Histopathological effects of anthrax lethal factor on rat liver

Abstract

Bacillus anthracis, the causative agent of anthrax, has become an increasingly important scientific topic due to its potential role in bioterrorism. The lethal toxin (LT) of B. anthracis consists of lethal factor (LF) and a protective antigen (PA). This study investigated whether only lethal factor was efficient as a hepatotoxin in the absence of the PA. To achieve this aim, LF (100 µg/kg body weight, dissolved in sterile distilled water) or distilled water vehicle were intraperitoneally injected once into adult rats. At 24 h post-injection, the hosts were euthanized and their livers removed and tissue samples examined under light and electron microscopes. As a result of LF application, hepatic injury – including cytoplasmic and nuclear damage in hepatocytes, sinusoidal dilatation, and hepatocellular lysis – became apparent. Further, light microscopic analyses of liver sections from the LF-injected rats revealed ballooning degeneration and cytoplasmic loss within hepatocytes, as well as peri-sinusoidal inflammation. Additionally, an increase in the numbers of Kupffer cells was evident. Common vascular injuries were also found in the liver samples; these injuries caused hypoxia and pathological changes. In addition, some cytoplasmic and nuclear changes were detected within the liver ultrastructure. The results of these studies allow one to suggest that LF could be an effective toxicant alone and that PA might act in situ to modify the effect of this agent (or the reverse situation wherein LF modifies effects of PA) such that lethality results.

• Anthrax
• electron microscopy
• histopathology
• lethal factor
• light microscopy
• rat

Introduction

Anthrax is an acute (septicemic) disease caused by Bacillus anthracis. Pathologies associated with anthrax include an increase in body temperature, swelling of the spleen, and dark tarry blood clot formation in sub-serosal tissues accompanied by sero-hemorrhagic infiltrations (Morens, 2003). The first experimental studies in relation to B. anthracis toxins were performed in the second half of the 20^th Century (Kolesnik et al., 1987). These showed that the failure of the lungs and blood vessels that becomes apparent after B. anthracis exposure was caused by a series of toxins. Other studies also reported that macrophages (and the central nervous system) were negatively affected (Bonventre et al., 1967). In the past decade, the possibility of the use of anthrax agents and/or their associated toxins as bioweapons, in addition to the fact that some people receiving anthrax spores during bioterrorism events, led to a surge in the importance of understanding what the causative toxins were and how they might act in situ (Khan et al., 2000).

Two major virulence factors of B. anthracis are known, i.e. the capsule and the toxin. With respect to the toxin itself, there are three different antigenic components – Factor I, edema factor (Edema Factor, EF), Factor II, protective antigen (Protective Antigen, PA), and Factor III, lethal factor (Lethal Factor, LF). Two or three of these factors have a combined effect. For example, in experimental animals, lethal toxin (LT or LTx) – consisting of PA + LF – is known to be responsible for the lethality of anthrax (Young & Collier, 2007). In this study, we wondered whether or not LF itself has any influence in the absence of other factors. To address this query, we investigated the hepatotoxic effects of LF via histopathological examination of the livers of adult rats that had been subjected to a single dose of the LF.

Materials and methods

Animals and treatments

For this study, 12 adult Sprague Dawley rats (200–250 g; 10-weeks-of-age) were obtained from the Experimental Research Center of Atatürk University. All rats were housed in specific pathogen-free facilities maintained at 22 [±2]°C with a 55 [±5]% relative humidity and a 12 h light/dark cycle. All rats had ad libitum access to standard rodent chow and filtered water over the course of the experiment. The Local Animal Care Committee of Atatürk University approved all experimental protocols utilized in the studies described herein.

After being acclimated for 1 week, the rats were allocated to two groups. Rats in the treatment groups (n = 6) were then intraperitoneally administered 100 µg Lethal Factor (LF)/kg body weight in a 20 µl volume or an equal volume of vehicle only. Anthrax LF (recombinant LF from B. anthracis) was purchased from Calbiochem (Merck, Germany, Catalog #176900) and dissolved in sterile distilled water to achieve the concentrations needed for injection(s). This dose was selected based upon studies by Moayeri et al. (2005). Twenty-four hours after the injection, a perfusion fixation procedure was performed on all rats by intra-cardiac administration of neutral formalin while the rats were under Sevorane® (Abbott, İstanbul, Turkey) anesthesia (inhalation). Thereafter, each liver was removed and sub-divided for use in the light and electron microscopy studies.

Tissue processing for light microscopic analyses

For the light microscopy studies, the isolated portion of the liver was fixed in 10% formalin solution for 48–55 h, then dehydrated in a graded alcohol series, immersed in xylene, and ultimately embedded in paraffin wax. Tissue sections were then prepared on a Leica RM2125RT microtome (Ser-Med, Ankara, Turkey). Sections (5-µm thick) were mounted onto glass slides and then stained with hematoxylin and eosin. After appropriate de-staining and cover-slipping of the slides, the tissues were examined and photographed on a BX 51 light microscope (Olympus Corp., Tokyo, Japan)

Electron microscopy

Electron microscopy was performed as previously reported in Aslan et al. (2006) and Kiki et al. (2007). In brief, liver samples from each rat (two tissue samples/rat) were fixed in 2.5% glutaraldehyde (in 0.1 M phosphate buffer), post-fixed in 1.5% phosphate-buffered osmium tetroxide, dehydrated in a graded acetone series, and then washed in propylene oxide. After dehydration, samples were embedded in fresh Araldite CY 212 (Agar, Cambridge, UK) and then cut into 70–80 nm thick sections for ultra-structural evaluation. The thin sections were stained in 2% (w/v) uranyl acetate and 0.4% (w/v) lead citrate solutions, and then examined on a 100 SX transmission electron microscope (Jeol, Tokyo). Twenty photographs were obtained/sample.

Results

Histological results – Light microscopy findings

In sections obtained from livers of control rats, a conventional healthy structure of liver at centri-lobular and portal areas was seen (Figure 1a and inset, respectively). In comparison, livers of rats that received a single LF dose evidenced enlargement of blood vessels and sinusoids that were dilated relative to the same structures in control rat tissues (Figures 1b–f). Disorderly hepatocyte cords and a loss of connection between hepatocytes were also noted (Figures 1c and f). Enlarged spaces between hepatocytes and between hepatocytes and sinusoidal endothelial cells were also an important finding (Figures 1c and d). Some of the cells that made up the cords did not have regular borders and were fragmented; several of these cells were also filled with vacuoles (Figure 1d) and, in some, perinuclear vacuoles were evident (Figure 1d).

Figure 1. Representative liver section light microscopy photographs. Sections of livers from control (a) and LF treated (b–f) adult rats. Thick arrowhead: smaller cells with eosinophilic cytoplasm and damaged nuclei; asterisk: dilated sinusoid; thick arrow: giant hepatocytes; waved arrows: peri-nuclear and cytoplasmic vacuoles; double asterisks: cellular debris; K: Kupffer cell; thin arrowhead: thinned hepatocyte cords. Eosinophilic-dyed infiltrates can be seen in (e); (a) shows a portal area structure in livers of control rats. Scale = 50 µm.

In addition, in the livers of LF-treated rats, cells with vacuoles or with a more eosinophilic cytoplasm (compared to in other cells) were also found within the liver parenchyma (Figures 1b and d); some of these had dark basophilic nuclei (Figures 1b, d and f). In addition, there were some inflammatory cells within the sinusoids and the number of sinusoids lining cells was increased (Figures 1b and e). Further, the nuclei of hepatocytes in LF-treated rats lost their rounded appearance and some were irregular-shaped (Figures 1b and d). Vein branches (and some sinusoids) containing eosinophilic-dyed infiltrates, as well as blood clots and red blood cells, were also seen with increasing frequency (Figure 1e). In this parenchyma, many enlarged Kupffer cells were encountered (Figure 1b) and macrophages bearing eosinophilic cytoplasm were detected in zones of dead cells. In the portal areas of these livers, eosinophilic-staining hepatocytes were apparent (Figure 1b) and hepatocytes displayed partial/total loss of cytoplasm.

Histological results – Electron microscopy findings

In the sections of livers from the adult control rats, healthy ultra-structures of the liver were evident (Figures 2a–d). In contrast, in liver sections from rats treated with LF, microvilli on hepatocytes were smoothed (Figure 2g) and the smooth endoplasmic reticuli in these cells were dilated and enlarged (Figure 3b). It was also noted that boundaries of mitochondria in these cells were irregular and organelle contents were dispersed in the cytosol (Figures 2g and 3b and e). Dense chromatin from the nuclei of liver cells was also seen (Figures 3a–d). The cytoplasm of the hepatocytes was also found to contain large stores of glycogen (Figure 3a). On the sinusoidal side of the hepatocytes, large lipid droplets were seen (Figures 2e and 3a). The distances between hepatocytes and between hepatocytes and sinusoids were also expanded (Figures 2f and g). In the LF rats, Kupffer cells were larger than in control rat livers, had more cytoplasm (Figures 2e and f), and contained many primary and secondary lysosomes (Figure 2f). In addition, some primary lysosomes and membranes on hepatocytes showed myelin degeneration (Figure 2e). Moreover, many erythrocytes, thrombocytes, and fibrin were detected in the liver sinusoids (Figure 3f). Lastly, the livers of these hosts also showed increases in levels of connective tissue fibers with peri-sinusoidal localization and expanded space of Disse (Figure 3e).

Figure 2. Representative liver section electron microscopy photographs. (a–d) Samples from control rats. n, hepatocyte nucleus; m, mitochondria; er, smooth endoplasmic reticulum; ger, rough endoplasmic reticulum; lz, lysosome; s, bile canalicule; gl, glycogen; j, junctional complex; He, hepatocytes; E, endothelial cell; si, sinusoid; er, erythrocyte; K, Kupffer cell; asterisk shows space of Disse. (e–g) Samples from LF-treated rats. L, lipid droplet; v, vacuole; n, hepatocyte nucleus; E, pycnotic endothelial cell nuclei; er, erythrocyte; sl, secondary lysosome in Kupffer cell; He, hepatocytes; m, mitochondria; black arrow: flattened hepatocyte microvilli; white arrow: electron-dense material in space of Disse similar to cytoplasmic particles containing cellular debris; black arrowheads: myelin-like membranes of primary lysosomes; asterisk shows enlarged distance between cells, including bile canalicule. Scale = 0.5 µm.

Figure 3. Representative liver section electron microscopy photographs (a–f). Samples from LF rats. L, lipid droplet; m, mitochondria; ger, rough endoplasmic reticulum; gl, glycogen deposits; st, cytoplasm; He, hepatocyte; E, endothelial cell; si, sinusoid; arrow: dilated smooth endoplasmic reticulum; er, erythrocyte; mf, myofibroblast; f, fibrin deposition, t, thrombocyte; black arrow, perisinusoidal localization of connective tissue fibers; asterisk: expanded space of Disse; n, nucleus. Scale = 0.5 µm.

Discussion

As this study sought to detect whether anthrax lethal factor (LF) alone could cause liver damage, livers from LF-treated rats were analyzed by light and electron microscopy (EM). For this protocol, intraperitoneal (IP) administration of LF was employed for easy standardization (Kandadi et al., 2012; Moayeri et al., 2003, 2005) In terms of duration of the experiment, our approach was based on studies performed earlier. For example, Warfel & D’Agnillo (2011) observed optimal effects from lethal toxin 24 h after administration. Similarly, Moayeri et al. (2003) detected significant histopathological changes in the liver 24 h after an IP dosing of LT.

In the light microscopy evaluations here, sinusoidal dilation was seen in tissues of LF-treated rats. In the literature, many studies note that anthrax toxins cause generalized vascular pathologies and organ failures (Grinberg et al., 2001; Moayeri et al., 2003); however, none as yet had reported sinusoidal dilatation in the liver. Only Duong et al. (2006) reported that congestion developed within hepatic portal vessels, sinuses, and central veins following treatment of mice with LT. The present studies also showed that clot formation and red blood cell accumulation in blood vessels was increased, and that endothelial cells had undergone hypertrophy, as a result of the single dose with LT. With regard to these findings, Hanna et al. (1993) reported that B. anthracis or its related toxins caused vascular permeability changes, inflammation, and increases in tumor necrosis factor (TNF)-α production. Interestingly, others have claimed the pathologic damage after exposure to the LT occurs independently from any change in expression of cytokines (see Moayeri et al., 2003).

Guarner et al. (2003) suggested that the toxic effects of LT might be caused by hemorrhage, pleural effusion, and vasculitis. Vascular dysfunction is an important component in the pathogenesis of anthrax, although the underlying cause of this dysfunction is not yet known (Cui et al., 2004; Fritz et al., 1995; Shieh et al., 2003). In the current study, we found evidence of increased Kupffer cell activity in response to hepatic inflammation. In addition, in liver sections of rats that underwent the single LF dose, cords of the hepatocytes were broken down. This might have evolved from hypoxic damage to hepatocellular proteins or necrosis among the host hepatocytes. Indeed, Moayeri et al. (2003) reported similar findings, i.e. when LF and PA formed lethal toxin (LT), hosts suffered vascular insufficiency, tissue hypoxia, and coagulative necrosis.

Studies about the effects of anthrax on liver pathology and vascular damage have not yet provided detailed histologic analyses at the cell organelle level (Hanna et al., 1993; Grinberg et al., 2001). The current study presents some novel findings, including vacuolar degeneration of hepatocytes and loss of intercellular connections. In addition, there were increases in glycogen content and large lipid droplets in the cytoplasm of hepatocytes of LF-treated rats. The increased lipid droplets and glycogen content may be signs of, respectively, cell degeneration (Cotran et al., 1998) or regeneration (Kudryavtseva et al., 1998). Moreover, mitochondrial degeneration in hepatocytes of LF-treated hosts could be related to increases in oxidative stress (Akçay et al., 1995; Chen et al., 2010; De Lisio et al., 2011). We believe a key reason for the cell degeneration seen in response to LF was formation of free radicals and subsequent damage that occurred as a result of the cells ultimately becoming hypoxic. Further biochemical and molecular analyses of the changes in these cells are needed to provide support for this line of thought.

Although LT formed by the combination of LF and PA is active in many cell types, it leads to lysis in only certain types (Singh et al., 1989). In the detailed histopathological assessments done in the current study, it was determined that hepatocytes lost their proper boundaries, gained an irregular appearance, and discharged cytoplasmic contents. Furthermore, their nuclei were irregular and associated chromatin was dissolved. These findings might indicate that these cells underwent lysis. It is known that LF causes lysis by protein degradation in some cell types, including antigen-presenting cells (Baldari et al., 2006). In our opinion, proteolytic activity induced by LF (Tang & Leppla, 1999) in macrophages is likely to occur in other cell types. In fact, some researchers have suggested that hepatocytes could be susceptible to lytic effects caused by LT (Moayeri et al., 2003). However, evidence of this is still lacking in the literature.

To date, EM studies that dealing with B. anthracis usually focused on the structure of the organism itself (and its spores) or its entry into cells. The study here is the first evaluation of the effects of anthrax components at the EM level. Liver samples from treated rats evidenced expansion in cisterns of the smooth endoplasmic reticulum, lipid vacuoles, irregular sinusoidal walls, and hypertrophic Kupffer cells (with eosinophilic cytoplasm). In addition, the Kupffer cells were large and had many secondary lysosomes. This finding suggested to us that activation of Kupffer cells was increased by LF. The expanded space of Disse seen in tissues from the LF-treated rats may have developed as a result of increased vascular permeability. This elevated permeability may also have been a factor contributing to the increased distances seen between liver cells (between both hepatocytes and between hepatocytes and sinusoidal cells) and contributed to the smoothening of the surfaces of the hepatocytes.

While the mechanism(s) underlying the observed effects on the livers of LF-treated hosts remain undefined, some research has indicated that LF causes dysfunction of cell permeability by blocking activation of the p38 kinase barrier-augmenting pathway (An et al., 2005; Liu et al., 2009, 2010). It was claımed by these authors that, as a result of this, cell cytoskeletal components were significantly altered by pmHSP27 over-expression. Langevin et al. (2005) reported that LF also caused reduced junctional expression of E-cadherin and cadherin trafficking to adherent junctions. Gong et al. (2008) noted that LT reduced total VE-cadherin gene and protein expression and changes in VE-cadherin expression correlated with the appearance of actin stress fibers. These changes occurred in concert with increases in phosphorylation of stress fiber-associated protein myosin light chain (p-MLC) and cleavage of Rho-associated kinase-1, the latter that is known to regulate tight junction integrity (by phosphorylation of claudin-5 and occludin). D’Agnillo et al. (2013) recently suggested that LT caused reductions in trans-endothelial electrical resistance and of claudin-5 levels in endothelial tight junctions. Loss of trans-endothelial electrical resistance and claudin-5 is known to precede formation of actin stress fibers, inter-cellular gaps, and adherent junction disorganization in cells/tissues (Warfel et al., 2005). Other studies have shown that LT could block cadherin localization to adherent junctions (Guichard et al., 2010). While most of the above-cited studies generally focused on effects of LT on endothelial cells, it would not be unimaginable that the same molecular events could also occur in hepatocytes and be responsible, in part, for much of the damage seen here in the livers of the LT-treated rats.

Conclusion

The current literature is mostly lacking in reports on the pathologies arising from LF exposure. Published studies usually mention effects of B. anthracis or double or triple combinations of their associated factors (Firoved et al., 2007; Turk, 2007). Although Kuo et al. (2008) used only LF in their studies; they did not note pathological effects. The most important result obtained in the current study was that the lethal factor of B. anthracis (without protective antigen) led to tissue (hepatic) damage. As similar results were noted in previous studies using LT (i.e. combination of PA + LF) (Moayeri et al., 2003), we conclude that LF alone – when provided as a single dose – can induce toxic changes in the liver of a host.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. This work was supported by funds from Ataturk University (#2007/13).

Acknowledgements

The authors want to thank Ozgen Vuraler for performing the electron microscopy process and Associate Professor Dr Ahmet Ozbek for kindly reading the slides.