IL-10 alters immunoproteostasis in APP mice, increasing plaque burden and worsening cognitive behavior

Abstract

Anti-inflammatory strategies are proposed to have beneficial effects in Alzheimer's disease. To explore how anti-inflammatory cytokine signaling affects Aβ pathology, we investigated the effects of adeno-associated virus (AAV2/1)-mediated expression of Interleukin (IL)-10 in the brains of APP transgenic mouse models. IL-10 expression resulted in increased Aβ accumulation and impaired memory in APP mice. A focused transcriptome analysis revealed changes consistent with enhanced IL-10 signaling and increased ApoE expression in IL-10-expressing APP mice. ApoE protein was selectively increased in the plaque-associated insoluble cellular fraction, likely because of direct interaction with aggregated Aβ in the IL-10-expressing APP mice. Ex vivo studies also show that IL-10 and ApoE can individually impair glial Aβ phagocytosis. Our observations that IL-10 has an unexpected negative effect on Aβ proteostasis and cognition in APP mouse models demonstrate the complex interplay between innate immunity and proteostasis in neurodegenerative diseases, an interaction we call immunoproteostasis.

Figure 1. IL-10 increases Aβ deposition in Tg CRND8 mice

A. Neonatal CRND8 mice were injected with either AAV2/1-IL-10 or AAV2/1-GFP (control) in the cerebral ventricles and analyzed after 6 months for Aβ plaque pathology using anti Aβ mAb 33.1.1, anti-ubiquitin and ThioS staining. Both Aβ staining protocols depict increased Aβ deposition in IL-10/Tg mice compared to controls whereas ubiquitin staining around individual plaques was unaltered. Scale Bar, 600μm (whole brain), 125μm (cortex and hippocampus). n=8-10mice/group for immunohistochemistry and n=6/group for ThioS staining. B. Quantification of Aβ plaque shows significantly increased amyloid plaque burden (immunostained with anti Aβ 33.1.1) and total number of ThioS cored plaques in IL-10/Tg mice compared to control/Tg mice. Data represents mean ± sem. n=8-10/group for immunohistochemistry and n=6 mice/group for ThioS staining. ***p<0.001, unpaired two-tailed t test). C. Biochemical analyses of sequentially extracted Aβ42 and Aβ40 levels by end-specific sandwich ELISA show significantly increased SDS soluble and formic acid extractable insoluble Aβ levels in IL-10/Tg compared to control/Tg mice. No change was detected in RIPA extracted Aβ42 and Aβ40. Data represents mean ± sem. n=6 mice/group. (*p<0.05, ***p<0.001, 2 way Anova with Tukey's multiple comparison test). D. Representative 82E1 immunoblots of 6 month old IL-10/Tg mice show no significant changes in Aβ oligomers compared to control/Tg mice. The left lane shows a representative aggregated Aβ42 preparation. A band migrating ∼8kDa is upregulated in select IL-10/Tg mice (asterisk, right). Molecular weight markers are indicated on the right (kDa). The lower panel represents the 82E1 blot re-probed with anti-actin antibody to depict loading amount. n=6 mice/group. E. Representative Z slice section analysis of 4G8 immunoreactive Aβ (Alexa Fluor 488nm) and Iba-1 labeled microglia (Alexa Fluor 594nm) shows increased Aβ accumulation in microglia surrounding amyloid plaques in IL-10/Tg mice. DAPI, in blue, denotes nucleus. Scale bar, 20.5μm. n=3 mice/group. See also Figures S1-S4.

Figure 2. IL-10 increases Aβ deposition in Tg2576 mice

A.8 month old Tg2576 mice were injected in the hippocampus with AAV2/1-IL-10 or AAV2/1-GFP (Control), aged for 5 months and analyzed at 13 months. Representative brain sections (cortex and hippocampus) stained with anti-Aβ mAb 33.1.1 show that Aβ plaque burden is increased in IL-10 expressing mice compared to control. No significant changes in ThioS stained cored plaques was observed. Scale Bar, 125μm (whole brain), 75μm (cortex and hippocampus). n=5-7 mice/group. B. Quantitative burden analysis of Aβ plaque deposits show significantly increased Aβ plaque immunoreactivity but no change in ThioS reactive plaques in IL-10 expressing mice compared to controls. Data represents mean ± sem. n=5-7 mice/group. *p< 0.05, Unpaired t test. C. Biochemical analyses of Gn-HCL solubilized Aβ42 and Aβ40 levels measured by ELISA show significantly increased insoluble Aβ42 levels in IL-10 expressing mice compared to controls but no significant differences in soluble Aβ levels (TBSx fraction) are seen. Data represents mean ± sem. n=5-7 mice/group. *p< 0.05, Unpaired t test. See also Figure S4.

Figure 3. IL-10 worsens cognitive impairment in TgCRND8 mice

A. Schematic representation of the stimuli parameters used in delay conditioning paradigms. On day 1, mice were exposed to two pairings of an auditory conditioned stimulus (CS) and a brief co-terminating foot shock (US). After a day of recovery (day 2), the memory of the association between the training context and the US was evaluated on day 3, and the memory of the association between the tone CS and US evaluated on day 4. B. Mean percent of freezing exhibited by IL-10 expressing and control groups during training. All mice showed comparable activity in the training chamber, pausing briefly during their exploration of the novel environment of the chamber before the onset of the first tone stimulus. The evaluation of freezing response immediately following the presentation of a foot shock during training showed that both the control/Tg and IL-10/Tg mice froze less as compared to control/nTg mice and IL-10/nTg mice respectively (n=8-12/group). **p < 0.01, MODLSD Bonferroni t-tests. C-D. In the context test, IL-10/Tg mice showed significantly lower rates of freezing compared to control/Tg mice as well as nTg littermates expressing GFP or IL-10 indicative of severe memory deficits (C). During the tone test, there was no difference in the freezing response of the mice in the modified context of the chamber at the stage preceding tone presentation (Pre CS, D). During tone presentation (Post CS, D), IL-10/Tg mice showed significantly weaker tone fear memory as compared to their control/nTg and IL-10/Tg littermates, while the control/Tg mice showed a trend in freezing decrease as compared to Control/nTg (p=0.07) (n=8-12 mice/group). *p<0.05, **p < 0.01, ***p<0.001, MODLSD Bonferroni t-tests.

Figure 4. IL-10 induced cognitive impairment is accompanied by reduced synaptic proteins

A-B. Synaptophysin, PSD95 and phosphorylated PSD95 (pSer295 PSD95) levels in 6 month old IL-10/Tg and control/Tg mice. Molecular weight markers are indicated on the left (kDa). All blots were simultaneously reprobed with anti-actin antibody to depict loading amount. Intensity analysis (mean ± sem) of immunoreactive bands of interest were normalized to β-actin (B). n=5 mice/group; *p<0.05, Unpaired two-tailed t test. C. Synaptophysin immunoreactivity was decreased in both the cortex and hippocampus of IL-10/Tg mice compared to control/Tg. Scale, 150μm, n=5/group.

Figure 5. IL-10 expression leads to an M2 phenotype and does not affect astrogliosis or plaque engagement of astroglia in TgCRND8 mice

A. Sustained IL-10 production leads to a partial M2 phenotype in 6 month Tg and nTg littermate brains in CRND8 mice. Quantitative RT-PCR demonstrates that IL-10/Tg mice have elevated levels of M2 phenotype marker, Ym-1, but show no significant changes in MRC1 or Arginase (Arg). Analysis of three M1 phenotype markers, IL-6, iNOS or IL-1β, showed no changes. n=2-3 mice/group; each sample tested in triplicate. (***p<0.001, 1-way Anova with post-hoc Tukey's test). B. Representative images of Iba-1 (microglia) and GFAP (astrocyte) immunoreactivity in intact hippocampus (top) and higher magnification of selected area of interest (bottom) from Control/Tg and IL-10/Tg mice is shown. Insets depict individual cells (high magnification) from corresponding low magnification panels. Scale Bar, 125μm (top), 25μm (bottom), 12.5 μm (insets, bottom). n=6mice/group. C. Representative immunoblot and densitometric analysis of normalized levels of GFAP and cd11b obtained from 6 month old IL-10/Tg and Control/Tg mice. Molecular weight markers are indicated on the left (kDa). The lower panels represent blots re-probed with anti-actin antibody to depict loading amount. n=5 mice/group; p>0.05, 1-way Anova with Tukey's multiple comparison test. Data represents mean ± sem. D. IL-10 expression does not alter astrocytic or microglial engagement around cored Aβ plaques. The number of astrocytes (GFAP-Cy3) and microglia (Dylight 594 conjugated Tomato Lectin) engaged closely with the ThioS reactive plaques were quantified by counting the number of DAPI positive nuclei (restricted by a circular area of 100μm diameter around each plaque core, as depicted by the white circle). Data represents mean±sd. Scale bar, 25μm, n=6-7mice/group, 10 plaques/mouse. See also Figure S5.

Figure 6. Recombinant IL-10 attenuates fAβ40 uptake by primary glia

A. List representing log ratio changes of differentially expressing genes in primary murine microglia treated with 10μM fAβ42 or IL-10, alone or in combination, following analysis by NanoString Inflammation GX array (q=0.05). n=3/group. A selected set of altered transcripts is shown as fold ratio change over naïve glia. n=2-3/treatment. B. IL-10 treatment decreases microglial phagocytosis of fAβ40 but does not affect fβ42 uptake. Flow cytometric analysis for the presence of Aβ42-Hilyte555 or Aβ40-Hilyte488 in primary mouse glia cells was done following exposure to Aβ or vehicle control for various times (15 minutes to 1 hour). Data represents percent of microglial population positive for 555nm or 488nm fluorescence (mean±sem). (*p<0.05, **p<0.01, unpaired 2-tailed t test). C. Representative pictograms depicting flow cytometric (FACS) and microscopic analysis of IL-10 or vehicle (Control) treated primary microglia internalizing fAβ40-488nm at different timepoints. See also Figure S6.

Figure 7. Transcriptome profile of TgCRND8 mice in response to IL-10

A. List representing log ratio changes of differentially expressing genes in IL-10/Tg mice compared to Control/Tg mice as analyzed by NanoString Inflammation GX array (q=0.05). n=3/group. B. Volcano plots highlighting differential expression of genes with IL-10 over-expression in IL-10/Tg compared to Control/Tg mice as analyzed by NanoString Neurodegeneration Custom Array. Orange circles indicate statistically significant changes with fold change <2 and blue circles represent fold changes > 2 (q < 0.05). n=6 mice/group. See Table S1, S2 for details. C. Representative rank-ordered list of genes that are differentially upregulated in response to IL-10 expression in Tg mice. See Table S2 for details. q<0.05; n=6mice/group. D. Network analysis of differentially expressed genes in TgCRND8 mice. All differentially expressed genes (q < 0.05) were submitted to IPA for canonical network analysis. Nine significant networks with known connections to AD are shown. Among the genes submitted (140 for APP mice), these 9 networks include 55 genes. Differentially expressed genes from the IL-10 network are all contained within the IL-10 sub-network circle. Blue edges connect genes to shared networks. Gold edges denote literature connections between ApoE and IL-10 canonical pathway genes. n=6 mice/group. See Table S4. E. DIRAC output of rank-ordered network alterations in response to IL-10 signaling in APP mice. Table S5 describes detailed information of the genes in each network affected. See also Figure S7, Tables S1-S5.

Figure 8. IL-10 induced ApoE is redistributed to insoluble plaque-associated cell fraction and impairs microglial uptake of Aβ

A-B. Representative anti ApoE immunoblots from sequentially extracted brain lysates-RIPA and SDS (A) and FA (B)- of TgCRND8 mice. Intensity analysis of immunoreactive bands of interest in the RIPA and SDS lysates were normalized to β-actin and that of FA lysates were normalized to dry weight of the hemibrain. Molecular weight markers are indicated (kDa). Data represents mean ± sem. n=6-7/group. ***p<0.001, 1-way Anova with Tukey's multiple comparison test. C-D. Representative ThioS stained Aβ plaque shows increased plaque associated ApoE in IL-10/Tg mice cortex (C). Representative GFAP stained sections also demonstrate selective increase in plaque associated ApoE in IL-10/Tg mice (D). DAPI represents cellular nuclei. n=3-6/group. E. Microglia grown in ApoE conditioned media internalize less fAβ40 compared to GFP conditioned media. Flow cytometric analysis for the presence of Aβ40-Hilyte555 was done following exposure to Aβ for 30 minutes. Representative data from two experiments is presented (**p<0.01, unpaired 2-tailed t test). F. Amyloid pull down assay shows aggregated Aβ42 specifically binds ApoE in media. Representative assay from two independent experiments. ChA silk, silkmoth chorion 1-51 polypeptide; CS 35, bacterial cold shock polypeptide 1 -35. G. Mechanistic insights into the pro-amyloidogenic effects of IL-10. IL-10 can directly dampen microglial phagocytosis; ApoE, induced by IL-10, may inhibit astroglial phagocytosis and removal of Aβ by binding to Aβ in plaques. See also Figure S8.