The Vascular System as a Target of Metal Toxicity

Abstract

Vascular system function involves complex interactions among the vascular endothelium, smooth muscle, the immune system, and the nervous system. The toxic metals cadmium (Cd), arsenic (As), and lead (Pb) can target the vascular system in a variety of ways, ranging from hemorrhagic injury to subtle pathogenic remodeling and metabolic changes. Acute Cd exposure results in hemorrhagic injury to the testis, although some strains of animals are resistant to this effect. A comparison of Cd-sensitive with Cd-resistant mouse strains showed that expression of the Slc39a8 gene, encoding the ZIP8 transporter, in the testis vasculature endothelium is responsible for this difference. Endogenously, ZIP8 is a Mn²⁺/HCO₃⁻symporter that may also contribute to Cd damage in the kidney. Chronic Cd exposure is associated with various cardiovascular disorders such as hypertension and cardiomyopathy and it is reported to have both carcinogenic and anticarcinogenic activities. At noncytotoxic concentrations of 10–100nM, Cd can inhibit chemotaxis and tube formation of vascular endothelial cells. These angiostatic effects may be mediated through disruption of vascular endothelial cadherin, a Ca²⁺-dependent cell adhesion molecule. With regard to As, ingestion of water containing disease-promoting concentrations of As promotes capillarization of the liver sinusoidal endothelium. Because capillarization is a hallmark precursor for liver fibrosis and contributes to an imbalance of lipid metabolism, this As effect on hepatic endothelial cells may be a pathogenic mechanism underlying As-related vascular diseases. With regard to Pb, perinatal exposure may cause sustained elevations in adult blood pressure, and genetically susceptible animals may show enhanced sensitivity to this effect. Taken together, these data indicate that the vascular system is a critical target of metal toxicity and that actions of metals on the vascular system may play important roles in mediating the pathophysiologic effects of metals in specific target organs.

Toxic metals, such as cadmium (Cd), arsenic (As), and lead (Pb), pose serious risks to human health. The importance of these metals as environmental health hazards is readily evident from the fact that all three are ranked in the top 10 on the current Agency for Toxic Substances and Disease Registry Priority List of Hazardous Substances (ATSDR, 2005). As a result of the extensive use of these metals and their compounds in industry and consumer products, these agents have been widely disseminated in the environment. Because metals are not biodegradable, they can persist in the environment and produce a variety of adverse effects. Exposure to these metals can result in damage to a variety of organ systems (Hughes, 2002; Ibrahim et al., 2006; Jarup et al., 1998) and, in some cases, these metals also have the potential to be teratogenic and carcinogenic (Han et al., 2000; IARC 1987; Kitchin, 2001; Waalkes et al., 1992).

Even though the importance of metals as environmental health hazards is now widely appreciated, the specific mechanisms by which metals produce their adverse effects have yet to be fully elucidated. However, a growing volume of evidence indicates that many of the effects of metals may result from specific actions on various components of the vascular system. These recent advances in the field of metal toxicology have coincided with advances in the understanding of the intricate functioning of the vascular system. Studies over the past 25 years have revealed that the vascular system is much more than the body's “plumbing.” Rather than being a static series of pipes and tubes, the vascular system is extremely dynamic and plays a critical role in homeostasis and in regulating the function of all organs of the body. The functioning of the vascular system involves complex interactions among the vascular endothelium, vascular smooth muscle, the immune system, the nervous system, and even the local chemical/metabolic environment of individual organs (Galley and Webster, 2004; Gibbins et al., 2003; Hill et al., 2001; Triggle et al., 2003; Villar et al., 2006). Recent studies have shown that toxic metals can target the vascular system for both acute injury and disease promotion. These vascular effects contribute to a variety of pathologic conditions including edema, atherosclerosis, and hypertension. In addition, the vascular effects of the metals may play key roles in mediating the toxic actions of metals in specific organ systems (Navas-Acien et al., 2005b; Prozialeck et al., 2006) and in promoting tumor growth (Kamat et al., 2005; Liu et al., 2006; Soucy et al., 2003). In order to highlight some of the most recent work in this area, a symposium titled “The Vascular System as a Target of Metal Toxicity” was held at the 2007 Meeting of the Society of Toxicology (SOT) in Charlotte, NC. This symposium was cosponsored by the Metals, Mechanisms and Toxicologic and Exploratory Pathology Specialty Sections of the SOT. The purpose of this report is to summarize the work presented in that symposium.

The Vascular Endothelium in Cd-Induced Edema and Hemorrhagic Injury

The vascular endothelium consists of specialized epithelial-like cells that line the lumenal surface of all blood vessels and form the capillary networks that mediate the delivery of oxygen and nutrients to tissues of the body. Obviously, vascular endothelial cells would be exposed to any toxic metal(s) circulating in the blood stream and, if the metals are present at sufficiently high concentrations, the endothelial cells could be injured or killed. The resulting loss of endothelial barrier integrity would result in edema and hemorrhaging in various tissues. Indeed, acute exposure to high levels of many metals has been shown to cause hemorrhaging in tissues such as the lung. However, one metal that appears to be unique in its ability to injure vascular endothelial cells and alter vascular permeability is Cd.

The idea that the vascular endothelium is an important target of Cd toxicity stemmed from an observation by Alsberg and Schwartze (1919) almost 90 years ago when they reported that acute exposure to subcutaneously administered Cd in rats caused purple discoloration of the testes. This observation went largely unnoticed until the 1950's and 1960's when other investigators reported that Cd caused hemorrhaging of the testes in a wide variety of species (Chiquoine, 1964; Hoey, 1966; Kar and Das, 1960; Parizek and Zahor, 1956). Later studies showed that Cd produced this effect by causing the breakdown of the junctions between the endothelial cells of the testicular capillaries and venules, resulting in an increase in vascular permeability, followed by edema, hemorrhage, and testicular necrosis (Aoki and Hoffer, 1978; Fende and Niewenhuis, 1977; Gabbiani et al., 1974; Gunn and Gould, 1970; Sacerdote and Cavicchia, 1983). A great deal of morphologic and biochemical evidence indicated that these effects of Cd on microvascular permeability resulted from direct actions of Cd on the endothelial cells in this particular vascular bed. However, this also raised the intriguing question as to why the endothelial cells in the testis were sensitive to this effect of Cd, whereas the endothelial cells in most other vascular beds were not affected.

Identification of A Specific Metal Transporter that Conveys Cd Sensitivity to Vascular Endothelial Cells

Lucis and Lucis (1969) discovered that certain inbred strains of animals were resistant to Cd-induced testicular necrosis. Discovery of mutant mouse strains provided the foothold needed for solving this mystery through the use of genetics and genomics research. By screening inbred strains of mice and performing various genetic crosses, Taylor et al. (1973) defined genetically the so-called Cdm locus within a 24-centiMorgan (cM) segment of DNA that confers sensitivity versus resistance to Cd-induced testicular necrosis. Using recombinant inbred lines and several dozen microsatellite markers, Dalton et al. (2000) were able to decrease the distance of 24 cM to 0.64 cM on mouse chromosome 3. With the advent of new knowledge about the mouse genome, the 0.64 cM was further reduced to 880 kb; one of the three functional genes therein was discovered to be Slc39a8, encoding the 8-transmembrane ZIP8 transporter (Dalton et al., 2005). In retrovirally infected mouse fetal fibroblast cultures (rvZIP8 cells), the complementary DNA (cDNA)–expressed ZIP8 protein was shown to enhance Cd uptake by 10-fold and increase sensitivity to cell killing by 30-fold. By in situ hybridization, two Cd-sensitive mouse inbred strains exhibited high ZIP8 expression in vascular endothelial cells of the testis, whereas two Cd-resistant strains showed negligible expression in these cells. Endothelial cell injury results in vascular leakage, which includes red cell extravasation and platelet plugging, ultimately causing testicular ischemia, followed by necrosis. Interestingly, although striking differences in ZIP8 expression are found in endothelial cells of the testis vasculature between inbred strains of mice, ZIP8 total messenger RNA (mRNA) levels are widely distributed in many tissues and do not differ substantially between strains. This observation, in conjunction with data demonstrating no mRNA sequence alterations between sensitive and resistant inbred strains of mice, led to the hypothesis that differential endothelial cell expression of Slc39a8 in blood vessels of the testis is the consequence of a DNA variant site(s) within an intron or in a 5′- or 3′-flanking region cis to the Slc39a8 gene (Dalton et al., 2005).

In Hank's balanced-salt solution, ZIP8 in the rvZIP8 cells has a K_m of 0.62mM for Cd and 2.2μM for Mn uptake. In the Madin–Darby canine kidney polarized epithelial cell line, ZIP8 is localized on the apical surface. Cd or Mn uptake is absolutely dependent on HCO₃⁻ in the medium. The ZIP8 endogenous function thus appears to be a Mn²⁺/HCO₃⁻ symporter, which Cd is able to highjack and thus gain entrance into cells, using a HCO₃⁻ gradient (He et al., 2006).

The Cd-sensitivity trait is dominant over the Cd-resistance trait (Dalton et al., 2000; Taylor et al., 1973). Therefore, the Slc39a8 gene was isolated on a bacterial artificial chromosome (BAC) clone, derived from a BAC library constructed from the genome of the Cd-sensitive 129S6/SvEvTac (abbreviated 129) strain. This BAC was inserted into the resistant C57BL/6J (B6) mouse genome. The BAC insertion turned out to be successful in demonstrating the accumulation of ZIP8 mRNA and protein specifically in the testicular endothelial cells of the BTZIP8-3 BAC-transgenic line (Wang et al., 2007). This means that the Cd-sensitive 129 strain's Slc39a8 gene is successfully expressed in the BAC-transgenic mouse line having > 99.8% of the Cd-resistant B6 strain's genes. Moreover, the phenotype of Cd-resistance was shown to revert to sensitivity to Cd-induced testicular necrosis, thereby proving unequivocally that the Slc39a8 gene is indeed the Cdm locus (Wang et al., 2007). It should be noted that, this genetic difference in ZIP8 expression between inbred mouse strains occurs only in these specialized vascular endothelial cells that one finds in the testis; hence, the only test for the phenotype is in the intact animal and cannot be carried out in vitro or in cell culture.

The BTZIP8-3 mouse line, having three Slc39a8 gene copies derived from the 129 mouse plus the two normal Slc39a8 gene copies from the B6 mouse, therefore has five Slc39a8 genes total and is viable and fertile. Two additional BAC-transgenic lines, one having seven, and the other eight Slc39a8 gene copies, never produced offspring. Wang et al., (2007) thus concluded that having seven or more Slc39a8 gene copies causes infertility and/or very early embryolethality. ZIP8 is highly expressed in placenta, and also contributes to Zn²⁺ uptake. It is therefore likely that Zn perturbation by excessive ZIP8 expression can lead to infertility and/or very early embryolethality.

Testing the BTZIP8-3 line for Cd-induced testicular necrosis led to an unexpected bonus finding: acute renal failure occurred and actually preceded damage to the testis by several hours; a high abundance of ZIP8 mRNA and protein was demonstrated on the apical surface of the renal proximal tubular epithelial cells (Wang et al., 2007). It is therefore postulated that ZIP8 might be the most important metal transporter in causing Cd-induced renal metabolic acidosis and kidney damage, conditions often seen in human populations chronically exposed to environmental Cd.

Effects of Cd on Adherens Junctions and Vascular Endothelial Cadherin-Cadherin

Although the foregoing findings with the S1c39a8/ZIP8 transporter have provided significant insights into the reason why certain vascular beds and strains of animals are sensitive to Cd-induced increases in microvascular permeability, they do not explain the mechanism by which Cd acts on the endothelial cells to produce this effect. Results of early morphologic studies suggested that this increase in microvascular permeability involved specific changes in the ultrastructure of the adhering junctional complexes that mediate adhesion between the capillary endothelial cells (Niewenhuis et al., 1997; Peereboom-Stegeman and Jongstra-Spaapen, 1979; Sacerdote and Cavicchia, 1983), although the specific molecular targets on which Cd acted to produce these effects remained unknown. However, results of recent studies suggest that these microvascular effects of Cd may involve alterations in the function of the Ca-dependent cell adhesion molecule, VE-cadherin (vascular endothelial cadherin).

VE-cadherin is a member of the cadherins superfamily of Ca-dependent cell adhesion molecules (Goodwin and Yap, 2004; Koch et al., 2004; Nollet et al., 2000). Although E-cadherin is the dominant cadherin expressed in most epithelial cells, vascular endothelial cells primarily express VE-cadherin (Dejana et al., 2001; Lampugnani et al., 1992; Vincent et al., 2004). Both E-cadherin and VE-cadherin are single pass transmembrane proteins that are usually localized at adherens-type cell–cell junctions. The extracellular domain of the cadherin contains multiple Ca-binding sites, as well as the adhesive regions of the molecule. The intracellular domain is bound to β-catenin and several associated molecules that link the junctional complex to the actin cytoskeleton.

The finding that the cadherins might be targets of Cd toxicity stemmed from a series of observations by Prozialeck and coworkers (Prozialeck, 2000; Prozialeck and Niewenhuis, 1991a,b) who found that exposing cultured renal epithelial cells to 10–20μM Cd for 1–4 h caused the cells to separate from each other and change morphologically from epithelioid to rounded. This effect coincided with the loss of E-cadherin from the cell–cell contacts and a reorganization of the actin cytoskeleton. These effects differed from those produced by other metals such Hg (Prozialeck and Niewenhuis, 1991b), but resembled those that occurred when the cells were incubated in the presence of the Ca chelator ethylene glycol-bis (β-aminoethylether) N,N,N′,N′-tetraacetic acid (EGTA) (Prozialeck, 2000). Moreover, they occurred at Cd concentrations and times of exposure that did not cause the loss of cell membrane integrity or alter cellular levels of ATP or glutathione, suggesting that they represented relatively specific toxic actions of Cd on the E-cadherin–dependent junctions between the cells (Prozialeck, 2000). The disruption of cadherin-dependent cell junctions by Cd was not cell- or cadherin-specific, as similar effects were also observed on E- and N-cadherin junctions in several other types of epithelial cells (Prozialeck, 2000). Additional studies have shown that Cd has similar effects on VE-cadherin in vascular endothelial cells in culture (Prozialeck, 2000; Prozialeck et al., 2006). In addition, studies utilizing a murine model of Cd-induced pulmonary injury have shown that Cd causes a redistribution of VE-cadherin in vascular endothelial cells of the lung (Pearson et al., 2003). Together, these findings indicate that VE-cadherin may be a key target on which Cd acts to disrupt endothelial barrier integrity. Recently, Pereira et al. (2007) have recently found that As also disrupts VE-cadherin–dependent intercellular junctions in vascular endothelial cells and they suggested that this effect may contribute to the development of atherosclerosis.

Effects of Cd and As on Angiogenesis and Vascular Remodeling

Over the past two decades the process of angiogenesis has been the subject of considerable research. In light of this large volume of work, it is somewhat surprising that relatively little has been published regarding the effects of metals on angiogenesis. However, there is a growing volume of evidence indicating that certain toxic metals, most notably Cd and As, can have profound effects on angiogenesis.

The process of angiogenesis involves several steps each of which can be studied in vitro. These steps that are shown schematically in Figure 1 include: (1) basement membrane degradation; (2) endothelial cell migration away from the vessel in response to a chemoattractant gradient; (3) endothelial cell proliferation; and (4) morphogenesis into tube-like structures (Carmeliet, 2000; Hayden and Tyagi, 2004). In the mid 1990's, Kishimoto et al.(1996a; 1996b) reported that Cd inhibited proliferation, migration and tube formation by endothelial cells in culture. However, these studies were carried out in the absence of serum and utilized relatively high concentrations of Cd that also affected cell viability. In addition, some of the results suggested that Cd produced these effects by acting on the Matrigel matrix rather than the endothelial cells themselves (Kishimoto et al., 1996b). Recent studies from the Woods and Prozialeck laboratories have shown that Cd, at noncytotoxic concentrations, has direct effects on endothelial cell migration and tube formation (Prozialeck et al., 2006). These inhibitory effects clearly resulted from direct actions on the endothelial cells and were evident when the cells were exposed to concentrations of Cd as low as 0.1–1.0μM in the presence of serum, conditions that mimic the patterns of exposure of endothelial cells in vivo. Additional studies showed that these same levels of exposure resulted in a loss of VE-cadherin from the cell–cell contacts (Prozialeck et al., 2006).

Fig. 1.

The effects of Cd on the angiogenic process. The angiogenic process can be broken down into several sequential steps, which can be examined in vitro, including endothelial cell migration, proliferation, and tube formation. Above these steps, labeled in the figure, is the normal angiogenic process occurring in the absence of Cd. Endothelial cells migrate in response to a chemoattractant, they proliferate while extending the existing blood vessel, and ultimately they differentiate into tube-like structures. The latter process is known to involve VE-cadherin, which is involved in cell–cell adhesion as well as intercellular formation of the lumen (Yang et al., 1999). Below the labels, in the presence of Cd, we summarize evidence suggesting that vascular quiescence predominates in the presence of Cd. Studies from our lab and others (Kishimoto et al., 1996a) suggest that Cd inhibits endothelial cell migration, proliferation, and the process of tube formation. Moreover, our data suggest that VE-cadherin is sequestered away from the endothelial cell surface, which may cause the inhibition of tube formation.

In light of the fact that VE-cadherin plays a critical role in endothelial tube formation, these findings suggest that Cd disrupts angiogenesis by redistributing VE-cadherin from the endothelial cell surface and inhibiting cell migration and tube formation. Recently, Kolluru et al. (2006), have reported additional evidence showing that Cd can directly inhibit endothelial cell migration and tube formation, and they presented evidence that these effects may be associated with decreased nitric oxide (NO) production by the endothelial cells. Additional studies are needed to examine the possible relationships between the Cd-induced changes in NO metabolism and the alterations in VE-cadherin expression and localization.

In contrast to Cd, low levels of As(III) promote endothelial cell tube formation and angiogenesis in both cell culture and in vivo mouse models (Kamat et al., 2005; Kao et al., 2003; Liu et al., 2006; Soucy et al., 2005). Angiogenesis is the rate limiting step in tumor growth and the in vivo angiogenic effects of low dose As(III) promote tumor growth in mouse xenograph models (Kamat et al., 2005; Liu et al., 2006). However, higher doses of As(III) are toxic to endothelial cells (Roboz et al., 2000) and inhibit angiogenesis (Kao et al., 2003; Liu et al., 2006; Soucy et al., 2003). As noted previously, angiogenesis is a complex process of endothelial cell proliferation, migration, and vessel maturation (reviewed in Carmeliet, 2000; Hayden and Tyagi, 2004). Pathological angiogenesis is usually accompanied by recruitment of inflammatory and progenitor cells that elaborate growth factors to complete remodeling of the new vessel wall (Ruiz et al., 2006). A recent examination of As(III) effects on inflammatory angiogenesis demonstrated dose-dependent increases in CD45 positive leukocytes in Matrigel plugs implanted in mice exposed to 50–500 ppb (0.3–3.3μM) of As(III) through their drinking water, relative to plugs from control unexposed mice (Straub et al., 2007b). This increase in CD45 positive cells was prominent after 5 weeks of exposure and remained significant through 20 weeks of exposure (Soucy et al., 2005). These same exposures produced highly significant increases in the number of CD31/PECAM positive blood vessels (i.e., luminal structures containing red blood cells) in the Matrigel plugs (Soucy et al., 2005). An important finding in these studies was that the threshold for the angiogenic response was between 1 and 5 ppb (6 and 33nM) (Soucy et al., 2005), which is below the current drinking water arsenic MCL of 10 ppb (66nM).

A limitation of xenograph models and the mouse Matrigel models of angiogenesis is that both are inherently inflammatory in the mouse. This inflammatory potential is appropriate for modeling potentiation of tumor angiogenesis, but it may mask direct pathogenic vascular effects that promote the range of vascular diseases caused by environmental exposure to As(III). These vascular diseases include hypertension, atherosclerosis, coronary vessel disease, noncirrhotic portal hypertension, and possibly diabetes (Navas-Acien et al., 2005a).

To examine the effects of As(III) on an endogenous vascular bed, we focused studies on the specialized vasculature of the liver sinusoids (Straub et al., 2007a,b). The liver sinusoidal endothelial cells (LSEC) are both morphologically and phenotypically unique relative to macrovascular and microvascular endothelium. Their roles as the major scavenger cells and filters in the liver are facilitated by their fenestrations and loose intracellular connections, as well as their unique expression of scavenger receptors on a specialized microtubular network (Falkowska-Hansen et al., 2007). These receptors allow the LSEC to be the primary site of removal of all major circulating biological molecules, including modified proteins, polysaccharides, lipids, and nucleic acids (Falkowska-Hansen et al., 2007) and loss of this function has significant systemic consequences (McCuskey, 2006). Exposure of mice to As(III) in their drinking water caused defenestration and capillarization of the LSEC (Straub et al., 2007a,b). Capillarization is a process in which the normally fenestrated, discontinuous LSEC become a continuous endothelium with limited transendothelial cell transport due to loss of fenestrae, formation of tight intercellular endothelial junctions, and formation of a basement membrane (Braet and Wisse, 2002; Couvelard et al., 1993; Dubuisson et al., 1995; Xu et al., 2003). Capillarization precedes vascular remodeling of other liver vessels, such as the hepatic arterioles and the peribiliary vascular plexus causing the shunting of blood flow, vascular channel formation, and eventually liver fibrosis (Couvelard et al., 1993; DeLeve et al., 2004; Li et al., 2005).

Quantitative morphometric analysis revealed that 2- to 5-week exposures to 250 ppb As(III) decreased the average size of the fenestrae and eliminated gaps between cells to decrease overall sinusoid porosity (i.e., open space per unit area), relative to unexposed mice (Straub et al., 2007a). The surface of the As(III) exposed sinusoids also showed an increase in associated detritus and projections, some of which were microvilli from the underlying hepatocytes protruding through the remaining LSEC fenestrae. There were no zonal differences along the sinusoids for the effect of As(III) on porosity (Straub et al., 2007a) suggesting that an effect directly on the endothelial cells and not dependent on the metabolic function of the underlying hepatocytes. Quantitative immunofluorescent analysis demonstrated that junctional expression of CD31/PECAM-1 protein increased as porosity decreased (Straub et al., 2007a,b). Concomitant with the increase in PECAM-1, a laminin-1 positive basement membrane formed. These protein changes confirmed that As(III) promoted the formation of tight endothelial cell junctions and underlying matrix that are characteristic of continuous endothelium.

Further studies demonstrated that the effect of As(III) was both dose and time dependent. Exposure to As(III) at the current arsenic MCL of 10 ppb (66nM) reduced porosity by 20–30% and 50 ppb exposures caused almost complete loss of porosity. These decreases were partially apparent following 1 week of exposure, but were highly significant at 2 weeks. There were no additional differences in As(III) effects on porosity between 2 and 5 weeks (Straub et al., 2007b). It is important to note that during the 5-week experiments the sinusoidal porosity in the control mice also tended to decrease. This was expected because capillarization increases with age and is thought to contribute to the age-related risk for atherosclerosis (Cogger et al., 2004; Hilmer et al., 2005). Transmission electron microscopy revealed that as porosity decreased, the hepatocytes developed microvilli that clogged the space of Disse. After 5 weeks, but not 2 weeks, exposures, caveolae were apparent on the surface of the LSEC (Straub et al., 2007b). This increase in caveolae was associated with appearance of caveolin-1 protein. These data suggest that As(III) caused time-dependent changes in LSEC function from proficient scavengers to normal endothelium that use caveolae to transcytose macromolecules. Because caveolin-1 is a major scaffold for endothelial cell signaling proteins, the data suggest that prolonged As(III) exposure changes the signaling phenotype of the cells, as well.

The cellular mechanisms for the vascular effects of arsenic remain unresolved. However, As(III) has been shown to stimulate Rac1-GTPase, which mediates both endothelial shape change and nicotinamide adenine dinucleotide phosphate (reduced) oxidase (NOX)–dependent generation of oxidants (Qian et al., 2005; Smith et al., 2001). To investigate whether LSEC Rac1 was affected by As(III) in vivo, livers were perfused with a colloidal silica solution to specifically coat the LSEC membranes immediately after euthanizing the mice (Straub et al., 2007a). The colloid-bound membranes were separated from the rest of the liver cell membranes and then probed for Rac1 protein changes relative to changes in actin content. The analysis revealed that only the LSEC cell membranes from arsenic exposed mice contained Rac1 protein. To further investigate whether As(III) has a direct functional effect on LSEC Rac1, LSEC were isolated from unexposed mice and cultured in collagen-coated dishes. Addition of 1–5μM As(III) to the cultures promoted junctional PECAM-1 expression. Incubating the cells with apocynin, a Rac1 and NOX inhibitor (Klees et al., 2006), before adding As(III) prevented the As(III)-stimulated expression of PECAM-1. DeLeve et al. (2004) demonstrated that LSEC fenestrations are maintained by vascular endotheliam growth factor (VEGF)-stimulated NO production. Thus, As(III) stimulation of Rac1-dependent NOX superoxide generation and cytoskeletal regulation provide mechanisms for defenestration through NO quenching and cell shape change. This mechanism is schematically shown in Figure 2. More definitive studies are needed to resolve how As(III) stimulates LSEC Rac1 and to prove the role of Rac1 in mediating LSEC activation and capillarization. Nonetheless, these data suggest that low, noncytotoxic levels of As(III) signal through specific cellular pathways to functionally alter the phenotype of the endothelial cells in this important metabolic vascular bed.

Fig. 2.

Hypothetical scheme for AS(III)-stimulated remodeling of LSEC. Fenestrations in the normal LSEC Rac1-GTPase activity are maintained by constitutive VEGF-stimulated NO production As(III) stimulates LSEC Rac1-GTPase activity which increases NOX generated superoxide and regulates cell spreading. Thus, As(III) may promote defenestration and capillarization by quenching NO through peroxynitrite formation (ONOO) and cytoskeletal rearrangement. The capillarized endothelium is stabilized by increased junctional expression of PECAM and formation of a laminin-containing basement matrix.

Cd and Pb in Hypertension

Results of a large number of epidemiologic studies suggest that exposure to metals such as Cd and Pb may play a role in the development of hypertension (for reviews see Harlan, 1988; Nakagawa and Nishijo, 1996; Navas-Acien et al., 2005b; Pirkle et al., 1985; Satarug and Moore, 2004; Schwartz, 1988). For example, many epidemiologic studies have suggested a possible association between exposure to Cd and hypertension (Luoma et al., 1995; Satarug et al., 2003, 2005; Vivoli et al., 1989), although other studies showed no direct relationship between blood pressure and blood levels of Cd (Fontana and Boulos, 1986; Staessen et al., 1996, 2000). In addition, a large number of animal studies have shown that chronic exposure to Cd can lead to elevations in blood pressure (Nechay, 1978; Ohanian et al., 1978; Oner et al., 1995; Perry et al., 1977; Revis et al., 1981).

Although the cardiovascular system is not typically viewed as a primary target of Pb toxicity, high concentrations of Pb, such as might occur during occupational exposure, are toxic to both the heart and vascular smooth muscle. Furthermore, because the kidney is a primary organ of Pb toxicity, indirect cardiovascular effects could occur secondarily to renal injury. A number of studies have found correlations among occupational Pb exposure, nephrotoxicity, and increased arterial blood pressure (Harlan, 1988; Lal et al., 1991; Pirkle et al., 1985; Schwartz, 1988). However, in these studies blood Pb levels ranged from 100 to 120 μg/dl (4.7–5.8μM) (Gross, 1981; Kehoe, 1961). At these exposures, severe kidney damage has generally been present; thus, this hypertension is probably of renal origin.

Several more recent epidemiological studies suggest that Pb concentrations below Center for Disease Control limits, and within the realm of environmental exposure, may be a risk factor for hypertension. A study in which bone Pb levels, as determined by K-shell x-ray fluorescence, were compared with development of hypertension provides one of the stronger pieces of evidence for environmental Pb exposure as a risk factor in development of hypertension (Cheng et al., 2001). Some reports suggest that even transient exposure to Pb during childhood can have a long-term and delayed hypertensive effect (Hu, 1991). Studies in experimental animals also support a correlation between low-level exposure to Pb and development of hypertension (Aviv et al., 1980; Kopp et al., 1980; Perry et al., 1988; Vaziri et al., 1999). However, whether or not low-level exposure to Pb is actually a causal factor in hypertension remains controversial, as other epidemiological studies found no definitive connection between blood pressure and blood or bone Pb levels (Staessen et al., 1999). Moreover, the increases in blood pressure associated with Pb exposure, though consistent, are modest. Vupputuri et al. (2003) found that Pb exposure (blood Pb ≥ 5 μg/dl; 0.24μM) caused a 1.67 and 1.68 mmHg increase for black males, and a 3.48 and 2.22 mmHg increase in black females. Hypertension involves interactions among the cardiovascular, nervous, renal, and endocrine systems, and is expressed primarily in adulthood, whereas environmental-level exposure to Pb is generally without obvious symptoms and typically occurs during childhood. Thus, as noted by Vaziri et al. (1988), it may be difficult to make an association of a role for Pb in a disease in which the onset is far removed from the postulated exposure, with no noticeable intervening pathophysiology.

Epidemiological data in humans also indicate that certain subgroups are especially sensitive to Pb. Men (Glenn et al., 2003) and postmenopausal women (Nash et al., 2003) are more likely to develop hypertension with occupational Pb exposure than are premenopausal, adult women. Muntner et al. (2005) found that both non-Hispanic blacks and Mexican-Americans had a higher association of Pb with hypertension. One potential common explanation for these findings is that individuals predisposed to develop hypertension may also be more prone to increased risk of hypertensive responses to Pb.

To examine the role of perinatal Pb exposure on the development of adult hypertension in a “susceptible population” studies were undertaken using the “spontaneously” hypertensive rat (SHR) as a model. SHRs were exposed to Pb acetate in the maternal drinking water (100 ppm, 263μM) perinatally or perinatally and postweaning (100 ppm in drinking water). Blood Pb levels were undetectable in rats receiving water, or Pb perinatally only, and ranged from 0 to 12.5 μg/dl (60μM) in rats exposed to Pb postweaning. SHR became markedly hypertensive over the next 6 months as expected. At this point radiotelemetry transmitters were surgically implanted and the animals were allowed 7 days of surgical recovery. Subsequently, one week of 24 h/day radiotelemetric recordings of blood pressure, heart rate, and activity were made. Continuous 24-h average blood pressure measurements revealed that neither perinatal nor perinatal plus postweaning Pb exposure affected basal blood pressure. A more comprehensive study is underway using a wider range of Pb doses, and the comparative response of Pb-treated and untreated SHR rats to several commonly used antihypertensive drugs. Nonetheless, using the most sensitive method available for measuring blood pressure in rodents, and a genetically susceptible strain of rats, we were unable to demonstrate a statistically significant increase in basal blood pressure in this initial study.

There are many reports in the literature proposing possible physiological and cellular mechanisms of metal-induced hypertension based on animal and in vitro studies. Figure 3 highlights some of the reported mechanisms by which metals such as Cd and Pb could contribute to the development of hypertension. In the case of Cd, considerable evidence suggests that hypertensive effects result from complex actions on both the vascular endothelium and vascular smooth muscle. For example, Cd causes the release of a variety of proinflammatory mediators such as tumor necrosis factor alpha, from endothelial cells, see site 1 (Kaji, 2004; Mlynek and Skoczynska, 2005; Szuster-Ciesielska et al., 2000). In addition, Cd stimulates the release of antithrombolytic agents such as plasminogen activator inhibitor-1 and facilitates the adhesion of leukocytes and platelets to the endothelium, see site 2 (Hernandez and Macia, 1996; Kaji, 2004; Yamamoto et al., 1993). Pb may cause enhanced sympathetic nerve activity with increases in circulating epinephrine and norepinephrine levels (site 3) in conjunction with decreased density of vasodilating β₂ adrenergic receptors (site 4) (Chang et al., 1996, 1997; Tsao et al., 2000). Although acute exposure to Cd and Pb results in depressed plasma renin levels (Fleischer et al., 1980; Puri, 1992) chronic low-level exposure to Pb results in increased activity of angiotensin converting enzyme activity and increases in plasma renin, angiotensin II, and aldosterone levels (Boscolo and Carmignani, 1988; McAllister et al., 1971; Vaziri, 2002). Plasma kininase I and II levels are higher during Pb exposure. This can lead to decreases in plasma bradykinin levels resulting in a reduction in endothelial NO production (site 5) (Carmignani et al., 1999). In rodent models of Pb-induced hypertension there is an elevation in plasma concentration of the potent vasoconstrictor, endothelin-3, see site 6 (Khalil-Manesh et al., 1993). Furthermore, coronary microsvascular endothelial cells exposed to 2μM CdCl₂ exhibit increased secretion of endothelin-1 and angiotensin II (Kusaka et al., 2000). However, Cd was also found to antagonize the actions of endothelin (Koschel et al., 1995; Wada et al., 1991). There is very strong evidence that both Cd and Pb decrease the functional availability of the potent vasodilator NO (site 7), most likely through direct or indirect mechanisms involving oxidative stress (Bilgen et al., 2003; Grabowska-Maslanka et al., 1998; Kishimoto et al., 1994; Skoczynska and Martynowicz, 2005; Vaziri, 2002). Elevations of intracellular smooth muscle Ca could lead to increased arterial tone leading to hypertension. Pb is a well-known inhibitor of Na/K ATPase. Inhibition of this enzyme leads to elevations of intracellular Na resulting in increased intracellular Ca levels, see site 8 (Piccinini et al., 1977). Another mechanism by which metals can affect vascular smooth muscle is by altering protein kinase C (PKC) activity. Both Pb and Cd exposure are associated with increased PKC activity (Hwang et al., 2001; Washington et al., 2006). Furthermore, in isolated tissue baths, PKC and L-type voltage-gated Ca channel antagonists, when applied separately, significantly diminished Pb-induced vascular smooth muscle contraction, see site 9 (Watts et al., 1995). Lastly, Cd promotes the proliferation of vascular smooth muscle cells (site 10) and enhances the production of extracellular matrix components (site 11) that increase the stiffness of blood vessels (Abraham et al., 2000; Fujiwara et al., 1998; Jeong et al., 2000; Kaji, 2004).

Fig. 3.

Synopsis of proposed physiological and cellular mechanism of metal-induced hypertension. On the graph, sites 1–7 reflect endocrine or paracine vaso-active mediators/receptors, sites 8–11 reflect direct changes in vascular smooth muscle physiology and associated connective tissue caused by metal exposure. Abbreviations: αR = α adrenergic receptor; ACE = angiotensin converting enzyme; βR = β adrenergic receptor; E = epinephrine; ECM = extracellular matrix; IL = interleukin; NE = norepinephrine; NO = nitric oxide; PAI-1 = plasminogen activator inhibitor -1; PKC = protein kinase C; SMC = smooth muscle cells; TNF = tumor necrosis factor α.

These mechanisms reflect the multifactorial nature of hypertension with the obvious interdependence among several organ systems, as well as the fact that Pb and Cd have many and multifarious actions. It remains to be determined how all of these various effects of metal exposure summate to influence the net regulation of blood pressure.

Although the evidence in support of a role of metal-induced hypertension is strong, it is not conclusive, nor are the mechanisms by which Pb or Cd may act on the vascular system clear. Pb exposure may be more of a risk factor for certain “susceptible populations”, and thus may not be apparent in the general population. Inasmuch as hypertension remains a significant risk factor for other forms of cardiovascular disease, the propensity for environmental exposure factors to contribute to this risk is, potentially, an important public health concern.

Conclusion

The functioning of the vascular system involves complex interactions among multiple cell types, with each one utilizing a myriad of cellular signaling pathways that allow the vascular system to respond and adapt to ever-changing environmental conditions. Recent studies have provided new insights into the mechanisms by which metals can influence vascular function. Expression of the ZIP8 metal ion transporter (Slc39a8 gene) appears to be a key factor contributing to the selective toxicity of Cd in the endothelial cells of organs such as the testes and kidneys. At the cellular level, metals such as Cd and As have profound effects on the process of angiogenesis. These effects involve alterations in the function and expression of cell adhesion molecules such as VE-cadherin and PECAM-1, although the specific signaling pathways that mediate these actions have yet to be elucidated. Further research is needed, especially in the area of metal-induced hypertension, to determine the significance and the mechanism of the adverse effects. Clearly, much work remains to be done. It is our hope that this symposium overview will be useful to investigators in this field and help to provide a conceptual framework for future studies.