Manganese intoxication decreases the expression of manganoproteins in the rat basal ganglia: An immunohistochemical study

Brain Research Bulletin 74 (2007) 406–415

Research report

Manganese intoxication decreases the expression of manganoproteins in the rat basal ganglia: An immunohistochemical study M. Morello a , P. Zatta b , P. Zambenedetti c , A. Martorana a , V. D’Angelo a , G. Melchiorri a , G. Bernardi a,d , G. Sancesario a,d,∗ a

Department of Neuroscience, University of Rome Tor Vergata, Rome, Italy b Department of Biology, University of Padua, Italy c Anatomopathology Division and Brain Bank, Dolo General Hospital, Dolo, Venice, Italy d Santa Lucia Foundation, Rome, Italy Received 13 February 2007; received in revised form 12 June 2007; accepted 12 July 2007 Available online 2 August 2007

Abstract Manganese (Mn) is a cofactor for some metalloprotein enzymes, including Mn-superoxide dismutase (Mn-SOD), a mitochondrial enzyme predominantly localized in neurons, and glutamine synthetase (GS), which is selectively expressed in astroglial cells. The detoxifying effects of GS and Mn-SOD in the brain, involve catabolizing glutamate and scavenging superoxide anions, respectively. Mn intoxication is characterized by impaired function of the basal ganglia. However, it is unclear whether regional central nervous system expression of manganoproteins is also affected. Here, we use immunocytochemistry in the adult rat brain, to examine whether Mn overload selectively affects the expression of GS, Mn-SOD, Cu/Zn-SOD, another component of the SOD family, and glial fibrillary acid protein (GFAP), a specific marker of astrocytes. After chronic Mn overload in drinking water for 13 weeks, we found that the number and immunostaining intensity of GS- and Mn-SOD-positive cells was significantly decreased in the striatum and globus pallidus, but not in the cerebral frontal cortex. In addition, we found that GS enzymatic activity was decreased in the strio-pallidal regions but not in the cerebral cortex of Mn-treated animals. In contrast, Cu/Zn-SOD- and GFAPimmunoreactivity was unchanged in both the cerebral cortex and basal ganglia of Mn-treated rats. Thus, we conclude that in response to chronic Mn overload, a down-regulation of some manganoproteins occurs in neurons and astrocytes of the striatum and globus pallidus, probably reflecting the vulnerability of these regions to Mn toxicity. © 2007 Elsevier Inc. All rights reserved. Keywords: Glutamine synthetase; Mn-SOD; Cu/Zn-SOD; Astrocytes; Globus pallidus; Striatum; Cerebral cortex; Cytochemistry; Superoxide; Glutamate toxicity

1. Introduction Manganese (Mn) is an essential trace element in plant and animal life [11,20]. Although, it is known to be potentially toxic to the central nervous system (CNS) [1,5], its mechanism of action is complex and controversial. Mn toxicity is an occupational health hazard in miners and other workers [12,13,41]. Mn has also been identified as a nutritional contaminant in patients receiving long-term parenteral nutrition [14,21], and as an endogenous neurotoxin in patients with chronic liver fail-

∗ Corresponding author at: Department of Neuroscience, University of Rome Tor Vergata, Via Montpellier 1, 00133 Rome, Italy. Tel.: +39 06 20903013; fax: +39 06 72596022. E-mail address: [email protected] (G. Sancesario).

0361-9230/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.brainresbull.2007.07.011

ure [27]. Independent of the route of absorption, Mn overload acts as a toxicant to the brain, resulting in neurological disorders that are characterized by early psychotic symptoms and later on by irreversible parkinsonism, a syndrome known as manganism [5,12,13,35,42]. Unlike Parkinson’s disease, in which the neurodegeneration primarily occurs in the substantia nigra pars compacta, Mn intoxication in humans and animal models results in prominent neuronal loss and gliosis in the globus pallidus and the caudate putamen [42,43,58]. In the past, research focused on understanding the mechanism(s) of Mn neurotoxicity has relied on the identification of a wide spectrum of non-physiological functions gained by a high concentration of Mn in the brain. Early studies hypothesized that Mn toxicity invokes enhanced auto-oxidation of dopamine [1,2,25]. It has also been suggested that the oxidation state of Mn is an important factor that contributes to its cytotoxicity [11].

M. Morello et al. / Brain Research Bulletin 74 (2007) 406–415

Neurodegeneration in the striatum may be linked to a cascade of oxidative damage related to the ease with which Mn can be readily oxidized from the 2+ to the 3+ oxidation state [1,2,26]. Since Mn preferentially accumulates in the mitochondria of basal ganglia, it has been suggested that the mitochondria are target organelles for Mn toxicity [31,38]. In the mitochondria, Mn can disrupt calcium homeostasis and other mitochondrial functions [24,31,34]. Although a wide range of heterogeneous biochemical effects may be linked to Mn toxicity, few studies have investigated whether Mn intoxication affects the homeostasis of the manganoproteins, and their functions and distribution in the brain [18,53]. Mn is an essential cofactor for certain metalloenzymes or manganoproteins, involved in nitrogen and oxygen metabolism [11,52]. Indeed, Mn is specifically required for maximal catalytic activity in many cases. Within the CNS, Mn is a cofactor for manganoproteins such as glutamine synthetase (GS) and superoxide dismutase (Mn-SOD) [2]. GS, selectively expressed in astroglial cells [60], catalyzes the conversion of glutamate to glutamine, thereby preventing an increase in extracellular glutamate levels and glutamate-dependent overexcitation [11,54]. Region-selective distribution of GS has been associated with areas rich in glutamate innervation [40,51]. MnSOD is a mitochondrial enzyme that is predominantly localized to neurons [32]. It specifically regulates and detoxifies superoxide (O2 − ), an extremely powerful oxidant and by-product of cellular metabolism. It is possible that the expression and activity of manganoproteins may be regulated by changes in cellular Mn levels under physiological and pathological conditions. Here, we used immunocytochemistry to examine whether chronic Mn overload affected the expression of GS and Mn-SOD in the rat strio-pallidal complex and cerebral cortex, representing respectively Mn-vulnerable and Mn-resistant brain regions in human and nonhuman primates [39,41,59]. We also examined whether manganoproteins were more sensitive to Mn overload by analyzing the immunocytochemical expression of: a) copper/zinc superoxide dismutase (Cu/Zn-SOD), another component of the SOD family, that is present in neurons and in astrocytes [32]; and b) the glial fibrillary acid protein (GFAP), considered to be a sensitive marker of astroglial function. Finally, we evaluated whether chronic Mn overload affected GS enzymatic activity, since GS is reportedly very sensitive to changes in Mn concentrations [54]. Here, we report that an excess of Mn can impair some biological processes such as expression and activity of manganoproteins, that are normally regulated by Mn trace levels in the brain. 2. Experimental procedure 2.1. Animal protocols Forty male Wistar rats (Charles River, Como, Italy), weighing 85–100 g at the start of the experiment, were randomly divided into two groups and housed in stainless steel cages under stable conditions of humidity (60 ± 5%) and temperature (22 ± 2 ◦ C). They were fed on a standard pellet diet (RF18: Morini, Bologna, Italy), containing Mn (48 mg/kg), and were maintained on a 12 h light–dark cycle (light on from 6 a.m. to 6 p.m.; light off from 6 p.m. to 6

407

a.m.). The experimental protocols conformed to the guidelines of the European Union Council (86/609/EU) and were approved by the Institutional Animal Care and Use Committee of the University of Rome Tor Vergata. Chronic Mn intoxication was induced per os. Briefly, rats had free access to either normal drinking water, or to a solution of MnCl2 ·4H2 O (Sigma, Milan, Italy) in drinking water (20 mg/ml corresponding to about 100 mM of Mn2+ ) [8,30,48]. The average consumption of food (11–15 g/100 g body weight) and drinking water (10–12 ml/100 g body weight) were similar at different time intervals in the two groups during the course of the experiment, as was the average body weight (300–350 g) before sacrifice. The treatment lasted for 13 weeks, and then the rats were killed for chemical, biochemical or morphological studies. For morphological analysis, controls (n = 10) and Mn-exposed rats (n = 10) were deeply anaesthetized with chloral hydrate (400 mg/kg i.p.), followed by transcardial perfusion with the aid of a peristaltic pump (Minipuls 3® , Gilson, Middleton, USA). Initial perfusion consisted of 50 ml of saline and heparin (1%) at room temperature, followed by 250 ml of 4% paraformaldehyde in 0.1 M phosphate buffer (PB) (pH 7.4), at standard flow rate of 12–16 ml/min, and at a maximum pressure of 0.5 MPa. The brains were removed, postfixed in the 4% paraformaldehyde solution for 2 h at 4 ◦ C, then stored in PB overnight. The brains were transferred in PB with 30% sucrose for 48 h, frozen and sectioned in a cryostat (Cryostat Leitz-Leica 1720, Wetzlar, Germany) at 40-␮m thickness. For chemical and biochemical studies, the animals were killed by decapitation. The brain was rapidly removed, rinsed with cold physiological saline to remove any excess blood, then blotted dry with filter paper. The striatum, the globus pallidus, and the sensori-motor cortex were dissected out, and the tissues were either dehydrated overnight in the oven at 100 ◦ C or immediately frozen in liquid nitrogen and stored at −80 ◦ C until analysis.

2.2. Measurement of Mn levels in brain Samples of the different brain regions were digested by addition of four volumes of 65% HNO3 for one week at room temperature. The Mn content was determined in five Mn-treated and five control rats by an atomic absorption spectrophotometer (Perkin-Elmer, 2100) equipped with a graphite furnace with platform and a hollow cathode Mn lamp (absorption at 279.5 nm). The Mn standard used for atomic absorption was purchased from Carlo Erba (Milan, Italy). Palladium (0.005 mg) and Mg(NO3 )2 (0.003 mg) were used as matrix modifiers. An atomization temperature equal to 1900 ◦ C was reached. Data were expressed as ␮g/g dry weight of brain tissue, and Student’s t test was used to evaluate significant differences in the Mn content between control and Mntreated animals.

2.3. Measurement of GS activity GS activity was determined in tissue homogenates by measuring the ␥glutamyl hydroxamate formation in five Mn-treated and five control rats. The specific synthetase enzymatic activity in the “transferase” reactions was calculated in nM of ␥-glutamyl hydroxamate formed per mg of total protein/min. as described previously [6]. Briefly, the assay reaction mixture (Tris-HCl buffer solution 0.1 M, pH 6.8) contained the following: 50 mM imidazole–HCl, 50 mM NH2 OH, 100 mM L-glutamine, 0.5 mM MnCl2 , 0.2 mM ADP and 25 mM KH2 AsO4 . After 30 min incubation at 37 ◦ C, an equal volume of stop solution, consisting of 0.37 M FeCl3 , 0.3 M trichloroacetic acid and 0.6 M HCl, was added. The absorbance of the solution obtained was measured spectrophotometrically at 505 nm (Agilent 8453). A solution of ␥-glutamyl hydroxamate was used as a standard. A protein assay was performed on an aliquot of the homogenate, using the procedure described by Lowry et al. [33]. All chemicals were purchased from Sigma (Milan, Italy). Student’s t-test was used to evaluate significant differences in the GS activity between control and Mn-treated animals.

2.4. Histology, GS, Mn-SOD, Cu/Zn-SOD and GFAP immunoreactivities Coronal brain sections including the sensorimotor cerebral cortex, the caudate-putamen and globus pallidus (1.7 mm anterior to 1.4 mm posterior to

408

M. Morello et al. / Brain Research Bulletin 74 (2007) 406–415

Bregma, according to the brain atlas of Paxinos and Watson [44]) were either stained with cresyl violet, using standard procedures, to evaluate the general tissue morphology (in three Mn-treated and three control rats), or processed using immunocytochemistry techniques for GS, GFAP, Mn-SOD and Cu/ZnSOD (in seven Mn-treated and seven control rats). Free-floating brain sections were incubated overnight in goat anti-mouse GS monoclonal antibody (Sigma; 1:400), goat anti-rabbit Mn-SOD polyclonal antibody (The Binding Site, Birmingham, England; 1:500), goat anti-mouse Cu/Zn-SOD monoclonal antibody (The Binding Site; 1:500), or goat anti-mouse GFAP monoclonal antibody (Sigma; 1:400). The primary antibodies were detected using a biotinylated secondary antibody rabbit anti-goat IgG (Vector Laboratories, Vectastain ABC KIT, Burlingame, CA, USA; 1:100) for 2 h, followed by incubation with avidin-biotinperoxidase complex (DBA, Milan, Italy; 1:100) for 3 h at room temperature. The sections were then rinsed and reacted with 3-3-diaminobenzidine tetrahydrochloride (Sigma; 0.05%) and H2 O2 (0.003%) in PB solution. The sections were mounted on gelatin-chromalum-coated slides for light microscope examination. The specificity of the immunocytochemical reaction was confirmed by the absence of staining after omission of the different primary antibodies. To evaluate the morphological characteristics and the relative numbers of cells immunoreactive for GS, Mn-SOD, Cu/Zn-SOD and GFAP in the sensorimotor cerebral cortex, striatum and globus pallidus, we used a light microscope (Zeiss) equipped with a 20× objective. An expert operator, blind to the pharmacological treatment received by each animal, performed automatic morphological analysis using an image analyser system (Image Pro-Plus, Media Cybernetics 3.0). Counts of GS, Mn-SOD, Cu/Zn-SOD and GFAP immunoreactive cells were performed in three immunostained sections per immunoreactive marker per animal (n = 7 control and n = 7 Mn-treated rats) involving the crosssectional areas of the cerebral cortex, striatum and globus pallidus (1.7 mm anterior to 1.4 mm posterior to Bregma, according to the brain atlas of Paxinos and Watson [44]). The counting frame (4.5 × 104 ␮m2 ) was randomly placed on the first counting area that contained the dorsolateral part of the caudate putamen, the sensorimotor cortex, and the globus pallidus from both sides. Neuronal and glial immunoreactive cell numbers were assessed against the background staining by counting three adjacent grid areas in each region of interest. These values were expressed as the number of immunoreactive cells (mean ± S.D.) per unit test area for each region independent of the hemispheric laterality. The intensity of the GS, Mn-SOD, Cu/Zn-SOD and GFAP immunoreactivity was measured on the same sections used for cell counting (three immunostained sections per immunoreactive marker per animal, n = 7 control and n = 7 Mn-treated rats), following the same strategy for the sampling of unit test areas (4.5 × 104 ␮m2 ). The intensity of the immunoreactive signal was calculated from the densitometry of the immunoreaction product-dependent relative optical density (IOD) using a microcomputer-based image display system (Image Pro-Plus, Media Cybernetics 3.0). The results were expressed as an arbitrary unit of IOD (mean ± S.D.) per unit test area (4.5 × 104 ␮m2 ) ranging between 0 and 10, independent of the hemispheric laterality. The brightest level of IOD matched the background signals of the neuropil and was arbitrarily considered to be 0. The darkest level was considered equal to 10 and matched the maximum attainable content of immunoreaction product in the cell bodies. The Mann–Whitney U-test was used for statistical analysis of the GS, MnSOD, Cu/Zn-SOD and GFAP immunopositive cell numbers. Student’s t-test was used to evaluate significant differences in the intensity of the GS, MnSOD, Cu/Zn-SOD and GFAP immunoreactive products between control and Mn-treated animals. The source of the variance was determined with the Levene’s test. All analyses were performed using the SPSS version 13.0 for Windows (Chicago, Illinois, USA). Data are presented as mean ± S.D. Differences were considered to be statistically different when P < 0.05.

3. Results 3.1. Histological effects of Mn overload on the rat brain The content of Mn in the brain of control animals was approximately 1 ␮g/g of dry weight, ranging from 1 ␮g in the sensorimotor cortex to 1.5 ␮g in the globus pallidus. After chronic Mn treatment, the Mn content increased approximately

Fig. 1. Mn levels in different brain regions of control (n = 5) and Mn-treated (n = 5) rats. Values are expressed as mean ± S.D. of ␮g Mn/g tissue dry weight (*P < 0.01 vs. respective control).

three-fold in the cerebral cortex and striatum, and four-fold in the globus pallidus, as previously reported [8,16,57] (Fig. 1). Qualitative assessment of Nissl-stained sections from both control and Mn-treated animals revealed normal tissue structure (data not shown). In addition, a similar number of GFAP-positive cells was observed in control and Mn-treated animals (Fig. 2). Thus, after chronic Mn treatment, no overt signs of neuronal damage or reactive gliosis were detected in any of the brain regions of interest. 3.2. The effects of Mn overload on the distribution of astrocytes and GS in the cerebral cortex and basal ganglia Interestingly, we show that in control animals, the number of GS-positive cells detected in serial sections of the cerebral cortex, striatum and globus pallidus was equal to the number of GFAP-positive cells (Fig. 2), confirming that GS is an enzyme that is almost exclusive to astrocytes. In control animals, the number of GFAP- and GS-positive cells was constitutively higher in the globus pallidus (Fig. 2c) than in the other regions of interest (Fig. 2a and b), as reported previously [45,64]. However, the intensity of the GS-immunoreactivity (expressed as mean optical density per unit test area) was lower in the globus pallidus than in the other regions (Figs. 3 and 4). These findings confirmed that the GS regional content is proportional to the density of glutamatergic terminals, which is relatively low in the globus pallidus [22]. After chronic Mn treatment, both the number of GSimmunoreactive cells (Fig. 2) and the intensity of the GS-immunoreactivity (Figs. 3 and 4) were significantly reduced in the basal ganglia, especially in the globus pallidus (P < 0.01), but not in the sensorimotor cortex. These findings were confirmed by measuring the actual GS enzymatic activity in these brain regions (Fig. 5). After chronic Mn treatment, the GS activity was significantly decreased in the striatum (35%, P < 0.05) and the globus pallidus (47%, P < 0.01), whereas a non-significant reduction (10%) was observed in the sensorimotor cortex (Fig. 5). These data suggest that the effect of Mn-treatment on GS was more striking in the globus pallidus, where the number of astrocytes is constitutively higher and the GS content is constitutively lower.

M. Morello et al. / Brain Research Bulletin 74 (2007) 406–415

409

most likely glia, also exhibited light Mn-SOD immunoreactivity. In all layers of the sensorimotor cortex, a large number of neurons were Mn-SOD-positive (Fig. 6a). In the striatum, numerous medium-sized and large neurons showed strong MnSOD immunoreactivity (Fig. 6c) and in the globus pallidus (Fig. 6e), many fusiform or polygonal neurons showed intense staining for Mn-SOD. Thus, the number of Mn-SOD positive cells was proportional to the specific density of neurons in the different regions, so that the gradient of Mn-SOD positive cells was cerebral cortex > striatum > globus pallidus (Fig. 2). However, among these different structures the striatum was characterized by the most intense Mn-SOD immunostaining (Figs. 4 and 6), confirming previous results [32]. After Mn treatment, both the number (Fig. 2) and staining intensity (Figs. 4 and 6d and f) of Mn-SOD-positive cells were significantly decreased in the striatum as well in the globus pallidus compared with control animals. However, chronic Mn treatment did not affect the Mn-SOD immunostaining in the sensorimotor cortex (Figs. 2, 4 and 6b). 3.4. The effects of chronic Mn treatment on Cu/Zn-SOD immunoreactivity In the sensorimotor cortex (Fig. 7a) and the basal ganglia (Fig. 7b and c) of control animals, positive Cu/Zn-SOD immunoreactivity was observed in cells identifiable as neurons or glial cells. The Cu/Zn-SOD immunoreactivity was readily observed against the neuropil staining (Fig. 7). Moreover, the number (Fig. 2) and staining intensity (Figs. 4 and 7a, c and e) of Cu/Zn-SOD positive cells was higher in the sensorimotor cerebral cortex than in the striatum but was very low in the globus pallidus (Figs. 2 and 7) (P < 0.01). After chronic Mn treatment, the Cu/Zn-SOD immunoreactivity was unchanged in the sensorimotor cortex or the basal ganglia, both in terms of the number of positive cells and the intensity of immunostaining (Figs. 2, 4 and 7). 4. Discussion

Fig. 2. Number of cells immunoreactive for glial fibrillary acid protein (GFAP), glutamine synthetase (GS), Mn-superoxide dismutase (Mn-SOD), and Cu/Znsuperoxide dismutase (Cu/Zn-SOD) per unit test area in serial sections of the cerebral cortex (A), striatum (B) and globus pallidus (C) in control (n = 7) and Mn-treated (n = 7) rats. Values are expressed as means ± S.D. of number of cells per unit test area (4.5 × 104 ␮m2 ). *P < 0.05; **P < 0.01 vs. respective control.

3.3. The effects of chronic Mn treatment on Mn-SOD immunoreactivity Mn-SOD-positive cells in control rats were predominantly large in size, exhibited neuronal characteristics and had an intense, granular reaction product in the cytoplasm (Fig. 6a, c and e). Apart from the neuronal cells, other smaller cells

Mn neurotoxicity can result in a continuum of dysfunctions from behavioural abnormalities and biochemical effects to extrapyramidal motor symptoms and anatomical lesions in the basal ganglia, depending on species, age and sex, in addition to the severity, duration and route of intoxication [17,36,37,39]. While studies in rodents suggest a variable distribution of Mn throughout the brain and the occurrence of subtle behavioural changes, the effects of Mn on nonhuman primates more closely resemble the human in the expression of manganism [8,39,41]. Here, we examined the effects of a high level of sub-chronic Mn exposure in rats using an oral route of administration [8,30,48]. We confirm that this experimental paradigm can induce biochemical and cellular changes in the basal ganglia in rodents, without significant neuronal loss and gliosis during the time course of the experiment [8,48]. We show that the Mn sub-chronic intoxication per os in rats selectively decreases the immunoreactivity of the manganoproteins, GS and Mn-SOD, in

410

M. Morello et al. / Brain Research Bulletin 74 (2007) 406–415

Fig. 3. Low-power microphotographs illustrating glutamine synthetase (GS) immunoreactive cells in sensorimotor cortex (A and B), striatum (C and D) and globus pallidus (E and F) in a control (A, C and E) and Mn-treated (B, D and F) rat. Bar = 160 ␮m.

the basal ganglia, suggesting that this could be an early event in the complex process of Mn-dependent brain damage. 4.1. Effects of Mn accumulation on GS and GFAP Mn is normally present in the mammalian brain in very small quantities, averaging 1–2 ␮g/g dry weight [46]. Within the CNS, Mn accumulates mainly within astrocytes reaching a total concentration of 50–70 ␮M, of which 30–40% localizes to the cytoplasm [3,4,61]. Therefore, the high level of Mn accumulation in the globus pallidus may be due to a higher density of astrocytes in this brain region. Since the concentration of free cytoplasmic Mn(II) is close to the Kd for Mn(II) with GS, the maximum activating effect of Mn(II) on GS activity occurs only within a very narrow concentration range. Indeed, a slight excess in cytoplasmic Mn(II) has been shown to strongly inhibit the activity of GS [54,62]. We have demonstrated that after chronic oral overload the tissue content of Mn reached its maximum attainable levels, increasing by three to

four times in different regions of the brain [16]. It is therefore conceivable that under our experimental conditions, the Mn accumulation in astrocytes has reached a concentration range that is sufficiently high to account for the observed downregulation of GS activity. Alternatively, it is possible that the reduction in GS levels and activity we observe here, is a reaction to oxidative stress to which GS is exquisitely sensitive [49]. A previous study failed to demonstrate any changes in GS activity and GS protein levels in the cerebral and cerebellar cortex of neonatal rats exposed to oral Mn overload for three weeks [59]. In contrast, an in vitro study demonstrated a dosedependent increase in GS expression in cortical astrocytes after 24 h exposure to Mn [10]. Erikson et al. [18] reported a decrease in GS mRNA and protein levels in the striatum of neonatal rats inhalated with Mn. However, limited changes in GS protein or mRNA were found in numerous studies in rats following 90-day MnSO4 inhalation exposure despite instances of significant Mn accumulation in the striatum [53].

M. Morello et al. / Brain Research Bulletin 74 (2007) 406–415

411

Fig. 5. Glutamine synthetase (GS) activity in the sensorimotor cortex, striatum and globus pallidus of control (n = 5) and Mn-treated (n = 5) rats. GS activity (mean ± S.D.) was detected as ␥-glutamyl hydroxamate (nM) formation by 1 mg of total protein in 1 min. *P < 0.05; **P < 0.01 vs. respective control.

[50,51]. The basis for such regional sensitivity of GS in the brain to different factors is presently unknown. It is possible that the larger decrease in GS activity and immunoreactivity we observe in the globus pallidus after chronic oral Mn overload may be dependent both on higher accumulation of Mn and lower constitutive levels of GS compared to other brain regions. Since the GFAP immunoreactivity was unaffected by Mn overload in any brain region under our experimental conditions, these findings suggest that GS and GFAP expression may be independently regulated in astrocytes [50]. Although differences in experimental paradigms and duration of exposure to Mn should be emphasized, our results demonstrate that changes in the GS expression and activity after long-term Mn treatment in vivo affect the striatum and globus pallidus, but not the cerebral cortex, consistent with previous in vivo [18,59] and in vitro [10] studies. 4.2. Effects of Mn accumulation on Mn-SOD and Cu/Zn-SOD

Fig. 4. Immunoreaction product intensity for glial fibrillary acid protein (GFAP), glutamine synthetase (GS), Mn-superoxide dismutase (Mn-SOD), and Cu/Znsuperoxide dismutase (Cu/Zn-SOD) per unit test area in serial sections of the cerebral cortex (A), striatum (B) and globus pallidus (C) in control (n = 7) and Mn-treated (n = 7) rats. Values are expressed as means ± S.D. of IOD per unit test area (4.5 × 104 ␮m2 ). *P < 0.05, **P < 0.01 vs. respective control.

Here, we show that a reduction in GS activity, detected biochemically, is concordant with a reduced GS content, detected immunohistochemically, and that the globus pallidus is the most affected region. Region-selective changes in GS in the rat CNS have been observed not only after Mn overload but also after pharmacological treatment with gamma-vinyl GABA, and in pathological conditions following portocaval anastomosis

Only about 20% of the total SOD activity in the brain is due to neuronal Mn-SOD. About 80% is due to Cu/Zn-SOD activity, found in both neurons and astrocytes [23]. Accordingly, we found that in general the intensity of Cu/Zn-SOD immunoreactivity in the brain was higher than that of Mn-SOD. Moreover, the regional distribution of Mn-SOD and Cu/Zn-SOD was different. The striatum expressed the highest staining intensity for Mn-SOD, whereas the sensorimotor cortex expressed the highest staining intensity for Cu/Zn-SOD as previously described [32]. Here, we detected a low number of Cu/Zn-SOD-positive cells in the globus pallidus. Therefore, in contrast to other brain regions, the lack of detectable levels of Cu/Zn-SOD and MnSOD in many cells of the globus pallidus, under basal conditions and after Mn overload, may have consequences in particular pathological conditions leading to oxidative damage such as in manganism. In addition to the decreased GS expression and activity, we show that after chronic Mn treatment the striatum and the globus pallidus undergo a reduction in Mn-SOD protein levels, without accompanying changes in Cu/Zn-SOD and GFAP levels

412

M. Morello et al. / Brain Research Bulletin 74 (2007) 406–415

Fig. 6. Low-power microphotographs illustrating Mn superoxide dismutase (Mn-SOD) immunoreactive cells in sensorimotor cortex (A and B), striatum (C and D) and globus pallidus (E and F) in a control (A, C and E) and Mn-treated (B, D and F) rat. Bar = 160 ␮m.

in any of the brain regions examined here. Our results are in accordance with an in vitro study showing that Mn overexposure dose-dependently decreased total SOD activity in astrocytes [10]. Thus, from our observations, we can infer that Mn overload selectively decreases the expression of some important manganoproteins in the brain, including GS and Mn-SOD. Furthermore, these decreases are region-specific, with the striatum and globus pallidus being most affected. Further studies are required to clarify the mechanisms by which chronic Mn overload affects the expression of manganoproteins: whether GS and Mn-SOD undergo oxidation and rapid degradation leading to a decrease in both their activity and protein levels [49,59]; and/or whether GS and MnSOD expression can be affected by a feedback mechanism through a metal regulatory element at a transcriptional level [55].

4.3. Convergence of regional manganoproteins reduction and a low reservoir of Cu/Zn-SOD The decrease in GS and Mn-SOD levels is likely to correspond to a decrease in their functional activities. The reduced activity of GS in astrocytes may slow down the catabolism of glutamate to glutamine, participating in the impairment of the extracellular glutamate scavenger system demonstrated in manganism [7,9,15,19,28,29,47,56,61]. Moreover, it is possible that the reduced expression of Mn-SOD we observe after chronic Mn treatment both in the striatum and the globus pallidus neurons could affect their antioxidant capacity against superoxide (O2 − ). Interestingly, a reduction in mitochondrial Mn-SOD has been reported to exacerbate glutamate toxicity in cultured mouse cortical neurons [63]. To explain the increased vulnerability of the globus pallidus in chronic manganism, we should consider not only the effects

M. Morello et al. / Brain Research Bulletin 74 (2007) 406–415

413

Fig. 7. Low-power microphotographs illustrating Cu/Zn superoxide dismutase (Cu/Zn-SOD) immunoreactive cells in sensorimotor cortex (A and B), striatum (C and D) and globus pallidus (E and F) in a control (A, C and E) and Mn-treated (B, D and F) rat. Bar = 160 ␮m.

of Mn on the GS and Mn-SOD, but also the regional distribution of Cu/Zn-SOD. Since, the globus pallidus has a constitutive low reservoir of Cu/Zn-SOD, it may be particularly vulnerable to Mn overload and oxidative damage when Mn-SOD levels are decreased. A previous study supports this view, showing that after chronic Mn-treatment the globus pallidus neurons, but not the striatal neurons, manifest peculiar responses to glutamate: repeated applications of glutamate produced irreversible cell damage at concentrations that commonly promote desensitizing responses [48]. Therefore, under our experimental conditions, the neurotoxic potential of Mn may depend more on the regional impairment of homeostasis of the tissues, than on the intrinsic pro-oxidant capacity of Mn itself.

Mn-SOD in neurons. This decreased expression probably results in the partial loss of function of these enzymes in the striatum and globus pallidus. Secondary effects, such as oxidative and neurotoxic stress may follow in the basal ganglia, with impaired glutamate metabolism and reduced radical scavenger activity. The globus pallidus may be more vulnerable than other brain regions to chronic Mn treatment, due to its constitutive low levels of Cu/Zn-SOD. Conflict of interest None. Acknowledgements

5. Conclusions In conclusion, we show that chronic Mn treatment affects the expression of manganoproteins, including GS in astrocytes and

The authors thank Roberto Sorge (Laboratory of Biometry, University of Rome Tor Vergata) for performing statistical analysis. This investigation was supported by grants from the Italian

414

M. Morello et al. / Brain Research Bulletin 74 (2007) 406–415

Ministry of Health (N38/2march2006), and the Italian Ministry of Research and University (FIRB#RBNE03PX83). References [1] F.S. Archibald, C. Tyree, Manganese poisoning and the attack of trivalent manganese upon catecholamines, Arch. Biochem. Biophys. 256 (1987) 638–650. [2] M. Aschner, J.L. Aschner, Manganese neurotoxicity: cellular effects and blood-brain barrier transport, Neurosci. Biobehav. Rev. 15 (1991) 333–340. [3] M. Aschner, M. Gannon, H.K. Kimelberg, Manganese uptake and efflux in cultured rat astrocytes, J. Neurochem. 58 (1992) 730–735. [4] M. Aschner, K.E. Vrana, W. Zheng, Manganese uptake and distribution in the central nervous system (CNS), Neurotoxicology 20 (1999) 173– 180. [5] A. Barbeau, Manganese and extrapyramidal disorders, Neurotoxicology 5 (1984) 13–35. [6] I.S. Boksha, E.B. Tereshkina, G.S. Bubaeya, Glutamine synthetase and glutamine synthetase-like protein from human brain: purification and comparative characterization, J. Neurochem. 75 (2000) 2574–2582. [7] E.P. Brouillet, L. Shinobu, U. McGarvey, F. Hochberg, M.F. Beal, Manganese injection into the rat striatum produces excitotoxic lesions by impairing energy metabolism, Exp. Neurol. 120 (1993) 89–94. [8] P. Calabresi, M. Ammassari-Teule, P. Gubellini, G. Sancesario, M. Morello, D. Centonze, G.A. Marfia, E. Saulle, E. Passino, B. Picconi, G. Bernardi, A synaptic mechanism underlying the behavioural abnormalities induced by manganese intoxication, Neurobiol. Dis. 8 (2001) 419–432. [9] D. Centonze, P. Gubellini, G. Bernardi, P. Calabresi, Impaired excitatory transmission in the striatum of rats chronically intoxicated with manganese, Exp. Neurol. 172 (2001) 469–476. [10] C.J. Chen, S.L. Liao, Oxidative stress involves in astrocytic alterations induced by manganese, Exp. Neurol. 175 (2002) 216–225. [11] D.W. Christianson, Structural chemistry and biology of manganese metalloenzymes, Prog. Biophys. Mol. Biol. 67 (1997) 217–252. [12] J. Cooper, On the effects of black oxide of manganese when inhaled into the lungs, Br. Ann. Med. Pharmacol. 1 (1837) 41–42. [13] G.C. Cotzias, Manganese in health and disease, Physiol. Rev. 38 (1958) 503–532. [14] A. Ejima, T. Imamura, S. Nakamura, H. Saito, K. Matsumoto, S. Momono, Manganese intoxication during total parenteral nutrition, Lancet 339 (1992) 426. [15] K. Erikson, M. Ashner, Manganese causes differential regulation of glutamate transporter (GLAST) taurine transporter and metallothionein in cultured rat astrocytes, Neurotoxicology 23 (2002) 595–602. [16] K.M. Erikson, A.W. Dobson, D.C. Dorman, M. Aschner, Manganese exposure and induced oxidative stress in the rat brain, Sci. Total Environ. Rev. 334/335 (2004) 409–416. [17] K.M. Erikson, A.W. Dobson, D.C. Dorman, M. Aschner, Airborne manganese exposure differentially affects end points of oxidative stress in an age- and sex-dependent manner, Biol. Trace Elem. Res. 100 (2004) 49–62. [18] K.M. Erikson, D.C. Dorman, L.H. Lash, M. Aschner, Persistent alterations of biomarkers of oxidative stress resulting from combined in utero and neonatal manganese inhalation, Biol. Trace Elem. Res. 104 (2005) 151–163. [19] K.M. Erikson, R.L. Suber, M. Aschner, Glutamate/aspartate transporter (GLAST), taurine transporter and metallothionein mRNA levels are differentially altered in astrocytes exposed to manganese chloride, manganese phosphate or manganese sulphate, Neurotoxicology 23 (2002) 281–288. [20] M.L. Failla, Considerations for “optimal nutrition” for copper, zinc, manganese and molybdenum, Proc. Nutr. Soc. 58 (1999) 497–505. [21] J.M. Fell, A.P. Reynolds, N. Meadows, K. Khan, S.G. Long, G. Quaghebeur, W.J. Taylor, P.J. Milla, Manganese toxicity in children receiving long-term parenteral nutrition, Lancet 347 (1996) 1218–1221. [22] F. Fonnum, R.E. Paulsen, Comparison of transmitter amino acid levels in rat globus pallidus and neostriatum during hypoglycemia or after treatment with methionine sulfoximine or ␥-vinyl -aminobutyric acid, J. Neurochem. 54 (1990) 1253–1257.

[23] P. Gaetani, R. Rodriguez-y-Baena, S. Quaglini, R. Bellazzi, C. Caf´e, C. Torri, F. Marzatico, Experimental subarachnoid haemorrhage events related to antioxidant enzymatic systems and eicosanoid peroxide enhancement, Neurochem. Res. 19 (1994) 839–844. [24] C.E. Gavin, K.K. Gunter, T.E. Gunter, Manganese and calcium efflux kinetics in brain mitochondria: relevance to manganese toxicity, Biochem. J. 266 (1990) 329–334. [25] D.G. Graham, Catecholamine toxicity: a proposal for the molecular pathogenesis of manganese neurotoxicity and Parkinson’s disease, Neurotoxicology 5 (1984) 83–95. [26] D. Hamai, S.C. Bondy, Oxidative basis of manganese neurotoxicity, Ann. N.Y. Acad. Sci. 1012 (2004) 129–141. [27] R.A. Hauser, T.A. Zesiewicz, A.S. Rosemurgy, C. Martinez, C.W. Olanow, Manganese intoxication and chronic liver failure, Ann. Neurol. 36 (1994) 871–875. [28] A.S. Hazell, Astrocytes and manganese neurotoxicity, Neurochem. Int. 41 (2002) 271–277. [29] A.S. Hazell, M.D. Norenberg, Manganese decreases glutamate uptake in cultured astrocytes, Neurochem. Res. 22 (1997) 1443–1447. [30] J.C. Lai, M.J. Minski, A.W. Chan, T.K. Leung, L. Lim, Manganese mineral interactions in brain, Neurotoxicology 20 (1999) 433–444. [31] J.J. Liccione, M.D. Maines, Manganese-mediated increase in the rat brain mitochondrial cytochrome P-450 and drug metabolism activity: susceptibility of the striatum, J. Pharmacol. Exp. Ther. 248 (1989) 222–228. [32] J. Lindenau, H. Noack, H. Possel, K. Asayama, G. Wolf, Cellular distribution of superoxide dismutases in the rat CNS, Glia 29 (2000) 25–34. [33] O.H. Lowry, N.J. Rosebrough, A.L. Farr, R.J. Randall, Protein measurement with the Folin phenol reagent, J. Biol. Chem. 193 (1951) 265–275. [34] E.A. Malecki, Manganese toxicity is associated with mitochondrial dysfunction and DNA fragmentation in rat primary striatal neurons, Brain Res. Bull. 55 (2001) 225–228. [35] I. Mena, O. Marin, S. Fuenzalida, G.C. Cotzias, Chronic manganese poisoning: clinical picture and manganese turnover, Neurology 17 (1967) 128–136. [36] D. Mergler, M. Baldwin, S. Belanger, F. Larribe, A. Beuter, R. Bowler, M. Panisset, R. Edwards, A. De Geoffroy, M.P. Sassine, K. Hudnell, Manganese neurotoxicity, a continuum of dysfunction: results from a community based study, Neurotoxicology 20 (1999) 327–342. [37] M. Miele, P.A. Serra, G. Esposito, M.R. Delogu, R. Migheli, G. Rocchitta, M.S. Desole, Glutamate and catabolites of high-energy phosphates in the striatum and brain stem of young and aged rats subchronically exposed to manganese, Aging Clin. Exp. Res. 12 (2000) 393–397. [38] M. Morello, A. Canini, M. Grilli-Caiola, A. Martorana, P. Mattioli, G. Bernardi, G. Sancesario, Manganese detected by electron spectroscopy imaging and electron energy loss spectroscopy in mitochondria of normal brain cells, J. Trace Microsc. Techn. 20 (2002) 481–491. [39] M.C. Newland, Animal models of manganese’s neurotoxicity, Neurotoxicology 20 (1999) 415–432. [40] M.D. Noremberg, Distribution of glutamine synthetase in the rat central nervous system, J. Histochem. Cytochem. 27 (1979) 756–762. [41] L. Normandin, M. Panisset, J. Zayed, Manganese neurotoxicity: behavioural, pathological, and biochemical effects following various routes of exposure, Rev. Environ. Health 17 (2002) 189–217. [42] C.W. Olanow, P.F. Good, H. Shinoth, K.A. Hewitt, F. Vingerhoets, B.J. Snow, M.F. Beal, D.B. Calne, D.P. Perl, Manganese intoxication in the rhesus monkey: a clinical imaging, pathologic and biochemical study, Neurology 46 (1996) 492–498. [43] P.K. Pal, A. Samii, D.B. Calne, Manganese neurotoxicity: a review of clinical features, imaging and pathology, Neurotoxicology 20 (1999) 227–238. [44] G. Paxinos, C.R. Watson, The Rat Brain in Stereotaxic Coordinates, Raven Press, London, 1992. [45] G. Percheron, J. Yelnik, C. Francois, A Golgi analysis of the primate globus pallidus. III. Spatial organization of the striato-pallidal complex, J. Comp. Neurol. 227 (1984) 214–227. [46] J.R. Prohaska, Functions of trace elements in brain metabolism, Physiol. Rev. 67 (1987) 858–901. [47] A.A. Shukakidze, I.L. Lazriev, R.G. Khetsuriani, T.Z. Bikashvili, Changes in neuroglial ultrastructure in various parts of the rat brain during man-

M. Morello et al. / Brain Research Bulletin 74 (2007) 406–415

[48]

[49] [50]

[51] [52] [53]

[54]

[55]

[56]

ganese chloride poisoning, Neurosci. Behav. Physiol. 32 (2002) 561– 566. F. Spadoni, A. Stefani, M. Morello, F. Lavaroni, P. Giacomini, G. Sancesario, Selective vulnerability of pallidal neurons in the early phases of manganese intoxication, Exp. Brain Res. 135 (2000) 544–551. E.R. Stadtmann, Protein oxidation and aging, Science 257 (1992) 1220–1224. I. Suarez, G. Bodega, E. Arilla, B. Fernandez, Region-selective glutamine synthetase expression in the rat central nervous system following portocaval anastomosis, Neuropathol. Appl. Neurobiol. 23 (1997) 254–261. I. Suarez, G. Bodega, B. Fernandez, Glutamine synthetase in brain: effect of ammonia, Neurochem. Int. 41 (2002) 123–142. A. Takeda, Manganese action in brain function, Brain Res. Brain Res. Rev. 41 (2003) 79–87. M.D. Taylor, K.M. Erikson, A.W. Dobson, V.A. Fitsanakis, D.C. Dorman, M. Aschner, Effects of inhaled manganese on biomarkers of oxidative stress in the rat brain, Neurotoxicology 27 (2006) 788–797. G. Tholey, S. Bloch, M. Ledig, P. Mandel, F. Wedler, Chick brain glutamine synthetase and Mn2+ –Mg2+ interactions, Neurochem. Res. 12 (1987) 1041–1047. J. Thongphasuk, L.W. Oberley, T.D. Oberley, Induction of superoxide dismutase and cytotoxicity by manganese in human breast cancer cells, Arch. Biochem. Biophys. 365 (1999) 317–327. H. Toki, K. Namikawa, Q. Su, S. Kiyu-Seo, K. Sato, H. Kiyama, Enhancement of extracellular glutamate scavenge system in injured motor neurons, J. Neurochem. 71 (1998) 913–919.

415

[57] E.J. Underwood, Trace metals in human and animal health, J. Hum. Nutr. 35 (1981) 37–48. [58] M.A. Verity, Manganese neurotoxicity: a mechanistic hypothesis, Neurotoxicology 20 (1999) 489–497. [59] S. Weber, D.C. Dorman, L.H. Lash, K. Erikson, K.E. Vrana, M. Aschner, Effects of manganese (Mn) on the developing rat brain: oxidative stress related endpoints, Neurotoxicology 23 (2002) 169–175. [60] F.C. Wedler, R.B. Denman, Glutamine synthetase: the major Mn(II) enzyme in mammalian brain, Curr. Top. Cell. Regul. 24 (1984) 153– 169. [61] F.C. Wedler, H.W. Ley, Kinetic, ESR, and trapping evidence for in vivo binding of Mn(II) to glutamine synthetase in brain cells, Neurochem. Res. 19 (1994) 139–144. [62] F.C. Wedler, M.C. Vichnin, B.W. Ley, G. Tholet, M. Ledig, J.C. Copin, Effects of Ca(II) ions on Mn(II) dynamics in chick glia and rat astrocytes: potential regulation of glutamine synthetase, Neurochem. Res. 19 (1994) 145–151. [63] Y. Li, J. Copin, L.F. Reola, B. Calagui, G.T. Gobbel, S.F. Chen, S. Sato, C.J. Epstein, P.H. Chan, Reduced mitochondrial manganese-superoxide dismutase activity exacerbates glutamate toxicity in cultured mouse cortical neurons, Brain Res. 814 (1998) 164–170. [64] K. Zilles, F. Hajos, M. Kalman, A. Schleicher, Mapping of glial fibrillary acidic protein-immunoreactivity in the rat forebrain and mesencephalon by computerized image analysis, J. Comp. Neurol. 308 (1991) 340– 355.