Intracranial injection of AAV expressing NEP but not IDE reduces amyloid pathology in APP+PS1 transgenic mice

Abstract

The accumulation of β-amyloid peptides in the brain has been recognized as an essential factor in Alzheimer's disease pathology. Several proteases, including Neprilysin (NEP), endothelin converting enzyme (ECE), and insulin degrading enzyme (IDE), have been shown to cleave β-amyloid peptides (Aβ). We have previously reported reductions in amyloid in APP+PS1 mice with increased expression of ECE. In this study we compared the vector-induced increased expression of NEP and IDE. We used recombinant adeno-associated viral vectors expressing either native forms of NEP (NEP-n) or IDE (IDE-n), or engineered secreted forms of NEP (NEP-s) or IDE (IDE-s). In a six-week study, immunohistochemistry staining for total Aβ was significantly decreased in animals receiving the NEP-n and NEP-s but not for IDE-n or IDE-s in either the hippocampus or cortex. Congo red staining followed a similar trend revealing significant decreases in the hippocampus and the cortex for NEP-n and NEP-s treatment groups. Our results indicate that while rAAV-IDE does not have the same therapeutic potential as rAAV-NEP, rAAV-NEP-s and NEP-n are effective at reducing amyloid loads, and both of these vectors continue to have significant effects nine months post-injection. As such, they may be considered reasonable candidates for gene therapy trials in AD.

Figure 1. Study 1: Distribution of HA expression in the hippocampus (A–D) and frontal cortex (E–H) following intracranial administration of rAAV vectors, detected using an anti-HA antibody.

Panels A & E show NEP-n treated animals; panels B & F show NEP-m treated animals; panels C & G show NEP-s treated animals. Panels D & H show no positive staining in the contralateral uninjected left hippocampus and left anterior cortex, respectively. Scale bar = 120 µm.

Figure 2. Study 1: Aβ immunostaining is observed in mice throughout both the ipsilateral hippocampus (A, B and C) and ipsilateral anterior cortex (D, E and F).

Aβ staining in the hippocampus of animals that received intracranial injections of rAAV- NEP-n (B) or NEP-s (C) is reduced compared to staining in those animals that received injections of control vector rAAV- NEP-m (A). Aβ staining in the anterior cortex of mice that received intracranial injections of rAAV- NEP-s (F) or NEP-n (E) is also reduced compared to staining in mice that received control vector rAAV- NEP-m (D). Scale bar = 120 µm. Quantification of percent area of positive total Aβ staining is shown in the hemisphere ipsilateral to injection sites (G) and in the hemisphere contralateral to injection sites (H). NEP-n (n = 18), NEP-m (n = 15), NEP-s (n = 17). The (*) indicates significance compared to NEP-m with p<0.05; the (**) indicate significance compared to NEP-m with p<.001.

Figure 3. Study 1: Congophilic staining is observed in mice throughout both ipsilateral hippocampus (A, B and C) and ipsilateral anterior cortex (D, E and F).

Congophilic staining in the hippocampus of animals that received intracranial injections of rAAV- NEP-n (B) or NEP-s (C) is reduced compared to staining in those animals that received injections of control vector rAAV- NEP-m (A). Congophilic staining in the anterior cortex of mice that received intracranial injections of rAAV- NEP-s (F) or NEP-n (E) is also reduced compared to staining in mice that received control vector rAAV- NEP-m (D). Scale bar = 120 µm. Quantification of percent area of positive total Congophilic staining is shown in G (hemisphere ipsilateral to injection sites) and H (hemisphere contralateral to injection sites). NEP-n (n = 18), NEP-m (n = 15), NEP-s (n = 17). The (*) indicates significance compared to NEP-m with p<0.05.

Figure 4. Study 2: Aβ immunostaining is observed in mice throughout both the ipsilateral hippocampus (A, B and C) and ipsilateral anterior cortex (D, E and F).

Aβ staining in the hippocampus of animals that received intracranial injections of rAAV- IDE-n (B) or IDE-s (C) is unchanged compared to staining in those animals that received injections of control vector rAAV- GFP (A). Aβ staining in the anterior cortex of mice that received intracranial injections of rAAV- IDE-n (E) or IDE-s (F) is also unchanged compared to staining in mice that received control vector rAAV- GFP (D). Scale bar = 120 µm. Quantification of percent area of positive staining is shown in the hippocampus (G) and in the anterior cortex (H). No significant differences were observed. n = 8/group.

Figure 5. Study 2: Congophilic staining is observed in mice throughout both ipsilateral hippocampus (A, B and C) and ipsilateral anterior cortex (D, E and F).

Congophilic staining in the hippocampus of animals that received intracranial injections of rAAV- IDE-n (B) or IDE-s (C) is unchanged compared to staining in those animals that received injections of control vector rAAV- GFP (A). Congophilic staining in the anterior cortex of mice that received intracranial injections of rAAV- IDE-n (E) or NEP-s (F) is also unchanged compared to staining in mice that received control vector rAAV- GFP(D). Scale bar = 200 µm. Quantification of percent area of positive total congophilic staining is shown in G and H (hemisphere ipsilateral to injection sites). No significant differences were found. n = 8/group.

Figure 6. Study 3: Aβ immunostaining is observed in mice throughout both the hippocampus (A, B and C) and anterior cortex (D, E and F).

Aβ staining in the hippocampus of animals that received intracranial injections of rAAV- NEP-n (B) is reduced compared to staining in those animals that received injections of control vector rAAV- NEP-m (A). Aβ staining in the anterior cortex of mice that received intracranial injections of rAAV- NEP-n (E) or NEP-s (F) is also reduced compared to staining in mice that received control vector rAAV- NEP-m (D). Scale bar = 50 µm. Quantification of percent area of positive total Aβ staining is shown in and H. The (*) indicates significance compared to NEP-m mice with p<0.05; the (^∧) indicates significance compared to Tg control mice with p<0.05. The (#) indicates significance compared to NEP-m mice with p<0.01. Tg control (n = 9), NEP-n (n = 4), NEP-m (n = 7), NEP-s (n = 6).

Figure 7. Study 3: Congophilic staining is observed in mice throughout both the hippocampus (A, B and C) and anterior cortex (D, E and F).

Congophilic staining in the hippocampus of animals that received intracranial injections of rAAV- NEP-n (B) is reduced compared to staining in those animals that received injections of control vector rAAV- NEP-m (A). Congophilic staining in the anterior cortex of mice that received intracranial injections of rAAV- NEP-n (E) or NEP-s (F) is also reduced compared to staining in mice that received control vector rAAV- NEP-m (D). Scale bar = 50 µm. Quantification of percent area of positive total Aβ staining is shown in G and H. The (*) indicates significance compared to NEP-m mice with p<0.05, the (^∧) indicates significance compared to Tg control mice with p<0.05, and the (°) indicates significance compared to Tg control mice with p<0.01. Tg control (n = 9), NEP-n (n = 4), NEP-m (n = 7), NEP-s (n = 6).

Figure 8. Study 3: CD68 (A-C) and CD45 (D-F) immunostaining is observed throughout the hippocampus of study mice.

In the hippocampus, no significant differences are observed in the amount of total CD68 staining between the NEP-n, NEP-s, and NEP-m groups, but CD45 staining in NEP-s treated mice is greater than CD45 staining in NEP-m treated mice. Scale bar = 50 µm. Panels G and H present ANOVA analysis of the ratio of CD68 to congophilic staining, and of the ratio of quantitated CD45 to congophilic staining, respectively, in the hippocampus of study mice. The (*) indicates significance compared to NEP-m mice with p<0.05, and the (^∧) indicates significance compared to Tg control mice with p<0.05 or p<0.01 (°).Tg control (n = 9), NEP-n (n = 4), NEP-m (n = 7), NEP-s (n = 6).