A total of 182 male mice were used in this study across four separate cohorts (see Table 1). All WTs and Tg2576 mice within each individual cohort were from the same litters to act as appropriate littermate controls. Mice were individually housed due to their aggressive behaviour, as previously reported [40, 41]. Tg2576 mice expressing the human APP695Swe mutation were maintained on a hybrid C57Bl/6 × SJL background . Mice were housed on a 12-h light–dark cycle. Testing occurred during the light period (08:00–19:00). Mice were allowed ad libitum access to food and water throughout the duration of the experiment. All experiments were conducted in compliance with the UK Home Office under the Animal Scientific Procedures Act (1986) and EU regulations.
Diminazene aceturate (DIZE) was purchased from Santa-Cruz Biotechnology (CAS 908-54-3). DIZE was freshly prepared daily in 0.9% sterile saline to a concentration of 50 mg/ml. DIZE was administered to Tg2576 mice at 15 mg/kg/day by intraperitoneal (i.p.) injection using a BD MicroFine™ 0.5-ml insulin syringe. All other mice received equivalent volumes of vehicle (saline) injections. This dose was based on several previous studies showing that the optimal dose of DIZE in inducing ACE2 activity was 15 mg/kg/day [39, 43, 44] and a preliminary study of our own across a range of DIZE concentrations by IP that confirmed previous findings (data not shown).
A competitive ACE2 antagonist, based on the structure of a commercially available compound, C16 (MLN4760, Merck Millipore) was synthesised according to reported procedures , and purified by reverse phase flash chromatography. Purity and identity were confirmed by 1H and 13C NMR spectroscopy, polarimetry, high-resolution mass spectrometry and HPLC. The compound data were in good accordance with the literature and the inhibitor performed similar to the commercial inhibitor when tested in vitro using recombinant mouse ACE2 (R&D systems) (data not shown). The ACE2 inhibitor was co-administered with DIZE in relevant experiments, in an identical vehicle as DIZE, at 25 mg/kg/day, which was the dose most consistently used in previous C16 studies [46, 47].
In Experiment 1, we investigated whether DIZE influenced cognitive performance and Aβ-related pathology in male 13–15-month-old Tg2576 mice. An outline of the experimental design is shown in Fig. 1a. Mice from cohort 1 consisted of three groups: WT vehicle (n = 13), Tg2576 vehicle (n = 11) and Tg2576 DIZE (n = 11). Prior to group allocation, all mice were tested on the object-in-place (OiP) task (as described below) to establish a baseline of cognitive performance and to ensure that the Tg2576 groups had a comparable level of performance prior to drug administration. Following group allocation, all mice were weighed and six mice/group were used to assess mean arterial blood pressure (MABP) and heart rate (HR) at baseline (see also below). A 30-day DIZE/vehicle treatment period then commenced. Mice received daily i.p. injections in the morning and were rested for at least 1 h before any physiological/behavioural testing was performed. Mice were weighed and MABP and HR measurements were recorded weekly in the sub-groups of mice until the end of the treatment period. Behavioural assessment was again conducted starting with habituation on day 23 of drug treatment followed by the OiP task on day 28 and day 30 of treatment. Immediately following behavioural assessment at day 30, the mice were culled by cervical dislocation and the left and right hippocampi and cortices were dissected and snap frozen in liquid nitrogen and stored at − 80 °C until processed for downstream biochemical and pathological assessment. A total of three mice in cohort 1 died during the Experiment 1—one WT vehicle and two from the Tg2576 DIZE group (one on day 2 of administration and the second on day 29). Therefore, the final n for the WT vehicle group was n = 12, the Tg2576 vehicle group remained at n = 11 and the Tg2576 DIZE group was n = 9.
In Experiment 2, we aimed to (1) replicate our findings from Experiment 1 in a larger independent cohort (cohort 2) and (2) determine if the observed protective effects of DIZE were specific to ACE2 enhancement. The experimental design used identical dosing and behavioural testing methods as outlined for Experiment 1 (see Fig. 1a). However, in this experiment, we included an additional group of mice that were co-administered DIZE and C16 (an ACE2 inhibitor) to test the hypothesis that enhanced ACE2 activity was required for the protective effects of DIZE. A total of four mice died in cohort 2 (two mice administered DIZE and two administered DIZE + C16). Final group numbers in cohort 2 were, therefore, as follows: WT vehicle (n = 22), Tg2576 vehicle (n = 11), Tg2576 DIZE (n = 12), Tg2576 DIZE + C16 (n = 11). Cohorts 1 (Experiment 1) and 2 (Experiment 2) were used for discovery and replication studies, respectively, to test and verify the behavioural changes independently but brain tissue was collected and combined from both cohorts, given the identical experimental dosages and design, to allow for meaningful statistically powered biochemical assessment of changes in ACE2, Aβ, cytokines, and markers of astrocyte, microglial and vascular function.
In Experiment 3, we studied a separate cohort of mice (cohort 3) consisting of four groups: WT vehicle (n = 7), WT DIZE (n = 7), Tg2576 vehicle (n = 7) and Tg2576 DIZE (n = 7) (all 13–15 months old). The addition of WT DIZE group was designed to assess if DIZE affected WT mice behaviour and biochemistry in the same way as Tg2576 mice. The experimental procedure replicated that of Experiments 1 and 2. Upon completion of this study, mice brains were either snap frozen for biochemical assessment or processed for IHC and IF analyses.
In Experiment 4, we used a 4th group of mice (cohort 4) to investigate if long-term (10 weeks) DIZE administration (Tg-DIZE chronic) could protect against the development of Aβ pathology and cognitive impairment when administered to pre-symptomatic younger mice aged 9–10 months. To validate and extend our observations on the beneficial effects of DIZE in aged (symptomatic) Tg2576 mice, we compared the chronic treated (10 weeks) condition with a group that received DIZE acutely for 10 days (Tg-DIZE acute) at 12–13 months of age (n = 17) immediately prior to behavioural testing and tissue collection. The experimental procedure is summarised in Fig. 1b. We studied WT vehicle (n = 20), Tg2576 vehicle (n = 16) and Tg-DIZE chronic (n = 11) or Tg-DIZE (n = 17).
All mice were tested on the OiP task as previously described [48, 49]. A detailed description of the OiP task is presented in Supplementary Fig. 1, online resource. Briefly, prior to testing associative recognition memory, animals were habituated to the arena. Mice were allowed to explore the arena freely for 10 min on day 1 and day 2. Mice were then further habituated for two consecutive days to the arena containing four different objects for 10 min each day. Each object was approximately 15 cm from the walls of the arena and 25 cm apart from each other. This arrangement remained constant throughout the study. A different set of objects was used each day and across both time points tested. No sets of objects were re-used during habituation or testing. Each mouse received two rounds of behavioural testing; each test was separated by 48 h. The group of objects used and specific pair of objects that underwent a location switch and their order across days were counterbalanced within and between groups.
Time spent exploring the objects was recorded in both sample and test phases. Object exploration was defined according to the methods described previously by Ennaceur and Delacour (1988). In brief, object contact was defined as when an animal was within a 2 cm radius of the object and directly facing, sniffing, gnawing, but not climbing or sitting on, the objects. A discrimination ratio (DR) was used to provide an index of the mouse’s discrimination performance in the test phase that was independent of individual differences in object contact times; this was calculated as the time spent exploring objects in novel locations/the time spent exploring all objects. All reported contact times and DR scores were averaged across the two trials. We mitigated against the potential effects of repeat testing in our study by counterbalancing across object sets and switching the location of objects within and between groups. All objects were cleaned before each phase of testing to reduce the use of odour cues introduced from handling the objects.
Blood pressure measurement
Our aim was to use a dose of DIZE that influenced ACE2 activity but did not change blood pressure. The effect of DIZE (15 mg/kg/day) on MABP and HR was, therefore, assessed at weekly intervals over the entire course of the study in Experiment 1. Measurements were taken prior to DIZE treatment to determine a baseline MABP and then at weekly intervals until the end of the study. Six mice from each treatment group were habituated to the apparatus (Harvard Apparatus) and used throughout the study. To habituate animals for MABP and HR recordings, each mouse was given a 15-min trial/day for 4 days. Mice were placed in a mouse holder (Part no. 76-0184) underneath a heating unit (Part no.76-0178) in a dark room. The tail cuff and pulse transducer (Part no. 76-0432) were placed around the mouse’s tail and inflated 7–8 times/session. Following habituation, each mouse was placed in the holding unit and given 2–3 min to adapt to the dark room and warmth from the heating device, after which the tail cuff was inflated and MABP and HR was recorded from the blood pressure recording unit (Part no. 76-0173). Three recordings (that were averaged) were taken per mouse. Mice were then removed from the holding unit and returned to home-cages, after which the holding units were wiped clean with 70% ethanol wipes and allowed to dry thoroughly between mice, to prevent stress that may have been caused from residual scents from other males.
Biochemical measurements of markers of disease pathology
Immediately following behavioural testing, mice were culled by cervical dislocation and their brains were removed and dissected. In Experiments 1, 2 and 4, both hemispheres were dissected into cortex, hippocampus and frontal cortex and snap frozen at − 80 °C until protein extraction as previously described . In Experiment 3, the left hemisphere was dissected as above, snap frozen, and used for synaptosome extraction (see below). The contralateral hemisphere was fixed in 4%PFA in 0.1 M PBS for 12 h before being transferred to 25% sucrose in PBS for 48 h. Brains were then sliced using a vibratome and stored in cryoprotectant as previously described .
Synaptosome extractions were performed using Syn-PER™ synaptic protein extraction reagent (ThermoFisher, UK). Western blotting was performed using standard methods as described previously . Briefly, after protein quantification, 20 μg protein/sample was resolved on either a 10 or 7.5% polyacrylamide gel and detected with the relevant antibody (NR2B, pY1472-NR2B (Millipore), GluA1, pSer845 GluA1, PSD95 (Abcam), total ERK, phospho-ERK (Cell Signalling Technologies), NR1 (BD Biosciences), Total Tau (DAKO), PHF1 (p-Tau Ser396/404; a generous gift from P. Davies) and β-actin (Sigma).
Amyloid-β ELISA measurements
Soluble and insoluble fractions were extracted from right hippocampus and right cortex as previously described . Briefly, brain samples of each mouse were homogenised in 2% sodium dodecyl sulphate (SDS) using a Precellys 24‐Dual (Bertin technologies). Homogenate was centrifuged at 28,300 rpm for 1 h at 4 °C. The supernatant was carefully removed and stored at – 20 °C for later analysis as the “soluble” fraction. The insoluble pellet was further dissolved in 70% formic acid. Samples were centrifuged again at 28,300 rpm for 1 h at 4 °C. The supernatant was carefully removed and added 1:20 to a neutralising buffer (1 M Tris, 0.5 M Na2HPO4, pH 11) and stored at − 20 °C for subsequent analysis.
Enzyme-linked immunosorbent assay (ELISA) kits specific for human Aβ40 and Aβ42 (Invitrogen:#KHB3482 and #KHB3441) and Aβ43 (Tecan: RE59711) were used to quantify soluble and insoluble Aβ species according to the manufacturer’s instructions. Results were expressed as picograms (pg) per milligram (mg) of tissue. The mean from duplicate measurements are presented.
ACE1 and ACE2 enzyme activity measurements
To quantify the enzyme activity of ACE1 and ACE2, we used established fluorogenic assays that have previously been described [4, 5, 35, 54]. Brain tissue samples from the left hippocampus and left cortex were dissected and homogenised in 1% SDS lysis buffer using an automated Precellys tissue homogeniser. ACE1 and ACE2 activities were quantified using an ACE1 fluorogenic peptide Abz-FRK(Dnp)-P (Enzo Life Sciences) in the presence/absence of captopril (Enzo) or an ACE2 substrate peptide Mca-APK(Dnp) in the presence/absence of MLN4760 (Millipore). Specific enzyme activity was calculated by subtracting the fluorescence in the presence of the specific inhibitor from total fluorescence without inhibition.
ACE2 protein level measurement
Total protein levels of ACE2 were measured in the right hippocampus by ELISA according to the manufacturer’s instructions (Abcam: ab213843).
Angiotensin-I, -II and -(1–7) measurement
Levels of Ang-I, Ang-II, Ang-(1–7) were measured in hippocampal homogenates prepared in 1% SDS using in-house direct ELISA’s, as previously described for the quantification of peptides in human postmortem brain tissue and CSF [4, 5, 35, 36, 54]. In brief, serial dilutions of recombinant Ang-I, Ang-II or Ang-(1–7) (5000–78.125 pg/ml) (Abcam, Cambridge, UK), and mouse brain extracts (diluted 1 in 20 in PBS) were coated on a Nunc Maxisorp 96-well plate (ThermoFisher Scientific, Waltham, MA, USA) for 2 h at room temperature on a plate shaker (300 rpm), washed, and incubated with biotinylated detection antibodies [biotinylated anti-Ang-I (diluted 1 in 100 in PBS); biotinylated anti-Ang-II (diluted 1 in 500 in PBS) or biotinylated anti-Ang-(1–7) (diluted 1 in 100 in PBS) (all from Cloud-Clone, Wuhan, China)] for another 2 h with shaking. After washing and incubation with streptavidin:HRP (diluted 1 in 200 in 0.1% PBS:Tween-20) (R&D systems) for 20 min, the plates were washed and incubated with TMB substrate (R&D systems). The reaction was stopped after 20 min and absorbance was read at 450 nM using a FLUOstar OPTIMA plate reader (BMG Labtech, Aylesbury, BUCKS, UK). Samples were measured in duplicate and interpolated from the serial dilution of recombinant protein.
ELISA kits specific for mouse interleukin (IL)-1β, IL-6 and IL-10 and TNF-α (R&D Systems: DY401, DY406, DY417, DY410) were used to quantify cytokine levels in hippocampal soluble sample extracts. Results were expressed as pg/mg. Means from duplicate readings are presented.
Microglial (CD68 and IBA1) and astrocyte (GFAP) ELISA
The level of an astrocytic activation marker (GFAP) and two independent microglial markers (IBA1 and CD68) were measured in hippocampal samples homogenised in 1% SDS buffer using commercial ELISA kits following the manufacturer’s protocols: GFAP (samples diluted 1 in 2000) (Cat no: SEA068Mu), IBA1 (samples diluted 1:200 in PBS) (Cat no. SEC288Mu) (Cloud Clone, Wuhan, China) and CD68 (samples diluted 1 in 650 in PBS) (Cat no. OKEH03503, Aviva Systems Biology, San Diego, USA).
Vascular markers of tissue oxygenation and BBB leakiness
The level of VEGF was measured in hippocampal lysates diluted 1 in 40 in PBS using a commercially available ELISA kit (Cat no. DY493, R&D systems) (MA, USA). The ratio of MAG:PLP is a marker of tissue oxygenation previously developed in studies using human postmortem brain tissue [55, 56]. MAG was measured in hippocampal lysates diluted 1 in 650 in PBS using an in-house direct ELISA that showed species cross-reactivity in mice as previously described [55, 57]. PLP concentration was measured in hippocampal samples diluted 4 in 100 in PBS using a commercially available sandwich ELISA following the manufacturer’s instructions (Aviva Biological Systems, San Diego, USA).
Brains hemispheres were fixed in 4% paraformaldehyde in 0.1 M PBS/PFA for 12 h at room temperature. Brains were then transferred to 25% reagent-grade sucrose in 0.1 M PBS. Each brain remained in sucrose until sinking, indicating it was fully saturated (approximately 48 h), and was then sliced using a vibratome. 40-μm coronal sections of hippocampus were stored at − 20 C in an ethylene–glycol-based cryoprotectant (Sigma, UK) until they were used for immunohistochemical analysis.
Parenchymal Aβ imaging
Sections were washed in 3 × 5-min changes of PBS and 3 × 5-min washes in distilled water before being incubated for 10 min in formic acid. Sections were then washed 3 × 5 min in distilled water and equilibrated in PBS before being blocked for 20 min in 5% PBS/donkey serum (Sigma-Aldrich, Dorset, UK). Sections were incubated at 4°C overnight with mouse anti-human Aβ (4G8) at 1 in 250 in PBS (Biolegend, San Diego, CA, USA). Sections were washed 3 × 5 min in PBS and incubated at room temperature for 1 h with Alexa Fluor 488 donkey anti-mouse (ThermoFisher UK) at 1 in 500 in PBS. Sections were washed 3 × 5 min in PBS and mounted using Vectashield mounting medium with DAPI (Vector Labs, Peterborough, UK).
MasR immunofluorescent labelling
Sections were washed in 3 × 5-min changes of PBS and blocked for 20 min in 5% PBS/donkey or goat serum (Sigma-Aldrich, Dorset, UK). Sections were incubated at 4 C overnight with rabbit anti-MasR (1:100) (Cat no. AAR-013) (Alomone Labs, Tel Aviv, Israel). Sections were washed 3 × 5 min in PBS and incubated at room temperature for 1 h (1 in 500 in PBS) with Alexa Flour 488 donkey anti-rabbit (1 in 500 in PBS) (ThermoFisher, UK). Sections were washed 3 × 5 min and mounted using Vectashield mounting medium with DAPI (Vector Labs, Peterborough, UK).
GFAP immunolabelling of reactive astrocytes
Sections were washed in 3 × 5-min changes of PBS and blocked for 20 min in 5% PBS/goat serum (Sigma-Aldrich, Dorset, UK). Sections were incubated at 4°C overnight with chicken anti-GFAP (1:100) (Cat no. abs4674) (Abcam, Cambridge, UK). Sections were washed 3 × 5 min in PBS and incubated at room temperature for 1 h with Alexa Fluor 594 goat anti-chicken (1 in 500 in PBS) (ThermoFisher, UK). Sections were washed 3 × 5 min and mounted using Vectashield mounting medium with DAPI (Vector Labs, Peterborough, UK).
All statistical analyses were performed using IBM SPSS statistics (version 25). The behavioural data conformed to the assumptions of analysis of variance (ANOVA) and were analysed using a mixed measures design. Significant interactions were assessed using tests for simple main effects with Bonferroni correction. Western blot data were analysed using one-way ANOVA followed by Tukey’s post hoc analysis. ELISA data were analysed using either an independent-samples t test or one-way ANOVA with post hoc Tukey comparisons. All data were subject to Levene’s and Shapiro–Wilks tests for data normality prior to analysis. Appropriate transformations were carried out when necessary. Data generated from ELISA assays were quantified by comparing data to standard curves from each plate using GraphPad Prism®7 and normalised to total protein concentration.