Chapter 19: Toxic Responses of the Skin

As the body's first line of defense against external insult, the skin's enormous surface area (1.5-2 m²) is exposed routinely to chemicals and may inadvertently serve as a portal of entry for topical contactants. Recognizing the potential hazards of skin exposure, the National Institute of Occupational Safety and Health (NIOSH) characterized skin disease as one of the most pervasive occupational health problems in the United States. In 1982, NIOSH placed skin disease in the top 10 leading work-related diseases based on frequency, severity, and the potential for prevention. Data from the Bureau of Labor Statistics indicate that in 2004 skin disease attributed to workplace exposures accounted for nearly 16% of reported nonfatal occupational disease in private industry; incidence data indicate a rate of 4.4 cases per 10,000 or nearly 39,000 new cases per year. On that basis, the annual economic burden of dermatitis in workers in the USA has been estimated as ≈$1.2 billion (Blanciforti, 2010). Substantial reduction in the reported incidence has occurred in recent years thanks to workplace cleanup and better personal protective equipment. Nevertheless, improvements in prevention and management are needed for continued progress (Emmett, 2003). Skin conditions resulting from exposures to consumer products or occupational illnesses not resulting in lost work time are poorly recorded and tracked. Hence, the incidence of such skin diseases appears greatly underestimated.

NIOSH recently changed its skin notations to reflect more accurately the varied risks to workers after topical exposure (see DHHS(NIOSH) Publication No. 2009-147, www.cdc.gov/niosh). Previously, inhalation exposures were thought to produce the most risk to workers, with skin exposure being only a secondary exposure pathway. The new guidelines now reflect increased understanding of several pathways by which skin exposure can lead to disease: systemic toxicity via skin absorption (SYS), direct effects that damage the skin (DIR), and immune-mediated responses to chemicals that contact the skin (SEN). The new notations utilize these categories to designate hazard-specific skin notations, and have incorporated standard criteria to ensure consistency in how chemicals are designated. Determining the hazard potential of a particular chemical is based on physicochemical properties of the substance, toxicokinetic studies, epidemiological data, in vitro or in vivo laboratory testing, and in silico computational predictions. These criteria are specified in the chemical designation.

A large and highly accessible human organ, the skin protects the body against external insults to maintain internal homeostasis. Its biological sophistication allows it to perform a myriad of functions above and beyond that of a suit of armor. Physiologically, the skin participates directly in thermal, electrolyte, hormonal, metabolic, antimicrobial and immune regulation, without which a human would perish. Rather than merely repelling noxious physical agents, the skin may react to them with a variety of defensive mechanisms that prevent internal or widespread cutaneous damage. If an insult is severe or sufficiently intense to overwhelm the protective function of the skin, acute or chronic injury becomes readily manifest in various ways. The specific presentation depends on a variety of intrinsic and extrinsic factors including body site, duration of exposure, and other environmental conditions (Table 19-1).

Skin Histology

The skin consists of two major components—the outer epidermis and the underlying dermis, which are separated by a basement membrane (Fig. 19-1). The junction ordinarily is not flat but has an undulating appearance (rete ridges). In addition, epidermal appendages (hair follicles, sebaceous glands, and eccrine glands) span the epidermis and are embedded in the dermis. In thickness, the dermis comprises approximately 90% of the skin and has largely a supportive function. It has a high content of collagen and elastin secreted by scattered fibroblasts, thus providing the skin with elastic properties. Separating the dermis from underlying tissues is a layer of adipocytes, whose accumulation of fat has a cushioning action. The blood supply to the epidermis originates in the capillaries located in the rete ridges at the dermal–epidermal junction. Capillaries also supply the bulbs of the hair follicles and the secretory cells of the eccrine (sweat) glands. The ducts from these glands carry a dilute salt solution to the surface of the skin, where its evaporation provides cooling.

The interfollicular epidermis is a stratified squamous epithelium consisting primarily of keratinocytes. These cells are tightly attached to each other by desmosomes and to the basement membrane by hemidesmosomes. Melanocytes are interspersed among the basal cells and distributed in the papilla of hair follicles. In the epidermis, these cells are stimulated by ultraviolet light to produce melanin granules. The granules are extruded and taken up by the surrounding keratinocytes, which thereby become pigmented. Migrating through the epidermis are numerous Langerhans cells, which are important participants in the immune response of skin to foreign agents.

Keratinocytes of the basal layer comprise the germinative compartment. When a basal cell divides, one of the progeny detaches from the basal lamina and migrates outward. As cells move toward the skin surface, they undergo a remarkable program of terminal differentiation. They gradually express new protein markers and accumulate keratin proteins, from which the name of this cell type is derived. The keratins form insoluble intermediate filaments accounting for nearly 40% of the total cell protein in the spinous layer. At the granular layer, the cells undergo a striking morphological transformation, becoming flattened and increasing in volume nearly 40-fold. Lipid granules fuse with the plasma membrane at the granular layer/stratum corneum interface, filling the intercellular spaces of the stratum corneum with lipid, as opposed to the aqueous intercellular solution in the viable epidermis. This lipid performs a variety of defensive functions, both in preventing diffusion of water and ions out of the body, and preventing access of toxins and bacteria into the body. It thus constitutes the major barrier to toxin entry into the skin (Elias, 2005). Meanwhile, the plasma membranes of these cells become permeable, resulting in the loss of their reducing environment and consequently in extensive disulfide bonding among keratin proteins. Cell organelles are degraded, while a protein envelope is synthesized immediately beneath the plasma membrane. The membrane is altered characteristically by the loss of phospholipid and the addition of sphingolipid.

This program of terminal differentiation, beginning as keratinocytes leave the basal layer, produces the outermost layer of the skin, the stratum corneum. No longer viable, the mature cells (called corneocytes) are approximately 80% keratin in content. They are gradually shed from the surface and replaced from beneath. The process typically takes two weeks for basal cells to reach the stratum corneum and another two weeks to be shed from the surface. In the skin disease psoriasis, the migration of cells to the surface is nearly 10-fold faster than normal, resulting in a stratum corneum populated by cells that are not completely mature. In instances in which the outer layer is deficient due to disease or physical or chemical trauma, the barrier to the environment that the skin provides is inferior to that provided by normal, healthy skin.

Percutaneous Absorption

Until the turn of the century, the skin was believed to provide an impervious barrier to exogenous substances. Gradually, the ability of substances to penetrate the skin, though this process generally is very slow, became appreciated. During the past 50 years, the stratum corneum has been recognized as the primary barrier (Scheuplein and Blank, 1971). Diseases (eg, psoriasis) or other conditions (eg, abrasion, wounding) in which this barrier is compromised can permit greatly increased uptake of poorly permeable substances, as does removal of the stratum corneum by tape stripping or organic solvents such as acetone. The viable layer of epidermis provides a much less effective barrier, because hydrophilic chemicals readily diffuse into the intercellular water, while hydrophobic agents can partition into cell membranes, and each can diffuse readily to the blood supply in the rete ridges of the dermis.

Probably the best-known biological membrane barrier for this purpose, the stratum corneum prevents water loss from underlying tissues by evaporation. Its hydrophobic character reflects the lipid content of the intercellular space, approximately 15% of the total volume (Elias, 1992). The lipids, a major component being sphingolipids, have a high content of long-chain ceramides, removal of which seriously compromises barrier function as measured by transepidermal water loss. The stratum corneum ordinarily is hydrated (typically 20% water), the moisture residing in corneocyte protein, but it can take up a great deal more water upon prolonged immersion, thereby reducing the effectiveness of the barrier to chemicals with a hydrophilic character. Indeed, occlusion of the skin with plastic wrap, permitting the retention of perspiration underneath, is a commonly employed technique to enhance uptake of chemicals applied to the skin surface. Although penetration from the air is generally too low to be of concern, protection from skin uptake may be advisable for some compounds (eg, nitrobenzene) at concentrations high enough to require respirator use.

Finding the rate at which the uptake of agents through the skin occurs is important for estimating the consequences of exposure to many agents we encounter in the environment. Indeed, a regulatory strategy permitting bathing in water considered barely unfit for drinking was revised when it was realized that exposure from dermal/inhalation uptake during bathing could be comparable to that from drinking 2 L of the water (Brown et al., 1984). Uptake through the skin is now incorporated in pharmacokinetic modeling to estimate potential risks from exposures. The degree of uptake depends upon the details of exposure conditions, being proportional to solute concentration (assuming it is dilute), time, and the amount of skin surface exposed. In addition, two intrinsic factors contribute to the absorption rate of a given compound: its hydrophobicity, which affects its ability to partition into epidermal lipid, and its rate of diffusion through this barrier. A measure of the first property is the commonly used octanol/water partitioning ratio (K_ow). This is particularly relevant for exposure to contaminated water, as occurs during bathing or swimming. However, partitioning of a chemical into the skin is greatly affected by its solubility in or adhesion to the medium in which it is applied (including soil). Similarly, very hydrophobic compounds, once in the stratum corneum, may diffuse only very slowly into less hydrophobic regions below. The second property is an inverse function of molecular weight (MW) or molecular volume. Thus, hydrophobic agents of low MW permeate the skin better than those of high MW or those that are hydrophilic. For small molecules, hydrophobicity is a dominant factor in penetration.

Although only small amounts of chemicals may penetrate the stratum corneum, those of high potency may still be very dangerous. For example, hydrophobic organophosphorus and carbamate pesticides can be neurotoxic to humans and domestic animals by skin contact. Children and adolescents harvesting tobacco are especially susceptible to poisoning by contact with nicotine, a natural pesticide present in moisture on the leaves (McKnight and Spiller, 2005). Conditions of topical treatment of livestock for pest control must take into consideration not only the tolerance of the animals but also residues in meat and milk resulting from skin penetration. High level skin exposure to chemicals considered safe at low levels can be dangerous, evident from the nervous system toxicity and deaths of babies exposed to hexachlorophene mistakenly added to talcum powder for their diapers (Martin-Bouyer et al., 1982). Previous findings of N-nitrosamines that penetrate skin well (such as N-nitrosodiethanolamine) in cutting oils and cosmetics raised concern and led to their monitoring and to reduction of exposure.

Considerable empirical information has been collected on some compounds of special interest (including pharmaceuticals, pesticides, and pollutants) for use in quantifying structure–penetration relationships. From such information, relations can be obtained for skin penetration (P_cw) using empirically derived constants (C₁, C₂, C₃) that have the form shown below (Potts and Guy, 1992).

log P_cw = C₁ − C₂(MW) + C₃ log K_ow

Such relations describe steady-state conditions, in which a chemical leaves the stratum corneum at the same rate it enters. Because rates of transfer of very hydrophobic agents into the aqueous phase of the spinous layer are slow, saturation of the stratum corneum provides a depot, leading to continued penetration into the body for relatively long time periods after external exposure to a chemical stops. Recognition that such estimates overpredict uptake from short exposures has led to modeling of limited doses (Guy, 2010a).

Diffusion through the epidermis is considerably faster at some anatomical sites than others. A list in order of decreasing permeability gives the following hierarchy: foot sole > palm > forehead > abdomen (Scheuplein and Blank, 1971). Scrotal skin reportedly has the highest permeability for some topical chemicals (Fisher, 1989). Under ordinary conditions, absorption through the epidermal appendages is generally neglected, despite the ability of chemicals to bypass the stratum corneum by this route, because the combined appendageal surface area is such a small fraction of the total available for uptake. However, because loading of the stratum corneum is slow, penetration through the appendages can constitute an appreciable fraction of the total for short exposures. In some cases, the effects of appendages can even be dominant. For instance, benzo[a] pyrene penetrates the skin of haired mice several fold faster than that of hairless strains (Kao and Hall, 1987). Recent studies indicate the hair follicles serve as important routes of entry for hydrophilic agents and pharmaceuticals as well, a property that could be exploited for therapeutic purposes (Otberg et al., 2008).

Transdermal Drug Delivery

The ability of the stratum corneum to serve as a reservoir for exogenously applied agents is well illustrated by the recent development of methods for the delivery of pharmaceuticals. Application of drugs to the skin can produce systemic effects, a phenomenon observed unintentionally before the ability of the skin to serve as a delivery system was appreciated. For example, topical exposure of young girls to estrogens has led to reports of pseudoprecocious puberty, while in young or adult males, such exposure has produced gynecomastia (Amin et al., 1998). Transfer of topically applied hormonal medication from a treated to an untreated individual with consequent effects on the latter is also observed (Busse and Maibach, 2011).

Specially designed patches are currently in use to deliver at least 17 different drugs (including estradiol, testosterone, nitroglycerin, scopolamine, clonidine, fentanyl, and nicotine) for therapeutic purposes (Guy, 2010b), and others are under development. The advantages of this approach over oral dosing include providing a steady infusion for extended periods (typically one-seven days), thereby avoiding large variations in plasma concentration, preventing exposure to the acidic pH of the stomach, and avoiding biotransformation in the gastrointestinal tract or from first-pass removal by the liver. The contrast in plasma concentration kinetics between different methods of delivery is particularly evident for chemicals that are rapidly metabolized, such as nitroglycerin, which has a half-life of minutes. A variety of chemicals, chosen carefully to minimize irritation or allergenicity, have been incorporated into pharmaceutical preparations to enhance absorption and penetration. In addition, encapsulating a drug in small vesicles of phospholipid or nonionic surfactant or coating nanoparticles can improve delivery through the epidermis and hair follicles (Desai et al., 2010). Finally, various microneedle systems have been studied to deliver drugs, vaccines, and gene therapy across the skin. These systems may prove especially important in delivering vaccines in areas where refrigeration for conventional vaccines is not available (Prausnitz et al., 2009).

Measurements of Penetration

A pharmacokinetic approach with intact subjects has been commonly employed with experimental animals. To simplify determination of penetration kinetics, skin flaps may be employed and the capillary blood flow monitored to measure penetration. For this purpose, pig skin has particular utility (Riviere and Brooks, 2005). Because penetration through rodent skin is usually faster than through human skin, the former can provide an overestimate for behavior of the latter. Without verification using human skin, such measurements are subject to large uncertainties due to species differences in density of epidermal appendages, stratum corneum properties (eg, thickness, lipid composition), and biotransformation rates. A promising variation minimizing species differences is to use skin grafts on experimental animals for these measurements. Human skin persists well on athymic mice and retains its normal barrier properties.

For risk assessment and pharmaceutical design, the most useful subject for experimentation is human skin. Volunteers are dosed, plasma and/or urine concentrations are quantified at suitable intervals, and amounts excreted from the body are estimated. Past measurements of penetration often used ¹⁴C-labeled agents. This approach is not preferred, but use of isotopic labels now is readily feasible when coupled to ultrasensitive detection by accelerator mass spectrometry (Buchholz et al., 1999). For in vitro work, excised split-thickness skin can be employed in special diffusion chambers, though care is needed to preserve the viability of the living layer of epidermis. The chemical is removed for measurement from the underside by a fluid into which it partitions, thereby permitting continued penetration. Commonly employed is a simpler setup using cadaver skin with the lower dermis removed. This lacks biotransformation capability, but retains the barrier function of the stratum corneum. Accurate testing of percutaneous absorption of poorly soluble chemicals from environmental and pharmaceutical substrates requires attention to details of particle size, component complexes, vehicle, application rate, and skin contact. A long time goal has been to predict chemical permeability from chemical structure. Comparisons of predictive models with results from transdermal patches show good correlations, despite large discrepancies in fluxes of some compounds, and emphasize the importance of experimental details (Farahmand and Maibach, 2009).

Biotransformation

The ability of the skin to metabolize chemicals that diffuse through it contributes to its barrier function. In specific activity, phase I metabolism in the skin usually is only a small fraction (≈3%) of that in the liver (Rolsted et al., 2008), but it is capable of affecting the outcome of exposure. Biotransformation influences the biological activity of xenobiotics and topically applied drugs, leading to their degradation or their activation as skin sensitizers or carcinogens (Svensson, 2009). While overall activities are normally low, they can be induced (or suppressed). In fact, a large fraction of the pharmaceuticals used in clinical dermatology are cytochrome P450 inducers, inhibitors, or substrates (Ahmad and Mukhtar, 2004). For example, cytochromes P450 1A1 and 1B1 are inducible in the epidermis to high levels by crude coal tar (Smith et al., 2006), which is used in dermatological therapy. Well-known xenobiotic inducers of these isozymes include tetrachlorodibenzo-p-dioxin (TCDD), polycyclic aromatic hydrocarbons, and coplanar polychlorinated biphenyls. Exposure to such inducers could influence skin biotransformation and even sensitize epidermal cells to other chemicals that are not good inducers themselves, a phenomenon observable in cell culture (Walsh et al., 1995).

Biotransformation of a variety of compounds in the skin has been detected, including arachidonic acid derivatives, steroids, retinoids, amines, and polycyclic aromatic hydrocarbons, suggesting that multiple P450 activities are expressed. Using sensitive technologies, it is now evident that numerous distinct isozymes are expressed at widely varying levels. Evidence for expression of >30 CYP genes has been obtained by DNA microarray and real-time polymerase chain reaction (Hu et al., 2010; Luu-The et al., 2009). A recent survey of those in the CYP1-4 families indicated that half were expressed at substantially higher levels in differentiating keratinocytes (similar to spinous cells) than in basal-like cells (Du et al., 2006). Species differences are apparent in the amounts of P450 activities detectable. For example, measured ethoxycoumarin-O-deethylase activity is 20-fold higher in mouse than human (or rat) skin. Differences of such magnitude help rationalize the observation that the rate of penetration of ethoxycoumarin is sufficient to saturate its metabolism in some species such as human but not in others such as mouse or guinea pig (Storm et al., 1990). Beyond the cytochromes P450, other phase I enzymes expressed in the skin include flavin-dependent monooxygenases, aldehyde dehydrogenases, carboxylesterases, and glutathione peroxidases (Hu et al., 2010). To the extent that phase I and II metabolism influences sensitization to exogenous chemicals (Svensson, 2009), they may also help rationalize species differences in allergic response.

Enzymes participating in phase II metabolism are expressed in skin. Several isozymes each of UDP glucuronosyltransferase, sulfotransferase, and glutathione S-transferase have been detected in human (Hu et al., 2010; Luu-The et al., 2009) and rodent skin (Oesch et al., 2007). In general, this activity occurs at a lower rate than observed in the liver, but exceptions are evident, as in the case of quinone reductase (Khan et al., 1987). Different species express different relative amounts of the various isozymes, which could alter resulting target specificities or degrees of responsiveness. Glutathione S-transferase, for instance, catalyzes the reaction of glutathione with exogenous electrophiles or provides intracellular transport of bound compounds in the absence of a reaction. It also facilitates the reaction of glutathione with endogenous products of arachidonate lipoxygenation (leukotrienes) to yield mediators of anaphylaxis and chemotaxis, which are elements of the inflammatory response in skin. Of the first three major transferase forms characterized in liver, the dominant form in skin of humans and rodents is the P isozyme. A comparison of human with rodent skin indicates the former has higher glutathione S-transferase activity but lower glutathione content, suggesting that it may be less susceptible to low doses of electrophilic substrates, whereas rodent skin could be less susceptible at higher doses when glutathione is depleted in human but not mouse skin (Jewell et al., 2000).

A variety of other metabolic enzyme activities have also been detected in human epidermal cells, including sulfatases, β-glucuronidase, N-acetyl transferases, esterases, and reductases (Hu et al., 2010; Luu-The et al., 2009). In addition, the intercellular region of the stratum corneum has catabolic activities (eg, protease, lipase, glycosidase, and phosphatase) supplied by the lamellar bodies along with their characteristic lipid (Elias, 1992). The epidermis and pilosebaceous units are the most important drug targets and sites of toxic effects sites and, indeed, are the major sources of biotransformation activity in the skin. However, other cell types such as fibroblasts are known to participate in biotransformation, helping to rationalize observations that organotypic cell cultures containing underlying fibroblasts resemble natural skin in biotransformation better than those without fibroblasts (Gibbs et al., 2007; Hu et al., 2010; Luu-The et al., 2009). Such cultures, where keratinocytes float on collagen gels at the air–liquid interface, are better skin mimics than conventional ones using established keratinocyte lines. Their development has stimulated aspirations that these in vitro models can replace much animal testing of cosmetics and potential skin toxicants.

The report that cholesterol sulfotransferase is regulated by ligands of LXR and PPAR illustrates the potential for exogenous chemicals, including pharmaceuticals, to influence such activities and thereby the barrier function of the skin (Jiang et al., 2005), where penetration can be influenced by cell differentiation and biotransformation. The influence of hydroxysteroid dehydrogenases and microsomal reductase activities during percutaneous absorption is evident in studies on skin in organ culture. Biotransformation can activate prodrugs, an example being minoxidil, applied to prevent hair loss, which undergoes sulfation locally to its active form. To permit increased penetration of pharmaceuticals through the lipid barrier of the stratum corneum, prodrug design strategies have been developed that depend on epidermal esterase activity. Thus, hydrophobic esters of pharmaceuticals targeting the skin (eg, acne) could be applied to the skin surface, de-esterified during transit, and yield slow release of the active forms (Fang and Leu, 2006).

In the occupational arena, where records are compiled on large workforces, contact dermatitis is by far the largest category (≈90%) of compensated skin disease. Using eczema of the hand as a sentinel condition, since 80% of the total reported dermatitis occurs at that location (10% on the face), reveals a prevalence of 7% to 10% among workers. Attributed to better diagnosis, more accurate identification of offending chemicals, and more effective prevention and worker education, the fraction of afflicted workers recovering without impairment has improved nearly to 80% with proper management (Belsito, 2005). However, while certain conditions carry a favorable prognosis, others (eg, chronic cumulative irritant contact dermatitis or contact allergy to nickel, chromate, formaldehyde, or rubber) frequently result in chronic disease in which changing jobs is of limited or no benefit (Belsito, 2005; Emmett, 2003). Overall, contact dermatitis falls into the two major categories of irritant and allergic forms that share important features (Watkins and Maibach, 2009). Both involve inflammatory processes and can have indistinguishable clinical characteristics of erythema (redness), induration (thickening and firmness), scaling (flaking), and vesiculation (blistering) in areas of direct contact with the chemical. Biopsies from affected sites reveal a mixed-cell inflammatory infiltrate of lymphocytes and eosinophils and spongiosis (intercellular edema), but are insufficient to distinguish the two conditions from each other or from certain other common syndromes. The two can coexist. For example, since wet cement is alkaline and often contains chromates (commonly allergenic), chronic exposure can produce a composite response.

Irritant Dermatitis

Accounting for nearly 80% of contact dermatitis cases, this condition arises from the direct action of agents on the skin. A chemical in this category is anticipated to give an adverse reaction to anyone if the concentration is high enough and the exposure time long enough. Certain chemicals at sufficient concentration produce an acute irritation, sometimes called a second-degree chemical burn, that can even result in scarring in serious cases. These include strong acids, alkalies, and powerful oxidizing and reducing agents that substantially disrupt the cornified layer, producing cytotoxicity directly. Contact with a variety of plants can also have irritant effects (Modi et al., 2009). Exposure stimulates release of proinflammatory cytokines (IL1-α, IL1-β, and TNF-α) from keratinocytes. More common is chronic cumulative irritation from repeated exposures to mild irritants such as soaps, detergents, solvents and cutting oils. An example of eczema from cutting oil is shown in Fig. 19-2A. The chronic friction and production of small scale trauma can wear away the lipid barrier of the stratum corneum, leading gradually to further damage (loss of cohesion, desquamation) that facilitates penetration of exogenous chemicals and may be detectable as increased transepidermal water loss. In some cases, epidermal thickening occurs without much inflammation. Chronic exposure in the occupational setting often elicits a process of “hardening,” where the irritant response resolves. Of uncertain mechanism, it exhibits thicker granular layer and stratum corneum, increased permeability to irritants, altered expression of inflammatory mediators, and a parallel hardening of unexposed skin (Watkins and Maibach, 2009). In any case, increased penetrance at sites of barrier damage can facilitate exposure to chemicals that elicit subsequent toxic effects.

The skin at some anatomic sites is more sensitive than at other sites. Eyelids have a thin epidermis and are quite sensitive, for example, and the back is more sensitive than the forearm or the scalp of individuals with male pattern baldness (Zhai et al., 2004). Individuals vary greatly in sensitivity to irritant dermatitis. Fig. 19-2B shows an irritant reaction on the inside of the elbow on an atopic person. The incidence of atopic dermatitis is increasing rapidly in industrialized societies, and now comprises up to 20% of the pediatric population in these societies (Elias and Schmuth, 2009). Atopic individuals are the most sensitive to irritants and exhibit a propensity to produce specific IgE antibodies to allergens and typically suffer from hay fever. These individuals usually have a poorer prognosis than nonatopics and have a higher frequency of persistent dermatitis. The best preventive measure for atopics and others is to avoid exposure to contact irritants, but in practice this strategy is difficult to implement. That atopic dermatitis has a strong genetic component led to the realization that defects in the intermediate filament aggregating protein filaggrin are strongly associated with this condition and associated development of asthma (Palmer et al., 2006). That loss of filaggrin also causes ichthyosis vulgaris, a scaly skin condition with defective epidermal barrier function (Gruber et al., 2011), has raised the possibility that contributions from other genetic defects in barrier function may be revealed.

Information on the irritancy of chemicals toward human skin may be obtained as part of differential diagnosis by patch testing for allergic response. The skin of laboratory animals (mice, rats, rabbits, and guinea pigs) can be used for testing, but it is thinner and more sensitive than human skin to irritants. For development of new pharmaceuticals, cosmetics, and other consumer products, a great need exists for an in vitro system to determine the potential for irritant responses. Use of human epidermal cell cultures has been increasing as reconstructed epidermal and skin models come closer to the native differentiated state. For example, a recent study compared 50 chemicals for which data on 30 are available from patch testing (Tornier et al., 2006). The tests provided useful comparative data and, judging by viability (mitochondrial function), histology, and release of inflammatory mediators (IL-1α), suggested a parallel with natural skin. Such models offer advantages in convenience and cost, appear applicable to phototoxicity as well, and are more uniform in response than skin in the human population. Analyses of cellular responses by DNA microarray and proteomic methods are being explored. Although advanced, the state of maturation in such culture models is not complete, as seen by histology and barrier function, resulting in their greater sensitivity than the skin (Netzlaff et al., 2005). In addition, extrapolation of the models to cumulative insult dermatitis presents a challenge.

Chemical Burns

A chemical that is extremely corrosive can produce immediate coagulative necrosis resulting in considerable tissue damage with ulceration and sloughing. Sometimes referred to as a third degree chemical burn, the damage does not have a primary inflammatory component and thus may not be classified as an irritant reaction. Examples of burns from ethylene oxide and alkali are shown in Fig. 19-2C and D. If the chemical is not quickly and completely removed, damage to the skin may continue and, with increased access to the circulation, systemic injury can occur. Table 19-2 lists some important corrosive agents giving chemical burns in the occupational arena. Certain chemical warfare agents first used in combat a century ago, such as bis-(2-chloroethyl)sulfide (sulfur mustard) or 2-chlorovinyl dichloroarsine (Lewisite), are potent vesicants upon skin contact and produce considerable damage when inhaled. Sulfur mustard was used in numerous conflicts during the last century (Ghabili et al., 2011), and the threat of use remains (Shakarjian et al., 2010). A more immediate threat is chemical assault on individuals, a worldwide problem resulting in serious disfigurement and visual impairment (Milton et al., 2010), especially in the developing world (Karunadasa et al., 2010). Acids, typically those readily available (sulfuric, nitric, sometimes formic), are the most common agents, and the face is the most common target.

Allergic Contact Dermatitis

Allergic contact dermatitis is a delayed (T-cell mediated) hypersensitive reaction. To induce sensitization through the skin, chemical haptens generally penetrate the lipid barrier and, to be detected by the immune system, become attached to carrier proteins. The complete antigens are then processed by Langerhans cells (resident macrophages) and displayed on their surfaces with major histocompatibility complex II molecules. The Langerhans cells present the processed peptides to T helper type 1 cells in regional lymph nodes, thereby stimulating interleukin release and proliferation of the sensitive T helper cells. Over a one- to three-week period, memory T cells are thus generated and enter the circulation. Upon subsequent exposure to a specific antigen previously encountered, allergen presentation by the Langerhans cells results in a much greater response due to homing by the memory cells to the skin, their clonal proliferation and their release of cytokines chemotactic for inflammatory cells and stimulatory for their further production. Because this process takes time, the characteristic dermal infiltration and spongiosis result after a delay (latent period) of a half to several days (Mark and Slavin, 2006).

Thousands of chemicals have been reported to give rise to allergic contact dermatitis, many across a variety of occupations and consumer products. Table 19-3 lists some common contact allergens, two of which among others are shown in Fig. 19-3. Because most chemicals in the chemical universe are only weakly active or infrequently encountered, much effort has focused on finding the major allergens in the population by systematic patch testing of dermatology patients. Although not measuring sensitivity in the population at large, the results are quite useful. The panel of chemicals tested can vary with geographic location to accommodate local usage, or it can be directed to specific anatomic sites such as the foot (Holden and Gawkrodger, 2005). Panels also are adapted to emerging trends as new products appear and others decline in use. Table 19-4 lists the 21 agents most commonly giving positive reactions in adult subjects (ages 19–64) in recent patch tests conducted by the North American Contact Dermatitis Group (Warshaw et al., 2012; Yoo et al., 2010). The testing showed that individuals older than 64 exhibited rates similar to those in adults, both higher than those in children. Most such surveys reveal two-thirds of the subjects to be sensitive to at least one agent in the panel (Yoo et al., 2010). Increasing prevalence of reaction to nickel among younger subjects may reflect increasing exposure, including through body piercing (Schram et al., 2010). A number of agents (nickel, dichromate, p-phenylenediamine, and formaldehyde) have shown high prevalences of reactivity for several decades, while others, once thought innocuous, have more recently become recognized as reactive. For example, individuals exhibit sensitization to gold jewelry, dental gold, and gold coronary stents (Moller, 2010). The high prevalence of allergic contact dermatitis to metals has raised concerns about possible reactions to surgical implants containing them (Basko-Plluska et al., 2011), especially nickel (Schram et al., 2010). An example of contact allergy to dichromate in cement is shown in Fig. 19-2E.

Sensitization by ingredients in cosmetics is a common problem, and one that changes as the formulations evolve (Pascoe et al., 2010). As with other consumer products, reduction in use of the most prevalent allergenic chemicals and their replacement by less allergenic substitutes are advocated. Caution in using less characterized chemicals as replacements must be exercised, however, since their allergenicity may not become evident until they reach large populations of users, as has happened in several prominent cases (Uter et al., 2005). For instance, methylchloroisothiazolinone/methylisothiazolinone, used in cosmetics, was replaced with the biocide methyldibromo glutaronitrile, which did not cause allergic contact dermatitis in initial screens. Upon widespread use, however, the latter was shown to be a potent contact allergen (Kynemund Pedersen et al., 2004). Paradoxically, several chemicals that are reliably contact sensitizing are used therapeutically for papillomavirus-induced warts, skin cancer, and alopecia areata (Holzer et al., 2006).

Unlike contact irritants, where the response is generally proportional to the applied dose and time, contact allergens can elicit reactions at very small doses. Nevertheless, careful analysis from human and animal testing (Arts et al., 2006; Boukhman and Maibach, 2001) shows that a higher dose confers a greater likelihood of sensitization and that doses below a threshold for sensitization can have a cumulative effect. In addition, the dose required to elicit a reaction is lower after sensitization with a higher dose. Moreover, the dose dependence for sensitization displays nonlinearity, suggesting that the response of individual dendritic cells is sublinear, probably sigmoidal. Thus, more stimulation can produce a more than proportionally larger response, although at high doses saturation and sometimes even inhibition of the response become evident. This result emphasizes the importance of minimizing individual exposures. The findings also reveal a wide variation in human response to sensitization, which appears to have at least in part a genetic basis.

Diagnosis and Testing

When a patient exhibits allergic contact dermatitis, finding the responsible agent is important to avoid continued exposure. For this purpose, patch testing is commonly employed by procedures refined over many years of practice since it was first employed a century ago (Mark and Slavin, 2006). On the washed backs of patients, who are not currently exhibiting contact dermatitis or using corticosteroids or other immunosuppressives, are placed patches each containing a small amount of a potential allergen. Conveniently, many of the materials are commercially available at standardized concentrations too low to produce irritant reactions. Certain chemicals normally are not tested because they induce too strong a response (urushiol from poison ivy) or might produce sensitization (beryllium). After two to three days, during which time a maximal reaction usually develops, the patches are removed and sites of exposure are scored for degree of response. Relevance to the patient's actual environment must be considered so that exposure in daily life can be minimized to appropriate chemicals. Interpretation of the results and environmental modification should take into account the phenomenon of cross-sensitivity, where reactivity to a compound may be evident if it shares functional groups that have provoked sensitization in another compound. Fig. 19-4 illustrates the principle with three amine compounds, and Table 19-5 lists some common cross-reacting chemicals.

Animal testing to predict allergenicity has an extended history (Ngo and Maibach, 2010). A chemical is applied to intact or abraded skin or through intradermal injection with or without adjuvant to enhance sensitization. The reaction of the skin to subsequent challenge with the chemical is then observed and graded. This approach has successfully identified some strong sensitizers relevant to human exposures, but detection of weak sensitizers on a large scale is hampered by the usual difficulties in animal testing, including small animal numbers and limited experiment time to reduce expense. In addition, extrapolation of sensitivity measurements from laboratory animals to humans presents large uncertainties. Nevertheless, the local lymph node assay performed in mice has gained attention as a way to measure the pool of sensitized T cells by their proliferation in draining lymph nodes, illustrated by a comparison of potencies of Disperse Blue 106 and 2,4-dinitrochlorobenzene (Betts et al., 2005). Because sensitizers differ in potency by at least four orders of magnitude, a quantitative assay has a distinct advantage.

Increasing emphasis on reducing or eliminating animal use in toxicity testing, driven in part by regulatory initiatives, has stimulated development of integrated testing strategies, where predictions of toxic effects such as skin sensitization include physical chemical structural analysis and in vitro testing. The most important sensitization property of a chemical is its ability to form protein adducts, thus creating a complete antigen in the skin (Roberts and Aptula, 2008). Because chemicals forming protein adducts are generally electrophiles, their relative reactivity in this context likely correlates well with their ability to react with DNA and thus with their mutagenicity. Thus, information on the bacterial mutagenicity (and clastogenicity) of chemicals is valuable for estimating their sensitization potential (Mekenyan et al., 2010). Protein adduct formation may be modeled well for certain chemical classes such as Michael acceptors (eg, α,β-unsaturated carbonyls), which react with protein thiols, by their reactivity with glutathione (Roberts and Natsch, 2009; Schwoöbel et al., 2010), which also correlates well with their toxicity toward the model organism Tetrahymena pyriformis (Böhme et al., 2009). In such analyses, the contribution of metabolites of chemicals of interest merit consideration since, as in the case of diphenylthiourea (Samuelsson et al., 2011), they may be ultimate sensitizers.

Foreign body reactions, isolating invading substances that cannot be readily removed, occur infrequently toward a variety of agents introduced into the skin through injection or after laceration or abrasion. These can produce persistent lesions with abundant inflammatory cells resembling chronic infectious conditions (eg, tuberculosis, leprosy, leishmaniasis, and syphilis) and present diagnostic challenges (Del Rosario et al., 2005). In the case of silica or talc (a magnesium silicate), a resulting hard nodule may appear after a latent period of months or years as the original large particles disaggregate to assume a colloidal state. Injection of paraffin or mineral oil in the skin or contamination of wounds with starch powder cross-linked with epichlorohydrin for use in surgical gloves may also result in granulomatous reactions. Delayed allergic sensitization may occur with beryllium analogous to the reaction in the lung, and skin lesions have even been reported in individuals with life-threatening pulmonary exposure. An example of beryllium granuloma of the skin is shown in Fig. 19-2F. However, allergic sensitization may well contribute to the skin rashes that occur during acute beryllium exposure in a continuum of hypersensitivity reactions (Cummings et al., 2009). Cutaneous gadolinium deposition and fibrosis are seen in the nephrogenic systemic fibrosis syndrome evidently arising from gadolinium exposure (Wilford et al., 2010). Metallic mercury and zirconium compounds, formerly used in deodorants, and tattoo dyes (containing cobalt, chromium, mercury, lead, iron, cadmium, and manganese compounds) can also induce granulomatous reactions (Kaur et al., 2009) that in rare cases can be induced by intense light treatment (Tourlaki et al., 2010).

The ultraviolet and visible spectra of solar radiation reaching the earth extend from 290 to 700 nm. Wavelengths beyond this range are either filtered by the earth's atmosphere or are insufficiently energetic to cause cutaneous pathology. Adequate doses of artificially produced UVC (<290 nm) or X-rays can produce profound physical and toxicological skin changes. The protective skin pigment melanin, synthesized in melanocytes, absorbs a broad range of radiation from UVB (290-320 nm) through the visible spectrum. Other chromophores in the skin include amino acids, primarily tryptophan and to a lesser extent tyrosine, and their breakdown products (eg, urocanic acid), which absorb light in the UVB range. Biologically, the most significant chromophore is DNA, since damage from radiation can have lasting effects on the genetic information in target cells.

Adverse Responses to Electromagnetic Radiation

The most evident acute feature of ultraviolet radiation exposure is erythema (redness or sunburn). The minimal erythema dose (MED), the smallest dose of ultraviolet light needed to induce an erythematous response, varies greatly from person to person. Vasodilation responsible for the color change is accompanied by significant alterations in a variety of inflammatory mediators from injured keratinocytes and local inflammatory cells that may be responsible for some systemic symptoms associated with sunburn such as fever, chills, and malaise. UVB (290-320 nm) is the most effective solar band to cause erythema in human skin. A substantially greater dosage of UVA (320-400) reaches the earth compared to UVB (up to 100-fold); however, its efficiency in generating erythema in humans is about 1000-fold less than that of UVB. Both UVA and UVB have been implicated in the development of melanoma and nonmelanoma skin cancers. Because of its longer wavelength and greater depth of skin penetration, UVA is likely more responsible for long-term UV effects such as wrinkling, skin atrophy, and easy bruisability. Overt pigment darkening is another typical response to ultraviolet exposure. This may be accomplished by enhanced melanin production by melanocytes or by photo-oxidation of melanin. Tanning or increased pigmentation usually occurs within three days of ultraviolet light exposure, while photo-oxidation is evident immediately. The tanning response is most readily produced by exposure to UVB and may be induced, along with erythema and DNA repair, by DNA damage. Tanning serves to augment the protective effects of melanin in the skin over the long run, but in the short run the protection afforded appears insufficient to balance the damage sustained in acquiring it, especially in fair-skinned individuals (Sheehan et al., 2002).

Chronic exposure to radiation induces a variety of characteristic skin changes. For ultraviolet light, these changes accelerate or mimic aging, but the rate depends greatly on the baseline skin pigmentation of the individual. Lighter skinned people suffer from chronic skin changes with greater frequency than darker individuals, and locations such as the head, neck, hands, and upper chest are more readily involved due to their routine exposures. Pigmentary changes such as freckling and hypomelanotic areas, wrinkling, telangiectasias (fine superficial blood vessels), actinic keratoses (precancerous lesions), and malignant skin lesions such as basal and squamous cell carcinomas and malignant melanomas are all consequences of chronic exposure to ultraviolet light exposure. One significant pathophysiological response of chronic exposure to ultraviolet light is the pronounced decrease of epidermal Langerhans cells. Chronically sun exposed skin may have up to 50% fewer of these compared to photoprotected areas. This decrease may result in lessened immune surveillance of neoantigens on malignant cells and thus allow such a transformation to proceed unabated. For these reasons, gaining a tan, either naturally or through tanning salons, is not recommended (Lim et al., 2011). Exposures to ionizing radiation may produce a different spectrum of disease depending upon the dose delivered. Large acute exposures will result in local redness, blistering, swelling, ulceration, and pain. After a latent period or following subacute chronic exposures, characteristic changes such as epidermal thinning, freckling, telangiectasias, and nonhealing ulcerations may occur. Also, a variety of skin malignancies have been described years after skin exposure to radiation.

Aside from the toxic nature of electromagnetic radiation, natural and environmental exposures to certain bands of light are vital for survival. Ultraviolet radiation is critical for the conversion of 7-dehydrocholesterol to previtamin D3, without which normal endogenous production of vitamin D would not take place. Blue light in the 420 to 490 nm range can be lifesaving due to its capacity to photoisomerize bilirubin (a red blood cell breakdown product) in the skin. Infants with elevated serum bilirubin, potentially neurotoxic, have difficulty clearing this by-product because of its low water solubility, but treatment with blue light renders bilirubin more water soluble and markedly augments excretion. In addition, the toxic effects of ultraviolet light have been exploited for decades through artificial light sources for treatment of hyperproliferative skin disorders such as psoriasis.

Photosensitivity

An abnormal sensitivity to ultraviolet and visible light, photosensitivity may result from endogenous or exogenous factors (Bylaite et al., 2009). For instance, a variety of genetic diseases, such as xeroderma pigmentosum, impair the cell's ability to repair ultraviolet light-induced damage. The autoimmune disease lupus erythematosus also features abnormal sensitivity to ultraviolet light. In hereditary or chemically induced porphyrias, enzyme abnormalities disrupt the biosynthetic pathways producing heme, the prosthetic building block for hemoglobin, myoglobin, catalases, peroxidases, and cytochromes, leading to accumulation of porphyrin precursors or derivatives throughout the body, including the skin. These compounds in general fluoresce when exposed to light of 400 to 410 nm (Soret band), and in this excited state interact with cellular macromolecules or with molecular oxygen to generate toxic free radicals. A “constitutional” sensitivity to light (porphyria cutanea tarda) can be precipitated by alcohol, estrogens, or certain antibiotics in individuals with hereditary abnormalities in porphyrin synthesis, and an “acquired” sensitivity in general by hexachlorobenzene and mixtures of polyhalogenated aromatic hydrocarbons (Frank and Poblete-Gutiérrez, 2010).

Phototoxicity

Phototoxic reactions from exogenous chemicals may be produced by systemic or topical administration or exposure. In acute reactions, the skin can become red and blister within minutes to hours after ultraviolet light exposure. In an occupational setting, for example, exposing the skin to coal tar and sunlight can quickly produce a stinging sensation and elicit damage resembling a bad sunburn with hyperpigmentation. Phototoxic chemicals most commonly absorb ultraviolet light in the range of 320 to 400 nm (UVA), thereby assuming a higher energy excited triplet state, and either transfer an electron to other molecules or become reduced to form highly reactive free radicals (DeLeo, 2004; Moan and Peng, 2004). An oxygen-dependent photodynamic reaction commonly occurs as these excited molecules, returning to the ground state, transfer their energy to oxygen, forming highly reactive singlet oxygen. These reactive products are capable of damaging cellular macromolecules, notably unsaturated membrane lipids, and causing cell death. The resulting damage stimulates release of a variety of immune mediators from keratinocytes and local white blood cells that recruit more inflammatory cells to the skin and thus yield the clinical signs of phototoxicity. Agents most often associated with phototoxic reactions are listed in Table 19-6.

Psoralens are good examples of agents that produce phototoxicity largely without requiring oxygen radicals. Upon entering cells, they intercalate with the DNA and then form covalent adducts and cross-links when activated by UVA. The result is to inhibit DNA synthesis and repair and cell growth or survival. Psoralens (and other phototoxins) can be encountered in a variety of food plants, including limes and celery, producing phytophotodermatitis in occupational settings where such a food is routinely handled. An example of phototoxicity from lime juice is given in Fig. 19-2G. Psoralens in combination with UVA (PUVA) are used therapeutically to inhibit growth of keratinocytes and lymphocytes in hyperproliferative conditions such as psoriasis and T-cell lymphomas. In analogous fashion, though generating reactive oxygen, photodynamic therapy is a new approach that takes advantage of the selective uptake of a photosensitizing agent, usually a porphyrin precursor, into rapidly dividing cells such as in neoplasms. The photosensitizing agent is activated by laser or continuous wave light sources ranging from 400 to 600 nm or more, resulting in selective destruction of the neoplasm (Ibbotson, 2011).

Nails often suffer toxic reactions to drugs, most commonly to systemically administered agents such as antibiotics, less commonly to topically applied agents. While various types of nail pathology may result from therapy-induced changes in cell growth or division, changes in pigmentation or detachment of the various layers of the nail plate from each other (onycholysis) often result from photosensitive reactions, commonly phototoxic, less commonly photoallergic. Agents such as topical or systemic porphyrins, coal tar, and drugs such as the tetracycline family of antibiotics produce photodynamic toxicity, which requires generation of oxygen radicals. In contrast, topical or systemic psoralens produce onycholysis by a nonphotodynamic mechanism that does not require oxygen.

Photoallergy

A photoallergen elicits an allergic response by forming a complete antigen upon absorbing ultraviolet or visible light. Light stimulates the agent either to assume an excited state that can bind directly to a carrier protein or to yield a stable photoproduct that becomes conjugated to a carrier (Andreu et al., 2010). As in the case of allergic contact dermatitis, the complete antigen formed then is processed by Langerhans cells and presented to T cells. Upon recurrent exposure to an exogenous agent and light, a delayed hypersensitivity (Type IV) reaction ensues, leading typically to eczema (erythema, vesiculation, and itching). Like phototoxins, most photoallergens respond to light in the range of 320 to 400 nm (UVA). Light sensitivity usually subsides within days but may persist for several weeks if the agent is retained in the epidermis.

Use of halogenated salicylanilides and related antimicrobial agents in widely used consumer personal care products such as soap led to many thousands of cases of photoallergy in the 1960s and 1970s. When the most active agents were removed from the market, fragrances in cosmetics and sunscreens then became conspicuous sources of photoallergy a decade later. In parallel, certain antimicrobial agents were seen to be photoallergens (Kerr and Ferguson, 2010). Currently, active ingredients in sunscreens and antimicrobial agents are the most frequently recognized causes followed by medications, fragrances, plant derivatives, and pesticides (Victor et al., 2010). Although of low risk, these agents have become widely used as the risks of extensive sun exposure have become appreciated.

Photoallergy generally is distinguishable from phototoxicity, since the former results from delayed hypersensitivity, and amounts of agent too low to give a toxic response still suffice to elicit allergy. Agents causing phototoxicity, because reactivity is generated upon light exposure, may also be photoallergenic, but the latter type of response is much less common. Among individuals with a history of photosensitivity, a recent study reported that 11% exhibited photoallergy (Victor et al., 2010). Diagnosis is best performed by patch testing with and without light exposure of the treated surface to distinguish photocontact from contact allergy. Since the offending agent may not be obvious from the patient history due to the delay between exposure to the agent and sunlight and the symptoms, a panel of test agents may include some 50 common photoallergens as well as the patient's own sunscreen and personal care products. To assist in predicting risks of photoallergy, efforts have been made to derive important chemical features among existing photoallergens that account for their reactivity toward proteins. This information, coupled with assessment of physical properties such as aqueous/lipid partitioning, is anticipated to streamline testing of new products (Barratt, 2004).

Acne is a common affliction of the pilosebaceous units in the face, upper chest, and upper back. Found in many forms, this condition typically arises from blockage of the sebaceous duct leading from the gland to the hair follicle, resulting in retention of sebum and enlargement of the gland. In the most common form (acne vulgaris), androgen stimulation at puberty leads to high sebum production, hyperproliferation, and cornification of the ductal cells, plugging the orifice, and retention of corneocytes in the hair follicle lumen (Cunliffe et al., 2004). Proliferation of resident bacteria and inflammation typically result. Long-chain fatty acids can give an acne-like response in animal models and appear to do so in sensitive individuals exposed to cosmetics. Likely acting in a similar fashion, petroleum products (oils, coal tar) in the workplace can give rise to acneiform eruptions (Ancona, 1986). Insoluble cutting oils used in machining may have this effect, as illustrated in Fig. 19-2H. Such agents, which can contain chlorinated paraffins, stimulate excessive stratification of the ductal cells, preventing their disaggregation and blocking the flow of sebum.

Chloracne

The most disfiguring form of acne in humans, chloracne, is caused by exposure to halogenated aromatic hydrocarbons. Chloracnegenic agents such as polychlorinated biphenyls (PCBs) and naphthalenes are often accompanied by the highly potent polychlorinated dibenzofuran and dioxin contaminants readily formed during synthesis of these agents or chlorinated phenolic pesticides and can arise from combustion of the latter in fires and explosions. Chloracne is a rare disease but still occurs sporadically from industrial and environmental exposures (Gawkrodger et al., 2009; Passarini et al., 2010). Its recalcitrant nature and its preventability make it an important occupational and environmental illness. Typically, comedones and straw-colored cysts are present behind the ears, around the eyes, shoulders, back, and genitalia. In addition to acne, hypertrichosis (increased hair in atypical locations), hyperpigmentation, brown discoloration of the nail, conjunctivitis and eye discharge may be present. Since chloracne is a marker of systemic exposure, concurrent effects in the liver and nervous system may accompany the integumentary findings. Chloracne exhibits progressive degeneration of sebaceous units, transition of sebaceous gland cells to keratinizing cells and prominent hyperkeratosis in the follicular canal. Accounting for the distinctive effects on sebaceous glands, chloracnegens have been hypothesized to shift the differentiation of stem cells in the hair follicle to an epidermal program, which may contribute to their long lasting effects (Panteleyev and Bickers, 2006).

Since the original recognition of the syndrome a century ago, effects of chloracnegens on humans have been clearly demonstrated over the last 5 decades through industrial disasters. In 1953 a chemical plant in Ludwigshafen exploded, discharging 2,4,5-trichlorophenol; in 1976 in Seveso, Italy, a reactor explosion liberated tetrachlorodibenzo-p-dioxin (TCDD); and in 1968 and 1979 in Japan and Taiwan, respectively, rice cooking oil was contaminated with PCBs, polychlorinated dibenzofurans and related polychlorinated aromatic hydrocarbons. Individual poisonings involving remarkably high TCDD body burdens have not been lethal despite severe chloracne (Geusau et al., 2001; Sorg et al., 2009). Skin toxicity can appear within months after exposure and, depending on the severity, may either be cleared quickly or remain manifest even decades after exposure ceases, reflecting the high lipid solubility, recalcitrance to metabolic clearance, and thus long half-life in humans of the responsible agents. The highly potent TCDD has a half-life in humans of ≈ 8 years (Geyer et al., 2002), and highly chlorinated dibenzofurans and biphenyls are similarly persistent.

Chlorinated dioxins and dibenzofurans have substantial and reproducible effects on cellular function. TCDD is one of the most potent known inducers of CYP1A1 by virtue of its high affinity for the Ah receptor stimulating its expression. The fate of individuals afflicted with chloracne has been of particular interest in view of the suggestion they may be more prone to other adverse health effects of exposure. So far, however, chloracne has not proven to be a sentinel biomarker for cancer (Pesatori et al., 2009). Since polymorphisms in Ah receptor sequence confer differences in sensitivity to health effects in animals, analogous polymorphisms have been sought in the human population; Ah receptor polymorphisms have been identified, but as yet they have not been shown to affect responses to the receptor ligands (Okey et al., 2005). In the mouse, skin symptoms analogous to chloracne in humans are dependent on the hr locus, where the (−/−) genotype is susceptible (Poland and Knutson, 1982), although subclinical acne-like symptoms have now been observed in B6C3F1 mice (Ramot et al., 2009). An analogous locus has been identified in humans that governs hair loss, where mutations give a phenocopy of the generalized atrichia from loss of vitamin D receptor activity (Miller et al., 2001), but a functional allele does not prevent susceptibility to chloracne.

Melanocytes help protect the skin from harmful effects of ultraviolet light by producing the insoluble polymeric pigment melanin starting with the action of tyrosine hydroxylase. As indicated in Table 19-7, a variety of agents can interfere with normal pigmentation to yield either excessive or reduced amounts of melanin. An acquired condition of generalized pigmentation loss, leukoderma (vitiligo) has a genetic basis but is triggered by environmental influences, particularly involving generation of reactive oxygen (Guerra et al., 2010). Thought likely to be of autoimmune origin, this condition is characterized by loss of melanocytes. In the occupational arena, depigmentation is well known to occur through exposure to phenols, catechols, quinones, and related compounds, several of which are shown in Fig. 19-5, that appear toxic to melanocytes. An illustration of depigmentation caused by antioxidant exposure in rubber manufacture is given in Fig. 19-2I. In case reports, the cancer chemotherapeutic imatinib mesylate has been found to give pigment loss in the epidermis, probably through its inhibition of c-Kit tyrosine kinase signaling essential for melanocyte function (Arora et al., 2004), and paradoxically to give hyperpigmentation in the hard palate (Mattsson et al., 2011) and repigmentation of gray hair (Etienne et al., 2002). In sub-Saharan Africa and Asia, use of skin-lightening creams containing hydroquinone, corticosteroids, or mercurials is common among women. These agents produce their effects by inhibiting tyrosine hydroxylase activity or expression, but the treatments can give nonuniform effects or discoloration (ochronosis) and frequently are accompanied by undesirable cutaneous consequences including allergic contact dermatitis. Moreover, they pose risks of serious systemic effects, particularly the mercurials (nephropathy, neuropathy) and corticosteroids (adrenal suppression) (Ladizinski et al., 2011). By contrast, hyperpigmentation is well known to result from exposure to phototoxic agents including coal tar, coumarin derivatives found in perfumes and certain food such as limes (shown in Fig. 19-2G) and food plants (parsley, celery), dyes in cosmetics, and elements such as lead, bismuth, and arsenic (Table 19-7). An example of excessive pigmentation from mercaptobenzothiazole is shown in Fig. 19-2J. In addition, phototoxic or photosensitive reactions commonly occur with some oral medications, most notably diuretic, antibacterial, and nonsteroidal anti-inflammatory drugs (NSAIDs) (Moore, 2002).

For those allergens to which IgE antibodies have been elicited by previous or ongoing exposure, subsequent contact can lead to development of hives, typically in minutes, through an immediate type I hypersensitivity reaction. Hives are raised wheals that usually itch or sting and may appear reddish. Generally disappearing within hours and rarely lasting longer than a day or two, the symptoms result from degranulation of cutaneous mast cells by liganded IgE, leading to release of histamine and other vasoactive substances. Food allergies (Sicherer and Sampson, 2010) and pharmaceuticals (Limsuwan and Demoly, 2010) are major causes of acute urticaria, but many other causes are known (Table 19-8). Some agents (eg, opiates) can bring about direct release of histamine from mast cells without antibody mediation, while others (nonsteroidal anti-inflammatories) may do so through effects on arachidonic acid metabolism or by uncertain mechanisms. Responses lasting longer than six weeks, if not attributable to allergen exposure, frequently have an autoimmune basis such as development of autoantibodies against the high affinity IgE receptor. Lifetime prevalence of urticaria has been estimated as 9%, of which nearly 2% occurs chronically (Zuberbier et al., 2010).

Contact urticaria in an occupational setting can arise from exposure to plant or animal proteins and appears more common in atopic individuals, who are especially sensitive to irritants. Among the numerous occupations where this response occurs include hairdressers and those involving routine handling of food, plant, or animal products (Doutre, 2005). The common occupational problem of allergic contact dermatitis to rubber gloves, especially in the health care industry, reflects the variety of accelerators (carbamates, thiurams, 2-mercaptobenzothiazole, and 1,3-diphenylguanidine) and antioxidants (p-phenylenediamines) they contain (Cao et al., 2010). The widespread use of latex rubber in medical supplies and devices, as well as gloves, has resulted in widespread sensitization to the latex proteins they contain, cross-reaction with which may even be responsible for certain food allergies (Santos and Van Ree, 2011). The latex proteins have a propensity to induce immediate type I hypersensitive reactions, where the response can range from a mild skin reaction to anaphylaxis and death (Reines and Seifert, 2005). In addition, powder used in the gloves can adsorb the allergenic proteins and distribute them through the air. Modification of the manufacturing process to produce low-protein, low-allergenic, powder free gloves has led to reappraisal of their replacement by nitrile and other gloves (Palosuo et al., 2011).

A life-threatening response of anaphylactic shock, as from latex allergy, may occur from IgE-mediated massive release of histamine and other vasoactive agents from the mast cells upon systemic exposure to allergens. Certain food allergies (eg, to nuts, fish, and shellfish) are capable of producing this extreme response. Among pharmaceuticals, the best known example is penicillin, where adverse reactions must be managed carefully to avoid serious consequences (Gruchalla and Pirmohamed, 2006). Other antibiotics that can produce anaphylaxis include cephalosporins, sulfonamides, macrolides, vancomycin, tetracyclines, and fluoroquinolines; all but the last two also can produce toxic epidermal necrolysis. Insect allergens such as those in bee and wasp stings are well known to be capable of giving an anaphylactic response.

Toxic epidermal necrolysis is a rare life-threatening skin disease with an incidence of one to two cases per million person-years except among AIDS patients, where it is 1000-fold higher (Harr and French, 2010). At the most severe end of a spectrum of adverse cutaneous hypersensitivity reactions to drugs, this syndrome involves detachment of ≥30% of the epidermal surface from the dermis, commonly accompanied by severe erosions of mucous membranes, and has a fatality rate ≈30%. The Stevens–Johnson syndrome is a milder but still serious version of nearly the same phenomenon with up to 10% of the skin surface area affected and with a proportionally lower fatality rate. The two syndromes overlap when 10% to 30% of the skin surface is involved. Toxic epidermal necrolysis commonly resembles an upper respiratory tract infection in the first several days (fever, cough, sore throat, and malaise), but prompt diagnosis when the cutaneous lesions become evident several days later improves survival chances. Ongoing exposure to drugs first encountered during the previous month is stopped and treatment is initiated similar to that for burns, where fluid and electrolyte replacement, protection from infection, and nutritional assistance are critical. Nearly 200 drugs have been reported to cause this syndrome with major contributors being anticonvulsants, nonsteroidal anti-inflammatories, antibacterial sulfonamides, allopurinol, and nevirapine. Mechanisms leading to this idiosyncratic drug reaction are under scrutiny to prevent its occurrence or at least to optimize treatment. Studies of genetic predisposition have revealed striking correlations between HLA genotype and susceptibility to carbamazepine, but other reports and a recent genome wide association study raise the possibility that susceptibility may depend on the agent and ethnic background (Ahmadi, 2011). A characteristic feature of the syndrome is the large-scale apoptosis of epidermal keratinocytes. Candidates for mediating apoptosis through cell surface death receptors include tumor necrosis factor and FAS ligand, which appear elevated; in addition, drug-sensitized natural killer and cytotoxic T lymphocytes, secreting granulysin, and other components of the innate immune response may participate in inducing keratinocyte death (Bellon and Blanca, 2011). Effectiveness of treatments has been difficult to evaluate, but promising approaches involve immunosuppression (cyclophosphamide, ciclosporin) or blockage of death receptors using intravenous immunoglobulin therapy (Ardern-Jones and Friedmann, 2011).

Radiation

Radiation from ionizing wavelengths to ultraviolet wavelengths has been shown to cause skin cancer. Shortly after the discovery of radioactive elements at the turn of the 20th century, it was observed that X-rays could cause severe burns, squamous cell carcinoma, and basal cell carcinomas. X-ray-induced nonmelanoma skin cancers (NMCS) continued to be observed throughout the 20th century, as X-rays were used therapeutically until the mid-20th century for a variety of skin diseases (acne, atopic dermatitis, psoriasis, and tinea). Although NMSC from X-rays are now uncommon, dermal atrophy or sclerosis still is seen as sequelae of radiodermatitis, which sometimes develops after X-ray treatment of internal malignancies.

UV-Induced Skin Cancer

Most skin cancers in the United States now are UV-induced. The most common UV-induced skin cancers are NMSC and cutaneous malignant melanoma. NMSC are increasing in incidence and now number more than 1.2 million cases each year (Neville et al., 2007). The incidence of malignant melanoma also is increasing rapidly, faster than any other malignancy (Rigel, 2008). Although melanoma occurs with much less frequency than NMSC, it accounts for most of the deaths due to skin cancer in the USA (Balk, 2011). An important new source of UV radiation is the cosmetic tanning industry (Schulman and Fisher, 2009). The American Cancer Society, The Centers for Disease Control, the American Academy of Dermatology, and the National Council on Skin Cancer Prevention all have recommended that sun tanning and tanning beds be avoided (Balk, 2011). In addition to UV exposure, particularly blistering sunburns, as a risk factor for malignant melanoma and NMSC, genetic variations in the melanocortin 1 receptor predict increased skin cancer risk, independent of hair type or skin color (Lynde and Sapra, 2010).

UVB (290-320 nm) induces pyrimidine dimers and 8-oxoguanine modifications, thereby eliciting mutations in critical genes (Melnikova and Ananthaswamy, 2005; Nishigori, 2006). The p53 tumor suppressor gene is a major target in which damage occurs early and is detectable in most resulting squamous cell carcinomas. Because the p53 protein arrests cell cycling until DNA damage is repaired and may induce apoptosis, its loss destabilizes the genome of initiated cells and gives them a growth advantage. UV light also has immunosuppressive effects that may help skin tumors survive. Nonmelanoma skin cancer incidence is highest in the tropics and is highest in pale-complexioned whites, particularly at sites on the head and neck that receive the most intense exposure. Individuals with xeroderma pigmentosum, who are deficient in repair of pyrimidine dimers, must scrupulously avoid sun exposure to prevent the occurrence of premalignant lesions that progress with continued exposure. Even when it does not cause cancer in normal individuals, sun exposure leads to premature aging of the skin. For this reason, sunbathing is discouraged and the use of sunblock lotions is encouraged, especially those that remove wavelengths up to 400 nm. Protection from solar (and other) radiation as well as chemical carcinogens is important in the occupational arena (Gawkrodger, 2004).

Polycyclic Aromatic Hydrocarbons

A landmark epidemiological investigation by Percival Pott in 1775 connected soot with the scrotal cancer prevalent among chimney sweeps in England. Since that time, substances rich in polycyclic aromatic hydrocarbons (coal tar, creosote, pitch, and soot) have become recognized as skin carcinogens in humans and animals (Voelter-Mahlknecht et al., 2007). With their variable content of polycyclic aromatic hydrocarbons and nitrosamines, mineral oils have a long history of causing skin cancer in the occupational arena among, for example, metalworkers, cotton spinners, and jute processors (Tolbert, 1997). Polycyclic aromatic compounds alone are relatively inert chemically, but they would tend to accumulate in membranes and thus perturb cell function if they were not removed. They are hydroxylated by a number of cytochrome P450 isozymes, primarily 1A1 and 1B1 in epidermal cells, and conjugated for disposal from the body. Oxidative biotransformation, however, produces electrophilic epoxides that can form DNA adducts. Phenols, produced by rearrangement of the epoxides, can be oxidized further to electrophilic quinones that generate reactive oxygen species and lead to DNA adducts, oxidized bases, and abasic sites (Park et al., 2006). Occupations at risk of skin cancer from exposure to these compounds (eg, roofing) often involve considerable sun exposure, an additional risk factor. Concern persists regarding the use of coal tar for pharmaceutical purposes (Thami and Sarkar, 2004). The combination of coal tar and UV light is useful in treating severe psoriasis, since the toxicity reduces the excessive turnover rate of keratinocytes that characterizes this disease. Repeated treatments are necessary, raising the specter of elevated risk of nonmelanoma skin cancer. However, this deleterious outcome has not been manifest clinically (Roelofzen et al., 2010). Using longer wave UVA with phototoxic psoralen derivatives (PUVA) in place of coal tar in such protocols is much less messy and avoids UVB, but produces DNA adducts and a persistent elevation of nonmelanoma skin cancer risk (Nijsten and Stern, 2003).

As a major regulator of induction of CYP1A1, and to a lesser extent CYP1B1, the aryl hydrocarbon receptor (AhR) plays a critical role in the response of epidermal cells to polycyclic aromatic hydrocarbons (Swanson, 2004). For example, mice in which the AhR gene has been functionally ablated are insensitive to skin cancer from topical application of benzo[a]pyrene (Shimizu et al., 2000). By contrast, a constitutively activated AhR expressed in mouse epidermal keratinocytes was not seen to induce neoplasia, but it did result in inflammatory skin lesions reminiscent of contact dermatitis elicited by polycyclic aromatic hydrocarbons in humans (Tauchi et al., 2005). A common AhR polymorphism reduces the sensitivity in mice by an order of magnitude to toxic effects of ligands for this receptor, while in the rat an AhR variant with more subtle transactivation properties bestows dramatic resistance to lethal effects of TCDD without altering CYP1A1 inducibility (Okey et al., 2005). Screening efforts in humans have not yet revealed AhR polymorphisms that alter human responses to receptor ligands, but polymorphisms in cytochromes P450 do exist that are regulated by the AhR and could plausibly affect skin sensitivity or function. An example is CYP2S1, which metabolizes retinoic acid (Saarikoski et al., 2005).

Mouse Skin Tumor Promotion

Through the work of numerous investigators over the past century, initially studying substances rich in polycyclic aromatic hydrocarbons, mouse skin has been developed as an important target for carcinogenicity testing (Rubin, 2001). The observed incidence of squamous cell carcinomas has been helpful in providing a biological basis for conclusions from epidemiological studies. For instance, mouse skin carcinogenicity of tobacco smoke condensate and constituents strongly supported the conclusion that tobacco smoke is carcinogenic in humans. Carcinogenicity in mouse skin is taken as evidence of a carcinogenic risk for humans generally, including for the lung (Walaszek et al., 2007). Much has been learned about the pathogenesis of squamous cell carcinomas in mouse skin that does have general applicability to human squamous cell carcinomas of the skin or other anatomic sites, a learning process now augmented with mouse strains engineered for over-expression or ablation of genes of interest (Abel et al., 2009).

Initial fractionation of coal tar showed that the carcinogenic polycyclic aromatic hydrocarbons thereby purified and identified accounted for only a small fraction of the total carcinogenicity (Rubin, 2001). This puzzling observation led to the realization that other constituents, by themselves not carcinogenic, acted as cocarcinogens. Some agents, called tumor promoters, were found effective even when applied to the mouse skin long after a single treatment with a carcinogen applied at a dose too low to give cancer alone. The latter, inducing tumors by itself at sufficiently high doses, was a “complete” carcinogen with the salient property of being genotoxic. One explanation for the commonly observed nonlinear response is that a large dose of a complete carcinogen does more than initiate cancer by damaging the DNA in cells. It is also toxic, killing some cells and thereby stimulating a regenerative response in the surviving basal cells. Tumor promoters are agents that do not cause cancer themselves, but induce tumor development in skin that has been initiated by a low dose of a carcinogen. Their promoting power generally is parallel to their ability to give sustained hyperplasia of the epidermis with continued treatment by a variety of signaling pathways (Rundhaug and Fischer, 2010). Selective stimulation of tumor growth is envisioned to occur from differential stimulation of initiated cells or due to the insensitivity of initiated cells to toxicity or to terminal differentiation induced in uninitiated cells by the promoter. Continuing efforts to understand the promotion process are envisioned to assist development of short-term tests that predict promoting potential (Curtin et al., 2006).

An advantage of the experimental model of mouse skin carcinogenesis is the ability to separate the neoplastic process into stages of initiation, promotion, and progression, and the analysis can be extended to more complicated multihit, multistage models (Owens et al., 1999). In the simplest model, the skin is treated once with a low dose of an initiator, a polycyclic aromatic hydrocarbon, for example. The skin does not develop tumors unless it is subsequently treated with a promoter, which must be applied numerous times at frequent intervals (eg, twice per week for three months). Application of the promoter need not start immediately after initiation, but if it is not continued long enough, or if it is applied before or without the initiator, tumors do not develop. A consequence of promotion then is a tendency to linearize the dose–response curve for the initiator. Although an important aspect of promotion is its epigenetic nature, papillomas arising from promotion characteristically are aneuploid. Mutations in the c-Ha-ras gene are commonly found in papillomas, particularly those initiated by polycyclic aromatic hydrocarbons. Eventually some of the resulting papillomas become autonomous, continuing to grow without the further addition of the promoter. Genetic damage accumulates in the small fraction of tumors that progress to malignancy.

A number of natural products are tumor promoters, many of which alter phosphorylation pathways. The best-studied example, and one of the most potent, is the active ingredient of croton oil, 12-O-tetradecanoylphorbol-13-acetate. This is a member of a diverse group of compounds that give transitory stimulation followed by chronic depletion of protein kinase C in mouse epidermis (Fournier and Murray, 1987). Another group of agents, an example of which is okadaic acid, consists of phosphatase inhibitors. Compounds acting by other routes are known, including thapsigargin (calcium channel modulator) and benzoyl peroxide (free radical generator). Sensitivity to tumor promotion is an important factor in the relative sensitivity to skin carcinogenesis among different mouse strains and even among other laboratory animal species. An intriguing example, TCDD is 100-fold more potent than tetradecanoylphorbol acetate in certain hairless mouse strains but virtually inactive in some nonhairless strains (Poland et al., 1982). Much effort is currently focusing on the signal transduction pathways that are perturbed by treatment with such chemicals. The finding that chromate (a human lung carcinogen) in the drinking water enhances tumor yield from ultraviolet light in mouse skin suggests that generation of reactive oxygen can be effective and raises concern about the Cr(VI) drinking water standard (Davidson et al., 2004).

The desire to reduce the cost and improve the effectiveness of cancer testing in animals has led to development of transgenic mice with useful properties. The Tg.AC strain, for example, exhibits a genetically initiated epidermis (Humble et al., 2005). Integrated in this mouse genome is a v-Ha-ras oncogene driven by part of the ζ-globin promoter, which in this context fortuitously drives expression only in epidermis that is wounded or treated with tumor promoters. The mice display enhanced skin sensitivity to a number of nongenotoxic carcinogens and have low backgrounds of spontaneous tumors over a 26-week treatment period. They also respond to genotoxic carcinogens, which target genes that cooperate with the ras oncogene in neoplastic development (Owens et al., 1995). Tg.AC mice can be employed for testing in combination with a different mouse strain that has one p53 allele inactivated, and thus displays enhanced sensitivity to genotoxic carcinogens in other tissues in addition to the skin. These genetically modified mice promise to speed up testing using fewer animals and to reduce false-positive results arising from either strain-/species-specific idiosyncrasies or high doses that could elicit secondary responses not relevant to human exposures. Although replacement of the very expensive two-year rodent carcinogenesis assay is not yet feasible, these genetically modified mice have reduced the number of such assays and assisted pharmaceutical evaluation (Jacobs, 2005).

Arsenic

An abundant element in the earth's crust that is encountered routinely in small doses in the air, water, and food, arsenic has long been known to affect human health (Hughes et al., 2011). In earlier times, high exposures occurred from medications such as Fowler's solution (potassium arsenite) and from pesticides such as sodium arsenite and lead arsenate; the latter pesticides have been phased out in the USA but monomethylarsenate is still used in considerable (though decreasing) volumes in agriculture. Around the world, substantial exposures can occur as a result of mining operations to individuals who breathe the dust from ores or smelting fumes or to communities encountering leachate from tailings into waterways. Glass manufacture and food dried using coal with high arsenic content can also contribute to high exposures, and the environmental burden is increased by disposal of large volumes of wood preserved with chromated copper arsenate and of litter from poultry fed organic arsenical antibiotics. Well water derived from geological formations with high arsenic content is now recognized as a major health concern, highlighted by populations in the tens of millions at risk in Bangladesh and India (Tapio and Grosche, 2006). Previous studies of populations in southwestern Taiwan revealed an association of high arsenic exposure with altered skin pigmentation and hyperkeratosis of the palms and soles, blackfoot disease from impaired circulation reflecting endothelial cell damage, and carcinomas of the skin and several other organs (bladder, lung, and liver). More recent concentration dependence studies in Bangladesh show skin hyperpigmentation and keratoses occurring with drinking water concentrations close to 50 ppb (Argos et al., 2011), the former USA drinking water standard. Synergistic effects with arsenic have now been reported with smoking, sun exposure, and fertilizer use (Melkonian et al., 2011). Recognition that carcinomas of the skin and internal organs occur in humans drinking water with concentrations less than an order of magnitude higher than 50 ppb contributed to reducing the USA standard to 10 ppb as of 2006.

Due to continuing debate over the dependence of health effects on arsenic at low concentration, much effort is devoted to understanding arsenic action. Numerous mechanisms have been proposed by which arsenic serves as a carcinogen and gives other pathological effects, of which the majority could plausibly involve binding of trivalent forms to protein sulfhydryls (Kitchin and Wallace, 2008). Methylation has been considered a detoxification method, because the observed pentvalent mono- and dimethyl arsenates isolated in urine from exposed humans and animals are much less toxic than the inorganic forms. Wide species differences in methylation capability even among primates have raised the possibility that other detoxification pathways exist, and observations that the methylated forms are also found at low levels in the highly toxic trivalent state in animal tissues and in human urine have prompted re-evaluation of their roles. Exploring how the recently identified As(III)-methyltransferase influences toxic effects is anticipated to help clarify this important issue (Hughes et al., 2011). For the epidermis, where little methylation has been demonstrated, effects of the inorganic forms may dominate. As shown by many labs, effects of arsenate (the most common form in drinking water) in vivo or in culture are mediated by reduction to arsenite (Patterson et al., 2003).

Arsenic is a weak mutagen at best in bacteria and mammalian cells, but it does induce DNA deletions and chromosomal aberrations and acts as a comutagen, possibly by interfering with DNA repair (Rossman et al., 2004). Consistent with this view, high arsenic exposure from drinking water appears to promote skin lesions in users of betel nut (McCarty et al., 2006), a source of genotoxic nitrosamines known to be activated by cytochrome P450 activities in keratinocytes of the oral cavity (Miyazaki et al., 2005). An important contributing factor to such phenomena, limiting DNA repair, could be arsenite inactivation of poly(ADP-ribose) polymerase-1 by binding to its zinc finger motifs (Zhou et al., 2011). Similarly, in mice where the skin is irradiated with ultraviolet light, arsenic acts as a cocarcinogen (Rossman et al., 2004), and in Tg.AC mice, expressing an activated ras oncogene in the epidermis, arsenic enhances phorbol ester tumor promotion (Germolec et al., 1997). Among the proposed mechanisms of arsenic action, generation of reactive oxygen in target cells is a likely contributor to such carcinogenic and other pathological effects. This phenomenon has been proposed to result from activation of NADPH oxygenases (Cooper et al., 2009), damage to mitochondria (Liu et al., 2005), or, in the case of the methylated forms, (auto)oxidation (Kitchin and Ahmad, 2003). Pursuant to the demonstration that inorganic arsenic is a transplacental carcinogen for various tissues in mice (Waalkes et al., 2004), further investigation revealed that this treatment enlarged stem cell compartments, including in the epidermis (Tokar et al., 2011). Consonant with this observation, arsenite prevents differentiation and preserves stem cell character in human keratinocyte culture (Reznikova et al., 2010).