- Email: [email protected]
The PARP inhibitor benzamide protects against kainate and NMDA but not AMPA lesioning of the mouse striatum in vivo
Brain Research 996 (2004) 1 – 8 www.elsevier.com/locate/brainres
Research report
The PARP inhibitor benzamide protects against kainate and NMDA but not AMPA lesioning of the mouse striatum in vivo Cristina Cosi *, Karen Guerin, Marc Marien, Wouter Koek, Karin Rollet Divisions de Neurobiologie I et II, Centre de Recherche Pierre Fabre, 17 avenue Jean Moulin, 81106 Castres, France Accepted 8 September 2003
Abstract Overactivation of poly(ADP-ribose) polymerase (PARP) in response to genotoxic insults can cause cell death by energy deprivation. We previously reported that neurotoxic amounts of kainic acid (KA) injected into the rat striatum produce time-dependent changes in striatal PARP activity in vivo. Here, we have investigated the time-course of KA-induced toxicity and the effects of the PARP inhibitor benzamide on KA, AMPA and NMDA neurotoxicities in vivo, by measuring changes in the volume of the lesion and in NAD+ and ATP levels induced by the intra-striatal injection of these excitotoxins in C57Bl/6N mice. The KA-induced lesion volume was dependent on the amount of toxin injected and the survival time. The lesion was well developed at 48 h and was almost undetectable after one week. KA produced an extensive astrogliosis at one week. Benzamide partially prevented both KA- and NMDA- but not AMPA-induced lesions when measured at 48 h after the treatment. The effects of benzamide appeared to be in part related to changes in energy metabolism, since KA produced decreases in striatal levels of NAD+ and ATP that were partially prevented by benzamide at 48 h and which returned to control levels at one week. NMDA did not affect NAD+ and induced little alteration in ATP levels. Benzamide had no effect on AMPA-induced decreases in either NAD+ or ATP levels at 48 h. These results (1) indicate that PARP overactivation and energy depletion could be responsible in part for the cellular demise during the development of the lesion induced by KA; (2) confirm that PARP is involved in NMDA but not AMPA toxicities; (3) suggest the existence of differences between KA and AMPA-mediated toxicities; and (4) provide further evidence supporting PARP as a novel target for new drug treatments against neurodegenerative disorders. D 2003 Elsevier B.V. All rights reserved. Theme: Neurotransmitters, modulators, transporters, and receptors Topic: Excitatory amino acids: excitotoxicity Keywords: Poly(ADP-ribose) polymerase; Excitotoxicity; Kainate; NMDA; AMPA; Benzamide; C57Bl/6N mouse
1. Introduction Poly(ADP-ribose) polymerase (PARP) is a DNA binding protein that uses NAD+ as substrate. PARP is activated by strand breaks in the DNA molecule that can be induced by DNA damaging agents, including free radicals [12]. When fully activated, for example by free radical-induced DNA damage, PARP can deplete cellular energy stores, under the form of NAD+ and ATP, predisposing the cell to death [3]. In vitro and in vivo studies using PARP inhibitors, including benzamide, and/or mice or cells with a disrupted PARP-1 gene, have indicated a participation of PARP in glutamate-[7] and N-methyl-D-aspartate (NMDA)-induced * Corresponding author. Tel.: +33-5-63-71-42-86; fax: +33-5-63-7143-63. E-mail address: [email protected] (C. Cosi). 0006-8993/$ - see front matter D 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2003.09.072
[21,27] neurotoxicities, cerebral ischemia [13,14,22,25], and 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) [6,9,20], and methamphetamine-induced [8,24] neurotoxicities. In particular, it has been shown that the PARP inhibitor benzamide is neuroprotective in C57Bl/6N mice against different types of neurotoxicities [8,9] without affecting body temperature [8]. Kainic acid-induced (KA) neurotoxicity is related to free radical formation via membrane lipid peroxidation and the arachidonic acid cascade and ATP depletion [11,15]. In particular, KA injected into the striatum of mice causes a rapid decline (by about 30%) in striatal levels of ATP [23]. We have previously shown that a nanomolar amount of KA injected into the striatum of the rat can produce timedependent changes in striatal PARP activity [10]. In the present study, we investigated whether the timerelated changes in striatal PARP activity induced by intra-
2
C. Cosi et al. / Brain Research 996 (2004) 1–8
striatal KA [10] were correlated or not with PARP-mediated cell death by energy depletion, by studying the time course effects of KA and the effects of the PARP inhibitor benzamide on lesion volume and striatal ATP and NAD+ levels. Since PARP-1 knock-out mice are completely resistant to NMDA-induced lesion but not a-amino-3-hydroxy-5-methyl-4-isoxazole-4-propionique acid (AMPA)-induced lesions [21], we also investigated if pharmacological inhibition of PARP with benzamide would protect C57Bl/6N mice against NMDA and AMPA neurotoxicities.
2. Materials and methods 2.1. Animals Animals were handled and cared for in accordance with the Guide for the Care and Use of Laboratory Animals (National Research Council, 1996) and the European Directive No 86/609. The experimental protocol was carried out in compliance with local ethical committee guidelines for animal research. Male C57Bl/6N mice (26 – 28 g, Charles River, Elbeuf, France) were housed in groups of 10 with free access to water and food, under a 12-h light/ dark cycle. 2.2. Excitotoxic lesions Twenty-four hours before surgery, mice were weighed and isolated. Mice were anesthetized with an intra-peritoneal (i.p.) injection of sodium pentobarbital (65 mg/kg). The body temperature was monitored by a rectal probe and maintained at 37 jC by a thermostated heating pad. Excitatory amino acids were dissolved in PBS (pH adjusted to 7.4 with 1 N NaOH). Mice were placed in a stereotaxic apparatus (Kopf model 900) and NMDA (20 nmol), AMPA (5.0 nmol) [21], kainic acid (0.5, 1.0 or 1.5 nmol), or vehicle (PBS 10 mM) were injected in a volume of 0.3 Al into the right striatum according to the following stereotaxic coordinates: 0.5 mm anterior to bregma, 2.0 mm lateral to bregma, and 3.5 mm ventral from the skull surface [16]. At the end of the surgery, a 1 ml subcutaneous (s.c.) injection of 0.9% NaCl was administrated to prevent dehydration, and mice were placed in a thermostated cage (25 jC) until they were completely awake. During the next 48 h, mice received two injections of 0.9% NaCl per day. The PARP inhibitor benzamide (160 mg/kg) was prepared in 0.9% NaCl containing 0.01% of carboxymethylcellulose. Benzamide or vehicle was administered i.p. 30 min before, and 3.5 h after the intrastriatal injection of the neurotoxin or vehicle [8,9]. All drugs were purchased from Sigma, France. 2.3. Estimation of lesion volume Mice were killed by decapitation at different times after the intrastriatal injections. Brains were dissected and frozen
in isopentane at 45 jC, then stored at 80 jC until processed. Brains were sectioned, starting from the anterior aspect of the corpus callosum, into 240 sequential rostrocaudal sections of 20-Am thickness. These sections encompassed both the entire span of the lesioned tissue and the entire striatum. One out of every 10 sections was collected on polylysine-coated slides, fixed in 4% paraformaldehyde (0 jC, 30 min) and stained with cresyl violet (1% cresyl violet acetate in 0.1 M acidic acid, pH 4). The area of the lesion, indicated by the lack of staining when compared to the contralateral (intact) striatum, was measured in each series of sections by means of an image analysis program (Scion Image, NIH) and the lesion volume was given by the sum of the serial lesion volumes calculated using the known distance between sections. Occurrence of variability for the extension of the lesion volume obtained with NMDA and AMPA, when compared with that reported previously [21] might be due differences in mouse strains and lesioning procedures image analysis procedures. 2.4. Immunocytochemistry for glial fibrillary acidic protein Immunocytochemistry for glial fibrillary acidic protein (GFAP) was performed on sections adjacent to those utilized for the estimation of the lesion volume. Sections were collected on polylysine-coated slides and fixed in 4% paraformaldehyde (0 jC, 30 min). After antigen retrieval by trypsin (0.05% in 0.1% CaCl2) and inactivation of endogenous peroxidases, slides were incubated for 1 h with 10% goat serum in PBS in order to block non-specific sites and then incubated at 4 jC overnight with mouse monoclonal antibody anti-GFAP (diluted 1:1000; Sigma). After washing with PBS, slides were incubated for 45 min at room temperature with goat-anti mouse secondary antibody peroxidase-conjugated (diluted 1:200; Chemicon International). After incubation with the peroxidase-antiperoxidase (diluted 1:200; Chemicon International) the antibody was revealed by 3V-diaminobenzidine oxidation. 2.5. Measurement of striatal ATP and NAD + levels Measurements of striatal ATP and NAD+ levels were measured as previously described [6]. Mice were killed by head-focused microwave irradiation (3.4 kW, 0.8 – 1.1 s; Sacronk Model 8000, SAIREM, Vaulx-en-Velin, France) at different times after the intra-striatal injection of the different neurotoxins. This method of sacrifice is essential for rapid and effective fixation of the brain in situ, to permit accurate determination of metabolites which are susceptible to post-mortem changes. Brains were then removed on ice (+4 jC) and striata were dissected and stored at 80 jC until processed. The samples were homogenized by brief sonication in a solution containing perchloric acid (0.2 M), EDTA (1 mM) and Na2S2O5 (1 mM), and were then centrifuged at 17 400g for 20 min at +4 jC. The supernatants were filtered through Millex HV 0.45 Am filters (Millipore, France) and
C. Cosi et al. / Brain Research 996 (2004) 1–8
3
analyzed by HPLC (Alliance 2690) with UV detection (Waters Model 486). Aliquots of 20 Al were automatically injected by a refrigerated (+8 jC) autosampler. Absorbance wavelength for detection was set at 260 nm. Values were expressed in nmol per mg of protein. Proteins were measured according to Lowry.
3. Results 3.1. Dose-dependency of the kainic acid induced-lesion Since previous studies using AMPA or NMDA indicated that a well-developed lesion volume is obtained 48 h after the intra-striatal injection of the toxins [2], the present study evaluated the effects of different doses of kainic acid (KA) on the volume of the lesion at this time point (Fig. 1). Intrastriatal injection of KA produced a lesion whose volume was dose-dependent. KA (0.5 nmol) produced a lesion volume of 0.50F0.10 mm3. Maximal lesion volume was obtained with KA at 1.5 nmol (6.14F1.39 mm3), but this was not significantly different from the lesion obtained with 1.0 nmol (4.39F1.33 mm3). 3.2. Time-dependency of kainic acid induced lesion Kainic acid (1.25 nmol) caused a lesion whose development depended on time (Fig. 2). At 1 h no lesion was detectable. At 6 h a lesion volume of 0.59F0.25 mm3 was visible. At 24 h the lesion was well developed and the volume measured 1.54F0.27 mm3. At 48 h the lesion volume was 1.78F0.36 mm3. At 1 week after the injection of KA, the lesion volume had decreased to 0.08F0.03 mm3.
Fig. 1. Dose-related changes in the volume of the lesion 48 h after the intrastriatal injection of kainic acid (KA) into the right striatum of C57Bl/6N mice. Values, in mm3, are meansFS.E.M. of four animals per group. *p<0.05 compared to KA (0.5 nmol), Kruskal – Wallis ANOVA followed by Mann – Whitney U-test.
Fig. 2. Time-related changes in the volume of the lesion induced by the intra-striatal injection of KA (1.25 nmol) in C57Bl/6N mice. Values, in mm3, are meansFS.E.M. of six to seven animals per group. *p<0.05 compared to 6 h and compared to 1 week, Kruskal – Wallis ANOVA followed by Mann – Whitney U-test.
3.3. GFAP immunostaining In PBS injected animals, at 6 h after injection, a lightly labeled diffuse pattern of GFAP immunopositive cells with stellated shape was observed around the lateral ventricles, within the corpus callosum and the septal area, and in the proximity of the needle tract. At 48 h after PBS injection, strongly labeled astrocytes were found in the proximity of the needle tract and adjacent to the lateral ventricle on the injected side. A more diffuse astrocyte reaction was found also in the septum, corpus callosum, anterior commissure and ventral thalamus. At 1 week after PBS injection, astrocytes were accumulated along the needle tract. In KA injected (1.5 nmol) animals, strongly labeled GFAP-positive cells were observed at 6 h in the same areas as in the PBS injected animals at 6 h. These cells presented few arborizations. At 48 h a diffuse astrogliosis was observed in both hemispheres with widely distributed strongly labeled astrocytes in the area around the lesion. At 48 h, a dense distribution of strongly labeled-GFAP positive astrocytes was observed in the intermediate layer of the parietal and pyriform cortices adjacent to the lateral ventricle ispilateral to the lesion. At 1 week after KA injection (Fig. 3, upper and lower panels), the area corresponding to the lesion was filled with strongly labeled GFAP-positive astrocytes, larger in size than those observed at 48 h and with long interconnecting extensions. A more diffuse astrogliosis was observed throughout both hemispheres. Treatment with benzamide did not affect the patterns of distribution of the astrocytes either in the PBS or in the KA injected animals, at 48 h. 3.4. Benzamide affords partial protection against kainateand NMDA-but not AMPA-induced neurotoxicities The effects of benzamide on KA-, NMDA- and AMPAinduced lesions were investigated by measuring changes in
4
C. Cosi et al. / Brain Research 996 (2004) 1–8
Fig. 3. Kainic acid-induced astrogliosis. Glial fibrillary acidic protein (GFAP) immunostaining at 10 days after KA injection (1.5 nmol) into the right striatum. Upper panel: injected side with numerous strongly stained GFAP immunopositive cells. Scale bar =75 Am. Lower panel: highmagnification micrographs showing hypertrophic astrocytes. Scale bar=3 Am.
lesion volume at 48 h after the intra-striatal injections of the toxins and the i.p. treatment with benzamide (Figs. 4 and 5). Benzamide administered i.p. 30 min before and 3.5 h after the injection of 1.0 nmol of KA reduced the volume of the lesion by 59%, from 2.22F0.36 to 0.92F0.32 mm3 (Fig. 4A). The intra-striatal injection of 20 nmol of NMDA produced a lesion volume of 3.40F0.72 mm3 at 48 h (Fig. 4B). Benzamide treatment afforded partial protection against the toxicity of NMDA, allowing a 62% reduction of the volume of the lesion (1.31F0.38 mm3). The intra-striatal injection of 5.0 nmol of AMPA produced a lesion volume of 1.96F0.34 mm3 at 48 h. Benzamide treatment provided no protection against AMPAinduced excitotoxicity (Fig. 4C). 3.5. Time-course of kainate-induced changes in striatal levels of ATP and NAD+ The time-course of effects of intra-striatal injection KA (1 nmol) on striatal ATP and NAD+ levels were investigated in mice sacrificed by head focused microwave irradiation at
Fig. 4. Benzamide treatment decreases the volume of the lesion induced by intrastriatal injection of (A) KA (1.0 nmol) and (B) NMDA (20 nmol) but not (C) AMPA (5 nmol) in mice. Values, in mm3, are meansFS.E.M. of eight animals per group. *p<0.05, Mann – Whitney U-test.
C. Cosi et al. / Brain Research 996 (2004) 1–8
5
than that obtained with KA ( 15%) at the same time point (Table 1).
3.6. Effects of benzamide on excitotoxin-induced changes in striatal levels of ATP and NAD + KA, NMDA and AMPA differentially affected striatal levels of ATP and NAD+ at 48 h when compared to the striata injected with PBS in controls. KA (1 nmol) caused a 15% decrease in ATP levels and reduced NAD+ by 16% when compared to the PBS injected controls (Table 1A), whereas AMPA produced a significant and pronounced decrease in NAD+ ( 26% compared to the PBS injected controls) (Table 1B). No changes in striatal NAD+ and ATP were detected at 48 h after NMDA intrastriatal administration. Benzamide (160 mg/kg i.p.) administrated 30 min before and 3.5 h after KA injection completely prevented the decrease in ATP, and partially reduced the decrease of
Fig. 5. Cresyl violet stained sections of the brains of mice injected in the right striatum with either PBS or excitotoxin and treated with benzamide (2160 mg/kg i.p.). (Right panel) or vehicle (left panel). Benzamide provided partial neuroprotection against KA and NMDA-induced toxicities, but not against AMPA-induced toxicity. The lack of staining with cresyl violet indicates the area of the lesion.
different times after KA injection (Fig. 6A and B). KA caused a rapid and transient decrease in ATP levels (Fig. 6B) in the KA-injected striatum compared to the PBS injected control. ATP was reduced by 31% at 1 h, at which time a smaller ( 9%) decrease of NAD+ was also observed (Fig. 6B). NMDA (20 nmol) decreased striatal ATP by 7% at 1 h, compared to the side contralateral to the injection, and did not affect NAD+. In the KA treated animals, ATP and NAD+ were both back to control levels at 6 h. A second phase of decreases followed at later times: NAD+ levels were reduced at 24 h ( 21%) and at 48 h ( 15%) (Fig. 6A) and were paralleled by decreases in striatal ATP ( 19% and 10%, at 24 and 48 h, respectively) (Fig. 6B). NAD+ and ATP deficits recovered to control levels after 1 week. At 48 h after the injection of NMDA (20 nmol) no significant changes were detected in the levels of ATP and NAD+ (Table 1C). On the contrary, at 48 h AMPA (5 nmol) produced a decrease in NAD+ ( 26%) which was greater
Fig. 6. Time-course of changes in striatal levels of (A) NAD+ and (B) ATP following injection of 1.0 nmol of KA or PBS into the right striatum of C57Bl/6N mice killed by head focused microwave irradiation. Levels of NAD+ and ATP were measured by HPLC with UV detection. Values are meansFS.E.M. of 5 – 16 animals per group and are expressed in nmol per mg of protein. *p<0.05, **p<0.01 Mann – Whitney U-test, KA-injected versus PBS-injected striata. ., KA injected striata; E, PBS injected striata.
6
C. Cosi et al. / Brain Research 996 (2004) 1–8
Table 1 Benzamide (2 160 mg/kg i.p.) prevents the decreases in striatal NAD+ and ATP induced by local injection of KA (nmol) (A) but not the decreases in striatal NAD+ and ATP induced by local injection of AMPA (5 nmol) (B), at 48 h (C) Lack of effect of benzamide or local injection of NMDA (20 nmol) on striatal levels of NAD+ and ATP Treatments Intrastriatal + i.p.
NAD+ (nmol/mg protein), mean F S.E.M.
ATP (nmol/mg protein) mean F S.E.M.
A PBS + vehicle KA + vehicle PBS + benzamide KA + benzamide
3.893 F 0.056 3.257 F 0.140*** 4.037 F 0.164 3.572 F 0.061**,#
23.45 F 0.615 19.99 F 1.141** 24.73 F 0.897 22.41 F 0.828
B PBS + vehicle AMPA + vehicle PBS + benzamide AMPA + benzamide
4.432 F 0.072 3.266 F 0.259* 4.327 F 0.056 3.133 F 0.321***
25.12 F 0.583 22.92 F 1.096 23.43 F 0.617 19.96 F 1.688
C PBS + vehicle NMDA + vehicle PBS + benzamide NMDA + benzamide
3.498 F 0.095 3.408 F 0.095 3.616 F 0.107 3.462 F 0.101
21.65 F 1.028 20.58 F 0.732 22.39 F 0.63 22.16 F 0.633
Values are mean F S.E.M. of 7 – 10 animals per group and are expressed in nmol per mg of protein. Kruskal – Wallis ANOVA followed by Mann – Whitney U-test. Abbreviations: i.p.; intraperitoneal. PBS, phosphatebuffered saline. * p < 0.05 vs. [PBS + vehicle] groups. ** p < 0.01. *** p < 0.001. # p < 0.05 vs. [KA + vehicle] groups.
NAD+ at 48 h (Table 1A). Benzamide did not affect the AMPA-induced decrease of NAD+ at 48 h (Table 1B). Benzamide by itself had no effects on striatal ATP and NAD+ levels in sham-operated mice at 48 h (Table 1).
4. Discussion The mechanisms of KA-induced neurotoxicity are not entirely understood but they appear to be associated with massive formation of free radicals due to membrane lipid peroxidation and to impairment of energy metabolism [15], events known to be involved in PARP-mediated cell death [17] Consistent with an earlier report [23], we found that the intrastriatal injection of KA in the rat produced an acute decrease in ATP and an observed a smaller decrease of NAD+. We have previously shown in the same model that KAinduced neurotoxicity involves a time-dependent activation of PARP [10]: no change in PARP activity at 6 h, an increase of at 24 h, and an increase in both PARP activity and protein levels after 1 week. In the present study, the KAinduced lesion volume, as measured by Nissl staining, was barely detectable at 6 h, was well-developed at 24 and 48 h and was no longer evident after 1 week. These changes
paralleled the effects of KA on striatal NAD+ and ATP levels: no effect at 6 h, decreases at 24 and 48 h, and recovery at 1 week. The magnitude of these effects (15 – 20%) might be underestimated since the dissected striatum includes not only the «penumbral area» around the lesion, where dying cells depleted of ATP and NAD+ are located, but also undamaged tissue outside the lesion area. These results suggested a possible correlation, at least during the first three days following intrastriatal injection of KA, between the increase of PARP catalytic activity [10], NAD+ and ATP decreases, and cell death (present study). To further investigate this possible correlation, we examined the effects of the PARP inhibitor benzamide on the lesion volume and on the deficits of striatal NAD+ and ATP induced by KA at 48 h. At concentrations approaching 5 mM, benzamide has been reported to inhibit glucose metabolism and DNA synthesis. However, when used at 0.1– 1 mM it is a selective inhibitor of PARP (see references in Ref. [6]). In our previous studies benzamide was measured in the striatum of C57BL mice after i.p. injection of the neuroprotective dose of 160 mg/kg i.p. [6,8]; at this dose, the compound was found to achieve a maximal brain concentration of 0.64 mM at 30 min post-injection, and was still present in the brain 2 h after the injection (see references in Ref. [6]). In the present study, benzamide, used at the same dose and injection protocol that affords partial protection against MPTP and methamphetamine neurotoxicities [8,9], partially prevented both the development of the striatal lesion and the decrease NAD+, and completely prevented the drop of striatal ATP observed at 48 h after KA injection, without affecting energy metabolites in sham-operated animals. Since the AMPA-induced lesion and NAD+ decrease are resistant to benzamide treatment, it cannot be ruled out that the residual lesion volume and NAD+ decrease after benzamide treatment that persisted in KA lesioned striata could be related to AMPA receptor activation. Taken together with the increase of PARP activity at 24 h [10], these data are consistent with a causative role of PARP overactivationdependent energy depletion during the early development of the KA lesion. At later times, PARP-mediated mechanisms other than energy depletion may be involved in the observed disappearance of the lesion and the recovery of energy levels. In a previous study [10], striatal PARP activity and protein levels were found to be markedly increased at 1 week following intrastriatal KA injection. PARP activity and protein increase during proliferation and differentiation of astrocytes in culture [4] and systemic administration of KA induces astrogliosis [1]. In the present study GFAP-positive cells were detected from the earliest time points of KA toxicity, and increased in number and in staining intensity over time. At ten days, the presence of reactive astroglia at the site of the lesion might suggest a restoration of energy levels by these cells in support of the neurons surrounding the glial scar. Further studies are needed to investigate if the recovery
C. Cosi et al. / Brain Research 996 (2004) 1–8
of energy metabolism is due to astroglial cells and whether this is associated with functional recovery. Since PARP-1 knock-out mice are highly resistant to NMDA-induced neurotoxicity, we investigated if the NMDA-induced lesion in wild type mice could be prevented by treatment with benzamide, and if the eventual protective effects could be correlated with preservation of energy metabolism. NMDA, at the same dose as used by Mandir et al. [21], induced a lesion whose volume was reduced by treatment with benzamide, as measured at 48 h. However, NMDA had little or no effect on ATP and NAD+, at 1 h or at 48 h, in apparent contrast with the finding of Mandir et al. [21], where an increase in poly(ADP-ribose) was found during the first 24 h following NMDA injection in wild type mice. Nevertheless, NMDA produced a lesion volume approximately 50% larger than that induced by KA, without any expected proportional decrease in energy metabolites at 48 h. It cannot be ruled out that NMDA could cause energy decrease at times other then those observed in the present study (1 and 48 h), as suggested by a 7% decrease in ATP observed after NMDA injection. Also, the time course of energy recovery (possibly by astrogliosis) might occur faster in NMDA toxicity than in KA toxicity, which could explain the lack of energy depletion at 48 h after NMDA administration. On the other hand, these findings can also suggest that other PARP-mediated mechanisms independent of energy depletion and prevented (at least in part) by inhibition of PARP catalytic activity might be prevalent in NMDAinduced neurotoxicity in vivo. In fact, the increase in poly(ADP-ribose) [21] could have been due to an increased expression of PARP protein that is one of the major acceptors of the polymer of ADP-ribose. Regulation of gene expression by PARP through protein –protein interactions with transcription factors such as NF-kappaB, might represent an important mechanism of inflammation-mediated secondary damage in excitotoxicity [18] and in particular in NMDA-induced toxicity [26]. Moreover, this property of PARP to regulate gene expression, by forming multiprotein complexes with transcription factors independent of [19] or dependent on [5] auto-ADP-ribosylation of PARP, might also explain the strong resistance of PARP-1 knock-out mice to cerebral ischemia [13] and NMDA excitotoxicity [21]. Although it is currently difficult to discriminate between AMPA and KA receptors for lack of selective agonists and antagonists, studies utilizing mice deficient in specific KA receptor subunits (GluR5 and GluR6) have resolved the existence of AMPA and KA receptor entities (see Ref. [15] for a review) In our study, AMPA administered at the same dose used by Mandir et al. [21] induced a lesion which was not prevented by benzamide, in agreement with the lack of protection by PARP-1 / phenotype against AMPA toxicity. Early treatment with benzamide did not prevent AMPAinduced NAD+ and ATP depletions at 48 h suggesting that PARP activation is not implicated in the control of energy metabolism in the initiation of AMPA toxicity. Changes in NAD+ and ATP could be due to other mechanisms unrelated
7
to PARP and insensitive to benzamide, e.g. impairment of mitochondrial respiration. On the other hand, PARP is not activated in the first 24 h following AMPA administration in wild type mice [21], consistent with a lack of involvement of PARP in AMPA toxicity, at least during the first 24 h. Alternatively, since PARP inhibition by benzamide covers only the early phases of neurodegenerative process that is quantified at 48 h, it cannot be ruled out that differences in the time-course of neurotoxicity for KA and AMPA may account for the lack of effect of benzamide on AMPA neurotoxicity. While further studies are needed to investigate the different PARP-mediated mechanisms induced by overactivation of glutamate receptor subtypes, these results indicate that PARP overactivation and energy depletion could be responsible in part for the cellular demise during the development of the lesion induced by KA, confirm that PARP is involved in NMDA but not AMPA toxicities, and suggest the existence of differences between KA and AMPA toxicities. Furthermore, the results here obtained with benzamide are similar to those reported for the PARP-1 / phenotype, suggesting that neuroprotective action of benzamide is mediated by selective inhibition of PARP-1. This study provides further evidence supporting PARP as a novel target for new drug treatments against neurodegenerative disorders.
Acknowledgements The authors gratefully acknowledge the excellent technical assistance of N. Leduc, A.-M. Ormie`re, J. Floutard and O. Marteau.
References [1] C.A. Altar, M. Baudry, Systemic injection of kainic acid: gliosis in the olfactory and limbic brain regions quantified with [3H] PK 11195 binding autoradiography, Exp. Neurobiol. 109 (1990) 333 – 341. [2] C. Ayata, G. Ayata, H. Hara, R.T. Matthews, M.F. Beal, R.J. Ferrante, M. Endres, A. Kim, R.H. Christie, C. Waeber, P.L. Huang, B.T. Hyman, M.A. Moskowitz, Mechanisms of reduced striatal NMDA excitotoxicity in type I nitric oxide synthase knock-out mice, J. Neurosci. 17 (1997) 6908 – 6917. [3] N.A. Berger, Poly(ADP-ribose) in the cellular response to DNA damage, Radiat. Res. 101 (1985) 4 – 15. [4] G.M. Chambert, P.C. Niedergang, F. Hog, M. Partisani, P. Mandel, Poly(ADPR) polymerase expression and activity during proliferation and differentiation of rat astrocyte and neuronal cultures, Biochim. Biophys. Acta 1136 (1992) 196 – 202. [5] W.J. Chang, R. Alvarez-Gonzales, The sequence-specific DNA binding of NF-kappa B is reversibly regulated by the automodification reaction of poly(ADP-ribose) polymerase 1, J. Biol. Chem. 276 (2001) 47664 – 47670. [6] C. Cosi, M. Marien, Decreases in mouse brain NAD+ and ATP induced by 1-methyl-4-phenyl-1, 2,3,6-tetrahydropyridine (MPTP): prevention by the poly(ADP-ribose) polymerase inhibitor, benzamide, Brain Res. 809 (1998) 58 – 67.
8
C. Cosi et al. / Brain Research 996 (2004) 1–8
[7] C. Cosi, H. Suzuki, D. Milani, L. Facci, M. Menegazzi, G. Vantini, Y. Kanai, S.D. Skaper, Poly(ADP-ribose) polymerase: early involvement in glutamate-induced neurotoxicity in cultured cerebellar granule cells, J. Neurosci. Res. 39 (1994) 38 – 46. [8] C. Cosi, P. Chopin, M. Marien, Benzamide, an inhibitor of poly (ADP-ribose) polymerase, attenuates methamphetamine-induced dopamine neurotoxicity in the C57Bl/6N mouse, Brain Res. 735 (1996) 343 – 348. [9] C. Cosi, F. Colpaert, W. Koek, A. Degryse, M. Marien, Poly(ADPribose) polymerase inhibitors protect against MPTP-induced depletions of striatal dopamine and cortical noradrenaline in C57B1/6 mice, Brain Res. 729 (1996) 264 – 269. [10] C. Cosi, E. Cavalieri, A. Carcereri de Prati, M. Marien, H. Suzuki, Effects of kainic acid lesioning on poly(ADP-ribose)polymerase (PARP) activity in the rat striatum in vivo, Amino Acids 19 (2000) 229 – 237. [11] J.T. Coyle, P. Puttfarcken, Oxidative stress, glutamate, and neurodegenerative disorders, Science 262 (1993) 689 – 695. [12] G. De Murcia, V. Schreiber, M. Molinete, B. Saulier, O. Poch, M. Masson, C. Niedergang, J. Menissier de Murcia, Structure and function of poly(ADP-ribose) polymerase, Mol. Cell. Biochem. 138 (1994) 15 – 24. [13] M.J. Eliasson, K. Sampei, A.S. Mandir, P.D. Hurn, R.J. Traystman, J. Bao, A. Pieper, Z.Q Wang, T.M. Dawson, S.H. Snyder, V.L. Dawson, Poly(ADP-ribose) polymerase gene disruption renders mice resistant to cerebral ischemia, Nat. Med. 3 (1997) 1089 – 1095. [14] M. Endres, Z.Q. Wang, S. Namura, C. Waeber, M.A. Moskowitz, Ischemic brain injury is mediated by the activation of poly(ADPribose)polymerase, J. Cereb. Blood Flow Metab. 17 (1997) 1143 – 1151. [15] A.A. Farooqui, Y.W. Ong, X.-R. Lu, B. Halliwell, A.L. Horrocks, Neurochemical consequences of kainate-induced toxicity in brain: involvement of arachidonic acid relase and prevention of toxicity by phospholipase A2 inhibitors, Brain Res. Rev. 38 (2001) 61 – 78. [16] K.B.J. Franklin, G. Paxinos, The Mouse Brain in Streretaxic Coordinates, Academic Press, San Diego, CA, 1997.
[17] C.H. Ha, S.H. Snyder, Poly(ADP-ribose) polymerase-1 in the nervous system, Neurobiol. Dis. 7 (2000) 225 – 239. [18] C.H. Ha, L. Hester, S.H. Snyder, Poly(ADP-ribose)polymerase-1 dependence of stress-induced transcription factors and associated gene expression in glia, Proc. Natl. Acad. Sci. 99 (2002) 3270 – 3275. [19] P.O. Hassa, M.O. Hottiger, A role of poly (ADP-ribose)polymerase in NF-kappaB transcriptional activation, Biol. Chem. 380 (1999) 953 – 959. [20] A.S. Mandir, S. Przedborski, V. Jackson-Lewis, Z.Q. Wang, C.M. Simbulan-Rosental, B.E. Hoffman, D.B. Guastella, V.L. Dawson, T.M. Dawson, Poly(ADP-ribose)polymerase activation mediates 1methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-induced parkinsonism, Proc. Natl. Acad. Sci. 96 (1999) 5774 – 5779. [21] A.S. Mandir, M.F. Poitras, A.R. Berliner, W.J. Herring, D.B. Guastella, A. Feldman, G.G. Poirier, Z.Q Wang, T.M. Dawson, V.L. Dawson, NMDA but not non-NMDA excitotoxicity is mediated by Poly(ADP-ribose)polymerase, J. Neurosci. 20 (2000) 8005 – 8011. [22] K. Plaschke, J. Kopitz, M.A. Weigand, E. Martin, H.J. Bardenheuer, The neuroprotective effect of cerebral poly(ADP-ribose)polymerase inhibition in a rat model of global ischemia, Neurosci. Lett. 284 (2000) 109 – 112. [23] K.C. Retz, J.T. Coyle, Kainic acid lesion of mouse striatum: effects on energy metabolites, Life Sci. 27 (1980) 2495 – 2500. [24] P. Sheng, C. Cerruti, J.L. Cadet, Methamphetamine (METH) causes reactive gliosis in vitro: attenuation by the ADP-ribosylation (ADPR) inhibitor, benzamide, Life Sci. 55 (1994) L51 – L54. [25] K. Takahashi, A.A. Pieper, S.E. Croul, J. Zhang, S.H. Snyder, J.H. Greenberg, Post-treatment with an inhibitor of poly(ADP-ribose) polymerase attenuates cerebral damage in focal ischemia, Brain Res. 829 (1999) 46 – 54. [26] O. Ulrich, A. Diestel, I.Y. Eyu¨poglu, R. Nitsch, Regulation of microglia expression of integrins by poly(ADP-ribose) polymerase, Nat. Cell Biol. 3 (2001) 1035 – 1042. [27] J. Zhang, V.L. Dawson, T.M. Dawson, S.H. Snyder, Nitric oxide activation of poly(ADP-ribose) synthetase in neurotoxicity, Science 263 (1994) 687 – 689.
Research report
The PARP inhibitor benzamide protects against kainate and NMDA but not AMPA lesioning of the mouse striatum in vivo Cristina Cosi *, Karen Guerin, Marc Marien, Wouter Koek, Karin Rollet Divisions de Neurobiologie I et II, Centre de Recherche Pierre Fabre, 17 avenue Jean Moulin, 81106 Castres, France Accepted 8 September 2003
Abstract Overactivation of poly(ADP-ribose) polymerase (PARP) in response to genotoxic insults can cause cell death by energy deprivation. We previously reported that neurotoxic amounts of kainic acid (KA) injected into the rat striatum produce time-dependent changes in striatal PARP activity in vivo. Here, we have investigated the time-course of KA-induced toxicity and the effects of the PARP inhibitor benzamide on KA, AMPA and NMDA neurotoxicities in vivo, by measuring changes in the volume of the lesion and in NAD+ and ATP levels induced by the intra-striatal injection of these excitotoxins in C57Bl/6N mice. The KA-induced lesion volume was dependent on the amount of toxin injected and the survival time. The lesion was well developed at 48 h and was almost undetectable after one week. KA produced an extensive astrogliosis at one week. Benzamide partially prevented both KA- and NMDA- but not AMPA-induced lesions when measured at 48 h after the treatment. The effects of benzamide appeared to be in part related to changes in energy metabolism, since KA produced decreases in striatal levels of NAD+ and ATP that were partially prevented by benzamide at 48 h and which returned to control levels at one week. NMDA did not affect NAD+ and induced little alteration in ATP levels. Benzamide had no effect on AMPA-induced decreases in either NAD+ or ATP levels at 48 h. These results (1) indicate that PARP overactivation and energy depletion could be responsible in part for the cellular demise during the development of the lesion induced by KA; (2) confirm that PARP is involved in NMDA but not AMPA toxicities; (3) suggest the existence of differences between KA and AMPA-mediated toxicities; and (4) provide further evidence supporting PARP as a novel target for new drug treatments against neurodegenerative disorders. D 2003 Elsevier B.V. All rights reserved. Theme: Neurotransmitters, modulators, transporters, and receptors Topic: Excitatory amino acids: excitotoxicity Keywords: Poly(ADP-ribose) polymerase; Excitotoxicity; Kainate; NMDA; AMPA; Benzamide; C57Bl/6N mouse
1. Introduction Poly(ADP-ribose) polymerase (PARP) is a DNA binding protein that uses NAD+ as substrate. PARP is activated by strand breaks in the DNA molecule that can be induced by DNA damaging agents, including free radicals [12]. When fully activated, for example by free radical-induced DNA damage, PARP can deplete cellular energy stores, under the form of NAD+ and ATP, predisposing the cell to death [3]. In vitro and in vivo studies using PARP inhibitors, including benzamide, and/or mice or cells with a disrupted PARP-1 gene, have indicated a participation of PARP in glutamate-[7] and N-methyl-D-aspartate (NMDA)-induced * Corresponding author. Tel.: +33-5-63-71-42-86; fax: +33-5-63-7143-63. E-mail address: [email protected] (C. Cosi). 0006-8993/$ - see front matter D 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2003.09.072
[21,27] neurotoxicities, cerebral ischemia [13,14,22,25], and 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) [6,9,20], and methamphetamine-induced [8,24] neurotoxicities. In particular, it has been shown that the PARP inhibitor benzamide is neuroprotective in C57Bl/6N mice against different types of neurotoxicities [8,9] without affecting body temperature [8]. Kainic acid-induced (KA) neurotoxicity is related to free radical formation via membrane lipid peroxidation and the arachidonic acid cascade and ATP depletion [11,15]. In particular, KA injected into the striatum of mice causes a rapid decline (by about 30%) in striatal levels of ATP [23]. We have previously shown that a nanomolar amount of KA injected into the striatum of the rat can produce timedependent changes in striatal PARP activity [10]. In the present study, we investigated whether the timerelated changes in striatal PARP activity induced by intra-
2
C. Cosi et al. / Brain Research 996 (2004) 1–8
striatal KA [10] were correlated or not with PARP-mediated cell death by energy depletion, by studying the time course effects of KA and the effects of the PARP inhibitor benzamide on lesion volume and striatal ATP and NAD+ levels. Since PARP-1 knock-out mice are completely resistant to NMDA-induced lesion but not a-amino-3-hydroxy-5-methyl-4-isoxazole-4-propionique acid (AMPA)-induced lesions [21], we also investigated if pharmacological inhibition of PARP with benzamide would protect C57Bl/6N mice against NMDA and AMPA neurotoxicities.
2. Materials and methods 2.1. Animals Animals were handled and cared for in accordance with the Guide for the Care and Use of Laboratory Animals (National Research Council, 1996) and the European Directive No 86/609. The experimental protocol was carried out in compliance with local ethical committee guidelines for animal research. Male C57Bl/6N mice (26 – 28 g, Charles River, Elbeuf, France) were housed in groups of 10 with free access to water and food, under a 12-h light/ dark cycle. 2.2. Excitotoxic lesions Twenty-four hours before surgery, mice were weighed and isolated. Mice were anesthetized with an intra-peritoneal (i.p.) injection of sodium pentobarbital (65 mg/kg). The body temperature was monitored by a rectal probe and maintained at 37 jC by a thermostated heating pad. Excitatory amino acids were dissolved in PBS (pH adjusted to 7.4 with 1 N NaOH). Mice were placed in a stereotaxic apparatus (Kopf model 900) and NMDA (20 nmol), AMPA (5.0 nmol) [21], kainic acid (0.5, 1.0 or 1.5 nmol), or vehicle (PBS 10 mM) were injected in a volume of 0.3 Al into the right striatum according to the following stereotaxic coordinates: 0.5 mm anterior to bregma, 2.0 mm lateral to bregma, and 3.5 mm ventral from the skull surface [16]. At the end of the surgery, a 1 ml subcutaneous (s.c.) injection of 0.9% NaCl was administrated to prevent dehydration, and mice were placed in a thermostated cage (25 jC) until they were completely awake. During the next 48 h, mice received two injections of 0.9% NaCl per day. The PARP inhibitor benzamide (160 mg/kg) was prepared in 0.9% NaCl containing 0.01% of carboxymethylcellulose. Benzamide or vehicle was administered i.p. 30 min before, and 3.5 h after the intrastriatal injection of the neurotoxin or vehicle [8,9]. All drugs were purchased from Sigma, France. 2.3. Estimation of lesion volume Mice were killed by decapitation at different times after the intrastriatal injections. Brains were dissected and frozen
in isopentane at 45 jC, then stored at 80 jC until processed. Brains were sectioned, starting from the anterior aspect of the corpus callosum, into 240 sequential rostrocaudal sections of 20-Am thickness. These sections encompassed both the entire span of the lesioned tissue and the entire striatum. One out of every 10 sections was collected on polylysine-coated slides, fixed in 4% paraformaldehyde (0 jC, 30 min) and stained with cresyl violet (1% cresyl violet acetate in 0.1 M acidic acid, pH 4). The area of the lesion, indicated by the lack of staining when compared to the contralateral (intact) striatum, was measured in each series of sections by means of an image analysis program (Scion Image, NIH) and the lesion volume was given by the sum of the serial lesion volumes calculated using the known distance between sections. Occurrence of variability for the extension of the lesion volume obtained with NMDA and AMPA, when compared with that reported previously [21] might be due differences in mouse strains and lesioning procedures image analysis procedures. 2.4. Immunocytochemistry for glial fibrillary acidic protein Immunocytochemistry for glial fibrillary acidic protein (GFAP) was performed on sections adjacent to those utilized for the estimation of the lesion volume. Sections were collected on polylysine-coated slides and fixed in 4% paraformaldehyde (0 jC, 30 min). After antigen retrieval by trypsin (0.05% in 0.1% CaCl2) and inactivation of endogenous peroxidases, slides were incubated for 1 h with 10% goat serum in PBS in order to block non-specific sites and then incubated at 4 jC overnight with mouse monoclonal antibody anti-GFAP (diluted 1:1000; Sigma). After washing with PBS, slides were incubated for 45 min at room temperature with goat-anti mouse secondary antibody peroxidase-conjugated (diluted 1:200; Chemicon International). After incubation with the peroxidase-antiperoxidase (diluted 1:200; Chemicon International) the antibody was revealed by 3V-diaminobenzidine oxidation. 2.5. Measurement of striatal ATP and NAD + levels Measurements of striatal ATP and NAD+ levels were measured as previously described [6]. Mice were killed by head-focused microwave irradiation (3.4 kW, 0.8 – 1.1 s; Sacronk Model 8000, SAIREM, Vaulx-en-Velin, France) at different times after the intra-striatal injection of the different neurotoxins. This method of sacrifice is essential for rapid and effective fixation of the brain in situ, to permit accurate determination of metabolites which are susceptible to post-mortem changes. Brains were then removed on ice (+4 jC) and striata were dissected and stored at 80 jC until processed. The samples were homogenized by brief sonication in a solution containing perchloric acid (0.2 M), EDTA (1 mM) and Na2S2O5 (1 mM), and were then centrifuged at 17 400g for 20 min at +4 jC. The supernatants were filtered through Millex HV 0.45 Am filters (Millipore, France) and
C. Cosi et al. / Brain Research 996 (2004) 1–8
3
analyzed by HPLC (Alliance 2690) with UV detection (Waters Model 486). Aliquots of 20 Al were automatically injected by a refrigerated (+8 jC) autosampler. Absorbance wavelength for detection was set at 260 nm. Values were expressed in nmol per mg of protein. Proteins were measured according to Lowry.
3. Results 3.1. Dose-dependency of the kainic acid induced-lesion Since previous studies using AMPA or NMDA indicated that a well-developed lesion volume is obtained 48 h after the intra-striatal injection of the toxins [2], the present study evaluated the effects of different doses of kainic acid (KA) on the volume of the lesion at this time point (Fig. 1). Intrastriatal injection of KA produced a lesion whose volume was dose-dependent. KA (0.5 nmol) produced a lesion volume of 0.50F0.10 mm3. Maximal lesion volume was obtained with KA at 1.5 nmol (6.14F1.39 mm3), but this was not significantly different from the lesion obtained with 1.0 nmol (4.39F1.33 mm3). 3.2. Time-dependency of kainic acid induced lesion Kainic acid (1.25 nmol) caused a lesion whose development depended on time (Fig. 2). At 1 h no lesion was detectable. At 6 h a lesion volume of 0.59F0.25 mm3 was visible. At 24 h the lesion was well developed and the volume measured 1.54F0.27 mm3. At 48 h the lesion volume was 1.78F0.36 mm3. At 1 week after the injection of KA, the lesion volume had decreased to 0.08F0.03 mm3.
Fig. 1. Dose-related changes in the volume of the lesion 48 h after the intrastriatal injection of kainic acid (KA) into the right striatum of C57Bl/6N mice. Values, in mm3, are meansFS.E.M. of four animals per group. *p<0.05 compared to KA (0.5 nmol), Kruskal – Wallis ANOVA followed by Mann – Whitney U-test.
Fig. 2. Time-related changes in the volume of the lesion induced by the intra-striatal injection of KA (1.25 nmol) in C57Bl/6N mice. Values, in mm3, are meansFS.E.M. of six to seven animals per group. *p<0.05 compared to 6 h and compared to 1 week, Kruskal – Wallis ANOVA followed by Mann – Whitney U-test.
3.3. GFAP immunostaining In PBS injected animals, at 6 h after injection, a lightly labeled diffuse pattern of GFAP immunopositive cells with stellated shape was observed around the lateral ventricles, within the corpus callosum and the septal area, and in the proximity of the needle tract. At 48 h after PBS injection, strongly labeled astrocytes were found in the proximity of the needle tract and adjacent to the lateral ventricle on the injected side. A more diffuse astrocyte reaction was found also in the septum, corpus callosum, anterior commissure and ventral thalamus. At 1 week after PBS injection, astrocytes were accumulated along the needle tract. In KA injected (1.5 nmol) animals, strongly labeled GFAP-positive cells were observed at 6 h in the same areas as in the PBS injected animals at 6 h. These cells presented few arborizations. At 48 h a diffuse astrogliosis was observed in both hemispheres with widely distributed strongly labeled astrocytes in the area around the lesion. At 48 h, a dense distribution of strongly labeled-GFAP positive astrocytes was observed in the intermediate layer of the parietal and pyriform cortices adjacent to the lateral ventricle ispilateral to the lesion. At 1 week after KA injection (Fig. 3, upper and lower panels), the area corresponding to the lesion was filled with strongly labeled GFAP-positive astrocytes, larger in size than those observed at 48 h and with long interconnecting extensions. A more diffuse astrogliosis was observed throughout both hemispheres. Treatment with benzamide did not affect the patterns of distribution of the astrocytes either in the PBS or in the KA injected animals, at 48 h. 3.4. Benzamide affords partial protection against kainateand NMDA-but not AMPA-induced neurotoxicities The effects of benzamide on KA-, NMDA- and AMPAinduced lesions were investigated by measuring changes in
4
C. Cosi et al. / Brain Research 996 (2004) 1–8
Fig. 3. Kainic acid-induced astrogliosis. Glial fibrillary acidic protein (GFAP) immunostaining at 10 days after KA injection (1.5 nmol) into the right striatum. Upper panel: injected side with numerous strongly stained GFAP immunopositive cells. Scale bar =75 Am. Lower panel: highmagnification micrographs showing hypertrophic astrocytes. Scale bar=3 Am.
lesion volume at 48 h after the intra-striatal injections of the toxins and the i.p. treatment with benzamide (Figs. 4 and 5). Benzamide administered i.p. 30 min before and 3.5 h after the injection of 1.0 nmol of KA reduced the volume of the lesion by 59%, from 2.22F0.36 to 0.92F0.32 mm3 (Fig. 4A). The intra-striatal injection of 20 nmol of NMDA produced a lesion volume of 3.40F0.72 mm3 at 48 h (Fig. 4B). Benzamide treatment afforded partial protection against the toxicity of NMDA, allowing a 62% reduction of the volume of the lesion (1.31F0.38 mm3). The intra-striatal injection of 5.0 nmol of AMPA produced a lesion volume of 1.96F0.34 mm3 at 48 h. Benzamide treatment provided no protection against AMPAinduced excitotoxicity (Fig. 4C). 3.5. Time-course of kainate-induced changes in striatal levels of ATP and NAD+ The time-course of effects of intra-striatal injection KA (1 nmol) on striatal ATP and NAD+ levels were investigated in mice sacrificed by head focused microwave irradiation at
Fig. 4. Benzamide treatment decreases the volume of the lesion induced by intrastriatal injection of (A) KA (1.0 nmol) and (B) NMDA (20 nmol) but not (C) AMPA (5 nmol) in mice. Values, in mm3, are meansFS.E.M. of eight animals per group. *p<0.05, Mann – Whitney U-test.
C. Cosi et al. / Brain Research 996 (2004) 1–8
5
than that obtained with KA ( 15%) at the same time point (Table 1).
3.6. Effects of benzamide on excitotoxin-induced changes in striatal levels of ATP and NAD + KA, NMDA and AMPA differentially affected striatal levels of ATP and NAD+ at 48 h when compared to the striata injected with PBS in controls. KA (1 nmol) caused a 15% decrease in ATP levels and reduced NAD+ by 16% when compared to the PBS injected controls (Table 1A), whereas AMPA produced a significant and pronounced decrease in NAD+ ( 26% compared to the PBS injected controls) (Table 1B). No changes in striatal NAD+ and ATP were detected at 48 h after NMDA intrastriatal administration. Benzamide (160 mg/kg i.p.) administrated 30 min before and 3.5 h after KA injection completely prevented the decrease in ATP, and partially reduced the decrease of
Fig. 5. Cresyl violet stained sections of the brains of mice injected in the right striatum with either PBS or excitotoxin and treated with benzamide (2160 mg/kg i.p.). (Right panel) or vehicle (left panel). Benzamide provided partial neuroprotection against KA and NMDA-induced toxicities, but not against AMPA-induced toxicity. The lack of staining with cresyl violet indicates the area of the lesion.
different times after KA injection (Fig. 6A and B). KA caused a rapid and transient decrease in ATP levels (Fig. 6B) in the KA-injected striatum compared to the PBS injected control. ATP was reduced by 31% at 1 h, at which time a smaller ( 9%) decrease of NAD+ was also observed (Fig. 6B). NMDA (20 nmol) decreased striatal ATP by 7% at 1 h, compared to the side contralateral to the injection, and did not affect NAD+. In the KA treated animals, ATP and NAD+ were both back to control levels at 6 h. A second phase of decreases followed at later times: NAD+ levels were reduced at 24 h ( 21%) and at 48 h ( 15%) (Fig. 6A) and were paralleled by decreases in striatal ATP ( 19% and 10%, at 24 and 48 h, respectively) (Fig. 6B). NAD+ and ATP deficits recovered to control levels after 1 week. At 48 h after the injection of NMDA (20 nmol) no significant changes were detected in the levels of ATP and NAD+ (Table 1C). On the contrary, at 48 h AMPA (5 nmol) produced a decrease in NAD+ ( 26%) which was greater
Fig. 6. Time-course of changes in striatal levels of (A) NAD+ and (B) ATP following injection of 1.0 nmol of KA or PBS into the right striatum of C57Bl/6N mice killed by head focused microwave irradiation. Levels of NAD+ and ATP were measured by HPLC with UV detection. Values are meansFS.E.M. of 5 – 16 animals per group and are expressed in nmol per mg of protein. *p<0.05, **p<0.01 Mann – Whitney U-test, KA-injected versus PBS-injected striata. ., KA injected striata; E, PBS injected striata.
6
C. Cosi et al. / Brain Research 996 (2004) 1–8
Table 1 Benzamide (2 160 mg/kg i.p.) prevents the decreases in striatal NAD+ and ATP induced by local injection of KA (nmol) (A) but not the decreases in striatal NAD+ and ATP induced by local injection of AMPA (5 nmol) (B), at 48 h (C) Lack of effect of benzamide or local injection of NMDA (20 nmol) on striatal levels of NAD+ and ATP Treatments Intrastriatal + i.p.
NAD+ (nmol/mg protein), mean F S.E.M.
ATP (nmol/mg protein) mean F S.E.M.
A PBS + vehicle KA + vehicle PBS + benzamide KA + benzamide
3.893 F 0.056 3.257 F 0.140*** 4.037 F 0.164 3.572 F 0.061**,#
23.45 F 0.615 19.99 F 1.141** 24.73 F 0.897 22.41 F 0.828
B PBS + vehicle AMPA + vehicle PBS + benzamide AMPA + benzamide
4.432 F 0.072 3.266 F 0.259* 4.327 F 0.056 3.133 F 0.321***
25.12 F 0.583 22.92 F 1.096 23.43 F 0.617 19.96 F 1.688
C PBS + vehicle NMDA + vehicle PBS + benzamide NMDA + benzamide
3.498 F 0.095 3.408 F 0.095 3.616 F 0.107 3.462 F 0.101
21.65 F 1.028 20.58 F 0.732 22.39 F 0.63 22.16 F 0.633
Values are mean F S.E.M. of 7 – 10 animals per group and are expressed in nmol per mg of protein. Kruskal – Wallis ANOVA followed by Mann – Whitney U-test. Abbreviations: i.p.; intraperitoneal. PBS, phosphatebuffered saline. * p < 0.05 vs. [PBS + vehicle] groups. ** p < 0.01. *** p < 0.001. # p < 0.05 vs. [KA + vehicle] groups.
NAD+ at 48 h (Table 1A). Benzamide did not affect the AMPA-induced decrease of NAD+ at 48 h (Table 1B). Benzamide by itself had no effects on striatal ATP and NAD+ levels in sham-operated mice at 48 h (Table 1).
4. Discussion The mechanisms of KA-induced neurotoxicity are not entirely understood but they appear to be associated with massive formation of free radicals due to membrane lipid peroxidation and to impairment of energy metabolism [15], events known to be involved in PARP-mediated cell death [17] Consistent with an earlier report [23], we found that the intrastriatal injection of KA in the rat produced an acute decrease in ATP and an observed a smaller decrease of NAD+. We have previously shown in the same model that KAinduced neurotoxicity involves a time-dependent activation of PARP [10]: no change in PARP activity at 6 h, an increase of at 24 h, and an increase in both PARP activity and protein levels after 1 week. In the present study, the KAinduced lesion volume, as measured by Nissl staining, was barely detectable at 6 h, was well-developed at 24 and 48 h and was no longer evident after 1 week. These changes
paralleled the effects of KA on striatal NAD+ and ATP levels: no effect at 6 h, decreases at 24 and 48 h, and recovery at 1 week. The magnitude of these effects (15 – 20%) might be underestimated since the dissected striatum includes not only the «penumbral area» around the lesion, where dying cells depleted of ATP and NAD+ are located, but also undamaged tissue outside the lesion area. These results suggested a possible correlation, at least during the first three days following intrastriatal injection of KA, between the increase of PARP catalytic activity [10], NAD+ and ATP decreases, and cell death (present study). To further investigate this possible correlation, we examined the effects of the PARP inhibitor benzamide on the lesion volume and on the deficits of striatal NAD+ and ATP induced by KA at 48 h. At concentrations approaching 5 mM, benzamide has been reported to inhibit glucose metabolism and DNA synthesis. However, when used at 0.1– 1 mM it is a selective inhibitor of PARP (see references in Ref. [6]). In our previous studies benzamide was measured in the striatum of C57BL mice after i.p. injection of the neuroprotective dose of 160 mg/kg i.p. [6,8]; at this dose, the compound was found to achieve a maximal brain concentration of 0.64 mM at 30 min post-injection, and was still present in the brain 2 h after the injection (see references in Ref. [6]). In the present study, benzamide, used at the same dose and injection protocol that affords partial protection against MPTP and methamphetamine neurotoxicities [8,9], partially prevented both the development of the striatal lesion and the decrease NAD+, and completely prevented the drop of striatal ATP observed at 48 h after KA injection, without affecting energy metabolites in sham-operated animals. Since the AMPA-induced lesion and NAD+ decrease are resistant to benzamide treatment, it cannot be ruled out that the residual lesion volume and NAD+ decrease after benzamide treatment that persisted in KA lesioned striata could be related to AMPA receptor activation. Taken together with the increase of PARP activity at 24 h [10], these data are consistent with a causative role of PARP overactivationdependent energy depletion during the early development of the KA lesion. At later times, PARP-mediated mechanisms other than energy depletion may be involved in the observed disappearance of the lesion and the recovery of energy levels. In a previous study [10], striatal PARP activity and protein levels were found to be markedly increased at 1 week following intrastriatal KA injection. PARP activity and protein increase during proliferation and differentiation of astrocytes in culture [4] and systemic administration of KA induces astrogliosis [1]. In the present study GFAP-positive cells were detected from the earliest time points of KA toxicity, and increased in number and in staining intensity over time. At ten days, the presence of reactive astroglia at the site of the lesion might suggest a restoration of energy levels by these cells in support of the neurons surrounding the glial scar. Further studies are needed to investigate if the recovery
C. Cosi et al. / Brain Research 996 (2004) 1–8
of energy metabolism is due to astroglial cells and whether this is associated with functional recovery. Since PARP-1 knock-out mice are highly resistant to NMDA-induced neurotoxicity, we investigated if the NMDA-induced lesion in wild type mice could be prevented by treatment with benzamide, and if the eventual protective effects could be correlated with preservation of energy metabolism. NMDA, at the same dose as used by Mandir et al. [21], induced a lesion whose volume was reduced by treatment with benzamide, as measured at 48 h. However, NMDA had little or no effect on ATP and NAD+, at 1 h or at 48 h, in apparent contrast with the finding of Mandir et al. [21], where an increase in poly(ADP-ribose) was found during the first 24 h following NMDA injection in wild type mice. Nevertheless, NMDA produced a lesion volume approximately 50% larger than that induced by KA, without any expected proportional decrease in energy metabolites at 48 h. It cannot be ruled out that NMDA could cause energy decrease at times other then those observed in the present study (1 and 48 h), as suggested by a 7% decrease in ATP observed after NMDA injection. Also, the time course of energy recovery (possibly by astrogliosis) might occur faster in NMDA toxicity than in KA toxicity, which could explain the lack of energy depletion at 48 h after NMDA administration. On the other hand, these findings can also suggest that other PARP-mediated mechanisms independent of energy depletion and prevented (at least in part) by inhibition of PARP catalytic activity might be prevalent in NMDAinduced neurotoxicity in vivo. In fact, the increase in poly(ADP-ribose) [21] could have been due to an increased expression of PARP protein that is one of the major acceptors of the polymer of ADP-ribose. Regulation of gene expression by PARP through protein –protein interactions with transcription factors such as NF-kappaB, might represent an important mechanism of inflammation-mediated secondary damage in excitotoxicity [18] and in particular in NMDA-induced toxicity [26]. Moreover, this property of PARP to regulate gene expression, by forming multiprotein complexes with transcription factors independent of [19] or dependent on [5] auto-ADP-ribosylation of PARP, might also explain the strong resistance of PARP-1 knock-out mice to cerebral ischemia [13] and NMDA excitotoxicity [21]. Although it is currently difficult to discriminate between AMPA and KA receptors for lack of selective agonists and antagonists, studies utilizing mice deficient in specific KA receptor subunits (GluR5 and GluR6) have resolved the existence of AMPA and KA receptor entities (see Ref. [15] for a review) In our study, AMPA administered at the same dose used by Mandir et al. [21] induced a lesion which was not prevented by benzamide, in agreement with the lack of protection by PARP-1 / phenotype against AMPA toxicity. Early treatment with benzamide did not prevent AMPAinduced NAD+ and ATP depletions at 48 h suggesting that PARP activation is not implicated in the control of energy metabolism in the initiation of AMPA toxicity. Changes in NAD+ and ATP could be due to other mechanisms unrelated
7
to PARP and insensitive to benzamide, e.g. impairment of mitochondrial respiration. On the other hand, PARP is not activated in the first 24 h following AMPA administration in wild type mice [21], consistent with a lack of involvement of PARP in AMPA toxicity, at least during the first 24 h. Alternatively, since PARP inhibition by benzamide covers only the early phases of neurodegenerative process that is quantified at 48 h, it cannot be ruled out that differences in the time-course of neurotoxicity for KA and AMPA may account for the lack of effect of benzamide on AMPA neurotoxicity. While further studies are needed to investigate the different PARP-mediated mechanisms induced by overactivation of glutamate receptor subtypes, these results indicate that PARP overactivation and energy depletion could be responsible in part for the cellular demise during the development of the lesion induced by KA, confirm that PARP is involved in NMDA but not AMPA toxicities, and suggest the existence of differences between KA and AMPA toxicities. Furthermore, the results here obtained with benzamide are similar to those reported for the PARP-1 / phenotype, suggesting that neuroprotective action of benzamide is mediated by selective inhibition of PARP-1. This study provides further evidence supporting PARP as a novel target for new drug treatments against neurodegenerative disorders.
Acknowledgements The authors gratefully acknowledge the excellent technical assistance of N. Leduc, A.-M. Ormie`re, J. Floutard and O. Marteau.
References [1] C.A. Altar, M. Baudry, Systemic injection of kainic acid: gliosis in the olfactory and limbic brain regions quantified with [3H] PK 11195 binding autoradiography, Exp. Neurobiol. 109 (1990) 333 – 341. [2] C. Ayata, G. Ayata, H. Hara, R.T. Matthews, M.F. Beal, R.J. Ferrante, M. Endres, A. Kim, R.H. Christie, C. Waeber, P.L. Huang, B.T. Hyman, M.A. Moskowitz, Mechanisms of reduced striatal NMDA excitotoxicity in type I nitric oxide synthase knock-out mice, J. Neurosci. 17 (1997) 6908 – 6917. [3] N.A. Berger, Poly(ADP-ribose) in the cellular response to DNA damage, Radiat. Res. 101 (1985) 4 – 15. [4] G.M. Chambert, P.C. Niedergang, F. Hog, M. Partisani, P. Mandel, Poly(ADPR) polymerase expression and activity during proliferation and differentiation of rat astrocyte and neuronal cultures, Biochim. Biophys. Acta 1136 (1992) 196 – 202. [5] W.J. Chang, R. Alvarez-Gonzales, The sequence-specific DNA binding of NF-kappa B is reversibly regulated by the automodification reaction of poly(ADP-ribose) polymerase 1, J. Biol. Chem. 276 (2001) 47664 – 47670. [6] C. Cosi, M. Marien, Decreases in mouse brain NAD+ and ATP induced by 1-methyl-4-phenyl-1, 2,3,6-tetrahydropyridine (MPTP): prevention by the poly(ADP-ribose) polymerase inhibitor, benzamide, Brain Res. 809 (1998) 58 – 67.
8
C. Cosi et al. / Brain Research 996 (2004) 1–8
[7] C. Cosi, H. Suzuki, D. Milani, L. Facci, M. Menegazzi, G. Vantini, Y. Kanai, S.D. Skaper, Poly(ADP-ribose) polymerase: early involvement in glutamate-induced neurotoxicity in cultured cerebellar granule cells, J. Neurosci. Res. 39 (1994) 38 – 46. [8] C. Cosi, P. Chopin, M. Marien, Benzamide, an inhibitor of poly (ADP-ribose) polymerase, attenuates methamphetamine-induced dopamine neurotoxicity in the C57Bl/6N mouse, Brain Res. 735 (1996) 343 – 348. [9] C. Cosi, F. Colpaert, W. Koek, A. Degryse, M. Marien, Poly(ADPribose) polymerase inhibitors protect against MPTP-induced depletions of striatal dopamine and cortical noradrenaline in C57B1/6 mice, Brain Res. 729 (1996) 264 – 269. [10] C. Cosi, E. Cavalieri, A. Carcereri de Prati, M. Marien, H. Suzuki, Effects of kainic acid lesioning on poly(ADP-ribose)polymerase (PARP) activity in the rat striatum in vivo, Amino Acids 19 (2000) 229 – 237. [11] J.T. Coyle, P. Puttfarcken, Oxidative stress, glutamate, and neurodegenerative disorders, Science 262 (1993) 689 – 695. [12] G. De Murcia, V. Schreiber, M. Molinete, B. Saulier, O. Poch, M. Masson, C. Niedergang, J. Menissier de Murcia, Structure and function of poly(ADP-ribose) polymerase, Mol. Cell. Biochem. 138 (1994) 15 – 24. [13] M.J. Eliasson, K. Sampei, A.S. Mandir, P.D. Hurn, R.J. Traystman, J. Bao, A. Pieper, Z.Q Wang, T.M. Dawson, S.H. Snyder, V.L. Dawson, Poly(ADP-ribose) polymerase gene disruption renders mice resistant to cerebral ischemia, Nat. Med. 3 (1997) 1089 – 1095. [14] M. Endres, Z.Q. Wang, S. Namura, C. Waeber, M.A. Moskowitz, Ischemic brain injury is mediated by the activation of poly(ADPribose)polymerase, J. Cereb. Blood Flow Metab. 17 (1997) 1143 – 1151. [15] A.A. Farooqui, Y.W. Ong, X.-R. Lu, B. Halliwell, A.L. Horrocks, Neurochemical consequences of kainate-induced toxicity in brain: involvement of arachidonic acid relase and prevention of toxicity by phospholipase A2 inhibitors, Brain Res. Rev. 38 (2001) 61 – 78. [16] K.B.J. Franklin, G. Paxinos, The Mouse Brain in Streretaxic Coordinates, Academic Press, San Diego, CA, 1997.
[17] C.H. Ha, S.H. Snyder, Poly(ADP-ribose) polymerase-1 in the nervous system, Neurobiol. Dis. 7 (2000) 225 – 239. [18] C.H. Ha, L. Hester, S.H. Snyder, Poly(ADP-ribose)polymerase-1 dependence of stress-induced transcription factors and associated gene expression in glia, Proc. Natl. Acad. Sci. 99 (2002) 3270 – 3275. [19] P.O. Hassa, M.O. Hottiger, A role of poly (ADP-ribose)polymerase in NF-kappaB transcriptional activation, Biol. Chem. 380 (1999) 953 – 959. [20] A.S. Mandir, S. Przedborski, V. Jackson-Lewis, Z.Q. Wang, C.M. Simbulan-Rosental, B.E. Hoffman, D.B. Guastella, V.L. Dawson, T.M. Dawson, Poly(ADP-ribose)polymerase activation mediates 1methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-induced parkinsonism, Proc. Natl. Acad. Sci. 96 (1999) 5774 – 5779. [21] A.S. Mandir, M.F. Poitras, A.R. Berliner, W.J. Herring, D.B. Guastella, A. Feldman, G.G. Poirier, Z.Q Wang, T.M. Dawson, V.L. Dawson, NMDA but not non-NMDA excitotoxicity is mediated by Poly(ADP-ribose)polymerase, J. Neurosci. 20 (2000) 8005 – 8011. [22] K. Plaschke, J. Kopitz, M.A. Weigand, E. Martin, H.J. Bardenheuer, The neuroprotective effect of cerebral poly(ADP-ribose)polymerase inhibition in a rat model of global ischemia, Neurosci. Lett. 284 (2000) 109 – 112. [23] K.C. Retz, J.T. Coyle, Kainic acid lesion of mouse striatum: effects on energy metabolites, Life Sci. 27 (1980) 2495 – 2500. [24] P. Sheng, C. Cerruti, J.L. Cadet, Methamphetamine (METH) causes reactive gliosis in vitro: attenuation by the ADP-ribosylation (ADPR) inhibitor, benzamide, Life Sci. 55 (1994) L51 – L54. [25] K. Takahashi, A.A. Pieper, S.E. Croul, J. Zhang, S.H. Snyder, J.H. Greenberg, Post-treatment with an inhibitor of poly(ADP-ribose) polymerase attenuates cerebral damage in focal ischemia, Brain Res. 829 (1999) 46 – 54. [26] O. Ulrich, A. Diestel, I.Y. Eyu¨poglu, R. Nitsch, Regulation of microglia expression of integrins by poly(ADP-ribose) polymerase, Nat. Cell Biol. 3 (2001) 1035 – 1042. [27] J. Zhang, V.L. Dawson, T.M. Dawson, S.H. Snyder, Nitric oxide activation of poly(ADP-ribose) synthetase in neurotoxicity, Science 263 (1994) 687 – 689.