The intracellular angiotensin system buffersdeleterious effects of the extracellular paracrine system

OPEN The intracellular angiotensin system buffers deleterious effects of the extracellular paracrine system Begoña Villar-Cheda 1,2,5 , Maria A Costa-Besada 1,2,5 , Rita Valenzuela 1,2 , Emma Perez-Costas 3 , Miguel Melendez-Ferro 4 and Jose L Labandeira-Garcia* ,1,2 The ‘classical’renin–angiotensin system (RAS) is a circulating system that controls blood pressure. Local/paracrine RAS, identified in a variety of tissues, including the brain, is involved in different functions and diseases, and RAS blockers are commonly used in clinical practice. A third type of RAS (intracellular/intracrine RAS) has been observed in some types of cells, including neurons. However, its role is still unknown. The present results indicate that in brain cells the intracellular RAS counteracts the intracellular superoxide/H 2 O 2 and oxidative stress induced by the extracellular/paracrine angiotensin II acting on plasma membrane receptors. Activation of nuclear receptors by intracellular or internalized angiotensin triggers a number of mechanisms that protect the cell, such as an increase in the levels of protective angiotensin type 2 receptors, intracellular angiotensin, PGC-1αand IGF-1/SIRT1. Interestingly, this protective mechanism is altered in isolated nuclei from brains of aged animals. The present results indicate that at least in the brain, AT1 receptor blockers acting only on the extracellular or paracrine RAS may offer better protection of cells. Cell Death and Disease (2017) 8, e3044; doi:10.1038/cddis.2017.439; published online 7 September 2017 The ‘classical’renin–angiotensin system (RAS) is a circulating humoral system that controls blood pressure. The actions of angiotensin II (AII), the most important effector peptide in RAS, are mediated by two main cell receptors: AII type 1 and 2 (AT1 and AT2). AT1 receptors mediate major effects of the system, and it is generally considered that protective AT2 receptors antagonize the pro-oxidative effects of AT1 receptors. 1 More recently, local or tissue RAS has been identified in a variety of tissues, including the central nervous system. 2 It is now known that the local brain RAS is involved in different brain functions, and also appears to be altered in some disorders. 3,4 In previous studies, we have demonstrated the presence of a local RAS in the substantia nigra pars compacta (SNc) and striatum of rodents and primates, including humans. 5–7 This local RAS modulates dopamine release 8,9 possibly via mutual regulation between dopamine and angiotensin receptors. 10–12 However, dysregulation of these interactions exacerbates neuroinflammation, oxidative stress and dopaminergic neuron death. 13,14 In addition, immunohistochemical studies have revealed an apparent intracellular localization of several RAS components in different types of cells, including dopaminergic neurons and glial cells of mammals, including non-human primates and human. 5,15,16 However, the role of the intracellular RAS, and particularly the nuclear components of the RAS, is still unknown. In the present study we investigated the presence and possible role of major RAS components in brain cell nuclei. In particular, we investigated the possible effects of nuclear angiotensin receptors in the transcription of other components of the intracellular RAS and of several proteins that act as major regulators of the mitochondrial function. Our experiments were carried out in rats, knockout (KO) and transgenic mice, as well as in the MES 23.5 dopaminergic neuron cell line. A major difficulty in elucidating the role of intracellular RAS is to separate the responses induced by intracellular AII through intracellular or nuclear receptors from those induced by extracellular AII through activation of cell surface receptors. To overcome this technical difficulty, we investigated the effect of AII both in cells and in isolated nuclei. Our results show that nuclear angiotensin receptors control key events for nuclear–mitochondrial interaction and neuronal survival. Results Localization of AT1 and AT2 receptors in the nucleus of dopaminergic neurons. Our previous immunohistochemical and laser confocal microscopy studies have revealed immunolabeling for AT1, AT2 receptors and angiotensinogen in the nuclei of SNc dopaminergic neurons and MES 23.5 dopaminergic neuron cell line. 5 In the MES 23.5 dopaminergic neuron cell line, labeling for AT1 and AT2 receptors also colocalized with the nuclear marker Hoechst 33342 (Figures 1a and b). Samples for electron microscopy were 1 Laboratory of Neuroanatomy and Experimental Neurology, Department of Morphological Sciences, Faculty of Medicine, CIMUS, University of Santiago de Compostela, Santiago de Compostela, Spain; 2 Networking Research Center on Neurodegenerative Diseases (CIBERNED), Spain; 3 Division of Pediatric Nephrology, Department of Pediatrics, University of Alabama at Birmingham, Birmingham, AL, USA and 4 Division of Pediatric Surgery, Department of Surgery, University of Alabama at Birmingham, Birmingham, AL, USA *Corresponding author: JL Labandeira-Garcia, Laboratory of Neuroanatomy and Experimental Neurology, Department of Morphological Sciences, Faculty of Medicine, CIMUS, University of Santiago de Compostela, Santiago de Compostela 15782, Spain. Tel: +34 881812223; Fax: +34 881812378; E-mail: [email protected] 5 These authors contributed equally to this work. Received 15.3.17; revised 25.7.17; accepted 31.7.17; Edited by A Verkhratsky Citation: Cell Death and Disease (2017) 8, e3044; doi:10.1038/cddis.2017.439 Official journal of the Cell Death Differentiation Association www.nature.com/cddis obtained from the densely packed dopaminergic cell clusters of the rat SNc, where immunoelectron microscopy confirmed the labeling for both AT1 and AT2 receptors in the nuclear membranes of dopaminergic neurons (Figures 1c and d). Nuclei were isolated from the rat nigral region in the ventral mesencephalon (Figures 1e and f), and the quality of the isolation was demonstrated by the absence of markers of the cytosol fraction such as α-tubulin, and membrane markers Figure 1 Nuclear AT1, AT2 receptors and NOX4 in nuclei from MES 23.5 dopaminergic neurons and rat nigral region. (aand b) MES 23.5 dopaminergic neurons showing triple immunolabeling for the dopaminergic marker (TH), the nuclear marker (Hoechst), and AT1 (a) or AT2 (b) receptors. (cand d) Electron microscopy of AT1 and AT2 labeling (white arrows) in nuclei and nuclear membranes (between black arrowheads) of rat dopaminergic neurons. (eand f) Nuclei isolated from the rat nigral region in the ventral mesencephalon; the integrity of nuclei was confirmed by microscopic examination with phase contrast (e) and Hoechst staining (f); areas boxed in (eand f) are magnified in (e1 and f1), respectively. (g) WB of whole homogenate (WH) and isolated nuclei (N) from the nigral region showing the expression of AT1 and AT2 receptors, Nox4, as well as different compartment markers used to assess the purity of the nuclei isolation (HDAC2 as a nuclear marker; tubulin as a cytosol marker; and Na + /K + -ATPase as plasma membrane marker). Note the higher expression of AT1 and lower expression of AT2 in the nucleus compared to receptors in WH (gand h). Data are mean ±S.E.M. *Po0.05 compared to WH. Student’st-test (n=3–4). Scale bars: 150 μm(aand b), 0.5 μm(cand d) and 50 μm(eand f) Nuclear angiotensin receptors B Villar-Cheda et al 2 Cell Death and Disease such as Na + /K + -ATPase. In addition, the presence of the nuclear fraction was confirmed with the use of the nuclear marker histone deacetylase 2 (HDAC2; Figure 1g). The two main angiotensin receptor types (i.e., AT1 and AT2) were observed in the nuclear fraction. There was a clear difference in the abundance of each receptor subtype; the AT2 receptor appeared less abundant in nuclear than in total cell homogenates, while the opposite occurred for the AT1 receptor (Figures 1g and h). We also observed the presence of Nox4, which is a major cell source of intracellular ROS in the nuclear fraction (Figures 1g and h). Effects of AII on nuclear RAS components. Twenty-four hours after transient transfection of the MES 23.5 dopaminergic neuron cell line, AT1 and AT2 receptors labeled with fluorescent proteins (enhanced green fluorescent protein, Figure 2 Presence of fluorescence-tagged angiotensin receptors and fluorescent AII in nuclei. Colocalization (c,fand i) of the fluorescent nuclear marker Hoechst (a,dand g) with AT1-EGFP (b), AT2-YFP (e) or AII-Rhod (h). WB analysis of GFP-/YFP-tagged protein in MES 23.5 transfected (+) and not transfected ( −) cells showing the specificity of the common anti-GFP/YFP antibody (j). Treatment of MES 23.5 dopaminergic neuron cell line with AII increased nuclear AT1-EGFP receptor protein but not nuclear AT2-YFP receptor protein relative to nuclei of control cells (kand l). Data are mean ±S.E.M. *Po0.05 compared to control. Student’st-test (n=3–4). AT2-YFP, AT2 tagged to YFP; AT1- EGFP, AT1 tagged to EGFP; AII-Rhod, rhodamine-fluorescent AII. Scale bar: 20 μm(a–i) Nuclear angiotensin receptors B Villar-Cheda et al 3 Cell Death and Disease EGFP; yellow fluorescent protein, YFP) were located both at the plasma membrane and intracellularly, and they colocalized with the nuclear marker Hoechst 33342 (Figures 2a–f). In a second series of experiments, cultures of the MES 23.5 dopaminergic neuron cell line were treated with rhodamineconjugated AII to investigate the localization of AII within the nucleus. The fluorescent AII was internalized and colocalized with the nuclear marker Hoechst 33342, which was evident after 30 min, and persisted for 24 h after treatment (Figures 2g–i). Interestingly, treatment of the MES 23.5 dopaminergic neuron cell line with 100 nM AII increased nuclear AT1- EGFP fluorescence in these cells relative to nuclei from control cells. However, the same treatment did not induce any significant increase in the levels of nuclear AT2-YFP fluorescence relative to control cells. These effects were confirmed by western blot (WB) in samples from isolated nuclei (Figures 2j–l). RT-PCR analysis of brain-isolated nuclei revealed that treatment with 100 nM AII induces a significant increase in angiotensinogen and AT2 receptor mRNA expression (Figures 3a and b), as well as an increase in the expression of mRNA for renin and prorenin/renin receptors (PRRs; Figures 3c and d). In contrast, the levels of AT1 receptor mRNA did not change significantly (Figure 3e). The increases in mRNA expression induced by All for the abovementioned RAS components were inhibited by simultaneous treatment with the AT1 receptor antagonist losartan, which indicates that these effects are mediated via nuclear AT1 receptors. This was confirmed by treatment of isolated nuclei from AT1 and AT2 KO mice. AII induced the abovementioned effects in AT2 KO, but not in AT1 KO mice (Figures 3f and g). The effects of AII on brain-isolated nuclei were confirmed in nuclei from the MES 23.5 dopaminergic neuron cell line, in which AII treatment also increased mRNA expression for AT2 receptors and angiotensinogen, but not mRNA expression for AT1 receptors (Figure 3h). Effects of AII on the transcription of PGC-1α, IGF-1 and SIRT1 in isolated nuclei. Treatment of isolated nuclei with 100 nM AII led to an ~ 60% increase in mRNA expression for PGC-1αand a twofold increase in mRNA expression for IGF-1. These increases were inhibited by the simultaneous treatment with the AT1 receptor blocker losartan, suggesting that AII acts via nuclear AT1 receptors (Figures 4a and c). When nuclei isolated from KO AT1 and KO AT2 mice were treated with All the effect was observed in nuclei from KO AT2 mice, but was absent in nuclei from KO AT1 mice (Figures 4b–d). Isolated nuclei were also treated with AII to assess its effects on mRNA expression for SIRT1. However, the increase in SIRT1 mRNA induced by All was not significant (Figure 4e). To further investigate a possible interaction between SIRT1 and the nuclear RAS we investigated the expression of AT1 and AT2 receptors in nuclei isolated from transgenic mice overexpressing SIRT1. Interestingly, these mice showed a significant decrease in the levels of nuclear AT1 receptors, but a significant increase in the levels of nuclear AT2 receptors (Figure 4f). The results observed in nuclei isolated from the brain nigral region were also observed in nuclei isolated from the MES 23.5 dopaminergic neuron cell line, where AII treatment also produced an increased mRNA expression for PGC-1αand IGF-1, but a nonsignificant increase in mRNA expression for SIRT1 (Figure 4g). Effects of AII on superoxide/H 2 O 2 , calcium and nitric oxide levels in isolated nuclei. Treatment of brain-isolated nuclei with AII induced a significant increase in the levels of nuclear superoxide/H 2 O 2 , which was inhibited by simultaneous treatment with the AT1 receptor blocker losartan, but not by the simultaneous treatment with the AT2 receptor blocker PD123,319 (Figure 5a). The increase in superoxide/ H 2 O 2 induced by All was also inhibited by the simultaneous treatment with the antioxidant N-acetyl-cysteine (NAC) or the simultaneous treatment with the Nox4 inhibitor (diphenyleneiodonium, DPI; Figure 5b). Altogether these results suggest that at the nuclear level, AII activates the AT1/Nox4 axis to induce the release of superoxide/H 2 O 2 . The role of AT1 and AT2 receptors in AII-induced nuclear superoxide/ H 2 O 2 release was confirmed in isolated nuclei from AT1 and AT2 KO mice (Figure 5c). Levels of superoxide/H 2 O 2 were significantly lower in nuclei from KO AT1 mice than from wildtype (WT) mice. Nuclei from KO AT2 mice showed significantly higher levels of superoxide/H 2 O 2 than nuclei from WT mice. Treatment of nuclei from KO AT1 mice with AII induced a significant additional decrease in the levels of nuclear superoxide/H 2 O 2 (possibly as a result of the effect of AII on nuclear AT2 receptors). Treatment of isolated nuclei from KO AT2 mice with AII induced an additional increase in the levels of nuclear superoxide/H 2 O 2 . Altogether these results reveal an opposite effect of AT1 (marked increase), and AT2 (slight decrease) nuclear receptors on the levels of nuclear superoxide/H 2 O 2 . The observations in brain nuclei were confirmed in nuclei isolated from the MES 23.5 dopaminergic neuron cell line, in which AII treatment also induced a significant increase in the levels of nuclear superoxide/H 2 O 2 . This effect was inhibited by the simultaneous treatment with the AT1 receptor blocker losartan or the Nox4 inhibitor DPI, but not by the simultaneous treatment with the AT2 receptor antagonist PD123,319 (Figure 5d). The AII-induced increase in AT2 receptor or PGC-1α mRNAs (Figures 3 and 4) was not blocked by the simultaneous treatment with the antioxidant NAC or the Nox inhibitor DPI, at least in the present experimental conditions. This suggests that the AII-induced increase in nuclear levels of superoxide/H 2 O 2 does not mediate the abovementioned effects on transcription (Figures 5e and f). Treatment of isolated nuclei with AII led to increased levels of nuclear calcium, which were inhibited by treatment with the AT1 receptor blocker losartan, indicating that the effect is mediated by AT1 receptors (Figure 6a). Interestingly, the simultaneous treatment of isolated nuclei with the IP3 receptor inhibitor 2-aminoethyl diphenylborinate (2-APB) inhibited the AII-induced increase in AT2 and PGC-1αmRNAs, suggesting that these effects are mediated by Ca 2+ via nuclear IP3·receptors (Figures 6b and c). Treatment of isolated nuclei with AII and the AT1 blocker losartan induced a significant increase in nitric oxide (NO) levels, which was inhibited by simultaneous treatment with the Nuclear angiotensin receptors B Villar-Cheda et al 4 Cell Death and Disease AT2 receptor antagonist PD123,319 (Figure 6d). This indicates that nuclear AT2 receptors mediate the AII-induced production of nuclear NO. The AII-induced increase in nuclear NO was also inhibited by simultaneous treatment with the NO synthase (NOS) inhibitor, N-nitro-L-arginine methyl ester (L- NAME), which indicates that the nuclear NO synthase involved in this process. Effects of AII on isolated nuclei from aged rats. Nuclei isolated from brains of aged rats showed a significant decrease in the levels of both AT1 and AT2 receptors (Figure 7a). Interestingly, treatment of these nuclei with 100 nM AII did not induce any significant increase in mRNA expression for IGF-1, PGC-1αor AT2 receptors (Figure 7b). Similarly, the AII-induced increase in the levels of superoxide/ Figure 3 Effect of angiotensin (AII) on transcription of different RAS components. Treatment of rat-isolated nuclei with AII induced an increase in the expression of mRNA for AT2 (a), angiotensinogen (ANG; b), renin (c) and PRRs (d), which was inhibited by simultaneous treatment with the AT1 receptor antagonist losartan (los), indicating that these effects are mediated via nuclear AT1 receptors. In contrast, the levels of AT1 receptor mRNA did not change significantly after AII administration (e). The role of AT1 receptors in these effects was confirmed in isolated nuclei from AT1 and AT2 KO mice treated with AII (fand g). The effects of AII on brain-isolated nuclei were also observed in nuclei from the MES 23.5 dopaminergic neuron cell line (h). Data are mean ±S.E.M. *Po0.05 compared to control, # Po0.05 compared to the group treated with AII. One-way analysis of variance and Holm–Sidak post hoc test (a–e) and Student’st-test (f–h)(n=4–8) Nuclear angiotensin receptors B Villar-Cheda et al 5 Cell Death and Disease H 2 O 2 was not statistically significant in these nuclei (Figure 7c). As observed in nuclei from young KO AT1 mice (Figure 5c), the levels of nuclear superoxide/H 2 O 2 were significantly lower in aged KO AT1 mice than in WT aged mice. Treatment with AII did not induce any significant increase in the levels of superoxide/H 2 O 2 in nuclei from aged KO AT1 mice, as previously observed in young KO AT1 mice (Figure 7d). Discussion We report here the presence of AT1 and AT2 receptors as well as Nox4 in isolated nuclei from brain cells. AT1 receptors were more abundantly expressed in isolated nuclei than in the total cell homogenate, while the opposite occurred for AT2 receptors. Interestingly, we recently observed the opposite in isolated mitochondria, where AT2 are much more abundant than AT1 receptors. 17 Treatment of the MES 23.5 dopaminergic neuron cell line with AII induced an increase in the expression of nuclear AII and AT1 receptors. In contrast, treatment of isolated nuclei with AII induced, via AT1 activation, an increase in AT2 mRNA but not AT1 mRNA transcription. The presence of AT1 receptors in the nucleus has been observed in several types of peripheral cells, although the function of nuclear angiotensin receptors was Figure 4 Effect of angiotensin (AII) on transcription of PGC-1α, IGF-1 and SIRT1. Treatment of isolated nuclei with AII induced an increase in the expression of mRNA for PGC-1α(a) and IGF-1 (c), which was inhibited by the simultaneous treatment with the AT1 receptor antagonist losartan (los), indicating that these increases are mediated via nuclear AT1 receptors. This was confirmed by treatment of isolated nuclei from AT1 and AT2 KO mice with AII (band d). The expression of SIRT1 mRNA did not change significantly after treatment of isolated nuclei with AII (e). However, nuclei isolated from mice overexpressing SIRT1 (SIRT1-TG) showed a decrease in the expression of nuclear AT1, and an increase in the levels of AT2 receptors (f). The effects of AII on brain-isolated nuclei were also observed in nuclei from the MES 23.5 dopaminergic neuron cell line (g). Data are mean ±S.E.M. *Po0.05 compared to control or WT (f), # Po0.05 compared to the group treated with AII. One-way analysis of variance and Holm–Sidak post hoc test (a,cand e), and Student’st-test (b,d,fand g)(n=4–8) Nuclear angiotensin receptors B Villar-Cheda et al 6 Cell Death and Disease not clarified. 18,19 In renal cells, a rapid internalization of the AII–AT1 receptor complex, via receptor-mediated endocytosis, has been reported. 20–22 AT2 receptors lack a canonical nuclear localization sequence as AT1 receptors do, and generally they are not internalized following ligand activation in vascular or renal cells. 23,24 It has been suggested that the intracellular RAS may serve to amplify events that are initiated in the plasma membrane, leading to signal amplification in successive steps with reutilization and minimal waste. 18,25 Quite the opposite, the results of the present study suggest that the intracellular, and particularly the nuclear RAS, may counteract and modulate the deleterious effects of the membrane AII/AT1 activation (Figure 8). Extracellular (i.e., paracrine) AII induces activation of the membrane AT1–Nox2 axis, with the generation of intracellular superoxide/H 2 O 2 and oxidative stress, but also induces the internalization of the AII– AT1 receptor complex to the nucleus. We observed that activation of nuclear AT1 receptors by intracellular AII triggers a number of mechanisms that may protect cells against oxidative stress. These mechanisms include an increase in the levels of AT2 receptors and intracellular angiotensin, as well as of PGC-1αand IGF-1. Interestingly, this protective mechanism is altered in isolated nuclei from brains of aged animals. Figure 5 Effect of angiotensin (AII) on nuclear superoxide/H 2 O 2 production, AT2 and PGC-1αmRNA expression. Treatment of isolated nuclei with AII increased the levels of nuclear superoxide/H 2 O 2 (i.e., DHE fluorescence; a–d), which were inhibited by the simultaneous treatment with the AT1 receptor antagonist losartan (a; los), the antioxidant NAC and the Nox inhibitor DPI (b), but not by the AT2 receptor antagonist PD123,319 (PD) (a). Involvement of AT1 in these effects was confirmed by treatment of nuclei from AT1 and AT2 KO mice with AII, which also revealed an opposite effect of nuclear AT1 (increase) and AT2 (decrease) receptors on the levels of nuclear superoxide/H 2 O 2 (c). The effects of AII on brain nuclei were also observed in nuclei from the MES 23.5 dopaminergic neuron cell line (d). However, the effects of AII on AT2 (e) and PGC-1α(f) mRNA expression were not inhibited by the antioxidant NAC or the Nox inhibitor DPI. Data are mean ±S.E.M. *Po0.05 compared to control or WT (c), & compared to KO AT1, # Po0.05 compared to group treated with AII (a,band d) or KO AT2 (c). One-way analysis of variance and Holm–Sidak post hoc test (n=4–10) Nuclear angiotensin receptors B Villar-Cheda et al 7 Cell Death and Disease Effects of nuclear AII/AT1 activation on nuclear RAS components. Activation of nuclear AT1 receptors by AII induces an increase in the expression of AT2 receptor mRNA. This may lead to an increase in the levels of protective AT2 receptors that traffic to different cell structures such as mitochondria and membrane. A compensatory increase in the opposite or protective arm of the RAS (i.e., AII/AT2) has been observed in different studies after increasing the activity of the membrane AII/AT1/Nox2/superoxide axis. 26–28 Consistent with this, we have recently shown that AT2 receptors are particularly abundant in mitochondria and that mitochondrial AT2 receptors, via NO, downregulate mitochondrial respiration to counteract oxidative stress in neurons. 17 In the present study, we observed that nuclear AT2 receptor activation increases the levels of nuclear NO, and our observations in AT1 and AT2 KO mice suggest that nuclear AT2 receptors modulate the effects of nuclear AT1 receptors. In addition to an increase in AT2 mRNA expression, activation of AT1 nuclear receptors by AII induced an increase in the expression of mRNA for angiotensinogen, renin and renin–prorenin receptors, which indicates that an increase in the synthesis of intracellular angiotensin may further act on intracellular AT2 receptors to counteract oxidative damage. Effects of nuclear AII/AT1 activation on transcription of PGC-1α.Activation of nuclear AT1 receptors also induced an increase in mRNA expression for PGC-1α. Members of the PGC-1 family have emerged as master transcriptional regulators of the mitochondrial function that improve or rescue mitochondrial dysfunction. 29–32 In dopaminergic neurons, PGC-1αhas been shown to be essential for survival, possibly through the maintenance of mitochondrial function, as conditional PGC-1αKO mice show a significant loss of this type of neurons, 33,34 while overexpression of PGC-1αprotects dopaminergic neurons. 33–37 The present results show that activation of nuclear AT1 receptors by AII induces an increase in the expression of PGC-1αthat may counteract the prooxidative effects of activation of the surface membrane AT1 receptor. Nuclear AT2 receptors modulate this process by counteracting the effects of nuclear AT1 receptors. Effects of nuclear AII/AT1 activation on transcription of IGF-1 and SIRT1. We also investigated the effect of activation of nuclear AT1 receptors on two additional factors (i.e., IGF-1 and SIRT1) that have been shown to modulate mitochondrial function and to interact with the RAS in studies performed at tissue and cellular levels. 38,39 IGF-1 has been found to increase survival of dopaminergic neurons, 40–42 and Figure 6 Effect of angiotensin (AII) on nuclear calcium and NO levels. Treatment of rat-isolated nuclei with AII induced a significant increase in nuclear calcium (a), NO levels (d), which were inhibited by the simultaneous treatment with the AT1 receptor antagonist losartan (los) and the AT2 receptor antagonist PD123,319 (PD), respectively. Treatment of isolated nuclei with AII also induced an increase in the expression of mRNA for AT2 (b) and PGC-1α(c), which was inhibited by the simultaneous treatment with the IP3 receptor blocker 2-APB, suggesting that Ca 2+ signaling is involved in these transcriptional changes. The AII-induced increase in nuclear levels of NO was inhibited by the NOS inhibitor L-NAME (d). Data are mean ±S.E.M. *Po0.05 compared to control, # Po0.05 compared to the group treated with AII, & compared to the group treated with AII+los. Oneway analysis of variance and Holm–Sidak post hoc test (n=4–8) Nuclear angiotensin receptors B Villar-Cheda et al 8 Cell Death and Disease activation of nuclear AT1 receptors increases the expression of IGF-1 mRNA. The cytoprotective effects of IGF-1 have been related to mitochondrial protection, leading to a reduction of free radical production, oxidative damage and apoptosis. 43,44 IGF-1 has been suggested to inhibit cell death triggered by oxidative stressors via increasing expression of SIRT1. 45 SIRT1 is present in the nucleus and cytoplasm. It deacetylates several proteins involved in cell survival, metabolism and stress response, and stimulates resistance to oxidative stress in several types of cells. 46 Several studies lead to the conclusion that SIRT1 exerts its control over mitochondrial function mainly by regulating PGC-1α activity. 30,31,47 In a recent study, we observed that AII induced an increase in SIRT1 expression in dopaminergic neurons and microglia. 38 However, the activation of AT1 nuclear receptors in isolated nuclei did not induce any significant increase in SIRT1 mRNA. This suggests that certain components located in the cytoplasm are involved in the increase in expression of SIRT1 induced by membrane AII/AT1 activation. The abovementioned increase in expression of IGF-1 may be a possible mechanism, as it has been observed that IGF-1 induced an increase in SIRT1 expression in cardiomyocytes. 45 The involvement of nuclear angiotensin receptors in the regulation of SIRT1 levels is suggested by our observations in isolated nuclei from transgenic mice overexpressing SIRT1. Overexpression of SIRT1 led to a decrease in expression of nuclear AT1 receptors and an increase in expression of AT2 nuclear receptors, which suggests a possible feedback regulation of the abovementioned mechanism. This is also consistent with observations showing that overexpression of SIRT1 in muscle cells downregulates PGC-1α. 48 In summary, activation of nuclear AT1 receptors by AII may trigger a compensatory mechanism by increasing both PGC-1αand SIRT1 via IGF-1. Effects of nuclear AII/AT1 activation on nuclear calcium and superoxide/H 2 O 2 levels. We have observed that activation of nuclear AT1 receptors induces an increase in nuclear superoxide/H 2 O 2 as well as in Ca 2+ levels, which have been shown to activate signaling pathways that influence gene expression in different types of cells. Nuclear Ca 2+ signaling is an important regulator of gene transcription, 49 and IP3 receptors are involved in Ca 2+ signaling. 50 Calcium has been shown to directly bind Figure 7 Nuclear AT1 and AT2 receptors in aged rats and mice. Nuclei isolated from brains of aged rats showed a significant decrease in the levels of both AT1 and AT2 receptors (a). Treatment of nuclei with AII did not induce any significant increase in AT2, angiotensinogen (ANG) and PGC-1αmRNA expression (b), or in the levels of nuclear superoxide/H 2 O 2 (c). Levels of nuclear superoxide/H 2 O 2 were significantly lower in aged KO AT1 mice than in WTaged mice, and treatment with AII did not induce any significant increase in superoxide/H 2 O 2 in nuclei from aged KO AT1 mice (d). Data are mean ±S.E.M. *Po0.05 compared to the corresponding control, # Po0.05 compared to aged WTor aged WT treated with AII. One-way analysis of variance and Holm–Sidak post hoc test (band d) and Student’st-test (aand c)(n=3–8) Nuclear angiotensin receptors B Villar-Cheda et al 9 Cell Death and Disease