Independent actions by HIV-1 Tat and morphine to increase recruitment of monocyte-derived macrophages into the brain in a region-specific manner

Abstract

Despite advances in the treatment of human immunodeficiency virus (HIV), approximately one-half of people infected with HIV (PWH) experience neurocognitive impairment; opioid use disorder (OUD) can exacerbate the cognitive and pathological changes seen in PWH. HIV increases inflammation and immune cell trafficking into the brain; however, less is known about how OUD affects the recruitment of immune cells. Accordingly, we examined the temporal consequences of HIV-1 Tat and/or morphine on the recruitment of endocytic cells (predominantly perivascular macrophages and microglia) in the dorsal striatum and hippocampus by infusing multi-colored, fluorescently labeled dextrans before and after exposure. To address this question, transgenic mice that conditionally expressed HIV-1 Tat (Tat+), or their control counterparts (Tat−), received three sequential intracerebroventricular (i.c.v.) infusions of Cascade Blue-, Alexa 488-, and Alexa 594-labeled dextrans, respectively infused 1 day before, 1-day after, or 13-days after morphine and/or Tat exposure. At the end of the study, the number of cells labeled with each fluorescent dextran were counted. The data demonstrated a significantly higher influx of newly-labeled cells into the perivascular space than into the parenchyma. In the striatum, Tat or morphine exposure increased the number of endocytic cells in the perivascular space, while only morphine increased the recruitment of endocytic cells into the parenchyma. In the hippocampus, morphine, but not Tat, increased the influx of dextran-labeled cells into the perivascular space, but there were too few labeled cells within the hippocampal parenchyma to analyze. Collectively, these data suggest that HIV-1 Tat and morphine act independently to increase the recruitment of endocytic cells into the brain in a region-specific manner.

Introduction

Despite the advent of combination antiretroviral therapy (cART), up to one-half of people infected with HIV (PWH) may experience neurocognitive dysfunction [1,2]. Furthermore, opiate abuse exacerbates the neurocognitive impairment of HIV and its associated pathologic changes. HIV-1 infection is associated with blood-brain barrier (BBB) impairment, inflammation, and increased monocyte migration into the central nervous system (CNS) [3–12]. Opiate abuse exacerbates the neuropathology of HIV, in part, by enhancing glial activation, neurotoxicity, vascular leakiness, and increased macrophage entry into the CNS [13,14,23,15–22].

HIV-1 Tat, the trans-activator of transcription, is a protein secreted by HIV-infected cells. Unlike other HIV viral proteins, Tat levels within the cerebrospinal fluid (CSF) often remain elevated even in virally-suppressed PWH [24,25]. Furthermore, Tat is biologically active [24,25] and Tat levels are significantly associated with drug use and abuse [24,25]. In vitro studies have demonstrated that HIV infection [68] and exposure to Tat [26] increase the transmigration of monocytes across the BBB through increases in production and sensitivity to the chemokine C-C motif ligand 2 (CCL2). Furthermore, although depletion of CD4+ T cells is a primary cause for terminal progression into acquired immunodeficiency syndrome (AIDS), increased monocyte recruitment is a significantly better predictor of disease progression than CD4+ counts in simian immunodeficiency virus (SIV)-infected adult macaques [27]. Accumulation of perivascular macrophages during HIV infection is a hallmark of HIV encephalitis [28–30] and pathologic changes in brains of patients with HIV-associated dementia correlate with an increase in activated bone-marrow derived cells, including macrophages [31]. HIV infection increases the expression of junctional adhesion molecule-A (JAM-A) and activated leukocyte cell adhesion molecule (ALCAM) on monocytes and is at least one mechanism by which the enhanced transmigration occurs [32]. In addition, using a Tat transgenic mouse model, our lab expanded upon previous work by demonstrating that Tat exposure results in increased migration, as well as the phagocytic activity, of monocytes and macrophages within the brain [33].

Opiates, such as morphine, may also modulate monocyte infiltration into the CNS, although the directionality of these changes may depend on context. One study found that SIV-infected macaques treated with morphine have increased monocyte migration into and viral loads within the brain in comparison to those animals not exposed to morphine [34]. A more recent study demonstrated a differential impact of morphine on viral reservoir size. Morphine exposure resulted in a decreased reservoir size within the lymph nodes but an increased viral reservoir size within the CNS. This was thought, in part, to be due to increased CNS infiltration of infected monocytes [35]. In a rat model, chronic morphine exposure and/or HIV-1 Tat exposure significantly increased immune cell trafficking into the CNS in response to bacterial infection by S. pneumoniae [36]. Morphine also increased the number of infiltrating macrophages and microglia at the site of injury in a spinal cord injury model [37]. Other studies, however, have concluded that opioid exposure either has no effect or decreases chemokine-stimulated monocyte migration in in vitro chemotaxis assays [38–40]. To date, the effects of Tat and morphine co-exposure on immune cell recruitment and function (endocytic activity) have not been assessed in vivo.

The purpose of this study was to examine the effects of HIV-1 Tat and/or morphine on recruitment of new endocytic cells (predominantly macrophages and microglia) within the brain. New endocytic cells were identified by the engulfment of uniquely colored, fluorescent dextrans infused intracerebroventricularly (i.c.v.) 1- or 13-days following morphine and/or Tat exposure. We hypothesized that morphine co-exposure with Tat would increase the recruitment of monocyte-derived macrophages from the blood into the brain in a region–specific manner in Tat transgenic mice. In this paper, recruitment is defined as the appearance of new macrophages resulting from 1) the infiltration of circulating monocyte-derived macrophages into the CNS (monocyte-derived macrophages), 2) the recruitment of previously quiescent (‘M0’) microglia, or 3) newly labeled cells because of perivascular macrophage or microglial proliferation. To examine recruitment of endocytic cells, HIV-1 Tat (Tat+) or (Tat−) mice received three sequential i.c.v. infusions of fluorescently labeled-dextrans (infused 1 day prior to Tat induction, 1-day post-induction, and 13-days post-induction) with or without morphine exposure. The number of cells labeled with each fluorescent dextran was counted in the striatum and hippocampus. We, and others, have previously demonstrated that the striatum is preferentially vulnerable to Tat and morphine co-exposure [33,41,42], while the hippocampus displays less overt pathology although significant impairments to hippocampal function are demonstrable [43].

Material and Methods

The use of mice in these studies was approved by the Institutional Animal Care and Use Committee at Virginia Commonwealth University and the experiments were conducted in accordance with ethical guidelines defined by the National Institutes of Health (NIH Publication No. 85–23). Adult, female mice (approximately 70 days of age) were utilized for these initial studies given their capacity for a more dynamic neuroimmune response to a range of insults as compared to adult males [44,45]. Mice were generated in the vivarium at Virginia Commonwealth University and either expressed an HIV-1 tat transgene (Tat+) or were controls that lacked the tat transgene (Tat−). Tat+ mice conditionally expressed the HIV-1 Tat_1–86 protein in a nervous system-targeted manner via a glial fibrillary acidic protein (GFAP)-driven, Tet-on promoter, which is activated by consumption of chow containing the tetracycline-class antibiotic, doxycycline (Dox), which, unlike tetracycline, crosses the blood-brain barrier. Tat− controls expressed only the tetracycline-responsive rtTA transcription factor as described [41,46]. To induce Tat expression, mice were placed on Dox-containing chow (Dox Diet #2018; 6 g/kg) obtained from Harlan Laboratories (Madison, WI). Mice were housed 4–5/cage and were maintained in a temperature- and humidity-controlled room on a 12:12 h light/dark cycle (lights off at 18:00 h) with ad libitum access to food and water.

Tat Induction and Morphine Administration

For the macrophage recruitment studies, mice underwent osmotic minipump surgery on the same day as Tat induction was initiated. Briefly, mice were anesthetized with isoflurane (4% induction, 2% maintenance). A small mid-scapular entry was made through the skin and an ALZET^® osmotic pump (2002 model, flow rate 0.5 μL/h, 14-day delivery time) delivering placebo or 0.77 mg morphine/day was implanted. Bupivacaine was applied to all surgical sites immediately after implantation. For the chemokine analysis studies, Tat expression was induced by Dox-containing chow for 14 days, followed by a 5-day washout period. Studies, including our unpublished work, indicate continued elevation of Tat mRNA, including in the striatum and hippocampus, for at least three weeks after withholding doxycycline [47]. After the washout period, osmotic pumps (2001 model, flow rate 1 μL/h) were implanted, to deliver placebo or morphine at a dose of 1.5 mg/day. Following 5 days of morphine (or placebo) exposure, mice were euthanized, and brains were immediately harvested, striatum and hippocampus were dissected and frozen at −80 °C until chemokine analysis. All drug preparations were made in batches to minimize dosing variability.

i.c.v. Infusion of Dextrans

Mice were anesthetized and intraventricular infusions into the cerebrospinal fluid (CSF) were performed through a guide cannula (Charles River, 26 gauge) that was stereotaxically implanted using a Digital Lab Standard “Just for Mice^™” dual stereotaxic device (AP: −0.5 mm, Lat: ±1.6 mm, DV: −2 mm) as previously described [33]. To maintain patency, a dummy cannula was inserted into the guide cannula and secured with a nylon cap. To allow for adequate recovery prior to labeling of phagocytes, the first infusions did not occur until 2 days after cannulation, and Tat was not induced until 3 days after cannulation. Following surgery, mice were monitored to ensure weight gain, muscle tone, proper neurological response, and general health [48].

Use of Different Colored Dextrans to Discern Endocytic Cell Populations before and after Morphine and Tat Exposure

Fluorescently labeled dextrans (10 kDa) were administered via slow, bilateral, infusion (1.5 μL/side) aimed at the lateral ventricles using a 30-gauge Hamilton micro-syringe and polyurethane catheter tubing (each infusion performed over 60 secs). After injection, hydrophilic tracers diffuse within the CSF along perivascular spaces. Uptake by endocytic cells leads to dextran internalization and stable labeling which can be observed throughout the brain [49]. The markers are taken up within hours and remain stable within the cells for up to eight weeks [49]. To measure differences in the infiltration of cells in response to Tat and/or morphine, endocytic cells were sequentially labeled with Cascade Blue-labeled dextran (Tracer #1) 1-day prior start of Tat induction with doxycycline and/or morphine treatment (Fig. 1). Tracers #2 (Alexa Fluor^® 488-conjugated dextrans) and #3 (Alexa Fluor^® 594-conjugated dextrans), were injected into the cannula on day 1 and day 13 after initiation of Tat induction and morphine introduction (if applicable). After 14 days of Tat induction, mice were perfused with neutral pH, buffered paraformaldehyde (4%) and the tissue processed for fluorescence microscopy. Frozen coronal slices (40 μm-thickness; obtained 0.845–1.245 mm from Bregma) were counterstained with TO-PRO-3+ nuclear stain (Thermo Fisher; far red fluorescence). We have previously seen that endocytosed dextrans are visible throughout the majority of the macrophage/microglial somata (identified by ionized calcium binding adaptor molecule (Iba-1) counter-staining) making counting unambiguous [33]. Since Iba-1 does not discern macrophages from microglia [50], we relied on cellular morphology, the position of the cell body in relation to a capillary/venule (in 3D-reconstructed confocal images), and (to a lesser extent) the degree of dextran labeling as criteria for differentiating macrophages and microglia. Macrophages were identified as having a (i) round/ovoid cell bodies (ii) without cytoplasmic processes, (iii) immediately adjacent blood vessels, and (iv) tended to be more extensively labeled with dextrans than microglia. Microglia were identified by (i) multiple, (ii) ramified, cytoplasmic processes, and (iii) cell body distant from blood vessels. Juxtavascular microglia with ramified processes and at least 30% of their soma perimeter associated with blood vessels [51] were rarely observed and would have been scored as macrophages by these criteria.

Fig 1.

Experimental design

The number of blue, green and/or red fluorescent dextran-labeled cells in each treatment group were determined morphologically by counting TO-PRO-3+ cells visualized using a Zeiss LSM 700 confocal microscope (63×, 1.38 NA objective) in sequential fields until a criterion of 600 cells/mouse was met. Cells were sampled and the proportion of cells labeled with one or more colored dextrans versus unlabeled cells recorded. For striatum, Tat−/placebo, n = 8; Tat+/placebo, n = 6; Tat−/morphine, n = 6; Tat+/morphine, n = 5 (Figs. 3, 5). For hippocampus, Tat−/placebo, n = 7; Tat+/placebo, n = 3; Tat−/morphine, n = 4; Tat+/morphine, n = 5 (Fig. 4).

Fig 3.

Effects of Tat and morphine exposure in striatum on endocytic cell recruitment into the perivasculature and the parenchyma. The stable population (blue/green/red triple-label) (A, D) represents cells that were actively endocytic before exposure to Tat or morphine and remained endocytic throughout the experiment. The stable population in the perivasculature and the parenchyma was not significantly different between treatment groups. The newly recruited cells (dual-labeled with green/red) represent cells that were rapidly recruited to the perivascular space (B) or parenchyma (E) with 24 hours of exposure to Tat and/or morphine. These cells remained endocytic for the remainder of the experiment. Upon exposure, Tat and morphine independently increased the recruitment of cells within the perivascular space (B), (^#p < 0.05 Tat main effect, *p < 0.05 morphine main effect). Within the parenchyma, there was a significant increase in newly recruited cells in morphine-exposed mice, irrespective of Tat status, (^#p < 0.05). Tat and morphine also had significant main effects on the relative proportion of newly recruited cells (ratio of newly recruited cells to stable cells) within the perivascular space (C), (^#p < 0.05 Tat main effect, *p < 0.05 morphine main effect). The proportion of newly recruited cells (represented by the ratio of newly recruited to stable cells) was not significantly influenced by Tat or morphine within the parenchyma (F). The number of blue, green and/or red fluorescent dextran-labeled cells in each group were determined by counting TO-PRO-3+ cells visualized using confocal microscopy in sequential fields until a criterion of 600 cells/mouse was met. Cells from at least n = 5 mice/treatment group were sampled and the proportion of cells labeled with one or more colored dextrans versus unlabeled cells recorded. Cells from at least n = 5 mice/treatment group were sampled. Data are presented as means ± SEM.

Fig 5.

Increased infiltration of newly activated endocytic cells is influenced by Tat within the perivasculature of the striatum. Exposure to Tat significantly increased recruitment into the parenchyma (Tat+/placebo as compared to Tat−/placebo, p = 0.002) (A). The relative proportion of newly recruited cells (newly recruited to stable cell ratio) of striatal endocytic cells in the perivascular space recruitment strongly predicts recruitment within the parenchyma (B). Cells from at least n = 5 mice/treatment group were sampled.

Fig 4.

Effects of Tat and morphine exposure in hippocampus on number of endocytic cell recruitment. The stable population (triple-label, blue/green/red) was similar among the experimental groups in both the perivascular space and parenchyma. Within the perivascular space, the recruitment of newly-labeled endocytic cells (green/red dual-label) was significantly increased by morphine exposure, irrespective of Tat status (B), (^#p < 0.05). Tat status did not significantly influence recruitment. The relative proportion of newly recruited cells (the ratio of newly-labeled cells to stable cells) within the perivascular space Tat−/morphine mice was significantly increased as compared to Tat−/placebo (*p<0.05) and Tat+/placebo (*p < 0.05) (C). Tat expression significantly attenuated the increase (^#p < 0.05, Tat−/morphine vs Tat+/morphine). There were too few labeled cells across the groups within the parenchyma of the hippocampus to analyze (E). Cells from at least n = 3 mice/treatment group were sampled. Data are presented as means ± SEM.

Chemokine Expression

To measure the expression of specific chemokines in the striatum and hippocampus, samples were homogenized, and supernatants were collected. Protein concentrations were quantified using the BCA protein assay (Pierce, Rockford, IL). Sample preparation and chemokine analysis followed the manufacturer’s instructions (LEGENDplex^™ Mouse Proinflammatory Chemokine Panel, BioLegend, San Diego, CA). Briefly, sample lysates were vortexed and a volume containing 12.5 μg of protein or standard was loaded into the filter plate wells. Each well was then incubated with 12.5 μL of chemokine antibody conjugated beads for 2 h, and then incubated with biotinylated detection antibodies for 1 h on a shaker at 500 rpm. After shaking, streptavidin-PE (SA-PE) was added to the wells to amplify the signal, and the filter plate was again placed on the shaker for 30 minutes at 500 rpm. After SA-PE incubation, the filter plate was removed from the shaker, washed, sealed, and refrigerated for 24 hr. Prior to flow cytometer analysis on the Cytek^® Aurora (Cytek^® Biosciences, Fremont, CA), the samples were washed with the wash buffer, resuspended on the shaker for 5 minutes and the plate was run on the Cytek^® Aurora for analysis. Data was retrieved using the LEGENDplex^™ Data Analysis Software Suite powered by Qognit (San Diego, CA). All samples were run in duplicate. The limit of detection for CCL2/MCP-1 was 12.48 pg/mL, for CCL3/MIP-1α was 1.38 pg/mL, and 6.78 pg/mL for chemokine (C-X-C) motif ligand 1 (CXCL1)/KC.

Statistical Analysis

For the macrophage recruitment studies, data were analyzed via separate two-way ANOVAs with drug condition (placebo or morphine minipump) and genotype condition (Tat− or Tat+) as between-subjects factors. Following main effects, group differences were delineated via Fisher’s Protected Least Significant Difference post-hoc test. Interactions were delineated via simple main effects and main effect contrasts with alpha corrected for family-wise error. All analyses were considered significant when p < 0.05.

For the chemokine analysis, the LEGENDplex^™ Data Analysis Software Suite was used to create five parameter logistic standard curves and fit each well to the standard curves. Data from duplicate wells were averaged and analyzed by analysis of variance (ANOVA) followed by a Bonferroni correction for multiple comparisons to identify significant interactive and main brain region/treatment exposure effects using GraphPad Prism Version 9.4.0 (GraphPad Software, LLC.)

Results

To determine baseline numbers of resident endocytic cells, Cascade Blue-labeled dextrans were i.c.v.-injected 24 h before Tat and/or morphine exposure. To examine the extent that Tat- and/or morphine exposure recruited new cells into the brain (Cascade-Blue-negative), Alexa 488 and Alexa 594, respectively, were administered 1 day and 13 days following exposure (Figs. 1, 2). We considered Cascade Blue-labeled cells to have been residing within the CNS or perivascular space before the onset of Tat and/or morphine exposure, while cells lacking blue dextrans but possessing green/red or red only labels would be assumed to be newly recruited to the CNS. However, only triple-labeled (blue/green/red) and dual-labeled (green/red) cells (no red only or other color combinations) were observed, suggesting that when an effect was observed with Tat and/or morphine, the cell recruitment only occurred within the first 24 hours of exposure and, that once a cell was actively endocytotic, it remained active throughout the course of the experiment.

Fig 2.

Localization of labeled i.c.v.-injected dextrans into Iba-1-immunoreactive macrophages and microglia, as well as spurious localization of dextrans in non-Iba-1-positive cells, in the dorsal striatum 24 h before, and 1 and 14 days after Tat and/or morphine exposure. Cascade Blue- (blue), Alexa 488- (green), and Alexa 594 (red)- fluorescently labeled dextrans, respectively, were injected 1 day prior, 1 day after, and 13 days following sustained Tat and/or morphine exposure, colocalized with Iba-1 immunofluorescence (Alexa 647) (magenta) and visualized by confocal microscopy to assess endocytosis by resident (Cascade Blue dextran-labeled) and newly recruited (non-Cascade Blue dextran-labeled) cells. (A) Dextran-labeled perivascular cells were almost exclusively Iba-1 immunoreactive and typically display all three dextran labels (white arrowheads; A) indicating resident cells, while some Iba-1 positive macrophages/microglia were unlabeled (black arrowheads; A). (B) Ramified/quiescent Iba-1-immunoreactive microglia were sparsely labeled with Cascade Blue and Alexa 488- and 594-dextran labels (white arrowheads; B) suggesting little turnover. Sparse dextran labeling not associated with Iba-1-positive cells or processes are presumed to be associated with astrocytes or the extracellular matrix [56]. (A and B) The dotted lines (············) indicate approximate edge of the capillaries/post-capillary venules, while intermittent dashed lines (· · · · · · ·) indicate the approximate edge of a blood vessel partially outside the plane of section. White mater tracts within the striatum are indicated (*). Representative samples from ≥ n = 4 mice per group. All images are the same magnification. Scale bars = 20 μm.

Dextran-labeled perivascular cells were almost exclusively Iba-1 immunoreactive macrophages and typically displayed all three dextran labels (Fig. 2A) indicating that they were resident throughout the experiment and that they remained endocytic. Some Iba-1 positive juxtavascular macrophages/microglia were unlabeled, indicating inactivity (Fig. 2A). Ramified, Iba-1-immunoreactive microglia were sparsely labeled with Cascade Blue, and typically both the Alexa 488- and 594-dextran labels and their cell bodies were largely found in the parenchyma, not in direct association with blood vessels (Fig. 2B). Some sparse dextran label was not associated with Iba-1-immunoreactive macrophages or microglia and presumed to be internalized by astrocytes [52–55] or bound within the extracellular matrix [56–58]. Typically, the non-Iba-1 associated dextran labels only consisted of single punctum/foci, but they were consistently present in the striatum (Fig. 2B) and hippocampus (not shown).

There were no significant differences in the resident (stable; blue/green/red triple-label) population of endocytic cells in the perivasculature or parenchyma of the striatum (Fig. 3A, D). However, upon exposure, Tat and morphine independently increased the infiltration of cells within the perivascular space of the striatum. There was a significant main effect for the percentage of new endocytic cells to increase with Tat [F_(1,21) = 7.99, p < 0.05]. There was also a main effect with morphine [F_(1,21) = 53.99, p < 0.05] (Fig. 3B). Within the parenchyma of the striatum, there was a main effect for morphine to increase the percentage of new endocytic cells (green/red dual-label), irrespective of Tat status, as compared to placebo-exposed mice [F_(1,21) = 5.21, p < 0.05] (Fig. 3E). However, no significant interactions were observed with Tat and morphine co-exposure. No cells were singly labeled with red, suggesting that any morphine or Tat-induced recruitment occurred within the first day after exposure.

Within the hippocampus, the stable population (blue/green/red triple-label) was not different between treatment groups in either the perivascular space or parenchyma. Furthermore, the number of cells that engulfed the dextrans at baseline was very low in the parenchyma (Fig. 4A, D), indicating an intact BBB. The recruitment of newly labeled endocytic cells (green/red dual-label) into the perivascular space was influenced by morphine exposure but not by Tat. The morphine effect was a main effect; there was a higher percentage of fluorescently labeled cells from the morphine-exposed mice, regardless of Tat status, as compared to mice that were not exposed to morphine [F_(1,15) = 45.82, p < 0.05] (Fig. 4B). There were too few labeled cells across the groups within the parenchyma of the hippocampus to analyze (Fig 4E).

When the data were expressed as a ratio of newly labeled cells (green/red dual-label) to stable cells (blue/green/red triple-label), representing the relative proportion of newly recruited cells that may include newly endocytotically active cells, within the perivascular space of the striatum, main effects for morphine and for Tat were observed. There were significant increases in the proportion of newly recruited cells upon exposure to Tat [F_(1,42)=6.64, p < 0.05] and, irrespective of Tat status, there were also increases upon morphine exposure [F_(1,42) = 15.215, p < 0.05] (Fig. 3C). Within the parenchyma of the striatum, this ratio was not significantly influenced by Tat or morphine exposure (Fig. 3F). Compartmental effects within the striatum were observed, with significant increases in the proportion of newly recruited cells in the perivascular space as compared to the parenchyma [F_(1,42) = 6.08, p < 0.05], data not shown.

The relative proportion of newly recruited cells within the perivascular space of the hippocampus in Tat−/morphine mice was significantly increased as compared to Tat−/placebo (p < 0.05) and Tat+/placebo (p < 0.05). In comparison to the Tat−/morphine group, exposure to Tat significantly attenuated the recruitment of new cells (p < 0.05) (Fig. 4C).

The proportion of newly labeled parenchymal cells to newly labeled perivascular cells is presumed to measure recruitment of newly activated endocytic cells through a compromised BBB and into the parenchyma. In the striatum, Tat exposure alone significantly increased recruitment into the parenchyma (Tat+/placebo as compared to Tat−/placebo, p = 0.002) (Fig. 5A). Morphine also tended to increase the proportion of newly-labeled cells within the parenchyma in Tat− (p = 0.059) or Tat+ (p = 0.07) mice but large variance kept this from reaching statistical significance.

Simple regression revealed that recruitment in the perivascular and parenchymal spaces of the striatum were significantly and positively correlated, accounting for 37.5% of the variance in endocytic cell recruitment within these tissue compartments (r = 0.61, R² = 0.375, p = 0.001) (Fig. 5B).

The striatum also displayed significantly higher levels of the chemokine, CCL3, than in the hippocampus [F_(1,32) = 3.648, p < 0.05] (Fig. 6). In contrast, no significant differences were observed in chemokine levels (CCL2, CCL3, or CXCL1) upon Tat induction or morphine exposure within each brain region (Fig. 6). The other chemokines were below the limit of detection.

Fig 6.

Chemokine levels in the striatum and hippocampus in Tat and morphine treated mice. CCL3 levels within the striatum were significantly higher than in the hippocampus, regardless of Tat status or morphine exposure (*p < 0.05, brain region main effect) (A). There were no significant differences in CCL2 (B) or CXCL1 (C) observed. Data is represented as means ± SEM. Sample sizes were Tat−/placebo, n = 6; Tat+/placebo, n = 4; Tat−/morphine, n = 5; Tat+/morphine, n = 5.

Discussion

Throughout this paper, recruitment is defined as the appearance of new macrophages resulting from 1) the infiltration of circulating monocyte-derived macrophages into the CNS, 2) the recruitment of previously quiescent (‘M0’) microglia and/or 3) cells recently generated by microglial or perivascular macrophage proliferation.

Striatum: Tat and morphine displayed independent effects on endocytic cell recruitment within the striatum.

Prior studies have demonstrated an association with SIV or HIV infection and increased activation of monocytes [59], accumulation of perivascular macrophages [60], and also increased monocyte/macrophage turnover in tissues [27,61,62]. In this study, within the first 24 hours of exposure, Tat and morphine independently increased the number of newly labeled cells as well as the relative proportion of newly labeled cells [ratio of newly labeled cells (green/red) to stable cells (blue/green/red)] within the perivascular space. Within the parenchyma of the striatum, however, only morphine exposure resulted in significant increases in the number of newly recruited cells; Tat exposure trended toward increases (p = 0.0987), but this did not reach significance. The effects observed in the striatum are consistent with many other studies demonstrating opioid-specific increases in immune cell infiltration in the CNS [23,34,36,37,63–65], even though there are a few studies in which morphine has failed to stimulate immune cell migration [36,66].

Dextran internalization within the brain is mediated by microglia or CNS-associated macrophages, typically found in the boundary regions, such as the perivascular space, meninges and choroid plexus [54]. Perivascular macrophages are located within the perivascular, or Virchow-Robin, space. The perivascular space surrounds blood vessels and is a distinct compartmental interface between blood, CSF, and brain parenchyma [67,68]. Under normal homeostatic conditions, there is limited infiltration of peripherally circulating monocytes into the CNS; however, with inflammation, injury, viral protein exposure, or disease, the rate of monocyte infiltration is increased and the perivascular space is the primary site of infiltration for these new macrophages [27,28,30,60,62,69–71]. Local proliferation of macrophages and microglia can also occur during inflammation or in response to injury, albeit at slower rates [70,72,73]. From the present study, it is unclear whether the increase in labeled cells (which we have termed recruitment) is due to increased infiltration of peripheral monocytes, local microglial or perivascular macrophage proliferation, or increased activation of previously quiescent microglia. Astrocytes can also display endocytic activity under certain conditions [54,74]. However, within the perivascular space, it is presumed that the increase in endocytic cells is because of infiltration of circulating monocytes. Within the parenchyma, however, microglia are widely distributed, constituting up to 12% of CNS cells [71] and, therefore, it is more likely that the observed endocytic cells within the parenchyma were a mixture of mainly microglia with an occasional infiltrating macrophage—although cell-specific markers are needed for confirmation. So, it is quite possible that the increases observed in the perivascular space and the parenchyma are a result of a combination of factors; increased infiltration of monocyte-derived macrophages from the periphery as well as local proliferation of both microglia and perivascular macrophages.

Microglia are heterogeneous and can assume many distinctive phenotypes that change over time. At least nine transcriptionally-distinct microglial phenotypes have been identified [75,76], which include states involved in homeostasis, interferon response, antigen presentation and proliferation [76]. Additionally, microglia can change phenotypically depending on age, brain region, brain activity, alterations in their extracellular milieu and in association with specific neurodegenerative diseases [75–78]. In our study, the increased population of endocytic cells observed in the striatal parenchyma within the first 24 h following Tat induction and/or morphine exposure may suggest increased activation, proliferation, and/or recruitment of microglia. However, despite the ability to optically section the tissue using confocal microscopy, at least a portion of the cells presumed to be in the parenchyma could have been associated with adjacent blood vessels. Accordingly, a small number of the endocytic cells identified as parenchymal may have been perivascular macrophages. The current experiment does not allow for full discrimination of the cell types.

No interactive effects between Tat and morphine were observed within the striatum in these studies. Others, however, have observed opioid-HIV or opioid-HIV viral protein interactive effects within the CNS. Morphine exposure in SIV-infected rhesus macaques increased CNS infiltration of monocytes, and morphine exposure also was associated with increased SIV viral loads in the CNS and with the development of encephalitis [34]. Furthermore, a recent follow-up study found that morphine increased the size of the SIV reservoir within the CNS but decreased the reservoir size within lymph nodes, suggesting that morphine may differentially modulate SIV/HIV reservoirs throughout the body [35]. In vitro studies have demonstrated that morphine enhances Tat effects on microglial activation and enhances Tat-induced chemokine and inflammatory cytokine expression [79]. Additionally, in a S. pneumoniae model of systemic infection, chronic morphine exposure and/or HIV-1 Tat exposure induced the migration of immune cells from the periphery into the CNS [36]. Differences observed here could reflect inherent differences in the experimental models and differences in the relative doses and duration of morphine used in each study.

Hippocampus: Morphine, but not Tat, induced recruitment of endocytic cells within the perivascular space.

The morphine effects on the recruitment of newly labeled endocytic cells occurred irrespective of Tat status. Within the hippocampal parenchyma, however, observations of newly labeled cells were rare regardless of treatment. The paucity of newly labeled cells may suggest 1) a lack of BBB damage within the hippocampus, which would minimize the translocation of endocytic cells from the perivasculature into the parenchyma, intrinsic differences in 2) cell types or 3) bioavailability of i.c.v. infused dextrans in the perivascular and the parenchymal compartments, or 4) differences in the timing of activation between cells in the perivasculature and parenchyma of the hippocampus. CSF in the ventricular compartment is confluent with the extracellular space (ECS) of the parenchyma and with the perivascular space, although the CSF in the perivascular space will also be diluted interstitial fluid (exudate) from blood vessels [80] and the physical and physiological constraints to dextran availability are likely to differ markedly [81].

A positive correlation between macrophage turnover rate and brain histopathology has previously been reported; the higher the turnover rate, the greater the damage within the brain [61,62,82,83]. When examining the recruitment of endocytic cells within the hippocampus, there was an observed Tat and morphine interaction within the perivasculature. The relative proportion of newly recruited cells was increased by morphine exposure; the Tat−/morphine group had increased recruitment as compared to either placebo group (Tat−/placebo and Tat+/placebo). Interestingly, Tat attenuated the morphine effects on recruitment (Tat+/morphine group had significantly lower index of recruitment of newly-labeled cells than the Tat−/morphine group). Within the parenchyma, observations of new infiltrating cells were rare and therefore statistical analysis could not be performed for recruitment to the parenchyma.

Differential responses in the striatum versus hippocampus.

Tat exposure increased recruitment to the perivascular space in the striatum but not in the hippocampus. This might reflect inherent differences between the brain regions. Astrocytes, for example, demonstrate brain-region specific differences in cytokine/chemokine secretion under basal conditions as well as following Tat induction [84]. HIV and/or opioid exposure can increase chemokines in vulnerable CNS areas establishing gradients that can promote leukocyte adhesion to endothelial cells and subsequent migration into the perivascular space or possibly the brain parenchyma [85,86].

Although our data demonstrated a regional effect, we did not find significant changes in chemokine levels upon Tat or morphine exposure. This contrasts with other studies reporting that Tat exposure can increases chemokine levels. In intact rodent models and in astrocytes isolated from rodents, HIV-1 Tat increases CCL5, a chemokine ligand of CCR5, [36,87–90]. Other chemokines, such as CCL4 [87,88], CCL3 [87], CXCL1 [87,88] and CCL11 [87,88] are also elevated upon Tat exposure. These discrepancies could result from differences in the dose or duration of exposure in study design, as well as the model system used. Chemokine increases tend to be transient, triggering cascades of alternate inflammatory mediators and can resolve depending on the nature and duration of the insult [91,92]. We previously found that allostatic mechanisms appear to cause chemokines to return to steady-state levels following sustained Tat exposure [93]. For this reason, we speculate that transient increases in chemokines preceded and triggered the infiltration of inflammatory cells into the brain but are no longer evident after constant exposure to Tat for three weeks and to morphine for 5 days in the present study.

The impact of morphine exposure on chemokine secretion appears to be more complex. Morphine alone can have minimal effect on CCL5 levels [36,87]. However, withdrawal from morphine [87] or morphine exposure in the context of systemic bacterial infection [36] significantly increases in CCL5. The data herein is consistent with prior reports; we also did not observe changes in chemokines after 5 days of continuous morphine exposure.

This study did demonstrate significantly higher levels of the chemokine CCL3 within the striatum compared to the hippocampus. Other studies have also demonstrated regionality of chemokine response, with increases in CCL5 in the striatum but decreased CCL5 in the thalamus in response to Tat exposure [88].Therefore, in the present study, region-specific differences in chemokine production may differentially promote monocyte migration. Alternatively, the data may represent known regional differences in Tat expression observed in other studies using this model [94]. Using a different Tat-transgenic mouse model, others have observed regional differences in astrocyte activation, neuronal integrity, and epigenetic changes in response to Tat induction [95]. There could also be time-dependent differences in Tat induction in response to doxycycline between the striatum and hippocampus which may explain our observed data.

Newly infiltrating endocytic cells from perivasculature to parenchyma.

The ratio of newly labeled cells (green/red dual-label) in the parenchyma versus the perivasculature is assumed to measure infiltration of new cells into the parenchyma through a compromised BBB. In these studies, Tat exposure significantly increased the infiltration of newly active cells into the parenchyma of the striatum, suggesting that Tat exposure resulted in BBB damage, allowing for increased infiltration. Previously, we demonstrated that Tat exposure damages the BBB and can increase the paracellular flux of labeled tracer molecules into the brain [33]. We also demonstrated that Tat exposure increases the number of endocytic cells within both the perivascular space and the parenchyma [33]. Morphine exposure also causes BBB damage, disrupts the tight junction protein, zonula occludens-1 (ZO-1), with subsequent increases tracer leakage within the brain [22]. Other studies also have demonstrated decreases in other junctional proteins, occludin and claudin-, and increases in junctional adhesion molecule-2 (JAM-2) and the efflux transport protein, P-glycoprotein. These changes were associated with increases in inflammatory cytokine secretion and increased BBB permeability [85,96]. In the present studies, morphine exposure tended to increase the ratio of (newly labeled) parenchymal macrophages to perivascular macrophages, however, this did not reach statistical significance. In the hippocampus, newly labeled cells were not observed within the parenchyma under either experimental condition, suggesting that BBB permeability is lower within the hippocampus than the striatum.

The specific route by which dextrans are internalized within the cell can vary depending on size [97], surface receptors [98], cell types [98] and charge [99]. For example, in one study, the 70 kDa dextrans are internalized more predominantly by macropinocytosis whereas the 10 kDa dextrans were internalized by more general fluid endocytosis, which includes both micropinocytosis and macropinocytosis [97]. Investigation of the different routes/mechanisms of dextran internalization was beyond the scope of this study and, therefore, the more general terms of endocytosis or internalization were used throughout.

Conclusions.

These data support the hypothesis that HIV-1 Tat and morphine can influence endocytic cell recruitment and accumulation within the brain, although their effects were shown to be largely independent of each other and were also brain region- and brain compartment-specific. Future studies focusing on these regional differences to gain a better understanding of the relative role of monocyte infiltration versus local proliferation in endocytic cell recruitment are warranted.

Highlights.

• Tat and morphine increase macrophage recruitment within perivasculature of striatum

• Within the striatum, Tat and morphine effects are independent effects

• Morphine, but not Tat, increases macrophage recruitment in hippocampus

• Recruitment in the perivascular space strongly predicts recruitment into parenchyma