Male C57BL/6 mice were obtained from Charles River Laboratories Inc. (Wilmington, MA, USA). MiR-155 KO mice (Dagan et al., 2012), strain B6.Cg-Mir155tm1.1Rsky/J, were purchased from Jackson labs (Bar Harbor, MA, USA). C57BL/6 mice were used as wild type (WT) controls for miR-155 KO experiments. All mice were group housed in a 12 h light-dark cycle and fed ad libitum. All procedures and protocols were approved by the Institutional Animal Care and Use Committee of the University of Nebraska Medical Center and conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Seven to nine week-old male mice were anesthetized using 5% inhaled isoflurane, their heads were shaved and placed in a Kopf stereotaxic head frame where they were maintained under anesthesia using 2% isoflurane. A 4-mm craniotomy was performed midway between lambda and bregma on the left side. The exposed dura was then impacted using a Precision Systems and Instrumentation TBI-0310 (Fairfax Station, VA, USA) at a speed of 3.5 m/s with a 200 ms dwell time. This procedure was similar to previous reports (Mattson and Scheff, 1994; Smith et al., 1995). Injury depths of 0.5 and 1.0 mm were used to simulate moderate and severe injuries, respectively. After injury, Surgicel (Johnson and Johnson, Dallas, TX, USA) was placed over the injury site, the removed skull was adhered back to its original place with dental cement and the wound closed with wound clips. 0.5% bupivacaine with 1:200,000 epinephrine was used as an analgesic. Naïve animals were not exposed to injury, craniotomy or anesthesia.

Paraffin-embedded tissues were used for Luxol fast blue, terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) and fluorescent in situ hybridization (FISH). Brains were submersion fixed in 4% PFA overnight before processing; staining was performed on 5 μm sections. Iba1 and fluoro-jade C (FJC) staining were performed on fixed-frozen tissues. Transcardial perfusion was performed on anesthetized mice with cold PBS and 2.5% sucrose followed by PBS with 2.5% sucrose and 4% PFA. Brains were submersion fixed an additional 24 h, cryoprotected in 30% sucrose, frozen with 2-methyl butane on dry ice and stored at −80°C. A 4 mm coronal section, including the injury site was sectioned into 50 μm slices and every fifth section was collected in 0.1 M phosphate buffer.

After hydration to 95% ethanol, slides were stained with filtered 0.1% Luxol fast blue in a solution of 0.5% acetic acid at 60°C overnight. The next day slides were rinsed in 95% ethanol followed by distilled water then differentiated in 0.05% lithium carbonate for 1 min and 70% ethanol for 1 min. Slides were counterstained with 0.5% cresyl violet for 30 min at 60°C, rinsed in distilled H₂O, and differentiated in 95% ethanol for 5 min. Coverslips were mounted using Cytoseal (Thermo, Waltham, MA, USA). Slide scanning was performed by the UNMC tissue sciences facility using a Ventana’s Coreo Au Slide Scanner at 40× magnification.

ApopTag Peroxidase in situ apoptosis detection kit (Millipore, Temecula, CA, USA) was used for TUNEL staining according to manufacturer’s directions in conjunction with TSA plus cyanine 5 kit (PerkinElmer, Waltham, MA, USA). Briefly, hydrated slides were treated with 20 μg/mL proteinase K, then terminal deoxynucleotidyl transferase (TdT) was used to transfer digoxigenin (DIG) labeled nucleotides onto fragmented DNA. The tail of DIG labeled nucleotides was then detected using an anti-DIG antibody conjugated to peroxidase. A TSA plus cyanine 5 kit (PerkinElmer, Waltham, MA, USA) was used to develop the fluorescent stain and nuclei were stained with DAPI.

ISH was performed as described previously (Chaudhuri et al., 2013). In brief, double DIG-labeled LNA probes (Exiqon, Vedbaek) were used in conjunction with anti-DIG-POD, Fab fragments (Roche, Basel) and TSA plus cyanine 5 (PerkinElmer, Waltham, MA, USA). To reduce background, a 10 min 3% H₂O₂ peroxidase quenching step was added to the original protocol. MAP2 antibody (ab5392, abcam, Cambridge) was used at 1:1000 dilution; nuclei were stained with DAPI.

Free-floating cryosections were incubated with 3% H₂O₂ to quench endogenous peroxidase. After thorough washing sections were blocked with 10% normal goat serum (NGS) +0.3% Triton-X100 in PBS. Tissues were incubated with Iba1 (Wako, Kampenhout; 1:500) in 0.3% Triton-X100 +3% NGS in PBS at 4°C overnight. ImmPRESS HRP Anti-Rabbit IgG (Vector labs, Burlingame) was used as a secondary and DAB Plus substrate system (Thermo, Waltham, MA, USA) was used for developing. Sections were then mounted on gelatin-coated slides, allowed to dry overnight, incubated in xylene for 1 min, and mounted with Cytoseal (Thermo, Waltham, MA, USA). Slides were imaged using a light microscope at 100× magnification. Images were quantified using ImageJ (Schneider et al., 2012) to calculate the percent area covered by Iba1 staining in the hippocampus and cortex.

Free-floating cryosections were mounted on gelatin coated slides and air-dried overnight. Slides were incubated in 0.06% potassium permanganate followed by 0.0001% FJC (Histochem, Jefferson, AR, USA) and 0.0001% DAPI (Sigma, St. Louis, MO, USA) in 1% acetic acid. Tissue was then dried in a 60°C oven and incubated in xylene for 1 min, then coverslips were mounted with DPX (Sigma, St. Louis, MO, USA). For each brain, three slices between bregma −2.5 and −1.5 were imaged for quantification. FJC positive cells were counted using ImageJ (Schneider et al., 2012) by a blinded observer. The average number of fluoro-jade positive cells per slice is reported.

Tissue was isolated from the hippocampus and RNA was extracted using TRIzol (Life Technologies, Carlsbad, CA, USA) according to manufacturer’s instructions. TaqMan miRNA and Gene Expression Assays (Life Technologies, Carlsbad, CA, USA) were used for cDNA synthesis and real time PCR according to manufacturer’s instructions. Small nuclear RNA U6 (U6) was used as a control for miRNA studies and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as a control for mRNA studies. Delta-delta Ct method was used to calculate fold change (2^{−((Ct_miRNA − Ct_U6)_exp − (Ct_miRNA − Ct_U6)_control)}) or (2^{−((Ct_mRNA − Ct_GAPDH)_exp − (Ct_mRNA − Ct_GAPDH)_control)}). Statistical significance was determined using ΔCt values (Ct_miRNA − Ct_U6) or (Ct_mRNA − Ct_GAPDH).

Protein was isolated from ipsilateral hippocampal tissue using TRIzol (Life Technologies, Carlsbad, CA, USA) according to manufacturer’s instructions. Equal amounts of protein samples were separated by SDS-PAGE and transferred to a nitrocellulose membrane. The membrane was blocked in Superblock (Pierce, Waltham, MA, USA) and incubated overnight at 4°C with primary antibodies. After washing, membrane was incubated with infrared secondary antibodies (LI-COR, Lincoln, NE, USA) for 1 h at room temperature. Imaging and densitometry was performed with an Odyssey Imaging System (LI-COR, Lincoln, NE, USA). The following antibodies and concentrations were used for Western blotting: suppressor of cytokine signaling 1 (SOCS1; 1:1000, 38-5200, Thermo, Waltham, MA, USA), caspase 3 (1:1000, 9662, Cell Signaling, Danvers, MA, USA), beta-actin (AM1829B, 1:20,000, Abgent, San Diego, CA, USA). For statistical analysis, each sample was normalized to beta-actin.

For comparisons between two groups an unpaired two-tailed student’s T-test was used to calculate significance. For comparisons between multiple groups two-way ANOVAs followed by Bonferroni post hoc tests were employed. For testing significance of multiple parameters, such as the expression of several genes, between two groups a false discovery rate calculation was used. All statistical tests were performed with GraphPad Prism (La Jolla, CA, USA). P < 0.05 was considered significant. For data represented in graphs, the mean ± SEM is shown. A minimum of three biological replicates was used for all experiments.

CCI produced a cortical lesion, and white matter damage to the underlying corpus callosum and alveus was evident later time points (Figure 1A). Given that apoptosis can occur in TBI and may be influenced by the miRNAs of interest, we examined the level of apoptosis in the cortex and hippocampus. Importantly, apoptosis in the hippocampus is limited compared to the cortical lesion boundary, which shows high levels of TUNEL staining (Figure 1B). To confirm that inflammatory signaling was present in the hippocampus, we measured the expression of pro-inflammatory cytokines IL-1β, TNFα and IL-6 in the hippocampus 1, 3, 7 and 14 days after injury by qPCR. As expected based on previous reports, expression of pro-inflammatory cytokines was increased acutely after CCI in the injured (ipsilateral) compared to the uninjured (contralateral) hippocampus (Figure 1C). This indicated that the hippocampus was a suitable region for studying regulation of inflammation by miRNAs while limiting the confounding role of apoptosis.

Expression of miR-155 (UUAAUGCUAAUUGUGAUAGGGGU), miR-21(UAGCUUAUCAGACUGAUGUUGA), miR-223 (UGUCAGUUUGUCAAAUACCCCA) and miR-146 (UGAGAACUGAAUUCCAUGGGUU) was examined in hippocampi of the injured (ipsilateral) and the uninjured (contralateral) hemisphere from the same animals at 1, 3, 7 and 14 days after moderate CCI and in naïve controls (Figure 1D). In naïve animals, there was no difference in miRNA expression between hemispheres. CCI significantly increased expression of miR-155, miR-21 and miR-223 in the ipsilateral compared to the contralateral hippocampus by two-way ANOVA. Bonferroni post hoc analysis showed that the difference in miR-155 reached statistical significance at day 1, 3 and 7 whereas, expression of miR-21 was significantly increased at 14 days after CCI (P = 0.035), but not at earlier time points, day 1, 3 and 7. The post hoc test did not reveal significance for miR-223 expression at these time points. Expression of miR-146 was not altered by CCI at the time points examined by two way ANOVA (P = 0.16). Increasing the injury severity from moderate to severe, 0.5 mm and 1.0 mm depth respectively, revealed significantly increased expression of miR-155, miR-21 and miR-223 at 3 days post-injury (DPI) as determined by two-way ANOVA. Bonferroni post hoc analysis showed that miR-155, miR-223 and miR-21 were all elevated in the ipsilateral as compared to the contralateral hemisphere 3 days after severe CCI, whereas only miR-155 was also elevated after moderate CCI (Figure 1E). In summary, inflammation-associated miRNAs are elevated in the ipsilateral hippocampus after TBI and in general showed a more prolonged increase compared to the expression of cytokines IL-1β, TNFα and IL-6, which peaked 1 day after injury. Of the miRNAs examined, miR-155 alone was significant at all time points measured up to 7 days after moderate injury. Considering the strong induction of miR-155, the function of miR-155 in TBI was further investigated.

To examine the cell type-specific localization of miR-155 in the injured brain, we performed FISH. Examining miR-155 by FISH revealed prolonged elevation of miR-155 at 1, 3 and 7 days after moderate CCI as well as 3 days after severe CCI in the hippocampus, similar to that observed by qPCR (Figures 2A,B). Expression of miR-155 was also observed in the boundary zone of the cortex (Supplementary Figure S1A). Importantly, no fluorescent signal was detected in miR-155 knockout (KO) mice, confirming FISH specificity (Figure 2B). Interestingly, expression of miR-155 was largely nuclear and present at high levels in the laminar layers of the hippocampus, structures composed primarily of neurons. Co-staining for miR-155 and neuronal marker MAP2 confirmed that miR-155 was expressed at high levels in neuronal nuclei in both the hippocampus and cortex (Figure 2C, Supplementary Figure S1B). Together with qPCR data, FISH clearly shows elevation of miR-155 after CCI.

In order to understand the function of miR-155 in regulating the immune response in TBI, we subjected miR-155 KOs and WT mice to CCI. Three days after CCI, RNA was isolated from ipsilateral hippocampi and expression of cytokines and chemokines were measured by qPCR. Unexpectedly, there were no differences in the expression of IL-6, IL-1β or TNFα (Figure 3A), which are regulated by miR-155 in other systems. However, miR-155 KO mice had decreased expression of 6/9 IFN related genes measured (FDR < 0.05), including IFNα2, IFNα4, IFNα5 and IFNβ1, as well as IFN regulatory factor 1 (IRF1) and IFN induced chemokine CXCL10 (Figure 3B). Overall, we found a reduction of the IFN response in miR-155 KO mice after TBI without changes in other TBI induced inflammatory cytokines.

To examine the importance of miR-155 in microglial activation in the context of TBI, we examined the levels of microglial marker Iba1 in hippocampal slices 3 days after CCI in miR-155 KO and WT mice. In miR-155 KO mice we observed an increase rather than the expected decrease in Iba1 staining in the ipsilateral hippocampus compared to control mice (P = 0.046 by Student’s T-test; Figures 3C,D). These data suggest that miR-155 is not essential for the activation of microglia after TBI and implies that a dampened microglial response is not responsible for the reduction in IFN signaling observed in miR-155 KO mice.

We hypothesized that previously validated miR-155 targets SOCS1 and SHIP1 mediated miR-155 regulation of IFN signaling in the context of TBI. However, compared to the expression found in WT mice, no increase in SOCS1 was observed in miR-155 KO mice at the mRNA or protein level after TBI (Figures 3E,F). In fact there was a significant decrease in SOCS1 mRNA expression in miR-155 KO mice (P = 0.007 by student’s T-test). Therefore, miR-155 regulation of IFN signaling in TBI is likely independent of SOCS1. We also tested SHIP1, a second target of miR-155 known to regulate IFN signaling and found no significant change in mRNA expression by qPCR. In summary, miR-155 KO mice show a decreased expression of type 1 IFNs and IFN regulated genes without evident decreases in microglial activation or increased SOCS1 or SHIP1 expression.

Since miR-155 was highly increased in neuronal nuclei, the effect of miR-155 ablation on neuronal degeneration and apoptosis in the hippocampus was measured. Fluoro-jade C (FJC) staining for degenerating neurons showed evidence of neurodegeneration both in the cortex and hippocampus. In the hippocampus FJC positive cells were found predominately in the Dentate gyrus (DG). The DG in miR-155 KO mice had significantly more FJC positive cells relative to controls (P = 0.046 by Student’s T-test; Figures 4A,B), indicating that miR-155 KO mice had increased hippocampal neurodegeneration. Consistent with TUNEL staining of the hippocampus, there were low levels of caspase 3 cleavage in both groups, 10% and 20% for WT miR-155 KO mice respectively, a difference that was not significant (P = 0.20 by Student’s T-test; Figures 4C,D). Overall we found increased hippocampal neurodegeneration in miR-155 KO mice, without significant evidence of increased apoptosis. Figure 5 graphically summarizes our main findings on miR-155 in TBI.

We have identified miR-155 as a TBI induced miRNA that promotes the type 1 IFN response. In the context of TBI, miR-155 is strongly expressed in neuronal nuclei, suggesting that it may regulate gene expression through non-canonical mechanisms. Increased neurodegeneration in miR-155 KO mice points to a neuroprotective role for miR-155 in TBI.