3.4. Common mechanism of action
Human health risk assessments for pesticides, including OPs, have been routinely performed using the reference dose (RfD) or acceptable daily intake (ADI) approaches, which define an acceptable level of human exposure to compounds, that would be without appreciable risks of deleterious effects during a lifetime of exposure to the general population, including sensitive subgroups. No-observed-adverse-effect-levels (NOAELs) are derived from subchronic or chronic animal studies identifying the most sensitive adverse effect; division of the NOAEL by uncertainty factors (generally based on interspecies extrapolation and intraspecies variability) allow the determination of the RfD (or ADI). Even though an individual may be exposed to more than one OP at any given time (e.g., through consumption of multiple residues on foods), risk characterization has been based on toxicity of, and exposure to, individual chemicals [93]. With the enactment by the US Congress of the Food Quality Protection Act (FQPA) in 1996, the US Environmental Protection Agency (US EPA) is now required to conduct combined risk assessments for pesticides showing a “common mechanism of toxicity” [65]. In essence, if two or more pesticides are considered to act through a common mechanism, then cumulative effects of co-exposure would have to be considered in the evaluation and setting of tolerance levels.
In case of OPs, all indeed act by phosphorylating AChE and eliciting a spectrum of cholinergic effects [94]. The US EPA has thus developed a revised cumulative risk assessment for OPs, based on the potency in inhibiting AChE and on the assumption of dose additivity [95]. A decision to perform such combined risk assessments on all OPs based on the interaction with AChE as the common mechanism of action, and hence, of toxicity, may potentially restrict the use of some OPs, particularly when multiple residues of OPs are found in certain foods and when aggregate exposures from nondietary routes are also considered, as required by FQPA [93]. Determining whether all OPs have the same mechanism of toxicity or if sufficient differences in toxic mechanisms exist to allow subclassification, is thus a current topic of much debate. A few years ago, an ILSI working group formulated six hypotheses for possible subclassifcation of OPs, based on differences in OP metabolism, distribution and molecular targets [94]. However, all hypotheses were rejected, and the working group conducted that all “OPs should be considered to act as a common mechanism of toxicity if they inhibit AChE by phosphorylation and elicit any spectrum of cholinergic effects” [94]. Nevertheless, it is being argued that available information are not sufficient to allow a meaningful cumulative risk assessment for OPs. Some of the arguments are as follows: 1) OP exposures causing similar levels of AChE inhibition may induce a different spectrum of toxic signs; 2) additivity cannot be always assumed, as one OP may modify (increase or decrease) the toxicity of another OP; 3) non-cholinergic targets (i.e., targets other than AChE) exist that may be involved in the differential toxicity of OPs and may modify cholinergic toxicity [57], [93] and [95]. The latter argument is of particular interest because it may hinge on various aspects discussed in this section, such as chronic low toxicity and developmental toxicity of OPs, and will be discussed more in detail below (see Section 3.6). Yet, with regard to a common mechanism of action for OPs, the question appears to be still open.
3.5. OP-induced delayed neurotoxicity
In a previous section, NTE has been described as the serine esterase representing the primary target for OPIDP. NTE is defined as the phenylvalerate hydrolyzing activity that is resistant to paraoxon but is inhibited by mipafox, a neuropathic OP [14]. This is a “working definition” that has proven useful for research on OPIDP over the years [11], but does not shed light on the nature and physiological action of the enzyme, nor on its clear role in OPIDP, aside of that of being the target for initiators. Molecular research in the past few years and the use of transgenic animal models have provided some novel and at times unexpected information.
NTE is a member of a novel protein family represented in organisms from bacteria to man; of its 1327 amino acids, 200 near the C-terminal are highly conserved, and the active site serine (S966) lies within this region [96]. Atkins and Glynn expressed an active site region corresponding to residues 727–1216 in E. coli and referred it as NEST (NTE-esterase domain) [97]. This made it possible to study substrate specificity, which showed preferential hydrolysis of lipids [98]. A more recent study showed that NTE in mammalian cells and its homologue in yeast degrade phosphatidylcholine to glycerophosphocholine, thereby substantiating a primary role in lipid homeostasis [99]. Quistad et al. provided evidence that NTE may be a lysophospholipase (LysoPLA), which hydrolyzes lysolecithin, a major membrane phospholipid which has demyelinating properties [100]; neuropathic compounds were found to inhibit NTE and NTE-LysoPLA to the same extent, both in vitro and in vivo [100] and [101].
The primary sequence of human NTE has a 41% identity with the Swiss Cheese Protein (SWS) in neurons of Drosophila. In the developing nervous system of Drosophila SWS is possibly involved in cell–cell communication between neurons and glia, but in SWS mutants, aberrant cell–cell interactions lead to apoptotic neuronal and glial death, with extensive vacuolation in the brain (hence the name) [102]. Deletion of NTE gene was found to be lethal in mice at an early embryonic age [103] and [104]. Though embryonic lethality of NTE was indicated to be due to placental failure and impaired vasculogenesis, rather than loss of NTE expression by neurons [104], this nevertheless precluded the study of the role of NTE in the adult brain. A conditional mutant NTE strain where NTE was deleted in neuronal tissue did not show embryolethality and provided viable offspring [105]. These animals (Nes-cre: NTEfl/fl) had less than 10% of NTE activity, but no NTE protein could be detected by Western blot. AChE activity was normal. When the animal brains were examined at 3–4 months of age, significant vacuolization and a dramatic redistribution of the rough endoplasmic reticulum was found in the hippocampus and the thalamus. Loss of Purkinje cells in the cerebellum was also present, with parallel motor deficits in a rotarod test [105].
Additional studies in Drosophila added further interesting information. As said, SWS mutants show vacuolization and degeneration of neurites. This could be reversed by murine NTE, indicating that the mouse protein (and most likely the human NTE protein, which is 96% identical to the murine one) is a true functional ortholog of fly SWS [106]. In Drosophila, SWS is expressed in neurons and to a minor extent in glia, and is localized in the endoplasmic reticulum; a similar distribution had been found for NTE in mouse brain, though no NTE had been detected in glia [106] and [107]. Phosphatidylcholine levels were increased in SWS mutants, substantiating the role of SWS/NTE in its hydrolysis [99], and in the regulation of normal lipid composition in the CNS [106].
While loss of both NTE alleles resulted in embryolethality, heterozygous (Nte+/−) mice were viable and fertile [103]. Nte+/− mice had 40% less brain NTE activity than wild-type mice, but equal brain AChE activity. Nte+/− mice also displayed a phenotype of hyperactivity, which is not directly relevant to OPIDP, and were more sensitive to the acute toxicity of ethyl octophosphoro-fluoridate (EOPF), an OPIDP-causing OP [103]. Following administration of a low dose of EOPF, Nte+/− mice showed a further increased locomotor activity, but NTE activity was not quantified in these animals [103].
Do these studies contribute to our understanding of OPIDP? Unfortunately, so far only to a very limited extent. Clearly, NTE seems to play a role in membrane lipid metabolism, and may be involved in intra-neuronal membrane trafficking and lipid homeostasis. Whether its lysoPLA activity may be involved in OPIDP remains to be clarified, as this enzymatic activity is also affected by chemicals that do not cause OPIDP. Accumulated lysolecithin causes demyelination of neuronal sheaths, often accompanied by neuronal lesions [100], while changes in axonal morphology seem to be the early morphological events in OPDIP, with secondary attenuation of the myelin sheath [12]. Studies in genetically modified mice have provided some interesting insights in the physiological role of NTE. NTE appears to be required for normal blood vessel and placental development [103] and [104], and absence of brain NTE results in neuronal degeneration and loss of endoplasmic reticulum in various brain areas [105], two effects not directly related to OPIDP. Furthermore, mice, though an excellent tool for transgenic research, are resistant to OPIDP [18]. Thus, the most crucial issues in the mechanisms of OPIDP development and progression still remain obscure [13].