Abstract Chronic microglial activation and associated neuroinflammation are key factors in neurodegenerative diseases including HIV-associated neurocognitive disorders. Colony stimulating factor 1 receptor (CSF1R)-mediated signaling is constitutive in cells of the myeloid lineage, including microglia, promoting cell survival, proliferation, and differentiation. In amyotrophic lateral sclerosis and Alzheimers disease, CSF1R is upregulated. Inhibiting CSF1R signaling in animal models of these diseases improved disease outcomes. In our studies, CNS expression of the CSF1R ligand, colony-stimulating factor 1 (CSF1) was significantly increased in a SIV/macaque model of HIV CNS disease. Using a Nanostring nCounter immune panel, we found CSF1 overexpression was strongly correlated with upregulation of microglial genes involved in antiviral and oxidative stress responses. Using in situ hybridization, we found that CSF1R mRNA was only present in Iba-1 positive microglia. By ELISA and immunostaining with digital image analysis, SIV-infected macaques had significantly higher CSF1R levels in frontal cortex than uninfected macaques (p = 0.018 and p = 0.02, respectively). SIV-infected macaques treated with suppressive ART also had persistently elevated CSF1R similar to untreated SIV-infected macaques. Coordinate upregulation of CSF1 and CSF1R expression implicates this signaling pathway in progressive HIV CNS disease. Colony stimulating factor 1 (CSF1), Colony stimulating factor 1 receptor (CSF1R), Human immunodeficiency virus (HIV), Macaque, Microglia, Neurodegeneration, Simian immunodeficiency virus (SIV) INTRODUCTION Microglia, comprising parenchymal microglia and perivascular macrophages in the CNS, play a central role in multiple chronic neurodegenerative diseases including Alzheimers disease (AD), amyotrophic lateral sclerosis (ALS), and human immunodeficiency virus (HIV)-associated neurocognitive disorder (HAND) (1–4). In these diseases, immune activation of microglia may be pivotal in regulating neuroinflammation, a process likely dependent on disease stage (5). Knockout and inhibition studies in rodent models have shown that colony stimulating factor 1 receptor (CSF1R) signaling is essential for microglial survival, proliferation, and differentiation (6–9). Upregulation of CSF1R by microglia has been identified in AD, ALS, and other neuroinflammatory diseases including prion diseases (10–2). Of particular note, CSF1R blockade improved disease outcomes in mouse models of late-stage AD and ALS (13, 14), demonstrating that microglial immune responses can be either beneficial or harmful depending on context. CSF1R has 2 known ligands: Colony stimulating factor 1 (CSF1) and IL34. Both ligands are expressed in the CNS, but IL34 appears to be more essential to homeostatic regulation in specific areas of the brain, including the cerebral cortex, basal ganglia, and hippocampus (15). Interestingly, previous work suggests that CSF1 is more inducible than IL34 in neurodegenerative diseases (14, 16) and therefore may be the more important CSF1R ligand in chronic neurodegeneration (17). Persistent CNS inflammation including microglial immune activation is thought to drive the pathogenesis of HAND (3, 18). HIV can infect cells of the myeloid lineage, including microglia, thereby stimulating or dysregulating the neuroimmune response in HAND (3, 4). Previous work in a rhesus macaque SIVmac251 model of HIV CNS infection has shown that increased CSF1 expression occurs in CD163-positive cells (19), cells that accumulate in perivascular cuffs and microglial nodules in SIV encephalitis (16). In this report, we show that CSF1 expression increased in the SIV/pigtailed macaque model of HIV CNS disease whereas IL34 expression did not change with SIV infection. CSF1 overexpression was significantly correlated with microglial genes involved in the antiviral response to SIV, and in response to oxidative stress. In addition to upregulation of the ligand CSF1, CSF1R expression in microglia also increased with infection, implicating the CSF1-CSF1R signaling pathway in HIV CNS disease. MATERIALS AND METHODS Animals All experiments were performed on samples collected at necropsy from male juvenile pigtailed macaques (Macaca nemestrina). Samples were stored at –80 °C prior to use. Animals were either (1) mock inoculated with Lactated Ringers Solution, (2) inoculated intravenously with the neurovirulent molecular clone SIV/17E-Fr and the immunosuppressive swarm SIV/DeltaB670 (SIV), or (3) SIV inoculated and treated with an ART regimen (SIV + ART) (20–2). Untreated SIV-infected animals were euthanized at 21 days postinfection (p.i.), or 84 days p.i. unless the animal began showing symptoms of AIDS before the end point of the respective study. SIV-infected animals that were treated with ART (n = 7) were fully suppressed in CSF and plasma by 56 days p.i., and were euthanized 180 days p.i., after ∼120 days of ART suppression. SIV + ART animals were treated daily with a subcutaneous injection of 2.5 mg/kg Dolutegravir (ViiV Healthcare US, Raleigh, NC), 20 mg/kg PMPA, and 40 mg/kg FTC (Gilead, Foster City, CA). Four of the 7 treated animals were also given 20 mg of oral Maraviroc (ViiV Healthcare US) daily. This difference in treatment regimen did not cause a significant difference in time to suppression in either CSF or plasma; there was also no significant difference in the CNS-specific measures performed for this study. As this study was conducted retrospectively, all samples were not available for each animal, hence minor variation in animal group size in different experiments. At necropsy, all animals were perfused with phosphate-buffered saline. Brains were harvested and sectioned coronally at necropsy. All samples were either immersion fixed in 10% neutral buffered formalin, Streck tissue fixative, or flash-frozen. Both basal ganglia and frontal cortex sections were utilized in this study. All animal studies were approved by the Johns Hopkins Animal Care and Use Committee. RNA Measurements The Nanostring nCounter gene expression panel used for analysis included 249 genes of interest (23). The geometric means of 4 housekeeping genes were used to normalize counts. mRNA counts of genes of interest were then normalized to the geometric mean of the 4 housekeeping genes (GAPDH, RPL13A, SDHA, and TBP), as well as the geometric mean of spike-in controls. The limit of detection was determined by taking the mean of 3 negative controls plus 2 standard deviations as detailed previously (23). RNA for this panel was isolated from basal ganglia sections harvested at necropsy from uninfected and SIV-infected (84 days p.i.) macaques as described previously (23). One hundred nanogram of RNA was used for input and 100 ng of MS2 phage RNA (Roche, Basel, Switzerland) run in triplicate served as a negative control. For quantitative RT-PCR, RNA was isolated from grey and white matter frontal cortex flash frozen at –80 °C. Tissue was placed in Fast RNA tubes with lysing matrix D (MP Bio, Solon, OH) and RNA STAT-60. Each tube was homogenized for 30 seconds using the Fast Prep-24 homogenizer (MP Bio). Tubes were then incubated at room temperature for 5 minutes, and 200 μL of chloroform was added. Tubes were then shaken for 15 seconds and incubated at room temperature for 3 minutes. Samples were then centrifuged at 14 000g for 15 minutes at 4 °C. The aqueous portion was then added to a fresh tube with 500 μL of ice-cold 2-isopropanol and vortexed. After an overnight incubation at –20 °C, tubes were centrifuged at 14 000g for 15 minutes at 4 °C and the isopropanol was removed. The pellet was then washed in ice-cold 70% ethanol. Tubes were centrifuged at 10 000g for 5 minutes at 4 °C and the ethanol was removed. The pellet was then allowed to air dry for ∼15 minutes at room temperature. RNA isolation was completed using the Qiagen RNeasy kit (Qiagen, Frederick, MD) according to the manufacturer’s protocol. Quality and concentration of the isolated RNA was determined using Nanodrop. One microgram of RNA was added to each RT reaction. The High Capacity cDNA reverse transcription kit was used (Applied Biosystems, Carlsbad, CA) with samples run in duplicate with a no reverse transcriptase control for each as well as no template controls. Reverse transcription was performed with the PTC-200 (MJ Research, Port Republic, NJ). The samples were held at 25 °C for 10 minutes to anneal, then 37 °C for 120 minutes for reverse transcription, and finally, 85 °C for 5 minutes to inactivate the reverse transcriptase. Samples were held at 4 °C overnight before being stored at –20 °C. Four microliter of cDNA was used for qPCR per sample. Each was run in duplicate with no template controls and no reverse transcriptase controls. The TaqMan Universal Master Mix II or the Gene Expression Master Mix was used (Applied Biosystems) with CSF1 (cat. Rh02621778_m1) and IL34 (cat. Rh01050928_m1) probes (Applied Biosystems); all counts were normalized to 18S ribosomal RNA and reported as ΔΔCt (cycle threshold). CSF1R Cellular Localization and Measurements CSF1R RNA was visualized in cells by in situ hybridization. Combined CSF1R ISH—Iba-1 IHC double staining was performed on frontal cortex using the Leica Bond RX automated system (Leica Biosystems, Richmond, IL). Tissue was fixed in 10% neutral buffered formalin for 24 hours and embedded in paraffin before sectioning at 5 μM. A CSF1R probe (cat. 310818, Advanced Cell Diagnostics, Newark, NJ) was used with the RNAScope 2.5 LS Assay-RED Kit according to manufacturer’s protocol. Epitope retrieval was performed by heating to 95 °C for 20 minutes in EDTA-based ER2 buffer (Leica Biosystems). Anti-Iba-1 antibody (cat. 019-19741, Wako, Richmond, VA) was diluted 1:500. Slides were counterstained with hematoxylin. CSF1R IHC was performed using indirect, alkaline phosphatase-based immunostaining on Streck tissue fixative-fixed frontal cortex sections using the Bond RX automated system with the Bond Polymer Refine Red kit (Leica Biosystems). Each slide was heated to 95 °C for 20 minutes in EDTA-based ER2 buffer for heat-induced epitope retrieval. Anti-CSF1R (Santa Cruz Technologies, cat. sc-692, Santa Cruz, CA) was used as the primary antibody (1:50 dilution). Positive immunoreactivity was visualized by labeling with the Bond Polymer Refine Red kit (alkaline phosphatase) (Leica Biosystems). Slides were counterstained with hematoxylin. Image acquisition and analysis to measure CSF1R was performed with Nikon Elements software (Nikon, Melville, NY) by generating composite images composed of 3 × 12 200× high-power fields encompassing both grey and white matter. A threshold for positive staining was established using a set of blinded images and applied to all images to calculate the area fraction (%ROI) representing positive staining. Regions of interest (ROIs) were drawn around grey and white matter separately. The median size of the areas analyzed for grey and white matter were 2.6 × 106 μm2 and 8.6 × 105 μm2, respectively. CSF1R ELISAs were performed on frontal cortex homogenates prepared from 50 mg of tissue placed in 200 μL of 1× cell lysis buffer (Cell Signaling and Technologies, Danvers, MA) with added protease and phosphatase inhibitors (Roche, Basel, Switzerland) and homogenized with a hand-held homogenizer. 300 μL of 1× cell lysis buffer was then added for a total of 500 μL. Samples were then sonicated using a cup sonicator 5 times for 20 seconds with 15-second breaks. Tubes were then rotated at 4 °C for 2 hours and pelleted at 4 °C for 5 minutes at 14 K rpm. Supernatant was collected and used. Protein concentrations of these homogenates were determined using the Qubit protein assay kit according to manufacturer’s protocol (Invitrogen, Carlsbad, CA). These concentrations were validated by running 5 μg of protein of each sample on a Criterion Stain Free Tris–HCl 4%–20% separation gel and read using the Gel Doc EZ Imager (Bio-Rad, Hercules, CA). The Biomatik Human MCSFR ELISA kit (Biomatik, Wilmington, DE) was used per manufacturer’s protocol. Thirty micrograms of protein from each sample were added in duplicate to antibody precoated wells. TMB solution was allowed to develop for 15 minutes at room temperature. Wells were read at 450 nm immediately after the addition of stop solution. Statistical Analysis Prism was used to calculate all statistics except the Benjamini–Hochberg correction, which was applied using the p.adjust function in R. The Benjamini–Hochberg false discovery rate was set to 0.05. Nonparametric statistical tests were used (Mann–Whitney and Spearman correlation). Statistical significance was attributed to p values less than or equal to 0.05. RESULTS CSF1 Expression Increased in SIV-Infected Pigtailed Macaques Alterations in expression of both CSF1R ligands CSF1 and IL34 have been reported in vivo in an SIVmac251/rhesus macaque model and with HIV infection in vitro (16, 17, 24, 25) To evaluate CSF1 expression in the SIV/pigtailed macaque model of HIV CNS disease, we analyzed a Nanostring nCounter dataset including 123 genes above the limit of detection (Supplementary Data Table S1) obtained using RNA isolated from basal ganglia from SIV-infected macaques euthanized 21 days p.i. (n = 6) or 84 days p.i. (n = 13). Uninfected pigtailed macaques (n = 10) served as controls (23). CSF1 mRNA levels were significantly increased in SIV-infected pigtailed macaques at 84 days p.i. compared with uninfected controls (p = 0.003, Mann–Whitney, Fig. 1). To determine whether development of SIV encephalitis was related to CSF1 expression levels, the 84 day p.i. SIV-infected group was subdivided into SIV-infected animals with (n = 8) or without SIV encephalitis (n = 5). CSF1 mRNA expression was significantly increased in the animals with SIV encephalitis versus those without encephalitis (p = 0.002, Mann–Whitney, Fig. 1). FIGURE 1. View largeDownload slide Nanostring analysis of CSF1 mRNA expression in SIV-infected pigtailed macaques. CSF1 mRNA counts in basal ganglia of SIV-infected macaques were significantly increased versus uninfected animals. CSF1 mRNA expression was also significantly higher in animals with SIV encephalitis compared with SIV-infected animals without encephalitis (encephalitis = ▲; no encephalitis = ▽). FIGURE 1. View largeDownload slide Nanostring analysis of CSF1 mRNA expression in SIV-infected pigtailed macaques. CSF1 mRNA counts in basal ganglia of SIV-infected macaques were significantly increased versus uninfected animals. CSF1 mRNA expression was also significantly higher in animals with SIV encephalitis compared with SIV-infected animals without encephalitis (encephalitis = ▲; no encephalitis = ▽). To determine whether CSF1 expression was altered during the asymptomatic phase of infection, CSF1 mRNA counts in basal ganglia of SIV-infected pigtails euthanized at 21 days p.i. were compared with uninfected and 84 day groups. The asymptomatic SIV group had higher CSF1 expression than uninfected controls (p = 0.09, Mann–Whitney; median = 196.5 mRNA copies) comparable to the terminal nonencephalitic SIV-infected animals sacrificed at 84 days p.i (median = 122 mRNA copies). This intermediary level of CSF1 expression suggests that CSF1 alterations develop early in the course of disease; analysis of larger groups is needed to confirm this finding. Further analysis of the Nanostring nCounter dataset revealed differential expression of 44 additional genes in the basal ganglia of SIV-infected pigtailed macaques sacrificed 84 days p.i. (n = 13) compared with uninfected controls (n = 10). Fold change in expression was calculated by dividing the median expression of the SIV 84 day p.i. mRNA counts by that of the uninfected controls. In cases of downregulation, the value of uninfected median counts divided by the SIV-infected counts was a negative number. Statistical significance was calculated by Mann–Whitney with a Benjamini–Hochberg adjustment (p ≤ 0.05). If any gene had mRNA counts below the limit of detection, their values were set at the limit of detection for use in Mann–Whitney tests. These values were also used to identify strength of correlations between differentially expressed genes and CSF1 RNA levels. Twenty-nine of the 45 genes found to be differentially expressed with SIV were highly correlated with CSF1 mRNA expression levels. A subset of these genes grouped functionally is represented in the Table; the complete list of genes above the limit of detection is listed in Supplementary Data Table S1. Correlation was determined using Spearman’s rank analysis with a Benjamini–Hochberg adjustment (p ≤ 0.05). CSF1 mRNA levels were significantly positively correlated with genes involved in receptor tyrosine kinase signaling, including STAT3. As CSF1R is the only known receptor of CSF1, the correlation between CSF1 RNA with RTK genes indicates that increased CSF1 promotes CSF1R signaling. CSF1 expression was also positively correlated with microglial markers. Of particular interest, CD163 and CD68 expression on microglia have been associated with immune activation and phagocytosis (26). CSF1 mRNA levels were also strongly correlated with ALOX5AP, a microglial protein essential in the production of leukotrienes that serves as a positive regulator of inflammation. As such, ALOX5AP activity has been shown to increase neuronal damage in vitro (27). These correlations taken together suggest a regulatory relationship between CSF1R signaling and regulation of microglial immune responses in SIV infection. TABLE. With SIV Infection, a Subset of Genes Expressed in the Brain Was Significantly Correlated With CSF1 mRNA Counts Uninfected Versus SIV mRNA Counts Correlation With CSF1 mRNA Gene Fold Change SEM p Value Spearman R p Value RTK signaling STAT1 7.9 U: 20.49 0.002 0.60 0.051 S: 468.6 STAT3 2.1 U: 24.88 0.05 0.96 0.002 S: 110.4 STAT5a 2.1 U: 7.42 0.01 0.82 0.005 S: 19.36 TYK2 1.3 U: 8.67 0.04 0.63 0.04 S: 20.24 Mφ/microglia CD14 4.2 U: 10.64 0.002 0.87 0.003 S: 82.75 CD16* 6.9 U: 0.70 *0.002 0.71 0.016 S: 13.49 CD68* 9.0 U: 4.05 *0.009 0.80 0.006 S: 82.22 CD163 13.8 U: 8.80 0.004 0.72 0.01 S: 316 ALOX5AP 3.1 U: 11.78 0.009 0.75 0.01 S: 85.05 Immune activation CCL2 50.6 U: 2.78 0.002 0.84 0.004 S: 288.3 IRF1 17.7 U: 3.33 0.002 0.92 0.02 S: 137.3 IRF3 1.2 U: 41.34 0.04 0.72 0.02 S: 47.8 MX1 48.9 U: 42.88 0.002 0.74 0.01 S: 1957 Oxidative stress SOD2 7.2 U: 112.2 0.009 0.91 0.002 S: 5744 GPX1 2.2 U: 104.4 0.006 0.74 0.005 S: 191.3 GSTZ1 1.4 U: 12.59 0.054 0.82 0.004 S: 12.16 Uninfected Versus SIV mRNA Counts Correlation With CSF1 mRNA Gene Fold Change SEM p Value Spearman R p Value RTK signaling STAT1 7.9 U: 20.49 0.002 0.60 0.051 S: 468.6 STAT3 2.1 U: 24.88 0.05 0.96 0.002 S: 110.4 STAT5a 2.1 U: 7.42 0.01 0.82 0.005 S: 19.36 TYK2 1.3 U: 8.67 0.04 0.63 0.04 S: 20.24 Mφ/microglia CD14 4.2 U: 10.64 0.002 0.87 0.003 S: 82.75 CD16* 6.9 U: 0.70 *0.002 0.71 0.016 S: 13.49 CD68* 9.0 U: 4.05 *0.009 0.80 0.006 S: 82.22 CD163 13.8 U: 8.80 0.004 0.72 0.01 S: 316 ALOX5AP 3.1 U: 11.78 0.009 0.75 0.01 S: 85.05 Immune activation CCL2 50.6 U: 2.78 0.002 0.84 0.004 S: 288.3 IRF1 17.7 U: 3.33 0.002 0.92 0.02 S: 137.3 IRF3 1.2 U: 41.34 0.04 0.72 0.02 S: 47.8 MX1 48.9 U: 42.88 0.002 0.74 0.01 S: 1957 Oxidative stress SOD2 7.2 U: 112.2 0.009 0.91 0.002 S: 5744 GPX1 2.2 U: 104.4 0.006 0.74 0.005 S: 191.3 GSTZ1 1.4 U: 12.59 0.054 0.82 0.004 S: 12.16 Uninfected versus SIV p values were calculated using normalized mRNA counts (Mann–Whitney). Correlations were calculated only using SIV mRNA counts. All p values were adjusted using Benjamini–Hochberg correction. * Values below the limit of detection were set at the limit of detection to calculate the conservative p value. Given these correlations between CSF1 and microglial immune activation, the finding that CSF1 mRNA expression also highly correlated with genes essential in immune signaling, particularly those involved in the antiviral response such as CCL2 and Mx1, was also consistent with a pro-inflammatory environment in the CNS. Correlations with CSF1 also extended to genes associated with oxidative stress including SOD2, further supporting a relationship between CSF1R signaling and a functional immune response in the CNS. CSF1 and IL34 mRNA Expression in SIV-Infected Pigtailed Macaque Frontal Cortex To validate the finding of increased CSF1 expression from the Nanostring nCounter analyses and to expand measurements to include both CSF1R ligands in grey and white matter, quantitative RT-PCR was performed to quantify CSF1 and IL34 expression using samples from frontal cortex (Fig. 2). CSF1 expression was increased in both grey and white matter frontal cortex samples (Fig. 2, upper row). In contrast, IL34 did not change with SIV infection in either grey or white matter (Fig. 2, lower row). This ligand-specific difference has been reported in multiple CNS inflammation models (14, 16), and implies that CSF1 is selectively induced during chronic neuroinflammation whereas IL34 expression is not altered. FIGURE 2. View largeDownload slide CSF1 mRNA significantly increased with SIV infection. CSF1 mRNA expression measured by quantitative RT-PCR significantly increased in SIV-infected macaques (frontal cortex) compared with uninfected animals in both grey and white matter (top row). In contrast, there was no significant difference in IL34 expression in either grey or white matter (no encephalitis = ▽; encephalitis = ▲;). FIGURE 2. View largeDownload slide CSF1 mRNA significantly increased with SIV infection. CSF1 mRNA expression measured by quantitative RT-PCR significantly increased in SIV-infected macaques (frontal cortex) compared with uninfected animals in both grey and white matter (top row). In contrast, there was no significant difference in IL34 expression in either grey or white matter (no encephalitis = ▽; encephalitis = ▲;). CSF1R Expression Is Increased in Frontal Cortex White Matter of SIV-Infected Pigtailed Macaques Given upregulation of CNS CSF1 in our SIV/macaque model of HIV CNS disease, we next evaluated the CSF1 receptor, CSF1R, in the CNS. To identify the cell populations expressing CSF1R, in situ hybridization to detect CSF1R RNA was performed on macaque brain sections (Fig. 3). CSF1R mRNA was present in both Iba-1 positive perivascular macrophages and parenchymal microglia (Fig. 3B, C). CSF1R-Iba1 double-positive cells were distributed across both grey and white matter, including perivascular and subpial locales, suggesting that both resident and infiltrating myeloid cells express CSF1R. Of particular interest, highest numbers of CSF1R-Iba1 double-positive cells were present in perivascular cuffs and microglial nodules. FIGURE 3. View largeDownload slide Constitutive and induced CSF1R expression was restricted to microglia. (A)CSF1R ISH bright-field shows positive staining (red) in both parenchymal microglia (arrows) and perivascular macrophages (arrowheads). (B)CSF1R RNA (red) colocalizes with Iba1-positive microglia (brown). Double stained cells are denoted by arrowheads. (C) Overlaid CSF1R ISH fluorescence image of (B) confirmed localization of CSF1R RNA (arrowheads, red) specifically to microglia. FIGURE 3. View largeDownload slide Constitutive and induced CSF1R expression was restricted to microglia. (A)CSF1R ISH bright-field shows positive staining (red) in both parenchymal microglia (arrows) and perivascular macrophages (arrowheads). (B)CSF1R RNA (red) colocalizes with Iba1-positive microglia (brown). Double stained cells are denoted by arrowheads. (C) Overlaid CSF1R ISH fluorescence image of (B) confirmed localization of CSF1R RNA (arrowheads, red) specifically to microglia. As upregulated CSF1R protein expression has been associated with several neuroinflammatory conditions, we visualized CSF1R protein using immunohistochemical staining. CSF1R immunostaining was markedly upregulated in white matter compared with uninfected controls (Fig. 4). This increase was most pronounced within perivascular cuffs composed of macrophages that are characteristic of SIV encephalitis (Fig. 4, middle panel, inset). Increased CSF1R staining was associated with a change in microglial morphology. In the SIV and SIV + ART groups, microglia were larger with broad processes as compared with uninfected animals. CSF1R immunostaining was significantly increased in white matter of SIV-infected macaques compared with uninfected animals (p = 0.02, Mann–Whitney; median %ROI positive for CSF1R in SIV = 0.87 [n = 8 animals] versus 0.50 in uninfected macaques [n = 7 animals]). CSF1R immunostaining was also increased in grey matter of SIV-infected macaques compared with uninfected animals (p = 0.054, Mann–Whitney; median %ROI positive for CSF1R in SIV = 0.99 versus 0.53 in uninfected macaques). In previous studies, changes in the white matter have been the most predictive of HAND development in HIV-infected patients (28). This further suggests a role for CSF1R expression and its signaling in the pathogenesis of neurologic disease in SIV infection. FIGURE 4. View largeDownload slide Elevated CSF1R protein in both untreated and ART-treated SIV-infected pigtailed macaques. Constitutive CSF1R staining in the frontal cortex of an uninfected untreated macaque (left panel). Both perivascular macrophages and parenchymal microglia were positively immunostained for CSF1R (red). A marked increase in CSF1R immunostaining developed with SIV infection (middle panel). Inset: Perivascular cuff of macrophages with high CSF1R expression. CSF1R immunostaining remained elevated in microglia in SIV-infected macaques receiving suppressive ART (right panel; all images from frontal cortex; 400×). FIGURE 4. View largeDownload slide Elevated CSF1R protein in both untreated and ART-treated SIV-infected pigtailed macaques. Constitutive CSF1R staining in the frontal cortex of an uninfected untreated macaque (left panel). Both perivascular macrophages and parenchymal microglia were positively immunostained for CSF1R (red). A marked increase in CSF1R immunostaining developed with SIV infection (middle panel). Inset: Perivascular cuff of macrophages with high CSF1R expression. CSF1R immunostaining remained elevated in microglia in SIV-infected macaques receiving suppressive ART (right panel; all images from frontal cortex; 400×). Immunostaining results for CSF1R expression were confirmed by performing ELISA measurements on frontal cortex white matter homogenates. CSF1R protein levels were significantly increased in SIV-infected pigtailed macaques compared with uninfected macaques (p = 0.018, Mann–Whitney, Fig. 5). Additionally, there was no difference in CSF1R protein levels between the SIV and the SIV + ART groups (p > 0.999, Mann–Whitney). Collectively, these data demonstrate that CNS CSF1R expression by microglia is upregulated in SIV-infected macaques corresponding with increases in its ligand CSF1. FIGURE 5. View largeDownload slide Elevated CNS CSF1R Expression in SIV and SIV + ART. Measurements of CSF1R protein levels in frontal cortex by ELISA showed a significant increase in CSF1R concentration in the SIV-infected group (Mann–Whitney). With suppressive ART, CSF1R levels in frontal cortex remained elevated (uninfected = •; SIV encephalitis = ▲; SIV + ART =▪). FIGURE 5. View largeDownload slide Elevated CNS CSF1R Expression in SIV and SIV + ART. Measurements of CSF1R protein levels in frontal cortex by ELISA showed a significant increase in CSF1R concentration in the SIV-infected group (Mann–Whitney). With suppressive ART, CSF1R levels in frontal cortex remained elevated (uninfected = •; SIV encephalitis = ▲; SIV + ART =▪). DISCUSSION In this study based in the SIV/pigtailed macaque model of HIV CNS disease, we found that CSF1 expression is significantly increased in the CNS in both grey and white matter with SIV infection. This increase was most pronounced in SIV-infected animals that developed SIV encephalitis. CSF1 mRNA levels in the brain were strongly correlated with upregulation of immune activation and oxidative stress genes. The increase in CSF1 expression in our pigtailed macaque model of HIV CNS disease is in contrast to the work done in the rhesus macaque model of HIV CNS disease where a decrease in CSF1 expression was observed (16). Interestingly, although IL34, the other ligand that binds CSF1R, is primarily expressed in the CNS (15), IL34 mRNA expression did not change with SIV infection. Similarly, a lack of a change in IL34 expression has been shown in other models of neurodegenerative diseases including SIV infection of rhesus macaques (14, 16). Other groups have also noted that high concentrations of IL34 in media promote ramified microglial morphology in vitro, suggesting that IL34 promotes a resting state (29). Together, these findings suggest that induction of CSF1 rather than IL34 is the key driver of increased CSF1R signaling during neuroinflammatory processes including HIV-induced CNS inflammation. This is further supported by the strong correlation between CSF1 and receptor tyrosine kinase signaling in the SIV/pigtailed macaque model. Both CSF1 and IL34 have been evaluated in a SIV rhesus macaque model, but CSF1R has not been examined in detail. CSF1R is expressed by microglia, including parenchymal microglia and perivascular macrophages (6, 8, 9). Like the ligand CSF1, CSF1R protein levels were significantly increased in the CNS of SIV-infected pigtailed macaques. Surprisingly, CSF1R expression by microglia also remained elevated in ART-treated SIV-infected pigtailed macaques despite sustained suppressive therapy. The significant increase in CSF1R expression shown by both CSF1R ELISA and digital image analysis of CSF1R immunostaining may represent sustained microglial immune activation, increased trophic demand from stressed microglia, or a combination of these. Together, these findings suggest that CSF1 and CSF1R may also remain upregulated in the CNS of HIV-infected patients treated with ART and could contribute to the sustained CNS inflammation suspected to underpin HAND. Although it has been suggested that CSF1R is an M2 marker (30), the M1/M2 polarization paradigm may not be applicable to microglia (31). As CSF1 expression correlates with both M1 and M2 markers, notably CD14 and CD68 respectively, as well as genes associated with driving M1 and M2 polarization (32), it follows that CSF1R is not representative of either M1 or M2 immunophenotypes. This activation axis is a fluid spectrum, and it is likely that any given microglial cell can behave in an equally fluid manner in response to insult. It is possible that CSF1R signaling plays an anti-inflammatory role, in which case coordinate upregulation of CSF1 and its receptor, CSF1R, could represent a CNS response to mitigate the strong pro-inflammatory response to SIV. An alternate possibility is that CSF1R signaling indirectly promotes the pro-inflammatory environment by increasing the survival and proliferation of microglia. In light of studies showing damage by removing CSF1R-expressing microglia at the beginning of disease progression (33) and others showing the opposite effect later on in disease (6, 13, 14), it is possible that CSF1R signaling may be protective at the initiation of disease, but becomes damaging with persistent sustained immune stimulation. Evidence of a relationship between CSF1R over-expression, increased signaling through CSF1R, and the pathogenesis of neurodegenerative diseases such as AD and ALS in mouse models (6, 13, 14) suggests that this CSF1R immunoregulatory role extends beyond HIV to chronic CNS immune activation in general. However, considering the strong correlations between CSF1 expression and the genes associated with the antiviral response in particular, upregulation of CSF1R in SIV-infected macaques may also be linked to SIV replication within the CNS. Whether upregulation of CSF1R represents a direct immune response to virus, an enhanced demand for trophic support by damaged microglia, or a combination of these remains to be determined. Given these findings and our study results, inhibiting CSF1R signaling may prove to be an effective therapy for chronic inflammation in the CNS, such as that seen in HAND. Future studies testing the ability of CSF1 receptor blockade to deplete microglia in SIV-infected macaques receiving suppressive ART will enable us to determine whether microglia depletion alters persistent immune activation present in SIV-infected macaques receiving ART to model HIV CNS disease. Moreover, CSF1R blockade to target and selectively deplete brain macrophage populations harboring latent, replication competent HIV will be a novel approach to decrease or eradicate HIV CNS reservoirs. ACKNOWLEDGMENTS We thank Megan McCarron, Ken Witwer, Lisa Mangus, Sarah Beck, and the Retrovirus laboratory team for expert assistance. We acknowledge the helpful insights and thoughtful discussions provided by Drs. Feilim MacGabhann, H. Benjamin Larman, and Norman Haughey. We also thank the Johns Hopkins Pathobiology doctoral program for guidance and support. REFERENCES 1 Lall D, Baloh RH. Microglia and C9orf72 in neuroinflammation and ALS and frontotemporal dementia. J Clin Invest 2017; 127: 3250– 8 http://dx.doi.org/10.1172/JCI90607 Google Scholar CrossRef Search ADS PubMed 2 Ransohoff RM, El Khoury J. Microglia in health and disease. (Review). Cold Spring Harb Perspect Biol 2015; 8: a020560 Google Scholar CrossRef Search ADS PubMed 3 Saylor D, Dickens AM, Sacktor N, et al. HIV-associated neurocognitive disorder – pathogenesis and prospects for treatment. Nat Rev Neurol 2016; 12: 234– 48 http://dx.doi.org/10.1038/nrneurol.2016.27 Google Scholar CrossRef Search ADS PubMed 4 Williams DW, Veenstra M, Gaskill PJ, et al. Monocytes mediate HIV neuropathogenesis: Mechanisms that contribute to HIV associated neurocognitive disorders. Curr HIV Res 2014; 12: 85– 96 http://dx.doi.org/10.2174/1570162X12666140526114526 Google Scholar CrossRef Search ADS PubMed 5 Dheen ST, Kaur C, Ling EA. Microglial activation and its implications in the brain diseases. Curr Med Chem 2007; 14: 1189– 97 http://dx.doi.org/10.2174/092986707780597961 Google Scholar CrossRef Search ADS PubMed 6 Elmore MR, Najafi AR, Koike MA, et al. Colony-stimulating factor 1 receptor signaling is necessary for microglia viability, unmasking a microglia progenitor cell in the adult brain. Neuron 2014; 82: 380– 97 http://dx.doi.org/10.1016/j.neuron.2014.02.040 Google Scholar CrossRef Search ADS PubMed 7 Erblich B, Zhu L, Etgen AM, et al. Absence of colony stimulation factor-1 receptor results in loss of microglia, disrupted brain development and olfactory deficits. PLoS ONE 2011; 6: e26317 Google Scholar CrossRef Search ADS PubMed 8 Sherr CJ. Colony-stimulating factor-1 receptor. Blood 1990; 75: 1– 12 Google Scholar PubMed 9 Stanley ER, Chitu V. CSF-1 receptor signaling in myeloid cells. Cold Spring Harb Perspect Biol 2014; 6(6). 10 Akiyama H, Nishimura T, Kondo H, et al. Expression of the receptor for macrophage colony stimulating factor by brain microglia and its upregulation in brains of patients with Alzheimer's disease and amyotrophic lateral sclerosis. Brain Res 1994; 639: 171– 4 http://dx.doi.org/10.1016/0006-8993(94)91779-5 Google Scholar CrossRef Search ADS PubMed 11 Gomez-Nicola D, Fransen NL, Suzzi S, et al. Regulation of microglial proliferation during chronic neurodegeneration. J Neurosci 2013; 33: 2481– 93 http://dx.doi.org/10.1523/JNEUROSCI.4440-12.2013 Google Scholar CrossRef Search ADS PubMed 12 Mitrasinovic OM, Murphy GMJr. Microglial overexpression of the M-CSF receptor augments phagocytosis of opsonized Abeta. Neurobiol Aging 2003; 24: 807– 15 http://dx.doi.org/10.1016/S0197-4580(02)00237-3 Google Scholar CrossRef Search ADS PubMed 13 Dagher NN, Najafi AR, Kayala KM, et al. Colony-stimulating factor 1 receptor inhibition prevents microglial plaque association and improves cognition in 3xTg-AD mice. J Neuroinflammation 2015; 12: 139 http://dx.doi.org/10.1186/s12974-015-0366-9 Google Scholar CrossRef Search ADS PubMed 14 Martinez-Muriana A, Mancuso R, Francos-Quijorna I, et al. CSF1R blockade slows the progression of amyotrophic lateral sclerosis by reducing microgliosis and invasion of macrophages into peripheral nerves. Sci Rep 2016; 6: 25663 Google Scholar CrossRef Search ADS PubMed 15 Zelante T, Ricciardi-Castagnoli P. The yin-yang nature of CSF1R-binding cytokines. Nat Immunol 2012; 13: 717– 9 http://dx.doi.org/10.1038/ni.2375 Google Scholar CrossRef Search ADS PubMed 16 Gerngross L, Fischer T. Evidence for cFMS signaling in HIV production by brain macrophages and microglia. J Neurovirol 2015; 21: 249– 56 http://dx.doi.org/10.1007/s13365-014-0270-6 Google Scholar CrossRef Search ADS PubMed 17 De I, Nikodemova M, Steffen MD, et al. CSF1 overexpression has pleiotropic effects on microglia in vivo. Glia 2014; 62: 1955– 67 http://dx.doi.org/10.1002/glia.22717 Google Scholar CrossRef Search ADS PubMed 18 Spudich SS. Immune activation in the central nervous system throughout the course of HIV infection. Curr Opin HIV AIDS 2016; 11: 226– 33 http://dx.doi.org/10.1097/COH.0000000000000243 Google Scholar CrossRef Search ADS PubMed 19 Gerngross L, Lehmicke G, Belkadi A, et al. Role for cFMS in maintaining alternative macrophage polarization in SIV infection: Implications for HIV neuropathogenesis. J Neuroinflammation 2015; 12: 58 http://dx.doi.org/10.1186/s12974-015-0272-1 Google Scholar CrossRef Search ADS PubMed 20 Beck SE, Kelly KM, Queen SE, et al. Macaque species susceptibility to simian immunodeficiency virus: Increased incidence of SIV central nervous system disease in pigtailed macaques versus rhesus macaques. J Neurovirol 2015; 21: 148– 58 http://dx.doi.org/10.1007/s13365-015-0313-7 Google Scholar CrossRef Search ADS PubMed 21 Beck SE, Queen SE, Metcalf Pate KA, et al. An SIV/macaque model targeted to study HIV-associated neurocognitive disorders. J Neurovirol 2017 Oct 3 [Epub ahead of print]. 22 Beck SE, Queen SE, Witwer KW, et al. Paving the path to HIV neurotherapy: Predicting SIV CNS disease. Eur J Pharmacol 2015; 759: 303– 12 http://dx.doi.org/10.1016/j.ejphar.2015.03.018 Google Scholar CrossRef Search ADS PubMed 23 Meulendyke KA, Ubaida-Mohien C, Drewes JL, et al. Elevated brain monoamine oxidase activity in SIV- and HIV-associated neurological disease. J Infect Dis 2014; 210: 904– 12 http://dx.doi.org/10.1093/infdis/jiu194 Google Scholar CrossRef Search ADS PubMed 24 Gruber MF, Weih KA, Boone EJ, et al. Endogenous macrophage CSF production is associated with viral replication in HIV-1-infected human monocyte-derived macrophages. J Immunol 1995; 154: 5528– 35 Google Scholar PubMed 25 Kalter DC, Nakamura M, Turpin JA, et al. Enhanced HIV replication in macrophage colony-stimulating factor-treated monocytes. J Immunol 1991; 146: 298– 306 Google Scholar PubMed 26 Sasaki A. Microglia and brain macrophages: An update. Neuropathology 2017; 37: 452– 64. Google Scholar CrossRef Search ADS PubMed 27 Klegeris A, McGeer PL. Toxicity of human monocytic THP-1 cells and microglia toward SH-SY5Y neuroblastoma cells is reduced by inhibitors of 5-lipoxygenase and its activating protein FLAP. J Leukoc Biol 2003; 73: 369– 78 http://dx.doi.org/10.1189/jlb.1002482 Google Scholar CrossRef Search ADS PubMed 28 Cardenas VA, Meyerhoff DJ, Studholme C, et al. Evidence for ongoing brain injury in human immunodeficiency virus-positive patients treated with antiretroviral therapy. J Neurovirol 2009; 15: 324– 33 http://dx.doi.org/10.1080/13550280902973960 Google Scholar CrossRef Search ADS PubMed 29 Muffat J, Li Y, Yuan B, et al. Efficient derivation of microglia-like cells from human pluripotent stem cells. Nat Med 2016; 22: 1358– 67 http://dx.doi.org/10.1038/nm.4189 Google Scholar CrossRef Search ADS PubMed 30 Ohashi T, Aoki M, Tomita H, et al. M2-like macrophage polarization in high lactic acid-producing head and neck cancer. Cancer Sci 2017; 108: 1128– 34 http://dx.doi.org/10.1111/cas.13244 Google Scholar CrossRef Search ADS PubMed 31 Ransohoff RM. A polarizing question: Do M1 and M2 microglia exist? Nat Neurosci 2016; 19: 987– 91 Google Scholar CrossRef Search ADS PubMed 32 Martinez FO, Gordon S. The M1 and M2 paradigm of macrophage activation: Time for reassessment. F1000Prime Rep 2014; 6: 13 Google Scholar CrossRef Search ADS PubMed 33 Jin WN, Shi SX, Li Z, et al. Depletion of microglia exacerbates postischemic inflammation and brain injury. J Cereb Blood Flow Metab 2017; 37: 2224– 36 http://dx.doi.org/10.1177/0271678X17694185 Google Scholar CrossRef Search ADS PubMed © 2018 American Association of Neuropathologists, Inc. All rights reserved.
Journal of Neuropathology & Experimental Neurology – Oxford University Press
Published: Mar 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera