Amifostine Protects Bone Marrow from Benzene-Induced Hematotoxicity in Mice

Abstract

Benzene is one of the most widely used industrial chemical agents. Long-term benzene exposure causes bone marrow aplasia and leads to a wide range of hematopoietic disorders including aplastic anaemia (AA). There are currently no effective approaches to protect people from benzene-induced hematotoxicity and AA. In addition, current treatments for AA have limitations with short- and long-term risks. Protective agents and new therapeutic approaches, therefore, are needed to prevent and treat the disease. Amifostine is a well-known cytoprotective agent and has been widely used in clinical for protecting normal tissues from the toxic effects of chemotherapy and radiotherapy. The authors utilized an established mouse model to determine the protective effect of amifostine on benzene-induced bone marrow hematotoxicity. Whole-blood cell count, morphological and histopathological alterations in the bone marrow and spleen, as well as the production of inducible toxic oxidative species were examined and compared among the mouse groups. Amifostine treatment in benzene-exposed mice significantly improved blood cell counts, and morphological and histopathological signs of hematotoxicity in the bone marrow as well as in the spleen. Moreover, amifostine prevented benzene-induced bone marrow and spleen cell apoptosis and rescinded the inhibition of cell proliferation induced by benzene exposure. Finally, amifostine significantly inhibited the levels of reactive oxidative species and lipid peroxidation induced by benzene exposure. These data suggest that amifostine appears to have substantial protective effect on benzene-induced bone marrow hematotoxicity.

Aplastic anemia (AA) is a rare, potentially life-threatening failure of hemapoiesis characterized by pancytopenia and bone marrow aplasia (Brodsky and Jones 2005; Young 2002). The incidence of the disease in Asia is two- to threefold higher than that in the Western countries (Issaragrisil et al. 2006). Most cases of AA are acquired in addition to unusual inherited forms of the disease. Although acquired AA is defined as an autoimmune disorder mediated by cytotoxic T lymphocytes that destruct hematopoietic stem cells (Smith 1996; Zeng et al. 2001), it can result from direct physical and chemical destruction of hematopoietic cells in the bone marrow.

Benzene is one of the most widely used industrial chemical agents. Long-term benzene exposure causes a wide range of hematopoietic disorders including AA characterized by severe bone marrow aplasia (Smith 1996). A recent epidemiological study from Thailand demonstrated that more than 4 days’ total exposure to benzene was associated with a 3.5-fold increase in relative risk for developing AA (Issaragrisil et al. 2006). The positive association of benzene exposure with hemototoxicity was further confirmed by several other studies. Paci et al. and Yin et al. reported that increase in cases of marrow failure and a higher incidence of AA were observed among Italian shoe workers (Paci et al. 1989) and Chinese workers in a variety of industries (Chen and Chan 1999; Yin et al. 1987, 1996) exposed to high levels of benzene contained in the adhesives used in the shoe-making process and other industrial organic solvents. In addition to the industrial exposure, benzene exposure can also occur from cigarette smoking, combusting and evaporation of gasoline, and automobile emissions. Therefore, benzene-exposure becomes an increasing problem of environmental health to general public worldwide (Gist and Burg 1997). Moreover, recent study revealed that hematotoxicity can be induced in individuals exposed to benzene at or below the current U.S. occupational standard of 1 ppm (Lan et al. 2004). Despite the increasing risk of benzene exposure, there are currently no effective and feasible approaches to protect the industrial workers and other individuals from benzene-induced hematotoxicity. On the other hand, current treatments for acquired AA, such as allogeneic bone marrow transplantation and immunosuppressive therapy, have limitations with short- and long-term risks. Moreover, significant side effects, such as fatal fungal infection after immunosuppression, remain to be a major problem (Marsh 2005). Protective agents and new therapeutic approaches with higher efficacy and lower side effects are also needed to treat the disease.

Amifostine is a well-known cytoprotectective agent and has been used in clinical for protecting normal tissues from the toxic effects of chemotherapy and radiotherapy. At the molecular level, amifostine affects redox sensitive transcription factors, gene expression, chromatin stability, and enzymatic activity. At the cellular level, it plays roles in the regulation of cell growth and cell cycle progression (Culy and Spencer 2001). In the present study, we demonstrate that amifostine protects the bone marrow hematopoietic progenitor cells from benzene-induced apoptosis in vivo. In addition, amifostine can inhibit benzene-induced oxidative reaction in mice. These data imply a substantial protective effect of amifostine on benzene-induced hematotoxicity.

DESIGN AND METHODS

Animals

A total of one hundred male CD1 (8 to 12 months old) mice were used in the study. Mice had 10 to 17 days of acclimation, within 8-m³ inhalation chambers, breathing ambient air. The animals were placed in cages (five mice per cage) where food and water were available ad libitum. Benzene-induced hematotoxicity mouse model was established according to the published methods (Farris et al. 1997; Velasco Lezama et al. 2001).

Experimental Design

Mice were randomly divided into four groups (25 mice in each group) including two control groups and two experimental groups. Group 1, mice were injected with corn oil (2 ml/kg) only as control; group 2, mice were treated with 2 ml/kg (equal to 1940 mg/kg body weight) benzene mixed with equal volume of corn oil; group 3, mice were injected with amifostine (200 mg/kg body weight), followed by corn oil as second control; group 4, mice were received injections of amifostine (200 mg/kg body weight) 30 min prior to benzene treatment (2 ml/kg body weight mixed with equal volume of corn oil). Benzene–corn oil mixture was injected subcutaneously in the dorsal region, and amifostine (dissolved in saline) was given via intraperintoneal (IP) injections. All treatments were three times a week for a total of 8 weeks.

Tissue Sampling

All mice were killed by CO₂ overdose. Blood samples (0.5 ml) for cell counting were taken from the posterior vena cava and anticoagulated with 1.5 mg dipotassium EDTA/ml of blood. Tibial sections (5 to 10 μm thick) were prepared and mounted on glass slides for hematoxylin and eosin (H&E) and immunohistochemical staining. Bone marrow smear was prepared from the contents of one tibia and subjected to Romanowsky staining as described previously (Kinsey and Watts 1988). Bone marrow cellularity was assessed by injection of 1 ml of saline solution into the femoral medullary channel, and cell suspensions were diluted with Turk’s solution (1:20), as described previously (Turton et al. 2006; Velasco Lezama et al. 2001). For bone marrow progenitor cell assay, the contents of one femur were flushed into 1.0 ml of Iscove’s modified Dulbecco’s medium (IMDM) under sterile conditions.

Spleen was removed, weighted, placed in Bouin’s solution, and washed with a 1:1 mixture of 100% ethanol and xylene, followed by washing in absolute xylene twice. The spleen was sectioned transversely with a razor blade and with each half held with forceps, the cut surface was lightly touched repeatedly on a glass microscopic slide and subjected to H&E staining and immunohistochemical analysis. The thickness of the spleen sections was 4 μm.

Sample Analysis

Blood Cell Counts

Methods for blood cell counts have been described previously (Velasco Lezama et al. 2001). Briefly, whole blood was diluted in Gower’s liquid (1:200) in red cell pipettes for erythrocyte (RBC) counts, and blood was diluted in Turk’s solution (1:20) for leukocyte (WBC) quantification. Cells were counted with a hemacytometer by light microscopy. Blood was diluted with 1% ammonium oxalate (1:200) for platelet quantification using a phase-contrast microscope. Blood smears were prepared and stained with Wright’s stain and new methylene blue for differential leukocyte and reticulocyte counts, respectively. Hemoglobin concentration was determined by the cyanometa-hemoglobin method (Williams et al. 1991).

Bone Marrow Cellularity

The femoral bone marrow suspension was used to obtain a nucleated cell count from the basophil channel of the H*1. Because the presence of fat particles was evident in the analyzer cytogram, correction of the counts was performed according to Bentley et al. (1995).

Bone Marrow Progenitor Cell Analysis

Femoral bone marrow cell suspension in IMDM (0.3 ml) was resuspended in 3.0 ml of Methocult GF M3434 culture medium (Life Technology) containing 0.9% methylcellulose, 15% fetal bovine serum (FBS), 0.1 mM mercaptoethanol (Sigma, St. Louis, MO), 2 nM L-glutamine, 1% bovine serum albumin (BSA), 10 μM/ml insulin (Sigma), 200 μg/ml human transferrin, 10 ng/ml recombinant murine interleukin (IL)-3, 10 ng/ml recombinant human IL-6 (R&D System), 50 ng/ml recombinant murine stem cell factor (Stem Cell Technology, Vancouver, Canada), and 3 U/ml recombinant murine erythropoietin (Amgen, Thousand Oak, CA). The tubes were vortexed and allowed to stand for 5 min, at 37°C. Cells were dispensed into culture plates and incubated at 37°C. Erythroid burst-forming units (BFU-E) were counted after 10 days of incubation. Erythroid colony-forming units (CFU-E), and granulocyte-macrophage colony-forming units (CFU-GM) were counted after 12 days of incubation.

Assessment and Enumeration of Bone Marrow Cells

Bone marrow smears were stained with May-Grunwald-Giemsa (MGG) stain as described by Dacie and Lewis (1975) and examined microscopically for morphological assessment and the enumeration of myeloid, erythroid, lymphoid, and megakaryocytes in 200 cell-differential counts.

TUNEL Assay

TUNEL (deoxynucleotidetransferase-mediated dUTP nick end labeling) assay was conducted using the ApopTag in situ apoptotic detection kit (Oncor, Gaithenberg, MD) according to the manufacturer’s instruction. The labeled cells were examined using a fluorescent microscope. Any yellow or green or brown staining cells were considered as a single countable apoptotic cell. The percentage of apoptotic cells (from a total of 500 cells) was estimated and scored under a fluorescent microscope examined by two blinded observers on three separate occasions in a coded manner.

Immunohistochemistry

Immunohistochemical staining for the expression of PCNA was carried out as previously described (Liu et al. 2000). Briefly, tissue specimens were deparaffinized, treated with 3% hydrogen peroxide, followed by incubation with the appropriate blocking serum and incubated with primary antibodies against PCNA (BD Biosciences, San Diego, CA) for the detection of cell proliferation. Staining was carried out using the avidin-biotin complex method with reagents from Vector Labs (Burlingame, CA). The intensity and extent of positivity of every stained specimen were estimated by the percentage of positive-stained cells (from a total of 500 cells) scored under a light microscope examined by two blinded observers on three separate occasions in a coded manner.

Reactive Oxygen Species and Lipid Peroxidation Assays

Intracellular reactive oxygen species (ROS) levels were determined in bone marrow cells harvested from femurs of the control versus treated mice, and assayed by measuring the oxidative conversion of cell permeable 2′,7′-dichlorofluorescein diacetate (DCFH-DA) to fluorescent dichlorofluorescein (DCF). ROS assay kit was purchased from Beyotime (Jiangsu, China) and used according to the manufacturer’s instructions. Lipid peroxidation levels in control vs. treated mice were assayed by measuring plasma malondialdehyde (MDA) levels using the thiobarbituric acid (TBA) method as described previously (Rao et al. 1989). The MDA-TBA adducts formed in mice plasma were separated and quantified by HPLC with fluorometric detection.

Statistics

Results were expressed as mean ± SEM. One-way analysis of variance (ANOVA) was used to compare each parameter. Comparisons between groups were performed using post hoc Bonferroni t test when the analysis of variance (ANOVA) test was statistically significant. p< .05 was considered significant.

RESULTS

Protective Effect of Amifostine on Benzene-Induced Losses of Blood Cell Counts

Benzene-induced hematotoxicity causes significant decreases in the numbers of RBC and WBC in whole blood (Farris et al. 1997; Velasco Lezama et al. 2001). We examined cell counts of WBC, RBC, platelet, and reticulocyte, as well as the content of hemoglobin in whole blood derived from the mice with various treatments. As shown in Tab. 1, mice exposed to benzene for 8 weeks showed a significant decrease in the counts of all types of blood cells and a lower concentration of hemoglobin, as compared to control group. In contrast, treatment with amifostine in benzene-exposed resulted in significant increases in blood cell counts and hemoglobin levels, suggesting a protective effect of amifostine on benzene-induced reductions of peripheral blood cells and hemoglobin.

Amifostine Prevents Bone Marrow from Benzene-Induced Damage

Severe bone marrow hypoplasia is an important feature of benzene-induced hematotoxicity (Farris et al. 1997; Velasco Lezama et al. 2001). We examined and compared morphological and pathological alterations in the bone marrow among the mouse groups. Continuous exposure to benzene for 8 weeks significantly reduced the numbers of nucleated bone marrow cells (Fig. 1A ) and megakaryocytes (Fig. 1B ). In contrast, administration of amifostine to benzene-exposed mice led to a significant improvement of these cells from benzene-induced losses. We also examined the composition of the bone marrow cells, and observed that the percentage of granulocytic precursors and neutrophils in total bone marrow cells was significantly reduced from 50% (control mice) to 26.5%, this decrease was accompanied by increases in erythroid precursors (from 19.5% to 30.8%) and lymphocytes (from 28.1% to 37.3%) in benzene-treated mice. However, no significant differences were seen in these parameters (versus control) between benzene-exposed mice treated with amifostine and control mice (Fig. 1C ).

We next determined the effects of amifostine on bone marrow hematopoietic progenitor cell replication and the potential of further differentiation to erythropoietic and granulopoietic lineages. As shown in Fig. 1D , benzene exposure induced significant decreases in both BFU-E and CFU-E. Benzene exposure also suppressed CFU-GM in the bone marrow. In contrast, treatment with amifostine in the benzene-exposed mice significantly increased BFU-E, CFU-E, and CFU-GM units to the levels similar to the control values. These results suggest a role for amifostine in preventing benzene-induced damage in bone marrow hematopoietic progenitor differentiation.

Histomorphometric analysis (H&E staining) on tibial sections was next performed, and revealed that benzene-exposed mice showed a significant reduction in bone marrow hematopoietic cells, and this reduction was accompanied by increased number of fat cells (Fig. 2). In addition, Romanowsky-stained bone marrow smears, derived from benzene-exposure mice (Fig. 3b ) displayed a variety of nuclear/cytoplasmic dyscrasias, including nuclear and cytoplasmic blebbing, vacuolization, atypical mitotic figures, and significant depletion of nucleated bone marrow cells and other hematopoietic progenitors. There were no megakaryocytes presented in the bone marrow smears. In contrast, treatment with amifostine in benzene-exposed mice showed significant alleviation in benzene-induced hematotoxicity in bone marrow damage. As shown in Fig. 3d , amifostine treatment in benzene-exposed mice caused an increase in the numbers of normoblasts, myeloid, and other nucleated bone marrow cells. Romanowsky-stained bone marrow smears also showed increased numbers of erythoid cells, erythoid precursors, and megakaryocytes.

Amifostine has been reported to play important roles in protecting normal cells from apoptosis and stimulating proliferation of several cell lines (Culy and Spencer 2001). We further determined whether amifostine can protect bone marrow cell from benzene-induced apoptosis and promote cell proliferation. As shown in Fig. 4A , TUNEL assay demonstrated that whereas exposure to benzene induced an increase in bone marrow cell apoptosis, as compared to control, amifostine treatment significantly reduced the number of apoptotic cells in the bone marrow, suggesting a role for amifostine in protecting bone marrow cells from benzene-induced apoptosis. Immunohistochemical staining [using Proliferating Cell Nuclear Antigen (PCNA) antibodies] revealed increased counts of positive-stained bone marrow cells (Fig. 4B ), suggesting that amifostine is able to rescind the inhibition of bone marrow cell proliferation induced by benzene exposure.

Amifostine Protects Spleen from Benzene-Induced Damage

Spleen normally regulates peripheral erythrocytes and platelet counts by removing aged or damaged cells (Tavasoli 1991). Exposure to benzene causes serious spleen damage and malfunction. We determined whether amifostine is able to alleviate spleen’s damaging from benzene-induced toxicity. General examination on the spleens derived from the benzene-exposed mice demonstrated significant reduction in size (25% decrease over control) and weight (48% decrease over control) of spleen, whereas mice treated with both amifostine and benzene showed a significant improvement in these parameters (Fig. 5A ). Histopathologic analysis revealed that the spleens from benzene-exposed mice were shrunken and collapsed with wrinkled capsule. We also observed an increased number of megakaryocytes in these spleens, which represented extramedullary hematopoiesis secondary to anemia, suggesting a compensatory reaction, presumably, resulted from the inhibition of bone marrow function in these mice. In contrast, benzene-exposed mice treated with amifostine were demonstrated significant improvement in histopathologic signs of spleen evidenced by increased units of well-organized red pulp (Fig. 5B ). We further examined spleen cell apoptosis and proliferative index in control versus treated mice. TUNEL assay revealed that benzene exposure resulted in an increased spleen cell apoptosis, whereas mice received a combination treatment with benzene and amifostine showed a significant reduction in the number of apoptotic cells (Fig. 4A ). In addition, immunohistochemical analysis demonstrated that although amifostine alone had no effect on spleen cell proliferation, as determined by staining with PCNA antibodies, it rescinded the inhibition of spleen cell proliferation induced by benzene exposure (Fig. 4B ).

Amifostine Inhibits Inducible Reactive Oxygen Species and Lipid Peroxidation by Benzene Exposure

Benzene-induced hematotoxicity is related to the metabolism benzene to its reactive intermediates (Hiraku and Kawanishi 1996; Smith 1996). These metabolites can travel to the bone marrow where they are activated by peroxidase and other enzymes to produce highly reactive free radicals including active oxygen species and other toxic species (Eastmond, Smith, and Irons 1987; Hiraku and Kawanishi 1996; Subrahmanyam et al. 1991). We examined the levels of reactive oxygen species (ROS) and lipid peroxidation (plasma MDA) in the mice of each group. As demonstrated in Fig. 6, benzene-exposed mice had significant increases in both intracellular ROS levels of the bone marrow cells and plasma MDA production, as compared to the control groups. In contrast, no demonstrable induction in these parameters was observed in the benzene-exposed mice that were treated with amifostine. These results suggest an additional mechanism underlying the protective effect of amifostine on benzene-induced hematotoxicity by inhibiting the production of inducible toxic oxidative species.

DISCUSSION

Benzene, a commonly used industrial solvent and a component of gasoline, has been reported to induce several hematopoietic disorders, including leukemia and AA (Farris et al. 1997). Although long-term and high-dose benzene exposures have been known to have toxic effects on the blood and bone marrow (Aksoy 1989), recent study indicates that hematotoxicity can be induced by exposure to very low levels (at or below the current U.S. occupational standard of 1 ppm) of benzene (Lan et al. 2004). Exposure to benzene now occurs worldwide and is becoming a serious problem to human health, particularly to workers in the oil shipping, shoe manufacture, automobile repair, and other industries. Effective pharmacological intervention is, therefore, needed to prevent the disease. In addition, new therapeutic approaches with higher efficacy and less side effects are needed to treat the disease.

A mouse model for benzene-induced hemototoxicity was established previously by subcutaneous administration of benzene (Farris et al. 1997; Velasco Lezama et al. 2001). In the present study, we utilized this in vivo model to investigate the protective effect of amifostine, a well-known cytoprotector, on benzene-induced hematotoxicity. Although mice showed the signs of persistent hematotoxicity after the 8 weeks of benzene exposure, amifostine treatment significantly increases blood cell counts, spleen weights, and alleviates benzene-induced morphological and histopathological damages in the bone marrow and spleen. Moreover, our data show that amifostine significantly suppresses benzene-induced cell apoptosis and prevents benzene-induced inhibition of cell proliferation in the bone marrow and spleen of the mice. Finally, we demonstrate that amifostine is able to inhibit benzene-induced increase in both intracellular ROS and plasma MDA, suggesting another mechanism underlying protective effect of amifostine on benzene-induced hematotoxicity by inhibiting the production of inducible toxic oxidative species. These observations suggest that amifostine exists substantial, if not fully, protective effects on benzene-induced hematotoxicity.

In the present study, we observed that the major damage induced by benzene exposure occurred in the bone marrow of the mice, consistent with previous reports (Farris et al. 1997; Marsh and Testa 2000; Velasco Lezama et al. 2001; Young 2002). Our data indicate that one of the most significant effects of amifostine on protecting mice from benzene-induced damage also occurred in the bone marrow environment, as evidenced by increased number of bone marrow nucleated cells, promoted formation of colony-forming progenitors of mature erythroid cells, granulocytes, and megakaryocytes, and reduced apoptotic rate of bone marrow cells. In addition, we demonstrate that amifostine inhibites benzene-induced elevation of intracellular ROS and plasma lipid peroxidation levels. These observations are in agreement with the previous reports that the critical roles of amifostine in protecting normal tissues from radiotherapy and chemotherapeutic agents are to modify the DNA target of the agents and to scavenge excess oxygen free radicals induced by toxic substances (Culy and Spencer 2001; List et al. 1996), and indicate that the major mechanism underlying protective effect of amifostine is to prevent bone marrow cells from oxidative DNA damage and apoptosis induced by benzene metabolites and rescinded the inhibition of bone marrow cell proliferation in the presence of benzene.

The in vivo model used in this study enabled us to explore a novel role for amifostine in protecting benzene-induced hematotoxicity. Our data indicate that amifostine significantly alleviates benzene-induced hematotoxicity in mice. Although reduction in benzene exposure using engineering controls such as substitution for benzene by other solvents, to limit exposure is the preferred procedure, the data derived from this study provide a rationale to use amifostine as an alternative, or supplementing, approach to prevent benzene-induced hematotoxicity. Further studies with various treatment schema, however, are needed to establish the optimal doses and timing for the drug. In addition, more studies are needed to determine if animals already impacted by benzene could be treated with amifostine to reverse the effects of benzene. These studies will determine whether amifostine can be used as a protective and/or therapeutic agent, either alone or in combination with other forms of treatment to enhance their efficacy.