Up-regulation of microRNA-142 in simian immunodeficiency virus encephalitis leads to repression of sirtuin1

Abstract

MicroRNA (miR)-142 is up-regulated in the brain in HIV and SIV encephalitis (SIVE). We identified the cell types where miR-142 is up-regulated and its relevant downstream target. Fluorescent in situ hybridization combined with immunofluorescent labeling revealed that miR-142-3p and -5p are expressed within hippocampal neurons and myeloid cells in SIVE. Sirtuin1 (SIRT1) was predicted as a potential miR-142 target by analysis of its 3′-UTR and bioinformatic analysis of factors linked to altered hippocampal gene expression profile in SIVE. Overexpression of pre-miR-142 in HEK293T cells led to a 3.7-fold decrease in SIRT1 protein level. Examination of the individual effects of miR-142-5p and miR-142-3p through overexpression and inhibition studies revealed that significant effects on SIRT1 occurred only with miR-142-5p. Luciferase reporter assays revealed a 2.3-fold inhibition of expression due to interaction of miR-142 with the SIRT1 3′-UTR, mutation analysis revealed that only the miR-142-5p target site was active. MiR-142 expression in primary human neurons led to a small (1.3-fold) but significant decrease in SIRT1 protein level. Furthermore, qRT-PCR revealed up-regulation of miR-142-3p (6.4-fold) and -5p (3.9-fold) and down-regulation of SIRT1 (33-fold) in macrophages/microglia from animals with SIVE. We have therefore elucidated a miR-mediated mechanism of regulation of SIRT1 expression in SIVE.—Chaudhuri, A. D., Yelamanchili, S. V., Marcondes, M. C. G., Fox, H. S. Up-regulation of microRNA-142 in simian immunodeficiency virus encephalitis leads to repression of sirtuin1.

Small noncoding RNAs, particularly microRNAs (miRs), have emerged as important translational regulators (1–3). Altered expression of miRs, and consequentially of their targets, has now been documented in various diseases, including those of the central nervous system (CNS; refs. 4, 5). HIV-associated neurocognitive disorder (HAND) is one such disease, and neurodegeneration in HAND is secondary to CNS inflammation. The pathological hallmarks of the most extreme form of the disease include astrogliosis, microgliosis, presence of multinucleated giant cells, and loss of dendrites and synapses (6–8), collectively termed HIV encephalitis (HIVE). These features are very well recapitulated in rhesus macaque model, simian immunodeficiency virus encephalitis (SIVE; ref. 9). In the CNS, productive HIV infection is harbored within microglia and macrophages. Although the virus does not infect neurons, inflammatory cytokines and chemokines released from infected macrophages and microglia, as well as HIV gene products (e.g., Tat, gp120) have deleterious effects on them, altering neuronal gene and protein expression (10–13).

To determine whether miRs could be the upstream factors driving such changes in neuronal gene and protein expression in HIVE/SIVE, we and others have compared the brain miR expression profile in HIVE/SIVE to uninfected control samples (14, 15). Up-regulation of miR-142 was detected in the frontal lobe white matter (15) as well as in the hippocampus and caudated nucleus (neuron-rich regions affected in the disease) (14). Unlike many miRs that have a single mature product, for miR-142 both the strands (-3p and -5p) can be incorporated into the RISC and are functionally active, having their own sets of target mRNAs. MiR-142 is expressed primarily in hematopoietic tissues (16), where its role in differentiation of T-lymphocytes and myeloid cells has been extensively investigated (17–19). Although miR-142 has been reported to be up-regulated in striatal neuronal postsynaptic densities following cocaine treatment (20) and in neurons following nerve crush injury (21), its functions in the brain is not clear.

Therefore, the goal of the present study was to determine the cell type in which miR-142 is expressed in SIVE brain and to determine its downstream targets relevant to the disease. We found that miR-142-3p and -5p are expressed within neurons as well as myeloid cells in the SIVE brain. Furthermore, sirtuin1 (SIRT1), a NAD-dependent protein deacetylase, was validated as a direct miR-142-5p target. We also provide evidence that miR-142 can regulate SIRT1 expression in the cell types where it was found to be expressed in the disease, i.e., in primary human neurons and in macrophages/microglia from rhesus macaques with SIVE. Given the multifaceted role of SIRT1 in preventing aging, inflammation and neurodegeneration, its down-regulation by miR-142 could be an important contributing factor in the pathogenesis of HIVE/SIVE.

MATERIALS AND METHODS

Fluorescent in situ hybridization (FISH) and immunofluorescent (IF) labeling

FISH and IF was performed as described previously (22) with minor modifications. First, formalin-fixed paraffin-embedded sections were deparaffinized. For combined FISH and IF, this was followed by antigen retrieval using 0.01 M citrate buffer and postfixation using 0.16 M l-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC; Sigma-Aldrich, St. Louis, MO, USA) to prevent loss of small RNAs. The sections were incubated with hybridization buffer (50% formamide; 10 mM Tris-HCl, pH 8.0; 200 μg/ml yeast tRNA; 1× Denhardt's solution; 600 mM NaCl; 0.25% SDS; 1 mM EDTA; and 10% dextran sulfate) for 1 h at 37°C in a humidified chamber for prehybridization. They were then incubated overnight at 37°C with locked nucleic acid (LNA)-modified, 5′- and 3′-digoxigenin-labeled miR-142-3p, miR-142-5p, or scrambled miR probe (Exiqon, Woburn, MA, USA) at a concentration of 4 pmol of probe per 100 μl of hybridization buffer. The sequences of all the probes used are listed in Table 1. Stringency washes were performed with 2× and 0.2× SSC (Invitrogen, Carlsbad, CA, USA) at 42°C. The hybridization and wash temperatures were optimized in preliminary experiments. The sections were then blocked with a solution of 1% BSA, 3% normal goat serum in 1× PBS for 1 h at room temperature, followed by incubation with anti-digoxigenin peroxidase antibody (1:100 in blocking buffer; Roche Applied Science, Mannheim, Germany) overnight at 4°C. For combined FISH and IF, coincubation with either anti-microtubule-associated protein 2 (MAP2; 1:1500; Sternberger Monoclonals Inc., Baltimore, MD, USA) or anti-CD163 (1:100; Vector Labs, Burlingame, CA, USA) and anti-glial fibrillary acidic protein (GFAP; 1:2000; Dako, Glostrup, Denmark) was performed at this step. The following secondary antibodies were used: Alexa Fluor 568 goat anti-mouse, 568 donkey anti-rabbit and 488 goat anti-mouse IgG (1:400; Invitrogen, Carlsbad, CA, USA). This was followed by signal amplification using tyramide signal amplification Cy5 kit (Perkin Elmer, Waltham, MA, USA) according to the manufacturer's protocol. The slides were mounted in Prolong gold antifade reagent with DAPI (Invitrogen). The sections were imaged in Zeiss Observer.Z1 microscope equipped with a monochromatic Axiocam MRm camera using Axiovision 40 v.4.8.0.0 software (Carl Zeiss, Oberkochen, Germany). The following colors were assigned to the fluorescent signals using the Axiovision software: red for MAP2 or CD163, magenta for GFAP, green for Cy5, blue for DAPI.

Table 1.

Probes used for FISH

miR	Probe sequence, 5′–3′
miR-142-3p	/5Dig/TCCATAAAGTAGGAAACACTACA/3Dig/
miR-142-5p	/5Dig/AGTAGTGCTTTCTACTTTATG/3Dig/
U6	/5Dig/CACGAATTTGCGTGTCATCCTY/3Dig/
Scramble-miR	/5Dig/GTGTAACACGTCTATACGCCCA/3Dig/

SIV/rhesus monkey model

Samples from SIV-infected rhesus monkeys that developed SIVE, and from uninfected control monkeys, were obtained from previous studies performed under approval from the Institutional Animal Care and Use Committees of The Scripps Research Institute and the University of Nebraska Medical Center, following U.S. National Institutes of Health (NIH; Bethesda, MD, USA) guidelines. All animals were initially tested to be free of SIV, type D simian retrovirus, and Macacine herpesvirus 1. For SIV infection, animals were intravenously inoculated with a cell-free viral stock, derived from SIVmac251 (23, 24). Animals received 0.25 ml of the stock, diluted into RPMI 1640 for injection, containing 5 ng/ml of p27 (gag) antigen. For animals used in this study, the infection was allowed to follow its natural course, and animals were euthanized when they showed signs of simian AIDS. At necropsy, all animals were perfused with PBS containing 1 U/ml heparin to remove blood-borne cells from the brain, and samples were taken for histopathological and molecular studies, as well as isolation of macrophage/microglia (below). Those in which pathological examination revealed multinucleated giant cells, microglial nodules, and infiltration of macrophages into the brain were classified as having SIVE; samples from these animals, as well as control uninfected animals prepared in a similar manner, were used for this study.

Isolation of macrophages/microglia from brain

Macrophages/microglia from the brain were isolated as described previously (25). In brief, minced brain fragments were subjected to tissue homogenization, and cells were isolated on a percoll gradient. CD11b-positive cells were purified immunomagentically and purity assessed by flow cytometry. Approximately 90% of the cells used were positive for CD11b.

Real-time quantitative RT-PCR (qRT-PCR) for miR-142 and SIRT1

TaqMan mature miR assays for miR-142-3p and -5p (Applied Biosystems, Foster City, CA, USA) were used to quantify the respective miRs according to the manufacturer's protocol. U6 small nuclear RNA (snRNA) was used as housekeeping control. Relative expression was calculated as described previously (14). For SIRT1 qRT-PCR, the RT² SYBR green qPCR master mix (SABiosciences, Valencia, CA, USA) was used with the following primers: forward, 5′-GCAACAGGCCCCTGATTACA-3′, reverse, 5′-AGCACAAACACATCATGCAAAT-3′. GAPDH was used as housekeeping control.

In silico prediction of miR-142 targets

The microRNA.org August 2010 release (http://www.microrna.org) was used to predict miR-142-5p and -3p targets, and targets common to both strands were determined. Only targets with miRSVR score ≤ −0.1 were considered for this purpose.

Ingenuity pathway analysis (IPA)

Differentially expressed genes (DEGs) from microarray results comparing gene expression in SIVE and uninfected rhesus macaque hippocampus (original data in Gene Expression Omnibus, GSE13824, DEG, as listed in Supplemental Table 2 in ref. 26) were subjected to ingenuity upstream regulator analysis (Ingenuity Systems, Redwood, CA, USA; http://www.ingenuity.com/). Differentially regulated pathways were filtered for transcription regulators that had an absolute score of z > 1.

Cell lines and primary neuron culture

HEK293T cells were grown in Dulbecco's modified Eagle medium (DMEM), high glucose with glutamax (Gibco; Invitrogen) containing 10% heat inactivated fetal bovine serum (FBS; (Thermo Scientific, Rockford, IL, USA). K562 cells were grown in Roswell Park Memorial Institute 1640 medium containing l-glutamine (Cellgro, Manassas, VA, USA), supplemented with 10% FBS. All cell lines were maintained at 37°C in 95% air and 5% CO₂.

For primary human neuron culture, fetal brain tissue (Birth Defects Laboratory, University of Washington, Seattle, WA, USA) was obtained in full compliance with the ethical guidelines of the NIH and under Institutional Review Board approval from the University of Washington and the University of Nebraska Medical Center. Tissue was incubated with 0.25% trypsin for 30 min, neutralized with 10% FBS, and further dissociated by trituration. The resulting single-cell suspension was cultured on poly-d-lysine coated plates in neurobasal medium supplemented with 0.5 mM l-glutamine, 500 U/ml penicillin, and 500 μg/ml streptomycin and B27 supplement (Gibco; Invitrogen). Neurons were cultured in vitro for 11 d before any experiment; to confirm purity of the primary neuron culture, they were stained for MAP2 (Sternberger Monoclonals).

Transient overexpression and inhibition of miR-142

The pEP-miR-142 plasmid (Cell Biolabs, San Diego, CA, USA) was used for overexpression of precursor miR-142 (pre-miR-142), and pEP-miR-null plasmid was used as control. MiRVana microRNA mimics miR-142-3p and miR-142-5p (Invitrogen) were used for selective overexpression of either of the mature strands. For inhibition of miR-142 activity, hsa-miR-142-3p and -5p 5′-fluorescein labeled power inhibitors (Exiqon, Woburn, MA, USA) were used. XtremeGene HP transfection reagent (Roche Applied Science) was used for transient transfections.

Preparation of miR-142-expressing lentivirus and transduction of primary human neurons

MiR-142-expressing lentivirus was prepared as described previously (27). In brief, the pre-miR-142 fragment from pEP-miR-142 expression vector was subcloned into the Pac1/NheI cloning sites of the FUGW lentivirus backbone vector. HEK293T cells were transfected with miR-142-FUGW, Δ8.9, and vesicular stomatitis virus G protein using XtremeGene HP transfection reagent (Roche Applied Science) according to manufacturer's protocol. Cell supernatant containing virions was collected at 48 and 72 h after transfection and concentrated by ultracentrifugation. FUGW without pre-miR-142 insert was used as control. Lentivirus titer was determined using HIV p24 ELISA assay (Express Biotech International, Thurmont, MD, USA). Primary human neurons were grown in vitro for 11 d and transduced with miR-142 or control lentivirus at a concentration of 5 × 10⁶ lentiviral particles/ml. Successful transduction was confirmed by visualizing GFP expression. At 6 d after transduction, cells were collected for Western blotting and qRT-PCR.

3′-UTR luciferase reporter assay

HEK293T cells were cotransfected with SIRT1 3′-UTR luciferase reporter vector (Switchgear Genomics, Menlo Park, CA, USA) and either miR-142 or miR-null plasmid, using XtremeGene transfection reagent. At 48 h after transfection, luciferase activity was measured using LightSwitch Luciferase Assay Reagent (Switchgear Genomics). Empty 3′-UTR vector was used as positive control. Site-directed mutagenesis of SIRT1 3′-UTR at the miR-142-5p and -3p binding sites was carried out using QuickChange Lightning site-directed mutagenesis kit (Stratagene, La Jolla, CA, USA) and the following primers.

For mutating miR-142-5p binding site: forward, 5′-CAGCTGCAAAAGCTTCTAGTCTTTCAAGAAGTTCATAACCCGGGAAATTGCACAGTAAGCATTTATTTTTCAGACCATTT-3′, reverse, 5′-AAATGGTCTGAAAAATAAATGCTTACTGTGCAATTTCCCGGGTTATGAACTTCTTGAAAGACTAGAAGCTTTTGCAGCTG-3′.

For mutating miR-142-3p binding site: forward, 5′-CTACTTATAAGATGTCTCAATCTGAATTTATTTGGCTGGATCCAAGAATGCAGTATATTTAGTTTTCCATTTGCATGATG-3′, reverse, 5′-CATCATGCAAATGGAAAACTAAATATACTGCATTCTTGGATCCAGCCAAATAAATTCAGATTGAGACATCTTATAAGTAG-3′.

Western blotting

Whole-cell lysates were prepared using RIPA buffer (50 mM Tris/HCl, pH 8; 150 mM NaCl; 1% Nonidet P-40; 0.5% sodium deoxycholate; and 0.1% SDS) and protein quantification was carried out using Pierce BCA protein assay (Thermo Scientific). Protein (5–15 μg) was loaded in each lane of NuPAGE 4–12% Bis-Tris gels (Invitrogen). Separated proteins were transferred onto nitrocellulose membranes using iBlot (Invitrogen). The membranes were blocked in SuperBlock (TBS) blocking buffer (Thermo Scientific) and then incubated overnight at 4°C with primary antibody. The following primary antibodies were used: rabbit polyclonal SIRT1 (1:200; Santa Cruz Biotechnology, Santa Cruz, CA, USA) and rabbit polyclonal actin (1:5000; Sigma-Aldrich). This was followed by incubation with secondary antibody, HRP-conjugated anti-rabbit IgG (1:20,000; Thermo Scientific), for 1 h at room temperature. Blots were developed using SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific), imaged and quantified using Carestream MI software.

Statistical analysis

Statistical analysis was performed using Prism software (GraphPad, La Jolla, CA, USA). Student's t test or 1-way ANOVA followed by Bonferroni multiple comparisons or Dunnett's post hoc test were performed as applicable.

RESULTS

MiR-142 is expressed in neurons and myeloid cells in SIVE

In a previous study, we had determined that miR-142-3p and -5p are up-regulated in caudate nucleus and hippocampus in SIVE brain compared to uninfected control brain (14). To confirm this, we performed FISH for miR-142-3p and -5p in hippocampal sections from rhesus macaques with SIVE and uninfected control macaques. We could detect expression of these miRs in the dentate gyrus region of the SIVE hippocampal sections (Fig. 1). As a positive control, U6, a noncoding snRNA, showed abundant signal in most cells in the tissue; as a negative control, a scrambled miR probe did not show any hybridization in these sections. In uninfected control, hippocampal sections miR-142-3p and -5p were below the detection limit, although U6 could still be detected (Fig. 1).

Figure 1.

FISH for miR-142-3p and -5p in hippocampal sections from rhesus macaques with SIVE and uninfected control macaques. Expression of miR-142-3p and -5p (green) was detected in the dentate gyrus region of the hippocampus in SIVE sections. In uninfected control sections, miR-142-3p and -5p were below the detection limit. Expression of the positive control U6 was found in both SIVE and uninfected hippocampus. No signal was detected with the negative control scrambled miR probe. DAPI (blue) was used to label nuclei. Scale bars = 20 μm.

Next, to determine the cell type in which miR-142 is expressed in SIVE, we performed FISH combined with IF labeling for cell type markers. In SIVE hippocampal sections, both miR-142-3p and -5p were localized within MAP2-labeled neurons in the dentate gyrus (Fig. 2A) and the CA4 region (Fig. 2B). Expression of both these miRs was below the detection limit in uninfected control hippocampal sections. Hybridization with the scrambled miR probe did not reveal any signal.

Figure 2.

Combined FISH and IF for miR-142 and MAP2. Both miR-142-3p and -5p (green) appear to localize within MAP2 (red)-labeled neurons in SIVE sections in the dentate gyrus (A) and CA4 region (B) of the hippocampus. Scrambled miR probe was used as negative control for hybridization. DAPI (blue) was used to label nuclei. MiR-142-3p and -5p probes did not show any hybridization in uninfected rhesus macaque hippocampal sections. Scale bars = 20 μm.

While MiR-142 has been detected in neurons following nerve crush injury (21) and in striatal postsynaptic densities (20), is it best known for its enrichment in cells of the hematopoietic lineage and activity in myeloid lineage cell differentiation (18, 19, 28). In HIVE/SIVE, macrophages infiltrate the brain in perivascular regions and, likely along with endogenous microglia, are found intraparenchymally in nodules. We therefore performed FISH coupled with IF labeling in regions of the brain where inflammatory lesions were present. Indeed, both miR-142-3p and -5p partially colocalized with CD163-labeled macrophages/microglia (Fig. 3), as well as being expressed in cellular processes surrounding the lesions. No colocalization was observed with the astrocyte marker glial fibrillary acidic protein (GFAP). The negative control scrambled miR probe and examination of the miR-142 probes in uninfected brain sections did not reveal signals.

Figure 3.

Combined FISH and IF for miR-142, GFAP, and CD163. In SIVE sections in brain regions that had large number of microglial nodules, miR-142-3p and -5p (green) partially colocalized with the macrophage/microglia marker CD163 (red). No colocalization of either probe was observed with GFAP (magenta), an astrocyte marker. DAPI (blue) was used to label nuclei, and a scrambled miR probe was used as negative control for hybridization. Scale bars = 20 μm.

MiR-142 target prediction

To identify downstream miR-142-3p and -5p targets, we used the microRNA.org August 2010 release. This method of miR target prediction was used because, in addition to seed match, binding energy, and sequence conservation, it also aligns the entire miR to the target 3′-UTR and takes into consideration stabilizing complementary base-pairing outside the seed region, therefore minimizing false-negative predictions (29, 30). In addition, it gives each miR-target 3′-UTR pair a score, (mirSVR score) that represents the probability of down-regulation (31). For our analyses, we used only the predicted targets that had a recommended mirSVR score ≤−0.1. Using this method, we found 4922 predicted targets for miR-142-5p and 1904 for miR-142-3p, as well as 816 common target genes.

Given the large number of targets for miR-142-3p and -5p, we first wanted to determine which of these could be relevant to SIVE pathogenesis. To do so, we analyzed microarray data from a previous study (26), comparing gene expression in SIVE and uninfected hippocampus. Using IPA, we identified transcription regulators whose predicted altered activity corresponded to the changes in gene expression found in this study (Table 2). Among the 7 transcription regulators that met our inclusion criteria, 3 that were predicted to be activated in SIVE hippocampus (TP53, STAT3, and JUN) are regulated by the NAD-dependent deacetylase SIRT1. SIRT1 deacetylates and down-regulates the transactivation functions of TP53 (32), STAT3 (33), and JUN (34). Interestingly, SIRT1 was also predicted as a target for miR-142-3p and -5p in our in silico target prediction analysis. Therefore, we hypothesized that miR-142 contributes to the altered gene expression profile in SIVE by down-regulating SIRT1, which could lead to the predicted increase in activity of the above transcription regulators.

Table 2.

IPA analysis of differentially expressed genes in SIVE hippocampus compared to uninfected control, filtered for transcription regulators that had a z score with an absolute value >1

Upstream regulator	Bias-corrected z score	P value of overlap
TP53	1.570	1.20E-15
HMGB1	1.443	8.21E-08
SMARCB1	1.438	2.11E-07
STAT3	1.360	6.45E-16
JUN	1.029	1.16E-09
KDM5B	–1.288	1.32E-03
MYC	–1.895	2.36E-05

Overexpression of miR-142 decreases and its inhibition increases SIRT1 expression

To test our hypothesis that SIRT1 expression is down-regulated by miR-142, we transfected HEK293T cells with pEP-miR-142 plasmid, which contains the entire pre-miR-142 sequence and can be processed inside the cells to generate miR-142-3p and -5p. The pEP-miR-null transfected HEK293T cells were used as control. Cells were harvested 48, 72, and 96 h post-transfection, and miR-142 expression was confirmed by qRT-PCR. Western blot for SIRT1 revealed 3.7- and 2.5-fold decrease in SIRT1 expression at the 72- and 96-h time points after miR-142 transfection, respectively, but this decrease was significant (P<0.05) only at the 72-h time point (Fig. 4A, B).

Figure 4.

MiR-142 regulates SIRT1 protein expression. A) Representative Western blot showing reduction in SIRT1 after pre-miR-142 overexpression in HEK293T cells for 72 and 96 h. B) Quantification of Western blot from 3 independent experiments revealed a decrease in SIRT1 protein levels at 72 h (−3.7-fold) and 96 h (−2.5-fold) after miR-142 transfection in HEK293T cells compared to miR-null transfection. *P < 0.05; 1-way ANOVA with Bonferroni's multiple correction test. C) Representative Western blot showing reduction in SIRT1 after transfection of miR-142-5p in HEK293T cells. D) Quantification of Western blot from 3 independent experiments revealed that the reduction of SIRT1 levels after only after miR-142-5p mimic transfection (−1.6-fold change) was statistically significant. NT, nontransfected; Cont-mim, control miR mimic; 3p-mim, miR-142-3p mimic; 5p-mim, miR-142-5p mimic. **P < 0.01; 1-way ANOVA with Bonferroni's multiple correction test. E) Representative Western blot showing increase in SIRT1 after inhibition of miR-142-5p in K562 cells. F) Quantification of Western blot from 3 independent experiments revealed that inhibition of only miR-142-5p increased SIRT1 protein level significantly (1.33-fold change). Cont-in, control miR inhibitor; 3p-in, miR-142-3p inhibitor; 5p-in, miR-142-5p inhibitor. All Western blots were normalized to β-actin. Y axis represents mean expression level. Error bars = sem. *P < 0.05; unpaired t test.

Next, we determined the contribution of the two miR-142 strands (-3p and -5p) to inhibition of SIRT1 expression. For this purpose we transfected HEK293T cells with mimics for mature miR-142-3p and miR-142-5p. Cells were harvested 72 h after transfection, and the respective miR expression verified by qRT-PCR. Western blotting revealed a significant decrease in SIRT1 expression after miR-142-5p transfection (fold change −1.6, P<0.01), whereas a significant change was not found after miR-142-3p transfection (Fig. 4C, D), leading us to believe that the miR-142-5p binding site on the SIRT1 3′-UTR likely dominated.

To further verify this miR-142-SIRT1 interaction, we next inhibited endogenous miR-142 activity and determined whether SIRT1 is expression is elevated as a consequence. As neither HEK293T cells nor primary human cultured neurons express miR-142, we chose the hematopoietic cell line K562 cells for this purpose. Transfection of hsa-miR-142-3p and -5p inhibitors in this cell line followed by Western blotting revealed, as predicted, a significant increase in SIRT1 levels only after inhibition of miR-142-5p (fold change 1.3, P<0.05; Fig. 4E, F).

SIRT1 is a direct miR-142-5p target

Our in silico target prediction had indicated that both miR-142-3p and -5p have binding sites in the SIRT1 3′-UTR (Fig. 5A). Of note, the miR-142-5p binding site is conserved only in humans, chimpanzee, and rhesus macaque (Fig. 5B). To examine whether miR-142 can act through these sites in the SIRT1 3′-UTR and inhibit translation, we performed 3′-UTR luciferase reporter assay. Cotransfection of a SIRT1 3′-UTR luciferase reporter vector with miR-142 in HEK293T cells yielded luciferase activity that was 2.3-fold lower than that of the same reporter cotransfected with a control, miR-null (Fig. 5C). To determine conclusively whether either the miR-142-3p or -5p binding site on the SIRT1 3′-UTR is functionally dominant we mutated each binding site individually and also generated a double mutant (Fig. 5A). Mutation of the miR-142-3p binding site did not affect the ability of miR-142 to inhibit SIRT1 translation, as evidenced by a 2.8-fold decrease in luciferase activity. However, miR-142 could no longer inhibit translation with the miR-142-5p mutant or with the double-mutant 3′-UTRs (Fig. 5C). This finding conclusively proved that SIRT1 is a bona fide miR-142-5p target, corroborating our miR-142-5p overexpression and inhibition studies described previously.

Figure 5.

SIRT1 is a direct miR-142-5p target. A) Alignment details of miR-142-3p and -5p on the SIRT1 3′-UTR. The miR seed sequences and their complementary sequences on the 3′-UTR are in italics. These complementary sequences were mutated to generate SIRT1 3′-UTRs that could not bind to miR-142-3p (SIRT1 3′-UTR-3pMut), or miR-142-5p (SIRT1 3′-UTR-5pMut) or both (SIRT1 3′-UTR-DM). B) SIRT1 3′-UTR sequence in different species showing that the miR-142-5p seed match is conserved only within human, chimpanzee, and rhesus macaque. C) 3′-UTR luciferase assay showing that miR-142 transfection reduces the translation of luciferase reporter upstream of SIRT1 3′-UTR, compared to miR-null transfection (−2.3-fold change). Mutating the miR-142-3p binding site did not alter this function (−2.8-fold change). MiR-142 could not reduce expression of the luciferase reporter after mutation of miR-142-5p binding site or in the double mutant. Empty 3′-UTR was used as negative control. The experiment was repeated 3 times. Y axis represents mean luciferase activity after miR-142 transfection relative to miR-null transfection. Error bars = sem. *P < 0.05; 1-way ANOVA with Dunnett's test.

MiR-142 represses SIRT1 expression in primary neurons

As our FISH and IF experiments had demonstrated that miR-142 is expressed in neurons in SIVE, we wanted to confirm that it can regulate SIRT1 expression in this cell type. Therefore, we transduced primary human neurons from 5 individual donors with miR-142 or control lentivirus. Human neurons (as opposed rodent neurons) were used because the miR-142-5p binding site on the SIRT1 3′-UTR, which was determined as the functional site in our in vitro studies, is conserved only within primates. At 6 d after transduction, cells were harvested for Western blotting. While average transduction efficiency was just 35—45%, we could still detect a small but significant decrease in SIRT1 levels due to miR-142 in the primary neurons (mean fold change −1.3, P<0.05; Fig. 6A).

Figure 6.

A) Decrease in SIRT1 protein levels was also observed in human neurons transduced with miR-142 expressing lentivirus. Experiment was repeated in human neurons from 5 donors. Representative Western blots from 3 human neuron donors are shown. Mean fold change of SIRT1 expression = −1.3, P < 0.05; paired t test. B) MiR-142-3p is up-regulated in SIVE macrophages/microglia compared to control (6.4-fold change; n=4 for uninfected samples, n=3 for SIVE). C) MiR-142-5p is up-regulated in SIVE macrophage/microglia compared to control (fold change=3.9, n=4 for uninfected samples, n=3 for SIVE). D) SIRT1 mRNA level is reduced in SIVE macrophages/microglia compared to control (−33-fold change; n=4). Y axis represents mean expression level in a logarithmic scale. Error bars = sem. *P < 0.05; unpaired t test.

SIRT1 is down-regulated and miR-142 is up-regulated in macrophages/microglia from SIVE animals

In addition to hippocampal neurons, miR-142 expression was also detectable in macrophage/microglia nodules in SIVE. To determine whether miR-142 can down-regulate SIRT1 in these cell types and to confirm that the miR-142-SIRT1 pathway is operant in the disease context, we performed qRT-PCR on macrophage/microglia RNA samples from uninfected and SIVE rhesus macaques. Indeed miR-142-3p and -5p were up-regulated 6.4- and 3.9-fold, respectively, in macrophages/microglia from SIVE animals (Fig. 6B, C). In these SIVE samples, SIRT1 was found to be down-regulated by 33-fold compared to uninfected control (Fig. 6D). Since our in vitro assays confirmed that SIRT1 is a direct miR-142-5p target, this large decrease in SIRT1 expression likely can be attributed, at least in part, to the increase in miR-142-5p expression in SIVE.

DISCUSSION

In this study, we report that miR-142 is expressed in hippocampal neurons and brain myeloid cells in SIVE. Overexpression of this miR led to decrease in SIRT1 protein level in cell lines and primary human neurons. In macrophages/microglia from rhesus macaques with SIVE, up-regulation of miR-142 was accompanied by a concomitant down-regulation of SIRT1. In addition, we provide conclusive evidence that SIRT1 is a direct miR-142-5p target.

Differential expression of miRs in the brain during HIV infection has been reported in several studies (14, 15, 35, 36). However, only two studies have quantified changes in brain miR expression profile in HIVE brain, and only one has done so in SIVE (14, 15). MiR-142 (both the -3p and -5p forms) was the only miR identified to be up-regulated by both these studies. To unravel the role of miR-142 in the brain in SIVE, we first determined the cell types in which it is up-regulated. Simultaneous FISH and IF was used for this purpose. MiR-142-3p and -5p were found to be expressed in neurons and macrophage/microglia nodules in SIVE. While miR-142 has been extensively studied in the hematopoietic system, where it is expressed in T-lymphocytes and myeloid cells, its expression and function in neurons is less well established. In this context, up-regulation of miR-142 was reported in striatal postsynaptic densities following cocaine treatment (20) and in the retina in animal models of retinitis pigmentosa (37). A recent study demonstrated an increase in miR-142 expression in neurons following peripheral nerve crush (21). The pathological feature common to both the latter conditions is inflammation, chronic in case of retinitis pigmentosa and acute in nerve crush injury (38–40). MiR-142 therefore appears to be up-regulated in response to inflammation. Neuroinflammation, characterized by excessive activation of astrocytes and microglia, accompanied by infiltration of macrophages, is a hallmark of HIVE and SIVE and could be the cause of the observed up-regulation of miR-142.

Next, we performed in silico search for putative miR-142 targets using microRNA.org. Of the targets common to both miR-142-3p and -5p, we chose SIRT1 for further examination. SIRT1 is a NAD-dependent protein deacetylase well known for its role in mediating the longevity promoting effects of calorie restriction (41). Neuroprotective roles of SIRT1, both in vitro and in vivo, have been well documented (42) and brain-specific knockout of SIRT1 impairs memory and synaptic plasticity (43). SIRT1 has been found to have effects in a number of models of neurodegenerative diseases. In cellular and animal models of Alzheimer's disease, SIRT1 overexpression or activation leads to reduced beta amyloid production and toxicity (44, 45). In addition, SIRT1 deacetylates microtubule binding protein tau and promotes its clearance, thus preventing formation of neurotoxic tangles (46). In Huntington's disease, SIRT1 overexpression ameliorates and knockout exacerbates the neurotoxic effects of mutant HTT (44, 47). Overexpression of SIRT1 in a mouse model of Parkinson's disease led to reduction in alpha synuclein aggregates and increased life span, while SIRT1 knockout had the opposite effects (48). Dysregulation of autophagy has also been linked to a number of neurodegenerative conditions, and SIRT1 may promote neuronal survival by regulating autophagy. Overexpression of SIRT1 leads to induction of autophagy, and SIRT1 knockdown interfered with the same (49); similarly inhibition of SIRT1 by the chemical inhibitor Sirtinol led to impairment of autophagy (50).

Interestingly, SIRT1 has also been shown to interact with RNA binding proteins. TDP43, a constituent of the intracellular protein aggregates that occur in amyotrophic lateral sclerosis and fronto-temporal lobar dementia, forms a complex with the RNA-binding proteins Staufen1 and fragile-X-mental retardation protein, which bind to the 3′-UTR of SIRT1 mRNA and stabilize it (51). Consequently, loss of function of TDP43 leads to decrease in SIRT1 levels due to increased degradation of its mRNA. In a similar manner, in response to DNA damage, the RNA-binding protein HuR binds to the 3′-UTR of SIRT1 mRNA and increases its half-life (52). The resultant increase in SIRT1 protein level leads to deacetylation of NBS1, a protein that is involved in repair of DNA double-strand breaks (53).

SIRT1 regulates several pathways that promote cell survival. By deacetylating TP53, FOXO3a, and Ku70, SIRT1 reduces expression of proapototic proteins and increases that of antiapoptotic proteins, thereby preventing apoptosis in response to oxidative or metabolic stress (54). Furthermore, SIRT1 also has antiinflammatory effects that contribute to its neuroprotective and antiaging roles. SIRT1 knockout in several cell types (including myeloid cells) led to increased expression and activity proinflammatory cytokines such as TNF-α, IL1β, and IL6, and such molecules are key candidates for pathogenesis of SIVE (55–57). Thus many mechanisms exist by which SIRT1 can influence the pathogenesis of neurodegenerative disorders.

Located at the intersection of anti-inflammatory, antiaging, and neuroprotective pathways, SIRT1, therefore, appears to be an ideal molecule for mediating disease pathogenesis in SIVE. Although, the expression, activity or role of SIRT1 in HIVE/SIVE has not been investigated to date, several studies have shown that SIRT1 expression and/or activity can be altered by HIV proteins (58–61). Tat directly binds to the deacetylase domain of SIRT1 and inhibits its activity in T cells (61). It also induces the expression of miRs-34a and 217, which target SIRT1 3′-UTR reducing protein levels (59, 60). This activity would be especially relevant in the brain, where macrophages are the primary infected cells. The miR-induced reduction of SIRT1 would thus increase Tat activity, resulting in increased HIV expression and replication. SIRT1 can in turn affect HIV transcription by deacetylating the viral protein Tat (62–64). In addition, IPA analysis of our previous hippocampal gene expression data predicted that pathways downstream of SIRT1 are perturbed in SIVE, where the SIRT1 substrates TP53, STAT3, and JUN were predicted to be activated in encephalitis. As SIRT1 is known to deacetylate and inactivate these transcription regulators (32–34), decrease in SIRT1 expression in SIVE can explain the predicted increase in their activity.

The mirSVR scores for the miR-142-3p and -5p binding sites on the SIRT1 3′-UTR were −0.1285 and −0.8442, respectively, implying that the miR-142-5p had a higher probability of down-regulating SIRT1. Interestingly, this more efficient miR-142-5p binding site on the SIRT1 3′-UTR is conserved only within primates (Fig. 5B). Indeed, through a number of complementary assays, we determined that SIRT1 is a target of miR-142, primarily through this miR-142-5p seed sequence recognition site. Reduction of SIRT1 protein level due to miR-142 overexpression was also confirmed in primary neurons, the cell type in which miR-142 was found to be up-regulated in SIVE. Also in accordance with our in vitro studies, we found that in macrophage/microglia from SIVE rhesus macaques, miR-142 and SIRT1 are, respectively, up- and down-regulated compared to uninfected controls. However, it is certainly possible that other pathways may be acting in concert with miR-142 to regulate SIRT1 levels. In this context, two other miRs, miR-34a and -217, which are induced by the HIV protein Tat, have been reported to down-regulate SIRT1 (59, 60).

In summary, the present study identified that miR-142-5p and -3p are expressed in hippocampal neurons and CD163-positive macrophage/microglial nodules in SIVE. Overexpression of miR-142 led to a decrease in SIRT1 expression, and SIRT1 is a direct miR-142-5p target. In addition, miR-142 also repressed SIRT1 protein expression in primary human neurons. In macrophages/microglia from rhesus macaques with SIVE, miR-142 up-regulation was associated with a corresponding decrease in SIRT1 expression. The decrease in SIRT1 level due to miR-142 up-regulation in the brain could directly contribute to the inflammatory and neurodegenerative manifestations of SIVE. Activation of SIRT1 or inhibition of miR-142 could help ameliorate such manifestations and further investigation into miR-142 and SIRT1 regulated cellular pathways that could be altered in SIVE is therefore warranted.