Although circulatory shock related to lethal toxin (LeTx) may play a primary role in lethality due to Bacillus anthracis infection, its mechanisms are unclear. We investigated whether LeTx-induced shock is associated with inflammatory cytokine and nitric oxide (NO) release. Sprague-Dawley rats with central venous and arterial catheters received 24-h infusions of LeTx (lethal factor 100 μg/kg; protective antigen 200 μg/kg) that produced death beginning at 9 h and a 7-day mortality rate of 53%. By 9 h, mean arterial blood pressure, heart rate, pH, and base excess were decreased and lactate and hemoglobin levels were increased in LeTx nonsurvivors compared with LeTx survivors and controls (diluent only) (P ≤ 0.05 for each comparing the 3 groups). Despite these changes, arterial oxygen and circulating leukocytes and platelets were not decreased and TNF-α, IL-1β, IL-6, and IL-10 levels were not increased comparing either LeTx nonsurvivors or survivors to controls. Nitrate/nitrite levels and tissue histology also did not differ comparing LeTx animals and controls. In additional experiments, although 24-h infusions of LeTx and Esherichia coli LPS produced similar mortality rates (54 and 56%, respectively) and times to death (13.2 ± 0.8 vs. 11.0 ± 1.7 h, respectively) compared with controls, only LPS reduced circulating leukocytes, platelets, and IL-2 levels and increased TNF-α, IL-1α and -1β, IL-6, IL-10, interferon-γ, granulocyte macrophage-colony stimulating factor, RANTES, migratory inhibitory protein-1α, -2, and -3, and monocyte chemotactic protein-1, as well as nitrate/nitrite levels (all P ≤ 0.05 for the effects of LPS). Thus, in contrast to LPS, excessive inflammatory cytokine and NO release does not appear to contribute to the circulatory shock and lethality occurring with LeTx in this rat model. Although therapies to modulate these host mediators may be applicable for shock caused by LPS or other bacterial toxins, they may not with LeTx.
- Bacillus anthracis
inhalational Bacillus anthracis infection is a highly lethal disease and a major bioterrorism health threat today (3, 10, 21, 22). In the recent outbreak of inhalational B. anthracis in the United States, death was complicated by progressive shock despite treatment with appropriate antibiotics and aggressive hemodynamic and pulmonary support (3, 22). Over the past 20 years considerable work has shown the important role an excessive host inflammatory response plays in the pathogenesis of shock related to both gram-positive and gram-negative bacteria and the toxins they produce (7, 14, 30). To what extent the pathogenic events leading to shock and organ injury during B. anthracis infection resemble those of other bacteria is unknown.
Production of lethal toxin (LeTx) appears central to the pathogenesis of B. anthracis (18, 24). LeTx consists of two components, protective antigen (PA) necessary for the receptor-specific uptake of toxin by host cells and lethal factor (LF) the enzymatic moiety. Administration of purified preparations of LeTx is lethal in several different animal models (9, 25). Strains of B. anthracis in which either PA or LF have been inactivated are 1,000-fold less lethal than fully active strains (29). Finally, prophylactic treatment with antibody to either PA or LF alone is protective in animal models of infection (19, 20).
LF is a zinc protease that cleaves and inactivates several members of the mitogen-activated protein kinase kinases (MAPKKs) family (4, 17, 39). Although the molecular events after cleavage of MAPKK are unclear, LeTx causes lysis of macrophages from some inbred mice but not other cells (12). It has been proposed that macrophage lysis by LeTx may result in the release of inflammatory cytokines and subsequent activation of other injurious host mediators (13). However, data supporting a role for an excessive inflammatory response contributing to the pathogenesis of LeTx have been inconsistent. In some studies, administration of LeTx caused the release of TNF-α and IL-1β from RAW 264.7 murine macrophages and TNF-α from ICR mouse peritoneal macrophages (13, 33). Also, macrophage depletion or administration of IL-1β receptor antagonist was protective in BALB/c mice challenged with intravenous LeTx injections (13). However, in other studies, low LeTx doses were not associated with TNF-α or IL-1β release from RAW 264.7, J774A.1, or IC-21 cells (8). Also, LeTx was found to inhibit TNF-α and nitric oxide release in response to LPS and interferon (IFN)-γ in RAW 264.7 and MT2 cells (28). Finally, intraperitoneal challenge with lethal doses of LeTx in BALB/cJ and C57BL/6J mice did not result in sustained cytokine increases (23).
To investigate the pathogenesis of shock and organ dysfunction during anthrax sepsis, we developed a rat model in which LeTx is infused over a 24-h period. Traditional models studying LeTx action in this species have typically employed the Fischer rat and a rapid bolus of LeTx (2, 6). In these models death typically occurs within 1-2 h. In the present model, however, using Sprague-Dawley animals, doses of LeTx could be titrated so that lethality did not occur until 9-21 h after the initiation of toxin infusion. This longer time to death in combination with serial measures provided by indwelling intra-arterial and venous catheters allowed us to compare cardiopulmonary changes and intravascular inflammatory cytokine and nitrate/nitrite levels in survivors and nonsurvivors challenged with LeTx. Then, to determine how the effects of LeTx compare with a bacterial toxin with pathogenic effects directly related to inflammatory cytokine and nitric oxide production, we challenged animals with similarly lethal 24-h infusions of either LeTx or Escherichia coli LPS (30). Finally, to investigate the applicability of our model to other rat strains, we compared the effects of continuous LeTx infusion in Sprague-Dawley and Fischer rats.
MATERIALS AND METHODS
Animal care. The protocol used in this study was approved by the Animal Care and Use Committee of the Clinical Center of the National Institutes of Health. During the study, every effort was made to minimize animal suffering. The research protocol required the veterinarian staff or principal investigators to euthanize any animal that experienced unexpected pain or distress.
Study design. Sprague-Dawley or Fischer rats (n = 491 or 35, respectively) weighing 180-250 g with previously placed arterial and central venous catheters were anesthetized with ketamine (40 mg/kg) and xylazine (10 mg/kg) (Fig. 1). They were then randomized to receive LeTx with doses of LF in combination with PA ranging from 10 with 20 μg/kg body wt, respectively, to 5,000 with 10,000 μg/kg body wt, respectively, as either an infusion over 24 h or an intravenous bolus over 30 s (Table 1). Throughout this paper, doses of LeTx are designated as a multiple of the lowest doses of LF and PA tested (i.e., 1× is 10 μg/kg body wt LF and 20 μg/kg body wt PA) (Table 1). Control animals received diluent only (i.e., 0×, PBS with rat albumin 25 μg/ml). In experiments with Fischer rats, animals were only challenged with infused doses of LeTx of 1×,5×,or10× (Table 1). All animals received equivalent total volumes of diluent with or without LeTx as a bolus or infusion where appropriate. Immediately before and at 1- to 3-h intervals during LeTX infusion, animals had mean arterial blood pressure (MBP) and heart rates (HR) measured. At 3, 9, and 24 h after the start of LeTx infusion, some animals had serial arterial blood gases (ABGs), complete blood counts (CBCs) with differentials, and serum for cytokine measures obtained (11). At 6 or 24 h, other animals receiving LeTx or control infusions were randomly selected to be anesthetized to have blood obtained for serum nitrate/nitrite levels and then to be killed to have lung and peritoneal lavage performed and to have heart, lung, liver, and kidney tissue harvested for histological assessment. All animals not euthanized for tissues were observed for 168 h. In separate experiments, Sprague-Dawley rats were randomly assigned to receive similarly lethal 24 h infusions of either B. anthracis LeTx (10×) or E. coli 0111:B4 LPS (3 mg/kg body wt, Sigma, St. Louis, MO) and to have hemodynamic measures, ABGs, CBCs, serum cytokine, and nitrate/nitrite levels performed at 3, 9, and 24 h. Numbers of animals assigned to individual experiments are outlined in Table 1.
LeTx and LPS preparations. PA and LF were recombinant proteins prepared from B. anthracis under LPS-free conditions as previously described (27, 31). B. anthracis does not contain LPS, and functional tests for evidence of endotoxin in PA and LF preparations, including inflammatory cytokine production by macrophages in vitro or in mice in vivo, were negative. Based on studies assessing the concentration of LeTx in guinea pigs in the terminal stages of B. anthracis infection, the dose of LeTx employed in most of our experiments (i.e., 10×) in rats is not in excess of what would be expected with a lethal dose of live bacteria in this animal model (29a). LPS from E. coli 0111:B4 was purchased from Sigma. Rat albumin (25 μg/ml) was used to maintain the stability of PA and LF in PBS. Similar treatment was applied to LPS infusions. In all experiments, volumes infused of either diluent as control or the differing toxin concentrations of LeTx or LPS were the same.
Hemodynamic, ABG, and CBC measures. Catheters protected with a coiled spring (Coiled Tether, Lomir, Malone, NY) were attached to exteriorized arterial and central venous access ports on each animal. Central venous catheters were attached via three-way stopcocks to a syringe pump (PHD 2000 syringe Pump, Harvard Apparatus, Holliston, MA) to provide LeTx as an infusion. Arterial catheters were connected to transducers (Pressure Transducer, Maxxim Medical, Athens, TX) to determine arterial blood pressure (systolic, diastolic, and mean) and HR (BioSystem XA; BUXCO, Troy, NY) and for blood collection. After equilibration, continuous measures of each hemodynamic parameter were collected for a 5-min period, and the mean of the measurements for that period was recorded. Arterial blood was collected for blood gas analysis and lactate measures (iSTAT Clinical Analyzer, Abbott, East Windsor, NJ) or complete blood cell count and white blood cell differentials. Alveolar-to-arterial oxygen gradients (A-aO2) and arterial base excesses (ABE) were calculated using standard formulas.
Cytokine measures. In initial experiments with LeTx alone, cytokine measures, including TNF-α, IL-1α and -1β, IL-2, IL-4, IL-6, IL-10, IFN-γ, and granulocyte macrophage colony stimulating factor (GM-CSF) were made using a protein multiplex immunoassay system (Rat 9-Bio-Plex Cytokine Array System, Bio-Rad Laboratories, Hercules, CA). In experiments comparing the effects of LeTx and LPS, the cytokines TNF-α, IL-1α and -1β, IL-2, IL-4, IFN-γ, GM-CSF, three migratory inhibitory proteins (MIP-1α, MIP-2, MIP-3α), monocyte chemoattractant protein (MCP)-1, and regulated on activation, normal T-cell expressed and secreted (RANTES) were measured using the Searchlight Proteome Array Multiplex system (Pierce, Rockford, IL).
Nitrate/nitrite measures. As previously described, plasma was precipitated with two volumes of cold ethanol (-20°C), followed by centrifugation (41). The supernatant was then mixed with vanadium chloride (5 ml of 0.4% in 1 N HCl) and heated to 90°C in a purge vessel through which helium was continuously bubbled. The resulting nitric oxide from nitrate, nitrite, and S-nitrosothiol species in the samples was detected with a nitric oxide analyzer (Sievers 280, Boulder, Colorado). Measurements were based on comparisons to control samples with known concentrations of nitrate.
Lung and peritoneal lavage, lung wet-to-dry weight ratios, and tissue histology measures. After lung removal, lavage was performed by cannulating the trachea with a blunt needle and sequentially lavaging with four 3-ml aliquots of PBS that were combined. Peritoneal lavage was performed by injecting 20 ml PBS followed by aspiration. The lavaged fluid was centrifuged, the supernatant was removed, and the cell pellet was resuspended. Cell counts were performed with an electronic cell counter (Z1 Coulter Particle Counter, Coulter Electronics, Miami, FL). Slides for differential cell counts were prepared (Cytospin 4, Thermo Shandon, Pittsburgh, PA) and stained. Cell differentials were performed on each slide. Lavage supernatants were passed through a 45-μm filter, and protein determinations were made using a Folin-Lowry technique.
Liver, kidney, heart, and two lung lobes were removed from randomly selected animals at 6 or 24 h. One lung lobe was air dried, and wet-to-dry weight ratios were calculated as previously described (11). Other tissue was fixed in 10% formalin and stained with hemotoxylin and eosin for reading by a pathologist blinded to study groups. Organs were graded numerically as to the presence and degree of vascular congestion and of extravascular hemorrhage, necrosis, and elements of an inflammatory cellular infiltrate, both acute and chronic.
In vitro macrophage cytotoxicity assays. Thioglycolate-induced peritoneal macrophages were harvested from BALB/cJ mice and noninduced resident cells from Sprague-Dawley and Fischer rats by saline lavage. These macrophages or RAW 264.7 cells were plated in 96-well plates in DMEM supplemented with 10% fetal bovine serum, 10 mM HEPES, 2 mM Glutamax I, and 50 μg/ml gentamicin for 4 h. LeTx (PA + LF) was added at a range of concentrations (0-2,000 ng/ml) in duplicate, and cells were incubated at 37°C for 2 h. Cell viability was assessed by the 3-(4,5-dimethyl-2-thiazyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) as previously described (37). Viability was a calculated as a percentage of medium-treated cells.
Statistics. The relationships between infused LeTx doses and final mortality rates were analyzed by a Spearman Rank test. A Cox proportional hazards model was used to compare survival rates with differing LeTx doses vs. diluent only, with LeTx administered as an infusion vs. a bolus, with infusion of LeTx vs. LPS, and with LeTx infusion in Sprague-Dawley vs. Fischer animals. Laboratory parameters were compared with ANOVA taking into account time, LeTx dose, and type of toxin. Alternative Tukey test was applied when necessary. Variables were transformed where appropriate. Data in Figs. 5 and 6 comparing the effects of LeTx and LPS are shown as differences from control. P ≤ 0.05 was considered statistically significant. Data are presented as means ± SE.
Comparison of the effects of LeTx infusion vs. bolus on mortality rates in Sprague-Dawley rats. LeTx doses ranging from 0× (control) up to 500×were administered as 24-h infusions (Table 1). Increasing doses of LeTx resulted in increased overall mortality rates (P < 0.001, r = 0.89) and decreased time to death in nonsurvivors (P < 0.0001, r = -0.70). LeTx infusion with doses of 5×, 7.5×, or 10× resulted in mortality rates of 10-45% (31% overall) with time to death in nonsurvivors ranging from 9 to 21 h (13.6 ± 1.5 h overall). Compared with infusion, administration of these same three doses of LeTx administered as a 30-s intravenous bolus resulted in higher mortality rates of 50-75% (56% overall) with shorter time to death in nonsurvivors of 1-3 h (overall 2.0 ± 0.3 h) (P = 0.02 and P < 0.0001 for the differing effects of infusion vs. bolus of LeTx on overall mortality rate and time to death, respectively) (Fig. 2). Doses of LeTx administered as an infusion (10×) that produced similar [P = nonsignificant (NS)] mortality rates compared with a bolus (5×) still resulted in significant increases in time to death (11.4 ± 1.5 vs. 2.7 ± 1.2 h, P < 0.0001).
Laboratory measures in Sprague-Dawley rats receiving LeTx infusion or control. To identify pathophysiological events associated with death during 24-h LeTx infusions in Sprague-Dawley animals, selected laboratory parameters were investigated serially in survivors and nonsurvivors receiving LeTx and in controls receiving diluent only. This analysis was performed with parameters collected up to 9 h after the start of a 10× dose of LeTx infusion, the time at which death was first evident in the nonsurvivor group. Other measures were obtained in animals randomly selected for death at 6 h and were used to assess the effects of LeTx alone without differentiating survivors and nonsurvivors.
From 0 to 9 h after the start of LeTx infusion, MBP and HR decreased in nonsurvivors with LeTx compared with survivors and controls (P < 0.0001 for the differences in each parameter comparing nonsurvivors, survivors, and controls over the 9 h) (Fig. 3). These decreases in nonsurvivors were even more striking since gradual clearance of initial anesthetic in controls and survivors resulted in increases in these two parameters over this 9-h period. Although not different early at 3 h, later at 9 h arterial pH and base excess were lower and lactate and hemoglobin were higher in nonsurvivors compared with survivors receiving LeTx and controls (P ≤ 0.01 for the differences in each parameter comparing nonsurvivors, survivors, and controls at 9 h) (Fig. 4). At these same time points, alveolar to arterial oxygen gradients, and circulating leukocytes and platelets (Table 2) did not differ significantly comparing nonsurvivors and survivors receiving LeTx and controls (P = NS for each). None of these parameters at 24 h were significantly different comparing controls and survivors receiving LeTx (P = NS for each, data not shown).
Using the Bio-Plex Assay, TNF-α, IL-6, and IL-10 at both 3 and 9 h and IL-1β at 3 h were similar comparing nonsurvivors and survivors receiving LeTx and controls (P = NS for each) (Table 3). At 9 h, IL-1β in nonsurvivors and survivors receiving LeTx was reduced compared with controls (P = 0.01 for the effect of LeTx vs. control). IL-2, IL-4, IFN-γ, and GMCSF were not detected. In animals sampled before death at 6 h compared with controls, those receiving LeTx did not have significant differences in nitrate/nitrite levels (27.2 ± 3.9 vs. 27.9 ± 3.7 μmol/l, respectively, P = NS). None of these measures at 24 h was significantly different comparing controls (all survivors) and survivors receiving LeTx (P = NS for each, data not shown).
After death and inspection, animals receiving LeTx were noted in almost all cases to have pleural serosanguinous fluid collections not found in control animals. Compared with controls, animals receiving LeTx had increases in lung lavage protein that approached being significant [2.64 ± 0.09 vs. 3.04 ± 0.21 log (mg/ml)] (P = 0.11 in controls vs. LeTx animals for 6 and 24 h combined). Compared with controls, animals receiving LeTx did not have significant differences in histology scores for liver, kidney, heart, or lung at either of the time points measured (P = NS for all, Table 4). Furthermore, neither lung lavage or wet-to-dry weight ratios differed between these groups (P = NS for all, data not shown).
Comparison of the effects of similarly lethal LeTx and LPS infusions in Sprague-Dawley rats on mortality rates and laboratory measures. Twenty-four hour infusions of LeTx (10×) and LPS (3 mg/kg) produced similar mortality rates (54 vs. 56%) and time to death (13.2 ± 0.8 vs. 11.0 ± 1.7 h) (P = NS for both comparing the effects of LeTx vs. LPS). Compared with controls, from 0 to 9 h, nonsurvivors had decreases (mean ± SE changes from control averaged over the 9-h time period) in MBP and HR with both LeTx (-8 ± 2 mmHg and -30 ± 7 beats/min) and LPS (-6 ± 2 mmHg and -12 ± 7 beats/min) (all P < 0.05 except for HR with LPS that was P = 0.11). In survivors, MBP was not altered with either LeTx or LPS (-1 ± 2 vs. 2 ± 2 mmHg, respectively, P = NS), whereas HR was decreased with the LeTx (-20 ± 7 beats/min) and increased with the LPS (22 ± 7 beats/min) (P = 0.05 for the differing effect of LeTx vs. LPS on HR in survivors from 0 to 9 h combined).
Similarly in both survivors and nonsurvivors at either 3 or 9 h or both, compared with controls, LPS but not LeTx decreased circulating neutrophils, lymphocytes, and monocytes and serum IL-2 levels (Searchlight Proteome Array Multiplex System) and increased serum TNF-α, IL-1α, IL-1β, IL-6, IL-10, GM-CSF, IFN-γ, RANTES, MIP-1, -2, and -3α, and MCP-1 levels (P < 0.05 and P = NS for the effects of LPS and LeTx, respectively, for each parameter) (Fig. 5, Table 5; data for MIP-2 and -3α not shown). IL-4 levels were not detectable in any group. Compared with controls, similarly in both survivors and nonsurvivors at 3 and 9 h, LPS but not LeTx increased serum nitrate/nitrite levels (P < 0.001 and P = NS for the effects of LPS and LeTx, respectively) (Fig. 6, Table 5). In survivors at 24 h, changes with LPS in nitrate/nitrite levels and all cytokines except IL-1α and IL-2 persisted (P ≤ 0.05 for all, Table 5).
Comparison of the effects of LeTx infusion on mortality rates in Sprague-Dawley vs. Fischer rats. Twenty-four hour infusions of LeTx in doses of 5 or 10×produced mortality rates that were greater in Fischer (72%) compared with Sprague-Dawley (18%) rats (P < 0.0001 for the effect of these 2 doses of LeTx combined on mortality rates in Sprague-Dawley vs. Fischer rats). However, with these same doses of LeTx, the time to death in nonsurvivors was similar (P = NS) comparing the Fischer (13.0 ± 1.2 h) and Sprague-Dawley (15.4 ± 2.5 h) rats. Also, infusions of LeTx with dose 1× did not cause mortality in either strain (n = 5 per group, data not shown).
Comparison of the in vitro cytotoxic effects of LeTx on Sprague-Dawley and Fischer rat and BALBc/J mouse macrophages and RAW 264.7 Cells. Macrophages from Sprague-Dawley rats were resistant (viability = 93.1 ± 2.7%) to LeTx (1,000 ng/ml) while BALB/cJ and RAW 264.7 macrophages were completely sensitive (viability = 0). Macrophages from Fischer rats exhibited a degree of resistance (viability = 64.6 ± 8.1%) that was variable between animals but signifi-cantly high compared with Balb/cJ and RAW 264.7 macrophages (P < 0.001) (data not shown).
During the development of this model of anthrax sepsis, we found that twenty-four hour infusions of LeTx resulted in dose-ordered increases in mortality rates in Sprague-Dawley rats. In contrast to bolus administration, infusion of LeTx in doses producing mortality rates of 10-45% did not result in death until 9 h and permitted a comparison of changes in nonsurvivors and survivors. Over the initial 9 h of infusion, LeTx caused reductions in blood pressure, but these decreases were much greater in nonsurvivors than survivors. During this time period, hypotension in nonsurvivors was associated with reductions in HR, the presence of acidosis, and increases in hemoglobin levels but was not associated with hypoxemia or increases in serum inflammatory cytokine levels. Furthermore, nitrate/nitrite levels and inflammatory tissue injury did not appear to be increased with doses of LeTx producing hypotension and lethality. In additional confirmatory studies, we showed that lethality and shock with LPS but not LeTx was associated with marked reductions in circulating leukocytes and platelets and increases in inflammatory cytokine and nitric oxide levels. Finally, to study the applicability of this infused LeTx model to other rat strains, we showed that similar doses of LeTx produced lethality with the same time courses in Fischer and Sprague-Dawley rats, although overall mortality rates were greater in the former than the latter.
Thus cardiovascular dysfunction manifested as hypotension and reduced tissue perfusion with lactic acidosis appeared to play a primary role in lethality during LeTx infusion in the present model. This is in contrast to many prior preclinical studies that suggested that the cardiovascular changes occurring with LeTx were secondary to pulmonary changes and only occurred very close to the time of death (2, 6). These pulmonary changes were believed related either to a primary effect of LeTx on the lung or to its effects on the central nervous system (2, 6, 32, 38). This did not appear to be the case in the present model because both oxygenation and carbon dioxide levels were unchanged from controls at a time when hypotension was already marked. Rapid administration of anthrax toxin or inability to serially monitor hemodynamics in prior studies or other factors may have resulted in the differences between our observations and previous ones. However, our findings are consistent with original speculation regarding the pathogenic effects of anthrax toxin (35), as well as by reports of nonsurvivors in the recent series of patients presenting with anthrax infection after exposure to contaminated mail (3, 22). In these patients shock was not only a common finding early in presentation but also as disease progressed (3, 22). In this regard, the slow administration of LeTx in our model may better simulate the progressive increases in toxin levels that have been demonstrated to occur in animals and would be expected in patients during live bacterial infection (15, 16). Finally, the present findings are similar to recent studies in C57BC/6J and BALB/cJ mice in which intraperitoneal administration of LeTx caused histological changes in the liver consistent with a shocklike state (23). This was first evident at 18-24 h but progressed through the subsequent 72-h period. Although physiological measures were not available from that study, lung histology showed little evidence of injury, suggesting that hepatic injury was a result of local hypoxic conditions likely due to a state of hypoperfusion. Although hepatic injury was not evident on histology in rats in the present study, this may have been due either to the more rapid time to death in the model or to earlier tissue procurement. Alternatively, differences in the species studied, the method of toxin administration, or other factors may have resulted in the differing histological findings.
Several mechanisms may have contributed to hypotension with LeTx infusion. In light of the reductions in HR in nonsurvivors, it is possible that LeTx had a direct negative chronotropic effect in this model. Supporting this, survivors receiving LeTx had decreases in HR compared with those receiving LPS, a toxin recognized to alter peripheral vascular function. Other bacterial toxins from both gram-negative and gram-positive bacteria have been shown to directly depress cardiac nodal and conduction tissue (42). It is also possible that independent of direct negative chronotropic effects, LeTx depressed myocardial function itself. This has been described with other bacterial toxins (26). However, in the present study in rats as well as in a prior study in mice, myocardial histology after LeTx challenge did not show evidence of injury that would provide a basis for direct chronotropic or inotropic effects (23). Furthermore, prior animal studies have not attributed negative inotropic or chronotropic effects to LeTx (2, 6, 15, 35, 38). Finally, patients with shock who were intensively monitored during the recent outbreak of B. anthracis were not reported to demonstrate uncharacteristic reductions in HR (3, 21, 22). Decreases in HR in the present study may have been a result of rather than a cause of the observed hypotension. Further studies including ones that maintain either HR or blood pressure constant during LeTx challenge will be necessary to clarify this relationship.
It is possible that LeTx may directly effect the peripheral vasculature leading to hypotension. Loss of vascular integrity due to LeTx would be consistent with the higher hemoglobin concentrations noted in nonsurvivors. Reductions in hemoglobin between 3 and 9 h in controls and survivors likely resulted from the repetitive blood draws that were done and with hemodilution due to continuous infusion with LeTx and diluent. In nonsurvivors, however, these reductions may have been counteracted by vascular leak and hemoconcentration. Consistent with vascular leakage, serosanguinous fluid collections were evident in the thoracic space in animals receiving LeTx but not in normal animals. Similar fluid collections were noted in mice challenged with LeTx, and recurrent pleural effusions were a prominent part of the clinical course in patients presenting with B. anthracis in the recent outbreak (3, 21-23). Consistent with a loss of vascular integrity in the present study, there were modest increases in lung lavage protein concentrations with LeTx. Loss of vascular integrity also appeared important in earlier models exploring the pathogenic effects of LeTx (2, 6, 35). However, we cannot exclude the possibility that changes in hydrostatic pressure or other mechanisms resulted in the observed changes in rats. Additional studies such as ones directly testing vascular permeability and cardiac filling pressures will be necessary to differentiate these mechanisms.
Alternatively, hypotension in the present study could have been related to a primary effect of LeTx on peripheral arterial vasoconstrictor function. If this were the case, however, hypotension with LeTx did not appear related to the excessive inflammatory cytokine production and nitric oxide release believed important in shock related to other types of bacterial toxin like endotoxin (7). In contrast to challenges with similarly lethal doses of LPS, TNF-α, IL-1β, and nitrate/nitrite levels were not increased with LeTx challenge. Furthermore, reductions in circulating leukocytes and platelets that typically accompany activation of an intravascular inflammatory response and that were present with LPS treatment did not occur with LeTx administration. Finally, peritoneal and lung lavage samples and histology specimens from several different organs did not show localized inflammatory injury with LeTx. These findings are at odds with those suggesting that agents designed to selectively inhibit components in the host inflammatory response may be protective during challenge with anthrax toxin (13). However, they are consistent with in vitro macrophage studies and in vivo mouse studies in which LeTx stimulation was not associated with inflammatory cytokine release (8, 23, 28). In contrast to the lack of effect LeTx appeared to have on cytokine and nitric oxide release, in other studies it has been shown to depress glucocorticoid receptor activity and to activate phospholipase A2 activity. These two actions could both interfere with normal vascular tone and provide an alternate mechanism for hypotension with LeTx (1, 34, 36, 40).
The Fischer rat has frequently been employed for studies with LeTx due to its recognized sensitivity to this toxin (2, 6). In the present investigation, however, similar doses of toxin were lethal in both Sprague-Dawley and Fischer animals, with time courses in nonsurvivors that were very similar. Thus alternate strains of rats to the Fischer animal appear applicable to the study of pathogenic events associated with LeTx. It was previously reported that macrophage sensitivity to LT was a requirement for animal susceptibility in the mouse model of toxin pathogenesis (13). In the present study, however, macrophages from both Sprague-Dawley and Fischer rats demonstrated resistance to the in vitro effects of LeTx compared with sensitive BALB/cJ mouse macrophages and RAW264.7 macrophages. These studies are consistent with others indicating that macrophage susceptibility may not play an essential role in LeTx-mediated toxicity (23). However, potential differences in the resistance exhibited by cells from Sprague-Dawley and Fischer animals deserves further study.
In summary, the present findings suggest that LeTx is capable of producing circulatory shock independent of its effects on pulmonary function and via mechanisms that do not induce excessive inflammatory cytokine and nitric oxide release. Although inhibition of the inflammatory response and nitric oxide may be applicable for shock related to LPS or other bacterial toxins, this may not be the case for LeTx during B. anthracis infection. Understanding the mechanisms underlying the circulatory shock caused by LeTx, however, will be important both for determining the usefulness of conventional hemodynamic therapies and for the development of more effective ones. For example, whether fluid administration would have beneficial or harmful effects in this model may depend on the degree to which loss of vascular integrity as opposed to abnormal vasoconstrictor function contributes to the hypotension that was observed. While aggressive fluid support might have little effect in the former, it would presumably be beneficial with the latter. This model provides a tool to address such questions.
We thank D. Stephany for technical assistance, M. Gladwin and his laboratory for nitric oxide determinations, and J. Candotti for preparation of the manuscript.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society