We have discovered a non-AT1, non-AT2 angiotensin binding site in rodent and human brain membranes, which, based on its pharmacological/biochemical properties and tissue distribution, is different from angiotensin receptors and key proteases processing angiotensins. In this study, the novel angiotensin binding site was localized to a specific brain cell type by using radioligand receptor binding assays. Our results indicate that the novel binding site is expressed in mouse primary cortical neuronal membranes but not in primary cortical astroglial and bEnd.3 brain capillary endothelial cell membranes. Whole-cell binding assays in neurons showed that the binding site faces the outer side of the plasma membrane. Consistent with our previous observations, the novel binding site was unmasked by the sulfhydryl reagent p-chloromercuribenzoate. This effect had a bell-shaped curve and was reversed by reduced glutathione, indicating that the function of the binding site might be regulated by the redox state of the environment. Density of the novel binding site measured by saturation binding assays was significantly increased in neuronal membranes of cells challenged in four in vitro models of cell death (oxygen-glucose deprivation, sodium azide-induced hypoxia, N-methyl-d-aspartate neurotoxicity, and hydrogen peroxide neurotoxicity). In addition, our in vivo data from developing mouse brains showed that the density of the novel angiotensin binding site changes similarly to the pattern of neuronal death in maturating brain. This is the first time that evidence is provided on the association of the novel angiotensin binding site with neuronal death, and future studies directed toward understanding of the functions of this protein are warranted.
The brain renin-angiotensin system (RAS) is one of the independently regulated, local angiotensin systems most known for its role in cardiovascular and hydromineral balance regulation in both health and disease (McKinley et al., 2003; Cuadra et al., 2010). Although substantial progress in understanding these functions of the brain RAS has been made, many questions remain unanswered regarding the precise nature and role of the system (Saavedra, 2005; Karamyan and Speth, 2007a; Phillips and de Oliveira, 2008; Speth and Karamyan, 2008). For example, almost all key components of the brain RAS, including renin and its recently identified receptor angiotensinogen, angiotensin-converting enzyme and its human homolog (ACE2), and angiotensin II (AngII) and its classic receptors, are found not only in brain areas involved in cardiovascular regulation but also in regions playing a role in neurogenesis, plasticity and memory, cognition and analysis of new information, reward and sensation of pleasure, motor coordination, etc. (McKinley et al., 2003; Phillips and de Oliveira, 2008; Lazartigues, 2009). Thus, it is believed that the role of the brain RAS is not associated exclusively with cardiovascular and hydromineral regulation, but a number of other processes/functions are regulated by this system (McKinley et al., 2003; Phillips and de Oliveira, 2008; Lazartigues, 2009).
In the course of radioligand binding studies of brain AngII receptors, we discovered a novel non-AT1, non-AT2 binding site for angiotensins in rat, mouse, and human brain membranes (Karamyan and Speth, 2007b; Karamyan et al., 2008a,b). This binding site is insensitive to blockade by specific type 1 (AT1) and type 2 (AT2) angiotensin receptor antagonists and is present in the brains of mice lacking these and Mas [the hypothesized receptor for angiotensin (1–7)] receptors (Karamyan and Speth, 2007b; Karamyan et al., 2008a). Moreover, the novel angiotensin binding site is not present in rodent tissues abundant with angiotensin receptors, such as liver, adrenal, and kidney (Karamyan and Speth, 2007b; Karamyan et al., 2008a) and is abundantly distributed in the rat brain, including nuclei involved in both cardiovascular and noncardiovascular functions (Karamyan and Speth, 2008). Finally, a unique feature of the novel angiotensin binding site is that it is unmasked (i.e., able to bind angiotensins) in the presence of optimal concentrations of organomercurial sulfhydryl reagents p-chloromercuribenzoate (PCMB) or p-chloromercuri-benzenesulfonate (PCMPS). This effect is reversed by the disulfide-reducing reagents dithiothreitol and 2-mercaptoethanol, indicating involvement of cysteine residues (thiol groups) in unmasking and function of the novel angiotensin binding site (Karamyan and Speth, 2007b). Currently, the identity of this protein is unknown, and we have information only about the molecular size and some biochemical properties of this protein (Karamyan et al., 2010). Efforts are being made toward its identification.
Pharmacological specificity of the novel angiotensin binding site has been studied in detail in both rodent and human brain membranes and was not in the scope of the current study (Karamyan and Speth, 2007b; Karamyan et al., 2008a,b). The purpose of this study was to locate the novel angiotensin binding site to specific brain cell types by using mouse brain cultures, estimate the density of the binding site in hosting cell types, and get more clues on the unmasking mechanisms of this protein. In addition, we provide evidence (in vitro and in developing brain tissue) pointing to an association of the novel angiotensin binding site with neuronal death.
Materials and Methods
Timed pregnant CD-1 female mice (Charles River Laboratories, Inc., Wilmington, MA) were used for the isolation of primary brain cultures. For developmental studies we used CD-1 mouse embryos and in-house-born pups of both genders. All animals were maintained in a 12-h light/dark cycle and fed ad libitum. Animal procedures were carried out using a protocol approved by the Texas Tech University Health Sciences Center Institutional Animal Care and Use Committee.
Mouse primary cortical neurons were isolated and cultured according to Mattson et al. (1995) with some modifications. In brief, cerebral cortices were obtained from E16 embryos (CD-1 mice; Charles River Laboratories) and dissected in Hank's balanced salt solution (HBSS) without Ca2+ and Mg2+ supplemented with 10 μg/ml gentamycin. Dissected pieces of cortices (free of meninges) were digested in trypsin (+deoxyribonuclease in HBSS) for 15 min at room temperature, neutralized with trypsin inhibitor, and washed three times with HBSS. Dissociated cell suspensions were transferred into 100-mm Petri dishes or six-well plates (0.5–0.6 × 104 per cm2 surface area) coated with polyethylenimine and cultured in Neurobasal medium (Invitrogen, Carlsbad, CA) supplemented with 1.3 mM l-glutamine, 25 μg/ml gentamicin, and 2% B27 (Invitrogen) at 37°C in a humidified atmosphere of 5% CO2 in air. After overnight incubation half of the medium was replaced with fresh Neurobasal medium (+supplements), and the cells were cultured for 12 more days without renewal of the medium (also referred to as the old medium group). Unless otherwise mentioned, all experiments used neurons grown in these conditions. Primary cortical neurons isolated by this approach yielded ∼95% pure neuronal cultures determined by immunocytochemical analysis using specific neuronal (anti-neuron-specific β III tubulin antibody; Abcam Inc., Cambridge, MA), astrocyte (anti-glial fibrillary acidic protein antibody; Cell Signaling Technology, Danvers, MA), and nuclear (4′,6-diamidino-2-phenylindole; Invitrogen) markers.
Mouse primary cortical astroglial cells were cultured as described by Keller et al. (1996). In brief, cerebral cortices were obtained from 1-day-old CD-1 mice, dissected, and treated as detailed above for primary neurons. Dissociated cell suspensions were transferred into 75-mm2 cell culture flasks and cultured in Dulbecco's modified Eagle's medium (with high glucose, sodium pyruvate, and l-glutamine) supplemented with 10% fetal bovine serum and penicillin/streptomycin (100 U/0.1 mg per ml, respectively), at 37°C in a humidified atmosphere of 5% CO2 in air. Confluent cultures (∼10 days) were subcultured in 100-mm Petri dishes for another 5 to 7 days and collected upon reaching confluence. The purity of astroglial cultures (>95%) was estimated by immunocytochemical analysis using specific astrocyte (anti-glial fibrillary acidic protein antibody; Cell Signaling Technology), microglial (anti-ionized calcium binding adaptor molecule 1 antibody; Abcam Inc.), and nuclear (4′,6-diamidino-2-phenylindole; Invitrogen) markers.
The bEnd.3 mouse brain capillary endothelial cell line was purchased from the American Type Culture Collection (Manassas, VA) and maintained according to the vendor's recommended protocol. The choice of a cell line over primary culture was determined by very small harvesting volumes of primary brain capillary endothelial cells from mice and the need of large amounts of membranes for receptor binding assays. bEnd.3 cells are one of the few cells commonly used for in vitro studies of blood-brain barrier functions (Huppert et al., 2010).
On the day of collection, all cells were washed twice by phosphate-buffered saline, scraped, and kept frozen in −80°C before their use in radioligand binding experiments.
Radioligand Binding Assays.
125I-Sar1-Ile8-AngII (125I-SI-AngII) was purchased from the Peptide Radioiodination Service Center of the University of Mississippi (University, MS) and American Radiolabeled Chemicals (St. Louis, MO). Receptor binding studies in neuronal membranes were carried out using established procedures (Karamyan and Speth, 2007b; Karamyan et al., 2008a) except that sonication (instead of homogenization) was used for preparation of membranes, and 5,7-diethyl-3,4-dihydro-1-[[2′-(1H-tetrazol-5-yl)[1,1′- biphenyl]-4-yl]methyl]-1,6-naphthyridin-2(1H)-one hydrochloride (ZD7155) (instead of losartan; Tocris Bioscience, Ellisville, MO) was used to block the AT1 angiotensin receptors. Unless otherwise mentioned, 10 μM final assay concentration of PCMB was used in all of our binding studies. The protein concentration was determined by the method of Bradford (1976) using bovine serum albumin as standard. Whole-cell binding assays on neurons were carried out in six-well plates in 50 mM Tris-HCl assay buffer, pH 7.2, containing 140 mM NaCl, 100 μM EDTA, and 100 μM o-phenanthrolin (Demaegdt et al., 2008) in 800-μl final volume. 125I-SI-AngII was used at ∼0.6 nM concentration ± 10 μM AngII for nonspecific binding in the presence of 1 μM ZD7155 (an AT1 angiotensin receptor antagonist), 10 μM 1-[[4-(dimethylamino)-3-methylphenyl]methyl]-5-(diphenylacetyl)-4,5,6,7-tetrahydro-1H-imidazo[4,5-c]pyridine-6-carboxylic acid ditrifluoroacetate (PD123319) (an AT2 angiotensin receptor antagonist), and 10 μM final assay concentration of PCMB. After 1-h incubation and three washes with 1 ml of phosphate-buffered saline, cells were dissolved in 750 μl of 1 M NaOH for 30 min at 24°C and transferred into cell culture tubes. The wells were washed with 250 μl of water and transferred into corresponding cell culture tubes, and radioactivity was measured in a gamma counter (Wizard2 2470; PerkinElmer Life and Analytical Sciences, Waltham, MA).
In the reversal experiments using glutathione (GSH) analogs, neuronal membranes or whole neurons were first pretreated with PCMB, and then after at least 5 min, one of the glutathione analogs and the radioligand were added onto the membranes or cells. The specific binding of the radioligand was estimated in the same way as in the regular binding experiments.
In a set of experiments, primary neurons were treated with the glutathione biosynthesis inhibitor l-buthionine-(S,R)-sulfoximine (150 μM for 24 h), and specific binding of 125I-SI-AngII was estimated in the presence or absence of PCMB in neuronal membranes (in the presence of AT1 and AT2 angiotensin receptor antagonists).
In Vitro Models of Cell Death.
Primary cortical neurons used in this set of experiments were grown in conditions where 1/3 of the Neurobasal medium (+ supplements) was refreshed twice a week. Four well established in vitro models of neuronal death were used in our study. For oxygen-glucose deprivation/reoxygenation (model of ischemia-reperfusion injury; Abbruscato et al., 2004) 12-in vitro-day-old neurons were deprived of oxygen and glucose in Earle's balanced salt solution (EBSS) without glucose in oxygen-free N2/CO2 (95%/5%) atmosphere (hypoxia chamber; Billups-Rothenberg, Del Mar, CA) at 37°C for 3 h. Thereafter, the medium was replaced by Neurobasal medium (+ supplements), and 24 h later the cells were collected for binding assays. NMDA neurotoxicity (model of excitotoxicity; Hewett et al., 2000) was carried out in the following way. In brief, 12-in vitro-day-old neurons were treated with 0.5 mM NMDA in HBSS for 30 min (at 37°C in a humidified atmosphere of 5% CO2 in air), followed by replacement of medium with Neurobasal medium (+ supplements). After 24 h the cells were collected for binding assays. Sodium azide neurotoxicity (model of hypoxic injury; Marino et al., 2007) was induced similarly to NMDA neurotoxicity except that 3 mM sodium azide was used. Hydrogen peroxide neurotoxicity (model of oxidative injury; Whittemore et al., 1995) was performed similarly to NMDA and sodium azide, except that 30 μM hydrogen peroxide was used.
Cell Viability Assays.
Alamar blue (indicator of metabolic activity of cells; White et al., 1996), lactate dehydrogenase (indicator of plasma membrane integrity; Hewett et al., 2000), and MTT (indicator of mitochondrial function) assays were used to estimate viability of neurons challenged in the above-mentioned models. Measurements were performed 24 h after challenging neurons (grown in 12-well plates) in the in vitro models, according to manufacturers' recommended protocols (Alamar blue, BioSource International, Camarillo, CA, Invitrogen; lactate dehydrogenase/cytotoxicity detection kit, Roche Diagnostics, Indianapolis, IN; MTT/in vitro toxicology assay kit, MTT based, Sigma-Aldrich, St. Louis. MO).
Determination of Bmax (fmole of radioligand bound per mg protein), Kd, and IC50 values were carried out by using one-site saturation [Y = Bmax × X/(Kd + X)] and competition [Y = bottom + (top − bottom)/(1 + 10^(X − LogIC50))] binding models of Prism software (GraphPad Software, Inc., San Diego, CA). Values reported were significantly different from zero and are presented as mean ± S.E.M. Ki value for AngII was determined by using the Cheng-Prusoff equation (Cheng and Prusoff, 1973): Ki = IC50/(1 + H/Kd), where H is the radioligand concentration and Kd is the affinity of the radioligand. Kon and Koff values and association and dissociation half-times were calculated by using one-phase association [Y = Y0 + (plateau − Y0) × (1 − exp(−K × X))] and dissociation [Y = (Y0 − plateau) × exp(−K × X) + plateau] models of Prism software. Comparison of Bmax and Kd values obtained from binding assays in membrane preparations of neurons challenged in in vitro models of cell death and in developing brains was carried out by using one-way analysis of variance followed by the Tukey-Kramer test. The results of cell viability assays were compared in the same way. Values reported are mean ± S.E.M.
The expression of the novel angiotensin binding site in membrane preparations of mouse primary cortical neurons was tested by radioligand receptor binding assays using 125I-SI-AngII in the presence of specific AT1 and AT2 angiotensin receptor antagonists ZD7155 and PD123319, respectively. First, a 150 μM final assay concentration of PCMB (similar to our previous studies in rodent and human brain membranes) was applied to unmask the binding site in neuronal membranes (Fig. 1a). Next, the optimal range of PCMB concentrations for unmasking of the binding site in primary neuronal membranes was determined. Specific binding of 125I-SI-AngII (10 μM AngII displaceable) as a function of PCMB concentrations had a bell shape, with optimal concentrations of PCMB ranging from 10 to 100 μM (Fig. 1b). Based on these results, 10 μM PCMB was used in all our subsequent experiments. Under the same experimental conditions, unlabeled AngII showed ∼100 nM affinity (Ki value) for the novel angiotensin binding site in neuronal membranes (Fig. 1c). Therefore, 10 μM final concentration of AngII was used to displace 125I-SI-AngII and determine its nonspecific binding in subsequent binding experiments.
To properly identify the specific radioligand binding and estimate the density of the novel binding site, correlation of the membrane/receptor concentration with the extent of specific binding of the radioligand was determined (Hoffman and Lefkowitz, 1980). Specific binding of 125I-SI-AngII (10 μM AngII displaceable) to the novel angiotensin binding site was linear to the neuronal membrane protein concentrations from 50 to 300 μg/ml in the assay medium (data not shown). For this reason, 200 to 300 μg/ml neuronal membranes was used in our future experiments.
Association rate of 125I-SI-AngII (∼0.6 nM) to the novel angiotensin binding site in neuronal membranes displayed a half-time of ∼15 min at 24°C, reaching a steady state at ∼90 min (Fig. 2a). The dissociation half-time of 125I-SI-AngII from the binding site was ∼76 min at 24°C (Fig. 2b, 68%) of binding being nondisplaceable at 180 min, indicating a possible pseudo-irreversible binding. It is noteworthy that the kinetic dissociation constant calculated from these experiments (0.2 nM) was similar to the equilibrium dissociation constant obtained from saturation binding studies (Fig. 3).
The Bmax value (128 ± 6 fmol/mg protein) for the novel angiotensin binding site in primary cortical neuronal membranes was calculated from saturation binding experiments conducted at steady-state conditions (2-h incubation at 24°C), based on the results of time course association experiments (Fig. 2a). The equilibrium dissociation constant (Kd) for 125I-SI-AngII was 0.65 ± 0.1 nM. There was no significant saturable binding of the radioligand to the classic AT1 and AT2 angiotensin receptors in these neuronal membranes (data not shown).
Saturation binding experiments in membrane preparations obtained from mouse primary cortical astroglial cells and the bEnd.3 mouse brain capillary endothelial cell line did not show any significant saturable binding of the radioligand under the same experimental conditions (Fig. 3).
High-affinity binding of 125I-SI-AngII to the novel angiotensin binding site was also observed in intact primary cortical neurons using both PCMB and its sulfonic acid derivative PCMPS (Fig. 4). Similar to membrane preparations, there was no binding to the classic AT1 and AT2 angiotensin receptors in these cells (Fig. 4).
Figure 5 summarizes the effects of reduced and oxidized glutathione and their structural analog γGlu-Abu-Gly (cysteine residue is replaced with 2-aminobutyric acid, Abu; Phoenix Pharmaceuticals, Belmont, CA) on the unmasking effect of PCMB in primary neuronal membranes (Fig. 5a) and whole neurons (Fig. 5b).
Depletion of intracellular levels of reduced glutathione by its biosynthesis inhibitor l-buthionine-(S,R)-sulfoximine in neurons did not unmask the novel binding site in neuronal membranes, and the presence of PCMB was still required for high-affinity binding of 125I-SI-AngII (data not shown).
The density of the non-AT1, non-AT2 angiotensin binding site was also determined in membrane preparations obtained from primary cortical neurons (in the presence of 1 μM ZD7155, 10 μM PD123319, and 10 μM PCMB) challenged in four well established in vitro models of neuronal death: oxygen-glucose deprivation/reoxygenation (model of ischemia-reperfusion injury; Abbruscato et al., 2004), NMDA neurotoxicity (model of excitotoxicity; Hewett et al., 2000), sodium azide neurotoxicity (model of hypoxic injury; Marino et al., 2007), and hydrogen peroxide neurotoxicity (model of oxidative injury; Whittemore et al., 1995). As presented in Fig. 6, the density (Bmax value) of the novel binding site was significantly increased in neuronal membranes of cells challenged in all four in vitro models. In addition, the density of the novel angiotensin binding site was significantly increased in neurons from the old medium group (Fig. 6), where the medium was not refreshed throughout the culturing period (see Materials and Methods). It is noteworthy that the averaged Kd values did not differ significantly among the groups, and the neurotoxic insults alone did not unmask the binding site in neuronal membranes (the presence of PCMB was required for unmasking).
Viability of neurons was measured 24 h after challenging the cells in the above-mentioned models by using lactate dehydrogenase, Alamar blue, and MTT assays. All assays indicated significantly decreased viability of neurons in all experimental groups (except the old medium group in the lactate dehydrogenase assay) compared with the fresh medium group (Fig. 7).
Saturation binding experiments in membrane preparations of mouse forebrains (in the presence of 1 μM ZD7155, 10 μM PD123319, and 150 μM PCMB), obtained at different stages of development, indicated that the density (Bmax value) of the novel angiotensin binding site gradually increases from E14 to P10, followed by a significant drop at P21 and a maintained level in 9- to 12-week-old animals (Fig. 8). The averaged Kd values did not differ significantly among the groups.
The initial purpose of this study was to locate the novel non-AT1, non-AT2 angiotensin binding site to specific cell types by using mouse brain cultures and radioligand receptor binding assays. Consistent with previous observations reported in rodent and human brain membranes, the 150 μM final assay concentration of PCMB unmasked the binding site in neuronal membranes (Fig. 1a). Concentration-effect curve for unmasking had a bell shape, with optimal concentrations of PCMB ranging from 10 to 100 μM (Fig. 1b). It is noteworthy that unmasking of the binding site (i.e., ability to bind angiotensins) by sulfhydryl reagents is one of the unique features of this protein (Karamyan and Speth, 2007b). Organomercurials and particularly PCMB are the most specific sulfhydryl agents forming an easily reversible mercaptide bond with the thiol group (Rothstein, 1970). A probable explanation for our observations is that unmasking of the binding site is associated with interaction of PCMB with the thiol group of cysteines in this protein (Karamyan and Speth, 2007b). In other words, the state of reduced (−SH) and oxidized (−SR) thiols in the binding site probably determines the conformation at which this protein binds angiotensins with high affinity. Furthermore, the bell-shaped concentration-effect curve of PCMB indicates fine regulation of this process (Vyas et al., 2002), because only a certain number of thiols need to be oxidized/modified for the binding site to be efficiently unmasked and able to bind angiotensins.
Saturation binding experiments at steady-state conditions indicated high-affinity, saturable binding of 125I-SI-AngII to the binding site in neuronal membranes but not in membrane preparations of astroglial and bEnd.3 cells (Fig. 3), suggesting that this protein is expressed mainly in neurons.
The presence of the binding site was confirmed in intact neurons by using both PCMB and PCMPS, indicating that the protein is located on the plasma membrane and is accessible for its ligands in an extracellular environment (Fig. 4). Considering that PCMPS is a cell-impermeable analog (Rothstein, 1970; Vyas et al., 2002), these data also indicate that the thiol group involved in unmasking is facing the cell surface.
Next, we performed experiments using GSH analogs to further confirm the involvement of thiol groups in the unmasking of the binding site. Addition of reduced GSH to neuronal membranes or intact neurons preincubated with PCMB prevented the unmasking with IC50 values of 5 to 15 μM (Fig. 5). It is noteworthy that oxidized glutathione (GSSG) and γGlu-Abu-Gly (both lack the thiol group) did not reverse the unmasking effect of PCMB (Fig. 5). These results confirm that unmasking of the binding site is specifically associated with modification of thiol groups in this protein. Considering that reduced GSH does not cross the plasma membrane (Skalska et al., 2009), these results further suggest that the thiol group responsible for unmasking faces the cell surface. In addition, these results indicate that the process of unmasking is reversible and may be regulated upon changes in extracellular levels of GSH in the brain. In other words, physiological/high extracellular concentrations of GSH favor the masked state of the binding site, whereas decreased levels of GSH (e.g., after oxidative stress) favor its unmasking and function. It is noteworthy that extracellular/cerebrospinal fluid concentrations of GSH in normal/healthy brain are in the range of 5 to 15 μM (opposite to millimolar intracellular concentrations), which change at different adaptive/pathophysiological conditions (Yang et al., 1994; Lada and Kennedy, 1997). To further study the unmasking mechanism of the binding site, we used the glutathione biosynthesis inhibitor l-buthionine-(S,R)-sulfoximine to deplete intracellular levels of reduced glutathione in neurons. This treatment did not unmask the novel binding site in neuronal membranes, and the presence of PCMB was still required for high-affinity binding of 125I-SI-AngII.
It is important to mention that reversible modification (oxidation/reduction) of the thiol group of cysteines (also known as reactive cysteines/thiols, cysteine/redox switches) is increasingly appreciated as an essential posttranslational modification of proteins (Janssen-Heininger et al., 2008; Jones, 2008). Our results suggest that the novel angiotensin binding site is likely to be a similar redox-sensitive protein located on the plasma membrane of neurons, which depending on the redox state of the extracellular environment in the brain may be unmasked to interact with angiotensins.
The ubiquitous role of oxidative stress in cell death and redox sensitivity of numerous proteins involved in this process is well documented (Calabrese et al., 2010). To further investigate a possible association of oxidative stress with the binding site, we carried out experiments to estimate density of this protein in membrane preparations of dying neurons. Neurons challenged in four in vitro models of cell death and grown in conditions with limited availability of nutrients (old medium) showed significantly increased density of the binding site (Fig. 6). It is noteworthy that the neurotoxic insults did not unmask the binding site in neuronal membranes, and the presence of PCMB was required for unmasking. Parallel cell viability assays indicated that the viability of neurons was significantly decreased in all groups compared with the fresh medium group (Fig. 7). It is noteworthy that these data suggest that any stressor reaching the threshold to initiate cell death probably will result in up-regulation of the binding protein. Similar Bmax values between the experimental groups indicate that up-regulation of this protein is a regulated process with an upper limit independent of the potency of a stressor.
Finally, we determined the density of the binding site in developing mouse brains. Our results indicated a gradual increase in Bmax values in mouse forebrain membranes from E14 to P10, followed by a dramatic drop in P21 animals and similar levels in 9- to 12-week-old animals (Fig. 8). It is noteworthy that the pattern of developmental changes in the density of the binding site (Fig. 8) is very similar to the occurrence of neuronal death in developing brain (Ferrer et al., 1992; Naruse and Keino, 1995). The number of dead neurons in rat primary visual cortex is low at birth, which increases from P2, peaking at the end of the first week, and decreases during the second week followed by low numbers at the end of the first month (Ferrer et al., 1992; Naruse and Keino, 1995). Clearly, these in vivo observations complement our in vitro data and support the association between neuronal cell death and the binding site. Moreover, these results clarify our previous observations in adult rat brain where distribution of the binding site was studied by in vitro receptor autoradiography (Karamyan and Speth, 2008). In the latter study, among brain regions with the highest radioligand binding were the olfactory bulb (highest compared with other brain regions), ventricle wall (throughout brain), and dentate gyrus. It is noteworthy that it was shown that the frequency of apoptosis in adult rat brain is up to 100 times higher in olfactory bulb, ventricle wall, and dentate gyrus (in decreasing order) compared with other brain areas (Biebl et al., 2000; Kuhn et al., 2005).
Our results provide the first evidence of association of the novel angiotensin binding site with a (patho)physiological process, suggesting a potential role of this protein in neuronal cell death. Although the exact function of the binding site is unknown, many features of this protein, e.g., high affinity and pharmacological specificity for physiologically relevant ligands, preservation in a number of species including humans, accessibility of ligands to the binding site from extracellular environment, and direct association of this protein with neuronal death, strongly favor the functional significance of this protein. This proposal is also supported by one of the primary concepts of functional genomics postulating that nonfunctional receptors/binding sites are evolutionarily discarded (Civelli, 1998).
It is noteworthy that the link between AngII and oxidative processes is well documented. Numerous studies have established that reactive oxygen species are products of AngII signaling via the AT1 receptor (Griendling et al., 1994; Zimmerman et al., 2004). A direct role of redox-sensitive molecules in signaling pathways downstream of the AT1 receptor (Tabet et al., 2008) and in negative regulation of AT1 receptor-mediated effects (Harrison and Sumners, 2009) have also been demonstrated. In addition, the modulatory role of the AT1 and AT2 receptors in cell proliferation, differentiation, repair, and apoptosis has been studied and debated (Kaschina and Unger, 2003).
In summary, our study provides a number of important findings about the non-AT1, non-AT2 angiotensin binding site, which is expressed primarily in neurons facing the outer side of the plasma membrane. The unique pattern of unmasking of this binding site by specific sulfhydryl reagents and reversal of this effect by reduced GSH suggest that the function of this protein probably depends on the redox state of the extracellular environment in the brain. In other words, it is hypothesized that the state of reduced and oxidized thiol groups in this protein determines its proper conformation for high-affinity binding of angiotensins. Finally, evidence is provided for the first time (in vitro and in developing brain) to support a potential role of the binding site in neuronal death.
Two important questions that are not clear from our study involve the direct function of the binding site in neuronal death and the mechanisms of unmasking of this protein in (patho)physiological conditions without use of exogenous sulfhydryl compounds. These questions warrant future studies and should provide insights into the way in which the novel binding site affects brain angiotensinergic activity and plays a role in neuronal cell death.
We thank Naomi Wangler for help with primary brain cultures.
This work was supported by Texas Tech University Health Sciences Center School of Pharmacy start-up funds (to V.T.K.).
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
- renin-angiotensin system
- angiotensin II
- Sar1-Ile8 angiotensin II
- 5,7-diethyl-3,4-dihydro-1-[[2′-(1H-tetrazol-5-yl)[1,1′-biphenyl]-4-yl]methyl]-1,6-naphthyridin-2(1H)-one hydrochloride
- 1-[[4-(dimethylamino)-3-methylphenyl]methyl]-5-(diphenylacetyl)-4,5,6,7-tetrahydro-1H-imidazo[4,5-c]pyridine-6-carboxylic acid ditrifluoroacetate
- 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide
- oxidized glutathione
- Hanks' buffered salt solution
- Earle's balanced salt solution
- embryonic day n
- postnatal day n
- lactate dehydrogenase.
- Received June 12, 2010.
- Accepted September 21, 2010.
- Copyright © 2010 by The American Society for Pharmacology and Experimental Therapeutics