|Home | About | Journals | Submit | Contact Us | Français|
Limited evidence suggests that brain cytosolic phospholipase A2 (cPLA2), which selectively releases arachidonic acid (AA) from membrane phospholipids, and cyclooxygenase-2 (COX-2), the rate-limiting enzyme for AA metabolism to prostanoids, change as a function of normal aging. In this study, we examined the protein levels of cPLA2 and COX-2 enzymes in hippocampus, frontal pole and cerebellum from young (2–5 year-old), middle-aged (8–11 year-old) and old (23 year-old) male and female Rhesus monkeys. In the cerebellum, cPLA2 protein level was higher in the young brain as compared to levels seen at both middle-aged and old. Similarly, in the frontal pole, the young brain showed a higher level of COX-2 protein as compared to the levels seen at both older ages. For both, once an animal reached 8–11 years of age the levels appeared to remain relatively constant over the next decade. Immunohistochemistry of COX-2 protein within the brain demonstrated no significant change in the localization to neurons within the frontal pole. In the young brain, the distribution of a low level of COX-2 protein within numerous neurons was different than the decreased number of neurons stained at a greater intensity in the adult brain. Based on the previous reports of localization of cPLA2 and COX-2 at post-synaptic sites in neurons results from the current study suggest that the elevated protein levels of the two enzymes seen in the younger brain is related to the greater potential for synaptic plasticity across multiple neurons as a function of age and that cPLA2 and COX-2 may be considered as post-synaptic markers.
Cytosolic phospholipase A2 (cPLA2), an 85 kDa calcium dependent PLA2, initiates the arachidonic acid (AA) cascade by releasing AA from the sn-2 position of membrane phospholipids . Released AA can be further metabolized to bioactive eicosanoids - prostaglandins, thromboxanes, and leukotrienes - via cyclooxygenases (COX), lipoxygenases and epoxygenases, respectively . Of COX isoforms, COX-1 is constitutively expressed and produces eicosanoids for normal physiological function and COX-3, a recently discovered splice variant of COX-1, may be involved in pain mediation [5,15]. COX-2 is an isoform also constitutively expressed in brain and spinal cord and involved in various neuronal functions including synaptic signaling and neurobehavioral functioning, as well as in the regulation of cerebral blood flow . As an inducible protein, COX-2 is regulated by growth factors, tumor promoters, cytokines, mitogens, glucocorticoids, and has been implicated in related processes including inflammation. Prostanoids can modulate synaptic transmission, neurotransmitter release, the hypothalamic-pituitary-adrenal axis, the sleep/wake cycle, appetite, fever, pain and behavior [27,36]. cPLA2 and COX-2 often are co-expressed in different tissues, suggesting that cPLA2 can be responsible for generating the substrate (AA) for COX-2 for prostaglandin production in vivo . Additionally, cPLA2 and COX-2 immunoreactivities are reported to be co-localized on plasma membrane of Purkinje cells of the Rhesus monkey cerebellum  and at post synaptic sites in rat cerebral cortex .
Given the involvement of the AA cascade in inflammatory processes, a number of studies have suggested an alteration with the process of normal physiological aging and in the pathophysiological changes associated with a number of age-related diseases. Few rodent studies offer support for this hypothesis, however the data are less than convincing. Recent studies by Hayek et al demonstrated no significant difference in the low level of spontaneous PGE2 production in peritoneal macrophages obtained from rats 6 months to 2 years of age. Upon stimulation with lipopolysaccharide however, COX activity was higher in macrophages from the old rats as compared to the young, with an associated increase in COX-2 protein synthesis and mRNA expression . Neuroinflammation and the upregulation of both cPLA2 and COX-2 has been implicated in age-related neurodegenerative diseases, including Alzheimer’s disease  yet the experimental data has been less than conclusive of what role these enzymes may play in the process of brain aging. Some studies indicate no differences in COX-2 mRNA and protein levels in the rat brain in relation to age [2,24]. However, a significant decrease in COX-2 mRNA levels in the cortex has been reported in the aged rat . Examination of changes localized to neurons showed an increase of COX-2 protein in aged rats [24,31]. Based upon the induction of COX-2 by inflammatory signals, studies examining transgenic mice developed as models of neurodegenerative disease have reported age related increases in COX-2 protein in transgenic “beta-amyloid precursor protein” Swedish mutation mice leading to the elevated production of pro-inflammatory eicosanoids .
More recent data however, suggests that COX-2 is not exclusively related to inflammation but rather that it may play a critical role in brain development and synaptic signaling. COX-2 immunoreactivity is present in dendritic spines, which are involved in synaptic signaling . The ontogenic profile for COX-2 is consistent with critical periods for activity-dependent synapse formation and remodeling and thus, may play a role in cortical postsynaptic signaling [32,55]. Similarly, cPLA2 immunoreactivity in rat brain is preferentially localized at post-synaptic membranes in neuronal somata and dendrites . COX-2 expression is elevated in the newborn rodent brain cortex and periventricular white matter and accounts for a greater percentage of total COX-activity as compared to the adult . Evidence indicates that levels of drebrin, an actin-binding protein localized mainly in dendritic spines, is significantly decreased during healthy aging . In the current study, we compared the protein expression of cPLA2 and COX-2 in specific brain regions associated with cognitive function and motor coordination (frontal pole, hippocampus, and cerebellum) in normal healthy Rhesus monkeys at three distinct ages: young (2–5 years), middle aged adult (8–11 years) and old (23 years). Based on the evidence of loss of dendritic spines , drebrin , and reduced markers for neurotransmission [11,34,46] with aging suggesting synaptic loss, we hypothesize that cPLA2 and COX-2 protein expression, localized at post-synaptic sites, is decreased in old compared to young monkeys.
The study conformed to the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 80–23), and was approved by the NIH Animal Care and Use Committee. Male and female Rhesus monkeys (three 2–5 year-old, five 8–11 year-old, and five 23 year-old) were anaesthetized with 10 mg/kg body weight of ketamine, or 3 mg/kg body weight of telazol with an intravenous line in the femoral vein, then killed by an overdose of sodium pentobarbital. The brains were rapidly excised; the cerebellum and brainstem were removed, and the cerebral hemispheres were dissected along the midsagittal plane and each hemisphere was cut into 1-cm sequential coronal sections. From one hemisphere, the frontal pole, hippocampus, and cerebellum were isolated and immediately frozen on dry ice and stored at −80 °C. For immunohistochemistry, the 1-cm sections were fixed by immersion in 10% neutral buffered formalin. Health status of the monkeys was monitored according to NIH guidelines. All monkeys were healthy based on prior repeated veterinary examination and clinical blood chemistry. No gross organ changes were noted during necropsy.
Approximately 250 mg (wet weight) of each dissected brain region - frontal pole, hippocampus, and cerebellum - was homogenized in a buffer containing 25 mM Tris, 150 mM NaCl, 5 mM EGTA, protease inhibitor cocktail tablet (Roche, Mannheim, Germany) and 10 μM pepstatin A (Sigma, St. Louis, MO). The cytosolic fraction was obtained from the homogenate by centrifugation at 18,000 × g for 10 min at 4 °C, and its protein concentration was determined using the Bradford method .
Ninety lg of cytosolic protein was loaded on Bio-Rad Criterion gels (Bio-Rad, Hercules, CA) and transferred to a nitrocellulose membrane as directed by the manufacturer. Western blotting was carried out as reported  using primary antibodies for COX-2 (polyclonal, 1:500, Cayman), cPLA2 α (1:250, monoclonal, Santa Cruz Biotechnology, Santa Cruz, CA), or actin (1:10,000, Sigma) followed by a secondary antibody conjugated with horseradish peroxidase (1:1000, Bio-Rad). Immunoblots were visualized by a chemiluminescence reaction (Pierce, Rockford, IL). The protein level was quantified by measuring the integrated optical density of the bands, after background subtraction by AlphaEase Stand Alone software (Alpha Innotech, San Leandro, CA). Image analysis of the blots was performed on optical density-calibrated images captured with a videocamera (Alpha Innotech). Only bands below the saturation limit were analyzed.
Based upon western blotting protein analysis, the frontal lobe and the cerebellum collected from the contralateral hemisphere of the brains of the young (2–5 years) and the old (23 years) monkeys were processed for immunohistochemistry. Formalin-fixed tissue was rinsed with phosphate buffer followed by dehydration in a series of graded alcohols, and processed using a RHS-1 Microwave Vacuum Histoprocessor (Milestone, Shelton CT). Tissue was embedded in paraffin and 10-μm sections were cut and collected on ProbeOn+ slides (Fisher Scientific, Norcross, GA). Sections were deparaffinized, rehydrated through a graded series of ethanols and 1x Automation buffer, and treated with 3% H2O2 for 15 min to inhibit endogenous peroxidase activity, and then washed in 1X Automation buffer (AB; Biocare Medical, Walnut Creek, CA) for 5 min. Slides were transferred to 0.01 M citrate buffer (pH 6.0) and subjected to heat-induced epitope retrieval using a decloaking chamber (Biocare Medical) followed by two 5-min washes in 1X AB. Endogenous biotin activity was quenched with an avidin/biotin blocking kit (Vector Laboratories, Burlingame, CA). Sections were incubated at room temperature for 1 h with a mouse anti-COX monoclonal antibody (1:50; Cayman), followed by incubation with a biotinylated anti-mouse and incubation with an avidin-biotin complex (Vectastain® Elite kit, Vector Laboratories) prior to development with 3,3′-diaminobenzidine (DAB, Vector Labs) and counterstained with modified Harris hematoxylin. Astrocytes were detected by glial fibrillary acidic protein (GFAP) binding a major intermediate filament protein of glial cells and major cytoskeletal structure in astrocytes .Sections were stained with a rabbit anti-cow GFAP antibody (Dako, Carpinteria, CA) at 1:1500 for 30 min, followed by a secondary antibody (Dako) labeled streptavidin-biotin (LSAB) kit with HRP for 15 min.
All final dilutions were determined via optimization studies for detection specificity and decreased background, using various dilutions and negative control serum or isotype control. Light microscopy was performed on sections with a Leica DMRBE microscope (Wetzlar, Germany) equipped with differential interference contrast (DIC/Nomarski) optics, epifluorescence, and motorized Z-control. Digital images were acquired using a SpotRT™ cooled, charged-couple device camera (Diagnostic Instruments, Sterling Heights, MI) under the control of Metamorph™ software (Universal Imaging Co., Downingtown, PA). In all our experiments, serial concentrations of the antibody and two concentrations of the secondary were always run, as were negative controls at each concentration on concurrent sections from each animal.
For each brain region examined, statistical comparisons between groups for either cPLA2 or COX-2 protein levels were performed with one-way ANOVA. If one-way ANOVA yielded a p < 0.05, it was followed by a post-hoc Tukey’s multiple comparisons test. Significance was taken as p ≤ 0.05. In the absence of an overall statistical significance, post-hoc tests were not performed and results were considered as not significantly different. Data are presented as mean ± SEM.
Comparison of protein levels for cPLA2 and COX-2 showed regional and age related differences. In the young brain, cPLA2 protein levels were similar across the three brain regions, the frontal pole, hippocampus, and cerebellum (Fig 1A). In the hippocampus, the levels were similar across all three ages while in the frontal pole and the cerebellum lower levels were seen in the adult. This decrease reached statistical significance in the cerebellum (One way ANOVA’s F value = 7.81; p value = 0.009; Tukey’s post-hoc p < 0.05; Fig 1A, B).
A similar pattern was seen for COX-2 protein levels (Fig 2A). In the young brain, levels were similar across the three brain regions. In both the hippocampus and the cerebellum, protein levels did not change with age (Fig 2A) yet, in the frontal pole, lower protein levels were seen in the adult (One way ANOVA’s F value = 7.59; p value = 0.01; Tukey’s post-hoc test p < 0.05; Fig 2A). These levels were maintained in the aged brain (Fig 2A). Each individual region was tested at least twice with comparable results. Protein levels of β-actin showed no significant change between the samples in the different brain regions and the values were used to normalize for possible unequal loading of the gels. The ratios of cPLA2 or COX-2 protein level to actin were reported as percent of the ratio obtained in the young group.
A general histopathological evaluation of the brain regions suggested no overt cell death and no increase in infiltrating cells or microglia with age. With age, the general cellular morphology of GFAP positive astrocytes showed an increase and thickening of processes (Fig. 3).
COX-2 immunoreactivity was localized to neurons and not present in astrocytes with immunoreactivity present in large neurons in the allocortical regions. In the rodent brain, COX-2 immunoreactivity is located in the dendrites of excitatory neurons . In the current study, immunoreactivity in the frontal pole was present in neurons and extended into the neuronal processes in the young brain (Fig 4B,D). In the older animal, the immunoreactivity was primarily localized to the neuronal cell body with little evidence of staining within projections (Fig 4A,C). Qualitatively, a greater number of neurons were positively stained for COX-2 in the young brain than seen in the aged brain however, the less number of neurons appeared to show a greater staining intensity with age (Fig 4A,C). While previous studies report a prominent localization of COX-2 in the rat cortex and hippocampus [12,28], the current study showed that staining was also evident in the cerebellum of the non-human primate. Purkinje neurons of the cerebellum were immunopositive for COX-2 with staining localized to the cytoplasm of the cell body and extending into the processes. This distribution was seen in both the young (Fig. 5A) and the old (Fig. 5B) brains. As in the frontal pole, qualitatively, a greater number of neurons were positively stained for COX-2 in the young brain than in the old brain (Fig. 5C,D). This staining pattern was not observed in the negative controls and was seen at both lower and higher staining intensities with the respective antibody concentrations.
AA is released by PLA2 from membrane phospholipids and then becomes a substrate for COX, and other AA oxygenating enzymes that generate bioactive eicosanoids, which are involved in a wide range of signaling functions, as described in the Introduction . In the brain, COX-2 has been suggested to be functionally coupled with cPLA2 to synthesize prostaglandins [8,9,40]. Some studies in rodents suggest that the cPLA2/COX pathway in brain is affected in normal aging [2,14,33,45]. We found that cPLA2 protein levels were decreased in the cerebellum of 8–11 and 23 year-old monkeys, compared to the 2–5 year-old monkeys. In contrast, COX-2 protein level was significantly decreased in the frontal pole from 8–11 and 23 years old monkeys compared to the young monkeys. This decrease was confirmed by immunohistochemical analysis, which showed fewer immunopositive neurons as a function of age. In the frontal pole of old brains, COX-2 immunoreactivity was localized only to the cytoplasm of neurons, but it also was extended to neuronal processes in the young brain. This evidence for a decrease in COX-2 protein level in the frontal pole is consistent with the “frontal aging hypothesis” proposed by many researchers (reviewed in ). This hypothesis argues that the earliest and most severe age-related changes in cognitive functions and in neurotransmitters levels are found in the frontal lobes. In this regard, old Rhesus monkeys (25–32 years old) showed significant cognitive impairments and a significant decrease in α1-adrenergic and α2-adrenergic receptor binding in the prefrontal cortex .
The observed age-related decreases in cPLA2 and COX-2 protein levels may be a consequence of synaptic dysfunction, whereas it is unlikely to be determined by neuronal loss. Indeed, recent reports using modern stereological cell-counting methods did not show any significant age-related loss of hippocampal neurons in rats , mice , monkeys [30,41], or humans . Peters and Sethares reported that even with significant thinning with age of cortical layer 1, accompanied by a loss of some dendrites and spines and a decrease in the frequency of synapses, in the occipital and prefrontal cortex in Rhesus monkeys, there was no significant age-related changes in the number of neurons, astrocytes, or microglia and oligodendrocytes in the layer . In contrast, Smith and colleagues described significant reductions in the number and size of immunolabeled neurons in subcortical cholinergic basal forebrain regions in old Rhesus monkeys, suggesting that atrophy contributes to age-related cognitive decline in primates, and that subcortical brain regions are more vulnerable to the effects of aging than cortical regions . Furthermore, Kemper and colleagues reported a significant age related reduction in both cell packing density and in the total number of neurons in nucleus centalis superior of Rhesus monkey brain . Purkinje neuronal cell loss in the cerebellum also has been suggested to be involved in normal aging . Our results indicate astrocyte hypertrophy in both cerebellum and frontal lobe of aged compared to young monkeys, as shown by GFAP immunostaining. These results are in agreement with previous reports showing increased GFAP in the human entorhinal cortex and in the hippocampus with aging  and in the hippocampus and striatum of aged rats .
cPLA2 and COX-2 enzymes have been shown to be localized at post-synaptic sites in neurons [12,28,39] where they are involved in the stimulatory activity of excitatory neurons. The high levels of these proteins maintained over the three brain regions in the young animal may be reflective of the higher level of synaptic changes and remodeling that occurs during brain maturation. Maintenance of these protein levels within the hippocampus over the lifespan may also be reflective of the dynamic nature of this brain region and the requirements for continued synaptic remodeling.
Since COX-2 appears as an enzyme with dual roles, activity dependent neuronal plasticity and neuroinflammation, the lack of a prominent microglia response occurring in the brain regions examined suggests that protein levels in the normal brain are reflective of neuronal activity. If this is the case, than one could propose that the continued synaptic activity and remodeling occurring within the hippocampus and cortex would serve to maintain protein levels over time. The higher levels in the cerebellum of the younger animal may also reflect the dynamic synaptic remodeling occurring during the maturation of this brain region.
Given the close pathway interactions of these two enzymes, the lack of a correlation between the two with regards to brain region is interesting. While there are higher levels of these proteins in the young, regional decrease in protein levels may reflect synaptic changes, and particularly the loss or dysfunction of dendritic spines occurring in aging . In this regard, synaptic density has been shown to be decreased in the hippocampus from old compared to young subjects . The number of synapses/neurons also was significantly decreased in the hippocampus and the cerebellum . A reduced number of synapses (at least 30%) has been reported in the cerebellar circuitry of old as compared to young rats . A significant decline in levels of synaptic proteins, particularly drebrin, involved in structural plasticity (remodeling) of axons and dendrites was found also in aging human brain . Synaptic proteins such as synaptophysin, synaptobrevin, synaptotagmin, synaptosomal-associated protein 25 syntaxin 1/HPC-1 and drebrin are decreased during aging also in rat brain .
The decrease in regional protein levels of cPLA2 or COX-2 during aging that we report in this study, which could lead to subsequent alteration of the AA metabolic pathway to bioactive prostaglandins, may contribute to the age-related cognitive decline and synaptic dysfunction. AA and its products play an important role in synaptic transmission, functional hyperemia, blood flow regulation (reviewed in:[7,31,38]). Additionally, several G proteins-coupled neurotransmitter receptors, including dopaminergic D2 , adrenergic α2 , and serotoninergic 2A/2C , can promote AA release from membrane phospholipids. Data suggest that COX-2 may also be required for brain muscarinic signaling. In this regard, COX-2 deficient mice do not show any increase in AA incorporation in response to arecoline (Basselin, personal communication), and activation of muscarinic receptors by arecoline coupled to cPLA2 [3,22,23] leads to increased generation of PGE2 . Indeed, evidence suggests that certain aspects of brain cholinergic function are diminished with advancing age . In brain, AA and its metabolites are thought to act as intracellular second messengers, suggesting that receptor-dependent potentiation of AA release participates in neuronal signaling .
The decrease in cPLA2 and COX-2 protein levels that we observe in this study becomes significant in the middle aged group. This raises the intriguing possibility that age-related synaptic dysfunction may start to occur at an earlier age than previously considered. In this regard, the number of nigral neurons immunoreactive for nuclear receptor-related factor 1 (Nurr1), involved in dopaminergic signaling, has been shown to be reduced in middle-aged and aged individuals relative to young subjects. The loss of Nurr1-immunoreactive neurons was highly correlated with a similar decline in tyrosine hydroxylase-immunoreactive neuron number in middle-aged and in aged subjects . Similarly, a significant decrease in the number of nigral neurons immunoreactive for guanosine triphosphate cyclohydrolase I (GTPCHI), a critical enzyme in catecholamine function, has been reported in middle age and aged cases relative to young cohorts .
cPLA2 and COX-2 enzymes have been shown to be localized at post-synaptic sites in neurons [28,40]. The regional specificity with regards to the decrease in these two proteins with age, and the lack of a decrease in the hippocampus suggests their correlation as synaptic markers with the reported age-related synaptic changes [18,24,48]. Such a relationship in the normal brain warrants further investigation as does the additional role for these enzymes in the pathophysiology of neuroinflammation in some age-related neurodegenerative disorders, such as Alzheimer’s and Parkinson’s disease, where cPLA2 and COX-2 enzymes would also be markers of the pathological neuroinflammatory process.
This research was entirely supported by the Intramural Research Program of the NIH (National Institute on Aging, National Institute of Environmental Health Sciences, and National Institute of Child Health and Human Development). We thank Drs. Stanley Rapoport, Christopher Toscano, and Saba Aid for helpful comments and suggestions.