A limited role for microglia in antibody mediated plaque clearance in APP mice

Abstract

Amyloid-beta (Abeta) accumulation in senile plaques is a hallmark of Alzheimer's disease (AD). Immunotherapy is a leading approach for amyloid clearance, despite the early termination of the Elan clinical trial with active immunization due to a few cases of meningoencephalitis. The mechanisms of immunotherapy-mediated amyloid clearance and this deleterious side effect are largely unknown. While clearance of Abeta probably results in part from microglia-mediated inflammation, it can be microglia independent. Therefore, establishing the role of microglia in Abeta clearance is important for the treatment of AD. We analyzed the effects of direct microglia activation and inhibition on antibody-mediated Abeta clearance. Robust microglia activation with interferon-gamma led to modest Abeta clearance alone but did not potentiate antibody-mediated clearance. Microglia elimination/inactivation with immunotoxin or minocycline only partially limited antibody-induced Abeta clearance suggesting that although there is a role for microglia in Abeta clearance, it does not account for the majority of the effect observed after anti-Abeta antibody treatment.

Figure 1

Effect of direct microglia activation and inhibition on clearance of senile plaques in vivo. Mice were treated with locally applied IFN-γ (15 μl, 50 ng/ml) and/or anti-Aβ antibody 10D5 (25 μl, 1.2 mg/ml). Senile plaque clearance was determined with longitudinal multiphoton imaging 3 and 7 days after treatment. Data are representative of 3–5 mice. Two-way ANOVA was used to detect statistically significant treatmentXsession effect ([F_(3,2)]=4.832; p<0.001] and treatment differences were detected by one-way ANOVA followed by Tamhane test. A) Measurements on day 3 (n= 77–232) F_[3,643]=19.061; *p<0.001 vs. Control group; †p<0.001 vs. 10D5 group. B) Measurements on day 7 (n= 42–160) F_[3,465]=9.878; *p<0.001 vs. Control group; †p<0.001 vs. 10D5 group. The same approach was used to determine the effect of microglia inhibition on 10D5 induced senile plaque clearance. Anti-Mac-1 saporin was locally applied (30 μl, 0.5 mg/ml) prior to 10D5 application. Minocycline was administered i.p. daily (90 mg/Kg) for 3 days prior 10D5 local application, and second group received minocycline 3 days prior to antibody application followed by 3 more days after 10D5 treatment. Data are representative of 3–5 mice. Two way ANOVA revealed a statistically significant treatmentXsession effect [F_(4,2)]=4.839; p<0.001]. Treatment differences were detected by one-way ANOVA followed by Tamhane test. C)Measurements on day 3 (n= 149–328) F_[4,1085]=10.922; *p<0.001 vs. Control group; †p<0.001 vs. 10D5 group. D) Measurements on day 7 (n= 122–350) F_[4,930]=7.928; *p<0.001 vs. Control group; †p<0.001 vs. 10D5 group.

Figure 2

Representative in vivo images acquired with multiphoton microscopy illustrating the effect of the selected treatments on Aβ clearance induced by 10D5 in APPswe/PS1dE9 mice. In control animals, existing plaques remain, and several new plaques appear within 1 week. In contrast, many plaques disappear in anti-Aβ-antibody treated animals. IFN- γ treated mice show that some senile plaques disappear after 7 days. When animals were treated with 10D5+IFN-γ the effect on plaque clearance was similar to that observed with 10D5 alone. Anti-Mac-1-saporin limits Aβ clearance induced by 10D5. A similar reduction in the effectiveness of 10D5 was observed after inhibiting microglia with minocycline administered for 3 or 6 days. White arrows point at plaques reduced or cleared on day 7. Blood vessels (blue) are imaged with fluorescence angiography (texas red dextran iv), while both vascular and parenchymal amyloid deposits are labeled with methoxy-XO4 (red). This image is a maximum intensity projection of a 3-dimensional volume that is 615×615 microns square and 200 microns deep. Scale bar=100 μm.

Figure 3

Effect of selected treatments on microglia numbers measured by stereological counting after immunostaining with anti Iba-1 antibody. Anti-Aβ 10D5 antibody (25 μl, 1.2 mg/ml locally applied) significantly increased microglia activation. Anti-Mac-1-saporin (30 μl, 0.5 mg/ml) locally applied before 10D5 application and minocycline (90 mg/Kg i.p.) daily administered for 6 days (days before and 3 days after local application of 10D5) reduced microglia activation. Although a similar profile was observed when minocycline was administered only for 3 days before 10D5 application, it did not reach statistical significance. IFN-γ (15 μl, 50ng/ml) alone or in combination with 10D5 showed a significant increase of microglia activation when compared to control values and 10D5 alone. Data are representative of 3–4 animals and differences are detected by one-way ANOVA followed by Tamhane test: microglia activation (F_[3,8]=22.548; *p<0.001 vs. 10D5 group; †p<0.001 vs. control group),microglia inactivation (F_[4,11]=6.557; *p=0.006 vs. 10D5 group).

Figure 4

Illustrative example of the effect of the selected treatments on microglia activation. In untreated APPswe/PS1dE9 transgenic animals, limited microglia activation was observed. Topical application of anti-Aβ antibody (10D5, 25 μl, 1.2 mg/ml) resulted in an increase in microglia activation. A remarkable increase of microglia activation is observed after local IFN-γ (15 μl, 50 ng/ml) alone or in combination with 10D5. A reduction in microglia activation was observed after topical application of anti-Mac-1-saporin (30 μl, 0.5 mg/ml) followed by 10D5 local application. Systemic administration of minocycline (90 mg/Kg i.p.) for 3 or 6 days in combination with 10D5 also reduced microglia activation. Microglia were immunostained using anti-Iba 1 (red) and senile plaques were stained with thioflavin S (blue). Scale bar=50 μm.