Nitric oxide production contributes to Bacillus anthracis edema toxin-associated arterial hypotension and lethality: ex vivo and in vivo studies in the rat

The most important aspect of the present study is the finding that the hypotensive and lethal effects of Bacillus anthracis edema toxin are mediated in part via the production of nitric oxide (NO) and that coadministration of an NO synthase inhibitor with edema toxin challenge has a beneficial survival effect.

Abstract

We showed previously that Bacillus anthracis edema toxin (ET), comprised of protective antigen (PA) and edema factor (EF), inhibits phenylephrine (PE)-induced contraction in rat aortic rings and these effects are diminished in endothelial-denuded rings. Therefore, employing rat aortic ring and in vivo models, we tested the hypothesis that nitric oxide (NO) contributes to ET's arterial effects. Compared with rings challenged with PA alone, ET (PA + EF) reduced PE-stimulated maximal contractile force (MCF) and increased the PE concentration producing 50% MCF (EC₅₀) (P < 0.0001). Compared with placebo, l-nitro-arginine methyl-ester (l-NAME), an NO synthase (NOS) inhibitor, reduced ET's effects on MCF and EC₅₀ in patterns that approached or were significant (P = 0.06 and 0.03, respectively). In animals challenged with 24-h ET infusions, l-NAME (0.5 or 1.0 mg·kg⁻¹·h⁻¹) coadministration increased survival to 17 of 28 animals (60.7%) compared with 4 of 27 (14.8%) given placebo (P = 0.01). Animals receiving l-NAME but no ET all survived. Compared with PBS challenge, ET increased NO levels at 24 h and l-NAME decreased these increases (P < 0.0001). ET infusion decreased mean arterial blood pressure (MAP) in placebo and l-NAME-treated animals (P < 0.0001) but l-NAME reduced decreases in MAP with ET from 9 to 24 h (P = 0.03 for the time interaction). S-methyl-l-thiocitrulline, a selective neuronal NOS inhibitor, had effects in rings and, at a high dose in vivo models, comparable to l-NAME, whereas N′-[3-(aminomethyl)benzyl]-acetimidamide, a selective inducible NOS inhibitor, did not. NO production contributes to ET's arterial relaxant, hypotensive, and lethal effects in the rat.

NEW & NOTEWORTHY

The most important aspect of the present study is the finding that the hypotensive and lethal effects of Bacillus anthracis edema toxin are mediated in part via the production of nitric oxide (NO) and that coadministration of an NO synthase inhibitor with edema toxin challenge has a beneficial survival effect.

bacillus anthracis edema toxin (ET) contributes to the pathogenesis of shock and lethality occurring with anthrax infection. Infusion of ET in healthy animals produces hypotension and lethality while ET inhibition improves survival in B. anthracis infected animal models (10, 33, 49). ET is a binary type toxin comprised of protective antigen (PA), the component necessary for host cell uptake, and edema factor (EF), the toxin's toxic moiety (50). Edema factor is a calmodulin-dependent adenyl-cyclase that increases intracellular cAMP levels (27). Understanding the mechanisms underlying ETs pathogenic effects may improve the management of anthrax infection and shock in the future.

We previously hypothesized that ET-mediated increases in intracellular cAMP levels would produce arterial relaxation, a potential basis for this toxin's hypotensive and lethal effects. In experiments we showed that ET inhibited phenylephrine (PE)-stimulated rat aortic ring contraction and reduced the force of rings already contracted with PE (30). ET stimulated cAMP production in aortic tissue, and adefovir, a nucleoside that blocks cAMP production by EF, inhibited both the arterial relaxant effects of ET as well as its associated cAMP increases. Together these findings supported the possibility that ET-stimulated cAMP production contributes to the hypotension this toxin produces in in vivo models and to hypotension during anthrax infection.

Of note in our prior experiments, the effects of ET on decreasing the maximal contractile force (MCF) aortic rings generated in response to PE and on decreasing the sensitivity of rings to PE were significantly reduced in rings in which the endothelium had been removed (i.e., denuded rings) (30). These findings suggested that ET's inhibition of arterial contractile function might be mediated in part by release of an endothelial derived relaxant factor (EDRF). Nitric oxide (NO) is a potent EDRF, which has been associated with the pathogenic effects of several bacterial toxins (4, 6, 37). Therefore, using both an ex vivo rat aortic ring model and an in vivo rat model, in combination with three nitric oxide synthase (NOS) inhibitors with the potential to inhibit one or more of the three NOS isoforms [endothelial (eNOS), neuronal (nNOS), and inducible (iNOS)], we tested the hypothesis that NO production contributes to ET's hypotensive and lethal effects. The three NOS inhibitors studied were as follows: l-nitro-arginine methyl ester (l-NAME), often described as a nonselective NOS inhibitor; S-methyl-l-thiocitrulline (SMTC), a selective nNOS inhibitor; and N′-[3-(aminomethyl)benzyl] acetimidamide (1400W), a selective iNOS inhibitor (2, 16, 19, 46). We also compared the effects of ET on the contractile force of rings isolated from wild-type (WT) vs. eNOS knockout (eNOS-KO) mice.

MATERIALS AND METHODS

Animal care.

This study was reviewed and approved by the Animal Care and Use Committee of the Clinical Center of the National Institutes of Health.

Assay of PE-stimulated rat aortic ring contraction following treatment with ET and NOS inhibitors.

On each experiment day, four aortic rings were prepared (see methods below) from Sprague-Dawley rats and then incubated for 60 min with either ET (PA + EF) or PA challenge alone, combined with placebo or a NOS inhibitor, either l-NAME, SMTC, or 1400W. Daily experiments included at least one PA with placebo and one ET with placebo ring to serve as controls. As placebo, rings not treated with an inhibitor were incubated with the same concentration and quantity of diluent in which the inhibitor being studied the same day was administered in. After incubation and following washing and equilibration (30 min), the contractile force of each ring to increasing concentrations of PE (10⁻⁹ to 10⁻⁵ M) was measured. Concentrations of ET employed in these studies (i.e., EF 800 or 1,200 ng/ml combined with PA 1,600 or 2,400 ng/ml) were comparable to ones we employed previously (30). The concentrations investigated of l-NAME (0.03, 0.3, 3.0, and 30.0 μM), SMTC (25, 50, and 100 μM), and 1400W (1, 10, or 100 μM) were based on previously reported dosing studies or on pilot studies during the present investigation (16, 19, 46). Four studies were performed. Study 1 investigated the effect of increasing l-NAME doses in ET-challenged rings only. To account for the potential spontaneous release of NO during the contraction of intact rings (i.e., release not related to ET itself), studies 2, 3, and 4 compared the effects of l-NAME, SMTC, or 1400W, respectively, to placebo in both PA- and ET-challenged rings (see Statistical analysis).

In daily experiments, aortic rings were prepared and their baseline contractile force was tested as previously described (30). Male, 8- to 10-wk-old Sprague-Dawley rats weighing 250 g on average were anesthetized with isoflorane and their thoracic aortas were isolated, removed, and cleaned of adventitia and fat (3, 7, 44). Cross sections of aorta 4 mm in length (aortic rings) were cut and mounted between two tungsten hooks in baths containing modified Krebs-Henseleit (KH) buffer (21) at 37°C and continuously bubbled with 95% O₂-5% CO₂ at pH 7.4. Suspended rings were connected to an isometric transducer and their contractile responses were recorded on a polygraph (Lab Chart; ADInstruments, Colorado Springs, CO). Rings were equilibrated until their resting tension stabilized at 1.5 g. After equilibration, the functional integrity of each ring was assessed by measuring its contractile response to 10⁻⁷M phenylephrine (PE) followed by its relaxation with 10⁻⁵M acetylcholine (ACH). A tension increase of ≥1 g to PE and subsequent relaxation of >90% to ACH were considered evidence of a ring's functional integrity. Rings not meeting these criteria were not studied. After a subsequent 30-min equilibration period, each ring's contractile response was measured by exposure for 4 min to 60 mM KCl. This measure was performed three times and the highest value was determined to be the ring's peak contractile force. In subsequent PE dose-response experiments, the MCF of each ring was calculated as the percentage of its peak contraction with 60 mM KCl. The effective dose of PE producing 50% of the MCF was designated the half-maximal effective PE concentration (EC₅₀, × 10⁻⁸M). MCF and EC₅₀ were calculated by standard curve analysis with SigmaPlot software (Systat Software, San Jose, CA).

Assays of lethality, NO production and hypotension in rats treated with ET and NO inhibitors.

Each week, 12 male, 8- to 10-wk-old Sprague-Dawley rats weighing 250 g on average, with indwelling carotid arterial and jugular venous catheters, were first assigned blindly using previously prepared assignment cards, to receive challenge with either ET (i.e., EF combined with PA in PBS diluent) or PBS only as 24-h continuous infusions. Assignment was done in a 2 to 1 ratio to ET and PBS groups, respectively, to account for the mortality expected with ET. Animals were then again assigned blindly to treatment with a dose of a particular NOS inhibitor or placebo, as a 24-h infusion concurrent with ET or PBS challenge. This design allowed us to account for any effect a NOS inhibitor might have on constitutive NO release not related to ET itself (see Statistical analysis). Three studies consisted of groups of experiments devoted to the investigation of differing doses of the same NOS inhibitor (study 1 for l-NAME, study 2 for SMTC, and study 3 for 1400W). The dose of ET employed in each study (i.e., EF 400 μg/kg combined with PA 800 μg/kg in 24 h) was the same as one shown to produce a >50% lethality rate in our prior studies. The doses investigated of l-NAME (0.5 and 1.0 mg·kg⁻¹·h⁻¹), SMTC (0.3, 0.9, or 1.5 mg·kg⁻¹·h⁻¹), or 1400W (0.175, 0.525, or 1.575 mg·kg⁻¹·h⁻¹) administered over 24 h were based on previously reported dosing or on pilot studies performed in the present investigation (13, 18, 19, 25). As placebo, animals not treated with an inhibitor received the same volume of PBS diluent in which the corresponding NOS inhibitor was administered.

In weekly experiments, measures were obtained as previously described (10, 29, 41). At the start of experiments, protected catheters were attached to exteriorized arterial and central venous access ports on each animal (10, 29, 41). Central venous catheters were attached via three-way stopcocks to a syringe pump to provide ET or PBS and NOS inhibitors or placebo as infusions. Arterial catheters were connected to transducers to determine arterial blood pressure and for blood sampling. Every 1 h during the 24-h challenges and treatments, animals had mean arterial blood pressure (MAP) measured. At 4, 8, or 24 h after the start of the infusions, animals had 0.5 ml of blood drawn for cAMP levels (studies with l-NAME only) and NO levels as previously described (30, 41). Withdrawn blood was replaced with a similar volume of normal saline. After the 24-h challenge and treatment period, animals were disconnected from monitors and observed for an additional 144 h (168 h total from the start of study) and animals alive at the end of this period were considered survivors.

cAMP and NO assays.

Plasma cAMP was measured with a cAMP Chemiluminescent Immunoassay Kit (Arbor Assays, Ann Arbor, MI). Plasma nitrite/nitrate levels were measured with a fluorometric assay kit (Cayman Chemical, Ann Arbor, MI).

Assay of PE-stimulated aortic ring contraction following treatment with ET comparing rings from WT and eNOS-KO mice.

Aortic rings were prepared from eNOS gene knockout (eNOS-KO) mice (n = 7, B6.129P2-Nos3^tm1Unc/J; stock no. 002684; Jackson Laboratory) and WT animals (n = 7; C57BL/6J; Jackson Laboratory). Methods for the isolation and preparation of these mouse rings were similar to those described above for studies in rats. On each experiment day, two aortic rings were prepared from a WT animal and two from an eNOS-KO animal. Then, one ring from each type of animal was treated with ET (EF 3,200 ng/ml and PA 6,400 ng/ml) and the other with PA alone. Following treatment with ET or PA, rings were challenged with increasing concentrations of phenylephrine and contractile force recorded, also as described above.

Reagents.

Toxin components (PA and EF) were prepared from recombinant protein components. For the aortic ring studies, these components were derived from Escherichia coli (8, 10, 11, 26). For the in vivo studies, components were derived B. anthracis (EF-A; Ref. 28). The lipopolysaccharide (LPS) contents of the E. coli derived PA and EF were 0.001 and 0.006 ng/μg, respectively. In the B. anthracis toxin preparations, LPS activity was undetectable. ET was comprised of EF with PA in ratios of 1:2 on the basis of weight. The three NOS inhibitors [l-NAME (Sigma-Aldrich, St. Louis, MO); SMTC (Sigma Aldrich); and 1400W (Sigma-Aldrich; Cayman Chemical, Ann Arbor, MI)] were all mixed in PBS for aortic ring and in vivo studies.

Statistical analysis.

All continuous data were analyzed using PROC Mixed in SAS Version 9.3 software (SAS Institute, Cary, NC). In aortic ring experiments, MCF and EC₅₀ measures were analyzed with two-way ANOVA accounting for treatment (NOS inhibitor or placebo) and challenge (ET or PA) to determine the effect of the NOS inhibitors, or one-way ANOVA to compare any two groups in each study or any two doses of l-NAME in the dose-response study (study 1 in the ex vivo aortic ring studies). In in vivo studies, a Cox proportional hazard model was used to analyze the effect of treatment on survival accounting for treatment dose, final outcome. and survival time. Treatment doses were combined for analysis when the dose effect was not statistically different to increase the power of the analysis and limit the number of animals required for study. The survival effect was expressed as the hazards ratio of survival (95% confident interval). For laboratory parameters, a three-way repeated-measures ANOVA was used to analyze the effect of the NOS inhibitors accounting for treatment, challenge, and time point, or one-way ANOVA to compare any two groups at individual time points or periods. Since the effect of l-NAME on MAP varied by time after ET or PBS challenge and to parallel the time points laboratory tests were performed at (4, 8, and 24 h after the start of challenge and treatment), MAP data were analyzed from 1 to 4 h, 5 to 8 h, and 9 to 24 h in each of the three in vivo studies. To investigate the effects of ET on NO-mediated events independent of the spontaneous or constitutive release of NO in either aortic rings or in vivo studies respectively, the effects of ET in placebo-treated rings or rats (i.e., ET with placebo treatment minus PA or PBS control with placebo treatment) were compared with the effects of ET in NOS inhibitor-treated rings or rats (i.e., ET with NOS inhibitor treatment minus PA or PBS control with NOS inhibitor treatment). Relationships between ET-associated NO levels and either blood pressure levels in animals overall all or survival times in nonsurvivors alone, were investigated with linear regression. Throughout, data are presented as mean ± SE. P ≤ 0.05 was considered significant.

RESULTS

Effects of l-NAME in ex vivo and in vivo studies.

An initial ex vivo study using the aortic ring model (ex vivo-study 1) examined the effect of increasing l-NAME doses on the arterial relaxant effects of ET. In aortic rings challenged with PA without EF and treated with placebo, stimulation with increasing PE concentrations produced a mean (±SE) maximal contractile force (MCF) of 87.4 ± 7.1 while the effective concentration of PE producing 50% of the MCF (EC₅₀, × 10⁻⁸ M) was 10.5 ± 5.2 (Fig. 1, A–C). Compared with PA and placebo, ET challenge with placebo treatment, decreased the MCF (56.2 ± 7.1) and increased the EC₅₀ (40.4 ± 5.2; P ≤ 0.003 for the comparisons). In ET-challenged rings, compared with placebo, l-NAME doses of 0.3, 3, and 30 μM, but not 0.03 μM, each significantly increased MCF (81.2 ± 10.0, 98.6 ± 9.2, and 102.0 ± 10.9, respectively, P ≤ 0.05), while l-NAME doses 3 and 30 μM, but not 0.03 or 0.3 μM, each significantly decreased the EC₅₀ (18.3 ± 6.8 and 16.3 ± 8.0) (P = 0.01).

Fig. 1.

This figure compares the mean (±SE) contractile force aortic rings generated during stimulation with increasing phenylephrine (PE) concentrations (A), the mean (±SE) maximal contractile force (MCF) developed during PE stimulation (B), and the mean (±SE) estimated concentration of PE producing 50% of the MCF (EC₅₀; C) after pretreatment with protective antigen (PA) (1,600 ng/ml) combined with placebo or with edema toxin (ET) (800 ng/ml of edema factor and 1,600 ng/ml of PA) combined with either placebo or l-nitro-arginine methyl-ester (l-NAME) in increasing concentrations of 0.03, 0.3, 3, or 30 μM. Contractile force in A is shown as the percentage of the force rings generated after exposure to 60 mM KCl during initial assessment of the rings. The number of rings employed to calculate MCF and EC₅₀ is shown under B and C. The brackets and P values demonstrate the levels of significance for comparisons of respective groups of rings.

To further examine the influence of ET on NO-induced contractile dysfunction while controlling for the possible release of NO due to ring contraction with PE stimulation, a second ex vivo study (ex vivo-study 2) examined treatment with placebo or l-NAME in both PA- and ET-challenged rings. The effects of the two l-NAME doses tested (3 and 30 μM) did not differ significantly (P ≥ 0.26) and these were combined in analysis. Compared with PA-challenged rings, in both placebo and l-NAME-treated rings, ET decreased MCF (P < 0.0001 for both) and increased EC₅₀ (P < 0.0001 for both; Fig. 2, A and B). Compared with placebo, in both PA and ET-challenged rings, l-NAME increased MCF (P = 0.004 and P < 0.0001, respectively) and decreased EC₅₀ (P = 0.03 and <0.0001, respectively) (Fig. 2, A and B, and Table 1). After we controlled for the effect of l-NAME on NO release due to PE-stimulated muscle contraction alone in PA-challenged rings (see materials and methods), compared with placebo, l-NAME reduced the effect of ET on decreasing MCF (−40.3 ± 5.0 vs. −27.1 ± 4.9) and on increasing EC₅₀ (19.4 ± 2.4 vs. 12.0 ± 2.3) in patterns that approached or were significant (P = 0.06 and 0.03 respectively) (Fig. 2, C and D).

Fig. 2.

This figure shows the mean (±SE) maximal contractile force (MCF) rings developed during phenylephrine (PE) stimulation (A) and the mean (±SE) estimated concentration of PE producing 50% of the MCF (EC₅₀; B) after pretreatment with protective antigen (PA) (1,600 ng/ml) or edema toxin (ET) (800 ng/ml of edema factor and 1,600 ng/ml of PA), each combined with either placebo or l-nitro-arginine methyl-ester (l-NAME) (data for l-NAME doses of 3 or 30 μM combined, see results). C and D: mean effects of ET on MCF and EC₅₀ in placebo vs. l-NAME-treated rings (see materials and methods). The bars in C and D reflect the differences between placebo and ET denoted by the arrows in A and B. The number of rings employed to calculate MCF and EC₅₀ is shown under A and B. The brackets and P values demonstrate the levels of significance for comparisons of respective groups of rings.

Table 1.

Effects of l-NAME, SMTC, and 1400W on MCF and EC₅₀ in PA- or ET-challenged aortic rings

	Treatment
Study/Challenge	Type	Doses combined, μM	MCF, %	EC50, ×10−8 M
Study 1
PA	l-NAME	3, 30	15.5 ± 4.5	−5.3 ± 1.0
ET	l-NAME	3, 30	28.5 ± 6.1	−12.6 ± 3.4
Study 2
PA	SMTC	25, 50, 100	9.2 ± 3.6	−6.2 ± 1.9
ET	SMTC	25, 50, 100	12.5 ± 5.3	−19.6 ± 5.9
Study 3
PA	1400W	1, 10, 100	17.8 ± 8.6	−2.7 ± 1.1
ET	1400W	1, 10, 100	23.9 ± 10.0	−14.9 ± 8.5

Values are means ± SE.

PA, protective antigen; ET, edema toxin; l-NAME, l-nitro-arginine methyl-ester; SMTC, S-methyl-l-thiocitrulline; 1400W, N′-[3-(aminomethyl)benzyl]-acetimidamide; MCF, maximal contractile force; EC₅₀, effective phenylephrine concentration producing 50% of the MCF.

An in vivo study (in vivo-study 1) then investigated the effects of l-NAME treatment in doses of 0.5 or 1.0 mg·kg⁻¹·h⁻¹ on survival and on cAMP, NO, and blood pressure levels in PBS- or ET-challenged animals. All animals challenged with 24-h PBS infusion and treated with either placebo or l-NAME survived for 168 h (4 animals each receiving placebo or l-NAME at 0.5 mg·kg⁻¹·h⁻¹ and 10 animals each receiving placebo or l-NAME at 1.0 mg·kg⁻¹·h⁻¹). In animals challenged with 24-h ET infusions, compared with placebo, l-NAME at 0.5 mg·kg⁻¹·h⁻¹ increased the number of animals surviving [1 survivor of 8 total (12.5%) vs. 5 of 8 (62.5%)] as did l-NAME at 1.0 mg·kg⁻¹·h⁻¹ [3 of 19 (16.7%) vs. 12 of 20 (58.3%)]. The effects of the low and high l-NAME doses on increasing the hazards ratios of survival (95% confidence interval) [2.67 (0.68, 10.46), P = 0.14 and 2.51 (1.13, 5.59), P = 0.045] did not differ significantly (P = 0.97). Across the two doses, l-NAME increased survival significantly with ET challenge [2.52 (1.27, 5.00), P = 0.01] (Fig. 3A). In subsequent analysis, to increase the power to identify effects related to l-NAME and limit animal use, data with the two l-NAME doses were combined.

Fig. 3.

A: the proportion of animals surviving following the start of 24-h challenges with PBS or edema toxin (ET) and concurrent treatment with 24-h infusions of either placebo or l-nitro-arginine methyl-ester (l-NAME). B and C: the mean (±SE) plasma cAMP and nitric oxide (NO) levels, respectively, at 4, 8, or 24 h in animals after the start of 24-h challenges with PBS or edema toxin (ET) and concurrent treatment with 24-h infusions of placebo or l-nitro-arginine methyl-ester (l-NAME). For all parameters, data are shown for l-NAME doses of 0.5 and 1.0 mg·kg⁻¹·h⁻¹ combined (see results).

Compared to PBS challenge, ET increased cAMP levels (P < 0.0001 for the overall effect of ET). Compared with placebo, l-NAME did not alter cAMP levels significantly in either PA- or ET-challenged rats (P = 0.70 and 0.55 for the overall effects, respectively; Fig. 3B). Different from cAMP, compared with PBS challenge, ET increased NO levels at 24 h but not earlier (P = 0.001 for the time interaction; Fig. 3C). Also different from cAMP, compared with placebo, while l-NAME treatment did not alter NO levels measurably with PBS challenge at any time point (P ≥ 0.35 at each time point), it decreased these with ET challenge at 4 and 24 h (P = 0.05 and P < 0.0001, respectively) and in a pattern that increased over time (P = 0.001 for the time interaction). Finally, compared with PBS, 24-h ET challenges progressively decreased mean arterial blood pressure (MAP) in animals receiving placebo and l-NAME (P < 0.0001 for the time interactions for each) (Fig. 4A). Compared with placebo, l-NAME treatment progressively increased MAP with both PBS and ET challenge (P = 0.05 and P < 0.0001 for the time interactions, respectively). After we controlled for the possible effect of constitutive NO release on MAP in PBS-challenged animals (see materials and methods), compared with placebo, l-NAME, while not significantly altering the effects of ET on decreasing MAP from 0 to 4 h or 5 to 8 h (P ≥ 0.32), progressively reduced these decreases from 9 to 24 h (P = 0.03 for the time interaction) although these changes were small ones (Fig. 4B).

Fig. 4.

A: the serial mean (±SE) hourly mean arterial blood pressures (MAP) in animals beginning immediately before and then after the start of 24-h challenges with PBS or edema toxin (ET) and concurrent treatment with 24-h infusions of placebo or l-nitro-arginine methyl-ester (l-NAME) (data shown for l-NAME doses of 0.5 and 1.0 mg·kg⁻¹·h⁻¹ combined, see results). B: comparison of the mean effects of ET on MAP in placebo vs. l-NAME-treated animals (see materials and methods) plotted from a value of 0 immediately before challenge and treatments were initiated (MAP did not differ across the 4 groups at time 0, P = ns for all comparisons). Comparisons of the effects of ET with placebo vs. l-NAME were done from 0 to 4 h, 5 to 8 h, and 9 to 24 h to correspond to the time points at which plasma NO levels were obtained (i.e., 4, 8, and 24 h).

Effects of SMTC in ex vivo and in vivo studies.

In an ex vivo study (ex vivo-study 3), we next examined the effects of SMTC on ET's arterial relaxant effects. The effects of the three doses of SMTC tested (25, 50, and 100 μM) on MCF and EC₅₀ did not differ in either PA- or ET-challenged rings (P ≥ 0.16) and these doses were combined for analysis. Compared with PA-challenged rings, in both placebo- and SMTC-treated rings, ET decreased MCF (P < 0.0001 for both) and increased EC₅₀ (P < 0.0001 for both) (Fig. 5, A and B). Compared with placebo, in both PA- and ET-challenged rings, SMTC increased MCF (P = 0.02 and P = 0.001, respectively) and decreased EC₅₀ (P = 0.05 and P < 0.0001, respectively) (Fig. 5, A and B, and Table 1). After we controlled for the effect of SMTC on NO release due to PE-stimulated muscle contraction alone in PA-challenged rings (see materials and methods), compared with placebo, SMTC, while not significantly altering ET's effect on MCF (−24.9 ± 3.7 vs. −21.5 ± 3.7, P = 0.52) did reduce ET's effect on increasing EC₅₀ (26.9 ± 3.0 vs. 13.4 ± 3.0, P = 0.003; Fig. 5, C and D).

Fig. 5.

This figure shows the mean (±SE) maximal contractile force (MCF) rings developed during PE stimulation (A), and the mean (±SE) estimated concentration of PE producing 50% of the MCF (EC₅₀; B) after pretreatment with protective antigen (PA) (1,600 ng/ml) or edema toxin (ET) (800 ng/ml of edema factor and 1,600 ng/ml of PA) combined with either placebo or S-methyl-l-thiocitrulline (SMTC) (data for SMTC doses 25, 50 and 100 μM were combined, see results). C and D: comparison of the mean effects of ET on MCF and EC₅₀ in placebo vs. SMTC-treated rings (see materials and methods). The bars in C and D reflect the differences between placebo and ET denoted by the arrows in A and B. The number of rings employed to calculate MCF and EC₅₀ is shown under A and B. The brackets and p-values demonstrate the levels of significance for comparisons of respective groups of rings.

An in vivo study (in vivo-study 2) then investigated the effects of SMTC treatment in doses of 0.3, 0.9, or 1.5 mg·kg⁻¹·h⁻¹ on survival and on NO and blood pressure levels in ET- or PBS-challenged animals. All animals challenged with 24-h PBS infusions and treated with either placebo or any of the three SMTC doses survived for 168 h. In animals challenged with 24-h ET infusions, compared with placebo, SMTC doses of 0.3 and 0.9 mg·kg⁻¹·h⁻¹ did not significantly alter survival [7 survivors of 12 (58.3%) with placebo vs. 5 survivors of 11 (45.5%) with SMTC at 0.3 mg·kg⁻¹·h⁻¹, P = 0.59; 1 survivor of 11 (9.1%) with placebo vs. 1 survivor of 12 (8.3%) with SMTC at 0.9 mg·kg⁻¹·h⁻¹ SMTC, P = 0.91] even when combined (P = 0.78). However, SMTC at 1.5 mg·kg⁻¹·h⁻¹ did increase survival [2 survivor of 19 (10.5%) with placebo vs. 9 of 20 with SMTC (45.0%)] and significantly increased the hazards ratio of survival 2.63 [(1.22, 5.67); P = 0.01; Fig. 6A]. Subsequent analysis examined the effect of SMTC at 1.5 mg·kg⁻¹·h⁻¹ on NO and blood pressure levels.

Fig. 6.

A: the proportion of animals surviving following the start of 24-h challenges with PBS or edema toxin (ET) and concurrent treatment with 24-h infusions of either placebo or S-methyl-l-thiocitrulline (SMTC) at 1.5 mg·kg⁻¹·h⁻¹. B: mean (±SE) plasma NO levels at 4, 8, or 24 h in animals after the start of 24-h challenges with PBS or edema toxin (ET) and concurrent treatment with 24-h infusions of placebo or SMTC (1.5 mg·kg⁻¹·h⁻¹).

Compared with PBS, ET increased NO levels at 24 h but not earlier (P = 0.01 for the time interaction; Fig. 6B). Compared with placebo, SMTC did not alter NO levels significantly at any time point in PBS-challenged animals (P ≥ 0.23). In ET-challenged animals, NO levels appeared decreased in SMTC-treated animals at 24 h, but this was not significant (P = 0.11). Compared with PBS, 24-h ET challenges produced progressive decreases in MAP in both placebo and SMTC-treated animals (P < 0.0001 for the time interactions for both) (Fig. 7A). Compared with placebo, SMTC treatment significantly increased MAP with ET but not PBS challenge (P = 0.002 and P = 0.29, respectively). Compared with placebo, SMTC, while not significantly altering the effects of ET on decreasing MAP from 1 to 4 h or 5 to 8 h (P ≥ 0.86), progressively reduced these decreases from 9 to 24 h (P = 0.01 for the time interaction; Fig. 7B).

Fig. 7.

A: the serial mean (±SE) hourly mean arterial blood pressures (MAP) in animals beginning immediately before and then after the start of 24-h challenges with PBS or edema toxin (ET) and concurrent treatment with 24-h infusions of placebo or S-methyl-l-thiocitrulline (SMTC at 1.5 mg·kg⁻¹·h⁻¹). B: comparison of the mean effects of ET on MAP in placebo vs. SMTC-treated animals (see materials and methods) plotted from a value of 0 immediately before challenge and treatments were initiated (MAP did not differ across the 4 groups at time 0, P = ns for all comparisons). Comparisons of the effects of ET with placebo vs. SMTC were done from 0 to 4 h, 5 to 8 h, and 9 to 24 h to correspond to the time points at which plasma NO levels were obtained (i.e., 4, 8, and 24 h).

Effects of 1400W in ex vivo and in vivo studies.

In a fourth ex vivo study (ex vivo-study 4), we examined the effects of 1400W on ET's arterial relaxant effects. The effects of the three doses of 1400W tested (1, 10, and 100 μM) on MCF or EC₅₀ did not differ significantly in either PA- or ET-challenged rings (P ≥ 0.18) for all except the lowest dose of 1400W, which had less effect on MCF at P = 0.06, and these doses were combined for analysis. Compared with PA-challenged rings, in both placebo- and 1400W-treated rings, ET decreased MCF (P < 0.0001 for both) and increased EC₅₀ (P < 0.0001 and P = 0.02, respectively) (Fig. 8, A and B). Compared with placebo treatment in PA- and ET-challenged rings, 1400W treatment increased MCF with both challenges (P = 0.01 and P = 0.001, respectively) and decreased EC₅₀ with both although this was only significant with ET (P = 0.01) and not with PA (P = 0.64; Fig. 8, A and B, and Table 1). After we controlled for the effect of 1400W on NO release due to PE-stimulated muscle contraction alone in PA-challenged rings (see materials and methods), compared with placebo, 1400W did not significantly alter ET's effect on either MCF (−38.0 ± 6.5 vs. −31.8 ± 6.5, respectively P = 0.50) or EC₅₀ (25.9 ± 5.7 vs. 13.6 ± 5.7, respectively, P = 0.13; Fig. 8, C and D).

Fig. 8.

This figure shows the mean (±SE) maximal contractile force (MCF) rings developed during PE stimulation (A) and the mean (±SE) estimated concentration of PE producing 50% of the MCF (EC₅₀; B) after pretreatment with protective antigen (PA) (1,600 ng/ml) or edema toxin (ET) (800 ng/ml of edema factor and 1,600 ng/ml of PA) combined with either placebo or N′-[3-(aminomethyl)benzyl] acetimidamide (1400W) (data for 1400W doses of 1, 10, and 100 μM combined, see results). C and D: comparison of the mean effects of ET on MCF and EC₅₀ in placebo- vs. 1400W-treated rings (see materials and methods). The bars in C and D, reflect the differences between placebo and ET denoted by the arrows in A and B. The number of rings employed to calculate MCF and EC₅₀ is shown under A and B. The brackets and P values demonstrate the levels of significance for comparisons of respective groups of rings.

An in vivo study (in vivo-study 3) then investigated the effects of 1400W treatment in doses of 0.175, 0.525 or 1.575 mg·kg⁻¹·h⁻¹ on survival and on NO and blood pressure levels in ET- or PBS-challenged animals. All animals challenged with 24-h PBS infusions and treated with either placebo or any of the three 1400W doses survived for 168 h. In animals challenged with 24-h ET infusions, compared with placebo, neither the 0.175, 0.525 or 1.575 mg·kg⁻¹·h⁻¹ dose of 1400W altered survival significantly [1 survivor of 8 (12.5%) with placebo vs. 1 survivor of 7 (14.3%) with 1400W 0.175 mg·kg⁻¹·h⁻¹; 2 survivors of 4 (50.0%) with placebo vs. 2 survivors of 4 (50.0%) with 0.525 mg·kg⁻¹·h⁻¹1400W; and 7 survivors of 21 (33.3%) with placebo vs. 8 survivors of 19 (42.1%) with 1.575 mg·kg⁻¹·h⁻¹1400W]. Because the hazards ratios of survival did not differ significantly (P = 0.90) comparing the increasing 1400W doses [0.89 (0.30, 2.67), P = 0.84; 0.97 (0.14, 7.00), P = 0.98; 1.03 (0.47, 2.27), P = 0.94], additional analysis examined the effects of the agent on NO levels and blood pressure after combining across the three doses.

Compared with PBS, ET increased NO levels at 24 h but not earlier (P < 0.0001 for the time interaction) (data not shown). At 24 h, compared with placebo, 1400W decreased NO levels in ET- but not PBS-challenged animals (P = 0.01 and P = 0.70, respectively) (data not shown). Compared with PBS, 24-h ET challenges produced progressive decreases in MAP in both placebo and SMTC-treated animals (P < 0.0001 for the time interactions, data not shown). Compared with placebo, 1400W did not alter MAP significantly in either PBS- or ET-challenged animals (P = 0.68 and P = 0.24, respectively, data not shown).

Relationship between NO and blood pressure levels and survival times at 24 h in untreated animals.

To further examine the relationship between ET-associated hypotension, NO production and mortality, we investigated these parameters in untreated animals across the three in vivo studies. Mortality with ET challenge occurred primarily between 24 and 48 h after the start of toxin. While 13 animals died up to and including 24 h (16.7%), 62 died after 24 h and up to and including 48 h (79.5%), and 3 died after 48 h (3.8%). We therefore examined the relationship between NO and blood pressure levels at 24 h immediately before the onset of most mortality and when recorded NO levels were the greatest and blood pressures the lowest with ET. Across untreated animals with available data (n = 78), there were modest but significant relationships between increasing NO levels and decreasing MAP (r = 0.42, P = 0.0002) and systolic arterial blood pressure (SAP) (r = 0.48, P < 0.0001). Furthermore, across nonsurvivors not receiving treatment (n = 57), there were modest but significant relationships between decreased MAP or SAP at 24 h and decreased survival times (r = 0.47, P = 0.002 and r = 0.43, P = 0.001, respectively). However, while NO levels were slightly higher in nonsurvivors with shorter survival times, this relationship was not significant (r = 0.15, P = 0.28).

Comparison of the effects of ET on the contractile force of aortic rings from eNOS-KO vs. WT mice.

To further examine whether ET's NO-mediated effects were related to the eNOS isoform, studies were conducted in aortic rings from WT and eNOS-KO mice. The MCF with phenylephrine in PA-treated rings was similar comparing rings prepared from WT and eNOS-KO mice (P = 0.75; Fig. 9). Compared with PA, treatment with ET reduced the MCF generated with phenylephrine in a pattern that was close to significance in rings from WT (P = 0.06) but not eNOS-KO mice (P = 0.66). Furthermore, the MCF with ET treatment was greater in rings from eNOS-KO animals compared with WT animals in a pattern that also approached significance (P = 0.08). The effects of PA and ET treatment on EC₅₀ in rings from WT and eNOS-KO animals did not differ significantly (P ≥ 0.40, data not shown).

Fig. 9.

This figure compares the mean (±SE) contractile force aortic rings generated during stimulation with increasing phenylephrine (PE) concentrations (A) and the mean (±SE) maximal contractile force (MCF) developed during PE stimulation (B) in rings from wild-type (WT) or endothelial nitric oxide synthase knockout (eNOS-KO) mice after pretreatment with protective antigen (PA) or with edema toxin (ET). Contractile force in A is shown as the percentage of the force rings generated after exposure to 60 mM KCl during initial assessment of the rings. The number of rings employed in each group is shown in A. The brackets and P values demonstrate the levels of significance for comparisons of respective groups of rings.

DISCUSSION

The present findings support the possibility that NO production contributes to the arterial hypotension and lethality ET produces in toxin and anthrax bacteria-challenged animal models. In aortic ring studies here, l-NAME reduced ET's effects on decreasing MCF and increasing EC₅₀. In in vivo studies l-NAME treatment reduced ET-associated increases in circulating NO levels at later time points. Furthermore, NO inhibition with l-NAME progressively reduced ET's hypotensive effects at later time points and significantly reduced ET-associated lethality. In aortic ring studies, SMTC reduced ET's effects on increasing EC₅₀, and in in vivo studies, at the highest dose studied, it reduced ET's hypotensive effects at later time points and ET-associated lethality. In both ex vivo and in vivo studies, 1400W had no significant effect on ET-associated changes except for reductions in NO at 24 h. In untreated animals from the three in vivo studies, at 24 h before the onset of most lethality with ET, there was a significant relationship between increased NO levels and decreased blood pressure.

Our findings and others support the possibility that ET-mediated cAMP release and its effects on NO production contribute to arterial relaxation and hypotension with ET. We previously showed that adefovir, a nucleoside that selectively inhibits EF-stimulated cAMP production, prevented the relaxant effects of ET in aortic rings (30). Early studies by others showed that isoprenaline and forskolin stimulated increases in cAMP, inhibited norepinephrine-induced aortic ring contraction (20). The actions of both agents were reduced in denuded aortic rings and in rings treated with l-NAME, a NOS inhibitor. More recent studies showed that β2-adrenoreceptor stimulation relaxed rat aorta through both NO-dependent and -independent mechanisms, both of which were mediated through PKA, one of the key effectors increased cAMP levels act on (15). Work in coronary blood vessels also showed that cAMP stimulates eNOS phosphorylation and activation through a PKB-mediated mechanism (53). Very recent studies showed that the endothelium-dependent component of forskolin/cAMP-induced relaxation of rat aortic rings is partially mediated by an increase in endothelial NO release due to both PKA and exchange protein directly activated by cAMP (Epac)-stimulated eNOS activity (17). Thus, while the present study does not directly link cAMP production to ET-mediated NO production and the hemodynamic changes noted here, our prior findings and these present ones, together with literature such as that cited here, strongly support this association.

Several lines of evidence from the present experiment suggest that eNOS was the isoform ET mediated its effects through in this rat model. l-NAME had the clearest effects in both the aortic ring and in vivo studies. While the l-NAME doses increasing blood pressure and survival in vivo were not selective ones, those inhibiting ET's effects on MCF and EC₅₀ in aortic rings were low and in a range reported to be eNOS selective (22, 46, 52). Consistent with a contribution of eNOS to ET's arterial effects, ET reduced MCF in aortic rings from WT but not eNOS-KO mice. Although SMTC, an nNOS inhibitor, did decrease EC₅₀ in the aortic ring model, it did not alter MCF and its in vivo effects were at doses that are not nNOS selective (23, 52). In both the aortic ring and in vivo models, 1400W, a selective iNOS inhibitor, did not alter arterial function, blood pressure, or survival. The highest dose of 1400W employed in vivo was three to four times greater than doses previously shown to increase blood pressure in endotoxin-challenged animals (14, 40, 48).

While l-NAME and higher dose SMTC did reduce hypotension with ET challenge from 9 to 24 h, these reductions did not appear as great as the increases in survival noted with the two treatments. These differences may relate to the fact that while blood pressure was only measured during the 24-h ET was delivered, lethality with ET occurred predominantly from 24 to 48 h. In rat and canine models, 24-h ET challenges reduced blood pressure up to 48 and 72 h after the start of toxin, respectively (29, 49). It is therefore possible that even though NOS inhibitors were stopped at 24 h, their effects on inhibiting the hypotensive effects of ET persisted and were greater between 24 and 48 h. It is noteworthy though that at 24 h in untreated animals, although there was a relationship between increased NO levels with ET and decreased blood pressure, and, in nonsurvivors, a relationship between decreased blood pressure and shortened survival time, there was no clear relationship between NO levels and subsequent survival times.

It is also possible that l-NAME and higher dose SMTC had protective effects with ET challenge beyond hemodynamic ones. For example, excessive NO production can interact with molecular oxygen and superoxide anion to produce reactive nitrogen species that can alter cellular function and produce tissue injury (24). NO production can also inhibit mitochondrial respiration leading to tissue injury and organ dysfunction (9).

Based on results with the NOS inhibitors studied, the effects of ET on reducing the responsiveness of aortic rings to PE and on reducing blood pressure in vivo were only partially related to the release of NO. Therefore, other mechanisms must also contribute to ET's arterial relaxant effects. In this regard, ET-stimulated increases in cAMP could influence any of the major pathways controlling vascular tone that rely on PKA or Epac activity including the following: decreases in intravascular calcium stores; hyperpolarization of smooth muscle cells; and reductions in the sensitivity of the contractile mechanism itself (36, 38). These findings raise the possibility that vasopressor agents such as phenylephrine might also have benefit with ET-associated shock.

In in vivo studies, the marked and persistent increases in cAMP levels were probably derived not only from vascular tissue but from other organs as well. The two anthrax toxin receptors [ATR1 (TEM8) and ATR2 (CMG2)] that mediate host cellular uptake of EF are widely distributed in different tissues including among others myocardial, arterial, venous, pulmonary, hepatic, renal, gastrointestinal, and skin (31). To what extent arterial as opposed to nonarterial cAMP production contributed to the present findings is not clear. What is apparent is that the effects of the NOS inhibitors in the present study were not related to inactivation of EFs adenyl-cyclase activity. Levels of cAMP in in vivo studies were increased in ET-challenged animals similarly with both placebo and NOS inhibitor.

Although the present study implicates NO production in the hypotensive and lethal effects of ET, whether NOS inhibition would have a role clinically during anthrax-associated shock is unclear. Importantly, NOS inhibitors were administered along with toxin infusion in this study, a method limiting any conclusion as to the actual clinical benefit of NOS inhibitors for anthrax. Several other issues might also preclude consideration of NOS inhibitors for anthrax. First, tachyphylaxis to NO or its production could decrease the effectiveness of NOS inhibitors (1). Second, prior attempts to inhibit NOS in septic shock related to nonanthrax bacteria, while improving blood pressure, actually worsened outcome, possibly due to harmful effects on afterload and cardiac function (32). Third, NO production has been noted to be important for host defense against anthrax and its inhibition might have detrimental effects during anthrax infection (43). Despite these issues, since both ET and the inflammatory response elicited by anthrax cell wall can contribute to excessive NO production and possibly produce levels greater than those encountered with other bacteria types, there may still be a role for judicious NOS inhibition during vasopressor resistant shock with anthrax infection (12, 42). Of note as well, bacterial NOS (bNOS)-mediated NO production by B. anthracis has been shown to be important for this bacteria's propagation (47). Since ET stimulation of host NO production might contribute to such propagation, its inhibition could also have protective effects. These questions can only be addressed in animal models that employ live anthrax bacteria challenge and that incorporate the type of conventional antibiotic and hemodynamic support employed in patients.

The present study did not examine the effects NOS inhibition with lethal toxin (LT), which also contributes to the pathogenesis of anthrax infection. However, in prior studies we did not find that LT altered arterial function in the aortic ring model (30). Furthermore, studies in mice showed that nNOS actually had a protective effect on myocardial function (35). Thus NO production may have very different roles in the pathogenic effects of ET and LT, making the clinical use of NOS inhibitors for anthrax infection even more problematic.

This study has several limitations. First, it is possible that early reductions in blood pressure with ET secondarily stimulated NO production, which NOS inhibition counteracted. For example, hemorrhagic shock was shown to stimulate pulmonary NO production in a rat model (39). However, as was the case with hemorrhagic shock, iNOS would be the most likely NOS isoform activated in response to systemic hypotension, and in the present study, 1400W, a selective iNOS inhibitor, had no effect on either blood pressure or survival. Second, even within the same animal species, there are differences in arginine metabolism comparing strains and sexes, which may alter NO production (34). It is also known that NO production rates differ comparing rodents and human (45, 51). Thus whether NO production would play the same role in shock related to ET during anthrax infection in humans as it does in this rat model is unclear. Third, there were small amounts of LPS detected in the PA and EF components employed in the aortic ring studies, which could have contributed to NO production. However, these amounts of LPS were extremely small. In prior studies, doses 20 times greater than these did not produce measurable changes in NO levels in animals (12). Also, in the present aortic ring studies, LPS would have been present in the PA-challenged control rings as well as ET-challenged ones.

Shock during anthrax infection is resistant to conventional hemodynamic support and is associated with a mortality rate greater than with other bacteria types (5). The present findings support the possibility that NO production contributes to the hypotensive and lethal effects of B. anthracis ET and provide one basis for the apparent resistance to vasopressor agents noted clinically during invasive infection. Based on the striking beneficial effect NOS inhibitors had on survival with ET challenge, additional investigation into the potential therapeutic benefit of such agents in models employing ET and LT and in bacteria-challenged models is warranted. However, other mechanisms not identified here, also appear to have a role in ET's arterial and hypotensive effects and these require investigation as well.

GRANTS

This research was supported by the Intramural Programs of the National Institutes of Health, Clinical Center, Critical Care Medicine Department and the National Institutes of Allergy and Infectious Diseases.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Y.L., X.C., and P.Q.E. conception and design of research; Y.L., W.X., L.O., H.S.-K., and Y.F. performed experiments; Y.L., X.C., and P.Q.E. analyzed data; Y.L., X.C., and P.Q.E. interpreted results of experiments; Y.L., X.C., and P.Q.E. prepared figures; Y.L., X.C., and P.Q.E. drafted manuscript; Y.L., X.C., D.S., M.M., S.H.L., and P.Q.E. edited and revised manuscript; Y.L., X.C., and P.Q.E. approved final version of manuscript.