Cerebrospinal Fluid Changes in the Renin-Angiotensin System in Alzheimer’s Disease

Abstract

Observations in autopsied brain tissue indicate that overactivation of the classical renin-angiotensin system (cRAS) and underactivity within regulatory RAS pathways (rRAS) are associated with pathology in Alzheimer’s disease (AD). The primary aim of this study was to investigate whether cerebrospinal fluid (CSF) markers of RAS are altered in AD in relation to established CSF markers of disease pathology (lower Aβ₄₂ and elevated tau) and CSF markers of capillary dysfunction. We studied 40 controls and 40 AD cases grouped according to a biomarker profile (i.e., AD cases t-tau>400 pg/mL, pTau >60 pg/mL, and Aβ₄₂ <550 pg/mL). Angiotensin-II converting enyme-1 (ACE1) and ACE2 enzyme activity was measured using fluorogenic peptide substrates; sPDGFRβ and albumin level by sandwich ELISA; and angiotensin-I (Ang-I), Ang-II, and Ang-(1-7) by direct ELISA. CSF Aβ₄₂, total, and phosphorylated tau levels were previously measured by INNOTEST sandwich ELISA. CSF ACE1 activity was significantly elevated in AD (p = 0.008) and positively correlated with ACE2 in AD (r = 0.420, p = 0.007). CSF ACE1 weakly correlated with t-tau (r = 0.294, p = 0.066) and p-tau (r = 0.329, p = 0.038) but not with Aβ₄₂ in the controls but not in AD. ACE1 correlated positively with sPDGFRβ (r = 0.426, p = 0.007), a marker of pericyte injury, and ACE2 correlated positively with albumin (r = 0.422, p = 0.008), a marker of blood-brain barrier integrity. CSF Ang-I, Ang-II, and Ang-(1-7) levels were unchanged in AD. This cross-sectional CSF study indicates RAS dysfunction in relation to capillary damage in AD.

Keywords

Alzheimer’s disease angiotensin-II angiotensin-(1-7)angiotensin-II converting enyme-1 (ACE1)angiotensin-II converting enyme-2 (ACE2)cerebrospinal fluid renin-angiotensin system

INTRODUCTION

Alzheimer’s disease (AD) is characterized neuropathologically by the deposition of extracellular amyloid-β (Aβ) within senile plaques and the accumulation of intracellular tau within neurofibrillary tangles [1]. Recent imaging, cerebrospinal fluid (CSF) biomarker studies, and disease modelling indicate that cerebrovascular dysfunction, including reduced cerebral blood flow, neurovascular uncoupling, and blood-brain barrier (BBB) breakdown, are major contributors to cognitive decline and disease pathology in the early stages of AD [2 –4]. A recent study reported that BBB leakiness and a marker of pericyte injury, indicated by increased CSF levels of soluble platelet-derived growth factor receptor β (sPDGFRβ), are sensitive markers of cognitive decline in the very early stages of AD independently of CSF Aβ and tau [5]. Together, these data support a growing hypothesis that cerebrovascular dysfunction is a major and underestimated contributor to cognitive decline and disease pathology particularly in the early stages of AD [6].

The renin-angiotensin-system (RAS) is a hormonal system that regulates systemic blood pressure. The RAS is expressed and functions independently within the brain and is dysfunctional in AD (reviewed in [7, 8]). Studies in human postmortem brain tissue indicate that overactivation of the classical RAS (cRAS) pathway is associated with elevated Aβ and tau pathology in AD [9 –12]. Angiotensin-converting enzyme-1 (ACE1) activity, the level of angiotensin-II (Ang-II), and the expression of angiotensin-II type 1 receptor (AT1R) are all increased in AD [9 , 11–13]. Cerebroventricular infusion of Ang-II, the main effector peptide of the RAS, accelerates Aβ production (via proteolytic cleavage of AβPP) and promotes the accumulation of disease-modified tau in Sprague Dawley rats in association with cognitive decline [14, 15]. Alternative regulatory RAS (rRAS) pathways exist within the brain that counteract the cRAS axis [16 –18]. We recently reported that ACE2 activity, which is primarily responsible for converting Ang-II to Ang-(1-7) [19, 20], was reduced by almost 50% in the mid-frontal cortex in AD and was inversely related to elevated parenchymal AB and tau load [10] and elevated ACE1 activity.

ACE1 inhibitors (ACE-Is) and angiotensin-II type 1 receptor (At1R) blockers (ARBs) that target the cRAS pathway have been shown to reduce the incidence [21 –24], delay cognitive decline [25], and improve cognition in AD [26]. The protective effects of RAS-blockers in AD are associated with lower CSF-tau [27] and elevated CSF-Aβ (indicative of reduced parenchymal Aβ deposition) [28] and reduced Aβ and tau pathology [29, 30]. Together, these data highlight an imbalance in brain RAS in AD comprising cRAS overactivation that is related to disease pathology that is, in part, associated with reduced rRAS activity.

ACE1 and ACE2 are central effectors of the cRAS and rRAS pathways and are responsible for the production and catabolism of Ang-II respectively. ACE1 and ACE2 are single-span membrane-bound ectoenzymes [31] that are predominantly expressed in endothelial and smooth muscle cells throughout the body but are also localized within neuronal cells [32, 33] and non-neuronal glial cells within the central nervous system [34, 35]. Shedding and solubilization of ACE1 [36] and ACE2 [37] results in the release of the active enzyme in various biological fluids including serum and CSF. ACE1 has previously been measured in CSF in AD, but findings are inconsistent with some studies reporting increased CSF ACE1 [38, 39], others reduced ACE1 [40 –42], and one study showing no change [43]. In contrast, ACE2 activity in CSF remains relatively uncharacterized with only a single study to-date showing unaltered CSF ACE2 in AD [41].

The major aim of this study was to investigate if CSF ACE1 and ACE2 activity are altered in AD, and if so, whether RAS CSF changes were associated with established CSF markers of disease pathology (i.e., increased t-tau and p-tau and reduced Aβ₄₂) and novel CSF markers of capillary damage including BBB breakdown (CSF albumin level) and pericyte injury (sPDGFRβ).

METHODS

Study cohorts

We studied CSF from 40 control and 40 AD cases that were approximately matched for age (69.1 y±12.1 in controls and 76.3 y±6.1 in AD) and distribution of sex across cohorts (23M:15F in controls and 25:15F in AD) (Table 1). Measurements of CSF-Aβ₄₂, t-tau, and p-tau using commercially available sandwich enzyme-linked immunosorbent assays (ELISAs) (INNOTEST, Fujirebio, Ghent, Belgium) had previously been determined. All AD patients had abnormal CSF levels of the core AD biomarkers (t-tau >400 ng/L, p-tau >60 ng/L, and Aβ₄₂ <550 ng/L), while controls had normal levels. The cut-off values used in this study are inline with current clinical practice and closely resemble those outlined by Hansson et al. [44]. The present assays were performed on de-identified leftover aliquots from clinical diagnostic CSF samples and complied with the Swedish Biobank law (Biobanks in Medical Care Act). All procedures were previously approved by the Ethical Committee at the University of Gothenburg (Sweden). Information on cognitive assessment and medical histories, including potential use of RAS-targeting or other classes of anti-hypertensives, were not collected for participants.

Table 1

Demographics of the CSF cohort

	n	Age (y)±SD	Sex (M:F)	Aβ₄₂ (ng/L)	t-tau (ng/L)	p-tau (ng/L)
Control	40	69.1±12.1	23:16	825.3±32.4	235.4±11.6	40.6±1.6
AD	40	76.3±6.1	25:15	431.3±12.2	724.1±35.4	84.6±3.1

CSF ACE1 fluorogenic activity assay

ACE1 activity was measured in CSF using an ACE1 specific FRET peptide substrate (Abz-FRK(Dnp)-P) (Biomol International, Exeter, UK) as previously described [10 , 45]. Recombinant ACE1 (1000-31.1 ng/ml) or CSF (5 μl CSF diluted in 45 μl assay buffer) was incubated with ACE1 substrate (10 μM diluted in assay buffer) for 2.5 h at 26°C. Fluorescence was measured with excitation at 320 nm and emission at 405 nm in a plate reader (FLUOstar, BMG Labtech, Aylesbury, UK). All cases were run in duplicate in the presence and absence of captopril for 10 min at 26°C prior to the addition of the FRET substrate; we previously showed that pre-incubation with captopril blocked ACE1 activity by over 90% [11, 13]. ACE1 activity was determined by subtracting the fluorescence in the captopril-inhibited from the untreated wells. A serial dilution of recombinant human ACE1 and measurements on carry-over samples across all plates was used to minimize variation between plates. ACE1 activity was expressed as relative fluorescence units (r.f.u).

CSF ACE2 fluorogenic activity assay

ACE2 activity was measured using the ACE2 specific FRET substrate (Mca-APK(Dnp) (Enzo Life Sciences, Exeter, UK) as previously described [10]. Recombinant ACE2 (440 – 6 ng/ml) (R&D systems, Cambridge, UK) or CSF (50 μl undiluted) was incubated with the FRET substrate (10 μM) diluted in assay buffer (75 mM Tris, 1M NaCl, pH7.5) for 3 h at 37°C. Recombinant ACE2 and CSF samples were assayed in duplicate in the presence or absence of an ACE2 specific inhibitor (MLN4760) (10 μM) (Merck, Nottingham, UK) diluted in distilled water for 10 min at 37°C prior to the addition of the fluorogenic substrate. ACE2 activity was determined by subtracting the fluorescence in the inhibited from the untreated wells. Cleavage of the ACE2 FRET peptide was measured using a microplate reader (BMG Labtech, Aylesbury, UK) at excitation/emission wavelength 330/390 nm. A serial dilution of recombinant human ACE2 and measurements on carry-over samples across all plates was used to minimize variation between plates. ACE2 activity was expressed in relative fluorescence units (r.f.u).

CSF angiotensin-I, angiotensin-II and angiotensin-(1-7) measurement by direct ELISA

Ang-I, Ang-II, and Ang-(1-7) concentrations were measured by direct ELISA as previously described [9, 10]. Serial dilutions of recombinant human Ang-I, Ang-II, and Ang-(1-7) (5000-78.125 pg/ml) (Abcam, Cambridge, UK) or CSF samples (100 μl undiluted) were incubated for 2 h in clear high-binding capacity NUNC maxisorp plates (ThermoFisher Scientific, Waltham, MA, USA) at 26°C with shaking. The wells were washed five times in PBS with 0.05% tween-20 and blocked for 1 h in 1 % PBS: BSA (Sigma Aldrich, Dorset, UK). After another five washes, the wells were incubated with either biotinylated anti-human Ang-I (diluted 1 in 100 in PBS), biotinylated anti-human Ang-II (diluted 1 in 500 in PBS), or biotinylated anti-human Ang-(1-7) (diluted 1 in 100 in PBS) (all from Cloud-Clone, Wuhan, China), for 2 h at 26°C with shaking, followed by a further wash step. Plates were incubated with streptavadin:HRP (1:200) in PBS:0.01% Tween-20 for 20 min at room temperature in the dark and TMB substrate (R&D systems) was added after a further wash and left to develop in the dark for 20 min. The reaction was stopped following the addition of 2 N sulphuric acid and the absorbance was read at 450 nm using a FLUOstar OPTIMA plate reader (BMG labtech, Aylesbury, BUCKS, UK). Concentrations were interpolated for each sample measured in duplicate following interpolation from respective standard curves generated by serially diluting recombinant standards. We previously demonstrated specificity of each assay for the respective target with minimal cross reactivity with other closely-related angiotensin peptides [9, 10].

ELISA measurement of CSF sPDGFRβ

CSF soluble PDGFRβ (sPDGFRβ) level was measured using a commercially available sandwich ELISA (Invitrogen Cat no EHPDGFRB) (ThermoFisher Scientific, Loughborough, UK) following the manufacturer’s protocol. CSF samples (100 μl) were loaded without dilution. Standards, samples, and blanks were added in duplicate. Absorbance was read at 450 nm following the addition of 2 N sulfuric acid in a FLUOstar OPTIMA plate reader (BMG labtech, Aylesbury, UK). sPDGFRβ concentration was interpolated from the standard curve for each case, derived from the serial dilution of recombinant PDGFRβ (18000-24 pg/ml). The average of duplicate measurements for each sample is presented.

ELISA measurement of CSF albumin

CSF albumin level was measured using a commercially available sandwich ELISA (Cat no 108788) (Abcam, Cambridge, UK) following the manufacturer’s protocol. CSF samples were diluted 1 in 2000 in proprietary dilution buffer supplied with the kit. Standards, samples and blanks were added in duplicate. Absorbance was read at 450 nm in a FLUOstar OPTIMA plate reader (BMG labtech, Aylesbury, UK). Albumin concentration in each sample was interpolated from the standard curve for each case, derived from the serial dilution of recombinant human albumin (200-3.125 ng/ml).

Statistical analysis

Unpaired two-tailed t-tests or ANOVA with Bonferroni post-hoc analysis was used for comparisons between groups, and Pearson’s test was used to assess linear correlation with the help of SPSS version 24 (SPSS, Chicago) and GraphPad Prism version 6 (GraphPad Software, La Jolla, CA). p-values <0.05 were considered statistically significant.

RESULTS

ACE1 enzyme activity is elevated in CSF from AD patients and correlates positively with ACE2 activity

CSF ACE1 activity was significantly higher in AD (2,148 rfu±153.8) compared to controls (1,635 rfu±110.9) (p = 0.008) (Fig. 1A). CSF ACE2 activity was unchanged in AD (360.1 rfu±24.8) compared to controls (318.1 rfu±16.3) (p = 0.169) (Fig. 1B). We previously reported an inverse relationship between ACE1 and ACE2 in the mid-frontal cortex in AD [10]. We examined the relationship between ACE1 and ACE2 in CSF and found a positive correlation in AD (r = 0.420, p = 0.007) but not in controls (r = 0.190, p = 0.239) (Fig. 1C).

Fig.1

ACE1 and ACE2 activity in cerebrospinal fluid (CSF) in Alzheimer’s disease (AD). A,B) Bar charts showing ACE1 and ACE2 enzyme activity in CSF in AD compared to age-matched controls. Bars represent the mean level and standard error of the mean. **p<0.01. C) Scatterplot showing relationship between CSF ACE1 and ACE2 grouped into AD and controls. The best-fit linear regression lines and 95% confidence intervals are superimposed. Each point in the scatterplot indicates a single individual. ^**p < 0.01; ^***p < 0.001

CSF ACE1 activity did not vary according to age in either controls (r = 0.056, p = 0.732) or AD (r = 0.114, p = 0.489) (Fig. 2A); however, CSF ACE2 activity correlated positively with age in controls (r = 0.423, p = 0.0065) but not in AD (p = 0.172, p = 0.290) (Fig. 2B). ACE1 and ACE2 activity did not vary according to sex (Fig. 2C, D).

Fig.2

Relationship between CSF ACE1 and ACE2 enzyme activity in relation to age and sex in cerebrospinal fluid (CSF) in normal aging and Alzheimer’s disease (AD). A,B) Scatterplots showing the relationship between ACE1 and ACE2 in normal aging (gray) and AD (black). ACE1 did not correlate with age in either cohort whereas a positive correlation (r = 0.423, p = 0.0065) was observed between ACE2 and age in controls only. Each point in the scatterplot indicates a single individual. The best-fit linear regression lines and 95% confidence intervals are superimposed. C,D) Bar charts show CSF ACE1 and ACE2 activity in control and AD groups according to sex. Bars represent the mean level and standard error of the mean.

CSF ACE1 and ACE2 enzyme activity are not strongly associated with established CSF markers of disease pathology in AD

We previously reported that increased ACE1 and reduced ACE2 activity are associated with parenchymal Aβ and tau load in mid-frontal cortex in AD [10 , 13]. We examined the relationship between CSF ACE1 and ACE2 activity with established CSF markers of disease pathology (t-tau, p-tau, and Aβ₄₂). CSF ACE1 activity did not correlate with CSF t-tau in either control (r = 0.294, p = 0.066) or AD group (r=–0.01, p = 0.945) (Fig. 3A) but was weakly positively correlated with CSF p-tau in controls (r = 0.329, p = 0.038) but not AD (r = 0.135, p = 0.413) (Fig. 3B). CSF ACE1 did not correlate with Aβ₄₂ in either control (r = 0.216, p = 0.179) or AD groups (r = –0.05, p = 0.761) (Fig. 3C). CSF ACE2 did not correlate with CSF-p-tau, CSF-t-tau, or CSF-Aβ₄₂ in either the AD or control groups (Table 2)

Fig.3

Relationship between cerebrospinal fluid (CSF) ACE1 and established CSF markers of disease pathology in Alzheimer’s disease (AD). A-C) Scatterplots showing the relationship between ACE1 activity and total tau (t-tau), phosphorylated tau (p-tau), and Aβ₄₂ in CSF in AD and age-matched controls. ACE1 enzyme activity weakly correlated with p-tau (r = 0.329, p = 0.038) in controls but not AD but all other correlations were not statistically significant. Each point in the scatterplot indicates a single individual. The best-fit linear regression lines and 95% confidence intervals are superimposed.

Table 2

A relationship between CSF ACE2 activity and markers of disease pathology in CSF was not observed in Alzheimer’s disease (AD)

	t-tau (ng/L)	p-tau (ng/L)	Aβ₄₂ (ng/L)
AD alone	r = 0.113	r = 0.076	r = –0.063
(N = 40)	p = 0.487	P = 0.643	P = 0.700
Control alone	r = 0.245	r = 0.183	r = 0.140
(N = 40)	p = 0.128	P = 0.259	P = 0.390

CSF ACE1 and ACE2 activity correlates with markers of capillary damage

ACE1 is enriched in cerebral vessels [46], and there is evidence of vascular breakdown in AD [47] that together are suggestive of a link between RAS activation and vascular dysfunction in AD. Here we examined the relationship between CSF ACE1 and ACE2 with albumin, an established CSF marker of BBB leakiness, and with soluble PDGFRβ, a marker of pericyte injury that is elevated in AD [5]. A positive correlation was observed between CSF ACE1 enzyme activity and CSF sPDGFRβ level in AD cases alone (r = 0.426, p = 0.007) but not in controls (r = 0.112, p = 0.482) (Fig. 4A). CSF ACE2 in contrast did not correlate with CSF sPDGRβ in either control (r = 0.047, p = 0.778) or AD (r = 0.121, p = 0.483) (Fig. 4B) but did correlate positively with CSF albumin in AD (r = 0422, p = 0.008) but not in controls (r = 0.179; p = 0.282) (Fig. 4D). CSF ACE1 activity did not correlate with CSF albumin in either controls (p = 0.038, p = 0.820) or AD (r = 0.129; p = 0.439) (Fig. 4C).

Fig.4

Relationship between cerebrospinal fluid (CSF) ACE1 and ACE2 with CSF markers of vascular dysfunction in Alzheimer’s disease (AD). A-D) Scatterplots showing relationship between CSF ACE1 and ACE2 enzyme activity with markers of pericyte loss (elevated sPDGFRβ) and blood-brain barrier breakdown (elevated albumin) in AD. ACE1 positively correlated with sPDGFRβ (r = 0.426, p = 0.007) and ACE2 positively correlated with CSF albumin (r = 0.422, p = 0.008) in AD. Each point in the scatterplot indicates a single individual. The best-fit linear regression lines and 95% confidence intervals are superimposed.

CSF angiotensin-I, angiotensin-II, and angiotensin-(1-7) level was unaltered in AD

Ang-I, Ang-II, and Ang-(1-7) levels were unchanged between AD and controls in CSF (Fig. 5) and did not correlate with Aβ, t-tau, and p-tau level (data not shown).

Fig.5

Angiotensin peptides are unchanged in cerebrospinal fluid (CSF) in Alzheimer’s disease (AD). Bar charts showing angiotensin-I, angiotensin-II, and angiotensin-(1-7) level in a subset of CSF samples (20 AD versus 20 controls) in AD. Bars represent the mean level and standard error of the mean.

DISCUSSION

In this study, we explored disease-related changes in CSF markers of RAS in relation to established CSF markers of disease pathology (lower Aβ₄₂ and elevated t-tau and p-tau) and markers of capillary damage (sPDGFRβ and albumin) in clinical diagnostic CSF samples in AD. Our observational study indicates that CSF ACE1 activity is elevated in AD. CSF ACE1 was positively correlated with CSF ACE2 in AD. We also observed that CSF ACE2 activity was positively associated with age in the control cohort suggesting it is elevated as a result of normal aging. In contrast to our recent findings in human autopsy brain tissue, there was little evidence of a relationship between CSF ACE1 and ACE2 with CSF t-tau, p-tau, or Aβ₄₂. We did, however, find moderately strong associations between CSF ACE1 activity and sPDGFRβ, a marker of pericyte damage, and between CSF ACE2 activity and CSF albumin level, a marker of BBB leakiness, within the AD cohort only. Angiotensin peptide concentrations were expressed at low levels in CSF and were unaltered in AD. These novel findings indicate that disease-related changes in RAS are associated with markers of capillary dysfunction in AD. Together, this study provides further mechanistic insights and improves our current understanding of the role of RAS in AD.

An imbalance in RAS resulting in overactivation of the cRAS (disease-associated) pathway and reduced activity in rRAS, is observed in human brain tissue in association with disease pathology in AD [9–11 , 13]. In this study, CSF ACE1 activity was increased in AD supporting findings from previous studies in postmortem tissue [11, 13] and the majority of CSF studies [38, 39]. CSF ACE2 activity was unchanged in AD supporting a recent study [41], but it should be noted that CSF ACE2 activity was higher in AD albeit non-significantly in this study and may have failed to reach significance due to the relatively small cohort size. Interestingly, ACE2 was increased in association with aging in the controls but not in AD, potentially indicating that ACE2 changes precede changes associated with disease pathology. CSF ACE1 activity correlated positively with CSF ACE2 activity whereas we previously reported the opposite in brain tissue, i.e., elevated ACE1 and reduced ACE2 [10]. ACE1 and ACE2 are predominantly membrane bound but both can be cleaved at their respective transmembrane domains and are released extracellularly (solubilized) in an active form [48, 49]. Shedding of membrane-bound ACE2 by ADAM-17, resulting in loss of membrane-bound ACE2, has previously been implicated in the pathogenesis of neurogenic hypertension [50] and could also account for the differences between brain and CSF activity in AD. ACE1 in contrast is elevated in both brain tissue and CSF, it is currently unclear whether overproduction of ACE2 or post-translational modifications are responsible for elevated ACE1 in AD.

We previously reported that ACE1 correlated positively with Aβ in AD, and reduced ACE2 activity correlated inversely with Aβ and tau in human brain tissue [10 , 13]. An association between CSF ACE1 and the ratio of p-tau₁₈₁:Aβ₄₂ has previously been reported in two separate cohorts: the Knight Alzheimer’s Disease Research Center (N = 311) and the Alzheimer’s Disease Neuroimaging Initiative (N = 293) [51]. In contrast, we did not find strong evidence to support similar changes between CSF RAS and markers of disease pathology in CSF. This could be due to many factors and potential limitations in the current study including the small cohort size and/or the fact the groups were based on CSF biomarkers levels, rather than clinical or pathological diagnosis. We also have no information on hypertension status or use of RAS-targeting anti-hypertensives that may conceal a potential relationship. It is also unclear and difficult to interpret why we observed a positive correlation between CSF ACE1 and p-tau in controls but not AD. It should, however, also be noted that changes in CSF reflect global changes throughout the brain and do not necessarily reflect specific regional changes previously reported within the brain. Furthermore, this is also an observational cross-sectional study and does not provide any clues as to the temporal sequence of RAS activation in relation to disease pathology. Further studies are required to address these current limitations and confirm or refute our current findings.

In contrast, we have found stronger evidence of a relationship between RAS overactivation and capillary breakdown indicated by moderate associations between CSF ACE1 and ACE2 enzyme activities with CSF markers of capillary damage in AD, namely pericyte damage (elevated CSF sPDGFRβ) and BBB breakdown (elevated CSF albumin), respectively. The microvasculature has been shown to degenerate in AD [47] and ACE1 activity is highly enriched in brain microvessels [46] that together are suggestive of a possible involvement of RAS and vascular disruption in AD. Recent studies also indicate that vascular dysfunction, including BBB breakdown and pericyte loss, are major contributors to cognitive decline and disease pathology in AD [2, 52]. Increased solubilization of PDGFRβ, a marker of pericyte injury (although PDGFRβ is also expressed by sub-populations of reactive glia), has recently been reported in CSF and is associated with BBB leakiness in the aging and diseased hippocampus [29], and has been shown to be a significant predictor of cognitive decline independently of Aβ and tau in the early stages of disease [5]. Our data indicate a relationship between RAS activation and both pericyte loss and BBB function. This potential link is supported by indirect evidence of a relationship between RAS and BBB permeability in vitro and in vivo [54]. Ang-II is a potent vasoconstrictor and mediates pericyte contraction and vessel constriction in vessel extracts from rat retina via AT1R activation [55]. Pericyte contraction and subsequent vessel constriction may be a major contributor to cerebral hypoperfusion in AD [56] and is likely to also lead to pericyte injury and release of soluble PDGFRβ. Collectively our findings indicate a potential relationship between RAS activation and vascular dysfunction in AD that warrants further investigation.

Much less is known regarding changes in angiotensin peptides in AD. We recently reported that Ang-II and Ang-III, despite short half-lives, were detectable in postmortem brain tissue, and were elevated in AD in relation to markers of disease pathology [9]. Despite changes in brain tissue, the levels of Ang-I, Ang-II, and Ang-(1-7) were unchanged in CSF and did not correlate with CSF markers of disease or vascular function. The difference between brain tissue and CSF may reflect the relative instability and shorter half-life of angiotensin peptides in CSF compared to brain. Another possibility that cannot be addressed in this study is that CSF changes in angiotensin peptide level may first require acute or chronic and sustained changes in the activity of brain ACE1 and ACE2.

This study has several limitations which we have already mentioned, including the criteria used to group AD cases, the small cohort sizes, and the lack of clinical details relating to the de-identified AD cases, including cognitive assessment and disease status markers. This is also an observational cross-sectional study and does not provide clues as to the timing and regional changes in RAS in relation to disease and vascular markers and it is also unclear what percentage of the participants were prescribed RAS-targeting anti-hypertensives. Despite these limitations, which will need to be carefully addressed in future studies, this study still provides novel insights into the potential relationship between RAS activation and capillary damage in AD and provides further evidence of a link between RAS and the pathogenesis of AD.

Footnotes

ACKNOWLEDGMENTS

The South West Dementia Brain Bank is part of the Brains for Dementia Research program, jointly funded by Alzheimer’s Research UK and Alzheimer’s Society, and is supported by BRACE (Bristol Research into Alzheimer’s and Care of the Elderly) and the Medical Research Council.

Authors’ disclosures available online ().

References

Gorevic

, Goni

, Pons-Estel

, Alvarez

, Peress

, Frangione

(1986) Isolation and partial characterization of neurofibrillary tangles and amyloid plaque core in Alzheimer’s disease: Immunohistological studies. J Neuropathol Exp Neurol 45, 647–664.

Iturria-Medina

, Sotero

, Toussaint

, Mateos-Perez

, Evans

, Alzheimer’s Disease Neuroimaging Initiative (2016) Early role of vascular dysregulation on late-onset Alzheimer’s disease based on multifactorial data-driven analysis. Nat Commun 7, 11934.

Montagne

, Barnes

, Sweeney

, Halliday

, Sagare

, Zhao

, Toga

, Jacobs

, Liu

, Amezcua

, Harrington

, Chui

, Law

, Zlokovic

(2015) Blood-brain barrier breakdown in the aging human hippocampus. Neuron 85, 296–302.

Sweeney

, Kisler

, Montagne

, Toga

, Zlokovic

(2018) The role of brain vasculature in neurodegenerative disorders. Nat Neurosci 21, 1318–1331.

Nation

, Sweeney

, Montagne

, Sagare

, D’Orazio

, Pachicano

, Sepehrband

, Nelson

, Buennagel

, Harrington

, Benzinger

TLS

, Fagan

, Ringman

, Schneider

, Morris

, Chui

, Law

, Toga

, Zlokovic

(2019) Blood-brain barrier breakdown is an early biomarker of human cognitive dysfunction. Nat Med 25, 270–276.

Sweeney

, Montagne

, Sagare

, Nation

, Schneider

, Chui

, Harrington

, Pa

, Law

, Wang

DJJ

, Jacobs

, Doubal

, Ramirez

, Black

, Nedergaard

, Benveniste

, Dichgans

, Iadecola

, Love

, Bath

, Markus

, Salman

, Allan

, Quinn

, Kalaria

, Werring

, Carare

, Touyz

, Williams

SCR

, Moskowitz

, Katusic

, Lutz

, Lazarov

, Minshall

, Rehman

, Davis

, Wellington

, Gonzalez

, Yuan

, Lockhart

, Hughes

, Chen

CLH

, Sachdev

, O’Brien

, Skoog

, Pantoni

, Gustafson

, Biessels

, Wallin

, Smith

, Mok

, Wong

, Passmore

, Barkof

, Muller

, Breteler

MMB

, Roman

, Hamel

, Seshadri

, Gottesman

, van Buchem

, Arvanitakis

, Schneider

, Drewes

, Hachinski

, Finch

, Toga

, Wardlaw

, Zlokovic

(2019) Vascular dysfunction-The disregarded partner of Alzheimer’s disease. Alzheimers Dement 15, 158–167.

Kehoe

(2018) The coming of age of the angiotensin hypothesis in Alzheimer’s disease: Progress toward disease prevention and treatment? J Alzheimers Dis 62, 1443–1466.

Kehoe

, Miners

, Love

(2009) Angiotensins in Alzheimer’s disease - friend or foe? Trends Neurosci 32, 619–628.

Kehoe

, Hibbs

, Palmer

, Miners

(2017) Angiotensin-III is Increased in Alzheimer’s disease in association with amyloid-beta and tau pathology. J Alzheimers Dis 58, 203–214.

10.

Kehoe

, Wong

, Al Mulhim

, Palmer

, Miners

(2016) Angiotensin-converting enzyme 2 is reduced in Alzheimer’s disease in association with increasing amyloid-beta and tau pathology. Alzheimers Res Ther 8, 50.

11.

Miners

, Ashby

, Van Helmond

, Chalmers

, Palmer

, Love

, Kehoe

(2008) Angiotensin-converting enzyme (ACE) levels and activity in Alzheimer’s disease, and relationship of perivascular ACE-1 to cerebral amyloid angiopathy. Neuropathol Appl Neurobiol 34, 181–193.

12.

Savaskan

, Hock

, Olivieri

, Bruttel

, Rosenberg

, Hulette

, Muller-Spahn

(2001) Cortical alterations of angiotensin converting enzyme, angiotensin II and AT1 receptor in Alzheimer’s dementia. Neurobiol Aging 22, 541–546.

13.

Miners

, Ashby

, Baig

, Harrison

, Tayler

, Speedy

, Prince

, Love

, Kehoe

(2009) Angiotensin-converting enzyme levels and activity in Alzheimer’s disease: Differences in brain and CSF ACE and association with ACE1 genotypes. Am J Transl Res 1, 163–177.

14.

Tian

, Zhu

, Xie

, Shi

(2012) Central angiotensin II-induced Alzheimer-like tau phosphorylation in normal rat brains. FEBS Lett 586, 3737–3745.

15.

Zhu

, Shi

, Zhang

, Wang

, Liu

, Chen

, Tong

(2011) Central angiotensin II stimulation promotes beta amyloid production in Sprague Dawley rats. PLoS One 6, e16037.

16.

Ferrario

, Trask

, Jessup

(2005) Advances in biochemical and functional roles of angiotensin-converting enzyme 2 and angiotensin-(1-7) in regulation of cardiovascular function. Am J Physiol Heart Circ Physiol 289, H2281–2290.

17.

Santos

, Campagnole-Santos

, Andrade

(2000) Angiotensin-(1-7): An update. Regul Pept 91, 45–62.

18.

Santos

, Simoes e Silva

, Maric

, Silva

, Machado

, de Buhr

, Heringer-Walther

, Pinheiro

, Lopes

, Bader

, Mendes

, Lemos

, Campagnole-Santos

, Schultheiss

, Speth

, Walther

(2003) Angiotensin-(1-7) is an endogenous ligand for the G protein-coupled receptor Mas. Proc Natl Acad Sci U S A 100, 8258–8263.

19.

Elased

, Cunha

, Marcondes

, Morris

(2008) Brain angiotensin-converting enzymes: Role of angiotensin-converting enzyme 2 in processing angiotensin II in mice. Exp Physiol 93, 665–675.

20.

Vickers

, Hales

, Kaushik

, Dick

, Gavin

, Tang

, Godbout

, Parsons

, Baronas

, Hsieh

, Acton

, Patane

, Nichols

, Tummino

(2002) Hydrolysis of biological peptides by human angiotensin-converting enzyme-related carboxypeptidase. J Biol Chem 277, 14838–14843.

21.

Davies

, Kehoe

, Ben-Shlomo

, Martin

(2011) Associations of anti-hypertensive treatments with Alzheimer’s disease, vascular dementia, and other dementias. J Alzheimers Dis 26, 699–708.

22.

, Lee

, Whitmer

, Kivipelto

, Lawler

, Kazis

, Wolozin

(2010) Use of angiotensin receptor blockers and risk of dementia in a predominantly male population: Prospective cohort analysis. BMJ 340, b5465.

23.

Yasar

, Xia

, Yao

, Furberg

, Xue

, Mercado

, Fitzpatrick

, Fried

, Kawas

, Sink

, Williamson

, DeKosky

, Carlson

, Ginkgo Evaluation of Memory (GEM) Study Investigators (2013) Antihypertensive drugs decrease risk of Alzheimer disease: Ginkgo Evaluation of Memory Study. Neurology 81, 896–903.

24.

Wharton

, Zhao

, Steenland

, Goldstein

, Schneider

, Barnes

, Gearing

, Yasar

(2019) Neurofibrillary tangles and conversion to mild cognitive impairment with certain antihypertensives. J Alzheimers Dis 70, 153–161.

25.

Wharton

, Goldstein

, Zhao

, Steenland

, Levey

, Hajjar

(2015) Modulation of renin-angiotensin system may slow conversion from mild cognitive impairment to Alzheimer’s disease. J Am Geriatr Soc 63, 1749–1756.

26.

Ohrui

, Tomita

, Sato-Nakagawa

, Matsui

, Maruyama

, Niwa

, Arai

, Sasaki

(2004) Effects of brain-penetrating ACE inhibitors on Alzheimer disease progression. Neurology 63, 1324–1325.

27.

Hajjar

, Levey

(2015) Association between angiotensin receptor blockers and longitudinal decline in tau in mild cognitive impairment. JAMA Neurol 72, 1069–1070.

28.

Nation

, Ho

, Yew

, Alzheimer’s Disease Neuroimaging I (2016) Older adults taking AT1-receptor blockers exhibit reduced cerebral amyloid retention. J Alzheimers Dis 50, 779–789.

29.

Wharton

, Zhao

, Steenland

, Goldstein

, Schneider

, Barnes

, Gearing

, Yasar

(2019) Neurofibrillary tangles and conversion to mild cognitive impairment with certain antihypertensives. J Alzheimers Dis 70, 153–161.

30.

Hajjar

, Brown

, Mack

, Chui

(2012) Impact of Angiotensin receptor blockers on Alzheimer disease neuropathology in a large brain autopsy series. Arch Neurol 69, 1632–1638.

31.

Erdos

, Skidgel

(1987) The angiotensin I-converting enzyme. Lab Invest 56, 345–348.

32.

Doobay

, Talman

, Obr

, Tian

, Davisson

, Lazartigues

(2007) Differential expression of neuronal ACE2 in transgenic mice with overexpression of the brain renin-angiotensin system. Am J Physiol Regul Integr Comp Physiol 292, R373–381.

33.

Lin

, Chen

, Zhang

, Chen

, Lin

, Morris

(2008) RNA interference shows interactions between mouse brainstem angiotensin AT1 receptors and angiotensin-converting enzyme 2. Exp Physiol 93, 676–684.

34.

Turner

, Hooper

(2002) The angiotensin-converting enzyme gene family: Genomics and pharmacology. Trends Pharmacol Sci 23, 177–183.

35.

Gallagher

, Chappell

, Ferrario

, Tallant

(2006) Distinct roles for ANG II and ANG-(1-7) in the regulation of angiotensin-converting enzyme 2 in rat astrocytes. Am J Physiol Cell Physiol 290, C420–426.

36.

Beldent

, Michaud

, Wei

, Chauvet

, Corvol

(1993) Proteolytic release of human angiotensin-converting enzyme. Localization of the cleavage site. J Biol Chem 268, 26428–26434.

37.

Tipnis

, Hooper

, Hyde

, Karran

, Christie

, Turner

(2000) A human homolog of angiotensin-converting enzyme. Cloning and functional expression as a captopril-insensitive carboxypeptidase. J Biol Chem 275, 33238–33243.

38.

, Ohrui

, Maruyama

, Tomita

, Nakayama

, Higuchi

, Furukawa

, Arai

(2006) ACE activity in CSF of patients with mild cognitive impairment and Alzheimer disease. Neurology 67, 1309–1310.

39.

Mateos

, Ismail

, Gil-Bea

, Leoni

, Winblad

, Bjorkhem

, Cedazo-Minguez

(2011) Upregulation of brain renin angiotensin system by 27-hydroxycholesterol in Alzheimer’s disease. J Alzheimers Dis 24, 669–679.

40.

Jochemsen

, Teunissen

, Ashby

, van der Flier

, Jones

, Geerlings

, Scheltens

, Kehoe

, Muller

(2014) The association of angiotensin-converting enzyme with biomarkers for Alzheimer’s disease. Alzheimers Res Ther 6, 27.

41.

Rocha

, Toledo

, Corgosinho

LTS

, de Souza

, Guimaraes

, Resende

EPF

, Braz

NFT

, Gomes

, Simoes

ESAC

, Caramelli

, Teixeira

(2018) Cerebrospinal fluid levels of angiotensin-converting enzyme are associated with amyloid-beta42 burden in Alzheimer’s disease. J Alzheimers Dis 64, 1085–1090.

42.

Zubenko

, Volicer

, Direnfeld

, Freeman

, Langlais

, Nixon

(1985) Cerebrospinal fluid levels of angiotensin-converting enzyme in Alzheimer’s disease, Parkinson’s disease and progressive supranuclear palsy. Brain Res 328, 215–221.

43.

Konings

, Kuiper

, Scheltens

, Grijpma

, van Pelt

, Wolters

(1993) Re-evaluation of cerebrospinal fluid angiotensin-converting enzyme activity in patients with ‘probable’ Alzheimer’s disease. Eur J Clin Chem Clin Biochem 31, 495–497.

44.

Hansson

, Zetterberg

, Buchhave

, Londos

, Blennow

, Minthon

(2006) Association between CSF biomarkers and incipient Alzheimer’s disease in patients with mild cognitive impairment: A follow-up study. Lancet Neurol 5, 228–234.

45.

Miners

, van Helmond

, Raiker

, Love

, Kehoe

(2010) ACE variants and association with brain Abeta levels in Alzheimer’s disease. Am J Transl Res 3, 73–80.

46.

Kalaria

, Harik

(1988) Adenosine receptors and the nucleoside transporter in human brain vasculature. J Cereb Blood Flow Metab 8, 32–39.

47.

Hase

, Ding

, Harrison

, Hawthorne

, King

, Gettings

, Platten

, Stevenson

, Craggs

LJL

, Kalaria

(2019) White matter capillaries in vascular and neurodegenerative dementias. Acta Neuropathol Commun 7, 16.

48.

Lambert

, Yarski

, Warner

, Thornhill

, Parkin

, Smith

, Hooper

, Turner

(2005) Tumor necrosis factor-alpha convertase (ADAM17) mediates regulated ectodomain shedding of the severe-acute respiratory syndrome-coronavirus (SARS-CoV) receptor, angiotensin-converting enzyme-2 (ACE2). J Biol Chem 280, 30113–30119.

49.

Sadhukhan

, Sen

, Ramchandran

, Sen

(1998) The distal ectodomain of angiotensin-converting enzyme regulates its cleavage-secretion from the cell surface. Proc Natl Acad Sci U S A 95, 138–143.

50.

Xia

, Sriramula

, Chhabra

, Lazartigues

(2013) Brain angiotensin-converting enzyme type 2 shedding contributes to the development of neurogenic hypertension. Circ Res 113, 1087–1096.

51.

Harari

, Cruchaga

, Kauwe

, Ainscough

, Bales

, Pickering

, Bertelsen

, Fagan

, Holtzman

, Morris

, Goate

, Alzheimer’s Disease Neuroimaging Initiative (2014) Phosphorylated tau-Abeta42 ratio as a continuous trait for biomarker discovery for early-stage Alzheimer’s disease in multiplex immunoassay panels of cerebrospinal fluid. Biol Psychiatry 75, 723–731.

52.

Nation

, Sweeney

, Montagne

, Sagare

, D’Orazio

, Pachicano

, Sepehrband

, Nelson

, Buennagel

, Harrington

, Benzinger

TLS

, Fagan

, Ringman

, Schneider

, Morris

, Chui

, Law

, Toga

, Zlokovic

(2019) Blood-brain barrier breakdown is an early biomarker of human cognitive dysfunction. Nat Med 25, 270–276.

53.

Matsugi

, Chen

, Anderson

(1997) Contractile responses of cultured bovine retinal pericytes to angiotensin II. Arch Ophthalmol 115, 1281–1285.

54.

Fuchtbauer

, Toft-Hansen

, Khorooshi

, Owens

(2010) Expression of astrocytic type 2 angiotensin receptor in central nervous system inflammation correlates with blood-brain barrier breakdown. J Mol Neurosci 42, 89–98.

55.

Kawamura

, Kobayashi

, Li

, Yamanishi

, Katsumura

, Minami

, Wu

, Puro

(2004) Effects of angiotensin II on the pericyte-containing microvasculature of the rat retina. J Physiol 561, 671–683.

56.

Nortley

, Korte

, Izquierdo

, Hirunpattarasilp

, Mishra

, Jaunmuktane

, Kyrargyri

, Pfeiffer

, Khennouf

, Madry

, Gong

, Richard-Loendt

, Huang

, Saito

, Saido

, Brandner

, Sethi

, Attwell

(2019) Amyloid beta oligomers constrict human capillaries in Alzheimer’s disease via signaling to pericytes. Science 365, eaav9518.