• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Mar 17, 2009; 106(11): 4501–4506.
Published online Feb 25, 2009. doi:  10.1073/pnas.0813404106
PMCID: PMC2647980

Intracerebroventricular amyloid-β antibodies reduce cerebral amyloid angiopathy and associated micro-hemorrhages in aged Tg2576 mice


Although immunization against amyloid-β (Aβ) holds promise as a disease-modifying therapy for Alzheimer disease (AD), it is associated with an undesirable accumulation of amyloid in the cerebrovasculature [i.e., cerebral amyloid angiopathy (CAA)] and a heightened risk of micro-hemorrhages. The central and peripheral mechanisms postulated to modulate amyloid with anti-Aβ immunotherapy remain largely elusive. Here, we compared the effects of prolonged intracerebroventricular (icv) versus systemic delivery of anti-Aβ antibodies on the behavioral and pathological changes in an aged Tg2576 mouse model of AD. Prolonged icv infusions of anti-Aβ antibodies dose-dependently reduced the parenchymal plaque burden, astrogliosis, and dystrophic neurites at doses 10- to 50-fold lower than used with systemic delivery of the same antibody. Both icv and systemic anti-Aβ antibodies reversed the behavioral impairment in contextual fear conditioning. More importantly, unlike systemically delivered anti-Aβ antibodies that aggravated vascular pathology, icv-infused antibodies globally reduced CAA and associated micro-hemorrhages. We present data suggesting that the divergent effects of icv-delivered anti-Aβ antibodies result from gradually engaging the local (i.e., central) mechanisms for amyloid clearance, distinct from the mechanisms engaged by high doses of anti-Aβ antibodies that circulate in the vasculature following systemic delivery. With robust efficacy in reversing AD-related pathology and an unexpected benefit in reducing CAA and associated micro-hemorrhages, icv-targeted passive immunotherapy offers a promising therapeutic approach for the long-term management of AD.

Keywords: Alzheimer disease, passive immunotherapy, vascular amyloid, microglia, peripheral sink

Accumulation of the amyloid-β (Aβ) peptide plays a pivotal role in the pathogenesis of Alzheimer disease (AD) (1, 2; also see ref. 3). Accordingly, immunization against Aβ has offered a promising approach toward the therapeutic management of AD (46). In animal models of AD, both active and passive anti-Aβ immunotherapies improve cognitive function and clear the parenchymal accumulation of amyloid (plaques) in the brain (713). The discontinued phase II trial of active immunization with aggregated Aβ as an immunogen (AN1792) along with the QS-21 adjuvant demonstrated some clinical benefit in AD subjects who developed a robust anti-Aβ antibody response (14, 15). The autopsy reports of 4 subjects from this trial who survived up to 20 months after their first immunization revealed focal clearance of parenchymal plaques in the cortex (1619). In contrast, the accumulation of amyloid in cerebral blood vessels, also known as cerebral amyloid angiopathy (CAA), was more severe than observed in non-immunized patients at a similar stage of AD, with multiple cortical hemorrhages evident in one case (17, 19, 20). Recent preclinical studies illustrate that passive immunizations with antibodies that target amyloid plaques and reverse the cognitive deficit also increase CAA in transgenic mouse models of AD (21, 22). In addition, CAA-associated micro-hemorrhages were augmented with such immunizations in transgenic AD models (2124). These effects of anti-Aβ immunotherapy present a significant risk in light of the fact that CAA is associated with pathological abnormalities and cognitive impairment in the elderly population, and is prevalent in >80% of patients with AD (25, 26).

The mechanisms proposed for the effects of anti-Aβ immunotherapy on amyloid accumulation remain elusive, generally relying on systemic regimens that yield high titers of anti-Aβ antibodies in the peripheral circulation (6, 27). High antibody titers bind to substantial amounts of Aβ in the bloodstream, which is believed to shift the overall Aβ equilibrium and create a “peripheral sink” that facilitates an efflux of Aβ from the brain (11, 28). Further, high doses of antibody in the periphery are required because of the low-level (≤0.1%) penetration of antibody across the blood–brain barrier to effectively engage the local (i.e., central) mechanisms for clearing the cerebral amyloid (10, 27, 29). Recent studies demonstrate that a single intracerebroventricular (icv) injection of anti-Aβ antibodies is able to prevent the Aβ-induced impairment of synaptic plasticity in the hippocampus (30), and also transiently reverse the memory deficit in a transgenic AD mouse model (31). Therefore, to gain further insight into the actions of anti-Aβ antibodies when present in the cerebrospinal fluid (CSF), we compared the effects of prolonged icv versus systemic delivery of antibodies on the behavioral as well as pathological changes in the aged Tg2576 mouse model of AD. Our results not only help to better define the mechanisms underlying immunotherapy-induced changes in amyloid, but also point to icv delivery as a superior therapeutic route for delivering anti-Aβ antibodies to the brain that can significantly reverse behavioral deficits and reduce AD-related pathological changes, and importantly, also reduce CAA and associated micro-hemorrhages.


icv And Systemic Anti-Aβ Antibodies Reverse Cognitive Decline and Clear the Parenchymal Plaques as Well as Associated Neuropathology in Aged Tg2576 Mice.

We used a mouse monoclonal IgG1, 6E10, that recognizes the N terminus of human Aβ and binds to the monomer, parenchymal plaques, and CAA (19, 32, 33). In this respect, the Aβ-binding properties of 6E10 are similar to the properties of anti-Aβ antibodies generated in subjects immunized with AN1972 in the aforementioned active immunotherapy trial (14, 34, 35). In addition, 6E10 targets the extra-neuronal soluble oligomer Aβ*56 and intraneuronal Aβ, both of which are implicated in the decline of cognitive function (31, 36, 37).

Male 16- to 18-month-old Tg2576 mice were implanted with osmotic mini-pumps to enable prolonged icv infusion of 6E10 (anti-Aβ IgG1) or a non-relevant isotype-control antibody (control IgG1). In vitro, pump-mediated release of 6E10 was verified to be consistent and stable over the course of 5 weeks [supporting information (SI) Fig. S1]. A total of 0.2 mg (at a maximum concentration of 1 mg/mL) or 0.04 mg (diluted to 0.2 mg/mL) 6E10 was delivered icv in the study. Other groups of Tg2576 mice received weekly i.p. injections of control IgG1 or 6E10, at a dose of 10 mg/kg (22, 38), for systemic delivery of a total of 2 mg antibody, over 5.5 weeks. At termination, plasma levels of 6E10 were 30.2 ± 4.5 μg/mL (mean ± SEM; n = 8) for the systemic group but were below the limits of ELISA detection (≤0.01 μg/mL) in the icv treatment groups.

Two days before euthanizing the animals, all mice were assessed for contextual memory by using a fear-conditioning paradigm that reveals a deficit in the Tg2576 mice as early as 9 months of age (32). As expected, aged Tg2576 mice that received control IgG1 via either systemic or icv routes demonstrated a robust deficit in contextual memory compared with WT (Fig. 1). In contrast, the transgenic mice systemically or centrally (i.e., icv) treated with 6E10 exhibited a reversal of memory deficit in this paradigm (Fig. 1). When assessed for cued fear-conditioning, the control IgG1-treated transgenic mice were equivalent to WT (P = 0.971), as also reported earlier (32); none of the 6E10 treatments influenced this response. Collectively, genotype- and treatment-induced changes in fear conditioning were specific to the context and not caused by an inability of mice to detect the cue (i.e., conditioned stimulus), foot-shock (i.e., unconditioned stimulus), or to exhibit a freezing response. The Tg2576 mice showed a trend for hyper-locomotion relative to WT (P = 0.085); this activity was also unaltered by any of the systemic or icv treatment groups (data not shown).

Fig. 1.
Behavioral improvement upon icv infusion of low-dose anti-Aβ IgG1 or systemic delivery of a relatively high dose of the same IgG in aged Tg2576 mice. Aged (16–18 months) Tg2576 mice received control IgG1 (white bars) or 6E10 (gray bars) ...

Treatment-induced changes in amyloid accumulation were quantified using a series of sections, per mouse brain, stained with the Campbell-Switzer protocol to reveal diffuse as well as dense-core fibrillar parenchymal plaques (19, 39, 40). In the transgenic mice treated with control IgG1, plaques were abundant throughout the cerebral cortex (Fig. 2A, D, and F), with intense staining evident in the frontal and entorhinal cortices, as previously described for Tg2576 mice of this age (41). The hippocampus was also extensively decorated by plaques, with a high density observed in the cornu ammonis and subiculum areas (Fig. 2, B, D, and F). Systemic treatment with 6E10 resulted in a significant reduction of parenchymal plaques in the cerebral cortex (by 57%) and hippocampus (by 52%) of the transgenic mice (Fig. 2). Similarly, icv infusion of 10- to 50-fold lower doses of 6E10 dose-dependently cleared the parenchymal plaque burden from the cerebral cortex (up to 64%) as well as the hippocampus (up to 62%) of the Tg2576 mice (Fig. 2). Regardless of the route of delivery, 6E10-mediated clearance of parenchymal plaques was bilateral and uniform across the two hemispheres of the brain (Fig. S2). Densitometric analysis of only the dense-core fibrillar plaques revealed that such plaques represented ≤10% of the total parenchymal plaques and were also significantly reduced in the cerebral cortex and hippocampus of transgenic mice following systemic or icv treatments with 6E10 (Fig. S3).

Fig. 2.
Decrease in parenchymal plaques following prolonged icv or systemic treatment with anti-Aβ IgG1 in aged Tg2576 mice. Transgenic mice received control IgG1 (white bars) or 6E10 (gray bars) via systemic or icv routes. Numbers within bars denote ...

Consistent with the clearance of parenchymal plaques, we observed a significant decrease in astrocytic clusters surrounding the dense-core fibrillar parenchymal plaques (by 29%) in mice that received systemic injections of 6E10 (Fig. S4). Likewise, a significant and dose-dependent reduction in astrocyte clusters (up to 47%) was evident in mice that received prolonged icv administrations of 6E10. Quantitative analysis of the dystrophic neurites that comprises of swollen distorted axons and dendrites surrounding the dense-core fibrillar parenchymal plaques also demonstrated a significant reduction (up to 47%) in both the icv and systemic 6E10 groups (Fig. S4).

Prolonged Treatment with icv Versus Systemic Anti-Aβ Antibodies Produces Contrasting Changes in CAA and Associated Micro-Hemorrhages.

In addition to developing several other manifestations similar to AD (41, 42), Tg2576 mice represent one of the well characterized models for an age-dependent progression of CAA, exhibiting a robust vascular pathology at ≥16 months of age (43, 44). CAA was clearly discernible as multifocal and circumferential deposition of amyloid within blood vessels residing in the leptomeninges, parenchyma, and hippocampal-thalamic interface of the Tg2576 mouse brain. Quantification was performed using the same series of brain sections that were stained to reveal both parenchymal plaques and CAA, as previously described (45) (SI Text). Systemic treatment with 6E10 markedly increased CAA throughout the cerebral cortex and hippocampus by 55% and 70%, respectively (Fig. 3 A–D). In contrast, icv delivery of 6E10 globally reduced CAA both in the cerebral cortex and hippocampus by up to 45% and 40%, respectively (Fig. 3 A–C and E). Both the systemic and icv 6E10 produced uniform changes in CAA across the two cerebral hemispheres (Fig. S5).

Fig. 3.
Contrasting effects of icv versus systemic anti-Aβ IgG1 on CAA in aged Tg2576 mice. Mice chronically received control IgG1 (white bars) or 6E10 (gray bars) via systemic or icv routes. Numbers within bars denote the total dose in milligrams. Quantitative ...

Analogous to its influence on CAA, systemic 6E10 significantly increased the frequency of cerebral hemorrhages by 96% compared with the corresponding control IgG1 group (P < 0.001; Fig. S6). In contrast, the frequency of cerebral micro-hemorrhages was either unchanged (0.2-mg dose) or significantly decreased by 28% (0.04-mg dose; P < 0.05) following icv infusions of 6E10 (Fig. S6). The severity of hemorrhages was also reduced by 21% with icv infusion of the 0.04-mg dose of 6E10 (P < 0.05; Fig. S6).

icv-Delivered Anti-Aβ Antibodies Engage Central, but Not Peripheral, Mechanisms for Clearing the Cerebral Amyloid.

Microglial activation was assessed based on the postulated mechanism that anti-Aβ antibody penetrates into the brain and stimulates the recruitment of microglia for phagocytosis of Aβ in plaques (10, 22, 38, 46). Parenchymal penetration of icv- or systemically delivered 6E10 was confirmed by co-localization of immunohistochemical staining for IgG with Congo red delineation of amyloid plaques (Fig. 4 A and B). Subsequently, we immunostained the sections for ionized calcium-binding adaptor molecule 1, which revealed activated microglia as intensely stained cells with relatively short processes surrounding the deposited amyloid in control IgG1-treated transgenic mice (Fig. 4D). Microglial activation was quantified as the ratio of percent cortical area occupied by active microglia to the percent cortical area occupied by plaques (38). A significant induction (70%) of plaque-enclosing reactive microglia was observed upon systemic treatment with 6E10 relative to control IgG1 (Fig. 4C). Likewise, icv infusion of 6E10 resulted in a dose-dependent recruitment of active microglia around plaques by 71% and 137% at 0.04-mg and 0.2-mg doses, respectively (Fig. 4 C–E). Analysis of sections stained for another microglial marker (CD68) also revealed a significant increase in microglial activation around plaques following icv or systemic treatment with 6E10 (Fig. S7). Given these results, it seems unlikely that microglial activation contributes toward the contrasting effects of icv versus systemic anti-Aβ immunotherapy on CAA and cerebral hemorrhages.

Fig. 4.
Central and/or peripheral actions of icv versus systemically delivered anti-Aβ IgG1. Tg2576 mice received control IgG1 (white bars) or 6E10 (gray bars) via systemic or icv routes. Numbers within or above bars denote the total dose in milligrams. ...

We evaluated a possible role for the “peripheral sink” mechanism, wherein an anti-Aβ antibody facilitates efflux of cerebral Aβ across the microvasculature of the brain into the systemic circulation (11, 28). Accordingly, we collected plasma samples immediately before euthanizing the mice and performed an ELISA that detected Aβ, both free and bound to the in vivo-administered 6E10. Interestingly, plasma Aβ was significantly elevated by systemic, but not icv, treatment with 6E10 (Fig. 4F), suggesting that vascular clearance of cerebral Aβ is related to the susceptibility to CAA and associated micro-hemorrhages.

The Rate of Parenchymal Plaque Clearance Influences CAA and Associated Micro-Hemorrhages.

Given the likelihood that immunotherapy-solubilized amyloid redistributes from the parenchymal plaques to the cerebral vasculature (19, 21, 47), it is possible that the amyloid clearance from prolonged icv infusions of low-dose anti-Aβ antibodies is sufficiently gradual that it fails to substantially accumulate in the cerebral vasculature and result in CAA. We aimed to test this hypothesis by delivering bolus icv injections of relatively high doses of anti-Aβ antibodies to briefly flood the CSF and maximize the movement of antibody into the parenchyma for clearance of amyloid (48, 49). Aged Tg2576 mice received a single icv injection of 6E10 at a dose (0.5 μg) calculated to acutely provide an antibody concentration in the CSF equivalent to that maintained during the prolonged infusion study. Additional groups of transgenic mice received an acute icv injection of 12- to 24-fold higher doses of 6E10, 6 and 12 μg, respectively. Mice were euthanized at 1, 4, or 8 weeks after injection and serial brain sections were analyzed for parenchymal plaques, CAA, and associated micro-hemorrhages. A dose- and time-dependent decrease in parenchymal plaques was observed relative to those in control IgG1-treated transgenic mice (Fig. S8). The lowest dose (0.5 μg) of 6E10 injected into the CSF did not alter the parenchymal plaque burden, even in the first week after injection. In contrast, 6 μg 6E10 significantly reduced parenchymal plaques by 51% (P < 0.01) and 42% (P < 0.05) at 1 and 4 weeks after injection, respectively, but had no effect at 8 weeks. The highest injected dose (12 μg) of 6E10 resulted in significant reductions of parenchymal plaques by 72% (P < 0.001), 51% (P < 0.01), and 24% (P < 0.001) at 1, 4, and 8 weeks after injection, respectively. Analysis of the dense-core fibrillar plaques revealed a significant reduction (by 74%; P < 0.05) only at 1 week after the bolus injection of the highest dose of 6E10 (Fig. S8). Of interest, CAA was also significantly increased (by 68%; P < 0.05) at 1 week after the injection of the highest dose of 6E10; this effect reduced at 4 weeks, and was completely resolved at 8 weeks after injection (Fig. S8). Corresponding to the increase in CAA, a transient increase in cerebral micro-hemorrhages (by 107%; P < 0.01) was noted only with the highest dose of 6E10 injected into the CSF. Irrespective of the treatment, Aβ levels were unaltered and 6E10 was undetectable in the plasma samples collected during euthanasia of mice.


Therapeutic strategies to treat CAA (25, 26, 50) are warranted because of the high prevalence of CAA (>80%) in patients with AD, and its adverse consequences. We demonstrate that, in striking contrast to an increase in CAA and micro-hemorrhages occurring with systemic delivery of the Aβ N terminus-binding IgG1, icv infusion of the same antibody reduces CAA and associated micro-hemorrhages.

Preliminary data from the phase I and II AN-1972 clinical trials suggest that immunotherapy may be effective early in AD, well before amyloid is extensively deposited and an advanced degree of vasculature is compromised (19, 51). Recent preclinical studies provide evidence that a reduction of both parenchymal plaques and CAA is achieved upon icv injection of an adeno-associated viral vector that expresses anti-Aβ single-chain variable fragments in neonatal transgenic CRND8 mice (52), or by local application of antibodies onto the cortex of Tg2576 mice at 10 to 13 months of age, when CAA follows a linear mode of progression (53). Our experiments in aged Tg2576 mice, with near-saturating levels of CAA and an abundant plaque pathology (4144), augment these findings and demonstrate that icv-targeted delivery of anti-Aβ antibodies not only reduces the behavioral deficit, parenchymal plaques, and associated pathology (eg, astrogliosis, dystrophic neurites), but also provides a previously unexpected opportunity to mitigate CAA and associated micro-hemorrhages despite the advanced disease stage.

We attempted to elucidate why systemic versus icv delivery of anti-Aβ antibodies had contrasting effects on CAA and cerebral hemorrhages, although both reversed cognitive decline and reduced the parenchymal plaque burden. Consistent with a central mechanism of action, both systemic- and icv-delivered anti-Aβ antibodies co-localized with amyloid plaques in the brain. Activated microglia were present around plaques in both treatment groups, as observed in autopsy reports from the active immunization clinical trial (1618, 20). Several studies have proposed a microglial phagocytosis mechanism of amyloid clearance, in which antibodies bound to Aβ also interact with Fc receptors (FcR) on microglial cells to initiate microglial phagocytosis of antibody-bound Aβ and subsequently clear amyloid plaques (10, 38, 46, 54). However, we and others have previously demonstrated a reduction of the parenchymal plaque burden by mechanisms that are independent of microglial function (5558). In the current study, microglia were activated with both systemic- and icv-delivered anti-Aβ IgG1, and may partially mediate the clearance of parenchymal plaques in both cases. However, as CAA was increased with systemic administration and decreased with icv administration of anti-Aβ IgG1, a significant role for microglia in mediating the effects of systemic or icv anti-Aβ antibodies on CAA appears unlikely.

Our results help to differentiate to what degree systemic versus icv delivery of antibodies engages the peripheral sink mechanism, wherein increases in plasma Aβ may reflect an efflux of cerebral Aβ across the microvasculature of the brain into the systemic circulation (11, 28). Whereas both systemic- and icv-delivered anti-Aβ antibodies entered the brain parenchyma, only the systemic-delivered antibodies were detected at substantial levels in the plasma to engage the peripheral sink mechanism. That said, a recent study has suggested that an increase in plasma Aβ (postulated to represent an efflux of cerebral Aβ) may simply be caused by an increase in the half-life of plasma Aβ when bound to a systemically administered anti-Aβ antibody (27). Regardless of whether the peripheral sink mechanism mediates the outcomes of systemic anti-Aβ immunotherapy, it is unlikely to play a role in the reduction of parenchymal plaques and CAA by icv-infused anti-Aβ IgG1. If anything, antibodies circulating in the CSF may create a “central sink” that establishes a mass action gradient to pull Aβ from the brain. The Aβ-antibody complex would presumably be resorbed within the CSF via the arachnoid granulations, a fundamentally different pathway that spares the cerebral vasculature except when administered at very high doses, as demonstrated from the bolus administration experiments.

Systemically delivered anti-Aβ antibodies are also likely to clear cerebral amyloid via the CSF route, although to a much lesser extent than the icv-delivered antibodies. Recent studies have demonstrated a spike in CSF anti-Aβ antibody concentration, which represents only a small fraction (≤1%) of the plasma concentration achieved following systemic delivery (27, 59). In our study, plasma concentrations of ≈200 nM were achieved after systemic delivery of 6E10. Accordingly, antibody concentrations in the CSF would be estimated at ≤2 nM in the systemic group, less than the CSF antibody concentrations of ≈10 nM or 50 nM maintained in the 2 prolonged icv infusion groups. Whereas the small fraction of systemically delivered anti-Aβ antibodies in the CSF may participate in amyloid clearance, antibodies circulating in the vasculature may be responsible for the increase in CAA. Although their overall penetration in the brain is negligible (≤0.1%) (10, 27, 29), amyloid-binding anti-Aβ antibodies, when circulating in the vasculature, may be able to associate with CAA (24). Such association could increase with increasing concentrations of antibody in the vasculature and/or increasing levels of preexisting CAA. The resulting antibody-CAA complex may then serve as a “seed” for mobilizing Aβ from the parenchyma and/or the systemic circulation to further enhance CAA and associated micro-hemorrhages. In agreement with this notion, Racke et al. (24) found that CAA-associated micro-hemorrhages were increased in aged PDAPP transgenic mice upon systemic delivery of an N terminus anti-Aβ antibody that bound to CAA but not by a mid-domain antibody that failed to bind CAA. Here, we demonstrate that icv delivery of a CAA-binding N terminus anti-Aβ antibody effectively circumvents the requirement of achieving high antibody concentrations in the vasculature for therapeutic efficacy, and more importantly, reverses the induction of CAA and associated micro-hemorrhages that is otherwise observed with systemic delivery of the same antibody in aged Tg2576 transgenic mice.

The contrasting results of systemic versus icv delivery of anti-Aβ antibodies reveal some of the complexity in regulating CAA in the context of parenchymal plaque clearance. Although the Aβ composition of parenchymal plaques typically differs from that of CAA, the antibody-solubilized Aβ may redistribute from the parenchymal plaques to the cerebral vasculature in the process of its perivascular efflux from the brain (19, 21, 47). The autopsy reports of AN1792-immunized patients provide some supporting evidence, in that the Aβ composition of their leptomeningial and cortical vasculature was found to be more reminiscent of Aβ in the parenchymal plaques (19). Further, an increase in CAA was noted only in specific brain areas that exhibited the highest extent of parenchymal plaque clearance (19, 20). In Tg2576 mice, a time-dependent reduction of dense-core fibrillar parenchymal plaques has been shown to parallel a time-dependent induction of CAA (21). Similarly, we observed a transient increase in CAA upon enhancing the clearance of parenchymal plaques, especially the dense-core fibrillar plaques, by acutely injecting a high dose of 6E10 into the CSF. Conversely, CAA was significantly decreased upon long-term icv infusions of much lower doses of the same antibody, suggestive of a relatively gradual clearance of amyloid (from both the parenchymal plaques and CAA) with this protocol of icv antibody delivery.

icv-Delivered anti-Aβ antibodies may engage a combination of central, rather than peripheral, mechanisms to reduce the parenchymal plaque burden and CAA. One mechanism may be the FcR-mediated microglial phagocytosis. Another may be an interaction of FcRs on the cerebral microvasculature (neonatal FcRs) with the antibody-bound Aβ, facilitating its clearance from the brain (60). However, given the efficacy of the F(ab′)2 fragment alone (55), FcRn may only partially mediate the effects of icv anti-Aβ antibodies. An earlier study precludes another postulated mechanism, namely disaggregation of amyloid, for clearance by 6E10 (61). Rather, it is likely that icv antibodies enter the brain parenchyma and alter a select pool of Aβ species, such as Aβ*56 and intraneuronal Aβ (31, 36), which in addition to influencing cognitive function, may also affect plaque formation. A reduction in Aβ*56 has been previously demonstrated upon central administration of anti-Aβ antibodies in the Tg2576 mice (62). Although intraneuronal Aβ is difficult to detect in the Tg2576 mice, future studies in the 3xTg-AD mice that substantially accumulate this species of Aβ would help determine its relevance in the outcomes of icv anti-Aβ immunotherapy (31).

In summary, we demonstrate that prolonged icv infusion of anti-Aβ IgG1 effectively alleviates the behavioral and pathological impairments in the aged Tg2576 mouse model of AD. More importantly, icv delivery of anti-Aβ IgG1 provides an unexpected benefit in reducing CAA and associated micro-hemorrhages, which present a significant concern in the use of systemic immunotherapy for treating AD. Given its efficacy at substantially lower doses, icv-targeted passive immunotherapy would offer a superior approach for the long-term management of AD and potentially other neurodegenerative disorders, such as Parkinson disease, in which site-directed delivery of neurotherapeutic agents is warranted.

Materials and Methods

Mice and Experimental Design.

Male Tg2576 mice on B6/SJL background were acquired along with corresponding WT (41; Taconic), and were single-caged in a humidity- and temperature-controlled room with 12-h light/dark cycle. All procedures were performed during the light cycle and were approved by an institutional animal care and use committee at WuXi AppTec, authorized by the American Association for the Accreditation of Laboratory Animal Care.

Forty transgenic mice of age 16 to 18 months were equally distributed among 5 treatment groups in the prolonged icv versus systemic antibody study. Ten WT mice served as phenotype controls for behavioral and pathological outcomes. In addition, 44 transgenic mice of age 12 to 16 months were evenly assigned among 11 treatment groups in the acute icv antibody study.

Prolonged icv infusions were performed using osmotic mini-pumps (Alzet model 2004; Durect) filled with preservative-free monoclonal control IgG1 (that recognizes domain 1 of the rat but not mouse CD4 protein) or 6E10 (Covance) and surgically implanted in mice as previously described (63). The brain-infusion cannula, connected to the osmotic mini-pump via a catheter, was stereotaxically implanted for infusion into the lateral cerebral ventricle (anteroposterior, −0.4 mm; mediolateral, 1 mm; dorsoventral, −2.3 mm relative to bregma). Acute icv injections were also stereotaxically performed using the same coordinates, with the help of a pump-driven Hamilton microsyringe (no. 7654–01) fitted with a 33-gauge stainless-steel needle. In this case, antibodies were infused in a volume of 12 μL at 0.5 μL/min. Mice in the systemic groups received weekly i.p. injections of 10 mg/kg control IgG1 or 6E10.

At termination (4 days after the last injection in the systemic treatment group), mice were anesthetized by i.p. injection of a mixture of ketamine (100 mg/kg) and xylazine (10 mg/kg). The osmotic mini-pump/catheter assembly, if any, was explanted to verify the infusion of IgG1 for an additional 2 days in vitro. A blood sample was withdrawn for ELISAs, and mice were then trans-cardially perfused with ice-cold saline solution followed by paraformaldehyde (4% in saline solution) to harvest organs for histologic examination.

Behavioral Assessment.

Two days before termination, a fear-conditioning protocol with 5 pairings of conditioned-unconditioned stimuli was used for assessment of contextual and cued memory, as previously described (32) (SI Text). Locomotor activity was assessed by placing the mice in a cubical arena (40 cm3) and measuring the horizontal distance traveled for up to 1 h using the Limelight video tracking system (Coulbourn Instruments).

Histological Analyses.

Brains were processed using MultiBrain Technology (NeuroScience Associates) and coronally sectioned at 35 μm thickness. A series of 18 to 20 sections, equally spaced at 350-μm intervals across the entire cerebrum, were used for each stain. Accordingly, analysis for each stain in the cerebral cortex involved 18 to 20 observations per hemisphere per animal, and for the hippocampus involved 7 to 9 observations per hemisphere per animal.

The Campbell-Switzer stain was used to visualize amyloid plaques in the parenchyma and CAA. Positive staining for plaques was verified using immunohistochemical detection of Aβ (rabbit polyclonal anti-pan-Aβ, 1:12,000; Biosource) and the Congo red stain. Diaminobenzidine-enhanced Perls hemosiderin and deOlmos amino-cupric silver stains were used for analyses of cerebral micro-hemorrhages and dystrophic neurites, respectively (24, 64). Astrocytosis and microglia activation around amyloid plaques were visualized with immunohistochemistry using rabbit polyclonal anti-glial fibrillary acidic protein (1:7,500; Dako), and rabbit polyclonal anti-ionized calcium-binding adaptor molecule 1 (1:15,000; Wako) or biotin-conjugated rat monoclonal anti-CD68 (1:400; AbD Serotec), respectively. Immunohistochemical detection of IgG counterstained with Congo red was performed as previously described (38) to reveal in vivo-administered 6E10 bound to amyloid plaques.

The Visiopharm Integrator System (Visiopharm) was used to quantify the percent area of cerebral cortex or hippocampus occupied by the stain depicting parenchymal plaques, CAA, dystrophic neurites, or activated microglia in brain sections. Contours were drawn to delineate the cerebral cortex and hippocampus in each hemisphere of the brain. The threshold for specificity of signal (based on density of staining) was determined and kept constant throughout analyses for each stain. Two individuals scored the number of astrocytic events surrounding amyloid plaques, and cerebral hemorrhages. The protocol for quantifying cerebral hemorrhages was similar to that previously described (23) (SI Text).


Diluted plasma samples were assayed for Aβ in triplicate using the BetaMark x-40 ELISA kit (Covance). In vivo-administered 6E10, if any, was confirmed not to interfere with the kit for analyses of Aβ in these samples. The ELISA for quantifying 6E10 levels (SI Text) was also previously verified in the presence of Aβ spiked up to 4-fold higher levels than detected in the plasma samples from this study.

Statistical Analyses.

All analyses were performed blind. An average score was used when 2 individuals scored the outcomes; the inter-scorer reliability ranged from 0.801 to 0.972 (P < 0.0001). Data are expressed as mean ± SEM. Statistical analyses were performed as appropriate with GraphPad Prism 4.0 (GraphPad) or SigmaStat 3.5 (Systat); differences were considered to be significant at P < 0.05.

Supplementary Material

Supporting Information:


We thank Jennifer Heisel and Daniel Beisang for their excellent technical assistance, and Drs. Dave Morgan and Eric Burright for their helpful comments on the manuscript. This work was funded by Medtronic, Inc.


Conflict of interest statement: All authors are either current or past employees of Medtronic, Inc.; this work was funded by Medtronic, Inc.

This article contains supporting information online at www.pnas.org/cgi/content/full/0813404106/DCSupplemental.


1. Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer's disease: progress and problems on the road to therapeutics. Science. 2002;297:353–356. [PubMed]
2. Golde TE, Eckman CB, Younkin SG. Biochemical detection of Aβ isoforms: implications for pathogenesis, diagnosis, and treatment of Alzheimer's disease. Biochim Biophys Acta. 2000;1502:172–187. [PubMed]
3. Lee HG, et al. Amyloid-β in Alzheimer disease: the null versus the alternate hypotheses. J Pharmacol Exp Ther. 2007;321:823–829. [PubMed]
4. Schenk D. Amyloid-β immunotherapy for Alzheimer's disease: the end of the beginning. Nat Rev Neurosci. 2002;3:824–828. [PubMed]
5. Brody DL, Holtzman DM. Active and passive immunotherapy for neurodegenerative disorders. Annu Rev Neurosci. 2008;31:175–193. [PMC free article] [PubMed]
6. Vasilevko V, Cribbs DH. Novel approaches for immunotherapeutic intervention in Alzheimer's disease. Neurochem Int. 2006;49:113–126. [PubMed]
7. Schenk D, et al. Immunization with amyloid-β attenuates Alzheimer-disease-like pathology in the PDAPP mouse. Nature. 1999;400:173–177. [PubMed]
8. Janus C, et al. Aβ peptide immunization reduces behavioural impairment and plaques in a model of Alzheimer's disease. Nature. 2000;408:979–982. [PubMed]
9. Morgan D, et al. Aβ peptide vaccination prevents memory loss in an animal model of Alzheimer's disease. Nature. 2000;408:982–985. [PubMed]
10. Bard F, et al. Peripherally administered antibodies against amyloid β-peptide enter the central nervous system and reduce pathology in a mouse model of Alzheimer disease. Nat Med. 2000;6:916–919. [PubMed]
11. DeMattos RB, et al. Peripheral anti-Aβ antibody alters CNS and plasma Aβ clearance and decreases brain Aβ burden in a mouse model of Alzheimer's disease. Proc Natl Acad Sci USA. 2001;98:8850–8855. [PMC free article] [PubMed]
12. Dodart JC, et al. Immunization reverses memory deficits without reducing brain Aβ burden in Alzheimer's disease model. Nat Neurosci. 2002;5:452–457. [PubMed]
13. Frazer ME, et al. Reduced pathology and improved behavioral performance in Alzheimer's disease mice vaccinated with HSV amplicons expressing amyloid-beta and interleukin-4. Mol Ther. 2008;16:845–853. [PMC free article] [PubMed]
14. Hock C, et al. Antibodies against β-amyloid slow cognitive decline in Alzheimer's disease. Neuron. 2003;38:547–554. [PubMed]
15. Gilman S, et al. Clinical effects of Aβ immunization (AN1792) in patients with AD in an interrupted trial. Neurology. 2005;64:1553–1562. [PubMed]
16. Nicoll JA, et al. Neuropathology of human Alzheimer disease after immunization with amyloid-β peptide: a case report. Nat Med. 2003;9:448–452. [PubMed]
17. Ferrer I, Boada Rovira M, Sanchez Guerra ML, Rey MJ, Costa-Jussa F. Neuropathology and pathogenesis of encephalitis following amyloid-β immunization in Alzheimer's disease. Brain Pathol. 2004;14:11–20. [PubMed]
18. Masliah E, et al. Aβ vaccination effects on plaque pathology in the absence of encephalitis in Alzheimer disease. Neurology. 2005;64:129–131. [PubMed]
19. Patton RL, et al. Amyloid-β peptide remnants in AN-1792-immunized Alzheimer's disease patients: a biochemical analysis. Am J Pathol. 2006;169:1048–1063. [PMC free article] [PubMed]
20. Nicoll JA, et al. Aβ species removal after Aβ42 immunization. J Neuropathol Exp Neurol. 2006;65:1040–1048. [PubMed]
21. Wilcock DM, et al. Passive immunotherapy against Aβ in aged APP-transgenic mice reverses cognitive deficits and depletes parenchymal amyloid deposits in spite of increased vascular amyloid and microhemorrhage. J Neuroinflammation. 2004;1:24. [PMC free article] [PubMed]
22. Wilcock DM, et al. Deglycosylated anti-amyloid-β antibodies eliminate cognitive deficits and reduce parenchymal amyloid with minimal vascular consequences in aged amyloid precursor protein transgenic mice. J Neurosci. 2006;26:5340–5346. [PubMed]
23. Pfeifer M, et al. Cerebral hemorrhage after passive anti-Aβ immunotherapy. Science. 2002;298:1379. [PubMed]
24. Racke MM, et al. Exacerbation of cerebral amyloid angiopathy-associated microhemorrhage in amyloid precursor protein transgenic mice by immunotherapy is dependent on antibody recognition of deposited forms of amyloid β J Neurosci. 2005;25:629–636. [PubMed]
25. Greenberg SM, Gurol ME, Rosand J, Smith EE. Amyloid angiopathy-related vascular cognitive impairment. Stroke. 2004;35:2616–2619. [PubMed]
26. Jellinger KA. Alzheimer disease and cerebrovascular pathology: an update. J Neural Transm. 2002;109:813–836. [PubMed]
27. Levites Y, et al. Insights into the mechanisms of action of anti-Aβ antibodies in Alzheimer's disease mouse models. FASEB J. 2006;20:2576–2578. [PubMed]
28. Lemere CA, et al. Evidence for peripheral clearance of cerebral Aβ protein following chronic, active Aβ immunization in PSAPP mice. Neurobiol Dis. 2003;14:10–18. [PubMed]
29. Banks WA, et al. Passage of amyloid β protein antibody across the blood-brain barrier in a mouse model of Alzheimer's disease. Peptides. 2002;23:2223–2226. [PubMed]
30. Klyubin I, et al. Amyloid β protein immunotherapy neutralizes Aβ oligomers that disrupt synaptic plasticity in vivo. Nat Med. 2005;11:556–561. [PubMed]
31. Billings LM, Oddo S, Green KN, McGaugh JL, LaFerla FM. Intraneuronal Aβ causes the onset of early Alzheimer's disease-related cognitive deficits in transgenic mice. Neuron. 2005;45:675–688. [PubMed]
32. Dineley KT, Xia X, Bui D, Sweatt JD, Zheng H. Accelerated plaque accumulation, associative learning deficits, and up-regulation of α7 nicotinic receptor protein in transgenic mice co-expressing mutant human presenilin 1 and amyloid precursor proteins. J Biol Chem. 2002;277:22768–22780. [PubMed]
33. He P, et al. Deletion of tumor necrosis factor death receptor inhibits amyloid β generation and prevents learning and memory deficits in Alzheimer's mice. J Cell Biol. 2007;178:829–841. [PMC free article] [PubMed]
34. Hock C, et al. Generation of antibodies specific for β-amyloid by vaccination of patients with Alzheimer disease. Nat Med. 2002;8:1270–1275. [PubMed]
35. Lee M, et al. Aβ42 immunization in Alzheimer's disease generates Aβ N-terminal antibodies. Ann Neurol. 2005;58:430–435. [PubMed]
36. Lesne S, et al. A specific amyloid-β protein assembly in the brain impairs memory. Nature. 2006;440:352–357. [PubMed]
37. Tampellini D, et al. Internalized antibodies to the Aβ domain of APP reduce neuronal Aβ and protect against synaptic alterations. J Biol Chem. 2007;282:18895–18906. [PubMed]
38. Wilcock DM, et al. Passive amyloid immunotherapy clears amyloid and transiently activates microglia in a transgenic mouse model of amyloid deposition. J Neurosci. 2004;24:6144–6151. [PubMed]
39. Callahan MJ, et al. Augmented senile plaque load in aged female β-amyloid precursor protein-transgenic mice. Am J Pathol. 2001;158:1173–1177. [PMC free article] [PubMed]
40. Van Vickle GD, et al. TgCRND8 amyloid precursor protein transgenic mice exhibit an altered γ-secretase processing and an aggressive, additive amyloid pathology subject to immunotherapeutic modulation. Biochemistry. 2007;46:10317–10327. [PubMed]
41. Hsiao K, et al. Correlative memory deficits, Aβ elevation, and amyloid plaques in transgenic mice. Science. 1996;274:99–102. [PubMed]
42. Kawarabayashi T, et al. Age-dependent changes in brain, CSF, and plasma amyloid (β) protein in the Tg2576 transgenic mouse model of Alzheimer's disease. J Neurosci. 2001;21:372–381. [PubMed]
43. Robbins EM, et al. Kinetics of cerebral amyloid angiopathy progression in a transgenic mouse model of Alzheimer disease. J Neurosci. 2006;26:365–371. [PubMed]
44. Domnitz SB, et al. Progression of cerebral amyloid angiopathy in transgenic mouse models of Alzheimer disease. J Neuropathol Exp Neurol. 2005;64:588–594. [PubMed]
45. Wilcock DM, Gordon MN, Morgan D. Quantification of cerebral amyloid angiopathy and parenchymal amyloid plaques with Congo red histochemical stain. Nat Protoc. 2006;1:1591–1595. [PubMed]
46. Bard F, et al. Epitope and isotype specificities of antibodies to β-amyloid peptide for protection against Alzheimer's disease-like neuropathology. Proc Natl Acad Sci USA. 2003;100:2023–2028. [PMC free article] [PubMed]
47. Morgan D. Mechanisms of Aβ plaque clearance following passive Aβ immunization. Neurodegener Dis. 2005;2:261–266. [PubMed]
48. Chauhan NB, Siegel GJ, Lichtor T. Distribution of intraventricularly administered anti-amyloid β peptide (Aβ) antibody in the mouse brain. J Neurosci Res. 2001;66:231–235. [PubMed]
49. Chauhan NB, Siegel GJ. Reversal of amyloid β toxicity in Alzheimer's disease model Tg2576 by intraventricular anti-amyloid β antibody. J Neurosci Res. 2002;69:10–23. [PubMed]
50. Rensink AA, de Waal RM, Kremer B, Verbeek MM. Pathogenesis of cerebral amyloid angiopathy. Brain Res Brain Res Rev. 2003;43:207–223. [PubMed]
51. Holtzman DM. Alzheimer's disease: moving towards a vaccine. Nature. 2008;454:418–420. [PubMed]
52. Levites Y, et al. Intracranial adeno-associated virus-mediated delivery of anti-pan amyloid β, amyloid β40, and amyloid β42 single-chain variable fragments attenuates plaque pathology in amyloid precursor protein mice. J Neurosci. 2006;26:11923–11928. [PubMed]
53. Prada CM, et al. Antibody-mediated clearance of amyloid-β peptide from cerebral amyloid angiopathy revealed by quantitative in vivo imaging. J Neurosci. 2007;27:1973–1980. [PubMed]
54. Wilcock DM, et al. Microglial activation facilitates Aβ plaque removal following intracranial anti-Aβ antibody administration. Neurobiol Dis. 2004;15:11–20. [PubMed]
55. Thakker DR, et al. Time- and dose-dependent clearance of cerebral amyloid β load via intracerebroventricular passive immunotherapy in the Tg2576 mouse model of Alzheimer's disease. Soc Neurosci Abstr. 2007:485.10.
56. Bacskai BJ, et al. Non-Fc-mediated mechanisms are involved in clearance of amyloid-β in vivo by immunotherapy. J Neurosci. 2002;22:7873–7878. [PubMed]
57. Das P, et al. Amyloid-β immunization effectively reduces amyloid deposition in FcRγ-/- knock-out mice. J Neurosci. 2003;23:8532–8538. [PubMed]
58. Wilcock DM, et al. Intracranially administered anti-Aβ antibodies reduce β-amyloid deposition by mechanisms both independent of and associated with microglial activation. J Neurosci. 2003;23:3745–3751. [PubMed]
59. Klyubin I, et al. Amyloid β protein dimer-containing human CSF disrupts synaptic plasticity: prevention by systemic passive immunization. J Neurosci. 2008;28:4231–4237. [PMC free article] [PubMed]
60. Deane R, et al. IgG-assisted age-dependent clearance of Alzheimer's amyloid β peptide by the blood-brain barrier neonatal Fc receptor. J Neurosci. 2005;25:11495–11503. [PubMed]
61. Mamikonyan G, et al. Anti-Aβ1–11 antibody binds to different β-amyloid species, inhibits fibril formation, and disaggregates preformed fibrils but not the most toxic oligomers. J Biol Chem. 2007;282:22376–22386. [PMC free article] [PubMed]
62. Ma QL, et al. Antibodies against β-amyloid reduce Aβ oligomers, glycogen synthase kinase-3β activation and tau phosphorylation in vivo and in vitro. J Neurosci Res. 2006;83:374–384. [PubMed]
63. Thakker DR, et al. Neurochemical and behavioral consequences of widespread gene knockdown in the adult mouse brain by using nonviral RNA interference. Proc Natl Acad Sci USA. 2004;101:17270–17275. [PMC free article] [PubMed]
64. Holtzman DM, et al. Apolipoprotein E isoform-dependent amyloid deposition and neuritic degeneration in a mouse model of Alzheimer's disease. Proc Natl Acad Sci USA. 2000;97:2892–2897. [PMC free article] [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...


  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem Substance links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...