Effector function of anti-pyroglutamate-3 Aβ antibodies affects cognitive benefit, glial activation and amyloid clearance in Alzheimer’s-like mice

A passive immunization study was conducted in male, plaque-rich APP_SWE/PS1ΔE9 transgenic (Tg) mice on a C57BL/6 J background starting at ~ 12 mo of age. APP_SWE/PS1ΔE9 mice express two human genes of familial AD, the APP^K594N/M595L Swedish and Presenilin 1 delta E9 (PS1ΔE9) (deletion of exon 9) under a mouse prion promoter [27]. Original Tg breeders were obtained from The Jackson Laboratory (Bar Harbor, ME) and were maintained in our colony by crossing male APP_SWE/PS1ΔE9 Tg mice with female C57BL/6 J mice. All animal protocols were approved by the Harvard Medical Area Standing Committee on Animals, and studies were performed in accordance with all state and federal regulations. The Harvard Medical School animal management program is accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care International and meets all National Institutes of Health standards as demonstrated by an approved Assurance of Compliance (A3431-01) filed at the Office of Laboratory Animal Welfare.

Another amyloid mouse model, hAPPSL;hQC mice, which are double transgenic for the human APP gene containing Swedish and London mutation and human glutaminyl cyclase (QC) [19], were also used to assess the efficacy of 07/2a IgG2a and 07/2a-k IgG2a antibodies. Animals were housed in individually ventilated cages on standardized rodent bedding supplied by Rettenmaier Austria GmbH & Co.KG (Vienna, Austria). Mice were kept in the Association for Assessment and Accreditation of Laboratory Animal Care-accredited animal facility of QPS Austria GmbH (previously JSW Lifesciences, GmbH, Grambach, Austria). Animal studies conformed to the Austrian guidelines for the care and use of laboratory animals and were approved by the Styrian government, Austria.

Multiple murine anti-pyroglutamate-3 Aβ antibodies of different IgG isotypes were used in these studies, all of which were generated and provided by Vivoryon Therapeutics AG (formerly Probiodrug AG, Halle, Germany). These include murine IgG1 (07/1) and IgG2a (07/2a) isotype versions of the same 07 monoclonal antibody targeting the N-terminally truncated and modified (by cyclization by QC) Aβ. In an effort to avoid inflammation and possibly ARIA, a CDC-mutant (K322A) version of the murine IgG2a isotype antibody (07/2a-k), which exhibits less complement activation due to its inability to bind C1q, was engineered and provided by Vivoryon Therapeutics AG. The respective IgG1 isotype control antibody was purchased from Thermo Fisher Scientific; the IgG2a isotype control was produced by Fraunhofer IZI-MWT.

In addition, the following studies included a purified β-amyloid 1–15 monoclonal antibody raised against dityrosine cross-linked human Aβ1–40, named 3A1, which was kindly provided by Dr. Brian O’Nuallain and is now commercially available from BioLegend. 3A1 does not recognize pGlu3 Aβ or APP [12].

A total of 80 male 12-mo-old APP_SWE/PS1ΔE9 mice were utilized for the first in vivo study. Prior to the start of immunization, five APP_SWE/PS1ΔE9 Tg mice (avg. 12.3 mo ± 0.08) were sacrificed as baseline controls to assess plaque burden at commencement of treatment. The remaining 75 mice were divided into four groups and received the following treatments: 250 μl sterile phosphate buffered saline (PBS) (n = 14; originally n = 15 but one removed due to incorrect genotype); avg. 12.2 mo ± 0.77), 300 μg 3A1, a general Aβ IgG1 mAb (n = 15; avg. 12.2 mo ± 0.8), 300 μg 07/1, a pGlu-3 Aβ IgG1 mAb (n = 15; avg. 12.2 mo ± 0.55) or 300 μg 07/2a, a pGlu-3 Aβ IgG2a mAb (n = 15; avg. 12.3 mo ± 0.35). A group of age- and gender-matched wildtype (Wt) littermates were injected with 250 μl sterile PBS (n = 15; avg. 12.3 mo ± 0.91) and served as behavioral controls. Mice were treated weekly with a total volume of 250 μl antibody or PBS via intraperitoneal (i.p) injection for 16 weeks (i.e., from 12 to 16 mo of age).

Behavioral testing of all mice was performed starting at ~ 15 mo of age, during which time, mice continued to be treated. Open field (to measure spontaneous locomotor activity and anxiety), contextual fear conditioning (CFC) (to measure associative learning and fear memory), and water T-maze (to assess spatial learning and memory) tests were performed as described [12]. Mice were euthanized at 16 mo of age, 1 week after the last treatment.

An additional immunotherapy study with 29 mice in total was conducted in 8-mo-old hAPPSL;hQC mice examining the in vivo effects of 07/2a IgG2a and 07/2a-k IgG2a-k. These mice were divided into three groups; PBS vehicle treated (n = 7), 07/2a (150 μg) (n = 8), 07/2a (500 μg) (n = 7) and 07/2a-k (500 μg) (n = 7) and given a weekly i.p injection for 16 weeks.

Mice were euthanized and perfused and brain and plasma were harvested prior to dosing (baseline) or after i.p injections for 16 weeks as previously described in [16] and in Additional file 1 (S1.1).

Ten- and 20-μm-thick brain sagittal cryosections were cut with a Leica CM1850 cryostat and mounted on Colorfrost Plus slides (Fisher Scientific) or Poly-L-lysine-coated 12-mm glass coverslips (Corning) for immunohistochemistry (IHC) and an ex vivo phagocytosis assay, respectively. Neuropathological analysis of Aβ plaque burden and associated gliosis was carried out as previously described in [16] and in Additional file 1: (S1.2). Fibrillar amyloid was stained with Thioflavin S dye as previously described in [28] and in Additional file 1: (S1.2). Brain tissue without the cerebellum was homogenized. Aβ and cytokine levels in homogenates were measured by ELISA as outlined in Additional file 1: (S1.3).

The concentrations of the exogenous antibodies (3A1, 07/1, and 07/2a) in the brain homogenates and the terminal plasma samples (1 week post-final injection) were measured as described [12], using streptavidin-coated 96-well plates coated in biotinylated Aβ peptide as described in Additional file 1: (S.1.4).

Primary microglia (PMG) were prepared from the cortices of mouse pups at postnatal day 5 and cultured as described [29, 30]. The murine N9 immortalized microglial cell line was generated by Dr. Paola Ricciardi-Castagnoli [31] and was kindly provided to us by Dr. Joseph El Khoury (Massachusetts General Hospital) with her permission.

Aged (20-mo-old), plaque-rich APP_SWE/PS1ΔE9 unfixed mouse brains were gently frozen in liquid nitrogen vapors, sagittally sectioned to 20 μm, and mounted on 12 mm round poly-l-lysine-coated cover slips and placed in 24-well tissue culture plates for the PMG cell assay or on 18 mm round poly-l-lysine-coated cover slips and placed in 12-well tissue culture plates for the N9 cell assay. IHC was performed to confirm tissue was plaque-rich with an abundance of pGlu3-Aβ epitope available. Serial sections were incubated with or without antibodies for 1 h at 37 °C in 5% CO₂: anti-pGlu3 Aβ mAbs 07/1 IgG1, 07/2a IgG2a, and 07/2a-k IgG2a; anti-Aβ 3A1 IgG1; or isotype control IgG1 or isotype control IgG2a, at optimized concentrations of 10 μg/ml for the PMG cell assay or 15 μg/ml for the N9 cell assay. The assay was carried out as described by [32] and in Additional file 1 (S1.5).

The binding affinities of 07/2a and 07/2a-k to the murine Fcγ receptors (CD16, CD32 and CD64) of J774 macrophages were determined with a conventional ELISA (STC Biologicals, Boston, MA). Recombinant murine CD16, CD32, or CD64 were coated onto plates and bound 07/2a and 07/2a-k antibodies were detected with an anti-mouse Fab-HRP and quantitated by colorimetric readings. Dissociation constant (K_D) values were derived from fitted IC₅₀ values using fixed bottom (0%) and top (100%) values, as well as a slope value of 1.

In vivo microPET imaging was carried out as described in Liu et al. [25] and Additional file 1 (S1.6). Briefly, 4-mo- and 16-mo-old male APP/PS1dE9 mice underwent baseline imaging at day 0 followed by an intravenous tail vein injection of 500 μg of 3A1 mAb, 07/1 mAb, 07/2a mAb, 07/2a-k mAb, or PBS the next day. Follow-up microPET scans were performed 3 days and 30 days post-injection.

With the exception of the behavioral data, statistical analyses were conducted with Prism 5 (GraphPad) using two-way ANOVA with Bonforroni post hoc test for PET imaging analysis or one-way ANOVA with Newman-Keuls post hoc test for all the other group analyses. Where indicated, Student’s t test was performed for some analyses. For behavioral data, StatView (Version 5.0) was used along with Fisher’s PLS. A p value of < 0.05 was considered significant, and all data are expressed as the mean ± SEM, unless otherwise stated.

Behavioral testing was initiated at 15 mo of age, approximately 1 month prior to the mice receiving their 16th and final weekly i.p. injection. To control for non-specific effects on learning and memory, Wt littermate mice received injections of PBS. Following 13 weeks of antibody or PBS administration by i.p injection, mice were placed in an open field arena for measurement of the effects of passive immunotherapy on locomotor activity and anxiety. Total distance traveled and the percent distance traveled in the center of the field was recorded over 60 min. Antibody and PBS-treated Tg mice were compared to PBS-treated Wt control mice. As expected based on our previous studies, mAb and PBS-injected Tg mice were more active (i.e., more total distance traveled) than Wt PBS-injected mice in the first 30 min of the test session; however, no differences were observed between groups during the last 30 min of the test session (Fig. 1a). Therefore, mAb treatment did not affect locomotor activity. There was a significant decrease in percent distance traveled in the center of the open field in the Tg PBS-injected mice compared to Wt PBS controls (Fig. 1b) demonstrating a genotype-specific increase in anxiety-like behavior in APP_SWE/PS1ΔE9 mice at 15 mo of age. There was a strong trend for an increased percent distance traveled in the center in the mice treated with 07/1 mAb (p = 0.54); however, this did not reach significance (Fig. 1b). No differences were observed in the 07/2a-treated Tg mice in the open field testing. CFC testing did not show any differences in fear-associated learning or memory between groups (data not shown).

A water T-maze (WTM) test was used to assess antibody treatment effects on spatial learning and memory and cognitive flexibility in the APP_SWE/PS1ΔE9 mice. The number of correct responses to find the platform made by the PBS-injected Tg mice in the acquisition (p < 0.05) and reversal (p < 0.05) phases was significantly fewer than what was recorded by the Wt PBS-injected mice demonstrating that the APP_SWE/PS1ΔE9 mice had learning and memory deficits at 15 mo of age (Fig. 1c–e). These deficits were attenuated in the Tg mice that were treated with 07/2a mAb, which instead showed a significant increase in the percent of correct responses in both the acquisition and reversal tests compared to the PBS-injected Tg control mice (Fig. 1d, e).

Antibody and PBS-injected mice were sacrificed 1 week following the final injection at 16 mo of age. Tissues were harvested and analyzed for neuropathological and biochemical changes in the CNS and periphery after immunotherapy. In order to assess the potential differences between 07/1 and 07/2a on Aβ clearance in vivo, we examined mouse brain sections using quantitative immunohistochemistry. Aβ plaque burden was assessed in the brains of the five APP_SWE/PS1ΔE9 Tg mice sacrificed at ~ 12 mo of age as baseline controls. Immunostaining of fixed brain tissue sections with 07/2b (pGlu-3 Aβ IgG2b), R1282 (general Aβ), and 82E1 (Aβ1-x) showed immunoreactivity (IR) in the hippocampus (HC) and cortex (CTX) (Fig. 2a–h, m–p) in all APP_SWE/PS1ΔE9 Tg mice. Thioflavin S-positive fibrillar amyloid plaques were also observed in the HC and CTX (Fig. 2i–l). Plaque burden increased from baseline (12 mo) to 16 mo of age.

Pyroglu-3 Aβ IR, measured by 07/2b mAb, to avoid cross-reactivity between the immunizing mAb (07/1 or 07/2a) and secondary antibody, was reduced dramatically by 80% in the HC (p < 0.001) and by 81% in the CTX (p < 0.05) of the 07/2a-treated Tg mice compared to PBS-injected mice (Fig. 2a–d, q, u). 07/2a-specific significant reductions were also observed for general Aβ deposition in the HC (p < 0.05) and CTX (p < 0.001) where R1282 IR was reduced by 20% and 51%, respectively, compared to Tg PBS-injected mice (Fig. 2e–h, r, v). Quantification of Aβ1-x measured by 82E1 IR showed no significant differences between the antibody treatment groups and PBS-injected mice in the HC (Fig. 2t). However, there was a 35% reduction (p < 0.05) of 82E1 IR in the CTX of 07/2a-treated mice compared to PBS-injected mice (Fig. 2m–p, x). Thioflavin S-positive fibrillar amyloid plaques were reduced by 62% (p < 0.05) in HC and 23% (p < 0.05) in CTX in 07/2a-treated Tg mice compared to PBS controls (Fig. 2i–l, s, w). There were no significant differences in Aβ load as analyzed by one-way ANOVA between PBS-injected Tg mice and 07/1 and 3A1 mAb-treated mice in either the HC or CTX.

Passive immunization against Aβ in mouse models has demonstrated an increase in cerebral microhemorrhages hypothesized to be due to movement of parenchymal plaques to the vasculature for clearance following antibody treatment [33, 34] or possibly because of direct binding of antibodies to existing cerebral amyloid angiopathy [35]. Therefore, we sought to quantify the occurrence of microhemorrhages in this study. Hemosiderin staining was performed for the detection of microhemorrhages across the entirety of three equidistant sagittal cross sections of the mouse brain by three independent investigators blinded to the treatment of each mouse. No significant differences were observed between antibody treatment groups and PBS controls (Fig. 3h) indicating that none of the antibodies increased the incidence of microhemorrhage.

Biochemical analysis of guanidine-HCl-extracted Aβ from whole brain (excluding cerebellum) homogenates demonstrated a significantly reduced concentration of pGlu-3Aβ in the insoluble fraction from 07/2a-treated Tg mice compared to Tg PBS control mice (p < 0.05) (Fig. 3a). The levels of insoluble Aβx-42 were significantly lower in the 07/2a- and 3A1-treated mice compared with Tg PBS-injected mice (p < 0.05) (Fig. 3b). In addition, 07/2a-treated mice had significantly lower Aβx-40 (p < 0.05) and Aβx-38 (p < 0.05) in the insoluble Aβ fraction relative to Tg PBS control mice (Fig. 3c, d). A trend for reduction in Aβx-40 and Aβx-42 was observed by 3A1-treated mice compared to PBS-injected Tg mice and was significant when compared to PBS controls by Student’s t test (p < 0.05) (Fig. 3c, d). There were no significant changes in soluble Aβx-40 and Aβx-42 levels in the T-PER extracted Aβ homogenates (data not shown) and the pGlu-3 Aβ and Aβx-38 levels in this fraction were below the detection limit for their respective Aβ ELISAs.

Reduction of Aβ burden by altering the equilibrium between CNS and plasma Aβ, otherwise known as the “peripheral sink” hypothesis, has been demonstrated in previous immunotherapy studies as a mechanism of plaque reduction in the brain [4, 36]. To investigate if there were changes in peripheral Aβ levels in the antibody-treated Tg mice, terminal plasma was collected from all mice and Aβ was measured by an MSD immunoassay. Tg mice treated with 3A1 showed dramatic increases in Aβx-42 (p < 0.01), Aβx-40 (p < 0.001), and Aβx-38 (p < 0.001) levels compared with Tg PBS control mice (Fig. 3e–g). Interestingly, the pGlu-3 Aβ mAbs, 07/1 and 07/2a, showed no significant changes in Aβ levels compared to Tg PBS control mice suggesting less antibody binding in the periphery and, perhaps, an alternative mechanism of plaque removal. Similar to our prevention study [12], we did not see detectable levels of pGlu-3 Aβ in the plasma in these APP_SWE/PS1ΔE9 Tg mice (data not shown). We also examined cytokine and chemokine levels in the plasma of the mice in this study to determine if the treatments had any pro- or anti-inflammatory effects; however, after measuring for changes in IFNγ, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12p70, KC-GRO, and TNFα levels by MSD immunoplatform (Additional file 1: Figure S1), only KC-GRO (also known as CXCL1 chemokine) levels demonstrated a difference with 07/2a-treated mice showing a significant reduction (p < 0.05) compared to PBS-injected transgenic mice (Additional file 1: Figure S1H).

The amount of 07/1, 07/2a, and 3A1 exogenous antibodies present at sacrifice in the CNS and plasma were measured by ELISA in the T-PER soluble brain homogenates (Additional file 1: Figure S2A-C) and plasma (Additional file 1: Figure S2D-F). We observed an increase in plasma pGlu-3 Aβ antibody levels in 07/1 (~ 52,000 ng/mL) and 07/2a-treated mice (~ 58,000 ng/mL) (Additional file 1: Figure S2E-F). No antibodies recognizing pGlu3–12 Aβ were detected in plasma from 3A1 or PBS-injected mice, supporting 07/1 and 3A1 mAb’s specificities (Additional file 1: Figure S2E-F). In contrast, circulating antibodies that recognized Aβ1–18 were present in 3A1-immunized mouse plasma (~ 82,000 ng/mL) (Additional file 1: Figure S2D) but not in either of the pGlu-3 Aβ immunization groups or PBS-injected Tg controls (Additional file 1: Figure S2D). In parallel, increases in exogenous antibody levels, albeit at lower concentrations, were observed in the T-PER soluble brain fractions from saline-perfused mice. This demonstrates that a fraction of the exogenous antibodies administered in the periphery were able to penetrate the CNS. In the brain, ~ 250,000 pg/mL antibodies that recognized pGlu3–12 Aβ from 07/1 immunized Tg mice and ~ 425,000 pg/mL antibodies from 07/2a immunized mice were detected (Additional file 1: Figure S2B-C). No pGlu3-Aβ-specific antibodies were detected in soluble brain homogenates from 3A1-immunized or PBS-injected mice (Additional file 1: Figure S2B-C). Although lower than the amount of Aβ1–18-recognizing antibodies that were detected in the plasma, soluble brain homogenates from 3A1-immunized mice had concentrations of 40,000 pg/mL antibodies (Additional file 1: Figure S2A). There was a very small amount of Aβ1-18-recognizing antibodies in brain in the 07/1-treated mice but no appreciable amounts from 07/2a- or PBS-injected Tg mice detected (Additional file 1: Figure S2A). We also carried out immunohistochemical analysis on brain sections with biotinylated anti-mouse IgG1 and anti-mouse IgG2a secondary antibodies alone to detect the immunizing antibodies in brain. Adjacent sections were stained with R1282 and pGlu-3 Aβ 07/2b antibodies to identify plaques. Biotinylated IgG1 secondary antibody detected a subset of cerebral plaques in 3A1-treated and 07/1-treated mice while biotinylated IgG2a secondary antibody recognized a subset of plaques in 07/2a-treated mice (Additional file 1: Figure S3).

To evaluate if there were changes in glial accumulation in HC and CTX of these mice, IR with a marker for microglia and macrophage, Iba1, was quantified on three sections at equidistant planes. A significant increase in Iba1-positive staining was observed in the mice treated with 07/2a compared to PBS-injected Tg mice in both the HC (p < 0.05) and the CTX (p < 0.05) but not observed in the groups treated with 07/1 and 3A1 (Fig. 4a–e, k, l). We also measured IR in these treatment groups with a microglial and macrophage marker of phagocytosis, CD68. We observed from analysis of this phagocytic marker that 07/2a was the only mAb treatment to demonstrate a significant increase compared to PBS control Tg mice in the HC (Fig. 4m; p < 0.01) and CTX (Fig. 4n; p < 0.05). We carried out double immunofluorescence (IF) staining on the tissue with a general Aβ marker, R1282, and CD68 and measured, specifically in the hippocampus, the number of CD68-positive cells surrounding plaques. We found that 07/2a-treated Tg mice had a significant amount of CD68-positive, plaque-associated cells as semi-quantitatively measured compared to the PBS control mice (Fig. 4o, s).

In order to address the mechanism of plaque removal by 07/1, 07/2a, and 07/2a-k by potentially mediating microglial phagocytosis of Aβ, we assessed plaque clearance by microglia from unfixed, frozen brain tissue of aged APP_SWE/PS1ΔE9 Tg mice in an ex vivo phagocytosis assay. Previous studies have demonstrated that Aβ antibody added to tissue sections, binds to plaques, and initiates plaque clearance through Fc-receptor-mediated phagocytosis [4, 5]. Prior to initiation of the assay, we first determined whether the pGlu-3-42 Aβ epitope was accessible in the unfixed frozen tissue from a 20-mo-old APP_SWE/PS1ΔE9 mouse. Robust immunofluorescence detection with 07/1 mAb was observed in the cortex and hippocampus, with staining to a lesser extent observed in the cerebellum (Additional file 1: Figure S4A-C), suggesting sufficient pGlu-3-42 antigen was present to allow for Fc-receptor-mediated phagocytosis. Ex vivo phagocytosis assays were conducted, and Aβ plaque load/levels were measured by either immunofluorescent staining or ELISA, and the results were expressed as a ratio of the relevant isotype control. There were no significant differences in R1282 IR (data not shown) between tissue that had 24 h incubation with media alone (no cells) (Fig. 5a) and incubation with PMG (Fig. 5b), suggesting that without the addition of an Aβ antibody, phagocytosis is negligible. Similar R1282 IR levels were also seen in the tissue pre-incubated with IgG isotype control antibodies and PMG compared with PMG alone, eliminating non-specific Fc-receptor binding initiating clearance (Fig. 5d, g). Following 24 h incubation of the tissue with PMG, there was a significant reduction in R1282-positive staining present in the tissue pre-incubated with 07/2a mAb compared to incubation with IgG2a isotype control (p < 0.001) whereas exposure with the pGlu-3 Aβ IgG1 and 07/2a-k mAbs did not promote a significant reduction (Fig. 5a–h). Analysis of the guanidine-HCl insoluble Aβ homogenates by ELISA demonstrated that there was a significant decrease in pGlu-3 Aβ following incubation with the tissue with both 07/1 (p < 0.05) and 07/2a (p < 0.05) (Fig. 5i). Aβx-40 levels were only reduced with exogenous addition of the IgG2a antibodies, 07/2a (p < 0.05) and 07/2a-k (p < 0.01) (Fig. 5j), whereas there were no significant differences in Aβx-42 levels observed following incubation of any of the Aβ mAbs in the ex vivo phagocytosis assay (Fig. 5k). These results were mostly confirmed in another set of ex vivo phagocytosis assays using N9 cells, a murine immortalized microglial cell line (Additional file 1: Figure S5).

An accompanying in vitro assessment of the Fcγ receptor binding of 07/2a and 07/2a-k to isolated CD16, CD32, and CD64, i.e., the Fcγ receptor components, revealed no profound differences of the two antibodies (Table 1). In addition, binding affinities to pGlu-3 Aβ were similar between 07/1 (10.2 nM), 07/2a (9.8 nM), and 07/2a-k (8.08 nM).

We next compared two doses of the 07/2a antibody (150 and 500 μg) and one dose of 07/2a-k (500 μg) for their ability to clear Aβ in a plaque-rich AD model. Overall, Aβ clearance was similar between 07/2a-k, compared to that seen with 07/2a (Fig. 6). pGlu-3 Aβ (p < 0.05) and general Aβ (p < 0.01) immunoreactivity in the hippocampus showed significant reductions by 07/2a-k, similar to that of 07/2a-treated mice (Fig. 6a, b). The plaque-lowering potential of the CDC mutant antibody was further confirmed in the TBS soluble fraction isolated from the hemibrain of the treated mice and potentially containing toxic oligomers (Fig. 6c), although insoluble Aβ levels were not significantly altered. 07/2a IgG2a immunotherapy in these mice demonstrated a dose-dependent effect in Aβ reduction (Fig. 6a–c).

Different Ig isotypes of antibody used in passive immunization studies can have various influences on microglial activation, which may result in different mechanisms of Aβ clearance. Thus, it is important to compare different Ig isotypes against the same target for their effect on clearing Aβ and inducing transient and/or chronic neuroinflammation. We performed microPET imaging using a second-generation TSPO (translocator protein, a marker for activated microglia) PET tracer, ¹⁸F-GE180, to visualize the differences in microglial activation prior (baseline, day 0) to and following i.p. passive immunization (days 3 and 30) with 3 anti-pyroGlu3 Aβ mAbs (07/1 IgG1, 07/2a IgG2a, and 07/2a-k, a CDC mutant version of 07/2a IgG2a to avoid complement activation), a general Aβ IgG1 mAb (3A1), or PBS in young (pre-plaque, 4-mo-old) or aged (plaque-rich, 14–16-mo-old) male APP_SWE/PS1ΔE9 Tg mice.

The time-activity curve (TAC) of ¹⁸F-GE180 uptake in the hippocampus was analyzed using CT/PET/MRI co-registration. There were no differences in hippocampal TAC of ¹⁸F-GE180 at day 0 (baseline), 3, or 30 in 4-mo-old and 16-mo-old APP_SWE/PS1ΔE9 mice with PBS injection (Fig. 7a, f). Following a single injection of 07/1 mAb, a significant increase in hippocampal ¹⁸F-GE180 uptake was found at day 30 versus baseline in 4-mo-old (F_(1,8) = 5.62; p < 0.05) (Fig. 7c) and 16-mo-old (F_(1,8) = 10.54; p < 0.05) mice (Fig. 7h). Four-month-old APP/PS1dE9 mice injected with 07/2a antibody showed overt elevation in ¹⁸F-GE180 uptake at day 3 (F_(1,8) = 3.52; p = 0.097, trend) and day 30 (F_(1,8) = 5.62; p < 0.05) compared to baseline (Fig. 7d), while 16-mo-old APP/PS1dE9 mice treated with the same antibody showed a significant increase in hippocampal ¹⁸F-GE180 uptake at day 3 (F_(1,8) = 7.28; p < 0.05) but a decrease at day 30 (F_(1,8) = 15.84; p < 0.01) (Fig. 7i). No significant changes in hippocampal ¹⁸F-GE180 uptake were observed at day 3 or day 30 at either age in mice treated with 07/2a-k mAb (Fig. 7e, j). Although the 3A1-injected 4-mo-old APP/PS1dE9 mice showed a consistent reduction of ¹⁸F-GE180 hippocampal uptake at day 3 (F_(1,8) = 18.0; p < 0.01) and 30 versus baseline (F_(1,8) = 3.77; p = 0.088, trend), these mice had abnormally high baseline tracer uptake (Fig. 7b). Similar results were observed in the whole brain TAC of ¹⁸F-GE180 uptake (Additional file 1: Figure S6).