Keywords

1 Introduction

The concept posed in this paper is that perinatal insult can influence later-life neural survival or later-life susceptibility to toxic challenge. Also, perinatal insult can alone result in lifelong neural and behavioral abnormalities in humans, able to be modeled by appropriate treatments of animals. The series of topics highlighted here provide support for what was once a concept, but which is now recognized fact, supported by experimental and observational data in animal subjects and humans.

There are an encyclopedic number of studies demonstrating that perinatal exposure of noxious or seemingly innocuous agents alter the pattern of neural ontogenetic development and produce permanent neuroanatomical, neurochemical, and/or behavioral abnormalities. Examples of this are provided for what might be termed selective neurotoxins.

One general and surviving notion has been that the neurodegenerative and, for that matter, even the neurodevelopmental psychiatric disorders are induced by specific agents that degenerate one or more clearly defined population(s) of neurons, circuits, and/or regions, e.g., dopaminergic neurons in parkinsonism. However, the much more prevalent issue of senescence is a wider phenomenon affecting cells throughout the body, including spontaneous dying of neurons, as per pars compacta substantia nigra (SNpc) dopaminergic neurons and onset of Parkinson’s disease (PD) (Rodriguez et al. 2015). Neuropsychiatric disorders, such as schizophrenia, attention deficit/hyperactivity disorder (ADHD), autism and depression, and Lesch–Nyhan disease (LND), stem from abnormalities and disruptions, both genetic and environmental, of the normal courses of the developmental cycles (Grados et al. 2014; Groves et al. 2014). For example, exposure to femtomolar concentrations (fM, 10−15 mol/dm3 or 10−12 mol/m3) of fragrances (generally sweet or pleasant smell) results in morphological changes at the light microscopic level in fetal neuroblastoma cell lines pertaining to reduced oxytocin-positive and arginine vasopressin-positive neurons in male but not female neuroblastoma cell lines (Sealey et al. 2015). Stressor exposure during early life has the potential to increase an individual’s susceptibility to a number of neuropsychiatric conditions such as mood and anxiety disorders and schizophrenia in adulthood. Epigenetic processes exert cellular/tissue-specific changes in regulating expression of genes, providing potential biomarkers for examining the developmental trajectory of early stress-induced susceptibility to adult neuropsychiatric and/or neurologic disorders (Ibi and González-Maeso 2015; Marco et al. 2016). With the proliferation of gene-based models and etiology-based models for studying brain disorders (Bezard et al. 2013), the predictability of human and animal in vivo outcomes for neurotoxicity and retardation of developmental trajectories proceeds apace.

Under conditions of chronic inflammation, “mediator” molecules like cytokines may be disadvantageous to organism development over prolonged or exaggerated periods. Neuroprotective or neurotoxic outcomes evolving from interactions between cytokines and/or metabolites of tryptophan catabolism, the neuroactive kynurenines, partly influenced by corticosteroid action, contribute to the fate of several signaling pathways, e.g., serotonergic, dopaminergic, and glutamatergic transmissions, and receptor functions such as N-methyl-D-aspartate receptor (NMDA-R) or α7-nicotinic acetylcholine receptor (Myint 2013). For instance, altered kynurenine metabolism is implicated in the pathogenesis of Alzheimer’s disease (AD), PD, and Huntington’s disease (HD), whereas the metabolites and key enzymes, analogs of the metabolites, and small-molecule enzyme inhibitors, preventing the formation of neurotoxic compounds, confer both neuroprotective and therapeutic properties (Tan et al. 2012). Inflammatory mediators activate the kynurenine metabolic pathway and immobilize the production of neuroactive metabolites, thereby initiating a pathogenic cascade with neuropsychiatric consequences (Allison and Ditor 2015; Brundin et al. 2015; Meier et al. 2015). According to the (genetic) Vulnerability-stress-inflammation developmental notion of schizophrenia, stress generates immune alterations (proinflammatory cytokines) that influence dopaminergic, serotonergic, noradrenergic, and glutamatergic neurotransmission through the activation of the enzyme indoleamine 2,3-dioxygenase (IDO) of tryptophan/kynurenine metabolites, leading to kynurenic acid, with the concomitant activation of microglia, a veritable cascade of neuroinflammatory events (Müller et al. 2015).

1.1 Developmental Inflammatory Processes

Immune activation through prenatal or early postnatal exposure to viruses or bacterial products (e.g., lipopolysaccharide (LPS)) consistently impairs brain development and influences behavioral, emotional, and cognitive functional domains (Kirsten et al. 2013; Xia et al. 2014; Zhu et al. 2014a, b) with consequences for the pathogenesis of neuropsychiatric conditions (Delany et al. 2015; Kelly et al. 2015; Mossakowski et al. 2015; Pariante 2015). Maternal inflammation is similarly reflected in neuroinflammatory events resulting in structural and functional disturbances to the developing offspring brain.

Prenatal LPS-induced reorganization of the dendritic architecture was found in both L2PC-A and L2PC-B types, predominantly in the L2PC-A type in mouse offspring (Gao et al. 2015); there was also a differential alteration of intrinsic electrophysiological properties of the two L2PC types. As the resting membrane potential of L2PC-A neurons became hyperpolarized, these neurons were less excitable, whereas the resting membrane potential of L2PC-B neurons was partially depolarized and more excitable. Thus, morphological and electrophysiological abnormalities were linked to pyramidal neuron dysfunction stemming from inflammatory events during pregnancy. Parental microglia-induced neuroinflammation, triggered by bacterial or viral infections, may induce features of neuropsychiatric/neurologic disorders, such as ADHD, schizophrenia, and autism in offspring (Byrnes et al. 2009). In mice exposed prenatally to LPS at gestational days 15 and 17, there was downregulation of peripheral benzodiazepine receptors (PBRs), mediated by the activation of mGluR5 in astrocytes (Arsenault et al. 2015). In addition, the mGluR5–PBR interaction in a mouse model of schizophrenia (Basta-Kaim et al. 2015; Wischhof et al. 2015) was applicable to brain disorder pathophysiology. Thus, LPS-driven ontogenetic effects at mGluR5 have implications in later-life onset of neuropsychiatric disorders.

Inflammatory cytokines are able to affect neuronal ontogeny indirectly by acting at glia and subverting their imbued neuroprotective action to one that is adverse for neurons. By this means, gestational inflammation can indirectly affect neural function—and thereby pose a risk for later age development of neurological or psychiatric disorders (Fukushima et al. 2015; Jo et al. 2015; Steardo et al. 2015). A number of mechanisms may come into play: (i) stimulation of the phagocyte NADPH oxidase (PHOX) to produce superoxide and derivative oxidants, (ii) expression of inducible nitric oxide synthase (iNOS) that produces NO and derivative oxidants, (iii) release of glutamate and glutaminase, (iv) release of tumor necrosis factor alpha (TNF-α), (v) release of cathepsin B, (vi) phagocytosis of stressed neurons, and finally, (vii) decreased release of nutritive brain-derived neurotrophic factor (BDNF) and insulin-like growth factor-1 (IGF-1) (Brown and Vilalta 2015).

Despite all the evidence that neuroinflammation and reactive gliosis feature prominently in most brain and CNS disorders, the notion of glial cells as passive responders to neuronal damage rather than drivers of synaptic dysfunction is changing. Glia have active signaling activity with neurons and influence synaptic development, transmission, and plasticity by mobilizing a plethora of secreted and contact-dependent signals (Chung et al. 2015). Reactive astrogliosis, a feature of AD, presents a continuum of neuropathological processes with accompanying morphological, functional, genetic, and epigenetic events (Jain et al. 2015; Pekny et al. 2014; Steardo et al. 2015; Verkhratsky et al. 2015). Calcium, proteoglycan, TGF-β, NFκB, and complement mediate the neuron–glia interactions under physiological and neurodegenerative states (Lian and Zheng 2015). Although the influences of astrocytes on the aging process are more suspected than implicated, they appear to adopt different functions dependent on disease progression and the extent of accompanying parenchymal inflammation. Astrocytes enable clearance of Aβ and restrict the spread of inflammation in brain, yet astrocytes promote neurodegeneration in AD by releasing neurotoxins and negating crucial metabolic roles (Birch 2014). Using an experimental model of small subcortical infarcts in mice for studying pathophysiological changes in the corticospinal tract and assessing long-term neurologic outcomes and behavioral performance, Uchida et al. (2015) administered the vasoconstrictor peptide, endothlin-1 (ET-1), and the NOS inhibitor N(G)-nitro-L-arginine methyl ester (L-NAME), into the internal capsule of mice. At two months, they observed a loss of axons and myelin surrounded by reactive gliosis in the region of the injection and severe neurological deficits.

1.2 Perinatal Insult and Neurologic Neurodegenerative Disorders

The Latent Early-life Associated Regulation (LEARn) model poses environmental exposures as “hits,” which, when sufficient in strength and/or number as a fetal insult, leads to altered neural development and later-life disorder or susceptibility to disorder (Lahiri et al. 2009)—giving support to the “developmental origins of health and disease” (DOHaD) hypothesis (Barker 2007). This topic has been recently reviewed, in reference to perinatal insult and the ultimate development of neurodegenerative disorders (Tartaglione et al. 2016).

For example, perinatal exposure of mice or monkeys to lead (Pb) results in later-life cognitive deficits accompanied by the upregulation of amyloid precursor protein (APP), Aβ deposits, and phosphorylated tau in brain—features of Alzheimer’s disease (AD) (Bihaqi and Zawia 2013; Bihaqi et al. 2014). Similarly, perinatal exposure of rats to lead (Pb) leads to a similar pattern of deficits along with an increase in the brain level of 8-hydroxy-2ʹ-deoxyguanosine (oxo8dG), a major DNA oxidation metabolite reflecting oxidative stress (Bolin et al. 2006). Other heavy metals (arsenic, cadmium) and pesticide exposure during perinatal development produce similar dysfunctions in animals (Baldi et al. 2011; Ashok et al. 2015). This series of examples supports the contention that early-life insults can have permanent effects in brain and behavior.

In an analogous manner, perinatal exposure or treatment with iron leads to behavioral indices of PD in mice (Fredriksson et al. 1999, 2000) and rats (Dal-Pizzol et al. 2001), with effects thought to be associated with observed oxidative stress in brain. Manganese (Mn) had a direct effect but also increased the susceptibility of brain to later-life toxic insult (Cordova et al. 2012).

These examples give credence to the likelihood that there are multiple kinds of perinatal insults that produce lifelong neural dysfunctions, some of which lead to a greater incidence of neurological, neurodegenerative, and psychiatric disorders in humans.

1.3 Neurotrophins and Neuronal Development

In a long series of studies beginning in the first half of the twentieth century, R. Levi-Montalcini discovered that there were proteins termed neurotrophins that were essential for the development of the nervous system. One of these neurotrophins, nerve growth factor (NGF), was shown to promote the growth and development of the sympathetic nervous system during ontogeny, now known to act by regulating the expression of genes associated with axonal growth and synaptogenesis (Miller and Kaplan 2001). NGF likewise has a prominent effect on the maintenance and development of cholinergic nerves in basal forebrain (Niewiadomska et al. 2009). Impaired cleavage of proNGF to NGF has been suggested as one of the possible causes of degeneration of basal forebrain cholinergic nuclei in AD (Tuszynski and Blesch 2004). Reduced neuronal responding to NGF is another of many other possibilities related to the loss of cholinergic nerves in AD (Cooper et al. 1994). The multifactorial effect of NGF on the nervous system and on the immune system development has been recently reviewed (Bracci-Laudiero and De Stefano 2016).

Synthesis of NGF in brain cells and in the peripheral nervous system is upregulated by the catecholamines (Barra et al. 2014; Hasan and Smith 2014; Sygnecka et al. 2015), which is in keeping with the physiological relation between the level of NGF mRNA and the density of innervation in the peripheral sympathetic nervous systems (Furukawa 2015). NGF is essential for the survival and functional maintenance of forebrain cholinergic neurons projecting mainly to the cortex and hippocampus (Hohsfield et al. 2014; Iulita and Cuella 2014; Perez et al. 2015), with particular importance for the relative levels of pro-NGF and mature NGF. Thus, for example, diabetic encephalography has been characterized by deteriorations in the maturation of NGF (Soligo et al. 2015). NGF increases low-density lipoprotein receptor levels in PC6.3 cells and in cultured septal neurons from embryonic rat brain (Do et al. 2015), indicating that NGF and simvastatin, which is used to decrease unhealthy lipid levels, stimulates lipoprotein uptake by neurons with a positive effect on neurite outgrowth. Increases in low-density lipoprotein receptors and lipoprotein particles in neurons may exert a functional role during the brain development, as well as in neuroregenerative processes and following traumatic brain injuries. Although aging is a normal physiological process accompanied, more often than not, by deteriorations in certain cognitive domains, alterations in the levels of neurotrophic factors NGF, BDNF, and GDNF (glia-derived neurotrophic factor) are implicated in this decline, which implicates lowered neurotrophic levels in the pathogenesis of AD and other age-related disorders (Budni et al. 2015).

2 Actions and Mechanisms of Selective Neurotoxins

2.1 6-Hydroxydopamine

6-Hydroxydopamine (6-OHDA), the first selective neurotoxin to come into common use, was discovered in the late 1960s by H Thoenen and JP Tranzer during their search for norepinephrine (NE) analogs that might provide dark osmophilic “staining” of noradrenergic nerves during the electron microscopic observation (Thoenen and Tranzer 1968a, b). 5-Hydroxydopamine (5-OHDA) fulfilled that criterion, but 6-OHDA to their surprise produced overt destruction of noradrenergic nerves and its action was selective, leaving surrounding tissues and other nerves intact. 6-OHDA was eventually found to produce its neurotoxicity by generating intraneuronal oxidative stress and by an action on mitochondrial cytochromes, thereby blocking ATP formation and energy depletion of neurons (Cohen and Heikkila 1974). Later, 6-OHDA neurotoxicity was extended to dopaminergic nerves, as well (Ungerstedt 1968, 1971).

6-OHDA has found extensive use in neuroscience research, being cited (as “6-hydroxydopamine OR 6-OHDA”) in ~12,000 papers in PubMed. 6-OHDA is a useful agent for uncovering effects of noradrenergic and dopaminergic nerves and for studying neurotoxic processes and mechanisms and reactive neuroprotective strategic mechanisms of these nerves. As a neurotoxin, 6-OHDA destruction of pars compacta SNpc in adult species (rodents, non-human primates) is of value for producing animal modeling of PD. As a neuroteratogen—6-OHDA administration during ontogeny—6-OHDA has effectively modeled several neural disorders including PD, ADHD, and LND, each of which is described subsequently.

2.2 6-Hydroxydopa

Following the discovery of 6-OHDA as a neurotoxin, 6-hydroxydopa (6-OHDOPA) was developed with the rationale that (1) 6-OHDOPA would be able to cross the blood–brain barrier (6-OHDA does not), (2) to be decarboxylated to 6-OHDA in brain; thus, 6-OHDOPA would actually be a protoxin, and (3) 6-OHDOPA-derived 6-OHDA would then destroy noradrenergic and/or dopaminergic nerves deep in brain, (4) while obviating unintentional damage to other nerves which would otherwise occur during injection of 6-OHDA per se into brain (Ong et al. 1969; Berkowitz et al. 1970). Subsequently, 6-OHDOPA was confirmed as a neurotoxin, able to produce destruction to noradrenergic sympathetic nerves (Kostrzewa and Jacobowitz 1972; Sachs and Jonsson 1972a, b) and noradrenergic nerves in brain (Jacobowitz and Kostrzewa 1971; Kostrzewa and Jacobowitz 1973; Zieher and Jaim-Etcheverry 1973, 1975a, b). 6-OHDOPA also proved to be a unique neuroteratologic agent, able to destroy noradrenergic nerves in brain (Kostrzewa and Harper 1974)—with preference for locus coeruleus nuclei (Kostrzewa and Harper 1974; Tohyama et al. 1974a, b; Clark et al. 1979) and the dorsal bundle providing noradrenergic innervation to dorsal brain (Kostrzewa and Garey 1976, 1977)—while leaving dopaminergic innervation to rodent striatum virtually intact (Kostrzewa et al. 1988). This specificity of 6-OHDOPA for noradrenergic nerves provided a unique advantage in mapping noradrenergic nerves in brain in the 1970s (Jacobowitz and Kostrzewa 1971; Sachs et al. 1973).

6-OHDOPA, however, had specifically low potency and also lethality at high dose (Kostrzewa and Garey 1976). Part of the lethal effect may reside in additional agonist action of 6-OHDOPA at non-NMDA glutamatergic receptors (Rosenberg et al. 1991). At the time of its discovery 35 years ago, 6-OHDOPA was useful as a selective noradrenergic neurotoxin. Important discoveries were made by the use of this neurotoxin on noradrenergic systems in brain, including mapping of the dorsal noradrenergic bundle to forebrain, cerebellum, and spinal cord. 6-OHDOPA likewise was useful in uncovering the labile nature of locus coeruleus neurons. At this time, the inherent limitations of 6-OHDOPA relegate it to secondary status; tagged antibodies for marker enzymes (e.g., immunotoxin for dopamine-β-hydroxylase) also provide a more advantageous means to assess noradrenergic nerves. 6-OHDOPA mechanisms and actions were recently reviewed (Kostrzewa 2016).

2.3 DSP-4

DSP-4 [N-(2-chloroethyl)-N-ethyl-2-bromobenzylamine] is another neurotoxin discovered by S Ross and colleagues in the early 1970s during their search for bretylium-related compounds (Ross et al. 1973; Ross and Renyi 1976). DSP-4 was initially found to cross the blood–brain barrier and cyclize to a reactive aziridinium targeted to the NE transporter (NET) and taken up primarily by locus coeruleus noradrenergic nerves, leading to NE depletion (Jonsson et al. 1981, 1982) and overt destruction (Lyons et al. 1989). DSP-4 was recently reviewed (Bortel 2014; Nowak 2016; Ross and Stenfors 2015).

As a neuroteratogen, DSP-4 has relatively selective action on locus coeruleus projections to neocortex, hippocampus, cerebellum, and spinal cord, while leaving peripheral sympathetic nerves relatively unaffected (Zieher and Jaim-Etcheverry 1975a). Typically, reactive sprouting of noradrenergic innervation to hindbrain and cerebellum occurs consequent to relative inactivation or destruction of locus coeruleus-derived innervation to neocortex, hippocampus and spinal cord (Jonsson et al. 1981, 1982; Dabrowska et al. 2007; Bortel et al. 2008; Sanders et al. 2011). Effects are lifelong. DSP-4 has been used to study the neurotoxic and neuroprotective mechanisms of noradrenergic neurons and to determine the association between early loss of noradrenergic innervation and brain and behavioral outcomes.

2.4 Co-administration of DSP-4 and MPTP

When noradrenergic nerves are lesioned with DSP-4 prior to MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine) treatment of C57/Bl6 mice, dopaminergic lesioning is enhanced (Fornai et al. 1997). Prenatal iron (Fe2+, 7.5 mg/kg, on postnatal days 19–12) further exacerbates the effects on dopaminergic neurons and extent of movement disordering produced by the combination of DSP-4 and MPTP (Archer and Fredriksson 2006). Noradrenergic neuronal dysfunction is considered to add to the motor dysfunction in PD, as indicated by enhanced dysfunction of the subthalamic nucleus following the combination of DSP4 and 6-OHDA (Wang et al. 2014). This topic was recently reviewed (Archer 2016b).

2.5 Methamphetamine

The AMPH analog methamphetamine (METH) replicates many of the effects of AMPH. METH (like AMPH), with high affinity for the NET, DAT (dopamine transporter), and the serotonin (5-HT) transporter SERT, is accumulated by these nerves to evoke non-exocytotic release of NE, DA, and 5-HT (Sitte and Freissmuth 2010). Acute effects reflect sympathomimetic and serotoninergic actions at their respective receptor sites (de la Torre et al. 2000). Chronic METH is associated with neurotoxicity (Seiden et al. 1976), being related to acute METH-induced hyperthermia, promotion of reactive oxygen species (ROS), and excitotoxicity (Krasnova and Cadet 2009). Neuroteratologic effects of METH are expressed in large part by a spectrum of behavioral alterations, as outlined later in this paper.

2.6 Domoic Acid

Domoic acid, an agonist at AMPA/kainate-R, is an excitotoxin in high dose (Verdoorn et al. 1994; Tasker et al. 1996), producing neuronal loss and astrocytosis in hippocampus and amygdala, as well as prefrontal cortex and thalamus (Teitelbaum 1990). In both humans and laboratory rodents, neuropathological disturbances are characterized also by reactive gliosis and loss of neurons detected from 24 h onward and most severely after a week or two (Ananth et al. 2001, 2003). Immunohistochemical and histopathologic evidence of chronic inflammation from rats treated with domoic acid indicated severe neuronal degeneration, astrocytosis, microgliosis, universal NOS expression, and dystrophic calcification from 5 days to 54 days after administration (Vieira et al. 2015).

When administered to perinatal rats, domoic acid produces head tremor with vacuous chewing, “wet-dog shakes,” circling, forelimb tremor, hindlimb hyperextension, and hind-paw biting. At very high dose myoclonic jerks and clonic-tonic convulsions (Xi et al. 1997; Doucette et al. 2000).

When administered in low dose during the second week of postnatal ontogeny, these rats in adulthood displayed prominent cell loss in hippocampal CA1 and CA3 regions (Doucette et al. 2004; Bernard et al. 2007), a reduction in GABA neurons (Gill et al. 2010) with prominent mossy fiber sprouting (Holmes et al. 1999), and behaviorally, these rats, as adults, had stage 2/3 seizure (Racine 1972) induced by a novel/stressful environment (Doucette et al. 2004). Other behavioral deficits have been noted (Pérez-Gómez and Tasker 2014). The neurotoxic and behavioral outcome of perinatal domoic acid was recently reviewed (Pérez-Gómez and Tasker 2014; Doucette and Tasker 2016).

2.7 192 IgG-Saporin

The immunotoxin 192 IgG-saporin consists of the monoclonal antibody 192 IgG conjugated to the ribosome-inactivating protein (RIP) saporin. In perinatal rats, 192 IgG targets the low-affinity rat NGF receptor (p75NGF), which is expressed solely on cholinergic neurons in the nucleus basalis magnocellularis (NBM) and diagonal band of Broca (DBBh) in rat basal forebrain. Saporin, being then internalized by receptor-mediated endocytosis, travels by retrograde axonal transport to the neuronal perikaryon and inactivates ribosomes to inhibit protein synthesis, leading to neuronal cell death (Wenk et al. 1994; Leanza et al. 1995; Pappas et al. 1996; Robertson et al. 1998). Perinatal IgG-saporin selectively destroys 70–75 % of cholinergic in NBM/DBBh (Leanza et al. 1996), leading to ~70 % cholinergic denervation of hippocampus. In contrast to 192 IgG-saporin treatment of adult rats, which also destroys cerebellar Purkinje cells (Leanza et al. 1995; Waite et al. 1995; De Bartolo et al. 2009, 2010), perinatal 192 IgG-saporin spares Purkinje cells which have a lower expression of p75NGF.

192 IgG-saporin reduces ultrasonic vocalization (Kehoe et al. 2001; Ricceri et al. 2007) and impairs passive avoidance learning in rat pups (Ricceri et al. 2002). In adulthood, 192 IgG-saporin-lesioned rats spent less time exploring a novel environment (Ricceri et al. 1997; Scattoni et al. 2003), but otherwise there was limited impairment in learning and memory (Leanza et al. 1996; Pappas et al. 1996) except at 22 months (Pappas et al. 2005). The perinatal effects of IgG-saporin were recently reviewed (Petrosini et al. 2014, 2016).

2.8 Quinpirole

Acute administration of the DA D2-R agonist quinpirole produces several short-lived behavioral effects, the most prominent being yawning with penile erection in male rats (Kostrzewa and Brus 1991). However, when administered to rats once a day for several days during postnatal ontogeny, quinpirole produces DA D2-R supersensitization that is manifested as enhanced yawning, locomotor activity, altered pain threshold, and vertical jumping with paw treading (Kostrzewa et al. 1991, 1993a, b; Kostrzewa 1995; Kostrzewa and Kostrzewa 2012). DA D2-R supersensitivity persists for the duration of the life span (Oswiecimska et al. 2000) and is associated with enhanced AMPH-induced release of DA in rat striatum (Nowak et al. 2001; Cope et al. 2010).

Rats that had been quinpirole-primed during the first week or more of postnatal life have cognitive impairment (Brown et al. 2002) and deficits in prepulse inhibition (Maple et al. 2015). These behavioral effects are accompanied by a reduced brain level of BDNF (Brown et al. 2008; Thacker et al. 2006) and reduced expression level of the regulator of G-protein-signaling (RGS) RGS 9 gene which functions to terminate D2-R agonist action. Because these effects are largely attenuated by olanzapine (Thacker et al. 2006), rats with permanent D2-R supersensitivity have been posited as an animal of schizophrenia (Brown et al. 2012; Brown and Peterson 2016; Kostrzewa et al. 2016c; Maple et al. 2015).

3 Animal Modeling with Neuroteratologic Agents

3.1 Rodent Model of PD Produced by Perinatal 6-OHDA Treatment

Perinatal intracerebral (icv) treatment of rats with 6-OHDA (134 μg, half on each side) produces near-total lesioning of SNpc and lifelong near-total dopaminergic denervation of striatum. Acutely there is no discernible behavioral effect, and rats develop into adulthood with no motor deficit. Permanent serotoninergic hyperinnervation of striatum occurs as rats develop into adulthood (Berger et al. 1985; Snyder et al. 1986). Repeated DA D1 agonist treatments in adulthood prime DA D1-R, which remain supersensitized for the remainder of life, while DA D2-R are less affected (Breese et al. 1984a, 1985a, b, 1987; Criswell et al. 1989; Hamdi and Kostrzewa 1991; Kostrzewa and Gong 1991; Gong et al. 1992, 1993a, 1994; see Kostrzewa 1995; Kostrzewa et al. 1998). The perinatal 6-OHDA-lesioned rats represents a suitable animal model of severe PD (Kostrzewa et al. 2006; Kostrzewa et al. 2016b).

An alternative animal model of PD is produced by administering 6-OHDA unilaterally in adult rats to lesion the SNpc; effectiveness of known and putative anti-parkinsonian agents can be assessed by counting the numbers of rotations produced by those treatments (Ungerstedt 1971). Bilateral 6-OHDA adulthood treatment would produce aphasia, adipsia, lack of grooming, and immobility—with consequent death in a matter of day, except with prolonged special care. Still, these rats remain fragile.

In contrast to adulthood 6-OHDA-lesioned rats, the perinatally 6-OHDA-lesioned rats provide immense advantages for assessing anti-parkinsonian agents. Because perinatally 6-OHDA-lesioned rats are able to eat, drink, groom, and ambulate—as per intact controls, even in the relative absence of SNpc dopaminergic neurons, this neurochemical/neuroanatomical model of PD is behaviorally robust and demonstrates ambulatory enhancement when treated with anti-parkinsonian agents; motor dyskinesia produced by high-dose l-DOPA is able to be discerned (Kostrzewa et al. 2006; Kostrzewa et al. 2016b). These rats have been used to assess the elevation of tissue levels of striatal DA after acute l-DOPA treatment (Kostrzewa et al. 2002, 2005); also the effect of acute AMPH on striatal exocytosis (Nowak et al. 2005); and the effect of acute l-DOPA on striatal levels of ROS (Kostrzewa et al. 2000, 2002; Nowak et al. 2010). The perinatal 6-OHDA-lesioned rat, as a modeling of PD, is described in detail in a recent paper (Kostrzewa et al. 2016b).

3.2 Exercise Effectiveness in Improving Behavioral Deficits in a Rodent Model of PD

Physical exercise has proven to be effective and is a recommended alternative for ameliorating, even reversing motor and behavioral dysfunctions in neurodegenerative disorders (Archer 2011, 2012, 2014; Archer et al. 2014a, b; Archer and Garcia 2015; Archer and Kostrzewa 2012; Archer et al. 2011a, b, 2014a, b). In a rodent model of PD, exercise produced profound ameliorative effects (Archer and Fredriksson 2010, 2012, 2013; Archer et al. 2011a, b, 2014a, b; Fredriksson et al. 2011). Exercise is a particularly useful intervention in PD patients in sedentary occupations. The several links between exercise and quality of life, disorder progression and staging, risk factors, and symptom biomarkers in PD all endow a promise for improved prognosis. Nutrition provides a strong determinant for disorder vulnerability and prognosis, with fish oils and vegetables with a Mediterranean diet offering both protection and resistance, whereas exercise increases synaptic strength and influences neurotransmission. Nevertheless, the heterogeneity of exercise/activity programs, including stretching, muscle strengthening, balance, postural exercises, occupational therapy, cueing, and/or treadmill training, remains an issue and consensus concerning the optimal approach (Abbruzzese et al. 2015; but see also Uhrbrand et al. 2015). Three factors determining the effects of exercise on disorder severity of patients may be presented: (i) exercise effects on motor impairment, gait, posture, and balance; (ii) exercise reduction of oxidative stress, stimulation of mitochondrial biogenesis, and upregulation of autophagy; and (iii) exercise stimulation of dopaminergic neurochemistry and trophic factors.

Running-wheel performance, as measured by distance run by control and parkinsonian-modeled mice from different treatment groups, was related to dopaminergic system integrity, indexed by striatal DA levels (Archer and Kostrzewa 2016). Support for these notions (regarding the almost finite advantages to be gleaned from exercise) continues to emerge. Exercise triggers plasticity-related events in the human PD brain, such as corticomotor excitation, increases in gray matter volume, and an elevation in BDNF levels (Hirsch et al. 2015). Finally, both nutrition and exercise may facilitate positive epigenetic outcomes, such as lowering the dosage of l-DOPA required for a therapeutic effect. Exercise, as a potent epigenetic regulator, implies a potential to counteract pathophysiological processes and alterations, notwithstanding a paucity of understanding in the underlying molecular mechanisms and dose–response relationships (Archer 2015).

3.3 Rodent Model of ADHD Produced by Perinatal 6-OHDA with Adulthood 5,7-DHT Lesions

In the 1970s, B Shaywitz and colleagues produced an animal of “minimal brain dysfunction,” akin to today’s nomenclature for ADHD, by 6-OHDA lesioning of perinatal rats. These rats demonstrate attentional deficits with spontaneous hyperlocomotor activity, each of which is attenuated by acute AMPH treatment (Shaywitz et al. 1976a, b). Over the past 40 years, this has remained the gold standard for rodent modeling of ADHD.

A variation of this model consists of perinatal 6-OHDA lesioning (134 μg, half on each side), followed by adulthood (10 weeks of age) lesioning with 5,7-dihydroxytryptamine (5,7-DHT, 75 μg icv). Treatment of 6-OHDA-lesioned rats with 5,7-DHT had the effect of reducing striatal serotoninergic hyperinnervation by 30 % and suppressing D1-R supersensitivity while enhancing 5-HT2C-R sensitivity. Behaviorally, these rats displayed enhanced hyperlocomotor activity (vs rats lesioned solely with 6-OHDA), and this activity was attenuated by AMPH (Kostrzewa et al. 1994). Moreover, this animal model of ADHD was able to discern the ability of m-chlorophenylpiperazine (mCPP), a 5-HT agonist, to suppress the hyperlocomotor activity and thereby indicate a new approach toward ADHD treatment (Brus et al. 2004). In vivo microdialysis study indicates that the activity-suppressant effects of AMPH and mCPP are unrelated to exocytosis of striatal DA and 5-HT (Nowak et al. 2007). The higher level of hyperlocomotor activity in rats with the dual 6-OHDA + 5,7-DHT lesions represents a more robust model of ADHD in testing agents with the potential for ADHD treatment (Paterak and Stefański 2014; Kostrzewa et al. 2008). This animal model for ADHD was recently reviewed (Kostrzewa et al. 2016a). A non-pharmacological approach toward abating features of ADHD has been demonstrated (Archer and Kostrzewa 2012).

3.4 ADHD and NMDA-R Systems

An imbalance between central inhibitory/excitatory neurotransmitters and relative activity/connectivity between brain regions, with concomitant disturbances of higher cognitive function, is considered to reflect the pathogenesis of ADHD (He et al. 2015; Mohl et al. 2015; Monden et al. 2015; Roman-Urrestarazu et al. 2015).

Dysfunction of the default-mode network in ADHD patients is considered together with some of the animal models used to examine the neurobiological aspects of ADHD. Much evidence indicates that compounds/interventions that antagonize/block glutamate receptors and/or block glutamate signaling during the “brain growth spurt” (or in the adult animal model) may induce functional and biomarker deficits. Mice treated with glutamate receptor antagonist (MK-801, dizocilpine; ketamine) during the “brain growth spurt” fail to display exploratory activity when placed in a novel environment (the test cages) and later fail to adapt to the environment with locomotor suppression, implying a cognitive dysfunction. A disturbance of glutamate signaling during a critical stage of neural ontogeny may contribute to the ADHD pathophysiology. In a functional magnetic resonance imaging (fMRI) study of executive functioning in ADHD adults and matched controls, it was observed that in people with ADHD, there was a failure of deactivation of the medial prefrontal cortex (Salavert et al. 2015). In another study of ADHD adults, using a rest-to-take switching task, there was a disturbed reinitiation of a rest state.

“Hot” and “cool” cognitive functions present a dichotomy within executive function whereby the former refers to affective domains and the latter to cognitive domains (Doebel and Zelazo 2013; Hongwanishkul et al. 2006; Zelazo et al. 2003, 2004). Top-down processes that operate in more affectively neutral contexts have been termed “cool” executive functioning, whereas those operating in motivationally and emotionally significant situations are referred to as “hot” (Zelazo and Carlson 2012). ADHD children exhibited “cool” executive function deficits which appeared to be unrelated to comorbid oppositional defiant disorder (Antonini et al. 2015). Finally, Babenko et al. (2015) have highlighted the intricate interplay between prenatal stress exposure, associated changes in miRNA expression, and DNA methylation in placenta and brain with possible links to greater risks for incidence of ADHD later in life. The association of studies with NMDA-R antagonists and ADHD has been reviewed recently (Archer 2016a; Archer and Garcia 2016).

3.5 Rodent Model of Lesch–Nyhan Disease Produced by Perinatal 6-OHDA Treatment

Lesch–Nyhan disease (LND), a relatively rare neuroteratologic disorder attributable to a mutation in the HPRT 1 gene, is characterized by deficiency in hypoxanthine–guanine phosphoribosyltransferase (HGPRT). Abnormality in purine recycling leads to high serum levels of uric acid, the end product of purine metabolism, and gout—deposition of uric acid crystals in joints and soft tissue. Neurological symptoms represent a range of stages from mild to severe, but often being associated with self-biting and self-mutilation (Abel et al. 2014; Fu et al. 2015; Schroeder et al. 2001). In five different strains of mice with an HPRT gene knockout—characterized by one of two different HPRT gene mutations (Jinnah et al. 1999)—the nigrostriatal dopaminergic tract was found to be incompletely developed and the striatum had both reduced DA content and increased oxidative stress (Visser et al. 2001). While the HPRT-deficient mouse represents a viable model for the enzymatic deficiency in LND, the behavioral counterpart representing self-mutilation, however, is better modeled in rats that were perinatally lesioned with 6-OHDA (Breese et al. 1984b, 1986, 1989, 1990a, b; 1994; 2005). In these rats, DA D1-R are overtly supersensitive (for some behaviors) (Kostrzewa and Gong 1991; Kostrzewa et al. 1992; Gong et al. 1993a, b; 1994) and are further able to be supersensitized by repeated treatments with l-DOPA or a D1-R agonist—a priming process (Breese et al. 1984a, 1985a, b, 1987). When perinatal 6-OHDA-lesioned rats are acutely treated as adults with l-DOPA or with a DA D1-R agonist, there is prominent self-biting and self-mutilation that can be counteracted with a DA D1-R antagonist (see Wong et al. 1996; Papadeas and Breese 2014). Curiously, LND individuals have a DA deficiency in basal ganglia (as per 6-OHDA rats), and this apparently accrues from inadequate development of dopaminergic innervation (Göttle et al. 2014). The perinatal 6-OHDA-lesioned rat as a model of LND has recently been reviewed (Knapp and Breese 2016).

3.6 Permanent Animal Model of Tardive Dyskinesia

Tardive dyskinesia (TD) is a movement disorder produced in primates and other mammalian species by repeated treatments, over a period of months, with a DA D2-R antagonist. In humans, the D2-R antagonist is a common feature of antipsychotic agents used to treat schizophrenia. TD presents as involuntary repetitive purposeless movements, most often of the lower face—resembling someone chewing gum and sometimes also with tongue thrusting (Casey 1987; Jeste and Caligiuri 1993). In rats, TD is most reasonably produced by including haloperidol or other D2-R antagonist in the drinking water (Waddington et al. 1983; Waddington 1990). After a period of ~3 months, these rats, behaviorally, display spontaneous purposeless (vacuous) chewing movements (VCMs) which persist for as long as the D2-R antagonist is present in the drinking water. After withdrawal of the D2-R antagonist from drinking water, VCMs gradually disappear over a period of 4 to 6 weeks. This latter feature in rats—relating to the regression of TD upon D2-R antagonist withdrawal—contrasts with human TD, in which the TD persists and is often permanent even after the D2-R antagonist withdrawal.

In an attempt to produce a permanent model of TD, rats were first lesioned as perinates with 6-OHDA (134 μg, half on each side). When these rats (and controls) reached adulthood, haloperidol was added to the drinking water for a period of nearly one year. While intact control rats developed TD (i.e., increased number of VCMs) after ~3 months, 6-OHDA-lesioned rats developed TD only after 2 months. Moreover, the number of VCMs in haloperidol/6-OHDA rats was 2-fold greater than the number of VCMs in haloperidol/intact control rats. Significantly, after the removal of haloperidol from drinking water (i.e., haloperidol withdrawn stage), VCMs gradually disappeared in haloperidol/intact rat over a period of ~2 months, while VCMs persisted in 6-OHDA-lesioned rats, at the same elevated level and until the experiment ended 8 months later. At that time, it was determined that the D2-R number (i.e., Vmax) had been increased during the haloperidol phase and that D2-R number had reverted to normal by 8 months—signifying that numbers of VCMs were unrelated to numbers of strital D2-R (Huang et al. 1997).

The advantage of persistent VCMs in the withdrawal phase is that it becomes possible to test agents that might have the ability to suppress VCMs. To this end, it was found that agonists and antagonist at both the D2-R and D1-R had no effect, nor did agonists or antagonists at a number of other types of receptors. Only antagonists at 5-HT-R attenuated VCMs in rats in the withdrawal phase, and the common feature of each of these antagonists was that they have affinity for the 5-HT2C-R, a likely site that can be targeted to reduce TD in humans during the antipsychotic withdrawal phase (Kostrzewa et al. 2007). This animal model of TD is described in detail in a recent paper (Kostrzewa and Brus 2016).

3.7 Valproate Modeling of Autism Spectrum Disorder

Prenatal/postnatal/perinatal etiologies, ranging from exposures involving drugs to infections, as well as genetic factors, are complicit in autism spectrum disorder (ASD) that affects roughly 1–2 % of all children, according to the current analyses (Pelly et al. 2015). Several maternal diseases during pregnancy are linked to ASD, pregestationally and/or gestationally, including diabetes mellitus, maternal infections (i.e., rubella, cytomegalovirus), prolonged fever, and maternal inflammation, inducing changes in a variety of inflammatory cytokines (Ornoy et al. 2015); among external agents affecting ASD outcome are drugs such as valproic acid (VPA), the anticonvulsant agent and mood stabilizer, and antiepileptic compounds (Kulaga et al. 2011; Jacobsen et al. 2014). VPA is associated with poorer longer-term child developmental outcomes (Galbally et al. 2010).

Several aspects of animal models, generally and specifically pertaining to ASD, are scrutinized and surveyed, including construct validity, face validity, ASD-like behavioral and neurochemical alterations, histone deacetylase inhibition which elevates ROS, oxidative stress, and the status of experimental models and mitigating factors. These above processes relate to an altered epigenetic landscape in ASDs via altered methylation/hydroxymethylation patterns, local histone modification patterns, and chromatin remodeling (Banerjee et al. 2014; Grayson and Guidotti 2015; Siniscalco 2015).

ASD is characterized by deficits in social interaction and restricted or repetitive behaviors, but often accompanied by other behavioral (e.g., aggression), intellectual (e.g., lower IQ), neurological (e.g., epilepsy), or psychiatric (e.g., anxiety, depression) symptoms (Levy et al. 2009). The antiepileptic drug valproate (VPA), when used clinically to treat epilepsy and bipolar disorder in pregnant women (Lloyd 2013), is associated with a 4 % risk for offspring to develop ASD (Christianson et al. 1994; Christensen et al. 2013), with the incidence being 4–5 times greater in males (Wingate et al. 2014). Several types of animal models of ASD have been produced, but the most common model is produced by VPA treatment of perinatal rats (Rodier et al. 1997; Ranger and Ellenbroek 2016).

When pregnant rats are treated with VPA on gestation day 12, the time of fetal neural tube closure (Kim et al. 2011), the brain of offspring has notable abnormalities, including increased neocortical thickness with a higher number of cortical neurons (Sabers et al. 2015), reduced spine density in the hippocampus (Takuma et al. 2014), hyperserotonemia (Narita et al. 2002), and other defects. Behaviorally in rats and mice, there is hyperactivity, repetitive behaviors, and social deficits (Kim et al. 2014), resembling the behavioral spectrum in humans with ASD.

VPA is thought to act by inhibiting histone deacetylase (Phiel et al. 2001), resulting in hyperacetylated histones and associated increased transcriptional activity of multiple genes (Lloyd 2013), which is thought to account for the neuroteratologic effects. Secondarily, VPA increases the production of ROS in brain (Winn and Wells 1999), which may be detrimental to DNA integrity.

Animal modeling of ASD by VPA has been reviewed recently (Roullet et al. 2013; Ranger and Ellenbroek 2016).

4 Perinatal Insults that Model Psychosis Schizophrenia

There are a plethora of agents that, when administered to animals during ontogenetic development, model features of schizophrenia in the adulthood stage. Some of the more common agents having such an effect include epidermal growth factor (EGF) and its homologue neuregulin (NRG-1), METH, phencyclidine (PCP), and quinpirole. Details regarding these substances and their respective roles in animal modeling of psychosis and schizophrenia are described in the following section.

4.1 Epidermal Growth Factor and Schizophrenia Modeling

When administered to perinatal rats and mice, both EGF and NRG-1 produce adulthood effects that mirror some of the features common in schizophrenia: PPI deficit, altered sensorimotor gating and social interaction, exploratory suppression, cognitive deficit, sensitization to psychostimulants (METH; MK-801, dizocilpine), and other behavioral effects (Sotoyama et al. 2011, 2013; Sakai et al. 2014). Most deficits are reversed by atypical antipsychotics such as clozapine and risperidone but not by typical antipsychotics such as haloperidol (Sotoyama et al. 2013). Yet, in the EGF and NRG-1 models, learning is not compromised, as demonstrated by testing for context fear learning and passive avoidance learning (Futamura et al. 2003; Tohmi et al. 2005).

EGF is thought to exert its major effect on dopaminergic neurons in the SN, increasing dopaminergic activity in the globus pallidum (Sotoyama et al. 2011), while NRG-1 is more selective for dopaminergic neurons in the VTA (Abe et al. 2009; Iwakura et al. 2011a, b), producing enhanced dopaminergic activity in the prefrontal cortex (Kato et al. 2011). ROS formation is considered as a primary process in mediating these effects, as antioxidants suppress some of the adulthood behavioral deficits (Mizuno et al. 2008, 2010). This topic has been reviewed recently (Nagano et al. 2016).

4.2 Phencyclidine and Schizophrenia Modeling

In rodents, prolonged non-competitive NMDA-R antagonism by ketamine or PCP evokes a change in biomarkers in brain accompanied by a spectrum of behavioral activities that model schizophrenia—with non-classical antipsychotics acutely reversing many of the deficits (Barnes et al. 2015; Pyndt Jørgensen et al. 2015). Acute and subchronic treatments with PCP affect differentially the neuronal activity of different brain regions: basal DA, but not serotonin. Output in the medial prefrontal cortex is markedly reduced, and tyrosine hydroxylase expression in the ventral tegmental area is decreased, thereby accounting in part for concomitant behavioral alterations expressed through locomotor sensitization and cognitive deficits (Castañé et al. 2015).

Perinatal administration of the NMDA-R antagonist PCP to rodents produces a spectrum of neuropathological and behavioral effects that model some of the features of schizophrenia. Disruption of glutamate signaling during ontogeny by PCP is thought to impede development of the GABAergic system in brain (Ben-Ari et al. 1997; Le Magueresse and Monyer 2013), resulting in an overall imbalance in neuronal excitation and inhibition in brain in adulthood (Hoftman and Lewis 2011). In perinatal PCP-treated rats and mice, there is an adulthood reduction in fast-spiking GABAergic interneurons in medial prefrontal cortex, nucleus accumbens, and hippocampus (Nakatani-Pawlak et al. 2009; Kaalund et al. 2013; Radonjic et al. 2013; Kjaerby et al. 2014), mimicking reduced GABAergic interneuronal activity in the brain of schizophrenic patients (Reynolds et al. 2004). In the nucleus accumbens, there is also a prominent reduction in dendritic spine density of spiny neurons (Nakatani-Pawlak et al. 2009). Anatomic and neurochemical changes in PCP rodents include a decrease in the number of parvalbumin-positive cells and spine density in the frontal cortex, nucleus accumbens, and hippocampus (Nakatani-Pawlak et al. 2009). Also in brain, glutathione and antioxidant defenses are reduced (Radonjic et al. 2010; Stojkovic et al. 2012).

Behaviorally, there are cognitive deficits in the adulthood rodents that were treated perinatally with PCP, as demonstrated by impaired working memory (Morris water maze testing (Sircar 2003) and rate of learning (delayed spontaneous alternation task) (Wang et al. 2001), sensorimotor dysfunction (deficit in prepulse inhibition) (Anastasio and Johnson 2008; Broberg et al. 2010, 2013; Chen et al. 2011; Kjaerby et al. 2013), social withdrawal (White et al. 2009), reduced attention in a social novelty discrimination paradigm (Terranova et al. 2005), and executive function (attentional set-shifting task for executive function) (Broberg et al. 2008). Many of the behavioral deficits are reversed by atypical antipsychotics. This topic was recently reviewed (Neill et al. 2014; Grayson et al. 2016).

4.3 Methamphetamine and Schizophrenia Modeling

METH, used and abused illicitly as an aphrodisiac and euphoriant, produces elevated mood, increased alertness and concentration, “energy” in fatigued individuals, and reduced appetite and promotes (initial) weight loss at lower doses, whereas at higher doses the drug induces psychosis, affective disorders, and rhabdomyolysis (Ago et al. 2006; De Carolis et al. 2015; Harro 2015; Mouton et al. 2015). METH use by pregnant women is associated with cognitive, attentional, and mood dysfunctions in offspring (Hrebíčková et al. 2014; McDonell-Dowling and Kelly 2015; Smith et al. 2015).

Ontogenetic effects of METH are diverse and heavily reliant on gestational age in terms of long-lived alterations in behavior, epigenetic expression, neuronal organization, and overall neurotransmission and receptor parameters (Roos et al. 2015; Vrajová et al. 2014). The prenatal effects on cognitive and emotional behavior provide evidence of drastic disruptions of normal behavioral patterns (Fialová et al. 2015; Malinová-Ševčíková et al. 2014; Šlamberová et al. 2014, 2015). Long-term behavioral alterations induced by chronic METH use imply alterations in gene and protein expression within specific brain subregions involved in the reward circuitry and accompanied by major epigenetic modifications—histone acetylation and methylation (Desplats et al. 2014; Godino et al. 2015). Although epigenetic changes have not as yet been detected following prenatal METH exposures, these findings are awaited (Cadet 2014; Cadet and Jayanthi 2013).

Perinatal METH treatment has a range of effects on adulthood behaviors in rodents, depending upon whether METH is pre- and/or postnatal (Graham et al. 2013; Jablonski et al. 2016). Postnatal METH treatment in the range of birth through the postweaning period has the most pronounced effects, generally suppressing adulthood spontaneous locomotor activity and increasing acoustic startle reactivity (Vorhees et al. 2009). Given at the critical postnatal period, METH produces learning impairment and spatial memory impairment (Vorhees et al. 1994a, b, 2009).

Perinatal METH produces a persistent reduction in brain levels of DA and 5-HT, inhibiting tyrosine hydroxylase activity (Ricaurte et al. 1982; Bowyer et al. 1998), also 5-HT transporters (Kokoshka et al. 1998), and also other neurotransmitter systems. This topic was recently reviewed (Bisagno and Cadet 2014; Jablonski et al. 2016).

4.4 Quinpirole and Schizophrenia Modeling

Repeated daily postnatal quinpirole treatments of rats produce permanent DA D2-R supersensitivity (Kostrzewa 1995; Kostrzewa et al. 2003, 2004, 2008, 2011, 2016c). In adulthood, these rats display enhanced D2-R agonist-evoked behaviors and a spectrum of behavioral alterations. Rats exhibit improved active avoidance responding (Brus et al. 1998b), learning and memory deficits (Brus et al. 1998a) in the Morris water maze task, on place, and on match-to-place versions of this task (Brown et al. 2002, 2004a, 2005), and a deficit in prepulse inhibition (PPI) to acute startle (Maple et al. 2007). In the hippocampus on these rats, BDNF and NGF were reduced (Thacker et al. 2006; Maple et al. 2007), while in the striatum, nucleus accumbens, and frontal cortex expression of RGS9, a transcript regulating G-protein coupling to the D2-R was reduced (Maple et al. 2007). Long-term olanzapine treatment reversed the cognitive deficits, reversed the PPI deficit, and normalized otherwise reduced BDNF and NGF levels in hippocampus (Thacker et al. 2006; Maple et al. 2007) and RGS9 expression (Maple et al. 2007). Because nicotine likewise reverses D2-R supersensitization, drugs acting on α7 nAChRs (e.g., nicotine) have been suggested for the treatment of schizophrenia (Tizabi et al. 1999; Brown et al. 2004b, 2006; Perna and Brown 2013). Quinpirole modeling of schizophrenia was recently reviewed (Kostrzewa et al. 2016a, b, 2016c; Brown and Peterson 2016).

4.5 Stress and Neuropsychiatric Disorders

Prenatal restraint stress (PRS) during the last week of gestation is associated with postweaned offspring displaying attentional deficits, increased anxiety, impaired spatial learning (Lemaire et al. 2000) and deficit in working memory (Maccari et al. 2003), reduction in social play behavior, increased latency in approaching a novel object (Laviola et al. 2004), and a syndrome complex resembling features of ASD (see Weinstock 2008). Clearly, glucocorticoids are implicated in these outcomes. Disruption in the circadian rhythm also has analogous effects to PRS, as each is posed as a means to model psychiatric disorders (Marco et al. 2016).

4.6 Genetic Model of Alzheimer’s Disease

In the laboratory mouse model for AD, APPswe/PS1dE9, with mutant transgenes of APP and presenilin-1 (PS1), chronic inflammation provokes amyloid plaque formation as early as 4 months of age, with numbers of plaques increasing with aging (Ruan et al. 2009). CD11b-positive microglia clusters appeared in hippocampus and neocortex at the same period of development and these also proliferated with age. Clustered glial fibrillary acidic protein (GFAP)-positive astrocytes were observed in hippocampus and cortex after six months of age and became more numerous with aging. Astrocytes appear to be central to AD pathophysiology since the β-amyloid peptide Aβ suppresses cholinergic innervation and synaptic function, subsequent to astrocytic glutamate gliotransmission. Further, Aβ causes neuronal hyperexcitability (Hertz et al. 2015). Other developmental animal models of AD are expected to be introduced and to become more commonplace.

5 Conclusion

Neurotoxins have become paramount in exploring neuronal function in relation to neuroscience research and, in particular, in animal modeling of neurological, psychiatric, and behavioral dysfunctional states. This concise review highlights the mechanisms and action of the most commonly used neurotoxins and reviews the use of individual neurotoxins in animal modeling of PD, ADHD, LND, autism, TD, and psychotic and schizophrenic states. The influence of neurotrophins, EGF in particular, on ontogenetic is outlined, and the influence of perinatal stress as well as disrupted circadian cycling on neuronal ontogeny is described. Animal modeling of human disorders is likely to be used to an ever greater extent and through use of neurotoxins yet to be discovered.