Hidrogeoquímica: Salinidad intersticial, pH, Redox, variables de los sedimentos . 21

It has been postulated that regression results from brain damage after birth by environmental toxicants, particularly in light of evolving evidence of gliosis and neuronal loss in autism (Kern and Jones, 2006). The increasing incidence of autism parallels progressive contamination of the environment (Grandjean and Landrigan, 2006; Lathe, 2008), and autism rates correlate with presence of toxic landfi ll sites (Ming et al., 2008). A preliminary epidemiological study linked the incidence of autism most strongly to the estimated environmental concentrations of cadmium and mercury (Windham et al., 2006), and another suggested that proximity to point sources of environmental mercury associates with autism (Palmer et al., 2009). Greater cumulative exposure to mercury in autism is suggested by elevated dental concentrations of mercury (Adams et al., 2007), and urinary porphyrin abnormality, which correlates with autistic symptoms (Geier et al., 2008). A possible etiological role for mercury in autism (Bernard et al., 2001) is unproven (Institute of Medicine, 2004; Nelson and Bauman, 2003).

The very timing of regression in the second year is potentially signifi cant, because the BBB, which shields brain from blood-born toxicants, is mature in humans by about 1 year of age. Select regions of brain known as the circumventricular organs (CVO) fail to develop BBB, so they are potential portals for toxicants otherwise impeded by BBB after 1 year. The CVO are located conspicuously within or around the subcortical brainstem, defi ned here broadly to include medulla, pons, midbrain, and diencephalon (Figure 9.1). An extensive neural network interconnects multiple CVO and receives input from the viscera.

A CVO of particular interest, the area postrema (AP), is the so-called “emesis center” of the medulla. AP is one of the most highly vascularized regions of brain (Wislocki and Putnam, 1920), and blood fl owing through its capillaries has very long residence time in comparison to other regions of brain (Gross et al., 1990). The AP is

a central site of action for circulating angiotensin II (Kim et al., 2008). A rich network of axons emerges from the AP, including serotonergic connections to the pons.

AP-lesioned animals consume remarkable amounts of water (Curtis et al., 1996) or concentrated salt water (Johnson and Edwards, 1991)—an interesting observation in light of published reports of increased water consumption in autistic children (Terai et al., 1999) and an 8% (2,090/25,637) incidence of salt-craving in autistic children surveyed by ARI. Food aversions and cravings for carbohydrates and bland diets result from ablation of AP (Edwards et al., 1997), and although inadequately quantifi ed in autism, are commonly reported by parents. Flavor aversion after cad- mium exposure is reversed by dimercaptosuccinic acid (Peele et al., 1988), a chelating agent which does not cross the BBB, but which reportedly improved behavior in a number of children with autism (Geier and Geier, 2006).

AP is proximate to the dorsal motor nucleus of the vagus (DMV) and the nucleus tractus solitarius (NTS). The three structures comprise the DVC, which mediates autonomic function of the cervical, thoracic, and abdominal viscera (Figure 9.2). Ablation of the DMV blocks viscerosecretory as well as visceromotor function. The NTS receives abundant viscerosensory input via cranial nerves VII, IX, and X, integrates peripheral and central signals, and sends both sympathetic and parasym pathetic efferents to the viscera. Conceivably, reported rapid improvement in children with autism after secretin infusion (Horvath et al., 1998) relates to NTS, where binding of intravenous secretin is most prominent (Yang et al., 2004).

DMV is the consistent site of initial pathology in PD, which ascends in stages to outlying brainstem and distal structures including neocortex (Figure 9.3). Not surprisingly, digestive symptoms are frequent in PD and occasionally dominate the clinical picture (Spellman and Warner, 1977). Disorders of gastrointestinal motility are prominent in PD (Cersosimo and Benarroch, 2008). The gastroe- sophageal sphincter is lax, and 61% of patients with PD complained of esophageal

FIGURE 9.1 The CVO, in proximity to primitive brainstem (shaded), remain unprotected by BBB after its maturation in humans by about 12 months of age.

Pineal Subfornical organ Organum vasculosum Median eminence Posterior pituitary Area postrema

symptoms (Bassotti et al., 1998). Many environmental risk factors have been identifi ed for PD (Onyango, 2008).

Years after autistic regression, reported physical changes in brain are evident in widespread areas not limited to brainstem. Plausible mechanisms exist within our model for ramifying changes after initial brainstem injury. Brainstem injury itself may disturb development of higher structures (Geva and Feldman, 2008; Tanguay and Edwards, 1982). The spread of pathology from brainstem CVO conceivably involves diffusion of toxicants or reactive infl ammatory cytokines. Elevated cytokine levels in cerebrospinal fl uid (CSF) of subjects with autism (Vargas et al., 2005) plausibly result from toxicant-induced production in CVO, two of which—AP and median eminence (ME)—lack tight junctions with CSF (Broadwell et al., 1983).

Experimentally, infl ammatory cytokine diffuses from lateral ventricle along white matter nerve bundles of the corpus callosum, external capsule, and striatum all the way to amygdala (Vitkovic et al., 2000). Cytokine injected in striatum exac- erbates excitoxicity in cortex (Lawrence et al., 1998). The pattern of infl ammatory cytokine diffusion along nerve bundles suggests an anisotropic diffusion pathway in small channels located outside myelinated axons (Agnati et al., 1995).

9.4 TOXICANT ACCUMULATION AND INJURY IN CVO

As will be shown, cadmium, monosodium glutamate (MSG), and paraquat exem- plify a class of neurotoxicants, which do not readily cross the BBB and which preferentially accumulate in areas of brain unprotected by BBB. The data we will discuss shows that mercury in its inorganic form poorly crosses BBB and therefore

FIGURE 9.2 Cross section of luxol-blue stained medulla at level of area postrema (AP), which adjoins the dorsal motor nucleus of the vagus (DMV) and the NTS in the fl oor of the fourth ventricle (IV). AP lacks tight BBB and tight CSF barrier and is envisioned as key por- tal for neurotoxicants in autistic regression.

IV

AP

DMV

concentrates in CVO, and also that administration of mercury in the organic form— which is commonly understood to traverse the BBB—results in accumulation of inorganic mercury in CVO, with long-term residence after remote conversion and redistribution via blood. A review of the literature will make it clear that oxidative mechanisms of cytotoxicity are prominent for each of the aforementioned toxicants.

Cadmium injected intravenously into adult rats accumulated only in regions out- side the BBB, including AP and pineal, but did not appear elsewhere in the brain (Arvidson, 1986; Arvidson and Tjalve, 1986). Cadmium induces oxidative stress,

FIGURE 9.3 (See color insert following page 200.) The ascending pathology of Parkinson’s disease occurs in six recognizable stages, beginning in the DMV of the medulla. Progressive shading in table (A) corresponds to the like-shaded anatomic regions represented in diagram (B). (From Braak, H. et al., Cell. Tissue Res., 31, 121, 2004. With permission.)

(A) (B) Presymptomatic phase Threshold 1 2 3 4 5 6 Symptomatic phase Neocortex, primary, secondary Neocortex, high order association Mesocortex, thalamus Substantia nigra, amygdala Gain setting nuclei Dorsal motor × nucleus Stages of the PD-related path, process

histopathological damage, and increased lipid peroxidation in brain tissue (Mendez- Armenta et al., 2003; Santos et al., 2005). In rats, abnormal auditory brainstem response, cytokine elevation, and apoptosis from cadmium exposure are blocked by administration of antioxidants (Kim et al., 2008). The oxidative effect of cadmium appears to be due to inhibition of complexes II and III of the mitochondrial electron transport chain (Wang et al., 2004).

Paraquat administered subcutaneously to adult rats accumulated only in areas that lie outside the BBB, including AP and pineal (Naylor et al., 1995). Paraquat neurotoxicity, which involves increased tumor necrosis factor (TNF) alpha and increased superoxide production by microglia, is mediated by oxidative stress (Wu et al., 2005). In hepatocytes, herbicides uncouple oxidative phosphorylation, inhibit complexes II and IV, deplete glutathione (GSH), and cause mammalian cell death by inducing lipid peroxidation (Palmeira et al., 1995). Paraquat is suggested as a risk factor for PD (Wu et al., 2005; Peng et al., 2007).

MSG added to food is a common source of concentrated free glutamate (Walker and Lupien, 2000), which in excess is excitotoxic. Glutamic acid decarboxylase levels in brain are noted to be much lower in autism (Fatemi et al., 2002). MSG administered to neonatal rats resulted in widespread neuronal necrosis in the arcuate nucleus (Rascher and Mestres, 1980). Administration of MSG after 1 month of age resulted in injury to the ME—a CVO unprotected by BBB—but arcuate neurons were spared, consistent with closure of the BBB (Peruzzo et al., 2000).

Subcutaneous injections of relatively small amounts of MSG (0.1–0.5 mg/g body weight) in 7 day old mice reduced glutamic acid decarboxylase activity in hippocam- pus and cerebellum at 60 days (Urena-Guerrero et al., 2003). But greatest MSG toxicity was found in areas outside the BBB, particularly the ME, which receives axon terminals from the nearby arcuate nucleus and other hypothalamic secretory neurons (Meister et al., 1989).

A single subcutaneous dose of MSG to adult rats induced severe ultrastructural alterations in acetylcholine-positive AP neurons, which transport acetylcholinest- erase to the NTS (Karcsu et al., 1985). Intraperitoneal administration of MSG to rats resulted in oxidative changes in CVO, including lipid peroxides, which persisted for long periods after exposure (Bawari et al., 1995; Singh et al., 2003).

As reviewed by McGinnis (2001), worrisome levels of inorganic mercury exist in domestic water supplies, carbon-burning emissions, and municipal sludge used as fertilizer—and individual inorganic mercury ingestion can be much greater than expected. Mercury cell chlor-alkali plants are used to produce many food products, including high fructose corn syrup, which may contain mercury (Dufault et al., 2009). Ingested inorganic mercury avidly binds intestine, causing infl ammation at low nanomolar concentrations; fractional systemic absorption of ingested inorganic mercury is evidenced by Pink disease, a profound neurological disease in young children who ingested inorganic mercury found in teething powders (McGinnis, 2001). Autopsy reports for Pink disease include reference to vagal degeneration and “small cell” (microglial?) infi ltration in the brainstem (Wyllie and Stern, 1931).

By intramuscular injection, inorganic mercury accumulated largely in AP and brainstem motor nuclei (Arvidson, 1992), and was prominent in AP 16 days after injection (Nordberg and Sereniu, 1969). Inorganic mercury added to drinking water

of rats concentrated in the motor nuclei of brainstem and deep nuclei of cerebellum (Moller-Madsen and Danscher, 1986).

Organic mercury, as found in fi sh and as preservative in some vaccines (Poling 2008), passes the BBB readily. Most organic mercury is converted to inorganic mer- cury for excretion in feces, but some recirculates to increase inorganic mercury con- centration in blood (Havarinasab et al., 2007). Once in the brain, inorganic mercury persists for years (Vahter et al., 1994). It is the inorganic form which associates with immune stimulation (Havarinasab et al., 2007) and increased microglia in brain (Geier and Geier, 2007). Notably, amalgam removal decreased plasma and red-cell inorganic mercury levels by 73% (Halbach et al., 2008).

The absence of BBB increases accumulation of inorganic mercury during and after chronic methyl mercury (organic) exposure. Methyl mercury was administered orally for 18 months to adult female primates, and mercury levels were measured in six regions of brain, including pituitary, a CVO lacking BBB. Inorganic mercury concentrations increased on average in the six areas 30-fold at 6 months and 60-fold at 18 months, but by far the highest concentrations of mercury—largely inorganic— were achieved in pituitary. In animals in which methyl mercury was discontinued at 6 months, inorganic levels continued to climb in the pituitary—doubling between 6 and 12 months—but not in regions with BBB. Mercury in control animals was undetectable in most regions, but appeared higher in pituitary. One or two primates in each exposure group had distinctly higher or lower fractions of inorganic mercury across brain regions, assumed secondary to individual variations in demethylation (Vahter et al., 1994).

9.5 FINDINGS CONSISTENT WITH BRAINSTEM