Binding to voltage gate sodium channels during development leads to cognitive impairment

Iris Mangas, Antonio F. Hernandez, Kevin Crofton, Mary Gilbert, Martin Paparella, Anna Price, Tim Shafer, Laura Martino, Martina Panzarea, Andrea Terron

This AOP describes one adverse outcome that may result from the binding of xenobiotics to Voltage Gate Sodium Channels (VGSC) during mammalian development. Binding to VGSC, the molecular-initiating event (MIE; KE1353), results in disruption of sodium channel gate kinetics (KE1977) and consequently to disruption of action potential generation (KE1983); this leads to a subsequent alteration in neurotransmission at all lifestages, but with additional consequences when it occurs during development. Neurotransmitter release is essential for neural activity and neural activity is critical for normal brain development. Disruption of neural activity during development in many brain regions including the hippocampus can negatively impact both neuroanatomy, neurophysiology, and ultimately neurological function. Therefore, chemicals that bind with VGSCs to thwart or augment neurotransmission have the potential to cause adverse effects on the developing brain.  When this occurs in the developing hippocampus, it can ultimately lead to impairments in cognitive function. Herein, we discuss the implications of developmental VGSC binding, disruption of action potential generation and neurotransmission during brain development, altered hippocampal anatomy, function, and ultimately higher cognitive processing controlled by the hippocampus. The physiology of VGSC and its essentiality for neurotransmitter release is well known across species. The hippocampus is known to be critically involved in cognitive function, including learning and memory. The adverse consequences of a chemical interference at the VCSC will depend both on severity, duration, and developmental timing, indicating that exposure could produce different effects at different developmental windows of exposure. It is important to note that this could also occur in other areas of the brain as VGSC are foundational to the structure and function of all neurons. Here we focus on the hippocampus because of its well-known ties to cognition, and downstream outcome of concern for many chemical exposures, but there is less empirical evidence and biological knowledge on the adverse consequences in other brain areas, The overall weight of evidence for this AOP is strong. Gaps in our understanding include the specific critical developmental windows and the quantitative relationship of binding to VGSC and subsequent disruption and cognitive function. Although quantitative information is limited at all levels of KERs, a number of regulatory applications of this AOP for DNT assessment have been identified.

This AOP was originally started as a case study for an evidence-based AOP informed IATA for a single chemical developmental neurotoxicity hazard characterization. This case study was developed to support human health hazard characterization of the pyrethroid pesticidal active substance, deltamethrin, and as a proof of concept on the applicability of the data provided in the Developmental Neurotoxicity In vitro Battery. The goal of the IATA exercise was to apply mechanistic understanding of toxicity pathways for regulatory decision making (DNT IVB OECD, 2023). A Plant Protection Products Panel EFSA Working Group of experts was established for the development of the IATA and the outcome was published in OECD IATA Case Study Program (OECD IATA Cs n362., 2022). The IATA culminated with strong empirical evidence that deltamethrin interacts with the biological target (MIE) and may subsequently cascade through a series of measurable KEs, ultimately resulting in an adverse health outcome (AO). In this IATA Key Event Relationships (KERs) provided evidence for causality using experimental data only for deltamethrin. Although the MIE and early KEs had strong empirical support, the more downstream KEs and KERs did not.  The empirical support was weak or moderate for most of the downstream KERs and when limited to evidence from only a single chemical, most KERs were not adjacent.

To develop a tool to apply mechanistic understanding for human health protection, a stressor agonistic AOP supported by a strong weight of evidence was developed.  To do so several literature review tools were used in a step wise approach. Artificial Intelligence Tools (i.e., topic modeling) were employed to identify additional   essential Key Events and to increase the scope of empirical evidence. Mapping of the literature landscape analysis culminated in identification of a new downstream KE.  A chemically agnostic AOP was then developed by conducting both systematic broad and focused literature searches (i.e., searching literature using systematic search terms and, most importantly, providing a transparent description of how the literature was searched and selected) in order to collect empirical evidence in support of proposed KERs.  Finally, in line with the AOP development handbook, a search was done in the AOP Wiki to find existing KEs and KERs that may be common to those in the newly developed AOP. Several downstream KEs and KERs originally described and published in AOP 42, were adopted. Finally, expert discussions of the following contributors fostered further development on this new AOP: Iris Mangas, Antonio F. Hernandez, Kevin Crofton, Mary Gilbert, Martin Paparella, Anna Price, Tim Shafer, Laura Martino, Martina Panzarea, Andrea Terron.

Determination of confidence in the overall AOP as a basis to support specific regulatory application relies on the biological plausibility, empirical support, and quantitative understanding of the KERs, as well as the evidence supporting essentiality of the KEs. Table 1 provides an overall summary of the weight of evidence based on the evaluations of the individual linkages from the Key Event Relationship pages. It indicates how biological plausibility and empirical evidence improved with the new work reported here (e.g. from moderate to strong of biological plausibility of KER749.

Please, refer to Appendix B1. Statistical Analysis report for the description of the methodology and individual assessment in the Expert Knowledge Elicitation.

Table 1. Summary table with the assessment of the relative level of confidence in the overall AOP based on rank ordered weight of evidence elements and Expert Knowledge Elicitation.  KER1. Binding to VGSC leads to disruption of sodium channel gate kinetics; KER2. Disruption of sodium channel gate kinetics leads to disruption of action potential; KER3. Disruption of action potential leads to altered neurotransmission during development; KER4. Altered neurotransmission during development leads to hippocampal anatomy altered; KER749. Hippocampal anatomy altered leads to hippocampal physiology altered; KER748. Hippocampal physiology altered leads to cognitive function decrease.

The development of the new KE (KE 2005) referred to us ‘Altered neurotransmission during development’ has been a critical knowledge compilation in this AOP. The regulatory relevance of the in vitro testable KE “Altered neurotransmission during development” is now further supported by the characterization of mechanistic KERs within this augmented AOP. These KERs link the KEs “altered hippocampal anatomy” and “altered hippocampal physiology” to the AO “decreased cognitive function”. The biological plausibility and empirical evidence supporting these KERs have been assessed as moderate to strong.

This KE occurs in all life stages. As the balance between excitatory and inhibitory neurotransmission shapes hippocampal circuitry, any perturbation of this balance can lead to abnormal network activity (Cherubini et al., 2021). The methods and test systems used to measure abnormal network activity aresimilar in developmental or adult life stages. In addition, upstream KEs and pathways occurring in all life stages can lead to alteration in neurotransmission. However, in this AOP, KE 2005 pertains to the developmental period, because it is recognized that the biological and toxicological consequences can be different when disruption of neurotransmission occurs during development, versus adult life stages. It is well recognized that an infant’s brain contains more neurons at birth than that of an adult, and the developing brain undergoes remarkable remodelling to achieve mature neural circuits via processes like apoptosis and synaptic pruning. As brain development matures further activity-dependent remodelling will strengthen circuits that prove more relevant and weaken others that are less frequently used. Such remodelling is more prominent in the first 2 years of life in humans and again during adolescence, with neural activity being a key driver for synaptic pruning. Thus, disruption of the formation of precise neural circuits during critical stages of brain development (i.e., perinatal and adolescence) may underlie neurodevelopmental disorders (Faust et al., 2021; see KER4 description and life stage applicability).

In mammals, it is well known that activity-dependent neuronal remodelling and the timing of this process depends on the brain region and cellular subtypes. This AOP focused on the hippocampal region (as detailed in Appendix C) since hippocampal circuits have been more extensively studied, particularly in relation to regulated chemicals. Furthermore, the hippocampus has been causally linked to measurable AOs (e.g., learning and memory) in rodent models. While several model circuits for studying activity-dependent neuronal remodelling are available for many brain regions, future work is required to develop an AOP and KERs for other brain areas.

There are uncertainties in this new chemically agnostic AOP, including but not limited to knowledge gaps regarding quantitative relationships between KEs and the subsequent adverse impacts on cognitive functions, species extrapolation issues common to all animal based AOPs, and possible lower sensitivity of rodent cognition models commonly used in regulatory studies. This also holds true for developing AOPs for other brain regions, since this AOP is focused on the AO of altered hippocampal-based cognitive function. However, regarding the utility of the MEA-based neural network formation (NNF) assay (OECD, 2023) for use in chemical regulation, it is important to note that it uses cortical cell cultures.  Thus, an effect on the NNF assay may not necessarily correlate with changes in the hippocampal-based spatial cognitive tests commonly used in regulatory in vivo DNT studies. It is biologically highly plausible that disturbed cortical cell based NNF generalizes to an adverse effect within other brain regions. Acute exposure to deltamethrin does alter function of hippocampal networks grown on MEAs (Meyer et al., 2008), supporting the possibility of generalization of a similar effect if hippocampal neurons were to be used in the NNF in place of cortical neurons. For chemical regulation the derivation of relevant PoDs is more important than the prediction of any specific neurodevelopmental disorder at the organism level.

The newly developed KE 2005 can be measured using many methodologies that examine neural connectivity (i.e., neurotransmission), including the in vitro NNF assay. A standardized NNF test system to assess the potential impact of chemical exposure on neural network formation and function has been developed using rodent cortical neurons (Frank et al., 2017). This NNF assay is considered valid, biologically relevant and reliable by OECD (OECD, 2023) and the US EPA (US EPA, 2020a,b). An analysis for the regulatory use of the rodent primary cortical cell-based NNF assay and the additional 16 in vitro DNT assays has also been performed, and this may be contextualized with the uncertainties for in vivo data based uncertainties (Paparella et al., 2020).

The NNF assay represents a developing and relatively complex in vitro multi-cellular test system that includes many key neurodevelopmental processes, and provides a readout of neurophysiological function measured by changes in synaptic activity (i.e, general network activity, network bursting, network connectivity). If such activity is disturbed, it is likely caused by one or more upstream KEs (in this linear AOP or in a potential AOP network) that have previously been disturbed and not compensated at the (multi)cellular level. If these functional in vitro changes are large enough, they will disrupt neurological functions in an organism, ultimately eliciting negative effects. Within experimental systems, a positive response often holds greater regulatory relevance than a negative one, be it in vitro or in vivo. This is because none of these systems fully encompass all aspects of human higher cerebral functions, or the characteristics exhibited by humans in their natural state, including aspects such as metabolism, kinetics, molecular and cellular characteristics. Thus, positive effects observed in in vitro models, or in rodent in vivo studies, should be considered indicators of toxicity. Their impact in real life human conditions depends on additional factors such as (epi)genetic background, socioeconomic status, diet, lifestyle, stress, infections and chemical co-exposures.

The direct regulatory relevance of these disruptions, if integrated with other toxicological information, can be used to derive a PoD, which will be the basis for setting a health-based guidance value. To facilitate regulatory use in decision making, an agreed tiered testing strategy approach for use of in vitro data would be helpful. This approach should include the MEA-based NFF assay as well as the remaining assays in the DNT IVB together with interpretive guidance on the MEA/NFF outcomes for quantitative human health risk assessment.

Bal-Price, A. K., Suñol, C., Weiss, D. G., van Vliet, E., Westerink, R. H. S., & Costa, L. G. (2008). Application of in vitro neurotoxicity testing for regulatory purposes: Symposium III summary and research needs. NeuroToxicology, 29(3), 520-531. https://doi.org/10.1016/j.neuro.2008.02.008

Berret E, Barron T, Xu J, Debner E, Kim EJ and Kim JH, 2017. Oligodendroglial excitability mediated by glutamatergic inputs and Nav1.2 activation. Nature Communications, 8(1), 1–15. https://doi.org/10.1038/s41467–017–00688–0

Catterall, W. A., Cestèle, S., Yarov-Yarovoy, V., Yu, F. H., Konoki, K., & Scheuer, T. (2007). Voltage-gated ion channels and gating modifier toxins. Toxicon : official journal of the International Society on Toxinology, 49(2), 124–141. https://doi.org/10.1016/j.toxicon.2006.09.022

Chahine M (ed.), 2018. Voltage-gated Sodium Channels: Structure, Function and Channelopathies.246. Springer.

Cherubini E, Di Cristo G, Avoli M. Dysregulation of GABAergic Signaling in Neurodevelomental Disorders: Targeting Cation-Chloride Co-transporters to Re-establish a Proper E/I Balance. Front Cell Neurosci. 2022 Jan 5;15:813441. doi: 10.3389/fncel.2021.813441

Faust, T.E., Gunner, G. & Schafer, D.P. Mechanisms governing activity-dependent synaptic pruning in the developing mammalian CNS. Nat Rev Neurosci 22, 657–673 (2021). https://doi.org/10.1038/s41583-021-00507-y

Frank CL, Brown JP, Wallace K, Mundy WR, and Shafer TJ, 2017. From the Cover: Developmental Neurotoxicants Disrupt Activity in Cortical Networks on Microelectrode Arrays: Results of Screening 86 Compounds During Neural Network Formation. Toxicological Sciences, 160(1), 121–135. https://doi.org/10.1093/toxsci/kfx169

Káradóttir R, Hamilton NB, Bakiri Y and Attwell D, 2008. Spiking and nonspiking classes of oligodendrocyte precursor glia in CNS white matter. Nature Neuroscience, 11(4), 450–456. https://doi.org/10.1038/nn2060

Meacham CA, Brodfuehrer PD, Watkins JA and Shafer TJ , 2008. Developmentally-regulated sodium channel subunits are differentially sensitive to alpha-cyano containing pyrethroids. Toxicology and applied pharmacology, 231(3), 273–281. https://doi.org/10.1016/j.taap.2008.04.017

Narahashi T, 1996. Neuronal ion channels as the target sites of insecticides. Pharmacology and Toxicology, 79(1), 1–14

OECD Initial Recommendations on Evaluation of Data from the Developmental Neurotoxicity (DNT) In-Vitro Testing Battery; Series on Testing and Assessment No. 377. 2023. Available at: https://one.oecd.org/document/ENV/CBC/MONO(2023)13/en/pdf

Paparella M, Hougaard Bennekou S, and Bal-Price A, 2020.An analysis of the limitations and uncertainties of in vivo developmental neurotoxicity testing and assessment to identify the potential for alternative approaches. Reproductive Toxicology, 96; 327-336. https://doi.org/10.1016/j.reprotox.2020.08.002

Shafer TJ, Meyer DA and Crofton KM, 2005. Developmental neurotoxicity of pyrethroid insecticides: critical review and future research needs. Environmental Health Perspectives, 113(2), 123–136. https://doi.org/10.1289/ehp.7254

Smith TJ and Soderlund DM, 1998. Action of the pyrethroid insecticide cypermethrin on rat brain IIa sodium channels expressed in xenopus oocytes. Neurotoxicology, 19(6), 823–832.

Smith TJ and Soderlund DM, 2001. Potent actions of the pyrethroid insecticides cismethrin and cypermethrin on rat tetrodotoxin-resistant peripheral nerve (SNS/PN3) sodium channels expressed in Xenopus oocytes. Pesticide Biochemistry and Physiology, 70(1), 52–61. https://doi.org/10.1006/pest.2001.2538

Smith TJ, Lee SH, Ingles PJ, Knipple DC and Soderlund DM, 1997. The L1014F point mutation in the house fly Vssc1 sodium channel confers knockdown resistance to pyrethroids. Insect Biochemistry and Molecular Biology, 27(10), 807–812. https://doi.org/10.1016/S0965–1748(97)00065–9

Smith, E and Jonides, J. (1997). Working Memory: A View from Neuroimaging. Cognitive Psychology, 33:5-42.

Soderlund DM, Clark JM, Sheets LP, Mullin LS, Piccirillo VJ, Sargent D,and Weiner ML, 2002. Mechanisms of pyrethroid neurotoxicity: implications for cumulative risk assessment. Toxicology, 171(1), 3–59.

Tan J and Soderlund DM, 2009. Human and rat Nav1.3 voltage-gated sodium channels differ in inactivation properties and sensitivity to the pyrethroid insecticide tefluthrin. Neurotoxicology,30(1),81-9. doi: 10.1016/j.neuro.2008.10.008.

Trainer VL, McPhee JC, Boutelet-Bochan H, Baker C, Scheuer T, Babin D,and Catterall WA, 1997. High affinity binding of pyrethroids to the α subunit of brain sodium channels. Molecular Pharmacology, 51(4), 651–657

US EPA., 2020a., Agency Issue Paper: Use of New Approach Methodologies to Derive Extrapolation Factors and Evaluate Developmental Neurotoxicity for Human Health Risk Assessment. EPA HQ OPP 2020-0263

US EPA., 2020b., Peer Review of the Use of New Approach Methodologies (NAMs) to Derive Extrapolation Factors and Evaluate Developmental Neurotoxicity for Human Health Risk Assessment. FIFRA Scientific Advisory Panel Meeting Minutes and Final Report No. 2020-02

Vais H, Atkinson S, Eldursi N, Devonshire AL, Williamson MS, Usherwood PN. A single amino acid change makes a rat neuronal sodium channel highly sensitive to pyrethroid insecticides. FEBS Lett. 2000 Mar 24;470(2):135-8.https://doi.org/10.1016/S0014-5793(00)01305-3

Vais H, Williamson MS, Devonshire AL, and Usherwood PN, 2001. The molecular interactions of pyrethroid insecticides with insect and mammalian sodium channels. Pest Management Science ,57(10),877-88. https://doi.org/10.1002/ps.392Wang et al., 2001