|Home | About | Journals | Submit | Contact Us | Français|
Exposure to environmental lead (Pb) is a mild risk factor for Amyotrophic Lateral Sclerosis (ALS), a paralytic disease characterized by progressive degeneration of motor neurons. However, recent evidence has paradoxically linked higher Pb levels in ALS patients with longer survival. We investigated the effects of low-level Pb exposure on survival of mice expressing the ALS-linked superoxide dismutase-1 G93A mutation (SOD1G93A). SOD1G93A mice exposed to Pb showed longer survival and increased expression of VEGF in the ventral horn associated with reduced astrocytosis. Pretreatment of cultured SOD1G93A astrocytes with low, non toxic Pb concentrations upregulated VEGF expression and significantly abrogated motor neuron loss in co-culture, an effect prevented by neutralizing antibodies to VEGF. The actions of Pb on astrocytes might explain its paradoxical slowing of disease progression in SOD1G93A mice and the improved survival of ALS patients. Understanding how Pb stimulates astrocytic VEGF production and reduces neuroinflammation may yield a new therapeutic approach for treating ALS.
Amyotrophic lateral sclerosis (ALS) is an adult-onset neurodegenerative disorder characterized by the selective death of motor neurons, resulting in progressive paralysis and death. The cause of the disease remains unknown for the majority of sporadic cases of ALS. The discovery of missense mutations in the gene encoding the antioxidant enzyme Cu/Zn superoxide dismutase-1 (SOD1) in a subset of patients with familial ALS (Rosen et al. 1993) stimulated research in transgenic animal models expressing different SOD1 mutations though the pathogenic mechanism underlying the toxicity of SOD1 mutants remains controversial.
Studies in transgenic mice have shown evidence for a non-cell autonomous toxicity of SOD1 mutations. Decreasing mutant SOD1 expression in astrocytes or microglia (Boillée et al, 2006; Yamanaka et al, 2008), and grafting wild-type macrophages to mutant SOD1 mice (Beers et al, 2006), increase the lifespan of mutant SOD1 mice. In addition, in vitro studies have shown that astrocytes expressing the SOD1G93A mutation exert toxic influence to non-transgenic motor neurons (Vargas et al. 2006; Cassina et al. 2008; Nagai et al. 2007; Di Giorgio et al. 2007), suggesting that these glial cells can play a disease-modifying role in ALS pathogenesis (Barbeito et al. 2004).
Occupational or environmental exposure to a variety of toxicants including heavy metals has been evaluated as potential causes of ALS. Exposure to the heavy metal lead (Pb) and in particular blood and bone content of Pb was associated with an increased risk of ALS (Kamel et al. 2005). More recently, epidemiological data also showed that Pb blood and bone levels positively correlate with longer survival in ALS patients after diagnosis (Kamel et al. 2008), suggesting Pb exposure may paradoxically delay the disease progression.
Pb is a widely spread environmental heavy metal with no known specific biological function. Pb has been shown to induce acute and chronic neurotoxicity, particularly during CNS development (Cory-Slechta et al. 1995; Bellinger DC, 2008). Both neuronal and glial cells seem to be affected in Pb neurotoxicity (Tiffany-Castiglioni and Qian 2001). However, the toxicity of Pb on spinal cord motor neurons and astrocytes is presently unknown as is its potential role in the etiology of ALS.
Experimental evidence supports the view that astrocytes can sequester and buffer Pb in the CNS, preventing further diffusion of the metal to the neuronal compartment and subsequent neurotoxicity or altered synaptic transmission (Tiffany-Castiglioni et al. 1993, 2001). In particular, astrocytes are the cells that preferentially induce cytoprotective and antioxidant gene expression in response to Pb (Cabell et al. 2004). Thus, available evidence supports the view that astrocytes are key targets of Pb and respond to it by inducing neuroprotective pathways.
We began investigating the hypothesis that Pb would accelerate ALS in the SOD1G93A mice, but were surprised to find increased survival and reduced astrocytic reactivity. We further investigated this paradoxical finding in astrocyte/motor neuron co-cultures and found an up-regulation of vascular endothelial growth factor (VEGF) after Pb treatment. Our results suggest that Pb activates a novel pathway able to reduce neuroinflammation and slow neurodegeneration in ALS.
Culture media and serum were obtained from Invitrogen (Carlsbad, CA). All other reagents were from Sigma unless otherwise specified.
Procedures using laboratory animals were in accordance with international guidelines and were approved by the Institutional Animal Committee. Sprague-Dawley SOD1G93A L26H rats were kindly provided by Dr David S. Howland (Wyeth Research, Princeton, NJ, USA) (Howland et al. 2002). Transgenic ALS mice carrying the G93A mutation for human SOD1, strain B6SJL-TgN(SOD1-G93A)1Gur (Gurney et al. 1994), were obtained from Jackson Laboratory (Bar Harbor, ME, USA) and genotyped as previously described (Vargas et al. 2005). Mice were housed under controlled conditions with free access to food and water. Male transgenic (n=8 per group) and non-transgenic (n=7 per group) littermates were divided randomly into the following groups: A) control group which received sodium acetate in drinking water, at a concentration of 200 ppm, B) Pb treatment group, administered with Pb acetate (PbAc) in drinking water at a concentration of 200 ppm. The treatment was performed from weaning (21-25-days-old) to death. Animals were observed weekly for onset of disease symptoms, as well as progression to death. Onset of disease was scored as the first observation of abnormal gait or overt hind limb weakness. End-stage of the disease was scored as complete paralysis of both hind limbs and the inability of the animals to right them after being turned on a side.
Blood was collected at disease onset (between days 90-95). Pb levels were analyzed by Electrothermal Atomic Absorption Spectrometry (ETAAS) with a VARIAN SpectrAA-55. Analytical conditions were validated with quality assurance/quality control (QA/QC) procedure standards guidelines (Parsons 1997; CDC 2005).
Primary rat spinal cord astrocyte cultures were prepared from transgenic SOD1G93A and non-transgenic 1-day-old pups, genotyped by PCR, as previously described (Vargas et al. 2006; Cassina et al. 2002). Briefly, cells were plated at a density of 2 × 104 cells/cm2 in 35-mm Petri dishes or 24-well plates (Nunc, Naperville, IL, USA) and maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum, HEPES (3.6 g/L), penicillin (100 IU/mL) and streptomycin (100 μg/mL). Astrocyte monolayers were >98% pure as determined by GFAP immunoreactivity.
Motor neuron cultures were prepared from embryonic day 15 (E15) rat spinal cord by a combination of metrizamide gradient centrifugation and immunopanning with the monoclonal antibody IgG192 against p75 neurotrophin receptor as described (Cassina et al. 2002), then plated on rat astrocyte monolayers at a density of 300 cells/cm2 and maintained for 48 h in L15 supplemented medium as described (Cassina et al. 2002).
Confluent astrocyte monolayers were changed to L15 supplemented media prior treatment. Stock solution of PbAc (Baker, US) was prepared in distilled water and directly applied to astrocyte monolayers at the indicated concentrations. For co-culture experiments, astrocyte monolayers were treated with PbAc for 24 h and motor neurons were plated in fresh L15 supplemented media, after washing twice with Dulbecco's PBS. For neutralization of VEGF biological activity, anti-rat VEGF antibody (1μg/ml, AF564, R&D Systems) was added to the culture media 2 hs after motor neuron plating. Motor neuron survival was assessed after 48 h by directly counting all p75NTR immunoreactive cells displaying neurites longer than 4 cell bodies diameter (Cassina et al. 2002).
Protein samples extracted from astrocyte monolayers or spinal cord homogenates, were resolved on 12% SDS–polyacrylamide gel and transferred to nitrocellulose membrane (Amersham, Piscataway, NJ, USA) as described (Vargas et al. 2006). VEGF primary antibody (Santa Cruz Biotechnology) was used at 1:250 dilution. Equal protein load was verified by β-actin immunodetection (Sigma, St. Louis, MO) after stripping of the membrane. For spinal cord western blots, five transgenic mice at the age of symptoms onset and the same number of age-matched non-Tg mice were used. Spinal cords were removed and immediately frozen in liquid nitrogen, and processed in SDS 1% plus Complete Protease Inhibitor Cocktail (Roche) for protein extraction.
Mice were transcardially perfused with 4% paraformaldehyde fixative in PBS under deep anesthesia. The spinal cords were post fixed and embedded in Paraplast. Serial sections (5 μm-thick) were submitted to antigen retrieval (microwave in citrate buffer) followed by blockage of endogenous peroxidase activity prior to immunohistochemistry. After permeabilization (0.25 % Triton X-100 in 0.1M PBS) and blocking unspecific binding (10% goat serum, 2% BSA, 0.25 % Triton X-100 in PBS), slides were incubated with primary antibodies diluted in blocking solution for 48h at 4°C. The primary antibodies were mouse monoclonal anti-VEGF (1:100, Santa Cruz Biotechnology) and rabbit polyclonal anti-GFAP (1:600, Sigma). Sections were washed and exposed to secondary antibodies (1-2 hours at room temperature), Alexa Fluor488 conjugated goat anti-rabbit (Molecular Probes; 1,5 μg/ml) or anti-mouse biotinylated antibody (Jackson Immunoresearch, 1:400), followed by streptavidin-horseradish peroxidase (Jackson Immunoresearch, 1 μg/ml in Tris-HCl, pH 7.4) and developed with 0.5 mg/ml diaminobenzidine plus 0.01% (v/v) hydrogen peroxide in 0.05 M Tris-HCl. Controls were performed by omitting the primary antibody. Images were obtained using an Olympus FV300 confocal microscopy. Quantification of astrogliosis in the lumbar spinal cord was performed on images obtained from every fifth section (n=15 from each group) using ImageJ software (NIH). Ventral horn area occupied by GFAP immunofluorescence was measured and expressed as a percentage of total ventral horn area in each section.
Survival curves were compared by Kaplan-Meier analysis with the Log-rank test using the GraphPad PRISM software, version 4.03. All culture assays were performed in duplicate and each experiment was repeated at least three times. Quantitative data were expressed as mean ± SEM and ANOVA and Student's t test were used for statistical analysis, with p < 0.05 considered significant. When the normality test failed, comparison of the means was performed by one-way ANOVA on ranks followed by the Kruskal–Wallis test. Data from GFAP were analyzed using a one way ANOVA for repeated measurements, and compared by all pairwise multiple comparison procedures (Holm-Sidak method). All statistics computations were performed using the Sigma Stat System (1994, Jandel Scientific, San Rafael, CA, USA), or GraphPad InStat software, version 3.06.
Lead exposure increased survival of transgenic SOD1G93A mice. SOD1G93A and non-transgenic mice were treated from weaning until death with a low concentration of PbAc or sodium acetate (vehicle) in drinking water (Fig.1). Chronic exposure to Pb did not significantly modify disease onset (vehicle: 101±14 days; PbAc: 110±11 days) but significantly increased life expectancy by 12 days (vehicle n=7: 125±1; PbAc n=8: 137±2 days) (Fig.1 A and B). The Pb treatment caused a 13-fold increase in blood Pb levels as analyzed by electrothermal atomic absorption spectrometry at disease onset. Metal blood concentration in Pb-treated group (27±7.4 μg/dl vs. 2.0±1.0 μg/dl in control mice) is below the accepted level for occupational Pb exposure (40 μg/dL) under Occupational Safety and Health Administration (OSHA) standards (Roscoe et al. 2002). Accordingly, exposed mice behaved normally and did not show apparent signs of Pb intoxication, such as weight loss, disease or premature death, when compared to vehicle-treated mice.
Pb exposure decreased astrocyte GFAP levels in the ventral spinal cord. A group of mice were sacrificed at the first signs of disease onset, as described in methods, to perform microscopic analysis of the spinal cord. There were no apparent differences in motor neuron number or morphology with respect to mice treated with vehicle only. However, Pb-treated SOD1G93A mice displayed a marked reduction in astrocyte reactivity, determined by GFAP immunolabeling, when compared to vehicle-treated group (Fig. 2 A). Quantification of GFAP immunofluorescence showed about 50% reduction in GFAP staining in Pb-treated mouse spinal cord, compared to control littermates (Fig. 2 B).
Pb exposure was previously found to induce VEGF gene expression in immortalized human fetal astrocytes (Hossain et al. 2000). Thus, we analyzed VEGF expression in spinal cord of Pb- and vehicle-treated mice. Western blot analysis of spinal VEGF showed basal increase of VEGF levels in SOD1G93A mice compared to non transgenic mice, which was further elevated by 36% following Pb treatment (Fig. 3 A). Spinal cord VEGF levels in non-transgenic mice were below the detection level in the current analytical conditions and were not increased by Pb.
In agreement with western blot analysis, VEGF immunohistochemistry showed an increase of VEGF immunoreactivity in the ventral spinal cord of SOD1G93A mice treated with Pb (Fig. 3 B). In particular, VEGF immunoreactivity increased in the neuropil around motor neurons. In comparison, VEGF immunoreactivity in non-transgenic mice was limited to the motor neuron cell bodies and was not overtly increased by Pb (Fig. 3 B).
Because Pb induced VEGF expression in the neuropil of SOD1G93A spinal cord mice, we further investigated whether Pb induced VEGF in astrocytes cultures. VEGF expression was basally increased in SOD1G93A astrocyte cultures when compared to non-transgenic astrocytes. In addition, a 24h Pb exposure (10 μM) increased VEGF levels by 2-fold in SOD1G93A astrocytes (Fig. 4). VEGF expression was not detected in non-transgenic astrocyte cultures, which remained below the detection limit even following Pb treatment (Fig. 4).
We have employed our previously validated co-culture assay (Pehar et al. 2004) to analyze whether Pb modulated the ability of SOD1G93A astrocytes to promote motor neuron death. Pb exposure to astrocyte monolayers for up to 2 days did not induce cell death or apparent morphological changes in astrocytes as assayed by microscopic analysis, GFAP immunofluorescence and MTT (data not shown). In contrast, isolated embryonic motor neurons maintained with GDNF were vulnerable to Pb applied to the culture media at concentrations ≥ 10 μM (Fig. 5 A). Pre-exposure of SOD1G93A bearing astrocytes to 10 μM Pb for 24 h caused astrocytes to increase motor neuron survival from 64% (vehicle) to 96% (PbAc), thus partially reverting SOD1G93A astrocyte-mediated neurotoxicity (Fig 5 B). This protective effect was prevented by VEGF-blocking antibodies added to the co-cultures (Fig 5 B) while vehicle-treated cocultures were unaffected by VEGF-blocking antibodies (Fig 5 B) suggesting that VEGF is necessary for protection. In contrast, pre-exposure of non-transgenic astrocytes to Pb made them unable to support motor neuron survival (Fig 5 C) further indicating the specific effect of Pb on SOD1G93A bearing astrocytes.
Paradoxically, low-level Pb exposure was protective in a severe model of ALS, which correlated with higher levels of VEGF in spinal cord. The present finding was unexpected, since Pb is a well-known environmental neurotoxicant and a risk factor for ALS (Kamel et al. 2005). We provide evidence that Pb decreases the toxicity of dysfunctional astrocytes occurring in familial ALS by inducing VEGF in SOD1G93A bearing astrocytes. This response was specific for transgenic astrocytes since VEGF levels in non-transgenic astrocytes were below the limits of detection. The proposed mechanism elicited by Pb might explain recent epidemiological data showing that Pb blood and bone levels positively correlate with longer survival in ALS patients after diagnosis (Kamel et al. 2008). Hence, Pb exposure may potentially have opposing actions during the course of ALS, initially promoting the degeneration of motor neurons but later abrogating damage and neuroinflammation mediated by dysfunctional glia (Fig. 6).
Pb blood levels after chronic exposure of low-level Pb in drinking water were comparable to those reported in humans following occupational exposition (Roscoe et al. 2002; Kosnett et al. 2007). These levels did not produce overt Pb toxicity in mice after 90-100 days of treatment. SOD1G93A mice exposed to Pb developed as their littermates receiving vehicle and disease onset was not statistically different. But survival was significantly increased, suggesting a disease-modifying effect of Pb. Studies in a mouse model of inherited ALS have suggested that onset is related to motor neuron function, while progression depends upon astrocytes and microglia (Yamanaka et al 2008, Boillee et al. 2006). This is consistent with two pathogenic processes responding differently to Pb intoxication.
Protective Pb effects were evidenced by a dramatic reduction of GFAP-immunoreactive astrocytes, indicating a potential downregulation of ALS-linked neuroinflammation (Levine et al. 1999; Barbeito et al. 2004). Because ALS progression is regulated by neuroglia (Yamanaka et al. 2008), progressive Pb accumulation in astrocytes may reduce their deleterious activation. VEGF administration to SOD1G93A mice also reduce glial reactivity (Zheng et al. 2007), consistent with the concept of reducing neuroinflammation.
Astrocytes survive accumulating Pb to concentrations that are detrimental to neurons (Tiffany-Castiglioni et al. 1993). In agreement, both cultured SOD1G93A and non-transgenic astrocytes are resistant to moderate Pb concentrations that otherwise are toxic for isolated cultured motor neurons. However, preconditioning with Pb caused non-transgenic astrocytes to become neurotoxic to co-cultured motor neurons, suggesting it can also elicit inflammatory-like activation of astrocytes comparable to other stressful stimuli (Cassina et al. 2002). In contrast, motor neurons co-cultured on SOD1G93A astrocytes pre-treated with Pb were more likely to survive than on vehicle-treated transgenic astrocytes. These results further suggest that astrocytes bearing the SOD1G93A mutation are regulated differently from non-transgenic astrocytes.
One distinguishing characteristic of Pb-induced VEGF expression is that it specifically occurs in SOD1G93A astrocytes. Non-transgenic astrocytes do not express detectable VEGF levels in response to Pb. VEGF is classically induced by activation of hypoxia-inducible factor-1 (HIF-1) mediated by hypoxic stress and hypoxia-induced generation of reactive oxygen species (Xie et al. 2004; Chandel et al. 2000). However, VEGF expression is also known to be induced by Pb via a PKC/AP1-dependent and HIF-1-independent signaling pathway in an astrocytic cell line (Hossain et al. 2000). Decreased VEGF has been implicated in the pathogenesis of ALS (Oosthuyse et al. 2001; Lambrechts et al. 2003; Storkebaum et al. 2005; Wang et al. 2007). VEGF protects motor neurons in culture and in organotypic cultures (Li et al. 2003; Tolosa et al. 2008; Van den Bosch et al. 2004). Furthermore, VEGF was shown to be released by astrocytes (Van den Bosch et al. 2004) and exert a potent trophic activity for motor neurons by activating VEGFR-2 (Storkebaum et al. 2005). Interestingly, our results indicate that Pb increased VEGF expression in the spinal cord of SOD1G93A mice. Basal VEGF expression was greater in untreated SOD1G93A mice compared to non-transgenic controls as previously observed (Murakami et al. 2003). Nevertheless, the protection is appreciated when basal levels are increased suggesting that additional protective mechanisms may be enhancing VEGF release or bioavailability. Additionally, Pb may modulate ALS pathogenesis by affecting other physiological mechanisms. For example, it has been shown that Pb may decrease SOD activity in lead-exposed rats (Babu et al. 2007), which could have beneficial effects on familiar ALS models. Furthermore, at the cellular level, Pb potently stimulates the antioxidant enzyme hemoxygenase-1 expression in astrocytes (Cabell et al. 2004), suggesting increased antioxidant defenses and cytoprotective pathways that may also contribute to neuronal protection.
Understanding how Pb stimulates VEGF production in non-neuronal cells by the ALS disease process can reveal a simpler therapeutic route than gene therapy approaches under current development.
A. de León, P. Díaz-Amarilla, L. Lafon and N.Guelfi for invaluable help with genotyping, maintenance of the mice colony and lead administration to mice. Dr. Ximena Aguirrezábal for assistance with statistical analysis of survival studies. This work was supported by grants ES00240, AT002034, and NS058628 from the National Institutes of Health and from funding by the Linus Pauling Institute.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.