Chapter 21: Toxic Responses of the Endocrine System

Higher animals, including humans, have developed the ability to regulate their internal environment, independent of wide fluctuations in external factors, in the form of endocrine systems. An endocrine system consists of (1) an endocrine gland that secretes a hormone, (2) the hormone itself, and (3) a target tissue that responds to the hormone. The classical definition of a hormone is a chemical substance produced by a ductless endocrine gland and secreted into the blood, which carries it to a specific target organ to produce an effect (Porterfield, 2001). In addition to the humoral communication regulated by endocrine systems, the nervous system also regulates overall bodily functions (Hedge et al., 1987). The two systems are intimately interconnected and normally work in close concert. One direct point at which the two systems interface involves “neuroendocrine” cells, which are special types of neurons capable of secreting humoral substances (hormones) in response to synaptic input (neurotransmitters).

The hormone-producing glands of humans include the pituitary (hypophysis), the thyroid and parathyroids, the adrenals, the gonads, and the pancreas. The mechanisms by which endocrine glands synthesize, store, and secrete hormones depend on the chemical properties of the hormone. There are primarily three chemical classes of hormones: amino acid derivatives, peptides and proteins, and steroids. The amino acid derivatives include the catecholamines, epinephrine, and norepinephrine (produced in the adrenal medulla); and the thyroid hormones, triiodothyronine (T₃) and thyroxine (T₄). A large number of hormones are peptides or proteins, such as the neurohormones of the posterior pituitary, tropic hormones of the anterior pituitary, and pancreatic hormones. Steroids are produced by the adrenal cortex, the testes, the ovaries, and in pregnancy, the placenta. These hormones are derivatives of cholesterol. In general, designation of a hormone as hydrophilic (peptides and proteins, catecholamines) or hydrophobic (thyroid hormones, vitamin D, and steroids) provides information useful in understanding the synthesis, secretion, transport, and target cell mechanism of action of that particular hormone (Table 21-1).

Circulating hormone levels are influenced by the rate of secretion from the endocrine gland, rate of metabolism, and method by which hormone is transported in the blood. The rate of secretion is regulated in the endocrine gland by a variety of physiological factors. The rate of hormonal metabolism or degradation in the bloodstream determines its half-life (t_1/2) in blood. The cells of the endocrine gland typically contain only a small store of synthesized hormone, although one noted exception is the thyroid that stores hormone for several months. Therefore, in response to a specific stimulus, the cells synthesize and secrete hormone directly into the bloodstream where they are normally found in very low concentrations (10⁻⁸ to 10⁻¹² M). Although most hormones are carried by the blood, they are not all transported in the same way. Hydrophilic hormones (peptides and proteins) are freely dissolved in the plasma and do not require a specialized transport system. In contrast, given the limited aqueous solubilities of hydrophobic hormones (steroids, thyroid hormones), it is not possible for them to exist in free solution in the aqueous environment of the blood so they circulate bound to specialized serum binding proteins or to albumin. The two major functions of serum binding proteins are to (1) solubilize the hormone for transport in the blood, and (2) extend the life of the hormone in circulation (t_1/2) by protecting the small molecules from enzymatic degradation in the blood stream. The extent to which a hydrophobic hormone is protein bound varies from one hormone to another. However, in all cases, the total hormone pool is predominantly in the bound form. This is important because it is generally accepted that only the pool of free hormone is biologically active. Therefore, binding proteins also provide a large hormone reserve that can be called upon to replenish the free pool. It is the magnitude of the free pool rather than the total pool of hormone that is monitored and adjusted to maintain normal endocrine function.

Hormones elicit a wide variety of biological responses. For virtually all hormones, the initial step in this process is binding of the hormone to a receptor in its target tissue. Those receptors can be located on the plasma membrane (peptide, protein, and catecholamine hormones) or have their action in the cellular nucleus (steroids, thyroid, and vitamin D hormones). The interaction of the hormone with its receptor initiates a chain of intracellular events leading to the physiological response specific to that hormone. For each endocrine system in this chapter, the sections first briefly describe the anatomy and physiology of the normal endocrine gland because such a background is important for understanding the effects of xenobiotics on that gland. This is followed by a description of some effects of environmental chemicals on various aspects of the system’s function. Mechanistic information is included whenever possible to aid in the interpretation of findings and to assess their potential for human risk.

Anatomy and Physiology

The pituitary may be divided into four major subdivisions. The pars distalis (adenohypophysis or anterior pituitary) forms the largest subdivision. The pars intermedia forms the thin cellular zone between the adenohypophysis and neurohypophysis. The pars tuberalis (stalk) is a very small subdivision, highly vascularized, and is not known to secrete hormones. Finally, the pars nervosa (neurohypophysis or posterior pituitary) forms the second largest subdivision of the pituitary. There are two basic types of communication between the hypothalamus and pituitary (Fig. 21-1). (1) The adenohypophysis receives vascular supply from the median eminence. The communication between the two is endocrine in that peptide hormones are released in nerve endings of tuberoinfundibular neurons, which are neuroendocrine cells located in the median eminence. Once secreted, the peptides are transported in the blood via the pars tuberalis to the anterior pituitary. (2) The neurohypophysis receives neuronal communication from the supraoptic nucleus (SON) and paraventricular nucleus (PVN). The type of communication is neuroendocrine, the cell bodies are located in the SON and PVN, and their axons stretch through the median eminence of the hypothalamus, span the infundibular stalk, and terminate in the posterior lobe (Porterfield, 2001). These cells are called magnocellular neurons and are neuroendocrine cells. Therefore, functionally and anatomically, the posterior pituitary is an extension of the hypothalamus. In fact, the hormones produced in the posterior pituitary are actually synthesized in the neuroendocrine cell bodies located in the hypothalamus (SON and PVN).

There are a variety of mechanisms that regulate the rate of release of hormones in the neuroendocrine pathway. Superimposed on endogenous rhythms of hormone release are a number of feedback mechanisms. To maintain appropriate homeostasis, the endocrine organ must constantly monitor systemic hormone concentrations, or some function of it. This awareness is provided in the form of negative feedback loops. By this mechanism, hormones produced in the target organ feedback to reduce hypothalamic peptide and/or anterior pituitary hormone release (Hedge et al., 1987). For example, high circulating levels of cortisol produced in the adrenal cortex will inhibit corticotrophin releasing hormone (CRH) release from the hypothalamus, and adrenocorticotropic hormone (ACTH) release from the pituitary.

Most of the blood supply from the hypothalamus to the pars distalis (adenohypophysis, anterior pituitary) is derived from a portal system (hypothalamo-hypophyseal vessels), which arises from a capillary network in the median eminence, transports down the pituitary stalk, then terminates in the anterior lobe. Blood flowing from the hypothalamus to the pituitary delivers hypothalamic peptides that regulate anterior pituitary hormone secretion. There are six major protein hormones secreted by five distinctive cell types in the anterior pituitary. Table 21-2 provides a summary of some of the characteristics of these five cell types.

Each type of endocrine cell in the adenohypophysis is under the control of a specific releasing hormone from the hypothalamus (Fig. 21-2). These releasing hormones are small peptides synthesized and secreted by tuberoinfundibular neurons of the hypothalamus. They are transported by short axonal processes to the median eminence, where they are released into capillaries and conveyed by the hypophyseal portal system to specific trophic hormone-secreting cells in the adenohypophysis. Each hormone stimulates the rapid release of preformed secretory granules containing a specific trophic hormone. Specific releasing hormones have been identified for thyroid-stimulating hormone (TSH), follicle-stimulating hormone (FSH) and luteinizing hormone (LH), adrenocorticotropic hormone (ACTH), and growth hormone (GH). Prolactin (PRL) secretion is stimulated by a number of factors, the most important of which appears to be thyrotropin-releasing hormone (TRH). Dopamine serves as the major prolactin-inhibitory factor to suppress PRL secretion. Another hypothalamic release-inhibitory factor is somatostatin, which inhibits secretion of growth hormone (Hadley and Levine, 2007). Regulation of the thyroid gland by TSH and the adrenal cortex by ACTH will be discussed in this chapter. Regulation of the gonads by LH and FSH is discussed in Chap. 20.

The pars nervosa (neurohypophysis; posterior pituitary) in the human secretes two important hormones. These are vasopressin, VP (or antidiuretic hormone, ADH) and oxytocin. They are nonapeptides that share a strong structural homology. ADH plays an important regulatory role in water conservation and maintenance of body fluid osmolality, blood volume, and pressure. ADH (VP) has two major effects corresponding to its two names; (1) it enhances reabsorption of water by the distal tubule and collecting ducts of the kidney producing an antidiuretic effect, which decreases osmotic pressure of the blood, and (2) it causes contraction of vascular smooth muscle producing a generalized pressor effect (Porterfield, 2001). Oxytocin stimulates contraction of smooth muscles located in the uterine myometrium to regulate fetal parturition, and in breast alveoli to regulate milk let-down during lactation.

Pituitary Toxicity

Few studies have focused on the effects of xenobiotics on the pituitary gland (Stefaneanu and Kovacs, 1991; Capen, 2001). However, studies consistently show that heavy metals may target pituitary gland structure or function (Caride et al., 2010a; Lafuente and Esquifino, 1999). Cadmium affects many different cell types in the pituitary gland (Caride et al., 2010a; Lafuente and Esquifino, 1999). It affects the lactotrophs by inhibiting prolactin secretion after long-term or acute exposure (Lafuente and Esquifino, 1999). The ability of cadmium to inhibit prolactin secretion may depend on the route of administration and age of the animals. In adult rats, oral and subcutaneous administrations of cadmium inhibit prolactin secretion, whereas only oral administration inhibits prolactin in pubertal animals (Lafuente and Esquifino, 1999). Cadmium also affects the gonadotrophs by inhibiting LH and FSH secretion in male and female rats (Lafuente and Esquifino, 1999). This effect of cadmium on the gonadotrophs may depend on the route of exposure because oral, but not subcutaneous administration of cadmium inhibits gonadotropin levels (Lafuente and Esquifino, 1999). Further, the ability of cadmium to affect gonadotrophs may differ with age. Studies indicate that the degree of cadmium-induced inhibition of LH and FSH is higher in older rats compared to younger rats (Lafuente and Esquifino, 1999). Similarly, cadmium affects the corticotrophs by altering ACTH levels in a manner that differs with age (Lafuente and Esquifino, 1999). Cadmium exposure increases ACTH levels in rodents exposed during puberty and decreases ACTH levels in animals exposed during adulthood. Further, cadmium may also affect the somatotrophs (Lafuente and Esquifino, 1999). Studies indicate that acute exposure to cadmium decreases circulating GH levels, while treatment for a longer period (14 days) increases circulating GH levels (Lafuente and Esquifino, 1999). One study evaluated the effect of chronic exposure (30 days in drinking water) to cadmium in male Sprague–Dawley rats on circadian release of ACTH, GH, and TSH (Caride et al., 2010b). Compared with controls, cadmium delayed the 24-hour peak of ACTH release and eliminated that of GH. Interestingly, mean daily circulating levels of ACTH and TSH were increased by the metal. These findings supported that cadmium has chronotoxic effects at the level of the pituitary.

Cadmium is not the only heavy metal that affects the pituitary. Lead and mercury inhibit secretion of LH and FSH from the gonadotropes (Sikka and Wang, 2008). Further, high lead levels have been associated with increases in ADH secretion in case reports (Suarez et al., 1992).

Environmental contaminants such as polychlorinated biphenyls (PCBs) and polybrominated diphenyl ethers (PBDEs) also affect the pituitary gland (Sikka and Wang, 2008). PCBs inhibit release of LH and FSH from gonadotrophs (Sikka and Wang, 2008). They also affect the thyrotrophs by inhibiting TSH secretion (Patrick, 2009; Zoeller, 2010).

Several pesticides cause adverse effects on the pituitary (Lafuente et al., 2000; Reuber, 1984; Sikka and Wang, 2008; Wozniak et al., 2005). The insecticide dimethoate causes pituitary tumors in male and female rats (Reuber, 1984). The organochlorine pesticide methoxychlor increases prolactin levels as well as LH levels (Lafuente et al., 2000). Both dieldrin and endosulfan stimulate prolactin secretion from the pituitary cell line GH3 via a mechanism that involves increasing calcium, which stimulates the release of prolactin (Wozniak et al., 2005). Measurements were made in pubertal male Sprague–Dawley rats following prenatal and lactational exposure to endosulfan. Expression of mRNA encoding prolactin, LH, GH, and TSH was downregulated, whereas secretion of LH and GH was stimulated and prolactin and TSH were inhibited (Caride et al., 2010c). The herbicide atrazine has been demonstrated in vitro to also have direct pituitary effects by binding to the growth hormone releasing hormone receptor and inhibiting GH expression at the mRNA and protein levels (Fakhouri et al., 2010).

Several phytoestrogens affect pituitary cells (Wozniak et al., 2005). Coumestrol reduces pulsatile LH secretion and suppresses the pituitary response to exogenous gonadotropin releasing hormone (GnRH) in female rats (Dickerson and Gore, 2007; Gore, 2011). Acute genistein exposure suppresses pulsatile LH release in ovariectomized ewes (Dickerson and Gore, 2007). Chronic genistein exposure increases the pituitary size in rodents (Dickerson and Gore, 2007). Bisphenol A (BPA) increases LH secretion in adult male rats and GH content and receptors in juvenile rainbow trout (Aluru et al., 2010; Tohei et al., 2001).

A variety of chemicals used in industrial processes alter pituitary structure or function (Verma and Rana, 2009). Flame retardants (tetrabromobisphenol A and tetrachlorobisphenol A) increase proliferation in GH3 cells, a pituitary cell line (Kitamura et al., 2002). The plasticizer BPA increases GH content in rainbow trout (Aluru et al., 2010). 2-Mercaptobenzothiazole, which is used in rubber products and can come into contact with drinking water, causes pituitary tumors in chronically exposed rats and mice (Whittaker et al., 2004).

Finally, a few pharmaceutical agents have been linked to changes in the pituitary. Cyanamide, a chemical used in the treatment of alcoholics, increases pro-opiomelancortin mRNA in the anterior pituitary when coadministered with ethanol (Kinoshita et al., 2001). A safety evaluation of RO2910, a drug for treatment of HIV infection, has shown that RO2910 causes hypertrophy in the thyrotrophs (Zabka et al., 2011). The neuroleptic agent sulpiride causes the release of prolactin from the anterior pituitary in the rat and stimulates DNA replication (Capen, 2001). The administration of clomiphene prevents the stimulation of DNA synthesis produced by sulpiride, but does not affect prolactin release from the gland (Capen, 2001). These findings suggest that the intracellular prolactin content of the pituitary plays a role in the regulation of DNA synthesis through a mechanism mediated by estrogens (Capen, 2001).

The adrenals are two small glands situated on the superior poles of the kidneys (Fig. 21-3). The major physiological role of the adrenals is management of stress. Each adrenal gland is divided into two morphologically and functionally distinct regions. The adrenal cortex is the more exterior region, and the interior region is the medulla. The cortex synthesizes and secretes steroid hormones that function to regulate salt and fluid balance, glucose homeostasis, and a long-term response to stress. The medulla can be likened to a large postganglionic neuron, and secretes the catecholamines, epinephrine and norepinephrine, upon stimulation by preganglionic cholinergic sympathetic fibers. The medullary catecholamines combine with norepinephrine from sympathetic nerve endings to mediate a very rapid time frame of response to stress (Hadley and Levine, 2007). Because of the anatomic proximity of the two regions of the adrenal glands, blood from the cortex perfuses the medulla. This provides a functional association because cortisol secreted by the cortex stimulates expression of phenylethanolamine-N-methyltransferase (PNMT), the enzyme that converts norepinephrine to epinephrine in the medulla.

The adrenals have not been as widely studied in toxicology as other endocrine glands, even though it has been documented to be the most common toxicological target of all (Ribelin, 1984). In fact, the relative frequency of effects on endocrine systems reported from in vivo toxicology studies has been adrenal > testes > thyroid > ovary > pancreas > pituitary > parathyroid, with the adrenal cortex most frequently targeted (Harvey et al., 2007). This relative lack of attention in toxicology is surprising considering the critical role of the adrenals in mediating endocrine homeostasis throughout the body.

As seen in Fig. 21-3, the adrenal cortex can be divided at the cellular level into three separate zones, the zona glomerulosa (outermost), the zona fasciculata (intermediate), and the zona reticularis (innermost). The zona glomerulosa is responsible for production of the mineralocorticoid hormone, aldosterone. The glucocorticoid hormones, cortisol and corticosterone, are both secreted by the zona fasciculata and, to some extent the zona reticularis. The zona reticularis also secretes the androgens, dehydroepiandrosterone, and androstenedione (Porterfield, 2001). These adrenal androgens are thought to be involved in the onset of puberty and serve a function in postmenopausal women; however, they are of relatively little biological significance under normal circumstances.

The adrenal cortex regulates many physiological functions such as the immune system, inflammation, water and electrolyte balance, carbohydrate and protein metabolism involving such target organs as the liver, kidney, heart, bone, and nervous system (Harvey, 2010). The adrenal cortex is predisposed to the toxic effects of xenobiotic chemicals for two apparent reasons. First, the adrenal cortical cells of most animal species contain large stores of lipids used primarily as substrate for steroidogenesis. Many adrenal cortical toxic compounds are lipophilic and, therefore, can accumulate in these lipid-rich cells. Second, adrenal cortical cells express enzymes involved in steroidogenesis, including those of the cytochrome P450 family, which are capable of metabolizing xenobiotic chemicals. A number of toxic xenobiotic chemicals serve as pseudosubstrates for these enzymes and can be metabolized to reactive toxic species. These reactive compounds can cause direct toxicity by covalent interactions with cellular macromolecules, or through lipid peroxidation or the generation of free radicals (Hinson and Raven, 2006).

Steroidogenesis

Adrenal steroids are synthesized from cholesterol by specific enzyme-catalyzed reactions that involve a complex shuttling of steroid intermediates between the mitochondria and endoplasmic reticulum. Histologically, adrenal cortical cells are characterized by an abundance of lipid droplets, mitochondria, and smooth endoplasmic reticulum. Lipid droplets contain cholesterol, the precursor substrate for steroid production. The specificity of mitochondrial hydroxylation reactions in terms of precursor acted upon and the position of the substrate that is hydroxylated is confined to a specific cytochrome P450. The common biosynthetic pathway from cholesterol is the formation of pregnenolone, the basic precursor for the three major classes of adrenal steroids (Fig. 21-4).

In the zonae fasciculata and reticularis, pregnenolone is first converted to progesterone by two microsomal enzymes. Three subsequent hydroxylation reactions involve carbon atoms at the 17, 21, and 11 positions. The resulting steroid is cortisol, which is the major glucocorticoid in teleosts, hamsters, dogs, and nonhuman and human primates. Corticosterone is the major glucocorticoid produced in amphibians, reptiles, birds, rats, mice, and rabbits. It is produced in a manner similar to the production of cortisol, except that progesterone does not undergo 17α-hydroxylation and proceeds directly to 21-hydroxylation and 11β-hydroxylation.

In the zona glomerulosa, pregnenolone is converted to aldosterone by a series of enzyme-catalyzed reactions similar to those involved in cortisol formation; however, the cells of this zone lack the 17α-hydroxyprogesterone that is required to produce cortisol. Therefore, the initial hydroxylation product is corticosterone. Some of the corticosterone is acted on by 18-hydroxylase to form 18-hydroxycorticosterone, which in turn interacts with 18-hydroxysteroid dehydrogenase to form aldosterone. Since 18-hydroxysteroid dehydrogenase is expressed only in the zona glomerulosa, it is not surprising that only this zone has the capacity to produce aldosterone.

Glucocorticoids

To better appreciate the wide-ranging effects of glucocorticoids, the effects of the hormone can be divided into normal physiological and pharmacological effects, since cortisol is therapeutically quite effective and widely prescribed. The physiological effects include (a) hepatic glucose production, gluconeogenesis (liver), (b) protein catabolism (skeletal muscle), (c) fat catabolism (adipose tissue), (d) increased bone resorption, (e) altered mood (CNS), (f) increased gastric acidity (g.i. tract), and (g) PNMT synthesis (adrenal medulla). Therapeutically, at pharmacological levels the effects of cortisol include (a) preventing vascular collapse during overwhelming stress, (b) providing an anti-inflammatory effect, and (c) invoking immunosuppression.

Adrenocortical Toxicity

The zonae fasciculata and reticularis appear to be the principal targets of xenobiotic chemicals in the adrenal cortex. Classes of chemicals known to be toxic for the adrenal cortex include short chain (three or four carbons) aliphatic compounds, lipidosis-inducers, and amphiphilic compounds (Yarrington et al., 1981, 1985). In general, compounds seen to frequently produce necrosis, particularly in the zonae fasciculate and reticularis, include 7,12-dimethyl benz[a]anthracene (DMBA), acrylonitrile, hexadimethrine bromide, polyanethosulfonate (along with amino-caprionic acid), thioacetamide, and basic polyglutamic acid (Szabo and Lippe, 1989; Colby et al., 1994). By comparison, lipidosis inducers can cause accumulations, often coalescing, of neutral fats, which may be of sufficient quantity to cause a reduction or loss of organellar function and eventual cell destruction. Examples of chemicals that directly target glucocorticoid secretion in the adrenal cortex include dimethoate, ketoconazole, spironolactone, efonidipine, mibefradil, 1-aminobenzotriazole, and various PCBs (Colby et al., 1995; Harvey, 2010). Lesions that are caused may be classified as follows: endothelial damage (eg, acrylonitrile), mitochondrial damage (eg, DMNM, o,p′-DDD, amphenone), endoplasmic reticulum disruption (eg, triparanol), lipid aggregation (eg, aniline), lysosomal phospholipid aggregation (eg, chlorophentermine), and secondary effects due to embolization by medullary cells (eg, acrylonitrile). Tricresyl phosphate (TCP) and other triaryl phosphates cause a defect in cholesterol metabolism by blocking both uptake from serum and storage pathways. An inhibition of cytosolic neutral cholesteryl ester hydroxylase (nCEH) by triaryl phosphate results in the progressive accumulation of cholesteryl ester in the form of lipid droplets in the cytoplasm of adrenal cortical cells.

Biologically active cationic amphiphilic compounds produce a generalized phospholipidosis that involves primarily the zonae fasciculate and reticularis and produce microscopic phospholipid-rich inclusions. These compounds affect the functional integrity of lysosomes, which appear ultrastructurally to be enlarged and filled with membranous lamellae of myelin figures. Examples of compounds known to induce phospholipidosis include chloroquine, triparanol, and chlorophentermine.

In addition, there is a miscellaneous group of chemicals that affects hydroxylation and other functions of mitochondrial and microsomal fractions (eg, smooth endoplasmic reticulum) in the adrenal cortex. Examples of these compounds include o,p′-DDD and α-(1,4-dioxido-3-methylquinoxalin-2-yl)-N-methylnitrone (DMNM). Other compounds in this miscellaneous category cause their effects by means of cytochrome P450 metabolism and the production of toxic metabolites. A classic example is the activation of carbon tetrachloride, resulting in lipid peroxidation and covalent binding to cellular macromolecules of the adrenal cortex (Colby, 1988).

Adrenocortical toxicity can also involve increased secretion of endogenous glucocorticoids. Compounds that have been shown to cause this include ethanol, chlordecone, carbon disulfide, cannabinoids, cocaine, amitriptyline, and cytotoxic anticancer drugs (Harvey, 2010). Furthermore, pharmacological treatment with glucocorticoid agonists that have been widely used as anti-inflammatory agents can produce symptoms that resemble Cushing’s syndrome (Harvey, 2010).

There have been documented cases of iatrogenically (physician-caused) induced adrenocortical toxicity. For example, the anesthetic etomidate and the anticonvulsant aminoglutethimide have been demonstrated to produce significant morbidity and mortality because of their potency as steroidogenic enzyme inhibitors. Aminoglutethimide inhibits CYP11A (side chain cleavage enzyme; Johansson et al., 2002), whereas etomidate blocks CYP11B1 (CYPβ/18; Hinson and Raven, 2006). This iatrogenic suppression of steroidogenesis in the adrenal cortex can result in Addisonian crisis (which can be associated with fatigue, cardiovascular collapse, and death).

In Vitro Toxicity

In addition to in vivo testing for effects of xenobiotics on the adrenal cortex, much recent work has used an in vitro approach to identify the molecular targets of adrenocortical toxicity. Of particular usefulness has been the human adrenocortical carcinoma-derived NCI-H295R cell line. This cell line expresses all key enzymes necessary for steroidogenesis, and it produces all of the major steroids including progesterone, androgens, estrogens, glucocorticoids, and aldosterone (Harvey et al., 2007). These cells express functional ACTH receptors as well as functional receptors for CRH, angiotensin II, vasoactive intestinal peptide, atrial natriuretic peptide, LH, human chorionic gonadotropin (hCG), tumor necrosis factor, and activin A. They respond to forskolin and isobutylmethylxanthine (cAMP induction) as well as dibutyryl cyclic AMP with stimulated corticosterone synthesis. This cell line has proven useful for identification of specific steroidogenic enzymes that are targeted by xenobiotics. A recent review article provides a useful table listing over 60 compounds that have been shown to induce functional adrenocortical and steroidogenic toxicity along with the specific enzymes targeted (Harvey et al., 2007). Therefore, the H295R cell line is a versatile tool for assessment of adrenocortical function and steroidogenesis in general. Because the cell line is derived from a human source, it is also very worthwhile for hazard risk assessment.

Serum Binding Proteins

Cortisol and corticosterone are transported in blood by corticosteroid binding globulin (CBG), also called transcortin (Hedge et al., 1987). When bound to CBG, the steroid is biologically inactive and cannot have its effects on target tissues, or provide negative feedback to the hypothalamic–pituitary–adrenal (HPA) axis. Thus, a chemical affecting CBG could alter the balance between free and bound hormone, and impact its availability in target tissues. Nonsteroidal anti-inflammatory drugs (NSAIDS) have been reported to decrease the binding capacity of CBG by a mechanism other than simple displacement of bound glucocorticoid (Harvey, 2010).

Target Tissue Receptors

In addition to direct effects on steroidogenesis, chemically induced changes in adrenal function can result from blockage of the action of adrenocorticoids at peripheral sites. Adrenocortical steroids and HPA axis hormones exert their effects through receptors located in target tissues throughout the body. These receptors can be upregulated or downregulated by the action of xenobiotic compounds. For example, hexachlorobenzene reduces hepatic glucocorticoid receptors in rats (Lelli et al., 2007), and aminoglutethimide downregulates adrenocortical cell ACTH receptors (Fassnacht et al., 1998). Other chemicals can increase adrenocortical cell ACTH receptors (Li and Wang, 2005). In the first mechanism, many antisteroidal compounds (antagonists) act by competing with or binding to steroid hormone receptor sites; thereby, either reducing the number of available receptor sites or by altering their binding affinity. Spironolactone, an antimineralocorticoid, is an example of a peripherally acting adrenal cortical hormone antagonist (Los and Colby, 1994).

Neuroendocrine Regulation

The zonae fasciculata and reticularis of the adrenal cortex are under tropic control by the pituitary hormone, ACTH, which stimulates them to produce cortisol. The level of output of cortisol by the adrenal is almost entirely determined by the intensity of the ACTH stimulus. Increased cortisol produced then provides long-loop negative feedback on the hypothalamus and anterior pituitary and decreases CRH and ACTH secretion, respectively (Fig. 21-5). Stress is a major factor that can override the negative feedback control system and stimulate cortisol secretion (Hedge et al., 1987). Persistent exposure of the adrenal cortex to high levels of ACTH during chronic stress can result in adrenocortical hypertrophy. Adrenal androgens are also produced in ACTH-sensitive cells; therefore, abnormal conditions of stimulation of these cells by ACTH, such as Cushings’ disease, can also raise androgen secretion by the cortex. Conversely, conditions of reduced exposure of the cortex to its tropic hormone, ACTH, can result in adrenal atrophy.

Drugs that affect neuroendocrine function can indirectly affect adrenal function. For example, synthetic glucocorticoids, valproic acid, bromocriptine, cyproheptadine, ketanserin, ritanserin, somatostatin analogs, 4′-thio-beta-d-arabinofuranosylcytosine, hexachlorobenzene, alcohol, and caffeine have been shown to suppress ACTH or CRH secretion (Harvey, 2010). ACTH secretion can also be pharmacologically stimulated by, for example, caffeine, methylxanthines, adenosine analogs, 3,4-methylenedioxymethamphetamine, and di-2-ethylhexyl phthalate (Harvey, 2010). Exogenous steroids can disrupt normal function and structure of the adrenal cortex. Agonists will cause negative feedback inhibition of ACTH secretion in the pituitary and will result in atrophy of the zonae fasciculata and reticularis. On the other hand, antagonists of steroids will block hormone action in target tissues and on negative feedback, which will lead to increased ACTH secretion, and differential hyperplasia of the cortex. Generally, testing of compounds that have a neuroendocrine site of action is only likely to be detected by in vivo studies. Because of the tropic effects of ACTH, adrenal enlargement (hypertrophy) usually results from stimulation of the HPA axis (stress or direct ACTH stimulation), or loss of feedback inhibition. Thus, adrenal hypertrophy can distinguish a neuroendocrine site of action from direct adrenocortical inhibition/suppression. Other physiological endpoints can be evaluated during testing for HPA axis stimulation. A one-month toxicology study of corticosterone administration in rats at high but physiologically relevant levels (approximating stress) observed reduced body weight gain, and lower thymus, adrenal, prostate, and seminal vesicle weights (Harvey et al., 1992). The body and thymus weight effects were attributed directly to high corticosterone, and reduced prostate and seminal vesicle weights to the inhibition of LH and testosterone by corticosterone. These effects in combination with enlarged adrenals (resulting from ACTH stimulation) could be useful as indicators to rule out primary adrenocortical insufficiency (Harvey et al., 2007).

Animal Testing

Special considerations for animal testing in adrenal toxicity were outlined in a recent review (Harvey et al., 2007). A number of things should to be taken into account. To avoid increased glucocorticoid levels resulting from the stress of handling, sham dosing should be conducted seven to 10 days prior to the onset of dosing. If females are used, their stage of estrous cycle needs to be coordinated. Blood samples should be taken at the same time of day and should be completed within three minutes of cage disturbance to ensure nonstressed control values. Animal dosing should be via oral gavage (or the route of human exposure) to provide adequate translational relevance. A particularly useful method of testing of chemicals for effects on adrenocortical competence is the ACTH stimulation test in which ACTH is administered subcutaneously following a predetermined period of exposure of test animals to the compound of interest. Blood samples can be collected within one to three hours for assessment of glucocorticoid levels. A direct impact of the compound on the adrenal gland will be assumed if ACTH fails to elevate corticosterone levels to those seen in controls.

Mineralocorticoids

The adrenals are essential to life, mainly because of the salt-retaining function of the z. glomerulosa, but to a lesser extent, because of the stress management role of the adrenal cortex. Loss of mineralocorticoid production by the cortex results in a life-threatening retention of potassium and hypovolemic shock associated with the excessive urinary loss of sodium, chloride, and water. Aldosterone, the most important mineralocorticoid, participates in the regulation of renal sodium and potassium balance, thereby affecting blood pressure homeostasis (Hadley and Levine, 2007). Aldosterone promotes sodium reabsorption and increases the excretion of potassium and hydrogen ions by the kidney (Fig. 21-6). The ratio of the concentrations in the urine of Na⁺/K⁺ decreases when blood levels of aldosterone increase, and vice versa. All routes of loss of sodium and potassium from the body are controlled by aldosterone. Thus, an increase in the concentration of aldosterone in blood also decreases the Na⁺/K⁺ ratio of saliva, sweat, and feces.

In theory, chemicals that target steroidogenic enzymes in the pathway to aldosterone secretion could affect both aldosterone and glucocorticoid (StAR, CYP11A1, and CYP21) or only aldosterone synthesis and secretion (CYP11B1; see Harvey et al., 2007).

The aldosterone-producing cells of the zona glomerulosa are controlled separately by other tropic substances, but may require permissive concentrations of ACTH. Unlike the cells of the zonae fasciculata and reticularis, regulation of release from the zona glomerulosa is not primarily via ACTH. The renin–angiotensin system is the primary regulator of aldosterone secretion in the adrenal cortex (Porterfield, 2001). Renin is a proteolytic enzyme synthesized and secreted in the renal juxtaglomerular apparatus (JGA). Release of renin from the JGA is stimulated by decreases in blood pressure or volume, or reduced plasma sodium. Once released, the enzyme (renin) encounters renin substrate (angiotensinogen) of hepatic origin, which is already in circulation and serves as a prohormone reservoir for angiotensin. In the blood stream, renin cleaves angiotensinogen to become angiotensin I (AgI, a decapeptide), and angiotensin converting enzyme (ACE) further cleaves AgI to produce angiotensin II (AgII). This octapeptide (AgII) is one of the most potent known physiological vasoconstrictors. However, it also acts as a tropic hormone for the zona glomerulosa cells of the adrenal cortex to stimulate synthesis and secretion of aldosterone. Because AgII constricts resistance vessels in the body to increase blood pressure, and decrease vascular capacity, this contributes to a feedback mechanism to decrease secretion of renin by JGA cells. Further, by stimulating aldosterone secretion (long loop feedback), the resulting increase in Na⁺ and water retention returns blood volume and osmolality to normal. This further inhibits release of renin, providing a physiological feedback loop of regulation.

Fetal Adrenal

The adrenal cortex in the human fetus differs both structurally and functionally from that in the adult. A specialized fetal adrenal cortex exists in primates during late gestation (Mesiano and Jaffe, 1997). The cortex is composed of large polyhedral cells that produce abundant cortisol and estrogen precursors. The hormones secreted by the cortex are important for normal development of the fetus, and the steroid precursor dehydroepiandrostreone is converted to estriol by the placenta. The cells of the fetal cortex are produced in the outer cortex and migrate medially, where they undergo hypertrophy and eventually apoptosis. After birth, there is a rapid regression, apoptosis, and lysis of the fetal cortex with dilation of cortical capillaries and replacement by the typical three cortical zones. The fetal adrenal cortex is proportionately much larger than the adult gland because of the large size of the fetal zone. It is important not to misinterpret this as a lesion in neonatal primates since it represents physiological replacement of the fetal cortex with the definitive postnatal adrenal cortex.

X-Zone

Similar to the fetal cortex in primates, the X-zone in the mouse adrenal cortex is also a unique physiological phenomenon. In contrast to the fetal cortex of primates, the X-zone develops postnatally in the inner cortex of mice and is fully formed at weaning. Its function is unknown, but it may be similar to the fetal zone in primates. After weaning, the X-zone degenerates at variable rates, depending on the sex of the mouse. In male mice, the X-zone undergoes degeneration at puberty with accumulation of intracellular fat globules. In females, the zone undergoes slow regression and degeneration during the first pregnancy. As with the fetal zone in primates, it is important not to misinterpret the degeneration associated with regression of the X-zone in mice as a lesion.

Because it is classified as a specialized postganglionic (sympathetic) neuron, the adrenal medulla is a functional extension of the nervous system and is also called the sympathoadrenal system. Thus, the adrenal medulla comprises a branch of the sympathetic division of the autonomic nervous system. Located in the center of the adrenal gland, the physiological role of the medulla is to help maintain the constant internal environment of the body. The medulla is composed of cells called pheochromoblasts (specialized sympathetic ganglia), also known as chromaffin cells, which are the site of catecholamine synthesis and secretion. Chromaffin cells derive their name from the chromaffin reaction, the formation of a brown color due to oxidation of catecholamine stores (Tischler et al., 2010). These cells are true neuroendocrine cells, which provide a direct interface between the nervous and endocrine systems. That is, sympathetic, cholinergic stimulation of the cell bodies results in secretion of catecholamines, which behave as hormones by entering the circulation and producing true endocrine effects throughout the body. Secretory granules in the chromaffin cells are the site of storage for the catecholamines, which are rapidly released upon receipt of neuronal input from a preganglionic neuron along the splanchnic nerve pathway. As a result of this stimulation, the catecholamines, epinephrine, and norepinephrine are released into the circulation. Peptides are also reported to reside in chromaffin cells. These include chromogranins and the neuropeptides enkephalin (Schultzberg et al., 1978), neuropeptide Y (NPY, Varndell et al., 1984), substance P (Kuramoto et al., 1985), vasopressin and oxytocin (Hawthorn et al., 1987), galanin (Holgert et al., 1994), and neurotensin (NT, Holgert et al., 1994).

Sympathetic Response

The general functions of the sympathetic division of the autonomic nervous system can be summarized as follows: (a) ensuring reciprocity to counteract and balance the tonic effects of parasympathetic stimulation to visceral structures, (b) assisting in the maintenance of the steady state functions of the body (digestion, secretion, vasomotor tone, etc), and (c) assisting in the mobilization of body reserves to meet unusual or emergency situations (fear, fright, injury, etc). Dr. Walter Cannon in 1929 was the first to suggest that activation of the sympathetic nervous system prepares an animal for “fright, fight, or flight.”

Catecholamines

The sites of catecholamine synthesis in the body are the CNS, postganglionic sympathetic neurons, and the adrenal medulla. It is the adrenal medulla that is the major site of epinephrine production. In fact, 80% of the medullary output of catecholamines in humans is epinephrine. In the catecholamine biosynthetic pathway, tyrosine is acted on by tyrosine hydroxylase to produce dopa, which is converted to dopamine by dopa decarboxylase. Dopamine in turn is acted on by dopamine β-hydroxylase to form norepinephrine, which is converted to epinephrine by PNMT (Hedge et al., 1987). Tyrosine hydroxylase is the principal rate-limiting step in catecholamine synthesis. The conversion of tyrosine to dopa and dopamine occurs within the cytosol of chromaffin cells. Dopamine then enters the chromaffin granule, where it is converted to norepinephrine. Norepinephrine leaves the granule and is converted to epinephrine in the cytosol, and epinephrine, reenters and is stored in the chromaffin granule. In contrast to the synthesis of catecholamines which occurs in the cytosol, neuropeptides and chromogranin-A proteins are synthesized in the granular endoplasmic reticulum and are packaged into granules in the Golgi apparatus. Release of the catecholamines is stimulated by acetylcholine from cholinergic preganglionic neurons. Catecholamine release is stimulated by acetylcholine deriving from preganglionic neurons along the splanchnic nerve. Acetylcholine activates nicotinic and muscarinic cholinergic receptors, which cooperatively stimulate the secretory response by increasing a Ca⁺⁺ flux and activating protein kinase C. This flux involves Ca⁺⁺ from both extracellular influx and intracellular stores. Physiological activators of release include decreased blood pressure, decreased blood glucose, decreased oxygen availability, stress or anxiety, cold, exercise, and postural hypotension. Rather than staining for the chromaffin reaction to identify catecholamine-producing cells, more recently, immunostaining for catecholamine-synthesizing enzymes tyrosine hydroxylase (TH), dopamine-β-hydroxylase (DBH), and PNMT is used to identify individual cells that produce catecholamines, and whether they synthesize epinephrine (Tischler et al., 2010).

Catecholamines generally affect all tissues. However, the most pronounced effects are on the (a) heart, (b) liver, (c) skeletal muscle, (d) adipocytes, (e) vascular smooth muscle, and (f) bronchial smooth muscle.

Adrenergic Receptors

Catecholamine actions are mediated through interactions with specific classes of adrenergic receptors on target cell membranes. There are two major types of these receptors, known as α and β adrenergic receptors. The α and β classification is further divided into at least two subtypes of each, α₁ and α₂, and β₁ and β₂ (Table 21-3). Each type of receptor has its own unique pattern of distribution throughout the body (Porterfield, 2001). The relative number of each receptor type in each target organ determines, in part, the nature of the response of the organ to the catecholamines. For example, α₂ receptors predominate on insulin-secreting pancreatic B cells, whereas β receptors predominate on glucagon-secreting pancreatic A cells. The net effect is that catecholamines inhibit insulin (decreased cAMP production) and stimulate glucagon (increased cAMP production) release so that glucagon can facilitate an anti-insulin effect by increasing blood glucose. Epinephrine and norepinephrine display similar affinities for the different receptor types, with the exception of β₂ receptors that bind epinephrine 10 times greater than norepinephrine. As a result, epinephrine participates more than norepinephrine in mediating its metabolic effects. Therefore, receptor type variation on target tissues contributes to the diversity with which the sympathetic response exerts its specific effects.

General Toxicity

Examples of specific chemicals that target chromaffin cells include toxins that block voltage-gated ion channels (Alvarez et al., 2008), and bacterial toxins, which block exocytosis of secretory granules, thereby, preventing catecholamine release (Gasman et al., 1999). The effects of these toxins on whole animal physiology are relatively unimportant due to their more global and lethal effects on the nervous system. Furthermore, the relevance of the adrenal medulla as a target organ in toxicology is highly dependent on the species, strain, and sex of the animals studied (Tischler et al., 2010). The most common pathological changes seen in the adrenal medulla in toxicological studies involve proliferative lesions classified as nodular hyperplasia, although degenerative changes can also occasionally be observed.

Pheochromocytoma

The adrenal medulla can undergo a series of proliferative changes ranging from diffuse hyperplasia to benign and malignant neoplasms. The latter neoplasms have the capacity to invade locally and metastasize to distant sites. Larger benign adrenal medullary proliferative lesions are designated pheochromocytomas.

Pheochromocytomas are relatively rare in humans (Greim et al., 2009). The incidence is about 1/100,000, and both sexes are affected equally. The peak incidence is ages 40 to 50, and the tumors occur bilaterally in only 10% of the cases. About 10% are malignant, 10% occur outside of the adrenal gland, and 10% are considered to be hereditary (Gimm, 2005). Pheochromocytomas in humans and rats are both composed of chromaffin cells with variable numbers of hormone-containing secretory granules. There appears to be a striking species difference in the response of medullary chromaffin cells to mitogenic stimuli, with rats being very sensitive compared to humans. In humans, pheochromocytomas are uncommon except in patients with inherited clinical syndromes of multiple endocrine neoplasia (MEN). These tumors in rats usually do not secrete excess amounts of catecholamines, whereas human pheochromocytomas episodically secrete increased amounts of catecholamines, leading to hypertension and other clinical disturbances. Further, no data are available to suggest that pheochromocytomas may be inducible in humans. In general, they are usually indolent tumors in humans and mice. These observations have increasingly led regulatory agencies to diminish the need of investigating rat pheochromocytomas for purposes of risk assessment (Tischler et al., 2010).

Proliferative lesions occur in the adrenal medulla with high frequency in many strains of laboratory rats. The incidence of these lesions varies with strain, age, sex, diet, and exposure to drugs and a variety of environmental agents. Studies from the NTP historical database of two-year-old F344 rats have reported that the incidence of pheochromocytomas was 17.0% and 3.1% for males and females, respectively. Malignant pheochromocytomas were detected in 1% of males and 0.5% of females. In addition to F344 rats, other strains with a high incidence of pheochromocytomas include Wistar, NEGH (New England Deaconess Hospital), Long-Evans, and Sprague–Dawley. Pheochromocytomas are considerably less common in Osborne–Mendel, Charles River, Holtzman, and WAG/Rij rats. Most studies have revealed a higher incidence in males than in females. Crossbreeding of animals with high and low frequencies of adrenal medullary proliferative lesions results in F1 animals with an intermediate tumor frequency. For reasons not well understood, spontaneous pheochromocytomas are largely observed in aging male laboratory rats, but are relatively rare in mice (Greim et al., 2009). The differences between rats and mice have led to the suggestion that the mouse is a more suitable model for humans as regards adrenal medullary risk assessment. These medullary lesions, more frequently observed in males, can be induced by a variety of xenobiotic agents (Tischler et al., 2010).

Pheochromocytomas in rats differ from those in all other species in that they are common, often bilateral, and can be induced by many chemicals. An extensive listing of chemicals that are effective at producing proliferative lesions in rats and mice has been provided in reviews by Greim et al. (2009) and Tischler et al. (2010). Warren et al. (1966) studied over 700 pairs of rats and found that more than 50% of irradiated male rats developed adrenal medullary tumors. Substances that have been shown in NTP studies to induce pheochromocytoma in rats include metal compounds, halogenated and nonhalogenated aliphatic or aromatic hydrocarbon, aromatics, aromatic amines as well as other dyestuff, pesticides, and pharmaceuticals (Greim et al., 2009). Vitamin D is the most powerful mitogenic stimulus to cause chromaffin cell proliferation in the adrenal medulla in rats (Greim et al., 2009). Because the vitamin D effect has been seen in vivo, but not in vitro, it is thought to result from impaired calcium homeostasis, resulting in hypercalcemia.

In animals, there is no indication as yet that pheochromocytomas are induced by chemicals working through genotoxic mechanisms (Ozaki et al., 2002). Proliferation of chromaffin cells in male rats occurred following exposure to aromatic amines such as p-chloroaniline (Chhabra et al., 1991). Rather, proliferation of chromaffin cells in rats can be stimulated by chemicals that induce uncoupling of mitochondrial respiration, hypoxia, disturbances in calcium homeostasis, acute stress, and overfeeding (Greim et al., 2009). Many of the chemicals that induce pheochromocytoma in animals are uncouplers of oxidative phosphorylation, for example, acrylamide (Howland and Alli, 1986) or furan (Mugford et al., 1997).

Unlike in humans, in which they secrete high levels of epinephrine, rat pheochromocytomas are usually characterized by predominant or exclusive production of norepinephrine. Chemicals that induce pheochromocytomas in rats are pharmacologically diverse, and usually stimulate chromaffin cell proliferation by neutrally derived signals, or agents that regulate catecholamine synthesis and release. An important class of compounds that affects the rat adrenal medulla includes sugars and sugar alcohols, such as lactitol, xylitol, and sorbitol (Lynch et al., 1996).

Both nicotine and reserpine have been implicated in the development of adrenal medullary proliferative lesions in rats. Both chemicals act by a shared mechanism, because nicotine directly stimulates nicotinic acetylcholine receptors whereas reserpine causes a reflex increase in the activity of cholinergic nerve endings in the adrenal. A short dosing regimen of reserpine administration in vivo stimulates proliferation of chromaffin cells in the adult rat, and the mechanism may involve a reflex increase in neurogenic stimulation via the splanchnic nerve.

Roe and Bar (1985) have suggested that environmental and dietary factors may be more important than genetic factors as determinants of the incidence of adrenal medullary proliferative lesions in rats because the incidence of adrenal medullary lesions can be reduced by lowering the carbohydrate content of the diet. Several of the chemicals that increase the incidence of adrenal medullary lesions, such as sugar alcohols, increase absorption of calcium from the gut. Calcium ions as well as cyclic nucleotides and prostaglandins may act as mediators capable of stimulating both hormonal secretion and cellular proliferation. In summary, three dietary factors have been suggested to lead to an increased incidence of adrenal medullary proliferative lesions in chronic toxicity studies in rats (Roe and Bar, 1985). These are excessive intake of (1) food associated with feeding ad libitum; (2) calcium and phosphorus; and (3) other food components (eg, vitamin D and poorly absorbable carbohydrates), which increase calcium absorption.

Tischler et al. (1995) reported that adult rat chromaffin cells had a marked increase in bromodeoxyuridine (BrdU)-labeled nuclei in vitro following the addition of forskolin (activator of adenylate cyclase) and phorbol myristate (PMA, activator of protein kinase C), whereas mouse chromaffin cells had a minimal response to the same mitogens. This striking difference in sensitivity to mitogenic stimuli may explain the lower frequency of adrenal medullary proliferative lesions in mice compared to many rat strains. The human adrenal medulla, as in mice, has a low spontaneous incidence of proliferative lesions of chromaffin cells. Human chromaffin cells also failed to respond to a variety of mitogenic stimuli in culture (Tischler and Riseberg, 1993). These findings and others suggest that chromaffin cells of the rat represent an inappropriate model to assess the potential effects of xenobiotic chemicals on chromaffin cells of the human adrenal medulla.

A relationship exists between adenohypophyseal (anterior pituitary) hormones and the development of adrenal medullary proliferative lesions. For example, the long-term administration of growth hormone is associated with an increased incidence of pheochromocytomas as well as the development of tumors at other sites. Prolactin-secreting pituitary tumors, which occur commonly in many rat strains, also play a role in the development of proliferative medullary lesions. In addition, several neuroleptic compounds that increase prolactin secretion by inhibiting dopamine production have been associated with an increased incidence of proliferative lesions of medullary cells in chronic toxicity studies in rats.

In long-term animal studies, pheochromocytomas often are accompanied by tumors or toxic effects in other organs. They often are seen in cases involving renal, lung, and hepatic toxicity, in addition to endocrine disturbances (Greim et al., 2009). They are associated with the following conditions: hypoxia, uncoupling of oxidative phosphorylation, disturbances of calcium homeostasis, or disturbances of the hypothalamic endocrine axis (Greim et al., 2009). Additionally, pheochromocytomas often develop following treatment with substances that target enzymes in catecholamine synthesis, receptor tyrosine kinase (RET), hypoxia-inducible factor (HIF), succinate dehydrogenase, or fumarate hydratase. As with humans, pheochromocytomas can also occur in mice and rats in conjunction with MEN. This was seen in male rats after administration of propyl gallate (NTP, 1982).

In Vitro Testing

A commonly employed cell line used in neurobiology is the PC12 pheochromocytoma line derived from a rat adrenal medullary tumor (Greene and Tischler, 1976). These cells respond to nerve growth factor (NGF) by ceasing replication, extending neurites, and increasing a number of neuronal markers. This cell line is widely used in neurotoxicological studies. However, in using them, it must be considered that rat PC12 cells are representative only of a subset of chromaffin cells, and may differ in some ways from normal and neoplastic chromaffin cells of other species. At this time, there are no human chromaffin cell lines, despite numerous efforts to establish such a cell line (Tischler et al., 2010).

The PC12 cell line has been useful in determining intracellular mechanisms at the molecular level that are involved in chromaffin cell signaling and proliferation. Studies with PC12 cells have shown that substances which inhibit mitochondrial function (cyanide, rotenone) or uncouple oxidative phosphorylation (dinitrophenol, p-trifluoromethoxyphenyl hydrazone) stimulate catecholamine secretion in the same way as occurs under hypoxic conditions. This is thought to be dependent upon Ca⁺⁺ influx through voltage-gated channels (Taylor et al., 2000). Proliferation of chromaffin cells can also be induced in vitro by activators of adenylate cyclase that stimulates neuropeptide receptors, as well as stimulators of protein kinase C, which stimulates muscarinic cholinergic receptors.

General Anatomy

The thyroid gland consists of two lobes of endocrine tissue located just below the larynx on each side of the trachea with an isthmus connecting the two lobes (Fig. 21-7; Capen and Martin, 1989; Capen, 2001). A second cell type is also present, the C-cells, or parafollicular cells composing the intrafollicular spaces (Capen and Martin, 1989; Capen, 2001). The C-cells synthesize and secrete calcitonin (CT), a hormone involved in calcium homeostasis (see “Parathyroid gland” section for more details).

The thyroid secretes two hormones known as thyroxine (T₄) and triiodothyronine (T₃) (Capen and Martin, 1989; Capen, 2001; Hedge et al., 1987). Both of these hormones are produced in epithelial cells in the basic functional unit of the thyroid known as the follicle (Fig. 21-8). Each follicle consists of a sphere of epithelial cells surrounding a colloidal core. The colloid material is composed of the glycoprotein thyroglobulin (TGB), which acts as a storage depot for T₄ and T₃ (Capen and Martin, 1989; Capen, 2001). In humans, about three months of thyroid hormone is stored as TGB in the colloid (Hedge et al., 1987). This represents the largest reserve of any stored hormone in the body. Generally, the level of secretory activity of follicular cells can be estimated as a direct function of their height. Cells involved in synthesizing thyroid hormone are columnar in shape, whereas quiescent cells are cuboidal (Capen and Martin, 1989; Capen, 2001).

T₄ and T₃ are important regulators of overall metabolism, and their effects are regulated within a long time frame (Capen, 2001; Hedge et al., 1987). Essentially all tissues are to some degree targets for thyroid hormone. However, the primary target tissues for thyroid hormone include the liver, kidney, heart, brain, pituitary, gonads, and spleen (Capen and Martin, 1989; Capen, 2001; Zoeller et al., 2007).

Some studies indicate that xenobiotics directly affect the structure of the thyroid gland (Capen, 2001). For example, some environmental chemicals such as heavy metals and red dye #3 are known to decrease the size of the colloid space (Bronnikov et al., 2005; Capen and Martin, 1989; Capen, 2001). This is thought to reduce the space required for storing hormones, leading to an impaired ability of the thyroid gland to synthesize and store thyroid hormones. Chemicals such as PCBs are known to alter the appearance of the epithelial cells so that they are hypertrophic or hyperplastic in nature (Capen, 2001; Langer, 2010). This is thought to lead to excessive thyroid hormone production.

Thyroid Hormone Structure and Synthesis

Thyroid hormones are composed of two modified, covalently linked tyrosine amino acids (Fig. 21-9, Capen, 2001; Hedge et al., 1987). Each of the aromatic rings of the tyrosines contains one or two iodides. T₄ contains two iodides on each aromatic ring for a total of four, while T₃ contains two iodides on the tyrosine closest to the amino acid moiety (amino and carboxy groups), and one iodide on the outer aromatic ring (Capen, 2001; Hedge et al., 1987). The iodide is derived from dietary intake and is required for biologic activity. While the thyroid gland synthesizes and secretes both T₄, and T₃, it primarily releases T₄. In fact, about 90% of thyroid hormone secreted by the thyroid gland is in the form of T₄ in humans (Hedge et al., 1987).

TGB, a glycoprotein containing large numbers of tyrosine amino acid residues, is synthesized in the epithelial cell and serves as the backbone for thyroid hormone synthesis (Capen, 2001; Hedge et al., 1987). Iodine in the form of iodide (I⁻) is actively transported into the epithelial cell, where it is oxidized to I₂ by thyroid peroxidase. At the apical membrane, I₂ combines with tyrosine residues on TGB to form monoiodotyrosine (MIT) and diiodotyrosine (DIT), which remain attached to TGB (Capen, 2001; Hedge et al., 1987). Coupling between MIT and DIT occurs such that a combined MIT and DIT forms T₃; whereas a combined DIT and DIT forms T₄ (Hedge et al., 1987; Capen, 2001).

The iodinated TGB is stored in the follicular lumen as colloid until the thyroid gland is stimulated to secrete hormone (Capen, 2001; Hedge et al., 1987). Upon stimulation, iodinated TGB is endocytosed into the epithelial cell and transported in the direction of the basal membrane where lysosomal enzymes hydrolyze peptide bonds to release T₃ and T₄ for passive diffusion into the circulation (Capen, 2001; Hedge et al., 1987). Remaining MIT and DIT are recycled in the epithelial cell for synthesis of new TGB (Capen, 2001; Hedge et al., 1987).

T₄ from the thyroid gland can be peripherally converted to T₃ (active hormone) or reverse T₃ (rT₃, the inactive metabolite) then successively deiodinated by the monodeiodinases (Capen, 2001; Hedge et al., 1987). About 40% of circulating T₄ is metabolized to T₃ by 5′-monodeiodinase, and 40% is converted to rT₃ by 5-monodeiodinase (Capen, 2001; Hedge et al., 1987; Zoeller et al., 2007). Additionally, about two-thirds of T₃ and all rT₃ in circulation are produced from T₄ by peripheral conversion (Capen, 2001; Hedge et al., 1987). As a result, the ratio of circulating T₄ and T₃ does not reflect the ratio of these two substances when they were released from the thyroid gland. Based on a variety of observations, it appears that circulating T₄ levels provide a “sink” of prohormone that can serve as a ready supply for peripheral conversion to T₃ (the active form; Hedge et al., 1987).

Several studies indicate that xenobiotics can interfere with thyroid gland function by adversely affecting the process of thyroid hormone synthesis (Jugan et al., 2010; Kortenkamp, 2008; Lynch et al., 2002; Mastorakos et al., 2007; Sauvage et al., 1998). For example, environmental chemicals such as perchlorate, chlorate, and bromate inhibit uptake of iodide and thus, decrease thyroid hormone synthesis (Crofton, 2008). Some chemicals such as genistein and daidzein in soy products, thionamides, and substituted phenols can inhibit thyroid peroxidase, blocking incorporation of iodide into TGB (Delclos and Newbold, 2007; Doerge and Chang, 2002). Further, some chemicals may interfere with monodeiodinases, leading to decreased levels of thyroid hormones. Specifically, studies have shown that red dye #3 and propylthiouracil inhibit 5′-monodeiodinase, leading to reduced serum levels of T₃ (Capen, 2001; Crofton, 2008).

Thyroid Hormone Binding Proteins

Once released into the blood, thyroid hormones are rapidly bound to high affinity serum binding proteins (Capen, 2001; Hedge et al., 1987). The result is that less than 1% of the T₃ (99.7% bound) and less than 0.1% of the T₄ (99.97% bound) are free in circulation (Hedge et al., 1987). Only the small unbound fraction of the total hormone pool has access to receptors in target cells, and thus only the unbound fraction can exert biological activity.

There are three types of thyroid hormone binding proteins: thyroid binding globulin (TBG), thyroxine-binding prealbumin (TBPA), and albumin (Hedge et al., 1987; Capen, 2001). TBG binds about 80% of the thyroid hormones, whereas TBPA and albumin each bind about 10% of the thyroid hormones (Hedge et al., 1987). Environmental chemicals such as the PCBs and PBDEs are known to displace thyroid hormones from serum binding proteins (Jugan et al., 2010; Patrick 2009; Yamauchi and Ishihara, 2006). The displacement of thyroid hormones from the binding proteins often leads to a rapid decline in serum thyroid hormone levels.

Thyroid Hormone Receptors

Thyroid hormones act by binding to thyroid hormone receptors (TRs) (Zoeller, 2005). TRs are members of the nuclear receptor superfamily of ligand-inducible transcription factors. TRs can form homodimers or heterodimers with other nuclear hormone receptors such as the retinoid X receptor (Zoeller, 2005). The homodimers and heterodimers bind to thyroid hormone response elements located in target genes and interact with coactivators and corepressors to regulate transcription. In humans, thyroid hormone receptors are the products of two genes that encode three thyroid hormone receptor isoforms known as TRα, TRβ1, and TRβ2 (Zoeller, 2005). While all three isoforms are present in most tissues, their expression differs spatially and temporally during development (Zoeller, 2005). TRα is abundant in the brain, heart, and immune system, whereas TRβ1 is particularly expressed in the brain, liver, and kidney (Zoeller, 2005).

Environmental chemicals can interfere with thyroid hormone binding to TRs and thyroid-hormone related transcription at multiple levels (Jugan et al., 2010; Kitamura et al., 2002; Patrick, 2009; Zoeller, 2005). First, some chemicals such as PBDEs can bind directly to TRs and induce either agonistic or antagonist effects (Zoeller, 2005). Interestingly, some PBDE congeners have different affinities for TRα and TRβ, whereas some congeners can bind to more than one isoform. Second, some environmental chemicals interfere with thyroid hormone binding to receptors via indirect mechanisms (Zoeller, 2005). In such cases, it is thought that the chemicals exert their effects by promoting coactivators or inhibiting corepressors. For example, BPA impairs thyroid hormone action by inhibiting T₃ binding to TR and by recruiting the nuclear corepressor N-CoR to the TR, resulting in repression of transcription (Patrick, 2009; Zoeller, 2005). Similarly, some PCBs are able to suppress TR/coactivator (SRC-1) complex-mediated transactivation, leading to suppression of TR-mediated transcription (Crofton, 2008). Third, some xenobiotics can interfere with cross talk between TRs and other nuclear receptors. For example, hydroxyl-PCBs can partially dissociate the heterodimer TR/retinoic acid receptor from the T₃-response element (Crofton, 2008). Finally, some chemicals such as BPA and phthalates can inhibit expression of TRs (Patrick, 2009).

Thyroid Hormone Clearance

The main pathway for clearance of thyroid hormones from the serum is via conjugation to glucuronic acid or sulfate by phase II enzymes such as glucuronyl transferases (UDPGTs) and sulfo transferases (SULTs), respectively (Patrick, 2009; Zoeller, 2010). The metabolites then are transported across plasma membranes for elimination by phase III transporters, including the multidrug resistance protein 1 and the multidrug resistance-associated protein 2 (MRP2).

Some studies indicate that some xenobiotics may increase the clearance of thyroid hormones from the serum, limiting the availability of thyroid hormones to act on tissues and often resulting in symptoms of hypothyroidism (Brouwer et al., 1998; Yamauchi and Ishihara, 2006). For example, coplanar and noncoplanar PCB congeners have been shown to induce UDPGTs and SULTs, resulting in low serum T₄ levels (Crofton, 2008; Patrick, 2009). In contrast, a few studies have shown that xenobiotics such as pentachlorophenol or triclosan may inhibit SULTs, increasing the availability of thyroid hormones to act on tissues (Crofton, 2008; Jekat et al., 1994; Patrick, 2009). Studies have also shown that xenobiotics such as dioxin, rifampicin, and phenobarbital may decrease the transport of thyroid hormones into the brain and liver by inhibiting phase III transporters (Crofton, 2008).

Regulation of Thyroid Hormone Release

Thyroid hormone secretion is regulated by thyroid-stimulating hormone (TSH, thyrotropin) from the anterior pituitary gland (Capen, 2001; Hedge et al., 1987). The rate of release of TSH is under a hypothalamic–pituitary–thyroid regulatory axis involving negative feedback (Fig. 21-10). Specifically, a hypothalamic releasing hormone known as TRH directly stimulates release of TSH from the anterior pituitary gland (Hedge et al., 1987). TSH then increases secretion of T₄ and T₃ from the thyroid gland. In turn, T₄ and T₃ can feedback to the anterior pituitary to inhibit TSH release and they can feedback to the hypothalamus to inhibit TRH release. Interestingly, most inhibition by thyroid hormones is by T₄, although T₃ can also provide some degree of inhibition (Hedge et al., 1987). Further, the feedback effect is mediated primarily at the level of the anterior pituitary, although some degree of negative feedback occurs at the level of the hypothalamus (Hedge et al., 1987).

In addition to these methods of regulating thyroid hormone secretion, there is a circadian rhythm of TRH and TSH release, with a decrease following the onset of sleep (Hedge et al., 1987). Overall, the thyroid axis responds rapidly at the hypothalamo-pituitary unit, but beyond this level, the system is governed by processes that have extremely long time constants. Therefore, the long half-lives of thyroid hormones in circulation dampen the diurnal rhythm that is obvious in TSH levels. As a result, this rhythm is not reflected in circulating thyroid hormone concentrations. This contributes to the “sluggishness” of the system.

Xenobiotics have been shown to alter the ability of the hypothalamic–pituitary–thyroid regulatory axis to control thyroid hormone levels (Patrick, 2009; Zoeller, 2005). For example, chemicals such as PBDEs may increase TSH levels, leading to increased levels of T₄ and T₃ (Darnerud et al., 2001). Alternatively, xenobiotics may inhibit TRH or TSH levels, leading to decreased levels of T₄ and T₃ (Patrick, 2009).

Physiological Effects

Thyroid hormones influence nearly every tissue in the body, in a variety of ways (Capen, 2001; Hedge et al., 1987; Patrick, 2009). In spite of complexities of hormone action on target tissues, it is reasonably accurate to simply view thyroid hormone as the primary determinant of the overall metabolic rate of the body. The effects are exaggerated in states of thyroid excess (hyperthyroidism) or deficiency (hypothyroidism) and produce the clinical and biochemical manifestations of these disorders. In general, thyroid hormone stimulates both anabolic and catabolic biochemical pathways; however, its over-riding effect is catabolism (energy mobilization).

Thyroid hormone also produces significant effects on growth and development (Capen, 2001; Hedge et al., 1987; Zoeller et al., 2002; Zoeller, 2005). It is essential for normal development of the CNS and for maturation of the skeleton (Dickerson and Gore, 2007; Gore, 2011; Zoeller et al., 2002). A deficiency of thyroid hormone in early life leads to a delay in development of the brain in animal models and humans (Dickerson and Gore, 2007; Gore, 2011; Zoeller et al., 2002). Brain development is especially dependent on thyroid hormone during the first several months after birth in humans (Zoeller, 2005). If thyroid hormone levels are inadequate during this period, severe irreversible mental retardation in the form of cretinism occurs. Early diagnosis and immediate replacement with thyroid hormones can prevent these effects, and are; therefore, essential.

Thyroid Toxicity

Given the influence of thyroid hormones on numerous tissues in the body, it is not surprising that xenobiotics that affect thyroid hormone levels often cause symptoms of hypothyroidism, hyperthyroidism, or lead to significant impairment in brain development and function. Below are some specific examples of environmental chemicals that have been shown to affect thyroid hormone levels through a variety of mechanisms, resulting in adverse physiological outcomes.

PCBs

PCBs are some of the best characterized thyroid disrupting chemicals (Boas et al., 2009; Dickerson and Gore, 2007; Gore, 2011; Jugan et al., 2010; Patrick, 2009; Zoeller, 2010). PCBs are industrial chemicals that were widely used in capacitors and transformers. They are made of two phenyl rings with varying degrees of chlorination, resulting in 209 different congeners. The production of PCBs was banned in the 1970s, but unfortunately, PCBs are persistent chemicals and thus, they are still routinely found in humans and wildlife blood samples and tissues.

PCBs are known to interfere with the thyroid system in a manner that leads to serious neurocognitive effects (Gore, 2011; Porterfield, 2000; Zoeller et al., 2002). Several human studies indicate that PCB exposure in prenatal life is associated with lower full-scale and verbal IQ scores and less short-term and long-term memory and attention in postnatal life (Porterfield, 2000; Zoeller et al., 2002). Rodent studies indicate that prenatal PCB exposure increases hyperactivity (Porterfield, 2000; Zoeller et al., 2002). Further, rodent studies indicate that PCB exposure causes hearing disorders because the developing auditory system is sensitive to thyroid hormones (Porterfield, 2000; Zoeller et al., 2002).

While PCBs may exert negative effects on animals and humans via several mechanisms, the most common pathway is thought to include PCB inhibition of thyroid hormone levels or activity (Gore, 2011; Patrick, 2009). Several studies indicate that PCBs decrease the levels of thyroid hormone by inhibiting thyroid hormone synthesis and/or increasing the metabolism of thyroid hormones by increasing phase II enzymes (Zoeller, 2010). Further, some studies indicate that PCBs interfere with thyroid hormone action by inhibiting the binding of thyroid hormones to binding proteins or blocking the ability of thyroid hormones to bind to TRs (Zoeller, 2010).

PBDEs

PBDEs are also well-known thyroid disrupting chemicals (Boas et al., 2009; Jugan et al., 2010; Zoeller, 2010). These chemicals are flame retardants that are used in a variety of products, including electric equipment, clothing, furniture, carpeting, plastics, and paints. PBDEs are not chemically bound to products in which they are used; thus, they can leach from the products into human and wildlife tissues and into the environment over time.

The structure of PBDEs often resembles that of PCBs. Thus, it is not surprising that many of the toxic effects of PBDEs are similar to those elicited by PCBs. Like PCBs, PBDEs have been shown to inhibit thyroid hormone levels and/action, leading to serious neurocognitive deficits. PBDEs inhibit thyroid hormone levels by inducing hepatic phase II enzymes, resulting in increased metabolism of circulating thyroid hormones (Zoeller, 2010). They also can downregulate proteins required for transporting thyroid hormone into target cells and they can bind to TRs, blocking the ability of thyroid hormones to bind to TRs (Zoeller, 2010).

Perchlorate

Perchlorate is another thyroid disrupting chemical (Crofton, 2008; Jugan et al., 2010; Patrick, 2009; Zoeller, 2010). This chemical is widely used as a rocket propellant as well as a chemical in fireworks and airbag deployment systems. Perchlorate is also used in pharmaceutical industries. It is a highly stable and water-soluble compound and thus, it is known to persist in the environment, particularly in the water and food supply. While less is known about the effects of perchlorate on the thyroid system than about PCBs and PDBEs, a few studies indicate that perchlorate exposure inhibits thyroid hormone levels, possibly leading to hypothyroid-like outcomes (Crofton, 2008; Patrick, 2009). The mechanism of action of perchlorate is thought to be primarily by reducing iodide uptake, which ultimately reduces thyroid hormone synthesis (Patrick, 2009). It is important to note, however, that the effects of perchlorate on the thyroid gland have not been well studied and the results to date in humans are equivocal.

Pesticides

Pesticides such as dichlorodiphenyltrichloroethane (DDT) and hexachlorobenzene (HCB) are known thyroid disrupting chemicals (Boas et al., 2009). While both DDT and HCB have been banned in many countries, they persist in the environment due to their long environmental half-lives and thus, they may place humans and wildlife at risk for thyroid disorders. Pesticide mixtures containing DDT have been shown to increase thyroid volume and to induce antibodies that attack the thyroid gland, resulting in autoimmune thyroid disease (Boas et al., 2009; Crofton, 2008; Patrick, 2009). Further, DDT has been shown to inhibit TSH receptors, blocking the ability of TSH to induce secretion of T₄ and T₃ and in turn, resulting in low circulating T₄ and T₃ levels (Boas et al., 2009; Patrick, 2009). HCB has been shown to interfere with thyroid function by blocking the ability of thyroid hormones to bind to TR (Boas et al., 2009; Crofton, 2008; Patrick 2009).

Perfluorinated Chemicals

Perfluorinated chemicals (PFCs) are a family of chemicals used in many products due to their surface protection properties. Such products include stains and oil-resistant coatings, floor polishes, and insecticides. Some studies have shown that a PFC known as perfluorooctane sulfonate (PFOS) decreases T₄ levels in pregnant dams as well as their pups (Boas et al., 2009). The PFC known as perfluorooctanic acid (PFOA) has been shown to decrease T₃ levels. While the mechanisms by which PFOS and PFOA decrease thyroid hormone levels are not completely clear, studies suggest that they upregulate phase II metabolic enzymes in the liver and increase deiodinases in the thyroid gland (Boas et al., 2009).

Bisphenol A

BPA is a suspected possible thyroid hormone disrupting chemical (Boas et al., 2009; Zoeller, 2010). It is primarily used in the manufacture of polycarbonate plastics such as those used in baby bottles, toys, and food containers. It is also a component of dental sealants and the linings of food cans. BPA can leach out of products and enter the blood and organs. In fact, recent studies indicate that over 95% of human urine samples contain BPA, indicating continuous and widespread exposure to this chemical. Given the vast exposure of humans to BPA, it is important to consider its potential effects on the thyroid gland. Some laboratory studies have shown that BPA blocks T₃ action by antagonizing the binding of T₃ to its receptor (Zoeller, 2010). Further, some studies have shown that BPA inhibits T₃-mediated gene expression in cell lines (Zoeller, 2010). While the effects of BPA-induced inhibition of thyroid hormone action in humans are unclear, several studies suggest that BPA leads to symptoms of hypothyroidism or thyroid resistance syndrome in animal models. Further, prenatal BPA exposure has been shown to cause attention deficit hyperactivity disorder like symptoms in rodents in postnatal life (Zoeller, 2010). This effect is thought to be due to the BPA-induced inhibition of normal thyroid function, which in turn affects normal development of the neurological system (Zoeller, 2010).

Phthalates

A few recent studies suggest that phthalates may act as thyroid disrupting agents; however, they are not thought to do so to the same degree as PCBs and PBDEs (Boas et al., 2009; Jugan et al., 2010). Phthalates are used to improve the flexibility of plastics used in a variety of products, including toys, medical tubing, plastic bottles, and cosmetics. To date, a few small human studies have shown that phthalate exposures may alter the levels of T₄ and T₃ levels in adult men and pregnant women (Jugan et al., 2010). Interestingly, a by-product of phthalates produced by gram-negative bacteria has been shown to inhibit thyroperoxidase, an enzyme required for thyroid hormone synthesis (Jugan et al., 2010). This results in low thyroid hormone levels and to symptoms of hypothyroidism in humans. A few rodent studies have shown that a phthalate known as di-n-butyl phthalate decreases T₃ and T₄ in a dose-dependent manner (Boas et al., 2009; Jugan et al., 2010). The consequences of phthalate-induced changes in thyroid hormone levels in humans or rodents are unclear at this time and should be investigated in future studies.

General Anatomy

The parathyroid glands are embedded in the surface of the thyroid gland (Fig. 21-7) (Capen and Rosol, 1989; Hedge et al., 1987). Humans have four parathyroid glands, which are located on the back of side of the thyroid gland (Hedge et al., 1987). The parathyroid glands are composed of chief cells and oxyphil cells (Capen and Rosol, 1989; Hedge et al., 1987). The main function of chief cells is to produce parathyroid hormone (PTH), whereas the function of the oxyphil cells is unknown (Capen and Rosol, 1989; Hedge et al., 1987). In humans, oxyphil cells are absent at birth, appear around puberty, and increase in number with age. Thus, it is thought that they may represent structurally and functionally modified chief cells.

The parathyroid glands are critical for life (Capen and Rosol, 1989). This is largely because PTH helps maintain normal plasma calcium levels (Fig. 21-11) (Capen and Rosol, 1989; Hedge et al., 1987). Calcium is required in optimal concentrations for many of life’s fundamental processes: fertilization, vision, locomotion-muscle contraction, nerve conduction, blood clotting, exocytosis, cell division, and the activity of a number of enzymes and hormones (Hedge et al., 1987). Therefore, the concentrations of calcium in the cellular and extracellular fluids must be maintained at a constant value. When the parathyroid glands are removed or damaged, PTH levels drop, causing a major drop in circulating calcium levels. In turn, this can lead to tetanic convulsions and death (Capen and Rosol, 1989; Hedge et al., 1987).

Parathyroid Toxicity

Xenobiotic exposures may alter the structure of the parathyroid gland (Capen and Rosol, 1989). In some cases, chemicals such as the anticancer drug l-asparaginase cause death of parathyroid cells (Capen and Rosol, 1989; Capen, 2001). Specifically, studies have shown that l-asparaginase selectively destroys chief cells in rabbits. This results in a reduced size and capacity of the parathyroid to release PTH, which eventually leads to tetany and death. Studies have also shown that sublethal doses of heroin cause degenerative changes in the rat parathyroid gland (Barai et al., 2009). These changes are characterized by cytoplasmic vacuolization, pyknotic nuclei in the chief cells, and dying Golgi complexes and mitochondria.

Many xenobiotic exposures have been shown to increase the size of the parathyroid gland. Lead exposure has been shown to significantly increase parathyroid gland weight (Szabo et al., 1991). Some pesticides such as rotenone, malathion, and malaoxin increase proliferation of the chief cells and thus, increase the size of the parathyroid gland (Abdo et al., 1988; Reuber, 1985). This scenario often results in parathyroid gland cancer. Further, fungicides such as hexachlorobenzene and bis(tri-n-butyltin)oxide, the antifreeze ethylene glycol, the broad-spectrum germicide o-benzyl-p-chlorophenol, and the diuretic drug hydrochlorothiazide have been shown to induce proliferation and/or adenomas in rat parathyroid glands (Andrews et al., 1989; Arnold et al., 1986; Bucher et al., 1990; DePass et al., 1986; Lijinsky and Reuber, 1987; National Toxicology Program, 1994; Wester et al., 1990). Similarly, irradiation and coumarin exposure induce parathyroid gland adenomas in rodents (National Toxicology Program, 1993). It is unclear whether all of the effects of xenobiotics are due to a primary effect of the chemical on the parathyroid gland or due to effects on the kidney or bones, which in turn alter the parathyroid gland (secondary effects).

PTH Structure and Synthesis

PTH is a polypeptide hormone that is derived from a precursor molecule called preproparathyroid hormone (Fig. 21-12; Potts, 2005). In humans, preproparathyroid hormone is composed of 115 amino acid residues. The preproparathyroid hormone is cleaved by trypsin-like proteases in the Golgi zone of the chief cells to become proparathyroid hormone, which is 90 amino acids in size. Six amino acid residues are then removed from the proparathyroid hormone by proteases in the Golgi to make PTH (Capen, 2001). PTH is then packaged into secretory vesicles that migrate to the periphery of the cell, where the hormone is secreted via vesicular exocytosis (Capen, 2001; Potts, 2005). Interestingly, the structure of this hormone is highly conserved among many species, including humans, cows, pigs, dogs, rats, and chickens (Potts, 2005).

A few studies indicate that xenobiotics may interfere with the normal synthesis of PTH. Metals such as aluminum and cadmium have been shown to inhibit PTH secretion in a variety of species, including humans (Cannata et al., 1983; Jeffery et al., 1996; Kido et al., 1991; Rignell-Hydbom et al., 2009). Similarly, alcohol consumption has been shown to decrease PTH levels in pregnant rats (Keiver and Weinberg, 2003). Lithium, a drug widely used to manage manic-depressive illnesses, has been associated with a rise in PTH levels as well as abnormally high calcium levels (Matsis et al., 1989). Interestingly, a longitudinal study on Gulf War I veterans showed that depleted uranium exposure was associated with significant decreases in serum levels of PTH as well as significant increases in urinary calcium (McDiarmid et al., 2011; Tissandie et al., 2006).

PTH Receptors

The PTH receptor is a single G-protein-coupled receptor called the PTH/PTHrP receptor (PTHR1) (Potts, 2005). This receptor is the primary receptor that mediates the traditional actions of PTH. However, it is thought other receptors exist that bind to portions of PTH or proparathyroid hormone. In humans, the gene for the PTHR1 is located on chromosome 3 and consists of 14 exons. PTHR1 has been extensively localized in bone and kidney cells (Potts, 2005).

Limited information is available about whether xenobiotics bind PTHR1 and exert agonistic or antagonistic actions. However, one study shows that xenobiotics may alter the expression of PTHR1. Specifically, studies have shown that binge alcohol drinking significantly decreases expression of PTHR1 in male rats (Callaci et al., 2009).

PTH Clearance

PTH is primarily metabolized by the liver and the kidney (Capen and Rosol, 1989). It is thought that the hormone is cleaved into two major fragments (N-terminal and C-terminal) quite quickly by metabolic enzymes. In cows, for example, the half-life of PTH is estimated to be between three and four minutes. In dogs, the half-life of PTH is between five and six minutes, with about 60% of PTH metabolism occurring in the liver.

It is possible that xenobiotics increase the metabolism of PTH and that this leads to abnormally low levels of PTH. Alternatively, it is possible that xenobiotics decrease the metabolism of PTH and that this leads to high levels of PTH. Unfortunately, these issues have not been examined in detail and it is unclear which xenobiotics, if any, directly affect PTH metabolism.

Physiological Effects

The main physiological role of the parathyroid gland is to control circulating calcium levels (Capen and Rosol, 1989; Capen, 2001; Hedge et al., 1987). Given the importance of maintaining normal levels of circulating calcium, multicellular organisms have evolved a complex system of controls to insure this constancy. Thus, PTH works in concert with CT and vitamin D (Capen, 2001; Hedge et al., 1987). CT is secreted from the C cells of the thyroid gland (Capen, 2001; Hedge et al., 1987). Vitamin D is produced from precursors in the skin, and it is obtained from the diet (Capen, 2001; Hedge et al., 1987).

PTH serves to increase circulating calcium levels by increasing the release of calcium from bone (Capen, 2001; Hedge et al., 1987). Bone is remodeled continuously during adulthood by the resorption of old bone by osteoclasts and the subsequent formation of new bone by osteoblasts (Hedge et al., 1987). These two events are responsible for renewing the skeleton, while maintaining its anatomical and structural integrity. Under normal conditions, bone remodeling proceeds in cycles in which osteoclasts adhere to bone and subsequently remove it by acidification and proteolytic digestion. Shortly after the osteoclasts have left the resorption site, osteoblasts invade the area and begin the process of forming new bone by secreting osteoid (a matrix of collagen and other proteins), which is eventually mineralized (Capen, 2001; Hedge et al., 1987). In regulating bone remodeling, PTH binds to osteoblasts and increases IL-6 which, along with other cytokines, causes osteoclast differentiation. In turn, this demineralizes bone and releases calcium from the bone into circulation (Capen, 2001; Hedge et al., 1987).

PTH also serves to increase calcium levels by increasing the tubular reabsorption of calcium by the kidney (Capen and Rosol, 1989). Further, it inhibits the renal reabsorption of phosphate, which aids in increasing the solubility of calcium (Capen, 2001; Hedge et al., 1987). PTH also enhances magnesium reabsorption, inhibits bicarbonate ion reabsorption, and blocks exchange of sodium ions by the tubules (Capen, 2001; Hedge et al., 1987). These actions of PTH result in metabolic acidosis, which favors removal of calcium from plasma proteins and bones (Capen, 2001; Hedge et al., 1987). In turn, this increases circulating levels of ionized calcium (Capen, 2001; Hedge et al., 1987).

CT reduces circulating calcium levels by reversing the action of PTH on bone resorption (Capen and Rosol, 1989). When circulating calcium levels become high, CT is secreted by the C cells of the thyroid (Capen, 2001; Hedge et al., 1987). CT serves to prevent hypercalcemia by shutting down efflux of calcium from bone. It also negatively regulates PTH to prevent kidney calcification (Capen, 2001; Hedge et al., 1987).

Vitamin D₃ (cholecalciferol) is a steroid-like compound that is essential for calcium absorption in the gastrointestinal tract (Capen, 2001; Hedge et al., 1987). It is derived from cholesterol, and the active form is produced from a precursor, 7-dehydrocholesterol. Exposure of the skin to ultraviolet light causes formation of vitamin D, which is biologically inert and must be activated by two sequential hydroxylations (Capen, 2001; Hedge et al., 1987). The first hydroxylation occurs in the liver, and the second occurs in the kidney (Capen, 2001; Hedge et al., 1987). The second conversion is stimulated by PTH activation of 1-hydroxylase enzyme activity. Vitamin D also serves to inhibit PTH actions and build bone (Capen, 2001; Hedge et al., 1987).

Some xenobiotics such as pesticides and fungicides can cause excessive PTH secretion by the parathyroid gland and lead to hyperparathyroidism (Capen, 2001). However, it is important to note that most excessive PTH secretion is due to a parathyroid adenoma that is no longer under the negative feedback control of calcium. The traditional symptoms of PTH excess have been characterized as “stones, bones, and groans” for several reasons (Hedge et al., 1987). First, excessive PTH levels lead to hypercalcemia, hypercalciuria, and hypophosphatemia, which often lead to the formation of renal stones. The constant supply of calcium released via unchecked bone resorption must eventually be excreted, and even though PTH normally elicits calcium reabsorption from the kidney, the consistently elevated calcium exceeds the renal threshold for reabsorption such that calcium leaks into the urine. The hallmark of this disease is abnormally increased bone resorption, leading to severe bone pain.

Xenobiotic exposures such as those to heavy metals may cause low PTH secretion and lead to hypoparathyroidism (Long et al., 1992; Pounds, 1984). This condition leads to a very densely calcified skeleton, hypocalcemia, and hyperphosphatemia. Of great concern is that hypoparathyroidism often leads to tetany and death (Capen, 2001).

Regulation of PTH Release

PTH release is controlled by circulating calcium levels (Capen and Rosol, 1989). The cells in the parathyroid gland, kidney, and other cells that respond to calcium possess recognition sites for circulating calcium levels known as calcium sensors or receptors (Hedge et al., 1987). Recently, the calcium sensor or calcium receptor on the parathyroid cell was cloned and determined to belong to the 7-transmembrane class of G-protein-coupled receptors linked to phospholipase C.

When the calcium receptors in the parathyroid gland sense low calcium levels, they stimulate the parathyroid gland to release PTH (Hedge et al., 1987). PTH then functions to raise plasma calcium primarily by stimulating bone resorption and secondarily by enhancing renal calcium reabsorption. Further, PTH stimulates the metabolism in the kidney of vitamin D to its active hormonal form, 1,25-dihydroxyvitamin D₃ (1,25(OH)₂D₃) or 1,25-dihydroxycholecalciferol (Hedge et al., 1987). This last effect of PTH shifts calcium recovery from the skeletal reserve and the kidney in the acute situation to intestinal absorption mediated by 1,25(OH)₂D₃ in the chronic situation, thereby sparing mineralized bone (Hedge et al., 1987).

In general, xenobiotic exposures that interfere with the regulation of PTH could cause problems in controlling calcium levels. As mentioned above, lithium has been shown to increase PTH levels and in turn, increase circulating levels of calcium. Depleted uranium exposure has been associated with low PTH and consequently, abnormal levels of calcium in the urine (McDiarmid et al., 2011).

The pancreas is an organ that has both endocrine and nonendocrine functions. The acinar or exocrine portion of the pancreas is concerned primarily with the regulation of gastrointestinal function. Scattered among the pancreatic acini are the endocrine units of the pancreas, the Islets of Langerhans (Porterfield, 2001). The Islets of Langerhans comprise only 1% to 2% of the weight of the pancreas. The major physiological function of the endocrine pancreas is to serve as the primary homeostatic regulator of fuel metabolism, particularly circulating glucose. Islet cells are sensors of glucose homeostasis (maintaining balance by regulation and counterregulation) that respond to changes in their nutrient and hormonal environment. The hormones secreted by islets have major effects directly on the liver. Three major cell types within the endocrine pancreas are known to produce the hormones involved in this regulation. The most abundant cell type is the beta cell (β), the site of synthesis and secretion of insulin. Glucagon is produced by the alpha cell (α) and the delta cell (δ) is the site of somatostatin synthesis (Hadley and Levine, 2007). It is likely that a functional relationship exists between the various cell types of the islets because it is known that both glucagon and somatostatin affect insulin secretion, and that somatostatin also influences glucagon secretion.

Role of the Liver in Glucose Production

Energy for cellular metabolism can be derived from fatty acids (β oxidation) or glucose (glycolysis, TCA cycle) in the blood. The liver is the primary contributor to increasing blood glucose levels. Fig. 21-13 summarizes physiological sources of energy for cellular metabolism.

Pancreatic Hormones

Insulin

The overall effects of insulin are to stimulate anabolic processes (energy storage). Specifically, insulin functions to lower blood levels of glucose, fatty acids, and amino acids and to promote their conversion to the storage form of each: glycogen, triglycerides, and protein, respectively. A number of factors affect the rate of secretion of insulin (Hedge et al., 1987). The most powerful physiological stimulus is increased circulating blood glucose. In addition, an increase in the concentration of amino acids (especially arginine and leucine) and ketone bodies in blood also increase the rate of secretion of insulin. Glucagon and the gastrointestinal peptides gastrin, secretin, gastric inhibitory polypeptide also stimulate release of insulin. Conversely, insulin secretion is inhibited by hypoglycemia, epinephrine, and norepinephrine via α₂-adrenergic receptors (inhibition of cAMP production) and somatostatin.

The variety of physiological responses to insulin include (a) increased cellular glucose uptake (in most tissues), (b) lower blood glucose levels, (c) stimulated glycogen synthesis (liver, muscle), (d) stimulated glycerol production (adipose tissue), (e) increased amino acid uptake (liver, muscle), (f) inhibited lipolysis (adipose tissue), and (g) stimulated protein synthesis (replication, transcription, and translation), a mitogenic response. As regards the pathophysiology of insulin, hypersecretion produces hypoglycemia and hyposecretion produces diabetes mellitus.

Glucagon

Glucagon is the primary hormone with action counterregulatory to insulin, because it stimulates catabolic processes (energy mobilization) to prevent hypoglycemia (Hedge et al., 1987). The most powerful physiological stimulus of secretion of glucagon is reduced circulating blood glucose. Thus, as blood glucose levels fall (hypoglycemia), glucagon secretion increases in an attempt to restore normal homeostasis. Conversely, an increase in blood glucose levels inhibits glucagon secretion. In addition to circulating levels of glucose, glucagon secretion is regulated by other factors. The release of glucagon is stimulated by epinephrine and norepinephrine (via β-adrenergic receptors; stimulation of cAMP production), and by the amino acids, arginine, leucine, and alanine (unless accompanied by glucose ingestion). Conversely, glucagon secretion is inhibited by insulin, and somatostatin.

The physiological responses to glucagon occur mostly in the liver with a stimulation of glycogenolysis, gluconeogenesis (conversion of amino acids and glycerol to glucose), lipolysis, and ketogenesis (over a long time). Additionally, glucagon stimulates the secretion of insulin in pancreatic β cells. Due to the ability of other counterregulatory hormones (epinephrine, growth hormone, and cortisol) to compensate for a deficiency of glucagon, there are no significant pathological conditions associated with abnormal glucagon secretion.

Somatostatin

Somatostatin was first isolated from the hypothalamus; its role in regulation of neuroendocrine function is to inhibit secretion of growth hormone in the anterior pituitary. After its identification in hypothalamic tissue, somatostatin was found in other cells of the brain, in various parts of the gastrointestinal tract, and in the δ cells of the pancreas. The generalized function of somatostatin appears to be as a hormone release inhibitor. Its physiological role within the pancreas is unknown; however, it inhibits secretion of insulin and glucagon (paracrine effect), and inhibits its own secretion (autocrine effect).

Interactions of Release

In addition to individual regulation of secretion of insulin and glucagon, the interactions of their release as relates to overall nutrient homeostasis of the individual must be considered. Although glucagon and insulin exert opposing effects on carbohydrate metabolism, they act in concert to preserve normoglycemia in the face of any perturbations that might tend to elevate or lower blood glucose (Hedge et al., 1987). Many substances that influence insulin secretion also affect glucagon secretion but usually in the opposite direction. Fig. 21-14 shows how insulin and glucagon interact to maintain glucose homeostasis (normoglycemia).

Insulin and glucagon exert opposing effects on various metabolic processes (Table 21-4). Therefore, many investigators like to think of the insulin-to-glucagon ratio in blood as an important determinant of the overall metabolic status. Thus, when there is a high ratio of insulin to glucagon, the effects of insulin dominate, producing a relatively anabolic state. When the ratio of insulin to glucagon is low, a catabolic state exists.

Metabolic Responses in Diabetes

Diabetes mellitus is the result of inadequate insulin action. It is usually manifest in one of two forms. Type 1 (insulin-dependent) results from autoimmune-based destruction of pancreatic β cells. Type 2 involves end organ insensitivity or resistance to insulin (non-insulin-dependent). In type 2 diabetes, insulin levels also eventually drop due to extended stress placed on pancreatic β cells. As a result of the insufficient physiological insulin action, reduced glucose removal from plasma causes hyperglycemia and a variety of metabolic alterations result. The net effect is that of increased counter-regulatory hormone dominance. These effects are summarized in Fig. 21-15. The major clinical effects in poorly controlled diabetes usually relate to progressive deterioration of function in a variety of tissues. Morbidity and mortality resulting from this disease make diabetes among the most costly of chronic diseases (Fischer, 2010). There is a genetic component associated with diabetes; however, studies in identical twins have suggested there is also most likely an environmental component to the disease. Diet, viral disease, and exogenous chemical substances are thought to trigger a genetic predisposition to its development (Fischer, 2010).

Fig. 21-16 summarizes the clinical consequences of diabetes mellitus in which the liver becomes an overproducer of glucose resulting in glucose excretion from the kidneys. This results not only from lack of insulin action, but increased glucagon secretion (which is insulin-dependent in pancreatic α cells).

Pancreatic Toxicity

The insulin-secreting β cells are particularly sensitive to chemical attack, compared with the glucagon-secreting α and somatostatin-secreting δ cells (Malaisse et al., 1982). Further, the clinical consequences of insulin deficiency are physiologically more severe than those that would result from glucagon deficiency because the other counterregulatory hormones that oppose insulin action can compensate for reduced glucagon regulation. For those reasons, there is relatively little information about chemicals that affect α or δ cells, thus, this section will focus mainly on β-cell effects. Two chemicals that have been widely used to generate animal models of diabetes are alloxan and streptozotocin (STZ). Both of these selectively destroy pancreatic β cells, thereby causing insulin insufficiency (Scarpelli, 1989; Pisarev et al., 2009; Fischer, 2010; Adeghate, 2010). The structures of these chemicals are very different. Alloxan is a cyclic urea derivative, whereas STZ is an N-methylnitrosourea containing a deoxyglucose moiety. Each of these, however, can produce β cell destruction following a single intraperitoneal injection.

The mechanism by which alloxan is thought to have cytotoxic effects involves the generation of reactive species of oxygen, particularly the hydroxyl free radical (OH^•). Isolated intact pancreatic islets synthesize and secrete insulin in response to high glucose (Fischer, 2010). A short exposure of isolated pancreatic islets to alloxan eliminates glucose-stimulated insulin secretion. There is support for this mechanism because this effect can be completely inhibited by pretreatment of islets with superoxide dismutase (SOD), catalase, diethylenetriaminepentaacetic acid, or dimethylurea (Fischer and Hamburger, 1980a, b). Furthermore, pancreatic islets have been shown to be deficient in the important protective factor, glutathione peroxidase (Malaisse et al., 1982). This probably contributes to the selectivity of alloxan for pancreatic β cells.

Two separate mechanisms are thought to be involved in the effects of STZ on β cells (Fischer, 2010). A single high dose of STZ causes generalized cytotoxicity and cell death, whereas low, multiple doses cause nonlethal cell damage, which results in an immune-based inflammatory response leading to cell death. High dose exposure to STZ causes alkylation of proteins and toxicity in β cells, where it decomposes to release nitric oxide. Thus, alkylation and nitric oxide-mediated injury could play a role in STZ-induced injury. Additionally, high-dose exposure to STZ causes a decrease in cellular levels of nicotinamide adenine dinucleotide (NAD). Part of the selectivity of STZ for pancreatic β cells is thought to be due to its glucose moiety, which is taken up via the glucose transporter GLUT-2 (Schnedl et al., 1994).

A common target of alloxan and STZ in pancreatic β cells is DNA (Uchigata et al., 1982). There are data to support that DNA damage occurs, poly(ADP-ribose) synthetase is activated, polyadenylation increases, and NAD declines. Pretreatment of rat pancreatic islets with the poly(adenosine diphosphate ribose) inhibitor nicotinamide prior to incubation with alloxan or STZ protects them against inhibition of proinsulin synthesis. However, in rats given these combined treatments, several months later the animals exhibit islet cell tumors (Yamagami et al., 1985).

Multiple exposure of animals to low doses of STZ demonstrates evidence of an immune system component (Fischer, 2010). There is a progressive series of changes involving infiltrating lymphocytes and macrophages in the islets, which produce inflammatory destruction of β cells 10 to 15 days after the initiation of treatment (Like et al., 1978).

Daily oral administration of the antihistamine–antiserotonin drug cyproheptadine (CPH) to rats causes progressive ultrastructural changes in the endoplasmic reticulum of β cells within two to three days of treatment (Fischer, 2010). This is accompanied by loss of insulin secretory granules. This is followed by swelling and vesiculation of the ER. The loss of pancreatic insulin appears to be due to a direct effect of CPH on insulin biosynthesis (Miller et al., 1993). However, CPH also inhibits insulin secretion from β cells. This has been attributed to an effect of CPH on calcium influx into β cells.

CPH can be converted in rats to an epoxide metabolite desmethylcyproheptadine-10,11 (DMCPH)-epoxide. This is a major metabolite of CPH found in urine and pancreatic tissue (Fischer, 2010). Decreasing the production of this metabolite using inhibitors of drug metabolism reduces pancreatic insulin depletion caused by CPH (Chow et al., 1988). Thus, DMCPH is likely the most toxic form of the chemical.

There appear to be big species differences in β-cell toxicity. Human β-cell destruction has been demonstrated with the rodenticide Vacor (Prosser and Karam, 1978) and during treatment with the drug pentamidine (Bouchard et al., 1982). Whereas humans develop Vacor-induced diabetes, no sensitive animal has been found. Conversely, unlike in rats, STZ, when used as chemotherapy, has not caused diabetes in humans (Fischer 2010).

Insulin Resistance

Insulin resistance and defective function of pancreatic β cells usually occur sometime before the development of type 2 diabetes. Glucocorticoid treatment induces insulin resistance and enhances insulin secretion in rodents and humans (Rafacho et al., 2010). In that study, dexamethasone treatment of rats for five days resulted in hyperinsulinemia at the end of dosing, whereas 10 days following cessation of dexamethasone treatment hyperinsulinemia and insulin resistance had resolved. This provides important information as regards the therapeutic use of glucocorticoids in humans.

In a study investigating nondiabetic residents living near a deserted pentachlorophenol and chloralkali factory in Taiwan, insulin resistance was associated with increased circulating levels of dioxins and mercury (Chang et al., 2011). Furthermore, increased levels of dioxins and mercury combined were associated with even greater insulin resistance. Thus, simultaneous exposure to dioxins and mercury appears to enhance the risk of insulin resistance.

BPA, an endocrine-disruptor, is a chemical compound in the manufacture of polycarbonate plastics that is widespread in the environment. BPA exposure of pregnant mice resulted in increased insulin, leptin, triglyceride, and glycerol levels and greater insulin resistance in mothers four months postpartum (Alonso-Magdalena et al., 2010). Further, male offspring of the exposed mothers, at six months of age, demonstrated reduced glucose tolerance and increased insulin resistance. Thus, BPA may represent a risk factor for diabetes in exposed humans.

In Vitro Testing

Several cell lines are available for testing of insulin secretion. Pancreatic β-cell-derived RINm5F cells were exposed to a combination of the cytokines, IL-1β, TNF-α, and IFNγ to simulate type 1 diabetes mellitus conditions (Gurgul-Convey et al., 2011). This study showed that hydrogen peroxide produced by these cytokines reacted in the presence of trace metal Fe⁺⁺ with nitric oxide to form highly toxic hydroxyl radicals. The authors concluded that proinflammatory cytokine-mediated β-cell death is due to nitro-oxidative stress-mediated hydroxyl radical formation in the mitochondria.

RINm5F cells were also used to investigate the role of oxidative stress in inorganic arsenic (iAs) exposure (Lu et al., 2011). A number of proapoptotic mitochondrial and cytosolic markers were investigated and found to be elevated during β-cell toxicity. The iAs-induced apoptosis and its cellular signaling events could be reversed by the antioxidant N-acetylcysteine. Therefore, it was concluded that iAs-induced oxidative stress causes apoptosis in β cells, by the mitochondria-dependent and ER stress-triggered signaling pathways. Rat-derived INS-1(832/13) cells that secrete insulin in response to glucose were also studied for the effects of exposure to iAs (Fu et al., 2010). Reactive oxygen species derived from glucose metabolism provide a metabolic signal for glucose-stimulated insulin secretion from pancreatic β cells. Exposure of these cells to low levels of iAs increased an antioxidant-mediated response, and dampened glucose-stimulated insulin secretion. These findings support that low levels of arsenic can induce a cellular adaptive oxidative stress response, and disrupt β-cell function. The apoptotic as opposed to oxidative stress effects of iAs between these two studies seem to be disparate. However, two different cell lines were used, and the former study incubated the cells with higher levels of iAs (2, 5 μM), compared with the latter (0.05–0.5 μM). It is possible that oxidative stress induced at the higher levels of iAs exposure (causing cell death) is sufficient to overwhelm that which, at lower levels, plays a regulatory role in insulin secretion (inhibition). This observation suggests that the nature of the cellular response depends on the level of exposure.