CCL2 accelerates microglia-mediated Abeta oligomer formation and progression of neurocognitive dysfunction

Abstract

Background: The linkages between neuroinflammation and Alzheimer's disease (AD) pathogenesis are well established. What is not, however, is how specific immune pathways and proteins affect the disease. To this end, we previously demonstrated that transgenic over-expression of CCL2 enhanced microgliosis and induced diffuse amyloid plaque deposition in Tg2576 mice. This rodent model of AD expresses a Swedish beta-amyloid (Abeta) precursor protein mutant.

Methodology/principal findings: We now report that CCL2 transgene expression accelerates deficits in spatial and working memory and hippocampal synaptic transmission in beta-amyloid precursor protein (APP) mice as early as 2-3 months of age. This is followed by increased numbers of microglia that are seen surrounding Abeta oligomers. CCL2 does not suppress Abeta degradation. Rather, CCL2 and tumor necrosis factor-alpha directly facilitated Abeta uptake, intracellular Abeta oligomerization, and protein secretion.

Conclusions/significance: We posit that CCL2 facilitates Abeta oligomer formation in microglia and propose that such events accelerate memory dysfunction by affecting Abeta seeding in the brain.

Figure 1. Accelerated spatial learning impairment in APP/CCL2 mice.

The same animals were tested by RAWM test at two different time points: 2–3 (A and C) and 8–9 months of age (B and D). The numbers of animals were 9 APP/CCL2, 10 CCL2, 8 APP, and 10 WT mice (all littermates from cross breeding of APP and CCL2 hemizygotes). The compiled average errors (A and B) and latency (C and D) of Trial 1-5 are shown. *, ** or *** denotes p<0.05, 0.01 or 0.001 versus WT, #, ## or ### denotes p<0.05, 0.01 or 0.001 versus CCL2, and $, $$ or $$$ denotes p<0.05, 0.01 or 0.001 versus APP group, respectively, as determined by two-way ANOVA and Bonferroni posttests.

Figure 2. Paired-pulse facilitation and input-output responses recorded in the CA1 region of mouse hippocampal slices.

WT (n = 15 slices from 8 animals), CCL2 (n = 13 slices from 7 animals), APP (n = 12 slices from 7 animals) and APP/CCL2 mice (n = 14 slices from 7 animals) at 5–6 months of age were tested. (A) Superimposed sample traces of field excitatory postsynaptic potentials (fEPSPs) recorded in the CA dendritic field in response to twin pulse stimulation of Schafer-collateral pathway at various interpulse intervals (as shown in B) in WT and APP/CCL2 mice. (B) Graph plots of averaged paired-pulse facilitation (PPF) ratio at 20-, 40-, 80-, 160-, and 320-msec interpulse intervals in 5–6-month-old WT, CCL2, APP and APP/CCL2 mice. A significant difference between WT and APP/CCL2 mice (*p<0.05) was found at 20-, 160-, and 320-msec intervals and WT versus APP (#p<0.05) at 20-msec intervals as determined by ANOVA and Newman-Keuls post hoc (p<0.05). (C) Representative traces of superimposed fEPSPs, evoked by single pulse stimulations at various intensities ranging from 30 to 300 µA with an increment of 30 µA, were recorded from slices of WT, CCL2, APP and APP/CCL2 mice, respectively. (D) Averaged input-output curves generated using different stimulus intensities ranging from 0–300 µA in 30 µA increments as shown in C. Significant difference between WT and APP/CCL2 (*p<0.05) at the stimulus intensity of 270 µA was determined by ANOVA and Newman-Keuls post hoc.

Figure 3. Enhanced Aβ oligomer formation and deposition in APP/CCL2 mice.

(A) Frozen sections (10-µm thickness) of temporal cortex of APP and APP/CCL2 mice at 9 and 14 months of age were immunostained by Aβ antibody and visualized by DAB staining. Scale bar, 400 µm. (B and C) Quantification of cortical Aβ staining in APP and APP/CCL2 mice at 9 months of age (n = 5 per group). Percent area occupied by immunoreactivity (C) and numbers of plaques were measured. *** denotes p<0.001 vs APP/CCL2 as determined by Student's t-test. (D) APP/CCL2 and APP mice were sacrificed at 9 months of age and 2% SDS-soluble fractions were subjected to dot blot assay for Aβ oligomer using NU-1 as described (each brain sample per dot). Top panel is a representative dot blot image (3 representative dots per group out of n = 6 per group). Bottom panel shows quantification of band luminescent intensities (n = 6 per group). * denotes p<0.05 vs APP/CCL2 as determined by Student's t-test. (E–G) Identification of Aβ oligomers in the brains of APP and APP/CCL2 mice assessed by immunoprecipitation of Aβ oligomers from extracellular-enriched fraction using NU-1 anti-Aβ oligomer monoclonal antibody and Western blot using biotinylated 6E10 anti-Aβ monoclonal antibody. Oligomerized synthetic human Aβ1–42 peptide was used as a size marker and positive control of Aβ oligomers (hAβ42, left lane). Arrows indicate respective migration positions of trimers (3-mer), pentamer (5-mer in box), hexamers (6-mer in box), octamer (8-mer), and nonamers (9-mer). Band luminescent intensities for trimers (F) and pentamers (G) were quantified by Image J software. * or *** denotes p<0.05 or 0.001 as determined by Student's t-test, respectively.

Figure 4. Accumulation of astrocytes and microglia to Aβ deposits.

Frozen sections (10-µm thickness) of temporal cortex of APP and APP/CCL2 mice at 9 months of age were subjected to double immunofluorescence for GFAP (A–B, green) or IBA1 (C–D, green) and Aβ (NU-1) (E-H, red). (I–L) Merged images. Scale bar, 50 µm.

Figure 5. Aβ aggregation in mouse primary microglia.

Mouse primary microglia (5×10⁴ per well) were seeded in 96-well tissue culture plates, and incubated with monomeric Aβ42 (1 or 10 µM) in phenol red-free DMEM for 0 (-Aβ), 1, 24, and 72 h, followed by washing, fixation, and immunocytochemistry with NU-2 anti-Aβ oligomer antibody (green) and Hoechst 33342 for nuclear staining (blue). (A) Representative phase and merged immunofluorescet images after incubation with 10 µM Aβ42. Scale bar, 100 µm. (B, C) Fluorescent intensity of Aβ (B) or nuclear (C) signals were quantified using fluorometer (Ex/Em 488 nm/519 nm for Alexa 488, and 350 nm/461 nm for Hoechst 33342, respectively, n = 3 per group). −1° Ab represents negative control staining of microglia for Aβ (B) or nuclear (C) staining using secondary antibody (Alexa 488 anti-mouse IgG) and Hoeschst 33342 for nuclear staining. * or *** denotes p<0.05 or 0.001 versus 0 h incubation, respectively.

Figure 6. CCL2 and TNF-α enhances Aβ aggregation in mouse microglia.

(A) Primary mouse microglia were incubated without Aβ (-Aβ), with Aβ (other panels), and co-incubated with 10 ng/ml CCL2, with 10 ng/ml TNF-α, with 1 µg/ml neutralizing anti-CCL2 antibody (CCL2 mAb), or co-incubated with 10 ng/ml CCL2 and 1 µg/ml CCL2 mAb (CCL2+CCL2 mAb) for 1 h, followed by immunofluorescence with NU-2 anti-Aβ oligomer antibody (green) and Hoechst33342 for nuclear staining (blue). Merged captured images were shown. Insets are high-magnification. Scale bar, 100 µm. (B) Flurolometric quantification of Aβ oligomer fluorescent signal (Ex/Em = 488 nm/519 nm) normalized by the nuclear staining signal (Ex/Em = 350 nm/461 nm, n = 3 per group). White or black bars represent incubation of microglia with no Aβ (white, negative control) or with 10 µM Aβ42 (black) Aβ, respectively. *** and ### denote p<0.001 versus Aβ group, and versus CCL2 group, respectively.

Figure 7. CCL2 accelerates intracellular Aβ microglial oligomer formation.

(A, B) Quantification of Aβ oligomers in mouse microglia (A), or secreted from mouse microglia (B), incubated without Aβ (-Aβ), with Aβ (+Aβ), and co-incubated with CCL2 (10 ng/ml, Aβ+CCL2) for 1 h, assessed by dot blot. Band luminescent intensities were quantified by Typhoon Phosphoimager. ** denotes p<0.01 as determined by Student's t-test. (C) Identification of Aβ oligomers in mouse microglia assessed by Western blot. Synthetic human Aβ1–42 peptide (hAβ42) was used as a size marker and positive control (left lane). Arrows indicate respective migration positions of monomers (1-mer), dimers (2-mer), tetramers (4-mer), hexamers (6-mer) and dodecamers (12-mer). Band luminescent intensities for monomers (D), dimers (E), tetramers (F), and dodecamers (G) were quantified by Image J software. *, **, or *** denotes p<0.05, 0.01, or 0.001 as determined by Student's t-test, respectively.