Amyloid-β Peptides Activate α1-Adrenergic Cardiovascular ReceptorsNovelty and Significance
Alzheimer disease features amyloid-β (Aβ) peptide deposition in brain and blood vessels and is associated with hypertension. Aβ peptide can cause vasoconstriction and endothelial dysfunction. We observed that Aβ peptides exert a chronotropic effect in neonatal cardiomyocytes, similar to α1-adrenergic receptor autoantibodies that we described earlier. Recently, it was shown that α1-adrenergic receptor could impair blood–brain flow. We hypothesized that Aβ peptides might elicit a signal transduction pathway in vascular cells, induced by α1-adrenergic receptor activation. Aβ (25–35) and Aβ (10–35) induced a positive chronotropic effect in the cardiac contraction assay (28.75±1.15 and 29.40±0.98 bpm), which was attenuated by α1-adrenergic receptor blockers (urapidil, 1.53±1.17 bpm; prazosin, 0.30±0.96 bpm). Both Aβ peptides induced an intracellular calcium release in vascular smooth muscle cells. Chronotropic activity and calcium response elicited by Aβ (25–35) were blocked with peptides corresponding to the first extracellular loop of the α1-adrenergic receptor. We observed an induction of extracellular-regulated kinase 1/2 phosphorylation by Aβ (25–35) in Chinese hamster ovary cells overexpressing α1-adrenergic receptor, vascular smooth muscle cells, and cardiomyocytes. We generated an activation-state–sensitive α1-adrenergic receptor antibody and visualized activation of the α1-adrenergic receptor by Aβ peptide. Aβ (25–35) induced vasoconstriction of mouse aortic rings and in coronary arteries in Langendorff-perfused rat hearts that resulted in decreased coronary flow. Both effects could be reversed by α1-adrenergic receptor blockade. Our data are relevant to the association between Alzheimer disease and hypertension. They may explain impairment of vascular responses by Aβ and could have therapeutic implications.
- α1-adrenergic receptor
- Alzheimer disease
- amyloid-β peptides
- signal transduction
Alzheimer disease (AD) features the accumulation of amyloid-β (Aβ) in neuritic plaques, brain parenchyma, and walls of cerebral vessels.1,2 Aβ peptides are the cleavage products of the transmembrane amyloid precursor protein and are naturally produced within the central nervous system throughout life.3 Aβ are degraded, chemically modified, and cross-linked, thereby increasing their relative insolubility, stability, and toxicity.4 Aβ peptides from amyloid deposits in brain and cerebrospinal fluid vary in amino acid composition, ranging from the full-length Aβ (1–40) and Aβ (1–42) peptides to shorter carboxy-terminal Aβ peptides, as well as aminoterminal truncated species.5 The ability to affect cognitive processes is typical not only of the full-length peptide but also of several Aβ fragments, in particular the undecapeptide Aβ (25–35).6,7 The fragment consists of amino acid residues 25 to 35 in Aβ (GSNKGAIIGLM) that can form large β-pleated sheet fibrils similar to those obtained by full-length Aβ. Aβ (25–35) exhibits a potent vasoconstrictor effect in rat skin microvasculature and elicits endothelial dysfunction.8 Aβ (25–35) also induces interleukin 1β release from macrophages and microglia.9
AD may be aggravated by vascular mechanisms.10,11 Aβ peptides can enhance vasoconstriction when deposited in microvessels,12 and cause abnormal vascular reactivity even without vascular deposition.13,14 Epidemiological data associate hypertension with AD.14 In preliminary observations, we observed that several Aβ peptides influence the spontaneously beating rate of neonatal rat cardiomyocytes15 in a fashion similar to autoantibodies we described earlier that stimulate the α1-adrenergic receptor (α1-AR) in patients with refractory hypertension.16 Furthermore, Karczewski et al17 used MRI and observed a pathological role of α1-AR in the brain vasculature. These observations caused us to pursue the mechanisms by which Aβ influences cardiovascular cells.
Generation of an Activation State–Sensitive α1A-Adrenergic Receptor Antibody
We generated antibodies directed against the activated form of the α1A-adrenergic receptor (α1A-AR) by adapting an earlier method for other G-protein–coupled receptors.18 Accordingly, the peptide ASDSSNCTQP (Biosyntan; Berlin, Germany) was chosen from the midportion of the aminoterminal tail of human α1A-AR (aa 8–17; NP_150646, National Center for Biotechnology Information). The peptide was coupled to keyhole limpet hemocyanin (Calbiochem) using the glutaraldehyde method. New Zealand white rabbits were immunized with the conjugated peptide according to standard procedures. Collected antisera were affinity-purified onto the same peptide immobilized to ω-aminohexyl agarose (Sigma-Aldrich).
Cardiac Contraction Assay, Preparation of Human α1-AR Antibody IgG Fractions
Isolation and cultivation of neonatal heart cells were performed as described previously.16 For detection of chronotropic activity of various Aβ peptides, Aβ (1–15), Aβ (25–35), Aβ (10–35), Aβ (1–40), and Aβ (1–42) obtained from antibodies-online were added to neonatal rat cardiomyocytes at a concentration of 0.1 µmol/L. To show activation of a G-protein–coupled receptor and identify the receptor, β1 (bisoprolol, 1 µmol/L) and β2 (ICI 118.551, 0.1 µmol/L) adrenergic receptor antagonist, angiotensin II receptor type 1 (AT1) receptor blocker, 1 µmol/L irbesartan, α1A-AR blockers, 1 µmol/L prazosin, and 1 µmol/L urapidil were added. For the neutralization experiments, synthetic peptides, P37-LGYWAF, P38-WAFGRV, P39-GRVFCN, P40-APEDET (Biosyntan), corresponding to the sequence of the first extracellular loop (P37, P38, P39) and the second extracellular loop (P40) of the human α1Α-AR, were each added with a 2000-fold excess. The isolation of immunoglobulin fractions containing human α1-AR antibodies against the first extracellular loop from serum samples was completed as described earlier.16 The immunoglobulin fractions were added to neonatal rat cardiomyocytes at a dilution of 1:40. The beating rate was counted for 15 seconds and 5, 60, and 240 minutes after the addition of Aβ peptides, antagonists, α1-AR antibodies, or synthetic peptides corresponding to the sequence of α1Α-AR.
Isolation of Adult Rat Ventricular Cardiomyocytes
Single ventricular myocytes were isolated enzymatically from adult male Sprague-Dawley rats (≈250 g) as described in Lamounier-Zepter et al.19 Cardiomyocytes were seeded on laminin-coated slides for confocal microscopy and treated with 1 µmol/L phenylephrine and 0.1 µmol/L of various Aβ peptides for 5 minutes, respectively.
Cell Culture, Western Blot, Confocal Microscopy, and Calcium Measurements
Chinese hamster ovary (CHO) cells and CHO cells stably expressing the human α1-AR were cultured in F12 HAM medium supplemented with glutamine, 10% FCS, and 1% penicillin/streptomycin. Rat vascular smooth muscle cells (VSMCs) were cultured in SmBM (Lonza) medium supplemented with SmGM-2 Single Quots (Lonza), and 1% penicillin/streptomycin. For determination of extracellular-regulated kinase (ERK) 1/2 phosphorylation, cells were treated with 1 µmol/L phenylephrine and 0.1 µmol/L of various Aβ peptides, respectively. For inhibition experiments, 1 μmol/L of urapidil was added. Western blot experiments were performed as described earlier.16 ERK1/2- and α1A-AR activity in different cell types was detected by immunfluorescence with specific antibodies against phosphorylated ERK (9101, Cell Signaling, 1:100) and the activated α1A-AR (88 µg/mL; 1:25) and then exposed to the secondary antibody (Alexa-488–conjugated antirabbit IgG, Life Technologies, 1:200) by immunfluorescence. For microscopy, the preparation was mounted with mounting medium (Aqua Polymount, Polyscience) under a glass coverslip and analyzed with a Nikon-Diaphot microscope. A confocal imaging system (MRC 1024 Bio-Rad, Munich, Germany) with an ultraviolet source and an argon/krypton laser was used. At least 30 cells from each of ≥3 independent experiments were examined under each experimental condition. For each set of experiments, identical settings for power of the light source, confocal aperture, gain, and black level were used. Immunofluorescence was evaluated by confocal microscopy by using pseudocolors to encode fluorescence intensity.
Intracellular calcium levels in rat VSMCs were measured with the Fluo-4 Direct Calcium Assay Kit (Life Technologie) according to the manufacturer protocol. Measurements started after the addition of 1 µmol/L phenylephrine, 0.1 µmol/L of various Aβ peptides or 2 µg/mL α1AR-autoantibody, respectively. Human IgG containing activating α1-AR antibodies were obtained from patients requiring ≥3 medication classes to achieve goal control blood pressure values as described earlier.16 For the identification of receptor-binding sites, various synthetic peptides corresponding to the sequence of the first extracellular loop (P37, P38, P39), and the second extracellular loop (P40) of the human α1Α-AR were preincubated with a 100-fold excess with 0.1 µmol/L Aβ (25–35) for 30 minutes at 4°C and added, respectively.
Ex Vivo Vascular Activity Analyses
All procedures were performed according to guidelines from the American Physiological Society after appropriate approval. Male wild-type C57BL/6 mice (25–30 g, 8–12 weeks) were used for the measurements of isometric contractions and were euthanized under isoflurane. Mouse thoracic aortas were removed, quickly transferred to cold (4°C), oxygenated (95% O2/5% CO2) Krebs–Henseleit buffer solution, and dissected into 2-mm rings whereby perivascular fat and connective tissue were removed. Rings were incubated in medium 199 supplemented with 1% penicillin/streptomycin and kept at 37°C for 24 hours in the presence or absence of 1 µmol/L Aβ (25–35). For inhibition experiments, 1 μmol/L of prazosin was added. Each ring was positioned between 2 stainless steel wires (diameter, 0.0394 mm) in a 5-mL organ bath of a Mulvany Small Vessel Myograph (DMT 610 M; Danish Myo Technology, Denmark).20 The organ bath was filled with Krebs–Henseleit buffer solution. The bath solution was continuously oxygenated with a gas mixture of 95% O2 and 5% CO2, and kept at 37°C (pH 7.4). The rings were placed under tension of 0.3 g.20 The software Chart5 (AD Instruments Ltd, Spechbach, Germany) was used for data acquisition and display. The rings were precontracted with 60 mmol/L KCl and equilibrated until a stable resting tension was acquired. Vessels were exposed to increasing cumulative doses of 5-hydroxytryptamine (3–1000 nmol/L). Drugs were added to the bath solution if not indicated otherwise. Tension is expressed as tension/length (mN/mm).
Male Sprague-Dawley rats (250 g, 11 weeks) were injected intraperitoneally, with 200 I.U. of Heparin-Na in PBS, and anesthetized with 20 mg Pentobarbital in 1 mL distilled water. After complete narcosis was reached, they were euthanized and hearts were rapidly excised. After placing them in ice-cold modified Krebs–Henseleit buffer, the aorta was cannulated. The mounted hearts were perfused in a retrograde fashion at constant pressure (70 mm Hg) with continuously aerated (95% O2–5% CO2) modified Krebs–Henseleit buffer containing (in mmol/L) 118 NaCl, 4.7 KCl, 1.2 MgSO4, 1.8 CaCl2, 22.45 NaHCO3, 0.23 KH2PO4, and 11.1 glucose, at a temperature of 37°C. The hearts were stabilized for 15 minutes and then treated for 20 minutes with 0.1 µmol/L urapidil or distilled water followed by 25 minutes of perfusion with 0.1 µmol/L Aβ (25–35) or distilled water as control. Hugo Sachs Electronic software was used for the data acquisition.
Data are presented as mean±SEM. Group differences were analyzed by t test, Mann–Whitney U test, or ANOVA with Bonferroni post hoc test, as appropriate. A value of P<0.05 was considered statistically significant.
Aβ (25–35) increased the beating rate of neonatal rat cardiomyocytes in a time-dependent manner, whereas the negative control peptide with a reverse sequence of Aβ (25–35) did not change the frequency of spontaneously beating neonatal cardiomyocytes (Figure 1A). We found that the beating rate increased ≤4 hours (P≤0.005). The chronotropic effect of Aβ (25–35) was blocked by prazosin and urapidil, whereas β1 and β2 adrenergic receptor antagonists and angiotensin II AT1 receptor blocker had no effect (Figure 1B). Dose–response curves of Aβ (25–35) were generated (10−6 to 10−10 mol/L) in one log order concentrations and showed a typical sigmoid curve (Figure 1C). The efficacy (Emax) and potency (EC50) parameters are derived from these data. The maximal chronotropic effect (Emax) of Aβ (25–35) was a 30.82±1.22 increase in number of bpm, and the EC50 value for Aβ (25–35) was 7.4±1.7 nmol/L. A maximal stimulatory effect was observed at a concentration of 1 µmol/L of Aβ (25–35). Using various Aβ peptides (Figure 1D) in cardiac contraction assay, we observed positive or negative chronotropic activity depending on the amino acid sequence of Aβ peptides (Figure 1E). Although Aβ (25–35) and Aβ (10–35) increased the cardiomyocyte contraction rate, peptides Aβ (1–15), Aβ (1–40), and Aβ (1–42) led to decreased beating rate (P≤0.0001).
Treatment of rat VSMCs with phenylephrine, Aβ (25–35), Aβ (10–35), and α1-AR autoantibodies led to an immediate increase in intracellular Ca2+, whereas Aβ (1–15), Aβ (1–40), and Aβ (1–42) resembled untreated VSMCs. Aβ (25–35) induced a similar increase in intracellular Ca2+ level to that observed by the stimulation with phenylephrine and α1-AR antibodies (Figure 1E). Aβ (10–35) resulted in a weaker intracellular calcium release compared with Aβ (25–35). For mapping of the binding site of Aβ (25–25) to α1-AR, we used synthetic peptides corresponding to the sequence of the first extracellular loop (P37, P38, P39), and the second extracellular loop (P40) of the human α1Α-AR (Figure 2A). Epitope α1A-AR mapping by cardiac contraction assay was performed with overlapping peptide sequences (Figure 2B). For the first extracellular loop, the peptides P38 (WAFGRV) and P39 (GRVFCN) blocked the spontaneous beating rate response of Aβ (25–35) completely, whereas the peptide sequence P37 (LGYWAF) from the first extracellular loop and P40 (APEDET) and other overlapping peptides (data not shown) from the second extracellular loop showed no effect. Importantly, P38 and P39 from the first extracellular loop also abolished the calcium response of Aβ (25–35; Figure 2C).
We next examined the effects of Aβ peptides on ERK 1/2 signal transduction. Using phosphospecific immunoblotting, we observed that phenylephrine and Aβ (25–35) induced ERK 1/2 phosphorylation in α1AR-harboring CHO cells (Figure 3A). The incubation of VSMCs with various Aβ peptides also resulted in ERK 1/2 phosphorylation (Figure 3A, right). In situ imaging of ERK 1/2 signaling was performed (Figure 3B and 3C). Aβ (25–35) induced ERK activation in a time-dependent manner in neonatal cardiomyocytes and in α1A-AR harboring CHO cells but not in wild-type CHO cells (Figure 3B). In adult cardiomyocytes and VSMCs, Aβ (25–35) elicited ERK signaling similar to phenylephrine, whereas all other Aβ peptides induced minor or no ERK activation (Figure 3C).
We next generated an activation-state–sensitive α1A-AR antibody and confirmed α1A-AR activation by Aβ (25–35) and Aβ (10–35; Figure 4A). This conformational-dependent antibody did not react with the nonstimulated α1A-AR (Figure 4A). However, when the α1A-AR was activated by phenylephrine, we obtained positive staining. This staining could be blocked by urapidil. Aβ (25–35) and Aβ (10–35) induced strong α1A-AR staining that was also blocked by urapidil. None of the other Aβ peptides led to receptor activation (data not shown).
The influence of Aβ (25–35) on vasoconstriction of mouse aortic rings was studied in the presence and absence of prazosin. The vasoconstriction to 5-hydroxytryptamine was significantly increased (P≤0.001) in aortic rings by Aβ (25–35) pretreatment (Figure 4B, left). This vasoactive effect was abolished in the additional presence of prazosin during the Aβ (25–35) pretreatment. In a further ex vivo model of Langendorff-perfused rat hearts, Aβ (25–35) induced vasoconstriction of coronary arteries that resulted in decreased coronary flow (Figure 4C). Rat hearts preincubated for 20 minutes with urapidil showed a significant higher (P≤0.005) coronary flow compared with Aβ (25–35) perfused hearts without urapidil pretreatment. Similar results were achieved in middle cerebral artery segments (Figure S2). As expected, there was a significant effect of concentration on responses in all middle cerebral artery segments. Incubation with Aβ (25–35) significantly enhanced the middle cerebral artery response to α1 agonist phenylephrine (P=0.0003) but not to KCl.
We found that Aβ peptides can activate the α1A-AR in a fashion similar to α1A-AR activating antibodies that we described earlier.16 The activation was robust, and the binding site on α1A-AR was specifically identified. We showed that the activation could be relevant because calcium signaling and ERK phosphorylation occurred in VSMCs. Using an activation-state–sensitive α1-AR antibody, we could confirm activation of the α1-AR by Amyolidβ. Aβ (25–35) induced vasoconstriction of mouse aortic rings and in coronary arteries in Langendorff-perfused rat hearts that resulted in decreased coronary flow. Both effects could be reversed by α1-AR blockade. Our data extend the effects of Aβ peptides to the vasculature outside the brain in a functional and structural fashion and could in part explain how Aβ can adversely influence blood vessel function.2
Aβ has been identified in media and adventitia of small arteries and capillaries of cerebral and peripheral vessels earlier.11,13,21 Investigators have proposed that vascular dysfunction is an early component of AD pathology.11,13,21,22 In vitro data showed that soluble Aβ can cause abnormal vascular reactivity of peripheral arteries in the absence of vascular deposition or vessel wall dysfunction.12,23 We suggest that α1-AR activation could play a role in this process. Surprisingly the kinetics of the α1-AR activation by Aβ (25–35) is quite distinct from that of natural ligands. The α1-AR activation elicited by Aβ (25–35) lasts for hours, whereas the effects of conventional agonists are much shorter.
It has been shown by several groups that soluble Aβ can cause abnormal vascular reactivity.11,12 Amyloid-β–induced vascular dysfunction might be an early step in Aβ diseases and could even precede significant Aβ deposition. Transgenic mice overexpressing amyloid precursor protein show attenuated cerebrovascular reactivity to endothelium-dependent vasodilators.11,22 We show for 3 different vascular beds, namely middle cerebral artery, aorta, and coronary arteries that application of exogenous Aβ (25–35) to normal blood vessels ex vivo causes endothelium-dependent vasoconstriction, which can be blocked by α1-adrenoceptor blockers.
Acute application of Aβ peptide to the cerebral cortex of normal mice reproduces the cerebrovascular effects observed in transgenic mice, linking the Aβ peptide to the mechanisms of the vascular dysfunction.24 Using the apolipoprotein E–deficient mouse model for atherosclerosis, Van de Parre et al showed an involvement of Aβ in cardiovascular disease. Amyloid precursor protein is upregulated in atherosclerosis-prone regions of aorta and intensified atherogenesis.25,26 We observed amyloid deposits in human atherosclerotic plaques suggesting a link between Aβ and vascular dysfunction beyond their role in the central nervous system.
The Honolulu Heart Program/Honolulu Asia Aging Study showed that the Aβ-related risk for AD was increased when blood pressure was higher.27 The authors postulated that hypertension could compromise vascular integrity leading to cerebral amyloid angiopathy and impaired Aβ clearance from the brain.
Recently, several research groups demonstrated Alzheimer pathology in 2 established mouse models of hypertension.15,25,28 Using transverse aortic constriction and angiotensin II infusion, cerebral amyloid deposition was observed as early as 4 weeks after the insult in both models.15 They observed an increased permeability of the blood–brain barrier with albumin leakage in the cortex. In several brain areas controlling cognitive functions, such as the cortex, soluble oligomers and intermediate amyloids were recognized. These data indicated that increased blood pressure levels can be mechanistically related to the pathophysiology of AD.28 They implicated the receptor for advanced glycation end products in the process.28 Several mouse models for AD have been established and characterized; however, it is unclear whether they develop hypertension or hypertension-induced end-organ damage. Moreover, passive transfer of Aβ protein in vivo will be important to show in vivo relevance.29
In the process of AD, Aβ peptides are degraded and modified. Shorter carboxy-terminal truncated Aβ peptides are generated from the full-length Aβ (1–42) and Aβ (1–40) peptides. These truncated Aβ peptides are more toxic and insoluble.30 From our experiments, we cannot differentiate whether the observed effects leading to impairment of blood vessel function are mediated by paracrine effect or by changing into an aggregated, insoluble form. In AD, the fragment consisting of residues 25 to 35 (GSNKGAIIGLM) has been shown to form large sheet fibrils, essentially similar to those obtained by the full-length Aβ.31 Thus, Aβ (25–35) is a disease-related short peptide that represents the actual biologically active region of Aβ. Aβ (25–35) is therefore particularly useful in studying amyloid formation and elongation.31 Aβ (25–35) show potent vasoconstrictor effect in microvasculature and blunts the vasodilator response of substance P and other vasodilators.6,8 The evidence we presented here gives Aβ (25–35) an additional dimension, namely a direct, G-protein receptor–mediated vascular effect.
Aβ (25–35) robustly activates the α1-AR in VSMCs by specifically targeting the first extracellular loop of the receptor. This state of affairs could link hypertension and cardiovascular disease to AD pathogenesis. Moreover, the findings could have therapeutic implications and suggest that site-specific blocking α1-AR could have a salubrious influence on AD progression beyond blood pressure–lowering effects.
Sources of Funding
The work was partially funded by the Deutsche Forschungsgemeinschaft grant HE6249/1-1.
The online-only Data Supplement is available with this article at http://hyper.ahajournals.org/lookup/suppl/doi:10.1161/HYPERTENSIONAHA.113.01348/-/DC1.
- Received March 8, 2013.
- Revision received March 27, 2013.
- Accepted August 5, 2013.
- © 2013 American Heart Association, Inc.
- Thinakaran G,
- Koo EH
- Hardy J,
- Selkoe DJ
- Smith EE,
- Greenberg SM
- Elias MF,
- Davey A
- Gupta A,
- Décaillot FM,
- Gomes I,
- Tkalych O,
- Heimann AS,
- Ferro ES,
- Devi LA
- Lamounier-Zepter V,
- Ehrhart-Bornstein M,
- Karczewski P,
- Haase H,
- Bornstein SR,
- Morano I
- Fésüs G,
- Dubrovska G,
- Gorzelniak K,
- Kluge R,
- Huang Y,
- Luft FC,
- Gollasch M
- Iadecola C,
- Park L,
- Capone C
- Shah NS,
- Vidal JS,
- Masaki K,
- Petrovitch H,
- Ross GW,
- Tilley C,
- DeMattos RB,
- Tracy RP,
- White LR,
- Launer LJ
- Carnevale D,
- Mascio G,
- D’Andrea I,
- Fardella V,
- Bell RD,
- Branchi I,
- Pallante F,
- Zlokovic B,
- Yan SS,
- Lembo G
- Ma B,
- Nussinov R
Novelty and Significance
What Is New?
The present study shows for the first time that amyloid-β peptides induce a signal transduction pathway in cardiovascular cells via activation of α1-adrenergic receptor.
What Is Relevant?
The present study provides further evidence of the association between hypertension and Alzheimer disease. Our data provide a mechanism for the observed impairment of blood vessel function by amyloid-β peptides.
We found that amyloid-β peptides can activate the α1-adrenergic receptor and specifically identified the binding site. The peptides induce calcium signaling and extracellular-regulated kinase phosphorylation in vascular cells and present a novel mechanism by which amyloid-β peptides induce vascular dysfunction. These findings support the idea that vascular pathology is an early component of Alzheimer disease.