点击显示 收起
Institut National de la Santee et de la Recherche Meedicale Unitee 679, Experimental Neurology and Therapeutics, Hpital de la Salpe簍rieere, Paris, (S.T., B.S.L., E.C.H., P.P.M.)
Centre de Recherche Pierre Fabre, Castres, France (M.M., F.C., P.P.M.)
Abstract
We have developed a model system in which rat basal forebrain cholinergic neurons degenerate progressively when maintained in culture conditions that make them susceptible to low-level oxidative stress. In this study, we showed that cholinergic neurons identified by acetylcholinesterase cytochemistry or choline acetyl transferase immunocytochemistry are rescued efficiently by the neurotransmitter noradrenaline (NA). The effect of NA required neither adrenoceptor activation nor intracellular accumulation. NA operated via a mechanism that precluded activation of a cell death pathway in which reactive oxygen species (ROS) and proapoptotic caspases were crucially involved. It is noteworthy that NA remained protective even when applied late in the degenerative process but before intracellular ROS began to increase. The high efficacy of iron chelators and catalase in preventing the death of cholinergic neurons in this model suggested that NA neutralized the effects of hydroxyl radicals produced through a Fenton-type reaction. Pyrocatechol [the diphenolic moiety of NA] was sufficient in itself to prevent ROS production and cholinergic cell demise, indicating that the catechol structure was instrumental for the neuroprotective function of NA. Therefore, the noncatecholic neurotransmitter GABA failed to prevent neurodegeneration. Nerve growth factor and brain derived neurotrophic factor, two trophic peptides for septal cholinergic neurons, did not afford protection by themselves and did not improve neuroprotection provided by NA. However, in the presence of NA, they both retained their efficacy to stimulate cholinergic parameters. These data indicate that NA-based therapeutic strategies may be of interest in such neurodegenerative conditions as Alzheimer's disease, where progressive cholinergic deficits occur.
The locus ceruleus (LC) provides the major source of noradrenaline (NA)-containing axons that are distributed widely throughout the central nervous system (Berridge and Waterhouse, 2003). This ubiquitous distribution is consistent with a prominent role for NA in a variety of central nervous system functions and behaviors that include locomotor control, cognition, motivation, and attention (Berridge and Waterhouse, 2003). In addition to these established transmitter signaling functions, NA evidently operates as an endogenous neurotrophic substance during development and adulthood (Meier et al., 1991; Colpaert, 1994; Goldstein, 2000; Marien et al., 2004). For instance, pharmacological agents that mimic or increase NA release, such as 2 adrenoceptor antagonists and amphetamine-related compounds, reduced neuronal death in experimental models of cerebral ischemia and improved behavioral and functional recovery after brain injury (Goldstein, 2000). Furthermore, increasing extracellular levels of NA by genetic invalidation of the NA transporter or by treatment with specific uptake inhibitors conferred protection to dopaminergic neurons in a mouse model of Parkinson's disease (Rommelfanger et al., 2004). On the other hand, lesions of the LC-NA system were shown to induce synaptic loss (Matsukawa et al., 2003), aggravate postischemic cerebral damage (Goldstein, 2000), increase brain inflammatory responses (Galea et al., 2003), or worsen neurological deficits in animal models of Parkinson's disease (Marien et al., 1993).
Although numerous theories for the cause of Alzheimer's disease (AD) and its symptoms have been postulated, it has most often been characterized as a disorder of progressive cholinergic deterioration. Studies of postmortem human brain have revealed a severe depletion of choline acetyltransferase (ChAT), the biosynthetic enzyme for acetylcholine, in the cortex and hippocampus, concomitant with the reduction in the number of basal forebrain cholinergic neurons (Mufson et al., 2003). A number of observations suggest that the progressive nature of these cholinergic deficits might be influenced by the LC-NA system: 1) synapses between noradrenergic nerve terminals and cholinergic neurons are frequently encountered throughout basal forebrain areas, including the vertical and horizontal limb of the diagonal band nuclei and the nucleus basalis magnocellularis (Fort et al., 1995); 2) there is a remarkably early and profound loss of NA cells in the LC in AD (Zarow et al., 2003); 3) stimulation of LC-NA output by the 2 adrenoceptor antagonist, dexefaroxan, affords protection in a retrograde lesion paradigm of basal forebrain cholinergic degeneration (Debeir et al., 2002; Chopin et al., 2004).
To examine the neuroprotective potential of NA on basal forebrain cholinergic neurons, we have developed a model system in which rat septal cholinergic neurons in culture degenerate progressively when maintained in conditions that make them susceptible to low-level oxidative stress. In particular, our aim was to use this model 1) to determine whether NA can prevent cholinergic cell death in vitro and to quantify its effects, 2) to explore the possible mechanisms underlying this protection, and 3) to examine whether trophic factors that are known to stimulate cholinergic differentiation and survival could modulate these effects.
Materials and Methods
Peptides and Pharmacological Agents. The L and D stereoisomers of NA (tartrate salts) were from Fluka (Saint Quentin Fallavier, France). Nerve growth factor 2.5S (NGF; Grade II) and brain-derived neurotrophic factor (BDNF) were purchased from Alomone Labs (Jerusalem, Israel) and R&D Systems (Minneapolis, MN), respectively. Synthetic peptide inhibitors of caspases, including carbobenzoxy-Asp-Glu-Val-Asp fluoromethylketone (Z-DEVD-FMK), Z-Ile-Glu-Thr-Asp-FMK (Z-IETD-FMK), and Z-Leu-Glu-His-Asp-FMK (Z-LEHD-FMK) were obtained from VWR International (Fontenay/Bois, France). The fluorogenic caspase substrate CaspACE-fluorescein-isothiocyanate-Val-Ala-Asp-FMK (FITC-VAD-FMK) was obtained from Promega (Madison, WI). Unless specified, other pharmacological and biochemical agents were purchased from Sigma/RBI-Aldrich (Saint Quentin Fallavier, France). [1-14C]-acetyl-coenzyme-A and L-[7,8-3H]-NA were obtained from Amersham Biosciences (Orsay, France).
Cultures of Septal Cholinergic and LC Noradrenergic Neurons. Animals were treated in accordance with the National Institutes of Health's Guide for the Care and Use of Laboratory Animals, European Directive no. 86/609, and the guidelines of the local institutional animal care and use committee. Septal cultures were prepared from embryonic day 15.5 Wistar rat embryos (Janvier Breeding Center, Le Genest St. Isles, France). The septal area was dissected as described by Hartikka and Hefti (1988). The dissected area contained cholinergic neurons that give rise to the medial septal nucleus and the nucleus of the diagonal band of Brocca (Hartikka and Hefti, 1988). Tissue pieces were triturated mechanically without enzymatic treatment using modified L15 medium (Hartikka and Hefti, 1988). Cultures were plated at a density of 2.5 to 3 x 105 cells/cm2 and were maintained initially in 500 e of N5 medium (Kawamoto and Barrett, 1986) supplemented with 5 mM glucose, 5% horse serum, and 0.5% fetal calf serum. Glial cell proliferation was prevented by gradually increasing the concentration of the antimitotic cytosine arabinoside (araC) according to the culture age [day in vitro (DIV)] as follows: 1.5 e (DIV 0), 3 e (DIV 1eC2), and 5 e (DIV 3eC5). To generate low-level oxidative stress, the N5 medium was replaced at DIV 10 with a chemically defined serum-free medium containing a trace amount of ferrous iron (0.5 e). This medium consisted of minimal essential medium with Earle's salts/Ham's F-12 nutrient mixture (2:1; Invitrogen, Cergy Pontoise, France) supplemented with insulin (10 e/ml), glucose (30 mM), and penicillin/streptomycin (100 U/ml). Because neurons were grown in an astrocyte-poor environment that favors excitotoxic stress, the noncompetitive NMDA receptor antagonist dizocilpine (MK-801; 3 e) was also added to the culture medium from DIV 3 until the end of the culture period. Unless specified, 50 ng/ml NGF was used to supplement the culture media to stimulate cholinergic differentiation. Some experiments were also performed with BDNF at a final concentration of 20 ng/ml. The cultures were fed daily by replacing 350 e of the iron-containing culture medium. They were then supplemented with appropriate pharmacological treatments including NA and adrenoceptor agonists/antagonists. Cholinergic neurons in control cultures represented 1% of all septal neurons.
LC cultures were obtained from embryonic day 14 Wistar rat embryos. The area dissected corresponded to the whole proximal rhombencephalic ring between the distal part of the mesencephalic flexure and the proximal part of the pontine flexure (Holm et al., 2003). After trituration, the cells were plated at a density of 3 to 4 x 105 cells/cm2 and then maintained for 9 to 10 days in a culture medium consisting of minimal essential medium with Earle's salts /Ham's F-12 nutrient mixture (1:1) supplemented with fetal calf serum (0.5%), N2 cocktail (1x; Invitrogen), BDNF (20 ng/ml), glucose (30 mM), penicillin/streptomycin (100 U/ml), and 3 e MK-801. araC (1.5 e) was added between DIV 1 and 3 to limit glial cell proliferation.
Acetylcholine Esterase Cytochemistry. Acetylcholinesterase (AChE) cytochemistry was performed as described previously (Michel and Agid, 1995). In brief, after fixation with 4% formaldehyde and three washes with Dulbecco's phosphate-buffered saline (D-PBS) the cultures were incubated at 4°C in 50 mM acetate buffer, pH 5.0, containing 4 mM acetylthiocholine iodide, 2 mM copper sulfate, 10 mM glycine, and 10 mg/ml gelatin to prevent diffusion of the reaction products. Nonspecific choline esterases were inhibited by 0.2 mM ethopropazine. After 3 to 4 days, the gelatin was dissolved by briefly incubating the cultures at 37°C. The cultures were rinsed with distilled water and exposed for 30 min in 10% potassium ferricyanide. The reaction was stopped by washing with distilled water.
Choline Acetyltransferase Immunodetection. Cultures fixed with 4% formaldehyde were incubated for 4 days at 4°C with a ChAT monoclonal antibody (1:50 in D-PBS) derived from a rat hybridoma (Diasorin, Stillwater, MN). Subsequent incubations were performed at room temperature with a biotinylated anti-rat Ig antibody from goat (1:100 in D-PBS; Amersham Biosciences, Orsay, France), followed by amplification with a preformed avidin-biotin horseradish peroxidase complex (Vectastain; Vector Laboratories, Burlingame, CA). The peroxidase was revealed by incubation with a solution of diaminobenzidine (0.5 mg/ml) in D-PBS containing 0.0075% hydrogen peroxide (H2O2). The reaction was stopped by two rapid washes with D-PBS.
Measurement of ChAT Activity. ChAT activity was measured by the method described by Hartikka and Hefti (1988) with minor modifications. In brief, the cultures were washed three times with D-PBS and lysed at 4°C with 250 e/well of 50 mM Tris-HCl buffer, pH 6, containing 0.3% Triton-X-100. Aliquots (50 e) of the lysate were mixed with 100 e of a solution containing 30 mM choline chloride, 0.6 mM eserine hemisulfate, 900 mM NaCl, and 150 mM Na2HPO4. The reaction was initiated by addition of 50 e of a 10 mM sodium phosphate buffer, pH 7.4, containing [1-14C]acetyl-coenzyme-A diluted with unlabeled acetyl-CoA to give a final substrate concentration of 20 e (specific activity, 4 mCi/mmol). Radioactivity corresponding to the reaction product ([14C]ACh) was measured by organic liquid scintillation counting. Note that the rate of ChAT activity was linear with time and amount of cell homogenate added.
Quantification of Reactive Oxygen Species. The production of mitochondrial reactive oxygen species (ROS) was quantified using dihydrorhodamine 123 (DHR-123; Molecular Probes, Montluon, France) according to the method described by Troadec et al. (2002). In brief, septal cultures were exposed for 30 min to 50 e DHR-123, washed three times, and then maintained in serum-free medium supplemented with the test treatments. The fluorescent signal, visualized by epifluorescence microscopy (excitation at 488 nm; emission at 515 nm), was quantified using the Simple-PCI software from C-Imaging Systems (Cranberry Township, PA) and a Nikon TE-300 inverted microscope equipped with an ORCA-ER digital camera (Hamamatsu Photonics, Massy, France). Fluorescent images of randomly chosen fields (six to eight in each culture condition) were acquired with a 63x fluorescent objective. The average pixel intensity over the surface of each cell body was determined under the different test conditions. Background fluorescence was subtracted from raw data, and the results were expressed as a percentage of the mean fluorescence intensity per cell in control cultures. A minimum of 60 cells was analyzed under each test condition. In some experiments, ROS were measured specifically in the population of magnocellular cholinergic neurons identified on the basis of their larger size under phase contrast microscopy optics.
Detection of Caspase Activation. CaspACE-FITC-VAD-FMK in situ marker was added directly to the cultures at a final concentration of 10 e in the presence of the test treatments. After incubation for 20 min at 37°C, the cultures were washed three times and then maintained with the test treatments during examination. The fluorescent signal was visualized by epifluorescence microscopy using the same filters as for DHR-123. Fluorescent cells were counted in 20 to 30 fields using a 40x objective.
Quantification and Visualization of [3H]NA Uptake. Intraneuronal accumulation of NA via active transport was monitored in septal and LC cultures using a protocol adapted from Troadec et al. (2002). In brief, after preincubation for 10 min in 500 e of PBS containing 5 mM glucose and 100 e ascorbic acid, the uptake was initiated by addition of 50 nM [3H]NA (37 Ci/mmol) to the cultures and terminated after 15 min by removal of the incubation medium followed by two rapid washes with ice-cold PBS. Cells were scraped off the culture wells and counted by liquid scintillation counting. When the accumulation of [3H]NA was visualized by microautoradiography, the incubation time with [3H]NA was extended to 30 min. After two rapid washes with PBS, the cultures were fixed for 20 min in PBS containing 0.5% glutaraldehyde and 4% formaldehyde, and dehydrated with ethanol. The incorporation of the tritiated label was detected with the Hypercoat LM-1 emulsion (Amersham Biosciences) after an exposure of 8 to 10 days in the dark at 4°C. Blockade of NA saturable transport was obtained in the presence of 0.5 e desipramine.
Statistical Analysis. Comparisons between two groups were performed with Student's t test. Multiple comparisons against a single reference group were made by one-way analysis of variance followed by Dunnett's test. When all pairwise comparisons were made, the Student-Newman-Keuls test was used. S.E.M. values were derived from at least three independent experiments.
Results
NA Provides Long-Lasting Protection to Cholinergic Neurons. In the model system used here, rat septal cholinergic neurons were maintained initially for 10 days in a culture environment that favored their maturation and differentiation, and then exposed to trace amounts of ferrous iron to stimulate low-level oxidative stress. In the first 3 to 4 days after the initiation of the iron treatment, no noticeable morphological changes were detectable, and the number of cholinergic neurons assessed by AChE cytochemistry remained virtually unchanged (Fig. 1A). After DIV 14, however, AChE+ neurons began to degenerate progressively such that >90% had died by DIV 16 (Fig. 1A). When NA (5 e) was applied from DIV 10 onward, the loss of AChE+ neurons was completely prevented (Figs. 1A and 2, top). The survival-promoting effects of NA could be extended to at least DIV 20. When the addition of iron was made at DIV 13 instead of DIV 10, the loss of cholinergic neurons did not occur between DIV 14 and 16, as observed previously, but between DIV 18 and 20, which suggested that 1) neuronal demise was not dependent on a specific stage of maturation of the cultures and 2) a period of at least 3 to 4 days was required to initiate the degenerative process in the presence of the transition metal. In this paradigm, NA again provided robust protection to cholinergic neurons (Fig. 1A). Note that in subsequent experiments made with septal cultures, treatments with iron were initiated at DIV 10.
AChE cytochemistry is a reliable and sensitive marker to detect and quantify cholinergic neurons in septal cultures (Hartikka and Hefti, 1988; Michel and Agid, 1995). However, to confirm and extend our initial observation, we also counted ChAT+ cells at DIV 17 (i.e., at a time when extensive cell loss was occurring in control cultures exposed to iron first at DIV 10). ChAT+ cell counts revealed that the survival promoting effect of NA was concentration dependent with an EC50 of 0.3 e (Fig. 1B). Maximal effects were observed in the range of concentrations between 1 and 5 e (Fig. 1B). ChAT+ neurons exposed to NA were well differentiated (Fig. 2, bottom) and could be distinguished from other septal neurons on the basis of their size; i.e., the widest diameter of cholinergic neurons was, on average, 22.7 ± 0.4 e, whereas that of noncholinergic cells was 12.8 ± 0.4 e. Thus, NA preserved not only the survival but also the key morphological features of this population of neurons. In comparison degenerating cholinergic cells remaining in control cultures had shrunken cell bodies, and their neurites were either fragmented or absent (Fig. 2). Consistent with the results obtained by cell counting, NA prevented the loss in ChAT enzymatic activity associated with neuronal demise (Fig. 1B), indicating that cholinergic neurons rescued by NA remained functional. Finally, phase contrast examination of the cultures revealed that NA was also effective in providing neuroprotection to noncholinergic (mostly GABAergic) neurons in septal cultures (data not shown).
The Effects of NA on Cholinergic Neuronal Survival Are Independent of Adrenoceptors. We compared the neuroprotective potential of NA with that of selective agonists of adrenoceptors. Neither 1, 2 (phenylephrine and UK14,304, respectively), 1, nor 2 (xamoterol and procaterol, respectively) adrenoceptor agonists, each used at 5 e, were able to substitute for NA treatment (Fig. 3A). In addition, no survival-promoting effect was detectable when phenylephrine and UK14,304 or xamoterol and procaterol treatments were combined. Therefore, the protective effect of NA (2 e) was antagonized neither by phentolamine (2 e) nor by propanolol (2 e), two prototypical antagonists of - and -adrenoceptors, respectively, indicating that NA acted by a mechanism that presumably circumvented adrenoceptors (Fig. 3B). There were two other strong arguments in favor of this hypothesis: 1) the D stereoisomer of the neurotransmitter, which is not active at the adrenoceptor site (Hearse and Sutherland, 1999), was as protective as the naturally occurring L form (i.e., the form used throughout the manuscript except when specified), and 2) pyrocatechol, a compound that reproduces the structure of the diphenolic moiety of NA but lacks significant binding affinity for adrenoceptors (Vauquelin et al., 1979), mimicked in a concentration-dependent manner the protective effects of the neurotransmitter for cholinergic neurons (Fig. 3C).
NA Rescues Cholinergic Neurons by Preventing Oxidative Stress. Because the neuroprotective effect of NA for cholinergic neurons seemed to be independent of adrenoceptor activation, we examined the nature of other putative mechanisms of action. The survival promoting activity of NA was mimicked by Trolox (10 e; RBI/Sigma), a cell-permeable, water-soluble derivative of vitamin E, suggesting that NA itself behaved as an antioxidant (Fig. 4A). The use of the fluorescent probe DHR-123 revealed that NA was indeed as potent as Trolox in reducing ROS production in degenerating septal neurons (Fig. 4B) and in particular in the population of magnocellular cholinergic neurons (Fig. 4C). Pyrocatechol and the D stereoisomer of NA, which were as potent as NA in protecting cholinergic neurons (Figs. 3 and 4A) were also highly effective in reducing intracellular ROS production in our culture model (Fig. 4B). However, two other antioxidants, vitamin C 2-phosphate (100 e), a stable form of vitamin C, and N-acetyl cysteine (NAC; 5 mM), the precursor of reduced glutathione, both failed to provide neuroprotection (Fig. 4A). In addition, vitamin C and NAC were also totally ineffective in reducing oxidative stress in the present paradigm (Fig. 4B). Finally, GABA, a neurotransmitter that is structurally unrelated to NA in that it does not possess a catechol ring, was totally ineffective in reducing ROS levels and in rescuing cholinergic neurons from degeneration.
Besides Trolox and pyrocatechol, iron chelators such as desferrioxamine (10 e) and apotransferrin (100 e/ml) reproduced the neuroprotective action of NA (Fig. 5A). Neuroprotection with desferrioxamine and apotransferrin was associated with a decrease in the production of intracellular ROS similar to that obtained with NA (5 e) (Fig. 5B). This demonstrates that NA prevented a degenerative process initiated by iron, which is present at a low concentration in our culture medium. Consistent with this observation, cholinergic neurons were resistant to degeneration when maintained with the same culture medium formulated without iron (data not shown). It is interesting that catalase (300 IU/ml), an enzyme known to prevent iron-catalyzed Fenton-type chemistry, afforded a degree of protection and a reduction in the emission of intracellular ROS similar to that provided by NA (Fig. 5).
NA Provides Neuroprotection but Is Not Accumulated Intracellularly. Blocking NA uptake with desipramine, a selective inhibitor of the high-affinity NA transport system, was reported previously to prevent the trophic effects of NA in another model system (Zhang et al., 1997). This led us to speculate that the neuroprotective and thus antioxidant-like effect of NA might result from its accumulation into cholinergic neurons by active transport. Such a mechanism, however, was apparently not involved in the present model system. Indeed, in septal cultures, there was no detectable desipramine-sensitive NA uptake, in contrast to what we observed in a culture system that contains NA neurons (i.e., LC cultures) (Fig. 6A). Consistent with these observations, microautoradiographic studies revealed that the tritiated label was virtually absent from cholinergic neurons and other neurons in septal cultures (Fig. 6, B and C) but readily detectable in LC NA neurons and their processes (Fig. 6D). Not surprisingly, desipramine (0.5 e) was unable to reduce the survival promoting action of NA (1 e) for cholinergic neurons (Fig. 6E). No positive label was detectable in cholinergic neurons or in other neurons if the treatment with 50 nM [3H]NA was applied persistently (DIV 10eC17) to septal cultures maintained concomitantly with Trolox (10 e) to prevent neuronal cell demise (data not shown).
At What Stage of the Degenerative Process Does the Presence of NA Become Critical As a Protective Factor Experiments were carried out to determine whether the onset of cholinergic cell loss was preceded by a progressive accumulation of intracellular ROS (i.e., before DIV 14) and whether the antioxidant effect of NA was detectable during this time. ROS were measured in DIV 13 and 15 cultures in the absence and presence of NA (5 e). An increased production of ROS was observed only when the cell death process was truly engaged and not before (Fig. 7A); therefore, the antioxidant effect of NA was evident only during the stage when cell demise was occurring in the cultures. This might explain why NA could be protective to cholinergic neurons even if its application was delayed up to DIV 15 (Fig. 7B).
Protection of Cholinergic Neurons by NA Occurs via Inhibition of a Caspase-Dependent Mechanism. To further characterize the neuroprotective mechanisms of NA in our model, a possible involvement of proapoptotic caspases in cholinergic neuronal death was examined. For that reason, we compared the survival promoting effect of NA to that of several caspase inhibitors; Z-DEVD-FMK (100 e), a broad-range cell permeable caspase inhibitor; Z-IETD-FMK (50 e), a specific inhibitor of caspase-8; and Z-LEHD-FMK (100 e), an inhibitor that is relatively specific for caspase-9. Z-IETD-FMK mimicked completely the neuroprotective effect of NA, whereas Z-DEVD-FMK was partially protective and Z-LEHD-FMK was completely ineffective (Fig. 8A). It is noteworthy that the neuroprotective action of Z-IETD-FMK was not additive to that produced by NA (data not shown), suggesting that both compounds acted by preventing a common cell death pathway.
A more direct approach to test for caspase involvement was taken by exposing the cultures to CaspACE FITC-VAD-FMK (10 e), a cell-permeable pan-caspase fluorogenic substrate. Visual examination of control cultures revealed the presence of a strong fluorescent signal in 8 to 12% of neuronal cells (Fig. 8, B and C), particularly in those neurons exhibiting evident morphological changes. The number of cell bodies labeled by CaspACE FITC-VAD-FMK was dramatically reduced by the presence of NA or Z-IETD-FMK, attesting to the reduction of caspase activity by the neurotransmitter (Fig. 8, B and C). It is interesting that the use of DHR-123 revealed that caspase inhibition by Z-IETD-FMK was also very effective in preventing ROS production in septal neurons and more specifically so in the population of magnocellular neurons exhibiting the typical morphology of cholinergic neurons (Fig. 8, D and E).
Is the Neuroprotective Action of NA Modulated by NGF In our model, the neurotrophin NGF was provided under basal conditions to stimulate the differentiation of cholinergic neurons. To determine whether NGF influenced the neuroprotective action of NA, we assessed the efficacy of a treatment with NA (DIV 10eC17) on cholinergic neurons maintained throughout their lifespan in the absence of the neurotrophin. NA had the same efficacy in preserving cholinergic neuronal numbers in the absence of NGF (Fig. 9A), suggesting that the survival promoting action of NA was not dependent on NGF. It is noteworthy that when NA was absent from the culture medium, NGF was totally ineffective in protecting cholinergic neurons from degeneration. However, NGF dramatically increased the proportion of cholinergic neurons expressing high levels of ChAT in NA-treated cultures, indicating that the capacity of this factor to stimulate ChAT expression was dependent on the presence of the neurotransmitter. The proportion of cells expressing high levels of ChAT was the same regardless of whether the NA treatment in the absence of NGF was initiated at DIV 1 or DIV 10 (data not shown). In line with these findings, ChAT enzymatic activity was at least 3-fold higher in the presence of NGF and NA compared with NA alone (Fig. 9B). Furthermore, the cell bodies of cholinergic neurons treated with NGF and NA were larger than those treated with NA alone (Fig. 9C).
BDNF, another neurotrophin known to promote the development and maintenance of cholinergic neurons (Knusel et al., 1992), was examined under similar experimental conditions to determine whether it, too, could influence the protective effects of NA. Similar to what was observed in NGF/NA-treated cultures, the large majority of cholinergic neurons rescued by cotreatment with BDNF and NA expressed high levels of the ChAT enzyme (Fig. 9, A and C). These neurons, however, were smaller than those treated with NGF and NA (Fig. 9C). Consistent with this finding, BDNF was also less effective than NGF in stimulating ChAT activity in the presence of NA (Fig. 9B). It is noteworthy that BDNF, like NGF, failed to increase cholinergic cell survival in the absence or the presence of NA (Fig. 9A).
Discussion
In this study, we demonstrated that NA provided sustained and robust neuroprotection in a model system of septal cholinergic cell death in vitro. It is noteworthy that rescued cholinergic neurons were functional and well preserved morphologically. Neuroprotection by NA was atypical in that it did not require adrenoceptor activation. NA evidently operated through a mechanism that precludes activation of a ROS-dependent cell death pathway. Neither NGF nor BDNF alone was able to mimic the protective action of NA, although both neurotrophins seemed dependent on the presence of the neurotransmitter to reinforce the expression of the cholinergic phenotype.
NA Protects Cholinergic Neurons from Oxidative Stress Caused by a Fenton-Type Reaction. Several observations indicate that NA operates in our model by reducing oxidative stress: 1) cholinergic neurons were protected efficiently by an analog of vitamin E Trolox and by pyrocatechol, a diphenolic compound with a strong antioxidant capacity (Troadec et al., 2002); 2) the rescue of cholinergic neurons was also achieved by desferrioxamine and apotransferrin, two molecules known to sequester iron, a transition metal catalyst for the Fenton reaction that allows the conversion of H2O2 into highly reactive hydroxyl radicals (Halliwell and Gutteridge, 1984); and 3) detoxification of H2O2 by catalase was strongly neuroprotective. The finding that NA reduced ROS emission to a large extent within septal neurons, particularly the population of cholinergic neurons, confirmed that NA was operating by preventing oxidative stress or its consequences. The antioxidant function of NA was apparently not mediated by activation of membrane located adrenoceptors. It seemed to be related, however, to the presence of a catechol moiety in its chemical structure because 1) its effects were reproduced in a concentration-dependent manner by the structural congener pyrocatechol (the o-catechol moiety of NA), and 2) GABA, a neurotransmitter with no catechol ring, was totally ineffective in reducing ROS levels and rescuing cholinergic neurons.
NA Operates via an Extracellular Mechanism. Both the protective and antioxidant effects of NA for cholinergic neurons were mimicked by long-term treatment with catalase, an enzyme that neutralizes ROS exclusively in the extracellular compartment because it cannot penetrate the plasma membrane of intact cells (Muzykantov, 2001). Several observations would suggest that the site of the antioxidant effect of NA was, like that of catalase, extracellular. First, neuroprotection by NA was independent of adrenoceptors, ruling out an antioxidant effect resulting from the activation of intracellular signaling pathways. Second, accumulation of NA via active transport has been described in peripheral nervous system developing cholinergic neurons (Habecker et al., 2000), but we found that this mode of transport was absent in basal forebrain cholinergic neurons. Third, the possibility that the accumulation of NA or its potential metabolites occurred by passive transport is also unlikely because we failed to detect substantial amounts of radioactive label within cholinergic neurons as well as other septal neurons, even after long-term exposure to [3H]NA. Therefore, we may surmise that the action of the neurotransmitter consisted in preventing the propagation of oxidative stress at the level of the outer plasma membrane, thus precluding ROS from being generated intracellularly. Consistent with this view, NA has been shown to prevent the peroxidation of membrane lipids in cell-free systems (Andorn and Pappolla, 2001). We cannot exclude the possibility that iron scavenging by the neurotransmitter represented the underlying chemical mechanism. Based on previous studies performed with other catechol derivatives (Stites et al., 2000), however, it is also conceivable that redox-cycling between NA and its autoxidation product NA-quinone participated in neuroprotection.
When Do Cholinergic Neurons Require the Presence of NA to Survive We were surprised to find that the death of cholinergic neurons did not occur immediately after addition of iron to the cultures at DIV 10 but only after a delay of 3 to 4 days. It may be that cholinergic neurons become susceptible to oxidative stress upon reaching some critical stage just before DIV 15 to 16 in culture, although this seems unlikely because a comparable delay before neurodegeneration was observed when the treatment with iron was postponed by several days. On the other hand, as a consequence of the treatment with iron, intracellular ROS might accumulate progressively before reaching levels sufficient to kill cholinergic cells. If this were the case, one would expect NA to suppress intracellular ROS production during that time. Arguing against this hypothesis, ROS levels were identical in NA-treated and control cultures before the onset of cholinergic cell degeneration. It cannot be excluded that NA protected the plasma membrane of cholinergic neurons from oxidative stress-mediated damage long before an increase in ROS levels was detectable intracellularly. The fact that NA was still able to effectively interrupt the degenerative process even when applied several days after placing the cultures in a ROS-producing environment indicates that these alterations if they occurred were still reversible in the presence of the neurotransmitter.
NA Reduces Intracellular ROS Production via Inhibition of a Caspase-Dependent Mechanism. Two arguments suggest that caspases [i.e., proteases that mediate apoptotic cell death (Hengartner, 2000)] were crucially involved in the death pathway inhibited by NA: 1) Z-IETD-FMK, a selective inhibitor of caspase-8, and Z-DEVD-FMK, a broad-range caspase inhibitor reproduced totally and partially the protective effect of NA, respectively; and 2) NA itself prevented caspase activation. Given that the site of action of NA was presumably extracellular, caspase inhibition produced by the neurotransmitter was most probably indirect. Activation of caspase-8 and other caspases is frequently caused by stimulation of the death domain receptor Fas (FasR) by its cognate ligand FasL (Qiu et al., 2002), suggesting that NA might prevent a FasR-related mechanism. This seems unlikely, however, because cholinergic neurons were resistant to a soluble recombinant FasL protein, and a FasL recombinant neutralizing antibody failed to reproduce neuroprotection by NA (data not shown). It is worth noting that the reduction of oxidative stress produced by the inhibitor of caspase-8 in cholinergic neurons was similar to that obtained with NA, which indicates that intracellular ROS were probably generated downstream of the level at which the caspase-dependent mechanism intervenes. It is interesting that a recent study has linked mitochondrial ROS production during apoptosis to caspase-dependent cleavage of a complex I subunit of the mitochondrial electron transport chain (Ricci et al., 2004).
NA Provides Optimal Neuroprotection Even in the Absence of NGF. Unlike dopaminergic neurons, which were protected only partially against oxidative stress by NA [40% of them were rescued after 10 days of treatment (Troadec et al., 2001)], virtually all septal cholinergic neurons survived to ROS-mediated damage in the presence of the neurotransmitter. Such a difference may lie in the fact that NGF present continuously for the differentiation and maturation of cholinergic neurons may have already enhanced the survival promoting effects of NA in the same way that cAMP-elevating agents improved the effects of the neurotransmitter for dopaminergic neurons in mesencephalic cultures (Troadec et al., 2002). Supporting this view, NGF can attenuate the deleterious consequences of oxidative stress-mediated neuronal injuries in some culture models (Salinas et al., 2003); on the other hand, mitochondrial-derived ROS are produced in NGF-responsive neurons that are deprived of the neurotrophin (Kirkland and Franklin, 2003). To determine whether NGF improved the neuroprotective action of NA on cholinergic neurons, we performed experiments with cultures maintained throughout their lifespan without the neurotrophin. Despite the absence of NGF, NA alone was sufficient to rescue virtually all cholinergic neurons susceptible to degeneration. This indicates that NGF was not essential for survival promotion by NA. Consistent with this observation, NGF was unable to promote cholinergic cell survival or to reduce ROS production in the absence of NA. Therefore, the presence of NGF cannot account for the high efficacy of NA to rescue septal cholinergic neurons.
NGF and BDNF Retain Their Capacity to Stimulate Cholinergic Parameters in the Presence of NA. Even though NGF lacked survival-promoting effects either alone or in the presence of NA, it did improve the impact of NA by increasing ChAT expression and cell size in surviving cholinergic neurons. BDNF, another neurotrophin known to function as a trophic factor for cholinergic neurons (Knusel et al., 1992), had an effect that was similar to but somewhat less pronounced than that of NGF. That exogenous NGF and BDNF were able to stimulate cholinergic parameters exclusively in the presence of NA is noteworthy, because a number of diphenolic antioxidants have been shown to interfere with the function of some regulatory peptides or cytokines (Zi et al., 1998). Our results are more consistent with studies showing that NA can act cooperatively with trophic peptides, such as glial cell line-derived neurotrophic factor and ciliary neurotrophic factor on specific populations of neurons in the brain (Louis et al., 1993; our unpublished data).
Recent studies support the idea that proapoptotic caspases (Rohn et al., 2001; Cribbs et al., 2004) and ROS-mediated reactions (Casadesus et al., 2004; Pratico and Sung, 2004) are crucially involved in neurodegeneration in AD. Our results showing that NA can substantially reduce both caspase activation and ROS production in cholinergic neurons, one of the neuronal populations most at risk in AD, provide support to the concept that compounds that stimulate NA output may have therapeutic utility in the treatment of this pathology.
doi:10.1124/mol.104.007864.
References
Andorn AC and Pappolla MA (2001) Catecholamines inhibit lipid peroxidation in young, aged and Alzheimer's disease brain. Free Radic Biol Med 31: 315eC320.
Berridge CW and Waterhouse BD (2003) The locus coeruleus-noradrenergic system: modulation of behavioral state and state-dependent cognitive processes. Brain Res Rev 42: 33eC84.
Casadesus G, Smith MA, Zhu X, Aliev G, Cash AD, Honda K, Petersen RB, and Perry G (2004) Alzheimer disease: evidence for a central pathogenic role of iron-mediated reactive oxygen species. Alzheimers Dis 6: 165eC169.
Chopin P, Debeir T, Raisman-Vozari R, Colpaert FC, and Marien MR (2004) Protective effect of the alpha2-adrenoceptor antagonist, dexefaroxan, against spatial memory deficit induced by cortical devascularization in the adult rat. Exp Neurol 185: 198eC200.
Colpaert F (1994) Noradrenergic mechanisms in Parkinson's disease: a theory, in Noradrenergic Mechanisms in Parkinson's Disease (Briley M, Marien M, eds), pp 225eC254, CRC Press, Boca Raton, FL.
Cribbs DH, Poon WW, Rissman RA, and Blurton-Jones M (2004) Caspase-mediated degeneration in Alzheimer's disease Am J Pathol 165: 353eC355.
Debeir T, Marien M, Chopin P, Martel JC, Colpaert F, and Raisman-Vozari R (2002) Protective effects of the alpha2-adrenoceptor antagonist, dexefaroxan, against degeneration of the basalocortical cholinergic system induced by cortical devascularization in the adult rat. Neuroscience 115: 41eC53.
Fort P, Khateb A, Pegna A, Me筯lethaler M, and Jones BE (1995) Noradrenergic modulation of cholinergic nucleus basalis neurons demonstrated by in vitro pharmacologial and immunohistochemical evidence in the guinea pig brain. J Neurosci 7: 1502eC1511.
Galea E, Heneka MT, Dello Russo C, and Feinstein DL (2003) Intrinsic regulation of brain inflammatory responses. Cell Mol Neurobiol 23: 625eC635.
Goldstein LB (2000) Effects of amphetamines and small related molecules on recovery after stroke in animals and man. Neuropharmacology 39: 852eC859.
Habecker BA, Klein MG, Cox BC, and Packard BA (2000) Norepinephrine transporter expression in cholinergic sympathetic neurons: differential regulation of membrane and vesicular transporters. Dev Biol 220: 85eC96.
Halliwell B and Gutteridge JM (1984) Free radicals, lipid peroxidation and cell damage. Lancet 2: 109519.
Hartikka J and Hefti F (1988) Development of septal cholinergic neurons in culture: plating density and glial cells modulate effects of NGF on survival, fiber growth and expression of transmitter-specific enzymes. J Neurosci 8: 2967eC2985.
Hearse DJ and Sutherland FJ (1999) Catecholamines and preconditioning: studies of contraction and function in isolated rat hearts. Am J Physiol 277: H136eCH143.
Hengartner MO (2000) The biochemistry of apoptosis. Nature (Lond) 407: 770eC776.
Holm PC, Rodriguez FJ, Kresse A, Canals JM, Silos-Santiago I, and Arenas E (2003) Crucial role of TrkB ligands in the survival and phenotypic differentiation of developing locus coeruleus noradrenergic neurons. Development 130: 3535eC3545.
Kawamoto JC and Barrett JN (1986) Cryopreservation of primary neurons for tissue culture. Brain Res 384: 84eC93.
Kirkland RA and Franklin JL (2003) Prooxidant effects of NGF withdrawal and MEK inhibition in sympathetic neurons. Antioxid Redox Signal 5: 635eC639.
Knusel B, Beck KD, Winslow JW, Rosenthal A, Burton LE, Widmer HR, Nikolics K, Hefti F (1992) Brain-derived neurotrophic factor administration protects basal forebrain cholinergic but not nigral dopaminergic neurons from degenerative changes after axotomy in the adult rat brain. J Neurosci 12: 4391eC4402.
Louis JC, Magal E, Burnham P, and Varon S (1993) Cooperative effects of ciliary neurotrophic factor and norepinephrine on tyrosine hydroxylase expression in cultured rat locus coeruleus neurons. Dev Biol 155: 1eC13.
Marien M, Briley M, and Colpaert F (1993) Noradrenaline depletion exacerbates MPTP-induced striatal dopamine loss in mice. Eur J Pharmacol 236: 487eC489.
Marien MR, Colpaert FC, and Rosenquist AC (2004) Noradrenergic mechanisms in neurodegenerative diseases: a theory. Brain Res Rev 45: 38eC78.
Matsukawa M, Nakadate K, Ishihara I, and Okado N (2003) Synaptic loss following depletion of noradrenaline and/or serotonin in the rat visual cortex: a quantitative electron microscopic study. Neuroscience 122: 627eC635.
Meier E, Hertz L, and Schousboe A (1991) Neurotransmitters as developmental signals. Neurochem Int 19: 1eC15.
Michel PP and Agid Y (1995) Death of septal cholinergic neurons produced by chronic exposure to glutamate is prevented by the noncompetitive NMDA receptor/channel antagonist, MK-801: role of nerve growth factor and nitric oxide. J Neurosci Res 40: 764eC775.
Mufson EJ, Ginsberg SD, Ikonomovic MD, and DeKosky ST (2003) Human cholinergic basal forebrain: chemoanatomy and neurologic dysfunction. J Chem Neuroanat 26: 233eC242.
Muzykantov VR (2001) Targeting of superoxide dismutase and catalase to vascular endothelium. J Control Release 71: 1eC21.
Pratico D and Sung S (2004) Lipid peroxidation and oxidative imbalance: early functional events in Alzheimer's disease. J Alzheimers Dis 6: 171eC175.
Qiu J, Whalen MJ, Lowenstein P, Fiskum G, Fahy B, Darwish R, Aarabi B, Yuan J, and Moskowitz MA (2002) Upregulation of the Fas receptor death-inducing signaling complex after traumatic brain injury in mice and humans. J Neurosci 22: 3504eC3511.
Ricci JE, Munoz-Pinedo C, Fitzgerald P, Bailly-Maitre B, Perkins GA, Yadava N, Scheffler IE, Ellisman MH, and Green DR (2004) Disruption of mitochondrial function during apoptosis is mediated by caspase cleavage of the p75 subunit of complex I of the electron transport chain. Cell 117: 773eC786.
Rohn TT, Head E, Nesse WH, Cotman CW, and Cribbs DH (2001) Activation of caspase-8 in the Alzheimer's disease brain. Neurobiol Dis 8: 1006eC1016.
Rommelfanger KS, Weinshenker D, and Miller GW (2004) Reduced MPTP toxicity in noradrenaline transporter knockout mice. J Neurochem 91: 1116eC1124.
Salinas M, Diaz R, Abraham NG, Ruiz de Galarreta CM, and Cuadrado A (2003) Nerve growth factor protects against 6-hydroxydopamine-induced oxidative stress by increasing expression of heme oxygenase-1 in a phosphatidylinositol 3-kinase-dependent manner. J Biol Chem 278: 13898eC13904.
Stites TE, Mitchell AE, and Rucker RB (2000) Physiological importance of quinoenzymes and the o-quinone family of cofactors. J Nutr 130: 719eC727.
Troadec JD, Marien M, Darios F, Hartmann A, Ruberg M, Colpaert F, and Michel PP (2001) Noradrenaline provides long-term protection to dopaminergic neurons by reducing oxidative stress. J Neurochem 79: 200eC210.
Troadec JD, Marien M, Mourlevat S, Debeir T, Ruberg M, Colpaert F, and Michel PP (2002) Activation of the mitogen-activated protein kinase (ERK) signaling pathway by cyclic AMP potentiates the neuroprotective effect of the neurotransmitter noradrenaline on dopaminergic neurons. Mol Pharmacol 62: 1043eC1052.
Vauquelin G, Bottari S, Kanarek L, and Strosberg AD (1979) Evidence for essential disulfide bonds in 1-adrenergic receptors of turkey erythrocyte membranes. Inactivation by dithiothreitol. J Biol Chem 254: 4462eC4469.
Zarow C, Lyness SA, Mortimer JA, and Chui HC (2003) Neuronal loss is greater in the locus coeruleus than nucleus basalis and substantia nigra in Alzheimer and Parkinson diseases. Arch Neurol 60: 337eC341.
Zhang JM, Dix J, Langtimm-Sedlak CJ, Trusk T, Schroeder B, Hoffmann R, Strosberg AD, Winslow JW, and Sieber-Blum M (1997) Neurotrophin-3- and norepinephrine-mediated adrenergic differentiation and the inhibitory action of desipramine and cocaine. J Neurobiol 32: 262eC280.
Zi X, Grasso AW, Kung HJ, and Agarwal R (1998) A flavonoid antioxidant, silymarin, inhibits activation of erbB1 signaling and induces cyclin-dependent kinase inhibitors, G1 arrest and anticarcinogenic effects in human prostate carcinoma DU145 cells. Cancer Res 58: 1920eC1929.