structural similarity of the chemicals. It shows that obviously small changes in the chemical structure, in the case of 2-phenoxyethanol probably the aromatic phenylester group, are responsible for selective effects on receptor function. The primary target site could be the channel proteins themselves, the factors that regulate channel activities, or the surrounding lipid bilayer. Since

Fig. 3 Inhibition of N-methyl-D-aspartate (NMDA)-induced membrane currents by different concentrations of ethylene glycol monophenyl ether (2-phenoxyethanol, solution No 5). Each point represents mean + SEM of 7 experiments (7 oocytes 3 frogs). Inset: Original recordings from one Xenopus oocyte are superimposed. Concentration of NMDA 100 umol/l; ethylene glycol monophenyl ether was administered in concentrations of 1—500 umol/l. The currents are shown as a fraction of the current obtained in the absence of GE

the effect of 2-phenoxyethanol occurs immediately and is restricted to the NMDA receptor, whereas KA an GABA receptors are not impaired by this substance, it is likely that this compound interacts directly with the NMDA receptor and not via a nonspecific interaction, for example with the surrounding neuronal membrane. This is further supported by the findings that the other glyco ethers also have no effects on the function of the different receptors nor on the membrane potential of the oocytes

The effect of 2-phenoxyethanol on NMDA receptor indicates a strong neurotoxic potential of this compound. Generally, impairment of NMDA receptors may lead to several neurological disorders, including a general decrease in neuronal activity and disturbances in the processes of learning and plasticity in man (Stoltenburg Didinger 1994). This is in line with observations on three patients, showing an immediate and delayed neurotoxic effect of this compound. 2-phenoxyethanol, used as a anaesthetic for handling small fish at a salmon hatchery caused three women to experience headache and symptoms of intoxication during use, followed by diminished sensation and strength of hands and fingers. After 1 to years of exposure, the patients manifested gradual onset of symptoms of impairments of central cognitive functions that persisted (Morton 1990).

There are no data available concerning the concentration of ethylene glycol monophenyl ether or other G compounds in brain tissue after exposure. Therefore, the dose-response data determined in the present investigation can hardly be related to the neurotoxic symptom described by Morton (1990). In general, GE neurotoxicity is dependent on many factors, including species, age, dose, and route of administration. Ahmed and co- workers showed in animal experiments that ethylene glycol monomethyl ether and/or its metabolites could rapidly cross the blood-brain barrier at an early time period after treatment, but a considerable amount was eliminated 24 h after treatment (Ahmed et al. 1983) From the case report above one can conclude that ethylene glycol monophenyl ether also enters the brain tissue and, during longer periods of exposition, may reach concentrations, which are able to impair brain function. This may be to block NMDA-dependent synaptic transmission.

It could not be excluded that, besides 2-phenoxyeth anol, metabolites of this substance are responsible for the neurotoxic effects, since all GE undergo extensive in vivo metabolism. Alkoxyacetic acid metabolites appear to be also responsible for the toxic effects of several GE in vivo (Bartnick et al. 1987; Foster et al. 1984; Ghan ayem et al. 1987; McGregor et al. 1983; Moss et al 1985). It is possible that other metabolites of the different GE compounds have special effects on the metabolism of neurons or glial cells, which are in many aspects different to that of oocytes. However, the data presented in this study indicate that not only GE metabolites but also parent molecules may be responsible for neurotoxic effects. In addition, it has been taken into account that metabolites of other GE parent compounds, found not to be effective in the present study may also exert effects on excitatory or inhibitory receptors. In summary, 2-phenoxyethanol has been show to counteract the effects of NMDA. This could provide basis for explanation of neurotoxic signs described in the literature. Nevertheless, the effects of metabolites of 2-phenoxyethanol, especially phenoxyacetic acid, on glutamate receptors need to be further investigated.

References

Stoltenburg-Didinger G (1994) Neuropathology of the hippocampus and its susceptibility to neurotoxic insult. NeuroToxicology 15: 445—450