Arsenite Increases Vasoconstrictor Reactivity in Rat Blood Vessels: Role of Endothelial Nitric Oxide Function

Abstract

Arsenite has been shown to inhibit endothelium-dependent, nitric oxide-mediated vasodilation in vitro. This study investigated the effects of arsenite on vascular reactivity in vivo. Saline or sodium arsenite (6 mg kg⁻¹) was administered intravenously in Wistar-Kyoto rats for 4 h. As compared to saline, arsenite significantly increased vasoconstrictor responses to phenylephrine in both rat isolated aorta and renal arteries examined in tissue bath. This change was diminished after preincubation of the tissues with the nitric oxide synthase inhibitor N ^G-nitro-l-arginine methyl ester, which increased phenylephrine-induced vasoconstriction to a similar extent as arsenite. In contrast, acetylcholine-induced vasodilation, mediated by nitric oxide in the aorta and by an endothelium-derived hyperpolarizing factor in renal arteries, was not affected by arsenite. Arsenite induced expression of heat shock proteins Hsp72, Hsp32, and Hsp90, but endothelial nitric oxide synthase expression was not changed. The effects of arsenite on vasoreactivity were unlikely to be mediated by heat shock protein induction, because blockade of heat shock protein induction had little effect on the increased vasoconstriction in vessels from arsenite-treated animals. Our study suggests that in vivo arsenic treatment increases vasoconstrictor reactivity by compromising basal endothelial nitric oxide function, which is not caused by altered endothelial nitric oxide synthase expression.

Chronic exposure to arsenic through food or water is related to increased incidence of vascular diseases in humans (Engel and Smith 1994), and evidence have suggested that arsenic may contribute to the development of atherosclerosis (Wang et al. 2002; Simeonova et al. 2003), hypertension (Chen et al. 1995), and is-chemic stroke (Chiou et al. 1997). The mechanisms by which arsenic compounds induced cardiovascular disease are not totally clear, and studies have demonstrated that these may involve activation of inflammatory responses (Bunderson, Coffin, and Beall 2002; Tsai et al. 2002), increased oxidative stress (Smith, Klei, and Barchowsky 2001), increased platelet activity and thrombosis (Lee et al. 2002), and aberrant cell mitogenesis and apoptosis (Barchowsky et al. 1999b; Yeh et al. 2003).

Regulation of vasomotor activity by neuronal and humeral mediators is an important mechanism to maintain the homeostasis of the circulatory system, which controls regional blood flow and systemic blood pressure. Vascular tone is under strict regulation by endogenous vasoactive factors, of which norepinephrine and nitric oxide (NO) are the predominant vasoconstrictor and vasodilator respectively. Norepinephrine is released from perivascular adrenergic nerve endings and contracts vascular smooth muscle by activating α adrenoreceptor and subsequent increase in intracellular Ca²⁺ (Docherty 1998). NO is largely produced from vascular endothelial cells by endothelial nitric oxide synthase (eNOS) and relaxes smooth muscle by activation of guanylyl cyclase and accumulation of intracellular cyclic guanosine monophosphate (cGMP) (Moncada, Palmer, and Higgs 1991). A recent study by Lee et al. has shown that treatment of rat isolated aorta with sodium arsenite for 14 h in vitro markedly reduced endothelium-dependent, NO-mediated vasodilation induced by acetylcholine, and this effect of arsenite might be attributable to a direct inhibition of eNOS activity and impairment of the smooth muscle responsiveness to NO (Lee et al. 2003). Although these authors also provided evidence that in vivo administration of arsenite blunted acetylcholine-induced reduction in systemic blood pressure, indicating that NO-mediated vasodilator response was impaired by arsenite, direct information about the effects of arsenic on vasomotor function in vivo is still unavailable.

Moreover, arsenite is a potent inducer of heat shock proteins (Hsp), which are intracellular chaperones expressed in response to various cellular stresses, such as heat, hypoxia, oxidative stress and toxins (Benjamin and McMillan 1998). These proteins have extensive cytoprotective effects (Jaattela 1999). Some studies indicate that Hsp may also be involved in modulating vasoreactivities. For example, in isolated rat aorta, in vitro incubation with arsenite attenuated phenylephrine-induced vaso-constrictions through increased generation of carbon monoxide by Hsp32 (Caudill et al. 1998), although this effect is still inconclusive (Knoepp et al. 2000). Whether the observed inhibition of NO-mediated vasodilation by arsenite (Lee et al. 2003) is associated with Hsp expression is unknown.

Therefore, in this study we set out to (1) characterize the effects of in vivo treatment with sodium arsenite on the vasore-activity in two different blood vessels, i.e., aorta and renal artery of rat, using ex vivo tissue bath method; and (2) clarify whether induction of Hsp is involved in the regulation of vasoreactivity by arsenite.

MATERIALS AND METHODS

Animals and Tissue Preparations

All procedures described below are in accordance with the Animal Experimentation Ethics Guidelines of the National Health and Medical Research Council of Australia and Howard Florey Institute. Male Wistar-Kyoto rats (200 to 300 g) were maintained on standard chow and water ad libitum on a 12-h light/dark cycle. Sodium arsenite was dissolved in normal saline and given at 6 mg kg⁻¹, intravenously (i.v.) (Hauser et al. 1996). Normal saline was used as control. Four hours after injection, rats were sacrificed, and the thoracic aorta and branches of renal arteries (outer diameter of 250 to 350 μm) were isolated in cold Kreb’s physiological salt solution (PSS) containing (in mM) NaCl 118, KCl 4.7, KH₂PO₄ 1.2, MgSO₄ 1.2, NaHCO₃ 25, CaCl₂ 2.5, and D-glucose 11.1.

Vasoreactivity Studies

The aorta was cut into four 3- to 5-mm segments, suspended under 2 g resting tension in organ bath filled with 15 ml PSS (37°C) that was continuously gassed with 5% CO₂/95% O₂. Renal artery rings were studied in a four-chamber wire myograph (Model 610M; J.P. Trading I/S, Denmark) as previously described (Jiang and Dusting 2001). The renal arteries were normalized to 90% of the inner circumference corresponding to 100 mm Hg blood pressure. The isometric tension of the vessel wall was recorded with a MacLab system (AD Instruments, Australia). After 20 to 30 min of equilibration, vessels were contracted with KCl (60 mM) to determine the maximal vaso-constrictor response. After washing, a cumulative concentration response curve to phenylephrine was constructed. To study endothelium-dependent vasorelaxation, vessel rings were pre-contracted with phenylephrine (1 μM for both aorta and renal arteries) and a cumulative concentration response curve to acetylcholine was constructed. Endothelium-independent vasodilation was induced with the NO donor sodium nitroprusside. After each concentration response curve, vessels were thoroughly washed with PSS to allow them to return to baseline tension and were allowed to equilibrate for 20 to 30 min.

Total RNA isolation

Total RNA was extracted using the guanidinium isothiocyanate method (Chomczynski and Sacchi 1987). Briefly, tissues were homogenized in 10 volumes ice-cold denaturing buffer containing 4 M guanidine thiocyanate, 0.025 M sodium citrate, 0.1 M 2-mercaptoethanol, and 0.5% Sarkosyl. Then 0.1 volume of sodium acetate (2 M, pH 4.0), 1 volume of acid phenol, and 0.2 volume of chloroform:isoamylalcohol (49:1) were added. Samples were mixed by inversion and stored on ice for 20 min. After centrifugation at 7500 rpm for 20 min at 4°C, the upper aqueous phase was transferred to a fresh tube, thoroughly mixed with acid phenol and chloforom:isoamylalcohol. This was repeated three times. Following overnight precipitation at −20°C, the RNA pellets were washed three times in RNase-free 70% ethanol, air dried, and resuspended in RNase-free water and stored at −20°C. The yield and purity of total RNA were analyzed using a DU-50 spectrophotometer at wavelengths of 260 nm (RNA) and 280 nm (protein).

Quantification of mRNA by Real-Time Polymerase Chain Reaction (PCR)

Total RNA was reverse-transcribed to cDNA using TaqMan reverse transcription reagents (Applied Biosystems) following the manufacturer’s instructions. The final reverse transcription reaction mix consisted of 1× TaqMan reverse transcription buffer, 5.5 mM MgCl₂, 500 μM dCTP, dATP, dGTP, and dTTP each, 2.5 μM random hexamers, 4 U μl⁻¹ RNase inhibitor, and 1.25 U μl⁻¹ MultiScribe reverse transcriptase. After 10 min of incubation at room temperature, reverse transcription was performed at 48°C for 30 min, followed by 95°C for 5 min. To identify possible genomic DNA contamination, samples transcribed without reverse transcriptase were used as negative control. The real-time PCR reactions were performed in the ABI Prism 7700 system (Applied Biosystems) using dual-labeled probes (5′FAM-3′TAMRA for test genes or 5′VIC-3′TAMRA for 18s housekeeping gene). The final PCR reaction mix (total volume 25μl) contained the optimal cDNA template concentration, 1× TaqMan Universal PCR master mix (Applied Biosystems), and primers and probe at the optimal concentrations. Thermal cycler parameters were 2 min at 50°C, 10 min at 95°C, and 40 cycles of 95°C for 30 s and 60°C for 1 min. Sequences of the forward primer, reverse primer, and probe for different genes are

Hsp72: AGGTTGCATGTTCTTTGCGTT; GGTGGCAGTGCTGAGGTGTT; CAGGAAGGAAACACCATTTTT-ACACAGCTACTTAGATT;

Hsp32: GGATCCTGGACTCACCGTTCTA; TCGTCTGAAGTTGGGAATTGATG; CACTCCGCGCCACCAAGAGGG;

Hsp90: CAGCCAACATGGAACGGATT; CATGTAGCCCATTGTCGAGTTGT; CCGCAGTGCCTGGGCCTTCA;

eNOS: CCGGGACTTCATCAATCAGTACT; TGTGGCTGCCACCTCAG; TCCACTTCCTGAAGCCGCTGCTCATGAGCCTG-GGA;

18s: CGGCTACCACATCCAAGGAA; GCTGGAATTACCGCGGCT; TGCTGGCACCAGACTTGCCCTC.

Drugs

The following drugs were used: acetylcholine perchlorate (BDH Chemicals), arsenite solution (Merck), N ^G-nitro-l-arginine methyl ester hydrochloride (l-NAME; Sigma), l-phenylephrine hydrochloride (Sigma), sodium nitroprusside (Sigma).

Data Analysis

Contractions to phenylephrine were expressed as percentages of the maximal contractions to 60 mM KCl. Vasodilations to acetylcholine were expressed as percentage reductions of phenylephrine-induced tone. The pEC₅₀ (which is the negative logarithm of the concentration required to produce 50% of the maximal response) values were calculated using Graph-Pad Prism software. Data are expressed as mean ± standard error of the mean (SEM). The mean data were analyzed with one-way analysis of the variance (one-way ANOVA) followed by Newman-Keuls t test. A value of p < .05 was regarded as statistically significant.

RESULTS

Vasoconstrictor Responses Were Enhanced by Arsenite

In the aortic rings with intact endothelium, contractions induced by 60 mM KCl were not altered by arsenite, being 25.1 ± 1.9 mN in control (n = 7) and 21.4 ± 1.7 mN (p > .05) in arsenite-treated animals (n = 8). Contractions induced by phenylephrine (10⁻⁹ to 10⁻⁵ M) were significantly increased by arsenite treatment (Figure 1). Values for pEC₅₀ and maximal response (R _max) are shown in Table 1. In rat renal arteries with in-tact endothelium, similarly, KCl-induced contractions were not significantly altered by arsenite (17.5 ± 1.5 mN in control and 15.8±0.9 mN in arsenite-treated animals, n = 7 and 6, p > .05), whereas the contractions induced by phenylephrine (10⁻⁸ to 10⁻⁴ M) were significantly increased by arsenite (Figure 1 and Table 1).

Effects of Arsenite Depended on the Presence of Basal NO

In aortic rings pretreated with the NOS inhibitor l-NAME (100 μM) in vitro, 60 mM KCl-induced contractions were not different between saline and arsenite-treated animals (24.1 ± 1.9 mN and 20.9 ± 2.1 mN, respectively, n = 7 and 6, p > .05), whereas the difference in phenylephrine-induced contractions between saline and arsenite-treated animals was abolished by l-NAME (Figure 1 and Table 1). In renal arteries pretreated with l-NAME, 60 mM KCl–induced contractions were not different between the two groups (16.4 ± 1.1 mN in saline- and 15.4 ± 1.0 mN in arsenite-treated animals, n = 6, p > .05). Consistent with the findings in aorta, the difference in phenylephrine-induced contractions between saline and arsenite-treated animals was abolished by l-NAME (Figure 1 and Table 1). We also compared the effects of l-NAME in vessels from saline and arsenite-treated animals. In the aorta, l-NAME significantly increased the contractions to phenylephrine in the saline group, whereas it had no effect in the arsenite group (Table 1). In the renal arteries, l-NAME significantly increased phenylephrine-induced contractions in vessels from both the saline and arsenite group (Table 1). Contractions to 60 mM KCl were not changed by l-NAME in none of the above situations.

Endothelium-Dependent Relaxation Responses

In aortic rings precontracted with phenylephrine (1 μM), acetylcholine (10⁻⁹ to 10⁻⁵ M) produced endothelium-dependent vasodilations in a concentration-dependent manner. This response was abolished by l-NAME (100 μM). There was no difference in acetylcholine-induced vasodilations between saline- and arsenite-treated animals (Figure 2). The endothelium-independent vasodilations induced by the NO donor sodium nitroprusside (10⁻⁹ to 10⁻⁵ M) were not affected by arsenite either (data not shown). Unlike those in aorta, the endothelium-dependent vasodilations induced by acetylcholine (10⁻⁹ to 3 × 10⁻⁶ M) in the renal arteries precontracted with phenylephrine (1 μM) were not affected by l-NAME treatment. Arsenite had little effect on this l-NAME-resistant response in renal arteries (Figure 2).

Previously we have demonstrated that in small renal arteries from normal Wistar-Kyoto rats, acetylcholine-induced vasodilations are largely dependent upon hyperpolarization mechanisms that are initiated in the endothelium but do not depend upon NO release, and this response can be abolished by l-NAME plus high-K⁺ (60 mM) PSS (Jiang and Dusting 2001). Therefore we examined acetylcholine-induced vasodilations in renal arteries that were precontracted with high-K⁺ PSS, which eliminates the hyperpolarizing component in these vessels. In renal arteries from both saline- and arsenite-treated rats, acetylcholine-induced vasodilations were similarly reduced in K⁺-contracted tissues as compared to those in phenylephrine-contracted tissues, and were completely abolished by l-NAME (100 μM). In K⁺-contracted tissues, there was not a significant difference in acetylcholine-induced vasodilations between the saline and arsenite groups (Figure 3).

Effects of Arsenite on Vasoreactivity Did Not Depend on Hsp Expression

Treatment with arsenite significantly increased mRNA expression of Hsp72, Hsp32, and Hsp90 in the rat aorta (Figure 4a ), whereas the eNOS mRNA expression was not altered (data not shown). To investigate the role of Hsp induction in the effects of arsenite on vasoreactivity, we cotreated the rats with arsenite plus quercetin, which has been shown to suppress heat- and arsenite-induced Hsp expression (Hosokawa et al. 1990; Koishi et al. 1992; Elia and Santoro 1994). We found that concomitant treatment with quercetin suppressed the induction of Hsp by arsenite in the aorta (Figure 4a ), but quercetin had little effect on arsenite-induced increase of the vasoconstriction responses to phenylephrine (Figure 4b ).

DISCUSSION

Previous studies did not provide unequivocal information about how arsenite exposure of vascular tissues modulates adrenoreceptor-mediated vasoconstrictions (Caudill et al. 1998; Knoepp et al. 2000). In this study we demonstrated that in vivo administration of arsenite significantly increased phenylephrine-induced vasoconstrictor responses ex vivo in both large conduit (aorta) and small muscular (renal arteries) vessels. In both aorta and renal arteries from control animals, inhibition of basal en-dothelial NO release with l-NAME significantly increased the contractions to phenylephrine, indicating that basal endothelial NO release tonically modulates vascular tone in response to adrenoreceptor agonists in both vessels. However, in tissues pre-treated with l-NAME, the difference in phenylephrine-induced contractions between saline and arsenite-treated animals was abolished, indicating that the increase in the contraction to phenylephrine was NO-dependent. Moreover, l-NAME failed to increase phenylephrine-induced contractions in the aorta from arsenite-treated animals, in contrast to its effect in control tissues. Taken together these results suggest that arsenite-induced hyperreactivity of vascular tissues to phenylephrine is associated with an attenuation of the basal NO availability in the arteries.

In contrast to the study by Lee and colleagues (Lee et al. 2003), we did not observe an obvious alteration of endothelium-dependent vasodilation induced by acetylcholine in either aorta or the renal artery. It was noted that in renal arteries from both control and arsenite-treated animals, acetylcholine-induced vasodilations were not affected by NOS inhibition with l-NAME, but were abolished by the combination of l-NAME and high-K⁺ PSS, of which the latter readily eliminates smooth muscle hyperpolarization-mediated relaxation (Vanhoutte 2004). In K⁺-contracted vessels, the vasodilations to acetylcholine were considerably reduced as compared to phenylephrine-contracted tissues. These results suggest that in the renal circulation acetylcholine-induced arterial relaxation was largely dependent on an endothelium-derived hyperpolarizing mechanism (Jiang and Dusting 2001), whereas the role of endothelial NO appears to be redundant in this response. In the presence of l-NAME, there was no difference between the vasodilation in control and arsenite-treated animals, suggesting that arsenite had minimal effect on endothelium-derived hyperpolarizing factor–mediated responses (Vanhoutte 2004). On the other hand, acetylcholine-induced vasodilation in the aorta was abolished by l-NAME, demonstrating that NO is the predominant mediator. We speculate that the discrepancy in the effects of arsenite on endothelium-dependent vasodilation in the aorta between the study by Lee et al. (2003) and ours may be because of the difference in the duration of arsenite treatment (14 versus 4 h), and effects of long-term arsenite exposure on vascular functions in vivo warrant further studies. We did not, however, measure the blood concentration of arsenite obtained after intravenous administration, thus it is difficult to directly compare the local arsenite concentrations achieved in vivo with those used in vitro. Nonetheless, these data clearly demonstrated that exposure of blood vessels to arsenite impaired endothelial NO bioavailability. Moreover, given that NO is an important endogenous factor with multiple vascular protective actions, endothelial NO dysfunction may represent a mechanistic link between arsenic intoxication and the observed increase in vascular disease risks.

Possible mechanisms that are responsible for the compromised endothelial NO release are implied by several lines of evidence. Firstly, acute arsenite treatment (<2 h) in endothelial cells increased superoxide generation (Barchowsky et al. 1999a), and this may be caused by activation of NADPH oxidase (Lynn et al. 2000; Smith, Klei, and Barchowsky 2001), which has been suggested to be a major source of superoxide in vascular tissues and have critical roles in normal cell physiology and in development of a variety of vascular diseases (Griendling, Sorescu, and Ushio-Fukai 2000; Jiang, Drummond, and Dusting 2004). Secondly, arsenite may have a direct inhibitory effect on eNOS (Lee et al. 2003). However, this action is unlikely to be involved in our observations because in neither aorta precontracted by phenylephrine nor renal arteries precontracted by K⁺, arsenite had any effect on acetylcholine-induced vasodilations. Moreover, we found that eNOS expression was not changed by arsenite treatment. Thirdly, it was reported that chronic (18 weeks) ingestion of arsenic in rabbits resulted in a significant reduction of systemic basal NO release, caused by enzymatic uncoupling of NOS with a subsequent enhancement of reactive oxygen species (Pi et al. 2003). These authors demonstrated that hepatic levels of tetrahydrobiopterin (BH₄), a cofactor for NOS, were markedly reduced in arsenate-exposed rabbits. However, the importance of this mechanism in acute arsenic treatment as in the present study is unclear. Fourthly, there is evidence that arsenite may induce endothelial cell apoptosis (Yeh et al. 2003).

We found that following arsenite injection, the mRNA of Hsp72, Hsp32, and Hsp90 were markedly elevated in vascular tissues, indicating a global Hsp induction, which may interact with NO signaling in different ways. For example, Hsp90 has been shown to be closely involved eNOS function (Garcia-Cardena et al. 1998). There is evidence that disruption of the interaction between heat shock protein and eNOS blocks acetylcholine-induced vasodilation of isolated rat aorta (Garcia-Cardena et al. 1998). On the other hand, Hsp27 has been shown to reduce the responsiveness of smooth muscle to NO (Knoepp et al. 2000). Moreover, Hsp70 may enhance inducible NOS (iNOS) expression in response to cytokines and increase subsequent NO release (Bellmann et al. 2000). In the present study, however, the effects of arsenite on vasoreactivity were unlikely to be mediated by Hsp induction, since coadministration with quercetin, which nearly abolished Hsp induction by arsenite, had little effect on the increased vasoconstriction in vessels from arsenite-treated animals.