Advertisement
Research ArticleNEUROPHARMACOLOGY

Protection and Reversal of Excitotoxic Neuronal Damage by Glucagon-Like Peptide-1 and Exendin-4

TracyAnn Perry, Norman J. Haughey, Mark P. Mattson, Josephine M. Egan and Nigel H. Greig
Journal of Pharmacology and Experimental Therapeutics September 2002, 302 (3) 881-888; DOI: https://doi.org/10.1124/jpet.102.037481

Abstract

Glucagon-like peptide-1 (7–36)-amide (GLP-1) is an endogenous insulinotropic peptide that is secreted from the L cells of the gastrointestinal tract in response to food. It has potent effects on glucose-dependent insulin secretion, insulin gene expression, and pancreatic islet cell formation. In type 2 diabetes, GLP-1, by continuous infusion, can normalize blood glucose and is presently being tested in clinical trials as a therapy for this disease. More recently, GLP-1 has been found to have central nervous system (CNS) effects and to stimulate neurite outgrowth in cultured cells. We now report that GLP-1, and its longer-acting analog exendin-4, can completely protect cultured rat hippocampal neurons against glutamate-induced apoptosis. Extrapolating these effects to a well defined rodent model of neurodegeneration, GLP-1 and exendin-4 greatly reduced ibotenic acid-induced depletion of choline acetyltransferase immunoreactivity in basal forebrain cholinergic neurons. These findings identify a novel neuroprotective/neurotrophic function of GLP-1 and suggest that such peptides may have potential for halting or reversing neurodegenerative processes in CNS disorders, such as Alzheimer's disease, and in neuropathies associated with type 2 diabetes mellitus.

Although GLP-1 is produced by cells in the intestines and was discovered because of its effects on glucose metabolism (Doyle and Egan, 2001), recent studies have shown that GLP-1 can decrease feeding by acting on specific receptors in the brain. When injected intracerebroventricularly, GLP-1 dramatically reduces food and water intake (Gunn et al., 1996; Tang-Christensen et al., 1996; Turton et al., 1996; Wang et al., 1998) and body weight (Donahey et al., 1998;Meeran et al., 1999) in rats. The GLP-1 receptor antagonist exendin (9–39) inhibits the effects of GLP-1 on food and water intake, suggesting that GLP-1 receptors are involved in satiety (Turton et al., 1996). However, the effects of GLP-1 on feeding are not sustained, and mice lacking GLP-1 receptors are lean, eat normally, and do not develop obesity either with aging or after several months of high-fat intake (Scrocchi et al., 1996; Scrocchi and Drucker, 1998). Several studies have demonstrated GLP-1 receptor expression in both the rodent (Goke et al., 1995; Shughrue et al., 1996) and human (Wei and Mojsov, 1995; Satoh et al., 2000) brain. GLP-1 receptors appear to be distributed primarily in the hypothalamus, thalamus, brainstem, lateral septum, subfornical organ, and the area postrema. However, specific binding sites for GLP-1 have also been detected in neurons in the caudate-putamen, cerebral cortex, hippocampus, and cerebellum (Campos et al., 1994; Calvo et al., 1995; Goke et al., 1995). It remains to be established whether GLP-1 is produced by neural cells, but it has been shown that GLP-1 present in the bloodstream can enter the brain (Orskov et al., 1996).

We have recently shown that GLP-1 receptor activation induces neurite outgrowth in PC12 cells and SK-N-SH human neuroblastoma cells by a mechanism involving the second messenger cyclic AMP (Perry et al., 2002). Signals that stimulate cyclic AMP production can protect neurons against death in various paradigms. For example, corticotropin-releasing hormone and urocortin can protect cultured hippocampal neurons against death induced by glutamate and amyloid β-peptide (Pedersen et al., 2001, 2002).

In the present study, we have further examined the neurotrophic properties of GLP-1 and its longer-acting analog exendin-4 (Greig et al., 1999; see Perry et al., 2002 for amino acid sequences of GLP-1 and exendin-4) in cultured hippocampal neurons and in a well established rodent model of neurodegeneration. Exendin-4 has been shown to bind to the known GLP-1 receptor in pancreatic β cells (Goke et al., 1993;Thorens et al., 1993) and has several advantages over GLP-1: it has a higher potency than GLP-1, its half-life is approximately 120 min, and it maintains higher plasma levels of insulin over a longer time duration than GLP-1 (Ryan et al., 1998; Greig et al., 1999; Egan et al., 2002). In line with our demonstration that rat hippocampal neurons express functional GLP-1 receptors, we have tested the hypothesis that GLP-1 and exendin-4 can protect against glutamate-induced apoptosis and restore cholinergic marker activity in adult rats following nonselective excitotoxic basal forebrain damage. For several decades, lesion models of Alzheimer's disease in the rat have encompassed one aspect of the human condition, that is, the loss of cholinergic neurons found in the medial septum and nucleus basalis of Meynert (or basal nucleus in rodents). Such models have relied on immunohistochemical correlates of cholinergic phenotype to identify lesion- and treatment-induced changes. Although treatment with growth factors is well documented to protect against lesion-induced cholinergic cell death (Haroutunian et al., 1986; Mandel et al., 1989), recent observations suggest that such neurons may simply down-regulate their phenotype, rather than die, and that treatment with such growth factors may rescue cholinergic neurons (Haas and Frotscher, 1998; Weis et al., 2001). We propose that GLP-1 and exendin-4 possess neurotrophic capabilities and, as such, offer the possibility for restoring the cholinergic phenotype following partial excitotoxic damage within the basal nucleus of the rat. Although originally developed for their insulinotropic properties, GLP-1 and exendin-4 may have potential for reversing or halting neurodegenerative processes observed in central nervous system disorders such as severe epileptic seizures and Alzheimer's disease and in neuropathies associated with type 2 diabetes mellitus.

Materials and Methods

Cell Culture.

Hippocampal neuronal cultures were prepared from 18-day-old embryonic Sprague-Dawley rats using methods similar to those described previously (Mattson et al., 1995). Briefly, cells were dissociated by mild trypsinization and trituration and plated in minimal essential medium containing 10% fetal bovine serum and 1% antibiotic solution (104 U/ml penicillin G, 10 mg/ml streptomycin, and 25 μg/ml amphotericin B; Sigma-Aldrich, St. Louis, MO). Hippocampal neurons were plated at a density of 100,000 cells/ml on 25-mm-diameter poly-d-lysine-coated, glass coverslips. Three hours after plating, the medium was replaced with serum-free Neurobasal medium containing 1% B-27 supplement (Invitrogen, Carlsbad, CA). Immunofluorescence staining for mitogen-activated protein-2 (neurons) and glial fibrillary acidic protein (GFAP; astrocytes) showed that more than 98% of the cells were neurons, and the remainder were predominantly astrocytes. Cultures were used within 7 to 10 days of plating.

Competitive Binding of GLP-1 to GLP-1 Receptor in Hippocampal Neurons.

Binding studies were performed as described byMontrose-Rafizadeh et al. (1997b). Duplicate hippocampal neuronal cultures were washed in 0.5 ml of binding buffer and subsequently incubated in 0.5 ml of buffer containing 2% bovine serum albumin, 17 mg/l diprotin A (Bachem California, Torrance, CA), 10 mM glucose, 0.001 to 1000 nM GLP-1, and 30,000 cpm of 125I-GLP-1 (Amersham Biosciences UK, Ltd., Little Chalfont, Buckinghamshire, UK), overnight at 4°C. At the end of the incubation, the supernatant was discarded, and the cells were washed three times in ice-cold phosphate-buffered saline (PBS) and incubated at room temperature with 0.5 ml of 0.5 M NaOH and 0.1% SDS for 10 min. Radioactivity in cell lysates was measured in an Apec-Series gamma counter (ICN Biomedicals, Inc., Costa Mesa, CA). Specific binding was determined as the total binding minus the radioactivity associated with cells incubated in the presence of a large excess of unlabeled GLP-1 (1 μM). The GLP-1 concentration associated with 50% binding, EC50, was determined by logit plot analysis.

Intracellular cAMP Determination.

To demonstrate the presence of functional GLP-1 receptors, cyclic AMP was measured according to the method of Montrose-Rafizadeh et al. (1997a). Triplicate hippocampal neuronal cell cultures were treated with 10 nM GLP-1 and harvested at 5-min intervals after the onset of drug treatment for a total period of 30 min. Cells harvested at the start of drug treatment (0 min) were used for baseline levels of cAMP.

Cell Survival.

The fluorescent DNA binding dye Hoescht 33342 was used to measure apoptotic cell death. Neurons were incubated in Locke's buffer with GLP-1 (10 nM) or exendin 4 (0.3 μM) alone, and in combination with glutamate (10 μM) for 16 h. The concentration of GLP-1 used was based on the EC50value derived from the binding experiment, which was demonstrated to stimulate the release of cAMP, and which induced differentiation without causing cell death in our previous neuronal cell studies (Perry et al., 2002). Exendin-4 is well documented to bind at the GLP-1 receptor (Goke et al., 1993, 1995), and we have based our dose on that used previously in cultured neuronal cells (Perry et al., 2002). Cells were fixed in a solution of 4% paraformaldehyde in PBS, and membranes were permeabilized with 0.2% Triton X-100. Following incubation with Hoechst 33342 (1 μM) for 30 min, nuclei were visualized under epifluorescence illumination (340-nm excitation, 510-nm barrier filter) using a 40× oil immersion objective. Approximately 200 cells were counted in at least three separate dishes for each treatment condition, and experiments were repeated at least twice. Cells were considered apoptotic if nuclear DNA was fragmented or condensed, whereas cells with nuclear DNA of a more diffuse and uniform distribution were considered viable. At the time of counting, the investigator was unaware of the identity of the treatment groups. The percentage of cells with condensed or fragmented nuclei was determined in each culture.

Animals and Surgical Procedures.

The study was undertaken using a total of 35 adult male Fischer-344 rats weighing approximately 300 g each. The animals were housed under controlled light/dark and temperature conditions with food and water available ad libitum. Rats were anesthetized with ketamine (90 mg/kg) and acepromazine (0.91 mg/kg). Stereotaxic surgery was carried out as described (Perry et al., 2001). Ibotenic acid dissolved in 0.1 M PBS was infused unilaterally into the left lateral branch of the forebrain bundle, which was referred to as the basal nucleus by Paxinos and Watson (1998), at 10 μg/μl (0.5 μl, two sites). Prior pilot examination of the efficacy of this particular batch of toxin demonstrated that this dose produced a 60% loss of choline acetyltransferase (ChAT)-positive immunoreactivity in the basal forebrain, with a comparable loss of projections to the cortex. Although ibotenic acid is well documented to produce nonselective damage, in this particular series of experiments we elected to use cholinergic neurons as a marker to quantify the degree of degeneration. The rationale for the experimental parameters included the following. First, by sparing a proportion of cell bodies, we ensured that the infused peptide would have a residual population on which to exert possible trophic effects. Second, unilateral rather than bilateral damage was induced such that contralateral comparisons could be made within individual animals rather than simply between groups. A second series of animals receiving infusions of vehicle were used as controls. Each infusion was performed over 2.5 min, and an additional 2.5 min were allowed for diffusion before the cannula was retracted. After 2 weeks, animals were reanesthetized and stereotaxically implanted with an intracerebroventricular cannula into the right lateral ventricle (anterior-posterior −0.8 mm, lateral to bregma +1.4 mm, ventral to dura −4.0 mm).

The cannulae were attached via a catheter to an osmotic minipump (Alza Pharmaceuticals, Mountain View, CA). Based on previous in vivo dose-response data from our laboratory, pumps were filled with 2 × 10−8 M GLP-1, 2 × 10−9 M exendin-4, or vehicle [artificial cerebrospinal fluid (aCSF)]. Both peptides were diluted in vehicle. The pumps were set to deliver 0.25 μl/h over 14 days (total of 5.54 ng of GLP-1 or 0.7 ng of exendin-4). The brain infusion kits were assembled 5 to 6 h prior to implantation and placed in sterile saline at 37oC. The minipumps were inserted into a subcutaneous pocket between the shoulder blades, the wounds sutured, and the animals allowed to recover. Animals receiving infusions of GLP-1 or exendin-4 became modestly aphagic and adipsic as a result of the insulinotropic nature of the peptides, which resulted in a slight drop in body weight. This was recouped within 3 to 4 days with the administration of twice daily fluids (0.9% saline) and soft diet, and by the time of sacrifice there were no differences in body weight between the groups. On expiry of the minipumps (14 days), animals were terminally anesthetized with 0.1 mg/kg sodium pentobarbitone and transcardially perfused with 100 to 150 ml of PBS (pH 7.4), followed by 250 to 350 ml of 4% paraformaldehyde solution in PBS, at a constant pressure of 100 mm Hg over a period of 15 to 20 min. The brains were taken for immunocytochemical assessment and quantification of the lesion-induced damage and any resulting effects of peptide infusion on the cholinergic contingent of the basal forebrain.

Immunohistochemistry.

Adjacent coronal brain sections were taken at 40-μm thickness, through the lesion area, and processed free-floating for ChAT using the polyclonal goat anti-ChAT antibody at 1:100 dilution (Chemicon International Inc., Temecula, CA), and for GFAP using the polyclonal rabbit anti-GFAP antibody at 1:750 dilution (Chemicon). Visualization of positive immunoreactivity was carried out using an avidin-biotin/horseradish peroxidase protocol. In addition, one series of sections was stained for acetylcholinesterase activity, as a histochemical marker for cholinergic neurons of the basal forebrain using a modified method by Geula and Mesulam (1989). An additional series of sections was mounted onto gelatin-coated slides and stained with cresyl violet to visualize cell bodies.

Quantification and Statistical Analyses.

In the cell culture studies, the percentage of cells undergoing apoptosis as a result of each treatment condition were subjected to ANOVA using StatView statistical software (StatView, Cary, NC). Following significant main effects, a posteriori comparisons of treatment versus control were made using Tukey's honestly significant difference (HSD) test, using the pooled ANOVA error term and degrees of freedom.

For the in vivo studies, ChAT-positive immunoreactive cell bodies in the forebrain area were visualized under 100× magnification and manually counted on both sides. The raw counts were corrected with theAbercrombie (1946) formula for an estimate of the total number of cell bodies in the area. Characterization of the cell loss in the basal nucleus as a result of the ibotenic acid lesion was made by comparison of left (lesion side) relative to right (infusion side) counts, and the data presented as the percentage of the change. The animal names were coded such that cell counts were formally conducted blind to the experimental condition. Differences were subjected to analysis of variance and a posteriori comparisons using Tukey's HSD test as described above. Significance was accepted at p < 0.05 for all statistical analyses.

Results

GLP-1 and Exendin-4 Protect Cultured Hippocampal Neurons against Cell Death Induced by Glutamate.

Binding of125I-GLP-1 to cultured hippocampal neurons was displaced, concentration dependently, by unlabeled GLP-1 (Fig.1A). The concentration of GLP-1 required to displace 50% bound 125I-GLP-1 was determined by logit plot analysis and required a concentration of 14 nM GLP-1 (r = −0.999) in cultured hippocampal neurons.

Figure 1
Figure 1

Displacement of 125I-GLP-1 binding with cold GLP-1 (A), GLP-1-stimulated release of cAMP (B), and protection against glutamate-induced apoptosis (C) in cultured hippocampal neurons. 125I-GLP-1 binding to intact cultured hippocampal neurons was competed with various concentrations of GLP-1. The data are normalized to maximum values obtained in the presence of 1 μM GLP-1. Each data point represents the mean of two experimental values and is presented as the percentage of maximum binding in the absence of cold peptide. cAMP levels were assayed over 30 min of incubation with 10 nM GLP-1 (B). Vertical error bars represent ± standard error of the mean of three individual experimental values. Treatment with 10 nM GLP-1 or 0.3 μM exendin-4 completely protected against the apoptotic effects of 10 μM glutamate (C). Cultures were treated overnight, fixed with 4% paraformaldehyde, and stained with Hoechst 33342. The number of apoptotic nuclei was counted, and the values are presented as the pooled mean of six individual dishes per treatment condition. Vertical error bars represent ± standard error of the difference between the means. Significant difference from control: ★★, p < 0.01, and ★★★,p < 0.001.

Treatment of cultured hippocampal neurons with 10 nM GLP-1 evoked an increase in cAMP production (Fig 1B). Activation of the GLP-1 receptor has previously been shown to stimulate adenylyl cyclase, leading to an increase in intracellular cAMP. There was a maximal 2- to 3-fold increase in cAMP levels within 15 min of stimulation, which returned to near baseline within 30 min. One-way ANOVA demonstrated significant main effects (F = 9.45, df = 6,20;p < 0.001) of treatment on cAMP production. Subsequent multiple comparisons using Tukey's HSD test revealed significant increases in cAMP production after 10 (p < 0.01) and 15 (p < 0.001) min. These data demonstrate that primary hippocampal neurons express functional GLP-1 receptors, making them an appropriate in vitro system in which to study potential protective and trophic effects of these peptides.

In light of our prior demonstration that GLP-1 and exendin-4 stimulate neurite outgrowth in cultured neural tumor cell lines (Perry et al., 2002), we examined whether such peptides could protect against excitotoxic cell death. Primary hippocampal neurons were treated overnight with 10 μM glutamate. After fixation, the cells were stained with Hoechst 33342, and the number of apoptotic cells was counted. In cells cultured in medium alone, 23% of the neurons exhibited apoptotic nuclei. Glutamate treatment produced 73% apoptosis (Fig. 1C). Concurrent treatment with either 10 nM GLP-1 (24% apoptotic cells) or 0.3 μM exendin-4 (25% apoptotic cells) completely protected against cell death (Fig. 1B). Treatment with GLP-1 or exendin-4 alone did not produce any increase in the percentage of apoptotic cells (20 and 23%, respectively) beyond that of control levels. The values represent the pooled means of six individual experiments. One-way ANOVA demonstrated statistically significant differences in the extent of cell death between each insult (F = 35.31, df = 5,36; p < 0.001), and subsequent multiple comparison using Tukey's HSD test (Tukey statistic Tc = 14.91 and 18.165) revealed significant increases in the percentage of apoptotic cells following glutamate treatment (p < 0.01, compared with controls). Concurrent treatment of the cultures with GLP-1 or exendin-4 significantly protected against glutamate-induced cell death (bothp < 0.01, compared with glutamate alone). No significant differences existed between the concurrent glutamate/peptide cultures and controls, demonstrating complete protection of neurons against the effects of glutamate.

Chronic Treatment with GLP-1 or Exendin-4 Attenuates the Ibotenic Acid-Induced Cholinergic Marker Deficit in Adult Rats.

Choline acetyltransferase immunoreactivity was used as a marker for cholinergic neurons throughout the basal forebrain. The ChAT antibody stained numerous, large, multipolar neurons, with a similar size and distribution to the acetylcholinesterase-positive cells. The immunocytochemical staining had low background and provided a clear picture of the cell morphology (Fig. 2). Injection of ibotenic acid with subsequent infusion of vehicle resulted in a substantial (43%) loss of ChAT-immunoreactive neurons (Fig.3, bar 4) over an approximately 1-mm radius from the injection site in the left basal nucleus, which was within the realm of that predicted from the pilot study. The sham-operated control group receiving vehicle infusion showed an increase in the percentage of ChAT-positive cell bodies in the left basal nucleus relative to the right basal nucleus (Fig. 3, bar 1). Clearly, this was unexpected and will be addressed later in this section. Ibotenate-lesioned animals that received GLP-1 or exendin-4 infusions had a decreased loss of ChAT-immunoreactive cell bodies in the left basal nucleus relative to those lesioned animals that received vehicle infusion. More specifically, infusion of exendin-4 produced a decrease in the loss of ChAT-immunoreactive cell bodies in the left basal nucleus from 43%, as was apparent following vehicle infusion, to just 24% below that of the right basal nucleus (Fig. 3, bar 5). Furthermore, GLP-1 infusion resulted in more striking reversal effects, decreasing the loss of ChAT-positive immunoreactive cell bodies in the left basal nucleus to just 6% below that of the right basal nucleus (Fig. 3, bar 6). Standard ANOVA demonstrated an overall significant effect of treatment condition (F = 21.363, df = 5,28; p < 0.001). Multiple comparisons of peptide versus vehicle treatment (Tc = 14.14 and 19.71) revealed significant improvements in ChAT immunoreactivity in the left basal nucleus following infusion of exendin-4 (p < 0.05) and GLP-1 (p < 0.01) after an ibotenic acid lesion. Although infusion of GLP-1 following an ibotenic acid lesion decreased the ChAT-positive cell loss in the left basal nucleus to produce near-equal values with the right side, the overall percentage of difference was still significantly lower than that of the sham vehicle group (p < 0.001). This is likely due to the perceptible increase in ChAT immunoreactivity in the sham group receiving vehicle infusion (Fig. 3, bar 1).

Figure 2
Figure 2

Choline acetyltransferase and glial fibrillary acidic protein immunoreactivity in the basalis nucleus. ChAT-positive immunoreactivity, ipsilateral (A and C) and contralateral (B and D), to a partial ibotenic acid lesion is shown. Panels A and B, and C and D depict the left and right basal nucleus from individual animals that received vehicle infusion and GLP-1 infusion, respectively. ChAT-positive immunoreactivity in the ipsilateral basal nucleus in an animal that received vehicle infusion (A) was substantially lower than that in the ipsilateral basal nucleus in an animal that received GLP-1 infusion (C). GFAP immunoreactivity, a marker for reactive astrocytes produced in response to injury, demonstrated areas of positive immunoreactivity surrounding the site of cannula implantation and lining of the lateral ventricle in the vicinity of the site of infusion. Interestingly, infusion of GLP-1 produced a more elevated glial reaction in the basal nucleus on the infusion side (F) than that apparent as a result of the lesion (E) or after vehicle infusion.

Figure 3
Figure 3

Percentage of difference in the Abercrombie-corrected number of ChAT-immunoreactive cell bodies in the ipsilateral basal nucleus (lesion side) relative to the intact contralateral basal nucleus in sham and ibotenic acid animals receiving i.c.v. infusion of vehicle (aCSF), exendin-4, or GLP-1. Vertical error bars represent the standard error of the difference between the means. Significant difference from ibotenic acid vehicle group: ★, p< 0.05 and ★★, p < 0.01.

Separating the left and right ChAT-positive immunoreactive neuronal counts for the sham vehicle group (Table1) revealed a significant difference between left (586 ± 32) and right (478 ± 40) basal nuclei ChAT-positive cell counts (p < 0.01). Pressure effects from cannula implantation and treatment delivery may account for the apparent decrease in ChAT immunoreactivity in the right basal nucleus. These observations suggest that any disturbance of tissue integrity, however mild or nonspecific, can produce a functional disruption of ChAT immunoreactivity. Furthermore, such effects may account for the lower than anticipated percentage of loss in ChAT-positive immunoreactivity in the ibotenic acid group receiving vehicle infusion (Fig. 3, bar 4). To examine this further, rather than comparing “within” groups (i.e., left versus right), “between”-groups comparisons of left basal nucleus counts (F = 6.136, df = 5,28; p < 0.001; Table 1) for the sham aCSF and ibotenic acid aCSF groups (586 ± 32 and 260 ± 28, respectively) revealed a 56% loss in ChAT immunoreactivity, reflecting closely our pilot examination of ibotenic acid efficacy. This further supports the contention that nonspecific damage in the right basal nucleus produced decreases in the number of ChAT-immunoreactive cell bodies, affecting the overall percentage of loss when comparisons were made within individual experimental groups. In addition, there was no significant difference between groups when right basal nucleus ChAT-positive cell counts were analyzed separately (F = 0.512, df = 5,28; p > 0.05; Table 1), implying that such disruption of tissue integrity affected all experimental groups equally.

View this table:
Table 1

Abercrombie-corrected ChAT-positive cell counts in the basal nucleus

Discussion

The present results indicate that GLP-1 and exendin-4 may have novel therapeutic potential for the treatment of neurodegenerative disorders. We have demonstrated that GLP-1 and exendin-4 provide protection against cell death induced by glutamate neurotoxicity, which is mediated by calcium influx throughN-methyl-d-aspartate receptor channels in cultured hippocampal neurons (Mattson et al., 1995). Moreover, using a well established model of neurodegeneration in the rat, GLP-1 and exendin-4 demonstrated significant reversal of the cholinergic marker deficit in the basal forebrain induced by ibotenic acid.

GLP-1 and exendin-4 have previously been shown to possess neurite outgrowth-promoting properties in rat pheochromocytoma cells and in a human neuroblastoma cell line (Perry et al., 2002). We demonstrated that primary hippocampal neurons exhibit a robust increase in cAMP levels when exposed to a single dose of GLP-1, indicating the presence of functional GLP-1 receptors in those neurons. Similar to our demonstration in PC12 and SK-N-SH cells, cAMP is rapidly released after activation with a single dose of GLP-1, reaching maximal levels within 15 min. Cyclic AMP levels decrease back to baseline within 30 min, regardless of the continuous stimulation of the GLP-1 receptor. This “pulsatile” synthesis can be restored following a suitable “recovery phase”, which is a classic response to GLP-1 in β cells of the pancreas.

When cultured hippocampal neurons were subjected to excitotoxic insult, GLP-1 and exendin-4 were able to provide complete protection from the glutamate-induced cell death, as has been shown by other neurotrophic factors (Cheng et al., 1994; Mattson et al., 1995). These results suggest that such peptides may play a significant role in protecting hippocampal neurons against excitotoxic damage and potentially other types of brain injury. The dose range of the peptides used in these studies was focused around the EC50 value for GLP-1 derived from the binding study and the doses we have previously used to examine trophic properties in neuronal-like cells as well as those currently used in ongoing experiments related to pancreatic β cell function. Indeed, all doses of peptides used for the in vivo studies reflect those in current clinical investigations of the efficacy of GLP-1 and exendin-4 in patients with type 2 diabetes mellitus (Perfetti et al., 2000; Doyle and Egan, 2001; Larsen et al., 2001), together with the relative potencies of the peptides for the GLP-1 receptor.

Following translation of the in vitro effects to a well established model of neurodegeneration in the rat, we have shown complete amelioration of the ibotenic acid-induced cholinergic marker deficit following infusion of GLP-1. Using ChAT-positive immunoreactivity as the marker for cholinergic cell bodies within the basal forebrain, treatment with exendin-4 significantly decreased the loss of ChAT-positive cell bodies in the lesioned area compared with the vehicle-treated lesion group. The GLP-1-mediated effect was more pronounced than the exendin-4 effect, affording essentially complete recovery of the lesion deficit.

Despite the GLP-1-mediated restoration of ChAT-positive neurons in the lesioned basal forebrain, comparison with the sham-operated vehicle-treated group showed a significant difference. This appears to be due to the apparent increase in ChAT-positive immunoreactivity in the sham vehicle group (Fig. 3, in which the number of ChAT-positive neurons in the left basal nucleus is presented as a percentage of the number in the right basal nucleus). Analysis of separate left and right ChAT-immunoreactive cell body counts in the sham vehicle group suggested that there was a deficit/down-regulation of ChAT immunoreactivity in the right basal nucleus. Clearly, as the control group, the expectation would be to find equal numbers of cell bodies on the left and right sides. This suggests that disturbance of tissue integrity (as a result of cannula implantation and/or treatment delivery), however mild or nonspecific, may produce a disruption of certain markers within particular cell populations. In support of this,Torres and colleagues (1994) have demonstrated that injections of saline into the septum or diagonal band induce a loss of parvalbumin immunoreactivity in the vicinity of the injection, which was as extensive as the loss following a lesion. The lack of significant effects between groups when right basal nucleus ChAT-positive counts were analyzed separately demonstrates that all groups were affected equally and does not detract from the clarity of the GLP-1- and exendin-4-mediated amelioration of cholinergic neuron atrophy. In this series of experiments, the unilateral lesion deficit with contralateral i.c.v. cannula implantation allowed each animal to act as its own control (lesion side versus intact side). However, with the overall increase in the percentage of ChAT-positive cell bodies in the sham aCSF group (Fig. 3), it was necessary to analyze left and right basal nucleus cell body counts separately and confirm that this was not a spurious trophic phenomenon of aCSF infusion (such effects were not apparent in the sham GLP-1 or sham exendin-4 groups) but more likely due to disruption of tissue integrity common across all groups, as was the case. Future studies will address this issue with bilateral rather than unilateral lesion and cannula implantation parameters such that more reliable “between”-groups comparisons can be made. Regardless, this should not detract from the clarity of our peptide data. We have demonstrated complete recovery of cholinergic marker activity following an excitotoxic nonselective ibotenic acid lesion, which we can attribute to the trophic influence in the basal nucleus of GLP-1 treatment.

Taken together, the ability of GLP-1 to protect against glutamate-induced toxicity in cultured cells and to restore cholinergic marker function following excitotoxic-induced damage within the basal forebrain demonstrates that the beneficial effects of these peptides are likely not specific for excitotoxicity or indeed cholinergic neuronal death but may represent a trophic property for all neuronal GLP-1 receptor-expressing cells. Further studies are ongoing to establish whether other neuronal populations are affected by GLP-1.

The mechanisms by which GLP-1 and its naturally occurring analog, exendin-4, are able to exert beneficial effects within the denervated basal forebrain remain unclear. Previous studies have documented the presence and cellular localization of GLP-1 binding sites in the rodent (Kanse et al., 1988; Goke et al., 1995; Shughrue et al., 1996) and human (Satoh et al., 2000) brain. However, such studies have been severely hampered by the lack of a specific antibody against the GLP-1 receptor. As an alternative to direct immunocytochemical labeling, a number of experimental approaches have been used, which involve indirect techniques for the identification of GLP-1 receptors in the brain. First, Northern blot analysis of polymerase chain reaction-amplified GLP-1 receptor mRNA transcripts demonstrated high levels of the GLP-1 receptor mRNA in brainstem and cerebellum, whereas cortex and hypothalamus expressed lower levels (Campos et al., 1994). Second, binding studies using 125I-GLP-1 (Goke et al., 1995), further demonstrated the presence of the GLP-1 receptor in the subfornical organ, arcuate nucleus of the hypothalamus, interpeduncular nucleus, parategmental nucleus, inferior olive, and the nucleus of the solitary tract. Third, following cloning of the GLP-1 receptor (Thorens et al., 1993), in situ hybridization studies demonstrated mRNA expression in the basal portion of the frontal cortex, substantia innominata, nucleus accumbens, nucleus basalis magnocellularis, ventral pallidum, lateral septum (and, to a lesser degree, the medial septum), diagonal band of Broca, medial/central and basolateral nuclei of the amygdala, and the CA2 to CA3 region of the hippocampus (Merchenthaler et al., 1999) and, more recently, on glial cells following mechanical brain injury (Chowen et al., 1999). Good correlations have been found between the in situ studies and previous immunocytochemical observations on the location of GLP-1-immunoreactive cells (Jin et al., 1988). Although the highest density of GLP-1 receptors seems to be in circumventricular areas, where generally large numbers of peptide receptors are located, there are moderate to low densities elsewhere in the brain. In particular, Merchenthaler et al. (1999) specify GLP-1 receptor mRNA-expressing cells in basal forebrain areas and the CA2 to CA3 region of the hippocampus.

In summary, from our demonstration that GLP-1 binds and stimulates cAMP production in primary hippocampal neurons, protects against apoptotic cell death in the same hippocampal cells, and restores cholinergic marker function following excitotoxic damage in the basal forebrain, it seems highly probable that GLP-1 receptor expression has a cholinergic contingent. Although we have demonstrated the functional presence of GLP-1 receptors on cultured hippocampal neurons (intracellular cAMP increased after GLP-1 stimulation), such cells are embryonic, and we cannot conclude that they perfectly represent the phenotype of cholinergic cell bodies in the basal nucleus of adult rats. Without a selective antibody to the GLP-1 receptor (the currently available polyclonal antibody produced extensive nonspecific staining in these studies and was considered not selective for the GLP-1 receptor), it remains difficult to establish the precise location of the GLP-1 receptor or the mechanism by which GLP-1 and exendin-4 exert their beneficial effects. Nevertheless, our findings suggest that these peptides may offer the possibility for rescue of damaged neurons in either the central or peripheral nervous systems associated with neurodegeneration. Our ongoing studies are focused on elucidating the functional effects and mechanisms of action of GLP-1 and related analogs within these systems.

Footnotes

  • This work was supported by the Intramural National Institute on Aging.

  • DOI: 10.1124/jpet.102.037481

  • Abbreviations:
    GLP-1
    glucagon-like peptide-1 (7–36)-amide
    PBS
    phosphate-buffered saline
    ChAT
    choline acetyltransferase
    GFAP
    glial fibrillary acidic protein
    ANOVA
    analysis of variance
    HSD
    honestly significant difference
    aCSF
    artificial cerebrospinal fluid
    • Received April 11, 2002.
    • Accepted May 9, 2002.

References

  1. ↵
    1. Abercrombie M
    (1946) Estimation of nuclear population from microtome sections. Anat Rec 94:239–247.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Calvo JC,
    2. Yusta B,
    3. Mora F,
    4. Blazquez E
    (1995) Structural characterization by affinity cross-linking of glucagon-like peptide-1(7–36)amide receptor in rat brain. J Neurochem 64:299–306.
    OpenUrlPubMed
  3. ↵
    1. Campos RV,
    2. Lee YC,
    3. Drucker DJ
    (1994) Divergent tissue-specific and developmental expression of receptors for glucagon and glucagon-like peptide-1 in the mouse. Endocrinology 134:2156–2164.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Cheng B,
    2. Goodman Y,
    3. Begley JG,
    4. Mattson MP
    (1994) Neurotrophin-4/5 protects hippocampal and cortical neurons against energy deprivation- and excitatory amino acid-induced injury. Brain Res 650:331–335.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Chowen JA,
    2. de Fonseca FR,
    3. Alvarez E,
    4. Navarro M,
    5. Garcia-Segura LM,
    6. Blazquez E
    (1999) Increased glucagon-like peptide-1 receptor expression in glia after mechanical lesion of the rat brain. Neuropeptides 33:212–215.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Donahey JC,
    2. van Dijk G,
    3. Woods SC,
    4. Seeley RJ
    (1998) Intraventricular GLP-1 reduces short- but not long-term food intake or body weight in lean and obese rats. Brain Res 779:75–83.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Doyle ME,
    2. Egan JM
    (2001) Glucagon-like peptide-1. Recent Prog Horm Res 56:377–399.
    OpenUrlCrossRefPubMed
  8. ↵
    1. Egan JM,
    2. Clocquet AR,
    3. Elahi D
    (2002) The insulinotropic effect of acute Exendin-4 administration to humans: comparison of nondiabetic state to type 2 diabetes. J Clin Endocrinol Metab 87:1282–1290.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Geula C,
    2. Mesulam MM
    (1989) Cortical cholinergic fibers in aging and Alzheimer's disease: a morphometric study. Neuroscience 33:469–481.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Goke R,
    2. Fehmann HC,
    3. Linn T,
    4. Schmidt H,
    5. Krause M,
    6. Eng J,
    7. Goke B
    (1993) Exendin-4 is a high potency agonist and truncated exendin-(9–39)-amide an antagonist at the glucagon-like peptide 1-(7–36)-amide receptor of insulin-secreting beta-cells. J Biol Chem 268:19650–19655.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Goke R,
    2. Larsen PJ,
    3. Mikkelsen JD,
    4. Sheikh SP
    (1995) Distribution of GLP-1 binding sites in the rat brain: evidence that exendin-4 is a ligand of brain GLP-1 binding sites. Eur J Neurosci 7:2294–2300.
    OpenUrlCrossRefPubMed
  12. ↵
    1. Greig NH,
    2. Holloway HW,
    3. De Ore KA,
    4. Jani D,
    5. Wang Y,
    6. Zhou J,
    7. Garant MJ,
    8. Egan JM
    (1999) Once daily injection of exendin-4 to diabetic mice achieves long-term beneficial effects on blood glucose concentrations. Diabetologia 42:45–50.
  13. ↵
    1. Gunn I,
    2. O'Shea D,
    3. Turton MD,
    4. Beak SA,
    5. Bloom SR
    (1996) Central glucagon-like peptide-I in the control of feeding. Biochem Soc Trans 24:581–584.
    OpenUrlFREE Full Text
  14. ↵
    1. Haas CA,
    2. Frotscher M
    (1998) Role of NGF in axotomy-induced c-Jun expression in medial septal cholinergic neurons. Int J Dev Neurosci 16:691–703.
    OpenUrlPubMed
  15. ↵
    1. Haroutunian V,
    2. Kanof PD,
    3. Davis KL
    (1986) Partial reversal of lesion-induced deficits in cortical cholinergic markers by nerve growth factor. Brain Res 386:397–399.
    OpenUrlPubMed
  16. ↵
    1. Jin SL,
    2. Han VK,
    3. Simmons JG,
    4. Towle AC,
    5. Lauder JM,
    6. Lund PK
    (1988) Distribution of glucagonlike peptide I (GLP-I), glucagon and glicentin in the rat brain: an immunocytochemical study. J Comp Neurol 271:519–532.
    OpenUrlCrossRefPubMed
  17. ↵
    1. Kanse SM,
    2. Kreymann B,
    3. Ghatei MA,
    4. Bloom SR
    (1988) Identification and characterization of glucagon-like peptide-1 7–36 amide-binding sites in the rat brain and lung. FEBS Lett 241:209–212.
    OpenUrlCrossRefPubMed
  18. ↵
    1. Larsen J,
    2. Hylleberg B,
    3. Ng K,
    4. Damsbo P
    (2001) Glucagon-like peptide-1 infusion must be maintained for 24 h/day to obtain acceptable glycemia in type 2 diabetic patients who are poorly controlled on sulphonylurea treatment. Diabetes Care 24:1416–1421.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    1. Mandel RJ,
    2. Gage FH,
    3. Thal LJ
    (1989) Spatial learning in rats: correlation with cortical choline acetyltransferase and improvement with NGF following NBM damage. Exp Neurol 104:208–217.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Mattson MP,
    2. Lovell MA,
    3. Furukawa K,
    4. Markesbery WR
    (1995) Neurotrophic factors attenuate glutamate-induced accumulation of peroxides, elevation of intracellular Ca2+ concentration and neurotoxicity and increase antioxidant enzyme activities in hippocampal neurons. J Neurochem 65:1740–1751.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Meeran K,
    2. O'Shea D,
    3. Edwards CM,
    4. Turton MD,
    5. Heath MM,
    6. Gunn I,
    7. Abusnana S,
    8. Rossi M,
    9. Small CJ,
    10. Goldstone AP,
    11. et al.
    (1999) Repeated intracerebroventricular administration of glucagon-like peptide-1-(7–36) amide or exendin-(9–39) alters body weight in the rat. Endocrinology 140:244–250.
    OpenUrlCrossRefPubMed
  22. ↵
    1. Merchenthaler I,
    2. Lane M,
    3. Shughrue P
    (1999) Distribution of pre-pro-glucagon and glucagon-like peptide-1 receptor messenger RNAs in the rat central nervous system. J Comp Neurol 403:261–280.
    OpenUrlCrossRefPubMed
  23. ↵
    1. Montrose-Rafizadeh C,
    2. Wang Y,
    3. Janczewski AM,
    4. Henderson TE,
    5. Egan JM
    (1997a) Overexpression of glucagon-like peptide-1 receptor in an insulin-secreting cell line enhances glucose responsiveness. Mol Cell Endocrinol 130:109–117.
    OpenUrlCrossRefPubMed
  24. ↵
    1. Montrose-Rafizadeh C,
    2. Yang H,
    3. Rodgers BD,
    4. Beday A,
    5. Pritchette LA,
    6. Eng J
    (1997b) High potency antagonists of the pancreatic glucagon-like peptide-1 receptor. J Biol Chem 272:21201–21206.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Orskov C,
    2. Poulsen SS,
    3. Moller M,
    4. Holst JJ
    (1996) Glucagon-like peptide I receptors in the subfornical organ and the area postrema are accessible to circulating glucagon-like peptide I. Diabetes 45:832–835.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Paxinos G,
    2. Watson C
    (1998) The Rat Brain in Stereotaxic Coordinates. (Academic Press, NSW, Australia).
  27. ↵
    1. Pedersen WA,
    2. McCullers D,
    3. Culmsee C,
    4. Haughey NJ,
    5. Herman JP,
    6. Mattson MP
    (2001) Corticotropin-releasing hormone protects neurons against insults relevant to the pathogenesis of Alzheimer's disease. Neurobiol Dis 8:492–503.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Pedersen WA,
    2. Wan R,
    3. Zhang P,
    4. Mattson MP
    (2002) Urocortin, but not urocortin II, protects cultured hippocampal neurons from oxidative and excitotoxic cell death via corticotropin-releasing hormone receptor type I. J Neurosci 22:404–412.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Perfetti R,
    2. Zhou J,
    3. Doyle ME,
    4. Egan JM
    (2000) Glucagon-like peptide-1 induces cell proliferation and pancreatic-duodenum homeobox-1 expression and increases endocrine cell mass in the pancreas of old, glucose-intolerant rats. Endocrinology 141:4600–4605.
    OpenUrlCrossRefPubMed
  30. ↵
    1. Perry T,
    2. Hodges H,
    3. Gray JA
    (2001) Behavioural, histological and immunocytochemical consequences following 192 IgG-saporin immunolesions of the basal forebrain cholinergic system. Brain Res Bull 54:29–48.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Perry T,
    2. Lahiri DK,
    3. Chen D,
    4. Zhou J,
    5. Shaw KT,
    6. Egan JM,
    7. Greig NH
    (2002) A novel neurotrophic property of glucagon-like peptide 1: a promoter of nerve growth factor-mediated differentiation in PC12 cells. J Pharmacol Exp Ther 300:958–966.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Ryan AS,
    2. Egan JM,
    3. Habener JF,
    4. Elahi D
    (1998) Insulinotropic hormone glucagon-like peptide-1-(7–37) appears not to augment insulin-mediated glucose uptake in young men during euglycemia. J Clin Endocrinol Metab 83:2399–2404.
    OpenUrlCrossRefPubMed
  33. ↵
    1. Satoh F,
    2. Beak SA,
    3. Small CJ,
    4. Falzon M,
    5. Ghatei MA,
    6. Bloom SR,
    7. Smith DM
    (2000) Characterization of human and rat glucagon-like peptide-1 receptors in the neurointermediate lobe: lack of coupling to either stimulation or inhibition of adenylyl cyclase. Endocrinology 141:1301–1309.
    OpenUrlCrossRefPubMed
  34. ↵
    1. Scrocchi LA,
    2. Brown TJ,
    3. MaClusky N,
    4. Brubaker PL,
    5. Auerbach AB,
    6. Joyner AL,
    7. Drucker DJ
    (1996) Glucose intolerance but normal satiety in mice with a null mutation in the glucagon-like peptide 1 receptor gene. Nat Med 2:1254–1258.
    OpenUrlCrossRefPubMed
  35. ↵
    1. Scrocchi LA,
    2. Drucker DJ
    (1998) Effects of aging and a high fat diet on body weight and glucose tolerance in glucagon-like peptide-1 receptor −/− mice. Endocrinology 139:3127–3132.
    OpenUrlCrossRefPubMed
  36. ↵
    1. Shughrue PJ,
    2. Lane MV,
    3. Merchenthaler I
    (1996) Glucagon-like peptide-1 receptor (GLP1-R) mRNA in the rat hypothalamus. Endocrinology 137:5159–5162.
    OpenUrlCrossRefPubMed
  37. ↵
    1. Tang-Christensen M,
    2. Larsen PJ,
    3. Goke R,
    4. Fink-Jensen A,
    5. Jessop DS,
    6. Moller M,
    7. Sheikh SP
    (1996) Central administration of GLP-1-(7–36) amide inhibits food and water intake in rats. Am J Physiol 271:R848–R856.
    OpenUrlAbstract/FREE Full Text
  38. ↵
    1. Thorens B,
    2. Porret A,
    3. Buhler L,
    4. Deng SP,
    5. Morel P,
    6. Widmann C
    (1993) Cloning and functional expression of the human islet GLP-1 receptor. Demonstration that exendin-4 is an agonist and exendin-(9–39) an antagonist of the receptor. Diabetes 42:1678–1682.
    OpenUrlAbstract/FREE Full Text
  39. ↵
    1. Torres EM,
    2. Perry TA,
    3. Blockland A,
    4. Wilkinson LS,
    5. Wiley RG,
    6. Lappi DA,
    7. Dunnett SB
    (1994) Behavioural, histochemical and biochemical consequences of selective immunolesions in discrete regions of the basal forebrain cholinergic system. Neuroscience 63:95–122.
    OpenUrlCrossRefPubMed
  40. ↵
    1. Turton MD,
    2. O'Shea D,
    3. Gunn I,
    4. Beak SA,
    5. Edwards CM,
    6. Meeran K,
    7. Choi SJ,
    8. Taylor GM,
    9. Heath MM,
    10. Lambert PD,
    11. et al.
    (1996) A role for glucagon-like peptide-1 in the central regulation of feeding. Nature (Lond) 379:69–72.
    OpenUrlCrossRefPubMed
  41. ↵
    1. Wang T,
    2. Edwards GL,
    3. Baile CA
    (1998) Glucagon-like peptide-1 (7–36) amide administered into the third cerebroventricle inhibits water intake in rats. Proc Soc Exp Biol Med 219:85–91.
    OpenUrlAbstract/FREE Full Text
  42. ↵
    1. Wei Y,
    2. Mojsov S
    (1995) Tissue-specific expression of the human receptor for glucagon-like peptide-I: brain, heart and pancreatic forms have the same deduced amino acid sequences. FEBS Lett 358:219–224.
    OpenUrlCrossRefPubMed
  43. ↵
    1. Weis C,
    2. Marksteiner J,
    3. Humpel C
    (2001) Nerve growth factor and glial cell line-derived neurotrophic factor restore the cholinergic neuronal phenotype in organotypic brain slices of the basal nucleus of Meynert. Neuroscience 102:129–138.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Download PDF
Article Alerts
Email Article
Citation Tools

Research ArticleNEUROPHARMACOLOGY

Protection and Reversal of Excitotoxic Neuronal Damage by Glucagon-Like Peptide-1 and Exendin-4

TracyAnn Perry, Norman J. Haughey, Mark P. Mattson, Josephine M. Egan and Nigel H. Greig
Journal of Pharmacology and Experimental Therapeutics September 1, 2002, 302 (3) 881-888; DOI: https://doi.org/10.1124/jpet.102.037481
Twitter logo Facebook logo Mendeley logo

Jump to section

Advertisement