The pro-inflammatory microRNA miR-155 influences fibrillar β-Amyloid1-42 catabolism by microglia

Abstract

Microglia are the innate immune cells of the central nervous system that adopt rapid functional changes in response to Damage Associated Molecular Patterns, including aggregated β-Amyloid (Aβ) found in Alzheimer's disease (AD). microRNAs (miRNAs) are post-transcriptional modulators that influence the timing and magnitude of microglia inflammatory responses by downregulating the expression of inflammatory effectors. Recent studies implicate miR-155, a miRNA known to regulate inflammatory responses, in the pathogenesis of neurodegenerative disorders including multiple sclerosis, ALS, familial Parkinson's disease, and AD. In this work, we asked if miR-155 expression in microglia modifies cellular behaviors in response to fibrillar Aβ_1-42 (fAβ_1-42 ), in vitro. We hypothesized that in microglia, miR-155 expression would impact the internalization and catabolism of extracellular fAβ_1-42 . Primary microglia stimulated with lipopolysaccharide demonstrate fast upregulation of miR-155 followed by delayed upregulation of miR-146a, an anti-inflammatory miRNA. Conditional overexpression of miR-155 in microglia resulted in significant upregulation of miR-146a. Conditional deletion of miR-155 promoted transit of fAβ_1-42 to low-pH compartments where catabolism occurs, while miR-155 overexpression decreases fAβ_1-42 catabolism. Uptake of fAβ_1-42 across the plasma membrane increased with both up and downregulation of miR-155 expression. Taken together, our results support the hypothesis that inflammatory signaling influences the ability of microglia to catabolize fAβ_1-42 through interconnected mechanisms modulated by miR-155. Understanding how miRNAs modulate the ability of microglia to catabolize fAβ_1-42 will further elucidate the role of cellular players and molecular crosstalk in AD pathophysiology.

Keywords: catabolism; fibrillar Aβ1-42; miR-146a; miR-155; miRNAs; primary microglia.

FIGURE 1. Microglia respond to LPS with phased induction of miR-155 and miR-146a in a pattern similar to macrophages.

Primary mouse neonatal microglia were plated and exposed to 100 ng/ml of LPS for one of a set of time points, then total RNA was extracted and analyzed for changes in total microRNA expression. Quantification of (a) miR-155 and (b) miR-146a expression over time post LPS exposure demonstrates that miR-155 is upregulated quickly after LPS exposure while miR-146a is delayed and appears to be upregulated after miR-155 induction. This pattern matches that seen in peripheral macrophages. (One-Way ANOVA with Bonferroni's correction for multiple comparisons. For non-treated (NT) v. miR-155: Day 1: p < .0001, Day 2: p < .0001, Day 3: p = .0016, Day 4: p = n.s. and Day 5 p = n.s. (not significant = n.s.). For miR-146a: Day 1: p = n.s., Day 2: p = n.s., Day 3: p = n.s., Day 4: p = .0441, Day 5: p = .0018. Key: *** = p < .0001, ** = p < .001

FIGURE 2. LPS influences microglia lysosomal transit of fibrillar Aβ_1-42 without significantly shifting acidity of low-pH organelles.

Wild-type (WT) microglia were plated and treated with LPS (100 ng/ml) for 24 hr. Then cells were exposed to 1 μM fAβ_1-42-pHrodo for 6 hr or Lysotracker Deep Red for 30 min prior to FACS analysis. (a) fAβ_1-42 -pHrodo (PE) molecules localized to low-pH compartments were detected using flow cytometry (BD LSRII). The Geometric Mean of fAβ_1-42 -pHrodo were calculated from biological replicates and visually appear to have lower fluorescence after exposure to LPS. (b) LPS treatment significantly downregulated transit of fAβ_1-42 -pHrodo to lysosomal compartments where it would fluoresce (Unpaired t-test, p = .0039). (c) Lysotracker Deep Red (APC) representative histograms. (d) 24 hr of LPS (100 ng/ml) treatment did not significantly alter acidity of low-pH organelles (Unpaired t-test, p = .4139, n.s.). Key: ** = p < .01

FIGURE 3. LPS stimulation in primary microglia impacts intracellular handling of Aβ_1-42 to low-pH compartments and at the plasma membrane.

(a) Representative images of WT microglia exposed to fAβ_1-42 -pHrodo or fAβ_1-42 −488 at 6 hr (where fluorophores do not colocalize) and 24 hr (where fluorophores colocalize). WT microglia were plated and treated with 100 ng/ml of LPS for 24 hr. Then, cells were treated with 1 μM of fAβ_1-42 tagged with (b) pHrodo or (c) a non-pH sensitive fluorophore (488) for 6, 12, 18, and 24 hr. Total internalization and levels of fAβ_1-42 located in low-pH compartments was detected using flow cytometry (BD LSRII). Geometric mean of biological replicates (n = 3) were analyzed. No change in shift to low-pH compartments was detected using fAβ_1-42 -pHrodo (all n.s.). For overall internalization detected with fAβ_1-42 −488 there was a significant effect of LPS treatment with increased internalization over 24 hr (Two-Way ANOVA with Tukey's Multiple Comparisons correction: Treatment: F (1,4) = 55.44, p = .0017). While the 6 hr and 24 hr timepoints appear to have particularly significant increases in internalization and the interaction of treatment by time was significant (F (3,12) = 3.577, p = .0469) the post hoc comparisons did not remain significant after multiple comparison correction

FIGURE 4. In the absence of a strong inflammatory stimulus, overexpression of miR-155 is sufficient to induce expression of miR-146a.

(a) Graphic showing workflow of conditional miR-155 overexpression (OX) in primary mouse microglia. mBIC-FSF neonatal microglia were plated and treated with rAAV2-Cre or rAAV2-Control (RFP or GFP; “Control”) for 5 days. Then total RNA was extracted and analyzed for changes in miR-155 and miR-146a expression. For (b) miR-155, OX increased in AAV-treated cells above both control and non-treated levels. One-way ANOVA, (F (2,9)=139.1, p < .0001) with Tukey's multiple comparisons. Tukey's Multiple Comparisons corrections: NT (non-treated) v. rAAV2-Cre: p < .0001; rAAV2-Control (RFP or GFP) v. rAAV2-Cre: p < .0001. (c) Without an inflammatory stimulus, but with miR-155 induction, miR-146a expression was also increased indicating that miR-155 regulates miR-146a expression in microglia. This pattern of expression is similar to peripheral macrophages. One-way ANOVA with Tukey's multiple comparisons correction; NT (non-treated) v. rAAV2-Cre: p < .0001; rAAV2-Control (RFP or GFP) v. rAAV2-Cre: p < .0001. Key: **** = p < .001

FIGURE 5. Microglia internalization and transit of fAβ_1-42 to low pH compartments in the endosome/lysosome pathway is influenced by miR-155.

Microglia from miR-155^flx/flx or mBIC-FSF mice were isolated, plated, and transfected with rAAV2-Cre for 7 days, then RNA was isolated and assayed by qPCR for changes in miRNA expression (unpaired t-test). (a) miR-155 CKO showed a significant decrease in miR-155 expression (p = .0048). (b) miR-155 OX showed a significant increase in miR-155 expression (p = .0335). These indicate that the virus is working as expected in microglia. Microglia were then treated with 1 μM fAβ_1-42 for 6 hr. Lysosomal fAβ_1-42 levels were detected using flow cytometry (BD LSRII). (c) Representative plots of fluorescence detected from each of the cell types analyzed. (d) We observed significant differences between all three conditions indicating that altering miR-155 expression changes microglia internalization and transit of fAβ1-42 to low pH compartments. One-way ANOVA (F (2,11)=76.19, p = .0001). Tukey's corrected multiple comparisons: Control v. miR-155 OX, p = .0001; miR-155 OX v. miR-155 CKO, p < .0001; Control v. miR-155 CKO: p = .0002. Key: **** = p < .001, *** = p < .01

FIGURE 6. Conditional manipulation of miR-155 expression impacts fAβ_1-42 catabolism in low-pH compartments and internalization at the plasma membrane.

Microglia isolated from WT, miR-155 flx, or mBIC-FSF mice, plated, and infected with rAAV2-Cre for 7 days, then exposed to fAβ_1-42 -pHrodo or fAβ_1-42 −488 for 12, 18, and 24 hr. Total internalization and levels of fAβ_1-42 located in low-pH compartments was detected using flow cytometry (BD LSRII). Geometric mean of biological replicates (n = 3) were analyzed. (a) Using a Two-Way ANOVA we observed an interaction of genotype and time (F (6,15) = 8.827, p = .0003) of miR-155 CKO and fAβ_1-42 -pHrodo exposure over time. A significant shift to low-pH compartments was detected in miR-155 CKO microglia using fAβ_1-42 -pHrodo at 6 hr compared to miR-155 OX microglia (Tukey's corrected multiple comparisons, p = .0001). (b) Using a two-way ANOVA we observed an interaction of genotype and time (F (6,18) = 4.752, p = .0046). Tukey's corrected multiple comparisons showed a significant increase in fAβ_1-42-488 internalization relative to control by miR-155 CKO microglia (p = .0388) and miR-155 OX microglia (p = .0287) at 6 hr. miR-155 CKO microglia showed non-significant trends in fAβ_1-42-488 internalization at 12 hr compared to control (p = .052) or miR-155 OX microglia (p = .2040). At 18 hr miR-155 CKO microglia peak in fAβ_1-42-488 internalization compared to miR-155 OX microglia (p = .0214), with a plateau thereafter. Key: **** = p < .0001, * = p < .05